Tree and pasture productivity in Pseudotsuga menziesii (Mirb.) Franco silvopastoral systems fertilized with sewage sludge

Rosa Mosquera-Losada

In Galician silvopastoral systems (northwest Spain) fertilization with sewage sludge could enhance tree and pasture productivity which is limited by soil acidity. The effect of sewage sludge on tree growth and the pasture production is different depending on the process used to stabilize the sewage sludge. The aim of the present study was to evaluate the effect of fertilization with municipal sewage sludge, which has been stabilized using anaerobic digestion, composting, and pelletization, on tree and pasture productivity compared to control treatments (mineral and no fertilization) in a silvopastoral system under Pseudotsuga menziesii (Mirb.) Franco. Mineral fertilization increased the annual pasture production and reduced the tree heights due to the competition by the nutrients generated between pasture and trees. However, tree height was increased by the application of pelletized sewage sludge applied in split doses.

EGF at 50: The Future of European Grasslands Edited by A. Hopkins R.P. Collins M.D. Fraser V.R. King D.C. Lloyd J.M. Moorby P.R.H. Robson VOLUME 19 GRASSLAND SCIENCE IN EUROPE EGF at 50: The Future of European Grasslands Proceedings of the 25th General Meeting of the European Grassland Federation Aberystwyth, Wales 7-11 September 2014 Edited by A. Hopkins R.P. Collins M.D. Fraser V.R. King D.C. Lloyd J.M. Moorby P.R.H. Robson Published by Organising Committee of the 25th General Meeting of the European Grassland Federation IBERS, Aberystwyth University, Gogerddan, SY23 3EE, UK Copyright © 2014 IBERS All rights reserved. Nothing from this publication may be reproduced, stored in computerised systems or published in any form or any manner, including electronic, mechanical, reprographic or photographic, without prior written permission from the publisher. The individual contributions in this publication and any liabilities arising from them remain the responsibility of the authors. ISBN 978-0-9926940-1-2 Printed by Gomer Press Ltd. Parc Menter Llandysul Llandysul Ceredigion SA44 4JL Wales Distributed by European Grassland Federation EGF W. Kessler – Federation Secretary c/o Agroscope Reckenholzstrasse 191 CH-8046 Zürich, Switzerland Email: fedsecretary@europeangrassland.org Organising Committee President Athole Marshall AU - IBERS General Secretary Vicky King AU - IBERS Members Julie Morgan Aberystwyth University Florian Mikal Mi Mikal Aberystwyth University Sheena Duller AU – IBERS Michael Lee University of Bristol Alan Lovatt AU - IBERS Matthew Lowe AU - IBERS Christina Marley AU - IBERS Heather McCalman AU - IBERS Huw Powell AU - IBERS James Vale AU - IBERS Chair Rosemary Collins AU - IBERS Members Mariecia Fraser AU - IBERS David Lloyd AU - IBERS Jon Moorby AU - IBERS Paul Robson AU - IBERS Scientific Committee External Reviewer Alan Hopkins GES Consulting, Exeter, UK Supporters and Sponsors Aberystwyth University British Grassland Society Stapledon Memorial Trust Germinal Holdings Limited Haldrup Waitrose Waitrose Farming Partnership Yara Limited Foreword The title of this, the 25th General Meeting of the European Grassland Federation, is ‘EGF at 50: the future of European Grasslands'. Fifty years is an important milestone in the life of any organisation, so this theme encompasses an element of ‘taking stock’ as well as looking to the future. The EGF was set up in 1964 as a forum for research workers, advisors, teachers, farmers and policy makers with active interest in all aspects of grasslands in Europe. Its objectives are to facilitate and maintain close contact between grassland organizations in Europe, to promote the interchange of scientific and practical experience between grassland experts and to initiate conferences and other meetings on all aspects of grassland production and utilization in Europe. We hope that this conference will make a useful contribution to furthering the aims of the EGF. We have great pleasure in extending a warm welcome to all delegates to Aberystwyth, the location of this 25th General Meeting. Aberystwyth has been strongly associated with grassland agriculture, and in particular the breeding of forage and cereal crops, since the foundation of the Welsh Plant Breeding Station (WPBS) in 1919. The influential grassland scientist and environmentalist Sir George Stapledon was its first Director, a post he held from 1919 to 1942. Stapledon argued that grasslands were at the heart of successful agriculture, which in turn was at the heart of the UK's economic and spiritual well-being. For many years his vision was allied with the requirement to increase production from all types of grassland. One of the ways in which he had a direct impact on grassland production was through the development in WPBS of the ‘S’ varieties of grasses, clovers and oats. The highly successful S23 perennial ryegrass was launched in 1933 and S184 white clover was first marketed in 1942. Stapledon was a strong advocate of the use of grass-legume mixtures as leys, realising the importance of nitrogen transfer from legume to grass to cereal. Varieties such as the ‘S’ types enabled grassland to support higher levels of stocking, and therefore production per unit area. The political/social environment in which agriculture operates in the UK and other European countries has changed dramatically since Stapledon’s time, and the very necessary focus on increasing productivity that prevailed during the early and middle twentieth century has now widened to incorporate environmental concerns. Overproduction has been checked by a number of political and economic drivers, and the emphasis is now firmly on sustainability. It hardly needs to be stated nowadays that grassland fulfils a truly multifunctional purpose, supplying forage for animals, regulating water flows, storing carbon, preventing soil erosion, providing habitats for species in all trophic levels, and playing an important cultural role in society. But it could be argued that public awareness of the real value of grassland is still evolving, and there is much room for greater engagement of grassland science and scientists with wider society. The breadth of topics covered by the submissions to this conference shows that grassland scientists across Europe and beyond are actively engaging with grassland as a multifunctional entity, and research on all the issues mentioned above is represented by high quality posters and theatre papers. The opening session of the conference comprises three plenary papers providing an overview of European grassland research in Nordic, Temperate and Mediterranean regions. This is followed by plenary and submitted papers grouped into five Themes, plus one ‘extra’ Theme on forage crop improvement which was added to accommodate many high quality contributions on this topic. We hope that EGF 2014 will be characterised by fruitful discussions and enjoyable social interactions between grassland scientists, farmers, advisors and sponsors. On behalf of the Organising Committee we express our sincere gratitude to the many people who have contributed to the success of this conference – the Executive Committee and Secretary of the EGF, the external reviewer, Alan Hopkins, Aberystwyth University Conference Office, the IBERS staff responsible for the many different organisational aspects, the farmers who hosted mid-conference tours, the British Grassland Society who organised the post-conference tour, Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands vii technicians and staff of the Arts Centre, entertainers at the various social events, official speakers at the opening ceremony and conference dinner, poster competition judges, and of course the conference session chairs, speakers (plenary and submitted) and poster presenters. We also thank our sponsors for their generous support. Financial input from the Stapledon Memorial Trust met the full cost of printing the conference Proceedings and sponsored the early-career scientist master classes, and the British Grassland Society sponsored the Opening Session and social event. We gratefully acknowledge this support. Finally, we thank you, the delegates, for attending this conference. We hope you will enjoy your time here and return to your countries with warm memories of Aberystwyth. Athole Marshall President European Grassland Federation Vicky King Secretary Organising Committee Rosemary Collins Chair Scientific Committee Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands viii Table of Contents Foreword Special paper ‘The European Grassland Federation at 50: past, present and future’ The European Grassland Federation at 50: past, present and future Prins W.H. and Kessler W. ................................................................................................. 3 European grasslands overviews invited papers European grasslands overview: Nordic region Helgadóttir Á., Frankow-Lindberg B.E., Seppänen M.M., Søegaard K. and Østrem L... 15 European grasslands overview: temperate region Huyghe C., De Vliegher A. and Goliński P. ..................................................................... 29 European grasslands overview: Mediterranean region Cosentino S.L., Porqueddu C., Copani V., Patanè C., Testa G., Scordia D and Melis R. .......................................................................................................................................... 41 Theme 1 ‘Climate change: mitigation and adaptation’ Theme 1 invited papers Synergies between mitigation and adaptation to climate change in grassland-based farming systems Del Prado A., Van den Pol-van Dasselaar A., Chadwick D., Misselbrook T., Sandars D., Audsley E. and Mosquera-Losada M.R. ........................................................................... 61 The role of grassland in mitigating climate change Soussana J.-F., Klumpp K. and Ehrhardt F. .................................................................... 75 Theme 1 submitted papers Reducing greenhouse gas emissions in silage production with oxygen barrier film Wheelton P., Wilkinson J.M., Van Schooten H., Jan Ten Hagen P. and Wigley S........... 91 Effects of fertilization and soil compaction on nitrous oxide (N2O) emissions in grassland Sturite I., Rivedal S. and Dörsch P. .................................................................................. 94 Modelling livestock and grassland systems under climate change Kipling R.P., Saetnan E., Scollan N.D., Bartley D., Bellocchi G., Hutchings N.J., Dalgaard T. and Van den Pol-van Dasselaar A. .............................................................................. 97 Multiple regression analysis of the relationship between bioclimatic variables and grazing season length on European dairy, beef and sheep farms Phelan P., Morgan E.R., Rose H. and O’Kiely P. .......................................................... 100 Performance of legumes for potential use in pasture swards under conditions of periodic water limitation Breitsameter L., Küchenmeister K., Küchenmeister F., Wrage-Mönnig N. and Isselstein J. ........................................................................................................................................ 103 Drought effects on herbage production of permanent grasslands in northern Germany Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands ix Hoffstätter-Müncheberg M., Merten M., Isselstein J., Kayser M. and Wrage-Mönnig N. ........................................................................................................................................ 106 The effect of drought on the depth of water uptake of deep- and shallow-rooting grassland species Hoekstra N.J., Finn J.A., Hofer D., Suter M. and Lüscher A. ........................................ 109 CLIMAGIE: A French INRA project to adapt grasslands to climate change Durand J.-L., Ahmed L., Andrieu B., Barre P., Combes D., Cruz P., Decau M.L., Enjalbert J., Escobar-Gutiérrez A., Fort F., Frak E., Ghesquière M., Gastal F., Goldringer I., Hazard L., Jouany C. Julier Koubaiti B., Litrico I., Louarn G., Meuriot F., Morvan Bertrand A., Picon Cochard C., Pottier J., Prudhomme M.P., Sampoux J.P., Volaire F., Zaka S. and Zwicke M..................................................................................................... 112 Comparison of temperature responses of different developmental processes in Medicago sativa L. and Festuca arundinacea Schreb. Zaka S., Ahmed L.Q., Escobar-Gutiérrez A.J., Durand J.-L. and Louarn G. ................ 115 Qualitative overview of mitigation and adaptation options in livestock systems Van den Pol-van Dasselaar A. and Bannink A. .............................................................. 119 Genetic diversity of Lolium perenne L. in the response to temperature during germination Ahmed L.Q., Durand J.-L., Louarn G., Fourtie S., Sampoux J.-P. and Escobar-Gutiérrez A. J. ................................................................................................................................. 122 Time of ploughing affects nitrous oxide emissions following renovation and conversion of permanent grassland Biegemann T., Loges R., Poyda A. and Taube F. ........................................................... 125 Comparing nitrous oxide emissions from white clover-ryegrass pasture with swards receiving applied synthetic fertilizer Hyland J.J., Jones D.L., Chadwick D. and Williams A.P. .............................................. 128 Theme 1 posters Impact of climate change on grassland productivity and forage quality in Austria Poetsch E.M. , Asel A. , Schaumberger A. and Resch R. ................................................. 139 Effect of climatic changes on grassland growth, water condition and biomass – the FINEGRASS project Dąbrowska-Zielińska K., Goliński P., Jørgensen M., Mølmann J. and Taff G. ............. 142 Generating carbon credits from perennial forage species crops in the Mediterranean region: the case of Phalaris aquatica L. Pappas I.A., Papaspyropoulos K.G., Karachristos C.N. and Christodoulou A.S. ......... 145 Agroforestry systems: an option for mitigation and adaptation to overcome global climate change Mosquera-Losada M.R. and Rigueiro-Rodríguez A. ...................................................... 148 Drought tolerance of the Lolium multiflorum-Festuca arundinacea introgression forms Perlikowski D., Pawłowicz I., Zwierzykowski Z., Zwierzykowski W., Paszkowski E. and Kosmala A....................................................................................................................... 151 Effect of water stress on Lotus corniculatus L. nutritive value at different stages of maturity Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands x Georganta A., Parissi Z.M., Kyriazopoulos A.P., Abraham E.M. and Lazaridou M. .... 154 Impact of limited irrigation on water economy and photosynthetic performance of Lotus corniculatus Kostopoulou P., Karatassiou M., Lazaridou A., Lazaridou M. and Patakas A. ............ 157 Drought resistance of selected forage legumes for smallholder farmers in East Africa Wrage-Mönnig N., Mutimura M., Kigongo J., Paul B.K., Isselstein J. and Maass B.L. 160 Water use efficiency of tall fescue (Festuca arundinacea Schreb.) and perennial ryegrass (Lolium perenne L.) under different management intensity Pardeller M., Schäufele R., Pramsohler M. and Peratoner G. ...................................... 163 Important differences in yield responses to simulated drought among four species and across three sites Hofer D., Suter M., Hoekstra N.J., Haughey E., Eickhoff B., Finn J.A., Buchmann N. and Lüscher A. ....................................................................................................................... 166 Assessment of the nutritive value and methanogenic potential of two cultivars of Lotus corniculatus L. and Lotus uliginosus Schkuhr. Marichal M. de J., Piaggio L., Crespi R., Arias G., Furtado S., Cuitiño M.J. and Rebuffo M. .................................................................................................................................... 169 Improvement of the digestibility of tall fescue (Festuca arundinacea Schreb.) inspired by perennial ryegrass (Lolium perenne L.) Baert J. and Van Waes C. ............................................................................................... 172 Dry matter yield and digestibility of five cool season forage grass species under contrasting N fertilizations Cougnon M., Baert J. and Reheul D............................................................................... 175 Grazing season length on dairy, beef and sheep farms in Europe Phelan P., Morgan E., Rose H. and O’Kiely P. ............................................................. 178 Effects of mild heat stress periods on milk production, milking frequency and rumination time of grazing dairy cows milked by a mobile automatic system Lessire F., Hornick J.L. and Dufrasne I. ........................................................................ 181 Interactive effects of Epichloë endophytes and plant origin on mineral content in Festuca rubra Vázquez de Aldana B.R., Helander M., Zabalgogeazcoa I., García-Ciudad A.,............ 184 Soil carbon status of survived and restoring wood pasture in the protected area Natura 2000 Slepetiene A., Slepetys J., Liaudanskiene I., Jokubauskaite I., Stukonis V. .................... 187 Plant succession and soil carbon sequestration potential of abandoned arable fields in a subhumid Mediterranean environment Karakosta C., Pappas I.A. and Papanastasis V.P. ......................................................... 191 Theme 2 ‘Grasslands and ecosystem services’ Theme 2 invited papers Functions of grassland and their potential in delivering ecosystem services Isselstein J. and Kayser M. ............................................................................................. 199 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands xi Comparing synthetic and natural grasslands for agricultural production and ecosystem service Humphreys M.W., O’Donovan G. and Sheehy-Skeffington M. ...................................... 215 Theme 2 submitted papers Single species and mixed grazing regimes to restore Nardus stricta moorland Critchley C.N.R., Griffiths J.B. and Clarke A. ............................................................... 233 Integrating biodiversity conservation with grassland farming: extensive cattle grazing and farmland birds Buckingham D.L., Brook A.J., Eschen R., Maczey N., Wheeler K. and Peach W.J. ...... 236 Spatial soil variation on the North Wyke Farm Platform Shepherd A., Harris P., Griffith B., Noacco V., Ramezani K., Tuominen E. and Eludoyin A. ..................................................................................................................................... 239 “Virtual grassland”: an individual-based model to deal with grassland community dynamics under fluctuating water and nitrogen availability Louarn G., Escobar-Gutiérrez A., Migault V., Faverjon L. and Combes D. ................. 242 Towards a new potassium fertilization recommendation in the Netherlands Holshof G. and van Middelkoop J.C. ............................................................................. 245 Increasing perennial ryegrass (Lolium perenne) yields using commercial bio-inocula on a phosphate-limited soil Owen D., Williams A.P., Griffith G.W. and Withers P.J.A............................................. 248 Long-term effects of extensification regimes on soil and botanical characteristics of improved upland grasslands Pavlů V., Pavlů L. and Fraser M.D. ............................................................................... 251 How do pre-sowing disturbance and post-establishment management affect restoration progress in ex-arable calcareous grassland? Wagner M., Bullock J.M., Meek W.R., Walker K.J., Stevens C.J., Heard M.S. and Pywell R.F. ................................................................................................................................. 254 Sown biodiverse pastures as a win-win approach to reverse the degradation of Mediterranean ecosystems Teixeira R.F.M., Proença V., Valada T., Crespo D. and Domingos T. .......................... 258 Fauna-flora relationships within improved upland grasslands managed under alternative extensification regimes Rosa García R. and Fraser M.D. ................................................................................... 261 The effects of agricultural forages on soil biology – linking the plant-soil-invertebrate ecosystem Crotty F.V., Fychan R., Scullion J., Sanderson R. and Marley C.L. .............................. 267 Managing grasslands to mitigate flooding risk Newell Price J.P., Balshaw H. and Chambers B.J. ........................................................ 270 Developing an in situ sensor for real time monitoring of soil nitrate concentration Shaw R., Williams A.P., Miller A. and Jones D.L. ......................................................... 273 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands xii Theme 2 posters PastureBase Ireland – the measurement of grass dry matter production on grassland farms Griffith V., O’Donovan M., Geoghegan A. and Shalloo L. ............................................ 279 Estimation of grassland production with a new land classification system in Hungary Hoffmann R., Keszthelyi S. and Pál-Fám F. ................................................................... 282 Influence of nitrogen fertilizers on yield and digestibility of grass Adamovics A. and Platace R........................................................................................... 285 Forage yield and protein content of five native species from Lanzarote (Canary Islands) Chinea E., Batista C., García-Ciudad A. and García-Criado B. ................................... 288 Effect of manure enriched with clinoptilolite on pasture yield and quality Simić A., Rakić, V., Marković, J., Dželetović Ž., and Živanović I. ................................. 291 Influence of long-term organic and mineral fertilization on Festuca rubra L. grassland Păcurar F., Rotar I., Balazsi A., and Vidican R. ............................................................ 294 Effects of low-input treatments on Agrostis capillaris L. - Festuca rubra L. grasslands Rotar I., Păcurar F., Balázsi Á., Vidican R., Mălinaş A. ............................................... 298 Influence of fertilization on the biodiversity of Festuca rubra L. and Agrostis capillaris L. grassland Samuil C., Vintu V., Popovici C.I. and Stavarache M. ................................................... 302 The effect of organic fertilization on Agrostis capillaris L. and Festuca rubra L. grasslands from the Romanian Eastern Carpathians Vintu V., Chidovet S., Samuil C. and Stavarache M....................................................... 306 Herbage Recommended List applicability to low inorganic nitrogen (N) production systems Matthews J.M., Genever E., McConnell D. and Kerr S. ................................................. 309 Effect of mineral fertilization on yield and quality of grassland ecosystem Agrostietum vulgaris Vuckovic S., Simic A., Jovanovic M., Cupina B.. and Krstic D. ..................................... 312 Impact of surface fertilization on dehydrogenase activity in grassland soil Tampere M., Kauer K., Keres I., Laidna T., Loit E., Parol A., Selge A., Viiralt R. and Raave H. .................................................................................................................................... 315 Reduction of soft rush (Juncus effusus L.) by a combination of trimming and grazing Nielsen A.L., Hald A.B. and Nissen T. ............................................................................ 318 Prospects for biological control of Rumex obtusifolius using a native clearwing moth Hahn M.A., Häfliger P., Schaffner U. and Lüscher A. ................................................... 321 Effect of different grazing regimes on the coverage of Taraxacum spp. under a long-term grazing experiment Supek Š., Pavlů V. , Ludvíková V., Pavlů L., Gaisler J. and Hejcman M. ...................... 324 Emergence and survival of Rumex OK-2 (Rumex patientia x Rumex tianschanicus) in grasslands under different management conditions Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands xiii Hujerová R. , Gaisler J. , Pavlů L. , Pavlů V. and Hejcman M. ..................................... 327 Mixed cropping of grass and alfalfa to reduce weed growth Surault F., Julier B. and Huyghe C. ............................................................................... 330 Impact of site conditions on natural and fodder value of meadow-pasture communities with different contributions of Urtica dioica L. Strychalska A., Kryszak A., Kryszak J. and Jankowski K. .............................................. 333 Tree and pasture productivity in Pseudotsuga menziesii (Mirb.) Franco silvopastoral systems fertilized with sewage sludge Ferreiro-Domínguez N., Rigueiro-Rodríguez A. and Mosquera-Losada M.R. .............. 336 Improved light availability of legumes in moderately N-fertilized mixed swards Frankow-Lindberg B.E. and Wrage-Moennig N. ........................................................... 339 Nitrogen application strategies to mixed grass-legume leys Frankow-Lindberg B.E. and af Geijersstam L. .............................................................. 342 Potential of short-term nitrogen transfer between Trifolium repens and the grasses Festuca gr. rubra and Brachypodium pinnatum in highland grasslands Canals R.M., San Emeterio L., Gutiérrez R. and Juaristi A. .......................................... 347 Root architecture of interspecific hybrids between Trifolium repens L. and Trifolium ambiguum M. Bieb. and their potential to deliver ecosystem services Marshall A.H., Lowe M. and Sizer-Coverdale E. ........................................................... 350 Interactive N supply and cutting intensity effect on canopy height at 95% light interception Pontes L. da S., Baldissera T.C., Barro R.S., Giostri A.F., Stafin G., Santos B.R.C., Porfírio-da-Silva V. and Carvalho P.C. de F. ................................................................ 353 Interactive N supply and cutting intensity effect on leaf nutritive value of C4 grasses Pontes L. da S., Giostri A., Barro R.S., Baldissera T.C., Carpinelli S., Guera K.C.S. and Carvalho P.C. de F. ........................................................................................................ 356 Forage selection and animal performance of grazing heifers on semi-natural fen grassland Müller J. and Sweers W. ................................................................................................. 359 Breed type differences in hoof volume in beef suckler cows Fraser M.D. and Vale J.E............................................................................................... 362 Changes in Koniks' diet due to vegetative season, years and social behaviour Chodkiewicz A. and Stypiński P. .................................................................................... 364 Long-term stability of sward patch structure under different intensities of cattle grazing Tonn B., Wrage-Mönnig N. and Isselstein J. .................................................................. 367 Impact of long-term extensive use of permanent grasslands on their provisioning service Kizeková M., Kanianska R., Makovníková J., Beňová D., Čunderlík J. and Jančová Ľ.370 Herbage yield and quality of a limestone grassland managed differently for 30 years Seither M......................................................................................................................... 373 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands xiv BIOECOSYS: towards the development of a decision support tool to evaluate grassland ecosystem services Campion M., Ninane M., Hautier L., Dufrêne M. and Stilmant D. ................................ 376 Grassland biodiversity: how we might meet international commitments Peel S. ............................................................................................................................. 379 Resilience of Mediterranean ecosystems: tree and management effects on variability of herbaceous pastures in a dry year López-Sánchez A., San Miguel A. and Roig S. ............................................................... 382 Soil organic carbon and nitrogen stocks affected by grazing intensity in temperate permanent grassland Nüsse A.M., Linsler D., Kaiser M., Isselstein J. and Ludwig B...................................... 385 Effects of biomass of perennial grasses and legumes on soil carbon Skuodiene R. and Tomchuk D. ........................................................................................ 388 Soil organic carbon characteristics under different intensities of grassland management Karabcová H., Mičová P. and Fiala K. .......................................................................... 391 Efficacy of the agrosteppe method for restoring eroded lands Dzybov D.S. and Starodubtseva A.M. ............................................................................. 394 Zonal strategy for sward renovation by total reseeding based on research results Ene T.A., Mocanu V., Mocanu V., Ciopata A.C. and Cardasol V. ................................. 397 Magnesium content in soil and selected layers of upland grassland biomass Grygierzec B., Kasperczyk M. and Szewczyk W. ............................................................ 400 Effects of previous cropping and establishment method on mineral concentration of wholeplant spring wheat Fychan R., Scott M.B., Davies J.W., Crotty F.V., Sanderson R. and Marley C.L. ......... 404 Effect of soil amendment in the cultivation of selected grass species Sosnowski J., Jankowski K., Kolczarek R. and Wiśniewska-Kadźajan B. ...................... 407 Milk production and profitability in relation to size of grassland farms Jankowski K., Sosnowski J., Wiśniewska-Kadżajan B. and Kolczarek R. ...................... 411 Meadow apophytes in segetal communities Skrzyczyńska J., Ługowska M., Skrajna T., Rzymowska Z., Jankowska J. and Sosnowski J. ........................................................................................................................................ 414 Effectiveness of grassland management and mechanical methods for the weed control of Colchicum autumnale in permanent meadows Peratoner G., Figl U., Florian C., Klotz C. and Gottardi S. .......................................... 418 Issues regarding the genus Fusarium in permanent grassland Nedělník J., Palicová J., Hortová B. and Strejčková M. ................................................ 421 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands xv Theme 3 ‘Novel uses of grassland, including bioenergy and biorefining’ Theme 3 invited papers Novel products from grassland (bioenergy & biorefinery) Thumm U., Raufer B. and Lewandowski I. ..................................................................... 429 Grasslands for forage and bioenergy use: traits and biotechnological implications Barth S., Jones M., Hodkinson T., Finnan J., Klaas M. and Wang Z.-Y. ....................... 438 Theme 3 submitted papers Bioenergy potential of meadows of Ukraine Petrychenko V., Kurhak V. and Rybak S. ....................................................................... 453 Permanent grassland for anaerobic digestion: a novel insight into management–methane yield relations Herrmann C., Heiermann M., Schmidt F. and Prochnow A. ......................................... 456 Potential use of native Piptatherum miliaceum (L.) Coss. for forage production and bioenergy Porqueddu C., Sulas L., Re G.A., Sanna F., Franca A. and Melis R.A.M...................... 459 Second generation bioethanol production from Phalaris aquatica L. energy crop Pappas I.A., Kipparisides C. and Koukoura Z. .............................................................. 462 Hydrothermal processing of rush (Juncus spp.) and bracken (Pteridium aquilinum) dominant biomass from semi-natural landscape management Corton J., Ross A.B. , Lea-Langton A.R. , Donnison I.S. and Fraser M.D. .................. 465 The yield and variation of chemical composition of cocksfoot biomass after five years of digestate application Tilvikiene V., Kadziuliene Z., Dabkevicius Z., Šarūnaitė L., Šlepetys J., Pocienė L., Šlepetienė A. and Ceceviciene J. .................................................................................... 468 Evaluating sample preparation method effects on the specific methane yield of pre-and postensilage grass in an in vitro batch anaerobic digestion assay Nolan P., Doyle E.M. and O’Kiely P.............................................................................. 471 Theme 3 posters Area-specific bioenergy potentials from European floodplain grasslands – the DANUBENERGY project Bühle L., Hensgen F., Goliński P. and Wachendorf M. ................................................. 477 Permanent grasslands under different management as potential source of biomass for combustion in the Czech Republic Štýbnarová M., Mičová P., Karabcová H. and Látal O. ................................................ 480 What is the biomethane production potential of the available grassland biomass resource in Ireland? McEniry J., O’Kiely P., Wall D.M. and Murphy J.D. .................................................... 483 Evaluation of biomass yield of energy crops using waste products as fertilizers Rancane S., Gutmane I., Berzins P., Stesele V. and Dzene I. ......................................... 486 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands xvi Utilization of reed canary grass in pellet production Platace R. and Adamovics A........................................................................................... 489 Can specific methane yield of perennial ryegrass be reliably predicted? Herrmann A., Techow A., Kluß C., Loges R. and Taube F. ........................................... 492 Phytoestrogen content in clover (Trifolium spp.) and in grass stands depending on treatment and storage Řepková J., Nedělník J., Krtková V., Schulzová V., Novotná H., Hajšlová J. and Jakešová H. .................................................................................................................................... 495 Demand for K and P in reed canary grass (Phalaris arundinacea) during the harvest years Palmborg C., Lindvall E. and Gustavsson A.-M. ........................................................... 498 Organic seed production of yellow oat grass – preliminary results Macháč R. ....................................................................................................................... 502 Theme 4 ‘Livestock production’ Theme 4 invited papers Quality and authenticity of grassland products Moloney A.P., Monahan F.J. and Schmidt O. ................................................................ 509 Sustainable intensification of grass-based ruminant production Baumont R., Lewis E., Delaby L., Prache S. and Horan B. ........................................... 521 Theme 4 submitted papers Plant or animal needs - how to determine the optimal N intensity of grassland? Herrmann A., Techow A., Kluß C. and Taube F. ........................................................... 535 Energy expenditure of two grazing Holstein cow strains Schori F., Thanner S., Görs S., Metges C.C., Bruckmaier R.M. and Dohme-Meier F. . 538 Effects of mechanically separated dairy cow slurry on grazing performance Henry C.A., Lee M.A., McConnell D.A., Wood B.L. and Roberts D.J. .......................... 541 Effect of growth stage on the phosphorus content of grass, and on phosphorus excretion on dairy farms Van Middelkoop J.C., Holshof G. and Plomp M. ........................................................... 544 Effects of concentrate levels on milk production and traffic of cows milked by a mobile automatic milking system on pasture Lessire F., Hornick J.L. and Dufrasne I. ........................................................................ 547 Relationship between fatty acid content and nutritive value of perennial ryegrass (Lolium perenne) Morgan S.A., Huws S.A., Tweed J.K.S., Hayes R.C. and Scollan N.D. .......................... 550 Can we use the fatty acid composition of bulk milk to authenticate the diet composition? Martin B., Coppa M., Chassaing C., Agabriel C., Borreani G., Barcarolo R., Baars T., Kusche D., Harstad O.M., Verbič J., Golecký J. and Ferlay A. ..................................... 553 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands xvii Effect of dietary supplementation on milk production and milk composition of grazing dairy cows in late lactation Reid M., O’Donovan M., Lalor S.T.J., Bailey J.S., Elliott C.T., Watson C.J. and Lewis E. ........................................................................................................................................ 556 Combining robotic milking and grazing Brocard V., Huneau T., Huchon J-C. and Dehedin M. .................................................. 559 GPS tracking of Old Norwegian ewes on a coastal heathland-dominated island Lind V. and Bär A. .......................................................................................................... 563 Nutritive value of leaf fodder from the main woody species in Iceland Hejcman M., Hejcmanová P., Pavlů V., and Thorhallsdottir A.G. ................................ 566 Sensory quality and authentication of lamb meat produced from legume-rich forages Devincenzi T., Prunier A., Nabinger C. and Prache S. .................................................. 569 Dynamics of dry matter intake in livestock production systems in the Netherlands Van den Pol-van Dasselaar A., Nolles J.E., Philipsen A.P. and Stienezen M.W.J. ....... 573 Cutting strategy of a five-cut system in different grassland mixtures Søegaard K. .................................................................................................................... 576 Theme 4 posters Conserving high moisture spring field bean (Vicia faba L.) grains O’Kiely P., Stacey P. and Hackett R. ............................................................................. 583 Fava bean-rapeseed intercrop as a sustainable alternative to Italian ryegrass: production, forage quality and soil fertility evolution Jiménez J.D., Vicente F., Benaouda M., Soldado A. and Martínez-Fernández A. ......... 587 Fava bean-rapeseed and maize silages growing under organic fertilization as a sustainable alternative for dairy cow feeding Jiménez J.D., Martínez-Fernández A., González A., Soldado A., de la Roza-Delgado B. and Vicente F. ................................................................................................................. 590 Effect of harvest and ensiling on different protein fractions in three different legumes Wyss U., Girard M., Grosse Brinkhaus A., Arrigo Y., Dohme-Meier F. and Bee G. ..... 593 Nutritive value evaluation of some grasses and legumes for ruminants Tomić Z., Bijelić Z., Mandić V., Simić A., Ruzić-Muslić D., Stanišić N. and Maksimović N. ........................................................................................................................................ 597 Forage quality in legumes and non-leguminous forbs Elgersma A., Søegaard K. and Jensen S.K. .................................................................... 600 Feed value of restrictedly and extensively fermented organic grass-clover silages from spring and summer growth Bakken A.K., Vaga M., Hetta M., Randby Å.T. and Steinshamn H. ............................... 603 Feeding, mycological, and toxicological quality of haylage Nedelnik J., Strejckova M., Cholastova T., Both Z., Palicova J. and Hortova B. .......... 606 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands xviii Mycotoxin and chemical characteristics of silages collected from horizontal silos on farms in Co. Meath, Ireland - a pilot study McElhinney C., Danaher M., Elliott C. and O'Kiely P. ................................................. 610 Prediction of energy content of grass silages depending on grass and ensiling conditions Pickert J. and Weise G. .................................................................................................. 613 Predicting organic matter digestibility by two enzymatic in vitro methods Beecher M., Baumont R., Aufrère J., Boland T.M., O’Donovan M., Galvin N., Fleming C. and Lewis E. ................................................................................................................... 616 Carbon sequestration in silage maize as affected by N fertilization Herrmann A., Böttger F., Lausen P. and Taube F. ........................................................ 619 Accuracy of forage intake estimation with three different indirect prediction models Salas-Reyes I.G., Martínez-Fernández A., Morales-Almaráz E., Jiménez J.D., AlbarránPortillo B., de la Roza-Delgado, B. and Vicente F. ....................................................... 622 Compatibility of using TiO2 and the faecal near-infrared reflectance spectrometry for estimation of cattle intake Vandermeulen S., Decruyenaere V., Ramirez-Restrepo C. and Bindelle J. ................... 625 Dry matter intake and in vivo digestibility of four perennial ryegrass cultivars Garry B., O’Donovan M., Boland T.M. and Lewis E. .................................................... 628 Accurate monitoring of the rumination behaviour of cattle using IMU signals from a mobile device Andriamandroso A.L.H., Lebeau F. and Bindelle J........................................................ 631 Energy consumption and greenhouse gas emissions of DAIRYMAN farms in South-West Germany Jilg T., Herrmann K., Hummler T. and Elsaesser M. ..................................................... 634 The DAIRYMAN-Sustainability-Index (DSI) as a tool for comparing dairy farms Elsaesser M., Herrmann K. and Jilg T. .......................................................................... 638 Dairy system sustainability in relation to access to grazing: a case study Decruyenaere V., Herremans S., Visser M., Grignard A., Jamar D., Hennart S., Campion M. and Stilmant D........................................................................................................... 641 Beef productivity on the North Wyke Farm Platform in two baseline years Thompson J.B., Orr R.J., Dungait J., Murray P.J. and Lee M.R.F. ............................... 644 Milk production of sheep fed on preserved forage in winter and grazing in spring Stoycheva I., Kirilov A. and Simeonov M. ...................................................................... 647 Evaluation of a home-grown crimped lupin and barley concentrate feed for finishing lambs Marley C.L., Jones H., Theobald, V., Sanderson R. and Fychan R. .............................. 651 Optimal base temperature for computing growing degree-day sums to predict forage quality of mountain permanent meadow in South Tyrol Romano G., Schaumberger A., Piepho H.-P., Bodner A. and Peratoner G. .................. 655 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands xix Added value chain of the dairy industry and its development in Central Switzerland Hofstetter P. .................................................................................................................... 658 Economics of grazing Van den Pol-van Dasselaar A., Philipsen A.P. and de Haan M.H.A. ............................ 662 Does early spring grazing stimulate spring grass production? van Eekeren N., Rietberg P., Iepema G. and de Wit J. ................................................... 665 Use of milk fatty acid composition to authenticate cow diets Coppa M., Revello-Chion A., Giaccone D., Comino L., Tabacco E. and Borreani G. .. 668 Potential lipid markers of plant species from grasslands to authenticate mountain dairy foods Barron L.J.R., Aldezabal A., Valdivielso I., Bustamante M., Amores G., Virto M., Ruiz de Gordoa J.C. and de Renobales M................................................................................... 671 Is phytanic acid a suitable marker for authentication of milk and dairy products from grassfed cows or organic farming systems? Capuano E., Elgersma A., Tres A. and Ruth S.M. van ................................................... 674 Potential of fertilized grass clover swards to produce adequate herbage to support dairy cow milk production in high stocking rate grass based systems Egan M.J., Lynch M.B. and Hennessy D. ....................................................................... 677 Feeding strategies and feed self-sufficiency of dairy farms in the lowland and mountain area of Western Switzerland Ineichen S., Piccand V., Chevalley S., Reidy B. and Cutullic E. .................................... 680 Weather effects and cattle behavioural characteristics Halasz A. and Nagy G. ................................................................................................... 683 Relationship between the composition of fresh grass-based diets and the excretion of dietary nitrogen from dairy cows Moorby J.M. ................................................................................................................... 686 Theme 5 ‘MultiSward’ Theme 5 invited papers Multi-species swards and multi scale strategies for multifunctional grassland-base ruminant production systems: An overview of the FP7-MultiSward project Peyraud J.L., van den Pol–van Dasselaar A., Collins R.P., Huguenin-Elie O., Dillon P., Peeters A. ........................................................................................................................ 695 Theme 5 submitted papers Biomass production in multispecies and grass monoculture swards under cutting and rotational grazing Collins R.P., Delagarde R. and Husse S. ....................................................................... 719 Nitrogen capture in mixed swards benefits from temporal complementarity among species Husse S., Huguenin-Elie O., Buchmann N. and Lüscher A. ........................................... 722 Holshof G. and van den Pol–van Dasselaar A. .............................................................. 725 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands xx Interest of multi-species swards for pasture-based milk production systems Roca-Fernández A.I., Peyraud J.L., Delaby L., Lassalas J. and Delagarde R. ............. 728 Influence of ryegrass alone or blended with clover and chicory on feed intake and growth performance of steers Morel I., Schmid E., Soney C., Aragon A. and Dufey P.-A. ........................................... 731 Associative effects between forage species on intake and digestive efficiency in sheep Niderkorn V., Martin C., and Baumont R. ...................................................................... 734 Effects of restricting access time to pasture on late lactation dairy cow production Kennedy E., Garry B., Ganche E., O’Donovan M., Murphy J.P., and Hennessy D. ..... 737 Theme 5 special paper Grassland term definitions and classifications adapted to the diversity of European grasslandbased systems Peeters A., Beaufoy G., Canals R.M., De Vliegher A., Huyghe C., Isselstein J., Jones G., Kessler W., Kirilov A., Mosquera-Losada M.R., Nilsdotter-Linde N., Parente G., Peyraud J.-L., Pickert J., Plantureux S., Porqueddu C., Rataj D., Stypinski P., Tonn B., van den Pol – van Dasselaar A., Vintu V. and Wilkins R.J. ............................................................... 743 Theme 5 submitted papers Roles and utility of grasslands in Europe De Vliegher A., Van Gils B. and van den Pol-van Dasselaar A. .................................... 753 An indicator-based tool to assess environmental impacts of multi-specific swards Plantureux S., Dumont B., Rossignol N., Taugourdeau S. and Huguenin-Elie O. ......... 756 Assessment of ecosystem services provided by grasslands and grassland-based systems by indicators: a regional perspective Peeters A., Stolze M., Goliński P., Scimone M., Moakes S., Thorne F. and Plantureux S. ........................................................................................................................................ 759 Threats and opportunities for European grassland areas under different market and policy scenarios Hecht J., Moakes S., Offermann F. and Peeters A. ........................................................ 763 Appreciation of the functions of grasslands by European stakeholders van den Pol-van Dasselaar A., Goliński P., Hennessy D., Huyghe C., Parente G. and Peyraud J.-L. .................................................................................................................. 766 Theme 5 posters Effect of grassland management in autumn on the mineral N content in soil De Vliegher A. and Vandecasteele B. ............................................................................. 773 Impact of plant diversity, with equal number of grass and legume species, on sward productivity and legume content under contrasted mowing management in a low input system Jamar D., Clement C., Seutin Y., Planchon V., Campion M. and Stilmant D. ............... 776 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands xxi The effect of different fodder galega-grass mixtures and nitrogen fertilization on forage yield and chemical composition Meripõld H., Lättemäe P., Tamm U. and Tamm S. ........................................................ 780 Grass only and grass-white clover (Trifolium repens L.) swards: herbage production and white clover performance Egan M., Enriquez-Hidalgo D., Gilliland T., Lynch M.B. and Hennessy D. ................. 783 The persistence of perennial ryegrass cultivars (Lolium perenne L.) in binary mixtures with white clover (Trifolium repens L.) under grazing Gregis B. and Reidy B. ................................................................................................... 786 Grass-only and grass-white clover (Trifolium repens L.) swards: dairy cow production Enriquez-Hidalgo D., Egan M., Gilliland T., Lynch M.B. and Hennessy D. ................. 789 Effect of grass-only compared to grass-white clover swards on cow rumen function and methane emissions Enriquez-Hidalgo D., Lewis E., Gilliland T. and Hennessy D. ...................................... 792 Animal choice for grass-based systems Delaby L., Hennessy D., Gallard Y. and Buckley F........................................................ 795 Effect of sheep breed on lamb production from lowland pasture under continuous stocking Goliński P., Golińska B. and Biniaś J. ........................................................................... 798 Appreciation of the functions of grassland by Belgian stakeholders De Vliegher A., Dufrasne I., Schellekens A., Peeters A., Van den Pol-van Dasselaar A. ........................................................................................................................................ 801 Appreciation of the functions of grassland by Dutch stakeholders Van den Pol-van Dasselaar A. and Stienezen M.W.J. .................................................... 804 Appreciation of the functions of grassland by French stakeholders Huyghe C., Peyraud J.-L., Brocard V., Van den Pol-van Dasselaar A. ......................... 807 Appreciation of the functions of grassland by Irish stakeholders Hennessy D. and Van den Pol-van Dasselaar A. ........................................................... 810 Appreciation of the functions of grassland by Italian stakeholders Parente G. and Van den Pol-van Dasselaar A. .............................................................. 813 Appreciation of the functions of grassland by Polish stakeholders Goliński P., Van den Pol-van Dasselaar A. and Golińska B.......................................... 816 Theme 6 ‘Approaches to forage crop improvement’ Theme 6 submitted papers Genomic characterization of survivor populations of red clover by GBS Ergon Å. and Rognli O.A. ............................................................................................... 823 Towards genomic selection in perennial ryegrass genetic improvement Skøt L., Grinberg N.F. , Lovatt A., Hegarty M., Macfarlane A., Blackmore T. Armstead I., King R.D. and Powell W.A., ........................................................................................... 826 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands xxii Prospects for introducing genomic selection into forage grass breeding Fè D., Ashraf B., Byrne S., Czaban A., Roulund N., Lenk I., Asp T., Greve Pedersen M., Janss L., Jensen J. and Jensen C.S. ................................................................................ 830 Population selection within perennial ryegrass cultivars under simulated grazing Cashman P., Gilliland T.J., O’Donovan M. and McEvoy M.......................................... 833 Genetic gain in yield of perennial ryegrass (Lolium perenne), Italian ryegrass (Lolium multiflorum Lam.) and hybrid ryegrass (Lolium x boucheanum Kunth) cultivars in Northern Ireland Recommended Lists 1972-2013 McDonagh J., McEvoy M., O’Donovan M. and Gilliland T.J. ...................................... 836 Variation in the reproductive development of perennial ryegrass (Lolium perenne) cultivars Wims C.M., Lee J.M., Rossi L. and Chapman D.F. ........................................................ 840 Pasture profit index: updated economic values and inclusion of persistency McEvoy M., McHugh N., O’Donovan M., Grogan D. and Shalloo L. ........................... 843 Theme 6 posters The effect of resistance to mildew infection on ruminal fermentation of Lolium perenne Claes J., Davies T.E., Rees Stevens P., Wilkinson T., Mur L.A.J. and Kingston-Smith A.H. ........................................................................................................................................ 849 Disease resistance in red clover (Trifolium pratense L.) to stem nematodes and Sclerotinia Lowe M., Kelly R., Skøt L. and Mizen K.A. .................................................................... 852 Selection of white clover (Trifolium repens L.) for improved phosphorus use efficiency Lloyd D.C., Vale J.E. and Marshall A.H. ....................................................................... 855 Selection of contrasting cold-tolerant white clover genotypes from twenty-eight populations naturalized in southern Chile and Argentina Acuña H., Inostroza L. and Pino M.T. ............................................................................ 858 Analysis of changes in population structure over time in components of multi-species swards Kelly R., Skøt L., Skøt K.P. and Collins R.P. .................................................................. 861 Temporal genetic shifts in mono- and bi-specific swards of perennial ryegrass and red clover Cnops G., Muylle H., De Vliegher A., Vleugels T. and Roldán-Ruiz I. .......................... 864 Persistence of red clover (Trifolium pratense L.) varieties in mixed swards over four harvest years Marshall A.H., Lowe M. and Vale J.E. ........................................................................... 867 Developing an optimal sampling strategy to assess the quality of perennial ryegrass varieties on a national variety evaluation scheme Burns G.A., O’Kiely P., Grogan D., Conaghan P. and Gilliland T.J. ........................... 870 Screening reveals opportunities for high sugar cultivars of Lolium perenne L. Suter D., Hofer D. and Lüscher A. ................................................................................. 874 The influence of autumn closing date and spring opening date on herbage production and quality in spring and throughout the growing season Lawrence D., O’Donovan M., Boland T.M. and Kennedy E. ......................................... 877 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands xxiii Increasing protein yields from grassland by reseeding of legumes Elsaesser M., Engel S., Breunig J. and Thumm U. ......................................................... 880 Change in birdsfoot trefoil (Lotus corniculatus L.) nutritive value with stem elongation, flowering and pod formation Hunt S.R., Griggs T.C. and MacAdam J.W..................................................................... 884 Studies on forage quality of weed species in subalpine meadows in the Southeastern Carpathians of Romania Ciopată A-C., Maruşca T., Oprea G., Mocanu V., Blaj V.A. and Haş E.C.................... 887 Index of Authors .................................................................................................................. 891 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands xxiv Special paper ‘The European Grassland Federation at 50: past, present and future’ Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 1 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 2 The European Grassland Federation at 50: past, present and future Prins W.H.1 and Kessler W.2 1 Hollandseweg 382, NL-6705BE Wageningen, The Netherlands 2 Agroscope, Reckenholzstrasse 191, CH-8046 Zürich, Switzerland Corresponding author: prinsw@upcmail.nl (www.europeangrassland.org) Abstract In 1963 the European Grassland Federation (EGF) was founded in the UK. The origin and membership over fifty years are described. The changes in grassland research and development are reflected in the subjects of the General Meetings and Symposia. The successful development in more recent years of Working Groups, Master Classes and Workshops is described. The links to the European Union and the future direction of the EGF are explored. The EGF continues to play an important role as a non-political forum for exchanging and communicating the results of grassland science in Europe and in bringing grassland scientists together. Future challenges for the EGF are discussed. Keywords: grassland research, conferences, publications, education, European Union Introduction In the years following World War II, scientists contacted each other bilaterally at occasional European conferences and also at International Grassland Congresses. To improve contacts, in 1963 the European Grassland Federation (EGF) was officially established at a special Symposium at Hurley, UK, organized by the British Grassland Society. At this inaugural meeting, representatives in attendance from grassland societies and countries in Europe agreed to the formation of the EGF, adopted the constitution and appointed the first Executive Committee charged with organization of the first General Meeting in The Netherlands in 1965 (Powell et al., 1995). Since that time important changes in the European political scene have taken place, and these have affected the EGF. This article reviews important developments in EGF over 50 years, with particular attention to the years since 2003. The history of the first 40 years was described in full by Prins (2004) and is also available from the EGF website (www.europeangrassland.org). EGF organization The objectives of the EGF were and still are: to facilitate and maintain close contact between European grassland organizations, to promote the interchange of scientific and practical experience between grassland experts, and to initiate symposia and other meetings between European grassland organizations. As was explained before (Prins, 2004), the EGF has a simple structure with membership open to national or representative grassland organizations in Europe. By 2013 thirty-one countries had become Full Members (Table 1). For those countries without a national or representative organization, individual grassland workers may become Corresponding Members to represent their country. In 2013 there were four corresponding members (Table 1). Membership of the EGF is gratis. Contact between EGF and grassland specialists throughout Europe is maintained through country representatives who act as intermediaries between their national colleagues and the Federation Secretary. Europe has been divided into seven regions and each region is represented in the Executive Committee which manages the affairs of the Federation. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 3 Table 1. Representation of European countries in seven regions: thirty-one full members and four corresponding members* as of 2013. EGF region Members and corresponding members* Central Europe Austria, Czech Republic, Germany, Hungary, Slovakia, Switzerland Western Europe Belgium, France, Ireland, The Netherlands, United Kingdom, Luxembourg* Northern Europe Denmark, Finland, Norway, Iceland, Sweden North-eastern Europe Estonia, Lithuania, Poland, Latvia* Eastern Europe Bulgaria, Romania, Russia, Ukraine South-eastern Europe Bosnia-Herzegovina, Croatia, Slovenia, Serbia, Macedonia* Southern Europe Greece, Italy, Portugal, Spain, Albania* The EGF has Honorary Life Presidents who help to take care of continuity. Dr W. Davies, the great stimulator and founding father of the EGF, was voted the first Honorary Life President at the inaugural meeting in 1963 and he occupied that position until his death in 1968, playing an active role in the EGF (Powell et al., 1995). Three more founding fathers, namely Prof. 't Hart, Dr. Järvi and Dr. Caputa were elected in 1980 and 1982 (Table 2). Not until 2000 were any further Honorary Life Presidents elected. To make the choice more objective, the Executive Committee decided that candidates should be able to show a set of achievements within EGF (Prins, 2004). Since then, seven more have been elected (Table 2). They are expected to attend meetings of the Executive Committee and to support and advise the Federation Secretary and Organizing Committees. Table 2. Honorary Life Presidents since 1963. Name Country Year elected Year of death Dr. W. Davies UK 1963 1968 Prof. M.L. 't Hart The Netherlands 1980 2005 Prof. N.G. Andreev USSR 1980 1996 Dr. V. Järvi Finland 1982 1987 Dr. J. Caputa Switzerland 1982 1992 Dr. J. Frame UK 2000 2006 Prof. G. Blagoveschensky Russia 2000 Prof. L. 't Mannetje The Netherlands 2002 Prof. R.J. Wilkins UK 2002 Prof. J. Nösberger Switzerland 2004 Dr. W.H. Prins The Netherlands 2004 Prof. J. Parente Italy 2010 2008 Scientific Advisory Board To assist the Executive Committee with strategy and focus on the right themes, subjects, authors and keynote speakers, a Scientific Advisory Board was established in 2012. The Board consists of independent senior researchers who are well known and actively involved with Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 4 grassland research. The Board does not meet, but is consulted by the Federation Secretary on an ad hoc basis. Presently the Board comprises some 13 experts. The EGF Secretariat and Fund In the early years EGF business was handled by the Secretary of the current Organizing Committee of the General Meeting. Thus the EGF secretariat changed hands every two years. This hampered the contact between member countries and within the Executive Committee. As from 1973, a voluntary continuing Federation Secretary was elected. Through the years this post was filled by Dr. R. Tayler (UK, 1973-1978), Dr. J.W. Minderhoud (The Netherlands, 1978-1985), Dr. W.H. Prins (The Netherlands, 1985-2004) and Dr. W. Kessler (Switzerland, since 2004). With the growth of EGF, the decision to ensure a high and consistent standard in the level of EGF conferences and proceedings, as well as to promote EGF as European forum of grassland science, the task of the Federation Secretary has become more and more important. These days communication is mostly through e-mail, with general information distributed via the EGF website mentioned above. However, the Secretary has to ensure that important messages and documents are printed and kept for the archives of the Federation. As country membership is free, secretarial costs of EGF administrative affairs had to be covered by conference budgets. In the early years this often meant that the costs were met by the Secretary’s employer. However, in 1982 in Reading, a fund was founded so as to facilitate continuing activities of the EGF. Money became available from the financial budgets of EGF conferences via a levy of 10 Swiss Francs (SFr) per paying participant, as part of the registration fee. From the start of the EGF Fund, the Swiss Grassland Society (AGFF) has handled the fund, for which EGF is grateful. The EGF Fund is supplemented by the net proceeds of sales of EGF Proceedings in the series 'Grassland Science in Europe', presently carried out through the good office of the Federation Secretary in Zürich. General Meetings and Symposia From the start it was decided to use only English for general communication at EGF conferences, both orally and in writing. The EGF can be proud to have opted for this unifying solution and is happy that native English-speaking colleagues are always prepared to assist with 'anglicising' the conference papers. These conferences are General Meetings lasting four days, intended for a wide group of grassland people, and Symposia of three days, for specialist topics. The General Meetings include scientific and social programmes and a Business Meeting. Table 3 lists the General Meetings since 1965. The scientific programmes of the first ten meetings generally concerned the production, quality, forage conservation and economics of native and sown grasslands as affected by climate, species, soil and nutrition. From the 11th General Meeting in 1986, the topics reflected the wider implications of grassland use, e.g. energy, sustainability, society, ecology, biodiversity and multi-functional grasslands. At the first General Meeting in Wageningen some 100 participants were present. Through the years the number has increased and nowadays about 300–400 people are expected to attend the General Meetings. The record was 600 in La Rochelle (2002). Naturally, the majority originate from Europe, but participants from outside are welcomed and there are regular attendants from North America and Japan. For specialist topics EGF invites member nations to organize symposia. In the early years these were occasional symposia, mostly dealing with conventional grassland farming. They featured mainly agronomic subjects (Hurley, 1963; Aberdeen, 1968; Wageningen, 1980 and 1987; Graz, 1991) or were more focused towards economics (Versailles, 1965) or forage conservation (Brighton, 1979; Braunschweig, 1991). Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 5 Table 3. EGF General Meetings, 1963–2013. 1st Nitrogen and Grassland. Wageningen, The Netherlands. June-July 1965 2nd Evaluation of grassland production. Versailles, France. May 1967 3 rd Conservation of grassland products. Braunschweig, Germany. June 1969 4 th Comparison between natural and artificial grassland. Lausanne, Switzerland. June 1971 5th 6 th 7th Quality of herbage. Uppsala, Sweden. June 1973 Pasture and forage production in seasonally arid climates. Madrid, Spain. April 1975 Constraints to grass growth and grassland output. Gent, Belgium. June 1978 8 th Forage production under marginal conditions. Zagreb, Yugoslavia. June 1980 9 th Efficient grassland farming. Reading, UK. September 1982 10th 11 th Impact of climate on grass production and quality. Ås, Norway. June 1984 Grassland – facing the energy crisis. Setubal, Portugal. May 1986 12th Grassland and animal production, now and in the future. Dublin, Ireland. July 1988 13th Soil–grassland–animal relationships. Banská Bystrica, Czechoslovakia. June 1990 th Sustainable production from grassland. Lahti, Finland. June 1992 15th Grassland and society. Wageningen, The Netherlands. June 1994 14 th Grassland and land use systems. Grado, Italy. September 1996. Grassland Science in Europe, Volume 1 17th Ecological aspects of grassland management. Debrecen, Hungary. May 1998. Grassland Science in Europe, Volume 3 18th Grassland farming: balancing environmental and economic demands. Aalborg, Denmark. May 2000. Grassland Science in Europe, Volume 5 19th Multi-function grasslands: quality forages, animal products and landscapes. La Rochelle, France. May 2002. Grassland Science in Europe, Volume 7 20th Land use systems in grassland dominated regions. Luzern, Switzerland. June 2004. Grassland Science in Europe, Volume 9 21st Sustainable grassland productivity. Badajoz, Spain. April 2006. Grassland Science in Europe, Volume 11 22nd Biodiversity and animal feed; future challenges for grassland production. Uppsala, Sweden. June 2008. Grassland Science in Europe, Volume 13 23rd Grassland in a changing world. Kiel, Germany. August-September 2010. Grassland Science in Europe, Volume 15 24th Grassland – a European resource? Lublin, Poland. June 2012. Grassland Science in Europe, Volume 17 16 From 1997 onwards, EGF Symposia have been organized regularly in alternate years between General Meetings. Table 4 shows that, since then, less attention has been paid to conventional grassland production, with precedence given to subjects such as grassland biodiversity (Poland, 1997; Estonia, 2005), grassland and woody plants (Greece, 1999), organic grassland farming (Germany, 2001), effects on environment and economy (Bulgaria, 2003; Belgium, 2007; Austria, 2011; Iceland, 2013) and alternative functions (Czech Republic, 2009). Generally about 150-200 people are expected to attend the symposia, and the record was 280 at the Brighton, UK symposium (1979). Many symposia are also attended by participants from outside Europe. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 6 Table 4. EGF Symposia, 1963–2013. Symposia 1st 2 nd 3rd The agronomic evaluation of grassland. Hurley, UK. September 1963 Economic problems relating to the production of forages. Versailles, France. March 1965 Hill land productivity. Aberdeen, UK. June–July 1968 4 th Forage conservation in the ‘80s. Brighton, UK. November 1979 5 th The role of nitrogen in intensive grassland production. Wageningen, The Netherlands. August 1980 6th Animal manures on grassland and fodder crops: fertilizer or waste? Wageningen, The Netherlands. August–September 1987 7th Forage conservation towards 2000. Braunschweig, Germany. January 1991 8 th Grassland renovation and weed control in Europe. Graz, Austria. September 1991 9th Management for grassland biodiversity. Lomza, Poland. May 1997. Grassland Science in Europe, Volume 2 10th Grasslands and woody plants in Europe. Thessaloniki, Greece. May 1999. Grassland Science in Europe, Volume 4 11th Organic grassland farming. Witzenhausen, Germany. July 2001. Grassland Science in Europe, Volume 6 12th Optimal forage systems for animal production and the environment. Pleven, Bulgaria. May 2003. Grassland Science in Europe, Volume 8 13th Integrating biodiversity and efficient grassland farming, Tartu, Estonia. August 2005. Grassland Science in Europe, Volume 10 14th Permanent and temporary grassland: plant, environment, economy. Gent, Belgium. September 2007. Grassland Science in Europe, Volume 12 15th Alternative functions of grassland. Brno, Czech Republic. September 2009. Grassland Science in Europe, Volume 14 16th Grassland farming and land management systems in mountainous regions. Gumpenstein, Austria. August 2011. Grassland Science in Europe, Volume 16 17th The role of grasslands in a green future – threats and perspectives in less favoured areas, Iceland. June 2013. Grassland Science in Europe, Volume 18 Events for young scientists, special workshops and project meetings Since 2002, EGF General Meetings regularly include special offers, such as seminars or master classes, for scientists younger than 35 years, active in research, resident of an EGF country, with an accepted contribution for the conference and thus eligible for a reduced registration fee. Moreover special workshops organized outside the scientific programmes of EGF conferences provide the opportunity to discuss subjects between researchers. Sometimes policy makers from the EU have participated in these workshops. Examples are the recent workshops held in Poland (2012) on: i) Grassland Production in Europe, about the information needs of Eurostat as well as the MultiSward Book Project, and on ii) EGF Resolutions on the Common Agricultural Policy (CAP) Reform of 12 October 2011. The annual EGF conferences are also on offer, and have been used intensively for years, for project management and project team meetings of European research projects of various programmes (Interreg, COST, EU Seventh Framework Programme for Research FP7, EU Marie Curie Actions etc.). Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 7 These additional events connected to EGF conferences have proved to be very useful for the attendants, not only for technical reasons, but also for social reasons by extending and strengthening their networks. From Conference Proceedings to the series Grassland Science in Europe The Proceedings of General Meetings and Symposia have all appeared in print. At first each Organizing Committee took care of its own proceedings and was thus influenced by local editing and printing procedures. In the 1990s it was decided to opt for more standardization. All papers were read and checked by at least two referees and native English-speaking colleagues started to help with ‘anglicising’ the manuscripts. The final standardization took place with the proceedings appearing as consecutive volumes in the series 'Grassland Science in Europe' for both General Meetings (Table 3) and Symposia (Table 4). For 2014 it is already Volume 19 in this series. Discussions are taking place as to whether or not EGF should continue to produce paper copies of the Proceedings. These have been available on compact disc and on the EGF website since 2004. As mentioned above, EGF welcomes participants from outside Europe to attend the conferences. However, papers from outside are only accepted if at least one author is from Europe and the work is relevant to Europe. Grass and Forage Science, the official journal of EGF In 1996 the British Grassland Society (BGS) and EGF joined hands in the publication of Grass and Forage Science. This cooperation, with the financial responsibility held by BGS, has contributed to the journal becoming much more international in its scope and focus. In the 1990s, papers submitted to Grass and Forage Science were predominantly from Englishspeaking countries, with very few from elsewhere in Europe. By 2013, around one-third of papers published in Grass and Forage Science were from authors based in mainland Europe. Many European countries are also now represented in the journal's Advisory Editors and Associate Editors. The journal has also increased its presence and attractiveness to authors and readers in other grassland areas of the world, notably China and Latin America. Publication of the review paper on international grazing terminology in March 2011 was an important milestone in the journal's increasingly international appeal. Other important review papers, including some that have been adapted from plenary papers presented at recent EGF General Meetings and Symposia, have also contributed to the journal receiving increased citation metrics and downloads from the journal website. The EGF is grateful to BGS for taking care of the management of Grass and Forage Science. Working Groups In the early years EGF had already set up selected committees on specialist topics like 'Methods of Forage Production Experiments' or 'Nomenclature'. Since 2002 EGF has acted as an umbrella for Working Groups of scientists from different European countries. These groups are welcome to organize their own meetings at EGF conferences and report at each EGF General Meeting. Recently they have started to report also on the EGF website. There are four active groups and one group has finished its work. The groups are listed below: - Grassland re-sowing and grass-arable rotations (2002-) - Dairy farming systems (2004-) - Analytical methods for measuring fatty acids in forage (2006-2010) - Grazing (2008-) Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 8 - Semi-natural grasslands (2011-). More details can be found at the EGF website. The modern EGF conference provides a good forum for the encouragement of these and other multi-national and multi-disciplinary projects and for reporting their progress. This contrasts with the early days of EGF, when most papers reported the results of work carried out in one institute in one country and there were no working groups. European Union (EU) The EGF is the only European-wide non-political body that organizes regular meetings to discuss scientific and socioeconomic issues related to European grassland science, management and farming in the broadest sense. The Federation realized that it was in a position to provide expert information required for the evolution of EU policies, whilst the decisions of the EU would affect the adoption of technologies being developed by EGF scientists. Efforts have been made over 20 years to increase contact between EGF and the EU. All General Meetings since 2002 have had plenary papers concerning the effects of EU policies on grassland. EU-officials came to EGF conferences and notably in 2008 the session on ‘Grasslands – Challenge for the Future’ included an address by the EU Commissioner for Agriculture and Rural Development. Exchanges between officials of EGF and the EU have occurred since 1997, but it was not until 2011 that any continuing relationships developed. It was agreed with EU officials that workshop sessions would be held at the General Meeting in Poland in 2012 with EU participation and that the Federation would supply input to the debate on evolving EU policies for grassland. A Working Group met in Lublin and produced a series of resolutions that were adopted at the Business Meeting and submitted to the EU in July 2012. Following this, the EGF was asked to provide advice to Eurostat on grassland. Prof. A. Peeters led a Working Group that submitted to Eurostat late in 2013 a paper entitled ‘Grassland Term Definitions and Classifications Adapted to the Diversity of European Grassland-Based Systems’. In another significant development in October 2012, the EU invited EGF to nominate a representative on the Steering Board for the European Innovation Partnership on Agricultural Productivity and Sustainability (EIP-AGRI). After that, the Federation Secretary and Prof. A. Peeters were decisively involved in the EU’s development of the Strategic Implementation Plan for EIP-AGRI. It is too early to identify policy decisions that have arisen directly from inputs from EGF, but contacts are now well established and should enable more rapid and effective use of findings from our scientists in the evolution of EU policies. In addition to these contacts at the policy level, EGF plays an important part in the execution and reporting of multi-national research programmes funded by the EU. The personal relationships between scientists at EGF meetings have been important in developing consortia to bid for and undertake these programmes. Increasingly, not only are early results reported at EGF meetings, but also working meetings of consortia are often attached to EGF meetings. The future: challenges and perspectives The paper on EGF History written ten years ago (Prins, 2004) included a number of statements on 'The Future'. We now comment on the extent to which our aspirations in these areas have been realized. Some new challenges are added. 1. ‘Grasslands are now used increasingly for multi-functional purposes rather than solely for animal production. Increasing involvement of scientists from basic, social and environmental sciences is therefore to be expected’. In the past ten years, indeed more participation from non-agronomic scientists took place in the conferences, which was, of course, related to the conference themes. Certainly more Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 9 speakers from outside the agronomy scene were invited to present a plenary paper. There is an increasing interest in how to meet the challenges of the various services which grassland provides for the human well-being and for its contribution to natural capital and environmental services. Furthermore, the evident atmospheric changes require new science-based information on how to support the preparedness of agriculture for climate change. For the future, the effort by EGF to reach scientists outside the regular agronomy world should be increased, the more so because some renowned grassland research institutes have been substantially reduced or even closed and some university chairs of grassland science have been discontinued. New grassland-relevant scientific input has originated in institutes of ecology and other scientific societies offer platforms for the discussion of grassland issues, but a professional link to the agricultural relevance is often missing. The actual scientific environment for grassland science has substantially expanded; it requires strong collaborations among many disciplines, including agronomists, animal scientists, soil scientists, ecologists, economists, sociologists and modellers. Closer cooperation, both between scientists in different disciplines and between different Scientific Societies, would produce 'win-win' situations. 2. ‘In order to provide emphasis on new and important areas, and to facilitate in-depth studies, the EGF encourages the formation of Special Interest Groups. There are many advantages of such Groups convening during or around EGF conferences.’ The section above on Working Groups shows that good progress has been made in this initiative. For the future, this effort should be maintained but EGF should try to involve more scientists from outside the disciplines that represented the previous core of EGF, as mentioned under 1. 3. ‘The EGF is the only European-wide non-political body that organizes regular meetings to discuss scientific and socioeconomic issues related to European grassland science, management and farming in the broadest sense. In order to spread the social and political impact of the knowledge and understanding of grassland matter, the Secretariat of the EGF plans to issue press statements to the general and farming media in countries in Europe, before and after meetings.’ The planned press statements have not appeared regularly; the EGF, run by volunteers, should find a way to get this organized properly. A success is that, since 2009, at the end of every conference a senior scientist summarizes the results and this Synthesis then appears on the EGF website. The following are new statements regarding challenges and perspectives: 4. There is a growing tendency to base important long-term policy decisions in Europe on results of short-term research. The EGF should send a wake-up call that these decisions can be dangerous and often misleading. The EGF stresses the need for long-term experiments in strategic agro-ecological systems. In the past those experiments repeatedly demonstrated that long-term responses differed markedly from short-term responses. 5. With the increasing world population and use of agricultural products for other purposes than food or feed, European politics may require that farmers should become more autonomous in the supply of feed, particularly protein, for the livestock sector. As a consequence grassland farming may be confronted with complex challenges which have to be solved by agricultural and environmental scientists together with the social science communities. 6. There is concern that the education of future grassland experts should be improved in most European countries. In this, EGF may be able to contribute by facilitating networking and exchange of information, not only via the usual conferences, working groups, master classes and workshops, but also by facilitating internships, mentors and career resources for the mutual benefit of early-career scientists, experts and interested institutions. Such actions will certainly support the efforts of the EGF to be perceived as the voice of grassland issues in Europe. For decades EGF has played an integral role in advancing grassland research and management. It continues to anticipate new challenges and is prepared to address emerging issues in a manner Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 10 that is science-based, comprehensive and socially acceptable. The update of the core values of the Federation will be expressed in a new mission statement of EGF that is being drafted in consultation with the Scientific Advisory Board. Acknowledgements Thanks are due to Mr. A. Hopkins, Prof. R.J. Wilkins and Prof. J. Nösberger for dealing with the sections on Grass and Forage Science, European Union and Future as well as for their general comments, suggestions and English language revisions. References Powell R.A., Corrall A.J. and Corrall Rosemary G. (1995) A history of the British Grassland Society, 1945–1995. In: Pollott G.E. (ed.) Grassland into the 21st Century. Occasional Symposium No. 29 of the British Grassland Society, pp. 2-30. Reading: British Grassland Society. Prins W.H. (2004) A history of the European Grassland Federation, 1963-2003. Grass and Forage Science 59, 2-7. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 11 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 12 European grasslands overviews invited papers Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 13 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 14 European grasslands overview: Nordic region Helgadóttir Á.1, Frankow-Lindberg B.E.2, Seppänen M.M.3, Søegaard K.4 and Østrem L.5 1 Department of Land and Animal Resources, Agricultural University of Iceland, Árleyni 22, IS-112 Reykjavík, Iceland 2 Department of Crop Production Ecology, Swedish University of Agricultural Sciences, P.O. Box 7043,SE-750 07 Uppsala, Sweden 3 Department of Agricultural Sciences, P.O. Box 27, FI-00014 University of Helsinki, Finland 4 Department of Agroecology, Aarhus University, P.O. Box 50, DK-8830 Tjele, Denmark 5 Bioforsk - Norwegian Institute for Agricultural and Environmental Research, Fureneset, NO-6967 Hellevik i Fjaler, Norway Corresponding author: aslaug@lbhi.is Abstract The Nordic countries stretch over a large geographic area and, hence, conditions for plant growth vary considerably across the region. The western regions along the North Atlantic Ocean enjoy a mild maritime climate whereas continental climate prevails in the eastern part. Temperature variations have implications for the length of the growing season and Accumulated Day Degrees, thus influencing the timing and number of harvests in different areas across the region. Grasslands in the Nordic region are a diverse group and their relative importance differs greatly across countries. Thus, natural grasslands are extensive in Iceland and play an important role for livestock production, whereas they are much less significant in the other countries. Semi-natural grasslands are used for grazing but they are also important for maintaining biodiversity and landscape types. Forage for winter fodder and dairy in summer is obtained from cultivated grasslands of which short-term leys dominate in the southern regions. Milk and beef is the dominant produce, particularly in Denmark, whereas lamb and horse meat is a significant part of the production in Iceland. The expected climate change at northern latitudes will result in a longer growing season and higher temperatures during the growing season, both of which may lead to increased biomass production potential. However, new types of stresses may offset the potential gain. This will have various implications for adapting forage cultivars to the changing conditions both through breeding and different management schemes. Keywords: agricultural production, climate, forage mixtures, local adaptation, soil Geography and the natural environment The Nordic region consists of the Nordic countries, the Faroe Islands, Greenland and Åland. The present paper will primarily focus on grasslands in the five Nordic countries: Denmark, Finland, Iceland, Norway and Sweden. These countries stretch over a large geographic area from 54° 58' N in the south of Denmark to 71° 08' N in Northern Norway, and from 24° 32' W in the west of Iceland to 31° 35' E on the eastern border of Finland with Russia (Figure 1). There are currently around 25 million inhabitants, making the region one of the most sparsely populated parts of the world. Iceland thus has around 3 inhabitants per km-2, Sweden, Norway and Finland each have 16–23 inhabitants km-2 whereas Denmark has 130 inhabitants km-2, a value close to the EU mean. Climatic conditions Conditions for plant growth vary considerably across the region, reflecting its wide geographic spread. The Gulf Stream brings warm sea from the Gulf of Mexico into the North Atlantic Ocean resulting in a milder climate than could otherwise be expected from the geographic location. However, various factors can affect the temperature as indicated for example by the Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 15 10 °C isotherm line in July, which falls below the Arctic Circle in Iceland but stretches well beyond in Northern Scandinavia (Figure 1). The difference in daily mean temperature during the warmest period in summer between the warmest (e.g. Helsinki, FI) and coldest agricultural areas (e.g. Akureyri, IS) is thus around 7 °C and temperature can vary up to 5 °C between coastal areas (Reykjavík, IS) and inland (Kajaani, FI) at the same latitude during summer (Figure 2). Temperature variations have implications for the length of the growing season and Accumulated Day Degrees, thus influencing the timing and number of harvests in different areas across the region. The growing season can be defined as the number of days from when the daily mean temperature exceeds 5 °C in spring and falls below 5 °C in autumn (SNP, 1992). The length of the growing season varies thus from around 230 days in the south west, to around 125 days in the northern-most and high altitude regions where agriculture is at its limit (Table 1). Similarly Accumulated Day Degrees (Tsum >5 °C) are currently only around 640 in Reykjavík (IS) compared to 1732 in Tranebjerg (DK), and 1020 – 1150 at Værnes (NO), Holmögadd (SE) and Kajaani (FI), which are at comparable latitudes to Reykjavík. Figure 1. The Nordic region showing the Arctic Circle (broken line) and the 10 °C isotherm for July (solid line). Letters show geographic locations of sites listed in Table 1. Temperature conditions during autumn and winter vary similarly across the region, both from north to south and from coast to inland at the same latitude (Figure 2). In the more maritime regions winters are often relatively mild with mean temperatures during the coldest month varying from around 2 °C in Sola (NO) and Tranebjerg (DK) to -3.5 °C in Tromsø (NO). Here, winters are characterised by unstable snow cover and frequent freeze-thaw cycles. In the more continental regions to the east of the Norwegian mountain range temperatures during winter are, on the other hand, much lower ranging e.g. from around -5 °C in Helsinki (FI) to -13 °C in Sodakylae (FI) resulting in much more stable winters. Such varying conditions during winter have different implications for winter damages of grasslands across the region. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 16 Soil conditions Most soil types in the Nordic countries are formed on glacial materials from the last ice age (European Commission, 2005). Large areas of Sweden and Finland are covered by a continuous layer of glacial till dominated by podzols but these tend to be weakly developed in the northernmost parts. In Norway, on the other hand, the glacial till is thin and discontinuous, exposing large areas of bare rock, and varying both in texture and nutrient value. Smaller areas of sandy glacio-fluvial deposits can be found in all three countries. In the low-lying coastal areas along the Gulf of Bothnia silty and clayey deposits are common, while marine and river deposits around fjords and in valley bottoms are most important parent materials for the agricultural soils in the middle and southern parts of Norway. Peat covers around 15% of the area in Finland and Sweden, while in Norway peat soils constitute a considerable part of agricultural land in the western and northern part, all of which is used for grassland. Peat soils are acidic, of low nutrient value and have to be drained. Land use in Sweden is dominated by forestry (56% of land) while agriculture covers approximately 7% of the area. The best soils are mainly used for growing cereals and oil crops, while grassland production most often occurs on the better soils in forested areas, or poorer soils in otherwise crop dominated areas. In Denmark Podzol forms a continuous block over nearly the whole western half of Jutland. The sandy Podzols in the far west are unfertile and limit agricultural activity unless heavily fertilized and limed. These regions were almost all covered by heath until the beginning of the 20th century. In the eastern part of the country Haplic Luvisols, the most productive soils, predominate. Dairy farms are primarily placed on loamy-sand or coarse sandy soil where the plant available water is around 90 and 60 mm, respectively. Therefore most of the intensive managed grasslands on sandy soils are irrigated at drought stress. Soils in Iceland originate from parent material of recent volcanic origin, which consists mostly of basaltic tephra, and are classified as Andosols (Arnalds, 2004). Such soils generally contain a range of pore sizes that can retain large amounts of water. They are high in organic C and N, and have a strong tendency to fix phosphorus. All these characteristics provide an excellent environment for root growth (Nanzyo et al., 1993). The role of grasslands in agricultural production and environmental protection Extent and importance of different types of grassland Grasslands in the Nordic region are a diverse group that can be classified in various ways depending on their origin, vegetation type and/or current land use. In this paper we will primarily focus on grasslands that play a role in animal husbandry. We will follow the definition of grassland as a term that ‘bridges pastureland and rangeland and may be either a natural or an imposed ecosystem. The vegetation of grassland in this context is broadly interpreted to include grasses, legumes and other forbs, and at times woody species may be present’ (Allen et al., 2011). In our discussion we will further divide grassland into three major types with respect to origin (Hejcman et al., 2012): (i) natural grasslands dominated by indigenous grasses and other herbaceous species; (ii) semi-natural grasslands created by longterm human intervention and with a wide range of species richness and herbage productivity; and iii) improved grasslands established with domesticated forage species that receive intensive management (fertilization, weed control, renovation). Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 17 Figure 2. Estimates of mean temperature for each day of the year for the period 1986-2005 during summer (T>5 °C) and winter (T<5 °C) for selected sites in the Nordic countries, going from south to north in the maritime West (left) and continental East (centre), and from West to East at the same latitute of around 64°N (right) (for information of geographic location, see Table 1 and Figure 1) (estimates based on an improved method from Björnsson et al., 2007). Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 18 Table 1. Geographic location, length of the growing season (T>5 °C) and Accumulated Day Degrees (° D; Tsum>5 °C) for selected sites in theNordic countries, based on estimates of mean temperature for each day of the year for the period 1986-2005 (Björnsson et al., 2007). Site A B C D E F G H I J K L Latitude Akureyri (IS) Bodø (NO) Helsinki (FI) Holmögadd (SE) Ivalo (FI) Kajaani (FI) Reykjavík (IS) Sodankylae (FI) Sola (NO) Tranebjerg (DK) Tromsø (NO) Værnes (NO) 65.7° N 67.3° N 60.3° N 63.6° N 68.4° N 64.3° N 64.1° N 67.4° N 58.5° N 55.9° N 69.7° N 63.5° N Longitude 18.1° W 14.4° E 25.0° E 28.8° E 27.3° E 27.7° E 21.9° W 26.7° E 05.3° E 10.6° E 18.9° E 12.9° E Growing season Period No. of days 9 May – 1 Oct. 146 2 May – 19 Oct. 171 21 April – 18 Oct. 181 13 May – 20 Oct. 161 21 May – 22 Sept. 125 6 May – 3 Oct. 150 4 May – 9 Oct. 159 17 May – 21 Sept. 128 3 April – 17 Nov. 229 2 April – 21 Nov. 234 16 May – 1 Oct. 139 17 April – 21 Oct. 188 °D 669 900 1379 1028 718 1020 642 785 1391 1732 659 1145 Natural grasslands Virtually all of Icelandic grasslands are either rangeland or pastureland and therefore important for livestock production. There are discrepancies as to how grassland in Iceland has been defined. In connection with the land-use related GHG emission reported to the UN-Framework Convention on Climate Change, the Icelandic land-use category Grassland is defined as all land with > 20% coverage of vascular plants, and not included under categories Forest land, Cropland, Wetland or Settlement, and estimated to be 53,000 km2 or 51% of the country (Gudmundsson et al., 2013). On the other hand, according to results obtained from the Icelandic Farmland Database, which classifies the surface of the country into twelve different classes based on satellite images and field research (Arnalds and Barkarson, 2003), only around 43,000 km2 or 42% of the country is vegetated (Arnalds, 2011). In the present context two of these classes could be truly classified as natural grassland; ‘grassland’ being 2,375 km2 or 2.3% of the country and ‘rich heath’ 6,843 km2 or 6.6% or a total of 9,218 km2 (Table 2). Table 2. The extent of different types of grasslands (in thousand hectares) in the Nordic region. UAA1 Denmark Finland Iceland Norway Sweden 1 2,645 2,285 130 990 3,030 Cultivated grasslands 332 650 120 470 1,100 Semi-natural grasslands 200 40 42 177 440 Natural grasslands 0 0 922 ? 0 Utilizable Agricultural Area These areas are characterized by low-growing vegetation of grasses, sedges, forbs and woody perennials. They are the most valuable part of the rangelands and used for extensive summer grazing of sheep and horses from the end of June to early September when the animals are rounded up. Around 40% of these natural grasslands are found above 200 m and each local community has grazing rights in the common grazing areas of the highlands. Roughly half of the required feed for sheep and a large part of feed for horses is obtained from summer grazing on the natural grassland. The grazing resources of outlying fields in Norway are significant and these are currently being surveyed. Preliminary data for a grazing period of 100 days year-1 show that about 760,000 FEm (milk feeding units) may be utilized (Rekdal, 2013). Natural grassland rarely occurs in Sweden, Finland and Denmark. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 19 Semi-natural grasslands The Icelandic Farmland Database defines ‘cultivated land’ as being 1,723 km2 or 1.7% of the country (Arnalds, 2011). Of this it is estimated that around 42,000 ha are ‘abandoned crop land’ (Gudmundsson et al., 2013). Most if not all of this consists of old hayfields that have been taken out of use, mostly because of structural changes in agricultural production over the last few decades. In the present context these old grass fields can be defined as semi-natural grasslands in Iceland (Table 2). Some of these have been used for hay making over centuries but mostly they originate from the cultivation era that started around 100 years ago (Helgadóttir et al., 2013). Even though most of the fields were originally sown with improved grass cultivars they are now dominated by indigenous species. Such ecotypes may become valuable for future breeding (Rognli et al., 2013a). In Norway grasslands are defined as areas covered with grass that may be mechanically harvested or grazed but never ploughed and that may be cultivated by fertilization, harvested mechanically and improved by selected species (CPA, 2013). Grazing lands are assigned to areas with at least 50% grass cover and annual grazing. Trees, stumps, and rocks may be present but grazing is considered the most important land use form either as surface-cultivated grass leys or unimproved grazing land (NFLI, 2011). Statistics Norway (2014) further defines the two types of grassland management: (1) surface-cultivated pastures have shallow topsoil layers, often with surface rocks, and can be mechanically harvested but are not ploughed, and (2) unimproved grazing land is never mechanically harvested (or ploughed) but only grazed and can be considered semi-natural landscapes. For the second class at least half of the area should be covered by grasses or palatable herbs and enclosed by a fence or a natural barrier. It must also be grazed or harvested at least once a year to be eligible for subsidy support (Kynding Borgen and Hylen, 2013). Such areas will also include active summer farms that still can be found in specific regions. In 2012 semi-natural grasslands amounted to 177,000 ha of which 20,000 ha were surface-cultivated pastures and 156,000 ha were unimproved grazing land (Statistics Norway, 2014) (Table 2). In the past, Sweden had large areas of semi-natural grasslands, culminating at around 1.6 million ha at the end of the 18th century (Statistics Sweden, 2013), that were mainly used to provide winter fodder. Summer grazing took place in the forests. Today, only some 440,000 ha of these semi-natural grasslands remain, almost all of which are now used for grazing (Table 2). They play an extremely important role for the protection of biodiversity in the Swedish landscape, and farmers can receive subsidies if they are managed according to a set of rules aimed at the preservation of biodiversity (no fertilization, specified grazing regimes). The output in terms of animal produce is low but the marketing of beef to a premium price from cattle that have grazed in these areas is increasing. In Denmark semi-natural grasslands cover 200,000 ha and are either not ploughed at all or only rarely. Some of these are on wet organic or sandy soils along rivers, near the coastline or on old hilly grasslands. This is therefore a very heterogeneous group with respect to yield, species composition and use. The aim of the authorities is to preserve these permanent grasslands in order to enhance a landscape of light and open (non-forest) river valleys and to maintain biodiversity. Strict rules apply to the management of these grasslands. Semi-natural grasslands are of minor significance in Finland and most of these can be found in coastal areas where the aim is to keep the landscape open to encourage bird life and natural meadows (Table 2). Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 20 Cultivated grasslands The total area of cultivated land in Iceland in current use is around 130,000 ha (Table 2), of which approximately 45% is on drained wetland (Hallsdóttir et al., 2012). Around 90% of this area is used for permanent grass fields, most of which are older than five years and the remainder is under annual crops of which 5,000 ha are currently used for barley. Grassland or cropland constitutes about 3.0% of total land area in Norway, and in 2012 total agricultural land amounted to 990,000 ha of which cultivated leys constituted 470,000 ha or 73% of the total grassland area. Grass leys in rotation with other crops are classified as cropland so the grassland area is actually higher than shown by Statistics Norway (2014) (Kynding Borgen and Hylen, 2013). In 2008, 60% of the grassland was owned and 40% rented land (STM20112012). In Sweden, short-term leys on arable land dominate the production of forage for winter feeding and grazing. Today they cover approximately 1.1 million ha, or 45% of the arable land area (Statistics Sweden, 2013). Cultivation of forage maize occurs on approximately 20,000 ha and is increasing. In Finland cultivated grasslands cover 650,000 ha or 28% of the utilizable agricultural area. Cultivated grasslands are around 330,000 ha or 13% of the utilizable agricultural area in Denmark and include grass and clover in grass-arable rotation. Other main forage crops are maize and cereal and pea whole-crop, which constitute 184,000 and 54,000 ha, respectively (Statistics Denmark, 2012). Management of cultivated grasslands and their production potential The livestock The number of farm animals varies considerably across countries both in absolute terms and relative to population size (Table 3). Table 3. Number of farm animals (total no. in thousands and per thousand inhabitants) and main agricultural produce from grassland in the Nordic countries in 2012. Country Dairy cows Total ×1000 Per 1000 inhabitants Total ×1000 Horses Per 1000 inhabitants Total ×1000 Goats Per 1000 inhabitants Total ×1000 Per 1000 inhabitants Denmark 568 101 160 29 60 11 ? ? Finland 284 52 130 24 31 6 5 <1 Iceland 25 77 476 1480 77 240 1 3 Norway 233 46 906 144 36 7 35 7 Sweden 348 37 610 64 363 39 5 <1 Country Milk Total, ML Denmark 1 Sheep Lamb meat, t L per inhabitant 49091 877 Finland 2188 Iceland 127 Norway Sweden Total Beef, t Per 1000 inhabitants Total ×1000 Horse meat, t Per 1000 inhabitants 2000 <1 142 25 404 950 <1 81 396 9920 31 4 1531 305 17500 3 2861 300 5030 <1 Total Per 1000 inhabitants 1000 <1 15 487 <1 13 1510 5 78 16 140 <1 121 13 1160 <1 Of this 9% is organically produced The numbers of both sheep and horses per inhabitant are highest in Iceland whereas dairy cows are of greatest importance in Denmark. In Iceland dairy cows are usually housed from early September to late May and are grazed on cultivated grass fields during summer. With the Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 21 advent of automatic milking it is though getting more common to keep milking cows inside for the whole year, with access to grazing. Most sheep farmers house their flocks from November to May. Riding horses are similarly housed from November to May but otherwise the Icelandic horse is, by tradition, kept outdoors all year around. Dairy cows are mainly fed indoors all year in Norway. The -razing period for stabled dairy cows is 8 weeks, whereas in loose-housing dairy cows should, from 2014, have the possibility to 'exercise and move around' outdoors for at least 8 weeks during the summer period. In some regions, where the main calving time is during autumn, dry cows may be grazed during summer. The housing of sheep varies much across the country depending on the duration of winter. Sheep will normally be kept indoors for one to three weeks after lambing (mid-April–mid-May), followed by some weeks of onfarm grazing before being taken to the areas of summer grazing. The sheep return to the farm in early/mid-September. The Old Norse sheep breed (Ovis aries) (3% of total sheep number) is being grazed all-year round, however, with strict regulations for animal welfare to secure feeding possibilities during harsh winters. Dairy cows are the main consumers of forage in Sweden and Finland. Grazing is compulsory by law for dairy cows and provides forage from May/June until the end of August/September, depending on latitude. However, supplementary feeding of dairy cows during summer is very common. In Denmark conventional dairy cows (+heifers and calves) are mostly kept in the stable the whole year, whereas the organic dairy cows are grazed at least 150 days for a minimum of 6 hours a day, in accordance with the rules. Harvest management Most of the fodder produced for winter feeding is ensiled either with or without additives in round bales or bunker silos in all countries. Dairy farmers in Iceland aim to take the first cut around the heading of timothy, which may occur as early as the middle of June in the more favourable regions. A second cut is usually taken on these fields around the middle of August. On sheep farms the first harvest is commonly delayed until early July and the fields are used for grazing in early spring and again in autumn for fattening the lambs prior to slaughter, rather than taking a second cut. Where two cuts are taken from hay fields about two-thirds of the fertilizer is applied in spring and one-third after first cut. Otherwise all fertilizer is applied in spring. Nitrogen inputs vary around 100 – 140 kg ha-1. In addition most farmers use animal manure to its fullest potential. In Norway fields are cut three times in the more favourable regions, in early June, late July and mid-September, whereas only two cuts are taken in the middle of June and again in the middle of August where the growing season is shorter. In Sweden and Finland winter feed is produced on intensively managed short-term leys with inputs of 150 – 250 kg N ha-1 and two to four cuts per season, depending on latitude. The aim is to cut the first harvest at the beginning of heading, or at >80% IVOMD (in vitro organic matter digestibility). In addition most farmers use animal manure to its fullest potential. This is also the case in Denmark. There dairy farmers aim to take the first cut at a certain IVOMD or NEL (net energy lactation) concentration, and they decide the cutting time from a herbage sample and an on-line prognosis which calculates the decrease in feeding value due to growth and the effect of weather on the decrease rate. The targeted nutritive value depends on the feeding ration on the farm, but is typically 80% IVOMD. The recommended cuts are five and four for red and white clover mixtures respectively. For beef cattle, sheep and goats there is less focus on nutritive value. The maximum N-application in intensive managed grasslands is around 240 and 360kg ha-1 in grass-clover and pure grass, respectively. In permanent grasslands, N varies from 0 – 130 kg ha-1, depending of production level, use and nature value. Most of the applied N comes from slurry, all of which must be injected. The only other possibility is acidification and using trailing hoses. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 22 Production capacity According to the Icelandic Agricultural Statistics (Statistics Iceland, 2013) the mean fodder yield obtained in 2011 was 4,012 and 2,720 FEm ha-1 for dairy and sheep farms, respectively (Statistics Iceland, 2013). The total fodder produced on-farm in the whole country was worth a total of around 54 million Euros in 2011, which was around 55% of the total fodder costs for the whole agricultural sector that year (Statistics Iceland, 2013). Most of this was roughage obtained from hayfields. Barley produced locally has been sufficient to supply around 40% of the concentrate use in the dairy sector. In Norway the mean fodder yield is estimated to be around 4,000 and 3,500 FEm ha-1 on dairy farms and sheep farms, respectively (Bakken and Johansen, 2014; A.K. Bakken, personal communication). There is no exact on-farm dry matter yield (DMY) registration, but estimated DMY and the amount of milk and meat being produced indicates a reduction in grass yield over the last few years. Important reasons are poor soil cultivation, need for drainage and use of heavier agricultural machinery. Increased conservation in round bales also may have caused increased loss due to wilting (Lunnan, 2012). Organic grassland (ley and grazed area) constituted <10 % of total area in 2012 (DEBIO, 2012). Imported feed constitutes 32% of the total feed and especially the protein fraction is increasing (SLF). If imported soya used for feed is included the imported feed is 38%. A higher degree of food self-sufficiency in Norway will require considerable changes of the diet from animal-based food to fish and plants grown in Norway (Bakken and Johansen, 2014). In Sweden the intensively managed short-term leys yield 7 – 12 tons dry matter ha-1, with a digestibility of >80% and a crude protein of around 15%. In Finland the yield is 5 – 10 tons dry matter ha-1. This fodder is primarily used for dairy and beef cattle. Forage for horses and suckler cows is less intensively cultivated, and is generally harvested at a later phenological stage. In Denmark the potential production is around 12,000 FE (90,000 MJ net energy), but due to different losses in the feed processing the net energy yield (energy consumed by the cows) was 8,410 and 2,560 FEm ha-1 on intensive grasslands (in crop rotation) and permanent grasslands, respectively (Statistics Denmark, 2012). For the whole country, home-grown fodder accounted for 79% and imported for 21% of the total requirements. For forage crops 97% was produced locally and 3% was imported. The main agricultural produce obtained from this fodder in the different countries can be seen in Table 3. Procurement of suitable plant material Breeding for local adaptation In a European context the main limiting factors for forage crops in the Nordic environment are a short and cool growing season and various winter stresses such as frost, ice encasement, low temperature fungi, prolonged snow cover, water-logging and low light intensity (Figure 2). Temperature and photoperiod are probably the two most important factors that control phenological development and play a decisive role for the adaptation of perennial fodder crops. This is of special importance for growth cessation of perennial crops in the autumn and obviously complicates the transfer of forage crops within the region. The photoperiod is directly related to latitude and can thus vary within the Nordic region from 14 to 24 hours and from 9 to 12 hours at the beginning and the end of the growing season, respectively. Temperature, on the other hand, can vary considerably at the same latitude, as discussed earlier, making transfer of material even more complicated. One way of dealing with such genotypeenvironment interactions is to calculate agroclimatic indices. These are commonly based on Accumulated Day Degrees, which are related to plant growth and development, and have been Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 23 applied successfully, together with indicies reflecting the risk of winter damage, for assessing overwintering of perennial crops in Eastern-Canada (Bélanger et al., 2002) and Norway (Thorsen and Höglind, 2010). When dealing with the whole Nordic region it is sensible to define agroclimatic areas that differ in climatic conditions for plant production (SNP, 1992). One attempt has been made to construct such agroclimatic zones based on results from variety trials with timothy across the whole Nordic region (Björnsson, 1993). The outcome was five internordic zones based on similarity of results and geographic considerations. Interestingly though, in a recent study of molecular variation (SSR markers) of local Nordic timothy populations and cultivars, only 6% of the variation was between populations, and no variation was found between cultivars and local populations (Rognli et al., 2013a). The results for timothy are in stark contrast to meadow fescue where local Nordic populations show clear geographic structuring based on molecular variation (Fjellheim and Rognli, 2005; Fjellheim et al., 2009). Similarly, Finne et al. (2000) found highly significant genotypic variation between local populations of white clover collected from a wide range of latitudes and altitudes in Norway. Breeding of grassland species for Denmark and different areas of Norway, Sweden and Finland has been carried out for over a 100 years. In the early days, natural selection in local material was the main method to extend the area in which a population performed well. Natural selection was likewise used to increase the resistance to various pests and diseases. Poly-crossing and progeny testing of individual plants selected for good performance followed by mass selection is nowadays the most common method. The original plant material may then be of indigenous or foreign origin, depending on the species in question. Icelandic agriculture has to a large extent depended on the import of forage cultivars. Not unexpectedly, the most suitable material has originated from the northern areas of Norway and Sweden. Northerly adapted plant material is generally characterized by low yields, slow regrowth potential after first cut and reasonable tolerance to frost and ice encasement. Timothy (Phleum pratense) is the only forage species for which cultivars have been bred specifically for Icelandic conditions. Selection of appropriate species Not surprisingly the choice of appropriate species varies depending on the geographic location. Timothy has been the major grass species in the Nordic region north of 60° N. Timothy is thus by far the most important fodder species in Iceland. It has a clear quality advantage over other grass species that are currently available for forage production in Iceland, both with respect to mean dry matter digestibility (DMD) and daily voluntary DM intake (see Helgadóttir et al., 2013). Similarly in Norway and Sweden and Finland, timothy is the only grass species that can be grown successfully almost anywhere because of its superior persistence (Larsen and Marum, 2006; Østrem et al., 2013). Where it is the main species in seed mixtures, the harvesting dates are mostly determined by the developmental stage of timothy. An increased regrowth capacity will, however, be required in a prolonged growing season. Meadow fescue (Festuca pratensis) is a well-adapted grass species and is frequently used in timothy-based mixtures in these countries, and in mixtures for combined cutting and grazing, smooth meadow-grass (Poa pratensis) is a winter-hardy species. Perennial ryegrass (Lolium perenne) has played a role in the more maritime regions south of 60° N. It has for many years been the most important grass species in Denmark because of high nutritive value (Søegaard et al., 2010) and it is also grown successfully in the southern parts of Sweden but here always in mixtures with other grasses as it is still too unreliable to be grown in monoculture. With its high biomass yield and regrowth capacity, and superior feed quality perennial ryegrass will undoubtedly become a promising option at higher latitudes with prolonged growing season and milder winters. The same is true for ×Festulolium hybrids (Østrem and Larsen, 2010). Festulolium holds a vast genetic variation (Ghesquière et al., 2010) and several Festulolium hybrids have been introduced in the last 10 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 24 years to Sweden, Denmark and Norway. Compared to perennial ryegrass it has an earlier spring growth and a lower nutritive value at the same date. Farmers therefore harvest mixtures with Festulolium in spring about one week earlier than mixtures with perennial ryegrass only. Targeted breeding goal for Festulolium is to understand the mechanisms behind the growth cessation to secure cold hardening and winter survival. For high latitudes the allotetraploid approach, including the full genome of both parental species, seems to be the most useful approach for exploiting the genetic variation, among others for winter survival, in the combination of L. perenne and F. pratensis (Østrem et al., 2013). The more deep-rooted tall fescue (Festuca arundinacea) might become a more important species also at high latitudes, due to possible drought incidences in spring/early summer in a changing climate (IPCC, 2013). When tall fescue is one of the parents, cultivars of Festulolium have proved a winter hardy alternative in the Nordic-Baltic region (Gutmane and Adamovics, 2008; Halling, 2012; Østrem et al., 2013). Red clover (Trifolium pratense) and white clover (T. repens) are the most important legume species in the Nordic countries (Marum, 2010). In Sweden red clover dominates followed by white clover and lucerne, which is an important species in the drier eastern parts of the country. In Finland red clover is the most important forage legume. In Denmark white clover was successfully reintroduced in the 1990s after the high-N period during 1960-1990. Red clover was reintroduced a few years later. The main reason for using clover in Denmark, Finland and Sweden is the well-known higher intake of clover mixtures, demand for reduced N-application and a general perception of better quality. Today the goal for nearly all intensively managed grasslands in Denmark is to have a high content of clover, which means at least 50% in the summer period. In organic farming clover has a significant impact on the whole crop rotation system. However, white clover fatigue has been a big challenge due to a high proportion of clover in the grazing area for the cows close to the stables (Søegaard and Møller, 2005). The importance of forage mixtures Species mixtures are the norm in all countries. Mixtures provide yield stability over time, as well as an extended period over which a good feeding quality can be maintained. A typical mixture for large areas where grassland farming is important would contain timothy, meadow fescue, perennial ryegrass, red and white clover in different proportions depending on location, soil condition, planned sward duration and the use of the herbage. In Sweden all-grass mixtures (timothy, meadow fescue and perennial ryegrass) are grown for the production of forage for horses. Recent pan-European experiments have shown that grass-legume mixtures are more productive and show greater yield stability over time than their individual components in monoculture irrespective of fertilizer treatment (Finn et al., 2013) and in the northern regions higher yields were not at the cost of poorer nutritive value (Sturludóttir et al., 2014). Future challenges Climate change Over the next 100 years the global temperature is expected to increase in the range 0.3–4.8 °C in the world, depending on climatic models, with the Arctic region warming more rapidly than the global mean (IPCC, 2013). In the Nordic countries this temperature increase is likely to be larger in northern areas than in the south, and most probably temperature will increase more during winter than summer with more of the precipitation falling as rain (Björnsson et al., 2008; Hanssen-Bauer et al., 2009). This will have various implications for adapting forage cultivars to the changing conditions both through breeding and different management schemes. The expected changes in climate at northern latitudes will result in a longer growing season, because of earlier spring and later autumn, and higher temperatures during the growing season, both of which may lead to increased biomass production potential. However, new types of Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 25 stresses may offset the potential gain, such as insufficient hardening conditions during autumn to prepare the plants for altered winter conditions. Higher temperature in the autumn at the same light intensity will mean that the forage species are cold hardened at later stages during autumn, calling for new breeding strategies. Also during the winter deacclimation and reacclimation are important challenges due to experienced and expected temperature increase during winter (January-April) which will pose stress on plants if warm spells occur during midwinter. In adapting forage plants to new conditions in the north, new variation in exotic material has to be looked for and the most promising material subsequently introgressed into present cultivars. As grassland agriculture in the region is a low value enterprise it would be desirable to breed material with as wide adaptation as possible so material could be used across larger areas within specific agroclimatic zones rather than focus on narrow adaptation to certain geographic regions. Such an approach would require considerable pre-breeding efforts in line with the already initiated Nordic Public Private Partnership for pre-breeding in perennial ryegrass (Rognli et al., 2013b). Climate change may also increase disease pressure in forage plants. This has already occurred for leaf spots in meadow fescue, whereas e.g. crown rust is expected to become a problem on perennial ryegrass. Breeding material of vulnerable species should therefore be tested under more southern growing conditions. Improved land- and nutrient-use efficiency The Nordic countries must rise to the pressing challenge of sustainable intensification of their agricultural production (The Royal Society, 2009). For grassland-based agriculture this may involve improving the quality of the herbage either through breeding or optimizing D-values through more precise management strategies, thus obtaining more feeding units per land unit, as higher temperatures during summer will, without changing the management, reduce quality because of increased lignification of cell walls and lower digestibility of organic matter (Buxton, 1996). One approach would also be to increase the use of perennial ryegrass as discussed earlier and benefit from its superior feed quality and productivity. It is also important to reduce the reliance on artificial fertilizers. An obvious way is to make better use of N-fixing species and it is therefore especially important to improve their adaptation in the more marginal regions. Forage mixtures have also been shown to enhance yields compared to their individual components in pure stand (Finn et al., 2013) but the mixtures have to be carefully designed in order to realize their full potential in order to improve resource complementarity and increase yield. References Allen V.G., Batello C., Berretta E.J., Hodgson J., Kothmann M., Li X., Mcivor J., Milne J., Morris C., Peeters A. and Sanderson M. (2011) An international terminology for grazing lands and grazing animals. Grass and Forage Science 66, 2-28. Arnalds Ó. (2004) Volcanic soils of Iceland. Catena 56, 3-20. Arnalds, Ó. 2011. Náttúrufar [Physiography]. In: Aradóttir Á.L. and Halldórsson G. (eds) Vistheimt á Íslandi [Ecological restoration in Iceland], Agricultural University of Iceland and Soil Conservation Service of Iceland, Reykjavík, Iceland, pp. 14-18. [In Icelandic]. Arnalds O. and Barkarson B.H. (2003) Soil erosion and land use policy in Iceland in relation to sheep grazing and government subsidies. Environmental Science and Policy 6, 105-113. Bakken A.K. and Johansen A. (2014) Matproduksjon på norske arealressursar med og utan import av fôrråvarer [Food production on Norwegian land with and without the import of raw feed stuffs]. Bioforsk FOKUS 9(2), 82. Bélanger G., Rochette P., Gastonguay Y., Bootsma A., Mongrain D. and Ryan D. (2002) Climate change and winter survival of perennial forage crops in Eastern Canada. Agronomy Journal 94, 1120-1130. Björnsson H. (1993) Zones for performance testing of timothy (Phleum pratense L.) in the Nordic countries. Acta Agriculturae Scandinavica, Section B, Soil and Plant Science 43, 97-113. Bjornsson H., Olason E.O., Jonsson T. and Henkirksen, S. (2007) Analysis of a smooth seasonal cycle with daily resolution and degree day maps for Iceland. Meteorologische Zeitschrift 16, 57-69. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 26 Björnsson H., Sveinbjörnsdóttir Á.E., Daníelsdóttir A.K., Snorrason Á., Sigurðsson B.D., Sveinbjörnsson E., Viggósson G., Sigurjónsson J., Baldursson S., Þorvaldsdóttir S. and Jónsson T. (2008) Hnattrænar loftslagsbreytingar og áhrif þeirra á Íslandi – Skýrsla vísindanefndar um loftslagsbreytingar [Global climate change and their effects in Iceland – Report of the scientific committee on climate change], Ministry for the Environment and Natural Resources, Reykjavík, Iceland, 118 pp. [In Icelandic]. Buxton, D.R. (1996) Quality–related characteristics of forages as influenced by environment and agronomic factors. Animal Feed Science Technology 59, 37-49. CPA (2013) National Inventory Report 2013 Norway. Climate and Pollution Agency, Oslo. Danmarks Statistik (2012) Statistisk Årbog 2012, 524 pp. DEBIO (2010). Statistikkhefte [Statistics Booklet]. http://www.debio.no European Commission (2005) Soil Atlas of Europe, European Soil Bureau Network. Office for Official Publication of the European Communities, Luxembourg, 128 pp. Finn J. et al. (2013) Ecosystem function enhanced by combining four functional types of plant species in intensively-managed grassland mixtures: a three-year continental-scale field experiment. Journal of Applied Ecology 50, 365-375. Finne M.A., Rognli O.A. and Schjelderup I. (2000) Genetic variation in a Norwegian germplasm collection of white clover (Trifolium repens L.). 1. Population differences in agronomic characteristics. Euphytica 112, 33-44. Fjellheim S., Pasakinskiene I., Grønnerød S., Paplauskiene V. and Rognli O.A. (2009) Genetic structure of local populations and cultivars of meadow fescue from the Nordic and Baltic regions. Crop Science 49, 200-210. Fjellheim S. and Rognli O.A. (2005) Genetic diversity within and among Nordic meadow fescue (Festuca pratensis Huds.) cultivars determined on the basis of AFLP markers. Crop Science 45, 2081-2086. Ghesquière M., Humphreys M.W. and Zwierzykovski Z. (2010) Festulolium. In: Boller B., Posselt U.K. and Veronesi F. (eds) Fodder Crops and Amenity Grasses, Handbook of Plant Breeding 5. Springer Science+Business Media, LLC 2010, pp 293-316. Gudmundsson J., Brink S.H. and Gísladóttir F. Ó. (2013) Preparation of a LULUCF land-use map for Iceland: Development of the Grassland layer and subcategories. Grassland Science in Europe 18, 102-104. Gutmane I., and Adamovics A. (2008) Analysis of Festulolium and hybrid ryegrass (Lolium x boucheanum) dry matter stability. Grassland Science in Europe, 13, 248-250. Halling M. (2012) Yield stability of Festulolium and perennial ryegrass in southern and central Sweden. Grassland Science in Europe 17, 118-120. Hallsdóttir B.S., Wöll C., Guðmundsson J., Snorrason A. and Thórsson J. (2012) Emissions of greenhouse gases in Iceland from 1990 to 2010. National Inventory Report 2012. Submitted under the United Nations Framework Convention on Climate Change and the Kyoto Protocol, Environmental Agency of Iceland, UST: 330, 293 pp. Hanssen-Bauer I., Drange H., Førland E.J., Roald L.A., Børsheim K.Y., Hisdal H., Lawrence D., Nesje A., Sandven S., Sorteberg A., Sundby S., Vasskog K. and Ådlandsvik B. (2009) Klima I Norge 2100. Bakgrunnsmateriale til NOU Klimatilplassing [Climate in Norway 2100. Background material to NOU Climate Adaptation], Norsk klimasenter, September 2009, Oslo, 148 pp. [In Norwegian]. Available online at: http://www.regjeringen.no/upload/MD/Kampanje/klimatilpasning/Bilder/NOU/Klima-Norge-2100 lavoppl 2opplag okt2009.pdf. Helgadóttir Á., Eythórsdóttir E. and Jóhannesson T. (2013) Agriculture in Iceland. Grassland Science in Europe 18, 30-43. Hejcman M., Hejcmanová P., Pavlů V. and Beneš J. (2012) Origin, history, management and plant species composition of grasslands in Central Europe – a review. Grassland Science in Europe 17 554-567. Kynding Borgen S. and Hylen G. (2013) Emissions and methodologies for cropland and grassland used in the Norwegian national greenhouse gas inventory. Report from the Climate Center Norwegian Forest and Landscape Institute, 38 pp. Larsen A. and Marum P. (2006) Breeding goals and possibilities in future timothy breeding. In: Sveinsson, T. (ed) Timothy productivity and forage quality – possibilities and limitations. Proceedings from NJF Seminar 384, Akureyri, Iceland, 10–12 August 2006, AUI Report No. 10, pp. 31-39. ISSN 1670-5785. Lunnan T. (2012) Avlingsutvikling i engdyrkinga [Yield developements in cultivated grasslands]. Bioforsk FOKUS 7(2) 141-142. Marum P. (2010) Legume breeding and adaptation to changing environments. In: Rasmussen J., Schacht M. and Helgadóttir Á. (eds) The potential of forage legumes to sustain a high agricultural productivity – a Nordic perspective. Proceedings from NJF Seminar 432, Hvanneyri, Iceland, 20–22 June 2010, NJF Report, Vol. 6, No. 3, pp. 13-16. http://www.njf.nu/filebank/files/20100803$084935$fil$93yaFekFtVspuHW8KAcG.pdf Nanzyo M., Dahlgren R. and Shoji S. (1993) Chemical characteristics of volcanic ash soils. In: Shoji S., Nanzyo M. and Dahlgren R. (eds) Volcanic Ash Soils: Genesis, Properties and Utilization, Elsevier, Amsterdam, The Netherlands, pp. 145-187. NFLI (2011) Landskogstakseringens Feltinstruks 2011 [National forest inventory field guide]. Norwegian Forest and Landscape Institute. [In Norwegian]. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 27 IPCC (2013) Summary for Policymakers. In: Stocker, T.F., D. Qin, G.-K. Plattner, M. Tignor, S. K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley (eds.) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA. Olsen O.B. (ed) (1993) Forskning for landbruk og samfunn i forandring [Agricultural Research for Society in Transition]. Statens forskingsstasjoner i landbruk, 64 pp. ISBN 82-991161-4-7. [In Norwegian]. Rekdal Y (2013) Beite i utmark – kvalitet og kapasitet [Grazing in extensive grasslands – quality and quantity]. Bioforsk FOKUS 8(2), 135-137. Rognli O.A., Fjellheim S., Pecetti L. and Boller B. (2013a) Semi-natural grasslands as a source of genetic diversity. Grassland Science in Europe 18, 303-313. Rognli O.A., Larsen A.S., Nielsen N.C., Helgadóttir Á., Aavola A., Persson C., Isolahti M., Asp T. and Marum P. (2013) The Nordic Public-Private Partnership (PPP) for pre-breeding in perennial ryegrass (Lolium perenne L.). Book of Abstracts of the 17th Symposium of the European Grassland Federation, Akrueyri, Iceland, 23-26 June 2013, p. 84. Søegaard K., Jensen S.K. and Sehested J. (2010) Vitaminer, mineraler og foderværdi af græsmarksarter [Vitamins, minerals and feed value of forage species]. Intern Rapport Husdyrbrug nr. 27, 15-20. [In Danish]. Søegaard K. and Møller K. (2005) White clover soil fatigue: an establishment problem on large and intensive farms. In: Murphy J.J. (ed) Utilisation of grazed grass in temperate animal systems. Proceedings of satellite workshop of the XXth International Grassland Congress, Cork, Ireland. Wageningen: Wageningen Academic Publishers ( ISBN 9076998760), p. 231. SNP (1992). Agroklimatisk kartlegging av Norden [Agroclimatic mapping of the Nordic countries], Samnordisk planteforedling, Skrifter och rapporter nr 5. [In Norwegian]. Statistics Denmark (2012) http://www.dst.dk/pukora/epub/Nyt/2013/NR397.pdf Statistics Iceland (2013). (http://www.statice.is) and (http://hag.is/pdf/2011/kaflar2011.pdf). Retrieved from www.statice.is on 30 December 2013. Statistics Norway (2014) (http://www.ssb.no/) Retrieved from www.ssb.no on 10 February 2014. Statistics Sweden (2013) https://www.jordbruksverket.se/download/18.765a35dc13f7d0bf7c4639/1372238974961/Kapitel+1+Historisk+j ordbruksstatistik.pdf and https://www.jordbruksverket.se/download/18.765a35dc13f7d0bf7c463b/1372238975402/Kapitel+3+%C3%85k erarealens+anv%C3%A4ndning.pdf Sturludottir E., Brophy C., Bélanger G., Gustavsson A.-M., Jørgensen M., Lunnan T. and Helgadottir A. (2014) Benefits of mixing grasses and legumes for herbage yield and nutritive value in Northern Europe and Canada. Grass and Forage Science, 69, 229-240. The Royal Society (2009) Reaping the beneﬁts: science and the sustainable intensiﬁcation of global agriculture, The Royal Society, London, United Kingdom, 73 pp. Thorsen S.M. and Höglind M. (2010) Assessing winter survival of forage grasses in Norway under future climate scenarios by simulating potential frost tolerance in combination with simple agroclimatic indices. Agricultural and Forest Meteorology 150, 1272-1282. Østrem L. and Larsen A. (2010) Fiber content and plant development in Festulolium. In: Huyghe C. (ed) Sustainable Use of Genetic Diversity in Forage and Turf Breeding. Springer, Netherlands, pp. 563-568. Østrem L., Volden B. and Larsen A. (2013). Morphology, dry matter yield and phenological characters at different maturity stages of ×Festulolium compared with other grass species. Acta Agriculturae Scandinavica, Section B, Soil and Plant Science 63, 531-542. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 28 European grasslands overview: temperate region Huyghe C.1, De Vliegher A.2 and Goliński P.3 1 INRA, Centre de Recherche Poitou-Charentes, 86600 Lusignan, France 2 Institute for Agricultural and Fisheries Institute Burg. Van Gansberghelaan 109 9820 Merelbeke, Belgium 3 Department of Grassland and Natural Landscape Sciences, Poznan University of Life Sciences, Dojazd 11, 60-632 Poznan, Poland Corresponding author: christian.huyghe@lusignan.inra.fr Abstract The temperate region, from Ireland to Poland through France, Benelux and Germany is characterized by an important contribution of permanent and temporary grasslands to the Utilized Agricultural Area, with a strong decreasing gradient from West to East of the region, where they provide a range of environmental services, especially in terms of preservation of plant biodiversity. There are contrasting acreages of grasslands under organic farming. The acreage of permanent grasslands has decreased, on average, over the last decades. The acreage of pure forage legume swards has also decreased in relation to the changes in annual cropping systems. The potential of biomass production from grasslands shows a strong West-to-East pattern. The whole forage and grassland systems are relevant to the changes in potential of animal performance, especially visible in dairy cows, that is showing a steady increase in genetic potential, a concentration around a dominating breed, and strong difference among countries in terms of monthly milk deliveries. There is a range of animal PDO products that ensure an economic valorisation of grassland-based production systems. Countries of the temperate regions also developed a very important plant breeding activity dedicated to grass and legume forage species. The present paper documents all these aspects for the various countries. Introduction Grasslands are a major component of the landscapes in temperate regions of Europe, as they are related to a very important economic activity of animal production. In the present paper, the temperate regions include oceanic, sub-oceanic and sub-continental climatic zones. We will analyse Ireland, the UK, France, the Benelux, Germany, Czech Republic, Slovakia and Poland. Because of the large variation in climate and soil conditions, and in history, both permanent and temporary grasslands are used, the latter including a varying proportion of forage legumes. In the temperate regions there is a very limited proportion of common lands, which contribute little to feed production, even though they can play a key role for the preservation of the environment in some areas. In this paper, we focus on the present acreage of permanent and temporary grasslands, the variation over the last decades and their production. We also describe the animals that are using this feed resource and document their number and their productivity and investigate how this influences the grassland acreage and its management. Temporary grasslands depend upon the seed sector and the potential genetic improvement of the varieties available for farmers. Alongside these supporting and provisioning services, we assess how the grasslands contribute to environmental services in the temperate regions. The data presented in this paper were collected during the Multisward project (www.multisward.eu) (Huyghe et al., 2014). Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 29 1. Grassland acreage and production in temperate European regions and changes over recent years Grasslands acreage and changes over decades Permanent and temporary grasslands contribute contrasting proportions of the utilized agricultural area (UAA) in the various countries of the temperate region, as defined in the present study (Figure 1). Overall, permanent grasslands account for 75% to 20% of the UAA, in Ireland and Poland respectively. The proportion of temporary grasslands is lower, as it ranges from 15% in Ireland to 2% in Poland. On average, this means that permanent and temporary grasslands are a key component of the agricultural landscape in the temperate region. There is a very strong gradient from the most oceanic countries where grasslands, and especially permanent grasslands, are very abundant, to the most continental countries. This is related to the potential of biomass production (see section below). 80 70 60 50 Permanent grasslands 40 Temporary Grasslands 30 20 10 0 IE UK LU NL BE FR SK DE CZ PL Figure 1. Share of permanent and temporary grasslands in 2009, expressed as a percentage of Utilized Agricultural Area in the countries of the temperate region considered. (Source: Eurostat, 2009). Organic farming is rapidly expanding in Europe and herbivore production based upon grasslands, both permanent and temporary, is the main production in organic farming. Indeed, this production is easier without mineral fertilizer and pesticides and makes it possible to combine with, and provide fertility to, annual crops. However, there are strong differences among countries, both expressed as a percentage of the grasslands and in absolute values (Figure 2). Acreage of permanent grasslands under organic management is large in Germany, UK, Czech Republic and France, while most organic forage crops (temporary grasslands and annual forage crops) are the most abundant in France, Poland and Germany. Since the 1970s, the proportion of permanent grasslands in the utilized agricultural area has slightly declined in France, Germany and Netherlands. But it has remained very stable in Poland and UK, and even increased slightly in Ireland (Figure 3). This is due to the role they play in the economic activity as a cheap feed source for herbivore production and to their ability to valorise poor soils or fragile environments. In countries where it has declined, however, the losses of permanent grasslands, expressed in hectares may reach large areas. For instance in France, the limited variation in %UAA, is in reality 3 Mha of grasslands lost between 1970 and 2007. Half of this loss is from abandonment and the other part is because of conversion to ploughed land and also as a result of urbanization (Huyghe et al., 2005). Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 30 700 600 (x 1000 ha) 500 400 Permanent grasslands 300 Forage crops on arable land 200 100 0 BE CZ DE IE FR LU NL PL SK UK Figure 2. Organic farming area under permanent pastures and green fodder in 2011 in the countries of the temperate region considered. (Source: Eurostat, 2011). The stability reported for the permanent grasslands is also true for temporary grasslands in all countries. The only exception is the acreage of lucerne, when grown as pure stands, which severely declined since it was a significant crops in early 1970s. 90 80 70 % UAA 60 50 40 30 20 10 0 1961 France 1965 1970 Germany 1975 1980 Ireland 1985 1990 Netherlands 1995 Poland 2000 2005 2007 United Kingdom Figure 3. Changes in the proportion of the permanent grassland area in the UAA (%) in some countries of the considered temperate region between 1961 and 2007. (Source: FAOSTAT and authors’ own calculations). Even if the proportion of permanent and temporary grasslands appears stable at a country level, this may mask significant changes at the local level. This is especially important for temporary grasslands, as they are linked both to the changes in animal production and the changes in the other crops of the rotations. One of the most striking examples is the spatial distribution of lucerne in the Seine Basin in France (Mignolet et al., 2012) (Figure 4). In this region, in 1970, lucerne was used both in mixed farms for feeding herbivores and in specialized cereals farms Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 31 where lucerne was used for dehydration. In 2010, the mixed farming mainly disappeared in the Seine Basin and lucerne is now only located where the dehydration factories are located. This clearly means that the contribution of lucerne, a perennial legume, is restricted to the sustainability of the cropping systems dominated by cereals. Moreover, this was accompanied by a strong reduction of animal production and a quick simplification of the rotations, down to 2 or 3 crops. Another example is the transfer from grassland to green maize in Brittany, Belgium, The Netherlands and Germany. In Belgium the grassland area decreased by 215 000 ha in the period 1970-2010, while the forage maize area increased from 18 000 ha to 176 000 ha in the same period (Anonymous, 2014). Figure 4. Lucerne production in the Seine Basin from 1970 to 2010. (Source: Schott et al., 2009; Catherine Mignolet, pers. comm.). Biomass production and utilization of feed The potential of aerial biomass from grasslands predominated by grasses was calculated for Europe, under the hypothesis of a good sward structure, a good nitrogen fertilization and a good water supply. These calculations were performed by Alain Peeters (Figure 5). The results differ slightly from the map drawn by Smit et al. (2008) by taking into account the soil quality. The striking feature is that the temperate region is the region with the highest potential for biomass production from grasslands, with, on average, good soil quality and adequate climate conditions. There is a slight West-to-East gradient, with higher potentials in the West due to the potential for longer growing seasons because of the oceanic climate. This can also be related to the range of species that may be used for sowing temporary grasslands or being the most productive in the permanent grasslands. Indeed, in the Western part of the temperate region, perennial ryegrass is the predominant grass species, while in the eastern part of the considered zone, meadow fescue, smooth-stalked meadow grass and timothy will be used, and cocksfoot and tall fescue are used in France in areas that have water stress in summer. Similarly, different legume species are used in mixtures with grasses. White clover, Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 32 suitable for grazing, is the most abundant in the Western countries, while red clover, not generally suitable for grazing, is often used in mixtures in the Eastern countries of the temperate region. A first consequence of this pattern is the mean livestock density (Figure 6). It is highest in the Benelux countries and lowest in Poland, Czech Republic and Slovakia. The lower values in these countries may also be due to the lowest levels of fertilisation on some farms, and its relevance to the lowest levels of soil fertility. The direct consequence of this pattern may be found in the management of grasslands and especially their use by the animals. Where the production potential is high, with a long growing season, then grazing is the favourite practice as it reduces the need to make stocks of conserved feeds and, as a consequence, production costs are reduced. In contrast, where growth is limited by cold winter or summer drought, the duration of the growing season is shorter and conserved feed (silage, hay, haylage) is produced and fed either in barns or as a supplementary feed directly in the paddocks, especially to suckler cows. Figure 5. Production potential (annual yields in t DM/ha) of mown and heavily fertilized grasslands. (Source : A. Peeters, own calculations). Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 33 3.5 3 2.5 2 LU/ha total fodder area 1.5 LU/ha grassland 1 0.5 0 Figure 6. Grazing livestock density in the countries of the temperate region in 2007. (Source: Eurostat and authors’ own calculations). 2. Animal production The first ecosystem service provided by grasslands is the provisioning service, as grasslands ensure the production of feed for ruminants. This diet, rich in cellulose and protein, may cover a large part, or all of the nutrient requirements of cattle and sheep. In temperate regions, cattle represent the major part of the livestock units, although there are also large numbers of sheep in Ireland and especially in the UK, where they contribute 10 and 30% of the total grazing livestock respectively. The numbers of dairy cows and beef-suckler cows are presented in Figure 7. With more than 3.5 million dairy cows, Germany and France have the largest numbers of dairy cows in Europe. The numbers of dairy cows in most countries have declined steadily since the establishment of milk quotas in 1984, and as a consequence of the high increase in milk yield per cow (Figure 9). The peculiarity of the temperate region is the high number of suckler cows (named as ‘other cows’ according to Eurostat). The number of suckler cows equals the number of dairy cows in Belgium, Ireland and UK and even exceeds it in France. All these countries have specialized meat breeds: Belgian Blue, Aberdeen Angus, Hereford, Limousin, Charolais, Blonde d’Aquitaine etc. The predominant dairy breed is the Holstein Friesian, while other dairy breeds as Normande, Montbeliarde and Braunvieh are also quite popular and have experienced increasing genetic merits that are close to those met in Holstein Friesian. As a consequence of the growing season of grasslands and the management of the herds, the monthly deliveries of milk vary greatly among countries. Figure 8 illustrates the most contrasting patterns among the temperate region. For Ireland, there is a strong peak production in spring and summer months, where the animals graze abundant feed and this is achieved thanks to spring calving. During the winter, because of the low feed resource available for grazing, the cows are dry. The milk industry has been adapted to this large peak-to-trough ratio. In contrast, France has a fairly flat distribution of monthly milk delivery. This has been achieved over the years thanks to a differential milk price between spring and winter, encouraging the farmers to produce milk in autumn and winter. The feed resource is then ensured by provision of conserved feed stocks, mainly silage, from either grass or maize. Maize silage has to be supplemented with soy meal. Indeed, when grazing grass swards, or mixtures of grass and legumes, the animals cover their needs in energy and protein. However, this may Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 34 not be the case for the very high yielding dairy cows, for whom the energy-protein balance in grazed grass may not be optimal. 4500 4000 3500 x 1000 3000 2500 Dairy cows 2000 Other cows 1500 1000 500 0 DE FR PL UK NL IE BE CZ SK LU Figure 7. Number of dairy and other cows in the countries of the temperate region in 2011. Source: Eurostat, 2012. Monthly deliveries of milk (x1000 tonnes) 2500 2000 1500 Ireland France 1000 500 0 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Figure 8. Changes in the monthly milk delivery of cow milk to the dairy industry in France and Ireland in 2011 (in 1000 tonnes). When fed with maize silage whose protein content is very low, supplementation with soy meal is compulsory. As a consequence, the agronomic This underlines the coherence between the grassland acreage and management, the extra feed and concentrate resources, the management Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 35 of dairy cows and the whole dairy industry sector. It also questions the genetic value of cows and their ability to valorise grasslands. In the course of Multisward project, the potential of hybrid breed for milk-solids production, performance for grazing and reproductive performance (intervals between calvings) was underlined. And it is an extra component of this coherent system for the optimized use of feed biomass produced by grasslands. However, the present trend in animal performance, especially milk production per cow, shows a constant increase in mean milk yield per cow and per year (Figure 9). The trend is very similar among countries of the temperate region, except Ireland where it is less steady. As a consequence, this increasing animal performance, a large part of which is due to increasing genetic merit, will not be fully relevant with a large share of the grasslands in the animal feed. A combination of grass and maize silage in a ratio, on DM basis varying between 60-40 to 4060, and complemented with concentrates, can fulfil the requirements of energy and protein for cows with a very high milk production. The trend of increasing performance is less pronounced in beef cattle, even though a regular increase in animal weight at slaughter is recorded. Beef cattle, in general heifers and suckler cows in particular, in all the countries of the temperate region, are mainly fed through grazing, with hay or grass silage during the winter in all the countries of the temperate region. They well valorise the permanent grasslands in many regions. In most intensive production systems, fattening bulls are kept in stables, and fed with maize silage and concentrates during the finishing period. 9000 Mean milk yield (l/cow/year) 8000 7000 France Poland 6000 Ireland UK Netherlands 5000 4000 3000 1975 1980 1985 1990 1995 2000 2005 2010 2015 Figure 9. Mean milk yield per dairy cow in France, UK, Netherlands, Ireland and Poland over the last decades. (Source: National statistics, ICAR database). Grassland-based dairy products have a very high value in most countries of the temperate regions. Many PDO cheeses are produced and give an additional market value. We could for instance name Comté or Beaufort in France produced in the Jura and Alps and fully based upon grazing or hay. It is also the case for Oscypek (PDO). This hard sheep’s milk cheese is produced Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 36 mainly in the Podhale and Tatra regions, and was once served as payment between farmers and senior shepherds. 3. Breeding and seed production The temperate region has a climate that is very favourable for the important breeding activity of forage species and also for seed production. Most large forage-breeding companies are based in this region, except DLF which is located in Denmark but it also has many trial sites in the temperate region. This situation occurs because the temperate region is a major European market and varieties must be well adapted, in terms of their phenology, potential for forage production, and resistance to foliar diseases. One of the most important tasks facing breeding of fodder grasses in this region was to develop genotypes from the Lolium genus for increased tolerance to drought and frost by way of introducing resistance genes using species from the Festuca genus by interspecific crossing (Zwierzykowski et al., 2004; Goliński et al., 2005). In conjunction with this situation, many public research institutes are also based in the temperate region (Teagasc, IBERS Aberystwyth, INRA Lusignan, University of Hohenheim, Wageningen University, University of Poznan, for instance). Seed production is also very important, both for forage grasses and legumes (Figure 10). Denmark is the only country that produces more grass seeds than any country of the temperate region, with 101,300 t on average between 2007 and 2009. Production is mainly devoted to grasses, with the exception of France and Czech Republic where there is also significant production of forage legumes, mainly lucerne and red clover. UNITED KINGDOM SLOVAKIA POLAND NETHERLANDS LUXEMBOURG Grasses IRELAND Legumes GERMANY FRANCE CZECH R. BELGIUM 0 5000 10000 15000 20000 25000 30000 Figure 10. Mean annual seed production in the various countries of the temperate region between 2007 and 2009 (in t). (Source: National Certification Agencies). 4. Environmental services of grasslands in the temperate regions In comparison with other European regions, the temperate zone has a fairly low number of Natura 2000 sites, but it hosts many areas of High Nature Value farmland, especially in Scotland and Wales, Massif Central and the Alps, South of Germany, and the eastern and southern parts of Poland. These regions are those where the grassland management is less intensive and amounts of biomass production are low. In the other grasslands, the high management intensity, both in terms of soil fertility or animal management, tends to lead to a fairly low level of species diversity. Indeed, the botanical composition of pasture vegetation is strongly influenced by Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 37 management (Štýbnarová et al., 2009). In the case of the remaining semi-natural pastures, botanical composition is the result of traditional agricultural activities such as haymaking or herding (Isselstein et al., 2005; Maurer et al., 2006; Peratoner et al., 2009). Plant speciesrichness declines as applications of fertilizer increase, especially nitrogen (e.g., Zechmeister et al., 2003) and phosphorus (Poozesh et al., 2008), even if, in some situations such as ultramafic soils of Tuscany, nutrient addition tended to increase the diversity (Ricotta et al., 2005). Nitrogen fertilizer levels — even when far below those applied on intensively managed grasslands — cause significant losses in sward plant diversity, with half of the number of plant species eliminated in response to fertilizations of between 20 and 50 kg N/ha/yr (Plantureux et al., 2005). P enrichment presented a greater threat to biodiversity than N enrichment in research on 132 semi-natural grasslands located along a gradient of nutrient availability and atmospheric N deposition. However, as N- and P-driven species-loss appeared to be independent, the results of this research suggest that simultaneously reducing N and P inputs is a prerequisite for maintaining maximum plant diversity (Ceulemans et al., 2013). The abundance of forage legumes in the swards will have a similar effect. The impact of grazing management on grassland biodiversity has been documented in many studies (e.g., Rook and Tallowin, 2005). Nutrient depletion can be accelerated by haymaking. However, atmospheric deposition of nutrients is increasing and such deposits are believed to slow down the effects that extensification has on biodiversity (Plantureux et al., 2005). Today, in regions with intensified agriculture such as the countries of the temperate region, semi-natural grasslands persist only on a low proportion of the total grassland area. Their preservation is a primary goal for nature conservation (Isselstein et al., 2005), such as through the Habitats Directive (1992) or the international treaty drawn up at the Ramsar Convention on Wetlands (1971). Semi-natural grasslands have persisted mainly on locations that are less suitable for agriculture because of biotic and abiotic constraints (Pärtel et al., 2005). Yet, areas of semi-natural grasslands still exist in Poland (Veen et al., 2001; Goliński and Golińska, 2011) where they must be preserved and possibly valorised economically, for instance through recreation of special and well-identified animal products (see above). Erosion is becoming a major issue for European soils (Souchère et al., 2003), as an effect of the change in land use. Grasslands are a key asset for limiting soil erosion because of their ability to reduce run-off, thanks to the permanent soil cover. Within the framework of good agricultural and environmental practices, permanent grassland is strongly recommended and in some regions is obligatory and financially supported on slopes with a minimal gradient. However, because of the reduction in the area of permanent grasslands in the temperate region countries, this issue will become increasing important and should be considered as a benefit of grasslands. Conclusions: the main drivers of the changes Grasslands are experiencing dramatic changes in the countries of European temperate region. All the changes are occurring in a very relevant way related to an increasing animal performance and to an increasing productivity of human work in larger farms. Beyond the search for a higher biomass productivity, which is possible in this region because of the soil and climate conditions, there is also a risk of grassland abandonment or ploughing, as herbivores, especially high genetic-merit dairy cows, are increasingly fed with maize silage supplemented with soya-based concentrate. The Common Agricultural Policy has been fully in line with this trend. Quotas, established in 1984, have provided security for dairy farmers and, in some countries, have slowed down the concentration of milk production and the dairy industry in the most favourable zones for biomass production, milk processing and export. The future disappearance of quotas will offer the possibility for a quick increase in mean production per dairy farm and, simultaneously a Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 38 reduction in the number of dairy farms, which will be concentrated in these favourable zones. These trends will, however, endanger the large regulating, provisioning and cultural ecosystem services provided by grassland and grassland-based production systems in the rest of the temperate region. As a consequence, it is very important that the future European and national policies of the countries in the temperate region take into account both the economic and environmental services provided by grasslands and grassland-based systems, and also consider the social aspects for farmers involved in herbivore production and of the whole of society which derives a huge benefit from grasslands. References Anonymous (2014) http://economie.fgov.be/nl/modules/publications/statistiques/economie/downloads/landbouw__landbouwgegevens.jsp Ceulemans T., Merckx R., Hens M. and Honnay O. (2013) Plant species loss from European semi-natural grasslands following nutrient enrichment – is it nitrogen or is it phosphorus? Global Ecology and Biogeography 22, 73-82. Goliński P. and Golińska B. (2011) Agri-environmental funding schemes – a tool for supporting the conservation of semi-natural grassland in Poland. Grassland Science in Europe 16, 592-594. Goliński P., Kozłowski S., Broda Z. and Mikulski W. (2005) Prospects of grass utilization in Poland and new approaches and goals in grass breeding. In: Zwierzykowski Z. and Kosmala A. (eds.) Recent advances in genetics and breeding of the grasses. Institute of Plant Genetics PAS, Poznań, Poland, pp. 147-162. Huyghe C., De Vliegher A., van Gils B. and Peeters A. (2014) Grassland and herbivore production in Europe and effects of common policies. Quae Editions, Paris, 290 pp. Huyghe C., Duru M., Peyraud J.L., Lherm M., Gensollen V., Bournoville R. and Couteaudier Y. (2005) Prairies et cultures fourragères : au carrefour des logiques de production et des enjeux environnementaux. Inra éditions, 209 pp. Isselstein J., Jeangros B. and Pavlu V. (2005) Agronomic aspects of biodiversity targeted management of temperate grasslands in Europe – A review. Agronomy Research 3, 139-151. Maurer K., Weyand A., Fischer M. and Stöcklin J. (2006) Old cultural traditions, in addition to land use and topography, are shaping plant diversity of grasslands in the Alps. Biological Conservation 130, 438-446. Mignolet C., Schott C., Benoît M. and Meynard J.M. (2012) Transformations des systèmes de production et des systèmes de culture du bassin de la Seine depuis les années 1970 : une spécialisation des territoires aux conséquences environnementales majeures. Innovations Agronomiques 22, 1-16. Pärtel M., Bruun H.H. and Sammul M. (2005) Biodiversity in temperate European grasslands: origin and conservation. Grassland Science in Europe 11, 1-13. Peratoner G., Figl U., Gottardi S., Bodner A., Werth E. and Kasal A. (2009) Temporal variability in the forage production of a protected area with heterogeneous vegetation types. Grassland Science in Europe 14, 547-550. Plantureux S., Peeters A. and McCracken D. (2005) Biodiversity in intensive grasslands: effect of management improvement and challenges. Agronomy Research 3, 153-164. Poozesh V., Ricotta C., Avena G. and Chiarucci A. (2005). Quantifying the effect of nutrient addition of the taxonomic disctinctness of serpentine vegetation. Plant Ecology 179, 21-29 Rook A.J. and Tallowin J.R.B. (2005) Grazing and pasture management for biodiversity benefit. Animal Research 52, 181-189. Schott C., Mignolet C. and Benoît M. (2009) Agriculture du bassin de la Seine : Découvrir l’agriculture du bassin de la Seine pour comprendre les enjeux de la gestion de l’eau. Nanterre (FRA) : Agence de l’Eau SeineNormandie. 79 p. Programme Piren - Seine, (5). http://www.sisyphe.upmc.fr/piren/book/978 Smit H., Metzger M. and Ewert F. (2008) Spatial distribution of grassland productivity and land use in Europe. Agricultural systems 98, 208-219. Souchère V., King C., Dubreuil N., Lecomte-Morel V., Le Bissonnais Y. and Chalat M. (2003) Grassland and crop trends: role of the European Union Common Agricultural Policy and consequences for runoff and soil erosion. Environmental Science and Policy 6, 7-16. Štýbnarová M., Hakl J., Krhovjáková J. and Pozdíšek J. (2009) Botanical composition of pasture sward influenced by intensity of utilization and mineral fertilisation. Grassland Science in Europe 14, 261-264. Veen P., Molnár Z., Pärtel M. and Nagy S. (2001) Grassland ecosystems in Central and Eastern Europe. Prepared in the framework of the High Level Conference on EU Enlargement, The relation between agriculture and nature management, 19 pp. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 39 Zechmeister H.G., Schmitzberger I., Steurer B., Peterseil J. and Wrbka T. (2003) The influence of land-use practices and economics on plant species richness in meadows. Biological Conservation 114, 165-177. Zwierzykowski Z. (2004) Amphiploid and introgression breeding within the Lolium-Festuca complex – achievements and perspectives. In: Yamada T. and Takamizo T. (eds.) Development of a novel grass with environmental stress tolerance and high forage quality through intergeneric hybridization between Lolium and Festuca, Nat. Agric. and Bio-orientated Res. Org., Japan, pp. 17-29. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 40 European grasslands overview: Mediterranean region Cosentino S.L.1, Porqueddu C.2, Copani V.1, Patanè C.3, Testa G.1, Scordia D.1 and Melis R.2 1 Dipartimento di Scienze delle Produzioni Agrarie e Alimentari (DISPA), Università degli Studi di Catania, Via Valdisavoia, 5, 95123 Catania, Italy 2 CNR-Istituto per il Sistema Produzione Animale in Ambiente Mediterraneo (ISPAAM) Traversa la Crucca 3, loc. Baldinca, 07100 Sassari, Italy 3 CNR-Istituto per la Valorizzazione del Legno e delle Specie Arboree (IVALSA) UOS di Catania, Via P. Gaifami 18, 95126 Catania, Italy Corresponding author: cosentin@unict.it Abstract The paper reviews the main traits of the grasslands in the Mediterranean basin, a global biodiversity hotspot where almost 22,500 endemic vascular plant species, more than four times the number found in all the rest of Europe, have been described. In this context the grasslandbased systems are no longer seen exclusively as livestock production enterprises but as multiple-use systems with important consequences for the global environment. Plants are subjected to very low rainfall during the spring-summer period (2-6 months) resulting in a 3001000 mm water deficit that negatively affects plant survival and crop production. For this reason the traits of species to cope with summer drought are discussed with special emphasis on drought escape in annual species and seed bank management, as well as drought survival in perennial grasses. The agronomic role and importance of legumes in the grassland mixtures are also described pointing out the grass-legume interactions. The prediction of the influence of the global warming on Mediterranean pasture grasses is analysed, as well as the contribution of grassland to environmental issues (multifunctionality). With this regard the results of the role of the grasslands on soil erosion control and carbon sequestration are reported. The conservation of biodiversity in grassland species is discussed. Finally, the use of grasslands as source of biomass for bioenergy and bio-refining are also presented. Keywords: drought escape, summer survival, multifunctionality, soil erosion, carbon sequestration, biodiversity, bioenergy Introduction There are five areas of the world with Mediterranean-type climates: the Mediterranean basin itself, South Africa, California, Chile and southern Australia. By far the largest is the Mediterranean basin, it covers approximately 2.5 millions km2; about half of the Mediterranean areas are located on the European side and over than 50% is represented by grasslands and rangelands (Eurostat, 2010). They are important ecosystems that have traditionally played an important role in the evolution of human societies (Jouven et al., 2010). The fact that the Mediterranean basin is a global biodiversity hotspot with an extremely high number of endemic plant species is strongly related to the long-term management practices. Livestock grazing is the major factor in these ecosystems, which in addition to the quantity and quality of forage, strongly affects vegetation dynamics, species, and landscape diversity (Perevolotsky, 2005). Animal production systems in Mediterranean Europe are dominated by small ruminants and beef cattle belonging to local races which, with respect to dairy cows, exploit better unfavourable areas and shrublands under all-year-round open-air grazing. Moreover, Mediterranean grassland-based systems are no longer seen exclusively as livestock production enterprises but as multiple-use systems with important consequences for the global environment. Elevation greatly affects local climate and ecology in this area and vegetation can be distinguished on the basis of thermal criteria (Barbero and Quezel, 1981). Plants are subjected Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 41 to very low rainfall during the spring-summer period (2-5 months) resulting in a 300-1000 mm water deficit (Volaire et al., 2009) that negatively affects plant survival and crop production. The stress imposed during summer months is the most important environmental constraints for plant growth in association with large intra- and inter-annual variability of rainfall distribution. This rainfall pattern induces a highly seasonal growth of plants during spring months (February-March/May-June) and reduced growth or dormancy of plants during summer months, with a second growth period occurring in autumn. In this paper we consider a recent classification proposed by Peeters et al., (2013) which defines grasslands as ‘land devoted to the production of forage for harvest by grazing/browsing, cutting, or both, or used for other agricultural purposes such as renewable energy production’. The conservation of grasslands is important not only as a feed source but also because they support biodiversity, contribute to the reduction of CO2 levels from atmosphere, as they act as a carbon sink, and generate several environmental and economic services such as prevention of fire risk, recreational activities and tourism (Carrillo et al., 2014). Therefore, the characterization of the environmental threats that affect European Mediterranean grasslands and the ability of species to cope with climate change is a fundamental step to increasing resilience of grassland. Designing resilient and sustainable grasslands: traits of species to cope with summer drought Drought escape in annual species and seed bank management Mediterranean semi-natural grasslands are dominated by annual species well adapted to the highly variable Mediterranean climate because they produce a huge amount of seeds that can survive for a long time in the soil seed bank that plays an important role in the composition and conservation of plant communities (Koukoura, 2007). The annual composition and abundance of species is subjected to wide fluctuations, depending on the number of seeds produced and their dormancy type and degree, the satisfaction of their germination requirements, the soil physical and microbiological characteristics, the pattern of environmental factors, grazing, burning and other soil disturbance factors that can influence the total amount of seed produced by the plant community (Long et al., 2014). For example, increasing air temperatures may affect seed bank through an increased soil moisture evaporation, which induces a reduced seedling survival and related species richness (Ooi, 2012). Drought escape is the main adaptive strategy that is exhibited by annual pasture or forage species surviving during the dry period as seed. These species frequently exhibit seed dormancy, which has been defined as an ‘internal condition of the seed that prevents its germination under even favourable water, thermal and gaseous conditions’ (Benech-Arnold et al., 2000). This condition of the seed to remain viable but quiescent allows the seed itself to persist in the soil for months or years until the environmental conditions permit germination (Baskin and Baskin, 2004). Temperature is the major environmental factor regulating seed dormancy and germination, and seasonal changes in temperature are responsible for the loss and cycling of dormancy (Hilhorst, 1998). In legumes, seed dormancy is determined by the presence of a water-impermeable coat. Hardseededness is believed to be controlled by both genetic and environmental factors occurring during seed development (Clua and Gimenez, 2003). In this regard, the results of a study conducted on Medicago rugosa Desr. and M. orbicularis Bartal, two annual pasture legumes quite common in the Mediterranean basin, revealed that the level of hardseededness may greatly depend on the intensity of plant water stress occurring during seed development and filling, and that more permeable and promptly germinating seeds could be produced by modifying the water balance (Norman et al., 2002; Patanè et al., 2008). The softening of hard seeds differs between and within species and in Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 42 many situations the level of hardseededness may be less important than the softening pattern, as found by Porqueddu et al. (1996) in M. polymorpha L. Hardseededness is responsible for the differences in seed germination rate through the year in natural conditions; however, it may represent a limit when a prompt germination is required, i.e. sown pasture. Both mechanical and chemical scarification are useful for the breakdown of hardseededness, but the outcome is influenced by seed traits. In an experiment, seeds of M. rugosa exhibited a 65% germination after 35 min soaking in a 70% concentrated sulphuric acid solution whilst seeds of M. orbicularis did not germinate under the same chemical treatment (Patanè and Bradford, 1993). These results were supported by the different thickness of the palisade layer of the seed coat, as observed at SEM (Patanè and Gresta, 2006), which is more than two-fold greater in M. orbicularis when compared to M. rugosa. In addition to meteorological factors, grassland management practices may also have huge effects on soil seed bank dynamics but their effects are controversial. Under grazing, the type of regime (continuous vs. seasonal), grazing pressure (low or heavy) and grazing season have different effects on seed bank density and on the different functional groups of plants, with heavy grazing being unfavourable to the seed bank of annuals if grazing is prolonged during the period of seed setting (Sternberg et al., 2003). Nonetheless, the maintenance of Mediterranean grasslands with a high degree of plant biodiversity depends on frequent grazing that controls the growth and predominance of a few dominant tall grass species (Sternberg et al., 2000). In general, intensive grazing seems to favour Mediterranean annual legumes, especially subclover-based pastures. Although winter grazing is advantageous to seed production, avoiding excessive predation of seed by livestock grazing stubble is also a critical aspect in legume persistence. The persistence is not only related to the seed yield but also to the pod and seed characteristics of each pasture legume, e.g. seed size, seed dispersal capacity and seed burial ability. Drought survival in perennial species Drought tolerance is expressed in slowly growing perennial plants (Chaves et al., 2003). Drought tolerance in perennial species is also of great agronomic importance. In fact, the production period of annuals is short and could be further compromised by climate change. The use of perennial forage species could extend the feeding season, increase yield during winter and also enable production in late spring in the presence of residual soil moisture (Annicchiarico et al., 2011; Annicchiarico et al., 2013; Volaire, 2008; Volaire et al., 2013). There are, however, very few cultivars that are suitable for Mediterranean severe drought conditions (Annicchiarico et al., 2013; Volaire et al., 2013). The required characteristics, as reported by Volaire et al. (2013), are not to grow during the drought period but survive across the droughts (Annicchiarico et al., 2011). In Mediterranean and temperate plants a combination of plant strategies to overcome drought are present (Volaire and Lelievre, 1997; Poirier et al., 2012). Plant adaptations include dehydration delay (avoidance), for example by increased root development and water uptake (Volaire and Lelievre, 2001), dehydration tolerance by a superior osmotic potential through solute accumulation and osmotic adjustment (Volaire, 2008) and plant summer dormancy, an adaptive response that enables survival during the most threatening seasons because it maintains viability of meristems and prevents regrowth during cases of occasional summer rainfall (Volaire et al., 2009; Lelièvre et al., 2011). Summer-dormant genotypes showed a superior survival after severe and repeated summer droughts (Norton et al., 2006; 2012). Summer dormancy is an endogenously controlled process that results in the cessation or reduction of leaf growth in summer and the senescence of mature herbage in some genotypes of perennial pasture species, such as cocksfoot (Dactylis glomerata), tall fescue (Festuca arundinacea), bulbous canary grass (Phalaris aquatica) and perennial ryegrass (Lolium perenne), and is expressed under high temperatures and long Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 43 photoperiods and, in the genotypes which have the trait, expression will occur independently of soil moisture level (Norton et al., 2006). Copani et al. (2012) carried out a field experiment in order to investigate the genetic diversity and the presence of summer dormancy in several wild Sicilian populations of cocksfoot. The results highlighted that a high level of summer dormancy was present in most populations; however, two populations showed an incomplete dormancy and two showed the absence of dormancy. The same authors observed a relationship between summer dormancy level and the microclimate of the collection site. Most of the summer-dormant populations came from sites with rainfall < 600 mm year-1 and a dry period > 120 days (Copani et al., 2012). Norton et al. (2012) suggested selection for summer dormancy and dehydration-tolerance genotypes prior to field screening, as this is highly likely to improve overall persistence. A method based on germination response to photoperiod to differentiate summer-dormant from summer-active types of perennial grasses has been developed by Malinowski et al. (2008). The authors observed that the germination response to photoperiod was associated with summer dormancy type and not necessarily with the climate where grasses developed. Moreover, Annichiarico et al. (2011) highlighted the need to select different adaption targets, plant types and genetic resources for the different environments. For instance, summer-dormant genotypes of D. glomerata subsp. hispanica have a prevailing interest for northern Africa. Conversely, non-dormant or incompletely dormant Mediterranean cultivars of D. glomerata subsp. glomerata are best adapted for southern Europe, especially when targeted to moderate crop duration (3–4 years). Nonetheless, completely summer-dormant germplasm could gain adaptive potential for Mediterranean-climate European regions in the future, to mitigate the effects of the predicted increase in drought due to climate change. The concurrent use of plants using different strategies to overcome drought is one of the adaptation strategies proposed by Kreyling et al. (2008), useful to establish permanent and multi-specific grasslands that ensure ecosystem stability, in order to enhance the sustainability of agricultural production and ecosystem services (Volaire et al., 2013). The role and use of legumes Naturally occurring forage legumes (annuals and perennials) are well-nodulated, and their root nodules are active in fixing N2. As a consequence, legumes contribute to increase fertility of soils, help to maintain the organic matter of soil and improve the physical conditions of soil, promoting efficient low input and low cost production systems and thereby reducing the need for inorganic fertilizers. This is true mostly for the European Mediterranean regions characterized by a favourable climate for legume growth and very poor soils (Porqueddu, 2001; Sulas, 2005). Testa and Cosentino (2009) evaluated, using the isotope dilution method, the amount of nitrogen fixation of three legumes (Medicago sativa, Trifolium subterraneum and Vicia faba subsp. Minor). The percentage of nitrogen derived from the atmosphere (Ndfa) ranged from 94% to 84% (respectively in the first and second year) in M. sativa, from 69% to 81% in T. subterraneum and from 91% to 93% in V. faba. Other similar researches carried out by Cosentino et al. (2003) highlighted high values of nitrogen derived from atmosphere in sulla and alfalfa, corresponding from 136 and 162 kg ha-1 of nitrogen fixed by sulla (in the first and second year, respectively) and 67 and 184 kg ha-1 of fixed nitrogen in alfalfa. Similar values were obtained by Sulas et al. (2009) that in Northwest Sardinia (Sassari) and Central Western Italy, obtained 113 kg ha-1 and 130 kg ha-1 of nitrogen fixed on Sulla coronaria (L.) Medik (sulla). Moreover, it is well known that legumes improve nutritional quality of forage, reducing the need for concentrate feeds, also if their nutritive value may be influenced by the presence of anti-quality factors (Papanastasis and Papachristou, 2000). In addition, if present in moderate concentrations, some secondary metabolites of legumes, i.e. condensed tannins, can enhance Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 44 forage nutritive value by playing a role in promoting amino-acid absorption in the intestine, decreasing nitrogen excretion, and they can reduce GHG emission in atmosphere and the loads of gastro-intestinal parasites (Piluzza et al., 2014). Legume forage can influence the quality of meat in cattle (Maughan et al., 2014) and increase daily average liveweight gain, linolenic and linoleic acids proportion and unsaturated to saturated fatty acids in lamb meat (Fraser et al., 2004). In the presence of a natural seed bank of pasture legumes in a grassland, a single or a yearlyrepeated phosphorus fertilization, without any over-seeding, may be sufficient to gain satisfactory productive results (Bullitta et al., 1989; Henkin and Seligman, 2000) and increase the resilience of the system. In the last forty years, the annual self-reseeding legumes selected in Australia have been also increasingly utilized in Mediterranean Europe, i.e. subterranean clovers and medics, mainly for the improvement of low quality native pastures in agro-pastoral systems, although these commercial varieties have sometimes been shown to be unfit for the different climatic conditions and management systems of southern Europe (Sulas, 2005, Porqueddu et al., 2010, Salis et al., 2012) and an effort is required to promote the selection and the seed multiplication of native genotypes. A review on annual self-reseeding legumes and their role in Mediterranean farming systems was done by Porqueddu and Gonzales (2006). Among perennial species, the well-known alfalfa (Medicago sativa L.) is a drought-tolerant perennial pasture legume that is valuable in many farming systems due to its ability to produce summer fodder. The deep taproot system allows alfalfa to access to deep soil moisture, tolerate long dry periods, and respond quickly to summer rainfall. These characteristics were confirmed by Testa et al. (2011) who evaluated the effect of harvest time and soil water content on three varieties of alfalfa in a four-year experiment. They found that alfalfa could be used for the improvement of qualitative and quantitative characteristics of forage in 'low mountain' areas of the Mediterranean, thanks to the slight decrease in yield and quality in rainfed conditions. Furthermore, this research highlighted that, by combining two harvest times, producers can ensure fresh forage from May to November. Yet, some perennial legumes are able to escape summer drought and regrow at first rain in autumn, as in the case of sulla and Onobrychis viciifolia Scop. (sainfoin), and their exploitation offers an opportunity to stabilize both grassland production and forage quality (Sulas, 2005; Demdoum et al., 2010; Re et al., 2014). More recently, Australian research has focused on deep-rooted and drought-tolerant perennial legumes (e.g., Caucasian clover, stoloniferous red clover, tallish clover, etc.), which also have a high feeding values that decline slowly with maturity. Recent research indicates that the Mediterranean Bituminaria bituminosa L. (syn. Psoralea bituminosa C.H. Stirton) has potential as a forage legume for disadvantaged areas of the Mediterranean (Martínez-Fernández et al., 2012). This perennial legume is drought tolerant due to a deep root system and other physiological adaptive traits (Castello et al., 2013). It grows and remains green all-year-round even during summer and autumn and is assumed to be tolerant of heavy grazing (Sternberg et al., 2006). The preliminary results obtained in different Mediterranean regions are very encouraging in view of the valorization of Psoralea as perennial forage legume for marginal rainfed areas (Real and Verbyla, 2010; Porqueddu et al., 2011). Growing Psoralea in permanent dense stands may provide alternative sources of highly nutritious forage (Reaside et al., 2013). Nonetheless, it has been estimated that approximately 70% of freshly harvested seeds of this species are hard and exhibit a threemonths primary dormancy, the causes of which are still unknown (Castello et al., 2013). Grassland mixtures and grass-legume interactions Under sub-optimal and variable environmental conditions, as occur in many pastoral farming systems of the Mediterranean basin, the maintenance of high levels of inter- and intra-specific diversity is essential to achieve satisfactory and persistent pasture swards, and also for the Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 45 control of weeds and replacement of forage plants damaged by drought or biotic adversity. When improvement of the native sward is needed, the use of mixtures of pasture species rather than a single species seems more appropriate (Dear and Roggero, 2003). Nonetheless, there has been little experimentation and there is little information available on production, biomass composition and effect of grazing on the persistence and environmental impact of mixedlegume swards (Rochon et al., 2004). There are many advantages connected to the use of forage mixtures in relation to agronomic (Cosentino et al., 2003), environmental and economic aspects (Malhi et al., 2002). Due to the utilization of nitrogen symbiotically fixed by legumes (Whitehead, 1995), grass crops, when grown in association with legumes, may improve forage dry matter and protein yield even if the efficiency of this utilization could be strongly influenced by genotypic and environmental factors (Cassaniti et al., 2003a; 2003b). Testa et al. (2006) studied the nitrogen transferred from a legume (alfalfa or Lotus corniculatus) in a binary mixture with a grass (Italian ryegrass) at different sowing rates. The estimation of % Ndfa in the grass component showed different behaviour in relation to legumes involved in the mixtures and its density. In alfalfa, the highest sowing density positively affected the amount of nitrogen derived from the atmosphere in the grass component. A coordinated field experiment across 31 European sites confirmed that a significant yield gain was obtained when four-species mixtures were used, yielding more than the highest-yielding monoculture at most sites, including the Mediterranean ones (Finn et al., 2013). Porqueddu and Maltoni (2007) found that the use of species (L. rigidum, D. glomerata, M. polymorpha, M. sativa) belonging to different functional groups (fast and low establishing grasses, fast and low establishing legumes) enabled the achievement of higher yields, a better seasonal forage distribution, and a better weed control, and higher forage quality and lower seasonal variation with respect to pure stands were assessed (Maltoni et al., 2007). In the Mediterranean environment the advantage of using mixtures of grass and legume was confirmed by Riggi et al. (2006), who tested legume and grass mixtures at different sowing ratios. A stable pasture mixture could be obtained also with a mixture of summer-dormant and summer-active species or varieties, so that the available moisture in soil could be exploited throughout the year (Norton and Volaire, 2012). The stability and persistence of each component in the mixture is another point to be considered when determining mixture composition. Several studies have shown that in mixtures based on annual medics, T. michelianum and subclover, the latter species prevails from the second year onwards (Porqueddu et al., 2004). Prediction of climatic changes on Mediterranean pasture grasses Climatic models forecast climate change will have a great impact on agricultural systems and production, with different intensity depending on the region. Warmer and drier conditions by 2050 are predicted for Europe. Southern Europe is likely to experience the largest yield losses (-25 % by 2080 under a 5.4 °C warming), with increased risks of failure especially for rain-fed summer crops (IPCC, 2013). Increased inter-annual variability may be another significant aspect of climate change, and this is of high ecological relevance. Climate change is considered likely to have a long-term great impact on all plants, primarily those whose persistence depends on the soil seed bank, e.g. grassland plants. This fact involves an assessment of biodiversity in response to future scenarios, in order to identify those species that are less susceptible to climatic changes. The adaptive capacity of these species, however, depends on several factors, including their adaptation speed to changed climatic conditions. However, a meta-analysis carried out by Dumont et al. (2014) did not reveal any variations in the response of grasses, forbs and legumes to elevated CO2, warming and drought under Mediterranean conditions. Climate change is thus not expected to directly affect the chemical composition of grassland species, but could shift grassland botanical composition. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 46 The strong relationship between climatic factors and seed dormancy and germination indicates that climatic changes will certainly have impact on seed banks, which ensure survival of the seed population, particularly under uncertain climates. Besides temperature, water is crucial for seed germination and its availability greatly affects seed dormancy and germination speed (Bradford, 2002). In a climatic change scenario it may be essential to predict seed germination under specific thermal and water environmental conditions, and modeling germination response to modified environmental conditions may contribute to increasing the accuracy of predicted climate change impacts on each plant species (Ooi, 2012). Thermal time, hydrotime and hydrothermal time models have been proposed to predict how any fraction of the seed population will respond to an environmental disturb or change, in relation to its physiological state (Bradford, 2002). To predict long-term consequences of climatic changes on species distribution and potential extinction, a deep study is required of relationships associating these changes with mechanisms regulating seed bank longevity in ecosystems where population dynamics are driven by environmental factors. Contribution of grassland to environmental issues (multifunctionality) Soil erosion control Mediterranean areas are susceptible to soil erosion due to orographic and climatic conditions. Rainfall is mainly concentrated in the autumn-winter period, thereby supporting the cultivation of autumn-winter crops. According to Plan Bleu (2003) the area of land subjected to soil erosion in the Mediterranean environment covers 1,309,000 km2, equal to 15% of the land of Mediterranean countries. The long dry period followed by heavy bursts of erosive rainfall, falling on steep slopes with fragile soils, resulting in considerable amounts of soil erosion that, especially, through the loss of organic matter and nutrients leads to a reduction of cultivable soil depth and soil fertility (Van Rompaey et al., 2005). This causes a loss of productivity, which initially is replaced with increased applications of fertilizers, but at the end this can lead to land abandonment (Cerdan et al., 2010). In this environment the growing cycle of winter cereals and annual forage crops (e.g. oats-vetch) determines a lack of soil covering during the first rains which occur in the early months of autumn and winter. These aspects are confirmed and highlighted by an ongoing research carried out by Cosentino et al. (2008, 2011) from 1996 in Sicily on a slope of 26-28%. In each year the effect of twelve cropping systems, among which some were meadows, were evaluated. The highest annual values of soil losses were observed in the annual tilled crops, 23 t ha-1 yr-1 in the average of the crop rotations legumecereal (d. wheat)-brassica, 15.5 t ha-1 yr-1 in the crop rotation d. wheat - d. wheat - set aside, and 11 t ha-1 yr-1 in the monoculture of d. wheat. Very low soil losses were observed in the plots managed with perennial crops, alfalfa (0.15 t ha-1 yr-1), Italian ryegrass in pure stand and in mixtures with T. subterraneum (1.8 t ha-1 yr-1) and tall fescue followed by subterranean clover (1.3 t ha-1 yr-1). The perennial Miscanthus and Medicago arborea reduced soil losses to 0.1 t ha-1 yr-1. Introducing in this environment the use of conservative techniques (sod seeding) in the cultivation of cereals, allows some reduction in soil losses (2 t ha-1 yr-1) maintaining the yield that is not different to that of conventional tillage. Porqueddu et al. (1994), in similar experiments carried out in Sardinia, showed that on a 30% slope, average annual soil losses were ten and twenty times higher in annual forage crop and bare soil respectively, compared to permanent grassland. In addition, annual forage crops did not give great advantage in terms of forage yield compared to that of fertilized permanent pasture on slopes. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 47 Carbon sequestration The role that natural rangelands play in the global carbon cycle is extremely important, accounting for 10–30% of the world’s total soil organic carbon (Brevik, 2012). As reported by Doran and Zeiss (2000) the thin layer of soil covering the surface of the earth is a major interface between agriculture and the environment and represents the difference between survival and extinction for most land-based life. In the soil, decomposition processes, as mediated by organisms in soil, play a predominant role in completing the cycle of life, in recycling of the building block nutrients to plants, and C as CO2 to the atmosphere (Doran, 2002). These assumptions were confirmed by a long-term field experiment carried out by Cosentino et al. (2013): in all the plots where perennial or forage crops were cultivated it was observed that there was a higher amount of carbon stored in the soil. Comparing the soil organic matter (SOM) content observed in two typical Mediterranean cropping systems involving durum wheat (two years of durum wheat followed by fallow) with plots managed with grasses (Lolium multiflorum Lam.) or forage legume (M. sativa), they observed a slight decrease in the SOM of the plots cultivated with durum wheat, from 1.34% to 1.18% (the average of the 0-40 cm soil layers) while a slight increase was observed in a plot managed for 7 years with L. multiflorum and then with Lolium intercropped with subterranean clover (from 1.58 to 1.66%, representing an equivalent of 2.6 t yr-1). In this plot a different behaviour was observed in relation to the depth of the soil layer. In the first layer (0-20 cm) an increase from 1.56% to 2.11% was observed, while a slight decrease from 1.59% to 1.22% was recorded in the deeper layer (21-40 cm). In plots with M. sativa, a continuous increase of SOM (about 29.5%) was observed during four years, corresponding to an average of 7.4 t ha-1 of CO2 stored. The benefit of the cultivation of perennial crops instead of tilled crops was highlighted in a plot managed only with perennial crops (Miscanthus from 1997 to 2001 and M. arborea from 2002 until today) with an average increase in the first 40 cm of soil, from 1.04% to 2.05% of SOM, corresponding to 6.8 t ha-1 of CO2 stored. Conservation of biodiversity The flora biodiversity of the Mediterranean Basin 'biodiversity hotspot' is outstanding, with 15,000 to 25,000 species, 60% of which are unique to the region. These semi-natural agroecosystems are threatened by three main factors: the abandonment of husbandry, the intensification of agricultural practices and global warming. Moreover, the occurrence of recurrent fires in some Mediterranean areas due to the accumulation of flammable shrubby vegetation is of special concern from the environmental point of view, as it dramatically alters soil characteristics and impacts on the local flora and fauna communities (Rosa García et al., 2010; Osoro et al., 2012). All these disturbance factors affect biodiversity, which constitutes the most important stability factor of ecosystems and agroecosystems (Sala et al., 2000; Duffy, 2002; Spehn et al., 2005; Rockström et al., 2009; Brussaard et al., 2010; Fontaine, 2011). Rosa García et al. (2013) underlined that halting and reversing the decline of permanent pastures, including ligneous pastures, is one of the biggest challenges for the maintenance of European biodiversity and wider ecosystem services, including the Mediterranean areas, and that the development of proper management strategies is fundamental. The latest EU biodiversity targets support the maintenance of many semi-natural permanent pastures in farmlands, not only within Natura 2000, and the maintaining, enhancing and restoring of ecosystem services (EC, 2011). There is growing evidence from several systems that important ecosystem processes, such as productivity and nutrient cycling, can be significantly related to the species richness of plants (Duffy, 2002). Sala et al. (2000) believe that the Mediterranean climate and grassland ecosystems are likely to experience the greatest proportional change in biodiversity because of the substantial influence of all drivers of biodiversity change. Spehn et al. (2005), analysing Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 48 the relationship between plant diversity and ecosystem functioning within the European BIODEPTH network of plant-diversity manipulation experiments, showed that altering biodiversity changes in the numbers and types of plant species and functional groups in experimental communities significantly affected all ecosystem processes examined over the investigated three-year period. The consistent effects of species richness on multi-functionality over and above those of climate and abiotic factors highlight the importance of plant biodiversity as a driver of multifunctionality in drylands. Temporal stability of the ecosystem increases with diversity, despite a lower temporal stability of individual species, because of both portfolio (statistical averaging) and over-yielding effects. Scientific results indicate that the reliable, efficient and sustainable supply of some foods (e.g. livestock fodder), biofuels and ecosystem services can be enhanced by the use of biodiversity (Tillman et al., 2006; Lüscher et al., 2008). Novel products from grassland (bioenergy and biorefining) As reported by Rӧsch et al. (2009), in Central Europe an increasing portion of the grassland is no longer needed for feeding cattle and is now attractive for energy purposes. In southern Europe this situation does not usually occur under rainfed conditions, due to the scarcity of forage availability. Owing to the high hemicellulose and cellulose and low lignin content, permanent grassland might be successfully used as feedstock for second-generation ethanol production. Generally, the more delayed the harvest the higher the structural polysaccharides (Prochnow et al., 2013), which are the primary substrates for ethanol production. In a typical Mediterranean grassland, the hemicellulose content decreases when harvest is delayed from June to September, while cellulose content increases (Martillotti et al., 1996). In order to calculate the theoretical ethanol yield (TEY) the equation reported in Scordia et al. (2014) may help to assess either the TEY from a dry feedstock tonne (kg ethanol DM t-1) and the TEY per unit land area (L ethanol ha1 ). Overall TEY, summing up cellulose and hemicellulose content, might rise from 566.7 to 599.3 kg ethanol DM t-1 moving from June to September cut. The biomass yield of unsown permanent grassland in the Mediterranean area is in the range of 1.0 to 5.0 t DM ha-1, leading to a minimum of 718.3 L of ethanol ha-1 to a maximum ethanol value of 3798.0 L ha-1. In Europe, the most investigated perennial grasses for energy purposes are switchgrass, reed canary grass, miscanthus and giant reed (Lewandowski et al., 2003; Cosentino et al., 2006; Cosentino et al., 2007; Monti et al., 2008, Scordia et al., 2013). Actual Mediterranean constraints and predicted climate change now require the identification of native perennial grasses that are able to use natural resources efficiently (www.optimafp7.eu). In this regard, Copani et al. (2013) are evaluating the physiological and productive responses of some native perennial grasses widespread in semi-arid Mediterranean area, such as Oryzopsis miliacea (L.) Asch. & Schweinf, Cymbopogon hirtus L. Janchen, Lygeum spartum L., Sorghum halepense L. (Pers.), or endemic in Sicily as Saccharum spontaneum L. spp. aegyptiacum (Willd.) Hackel, for bioenergy purposes. Saccharum was the highest yielding species both at first and second year harvest (9.8 and 18 t DM ha-1), while Sorghum, Cymbopogon and Oryzopsis showed no significant differences both at first and second year (2.7 and 5.5, 2.6 and 4.0, 3.5 and 3.6 t DM ha-1, respectively). In the same environment, Cosentino et al. (2012) showed that Saccharum reached yields higher than 30.0 t DM ha-1 when 50% or 100% ETm restitution was applied in an older stand and indicated this species as potential energy crop for this environment. A similar activity is being carried out in Sardinia with the native O. miliacea, Ampelodesmos mauritanica Th. Dur. & Schinz. and Hyparrhenia hirta (L.) Stapf, where these species are studied. From the preliminary results, O. miliacea seems the most promising species, showing high DMY and a favourable biomass allocation, mainly in tillers (Porqueddu et al., 2014 [these proceedings]). Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 49 Previous experiments showed that Miscanthus needs irrigation water to achieve high yields in the Mediterranean area (27 t DM ha-1 with 100% ETm restitution) (Cosentino et al., 2007). Results were corroborated by a long-term Miscanthus trial (17-years) grown in rainfed condition, which showed a marked yield reduction when rainfall decreased from 952 to 467 mm yr-1 (15.4 and 9.8 t DM ha-1, respectively). On the other hand, giant reed (16-year-old stand) grown in a similar long-term experiment seems to be more drought resistant than Miscanthus; indeed, biomass reduction resulted more contained as drought increased (20.2 and 17.7 t DM ha-1) (OPTIMA project, unpublished results). Given the multiple functions that grassland and grassland landscapes can provide for society, their use as a source of energy feedstock has to be carefully evaluated. Grassland might need economic support for it to be comparable to other high yielding species dedicated to biomass production (Leible et al., 2005). In this context, the use of grasslands for bioconversion or heating/energy production may be restricted to areas that cannot be ploughed, and in marginal environments in general after satisfying the requirements for livestock feeding (Peeters, 2009). Future challenges and perspectives The newly recognized multifunctional role of grasslands requires a renewal in grassland science, whose objectives should shift from the focus on the main function of grassland as a forage resource to a much broader concept of sustainable resource management involving environmental protection and conservation, livestock production and socio-economic development. This is especially true for the Mediterranean area, where land abandonment is a threat for grassland-based pastoral systems and the natural environment. The importance in soil protection and for carbon sequestration to prevent greenhouse gases emissions has been shown, as well as the possibility of bioenergy production from grassland as a renewable source of energy with low greenhouse emissions. Considering the importance of the Mediterranean basin as biodiversity hotspot, the opportunity to develop plant conservation strategies must be considered and risk maps at eco-regional scales should be used to inform stakeholders of grassland vulnerabilities, and to suggest management recommendations for ecologically significant areas expected to be sensitive to climate change. Despite their ecological, economic and social importance, grasslands still receive only limited scientific, political and media attention in relation to their conservation merits. This is mainly because they are widely perceived as bad and/or degraded lands suitable only for grazing, although in many Mediterranean marginal areas of Europe permanent grasslands are the basis of pastoral farming systems characterized by a low input management and the exploitation of local rustic breeds. The outcomes of these pastoral systems are farm products with special sensorial and nutritive qualities which are comparable to organic production. Therefore, the organization and full valorization of these products is desirable, as this can also play an important role in connecting rural and urban culture, and consequently rural people may be valued properly according to their importance for the actual society. Future multidisciplinary investigations on particular grasslands types and plant species components related to the quality and value of livestock products are needed. At the same time, a successful development of a European seed industry for well-adapted Mediterranean grassland species is needed to make effective the selection activities carried out by several public research institutions. Finally, more efforts in on-farm experimentation and knowledge transfer to farmers are required, with a special focus on the correct incorporation and management of legumes and grass-legume mixture in agricultural systems, which is necessary for their full exploitation. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 50 References Annicchiarico P., Peccetti L., Bouzerzour H., Kallida R., Khedim A. and Porqueddu C. (2011) Adaptation of contrasting cocksfoot plant types to agricultural environments across the Mediterranean basin. Environmental and Experimental Botany 74, 82-89. Annicchiarico P., Peccetti L., Abdelguerfi A., Bouzerzour H. and Kallida R. (2013) Optimal forage germplasm for drought-prone mediterranean environments. Field Crops Research 148, 9-14. Barbero M. and Quézel P. (1981) Classifying Mediterranean ecosystems in the Mediterranean rim countries and in Southwestern U.S.A. Symposium on Dynamics and Management of Mediterranean-type Ecosystems, June 2226, 1981, San Diego, California. (http://www.fs.fed.us/psw/publications/documents/psw_gtr058/psw_gtr058_2a_barbero.pdf) Baskin J.M. and Baskin C.C. (2004) A classification system for seed dormancy. Seed Science Research 14, 1-16. Benech-Arnold R. L., Sánchez R.A., Forcella F., Kruk B.C. and Ghersa, C.M. (2000) Environmental control of dormancy in weed seed banks in soil. Field Crops Research 67, 105-122. Bradford K.J. (2002) Applications of hydrothermal time to quantifying and modeling seed germination and dormancy. Weed Science 50, 248-260. Brevik E.C. (2012) Soils and climate change: gas fluxes and soil processes. Soil Horizons. Online at: https://www.soils.org/publications/sh/pdfs/53/4/12 Brussaard L., Caron P., Campbell B., Lipper L., Mainka S., Rabbinge R., Babin D. and Pulleman M. (2010) Reconciling biodiversity conservation and food security: scientific challenges for a new agriculture. Current Opinion in Environmental Sustainability 2, 34-42. Bullitta P., Bullitta S. and Roggero P.P. (1989) Agronomic methods to increase pasturelands production in Mediterranean marginal areas. Proceedings of XVI International Grassland Congress, Nice, France, 1989, 15911592. Carrillo, Y., Dijkstra, F.A., Le Cain, D., Morgan, J.A., Blumenthal, D., Waldron, S. and Pendall, E. (2014) Disentangling root responses to climate change in a semiarid grassland. Oecologia DOI 10.1007/s00442-0142912-z Cassaniti S., Cosentino S.L., Riggi E. and Copani V. (2003a) Produttività e asportazione di azoto da parte della coltura graminacea in erbai bifiti di loiessa e leguminose foraggere. Rivista di Agronomia 37, 79-88. Cassaniti S., Cosentino S., Riggi E., Copani V. and Testa G. (2003b) Effects of legume species on the N-uptake of barley (Hordeum vulgare L.) and dry matter yield of the respective binary grass/legume mixtures. Grassland Science in Europe 8, 449-452. Castello M., Croser J.S., Lulsdorf M.M., Ramankutty P., Pradhan A., Nelson M.N. and Real D. (2013) Breaking primary dormancy in seeds of the perennial pasture legume tedera (Bituminaria bituminosa C.H. Stirt. vars albomarginata and crassiuscula). Grass and Forage Science online at doi 10.1111/gfs 12107. Cerdan O., Govers G., Le Bissonnais Y., Van Oost K., Poesen J., Saby N., Gobin A., Vacca A., Quinton J., Auerswald K., Klik A., Kwaad F.J.P.M., Raclot D., Ionita I., Rejman J., Rousseva S., Muxart T., Roxo M.J. and Dostal T. (2010) Rates and spatial variations of soil erosion in Europe: A study based on erosion plot data. Geomorphology 122, 167-177 Chaves M.M., Maroco J.P. and Pereira J.S. (2003) Understanding plant responses to drought – from genes to the whole plant. Functional Plant Biology 30, 239-264. Clua A.A. and Gimenez D.O. (2003) Environmental factors during seed development of narrow leaved bird’sfoot-trefoil (Lotus tenuis) influences subsequent dormancy and germination. Grass and Forage Science 58, 333338. Copani V., Testa G., Lombardo A. and Cosentino S.L. (2012) Evaluation of populations of Dactylis glomerata L. native to Mediterranean environments. Crop and Pasture Science 63, 1124-1134. Copani V., Cosentino S.L., Testa G., Virgillito S. and Scordia D. (2013) Physiological and productive responses of native perennial grasses to environmental constraints in Mediterranean environments. In: 21st European Biomass Conference and Exhibition, 3-7 June 2013, Copenhagen, Denmark. Cosentino S.L., Cassaniti S., Gresta F., Copani V. and Testa G. (2003) Quantificazione dell’azotofissazione in sulla ed erba medica nei Monti Nebrodi. Rivista di Agronomia 37, 119-127. Cosentino S.L., Copani V., D’Agosta G.M., Sanzone E. and Mantineo M. (2006) First results on evaluation of Arundo donax L. clones collected in Southern Italy. Industrial Crops and Products 23, 212-222. Cosentino S., Patanè C., Sanzone E., Copani V. and Foti S. (2007) Effects of soil water content and nitrogen supply on the productivity of Miscanthus x giganteus Greef et Deu. in Mediterranean environment. Industrial Crops and Products 25, 75-88. Cosentino S., Mantineo M. and Copani V. (2008) Sod seeding and soil erosion in a semi-arid Mediterranean environment of South of Italy. Italian Journal of Agronomy 3, 47-48. Cosentino S., Scalici G., Ambra R. and Virgillito S. (2011) Sistemi colturali ed erosione nella collina interna Siciliana. Atti XL Convegno della Società Italiana di Agronomia. Teramo, 7-9 settembre. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 51 Cosentino S.L., Copani V., Scordia D. and Testa G. (2012) Valutazione di una nuova specie poliennale da biomassa lignocellulosica in relazione a diversi livelli irrigui. XLI Convegno Nazionale della Società Italiana di Agronomia, Bari, 19-21 Settembre 2012, 176-179. Cosentino S., Copani V., Testa G. and Scordia D. (2013) Long-term effects of agronomic practices on soil organic carbon and crop productivity in the internal hills of Sicily. Oral presentation. European Geosciences Union, General Assembly 2013. Vienna (Austria), 7-12 April. Dear B.S. and Roggero P.P. (2003) The present and potential role of perennial grass monocoltures and mixtures with annual legumes in Italian and southern Australian farming systems. In: Bennett S.J. (ed) New perennial legumes for sustainable agriculture. University of Western Australia Press, Crawley, Western Australia. Demdoum S., Munoz F. and Delgado I. (2010) Forage production of a collection of sainfoin over a three year period. Options Méditerranéennes 92,101-104. Doran J.W. (2002) Soil health and global sustainability: translating science into practice. Publications from USDAARS/UNL. Faculty. Paper 181. Doran J.W. and Zeiss M.R. (2000) Soil health and sustainability: managing the biotic component of soil quality. Applied Soil Ecology 15, 3-11. Duffy J.E. (2002) Biodiversity and ecosystem function: the consumer connection. Oikos 99, 201-219. Dumont B., Andueza D., Niderkorn V., Lüscher A., Porqueddu C. and Picon-Cochard C. (2014) A meta-analysis of climate change effects on forage quality in grasslands: perspectives for mountain and Mediterranean areas. Options Méditerranéennes (in press). http://epp.eurostat.ec.europa.eu/portal/page/portal/eurostat/home/ EC (2011) Communication from the Commission to the European Parliament, the Council, the Economic and Social Committee and the Committee of the Regions – our life insurance, our natural capital: an EU biodiversity strategy to 2020. COM (2011) 244 final. Brussels. Finn J.A., Kirwan L., Connolly J., Sebastià M.T., Helgadottir A., Baadshaug O.H., Bélanger G., Black A., Brophy C., Collins R.P., Čop J., Dalmannsdóttir S., Delgado I., Elgersma A., Fothergill M., Frankow-Lindberg B.E., Ghesquiere A., Golinska B., Golinski P., Grieu P., Gustavsson A.-M., Höglind M., Huguenin-Elie O., Jørgensen M., Kadziuliene Z., Kurki P., Llurba R., Lunnan T., Porqueddu C., Suter M., Thumm U. and Lüscher A. (2013) Ecosystem function enhanced by combining four functional types of plant species in intensively managed grassland mixtures: a 3-year continental-scale ﬁeld experiment. Journal of Applied Ecology 50, 365-375. Fontaine J. J. (2011) Improving our legacy: Incorporation of adaptive management into state wildlife action plans. Journal of Environmental Management 92, 1403-1408. Fraser M.D., Speijers M.H.M., Theobald V.J., Fychan R. and Jones R. (2004) Production performance and meat quality of grazing lambs finished on red clover, lucerne or perennial ryegrass swards. Grass and Forage Science 59, 345-356. Henkin Z. and Seligman N.G. (2000) Long term dominance of annual legume species after P application in a Mediterranean rangeland. Options Méditerranéennes, 45, 137-140. Hilhorst H.W.M. (1998) The regulation of secondary dormancy. The membrane hypothesis revisited. Seed Science Research 8, 77-90. IPCC (2013) Fifth assessment Report (AR5). Climate Change 2014: Impacts, Adaptation, and Vulnerability. Vol I and II. Jouven M., Lapeyronie P., Moulin C.H. and Bocquier F. (2010) Rangeland utilization in Mediterranean farming systems. Animal 4, 1746-1757. Koukoura Z. (2007) Natural grasslands – a case study in Greece. Soil Biology 10, 257-269. Kreyling C., Wenigmann M., Beierkuhnlein C. and Jentsch A. (2008) Effects of extreme weather events on plant productivity and tissue die-back are modified by community composition. Ecosystems 11, 752-763. Leible L., Kalber S. and Kappler G. (2005) Development of scenarios for the supply of biomass from agriculture and forestry in two model regions of Baden-Wurttemberg for the production of synthetic liquid fuels-Regional biomass scenarios. Study on behalf of DaimlerChrysler AG, Karlsruhe: Forschungszentrum Karlsruhe, 1-42. (in German.) Lelièvre F., Seddaiu G., Ledda L., Porqueddu C. and Volaire F. (2011) Water use efficiency and drought survival in Mediterranean perennial forage grasses. Field Crop Research 121, 333-342. Lewandowski I., Scurlock J.M.O., Lindvall E. and Christou M. (2003) The development and current status of perennial rhizomatous grasses as energy crops in the US and Europe. Biomass Bioenergy 25, 335-361. Long R.L., Gorecki M.J., Renton M., Scott J.K., Colville L., Goggin D.E., Commander L.E., Westcott D.A., Cherry H. and Finch-Savage W.E. (2014) The ecophysiology of seed persistence: a mechanistic view of the journey to germination or demise. Biological Reviews. doi: 10.1111/brv.12095 Lüscher A., Finn J.A., Connolly J., Sebastià M.T., Collins R.P., Fothergill M., Porqueddu C., Brophy C., Huguenin-Elie O., Kirwan L., Nyfeler D. and Helgadóttir A. (2008) Benefits of sward diversity for agricultural grasslands. Biodiversity, Special Issue on Biodiversity & Agriculture 9, 29-32. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 52 Maestre F.T., Quero J. L., Gotelli N. J., Escudero A., Ochoa V., Delgado-Baquerizo M., García-Gómez M., Bowker Matthew A., Soliveres S., Escolar C., García-Palacios P., Berdugo M., Valencia E., Gozalo B., Gallardo A., Aguilera L., Arredondo T., Blones J., Boeken B., Bran D., Conceição A. A., Cabrera O., Chaieb M., Derak M., Eldridge D. J., Espinosa C. I., Florentino A., Gaitán J., Gatica M. G., Ghiloufi W., Gómez-González S., Gutiérrez J. R., Hernández R. M., Huang X., Huber-Sannwald E., Jankju M., Miriti M., Monerris J., L. Mau R., Morici E., Naseri K., Ospina A., Polo V., Prina A., Pucheta E., Ramírez-Collantes D. A., Romão R., Tighe M., Torres-Díaz C., Val J., Veiga J. P., Wang D. and Zaady E. (2012) Plant species richness and ecosystem multifunctionality in global drylands. Science 335, 214-218. Malhi S.S., Zentner R.P. and Heier K. (2002) Effectiveness of alfalfa in reducing fertilizer N input for optimum forage yield, protein concentration, returns and energy performance of bromegrass-alfalfa mixtures. Nutrient Cycling in Agroecosystems 62, 219-227. Malinowski D.P., Belesky D.P., Kramp B.A., Ruckle J. , Kigel J. and Pinchak W.E (2008) A method to differentiate summer-dormant from summer-active tall fescue and orchardgrass accessions at germination stage. Australian Journal of Agricultural Research 59, 1092-1102. Maltoni S., Molle G., Porqueddu C., Connolly J., Brophy C. and Decandia M. (2007) The potential feeding value of grass-legume mixtures in dry Mediterranean conditions. In: A. Helgadottir and E. Pötsch (eds.). Final Meeting of COST Action 852, Raumberg-Gumpenstein (Austria) 30 August – 3 September 2006, pp. 149-152. Martillotti F., Bartocci S. and Terramoccia S. (1996) Tavole dei valori nutritivi degli alimenti di interesse zootecnico. Progetto editoriale multimediale realizzato dall’INEA nell’ambito del Programma Operativo “Sviluppo della divulgazione agricola e delle attività connesse”, Reg. CEE 2052/88, Obiettivo 1. A cura di Nencioni, M.C., Rubino, R. pp. 88-90. Martınez-Fernández D., Walker D. J., Romero P., Martınez-Ballesta M. C. and Correal E. (2012) The response of the leguminous fodder plant Bituminaria bituminosa to water stress. Journal of Agronomy and Crop Science 198, 442-451. Maughan B., Provenza F.D., Tansawat R., Maughan C., Martini S., Ward R., Clemensen A., Song X., Cornforth D. and Villalba J.J. (2014) Importance of grass-legume choices on cattle grazing behavior, performance and meat characteristics. Journal of Animal Science. doi:10.2527/jas.2013-7297 Monti A., Bezzi G., Pritoni G. and Venturi G. (2008) Long-term productivity of lowland and upland switchgrass cytotypes as affected by cutting frequency. Bioresource Technology, 99, 7425-7432. Norman H.C., Cocks P.C. and Galwey N.W. (2002) Hardseededness in annual clovers: Variations between populations from wet and dry environment. Australian Journal of Agricultural Research 53, 821-829. Norton M.R., Lelievre F. and Volaire F. (2006) Summer dormancy in Dactylis glomerata L., the influence of season of sowing and a simulated mid-summer storm on two contrasting cultivars. Australian Journal of Agricultural Research 57,1267-1277. Norton M.R. and Volaire F. (2012) Summer dormancy in Phalaris aquatica L., the influence of season of sowing and summer moisture regime on two contrasting cultivars. Journal of Agronomy and Crop Science 198, 1-13. Ooi M.K.J. (2012) Seed bank persistence and climate change. Review. Seed Science Research 22, 53-60. OPTIMA project (www.optimafp7.eu) – Optimization of perennial grasses for biomass production. Grant Agreement n. 289642. Osoro K., Ferreira L.M.M., Garcìa U., Rosa Garcia R. and Celaya R. (2012) Diet selection and performance of sheep and goats grazing on different heathland vegetation types. Small Ruminant Research 109, 119-127. Papanastasis V.P. and Papachristou T.G. (2000) Agronomic aspects of forage legumes: management and forage quality. In: Sulas L. (eds). Legumes for Mediterranean forage crops, pastures and alternative uses. Proceedings of the 10th Meeting of the Mediterranean Sub-Network of the FAO-CIHEAM Inter-Regional Cooperative Research and Development Network on Pastures and Fodder Crops, Sassari (Italy), 4-9 April. Cahiers Options Mediterraneennes (France), Plant Production and Protection Div. FAO, Rome (Italy) 45:113-126. Patanè C. and Bradford K.J. (1993) Seed dormancy in two Medicago species typical of natural pastures of Ragusa Plateau. Rivista di Agronomia 4, 412-418. Patanè C. and Gresta F. (2006) Germination of Astragalus hamosus and Medicago orbicularis as affected by seedcoat dormancy breaking techniques. Journal of Arid Environments 67, 165-173. Patanè C., Cosentino S.L. and Copani V. (2008) Hardseededness in two Medicago species as affected by water stress during seed development. Options Mediterranéennes 79, 345-348. Peeters A. (2009) Importance, evolution, environmental impact and future challenges of grasslands and grasslandbased systems in Europe. Grassland Science 55, 113-125. Peeters A., Beaufoy G., Canals R.M., De Vliegher A., Huyghe C., Isselstein J., Jones G., Kessler W., Kirilov A., Mosquera-Losada M.R., Nilsdotter-Linde N., Parente G., Peyraud J.L., Pickert J., Plantureux S., Porqueddu C., Rataj D., Stypinski P., Tonn B., van den Pol - van Dasselaar A., Vintu V. and Wilkins R. (2014) Grassland term definitions and classifications adapted to the diversity of European grassland-based systems. Grassland Science in Europe 19 (this volume). Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 53 Perevolotsky A. (2005) Integrating landscape ecology in the conservation of Mediterranean ecosystems. The Israeli experience. Israel Journal of Plant Science 53, 203-213. Piluzza G., Sulas L. and Bullitta S. (2014) Tannins in forage plants and their role in animal husbandry and environmental sustainability: a review. Grass and Forage Science 69, 32-48. Plan Bleu Centre d’activités régionales (2003) Threats to Soils in Mediterranean Countries. Document Review. Plan Bleu Papers 2. Poirier M., Durand J.L. and Volaire F. (2012) Persistence and production of perennial grasses under water deficits and extreme temperatures: importance of intraspecific vs. interspecific variability. Global Change Biology 18, 3632-3646. Porqueddu C. and Roggero P.P. (1994) Effetto delle tecniche agronomiche di intensificazione foraggera sui fenomeni erosive dei terreni in pendio in ambiente mediterraneo. Rivista di Agronomia 4, 364-370. Porqueddu C., Loi A. and Cocks P.S. (1996) Hardseededness and pattern of hard seed breakdown in Sardinian populations of Medicago polymorpha under field conditions. Journal of Agricultural Science 126, 161-168. Porqueddu C. (2001) Screening germplasm and varieties for forage quality: Constraints and potentials in annual medics. Proceedings of the XIV Eucarpia Medicago spp. Group Meeting on 'Quality in lucerne and medics for animal production'. Zaragoza and Llieida (Spain), 12-15 September 2001, pp. 89-98. Porqueddu C., Ovalle C., Del Pozo A., Fernández F. and Caredda S. (2004) Rehabilitation with annual pasture mixtures in different Mediterranean regions. Proc. of 11th Meeting FAO-CIHEAM 'Rangeland and pastures rehabilitation in Mediterranean areas', Djerba, Tunisia, 29 October-1st November 2002, pp.463-466 Porqueddu C. and Maltoni S. (2007) Biomass production and unsown species control in rainfed grass-legume mixtures in a Mediterranean environment. In: A. Helgadottir and E. Pötsch (eds.). Proceedings of the COST 852 final meeting, 30 August-3 September 2006, Raumberg-Gimpenstein, Austria. 41-44. Porqueddu C. and Gonzàles F. (2006) Role and potential of annual pasture legumes in Mediterranean farming systems. Pastos, XXXVI (2), 125-142. Porqueddu C., Franca A. and Sulas L. (2010) A second generation of pasture legumes: an opportunity for improving the biodiversity in farming systems of Mediterranean basin? Proceedings 13th Meeting of the subnetwork of the FAO-CIHEAM, “The Contributions of Grasslands to the Conservation of the Mediterranean Biodiversity" . Options méditerranéennes, CIHEAM, Series A N. 92, 241-246. Porqueddu C., Dettori G., Falqui A. and Re G.A. (2011) Bio-agronomic evaluation of fifteen accessions within Psoralea complex. Grassland Science in Europe 16, 374-376. Porqueddu C., Sulas L., Re G.A., Sanna F., Franca A. and Melis R.A.M. (2014) Potential use of native Piptatherum miliaceum (L.) Coss. for forage production and bioenergy. Grassland Science in Europe19 (this volume). Prochnow A., Heiermann M., Plöchl M., Linke B., Idler C., Amon T. and Hobbs P. (2009a) Bioenergy from Permanent Grassland - A Review: 1. Biogas. Bioresource Technology 100, 4931-4944. Prochnow A., Heiermann M., Plöchl M., Amon T. and Hobbs P.J. (2009b) Bioenergy from permanent grassland - A review: 2. Combustion. Bioresource Technology 100, 4945-4954. Prochnow A., Heiermann M. and Plöchl M. (2013) Permanent grasslands for bioenergy: factors affecting management and conversion efficiency. The role of grasslands in a green future: threats and perspectives in less favoured areas. Grassland Science in Europe 18, 514-521. Re G. A., Piluzza G., Sulas L., Franca A., Porqueddu C., Sanna F. and Bullitta S. (2014) Condensed tannins accumulation and nitrogen fixation potential of Onobrychis viciifolia Scop. grown in a Mediterranean environment. Journal of the Science of Food and Agriculture 94, 639-645. Real D. and Verbyla A. (2010) Maximizing genetic gains using a “plant” model in the Tedera (Bituminaria bituminosa var. albomarginata and var. crassiuscula) breeding program in Australia. Options Mediterraneennes A, 92:87-96. Reaside M.C., Nie Z.N., Clark S.G., Partington D.L., Behrendt R. and Real D. (2013) Evaluation of tedera [(Bituminaria bituminosa (L.) C.H. Stirton var. albomarginata)] as a forage alternative for sheep in temperate southern Australia. Crop and Pasture Science 63, 1135-1144. Riggi E., Testa G. and Cosentino A. (2006) Yield and relative indices of legume-grass forage mixtures in Mediterranean environment as affected by seeding ratios and legume genotypes. In: Bibliotheca Fragmenta Agronomica - IX ESA Congress. Book of Proceedings. Warsaw Agricultural University (WAU), 4-7 Settembre, Varsavia: Institute of soil Science and Plant Cultivation vol. 11, p. 583-584. PL ISSN: 0860-4088. Rochon J.J., Doyle C.J., Greef J.M., Hopkins A., Molle G., Sitzia M., Scholefield D. and Smith C.J. (2004) Grazing legumes in Europe: a review of their status, management benefits, research needs and future. Grass and Forage Science 59, 197-214. Rockström J., Steffen W., Noone K., Persson Å., Chapin III F.S., Lambin E.F., Lenton T.M., Scheffer M., Folke C., Schellnhuber H.J., Nykvist B., de Wit C.A., Hughes T., van der Leeuw S., Rodhe H., Sörlin S., Snyder P.K., Costanza R., Svedin U., Falkenmark M., Karlberg L., Corell R., Fabry V.J., Hansen J., Walker B., Liverman D., Richardson K., Crutzen P. and Foley J.A. (2009) A safe operating space for humanity. Nature 461, 472-475. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 54 Rosa Garcìa R., Ocharan F.J., Jàuregui B.M., Garcìa U., Osoro K. and Celaya R. (2010) Ground-dwelling arthropod communities present in three types of Cantabrian (NW Spain) heathland grazed by sheep and goat. European Journal of Entomology 107, 219-227. Rosa Garcìa R., Fraser M.D., Celaya R., Ferreira L.M.M., Garcìa U. and Osoro K. (2013) Grazing land and management and biodiversity in the Atlantic European heathlands – a review. Agroforestry Systems 87, 19-43. Rӧsch C., Skarka J., Raab K. and Stelzer V. (2009) Energy production from grassland – Assessing the sustainability of different process chains under German conditions. Biomass and Bioenergy 33, 689-700. Sala O. E., Stuart Chapin F., Armesto J.J., Berlow E., Bloompeld J., Dirzo R., Huber-Sanwald E., Huenneke L.F., Jackson R.B., Kinzig A., Leemans R., Lodge D.M., Mooney H.A., Oesterheld M., LeRoy Poff N., Sykes M.T., Walker B.H., Walker M. and Wall D.H. (2000) Global biodiversity scenarios for the Year 2100. Science 287, 1770-1774. Salis L., Sitzia M., Vargiu M., Mulè P., Re G.A, and Sulas L. (2012) Adaptation of Australian self-reseeding forage legumes to three environments of Sardinia. Options méditerranéennes 102, 265-269. Scordia D., Cosentino S.L. and Jeffries TW. (2013) Effectiveness of dilute oxalic acid pretreatment of Miscanthus x giganteus biomass for ethanol production. Biomass and Bioenergy 59, 540-548. Scordia D., Testa G. and Cosentino S.L. (2014) Perennial grasses as lignocellulosic feedstock for second generation bioethanol production in Mediterranean environment. Italian Journal of Agronomy [In press]doi: 10.4081/ija.2014.581 Spehn E.M., Hector A., Joshi J., Scherer-Lorenzen M., Schmid B., Bazeley-White E., Beierkuhnlein C., Caldeira M.C., Diemer M., Dimitrakopoulos P.G, Finn J.A., Freitas H., Giller P.S., Good J., Harris R., Högberg P., HussDanell K., Jumpponen A., Koricheva J., Leadley P.W., Loreau M., Minns A., Mulder C.P.H., O’Donovan G., Otway S.J., Palmborg C., Pereira J.S., Pfisterer A.B., Prinz A., Read D.J., Schulze E.D., Siamantziouras A.S.D., Terry A.C., Troumbis A.Y., Woodward F.I., Yachi S. and Lawton J.H. (2005) Ecosystem effects of biodiversity manipulations in European grasslands. Ecological Monographs 75, 37-63. Sternberg M., Gutman M., Perevolotsky A., Ungar E.D. and Kigel J. (2000) Vegetation response to grazing management in a Mediterranean herbaceous community: a functional group approach. Journal of Applied Ecology 37, 224-237. Sternberg M., Gutman M., Perevolotsky A. and Kigel J. (2003) Effects of grazing on soil seed bank dynamics: an approach with functional groups. Journal of Vegetation Science 14, 375-386. Sternberg, M., Gishri, N. and Mabjeesh S. (2006) Effects of grazing on functional traits of Bituminaria bituminosa in a Mediterranean grassland. Journal of Agronomy and Crop Science 192, 399-407. Sulas L. (2005) The future role of forage legumes in Mediterranean-climate areas. In: Reynolds S. and Frame J. (eds) Grasslands: Developments Opportunities Perspectives. FAO Science Publishers, Inc Plymouth UK, pp. 2954. Sulas L., Seddaiu G., Muresu R. and Roggero P.P. (2009) Nitrogen fixation of sulla under mediterranean conditions. Agronomy Journal 101, 1470-1478. Testa G., Riggi E. and Cosentino A. (2006) Uptake of biologically fixed nitrogen by grass in grass-legumes binary mixtures. In: Bibliotheca Fragmenta Agronomica - IX ESA Congress. Book of Proceedings. Warsaw Agricultural University (WAU), 4-7 Settembre, Varsavia: Institute of soil Science and Plant Cultivation, vol. 11, p. 427-428. Testa G. and Cosentino A.D. (2009) Biological nitrogen fixation by alfalfa, field bean and subterranean clover. Proceedings of the 16th Nitrogen Workshop 2009: Connecting different scales of nitrogen use in agriculture, p.109-110. Testa G., Gresta F. and Cosentino S.L. (2011) Dry matter and qualitative characteristics of alfalfa as affected by harvest times and soil water content. European Journal of Agronomy, 34, 144-152. Tillman D., Reich P.B. and Knops J.M.H. (2006) Biodiversity and ecosystem stability in a decade long grassland experiment. Nature 441, 629-632. Van Rompaey A., Bazzoffi P., Jones R.J.A and Montanarella L. (2005) Modelling sediment yields in Italian catchments. Geomorphology 65,157-169. Volaire F. and Lelievre F. (1997) Production, persistence, and water-soluble carbohydrate accumulation in 21 contrasting populations of Dactylis glomerata L. subjected to severe drought in the south of France. Australian Journal of Agricultural Research 48, 933-944. Volaire F. and Lelievre F. (2001) Drought survival in Dactylis glomerata and Festuca arundinacea under similar rooting conditions in tubes. Plant and Soil 229, 225-234. Volaire F. and Norton M. (2006) Summer dormancy in perennial temperate grasses. Annals of Botany 98, 927933. Volaire F. (2008) Plant traits and functional types to characterise drought survival of pluri-specific perennial herbaceous swards in Mediterranean areas. European Journal of Agronomy 29, 116-124. Volaire F., Norton M.R. and Lelièvre F. (2009) Summer drought survival strategies and sustainability of perennial temperate forage grasses in Mediterranean areas. Crop Science 49, 2386-2392. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 55 Volaire F., Barkaoui K. and Norton M. ( 2013) Designing resilient and sustainable grassland for a drier future: adaptive strategies, functional traits and biotic interactions. European Journal of Agronomy 52, 81-89. Whitehead D.C. (1995) Grassland Nitrogen. Wallingford UK: CAB International, p. 387. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 56 Theme 1 ‘Climate change: mitigation and adaptation’ Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 57 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 58 Theme 1 invited papers Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 59 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 60 Synergies between mitigation and adaptation to climate change in grasslandbased farming systems Del Prado A.1, Van den Pol-van Dasselaar A.2, Chadwick D.3, Misselbrook T.4, Sandars D.5, Audsley E.5 and Mosquera-Losada M.R.6 1 Basque Centre for Climate Change (BC3), Alameda Urquijo, 4, 4º-1ª /48008 Bilbao, Spain 2 Wageningen UR Livestock Research, 8219PH Lelystad, The Netherlands 3 School of Environment, Natural Resources and Geography, Bangor University, Bangor, Wales, UK 4 Rothamsted Research, North Wyke, Okehampton, Devon, EX20 2SB, UK 5 School of Applied Sciences, Cranfield University, Cranfield, Bedford MK43 0AL 6 Departamento de Producción Vegetal, Universidad de Santiago de Compostela, Spain Corresponding author: agustin.delprado@bc3research.org Abstract Climate change mitigation and adaptation have generally been considered in separate settings for both scientific and policy viewpoints. Recently, it has been stressed (e.g. by the latest IPCC reports) the importance to consider both mitigation and adaptation from land management together. To date, although there is already large amount of studies considering climate mitigation and adaptation in relation to grassland-based systems, there are no studies that analyse the potential synergies and tradeoffs for the main climate change mitigation and adaptation measures within the current European Policy context. This paper reviews which mitigation and adaptation measures interact with each other and how, and it explores the potential limitations and strengths of the different policy instruments that may have an effect in European grassland-based livestock systems. Keywords: Mitigation, adaptation, resilience, climate change, grassland-based, livestock Introduction In the last IPCC report (AR5-WGIII: IPCC, 2014b), for the first time, most of the terrestrial land comprising agriculture, forestry and other land use (AFOLU) was considered altogether. Moreover, it was also highlighted in the AFOLU chapter (IPCC, 2014b) the importance to consider the systemic feedbacks and interactions between mitigation and adaptation options from land management (Separate sub-section: 11.5). In grassland-based systems, however, the potential interactions between mitigation and adaptation options, compared with forest or arable systems, have received much less attention and this has been reduced to changes in carbon (C) stocks and pasture productivity. Changes in biogeochemical cycles (mainly C and N) and water cycles are expected to exert large impacts on livestock productivity and N and C emissions from grassland-based systems (i.e. CO2, CH4 and N2O). Climate change impacts on livestock will include effects of forage and feed quality and productivity, direct impacts of changes in temperature and water availability on animals, and indirectly through livestock disease increase (IPCC, 2014a). However, socio-economic changes are expected to have a still greater effect on mitigation and adaptation potentials (Schmidhuber and Tubiello, 2007). Climate mitigation options in grassland systems mainly include practices that increase soil C stocks or can help to reduce GHG (greenhouse gases) emissions at the soil, feed, animal or manure management level. However, at a wider context, demand-side measures (e.g. human dietary changes, reducing losses and wastes in the agro-food chain) or substitution of fossil fuels by biomass can also play an important role in mitigating climate change. Mitigation options in grassland-based systems need also to be addressed for their potential impact on all other ecosystem/environmental services provided by grasslands for current and future scenarios, as climate regulation is just one of the services amongst a varied list (e.g. food Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 61 production). Mitigation and adaptation in grassland-based systems are closely integrated through a network of feedbacks, synergies and risk of trade-offs. Mitigation measures may also be vulnerable to climate change or there may be possible synergies and trade-offs between mitigation and adaptation options (IPCC, 2014b). Many policies are directly (e.g. Kyoto Protocol) and indirectly (e.g. Common Agricultural Policy: CAP) affecting the potential success of implementing measures that both reduce GHG emissions and help to adapt grassland-based livestock systems to climate change. There is however a need for a more integrated and scientifically-based regulatory approach. Although there are already many studies that have reviewed measures regarding grasslandbased systems on GHG mitigation (e.g. Smith et al., 2007; Verge et al., 2007) or climate change adaptation (e.g. Bryan et al., 2009; Olesen et al., 2011; Tingem et al., 2009), only few sudies have assessed the benefits and trade-offs of their synergistic effects (e.g. Lal et al., 2011). The main objective of this paper is to provide a high-level assessment of synergies and tradeoffs for the main potential climate change mitigation and adaptation measures in grasslandbased livestock systems within the current European Policy context. Climate change mitigation now and in the future in grassland-based livestock systems The main aim of the mitigation options in grassland-based systems is to reduce emissions of CH4 or N2O and/or to increase soil C storage, especially by soil as grasslands account for 75% of C in the terrestrial ecosystems (Lal, 2005; Dresner et al., 2007). Recently, more or less comprehensive reviews on GHG mitigation from grassland-based systems have been produced (e. g. Project ANIMALCHANGE: Van den Pol-van Dasselaar, 2012; UNEP, 2013; Havlík et al., 2014; Del Prado et al., 2013a). Nitrous oxide is formed in the soil through nitrification and denitrification (Wrage et al., 2001) and controlled by a number of site-specific factors, including soil moisture content (Del Prado et al., 2006), temperature (Dobbie et al., 2001) and also, management factors such as fertilizer (Cardenas et al., 2010) and management of soil organic matter content (Mosquera-Losada et al., 2011a) and grazing (Van den Pol-van Dasselaar et al., 2008). Carbon sequestration also depends on edaphoclimatic conditions (Theng et al., 1989), the presence of trees (Mosquera-Losada et al., 2011a) and the organic matter quality and quantity (Mosquera-Losada et al., 2011a and 2011b). Methane can be produced via enteric fermentation, which depends greatly on the level of feed intake, the quantity of energy consumed and feed composition, or can be produced at the manure management level, which increases with temperature, and with increased biodegradability of the manure (Monteny et al., 2006). As highlighted in the last IPCC report on mitigation of climate change in the AFOLU sector, there is an emerging scientific activity on prediction of the likely impact of the climate change on the potential to reduce net GHG emissions (i.e. impacts on N2O and CH4 emissions and on the rates of C sequestration) in the AFOLU sector in general, and in grassland-based systems in particular. Mitigation options available today in the grassland-based farming sector may not be available or as effective with further global warming. Soil C storage has been shown to be vulnerable not only to climate change but more importantly, to changes in the disturbance regime, both natural and human-induced. Land use projections indicate large changes in land use and potentially this change will exacerbate the release of CO2 from soils including grasslands. Increasing temperatures, when water is not limiting, are expected to accelerate soil organic matter (SOM) decomposition rates but result in an increase of C returns through plant residues considering the CO2 fertiliser effect and the lengthening of the growing season (Bindi and Olesen, 2011). Increasing SOM will enhance soil C storage and may also increase above and belowground biomass production or at least improve yield stability (Pan et al., 2009). However, biological processes resulting in N2O emissions (i.e. denitrification) could be Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 62 stimulated by greater SOM. Moreover, increased variability and higher frequency of extreme events will negatively impact soil C storage, by both decreasing production levels and enhancing soil C losses. At the manure level, GHG emission changes are expected in relation to temperatures and sometimes indirect effects driven by changes in the composition of the feed (e.g. digestibility). Main climate change effects on European grassland-based livestock systems During the last century, the climate in Europe has changed more than in other areas of the world (IPCC, 2007). Compared to the pre-industrial era, when the mean annual temperature increased by 0.8 °C globally, it increased by 1.2 °C in Europe. Based on theoretical models, a further increase of 1.0–5.5 °C is expected by the end of the twenty-first century (Christensen et al., 2007). The increase in temperature has been most apparent in mountainous areas such as the Alps, which tend to have high biodiversity and where temperature increased by 2 °C during the twentieth century (EEA, 2009). This is twice the average temperature increase for the northern hemisphere. In addition, the quantity and distribution of precipitation have also changed in Europe during the twentieth century. Although there has been a 20% decrease in rainfall in southern Europe, there has been a 10–40% increase in rainfall in northern Europe. Furthermore, an increase in the frequency of extreme weather events is predicted across the European continent (EEA, 2008). The most important impacts of climate change on grassland-based farming systems in Europe are expected to be through changes in pasture productivity and forage quality, therefore potentially affecting the duration of the grass growing season and the forage supply to ruminants. An example about how different intra-year temperature and precipitation regimes affect total and also seasonal distribution of pasture was found by Mosquera-Losada and González-Rodríguez (1998) in dairy systems. This paper highlights the importance of having flexible grazing systems, which affects annual and instantaneous stocking rate. The intensification of the hydrological cycle caused by more intensive rainfall and longer dry periods is expected to result in higher risks of soil erosion and nutrient leaching in currently wet temperate climates. Changes in precipitation patterns in drier areas will lead to higher dependency on stored soil moisture storage and seasonality for supporting grass growth. Elevated concentrations of CO2 may also increase water use efficiency through reduced plant stomata aperture, but increase run-off risk through reduced plant transpiration, thus resulting in excess water at the land surface (Betts et al., 2007). Biological and physical processes regulating nutrient cycling in grassland-based farming systems may actually be more sensitive to extreme events rather than changes in average climatic conditions. For Europe, an increased risk of low forage production in summer due to severe summer droughts events is expected to be offset by the appearance of new opportunities for forage production in other seasons due to warming effects. For southern latitudes, higher evapotranspiration rates will negatively affect grass yield and the period of grass growth will be shorter unless the grassland is irrigated. In general, poorer grass nutritional qualities, e.g. lower grass digestibility, can also be expected. Potential synergies and tradeoffs between strategies to adapt to climate change and measures to mitigate GHG emissions GHG emission mitigation choices may further enhance or reduce resilience to climate variability and change in terms of ecosystem goods and services provision, and thus influence the potential of grassland-based systems to adapt to climate change. climate change may affect climate adaptation and mitigation strategies through changes in feed supply, animal diet composition, animal and plant breeding, soil management, enhancement of floral biodiversity and via more resistant and resilient production systems against climate change (e.g. agroforestry systems). Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 63 Climate change affecting feed supply (grazing and forage) Spring growth, provided that water resources for grass growth are available, and winter production may benefit from mild climate conditions. This can contribute to improvement in the farm´s degree of forage autonomy and security of livestock systems when facing more hazardous climate conditions (e.g. summer droughts) through the extension of the grazing season and the reduction of forage requirements (Graux et al., 2013). For example, forage resources usually stored for over-wintering livestock could be partially redistributed in summer to deal with increased risk of forage deficits (Graux et al., 2013). However, for southern latitudes and dates getting closer to the XXII century (e.g. UK: Del Prado et al., 2009) the projections suggest that grazing activity will be constrained due to too high temperatures and excessive drought in Europe. Extending grazing seasons by e.g. the presence of shelter/shade belts of trees would reduce the wind speed and therefore evapotranspiration (ETP). The presence of trees at low density would also increase the duration of the growing season due to their presence, which may partly reduce GHG emissions (Tackas and Frank, 2009) through improving soil N recovery by trees, but may also become hot-spots for N2O from overlapping urine patches, and soils could become eroded due to the action of hooves in camping areas used by livestock for shelter/shade. Extending the grazing season in some cases may also be limited by the bearing capacity of the soil driven by good soil structure degradation (e.g. poaching caused by trampling cattle or/and severe summer droughts, etc.) and therefore, it may, in some cases be impractical. Hence, avoiding compaction by traffic, tillage (Pinto et al., 2004) and grazing livestock (De Klein and Ledgard, 2005) may help to maintain grasslands in good conditions and also to reduce N2O emissions. N2O 1.40 1.20 1.00 0.80 NO3/ha CH4-enteric 2020 0.60 0.40 2020ADAPT NH3 CH4-manure SW Figure 1. Comparison between adapted (extending one month grazing) and un-adapted typical dairy farms in the south west England (2020) for GHG, NO3- leaching and NH3 emissions. Values for the adapted scenario <1 indicate a reduction in emissions (adapted from Misselbrook et al., 2013). Poorer grass nutritional qualities, e.g. lower grass digestibility, will lead to higher CH 4 emissions from enteric fermentation of cattle (Hart et al., 2009). Although if lower forage quality would mean reduced livestock ‘yields’ and/or quality then market reaction would be to import feeds. For those systems extending the grazing season (e.g. Figure 1: UK: Misselbrook et al., 2013), smaller volumes of manure storage could lead to a reduction in CH4 emissions from manure handling. However, the rise in average and extreme temperatures, should the frequency of manure removal from the storage remain unchanged or no additional structural measures are implemented to either lower the manure storage temperature or to aerate the slurry, could increase the amount of CH4 per kg of volatile solid of manure and therefore, lead to similar total CH4 emissions from manure management. Moreover, even if the CH4 emissions Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 64 from manure management were smaller, this value (in CO2 equivalents) would be partially offset or accentuated by an increase in N2O emissions promoted by the increase in grazing activity. Climate change affecting feed supply (purchased feed and different crop rotations) Changes in grassland productivity will affect either animal productivity or the amount of purchased feed required (Mosquera-Losada and González-Rodríguez, 1998). For semi-arid regions (e.g. south of Europe), a reduction in annual grass productivity will lead to lower animal productivity or will have to be compensated with a larger share of imported feedstock with associated monetary and environmental costs, which may translate into a potential loss of resilience in grassland-based livestock systems. In some of these regions tree presence to feed animals with acorns, could supply part of these needs (Moreno and Pulido, 2009). An increase in the establishment of rotations best suited to the area or crop rotations with legumes annual crops (Bryan et al., 2011) may also occur as an adaptation strategy. Some crops that currently grow mostly in southern Europe will become more suitable further north or in higher altitudes areas in the South. For example, forage maize, may become more common across in the boreal regions of Europe. Maize forage, however, tends to make the management system less flexible to inter-annual temperature/precipitation variations. Moreover, maize area cannot be used for grazing during the summer or autumn if no grass is available during this period. In contrast, grass areas can be open or harvested for silage if a restriction or an excess of grass production happens (Mosquera-Losada and GonzálezRodríguez, 1998). At the animal level, forage maize animal intake is generally promoted at the expense of grass due to a better balance between protein intake and soluble carbohydrates (e.g. through increasing starch concentration in the diet), which additionally may help to increase animal energy use efficiency and decrease CH4 emissions per kg DM intake. However, this CH4 reduction may be offset by larger N2O emission losses and a larger CO2 release of converting some grassland into arable land (Vellinga and Hoving, 2011). Conversely, converting crops to pasture has been found to reduce N2O emissions (Eagle et al., 2012) and also contribute to sequester soil C, especially in the first years after conversion. Leguminous species are well adapted to future conditions of climate change (Kreyling et al., 2012) considering that their optimum temperature is higher than non-leguminous crops and that they also have more positive responses to elevated concentrations of CO2 (Soussana and Lüscher, 2007) than non-legume species. In a situation with a larger share of mixed legume/grass pastures, in addition to presenting climate adaptive advantages over conventional pastures, these systems have lower requirements for N fertilizer through the use of biological N fixation of nodules on the roots of legumes, which would lead to energy savings and GHG emissions reductions from both fertilizer production and use (Del Prado et al., 2011; Zhang et al., 2013). Climate change affecting feed supply (use of by-products and alternative forages) Different by-products from agricultural, forestry, agro-industry and bioenergy activities can also be used for feeding ruminants as an adaptive response to forage supply seasonal constraints. Rinne et al. (2012) reviewed different by-products (e.g. camelina meal, tomato pomace) that are currently underutilized but that could potentially be used as feed for low input and organic dairy production systems. Those practices are currently used as part of some livestock systems at a regional level (Correal et al., 2009). These by-products vary in their geographical availability, nutritional value, their effect on rumen CH4 and N excretion (i.e. effect on GHG mitigation) and have logistic-related challenges. Environmentally speaking (e.g. GHG intensity), the use of some of these by-products as animal feed may not always be the best option in comparison with their use in bioenergy or for soil improvement. In this sense, Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 65 removal of crop residues from cropping systems for use in bioenergy, if this means that soil C contents are being depleted (e.g. straw: Liu et al., 2014), will bring large risks of negative impacts on adaptation measures and potentially, small or negligible positive effects on the reduction of net GHG emissions. Mitigation and adaptation conflicts may therefore appear as one chooses a particular use of the by-product or another. Other alternative forage supply may include tree leaves and shrubs, particularly in small-scale livestock farms with dry to semi-arid climates. Such species can alleviate feed shortages, or even fill feed gaps in the winter and especially in the summer, when grassland growth is limited or dormant due to unfavourable weather conditions (Papanastasis et al., 2008). Although some species have leaves with a low content of CP and a high content of fiber and contain high levels of secondary compounds such as tannins, alkaloids, saponins and oxalates which reduce the nutritive value of poor-quality diets, some of these compounds (e.g. condense tannins), when improved temperate forages are fed, can also have substantial benefits for ruminant productivity (i.e. reducing CH4) and health (Waghorn and McNabb, 2003). Moreover, there are also other species (e.g. Morus alba, Fraxinus excelsior, Betula alba) whose young leaves are rich sources of protein and fibre and generally used in the past to feed animals before modern techniques like fertilizer were used. Changes in fertiliser management, diet and genetics to increase N use efficiency Manipulating the diet (e.g. feeding nitrification inhibitors: Ledgard et al., 2008 or salt supplementation) during the grazing period has also been proposed as a means to reduce N2O emissions. Improving fertiliser efficiency, optimising methods, timing and rates of applications (Brown et al., 2005), using NH4+-based fertilisers rather than nitrate-based ones (e.g. Dobbie and Smith, 2003) and employing nitrification chemical inhibitors (e.g. Zaman et al., 2009) may also have a role in both mitigation (i.e. reduction of direct and indirect soil N2O emissions) and adaptation (through a better N use efficiency at the soil-plant level). New traits in animals and grasses may also assist farmers to both mitigate and adapt to climate change. Del Prado and Scholefield (2008), for example, using a farm modelling approach, evaluated the scope for different animal and plant genetic traits, some existing and other theoretical, to help reduce GHG emissions on UK dairy farms. More efficient animals in utilising N (Alford et al., 2006) have also been proposed to decrease the impact of urinary N during grazing. Some of the traits, e.g. improved N use efficiency in grasses (e.g. high sugar grasses: Wilkins et al., 2000) could actually be both potentially useful for climate mitigation and may also promote Climate adaptation as they may reduce GHG emissions from urinerelated N2O emissions and improve the quality of the forage, which may be beneficial in future scenarios where climate has a detrimental effect on grass nutritional properties. Soil management, plant biodiversity and new plant breeds to improve system resilience against environmental stress conditions and prevent soil erosion Other strategies to both mitigate and adapt to climate change may involve management practices that target directly to the soil, both improving the capacity to store water and to prevent soil erosion. By increasing the ability of soils to hold soil moisture and to better withstand erosion by enriching biodiversity through more diversified cropping systems, grassland systems will be able to sequester more soil C and also to better resist extreme events such as droughts and /or floods, both of which are projected to increase in frequency and severity in future warming climates (Rosenzweig and Tubiello, 2007). The measures for the conservation of soil moisture may also include changes in tillage practices. Reduced tillage, for example, increases the resilience to climate change through improved soil fertility and increased capacity for water retention in the soil. This improvement is expected in the long-term productivity potential when tillage is reduced (Olesen et al., 2011). Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 66 The reduced tillage at pasture reseeding promotes C sequestration and preservation in pastures and is considered to be more effective under conditions of water deficit (Alvaro -Fuentes et al., 2011). It leads also to significant savings in CO2 emissions produced by machinery. However, the impact on N2O emissions under different conditions is unclear (Estavillo et al., 2002; Pinto et al., 2004). Nitrous oxide emissions appear to be strongly influenced by soil water content immediately after nitrogen fertilization (Del Prado et al., 2006). In view of this dominant effect of a particular soil moisture level coinciding with tillage and fertilization, it is key to find the best timing for the renewal of pasture. Velthof et al. (2010), for example, considering average Dutch climatic season conditions, suggest that this pasture renewal should take place in spring rather than fall because Dutch autumn, compared with spring is generally wetter and N uptake by the reseeded grass is lower. The effect of reduced tillage has also been observed by increasing the periods between which a pasture is renewed. Vellinga et al. (2004), for example, found that although tillage increased N2O and CO2 in the intensively managed pastures in the studied year, in the long run, the renovation of pastures was more important to prevent deterioration in pasture quality and thus, to prevent from soil loss and large productivity losses. For areas which are subject to severe or extremely severe environmental stress conditions the establishment of a community of pastures formed by species that ensure ecological stability, both in ecosystem resistance and resilience, is key as an adaptation measure to climate change (Volaire et al ., 2014). Additionally the species composition of the pasture is expected to undergo changes, as for example, warming will favour C4 species over C3 species (Howden et al., 2008). Biodiversity should act as a safeguard of ecosystem functioning, thus promoting a more stable ecosystem to avoid fluctuations arising from adverse climatic fluctuations (Volaire et al., 2014). Promoting biodiversity could also have an effect on the mitigation potential of pastures and in some occasions of rumen methane. Considering that N remains one of the main elements that determines the diversity of plants, the application of less fertilizer should be a requirement to increase diversity in different floral species in grasslands (Mountford et al., 1993). This reduced input fertilizer would be necessarily associated with lower emissions of N2O per ha and potentially a greater amount of C accumulated in the soil. New grass breeds have already been tested to improve water use efficiency. For example, McLeod et al. (2013) tested in the UK a novel grass Festulolium hybrid capable to reduce runoff by 40-50% compared to a leading UK nationally recommended L. perenne cultivar and F. pratensis over a two year field experiment. The rapid growth and turnover of roots in the hybrids resulted in greater soil water storage capacity in the plots with observed lower rainfall runoff. This may, in turn, have significant effects on N2O emissions and soil C storage. Agroforestry systems Agroforestry is a well-founded example of mitigation and adaptation synergy (e.g. IPCC, 2014b; EU forest strategy: EU, 2013) since trees planted and grassland soils sequester C and tree and grassland products provide livelihood to communities, especially during drought years (Verchot et al., 2007). Agroforestry in general and silvopastoral systems in particular lead to greater resilience to climate change due to improved soil conditions and management efficiency in water use (Kumar et al., 2011). Its characteristics are able to reduce evapotranspiration and thus improve the maintainability of soil water (Tackas and Frank, 2009). These practices also have a great potential to offset GHG emissions through the sequestration of C in soil and tree biomass and avoiding the release of NO 3 leaching (indirect N2O emissions) (Rigueiro et al., 2009). Moreover, these systems also improve the N use efficiency of the system and offer large resilience against climate change stress conditions through the reduction of temperature of the system (Rigueiro et al., 2009). It can also help reduce erosion of adjacent fields handled more intensely (Verge et al., 2007). Policy implications Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 67 Climate mitigation policies and measures may exhibit synergies and risk trade-offs with climate adaptation (Bates et al., 2008). However, policies of mitigation and adaptation are often being considered in separate settings, resulting in potential conflicts. An integrated adaptation and mitigation framework is important to ensure that trade-offs between the two are minimized and synergies encouraged (Wreford et al., 2010). However, this is not easy as mitigation and adaptation may occur simultaneously, but differ in their spatial, timing and geographical characteristics (Smith and Olesen, 2010). Amongst the number of policies affecting climate change mitigation and adaptation in grassland-based systems in Europe, the newly reformed EU Common Agricultural Policy (CAP), in principle, has made a decisive move towards promoting a greener and climatically friendlier EU agricultural sector. The new CAP has introduced direct payments associated with different practices that, in some cases, are expected to enhance GHG mitigation and adaptation to climate change. Specifically, new payments within Pillar I associated with the diversification of crop rotations, maintaining permanent pastures and ensuring Ecological Focus Areas should be targeting, in part, climate-friendly or climate-smart agriculture. Permanent pasture maintenance is an important way to prevent N emissions through avoiding plough management and conversion of permanent grasslands into arable lands (EU Regulation 1307/2013). Leguminous species are mentioned explicitly in the areas of ecological interest (N-fixing species) but there is no special plan for their promotion. Other practices, such as those mentioned in previous sections, grazing, for example, is encouraged directly through the support of agroforestry systems and forests with fire risk areas (through the Rural Development Programme (Pilar II)), avoiding huge amounts of C release and through cross-compliance via for example promotion of good standards for animal welfare. Floristic biodiversity should also be encouraged but are not explicitly mentioned within the new PAC to safeguard ecosystem functioning against adverse climatic fluctuations. The replacement of permanent grasslands by forage maize is no longer allowed by the CAP as penalties are included in the last CAP if destruction above 5% is present. The new CAP, however, does not explicitly address the worrying import of feed in grassland-based intensive systems. In fact, in some countries, this is still indirectly encouraged through additional payments to more intensive systems. The CAP has been blamed for distortion of global markets in this sense. Khatun (2012) points at the absence of tariffs for animal feed as a key driver for fueling EU cheap imports of animal feed from Latin America and consequently, for the effect on land use, land use change, and forestry (LULUCF) outside of the EU and, thereby preventing from a huge potential for mitigating climate change by reducing emissions from deforestation and forest degradation (REDD+ programme) outside Europe. Policies, hence, can create both positive and perverse incentives for mitigation or adaptation (Wreford et al., 2010). A number of recent studies (e.g. Lassaletta et al., in press in Spain) suggest that many European Countries, either assisted by specific regulations (e.g. Kyoto, CAP) or fuelled by market pressures, are displacing large amounts of GHG emissions from their national primary sectors (e.g. grasslandbased farming) to other countries via agricultural goods importing. For example, cattle farming in Europe, whose feed system was traditionally based mainly on-farm forage (e.g. grass) production, in the recent decades has shift to heavily depend on cheap imported protein (e.g. soybean) from South America, resulting in a reduction of GHG emissions in the European GHG inventories but more than offsetting this potential mitigation by a consequential increase of GHG emissions by mainly land use change in South America. Much of these emissions are produced in non- Annex B countries and consequently, C leakage is being produced in Europe. Displacing agricultural productivity may indeed be an adaptation choice for countries, but this is certainly against securing Food Sovereignty and therefore, this jeopardizes the future resilience of the European food system. For example, if the conversion of annual crops to pasture is accompanied by a demand to grow annual crops outside Europe, this would not Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 68 represent a net mitigation but merely a shift in emissions and, in some cases, this would be an example of C and / or N2O leakage. Furthermore, a large part of the mitigation potential of grasslands is also subject to challenges in relation to effectiveness over different time-scales. For example, whereas certain types of mitigation activities (e.g. N2O reduction from reduced N fertilization, CH4 reduction in the rumen through animal diet changes, bioenergy) are effectively permanent since the emissions, once avoided, cannot be re-emitted (IPCC, 2014b), some activities that helped to sequester C (e.g. reducing tillage), can be reversible and non-permanent. Moreover, some of these practices to sequester soil C may also be constrained due to the saturation of grassland soils to sequester C indefinitely. Therefore protecting the large C stocks in grasslands should be an important management and policy target, rather than necessarily trying to increase the C stocks (Smith, in press) since it is easier and faster for soils to lose C that it is for them to gain C (Johnson et al., 2009). Mitigation options for any of the GHG gases must also be tailored to the specific soil, climatic and production system conditions (Bustamante et al., in press). There will be very few strategies that are universally applicable for all systems and under any climatic circumstances. All mitigation options certainly affect and are affected by the cycles of C and N. Nitrogen and C cycles are also currently decoupled for most intensive grassland systems (Soussana and Lemaire, 2014), these systems release by ruminants bound-C digestible as CO2 and CH4, and return digestible N in high concentrations (urine patches). The coupling / decoupling of C and N makes an added difficulty to analyze the effectiveness of mitigating measures as sometimes some of the measures that increase soil C storage, for example, addition of manure, can also increase losses of N2O by increasing soluble C in the system. In contrast, measures that promote the reduction of N2O can cause a net loss of C from the system through increased soil respiration (Scholefield et al., 2005). Moreover, some of the mitigation methods lead to pollution swapping (e.g. NH3 volatilization , leaching of NO3-), and losses in biological diversity and / or productivity (Del Prado and Scholefield, 2008), and also can cause numerous interactions between mitigation measures so that their effect in the case of using multiple measures simultaneously are not necessarily additive (Del Prado et al., 2010). Also, the reference unit to which GHG emissions relate within the CAP is commonly the forage area, which may not, in some cases, coincide with the preferred reference unit used by the agroindustry (C footprint or GHG per unit of product). The emphasis therefore seems to have been diverted from what the consumers and markets dynamics are essentially promoting. Preferably, one should consider more than one reference unit or functional unit (e.g. per hectare and per unit of output) at the same time to avoid conflicts of interpretation about what is true / false mitigation (Del Prado et al., 2010). Agroindustry generally uses the Life Cycle Assessment (LCA) as the methodology choice in order to report GHG emissions from the full cycle of the production of a food. A key element still unsolved is the way LCA assigns different amounts of GHG emissions to different goods according to its market-based value. Given a specific policy context, the farmer may choose among the most cost-effective and easier-toadopt options. Ecosystem services which currently have no market value may become valuable also in monetary terms in the future. Some farmers may, therefore, in the future also seek to maximize the ecosystem service value. Alternative methodologies are already suggesting that, for products that are produced through extensive and in some cases greener conditions, these emissions should be split according to not only market but non-market (e.g. ecosystem services) values (Ripoll-Bosch et al., 2013) as well. An important issue that may not be reflected in the new CAP and in other policies is the alarming growth tendency of feeding ruminants (e.g. dairy cattle: Del Prado et al., 2013b) with a greater amount of feed ingredients which could be used directly in the human food chain (e.g. cereals) (Eisler et al., 2014). This relates very significantly to the potential competitive Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 69 advantage that pasture-based livestock (ruminants generally are able to use low-quality plant biomass and that is inedible to humans) might have over another livestock (e.g. monogastric animals). Policies therefore should be useful to overturn this trend. Additionally, non-climate policies and regulations are already in place for other environmental issues (e.g. water quality, NH3) and have consistently assisted in reducing GHG emissions from the agricultural sector (e.g. EU: Velthof et al., 2014). Nitrate leaching losses, however, are expected to increase for numerous areas that are already constrained in their nutrient use by the EU Nitrates Directive (Anon, 1991) in Europe and for feed commonly used in animal diets, for example wheat (Olesen et al., 2007). This increase in NO3- leaching may trigger more stringent regulations and hence affect animal productivity and GHG emissions, which may challenge climate change adaptation also from a policy perspective. Research-oriented policies should and already have a role, for example, in encouraging the study of new grass varieties that can better adapt to climate change and also present properties that can increase the efficiency of use of nutrients and energy in the soil-plant-animal system. It is therefore imperative that all the policies, from the local to the global levels, are appropriately integrated with the policies relating to climate change, bioenergy, food, waste, research and health in order to promote a net reduction of GHG from the standpoint not only of production (supply) but also of demand in order to avoid possible market distortions and maladaptation practices at all levels. Acknowledgements This study was performed thanks to the support of the livestock and grassland theme (LiveM) of the FACCE-JPI knowledge hub and project MACSUR. MACSUR is funded through national funding bodies as part of the EU FACCE-JPI (Joint Programming Initiative for Agriculture, Climate Change, and Food Security). AGFORWARD and CICYT (Spanish Ministry) projects are also acknowledged. References Alford A. R., Hegarty R. S., Parnell P. F., Cacho O. J., Herd R. M. and Griffith, G. R. (2006) The impact of breeding to reduce residual feed intake on enteric methane emissions from the Australian beef industry. Australian Journal of Experimental Agriculture 46, 813-820. Álvaro‐Fuentes J., Easter M., Cantero‐Martinez C. and Paustian K. (2011). Modelling soil organic carbon stocks and their changes in the northeast of Spain. European Journal of Soil Science 62(5), 685-695. Anon. (1991) Council Directive concerning the protection of water against pollution caused by nitrates from agricultural sources. Official Journal of the European Communities (91/676/EEC) 8572/91. Bates B., Kundzewicz Z. W., Wu S. and Palutikof J. (2008) Climate change and water. Intergovernmental Panel on Climate Change (IPCC). Bell M. J., Eckard R. J. and Cullen B. R. (2012) The effect of future climate scenarios on the balance between productivity and greenhouse gas emissions from sheep grazing systems. Livestock Science 147, 126-138. Betts R. A., Boucher O., Collins M., Cox P. M., Falloon P.D., Gedney N., Hemming D. L., Huntingford C., Jones C. D., Sexton D. M. H. and Webb M.J. (2007) Projected increase in continental runoff due to plant responses to increasing carbon dioxide. Nature 448, 1037-1041. Bindi M. and Olesen J. E. (2011) The responses of agriculture in Europe to climate change. Regional Environmental Change 11(1), 151-158. Brown L., Scholefield D., Jewkes E.C., Lockyer D.R. and Del Prado A. (2005) NGAUGE: A decision support system to optimize N fertilization of British grassland for economic and environmental goals. Agriculture Ecosystems and Environment 109, 20-39. Bryan E., Deressa T. T., Gbetibouo G. A. and Ringler C. (2009) Adaptation to climate change in Ethiopia and South Africa: options and constraints. Environmental Science and Policy 12(4), 413-426. Bryan B. A., King D. and Ward J. R. (2011) Modelling and mapping agricultural opportunity costs to guide landscape planning for natural resource management. Ecological Indicators 11, 199-208. Bustamante M., Robledo-Abad C., Harper R., Mbow C., Ravindranat N. H., Sperling F., Haberl H., De Siqueira PintoA. and Smith P. (in press) Co-benefits, trade-offs, barriers and policies for greenhouse gas mitigation in the agriculture, forestry and other land use (AFOLU) sector. Global Change Biology. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 70 Cardenas L. M., Thorman R., Ashlee N., Butler M., Chadwick D., Chambers B., Cuttle S., Donovan N., Kingston H., Lane S., Dhanoa M. S. and Scholefield D. (2010) Quantifying annual N2O emission fluxes from grazed grassland under a range of inorganic fertiliser nitrogen inputs. Agriculture Ecosystems and Environment 136, 218226. Christensen J. H., Hewitson B., Busuioc A., Chen A., Gao X., Held R., Jones R., Kolli R. K., Kwon W. K., Laprise R., Magana Rueda V., Mearns L., Menendez C. G., Räisänen J,. Rinke A., Sarr A,. Whetton P., Arritt R., Benestad R., Beniston M., Bromwich D., Caya D., Comiso J., de Elia R. and Dethloff K. (2007) Regional climate projections, Climate Change, 2007: The Physical Science Basis. Contribution of Working group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change, University Press, Cambridge, Chapter 11. Correal E., Erena M., Ríos S., Robledo A. and Vicente M. (2009) Agroforestry systems in Southeastern Spain. In: Rigueiro-Rodríguez A., McAdam J. and Mosquera-Losada M.R. (eds) Agroforestry in Europe. Springer, pp. 183-210. De Klein C. M. and Ledgard S. (2005). Nitrous Oxide Emissions from New Zealand Agriculture - key Sources and Mitigation Strategies. Nutrient Cycling in Agroecosystems 72, 77-85. Del Prado A., Chadwick D., Cardenas L., Misselbrook T., Scholefield D. and Merino P. (2010) Exploring systems responses to mitigation of GHG in UK dairy farms. Agriculture, Ecosystems and Environment 136 (3-4), 318332. Del Prado A., Crosson P., Olesen J.E. and Rotz A. (2013a) Whole-farm models to quantify greenhouse gas emissions and their potential use for linking climate change mitigation and adaptation in temperate grassland ruminant-based farming systems. Animal 7 (suppl.2), 373-385. Del Prado A., Mas K., Pardo G. and Gallejones. P. (2013b) Modelling the interactions between C and N farm balances and GHG emissions from confinement dairy farms in northern Spain. Science of The Total Environment 465, 156-165. Del Prado A., Merino P., Estavillo J. M., Pinto M. and González-Murua, C. (2006) N2O and NO emissions from different N sources and under a range of soil water contents. Nutrient cycling in agroecosystems 74(3), 229-243. Del Prado A., Misselbrook T., Chadwick D., Hopkins A., Dewhurst R.J., Davison P., Butler A., Schröder J. and Scholefield D. (2011) SIMSDAIRY: A modelling framework to identify sustainable dairy farms in the UK. Framework description and test for organic systems and N fertiliser optimisation. Science of the Total Environment 409, 3993-4009. Del Prado A. and Scholefield D. (2008). Use of SIMS DAIRY modelling framework system to compare the scope on the sustainability of a dairy farm of animal and plant genetic-based improvements with management-based changes. Journal of Agricultural Science 146, 1-17. Del Prado A., Shepherd A., Wu L., Topp C., Moran D., Tolkamp B. and Chadwick D. (2009) Modelling the effect of climate change on environmental pollution losses from dairy systems in the UK. BC3 Working Paper Series 2009-07. Basque Centre for Climate Change (BC3). Bilbao, Spain. Dresner S., Ekins P., McGeevor K. and Tomei J. (2007) Forest and climate change global understandings and possible responses. In: Freer-Smith P.H., Boradmeadow M.S.J. and Lynch JM (eds.) Forestry and climate change CAB International, Wallingford, UK, pp. 38-48. Dobbie K. E and Smith K. A. (2001) The effects of temperature, water‐filled pore space and land use on N2O emissions from an imperfectly drained gleysol. European Journal of Soil Science 52(4), 667-673. Eagle A. J., Olander L. P. and Donald L. S. (2012) Chapter Three - Greenhouse Gas Mitigation with Agricultural Land Management Activities in the United States-A Side-by-Side Comparison of Biophysical Potential. Advances in Agronomy. Academic Press. Eisler M. C., Lee M. R., Tarlton J. F., Martin G. B., Beddington J., Dungait J.A., Greathead H., Liu J., Mathew S., Miller H., Misselbrook T., Murray P., Vinod V. K., Van Saun R. and Winter M. (2014) Agriculture: Steps to sustainable livestock. Nature 507 (7490), 32. Estavillo J. M., Merino P., Pinto M., Yamulki S., Gebauer G., Sapek A. and Corre W. (2002) Short term effect of ploughing a permanent pasture on N2O production from nitrification and denitrification. Plant and Soil 239(2), 253-265. European Union (EU) (2013) Communication from the Commission on an EU Forest Action Plan COM 2013 http://ec.europa.eu/agriculture/forest/strategy/communication_en.pdf Graux A. I., Bellocchi G., Lardy R. and Soussana J. F. (2013) Ensemble modelling of climate change risks and opportunities for managed grasslands in France. Agricultural and Forest Meteorology 170, 114-131. Hart K. J., Yañez-Ruiz D. R., Duval S. M., McEwan N. R. and Newbold C. J. (2008) Plant extracts to manipulate rumen fermentation. Animal Feed Science Technology 147, 8-35. Havlík P., Valin H., Herrero M., Obersteiner M., Schmid E., Rufino M. C., Mosnier A., Thornton P. K., Battcher H., Conant R. T., Frank S., Fritz S., Fuss S., Kraxner F. and Notenbaert A. (2014) Climate change mitigation through livestock system transitions. Proceedings of the National Academy of Sciences 111, 3709-3714. Howden S. M., Crimp S. J., and Stokes C. J. (2008) Climate change and Australian livestock systems: impacts, research and policy issues. Animal Production Science 48(7), 780-788. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 71 Johnson A. E., Poulton P.R and Coleman K. (2009) Soil organic matter: Its importance in sustainable agriculture and carbon dioxide fluxes. Advances in Agronomy 101, 1-57. Intergovernmental Panel on Climate Change (IPCC) (2007) Climate change: the physical science basis. In: Solomon S., Qin D., Manning M., Chen Z., Marquis M., Averyt K.B., Tignor M. and Miller H.L. (eds.), Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change Chapter 2. Cambridge University Press, United Kingdom. 2007. Intergovernmental Panel on Climate Change (IPCC) (2014a) Climate Change 2014: Impacts, Adaptation, and Vulnerability. IPCC WGII AR5 Food security and Food production systems. http://ipccwg2.gov/AR5/images/uploads/WGIIAR5-Chap7_FGDall.pdf . Intergovernmental Panel on Climate Change (IPCC) (2014b) Climate Change 2014: Mitigation of Climate Change. IPCC WGIII AR5. Agriculture, Forestry and Other Land Use (AFOLU). http://report.mitigation2014.org/drafts/final-draft-postplenary/ipcc_wg3_ar5_finaldraft_postplenary_chapter11.pdf Khatun K. (2012) Reform or Reversal; Implications of the Common Agricultural Policy (CAP) on Land Use, Land Use Change and Forestry (LULUCF) in Developing Countries. Conservation Letters 5. 99-106. Kreyling J., Thiel D., Simmnacher K., Willner E., Jentsch A. and Beierkuhnlein C. (2012) Geographic origin and past climatic experience influence the response to late spring frost in four common grass species in central Europe. Ecography 35(3), 268-275. Kumar A. B. M. and Nair P. K. R. (2011) Carbon Sequestration Potential of Agroforestry Systems: Opportunities and Challenges. Springer. Lal R. (2005) Forest soils and C sequestration. Forest Ecology and Management 220, 242-258. Lal R., Delgado J. A., Groffman P. M., Millar N., Dell C. and Rotz A. (2011) Management to mitigate and adapt to climate change. Journal of Soil and Water Conservation 66(4), 276-285. Lassaletta L., Aguilera E., Sanz-Cobeña A., Pardo G., Billen G., Garnier J. and Grizzetti, B. (in press) Leakage of nitrous oxide emissions within the Spanish agro-food system in 1961–2009. Mitigation and Adaptation Strategies for Global Change 1-20. Ledgard S.F., Menneer J.C., Dexter M.M., Kear M.J., Lindsey S., Peters J.S. and Pacheco D. (2008) A novel concept to reduce nitrogen losses from grazed pastures by administering soil nitrogen process inhibitors to ruminant animals: A study with sheep. Agriculture, Ecosystems and Environment 125, 148-158. Liu C., Lu M., Cui J., Li B. and Fang C. (2014) Effects of straw carbon input on carbon dynamics in agricultural soils: a meta‐analysis. Global Change Biology 20, 1366-1381. Macleod C. J. A., Humphreys M. W., Whalley W. R., Turner L., Binley A., Watts C. W., Skot L., Joynes A., Hawkins S., King I. P., O'Donovan S. and Haygarth, P. M. (2013) A novel grass hybrid to reduce flood generation in temperate regions. Scientific reports, 3. Misselbrook T., Del Prado A. and Chadwick D. (2013) Opportunities for reducing environmental emissions from forage-based dairy farms. Agriculture and Food Science 22, 93-107. Monteny G. J., Bannink A. and Chadwick D. (2006) Greenhouse gas abatement strategies for animal husbandry. Agriculture, Ecosystems and Environment 112(2), 163-170. Moreno G. and Pulido F. (2009) The functioning, management and persistence of Dehesas. In: RigueiroRodríguez A., McAdam J. and Mosquera-Losada M.R. (eds) Agroforestry in Europe. Springer: pp. 127-160. Mosquera-Losada M. R., Freese D., Rigueiro-Rodríguez A. (2011a) Carbon Sequestration Potential of Agroforestry Systems: Opportunities and Challenges, Advances in Agroforestry 8, 43-59 Mosquera-Losada M. R., Morán-Zuloaga D. and Rigueiro-Rodríguez A. (2011b) Effects of lime and sewage sludge on soil, pasture production and tree growth in a six year old Populus Canadensis Moench silvopastoral system. Journal of Plant Nutrition and Soil Science 174, 145-153. Mosquera-Losada M. R., González-Rodríguez A. (1998) Effect of annual stocking rates in grass and maize+rye system on production by dairy cows. Grass and Forage Science 53 (2), 95-108. Mountford J.O., Lakhani K.H. and Kirkham F.W. (1993) Experimental assessment of the effects of nitrogen addition under hay-cutting and aftermath grazing on the vegetation of meadows on a Somerset peat moor. Journal of Applied Ecology 30, 321-332. Olesen J. E., Trnka M., Kersebaum K. C., Skjelvåg A. O., Seguin B., Peltonen-Sainio P., Rossi F., Kozyra J. and Micale F. (2011) Impacts and adaptation of European crop production systems to climate change. European Journal of Agronomy 34, 96-112. Pan G., Smith P. and Pan W. (2009) The role of soil organic matter in maintaining the productivity and yield stability of cereals in China. Agriculture, Ecosystems and Environment 129(1), 344-348. Papanastasis V. P., Yiakoulaki M. D., Decandia M. and Dini-Papanastasi O. (2008) Integrating woody species into livestock feeding in the Mediterranean areas of Europe. Animal Feed Science and Technology 140(1), 1-17. Pinto M., Merino P., Del Prado A., Estavillo J. M., Yamulki S., Gebauer G. and Oenema O. (2004) Increased emissions of nitric oxide and nitrous oxide following tillage of a perennial pasture. Nutrient Cycling in Agroecosystems 70(1), 13-22. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 72 Rinne M., Dragomir C., Kuoppala K., Marley C., Smith J. and Yañez-Ruiz D. (2012) Novel and underutilized feed resources – potential for use in organic and low input dairy production. In: Rahmann G. and Godinho D. (eds.) Agriculture and Forestry Research (362), pp. 404-407. Rigueiro-Rodríguez A., McAdam J., Mosquera-Losada M. R. (2009) Agroforestry in Europe. Springer, pp. 127160. Ripoll-Bosch R., de Boer I. J. M., Bernues A. and Vellinga T. V. (2013) Accounting for multi-functionality of sheep farming in the carbon footprint of lamb: A comparison of three contrasting Mediterranean systems. Agricultural Systems 116, 60-68. Rosenzweig C. and Tubiello F. N. (2007) Adaptation and mitigation strategies in agriculture: an analysis of potential synergies. Mitigation and Adaptation Strategies for Global Change 12(5), 855-873. Schmidhuber J. and Tubiello F. N. (2007) Global food security under climate change. Proceedings of the National Academy of Sciences 104(50), 19703-19708 Scholefield D., Jarvis S. C., Brown L., Del Prado A., Hopkins A. and Cardenas L. (2005) Feed back and FeedForward Interactions between Climate Change and Grassland-Based Agriculture. Avoiding Dangerous Climate Change. Proceeding of the International Symposium on Stabilisation of Greenhouse Gases, Met Office, Exeter, UK, 1-3 February 2005. Smith P. (in press). Do grasslands act as a perpetual sink for carbon? Global Change Biology. Smith P., Martino D., Cai Z., Gwary D., Janzen H., Kumar P, McCarl B., Ogle S., O’Mara F., Rice C., Scholes B., Sirotenko O., Howden M., McAllister T., Pan G., Romanenkov V., Schneider U. and Towprayoon (2007) Policy and technological constraints to implementation of greenhouse gas mitigation options in agriculture. Agriculture, Ecosystems and Environment 118(1), 6-28. Smith P. and Olesen J. (2010) Synergies between the mitigation of, and adaptation to, climate change in agriculture. The Journal of Agricultural Science 148, 543-552. Soussana, J.-F. & Lemaire, G. (2014) Coupling carbon and nitrogen cycles for environmentally sustainable intensification of grasslands and crop-livestock systems. Agriculture, Ecosystems & Environment, 190, 9-17. Soussana J. F. and Lüscher A. (2007) Temperate grasslands and global atmospheric change: a review. Grass and Forage Science 62(2), 127-134. Takács V. and Frank N. (2009) The traditions, resources and potential of forest growing and multipurpose shelterbelts in Hungary. In: Rigueiro-Rodríguez A, McAdam J, Mosquera-Losada M. R. (eds) Agroforestry in Europe. Springer, pp. 415-438. Theng B. K. G., Tate K. R., Sollins P., Moris N., Madkarni N. and Tate R. (1989) Constituents of organic matter in temperate and tropical soils. In: Coleman DC, Oades JM, Uehara G (eds.) Dynamics of soil organic matter in tropical ecosystems. College of Tropical agriculture and Human Resources, University of Hawaii, Honolulu, pp. 5-32. Tingem M., Rivington M. and Bellocchi G. (2009) Adaptation assessments for crop production in response to climate change in Cameroon. Agronomy for sustainable development 29(2), 247-256. UNEP (2013). Drawing down N2O to protect Climate and the Ozone layer. A UNEP synthesis report. United Nations Environment Programme (UNEP), Nairobi, Kenya. http://www.unep.org/publications/ebooks/UNEPN2Oreport/ Van den Pol-van Dasselaar A. (2012). Qualitative overview of mitigation and adaptation options and their possible synergies & trade-offs. Deliverable 8.1 of EU project AnimalChange, 32 p. Van den Pol-van Dasselaar A., Vellinga T. V., Johansen A. and Kennedy E. (2008) To graze or not to graze, that’s the question. Grassland Science in Europe 13, 706-716. Vellinga T. V., Van den Pol-van Dasselaar A. and Kuikman P. J. (2004) The impact of grassland ploughing on CO2 and N2O emissions in the Netherlands. Nutrient Cycling in Agroecosystems 70(1), 33-45. Vellinga T. V. and Hoving I. E. (2011) Maize silage for dairy cows: mitigation of methane emissions can be offset by land use change. Nutrient cycling in agroecosystems 89(3), 413-426. Velthof G. L., Lesschen J. P., Webb J., Pietrzak S., Miatkowski Z., Pinto M., Kros J. and Oenema O. (2014)The impact of the Nitrates Directive on nitrogen emissions from agriculture in the EU-27 during 2000–2008. Science of The Total Environment 468, 1225-1233. Velthof G. L., Hoving I. E., Dolfing J., Smit A., Kuikman P. J. and Oenema O. (2010). Method and timing of grassland renovation affects herbage yield, nitrate leaching, and nitrous oxide emission in intensively managed grasslands. Nutrient Cycling in Agroecosystems 86, 401-412. Verchot L., M. van Noordwijk S., Kandji T., Tomich C., Ong A., Albrecht J., Mackensen C., Bantilan K., Anupama and Palm C. (2007) Climate change: linking adaptation and mitigation through agroforestry. Mitigation and Adaptation Strategies for Global Change 12, 1381-1386. Verge X. P. C., De Kimpe C. and Desjardins R. L. (2007) Agricultural production, greenhouse gas emissions and mitigation potential. Agricultural and Forest Meteorology 142(2), 255-269. Volaire F., Barkaoui K. and Norton M. (2014) Designing resilient and sustainable grasslands for a drier future: Adaptive strategies, functional traits and biotic interactions. European Journal of Agronomy 52, 81-89. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 73 Waghorn G. C. and McNabb W. C. (2003) Consequences of plant phenolic compounds for productivity and health of ruminants. Proceedings of the Nutrition Society 62, 383-392. Wilkins P.W., Allen D.K. and Mytton L. R. (2000) Differences in the nitrogen use efficiency of perennial ryegrass varieties under simulated rotational grazing and their effects on nitrogen recovery and herbage nitrogen content. Grass and Forage Science 55, 69-76. Wrage N., Velthof G. L., Van Beusichem M. L. and Oenema O. (2001) Role of nitrifier denitrification in the production of nitrous oxide. Soil Biology and Biochemistry 33(12), 1723-1732. Wreford A., Moran D. and Adger N. (2010) Climate Change and Agriculture: Impacts, Adaptation and Mitigation. OECD Publishing, Paris, France. Zaman M., Saggar S., Blennerhassett J. D. and Singh J. (2009) Effect of urease and nitrification inhibitors on N transformation, gaseous emissions of ammonia and nitrous oxide, pasture yield and N uptake in grazed pasture system. Soil Biology and Biochemistry 41, 1270-1280. Zhang W. F., Dou Z., He P., Ju X.T., Powlson D., Chadwick D., Norse D., Lu Y. L., Zhang Y., Wu L., Chen X. P, Cassman K.G. and Zhang F.S. (2013). New technologies reduce greenhouse gas emissions from nitrogenous fertilizer in China. PNAS 110, 8375-8380. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 74 The role of grassland in mitigating climate change Soussana J.-F.1,2, Klumpp K.1 and Ehrhardt F.2 1 INRA, Grassland Ecosystem Research, UR874, Clermont Ferrand, France. 2 INRA, Paris, France Corresponding author: Jean-Francois.Soussana@paris.inra.fr Abstract Grassland management has a large potential to mitigate livestock greenhouse gas (GHG) emissions at a low (or even negative) cost, by combining a moderate intensification and the restoration of degraded pastures. A synthesis of eddy flux covariance data shows, on average 213 site years, a mean net carbon storage (NCS) equal to 76 ± 11 gC m-2 yr-1 indicating a significant carbon sequestration in European grasslands. Through statistical modelling we show that this carbon sink activity is largely controlled by climatic and by management factors, with additional drivers from soil and vegetation types. Simple calculations show that pasture carbon neutrality (in CO2 equivalents, accounting for enteric CH4 and soil N2O emissions) can be obtained below a critical stocking rate (SR*). When climate is optimal for grassland canopy photosynthesis (10 °C and 1200 mm yr-1), SR* equals 0.85 and 1.2 LSU ha-1 over a 200 days grazing season without and with manure application, respectively. A low herbage use efficiency (ratio of intake to above-ground net primary productivity) comprised between 0.2 and 0.4 without and with manure application, respectively, is required to reach SR*. Suboptimal climate conditions lead to lower SR* and herbage use efficiency values without manure application, while manure supply moderates this decline. In contrast to manure application, mineral N fertilizer supply leads to minor changes in SR* values. Climate change affects the grassland carbon sink which is not permanent. A short-term change in temperature and precipitation has large implications for the GHG balance of temperate pastures. An exponential rise in GHG emissions per head is simulated with warming and a decline in precipitation affecting a pasture managed at the critical stocking rate under the current climate. The implications of these findings for grassland management are discussed. Keywords: Climate change; Pasture; Livestock; Carbon; Greenhouse Gas; Soil. Introduction Since the industrial revolution, cumulated anthropogenic CO2 emissions to the atmosphere reached 545 ± 85 Gigatons C (1 Gt C = 109 metric tons C). Land use change (including deforestation, afforestation and reforestation) contributed to one third of this amount. About half of the anthropogenic emissions since 1750 was removed from the atmosphere by sinks, and stored in the natural carbon cycle reservoirs. Vegetation biomass and soils not affected by land use change and land degradation stored 150 ± 90 Gt C and the ocean reservoir stored approximately the same amount (IPCC, 2013). The current mean global temperature is slightly above the temperature range experienced during the Holocene, which has seen the onset and expansion of agriculture since ca. 10,000 yrs BP (Marcott et al., 2013). By the end of the 21st century, the biosphere will experience unchartered conditions with a temperature rise between 1.5 and 4.5 °C compared to 1980-1999 and CO2 concentrations in the range 450-1000 ppm (IPCC, 2013). The entire agriculture, forestry and land use sector (AFOLU) contributes directly to 24 percent of total anthropogenic emissions, a value which has declined since 2005 given the reduction in tropical deforestation and in the intensity of GHG emissions per unit of agricultural production (IPCC, 2014). In a separate assessment, the FAO estimated greenhouse gas (GHG) emissions from livestock supply chains (from land use for feed production to the processing and transport of animal products) to represent globally 7.1 Gigatons of CO2 equivalents or 14.5 percent of Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 75 global anthropogenic GHG emissions. The main sources of emissions identified by the FAO are related primarily to the production and processing of animal feed: this corresponds to 45 percent of total emissions, 9 of which are related to the expansion of grazing and crop areas at the expense of forests. Next comes methane emissions from the digestive process in ruminant animals (39 percent), followed by emissions from manure, at 10 percent. The remainder comes from the processing and transportation of animal products (FAO, 2013). The grassland biome covers about one-quarter of the earth’s land area (Ojima et al., 1993). Except within eco-geographical regions where vegetation is maintained by climate and soil factors at herbaceous stage, most of the grasslands around the world are the result of livestock management avoiding encroachment by shrubs and trees (Lauenroth, 1979; Lemaire et al., 2005). At the global scale, grasslands were estimated to be a net C sink of about 0.5 PgC per year (Scurlock and Hall, 1998), but with considerable uncertainty. Grassland ecosystems hold large C reserves, mostly in soil organic matter. Historically, some of these soils have lost one-half to two-thirds of the original top soil organic carbon (SOC) pool with a cumulative loss of 3–4 tons C/ha. Especially in drylands, the depletion of soil C is accentuated by soil degradation and exacerbated by land misuse and soil mismanagement (Lal, 2004). Restoration of degraded lands and grazing land management have been shown to be key options for GHG mitigation in the AFOLU sector (IPCC, 2014) and the global soil organic carbon sequestration potential is estimated to be 0.01 to 0.3 Gt C/year for permanent pastures, which could potentially offset up to 4% of the global GHG emissions (Lal, 2004). Follett and Schuman (2005) reviewed grazing land contributions to C sequestration worldwide using 19 regions. A positive relationship was found, on average, between the C sequestration rate and the animal stocking density, which is an indicator of the pasture primary productivity. Based on this relationship, they estimate a 200 Megatons SOC sequestration per year on 3.5 billion ha of permanent pasture worldwide. Using national grassland resource dataset and NDVI (Normalized Difference Vegetation Index) time series data, Piao et al. (2009) estimated that C stocks of China’s grasslands increased over the past two decades by 117 and 101 g C/m2 per year. In Europe, Schulze et al. (2009) inferred a net C sink in grasslands of 57±34 gCm−2 yr−1 from a small sample of flux tower net CO2 exchange measurements, completed by C imports/exports at each site to estimate the carbon balance. When accounting for emissions of non- CO2 greenhouse gases (GHGs) such as methane (CH4) from grazing animals and nitrous oxide from soil nitrification/ denitrification, the European grasslands were estimated to be nearly neutral for their radiative forcing, with a net balance of −14±18 g CO2-Ceq m−2 y−1 (Schulze et al., 2010). Therefore, non CO2 emissions associated to grassland management by herbivores tend to offset the carbon sink which is currently observed in European grasslands. After presenting the grassland carbon cycle, we review the evidence for a carbon sink in European grasslands and address drivers and trade-offs with non CO2 emissions. Finally we discuss the risks of soil carbon losses associated to an increased climatic variability. 1. The carbon cycle in managed grasslands While there has been steady C accumulation in soils of many ecosystems over millennia (Schlesinger, 1990), it is usually thought that soil C accumulation capacity is limited (Six et al., 2002). Therefore, in a steady-state, non-disturbed soils should have attained C balance after several centuries (Lal, 2004). However, net primary productivity and soil respiration are currently affected by climate change in most regions of the world (Nemani et al., 2003), which implies that soil C stocks are unlikely to have reached equilibrium (Soussana et al., 2010c). The potential for sequestrating C in deep soil layers is also considered large (long residence time), but owing to the low influx of C within these horizons this process remains slow. The process-based ORCHIDEE-GM model was used to simulate the net carbon balance of Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 76 grasslands since 1900. Simulations show an increase in grassland carbon stocks over the last decades, with half of this effect attributed to changes in grassland management (reduction in livestock numbers) and the remainder explained by warming and rising atmospheric CO2 which have both contributed to increase grassland productivity (Chang et al., 2014). Soil C sequestration is reversible and C can be rapidly lost through a number of processes such as soil disturbance, vegetation degradation, fire, erosion, nutrient shortage and water deficit. Therefore, agricultural practices like ploughing, which mix soil layers and break soil aggregates, accelerate top soil organic C decomposition (Paustian et al., 1998, Conant et al., 2001). Changes in SOC through time are non-linear after a change in land use or in grassland management. A simple two parameters exponential model has been used to estimate the magnitude of the soil C stock changes, showing that C is lost more rapidly than it is gained after a change in land use (Soussana et al., 2004). As a result of periodic tillage and re-sowing, short-duration grasslands tend to have a potential for soil C storage intermediate between crops and permanent grasslands. Part of the additional C stored in the soil during the grassland phase is released when the grassland is ploughed up. The mean C storage increases in line with prolonging the lifespan of covers, that is, less frequent ploughing (Soussana et al., 2004). Under intensive grazing, up to 60% of the above-ground dry-matter production is ingested by domestic herbivores (Lemaire and Chapman, 1996). However, this percentage can be much lower under extensive grazing. The largest part of the ingested C is digestible and, hence, is respired shortly after intake. The non-digestible C (25 to 40% of the intake according to the digestibility of the grazed herbage) is returned to the pasture in excreta (mainly as faeces). The nature, frequency and intensity of disturbance play a key role in the C balance of grasslands. In a cutting regime, more than 80% of the above-ground primary production is harvested and exported as hay or silage, but part of these C exports may be compensated by organic C imports through farm manure and slurry application. Off-site C sequestration occurs whenever more manure C is produced by then returned to a grassland plot. The sum of on- and off-site C sequestration reached 129, 98 and 71 g C/m2 per year for grazed, cut and mixed European grasslands on mineral soils, respectively, however with high uncertainty (Soussana et al., 2010c). Long-term field observations show that even when plant material is incorporated in large amounts, the soil C content does not necessarily increase (Ammann et al., 2007). Recent results suggest that N-deficiency may lead to a soil C loss though competition between SOC decomposing and storing microbes (Fontaine and Barot, 2005). Accordingly, long-term C storage in terrestrial ecosystems depends also on the ability to sequester nutrients, which explains the lower C sequestration in grasslands of low productivity and fertility (Franzluebbers et al., 2007). Temperate grasslands have often been intensified by combinations of (i) an increased primary production through an improvement of the N-P-K status of vegetation, (ii) an increased stocking density for converting more efficiently herbage production into animal products, (iii) sowing, or over-sowing, of improved grass and legume species. Intensification has three contrasting effects for the carbon cycle of grasslands (Soussana and Lemaire, 2014): first, an increase in the net primary productivity; second, a decline in the amounts of organic carbon returned to the soil (Soussana et al., 2007); third, a possible decline in the turnover of soil organic matter when nutrients are in ample supply for soil microbes (reduced priming effect, Fontaine et al., 2007; 2011). Depending on the balance of these effects, the impacts on the soil carbon balance may vary. Grassland intensification also leads to increased emissions of N2O from fertilizers and biological N fixation, and to increased methane emissions from enteric fermentation. In comparison to an unmanaged control pasture, doubling the animal stocking density and supplying mineral N fertilizers led to increased net GHG emissions per unit area at an upland permanent pasture site in France (Allard et al., 2007). Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 77 However, during dry years, the moderately intensive grassland was more resilient in terms of carbon storage, emitted less GHGs, and provided increased cattle liveweight gains (Klumpp et al., 2011). Therefore, a moderate intensification of permanent pastures could provide an interesting combination of mitigation and adaptation. Management practices that enhance C sequestration in temperate grasslands were summarized in several reports (e.g. Soussana et al., 2004, 2010, Pellerin et al., 2013): - Grazing management (extension of grazing season, strip grazing, etc) systems that maximizes production, and increases carbon inputs and sequester carbon, while reducing GHG emissions, - Sowing improved species can lead to increased production through species that are better adapted to local climate, more resilient to grazing, more resistant to drought and able to enhance soil fertility (e.g. N-fixing crops), - Direct inputs of water, fertilizer and organic matter to enhance water and N balances, plant productivity and carbon inputs. However, inputs of water, N and organic matter all tend to require energy and can each enhance fluxes of N2O, which are likely to offset carbon sequestration gains, - Restoring degraded lands enhances production in areas with low productivity, increasing carbon inputs and sequestering soil organic carbon, - Including grass in the rotation cycle on arable lands can increase production return organic matter (when grazed as a forage crop), and reduce disturbance to the soil through tillage. Thus, integrating grasses into crop rotations can enhance carbon inputs and reduce decomposition losses of carbon, with benefits for carbon sequestration. 2. Methods for assessing the carbon balance of a grassland Two methods can be used for measuring the carbon balance of a grassland field: direct measurements of soil organic carbon stock change; carbon flux measurements allowing calculation the carbon balance. We briefly review below these two methods. Soil organic carbon stocks A number of studies have analysed effects of grassland and rangeland management on SOC stocks. Most studies concern only the top-soil (e.g. 0 to 30 cm), although C sequestration or loss may also occur in deeper soil layers (Fontaine et al., 2007). It is often assumed that impacts of management are greatest at the surface and decline with depth in the profile (Ogle et al., 2004). The uncertainties concerning the estimated values of C storage or release after a change in grassland management are still very high (estimated at 25 g C/m2 per year). Data from the National Soil Inventory of England and Wales obtained between 1978 and 2003 (Bellamy et al., 2005) show that C was lost from most top soils across England and Wales over the survey period. Nevertheless, rotational grasslands gained C at a rate of ca. 10 g C/m 2 per year. The Countryside Surveys of Great Britain are ecological assessments in UK that have taken place since 1978 (Firbank et al., 2003). In this survey, significant increases in soil C concentration, ranging from 0.2 to 2.1 g/kg per year, were observed in both fertile and infertile grasslands (CLIMSOIL, 2008). In Belgium, grasslands were reported either to be sequestering C in soils at rates of 22 or 44 g C/m2 per year (Lettens et al., 2005a; Goidts and van Wesemael, 2007, respectively), or losing C at 90 g C/m2 per year on podzolic, clayey and loam soils (Lettens et al., 2005b). However, soil bulk density was estimated from pedo-transfer functions in these studies, which adds to the uncertainty, as a small change in bulk density can result in a large change in stock of SOC (Smith et al., 2007). Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 78 Carbon fluxes in grassland ecosystems An alternative to the direct measurement of C stock changes in grasslands is to measure the net balance of C fluxes (net C storage, NCS) exchanged at the system boundaries. This approach provides a high temporal resolution and changes in C stock can be detected within one year. In contrast, direct measurements of stock change require several years or several decades to detect significant effects, given the high variability among samples. The net carbon storage (NCS) is the arithmetic sum of the C fluxes crossing the boundary of the field investigated: i) trace gases exchanged with the atmosphere (i.e. CO2, CH4, volatile organic compounds, VOC, and emissions during fires), ii) organic C imports (manures) and exports (harvests, animal products), iii) dissolved C lost in waters (dissolved organic and inorganic C) and lateral transport of soil C through erosion (Eq. 1): NCS = (NEE - FCH4-C - FVOC - Ffire) + (Fmanure - Fharvest - Fanimal-products) - (Fleach + Ferosion) (Eq. 1) where NEE is the net ecosystem exchange of CO2 between the ecosystem and the atmosphere, which is here conventionally positive for a C gain by the ecosystem. FCH4-C, FVOC and Ffire are trace gas C losses from the ecosystem (g C/m2 per year). Fmanure, Fharvest and Fanimal-products are lateral organic C fluxes (g C/m2 per year) which are either imported or exported from the system. Fleach and Ferosion are organic (and/or inorganic C losses in g C/m2 per year) through leaching and erosion, respectively. Nevertheless, depending on the system studied and its management, some of these fluxes can be neglected for NCS calculation. For instance, fire emissions by grasslands are very low in temperate regions like Europe (i.e. below 1 g C/m2 per year over 1997-2004), while they reach 10 and 100 g C/m2 per year in Mediterranean and in tropical grasslands, respectively (Van der Werf et al., 2006). Erosion (Ferosion) is also rather insignificant in permanent grasslands (e.g. in Europe), but can be increased by tillage in the case of sown grasslands. The global map of Ferosion created by Van Oost et al. (2007) indicates that grassland C erosion rates are usually below 5 g C/m2 per year, even in tropical dry grasslands (Van Oost et al., 2008). The total dissolved C loss by leaching was estimated by Siemens (2003) and Janssens et al. (2003) at 11±8 g C/m2 per year for Europe. This flux tends to be highly variable depending on soil (pH, carbonate) and climate (rainfall, temperature) factors and it could reach higher values in wet tropical grasslands, especially on calcareous substrate. VOC emissions by grassland systems are increased in the short-term by cutting and tend to be higher with legumes than with grass species (Davison et al., 2008). However, these C fluxes are usually small and can easily be neglected. Therefore, with temperate managed grasslands equation 1 can be simplified as (Allard et al., 2007): NCS = (NEE - FCH4-C) + (Fimport - Fexport - Fanimal-products) - Fleach (Eq. 2) With the advancement of micrometeorological studies of the ecosystem-scale exchange of CO2 (Baldocchi and Meyers, 1998), eddy flux covariance measurement techniques have been applied to measure NEE in grasslands and wetlands. Ruminant’s belched CO2 (digestive + metabolic CO2) emission at grazing (Pinares-Patino et al., 2007) is a component of NEE. FCH4 is the sum of the CH4-C flux exchanged (methane emission or oxidation) between the soil and the atmosphere and of the methane emission by the enteric fermentation of ruminants at grazing. At one site, in an upland pasture in Central France, soil carbon stock change was directly measured after 4 and 8 yrs of continuous eddy flux covariance CO2 measurements. Soil coring indicated, on average, a 35 % higher NCS than the eddy flux covariance estimate of NCS (Katja Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 79 Klumpp, unpublished). Therefore, direct soil measurements confirmed the occurence of soil carbon sequestration which was first determined from the eddy flux covariance measurements. 3. The net carbon balance of European grasslands A network of grassland sites equipped for eddy flux covariance measurements of CO2 exchanges with the atmosphere has been established in a range of national and EC funded research projects, starting with EC FP5 (Greengrass project, see Soussana et al., 2007; Carbomont project), followed by FP6 (CarboEurope IP) and by the newly established ICOS infrastructure (Integrated Carbon Observation System). Through collaborations with these projects in the EC FP7 AnimalChange project, we have collected data from 39 sites spanning the European continent and including grasslands, wetlands and moorlands with a diversity of managements (grazing, cutting, abandoned, with or without inorganic and organic N fertilization). Each site has run, over at least two years, eddy flux covariance measurements of CO2 exchanges with the atmosphere and measurements of organic C imports (manures) and exports (harvests, animal products). The annual carbon balance of these sites has been calculated according to Equation 2 and analyzed in order to determine the drivers of carbon sequestration across European grasslands. These sites are from 15 European countries, spanning highly contrasting regions from the Arctic to the Mediterranean and from the Atlantic to central Europe. Each site has been measured between 1 and 11 years, resulting in a total of 213 site years. Data include the site latitude and longitude, the vegetation type with three categories (permanent and sown grasslands, wetlands), the management type (grazing, cutting, mixed or abandoned), the annual N fertilizer supply (inorganic and organic), the annual means of air temperature and of precipitations and the annual Net Ecosystem Exchange (NEE) of CO2. In addition, the lateral fluxes of organic carbon induced by herbage harvests and by manure supplies have been recorded at all sites. Moreover, with grazed sites carbon intake and carbon emissions to the atmosphere as methane from enteric fermentation have been estimated from the animal type, live weight and mean annual stocking density based on IPCC Tier 1 methodology. Finally, the exports of C in animal products (meat and milk) have been estimated as in Soussana et al. (2010c). On average of the 213 site years data, the gross primary productivity (GPP, i.e. photosynthesis) reached 1218 ± 42.8 gC m-2 yr-1 (mean ± s.e.) with a mean net carbon storage (NCS) equal to 76 ± 11 gC m-2 yr-1 showing a significant carbon sequestration in European grasslands. According to a one-Sample Signed Rank Test , there is a statistically significant difference (P <0.001) between the NCS median and zero. The 95% range of the NCS median is 38 to 81 gC m-2 yr-1(Tab 2). Approximately one fourth of the site years NCS had negative values, showing net carbon release by the grasslands. Kruskal-Wallis One Way Analysis of Variance on Ranks did not indicate a statistically significant difference in NCS across grassland types and between cut only and grazed only pastures. Climate was a strong driver of GPP, with maximal GPP (GPPmax = 2250 gC m-2 yr-1) reached at a mean annual temperature of 10.0±0.5 °C and at a mean annual precipitation of 1240±80 mm yr-1. A Lorentzian model fitted to these climatic drivers explained 52% of the annual GPP variance across sites and years. There was no significant effect of management factors on annual net carbon storage, possibly because the management factors were confounded with climate factors. On average, net carbon sequestration reached 9.0±0.02% of GPP in the absence of a lateral flux of organic carbon and of N fertilizer supply (i.e. unmanaged grassland). Consistent with a previous report (Soussana et al., 2007), at a given GPP there was a highly significant (P<0.0001) increase in NCS with N fertilizer supply and with the net import of organic carbon (Lc, organic C balance between manure supply and exports by harvest and intake of digestible carbon, in kg C ha-1 yr-1). Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 80 NCS = (f2+kN.Ns).GPP + kC.Lc (Eq. 3) , with: Lc = Manure-Dig.Intake-Harvest Where, Ns is the N fertilizer supply (kg N ha-1 yr-1); f2, kN and kC are numerical constants and Dig is the digestibility of the ingested DM. Hence, there is a clear trade-off between C sequestration and herbage use by cutting and grazing in grasslands. This trade-off is stronger at cutting since non-digestible carbon is returned to the soil during grazing. In unmanaged grasslands, Eq. 3 shows that NCS is a constant fraction (f2 = 0.09) of GPP. In grazed only pastures, NCS is reduced by animal intake and is increased by manure supply. DM intake is a fraction of ANPP, the above-ground net primary productivity. This fraction (f1) is the herbage use efficiency. In turn, ANPP is a fraction f0 (one third on average) of GPP. Hence: NCS = ((f2+kN.Ns) /f0.f1).Intake+kC.(Manure-Dig.Intake) (Eq. 4) 4. Can we create carbon neutral pastures in Europe? For a grazed and unmanaged pasture developed on mineral soils, the IPCC Tier 1 method (IPCC, 2006) shows that emissions of CH4 and N2O are directly proportional to the intake of dry-matter by the grazing livestock. However, the Tier 1 method does not account for soil organic carbon stock changes (NCS) in pastures. The net GHG balance (kgCO2 equivalents ha1 yr-1) of a pasture can therefore be corrected taking into account NCS as calculated by Eq. 2: GHG = Intake.(fN. EN2O.wN2O + Dig. ECH4. wCH4) -NCS. MCO2 (Eq. 5) Intake is the annual DM intake by the grazing livestock (kg DM ha-1 yr-1), EN2O and ECH4 are the N2O and CH4 emission factors at grazing calculated from default IPCC Tier 1 values. wN2O and wCH4 are the warming potential of N2O and CH4 relative to CO2 per unit weight on a 100 yrs time horizon (298 and 25, respectively, IPCC, 2006). MCO2 is the molar weight ratio of CO2 to C (44/12). fN is the fraction of N in herbage DM. Combining equations 3 and 4: GHG = Intake.(fN.EN2O.wN2O+Dig.ECH4.wCH4-((f2+kN.Ns)/(f0.f1)-kC.Dig).fC.wCO2)+kc.Manure (Eq. 6) When there is no manure applied, Eq. 6 simplifies and the herbage use efficiency (f1*) for which GHG equals 0 (i.e. carbon neutral pasture) does not vary with intake and can be calculated as: f1*= ((f2+kN.Ns)/f0-kC.Dig).fC.wCO2 / (fN.EN2O.wN2O+Dig.ECH4.wCH4) (Eq.7) For this critical herbage use efficiency (f1*), intake can be calculated from GPP. Moreover, since GPP has been fitted to the mean annual temperature (T) and mean annual precipitation (P) using a Lorentzian model (GPP = GPPmax / f(T).f(P)), GPP can be estimated from climate conditions. Hence, assuming cattle grazing with a fixed DM intake per head (20 kg DM intake per day and per livestock unit, LSU), a critical stocking rate (SR*, LSU ha-1) leading to zero carbon emissions by a pasture can be calculated. These calculations for carbon neutral pastures were made considering three management options: i) no fertilization; ii) pastures supplied with mineral N fertilizers, ii) pastures supplied with manure (calculated from Eq. 6). The corresponding results are shown in Figure 1 for a range of T and P conditions covering the climatic conditions experienced by grasslands in Europe. Without N fertilizer supply, the critical cattle stocking rate (SR*) under optimal climate conditions can reach 0.85 LSU/ha over a grazing period of 200 days per year. With less favorable climate conditions, SR* declines down to 0.3 LSU/ha (Fig. 1). However, these calculations assume a constant duration of the annual grazing season. Assuming, more Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 81 realistically, a reduced grazing duration in unfavorable climates implies that SR* would not decline to the extent shown in Fig. 1a. A supply of one ton of fresh carbon as cattle manure with a C:N ratio of 14 leads to a rise in the critical stocking rate up to 1.2 LSU/ha (Fig. 1b), whereas the supply of mineral N fertilizer (100 kg N/ha) is relatively less efficient since SR* does not exceed 0.93 LSU/ha (Fig. 1c). While manure supply has a large impact on SR* in unfavorable climate (SR* with manure is always above 0.6 LSU/ha), mineral N fertilizer has a small negative impact on SR* in poor climates (minimum SR* of 0.26 LSU/ha) (Figure 1). Figure 1. Critical stocking rate (in LSU, livestock units per ha during a 200 days grazing season) leading to a zero GHG balance of pastures which are unmanaged (a), fertilized with manure (supplying one ton C and 70 kg N/ha) (b) and supplied with inorganic N fertilizer (100 kg N/ha) (c). These estimates of SR* are consistent with previous results showing that grazing and cutting intensity reduces carbon storage in grasslands (Soussana et al., 2007), which explains why SR* values are usually low and are reached mostly under extensive grazing conditions. Indeed, herbage use efficiency (f1, the ratio of herbage intake to above-ground net primary productivity) has a low value when carbon neutrality is reached in pastures. With unmanaged pastures, f1*=0.20 and this value is constant across the climatic space (Eq. 5). However, f1* derivation in fertilized pastures requires numerical solving and values across climatic conditions vary with manure application between 0.29 and 0.40, and with N fertilizer application between 0.18 and 0.22. These results show that carbon neutral pastures can be achieved at low livestock grazing density and confirm the concept of a critical livestock density proposed by Soussana and Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 82 Lemaire (2014) for environmentally sustainable grassland intensification. Further work will be required to apply this concept to realistic grazing systems by taking into account livestock needs for grassland roughage in winter (and in summer in the Mediterranean). Moreover, the above results are only valid on average. The exact SR* value may vary across grassland fields depending on the soil type, on the vegetation cover, on the current soil organic carbon stock and on the abundance of N-fixing legumes. For instance, drained organic soils are unlikely to store C even under low grazing density conditions (Soussana et al., 2010). The likely technical feasibility of carbon neutral pastures questions the possibility of voluntary carbon payments that could be offered to farmers managing ‘carbon neutral pastures’. Such payments would aim at compensating for the loss of profitability created by the limitation in animal stocking density below SR*. While this carbon mitigation option has a large technical potential, given the extent of grasslands in Europe, its cost would need to be assessed (see Pellerin et al., 2013) and compared to other options in the animal agriculture sector (e.g. biogas, changes in animal diets, etc.). This option could be combined with P fertilization in some depleted soils and with an increased use of pasture legumes which would further reduce N2O and CH4 emissions (Luescher et al., 2014). Silvo-pastoral systems which are currently being developed in the wet tropics could also be increasingly used to strengthen the grassland carbon sink in the temperate zone (Pellerin et al., 2013). 5. The climate sensitivity of grassland carbon stocks Grassland production is intimately linked to climate conditions and therefore highly exposed to climatic variability and climate change. Between 1980 and 1999, severe droughts have caused mortality rates in national herds of between 20% and 60% in several arid sub-Saharan countries (IPCC, 2007). The extreme drought and heatwave that hit Europe in the summer of 2003 was unprecedented since at least 1500. It caused a green fodder deficit of up to 60% in affected countries like France. In Switzerland fodder had to be imported from as far away as Ukraine. In Australia, the widespread six-year drought from 2001 to 2007 is considered the most severe in the nation’s history and had large negative impacts on livestock production. A further drying of large parts of the subtropics is likely by the end of this century (IPCC, 2013). For instance, in Europe, in the next 40 years, the risk of summers as warm as 2003 may increase by two orders of magnitude and may approach the norm by 2080 under high emission scenarios. Increased aridity and persistent droughts are projected in the twenty-first century for most of Africa, southern Europe and the Middle East, most of the Americas, Australia and South East Asia (Field et al., 2012). A number of these regions have a large fraction of their land use covered by grasslands and rangelands. Projected increases in climate variability and increases in the length of the dry summer period is likely to impact negatively on ground cover in Mediterranean climates and in drylands, increasing soil erosion risks (Crimp et al., 2010). A probabilistic risk analysis can be developed, by defining risk as the product of hazard probability (e.g. the probability of drought occurrence) and the response to hazard (Van Oijen et al., 2013). With this approach, a significant increase in exposure to summer drought risk was evidenced for French grasslands (Graux et al., 2013). Simulated future conditions show an increased inter-annual and seasonal variability of grassland production. Dairy production at grazing in summer is estimated to drop down below one-third of the current median value in four out of 30 years for 2070–2099, whereas similar shortfalls were not observed with the baseline climate (Graux et al., 2013). A detailed analysis of risks under the A1B emission scenario further shows that European grasslands could shift from a carbon sink to a carbon source for the atmosphere in extreme years (Van Oijen et al., 2014). The Agricultural Model Intercomparison and Improvement Project (AgMIP) has initiated a Coordinated Climate-Crop Modeling Project (C3MP) with a protocol first established for arable crops (Ruane et al., 2014). This protocol has been adapted to grassland systems Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 83 considering specific management recommendation, with first focus on temperate grasslands. The impact of climate change on future greenhouse gas emissions and removals from grassland systems is explored by utilizing site-calibrated models to provide projections under probabilistic climate change scenarios. These scenarios are defined by a combination of air temperature, precipitation and CO2 atmospheric rate changes. This protocol has been applied to a temperate grassland site, assuming a 200 days grazing season with mineral N fertilizer supply and a herbage use efficiency of 0.2 (close to f1* under current climate conditions). Results show that a rise in air temperature and in precipitation departs the pasture from carbon neutrality and leads to a large rise in GHG emissions per unit area (Fig. 2a). Moreover, as the pasture productivity declines with warming and reduced precipitation the GHG balance per head increases exponentially (Fig. 2b). Therefore, climatic variability has a strong potential to shift the carbon sink of grasslands to a carbon source. These results do not account, however, for the long term acclimation of grasslands to climate change through changes in vegetation which may increase pasture resilience. Figure. 2. Simulated GHG balance of a temperate grassland for a range of temperature and precipitation conditions (following the C3MP protocol, see Ruane et al., 2014). GHG balance was expressed per unit pasture area (a) and per head (b) assuming a 200 days grazing season with one cattle livestock unit. A herbage use efficiency value of 0.2 (close to f1*, see text, under current climate conditions) was assumed. Conclusion The carbon sink of European grasslands results from past changes in management (grassland fertilization, reduction in stocking density) and from global change (rise in atmospheric CO2 and warming). This carbon sink can be managed by grazing and by fertilization. A better understanding of the role of these drivers and of their interactions with soil and vegetation types may allow designing guidelines for carbon neutral pastures. Such carbon neutral pastures are extensive and extensification is currently not economically viable in the absence of funding from carbon markets or targeted agricultural subsidies. Priority should therefore be given to the restoration of degraded pastures which offers win-win possibilities by combining increases in plant productivity, in soil carbon stocks and in animal production. Such a scheme has already been successfully introduced in Portugal (see www.terraprima.pt/pt/) using phosphorus fertilization and species rich grass-legume mixtures. However, soil organic carbon stocks are vulnerable to changes in agricultural management and to climate change. Therefore, the mitigation of CH4 and N2O emissions from animal production is required at all stages of livestock supply chains in addition to efforts aiming at pasture restoration and transition towards carbon neutral pastures. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 84 Acknowledgements This research was supported financially by the AnimalChange project (Grant agreement number: FP7- 266018) References Allard V., Soussana J.F., Falcimagne R., Berbigier P., Bonnefond J.M., Ceschia E., D’hour P., Hénault C., Laville P., Martin C. and Pinares-Patino C. (2007) The role of grazing management for the net biome productivity and greenhouse gas budget (CO2, N2O and CH4) of semi-natural grassland. Agriculture, Ecosystems and Environment 121, 47-58. Ammann C., Flechard C.R., Leifeld J., Neftel A. and Fuhrer J. (2007) The carbon budget of newly established temperate grassland depends on management intensity. Agriculture Ecosystems and Environment 121, 5-20. Baldocchi D. and Meyers T. (1998) On using eco-physiological, micrometeorological and biogeochemical theory to evaluate carbon dioxide, water vapor and trace gas fluxes over vegetation: a perspective. Agricultural and Forest Meteorology 90, 1-25. Bellamy P.H., Loveland P.J., Bradley R.I., Lark R.M. and Kirk G.J.D. (2005) Carbon losses from all soils across England and Wales 1978-2003. Nature 437,245-248. CLIMSOIL 2008. Review of existing information on the interrelations between soil and climate change. Final report of the ClimSoil project, contract number 070307/2007/486157/SER/B1. European Commission, December 2008. Chang J. (2014) Amélioration de la représentation des systèmes prairiaux dans un modèle de dynamique de végétation globale. Thèse de Doctorat, Université de Versailles Saint-Quentin en Yvelines, France. Chapter 3, 180pp. Conant, R.T., Paustian, K. and Eliott, E.T. (2001) Impacts of periodic management andconversion into grasslands: effects on soil carbon. Ecological Applications 11, 343-355. Crimp, S.J., Stokes, C.J., Howden, S.M., Moore, A.M., Jacobs, B., Brown, P.R., Ash, A.J., Kokic, P. and Leith, P. (2010) Managing MDB livestock production systems in a variable and changing climate: challenges and opportunities. Rangelands Journal 32: 293-304. Davison B., Brunner A., Ammann C., Spirig C., Jocher M. and Neftel A. (2008) Cut-induced VOC emissions from agricultural grasslands. Plant Biology 10, 76-85. FAO (2013) Tackling climate change through livestock. A Global assessment of emissions and mitigation opportunities, FAO, 2013. Field C. B. et al. (2012) Special Report on Managing the risks of Extreme Events and Disasters to Advance Climate Change Adaptation (SREX), Cambridge University Press. Firbank L.G., Smart S.M., Crabb J., Crabb J., Critchley C.N.R., Fowbert J.W., Fuller R.J., Gladders P., Green D.B., Henderson I. and Hill M.O. (2003) Agronomic and ecological costs and benefits of set-aside in England. Agriculture Ecosystem and Environment 95, 73-85. Follett R.F. and Schuman G.E. (2005) Grazing land contributions to carbon sequestration. In ‘Grassland: a global resource’, Proceedings of the XXth International Grassland Congress, Dublin, Ireland (ed. DA McGilloway), pp. 265-277. Wageningen Academic Publishers, Wageningen, The Netherlands. Fontaine S., Barot S. (2005) Size and functional diversity of microbe populations control plant persistence and long‐term soil carbon accumulation. Ecology Letters 8, 1075-1087. Fontaine S., Barot S., Barre P., Bdioui N., Mary B. and Rumpel C. (2007) Stability of organic carbon in deep soil layers controlled by fresh carbon supply. Nature 450, 277-281. Fontaine, S., Henault, C., Aamor, a., Bdioui, N., Bloor, J. M. G., Maire, V., Mary, B., et al. (2011) Fungi mediate long term sequestration of carbon and nitrogen in soil through their priming effect. Soil Biology and Biochemistry 43, 86-96. Franzluebbers A.J. and Doraiswamy P. (2007) C sequestration and land degradation in Climate and Land degradation. Springer ed Berlin, pp 343-358. Goidts E. and van Wesemael B. (2007) Regional assessment of soil organic carbon changes under agriculture in Southern Belgium (1955–2005). Geoderma 141, 341-354. Graux, A.I.; Bellocchi, G.; Lardy, R.; Soussana, J-F. (2013) Ensemble modelling of climate change risks and opportunities for managed grasslands in France. Agricultural and Forest Meteorology 170, 114-131. IPCC (2006) Good practice guidance on land use change and forestry in national greenhouse gas inventories. IPCC, Institute for Global Environmental Strategies, Tokyo, Japan. IPCC (2007) Climate change: impacts, adaptation and vulnerability. In: Contribution of WG II to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge. IPCC (2013) Climate change 2013: The physical science basis. Cambridge University Press, Cambridge, 1535pp. IPCC (2014) Mitigation of climate change, Cambridge University Press (in press). Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 85 Janssens I.A., Freibauer A., Ciais P., Smith P., Nabuurs G.J., Folberth G., Schlamadinger B., Hutjes R.W.A., Ceulemans R., Schulze E.D., Valentini R. and Dolman A.J., 2003. Europe’s terrestrial biosphere absorbs 7 to 12% of European anthropogenic CO2 emissions. Science 300, 1538-1542. Klumpp K., Tallec, T., Guix N. and Soussana J.-F. (2011) Long-term impacts of agricultural practices and climatic variability on carbon storage in a permanent pasture. Global Change Biology 17, 3534-3545. Lal, R. (2004) Soil carbon sequestration impacts on global climate change and food security. Science, 304, 16231627. Lauenroth, W.K. (1979) Grassland primary production: North American Grasslands in perspective. In: French, N.R. (ed.), Perspectives in Grassland Ecology. Ecological Studies. Springer-Verlag, New York, pp. 3-24. Lemaire, G., Chapman, D. (1996) Tissue flows in grazed plant communities. In: Hodgson, J., Illius, A.W. (Eds.), The Ecology and Management of Grazing Systems. CAB International, Wallingford, UK, pp. 3-35. Lemaire, G., Wilkins, R., Hodgson, J. (2005) Challenge for Grassland Science: managing research priorities. Agriculture Ecosystems and Environment 108, 99-108. Lettens S., van Orshoven J., van Wesemael B., Muys B. and Perrin D. (2005a). Soil organic carbon changes in landscape units of Belgium between 1960 and 2000 with reference to 1990. Global Change Biology 11, 21282140. Lettens S., Van Orshovena J., van Wesemael B., De Vos B. and Muys B. (2005b) Stocks and fluxes of soil organic carbon for landscape units in Belgium derived from heterogeneous data sets for 1990 and 2000. Geoderma 127, 11-23. Lüscher et al. (2014) Potential of legume‐based grassland-livestock systems in Europe: a review. Grass and Forage Science 69, 206-228. Marcott, Shaun A.; Shakun, Jeremy D.; Clark, Peter U.; et al. (2013) A Reconstruction of Regional and Global Temperature for the Past 11,300 Years. Science 339, 1198-1201. Nemani, R.R., Keeling, C.D., Hashimoto, H., Jolly, W.M., Piper, S.C., Tucker, C.J. and Myneni, R.B. (2003) Climate-driven increases in global terrestrial net primary productionfrom 1982 to 1999. Science 300, 1560-1563. Ogle S.M., Conant R.T. and Paustian K., 2004. Deriving grassland management factors for a carbon accounting method developed by the Intergovernmental Panel on Climate Change. Environmental Management 33, 474-484. Ojima, D.S., Parton, W.J., Schimel, D.S., Scurlock, J.M.O. and Kittel, T.G.F. (1993) Modelling the effects of climatic and CO2 changes on grassland storage of soil C. Water, Air, and Soil Pollution 70, 643-657. Paustian, K., Cole, C.V., Sauerbeck, D., Sampson, N. (1998) CO2 mitigation by agriculture: an overview. Climatic Change 40, 135-162. Pellerin S., Bamière L., Angers D., Béline F., et al. (2013) Quelle contribution de l’agriculture française à la réduction des émissions de gaz à effet de serre? INRA (France), 90pp. Piao S., Fang J., Ciais P., Peylin P., Huang Y., Sitch S. and Wang T. (2009) The carbon balance of terrestrial ecosystems in China. Nature 458, 1009-1013. Pinares-Patino C.S., D’Hour P., Jouany J.P. and Martin C. (2007) Effects of stocking rate on methane and carbon dioxide emissions from grazing cattle. Agriculture Ecosystems and Environment 121, 30-46. Ruane, A.C., McDermid S., Rosenzweig C., Baigorria G.A., Jones J.W., Romero C.C., and Cecil L.D. (2014) Carbon-temperature-water change analysis for peanut production under climate change: A prototype for the AgMIP Coordinated Climate-Crop Modeling Project (C3MP). Global Change Biology 20, 394-407. Schlesinger, W.H. (1990) Evidence from chronosequence studies for a low carbonstorage potential of soils. Nature 348, 232-234 Schulze E.D., P. Ciais, S. Luyssaert, A. Freibauer, I.A. Janssens, J.F. Soussana et al. (2009) The Greenhouse Gas Balance of Europe: Methane and nitrous oxide compensate the carbon sink of EU-25. Nature Geosciences 2, 842850. Siemens J. (2003) The European carbon budget: a gap. Science 302, 1681-1681. Six, J., Conant, R.T., Paul, E.A. and Paustian, K. (2002) Stabilization mechanisms ofsoil organic matter: implications for C-saturation of soils. Plant and Soil 241,155-176. Smith P., Chapman S.J., Scott W.A., Black H.I.J., Wattenbach M., Milne R., Campbell C.D., Lilly A., Ostle N., Levy P.E., Lumsdon D.G., Millard P., Towers W., Zaehle Z. and Smith J.U. (2007) Climate change cannot be entirely responsible for soil carbon loss observed in England and Wales, 1978-2003. Global Change Biology 13, 2605-2609. Soussana J.F., Graux A.I. and Tubiello F.N. (2010a) Improving the use of modelling for projections of climate change impacts on crops and pastures. Journal of Experimental Botany 61, 2217-2228. Soussana J.F. and Tallec T. (2010b) Can we understand and predict the regulation of biological N2 fixation in grassland ecosystems? Nutrients Cycling in Agroecosystems 88, 197-213. Soussana, J.F., Tallec, T. and Blanfort, V. (2010c) Mitigating the greenhouse gas balance of ruminant production systems through carbon sequestration in grasslands. Animal 4:3, 334-350. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 86 Soussana, J.F., Allard, V., Pilegaard, K, Ambus, C., Campbell, C., Ceschia, E., et al. (2007) Full accounting of the greenhouse gas (CO2, N2O, CH4) budget of nine European grassland sites. Agriculture Ecosystems and Environment 212, 121-134. Soussana, J.F., Loiseau, P., Vuichard, N., Ceschia, E., Balesdent, J., Chevallier, T. and Arrouays, D. (2004) Carbon cycling and sequestration opportunities in temperate grasslands. Soil Use and Management 20, 219-230. Soussana, J.F. and Lemaire G. (2014) Coupling carbon and nitrogen cycles for environmentally sustainableintensification of grasslands and crop-livestock system. Agriculture, Ecosystems and Environment 190, 9-17. Van der Werf G.R., Randerson J.T., Giglio L., Collatz G.J., Kasibhatla P.S. and Arellano A.F. (2006) Interannual variability in global biomass burning emissions from 1997 to 2004. Atmospheric Chemistry and Physics 6, 34233441. Van Oijen M., et al. (2013) A novel probabilistic risk analysis to determine the vulnerability of ecosystems to extreme climatic events. Environmental Research Letters 8 (1). Van Oijen M et al. (2014) Impact of droughts on the C-cycle in European vegetation: 1 A probabilistic risk analysis using six vegetation models. Biogeosciences (in press). Van Oost K., Quine T.A., Govers G., De Gryze S., Six J., Harden J.W., Ritchie J.C., McCarty G.W., Heckrath G., Kosmas C., Giraldez J.V., da Silva J.R.M. and Merckx R. (2007) The impact of agricultural soil erosion on the global carbon cycle. Science 318, 626-629. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 87 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 88 Theme 1 submitted papers Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 89 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 90 Reducing greenhouse gas emissions in silage production with oxygen barrier film Wheelton P.1, Wilkinson J.M.2, Van Schooten H.3, Jan Ten Hagen P.3 and Wigley S.1 1 Bruno Rimini Ltd, 309 Ballards Lane, London N12 8LY, United Kingdom 2 School of Biosciences, University of Nottingham, Sutton Bonington Campus, Loughborough, LE12 5RD, United Kingdom 3 Wageningen UR Livestock Research, P.O.Box 65, 8200AB Lelystad, The Netherlands Corresponding author: j.mike.wilkinson@gmail.com Abstract Farm-scale bunker silos were filled with whole-crop maize (Zea mays L.) and their top surfaces covered with either two layers of standard low-density polyethylene film of 150μm thickness, following normal practice, or a single layer of low-density oxygen barrier (OB) film of 45μm thickness. Total weight of film used per silo was 241.5 kg for standard film and 43.4 kg for OB film. Primary energy used per silo for the manufacture of the films was 18.9 GJ for standard film and 3.39 GJ for OB film. Estimated global warming potential of film used per silo was 514.4 kg CO2e for standard film and 92.3 kg CO2e for OB film. Mean composition of samples of silage taken from the top 30 cm of each silo was similar between the two types of film. The use of OB film reduced primary energy use and greenhouse gas emissions associated with film by 82% without affecting silage composition adversely. Keywords: silage, polyethylene, greenhouse gases Introduction A bunker silo of 40 m length, 12 m width and 2.5 m height contains almost 20% of the original ensiled crop in the top 0.5 m. Studies with 127 farm-scale silos in the USA over a four-year period revealed that loss of organic matter (OM) during the storage period was 470 g kg-1 in the top 0.5 m of uncovered silos, compared to 113 g kg-1 for the same silage 0.5 m to 1 m from the top surface (Bolsen, 1997). These losses illustrate the importance of maintaining an effective barrier to both water and oxygen throughout the storage period. Covering silos with polyethylene film reduces losses by protecting the crop from the effects of wind and rain and also by reducing, but not preventing, oxygen permeation into the silo. Normal practice in northern Europe is to line the side walls with a single sheet of film that overlaps the periphery of the top surface, and to use two layers of film to cover the top surface itself. Global consumption of low-density polyethylene film for silage was 582.5 kt in 2012 (Wordpress, 2013). We estimate that 368 kt of polyethylene film are used annually worldwide to cover walled bunker and unwalled clamp silos and 156 kt of stretch-film to wrap baled silage. The production of low-density polyethylene is associated with the consumption of 78.1 MJ primary energy kg-1 and with a global warming potential (GWP) of 2.13 kg CO2e kg-1 (Plastics Europe, 2008). Additional energy is used in the recovery and recycling of film and initiatives have been launched in some member states of the European Union to encourage the recycling of agricultural plastics. For example, in France farmers are charged €65 t-1 of film to support a recycling programme (Comité Français des Plastiques en Agriculture, 2012). Losses of nutrients to the atmosphere through the aerobic deterioration of silage increase greenhouse gas (GHG) emissions per unit of animal product output and have a large negative impact on net return to labour and management (Van Schooten and Phillipsen, 2012). The use of oxygen barrier (OB) film to cover silos and bales is associated with reduced losses of organic matter during the storage period and increased aerobic stability of silage in the peripheral areas of silos, compared to standard polyethylene film (Wilkinson and Fenlon, 2013). In this paper Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 91 the hypothesis was tested on farm-scale silos that the use of OB film reduced GHG emissions in silage production compared to standard polyethylene film. Materials and methods Two adjacent walled bunker silos of 40 m length and 12 m width at the Waiboerhoeve Research Farm of Wageningen UR Livestock Research, Lelystad, The Netherlands, were filled with chopped whole-crop forage maize (Zea mays L.) between 15 and 20 October 2012. Both silos were filled to an average height of 2.2 m. Harvesting and ensiling equipment were identical for both silos and about 700 t of fresh crop were ensiled in each silo. The top surface of one silo was covered immediately with two sheets of standard low-density polyethylene film (RKW ProAgri®, Michelstadt, Germany), following normal practice on the farm. Each sheet was of 50 m length, 14 m width, 150 μm thickness and 0.92 g cm-3 density. A third sheet of the same standard polyethylene film (50 m length x 7 m width x 150 μm thickness, 0.92 g cm-3 density) was used to line the side walls. The top surface of the other silo was covered immediately with a single sheet of OB film comprising low density polyethylene co-extruded with ethylene vinyl alcohol (Supastop®, B Rimini Ltd, London, UK) of 50 m length, 14 m width, 45 μm thickness and 0.93 g cm-3 density. A second sheet of the same OB film (50 m length and 5 m width, 60 μm thickness and 0.93 g cm-3 density) was used to line the side walls. Woven polypropylene netting (Genap BV, ’s-Heerenberg, The Netherlands), weighed down by gravel bags, was placed above the top sheets of both silos. Six samples of silage of 1 kg fresh weight were taken for compositional analysis (BLGG AgroXpertus, Wageningen, The Netherlands) to 30 cm depth from the top surface and 2 m from the outer walls from each silo during the feed-out periods; on 26 June 2013 (251 days post-ensiling) for the silo covered with OB film and on 11 September 2013 (328 days post-ensiling) for the silo covered with standard film. Results Primary energy use and GHG emissions, estimated as GWP, associated with the two films are shown in Table 1. The use of OB film reduced total weight of film, primary energy use and GWP from film by 82% compared with standard film. Additional benefits of lower mass of film to be recycled, with further reductions in GHG, would also accrue. Table 1. Weight of film used per silo, associated primary energy use and global warming potential (GWP): Standard film compared to OB film Standard film OB film Total weight of film (kg per silo) 241.5 43.4 Primary energy @ 78.1 MJ kg-1 film (GJ) 18.9 3.39 GWP @ 2.13 kg CO2e kg-1 film (kg) 514.4 92.3 The mean composition of the silages in the top 30cm stored under standard polyethylene or OB film is shown in Table 2. Differences in mean composition between silages stored under standard and OB film were relatively small. With the exception of water soluble carbohydrates, lactic acid and acetic acid, coefficients of variation for compositional parameters were less than 10% for silages stored under both types of film. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 92 Table 2. Composition of silages (mean ± SD of six samples) in the top 30cm stored under either standard polyethylene or oxygen barrier film Standard film OB film 342.0 ± 1.7 335.7 ± 21.3 Ash 35.5 ± 2.7 41.3 ± 4.1 pH 4.1 ± 0.1 4.1 ± 0.2 Crude protein 64.0 ± 1.9 66.8 ± 1.3 Water soluble carbohydrates 13.3 ± 4.8 15.5 ± 5.0 Starch 400.8 ± 26.9 382.2 ± 31.2 Neutral detergent fibre 369.2 ± 34.0 367.0 ± 18.2 Acid detergent fibre 201.2 ± 19.3 205.3 ± 14.0 Digestible organic matter 726.0 ± 22.1 716.5 ± 14.6 Lactic acid 42.0 ± 5.1 31.8 ± 11.8 Acetic acid 13.2 ± 4.7 17.3 ± 2.1 Dry matter (DM, g kg-1) -1 Composition (g kg DM) Conclusions Covering ensiled forage maize with a single layer of OB film gave large reductions in primary energy and GHG associated with polyethylene film, compared with normal practice of covering with two layers of standard film. The composition of silage in the top 30 cm was similar between the two types of film. References Bolsen K.K. (1997) Issues of top spoilage losses in horizontal silos. In: Silage: Field to Feedbunk. Northeast Regional Agricultural Engineering Service Publication NRAES-99, pp.137-150. Comité Français des Plastiques en Agriculture (2012) L’augmentation concertée de l’éco-contribution pour les films plastiques agricoles. Communiqué de presse, Décembre 2012. http://www.plastiquesagriculture.com/PDF_telechargeables/CP_ecocontribution_dec_2012.pdf Plastics Europe (2008) Environmental Product Declaration of the European Plastics Manufacturers. Low density polyethylene. http://www.plasticseurope.org/Documents/Document/20100312112214-FINAL_LDPE_27040920081215-018-EN-v1.pdf Van Schooten H. and Phillipsen B. (2012) Grass silage management affecting greenhouse gas emissions and farm economics. In: Kuoppala, K, Rinne, M. and Vanhatalo, A (eds) Proceedings of the XVI International Silage Conference, Finland, pp. 126-127. Wilkinson J.M. and Fenlon J.S. (2013) A meta-analysis comparing standard polyethylene and oxygen barrier film in terms of losses during storage and aerobic stability of silage. Grass and Forage Science doi: 10.1111/gfs.12087 (In press, 2014, vol 69). Wordpress (2013) Agricultural film market is expected to reach USD 9.66 billion in 2019: Transparency Market Research. http://rahul28feb86.wordpress.com/2013/10/21/agricultural-film-market-is-expected-to-reach-usd-966-billion-in-2019-transparency-market-research/ Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 93 Effects of fertilization and soil compaction on nitrous oxide (N2O) emissions in grassland Sturite I.1, Rivedal S.2 and Dörsch P.3 1 Norwegian Institute for Agricultural and Environmental Research, Tjøtta, NO-8861 Tjøtta, Norway 2 Norwegian Institute for Agricultural and Environmental Research, Fureneset, NO-6967 Hellevik i Fjaler, Norway 3 University of Life Sciences Ås, NO-1432 Ås, Norway Corresponding author: Ievina.Sturite@bioforsk.no Abstract The effect of fertilization and soil compaction on nitrous oxide (N2O) emissions in pure grass and mixed grass-clover leys was assessed at Fureneset in western Norway. The experiment was divided into two measurement periods. The first period lasted from 30 April to 29 May, following the application of cattle slurry supplying 110 kg total N ha-1. The second period lasted from 3 to 30 July, following 60 kg total N ha-1 as mineral NPK applied after the first cut. Cattle slurry did not lead to a transient increase in N2O emissions, indicating a low mineralization rate at low temperatures in spring 2013. Soil compaction increased N2O emissions only in the mixed grass-clover ley. The application of mineral fertilizer after the first cut induced transiently high N2O emissions, which tended to be higher in compacted than noncompacted plots. The N2O emissions from mixed grass-clover ley were negligible during the second experimental period, indicating that clover can substitute for the input of mineral fertilizer and thus mitigate N2O emissions during the growing season. Keywords: cattle slurry, clover, gaseous emissions, grasses, mineral fertilizer, nitrogen. Introduction Agriculture is responsible for a large part of the atmospheric loading of nitrous oxide (N2O). The main drivers of N2O emissions are the soil microbial processes of nitrification and denitrification. Several studies indicate that N2O emission rates depend on soil physical environment and management practices. Soil compaction is a common form of soil structure degradation. Due to reduction of total soil porosity and changes in pore size distribution, the mineralization rate of nitrogen (N) and carbon can be reduced (Breland and Hansen, 1996) and air permeability and gas diffusivity altered (Ball et al., 2008). These factors may increase anaerobic microsites in soil that may lead to higher N2O emissions. The management of productive grasslands typically involves cutting and/or grazing, and fertilizer application. The addition of N-fertilizers not only stimulates growth of plants, but also increases the potential of direct N2O losses. The objective of this study was to determine the effect of fertilization and soil compaction on gaseous N2O emissions in pure grass and mixed grass-clover leys. Materials and methods The N2O flux measurements were carried out on third-year leys on sandy loam with pH 5.9 at Fureneset (61° 22' N 5° 24' E) in western Norway. The field trials were harvested twice a year. Two levels of two wheel-by-wheel passes with tractor traffic were introduced after each harvest: no traffic/ no soil compaction (NC) and traffic with heavy tractor (7 t)/compaction (C). Pure grass and grass-clover leys were included in the experiment. The dominating grass species were timothy (Phleum pratense L.) and meadow fescue (Festuca pratensis L.) with a visually determined content of approximately 80-95% of dry matter yield in the pure grass ley. In the mixed ley the content of red clover (Trifolium pratense L.) and white clover (Trifolium repens Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 94 L.) together was 20-30% at first cut and 40-60% at second cut. The content of grasses was reduced correspondingly. The experiment was divided into two measurement periods. The first N2O measurement period lasted from 30 April to 29 May. The first gas sampling occurred shortly before cattle slurry, supplying 110 kg total N ha-1, was applied to all experimental plots. Then gas samples were collected after two hours, and then at 1, 4, 6, 8, 13, 20 and 27 days after application. The second N2O measurement period lasted from 3 to 30 July. Half of the pure grass plots received 170 kg total N ha-1, of which 60 kg N ha-1 was applied after the first cut as mineral fertilizer. Similar to the previous measurement period, the first gas measurement took place shortly before fertilization and then gas samples were collected 1, 2, 5, 7, 9, 13, 16 and 27 days after application. N2O flux was measured using static chambers. One aluminium frame (52 cm x 52 cm x 25 cm) per plot was inserted to a depth of 10-12 cm shortly before the first N2O measurement and left in the soil throughout the measurement period. The frames had a groove filled with water to ensure air-tight connection with 20-cm high vented aluminium chamber. Air samples were taken from the chamber headspace at regular intervals (0, 15, 30, 45 min), using a 20-mL air-tight polypropylene syringe. The samples were injected into pre-evacuated 12-mL glass vials and analysed by a gas chromatograph. Data of actual air and soil temperature and precipitation during the experimental periods were taken from local weather station at Fureneset (Figure 1B and D). The experiment was a completely randomized split plot block design with three replicates. Analysis of variance (General Linear Model) was used according to a split-split plot model to evaluate the significance of tractor traffic (main plots), fertilization type/level (split plots) and seed mixture (split-split plots) on cumulative N2O emissions. Results and discussion The N2O fluxes, generally, were low after the cattle slurry had been applied in spring (Figure 1A). Surprisingly, no short-term N2O emission peaks were measured after the application. This is in contrast with previous studies under the same climate conditions (Rivedal et al., 2013). Relatively low soil temperatures at the beginning of the experiment (Figure 1B) probably resulted in low mineralization rates and low gaseous losses from applied cattle slurry. One day after the application, however, N2O emissions in the compacted grass-clover stands were significantly greater than in compacted and non-compacted pure-grass stands and in noncompacted grass-clover mixtures (Figure 1A). This pattern remained over the entire one-month experiment. These results suggest that more N was available for microbial transformation in grass-clover stands than in pure-grass stands in spring. As expected, the application of mineral fertilizer after the first cut induced transiently high N2O emissions (Figure 1C). Emissions were highest on the second day after the application and decreased significantly thereafter. Difference between fertilized and unfertilized treatments remained for ten days. The effect of soil compaction after the first cut on the N2O production was less, as expected. There was a significant interaction between the soil compaction and fertilizer application on the first day after fertilization. Limited precipitation before tractor traffic resulted in dry soils and may have limited the effect of altered soil structure. It has been demonstrated that more N2O is produced in response to soil drying and rewetting than to soil compaction (Beare et al., 2009). This experiment clearly showed that the presence of clover in the sward did not affect N2O production during the growing season (Figure 1C). This is in accordance with a recent study in Norway which found a statistically measurable impact of clover on N2O emissions only in a dry and warm year but not in a wet and cold year (Hansen et al., 2014). Thus, use of clover in leys may minimize direct N2O losses from synthetic fertilizers, depending on the annual weather conditions. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 95 Figure 1. N2O emissions from the non-compacted (NC) and compacted (C) pure grass and grass-clover mixture after application of 110 kg total N/ha as cattle slurry on 2 May (1A) and after application of 60 kg total N/ha as mineral N on 3 July (1C). Figures 1B and 1D show mean daily air and the soil temperature and sum of daily precipitation. Arrows denote the time of N application. Conclusion Cattle slurry applied in spring did not increase N2O emissions in pure grass and grass-clover stands in a cold year. However, the presence of clover had some impact on N2O flux, particularly in plots subjected to heavy tractor traffic the year before. Use of mineral fertilizer to pure grass after the first cut enhanced N2O emissions immediately after the application, and high fluxes lasted for approximately one week. N2O emissions from mixed grass-clover swards were negligible. Thus, inclusion of clover in a sward may minimize the risk of direct environmental pollution during the growing season. References Ball B.C., Crichton I. and Horgan G.W. (2008) Dynamics of upward and downward N 2O and CO2 fluxes in ploughed or no-tilled soils in relation to water-filled pore space, compaction and crop presence. Soil and Tillage Research 101, 20-30. Beare M.H., Gregorich E.G. and St-Georges P. (2009) Compaction effects on CO2 and N2O production during drying and rewetting of soil. Soil Biology and Biochemistry 41, 611-621. Breland T.A. and Hansen S. (1996) Nitrogen mineralisation and microbial biomass as affected by soil compaction. Soil Biology and Biochemistry 28, 655-653. Hansen S., Bernard M.E., Rochette P., Whalen J.K. and Dörsch P. (2014) Nitrous oxide emissions from a fertile grassland in Western Norway following the application of inorganic and organic fertilizers. Nutrient Cycling in Agroecosystems 98, 71-85. Rivedal S., Hansen S., Løes A.K. and Dörsch P. (2013) Utslepp av lystgass frå moldrik jord på Vestlandet. Bioforsk Fokus 8, 366-368. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 96 Modelling livestock and grassland systems under climate change Kipling R.P.1, Saetnan E.1, Scollan N.D.1, Bartley D.2, Bellocchi G.3, Hutchings N.J.4, Dalgaard T.4 and Van den Pol-van Dasselaar A.5 1 Institute of Biological, Environmental and Rural Sciences, Aberystwyth University, Gogerddan, Aberystwyth, UK, SY23 3EB 2 Moredun Research Institute, Pentlands Science Park, Penicuik, Midlothian, UK, EH26 0PZ 3 Grassland Ecosystem Research Unit, French National Institute for Agricultural Research, 5 chemin de Beaulieu, 63039 Clermont-Ferrand, France 4 Aarhus University, Department of Agroecology. Blichers Allé 20, DK-8830 Tjele, Denmark 5 Wageningen UR Livestock Research, P.O. Box 65, 8200 AB Lelystad, the Netherlands Corresponding author: rpk@aber.ac.uk Abstract The livestock and grassland theme (LiveM) of the MACSUR (Modelling Agriculture with Climate Change for Food Security) (www.macsur.eu) knowledge hub brings together partners from across Europe to develop a pan-European modelling capability in the area of livestock systems modelling of climate change adaptation and mitigation. Through the project, inventories of grassland, animal and farm-scale models, as well as datasets related to grasslands and livestock have been compiled. Model inter-comparisons have taken place for grassland models, and a model evaluation protocol is being developed. Farm-scale modellers are undertaking a model inter-comparison exercise, and the theme has formed links to related projects in order to bring together a more coherent livestock systems modelling community. The need for better knowledge exchange within the livestock research community has been highlighted within the project, and is a focus for further action. The knowledge hub concept creates an arena for collaboration between research groups, disciplines and projects essential for tackling complex global issues such as the impact of climate change on agriculture. Keywords: climate change, grasslands, integration, livestock systems, models Introduction The ‘knowledge hub’ concept within FACCE-JPI (Agriculture, Food Security and Climate Change Joint Programming Initiative) (www.faccejpi.com) is focussed on facilitating the creation of collaborative, inter-disciplinary structures for research, developing the coherence of purpose required to efficiently tackle urgent global and multi-sectoral problems such as climate change (Holzinger et al., 2012; Soussana et al., 2012). The MACSUR knowledge hub brings together 74 organizations from 17 European countries and Israel. It has the aim of developing a pan-European agricultural modelling capability, bringing together modelling teams within and between disciplines to improve the accuracy of predictions of the effects of climate change adaptation and mitigation on European agricultural systems. The project connects crop, trade, livestock and grassland modellers, and is focussed on collating, sharing and evaluating datasets for modelling use, developing methods of model inter-comparison, exploring ways to improve the impact and relevance of modelling outputs, and scaling up model predictions to the regional level. A previous scoping paper detailed the priorities and opportunities provided by MACSUR for crop modelling (Rötter et al., 2013). The paper presented here focuses on the LiveM theme, which deals with modelling livestock systems, including grasslands. Modelling such systems is complex, requiring the input of both physiological and management data, including for example the choice and nutritional constituents of feeds. An important challenge for LiveM is to develop awareness of the value of modelling to the wider livestock and grassland research community; for example, models can be used to demonstrate the potential real-world impacts Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 97 of experimental findings, aiding the successful communication of complex issues to stakeholders and policy-makers. Given the disparate nature of the livestock research community, the first priority for LiveM has been to develop links between research groups and between projects, facilitating the development of a mutual understanding of approaches and research needs. Materials and methods In order to achieve the aims of the LiveM theme within MACSUR, four work-packages (WPs) tackle different aspects of developing integrated modelling capabilities: WP1: Collation and exploration of datasets on animal disease, dairy cow bio-meteorology and C sequestration in grasslands. Production of an inventory of animal-scale livestock models, and the development of online databases to share information. WP2: Identifying grassland models and datasets, developing methods of data evaluation and model inter-comparison and creating a protocol for model evaluation. WP3: Identifying farm-scale models and creating an online inventory, undertaking model intercomparisons to assess the state-of-the-art in farm-scale modelling, and assessing the impact of mitigation policies on livestock systems. WP4: Examining methods of scaling up livestock models in order to investigate the regional impacts of climate change, including methods for stakeholder involvement. Figure 1. Overview of progress in LiveM. Note: acronyms: SOLID - Sustainable, Organic and Low Input Dairying project, AgMIP - Agricultural Model Intercomparison and Improvement Programme (Rosenzweig et al., 2012), TradeM – trade modelling theme. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 98 For all WPs, a central focus will be to contribute to cross-theme regional pilot studies, which bring together the methods and collaborations developed within each theme to model climate change scenarios in northern, central and southern Europe. Results and discussion Within LiveM, good progress has been made across all work-packages (Figure 1). Bellocchi et al. (2013a, b) provide more detail on the achievements of grassland modellers in WP2. Work to develop links between modellers and dataset holders has proved a challenge, so that the creation of online inventories of meta-data has emerged as a priority in the formation of a more cohesive research community. The development of links with related projects is an essential step in building the capacity to address complex global issues in a joined-up way, using the pooled resources of different research groups, projects and disciplines. Providing the space for such interactions is a key advantage of the knowledge hub set-up, which creates an arena of exchange from which future integrated research is generated. Conclusion There is an urgent need for livestock and grassland researchers to develop a more coherent approach to the complex challenges facing the sector, through the development of resources which facilitate increased understanding between groups and disciplines. The MACSUR knowledge hub is an important step in developing a joined-up approach to livestock research. The ultimate aim must be to realise the potential of a more integrated research community that effectively links experimental researchers, modellers, stakeholders and policy-makers. Acknowledgements MACSUR is funded through national funding bodies as part of the EU FACCE-JPI (Joint Programming Initiative for Agriculture, Climate Change, and Food Security). References Bellocchi G., Ma S, Köchy M. and Braunmiller K. (2013a) Datasets classification and criteria for data requirements. FACCE MACSUR Reports, 2. Bellocchi G., Ma S, Köchy M. and Braunmiller K. (2013b) Identified grassland-livestock production systems and related models. FACCE MACSUR Reports, 2. Holzinger F., Meyer S. and Polt W. (2012) European Joint Programming Initiatives. In: OECD (Ed) Meeting Global Challenges through Better Governance: International Co-operation in Science, Technology and Innovation, OECD Publishing, pp. 151-169. Rosenzweig C., Jones J.W., Hatfield J.L., Ruane A.C., Boote K.J., Thorburn P., Antle J.M., Nelson G.C., Porter C., Janssen S., Asseng S., Basso B., Ewert F., Wallach D., Baigorria G. and Winter J.M. (2012) The Agricultural Model Intercomparison and Improvement Project (AgMIP): Protocols and pilot studies. Agricultural and Forest Meteorology 170, 166-182. Rötter R.P., Ewert F., Palosuo T., Bindi M., Kersebaum K.C., Olesen J.E., Trnka M., van Ittersum M.K., Janssen S., Rivington M., Semenov M., Wallach D., Porter J.R., Stewart D., Verhagen J., Angulo C., Gaiser T., Nendel C. Martre P. and de Wit A. (2013) Challenges for agro-ecosystem modelling in climate change risk assessment for major European crops and farming systems. In: Impacts World 2013 Conference Proceedings, Potsdam Institute for Climate Impact Research, Potsdam, pp. 555-564. Soussana J-F., Fereres E., Long S.P., Mohrens F.G.M.J, Pandya-Lorch R., Peltonen-Sainio P., Porter J.R., Rosswall T. and Von Braun J. (2012) A European science plan to sustainably increase food security under climate change. Global Change Biology 18, 3269-3271. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 99 Multiple regression analysis of the relationship between bioclimatic variables and grazing season length on European dairy, beef and sheep farms Phelan P.1, Morgan E.R.2, Rose H.2 and O’Kiely P.1 1 AGRIC, Teagasc, Grange, Dunsany, Co. Meath, Ireland, 2 School of Biological Sciences, University of Bristol, Woodland Road, Bristol, United Kingdom Corresponding author: Paul.Phelan@teagasc.ie Abstract The ability of bioclimatic variables to predict the grazing-season length on European dairy, beef and sheep farms was tested using stepwise multiple regression. Nineteen bioclimatic variables were sourced from the BIOCLIM project and grazing-season length data were sourced from the 2010 EUROSTAT Survey on Agricultural Production Methods. The experimental units were 987 European NUTS regions (nomenclature of territorial units for statistics) of which 703, 774 and 857 had grazing season length recorded for dairy, beef and sheep farms respectively. Bioclimatic variables accounted for an R2 of approximately 0.60 for all three farm types, although the variables selected differed between sheep farms and dairy/beef farms. However, for all three farm types, cold weather limitations had the greatest effect on grazing season length, with the mean temperature of the coldest quarter resulting in R2 values of 0.55 and 0.53 on dairy and beef farms, respectively, and the minimum temperature in the coldest quarter resulting in an R2 of 0.52 for sheep farms. These results will enable some estimations of potential impacts of climate change on grazing management in Europe, although other sources of variation may need to be addressed first. Keywords: grazing, season, climate, dairy, beef, sheep Introduction Grazing-season length is an important component of many ruminant production systems. It can influence production cost, environmental impacts, animal welfare and livestock disease transmission. The BIOCLIM bioclimatic variables are biologically meaningful climate variables on annual trends, seasonality and extreme (potentially limiting) climatic factors that are widely used to predict ecological niches and species geographic distributions (Booth et al., 2014). These bioclimatic variables may also influence the management of livestock on commercial farms. For example, grazing-season length on farms may be influenced by these bioclimatic variables through their impact on grass growth, animal welfare and land trafficability. The objective of this study was to test the ability of bioclimatic variables to predict grazing season length in European regions using multiple regression analyses. Materials and methods Grazing season length data were sourced from the 2010 EUROSTAT Survey on Agricultural Production Methods (SAPM). Full SAPM methodological details are at: http://epp.eurostat.ec.europa.eu/statistics_explained/index.php/Glossary:Survey_on_agricultu ral_production_methods_(SAPM). Grazing-season length was defined as the number of months when livestock had at least some daily access to pasture. Weighted mean grazing season lengths for dairy, beef and sheep farm enterprises (standard EU farm typology: EU Commission Regulation No 1242/2008) were calculated for NUTS (nomenclature of territorial units for statistics) regions in 33 countries across Europe (NUTS 2 in Germany, NUTS 3 in the other 32 countries). There were 978 NUTS regions in total, of which 703, 774 and 857 had grazing-season length recorded on dairy, beef and sheep farms, respectively. Zero-grazing farms were excluded in order to focus only on farms that practised grazing. Gridded (1 km) Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 100 datasets on bioclimatic variables were downloaded from the WORLDCLIM website (www.worldclim.org) and processed using QGIS® to give regional means for each of the NUTS regions described above. There were 19 bioclimatic variables available (Table 1). The relationships between grazing-season lengths and these variables were then analysed using a stepwise multiple linear regression model with the GLMSELECT procedure in SAS®. Table 1. WORLDCLIM (www.worldclim.org) bioclimatic variables based on temperature (temp: °C) and precipitation (prec: mm) records between the years 1950 and 2000. BIO1: BIO2: BIO3: BIO4: BIO5: BIO6: BIO7: BIO8: BIO9: BIO10: Annual mean temp Mean diurnal temp range Isothermality (Bio2 ÷ Bio7) Temp seasonality (standard deviation) Max temp of warmest month Min temp of coldest month Temp annual range (Bio5 - Bio6) Mean temp of wettest quarter Mean temp of driest quarter Mean temp of warmest quarter BIO11 BIO12 BIO13 BIO14 BIO15 BIO16 BIO17 BIO18 BIO19 Mean temp of coldest quarter Annual prec Prec of wettest month Prec of driest month Prec seasonality (coefficient of variation) Prec of wettest quarter Prec of driest quarter Prec of warmest quarter Prec of coldest quarter Results and discussion Bioclimatic variables were significantly associated with grazing season length for all three farm types and it was primarily cold temperature limitations that had the largest effect on grazing season length (Table 2). For dairy and beef farms, BIO11 accounted for the majority of the final model R2 and BIO6 accounted for the majority of the R2 on sheep farms. Low temperatures can reduce grass growth and result in animals being housed for their welfare (Hahn, 1981) and could therefore restrict grazing season length. The difference in variables selected on sheep farms (BIO6) as opposed to dairy/beef farms (BIO11) may be due to the generally longer grazing season on sheep farms (as shown in the dependant mean in Table 2) and therefore the greater likelihood that more extreme measures of cold temperature would become the limiting factor. However, it should also be noted that BIO11 and BIO6 were very closely correlated with each other (R2 = 0.98) and that excluding BIO6 from the model resulted in it being replaced by BIO11 for sheep farms, with little change to the final R2 (0.615 to 0.610) or root mean square error (1.227 to 1.228) and no change to the final selection of other variables. The other bioclimatic variables that were selected by the model had a much smaller effect on the R2 values (Table 2). The biological significance of these variables for grazing season length are more difficult to interpret and, in some cases, may be questionable. Isothermality (BIO3) is a quantification of how large the diurnal temperature range is in comparison to the annual temperature range and it was generally highest in the south-western regions of Europe. However, there are no obvious biological reasons why it should increase grazing season length. In contrast, BIO15 reduced the grazing season on sheep farms, possibly because declining summer precipitation along with increasing winter precipitation could reduce grass growth rates and land trafficability. However, why this was significant for sheep farms and not dairy or beef farms remains unclear. The negative effect of BIO18 (Table 2) on grazing-season length is also surprising. BIO18 can be generally be interpreted as summer rainfall, which is unlikely to be a limiting factor for grazing-season length in most regions of Europe. However, BIO18 is greatest around the Alpine, Carpathian and Kjolen mountain ranges where grazing-season length may be limited by other factors such as cold temperatures and soil characteristics. It should be noted that grazing-season length as recorded by the SAPM does not include any measurements of daily access time to pasture or feed supplementation while at pasture. Therefore, grazing-season length does not directly reflect the importance of grazing for Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 101 milk/meat production on each farm because of feed supplementation within the grazing season. Furthermore, other factors such as land availability and land quality could also influence grazing-season length. Table 2. Multiple regression analysis of the effects of bioclimatic variables on regional grazing season length in Europe. Only variables with a significant effect (P < 0.05) were included in the final model (stepwise selection). Dairy farms a Effect (SE ) Intercept Beef farms R 2b a Effect (SE ) 3.06 (0.454) Sheep farms R 2b Effect (SEa) 4.72 (0.439) R2b 7.10 (0.491) Coefficients: BIO11 (mean temp of coldest quarter) 0.15 (0.019) 0.55 0.13 (0.019) 0.53 BIO6 (min temp of coldest month) 0.16 (0.019) 0.52 BIO15 (prec seasonality) -0.032 (0.003) 0.56 14.56 (1.714) 0.59 BIO3 (isothermality) BIO18 (prec of warmest quarter) 14.7 (1.526) 0.58 -0.005 (0.0007) 0.61 12.81 (1.468) c 0.56 -0.007 (0.0007) 0.60 c -0.006 (0.0007) 0.62c Number of regions 703 774 857 Dependant mean (months) 6.81 7.66 8.27 Model P value <.001 <.001 <.001 Root MSEd 1.15 1.21 a b 2 c Standard Error. Progression of model R as each variable was included. Final R2 of the model. 1.23 Conclusion A number of bioclimatic variables predicted grazing-season length and could account for up to approximately 0.60 of the variation across farm types. In particular, cold temperature limitations (BIO6 and BIO11) had the greatest effect and appeared to be the most biologically meaningful. These results may enable some estimations of the impact of climate change on grazing-season length in Europe. However, approximately 0.40 of the variation in grazingseason length on European dairy, beef and sheep farms was not explained by bioclimatic variables and the inclusion of non-climatic factors may therefore be required. Acknowledgements This study was funded by EU FP7 GLOWORM project (www.gloworm.eu). References Booth T.H., Nix H.A., Busby J.R. and Hutchinson M.F. (2014) BIOCLIM: the first species distribution modelling package, its early applications and relevance to most current MaxEnt studies. Diversity and Distributions 20, 19. Hahn G.L. (1981) Housing and management to reduce climactic impacts on livestock. Journal of Animal Science 52, 1. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 102 Performance of legumes for potential use in pasture swards under conditions of periodic water limitation Breitsameter L.1, Küchenmeister K.1, Küchenmeister F.1,3, Wrage-Mönnig N.2 and Isselstein J.1 1 Georg-August-Universität Göttingen, Department of Crop Sciences, Grassland Science, v.-Siebold-Str. 8, 37075 Göttingen, Germany. 2 Rhine-Waal University of Applied Sciences, Faculty of Life Sciences, Marie-Curie-Str. 1, 47533 Kleve, Germany. 3 Claas rain - Ingenieurbüro für landwirtschaftliche Feldberegnung, Breite Straße 87, 16727 Velten, Germany Corresponding author: Laura.Breitsameter@agr.uni-goettingen.de Abstract In Central Europe, the yield stability of grass-Trifolium repens mixtures may be at risk in the face of climate projections predicting periods of reduced rainfall or drought in summer, because yields of T. repens strongly decrease when water supply is limited. So far, there is little knowledge of the agronomic potential under drought conditions of alternatives to T. repens. In the present study we examined dry matter yield and water-use efficiency of five legume species that are currently of minor importance in pasture swards. Our data showed high dry matter yield stability and water-use efficiency for Medicago lupulina and indicate that this species warrants further consideration as an alternative to T. repens in pasture swards under conditions of limited summer precipitation. Keywords: Dry matter yield, agronomic water-use efficiency, intrinsic water-use efficiency, stable carbon isotope Introduction Climate projections for Central Europe predict periods of reduced rainfall or drought in summer. In temperate pastures, this may set the yield stability of grass-white clover (Trifolium repens) mixtures at risk due to the high water requirement of white clover (Rochon et al., 2004). Besides advances in breeding for white clover cultivars that feature an improved drought tolerance, the use of other legume species for pasture sward mixtures may offer an alternative. However, knowledge about their agronomic potential under drought conditions is limited. In the present study, we examine dry matter yield, water consumption and water-use efficiency of five legume species that are currently of minor importance in pasture swards. Material and methods The study was carried out from 2009 to 2011 at Göttingen, Germany, as a two-factorial (plant species; level of water supply: limited vs. well-watered control) container experiment in an unheated greenhouse. We used Lotus corniculatus L. (‘Bull’), Lotus uliginosus Schkuhr (wild seeds), Medicago lupulina L. (‘Ekola’), Medicago falcata L. (wild seeds), and Onobrychis viciifolia Scop. (‘Matra’), which were selected based on their indicator values of soil moisture requirement, forage quality and tolerance towards grazing (Dierschke and Briemle 2002), as well as Trifolium repens L. (‘Rivendel’), Lolium perenne L. (‘Signum’) and Dactylis glomerata L. (‘Donata’) as controls. The species were established in summer 2009 as pure stands in containers as a randomized block design with four replicates (64 containers of 30 L, filled with a homogeneous mixture of 20 kg air-dried sand, 5.5 kg air-dried compost soil and 0.9 kg vermiculite with a top layer of 1.5 kg compost as a seed bed). In spring 2010, L. uliginosus was re-sown due to total frost damage in winter. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 103 The relation of volumetric soil water content (regular weighing) and soil water tension was determined as a soil water retention curve with a pressure plate extractor. For the control treatment, the containers were re-watered to a water content of 25 vol. % (-0.03 MPa) once it dropped below 18 vol. % (-0.3 MPa). Water limitation was imposed during three periods in spring 2010 (moderate limitation), summer 2010 and spring 2011 (both severe limitation). For the moderate water limitation, the containers were left unwatered until three days after the first sward type (T. repens in most cases) had reached a soil water content of 10 vol. % (-1.5 MPa ), then re-watered to 25 vol. % and left to fall dry again in the same way a second time. For the severe water limitation, the containers were left unwatered until five days after the first sward type had reached a soil water content of 10 vol. %, re-watered to 25 vol. % and left to fall dry again in the same way two more times. The total water consumption of each container during each period was recorded. Between periods of water limitation, one re-growth at non-restricted water supply was allowed for. No fertilization was carried out. Dry matter (DM) yield was assessed by clipping the aboveground biomass at 3 cm stubble height at the end of each period. Herbage was dried at 60 °C for 48 hours and weighed. The agronomic water-use efficiency (aWUE) was calculated based on dry matter yield and total water consumption for each container. The harvested biomass was analysed for stable carbon isotope composition (13C/12C) with an isotope ratio mass spectrometer to calculate intrinsic water-use efficiency (iWUE, i.e. CO2 assimilation per stomatal conductance) according to Farquhar et al. (1989). An analysis of variance was calculated to determine the effects of the factors plant species and level of water supply on dry matter yield and water consumption within a period of water limitation. Results and discussion Water limitation significantly (P<0.001) decreased DM yields, but the extent of yield reduction differed distinctly among the tested species. Under water limitation, T. repens produced merely about half the yield of the control, whereas M. falcata and M. lupulina on average reached ≥ 60% of the yield of the control. In the tested grass species, yields under water limitation were > 80% of those of the control (Table 1). Agronomic WUE, on average, was higher in the Medicago ssp. and in T. repens than in the grasses. Values of this parameter decreased for most species at soil water content values < 10% (Figure 1, left). Soil water content was not as strongly correlated with aWUE as with iWUE. Intrinsic WUE consistently increased in all species with decreasing soil water content (Figure 1, right). The iWUE of M. lupulina and T. repens was comparatively large across the tested range of soil water content, whereas M. falcata and the grass species consistently showed comparatively small iWUE values. The differences in enrichment of 13C in the harvested biomass among the tested species may indicate different strategies, e.g. concerning the extent of stomatal closure at water limitation. In legumes, N fixation may explain higher WUE values as compared to grasses. Our data hint at strong inter-specific differences in WUE for the tested legume species that warrant further consideration. The results of this experiment have pointed out higher dry matter yield stability than in T. repens under conditions of severe water limitation for some legume species, and particularly for M. lupulina. This species also featured high values both of aWUE and iWUE. We therefore suggest further research on this species as an alternative to T. repens in pasture swards in response to predicted future climate change involving periods of limited precipitation in summer. Further research should additionally examine the contribution of the potential alternatives to T. repens regarding dry matter yield and forage quality in grass-clover mixed stands. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 104 Table 1. Mean dry matter (DM) yield and mean total water consumption of pure stands of six legume and two grass species under conditions of severe water limitation (4 to 10 vol. % H 2O in soil) and in the control treatment (18 to 24 vol. % H2O in soil). Data shown are means of two periods of severe water limitation in summer 2010 and spring 2011. Within columns, values followed by the same letter do not differ significantly at the P = 0.05 level. Species (abbreviation) L. corniculatus (Lc) L. uliginosus (Lu) M. falcata (Mf) M. lupulina (Ml) O. viciifolia (Ov) T. repens (Tr) D. glomerata (Dg) L. perenne (Lp) DM yield [g/container] water limited control 46.0 c 83.0 b 43.5 bc 84.6 b 44.6 c 70.1 ab 51.6 c 84.3 b 32.0 ab 56.9 ab 48.0 bc 94.9 b 33.2 a 35.5 a 31.9 a 37.2 a water consumption [l/container] water limited control 19.8 b 28.2 bc 19.5 b 27.4 bc 18.8 ab 21.3 ab 18.2 ab 22.6 ab 17.0 a 22.9 abc 19.8 b 29.9 c 19.8 b 19.9 a 18.8 ab 19.4 a Figure 1. Agronomic (left) and intrinsic (right) water-use efficiency of six legume and two grass species in response to volumetric soil water content [%]. Shown are four values per species: three for water-limited treatments for each of the three stress periods and one for the control (averaged over three periods). Values on the x-axes are minimal values over the respective periods. For species name abbreviations see Table 1. Acknowledgements This study was carried out in the frame of the research co-operation KLIFF and funded by the Ministry for Science and Culture of Lower Saxony, Germany. References Dierschke H. and Briemle G. (2002) Kulturgrasland. Wiesen, Weiden und verwandte Staudenfluren, Ulmer, Stuttgart, 239pp. Farquhar G.D., Ehleringer J.R. and Hubick K.T. (1989) Carbon isotope discrimination and photosynthesis. Annual Review of Plant Physiology 40, 503-537. Rochon J.J., Doyle C.J., Greef J.M., Hopkins A., Molle G., Sitzia M., Scholefield D. and Smith C.J. (2004) Grazing legumes in Europe: a review of their status, management, benefits, research needs and future prospects. Grass and Forage Science 59, 197-214. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 105 Drought effects on herbage production of permanent grasslands in northern Germany Hoffstätter-Müncheberg M.1, Merten M.2, Isselstein J.1, Kayser M.2 and Wrage-Mönnig N.3 1 Institute of Grassland Science, Georg-August-University, Göttingen, Germany D-37073 2 Institute of Grassland Science, Georg-August-University, Vechta, Germany D-49377 3 Faculty of Life Sciences, Hochschule Rhein-Waal, Kleve, Germany, D-47533 Corresponding author: monika.hoffstaetter-muencheberg@agr.uni-goettingen.de Abstract We investigated the influence of sward composition (dry weight by functional group) and nitrogen fertilization on the annual yields of drought-stressed meadows from three sites typical of northern Germany by generating artificial drought with rain-out shelters that held back, on average, 148 of 242 mm rainfall from the beginning of the vegetation period until the end of the last stress treatment. Yields decreased from on average 7510 to 7298 kg ha -1 a-1 when swards were exposed to drought stress, but this effect was enlarged and modified by fertilization, site and sward composition. Nitrogen-fertilization had a positive effect on productivity. The influence of sward composition was site specific and may rather have been a functional-group effect than the influence of diversity. Future productivity losses due to drought stress may be smaller than expected. Keywords: water limitation, yield, sward composition, nitrogen Introduction Larger climatic variability and frequent climate extremes like drought periods will characterize northern Germany’s future climate (Schär et al., 2004). More frequent and severe drought events are expected to have a negative impact on herbage production and fodder quality of permanent grassland (Beierkuhnlein et al., 2011). We hypothesize a negative effect of drought stress on herbage yield and mitigating influences of site, sward composition and fertilization on the size of its effect. In particular, the supposedly positive influence of diversity on grassland productivity (Tilman et al., 2001) still remains unclear where agriculturally managed permanent grasslands are concerned. Materials and methods In a three-year experiment we investigated the effects of drought (with and without rain-out shelters), sward composition (with and without reduction of dicot-species cover) and nitrogen fertilization (with 90 kg ha-1 or without) in a completely randomized block design with four replicates at three locations. Spring and summer drought events (five to six weeks long) were induced by installing rain-out shelters, each of 3.24 m², on three different grassland sites typical for northern Germany (south-eastern (SE) lowland (average -160 of a total 230 mm rainfall from the beginning of the vegetation period until the end of the second stress treatment), submountainous (ave. - 148 of 258 mm), and north-western (NW) lowland (ave. -136 of 237 mm)). Rainfall data are averages over 2011 and 2012. Sward composition was manipulated with herbicides to reduce dicot dry matter yield share from on average 14 to 4 g kg-1 to measure the influence of sward composition. Biomass samples were cut after each stress period and once again in autumn on 0.16 m² and at 7 cm stubble height. Herbage production was determined as accumulated dry weight production (drying at 60 °C for 48 h) per year (three cuts/year). Data were analysed by applying generalized mixed models that allowed for heteroscedasticity. Site, year and block were treated as random effects as they are nested and do not represent real treatments. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 106 Results and discussion The strongest factor effects on annual dry matter production were fertilization and site identity (Table 1). All sites showed enhanced productivity after N-fertilization. Generalizing over all sites, fertilization led to a larger biomass yield under drought-stress conditions, regardless of sward composition. Unfertilized swards had smaller biomass yields if swards were diverse. Grass-dominated, unfertilized swards did not show biomass differences between drought-stress levels. The three sites reacted differently to the treatments (Figure 1). If fertilized grassdominated lowland swards were exposed to drought stress, they had larger biomass yields than if exposed to rainfall. The unfertilized, grass-dominated lowland swards did not show any differences in yield regarding drought exposure. As fertilization leads to higher grass-tillering rates and thus higher sward density (Simon and Lemaire, 1987) that might decrease evaporation. Plants, e.g. Lolium perenne (Lucero et al., 2000), may also be able to increase their water-use efficiency if droughts occur. Figure 1. Productivity as annual dry matter yield (kg ha-1) on three sites (upper row) under the influence of drought stress (filling), sward composition (right side) and N-fertilization (lower axis). Data are derived from three years (2011 - 2013) of observation. The influence of diversity, however, was not found to be as important as suggested by reports in the literature (e.g. Tilman et al., 2001). The two diverse lowland swards showed dissimilar responses to drought stress, with a site-specific influence of fertilization (see Figure 1). The sub-mountainous swards did not show differences in yield regarding sward composition and drought stress. Sward type might only be influential if the dicot share in the biomass exceeds a certain threshold: our diverse sub-mountainous swards did not exceed eight forb species that delivered a mean share of 42 g kg-1 of annual biomass (data not shown). The lowland swards had a higher dicot share (average 50 - 100 g kg-1 of dry matter), even if grassGrassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 107 dominated. All trends seen in the annual yields (Figure 1) are already showing in the summed yields of the first two cuts (after the stress periods). That means the swards did not show compensational growth after the stress periods (data not shown) and all results of annual dry matter yield can be attributed to the first two cuts. This suggests a strong plasticity and high resilience to drought stress (as most swards did not show reduced biomasses if stressed), or soil water content was not the limiting factor, although soil water measurements (data not shown) at the densely rooted upper soil level (15 cm) indicated drought. Further works will clarify this. Table 1. Anova type III table from Linear Mixed Model of annual dry matter yield (kg ha -1 a-1) from three sites over 3 harvest years. The not significant (n.s.) variables were left in the model because omitting them deteriorated the model fit (unlike all other n.s. variables). Factor Sward composition (s) Fertilization (f) Drought stress (d) Year (y) f × site y × site s×d f×d y×d s×y F 1.62193 35.36478 2.50795 6.54821 6.40737 28.24634 18.40618 11.72089 2.51952 1.62961 Degrees of freedom 1 1 1 1 2 2 1 1 1 1 P 0.2039 < 0.001 0.1145 0.0111 0.0019 < 0.001 < 0.001 0.0007 0.1136 0.2029 Significance level n.s. *** n.s * ** *** *** *** n.s. n.s. Conclusion Future productivity losses due to drought stress might be smaller than expected, depending on the site conditions. Fertilization seems to have a supporting effect on drought resilience of lowland swards. Dicot richness can have a stabilizing and yield improving function as mentioned in literature (Tilman et al., 2001), but the dicot share of the biomass maybe needs to exceed a certain threshold to deliver benefits to productivity. Acknowledgements We thank the Ministry for Science and Culture of Lower Saxony for funding. References Beierkuhnlein C., Thiel D., Jentsch A., Willner E., and Kreyling J. (2011) Ecotypes of European grass species respond differently to warming and extreme drought. Journal of Ecology 99, 703-713. Schär C., Vidale P. L., Lüthi D., Frei C., Häberli C., Liniger M.A., and Appenzeller C. (2004) The role of increasing temperature variability in European summer heatwaves. Nature 427, 332-336. Lucero D.W., Grieu P. and Guckert A. (2000) Water deficit and plant competition effects on growth and wateruse efficiency of white clover (Trifolium repens, L.) and ryegrass (Lolium perenne, L.). Plant and Soil 227, 1-15. Simon J.C. and Lemaire G. (1987) Tillering and leaf area index in grasses in the vegetative phase. Grass and Forage Science 42, 373-380. Tilman D., Reich P.B., Knops J., Wedin D., Mielke T. and Lehman C. (2001) Diversity and productivity in a longterm grassland experiment. Science 294, 843-845. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 108 The effect of drought on the depth of water uptake of deep- and shallowrooting grassland species Hoekstra N.J.1, Finn J.A.1, Hofer D.2, Suter M.2 and Lüscher A.2 1 Teagasc, Johnstown Castle Environment Research Centre, Wexford, Ireland 2 Agroscope, Institute for Sustainability Sciences ISS, CH-8046 Zürich, Switzerland Corresponding author: Nyncke.Hoekstra@teagasc.ie Abstract Increased incidence of drought, as predicted under climate change, has the potential to disrupt grassland production, highlighting the need for the design of grassland management systems adapted to future climate-change scenarios. More-deeply rooted plants are more likely to survive extended periods of drought by accessing deeper soil layers that contain higher soil moisture levels. However, very little is known about the depth of water uptake of grassland species as affected by drought. Therefore, in this study, we used the natural abundance δ18O isotope method to assess the effect of drought on the depth of water uptake of two shallowrooting (Lolium perenne L. and Trifolium repens L.) and two deep-rooting species (Cichorium intybus L. and Trifolium pratense L.) in intensively managed grassland monocultures to test the following hypotheses: 1) drought will result in a shift of water uptake to deeper soil layers and 2) deep-rooting species take up a higher proportion of water from deeper soil layers relative to shallow-rooting species. The δ18O isotope method showed a large treatment effect on the proportional contribution of the 0-10 cm soil depth interval to plant-water uptake, which ranged from 0.08 to 0.82. As hypothesized, water uptake shifted to deeper soil layers under drought conditions, with the exception of T. pratense in monoculture. For three of the species, the depth of water uptake corresponded to the rooting depth classification, but the deep-rooting T. pratense actually showed reliance on shallow soil-water uptake. Keywords: δ18O natural abundance, Lolium perenne, Trifolium repens, Cichorium intybus, Trifolium pratense, water uptake Introduction Increased incidence of weather volatility and drought, as predicted to occur under future climate change, has the potential to disrupt grassland production. This highlights the need for designing grassland management systems for forage production which are adapted to future climate change scenarios. Here we focus on the plant functional trait of rooting depth as an adaptation option to drought conditions, as plants that are more deeply rooted are more likely to survive extended periods of drought by accessing deeper soil layers that contain higher soil moisture levels (Chaves et al., 2003). However, very little is known about the depth of water uptake of grassland species and how this is affected by drought, because:1) most grassland research has focussed on aboveground biomass as opposed to belowground measures, and 2) the abundance of roots is not necessarily equivalent to root activity, i.e. water uptake. Increased insight into the effect of drought on the depth of water uptake of different species may yield important information with which to predict both the effect of future climate change on grassland production, and to design agricultural systems with improved resistance and resilience to drought. The natural abundance of δ18O in soil and plant water can be used to measure the depth of water uptake of individual species (Durand et al., 2007). Therefore, the objective of this study was to assess the effect of experimentally induced drought on the depth of water uptake of two shallow-rooting and two deep-rooting species in intensively managed grassland monocultures by using the natural abundance δ18O isotope method. We tested the following hypotheses: 1) drought will result in a shift of water uptake to deeper soil layers, and 2) deep-rooting species Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 109 take up a higher proportion of water from deeper soil layers compared to shallow-rooting species. Materials and methods We selected four model species of which two were shallow-rooting (Lolium perenne L. and Trifolium repens L.) and two were deep-rooting species (Cichorium intybus L. and Trifolium pratense L.). In August 2011, monocultures of all four species were sown on 3 m × 5 m plots in Reckenholz, Switzerland. In 2012, plots were cut six times and received 200 kg N ha -1 yr-1 and enough P and K as to be non-limiting for intensively managed grassland. Using rainout shelters, half of the plots were subjected to a drought treatment of 10 weeks summer rain exclusion (spanning 2 regrowth periods). Natural abundance of δ18O in soil and plant water was used to assess the depth of water uptake of individual species. The lower evaporation rate of heavy isotopes increases the concentration of 18O in water at the soil surface. This results in a vertical gradient in isotopic composition of water in the soil and therefore the composition of plant xylem water is an indicator of the mean depth of water uptake (Durand et al., 2007). Approximately one week before the end of the drought period, stem bases (up to 1.5 cm above soil level) were collected from five to eight tillers per plot of all four sown species. At the same time and for each plot, three soil cores (2 cm diameter) were taken to 40 cm depth and divided into five segments (0-5, 5-10, 10-20, 2030 and 30-40 cm). Water from the soil and plant samples was extracted using cryogenic vacuum distillation. Water samples were analysed for oxygen 18 isotopes at the Boston University Stable Isotope Laboratory. We applied the IsoSource stable isotope mixing model (Phillips and Gregg, 2003) to quantitatively determine the proportional contribution of each of the sources (i.e. five soil depth intervals) to the plant stem water δ18O signature. All statistical analyses were carried out using the statistical software R. Results and discussion During the drought period, a total of 247 mm of rain was excluded from the drought plots, which corresponded to 21% of the total annual rainfall for 2012. There was a large treatment effect on the proportional contribution of the 0-10 cm soil depth interval to plant water uptake (PCWU0-10), which ranged from 0.08 to 0.82 across the four species (Figure 1a). This means that between 8% and 82% of the total water uptake was derived from the 0-10 cm soil depth interval, respectively, and the remainder (92% and 18% respectively) was taken from deeper soil layers (10-40 cm soil depth, Figure 1b). In line with hypothesis 1, the PCWU0-10 of L. perenne, T. repens and C. intybus was reduced by 53%, 40% and 69%, respectively, under drought compared to control conditions (Figure 1a), as a result of a proportional increase in water uptake from deeper soil layers (Figure 1b). In contrast, the PCWU0-10 of T. pratense increased by 33%, resulting in a significant (P< 0.05) species × drought interaction. The reduction in aboveground biomass under drought conditions did not appear to be related to increased depth of water uptake. In fact, the species with the deepest water uptake (C. intybus) was – at this experimental site – one of the species most affected by drought (Hofer et al., 2014; site Reckenholz). Therefore, there are probably different mechanisms that determine the resistance to drought in this system, such as species drought tolerance and nutrient availability. As expected, the two shallow-rooting species L. perenne and T. repens relied mainly on shallow soil water (PCWU0-10 = 0.53), whereas the deep-rooting C. intybus relied on water from deeper soil layers (PCWU0-10 = 0.16) (Figure 1). However, PCWU0-10 of the second deep-rooting species, T. pratense was, on average, 0.68, which is even higher than for the shallow-rooting species. This highlights the fact that the presence of roots is not necessarily equivalent to root Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 110 activity, underpinning the necessity of studies examining root activity. Instead, the presence of roots determines the potential access to different soil layers. a) 0-10 cm soil depth 1.0 b) 1.0 Control Drought 0.8 0.8 0.6 0.6 0.4 0.4 0.2 10-20 cm 20-30 cm 0.2 0.0 Lp Tr Ci Tp PCWU10-40 PCWU0-10 10-40 cm soil depth Species 30-40 cm 0.0 Lp Tr Ci Tp Species Figure 1. The proportional contribution to plant water uptake of a) the 0-10 cm soil depth interval (PCWU0-10) and b) the 10-40 cm soil depth interval (PCWU10-40, split up into 10-20, 20-30 and 30-40 cm soil depth intervals) for the two shallow-rooting species (Lp = L. perenne and Tr = T. repens) and deep-rooting species (Ci = C. intybus and Tp = T. pratense) under control (shaded bars) and drought (white bars) conditions. Error bars are SE, n = 2. Conclusion The natural abundance δ18O technique provided novel insights into the depth of water uptake of deep- and shallow-rooting grassland species and revealed large interspecific differences and shifts in response to drought. Acknowledgements NJH was funded by the Irish Research Council, co-funded by Marie Curie Actions under FP7. References Hofer D., Suter M., Hoekstra N. J., Haughey E., Eickhoff B., Finn J. A., Buchmann N. and Lüscher A. (2014) Important differences in yield responses to simulated drought among four species and across three sites. Grassland Science in Europe 19 (these Proceedings). Chaves M.M., Maroco J.P. and Pereira J.S. (2003) Understanding plant responses to drought - from genes to the whole plant. Functional Plant Biology 30, 239-264. Durand J.L., Bariac T., Ghesquière M., Biron P., Richard P., Humphreys M. and Zwierzykovski Z. (2007) Ranking of the depth of water extraction by individual grass plants, using natural 18O isotope abundance. Environmental Experimental Botany 60, 137-144. Phillips D. and Gregg J. (2003) Source partitioning using stable isotopes: coping with too many sources. Oecologia 136, 261-269. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 111 CLIMAGIE: A French INRA project to adapt grasslands to climate change Durand J.-L.1, Ahmed L.1, Andrieu B.2, Barre P.1, Combes D.1, Cruz P.3, Decau M.L.4, Enjalbert J.5, Escobar-Gutiérrez A.1, Fort F.3, Frak E.1, Ghesquière M.1, Gastal F.1, Goldringer I.5, Hazard L.3, Jouany C.3 Julier Koubaiti B.1, Litrico I.1, Louarn G.1, Meuriot F.4, Morvan Bertrand A.4, Picon Cochard C.6, Pottier J.6, Prudhomme M.P.4, Sampoux J.P.1, Volaire F.7, Zaka S.1 and Zwicke M.6 1 INRA URP3F, BP 80006, 86 600 Lusignan, France 2 INRA EGC, 78850 Grignon, France 3 INRA AGIR, BP 52627, 31326 Castanet-Tolosan, France 4 INRA EVA, BP 5186, 14032 Caen, 5 INRA URGV, ferme du Moulon, 91190 Gif-Sur-Yvette 6 INRA UREP, 234 avenue du Brézet, 63100 Clermont-Fd., 7 CNRS CEFE 1919, route de Mende, 34293 Montpellier 5. Corresponding author: jean-louis.durand@lusignan.inra.fr Abstract The research teams inside the INRA project CLIMAGIE aim to improve the production of knowledge and subsequent innovations for adapting grasslands to the risks of climate change that threaten the maintenance of the ecosystem services provided by grasslands. Using plant biodiversity should contribute to improve grassland resistance and resilience under high climatic constraints and low inputs. Up to now, establishment and maintenance of pluri-specific grasslands remain poorly controlled. Collaboration between communities and functional ecologists, ecophysiologists and quantitative geneticians will provide new rules for species and cultivars ecotypes assembling. We will build up a new framework to propose a range of solutions depending on pedoclimatic conditions and grassland functions, enabling farmers and breeders to cope with uncertainties attached to future climate scenarios. That framework will be tested experimentally and in silico with the models under construction in our teams. It will contribute to the definition of new ideotypes and breeding schemes of major species for plant breeding, in close collaboration with seed companies and end users through participatory selection programmes. Keywords: climate change, water status, temperature, grasses, legumes, grasslands, plant breeding, modelling Introduction The impacts of temperature, water deficits and CO2 on sown monospecific grasslands have been assessed using crop models (Durand et al., 2010). In most locations studied so far, grasslands might show resilience or resistance, or even produce more (though with greater irregularity) under various scenarios of climatic conditions including increase of CO2. However, the responses of plants to extreme droughts and heat waves are not well described in the current models. The future management of grasslands will be based on lower inputs (fertilization, water) requiring the use of more genetically diverse grasslands that can use resources more efficiently (Darwin, 1859; Huyghe and Litrico, 2008). There has been less investigation so far of intra-specific genetic variability, despite clear evidence for of importance (Sampoux et al., 2010; Poirier et al., 2012). Hence, both the ranges of climate conditions and genetic variability must be explored more deeply. Phenology and plant productivity responses to water, temperature and nitrogen in particular need to be re-assessed over the full ranges projected in the future. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 112 The project The multidisciplinary INRA research programme CLIMAGIE organizes the collaboration between ecologists and quantitative geneticians to provide new rules for assembling species and cultivar ecotypes. CLIMAGIE builds up a framework to propose a range of solutions depending on pedoclimatic conditions and grassland functions, enabling farmers and breeders to cope with uncertainties attached to future climate scenarios. Mediterranean conditions with harsh water deficits and frequent heat waves bring about a complete cessation of plant growth during summer. That situation contrasts with temperate conditions, where summer conditions allow for a minimum grassland production. A recent study suggested that the summer water balance (P-ET°) would be a relevant indicator to ascribe a particular region to one of these two domains (Poirier et al., 2012). Each type illustrates tradeoffs between ecophysiological structures and functions (Volaire et al., 2013). The intraspecific genetic variability of most important traits associated to each type are investigated in CLIMAGIE, especially phenology, vegetative shoot and root growth, spatial and temporal patterns of water extraction, water use efficiency (δ13C in foliage biomass), nitrogen fixation and absorption, fructan concentrations and their responses to high temperatures and water deficits. These responses are integrated under the conditions of dense swards, with intense competition for light, both in vivo using controlled experiments and in silico using models under construction in our teams. This contributes to the definition of new ideotypes and breeding schemes for major species (Lolium perenne, Festuca arundinacea, Dactylis glomerata, Trifolium pratense and Medicago sativa), in close collaboration with seed companies on the one hand, and directly with end users through participatory selection programmes on the other hand. Three integrated groups of tasks are defined, at the relevant levels of complexity, both in terms of objects and methods (Figure 1.) The tasks of the project are organized within three work packages: 1. Analysis of the genetic intra- and inter-specific variability of the physiological responses to temperatures and droughts in grassland species (legumes and grasses). In particular, the morphogenetic response of various populations in important grassland species to the full range of temperature (5-40 °C) is studied. The capture of water is investigated using (i) the relationships between root-system architecture and the drought resistance of populations, (ii) a new interpretative model of δ13C variations in the foliage of diverse cultivars growing in the same conditions, and (iii) water control and measurement facilities to relate whole plant surface temperatures to soil-plant water relations. The evolvability of grass populations under severe drought conditions was studied in Festuca arundinacea and Dactylis glomerata and the potential of such conditions to select elite populations is investigated. Integrated methodologies enable the genetic variability of water use, water use efficiency and summer dormancy to be tested. All these studies involve multidisciplinary research groups. 2. Modelling of the dynamics of the long-term production of sown grasslands. Three models will be tested for: (i) spatially explicit tillering of multispecies grass swards based on the SisFRT model (Lafarge and Durand, 2010), (ii) individual based competition including legumes and grasses, and (iii) simulation of sexual reproduction and transmission of traits to the next generation within swards. 3. Operational selection schemes, ideotypes. This includes (i) novel methodologies to assess and manage both ex situ and in situ genetic resources including biogeographical approaches to assess the presence of ecotypes in relation to local pedo-climatic conditions, and (ii) designing selection procedures for mixed-species sown grasslands. The project’s results will be presented at a meeting in autumn 2015 in Lusignan, France. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 113 Genetic analysis level (Method: from genomics to ecology through comparative ecophysiology and quantitative genetics) Plant genetic resources management and localisation (c ) Functioning of communities under future climate conditions and extreme events Variability of the forage and turf grass and legume species Definition of ideotypes and (s) breeding for Participatory and ex situ new cultivars breeding (E/c) Genetic control of adaptation and response of selection to harsh summers responses to T and severe water deficits of species and functional Rules to associate cultivars and ecotypes for adapted grasslands Demonstrative experiment on the interaction between intra- and interspecific variability (Ip) Complex grasslands indidivual-centered simulation modelling Individual plant (ip) Ecotypes/cultivars (E/c) Species (s) Communities ( c ) Diversity level (object: from gene to plant community) Figure 1. Description of the different groups of tasks in CLIMAGIE. (Arrows indicate the exchange of information between the tasks.) References Darwin C. (1859) On the origin of species. Modern library Paperback Edition (1998). Durand J.L. (2012) http://www.inra.fr/climagie. Durand J.L., Bernard F., Lardy R. and Graux I. (2010) Climate change and grassland: the main impacts. In Brisson N. and Levrault F. (eds) Green book of the CLIMATOR project - Climate change, agriculture and forests in France: simulations of the impacts on the main species, 181-190. Huyghe C. and Litrico I. (2008) Analysis of the relationship between the specific diversity and the agricultural value of pastures (study of the literature). Fourrages 194, 147-160. Lafarge M. and Durand J.L. (2010) Comment l’Herbe Pousse. Structures clonales et spatiales des graminées. QUAE. 182 p. Poirier M., Durand J.L. and Volaire F. (2012) Persistence and production of perennial grasses under water deficits and extreme temperatures: importance of intraspecific vs. interspecific variability. Global Change Biology 18, 3632-3646. Sampoux J.P., Baudouin P., Bayle B., Béguier V., Bourdon P., Chosson J.F., Deneufbourg F., Galbrun C., Ghesquière M., Noël D., Pietraszek W., Tharel B. and Viguié A. (2011) Breeding perennial grasses for forage usage: An experimental assessment of trait changes in diploid perennial ryegrass (Lolium perenne L.) cultivars released in the last four decades. Field Crops Research 123, 117-129. Volaire F., Barkaoui K. and Norton M. (2013) Designing resilient and sustainable grasslands for a drier future: Adaptive strategies, functional traits and biotic interactions. European Journal of Agronomy 52, 81-89. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 114 Comparison of temperature responses of different developmental processes in Medicago sativa L. and Festuca arundinacea Schreb. Zaka S., Ahmed L.Q., Escobar-Gutiérrez A.J., Durand J.-L. and Louarn G. INRA.UR4 P3F, Le Chêne – BP6, F-86600 Lusignan, France Corresponding author:serge.zaka@infoclimat.fr Abstract Temperature is one of the most important factors affected by climate change and is also one of the most important variables involved in the control of plant developmental processes. We measured germination, leaf appearance, coleoptile expansion, radicle expansion and leaf expansion rates in lucerne (Medicago sativa L.) and tall fescue (Festuca arundinacea Schreb.). All processes studied related to plant development, and expansive growth followed common Arrhenius-type responses curves on the wide range of temperature (5–40 °C) in each studied species. This result has important consequences for modelling of temperature effects associated with global changes. Keywords: temperature, development, growth, lucerne, tall fescue Introduction Temperature is one of the most important factors affected by climate change. The Intergovernmental Panel on Climate Change anticipates an increase of the global average temperature from +1.8 °C (1.1–2.9 °C) to +4.0 °C (2.4–6.4 °C) by 2100 (IPCC, 2007). Moreover, temperature is one of the most important variables involved in the control of plant developmental processes (i.e. plant phenology, organogenesis and expansive growth). Hence, the productivity and quality of grassland are expected to change in the near future in response to changes in climate (Brisson and Levrault, 2010; CLIMFOUREL, 2011; Ruget et al., 2013). In a context in which climate changes could hasten growth in spring and increase the risk of summer yield losses, it is essential to study more thoroughly the effects of temperature on developmental processes of pasture species. For three crop species, Parent and Tardieu (2010) showed in a meta-analysis of 12 literature references, that the germination rate, cell division rate, leaf initiation and appearance rate, leaf expansion rate, seedling expansion rate and the reciprocal of the duration of phonological phases followed a common Arrhenius-type response curve to temperature after normalization within each species. In the case of perennial pasture species, which have not been subjected to a long history of varietal selection, such a result has to be confirmed. The present study addresses this question on a temperate grass (Festuca arundinacea Schreb., tall fescue) and a legume (lucerne; Medicago sativa L.) species of broad geographic dispersion. Our objective is to analyse the short-term response curves of developmental processes (Table 1) for M. sativa (cv. Barmed and cv. Harpe) and F. arundinacea (cv. Soni and cv. Centurion) in the wide range of temperature 5–40 °C. Materials and methods Germination: After wet-stratification for F. arundinacea or scarification treatments for M. sativa, seeds were placed at constant temperature (from 5 to 40 °C) and darkness in germination chambers in 90 mm diameter Petri-dishes, on two sheets of Whatman (#3645 WM-France) paper imbibed with 5mL of deionized water. Seeds were monitored regularly. Coleoptile and radicle expansion rate: After wet-stratification or scarification treatments, seeds were placed in germination chambers at 25 °C and darkness. When the radicle or coleoptile length was >1mm, seeds were placed in growth chambers over sheets of blue blotterpaper at constant temperature (from 5 to 40 °C), high humidity, darkness and regularly irrigated Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 115 by deionized water. Using a digital camera, photos of the seeds were taken routinely. Coleoptile and radicle expansion rates were determined by image analysis. Three replicates of 30 seeds were used for each population at each temperature. Leaf appearance, stem expansion and leaf expansion rate: Experiments were conducted in growth chambers in fertirrigated pots. At first, an excessive number of plants are grown at 25 °C for 3 weeks. Then, they go through a selection process to keep a homogenous set of plants (i.e. 4.5–5.5 leaves for M. sativa and 2.5–3.5 leaves for F. arundinacea). The homogenous selection is then transferred to the studied temperatures (5, 10, 15, 20, 25, 30, 35 or 40 °C) with a moderate light intensity (400 to 500 µmol.m-2.s-1), a constant photoperiod (16h) and a low water vapour saturation pressure (<1.5 kPa). Regular measurements (Table 1) were performed on plants with a ruler. Measurements stopped at 10 leaves for M. sativa and at 6 leaves for F. arundinacea. Results are presented for temperature treatments at 10, 15, 20, 25, 30 and 35 °C. Table 1. Developmental processes presented in this study Process Measurements Plant development Germination rate Leaf appearance rate Expansive growth Coleoptile expansion rate Radicle expansion rate Stem expansion rate Leaf expansion rate Results and discussion All processes studied related to plant development (germination rate, leaf appearance rate) and expansive growth (coleoptile and radicle expansion rate, stem expansion rate, leaf expansion rate) followed common Arrhenius-type responses curves on the wide range of temperature (5– 40 °C) in each studied species (Figure 1A, 1B and 1D). In order to allow comparison between processes, data were normalized by the mean rate at the optimum. However, different responses to temperature can be observed between species. Plants of F. arundinacea have an optimum of development around 25 °C whereas for M. sativa this is round 30 °C (Figure 1C). Because we worked with populations, standard deviations were important especially at high temperatures. Conclusion In line with the results of Parent and Tardieu (2010), the results presented here suggest that a coordination might exist between developmental processes via a common response curve to temperature in each of the species studies. This result has important consequences for the breeding and modelling of temperature effects associated with global changes. Further processes are currently being analysed, including cell division, photosynthesis and nitrogen fixation. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 116 Figure 1. Figure 1A, Reponses of leaf appearance rate to temperature (cv. Harpe). Figure 1B, Reponses of radicle expansion rate to temperature (cv. Harpe). Figure 1C, Reponses of the leaf-4 expansion rate to temperature (cv. Centurion). Figure 1D, Normalized rates of responses to temperature of leaf appearance (10–35 °C) and radicle expansion rate (5–35 °C). Just normalized responses to temperature are presented here. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 117 References Brisson N. and Levrault F. (2010) Changement climatique, agriculture et forêt en France: simulations d'impacts sur les principales espèces. Livre Vert du projet CLIMATOR (2007-2010). s.l.:ADEME. CLIMFOUREL (2011) Accompagner l'adaptation des systèmes d'élevage périméditerranéens aux changements et aléas climatiques. Available at: http://www.psdr-ra.fr/IMG/pdf/Climfourel_4_pages_2011.pdf [25 09 2013]. IPCC (2007) Contribution of Working Groups I, II and III to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Core Writing Team: Pachauri, R.K. and Reisinger, A. (Eds.). Genève, Suisse: s.n. Parent B. (2010) Modelling temperature-compensated physological rates, based on the co-ordination of responses to temperature of developmental processes. Journal of Experimental Botany, 61, 2057-2069. Ruget F. et al. (2013) Actes des Journées de l'AFPF - 26-27 mars 2013 - Impact des changements climatiques sur les productions de fourrages (prairies, luzerne, maïs): variabilité selon les régions et les saisons. Paris, AFPF. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 118 Qualitative overview of mitigation and adaptation options in livestock systems Van den Pol-van Dasselaar A. and Bannink A. Wageningen UR Livestock Research, P.O. Box 65, 8200 AB Lelystad, the Netherlands Corresponding author: agnes.vandenpol@wur.nl Abstract During the last decades the effects of climate change have received a lot of attention. Many mitigation options have been tested and more recently these are studied in an integrated manner with adaptation options in the FP7-funded project AnimalChange. This paper provides a qualitative overview of mitigation and adaption options in livestock systems at the level of manure/fertilizer, soil, feed/crop and animal, and of their synergies and trade-offs between individual greenhouse gases (GHG). Many of these options are linked to grasslands. Keywords: adaptation, mitigation, synergies, trade-offs Introduction During the last decades the effects of climate change have received a lot of attention. The sources and sinks of GHG emissions have been identified and the variation in their size has been evaluated. Many mitigation options have been tested experimentally and the results have been documented in several reviews (e.g. Vergé et al., 2007). Models are developed to predict GHG emissions and to evaluate mitigation strategies. In a similar manner, adaptation options have been studied (e.g. Olesen et al., 2011) which is particularly important for areas that are most vulnerable to climate change. Research results show numerous interactions between mitigation and adaptation in the context of different environmental and socio-economic conditions. Generally, limited information is available on the quantification and comparison of synergies and trade-offs however, and few papers report on this (e.g. Smith and Olesen, 2010). The project AnimalChange aims to provide scientific guidance on the integration of adaptation and mitigation objectives and to design sustainable development pathways for livestock production in Europe, in northern and Sub-Saharan Africa and Latin America. The aim of the present study was to provide a qualitative overview of mitigation and adaptation options in livestock systems at the level of manure/fertilizer, soil, crop/feed and animal, and of their synergies and the trade-offs between individual GHG. Materials and methods The overview of mitigation and adaptation options and their interaction is presented as a matrix. It is based on a review of available literature, expert judgement and additional information provided by the project partners of AnimalChange (e.g. recent research which has not yet been published, information from other climate change related European projects). AnimalChange partners represent various countries in Europe, Africa and Latin America. The overview focuses on livestock production systems. Results and discussion Table 1 provides a qualitative overview of the most relevant mitigation and adaption options and of their synergies and trade-offs between GHG. The options are strongly linked to changes in the N and C cycles of the farming system. Four categories of options are distinguished in this paper: at the level of manure/ fertilizer, soil, crop/ feed and animal. Many options are linked to grasslands. Furthermore, there are many synergies and trade-offs between adaptation options and mitigation options. The effects of climate change may cause a reduced efficacy or applicability of mitigation strategies. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 119 MANURE / FERTILISER Fertilisation rate Fertiliser type Fertiliser application Cover slurry stores/manure heaps Manure cooling Manure treatment Filtering CH4 from barns SOIL Reduced/zero-tillage Prevent soil compaction Water management Irrigation Restoring degraded lands Pasture reclaiming/recovery Incorporation crop residues CROP / FEED Crop rotation Perennial crops Legumes and mixtures New pasture species Improved crop varieties Novel crops Cover crops Conversion to grass Reforestation Optimal forage management Biodiversity Plant breeding Use climate forecasting Different planting dates Conservation as a buffer Mixed versus single species grass Agroforestry Optimal grazing Increased feed digestibility Feed analysis Improving roughage quality More concentrates Improving grass quality Use of silage maize Additives in general Additive nitrate Matching supply and demand Supplemental feeding ANIMAL Rumen control via breeding Immunological control Less consumption animal products Increased production in general Incr prod extensive systems Incr prod intensive systems Animal breeding Animal management Animal manipulation Replacement rate cattle Cooling of animals Livestock mobility Animal health ect so nm itig Eff ati ect on so ( +) na d a Mi pta t.p tio ot. n( +) CH 4( Mi t.p /+/ ot. ++ ) N2 O( Mi -/+ t.p /+ + ot. ) CO 2 (-/+ Mi tig / + ati +) on var iab Im ilit po y (v rta ari nce abl i n e/ r Ad mi ob apt tig ust ati ati ) on on (lo v a w/ r ia Im bili po hig rta ty h) (va nce ria in a b Pr o le/ dap du rob tat ctiv u st ion ity ) ( Co i l mp ow sts /h i act (hi gh) s (gh/ /+) low Fut /?/ ure be ne me fit) asu . re / re Ap pli c ad yt abi ou l ity se by f ar Acc me ep rs tab (ea ili t sy/ y fo dif r fa ficu rm l t) ers (po or/ go od ) Eff on Op ti Gr o u p Table 1. Qualitative overview of mitigation and adaption options in livestock systems at the level of manure/fertilizer, soil, feed/crop and animal, and their synergies and trade-offs. + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + ++ + ++ + ++ + + - + + + + + + + + + + ++ + + + - ++ ++ + + + ++ ++ + + + + + + + + + + + + + + + + + + + + + + + + ++ ++ + + ++ + ++ + + + + ++ + + + + + + + + + + + - ++ + + + + ++ ++ ++ + + + + ++ + + - + + + + + + rob var rob rob rob rob var high high high low low high low rob low rob var var var rob rob rob high low low low high high low rob high rob rob rob var var var rob rob var var low low low low low low low low low low rob var rob rob rob rob rob rob var rob rob var high low high low high high high low low low high low var var rob rob rob rob var var var rob low low low high high high low low low high + + + var var rob low low high + + + + rob low + var low + + + var rob var rob var var rob var rob var low low high high high low low low low high + + + + + + + + + + + + rob high + + + + + var high rob var var high high low + + ben ready ben ready ben ready low ready high future ? ready high future dif dif dif easy dif dif dif good good good good poor good poor low low low low high low ? ready ready ready ready ready ready ready dif dif dif dif dif easy easy poor good good good poor good poor low low ben low low low low low high ben low high ben low low low high ben high high low high low low high high ben high ready ready ready future future future ready ready ready ready ready future ready ready ready ready ready ready ready ready ready ready ready ready future ready ready ready easy easy easy dif dif dif easy easy dif easy easy easy easy easy easy easy dif dif dif easy dif easy dif easy easy easy dif easy good good good good good poor poor good poor good poor good good good good good poor good good good good good good good poor poor good good high future high future ready ben future ben ready ben future high future ben ready high ready ben future high ready low ready low future dif dif easy dif easy dif dif dif easy dif easy dif dif poor poor poor good good good good good poor good good good good Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 120 It may lead to lower yields due to elevated temperatures and fluctuations in water availability. The interactions between food production, adaptation options and mitigation options are complex and often dependent on the local situation and detailed aspects of the farming system. This limits the applicability of generic results to analyse specific farming systems for example. There are also many constraints for implementation of options in agriculture (e.g. Smith et al., 2007; Smith and Olesen, 2010). Mitigation and adaptation options need, therefore, to be tailored to specific regional contexts. Furthermore, in many countries, especially in Africa, the impact of agriculture on climate is an issue that is of far less importance, because of socioeconomic reasons such as, for example, addressing famine (Vergé et al., 2007). A clear understanding of the consequences of options at field and animal scale is important because farmers have to make their day-to-day decisions at those scales. Simultaneously, it is important to have predictions of the synergies and trade-offs between GHG that are sufficiently accurate. Furthermore, it is important to realize that regional and global effects and decisions at the scale of field and animal affect each other. For example, the impact of rising food demands means, other things remaining equal, that a reduction in food production in a certain region would result in increased food production elsewhere. Within the project AnimalChange, the effects of mitigation and adaptation options will be studied to quantify the consequences of adaptation and mitigation options for a range of farm systems and regions. This will lead to a consolidated overview of tested mitigation and adaptation options, including the investigation of breakthrough options and their applicability range in terms of farming systems and agro-ecological zones and their net effect on productivity and GHG emissions. Conclusion Many adaptation and mitigation options are linked to grasslands. Since synergies and tradeoffs between GHG exist for adaptation and mitigation options, accurate predictions of the effects of these options are needed to tailor them in the context of specific farming conditions. Acknowledgements This study is part of the FP7 AnimalChange project (Grant Agreement 266018) and cofinanced by the Dutch ministry of Economic Affairs (KB-12-006.04-003). References Olesen J.E., Trnka M., Kersenbaum K.C., Skjelvag A.O., Seguin B., Peltonen-Sainio P., Rossi F., Kozyra J. and Micale F. (2011) Impacts and adaptation of European crop production systems to climate change. European Journal of Agronomy 34, 96-112. Smith P. and Olesen J.E. (2010) Synergies between the mitigation of, and adaptation to, climate change in agriculture. Journal of Agricultural Science 148, 543-552. Smith P., Martino D., Cai Z., Gwary D., Janzen H., Kumar P., McCarl B., Ogle S., O’Mara F., Rice C., Scholes B., Sirotenko O., Howden M., McAllister T., Pan G., Romanenkov V., Schneider U. and Towprayoon S. (2007) Policy and technological constraints to implementation of greenhouse gas mitigation options in agriculture. Agriculture, Ecosystems & Environment 118, 6-28. Vergé, X.P.C., De Kimpe C. and Desjardins R.L. (2007) Agricultural production, greenhouse gas emissions and mitigation potential. Agricultural and Forest Meteorology 142, 255-269. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 121 Genetic diversity of Lolium perenne L. in the response to temperature during germination Ahmed L.Q., Durand J.-L., Louarn G., Fourtie S., Sampoux J.-P. and Escobar-Gutiérrez A. J. INRA, UR4 P3F, Le Chêne – BP6, F-86600 Lusignan, France Corresponding author: abraham.escobar@lusignan.inra.fr Abstract Perennial ryegrass (Lolium perenne L.) is the major grass forage species grown in temperate regions worldwide. Temperature is one of the major factors controlling seed germination and, in the context of global climate change, breeding L. perenne varieties adapted to new ranges of temperature could be necessary. The objective of the work presented here was to analyse the genetic variability of perennial ryegrass in response to temperature during germination. It was observed that the responses of the ryegrass populations showed statistically significant differences (P<0.05). At least three groups of populations can be distinguished. The findings of this study suggest that high genetic variability exists within L. perenne for response to temperature during germination. This variability could be exploited to breed new varieties adapted to the new environmental conditions induced by global climate change. Keywords: breeding, grasslands, ryegrass Introduction Perennial ryegrass (Lolium perenne L.) is the major grass forage species grown in temperate regions worldwide. In Europe, grasslands cover at least 30% of the 160 Mha Agricultural Surface Area (ASA). In France, since 2006, grasslands cover 20-25% of its continental land area, which represents around 40% of its ASA. The agricultural use-value of grasslands depends both on the structure of their canopy and on their botanical composition. The annual time course of temperature is one of the most important factors affected by climate change. Further, using a number of scenarios, the Intergovernmental Panel on Climate Change anticipates an increase in global average temperature, in the range 1.8 °C to 4.0 °C, by 2100 (IPCC, 2007). Temperature is one of the major factors controlling plant developmental rates (i.e. plant phenology, organogenesis and expansive growth). Temperature is also important in controlling seed germination. It is well documented that maximum cumulative germination is highly temperature dependent (Bewley and Black, 1994). However, the responses of ryegrass to germination temperatures have only been described for a few populations and in a narrow range of temperatures. In the context of global change, breeding ryegrass that is adapted to new ranges of temperature could be necessary. Knowing the variability of responses to temperature by different accessions of L. perenne germplasm is an unavoidable first step towards such breeding. Thus, the objective of the work presented here was to analyse the genetic variability of ryegrass in response to temperature during germination. Materials and methods Eight populations of ryegrass were evaluated. Six were wild populations collected in different places in France (Table 1). The other two were varieties obtained by divergent selection at INRA-UR4 P3F, Lusignan, France. Seeds were obtained from the Centre de Ressources Génétiques des Espèces Fourragères (INRA-UR4 P3F, France) where they were conserved at 5 °C and 30% Relative Humidity (RH). The seed dry weight was recorded from four replicates of 100 kernels before a cold stratification treatment for seven days, intended to release dormancy. After stratification, four plastic Petri-dishes (90 mm diameter) with 100 seed each, were placed in darkness in growth chambers at the following constant temperatures: 5, 10, 15, 20, 25, 30, 35 or 40 °C. Germination counting was carried out at variable time intervals and Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 122 duration that depended on the temperature treatments. A seed was considered as germinated when its radicle or coleoptile was at least 2 mm out of the seed. Here, we report data on maximum germination percentage. For each population, a third degree polynomial was adjusted by the least squares method. Sequential ANOVA pairwise comparisons were performed between the best fit of a given population and the rough data of a second one. The probability of a calculated F value greater that a tabular F (Pr>F) was calculated and a comparison matrix was constructed. The optimal temperature for germination was estimated. Table 1. List of six wild populations of Lolium perenne L. collected in different sites in France and two varieties (P19 and H1) obtained by divergent selection at INRA-UR4 P3F, Lusignan, France. Information of the collection sites is included Populations Collection Altitud site e ACVF10214 Nailloux 200 ACVF10491 Bésignan 650 ACVF20010 Bosdarros 230 ACVF50013 Saulieu 320 ACVF50039 Lure 400 ACVF60016 Reims 140 P19 Le Chêne 144 H1 idem idem Latitude and Longitude 43° 21' 20" N 1° 37' 27" E 44° 19' 13.6" N 5° 19' 31" E 43° 12' 35" N 0° 21' 41" W 47° 16' 46" N 4° 13' 44" E 47° 41' 0" N 6° 30' 0" E 49° 15' 0.2" N 4° 2' 0" E 46° 24' 10" N 0° 4' 48" E idem Mean Mean Precipitation temperature temperature of warmest of warmest of coldest quarter (mm) quarter (°C) quarter (°C) Precipitation of coldest quarter (mm) 20.0 5.2 170 185 18.3 2.6 171 202 18.4 5.1 188 241 16.7 1.4 226 229 17.5 1.5 259 250 17.6 2.7 175 149 18.3 4.4 162 226 idem idem idem idem Results and discussion The novelty of this work comes from the wide range of temperatures evaluated (5 to 40 °C). Striking results show that no germination at all was observed at 40 °C for any of the eight populations under study. Thus, values of zero recorded at this temperature were excluded from the curve fitting. It was observed that the responses of the ryegrass populations showed statistically significant differences (P<0.05). Indeed, the shape of the best fits were different (Figure 1). At least three groups of populations can be distinguished. The first group is formed by two wild populations and the two varieties (ACVF10214, ACVF10491, P19 and H1). For this group, there was little effect of the extreme temperatures (5 and 35 °C) on the maximum germination percentage. The third degree polynomial fitting the data could be replaced by a parabola. The second group includes three populations (ACVF20010, ACVF50013 and ACVF50039). The shape of the curve is rather asymmetric showing little effect of low temperature (5 °C) and optimum temperatures for maximum germination below 12 °C. The response of population ACVF60016 is very different from the other two groups. The asymmetry of the curve is the opposite of that the second group. Germination was poor at low temperatures (5 and 10 °C), it peaked at 25 °C, and then declined. Overall, these results demonstrate that genetic variability exists within the L. perenne species, which could be exploited to breed new varieties adapted to new environmental conditions that may be induced by global climate change. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 123 80 60 60 40 40 20 ACVF10214 20 ACVF10491 0 0 80 80 60 60 40 40 20 20 ACVF20010 ACVF50013 0 0 100 Germination maximum (%) Germination maximum (%) 100 100 ACVF50039 80 ACVF60016 80 60 60 40 40 20 20 0 Germination maximum (%) Germination maximum (%) 100 0 100 Germination maximum (%) Germination maximum (%) 100 80 100 80 80 60 60 40 40 20 P19 20 H1 0 Germination maximum (%) Germination maximum (%) 100 0 0 5 10 15 20 25 30 Temperature (°C) 35 40 0 5 10 15 20 25 30 35 40 45 Temperature (°C) Figure 1. Maximum germination (%) of eight populations of Lolium perenne L. in response to constant temperature during germination. Conclusion The findings of this study suggest that high variability exists within the species L. perenne for response to temperature during germination. This should prompt physiologists to extend the analyses of response to temperature to other processes (Zaka et al., 2014, this congress) and plant breeders to collect and analyse populations of ryegrass from sites with extreme environmental conditions. We suggest that seed germination of populations from northerly and cold sites is improved by high temperatures and limited by colder temperatures, and vice versa for warm-adapted populations from the South. The variability discovered in this study should serve breeders in developing ryegrass varieties for the future. References Bewley J.D. and Black M. (1985) Seeds physiology of development and germination. Plenum Press. London. IPCC (2007) International Panel on Climate Change. Fourth Assessment Report (AR4). Cambridge University Press, Cambridge. Zaka S., Ahmed L.Q., Escobar-Gutiérrez A.J., Durand J.-L. and Louarn G. (2014) Comparison of temperature responses of different developmental processes in Medicago sativa L. and Festuca arundinacea Schreb. Grassland Science in Europe 19 (these Proceedings). Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 124 Time of ploughing affects nitrous oxide emissions following renovation and conversion of permanent grassland Biegemann T., Loges R., Poyda A. and Taube F. Institute for Crop Science and Plant Breeding, Christian-Albrechts-University, D-24118 Kiel, Germany Corresponding author: tbiegemann@gfo.uni-kiel.de Abstract Grassland ploughing carries the risk of increased nitrogen mineralization and hence of increased nitrous oxide (N2O) emissions. So far, only a few studies are available estimating N2O-emissions after grassland ploughing at different times of the year. In the presented study we determined N2O fluxes from autumn-ploughed and spring-ploughed grassland during two experimental years. As an alternative to grass reseeding, maize was cultivated after springploughing. All treatments comprised the factor nitrogen fertilization in the form of cattle slurry with 0 and 240 kg N ha-1year-1. Results showed increased N2O fluxes compared to the undisturbed control in all ploughed treatments. Grassland renovation in autumn resulted in highest N2O fluxes of 1641 µg m-2 h-1 and highest cumulative emissions of 21.3 kg N2O-N ha1 year-1. Freeze and freeze-thaw-cycles were identified as a major driver of increased N2O fluxes during winter when ploughing of grassland had been carried out in autumn. Hence, grassland ploughing in spring induced lower cumulative N2O-N losses reaching a maximum of 6.32 kg N2O-N ha-1 year-1 when grass or maize was sown afterwards. Cultivating maize after grassland resulted in slightly higher N2O-N losses compared to grass resowing in spring. Fertilizer effect on N2O-emissions was only significant after spring-time renovation. It is concluded that grassland ploughing, if necessary, should occur in spring in order to avoid additional N2O losses due to freeze and freeze-thaw-cycle related emissions in regions where those cycles are expected. Additional N2O losses due to cultivating maize instead of grass resowing seem to be acceptable, under favourable soil conditions, in the year of ploughing. Keywords: nitrous oxide, grassland renovation, land use change Introduction Continuous intensification in dairy farming and the increased demand for renewable energy sources, such as anaerobic digestion in biogas plants, has raised the need for biomass in many parts in North-west Europe. As a consequence, permanent grassland has been converted to arable land and the management intensity of remaining grassland has been considerably increased (Taube et al., 2014). In this case, grassland renovation is reported to be carried out every 5 to 10 years depending on soil type and environmental conditions (Velthof et al., 2010). However, grassland renovation and conversion is usually associated with soil management practices which will ultimately lead to a loss of soil carbon, of between 10 and 32 t C ha-1 two years after grassland ploughing, even though grass is sown afterwards (Linsler et al., 2013; Necpálová et al., 2013). Regarding greenhouse gas emissions, in addition to large losses of CO2, grassland ploughing can induce emissions of the potent GHG nitrous oxide (N2O) (Velthof et al., 2010). N2O is a soil-born greenhouse gas which is mainly produced as a byproduct and intermediate of nitrification and denitrification, respectively. Its production in soils is highly dependent on soil water content and temperature. Therefore, time of ploughing could play a key role for ploughing-related N2O emissions due to different environmental conditions and different growth characteristics of newly established plants competing for nitrogen with soil nitrifiers and denitrifiers. In this context, grass re-seeding could be advantageously compared to other inserted forage crops because of a more rapid establishment of grass. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 125 However, in ley farming systems, grass is usually replaced by other forage crops such as maize for a minimum of one year before grass is reseeded again. The aim of this study was to investigate if time of grassland ploughing could affect N2O emissions in the year following renovation, and if an inserted forage crop such as maize would increase total N2O losses compared to prompt resowing in the year of ploughing. Materials and methods The field experiment was established in a long-term field trial at the experimental farm 'Lindhof' (54° 27' N 9° 57' E) of Kiel University in Northern Germany. The long-term mean annual temperature is 8.9 °C and the long term average annual rainfall is 768 mm. The soil type is classified as sandy loam (pH 5.7) with 11% clay, 29% silt, 60% sand and 1.7% C org in the topsoil (0-30 cm). Permanent grassland plots, sown in 1994, were mulched, rotovated and ploughed to a soil depth of 25 cm in September 2010, 2011 and May 2011, 2012. Ploughed plots were harrowed and re-seeded with a 30 kg ha-1 standard grass mixture. In addition to grassland renovation in spring, ploughed plots were sown with maize in a row width of 0.75 m with a plant density of 10 plants per m². The undisturbed sward served as control treatments. All treatments comprised the factor N-fertilizer (0 and 240 kg N ha-1 year-1 as cattle slurry). Slurry application was carried out using trailing hoses to each of the four silage cuts in the N-fertilized treatment (80, 60, 60, 40 kg N ha-1). A non-N-fertilized-treatment served as control. Maize plots received the same total amount of nitrogen fertilizer shared out in three equal dressings. Nitrous oxide emissions were measured once a week for a period of twelve months, starting shortly before grassland ploughing occurred, using the static chamber method (Hutchinson and Mosier, 1981). For N2O measurement, pre-installed soil collars were closed with a gas-tight chamber (h=35 cm, V= 113 l) for 40 minutes and three gas samples were taken at 20-minute intervals. Samples were analysed in the laboratory for N2O concentrations using a gas chromatograph (model 7890a, Agilent technology Inc., Santa Clara, CA, USA). N2O fluxes were calculated for each treatment and replicate by linear regression between measured N 2O concentrations and time. The cumulative N2O-emissions were calculated by linear interpolation of measured daily fluxes. For statistical analysis the software R (2012) was used. The statistical model included the factors treatment, fertilizer level and year as well as all their interaction terms as fixed factors. For N 2O fluxes, soil temperature (at 5 cm depth), soil freeze-thaw-cycles, water-filled pore space and soil nitrate concentration were modelled as covariables. The plot was regarded as a random factor. Based on this model, an analysis of variances (ANOVA/ANCOVA) was used. Multiple contrast tests were conducted to compare the various levels of the influence factors. Results and discussion Grassland ploughing significantly increased N2O emissions, independently of time and slurry application, compared to the undisturbed control. In contrast to authors who observed higher N2O emissions after spring-ploughing (Velthof et al., 2010), we found that soil freezing and freeze-thaw-cycles were the main driving factors for high N2O fluxes during winter following autumn-ploughing. Hence, we observed maximum fluxes after grassland renovation in autumn in both experimental years. Significant differences in accumulated N2O-N emissions among the two experimental years were strongly correlated to temporal difference for soil freezing during winter (80 days with frozen soil and 15 freeze-thaw-cycles vs. 23 days of frozen soil and 2 freeze-thaw-cycles). N2O-fluxes after spring-ploughing were mainly associated with comparably high soil nitrate contents in the upper soil layer, whereas high N2O-N losses were inhibited due to low soil water contents during spring. Cultivating maize after spring-ploughing resulted in higher N2O-N losses (+ 2.35 kg N2O-N ha-1) compared to grass re-seeding. Higher N-losses could be explained by delayed maize crop establishment in spring and lower N-yields Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 126 compared to reseeded grass (184 kg N ha-1 vs. 257 kg N ha-1). Slurry application increased N2O-emissions significantly, when grass reseeding occurred in spring. Table 1. Cumulative N2O-N-emissions (12-months) following ploughing up of grassland in three different resowing treatments. Different capital letters indicate significant differences among treatments. Different lowercase letters shows significant differences among the fertilizer level (P<0.05). Mean values are shown (n≥3). Accumulated N2O (kg N2O-N ha-1) 12-months after soil tilling occurred Fertilizer Level Year Control 0N 240N mean 0N 240N mean 1. 1. 1. 2. 2. 2. 0.44Aa 1.40Aa 0.94A 0.59Aa 0.92Aa 0.75A Grassland Renovation (Autumn) 21.31Ca 23.89Ca 22.6D 6.24ABCa 4.89ABCa 5.56ABC Grassland Renovation (Spring) 1.96Ba 3.67Bb 2.81B 3.90Ba 5.39Bb 4.64B Grassland Conversion / maize (Spring) 4.06Ba 5.70Ca 4.88C 6.32Ba 8.24Ca 7.28C Conclusion If grassland renovation is necessary due to sward deterioration, ploughing and reseeding should be carried out early in the year to reduce the risk of N2O emissions. Slightly higher emissions due to cultivating high yielding maize instead of grass seem to be acceptable, combined with a delayed reseeding following harvest of maize. References Taube F., Gierus M., Hermann A., Loges R. and Schönbach P. (2014) Grassland and globalization – challenges for north-west European grass and forage research. Grass and Forage Science 69, 2-16. Velthof G.L., Hoving I.E., Dolfing J., Smit A., Kuikman P.J. and Oenema O. (2010) Method and timing of grassland renovation affects herbage yield, nitrate leaching, and nitrous oxide emission in intensively managed grasslands. Nutrient Cycling in Agroecosystems 86, 401-412. Linsler D., Geisseler D., Loges R., Taube F. and Ludwig B. (2013) Temporal dynamics of soil organic matter composition and aggregate distribution in permanent grassland after a single tillage event in a temperate climate. Soil and Tillage Research 126, 90-99. Necpálová M., Casey I., and Humphreys J. (2013) Effect of ploughing and reseeding of permanent grassland on soil N, N leaching and nitrous oxide emissions from a clay-loam soil. Nutrient Cycling in Agroecosystems 95, 305-317. Hutchinson G.L. and Mosier A.R. (1981) Improved soil cover method for field measurement of nitrous-oxide fluxes. Soil Science Society of America Journal 45, 311-316. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 127 Comparing nitrous oxide emissions from white clover-ryegrass pasture with swards receiving applied synthetic fertilizer Hyland J.J., Jones D.L., Chadwick D. and Williams A.P. School of Environment, Natural Resources and Geography, Bangor University, Gwynedd, LL57 2UW, United Kingdom. Corresponding author: afpe69@bangor.ac.uk Abstract Nitrous oxide (N2O) contributes 6% of overall global radiative forcing, and has a global warming potential 298 times greater than that of carbon dioxide over a 100-year time period. Agricultural soils are a major source of N2O, accounting for 35% of annual global emissions. N2O is produced in soils where the available nitrogen (N) exceeds plant requirements through nitrification, denitrification. To reduce the environmental impact of agriculture, it is imperative to implement mitigation measures that are both effective and deemed practical by farmers. The introduction of clover into grassland swards to fix atmospheric nitrogen and hence reduce the need for inorganic fertilizer inputs (and corresponding N2O emissions) is one such measure. The aim of this study is to compare N2O emissions from pasture of differing clover contents, to those of grass-only pastures receiving corresponding N inputs from synthetic fertilizer. N2O emissions are measured by a closed chamber technique from grass-clover swards with high and low clover contents, and from grass-only swards with high and low synthetic fertilizer input. The research hypothesis is that greater clover content will be conducive to achieving more sustainable grassland-based livestock production systems. Introduction Meeting the nutritional needs of an ever-increasing global population will likely create a greater demand for synthetic N fertilizers, consequently increasing N2O emissions (Reay et al., 2011). By assessing current trends in synthetic fertilizer, a 47% increase in global N2O emissions from agricultural soils is forecast by 2020 relative to 1990 levels (US EPA, 2006). Legumes offer many attributes which are conducive to environmentally sensitive agriculture. Characterized by their ability to ‘fix’ atmospheric N, legumes crops are capable of making an important contribution to the future sustainability of grassland systems by displacing dependency on synthetic fertilizers and lowering GHG emissions (Peyraud et al., 2009; Yan et al., 2012). To reduce agricultural GHG emissions, mitigation measures that are easily implemented are more likely to be adopted by farmers. In a recent study comparing 26 mitigation measures, the inclusion of legumes was deemed as being the most practical to adopt by farmers, while adjudged the most effective in reducing GHGs from grassland-based systems by experts (Jones et al., 2013). Grass-white clover systems are attractive at a farm level, as production levels are not compromised if managed appropriately. With clover dry matter (DM) production of ca. 20 - 30%, grass-white clover swards are likely to yield similar DM contents to that of a perennial ryegrass pasture receiving 150 - 200 kg ha-1 yr-1 (Andrews et al., 2007), and 70% of a pasture receiving 350-400 kg ha-1 yr-1 (Andrews et al., 2007; Defra, 2010). Increasing fertilizer costs are likely to make grass-clover systems an attractive alternative in attaining yields (Humphreys et al., 2012). It is estimated that by 2022, the annual UK abatement potential from increased use of legumes could be 0.026 kt N2O (MacLeod et al., 2010). Research postulates that N2O from leguminous residues to be similar to those from the application of synthetic N (Ghosh et al., 2002). The magnitude of emissions, however, is uncertain, with limited research on comparisons of N2O emissions from fertilizer- and white clover-based systems under similar conditions (Ledgard et al., 2009). Emission intensities are a useful and informative method for comparing emissions between pastures of different compositions (Hansen et al., 2014). This study compares N2O emissions from white cloverGrassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 128 based grassland systems by analysing swards of differing clover contents to those of grass pastures receiving corresponding N inputs. From this analysis we propose to determine the optimal sward composition in terms of their respective emissions intensities per unit of product. Methodology The trial is located in the lowland part of Bangor University’s Henfaes Research Station. The experiment is a randomized block design with five treatments and four replicates (Table 1). Each plot measures 25 m2 in area. The treatments are: 1) a control, 2) a grass-clover sward with 10% clover cover, 3) a grass-clover sward with 50% clover cover, 4) a grass-only sward with an application of 180 kg N/ha (the predicted N-fixation from a sward with 10% clover) (Defra, 2010), 5) a grass-only sward with an application of 300 kg N/ha (predicted N-fixation from a sward with 50% clover) (Defra, 2010). Fertilizer is to be applied in 3 split applications throughout the growing season. Nitrous oxide (N2O) fluxes are measured by a closed static chamber technique during the 2014 growing season. The chambers are made of polypropylene and fitted into polyethylene collars, which are inserted 5 cm into the soil at least 24 hours before gas samples are taken. Two chambers are assigned to each plot, with additional chambers allocated to a reference plot. Sampling is conducted weekly with an increase in frequency following N application. On each sampling day, flux measurements are conducted between 09:00 and 12:00. N2O emissions from separate urine-treated plots are also measured to simulate urine deposition from sheep fed the biomass from the five treatments described above. Gas samples are analysed using a gas chromatograph (GC) fitted with an electron capture detector (ECD). N2O concentrations at 0, 20, 40, and 60 min are used to estimate N2O flux (g N ha-1 d-1) for each chamber. Ancillary soil measurements are also made on each sampling date to get an overall picture of different N forms. N2O emissions are expressed as a function of DM yield, or emissions intensities. Results and discussion Both herbage intake, and performance, of livestock is coupled with feed of high clover contents (Wilkins et al., 1994; Ribeiro et al., 2003). Although having advantages in terms of feed characteristics, increased clover content is not correlated to an increase in N2O emissions (Klumpp et al., 2011). The study takes place during the growing season of 2014 and the research hypothesis is that clover content will be conducive in achieving a more sustainable approach to grassland-based livestock production systems. References Andrews M., Scholefield D., Abberton M., McKenzie B., Hodge S. and Raven J. (2007) Use of white clover as an alternative to nitrogen fertilizer for dairy pastures in nitrate vulnerable zones in the UK: Productivity, environmental impact and economic considerations. Annals of Applied Biology 151, 11-23. DEFRA (2010). Fertiliser manual RB209. London, UK: Department for Environment, Food and Rural Affairs. Ghosh S., Majumdar D. and Jain M. (2002) Nitrous oxide emissions from kharif and rabi legumes grown on an alluvial soil. Biology and Fertility of Soils 35, 473-478. Hansen S., Bernard M., Rochette P., Whalen J.K. and Dörsch P. (2014) Nitrous oxide emissions from a fertile grassland in western Norway following the application of inorganic and organic fertilizers. Nutrient Cycling in Agroecosystems 96, 1-15. Humphreys J., Mihailescu E. and Casey I. (2012) An economic comparison of systems of dairy production based on N‐fertilized grass and grass‐white clover grassland in a moist maritime environment. Grass and Forage Science 67, 519-525. Jones A., Jones D., Edwards-Jones G. and Cross P. (2013) Informing decision making in agricultural greenhouse gas mitigation policy: A Best–Worst scaling survey of expert and farmer opinion in the sheep industry. Environmental Science and Policy 29, 46-56. Klumpp K., Bloor J. M., Ambus P. and Soussana J. (2011) Effects of clover density on N 2O emissions and plantsoil N transfers in a fertilized upland pasture. Plant and Soil 343(1-2), 97-107. Ledgard S., Schils R., Eriksen J. and Luo J. (2009). Environmental impacts of grazed clover/grass pastures. Irish Journal of Agricultural and Food Research 48, 209-226. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 129 MacLeod M., Moran D., Eory V., Rees R., Barnes A., Topp C.F.E., et al. (2010). Developing greenhouse gas marginal abatement cost curves for agricultural emissions from crops and soils in the UK. Agricultural Systems 103, 198-209. Peyraud J., Le Gall A. and Lüscher A. (2009) Potential food production from forage legume-based-systems in Europe: An overview. Irish Journal of Agricultural and Food Research 48, 115-135. Reay D. S., Howard C.M., Bleeker A., Higgins P., Smith K., Westhoek H., et al. (2011). Societal choice and communicating the European nitrogen challenge. In: Sutton, M.A. et al. (eds.) The European Nitrogen Assessment, Cambridge: Cambridge University Press, pp 585-601. Ribeiro Filho H., Delagarde R. and Peyraud J. (2003) Inclusion of white clover in strip-grazed perennial ryegrass swards: Herbage intake and milk yield of dairy cows at different ages of sward regrowth. Animal Science 77, 499510. US EPA (2006) Emissions: 1990-2020. Office of Atmospheric Programs Climate Change Division US Environmental Protection Agency, Washington, June 2006. Wilkins R.J., Gibb M., Huckle C. and Clements A. (1994) Effect of supplementation on production by spring‐ calving dairy cows grazing swards of differing clover content. Grass and Forage Science 49, 465-475. Yan M. J., Humphreys J. and Holden N. (2012) The carbon footprint of pasture-based milk production: Can white clover make a difference? Journal of Dairy Science 96, 857-865. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 130 Theme 1 posters Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 131 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 132 Impact of climate change on grassland productivity and forage quality in Austria Poetsch E.M. 1, Asel A. 2, Schaumberger A.1 and Resch R.1 1 Agricultural Research and Education Centre (AREC) Raumberg-Gumpenstein, A-8952 Irdning, Austria 2 University of Natural Resources and Life Sciences, A-1180 Vienna, Austria Corresponding author: erich.poetsch@raumberg-gumpenstein.at Abstract Grassland farming is the main land use system in mountainous regions of Austria. Climate change scenarios assume an increase of temperatures of up to 2 °C for the next few decades and also expect reduced rainfall in summer for the alpine space. This change will have an impact on grassland productivity concerning yield and forage quality. To provide grassland farmers and policy makers with relevant information, a series of 27 field experiments was established along a strong gradient of site and climatic conditions in Austria. The sites were clustered into four different climate groups by means of long-term temperature and precipitation values. Multivariate analysis showed that grassland yield and forage quality were mainly affected by the factors year, climate group and management intensity. In 2003 extraordinary weather conditions with high temperatures and below-average rainfall strongly affected grassland productivity concerning dry matter yield but had no significant impact on forage quality. In arid regions considerable yield losses up to 30% occurred, whereas in humid regions yields even increased partly. Our findings clearly demonstrate that adaptation strategies to climate change on grassland have to consider spatial aspects. Keywords: drought periods, dry matter yield, adaptation strategies, spatial impact Introduction Permanent grasslands of different types cover an area of 1.44 million hectares, which is 50% of the total Austrian agricultural area. There are 60,000 grassland farmers, mostly running small- to medium-size enterprises, who focus mainly on the efficient use of farm manure and the production of high quality forage as the most relevant farm internal resources in mountainous regions. Changes in climatic conditions and climate variability, e.g. extreme weather events (heat waves, droughts, etc.) are likely to occur more frequently in different spatial and time scales in the future (Eitzinger et al., 2009). In order to respond in time it is of great interest for farmers, and also for policy makers, to receive basic information about the regional impact of climate change on grassland yield and forage quality (Meisser et al., 2013). Materials and methods A multi-site field experiment was established by AREC Raumberg-Gumpenstein on 27 different locations in Austria in the year 2002. The experimental design using three replicates included three cutting intensities (2, 3 and 4 cuts year-1) each with an appropriate level of fertilization (0.9, 1.4 and 2.0 LU ha-1) using slurry or stable manure + liquid slurry respectively, with additional mineral nitrogen fertilizer (50 kg ha-1 year-1) for the most intensive variant. The harvesting dates were adapted to the particular site conditions, which varied from 6.4 – 11.1 °C of average yearly temperature, from 548 – 1440 mm of annual precipitation and an altitude from 209 – 1100 m a.s.l. The experimental sites were clustered into four climate groups, based on average temperature and precipitation data. Multivariate statistical analysis were then carried out to identify the most relevant management and site factors which influence dry matter yield and quality parameters. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 139 Results and discussion Grassland dry matter yield was significantly influenced by the factors year, climate group and management intensity, which explained more than 90% of the observed variation. There was no significant difference between the two humid climate groups which showed higher yields than the sites in arid regions, with the lowest yields occurring under warm conditions (Table 1). A multiple comparison of means showed that in three years (2003, 2007 and 2011) of the total project period, significantly lower yields occurred. In 2003, above-average temperatures were combined with below-average rainfall in almost all parts of Austria and caused dramatic damage and losses in grassland and arable farming. Our analysis (Table 2) showed that there were great spatial differences concerning the impact on grassland yield in this extraordinary year (Schaumberger et al., 2012). Under humid/warm conditions no yield reduction was noticed, and under these humid/warm conditions even an increased yield occurred especially when grassland was cut twice (+10%) or three times (+12%) per year. In contrast, a strong average yield decline of 24% in arid/cold regions, and 29% in arid/warm regions was observed with no significant differences between the three tested cutting/fertilization intensities. Table 1. Dry matter yield of grassland under different climate conditions in Austria (average of the period 20022011; a, b – indicate significant differences between climate groups (P<0.05)) Climate groups DM yield (t ha-1 year-1) SD humid/warm n=549 humid/cold n=360 arid/warm n=783 arid/cold n=675 8.10a 8.32a 6.82b 7.55c +/- 2.734 +/- 1.696 +/- 2.723 +/- 2.395 Table 2. Dry matter yield of grassland under different climate conditions in Austria (average of the dry year 2003; a, b – indicate significant differences between climate groups (P<0.05)). Climate groups DM-yield (t ha-1 year-1) SD humid/warm n=63 humid/cold n=27 arid/warm n=81 arid/cold n=72 8.14a 8.96a 4.83b 5.73b +/- 2.457 +/- 0.931 +/- 2.482 +/- 1.816 The crude protein (CP) content was also significantly influenced by the factors year, climate group and management intensity explaining 63% of the variation. On average (2002-2011) the highest CP content was found under humid/warm conditions (121g kg DM-1), the lowest CP concentration occurred in the climate group arid/cold with 115g kg DM-1 (Table 3). Within the climate groups, a rising CP content occurred with increasing management intensity (2 cuts > 3 cuts > 4 cuts). In contrast to the decreasing dry matter yield, the CP concentration in the dry year 2003 was significantly higher in all climate groups (ranging from 133 – 146 g kg DM-1 with significantly highest values under dry/warm conditions) and also for all tested management intensities (Table 4). These results could also be found for the energy concentration (MJ net energy lactation kg DM-1) in forage that was mainly affected by management intensity (2 cuts > 3 cuts > 4 cuts) but to a lesser extent by weather conditions (Table 3). In 2003 the highest values for energy concentration were measured in cold areas with normally low temperatures, and the lowest energy concentration occurred under humid/warm conditions (Table 4). Even though forage quality was just slightly influenced by the extraordinary weather situation in the year 2003, considerable differences occurred for energy yield and crude protein yield under arid conditions, caused by strong yield decline in these climate groups. To provide the present numbers of livestock on farms with sufficient amount of energy and crude protein additional feedstuff is required to bridge this critical Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 140 shortage. Another challenge is to prevent or repair drought damage of the sward by renovation measures using seeds mixtures that are well adapted for dry conditions. Table 3. Crude protein and energy content of forage under different climate conditions in Austria (average of the period 2002-2011; a, b – indicate significant differences between climate groups (P<0.05)). Climate groups humid/warm humid/cold arid/warm arid/cold CP (g kg DM-1) 115.2a 117.7b 118.6bc 120.8c 4.53a 4.79c 4.65b 4.90c MJ NEL (kg DM-1) Table 4. Crude protein and energy content of forage under different climate conditions in Austria (average of the dry year 2003; a, b – indicate significant differences between climate groups (P<0.05)). Climate groups humid/warm n=549 humid/cold n=360 arid/warm n=783 arid/cold n=675 CP (g kg DM-1) 135.2a 134.0a 145.7b 131.9a 4.49a 4.87b 4.84b 4.92b MJ NEL (kg DM-1) Conclusion Our findings clearly indicate that the impact of climate change on grassland productivity in mountainous regions of Austria shows a strong spatial variability and therefore requires different strategies of adaptation. Whereas humid regions with sufficient water supply even benefit from higher temperatures, in arid regions considerable yield losses have to be taken into account. Forage losses can be compensated by the purchase of external feedstuffs for the short term. To counterbalance negative climate change impact on grassland in the long run, there is a need for increased use of seed mixtures which contain drought-tolerant species like lucerne and better-adapted grass and clover cultivars, but the use of irrigation systems also has to be considered seriously. References Eitzinger J., Kersebaum K.C. and Formayer H. (2009) Landwirtschaft im Klimawandel. Auswirkungen und Anpassungsstrategien für die Land-und Forstwirtschaft in Mitteleuropa. Agrimedia GmbH. Clenze, Deutschland. Schaumberger A., Pötsch E.M. and Formayer H. (2012) GIS-based analysis of spatio-temporal variation of climatological growing season for Austria. Grassland Science in Europe 17, 634-636. Meisser M., Deleglise C., Mosimann E., Signarbieux C., Mills R., Schlegel P., Buttler A. and Jeangros B. (2013) Auswirkungen einer ausgeprägten Sommertrockenperiode auf eine montane Dauerweide im Jura. Agrarforschung Schweiz 4, pp 476-483. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 141 Effect of climatic changes on grassland growth, water condition and biomass – the FINEGRASS project Dąbrowska-Zielińska K.1, Goliński P.2, Jørgensen M.3, Mølmann J.3 and Taff G.4 1 Institute of Geodesy and Cartography, Modzelewskiego 27, 02-679 Warsaw, Poland, 2 Department of Grassland and Natural Landscape Sciences, Poznan University of Life Sciences, Dojazd 11, 60-632 Poznan, Poland, 3 Bioforsk Norwegian Institute for Agricultural and Environmental Research, Center for Arctic Agriculture and Nature Use, Holt, 9269 Tromsø, Norway, 4 Norwegian Research Institute of Forestry and Landscape, 9269 Tromsø, Norway Corresponding author: pgolinsk@up.poznan.pl Abstract Project FINEGRASS aims at assessing how climate change affects grassland yield production in Poland and North Norway. It also aims to develop tools for monitoring grassland productivity at the national, regional and the individual grassland scale through the use of the newest and most innovative remote sensing and in-situ based methods. Grass biomass from various farmers’ fields will be related to satellite image analyses of the same fields to build an explanatory and predictive regression model to assess forage biomass based on grass species, environmental conditions and land management. Satellite images will be processed to create intra-year and inter-year time series of derived indices: Normalized Difference Vegetation Index (NDVI) and Enhanced Vegetation Index (EVI) will be taken from available cloud-free Landsat, ASTER, SPOT, and/or Sentinel-2 satellite imagery. Keywords: climate change, grassland productivity, remote sensing Introduction Climate change influences grassland productivity throughout Europe. The extremes of the weather in winter, often lack of snow cover together with low temperatures, as well as often occurrence of the increased air temperatures early in spring, cause shifts in phenology and disturbance in water balance of the grasslands areas, which influence the grass yield. The lack of precipitation and increase of temperature later in spring and summer also cause reductions in moisture, resulting in changing water conditions in some areas. Climate change affects grassland yield production in Poland as well as in North Norway. Winterkill is one of the main obstacles for forage production in North Norway. The projected climate change scenarios indicate that the frequency, degree, and length of winter warming events will increase, and may already have increased. These winter warming events can lead to complete snowmelt during winter, thereby increasing the risk for direct frost on plants due to exposure to ambient air and anoxia damage due to accumulation of ground-ice. In Norway, the growing season has been extended by more than one week during the last 30 years (from 1979-2008, compared with 1961-1990) (Karlsen et al., 2009). In Poland, the most important abiotic factor limiting grassland productivity is water shortage and distribution during the vegetative season. There is a lack of tools for efficiently monitoring the effects of climatic trends on grassland productivity on a regional or national level, and therefore there is a need for developing reliable methods for quantifying yields and collecting data on a large scale. Remote-sensing technology may be applied to approach this. Satellite-based radiometers are useful for measuring vegetation characteristics over time across larger areas. Satellite imagery has been used to assess forage production levels over large areas by calculating the Normalized Difference Vegetation Index (NDVI) (Smit et al., 2008) or the Enhanced Vegetation Index (EVI) Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 142 (Kawamura et al., 2005). This technology may also be used to assess the forage grassland production on farm, regional and national levels. The project FINEGRASS entitled 'Effect of climatic changes on grassland growth, its water conditions and biomass' will be conducted for the period of 36 months (Dec. 2013-Dec. 2016). The project aims at the assessment of the influence of climatic changes on grassland productivity as well as development of innovative tools for grassland management on the national, regional and the individual grassland scale. In the FINEGRASS project, the newest and most innovative remote sensing and in-situ based methods will be applied as well as interdisciplinary research will be conducted. The project partners consider that this project is an answer to the lack of reliable information about grassland growth conditions and the influence of meteorological conditions, on their status available in the existing databases and official documents, which are being elaborated through traditional methods. Materials and methods The overall scope of work as well as the relations between the major research activities within the project are presented on the Figure 1. Figure 1. The major project activities and the relations between them Meteorological data will be gathered from the last 20 years. Years with meteorological anomalies in temperature and precipitation will be noted at each site in Norway and Poland. In North Norway for years with extensive winterkill we will estimate the extent of winterkill using Landsat. To assess if the trend of earlier snowmelt and extended growing season reported in other studies (Karlsen et al., 2009) affect grassland productivity, we will use satellite data (MODISand AVHRR for dates prior to MODIS launch) to estimate dates of snowmelt and greening of vegetation for Tromsø. We will compare these data with historical phenological and grass yield data from the Bioforsk Holt research station in Tromsø. The vegetation indices estimated from NOAA/AVHRR data and MODIS data will be analysed in the Poland study sites, and the anomalies of these indices will be connected to the climatological anomalies. Based on satellite-derived surface temperature the method of estimation of heat fluxes applying the energy balance equation will be used (Dąbrowska-Zielińska et al., 2009). Ground Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 143 measurements from the different managed grasslands (soil moisture, LAI, carbon fluxes and biomass) will be the input into the models that are being derived for the description of grasslands conditions throughout Poland and Norway. To develop models based on NDVI for estimation of yield and protein content in the sward, we will build regression models to assess how NDVI values over the summer period relate to grass yields (ground truthing, as determined by planned field campaigns, and historical data), including information on how the fields were managed. We will incorporate the use of unmanned aerial vehicle (UAV) as a safeguard against the lack of potential satellite imagery for comparison with fieldwork campaigns. Proximal remote sensing combined with chlorophyll measurements will be used to assess protein content at multiple times in the summer, and correlated with satellite, UAV and NDVI. Discussion Climate changes in Poland and Norway may affect the grassland productivity positively and negatively. Higher temperatures with an extended growing season may bring about new opportunities for farmers in the North where forage production is limited by the short growing season. In Poland, this effect allows the possible extension of the grazing season for suckler cows (Goliński et al., 2013). Increased precipitation and more unstable winter conditions may also increase the vulnerability of forage production. Extreme winter warming events may lead to de-hardening of grasses or more problems with ice encasement, thus increasing the risk of winterkill (Jørgensen et al., 2010). Thus, the projected climate scenarios point in different directions. It is therefore necessary to build new and efficient methods that can be used to monitor the productivity of grasslands to understand trends and anomalies that are likely to continue into the future. This can help in planning for agricultural practices and offsetting financial risks on large scales. Acknowledgements The authors would like to thank the Polish-Norwegian Research Programme for co-financing of the FINEGRASS project. References Dąbrowska-Zielińska K., Budzyńska M., Lewiński S., Hoscilo A. and Bojanowski J. (2009) Application of remote and in situ information to the management of wetlands in Poland. Journal of Environmental Management 90, 2261-2269. Goliński P., Golińska B. and Biniaś J. (2013) Effect of extended grazing season of suckler cows on yield, quality and intake of sward. Grassland Science in Europe 18, 273-275. Jørgensen M., Østrem L. and Höglind M. (2010) De-hardening in contrasting cultivars of timothy and perennial ryegrass during winter and spring. Grass and Forage Science 65, 1, 38-48. Kawamura K., Akiyama T., Yokota H., Tsutsumi M., Yasuda T., Watanabe O., Wang G. and Wang S. (2005) Monitoring of forage conditions with MODIS imagery in the Xilingol steppe, Inner Mongolia. International Journal of Remote Sensing 26, 7, 1423-1436. Karlsen S.R., Høgda K.A., Wielgolaski F.E., Tolvanen A., Tømmervik H., Poikolainen J. and Kubin E. (2009) Growing-season trends in Fennoscandia 1982–2006, determined from satellite and phenology data. Climate Research 39, 275–286. Smit H.J., Metzger M.J and Ewert F. (2008) Spatial distribution of grassland productivity and land use in Europe. Agricultural Systems 98, 208-219. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 144 Generating carbon credits from perennial forage species crops in the Mediterranean region: the case of Phalaris aquatica L. Pappas I.A.1,2, Papaspyropoulos K.G.1, Karachristos C.N.3 and Christodoulou A.S.1 1 Laboratory of Forest Economics, Department of Forestry and Natural Environment, Aristotle University of Thessaloniki, 54124, Thessaloniki, Greece. 2 Laboratory of Rangeland Ecology, Department of Forestry and Natural Environment, Aristotle University of Thessaloniki, 54124, Thessaloniki, Greece. 3 F. Fasoulas – N. Mantzios L.P., Yloriki E.E., Thessaloniki, Greece. Corresponding author: pappas@cperi.certh.gr Abstract Agricultural practices, like agroforestry, low/no tillage and planting perennial forage crops, have an important potential to increase carbon sequestration as an option for climate change mitigation. The purpose of this study was to quantify the carbon stored in (a) aboveground biomass and, (b) soil organic matter of forage crops, on sites at different altitudes in Northern Greece, planted with the native Mediterranean perennial grass Phalaris aquatica L., in order to estimate the potential carbon credits. The results showed that P. aquatica, as a perennial forage crop, can provide an alternative agricultural management option for climate change mitigation in the Mediterranean region, due to the large impact on the amount of carbon that can be stored in aboveground biomass and sequestered into the soil. Keywords: soil organic carbon, biomass, forage species, climate change, carbon offsets. Introduction Land use, land-use change and forestry (LULUCF) are a fundamental part of the Kyoto Protocol (KP) (UNFCCC, 1998). Cropland management is an alternative that can be used to achieve the target that each developed country has under the KP (Ovando and Capparos, 2009). In addition, farming systems that increase soil C stocks are fundamental to the sustainability of agricultural production systems. Positive results related to soil C increments, and thus to soil chemical, physical and biological quality, and to mitigation of global warming potential by atmospheric CO2-C removal, have been obtained by adopting forage farming worldwide (Neal et al., 2013). However, the amount of carbon stored in soil and vegetation depends on local site conditions, plant species and land use management. The aim of this research was to quantify the current C stocks aboveground and in the soil of Phalaris aquatica L. crops at two sites with different altitudes in Northern Greece, in order to estimate the potential carbon credits generated from this land use. Materials and methods Biomass samples of the perennial grass Phalaris aquatica L., from rain-fed crops (>10 years old) were collected at the end of the growing season of 2012 and 2013 from two sites in Northern Greece. The sites were at Thermi (40ο30΄N, 23ο4΄E) and Chrysopigi (41ο10΄N, 23ο33΄E) with different altitudes (30 and 650 m, respectively). Biomass production was measured using 10 (1 x 1 m2) plots. Harvested biomass was dried at 60 oC for 48h and milled to 1 mm. The ash-free dry weight (organic matter in biomass) was determined by loss on ignition at 600 oC for 4h (Allen, 1989). Carbon content was calculated by dividing the percentage ash-free dry weight by 1.8 (Koukoura, 1998). Carbon offsets were estimated after the conversion of 1 t of organic carbon to CO2 t (by multiplying by 3.67, i.e., 44/12). Soil samples were collected at two depths, 0-15 cm and 15-30 cm, and soil properties were determined using common analysis methods. Total nitrogen and organic matter concentration Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 145 were measured by the K2Cr2O7 method using the modified Kjeldahl wet digestion procedure of Miller and Keeney (1982). Soil C (t ha-1) was calculated according to the formula: C (t ha-1)= [C (%) x soil depth (m) x 10000 x bulk density (g cm-3) x 10]/1000 (Alifragis, 2008). One-way Analysis of Variance was used to compare means in two sites. Further differences were evaluated with the LSD post hoc test, at P=0.05 (Kinnear and Gray, 2008). Results and discussion Mean organic matter concentration (kg t-1 DM) in aboveground biomass of P. aquatica crops was significantly higher at Chrysopigi than the Thermi site (Table 1). This difference could be attributed to different soil types: at the Chrysopigi site the soil is characterized as sandy loam, while at the Thermi site it is clay loam (Table 2). According to Burval (1997), biomass of the species P. arundinacea grown on high clay content soils had a high amount of ash. Furthermore, there were higher mean carbon stocks (t ha-1 DM) stored in the aboveground biomass of lowland crops than in upland crops, mainly due to significant yield differences (Pappas et al., 2012) rather than differences in the organic carbon concentration (kg t -1 DM), which correspond to higher mean carbon offsets (11.37 t CO2 ha-1). Table 1. Mean values (with s.d.) of ash, organic matter, organic carbon, biomass yield, organic carbon and CO 2 (t ha-1) stored in aboveground biomass of Phalaris aquatica L. crops grown over 2 years on sites at different altitudes. Site Year Ash (g kg-1 DM) Organic matter (kg t-1 DM) C (kg t-1 DM) Biomass yield (t ha-1 DM) C (t ha-1 DM) CO2 (t ha-1 DM) 2012 64 (1.1) 936 (1) 520 (0.6) 4.9 (0.2) 2.55 (0.6) 9.35(0.7) Chrysopigi 2013 87 (0.6) 913 (0.5) 507 (0.8) 4.9 (0.3) 2.49 (0.8) 9.12 (0.3) Mean 76 (13) 924 (13.3) 514 (7.4) 4.9 (0.5) 2.52(7.4) 9.24(1.3) 2012 117 (0.7) 883 (0.6) 491(0.4) 6.2(0.5) 3.04 (0.2) 11.17(0.5) Thermi 2013 98 (0.6) 902 (0.7) 501 (0.6) 6.3(0.4) 3.16 (0.3) 11.59 (0.3) Mean 107 (11) 893 (11) 496 (6.1) 6.25(0.3) 3.10 (0.6) 11.37 (0.2) Table 2. Mean of soil organic matter, organic carbon, organic carbon and CO 2 (t ha-1) sequestered in Phalaris aquatica L. crops grown on sites at different altitudes. Site Depth (cm) Soil type Bulk density (g cm-3) C:Ν C (%) C (t ha-1) CO2 (t ha-1) Chrysopigi (650 m) 0-15 15-30 Sandy Loam Sandy Loam 1.31 (0.17) 1.0 (0.08) 11.6 10.3 1.15 (0.1) 0.93 (0.05) 22.9 (1.98) 14 (0.69) 84.0(7.25) 51.4 (2.62) 0-15 Clay Loam 1.40 (0.15) 11.7 0.82 (0.08) 17.1 (1.85) 62.8 (6.92) Thermi (30 m) 15-30 Clay Loam 1.0 (0.12) 10.8 0.65 (0.07) 9.7 (1.06) 35.7 (3.90) High quantities of total organic carbon (t ha-1) were sequestered in the soil at 0-30 cm depth at both sites (Table 2). The C content was higher in the topsoil (0-15 cm) than in the 15-30 cm depth. The same trend has been reported by several studies evaluating carbon sequestration under different perennial grasses cultivated on agricultural land (Omonode and Vyn, 2006; Clifton-Brown et al., 2007; Christensen et al., 2009). Αccording to Zan et al., (2001) the conversion of agricultural land to perennial forage crops can be expected to increase C stored in above- and belowground biomass and in the soil organic matter because of their perennial nature and greater root biomass production. Total soil carbon sequestration (in t ha-1) was significant higher in the upland (36.9 t ha-1) than the lowland (26.8 t ha-1) crop, corresponding to higher carbon offsets (tonnes of CO2) per hectare (135.4 and 98.5 t ha-1 respectively). This was due to higher C content in both depths, indicating differences in litter decomposition rates between the two sites. Moreover, carbon sequestered in the aboveground biomass, and Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 146 especially in the soil, is stable for a long period of time and fulfils the criteria of additionality, permanence, leakage and double counting in order to be verified as carbon credits in voluntary carbon markets (Murril, 2008; Papaspyropoulos et al., 2013). Conclusion The growing of perennial C3 forage crops is a potential climate-change mitigation activity in the Mediterranean region, as these crops can store and sequester larger amounts of carbon in the aboveground biomass and into agricultural soils, as shown in the case of Northern Greece. These carbon offsets can be efficiently introduced as carbon credits in evolved carbon markets. Acknowledgements The research was supported financially by the Hellenic General Secretariat of Research and Technology (GSRT) in the framework of National Activity 'Cooperation 2011'. References Alifragis D.A. (2008) The Soil, Birth-Properties-Classification, Volume 1, Aivazi Editions, Thessaloniki, Greece, 55pp. Allen S.E. (1989) Chemical Analysis of Ecological Materials, 2nd ed., Blackwell Scientific, Oxford, UK. Burvall J. (1997) Influence of harvest time and soil type on fuel quality in reed canary grass (Phalaris arundinacea L.). Biomass and Bioenergy 12, 149-154. Christensen B.T., Rasmussen J. Eriksen J. and Hansen E.M. (2009) Soil carbon storage and yields of spring barley following grass leys of different age. European Journal of Agronomy 31, 29-35. Clifton-Brown J.C., Breuer J. and Jones M.B. (2007) Carbon mitigation by the energy crop, Miscanthus. Global Change Biology 13, 2296-2307. Kinnear P.R. and Gray C.D. (2008) SPSS 15 Made Simple. Psychology Press. Hove. Koukoura Z. (1998) Decomposition and nutrient release from C 3 and C4 plant litters in a natural grassland. Acta Oecologica 19, 115-123. Miller R.H and Keeney D.R. (1982) Methods of soil analysis. Part 2: chemical and microbiological properties, 2nd edn. American Society of Agronomy, Soil Science Society of America, Madison. 228 pp. Murrill J. (2008) The voluntary carbon offset market: a research-based assessment. Perspectives in Business, 5 (2) 31-36. Neal J.S., Eldridge S.M., Fulkerson W.J., Lawrie R. and Barchia I.M. (2013) Differences in soil carbon sequestration and soil nitrogen among forages used by the dairy industry. Soil Biology & Biochemistry 57, 542-8. Omonode R.A. and Vyn T.J. (2006). Vertical distribution of soil organic carbon and nitrogen under warm-season native grasses relative to croplands in west-central Indiana, USA. Agriculture, Ecosystems and Environment 117, 159-170. Ovando P. and Caparros A. (2009) Land use and carbon mitigation in Europe: A survey of the potentials of different alternatives. Energy Policy 37, 992-1003. Papaspyropoulos K.G., Karachristos C.N., Ioannou K., Pappas I.A., Gounaris N., Mandana V., Theocharis N., Lefakis P.and Christodoulou A.S. (2013) The operation of voluntary carbon offset markets with the implementation of natural and artificial forestry projects. In: Hellenic Forestry Society (eds) Conservationmanagement of Hellenic forests in the economic crisis period and the challenge of natural forestry. The 16th National Forestry Conference, Hellenic Forestry Society,Thessaloniki, Greece, pp. 932-940. Pappas I., Koukoura Z., Kyparissides C., Goulas Ch. and Tananaki Ch. (2012) Phalaris aquatica L. lignocellulosic biomass as second generation bioethanol feedstock. Grassland Science in Europe 17, 448-450. United Nations Framework Convention on Climate Change (UNFCCC), 1998. Kyoto Protocol to the United Nations Framework Convention on Climate Change. United Nations. Zan S.C., Fyles J.W., Girouard P. and Samson R.A. (2001) Carbon sequestration in perennial bioenergy, annual corn and uncultivated systems in southern Quebec. Agriculture, Ecosystems and Environment 86, 135-144. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 147 Agroforestry systems: an option for mitigation and adaptation to overcome global climate change Mosquera-Losada M.R. and Rigueiro-Rodríguez A. Crop Production Departament, Escuela Politécnica superior, Universidad de Santiago de Compostela, 27002-Lugo, Spain Corresponding author: mrosa.mosquera.losada@usc.es Abstract Agroforestry, a combination of a woody (shrub/tree) with an herbaceous component (crops/pasture) is considered an important tool to mitigate and adapt agrarian systems to global climate change. This fact is based on the capacity that AGF systems have to preserve C already accumulated in the woody component but also to increase C sequestration in a tree-less system when trees are planted. Nitrate and CO2 emissions can be reduced by the consumption of the biomass of the understory in forests with high fire risk, but also for the better use of fertilizers in more open systems that will contribute to mitigate the negative effects of this fertiliser inputs on GHG atmosphere release. Resilience is also improved as biodiversity in AGF systems is usually higher than in tree-less systems. Keywords: biodiversity, GHG, carbon sequestration, land use, land management Introduction Agroforestry (AGF) systems include sustainable land management practices that comprise at least two components, one woody (tree /shrub) and one herbaceous (grass /crop including forage), but may also involve livestock as a third component. Man is part of the AGF systems as he manages this semi-natural system attempting to enhance synergies between the different components, in order to achieve optimal use of resources such as light, water and nutrients. When establishing an AGF system, the main ecological factor to consider is the amount of the radiation that can reach the lower strata of the AGF system. Therefore, when an AGF system is established in areas without initial woody cover, we should consider aspects such as the density and distribution of the woody component. If the AGF systems are set on systems with an existing woody component, we focus on optimizing the density of the woody component to enhance the overall productivity (Mosquera-Losada et al., 2009). The use of AGF systems and relationship to reducing the effects of climate change are clear. The Kyoto Protocol highlights that activities related to change of use (reforestation deforestation (Art 3.3) and management of forest lands, agricultural, livestock and revegetation (Art 3.4), can be used to mitigate and reduce emissions of greenhouse gases (GHG) (UN, 1998). Burley et al. (2007) indicate that forest land management can reduce the effect of the emission of GHG through carbon (C) sequestration and through the promotion of reduced emissions from vegetation (reducing fire hazards, deforestation). European countries used these mechanisms (items 3.3., and 3.4) to meet the Kyoto targets (EEA 2009). AGF systems are currently included in Canada and USA, and in European policies to make traditional-recent systems (monocrops) more sustainable (Schoeneberger et al., 2012). Studies with aims of comparing AGF systems or exclusively agricultural land use reflected a clear advantage to the former (Nair et al., 2009; Holwett et al., 2011; Mosquera et al., 2011). Preservation and maintenance of already accumulated carbon AGF systems can contribute to preserve and maintain C stored in terrestrial ecosystems as they can avoid 1) complete deforestation of forestlands to grass or forage lands, and 2) they reduce forest fires. Total deforestation is one of the most important pathways that contribute negatively to the net GHG emissions, mainly because: (i) it involves the release of C stored in the tree Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 148 itself and from the soil and (ii) deforested area stops photosynthesis, thus C fixation is reduced. If AGF system practices (instead of complete deforestation) is established, a high volume of soil could be occupied by tree roots in lower soil horizons than roots of herbaceous crops, even very far away from the tree canopy; e.g. Spanish dehesa with < 50-75 trees /ha can have roots over 50 m from the tree trunk below the rhizosphere of the crop (Howlett et al., 2012; Mosquera-Losada et al., 2012). This is very important because roots are one of the main sources of soil carbon. Soil C represents >65% of the C sequestered in terrestrial ecosystems. Moreover, partial deforestation to increase forage/pasture growth can contribute to the conservation and maintenance of the C accumulated compared to systems that promote total deforestation C, without causing a significant reduction of crop production. For example, tree cover close to 55% results in a decline in crop yields around 50% in temperate conditions (Mosquera-Losada et al., 2009). Further, the use of machinery in AGF systems can be performed with low tree density or with tree distribution in groups or hedges. Implementation of silvopastoral systems reduces fire risk through the direct understory consumption, reducing the existing biomass (load fuel) so avoiding potential C emissions atmosphere. Carbon sequestration promotion Land use change The increase of C in a system caused by the introduction of a tree species in a treeless field is associated with the rate of tree growth (mainly linked to root development at deeper soil layers) and the effect it causes on the understory. It has to be taken into account also that tree species with high growth rates (e.g. poplar) sequester more C per year than those with low growth rates, but the number of years for tree harvest is lower in the fast-growing tree species and therefore the potential of GHG release from the stand and soil is earlier. Promotion of AGF systems instead of simpler (monocrop) systems is usually linked to high levels of biodiversity (alpha biodiversity) because of the spatial heterogeneity caused by the tree (shadow), and the grazing animals (trampling, selective grazing, irregular faeces distribution) at plot level, but also preserves and increase biodiversity (beta biodiversity) at a landscape level in those transhumant systems where woody component is important. It is noteworthy that the promotion and preservation of species biodiversity play a key role to promote adaptation of ecosystems to the impacts of climate change and therefore its resilience. Management Land management can also be a source of GHG. For example, tillage and ploughing as well as fertilization are usually associated with a release of GHG caused by soil mineralization or the inefficient use of fertilizers. However, woody components of AGF systems can absorb the nutrients released by decomposition after ploughing (different N forms) or after fertilizer applications. Modern AGF systems established with Juglans at low density have shown that grazing improves the amount of C sequestered in the soil, when compared with tillage for understory control. C decrease in ploughed areas is probably associated with the lack of vegetation during several months per year. Replacement of materials and fossil fuels One of the main reasons for the increasing atmospheric GHG concentrations is the use of fossil fuels releasing C stored over millions of years. That is why the use of biomass (from AGF activities such as pruning, clearing, thinning, logging) is promoted as a renewable energy source. Such uses may also be undertaken as part of AGF systems, and have a great tradition in the Spanish dehesas, where in order to increase fruit production and shaping the trees, pruning is performed and generates income from land (charcoal). Today, the existence of pruning tools with hydraulic sytems and rising platforms could facilitate this type of use Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 149 because they reduce the costs of labour, a major limiting factor for the use of biomass in developed countries as an energy source. In recent years the use of woody crops for biomass production for energy purposes has been encouraged in different countries of Europe. These systems used woody crops at high densities (alder, eucalyptus, poplar and acacia). Afterwards, mechanized harvesting is performed 3-6 years after development depending on the site quality, and regrowth for a new crop is allowed without cultivation. In Germany, these tree species have been combined with crops (herbaceous species). This type of AGF, called 'alley cropping,' uses strips 10 m wide of woody crop in combination with other strips of 15-20 m of arable crop (Mosquera et al., 2011). Woody species interspersed with crops in alley cropping systems have been shown to contribute significantly to increased C sequestration, and to biodiversity in the soil, compared to crop areas without woody species (Quinkenstein et al., 2009; Matos et al., 2011). Conclusions AGF systems are a land management strategy, which compared with others, may allow improved resilience of agro-ecosystems to climate change impacts. This improvement in the response to climate change is based on their capacity to preserve and maintain the C accumulated, increase the C sequestered and provide materials and fuels to replace fossil fuels. This greater adaptation capacity is mainly based on improvement of biodiversity and better use of resources (nutrients, light and water) compared with other land management. References Burley J., Ebeling J. and Costa P.M. (2007) C sequestration as a forestry opportunity in a changing climate. In: Freer-Smith P.H. et al. (eds.) Forestry and climate change. CAB International, UK, pp. 31-37. EEA (2009) Greenhouse gas emission trends and projections in Europe 2009. Tracking progress towards Kyoto targets. http://www.eea.europa.eu/publications/eea_report_2009_9 (accessed 12 Jan 2011) Howlett D., Moreno G., Mosquera-Losada M.R., Nair P.K.R. and Nair V.D. (2012) Soil carbon storage as influenced by tree cover in the Dehesa cork oak silvopasture of central-western Spain. Journal of Environmental Monitoring 13,1897-1904. Howlett D., Mosquera-Losada M.R., Nair P.K.R., Nair V.D., Rigueiro-Rodríguez A. (2011) Soil carbon storage in silvopastoral systems and treeless pasture in northwestern Spain. Journal Environment Quality 40, 1-8. Mosquera-Losada M.R., Moreno G., Pardini A., McAdam J.H., Papanastasis V., Burgess P.J., Lamersdorf N., Castro M., Liagre F., Rigueiro-Rodríguez A. (2012) Past, present and future of agroforestry systems in Europe. In: Nair P.K.R. and Garrity D. (eds.) Agroforestry - the future of global land use, Advances in Agroforestry 9. Matos E.S., Freese D., Mendonca E.S., Slazak A. and Huttl R.F. (2011) C, nitrogen and organic C fractions in topsoil affected by conversion from silvopastoral to different land use systems. Agroforestry Systems 81, 203-11. Mosquera-Losada M.R., Freese D., Rigueiro-Rodríguez A. (2011) Carbon sequestration in European AGF systems. In: Kumar B.M., Nair P.K.R. (eds.) Carbon sequestration potential of agroforestry systems: opportunities and challenges, Advances in Agroforestry 8. Mosquera-Losada M.R., McAdam J., Romero-Franco R., Santiago-Freijanes J.J., Rigueiro-Rodríguez A. (2009) Definitions and components of AGF practices in Europe. In: Rigueiro-Rodríguez A, McAdam J, MosqueraLosada M.R. (eds). Agroforestry in Europe. Current status and future prospects. Springer, UK, pp 3-20. Nair P.K.R., Kumar B.M., Nair V.D. (2009) Agroforestry as a strategy for C sequestration. Journal of Plant Nutrition and Soil Science 172,10-23. Quinkenstein Q., Wöllecke J., Böhm C., Grünewald H., Freese D., et al. (2009) Ecological benefits of the alley cropping AGF system in sensitive regions of Europe. Environmental Science and Policy 12, 1112-1121. Schoeneberger M., Bentrup G., Gooijer H., Soolanayakanahally R. et al. (2012) Branching out: AGF as a climate change mitigation and adaptation tool for agriculture. Journal of Soil and Water Conservation 67,128-136. UN (1998) Kyoto protocol to the United Nations Framework convention on climate change. Available at http://unfccc.int/resource/docs/convkp/kpeng.pdf (accessed 12 January 2011). Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 150 Drought tolerance of the Lolium multiflorum-Festuca arundinacea introgression forms Perlikowski D.1, Pawłowicz I.1, Zwierzykowski Z.1, Zwierzykowski W.1, Paszkowski E.2 and Kosmala A.1 1 Institute of Plant Genetics of the Polish Academy of Sciences, Strzeszyńska 34, 60-479 Poznań, Poland 2 DANKO Plant Breeding Ltd., Szelejewo Drugie 39, 63-820 Piaski, Poland Corresponding author: akos@igr.poznan.pl Abstract Italian ryegrass (Lolium multiflorum Lam.) has high forage quality but low tolerance to abiotic stresses. Tall fescue (Festuca arundinacea Schreb.) expresses higher tolerance, especially to drought conditions. The hybrids of both species and their introgression derivatives combine the complementary attributes of L. multiflorum and F. arundinacea. The main aim of this study was to evaluate the persistence of BC4 L. multiflorum-F. arundinacea introgression forms after long-term drought in simulated field conditions and their ability to recover after stress termination. Diploid and tetraploid introgression forms, together with Lolium controls, were subjected to tests for drought tolerance under ‘rain-out’ shelters in the field. Fresh matter yield after 14 weeks of drought and re-growth after two weeks of further re-watering for each genotype and average values of these parameters for the whole populations, were estimated. The population of diploid introgression forms demonstrated higher level of drought tolerance, compared to the population of diploid L. multiflorum. However, among the best diploid Lolium controls, genotypes with relatively high drought tolerance also existed, showing even higher values of the analysed parameters compared to the best BC4 forms. The populations of tetraploid introgression forms and controls demonstrated higher average fresh matter yield and re-growth, compared to the diploids. Keywords: Lolium multiflorum, Festuca arundinacea, drought tolerance, introgression Introduction Italian ryegrass (Lolium multiflorum Lam.) has relatively high forage quality but rather poor tolerance to abiotic stresses (Humphreys and Thomas, 1993). Conversely, tall fescue (Festuca arundinacea Schreb.) is characterized by high level of tolerance, especially to drought, but it cannot compete with Lolium in terms of productivity and quality in favourable environmental conditions (Kosmala et al., 2007, 2012). These two species can be hybridized, and their homoeologous chromosomes pair and recombine frequently in the intergeneric hybrids. Thus, it is possible to combine their complementary traits within a single genotype on the way of crossing (Humphreys and Pašakinskienė, 1996; Kosmala et al., 2007; Perlikowski et al., 2013). The current research is a part of the project focused on development of new introgression Lolium cultivars with improved tolerance to water deficit. The main aim of this study was to evaluate the persistence of L. multiflorum-F. arundinacea introgression forms after long-term drought in simulated field conditions and their ability to recover after stress termination. Materials and methods The intergeneric partially fertile pentaploid (2n = 5x = 35) L. multiflorum × F. arundinacea hybrid was produced by crossing autotetraploid (2n = 4x = 28) L. multiflorum with hexaploid (2n = 6x = 42) F. arundinacea. The hybrid was then backcrossed to L. multiflorum cultivars – diploid cv. Tur and tetraploid cv. Lotos. After three subsequent backcrosses of the backcross (BC) plants to diploid or tetraploid Lolium, the introgression BC4 populations were generated. After initial establishment of 100 diploid and tetraploid BC4 plants in the glasshouse, 35 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 151 randomly selected genotypes from each group were subjected to tests for drought tolerance. Simultaneously, 35 genotypes of L. multflorum cv. Tur and 35 genotypes of cv. Lotos, were used as controls. The tests were performed using simulated conditions under ‘rain-out’ shelters at Danko Plant Breeding Ltd., in Szelejewo, Poland. Each shelter was covered with foil protecting it against rainfall and was equipped with soil draining system. In May 2011, three clones of each tested genotype were planted randomly under three distinct rain-out shelters (‘dry’ shelters), and one clone was planted in the field with no protection against rainfall, as a control (‘irrigated’ shelter). During the experiment (14 weeks of drought followed by two weeks of re-watering; June – October 2011), soil humidity under the shelters was monitored. Three harvests were taken during drought period (after four, 10 and 14 weeks of drought) at a cutting height of 8 cm. Fresh matter (FM) yield after 14 weeks of drought and re-growth (RG; scale 0 – 9) after two weeks of re-watering were estimated. Fresh matter yield of each genotype after drought conditions, as well as RG after re-watering, were expressed as a mean score (arithmetic mean of three scores derived from three clones from three distinct ‘dry’ shelters). Differences in FM yield and RG between the genotypes were evaluated using Tukey's HSD tests. Moreover, for each analysed parameter also an average score for the whole population was calculated, involving all the genotypes of this population under ‘dry’ shelters. The measured parameters were treated as the indicators of drought tolerance for particular genotypes and populations. The results of plant selection with reference to drought tolerance of tetraploid introgression forms have been published already. Some selected tetraploid genotypes were used in more detailed physiological and molecular research to find the cell components crucial for tolerance development (Perlikowski et al., 2013). Herein, the results obtained for diploid genotypes (introgression forms and controls), are shown and discussed. Results and discussion The level of soil humidity under the ‘dry’ shelters and the ‘irrigated’ one on the day of drought initiation was similar. It decreased continuously under the dry’ shelters as drought conditions progressed, and after 14 weeks of drought it was lower, compared to the level revealed for the ‘irrigated’ shelter. Two weeks after re-watering initiation, soil humidity under ‘dry’ shelters increased to the level observed under the ‘irrigated’ shelter (data shown in Perlikowski et. al., 2013). The average FM yield per genotype in the population of diploid introgression forms after 14 weeks of drought reached the value of 14.8 g and it was higher, compared to the value calculated for the control population of L. multiflorum cv. Tur (10.6 g). However, a big variation with respect to that trait was also observed among the particular genotypes within each analysed group (the introgression forms and Lolium controls ranged from 1.5 to 35.3 g and from 0.0 to 42.8 g, respectively). Generally, this tendency was also observed for the RG parameter. After two weeks of re-watering, the average RG reached value of 3.9 (range: 1.6 – 6.7) and 2.9 (range: 0.3 – 8.3) per genotype for the introgression forms and Lolium controls, respectively. The values of FM yield and RG for the clone present under ‘irrigated’ shelter were always higher, compared to the average values calculated for the corresponding population under ‘dry’ shelters. As demonstrated earlier (Perlikowski et al., 2013), the populations of tetraploid introgression forms and L. multiflorum cv. Lotos were shown to be better with respect to the average FM yield after 14 weeks of drought, and the average RG parameter after two weeks of re-watering, compared to the diploids. The average FM yield reached the value of 31.7 g and 19.9 g, and re-growth – 5.2 and 4.0 per genotype for the population of tetraploid introgression forms and controls, respectively. The comparisons with relation to the RG parameter among the best genotypes selected within the populations of diploid and tetraploid introgression forms and controls are shown in Figure 1. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 152 re-growth [0-9] 8 6 4 2 0 Lm (2x) BC4 (2x) Lm (4x) BC4 (4x) the best genotypes within analysed populations 'dry' shelter 1 'dry' shelter 2 'dry' shelter 3 mean for three 'dry' shelters" average value per genotype for the whole population Figure 1. Re-growth after two weeks of re-watering (scale 0 – 9) of the selected diploid and tetraploid BC4 L. multiflorum-F. arundinacea introgression forms and L. multiflorum (Lm) controls. Conclusions The analysed population of diploid L. multiflorum-F. arundinacea introgression forms demonstrated higher level of drought tolerance, compared to the population of diploid L. multiflorum cv. Tur, manifested by higher average FM yield after 14 weeks of drought and higher average RG after re-watering per genotype. However, among the best diploid Lolium controls, genotypes with relatively high drought tolerance also existed, showing even higher values of the analysed parameters, compared to the best BC4 forms. As had been expected, the populations of tetraploid introgression forms and controls demonstrated higher average FM yield and RG, compared to the diploids. Acknowledgements The research was funded by the Polish Ministry of Agriculture and Rural Development (grant no. HOR hn-801-19/13). References Humphreys M.W. and Thomas H. (1993) Improved drought resistance in introgression lines derived from Lolium multiflorum × Festuca arundinacea hybrids. Plant Breeding 11, 155-161. Kosmala A., Perlikowski D., Pawłowicz I. and Rapacz M. (2012) Changes in the chloroplast proteome following water deficit and subsequent watering in a high and a low drought tolerant genotype of Festuca arundinacea. Journal of Experimental Botany 63, 6161-6172. Kosmala A., Zwierzykowski Z., Zwierzykowska E., Łuczak M., Rapacz M., Gąsior D. and Humphreys M.W. (2007) Introgression-mapping of the genes for winter hardiness and frost tolerance transferred from Festuca arundinacea into Lolium multiflorum. Journal of Heredity 98, 311-316. Perlikowski D., Kosmala A., Rapacz M., Kościelniak J., Pawłowicz I. and Zwierzykowski Z. (2013) Influence of short-term drought conditions and subsequent re-watering on the physiology and proteome of Lolium multiflorum/Festuca arundinacea introgression forms with contrasting levels of tolerance to long-term drought. Plant Biology doi: 10.1111/plb.12074. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 153 Effect of water stress on Lotus corniculatus L. nutritive value at different stages of maturity Georganta A.1, Parissi Z.M.2, Kyriazopoulos A.P.1, Abraham E.M.2 and Lazaridou M.3 1Department of Forestry and Management of the Environment and Natural Resources, Democritus University of Thrace, 193 Pantazidou str., 68200 Orestiada, Greece 2Laboratory of Range Science (236), Dept. of Forestry and Natural Environment, Aristotle University of Thessaloniki, 54124 Thessaloniki, Greece 3 Technological Educational Institute of Kavala, Faculty of Agriculture, Dept. of Forestry, 66100 Drama, Greece Corresponding author: pz@for.auth.gr Abstract The aim of this study was to estimate the effect of moderate water stress on the nutritive value of Lotus corniculatus L. at three phenological stages. Plants of a natural population from Drama, Greece were tested under two irrigation levels: 1) up to field capacity, and 2) 40% of field capacity. The plants were harvested at three phenological stages: early vegetative, flowering and start of fruit formation. Samples were analysed for Crude Protein (CP), Neutral Detergent Fibre (NDF), Acid Detergent Fibre (ADF) and Acid Detergent Lignin (ADL). Dry matter digestibility (DMD) was calculated. Moderate water stress led to decreased fibre concentration and increased the digestibility. It reduced CP content only at the early vegetative stage. It can be concluded that moderate water stress had slightly affected the nutritive value of Lotus corniculatus and probably in a positive way. Keywords: legumes, drought, phenological stages, feed quality Introduction Adequate water supply is a crucial factor for grassland forage production (Hopkins and Del Prado, 2007). It is well known that forage legumes differ in drought-stress sensitivity (Dierschke and Briemle, 2002) and limited water supply can have strong effects on their production (Foulds, 1978). However, knowledge about the influence of water stress on the nutritive value of legumes is limited and inconsistent (Kuchenmeister et al., 2013). Lotus corniculatus L. is a promising drought-resistant forage legume of high nutritive value (Escaray et al., 2012). Additionally, condensed tannins present in its leaves prevent bloating in ruminants and protect plant proteins in the rumen from degradation (Waghorn et al., l987). Its overall forage quality under drought conditions is better than that of Medicago sativa due to higher leaf:stem ratio, and delayed maturity (Peterson et al., l992). The aim of the present study was to investigate the effect of water stress to the nutritive value of L. corniculatus in different phenological stages. Materials and methods Plants of a natural population of L. corniculatus were collected from Drama, northern Greece, at various elevations (from 100-500 m) in September 2012. Thereafter, they were transplanted into plastic pots filled with organic matter at the farm of the Aristotle University of Thessaloniki, northern Greece, altitude 10 m asl. The climate of the area could be characterized as Mediterranean semiarid (mean annual precipitation 443 mm; mean annual temperature 15.5ο C). For this experiment, 32 uniform plants were selected and transplanted into pots (diameter 16 cm, height 45 cm), filled with medium-texture soil and placed under a transparent shelter in spring 2013. Drip irrigation was applied in two levels: 1) up to field capacity (W), and 2) 40% of field capacity (WL). Pots were placed in a completely randomized design. The plants were harvested three times at 4 cm above the soil surface, at the following phenological stages: 1st Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 154 cutting: 21 May (early vegetative), 2nd cutting: 5 June (flowering) and 3th cutting: 28 June (start of fruit formation). Aboveground biomass from every individual plant was oven-dried at 60o C for 48 h, ground through a 1 mm screen and analysed for Neutral Detergent Fibre (NDF), Acid Detergent Fibre (ADF), Acid Detergent Lignin (ADL) (Van Soest et al., 1991) using the ANKOM fibre analyser (ANKOM Technology Corporation, Macedon, NY, USA). Nitrogen was determined using the Kjeldahl procedure (AOAC, 1990), and crude protein was calculated as N content × 6.25. Dry matter digestibility (DMD) was calculated using the equation proposed by Oddy et al. (1983): DMD %=83.58-0.824 ADF% +2.626 N%. Two-way ANOVA of the data was performed using SPSS® statistical software v. 18.0 (SPSS Inc., Chicago, IL, USA), in order to determine differences among the phenological stages and water treatments. The LSD at the 0.05 probability level was used to detect the differences among means (Steel and Torrie, 1980). Results and discussion NDF, ADF and ADL (across phenological stages) of limited watered plants were significantly reduced and, as a consequence, the DMD was significantly increased, while the CP was not significantly affected (Table 1). ADF (across water-stress treatment) was significantly increased from the early vegetative to the fruit formation stage, while the opposite trend was recorded for CP and DMD (Table 1). Table 1. Chemical composition (g kg-1 DM) and DMD (%) of Lotus corniculatus samples in two irrigation levels and three phenological stages Water levels CP NDF ADF ADL DMD W 182.5a* 423.8a 286.6a 90.1a 68.5b WL 178.4a 385.8b 254.1b 83.6b 70.1a Phenological stage Early vegetative 208.8a 396.1a 259.7b 85.9a 71.1a Flowering 169.9b 414.8a 292.8a 89.2a 69.3b Fruit formation 162.6b 403.4a 291.6a 85.4a 67.4c * Different letters in each column within irrigation level or phenological stage indicate significance at P ≤ 0.05 NDF and ADL were not affected by the vegetative stage. Significant interaction was observed between the water treatment and the phenological stages only for CP (Figure 1). CP content was significantly higher in the well-watered plants only during the early vegetative stage. CP (g/kg DM) 250 200 150 100 b CP WL a c c c CP W c 50 0 Early vegetative Flowering Fruit formation Phenological stages Figure 1. Crude protein (g kg-1 DM) of Lotus corniculatus samples at three phenological stages and two irrigation levels: W (up to field capacity) and WL (40% of field capacity) Similar to our results, Kuchenmeister et al. (2013) reported that moderate water stress reduced NDF and ADF, while it had no severe effects on CP content. Many interacting factors including the stage of plant development, leaf:stem ratio, environmental conditions or availability of nutrients could affect fibre concentration (Buxton, 1996, Fulkerson et al., 2007). The availability of N depends on N fixation but the concentration in the plant also depends on the Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 155 amount of biomass production. Kuchenmeister et al. (2013) explained the reduced CP concentration under strong water stress to reduced N fixation. It seems that moderate stress did not affect N fixation. However, the lower CP content of water-stressed plants at the early vegetative stage could be associated with higher biomass production at this phenological stage. Conclusion Moderate water stress led to decrease fiber concentration and increase the digestibility. Moreover, it reduced CP content only in early vegetative stage. It can be concluded that moderate water stress had slightly affected the nutritive value of Lotus corniculatus probably in a positive way. Acknowledgments This research was co-financed by the European Union (European Social Fund – ESF) and Greek national funds (ARCHIMEDES III). References AOAC (1990) Official Methods of Analysis, 15th edn. Washington DC, USA: AOAC, p. 746. Buxton D.R. (1996) Quality-related characteristics of forages as influenced by plant environment and agronomic factors. Animal Feed Science and Technology 59, 37-49. Dierschke H. and Briemle G. (2002) Kulturgrasland. Wiesen, Weiden, und verwandte Staudenfluren. Ulmer Verlag, Stuttgart, pp. 212-217. Escaray F.J., Menendez A.B., Garriz A., Pieckenstain F.L., Estrella M.J., Castagno L.N., Carrasco P., Sanjuan J. and Ruiz O. (2012) Ecological and agronomic importance of the plant genus Lotus. Its application in grassland sustainability and the amelioration of constrained and contaminated soils. Plant Science 182, 121-133. Foulds W. (1978) Response to soil moisture supply in three leguminous species I. Growth, reproduction and mortality. New Phytologist 80, 535-545. Fulkerson W.J., Neal J.S., Clark C.F., Horadagoda A., Nandra K.S. and Barchia I. (2007) Nutritive value of forage species grown in the warm temperate climate of Australia for dairy cows: Grasses and legumes. Livestock Science 107, 253-264. Hopkins A. and Del Prado A. (2007) Implications of climate change for grassland in Europe: Impacts, adaptations and mitigation options: a review. Grass and Forage Science 62, 118-126. Küchenmeister K., Küchenmeister F., Kayser M., Wrage-Mönnig N. and Isselstein J. (2013) Influence of drought stress on nutritive value of perennial forage legumes. International Journal of Plant Production 7, 693-710. Oddy V.H., Robards G.E. and Low S.G. (1983) Prediction of in vivo matter digestibility from the fiber nitrogen content of a feed. In: G.E. Robards and R.G. Pakham (eds.) Feed information and animal production. Commonwealth Agricultural Bureaux, Australia, pp. 395-398. Peterson P.R., Sheaffer C.C. and Hall M.H. (1992) Drought effects on perennial forage legume yield and quality. Agronomy Journal 84, 774-779. Steel R.G.D. and Torrie J.H. (1980) Principles and procedures of statistics, 2nd edn. New York, USA: McGrawHill, 481 pp. van Soest P.J. Robertson J.B. and Lewis B.A. (1991) Methods for dietary fiber, neutral detergent fiber, and non starch polysaccharides in relation to animal nutrition. Journal of Dairy Science 74, 3583-3597. Waghorn G.C., Ulyatt M.J., John A. and Fisher M.T. (1987) The effect of condensed tannins on the site of digestion of amino acids and other nutrients in sheep fed on lotus. British Journal of Nutrition 57, 115-126. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 156 Impact of limited irrigation on water economy and photosynthetic performance of Lotus corniculatus Kostopoulou P.1, Karatassiou M.1, Lazaridou A.2, Lazaridou M.2 and Patakas A.3 1 Laboratory of Range Ecology, Department of Forestry and Natural Environment, Aristotle University of Thessaloniki, 54124 Thessaloniki, Greece, 2 Department of Forestry, Faculty of Agriculture, TEI of East Macedonia and Thrace, 66100 Drama, Greece, 3 Department of Business Administration of Food and Agricultural Enterprises, University of Patras, G. Seferi 2, 30100 Agrinio, Greece Corresponding author (P. Kostopoulou): giotakos@for.auth.gr Abstract Spatial and temporal aspect of drought phenomena in Mediterranean areas force plants to grow under conditions of water deficit. Resistance of a plant species to drought is usually expressed by higher water-use efficiency and maintenance of productivity at high levels. The aim of this study was to evaluate the physiological responses of Lotus corniculatus under water-stress conditions. Plants from a natural population of a semi-arid area of northern Greece (Drama) were selected and transplanted to pots. After a period of plant adjustment, two irrigation regimes were used: a) irrigation up to field capacity, and b) partial irrigation in order to maintain water deficit conditions in the soil. Measurements of water potential, assimilation rate, transpiration rate, stomatal conductance and chlorophyll fluorescence were conducted at four growth stages. Water-use efficiency was estimated as the ratio of assimilation to transpiration rate. Our results showed that limited irrigation of L. corniculatus did not significantly affect the photosynthetic and photochemical performance, as well as the wateruse efficiency. Therefore, L. corniculatus could be a suitable and valuable candidate for improving grassland vegetation in dry areas. Keywords: photosynthesis, stomatal conductance, water use efficiency, Lotus corniculatus Introduction Drought is a common phenomenon in Mediterranean areas, especially during summer, and has many adverse impacts on plants, inhibiting their growth (Asgharipour and Heidari, 2011). The effect of drought on growth and yield of a species depends upon the severity, duration and timing of water deficit. The degree of plant response to water stress varies among species. Plants develop morphological and physiological adaptation mechanisms, such as reduction of leaf area, stomatal closure and regulation of water potential, in order to minimize water loss and, therefore, to survive during periods of water stress (Passioura, 1997; Sarker et al., 2004; Karatassiou et al., 2009). Understanding these mechanisms and controlling water economy and plant resistance to drought could increase food productivity and quality. Available knowledge on the effect of drought stress on Lotus corniculatus is limited. Lotus corniculatus is a worldwide-distributed species that grows under a wide range of environmental conditions and can be found in different microclimates of northern Greece. It is considered to be of high palatability, suitable for vegetation improvement of grasslands, and is usually used for ecological restorations of soils affected by nutrient deficiency, salinity, drought, or contaminants (Escaray et al., 2012). The aim of the research reported in this paper was to determine the physiological responses of L. corniculatus under conditions of water deficit. Materials and methods Lotus corniculatus plants were collected from three different locations of Drama (TEI-labor houses, Horisti, Lydia) northern Greece in September 2012 and transplanted into small pots. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 157 In March 2013, 32 plants were transferred to larger pots (16 cm diameter and 45 cm height), filled with soil of medium texture. The pots were placed under a transparent shelter in the experimental area. Microclimatic conditions of the experimental area are shown in Table 1. Drip irrigation was applied at two levels: full irrigation up to field capacity (FI) and limited irrigation (40% of FI) (LI). We followed a completely randomized experimental design with four replicates. Measurements were performed during spring 2013 on four different dates corresponding to four growth stages: early vegetative, vegetative, flowering and start of fruit formation. The measurements were carried out between 9.30 and 12.00. Water potential (Ψ) of the upper part of the stem was measured, using a pressure chamber (SKPM 1400, Skye Instruments Ltd, Llandrindod Wells, UK) as the leaf of L. corniculatus is too small to be measured. Assimilation rate (A), stomatal conductance (Gs) and transpiration rate (E) were measured with a portable photosynthesis system (LCpro-SD, ADC Bioscientific Ltd, Hoddesdon, UK) on the abaxial leaf surface. Water-use efficiency (WUE) was estimated as the ratio of assimilation to transpiration rate. The ratio of variable to maximum chlorophyll fluorescence (Fv/Fm) was measured at dark-adapted for 20 min leaves, using a chlorophyll fluorometer (OS 30p+, OptiSciences Inc, Hudson, USA). Statistical analysis was performed using the SPSS statistical package (SPSS for Windows, standard version, release 17.0; SPSS, Inc., Chicago, USA). Analysis of variance (ANOVA) was used to compare the irrigation treatments throughout the growing season and at each plant growth stage. A significance level of 5% was used throughout. Results and discussion Among the physiological parameters tested, irrigation treatment significantly affected only the L. corniculatus water potential. Throughout the growing season, L. corniculatus plants under limited irrigation had significantly lower water potential (-1.18 MPa) (Table 1) compared to plants under full irrigation (-0.59 MPa). Plants under full irrigation exhibited higher but not significantly different values of assimilation rate, stomatal conductance, transpiration rate, WUE and Fv/Fm compared to plants under limited irrigation (Table 1). Table 1. Microclimatic conditions of the experimental area during the growing period. Dates 1 2 3 4 Temperature (oC) 32.6 27.7 24.3 30.6 RH (%) PPFD (μmol m-2s-1) VPD (KPa) 13.8 19.7 33.5 38.5 872 941 1304 822 4.24 2.98 2.02 2.70 Assimilation rate, stomatal conductance and WUE varied significantly during the different growth stages, being lower at the end of the growing season (data not shown), probably corresponding to the altered needs of the plant at each growth stage and/or to the environmental conditions. In addition, water potential was maintained at high levels until the middle of the growing season, sharply decreasing towards the end. In contrast, the transpiration rate and the chlorophyll fluorescence ratio Fv/Fm remained unchanged throughout the various growth stages. Our results showed that limited irrigation had rather a small effect on the photosynthetic mechanism of L. corniculatus, because the plants maintained their photochemical efficiency and their stomatal apparatus open, and therefore they continued to photosynthesize despite the low water-potential values. Nowadays, it is recognized that some plants are adapted to photosynthesize even under low water potential, using mechanisms that tend to maintain turgor, producing and conserving osmolytes in order to protect tissues from dehydration (Berkowitz, 1998; Jones, 2004). Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 158 Conclusion Limited irrigation did not significantly affect the photosynthetic and photochemical performance, as well as the water use efficiency, of L. corniculatus plants. Therefore, L. corniculatus could be considered as a suitable and valuable candidate for improving grassland vegetation in dry areas or for water-saving cultivation. Table 2. Lotus corniculatus physiological responses under two different irrigation treatments throughout the growing season (n=16). Treatment Full Irrigation Limited Irrigation Significance Physiological parameters Gs (mol m-2s-1) Ψ (MPa) 0.06±0.01 -0.59±0.23 A (μmol m-2s-1) 4.70±0.78 E (mmol m-2s-1) 1.71±0.30 3.81±0.71 1.49±0.13 0.04±0.01 ns ns ns WUE 3.06±0.50 Fv/Fm 0.81±0.01 -1.18±0.33 2.43±0.37 0.79±0.01 * ns ns Significance: * P<0.05; ns, not significant. Acknowledgments This research has been co-financed by the European Union (European Social Fund – ESF) and Greek national funds through the Operational Programme "Education and Lifelong Learning" of the National Strategic Reference Framework (NSRF) - Research Funding Programme: ARCHIMEDES III. Investing in knowledge society through the European Social Fund. References Asgharipour M.R. and Heidari M. (2011) Effect of potassium supply on drought resistance in sorghum: plant growth and macronutrient content. Pakistan Journal of Agricutlural Science 48(3), 197-204. Berkowitz G.A. (1998) Water and salt stress. In: Raghavendra A.S. (ed) Photosynthesis, a comprehensive treatise. Cambridge University Press, Cambridge. pp. 226-237. Escaray F. J., Menendez A. B., Garriz A., Pieckenstain F. L. et al. (2012) Ecological and agronomic importance of the plant genus Lotus. Its application in grassland sustainability and the amelioration of constrained and contaminated soils. Plant Science 182, 121-133. Jones H. (2004) What is water use efficiency? In: Bacon M.A. (ed) Water use efficiency in plant biology. Blackwell Publishing, Oxford. pp. 27-41. Karatassiou M., Noitsakis B. and Koukoura Z. (2009) Drought adaptation ecophysiological mechanisms of two annual legumes on semi-arid Mediterranean grassland. Scientific Research and Essays 4, 493-500. Passioura J.B. (1997) In: Belhassen E. (ed) Drought tolerance in higher plants: genetical, physiological and molecular biological analysis. Kluwer Academic Publishers, The Netherlands, pp.1-6. Sarker B.C., Hara M. and Uemura M. (2004) Comparison of response of two C3 species to leaf water relations, proline synthesis, gas exchange and water use under periodic water stress. Journal of Plant Biology 47, 33-41. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 159 Drought resistance of selected forage legumes for smallholder farmers in East Africa Wrage-Mönnig N.1,5, Mutimura M.2, Kigongo J.3, Paul B.K.4, Isselstein J.5 and Maass B.L.4 1 Faculty of Life Sciences, Rhine-Waal University of Applied Sciences, Kleve, Germany 2 Rwanda Agriculture Board, Kigali, Rwanda 3 National Livestock Resources Research Institute, Tororo, Uganda 4 International Center for Tropical Agriculture, Nairobi, Kenya 5 Institute of Grassland Science, Department of Crop Sciences, University of Goettingen, Goettingen, Germany Corresponding author: Nicole.Wrage@hochschule-rhein-waal.de Abstract To improve feed availability for smallholder farmers in East Africa, we investigated the biomass production and drought resistance of five forage legumes in semi-arid environments of Rwanda and Uganda. The crops were grown under rain-fed conditions (control) and with additional irrigation (irrigated) and harvested at two-monthly intervals four to five times. Before harvests, the youngest leaves of the crops were sampled for stable carbon isotope analysis for an indication of intrinsic water-use efficiency. This is the first time carbon isotope data are reported for these species. The total annual dry matter production was larger for the legumes grown in Uganda than in Rwanda. However, there were large differences among harvests, with the variability between smallest and largest harvests being similar for both countries. Carbon isotopic signatures were more enriched for samples from Rwanda, hinting at a larger intrinsic water-use efficiency under the local conditions. Canavalia brasiliensis had most enriched carbon signatures in both countries, coupled with acceptable biomass production, and should be further investigated for adaptation in smallholder farming systems. Keywords: yield, stable isotopes, δ13C, irrigation, intrinsic water-use efficiency, forage Introduction Having access to reliable forage of sufficient quality, especially during the dry season, still poses the main challenge to smallholder dairy producers in semi-arid areas of East Africa (Hall et al., 2007). Typically, the amount of milk gained as well as the length of the reproductive cycle of the animals depends on the amount and quality of forage available. To overcome the difficulties, zero-grazing systems have been developed. In contrast to most other crops, many forage species can also be grown on marginal lands and thus provide an opportunity for farmers to build a livelihood. Legumes can offer the extra benefit of improving the nitrogen-poor soils. However, drought-resistance is an important key trait of the plants to be used successfully. Therefore, in this study, five forage legumes were tested for their ability to provide biomass in field trials in Uganda and Rwanda. Measurements of the stable carbon isotopes (δ13C signatures) were used to assess the drought resistance of the legumes. The 13C signatures are directly proportional to the intrinsic water-use efficiency, i.e. the amount of carbon assimilation per stomatal conductance for water (Farquhar et al., 1989). We hypothesized that there would be differences among the legumes in biomass production during wet and dry conditions and in intrinsic water-use efficiency. Materials and methods Five forage legumes, namely Lablab purpureus, Desmodium uncinatum cv. Silver leaf, Desmanthus virgatus, Macroptilium bracteatum cv. Burgundy bean, and Canavalia brasiliensis, were grown in a completely randomized block design (plots of 3 m x 6 m, 1 m in between plots) with five replicates, with or without additional irrigation (by hand, if no Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 160 precipitation had fallen the previous day) in field sites in Uganda (National Livestock Resources Research Institute, Tororo district; annual rainfall 1130-1720 mm, AATF, 2009) and Rwanda (Rwanda Agriculture Board, Karama Research Station, Bugesera district; annual rainfall 845 mm, REMA, 2007). Planting was done in the rainy season (October 2012) at recommended rates and spacing. Four (Rwanda) or five (Uganda) harvests took place at twomonthly intervals until June 2013. Biomass of 1 m² was harvested 10 cm above the ground, and samples of about 200 g were oven-dried (60 °C for 48 hours) and weighed. Just before harvest, the youngest leaf of several plants was sampled for stable isotope analysis. Isotope measurements were done on an isotope ratio mass spectrometer (Delta plus Finnigan MAT, Bremen, Germany) coupled to an elemental analyser (NA2500 CE lnstruments, Rodano, Milano, ltaly) via an interface (Conflo lll Thermo Electron Cooperation. Bremen, Germany). Isotopic values are given as δ13C values (standard: V-PDB): δ13C (‰) = 1000 (13C/12Csample – 13 12 C/ Cstandard)/(13C/12Cstandard). Statistical analyses (ANOVA, repeated measures GLM per country, testing for normality and homogeneity of variances) were done with SPSS version 16. Results and discussion Average dry matter yields were significantly larger in Uganda than in Rwanda (Table 1). There were no significant differences in average biomass production among forage legumes or irrigation treatments; standard deviations were very large. The individual harvests showed large differences in biomass production within species (Fig. 1), with larger harvests in the rainy seasons (from March to May and September to November) than in the intermediate dry seasons, especially without extra irrigation. Individual harvests from Rwanda produced as much biomass as those from Uganda with similar variability (Fig. 1). Table 1. Average dry matter yields of the five tested legumes over four (Rwanda) or five (Uganda) harvests (g m 2 ). There were significant differences between countries, but not among species or irrigation treatments within countries. Rwanda Irrigated Not irrigated Uganda Irrigated Not irrigated L. purpureus D. uncinatum D. virgatus M. bracteatum Average dry matter yield [g m-2] C. brasiliensis 277 280 296 285 191 316 290 309 188 385 701 653 530 447 602 625 542 508 704 634 The carbon isotopic signatures were more enriched in samples from Rwanda than from Uganda (Fig. 1). This hints at a larger intrinsic water-use efficiency of the plants under conditions in Rwanda. As the biomass per harvest was not necessarily different between sites (Fig. 1), the larger efficiency was probably due to a smaller stomatal conductance, maybe due to climatic conditions or soil fertility. There were no clear relationships between biomass produced and δ13C signatures for individual species, per country, per irrigation treatment or over all data. There were, however, significant differences in isotopic signatures among species within countries. In both countries, irrespective of irrigation, C. brasiliensis had most enriched isotopic values, while at the same time producing a relatively good harvest at most times (Figure 1). This suggests an efficient intrinsic water use of this forage legume that should be further investigated, especially regarding the variability of biomass production over time and its relation to the fertility level of the soils. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 161 Figure 1. Carbon stable isotope signatures (δ13C in ‰) versus dry matter yield (g m-2) of five forage legumes over the first four harvests after sowing in Rwanda (R, closed symbols) and Uganda (U, open symbols). Shown are means per species and harvest with or without extra irrigation (eight points per species). Macroptilium bracteatum did not produce harvestable biomass during the first harvest in Rwanda Conclusion Despite a similar biomass production per individual harvest, stable carbon isotope signatures of the forage legumes tested were influenced by growing conditions in the countries, suggesting an influence on intrinsic water-use efficiency. Differences among species in biomass production and intrinsic water-use efficiency can be further exploited. Acknowledgements This study was funded by the German Federal Ministry for Economic Cooperation and Development (BMZ) through a grant to the International Center for Tropical Agriculture. References AATF [African Agricultural Technology Foundation] (2009) Baseline study of smallholder farmers in Striga infested maize growing areas of Eastern Uganda. Nairobi, Kenya: African Agricultural Technology Foundation. Farquhar G.D., Ehleringer J.R. and Hubick K.T. (1989) Carbon isotope discrimination and photosynthesis. Annual Review of Plant Physiology 40, 503-537. Hall A., Sulaiman V.R. and Bezkorowajnyj P. (2007) Reframing technical change: Livestock fodder scarcity revisited as innovation capacity scarcity. UNU-MERIT Working Paper 2008-004. United Nations University – Maastricht Economic and Social Research and Training Centre on Innovation and Technology, Maastricht. REMA [Rwanda Environment Management Authority] (2007) Pilot integrated ecosystem assessment of Bugesera, Prepared for the UNEP/UNDP/GOR Poverty and Environment Initiative Project (PEI). Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 162 Water use efficiency of tall fescue (Festuca arundinacea Schreb.) and perennial ryegrass (Lolium perenne L.) under different management intensity Pardeller M.1,2, Schäufele R.2, Pramsohler M.1 and Peratoner G.1 1 Research Centre for Agriculture and Forestry Laimburg, I-39040 Auer, Italy. 2 Lehrstuhl für Grünlandlehre, Technische Universität München, Alte Akademie 12, D-85350 Freising-Weihenstephan, Germany. Corresponding author: Giovanni.Peratoner@provinz.bz.it Abstract Drought periods have been repeatedly observed in the last decade in the southern margin of the Alps. An adequate choice of drought-tolerant forage species and cultivars is of pivotal importance in tackling this challenge. Tall fescue (Festuca arundinacea Schreb.) and perennial ryegrass (Lolium perenne L.) are regarded as having contrasting levels of drought tolerance, but less is known about their intrinsic water use efficiency (Wi). To this end, analyses of the Carbon isotope composition of leaf material were performed on these species (one cultivar of perennial ryegrass, and two cultivars of tall fescue differing in leaf roughness) in the course of a field trial aiming at the optimization of seed mixtures for permanent meadows in droughtendangered areas under different management intensities and at two altitude ranges. W i was found to be mainly affected by both genotype and altitude, with tall fescue showing higher Wi than perennial ryegrass. Keywords: drought, Festuca arundinacea, Lolium perenne, intrinsic water-use efficiency Introduction Drought periods were repeatedly observed in the last decade in the southern margin of the Alps. The use of forage species and cultivars able to tolerate drought and efficiently use the water available from precipitation and irrigation represents an important issue for tackling this challenge. Perennial ryegrass (Lolium perenne L.) is considered to be well adapted to intensive management and to produce high-quality forage, but is also known to require adequate water availability, whereas tall fescue (Festuca arundinacea Schreb.) is regarded as a droughttolerant grass with good yield potential, but with rapidly declining forage quality (Dietl et al., 1998). Carbon isotope discrimination and water use efficiency are highly (negatively) correlated because both are affected by the relationship between photosynthesis and stomatal conductance. Hence, analysis of carbon isotope composition is a suitable tool to investigate intrinsic water use efficiency (Wi) in grassland. However, to our knowledge, no information at a species level is known for these grasses. For this reason, specific measurements were undertaken in the course of a field study in order to gain knowledge of W i of tall fescue and perennial ryegrass. Specifically, the study tested whether the known difference in drought tolerance between the two species is related to a difference in W i and whether differences in Wi are maintained in different environments, here at two altitudes and with different management intensities. Materials and methods Leaf material of tall fescue and perennial ryegrass was obtained from a field trial established three years before at two experimental sites (Table 1) and aiming at optimizing a seed mixture, containing both perennial ryegrass and tall fescue, for permanent mountain meadows at drought-endangered, non-irrigated locations. Three factors were studied in this experiment: seed mixture (Fa40 and Fa60, containing the same species, but 40% and 60% seed weight of tall fescue respectively), management intensity (low: 2 cuts year-1 coupled to a fertilization Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 163 level equal to 2 livestock units ha-1; high: 3 cuts year-1 coupled to a fertilization level equal to 2.5 livestock units ha-1) and the experimental site (low altitude: San Genesio/Jenesien 835 m a.s.l.; high altitude: Falzes/Pfalzen 1205 m a.s.l.). The experimental design is a Latin rectangle with 3 replications and a plot size of 4 × 4 m. Table 1. Description of the experimental sites San Genesio/Jenesien and Falzes/Pfalzen. Experimental site Location Geographic coordinates Altitude (m a.s.l.) Slope (%) Aspect Low altitude San Genesio/Jenesien 46° 31' 25" N 11° 20' 22" E 835 26 S High altitude Falzes/Pfalzen 46° 49' 18" N 11° 53' 42" E 1205 31 S Within each plot, three genotypes were sampled: a cultivar of perennial ryegrass (Ivana), a rough-leafed cultivar of tall fescue (Kora) and a soft-leafed cultivar of tall fescue (Barolex). Two different cultivars of tall fescue with contrasting leaf roughness (and equally abundant in the seed mixtures) were included in the study because differences in competitive ability, yield potential and some parameters of forage quality have been previously shown to be related to the leaf roughness of tall fescue (Peratoner et al., 2010). On 28 and 29 August 2013 samples were taken at the high altitude and at the low altitude site, respectively. Each sample consisted of the youngest fully expanded leaf of 10 randomly selected plants per species and plot. The samples were oven-dried for 2 days at 60 °C and milled using a mortar mill (model RM 200, Retch, Haan, D). The samples were subsequently dried for 24 h at 40 °C, then amounts of 0.7 ± 0.05 mg were weighed into zinc cups and burned in an elemental analyzer (NA 1110, Carlo Erba, Milan, Italy) coupled to an isotope mass spectrometer (Delta Plus, Finnigan MAT, Bremen, D). As a control, a standard was measured with a known C/N ratio after every tenth sample. 13C:12C ratios of samples were used to compute Wi according to Köhler et al. (2010). Statistical data analysis was performed by means of a mixed model taking into account the genotype, the seed mixture, the management intensity and the design factors (lines and columns) as fixed factors. The genotype was considered to be a repeated factor with the plot as a subject. Post hoc comparisons were performed by Sidak test. A probability of P<0.05 was regarded as significant. Results and discussion Wi was found to be significantly affected by the genotype, the experimental site (both P<0.001) and also by their interaction (P<0.01). Tall fescue was found to have higher water use efficiency than perennial ryegrass at both sites, with the cultivar Barolex showing intermediate values between Kora and Ivana at the low altitude site (Table 2). This confirms field observations of better drought tolerance of tall fescue in comparison with perennial ryegrass, and contributes to explain these different attributes 13C was consistently lower at high altitude (i.e. carbon isotope discrimination was consistently higher at high altitude). This is in contrast with other findings which have shown a decrease of carbon isotope discrimination with altitude (e.g. Körner et al., 1988; Männel et al., 2007). However, these studies analysed different species at each altitude. Here, the same plant species were sampled, and thus, the observed change in Wi with altitude indicate the species response to a decrease in vapour pressure deficit. The altitudinal effect on Wi was stronger in perennial ryegrass than in both tall fescue cultivars. Again, this indicates a better adaption of tall fescue to conditions of high water demand. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 164 Table 2. Intrinsic water use efficiency (μmol mol-1) depending on genotype and management intensity at two contrasting altitudes. Means without upper case letters in common within each genotype and means without lower case letters in common within each experimental site significantly differ from each other. Experimental site Genotype (species, cultivar) perennial ryegrass cv. Ivana tall fescue cv. Barolex Management intensity tall fescue cv. Kora low high Low altitude 85.6 Ab 88.6 Aab 92.4 Aa 90.3 Aa 87.4 Aa High altitude 61.7 Bb 77.7 Ba 80.4 Ba 71.4 Ba 75.2 Ba A further significant interaction was observed between experimental site and management intensity (P<0.05). Nutrient supply has indeed been demonstrated to affect Wi in grassland (Köhler et al., 2012). However, while altitude effects on Wi were consistent with those already observed at the interaction genotype × experimental site, no significant difference depending on management intensity could be statistically detected by the post hoc test (Table 2). Another significant interaction was found between seed mixture and management intensity (P<0.05). The only significant difference was found at low management intensity, with the Wi of plots sown with Fa40 being higher than in the plots sown with Fa60 (83.4 μmol mol-1 and 78.4 μmol mol-1 respectively). Nutrient supply and differences in species composition have indeed been demonstrated to affect Wi in grassland (Köhler et al., 2012). However, the explanation of these effects deserves further experimental work. Conclusion The present findings show that the higher drought tolerance of tall fescue relative to perennial ryegrass is related to a higher intrinsic water use efficiency. The species differences in W i are consistent in environments differing in vapour pressure deficit. References Dietl W., Lehmann J. and Jorquera M. (1998) Wiesengräser. Landwirtschaftliche Lehrmittelzentrale, Zollikofen, CH, 192 pp. Köhler I.H., MacDonald A. and Schnyder H. (2012) Nutrient supply enhanced the increase in intrinsic water-use efficiency of a temperate seminatural grassland in the last century. Global Change Biology 18, 3367–3376. Köhler I.H., Poulton P.R, Auerswald K. and Schnyder H. (2010) Intrinsic water-use efficiency of temperate seminatural grassland has increased since 1857: an analysis of carbon isotope discrimination of herbage from the Park Grass Experiment. Global Change Biology 16, 1531–1541. Körner C., Farquhar G.D. and Roksandic Z. (1988) A global survey of Carbon isotope discrimination in plants from high altitude. Oecologia 74, 623-632. Männel T.T., Auerswald K. and Schnyder H. (2007) Altitudinal gradients of grassland carbon and nitrogen isotope composition are recorded in the hair of grazers. Global Ecology and Biogeography 16, 583–592. Peratoner G., Resch R., Gottardi S., Figl U., Bodner A., Werth E. and Kasal A. (2010) Competitiveness, yield and forage quality of soft and rough-leafed varieties of tall fescue (Festuca arundinacea Schreb.) in a mountain environment. In: Jambor V., Jamborová S., Vosynková B., Procházka P., Vosynková D. and Kumprechtová D. (eds) Conference proceedings of the 14th international symposium on forage conservation, Mendel University, Brno, CZ, pp. 124-126. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 165 Important differences in yield responses to simulated drought among four species and across three sites Hofer D.1, Suter M.1, Hoekstra N.J.2, Haughey E.2, Eickhoff B.1, Finn J.A.2, Buchmann N.3 and Lüscher A.1 1 Agroscope, Institute for Sustainability Sciences ISS, Forage Production and Grassland Systems, Reckenholzstrasse 191, CH-8046 Zürich. 2 Teagasc, Environment Research Centre, Johnstown Castle, Wexford, Ireland 3 Institute of Agricultural Sciences, ETH Zürich, Universitätstrasse 2, CH-8092 Zürich Corresponding author: daniel.hofer@agroscope.admin.ch Abstract Summer droughts are predicted to increase in frequency due to climate change. We evaluated drought resistance in intensively managed grassland by using four model species (Lolium perenne L., Cichorium intybus L., Trifolium repens L., Trifolium pratense L.). The species represented different functional types, these being defined as a combination of traits related to symbiotic dinitrogen (N2) fixation and rooting depth. A summer drought period of ten weeks with complete exclusion of precipitation was simulated in a common field experiment at three sites (Tänikon CH, Reckenholz CH, Wexford IE). Aboveground biomass production was impaired in the drought treatment at all three sites, the mean reduction (compared to a control) was 30% at Tänikon, 48% at Reckenholz, and 85% at Wexford. Different plant functional types varied in their drought resistance: N2 fixing species showed only 8% and 28% biomass reduction at Tänikon and Reckenholz, respectively, compared to 51% and 68% for the nonfixing species. At Wexford, however, only the deep-rooted species C. intybus was able to counteract drought to some degree (57% biomass reduction compared to 94% reduction for the other three species). This suggests that the three sites exerted a very different degree of drought stress on plants but that cropping N2 fixing and deep-rooted species can be an important management option under future climate conditions. Keywords: drought, biomass production, plant functional types, intensive grassland Introduction In Central and Southern Europe, summer drought spells are expected to occur more frequently due to climate change (Lehner et al., 2006) and to impair forage production in grassland (Gilgen and Buchmann, 2009). Intensively managed grassland with high yielding forage species can be susceptible to drought and farmers might experience considerable loss of income (Finger et al., 2013). Therefore, the current forage production of intensively managed grassland should be adapted to future climate conditions, which could be achieved by cropping forage species with functional traits that improve drought resistance. Here, we present results from a multi-site drought stress experiment that aimed at studying four plant species from different functional types, these being defined as the factorial combination of traits associated with the method of nitrogen acquisition (N2 fixing vs. non-fixing species) and the spatial pattern of root growth (deep- vs. shallow-rooted species). We hypothesized that N2 fixing and deep-rooted species would have an improved drought resistance compared to non-fixing and shallow-rooted species. We also tested whether drought resistance of the four plant functional types would differ among sites. Materials and methods A common field experiment was set up at Tänikon and Reckenholz (northern Switzerland) and Wexford (south-east Ireland). At each site, monoculture plots (3 m × 5 m) of four plant species were established. The species represented the following functional types: a non-fixing, Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 166 shallow-rooted species (Lolium perenne L.), a non-fixing, deep-rooted species (Cichorium intybus L.), an N2 fixing, shallow-rooted species (Trifolium repens L.) and an N2 fixing, deeprooted species (Trifolium pratense L.). A drought treatment was set up by simulating a summer drought period of ten weeks using rainout shelters (3 m × 5.5 m) that led to complete rain exclusion. The control treatment consisted of the actual climatic conditions and each combination of species and treatment was replicated three times per site. Aboveground biomass of the central strip (1.5 m × 5 m) of each plot was harvested six times per year (five at Wexford) at a cutting height of 7 cm (5 cm at Wexford) using a plot harvester. There were two re-growth periods during the drought treatment, and results from the second re-growth period are presented here. Differences in dry matter yield between control and drought treatment as well as among species were tested by analysis of variance (ANOVA). Results and discussion The simulated drought reduced aboveground biomass production significantly at all sites (P < 0.001 each, Table 1). Table 1: Mean (± s.e.) aboveground biomass production (kg DM ha-1 harvest-1) of each of four forage plant species in second harvest period during simulated drought and the % reduction under drought vs. control conditions. Biomass production among species and between drought vs. control conditions were tested by ANOVA (lntransformed). Three replicates per treatment. Tänikon CH Reckenholz CH Monocultures Control Wexford IE Monocultures Drought L. perenne 1355 (±98) C. intybus 1477 (±180) 935 (±78) T. repens 1763 (±34) 1523 (±74) 479 (±187) T. pratense 2841 (±103) 2791 (±150) Average %reduction Control 65 682 (±78) Monocultures Drought %reduction Control Drought %reduction 166 (±47) 75 668 (±146) 96 (±52) 86 37 2062 (±143) 787 (±40) 61 808 (±30) 348 (±183) 57 14 1197 (±129) 789 (±80) 34 1509 (±208) 33 (±19) 98 2 3232 (±193) 2551 (±259) 21 1013 (±115) 11 (±11) 99 30 48 85 ANOVA Variable Df F-value P-value F-value P-value F-value P-value Species (Sp) 3 29.0 < 0.001 41.7 < 0.001 1.3 0.340 Drought (Dr) 1 18.7 < 0.001 31.1 < 0.001 68.1 < 0.001 Sp x Dr 3 6.3 0.005 8.9 0.006 4.3 0.043 The drought effect was moderate at both Swiss sites but severe at the Irish site. At Tänikon and Reckenholz, drought reduced the average biomass by 30% and 48%, respectively, whereas at Wexford the average reduction was 85%. The four different plant functional types varied in their drought resistance. At the two Swiss sites, aboveground biomass production of the N2 fixing species T. repens and T. pratense was clearly less reduced under drought compared to that of the non-fixing species L. perenne and C. intybus (pooled contrasts: P = 0.005 at Tänikon, P = 0.001 at Reckenholz). Given the type of nitrogen acquisition (non-fixing or N2 fixing), the deep-rooted species C. intybus and T. pratense also tended to be less impaired by drought than the shallow-rooted species L. perenne and T. repens (P = 0.061 at Reckenholz, P = 0.104 at Tänikon). However, at Wexford, only C. intybus, the species with the deepest roots, was able to counteract drought to some degree (57% biomass reduction) and it performed significantly better than the other three species (P = 0.020; contrast of C. intybus vs. the other three), which collapsed almost completely due to drought (94% biomass reduction on average). Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 167 It might be hypothesized that these differences in pattern of species’ responses are related to the big differences in drought severity among sites. Under moderate drought, as at the Swiss sites, the N2-fixing species have some growth advantage because they have access to N despite restricted uptake from dry soil due to symbiotic N2 fixation. In contrast, the non-fixing species might suffer from reduced availability of mineral N under drought conditions. Such explanation would match results from previous studies investigating the drought response of plant functional types in temperate grassland (Lüscher et al., 2005; Gilgen and Buchmann, 2009). Under severe drought stress, as at the Irish site, N2 fixation is strongly impaired (Serraj et al., 1999) and deep-rooted species might better resist the stress (Ho et al., 2005, Gilgen et al., 2010) if they can take up water from deeper soil layers. However, more detailed analyses are needed that will i) allow a more refined assessment of induced drought stress across different sites with varying soils and climatic conditions (Vicca et al., 2012), ii) investigate the effect of drought severity on species’ responses, and recovery rates, iii) reveal the effect of drought stress on symbiotic N2 fixation, and iv) disentangle pure water limitation from a co-limitation of water and nutrients. Doing so should help identify traits for improved drought resistance and thereby selection of species for forage production under future climate conditions. Conclusion Our study showed varying species’ responses related to differences in drought severity. At least under moderate drought, N2 fixing as well as deep-rooted species showed mitigation of drought stress in intensively managed grassland. Cultivating mixtures including species from different plant functional types might therefore be a promising strategy to ensure future forage production under varying degree of drought severity. Acknowledgements The research leading to these results has been conducted as part of the AnimalChange project which received funding from the European Community's Seventh Framework Programme (FP7/ 2007-2013) under the grant agreement n° 266018. References Finger R., Gilgen A. K., Prechsl U. E. and Buchmann N. (2013) An economic assessment of drought effects on three grassland systems in Switzerland. Regional Environmental Change 13, 365-374. Gilgen A. K. and Buchmann N. (2009) Response of temperate grasslands at different altitudes to simulated summer drought differed but scaled with annual precipitation. Biogeosciences 6, 2525-2539. Gilgen A.K., Signarbieux C., Feller U. and Buchmann N. (2010) Competitive advantage of Rumex obtusifolius L. might increase in intensively managed temperate grasslands under drier climate. Agriculture Ecosystems & Environment 13, 15-23. Ho M. D., Rosas J. C., Brown K. M. and Lynch J. P. (2005) Root architectural tradeoffs for water and phosphorus acquisition. Functional Plant Biology 32, 737-748. Lehner B., Doll P., Alcamo J., Henrichs T. and Kaspar F. (2006) Estimating the impact of global change on flood and drought risks in Europe: A continental, integrated analysis. Climatic Change 75, 273-299. Lüscher A., Fuhrer J. and Newton P.C.D. (2005) Global atmospheric change and its effect on managed grassland systems. In: McGilloway D.A. (ed.) Grassland: a global resource, Wageningen Academic Publishers, Wageningen, The Netherlands, pp. 251-264. Serraj R., Sinclair T. R. and Purcell L. C. (1999) Symbiotic N 2 fixation response to drought. Journal of Experimental Botany 50, 143-155. Vicca S., Gilgen A.K., Serrano M.C., Dreesen F.E., Dukes J.S., Estiarte M., Gray S.B., Guidolott, G., Hoeppner S.S., Leakey A.D.B., Ogaya R., Ort D.R., Ostrogovic M.Z., Rambal S., Sardans J., Schmitt M., Siebers M., van der Linden L., van Straaten O. and Granier A. (2012) Urgent need for a common metric to make precipitation manipulation experiments comparable. New Phytologist 195, 518-522. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 168 Assessment of the nutritive value and methanogenic potential of two cultivars of Lotus corniculatus L. and Lotus uliginosus Schkuhr. Marichal M. de J.1, Piaggio L.2, Crespi R.1, Arias G.1, Furtado S.1, Cuitiño M.J.3 and Rebuffo M.3 1 Facultad de Agronomía, UdelaR, Garzón 780, Montevideo 12900, Uruguay 2 Secretariado Uruguayo de la Lana, Rambla B. Brum 3764, 11800, Montevideo, Uruguay 3 Estación Experimental INIA - La Estanzuela, Ruta 50, Km 11, Colonia, Uruguay. Corresponding author: mariadejesus.marichal@gmail.com Abstract The aim of this study was to compare the composition, in vitro gas production kinetics and methanogenic potential of L. corniculatus ‘INIA Draco’ and ‘San Gabriel’ and L. uliginosus 4n ‘LE205’ and 2n ‘LE306’. Trefoils were seeded in a complete randomized block design with four replicates. Crude protein and neutral and acid detergent fibre were quantified. Gas production kinetics and methane production (from 0 to 8 and 8 to 24 h) were determined separately using an in vitro gas production procedure. Both cultivars of L. uliginosus presented higher (P < 0.025) fibre fractions, lower (P < 0.010) gas production kinetics and produced less methane in the overall period of incubation than cultivars of L. corniculatus. Within species, cultivars presented similar fibre contents and production of methane; however, differences (P < 0.050) in gas production kinetics were observed and L. uliginosus ‘LE306’ presented lower CP than did ‘LE205’. Results of this study suggest that cultivars of L. corniculatus would provide higher degradable organic matter that would digest faster in the rumen than L. uliginosus. Differences between these two species in their methanogenic potential could be associated with differences in the amount and rate of fermentation of rumen degradable organic matter. Keywords: birdsfoot trefoil, lotus major, chemical composition, in vitro gas production Introduction In Uruguay, sheep and beef nutrition is based mainly on grazing of native pastures. These pastures have a highly heterogeneous production, seasonal distribution, and quality of forage throughout the year. This implies periods of sub-nutrition, which negatively affects the performance of animals. Introduction of perennial legumes into native pastures, particularly Lotus corniculatus and Lotus uliginosus, has been widespread and has been used to improve the availability and forage-quality of production systems under grazing. These legume species have been of special interest because the condensed tannins in their foliage can provide significant benefits for ruminant performance, health and environmental sustainability (Waghorn, 2008). Most published information on these legumes has focussed primarily on biomass production and seasonal distribution, and there is no information in terms of feeding value and methanogenic potential of domestic cultivars. The aim of this study was to assess variation in chemical composition, in vitro gas production kinetics and methanogenic potential of L. corniculatus ‘INIA Draco’ and ‘San Gabriel’ and L. uliginosus 4n ‘LE205’ and 2n ‘LE306’. Materials and methods Trefoils were seeded in 2008 at the National Institute of Agricultural Research ‘INIA La Estanzuela’, Colonia, Uruguay (34° 20' S 57° 41' W) in a complete randomized block design with 4 replicates. Forages were harvested on 27 October 2009. The phenologial stage was calculated using the criteria described by Hoffman et al. (1993). In all samples, total N (AOAC, 2007) and neutral and acid detergent fibre (with heat stable α-amylase in the neutral detergent Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 169 solution, expressed on an ash-free basis, aNDFom and ADFom, respectively) (Van Soest et al., 1991) were quantified. Gas production kinetics and methane production were determined separately using an in vitro gas production procedure (Theodorou et al., 1994). In this study the kinetics of gas production and the volume of gas produced were fitted using the NLIN PROC of SAS according to the model V = a (1-e-kd (t-L)) where 'V' = cumulative gas production at time t, 'a' = potential gas production; 'kd' = fractional rate of gas production and "L" = lag time. No lag times were identified. The concentration of methane in the gas produced from 0 to 8 and 8 to 24 h of incubation was measured by gas chromatography. All results were analysed using the PROC MIXED of SAS considering total gas and methane accumulated from 0 to 8 and 8 to 24 hours of incubation as repeated measures. Results and discussion Cultivars of L. corniculatus and L. uliginosus ‘LE205’ were at a similar (P > 0.80) phenological stage, in the transition from late vegetative to late bud stage, while L. uliginosus ‘LE306’ was at a later (mid-bloom) (P < 0.05) phenological stage (Table 1). Table 1. Phenological characterization, chemical composition (g kg DM -1), in vitro potential of gas production (ml g iOM-1) and fractional gas production rate (h-1), and in vitro total gas (ml g iOM-1) and methane (ml g iOM1 ) production of Lotus corniculatus ‘ INIA Draco’ and ‘San Gabriel’ and Lotus uliginosus ‘LE205’ and ‘LE206’. Species Lotus corniculatus Lotus uliginosus SEM P Cultivars ‘INIA Draco’ ‘San Gabriel’ ‘LE205’ ‘LE306’ PS 1.9b 2.1b 2.1b 3.2a 0.18 0.027 -1 b b a aNDFom, g kg DM 446.7 453.7 508.7 530.3a 15.5 0.009 -1 b b a a ADFom, g kg DM 321.7 336.7 380.0 432.0 16.5 0.012 CP, g kg DM-1 183.3a 163.7ab 184.3a 140.0b 8.2 0.023 In vitro gas production kinetics 'a', ml / g iOM 174.5b 186a 151c 159c 3.7 < 0.001 -1 a a b b kd, h 0.090 0.087 0.073 0.066 0.002 < 0.001 In vitro total gas and methane production Total gas (ml /g iOM) accumulated from: 0 to 24 h 228a 224a 204b 189c 3.3 < 0.001 a a b c 0 to 8 h 159 156 137 121 2.3 < 0.001 8 to 24 h 68 69 68 68 2.0 < 0.001 Methane accumulated (ml / g iOM) from: 0 to 24 h 12.46a 13.29a 9.01b 7.54b 1.01 < 0.001 a a b b 0 to 8 h 7.37 6.88 4.39 3.84 0.58 < 0.001 8 to 24 h 4.35b 6.35a 4.54b 3.51b 0.63 < 0.001 PS: phenological stage; aNDFom: amylase neutral detergent fibre (organic matter basis); ADFom: acid detergent fibre (organic matter basis); CP: crude protein; 'a': gas production potential; kd: fractional rate of gas production; iOM: incubated organic matter. a, b, c: In the rows, superscript letters indicate differences significant at P ≤ 0.05. Lotus corniculatus presented lower (P < 0.025) fibre fractions and higher (P < 0.010) gas production kinetics and methane production during the overall period of incubation (0 to 24 hours) than did L. uliginosus. The higher potential and rate of gas production in L. corniculatus suggested a greater degradability of the organic matter (OM) in the rumen. Both species produced high volumes of gas, and 64 to 70% of total gas was produced within the first 8 hours of incubation suggesting that their degradable OM would be fermented rapidly. Cultivars of L. corniculatus were similar (P > 0.152) in their chemical composition and methane production over the overall period of measurement; however, ‘INIA Draco’ had a lower (P = 0. 037) potential of gas production than ‘San Gabriel’. Cultivars of L. uliginosus presented similar (P > 0.145) fibre contents, gas production kinetics and overall methane production; however, ‘LE306’ presented lower (P = 0.029) CP than ‘LE205’. The lower CP could be explained by Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 170 the difference in the phenological stage of cultivars. Buxton (1996) reported that protein concentration declines with increasing maturity because, in legumes, leaves presented higher protein content than stems and the proportion of leaves represents a smaller proportion of the available herbage in forages of more advanced maturity. During the overall period of measurement, cultivars of L. corniculatus produced 500 ml kg iOM-1 (P < 0.035) more methane than did cultivars of L. uliginosus. The production of methane is associated with the nature and rate of fermentation of carbohydrates, particularly with the carbohydrates of the cell wall, which are considered to be the most methanogenic. These carbohydrates are quantified as aNDFom, a fraction that has been positively associated with methane production (Moss et al., 2000). In this study, the aNDFom fraction in L. uliginosus was larger than in L. corniculatus; however, the former produced lower methane, suggesting that in these species the production of methane would be associated more with the amount and rate of degradation of the rumen degradable OM than to the aNDFom content. This could be the result of a higher content of more astringent condensed tannins in L. uliginosus which could inhibit fibre-degrading bacteria, reducing the proportion of dietary energy lost as methane (Waghorn, 2008). Conclusions Results of this study suggest that both the tested cultivars of L. corniculatus would provide higher degradable organic matter, which would degrade faster in the rumen than cultivars of L. uliginosus. Differences between species in their methanogenic potential could be associated with variations in the amount and rate of fermentation of rumen degradable organic matter. References AOAC (2007) Official Methods of Analysis,18thedn. Association Official Analytical Chemists,Washington, D.C. Buxton D.R. (1996) Quality related characteristics of forages as influenced by plant environment and agronomic factors. Animal Feed Science and Technology 59, 37-49. Hoffman P.C., Sievert S.J., Shaver R.D., Welch D.A. and Combs D.K. (1993) In situ dry matter, protein, and fiber degradation of perennial forages. Journal of Dairy Science 76, 2632-2643. Moss A.R., Jouany J.P. and Newbold J. (2000) Methane production by ruminants: its contribution to global warming. Annales de Zootechnie 49, 231-253. Theodorou M.K., Williams B.A., Dhanoa M.S., McAllan A.B. and France J. (1994) A simple gas production method using a pressure transducer to determine the fermentation kinetics of ruminant feeds. Animal Feed Science and Technology 48, 185-197. Van Soest P.J., Robertson J.B. and Lewis B.A. (1991) Methods for dietary fiber, neutral detergent fiber, and nonstarch polysaccharides in relation to animal nutrition. Journal of Dairy Science 74, 3583-3597. Waghorn G. (2008) Beneficial and detrimental effects of dietary condensed tannins for sustainable sheep and goat production - Progress and challenges. Animal Feed Science and Technology 147, 116-139. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 171 Improvement of the digestibility of tall fescue (Festuca arundinacea Schreb.) inspired by perennial ryegrass (Lolium perenne L.) Baert J. and Van Waes C. Institute for Agricultural and Fisheries Research (ILVO), Plant Sciences Unit, Caritasstraat 21, 9090 Melle, Belgium Corresponding author: joost.baert@ilvo.vlaanderen.be Abstract Due to climate change there is an increasing interest in NW Europe for more drought-tolerant fodder grass species like tall fescue (Festuca arundinacea Schreb.). Tall fescue has a high yield potential but a low digestibility. In a pot experiment we determined the digestibility and the sugar, fibre and protein content of 300 non-vernalized single plants (12 cultivars × 25 plants per cultivar) from perennial ryegrass and tall fescue. Tall fescue had a lower digestibility and sugar content and a higher fibre and protein content than perennial ryegrass. The variation in sugar content was as high for tall fescue as for perennial ryegrass. The same positive correlation between digestibility and sugar content was found in tall fescue as in perennial ryegrass. Because of the lower sugar content in tall fescue there seems to be enough margin to improve its digestibility by increasing the sugar content without adverse effects on fibre and protein value. Keywords: Festuca arundinacea, Lolium perenne, digestibility Introduction Perennial ryegrass (Lolium perenne L.) is the most important fodder grass species on dairy farms in NW Europe. It is a high yielding species with a very good digestibility. However, during periods of drought, perennial ryegrass stops growing. Due to climate change more extreme weather conditions, including drought periods, are expected to occur (Parry et al., 2007). Therefore there is increasing interest in grass species that are more tolerant to drought stress. Tall fescue (Festuca arundinacea Schreb.) is not only more drought tolerant than perennial ryegrass but it also has a higher yield potential. Unfortunately, tall fescue lacks palatability and digestibility. For use on intensive dairy farms the feed quality of tall fescue needs to be improved. Leaves have a better digestibility than stems and therefore breeders should focus on the leaf component to improve digestibility (Beecher et al., 2013). Breeding leafy varieties without reheading is a first step. Also, in the vegetative stage there are differences in digestibility among varieties. In this study we compared the digestibility and its components of non-vernalized single plants of perennial ryegrass (Lp) and tall fescue (Fa) Based on this comparison we made some suggestions for the improvement of the digestibility of tall fescue. Materials and methods In February 2008, three hundred seeds of both Lp and Fa were sown in trays in the greenhouse. For each species, 25 seeds of each of 12 varieties or populations were used. In April 2008 the single plants were transplanted into 12 L pots, put on an open hardened field, and irrigated. An equalization cut was applied in May. Pots were further harvested in June and August. After each cut the pots were fertilized with 4 g of a 16-8-22 NPK fertilizer. At each harvest, fresh (FY) and dry weights of the single plants were measured and dry matter content (DMc) calculated. Dry matter digestibility (DMD) and contents of neutral detergent fibre (NDF), acid detergent fibre (ADF), water soluble carbohydrates (WSC) and crude protein (CP) were determined by NIRS (near infrared reflectance spectroscopy) on dried samples. The content of Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 172 protein digestible in the intestine (DP) was estimated by a multiple regression equation including DMD, CP, DM and NDF. The perennial ryegrass plants showed no stems at all in the sowing year. About 20% of the tall fescue plants already showed one or a few real stems at the first harvest. At the second harvest more than half of the tall fescue plants had a few stems. Therefore we present results only from the first harvest. Results and discussion The vegetative tall fescue had, on average, significantly (t-test, P<0.001) lower digestibility and sugar content but a higher fibre and crude protein content than perennial ryegrass (Table 1). There was no difference in digestible protein content. The generative plants of tall fescue had an even lower digestibility and sugar content and higher fibre content. The variation of the feeding quality components among the single plants of both perennial ryegrass and tall fescue was highest for the sugar content and lowest for the digestibility. Table 1. Means and coefficients of variation (cv) of fresh yield, dry matter content, digestibility, fibre, sugar and protein content for vegetative plants of perennial ryegrass (Lp) and both vegetative and generative plants of tall fescue (Fa). FY (g) DMc (%/FY) DMD (%/DM) NDF (%/DM) ADF (%/DM) WSC (%/DM) CP (%/DM) DP (%/DM) vegetative Lp (n=295) mean cv 116 18.2 26.1 7.7 79.4 3.9 37.8 7.9 18.9 9.8 33.3 13.0 9.8 13.2 8.4 4.4 vegetative Fa (n=230) mean cv 87 23.5 24.0 8.1 76.4 3.6 41.3 6.9 18.7 9.2 27.9 16.2 11.8 15.4 8.4 5.9 generative Fa (n=66) mean cv 121 23.1 24.6 6.6 71.0 5.5 46.5 7.7 22.6 11.7 25.4 16.9 10.5 14.9 7.8 7.7 The DMD was positively correlated with the WSC content and negatively with NDF and ADF (Table 2). Table 2. Correlation coefficients between DMD and FY, DMc, NDF, ADF, WSC, CP and DP for vegetative plants of perennial ryegrass (Lp) and tall fescue (Fa) Lp Fa FY 0.09 -0.17 ** DMc 0.51 *** 0.41 *** NDF -0.92 *** -0.84 *** ADF -0.81*** -0.79 *** WSC 0.88*** 0.74 *** CP -0.48*** -0.22 ** DP 0.67*** 0.49 *** In Figure 1 we ranked the plants from low to high digestibility and calculated trend lines for the other quality components. The course of the digestibility is parallel to the course of the WSC. In perennial ryegrass the genetic variation for WSC is higher than for DMD (Humphreys, 1989). Thanks to this large genetic variation, high sugar ryegrasses have been bred and improvement in the energy value has been achieved by increasing the sugar content (Humphreys et al., 2010). However, a very high WSC content in the grass may cause rumen acidosis. In tall fescue there is also a large variation in WSC. Because of the lower level of the WSC content in tall fescue there is a greater margin for improvement of the WSC without risk of acidosis. Also, the higher fibre content of tall fescue reduces this risk. WSC is often negatively correlated with CP (Ghesquiere et al., 2007). As the CP content in tall fescue is higher than in perennial ryegrass the increase in WSC will have a less negative effect on the crude protein content and a positive effect on the protein use and on the reduction of the excreted nitrogen. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 173 Figure 1. Trends of NDF, ADF, WSC, CP and DP for single plants of vegetative perennial ryegrass (left) and tall fescue (right) ranked from low to high digestibility. Conclusion The DMD in tall fescue is highly positively correlated with the WSC content, as is also the case in perennial ryegrass. There is also a lot of variation in WSC in tall fescue. As the average WSC content in tall fescue is much lower than in perennial ryegrass, there is scope for improving digestibility by increasing the WSC and also maintaining enough fibre and protein. References Beecher M., Hennessy D., Boland T.M., McEvoy M., O’Donovan M. and Lewis E. (2013) The variation in morphology of perennial ryegrass cultivars throughout the grazing season and effects on organic matter digestibility. Published online in Grass and Forage Science doi: 10.1111/gfs.12081. Ghesquiere A., Muylle H. and Baert J. (2008) Analysis of the water soluble carbohydrate content in an unselected breeding pool of perennial ryegrass. Proceedings of the 27th meeting of fodder crops and amenity grasses section of Eucarpia, Copenhagen, 2007. Humphreys M. (1989) Water soluble carbohydrates in perennial ryegrass breeding. III. Relationships with herbage production, digestibility and crude protein content. Grass and Forage Science 44, 423-430. Humphreys M., Feuerstein U., Vandewalle M., and Baert J. (2010) Ryegrasses. In: Boller B., Posselt U.K., and Veronesi F. (eds) Handbook of plant breeding 5, Fodder crops and amenity grasses, Springer, New York, pp. 211-260. Parry M., Canziani O., Palutikof J., van der Linden P. and Hanson C. (2007) Contribution of Working Group II to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge UK and New York, NY, USA. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 174 Dry matter yield and digestibility of five cool season forage grass species under contrasting N fertilizations Cougnon M.1, Baert J.2 and Reheul D.1 1 Department of Plant Production, University of Ghent, Proefhoevestraat 22, 9090 Melle, Belgium 2 ILVO Plant, Caritasstraat 21, 9090 Melle, Belgium Corresponding author: mathias.cougnon@ugent.be Abstract The yield potential of different varieties of five grass species from the Festuca and Lolium genera were compared under two N-fertilization regimes (190 kg N ha-1 or 300 kg N ha-1 yr-1) in two successive years under a cutting management. No N fertilization x species interaction was found. Hybrid ryegrass (Lolium x boucheanum Kunth.) was the highest-yielding species in the first year, and tall fescue (Festuca arundinacea Schreb.) in the second year. Meadow fescue (Festuca pratensis L.) was always the lowest-yielding species. Keywords: ryegrass, fescue, Festulolium, N fertilization, digestibility Introduction Grassland farming in North-West Europe is changing. First, N-fertilization is decreasing, mainly due to legal restrictions (EU directive 91/676/EEC). Second, the importance of grazing in dairy farming is decreasing for different reasons (Van den Pol-van Dasselaar et al., 2008), and consequently the importance of cut grass is increasing. Third, more dry summer spells are expected due to climate change. As there are differences in N-use efficiency (Gilliland et al., 2010) and in drought resistance of different forage grass species (Willman et al., 1998) it is pertinent to revaluate the performance of perennial ryegrass (Lolium perenne L.; Lp) - currently the most widely used species - in comparison with more drought-tolerant species that perform well under a cutting management. The aim of this trial was to compare the performance of five forage-grass species under two N levels in a cutting management. The hypothesis was that there is an N-fertilization × species interaction for the different species. Materials and methods A trial was established in September 2011 on a sandy loam soil in Merelbeke, Belgium, comprising two varieties of Lp (‘Indiana’ and ‘Toronto’), tall fescue (Festuca arundinacea Schreb.; Fa; ‘Barolex’ and ‘Callina’), hybrid ryegrass (Lolium x boucheanum Kunth.; Lh; ‘Melprius’ and a candivar), meadow fescue (Festuca pratensis L., Fp;, ‘Pardus’ and a candivar) and Festulolium (Fl; ‘Achilles’ and ‘Lueur’). The trial was a split-plot design with three replicates; individual plot size was 7.8 m². N fertilization was the main plot factor with two levels: high (300 kg N ha-1 yr-1) or low N (190 kg N ha-1 yr-1) and the varieties of the different species formed the subplot factor. Plots were sown at a density of 10000 germinable seeds per plot. P and K fertilization for high and low N fertilizations were respectively: 16 and 11 kg P ha-1 and 267 and 189 kg K ha-1. Five cuts were harvested, in both 2012 and 2013. The fresh yield was determined by harvesting the entire plot using a plot harvester (Haldrup, Logstor, Denmark). A random subsample of the harvested material was taken from each plot and dried for at least 24 h at 70 °C. Digestibility of all varieties was analysed with NIRS (calibrated with Tilley and Terry) on a single pooled sample per variety and cut in 2012. ANOVA of the dry matter yields (DMY) for both years were performed using the aov() function in R. The hierarchy of the split-plot design and the nesting of the varieties within species were taken into account in the model. Multiple comparisons of species averages within the N levels was performed using the TukeyHSD() function. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 175 Results and discussion There was, obviously, a positive effect of N-fertilization on the DMY (P < 0.01) in both years and there was also a significant species effect on DMY in both years (P < 0.01). There was no significant species × N fertilization interaction, either in 2012 (P = 0.43) or in 2013 (P = 0.17) and therefore the DMY ranking of the species did not change when N fertilization was increased from 190 kg N ha-1 yr-1 to 300 kg N ha-1 yr-1 (Figure 1). Figure 1. Interaction plots for dry matter yields (kg ha -1 yr-1) of Lolium x boucheanum (Lh), Festulolium (Fl), Lolium perenne (Lp), Festuca arundinacea (Fa) and Festuca pratensis (Fp) averaged over two varieties per species at two N fertilization levels in 2012 (a) and 2013 (b). a a a b Fa Fl Fp Lh Lp a c b a ab b 15000 15000 a 20000 20000 In 2012, the first year after establishment, Lh and Fl were overyielding Fp and Fa at both N fertilizations (Figure 2a). b b b bc bc c c 10000 Dry matter yield (kg/ha) b 0 5000 10000 0 5000 Dry matter yield (kg/ha) b 190 300 N-fertilization (kg/ha) (a) 190 300 N-fertilization (kg/ha) (b) Figure 2. Dry matter yield (kg ha-1 yr-1) of Lolium x boucheanum (Lh), Festulolium (Fl), Lolium perenne (Lp), Festuca arundinacea (Fa) and Festuca pratensis (Fp) averaged over two varieties per species at two N fertilization levels in 2012 (a) and in 2013 (b). Error bars: ± standard deviation. Bars with a different letter are significantly different (P < 0.05; Tukey test). Averaged over both N levels, the DMY of Lh, the highest yielding species, was 1.49 and 1.34 times the DMY of Fp and Fa respectively. In 2013, the second full production year, Fa overyielded all the other species at both N fertilizations (Figure 2b). Fp remained the lowest Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 176 yielding species. Although the yield of Lh and Fl remained superior to that of Lp in the second year, their sward densities started to decline. The yield pattern of Fa was in accordance with earlier observations: due to its slow early vigour it has a low yield in the first production year, but once established it is very high yielding and persistent (Cougnon et al., 2013). The yield deficit of Fa in the first year was compensated by its high yield in the second year: averaged over the two years and over the two N levels, DM yields of Fa, Lh and F1 were not significantly different (28796, 29277 and 28030 kg DM ha-1 respectively). The tested Lh and Fl varieties seemed appropriate for leys lasting for only 2-3 years. Fa had the right properties to become an important forage grass in long-lasting cut grassland: once established it had the highest yield potential. The absence of drought periods in the experimental period did not allow any confirmation of the supposed superior drought resistance of Festuca genes. The digestibility of Fa was consistently lower than that of Lp and Lh, while the digestibility of Lh was generally higher than that of Fl (Table 1). Table 1. Organic matter digestibility (%) in the first full production year. Method: NIRS (cal. Tilley and Terry); data are weighed averages of 5 cuts. (Species codes: see Figure 1.) Fa Fl Lh Fp Lp High N 72.1 75.1 79.3 80.1 79.8 Low N 73.0 76.5 81.0 79.4 80.7 Conclusions Our hypothesis was not confirmed: there was no N fertilization × species interaction for the tested species. Meadow fescue always had the lowest yield performance, Festulolium and hybrid ryegrass varieties were the winners in year 1, and tall fescue was the winner in year 2. References Cougnon M., Van Waes C., Baert J. and Reheul D. (2013) Performance and quality of tall fescue (Festuca arundinacea Schreb.) and perennial ryegrass (Lolium perenne L.) and mixtures of both species grown with or without white clover (Trifolium repens L.) under cutting management. Grass and Forage Science DOI: 10.1111/gfs.12102. Gilliland T.J., Farrell A.D., McGilloway D. and Grogan D. (2010) The effect of nitrogen on yield and composition of grass-clover swards at three sites in Ireland: A comparison of six commonly grown species. Grassland Science in Europe 15, 946-948. Wilman D., Gao Y. and Leitch M. (1998) Some differences between eight grasses within the Lolium-Festuca complex when grown in conditions of severe water shortage. Grass and Forage Science 53, 57-65. Van den Pol-van Dasselaar A., Vellinga T.V., Johansen A. and Kennedy E. (2008) To graze or not to graze, that's the question. Grassland Science in Europe 13, 706-716. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 177 Grazing season length on dairy, beef and sheep farms in Europe Phelan P.1, Morgan E.2, Rose H.2 and O’Kiely P.1 1 AGRIC, Teagasc, Grange, Dunsany, Co. Meath, Ireland 2 School of Biological Sciences, University of Bristol, UK Corresponding author: paul.jp.phelan@gmail.com Abstract Grazing season length and livestock stocking density on grassland are two important parameters for modelling grassland processes such as nutrient losses and parasite transmission. However, there is a lack of data available on these two important parameters for most European grazing systems. This study calculated weighted means of grazing season lengths and grazing stocking rates for dairy, beef and sheep farms in 986 regions in 32 European countries using results from the 2010 Eurostat Survey on Agriculture Production Methods (SAPM). Means for each country are presented here. Keywords: grazing, pasture, livestock density Introduction Grazing season length has particular implications for production costs, nutrient losses, greenhouse gas emissions and parasite transmission. For example, extending the grazing season may reduce production costs, increase nutrient losses and increase Ostertagia infection levels (Charlier et al., 2005; Webb et al., 2005; Dillon et al., 2008). However, data on grazing season length in various regions and farm systems are generally not available. The EU GLOWORM project (www.gloworm.eu) is modelling the effects of future climate change on grazing management and parasite transmission. This requires data on current grazing season lengths on farms across Europe. The objective of this study was to establish a database of grazing-season length and grazing-livestock stocking rates on grassland across Europe. Materials and methods Data on grazing season length were obtained from the 2010 Survey on Agricultural Production Methods (SAPM). The survey was conducted on a sample population (between 3% and 30%) of farms in Belgium, Cyprus, Croatia, Denmark, Finland, Germany, Greece, Hungary, Ireland, Latvia, Norway, Poland, Slovenia, Spain, Sweden, Switzerland and United Kingdom. It was conducted as part of farm census collection in Austria, Bulgaria, Czech Republic, Estonia, France, Iceland, Italy, Lithuania, Luxembourg, Malta, Montenegro, Netherlands, Portugal, Romania and Slovakia. Full SAPM methodological details are available at: http://epp.eurostat.ec.europa.eu/statistics_explained/index.php. Farm enterprises were classified as dairy, beef or sheep farms according to standard EU farm typology (EU Commission Regulation No 1242/2008). The collated SAPM data was obtained in categorical format with the six following categories for 'number of months with livestock grazing': 0 months, 1 to 2 months, 3 to 4 months, 5 to 6 months, 7 to 9 months and 10 months or more. This study calculated weighted means and weighted standard deviations (stdev) from the proportion of LU in each category for both on-farm grazing and commonage grazing for each farm system and region using the following equations: (i) weighted mean = category mean × (proportion of farms in each category ÷ total farms) (ii) weighted stdev = (category mean - weighted mean)2 × (proportion of farms in each category ÷ total farms) Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 178 Category means and farm type were not available for area of commonage. Therefore a total stocking rate was used (total grazing LU of farms that practised common grazing / total area of commonage (ha)) in place of category means in the above equations for common land. Results and discussion Calculated grazing season lengths and grazing stocking densities for each country are shown in Table 1. Creighton et al. (2011) conducted a questionnaire survey of dairy farmers in Ireland and found that farmers grazed their cows for approximately 8.6 months, similar to the 8.8 months for Irish Dairy farms in Table 1. Table 1. Grazing season length and stocking density on grazed grassland for dairy, beef and sheep farms across Europe. Standard deviations are shown in parenthesis. Grazing season length (months) Grazing stocking density (LU ha-1) Dairy Beef Sheep Dairy Beef Sheep AT Austria 4.1 (2.49) 4.0 (3.01) 5.6 (3.03) 1.6 (0.67) 1 (0.56) 0.9 (0.52) BE Belgium 6.8 (1.79) 6.3 (2.48) 6.4 (3.33) 6.2 (10.36) 4.8 (8.02) 3.4 (5.26) BG Bulgaria 8.0 (2.00) 8.7 (2.38) 8.4 (2.14) 0.7 (0.81) 0.4 (0.60) 0.3 (0.39) CH Switzerland 5.8 (2.30) 5.5 (2.18) 4.8 (1.77) 2.3 (2.00)* 2.1 (1.47)* 1.1 (1.18)* CY Cyprus 2.2 (3.79) 6.0 (6.82) CZ Czech Republic 4.3 (3.19) 7.7 (2.79) 8.6 (2.54) 3.1 (1.21) 0.7 (0.07) 0.8 (0.03) DE Germany 4.1 (3.25) 4.9 (4.13) 5.3 (4.71) 4.0 (0.74) 1.5 (0.28) 0.9 (0.10) DK Denmark 4.3 (2.94) 6.0 (3.76) 7.3 (3.94) 7.8 (2.48) 3.3 (1.02) 1.6 (0.30) EE Estonia 4.7 (1.51) 6.7 (1.61) 6.7 (1.55) 2.7 (0.79) 0.8 (0.07) 0.6 (0.07) EL Greece 4.9 (4.16) 9.1 (2.63) 8.9 (2.79) 0.7 (1.51) 1.0 (3.03) 0.6 (0.48) ES Spain 5.5 (4.77) 9.1 (2.93) 9.4 (2.77) 3.0 (1.48) 0.9 (0.48) 0.7 (0.33) FI Finland 3.3 (2.03) 2.9 (3.26) 4.3 (2.91) 5.4 (1.98) 3.4 (1.83) 1.7 (0.23) FR France 8.8 (1.82) 9.3 (1.73) 9.4 (2.22) 2.1 (0.35) 1.2 (0.18) 0.7 (0.17) HR Croatia 3.6 (3.78) 2.0 (3.45) 9.0 (2.37) 1.7 (1.12) 1.0 (0.86) 0.2 (0.72) HU Hungary 2.7 (3.21) 5.1 (3.90) 5.6 (3.71) 4.6 (6.48) 0.7 (0.59) 0.8 (0.32) IE Ireland 8.8 (1.60) 8.1 (1.72) 9.5 (2.30) 2.2 (0.26) 1.2 (0.19) 0.7 (0.28) IS Iceland 6.3 (2.55) 7.2 (2.90) 6.4 (2.78) 0.6 (0.15) 0.3 (0.18) 0.1 (0.03) IT Italy 1.2 (2.66) 3.6 (4.55) 8.4 (3.76) 1.8 (1.12) 0.9 (0.58) 0.9 (0.37) LT Lithuania 5.2 (1.48) 5.0 (1.98) 5.3 (1.95) 1.1 (0.22) 1.0 (0.04) 0.7 (0.05) LU Luxembourg 6.8 (1.25) 6.9 (1.68) 6.6 (1.61) 3.1 (0.06) 2.3 (0.10) 1.6 (0.24) LV Latvia 5.6 (2.07) 7.1 (2.48) 6.4 (2.28) 1.0 (0.07) 0.6 (0.04) 0.7 (0.07) ME Montenegro 3.3 (3.14) 1.8 (2.76) 3.7 (3.27) 1.2 (0.24) 0.9 (0.15) 1.9 (0.73) MT Malta 0.9 (2.62) * NL Netherlands 5.3 (2.71) 4.0 (3.95) 6.2 (4.21) 3.9 (1.03) 3.8 (1.39) 2.4 (0.61) NO Norway 2.6 (1.79) 3.4 (2.14) 3.5 (2.26) 3.6 (0.57) 2.6 (0.25) 2.0 (0.23) PL Poland 3.9 (2.63) 3.6 (3.11) 5.3 (2.36) 3.9 (1.07) 1.8 (0.69) 1.3 (0.37) PT Portugal 5.2 (5.36) 9.7 (3.22) 9.9 (2.75) 3.1 (2.94) 0.6 (0.33) 0.5 (0.32) RO Romania 4.6 (2.63) 4.2 (3.16) 5.8 (2.53) 0.7 (0.79) 0.5 (0.68) 0.8 (0.71) SE Sweden 5.6 (1.59) 6.3 (2.04) 6.7 (3.02) 4.4 (1.20) 1.9 (0.20) 1.0 (0.13) SI Slovenia 3.2 (3.46) 3.5 (3.36) 6.2 (2.79) 4.7 (1.50) 2.1 (0.56) 1.3 (0.42) SK Slovakia 4.7 (2.58) 4.3 (2.43) 6.0 (2.11) 1.3 (0.34) 0.6 (0.05) 0.5 (0.05) UK United 8.1 (2.23) 8.6 (2.51) 10.0 (2.39) 2.8 (0.22) 1.4 (0.23) 0.8 (0.13) Kingdom * Commonage grazing land area not available, on-farm stocking rate only. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 179 Kuusela (2004) stated that the grazing season in Finland was approximately 3-5 months, and the results in Table 1 for Finland are within that range. In Austria, Steinwidder et al. (2010) recorded grazing season lengths on six pilot dairy farms ranging from 5 to 7 months, this is considerably longer than the 4.1 months recorded here (Table 1), possibly because the focus of the former study was on developing low-input farming methods. In northern Spain, Mendarte et al. (2010) stated that grazing season length in the area started in May and ended in October, giving a grazing season length of approximately six months, similar to the results for dairy farms in this study (Table 1). One of the limitations of this study is that the data are derived from a categorical, rather than discrete dataset (see Materials and methods section). Therefore, within-category variation and distribution is not available. While this is not a major issue at country level, due to the large number of results and wide spread across categories, at the much smaller NUTS 3 level it can have a larger effect. Another limitation is that daily grazing duration is not accounted for. For example, a questionnaire survey in 2006 found that the percentage of farms that grazed dairy cows both day and night during the summer in Belgium, Germany, Sweden, Ireland and the UK, were 58, 26, 71, 98 and 87%, respectively (Bennema et al., 2010). Finally, the results are only for one year (2009) and therefore any anomalies (e.g. weather) within that year for each region need to be considered when interpreting the data. Conclusion Weighted mean grazing season length and grazing stocking density were calculated for dairy, beef and sheep farms in 986 NUTS 3 regions in 32 countries across Europe. Results for each country are presented in Table 1. Grazing season length varied widely between regions and farm types. This data set is currently being used to assess the potential future impacts of climate change on grazing management across Europe. References Charlier J., Claerebout E., Mûelenaere E. D. and Vercruysse, J. (2005) Associations between dairy herd management factors and bulk tank milk antibody levels against Ostertagia ostertagi. Veterinary parasitology 133, 91-100. Creighton P., Kennedy E., Shalloo L., Boland T. M. and O’Donovan M. (2011) A survey analysis of grassland dairy farming in Ireland, investigating grassland management, technology adoption and sward renewal. Grass and Forage Science 66, 251-264. Dillon P., Hennessy T., Shalloo L., Thorne F. and Horan B. (2008) Future outlook for the Irish dairy industry: a study of international competitiveness, influence of international trade reform and requirement for change. Inernational Journal of Dairy Technoogy 61, 16-29. Mendarte S., Ibarra A., Garbisu C., Besga G. and Albizu I. (2010) Use of portable NIRS equipment in field conditions to determine the nutritional value of mountain pastures. Grassland Science in Europe 15, 244-246. Steinwidder A., Starz W., Podstatzky L., Kirner L., Pötsch E.M., Pfister R. and Gallnböck, M. (2010) Changing towards a seasonal low-input pastoral dairy production system in mountainous regions of Austria – results from pilot farms during reorganisation. Grassland Science in Europe 15, 1012-1014. Webb J., Anthony S.G., Brown L., Lyons-Visser H., Ross C. and Cottrill B. (2005) The impact of increasing the length of the cattle grazing season on emissions of ammonia and nitrous oxide and on nitrate leaching in England and Wales. Agriculture, Ecosystems and Environment 105, 307–321. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 180 Effects of mild heat stress periods on milk production, milking frequency and rumination time of grazing dairy cows milked by a mobile automatic system Lessire F., Hornick J.L. and Dufrasne I. Animal Production Department, Faculty of Veterinary Medicine, University of Liège, Chemin de la Ferme, 6 B39, 4000 Liège, Belgium Corresponding author: isabelle.dufrasne@ulg.ac.be Abstract Grazing dairy cows milked by an automatic system (AS) experienced mild heat stress (HS) periods twice during the summer. The daily temperature humidity index (THI) during these periods were higher than 72. Milk production, as well as milking frequency, rumination time and milk fat to protein ratio (F/P) during these periods were compared with adjacent periods with mean THI of 61. The daily milking frequency, the total number of visits to the AS, and milk production were significantly higher in HS periods (2.12 vs. 1.97, 2.99 vs. 2.69, and 19.7 vs. 18.5 kg milk per cow, respectively). There were significant interactions between times and periods for milking frequency and number of visits, whereas the daily rumination time was significantly lower (339 vs. 419 min) and the F/P in milk tended to be decreased (1.17 vs. 1.23). These results could be explained by changes in cow behaviour during HS periods. Keywords: dairy cows, grazing, automatic milking system, heat stress, rumination Introduction Cows milked by automatic systems (AS) are most often confined indoors or have access to pasture only during the day in summer. However, grazing allows reduced feeding costs and it improves animal health and welfare. A mobile AS, as described by Dufrasne et al. (2010), allowing grazing of dairy cows in fragmented areas is thus advantageous for animals. At grazing, the cows can move more than when in the barn and they are exposed to the environmental conditions. During heat stress periods, it is known that feed intake can be reduced, especially when temperatures are above 25 or 26 °C (Rhoads et al., 2013). Little information exists on the effects of heat stress on grazing dairy cows milked by an AS. The aim of this study was to determine the effects of heat stress periods on the milk yield, milking frequency, fat to protein ratio in milk (F/P) and rumination time (RT) of grazing dairy cows milked by an AS located on the pasture. Materials and methods This study was carried out at the Experimental Farm of the University of Liège (Belgium). A herd of about 50 dairy cows grazed on 18 ha of permanent pastures and was milked by an AS Lely A3 next®. The grazing period began in April and ended in October. The cows grazed by strip grazing and two allocations per day were provided. The gate of the AS was manually changed twice per day, at 6:00 h and 16:00 h, to guide the cows on to the next allocation. The cows had to pass in the AS in order to benefit from the new allocation. In practice, they were fetched for the morning milking, allowing a daily survey of the animals by the herdsman. They came freely to access the AS when the gate was changed at the afternoon. Furthermore, they had free access to the AS at day and night times. The temperature humidity indexes (THI) were calculated according to Ingraham et al. (1979) and were used to define, post hoc, mild heat stress periods (HS) according to Armstrong (1994). Consequently, a period of 4 and 7 consecutive days were identified in July (J) and in August (A), respectively. These periods were characterized by a THI >72 during the day and 23.1 °C mean temperature. These two heat stress periods were compared to corresponding normal periods chosen close before and after Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 181 the period of heat stress - less than 9 days - for their similarity regarding the distance from the AMS and the availability in water in the grazed paddock. During normal periods, maximum THI was < 68 and mean temperature 16.3 °C. Water was always available near the AS in a tin (1000 L) and available in some grazed paddocks in individual bowls. Only animals present from the beginning till the end of these periods were taken into account. The total number of lactating cows was 45 in J and 47 in A. The lactation number and days in milk during the normal and HS periods were similar (Table 1). The pastures consisted mainly of perennial ryegrass and white clover. The grass heights were measured by an INRA rising plate meter at each entry and exit in the paddocks. Grass yield was measured with a mower, cutting strips of 10 meters long. Grass was sampled at each entry in order to determine chemical composition. Each cow received an amount of concentrate determined with respect to lactation stage. The cows were equipped with a HR-Tag neck collar recording rumination parameters and cow activity (SCR, Israel). The temperature, THI, distance from the paddock to the AMS, days in milk and lactation number were analysed according to a GLM by using THI conditions (normal and HS) and periods (J and A). Data electronically captured by the AS, i.e. milk production, milking frequency, milking visits (result of the sum of milkings, failed and refused milkings), F/P and RT (991 data) were analysed according a mixed model (SAS, 1999) including THI conditions, periods, lactation number, stages of lactation as fixed effects, and allowing an type 1 autoregressive covariance structure for measurements performed on each animal within THI conditions and periods. Results and discussion THI and temperature were significantly different and there was no interaction between THI conditions and periods for environmental parameters (Table 1). Table1. Some environmental and animal characteristics during periods of normal and heat stress conditions in lactating cows at pasture (mean and standard deviation) Temperature (°C) THI Distance (m) Days in milk Lactation number Normal 16.3±0.9 60.8± 1.4 190.5 ±79.1 183.5± 89.4 2.83± 1.78 Heat stress 23.1± 2.3 70.5± 2.9 187.8 ±40.6 182.9± 89.5 2.84 ±1.77 P value <0.0001 <0.0001 NS NS NS The grass composition in terms of crude protein, neutral detergent fibre, acid detergent fibre, water soluble carbohydrate (% in DM) and grass digestibility (%) were 16.3, 47.8, 27.3 and 79.7 in J, and 19.2, 49.6, 26.2 and 80.6 in A. The grass heights were 10.1 and 8.5 cm at entry and 3.1 and 3.1 cm at the exit, respectively in J and A respectively. The grass yield was 1509 and 1437 kg DM ha-1 in J and A respectively, the calculated sward availability was 14 kg DM per day and per cow. On average, the cows received 1.9 kg and 2.0 kg concentrate per day during normal and HS periods. Milking frequency and visits were significantly higher in HS (Table 2). There were significant interactions between THI conditions and periods, the longest periods in A showing no difference between N and HS (1.90 vs 1.98 milking per day and 2.67 vs. 2.51 milking visits respectively). The higher milk production in HS periods can be explained by the increased milking frequency. The cows were attracted to the AS to drink water from a large trough located near the AS during HS periods (unpublished observations). These observations did not confirm those of Spörndly and Wredle (2005) who reported no significant difference in milk yield, milking frequency or water intake between a group of cows with unlimited access to water and a group that had access to water only in the barn. In the present study, with an increase in THI, this behaviour probably increased the number of milking visits and milking frequency. The daily RT was considerably decreased in HS. A reduction of the RT Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 182 in cows suffering from mild to moderate heat stress in the barn was reported by Soriani et al. (2013). Table 2. Daily parameters recorded in cows exposed to normal and mild heat stress at pasture (mean and standard error). Milking frequency Milking visit Milk prod. (kg day-1) Rumination (min day-1) Milk F/P Normal 1.97±0.04 2.69±0.09 18.5±0.3 418.9± 8.7 1.23± 0.01 Heat stress 2.12± 0.04 2.99±0.09 19.7±0.4 339.2± 9.5 1.17 ±0.01 P value <0.01 <0.05 <0.01 <0.0001 <0.10 Decrease in RT is often associated to a reduction in dry matter intake. It seems that, within the conditions of this trial, this was not observed as the milk yield in HS cows was not reduced, and was even increased. F/P tended to be lower in HS. This can be related to a diminution in pH rumen explained by a decrease in saliva production resulting from RT reduction. Such a decrease in ruminal pH when environmental temperature was increased was also described by Mishra et al. (1970). Conclusion It appears from these results that rumination, milking frequency and milk performance of cows milked by an automatic system are affected by a mild heat stress at pasture. More studies are needed to study the impact of the length of HS on these parameters. References Dufrasne I., Robaye V., Knapp E., Istasse L. and Hornick J.L. (2012) Effects of environmental factors on yields and milking number in dairy cows milked by an automatic system located in pasture. Grassland Science in Europe 17, 231-233. Ingraham R.H., Stanley R.W. and Wagner W.C. (1979) Seasonal effects of tropical climate on shaded and nonshaded cows as measured by rectal temperature, adrenal cortex hormones, thyroid hormone, and milk production. American Journal of Veterinary Research 40,1792–1797. Soriani N., Panella G. and Calamari L. (2013) Rumination time during the summer season and its relationships with metabolic conditions and milk production. Journal of Dairy Science 96, 5082–5094. Sporndly E. and Wredle E. (2005) Automatic milking and grazing—Effects of location of drinking water on water intake, milk yield and cow behavior. Journal of Dairy Science 88, 1711–1722. Mishra M., Martz F.A., Stanley R.W., Johnson H.D., Campbell J.R. and Hildebrand E. (1970) Effect of diet and ambient temperature-humidity and ruminal pH, oxidation – reduction potential, ammonia and lactic acid in lactating cows. Journal of Animal Science 31, 1023-1028. Rhoads R.P., Baumgard L.H., Suagee J.K. and Sanders S.R. (2013) Nutritional interventions to alleviate the negative consequences of heat stress. Advances in Nutrition 4(3), 267-276. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 183 Interactive effects of Epichloë endophytes and plant origin on mineral content in Festuca rubra Vázquez de Aldana B.R.1, Helander M.2, Zabalgogeazcoa I.1, García-Ciudad A.1, García-Criado B.1 and Saikkonen K.3 1 IRNASA-CSIC, 37008 Salamanca, Spain 2 Section of Ecology, Department of Biology, University of Turku, 20014 Turku, Finland 3 MTT Agrifood Research Finland, Plant Production Research, 31600 Jokioinen, Finland Corresponding author: beatriz.dealdana@irnasa.csic.es Abstract Red fescue (Festuca rubra) is a perennial grass growing in a wide range of ecological conditions in grasslands. It is often asymptomatically infected by the systemic fungal endophyte Epichloë festucae in European grasslands. In this study we analysed the mineral concentration of infected and uninfected F. rubra plants collected from southern (Spain) and northern Europe (Faroe Islands and Finland), and grown in an experimental field in Spain. Our results showed differences in mineral concentrations between plants collected from different geographic locations and endophyte status of the plants. The endophyte increased the concentration of several nutrients in the plants from northern Finland. The northern fescue plants from Faroe and Finland had, on average, a greater concentration of most mineral elements (P, K, Ca, Mg, S, Fe, Mn, Cu, Ni, Na, Al and Cr) compared to plants from Spain. Keywords: red fescue, macronutrient, micronutrient, fungal endophytes Introduction Festuca rubra L. is a perennial grass very persistent and tolerant to a wide range of ecological conditions. This grass species is asymptomatically infected by the systemic fungal endophyte Epichlöe festucae across European grasslands from Spain (Zabalgogeazgoa et al., 1999) to northernmost Finland and Norway (Wäli et al., 2007). Neotyphodium and Epichloë endophytes are seed transmitted and can be beneficial for host grasses. Endophytes may increase herbivore resistance, and the tolerance of the host grass to abiotic stresses (Malinowski and Belesky, 2006). In some circumstances epichloid endophytes can increase the nutrient content of infected plants (Malinowski et al., 2000; Zabalgogeazcoa et al., 2006; Vázquez de Aldana et al., 2013). The effects of the endophyte on the host plant are variable and dependent on the fungal and plant genotypes as well as environmental conditions (Saikkonen et al., 1998). The objective of the present study was to determine the effect of Epichloë endophyte on the mineral content of F. rubra plants originally collected from different geographic locations across Europe, when grown under the same environmental conditions. Materials and methods Festuca rubra plants were collected from grasslands in North and South Europe: Faroe Islands, northern Finland (Utsjoki), and western Spain (Salamanca) in 2011. At each location plants were collected from two F. rubra populations and transported to the laboratory. The presence of the fungus Epichloë festucae in plants was verified by isolating the fungus from stem and leaf sheaths on potato and dextrose agar (Bacon and White, 1994). The frequencies of endophyte-infected plants were in the ranges of 40-80%, 4-71% and 52-68%, in Spain, Faroe Islands and northern Finland, respectively. Endophyte-infected (E+; n=95) and uninfected (E−; n=58) plants were transplanted in an experimental field in Spain (Salamanca) in 2012. The soil type in the field is an eutric chromic cambisol soil with neutral pH at the surface. Plants were watered during the first weeks from October to November to promote their establishment. In June 2013, aboveground biomass of Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 184 plants was harvested. Dried and ground samples were calcined (450 °C) and ashes dissolved in HCl:HNO3:H2O (1:1:8). The concentrations of mineral elements (P, K, Ca Mg, S, Fe, Mn, Zn, Cu, Na, Al, Co and Cr) were determined by inductively coupled plasma (ICP) spectroscopy. Differences between infected and uninfected plants were tested by one-way ANOVA for each of the three locations. Results and discussion Mineral concentrations between geographic plant origins and endophyte status of the plants differed. The endophyte effect on the mineral concentration was significant for several elements and origins but not for biomass production (Table 1). Table 1. Mineral content in endophyte infected (E+) and non-infected (E-) plants of Festuca rubra from North (Faroe Islands and Finland) and South Europe (Spain). Faroe Islands Element Finland Spain E- E+ P E- E+ P E- E+ P P (g kg-1) 3.49 4.42 0.341 3.26 3.73 0.060 2.74 2.84 0.584 -1 K (g kg ) 9.13 12.9 0.387 9.79 12.1 0.007 7.27 7.08 0.487 Ca (g kg-1) 2.51 2.91 0.978 2.50 3.10 0.009 1.67 1.63 0.991 -1 0.855 0.965 0.787 0.801 0.887 0.141 0.635 0.653 0.533 1.20 1.39 0.348 0.975 1.14 0.040 855 854 0.817 Fe (mg kg-1) 349 204 0.032 231 281 0.050 132 136 0.313 -1 Mn (mg kg ) 42.4 42.2 0.322 39.5 43.3 0.262 34.9 31.3 0.457 Zn (mg kg-1) Essentials Mg (g kg ) S (g kg-1) 21.8 20.4 0.552 19.1 21.5 0.078 18.5 19.6 0.318 -1 6.29 8.08 0.737 5.35 6.70 0.005 4.37 4.29 0.879 -1 3.55 3.06 0.518 2.60 2.54 0.828 1.80 2.02 0.103 0.291 0.253 0.298 0.185 0.196 0.352 0.208 0.209 0.783 Al (mg kg-1) 426 180 0.011 238 292 0.066 131 144 0.129 -1 0.090 0.141 0.325 0.101 0.142 0.286 0.022 0.072 0.253 Cr (mg kg-1) 10.5 5.15 0.112 5.17 5.59 0.721 2.84 3.73 0.017 -1 0.064 0.000 0.220 0.005 0.022 0.144 0.011 0.044 0.292 9.17 14.4 0.419 13.0 13.1 0.938 19.0 19.5 0.886 Cu (mg kg ) Ni (mg kg ) Beneficials Na (g kg-1) Co (mg kg ) Others Cd (mg kg ) Biomass (g plant-1) Plants from northernmost Finland had the largest differences between infected and uninfected plants: the endophyte increased the concentrations of P, K, Ca, S, Fe, Zn, Cu and Al. Uninfected plants from Faroe Islands had greater concentration of Al and Fe than infected plants. In the Spanish plants, Epichloë endophyte increased the Cr concentration, a non-essential element for the plant. Our results showed that Epichloe endophyte increased concentration of several elements in plants from Finland, but only Cr in plants from Spain. These results for Spanish plants differ from previous experiments with half-sib lines in which endophyte increased P and Zn content under field and greenhouse conditions (Zabalgogeazcoe et al., 2006; Vázquez de Aldana et al., Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 185 2013). The greater nutrient content of E+ plants from Finland in Spanish conditions might indicate that nutrient allocation patterns are different for plants adapted to northern habitats compared to those growing in a Mediterranean habitat. We found that plants from northern Europe (Faroe Islands and Finland) had, on average, a greater concentration of most mineral elements (P, K, Ca, Mg, S, Fe, Mn, Cu, Ni, Na, Al and Cr) than plants from southern Europe (Spain). Differences in Zn and Cd concentrations in plants from different origins were not statistically significant. Plants from Spain had greater biomass than northern-origin plants. Differences in nutrient concentrations may be affected by differences in the length and mean temperature of the growing season between original habitats. Thus, under more favourable climatic conditions (light and temperature), northern fescues adapted to short growing season and lower temperature may exhibit higher nutrient uptake and accumulation than southern fescues which are adapted to longer growing season and higher temperature. Acknowledgements This collaborative work has been funded by the International Network for Terrestrial Research and Monitoring in the Arctic (INTERACT), project AGL2011-22783 from the Spanish Ministerio de Economia y Competivividad and Academy of Finland project 137909. References Malinowski D.P., Alloush G.A. and Belesky D.P. (2000) Leaf endophyte Neotyphodium coenophialum modifies mineral uptake in tall fescue. Plant and Soil 227, 115-126. Malinowski D.P. and Belesky D.P. (2006) Ecological importance of Neotyphodium spp. grass endophytes in agroecosystems. Grassland Science 52, 1-14. Saikkonen K., Faeth S.H., Helander M. and Sullivan T.J. (1998) Fungal endophytes: A continuum of interactions with host plants. Annual Review of Ecology and Systematics 29. 319-343. Vázquez-de-Aldana B.R., García-Ciudad A., García-Criado B., Vicente-Tavera S. and Zabalgogeazcoa I. (2013) Fungal endophyte (Epichloë festucae) alters the nutrient content of Festuca rubra regardless of water availability. PLoS ONE 8(12): e84539. doi:10.1371/journal.pone.0084539. Wäli P.R., Ahlholm J.U., Helander M. and Saikkonen K. (2007) Occurrence and genetic structure of the systemic grass endophyte Epichloe festucae in fine fescue populations. Microbial Ecology 53, 20-29. Zabalgogeazcoa I., Vázquez-de-Aldana B.R., García-Criado B. and García-Ciudad A. (1999) The infection of Festuca rubra by the fungal endophyte Epichloë festucae in Mediterranean permanent grasslands. Grass and Forage Science 54, 91-95. Zabalgogeazcoa I., García-Ciudad A., Vázquez de Aldana B.R. and García-Criado B. (2006) Effects of the infection by the fungal endophyte Epichloë festucae in the growth and nutrient content of Festuca rubra. European Journal of Agronomy 24, 374-384. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 186 Soil carbon status of survived and restoring wood pasture in the protected area Natura 2000 Slepetiene A., Slepetys J., Liaudanskiene I., Jokubauskaite I., Stukonis V. Institute of Agriculture, Lithuanian Research Centre for Agriculture and Forestry, Instituto al. 1, Kedainiai distr., Lithuania Corresponding author: alvyra@dobilas.lzi.lt Abstract Soils of survived and restoring wood pasture in the EU protected areas (Natura, 2000) were investigated. Soil samples were collected in 2012 from the 0–10, 10–20 and 20–30 cm depths. The study showed significantly different quantities of SOC and carbon forms accumulated in the soil. The soil of restoring wood pasture contained more organic carbon in 0–30 cm soil layer (24.0 g kg-1) than in soil of the survived wood pasture (17.0 g kg-1). The soils of protected areas differed in organic carbon stability. The largest amount of chemically (C chem. resist., C MHA) and physically (Cclay) stabilized organic carbon forms, which represents the resistance to degradation and the possibility to sequestration, was established in the soil of restoring wood pasture. From an environmental viewpoint, the mentioned soil use is very important in order to accumulate carbon in the soil, to reduce the degradation of soil organic carbon and atmospheric CO2 emission. The content and relative share of labile carbon (CH20) in the soil of survived pasture was higher and increased with depth. Keywords: restoring pasture, wood pasture, soil carbon, water soluble carbon, stabile carbon Introduction Forested pastures and wooded meadows are habitats that have significantly decreased or almost disappeared. Patches of these habitats, once numerous in Lithuania, have survived in the central part in moraine plain landscape. These types of habitat are a reminder of the very distant past: they existed even before the emergence of agriculture and livestock production and were supported by herds of large wild herbivores such as bison and tarpan. After World War II, wooded pastures disappeared rapidly because of the ban on grazing in forests and due to largescale land reclamation. These wood pastures became overgrown with forest vegetation and the former diversity of the pasture vegetation declined. One of the tasks of nature protection is to preserve these pastures and to restore abandoned ones. The establishment of the Natura 2000 Network is one of the main actions undertaken at the European level to contribute to the maintenance of biodiversity (Bartula et al., 2011). The main functions of the soil – biomass production and biodiversity, source of raw materials, storing, filtering and transforming nutrients, substances and water, physical and cultural environment for humans – are all directly dependent on the carbon pool. In line with new trends in global research, with increasingly more investigations and quantitative determinations of carbon flows being performed in order to reduce the negative impact of human activity on the environment, more attention in research has been given to carbon-compound transformations in agricultural soils (Liaudanskienė et al., 2011). Agricultural grassland soils are comprehensively investigated in Lithuania, but there is no assessment of soil carbon composition in old wood pastures and in pastures under restoration. We hypothesize that differences in the amount and type of the organic matter inputs, and wood-pasture soil use affect the accumulation and stability of soil organic carbon. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 187 Materials and methods The soil Endocalcari-Endohypergleyic Cambisol (CMg-n-h-can) of the EU-protected area of wood pasture (Klampute) was investigated in this research. The soil of the survived wood pasture and of 'restoring wood pasture' (i.e. wood pasture that is under restoration) where old trees have been retained and shrubs were cut, was investigated. Soil samples were taken in 2012 from the 0–10, 10–20 and 20–30 cm depths with 6 boreholes per replicated plot. Three field replicates of wood pasture were investigated. For chemical analyses the soil samples were air-dried, visible roots and plant residues were manually removed, and the soil samples were then sieved through a 2-mm sieve and homogeneously mixed. For the analyses of SOC content and mobile humic acid fractions, an aliquot of the soil sample was passed through a 0.25 mm sieve. SOC content was determined by photometric procedure at the wavelength of 590 nm using the UV-VIS spectrophotometer (Cary 50; Varian, USA) after wet combustion according to Nikitin. Chemo-destructive fractionation was used for determination of stable (chemically resistant) carbon (Cchem resist.). Mobile humic acids (MHA) were extracted by 0.1M NaOH solution and determined according to Ponomariova and Potnikova. The data of chemical analyses were processed (P < 0.05) by the statistical program STAT ENG for EXCEL version 1.55. Each variable (n = 3) was displayed as mean ±standard error of the mean. Results and discussion Klampute is one of the areas (12 ha) where restoration of habitats is being carried out. The vegetation structure and species composition that are typical of Fennoscandian wood pastures (Mosquera–Losada et al., 2009) have survived in the southern part of the area. Single old oaks (Quercus robur L.) and large hawthorn (Crataegus spp.) shrubs, including intrusive herbaceous plant communities, grow there. The northern part of the area, formerly a Fennoscandian wood pasture, is overgrown with a dense forest. Species inherent to forest plant communities are entrenched in place of extinct meadow herbaceous plants. Evidence of the previously existing habitat can only be judged from the existence of isolated old trees (Quercus robur L., Tilia cordata Mill., Fraxinus excelsior L.). Tree crowns were matted, and growing stock closeness was 95% against the clearance. Picea abies (L.) H. Karst is quite widespread there, and this species is almost non-existent in the southern part of the area. Some shade and acidic soiltolerant plants have been found, including Oxalis acetosella L., mosses Eurhynchium angustirete (Broth.) T. J. Kop., Pleurozium schreberi (Brid) Mitt. In 2011, restoration of the habitat in the Klampute area was launched. Young trees that were non-specific to the habitat were cut down during the winter. The grass was mown in 2012 and 2013. It is planned to acquire cattle, which will be grazed on the restored wood pastures in the future. Restoration success will depend on the rational use of these pastures. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 188 The study showed significantly different quantity of SOC and carbon forms accumulated in the soil (Table 1). Table 1. The soil variables of survived and restoring pasture in the protected area, 2012 Site description Survived wood pasture Restoring wood pasture Soil layer (cm) SOC 0–10 27.3±1.0 10–20 CH20 CH20 % from SOC C chem. resist. g kg-1 0.49±0.03 1.79 16.8±1.1 0.34±0.02 20–30 7.0±1.0 0–30 17.0 0–10 C CMHA Cclay chem. resist. from SOC g kg-1 0.32±0.036 1.17 4.25±0.69 11.4±0.33 2.02 0.11±0.004 0.65 2.30±0.68 7.64±0.50 0.22±0.02 3.14 0.08±0.001 1.14 0.40±0.11 4.64±0.75 0.35 2.32 0.17 0.99 2.32 7.89 34.3±5.9 0.54±0.04 1.57 0.30±0.016 4.58 6.31±0.62 17.6±0.80 10–20 23.5±3.1 0.37±0.01 1.57 0.39±0.002 1.66 4.87±0.67 12.8±0.47 20–30 14.2±1.2 0.26±0.01 1.83 0.25±0.046 1.76 2.74±0.59 7.29±0.41 0–30 24.0 0.39 1.66 0.31 2.67 4.64 12.6 g kg-1 The soil of 'restoring wood pasture' contained more organic carbon (in the 0–30 cm soil layer): 24.0 g kg-1 compared with 17.0 g kg-1 in the soil of the surviving wood pasture. In the 0–10 cm soil layer of survived wood pasture, the most readily degradable water-soluble carbon accounted (1.79 g kg-1) for a larger relative share of the SOC (2.32%). In restoring wood pasture overgrown by forest, more carbon was physically stabilized through binding with clay minerals (Cclay). Humification processes were also most intensive in the topsoil layer (0–10 cm), where the concentration of mobile humic acids (CMHA) was the highest (6.31 g kg-1) compared with that in the survived pasture (4.25 g kg-1). The soils of the protected areas differed in organic carbon stability. All carbon indicators tested showed higher accumulation of SOC and chemically and physically stabilized carbon forms in the restoring wood pasture. One of the possible explanations for this is that the residues and roots of grasses decompose faster than residues and roots of broadleaf trees; this is likely due to the higher content of easily decomposable compounds found in grasses. Furthermore, the survived wood pasture was used, i.e. the grass has been mowed or grazed, and therefore there were less organic residues returned to the soil surface, whereas the re-naturalization was held in the restoring wood pasture. The soil of the survived wood pasture contained more labile, water-soluble carbon (CH20), which shows its higher predisposition to transformation. Conclusion The largest amount of chemically and physically stabilized organic carbon forms, which represents the resistance to degradation and the possibility for sequestration, was established in the soil of the restoring wood pasture. The relative share of chemically resistant carbon (Cchem. resist) was higher in the soil of the restoring wood pasture and it decreased with depth. From an environmental viewpoint, the soil use is very important in order to accumulate carbon in the soil, to reduce the degradation of soil organic carbon and the net atmospheric CO2 emission. Conversely, the content and relative share of labile carbon (CH20) in the soil of the survived wood pasture was higher and it increased with depth. Acknowledgements This research was funded by grant of the Research Council of Lithuania No. MIP-039/2012. References Bartula, M., Stojšić, V., Perić, R. and Kitnaes K.S. (2011) Protection of Natura 2000 habitat types in the Ramsar Site Zasavica Special Nature Reserve in Serbia. Natural Areas Journal 31, 349–357. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 189 Liaudanskienė I., Šlepetienė A. and Velykis A. (2011) Changes in soil humified carbon content as influenced by tillage and crop rotation. Zemdirbyste-Agriculture 98, 227–234 Mosquera-Losada M.R., McAdam J.H. and Romero-Franco R. (2009) Definition of components of agroforestry practices in Europe. In: Rigueiro-Rodriguez A., McAdam J.H., Mosquera-Losada M.R. (eds.) Agroforestry in Europe. Curent status and future prospects. Springer, Berlin, pp.3-19. Ponomareva V.V. and Plotnikova T.A. (1980) Humus and Soil Formation. Leningrad: Nauka. 258 p. (in Russian). Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 190 Plant succession and soil carbon sequestration potential of abandoned arable fields in a sub-humid Mediterranean environment Karakosta C.,1 Pappas I.A.2 and Papanastasis V.P.2 1 Ministry of Environment, Energy and Climatic Change, Chalkokondili 31, 10432 Athens E-mail: chkarako@hotmail.com 2 Laboratory of Rangeland Ecology, Aristotle University of Thessaloniki, 54124, Thessaloniki Corresponding author: chkarako@hotmail.com Abstract Abandonment of arable land is a major land use change in the Mediterranean region. The effect of this land use change on plant succession and on soil organic-carbon pools is an important factor for mitigating the increase of global greenhouse gases and climate change. The aim of this research was to quantify the soil carbon stocks of old fields along the secondary succession with different ages since abandonment (1-60 years) in a sub-humid Mediterranean climate of Northern Greece. The results revealed that soil is an important carbon sink of atmospheric CO2, with mean average carbon sequestration of 2.68 t ha-1 and carbon sequestration rate 0.04 t ha-1 yr-1. This function was closely related with plant succession characterized by an evolution from annual to perennial herbaceous functional groups while woody species appeared in later stages. Keywords: soil carbon sink, plant functional groups, biomass, climate change Introduction Land use changes have significant effects on greenhouse gas emissions and carbon stocks in soil and vegetation (Feddema et al., 2005). In most European countries, the transformed economies and social conditions of previous decades have resulted in significant land use changes leading to agricultural intensification, industrialization and migration of people from the rural areas (Alberti et al., 2008). As a consequence, areas of marginal agriculture were abandoned, thus initiating secondary succession of the vegetation. This has occurred particularly in the Mediterranean region, and in Greece such abandonment is widespread in the mountainous areas (Papanastasis, 2007). Soil organic matter (SOM) can be a source or sink for atmospheric CO2 depending on land use, management of soil, vegetation and water resources (Lal, 2009). The aim of this research was to quantify the current SOC stocks in old fields with different ages of abandonment. Materials and methods The research was conducted in Taxiarchis village located in the Holomontas mountain of Chalkidiki, North Greece. The climate is sub-humid Mediterranean. Seven fields were chosen (mean altitude 850 m), each reflecting a different period of abandonment of agricultural use, following a time series of ten years (abandonment age 1, 10, 20, 30, 40, 50 and 60 years each approximately). In each field, plant cover and species composition were measured along 5 transects, using the point method (Cook and Stubbendieck, 1986). Biomass was measured in 10 quadrats (1 x 1 m2 each for woody species and 50 x 50 cm2 each for herbaceous species). All biomass samples were oven dried at 60º C for 48 h. Soil samples were collected at a depth of 0-20cm. In the laboratory, soil total nitrogen and soil organic matter concentration were measured by the K2Cr2O7 method using the modified Kjeldahl wet digestion procedure of Miller and Keeney (1982). The soil organic C content was estimated on a percentage basis and converted to tons per hectare using bulk density and soil depth. All data analyses were conducted using the software package SPSS 11.0. Significant differences for all statistical tests were evaluated at the level of P≤0.05. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 191 Results and discussion Agricultural abandonment resulted in spontaneous colonization of the old fields by various plant functional groups, with annual grasses, legumes and broadleaved herbs dominating in the early stages and their perennial counterparts replacing them in later stages, and with woody species appearing in older stages (Table 1). According to Bonet and Pausas (2004), the recovery of woody vegetation in Mediterranean old fields is usually hindered by herbaceous competition. Biomass tended to increase almost linearly along succession until 50 years, which is consistent with other studies on secondary succession in mesic Mediterranean conditions (La Mantia, 2008), but reduced from the age of 60 years. Table 1. Mean average contribution (%) of functional groups and biomass (g m-2) in fields with different ages since abandonment Functional groups and biomass Age of abandonment (years) Annual 1 33.51a 10 20.40c 20 26.90b 30 8.34d 40 0.56e 50 2.71de 60 0.73e Perennial Biennial Annual 2.79e 0.59b 27.84a 6.96e 0.38b 16.03b 22.16d 0.9b 17.94b 46.78a 0.12b 5.04c 32.26c 0 1.44c 46.19ab 2.50a 2.42c 38.55b 0 2.39c Legumes Perennial Annual 23.62b 11.65a 46.80a 9.43ab 25.56b 6.54b 25.53b 2.08bc 22.84b 0.23c 17.59c 2.18bc 20.25b 2.47bc Woody species Perennial Shrubs 0 0 0 0 0 0 6.20b 5.91b 10.84a 6.08b 2.69c 22.78a 5.85b 6.56b Trees 0 0 141.03bc 0 0 0 146.74bc 0 0 0 113.18c 0 0 0 222.86ab 24.43c 0 25.75a 261.37a 10.12d 0.94b 0 187.42b 438.18a 8.09a 15.11b 53.84d 89.57b Grasses Broadleaved herbs Bracken fern Biomass Herbaceous Woody Soil carbon dynamics after abandonment is connected to the development of the natural vegetation through secondary succession processes (Kosmas et al., 2000). The high soil organic and nitrogen content found in the one-year-old field since abandonment perhaps indicates previous organic amendments (McLauchlan, 2006). Overall, organic carbon and nitrogen tended to increase from the early to the middle stage of abandonment, with a maximum reached at 30 years (Table 2). Table 2. Mean average values (and standard deviations) of soil data in fields with different ages since abandonment Soil data Age of abandonment (years) Bulk density (g cm-3) 1 0.22±0.03 10 0.27±0.04 20 0.20±0.02 30 0.25±0.05 40 0.23±0.02 50 0.28±0.03 60 0.27±0.04 Carbon (%) 2.43±0.56 1.51±0.42 3.08±0.77 2.96±0.47 2.06±0.28 1.96±0.66 2.13±0.47 Nitrogen(%) 0.22±0.04 0.13±0.03 0.27±0.08 0.32±0.09 0.22±0.03 0.17±0.06 0.14±0.05 C:Ν 11.29±1.08 12.05±1.15 11.39±0.84 9.41±2.84 9.58±1.89 11.52±4.19 15.22±4.4 2.2±0.64 2.4±0.90 2.37±0.33 3.48±0.59 3.10±0.88 2.05±0.57 2.69±0.62 Carbon (t ha-1) This could be attributed to the increased presence in the cover of herbaceous plant functional groups, especially perennial grasses, which are a significant contributor of C in the soil (Santos et al., 2011), while legumes usually increase the N availability due to N2 fixation and higher N input via litter decomposition (Cuesta et al., 2012). Thereafter, both carbon and nitrogen were reduced but remained stable until 60 years. This was maybe due to the increase of woody species in the vegetation composition, which contributed more biomass, as compared to herbaceous species (Table 1), but less soil organic carbon. According to Satti et al. (2003), species with slow growth rates such as Mediterranean woody species, exhibit high Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 192 concentrations of secondary compounds and chemical defences against herbivores, and have high lignin content which promotes slow decomposition (Castro et al., 2010). This is also supported by the higher C:Ν ratio found in the field of 60 years since abandonment. Soil carbon storage (t ha-1) followed the same trend, with carbon content showing a mean average value of 2.68 t ha-1 and a rate of 0.04 t ha-1 yr-1. According to Cuesta et al. (2012) carbon accumulation in soils under secondary succession after crop abandonment is slow in Mediterranean environments; however, soil C storage is considered a viable option to mitigate climate change (Lal, 2004). Conclusions The results indicate that abandoned arable fields in a sub-humid Mediterranean environment are carbon sinks and that organic carbon storage in the soil is closely related to plant succession characterized by an evolution from annual to perennial herbaceous functional groups with woody species appearing in later stages. References Alberti G., Peressotti A., Piussi P. and Zerbi G. (2008) Forest ecosystem carbon accumulation during a secondary succession in the Eastern Prealps of Italy. Forestry 81, 1-11. Bonet A. and Pausas J.G. (2004) Species richness and cover along a 60-year chronosequence in old-fields of southeastern Spain. Plant Ecology 174, 257-270. Castro H., Fortunel C. and Freitas H. (2010) Effects of land abandonment on plant litter decomposition in a Montado system: relation to litter chemistry and community functional parameters. Plant and Soil 333, 181-190. Cook C.W. and Stubbendieck J. (1986) Range Research: Basic Problems and Techniques. Society for Range Management, Colorado, 317 pp. Cuesta B., Rey Benayas J.M., Gallardo A., Villar-Salvador P. and González-Espinosa M. (2012) Soil chemical properties in abandoned Mediterranean cropland after succession and oak reforestation. Acta Oecologica 38, 5865. Feddema J., Oleson K., Bonan G., Mearns L., Buja L., Meehl G. and Washington W. (2005) The importance of land-cover change in simulating future climates. Science 310 (5754), 1674 – 1678. Kosmas C., Gerontidis S.T., Marathianou M. (2000) The effect of land use change on soils and vegetation over various lithological formations on Lesvos (Greece). Catena 40, 51–68. La Mantia T., Rühl J., Pasta S., Campisi D. and Terrazzino G. (2008) Structural analysis of woody species in Mediterranean old fields. Plant Biosystems 142, 462–471. Lal R. (2004) Soil carbon sequestration impacts on global climate change and food security. Science 304, 16231627. Lal R. (2009) Challenges and opportunities in soil organic matter research. European Journal of Soil Science 60, 159-168. McLauchlan K. (2006) The nature and longevity of agricultural impacts on soil carbon and nutrients: a review. Ecosystems 9, 1364-1382. Miller R.H and D.R. Keeney (1982) Methods of soil analysis. Part 2: chemical and microbiological properties, 2nd edn. American Society of Agronomy, Soil Science Society of America, Madison. 228 pp. Papanastasis V. P. (2007) Land abandonment and old field dynamics in Greece. In: Cramer V.A. and Hobbs R. J. (eds) Old fields: dynamics and restoration of abandoned farmland. Washington, D.C.: Island, pp. 225-246. Santos N.Z., Dieckow J., Bayer C., Molin R., Favaretto N., Pauletti V. and Piva J.T. (2011) Forages, cover crops and related shoot and root additions in no-till rotations to C sequestration in a subtropical Ferralsol. Soil and Tillage Research 111(2), 208-218. Satti P., Mazzarino M.J., Gobbi M., Funes F., Roselli L. and Fernandez H. (2003) Soil N dynamics in relation to leaf litter quality and soil fertility in north-western Patagonian forests. Journal of Ecology 91, 173-181. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 193 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 194 Theme 2 ‘Grasslands and ecosystem services’ Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 195 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 196 Theme 2 invited papers Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 197 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 198 Functions of grassland and their potential in delivering ecosystem services Isselstein J. and Kayser M. Georg-August-University Göttingen, Department of Crop Production, Grassland Science, Von-Siebold-Straße 8, 37075 Göttingen, Germany Corresponding author: jissels@gwdg.de Abstract Grassland ecosystems provide ample services that are vital to nature and mankind. The role and the potential of grasslands to provide services has always been valued by farmers, it is now being increasingly recognized by the society as a whole. Services result from complex networks of interrelated ecosystem processes and both, processes and services, occur at a range of spatial and temporal scales. Generally, society expects that grasslands provide a range of services at the same time. This is difficult as often one intended utilization will dominate the others. The relationship of many of these services is therefore characterized by tradeoffs. Farmers play a centre in balancing conflicting relationships. One of the best options is to improve management in order to reduce tradeoffs. Research is increasingly needed to support innovative management in this respect. However, the mitigation of tradeoffs is in many cases difficult and limited and it requires an in-depth understanding of teh functional relationship between services. Balancing persistent tradeoffs is a matter of setting priorities which requires a discourse finally leading to an understanding among the various stakeholders. The practicability of the ecosystem services concept is most challenging both for research and development. The present review introduces conceptual ideas on the role of the ecosystem services concept in grassland farming. Major services and tradeoffs among services are identified. The feasibility to balance tradeoffs is analyzed in more detail in two case studies. One study is dealing with the various ways to provide protein to dairy cows and with the implications for other ecosystem services. The second example deals with wet grassland farming and the conservation of rare meadow birds. Keywords: grassland, ecosystem function, biodiversity, water, climate, tradeoff Introduction The concept of ecosystem services has been developed to highlight the benefits of nature to mankind. De Groot (1987) stated that: nature provides regulation processes to maintain clean air, water and soil; it provides many resources (ranging from ore to wildlife products and genetic material) as well as space and a suitable substrate for many human activities (e.g. agriculture, recreation, etc.) and nature provides opportunities for reflection, aesthetic enjoyment and spiritual enrichment. De Groot (1987) hoped that the maintenance of environmental functions (goods and services) may serve as a unifying concept to provide a common long-term goal for both economists and conservationists and that the function-concept could be a useful instrument to merge ecological principles and economic procedures. In 1997 a team of scientists and economists estimated the economic value of the world’s natural ecosystems (Constanza et al., 1997). An early general review of ecosystem services by a range of experts in different topics was provided by Daily (1997) and further elaborated into a framework (Daily et al., 2000, Daily, 2000). Key issues of biodiversity, ecosystems stability, and ecosystems productivity were reviewed by Tilman (1999, 2000), Grime (1997), Chapin (2000), Kaiser (2000) and McCann (2000). The concept of tradeoffs, the reduction of one ecosystem service as a consequence of increased use of another, was described by Rodriguez et al., (2006) and was shown to be effective in pointing to ecological aspects of agricultural systems and to account for negative side effects of production (Power, 2010, Sanderson and Wätzold, 2010). In agricultural systems it is of interest i) how these systems depend on natural Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 199 ecosystems supporting them, ii) how producing goods (energy, protein, and fibre) and the supply of other services interact, and iii) how the identification and quantification of possible disservices and tradeoffs that are directly or indirectly related to production might affect the efficiency of production, affect other services of this system, and affect neighbouring natural ecosystems, water, soil and air. The following review is concerned with agricultural grassland systems in Europe. Even though the review is focusing on one type of land-use only, namely grassland, the situation is highly complex. Grasslands are found under widely varying environmental conditions and their functioning is characterized by multiple interactions of environmental and management factors. The provisioning of certain services is thus difficult to predict. The situation is further complicated by the fact that functions and services are scale dependent. Services are valued by stakeholders. As it is most unlikely that all possible tradeoffs among different ecosystem services can be resolved there is a need to focus on those services that are considered by stakeholders to be most important, i.e. the grand challenges. Certainly, the identification of those challenges is not the same among the different stakeholders and the challenges may vary depending on the region and the status of the natural and agricultural resources. Therefore, in this review, general interests are preferentially considered compared to very particular interests. The perception of the challenges among stakeholders is a result of their intentions, be it to produce milk and meat from grassland or to conserve the diversity of the grassland-related biota. Based on intentions, the stakeholders are setting priorities. The grand challenges, the intentions and the priorities will be used to set a framework for the analysis of ecosystem functions and services. After an introductory chapter on terms and basic concepts of ecosystem services we continue with describing major services from grasslands followed by an analysis of the potential to balance tradeoffs among services through adapted management. This procedure is illustrated in more detail in two case studies. Terms, definitions and concepts of ecosystem services In a report on the value of the world’s ecosystem services and natural capital, Constanza et al. (1997) gave a set of useful definitions to distinguish ecosystem functions, goods and services. Ecosystem functions refer to the habitat, biological or system properties or processes of ecosystems, for example primary production, nutrient cycling, or biodiversity (Constanza et al., 1997). Ecosystem goods and services represent the benefits humans derive, directly or indirectly, from ecosystem functions. Those benefits can be classified as provisioning services such as food and water, as regulating services that mitigate the effects of floods, droughts or land degradation, as supporting services such as soil formation or nutrient cycling and as cultural services that provide recreational, spiritual benefits (Millennium Ecosystem Assessment, 2005). Ecosystem services and functions usually do not show a one-to-one correspondence. They are not independent of one another and are often highly non-linear (Power, 2010). Multi-criteria assessment techniques are commonly used when systems with multiple functions and services are to be analysed. Results are then presented in ‘spider diagrams’ with as many axes as services from the system (DeFries et al., 2004). As an example, Figure 1 depicts different grassland farming situations with five main services with the production function being represented by axis 1. In a traditional pre-industrial grassland system (Figure 1A) only the production function was highly valued and intended, but low input and relatively low yields allowed ample provision of a range of other services. With the intensification of grassland farming (Figure 1B) the production increased markedly, almost approaching the site-specific potential, however at the expense of other services. Abandoning grasslands from agricultural use, which was a common phenomenon in many central and eastern European countries during the 1990s, ignores the production function completely (Figure 1C). While this may support Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 200 some functions and services, such as the stabilization of soils or the provision of clean water, it may be detrimental to other services such as biodiversity which has been shown to decrease with long term successional change (Kesting et al., 2009). Figure 1. Provision of ecosystem services from grassland in (A) a pre-industrial situation, (B) current intensive farming and (C) a grassland abandoned from agricultural use, ecosystem services vary from 0 (no service/disservice) to 100% (maximum service possible under the particular conditions), schematic. 1: production service (forage, milk, meat), 2-5: other services, such as water, climate, diversity, culture). In agricultural systems, humans make management choices which change the type, magnitude and relative mix of services provided by the system. These effects are termed tradeoffs and occur when the provision of one ecosystem service is reduced as a consequence of increased use of another service (Rodriguiz et al., 2006). Such tradeoffs are typically related to the production of goods as food, fibre and bioenergy, and affect regulating services as water purification, soil conservation, carbon storage or might result in direct disservices as water and air pollution by emissions (Power, 2010). The impact of tradeoffs can occur on a range of spatial and temporal scales: are the effects of the tradeoff felt locally, for example on-farm, or at a more distant location, does it occur instantly or later? Are the effects reversible and if so, how quickly can they be reversed (Power, 2010)? Presenting results in ‘spider diagrams’ reflects a descriptive approach of assessing multiple ecosystem services. They do not provide in depth information on the functional relationships. Simply describing the magnitude of services from a production system falls to short. This does not provide us with the sound understanding of functional relationships needed for managing and balancing tradeoffs and disservices. However, relationships among different services are often characterized by multiple interactions, which are difficult to analyze. In order to explore the nature of relationships among services, a simple situation with only two services is depicted in Figure 1. The relationship of these services may be linear or non-linear, and it may be of a mutually exclusive, a non-interacting or mutually beneficial character (Figure 2). A typical example for the mutually exclusive type of relationship is the relation between production and nitrogen emission risk. A high grass or livestock production is usually achieved through a high input of nitrogen fertilizer. In general, the higher the fertilizer input the higher are nitrogen losses to the environment in form of denitrification, ammonia volatilization and/or nitrate leaching. However, ample research on the nitrogen use efficiency of grassland systems over the last decades has shown that there is - through better management - a considerable potential to minimize this tradeoff without compromising the production function. This is demonstrated by the dashed line of Figure 2A; the amount of service 1 is increased without change of the amount of service 2, and vice versa. The ‘no-interaction’ type of relationship (B) might be explained by the water retention function of grassland and grass production: as long as the grass sward is intact and thus the soil covered by a permanent vegetation, surface water run off will Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 201 be negligible, almost irrespective of the yield potential of the sward. Examples for mutually supportive functions (Figure 2C), i.e. a full win-win situation, are rare. Research over the last 15 years on sown, so-called artificial grasslands with a defined set of species in different mixtures, has revealed a considerable production benefit from increased phytodiversity. In temporary grassland (leys) a mixture of only a few grasses and legumes can be sufficient to yield much of this biodiversity effect (Lüscher et al., 2008). However, older investigations on permanent grasslands do not confirm a mutually supportive relationship between phytodiversity and production, but rather suggest a humpback function of the D-type of Figure 2 (e.g. Oomes, 1992). More recent research on permanent grassland reports either no or even a negative relationship (Assaf et al., 2011; Schneider et al., 2011; Rose and Leuschner, 2012). Figure 2. Schematic representation of the different types of functional relationships between two ecosystem services. Ecosystem services vary from 0 (no service/disservice) to 100% (maximum service possible under the particular conditions) provision. Kind of relationship of services: (A) linear mutually exclusive, (B) linear no interaction, (C) linear mutually supportive, (D) non-linear. Solid line: current status, dashed line: changed relationship, e.g. through improved management, with increased total service, see also text. Grassland ecosystem services – The big five Production Grasslands are a main resource for agricultural production in Europe. According to the EU statistics they cover some 35% of Europe's utilized agricultural area and some 8% of the total European land surface (EUROSTAT, 2013). Recently, Peeters et al. (2014, this volume) suggested a new terminology of European grasslands. Grasslands are being defined as agricultural land that is covered mainly by grasses and the grass sward is either temporary or permanent, while permanent grassland has not been renewed for at least 10 years. The percentage of grassland of the utilized agricultural area in the different EU countries is highly variable and ranges from almost zero to almost 100% (Figure 3A). The reason why the grassland differs so much across the different countries is not readily identified. A comparison of agricultural land-use data did not reveal a clear effect of the size of the available area within a country. In general, permanent grassland sites in Europe can be found on agricultural land that is less suitable for arable farming. A main reason for this is often a surplus of water. This is confirmed by an obvious relationship between annual precipitation in a country and the percentage of permanent grassland of the utilized agricultural area (Figure 3B). The grassland area in Europe has been declining over the last decades. Between 1990 and 2006 the annual decline was -1.2%, albeit with a high variation between countries (EEA, 2010). Pressure on grassland habitats is still present. Grasslands have been converted mainly into afforested area but also into arable land. The conversion to arable land is particularly important on farms with intensive dairying and biogas production (mainly in Germany, Osterburg et al., 2011). Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 202 Figure 3. Ratio of the total grassland area (TGL) to the utilized agricultural area (UAA) in relation to (A) the UAA and (B) the average annual precipitation of the EU countries (Data source: EUROSTAT 2013, World Bank 2013, own calculations). Grasslands are used for ruminant and equine husbandry and, in some countries, for the production of renewable energy. In Europe, they make up some 75% of the fodder area (grassland plus forage from arable land). The stocking rate of ruminants (cattle and sheep) is on average 1.2 LU (SD 0.7) per ha fodder area, and that of dairy cows is 0.5 (SD 0.3). The stocking rates of equines amount to 0.06 LU (SD 0.06) per ha total forage area and 0.07 (SD 0.08) per ha total grassland area (EUROSTAT, 2013). The production of milk and meat per total forage area varies greatly among countries and is positively related with the stocking rates. The more animals are kept per fodder area the higher is the output of milk and meat. On average, 2216 kg (SD 2075 kg) of milk and 227 kg of meat from ruminants (SD 589 kg) is produced per ha fodder area. Milk and meat yield per fodder area are positively related (Figure 4). This means that a lower milk production is not necessarily compensated by a higher meat production. The forage potential of the fodder areas is obviously highly variable among the different countries, so that in some countries a considerably high livestock performance of both dairy and beef/sheep is possible whereas in other countries it is not. To some extent this is confirmed by modeling data on the grassland productivity in Europe. Smit et al. (2008) estimated grassland productivity to vary between 10 and more than 70 dt/ha in the different environmental zones of Europe. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 203 Figure 4. Production of milk per ha fodder area in relation to the production of meat (cattle, sheep) per ha fodder area in the EU countries (Data source: EUROSTAT 2013, own calculations). Biodiversity Grasslands in Europe harbour a huge biodiversity and they contribute most significantly to the overall diversity of the European agricultural landscapes. As an example, in Central Europe, almost one third of the total higher plant flora (i.e. some 1000 species) is related to grassland while the weed flora of the arable land only covers some 300 species (Ellenberg and Leuschner, 2010). The importance is also reflected by the fact that the grassland area is a major contributor to the high nature value farmland (EEA 2004). However, the biodiversity of grasslands is threatened and for several species groups (e.g., invertebrates, birds, amphibian, reptiles) decreasing trends could so far not be reversed; almost 50% of the grassland habitats are classified as being in an unfavourable conservation condition (EEA, 2013). A drastic example is that of grassland butterflies, as the referring diversity indicator has decreased by some 50% since 1990 (EEA 2013). There are several reasons of biodiversity losses, ranging from conversion of grasslands to arable land or woodland (including fallow land), amelioration of grassland sites through drainage or fertilization or intensification of grassland utilization. Halting losses has been given a high priority, not necessarily among farmers, but from the society perspective. There is an urgent need to make agricultural management more compatible with biodiversity concerns. The discussion is characterized by two main perspectives: (i) How can the diversity of grasslands be conserved and what is required from the agricultural management, and (ii) how can potential benefits from biodiverse systems be exploited for grassland farming. Strong management restrictions imposed on grasslands to maintain the nature conservation function has led to a decreasing interest of farmers in utilizing such land for livestock feeding and grasslands have been abandoned from agricultural use (Isselstein et al., 2005). However, there is research available that has shown potential of integrating grasslands with a priority in nature conservation into up-to-date farming systems (e.g. Wrage et al., 2011; Jerrentrup et al., 2014). Yet, more research and development from the agricultural side is needed to identify and implement grassland farming systems and measures that make use of the grasslands without compromising the nature conservation value. Livestock husbandry systems as well as renewable energy production might be considered in this respect (Wachendorf and Soussana, 2012). Whether and to what extent biodiversity can support grass production is still debated. With regard to phytodiversity, model systems have shown that particularly under low input conditions diversity is enhancing productivity. This was mainly attributed to a higher niche complementarity and an increased resource-use efficiency of Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 204 diverse over monospecific swards (Hector et al., 1999; Tilman et al., 2001; Weigelt et al., 2009). The biodiversity effect is specifically true at a low number of species, and the specific return of any extra species entering the sward is diminishing with a higher species number already being present (Hector et al., 1999). The recently finished EU research project MULTISWARD could confirm that simple sown mixtures offer production advantages over monospecific swards (various authors, this volume). Those effects are obviously due to the functional composition of grass swards rather than to the mere species number (e.g. Grime,1997; Dias et al., 2013). There are only very few examples available showing that a biodiversity effect would persist at a more intensive grassland management and at a higher level of manuring (e.g. Nyfeler et al., 2009). For permanent grasslands, evidence of biodiversity revenues in terms of production is still scarce. There is some indication that with a more diverse botanical composition of the swards the feeding value might increase (Seither et al., 2012). Climate Depending on site conditions and management, grasslands contribute to and mitigate greenhouse gas (GHG) emissions and thus play a key role in agriculture and climate change. A comprehensive review on the topic has been given by Conant (2010). There are three gases through which grassland soils take part in GHG fluxes. CO2 is fixed as organic matter in the soil mainly through plant growth and the formation of transient or stable soil organic compounds and it is released through mineralization mainly after a disturbance event. N2O emissions occur with an increased availability of mineral nitrogen in the soil. CH4 may be released from wet grasslands on organic soils but the indirect emission through grass-fed ruminants is more important (Conant, 2010). In general, agriculture is the largest single contributor to N2O and CH4 emissions to the atmosphere and correspondingly agriculture is expected to reduce emissions. Due to the complex nature of GHG fluxes between soil and atmosphere and the different ways, either direct or indirect via livestock, how grasslands account for emissions, mitigation strategies are not straightforward and single measures may reduce one emission pathway but at the same time may increase another one (Flessa et al., 2012). The amount of carbon sequestration to the soil depends on the grassland management and the soil water status. As a rule, the higher the soil water table the higher the carbon sequestration to the soil or - on organic soils - the lower the carbon release via mineralization of organic matter (Flessa et al., 2012). Maintaining and re-establishing permanent grasslands on organic soils rather than conversion to arable land is a most effective way of maintaining and increasing soil carbon content. It had been assessed that conversion of grassland to arable land would reduce the organic carbon content in the top soil layer (0-30 cm) by 30 to 36% in the long run (Flessa et al., 2012). Increasing grassland productivity through improved management has a potential to increase soil organic carbon (Conant, 2010) although the processes are not fully understood. Fornara et al. (2013) did not find an increased carbon content of grassland soils that had been fertilized with a balanced nutrient mix over 19 years as compared to an unfertilized control treatment. However, an unbalanced fertilization with nitrogen only increased soil organic carbon. It was assumed that nitrogen fertilization increased root growth. Increasing productivity by higher fertilization bears the risk of higher N2O emissions, thus counteracting benefits in carbon storage. N2O emissions are closely related to the nitrogen surplus both at farm and field level balances, which was shown among others by the EU Dairyman project (Aarts et al., 2013). A high nitrogen surplus is either due to heavy nitrogen fertilization or to high stocking rates that come along with high amounts of purchased concentrate and protein feed. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 205 Water Grasslands affect local and regional water cycles (Ehlers and Goss, 2003). Compared to woodlands, grasslands have a considerably higher groundwater recharge as water fluxes from the soil to the atmosphere through interception water are lower (Ehlers and Goss, 2003). The quality of leaching water is often higher under grassland as compared to arable land because of a lower mineral, in particular nitrate, load of the soil water. In temperate climates unproductive evaporational water losses are generally lower than transpirational ones because the soil surface is covered by grass vegetation and thus less exposed to evaporation. The amount of water transpired by the vegetation is directly related to production (de Wit, 1958). Increasing production, e.g. by fertilization, therefore increases water consumption by the grass vegetation (Husemann and Wesche, 1964); yet the water-use efficiency may rise due to a decreasing evaporation to transpiration ratio (Ehlers and Goss, 2003). In a recent publication on seminatural grassland, Rose et al. (2011) confirmed that fertilization increases production and transpiration and that this had a clear decreasing effect on the amount of groundwater recharge. Thus, the way grassland is managed has an impact on the soil water table and the landscape hydrology (Rose et al., 2011). A strong feature of grasslands is the water retention capacity. Well managed grass swards are most effective in protecting soils from water surface runoff and soil erosion. In hilly regions where the soils are prone to surface water runoff the percentage of grasslands in the land use is important for the likelyhood of river floodings. In a model approach, van der Ploeg et al. (1999) calculated an increasing risk of floodings of the river Rhine at Cologne as a result of a loss of permanent grassland through conversion to arable land by 6% over a period of 20 years. They concluded that the maintenance of grasslands should be aimed at in order to decrease the risks of floodings of the large rivers. Culture Grasslands provide aesthetic values, opportunities for recreation and nature education and are a part of an open preferably diversified landscape (Hönigová et al., 2012). Mostly, this applies not only to natural or semi-natural grasslands but to managed grassland as well, while the degree of the socio-cultural service decreases with intensity and loss of biodiversity (Lindemann-Matthies, 2010). Some aspects and implications related to agricultural production, environment and landscape have been extensively reviewed by Gibon (2005) and Quetier et al. (2010) among others. To assess the non-material benefits of people from cultural ecosystem services and how they might be affected by changes in other services, that could be an increase or decrease in the intensity of agricultural management resulting in an altered landscape, is probably the most difficult task. Information on this topic and especially on possible tradeoffs is comparably scarce. In the generally positive perception of grazed grassland, aspects of landscape, aesthetic values, biodiversity and concern for animal welfare are combined (Van den Pol-van Dasselaar, 2008). This can have economic consequences for farmers and consumers when pasture-grazed dairy products are put on the market that are either promoted or ask for higher prices. The challenge of balancing tradeoffs - Two case studies Balancing tradeoffs among ecosystem services sets an important framework for grassland management-related research of today as ecosystem services from grassland are more and more considered as being relevant for society (Hopkins and Wilkins, 2006). Tradeoffs are looked at from various viewpoints and with different foci. Sanderson and Wätzold (2010) in their review on ecosystem services point to a range of pressing issues, e.g. the stocking rate-livestock tradeoff, the forage quantity-quality dilemma, the production-carbon storage relationship or the economic consequences of balancing tradeoffs. They conclude that any new recommendations of policies, programmes or payment schemes must acknowledge and understand the related Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 206 tradeoffs and their interconnections – and that designing cost-effective policies is an important way to mitigate tradeoffs and use resources efficiently. In the following, two case studies are introduced to point at two quite different aspects of grassland management: firstly, intensive grassland management for provision of and protein (and energy) in dairy farming and, secondly, extensive grassland as a habitat for meadow bird breeding. Protein for milk production and grassland management Nitrogen is a key factor in grassland farming and in ruminant nutrition and through the provision of feed from grasslands these two compartments are closely related (for a recent comprehensive overview on this see Taube et al., 2014). Protein is often a limited resource on farms with high yielding dairy cows. Many farmers perceive grass protein as not fully meeting the dietary requirements of high performing dairy and rely on in-bought protein feed. These protein sources are of high quality and they yield extra milk. However, a nitrogen balance of the dairy ration often shows that the mere amount of nitrogen provided with the roughage was sufficient. Depending on how much protein is purchased and where it had been produced, ecosystem services from agricultural land are concerned. For example, if soyabean from overseas is used ecosystem services from the farmland where soyabean is fed as well as services from the arable land where the soyabean had been produced are affected. The latter being an example for a direct dependency of services at a global scale. Obviously, there are economic benefits for global business partners which can be explained by the comparative advantage of production and trade. However, there are also obvious tradeoffs of this practice in terms of emission risks (e.g. greenhouse gas mitigation on the dairy farm) or biodiversity conservation (e.g. losses of natural diversity where the feed is produced). Del Prado et al. (2013), using practical farm data in a modelling approach, state that dairy diet is a strong factor explaining differences in GHG emissions from milk production and that the amount of feed that could have been utilized for human nutrition is a good predictor of the carbon footprint. The purchase of concentrate and protein feed has long been shown to boost nitrogen surpluses of nutrient balances of dairy farms (e.g. Aarts et al., 1992, as one of the early papers) with the result that the grass and arable fields are burdened with heavy fertilization. Thus, there is a tradeoff between milk production (via increased purchased feed) and the nitrogen emission risk of dairy farming. However, this tradeoff between milk production and nitrogen efficiency is not a fully exclusive one as is shown with the solid line in Figure 2A. If nitrogen in the roughage is limited, feeding additional protein will increase the overall nitrogen efficiency because it improves the conversion of the energy contained in the roughage into milk product. In addition, this will also reduce the land-use for producing the dairy feed (Yan et al., 2013a). The tradeoff may also be diluted by improved techniques of manure treatment on the farm. These measures will increase production through increased efficiency of carbon and nitrogen fluxes in the soil – grass – ruminant system and the relationship between the services will rather follow the dashed line of Figure 2A. The extent of the global tradeoff between the production service of the dairy and the biodiversity service overseas is dependent on the land type that is allocated to grow soyabean, e.g. whether it is land that had recently been converted from tropical forest and extensively managed rangelands or it is established cropland. In order to encounter the global tradeoff it has been suggested to increase the use of domestic protein sources (Taube et al., 2014). In this respect, grasslands and grass clover leys have a huge potential as protein yields per ha grass exceed those of grain crops by far (Taube et al., 2014). In general, there are two strategies to increase the protein yield from grassland. The first strategy includes an increased nitrogen fertilization or fostering the growth of forage legumes (i.e. grass-clover systems). Clover-based systems without using additional mineral fertilizer have a clear advantage in terms of nitrogen emission risks, albeit with somewhat reduced grass yields. Compared to grass-nitrogen systems, losses via denitrification (N2O) and leaching Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 207 (NO3) are lower, i.e. there are fewer disservices with regard to GHG releases (Yan et al., 2013b) and ground water contamination with nitrate (Eriksen et al., 2010). The difference between grass-nitrogen and clover-based systems is likely to decrease with increasing stocking rates (Andrews et al., 2007), in particular when high stocking rates are based on bought-in feed. On dairy farms, nitrogen is circulating in great quantities in the system while only small amounts are released from the system via milk and meat (Whitehead, 1995) The remaining nitrogen has to be recycled in the soil-plant system. The higher the stocking rates and the higher the amount of milk being produced per ha, the larger the nitrogen flux. Grassland-based livestock production ranges from extensive, all-year grazing on permanent grassland to intensive housing in stables (indoor keeping) based on cut grassland, leys, maize and concentrates. Pathways for losses of N by leaching and gaseous emissions, mainly NH3 and N2O differ among systems. With housed livestock, dairy especially, and application of manures, NH3 losses can be substantial while N leaching is usually relatively small (Benke, 1992; Anger, 2001). Cuttle and Jarvis (2005) report larger NH3 losses from housing and manure storage than from application. Consequently, efficiency of N cycling can be mainly improved by technical solutions in reducing NH3 losses in storing and applying manures (Anger and Kühbauch, 1999). In grazed systems N leaching losses, mainly from high N inputs at urine spots, are almost inevitable. The extent of N leaching is related to the number and timing of expected urine spots, feeding of concentrates, and N input from mineral fertilizers and manure (White et al., 2001; Pleasants et al., 2007; Moir et al., 2011). Apart from plant uptake and N leaching, NH3 losses, denitrification and microbial immobilization are other sources or sinks for excreted N. The later in the grazing season, the higher the stocking rate and the additional mineral and organic fertilization with nitrogen, the higher the general total load of the system with N leaching (Benke, 1992; Wachendorf et al., 2006; Kayser et al., 2008). Gaseous losses in the form of N2O are usually higher under grazing than in cut grassland if conditions are the same otherwise (Oenema et al., 1997). Generally, and different than in cut systems, reduction in N losses in grazing systems can be achieved more by adapting and adjusting the management. By introducing a more reduced and temporally adjusted N-fertilization and a combination of cutting and grazing - temporal succession and changes in utilization - both, nitrate leaching and the formation of N2O might be reduced (Anger, 2001). In conclusion, at the farm scale nutrient efficiency is thus depending on how well the cycling of nutrients within the soil-plant-ruminant system is organized. Losses reduce the cost-effectiveness (profitability) of production, and, as emissions, contribute to environmental stress. The extent to what losses occur is to a great deal depending on management factors. Meadow birds and grassland management Lowland fens are important for biodiversity and for providing a range of other ecosystem services among which are landscape hydrology and carbon mitigation (Kratz and Pfadenhauer, 2001; Bragg and Lindsay, 2003). As a result of the intensification of agriculture pristine lowland fens are drastically reduced worldwide (Parish et al., 2008). Open landscape lowland grasslands are especially important as breeding habitats for meadow birds. Besides preservation of organic carbon in soil, landscape hydrology and aesthetics, reestablishment of specific plant communities providing a habitat for meadow-breeding birds is often a priority target of restoration and conservation of fen areas. Restoration measures commonly include rewetting, drastic restrictions on fertilizer use or typically a call for complete de-eutrophication and use as extensive grasslands and a later first cutting date or start of grazing (Klimkowska et al., 2010; Müller et al., 2010). Given this situation, it is obvious that the production function of grasslands suffer. Yields successively decrease with time, more so will the feeding value of the forage. Stocking rates and defoliation frequencies are low, the grazing period is limited due Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 208 to wet soils. Livestock performance from wet grasslands is therefore low and there is a definite priority of the environmental services. However, also within this production system there is an incentive to reduce the tradeoff between the production and the environmental function. If the grasslands would be abandoned from agricultural use, the other services would be at risk. Meadow-breeding birds require a habitat quality with a heterogeneous vegetation structure composed of sufficiently high proportions of open, short-grass areas in combination with patches of taller vegetation (Fondell and Ball, 2004; Verhulst, 2011). Grazing intensity determines defoliation intensity, soil compaction, local nutrient returns in the form of excreta, and avoidance, and is thus a key management variable influencing the structure and composition of pastures, i.e. patch type, scale and stability (Pavlu et al., 2006; Dumont et al., 2011). At a restricted stocking density usually undergrazing and selective grazing occur (Adler et al., 2001) creating patches of different sward height at different spatial scales and with a comparatively high stability (Rook et al., 2004, Tallowin et al., 2005). While providing an adequate habitat for meadow birds, enhancing biodiversity and ensuring soil protection agricultural use with heifers, oxen or suckler cows can only be economically viable if costs for maintenance are kept to a minimum (Mills et al., 2007). Due to selective grazing, animals might select diets of a better quality than the mean of the herbage on offer (Phillips, 2002; Rook et al., 2004). Therefore, with reduced stocking, even less-productive grassland might be used for efficient livestock farming (Isselstein et al., 2007). In investigations on extensive grazing with oxen on fen grassland in northwest Germany, Benke and Isselstein (2001) found relatively high individual daily live weight gains of 418–871 g head-1 with an average of 699 g head-1 during 1993–2000. The potential gross biomass growth was about 80 GJ NEL ha-1, while the net pasture performance amounted to 14.3 GJ NEL ha-1 in 1999 and 21.3 GJ NEL ha-1 in 2000. Thus, the grass leavings of about 80% in 1999 and 73% in 2000 were very high (Wrage et al., 2011). However, breeding success of meadow birds is also directly related to stocking rate. Müller et al. (2009) found a positive linear relationship between stocking rate and number of nests lost by trampling. True nest survival was reduced from 60% at a stocking rate of 1.5-3 LU ha-1 to 20% at 4.5 LU ha-1. Scale issues, e.g. the size of the field and the resulting influence of animal behaviour on habitat pattern, seemed to play a role, too. They conclude that up to a density of 2 LU ha-1 trampling losses of nests are still moderate and can be tolerated because of the positive effects of grazing on sward structure and nesting habitat (Müller et al., 2009). We might summarize as follows: the intention to protect meadow birds and in turn preserve and provide a suitable habitat for breeding and rearing can be seen as the main driver, and target ecosystem service as well, that determines grassland management in designated areas. As explained above this management entails extensive grazing and generally facilitates other connected ecosystem services as soil protection, biodiversity, recreation etc. Limited agricultural use (production of goods as an ecosystem service), reduced biomass and greatly varying forage quality can be identified as the main tradeoffs. Tradeoffs at the farm level might occur locally distant when other fields have to be farmed more intensively, with possible risks for groundwater and air pollution, to make up for the restricted use in the fen area. In time, other implications directly related to the restrictions of grassland management and nutrient input are imaginable which might be adversely affecting the botanical composition (Janssens et al., 1998) and forage quality for grazers but also the abundance of soil fauna as a feed source for meadow birds (Altenburg and Wymenga, 1998; Müller et al., 2010). For example, even if managed extensively, K leaching from fen grassland can be higher than input with rain, groundand surface water (Koppisch et al., 2001). Finally, a nutrient imbalance occurs between N, P and K, as N is still supplied by the mineralization of peat, while K becomes deficient (Pfadenhauer et al., 2001; Kayser and Isselstein, 2005). Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 209 Outlook: future developments and research needs The awareness that grasslands are multi-functional and provide a range of highly valued services in addition to the production function has been steadily increasing over the last two decades. European agricultural policy has responded to this perception and is placing more emphasis on the environmental functions of grasslands. The concept of ecosystem services could be used as a framework in which well-justified solutions for a grassland management that support a range of functions are developed. This is becoming even more important with the recently claimed requirement for ‘sustainable intensification’ which appears as the new paradigm of agriculture. While the concept of ecosystem services provides a valuable guiding principle and helps to better understand the tradeoffs among different services, its practicability is also limited. A major reason for this limitation is the high complexity of interacting processes that determine functions and services which make the concept difficult to apply. A common experience in this respect is that a deliberate change of one function of the system results in unexpected disfunctions and disservices in other parts. To approach such problems more research and development is required on the nature of relationships among different functions employing more complex experimental designs. Irrespective of the research requirements the concept of ecosystem services is already supporting the societal discourse on where to go and as a result of this, the intentions and priorities of stakeholders are becoming obvious. As an example, within the EU Dairyman project farmers had been offered a selection of management options to improve their farming situation in terms of several sustainability measures (Aarts et al., 2013). The response of farmers was documented. It could be shown that the farmers differ widely in their preferences for actions. These differences occurred mainly among different regions in Northwest Europe reflecting the wide variation in the physical and socio-economic conditions in the different regions. Likewise, if other stakeholders were asked to express their preferences, a broad variation would have appeared. As has been shown above, the clarity on intentions and preferences of stakeholders is a prerequisite to balance tradeoffs. References Aarts H.F.M., Biewinga E.E. and van Keulen H. (1992) Dairy farming systems based on efficient nutrient management. Netherlands Journal of Agricultural Science 40, 285-299. Aarts F., Grignard A., Boonen J., de Haan M., Hennart S., Oenema J., Lorinquer E., Sylvain F., Herrmann K., Elsaesser M., Castellan E. and Kohnen H. (2013) A practical manual to assess and improve farm performances. Report from the EU Dairyman project, pp.97. Adler P.B., Raff D.A. and Lauenroth W.K. (2001) The effect of grazing on the spatial heterogeneity of vegetation. Oecologia 128, 465-479. Altenburg W. and Wymenga E. (1998) Onderzoek naar de achteruitgang van Weidevogels in het Natuurreservaat De Gouden Bodem. A&W-rapport 170, Veenwouden. Andrews M., Scholefield D., Abberton M.T., McKenzie B.A., Hodge S. and Raven J.A. (2007) Use of white clover as an alternative to nitrogen fertilizer for dairy pastures in nitrate vulnerable farms in the UK. Annals of Applied Biology 151, 11-23. Anger M. (2001) Kalkulation der umweltbelastenden Stickstoffverluste auf Dauergrünland und Bewertung des nachhaltigen Stickstoffeinsatzes im Grünlandbetrieb. Habil.-Schrift Univ. Bonn. Anger M. and Kühbauch W. (1999) Modellrechnungen zur Senkung von Stickstoffverlusten und Nährstoffüberschüssen im Milchviehbetrieb auf Grünland. Agribiological Research 52, 85–96. Assaf T., Beyschlag W. and Isselstein J. (2011) The relationship between plant diversity and productivity in natural and in managed grasslands. Applied Ecology and Environmental Research 9, 157-166. Benke M. and Isselstein J. (2001) Extensive Landwirtschaft auf Niedermoorgrünland - Probleme und Chancen. In: Kratz R. and Pfadenhauer J. (eds) Ökosystemmanagement für Niedermoore, Strategien und Verfahren zur Renaturierung. Ulmer, Stuttgart Benke M. (1992) Untersuchungen zur Nitratauswaschung unter Grünland mittels der Saugkerzenmethode in Abhängigkeit von der Nutzungsart (Schnitt/Weide), der Nutzungshäufigkeit, der Bestandeszusammensetzung (mit/ohne Weißklee) und der Stickstoffdüngung, Kiel. Bragg O. and Lindsay R. (eds.) (2003) Strategy and action plan for mire conservation in Central Europe. Wetlands International Publication 18. Wetlands International, Wageningen, NL. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 210 Chapin III F.S., Zavaletta E.S., Eviner V.T. et al. (2000) Consequences of changing biodiversity. Nature 405, 234–242. Conant R.T. (2010) Challenges and opportunities for carbon sequestration in grassland systems. A technical report on grassland management and climate change mitigation. Integrated Crop Management 9, 2010, FAO, Rome, pp.59. Costanza R., dArge R., de Groot R., Farber S., Grasso M., Hannon B., Limburg K., Naeem S., O’Neil R.V., Paruelo J., Raskin R.G., Sutton P. and van den Belt M. (1997) The value of the world’s ecosystem services and natural capital. Nature 387, 253–260. Cuttle S.P. and Jarvis S.C. (2005) Use of a systems synthesis approach to model nitrogen losses from dairy farms in south-west England. Grass and Forage Science 60, 261–273. Daily G.C. (2000) Management objectives for the protection of ecosystem services. Environmental Science and Policy 3, 333–339. Daily G.C. (Ed.) (1997) Nature’s Services: Societal Dependence on Natural Ecosystems. Island Press, Washington, DC. Daily G.C., Söderqvist T., Aniyar S. et. al. (2000) The value of nature and the nature of value. Science 289, 395– 396. De Groot R.S. (1987) Environmental functions as a unifying concept for ecology and economics. The Environmentalist 7 (2), 105–109. Del Prado A., Mas K., Pardo G. and Gallejones P. (2013) Modelling the interactions between C and N farm balances and GHG emissions from confinement dairy farms in northern Spain. Science of the Total Environment 465, 156-165. de Wit C.T. (1958) Transpiration and crop yields. Verslagen van landbouwkundige onderzoekingen 59, pp. 88. DeFries R.S., Foley J.A., Asner G.P. (2004) Land-use choices: balancing human needs and ecosystem function Frontiers in Ecology and the Environment 2, 249-257. Dias A.T.C., Berg M.P., de Bello F., Van Oosten A.R., Bila K. and Moretti M. (2013) An experimental framework to identify community functional components driving ecosystem processes and services delivery. Journal of Ecology 101, 29-37. Dumont D., Carrère P., Ginane C., Farruggia A., Lanore L., Tardif A., Decuq F., Darsonville O. and Louault F. (2011) Plant–herbivore interactions affect the initial direction of community changes in an ecosystem manipulation experiment. Basic and Applied Ecology 12, 187–194. EEA - European Environment Agency (2004) High nature value farmland - Characteristics, trends and policy challenges. EEA report No 1/2004, pp.27. EEA - European Environment Agency (2010) EU 2010 biodiversity baseline. EEA Technical report No 12/2010, pp.124. EEA - European Environment Agency (2013) The European grassland butterfly indicator: 1990–2011. EEA Technical report No 11, pp.36. Ehlers W. and Goss M. (2003) Water dynamics in plant production. Wallingford, UK: CABI Publishing, pp.277. Ellenberg H. and Leuschner C (2010) Vegetation Mitteleuropas mit den Alpen, 6. Auflage, Ulmer, UTB Eriksen J., Ledgard S., Luo J., Schils R. and Rasmussen J. (2010) Environmental impacts of grazed pastures. Grassland Science in Europe 15, 880-890. EUROSTAT (2013) Agricultural statistics. http://epp.eurostat.ec.europa.eu/portal/page/portal/agriculture/introduction Flessa H., Müller D., Plassmann K., Osterburg B., Techen A.-K., Nitsch H., Nieberg H., Sanders J., Meyer zu Hartlage O., Beckmann E. and Anspach V. (2012) Studie zur Vorbereitung einer effizienten und gut abgestimmten Klimaschutzpolitik für den Agrarsektor. Landbauforschung, Sonderheft 361, pp.437. Fondell T.F. and Ball I.J. (2004) Density and success of bird nests relative to grazing on western Montana grasslands. Biological Conservation 117, 203–213. Fornara D.A., Banin L. and Crawley M.J. (2013) Multi-nutrient vs. nitrogen-only effects on carbon sequestration in grassland soils. Global Change Biology 19, 3848-3857. Gibon A. (2005) Managing grassland for production, the environment and the landscape. Challenges at the farm and the landscape level. Livestock Production Science 96, 11–31. Grime J.P. (1997) Biodiversity and Ecosystem Function: The Debate Deepens. Science 277, 1260–1261. Hector A., Schmid B., Beierkuhnlein B., Caldeira M.C., Diemer M., Dimitrakopoulos P.G., Finn J.A., Freitas H., Giller P.S., Good J., Harris R., Högberg P., Huss-Danell K., Joshi J., Jumpponen A., Körner C., Leadley P.W., Loreau M., Minns A., Mulder C.P.H., O'Donovan G., Otway S.J., Pereira J.S., Prinz A., Read D.J., SchererLorenzen M., Schulze E.-D., Siamantziouras A.-S.D., Spehn E.M., Terry A.C., Troumbis A.Y., Woodward F.I., Yachi S. and Lawton H.J. (1999) Plant diversity and productivity experiments in European grasslands. Science 286, 1123-1127. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 211 Hönigová I., Vačkář D., Lorencová E., Melichar J., Götzl M., Sonderegger G., Oušková V., Hošek M. and Chobot K. (2012) Report to the EEA – European Topic Centre on Biological Diversity. Prague: Nature Conservation Agency of the Czech Republic, pp 78. Hopkins A. and Wilkins R.J. (2006) Temperate grassland: key developments in the last century and future perspectives. Journal of Agricultural Science 144, 503–523. Husemann C. and Wesche J. (1964) Wasserhaushalt und Wasserbilanz eines lehmigen Sandbodens nach Untersuchungen in wägbaren Lysimetern. Zeitschrift für Kulturtechnik und Flurbereinigung 5, 364-389 Isselstein J., Griffith B.A., Pradel P. and Venerus S. (2007) Effects of livestock breed and grazing intensity on biodiversity and production in grazing systems. 1. Nutritive value of herbage and livestock Performance. Grass and Forage Science 62, 145–158. Isselstein J., Jeangros B. and Pavlu V. (2005) Agronomic aspects of biodiversity targeted management of temperate grasslands in Europe – A review. Agronomy Research 3, 139–151. Janssens F., Peeters A., Tallowin J.R.B., Bakker J.B., Bekker R.M., Fillat F. and Oomes M.J.M. (1998) Relationship between soil chemical factors and grassland diversity. Plant and Soil 202, 69-78. Jerrentrup S., Wrage-Mönnig N., Röver K.U. and Isselstein J. (2014) Grazing intensity affects insect diversity via sward structure and heterogeneity in a long-term experiment. Journal of Applied Ecology DOI: 10.1111/13652664.12244. Kaiser J. (2000) Rift over biodiversity divides ecologists. Science 289, 1282–1283. Kayser M. and Isselstein J. (2005) Potassium cycling and losses in grassland systems: a review. Grass and Forage Science 60, 213–224. Kayser M., Seidel K., Müller J. and Isselstein J. (2008) The effect of succeeding crop and level of N fertilization on N leaching after break-up of grassland. European Journal of Agronomy 29, 200–207. Kesting S., Wrage N. and Isselstein J. (2009) Herbage mass and nutritive value of herbage of extensively managed temperate grasslands along a gradient of shrub encroachment. Grass and Forage Science 64, 246-254. Klimkowska A., VanDiggelen R., Grootjans A.P. and Kotowski W. (2010) Prospects for fen meadow restoration on severely degraded fens. Perspectives in Plant Ecology, Evolution and Systematics 12, 245–255. Koppissch D., Roth S., Knapp M., Blankenburg J. and Eschner D. (2001) The effect of rewetting and agricultural management on the content of plant available nutrients in soil (in German). In: Kratz R. and Pfadenhauer J. (eds.) Ökosystemmanagement für Niedermoore. Verlag E. Ulmer, Stuttgart, pp. 92-104. Kratz R. and Pfadenhauer J. (2001) Ökosystemmanagement für Niedermoore, Verlag Eugen Ulmer, Stuttgart, Germany. Lindemann-Matthies P., Junge X. and Matthies D. (2010) The influence of plant diversity on people’s perception and aesthetic appreciation of grassland vegetation. Biological Conservation 143, 195–202. Lüscher A., Finn J.A., Connolly J., Sebastià M.T., Collins R., Fothergille M., Porqueddu C., Brophy C., HugueninElie O., Kirwan L., Nyfeler D. and Helgadóttir A. (2008) Benefits of sward diversity for agricultural grasslands. Biodiversity 9, 29-32. McCann K.S. (2000) The diversity – stability debate. Nature 405, 228–233. Millenium Ecosystem Assessment (2005) Ecosystems and human well-being. Synthesis. Washington DC: Island Press. Mills J., Rook A.J., Dumont B., Isselstein J., Scimone M. and Wallis De Vries M.F. (2007) Effect of livestock breed and grazing intensity on grazing systems: 5. Management and policy implications. Grass and Forage Science 62, 429–436. Moir J.L., Cameron K.C., Di H.J. and Fertsak U. (2011) The spatial coverage of dairy cattle urine patches in an intensively grazed pasture system. Journal of Agricultural Science 149, 473–485. Müller J., Kayser M. and Belting H. (2010) Concepts for nutrient management in nature conservation areas on organic soils. Grassland Science in Europe 15, 1075–1077. Müller J. and Kayser M. (2009) Transformation of arable land into wet grassland on a degraded low peat soil. Grassland Science in Europe 14, 535–538. Nyfeler D., Huguenin-Elie O., Suter M., Frossard E., Connolly J. and Lüscher A. (2009) Strong mixture effects among four species in fertilized agricultural grassland led to persistent and consistent transgressive overyielding. Journal of Applied Ecology 46, 683–691. Oenema O., Velthof G.L., Yamulki S. and Jarvis S.C. (1997) Nitrous oxide emissions from grazed grassland. Soil Use and Management 13, 288–295. Oomes M.J.M. (1992) Yield and species density of grasslands during restoration management. Journal of Vegetation Science 3, 271–274. Osterburg B., Laggner B., Nitsch H., Roggensorf W. and Röder N (2011) Analysis of grassland conversion into arable land in Northwest Germany. Grassland Science in Europe 16, 350-352 Parish F., Sirin A., Charman D., Joosten H., Minayeva T., Silvius M. and Stringer L. (Eds.) (2008) Assessment on Peatlands, Biodiversity and Climate Change: Main Report. Global Environment Centre, Kuala Lumpur and Wetlands International, Wageningen. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 212 Pavlu V., Hejcman M., Pavlu L., Gaisler J. and Nezerkova P. (2006) Effect of continuous grazing on forage quality, quantity and animal performance. Agriculture, Ecosystems and Environment 113, 349-355. Peeters A., Beaufoy G., Canals R.M., De Vliegher A., Huyghe Ch., Isselstein J., Jones G., Kessler W., Kirilov A., Mosquera-Losada M.R., Nilsdotter-Linde N., Parente G., Peyraud J.-L., Pickert J., Plantureux S., Porqueddu C., Rataj D., Stypinski P., Tonn B., van den Pol-van Dasselaar A., Vintu V. and Wilkins R. (2014) Grassland term definitions and classifications adapted to the diversity of European grassland-based systems. Grassland Science in Europe 19, this volume. Pfadenhauer J., Höper H., Borkowsky O., Roth S., Seeger T. and Wagner C. (2001) Entwicklung pflanzenartenreichen Niedermoorgrünlands. In: Kratz R. and Pfadenhauer J. (eds) Ökosystemmanagement für Niedermoore. Verlag E. Ulmer, Stuttgart, pp. 134–153. Phillips C. (2002) Nutritional behaviour. In: Phillips C. (ed) Cattle Behaviour and Welfare. Blackwell, Oxford, UK, pp. 123–151. Pleasants A.B., Shoten P.R. and Wake G.C. (2007) The distribution of urine deposited on a pasture from grazing animals. Journal of Agricultural Science 145, 81–86. Power A.G. (2010) Ecosystem services and agriculture: tradeoffs and synergies. Philosophical Transactions of the Royal Scociety, B 365, 2959–2971. Quetier F., Rivoal F., Marty P., de Chazal J., Thuiller W. and Lavorel S. (2010) Social representations of an alpine grassland landscape and socio-political discourses on rural development. Regional Environmental Change 10, 119–130. Rodríguez J.P., Beard Jr. T.D., Bennett E.M., Cumming G.S., Cork S.J., Agard J., Dobson A.P. and Peterson G.D. (2006) Trade-offs across space, time, and ecosystem services. Ecology and Society 11, 28 Rook A.J., Dumont B., Isselstein J., Osoro K., WallisDeVries M.F., Parente G. and Mills J. (2004) Matching type of livestock to desired biodiversity outcomes in pastures – a review. Biological Conservation 119, 137–149. Rose L., Coners H. and Leuschner C. (2011) Effects of fertilization and cutting frequency on the water balance of a temperate grassland. Ecohydrology 5, 64–72. Rose L. and Leuschner C. (2012) The diversity-productivity relationship in a permanent grasslands: negative diversity effect, dominant influence of management regime. Plant Ecology and Diversity 5, 265-274. Sanderson M.A. and Wätzold F. (2010) Balancing trade-offs in ecosystem functions and services in grassland management. Grassland Science in Europe 15, 639–648. Schneider M.K., Homburger H. and Lüscher A. (2011) Survey of the biodiversity-productivity relationship in Swiss summer pastures. Grassland Science in Europe 16, 487-489 Seither M., Wrage N. and Isselstein J. (2010) Sward composition and grazer species effects on nutritive value and herbage accumulation. Agronomy Journal 104, 497–506. Smit, H.J., Metzger, M.J. and Ewert, F., 2008: Spatial distribution of grassland productivity and land use in Europe. Agricultural Systems 98, 208-219. Tallowin, J.R.B., Rook A.J. and Rutter S.M. (2005) Impact of grazing management on biodiversity of grasslands. Animal Science 81, 193–198. Taube F., Gierus M., Hermann A., Loges R. and Schönbach P. (2014) Grassland and globalization – challenges for northwest European grass and forage research. Grass and Forage Science 69, 2–16. Tilman D. (1999) The ecological consequences of changes in biodiversity: a search for general principles. Ecology 80, 1455–1474. Tilman D. (2000) Causes, consequences and ethics of biodiversity. Nature 405, 208–211. Tilman D., Reich P.B., Knops J., Wedin D., Mielke T. and Lehman C. (2001) Diversity and productivity in a longterm grassland experiment. Science 294, 843-845. Van den Pol-van Dasselaar A., Vellinga T.V., Johansen A. and Kennedy E. (2008) To graze or not to graze, that’s the question. Grassland Science in Europe 13, 706–716. van der Ploeg R.R., Ehlers W. and Sieker F. (1999) Floods and other possible adverse environmental effects of meadowland area decline in former West Germany. Naturwissenschaften 86, 313-319 Verhulst J., Kleijn D., Loonen W., Berendse F. and Smit C. (2011) Seasonal distribution of meadow birds in relation to in-field heterogeneity and management. Agriculture, Ecosystems and Environment 142, 161–166. Wachendorf M. and Soussana J.F. (2012) Perspectives of energy production from grassland biomass for atmospheric greenhouse gas mitigation: Grassland Science in Europe 17, 425-435. Wachendorf M., Büchter M., Volkers K.C., Bobe J., Rave G., Loges R. and Taube F. (2006) Performance and environmental effects of forage production on sandy soils. V. Impact of grass understorey, slurry application and mineral N fertilizer on nitrate leaching under maize for silage. Grass Forage Science 61, 243–252. Weigelt A., Weisser W.W., Buchmann N. and Scherer-Lorenzen M. (2009) Biodiversity for multifunctional grasslands: equal productivity in high-diversity low-input and low-diversity high-input systems. Biogeosciences, 6, 1695–1706. White S.L., Sheffield R.E., Washburn S.P., King L.D. and Green J.T. (2001) Spatial and time distribution of dairy cattle excreta in an intensive pasture system. Journal of Environmental Quality 30, 2180–2187. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 213 Whitehead D.C. (1995) Grassland Nitrogen. CAB International, Wallingford, UK. Wrage N., Strodthoff J., Cuchillo M.H., Isselstein J. and Kayser, M. (2011) Phytodiversity of temperate permanent grasslands: Ecosystem services for agriculture and livestock management for diversity conservation. Biodiversity and Conservation 20, 3317–3339. Yan M.J., Humphreys J. and Holden N.M. (2013a) Life cycle assessment of milk production from commercial dairy farms: The influence of management tactics. Journal of Dairy Science 96, 4112-4124. Yan M.J., Humphreys J. and Holden N.M. (2013b) The carbon footprint of pasture-based milk production: Can white clover make a difference? Journal of Dairy Science 96, 857-865. World Bank (2013) Data, indicators, average precipitation in depth. http://data.worldbank.org/indicator/AG.LND.PRCP.MM. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 214 Comparing synthetic and natural grasslands for agricultural production and ecosystem service Humphreys M.W.1, O’Donovan G.2 and Sheehy-Skeffington M.3 1 Institute of Biological, Environmental and Rural Sciences, Aberystwyth University, Gogerddan, Aberystwyth, Ceredigion, Wales, SY23 3EE, UK 2 Department of Biological, Biomedical and Analytical Sciences, University of the West of England, Frenchay Campus, Coldharbour Lane, Bristol, BS16 1QY, UK 3 Plant Ecology Research Unit, Botany and Plant Science, National University of Ireland, Galway, University Road, Galway, Ireland Corresponding author: mkh@aber.ac.uk Abstract Whilst the concept of ecosystem service is relatively new, the importance and benefits of natural grasslands to the environment has been long established. These complex bio-diverse ecosystems, as well as sustaining rich communities of flora and fauna, provide a range of environmental benefits including water, nutrient, and carbon capture. However, the perpetuity of natural grasslands and their associated benefits are under increased threat from pressures to feed and house our increasing population, urban expansion, and also through climate change. Regular ploughing and resowing of grasslands has led to soil erosion, depletion of scarce nutrient resources, pollution of our waterways, and releases of harmful greenhouse gases, the latter in particular exacerbated by livestock agriculture. A response is necessary both to reduce negative impacts of agriculture on the environment and, wherever possible, to engineer a positive ecosystem service. The genomic and phenomic diversity available in grass and clover species, and further access to novel variants through hybridization with wild-type relatives with suitable technologies to assist in their selection, provide alternatives to current plant varieties and increased capacity for efficient and ‛climate-smart’ agricultural practice. Holistic approaches to plant breeding can produce varieties that both safeguard agricultural production and provide some valuable ecosystem service. Keywords: Festulolium, clover, semi-natural grasslands, grassland diversity, ecosystem services Introduction Among diverse grassland ecosystems, it is necessary to distinguish climatically determined grasslands, where water availability is insufficient for development of forest ecosystems and where natural vegetation remains in dynamic equilibrium with herbivores (Lauenroth, 1979), from the anthropogenically generated grasslands, located mainly within temperate climate regions, where woody vegetation is excluded and herbaceous plant communities maintained by appropriate human intervention and by livestock agriculture. It is possible to further divide the latter grassland type into long-term naturalized grasslands and those that are cultivated, which differ according to level of intensification. However, all grasslands are multifunctional and to different degrees are capable of playing important roles in agronomic, economic and social activities. Whilst an increase in grassland production in terms of its provision as safe, healthy and economically sustainable fodder for livestock consumption was viewed as an important priority for national security, and still remains a priority, it is also important to acknowledge the important role of grassland ecosystems in all our lives, from the air we breathe to the water we drink, and wherever possible, manage grassland ecosystems to maximize their potential to deliver environmental benefits. The concept of 'ecosystem services' was outlined within the Millennium Ecosystem Assessment (2005) (http://www.maweb.org/en/index.aspx). A recent review of 17 grassland biodiversity experiments revealed that 84% of 147 grassland Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 215 species displayed ecosystem functions, as measured, at least once (Isbell et al., 2011), and while species may appear redundant under one circumstance, they may otherwise function fully in another, leading to the use of species mixtures to help ensure the expression of multiple ecosystem services, particularly when challenged by changing environmental conditions. Similarly for grassland agricultural production, it is normal practice both in the UK and in Ireland to re-sow grasslands with variety mixtures to encourage resilience to predictable stresses, long growing seasons and ultimately sustainable yields. Whilst the ‘market value’ of grasslands to agriculture is obvious through the production of meat, wool and milk from grazing animals, they are being increasingly recognized at the global scale for their non-market contribution to carbon sequestration, prevention of soil erosion and for genetic conservation. Other more subtle ecosystem-supporting services include pollination processes, improved soil structure, decreased nutrient leaching and nitrogen fixation. Ecosystem services may be produced in situ, e.g. through biomass production and carbon sequestration, or adjacent to other habitats and organisms, e.g. preventing leaching of nutrients downhill into rivers, streams or wetlands, or a source of flowers for pollinators from other locations. Many grasslands are ‛highly improved’ and as a consequence have significantly reduced ecosystem services compared to semi-natural or natural grassland ecosystems. The plight of semi-natural and natural grasslands is well known, with a loss of at least 98% of our wild flower meadows in Britain over the last 50 years (Peterken, 2013). Their fragmentation and degradation have reduced their potential for ecosystem services significantly. For grasslands, as with forest ecosystems, plant-soil interactions involve different and complex cycling elements, particularly of carbon (C), nitrogen (N), and phosphorus (P). These operate (i) in plants, where N, P and C are combined in organic matter synthesis, accumulation and long-term sequestration in soils; and (ii) in soils, where microbes feed abundantly on C and recapture and recycle mineral N and P. In most circumstances, grasslands can sequester C, N and P for relatively long-term periods, so contributing to the atmospheric CO2 sink, and reducing release of N compounds into water and atmosphere and their associated environmental risks. Grasslands are land-use systems that can be very favourable for environmental preservation. However, this idealistic view has to be tempered, as livestock decouple C and N–P cycles, through their urine and faeces depositions and their methane (CH4) emissions, offsetting, sometimes to a significant extent, the beneficial effect of grassland vegetation–soil interactions. Alternative viewpoints persist between ecologists and crop geneticists as to what extent it is possible to reduce the detrimental impacts of grassland agriculture on the environment by combined use of best farming practice and the inclusion of new and more sustainable grassland crop varieties. Entrenched viewpoints are sustained by some crop geneticists’ naïve understanding of ecosystem complexity and certain field ecologists’ ignorance of the potential of new plant breeding technologies, as well as by others whose priorities, for whatever reason, may differ significantly. Whilst risking being considered naïve, an argument is presented herein where a ‛win-win’ scenario might be achieved, but, as authors with different viewpoints ourselves, we counter-balance the argument by specific reference to the greater service diversity obtained from semi-natural grasslands, both by their species diversity and the far more stable soil system they support. Through the incorporation of new and appropriate technologies and germplasm, it is proposed on the one hand that there are significant opportunities available to ‛untap’ potential for ecosystem service from novel grasslands, whilst also providing productive crops considered fit for agricultural purpose. However, before these are outlined it should be clarified that the authors consider any new option for grassland agriculture that might safeguard the needs of both agriculture and the environment, will not in any way displace current priorities to Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 216 safeguard existing semi-natural grassland ecosystems, whose complex and diverse benefits are not likely to be easily reproduced. Semi-natural grasslands and their ecosystem services The value of semi-natural grasslands for ecosystem service should never be underestimated. To paraphrase the British Ecological Society Bulletin Bogbean Bennet Cartoon-strip (http://homepages.abdn.ac.uk/g.j.pierce/pages/bbb.htm), any attempt to compensate for the destruction of a habitat by reproducing something similar elsewhere is like 'moving the Mona Lisa' by scraping away all the paint on the canvas and reconstructing it elsewhere: the paint may be the same, but the interconnectivity of the paint flakes is totally lost. Ecosystems are highly complex. For the most part we study only certain aspects such as carbon sequestration, productivity, or species composition; be it birds, mammals, invertebrates or plants. The soil may be studied for respiration (productivity), arthropods, or fungi, but the full ecosystem in toto is virtually impossible to describe and quantify. So the ability to recreate it is not, in reality, a feasible option. In Ireland, as in Britain, grasslands are a derived habitat, maintained by farming practice, either by grazing or cutting. The last 50 years or more have seen widespread intensification, largely driven by EU policies and incentives (Hickie et al., 1999). Grasslands were traditionally managed by low-intensity farming and semi-natural grasslands differed in species composition as a consequence of, e.g., soil pH or water content (O’Sullivan, 1982). The EU drive for productivity beginning in the 1970s led to the ploughing up of grasslands for rotation crops and reseeding for silage, with a concomitant reliance on chemicals, including fertilizer application. These changes resulted in the reduction of semi-natural grassland and therefore plant species diversity in a grassland landscape. In particular, the traditional extensive practice of cutting meadows for hay has declined substantially in recent decades (Peterken, 2013). Semi-natural grasslands have thus imperceptibly become very rare in Ireland, Britain and mainland Europe (Fuller, 1987; Baldock, 1989; Feehan, 2003; Sullivan et al., 2010). Their conservation is now considered a priority and forms part of EU-, as well as national policy; the EU Habitats Directive lists 31 semi-natural grassland habitats in Europe, and national programmes are focused on conserving the range of these habitats that still remain (Critchley et al., 2003; EEA, 2004; Stevens et al., 2010; O’Neill et al., 2013). Aside from the ethics of conserving increasingly rare species-rich habitats, it is now seen as important to quantify their contribution to the environment as ecosystem services. This relatively recent concept relates to the current need to put an economic value on our environment, since, if it affects us economically, it may induce governments to take action. The drive to be productive and increase profit margins is influenced in this way, but the economic losses resulting from agricultural intensification also require evaluation (Purvis et al., 2008; Dreschler et al., 2010). Individual plant traits in grassland studies have come to the fore in relation to ecosystem services from semi-natural and natural grasslands in Europe. A Europe-wide study named VESTA (Garnier et al., 2007) involving 11 European sites illustrates that experiments that can isolate direct effects of climate and land use from indirect effects, such as changes in community functional composition, advance our understanding of the role of plant traits as linkages between environmental change and ecosystem properties. Inclusion of environmental variables such as climate, soil, and disturbance in quantitative analyses makes it possible to test hypotheses about the pathways that determine ecosystem properties through modification of plant traits in communities. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 217 Value and productivity of species-rich swards Arguments for intensive highly-productive agriculture often overlook the fact that grasslands rich in biodiversity have been shown to be more productive than species-poor swards (Tilman et al., 1996; 1997), especially grasslands with high forb diversity and resultant complementarities of resource use (Hooper et al., 2005; Bullock et al., 2007). The number of legumes alone does not account for a yield increase; instead the wide range of temporal and spatial growth patterns exhibited by different forbs maximizes the use of resources and can increase fodder value (Hofmann and Isselstein, 2005). Forb diversity is a key factor in the provision of nutritious hay, particularly for horses that do not eat silage (Allison and Day, 1999). The high quality of the species-rich grasslands of the Burren is widely known and farmers from the Irish midlands move animals onto the High Burren to increase bone and muscle quality before fattening on richer soils in the spring (Dunford, 2002; Williams et al., 2009). The experimental addition of species number on prairie grassland experiencing a fluctuating climate created greater below-ground biomass and temporal stability that, in turn, enhanced ecosystem services such as fodder provision and biofuel production (Tilman et al., 2006a). Over ten years the experimental manipulation of native grassland perennial mixtures (Low Input High Diversity Grasslands, LIHD), of varied species number, demonstrated that these required much less pesticide and fertilizer applications than conventional monoculture crops such as corn and soybean, so were less costly to produce in terms of fossil energy inputs (Tilman et al., 2006b). Carbon sequestration in the grassland soils was very high, whereas that of the annual crops was negligible, and would remain so over time. The more diverse the grasslands, the more biomass they produced. As a result, high-diversity grasslands had increasingly higher bioenergy yields that were 238% greater than monoculture yields after a decade (Tilman et al., 2006b). The LIHDs were also more beneficial in that they could be grown on degraded soils where no monocultures could easily be grown. Using an agro-ecosystem approach, the full benefits of a balance between crops, pests and their predators can be evaluated (Altieri, 2008), in contrast to the relative efficiency of a more industrialized approach to agriculture, where the real costs of fuels and chemical application need to be factored in, as well as the relative pest vulnerability of monotypic crops, including intensively farmed grasslands. The advantages to the environment of organic farming must not be discounted (Niggli et al., 2007), especially as organic farming yields are often no less than those of non-organic farming, contrary to widely-held belief (Badgley et al., 2007). On the other hand, in relation to ecosystem services it is important to address the range of these that biodiversity can offer, including habitat, as well as species (flora and fauna) diversity (Council of the European Union, 2010). Pollinator services Forbs in a grassland sward are important, as these are predominantly insect-pollinated species (as opposed to wind-pollinated grasses) and therefore they provide both pollen and nectar for a range of dipteran and other invertebrates (Williams,1988; Branquart and Hemptinne, 2000; Peterken, 2013). These in turn provide essential pollination for crops, notably oilseed rape, but also fruit, in particular nectar-producing trees such as apple and pear (Free, 1993). The recent decline in bees, both the honey bee due to diseases, and bumble and solitary bees, has caused much concern (Williams, 1982; Williams et al., 1991; Carvell, 2002). This has been attributed in large part to loss and fragmentation of semi-natural habitat with services not available in arable fields (Carvell et al., 2006). The reduction of a sustained source of food for these species is likely to be a factor in their decline, since once early-flowering crops such as oilseed rape finish flowering, there must be a series of other species to provide food until the late autumn (Prŷs-Jones and Corbet, 1991; Carvell, 2002). The type of plant species also affects the Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 218 pollinator species, since early-flowering annuals provide less nectar and may be more suitable to syrphids and butterflies, but bumblebees require a greater nectar supply and favour perennial forbs (Fussell and Corbet,1991a). Thus reseeding arable or resown field margins may not benefit bumblebees, despite providing nectar-producing flowers (Fussell and Corbet, 1991a). Semi-natural grasslands not only benefit nectar- and pollen-feeeders, but enable groundnesting species such as bumblebees to establish colonies (Fussell and Corbet, 1991b). Ant-hill building species such as Lasius flavus also only survive in unploughed old pastures for the same reason (King, 1977). Remote sensing to measure ecosystem services It is becoming customary and deemed essential to employ new high-throughput technologies to enable the monitoring of large land areas for ecosystem service. Habitats are mapped for their potential for ecosystem services in accordance with definitions that arose out of the Millennium Ecosystem Assessment in 2005 (http://www.maweb.org/en/index.aspx). These are classified as either supporting (primary production, nutrient cycling and soil formation), provisioning (food, fuel, water), regulating (climate, flooding, disease) or cultural (aesthetic, spiritual, educational). Grassland habitats contribute value to all of these although, all too often, the cultural contribution is overlooked. In a study of the Spanish coast, the aesthetic contribution that was shared by a range of habitats was evaluated at >25% (Brenner et al., 2010), as proof of its value. The use of satellite images to identify habitats has been in use since the advent of the LANDSAT satellite series launched in 1982 (Xie et al., 2008). As more satellites became available and resolution improved, extensive studies have been undertaken on large territories for land-use mapping, for instance, the EU Corine Land Cover initiative (http://www.eea.europa.eu/publications/COR0-landcover). In the UK, there is a bespoke landcover/land-use mapping initiative using Landsat imagery by Centre for Ecology and Hydrology (CEH). This has identified several improved, semi-natural and natural grassland categories at the national scale. This is updated regularly, most recently in 2007 (http://www.ceh.ac.uk/landcovermap2007.html). However, for UK grasslands, the overall accuracy of the mapping can be very low, as little as 34% for some categories, due to the changing temporal nature of natural and semi-natural grasslands, and the impact of management practices such as grazing, silage removal and fertilizer application. The advantage of using remote sensing to assess habitat quantity and quality is that it is repeatable over short time scales and covers large areas. China has been estimating the value of grassland ecosystem services countrywide using remote sensing (Jiang, 2007), and using non-market services such as measures for O2 released, CO2 fixed, control of soil erosion, water storage, nutrient recycling, reduction of pollution, and Net Primary Productivity. Results showed that habitats such as shrub-meadow complexes and upland grasslands provided the largest ecosystem service, while desert steppe and alpine desert provided the lowest. Mapping at the landscape scale to define indicative habitat type and area can help with conceptual models of ecosystem services such as pollination. Classification of suitable satellite imagery can define grasslands of all types in terms of area and quality within a range of certainties. From these sources, patch and population-level attributes such as floral density, patch size and patch isolation can be determined, and will influence the interactions of the target plant communities with pollinators. A recent publication has promoted the use of a model called the Mobile-Agent-Based Ecosystem Service (MABES), where ecosystem services are reliant on a mobile organism to deliver a particular service (Kremen et al., 2007). One worked example is pollination, but the model is being adapted for other tasks such as pest control and seed dispersion. Pollination is becoming an increasingly important ecosystem function due to the decline in bee populations worldwide. From a landscape perspective, a recent meta-analysis Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 219 found a significantly negative effect of habitat fragmentation on pollination of plants, and a strong correlation of this effect with reproductive success (Aguilar et al., 2006). Thus the fragmentation, and more specifically the isolation, of species-rich grasslands has wider negative effects on the ecosystem services they provide. On a landscape scale, agricultural intensification results in increased size of arable fields, decreased crop and weed diversity, and the loss and fragmentation of valuable natural to semi-natural perennial habitats such as agroforestry, grasslands and old fields. However, with more holistic grassland management at an integrated landscape-scale some more positive effects of agriculture on pollinator communities may occur. For example, in regions where the presence of low-intensity agriculture increases rather than decreases habitat heterogeneity within the foraging range of bees (e.g. <2 km), such as farmed landscapes that include relatively small field sizes, mixed crop types within or between fields, and patches of non-crop vegetation, such as hedgerows, fallow fields, meadows, and semi-natural wood or shrub-lands, it may be beneficial for biodiversity and ecosystem services (Bignal and McCracken, 1996; Sullivan et al., 2011). Increased ecosystem services by grasslands needs to be properly managed and evaluated (Kremen et al., 2007). For instance, what is required to best configure natural habitats within agricultural landscapes to promote population persistence of bees? How do the economic costs of these management practices compare to the benefits from enhanced pollination? Breeding grasslands for crop production and carbon sequestration A major aspect of ecosystem service by grassland is their role in C sequestration and in mitigating the rise in atmospheric CO2 levels. Whether or not an ecosystem accumulates or loses carbon (both above and below ground) is a function of inputs and outputs. Sequestered carbon can be defined as the difference between gross primary productivity and ecosystem respiration, which in turn is the sum of plant respiration and heterotrophic respiration of nonphotosynthetic organisms. This has been termed net ecosystem productivity (NEP) (Chapin et al., 2006). The final rate of accumulation or loss of carbon in a particular ecosystem, in addition to NEP, will depend on external deposition of C (such as inputs of organic manures and dissolved C in rain water) and also losses through erosion, removal (harvesting) and nonbiological oxidation through fire or UV radiation (Lovett et al., 2006). Grass crop breeding that accomplishes increased growth and turnover, both above and below ground, provides enhanced opportunity for C deposition. The original source of soil organic matter (SOM) is plant biomass, both above-ground, which decays in the litter layer and can be incorporated into the soil profile, and below-ground through turnover of the root system and as root exudates (Bardgett et al., 2005). Various plant traits can affect the quantity and the quality, e.g. the C:N ratio of SOM, which in turn can affect its decomposition rates and the level of carbon retained as soil organic matter (De Deyn et al., 2008). Traits that increase the rate of growth of above ground biomass, such as increased rates of photosynthesis, are often associated with a shorter lifespan and high nutrient demands, but this is not always the case. Often associated with high rates of photosynthesis and biomass production is a higher quality of the litter, which may lead to a faster turnover rate (Aerts and Chapin, 2000). For perennial crops such as grasses, their strategies for persistence and survival over years requires entry into cycles of growth in spring and early summer followed by reduced foliar and rooting growth in autumn. This may include periods of total growth cessation in preparation for winter survival. These annual growth cycles and associated growth conditions will inevitably impact on the rates of plant C deposition into soils. New strategies in Irish grassland agriculture that aim to make use of rising winter temperatures encourage whole-year growth and production. However, these have risk as severe crop damage may result from any onset of a harsh winter, and in any case will likely affect rates of accumulation of SOC. There is often trade-off between the amounts of biomass produced and its decomposition, characteristics which may affect soil organic Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 220 pools. Slower growing plants, in nutrient-poor environments, contribute to soil organic matter more on the basis of the recalcitrance of their organic matter, with high C:N ratios leading to slower rates of decomposition. In ecosystems with sufficient light and nutrients, and with high plant growth rate, the proportion of non-harvested biomass, and management regime such as cutting or grazing (where C is returned to the soil via dung) are major factors influencing levels of soil carbon. Grazing (defoliation) intensity and nutrient supply affect the balance of the major C fluxes and are best characterized in terms of the mean leaf area sustained over the growing season. ‘Losses’ of matter in respiration (approximately 40–50% of ‘gross’ photosynthesis) increase with the gross C uptake, of which 25% is associated with the energy and mass inefficiency of synthesis of new tissues, and the remaining 25% with the ‘maintenance’ of existing biomass (Thornley and Johnson, 1990). In perennial ryegrass, the longevity of leaves is approximately 30 days at 15oC (Parsons et al., 2011). Crops have been developed with different traits, and altered biochemistry from the wild type, through selective breeding programmes, hence leading to different litter qualities. Grasses with higher water soluble carbohydrate (WSC) contents, known as high sugar grasses (HSG), have been the focus of breeding programmes at IBERS (formerly IGER) to improve grassland productivity for meat and milk production (Humphreys and Theodorou, 2001). The impact of HSG and their potential for C sequestration is being investigated in collaboration with Rothamsted Research in the UK, but currently the 'jury is out'. The increased content of simple carbohydrates of HSG, as opposed to more recalcitrant compounds such as lignin, presents less of a challenge to decomposer organisms, a trait that also makes them more palatable for domesticated animals, and so will be decomposed more readily to CO2 and make a lower contribution to soil carbon than more recalcitrant tissues. Among grass species it is the root litter that is more recalcitrant and offers a greater obstacle to decomposer organisms (Craine et al., 2005). The impact upon root litter quality of selective breeding for increased WSC content of shoot tissues is as yet unknown but whether or not there is a change in decomposability is likely to affect the positive or negative outcome this trait might have on SOM. Apart from root and shoot tissues, the other factor that can influence the sequestration potential of plants is root exudation, both the quantity and quality of the compounds that are exuded. It has been suggested that root exudation is governed by plant metabolic activity (Bardgett et al., 2005), with faster growing species producing more exudates. Again, the effect on root exudation of selection for a higher WSC content is little understood. Higher WSC in shoots may mean a higher concentration of WSC across all plant organs and in exudates. Alternatively, it could mean that the plant is more successful in partitioning WSC and concentrating them in shoot tissues and hence reducing concentration in roots and exudates. Soils contain about twice as much C as the atmosphere and many soils are potentially able to sequester more than they do currently (Smith and Fang, 2010). Increasing steady-state soil C by 15% (e.g. from 0.05g/g to 0.058 g/g), it is claimed (Kell, 2011) would lower atmospheric CO2 by 30%, leading to a large environmental benefit. Deep soil carbon is an important contributor to overall soil carbon stocks with grasses and trees acting as major sources (Chabbi et al., 2009; Harper and Tibbett, 2013). Gregory et al. (2011) estimated that 980 Mt of organic C is stored below 30cm depth in soils in England and Wales, approximately 50% of the total. Evidence from tropical savannah suggests that the planting of exotic deeper rooting plant varieties leads to significant increases in soil carbon (Fisher et al., 1994). It is helpful that a grass trait that has received recent attention by crop geneticists for a range of reasons is for deeper rooting, with focus particularly on Festulolium (Kell, 2011). These are hybrids or hybrid derivatives between any Festuca (fescue) and Lolium (ryegrass) species, designed for their combined complementary characters (Ghesquière et al., 2010). The potential of certain Festulolium hybrid combinations for root growth and turnover to enhance SOC in sub-soils is currently being investigated and it may extend the range and diversity of soil biota. The deepGrassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 221 rooting trait was originally selected as an aid to an improved drought resistance (Durand et al., 2007; Alm et al., 2011), but also provides other ecosystem services, such as improving soil water retention and hence reduced runoff (Gregory et al., 2010; Macleod et al., 2013). The cultivation of Festulolium with large and deep-rooting systems as an aid to increase the input of atmospheric CO2 into agricultural soils is being assessed currently (J. Dungait, pers. comm.). However, there is conflicting opinion on whether the stores of SOC in subsoil will be increased or decreased by the introduction of new C. In addition, in order for deep-rooting plants to have any significant effect, the soil must be deeper than the root depth of current varieties, and furthermore root growth must not be impeded due to compaction, nutrient limitation or waterlogging. On thin soils or where roots cannot penetrate, deeper rooting plants are unlikely to provide additional sequestration benefits. Also, some studies have pointed to the effect of 'priming' whereby fresh labile organic matter stimulates the decomposition of older soil organic matter through the soil profile (Kuzyakov, 2010) and more specifically at depth (Fontaine et al., 2007). What happens to soil C at depth during disturbances such as drying or ploughing, land use change etc. is still little understood (Gregory et al., 2011). From an agricultural perspective, it is essential that a balance between above- and belowground growth is achieved so that forage yields are not significantly compromised. Recent studies at IBERS have demonstrated that Festulolium hybrids involving combinations of L. multiflorum or L. perenne with either F. arundinacea var glaucescens or F. mairei have excellent agronomic performance, including high yields of forage together with high WSC, and DOMD. From the perspective of their C sequestration potential, they also produce deep root systems that match their high above-ground growth. The L. perenne x F. pratensis amphiploid variety Prior has deep rooting and significant root turn-over at depth (MacLeod et al., 2013) leading to potential ecosystem benefits in terms of improved soil porosity. Dungait (pers. comm.) has provided preliminary data supporting greater root depth and C deposition by Festulolium cv. Prior in soils, compared with that of perennial ryegrass. Impacts of climate on C sequestration There have been extensive reviews (e.g. Soussana and Lüscher, 2007) of the impact of multiple components of climate change, singly or in combination, on the fluxes of C and N. As the major resource for photosynthesis, the expectation borne out by field experimentation is that elevated CO2 will increase all plant-derived C fluxes into the system. An increased uptake and availability of C due to elevated CO2 has been shown to increase C:N ratios in tissues within pasture species (Poorter et al., 1997) and to increase the exudation from roots of labile C (Allard et al., 2006). However, over an 11-year period of CO2 enrichment of grazed pastures, Newton et al. (2010) found a gradual transfer of N from plant to soil pools which would undoubtedly be a constraint on plant growth responses to elevated CO2. Overall, models predict higher total soil C sequestration in grasslands at elevated CO2 (Pepper et al., 2005). A recent study on Oklahoma prairie grasslands has shown that complex interactions occur in microbial diversity in soils in reaction to warming and drought. Warming without drought had the effects of increasing the abundance of microbes, but had the effect of reducing their diversity. Warming with drought resulted in large reductions of microbial populations but diversity was not affected. Some microbial communities are very resilient and recover quickly after the stress is removed, but others are not, becoming dormant and requiring significant time to reinstate after the stress had been removed (Sheik et al., 2011). Soil microbial composition is affected directly by the accompanying plant species in grasslands. Grigulis, et al. (2013) showed that high biomass production by plants was accompanied by bacterial dominance in the soil, fast microbial processes and a competitive strategy. By contrast, plant traits exhibiting low biomass production were associated with an increase in fungal communities, and slow microbial turnover. The low biomass communities Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 222 were therefore conservative by nature, providing ecosystem services such as high nutrient retention but lower C sequestration, while the more exploitative plant communities provided greater Net Primary Productivity and more C sequestration. Redesigning grassland crops for improved adaptations to climate change and for ecosystem service Comparative studies of genetic and phenotypic diversity found in natural grasslands adapted to contrasting climates provides us with insights into the key mechanisms involved in resistance to specific abiotic stress extremes, and reveals components of their genetic control. By selection and transfer from adapted ecotypes of trait-specific alleles, centuries of evolved adaptations may frequently be utilized over a few generations to reproduce a comparable phenotype in a commercial cultivar. For a perennial grassland species, not only is cultivar performance and stability over years important, but also its ability to coexist within complementary species’ mixtures and its competitive capabilities against invasive species. An extensive literature has explored the basis of grass-clover competition, with both above and below-ground factors found to be important. Shading of clover by grass can occur, depending on clover leaf size, management and, crucially, soil nitrogen (N) status. There is thus a strong interaction between interspecific plant competition and soil, particularly for the clover-Rhizobium symbiosis. Hydraulic lift is another positive species interaction that could be incorporated in species mixtures. The phenomenon, reported initially amongst tree and shrub species, describes how shallow-rooting species can benefit from improved acquisition of water and nutrients because of the release of water and nutrients into the topsoil from deep-rooting neighbours (e.g., Snyder et al., 2008). Advances in plant breeding strategy at IBERS over recent years have provided productive grass and clover varieties that also safeguard the environment. In the forefront are the high sugar ryegrasses (HSG) described earlier, and also legumes designed to improve ruminant nutrition by more efficient protein conversion and thereby to reduce greenhouse gas emissions by livestock. Two possible strategies of increasing efficiency of conversion of forage N to microbial N have been used: (i) increasing the amount of readily available energy accessible during the early part of the fermentation; and (ii) providing a level of protection to the forage proteins, and thereby reducing the rate at which their breakdown products are made available to the colonizing microbial population. The HSG are examples of the former, where increased WSC has been shown to have a positive impact on meat yields (Lee et al., 2001) and milk production (Miller et al., 2001). A significant contributor to greenhouse gas emissions by livestock is the plant-mediated proteolysis that occurs as a consequence of the assorted stresses encountered by living cells of ingested forage while in the rumen (Kingston-Smith et al., 2010). The second strategy to improve ruminant nutrition has focused on this, with a role for Festulolium in protein protection having been identified (O’Donovan et al., 2013). Amphiploid hybrids involving either L. multiflorum or L. perenne together with Festuca arundinacea var glaucescens have highly significantly enhanced protein stability compared with their Lolium parents when subjected to rumen-simulated conditions. It is hypothesized that protective measures that have evolved in the fescue to combat heat stress in its natural Mediterranean habitat, also come into play in the rumen to provide protein protection and to provide greater time for breakdown by rumen-based microbial populations. Plant breeding initiatives to improve protein stability under rumen conditions and to reduce greenhouse gas emissions by livestock have also involved legumes, such as polyphenol oxidase (PPO) expression by red clover (Trifolium pratense) (Webb et al., 2013). Condensed tannins also help to stabilize protein as it passes through the rumen, reducing loss of protein and preventing bloat. Although absent from the leaves of white and red clover, they are present in Lotus corniculatus and Lotus uliginosus (Marshall et al., 2008). Extensive variation in Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 223 condensed tannin content of the leaves of Lotus has been found. The development of varieties with appropriate levels of condensed tannins, to reduce protein loss without reducing intake or the agronomic yield and persistence is a research objective. Developing grasses with increased resilience to climate change Reference has already been made of the potential of Festulolium for ecosystem service. There are several examples of natural Festulolium hybrids. For example, both Italian ryegrass and perennial ryegrass hybrids with meadow fescue occur naturally in the UK as Festulolium braunii and Festulolium loliaceum, respectively, and have evolved with certain adaptive capabilities superior to their parent species. Festulolium loliaceum for example, is found commonly in mature meadows in flood-prone highly water-logged soils (Humphreys and Harper, 2008). IBERS achieved a notable first in 2012 in gaining inclusion onto the UK National Recommended Varieties List of the Festulolium variety AberNiche, which is a modified F. braunii comprising circa 10% meadow fescue genome (based on genetic and cytological screens fescue-specific using DarT markers and genome in-situ hybridization (GISH) (Kopecky and Harper, unpublished). The variety was developed for its high forage production and improved winter hardiness, presumably derived from its cold-tolerant fescue progenitor, but has also demonstrated excellent resilience to heat and drought. The entry of variety AberNiche onto the UK National Recommended Varieties List heralds a new dawn for Festulolium breeding in the UK. Genes for drought resistance have been transferred successfully from both F. arundinacea var glaucescens and from F. arundinacea onto chromosome 3 of both L. multiflorum and L. perenne and have improved the water use efficiency of both ryegrasses by >80% (Humphreys et al., 2013). Breeders’ lines with these fescue gene complements are currently in trial for variety assessment. Impact of flooding and drought The use of Festulolium for flood mitigation has been recently proposed (Macleod et al., 2013). In the UK, the cost of flooding is significant; it has been estimated that the devastating floods of summer 2007 cost the UK economy £3.2b (Environment Agency). Excessive run-off erodes top-soils and soil organic matter, and depletes valuable nutrients, with negative impacts on water quality. Eutrophication of surface and ground-waters in England and Wales is estimated to cost £75-114m a-1 due to loss of amenity value, reduced biodiversity and increased costs of water treatment. The cost of damage to agricultural soil in England and Wales has been estimated as £264m a-1, and that of treating water contaminated with agricultural pollutants as £203m a-1 (UK Parliamentary Office of Science & Technology, 2006). The winter of 2013-14 brought record rainfall to the UK leading to extreme flooding events across largely grasslanddominated areas. Insurance losses for England and Wales are expected to amount to £1.2b. Over the UK as a whole £600m of crops were lost due to flooding in 2012. To help mitigate the worst effects of flooding, Festulolium hybrid combinations are proving effective. Macleod et al. (2013) demonstrated how L. perenne x F. pratensis variety Prior reduced rainfall surface run-off by 51% compared to an IBERS-bred elite perennial ryegrass variety. It is hypothesized that the root turn-over at depth of the Festulolium led to improved soil structure and porosity and to improved soil water retention and reduced overland run-off. In a new five-year project (2014-2018) called SUREROOT, funded jointly by BBSRC and industry, IBERS and Rothamsted Research are assessing the potential of Festulolium and clover varieties both independently and in combination for flood mitigation. The trial will include field assessments on farms at different UK locations, on contrasting soils, and with alternative livestock management practice. The project also employs use of IBERS new National Plant Phenomics Centre and the North Wyke Farm Platform to assess how modified root systems on individual plant genotypes may if reproduced at the field scale affect soil structure and water, nutrient, Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 224 and C run-off at the field scale. Similar to Festulolium cv Prior, white clover improves soil structure and drainage through improved soil aggregation, which increases soil porosity and water infiltration (Mytton et al., 1993; Holtham et al., 2007). Values of macroporosity for soil cores under perennial ryegrass and under white clover were reported as 24% and 45%, respectively, indicating a significant improvement in water drainage through use of white clover (Holtham et al., 2007). Festulolium for bioremediation on mine-spoil Whilst primarily for economic reasons, mining underground for coal in the UK is largely no longer practised, it has been replaced by open-cast mining in areas such as South Wales, a practice viewed as a controversial due to its unsightly look, the large land areas employed, and the potential damage to both landscape and environment. It is considered a priority in such locations that areas used for open-cast mining are restored to their natural grassland ecology as quickly as possible as soon as coal extraction has ceased. However, combinations of a challenging climate and growth inhibiting mine-spoils have restricted progress to land reclamation. However, new Festulolium combinations: L. perenne x F. pratensis, L. perenne x F. arundinacea var glaucescens, and L. perenne x F. mairei amphiploid hybrids are being assessed by IBERS in trials on over-burden mounds on open-cast mine workings in South Wales aimed at their bioremediation and their restoration to grasslands. The combination of stress tolerance, rapid establishment and growth, nutrient and water-use efficiency, and large strong root systems present in the Festulolium hybrids provide them with an advantage over native grasses. Their rapid root turn-over should provide new sources of C to the mine spoil and provide the foundation for indigenous UK grasses to colonize and eventually return the land finally to its natural condition. Conclusion The development of new grassland agro-ecosystems must take account not only of traditional values of production, disease resistance, persistence and forage quality but also their impact on their surrounding environment. National variety assessment needs to be amended to account for the potential benefits to accrue through the use of novel varieties that offer enhanced resilience against climate change and/or provisions for other ecosystem service. A cohesive strategy is required involving relevant stakeholders to inform, encourage, and reward farmers to grow grassland varieties that provide a specific ecosystem service such as flood mitigation; to provide the necessary guidance as to where to sow, and how to manage sufficiently to produce the optimal benefits. Anthropomorphic and natural grasslands can potentially be mutually advantageous. Whilst historically agricultural developments have diminished areas of natural grassland and have led to the loss of their traditional environmental provisions, a reverse result may also be achieved. Novel grassland varieties have potential to engineer conditions that allow the restoration of natural grassland ecosystems, e.g. in land reclamation from industrially contaminated 'brownfield sites'. They may also extend the depth of sub-soils and extend the range of soil biota by rooting deeper than indigenous species. Indeed they may in species mixtures increase the resilience of grassland ecosystems to climate change in situations where the indigenous species lack the necessary adaptations to the stresses likely to be encountered. The genetic resources of many semi-natural grassland species can be harnessed to bring both the productive and ‘unproductive’ grasslands more into line. Semi-natural grasslands should not be showcased and 'fossilised in time and place' but must be preserved as essential resources for genetic material for future crop improvement and for ameliorating the damage done by traditional productive systems. Restoration at the landscape scale is required to redress the balance and Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 225 provide more connectivity, more ecosystem services, and more multifunctional dynamism in grassland management. Whilst new grassland varieties are required to reduce our 'agricultural footprint', to safeguard the future of fragile rural communities and to sustain our livestock industry, it is essential that further safeguards are put in place to assist the perpetuity of our existing natural and seminatural grasslands. In the current drive to increase productivity, natural grasslands are in danger of being further eroded and fragmented, greatly reducing their biodiversity and their ecosystem service value. They it must be remembered provide a much more diverse range of essential ecosystem services than those available within agricultural crop based systems. Indeed, restoration of partially degraded grasslands should also form part of national objectives. Grasslands, both natural and cultivated, dominate our landscape and both have a vital role in safeguarding the environment on which we all depend. References Aerts R. and Chapin F. S. (2000) The mineral nutrition of wild plants revisited: A re-evaluation of processes and patterns. Advances in Ecological Research 30, 1-67. Aguilar R., Ashworth L., Galetto L. and Aizen M. (2006) Plant reproductive susceptibility to habitat fragmentation: review and synthesis through a meta-analysis. Ecology Letters 968-980. Allard V., Robin C., Newton P.C.D., Lieffering M. and Soussana J.F. (2006) Short-term and long-term effects of elevated CO2 on Lolium perenne root exudation and its consequences on soil organic matter turnover and plant N yield. Soil Biology and Biochemistry 38, 1178-1187. Allison K. and Day C. (1999) A Guide to Herbs for Horses. J. A. Allen & Co Ltd, London. Alm V., Busso C. S., Ergon Å., Rudi H., Larsen A., Humphreys M.W. and Rognli O.A. (2011) A QTL analyses and comparative genetic mapping of frost tolerance, winter survival and drought tolerance in meadow fescue (Festuca pratensis Huds.). Theoretical and Applied Genetics 123 (3), 369-382. Altieri M.A. (1998) Agroecology: The Science of Sustainable Agriculture. (2nd Ed.). London: IT Publications. Badgley C., Moghtader J., Quintero E., Zakem E., Jahi Chappell M., Avilés-Vázquez K., Samulon A. and Perfecto I. (2007) Organic agriculture and the global food supply. Renewable Agriculture and Food Systems 22 (2), 86108. Baldock D. (1989) Agriculture and Habitat Loss in Europe. WWF International, Gland, Switzerland. Branquart E. and Hemptinne J.L. (2000) Selectivity in the exploitation of floral resources by hoverflies (Diptera: Syrphinae). Ecography 23, 732–742. Bardgett R.D., Bowman W.D., Kaufmann R. and Schmidt S.K. (2005) A temporal approach to linking aboveground and belowground ecology. Trends in Ecology and Evolution 20, 634-641. Bignal E.M. and McCracken D.I. (1996) Low-intensity farming systems in the conservation of the countryside. Journal of Applied Ecology 33, 413-424. Brenner J., Jimenez J., Sarda R. and Garola A. (2010) An assessment of the non-market value of the ecosystem services provided by the Catalan coastal zone, Spain. Ocean and Coastal Management 27-38. Bullock J.M., Pywell R. F. and Walker K.J. (2007) Long-term enhancement of agricultural production by restoration of biodiversity. Journal of Applied Ecology 44, 6-12. Carvell C. (2002) Habitat use and conservation of bumblebees (Bombus spp.) under different grassland management regimes. Biological Conservation 103, 33-49. Carvell C., Roy D.B., Smart S.M., Pywell R.F., Preston C.D. and Goulson D (2006) Declines in forage availability for bumblebees at a national scale. Biological Conservation 132, 481-489. Chabbi A., Kögel-Knabner I. and Rumpel C. (2009) Stabilised carbon in subsoil horizons is located in spatially distinct parts of the soil profile. Soil Biology and Biochemistry 41, 256-261. Chapin F.S., Woodwell G.M., Randerson J.T. et al. (2006) Reconciling carbon-cycle concepts, terminology, and methods. Ecosystems 9, 1041-1050. Council of the European Union (2010) Biodiversity: Post-2010 EU and Global Vision and Targets and International ABS Regime-council Conclusions, vol. 7536/10, Brussels. Critchley C. N. R., Burke M. J.W. and Stevens D. P. (2003) Conservation of lowland semi-natural grasslands in the UK: a review of botanical monitoring results from agri-environment schemes. Biological Conservation 115, 263-278. De Deyn G.B., Cornelissen J.H.C. and Bardgett R.D. (2008) Plant functional traits and soil carbon sequestration in contrasting biomes. Ecological Letters 11, 516-531. Drechsler M., Wätzold F., Johst K. and Shogren J.F. (2010) An agglomeration payment for cost-effective biodiversity conservation in spatially structured landscapes. Resource and Energy Economics 32, 261-275. Dunford B. (2002) Farming and the Burren. Teagasc, Dublin. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 226 Durand J. L., Bariac T., Ghesquière M., Brion P., Richard P., Humphreys M.W. and Zwierzykowski Z. (2007) Ranking of the depth of water extraction by individual grass plants using natural 18O isotope abundance. Environmental and Experimental Botany 60, 137-144. Feehan J. (2003) Farming in Ireland: History, Heritage and Environment. University College Dublin, Dublin. Fisher M. J., Rao I. M., Ayarza M.A., Lascano C.E., Sanz J.I., Thomas R.J. and Vera R.R. (1994) Carbon storage by introduced deep-rooted grasses in the South American savannas. Nature 371, 236-238. Fontaine S., Barot S., Barré P., Bdioui N., Mary B. and Rumpel C. (2007). Stability of organic carbon in deep soil layers controlled by fresh carbon supply. Nature 450, 277-280. Free J. B. (1993) Insect Pollination of Crops. 2nd ed. Academic Press, London. Fuller R. M. (1987) the changing extent and conservation interest in lowland grasslands in England and Wales: a review of grassland surveys 1930-84. Biological Conservation 40, 281-300. Fussell M. and Corbet S. A. (1991a) Flower usage by bumble-bees: a basis for forage plant management. Journal of Applied Ecology 29, 451-465 Fussell M. and Corbet S.A. (1991b) Bumble-bee habitat requirements: a public survey. Actae Horticulturae 288,159-163. Garnier E., Lavorel E., Ansquer P., Castro H., Cruz P., Dolezal J., et al. (2007) Assessing the effects of land-use change on plant traits, communities and ecosystem functioning in grasslands: a standardized methodology and lessons from an application to 11 European sites. Annals of Botany 967-985. Ghesquière M., Humphreys M.W. and Zwierzykowski Z. (2010) Festulolium. Handbook on Plant Breeding, Eucarpia Fodder Crops and Amenity Grasses Section, Veronesi, F. Posselt, U. Beart, B. (Eds) Springer, ISBN:9781441907592, 5, (Chapter 12), pp 293-316. Gregory A. S., Webster C. P., Watts C. W., Whalley W. R., Macleod C. J. A., Joynes A., Papadopoulos A., Haygarth P. M., Binley A., Humphreys M. W., Turner L. B. and Skøt, L. (2010) Soil management and grass species effects on the hydraulic properties of shrinking soils. Soil Science Society of America Journal 74, 753761. Gregory A. S., Whitmore A., Kirk G. J. D., Rawlins B., Ritz K. and Wallace P. (2011) Review of the evidence base for the status and change of soil carbon below 15 cm from the soil surface in England and Wales: DEFRA. Grigulis K., Lavorel S., Krainer U., Legay N., Baxendale C., Dumont M., et al. (2013) Relative contributions of plant traits and soil microbial properties to mountain grassland ecosystem services. Journal of Ecology 50-62. Harper R.J. and Tibbett M. (2013) The hidden organic carbon in deep mineral soils. Plant Soil 368, 641-648. Hickie D., Smyth E., Bohnsack U., Scott S. and Baldock D. (1999) A report on the impact of agriculture schemes and payments on aspects of Ireland’s heritage. The Heritage Council, Kilkenny. Holtham D. A. L., Matthews G. P. and Scholefield D. (2007) Measurement and simulation of void structure and hydraulic changes caused by root-induced soil structuring under white clover compared to ryegrass. Geoderma 142,142-151. Hofmann M. and Isselstein J. (2005) Species enrichment in an agriculturally improved grassland and its effects on botanical composition, yield and forage quality. Grass and Forage Science 60, 136-145. Hooper D. U., Chapin III F. S., Ewel J. J., Hector A., Inchausti P., Lavorel S., Lawton J. H., Lodge D. M. et al. (2005) Effects of biodiversity on ecosystem functioning: a consensus of current knowledge. Ecological Monographs 75, 3-35 Humphreys M.O. and Theodorou M.K. (2001) Breeding and utilising forages for sustainable ruminant production. In: S. C. Jarvis (Ed). Progress in grassland science: Achievements and opportunities. Proceedings of an IGER Research Colloquium. Okehampton, Devon, UK: IGER North Wyke. Humphreys M.W., Macleod C. J. A., Whalley W. R., Turner L. B., Farrell M. S., Ghesquière M., and Haygarth P. M. (2013). Designing grass cultivars for droughts and floods. In: breeding strategies for sustainable forage and turf grass improvement. (Eucarpia Fodder Crops and Amenity Section Meeting, Dublin, September 2011), Barth S. Milbourne D. (eds) Springer, ISBN:9789400745551, 171-180. Humphreys M.W. and Harper J.A. (2008) Festulolium loliaceum, an understudied natural UK grass hybrid species that may provide benefits to UK grasslands withstanding the onsets of climate change. European Crop Wild Relative Newsletter January, (6). Jiang L. L. P., Qin W. X., Wang R. J., Xu B. and Lu Q. (2007) Estimation of grassland ecosystem services value of China using remote sensing data. Journal of Natural Resources 2. Kell D. B. (2011) Breeding crop plants with deep roots: their role in sustainable carbon, nutrient and water sequestration. Annals of Botany 108, 407–418. King T. J. (1977) The plant ecology of ant-hills in calcareous grasslands: II. Succession on the mounds. Journal of Ecology 65 (1), 257-278. Kingston-Smith A., Edwards J. E., Huws S. A., Kim E. J., and Abberton M. 2010. Plant-based strategies towards minimising “livestocks’ long shadow”. Proceedings of the Nutrition Society 2000, 69:1-8. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 227 Kremen C., Williams N., Aizen M., Gemmill-Herren B., LeBuhn G., Minckley R., et al. (2007). Pollination and other ecosystem services produced by mobile organisms: a conceptual framework for the effects of land-use change. Ecology Letters 299-314. Kuzyakov Y. (2010) Priming effects: Interactions between living and dead organic matter. Soil Biology and Biochemistry 42, 1363-1371. Lauenroth W. K. (1979) Grassland primary production: North American grasslands in perspective. In: French, N.R. (ed.) Perspectives in Grassland Ecology. Ecological Studies. Springer-Verlag, New York, pp. 3–24. Lee M. R., Jones E. L., Moorby J. M., Humphreys M. O., Theodorou M. K., MacRae J. C. and Scollan N. D. (2001). Production responses from lambs grazed on Lolium perenne selected for an elevated water-soluble carbohydrate concentration. Animal Research 50, 441-450. Lovett G., Cole J. and Pace M. (2006) Is net ecosystem production equal to ecosystem carbon accumulation? Ecosystems 9, 152-155. Macleod C. J. A., Humphreys M. W., Whalley W. R., Turner L. B., Binley A., Watts C. W., Skot L., Joynes A., Hawkins S. L., King I. P., O'Donovan S. A., and Haygarth P. M. (2013). A novel grass hybrid to reduce flood generation in temperate regions. Scientific Reports, 3: 1683 DOI: 10.1038/srep01683 Marshall A. H., Bryant D. N., Latypova G. A., Hauck B., Olyott P., Morris P. and Robbins M. P. (2008) A high throughput method for the quantification of proanthocyanidins in forage crops and its application in assessing variation in condensed tannin content in breeding programmes for Lotus corniculatus L. and L. uliginosus. Journal of Agriculture and Food Chemistry 56, 974-981. Miller L., Moorby J. M., Davies D. R., Humphreys M. O., Scollan N. D., MacRae J. C. and Theodorou M. K. (2001). Increased concentration of water soluble carbohydrate in perennial ryegrass (Lolium perenne L.): milk production from late lactation dairy cows. Grass and Forage Science 56, 383-394. Mytton L. R., Cresswell A., and Colbourn P. (1993) Improvement in soil structure associated with white clover. Grass and Forage Science 48, 84-90 Newton P. C. D., Lieffering M., Bowatte W. M. S. D., Brock S.C., Hunt C. L., Theobald P. W. and Ross D. J. (2010) The rate of progression and stability of progressive nitrogen limitation at elevated atmospheric CO 2 in a grazed grassland over 11 years of Free Air CO2 enrichment. Plant and Soil 336, 433-441 O’Donovan S. A., Kingston-Smith A. H. and Humphreys M.W. (2013) Understanding the genetic basis for slow plant-mediated proteolysis in Festulolium hybrids. In: Breeding strategies for sustainable forage and turf grass improvement. Eds S Barth, D Milbourne Pub Springer-Verlag Chapter 11 89-95 O'Neill F. H., Martin J. R., Devaney F. M. and Perrin P.M. (2013) The Irish semi-natural grasslands survey 20072012. Irish Wildlife Manuals, No. 78. National Parks and Wildlife Service, Department of Arts, Heritage and the Gaeltacht, Ireland. O'Sullivan A. (1982) The lowland grasslands of Ireland. Journal of Life Sciences, Royal Dublin Society 3:131142. Parsons A, Rowarth J., Thornley J. and Newton P. (2011) Primary production of grasslands, herbage accumulation and use, and impacts of climate change. In: Lamaire G., Hodgson J., Chabbi A. (eds) Grassland Productivity and Ecosystem Services. Chapter 1. Wallingford UK: CABI. Pepper D. A., Del Grosso S. J., McMurtrie R. E. and Parton W. J. (2005) Simulated carbon sink response of shortgrass steppe, tallgrass prairie and forest ecosystems to rising [CO2], temperature and nitrogen input. Global Biogeochemical Cycles 19, 1-20. Peterken G. (2013) Meadows. British Wildlife Collection No. 2. British Wildlife Publishing, Gillingham, UK. Poorter H., van Berkel Y., Baxter B., Bel M., den Hertog J., Dijkstra P., Gifford R.M., Griffin K.L., Roumet C. and Wong S.C. (1997). The effect of elevated CO 2 on the chemical composition and construction costs of 27 C3 species. Plant, Cell and Environment 20, 472-482 Prŷs-Jones O.E. and Corbet, S.A. (1991) Bumblebees 2nd ed. Springer, Slough, UK. Purvis G., Anderson A., Baars J.-R., Bolger T., Breen J., Connolly J., Curry J., Doherty P., Michael D., Finn J., et al. (2008) Biodiversity-monitoring, Functional Significance and Management for the Maintenance and Economic Utilisation of Biodiversity in the Intensively Farmed Landscape. Environmental Protection Agency, Wexford. Sheik C., Beasley W., Elshahed M., Zhou X., Luo Y., and Krumholz L. (2011) Effect of warming and drought on grassland microbial communities. ISME 1692-1700. Smith P. and Fang C. (2010) C cycle: a warm response by soils. Nature 464, 499-500 Snyder K. A., James J.J., Richards J. H., and Donovan L. A. (2008) Does hydraulic lift or night-time transpiration facilitate nitrogen acquisition? Plant Soil 306, 159-166. Soussana J.-F. and Lüscher A. (2007) Temperate grasslands and global atmospheric change: a review. Grass and Forage Science 62, 127-134 Stevens D. P., Smith S. L. N., Blackstock T. H., Bosanquet S. D. and Stevens J.P. (2010) Grasslands of Wales. University of Wales Press, Cardiff. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 228 Sullivan C., Bourke D., Sheehy Skeffington M., Finn J., Green S, Kelly S. and Gormally M. (2011) Modelling semi-natural habitat area on lowland farms in western Ireland. Biological Conservation 144: 1089-1099. Sullivan C.A., Sheehy Skeffington M., Gormally M.J., and Finn J.A. (2010) The ecological status of grasslands on lowland farmlands in western Ireland and implications for grassland classification and nature value assessment Biological Conservation 143, 1529-1539. Thornley J.H.M. and Johnson I.R. (1990). In: Plant and Crop Modelling. Clarendon Press, Oxford, UK. Tilman D., Hill J., and Lehman C. (2006a) Carbon-negative biofuels from low input high diversity grassland biomass. Science, 1598-1600. Tilman D., Reich P. and Knops J. (2006b). Biodiversity and ecosystem stability in a decade-long grassland experiment. Nature 629-632. Tilman, D., Wedin, D., and Knops, J. (1996) Productivity and sustainability influenced by biodiversity in grassland ecosystems. Nature 379, 718-20. Tilman, D., Lehman, C. L., and Thomson, K. T. (1997) Plant diversity and ecosystem productivity: theoretical considerations. Proceedings of the National Academy of Science, USA. 94, 1857-61. Webb J. K., Cookson A., Allison G., Sullivan M. L., and Winters A. L. (2013) Gene Expression Patterns, Localization, and Substrates of Polyphenol Oxidase in Red Clover (Trifolium pratense L.) Journal of Agricultural and Food Chemistry 61, 7421–7430 Williams B., Parr S., Moran J., Dunford B., and Ó Conchúir R. (2009) The Burren –farming for the future of the fertile rock. British Wildlife 21, 1-9. Williams P.H. (1982) The distribution and decline of British bumblebees (Bombus Latr.). Journal of Apicultural Research 121 (4) 236-245. Williams P.H. (1988) Habitat use by bumble bees (Bombus species) Ecological Entomology 13, 223-237. Williams P.H. Corbet S. A. and Osborne J.L. (1991) Bee-keeping, wild bees and pollination in the European Community. Bee World 72 (4) 170-180. Xie Y., Sha Z., and Yu M. (2008) Remote sensing imagery in vegetation mapping: a review. Journal of Plant Ecology 1(1) 9-23. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 229 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 230 Theme 2 submitted papers Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 231 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 232 Single species and mixed grazing regimes to restore Nardus stricta moorland Critchley C.N.R.1, Griffiths J.B.2 and Clarke A.2 1 ADAS UK Ltd, c/o Newcastle University, NEFG Offices, Nafferton Farm, Stocksfield, United Kingdom 2 ADAS UK Ltd, Canolafan Enterprise Park, Llanafan, Aberystwyth, United Kingdom. Corresponding author: nigel.critchley@adas.co.uk Abstract Upland heathlands provide a range of ecosystem services but have been degraded by high stocking densities, causing replacement of dwarf shrubs by grasses such as Nardus stricta. A long-term paddock-scale experiment is being conducted to assess the effects of sheep, cattle and mixed grazing regimes on vegetation and livestock performance on former dwarf shrub heathland dominated by N. stricta. Treatments of 1.0 and 1.5 ewes ha-1, 0.5 cattle ha-1 in summer, and mixed grazing comprising 1.0 ewes ha-1 plus 0.5 cattle ha-1 in summer, have been applied over ten years, including a two-year period with no grazing, representing a ‘pulsed’ grazing system. Across all treatments, N. stricta frequency declined and this continued during the two years with no grazing, but the decline was not mirrored by any increase in dwarf shrubs. In the first year when grazing was re-introduced, cattle and sheep liveweight gains were satisfactory for commercial production and at least comparable to gains during previous years of the experiment. Pulsed grazing regimes with sheep or cattle at low to moderate stocking densities could be viable, but with beneficial effects on vegetation only emerging over several years. Keywords: Nardus stricta, dwarf shrubs, upland heathland, sheep, cattle, mixed grazing Introduction Upland heathlands in the UK are an internationally important habitat, providing a range of ecosystem services including biodiversity and food from extensive livestock production, as well as flood regulation, carbon storage and recreation. Past high stocking densities, particularly of sheep in the second half of the 20th century, caused degradation of upland heathland in many parts of the UK. This resulted in a loss of dwarf shrub species such as Calluna vulgaris and Vaccinium myrtillus and their replacement by grasses such as Nardus stricta. Since the late 1980s, agri-environment schemes in the UK have encouraged farmers to reduce sheep numbers in an attempt to reverse this process. However, cattle consume N. stricta more readily than do sheep, and could play a positive role in altering competitive interactions between plant species. Rotational or ‘pulsed’ grazing systems, in which an area is left ungrazed for up to 3-4 years between periods of grazing, could also help to restore upland heathland (Gardner et al., 2009). A long term paddock-scale experiment has been set up to assess the effects of sheep, cattle and mixed grazing regimes on biodiversity and livestock performance when applied to an area of N. stricta acid grassland. Here we examine the effects on vegetation and livestock performance following a two-year period without grazing, as part of a long term pulsed-grazing system. Materials and methods The study site was 71 ha of N. stricta grassland at an elevation of 305–625 m above sea level, with annual rainfall c. 2000 mm, in mid-Wales, UK (52° 22' N 3° 46' W). Soils were Stagnopodzols with peaty top soils or shallow soils with bedrock at 30 cm. The area was formerly dwarf shrub heath but degraded by a long period of sheep grazing at c. 2.2–2.7 ewes per ha up to the mid-1990s. The vegetation resembled variants of the U5 N. stricta – Galium saxatile grassland (Rodwell, 1992). The site was divided into twelve paddocks and four grazing Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 233 treatments applied from spring 2003 to autumn 2010, allocated randomly to paddocks and replicated over three blocks. All livestock were then removed for two years before reintroducing the grazing treatments in 2013. Treatments were low sheep (LS; 1.0 Welsh Mountain ewes ha-1 for ten months per year + lamb from May to August), high sheep (HS; 1.5 Welsh Mountain ewes ha-1 for ten months per year + lamb from May to August), cattle only (CO; 0.5 Welsh Black 2-year-old heifers ha-1 for two months in July and August) and mixed sheep plus cattle grazing (SC; low sheep + cattle only regimes). A grid of 125 points at 75 m spacing was superimposed on the study site and vegetation recorded at each point from a fixed 1 m × 1 m quadrat subdivided into 100 cells of 10 cm × 10 cm. Local shoot frequency (presence in each 10 cm cell) and a grazing index (proportion of occupied cells in which grazed shoots were present) of four key species (N. stricta, V. myrtillus, C. vulgaris, Molinia caerulea) had been recorded annually from 2003 to 2006, and this was repeated in 2010 and 2012. The maximum sward height within each quadrat was recorded using a sward stick. Vegetation records were made in September–October. Liveweights were recorded for all heifers when turned onto and removed from the paddocks, and for ewes from the onset of grazing in June 2012 to weaning in August. Lambs were weighed before turning on to the paddocks in June and again at shearing in July and at weaning in August. All data were analysed using repeated-measures Analysis of Variance in Statistica v11© (Statsoft Inc., Tulsa, Oklahoma). Results and discussion Mean vegetation height increased across all treatments during the period without grazing, but only by c. 2 cm, just outside the limits of statistical significance (F1,6 = 5.2, P = 0.06). Height variability (coefficient of variation) appeared to decline overall (F1,6 = 5.2, P = 0.06). Background grazing levels from wild herbivores in 2012 were negligible, in contrast to the previous years with livestock when, for example, grazing indices were 21 – 23% on V. myrtillus in treatments with sheep, and 54 – 63% on M. caerulea in treatments with cattle. N. stricta frequency appeared to decline during the period with no grazing, from 37.6% in 2010 to 30.8% in 2012 (F1,6 = 5.5, P = 0.06). This occurred across all treatments (year x treatment interaction not significant) and continued a decreasing trend recorded previously, from 51.4% in 2003 to 36.8% in 2005 (Critchley et al., 2013) (Figure 1). Figure 1. Local frequency (means and standard errors) of N. stricta in the four grazing regimes. HS = high sheep; LS = low sheep; SC = sheep + cattle; CO = cattle only Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 234 This trend could be a long-term response to the reduced numbers of livestock on the site during the experiment, which is likely to increase competition on N. stricta from more palatable grasses (Gardner et al., 2009). M. caerulea was only present at low frequencies but showed a significant increase from 4.4% to 6.7% across all treatments during the no-grazing period (F1,6 = 11.5, P < 0.05). However, no changes or other effects were detected for C. vulgaris or V. myrtillus frequencies. In the first year when grazing was re-introduced (2013), cattle liveweights increased significantly (F1,20 = 120.3, P < 0.001) during the period on the paddocks. Weight gains were slightly greater in the CO treatment (0.9 kg day-1 +/- 0.40 st.dev.) than SC (0.6 kg day-1 +/0.20 st.dev.) (F1,20 = 5.4, P < 0.05), and were considered adequate for commercial production. Cattle liveweight gains in 2013 compared favourably with those during the period 2003 to 2010, when overall mean daily liveweight gains were 0.6 kg day-1 (CO) and 0.5 kg day-1 (SC). This might be attributable to an increase in availability of forage biomass following the two years without grazing, although younger heifers were used in 2013, with lower mean initial liveweights (334 kg) compared to the period 2003 to 2010 (401 – 536 kg), and more potential for growth. Ewe liveweights increased significantly (F1,59 = 6.1, P < 0.05) from the onset of grazing in June 2013 to weaning in August (15 g day-1 +/- 50.6 st.dev.) but there were no significant differences between treatments. Lamb liveweight increases from the time on the paddocks up until shearing in July were lower in the HS treatment (132 g day-1 +/36.7 st.dev.) than LS (163 g day-1 +/- 49.3 st.dev.) and SC (177 g day-1 +/- 57.5 st.dev.) (F2,58 = 5.6, P < 0.01). However, there were no significant treatment effects on liveweight gains between shearing in July and weaning in August, nor on actual weaning weights (25 kg +/- 4.1 st.dev.), which were comparable to those from 2004 to 2010 (c. 24 – 32 kg overall). Conclusion Initial results suggest that pulsed grazing regimes with sheep or cattle at low to moderate stocking densities could have some small beneficial effects on vegetation without compromising livestock performance but major changes are likely to take much longer. Acknowledgements The work is funded by the UK Department for Environment, Food and Rural Affairs. We are grateful to many colleagues who have run this experiment since its inception in 2002. References Critchley C.N.R., Griffiths J.B., Clarke A. and Davies O.D. (2013) Effects of sheep and cattle grazing on vegetation, invertebrates and livestock performance on an upland acid grassland. Aspects of Applied Biology 118, 145-152. Gardner S.M., Waterhouse T. and Critchley C.N.R. (2009) Moorland management with livestock: the effect of policy change on upland grazing, vegetation and farm economics. In: Bonn A., Allott T., Hubacek K. and Stewart J. (eds) Drivers of Environmental Change in Uplands, Routledge, Abingdon, UK, pp. 186-208. Rodwell J.S. (ed.) (1992) British Plant Communities, Volume 3 Grasslands and Montane Communities, Cambridge University Press, Cambridge, UK, 540 pp . Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 235 Integrating biodiversity conservation with grassland farming: extensive cattle grazing and farmland birds Buckingham D.L.1, Brook A.J.2, Eschen R.2, Maczey N.2, Wheeler K.3 and Peach W.J.1 1 RSPB Centre for Conservation Science, RSPB, The Lodge, Sandy, Bedfordshire, SG19 2DL United Kingdom, 2 CABI, Bakeham Lane, Egham, TW20 9TY, United Kingdom, 3 ADAS, Rosemaund, Preston Wynne, Hereford, HR1 3PG, United Kingdom. Corresponding author: david.buckingham@rspb.org.uk Abstract Extensively grazed pastures are a valuable habitat for invertebrate-feeding birds and thus form a key component of agri-environment schemes that aim to restore declining farmland bird populations. Extensive grazing measures are potentially costly to implement as agricultural outputs can be impaired. The conflicting demands for food production (or cost to the agrienvironment scheme) and for conservation measures to be effective can be reconciled if the trade-offs between their different requirements are understood. This paper describes how an effective, low-cost extensive grazing measure was developed, by experimentally manipulating cattle grazing pressure to promote selective grazing and increase structural heterogeneity. The solution proved to be simple and flexible. Managing stocking rates to maintain an average sward height in the range 9-12 cm provided valuable foraging habitats for threatened farmland birds, while keeping agricultural costs well within available conservation budgets. The measure has been adopted by the English agri-environment scheme. Keywords: agri-environment measures, birds, extensive grazing Introduction A high proportion of the UK farmland bird species that are in severe long-term decline require invertebrate prey to rear their young. Extensively grazed pastures are an important source of suitable prey and play an important role in maintaining reproductive output in species such as the yellowhammer (Emberiza citrinella) and cirl bunting (E. cirlus) (Buckingham, 2006). Extensively grazed pastures form a key component of agri-environment schemes aiming to maintain or restore farmland bird populations, but they are potentially costly because suitable management curtails livestock production. This paper describes the research behind the development of an effective but affordable extensive grazing measure. The optimization process was guided by learning how structural heterogeneity in the sward can simultaneously meet the conflicting requirements of foraging birds, their invertebrate prey and livestock. This structural heterogeneity was created by reducing grazing pressure to encourage cattle to graze selectively (Hofmann and Tallowin, 2004). Materials and methods Extensive grazing treatments were compared in a series of split-field experiments between 2006 and 2012. In each experiment, all treatments were replicated in each study field, in 1-ha paddocks separated by fences. The paddocks were grazed by cattle and stocking rates were manipulated to maintain target sward heights (TSH). Sward structure, biodiversity metrics (plant and invertebrate communities; bird usage) and livestock live-weight yields were measured at the paddock level to quantify treatment effects in each year of the experiment. Variation within paddocks was measured to investigate the role of structural heterogeneity. The first experiment (2006-2009) contrasted a commercially grazed control (TSH 6-8 cm, April-October) with early closure (cattle removed in mid July) at two levels of grazing pressure: moderate (TSH 7-9 cm) and lenient (TSH 12-15 cm) (Peach, 2010). The experiment Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 236 was conducted on 13 unfertilized, semi-improved, permanent pastures. A subsidiary experiment tested these treatments for two years on 8 agriculturally improved permanent pastures with a recent history of fertilizer applications ≥50 kg N ha-1 year-1. The second experiment (2010-2012) compared the same commercially grazed control with two treatments which were kept at the same intermediate TSH (9-12 cm), by grazing continuously or intermittently (as in a rotational paddock grazing system) (Buckingham, 2013). All fields were situated in Devon, SW England and received no fertilizers during the experiments. Results and discussion The first experiment showed that lenient grazing combined with early closure led immediately to increased invertebrate densities in the first year, with further cumulative increases over time (Peach, 2010; Eschen et al., 2012). During the breeding season, invertebrate-feeding birds significantly preferred the extensively managed treatments as they held the most prey taxa. These patterns were observed on semi-improved and improved pastures. However, the lenient early-closure treatment caused substantial live weight yield losses (32% in year 2, rising to 62% in year 4) and key conservation priority bird species (notably the Emberiza buntings) did not respond strongly to the treatments. Partitioning the treatment responses into their constituent management components showed that early closure caused the biggest yield reductions through the loss of late-season grazing. Both early closure and lenient grazing also increased less-productive grasses, particularly Holcus and Agrostis, at the expense of ryegrass (Lolium perenne) and white clover (Trifolium repens). Lenient grazing resulted in the largest increases in invertebrate densities. Contrary to expectation, structural heterogeneity fell over time on both extensively managed treatments and fell below levels required by foraging buntings, explaining why usage levels were so far below those observed in other grassland studies (Stephens et al., 2002; Buckingham, 2006). Early closure failed to provide an anticipated increase in the availability of large-bodied invertebrate prey taxa in the early part of the bunting breeding season. The extensive grazing treatments mainly benefited birds in the second half of the breeding season. Based on these results, early closure was abandoned because of its high agronomic costs and smaller biodiversity benefits, compared to lenient grazing. Grazing throughout the growing season at a lower TSH was predicted to reduce yield losses and sward deterioration and to increase structural heterogeneity in the sward, thereby increasing its utility to foraging buntings. An intermediate TSH (9-12 cm) was adopted in an attempt to retain elevated invertebrate prey densities and high usage by foraging and nesting skylarks, all of which were predicted to decline in shorter swards. The resulting prediction was tested in the second fieldscale experiment. The second experiment confirmed that intermediate TSH grazing significantly increased invertebrate densities and the utility of pastures to both buntings and skylarks (Buckingham, 2013). The agricultural costs of intermediate TSH grazing were low in the first two years (£3771 ha-1 year-1). However, the third year (2012) was adversely affected by prolonged high rainfall, leading to a sharp increase in yield losses (£139 ha-1 year-1) and reduced biodiversity benefits. It is not clear whether agricultural costs would remain low for longer, in the absence of extreme weather, but these figures places bounds on the likely costs in adverse conditions. In practice, there were numerous interruptions to the continuity of grazing, particularly during dry conditions in the first two years, so the continuous and intermittent treatments differed mainly in the duration and frequency of grazing interruptions. Both continuous and intermittent grazing treatments resulted in similar biodiversity benefits on intermediate TSH paddocks, but invertebrate densities and bird usage tended to be lower on the intermittent grazing treatment. There was no evidence that the intermittent treatment enhanced sward heterogeneity. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 237 Conclusion Having been proven effective and affordable, intermediate TSH grazing will be funded through the new English agri-environment scheme (NELMS: the New Environmental Land Management Scheme, beginning in 2016). The 9-12 cm TSH allows farmers to adapt their stocking levels to local conditions, including seasonality of grass growth and weather. Additional management flexibility is allowed where evidence from the experiments indicates that it will not compromise the biodiversity objectives. The principal uncertainty is how long the intermediate TSH swards can be maintained without swards or yields deteriorating, particularly if extreme weather occurs. The new measure allows farmers to remove a field from this option after two years, if necessary. The measure is then implemented on a replacement field, so that an agreed area of measure is present on the farm throughout the NELMS agreement. Where possible, farmers are encouraged to continue managing fields with intermediate TSHs for longer, as the biodiversity benefits are expected to be cumulative. The measure also allows continuous or intermittent (paddock) grazing and small applications of farmyard manure, as the evidence suggests that these changes do not impair the biodiversity benefits. The area of this measure that is required to benefit a bird species at the population level is unknown. However, a single field managed correctly will benefit bunting territories within 200-400 m (Stevens et al., 2002), so only a small area within a farm needs to be committed to this measure, for it to have an effect. Acknowledgments This work was funded by Defra (contracts BD1454, BD5206 & BD5207). We would like to thank the project steering group members Richard Brand-Hardy, Val Brown, Steve Peel and Phil Grice for their guidance throughout the project. References Buckingham D.L. (2006) The effects of food abundance, sward structure and management on foraging by yellowhammers on agricultural grasslands. PhD thesis, University of Reading. Buckingham D.L. (2013) Utility of lenient grazing of agricultural grassland to promote in-field structural heterogeneity, invertebrates and bird foraging. Report to Defra on project BD5207. Defra, London, UK. http://randd.defra.gov.uk/ Eschen R., Brook A.J., Maczey N., Bradbury A., Mayo A., Watts P., Buckingham D., Wheeler K. and Peach W.J. (2012) Effects of reduced grazing intensity on pasture vegetation and invertebrates. Agriculture, Ecosystems and Environment 151, 53-60. Hofmann M. and Tallowin J.R.B (2004) Sward height distribution and temporal stability on a continuously stocked, botanically diverse pasture. Grassland Science in Europe 9, 192-194. Stevens D.K., Donald P.F., Evans A.D., Buckingham D.L. and Evans J. (2002) Territory distribution and foraging patterns of cirl buntings (Emberiza cirlus) breeding in the UK. Biological Conservation 107, 307-313. Peach W.J. (2010) Modified management of agricultural grassland to promote in-field structural heterogeneity, invertebrates and bird populations in pastoral landscapes. Report to Defra on project BD1454. Defra, London, UK. http://randd.defra.gov.uk/ Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 238 Spatial soil variation on the North Wyke Farm Platform Shepherd A.1, Harris P.1, Griffith B.1, Noacco V.2, Ramezani K.2, Tuominen E.2 and Eludoyin A.3 1 Rothamsted Research, North Wyke, Okehampton, Devon EX20 2SB, United Kingdom 2 Faculty of Engineering, University of Bristol, Bristol, BS8 1TR, United Kingdom 3 Department of Geography, University of Exeter, Exeter, Devon EX4 4RJ, United Kingdom Corresponding author: anita.shepherd@rothamsted.ac.uk Abstract Rothamsted Research at North Wyke in the south-west of England hosts the Farm Platform, a unique research facility for sustainable grassland systems. Fifteen catchments are distributed between three farmlets, ear-marked for different future land uses from April 2013. In order to check that the catchments had been allocated equally between the three farmlets with regards to quality, the soil was surveyed for each catchment in June 2012. Soil parameters of bulk density, pH, total nitrogen, total carbon and organic matter were surveyed on a square 50 m grid system and catchments differed from poorer soils to better soils with respect to the parameters surveyed. Catchments with sampled means at either ends of the range displayed consistency, producing a lower coefficient of variation than catchments with sampled means in the middle of the 15-field range. The means and coefficient of variation showed that the allocation of catchments to each farmlet gave an even share of catchments to each future land use. Keywords: farm platform, soil nutrients, spatial variation, grassland Introduction Advanced technology in the field of auto-measurement has a major role to play in agri-research (Griffith et al., 2013), but a key requirement is to start with a knowledge of baseline conditions (Orr et al., 2011) before measuring the effects of treatments. Rothamsted Research at North Wyke hosts the Farm Platform which contains 15 catchments consisting of one or more fields on permanent pasture which are hydrologically isolated and auto-sampled for drainage runoff and soil moisture, and monitored for agronomic productivity of beef and sheep systems. The dominant soil type at North Wyke is Hallsworth and Halstow series clay (Harrod and Hogan, 2008) with an impermeable layer at 30 cm depth, and runoff is collected at the perimeters leading to a flume laboratory where water flow and chemical/physical properties are measured. Catchments were allocated to three farmlets (Figure 1) from April 2011 for two years, after which they began to change to different land managements. Materials and methods Unstratified regular square grid sampling of soil was conducted in June 2012 in all the catchments on a fixed grid of 50 m with known GPS coordinates and 243 soil samples were collected and analysed. Bulk density was determined in samples which were collected by pressing a cylindrical core with 5.5 cm diameter and 10 cm depth into the soil. Samples were sieved to separate vegetation and stones larger than 2 mm and the volume of these was then measured by displacement of water. The remaining soil was oven dried at 105˚C for 24 h and bulk density was calculated. Total nitrogen, total carbon and organic matter and pH were measured at each sampling point in 10 soil cores which were 2 cm diameter and 10 cm depth. These were mixed and sieved though 2 mm mesh, ground and oven dried for 24 h. Analysis was conducted using an elemental analyser (NA2000, Carlo Erba Instruments, Milan, Italy) and soil organic matter was determined by loss on ignition. Soil pH was determined by adding 25 ml of deionised water to 10 ml of soil and measuring the solution pH. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 239 Figure 1. The North Wyke Farm Platform showing grey-scale farmlets each with 5 numbered catchments containing one or more fields. Results and discussion The mean sampled bulk density (Table 1) varied from 0.8 (coefficient of variation 0.10) to 0.98 (c.v. 0.07) with a catchment range of c.v. from 0.05 to 0.13. The farmlet means ranged from 0.86 to 0.92. Soil pH varied from 5.28 (c.v. 0.02) to 6.15 (c.v. 0.02) with a catchment range of c.v. from 0.01 to 0.08. The farmlet means ranged from 5.49 to 5.68. Total nitrogen varied from 4.27 (c.v. 0.07) to 6.79 (c.v. 0.04) with a catchment range of c.v. from 0.06 to 0.20. The farmlet means ranged from 5.28 to 5.59. Total carbon varied from 31.24 (c.v. 0.04) to 58.80 (c.v. 0.07) with a catchment range of c.v. from 0.04 to 0.23. The farmlet means ranged from 46.04 to 46.65. Organic matter varied from 6.78 (c.v. 0.11) to 12.46 (c.v. 0.05) with a catchment range of c.v. from 0.04 to 0.23. The farmlet means ranged from 10.04 to 10.48. In all cases the catchments showing highest and lowest parameters displayed lower variations, hence the highest and lowest means were the most stable. The three Longlands fields included between them the highest bulk density and pH, with the lowest nitrogen carbon and organic matter. Meanwhile, fields used previously for dairying contained higher organic matter, total carbon and total nitrogen content; e.g. Lower Wheaty and Dairy North. Both the Longlands fields and the former dairying fields were shared amongst each farmlet, which was reflected in the farmlet nutrient results being evenly distributed. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 240 Table 1. Soil parameters sampled in 2012 and mean values per catchment * and per farmlet (with coefficient of variation). Catchments BD (g/cm3) pH Total N (g/kg) Total C (g/kg) FARMLET A Pecketsford & Little Pecketsford 0.92 (0.09) 5.47 (0.03) 6.05 (0.09) 49.84 (0.10) Great Field 0.97 (0.12) 5.88 (0.05) 4.28 (0.13) 34.33 (0.15) Poor Field & Ware Park 0.85 (0.07) 5.62 (0.08) 5.41 (0.15) 44.30 (0.17) Lower Wheaty 0.91 (0.04) 5.28 (0.02) 6.79 (0.04) 57.54 (0.05) Longlands East 0.96 (0.05) 6.15 (0.02) 4.37 (0.11) 35.98 (0.16) FARMLET B Burrows & Bottom Burrows 0.91 (0.09) 5.67 (0.06) 5.35 (0.11) 44.15 (0.12) Orchard Dean 0.92 (0.10) 5.79 (0.04) 5.82 (0.12) 51.74 (0.15) Golden Rove 0.94 (0.07) 5.59 (0.06) 5.84 (0.08) 50.15 (0.07) Dairy North 0.92 (0.03) 5.79 (0.02) 6.55 (0.05) 58.80 (0.07) Longlands South 0.98 (0.07) 5.48 (0.01) 4.56 (0.06) 38.08 (0.08) FARMLET C Higher & Middle Wyke Moor 0.84 (0.08) 5.38 (0.06) 4.53 (0.13) 41.21 (0.15) Lower Wyke Moor 0.80 (0.10) 5.61 (0.04) 5.23 (0.16) 48.30 (0.17) Dairy South & Dairy Corner 0.89 (0.07) 5.56 (0.04) 5.99 (0.10) 52.55 (0.20) Dairy East 0.91 (0.05) 5.40 (0.03) 6.25 (0.06) 55.12 (0.08) Longlands North 0.91 (0.13) 5.53 (0.01) 4.27 (0.07) 31.24 (0.04) PER FARMLET FARMLET A (n = 84) 0.91 (0.10) 5.64 (0.07) 5.28 (0.20) 43.41 (0.22) FARMLET B (n = 81) 0.92 (0.09) 5.68 (0.05) 5.59 (0.13) 48.04 (0.16) FARMLET C (n = 85) 0.86 (0.09) 5.49 (0.05) 5.28 (0.18) 46.65 (0.22) * Catchment = Fields that are hydrologically isolated to measure runoff flow and quality SOM (mg/litre) 11.20 (0.10) 7.73 (0.11) 10.83 (0.16) 11.82 (0.06) 9.35 (0.10) 8.93 (0.23) 12.08 (0.13) 10.97 (0.04) 12.46 (0.05) 8.76 (0.07) 9.41 (0.12) 9.76 (0.15) 11.75 (0.12) 11.99 (0.05) 6.78 (0.11) 10.04 (0.20) 10.48 (0.20) 10.33 (0.18) Conclusions Catchments varied in respect of the parameters surveyed. Fields used previously for dairying were generally richer in nutrients. The means and coefficient of variation confirm that the allocation of catchments to each farmlet gave an even range of soil quality to each of the three farmlets which will in future have different land use treatments. Acknowledgements The North Wyke Farm Platform has been developed with funding from the Biotechnology and Biological Sciences Research Council. References Griffith B.A., Hawkins J.M.B., Orr R.J., Blackwell M.S.A. and Murray P.J. (2013) The North Wyke Farm Platform: Methodologies used in the remote sensing of the quantity and quality of drainage water. International Grassland Congress, Sydney, Australia. Harrod T. R. and Hogan, D. V. (2008). The soils of North Wyke and Rowden. http://www.rothamsted.ac.uk/sites/default/files/SoilsNWRowden.pdf Orr R.J., Griffith B.A., Rose S., Hatch D.J., Hawkins J.M.B. and Murray P.J. (2011) Designing and creating the North Wyke Farm Platform. Catchment Science 2011, 14 – 16 September, Dublin, Ireland. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 241 “Virtual grassland”: an individual-based model to deal with grassland community dynamics under fluctuating water and nitrogen availability Louarn G., Escobar-Gutiérrez A., Migault V., Faverjon L. and Combes D. INRA, UR4 Plurdisciplinaire Prairies Plantes Fourragères (P3F), BP6, F-86600 Lusignan, France Corresponding author: gaetan.louarn@lusignan.inra.fr Abstract The “Virtual grassland” model is being developed to provide herbaceous grass and legume plant models and coupling methods with soil and light transfer modules on the OpenAlea platform. It aims at simulating the effects of competition and facilitation for light, water and nitrogen acquisition at the population and community levels in grasslands. Preliminary assessments of the coupled models suggest good qualitative behaviour with respect to light, defoliation and water responses. Quantitative model assessment is still ongoing. Keywords: architecture, grass, legume, community, dynamics, water stress Introduction Multi-specific grasslands are a major source of forages worldwide. Their agricultural use-value depends on the structure of their canopy and on their botanical composition. Both determine the quantity and the quality of the biomass harvested by grazing or mowing. In temperate grasslands, perennial grasses and forage legumes dominate the floristic composition of the grasslands and are they generally intended to be grown in mixtures because of their agronomic (feeding value) and ecological (resource capture and use) complementarities (Louarn et al., 2010). An appropriate species balance is, however, difficult to maintain in such nonequilibrium communities. The proportion of forage legumes fluctuates from year-to-year, within a year between harvests, and even within a single growth period as a result of species interactions and management. In spite of this agronomic significance, most grassland models developed so far ignore plant-plant interactions and do not simulate the community dynamics of grasslands. Efforts have been done mainly on white clover based mixtures for which a gain of understanding of coexistence conditions was achieved thanks to individual-based models (Schwinning and Parsons, 1996; Soussana and Oliveira Machado, 2000). However, these predictions may present a limited interest for other grass-legume combinations. Indeed, white clover has rather an atypical colonization strategy and persistence habit among temperate forage legumes. The exposition of shoot apex to defoliation and the lack of an efficient vegetative reproduction in most tap-rooted forage legumes are known to affect their population dynamic and could modify the impact that legumes have on nutrient cycling. Furthermore, for white clover, being grown predominantly in areas with limited water stress, water shortage effects have not been included in these models. In order to deal with these questions, the “Virtual grassland” model at (http://openalea.gforge.inria.fr/wiki/doku.php?id=packages:ecophysio:grassland) is being developed under the open source OpenAlea modelling platform (Pradal et al., 2009) to simulate the architecture of various grass and legume species and predict the effects of plant interactions for light, water and nitrogen at the population and community levels. Materials and methods The “Virtual grassland” model provides: - i) Two generic plant models accounting respectively, for grass (L-grass; Verdenal et al., 2008) and legume (L-egume) morphogenesis. These models simulate 3D plant architecture above- Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 242 and below-ground and provide a central data structure describing plant structure from organ to community scales (MTG); - ii) Coupling methods (in the form of OpenAlea nodes) between the MTG central data structure for plant architecture and environmental models to simulate competition for resources above- (light) and belowground (water and soil N). The light transfer (Sinoquet et al., 2001), gas exchange (Prieto et al., 2012) and soil models (adaptation of STICS soil module; Brisson et al., 2008) are available through the Openalea platform and were reused from previously published models. - iii) A graphical user interface in the form of a dataflow enabling to manage model input parameters and visualize simulation results (Figure 1). The two plant models are based on the L-system formalism (L-py software, Boudon et al., 2012) and aim at capturing variation in morphogenesis and C/N metabolism specific to each functional group. In grasses for instance, differences in the self-regulation of leaf growth, tiller development and primary root initiation enable to account for growth habit ranging from longleaf forage species/cultivars to short-leaf turfs. Similarly, in the legume functional group, different growth habit in terms of shoot growth (erect/prostrate), branching, and ability to develop adventitious roots/shoots enable to distinguish between colonization strategies ranging from herbs perennating by formation of a single taproot (such as alfalfa or red clover) to clonal patches resulting from rhizome spreading (creeping lucerne) or rooted shoots at the soil surface (white clover). Dry matter accumulation, water and N requirements of each plant in a community are assumed driven by light interception. Figure 1. Graphical user interface and 3D plants illustrating an alfalfa plant population growing in a drying soil. Results and discussion The coupled soil-plant-atmosphere model is currently being assessed for both the behaviour of isolated plants (Figure 2), and the dynamic of plant populations in pure stands and in binary species mixtures. Parameters for morphogenesis were identified on contrasting alfalfa and perennial ryegrass cultivars. Morphogenetic response to light competition, defoliation and water stress were shown to be qualitatively consistent. Increased light competition resulted in reduced tillering/branching (and overall plant leaf area development), reduced root elongation rate (and root volume) and decreased root length density. An interaction between shoot morphogenesis (defining light competition ability) and defoliation regime existed for the two functional groups, morphotypes with short leaves (grass) / procumbent shoots (legume) being relatively favoured by more intense defoliations. Soil water availability affected first local root development and then plant water status, shoot growth and C assimilation/partitioning (Figure Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 243 2). Sensitivity to water stress was shown to be strongly dependent to developmental stages, and in particular to the rooting depth at the beginning of the stress. Figure 2. Illustration of the effect of four water regimes on the shoot and root morphogenesis of an isolated alfalfa plant. Quantitative model assessment is still ongoing and implies the comparisons of population productivity, size structure, and N content between simulated and experimental stands under various conditions of light and water availability. Conclusion In its present state, the “Virtual grassland” model intends to provide an integrated framework for analyzing trade-offs between traits in multispecific grasslands and for helping in the identification of rules to formulate the composition of mixtures, considering interactions with both biotic and abiotic environments. Ultimately, it may also contribute to predict the effects of plant interactions for light, water and nitrogen on grassland community dynamics. References Brisson N., Launay M., Mary B. and Beaudoin N. (2008) Conceptual basis, formalisations and parameterization of the STICS crop model. Quae Ed. 297 p. Boudon F., Pradal C., Cokelaer T., Prusinkiewicz P. and Godin C. (2012). L-Py: an L-System simulation framework for modeling plant development based on a dynamic language. Frontiers in Plant Science 3, art 76. Louarn G., Corre-Hellou G., Fustec J., Lô-Pelzer E., Julier B., Litrico I., Hinsinger P. and Lecomte C. (2010) Déterminants écologiques et physiologiques de la productivité et de la stabilité des associations graminéeslégumineuses. Innovations Agronomiques 11, 79–99. Pradal C., Boudon F., Nouguier C., Chopard J., Godin C. (2009). PlantGL: a Python-based geometric library for 3D plant modelling at different scales”. Graphical Models 71, 1-21. Prieto J.A., Louarn G., Pena J.P., Ojeda H., Simonneau T. and Lebon E. (2012) A leaf gas exchange model that accounts for intra-canopy variability by considering leaf nitrogen content and local acclimation to radiation in grapevine (Vitis vinifera L.). Plant, Cell and Environment 35, 1313-1328. Schwinning S. and Parsons A.J. (1996b) A spatially explicit population model of stoloniferous N-fixing legumes in mixed pasture with grass. Journal of Ecology 84, 815–826. Sinoquet H., Le Roux X., Adam B., Ameglio T. and Daudet F.A. (2001) RATP: a model for simulating the spatial distribution of radiation absorption, transpiration and photosynthesis within canopies: application to an isolated tree crown. Plant, Cell & Environment 24, 395-406. Soussana J.F. and Oliveira-Machado A. (2000) Modelling the dynamics of temperate grasses and legumes in cut mixtures. In: Lemaire G., Hodgson J., de Moraes A., Nabinger C., de Carvalho P.C. (eds) Grassland ecophysiology andgrazing ecology. CAB International, Wallingford, pp 169-190. Verdenal A., Combes D. and Escobar-Gutiérrez A.J. (2008) A study of ryegrass architecture as a self-regulated system, using functional–structural plant modelling. Functional Plant Biology 35, 911–924 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 244 Towards a new potassium fertilization recommendation in the Netherlands Holshof G. and van Middelkoop J.C. Wageningen UR Livestock Research , PO Box 65, 8200 AB Lelystad, the Netherlands Corresponding author: gertjan.holshof@wur.nl Abstract The current potassium (K) recommendation in the Netherlands has been based on experiments from the past. The plots in those experiments were harvested at high yields and high N levels. Due to lower annual N levels and new soil extraction methods, the K fertilization recommendation should now be adapted. To obtain new information, in 2011 and 2012 a twoyear grass cutting experiment on peat, sand and clay was set up with three N × eight K levels. The results show there is a need for about 50 kg K ha-1 for the first cut, and for every following cut the rate should be based on the expected amount of K-export in the grass. The effect of the first K application is still present in later cuts, but later cuts should also receive K fertilization based on expected K export by the yield. There is no interaction between N and K fertilization on the DM yield, but there is an interaction effect on K content. High K fertilization in combination with high N levels will lead to a higher K content of the grass. Keywords: potassium, K-fertilization recommendation, DM yield, N×K interaction, K-content Introduction The common potassium (K) recommendation is based on research from 1930-1960, from experiments with high nitrogen (N) levels ('t Hart and van der Pauw, 1942). The recommendation was derived from data obtained at a harvest level of 4-6 t DM ha-1. In present grassland management harvest took place at 1.5 (grazed plots) to 4 ton DM ha-1. Recently, new extraction methods for K in the soil have become available, which will probably lead to an adapted K recommendation. Due to limited N fertilization, an interaction between lower N levels, K fertilization on DM yield, and K content is also expected. To collect data to support the adaptation, two parallel experiments were set up. One field experiment was set up to test the effect of different K-fertilization strategies in combination with two N levels. This experiment is described in this article. The main objective is to get information about the K response and the N × K interaction on DM yield and K content of the grass. (Another parallel experiment collects data from several small field experiments to analyse the relation between different soil types, soil parameters and K fertilization on the DM yield: this will be described in a separate report (Bussink et. al., in prep)). The outcome of both field experiments will provide new insights for an adapted potassium fertilization recommendation. Material and methods In 2011 and 2012 a field experiment was set up in the Netherlands on three soil types: peat, sand and young marine clay. Five cuts of grass were harvested in both years. The first cut was at two yield levels: about 1700 kg DM ha-1 and about 3500 kg DM ha-1. The following four cuts were harvested around every 4-5 weeks. Two N levels (180 and 360 kg N ha-1 year-1) and a control (no N fertilizer) were combined with eight K strategies. The N and K fertilization is given in Table 1. The experiment was set up with two replicates, consisting of 3N × 8K = 24 plots on all soil types. The cuts were mown with a special mower for experimental research. The fresh yield was weighed and a sample was dried at 1030C to calculate the DM yield. The differences between the DM yield on the plots were statistically analysed using residual maximum likelihood (REML: Harville, 1977) with the Genstat program (15th edition). The year effect was set as a random effect in the REML analysis. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 245 Table 1. Nitrogen (N) and Potassium (K) fertilization (kg ha -1) per cut and total 1 NK / cut no 2 3 4 5 Total 0 N0 0 0 0 0 0 60 N1 40 40 20 20 180 120 N2 80 80 40 40 360 0 K0 0 0 0 0 0 50 K1 0 0 0 0 50 50 K2 33 33 33 33 183 50 K3 66 66 66 66 315 100 K4 100 100 100 100 500 149 K5 0 0 0 0 149 149 K6 33 33 33 33 282 149 K7 66 66 66 66 415 Results and discussion As expected, there was a strong (P <0.001) N response, which interacted with the soil type (the strongest effect on clay). There also was a significant overall K response (P <0.001) which also interacted with the soil type (hardly any effect on clay). There was no interaction effect on the DM yield between the N and the K application. This contrasts with the results of Widdowson et al. (1966), who found an interaction at the highest N level (314 kg N) in their experiment, although there was no interaction at N levels of 209 and 105 kg N ha-1 year-1. Wolton et al. (1968) also found an interaction effect of N×K, but there was a large variation and the effect was not consistent. They saw no effect of K fertilization when no N was applied. This was is in contrast to the results of our experiment. We found K response at all N levels, including control plots without N fertilization. The effect of K fertilization on DM yield per cut was significant (P<0.001). This means that if a higher yield per cut is required, K application should be higher. Figure 1. DM yield per cut and annual at eight K-fertilization strategies on sand and 360 kg N ha-1 Figure 1 represents the results as an average over 2011 and 2012 on the sand location and for an N application of 360 kg N ha-1 year-1. On the peat location similar results were found. On sand and peat an application of 50 kg K ha-1 is recommendable. A noticeable conclusion is that Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 246 the effect of K application in spring will have a (residual) effect on DM yield in later cuts. This effect can be seen comparing 149-4x0 K with the control without K. Not only in the first cut was DM yield higher, but also in later cuts that were not fertilized with K there was a higher yield. On peat and sand soils, later cuts respond significantly and positively to K application, which can be seen by comparing 50-4x40 with 504x66; the last strategy results in a higher annual DM yield. It seems that there is an upper threshold (maximum K application); a higher application did not result in a higher yield, but only in a higher K content. The effect of K fertilization on young marine clay was minimal, due to the type of soil. There was a slight difference (on average over all the plots and years, 500 kg) in DM yield in both years, but the effects were the same in 2011 and 2012. There is a significant effect of K fertilization on the first cut at all three soil types. The results of the experiment showed a higher K content in the grass when higher applications of K were given. High K contents should be avoided because of the risk of grass tetany (hypomagnesaemia) and too high K-applications also lead to losses. Based on this experiment (and the parallel experiment), an adapted K fertilization recommendation should be followed up. Conclusions An application of 50 kg K ha-1 for the first cut proved to be the most suitable recommendation. Due to the parallel experiment, a further differentiation will need to be made, that takes account of pH, soil type and K content (Bussink et al., in prep). After the first cut, some K fertilization is recommended, although a residual effect of the K applied in spring was recorded. The application should be equal to the expected amount of K exported in the yield. No interaction between K and N application was found on the DM yield, but there was an interaction effect on the K content of the grass. Higher applications than the expected amount of K exported by the crop should be avoided. Acknowledgements This project was funded and supported by the Dutch Dairy Board. References Bussink D.W, Holshof G. and van Middelkoop J.C. (2014, in prep). Interaction between N and K fertilisation; to an adapted K fertilisation recommendation (working title). Hart M.L. ‘t and Van der Paauw F. (1942) Kalibemesting op grasland. Directie van de Landbouw, Landbouwvoorlichtingsdienst, Med. No. 30 (in Dutch). Harville D.A. (1977) Maximum likelihood approaches to variance component estimation and to related problems. Journal of the American Statistical Association 72, 320-338. DOI: 10.2307/2286796. Widdowson F.V., Penny A. and Williams R.J.B. (1966) An experiment measuring effects of N, P and K fertilizers on yield and N, P and K contents of grazed grass. Journal of Agricultural Science, Cambridge 67, 121-128. Wolton K.M., Brockman J.S., Brough D.W.T. and Shaw P.G. (1968) The effect of nitrogen, phosphate and potash fertilizers on three grass species. Journal of Agricultural Science, Cambridge 70, 195-202. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 247 Increasing perennial ryegrass (Lolium perenne) yields using commercial bioinocula on a phosphate-limited soil Owen D.1, Williams A.P.1, Griffith G.W.2 and Withers P.J.A.1 1 School of Environment, Natural Resources and Geography, Bangor University, Gwynedd, United Kingdom, 2 Institute of Biological, Environmental and Rural Sciences, Aberystwyth, Ceredigion, United Kingdom. Corresponding author: afpe4d@bangor.ac.uk Abstract Bio-inocula are seen as a potential way of increasing crop yields and the availability of plant nutrients, principally phosphorus – a major limiting nutrient. Bio-inocula contain a variety of plant growth promoting micro-organisms (PGPM) such as phosphate solubilizing bacteria and mycorrhizal fungi. Although many commercial products contain multiple strains of fungi and bacteria, the possibility of interspecific competition between fungal components could reduce inoculum effectiveness. This study aimed to examine this and the effect of commercial inocula on the yield of grass (Lolium perenne) grown on a phosphate-limited soil. Bacterial inocula significantly increased dry matter yield (3.51 t ha-1) over controls. While not significant, grass treated with the mixed fungal treatment showed highest phosphate use-efficiency (0.47 g mg P-1). The single fungal treatment had the highest root colonization (93.3% of roots colonized). Overall soil fungal diversity and evenness showed no significant differences across treatments. While initial results suggest some interspecific competition between fungal components of mixed inocula, further analysis is required to establish if this impacts on grass yields. Introduction Increased agricultural production will be required to meet global demand for feed, fibre, bioenergy and food. This will require increased inputs of key limiting nutrients such as phosphorus (P), and/or better utilization of P sources applied and existing soil P reserves (Cordell and White, 2011). Soil micro-organisms improve plant nutrient availability (Barea et al., 2002) through processes of mineralization and solubilization of organic and inorganic soil phosphate, respectively. Identification of specific phosphate mobilizing fungi and bacteria has been exploited commercially through the development of bio-inocula. Commercial inocula use multiple species and genera to help overcome host fungal specificity and improve economic viability, e.g. Lolium perenne has been shown to be dominated by the glomus genera (Gollotte et al., 2004). Co-inoculation with different arbuscular mycorrhizal fungi (AMF) species is generally thought to be more effective due to functional complementarity within taxa (Koide, 2000), reducing plant investment costs to acquire P. However, functional redundancy may lead to negative effects between fungal taxons (Hepper et al., 1988) and growth depressions have been observed when multiple species were added (Mickelson and Kaeppler, 2005). The aim of this study was to establish the potential yield gains using commercial fungal inocula on a low-P soil; the use of fungal inocula in combination with low soluble phosphate fertilizer; and explore potential interspecific competition between fungal components of a mixed bioinoculum. Methods A field trial was established at Bangor University’s Henfaes Research Station in July 2013. Soil phosphate was measured at ~ 8mg kg-1, placing the soil at Olsen extractable P index of 0. A commercial bacterial suspension (Bac) (GlensideGroup, West Lothian), commercial fungal consortia of 5 fungal species (MF) (Glomus SP., GlensideGroup) and a single fungal species Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 248 (SF) (Glomus intraradices BEG 72, Plantworks Ltd. Kent) were applied to replicated (n = 9) trial plots (1.8 m × 2 m) laid out in a replicated control strip design. Applications were made at ten times the manufacturer’s recommendations, to improve subsequent detection rates (recommended application rates of MF and SF were 1 kg ha-1). Lolium perenne seeds were dressed with each treatment and applied using a seed drill (34 kg ha-1). The bacterial suspension (250 ml ha-1) was diluted with water (150 L ha-1) and applied with a hand sprayer. The phosphate treatments were applied to each strip as three treatments of TSP, RP and No phosphate (n = 3). N, K and phosphate treatments were applied in the spring (March 2014) as per RB209 recommendations (DEFRA, 2013) for a P index soil of 0. The first cut biomass, 8 weeks after seeding, was measured for yield (t ha-1), phosphate content (PO4-P mg kg-1) and fungal root colonization (Trouvelot et al., 1986) quantifying frequency (Mf) and intensity (Mi). Increased phosphate acquisition was investigated using a low and high soluble phosphate source (rock phosphate and triple super phosphate respectively). PUE (a measure of dry matter forage per unit phosphate acquired within the above ground biomass), was calculated and used to determine if bio-inocula improved phosphate acquisition. Ion Torrent DNA sequencing was used to assess any changes in biological communities after fertilization. Soil samples taken ten days after seeding were frozen (-80 °C), after freeze-dried samples were ground (< 1 mm). DNA extractions were conducted using the MoBio Laboratories (Solana, CA, USA) PowerSoil DNA extraction kit. An Ion Torrent PGMTM was used to sequence the extract. Sequences were clustered to 97% similarity using USEARCH and identified using the Ribosomal Database Project (RDP). A spreadsheet of phyla and quantities was produced for analysis, and used to calculate Shannon diversity index and equitability index. Results and discussion Results indicated yield gains with bio-inocula. The bacterial inocula significantly increased yield over the control, likely due to the nitrogen fixing species it contained, as the site received no N. The PUE of the bacterial inocula was similar to that of the other treatments, suggesting that the increased growth was not due to increased P acquisition but another limiting factor. Root colonization did increase with treatments; interestingly, the SF inocula colonized the roots more often and more intensely. This could be an indication of the potential interspecific competition within inocula formulation, the SF yielding ~20% more than the MF. That said, the PUE of the MF treatment was the highest of the 3 treatments and it is possible that this will factor in any potential yield gains after the phosphate treatments are applied. DNA analysis showed no real changes in diversity or evenness; however, soil samples were taken 1 week after seeding and, as such, any microbial additions may not have had time to establish. Soil samples 12 weeks after seeding may be more informative. Furthermore, the database is currently being refined to reduce the number of unidentified species. Therefore, future proposed analysis may reveal significant fungal community shifts, which the colonization and PUE data hinted at. References Barea J., Azcon R. and Azcon-Aguilar C. (2002) Mycorrhizosphere interactions to improve plant fitness and soil quality. Antonie van Leeuwenhoek 81, 343-351. Cordell D. and White S. (2011) Peak phosphorus: clarifying the key issues of a vigorous debate about long-term phosphorus security. Sustainability 3, 2027-2049. DEFRA (Department for Food, Environment and Rural Affairs) (2013) Fertiliser Manual 2013. Online at www.gov.uk/government/publications/fertiliser-manual-rb209 Gollotte A., van Tuinen D. and Atkinson D. (2004) Diversity of arbuscular mycorrhizal fungi colonising roots of the grass species Agrostis capillaris and Lolium perenne in a field experiment. Mycorrhiza 14, 111-117. Heffer P. and Prud’homme M. (2010) Fertilizer Outlook 2010-2014. 78th IFA Annual Conference Paris (France), 31 May – 2 June, 2010. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 249 Hepper C.M., Azcon-Aguilar C., Rosendahl S. and Sen R. (1988) Competition between three species of Glomus used as spatially separated introduced and indigenous mycorrhizal inocula for leek (Allium porrum L.). New Phytologist 110, 207-215. Koide R.T. (2000) Functional complementarity in the arbuscular mycorrhizal symbiosis. New Phytologist 147, 233-235. Mickelson S.M. and Kaeppler S.M. (2005) Evaluation of six mycorrhizal isolates for their ability to promote growth of maize genotypes under phosphorus deficiency. Maydica 50, 137-146. Trouvelot A., Kough J.L. and Gianinazzi-Pearson V. (1986) Mesure du taux de mycorhization VA d’un système radiculaire. Recherche de méthodes d’estimation ayant une signification fonctionnelle. In: V. Gianinazzi-Pearson and S. Gianinazzi (eds.) Physiological and Genetical Aspects of Mycorrhizae. INRA Press, Paris, pp. 217-221. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 250 Long-term effects of extensification regimes on soil and botanical characteristics of improved upland grasslands Pavlů V.1,2, Pavlů L.2 and Fraser M.D.3 1 Crop Research Institute, Department of Plant Ecology and Weed Sciences, Grassland Research Station Liberec, Czech Republic, CZ460 01 2 Czech University of Life Sciences Prague, Faculty of Environmental Sciences, Department of Ecology, Czech Republic, CZ165 21 3 Institute of Biological, Environmental and Rural Sciences, Aberystwyth University, Plas Gogerddan, Aberystwyth, United Kingdom, SY23 3EB Corresponding author: V.Pavlů - email: grass@volny.cz Abstract To reveal key factors necessary for restoring previously intensively managed grasslands a longterm extensification experiment was established in Wales in 1994. The treatments were: sheep grazing (± lime); hay cut only (± lime); hay cut and aftermath sheep grazing (± lime); and sheep grazing plus fertilizer and lime application as a control. Nutrient concentrations in the soil together with plant species composition were assessed in 2012. The treatments which received lime had a higher pH and plant-available concentrations of Ca and Mg in the soil, whereas higher plant-available concentrations of P, K and organic C were recorded in the control treatment fertilized by NPK. Management was a key driver in terms of diversity of vascular plant species. Multivariate analysis identified three groups of treatments with similar plant species composition: control and grazed treatments as the first group; hay cut treatments as the second group; and hay cut plus aftermath grazing as the third group. Lime application had relatively little effect on plant species composition, but it decreased the hemiparasitic species Rhinanthus minor which can initiate further steps within the extensification process. The most effective management for grassland biodiversity restoration was hay cutting with aftermath grazing. Keywords: extensification, grazing, cutting, nutrients, vegetation Introduction Long-term application of NPK fertilizers increases plant productivity and soil nutrient concentrations but also reduces plant species richness, and this can result in the dominance of largely species-poor grasslands with little conservation value. High concentrations of plantavailable P in the soil are particularly associated with low species richness and dominance of highly productive species (Tallowin and Smith, 2001). The restoration of species-rich grassland on previously agriculturally improved grasslands has a number of specific abiotic and biotic constraints, including high nutrient contents in the soil and a lack of desirable species. In order to achieve restoration of species-rich grassland it is necessary to identify key stages and likely timescales, as well as practical management techniques to be undertaken. To explore these issues further, a long-term extensification experiment testing different management regimes was set up at the Pwllpeiran Upland Research Centre in 1994 (the Brignant plots) (Defra, 2006). Preliminary results showed that the cessation of fertilizer use and imposing of extensive management practices allowed moderately species-rich grassland communities to develop within a 10-year timescale. In this study we analyse extensification management effects on soil nutrients and plant species composition after18 years of treatment imposition. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 251 Materials and methods The experiment was established on previously improved and consequently intensively managed pasture situated within the Cambrian Mountains at an altitude of 310 m a.s.l. Grass species dominated in the sward in 1994; particularly Lolium perenne, at 58% cover. Soils at the site are free-draining typical brown podzolic soils, with the area receiving a mean annual rainfall of 1770 mm. The treatments applied were: i) sheep grazing with fertilizers and lime application as control (CO); ii) hay cutting followed by aftermath sheep grazing, with (HGL+) and without (HGL-) lime application; iii) hay cutting only, with (HL+) and without (HL-) lime application; iv) sheep grazing, with (GL+) and without (GL-) lime application. The control treatment was a continuation of the previous relatively intensive management practice of an annual application of 60 kg N fertilizer and 30 kg P fertilizer ha-1. The lime treatments received a single application of lime in 1998 with the intention of maintaining a soil pH of 6.0. Treatments are replicated three times in a randomized block design with individual plots 0.08 ha (hay cut only) or 0.15 ha (grazed) in size. Fifteen individual soil cores were taken to a depth of 0-7.5 cm from randomly located areas from within each plot in July 2012. Plant-available Ca, K, Mg and P were extracted by the Mehlich III method (Mehlich, 1984) and then determined by ICP-OES. Total N was analysed by the Kjeldahl method and organic C by means of colorimetry. Visual percentage cover of vascular plant species was estimated in 10 randomly located (0.4 m-2) quadrats per plot in July 2012. Redundancy analysis (RDA) in the CANOCO 4.56 program was used to evaluate multivariate vegetation and soil nutrient data. Results and discussion Management treatment explained 35.5% (P<0.001) and 60.6 % (P<0.001) of the variability of the soil nutrient content in the first and all axes, respectively. The first axis of RDA displayed a gradient of pH in the soil (Fig. 1a). The treatments which received lime application in 1996 had higher pH and plant-available content of Ca and Mg in the soil, whereas higher plantavailable contents of P, K and organic C were found in the CO treatment fertilized by NPK. These results concur with studies which have shown that residual effects of former fertilizer treatments can be identified even when applications were short term (Pavlů et al., 2011). Treatment effects explained 59.8% (P<0.007) and 81.1% (P<0.001) of the variability in plant species composition in the first axis, and all axes, respectively. RDA analysis of the vegetation data revealed that the first axis displayed a gradient of the defoliation management (Figure 1b). Three groups of treatments with similar plant species composition were recognized on the ordination diagram based on RDA analysis of data collected in 2012: CO, GL+ and GLtreatments as the first group; HL+ and HL- treatments as the second group; and HGL+ and HGL- as the third group. The first group was species poor (number of species <8) and promoted especially grasses that are adapted to frequent defoliation (e.g. Poa pratensis, Agrostis capillaris, Lolium perenne, Festuca rubra) and the nitrophilous species Urtica dioica. In the second group the species sensitive to defoliation were more abundant (e.g. Holcus mollis, Alopecurus pratensis, Cirsium palustre). The third group was the most species rich (number of species >13) and included forbs species adapted to frequent defoliation (e.g. Hypochoeris radicata, Potentila erecta, Taraxacum sp., Crepis capillaris) but which need hay management to enable seed reproduction. Rhinanthus minor was the only species suppressed by previous lime application in extensively managed treatments. This species had not been recorded previously within the plots and is important as it can suppress existing vegetation and consequently accelerate the introduction of the other competitively weak species occurring in the surrounding areas. Overall it appears that management treatment applied was the key driver Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 252 in terms of diversity of vascular plant species. Similar results were found in extensification experiments in Germany (Hejcman et al., 2010; Pavlů et al., 2011). Figure 1. Ordination diagram showing the results of RDA analysis: a) nutrient content in the soil (left); b) plant species composition (right). For treatment abbreviations see Materials and methods section. Species abbreviations are based on first four letters of genus and first four letters of species name. Conclusion Although former lime and NPK fertilizer applications were still detectable in the soil, their effect on botanical composition was relatively small. The key effect on plant species composition was the management treatment applied. The most effective management for grassland biodiversity restoration was hay cutting with aftermath grazing. Acknowledgements At the time of the study the plots were supported by the Countryside Council for Wales. Data collection and paper preparation were supported by the Stapledon Memorial Trust, the Czech Science Foundation (Project No 521/08/1131) and ESF & MEYS (CZ.1.07/2.3.00/30.0040). References Defra 2006 Managements to achieve botanical diversification of improved grassland by natural colonisation. BD1452. http://defra.gov.uk Hejcman M., Schellberg J. and Pavlů V. (2010) Long-term effects of cutting frequency and liming on soil chemical properties, biomass production and plant species composition of Lolio-Cynosuretum grassland after the cessation of fertilizer application. Applied Vegetation Science 13, 257-269. Pavlů V., Schellberg J. and Hejcman M. (2011) Cutting frequency vs. N application: effect of a 20-year management in Lolio-Cynosuretum grassland. Grass and Forage Science 66, 501-515. Tallowin J.R.B. and Smith R.E.N. (2001) Restoration of a Cirsion-Molinetum to an agriculturally improved site: a case study. Restoration Ecology 9, 167-178. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 253 How do pre-sowing disturbance and post-establishment management affect restoration progress in ex-arable calcareous grassland? Wagner M.1, Bullock J.M.1, Meek W.R.2, Walker K.J.3, Stevens C.J.4, Heard M.S.1 and Pywell R.F.1 1 NERC Centre for Ecology & Hydrology, Benson Lane, Wallingford, United Kingdom, OX10 8BB 2 60 Midfield Road, Humberston, Grimsby, United Kingdom, DN36 4TH 3 Botanical Society of Britain and Ireland, 97 Dragon Parade, Harrogate, United Kingdom, HG1 5DG 4 Lancaster Environment Centre, Lancaster University, Lancaster, United Kingdom, LA1 4YQ Corresponding author: mwagner@ceh.ac.uk Abstract Restoration of semi-natural calcareous grassland needs to overcome dispersal limitation and microsite limitation. The effectiveness of sowing a mix of ten target species in combination with various pre-sowing disturbance techniques and post-establishment management regimes in overcoming these limitations, and in promoting vegetation change towards the restoration target was tested in a field experiment. Our results illustrate the importance of bare ground for introducing target species of calcareous grassland into species-poor grassland, and the slowness of colonisation by additional target species from outside even when a site is situated close to high-quality grassland that could serve as source for additional target species. Keywords: bare ground, compositional similarity, ecological restoration, target species Introduction Species-rich lowland calcareous grassland in the UK has markedly declined in extent over the past century. To reverse this decline, efforts are made to restore such grassland through grassland creation and the diversification of species-poor grassland (Walker et al., 2004). Grassland restoration has to overcome two key limitations (Bakker and Berendse, 1999). Dispersal limitation occurs because unassisted re-colonization by target species tends to be slow, and depends on local presence of existing high-quality grassland. Microsite limitation, a lack of small-scale sites suitable for seed germination and seedling establishment, can be a serious obstacle because soil fertility is often considerably raised, and existing vegetation tends to be highly productive and dense. Generalist grassland species which tend to be competitive and capable of rapid clonal spread usually perform better under such conditions than specialist species that are adapted to low soil fertility and that depend more strongly for regeneration on seedling establishment in seasonally available gaps (Pywell et al., 2003). We investigated in a 4-year field experiment how conditions for introduction and short-term persistence of specialist species of calcareous grassland can be provided during restoration. Based on annual vegetation surveys, and using reference data from local calcareous grassland, this paper compares restoration progress among various experimental treatments. Materials and methods A split-plot experiment was set up in spring 2008 on species-poor grassland in the Pegsdon Hills (51° 57' N 0° 23' W) at the NE edge of the Chiltern Hills, SE England. The grassland was created in 1993 on previously arable land by over-sowing with a low-diversity grass mix and subsequent sheep grazing. In 2008, vegetation at the site mainly consisted of Arrhenatherum elatius, Agrostis stolonifera, Poa trivialis, Medicago lupulina and Trifolium repens. Close to the restoration site there were areas of species-rich calcareous grassland. Twelve main plots of 75 m × 50 m were set up in four replicate blocks to compare different management regimes, Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 254 and at the centre of each plot, a grid of 5 m × 5 m cells was laid out to implement different types of pre-sowing disturbance at the sub-plot level. Allowing for guard rows between experimental sub-plots, disturbance treatments were assigned to grid cells at random. Prior to disturbance, herbage was cut and removed. Disturbance treatments were (1) undisturbed control; (2) band-spraying with glyphosate to kill c. 50% of the vegetation; (3) power harrowing to create c. 70-80% bare ground; and (4) ‘ridge-and-furrow’, i.e. creation of a 1.5 m wide × 0.4 m high ridge by two-directional ploughing. After disturbance, all experimental subplots were over-sown on 17 May 2008 with a seed mixture containing ten species of calcareous grassland (Bromopsis erecta, Campanula glomerata, Carex flacca, Filipendula vulgaris, Helianthemum nummularium, Hippocrepis comosa, Pimpinella saxifraga, Stachys officinalis, Succisa pratensis and Thymus pulegioides). Sowing rates varied among species, ranging from 30 to 120 seeds m-2. In 2008, the site was uniformly managed with a hay-cut in late July and aftermath cattle grazing. From 2009 onwards, three management regimes were implemented at main-plot level: (1) summer hay-cut followed by autumn cattle grazing; (2) spring grazing with sheep, followed by summer hay-cut and autumn cattle grazing; and (3) spring grazing with sheep, followed by summer and autumn cattle grazing. From 2009 onwards, cover of vascular plant species was recorded annually before the summer hay-cut in two permanent 1 m x 1 m quadrats per sub-plot. Prior to statistical analyses, data from both quadrats was averaged. Quadrat data (n = 39) from eight calcareous grasslands located within 25 km from our site was used to define the restoration target. To explore vegetation change in across treatments in relation to this target, Non-metric Multidimensional Scaling (NMDS) based on the Relative Sorensen distance measure was carried out using PC-ORD V6.08 (McCune and Mefford, 2011). Similarity of vegetation in experimental treatments with target vegetation was further investigated with a repeated-measures linear mixed model, using PROC MIXED in SAS V9.3 (SAS, 2011). For this analysis, Relative Sorensen distances between each sub-plot in each year and quadrats from reference grassland were converted into percentage similarities, and within each year, the 39 similarity values per experimental sub-plot were averaged and log-transformed. Management regime, pre-sowing disturbance, year, and the various interactions between these entered the analysis as fixed factors, and year as repeated-measures factor. Blocks and main plots nested within blocks were specified as random effects. A similar analysis was carried out for total cover data of sown species, which was also log-transformed prior to analysis. The most appropriate covariance structure from a list including compound symmetric and various autoregressive structures was selected on the basis of the Akaike Information Criterion. Results and discussion According to the NMDS biplot (Figure 1), between 2009 and 2011, vegetation composition in experimental treatments has gradually become more similar to target grassland composition. Mixed-model analysis shows that rate of increase in similarity over time with the calcareous grassland target vegetation is significantly affected by management regime and initial disturbance (Figure 2A). This increase was more pronounced in ridge-and-furrow sub-plots and in harrowed sub-plots. Regarding management regimes, it was less pronounced under sheep grazing in spring in combination with summer and autumn cattle grazing. However, progress was generally slow, and after three years, even in the most successful treatment combinations, vegetation was only about 10% similar to the target. Increase in total cover of sown species (Figure 2B) varied similarly between treatments, suggesting that the observed increase in similarity to target community composition was largely driven by changes in sown species cover. A closer look at the vegetation data confirms that levels of colonisation by additional species were very low, although there were shifts in abundance of unsown resident Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 255 species. Slow colonisation in spite of proximity of high-quality source grassland was also found by Hutchings and Booth (1996). Figure 1. NMDS biplot of vegetation in experimental treatments 2009-2011 in relation to calcareous grassland (= target). Figure 2A: Similarity of experimental treatment vegetation to target vegetation. Figure 2B: Total cover of sown species. (HC: hay cut + autumn cattle; SCC: spring sheep + hay cut + autumn cattle; SHC: spring sheep + hay cut + autumn cattle.) Conclusion Our results highlight the importance of bare-ground creation for successfully introducing latesuccessional target species of calcareous grassland into species-poor existing swards. They also illustrate that colonization from nearby species-rich grassland is slow when unassisted, and that, in order to achieve reasonably fast restoration of species-rich calcareous grassland, higher levels of intervention may be required, involving either active introduction of a wider range of target species, which could take place in a staged manner (Pywell et al., 2003), and/or grazing regimes targeted at assisting processes of species immigration from nearby species-rich source grassland (Poschlod et al., 1998). References Bakker J.P. and Berendse F. (1999) Constraints in the restoration of ecological diversity in grassland and heathland communities. Trends in Ecology and Evolution 14, 63-68. Hutchings M.J. and Booth K.D. (1996) Studies on the feasibility of re-creating chalk grassland vegetation on exarable land. I. The potential roles of the seed bank and the seed rain. Journal of Applied Ecology 33, 1171-1181. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 256 McCune B. and Mefford M. J. (2011) PC-ORD. Multivariate Analysis of Ecological Data. Version 6.08. MjM Software, Gleneden Beach, OR, USA. Poschlod P., Kiefer S., Tränkle U., Fischer S. and Bonn S. (1998) Plant species richness in calcareous grasslands as affected by dispersability in space and time. Applied Vegetation Science 1, 75-90. Pywell R.F., Bullock J.M., Roy D.B., Warman L., Walker K.J. and Rothery P. (2003) Plant traits as predictors of performance in ecological restoration. Journal of Applied Ecology 40, 65-77. SAS (2011) SAS Release 9.3 for Windows. SAS Institute Inc., Cary, NC, USA. Walker K.J., Stevens P.A., Stevens D.P., Mountford J.O., Manchester S.J. and Pywell R.F. (2004) The restoration and re-creation of species-rich lowland grassland on land formerly managed for intensive agriculture in the UK. Biological Conservation 119, 1-18. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 257 Sown biodiverse pastures as a win-win approach to reverse the degradation of Mediterranean ecosystems Teixeira R.F.M.1, Proença V.2, Valada T.2, Crespo D.3 and Domingos T.2,4 1 University of Antwerp, Department of Biology, Research Group of Plant and Vegetation Ecology, Universiteitsplein 1, B-2610 Wilrijk, Belgium 2 IN+, Centre, for Innovation, Technology and Policy Research, Environment and Energy Scientific Area, DEM, Instituto Superior Técnico, University of Lisbon; Avenida Rovisco Pais, 1, 1049-001 Lisbon, Portugal 3 Fertiprado, Seeds and Nutrients, Ltd.; Herdade dos Esquerdos, 7450-250 Vaiamonte, Portugal 4 Terraprima, Environmental Services, Ltd.; Quinta de França (Borralheira), Caria, 6200-710 Teixoso, Portugal Corresponding author: vania.proenca@ist.utl.pt Abstract The system of ‘sown biodiverse permanent pastures rich in legumes’ started being developed in Portugal in the second half of the 1960s as an economically rational strategy to increase grassland productivity, by sowing mixes of up to 20 species/cultivars of legumes and grasses. Compared to natural pastures, the resulting semi-natural system provides higher yields of better quality pasture, allowing an increase of sustainable stocking rates, with environmental cobenefits. In this work we present a conceptual model of the environmental and economic effects of this system when compared with an alternative (natural pastures), complete with a summary of evidence supporting the claim that this system is an economic and ecological win-win solution that addresses many of the causes of land degradation in the Mediterranean while also recovering soil condition, ecosystem functions and services. Keywords: biodiversity engineering, ecosystem restoration, carbon sequestration, semi-arid land, sustainable intensification, soil organic matter Introduction ‘Sown Biodiverse Permanent Pastures Rich in Legumes’ (SBPPRL) is a system of engineered pastures that uses biodiversity as a leverage for pasture productivity. The main rationale behind this system is the introduction of species or varieties specific of arid and semi-arid regions, either absent or in lesser percentage in spontaneous grasslands (as, for example, some varieties of legumes), with the aim of establishing a functioning ecosystem with complementary ecological niches, and which is more productive than less-diverse grasslands. The higher productivity and feed quality of SBPPRL are paired by several other benefits to farmers, but also to society. SBPPRL have been described as a win-win solution (environmentally and economically) that combines the private interests of farmers with the public services provided, such as carbon sequestration through increases in soil organic matter (SOM), which in turn may help reverse the degradation of Mediterranean arid and semi-arid ecosystems. For example, the system won in 2013 the ‘World You Like Challenge’ of the European Commission (http://world-you-like.europa.eu/) for the best solutions to address climate change. It was described by the European Commissioner for Climate Action as ‘a perfect example of how practical solutions for climate action can also save money and create jobs and growth’. Since 2009, the Portuguese Carbon Fund (a financial instrument created to help the country comply with Kyoto targets) has been supporting the installation and maintenance of SBPPRL through a system of payments for carbon sequestration. More than 50 000 ha were sown under this programme, which represents an estimated amount of 1.54 million t of sequestered CO2 between 2008 and 2014. In addition, between 1996 and 2008 carbon Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 258 sequestration by SBPPRL was estimated as 3.5 million tons of CO2, at an approximate rate of 5 t of CO2 per hectare (Teixeira, 2010). The baseline system Grazing has been a central activity in the Mediterranean, shaping these ecosystems for millennia. Fire, clearing of shrubs and reducing forest density have been employed to maintain or reverse the process of vegetation succession (Bugalho et al., 2011). Today, these Mediterranean ecosystems occupy vast areas where soils are shallow, poor in organic matter and nutrients, and much exposed to soil erosion (Van-Camp et al., 2004). This marginal land has been used as grazing land to produce livestock, in what we designate here as natural pastures (NP). NP areas have historically been where SBPPRL are installed, and for comparison purposes we will consider them as the baseline alternative. Ecological and economic effects of SBPPRL The effects of SBPPRL relative to NP are shown in Figure 1, and can be qualitatively summed up as follows. Increased productivity in SBPPRL allows a sustainable increase in animal carrying capacity. Animals graze the plants, which have an annual life cycle. High plant productivity implies increased atmospheric carbon capture through photosynthesis. Part of the biomass produced is stored in soils due to the high density of yearly-renewed roots. Storage occurs in the form of non-labile (lower decomposition rates) soil organic carbon (SOC), which is part of the SOM pools. The increase of SOM improves soil nutrient availability and water holding capacity, thus increasing plant productivity and reducing surface runoff of water, which in turn decreases sediment loss and soil erosion. This effect is further strengthened by avoided soil mobilization due to seed bank persistence, which can last for more than 20 years in well managed pastures. In addition, decreasing water runoff and soil erosion contribute to reduce silting, eutrophication and contamination of surface waters outside the sown plot. Nitrogen fixation by legumes eliminates the need for synthetic nitrogen fertilizers, the production of which is highly energy demanding, and therefore responsible for high greenhouse gas emissions. Finally, the sustainable increase of the stocking rate and reduced fertilizer use, and the remuneration for ecosystem services, such as carbon sequestration, increase the economic viability of the farms while addressing both drivers of ecosystem degradation. Quantitative data already exist for some of the causal links in Figure 1. The first experiment to assess the effects of SBPPRL gathered data from rainfed pastures (SBPPRL and NP) on eight farms from 2001 to 2005 (Teixeira, 2010; Teixeira et al., 2011). Each plot’s soil and landscape type was approximately homogeneous, in terms of soil and previous use (Teixeira et al., 2011). Dry matter (DM) productivity is systematically 50 to 100% higher for SBPPRL, increasing every year and reaching 2-6 t DM per hectare (Teixeira, 2010). Average stocking rate is 1.0 Cattle Unit (1 CU ≈ 1 adult cow) per hectare and per year in SBPPRL and 0.43 CU ha-1 year-1 in NP. SOM concentration increases on average by 0.21 percent points (pp) per year in SBPPRL, during the first 10 years, and by 0.08 pp per year in NP (Teixeira et al., 2011). The 0.21 pp increase in SOM is equivalent to the sequestration of 6.5 t of CO2 ha-1 year-1 in the first 10 cm of topsoil and during 10 years after installation (Teixeira, 2010). The overall greenhouse gas (GHG) balance of SBPPRL, considering CO2 emissions from limestone application to correct soil acidity, N2O emissions from nitrogen cycle, CH4 emissions from livestock and CO2 sequestration by plants, is of 1.55-2.13 t of CO2 ha-1 year-1 (Teixeira, 2010), which verifies its role as a carbon sink. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 259 Figure 1. Relative effects of SBPPRL and NP. Arrows indicate causal relation. Green emoticons are the effects of SBPPRL and brown emoticons of NP. Direct ecologic effects (at the farm) are represented in pink, indirect effects in yellow and socioeconomic effects in blue. Two colour blends indicate two types of effect. Conclusion SBPPRL are sustainable semi-intensive systems for meat production applied in Portugal that maximize the bundle of services provided by grasslands, and contribute to counteract degradation drivers. SBPPRL provide better results than NP for most environmental performance indicators studied to date, while also delivering socioeconomic benefits. References Bugalho M.N., Caldeira M.C., Pereira J.S., Aronson J. and Pausas J.G. (2011) Mediterranean cork oak savannas require human use to sustain biodiversity and ecosystem services. Frontiers in Ecology and the Environment 9, 278–286. Teixeira R.F.M. (2010) Sustainable land uses and carbon sequestration: the case of sown biodiverse permanent pastures rich in legumes. PhD dissertation thesis, IST, Technical University of Lisbon, Lisbon. Teixeira R.F.M., Domingos T., Costa A.P.S.V., Oliveira R., Farropas L., Calouro F., Barradas A.M. and Carneiro J.P.B.G. (2011) Soil organic matter dynamics in Portuguese natural and sown Grasslands. Ecological Modelling 222, 993-1001. Van-Camp L., Bujarrabal B., Gentile A-R., Jones R.J.A., Montanarella L., et al. (2004) Reports of the Technical Working Groups Established under the Thematic Strategy for Soil Protection. EUR 21319 EN/3, Office for Official Publications of the European Communities, Luxembourg, 872 pp. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 260 Fauna-flora relationships within improved upland grasslands managed under alternative extensification regimes Rosa García R.1 and Fraser M.D.2 1 Servicio Regional de Investigación y Desarrollo Agroalimentario (SERIDA), P.O. Box 13, 33300 Villaviciosa, Asturias, Spain 2 Institute of Biological, Environmental and Rural Sciences (IBERS), Gogerddan, Aberystwyth, Ceredigion SY23 3EB, UK Corresponding author: entomteam@hotmail.com Abstract The maintenance of plant communities in grasslands depends on regular disturbances, but their sustainable management demands the integration of both floral and faunal components. This study tested the short and long-term impacts of sheep grazing and/or hay cutting (both with and without lime addition) on arthropod foliage fauna within Cambrian Mountain grasslands. Data were collected from replicate plots during the summer of 2013; twice before hay cutting and once just after. Fauna were sampled with sweep nets, and complementary data on sward height, flower numbers, and percentage of forbs and grasses were collected. Total fauna abundance was higher in grazed plots (as Symphypleona flourished there), while family richness was higher in swards managed for hay cuts. However, taxa-specific responses occurred. Certain Heteroptera abounded in grazed areas whereas Diptera or certain Coleoptera were linked to hay cut areas. Short-term effects reflected phenological changes (e.g. in Symphypleona or Cantharidae) and fauna reductions after hay cutting, when mostly Diptera remained. Fauna communities were associated with flower-rich and forb-dominated areas, although certain taxa favoured grazed and grass-dominated ones. Both hay cutting and hay cutting plus aftermath grazing were linked to higher levels of biodiversity by providing more suitable environmental conditions for the fauna. Keywords: arthropoda, sheep grazing, hay cut, management Introduction Grassland ecosystems depend on regular disturbances to prevent vegetation succession and maintain a particular plant community, although whenever simultaneous sustainability is expected, the strategies must integrate both floral and faunal components (WallisDeVries et al., 2002). Common management strategies for improved grassland in the UK uplands include grazing and/or cutting, plus fertilization and lime addition. European grasslands agrienvironment schemes encouraged modification of such strategies to arrest biodiversity loss, but their effectiveness has been challenged (Kleijn et al., 2003). The aim of the current study was to test the long-term and short-term effects of such management options on foliage arthropod fauna. Materials and methods The study was based on long-term plots at the Pwllpeiran Upland Research Centre. The replicated plots (known as the Brignant plots) were set up in 1994 on a previously improved pasture dominated by grass species in the Cambrian Mountain ESA. The experimental design consisted of a randomized block design with three blocks and four grassland management regimes imposed on individual plots 0.08 ha (hay cut only) or 0.15 ha (grazed) in size. The control (CO) was the continuation of normal intensive management (i.e. limed, fertilized and continually grazed by sheep). The treatments G (sheep grazing), H (hay cut) and HG (hay cut and aftermath grazing) represented different options for extensification, each one with (+) and without (-) lime addition. Fauna and flora data were recorded three times during the summer Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 261 of 2013: 18 July (S1), 16 August (S2) and 28 August (S3). The hay was harvested on 26 August. Foliage arthropods were captured using a sweep net along random transects (two or three per plot, depending on size); with fifty sweeps made per transect. Fauna were identified to order or family level (Coleoptera, Hemiptera and Araneae). Sward surface height, the percentage of forbs and grasses, and the number of flowers were also recorded for each plot at the time of sampling. Mixed-model ANOVA was used to test for treatment effects in the fauna and flora over time and linear ordination method redundancy analysis (RDA) to test the responses of arthropod communities to the management treatments and plant variables. Results and discussion A total of 62260 arthropods were collected and assigned to 13 orders. Symphypleona, Diptera, Hemiptera and Hymenoptera accounted for 33%, 23%, 21% and 14% of the catches respectively. Comparing the three periods, the total abundance changed over time and between treatments (P < 0.001). The catches in S2 were higher than in S1 and S3 (P < 0.001); with the lowest records in S3 reflecting the drastic decline in fauna on H and HG sites following the hay cut. Total fauna abundance was lower in G than in CO and H (P < 0.01) and lime addition had no effect on abundance and diversity. Focusing on S1 and S2, short-term (seasonal) and longterm (treatment effects) were observed. The total abundance was lower in G than in HG (P < 0.1), H and CO (P < 0.05). It increased in S2 as Symphypleona peaked in CO; in fact, this order was more abundant in CO than in the other treatments during both periods (P < 0.05). If Symphypleona was excluded from the analyses the global patterns were consistent along time: higher fauna catches in H and HG (P < 0.05) and G (P=0.055) than in CO. Taxa-specific responses were observed. Sap-sucking insects from Hemiptera abounded more in H than in HG and G (P < 0.001). Within this order, relevant families responded differently. Miridae were more abundant in H than in G and HG (P < 0.001), and Aphididae in H than in G (P < 0.001), whereas Cicadellidae peaked in G compared to H and HG (P < 0.05), and was also more abundant when lime was added (P < 0.001). Diptera flourished in HG (P < 0.01) and H (P < 0.05) compared to G. Within this order, a relevant group in pollination and pest control, Syrphidae, was clearly linked to HG and H sites than to the others (P < 0.001) whereas Tipulidae, a frequent food resource for birds, was indifferent to the treatments. The order Coleoptera was more abundant in H sites than in the rest (P < 0.001), largely due to the preference of flower visiting Cantharidae (P < 0.001) for these sites during S1. Araneae was globally indifferent to the treatments, although the catches of common inhabitants of flowers and leaves like Thomisidae peaked in H and HG compared to CO and G (P < 0.001 and P < 0.01 respectively), and also in sites with lime addition (P < 0.01), whereas agrobiont Linyphiidae did not differ between treatments. Family richness differed between treatments and was higher in H than in CO (P < 0.05) whereas Shannon index was higher in G than in CO (P < 0.05). Regarding the differences between S2 and S3 (i.e. before and after hay cut), total abundance changed between periods (P < 0.001), increasing sharply in CO, maintaining in G and dropping in H and HG during S3 (with no differences between these treatments) compared to CO and G (P < 0.001). Fauna richness and Shannon index was reduced in hay cut sites (where mostly Diptera dominated) compared to grazed ones (P < 0.001), where no differences were observed. Regarding the importance of vegetation variables for the arthropod communities, the model which included all of these as environmental variables and the replicate blocks as covariables was significant (P < 0.001) when tested separately for each period. In the long term, arthropod communities were more commonly linked to taller, flower-rich areas with higher percentage of forbs, although certain groups were associated with grass-dominated areas (Figure1a). By contrast, after the hay cut, as vegetation was removed in the areas with a greater percentage of forbs (H and HG), fauna was linked mostly to grass-dominated sites (Figure 1b). Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 262 0.6 (b) 0.8 (a) Sca flower Chrys Forbs Aph Thy The Can Nit Del Acr Antho Aph Cic Grass Smi StaPsyl Ant Lin Pti Lat Tet Sco Sil Lyc Cer -0.6 Tho Chry Nab Cur Ara Mir Phal Cix Tip Sca Forbs Coc Tip Hyd Lyg Tin -1.0 Sta Tab Cara Cic The Syr Height Cer Chrys Tet 1.0 Smi Thy Tho Nit Cur Acr Coc Del Cryp Rha Lat flowers Syr Lyc Lin Mir Height -0.6 Rha Grass -1.0 1.0 Figure 1: RDA analysis for flora-fauna relationships during the second (a) and third (b) periods. Blue lines correspond to the records of arthropod families and red ones to plant data. Standardized by sample norm and blocks included as covariables. Unrestricted permutations. 0.31 and 0.43 for all canonical axes in (a) and (b) respectively. Eigenvalues in axis one were 0.18 and 0.43 for (a) and (b) respectively. Conclusions Both short-term and long-term consequences of management options must be considered as they might lead to contrasting conclusions. In the long term, management for hay was linked to increased fauna richness through the provision of greater variability of plant resources, but in the short term, hay-cut areas suffered the highest fauna losses. Understanding the effect of land use on grassland communities contributes to the conservation of these often highly diverse habitats. Acknowledgements This study was funded by a Travelling Fellowship from the Stapledon Memorial Trust. References Kleijn D. and Sutherland W.J. (2003) How effective are European agri-environment schemes in conserving and promoting biodiversity? Journal of Applied Ecology 40, 947–969. Wallis de Vries M.F., Poschlod P. and Willems J.H. (2002) Challenges for the conservation of calcareous grasslands in northwestern Europe: integrating the requirements of flora and fauna. Biological Conservation 104, 265–273. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 263 Sewage sludge fertilization effects on Q. rubra and pasture production and flora biodiversity Mosquera-Losada M.R., Rigueiro-Rodríguez A. and Ferreiro-Domínguez N. Crop Production Departament, Escuela Politécnica Superior, Universidad de Santiago de Compostela 27002-Lugo, Spain. Corresponding author: mrosa.mosquera.losada@usc.es Abstract Fertilization is usually applied to increase land productivity, but this technique also affects biodiversity in silvopastoral systems. This study aims at evaluating the effect of different doses of sewage sludge on tree growth, pasture production and biodiversity during three years, at eight years after establishment of a silvopastoral system with Quercus rubra. Improving soil fertility increases grassland biodiversity until tree canopy cover is not complete. Keywords: agroforestry, sowing, afforestation, anaerobic digestion, composting, pelletization Introduction Fertilization is usually applied to increase land productivity, but this technique also affects biodiversity in a spatial and temporal scale in grasslands. The introduction of a tree in a silvopastoral system also modifies biodiversity as it initially generates heterogeneity (unshaded and shaded areas) but later on, when tree canopy completely covers the understorey, a new homogeneous situation is created. Broadleaf trees are generally used in silvopastoral systems, Quercus rubra being of high interest due to its high timber quality and tree growth rate. The combination of trees and grassland is usually more efficient in the uptake of nutrients from soil, the tree being a good tool to use and recycle nutrients applied to the understorey with fertilizers. The use of sewage sludge as fertilizer has been promoted in the EU in recent years due to the high nutrient content. Tree development, as well as soil fertility improvement, could modify the relationship among the different understorey species. This study aims at evaluating the effect of different doses of sewage sludge on tree growth, pasture production and biodiversity during three years, at eight years after the establishment of a silvopastoral system with Q. rubra. Materials and methods The experiment was established in Baltar, A Pastoriza (Lugo, Galicia, northwest Spain) at an altitude of 475 m above sea level. Pasture was sown with a mixture of Dactylis glomerata L. var. Artabro (12.5 kg ha-1), Lolium perenne L. var. Brigantia (12.5 kg ha-1) and Trifolium repens L. var. Huia (4 kg ha-1) in December 2004. Plants of Q. rubra Franco were planted at a density of 434 trees ha-1 after pasture sowing in 2001. The experimental design was a randomized complete block with three replicates and four treatments. Each experimental unit had an area of 368.64 m2 and 25 trees planted with an arrangement of 5×5 stems with a distance between rows of 4.8 m, forming a perfect square. Treatments consisted of four doses of anaerobic sewage sludge meaning that 0, 100, 200 and 400 kg total N ha-1 were applied in 2001, 2002 and 2003. Sewage sludge was applied superficially and the calculation of the required amounts was conducted according to the percentage of total N and dry matter contents (EPA, 1994) and taking into account the Spanish regulation (R.D 1310/1990) (BOE, 1990) regarding heavy metal concentration for sewage sludge application. Tree heights were measured with a graduated ruler in October 2008 and pasture production was determined by taking four samples of pasture per plot at random (0.3 × 0.3 m2) during the spring and winter from 2002 to 2008. In the laboratory, the pasture samples were dried (72 hours at 60ºC) and weighed to estimate dry matter production. Annual pasture production per year was calculated by summing the Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 264 consecutive harvests of the pasture production in that year. Before drying, the pasture samples were separated into the different species by hand. Data were analysed using ANOVA and differences between averages were shown by the LSD test, if ANOVA was significant. The statistical software package SAS (2001) was used for all analyses. Results and discussion Tree growth was significantly higher in treatments receiving 200 and 400 kg N total ha-1 than 0 or 100 kg N ha-1. However, pasture production ha-1 was not modified during the three years, due to the lack of residual effect of fertilization. Higher doses of sewage sludge significantly increased pasture production in the first years of the experiment (Ferreiro-Domìnguez et al. 2011). Pasture production was reduced in the treatment with 200 kg N total ha-1, probably explained because tree canopy cover was complete. Prod. pasto = -1.1643 h2 + 9.5264 h - 15.834 Pasture production R2 = 0.5507 de pasto (Mg Producción Pasture production Mg ha ha-1-1) 3,7 0N 3,6 400N 3,5 3,4 3,3 100N 3,2 200N 3,1 3,5 4,0 4,5 5,0 h (m) Species Riqueza Richness específica a a 8 nº especies Species Number a a ab a ab 7 ab 6 b b bc 5 c 4 3 0N 100N 200N 400N 0N 100N 200N 400N 0N 100N 200N 400N 2006 2007 2008 Tratamientos Treatments Figure 1. Tree height (m), annual pasture production (Mg ha -1) relationship in 2008 is shown in the upper figure. Lower figure shows species richness per treatment Different letters indicate significant differences between treatments. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 265 No fertilización (0N) (2007) 1 Va Rp Pl Or Bp Ue Ta Cer Ra Dg Hl Rn Bp Rp Card Ta Ra St Or Cer Lt Dg Va Rn Pl Hl Tr Ra Pl Lt Cen Rn Cer Tr Ro Dg Va Ra Bp Rp Gr Cer Pl Ta Rn Va Lp Ta Or Lp St Pl Card Gr Tr J Cer Va Ro Cen Hl Dg Aca Pr Se Lt Rn Pl 0 Gr 0 Card 0,2 0 Cer 0,2 Cap 0,4 0,2 Cen 0,6 0,4 Poa 0,6 0,4 Ro 0,6 Hl 0,8 Dg 0,8 Aca 400 kg total N ha-1 (400N) (2008) 1 0,8 Tr (2007) 0 Aca Tr Cap Ra St Card So Pl Cen Rp Cer Ro Rn Va Hl Dg 400 kg total N 1 ha-1 (400N) Cen (2006) Aca Tr Se Lt Or Card Cen 400 kg total N 1 Gr 0 Ro 0,2 0 Pl 0,2 St 0,4 0,2 Cer 0,6 0,4 Poa 0,6 0,4 Hl 0,6 Dg 0,8 Aca 0,8 ha-1 (400N) 200 kg total N ha-1 (200N) (2008) 1 0,8 Ro (2007) 0 Aca Se Ro Gr St Rp Cer Rll Hl Or Lt Ta Va Rn Cen Tr Pl 200 kg total N 1 ha-1 (200N) Dg (2006) Dg Aca Pr Cyn Rll Se Ue Ro Or Cer Tr Rn Aca 200 kg total N 1 Gr 0 Lt 0,2 0 Ta 0,2 Hl 0,4 0,2 Va 0,6 0,4 Poa 0,6 0,4 Cen 0,6 Pl 0,8 Dg 0,8 ha-1 (200N) 100 kg total N ha-1 (100N) (2008) 1 0,8 Hl 1 Aca St Se Gr Ra Va Tr Rll Or Cer Hl Rn Dg Cen 100 kg total N ha-1 (100N) (2007) Hl 1 0 Aca Ue Tr St Se Poa Ta Cen Dg Aca Pl 0 Lt 0,2 0 Rll 0,2 Or 0,4 0,2 Cer 0,6 0,4 Card 0,6 0,4 Rn 0,6 Hl 0,8 100 kg total N ha-1 (100N) (2006) No fertilización (0N) (2008) 1 0,8 Aca No fertilización (0N) (2006) 1 0,8 Figure 2. Abundance diagrams from 2006 to 2008 in the different treatments. Aca: Agrostis capillaris L., Bp: Bellis perennis L., Cap: Capsella bursa pastoris L., Card: Cirsium arvense L., Cen: Centaurea limbata Hoffmanns, Cer: Cerastium glomeratum Thuill, Cyn: Cynosurus cristatus L., Dg: Dactylis glomerata L., Gr: Geranium rotundifolium L., Hl: Holcus lanatus L., J: Juncus effusus L., Lp: Lolium perenne L., Lt: Lotus corniculatus L., Or: Ornithopus compressus L., Pl: Plantago lanceolata L., Poa: Poa pratensis L., Pr: Prunella vulgaris L., Ra: Rumex acetosa L., Rll: Rumex acetosella L., Rn: Ranunculus repens L., Ro: Rumex obtusifolius L., Rp: Raphanus raphanistrum L., Se: Senecio jacobaea L., St: Stellaria media L. (Vill), So: Sonchus oleraceus L., Ta: Taraxacum officinale Weber, Tr: Trifolium repens L, Ue: Ulex europaeus L., Va: Veronica agrestis L Understorey species richness was promoted by fertilization from 2001 to 2008. The exception was the treatment with 200 kg N ha-1 in the last year, probably due to the lack of the heterogeneity created by the tree, which covers the understorey by a 100%. The dominant species after 8 years of planting was Agrostis capillaris, which is usually adapted to soils with low fertility (treatments 0 and 100 kg N ha-1) and shade (200 kg N ha-1). However, this species is not dominant in the treatment with 400 kg N ha-1 due to the smaller amount of shade than in 200 kg N ha-1 and the advantage that high fertility provides the high dose of sewage sludge which promotes a co-dominance with species better adapted to higher soil fertility, like Holcus lanatus and Dactylis glomerata. Conclusion Improving soil fertility increases grassland biodiversity if tree canopy cover is not complete. References Ferreiro-Domínguez N., Rigueiro-Rodríguez A. and Mosquera-Losada MR (2011) Response to sewage sludge fertilisation in a Quercus rubra L. silvopastoral system: soil, plant biodiversity and tree and pasture production. Agriculture, Ecosystems and Environment 141, 49-57. SAS (2001) SAS/Stat User´s Guide: Statistics, SAS Institute Inc., Cary, NC, USA, 1223 pp. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 266 The effects of agricultural forages on soil biology – linking the plant-soilinvertebrate ecosystem Crotty F.V., Fychan R., Scullion J., Sanderson R. and Marley C.L. Institute of Biological, Environmental and Rural Sciences, Aberystwyth University, Gogerddan, Aberystwyth, United Kingdom, SY23 3EE Corresponding author: fec3@aber.ac.uk Abstract Soil biology is key to maintaining soil health, and soil health is fundamental to the sustainability of agricultural systems. Alternative forages have higher concentrations of essential nutrients, and different rooting patterns, potentially affecting soil-plant-animal interactions. Soil fauna have significant effects on belowground processes and are a vital part of carbon/nitrogen cycling, litter decomposition and the redistribution of nutrients. It is unknown how the soil food web will be affected by different forages, whilst all other environmental variables remain the same, under field conditions. An experiment was set up to test the hypothesis that alternative forages would alter the soil habitat leading to changes in soil biology. To investigate this, plots of chicory (Cichorium intybus), red clover (Trifolium pratense) white clover (Trifolium repens) and perennial ryegrass (Lolium perenne) were established in 2009. Plots were maintained over a three-year period, before soil biology samples were taken including soil mesofauna, nematodes, and earthworms. Significant differences were found between the forages and earthworm abundance, as well as some of the microarthropod groups. The implication of these results in relation to the soil food web and sustainable grassland systems is discussed. Keywords: Trifolium pratense, Trifolium repens, Cichorium intybus, soil food webs, earthworms Introduction Agricultural profitability and sustainability are key to farming successfully, and with increasing feed and fertilizer costs, the use of home-grown alternative forages have increased. Advances in silage technology have allowed alternative forages to be ensiled as high-protein winter forage for livestock (Wilkins and Jones, 2000), giving farmers an option which may reduce their reliance on bought-in concentrate feed. Different forages have been shown to alter soil chemical composition due to their effects on the physical, chemical and biological properties of soil, due to differences in rooting depth (Culman et al., 2010) and N-fixation (Rasmussen et al., 2012). Agricultural grasslands usually support a relatively stable and numerous soil biota that contribute to soil functioning and fertility (Murray et al., 2012). Earthworms are often referred to as ecosystem engineers (Jones et al., 1994), because of their large effect over the soil environment (Blouin et al., 2013). Microarthropods also have significant effects on belowground processes and contribute to the carbon/nitrogen cycle and litter decomposition (Osler and Sommerkorn, 2007). To begin to understand the effect of different alternative forages on soil biology, an experiment was set up to test the hypothesis that alternative forages would alter the soil habitat leading to changes in the soil biology, in plots within the same field maintained under the same environmental conditions. Materials and methods The experimental area was set up at IBERS, Aberystwyth University (52° 25' 59" N 4° 1' 30" W) in 2009, and was uniformly ploughed to the same depth and standardized in accordance with RB209 guidelines (DEFRA, 2010). Field plots (7.5 m × 12 m) of five treatments were established in a randomized replicate block design (n=4). Perennial ryegrass Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 267 (Lolium perene) (cv. Premium) without any inorganic N ha-1 (0N), perennial ryegrass plus 200 kg N ha-1 (200N), chicory (Cichorum intybus) (cv. Puna II), white clover (Trifolium repens) (cv. AberDai) and red clover (Trifolium pratense) (cv. Merviot) were established at sowing rates of 33, 33, 6, 6 and 16 kg ha-1, respectively. Between 2009 and 2012 the plots were mechanically harvested and the cut material removed; plots of ryegrass 200N and chicory had 200 kg N ha-1 annum-1, applied as ammonium nitrate. No N fertilizer was applied to the plots of red clover, white clover or ryegrass 0N. Soil P and K indices were maintained at indices of 2+ using a dressing of 0-24-24 inorganic fertilizer applied at a rate of 60 kg P2O5 and 60 kg K2O ha-1 annum-1. In autumn 2012 and spring 2013, soil biological measurements were taken. Nematode abundance was assessed from a sub-sample of 200 g soil taken from across each plot to a depth of 15 cm, extractions were according to the methods of Whitehead and Hemming (1965) for wet tray extraction. Microarthropods were sampled from three intact soil cores collected from each replicate plot and bulked inside Tullgren funnels. Arthropods were extracted into 70% alcohol, over a 7-day period, prior to being counted and identification using a microscope (following Crotty et al. (2014)). Earthworm abundance and diversity was measured through the excavation of a 30 × 30 × 30 cm soil block from each plot, and through hand sorting to obtain the earthworms, before identification to species and biomass measured. All data were analysed using the statistical package GenStat (VSN International Ltd., Hemel Hempstead, UK) 14th edition. Data were analysed by a general analysis of variance (ANOVA) as a split plot, with treatment as the main factor. Where applicable, multiple comparisons were made using the Student Newman Keuls test. Results and discussion After the three years had elapsed and soil biology populations monitored, we found significant differences in earthworm abundance (Figure 1, P < 0.001), with the white clover treatment having greater abundances than all other treatments. Increases in earthworm abundance are known to increase organic matter consumption, soil structure modification and water infiltration (Blouin et al., 2013), leading to greater nutrient mobility and improved soil health. We found no difference between nematode abundance between treatments (mean ± sem of 43 ± 2 total nematodes per gram dry soil). However, grouping the microarthropods into feeding guilds (sensu Crotty et al. (2014)) the two ryegrass treatments had significantly greater numbers of insect herbivores (P < 0.001), compared to the alternative forages, with ryegrass 200N having the greatest number (>5000 herbivorous microarthropods per m2). The other microarthropod feeding guilds (microbivores, detritivores, omnivores, micropredators and macropredators) did not differ among treatments. Herbivores are known to reduce crop yields; therefore, in agriculture a reduction in numbers is preferable. Ploughing is known to reduce the heterogeneity of a field site (Hendrix et al., 1986); doing this prior to the start of the experiment allowed the impact of the four different forages on soil biology to become more visible. We hypothesized that over the preceding three years after ploughing there would be a net movement of microarthropods (mesofauna, nematodes and earthworms) into or out of each of the treatments, depending on preference. Here, our results suggest the alternative forages, particularly white clover, are promoting a healthier soil environment in comparison to the ryegrasses. It is thought that the introduction of legumes has a key influence over how soil biota function, promoting soil structure, water retention, biodiversity and C and N storage (Murray et al., 2012). Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 268 Figure 1. Mean total earthworm abundance (autumn and spring sampling period combined) per m2 (±s.e; n=4). Significant treatment effects indicated by difference in letters (P < 0.05) analysed using ANOVA and SNK test. Conclusion Our study compares belowground soil food webs to assess the effect of growing different agricultural forage crops. Despite all plots being grown within close proximity to each other, within the same field, differences were found within the soil faunal groups. This suggests that over time there is an immigration / emigration from an area of soil dependent on crop type, and should be a consideration when implementing different sustainable farming methods. Acknowledgements This work is funded through the Rural Development Plan for Wales 2007–2013, funded by the Welsh Government and the European Agricultural Fund for Rural Development. References Blouin M., Hodson M.E., Delgado E.A., Baker G., Brussaard L., Butt K.R., Dai J., Dendooven L., Peres G., Tondoh J.E., Cluzeau D. and Brun J.J. (2013) A review of earthworm impact on soil function and ecosystem services. European Journal of Soil Science 64, 161-182. Crotty F.V., Blackshaw R.P., Adl S.M., Inger R. and Murray P.J. (2014) Divergence of feeding channels within the soil food web determined by ecosystem type. Ecology and Evolution 4, 1-13. Culman S.W., DuPont S.T., Glover J.D., Buckley D.H., Fick G.W., Ferris H. and Crews T.E. (2010) Long-term impacts of high-input annual cropping and unfertilized perennial grass production on soil properties and belowground food webs in Kansas, USA. Agriculture Ecosystems & Environment 137, 13-24. DEFRA (2010). Fertilizer Manual (RB209), 8th ed. The Stationery Office, Norwich, UK, p. 257. Hendrix P.F., Parmelee R.W., Crossley D.A., Coleman D.C., Odum E.P. and Groffman P.M. (1986) Detritus food webs in conventional and no-tillage agroecosystems. Bioscience 36, 374-380. Jones C.G., Lawton J.H. and Shachak M. (1994) Organisms as ecosystem engineers. Oikos 69, 373-386. Murray P.J., Crotty F.V. and Van Eekeren, N. (2012) Management of grassland systems, and soil and ecosystem services. In: Wall D.H., Bardgett R.D., Behan-Pelletier V. and Van der Putten W.H. (Eds.) Soil ecology and ecosystem services. Oxford University Press, Oxford, UK, p. 424. Osler G.H.R. and Sommerkorn M. (2007) Toward a complete soil C and N cycle: incorporating the soil fauna. Ecology 88, 1611-1621. Rasmussen J., Soegaard K., Pirhofer-Walzl K. and Eriksen J. (2012) N2-fixation and residual N effect of four legume species and four companion grass species. European Journal of Agronomy 36, 66-74. Whitehead A.G. and Hemming J.R. (1965) A comparison of some quantitative methods of extracting small vermiform nematodes from soil. Annals of Applied Biology 55, 25-38. Wilkins R.J. and Jones R. (2000) Alternative home-grown protein sources for ruminants in the United Kingdom. Animal Feed Science and Technology 85, 23-32. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 269 Managing grasslands to mitigate flooding risk Newell Price J.P., Balshaw H. and Chambers B.J. ADAS Gleadthorpe, Meden Vale, Mansfield, United Kingdom, NG20 9PD ADAS Rosemaund, Preston Wynne, Hereford, United Kingdom, HR1 3PG, UK Corresponding author: Paul.Newell-Price@adas.co.uk Abstract Grassland can be a significant source of surface runoff and can dominate flooding risk in some catchments. Mechanical loosening is often promoted as a means of improving soil structural condition and reducing surface runoff volumes. In 2010, four ‘high’ bulk density sites were selected to investigate the impact of shallower and deeper mechanical loosening, and interactions with the introduction of deep-rooting herbs and legumes. Double ring infiltrometers were used to measure both initial and saturated water infiltration rates in spring 2011, 2012 and 2013. Of all the treatments, deeper mechanical loosening had the greatest effect on both initial and saturated infiltration rates. Effects were greatest on the ‘medium’ soil sites, where deeper loosening resulted in saturated infiltration rates that were 4- to 10-fold greater than the un-loosened control (P<0.001). These results could have significant implications for the management of grassland soils in reducing surface runoff and flooding risk, particularly in catchments dominated by dairy and beef/sheep farms. Keywords: Flooding risk, infiltration, soil structure Introduction In England, more than 5 million properties are at risk of flooding (nearly 1 in 6 homes). As a response to the devastating flooding in 2007, the Pitt Review provided a series of recommendations for improving the way flood risk is managed. The review advised that Catchment Flood Management Plans should be established to achieve greater working with natural processes such as using farmland to slow the progress of water to sub-catchment outlets and to minimize runoff. Grassland can be a significant source of runoff and sediment (Collins et al., 2010) and there is a need to improve understanding of the implications of using mechanical loosening to increase water infiltration in grassland soils and potentially reduce peak flow in catchments that have a recognized flooding and erosion problem. Materials and methods In 2010, four ‘high’ bulk density sites were selected to test mechanical remediation of compaction and its interaction with the introduction of deep-rooting herbs and legumes (DHL) (Table 1). The following treatments were applied, using a split-plot randomized block design: i) an un-loosened and uncultivated control; ii) uncultivated with shallower (c. 20 cm) loosening; iii) uncultivated with deeper (c. 30-35 cm) loosening; iv) un-loosened and power harrowed without species introduction; v) power harrowed without species introduction with shallower (c. 20 cm) loosening; vi) power harrowed without species introduction with deeper (c. 30-35 cm) loosening; vii) power harrowed, DHL mix drilled and un-loosened; viii) power harrowed and DHL mix drilled with shallower (c. 20 cm) loosening; and ix) power harrowed and DHL mix drilled with deeper (c. 30-35 cm) loosening. There were four replicates of each treatment with individual plots measuring 8 × 16 m. Mechanical loosening was carried out using a grassland ‘top-soiler’ with leading discs followed by a set of tines and a single packer roller to the rear of the unit. Double-ring infiltrometers with an internal diameter of 600 mm (internal ring) were used to measure both initial and saturated water infiltration rates in spring 2011, 2012 and 2013. Water infiltration rates in each of the three harvest years were analysed Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 270 by analysis of variance (ANOVA) using GenStat® Release 13 (Payne et al., 2010) to discern significant effects of treatments. Results and discussion Of all the treatments, deeper mechanical loosening had the greatest effect on both initial and saturated infiltration rates. Effects were greatest on the ‘medium’ soil sites (Nafferton, Odstone and Aberbran). At Bicton, the only ‘sandy’ soil site (topsoil clay content < 18%), deeper loosening increased what were already relatively high saturated infiltration rates in 2011, but there was no effect in subsequent years. By contrast, at the three ‘medium’ soil sites, mechanical loosening reduced infiltration rates into a third year, post-loosening. Mean saturated water infiltration rates in 2011, 2012 and 2013 are presented for all three ‘medium soil’ sites in Figure 1. Both shallower and deeper loosening resulted in higher (P<0.001) saturated infiltration rates (compared with the un-loosened control) in 2011 and 2012. By late winter 2012-13 (c. 30 months post loosening), infiltration rates were still higher (P<0.001) on the deeper loosened treatment compared with the un-loosened control. Notably, both initial and saturated infiltration rates in late winter 2012-13 on all three treatments were lower than in previous years, which may be a reflection of the very wet 2012 season. Figure 1. Effect of mechanical loosening on saturated water infiltration rates (mm/hour) on medium grassland soils with ‘high’ bulk density. The saturated water infiltration rates on the unloosened plots compared favourably with ‘typical’ infiltration rates reported in MAFF Reference Book 441 (1982) of 150-500 mm/h for sandy soils (as at Bicton) and 30-50 mm/h for clay loam sites (Nafferton, Odstone and Aberbran). The above results, therefore, suggest that mechanical loosening (particularly deeper loosening) can increase water infiltration rates that are commonly found on managed grassland by 4- to 10-fold. This could have important implications for the management of grassland to increase slurry infiltration and for the reduction of flooding risk at the local scale. Within the time frame of this study, the seed mix had no effect on water infiltration rates. However, forbs and legumes can take a few years to establish so it is possible that rooting effects could be observed once plants have been well established for 4-5 years. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 271 Table 1. Baseline soil analysis results for the four experimental sites. Determinand Units Nafferton Odstone Aberbran Bicton pH unit 5.9 6.9 5.6 5.7 Cross-compliance soil type - ‘Sandy and light silty’ ‘Medium’ ‘Medium’ ‘Medium’ 12 17 19 43 [1] [2] [2] [3] 91 210 76 125 [1] [2+] [1] [2-] Extractable Phosphorus (P) Extractable Potassium (K) mg/l [ADAS Index] mg/l [ADAS Index] mg/l 170 145 77 68 [ADAS Index] [3] [3] [2] [2] Total Nitrogen (N) % dm 0.29 0.25 0.21 0.17 Organic Carbon1 % dm 2.9 2.3 2.1 1.7 Organic Matter2 % dm 5.0 4.0 3.7 3.0 Bulk density g/kg 1.36 1.44 1.27 1.30 Extractable Magnesium (Mg) Conclusion Mechanical loosening can have significant effects on soil physical properties that persist for at least 30 months post-loosening. These changes in soil physical structure are associated with dramatic 4- to 10-fold increases in water infiltration rate that can also persist for a similar amount of time. Deeper loosening (to c. 30-35 cm depth) resulted in greater increases in water infiltration rate than shallower loosening (to c. 20 cm). Overall, it is concluded that deeper loosening is a more effective treatment than shallower loosening in improving grassland soil structural condition and function. Acknowledgements The funding provided by the Department for Environment, Food and Rural Affairs (Defra) is gratefully acknowledged. References Collins A.L., Walling D.E., Webb L. and King P. (2010) Apportioning catchment scale sediment sources using a modified composite fingerprinting technique incorporating property weightings and prior information. Geoderma 155, 249-261. MAFF (1982) Techniques for measuring soil physical properties. Ministry of Agriculture Fisheries and Food Reference Book 441. HMSO, London. Payne R.W., Murray D.A., Harding S.A., Baird D.B. and Soutar D.M. (2010) Introduction to GenStat® for Windows™. VSN International, Hemel Hempstead, UK. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 272 Developing an in situ sensor for real time monitoring of soil nitrate concentration Shaw R.1, Williams A.P.1, Miller A.2 and Jones D.L.1 1 School of Environment, Natural Resources and Geography, Bangor University, Gwynedd, United Kingdom, 2 John Innes Centre, Norwich, Norfolk, United Kingdom Corresponding author: rory.shaw@bangor.ac.uk Abstract Improving nitrogen-use efficiency is key to increasing the sustainability of livestock farming systems. Better management of nitrogen (N) inputs and waste resources is needed if significant improvements are to occur. However, farmers are currently limited by a lack of suitable, fieldbased tools for soil analysis and are overly reliant on limited, computer-based approaches. This project aims to develop an ion-selective electrode (ISE) capable of in situ, real-time soil monitoring of soil nitrate (NO3-). Construction of the electrodes in the laboratory is simple, low-cost and reproducible and the ISEs conform to theoretical norms. Current and future work will focus on the testing the performance of the electrodes in soil solutions and soils, and comparing the results to a range of standard methods. Keywords: sustainable farming, nitrogen-use efficiency, ion-selective electrode, soil testing Introduction Optimizing the use of nitrogen (N) represents one of the major goals of sustainable livestock farming systems, from both an economic and environmental standpoint. While there have been thousands of studies investigating different management strategies for optimizing N use on farms, translating this research into practical management advice and subsequent adoption by farmers has often been unsuccessful. Consequently, as evidenced by numerous recent reports, there is no doubt that we have a long way to go before N is used efficiently within the UK livestock sector (Wilkins, 2008; Rees and Ball, 2010; Spiertz, 2010). The lack of significant improvement is partially due to a paucity of farmer-operated, field-based tools for soil analysis, and reliance on virtual (computer)-based approaches, which have limited adoption and lack precision (Cuttle and Jarvis, 2005). Ideally, adoption of simple field-based sensors that can monitor soil N in real-time will allow for more active management of fertilizer and waste resources, resulting in enhancement of fertilizer use efficiency, better timing of waste applications and reductions in environmental pollution. Ion-selective electrodes (ISEs) have the potential to be used for real-time, in situ soil monitoring and have the advantages of being relatively inexpensive and easy to use (De Marco et al., 2007; Sinfield et al., 2010). However, current ISEs are not sufficiently robust enough for soil sensing and are subject to drifting calibration parameters, fouling and short lifespans (De Marco et al., 2007). This project aims to develop a nitrate (NO3-) ISE which can overcome these challenges and bring about a step change in on-farm N management. Materials and methods ISEs function by measuring the potential difference between an electrode containing an ionselective membrane and a reference electrode, which is not affected by the target ion. The activity of the target ion is related to the potential difference by the Nernst equation, which states that a ten-fold increase in activity of the target ion will result in a 59.1 mV change in electrode output. We construct the electrodes in our laboratory using a simple protocol and easily sourced materials. Briefly, both the NO3- and reference electrodes consist of a 1250 µl pipette tip, into which a PVC-based membrane is cast. The pipette tips are then back filled with Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 273 an unspecified solution, a Ag/AgCl wire is inserted, and the tip sealed. The electrodes are then coupled with a millivolt meter or suitable data logger. Initial work has focused on characterizing the electrode’s basic properties, including response time, the effect of interfering ions, developing a temperature compensation calculation and standardizing the manufacturing protocol to obtain a reproducible calibration (Buck and Lindner, 1994). Currently, we are testing the sensors in soil solutions and comparing the results to a standard colorimetric analytical method (Miranda et al., 2001) in order to assess the accuracy and precision of the sensors for NO3- determination. In addition, we are monitoring soil NO3- levels in the laboratory and assessing the performance of the sensors against a range of soil sampling techniques, including small tension lysimeters, centrifuge drainage and conventional soil core analysis. Following completion of this work, the sensors will be deployed in a grass-clover field trial in March–August 2014 to determine the differences in soil nitrate dynamics between different clover densities and inorganic N fertilizer amendments. Results and discussion Electrodes 1.7 - 1.12 calibration 29-1-14 Construction of the electrodes has proved succesful and reproducible, as seen in Figure 1. 300 250 EMF (mV) 200 150 100 50 0 0 2 4 pNO3 6 - Figure 1. Calibration of NO3- ISE in standard NO3- solutions. Electromotive force (EMF) is the electrode output and pNO3- is the negative log of the NO3- activity. Data points represent means (n=6) ± SEM and the curve represents a modified Nernst equation. These electrodes have a limit of detection of 47 µM and a near nernstian slope of 61.7 mV dec1 Temperature effect on NO3 ISE . 250 200 mV 150 100 50 0 5 10 15 20 25 30 35 40 Temperature c Figure 2. The effect of temperature on NO3- ISE output at different concentration of NO3-. Data points represent means (n=6) ± SEM Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 274 Figure 2 shows how electrode output (mV) is affected by temperature, with the magnitude of the change dependent upon the NO3- concentration of the sample. This is expected as the Nernst equation contains a temperature term, so the electrode outut can be easily adjusted using a simple calculation. Chloride is a well known interfering ion (Miller and Zhen, 1991) and the effect on these electrodes was determined using the fixed interference method (Buck and Lindner, 1994). The electrodes were recalibrated in the prescence of 100 mM Cl- and the selectivity coefficient was found to be 0.04. Ongoing work and the results of the field trials will be reported on in September, as well as recommendations for industry as to how to take this promising technology further. Acknowledgements The authors would like to thanks Hybu Cig Cymru (HCC), the Engish Beef and Lamb Executive (EBLEX), DairyCo. and Quality Meat Scotland (QMS) for funding this project. References Buck R.P. and Lindner E. (1994) Recommendations for nomenclature of ion-selective electrodes - Iupac recommendations 1994. Pure and Applied Chemistry 66, 2527-2536. Cuttle S.P. and Jarvis S.C. (2005) Use of a systems synthesis approach to model nitrogen losses from dairy farms in south-west England. Grass and Forage Science 60, 262-273. De Marco R., Clarke G. and Pejcic B. (2007) Ion-selective electrode potentiometry in environmental analysis. Electroanalysis 19,1987-2001. Miller A.J. and Zhen R.G. (1991) Measurement of intracellular nitrate concentrations in Chara using nitrateselective microelectrodes. Planta 184, 47-52. Miranda K.M., Espey M.G. and Wink D.A. (2001) A rapid, simple spectrophotometric method for simultaneous detection of nitrate and nitrite. Nitric Oxide-Biology and Chemistry 5, 62-71. Rees R.M. and Ball B.C. (2010) Soils and nitrous oxide research. Soil Use and Management 26, 193-195. Sinfield J.V., Fagerman D. and Colic O. (2010) Evaluation of sensing technologies for on-the-go detection of macro-nutrients in cultivated soils. Computers and Electronics in Agriculture 70, 1-18. Spiertz H. (2010) Food production, crops and sustainability: restoring confidence in science and technology. Current Opinion in Environmental Sustainability 2, 439-443. Wilkins R.J. (2008) Eco-efficient approaches to land management: a case for increased integration of crop and animal production systems. Philosophical Transactions of the Royal Society B. 363, 517-525. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 275 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 276 Theme 2 posters Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 277 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 278 PastureBase Ireland – the measurement of grass dry matter production on grassland farms Griffith V., O’Donovan M., Geoghegan A. and Shalloo L. Animal & Grassland Research and Innovation Centre, Teagasc, Moorepark, Fermoy Co. Cork Corresponding author: vincent.griffith@teagasc.ie Abstract Ireland in 2015 will produce milk without the restriction of milk quota. The future of an efficient low-cost milk production system will depend on low-cost feed in the form of grazed grass being converted to milk. On many farms some level of grassland measurement is being completed. To date, in Ireland there has been no central database to retain and archive these data for research purposes. The creation of PastureBase Ireland (PBI) in January 2013 addresses this issue. PastureBase Ireland is a web-based grassland database which has an inbuilt grassland decision support tool. The aim of the system is to build a large national grassland database which will increase the level of grass utilization and production at farm level. A key aspect of PBI is that the grassland farmer inputs the data at farm-level, through measuring farm grass covers regularly on the farm. For this paper, a subset of farms using PBI in 2013 was selected; all farms had a high level of grassland measurement to ensure the data used were accurate. The average DM production on the farms was 12.5 t ha-1. There was significant variation in the amount of grass produced between farms, ranging from 16 t ha-1 to 8 t ha-1. There was also significant variation between the average number of grazings per paddock achieved between farms, ranging from 9.1- 4.1. The implications of the grass growth data, the underlying reasons for variations in grass growth between farms and the variation that occurs within farms are discussed. Keywords: PastureBase Ireland, grass, database, measurement Introduction The potential to produce between 12 and 16 t ha-1 grass dry matter (DM) over a long growing season is a major competitive advantage for Irish dairy farmers. Dillon et al. (2005) showed a strong relationship between total production costs and the grazed-grass proportion in the diet of the dairy cow in a number of countries. The average milk production cost was reduced by 1 c/L for a 2.5% increase in grazed grass in the dairy cow diet. The data also demonstrated that a considerable proportion of the dairy cow diet (>50%) must comprise grazed grass before a significant impact on production cost is realized. In recent years, grazing management strategies have been identified that increase the proportion of grazed grass in the dairy cow diet, which reduces the dependency on indoor feeding in Irish systems. The competitive advantage of grass-based production systems is expected to increase over the next few years due to higher costs associated with conserved and bought-in feed. There is a requirement to refine and develop grassland technologies while also identifying the factors on-farm that are reducing grass growth. The development of a national grassland database which incorporates both a front-end, where farmers can enter grassland data, and a database, to collect all grassland data (commercial and research) would provide valuable strategic data for the entire grassland industry (dairy, beef and sheep enterprises). With these goals in mind, Teagasc launched PastureBase Ireland (PBI) in January 2013. This was built from an in-house prototype. The database stores all grassland measurements within a common structure. PBI will allow the quantification of grass growth and DM production (total and seasonal) across different enterprises, grassland management systems, regions and soil types, using a common measurement protocol and methodology. The background data such as paddock soil fertility, grass/clover cultivar, aspect, altitude, reseeding history, soil type, Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 279 drainage characteristics and fertilizer applications are also recorded. All measurements are attached to individual paddocks within farms and PBI is used on both commercial and research farms. PastureBase Ireland will also, for the first time, link grass growth on farms to reliable Met Eireann weather data. The objective of this paper is to examine grass growth data from a subset of farms in the PBI database and to identify the variations in grass growth that occur between farms and the variations that occur within a farm. The paper will also examine the economic loss of underproducing paddocks and strategies that farmers can use to increase grass growth on farms. Materials and methods Thirty-eight commercial grassland farms using the PBI decision support tool were selected to estimate grass production. Grass production was calculated from 1 January 2013 to 10 December 2013 (full grazing season). The farms used within this analysis were selected based on number of individual farm grass measurements, which had to exceed 30 grass measurements across the grazing season. All farms selected measured grass growth on each paddock weekly from January to December. The herbage mass on each paddock was measured either by visual assessment (calibrated by cutting and weighing) (O’Donovan et al., 2002) or by rising plate meter (Castle, 1976). Paddock growth was only calculated when the time between measurements did not exceed 16 days; all farms selected achieved this standard throughout the year. If farms did not have measurements within the 16-days window they were removed from the analysis. Both grazing yield and silage yield (estimated on harvest date) were calculated separately and then combined to give total grass production. The number of actual grazings achieved during the grazing season was calculated within the recording system. Average grass production (t ha-1) was calculated across all farms and for different soil types. The data set was analysed in SAS (2003). PROC GLM was used to examine the effect of farm on total, grazing and silage DM production. Results and discussion Table 1 shows the grazing total DM yield, DM yield, silage DM yield and number of grazing’s achieved on the farms for the entire 38 farms. Mean grass production across the farms (which ranged across Ireland) was 12.5 t ha-1. The most productive farms produced up to 8 t ha-1 more than the lowest producing farms for both total and grazing DM yield. The number of actual grazing ranged from (9.1 - 4.1), which shows the level of grass utilization on the farms. The level of silage produced on some of these farms is low as farms are focusing on producing and utilizing grass as an in situ feed rather than as conserved forage. Table 1. The total, grazing, silage DM production and number of grazings achieved on 38 grassland farms in 2013 Total production Grazing production Silage production Number of grazings achieved (t ha-1) (t ha-1) (t ha-1) Mean 12.52 10.35 2.13 6.17 SED 0.48 0.56 0.45 0.329 CV 0.26 0.22 0.17 0.213 Sig *** *** *** *** Farms which produced high levels of grass tended to have less between-paddock variations (this is measured by having a low CV – coefficient of variation) in comparison with farms producing lower levels of grass DM. Highly productive farms have adopted strategies such as regular monitoring of soil fertility, targeting reseeding on low-producing paddocks, reducing poaching damage, and continuing to maintain high levels of perennial ryegrass within paddocks to maximize production from individual paddocks. It is clear that there is huge economic loss Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 280 within and between farms due to variation in grass production. Shalloo (2009) found that increasing grass utilization was worth €160 per t DM utilized, and explained 0.44% of variation in net profit on commercial farms. Conclusion Grass DM production on the selected farms in 2013 averaged 12.5 t ha-1 which is considered a high level of grass growth. Large variation exists between the highest producing and lowest producing farms. The development of PBI both as a decision support tool and a grassland database is a hugely significant step for the future of grassland production systems in Ireland. PBI has the potential to add significant value to the data collected by individual farmers. These data show that there is significant potential for farmers to grow more grass on farms; however, the underlying reasons for poor grass DM production on individual farms and paddocks will need to be properly investigated and understood to achieve this potential. The creation of PBI will greatly assist this process. References Castle M.E. (1976) A simple disc instrument for estimating herbage yield. Journal of the British Grassland Society 31, 37-40. Dillon P., Roche J.R., Shalloo L. and Horan B. (2005) Optimising financial return from grazing in temperate pastures. In: Utilization of grazed grass in temperate animal systems, Proceedings of a satellite workshop of the XXth International Grassland Congress, Cork, Ireland (ed. J.J.Murphy). Wageningen Academic Publishers, The Netherlands, pp. 131–148. O’Donovan M., Connolly J., Dillon P., Rath M. and Stakelum G. (2002) Visual assessment of herbage mass. Irish Journal of Agricultural and Food Research 41, 201-211. SAS (2003) SAS Institute. Cary, NC, USA. Shalloo L. (2009) Pushing the barriers on milk costs/ outputs. In: Proceedings of Teagasc National Dairy Conference 2009, Teagasc, pp. 19-38. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 281 Estimation of grassland production with a new land classification system in Hungary Hoffmann R., Keszthelyi S. and Pál-Fám F. Kaposvár University, Institute of Plant Science, H-7400 Kaposvár, Guba S. 40, Hungary Corresponding author: Richard Hoffman (email: erwiniar@gmail.com) Abstract The future way of grassland management is greatly affected by the new functions of grasslands in relation to the environment. The evaluation of grasslands in Hungary is not solved: it is hard to plan the yields and we can estimate them only a posteriori. The D-e-Meter grassland module is a new grassland evaluation method. Our aim was the replacement of the disused classification system of grasslands, and the introduction of a new evaluation method as a function of sustainable agricultural developing. This aspect could be found in the multifunctional European Agricultural model and also in rural development. Therefore, we have carried out field experiments in the quest of solving this task in Bőszénfa, Southwest Hungary. The measured and estimated DM yield values were compared to each other for 5 grassland stands. The results showed that although there are significant differences between the estimated and the measured DM yields, the D-e-Meter method could provide a good basis to plan the management. The utility of the D-e-Meter method was confirmed by the evaluation of grassland plant products. These results certified the effects of different climatic features on grassland productivity. Keywords: D-e-Meter, grassland, evaluation, yields, nutritive content Introduction Nowadays the usage of both arable and pasture land is controlled and has close connection to environmentally friendly farming. An important part of the possibilities of farm management is to know the actual quality and value of the fields (Várallyay, 2003). In the past, several scientists tried to construct methods for the estimation of the quality and value of grassland. Therefore, arable land evaluation and controlled farming are important issues. More classification systems of grasslands were applied in practice during the last decades, starting with Braun-Blanquet (1954), up to Ren et al. (2008). In Hungary, works of Máté (2003), Tóth et al. (2007) and Dér et al. (2008) are relevant. During the last ten years a new method, called D-e-Meter was carried out in Hungary, which hopefully will be suitable for the estimation of the characteristics of both arable land and grasslands (Tóth et al., 2007). While this method will be suitable for arable farming, we have the aim to elaborate an up-to-date method for the grasslands, too. Classification of the grasslands must be solved in some ways. Baskay-Tóth (1966) separated grasslands into 3 groups according to their use: pasture, meadow and combined grasslands. Horn and Stefler (1990) classified grasslands by the intensity of their usage and separated them into three groups: intensive, semi-intensive and extensive pasture. A new way of grassland classification was created by Dér et al. (2003), which separates the grasslands by the way of their usage and the type of it: (1). Productive grassland (1a. unfertilized or barely fertilized grasslands with medium productivity; 1b. frequently fertilized grasslands with high productivity) and (2). Area of outstanding natural beauty (2a. strictly sheltered grasslands; 2b. non-strictly sheltered, and other grasslands of outstanding area; 2c. soil-protecting grasslands). Our aims were the formation of a more accurate system in the Hungarian grassland classification, replacing of the outdated 'Aranykorona-system' by means of this method, which imply both economic features and management facilities of grassland management. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 282 Materials and methods Our experiments were carried out from autumn of 2005, with soil analyses, until autumn of 2008, at the Game Farming Centre of the Kaposvár University, Bőszénfa. The examined five grasslands were: Baltacim (6 ha), Egyenestető (23 ha), Kuti III. (9 ha), Pacsirta (19 ha) and Templom Dél (20 ha). The meteorological details were measured by the Hungarian Meteorological Service. We measured three samples/cuttings per year from 2006 to 2008. The dates of the cuts are given in Table 1. Table 1.Time of sampling (date, as year-month-days) First growth (1) Second growth (2) Third growth (3) 2006.05.1525. 2006.07.515. 2006.09.1525. 2007.05.0515. 2007.07.1525. 2007.09.2030. 2008.05.1020. 2008.07.1020. 2008.09.2510.05. The samples were cut in four repetitions using a quadrat frame and the weights of the samples were measured. The samples were analysed by dry matter (DM) contents, and the crude protein, crude fat, crude fibre, crude ash and N-free extract were quantified by Wendeei-analysis. The statistical analysis was done with SAS 9.1 software at 5% significance level (P≤0.05) by ANOVAs. D-e-Meter estimation method for grasslands is started from the potential DM values from the literature. The modifying factors were: quality factor, coverage factor, soil-water management, agro-ecological district, gradient category, sward establishment date, yearly climatic effect, intensification of farming and purpose of use (grazing or cutting). The complete methodology is compiled in Dér et al. (2008). Results and discussion The results of the measurements and the estimated values are compiled in Table 2. At ‘Baltacim’ we did not detect any significant difference between 2006 and 2007 in the fresh mass product per hectare. The DM yield per hectare and the crude protein yield per hectare also showed no significant differences between the experimental years. The drought of year 2007 had a great effect on arable land production, but grassland production was affected only in the second cut, so the other two cuts could equalize the yields. At ‘Kuti III’ we did not find any significant differences among the experimented yields. At ‘Pacsirta’ the annual change of the weather had a greater effect on the yields. In 2007 the unsuitable management had an effect too, so that the differences were significant in all the measured parameters. At ‘Templom Dél’ the differences were significant in the fresh mass yield in the experimental years. The levels of the yields were influenced by the weather and the land use management. Although there are significant differences between the estimated and the measured DM yields, the values show that the D-e-Meter method could give a good basis to plan the management. By improving this method by 0.20.4 t/hectare we can give estimations for the DM yield without significant differences. Drought has the greatest effect on the botanical composition, the cover of the plants and the DM content in grasslands. In dry years the total number of the species can dramatically decrease, so this creates empty spaces, and both the number and mass of pioneer species (mainly weeds) increase. The DM content of the grasslands increased caused by the drought year of 2007, but the crude protein per DM kilograms and the crude fibre per DM kilograms were almost the same. The yields of the grasslands are influenced to a greater extent by the weather of the year (especially the rainfall) than are the nutritive contents. By evaluating the yields we can say that grasslands with greater yields have greater effect by the changes of the Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 283 weather than by the management. Grasslands can compensate for the environmental effects, which have impact on one cutting, compared to the field cropping species. The deviation between the estimated and measured DM yields is 0.05-12 t/ha, which shows a high level of deviation, but greater margin of error is in the droughty year, so we have to correct this point of the method. Of course, we must correct the accuracy of the method by further dates of the production, but all the results certify the conduciveness of the D-e-Meter system in the near future. Table 2. Estimated and measured DM contents of the investigated grasslands, and the measured fresh mass and crude protein values. Area name Year Estimated DM Yield (t/ha) Measured Yield (t/ha) Baltacim 2006 5.4a 2007 Egyenestető Kuti III. Pacsirta Templom Dél DM Fresh mass (t/ha) Crude protein yield (kg/ha) 5.4a 16.5a 631a 4.8a 5.6b 14.3ab 568a 2008 4.8a 5.6a 13.8b 543a 2006 6.7a 5.8b 20.2a 630a 2007 6.0a 4.8b 14.3b 499b 2008 7.5a 7.7a 23.2c 828c 2006 4.5a 3.9b 13.0a 400a 2007 3.0a 4.0b 12.1a 437a 2008 4.5a 3.8b 11.7a 493a 2006 6.7a 6.1a 22.7a 825a 2007 2.0a 2.6b 8.2b 398b 2008 6.7a 7.0b 23.1a 808a 2006 7.1a 7.4a 23.6a 993a 2007 4.8a 5.5b 15.1b 604b 2008 5.4a 7.6b 17.6c 713b References Baskay-Tóth B. (1966) Legelő- és rétművelés. Mezőgazdasági Kiadó. Budapest. Braun-Blanquet J. (1954) Pflanzensociologie. Springer Verlag, Berlin. Dér F., Fábián T. and Hoffmann R. (2008) Zselici gyepek termésének és táplálóanyag-tartalmának vizsgálata, valamint a területek földértékelése a D-e-Meter rendszerben. Acta Pascuorum 6, 33-38. Dér F., Marton I., Németh T., Pásztor L. and Szabó J. (2003) A szántóföldi növénytermesztés és a gyepgazdálkodás helyzete és kilátásai az EU- csatlakozás után. Nemzeti Fejlesztési Hivatal, 83-142. Horn P. and Stefler J. (1990) Hagyományos és új állattenyésztési ágazatokban rejlő lehetőségek az eltérő ökológiai, piaci adottságok kihasználására. Állattenyésztés és Takarmányozás 39/1, 27-43. Máté F. (2003) Az aranykoronától a D-e-Meter számokig. Földminősítés és földhasználati információ 1, 145-152. Ren J.Z., Hu Z.Z., Zhao J., Zhang D.G., Hou F.J., Lin H.L. and Mu X.D. (2008) A grassland classification system and its application in China. The Rangeland Journal 30/2, 199-209. Tóth T., Vinogradov Sz., Hermann T., Speizer F. and Németh T. (2007) Soil bonitation and land valuation with D-e-Meter system as a tool of sustainable land use. Cereal Research Communications 35/2, 1221-1224. Várallyay Gy. (2003) A földminőség kifejezésének céljai és lehetőségei. Földminősítés és földhasználati információ 1, 81-98. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 284 Influence of nitrogen fertilizers on yield and digestibility of grass Adamovics A. and Platace R. Latvia University of Agriculture, Institute of Agrobiotechnology, Jelgava, Latvia Corresponding author: aleksandrs.adamovics@llu.lv Abstract Forage in Latvia is mainly produced from grasses. Research was conducted with 5 grasses: reed canary grass (RCG) var. ‘Bamse’ (Phalaris arundinacea L.), tall fescue var.‘Fawn’ (Festuca arundinacea Schreb.), festulolium ‘Hykor’ (L. multiflorum x F. arundinacea)x FA, timothy ‘Jumis’ (Phleum pratense L.), and meadow fescue ‘Arita’ (Festuca pratensis Huds.). Yield of dry matter in RCG biomass varied between 4.45 and 9.0 t ha-1, in tall fescue from 4.3 to 8.16 t ha-1, in festulolium from 2.5 to 7.36 t ha-1, in timothy from 2.81 to 7.78 t ha-1, and in meadow fescue from 2.79 to 7.36 t ha-1. The results also show a close negative correlation (r = 0.8729, < 0.05), between content of crude fibre in the biomass and digestibility, and also between content of acid detergent fibre (ADF) and digestibility (r = -0.9990, P < 0.05). Keywords: grasses, fertilizer norms, crude fibre, digestibility Introduction Sustainable livestock production systems are increasingly relying on production of high quality forage that can be produced on farm. Forage in Latvia, and other countries, is mainly produced from grasses. The value of plant green mass depends upon the digestibility of its organic matter; at the beginning of the vegetation period it comprises 80%, whereas at the end phase it falls to 50% or less. Plant digestibility is affected by lignin content, which links to ruminant crude fibre. Therefore, as plants mature, their digestibility of organic matter reduces. This phenomenon is clearly visible with the crude fibre (CF) fraction in respect to content of neutral detergent fibre (NDF) and the acid detergent fibre (ADF) content. Changes of grass yield under the influence of various factors and climatic conditions of Latvia have been studied by several researchers (Adamovics, 1998; Runce, 1999; Bumane et al., 2003; Gutmane et al., 2008). Protein content in the grass dry matter is reported to be largely influenced by the plant development phase and nitrogen fertilizers applied (Adamovics, 1998). To produce high quality fodder and acquire high yields per unit area, research studied then influence of nitrogen mineral fertilizers on yield and digestibility of grasses. Materials and methods Field trials were carried out in 2011-2012 at the Research and Study farm “Peterlauki” (56° 53' N 23° 71' E) of the Latvia University of Agriculture, on sod calcareous soils, with pHKCl 6.7, available P of 52 mg kg 1, K of 128 mg kg 1, organic matter content 21-25 g kg 1. Research was conducted with 5 grasses: reed canary grass (RCG) var. ‘Bamse’ (Phalaris arundinacea L.), tall fescue var. ‘Fawn’ (Festuca arundinacea Schreb.), festulolium ‘Hykor’ (L. multiflorum x F. arundinacea)x FA, timothy ‘Jumis’ (Phleum pratense L.), and meadow fescue ‘Arita’ (Festuca pratensis Huds.). Fertilizer levels used in the research (kg ha-1) were: N0P0K0 (control), P2O5–80 K2O–120 (F– background), F+N30, F+N60, F+N90, F+N120 (60+60), F+N150 (75+75), F+N180 (90+90). Seed sowing was: 1000 germinable seeds per 1 m2. The first cut pf the herbage dry matter (DM) yield was analysed in the Analytical Laboratory for Agronomy Research of Latvia University of Agriculture, for the following quality indices: crude fibre content (measured in compliance with LVS EN ISO 5498:1981), acid detergent fibre ADF (LVS EN ISO 13906: 2008) and digestibility – with cellulose method. The trial data were processed using correlation, regression and variance analyses (ANOVA) and descriptive Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 285 statistics with Microsoft Excel for Windows 2000 (Arhipova and Balina, 2006). The means are presented with their LSD test. Results and discussion Dry matter, t ha-1 Most of the DM yields were produced by the first cut of the grass. Average DM yield of the first cut in two years of production accounted for 62 -75% of the annual yield. Nitrogen fertilizers notably increased yield regardless the weather differences. The grass yield was influenced mainly by mineral-nitrogen fertilizers. The highest DM indicators were recorded for reed canary grass (9.06 t ha-1), tall fescue (8.16 t ha-1), timothy (7.78 t ha-1) and meadow fescue (7.36 t ha-1) with fertilizer level of F+N180(90+90), while that for festulolium (7.36 t ha-1) was with F+N120 (60+60) kg ha-1 (Figure1). Festulolium Timothy Nitrogen fertilizers Figure 1. Dry matter yield of the first cut Digestibility, % The highest crude fibre content was found in meadow fescue (39.30 ±0.32%); in the biomass of other species it was notably (P<0.05) lower, the lowest being in timothy (35.72 ± 0.79%). Relation between crude fibre content in biomass and the digestibility was determined using correlation analysis (Figure 2). y = -0.4728x + 75.081 R² = 0.762, r = 0.8729 Crude fibre, % (dry matter) Figure 2. Results of digestibility and crude fibre content correlation analysis The results obtained indicate close, negative correlation (r = 0.8729, P<0.05); thus digestibility reduces along with higher content of crude fibre in grass. Content of ADF has a significant effect on digestibility of fodder and on total energy value of the feed (Osītis, 2005). The highest ADF was recorded in festulolium (40.80 ±0.30%); in other species it was notably lower (P<0.05). The digestibility indicators varied notably among the grass species. The highest digestibility values were observed for timothy, tall fescue and festulolium (61.07 ± 0.62%, 60.01 ± 0.23% and 59.91 ± 0.11%, respectively), whereas the lowest was for meadow fescue Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 286 Digestibility, % (57.11 ± 0.23%). Correlation analysis helped to find out the closeness of the relation between ADF content of biomass and digestibility (Figure 3.). y = -0.786x + 89.158 R² = 0.9981, r = 0.9990 Acid detergent fibre, % (dry matter) Figure 3. Results of digestibility and ADF content correlation analysis Results of analysis show close, negative correlation (r = -0.9990, P < 0.05), thus as the ADF content in grass rises, the digestibility declines. Conclusions Nitrogen fertilizers increased the yield of dry matter. The highest DM yield in this research was found when grass received fertilizer at 180 kg N ha-1. There is close, negative correlation between ADF content in the grass biomass and the digestibility; also crude fibre content in biomass and digestibility show close, negative correlation. Higher crude fibre and ADF content in the grass decreased the digestibility of forage. Acknowledgements The research was supported by the grant of the Ministry of Agriculture of the Republic of Latvia, Agreement No. 2013/86. References Adamovics A. (1998) Zalaugi – stabilas lopbaribas bazes pamats (Grasses – stable base for livestock farming base). Latvijas Lopkopis un Piensaimnieks, 2, pp. 4-5.(in Latvian). Arhipova I. and Baliņa S. (2006) Statistika ekonomika (Statistics of the economy). Datorzinibu centrs, Riga, 352 lpp. (in Latvian). Berzins P., Bumane Sk. and Antonija A. (2001) Fosfora un kalija efektivitate ganibas atkariba no so uzturvielu nodrošinajuma augsne (Efficiency of phosphorus and potassium in pasture depending on supply of these nutrients in soil). Agronomijas Vestis, 3, pp.180 -185 (in Latvian). Bumane S., Bughrara S.S. (2003) Optimization of mineral nutrition for perennial ryegrass seed production. Proceedings of the Latvia University of Agriculture 8 (303), pp.70–73. Gutmane I. and Adamovich A. (2008) Analysis of sward management factors influencing Festulolium and Lolium × boucheanum yield formation. Latvian Journal of Agronomy No.10, pp. 117 – 122. Ositis U. (2005) Dzivnieku edinasana kompleksa skatijuma (Cattle feeding from complex viewpoint), Jelgava, Ozolnieki, 320 pp. (in Latvian). Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 287 Forage yield and protein content of five native species from Lanzarote (Canary Islands) Chinea E.1, Batista C.1, García-Ciudad A.2 and García-Criado B.2 1 Escuela Técnica Superior de Ingeniería Agraria, Universidad de La Laguna, Spain 2 Instituto de Recursos Naturales y Agrobiología (IRNASA-CSIC), Salamanca, Spain Corresponding author: echinea@ull.es Abstract The scarcity of fresh forage in the Canary Islands is a disadvantage for the development of cattle raising. Consequently, two-thirds of the food used for ruminants has to be imported. This dependence from the outside implies not only economic but also strategic problems for cattle exploitation. The objective of this work was to study the production and protein levels of five native valuable forage species from Lanzarote's Biosphere Reserve (Atriplex halimus, Bituminaria bituminosa var. albomarginata, Coronilla viminalis, Echium decaisnei and Lotus lancerottensis). This trial was developed in an experimental plot located on Lanzarote Island where three cuttings were carried out (winter and spring 2010 and winter 2011). The main forage production was obtained from B. bituminosa and the lowest from C. viminalis. On the other hand, the species with highest protein content were A. halimus and C. viminalis (166.4 g kg-1 and 153.2 g kg-1, respectively) while E. decaisnei showed the lowest protein mean value (111.0 g kg-1). The growth of these species could be of interest for dry zones. Keywords: sustainability, cattle raising, Canary Islands, drylands Introduction The aridity and drought periods in Lanzarote Island (Canary Islands, Spain) cause the development of an active desertification process affecting 31% of its surface. The use of native species for the revegetation of the soil minimizes this effect and offers a possible solution, since these plants show a series of favourable characteristics. These species show quick growth, deep roots and high drought tolerance, while restoring and regenerating soils thereby preventing erosion (Chinea et al., 2004). Furthermore, their growth can be a source of cattle feed during periods of forage scarcity. The sustainability of this type of agriculture can represent an attractive option for farming production in territories with limited resources of quality water, as occurs in most part of the Canary Islands. According to 2010 Integral Cattle Plan data, Canary Islands is the region of Spain with the highest shortage in cattle feed production. The aim of this work was to study the production and protein levels of five native valuable forage species from Lanzarote's Biosphere Reserve (Atriplex halimus, Bituminaria bituminosa var. albomarginata, Coronilla viminalis, Echium decaisnei and Lotus lancerottensis), to offer a system for the harnessing of these species with interest for arid and semi-arid regions. Materials and methods Seeds from wild populations in Lanzarote Island were collected in July 2008 and germinated in forest containers (February 2009). Subsequently, they were planted (July 2009) in an experimental plot with 1387 m2 located on the same island at 105 m above sea level (UTM X: 640.202; Y: 3.208.902). This plot showed a traditional mulching system and the soil comes from fertile plain-transported soils with clayish texture and alkaline pH (8.4), classified as Eutric Fluvisols. A random block design with four replications per species was established, employing sub-plots of 1.5 × 1.5 m. The plot was not fertilized during the experiment. Plants were irrigated with a dose of 1.33 mm/month/plant during the first year and 0.66 mm/month/plant the second year. The quality of water according to its electrical conductivity (EC = 0.54 mS cm-1) implies no employment restriction, while the relationship between the Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 288 sodium adsorption rate (SAR = 3.2) and EC showed a light to moderate infiltration problem (Ayers and Westcot, 1985). During the period of the experiment, a mean temperature of 21.1ºC and a rainfall of 122 mm were registered. For all the studied species, cuttings were made when the main regrowths were greater than 30 cm. Three cuttings were made in total: January 2010 (winter 2010), June 2010 (spring 2010) and January 2011 (winter 2011). After cutting, the weight of all the green vegetal material was determined and 1 kg of each replicate was selected. Later, the edible fraction (leaves, green regrowths, flowers and non-lignified stems with diameters less than 5 mm) was manually selected and separated from the non-edible fraction. The edible fraction was dried at 60ºC for 48 hours and the production of edible dry matter (EDM) was determined. Samples were analysed for crude protein (CP) using the Kjeldahl distillation method. Data were statistically analysed using analysis of variance (ANOVA), and least significant difference (LSD) test was used for means comparison (SPSS statistics 17). Results and discussion The highest EDM production was obtained from B. bituminosa (2.11±0.42 t ha-1 and cutting), followed by E. decaisnei (1.69±0.25 t ha-1 and cutting) and A. halimus (1.65±0.19 t ha-1 and cutting). Coronilla viminalis (0.77±0.19 t ha-1 and cutting) showed the lowest mean production (Table 1). Table 1. Mean values ± standard error of edible dry matter production (EDM) and crude protein (CP) per species. Species followed by different letters show significant differences (P < 0.05) EDM (t ha-1) CP (g kg-1) Atriplex halimus 1.65±0.19bc 166.4±6.8a Bituminaria bituminosa 2.11±0.42a 133.4±4.7b Coronilla viminalis 0.77±0.19c 153.2±11.2ab Echium decaisnei 1.69±0.25ab 111.0±8.9c Lotus lancerottensis 1.42±0.25abc 144.7±10.2ab The production of B. bituminosa was notably lower than that mentioned by Méndez (2000) (19.2 and 12.9 t ha-1), which reported a whole irrigation dose of 219 L m-2 and a high planting density (90000 plants ha-1) compared with that one used in the present work (4444 plants ha1 ). Considering seasonal cuttings, the highest EDM production was obtained in spring 2010 followed by winter 2010 and winter 2011 (Table 2). Significant differences among all seasons can be observed (P < 0.05). Table 2. Mean values ± standard error of edible dry matter production (EDM) and crude protein (CP) per season. Seasons followed by different letters show significant differences (P < 0.05). Winter 2010 Spring 2010 Winter 2011 EDM (t ha-1) 1.37±0.15b 2.14±0.28a 1.11±0.17c -1 a c CP (g kg ) 164.8±6.4 121.8±7.8 141.8±5.9b Atriplex halimus showed the highest CP levels (166.4 ± 6.8 g kg-1), while the lowest appeared in E. decaisnei (111.0±8.9 g kg-1) (Table 1). Significant differences were observed for the latter species compared with all the others. Coronilla viminalis (153.2±11.2 g kg-1), L. lancerottensis (144.7±10.2 g kg-1) and B. bituminosa (133.4±4.7 g kg-1) presented intermediate levels and no significant differences were observed among them. Additionally, no significant differences were determined regarding the CP content for C. viminalis, L. lancerottensis and A. halimus (Table 1). Therefore, it could be highlighted that leguminous species (C. viminalis and L. lancerottensis) showed similar CP content to that reported by García-Criado et al. (1986) for the reference forage, alfalfa (148.0 g kg-1). On the other hand, lower CP values were observed in B. bituminosa and E. decaisnei, while the highest was obtained in A. halimus. The highest CP concentration was obtained in winter 2010 followed by winter 2011, and the lowest mean values in spring 2010 (Table 2). Nitrogen content is higher when the growing conditions (temperature, light and water availability) are more favourable (McDonald et al., 1988) and N Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 289 content decreases with plant age. This explains why in winter, with the highest vegetative regrowth rates in this Island (Chinea et al., 2013), the highest CP contents are reached. Conclusions The studied species showed high protein content and therefore could be of interest as forage crop in arid zones with low amounts of irrigation and without need of fertilization. Acknowledgements This work has been financed by Fundación Biodiversidad (MARM) and Cabildo Insular de Lanzarote. Authors wish to acknowledge the collaboration of Ana Carrasco, María del Mar Duarte, Alejandro Perdomo and Francisco Pino. References Ayers R.S. and Westcot D.W. (1985) Water quality for agriculture. FAO. Irrigation and Drainage. Paper 29 Rev. 1, Roma, Italy. Chinea E., Mora J.L., García-Ciudad A. and García-Criado B. (2013) Forage production of Tagasaste (Chamaecytisus palmensis) and three Teline species cultivated in the Canary Islands (Spain), during a period of ten years. Revista de la Facultad de Agronomía. LUZ 30, 591-618. Chinea E., Rodríguez Rodríguez A. and Mora J.L. (2004) Erosion control on soil with shrub and endemic forage legumes from the Canary Islands. Revista de la Facultad de Agronomía. LUZ 21, 363-373. García-Criado B., García-Ciudad A., Rico M. and García-Carabias M.S. (1986) Composición químicobromatológica de alfalfa deshidratada destinada al comercio exterior. En: XXVI Reunión científica de la Sociedad Española para el Estudio de los Pastos. (SEEP), 2-6 Junio, 1986, Oviedo, Spain, pp. 71-87. McDonald P., Edwards R.A. and Greenhalgh J.F.D. (1988) Nutrición animal. 3ª Ed., Acribia, Zaragoza, Spain, 518 pp. Méndez P. (2000) El heno de Tedera (Bituminaria bituminosa): un forraje apetecible para el caprino. En: III Reunión Ibérica de pastos y forrajes; XL Reunión Científica de la Sociedad Española para el Estudio de los Pastos. (SEEP), 7-13 May 2000, Braganza-La Coruña-Lugo (Portugal-Spain), pp. 411-414. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 290 Effect of manure enriched with clinoptilolite on pasture yield and quality Simić A.1, Rakić, V.1, Marković, J.2, Dželetović Ž.,3 and Živanović I.1 1 Faculty of Agriculture, University of Belgrade, 11080 Zemun-Belgrade, Serbia 2 Institute for Forage Crops, Kruševac, Serbia 3 Institute for Appliance of Nuclear Energy, Zemun, Serbia Corresponding author: alsimic@agrif.bg.ac.rs Abstract Since legislation imposes restrictions on the use of mineral nitrogen-based fertilizers, traditional organic fertilizers (such is cattle manure) are being applied and are now becoming a significant factor for achievement required forage production. Cattle manure offers available NH4+ for pastures; however, ammonia losses through leaching or evaporation significantly reduce its fertilization efficiency in crop production. Therefore, there is a need for a binding agent that can preserve nitrogen reserves. It is known from the literature that natural microporous alumosilicates – zeolites – can retain many different chemical species. Our previous results show that the addition of 10% by weight of zeolite tuff (70% by weight of clinoptilolite) to the fresh cattle manures binds 90% of the ammonia present in cattle manure. In the present study, cattle manure enriched with clinoptilolite was applied as a fertilizer. The experiments were carried out on natural pastures in 2012/13 and included four different treatments: pure fermented manure (30 t ha-1); manure+zeolite (30 t ha-1+10% zeolite); nitrogen-containing mineral fertilizer (50 kg ha-1 N) and control. The fertilizers were applied in December 2012, except mineral fertilizer, which was applied in spring 2013. The dry matter (DM) contents were estimated and the chemical analyses of forage done for two cuts. The implication for the future use of zeolites in sustainable grassland systems are discussed. Keywords: clinoptilolite, manure, pasture, quality, yield Introduction Nitrogen (N) is crucially important as a yield-contributing element for most crop plants including forages. Restrictions in the use of mineral nitrogen in Europe (European Union Council Regulation No 834/2007) has changed the view on N fertilization of hill-mountainous grasslands of Serbia. Cattle manure as a fertilizer for pastures provides an available NH4+ source, but ammonia losses through leaching or evaporation significantly reduce efficiency of manure in crop production. Hence, there is a need for a binding agent that can preserve N reserves from manures. It is estimated that, for a period from excretion until collection to the farmyard manure (FYM) heap, that 40% of N is lost by evaporation during the warm half of year, while during a cold half of year 16% of N is lost (Moreira and Satter, 2006). After 6months storage in the FYM heap, N comprised in FYM is assumed to be 80% of the collected N, due to gaseous losses (volatilization and denitrification; Petersen et al., 1998). It has been shown previously that the addition of 10% (by weight) zeolite to fresh cattle manure increases retention of ammonia by 90% in comparison to the system without zeolite. Also, preliminary tests indicated that grass herbage yields obtained in a pot experiment containing ammonialoaded zeolite were higher compared to control pots (Simić et al., 2013). Kavoosia (2007) reported that clinoptilolite decreased the NH4-N loss from the soil and enabled an easier uptake of nitrogen by the plant. The objective of this study was to investigate and launch the sustainable ammonia source for pastures in Serbia, which is based on natural zeoliteclinoptilolite mixed with manure, compared mutually and with mineral N supply. It is expected that the application of zeolite with manure can increase N use efficiency. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 291 Materials and methods Field plots (5m x 2m) were established on natural pastures in 2012/13 included five different treatments: a) pure manure (30 t ha-1); b) manure+zeolite (30 t ha-1+10 wt.% zeolite); c) pure zeolite (3 t ha-1); d) nitrogen application by mineral fertilizer (50 kg ha-1 N) and e) control. The field trial was established in the vicinity of Šabac, Serbia, by the method of RCB design of plots in 4 replications. The zeolitic tuff (Zlatokop deposit, Vranjska Banja; containing 70% of clinoptilolite: Ca1.6Mg0.7K0.7Na0.3Al5.5Si26O72·23H2O, grain size in the range 0.063-0.1 mm) was used as the additive to manure in this work. Prior to application, fresh cattle manure was homogenously mixed with the natural zeolite – clinoptilolite, and fermented during 3 months. The treatments with zeolite, manure and a mixture zeolite+manure were applied in autumn. Spring nitrogen application of mineral fertilizer was performed at the beginning of the vegetation season. The plots were harvested in May and the dry matter (DM) of the harvests was measured. Neutral Detergent Fibre (NDF) and Acid Detergent Fibre (ADF) were determined according to procedure by Van Soest; protein fractions (true protein and NPN) were determined as described by Licitra et al. (1996). The main characteristics of the soils were determined: pH in CaCl2 was 5.07 while pH in H2O was 5.73; the contents of P2O5 and K2O were 19.8 mg kg-1 and 115.1 mg kg-1 respectively, total N was found to be 0.16%. Total DM yield and pasture DM quality in each of the two cuts were analysed by analysis of variance (ANOVA) and LSD test, in order to recognize significant effects of fertilization treatments. Results and discussion Fertilizer treatments affected yield, especially in the first cut: the yield was doubled by fertilization, in comparison with the control treatment (Table 1). Table 1. Pasture dry matter yield and forage quality per cut. I treatment-control, II treatment- zeolite (3 t ha-1), III treatment- manure (30 t ha-1), IV treatment- manure+zeolite (mixed) (30 t ha-1), V treatment- mineral nitrogen (50 kg ha-1); DM – dry matter (t ha-1), CP – crude proteins (% DM), CCl - crude cellulose (% DM), TP - true proteins (% of crude proteins), NPN-non protein nitrogen (% of crude proteins), NPN(%solP)-percentage of soluble protein that is non-protein nitrogen. Treatments I cut I II III IV V Treatments II cut I II III IV V DM yield CP CCl NDF ADF TP NPN 1.99c 2.38c 4.53a 4.11ab 3.31b DM yield Total yield 9.71a 9.98a 10.1a 9.86a 10.5a CP 31.9a 33.0a 32.7a 33.3a 32.0a CCl 62.8a 63.8a 63.9a 63.4a 62.3a NDF 38.4a 37.8a 38.7a 38.0a 38.7a ADF 73.6ab 74.2ab 64.8b 83.5a 69.6ab TP 26.4ab 25.8ab 35.2a 16.5b 30.4ab NPN NPN (%solP) 89.4ab 84.6ab 100a 65.7b 96.2a NPN (%solP) 1.02ab 0.91b 1.38a 0.97ab 0.96ab 3.01c 3.29c 5.91a 5.08ab 4.27b 9.04a 9.02a 9.72a 9.05a 9.77a 25.9a 24.4a 25.3a 25.0a 27.3a 69.5b 66.6ab 66.7ab 63.1a 66.8ab 37.1a 41.3a 38.8a 37.9a 41.0a 84.2ab 86.5ab 73.9b 91.0a 80.7ab 15.8ab 13.5ab 26.1a 9.0b 19.3ab 79.3a 71.5a 78.6a 85.3a 93.8a Manure showed some extended effect in the second cut, while the pure zeolite treatment had diminishing effects on yield in the arid conditions during summer regrowth. The most promising treatments are those with manure (enriched by zeolite or pure manure). Spring N application was also a good way to increase yield, whereas the pure zeolite and control treatments achieved the lowest yield. These results are in accordance with Gholamhoseini et al. (2013) who found that reducing of N leaching in the presence of natural zeolite, such as clinoptilolite, increases plant-available N and consequently increases N use efficiency. Also, Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 292 the present results obtained in Serbia confirm previously published data on natural zeolite combined with manure, which has been used successfully to remediate soils having unfavourable chemical properties as well as to enhance the crop yield on them (Glisic et al., 2009). It is very important to point out here that results obtained in this work indicate that the usage of zeolite-based fertilizer leads to an increase in the true protein content and decrease of non-protein nitrogen in grasses (Table 1), which can have a positive influence on forage digestibility. It may be concluded that the high true protein content of the plants grown on the plots with clinoptilolite can be caused by their ability to increase nitrogen uptake from soil due to the effect of clinoptilolite. It is in accordance with the conclusion of Gevrek et al. (2009) that the high protein content of the rice plants grown with clinoptilolite could be caused by their ability to increase nitrogen uptake. Conclusion The results obtained show that the use of cattle manure enriched with natural zeolite can be used as a fertilizer for pastures and this contributes to a preservation of nitrogen. The application of such a fertilizer may reduce the application of mineral N fertilizers on natural pastures. Based on the results presented here, natural zeolite can be recommended for agricultural purposes in terms of sustainable fertilizing and improving the system of cattle farm - manure - organic fertilizer for forage crops. Future studies should focus on including additional sites with different soil types in contrasting climatic areas. Acknowledgement This research is supported by the Norwegian Programme in Higher Education, Research and Development (Project: The use of natural zeolite (clinoptilolite) for the treatment of farm slurry and as a fertilizer carrier). References Gholamhoseini M., Ghalavand A., Khodaei-Joghan A., Dolatabadian A., Zakikhani H., Farmanbar E. (2013) Zeolite-amended cattle manure effects on sunflower yield, seed quality, water use efficiency and nutrient leaching. Soil & Tillage Research 126, 193-202. Gevrek M.N., Tatar Ö., Yağmur B. and Özaydin S. (2009) The effects of clinoptilolite application on growth and nutrient ions content in rice grain. Turkish Journal of Field Crops 14 (2), 79-88. Glisic I.P., Milosevic T.M., Glisic I.S. and Milosevic N.T. (2009) The effect of natural zeolites and organic fertilisers on the characteristics of degraded soils and yield of crops grown in western Serbia. Land Degradation and Development 20, 33–40. Kavoosia M. (2007) Effect of zeolite application on rice yield, nitrogen recovery and nitrogen use efficiency. Communications in Soil Science and Plant Analysis 38, 69-76. Licitra G., Hernandez T.M. and Van Soest P.J. (1996) Standarrdization of procedures for nitrogen fractionation of ruminant feeds. Animal Science and Feed Technology 57, 347-358. Moreira V.R. and Satter L.D. (2006) Effect of scraping frequency in a freestall barn on volatile nitrogen loss from dairy manure. Journal of Dairy Science 89 2579-2587. Petersen S.O., Lind A.M. and Sommer S.G. (1998) Nitrogen and organic matter losses during storage of cattle and pig manure. Journal of Agricultural Science 130, 69-79. Simić A., Milovanović J., Grbić Alibegović S., Raičević S., Rakić V., Krogstad T. and Rajić N. (2013) Zeolite as a binding agent for ammonia ions and as a soil additive. Part II: Effect on grass growth and quality. Proceedings of the 5th Serbian-Croatian-Slovenian Symposium on Zeolites, pp 96-99. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 293 Influence of long-term organic and mineral fertilization on Festuca rubra L. grassland Păcurar F., Rotar I., Balazsi A., and Vidican R. University of Agricultural Sciences and Veterinary Medicine Cluj-Napoca, Cluj-Napoca, Romania, 400372 Corresponding author: fpacurar@gmail.com Abstract Montane semi-natural grasslands are economically, ecologically and socially important, and therefore they require certain conservation measures. This paper summarizes the impact of organic, mineral and combined fertilization on the floristic composition of Festuca rubra L. grassland and includes results from an experimental field, which was established in the Apuseni Mountains (Romania). Fertilizer applications resulted in pronounced sward changes, depending on quantities applied. Even fertilization using low quantities changes the grassland type, so future conservation measures must consider this situation. Keywords: semi-natural grasslands, organic fertilization, mineral fertilization Introduction Fertilization may generate major changes in swards and lead to a rapid reduction in species richness. Plant diversity is influenced by fertilizer type and amount, and by soil and climate (Vîntu et al., 2011). Combining organic fertilization with mineral fertilization might be a good solution for maintaining the grasslands’ species richness and also be associated with an appropriate increase the dry matter yield (Samuil et al., 2012). Montane grasslands are important parts of the high natural-value domains, and in Europe they show great species richness. They could help increase local income from tourism and also be a seed source for restoring local biodiversity (Hopkins, 2011). Festuca rubra L. grasslands occur as fragmented areas in Central Europe, while in Carpathian Mountains they cover large areas. Their persistence considerably depends on their management. This paper’s goal is to emphasize the impact of organic, mineral and combined fertilization upon the floristic composition in order to identify optimum conservation solutions for Festuca rubra L. grasslands. Materials and methods An experimental field was set up in 2001 in Gheţari village, Apuseni Mountains (Romania) at 1130 m elevation. The plots were arranged using the random blocks method with 4 variants in 5 replications (T1 - control; T2 - 10 t ha-1 manure; T3 - 10 t ha-1 manure + 50N 25P 25K; T4 100N 50P 50K; T5 - 10 t ha-1 manure + 100N 50P 50K). The manure was provided by the cattle and horse husbandry. The inputs were applied each year in early spring. The floristic studies were performed according to the Braun-Blanquét method. The floristic data were processed by PC-ORD (McCune and Mereford, 2002). Vegetation ordination was done by Non-metric Multidimensional Scaling (NMS). NMS is generally the best ordination method for community data (Peck, 2010). The distance calculation was performed by Sorensen (Bray-Curtis) index. NMS was applied five times by the option Autopilot (Autopilotmode). In order to verify the final solution, the ordination was manually performed (6 axes, 250 runs, 250 iterations). The program recommended in both cases the two-dimensional representation to a stress in real data of 5.355 (P<0.01). Cumulative coefficient of determination for the correlations between ordination distances and distances in the original n-dimensional space was 0.909 (Axis 1 – 0.907, Axis 2 – 0.002). Also the MRPP (Multi–response Permutation Procedures) was used for testing the hypothesis of no differences between two or more groups of entities. The method implies the statistic test T which describes the separation between the groups. The P value is Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 294 useful for evaluating how likely it is that an observed difference is due to chance. The agreement statistic A describes within-group homogeneity, compared to the random expectation. A Monte Carlo test of significance is included. Results and discussion Axis 2 Festuca rubra L. plant community (control) is submitted to profound sward changes as a consequence of application of organic and mineral inputs application (Figure 1). T2R1 F.rubra (27.5%)-A.capillaris (11.25%) grassland Cynocris Manure 10 Mg ha-1 Carucarv T2R2 Tragprat T2R3 T2R4 Trifprat Trifrepe Planlanc Leucvulg Lotucorn Hieraura Hierpilo Prunvulg Thympule Anthodor Luzumult Knauarve Euphcarn T1R1 Cyperacea-Juncaeae Cardhall T3R3 Manure 10 Mg ha-1 + 50N25P25K kg ha-1 T3R1 T3R2 A.capillaris (27.5%)-F.rubra (11.25%) grassland T3R4 Achimill Troleuro Verocham Fabaceae Myossylv Taraoffi T5R3 Rhinmino Vicicrac Centpseu Trisflav Ranuacri Pimpmajo Crepbien A.capillaris (37.5%)-F.rubra (15.94%) grassland Colcautu Forbs T5R1 Stelgram Hypemacu Agrocapi T5R2 Camppatu Violtric Alchvulg Festrubr Axis 1 T5R4 Poaceae Rumeacet Lathprat T4R2 Festprat T4R1 100N75P75K kg ha-1 T4R3 A.capillaris (50%)-F.rubra (16.9%) grassland Gymncono Planmedi Poteerec F.rubra (27.5%) grassland Arnimont Polyvulg Carepall Carlacau Campabie T1R4 T1R2 Verooffi T1R3 T4R4 Figure 1. Influence of organic and mineral fertilization upon the floristic composition (T – treatments, R – repetitions) Treatment performance generated the fixation of specific plant communities. When 10 t ha-1 manure were applied, besides the dominant species Festuca rubra L., Agrostis capillaris L. also occurred and became co-dominant (11.25%) and a large difference between T1 and T2 plant communities was registered (P<0.01, Table 1). Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 295 Table 1. The pairwise comparison between treatments with MRPP (T 1, T2,…T5– treatment,T- the t test, A - group homogeneity, P - statistical significance). Comparison T A P T1vs. T2 -4.10145197 0.41940328 0.00613211 T1 vs.T3 -4.28029428 0.59504554 0.00613211 T1 vs. T4 -4.39304453 0.64634213 0.00574951 T1 vs. T5 -4.33791511 0.63250901 0.00593859 T2 vs. T3 -4.11856121 0.33504342 0.00583843 T2 vs. T4 4.41742256 0.59348974 0.00565491 T2 vs. T5 -4.35817008 0.49167661 0.00567896 T3 vs. T4 -4.31111198 0.55182529 0.00593038 T3 vs. T5 -4.04164213 0.37275911 0.00650907 T4 vs. T5 -4.19750892 0.42446389 0.00606754 An increase in fertilizer quantity was marked by a greater share of Agrostis capillaris L. and a reduction of Festuca rubra L., which became co-dominant (Figure 1). When mineral fertilizers (T4) were applied, the greatest proportion (50%) of Agrostis capillaris L. species was observed, whereas combining organic with mineral fertilizers (T3, T5) did not produced an extreme ratio between the co-dominant species. Among the plant communities generated by the three treatments (T3, T4, T5), pronounced floristic differences were observed (P<0.01, Table 1), despite the fact that the plant community contained the same species. When performing a floristic study in the central part of Apuseni Mountains, Gârda (2010) all three plant communities were encountered, Festuca rubra L. showing the lowest share. Related to the economic groups of plants, it was noticed in the present experiment that grasses are favoured when large quantities of fertilizer are applied (T4, T5), legumes respond well to organic (T2) and organic-mineral (T3) fertilization, sedges and rushes did not prefer fertilization and forbs show the higher share by administrating T2 and T3 (Figure 1). T2, T3, T4 treatments explain as statistically significant the floristic variation, whereas the particular effect of T5 is difficult to distinguish from T4 or T3. In the control plant community there are numerous nitrophobic species which decline in proportion when other treatments were applied, or even disappear from the sward (Arnica montana L., Veronica officinalis L., Carlina acaulis L., Campanula abietina Griseb., Gymnadenia conopsea (L.) R. Br. etc.). Certain species prefer organic fertilization (T2), such as: Cynosurus cristatus L., Tragopogon pratensis L., Cardaminopsis halleri (L.) Hayek., etc. Organic-mineral fertilization in moderate quantities (T3) favours the following species: Veronica chamaedrys L., Taraxacum officinale Weber ex F.H.Wigg., Trollius europeaeus L., Pimpinella major (L.) Huds. etc. When large quantities of mineral fertilizers (T4) and organic-mineral fertilizers (T5) were applied, Trisetum flavescens (L.) Beauv., Rumex acetosella L. s. l., Lathyrus pratensis L. and Festuca pratensis L. species became established in the sward. Conclusions Dominance of Festuca rubra L. in grasslands is fragile. An annual application of 10 t ha-1 of organic manure does not maintain the control plant community with Festuca rubra L. as the dominant species. The evolution of Festuca rubra L. grassland brings a decline in the share, or even disappearance, of some oligotrophic species. In order to identify optimum possibilities to Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 296 preserve Festuca rubra L. grassland, studies that test other possibilities to apply organic and mineral fertilizers are necessary. References Gârda N. (2010) Studiul unor elemente de landșaft montan (cu privire specială asupra ecosistemelor de pajiști din comuna Gârda de Sus, Munții Apuseni), Teză de doctorat, USAMV Cluj- Napoca, pp. 160-224. Hopkins A. (2011) Mountainous farming in Europe. Grassland Science in Europe 16, 8-12. McCune B. and Grace J.B. (2002) Analysis of Ecological Communities. Gleneden Beach, Oregon, USA: MJM Software Design, pp. 125-143. Peck J. (2010) Multivariate analysis for community ecologists: step-by-step using PC-ORD. Gleneden Beach, Oregon, USA: MJM Software Design,162 pp. Samuil C., Vîntu V., Sarbu C. and Popovici I. (2012) Influence de la fertilization sur la végétation et la production d’une prairie á Festuca valesiaca L. Fourrages 210, 151-157. Vîntu V., Samuil C., Rotar I., Moisuc A., Razec I. (2011) Influence of the management on the phytocoenotic biodiversity of some Romanian representative grassland types. Notulae Botanicae Horti Agrobotanici 39, 119125. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 297 Effects of low-input treatments on Agrostis capillaris L. - Festuca rubra L. grasslands Rotar I., Păcurar F., Balázsi Á., Vidican R., Mălinaş A. University of Agricultural Sciences and Veterinary Medicine Cluj-Napoca, Grassland and Forage Crops Department, No 3-5, Manastur Street, 400372, Cluj-Napoca, Romania Corresponding author: fpacurar@gmail.com Abstract Traditional grassland management in the Apuseni Mountains has led to high biodiversity of the grassland ecosystems. The greatest risk for the future is the abandonment of these grasslands. In order to maintain and conserve these grassland systems the identification of new sustainable alternatives is required that area also economically sustainable. An experiment was installed in the Apuseni Mountains in 2009 to compare the effects of mowing, mulching, organic fertilization with mulching, and abandonment on oligotrophic grasslands. The highest number of species was influenced by organic fertilization combined with mulching. To maintain Agrostis capillaris L. - Festuca rubra L. phytocoenosis, the following treatments can be taken into account: organic fertilization combined with mulching, mowing, mulching. Keywords: mulching, organic fertilizers, management, grassland, species composition Introduction Most of the species-rich semi-natural grasslands from different European regions, which are managed in traditional systems, are threatened either by intensification or abandonment. In the Apuseni Mountains (Romania) large areas of oligotrophic grasslands were maintained by applying an extensive and traditional management over a long period, where the most important links were organic fertilizers and mixed use (Gârda, 2010). The main economic resource (wood resources) of the region has suffered a drastic and continuous decline, which entails depopulation of the area. Despite the existence of subsidy programmes, areas of oligotrophic grasslands are abandoned. Thus, the identification of new, viable and low-cost alternatives is required in order to maintain and conserve these grassland ecosystems, combining traditional management elements with new ones for the region. Mulching could be an inexpensive alternative for the conservation of spices-rich semi-natural grasslands (Tonn and Briemle, 2010). The aim of the research was to follow the effects of mowing, mulching, organic fertilization combined with mulching, and abandonment, on plant-species composition of Agrostis capillaris L. - Festuca rubra L grassland and to identify new solutions for maintenance and conservation of these grassland ecosystems. Materials and methods The experiment started in 2009 in Poienile Ursului (1349 m), Ocoale Village, Gârda de Sus Commune, Apuseni Mountains, using a randomized block design with 7 treatments in 5 replications: T1 - control (mown 1 / year); T2 - mulch 1 / year; T3 - mulch 1 / year + 5 t ha-1 manure / year; T4 - mulch 1 / year + 5 t ha-1 manure / 2 years; T5 - mulch 1 / year + 10 t ha-1 manure / 2 years; T6 - mulch 1 / year + 10 t ha-1 manure / 3 years; T7 - abandonment. Cattle and horse manure was applied in early spring, according to the treatments. Floristic composition was interpreted by the Braun-Blanquet method and plots were then mulched once each year (August), the same time as the traditional mowing period. Floristic data processing was performed with PC-ORD version 6, which uses multivariate analysis of the data entered into the spreadsheet. This program focuses on nonparametric tools, on graphics, randomization tests, and bootstrapped confidence intervals for analysis of community data (McCune and Grace, 2011). For data processing and interpretation we used multidimensional scaling (NMS); Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 298 well suited to data coordination that are not normal, discontinuous, or otherwise questionable (Peck, 2010). This paper summarizes the results of the fifth experimental year, showing the effect of treatments on plant species composition of grassland. NMS was carried out several times in autopilot mode, in order to minimize the stress. Distance measurement was done with Sorensen (Bray - Curtis). The recommended solution for data presentation was tridimensional. The coefficient of determination (r2) for the correlations between ordination distances and distances in the original n-dimensional space was 0.740 (Axis1 - 0.421; Axis2 - 0.169; Axis3 0.151). NMS applied at plots level showed no significant results; for this reason we chose not to present these results within this paper. Results and discussion Changes in vegetation determined by different treatments, took place within the Agrostis capillaris L. - Festuca rubra L. phytocoenosis. Statistically significant experimental factors which explain the changes in floristic composition are: mowing, abandonment, mulching, organic fertilization combined with mulching (Fig.1a, Fig.1b). Axis 1 is correlated with organic fertilization combined with mulching (r = -0.395), Axis 2 with mowing (r = 0.317) and abandonment (r = - 0.317) and Axis 3 with organic fertilization combined with mulching (r = - 0.555) and mulching (r = 0.449). The gradual increase of organic fertilizers dose combined with mulching determines the increase in share of the following species: Agrostis capillaris L., Plantago lanceolata L., Rhinanthus minor L., Rumex acetosa L., Veronica chamaedrys L., etc. Similar results have been obtained by Gaisler et al. (2004), where the species Veronica chamaedrys L., Hypericum maculatum Crantz, etc. were favoured by mulching once a year and Agrostis capillaris L., Plantago lanceolata L., Rumex acetosa L., Trifolium pratense L. were favoured by mulching three times per year. Veronica chamaedrys L. presents a relative tolerance for shadow conditions and being able to resist also under a dense layer of litter (Pavlů et al., 2003) and Rhinanthus minor L. is characteristic of species rich semi-natural grasslands with low productivity (Ellenberg et al., 1991), being well known for taking advantage of hosts with high concentration in N (Ameloot et al., 2008). The application of organic fertilizers combined with mulching led to a decrease in the percent of participation of the following species: Arnica montana L., Centaurea mollis Waldst. et Kit., Festuca rubra L., Polygala vulgaris L., Silene nutans L., Thymus pulegioides L. Mowing favours Colchicum autumnale L., Euphrasia officinalis L., Ranunculus acris L. and Trollius europaeus L. The abandonment determines the accentuated cover of Hieracium aurantiacum L., Lotus corniculatus L., and Viola tricolor L. Abandonment, even though it did not show a strong influence on vegetation, is not considered as a viable solution because it has been scientifically proven that in the long term it may produce profound changes. In our experiment mulching and/or fertilization combined with mulching favour Trifolium repens L. and Trifolium pratense L. (Fig. 1b). Effects of mulching, in particular at the species level, are very low. Mulching should be an acceptable solution, especially for already abandoned grasslands, because it could prevent the return of woody species (Gaisler et al., 2011). Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 299 Axis 2 Knaudips Lotucorn Thympule Arnimont Silenuta Polyvulg Centmoll Hieraura Cirseris Primveri Galimoll Abandonment Alchvulg Agrocapi Festrubr Violtric Leucvulg Stelgram Verocham Trifrepe Veraalbu Ceraholo Rhinmino Fertilisation and mulching Polycomo Anthodor Planlanc Linucath Rumeacet Hypemacu Luzucamp Centpseu Trifprat Planmedi Axis 1 Mowing Ranuacri Troleuro Pimpmajo Euphoffi Colcautu Trifrepe Axis 3 Figure 1a. Influence of the treatments on plant species composition (Axis 1 and Axis 2) Centmoll Fertilisation and mulching Trifprat Mulching Rhinmino Planlanc Colcautu Cirseris Ranuacri Alchvulg Hieraura Linucath Agrocapi Violtric Luzucamp Stelgram Rumeacet Verocham Polyvulg Galimoll Euphoffi Ceraholo Leucvulg Primveri Lotucorn Anthodor Axis 1 Polycomo Troleuro Thympule Festrubr Planmedi Pimpmajo Hypemacu Knaudips Veraalbu Silenuta Arnimont Centpseu Figure 1b. Influence of the treatments on plant species composition (Axis 1 and Axis 3) Agrocapi - Agrostis capillaris L., Alchvulg - Alchemilla vulgaris L.; Arnimont - Arnica montana L.; Centmoll Centaurea mollis Waldst. et Kit.; Centpseu - Centaurea pseudophrygia C.A.Mey.; Ceraholo - Cerastium holosteoides Fr.; Cirseris - Cirsium erisithales (Jacq.) Scop.; Colcautu - Colchicum autumnale L.; Euphoffi Euphrasia officinalis L.; Festrubr – Festuca rubra L., Galimoll - Galium mollugo L.; Hieraura - Hieracium aurantiacum L.; Hypemacu - Hypericum maculatum Crantz; Knaudips - Knautia dipsacifolia Kreutzer; Leucvulg - Leucanthemum vulgare Lam.; Linucath - Linum catharticum L.; Lotucorn - Lotus corniculatus L.; Pimpmajo Pimpinella major (L.) Huds.; Planlanc - Plantago lanceolata L.; Planmedi - Plantago media L.; Polycomo Polygala comosa Schkuhr; Polyvulg - Polygala vulgaris L.; Primveri - Primula veris L.; Ranuacri - Ranunculus acris L.; Rhinmino - Rhinanthus minor L.; Rumeacet - Rumex acetosa L.; Silenuta - Silene nutans L.; Stelgram Stellaria graminea L.; Thympule - Thymus pulegioides L.; Trifprat - Trifolium pratense L.; Trifrepe - Trifolium repens L.; Troleuro - Trollius europaeus L.; Veraalbu - Veratrum album L., Verocham - Veronica chamaedrys L.; Violtric - Viola tricolor L. Conclusion After a study undertaken over a period of 5 years, the experimental factor that showed an influence on the greatest number of species was the organic fertilization combined with mulching, whereas mulching had the lowest influence. To maintain Agrostis capillaris L. Festuca rubra L. phytocoenosis, the following treatments can be taken into account: organic fertilization combined with mulching, mowing and mulching. A particular attention should be paid to organic fertilization combined with mulching, because intensification of this factor could have an influence on high number of species. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 300 References Ameloot E., Verlinden G., Boeckx P., Verheyen K. and Hermy M. (2008) Impact of hemiparasitic Rhinanthus angustifolius and R. minor on nitrogen availability in grasslands. Plant and Soil 311, 255-268. Ellenberg H., Weber H.E., Düll R., Wirth V., Werner W. and Paulissen D. (1991) Zeigerwerte von Pflanzen in Mitteleuropa. Scripta Geobotanica 18, 1 - 248. Gaisler J., Hejcman M. and Pavlů V. (2004) Effect of different mulching and cutting regimes, on the vegetation of upland meadow. Plant Soil Environment 50, 324 - 331. Gaisler J., Pavlů V. and Pavlů L. (2011) Effect of different extensive management treatments on the plant diversity of an upland meadow without forage utilization. Grassland Science in Europe 16, 577 - 579. Gârda N. (2010) The study of some mountainous landscape elements (with special regard to grassland ecosystems in Gârda de Sus Commune, Apuseni Mountains). PhD Thesis. McCune B. and Grace J.B. (2002) Analysis of Ecological Communities. Gleneden Beach, Oregon, USA: MJM Software Design, pp. 188-198. Pavlů V., Hejcman M., Pavlů L. and Gaisler J. (2003) Effect of rotational and continuous grazing on vegetation of the upland grassland in the Jizerské hory Mts., Czech Republic. Folia Geobotanica 38, 21 - 34. Peck J. (2010) Multivariate analysis for community ecologists: step-by-step using PC-ORD. Gleneden Beach, Oregon, USA: MJM Software Design,162 pp. Tonn B. and Briemle G. (2010) Minimum management intensity for maintaining and improving biodiversity of a mesotrophic semi-natural grassland. Grassland Science in Europe 15, 745 - 749. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 301 Influence of fertilization on the biodiversity of Festuca rubra L. and Agrostis capillaris L. grassland Samuil C.1, Vintu V.1, Popovici C.I.1 and Stavarache M.1 1 University of Agricultural Sciences and Veterinary Medicine Iasi 3, 700490, Iasi, Romania Corresponding author: csamuil@uaiasi.ro Abstract The importance of semi-natural grasslands in Romania is shown by the area they occupy (4·9 million ha, which is 33% of the total agricultural area of the country) and their comparatively high biodiversity. In terms of area occupied by natural grasslands in Europe, Romania occupies fifth position after France, Britain, Spain and Germany. In Romania, meadows belonging to this category occupy an area of approximately 1,600,000 hectares and give relatively low production. The production of this semi-natural grassland is influenced by some natural and management factors. The objective of this paper was to evaluate the effects of fertilization upon the sward, in order to recommend certain versions which have minor repercussions upon the plant diversity. The experiment was set up in 2006 in mountain grassland of Festuca rubra L. and Agrostis capillaris L. In the experiment the effects of management treatments was evaluated on productivity and biodiversity of the grasslands. The use of a fertilizer management regime based on small quantities of organic nutrients can lead to larger yields while, at the same time, improving biodiversity. Keywords: species richness, biodiversity, grassland, manure fertilization Introduction Semi-natural grasslands, traditionally used as forage for ruminants, are an important feature of land use in Europe, and cover more than a third of the European agricultural area (Pacurar et al., 2012. Their fertilization with manure is considered an appropriate management to conserve biodiversity value (Peeters et al., 2004). Several management factors may affect biodiversity of these grasslands including fertilization, overseeding, grazing and cutting management (Samuil et al., 2012). In Romania, grasslands are an important forage resource, but irrational management systems during the last periuod have led to their present state of degradation (Vintu et al., 2011). In this study, an experimental approach was used to evaluate the effects of management treatments on the biodiversity of Festuca rubra L. and Agrostis capillaris L. grassland. This paper presents the results of two experiments located at Pojorata, Suceava county, on natural grasslands of different floristic compositions. Materials and methods The experiment was performed on a meadow of Agrostis capillaris and Festuca rubra, at 707 m elevation on a slope of 20%, in Pojorata. The area has average temperatures of 6.3 oC and 708 mm total annual precipitation. During April to September average temperature is 12.8oC with 514 mm rainfall. In terms of its climate, the area is situated to the north-east extremity of the European Central Province, with a temperate climate-moderate-continental, and some influence of the eastern continental climate and of the northern boreal climate. In the area where this research was conducted, the permanent grassland occupies an area of 156,000 ha, with a different distribution according to altitude, slope, temperature and precipitation. The experiment was arranged in subdivided parcels, with four replications, and a plot size of 4 m x 5 m. Fertilization methods used and the amounts of nutrients supplied to the treatments were: A - dose of manure: A0 – 0 t ha-1, A1 – 10 t ha-1, A2 - 20 t ha-1, A3 - 30 t ha-1; B - period of application of manure: b0 – no manure, b1 - annually, b2 - every two years, b3 - every three years; C- mineral nitrogen which was applied every year at rates of: c1 – 30 kg ha-1, c2 - 50 kg Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 302 ha-1; and UAC - apparent coefficient of use (0·4 - for the manure applied annually; 0·45 - for the manure applied every two years; 0·55 - for the manure applied every three years). The chemical composition of 1000 kg of manure was 5.19 kg N, 2.83 kg P and 6.72 kg K. The manure was applied during the autumn, whereas the mineral nitrogen was applied in spring each year before vegetation growth began. The grassland was managed by cutting with a Bertolini 411 harvester, and swards mown to a height of 4-5 cm above the ground. The harvesting was performed by mowing throughout the heading stage of the dominant grass. The final yield was expressed as dry matter (t DM ha -1). Plant species composition analysis was performed using the number of species, the Shannon Wiener index, the Simpson index and the evenness index. In order to calculate these values, surveys were conducted at five points for each experimental parameter. We recorded presence/absence data and also the plant cover. The cover of all vascular plant species was visually estimated in each plot based on the Braun-Blanquet methodology. To eliminate edge effects, relevés were taken in the centre of each 4 m × 5 m plot within an area of 2 m × 3 m in mid-June 2013. The total number of vascular plant species was counted directly in the field (Cristea et al., 2004). Results and discussion The botanical composition was very weak and was represented by species of low forage value (Table 1). Table 1. Floristic composition and characterization of identified species (after Ellenberg et al., 1992). Indicators Ellenberg* Indicators Ellenberg* Species Species L T W R Tr L T W R Tr Agrostis capillaris 7 x x x 3 Trifolium repens 8 x x x 7 Anthoxanthum odoratum x x x 5 x Achillea millefolium 8 x 4 x 5 Arrhenatherum elatius 8 6 5 7 7 Alchemilla vulgaris 6 4 6 x 6 Brachypodium pinnatum 6 5 4 7 4 Carum carvi 8 4 5 x 6 Briza media 8 x x x 3 Leucanthemum vulgare 7 x 4 x 3 Cynosurus cristatus 8 5 5 x 4 Colchicum autumnale 5 5 6 7 x Dactylis glomerata 7 x 5 x 6 Filipendula vulgaris 8 7 4 x 3 Festuca pratensis 8 x 6 x 6 Galium verum 7 5 4 7 3 Festuca rubra x x 5 x x Hypericum perforatum 7 x 4 x x Holcus lanatus 7 6 6 x 5 Knautia arvensis 7 5 4 x 3 , 8 x x 2 x Plantago lanceolata 7 x x x x Poa pratensis 6 x 5 x x Plantago major 8 x 5 x 6 Trisetum flavescens 7 x x x 5 Plantago media 7 x 4 8 3 Anthylis vulneraria 8 x 4 8 3 Prunella vulgaris 7 x x 4 x Lotus corniculatus 7 x 4 7 4 Ranunculus polyanthemos 6 x 4 x 3 Medicago lupulina 7 5 4 8 x Rhinanthus rumelicus 7 5 4 x 3 Trifolium campestre 8 6 4 6 3 Rumex acetosa 8 x x x 6 Trifolium montanum 7 x 3 8 2 Taraxacum officinale 7 x 5 x 8 Trifolium pratense 7 x x x x Thymus pulegioides 8 x 4 5 3 Nardus stricta * Ecological indicators, according to Ellenberg et al., 1992: L = light value; T = temperature value; W = soil moisture value; R = soil (water) acidity (pH) value; Tr = trophic value Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 303 At the beginning of the experiment 38 species were identifies, of which 13 belonged to the family Poaceae, 7 to the Fabaceae, and 18 species belonged to other families. For these species some Ellenberg indicators were given. The following species were associated with the control: Achillea millefolium, Agrostis capillaris, Briza media, Filipendula vulgaris, Nardus stricta, Plantago lanceolata and Potentilla erecta. In contrast, Arrhenatherum elatius, Dactylis glomerata, Festuca rubra, Lotus corniculatus, Trifolium pratense and Trifolium repens were associated with the fertilizer N and manure treatments. The analysis of biodiversity parameters highlights (this was not tested statistically) that the number of species increased in all the variants where fertilizers were applied, compared to the control (Table 2). Table 2. Diversity parameters determined in 2013 Variant of fertilization Species richness (no.) Shannon index Shannon evenness Simpson index (D) A0b0 16 2.304 0.748 0.102 A1b1 20* 2.634* 0.823* 0.071* A2b2 27** 2.742* 0.754 ns 0.082* A3b3 18 ns 2.421 ns 0.814 ns 0.089 ns A2b1+A1b2+ A0b0 24* 2.534* 0.721 ns 0.092 ns A2b1+ A0b0+ A1b3 21* 2.427 ns 0.826* 0.086* A2b1+A1b2+ A1b3 19 ns 2.441* 0.832* 0.091 ns ns - non significant, *P<0.05, **P<0.01 The number of species increased from 16, to 18-27 in the variants that were fertilized. Shannon index increased from 2.304 for the A0b0 treatment to values between 2.441 and 2.742 for fertilized treatments. Eveness Shannon was 0.748 at A0b0 and between 0.721 and 0.832 for the fertilized treatments. Simpson index was from 0.102 at A0b0 and was between 0.071 and 0.091 for the fertilized treatments. The increase of the number of species is attributed to the application of fertilizers. The manure was a source of increase in the number of species, especially in the category 'plants from other botanical families' because of the pool of seeds that it contains. The low doses of manure, applied at different intervals, together with low doses of chemical fertilizers, contribute to conserve the number of species. Conclusions The results showed that some changes occurred in plant species and functional groups of plants under the different fertilization treatments. The increase of the number of species is possibly due to improved soil nutrient supply on these lands and, in addition, to species being introduced with applied manure. Using a fertilization management based on small amounts of organic and mineral fertilizers can be a solution which will contribute to the conservation of the biodiversity of these grasslands. The results of this study, in an area considered to be regionally representative for large parts of the mountains of Romania, indicate that fertilization treatments are able to maintain a high diversity of species. References Cristea V., Gafta D. and Pedrotti F. (2004) Fitosociologie. Editura Presa Universitara Clujeana, pp 164-166. Ellenberg H., Weber H., Düll R., Wirth V., Werner W. and Paulißen D. (1992) Indicator values of plants in Central Europe. Verlag Eich Gotze KG, D – 3400 Götingen, 50-75. Pacurar F., Rotar I., Bogdan Anca, Vidican R. and Dale L. (2012) The influence of mineral and organic long-term fertilization upon the floristic composition of Festuca rubra L.-Agrostis capillaris L. grassland in Apuseni Mountains, Romania. Journal of Food, Agriculture and Environment 10, 866-879. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 304 Peeters A., Maljean J.F., Biala K. and Bouckaert V. (2004) Les indicateurs de biodiversité pour les prairies: un outil d'évaluation de la durabilité des systèmes d’élevage. Fourrages 178, 217-232. Samuil C., Vintu V., Sirbu C. and Popovici C.I. (2012) Influence of fertilization on the vegetation and production of Festuca valesiaca L. grassland. Fourrages 210, 151-157. Vintu V., Samuil C., Rotar I., Moisuc Al., Razec I. (2011) Influence of the management on the phytocoenotic biodiversity of some Romanian representative grassland types. Notulae Botanicae Horti Agrobotanici ClujNapoca 39, 119-125. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 305 The effect of organic fertilization on Agrostis capillaris L. and Festuca rubra L. grasslands from the Romanian Eastern Carpathians Vintu V., Chidovet S., Samuil C. and Stavarache M. University of Agricultural Sciences and Veterinary Medicine Iasi, Romania Corresponding author: vvintu@uaiasi.ro Abstract Agrostis capillaris L. and Festuca rubra L. grasslands cover extended areas in Romania and represent a valuable source of fodder for bovines and sheep. The present paper presents the results of research conducted on a grassland from Bukovina, in the North of Suceava County, regarding the influence of manures in rates of 20-50 t ha-1 applied manually over the vegetal canopy, productivity and fodder quality. After three years of research there was an increase in the degree of vegetation cover from 87.3% to 99.9%, and an increase in the quantity of species from the Fabaceae - with increased vegetation cover from 3.5% to 37.8%. There was an improvement of fodder quality through the increase of CP content from 89.4 g kg-1 to 118.3 g kg-1, a decrease of NDF, ADF and ADL content, and a rise of the RFQ from 73 in the unfertilized control to 90-99 in the fertilized variants. Keywords: permanent grassland, organic fertilization, productivity, fodder quality Introduction The area covered with permanent grasslands in Romania amounts to over 4.8 m. ha, which represents 33% of the agricultural area. A large part of this area is situated in the mountainous regions, where grassland stands are the main food source for the animals. The increase in productivity and the improvement of fodder quality, with minimal effects for biodiversity and the environment, represent major objectives in the efficient use of these important natural resources (Cruz et al., 2002; Elsaesser et al., 2008). The quantity and the quality of livestock production depends largely on fodder quality, which in turn is influenced by the botanical composition, the vegetation stage at harvesting and the level of fertilization (Pozdisek et al., 2007; Rodrigues et al., 2007; Vintu et al., 2011). Materials and methods A monofactorial-type trial was established on Agrostis capillaris L. and Festuca rubra L. grassland in Putna, located in the North of Suceava County (47049' N, 25036' E; altitude 611 m asl). The trial consisted a randomized block design, with three replicates. The soil had 14.3 ppm P and 287 ppm K. Cpmposted and semi-composted manure, at rates of 20-30 t ha-1 and 40-50 t ha-1 respectively, were manually applied in early spring, and effects were monitored. Experimental variants were: V1-unfertilized control, V2-20 t ha-1, V3-30 t ha-1, V4-40 t ha-1, V550 t ha-1. The composted manure had a content of 46 g kg-1 N, 18 g kg-1 P2O5, 51 g kg-1 K2O, and the semi-composted manure had 33 g kg-1 N, 13 g kg-1 P2O5, 40 g kg-1 K2O. The Kjeldahl method was used for the determination of the CP, and for NDF, ADF and ADL the Van Soest method was used, while RFQ was determined through the relation proposed by Undersander and Moore (2002). Determinations of fodder quality were carried out on samples from the first cycle of harvesting, the data representing the average for the years 2010-2012. The statistical interpretation of the results was conducted by an analysis of variance (ANOVA), calculating the least significant difference and the square correlations between the manure doses and the followed indicators. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 306 Results and discussion The obtained data reveal significant differences regarding the participation of species from various groups as far as the vegetation cover is concerned. Unlike the control variant a larger participation was identified, of 61.4%, for species from the Poaceae and of 22.4% for forbs, the fertilized variants revealing a significant increase in the species ratio from the Fabaceae, from 3.5% to 33.5-37.8%, with a superior fodder quality and positive effects on fodder quality (Table 1). Table 1. Effect of organic fertilization on the degree of vegetation cover. Vegetation cover (%) V1(Control) V2 V3 V4 V5 *, **, *** - positive significant and º, ºº, ººº - negative significant at the 0.05, 0.01, 0.001 probability level, ns=unsignificant Grasses 61.4 53.2ºº 56.3º 57.0ns 54.1ºº P<0.05=4.4; P<0.01=6.4; P<0.001=9.6 Legumes 3.5 37.8*** 34.0*** 33.5*** 35.7*** P<0.05=5.7; P<0.01=8.5; P<0.001=12.5 Forbs 22.4 8.6ººº Total 87.3 99.6*** 99.9*** 99.9*** 99.9*** P<0.05=1.3; P<0.01=1.8; P<0.001=2.8 Species group 9.6ººº 9.4ººº 10.1ººº P<0.05=2.5; P<0.01=3.4; P<0.001=5.4 The analysis of data from Table 2 reveals that all fertilized variants recorded a very significant progress in the production of dry matter, with a growth of 135-250% compared to the control variant (V2, respectively V5); this emphasizes the importance of manure in the increase of productivity for Agrostis capillaris and Festuca rubra grasslands in the Romanian Carpathians. As far as fodder quality is concerned, there was a positive influence of organic fertilization on CP content, increasing from 89.4 g kg-1 DM in the control V1, to 118.3 g kg-1 DM in the V5 variant, with statistically significant increased growth for the use of 30-50 t ha-1 manure rates. Table 2. Influence of organic fertilization on the production of dry matter and the fodder quality. Treatments Dry NDF ADF ADL (kg ha-1) (g kg-1) (g kg-1) (g kg-1) (g kg-1) (% DM) (% DM) (% BW) V1 (control) 1.08 96.6 89.4 596.2 516.2 90.0 V2 254.2 99.7 556.800 455.20 80.2000 53.4* matter (t·ha-1) V3 2.55*** 3.38*** CP 349.5* CP 103.4* 537.9 000 517.3 000 453.5 0 451.2 00 DDM 48.7 TDN DMI RFQ 44.9 2.01 73 51.6** 2.14** 90* 84.6 00 53.6** 51.8** 2.23*** 94** 85.7 0 53.8** 52.1** 2.33*** 99** V4 3.78*** 407.4** 107.8* V5 3.80*** 449.5** 118.3** 548.0000 441.300 86.90 54.5** 53.2** 2.20** 95** P<0.05 0.46 190.1 13.8 21.9 44.4 3.1 3.3 4.0 0.09 13 P<0.01 0.66 282.3 20.1 31.8 64.6 4.6 4.8 5.8 0.13 19 P<0.001 0.99 421.2 30.2 47.8 96.9 7.2 7.2 8.7 0.20 29 CP-crude protein, NDF-neutral detergent fibre, ADF-acid detergent fibre, ADL-acid detergent lignin, DDMdigestible dry matter, TDN-total digestible nutrients, DMI-dry matter intake, RFQ-relative forage quality The increase in production and of fodder content in crude protein has led to quantities of 254.2449.5 kg ha-1 CP, compared to just 96.6 kg ha-1 CP for the control, with statistically ensured growth for the manure doses of 30-50 t ha-1. The fertilization also determined the reduction in NDF, ADF and ADL content for the fertilized variants compared to the unfertilized control, with statistically ensured differences. Thus the content of NDF decreased from 596.2 g kg-1 in DM in the sample variant, to 517.3 g kg-1 in DM in the V4, and the ADF content decreased Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 307 from 516.2 g kg-1 in the control variant, to 441.3 g kg-1 in the V5. The decrease of the NDF and ADF content has influenced the rise of the RFQ values in all the fertilized variants, with statistically ensured growth, the highest RFQ value of 99 being reached for the V4. The RFQ growth compared to the unfertilized control varied between 23 and 34% (V2, respectively V4). Moreover, increases were observed in the DDM, DMI and TDN values compared to the sample one in all the fertilization variants, with statistically ensured growth, underlining the positive role of organic fertilization in fodder digestibility, following the reduction in content from the cell walls. The analysis of the obtained data indicates that the regressions between the applied manure rates and the NDF and ADF content were negative, and with TDN values are positively correlated (Figure 1). Figure 1. Correlations between the applied manure rates and the fodder content of NDF, ADF and TDN Conclusions A rational organic fertilization and its sustainable use has determined a change in the composition of floristic composition, an increase in the degree of vegetation cover, in productivity and the significant improvement in fodder quality, by reducing the content in the cell walls, increasing digestibilitty and the content of crude protein. The Agrostis capillaris L. and Festuca rubra L. grasslands from the Romanian Carpathians have the potential to become an important resource of fodder through fertilization with 30-50 t ha-1 of manure, thereby ensuring production growth of 212-250% compared with that of the unfertilized control. References Cruz P., Duru M., Therond O., Theau J.P., Ducourtieux C., Jouany C., Al Haj Khaled R. and Ansquer P. (2002) Une nouvelle approche pour caracteriser les prairies naturelles et leur valeur d’usage. Fourrages 172, 335–354. Elsaesser M., Kunz H.G. and Briemle G. (2008) Strategy of organic fertilizer use on permanent grassland–results of a 22-year-old experiment on meadow and mowing-grassland. Grassland Science in Europe 13, 580-582. Pozdisek J., Stybnarova M., Svozilova M. and Latal O. (2007) Changes in chemical composition, digestibility and energy content in permanent grassland influenced by intensity of utilization and fertilization. Grassland Science in Europe 12, 70-73. Rodrigues A., Andueza D., Violleau S., Fefeu B., Picard F., Cecato U. and Baumont R. (2007) Effect of the fertilization on the feed value of permanent grassland. Grassland Science in Europe 12, 200-203. Undersander D. and Moore J. (2002) Relative fodder quality. UW Extension. Focus on Fodder vol. 4, No. 5. Vintu V., Samuil C., Sirbu C., Popovici I.C. and Stavarache M. (2011) Sustainable management of Nardus stricta L. grasslands in Romania’s Carpathians. Notulae Botanici Horti Agrobotanici 39 (2), 142-145. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 308 Herbage Recommended List applicability to low inorganic nitrogen (N) production systems Matthews J.M.1, Genever E.2, McConnell D.2 and Kerr S.1 1 NIAB, Huntingdon Road, Cambridge, CB3 0LE, UK 2 Agriculture and Horticulture Development Board, Stoneleigh Park, Kenilworth, CV8 2TL, UK Corresponding author: joanna.matthews@niab.com Abstract The Recommended Grass and Clover List (RGCL) trialling system is designed to evaluate varietal potential in which fertility issues are removed as a limiting factor. As plant nutrients are an essential part of plant development, with nitrogen (N) in particular being second only to photosynthesis in importance to potential yield, the key question is whether the expression of yield and quality traits under a high-input system are reflected in restricted nutrient supply systems. A series of field trials was established in 2011 across three United Kingdom locations. Six intermediate perennial ryegrass (IPRG) (Lolium perenne. L) varieties were selected with similar heading dates, and managed under a standard RGCL operating procedure while being tested under varying levels of N inputs (100, 200, 400 kg N ha-1). Fertilizer regimes significantly (P<0.001 and P<0.01) affected IPRG yields and forage quality respectively. The highest levels of yield and quality were obtained from the 400 kg N ha-1 treatment. Yield performance rankings of IPRG varieties varied with management regime and individual cuts. There were no significant N-level varietal interactions; therefore, preliminary indications from this ongoing study are that the RGCL system represents the varietal performance irrespective of N fertilization regime. Keywords: nitrogen, grass clover recommended list Introduction The current RGCL system for perennial ryegrass (L. perenne) assesses varieties’ genotype x phenotype interactions of environmental and management factors (Burns et al., 2013). The RGCL procedure utilizes high levels of inorganic fertilizer particularly N (c. 400 kg N ha -1). This allows varieties to ‘perform’ without nutrient constraints and to achieve their true potential, and therefore identifies varietal differences. As a result the current RGCL trials represent a high-input system; however, grassland systems are diverse, from highly intensive to extensive semi-natural pastures (ADAS, 2009). Agri–environmental schemes such as the Entry Level Scheme (ELS) encompass 60% of England’s farmed area. Within ELS there are options to reduce N inputs within areas of farmland (Natural England, 2013). Grass species display differences in root growth, turnover rates and architecture Humphreys, 2011; Sokolovic et al., 2013). Root morphology can significantly influence infiltration / percolation rates and N interception (Nichols and Crush 2007; Humphreys, 2011). Nichols and Crush (2007) demonstrated water percolation levels through the soil profile under hybrid (Lolium x boucheanum) and Italian ryegrass (Lolium multiflorum Lam) was half that of L. perenne, co-occurring with significant (P<0.05) differences in labelled 15N interception. Studies in L. perenne have identified cultivar differences in both growth habit (Sokolovic et al., 2013) and root morphology impacting on 15N interception (Nichols and Crush, 2007). A proportion of herbage cultivars within commercial grassland systems are not under high input systems. Cultivars also demonstrate differences in below-ground attributes. Therefore, varietal responses in terms of performance (yield/quality) under more restricted N supply may differ from the RGCL system. Preliminary work (NIAB, 2006) indicated some correlation between PRG variety performance under the current RGCL system and a low input system, and weak correlation under zero inorganic N. This study builds on the NIAB (2006) Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 309 preliminary work and focuses on the applicability of the RGCL system in the context of lower N production systems. Materials and methods Field trials were established in two sowing years in Yorkshire, Shropshire and Devon (on Aberford, Arrow and Teme soil series, respectively). The experimental design was a randomized block replicated four times and plot size was 6.5m2. It comprised six IPRG (L. perenne) diploid and tetraploid (T) varieties. Selected varieties were Premium (D), Rodrigo (D), Abergreen (D), Aubisque (T), Montova (T) and Seagoe (T). The varieties represent a relatively narrow heading-date range and a broader performance range in terms of yield and quality. Varieties were sown in accordance with their ploidy (25 ha-1 diploid, and 35 kg ha-1 tetraploid). Three annual fertilizer regimes (100, 200, 400 kg N ha-1) were applied across the trial, with spring and post-cutting applications. Harvesting regime follows a system of successive harvest years incorporating conservation management (Year 1 and 3; circa 5 cuts/year) and grazing management (Year 2; circa 8 cuts/year) as per the RGCL protocol. Assessments included yield (t DM ha-1) and quality (Digestibility value). All other fertility indices were maintained as per 'good agronomic practice' and designed not to be crop limiting. Data were analysed by analysis of variance and other appropriate statistical tests. Results and discussion N fertilization regimes significantly (P<0.001) influence grass yield under conservation and grazing management. The greatest yield was expressed under the high (400N) input system, consistent with established findings (St. Luce et al., 2011). Analysis of quality of late-season digestibility value demonstrated significant (P<0.01) quality improvement with higher N levels. This is potentially attributable to physiological changes within the plants at differing N levels. Drought stress in the 2013 season and nutrient deficiency could have induced a higher proportion of stems and reproductive growth in the lower-N treatments. Increased stem production increases neutral detergent fibre (NDF) and acid detergent lignin (ADL) content and a higher ADL/NDF ratio, resulting in modifications to the digestibility value (Tas et al., 2005). Results are consistent with results reported by the Tas (2006), whereby phenotype N regime, management and cultivars genotype influenced herbage yield and composition. The 2012 conservation-management annual yield demonstrated consistent ranking of varieties across the fertilizer regimes. Variety ranking within the 2013 grazing system (Table 1) displayed greater variability across the N-fertilization regime. Table 1. Ranking of perennial ryegrass varieties under grazing management (2013: annual yield in t DM ha -1). T denotes tetraploid varieties. 100 kg N ha-1 200 kg N ha-1 400 kg N ha-1 1 Abergreen Abergreen Abergreen 2 Premium Seagoe (T) Montova (T) 3 Seagoe (T) Rodrigo Seagoe (T) 4 Aubisque (T) Montova (T) Premium 5 Rodrigo Aubisque (T) Aubisque (T) 6 Montova(T) Premium Rodrigo Rank The Meehan and Gilliland (2013) study indicates that minor temperature shifts significantly (P<0.019) impacted on L. perenne yields. The second sowing-year grazing management data will provide an insight as to the extent of the impact of temperature on varietal performance. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 310 Ranking of cultivars fluctuated between N-fertilizer regimes and individual cuts, culminating in differences expressed within the annual grazing yields (Table 2) and between management regimes. The most pronounced ranking differences were expressed in the 2013 grazingmanagement system; however, there was no significant N x variety interaction with annual or individual-cut yields. This suggests that the performance of varieties (yield/quality) in terms of ranking is unilateral irrespective of fertilizer regime. Conclusion Nitrogen significantly influenced herbage yield and quality over the 2012 and 2013 season. Conservation management preliminary results under differing N fertilisation regimes appear to mimic RGCL ranking. Grazing management demonstrated the largest disparity from the RGCL ranking. Analysis demonstrates that no significant N regime x variety interactions, therefore the preliminary findings indicate the RGCL system represents lower fertiliser regimes. The second year sowings data and the 3rd year conservation management will provide a more robust data set to draw further conclusions. Acknowledgement This work was funded by the Agriculture and Horticulture Development Board, through its divisions of EBLEX and DairyCo. References ADAS (2009) Management guidelines for grassland in environmental schemes. UK: ADAS. Burns G.A., O’Kiely P., Grogan D., Conaghan P. and Gilliland T.J. (2013) Is the ranking of perennial ryegrass varieties nutritive value at individual cuts representative of the mean annual ranking? In: Proceedings of the British Grassland Society 11th Research Conference. 2-3 September 2013. Stoneleigh, UK: BGS. Humphreys M.W. (2011) Grass roots for improved soil structure and hydrology. In: IBERS Knowledge-Based Innovations No 4, pp 20-25. Aberystwyth UK: IBERS. Available online at http://www.aber.ac.uk/en/ibers/publications/ibers-knowledge-based-innovations-4/ Meehan E.J and Gilliland T.J (2013) Comparative annual forage yields in an unpredictable climate. In: Proceedings of the British Grassland Society 11 th Research Conference. 2-3 September 2013. Stoneleigh, UK: BGS. Natural England (2013) Entry Level Stewardship. [Online] Available from http://www.naturalengland.org.uk/ourwork/farming/funding/es/els/default.aspx [Date accessed 1/12/13] NIAB (2006) SID 5 Research Project Final Report. DEFRA, UK. Nicholas S.N and Crush J.R (2007) Selecting forage grasses for improved nitrate retention – a progress report. [Online] Available from www.grassland.org.nz/publications/nzgrassland_publication_169.pdf. accessed 1/12/13. Sokolovic D., Babic S., Radovic J., Milenkovic J., Lugic Z., Andjelkovic S. and Vasic T. (2013) Genetic variations of root characteristics and deep root production in perennial ryegrass cultivars contrasting in field persistency. Breeding strategies for sustainable forage and turf improvements, 275-281. St. Luce, M., Whalen, J.K., Ziadi N. and Zebarth B.J (2011) Nitrogen dynamics and indices to predict soil nitrogen supply in humid temperature soils. Advances in Agronomy 112, 55-102. Tas B., Taweel H.Z., Smit H.J., Elgersma, A., Dijkstra J. and Tamminga S. (2005) Effects of perennial ryegrass cultivars on intake, digestibility and milk yield in dairy cows. Journal of Dairy Science 88, 3240-3248. Tas B. (2006) Nitrogen utilization of perennial ryegrass in dairy cows. In: Elgersma A., Dijkstra J. and Tamminga S. (eds) Fresh herbage for dairy cattle. Wageningen Frontis series. Springer, Netherlands. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 311 Effect of mineral fertilization on yield and quality of grassland ecosystem Agrostietum vulgaris Vuckovic S.1, Simic A.1, Jovanovic M.2, Cupina B.3. and Krstic D.3 1 Faculty of Agriculture, University of Belgrade, 11080 Zemun-Belgrade, Serbia 2 Institute of Agricultural Economics, 11060 Belgrade, Serbia 3 Faculty of Agriculture, University of Novi Sad, 21000 Novi Sad, Serbia Corresponding author: marijanajovanovic21@gmail.com Abstract The main objective was to examine the influence of mineral fertilization on the production capacity of Agrostietum vulgaris-type meadow located in western Serbia, during 2005-2008. Mineral fertilizers with different NPK rates (0-500 kg ha-1) were applied. Results established that mineral fertilization at N200P150K150 provided the highest yield of dry matter (8.13 t ha-1). On unfertilized soils dry matter yields of the grass sward was substantially lower and disappearance of valuable grasses was observed. As opposed to the unfertilized plots, on plots fertilized with high NPK rates herbs and weeds diappeared. We examined the effects of fertilizer application on hay and protein yield to avoid economic losses from loss of applied fertilizers. Keywords: fertilization, meadow, DM yield, forage quality Introduction Natural meadows cover large areas in the hilly-mountain region of Serbia. They are of considerable importance for forage and soil utilization and protection. Because of poor management and careless utilization these grasslands are now degraded, with low production and poor quality. The use of fertilizer is an important factor in intensive grass-based dairy farming, as NPK affects dry matter yield and crude protein content of herbage. Efficient fertilization provides plants with nutrients at appropriate proportions and quantities which thereby enable maximum yield increase of crops with high biological and technological quality (Barabasz et al., 2002). One of the highest presented and economically important associations in Balkan peninsula, at least in hilly-mountainous regions, is Agrostietum vulgaris (Tomić et al., 2009a). The association Agrostietum vulgaris in Serbia could include 47 plant species, of which 11 (or 23.4%) are useful grasses, 15 (or 32%) are useful legumes, 3 or (6.4%) are useful; however, less useful, are 17 (or 36.17%) that are are bad and worthless, with only one or (2.13%) harmful, but no poisonous species. Agrostietum vulgaris has a realized production of herbage mass of 3.15 t ha-1, and of dry matter 1.10 t ha-1. Content of crude protein was 10.0%, crude lipids 2.3%, crude fibre 29.8%, and NFE 41.7%. It realized the lowest production with regard to green mass and dry matter. The objective of this study was to assess the effect of fertilization on the yield and quality of semi-natural meadow type Agrostietum vulgaris, with respect to agroecological and economical conditions. Materials and methods Examination was carried out during four years (2005-2008), on a semi-natural meadow dominated by Agrostietum vulgare in the western region of Serbia (near Valjevo city). The experiment was a randomized block design with four replications. It included six fertilizer rates (0, 150, 200, 300, 350 and 500 kg ha-1 NPK yr-1), which were applied in early April. The data determined in this experiment are: dry matter yield (DM), crude protein (CP), crude fibre (CF) and nitrogen free extracts (NFE) contents. Data were analysed as a factorial design by ANOVA; differences between means were determined by LSD. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 312 Results and discussion Effect of fertilizers on meadow-pasture vegetation is complex. The grass cover responds not only in terms of yield and quality but also in plant composition. Continued use of fertilizers changes the botanical composition of plant association. At the study site, Agrostis vulgaris, Festuca rubra, Dactylis glomerata, Festuca pratensis and Arrhenatherum elatius made the most important contribution to herbage yield in all the treatments. The rainfall over the four growing seasons varied between 755 mm and 930 mm (Table 1). In the first and third year, rainfall was 6 and 8% less than the long-term annual rainfall of 820 mm for the study area, as opposed to the second and last year with 13% and 0.5% on average more than the longterm average. Table 1. Dry matter production (t DM ha-1) nutrient use efficiency (NUE) and seasonal rainfall use efficiency (RUE) (kg CP ha-1mm-1) on Agrostietum vulgaris grassland at different fertilizer application rates of N, P and K for the 2005-2008 growing seasons. Least significant differences (LSD) are calculated at the 5% level Treatments 2005 2006 2007 2008 kg ha-1 DM NUE RUE DM NUE RUE DM NUE RUE DM NUE RUE N0P0K0 2.32 0 3.01 2.15 0 2.32 1.47 0 1.94 1.89 0 2.3 N50P50K50 5.20 19.2 6.75 4.52 15.78 4.86 3.20 11.5 4.24 4.59 18 5.57 N100P50K50 6.22 19.5 8.07 5.8 18.25 6.24 4.35 14.4 5.77 6.62 23.6 8.03 N100P100K100 6.25 13.1 8.1 5.85 12.33 6.3 4.7 10.8 6.23 6.39 15 7.76 N150P100K100 7.01 13.4 9.1 7.9 16.41 8.5 5.73 12.2 7.59 6.62 13.5 8.03 N200P150K150 8.27 12 10.73 8.11 11.91 8.72 7.11 11.3 9.41 9.03 14.3 10.96 LSD Average Precipitation 0.75 5.88 1.37 5.72 0.83 4.43 771 930 1.95 5.86 755 824 Increased fertilization led to a higher (P<0.05) production over the four seasons (Table 2). Table 2. Crude protein (CP), crude fibre (CF), ash and nitrogen free extracts (NFE) in Agrostietum vulgaris DM depending on fertilization (%) during the four-year period, 2005-2008. Main Effect Fertilizer treatment N0P0K0 N50P50K50 N100P50K50 N100P100K100 N150P100K100 N200P150K150 LSD 0.05 Years 2005 2006 2007 2008 LSD 0.05 CP CF Ash NFE 8.89 8.55 8.75 9.34 10.93 10.81 1.42 30.72 32.79 33.83 34.38 33.45 32.22 2.23 1.63 1.49 1.31 1.63 1.10 1.79 0.55 48.39 47.26 46.81 44.32 45.41 45.47 3.1 11.74 9.15 9.28 8.02 1.16 32.71 37.24 29.73 31.92 1.82 1.74 1.79 1.17 1.27 0.45 47.80 45.76 45.68 45.88 2.53 Compared with other seasons, the high rainfall during the 2005/06 season was insufficiently converted into dry material. This may be due to the poor seasonal distribution. The increase in production with equal N50P50K50 application rates differed significantly (P<0.05) from that of the control, but increase was more significant with additional N spring rates. With NPK fertilization the Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 313 DM production peaked in the third treatment (N100P50K50 with average NUE=19) in despite of higher NPK rates on the following treatments. This leads to the conclusion that the equal quantities of NPK were insufficient to ensure a rapid reaction in production or it may have a longer residual effect. Mean herbage yield increased from 1.96 t ha -1 DM (control plot) to 8.13 t ha-1 year-1. Figure 1 shows crude protein (CP) production (kg CP ha-1) and seasonal rainfall use-efficiency (RUE) (kg CP ha-1 mm-1) of Agrostietum vulgaris grassland at different NP fertilization rates and for the four growing seasons. In treatments N60 and N120, the CP content in grass DM increased by 0.14 and 2.66%, and CP yield per hectare by 98 and 226%, respectively, compared with N0 (at P60 K80 background). Other tests performed in Serbia showed that fertilizer N had a favourable effect on the yield, protein, ash and fat content while decreasing cellulose content (Vuckovic et al., 2005a). 1400.00 1.80 PY 2004/05 1.60 PY 2005/06 1200.00 PY 2006/07 RUE 2004/05 1.20 Protein yield RUE 2005/06 800.00 RUE 2006/07 1.00 RUE 2007/08 0.80 600.00 0.60 Rainfall use efficiency 1.40 PY 2007/08 1000.00 400.00 0.40 200.00 0.20 0.00 0.00 N0P0K0 N50P50K50 N100P50K50 N100P100K100 N150P100K100 N200P150K150 Fertilizer rate Figure 1. Crude protein production (kg CP ha-1) and seasonal rainfall use-efficiency (RUE) (kg CP ha-1 mm-1) on the Agrostietum vulgaris grassland at different fertilizer rates of N, P and for four growing seasons. Conclusion The data obtained in this study indicate that fertilization has a considerable influence on semi-natural meadows dominated by Agrostietum vulgare, regarding their DM yield and quality as well as botanical composition. The maximum four-year average DM yield of 8.13 t ha-1 and CP yield of 877.5 t ha-1 was achieved with the N200P150K150 rate (500 kg ha-1 year-1), but nutrient use efficiency was greatest with the N100P50K50. The increased application of NPK fertilizers improved seasonal rainfall use efficiency and affected crude protein and fibre content. Acknowledgements Projects: (1) 'Research, education and knowledge transfer promoting entrepreneurship in sustainable use of pastureland/grazing' (Norway); (2) 'Mitigations and Adoption of Climate Change for Different Land Uses in West Balkan: Agricultural Practices to Combat Climate Change - Education, Collaboration and Research' Norway; (3) 'Improvement of technology for fodder plants growing on arable land and grassland' TR31016. References Barabasz W., Albińska D., Jaśkowska M. and J. Lipiec J. (2002): Biological effects of mineral nitrogen fertilization on soil microorganisms. Polish Journal of Environmental Studies 11(3), 193-198. Vučković S., Ćupina B., Simić A., Prodanović S. and Živanović T. (2005a) Effect of nitrogen fertilization and undersowing on yield and quality of Cynosuretumcristati-type meadows in hilly-mountainous grasslands in Serbia. Journal of Central European Agriculture 6 (4), 515-520. Tomić Z., Bijelić Z. and Krnjaja V. (2009a) Analysis of grassland associations of Staraplanina mountain. Biotechnology in Animal Husbandry 25 (5-6), 451-464. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 314 Impact of surface fertilization on dehydrogenase activity in grassland soil Tampere M., Kauer K., Keres I., Laidna T., Loit E., Parol A., Selge A., Viiralt R. and Raave H. Estonian University of Life Sciences, Institute of Agricultural and Environmental Sciences, Kreutzwaldi 5, 51014 Tartu, Estonia Corresponding author: mailiis.tampere@emu.ee Abstract Soil microorganisms have a significant role in soil processes and their activity is largely influenced by fertilization. The aim of this study was to determine the effect of fertilization on grassland soil dehydrogenase activity, which is one of the most sensitive bioindicators connected to soil fertility and microbial activity. Soil samples were collected after fertilization with mineral fertilizer, cattle slurry and cattle slurry digestate throughout the growing period in 2013 from grassland. Dehydrogenase activity was determined colorimetrically after sample incubation with iodonitrotetrazolium chloride. We found that the impact of surface fertilization on dehydrogenase activity was low. The dehydrogenase activity depended on air temperature more than on fertilization Keywords: dehydrogenase activity, fertilizer, grassland, temperature Introduction Soil microbiological activity plays a key role in nutrient cycling. Its activity is essential in both the mineralization and transformation of organic matter and plant nutrients in soil (Dick and Tabatabai, 1993). Organic and inorganic fertilizers increase nutrient availability to plants, but at the same time they can affect the population, composition, and function of soil microorganisms (Marschner et al., 2003) and consequently soil enzymatic activities (Wolińska and Stępniewska, 2012). It has been found that organic fertilizers increase soil enzyme activities, while inorganic fertilizers have relatively less effect (Mijangos et al., 2006; Chu et al., 2007), but also that dehydrogenase activity increases by mineral fertilizer use (Chu et al., 2005). Dehydrogenase activity (DHA) is one of the most adequate and sensitive bioindicators, relating to soil quality, fertility (Wolińska and Stępniewska, 2012) and overall microbial activity (Salazar et al., 2011), because dehydrogenases occur intracellular in all living microbial cells (Wolińska and Stępniewska, 2012). DHA reflects metabolic ability of the soil and its activity is considered to be proportional to the biomass of the microorganisms in soil (Wolińska and Stępniewska, 2012). The objective of this experiment was to study the effect of fertilization on dehydrogenase activity in grassland soil. We hypothesized that the surface application of fertilizers on the grassland does not have strong and rapid impact on soil enzymatic activity. Materials and methods The experiment was conducted at the Estonian University of Life Sciences (58° 23' 32" N 26° 41' 31" E; elevation 60 m) in the year 2013 on a Stagnic Luvisol (WRB). The sward consisted of bluegrass (Poa pratensis) and red fescue (Festuca rubra L.). Treatments were: (i) control (no fertilizer was applied), (ii) mineral N-fertilizer (NH4NO3), (iii) cattle slurry, and (iv) cattle slurry digestate, in three replicates. Fertilizers were applied to the soil in quantities according to the nitrogen rate of 180 kg ha-1 in three equal splits on 3 May, 11 June and 30 July. The organic fertilizers application rate was calculated based on NH4-N content. Fertilizers were surface-applied. The soil was sampled throughout the growing period: one day after fertilizer application and one and two months after the third fertilizer application. Soil samples (50 g) Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 315 were taken with a soil auger at depth 10 cm and stored in a refrigerator at 4ºC. Dehydrogenase activity was determined according to Von Mersi and Schinner (1991) in triplicate per sample and expressed by the dry weight of soil. Samples were incubated with iodonitrotetrazolium chloride (INT) and the formation of iodonitrotetrazolium formazan (INTF) was measured colorimetrically. All calculations were performed using the statistical package Statistica 9.0 (StatSoft.Inc). The probability level was set at 0.05. Results and discussion Our results showed that the fertilization effect on DHA in grassland soil was not statistically significant (P = 0.19). The effect of organic fertilizers on DHA did not differ significantly from the effect of mineral fertilizer, although in the soil fertilized with organic fertilizers, DHA after the treatment was slightly higher than in the control and with the use of mineral – N, and 30 days after fertilization DHA in treatments with organic fertilizer was significantly (P < 0.05) higher when compared to the control (Table 1). In treatments with mineral-N, DHA was similar to the control or slightly lower. Earlier studies have shown highest DHA in soils treated with animal manure and the lowest in the unfertilized soil or in soil treated with mineral fertilizer (Parham et al., 2002) because addition of organic matter to the soil with organic fertilizers increases its microbial growth (Mijangos et al., 2006). Our study showed that organic fertilizer effect on DHA in grassland soil (averaged for 0 – 10 cm layer) was not significantly higher when compared to control and mineral fertilizer application. We speculate that this may be caused by the fertilizer application being on the surface of the grassland; therefore its influence did not reach deeper soil layers or the one day that remained between fertilization and sampling time was too short a period for the effect of fertilization to occur. We found that DHA depended more on air temperature than on fertilization and fertilizer type (r = 0.48, P < 0.05), and it increased in all treatments with increase in the temperature. Dehydrogenase enzyme exists only inside the viable microbial cells; therefore, its activity should be the highest at a temperature close to the optimum temperature (30 °C) for microorganism growth (Wolińska and Stępniewska, 2011). As a result, DHA was the highest at the time of third fertilization (P < 0.01), when air temperature was the closest (21 °C) to the optimum temperature for growth of microorganisms. There was no significant difference in DHA between the first two fertilization times; then the mean air temperatures were respectively 11.9 °C and 13.8 °C. Table 1. The effect of fertilization on DHA (µg INTF g-1DM h-1±se ) Treatment Time of fertilization Control Mineral N Digestate Cattle slurry 3 May 54.3 ± 10.7 67.1 ± 7.4 68.9 ± 3.8 77.5 ± 19.2 11 June 61.4 ± 6.2 51.0 ± 5,8 71.1 ± 8,1 76.0 ± 10.2 30 July 117.6 ± 21,6 102.0 ± 12.8 145.3 ± 5.3 138.5 ± 10.8 Fertilizer aftereffect since the third application 30 days later 61.1 ± 14.4 87.2 ± 14.3 122.4 ± 8.6 118.3 ±19.6 60 days later 100.5 ± 25.2 83.9 ± 0.3 103.9 ± 2.8 97.1 ± 9.0 Conclusion From our results it can be concluded that fertilization does not significantly affect microbial activity in grassland. More than fertilization and fertilizer type, the dehydrogenase activity depended on air temperature and it increased with the increase in temperature. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 316 References Chu H.Y., Hosen Y., Yagi K., Okada K. and Ito O. (2005) Soil microbial biomass and activities in a Japanese Andisol as affected by controlled release and application depth of urea. Biology and Fertility of Soils 42, 89–96. Chu H., Lin X., Fujii T., Morimoto S., Yagi K., Hu J. and Zhang J. (2007) Soil microbial biomass, dehydrogenase activity, bacterial community structure in response to long-term fertilizer management. Soil Biology and Biochemistry 39, 2971–2976. Dick W.A. and Tabatabai M.A. (1993) Significance and potential uses of soil enzymes. In: Metting F.B. (ed) Soil microbial ecology. Marcel Dekker, New York, pp. 95-127. Marschner P. (2003) Structure and function of the soil microbial community in a long-term fertilizer experiment. Soil Biology and Biochemistry 35, 453–461. Mijangos I., Pérez R., Albizu I. and Garbisu C. (2006) Effects of fertilization and tillage on soil biological parameters. Enzyme and Microbial Technology 40, 100–106. Parham J.A., Deng S.P., Raun W.R. and Johnson G.V. (2002) Long-term cattle manure application in soil I. Effect on soil phosphorus levels, microbial biomass C and dehydrogenase and phosphatase activities. Biology and Fertility of Soils 35, 328–337. Salazar S., Sanchez L., Alvarez J., Valverde A., Galindo P., Igual J., Peix A. and Santa- Regina I. (2011) Correlation among soil enzyme activities under different forest system management practices. Ecological Engineering 37, 1123-1131. Von Mersi W. and Schinner F. (1991) An improved and accurate method for determining the dehydrogenase activity of soils with iodonitrotetrazolium chloride. Biology and Fertility of Soils 11, 216-220. Wolińska A. and Stępniewska Z. (2011) Microorganisms abundance and dehydrogenase activity as a consequence of soil reoxidation process. In: Miransari M.( ed) Soil Tillage & Microbial Activities, Research Singpost, Kerala, India, pp. 111-143. Wolińska A. and Stępniewska Z. (2012) Dehydrogenase activity in the soil environment. In: Canuto R.A. (ed), Dehydrogenases, InTech, pp. 183-210. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 317 Reduction of soft rush (Juncus effusus L.) by a combination of trimming and grazing Nielsen A.L.1, Hald A.B.1 and Nissen T.2 1 Natlan, Agro Business Park, Niels Pedersens Allé 2, DK-8830 Tjele, Denmark 2 Organic Advisory Centre, Silkeborgvej 260, DK-8230 Åbyhøj, Denmark Corresponding author: ln@natlan.dk Abstract In wet grassland, soft rush (Juncus effusus L.) may be very dominant. This can make it difficult to obtain botanical diversity and a suitable habitat for meadow birds. When the swards are managed by grazing alone the soft rush is difficult to control because the animals have a low preference for this species. In this study different trimming strategies were compared in wet grassland grazed by young heifers and steers from dairy cattle. The four strategies were: 1) trimming in late April and in mid-October (early and late); 2) trimming late in April (early); 3) trimming in mid-October (late); 4) no trimming (none). No trimming and grazing resulted in an increase of soft rush measured on a DM-basis. The best effect was obtained with two trimmings a year, late April and mid-October. Management should be adjusted when the soft rush is reduced and the sward has become attractive to meadow birds. Keywords: wet grassland, cutting, management, peat soil, biodiversity Introduction On grasslands, ecosystem services include delivery of botanical diversity and habitats for meadow birds. On moist soil it may be a problem to maintain botanical diversity and short vegetation for meadow birds in swards dominated by soft rush. Many of these grasslands are managed by cattle grazing, but when soft rush is dominating, the forage quality can be very low. Among the common species on moist grassland in Denmark the lowest digestibility was found for soft rush with an IVOMD value of 384 in July. At the same time, IVOMD of many of the other herbs varied between 600-800 (Nielsen and Søgaard, 2000). Comparing the effect of two cuts a year (middle of July and beginning of September) with the effect of continuous summer grazing with young dairy cattle, through a five-year period and leaving the vegetation to a cut in the middle of July in the sixth year, showed three times as much soft rush in the grazed sward than in the cut sward (i.e. 62% vs. 20% on DM basis) (Buttenschøn and Nielsen, 2004). Therefore, it seems possible to control the soft rush dominance by cutting, and farmers may obtain a more balanced diet for cattle if they combine grazing with trimming. To comply with Danish rules for farmers funded to manage grass- and nature areas, there is a requirement for no mowing between 1 May and 20 June; thus, early trimming for rush can be chosen in late April. Trimming in October can be chosen to prevent growth of the evergreen soft rush and make it vulnerable to the influence of water through winter. This paper presents results from a three-year study on the effect on soft rush of combining different strategies of trimming in paddocks grazed with young cattle from a dairy farm. Materials and methods The paddocks on permanent grassland used for the experiment were located at Fussingø Manor in Denmark. Four paddocks, 4 ha each, were grazed by heifers or steers of Holstein Friesian, or similar dairy cattle. Grazing started in the middle of May with a stocking density of approximately 750 kg per ha. From August to October cattle were withdrawn successively according to the decrease in grass production. Each paddock was located at a riverside area with most of the sward on moist peat soil, but about 20% of each paddock was on higher and dry mineral soil. Soft rush occurred only on the moist part of the paddock, and here four Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 318 different strategies of trimming were compared: 1) trimming in late April and in mid-October (early and late); 2) trimming late in April (early); 3) trimming in mid-October (late); 4) no trimming (none). Trimming was carried out with a rough cutter pulled by an all-terrain vehicle (ATV) with twin wheels. In each paddock, in the moist part, four plots of 10m × 10m were laid out at fixed points for botanical analysis. They were trimmed in the same way as the rest of the paddock. Before trimming in April samples were taken by cutting to the same height as trimming (7 cm) at 10 × 1 m2. Representative subsample were sorted into living and dead biomass of soft rush, grass and other herbs. In May 2010 and 2012, botanical diversity was analysed using extended Raunkiær circles (Böcher and Bentzon, 1958) with three circles in each of the four plots per paddock. Compressed sward height was measured by a rising plate meter (30 × 30 cm; 3.8 kg m-2) at specific routes in the paddocks, and the type of vegetation was recorded at each measurement. In addition to soft rush, the main other species in the sward were common species as Festuca rubra, Poa trivialis, Holcus lanatus, Rumex acetosa, Ranunculus repens, Taraxacum sp., Lotus pedunculatus var. pedunculatus, Trifolium repens, Juncus articulatus, Cirsium palustre; as well as Dactylorhiza majalis ssp. majalis. During the years of the experiment the ground water level varied from 10 to 40 cm below surface in MayAugust, average around 20 cm. In autumn it varied from 5 to 20 cm below the surface, average around 10 cm. Results and discussion The results from compressed sward height showed that height in areas with pure grass was significantly lower than in similar areas of mixed grass and soft rush, and areas dominated by soft rush had the highest compressed sward height (Figure 1). The results confirm the need of the cattle to graze the more digestible part of the vegetation. Figure 1. Sward height measured by rising plate meter in midsummer, average of measurements in the three years, 767 observations. Different letters indicate significant difference in height for type of vegetation (P<0.001). Treatment effects on soft rush performance over years is shown in Figure 2. Statistical evaluation over years within the individual treatments showed that soft rush increased where no trimming was applied, and decreased by two yearly trimmings. The results from Raunkiær analyses showed that treatments did not change botanical diversity significantly over the relatively short period of the experiment (values not shown). Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 319 Figure 2. Percentage soft rush in swards managed with grazing and trimming, and the effect of trimming strategy on the soft rush content measured on a DM basis. Different letters indicate significant difference over year within strategy and type of biomass (P<0.05). Conclusions With no trimming, soft rush increased, as measured on a DM basis. The best effect of trimming was obtained with two trimmings a year: in late April and middle of October. When the soft rush is reduced the swards will become increasingly interesting for meadow birds, which should be considered in the future trimming strategy. Acknowledgements The assistance of farmer Steen Hareskov is greatly appreciated. The study has received grants from the European Union, the rural development programme from the Ministry of Food, Agriculture and Fisheries, and from the Foundation for Organic Agriculture. References Böcher T.W. and Bentzon M.W. (1958) Density determination in plant communities. Oikos 9. 35-56. Buttenschøn R.M. and Nielsen A.L. (2004) Regulation of soft rush. Park og Landskab Videnblade, 6.2-13. (In Danish). Nielsen A.L. and Søegaard K. (2000) Forage quality of cultivated and natural species in semi-natural grasslands. Grassland Science in Europe 5, 213-215. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 320 Prospects for biological control of Rumex obtusifolius using a native clearwing moth Hahn M.A.1, Häfliger P.2, Schaffner U.2 and Lüscher A.1 1 Agroscope, Institute for Sustainability Sciences, Reckenholzstrasse 191, 8046 Zurich, Switzerland 2 CABI, Rue des Grillons 1, 2800 Delémont, Switzerland Corresponding author: min.hahn@agroscope.admin.ch Abstract Broad-leaved dock (dock, Rumex obtusifolius) is among the most troublesome weeds in European grasslands and efficient non-chemical methods for its control are lacking. This study evaluated the ability of the native specialist insect Pyropteron chrysidiforme to attack dock using a mass release approach. Larvae of this moth mine into the roots of dock, which can negatively affect plant performance and induce mortality. At four agricultural grassland sites across Switzerland we applied insects at various developmental stages and levels of protection (eggs, larvae unprotected, larvae encapsulated) onto marked plants and examined the resulting infestation. High infestation rates were attained with the application of eggs (71% of plants infested), while larval applications were less successful. Moreover, differences among sites suggest that weather conditions and biotic interactions may influence the infestation success. These results show that successful infestation of dock by P. chrysidiforme can be attained under various conditions in agricultural grasslands, which is promising for biological control. However, higher infestation rates may be necessary for practical application and could be achieved by specific application techniques that reduce adverse environmental effects. Moreover, ongoing long-term field experiments are needed to evaluate the impacts of this method and its efficiency in controlling dock. Keywords: biological control, Pyropteron chrysidiforme, root herbivory, Rumex obtusifolius, specialist insects, weed Introduction Broad-leafed dock (dock, Rumex obtusifolius) is among the most problematic weeds in agricultural grasslands and, because it is avoided by cattle, may lower fodder quality and rapidly displace other plants (Cavers and Harper, 1964). In addition to specific grassland management practices to reduce the presence of dock (Hopkins and Johnson, 2002), direct control approaches are used, which often include the use of herbicides. In order to reduce concomitant adverse effects for the environment and health, non-chemical alternatives are needed (Zaller, 2004). However, several attempts of above-ground biological control of dock have been deemed ineffective, given the plant’s highly regenerative capacity from the roots. Therefore, a more promising approach may adopt the use of specialized root-feeding insects. This strategy was successfully implemented in a biological control programme using a Moroccan clearwing moth (Pyropteron doryliforme) against invasive dock species in Australia (Faithful, 2000). Larvae of these moths mine into the roots of dock and thereby negatively affect plant performance. In the framework of the Australian programme, a high potential to control dock was also attributed to the native European specialist root-feeding clearwing moth Pyropteron chrysidiforme (Scott and Sagliocco, 1991). To assess whether P. chrysidiforme may be a potential candidate for biological control of dock in Europe, we evaluated its capacity to attack dock under natural conditions in agricultural fields using a mass release approach. Specifically, we investigated the variations in infestation of P. chrysidiforme on dock following application of insects at various developmental stages and levels of protection (eggs, larvae Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 321 unprotected, larvae encapsulated). Additionally, we examined the potential influence of root biomass, vegetation density, and other site conditions. Materials and methods We conducted a field experiment at four agricultural grassland sites across Switzerland. In spring 2012, late-instar larvae of P. chrysidiforme were collected in natural populations in western Switzerland and subsequently reared in the lab to produce eggs and larvae. In June 2012, thirty R. obtusifolius plants per site (120 plants total) were permanently marked, and 10 plants per site were subjected to either application of eggs glued onto toothpicks (30 eggs / toothpick), transfer of unprotected larvae using a paintbrush (6 larvae / plant), or transfer of encapsulated larvae using metal syringes inserted into the root crown (6 larvae / plant). The four sites were regularly managed (cutting and/or grazing), except for a period of three to four weeks after the insect application, when the sites were protected from disturbance. From September to October 2012, we sequentially excavated the plants, determined root biomass and dissected the roots to record the infestation by larvae of P. chrysidiforme, which was evaluated as the percentage of plants infested and the number of larvae per infested plant. Differences in infestation among application techniques and sites as well as potential effects of vegetation density, root biomass, and harvesting dates, were analysed by mixed effects models using logit and poisson link (for infestation rate and number of larvae, respectively) with the function ‘lmer’ in the package ‘lme4’ (Bates, Maechler and Bolker, 2012) in R version 2.14.1 (R Development Core Team, 2013). Results and discussion On average, 54% of the treated plants were infested with at least one larva per root, which indicates that successful establishment of P. chrysidiforme on dock in agricultural grasslands can be achieved. Furthermore, infestation rates differed among application techniques (χ2=7.197, df=2, P< 0.027). Application of eggs resulted in higher infestation rates (71% of plants infested) compared to unprotected larvae (47%; χ2= 4.508, df=1, P(eggs vs. unprotected larvae) < 0.034) and encapsulated larvae (44%; χ2= 5.710, df=1, P(eggs vs. encapsulated larvae) < 0.017). One factor that may at least partly explain the difference between the egg and unprotected larval applications is predation of larvae by ants, which is likely to have less influence on smaller, newly hatched larvae from eggs (Pedrotta et al., unpublished results). Furthermore, we found significant differences in infestation rates among the four experimental sites (χ2= 13.741, df=3, P<0.003; Figure 1). Aside from different biotic interactions (i.e. predation by ants), other site conditions may also affect infestation success. In particular, the weather conditions at the time of insect applications at each site may be important to assure benign conditions for the larvae, as is shown in a complementary common garden experiment (Hahn et al., unpublished results). Compared to the transfer of unprotected larvae, application of eggs maybe generally be less prone to adverse environmental factors, because the higher number of larvae hatching from eggs and the longer time period in which hatching occurs may increase the chance of successful establishment of the larvae in the roots. The reasons for the low infestation rates associated with the encapsulated larval transfers are less clear; while encapsulation is expected to reduce adverse environmental effects, larvae may possibly suffer from the handling procedure. Other experimental factors such as root biomass and vegetation density did not affect infestation rates. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 322 Figure 1. Proportion of R. obtusifolius plants infested with P. chrysidiforme at the four experimental sites, as a result of application of eggs or unprotected or encapsulated larvae. In contrast to the number of infested plants, the number of larvae per root did not significantly differ among treatments and sites, and was neither significantly affected by root biomass, nor by vegetation density. On average, an infested root hosted 1.89 larvae of P. chrysidiforme, with an observed maximum of 5 larvae per root. This relatively constant number of larvae per root is consistent with previous studies, which reported about one larva per root (Scott and Sagliocco, 1991) and may be due to intra-specific competition. Nevertheless, observations have been made of more than 10 P. chrysidiforme larvae per root in natural infestations (Hahn, personal observation) and, therefore, other adverse environmental factors may also limit larval establishment in the roots. Conclusions Our study revealed that successful infestation of dock by P. chrysidiforme can be attained under various conditions in agricultural grasslands. This result is promising for the use of mass release of P. chrysidiforme for the control of dock. However, greater and more persistent infestation may be necessary for practical application. This could be achieved by specific application techniques that minimize adverse environmental effects. Moreover, ongoing long-term field experiments are needed to evaluate the impact of this method on plant and population performance over multiple years and thus, its efficiency in controlling dock. Acknowledgements We thank H. Müller-Schärer, University of Fribourg and Andermatt Biocontrol AG for support of this work. This work was funded by the Commission for Technology and Innovation CTI of the Swiss Confederation and the Swiss grassland society AGFF. References Bates D., Maechler M. and Bolker B. (2012) lme4: Linear mixed-effects models using S4 classes. R package version 0.999999-0. http://CRAN.R-project.org/package=lme4. Cavers P.B. and Harper J.L. (1964) Rumex obtusifolius L. and R. crispus L. Journal of Ecology 52, 737-766. Faithful I. (2000) Distribution of dock moth in Victoria. Under Control 12, 9-10. Hopkins A. and Johnson R.H. (2002) Effect of different manuring and defoliation patterns on broad-leaved dock (Rumex obtusifolius) in grassland. Annals of Applied Biology 140, 255-262. R Development Core Team (2013) R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria. Scott J.K. and Sagliocco J.L. (1991) Host-specificity of a root borer, Bembecia chrysidiformis (Lep.: Sesiidae), a potential control agent for Rumex spp. (Polygonaceae) in Australia. Entomophaga 36, 235-244. Zaller J.G. (2004) Ecology and non-chemical control of Rumex crispus and R. obtusifolius (Polygonaceae): a review. Weed Research 44, 414-432. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 323 Effect of different grazing regimes on the coverage of Taraxacum spp. under a long-term grazing experiment Supek Š.1, Pavlů V. 1,2, Ludvíková V.1, Pavlů L.1, Gaisler J.2 and Hejcman M.1,2 1 Czech University of Life Sciences Prague, Faculty of Environmental Sciences, Department of Ecology, Prague, Czech Republic 2 Crop Research Institute, Prague, Grassland Research Station Liberec, Czech Republic Corresponding author: supek@fzp.czu.cz Abstract The effect of grazing by heifers under different grazing managements on the cover of Taraxacum spp. was studied on an upland long-term grazing experiment in the Jizerské hory Mts. from 1998 to 2013 The following treatments were applied: (i) intensive grazing (IG); (ii) extensive grazing (EG); (iii) first cut in the late spring and intensive grazing aftermath (ICG); (iv) first cut in the late spring and extensive grazing aftermath (ECG) and (v) unmanaged grassland (U) as the reference area. A significant effect of treatment, year and interaction of treatment and year on the % cover of Taraxacum spp. was revealed. The highest % cover was recorded in the ICG treatment in 2013. The combination of first cut and grazing the aftermath (ICG, ECG) strongly promoted the % cover of Taraxacum spp. up to 2004. After that there was a slight increase of Taraxacum spp. cover under the ICG and IG treatments. The highest % cover of Taraxacum spp. during the whole experiment was in the ICG treatment in 2013. Extensification of grassland management could be used as a simple tool for reduction of Taraxacum spp. Keywords: heifer grazing, dandelion, cutting, management control Introduction Taraxacum spp. (Asteraceae) is a stemless perennial herb, native in Europe, and occupying a wide range of habitats especially pastures, lawns and meadows (Stewart-Wade et al., 2002). High plasticity and ecotype differentiation in ecophysiological traits allow this species to spread along wide environmental gradients for what it is considered to be one of the most aggressive invasive plants around the world (Molina-Montenegro et al., 2013). Although dandelion is classified as a serious agricultural weed in arable fields, its weediness in grassland is not so straightforward, because its high nutritive value and palatability (Marten et al., 1987) can increase the quality of pasture forage (Pavlů et al., 2006). The objective of the study reported in this paper is to evaluate the effect of different grazing intensities on the cover of the dandelion and to recommend which management can be used for dandelion control in grasslands. Materials and methods The study site was performed on experimental grassland in the Jizerské hory Mountains, 10 km north of the city of Liberec, Czech Republic. The long-term grazing experiment called 'Oldřichov Grazing Experimen' (OGE) was established in the spring of 1998 and was arranged in two randomized blocks (Pavlů et al., 2007). The following treatments were studied: intensive grazing (IG), first cut in the late spring and intensive grazing aftermath (ICG), extensive grazing (EG), first cut in the late spring and extensive grazing aftermath (ECG), and unmanaged grassland (U) as the reference area. Development of dandelion % cover was recorded in permanent 1 m × 1 m plots using a continuous grid of nine 0.33 m × 0.33 m subplots in four replications in each paddock. Dandelion covers in each subplots were visually estimated in early May each year from 1998 to 2013. The mean of nine subplots was used for statistical evaluation. Repeated measures Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 324 ANOVA was used to evaluate seasonal development of the cover of Taraxacum spp. One-way ANOVA was used to test the cover of Taraxacum spp. in the particular year. Results A significant effect of treatment, year, and interaction of treatment and year, on the cover of Taraxacum spp. was revealed (Table 1). Table 1 Results of repeated measurements ANOVA analyses for seasonal development of % cover of Taraxacum spp. for years 1998 – 2013. Taraxacum spp. % cover Effect Degree of freedom F-ratio P-value Treatment 4 126.5 <0.001 Year 15 15.8 <0.001 Treatment × Year 60 3.6 <0.001 Immediately after management was imposed in 1998 all types defoliation treatments supported an increase of Taraxacum spp. cover (Figure 1). Figure 1. Changes in % cover of Taraxacum spp. under different treatments for the years 1998-2013. P represents probability value obtained by one-way ANOVA for each year and P < 0.01 was for all analyses, n.s. – nonsignificant result. Significant differences (P < 0.05) according to the Tukey post hoc test are indicated by different letters The lowest cover was recorded in the unmanaged treatment (U) during the whole experiment. Up to 2004, Taraxacum spp. was promoted especially by the combination of cut and grazing aftermath (ICG, ECG); however, the differences among all managed treatments were small. In 2007 there was revealed a striking decrease of Taraxacum spp. in all defoliation plots. The highest decrease was recorded in the ECG treatment (from 29% to 13%), followed by IG and ICG (a decrease about 8%) and EG (7%). After that a slight increase of Taraxacum spp. cover was revealed under the ICG and IG treatments. The highest cover of Taraxacum spp. during the whole experiment was 32%, in the ICG treatment in 2013. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 325 Discussion The higher defoliation intensity under intensive grazing is more favourable for dandelion presence. Sparse canopy and more light at the soil surface under intensive management increase the possibility for germination of dispersed diaspores and of plants succeeding into the generative phase (Pykälä, 2005). Similarly, the most favourable treatment for Taraxacum spp. presence in the course of our experiment was that of cutting in late spring and intensive grazing aftermath. It shows that, in Central European conditions, the applied cutting management in late May or early June can allow dandelion to reproduce and consequently spread its seeds. For that reason early cutting seems to be an appropriate method for range management to control its dispersal and germination of seedlings (Martinkova et al., 2009). Conclusion Intensive grazing was the treatment that most promoted the abundance of Taraxacum spp. and this effect was strengthened by cutting in the late-flowering period. Therefore, an early cut at the beginning of the Taraxacum flowering period could lead to elimination of flowers and thus to reduced seed production. Extensification of grassland management could be used as a simple tool for decreasing the presence of Taraxacum spp. References Stewart-Wade S. M., Neumann S., Collins L. L. and Boland G. J. (2002) The biology of Canadian weeds. 117. Taraxacum officinale G.H.Weber ex Wiggers. Canadian Journal of Plant Science 82, 825–853. Molina-Montenegro M. A., Palma-Rojas C., Alcayaga-Olivares Y., Oses R., Corcuera L. J., Cavieres L. A. and Gianoli E. (2013) Ecophysiological plasticity and local differentiation help explain the invasion success of Taraxacum officinale (dandelion) in South America. Ecography 36, 718-730. Marten G. C., Sheaffer C. C. and Wyse D. L. (1987) Forage nutritive value and palatability of perennial weeds. Agronomy Journal 79, 980-986. Pavlů V., Gaisler J., Hejcman M. and Pavlů L. (2006) Effect of different grazing system on dynamics of grassland weedy species. Journal of Plant Diseases and Protection 20, 377-383. Pavlů V., Hejcman M, Pavlů L. and Gaisler J. (2007) Restoration of grazing management and its effect on vegetation in an upland grassland. Applied Vegetation Science 10, 375–382. Pykälä J. (2005): Plant species responses to cattle grazing in mesic semi-natural grassland. Agriculture, Ecosystems & Environment 108, 109-117. Martinková Z., Honěk A. and Pekár S. (2009) Seed availability and gap size influence seedling emergence of dandelion (Taraxacum officinale) in grasslands. Grass and Forage Science 64, 160–168. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 326 Emergence and survival of Rumex OK-2 (Rumex patientia x Rumex tianschanicus) in grasslands under different management conditions Hujerová R. 1, Gaisler J. 2, Pavlů L. 1, Pavlů V. 1,2 and Hejcman M. 1,2 1 Department of Ecology, Faculty of Environmental Sciences, Czech University of Life Sciences, CZ-165 21, Prague 6 - Suchdol, Czech Republic 2 Crop Research Institute Prague, Department of Plant Ecology and Weed Sciences, CZ-460 01, Liberec, Czech Republic Corresponding author: hujerova@fzp.czu.cz Abstract Emergence and survival of Rumex OK-2 was studied in the north of the Czech Republic (in an experimental garden in Liberec town) in 2013. Three frequencies of cutting were applied: no (0C), one (1C) and three (3C) cuts per year. Seeds of Rumex OK-2 were sown into the sward with different microsite conditions (no gap – gap; fertilizer application and no fertilizer application) in each treatment. The following plant characteristics were measured: number of emerged plants, number of surviving plants, plant height and numbers of leaves. Measurements were made three times per vegetation season (middle of June, end of July and end of September) before cutting. Plants of Rumex OK-2 emerged more in the treatments with gaps. Survival of Rumex OK-2 plants was connected with treatments with gap, especially in the second and the third cutting date. Keywords: weeds, cutting frequency, competition, fertilizers application, gap Introduction Many broad-leaved Rumex species are considered to be the most troublesome weeds in grasslands and arable land worldwide (Zaller, 2004). These plants often colonize grasslands as well as permanent agricultural crops (Novák, 1994; Brant et al., 2006), where they can survive for a long time (Martinková et al., 2009). A new forage and energy-crop hybrid, R. patienta × R. tianschanicus, registered as cv. Rumex OK-2 (hereafter referred as Rumex OK-2) was introduced into the Czech Republic about ten years ago (Usťak, 2007). Rumex OK-2 is described as a perennial (up to 10 years) stress-tolerant plant, characterized by high ecological plasticity, with cold and winter hardiness, tolerance to salt stress and increased humidity (Kosakivska et al., 2008). Before the introduction to the culture it was presented as a competitive species with a low possibility of invasibility (Usťak, 2007). However, it behaves as an invasive weed species, especially in road ditches covered by grasslands in the vicinity of the field where it was previously grown (Hujerová, 2013a). The response of mature plants to different cutting frequencies of Rumex OK-2 is very similar to that of Rumex crispus (Hujerová 2013b). In view of the above-mentioned knowledge we established a manipulative experiment where emergence and survival of Rumex OK-2 in grasslands under different management conditions were studied. Methods and materials A plot experiment was conducted in 2013 at the experimental garden of the Crop Research Institute, Grassland Research Station Liberec, in the northern part of the Czech Republic, under conditions of natural rainfall, temperature and daylight. Twenty seeds of Rumex OK-2 were sown into the sward in May 2013. Twelve factorial treatments were applied: i) three frequencies of cutting- no (0C), one (1C) and three (3C) cuts per year; ii) two levels of disturbance - gap and no gap; iii) two levels of nutrients - fertilizers application and no fertilizer application. The experiment was arranged in four complete randomized blocks with individual plot sizes of 0.5 m × 0.5 m. NPK fertilizer was applied in amounts of 100 kg N ha-1 52 kg K Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 327 ha-1 and 27 kg P ha-1 in 0 15 m × 0.15 m areas allocated in the middle of each plot. Seeds were sown in the same area. We recorded number of emerged plants, number of surviving plants, plant height and numbers of leaves. Measurements were made three times per season (middle of June, end of July and September) before cutting. In the first cutting term the Rumex plants were not defoliated, because they were smaller than cutting height. One-way ANOVA and repeated measures ANOVA were used to evaluate number of emerged plants, number of surviving plants, plant height and numbers of leaves. Results and discussion The number of emerged Rumex OK-2 plants was significantly divided into two groups according to disturbance. In the treatments without gap, up to one plant per plot was found, whereas in plots with gap there were from seven to nine plants after one-and-a-half months after sowing date (Figure1). Figure 1. Development of emerged Rumex OK-2 plants during one-and-a-half months after sowing. Legend: fertilized, non-fertilized, 0C, 1C, 3C, gap, --- no gap. There were no emerged plants found in the no-gap, non-fertilized, no-cutting treatment (NGaNFC0). On the other hand the highest number of emerged plants was in the gap, nonfertilized, no cutting treatment (GaNFC0). It confirmed the results of Carvers and Harper (1964) for Rumex crispus and R. obtusifolius that seed germination is possible when a gap occurs in the established sward. In the first cutting term there were only a few surviving plants in treatments without gaps but several times more of them in treatments with gaps (Figure 2). However, the number of surviving Rumex OK-2 plants significantly decreased in the second cutting term because of high competitive ability of the existing sward. After the third cut only a few of the Rumex OK-2 plants survived in gap treatments. However, due to its fast spring growth and similar tolerance to cutting as R. crispus has (Hujerová, 2013b) we can expect it surviving, with possible flowering and consequent seed production in the next vegetation. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 328 Figure 2. Number of survived plants R. OK-2 in three cutting terms. NGaNFC3, GaFC0, GaFC1, GaFC3, GaNFC0, NGaFC1, GaNFC1, NGaFC3, GaNFC3. NGaNFC1, Conclusion Sward disturbance is the main factor for Rumex OK-2 infestation into existing grasslands. Although in the course of the vegetation season plants of Rumex OK-2 are exposed to high competitive pressure of existing sward, some were still revealed at the end of vegetation season. These plants in the next vegetation seasons can become an important source of seeds and support its expansion into the surroundings. Rumex OK-2 has similar behaviour as other broadleaved docks in Central Europe, so we can expect its further spreading. Acknowledgments This research was supported financially by the Ministry of Agriculture of the Czech Republic (0002700604), Czech University of Life Sciences (CIGA 20124208) and by ESF & MEYS (CZ.1.07/2.3.00/30.0040) References Brant v., Svobodová M., Šantrůček J., Hlavičková D. (2006) The influence of plant covers of set-aside fields and their management on the weed spectrum. Journal of Plant Diseases and Protection 20 (Special Issue), 941-947. Carvers P.B. and Harper J.L. (1964) Biological flora of the British Isles. Rumex obtusifolius L. and Rumex crispus L. Journal of Ecology 52, 737-766. Hujerová R., Gaisler J., Mandák B., Pavlů L. and Pavlů V. (2013a) Hybrid of Rumex patientia and Rumex tianschanicus (Rumex OK-2) as a potentially new invasive weed in Central Europe. Grassland Science in Europe 18, 466-468. Hujerová R, Pavlů V, Hejcman M, Pavlů L and Gaisler J (2013b). Effect of cutting frequency on above- and belowground biomass production of Rumex alpinus, R. crispus, R. obtusifolius and the Rumex hybrid (R. patienta × R. tianschanicus) in the seeding year. Weed Research 53, 378–386. Kosakivska I., Klymchuk D., Negretzky V., Bluma D. and Ustinova A. (2008) Stress proteins and ultrastructural charackeristics of leaf cells of plants with different types of ecological strategies. General and Applied Plant Physiology Special Issue 34, 405-418. Martinková Z., Honěk A., Pekár S. and Štrobach J. (2009) Survival of Rumex obtusifolius L. in unmanaged grassland. Plant Ecology 205, 105-111. Usťak S. (2007) Pěstování a využití šťovíku krmného v podmínkách České republiky (Cultivation and use of fodder sorrel in conditions of Czech Republic). VÚRV, v.v.i., Praha. Zaller J.G. (2004) Competitive ability of Rumex obtusifolius against native grassland species: above- and belowground allocation of biomass and nutrient. Journal of Plant Diseases and Protection 19, 345-351. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 329 Mixed cropping of grass and alfalfa to reduce weed growth Surault F., Julier B. and Huyghe C. INRA, UR4, Unité de Recherche Pluridisciplinaire Prairies et Plantes Fourragères, BP80006, 86600 Lusignan, France Corresponding author: fabien.surault@lusignan.inra.fr Abstract Alfalfa crops, as most pure-stand crops, often require herbicide treatments. The introduction of grass species grown in mixture with alfalfa has been studied in a micro-plot design located at two sites and harvested with an infrequent cutting schedule. Eight meadows were studied: seven alfalfa-grass mixtures (seven grass species tested) and one pure alfalfa. At each cut during three years, the weed proportion in harvested biomass was measured by manual sorting of species. The weed proportion reached up to 60% of the harvested biomass in pure alfalfa, but on the alfalfa-grass mixtures, it was reduced by 75% at one site and 90% at the other site on average in the first six cuts. There was little difference in the proportion of weeds between the different grass species. However, weed proportion tended to be lower with the most aggressive grass species (perennial ryegrass and festulolium) than with the other species. These results indicate that herbicide spraying could be greatly reduced if alfalfa-grass mixtures were cultivated instead of pure alfalfa. In addition, alfalfa-grass mixtures could have even greater advantage than pure alfalfa in cereal rotations to decrease weed pressure. Keywords: alfalfa, grass, legume, mixture, weed Introduction Today, one of the challenges of agriculture is to reduce significantly the use of pesticides, particularly herbicides that contaminate rivers and aquifers. In addition, for some crops such as alfalfa (Medicago sativa), the successive withdrawals of approvals of active substances limit the chemical solutions for weed control. In a pure alfalfa stand grown without herbicide treatment at sowing, the proportion of weeds can represent a significant part of biomass during the first cuts (Spandl et al., 1999) and severely compete with the establishing alfalfa plants. With its erect growth habit and slow aerial growth after sowing or cutting, alfalfa does not cover between-row space during the establishment phase, in early spring or after a cut. These are key periods for the development of weeds in the canopy and their presence leads to losses in forage production and quality and may reduce crop persistency. The emergence of weed seedlings is related to the composition of the active seed bank available in the topsoil. Kruidhof et al. (2008) showed that, because of its limited ability to intercept the light radiation in the establishment phase, alfalfa was much less competitive with weeds than most grasses. The objective of this study was to determine whether the combination of grass species with alfalfa could significantly reduce the proportion of weeds in harvested forage. Several forage grass species with varying characteristics were tested. Materials and methods Two micro-plots (9 m²) trials consisting of binary alfalfa-grass mixtures were sown on two contrasting sites: Somme-Vesle (in summer 2006) in the North-East of France and Lusignan (in summer 2007) in Centre-West of France. The soil in Somme-Vesle is a shallow sandy loam with a pH of 8.3, while in Lusignan it is a deep sandy clay loam with a pH of 6.5. The seven grass-alfalfa mixtures are shown in Table 1. A pure alfalfa (ALF) was included in the Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 330 Table 1. List of binary alfalfa-grass mixtures and weight of seeds sown for each species (kg/ha). Species Grass species associated to alfalfa ALF-TF Tall fescue (Festuca arundinacea) ALF-OG Orchardgrass (Dactylis glomerata) ALF-MF Meadow fescue (Festuca pratensis) ALF-AB Alaska brome (Bromus sitchensis) ALF-TI Timothy (Phleum pratense) ALF-FL Festulolium (Festuca glaucescens x Lolium multiflorum) ALF-RG Ryegrass (Lolium perenne) Species in pure stand ALF Alfalfa (Medicago sativa) Variety Flexy Lupre Préval Hakari Barfleo Lueur Brest Weight of seeds Grass Alfalfa 8.9 14.5 7.3 14.5 7.3 14.5 13.2 14.5 3.0 14.5 11.0 14.5 11.0 14.5 Comete - 22.0 trials as a control. Sowing densities were defined to achieve an equal number of seeds per unit area of both components of binary mixtures and among mixtures. Sowing density of controls are the common practices. A phospho-potassium fertilization was applied during the winter that followed the sowing, with 600 kg/ha of K2O and 180 kg/ha P2O5. No nitrogen fertilization was applied and no weed control was done. Plots were harvested with a forage harvester (Haldrup) and cutting occurred approximately after 42 days of regrowth. The proportion of weeds in the plots was determined from a sample of about 500 g of fresh biomass at harvest time. All species were manually separated (alfalfa, grass, weeds), dried at 60 °C for 72 hours and weighed to determine the proportion of each species as a percentage of total dry matter. The data were subjected to analysis of variance with Statistica software and the results were compared by a Student-Newman-Keuls test. Results and discussion In Somme-Vesle, the pressure of the weed flora was high. The proportion of weeds in the harvested biomass averaged 5.8% but reached 61% (Table 2). In each year, the weed proportions were higher in the first spring cut than in the next cuts, with 24.6, 13.1 and 9.1% for the first, second and third year, respectively. In Lusignan, weed pressure was low. The weed proportion in the harvested biomass averaged 1.5% and reached a maximum of 18%. There was little difference between the first spring cuts and the other cuts. At both sites, the weed proportion was significantly higher in the pure alfalfa stand than in the alfalfa-grass mixtures in the four cuts of the first year and the first two cuts of the second year, but not in the next cuts of the second and third years. In these six first cuts, the weed proportion in the mixtures, compared to pure alfalfa, was reduced by 44 to 90% in Somme-Vesle and by 79 to 98% in Lusignan, depending on the grass species. Even if the weed proportion was low in the mixtures, there was difference between them. Perennial ryegrass and festulolium were the two grass species that most limited the growth of weeds in the first six cuts at both sites, where the frequencies of weeds were the highest. Both grass species are known to establish quickly and are considered as aggressive. In contrast, the highest weed proportion in mixture was observed with brome in Somme-Vesle and timothy in Lusignan. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 331 Table 2. Biomass yield (t dry matter/ha), weed proportion in the harvested biomass (%) in both sites and each cut (C) during the 3 years and significance of effects of grass species and block for each cut in analysis of variance. Species C1 Somme-Vesle Yield 4.7 ALF-TF 29.2ab ALF-OG 18.6ab ALF-MF 20.7ab ALF-AB 34.7ab ALF-TI 26.1ab ALF-FL 10.1a ALF-RG 9.6a ALF 61.3c Anova Species *** Block ns Species Lusignan Yield ALF-TF ALF-OG ALF-MF ALF-AB ALF-TI ALF-FL ALF-RG ALF Anova Species Block C1 4.5 1.7a 0.9a 1.2a 0.7a 2.0a 0.0a 0.0a 10.3b *** ns Year 1 C2 C3 3.2 0.9a 3.6a 1.9a 6.5a 3.8a 1.4a 0.2a 13.7b 2.9 - *** ns Year 1 C2 C3 3.7 0.3a 0.2a 0.3a 3.8a 2.0a 0.0a 0.2a 16.5b 3.0 0.0a 0.3a 4.2a 1.3a 5.0a 0.0a 0.3a 17.6b *** ns *** ns Year 2 C2 C3 C4 C1 2.4 0.0a 0.0a 0.0a 2.1a 0.0a 0.0a 1.4a 15.0b 4.8 11.0a 11.3a 0.7a 34.3b 8.5a 4.3a 3.5a 44.5b 2.4 0.2a 0.0a 0.0a 2.4b 0.0a 0.0a 0.0a 7.6c *** ns *** ns - C4 C1 *** ns Year 2 C2 C3 C4 1.7 0.0a 0.3a 0.2a 0.6a 2.0a 0.6a 0.3a 7.5b 6.1 0.8a 0.7a 1.4a 6.4b 2.2a 1.2a 0.0a 8.7b 6.6 0.3a 3.2a 1.5a 2.3a 0.7a 0.8a 0.5a 5.6a 4.8 0.0a 0.0a 0.3a 0.0a 0.2a 0.5a 0.2a 0.1a 1.4 0.0a 0.2a 0.1a 0.1a 0.0a 0.0a 0.3a 0.0a ns ns ns ns ns ns *** ns *** ns Year 3 C2 C3 C4 5.5 0.7a 6.5a 4.5a 27.6a 8.9a 6.4a 8.4a 18.8a 3.9 0.0a 0.0a 0.0a 1.1a 0.0a 4.2a 4.9a 0.0a 2.3 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 ns ns - C1 ns ns ns ns Year 3 C2 C3 8.4 0.3a 0.9a 0.4a 0.9a 2.3a 1.6a 0.4a 0.2a 5.1 0.0a 0.1a 3.3a 3.4a 0.8a 0.2a 0.0a 0.1a 3.4 0.2a 0.2a 0.0a 0.1a 0.1a 0.2a 0.0a 0.0a 1.3 0.0a 0.2a 0.0a 0.1a 0.1a 0.2a 0.6a 0.0a ns ns ns ns ns ns C4 C1 - 2.8 0 0 0 0 0 0 0 0 ns ns 3.7 0.0a 0.0a 0.0a 0.0a 1.1a 0.0a 0.3a 0.0a C4 Conclusion In the context of limiting the use of pesticides, the combination of a grass with alfalfa can significantly reduce the weed proportion in harvested biomass. This type of mixture contributes to limit or even eliminate herbicide spraying during the establishment of alfalfa. The interest of alfalfa cropping in cereal rotations to limit the development of weeds in the following crops is mainly related to a change in weed flora (Meiss et al., 2010). With the strong reduction of weeds in the various years of the alfalfa-grass mixtures, our study shows that this type of grassland would further reduce weed pressure, in addition to a change in weed flora. References Kruidhof H.M., Bastiaans L. and Kropff M.J. (2008) Ecological weed management by cover cropping: effects on weed growth in autumn and weed establishement in spring. Weed Research 48, 492-502. Meiss H., Médiène S., Waldhardt R., Caneill J., Bretagnolles V., Reboud X. and Munier-Jolain N. (2010) Perennial lucerne affects weed community trajectories in grain crop rotations. Weed Research 50, 331-340. Spandl E., Kells J.J. and Hesterman O.B. (1999) Weed invasion in new stands of alfalfa with perennial forage grasses and an oat companion crop. Agronomy Journal 39, 1120-1124. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 332 Impact of site conditions on natural and fodder value of meadow-pasture communities with different contributions of Urtica dioica L. Strychalska A.1, Kryszak A.1, Kryszak J.1 and Jankowski K.2 1 Department of Grassland and Natural Landscape Sciences, Poznan University of Life Sciences Dojazd 11; 60-632Poznań, Poland 2 Department of Grassland, Siedlce University of Natural Sciences and Humanities, Prusa 14; 08-110 Siedlce, Poland Corresponding author: agastr@up.poznan.pl; akryszak@up.poznan.pl Abstract The effect of habitat conditions on the occurrence of Urtica dioica and the resulting natural and fodder value of swards in meadow-pasture communities were assessed. Habitat conditions, i.e. water content, soil reaction, contents of available forms of phosphorus, potassium and magnesium as well as nitrate nitrogen were analysed using laboratory methods. Species richness, floristic diversity index and fodder-value score of the whole sward were determined in communities with different proportions of common nettle. The highest share of this species was found in well-watered habitats, on slightly acidic or acidic soils, abundant in nitrate nitrogen and low in magnesium and potassium. An increase in the contribution of U. dioica in the sward had a negative effect on its natural and fodder value. Keywords: competitiveness, meadow communities, habitat, utilization Introduction The preservation of semi-natural grassland communities requires their regular and sustainable mowing or pasture use. Both intensive use with excessive fertilization, and also negligence of grassland use, contribute to transformations in their floristic composition, and thus to changes in their fodder value expressed in their palatability, digestibility and chemical composition (Trzaskoś, 1995; Kryszak and Kryszak, 2005). At present, the transformation of floristic composition mainly consists of changes in the proportions of individual species in the sward of meadows and pastures, connected, e.g., with encroachment of new species exhibiting high adaptability (Genovesi, 2004). An example may be provided here by common nettle (Urtica dioica L.), a species increasingly often found in high proportions and in large clusters in swards of permanent grassland. The aim of this study was to indicate habitat conditions promoting the occurrence of Urtica dioica. The effect of the contribution of this species on the natural and fodder value of swards in meadow-pasture communities was also determined. Material and methods Habitat conditions promoting the occurrence of Urtica dioica were determined using laboratory methods: soil moisture content by gravimetry (over-dry method), soil reaction by H2O potentiometry, available potassium, phosphorus and magnesium forms were determined in 0.5 mol/dm3 HCl extract, while the Kjeldahl method was used to determine contents of nitrate nitrogen (Boratyński et al., 1988). Samples for chemical analyses were collected from a depth of 15-20 cm soil from canary grass, foxtail grass and soft grass meadows, ryegrass pastures and those with the dominance of Kentucky bluegrass and red fescue, differing in the share of common nettle: A – absent in the sward, B – < 1 % share in phytocenosis, C – > 1% share. Statistical analysis methods using Canoco for Windows 4.5 (ter Braak and Smilauer, 19972002) were applied in the analysis of results. In addition, based on the floristic composition described in 150 relevés of approximately 100 m2 each, the mean number of species in the Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 333 phytocenoses was determined along with the floristic diversity index (Shannon-Wiener), and fodder value of the whole sward according to Filipek (1973). Results and discussion Results of the study indicate that the occurrence of Urtica dioica depends on soil reaction, and its proportion was found to increase with a decrease in soil pH. Moreover, N-NO3 content influences the share of Urtica dioica in the sward of meadows and pastures. However, it is connected with habitat water contents. In habitats of well-watered canary grass meadows no marked dependence could be observed between the share of this species and N-NO3 content in the soil, which is connected with a deficit of available nitrogen forms. In contrast, in less-moist or in dry habitats a marked dependence was found for an increase in the share of Urtica dioica with an increase in N-NO3 content in the soil. These dependencies are confirmed by the analysis of distribution of variables presented in the diagram (Figure 1), indicating that the most important factors influencing the share of Urtica dioica include soil contents of nitrate nitrogen, water content and soil reaction. Similarly, in moderately moist habitats a trend was observed for an increase in the contribution of Urtica dioica with an increase in soil abundance of available phosphorus forms and a decrease in magnesium and potassium abundance. Explanations: A – without nettle B – share of nettle < 1% C – share of nettle > 1% Ph – canary grass meadow Al – foxtail meadows Ho – Holcus meadows Lo – ryegrass pasture sward Pp-Fr – pasture sward with Poa pratensis and Festuca rubra M – moisture content pH – soil pH in H2O Mg – absorbable form of magnesium (mg·kg-1) K – absorbable form of potasium (mg·kg-1) P - absorbable form of phosphorus (mg·kg-1) N-NO3 – nitrate nitrogen (mg·dm-3) Figure 1. Habitat diversity of plant communities depends on the share of common nettle Results of this study confirm the effect of the share of Urtica dioica on natural and fodder value of swards in the examined communities. In all the communities with a low proportion of Urtica dioica an increase was observed in the mean number of species, which was accompanied by an increase in floristic diversity manifested in the calculated values of the Shannon-Wiener index (Table 1). In contrast, an opposite relationship was observed at an increase in the share of this species (Table 1). Some authors explain this phenomenon by high competitiveness of Urtica dioica (Grime et al., 2007). Moreover, participation of Urtica dioica in the sward exceeding 1% results in a reduction of shares of grass species with high feeding value, such as Lolium perenne, Poa pratensis, Phleum pratense, Alopecurus pratensis and Arrhenatherum elatius, causing at the same time a deterioration of fodder value. A significant reduction of sward fodder value was observed in Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 334 foxtail grass meadows, where the fodder value index dropped from 7.51 (good sward) to 5.72 (poor sward). Table 1. Natural and useful characteristics of studied plant community Number of plant species, FVS Community and utilization H'* average in releves ** A 21 1.7 5.75 Canary grass meadows B 24 1.8 6.18 2-3 x cuting C 15 1.6 4.57 A 17 1.4 7.51 Foxtail meadows B 20 1.9 8.12 2-3 x cuting C 16 1.4 5.72 A 23 1.8 5.12 Holcus meadows B 25 1.9 5.31 1-2 x cuting C 18 1.7 4.82 A 19 1.8 8.09 Ryegrass pasture sward pasturage/grazing: B 22 1.8 8.29 2-3 LU ha-1 C 23 2.1 7.25 18 2.0 6.31 Pasture sward with Poa pratensis and A Festuca rubra B 22 2.1 6.74 pasturage/grazing: 1 LU ha-1 C 20 2.0 6.16 * H/ - floristic diversity index- according to Shannon-Wiener; ** FVS -fodder value score - (Filipek, 1973). Conclusion The occurrence of common nettle (Urtica dioica) is promoted by well-watered habitats, soils with slightly acidic or acid reaction, as well as high soil resources of nitrate nitrogen and low magnesium and potassium contents in soil. An increase in the share of common nettle in the sward has a negative effect on its natural and fodder value. References Boratyński K., Czuba R. and Goralski J. (1988) Chemia rolnicza, Państwowe Wydawnictwo Rolnicze i Leśne, Warszawa, 398 pp. Filipek J. (1973) Projekt klasyfikacji roślin łąkowych i pastwiskowych na podstawie liczb wartości użytkowej. Postępy Nauk Rolniczych 4, 58-69. Genovesi P. and Shine C. (2004) The European strategy on invasive alien species. Nature and Environment 137, 67 pp. Grime J.P., Hodgson J.G. and Hunt R. (2007) Comparative plant ecology: a functional approach to common British species, 2nd edn. Castlepoint Press, Dalbeattie, UK. Kryszak J. and Kryszak A. (2005) Floristic changes in meadow swards after suspension of utilization, Grassland Science in Europe 10, 272-275. Kryszak A., Kryszak J., Klarzyńska A. and Strychalska A. (2009) Influence of expansiveness of select plant species on floristic diversity of meadow communities. Polish Journal of Environmental Studies 18, 1203-1210. Rosnitschek-Schimmel I. (1982) Effect of ammonium and nitrate supply on dry matter production and nitrogen distribution in Urtica dioica. Zeitschrift für Pflanzenphysiologie, 108, 329-341. ter Braak C.J.F. and Smilauer P. (1997-2002) Biometris – Plant Research International. Wageningen. Trzaskoś M. (1995) Some aspects of fodder herbs occurence in different habitat of meadows. Annales Universitatis Mariae Curie-Skłodowska Lublin Sectio E, Supplement 56, 295-299. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 335 Tree and pasture productivity in Pseudotsuga menziesii (Mirb.) Franco silvopastoral systems fertilized with sewage sludge Ferreiro-Domínguez N., Rigueiro-Rodríguez A. and Mosquera-Losada M.R. Crop Production Departament, Escuela Politécnica Superior, Universidad de Santiago de Compostela 27002-Lugo, Spain Corresponding author: mrosa.mosquera.losada@usc.es Abstract In Galician silvopastoral systems (northwest Spain) fertilization with sewage sludge could enhance tree and pasture productivity which is limited by soil acidity. The effect of sewage sludge on tree growth and the pasture production is different depending on the process used to stabilize the sewage sludge. The aim of the present study was to evaluate the effect of fertilization with municipal sewage sludge, which has been stabilized using anaerobic digestion, composting, and pelletization, on tree and pasture productivity compared to control treatments (mineral and no fertilization) in a silvopastoral system under Pseudotsuga menziesii (Mirb.) Franco. Mineral fertilization increased the annual pasture production and reduced the tree heights due to the competition by the nutrients generated between pasture and trees. However, tree height was increased by the application of pelletized sewage sludge applied in split doses. Keywords: agroforestry, sowing, afforestation, anaerobic digestion, composting, pelletization Introduction In Spain, around 65% of sewage sludge is recycled through agricultural soils. This is due to its agronomic value as a source of plant nutrients and organic matter, and its soil-improving qualities (MMA, 2006). Sewage sludge could be used as fertilizer in Galician silvopastoral systems (northwest Spain) in which tree and pasture productivity is limited by low soil fertility as a result of increased acidity. In Europe, sewage sludge should be stabilized before being used as fertilizer. The stabilization process could cause differences in the mineralization rates (EPA, 1994) and, therefore, in tree growth and pasture production. Anaerobic digestion and composting are the most important types of sludge stabilization promoted by the EU (European Directive 86/278) (EU, 1986). However, both types of waste contain high proportions of water which could be reduced by 98% through pelletization of anaerobic sludge via thermic treatment; this reduction consequently reduces the storage, transport and spreading costs compared with anaerobic or composted sludge (Mosquera-Losada et al., 2010). The aim of the present study was to evaluate the effect of fertilization with municipal sewage sludge that has been stabilized using anaerobic digestion, composting, and pelletization, on tree and pasture productivity compared to control treatments (mineral and no fertilization) in a silvopastoral system under Pseudotsuga menziesii (Mirb.) Franco. Materials and methods The experiment was established in Baltar, A Pastoriza (Lugo, Galicia, northwest Spain) at an altitude of 475 m above sea level. Pasture was sown with a mixture of Dactylis glomerata L. var. Artabro (12.5 kg ha-1), Lolium perenne L. var. Brigantia (12.5 kg ha-1) and Trifolium repens L. var. Huia (4 kg ha-1) in December 2004. Plants of Pseudotsuga menziesii (Mirb.) Franco were planted at a density of 952 trees ha-1 after pasture sowing in February 2005. The experimental design was a randomized complete block with three replicates and five treatments. Each experimental unit had an area of 168 m2 and 25 trees planted with an arrangement of 5×5 stems, forming a perfect square. Treatments consisted of (a) no fertilization (NF); (b) mineral fertilization (MIN) with 500 kg ha-1 8:24:16 compound fertilizer Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 336 (N:P2O5:K2O) at the beginning of the growing season and 40 kg N ha-1 before first harvest; (c) fertilization with anaerobically digested sludge (ANA) with an input of 320 kg total N ha-1 before pasture sowing; (d) fertilization with composted sewage sludge (COM) with an input of 320 kg total N ha-1 before pasture sowing and (e) application of pelletized sewage sludge (PEL), which involves a contribution of 320 kg total N ha-1 split as 134 kg total N ha-1 just before pasture sowing in 2004 and 93 kg N ha-1 at the end of 2005 and 2006. Sewage sludge was applied superficially and the calculation of the required amounts was conducted according to the percentage of total N and dry matter contents (EPA, 1994) and taking into account the Spanish regulation (R.D 1310/1990) (BOE, 1990) regarding the heavy metal concentration for sewage sludge application. Tree heights were measured with a graduated ruler in October 2008 and pasture production was determined by taking four samples of pasture per plot at random (0.3 × 0.3 m2) in May and December 2008. In the laboratory, the pasture samples were dried (72 hours at 60ºC) and weighed to estimate dry matter production. Annual pasture production in 2008 was calculated by summing the consecutive harvests of the pasture production in that year. Data were analysed using ANOVA and differences between averages were shown by the LSD test, if ANOVA was significant. The statistical software package SAS (2001) was used for all analyses. Results and discussion In this study, tree height was lower in the MIN treatment than in the NF and PEL treatments (P<0.001) (Figure 1). Tree heights 100 a a ab 40 ab b 20 NF MIN Mg ha-1 cm 80 60 Annual pasture production 7.5 6.0 4.5 a b bc 3.0 ANA COM PEL NF MIN ANA c COM bc PEL Figure 1. Tree heights (cm) (a) and annual pasture production (t [Mg] ha -1) (b) under the different fertilizer treatments in 2008. NF: no fertilization, MIN: mineral; ANA: anaerobic sludge; COM: composted sludge and PEL: pelletized sludge. Different letters indicate significant differences between treatments. Vertical lines indicate mean standard error. However, annual pasture production was increased with mineral fertilization (MIN) compared with the other treatments (NF, ANA, COM and PEL) (P<0.05). These results demonstrate a clear competition between pasture and trees in the MIN treatment, probably due to the intermediate soil pH (water soil pH: 5.6) which did not limit pasture development and, therefore, allowed a positive response of pasture to the mineral fertilization which consequently reduced the tree growth. Moreover, in this experiment, the competition generated between pasture and trees was high because the trees were in the first development stages and, therefore, their roots occupied the same soil depth as that of the pasture roots. Several studies have described that different root depths increase the sustainability and efficiency of use fertilizers (Nair and Kalmbacher, 2005), which used to be lower than 40% in agronomic soils (Jarvis and Menzi, 2004). Other authors, such as Rigueiro Rodríguez et al. (2000) and Mosquera Losada et al. (2006) also observed similar results to those found in our study in the first years after the establishment of silvopastoral systems in agronomic soils with Pinus radiata D. Don. On the other hand, the positive effect of fertilization with pelletized sludge on tree height was previously observed by Rigueiro-Rodríguez et al. (2010) in silvopastoral systems established Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 337 under Fraxinus excelsior L. and could be explained because this type of sewage sludge was applied several times, thus facilitating the incorporation of sludge to soil and subsequent extraction of nutrients by trees. In general, organic fertilizers are characterized by their slow release of nutrients gradually over time due to their low mineralization rate, which is very important for trees as they are able to make better use of nutrients slowly released than pasture. Finally, it should be noted that PEL, besides increasing the tree height compared with MIN treatment, also reduced the application and storage costs compared with the ANA and COM due to its lower proportion of water, and therefore the use of this type of PEL sludge as fertilizer should be recommended. Conclusion Mineral fertilization increased annual pasture production and reduced tree heights due to the enhancement of the negative competition by nutrients generated between pasture and trees. However, the application of pelletized sewage sludge, split several times, implied an increase of the tree heights. References BOE (Spanish Official Bulletin) (1990) Royal Decree 1310/1990 29th October 1990 that regulates the use of sewage sludge, http://www.boe.es/boe/dias/1990/11/01/pdfs/A32339-32340.pdf (Access confirmed 2 January 2014). Ministerio Agricultura, Pesca y Alimentación, Madrid, Spain. EPA (Environmental Protection Agency) (1994) Land application of sewage sludge. A guide for land appliers on the requirements of the federal standards for the use of disposal of sewage sludge 40 CFR Part 503, Environment Protection Agency, Washington D.C., USA, 105 pp. EU (European Union) (1986) DOCE nº L 181 04/07/1986.Council Directive 86/278/EEC of 12 June 1986 on the protection of the environment and, in particular of the soil, when sewage sludge is used in agriculture. http://eurlex.europa.eu/LexUriServ/LexUriServ.do?uri=CELEX:31986L0278:ES:HTML (Accessed 2 Jan 2014). Jarvis S.C. and Menzi H. (2004) Optimising best practice for N management in livestock systems: meeting production and environmental targets. Grassland Science in Europe 9, 361–372. MMA (Spanish Environment Ministry) (2006) Integrated waste plan 2007-2015. http://www.mma.es/secciones/calidad_contaminacion/residuos/planificacion_residuos/pdf/borradorpnir_anexo5. pdf (Access confirmed 2 January 2014). Mosquera-Losada M.R, Fernández-Núñez E. and Rigueiro-Rodríguez A. (2006) Pasture, tree and soil evolution in silvopastoral systems of Atlantic Europe. Forest Ecology and Management 232, 135–145. Mosquera-Losada M.R., Muñoz-Ferreiro N. and Rigueiro-Rodríguez A. (2010) Agronomic characterization of different types of sewage sludge: policy implications. Waste Management 30, 492–503. Nair P.K.R. and Kalmbacher R.S. (2005) Silvopasture as an approach to reducing nutrient loading of surface water from farms. In: Mosquera-Losada M.R., McAdam J. and Rigueiro-Rodríguez A. (eds) Silvopastoralism and Sustainable Land Management, CAB International, Wallingford, UK, pp. 272–274. Rigueiro Rodríguez A., Mosquera Losada M.R. and Gatica Trabanini E. (2000) Pasture production and tree growth in a young pine plantation fertilization with inorganic fertilizers and milk sewage in northwestern Spain. Agroforestry Systems 48, 245-256. Rigueiro-Rodríguez A., Ferreiro-Domínguez N. and Mosquera-Losada M.R. (2010) The effects of fertilization with anaerobic, composted and pelletized sewage sludge on soil, tree growth, pasture production and biodiversity in a silvopastoral system under ash (Fraxinus excelsior L.). Grass and Forage Science 65, 248–25 SAS (2001) SAS/Stat User´s Guide: Statistics, SAS Institute Inc., Cary, NC, USA, 1223 pp. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 338 Improved light availability of legumes in moderately N-fertilized mixed swards Frankow-Lindberg B.E. 1 and Wrage-Moennig N.2 1 Department of Crop Production Ecology, Box 7043, Swedish University of Agricultural Sciences, SE-750 07 Uppsala, Sweden 2 Faculty of Life Sciences, Marie-Curie-Str. 1, Rhine-Waal University of Applied Sciences, 47533 Kleve, Germany Corresponding author: Bodil.Frankow-Lindberg@slu.se Abstract A field experiment with five grassland species was carried out according to a simplex design in southern Sweden. The species represented the functional groups grasses, legumes and forbs, and were grown in two mixture types where but one species differed between them. Light transmission through the canopy was measured before each of four harvests. The harvested material was sorted into species, ground and analysed for δ13C. The δ13C signature of all species but one was positively correlated with light transmission. There was also a positive correlation between the δ13C signature of all species except one and the grass proportion of the plant community. The results indicate that mixing species with different leaf morphologies improves light availability and assimilation of legumes in particular. Keywords: δ13C signatures, forb, grass, legume, light transmission Introduction Empirical evidence indicates a positive relationship between grassland diversity and yield in both extensively and intensively managed systems (e.g., Finn et al., 2013). One possible reason for this is complementary use of available resources among species, such as resource partitioning by legumes and non-legumes with respect to N acquisition (e.g., FrankowLindberg and Dahlin, 2013). Other complementarities may involve differences in the spatial arrangement of leaves between species in mixed plant communities (Anten and Hirose, 1999). Aboveground complementarities in light interception have so far received less attention than have belowground complementarities. The few relevant studies have been performed in extensively managed grasslands, where potentially tall-growing species have been able to fully exhibit their growth potential (e.g. Jumpponen et al., 2005). Plant δ13C signatures (i.e., the ratio of the stable isotopes of carbon, 12C, and 13C in plant leaves) are affected by environmental conditions such as light availability (i.e., more depleted δ13C signatures occur with poor light availability). The aim of the study was to evaluate the effect of functional group composition on the δ13C signatures of the species grown. Materials and methods A field experiment was established by drilling at Svalöv, Sweden (55° 55' N 13° 07' E, 55 m a.s.l.), in June 2007. The climate is cold–temperate with an annual mean temperature of 7.7 °C and annual mean precipitation of 700 mm. Twenty-two mixtures of L. perenne, P. pratense, T. pratense, and C. intybus (Mixture type 1), ten mixtures of L. perenne, P. pratense, T. pratense and M. sativa (Mixture type 2), four monocropped stands each of L. perenne, P. pratense and T. pratense, and two monocropped stands each of C. intybus and M. sativa were established according to a simplex design, as described in detail in Frankow-Lindberg and Dahlin (2013). All mixtures and monocropped stands were sown at two seeding densities, i.e., 100% (high density) and 50% (low density) of the recommended seeding rates used in official varietytesting in Sweden, corrected for actual germination rates. The seeding rates in the high-density monocropped stands were 14.2 (P. pratense), 28.3 (L. perenne), 26.1 (T. pratense), 8.4 (C. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 339 intybus), and 26.3 (M. sativa) kg ha–1. In total, 48 plots were arranged in a completely randomized design, with an individual plot size of 17 m2. In the harvest years, 100 kg of N ha– 1 yr–1 was applied in split dressings (i.e., 40 kg of N ha–1 in early spring and 20 kg of N ha–1 for each summer regrowth in 2009). The plots were harvested four times in 2009 (i.e., 20 May, 24 June, 29 July, and 2 Sept.). The light transmission through the canopy (i.e., percent of incoming light (PAR)) of each plot was recorded before each harvest using a LiCor Quantum sensor (1 m long, five readings per plot) connected to a Quantum meter (LI-189, LM 189; Li-Cor, Lincoln, NE). The sown fractions from all harvests were ground per species to pass through a 1 mm screen, sub-sampled by riffle splitting, ball milled, and finally analysed for 13C abundance, i.e., δ13C expressed relative to international standard V-PDB (Vienna PeeDee Belemnite) using a PDZ Europa ANCA-GSL interfaced to a PDZ Europa 20-20 isotope ratio spectrometer (Sercon Ltd., Cheshire, UK). Linear correlations were calculated: between individual species’ δ13C signatures as the dependent variable and (i) light transmission through the canopy and (ii) functional group proportions of the harvested biomass as the independent variables. These were performed as completely randomized repeated-measures analyses with variables for sown density and mixture type included as fixed factors. Interactions between the independent variables and the fixed factors and between the independent variables and harvest occasion were also included. Results and discussion Increasing light transmission through the canopy was positively correlated with the δ13C signatures of all species (P < 0.05) except C. intybus. Furthermore, there was a significant positive correlation between the δ13C signatures of all species except C. intybus and the grass proportion in the harvested biomass (P < 0.05; Fig. 1). -26.5 -27.5 0.0 0.5 1.0 0.0 0.5 1.0 0.0 0.5 1.0 -28.5 (a) -29.5 -30.5 δ13C -31.5 R2=0,25** R² = 0.02 R² = 0.08 -26.5 -27.5 0.0 0.5 1.0 0.0 0.5 1.0 0.0 0.5 1.0 -28.5 -29.5 -30.5 (b) R2 = 0.24* R2 = 0.36*** R2 = 0.27* -31.5 Grass proportion Figure 1. Correlations between grass proportion in the harvested biomass and species’ shoot δ 13C signatures of a) perennial ryegrass and b) red clover in the first three harvests when light transmission was measured the day before each harvest occasion. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 340 The positive correlation between δ13C signatures and light transmission was strongest for the two legumes, which suggests that, despite height differences between them, their more horizontal leaf arrangement was a disadvantage in the competition for light. It is often noted that the N2 fixation of legumes increases when they are grown in mixtures rather than monocropped stands, and this was also observed in the present experiment (Frankow-Lindberg and Dahlin, 2013). Part of this increase is likely due to the uptake of soil N by non-legume species, forcing legumes to increase N2 fixation (Nyfeler et al., 2011), but the improvement in light conditions for legumes in mixtures with grasses may also make more energy available for this energy-demanding process. Conclusion In conclusion, our results suggest that mixing species of contrasting leaf morphologies and biomass distribution contributed to (i) increased light uptake by mixtures over monocropped non-legumes and (ii) better light availability for legumes in mixtures than monocultures. Acknowledgements This work was funded by the Swedish Research Council for Environment, Agricultural Sciences and Spatial Planning, contract 2005-3470-4745-69 and by the Behms Fund. References Anten N.P.R. and Hirose T. (1999) Interspecific differences in above-ground growth patterns result in spatial and temporal partitioning of light among species in a tall-grass meadow. Journal of Ecology 87, 583–597. Finn J.A., Kirwan L., Connolly J. et al. (2013) Ecosystem function enhanced by combining four functional types of plant species in intensively-managed grassland mixtures: a three-year continental-scale field experiment. Journal of Applied Ecology 50, 365–375. Frankow-Lindberg B.E. and Dahlin A.S. (2013) N2 fixation, N transfer, and yield in grassland communities including a deep-rooted legume or non-legume species. Plant and Soil 370, 567–581. Jumpponen A., Mulder C.P.H., Huss-Danell K. and Högberg P. (2005) Winners and losers in herbaceous plant communities: insights from foliar carbon isotope composition in monocultures and mixtures. Journal of Ecology 93, 1136–1147. Nyfeler D., Huguenin-Elie O., Suter M, Frossard E. and Lüscher A. (2011) Grass–legume mixtures can yield more nitrogen than pure stands due to mutual stimulation of nitrogen uptake from symbiotic and non-symbiotic sources. Agriculture, Ecosystem and Environment 140, 155–163. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 341 Nitrogen application strategies to mixed grass-legume leys Frankow-Lindberg B.E.1 and af Geijersstam L.2 1 Department of Crop Production Ecology, Box 7043, Swedish University of Agricultural Sciences, SE-750 07 Uppsala, Sweden 2 Swedish Rural Economy and Agricultural Society, Flottiljvägen 18, SE-392 41 Kalmar, Sweden Corresponding author: Bodil.Frankow-Lindberg@slu.se Abstract A field experiment was established in 2011 with the aim to evaluate different N application strategies on yield of mixed grass-clover leys. The treatments were 0, 40, 90, 160 or 250 kg N ha-1, applied in single or split applications. The ley crop was harvested three times yearly in 2012 and 2013, and DM yield, botanical composition and digestibility were determined. The response to N application declined from the first to the second harvest year, despite a slight overall decrease in clover content. The economic return of N applications to the last harvest each year was negative for all treatments. Keywords: grass-clover, red clover, white clover, N response Introduction Legume-rich leys can yield approximately the same as all-grass leys fertilized with 200 kg nitrogen (N) ha-1 (Kornher, 1982). However, even though grass-legume leys are the most commonly sown leys in Sweden most farmers apply N either in the form of slurry and/or mineral N fertilizers. This boosts yield and contributes to a more balanced botanical composition. Todays’ seeding mixtures include more species than in the past, and there is a lack of Swedish data on how different N strategies affect yield, N response and digestibility of the typical legume-rich leys sown today. Therefore a field experiment was established to provide such data. Materials and methods A field experiment was established at Färjestaden, Sweden (56◦38 N, 16◦28 E) on 29 April 2011. The climate is cold-temperate with an annual mean temperature of 6.9◦C and an annual precipitation of 500 mm. The soil at the site was a sandy soil with a pH of 6.7 containing 2.7% organic matter and soluble phosphorus and potassium were 20.4 and 6.9 mg 100g-1 soil, respectively. The experimental plots received 19 kg of P and 199 kg of K in 2012, and 121 kg of K and 22 kg sulphur in 2013. The plots were undersown in spring barley which was harvested on 11 August 2011. The seeding mixture contained 14% Trifolium pratense (cvs. Nancy and Rajah), 5% T. repens (cv. Klondike), 38% Phleum pratense (cv. Lischka), 9% Lolium perenne (cvs. Kentaur and Foxtrot), and 34% Festulolium hybrid (cv. Hykor). The total seeding rate was 22 kg ha-1. The plots were harvested three times in 2012 (29 May, 9 July, 27 August) and 2013 (6 June, 16 July, 26 August). Samples for the determination of dry matter, botanical composition and digestibility were taken at all harvest occasions. The experimental treatments were different N-application strategies. Here we report data from five of the treatments, namely: (i) no N application, (ii) 40 or (iii) 90 kg N ha-1 in spring and thereafter no N applications, (iv) 90 kg N ha-1 in spring + 35 kg N ha-1 to each regrowth, and (v) 120 kg N ha-1 in spring + 65 kg N ha-1 to each regrowth. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 342 Results and discussion There was no loss of plants despite the very long and cold winter in 2012-2013. Thus, total yield was high in both harvest years and averaged 12.5 tons and 13.2 tons ha-1 in 2012 and 2013, respectively. Approximately 50% of the yield was harvested at the first harvest occasion. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 343 0 250, HY1:H1, 7.19 40 250 160, HY2:H1, 6.67 90, HY1:H1, 6.78 90, HY2:H1, 6.46 40, HY1:H1, 6.42 40, HY2:H1, 6.03 0, HY1:H1, 5.29 0, HY2:H1, 5.87 250, HY1:H2, 4.08 DM yield (tons ha-1) 160 250, HY2:H1, 6.85 160, HY1:H1, 6.57 160, HY1:H2, 3.64 90, HY1:H2, 3.06 40, HY1:H2, 3.12 0, HY1:H2, 3.01 90 250, HY1:H3, 2.78 250, HY2:H2, 3.89 250, HY2:H3, 3.61 160, HY2:H2, 3.54 160, HY2:H3, 3.46 90, HY2:H2, 3.32 90, HY2:H3, 3.18 40, HY2:H2, 3.33 40, HY2:H3, 3.36 0, HY2:H2, 3.15 0, HY2:H3, 3.13 40, HY1:H3, 2.75 0, HY1:H3, 2.66 160, HY1:H3, 2.57 90, HY1:H3, 2.47 Figure 1. The effect of the amount of N applied on the yield of DM in each harvest year (mean ± LSD) and at harvest occasion. HY1 and HY2 refer to the first and the second harvest years, respectively, and H1, H2 and H3 refer to the first, second and third harvest occasion, respectively. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 344 There were significant yield differences among treatments in the two first harvests in the first harvest year (P<0.001). The result was similar in the second harvest year, but with smaller differences among treatments. Applying N to the regrowths resulted in a significant yield increase (P<0.001) compared to the treatments that did not receive any N in the second harvest both years. In the third harvest, there was no significant difference among treatments in the first harvest year, and small, but significant, differences among some treatments in the second harvest year. The clover content in the harvested DM was high in both harvest years, even though it declined somewhat from the first to the second harvest year (Table 1). This was mainly due to the content of red clover, which declined somewhat over the second harvest year, while that of white clover increased in all treatments (data not shown), mainly in the treatments that did not receive any N application to the summer regrowths. The drop in clover content over time, however, was largest at the highest N applications rates, which might be the reason for the increase in DM yield response to N in the last cut 2013 in these treatments. Table 1. Dry matter yield response to nitrogen (kg DM kg N-1) and clover content at each harvest occasion (% of dry matter) 0+0+0 40+0+0 90+0+0 90+35+35 120+65+65 2012 kg DM kg N-1 Clover, % of DM H1 H2 H3 H1 H2 H3 41 59 74 14.2 33 57 76 28.3 19 42 66 16.6 18.0 -2.5 25 31 43 15.8 16.5 1.8 22 17 32 kg DM kg N-1 H1 H2 H3 8.9 4.0 6.6 11.1 9.4 8.2 11.4 7.4 2013 Clover, % of DM H1 H2 H3 38 53 61 35 56 58 23 44 57 19 31 32 17 34 23 The DM yield response to N application was larger in the first harvest year than in the second harvest year (Table 1), despite similar or lower clover contents in the second harvest year compared to the first harvest year. The N response was similar at the two first harvest occasions, and poorer in the last summer regrowth, in the treatments that received N to the regrowths. This can be explained by the decrease in grass contribution over the season in all treatments. The overall response to N application was 11.4 and 9.5 (160 kg N ha-1), and 12.4 and 8.8 (250 kg N ha-1) kg DM kg N-1, in the first and the second harvest years, respectively. This is a slight increase in relation to the responses obtained with seeding mixtures in the past, based on T. pratense, P. pratense and Festuca pratensis (Kornher, 1982). The in vitro digestibility of the crop was unaffected by the treatments in all harvests except the last harvest in the first harvest year, when the treatment receiving 250 kg N ha-1 had the highest digestibility (P<0.001). Assuming a value of the harvest of 0.14 € per kg DM, an N cost of 0.89 € per kg, and a spreading cost of 17.78 € per occasion, we find that the application of N was economic for the first two harvests in 2012. This was not the case in 2013, when the spring application of N was uneconomic to all treatments except the treatment with the highest application rate, and then barely so. However, the applications to the second harvest in 2013 yielded a positive economic return. In neither year was it economic to apply N to the last summer regrowth. Conclusion The response to N applications was strongest in the first harvest year. Applying high N rates the first harvest year resulted in a decline in clover content in the second harvest year. In neither harvest year was it economical to apply N to the last harvest. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 345 Acknowledgements This work was funded by Swedish Farmers’ Foundation for Agricultural Research through Sverigeförsöken References Kornher A. (1982) Vallskördens storlek och kvalitet. Inverkan av valltyp, skördetid och kvävegödsling. Grovfoder. Forskning – tillämpning, No. 1. In Swedish. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 346 Potential of short-term nitrogen transfer between Trifolium repens and the grasses Festuca gr. rubra and Brachypodium pinnatum in highland grasslands Canals R.M., San Emeterio L., Gutiérrez R. and Juaristi A. Dpto. Producción Agraria, UPNA, Campus Arrosadia s/n. 31006 Pamplona, Spain Correspondence to: rmcanals@unavarra.es Abstract This research was undertaken to determine the occurrence of short-term N transfer between the legume Trifolium repens and grasses Festuca gr. rubra and Brachypodium pinnatum. In Europe, F. rubra is always associated with species-rich grasslands whereas B. pinnatum may constitute, under particular conditions, monospecific covers. The hypothesis underlying this experiment is that species that are more reliant on short-term N transfer from legumes might be more prone to constitute multispecific covers (and vice-versa). Undisturbed blocks of vegetation and soil were collected in highland grasslands and transplanted to a greenhouse, where a 15N leaf-feeding experiment was applied to study the fate of the labelling. One month later, a high percentage of the 15N applied was recovered in soils (up to 25%, on average) but, despite the 15N exuded in the soil, neither F. rubra nor B. pinnatum were significantly 15N enriched. The high C/N of these soils and the occurrence of some outliers suggest that this research merits more study. Keywords: N acquisition, legume-grass interaction, semi-natural grassland, species-rich cover. Introduction In highland grasslands over Europe, grasses are the dominant taxa. In floristic inventories made in species-rich grasslands their occurrence often exceeds 90% of the total cover. The grasses constitute a diverse group of species that mostly share endemicity and adaptation to a harsh climate and frequent soil-nutrient constraints. The relationship that these dominant grasses establish with the rest of the species influences the final floristic diversity and community functionality. In grasslands of the Pyrenees, under given conditions of change, some grass species that belong to species-rich communities, such as Nardus stricta and Brachypodium pinnatum, may increase and displace other species, degrading the community. The process has also been reported in other European grasslands. The dominance of N. stricta has been associated with decreasing herbivory and with selective grazing by sheep (Sebastià et al., 2008) and the constitution of monospecific covers of B. pinnatum is linked to prescribed burnings and to low grazing pressure, among other factors (Canals et al., 2014). Both species develop a powerful rooting system and a dense turf, and have a rapid decline of digestibility, which make them less attractive to livestock. Although these traits may help to explain its potential to displace other species in the community, there are other successful, competitive and productive grasses, such as Festuca gr. rubra and Agrostis capillaris that, even being dominant and in conditions of poor grazing, do not cause a loss of community richness and diversity. Different reasons may help to explain what makes a given species more prone than others to displace other taxa in the community (or in contrast, what makes it more prone to coexistence). One hypothesis is that some species may be more reliant than others on neighbouring species, i.e., facilitation is relevant for species development. The assumption is that the species needs diversity or, at least, the occurrence of some given taxa (such as legumes), to develop successfully. Facilitation mechanisms are known to be very important in constrained environments, as those of highlands. The availability of critical nutrients for plant growth, such as nitrogen (N), is limited, since mineralization is restricted and microbiota compete efficiently for N (Jaeger et al., 1999). Plants have developed different strategies for an efficient NGrassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 347 acquisition in this situation. In this research we are interested in assessing whether a short-term N transfer (Figure 1) occurs between the legume Trifolium repens and the grasses B. pinnatum and F. rubra. Figure 1. Conceptual model of short-term N transfer pathways between a legume and a grass. Methods In early March 2012, thirty 20×20×20 cm undisturbed blocks of soil and vegetation containing the three species of the study were collected in a semi-natural grassland of the south-western Pyrenees (1100 m a.s.l.; 43º0/N, 1º10/W). The blocks were transported to UPNA greenhouse facilities and placed intact in rectangular pots that had a bottom layer of sand and turve. Plant development was ensured over the following months by regular watering with distilled water. On 24 May 2012 the isotopic 15N solution was applied to T. repens plants as described by Marty et al. (2009). Grasses were cut to 1-cm of the soil the day before to reduce 15N dilution, and 20-40 T. repens leaflets per pot were labelled by placing 1µL of (15NH4)2SO4 50mM (at 15N 99 atom. %) in two wounds of 2 mm2 at the abaxial face of leaflets. Nineteen pots were labelled and two pots were used as controls for 15N natural abundance (9 pots were discarded because of the disappearance of the legume or the two grasses). Thirty days after the labelling, plants were removed from the soil and separated into species and tissues (green aboveground and roots). Plant samples were rinsed three times with distilled water and dried at 65ºC for 48h for dry weight determinations. A composite sample of soils was obtained per pot, which was dried at 100ºC till constant weight. Then, plant and soil samples were finely ground and sent to SIRFUC Davis for isotopic analyses with a mass spectrometer (PDZ Europa ANCA-GSL, Sercon Ltd). Dependent variables considered in the statistical analyses were % N total, 15N enrichment (ug15N/g) and % 15N recovered: µg 15N/g = [% N total x (sample 15N – natural 15N)] x 100 % 15N recover = [(µg15N recovered/pot)/(µg15N initially applied)/pot] x 100 Data on % total N were analysed using a parametric split-plot design (species was the mainplot, tissue the subplot and pot the replicate). Plant 15N enrichment and % 15N recovered were analysed using a T test, which compared whether the average of enrichment/recovery differed significantly from zero. 15N enrichment of F. rubra leaves could not be transformed successfully and was analysed by the Wilcoxon rank procedure. Software used was SPSS Statistics and Statistix v.8. Results and discussion Total N in plants differed significantly between single treatments (species F=74.53, P<0.0001; tissues F= 212.86; P<0.001) and their interaction (F=22.53, P<0.001). As expected, the legume displayed the highest % N in plants. However, when considering tissues separately, aerial parts of T. repens and F. rubra had similar % total N, higher than those in B. pinnatum, whereas legume roots concentrated more N than roots of grasses. Regarding the 15N enrichment, values were different from zero for T. repens (troot=13.741 P<0.001, taboveground=6.667 P<0.001) and for soil (tsoil=3.747 P<0.001). A major recovery of Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 348 15 N occurred in soils, 24.8%±6.27 (average ± std error), followed by T. repens 11.91%±3.01, which indicated that the enriched solution penetrated within the plant and was exudated to the soil in a high percentage (almost 1/4 in the soil after one month). Despite 15N exudation, neither F. rubra nor B. pinnatum tissues displayed a significant 15N enrichment (F.rubra troot=0.16 P=0.869, Wilcoxon testaboveground P=0.841; B. pinnatum troot=-1.148 P=0.272, taboveground=0.266 P=0.795) (Figure 2). These results contrast with those of Marty (2009) who found higher 15N enrichments in F. eskia than in N. stricta in a comparable experiment. On the one hand, the high C/N ratio of our soil, around 15, suggests a rapid use of released N and a strong competition with microbes, which may outcompete grasses for the use of N exudates (Jaeger, 1999). On the other hand, the occurrence of an outlier in pot 20, where both T. repens and F. rubra exhibit the highest 15N enrichment (T. repens: 44.44 µg15N/g aboveground and 9.99 µg15N/g belowground, and F. rubra: 2.43 µg15N/g aboveground), leaves open the possibility of N transfer between both species, which deserves more study. Figure 2. A) Excess of 15N (µg/g) in aboveground and belowground tissues of F. rubra. B) Excess of 15N (µg/g) in aboveground and belowground tissues of B. pinnatum. Acknowledgements Financial support was provided by the Spanish Ministry of Science and Innovation (CGL 201021963 and CGL2011-29746), and by the European POCTEFA Programme (EFA34/08). References Canals R.M., Pedro J., Rupérez E. and San Emeterio L. (2014) Nutrient pulses after prescribed winter fires and preferential patterns of N uptake may contribute to the expansion of Brachypodium pinnatum (L.) Beauv in highland grasslands. Applied Vegetation Science. DOI: 10.1111/avsc.12088 Jaeger C., Monson R., Fisk M. and Schmidt S. (1999) Seasonal partitioning of nitrogen by plants and soil microorganisms in an alpine ecosystem. Ecology 80, 1883-1891. Marty C., Pornon A., Escaravage N., Winterton P. and Lamaze T. (2009) Complex interactions between a legume and two grasses in a subalpine meadow. American Journal of Botany 96, 1814-1820. Sebastià M.T., de Bello F., Puig L. and Taull M. (2008) Grazing as a factor structuring grasslands in the Pyrenees. Applied Vegetation Science 11, 215-223. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 349 Root architecture of interspecific hybrids between Trifolium repens L. and Trifolium ambiguum M. Bieb. and their potential to deliver ecosystem services Marshall A.H., Lowe M. and Sizer-Coverdale E. Institute of Biological, Environmental and Rural Sciences, Aberystwyth University, Gogerddan, Aberystwyth, Ceredigion, SY23 3EE, United Kingdom. Corresponding author: thm@aber.ac.uk Abstract The potential of grasslands to deliver ecosystem services and mitigate some of the impacts of climate change is increasingly being recognized. Backcross hybrids between the stoloniferous Trifolium repens L., and the related rhizomatous species T. ambiguum M. Bieb have been produced using T. repens as the recurrent parent. The differences between parental species and the backcrosses in root morphology were studied in 1-m long pipes. The parental species differed in root distribution and in root weight distribution, with root weight of T. ambiguum significantly greater than T. repens in the 0.1 m to 0.5 m root zone. The backcrosses exhibited root characteristics intermediate between the parents. The extent to which such differences in root architecture may influence soil structure and deliver ecosystem services is discussed. Keywords: Trifolium repens, interspecific hybrids, moisture stress, root distribution, ecosystem services. Introduction Adaptation of agriculture to predicted climate change scenarios is increasingly recognized as important for UK grassland production. Increasing winter rainfall leading to flooding and greater incidence of summer droughts are likely to be more common (Macleod et al., 2013). Breeding of improved forage varieties able to tolerate periods of variable rainfall and that can help alleviate the impacts of flooding is now a key objective of the IBERS forage breeding programmes. The most important forage legume component of temperate pastures is white clover (Trifolium repens L.), a nitrogen-fixing species that produces forage of high quality. Hybrids have been successfully developed between white clover and Caucasian clover (T.ambiguum M. Bieb) to introgress the rhizomatous trait from Caucasian clover into white clover (Marshall et al., 2001) as a strategy for improving tolerance of moisture stress. Comparison of the BC1 and BC2 hybrids with the parental species (Marshall et al., 2001) showed that the backcross hybrids maintained a higher leaf relative-water content (RWC) than white clover at comparable levels of soil moisture. The basis of this improved drought tolerance is unclear although preliminary studies suggest that the hybrids have a higher proportion of their DM yield in roots than T. repens (Marshall et al., 2001). The objective of this study was to analyse the root distribution of the parental species and backcross hybrids. Materials and methods Hybrid development and their morphological characterization were described previously (Marshall et al., 2001). Four genotypes of the T. repens, T. ambiguum, BC1 and BC2 hybrids were cloned to provide 6 plants of each genotype. Clonal plants of the parental species and hybrids were transplanted into plastic pipes (1 m deep × 15 cm diameter) with several drainage holes in the base into which was inserted a polythene tube filled with soil. The pipes were placed on a gravel bed within a glasshouse maintained at ambient temperature. After 10 weeks the polythene tube was removed from the pipe and the aboveground foliage cut with hand-held shears and the root column separated into 10-cm deep longitudinal sections. The roots within each of these sections were carefully removed by gently washing each section under running Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 350 water. The dry weight of the roots within each 10 cm section was then determined after drying at 80ºC for 24 hours in a forced draught oven. The experiment was established as a split-plot design with four replicate blocks, species as whole plots and genotypes as sub-plots. Results and discussion Within the relatively limited objectives of this experiment, the parental species T. repens and T. ambiguum differed in the distribution of roots within the soil profile and in root weight distribution. There was a significant difference between species in root dry weight up to depths of 0.5 m, and also significant differences between genotypes within species (Table 1). However, at depths below 0.5 m, differences between species were small and insignificant and are not shown. Table 1. Significance levels for effect of species and genotype within species on root dry weight at different depths Root depth (m) 0.1 0.2 0.3 0.4 0.5 Species * ** *** *** *** Genotypes within species *** ** ** ** NS NS, not significant; *P<0.05; ** P<0.01; ***P<0.001 Root dry weights of T. repens and T. ambiguum in the 0 to 0.1m root zone were comparable (Figure 1); however, in the subsequent 0.1 m zones, to a depth of 0.5 m, the root dry weight of T. ambiguum was significantly greater than that of T. repens (Figure 1). 0 5 10 Root depth (m) 30 35 40 s.e.d. 4.57 * 0.1 0.2 Root dry weight (g) 15 20 25 s.e.d. 0.41 ** T. repens T. ambiguum 0.3 0.4 0.5 s.e.d. 0.26 *** BC1 BC2 s.e.d. 0.21 *** s.e.d. 0.18 *** Figure 1. Root dry weight in 0.1 m sections of soil columns containing T. repens, T. ambiguum, BC1 and BC2 hybrids. NS, not significant; *P<0.05; ** P<0.01; ***P<0.001 Previous studies have shown that T. ambiguum has a higher proportion of dry matter in roots than T. repens and that these differences may be one factor that improves tolerance of soil moisture deficit. Apart from the 0 to 0.1 m root zone, where the BC2 had the greatest root dry weight, the root dry weight of the BC1 and BC2 hybrids were not significantly different and were generally intermediate between the two parental species. Significant differences in root dry weight were found between genotypes of T. repens and of the BC1 and BC2 hybrids up to a depth of 0.4 m but not between genotypes of T. ambiguum. This suggests there is extensive variation in root density between genotypes of T. repens and, Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 351 not surprisingly, this variation is also observed within the BC1 and BC2 hybrids. Further work is clearly needed to explore this within other backcross families and to study whether comparable results are replicated in field conditions and when in mixed swards with companion grasses. Recent studies have shown that grass species differ in root distribution (Macleod et al., 2013), and that such differences can contribute to improved soil porosity with improved water infiltration, delivering valuable ecosystem services particularly in alleviating the impact of flooding. This present study suggests there may be potential to exploit differences in root architecture within these clover hybrids to develop white clover varieties that will also deliver ecosystem services. Experiments to validate the impact of these new varieties in mixed swards will be the focus of future studies. Conclusions The parental species differed in root distribution and in root weight distribution, with root weight of T. ambiguum significantly greater than T. repens in the 0.1 m to 0.5 m root zone with the backcross hybrids exhibiting root characteristics intermediate between the parental species. Difference in root distribution is a factor influencing the extent to which plants are able to tolerate moisture stress but it may also have beneficial effects on soil porosity delivering ecosystem services. Acknowledgements This research was funded by the UK Department for the Environment, Food and Rural Affairs through the Sustainable Livestock LINK programme. References Macleod C.J.A., Humphreys M.W., Whalley W.R., Turner L., Binley A., Watts C.W., Skøt L., Joynes A., Hawkins S., King I.P., O’Donovan S. and Haygarth P.M. (2013) A novel grass hybrid to reduce flood generation in temperate regions. Scientific Reports| 3:1683 | DOI: 10.1038/srep01683 Marshall A.H., Rascle C., Abberton M.T., Michaelson-Yeates T.P.T. and Rhodes I. (2001) Introgression as a route to improved drought tolerance in white clover (Trifolium repens L.). Journal of Agronomy and Crop Science 187, 11-18. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 352 Interactive N supply and cutting intensity effect on canopy height at 95% light interception Pontes L. da S.1, Baldissera T.C.1,2, Barro R.S.3, Giostri A.F.1,2, Stafin G.1, Santos B.R.C.1, Porfírio-da-Silva V.4 and Carvalho P.C. de F.3 1 IAPAR - Agronomic Institute of Paraná, Ponta Grossa-PR, Brazil, 2 UFPR - Federal University of Paraná, Curitiba-PR, Brazil, 3 UFRGS - Federal University of Rio Grande do Sul, Porto Alegre-RS, Brazil, 4 Embrapa Florestas - Colombo-PR, Brazil. Corresponding author: laisepontes@iapar.br Abstract The aim of this study was to evaluate the impact of pasture management practices on leaf canopy height (LCH) at 95% light interception (LI), since this is a valuable strategy of defoliation frequency. These relationships were investigated over two years with monocultures grown in a fully factorial block design crossing six C4 pasture grass species, two cutting intensities and two N levels. We found variations on LCH at 95% LI, mainly across seasons, but also between treatments. The range of these variations was species-dependent. Therefore, in order to maintain a target IL level, grassland managers should cut or graze at different heights according to the species and seasons. N fertilization and an increase in cutting height above ground level can provide a sward structure that allows lower pre-cutting/grazing height and, consequently, shorter intervals between defoliations. Keywords: C4 grasses, defoliation frequency, response to management Introduction The limitations of adopting standard pre-defined rest periods to control rotational stocking strategies are well known, and highlight the importance of using a plant growth-based criterion to define intervals between successive grazings. This leads to the creation of new methods to manage the resources in regard of these aims than in terms of fixed rest periods, as is still prescribed in recommendations based on classic grazing systems. For instance, recent studies have adopted the criterion based on sward light interception, since it is effective for dealing with the variability of herbage accumulation throughout the year (e.g. with Megathirsus maximum cv. Mombaça, Da Silva et al., 2009), particularly with C4 grass pastures. As a result, ceiling yield with a better control of stem and dead material accumulation has been reached when 95% of light is intercepted by the canopy. Further, correlations between light interception (LI) and sward height have been performed in order to obtain an easy pasture management practice. A consistency seems to emerge of leaf canopy height (LCH) at 95% LI throughout the year, regardless of swards being vegetative or reproductive. However, variations on LCH were related with Tifton 85 (Silva et al., 2013) due post-grazing residues used. Therefore, in order to obtain a better understanding of these relationships, an experiment was set up with a broader range of C4 grass pastures, since these canopies displayed a great vertical heterogeneity, particularly when nutrients are limited. This paper compares the impact of management practices (cutting intensity and N supply) on the LCH of six C4 grass species at 95% LI. In order to do this, the species studied were grown in monocultures, and a simulated rotational defoliation by mechanical cutting was adopted in order to control more easily the extent of the plant parts removed as well as the height at the targeted IL. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 353 Materials and methods The study was based on six perennial C4 grasses (Axonopus catharinensis (Ac), Cynodon spp. hybrid Tifton 85 (Cs), Hemarthria altissima cv. Florida (Ha), Megathirsus maximus cv. Aruana (Mm), Paspalum notatum cv. Pensacola (Pn) and Urochloa brizantha cv. Marandu (Ub)) that are widely used in Brazilian livestock. The experimental site was located at the Agronomic Institute of Paraná, Ponta Grossa-PR (25° 07' 22" S 50° 03' 01" W). The species were planted in pure stands in 2010 (4.5 m²). The cutting intensity and N fertilizer treatments (zero vs. 300 kg N ha-1year-1) were started in January 2011. Plots were cut when the LI of swards were 95%. At this moment, the sward height was measured (10 measures per plot, only in leaves and using a sward stick), and the residual kept was 30% (C30, i.e. high cutting intensity) and 50% (C50, i.e. lower cutting intensity) of this LCH. LI was monitored regularly with a ceptometer (Decagon LP-80 AccuPAR) placed at ground level and above the grass canopy. Data collected over two consecutive years (from spring to fall 2011-2012 and 2012-2013) were evaluated. An analysis of variance was performed on data of height at 95% of canopy LI using R software (R Development Core Team, 2013) to test statistical significance of main factors: year, block, season (nested in each year), species, cutting intensity, N supply and their interactions (except with block). This analysis was performed using GLM model, assuming block to be a random effect and the others as the fixed effects. Prior to ANOVA, height was transformed by log to normalize the data. Results and discussion Outputs of the ANOVA for height at 95% of canopy LI are shown in Table 1. Table 1. Percentage of variance explained (VE) and statistical significance from the analysis of variance for height at 95% of canopy light interception. Df, degrees of freedom. Df VE P Year 1 4.2 *** Species 5 60 *** Block 2 ns Cutting intensity (CI) 1 1.5 *** N supply 1 <1 ** Season(Year) 2 7.2 *** Season(Year)×species 10 3.1 *** Species×N 5 <1 * Species×CI 5 <1 ** Season(Year)×N 5 <1 * Season(Year)×CI 5 <1 * Season(Year)×Year 2 1.2 *** Season(Year)×Year×Species 15 1.5 *** Season(Year)×N×Species 9 <1 ** *, P < 0.05; **, P < 0.01; ***, P < 0.001; ns, not significant. Only significant interactions for at least one variable are presented. Species was the large source of variation, explaining 60% of total variance, with M. maximus having the greatest height, while U. brizantha displayed the lowest value for this variable (Table 2). Since these swards have a wide range of plant morphology and structure, varying from prostrate (e.g. Ac) to tall tussock, erect growing plants (e.g. Mm), a specific height at this target IL level (i.e. 95%) can be expected. Year (37±0.78 and 42±0.86 cm, on average, for the first and second year, respectively) and mainly season had also important effect on the variable studied (Table 1, and see Table 2 for means). In addition, season × species interaction was the most important in terms of variance explained (Table 1). Hence, means per species within each season are shown in Table 2. A considerable variability was observed over growing seasons, except for U. brizantha and A. catharinensis (Table 2). Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 354 Table 2. Seasonal means for leaf canopy height (cm) at 95% of light interception for each species. Data are means (± s.e.) of two labels of nitrogen supply, two cutting intensities, three blocks and two years. Data show the season × species interaction. Species Spring Summer Fall Means Axonopus catharinensis 35 ± 1.34 ab 38 ± 0.97 a 33 ± 1.95 b 36 ± 0.71 CD Urochloa brizantha 27 ± 0.76 a 23 ± 0.64 b 22 ± 0.91 b 24 ± 0.47 E Hemarthria altissima 42 ± 1.90 b 53 ± 1.69 a 40 ± 1.82 b 46 ± 1.20 B Megathirsus maximus 62 ± 2.02 a 58 ± 1.60 b 45 ± 2.37 c 56 ± 1.26 A Paspalum notatum 34 ± 2.36 b 40 ± 0.82 a 32 ± 1.95 b 37 ± 0.87 C Cynodon spp. 35 ± 1.80 b 39 ± 0.81 a 28 ± 1.73 c 35 ± 0.84 D Means 41 ± 1.26 A 42 ±0.84 A 33 ± 0.92 B Values followed by the same letters within a line (uppercase letters for means) do not differ significantly. The majority of species had a higher height in summer with lower values in fall (Table 2). A reason for the increase in height in some seasons would probably have been caused by stem formation due plant maturity developmental stage. Therefore, in order to maintain 95% as a target IL level, grassland managers should cut or graze at different height for reproductive or vegetative growth according to the species. This generates variation in the management of the system between seasons of the year, but need careful control if sward state that optimizes herbage production is to be achieved effectively. There were species × N and species × cut intensity interactions for height at cutting date (Table 1). Fertilizer N significantly decreased the height at the 95% LI for M. maximus (- 7 cm) and A. catharinensis (-3 cm). An increase in cutting intensity (or decrease in residual height) was associated to an increase in the height at cutting date, mainly for M. maximus (+ 8 cm), U. brizantha (+ 4 cm), H. altissima (+ 5 cm) and Tifton 85 (+ 3 cm). Only P. notatum was not affected by both treatments. Our findings reveal that N fertilization and an increase in cutting height above ground levels help to reduce the leaf canopy height pre-cutting. This result can be attributed to a probably maintenance of the sward structure leafy and with a higher tiller density (Silva et al., 2013), which allows the canopy to intercept 95% of the incident light faster, i.e. with a lower height. In summary, variations for LCH at 95% LI were observed between treatments and seasons, with a large range for this last factor. For instance, while variations on LCH across seasons can reach 11 cm for Tifton 85 (see Table 2), changes due the treatments were around 3 cm (C30 = 37 ± 1.36 cm vs. C50 = 34 ± 1.04 cm). For this reason, a narrow pre-cutting or grazing height range (e.g. 28 ± 3.0 cm during the fall for Tifton 85) could be suggested for each species and according to its phenological stage or growing period, regardless management practices. However, the level of herbage depletion and N fertilization are also powerful tools for managing and controlling sward structure in order to optimize herbage intake by herbivores. Conclusions The results of the present study indicate that variations on LCH at 95% LI occurs between treatments, but mainly across seasons. Furthermore, the range of these variations was speciesdependent. Future studies could focus on the effect of these sward surface heights on ruminants’ performance. References Da Silva S.C. da, Bueno A.A. de O., Carnevalli R.A., Uebele M.C., Hodgson J., Matthew C., Arnold G.C. and Morais J.P.G. de (2009) Sward structural characteristics and herbage accumulation of Panicum maximum cv. Mombaça subjected to rotational stocking managements. Scientia Agricola 66, 1, 8-19. Silva W.L., Galzerano L., Reis R.A. and Ruggieri A.C. (2013) Structural characteristics and forage mass of Tifton 85 pastures managed under three post-grazing residual leaf areas. Revista Brasileira de Zootecnia 42, 238-245. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 355 Interactive N supply and cutting intensity effect on leaf nutritive value of C4 grasses Pontes L. da S.1, Giostri A.1,2, Barro R.S.3, Baldissera T.C.1,2, Carpinelli S.1, Guera K.C.S.1 and Carvalho P.C. de F.3 1 IAPAR - Agronomic Institute of Paraná, Ponta Grossa-PR, Brazil, 2 UFPR - Federal University of Paraná, Curitiba-PR, Brazil, 3 UFRGS - Federal University of Rio Grande do Sul, Porto Alegre-RS, Brazil. Corresponding author: laisepontes@iapar.br Abstract Monocultures of six warm-season perennial grasses were compared in a factorial field design of two N levels (zero and 300 kg N ha-1year-1) and two cutting intensities (i.e., the residual kept was 30% and 50% of the canopy height at 95% light interception). The main components of plant nutritive value were measured in two successive growing periods. The species studied exhibited a significant range on leaf lamina quality and leaf proportion. Overall, changes on these parameters were greatest for N fertilizer than cutting-intensity treatments. Keywords: light interception, nutritive value, response to management Introduction For forage crops, it is well established that the nutritive value is strongly affected by management factors through changes on the three components of plant quality, i.e. leaf and stem quality and leaf proportion (Duru et al., 2008). For instance, with tropical/subtropical grass species, a higher nutritive value has been achieved when 95% of canopy light interception (LI) is used as a cutting frequency in rotational stocking (Da Silva and Carvalho, 2005). This response is largely a consequence of a higher leaf:stem ratio, since leaf quality is higher than those of stems. However, differences in quality among species may be attributed to differences in morphological features of their leaves (Pontes et al., 2007), since these values can reflect the strategy of plant species whereby adaptation to variations in land use are achieved. Therefore, the leaf lamina itself plays an important role on plant species quality according to growing conditions, particularly in situations where swards are leafy. The aim of this paper was to examine the effect of cutting intensity and N fertilizer application on leaf nutritive value and leaf proportion of six perennial C4 grasses, growing in monoculture. The 95% LI criterion was adopted as a cutting frequency. Knowing how these nutritional characteristics vary with management may aid decision-making in order to have more effective grazing programmes. Materials and methods The study was based on six perennial C4 grasses (Axonopus catharinensis (Ac), Cynodon spp. hybrid Tifton 85 (Cs), Hemarthria altissima cv. Florida (Ha), Megathirsus maximus cv. Aruana (Mm), Paspalum notatum cv. Pensacola (Pn) and Urochloa brizantha cv. Marandu (Ub)) that are widely used in Brazilian livestock. The experimental site was located at the Agronomic Institute of Paraná, Ponta Grossa-PR (25° 07' 22" S 50° 03' 01" W). The species were planted in pure stands in 2010 (4.5 m²). The cutting intensity and N fertilizer treatments were started in January 2011. Plots were cut when the LI of swards were 95%. At this moment, the sward surface height was measured, and the residual kept was 30% (C30, i.e. high cutting intensity) and 50% (C50, i.e. lower cutting intensity) of this original canopy height. We simulated rotational defoliation by mechanical cutting in order to easily compare several species and control the extent of the plant parts removed. Further, two contrasting N levels (zero and 300 kg N ha-1year-1, N0 and N300, respectively) were compared. At the cutting date, laminas (youngest fully expanded lamina) were selected at random from vegetative tillers and cut off, Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 356 oven dried at 60 °C for 48h, and milled through a 1 mm screen. Our focus is concentrated at the leaf lamina level in order to have a more accurate prediction about the effect of the management practices on species quality, i.e. with no effect of plant components proportion. One sample per plot was collected over two consecutive growing periods (from spring to autumn 2011-2012 and 2012-2013). Each sample was analysed via near-infrared reflectance spectroscopy (NIRS) procedure (FOSS-NIRSystems 5000; CEPA laboratory, Passo FundoRS, Brazil) for crude protein (CP), dry-matter digestibility (DMD), neutral detergent fibre (NDF) and acid detergent fibre (ADF). Leaf proportion was measured from representative samples harvested above cutting height. Analyses of variance were performed using Statgraphics Centurion XV to test statistical significance of main factors: year, block, species, cutting intensity, N supply and their interactions (except with block). All response variables were analysed using GLM model, assuming year and block to be a random effect and the others as the fixed effects. Results and discussion Outputs of the ANOVA for nutritive value parameters are shown in Table 1. Table 1. Percentage of variance explained and statistical significance from the analysis of variance for crude protein (CP), neutral detergent fibre (NDF), acid detergent fibre (ADF), dry-matter digestibility (DMD) and leaf proportion (LP). df CP NDF ADF DMD LP Growing period (GP) 1 ns 9.3* ns ns ns Block 2 ns ns ns ns ns N 1 39*** 16*** ns ns 8.6*** Cutting intensity 1 1.6** ns ns ns 1.5** Species 5 27* 38* 50** 50*** 42*** GP×Species 5 5.2*** 6.8*** 3.2* 3.2* GP×N 1 1.5** 1.5** *, P < 0.05; **, P < 0.01; ***, P < 0.001; df, degrees of freedom; ns, not significant. Only significant interactions for at least one variable are presented. Species was a large source of variation for NDF, ADF, DMD and leaf proportion, and explained between 38 and 50% of total variance (Table 1). Means per species are shown in Table 2. Table 2. Percentage (±standard error) on leaves of crude protein (CP), neutral detergent fibre (NDF), acid detergent fibre (ADF) and dry-matter digestibility (DMD) for six perennial C4 grasses. LP, leaf proportion (%). Species CP NDF ADF Axonopus catharinensis 13±0.54c 67±0.53b 34±0.68b Cynodon spp. 16±0.55ab 74±0.80a 35±1.06b Hemarthria altissima 15±0.59abc 71±0.90ab 35±0.79b Megathirsus maximus 17±0.66a 70±0.64ab 35±1.13b Paspalum notatum 13±0.46bc 74±0.54a 40±0.61a Urochloa brizantha 16±0.38a 68±0.81b 27±0.57c Values followed by the same letters within a column do not differ significantly. DMD 62±0.53b 61±0.82b 62±0.62b 61±0.88b 58±0.47c 68±0.44a LP 60±2.35b 59±1.75bc 36±2.14d 58±2.35bc 54±3.06bc 81±2.08a The highest values of leaf proportion and DMD were achieved for U. brizantha (Table 2). This species had also the lowest ADF (27%). At the opposite, P. notatum had the lowest DMD (58%) and the highest ADF (40%) values in leaves, and H. altissima displayed the lowest leaf proportion (36%). The CP percentage of all species ranged from 13 (Ac) to 17% (Mm), which is above the minimum CP level required for rumen function (7.5%, Van Soest, 1994). The NDF percentages recorded in this experiment ranged from 67 (Ac) to 74% (Cs and Pn), which were similar to those reported by Soares et al. (2009) for Ac, Cs, Ha, Mm and Ub, and where the NDF average in leaves was 70%. For CP, N supply effect was more important than species effects, accounting for 39% of total variance. This result confirms a previous report that the CP concentration of herbage varies mainly with the nitrogen status on the plant (Lemaire and Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 357 Gastal, 1997). For all species, N supply significantly increased the CP (+3.5%) and leaf proportion (+11%), and decreased the NDF (-3.2%) in leaves. The higher leaf proportion at N300 could be a consequence of a probably higher tiller density and an increase in leaf size. The CP percentage was significantly higher in C50 (15.5±2.89%) than in C30 (14.8±2.92%), probably due a higher opportunity in increasing the N resorption efficiency. Further, a higher leaf proportion was also observed in C50 (62±1.58%) when compared to C30 (57±1.73%), because increasing the level of depletion of the pre-cutting canopy height leads to a marked increase in stem and dead material (data not shown). The NDF was significantly higher in the first growing period (72±4.32%) than in the second (70±3.94%). In relationship to the interactions, the year×species was the most important in terms of variance explained. These results highlighted a sampling date effect despite lower differences on leaf quality over time than stems (Duru et al., 2008). However, despite significant interactions between growing periods and species for all variables (Table 1), the quality rankings of species were consistent [coefficients of Spearman rank correlations of 0.65, 0.49, 0.59 and 0.59 for CP, NDF, ADF and DMD respectively (significant at P < 0.001)] across years. Therefore, significant year×species interactions seem to result from differences in order of magnitude and not from differences in species ranking. Conclusion Species choice, lenient cutting intensity and nitrogen fertilizer application are strategies that increase forage quality with a potential positive impact on ruminant performance. Acknowledgements The present work has been financially supported by CNPq (Repensa). We thank G. Stafin for his technical collaboration. References Duru M., Cruz P., Al Haj Khaled R., Ducourtieux C. and Theau J.P. (2008) Relevance of plant functional types based on leaf dry matter content for assessing digestibility of native grass species and species-rich grassland communities in spring. Agronomy Journal 100, 1622-1630. Lemaire G. and Gastal F. (1997) N uptake and distribution in plant canopies. P.3-44. In: Lemaire G. (ed) Diagnosis of the nitrogen status in crops. Springer Verlag, Berlin. Pontes L. da S., Soussana J.F., Louault F., Andueza D. and Carrère P. (2007) Leaf traits affect the above-ground productivity and quality of pasture grasses. Functional Ecology 21, 844-853. Silva S.C. da and Carvalho P.C.F. (2005) Foraging behaviour and herbage intake in the favourable tropic/subtropics. In: McGilloway D.A. (ed), Proceedings of XX International Grassland Congress, Dublin, pp. 81-95. Wageningen NL: Wageningen Academic Publishers. Soares A.B., Sartor L.R., Adami P.F., Varella A.C. and Mezzalira J.C. (2009) Influence of luminosity on the behavior of eleven perennial summer forage species. Brazilian Journal of Animal Science 38, 443-451. Van Soest P.J. (1994) Nutritional ecology of the ruminants, 2nd edn. Ithaca, NY, USA: Comstock Publishing Associates/Cornell University Press. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 358 Forage selection and animal performance of grazing heifers on semi-natural fen grassland Müller J. and Sweers W. Working Group Grassland and Forage Science, Faculty of Agricultural and Environmental Sciences – University of Rostock, Justus-von-Liebig-Weg 6, 18059 Rostock, Germany Corresponding author: juergen.mueller3@uni-rostock.de Abstract Grazing is a suitable management to maintain the biodiversity and habitat functions of restored fen grasslands. Nutritional requirements of the grazing livestock need to be considered and selective grazing might occur. In order to analyse the selective grazing and actual energy intake by grazing cattle, we implemented a pasture experiment on a percolation mire site in the northeast of Germany. On a 29-ha pasture, 23 heifers had free access to either mineral or fen areas. Standing crop and faeces of cattle were sampled biweekly from May to September at 24 points along a transect. Plant samples were analysed by NIRS. The nitrogen content of the faeces was used to estimate the digestibility of the forage selected by the grazing cattle. Sedgedominated areas of the fen section were not visited for grazing before the grass-dominated sections of the moraine parts were depleted. This behaviour limited the degree of energy selection and thus the daily liveweight gain of the heifers. These findings show the limits of energy selection on large-scale semi-natural fen pastures and provide a reason to modify the continuous stocking system. Keywords: animal performance, semi-natural grasslands, forage selection, forage quality Introduction Restored fen grasslands in many parts of Europe contribute to biodiversity (Koch and Jurasinski, 2014), supply habitat functions and have potential to store carbon (Dierssen and Nelle, 2006). Grazing has been described as a suitable management tool to maintain these important ecosystem services in general (Middleton et al., 2006). However, at the regional level, appropriate grazing practices have to be developed to meet the challenges of maintaining the natural resources while establishing socio-economic sustainable land-use systems in detail (Ostermann et al., 1998). In northeast Germany, large-scale and continuous grazing with cattle was regarded as a key tool to fulfil both requirements. Economic success of this low-cost strategy depends to a great extent on the animal performance per head since stocking rate is limited by restrictions. On semi-natural fen grassland the animal performance, a function of amount and quality of forage intake, is to a great degree determined by the botanical composition and the phenological stage of the standing crop (Bockholt and Buske, 1997). Under continuous stocking, the ability of the single animal to graze selectively is the most important adaption mechanism to meet the nutritional demands (Strodthoff, 2002). This is especially important in periods of decreasing nutritive value of the sward. This study aimed at analysing the degree of selective grazing and actual energy intake and related animal performance of grazing cattle during the grazing period. Results contribute to a better understanding of the processes underlying animal performance on continuously grazed semi-natural fen grasslands. Materials and method The grazing experiment was located on a river valley segment consisting of a percolation mire with an adjacent moraine slope, in northeast Germany. From 2011 to 2013, a 34-ha pasture was continuously stocked with 23-27 heifers. Stocking densities were 0.84, 1.03 and 1.08 livestock units (LU) per ha for the grazing seasons 2011, 2012, and 2013, respectively. Heifers had free Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 359 access at all times to the moraine slope or fen areas. However, in midsummer 2011 a flooding event prevented access to the peat parts for the remaining time of the grazing season. All animals were equipped with digital earmarks allowing for their identification at an automatic weighing system. Data were used to calculate the daily liveweight gain. The standing pasture crop was sampled biweekly from May to September at 24 points along a transect which included peat and mineral sections. Oven-dried forage samples (65 °C) were ground to 1mm and analysed by NIRS (MPA Bruker®) for nutritive value parameters including ELOS (enzyme soluble organic matter). In accordance with the reference standard methods, crude ash and a sample subset selected using spectral information were additionally analysed according to fit the NIRS calibrations to the experimental data set. Digestibility of the offered pasture growth was calculated in compliance with Weißbach et al. (1999). Additionally, we collected fresh cattle faeces biweekly and analysed the nitrogen content. This allowed us to estimate the digestibility of the selected forage according to Schmidt et al. (1999). The selectivity was graphically presented as the distance between the digestibilities of offered and realized forage using statistical and graphic tools of SigmaPlot 11.0 ®. Results and discussion Trends of forage digestibility over the three observed grazing seasons are presented in Figure 1. Except for the first grazing season 2011, the well-known general trend of decreasing forage values with grazing duration from spring to autumn became obvious. Figure 1. Digestibility of the organic matter of the standing crop (DOM offer) versus digestibility of the pasture diet (DOMdiet) at each of the three grazing seasons. Mean values are derived from 24 forage samples (DOM offer) and from 6 faecal charges (DOMdiet). Error bars indicate standard deviations of the DOM diet - means. A flooding event in midsummer of 2011 prevented the cattle from grazing in these sections. This explains the paradox of negative selection for energy in the first experimental year. Periods with higher DOMdiet-lines are characterized by successful selection of the offered Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 360 pasture growth for digestibility. Such periods can be observed mainly at the beginning of the late summer and to the end of the grazing season (Figure 1). Compared to the results of Strodthoff (2002), degree of forage selection for high energy content in our experiment was generally smaller than expected. Probably this was due to the dimension of our large-scale pasture and the strong vegetation gradient within. The initially avoided plants at Strodthoff’s small-scale pasture were mainly grasses. At our experimental site the fen sections were dominated by sedges, which provide much lesser potential for energy selection. Sedge-dominated areas were not visited for grazing until the grass-dominated sections had been depleted. Furthermore, long pathways and the anxiety to leave the herd seem to counteract an efficient diet selection. We did not observe a clear relationship between the degree of energy selection and the daily liveweight gain over the complete grazing season (data not shown). However, there are strong indications that the degree of energy selection becomes especially important in late summer and autumn. Conclusion This investigation shows the magnitudes and limits of energy selection on large-scale seminatural fen pastures and it provides a reason for modifying the widely used system of continuous stocking. Forcing the cattle into the sedge-dominated fen areas in late spring by fencing would provide both higher digestibility of the standing crop and a better selection base in the late summer under a free-access situation. References Bockholt R. and Buske F. (1997) Variationsbreite des Futterwertes von Niedermoorgrünland unter Berücksichtigung der häufigsten autochthonen Pflanzen (Variation in forage quality of the most important species of fen grasslands). Das Wirtschaftseigene Futter 43, 5-20. Dierssen K. and Nelle O. (2006) Zustand, Wandel und Entwicklung europäischer Moorlandschaften (Situation, change and development of European mire landscapes). Nova Acta Leopoldina 94 (346), 241-257. Koch M. and Jurasinski G. (2014) Four decades of vegetation development in a percolation mire complex following intensive drainage and abandonment Plant Ecology & Diversity DOI=10.1080/17550874.2013.862752 Middleton B.A., Holsten B. and Van Diggelen R. (2006) Biodiversity management of fens and fen meadows by grazing, cutting and burning. Applied Vegetation Science 9, 307-316. Ostermann O.P. (1998) The need for management of nature conservation sites designated under Natura 2000. Journal of Applied Ecology 35, 968-973. Schmidt L., Weißbach F., Hoppe T. and Kuhla S. (1999) Untersuchungen zur Verwendung der KotstickstoffMethode für die Schätzung des energetischen Futterwertes von Weidegras und zum Nachweis der der selektiven Futteraufnahme auf der Weide. Landbauforschung Völkenrode 3, 123-135. Strodthoff J. (2002) Dynamik von Narbenstruktur und Weideleistung auf extensiviertem Niedermoorgrünland (Dynamics of sward structure and animal performance on extensified low-peat fen grasslands). PhD thesis, University of Göttingen, Cuvillier Verlag, Göttingen, 98 pp. Weißbach F., Kuhla S., Schmidt L. and Henkels A. (1999) Schätzung der Verdaulichkeit und der Umsetzbaren Energie von Gras und Grasprodukten (Estimation of digestibility and metabolic energy of fresh grass and grass conservates). Proceedings of the Society of Nutrition Physiology 8, 72. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 361 Breed type differences in hoof volume in beef suckler cows Fraser M.D. and Vale J.E. Institute of Biological, Environmental and Rural Sciences, Aberystwyth University, Gogerddan, Aberystwyth, United Kingdom, SY23 3EE Corresponding author: mdf@aber.ac.uk Abstract Trampling by cattle can be an effective way of creating establishment sites for plant species of conservation interest, but can also lead to unwanted sward damage. Hoof size and shape could potentially differ between traditional and modern breeds of cattle, in turn influencing their relative impact upon soil and/or vegetation. In this study linear measurements were used to estimate hoof volume for two contrasting breeds of suckler cow: Limousin-cross cows (modern, high performance breed type) and Belted Galloway cows (native breed type). Data were collected for 20 Limousin-cross cows and 10 Belted Galloway cows. The hoof volumes of the fore and hind feet of the Belted Galloway cows were less than the equivalent values for the Limousin-cross cows. However, the smaller body size of the Belted Galloways meant the calculated load on the hooves was similar for the two breeds (0.371 vs 0.399 kg live weight cm-3 for the Limousin-cross and Belted Galloway cows respectively; s.e.d. 0.0180 kg live weight cm-3; ns). The results suggest that smaller breeds with smaller feet may have less impact on vegetation and soil types sensitive to damage and erosion, but that larger breeds with larger hoof volumes may be more appropriate where disturbance is desirable. Keywords: cattle, disturbance, conservation grazing Introduction Grassland communities are increasingly recognized as being disturbance-dependent. Trampling, particularly by cattle, can be an effective way of creating establishment sites for plant species of conservation interest within semi-natural vegetation communities (Mitchell et al., 2008). However, poaching (i.e. damage caused to the sward exposing excessive bare ground) contravenes the 'Standards of Good Agricultural and Environmental Condition' and can lead to substantial environmental losses, particularly on carbon-rich soils. Managing the impact of cattle hooves on grassland is therefore both an agricultural and environmental issue. Hoof size and shape could potentially differ between different breed types of cattle, in turn influencing their relative impact upon soil and/or vegetation, and thus their suitability as conservation grazers. In this study correlations between linear measurements of the hoof and hoof volume (Scott et al., 1999), together with body mass, were used to explore the potential impact of different cattle breeds. Materials and methods The surface area of the sole of cattle feet rarely represents the surface on which weight is borne (Scott et al., 1999), and hence measurements were made which would allow hoof volume to be calculated. These measurements were made within a week of the hooves being inspected and trimmed as required, to avoid results being distorted by over-grown claws. Data were collected from 20 Limousin-cross cows (modern suckler type) and 10 Belted Galloway cows (native suckler type). For each animal, measurements were made on the digits of the left forefoot and the right hind-foot (two digits per foot). The data collected per digit were: a) the distance along the proximal border of the coronary band from abaxial groove to flexure of the dorsal surface (CorBand), b) the distance from the abaxial groove, along the distal border or weight-bearing region of the claw, to the point of the toe (Base), and c) the height of the abaxial Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 362 groove from the proximal border of the coronary band to the base of the claw (AbaxGr). Hoof volume was then calculated using the formula below (Scott et al., 1999): Volume (cm-3) = (17.192 × Base) + (7.467 × AbaxGr) + (45.270 × CorBand) – 798.5 The live weight of the cattle was also recorded. Results and discussion The hoof volume of both the fore and hind feet of the Belted Galloway cows were less than the equivalent feet of the Limousin-cross cows (Table 1). For both breed types, front hooves were larger than hind hooves, with 61% and 58% of the animals’ total volume being found in the front hooves of the Belted Galloway and Limousin-cross cows respectively. These results correspond well to the finding that the forelimbs of cattle bear approximately 60% of the body weight (Fessl, 1968). Table 1. Hoof dimensions and volume of contrasting breed types of suckler cow. Limousin-cross Belted Galloway s.e.d. Prob CorBand (cm) 14.0 14.2 0.42 ns Base (cm) 28.8 25.8 0.96 ** AbaxGr (cm) 26.0 23.0 0.73 *** Volume (cm-3) 521 456 32.7 * CorBand (cm) 12.2 11.9 0.39 ns Base (cm) 25.8 23.4 0.93 * 23.5 19.9 0.78 *** 371 287 27.2 ** Front foot Hind foot AbaxGr (cm) -3 Volume (cm ) The Limousin-cross cows were significantly heavier than the Belted Galloway cows (650 vs 589 kg; s.e.d. 22.0 kg; P<0.05). However, the calculated load on the hooves of each breed was similar for the two breeds, at 0.371 and 0.399 kg live weight cm-3 for the Limousin-cross and Belted Galloway cattle respectively (s.e.d. 0.0180 kg live weight cm-3; ns). Conclusions The smaller hoof volume for the same load suggests that smaller breed types such as the Belted Galloway may have less impact on vegetation and soil types sensitive to damage and erosion. In contrast, larger breed types with larger hoof volumes may be more appropriate where disturbance is desirable. Acknowledgements This work was funded by the UK Department for Environment, Food and Rural Affairs References Fessl L. (1968) Biometric studies on the ground surface of bovine claws and the distribution of weight on the extremities. Zentralblatt fur Veterinarmedizin 15, 844-860. Mitchell R.J., Rose R.J. and Palmer S.C.F. (2008) Restoration of Calluna vulgaris on grass-dominated moorlands: The importance of disturbance, grazing and seeding. Biological Conservation 141, 2100-2111. Scott T.D., Naylor J.M. and Greenough P.R. (1999) A simple formula for predicting claw volume of cattle. The Veterinary Journal 158, 190-195. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 363 Changes in Koniks' diet due to vegetative season, years and social behaviour Chodkiewicz A. and Stypiński P. Department of Agronomy, Faculty of Agriculture and Biology, Warsaw University of Life Sciences - SGGW, Warsaw, Poland Corresponding author: anna.chodkiewicz82@gmail.com Abstract Extensive grazing of horses is conducted in many regions of the Northern hemisphere in order to maintain plant communities with high nature values as well as enabling the recovery of breeds of wild horses. The aim of these studies was to examine the influence of changes in the vegetative season, and of years and herd behaviour, on the percentage share of chosen groups of plants grazed by horses. The studies were conducted in the Biebrza National Park situated in north-eastern Poland. Direct visual observations of grazing of Polish primitive horses (Koniks) from two herds were made during three months (early April, late June and August) in 2009 and 2010. During observation, the first ten bites of the sward made by a chosen mare every 5 minutes were separated into the bitten species. Koniks ate mainly sedges; only in June grasses dominated in their diet with Carex sp. The share of specific groups of plants being grazed by Koniks is significantly influenced by weather conditions in the year, which changed the availability of species, and by the season itself, as well as by social behaviour of the animals. Keywords: diet of horses, grazing, Koniks, sedges Introduction The extensive grazing of horses is conducted in many regions of the Northern hemisphere in order to maintain plant communities with a high nature values as well as enabling the recovery of breeds of wild horses. These animals could have an ability to graze plants with low utility value (Prache et al., 1998). It is well known that, during the vegetative season, horses show high selectivity towards grazed habitats (Archer, 1973, Crane et al., 1997). Simultaneously, although horses prefer mainly grasses and others monocotyledonous species, their preferences towards plant species changes due to many biotic (e.g. availability, quality and growth stages of particular species), abiotic (e.g. season) and individual (e.g. experience) factors (Bailey and Provenza, 2008). Koniks - Polish primitive horses - are descendents of tarpan. Though Koniks graze on wetlands, there is still not enough knowledge about their grazing preferences in this environment. The aim of these studies was to examine the influence of changes in vegetative season, years and behaviours of herds on the percentage share of chosen groups of plants bitten by horses. Material and methods The studies were conducted in the Biebrza National Park situated in northeastern Poland, where Koniks have been kept since 2004. Polish primitive horses stayed there in two herds - named after the names of the stallions: Mrok (M) and Limanek (L). Direct visual observations of grazing horses were made by an observer during three months (early April, late June and August) in 2009 and 2010. Every herd was observed for three days in each term in two threehours periods: during 07-12 a.m. and 4-8 p.m. During observation, the first ten bites of the sward made by a chosen mare every 5 minutes were assigned to bitten species. A random (different) mare was chosen each day from the herd, with no repetitions. The collected data set was analysed with a mixed model that used REML methods (procedure MIXED in SAS 9.3.). Corrected averages and the least significant differences (LSD) were used for the interpretation of gathered data. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 364 Results and discussion The significant influences on the percentage share of grasses and sedges among plants bitten by horses were both their availability (years) and behaviour of the herds (Figures 1 and 2). In 2010 - a year in which inundation lasted longer than in 2009 - conditions were favourable for species connected with wet habitats, and therefore the Koniks grazed Carex sp. more often than in the first year of observation (Figure 1). Among plants bitten by horses from the M herd, it was found that there was about 15% higher percentage share of grasses, simultaneously with a lower share of sedges (Figure 2). It resulted from the larger territory of this herd and more frequent observation of horses grazing on little mid-forest meadows where Poacae sp. are dominant in the sward. The L herd stayed mainly in communities situated near Woźnawiejski Canal, where sedges dominated. Figure 1. Percentage share of groups of plants in Koniks' bites during vegetative season (two-year average; I = confidence interval, P ≤ 0.05) Figure 2. Percentage share of groups of plants in Koniks' bites (in years 2009-2010; I = confidence interval, P ≤ 0.05). Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 365 share among bitten plants [%] Limanek Mrok Figure 3. Percentage share of groups of plants in Koniks' bites split by herd ( I = confidence interval, P ≤ 0.05). Previous research has showed that Koniks could reduce undesirable species successfully, e.g. some sedges (Musielak and Rogalski, 2006; Chodkiewicz and Stypiński, 2011). It should be assumed that, as suggested by Vulink (2001), with better availability grasses would contribute a greater part in the diet of the Koniks. Conclusion The share of specific groups of plants being grazed by Koniks is significantly influenced by weather conditions in the year as this changes the availability of species, and by the season itself, as well as by social behavior of the animals. Sedges are dominant in the diet of the Koniks' at the beginning and at the end of the vegetative season, and this may result from delayed growth of grasses. References Archer M. (1973) The species preferences of grazing horses. Journal of the British Grassland Society 28, 123128. Bailey D.W. and Provenza F.D. (2008) Chapter 2a Mechanism determining large-herbivore distribution.In: H.H.T. Prins and F. van Langelveld (eds). Resource Ecology: spatial and temporal dynamics of foraging. Springer, pp. 7-28. Chodkiewicz A. and Stypiński P. (2011) Diet preferences of Koniks horses in disadvantaged areas: a case study from Biebrza National Park. Grasslands Science in Europe 16, 326-328. Crane K.K., Smith M.A. and Reynolds D. (1997) Habitat selection patterns of feral horses in southcentral Wyoming. Journal of Range Management 50, 374-380. Musielak D. and Rogalski M. (2004) The impact of extensive grazing of Polish Koniks on changes in vegetation cover of selected plant communities of coastal meadows. In: H. Czyż (eds) Salt grasslands and coastal meadow. Wydaw. AR, Szczecin, pp. 39–44. Prache S., Gordon I.J. and Rook A.J. (1998) Foraging behaviour and diet selection in domestic herbivores. Annals de Zootechnie 47, 335-345. Vulink J.T. (2001) Hungry herds: Management of temperate lowlands wetlands by grazing. Van Zee tot Land 66, Lelystad, pp. 394. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 366 Long-term stability of sward patch structure under different intensities of cattle grazing Tonn B.1, Wrage-Mönnig N.2 and Isselstein J.1 1 University of Goettingen, Department of Crop Sciences, Institute of Grassland Science, vonSiebold-Str. 8, 37075 Goettingen, Germany 2 Rhine-Waal-University of Applied Sciences, Faculty of Life Sciences, Kleve, Germany Corresponding author: btonn@gwdg.de Abstract In low-intensity grazing systems, selective defoliation by the grazing animals can lead to a mosaic pattern of short, frequently grazed, and tall, rarely grazed patches. If the spatial pattern of these patches is stable over time, this may result in divergent development of botanical composition and nutrient cycling. Using sward height data from permanent quadrats, we tested the hypothesis that, under continuous cattle stocking, patch grazing can create sward structures that are stable not only over the short term, but also over several years. For three grazing treatments with different stocking rates, the transitions between three sward height classes (short, medium and tall) were quantified for three time scales: seasonal, interannual and longterm transitions. With the exception of long-term transitions under the highest stocking rate, transitions between sward height classes were non-random, confirming the initial hypothesis of short- and long-term patch stability. A stable sward height pattern may enhance biodiversity and has methodological consequences for studying extensive pastures. Keywords: patch-grazing, spatial heterogeneity, sward structure, stocking rate Introduction In low-intensity grazing systems, selective defoliation by the grazing animals together with positive feedbacks between defoliation frequency and forage quality can lead to a mosaic pattern of short, frequently grazed patches and tall patches where defoliation intensity and frequency are low (so-called ‘patch-grazing’, Adler et al., 2001). If the spatial pattern of these patches is stable over time, divergent development of botanical composition and nutrient cycling is to be expected. To assess patch stability in a long-term cattle grazing experiment, we employed a classification with temporally variable class boundaries to account for seasonal and annual differences in sward productivity. We assumed the existence of two functionally different patch types, frequently grazed (‘short’) and rarely grazed (‘tall’). As their height distributions, however, overlap at intermediate sward heights (Rossignol et al., 2011), we included an intermediate sward height class to cover this sward height range. Using biannual sward height data from seven successive years, we tested the hypothesis that under continuous stocking with cattle at a range of stocking rates, patch grazing can create sward structures that are stable not only over the short term, but also over several years. Materials and methods An experiment comparing three grazing treatments in a triplicate randomized block design with a paddock size of 1 ha, was established in 2002 at the experimental farm Relliehausen of the University of Goettingen, located in the Solling Uplands 40 km northwest of Goettingen. The paddocks are grazed by heifers under continuous stocking with three grazing intensity treatments defined by target sward heights: moderate grazing (MG), target sward height 6 cm; lenient grazing (LG): target sward height 12 cm, and very lenient grazing (VLG), target sward height 18 cm (prior to 2005: 12 cm). Target sward heights are maintained in a put-and-take system with bi-weekly sward height measurements. Average stocking rates were 670, 370 and Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 367 250 kg ha-1year-1 in the treatments MG, LG and VLG, respectively, in the period 2005-2011. Since 2002, no fertilizers have been applied and swards were not cut. For further details, see Isselstein et al. (2007). From 2007 to 2013, compressed sward heights (CSH) were measured each spring before grazing, and each autumn at the end of the grazing season, at 10 randomly located permanent quadrats of 1 m² size per paddock. Per quadrat, five measurements were made, using a rising plate meter with a circular aluminium plate weighing 200 g and with a diameter of 30 cm. Mean values per quadrat were calculated. At every measurement date, each quadrat was classified into one of three sward height classes: short, medium or tall. For each date, the class boundaries were based on the 33rd and 67th percentiles of the CSH at the 90 quadrats, with short: CSH ≤ 33rd percentile, medium: 33rd percentile < CSH < 67th percentile and tall: CSH  67th percentile. To quantify the long-term stability of patches, the frequencies of transitions between sward height classes were calculated for all quadrats per grazing treatment. Three time scales were considered: seasonal transitions (spring to autumn within each year), interannual transitions (autumn to next autumn) and long-term transitions (autumn 2007 to autumn 2013). The null hypothesis that sward height class at any time is independent of sward height class at a defined previous time was tested separately for each grazing intensity for seasonal, yearly and long-term transitions, using Fisher’s exact test as implemented in the Software R (R Development Core Team, 2008). Results and discussion Over all sampling dates, the grazing treatment with the highest stocking rate, MG, had the highest proportion of sampling quadrats classified as ‘short’ (52% compared to 29% in LG and 22% in VLG) and the lowest proportion of quadrats classified as ‘tall (15% compared to 40% in LG and 46% in VLG). The proportion of quadrats classified as ‘medium’ was similar in all grazing treatments (33, 31 and 31% in treatments MG, LG and VLG, respectively). The class boundary between short and medium patches varied between a CSH of 4.2 and 8.5 cm, with a median of 5.8 cm. For the class boundary between medium and tall patches, the median was at a CSH of 9.4 cm, with a range of 7.3-12.3 cm. Both seasonal and interannual transitions between sward height classes were dependent on initial sward height class (Figure 1). The proportion of short patches becoming tall patches, and, with the exception of grazing treatment MG, the proportion of tall patches becoming short patches, did not exceed 10%. The long-term transitions between sward height classes from autumn 2007 to autumn 2013 only depended on initial sward height in the more extensive grazing treatments LG and VLG. Even at this time scale, transition frequency from short to tall patches and vice versa did not exceed 20% in any of the treatments. Within-year and interannual stability of sward patch structure under continuous cattle grazing have been reported previously (Rossignol et al., 2011; Tonn et al., 2013). Our results indicate that even over a period of seven years, patches can be similarly stable if stocking rates are low, thus confirming the initial hypothesis. Conclusion The existence of temporally stable sward height patterns in continually grazed cattle pastures can lead to within-paddock differentiation of vegetation and nutrient cycling between frequently and rarely defoliated patches. Such a functionally heterogeneous mosaic structure has the potential of increasing botanic and faunistic diversity. It should also be explicitly recognized in sampling design of nutrient cycling and biodiversity studies in these systems. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 368 Figure 1: Frequencies with which permanent quadrats belonging to different sward height classes (short, medium or tall) transitioned to each of these sward height classes, compared to the frequencies expected if transitions between sward height classes are random (dashed outline). Asterisks in the upper right corner of each panel mark significant differences between observed and expected frequencies (***: P < 0.001, *: P < 0.05). Time scale considered: (a) from spring to autumn within each year, (b) from one autumn to following one, (c) from autumn 2007 to autumn 2013. MG, LG, VLG: stocking rates (moderate, lenient and very lenient grazing, respectively). References Adler P.B., Raff D.A. and Lauenroth W.K. (2001) The effect of grazing on the spatial heterogeneity of vegetation. Oecologia 128, 465-479. Isselstein J., Griffith P.A., Pradel P. and Venerus S. (2007) Effects of livestock breed and grazing intensity on biodiversity and production in grazing systems. 1. Nutritive value of herbage and livestock performance. Grass and Forage Science 62, 145-158. R Development Core Team (2008) R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. Rossignol N., Chadoeuf J., Carrère P. and Dumont B. (2011) A hierarchical model for analysing the stability of vegetation patterns created by grazing in temperate pastures. Applied Vegetation Science 14, 189-199. Tonn B., Wirsig A., Kayser M., Wrage-Mönnig N. and Isselstein J. (2013) Patch-differentiation of vegetation and nutrient cycling in an extensive pasture system. Proceedings of the 22nd International Grassland Congress, 1619 September 2013, Sydney, pp. 921-924. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 369 Impact of long-term extensive use of permanent grasslands on their provisioning service Kizeková M.1, Kanianska R.2, Makovníková J.3, Beňová D.1, Čunderlík J.1 and Jančová Ľ.1 1 NAFC - Grassland and Mountain Agriculture Research Institute, Banská Bystrica, Slovakia 2 Matej Bel University, Faculty of Natural Sciences 3 NAFC – Soil Science and Conservation Research Institute Corresponding author: kizekova@vutphp.sk Abstract In last two decades, changes in grassland management have taken place in Slovakia. The application of agri-environmental measures of the Common Agricultural Policy, which is a main driver of grassland extensification, has resulted in restoration of biodiversity of permanent grasslands and has brought environmental benefits to society as well. Nevertheless, unilateral extensive use of permanent grasslands has begun to have a negative impact on some ecosystem services such as the provision of high-quality forage for cattle. This has consequently resulted in a shortage of beef in human nutrition in terms of Recommended Allowances of Foodstuffs. Keywords: ecosystem services, permanent grasslands, extensive management Introduction Apart from the ecosystem services that permanent grasslands provide for people and the environment, one the main ecosystem services should be the provision of quality herbage for herbivores and a consequent positive effect on human nutrition. The production of surpluses of animal products and environmental damage caused by overgrazing has led to reassessment of the functions and services of permanent grasslands. Although grasslands have always had an important role in the Common Agricultural Policy of the European Union, attention has been drawn to the several ecosystem services, such as grassland habitat maintenance and biodiversity conservation, erosion control and maintenance of soil fertility. Provisioning service of permanent grasslands, in terms of providing high quality herbage, has been suppressed to the background. Nevertheless, extensive use of permanent grasslands supported by agri-environmental payments has resulted in grasslands not being able to provide a full range of ecosystem services, including food security and human nutrition. The objective of this work is to analyse the current state in the use of permanent grasslands and show the possible negative impact of extensive grassland management on their ecosystem services. Materials and methods Slovakia is landlocked country in Central Europe with a total area of 49,036 km 2. Slovakia is divided into eight regions, each of which is named after its regional capital. Three regions (Bratislava, Trnava and Nitra) are located in the south-western part of the country, which is under favourable climatic conditions with the highest proportion of productive land dominated by arable cultivation. The other five regions are characterized by the presence of a high percentage of low-quality agricultural land, which translates into a high share of LessFavourable Areas (LFAs). Currently, the use of permanent grasslands takes place mainly in upland and mountain regions, which with more then 500,000 hectares comprise about 40% of the entire LFAs. The majority of permanent grasslands are found in three regions (Banská Bystrica, Žilina and Prešov). In order to assess current development in provisioning services of permanent grassland, in terms of amount and quality of forage for herbivores, the following indicators were used: grassland area, dry matter yield (DMY), number of livestock units and stocking rate. The basis for the evaluation was data of the Statistical Office of the Slovak Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 370 Republic available on-line from RegDat (http://px-web.statistics.sk/PXWebSlovak). The analysis was performed for the period from 2009 to 2011. Results and discussion As in other European countries, in Slovakia the number of herbivores, especially cattle, has been decreasing since the 1990s. As in 1990, there were more than 2 million head of cattle in Slovakia; at present the number has decreased to around 463 000 and has been steadily decreasing. A significant reduction of numbers of herbivores has negatively affected the use of grasslands in terms of the area of permanent grassland that is used area, and also the herbage amount and quality. Out of more than 800 000 hectares of permanent grasslands recorded by the Statistical Office of the Slovak Republic, only 507 845 hectares were used in 2011 (Ministry of the Agriculture and Rural Development of the Slovak Republic, 2012). Table 1. Development of the permanent grassland used area (ha) and dry matter yield (DMY: t ha -1) Region 2009 Area DMY Bratislava 4,807 1.78 Trnava 8,046 1.71 Trenčín 40,243 2.19 Nitra 8,068 1.48 Žilina 111,815 2.25 Banská Bystrica 131,719 1.69 Prešov 137,974 1.66 Košice 65,665 2.09 Slovak Republic 508,337 1.86 Source: Statistical Office of the Slovak Republic 2010 Area 4,112 8,429 40,967 7,653 110,505 123,168 132,646 61,924 489,404 DMY 1.65 1.73 2.31 2.25 2.38 2.11 1.82 1.91 2.02 2011 Area 4,774 8,742 41,813 8,719 112,090 128,470 138,618 64,619 507,845 DMY 2.31 1.84 2.40 1.97 2.33 1.93 1.67 2.00 2.06 Table 1 shows the stabilization of the permanent grassland used area at national level for the three successive years, 2009-2011. Regarding the use of permanent grasslands in regions, there is evidence of a small increase of permanent grasslands in lowland regions (Trnava and Nitra) whereas the grassland used area has decreased in the Banská Bystrica mountain region by 2.5%. The analysis of stocking rate (Table 2) identified a quite logical location of farming systems in Slovakia, with intensive livestock farming in lowland regions and, conversely, extensive use of permanent grasslands in regions with the highest share of LFAs and important grassland habitats. However, the figures on dry matter yield (Table 1) indicate extensive use of permanent grasslands in both favourable lowland and less-favourable upland and mountain regions. Despite the findings on the positive effect of species-rich permanent grasslands on both quality and nutritive value of animal products, an assessment on the impact of extensive grassland management on herbage quality at selected farms in Slovakia showed that herbage did not meet the nutritional and energy requirements of dairy and beef cattle (Jendrišáková and Kizeková, 2011). The consequence of extensive use of permanent grassland is that concentrates have to be used on the majority of cattle farms in Slovakia, which is problematic if environmental and economic sustainability is also considered. Chrastinová and Burianová (2012) also support these findings and reported loss-making performance of cattle farming since 2008. The steadily decreasing cattle density in upland and mountain regions has led to a sharp decline of carbon input from grazing animals, which has resulted in a decrease of soil organic carbon stocks on permanent grassland, in comparison with arable land (Bančíková et al., 2013). These findings indicate that long-term extensive management of permanent grasslands could negatively affect some of their supporting services. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 371 Table 2. Development of number of herbivores and stocking rates, years 2009-2011 Region Bratislava Trnava Trenčín Nitra Žilina Banská Bystrica Prešov Košice Slovak Republic 2009 Herbivore LU Cattle Sheep, goats, horses 19,069 92,045 56,117 79,715 92,933 749 1,304 7,736 4,224 18,428 100,955 115,084 61,892 617,810 28,886 18,620 10,611 90,558 Stocking rate of herbivore LU ha-1 permanent grasslands 4.1 11.6 1.6 10.4 1.0 1.0 1.0 1.1 1.4 2010 Herbivore LU Cattle Sheep, goats, horses 19,016 91,573 56,739 79,541 92,349 710 1,378 7,791 4,269 18,947 101,453 117,788 62,287 620,746 30,355 18,899 11,506 93,855 Stocking rate of herbivore LU ha-1 permanent grasslands 4.8 11.0 1.6 11.0 1.0 1.1 1.0 1.2 1.5 2011 Herbivore LU Cattle Sheep, goats, horses 19,949 90,007 56,655 78,527 92,737 667 1,278 7,874 4,291 19,057 99,472 115,062 61,016 613,425 30,333 18,571 11,322 93,393 Stocking rate of herbivore LU ha-1 permanent grasslands 4.3 10.4 1.5 9.5 1.0 1.0 1.0 1.1 1.4 Source: Statistical Office of the Slovak Republic, own calculation The most alarming result of the loss-making cattle farming has been a reduction in the consumption of meat and dairy products. In 2009, 2010 and 2011, the annual consumption of beef and veal meet was 4.4, 4.3 and 3.7 kg per inhabitant in Slovakia, which is only 24% of amount in the Recommended Allowances of Foodstuffs (Sitárová, 2010; Ministry of the Agriculture and Rural Development of the Slovak Republic, 2012). Conclusion In Slovakia, extensive management of permanent grasslands in both favourable and less favourable areas for nearly the past two decades has lead to degradation of their provisioning and supporting services. Acknowledgements The study was supported by the Slovak Research and Development Agency grant No. APVV0098-12 and by the research programme of the Ministry of Agriculture of the Slovak Republic. References Barančíková G., Makovníková J., Skalský R., Tarasovičová Z., Nováková M., Halás J., Koco Š. and Gutteková M. (2013) Changes in organic carbon pool in agricultural soils and its different development in individual agroclimatic regions of Slovakia. Agriculture (Poľnohospodárstvo) 59, 1-9. Chrastinová Z. and Burianová V. (2012) Economic efficiency of Slovak agriculture and its commodity sectors. Agricultural Economics 58, 92-99. Jendrišáková S. and Kizeková M. (2011) Grassland habitats conservation. Banská Bystrica : CVRV - VÚTPHP Banská Bystrica. 150pp. Ministry of the Agriculture and Rural Development of the Slovak Republic (2012). Report on Agriculture and Food Industry in the Slovak Republic 2011 (Green report). 36pp. RegDat. Regional Statistics Database of the Statistical Office of the Slovak Republic Sitárová T. (2010) Food Consumption in the SR. Statistical Office of the Slovak Republic. 29pp. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 372 Herbage yield and quality of a limestone grassland managed differently for 30 years Seither M. Agricultural Centre for cattle production, grassland management, dairy management, wildlife and fisheries Baden-Wuerttemberg (LAZBW), Atzenberger Weg 99, 88326 Aulendorf, Germany Corresponding author: Melanie.Seither@gmx.net Abstract In an observational study over 31 years, the goal was to investigate the effect of mulching vs. mowing and the influence of fertilization (no fertilization vs. different N-P-K and P-K levels) on plant diversity in limestone grassland. Here, the management effects on herbage yield and quality were examined. Yields increased with the level of P-K or N-P-K fertilization, indicating N and P limitation at the site. Low P-K fertilization led to the highest plant diversity of all treatments. It increased the yield by 1.2 t ha-1 compared to no fertilization and resulted in comparable metabolizable energy (ME) content. Since it meets both agricultural and nature conservation goals, the requirement for zero-fertilization of nature protection sites should therefore be reconsidered. Keywords: utilization, fertilization Introduction Limestone grasslands are among the most species-rich habitats in Europe and often comprise a multitude of endangered species (Niemelä and Baur, 1998). Therefore they are under nature protection, implying for the farmers that their management has to ensure the maintenance of the species richness (Fauna-Flora guideline 1992). From agricultural aspects, the management of these grasslands of low productivity is not profitable. For this reason, diverse limestone grasslands have declined to a large extent, in particular due to management changes or abandonment (Niemelä and Baur, 1998; Steiner, 2011). Above others the nutrient limitation in limestone grasslands is one reason for their high diversity (Köhler et al., 2001). On the other hand, forbs and legumes rely on a sufficient P and K supply (Magyar et al., 2008). Therefore, there might be fertilizer levels that both facilitate plant species richness and increase herbage productivity and quality. In this paper, the effects of mowing in comparison with the less labour-intensive mulching, and the influence of different fertilizer levels on herbage yield and quality are analysed. Materials and methods The site is located in the nature protection site ‘Filsenberg’ in Mössingen (Swabian Alb, federal state Baden-Württemberg, Germany) at 780 m above sea level. The mean yearly temperature is 6.0 - 6.5 °C, and annual mean precipitation is 850 mm. The soil is a calcareous brown-soil with a depth of around 30 cm, and the texture is sandy or clayey loam over limestone. The vegetation at the start of the observational study was a Gentiano vernae-Brometum (KUHN 1957), mown once a year. From the middle of the 20th century onwards it received low dosages of P fertilizer (‘Thomasmehl’: Ca3(PO4)2 (Ca2SiO4) containing Fe, Mn, Mg and Cr) or mineral mixed fertilizer. In 1983, an observational study with varied utilization and fertilization was conducted on plots of 126 m² in size. The treatments were: 1) mulching (Mul); 2) mowing without fertilization (M); 3) P-K 10-16 (M+PK1); 4) P-K 16-64 (M+PK2); 5) N-P-K 10-10-16 (M+NPK1); 6) NP-K 20-20-32 (M+NPK2); and 7) N-P-K 40-16-64 (M+NPK3; P: P2O5, K: K2O. All values are in kg ha-1 a-1). In early spring, plots were yearly fertilized with mineral fertilizer (N fertilizer Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 373 ‘Kalkammonsalpeter’: about 80% H4N2O3; P fertilizer ‘Novaphos’: 23% P and 8% S; K fertilizer ‘Kornkali’: 40% K2O, 6% MgO, 3% Na, 5% S). In this paper, treatment effects on dry matter yield and metabolizable energy (ME; GfE, 1997) are shown. Since there was spatial and temporal pseudoreplication (i.e. treatments were not replicated and the same plots were surveyed repeatedly), statistical analyses were not possible. Results and discussion Differential management did not affect the P and K content of the soil: in all plots the plantavailable P content of the soil maintained a very low level (CAL extractable P2O5: 2.0 ± 0.5 in 1986, 1.4 ± 0.5 in 2012) throughout the study, while the K content sank from moderate (20.1 ± 3.0) to low values (9.9 ± 0.5; CAL extractable K2O; values are means ± standard deviations across plots in mg 100 g-1 dry soil; sampling depth 0-10 cm) in all plots. Herbage yields varied within treatments during the study period. This was ascribed to yearly differences in precipitation. Yields of M+PK1 and of the unfertilized plot were up to 3.5 t ha-1 a-1, which is in the range found for limestone grasslands in other studies (Ryser et al., 1995; Hochberg and Zopf, 2011; Karrer, 2011). Mulching and M+PK1 reached comparable yields, somewhat larger than those of mowing without fertilization. Hochberg and Zopf (2011) found alternately mown and mulched plots to be more productive than mown plots, an effect likely due to the nutrient return via the mulch layer. The herbage yield was generally increased by fertilization, indicating a limitation of both N and P at the site. The highest productivity was achieved with M+NPK3. In this treatment the proportion of grasses increased during the trial period; on average between 2008 and 2013 a grass proportion of 85% was reached, compared to 55% in the mown unfertilized plot. In grass-based swards, generally higher herbage yields are reached than in forb-based swards (Magyar et al., 2008). Figure 1. Development of the herbage yield in periods of 5 or 6 years (2008-2013). The herbage energy content, lying in the range found for other limestone grasslands (Hochberg and Zopf, 2011) at the start of the observation, decreased considerably in all plots after twenty years. This was related to an increase in the grass proportion (data not shown). After 22 years, management measures differed in their effect on vegetation composition. M+PK1 proved to be the treatment that facilitated plant diversity the most (Briemle and Tonn, 2008). The plots with highest dicot proportions, the unfertilized mown plot and M+PK1, resulted in a higher ME and N content of the herbage than the other treatments (Table 1). This is ascribed to the lower decline in herbage quality in dicot- compared to grass-based swards (Briemle et al., 1991). The P and K content in the biomass were positively related to the applied amount of P or K fertilizer. Thus the fertilizer outbalanced the potential positive effect of high legume and forb proportions on the mineral content of the forage. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 374 Table 1. Metabolizable energy (ME) and N, P and K content in the herbage (mean values for the respective time periods; herbage analyses for the determination of ME were not available before 1989). na: no data available, as M+PK2 and M+NPK3 did not start until 1991. For treatment abbreviations see Methods section ME (MJ kg-1 DM) 1989- 2005-13 96 MUL 8.6 8.3 M 8.9 8.7 M+PK1 8.8 8.5 M+PK2 8.4 8.1 M+NPK1 8.7 8.3 M+NPK2 8.8 8.3 M+NPK3 8.7 8.1 N (g kg-1) 1983-87 2008-13 18.6 18.2 18.5 na 18.4 18.5 Na 14.7 15.9 15.8 15.9 14.5 14.2 13.8 P (g kg-1) 1983-87 2008-13 1.2 1.4 1.4 na 1.4 1.5 na 1.4 1.5 1.8 2.1 1.9 2.2 2.2 K (g kg-1) 1983-87 2008-13 14.8 13.9 14.6 na 14.7 14.8 na 17.3 15.0 16.1 20.5 17.1 17.9 20.0 Conclusion Treatment M+PK1 considerably increased the herbage yield and reached a level of herbage quality comparable to that of the unfertilized mown plot. As M+PK1 meets both agricultural and nature conservation goals, the requirement for zero fertilization, often advocated for nature protection sites, should therefore be reconsidered. Acknowledgements Thanks are given to Karin King and Sylvia Engel for vegetation analyses, Petra Hirsch for laboratory analyses and Katja Herrmann for the English revision. References Briemle G., Eickhoff D. and Wolf R. (1991) Mindestpflege und Mindestnutzung unterschiedlicher Grünlandtypen aus landschaftsökologischer und landeskultureller Sicht: Praktische Anleitung zur Erkennung, Nutzung und Pflege der unterschiedlichen Grünlandgesellschaften. Landesanstalt für Umwelt, Messungen und Naturschutz Baden-Württemberg, Karlsruhe, 160 pp. Briemle G. and Tonn B. (2008) Auswirkungen geringer mineralischer Düngung auf Pflanzenbestand und Biomasseproduktion eines artenreichen Halbtrockenrasens. Mitteilungen der Arbeitsgemeinschaft Grünland und Futterbau (9), 266–269. GfE (1997) Energie und Nährstoffbedarf landwirtschaftlicher Nutztiere, Empfehlungen des Ausschusses zur Schätzung des Energiegehaltes im Grundfutter. Gesellschaft für Ernährungsphysiologie - Ausschuss für Bedarfsnormen. Hochberg H. and Zopf D. (2011) Sustainable management of Mesobrometum without animals? Grassland Science in Europe 16, 407–409. Karrer G. (2011) Dynamics of biomass production in extensively managed meadows at the eastern edge of the Alps. Grassland Science in Europe 16, 598–600. Köhler B., Ryser P., Güsewell S. and Gigon A. (2001) Nutrient availability and limitation in traditionally mown and in abandoned limestone grasslands: a bioassay experiment. Plant and Soil 230, 323–332. Magyar E.I., Buchgraber K., Warner D. and Szemán L. (2008) Der Einfluss von Düngung und Nutzung auf die Entwicklung der Kräuter in Grünlandbeständen. Acta Botanica Hungarica 50, 143–158. Niemelä J. and Baur B. (1998) Threatened species in a vanishing habitat: plants and invertebrates in calcareous grasslands in the Swiss Jura mountains. Biodiversity and Conservation 7, 1407–1416. Ryser P., Langenauer R. and Gigon A. (1995) Species richness and vegetation structure in a limestone grassland after 15 years management with six biomass removal regimes. Folia Geobotanica 30, 157–167. Steiner L. (2011) Einfluss von Fragmentierung und Isolation auf die Populationsbiologie und Diasporenbank von Kalk-Magerrasen (Mesobrometum erecti) in Südwestdeutschland. Dissertation. Carl von Ossietzky Universität Oldenburg. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 375 BIOECOSYS: towards the development of a decision support tool to evaluate grassland ecosystem services Campion M.1, Ninane M.2, Hautier L.3, Dufrêne M.4 and Stilmant D.1 1 Walloon Agricultural Research Centre, Agriculture and Natural Environment Department, Belgium 2 Walloon Agricultural Research Centre, Production and Sectors Department, Belgium 3 Walloon Agricultural Research Centre, Life Sciences Department, Belgium 4 University of Liège, Forest, Nature and Landscape Department, Belgium Corresponding author: m.campion@cra.wallonie.be Abstract As underlined by the Millennium Ecosystem Assessment report (2005), it is necessary to take into account and preserve all the functions and connected services associated with ecosystems. Therefore, agroecosystems such as grasslands, covering near one-fifth (19.5%) of the European territory – rising to 50% of the Walloon utilized agricultural area – and providing important ecosystem services, have to be studied and managed as multifunctional units, opening new opportunities for valorization. This will allow the answering of societal expectations oriented towards more sustainable agriculture and improved use of natural resources. In this context, the final and main goal of the BIOECOSYS project is the development of a specific methodology and decision support system for the quantification and valuation of ecosystem services provided by the grassland ecosystem linked to its history and management, to its soil and climate context, to its location in the landscape and in the socio-ecosystem. To reach this objective it is necessary to produce integrated knowledge at the different levels of organization of grassland agroecosystems: (1) to quantify ecosystem services to integrate them in decisionmaking processes, and (2) to give a value (economic or not) to the services provided to guide decision-making choices. A first result of the project, a provisional scheme of grassland ecosystem functioning in relation to the services provided, is presented. Keywords: agroecosystem, management, valuation, DSS, ecosystem function Introduction During the last decades, the over-exploitation and degradation of ecosystems has been recognized in different reports (Millennium Ecosystem Assessment, 2005). In parallel, growing demand for agricultural products, international awareness of biodiversity loss (e.g, the 1992 United Nations Rio Earth Summit) and climate changes have led to reconsideration of agroecosystems. The current challenge is to maintain or restore these ecosystems, enabling them to produce enough food but also services to improve the environment and human wellbeing. This must be done not only by avoiding pollution but also by maintaining and increasing ‘ecosystem services’ such as public goods (maintenance of water and air quality) or environmental services (maintenance of biodiversity, carbon sequestration) (Lemaire et al., 2005; FAO, 2007). To reach these goals, agriculture must rely on the increasing scientific knowledge about agroecosystems, especially considering the ecosystem-services concept, which is very useful in the agricultural and public policies establishment (Lamarque et al., 2011). The concept of agroecosystems multifunctionality, translated into ecosystem services, provides a new framework to drive researches necessitating genuinely inter-disciplinary approach (Hervieu, 2002; Lemaire et al., 2005). This concept also represents a key element in the development of ecologically intensive agriculture which aims to optimize the use of agroecosystem functionality to produce more while preserving, or even enhancing, its environmental services (Bonny, 2011). In Europe, grasslands are essential ecosystems representing near one-fifth (19.5 %) of the European territory – rising to 50% of the Walloon Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 376 utilized agricultural area – and providing important ecosystem services such as forage production, erosion and resources regulation, etc. Nevertheless, grasslands are actually threatened by land conversion to crops and face a significant pressure. Despite the demonstrated abilities of grassland systems to provide numerous ecosystem services (Amiaud and Carrère, 2012), for the diversity of grassland agroecosystems there remains a need to provide precise information on these services linked to their management and location (Puydarrieux and Devaux, 2013). BIOECOSYS and its objectives In order to allow the inter-disciplinary approach that will be necessary, the expertise of several units of the Walloon Agricultural Research Centre will be mobilized in order to carry out this project. These units are (1) Farming systems, Territories and Information technology Unit, (2) Plant protection and Ecotoxicology Unit, (3) Soil fertility and Water protection Unit, (4) Food and feed quality Unit, and (5) Crop production systems Unit. The final and main goal of BIOECOSYS is the development of a specific methodology and decision support system for the quantification and valuation of ecosystem services provided by the grassland ecosystem linked to its history and management scheme, its soil and climate context, and its location in the landscape and in the socio-ecosystem. To reach this objective it is necessary to produce integrated knowledge at the different levels of organization of grassland agroecosystems. Several ecosystem services will be quantitatively studied at the field and the landscape scales, where the basic biogeochemical processes are acting, while the valuation of grassland ecosystem services will be evaluated at the regional scale, supporting socio-economic and political decisions (Lemaire et al., 2005; Hein et al., 2006). Grassland ecosystem services conceptualization: a first output Early reflections and bibliographic researches have resulted in a first draft of grassland ecosystem functioning in relation to the services provided. Firstly, the CICES classification was examined to identify the different ecosystem services provided by grasslands. We applied the methodology described by Lamanda (2012) to conceptualize the grassland system. These services were connected with the grassland ecosystem functioning. The grassland is schematized on the basis of its three main compartments: (1) the soil, (2) the vegetation cover and (3) the faunal composition, in which various processes are taking place. Several abiotic factors (e.g. topography, landscape, climate, etc.) influence the functioning of these processes resulting in chain reactions which alter the supply of ecosystem services. In parallel, agricultural practices have a demonstrated and variable impact on the provision of ecosystem services. These different management methods (mowing rate, grazing intensity, fertilization schemes, etc.) and their impacts on grassland ecosystem services have to be modeled in order to allow their integration in the decision support system. This conceptualization frame will be validated by three focus groups mobilizing expertise in different fields interconnected. This will allow us to give a relative importance and an orientation to the different interconnections highlighted. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 377 Figure 1: Conceptualization of grassland ecosystem in link to ecosystem services provided References Amiaud B. and Carrère P. (2012) La multifonctionnalité de la prairie pour la fourniture de services écosystémiques. Fourrages 211, 229–238. Bonny S. (2011) L’agriculture écologiquement intensive: nature et défis. Cahiers Agricultures, 20, 451–462. FAO (2007) La situation mondiale de l’alimentation et de l’agriculture: payer les agriculteurs pour les services environnementaux. FAO, Rome. Hein L., van Koppen K., de Groot R.S. and van Ierland E.C. (2006) Spatial scales, stakeholders and the valuation of ecosystem services. Ecological Economics 57, 209–228. Hervieu B. (2002) Multi-functionality: a conceptual framework for a new organization of research and development on grasslands and livestock systems. Grassland Science in Europe 7, 1–4. Lamanda N., Roux S., Delmotte S., Merot A., Rapidel B., Adam M. and Wery J. (2012) A protocol for the conceptualisation of an agro-ecosystem to guide data acquisition and analysis and expert knowledge integration. European Journal of Agronomy 38, 104–116. Lamarque P., Quétier F. and Lavorel S. (2011) The diversity of the ecosystem services concept and its implications for their assessment and management. Comptes Rendus Biologies 334, 441–449. Lemaire G., Wilkins R. and Hodgson, J. (2005) Challenges for grassland science: managing research priorities. Agriculture, Ecosystems and Environment 108, 99–108. Millennium Ecosystem Assessment (2005) Ecosystems and human well-being - Biodiversity synthesis. http://www.maweb.org/en/synthesis.aspx Puydarrieux P. and Devaux J. (2013) Quelle évaluation économique pour les services écosystémiques rendus par les prairies en France métropolitaine ( No. 37), Notes et études socio-économiques. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 378 Grassland biodiversity: how we might meet international commitments Peel S. Natural England, 25 Queen Street, Leeds, LS1 2UN, United Kingdom Corresponding author: Steve.Peel@naturalengland.org.uk Abstract England has very challenging targets to meet the UN Convention on Biological Diversity, whilst aspiring to produce more food. And with a low proportion of publicly owned land this has to be achieved largely from commercial farmland. Most grassland receives less than 50 kg ha-1 of N fertilizer and has high potential for increasing biodiversity as well as greater production. This requires: the 2% of grassland which is species-rich to be better managed and expanded by restoration and creation, for which we have good techniques; other priority habitats, for example grazing marshes, better managed to protect priority birds and other species; very small areas (1%) in corners and margins allowed to grow tall for invertebrate nest sites; winter seed for birds (minimum 2%) which could be provided by allowing ryegrass to set seed; some areas (2-5%) grazed leniently and not topped, to provide habitat for invertebrates which are food for birds; much more grassland sown with legumes and robust herbs, a small proportion being allowed to flower to provide for bees and other pollinators; protection of legumes and herbs by avoidance of overall use of broad-spectrum herbicides, controlling thistles and other injurious weeds by targeted treatment. Keywords: Grassland, biodiversity, habitat, invertebrates, birds, legumes. Introduction England has a very high population density exceeding 400 km-2 and one of the lowest proportions of publicly owned land in Europe (Cahill, 2002) so a high proportion of the biodiversity and ecosystem services has to be delivered from privately owned agricultural land. This is one of the reasons why English agri-environment schemes are among the most ambitious and best-funded in Europe. The UK government is committed to challenging actions to meet the UN CBD Aichi targets, whilst at the same time aspiring to produce more food (DEFRA, 2011). Here I consider how this might be achieved from grassland. Fertiliser inputs Inputs to grassland fell from a peak in the late 1980s to less than half of that by 2012 (Table 1). Only 21% receives >100 kg ha-1 and 41% receives zero N fertilizer (Table 2). It might be expected that declining N-use would be compensated by increasing reliance on legumes, but UK sales of clover seed fell from 900 t in 1981 to 400 t in 2004 (FERA, 2005). Since then data have not been available. Table 1. Use of fertilizer nitrogen, phosphate and potash (kg ha-1) on grassland in England and Wales 19732012. (British Survey of Fertilizer Practice, 2012) Year 1973 1986 2002 2012 Nitrogen 85 135 85 54 Phosphate 34 22 18 8 Potash 22 33 24 11 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 379 Table 2. Use of fertilizer nitrogen (kg ha-1) on grassland in England and Wales: percentage of grassland area by field application rate (British Survey of Fertilizer Practice, 2012). kg N/ha 0 1-49 50-99 100-199 >200 % 41 15 21 18 3 The current agri-environment scheme, Environmental Stewardship (ES), has options which do not allow fertilizer N. Approximately 400,000 ha (9%) of grassland is in such options and a further 450,000 ha (11%) is in options which allow a maximum of 50 kg ha-1 although often in practice receives less, or even zero. So it is clear that most grassland in England is managed quite extensively, even when not in agri-environment agreement. Biodiversity, and production Table 3 shows that 2% of grassland is rich in plants, and also in scarce invertebrate species. A further area of semi-natural grassland, mainly of coastal and floodplain grazing marsh, is of particular importance for scarce birds and other target species. Lawton et al. (2010) concluded that the most important actions required to protect and increase biodiversity are to better manage and enlarge these habitats, and to create or restore new core areas. We have good understanding of such management, and of soil suitability and techniques for restoration (Peel et al., 2012) from existing grassland and arable land. Restoration and creation have been targeted in ES at land with low phosphorus status and/or other limiting characteristics (Peel and Diack, 2007) and there is evidence that grassland meeting the minimum threshold of priority habitat can be created, sometimes in less than 10 years (Natural England, 2013), although it may take c.100 years to fully develop. Livestock output from priority habitat is low but individual livestock performance and health on species-rich neutral and calcareous grassland can be good, even from commercial breeds (DEFRA, 2014a). Table 3. Summary of biodiversity of enclosed grassland in England Descriptor Priority habitat Semiimproved grassland Improved grassland Area (‘000 ha) % of grassl and Typical higher plant species (m-2) Probable fertiliser N (kg ha-1) 104 2 16-40 0 1453 33 9-15 0-50 2856 65 <10 0-300+ Value for biodiversity Current Potential Very high Moderat e to very high Lowmoderate Some – condition of most can be improved. High – introduce more species, diversify sward structure. High – introduce legumes and robust herbs, allow to flower occasionally, no overall herbicide sprays. 4413 100 Sources of area data: UK National Ecosystem Assessment (2011); Countryside Survey (2009). But increasing priority habitat towards 3% of grassland is not enough to reverse the declines of formerly common pollinators, other invertebrates and farmland birds, nor in the longer term to enable them to adapt to climate change. This requires core areas to be better connected (Lawton et al., 2010) and grassland has a key role to play in making farmland more permeable to wildlife. Further plant species, and more diverse structure, can be added to the 33% of grassland that already has moderate botanical diversity. And on the 65% that is species-poor a wider range of grasses, more legumes (e.g. Trifolium pratense, T. repens, Lotus corniculatus) and robust herbs (e.g. Cichorium intybus, Plantago lanceolata, Centaurea nigra, Sanguisorba minor) can be sown (DEFRA, 2014b). Current output from improved and semi-improved Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 380 grassland is quite variable: herbage production and nutritive value could be improved by greater reliance on legumes (Lüscher et al., 2013). Conclusions and opportunities With a well-resourced agri-environment scheme, priority habitat could be better managed and expanded to protect plant species and some specialized invertebrates. But many invertebrates and farmland birds require a range of resources on which to feed and breed at a larger scale than can be provided in England by priority habitat alone. Semi-improved and improved grassland, in association with woodland, hedgerows, wetland and arable land, could provide these resources. It would need: Very small areas (1%), in field corners and margins, allowed to grow tall for invertebrate nest sites. Could be sited to also buffer watercourses. Winter seed for birds (min 2%). Could be provided from plots on arable areas. Otherwise by allowing ryegrass to set seed (DEFRA, 2010) Some areas (2-5%) grazed leniently and not topped, to provide habitat for invertebrates which are food for birds (DEFRA, 2013). Much more grassland sown with legumes and robust herbs. A proportion must flower to provide for bees and other pollinators, which provide pollination services to crops. Protect legumes and herbs: control thistles and other injurious weeds by targeted treatment, not overall use of broad-spectrum herbicides. Because grassland is currently relatively extensively managed, yet use of legumes is limited, these changes would not necessarily reduce, and might increase, livestock production. References Cahill K. (2002) Who Owns Britain and Ireland? Canongate Books 465pp. Countryside Survey (2009) England Results from 2007. NERC/Centre for Ecology & Hydrology, Department for Environment, Food and Rural Affairs, Natural England, 119pp. (CEH Project Number: C03259). DEFRA (2010) Grass silage as a new source of winter food for declining farmland birds. Project BD1455. DEFRA (2011) Biodiversity 2020: a strategy for England’s wildlife and ecosystem services. Department for Environment, Food and Rural Affairs: London. DEFRA (2013) Utility of lenient grazing of agricultural grassland to promote in-field structural heterogeneity, invertebrates and bird foraging. Project BD5207. DEFRA (2014a) Sustainable Management Systems for Unimproved Neutral Grassland. Project BD1460. DEFRA (2014b) Extension to WEB (Widescale Enhancement of Biodiversity). Project BD5208. FERA (2005) Seed Trader’s Annual Returns. Lawton J. (2010) Making Space for Nature: a review of England’s wildlife sites and ecological network. DEFRA. Lüscher A., Mueller-Harvey I., Soussana J.F., Rees R.M. and Peyraud J.L. (2013) Potential of legume-based grassland-livestock systems in Europe. Grassland Science in Europe 18, 3-29. Natural England (2013) NECR107 Peel S. and Diack I.A. (2007) Restoring botanical biodiversity in permanent grassland – a targeted pro-active approach. Grassland Science in Europe 12, 516–519. Peel S., Chesterton C., Cooke A.I., Jefferson R.G., Martin D., Smith B.M., Smith S. and Tallowin J.R.B. (eds) (2012). Restoring diverse grassland: what can be achieved, where, and what will it do for us? Aspects of Applied Biology 115. 193pp. Warwick: Association of Applied Biologists. UK National Ecosystem Assessment (2011) The UK National Ecosystem Assessment Technical Report. UNEPWCMC, Cambridge. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 381 Resilience of Mediterranean ecosystems: tree and management effects on variability of herbaceous pastures in a dry year López-Sánchez A.1,2, San Miguel A.2 and Roig S.1,2 1 ECOGESFOR (Ecology and Sustainable Forest Management Research Group), Technical University of Madrid, Ciudad Universitaria, s/n 28040 - Madrid, Spain, 2 Department of Silviculture and Pastures, Technical University of Madrid, Ciudad Universitaria, s/n 28040 - Madrid, Spain Corresponding author: aida.lopez.sanchez@gmail.com Abstract The presence of trees and adaptative management techniques play an important role within Mediterranean agroforestry ecosystems. Dehesas are good examples of Mediterranean 'High Natural Value' agriculture and produce important natural resources. Our work focuses on the influence of trees and main management techniques (stocking rate and site quality) on yield and alpha diversity of a Dehesa in Central Spain. We analysed total yield, Richness and Shannon-Wiener indexes within the herbaceous pastures under 16 holm oaks (Quercus ilex ssp ballota (Desf.) Samp) located in different management zones, in a very extreme dry year, with phosphoric fertilization, intensive stocking rate, medium stocking rate, and high soil quality or low soil quality. Tree canopy had a significant influence on herbaceous yield and alpha indexes (both were higher outside canopy influence). Phosphoric fertilization also had a significant effect on herbaceous yield and alpha diversity. Tree presence, livestock and soil management were shown to be important factors modifying the characteristics of herbaceous pastures in Mediterranean agroforestry systems, even in extreme climatic conditions. Keywords: Mediterranean management, yield, alpha diversity, tree-grass interaction Introduction The presence of trees plays an important role providing a wide range of important ecological functions and ecosystems services within Mediterranean agroforestry ecosystems (Manning et al., 2006). Within the influence area of their canopy and roots, trees can change pasture species composition, structure, spatial distribution and biomass (Scholes and Archer, 1997). These agroforestry systems are associated with different common techniques of management (e.g. intensive grazing, fertilization) affecting the extensive yield-diversity of pastures and tree-grass interactions. Grazing by large herbivores alters the composition and biomass of grasslands (Bartolome and McClaran, 1992). Fertilization with minerals suitable for organic agriculture can be a management and conservation tool for semi-natural grasslands (Pãcurar et al., 2012). Mediterranean Dehesas are good examples of Mediterranean 'High Natural Value' agriculture in Europe, assigned as 'Site of Community Importance' (43/92/EEC Directive), and they are producers of important natural resources. Therefore, considering management decisions over tree-grass interaction in ecosystem dynamics and functional studies is a key research line. Our work focuses on the influence of trees and main management techniques (stocking rate and site quality) on yield and alpha diversity of a Dehesa in Central Spain. Materials and methods The study area was located within a Dehesa in Central Spain (39ºN, 5ºW; 350 m asl). The climate is continental Mediterranean and soils are sandy (>80% sand), acidic, and poor in organic matter (<1%). Sixteen holm oaks were selected for the study, distributed in four zones (5 ha each) with different management procedures (phosphoric fertilization; intensive stocking rate; medium stocking rate and high soil quality, hereafter MH; and medium stocking rate and low soil quality, hereafter ML). All of them were grazed by sheep. The average canopy radius Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 382 (standard deviation) of trees is 6.2 m (1.4). At each tree, we located six sampling frames (50 × 50 cm), considering three positions according to the proximity of the trunk (beneath crown: 0.5 radius; edge crown: 1 radius; and beyond canopy’s influence: > 1 radius). In total, 96 sampling units were studied. In each one, we analysed floristic composition and abundance by species or morphospecies, and calculated Richness and Shannon-Wiener indexes to measure alpha diversity. In addition, we mowed aboveground herbaceous biomass at ground level at the end of May; this was taken immediately to the laboratory where it was dried at 80ºC until stabilized at constant dry weight. We obtained the total dry matter at each sampling unit. We used generalized linear mixed models (GLMM) to study relationships between total yield or alpha diversity measures and independent factors (type of management and distance from the trunk). The model averaging approach (Burnham and Anderson, 2002) was used. We used R programming environment (Version 3.0.2, R Foundation for Statistical Computing, Vienna, Austria. http://www.R-project.org) for data processing, analysis and presentation of results. Results and discussion Mediterranean areas are associated with a high topo-edaphic and climatic variability, and specifically inter-annual rainfall variability (Gea-Izquierdo et al., 2010). It means that some years could be very extremely dry, as our studied 2012 year (298.1 mm; 596.7 mm average of last 20 years). However, in spite of these dry conditions, we found differences in biomass and composition of the herb layer due to the presence of trees and different types of management. A total of 70 different plant species or morphospecies was recorded at the estate. Beneath crown positions there were significantly lower values for total yield (166.1 kg ha-1± 107.6) than at the edge-crown positions (233.3 kg ha-1± 155.0), which showed lower values than beyond-crown positions (494.8 kg ha-1± 217.6) (Table 1). Within Mediterranean silvopastoral systems and depending on climatic conditions, the influence of trees can be opposite (McClaran and Bartolome, 1989). Both alpha indexes were significantly higher outside canopies than under the influence of tree canopies (beneath- and edge-crown projection) (Table 1). However, trees provide different species composition, increasing the whole-system diversity in Mediterranean agroforestry systems (Marañón, 1986). The fertilized plot showed higher values for total yield and Richness than zones with intensive stocking rate and MH zones (without differences between them). Shannon-Wiener index was higher at the fertilized plot than MH and MP zones (Table 1). Phosphoric fertilization had a significant effect on yield and alpha diversity of the herbaceous layer, but year effect (rainfall) mainly determined the effect of phosphoric fertilization. We did not find differences for Shannon-Wiener index between fertilized and intensive grazing zones. The tree random effect scarcely modifies the yield or alpha diversity average of the herb layer among trees. Conclusion High annual rainfall fluctuations within Mediterranean agroforestry systems, as dehesas, provide large differences on interactions between ecosystems components. Even at extremely dry conditions we found notable differences due to the presence of trees and different management techniques. The tree effect (under, beneath and edge canopies influence) provided different species composition increasing herbaceous diversity of whole system. Some management practices as P2O5 fertilization or intensive sheep grazing had a significant effect on yield and alpha diversity of herbaceous layer. The presence of trees and appropriate management techniques maintain and conserve the quality of dehesas, increasing the resilience of the Mediterranean ecosystem. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 383 Table 1. Summary of the top GLMM fitted (delta<2) to analyse the differences on total yield and alpha diversity indexes. Coeff.: estimated coefficient average; SE: standard error average. Number of top models Response variable Fixed effects Importance 1 a Management 1.00 Total dry matter Factors Coeff. SE P MH -40178 14963 0.007 ML 31347 27094 0.247 -32255 15588 0.039 Ecotone 14394 6797 0.034 Outside 198869 45078 <0.001 MH -0.406 0.117 0.001 ML -0.160 0.109 0.143 I -0.226 0.111 0.042 Ecotone 0.170 0.108 0.118 Outside 0.520 0.101 <0.001 MH 0.137 0.044 0.002 ML 0.088 0.040 0.027 I 0.070 0.039 0.070 Ecotone -0.075 0.042 0.075 Outside -0.188 0.040 <0.001 I Distance b 1 Richness Management Distance c 2 Shannon Management Distance 1.00 1.00 1.00 0.53 1.00 a Error distribution (link): Gamma (power, l=2); Deviance explained: 15.0%; Dispersion: 0.31; AIC= 1230.6 Error distribution (link): Poisson (log); Deviance explained:19.0%; Dispersion: 0.25; AIC= 422.9 c Error distribution (link): Gaussian (inverse); Deviance explained:28.5.0%; Dispersion: 0.91; AICmin= 158.9 I: intensive stocking rate; MH: medium stocking rate and high soil quality; ML: medium stocking rate and low soil quality; Ecotone: samplings located at 1 radius of crown projection; Outside: samplings located beyond canopies’ influence. b Acknowledgements This study has been developed thanks to a PhD fellowship ‘FPU’ of Spain Ministry of Education, Culture and Sport to ALS. We thank staff of ‘Dehesón del Encinar’ for the access to the estate. References Bartolome J.W. and McClaran M.P. (1992) Composition and production of California oak savanna seasonally grazed by sheep. Journal of Range Management 45, 103-107. Burnham K.P. and Anderson D.R. (2002) Model selection and multimodel inference: a practical informationtheoretic approach. Springer, New York, USA, 488 pp. Gea-Izquierdo G., Allen-Díaz B., Miguel A.S. and Cañellas I. (2010) How do trees affect spatio-temporal heterogeneity of nutrient cycling in mediterranean annual grasslands? Annals of Forest Science 67, 112. Manning A.D., Fischer J. and Lindenmayer D.B. (2006) Scattered trees are keystone structures - implications for conservation. Biological Conservation 132, 311-321. Marañón T. (1986) Plant species richness and canopy effect in the savanna-like ‘dehesa’ of S.W. Spain. Ecología Mediterránea 12, 131-141. McClaran M.P., Bartolome J.W. (1989) Effect of Quercus douglasii (Fagaceae) on herbaceous understory along a rainfall gradient. Madroño 36, 141-153. Pãcurar R.S., Rotar I., Bogdan A.D., Vidican R.M. and Dale L.M. (2012) The influence of mineral and organic long-term fertilization upon the floristic composition of Festuca rubra L.-Agrostis capillaris L. grassland in Apuseni mountains, Romania. Journal of Food, Agriculture and Environment 10, 866-879. Scholes R.J. and Archer S.R. (1997) Tree-grass interactions in savannas. Annual Review Ecology, Evolution and Systematics 28, 517-544 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 384 Soil organic carbon and nitrogen stocks affected by grazing intensity in temperate permanent grassland Nüsse A.M.1, Linsler D.1, Kaiser M.1, Isselstein J.2 and Ludwig B.1 1 Department of Environmental Chemistry, University of Kassel, Nordbahnhofstr. 1A, 37213 Witzenhausen, Germany 2 Department for Crop Science, Göttingen University, von-Siebold-Str. 8, 37075 Göttingen, Germany Corresponding author: anja.nuesse@ini-kassel.de Abstract Stocks of organic C (Corg) and total N (Nt) in soils of grazed grassland ecosystems may be affected by the management intensity. However, respective data for grasslands of the temperate zone are scarce. Soil samples were taken in 0-10 cm, 10-25 cm and 25-40 cm depth from pasture plots of one ha in central Germany which were intensively (mean compressed pasture height (CPH): 6 cm), extensively (mean CPH: 12 cm), and minimally (mean CPH 18 cm) grazed. Because of the large grazing heterogeneity the 1-ha plots of different grazing intensity showed patches of high (18 cm), medium (12 cm), and low (6 cm) CPH. We took soil samples from three soil depths, and three plots as well as three sub-plots of different CPH. In the 10-25 cm soil depth, our results showed significantly higher Corg and Nt stocks for extensive grazing compared to minimal grazing, suggesting positive effects of moderate grazing intensity on Corg and Nt stocks of the subsurface mineral soil. Keywords: Grazing, management intensity, Corg stocks, Nt stocks Introduction The intensity of grassland management varies but it is often driven by the demand for forage production. The type and intensity of management may exert a marked influence on the soil organic C (Corg) and total N (Nt) stocks of pastures (Conant et al., 2001). Moreover, the Corg stored in grassland soils may be in conjunction with the Nt storage, which is important for the productivity of grassland ecosystems (Conant et al., 2001). However, previous studies show varying effects of grazing on Corg and Nt; many of these differences may be the result of variations in climate and soil properties (Derner et al., 2006). Consequently, it is not possible to extrapolate from global data sets or from data of arid or semi-arid environments to pasture conditions in the temperate zone. For the latter, there have been only few studies which have investigated the effect of the grazing intensity on soil Corg and Nt stocks. Consequently, the objectives of our study were to analyse in a first step the effects of three different grazing intensities on the stocks of Corg and Nt in three different soil depths of a permanent grassland in central Germany. Materials and methods The study site is located in central German approximately 60 km NE of Goettingen (Lower Saxony) at 250 m above sea level. The mean annual temperature is 8.2 °C and the long-term average annual precipitation is 879 mm (Şahin Demirbağ et al., 2008). The experimental field site is a mesotrophic, moderately species-rich hill grassland on a brown earth pelosol with an average number of plant species of 10.9 m-2 (Şahin Demirbağ et al., 2008). The vegetation type is a Lolio-Cynosuretum (Scimone et al., 2007). A grazing experiment with cattle and no input of fertilizers, pesticides or cutting was established in spring 2005 with following treatments: intensive grazing pressure (IGP), target mean compressed pasture height (CPH) of 6 cm extensive grazing pressure (EGP), target mean CPH of 12 cm Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 385 minimal grazing pressure (MGP), target mean CPH of 18 cm The different grazing intensities were determined by measuring the CPH in the vegetation period at intervals of two weeks and, if necessary, the stocking rates were subsequently increased or decreased to ensure the targeted CPH. Each treatment was carried out in three replications by dividing three blocks into three paddocks of 1 ha, with different grazing intensities (i.e., intensive, extensive, minimal). Because of the selective cattle grazing, the paddocks did not show a homogeneous CPH. To adress this issue, the paddocks were further subdivided into patches of different CPH (i.e., low: 0-6 cm, medium 6-12 cm, high 12-18 cm). Soil samples were taken from the patches (low, medium, high: two pseudo-field replicates) within each paddock (intensive, extensive, minimal) in three depths (0-10 cm, 10-25 cm and 25-40 cm). The samples were sieved <2 mm and stored at 4 °C. The concentrations of total C (Ct) and Nt in the bulk soil were determined by dry combustion using a CN elemental analyser (Elemtal Vario El, Heraeus, Hanau, Germany). The inorganic C concentration was measured with the Scheibler method following DIN 19682-13 (2009). The Corg content was calculated by substracting the inorganic C from the Ct content. The Corg and Nt stocks of the different soil layers were calculated for an equivalent mass of soil as suggested by Ellert and Bettany (1995) to take differences in bulk density of the respective soil layers into account (Jacobs et al., 2010). Statistical analyses were conducted with the statistic software R (R Development Core Team, 2010). The two pseudo-replicates per paddock were averaged to carry out statistical analyses for nine different variants (3 grazing intensities × 3 CPH). For the analyses of the effect of grazing intensity, all six values for one paddock were averaged. The data were analysed with one-way analyses of variance (ANOVA). Effects were considered significant for P ≤ 0.05. Results and discussion In the surface soil (0-10 cm) Corg stocks showed large variability and no significant differences between the treatments with different grazing intensities were found (mean ± standard deviation for the treatments in t ha-1 were 34 ± 7 for IGP, 39±10 for EGP and 33±10 for MGP). The Corg and Nt stocks in the soil depth 10-25 cm, however, were significantly lower for MGP (Corg: 32 ± 6 t ha-1; Nt: 3.3 ± 0.8 t ha-1) than for EGP (Corg: 45 ± 11 t ha-1; Nt: 4.2 ± 0.5 t ha-1). These grazing intensities did not show significant differences in Corg and Nt compared to IGP (Corg: 42 ± 11 t ha-1; Nt: 4 ± 0.8 t ha-1). For EGP, the Corg and Nt input due to the stimulating effects of grazing on photosynthetic rates (e.g. stimulation of root growth as well as tiller and leaf, inhibition of stem and flower, incorporation of dead plant material) may have been more pronounced than the destructive effects that grazing exerts on the sward, which is linked with losses of Corg and Nt (Frame and Laidlaw, 2011). For MGP, the growth stimulation and also incorporation of dead plant material due to trampling was presumably markedly less compared to EGP. Conclusion Extensive grazing pressure with a CPH of about 12 cm seemed to be the best management intensity for the studied pastures in a temperate climate, which established highest C org and Nt stocks in soil. References Conant R.T., Paustian K. and Elliot E.T. (2001) Grassland management and conversion into grassland: effects on soil carbon. Ecological Applications 11, 343–355. Derner J.D., Boutton T.W. and Briske D. (2006) Grazing and ecosystem carbon storage in the North American Great Plains. Plant and Soil 280, 77–90. Ellert B.H. and Bettany J.R. (1995) Calculation of organic matter and nutrients stored in soils under contrasting management regimes. Canadian Journal of Soil Science 75, 529–538. Frame J. and Laidlaw A.S. (2011) Improved grassland management. Marlborough UK: The Crowood Press Ltd. Jacobs A., Helfrich M., Hanisch S., Quendt U., Rauber R. and Ludwig B. (2010) Effect of conventional and Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 386 minimum tillage on physical and biochemical stabilization of soil organic matter. Biology and Fertility of Soils 46, 671–680. Şahin Demirbağ N.S., Röver K.-U., Wrage N., Hofmann M. and Isselstein J. (2008) Herbage growth rates on heterogeneous swards as influenced by sward-height classes. Grass and Forage Science 64, 12–18. Scimone M., Rook A.J, Garel J.P. and Sahin N. (2007) Effects of livestock breed and grazing intensity on grazing systems: 3. Effects on diversity of vegetation. Grass and Forage Science 62, 172–184. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 387 Effects of biomass of perennial grasses and legumes on soil carbon Skuodiene R. and Tomchuk D. Vezaiciai Branch, Lithuanian Research Centre for Agriculture and Forestry Gargzdu 29, Vezaiciai, Klaipėda distr., Lithuania Corresponding author: donata.tomchuk@yahoo.com Abstract Experiments were carried out with the aim assessing the effects of incorporation of perennial grasses and legumes: red clover (Trifolium pratense L.), white clover (Trifolium repens L.), alfalfa (Medicago sativa L.), timothy (Phleum pratense L.) as green manure on the amount of soil organic carbon. Ploughed-in biomass of red clover and alfalfa, second year of usage, provided 2.5 – 2.8 times more organic carbon than ploughed-in timothy, and 2.2 – 3.1 times more than from white clover. The amount of organic carbon depended on the plant's aboveground biomass and root mass (respectively r = 0.999** and r = 0.992**). The C:N ratio of the aboveground biomass of legumes was the lowest (10.0–14.0) and was more favourable for a rapid decomposition than of the aboveground and root mass of timothy (20.0–47.0). After two years of soil usage for crop cultivation, the amount of organic carbon in soil increased by 40.4-52.5% when timothy grass or legumes of the second year of usage were ploughed, and by 20.0–22.2% when grasses of the first year of usage were ploughed. At the end of the study the C:N ratio was favourable for humification of organic carbon. Keywords: perennial grasses, biomass, organic carbon, C:N ratio Introduction Residues of plant aboveground biomass and root biomass are the main source of soil organic carbon (Rasse et al., 2005). The rate and duration of accumulation of organic material partly depends on the initial amount of N and C in soil (Velthof and Oenema, 2001). In grassland ecosystems up to 98% of total C is sequestered below ground (in the rhizosphere) and sequestered carbon is less volatile than carbon from above ground (Schlesinger, 1977). Carbon and nitrogen mainly accumulate in the top 10 cm layer, where approximately 80% of plant roots are developed (Davies et al., 2001). The aim of this study was evaluate the biological value of biomass of perennial species on soil carbon changes in soils of western Lithuania. Materials and methods Field experiments were conducted in western Lithuania at the Vezaiciai branch of the Lithuanian Research Centre for Agriculture and Forestry (55º43/ N, 21º27/' E) in 2002-2007. Two analogous experiments were set up in 2002 and 2003. The soil of the experimental site was Albi – Edohypogleyic Luvisol, light loam on medium heavy loam. The agrochemical characteristics of the plough layer were as follows: pHKCl – 6.0-6.1, mobile P2O5– 104-199 mg kg-1 soil, K2O – 120-166 mg kg-1 soil, Ntotal 0.08-0.11%, Corg 0.90-1.05%. The experiments were conducted in the following crop rotation sequence: perennials – winter triticale (Triticosecale Wittm.) – spring barley (Hordeum vulgare L.). Perennials included: red clover, white clover, alfalfa, timothy. Perennial species were ploughed under as green manure at different stages of development, of and for first year (I) and second year (II) of usage. The biomass was chopped and shallowly incorporated during the beginning of flowering of legumes and at the beginning of ear emergence of timothy. After two weeks it was then deeply ploughed to 25 cm. No mineral fertilizers or plant protection products were used, in order to determine the biological value of the different preceding crops. Plant root mass was determined from 10 cm depth by the Katchinski monolith washing method. The mass of all plant residues and aboveground biomass were recalculated into dry matter. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 388 Soil samples were collected before the trial establishment and after ploughing the perennials to below the 0–20 cm depth. Available P2O5 and K2O were determined by the A-L method, total nitrogen by Kjeldahl, organic carbon by mineralizer ‘Heraeus’. Statistical analysis was carried out using ANOVA (** in tables indicates significance at P < 0.01). Results and discussion The amount of aboveground and root mass of the perennial species in the crop rotation depended largely on the biological properties of the plant species and the stage of development (Table 1). Aboveground mass of perennials from the first year of usage was higher than from the second year of the same species by, respectively: 10% for red clover, 33% for white clover, and 45% for timothy. The smallest amount of aboveground mass originated from the timothy of second year. Table 1. Dry matter yield of biomass, organic carbon and C:N ratio of perennial species in first year (I) and second year (II) of usage Perennials Timothy I Timothy II Red clover I Red clover II White clover I White clover II Alfalfa II LSD0.05 Dry matter yield Aboveground mass Root mass g m-2 g m-2 403.7 721.3 223.7 1016.7 1904.5** 660.0 1714.2** 1139.3** 898.89** 374.2 603.5 311.3 1051.0** 1799.4** 317.861 355.558 Amount of organic carbon and C:N Aboveground mass Root mass g m-2 C:N g m-2 C:N 152.2 26 223.8 47 86.4 23 312.7 36 688.5** 12 233.6 20 637.1** 14 416.3* 22 323.0* 14 130.2 23 231.0 14 108.1 24 372.2 10 627.0** 27 163.232 147.575 The greatest amount of root biomass was from perennials of the second year of usage (with the exception of white clover) and the greatest amounts were from alfalfa and red clover, second year of usage. Root biomass from alfalfa and red clover accounted, respectively, for 63.1 and 39.9% of the biomass used as green manure. The smallest amount of biomass (915 g m-2 of dry matter) was produced by white clover of the second year of usage. Their roots constituted 34% of the biomass used as green manure. The general amount of accumulated organic carbon in biomass of the tested perennials was found to be 35.2-38.6% in aboveground part and 29.537.0% in the roots. The greatest amount of organic carbon was incorporated into the soil by ploughing of red clover (1053 g m-2) and alfalfa (999 g m-2) of the second year of usage. It was 2.5 – 2.8 times more than by ploughing of timothy for different development stages, and 2.2 – 3.1 times more than by ploughing white clover (Table 1). The amount of organic carbon depended on the plant's aboveground biomass and root mass (respectively r = 0.999** and r = 0.992**). Carbon and nitrogen ratio (C:N) varied between plant species and between biomass origin (aboveground part or roots). Clover root biomass C:N ratio was similar, but significantly smaller than that of timothy. Before the experiments, soil analysis showed organic carbon amounts were from 0.90 to 1.05% (Table 2). At the end of the study, after cultivation of triticale and spring barley, the amount of organic carbon increased by up to 40.4 – 52.5%, when perennials of second year of usage were ploughed under, and by 20.0 – 22.2% after ploughing the perennials from first year of use. Maximum Corg increase was delivered from alfalfa and red clover. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 389 Table 2. Changes in organic carbon in soil (0-20 cm) for first year (I) and second year (II) of perennial species (T – timothy, RC – red clover, WC – white clover, A – alfalfa). Parameter of soil Corg before experiment the C:N Corg at the end of experiment C:N Perennial species RC II WC I TI T II RC I WC II A II 0.95±0.07 1.04±0.08 0.95±0.04 1.04±0.02 0.90±0.03 1.05±0.02 1.01±0.03 9.5 10.4 9.5 10.4 9.0 11.7 10.1 1.14±0.15 1.46±0.20 1.16±0.22 1.53±0.05 1.10±0.18 1.53±0.10 1.54±0.17 9.5 11.2 10.5 10.1 9.2 12.7 11.0 Prior to the experiments, the C:N ratio in the soil arable layer was 9.0 – 11.7. After two years the C:N ratio increased where white clover, alfalfa and timothy of second year of usage were ploughed under (respectively by 8.5, 8.9 and 7.7%) and red clover of the first year of usage (10.5%). This resulted because of an increase of organic carbon after cereal cultivation. Conclusions The tilled-in biomass of perennial grasses considerably increase the amount of organic carbon in soil and led to its accumulation in the topsoil layer. Maximum amount of sequestered organic carbon was observed after tilling-in of red clover (1053 g m-2) and alfalfa (999 g m-2) of second year of usage, which was 2.5-2.8 times greater than from timothy, and 2.2 to 3.1 times more than from white clover. After two years of crop cultivation the soil organic carbon content increased by 40.4-52.5% when second year perennials were ploughed under, and by 20.022.2% when the first year perennials were ploughed under. At the end of the research, the C:N ratio in the arable soil layer was favourable for humification of organic carbon in all variances. Acknowledgments The study was conducted in accordance with the long-term programme 'Plant biopotential and quality for multifunctional practice'. References Davies M.G., Smith K.A. and Vinten A.J.A. (2001) The mineralization and fate of nitrogen following ploughing of grass and grass-clover swards. Biology and Fertility of Soils 33, 423–434. Rasse D.P., Rumpel C. and Dignac M.F. (2005) Is soil carbon mostly root carbon? Mechanisms for a specific stabilisation. Plant and Soil 269, 341-356. Schlesinger W.H. (1977) Carbon balance in terrestrial detritus. Annual Review of Ecology and Systematics 8, 51– 81. Velthof G.L. and Oenema O. (2001) Effects of ageing and cultivation of grassland on soil nitrogen. Alterrarapport 399, 1–53. (http://edepot.wur.nl/26458) Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 390 Soil organic carbon characteristics under different intensities of grassland management Karabcová H., Mičová P. and Fiala K. Agroresearch Rapotin, Rapotin, Czech Republic Corresponding author: h.karabcova@centrum.cz Abstract The effect of intensive and extensive grassland management on hot-water soluble carbon (Chws) and water soluble carbon (Cws) was monitored. The experimental plots were located at 400 m above sea level on Cambisol soil type in the Czech Republic. Three treatments, each applying mineral and organic fertilizers with graded stocking rates and different mowing frequency, were conducted on the permanent grassland. Average Chws values ranged from 573 to 657 mg kg-1 and Cws values ranged from 210 to 258 mg kg-1. Hot-water soluble and water soluble carbon were influenced by different grassland management. The greatest amounts of Chws and Cws were measured under the 'medium intensive' management. Intensive management led to a decline in the amount of Chws and Cws. Keywords: hot-water soluble C, water soluble C, grassland, Cambisol Introduction Soil organic carbon is one of the most important parts of the soil due to its ability to affect plant growth, both a source of energy and a trigger for nutrient availability through mineralization (Edwards et al., 1999). The organic carbon comprises an easily degradable part which can be expressed as hot-water soluble carbon (Chws), and water soluble carbon (Cws). These properties are known as indicators of the amount of available soil C substrate. These properties are studied for their usefulness as soil quality indicators responding to changes in the rhizosphere caused by management practices (Ghani et al., 2003). Both parameters can be also influenced by soil type (Uchida et al., 2012) and by altitude (Kolář et al., 2003). Previous studies have reported about Chws and Cws changes and dynamics influenced by grazing (Haynes, 2000), fertilization (Ghani et al., 2003), mowing and mulching (Váchalová et al., 2013) in grasslands. The aim of this study was to evaluate the quantitave changes of soil organic carbon under different levels of grassland management (extensive, medium-intensive, intensive). Materials and methods The experimental plots were located in the northwest part of Moravia in the Czech Republic. The area is situated 400 m above sea level and the soil is sandy-loam, type Cambisol. Annual average air temperature is 7.7 °C and annual rainfall average is 693 mm. The locality is characterized by semi-natural permanent grassland. Three treatments, each applying 3 rates of mineral and organic fertilizers, were conducted on the permanent grassland as shown in Table 1. Experimental plots were arranged in a completely randomized block design with four replications. The plot size was 12.5 m2. In 2013 soil samples (from a depth 0.05–0.30 m) were taken to determine the content of hot-water soluble carbon Chws (Körschens et al., 1990) and water soluble carbon Cws (Váchalová et al., 2013). Statistical data analysis was undertaken using the statistical program SPSS 13.0 for Windows. Analysis of variance (ANOVA) was performed by statistical program. Mean statistical differences (95% significance level) were calculated by Tukey’s HSD test (P < 0.05). We tested homogeneity of variances by Cochran–C’s test. Correlation analysis was also used. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 391 Table 1. Experimental treatments Treatment Fertilization Livestock unit (LU) Amount of applied N No. of cuts Control without 0 0 2 cuts per year Extensive mineral and organic 0.9 LU ha-1 60 kg N ha-1 2 cuts per year Medium-intensive mineral and organic 1.4 LU ha-1 90 kg N ha-1 3 cuts per year mineral and organic -1 Intensive 2.0 LU ha 120 kg N ha -1 4 cuts per year Results and discussion The effect of intensive and extensive grassland management on Chws and Cws was monitored. The measured values of Chws and Cws were optimum for this soil type. Kolář et al. (2003) reported that optimum value of Chws ranged from 300 to 600 mg kg-1. The values of Cws were positively correlated with the values of Chws. Chosen forms of carbon content showed differences between the treatments (Table 2). Table 2. Average values of Chws and Cws for the particular treatments. a, b, c - values with the same letter are not statistically different within the column P < 0.05 Treatment Chws (mg kg-1) Cws (mg kg-1) Control 573a 210a Extensive 642b 238b Medium-intensive 657c 258c Intensive 637b 234b P 0.035 0.027 The differences in Chws and Cws were significant between control and all fertilized treatments. There were statistically significant differences in Chws and Cws between 'medium intensive' (3 cuts per year, 90 kg N ha-1) treatment and all the other treatments. The content of Chws was higher at the treatment with 'extensive' (2 cuts per year, 60 kg N ha-1) management in comparison to 'intensive' (4 cuts per year, 120 kg N ha-1) management but this trend was not statistically significant (P = 0.786). Greater intensity of grassland use (multiple mowing) led to a moderate decrease in Chws and Cws. Duffková et al. (2005) also reported about Chws and Cws decrease in connection with a higher frequency of mowing. These results corresponded with findings of Ghani et al. (2003) who published that intensive grassland management had negative impacts on Chws content. The greatest content of selected forms of soil carbon was measured at the 'medium intensive' treatment. Intensive management showed decrease of soil carbon. Conclusion Our study confirmed a positive effect of 'medium intensive' management on selected forms of soil organic carbon. Hot-water soluble and water soluble carbon were influenced by the different types of grassland management. The greatest amount of hot-water soluble and water soluble carbon was recorded under the 'medium intensive' level of management. Intensive management led to a decline in the amount of hot-water soluble and water soluble carbon. Acknowledgements This study was supported by the Ministry of Education, Youth and Sports, Czech Republic, project LG13019. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 392 References Duffková R., Kvítek T. and Voldřichová J. (2005) Soil organic carbon and nitrogen characteristics in differently used grasslands at sites with drainage and without drainage. Plant Soil Environment 51, 165-172. Edwards J.H., Wood C.W., Thurlow D.L. and Ruf M.E. (1999) Tillage and crop rotation effects on fertility status of a Hapludalf soil. Soil Science Society of America Journal 56, 1577-1582. Ghani A., Dexter M. and Perrot K.W. (2003) Hot-water extractable carbon in soils: a sensitive measurement for determining impacts of fertilisation, grazing and cultivation. Soil Biology and Biochemistry 35, 1231-1243. Haynes R.J. (2000) Labile organic matter as an indicator of organic matter quality in arable and pastoral soils in New Zealand. Soil Biology and Biochemistry 32, 211-219. Kolář L., Klimeš F., Ledvina R. and Kužel S. (2003) A method to detemine mineralization kinetics of a decomposable part of soil organic matter in the soil. Plant Soil Environment 49, 8-11. Körchens M., Schultz E. and Behm R. (1990) Hot water extractable carbon and nitrogen of soils as criteria of their ability for N-release. Zentralblatt für Mikrobiologie 145, 30-311. Uchida Y., Nishimura S. and Akiyama H. (2012) The relationship of water-soluble carbon and hot-water soluble carbon with soil respiration in agricultural fields. Agriculture, Ecosystems and Environment 156, 116-122. Váchalová R., Kolář L., Kobes M. and Váchal J. (2013) The effect of grassland management practises on differentiation of soil organic matter fractions. Advanced Crop Science 3, 472-478. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 393 Efficacy of the agrosteppe method for restoring eroded lands Dzybov D.S.1 and Starodubtseva A.M.2 1 Department of Field and Grassland Forage Production, Stavropol Agricultural Research Institute, Nikonova 49, Mikhailovsk, the Stavropol Territory, Russia, 356241 2 Department of Crop Production and Grassland Ecosystems, Russian State Agrarian University – Moscow Timiryazev Agricultural Academy, Timiryazevskaya 49, Moscow, Russia, 127550 Corresponding author: anastasia.starodubtseva@gmail.com Abstract The steppes have been experiencing severe anthropogenic stress for thousands of years, and they are largely used for arable farming. The growing concern for biodiversity led to the necessity of developing a habitat restoration method which would result in establishment of a plant community totally equivalent to the initial one. The natural process takes up to 150 years, including the undesirable primary stages of forbs’ association featuring an excessive propagation of the hazardous allergenic species Ambrosia artemisiifolia and other ill weeds. Sod planting has been found ecologically inadequate, because of the uneven sward and strong competition between the sod blocks and the seedlings of the new generation. However sowing a complete species mixture after ploughing was proved efficient. The problem of seed collection conditioned by the differences in ripening time of the target species was solved by harvesting the equal adjacent areas of seed donors in two or three stages at 25-30 days intervals. This technology and the swards created were termed ‘agrosteppe’. The typical plant community obtained as a result was stable and self-reproducing already in three years after establishment. Introduction Rehabilitating completely destroyed multicomponent natural ecosystems, such as steppes, meadows, prairies and savannas, takes 80-150 years of natural self-remediation via succession (demutation). The recovered steppes (grasslands, etc.) should be similar to the natural ecosystems in their composition and flora abundance, as well as in their vertical structure, resistance to anthropogenic stress, economic and aesthetic features such as biological productivity, functionality, and wildlife well-being. On agriculturally used sites this results in similarity to the properties of natural hay and pasture and to the dietary characteristics of the resulting animal products. Materials and methods The experimental site of 1100 m2 (11 ha) is located on eroded lands in the Stavropol Territory of Russia. It is situated on the 2-6° southern slope of undulating terrain. The sandy soil contains only 0.9-1.0% humus; annual precipitation is uneven, up to 450-500 mm. The land was intensively used as a pasture. After the degradation of the natural steppe vegetation, the site was converted into arable area for growing crops such as wheat, maize, sunflower, and grasses for hay. Previously to agrosteppe establishment, the territory was abandoned and inhabited by weed communities with forbs prevailing: Daucus carota, Ambrosia artemisiifolia and Artemisia vulgaris. The height of the upper layer reached 1.5 m and surface coverage amounted to 60%. The direct census was carried out in accordance with Oscar Drude’s plant abundance scale on fixed 100 m2 sites. The four Poaceae species made 18% of the plant community; the only two Fabaceae representatives amounted to 9%, and the 16 forbs species dominated at 73%. The latter group was mostly represented by harmful weeds, being a hazard both for the crops and for human well-being (Ambrosia artemisiifolia). The agrosteppe was established in July 1980. The method has this name because the steppe is reconstructed using conventional Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 394 agricultural machinery, tools and techniques for tillage, sowing and harvesting the complex natural seed mixture. Results and discussion During the first year after the sowing of the agrosteppe, the typical wildland species were accompanied by approximately 60% of weeds. In the next year, 1982, the diverse steppe perennials successfully competed with them and finally dominated in the swards. Since the third year on the agriculturally created young steppe the sward has been sustainably reproducing itself without human interference. Single forbs are found only on the disturbed sites near animal holes and burrows. The stabilized botanical portrait of the agrosteppe (Table 1) is mainly represented by the grass association: Festuca valesiaca + Stipa pulcherrima + Filipendula vulgaris. The upper layer is 0.75 m high, the dominant layer is 0.50 m, and the ground layer is at 0.30 m. The surface coverage is 90-100%. The plant diversity of the agrosteppe has reached 67 species, compared with the 22 of the degraded land. The species abundance is subject to slight natural fluctuations in accordance with the changeable weather conditions resulting in so called ‘clover years’ or ‘Stipa years’, and so on. Like a natural steppe community the established one shows considerable tolerance to the pyrogenic effects and revegetates the next year. In addition, since the very first years of existence, agrosteppe is inhabited by animal species, including pollinators (Apis spp., Bombax spp., etc.), birds, reptiles, rabbits, and later by foxes, and others. Thus, both the phytocoenosis and the zoocenosis of the steppe ecosystem are reproduced simultaneously. Table 1. Botanical composition of the fully developed agrosteppe (3rd year of life and on) Poaceae + Cyperaceae (19.4%) 1. Bromopsis riparia 2. Carex michelii 3. Cleistogenes bulgarica 4. Dactylis glomerata 5. Erytrigia intermedia 6. Festuca pratensis 7. Festuca rupicola 8. Festuca valesiaca 9. Koeleria cristata 10. Phleum phleoides 11. Poa bulbosa 12. Stipa pennata 13. Stipa pulcherrima Fabaceae (9%) 14. Amoria ambigua 15. Amoria montana 16. Amoria repens 17. Astragalus austriacus 18. Lotus caucasicus 19. Medicago minima 20. Medicago romanica 21. Onobryhis arenaria 22. Trifolium alpestre 23. Trifolium pratense 24. Vicia angustifolia Sp3 Sp2 Sp1 Sp1 Sp1 Sp2 Sp3 Cop3 Cop1 Sp3 Sp3 Sp1 Cop2 Sp3 Cop1 Sp1 Sp2 Sp2 Sp2 Sp3 Sp2 Sp3 Sp2 Cop1 Forbs (73%) 26. Achillea nobilis 27. Achillea setacia 28. Arenaria serpyllifolia 29. Dianthus ruprechtii 30. Echium russicum 31. Filipendula vulgaris 32. Geranium sanquineum 33. Hieracium echioides 34. Plantago lanceolata 35. Plantago media 36. Veronica austriaca 37. Ranunculus polyanthemos 38. Stachys atherocalyx 39. Salvia tesquicola 40. Thymus marshallianuc 41. Potentilla argentea 42. Potentilla adenophylla 43. Jurinea arachnoidea 44. Potentilla recta 45. Fragaria viridis 46. Euphorbia sequieriana 47. Euphorbia iberica 48. Euphorbia stepposa 49. Cerastium caespitosum 50. Verbascum phoeniceum Sp2 Cop1 Sp2 Sp3 Sp3 Cop2 Cop1 Sp3 Sp3 Sp3 Sp3 Sp2 Sp3 Sp1 Sp3 Sp3 Sp3 Sp3 Sol Sp3 Sp2 Sp1 Sp2 Sp3 Sp2 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 395 25. Vicia tenuifolia Sp3 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. Centaurea orientalis Tragopogon dasyrhynchus Orchis tridentata Veronica verna Muscari muscarimi Salvia verticillata Verbascum lychnitis Anemone sylvestris Dracocephalum austriacum Paeonia tenuifolia Myositis suaveolens Scabeosa ochroleuca Eryngium campestre Poterium polygamum Leontodon hispidus Gallium ruthenicum Pastinaca pimpinellifolia Sp3 Sp3 Sp2 Sp2 Sp3 Sp2 Sp1 Sp1 Sp3 Sp2 Sol Sp1 Sp1 Sp3 Sol Sp3 Sol Conclusion The established agrosteppe is perennial like that of the natural zonal steppe, and it is likewise self-sustaining without any human intervention. It becomes a seed donor for ecological restoration of unproductive desertified lands: 1 ha of the donor steppe provide for 7-10 ha of resown territory. This allows recovering low-productive and abandoned land, and reproducing the floral diversity of steppe exponentially. In two years the weed community is replaced by a typical diverse steppe vegetation with approximately 67 species per 100 m2, including valuable forage grasses (Festuca, Koeleria, Bromopsis, Dactylis, Phleum, etc.), legumes (Trifolium, Amoria, Medicago, Onobrychis, Lotus, etc.), and forbs (Plantago, Poterium, Filipendula, etc.). The created sward contains a number of medicinal plants (Thymus, Fragaria, Achillea, etc.), and conditions the propagation of rare and protected genera such as Stipa, Anemone, Paeonia, and Orchis. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 396 Zonal strategy for sward renovation by total reseeding based on research results Ene T.A.1, Mocanu V.1, Mocanu V.2, Ciopata A.C.1 and Cardasol V.1 1 Research-Development Institute for Grassland, Brasov, Romania 2 National Research-Development Institute for Soil Science, Agrochemistry and Environment, Bucharest, Romania Corresponding author: vasmocanu@yahoo.com Abstract In recent years, the Research-Development Institute for Grassland, Brasov has carried out experiments in the Fagaras depression, a large area in the centre of Romania. The purpose of the research purposes has been to find the most suitable solutions for improving degraded grasslands in accordance with the area conditions and a lower environmental impact. Based on research results, a zonal strategy for sward renovation using a total reseeding technique was established and extended on to larger areas. Keywords: grassland, research solutions, reseeding, Fagaras depression Introduction Grasslands are an essential element of sustainable farming systems, in terms of animal welfare, fodder provision, soil quality and the best use of less-productive land. The experimental fields were located in representative areas of the Fagaras depression, allowing a high degree of extrapolation of research results. Investigation has focused on establishing a favourable interaction between crop and livestock systems in this area, with good results in terms of natural-resource use (Mocanu et al., 2011).This region covers an agricultural area of over 111,000 ha, of which 49,925 ha (45%) are grasslands, at altitudes of 450–650 m above sea level (a.s.l.). There are especially mixed farms (crops and livestock), the most economic farming system way in the current conditions. Except for eutric and carbonatic soils located in the Olt Floodplain, the other soils in the Fagaras Depression (about 110,000 ha, or 90% of the basin area) have, in addition to nutrient deficiencies, some limitations induced by high acidity and large amounts of exchangeable Al3+ etc. The forages of permanent grassland are poor in terms of yield and quality because of different stages of degradation, with multiple reasons for this. Therefore, to achieve high yields with high nutritional value of feed and high conversion efficiency in livestock products, it is necessary to apply measures to improve these pastures. Materials and methods The area under intervention is located in the Dragus village, Fagaras Depression, which is property of the local association of animal breeders. It is on a permanent pasture, 25 ha in area, at 505 - 530 m a.s.l. The area is in an advanced stage of degradation caused by a bad management (absence of annual clearing and maintenance work, invasion of worthless species, irrational grazing, and no fertilization or correction of soil acidity etc.). The soil agrochemical study highlights an area with high acidity and with a poor nutrients supply. Based on this soil agrochemical study, research results and the current area conditions, a zonal strategy for sward renovation by total reseeding has been established. The works applied to improve these degraded grasslands were as follows:  liming 4 – 6 t ha-1 agricultural limestone, in autumn 2012;  total destruction of the old sward by heavy disc harrow (two perpendicular passes) in the autumn of 2012;  seedbed preparation by disc harrow (two passes perpendicularly) in the spring of 2013; Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 397  seeding with valuable perennial grass and legume mixtures, simultaneously with rolling before and after sowing, using a special grass sowing machine made by the ResearchDevelopment Institute for Grassland, Brasov (Mocanu and Hermenean, 2013);  fertilization with complex fertilizers N15P15K15, 350 kg ha-1. For tilling, the disc-harrow alternative was chosen due to the thickness of fertile topsoil. Because the improved grassland is to be used in a mixed system, grazing and cutting, a complex seed mixture was selected. This seed mixture is suitable for local area conditions and consists of (see Mocanu et al., 2013): Festuca pratensis 8 kg; Festuca arundinacea 13 kg; Dactylis glomerata 5 kg; Trifolium pratense 2 kg; Trifolium repens 2.5 kg; Lotus corniculatus 3.5 kg; in total 34 kg seed ha-1. Results and discussion Because of the terrain topography, the reseeded area is divided into three plots, as follows: 18 ha in plot I, 5.5 ha in plot II and 1.5 ha in plot III. There was good establishment of the forage crop, with good ground cover of sown species, average of cover was from 88% on plot I, to 92% on plot II (Table 1). The average feed yield after the first harvest by mowing the three plots was 3.9-4.3 t DM ha-1, while the output of the control plot was 1.0-1.3 t DM ha-1. The second cycle consisted of using cattle grazing, average yield ranging between 1.4 – 1.7 t DM ha-1. Therefore, in the first year of the sward establishment, total feed yield of improved pasture was between 5.3 and 6.0 t DM ha-1. Table 1. Percentage cover and botanical composition of sown and control plots. Species Percentage cover of sward Grasses, of which: Dactylis glomerata Festuca arundinaceea Festuca pratensis Festuca rubra Agrostis tenuis Poa pratensis Nardus stricta Forage legumes, of which Trifolium pratense Trifolium repens Lotus corniculatus Other grasses TOTAL Sown plot, participation rate, % I II III 88 90 92 74 53 35 43 30 25 12 12 5 7 5 3 6 3 1 6 3 1 23 45 64 11 30 45 5 6 10 7 9 9 3 2 1 100 100 100 Control plot, % 100 92 10 60 10 12 8 100 Some soil samples for each variant (plot I and II with three replicates, and plot III with two replicates) were analysed before (control plot) and after intervention. As a result of liming, strong acid reaction of the soil has improved noticeably, especially in the 0-10 cm depth, soil pH status increasing on average of 1.27 units, from 4.74 (high acidic) to 6.01 units (low acidic) in all three plots. After fertilizer application, there was an increase of more than 3 times in the phosphorus content (6-19 mg kg-1 PAL) in the 0-10 cm depth. In addition to some assessments of botanical composition (Table 1) there were determinations of the main nutritional parameters (Table 2): crude protein, crude fibre, cell wall constituents (NDF, ADF, ADL), crude ash and dry matter digestibility (DMD) and organic matter (OMD), by infrared spectroscopy technique (NIRS). Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 398 Table 2. The main nutritional parameters of feed obtained on plots, % DM. No. Variant Crude Protein % Crude Ash % Crude fibre % NDF % ADF % ADL % DMD % OMD % 1 Control plot 8.60 7.30 40.80 73.90 45.00 4.20 45.20 44.20 2 Plot I, average 15.54 9.46 29.13 57.93 33.10 3.33 69.43 66.26 3 Plot II, average 16.36 9.70 29.26 56.56 33.13 3.70 66.76 63.06 4 Plot III, average 18.25 10.65 25.80 52.40 30.90 3.55 71.45 68.90 5 Total, average 16.72 9.94 28.06 55.63 32.38 3.53 69.21 66.07 Crude protein content (CB %) was in the range 15.54-18.25%, which is characteristic of quality forage, while the control plot had small protein content (8.6%). The constituent content of cell walls: NDF, ADF, ADL had values characterizing a forage with medium to high feeding value. The digestibility coefficients of the analysed forage samples have presented also values for a good quality of forages provided by reseeded pastures. Conclusion The establishment of productive grassland depends greatly on providing the right conditions for seed germination, and for seedling and root growth, with the final aim to develop a dense sward. For a successful action, it is necessary to remove the reasons that have caused the grassland became of low quality. To know these reasons, several analyses concerning the stationary area conditions, soil characteristics, botanical composition of the old vegetation, and the climatic conditions are carried out. Further, based on analysis and research results, the best solution for improving the degraded grassland needs to be decided, taking into account the following factors: appropriate tillage system for destruction of the old vegetation, reseeding period, seedbed preparation, basic fertilization, seed mixture choice, seeding machinery and post-sowing management. References Mocanu V., Cardaşol V., Hermenean I. and Mocanu Victoria (2011) Tehnologie modernizată pentru culturile specifice Zonei Făgăraş în vederea instalării stării de agroclimax, Editura Capo-Lavoro, Braşov, România, 58 pp., ISBN 978-973-98711-9-8. Mocanu V. and Hermenean I. (2013) Mecanizarea lucrărilor agricole pe pajişti-Tehnologii, maşini şi echipamente, Editura Universităţii Transilvania, Braşov, România, 416 pp., ISBN 978-606-19-0237-8. Mocanu V., Cardaşol V., and Maruşca T. (2013) Zonal solutions for sustainable farming systems by restoring the multifunctionality of grasslands. 17th Symposium of European Grassland Federation, Akueyri Iceland, June 2013, Book of abstracts, p. 85. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 399 Magnesium content in soil and selected layers of upland grassland biomass Grygierzec B., Kasperczyk M. and Szewczyk W. Department of Grassland Management, Institute of Plant Production, University of Agriculture in Krakow, al. Mickiewicza 21, 31–120 Krakow, Poland Corresponding author: w.szewczyk@ur.krakow.pl Abstract Investigations were aimed at determining the magnesium (Mg) content in soils and selected layers of plants, depending on the management method (hay production, grazing) and nitrogen fertilization level with basic PK fertilization. A four-year period of management and fertilization resulted in lowered content of available Mg in soil samples collected from all objects. Increasing nitrogen fertilization led to elevated Mg concentration in the soil of the experiment used for hay production, whereas an opposite relationship was revealed in the experiment with grazing. Mineral fertilization caused the highest increase in Mg concentration in the harvested plants. The lowest extractable Mg was in the root layer between 0 and 5 cm for the N120P18K66 treatment in both experiments. Keywords: hay, grazing, fertilization, magnesium content Introduction Magnesium deficiency poses a serious problem in soils and plants of permanent grasslands (Szewczyk et al., 2007). The highest deficit of this macroelement occurs in acid soils developed from loose and weakly clay-sands, and in shallow, skeletal soils (Sapek, 2008), usually covered with grass vegetation used for pastures. The Mg content in soil and sward herbage is affected by nitrogen fertilization (Kulczycki, 2006). The research aimed at determining the Mg concentration in soils and plants of grasslands depending on the methods of their management and nitrogen fertilization level, against the background of phosphorus and potassium treatment. Materials and methods The research comprised two experiments conducted in 2007–2010. The experiments were located on a permanent grassland (49o 47' 23'' N; 19o 50' 47'' E; elevation 470 m a.s.l.), on acid brown soils developed from sandstone with the composition of light loam. The first experiment was used for hay production and pasture, in which the first regrowth was cut at the full-earing stage of the dominating grass species, and the other two regrowths were grazed by a flock of mountain sheep. The other experiment was utilized as pasture, grazed by mountain sheep four times every year. The grazing was in quarters and lasted for three days for the first and second regrowths, and two days for the third and fourth regrowths. A randomized block design with four replications was used. Two fertilizer variants were considered in the experiments according to the scheme in Table 1. Control treatments were cut and grazed but did not receive mineral fertilizers. Dry matter yield of the sward used for hay production was assessed by cutting plants from the 12-m2 plots, whereas the sward used as pasture was cut from an area of 1 m2 of each plot prior to each grazing. The plant samples from the replications were mixed and dry-mineralized in a muffle furnace at 450 OC. Magnesium was extracted by hot dilution in nitric acid (1:1). The contents of Mg in plant material were assessed by the ICP-AS method on JY 238 Ultrace. Prior to the onset of the experiment and after its completion, turf samples were collected from an area of 30 × 30 cm area and to a depth of 15 cm from all treatments of both experiments in order to determine the aboveground plant (stubble) mass, root mass and its distribution in the 0–5 cm and 5–15 cm layers. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 400 Table 1. Mineral fertilization scheme Management kg N ha-1 (doses before follow regrowths) kg P ha-1 (single dose in spring kg K ha-1 (doses before follow regrowths) 18 66 (33 + 33) 18 66 (33 + 33) 80 (48 + 32) Mixed 120 (60 + 30 + 30) 80 (20 + 20 + 20 + 20) Grazing 120 (30 + 30 + 30 + 30) The stubble layer consists of plants remaining after grazing and cutting to the height of 5 cm above ground. The following assessments were made in the sampled soil material: granular size composition by sieve method, 1 mol dm-3 KCl by potentiometer, organic carbon by Tiurin method modified by Oleksynowa, total nitrogen by Kjeldahl method using Kjeltec apparatus, available P and K by Egner-Rhiem AL-method, available Mg by atomic absorption spectrometry AAS after extraction in 0.0125 mol CaCl2 dm-3. Obtained results were verified statistically by means of ANOVA using Statistica 6.0 application. Significance of differences was verified by means of Tukey’s test on the confidence level of 0.05. Results and discussion After the four–year period of management and fertilization of the experimental treatments, an increase of P and K content by a decrease of assimilable Mg was registered in soils of all treatments (Table 2). Table 2. Some properties of soil analysis 4.1 Org. matter C g kg-1 soil 24.7 Control 4.2 20.8 N80P18K66 4.0 17.3 N120P18K66 3.9 LSD 0.05 Variant pH in KCl Initial state State after 4 years Total N 2.6 Experiment I 3.4 P 18 K mg kg-1 soil 61 Mg 42 9 52 40 4.0 15 49 36 18.5 3.7 12 43 32 n.s. 1.3 n.s. 3 6 7 Control 4.1 25.0 3.1 10 46 40 N80P18K66 4.1 17.2 4.0 12 58 31 N120P18K66 4.2 18.7 3.5 16 60 37 LSD 0.05 n.s. 2.8 n.s. 4 9 5 Experiment II n.s. – not significant Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 401 Harvested plants I Stubble II Root mass of 0–5 cm layer III Root mass of 5–15 cm layer IV 0 A) 0.5 1 1.5 Control N80PK Control N80PK 2 2.5 3 2.5 3 N120PK Variability coefficients: I ) V–7% II) V–17% III) V–4% IV) V–5% Harvested plants I Stubble II Root mass of 0–5 cm layer B) Root mass of 5–15 cm layer III IV 0 0.5 1 1.5 2 N120PK Variability coefficients: I) V–10% II) V–15% III) V–11% IV) V–4% Figure 1. Mean content of Mg (g kg-1) in the dry mass of the meadow and pasture flora in layers (A: Experiment I; B: Experiment II). In the case of Mg this negative phenomenon may result from the more intensive utilization and, at the same time, greater soil exhaustion of available Mg. Grygierzec et al. (2007) found similar dependencies. The method of management and level of nitrogen fertilization affected Mg content in the plant material (Figure 1). Harvested plant mass was most abundant in this macroelement. Under cutting and pasture management the Mg concentration in harvested plants increased with the increase in rate of fertilization. Similar quantities of Mg in whole plant material might have been influenced by its availability. The experiments were located on soil that was initially poor in potassium. As reported by Barszczewski and Ducka (2012), potassium and magnesium antagonism in uptake by grassland plants was not registered despite the fertilization with potassium. According to Sapek (2008) acidity of soils may significantly alter Mg uptake by meadow plants. Conclusions After a four-year period of utilization, lower content of assimilable magnesium was found in the soil in relation with its initial state. Harvested plants were the most abundant in magnesium, whereas a shallower root mass revealed the least Mg quantities. Nitrogen fertilization increased the content of magnesium in harvested plants, while it decreased in the stubble and in root mass of the 0–5 cm layer in both experiments. The content of magnesium detected in the root mass of the 0–5 cm and 5–15 cm layers under pasture management was lower than in the case of combined utilization (hay production and grazing). References Barszczewski J. and Ducka M. (2012) Management of potassium and quality changes of sward from irrigated meadow during the long-term study. Journal of Research and Applications in Agricultural Engineering 57(3) (in Polish with English summary). Grygierzec B., Sołek–Podwika K. and Radkowski A. (2007) Contents of magnesium in soils and sward of mountain grasslands depending on the land use and nitrogen fertilization level. Zesz. Prob. Post. Nauk Roln. Warszawa, z. 520, 611-617 (in Polish with English summary). Kulczycki G. (2006) The influence of different potassium and nitrogen fertilization on the plant yield and light soil properties. Zesz. Nauk. UP we Wrocławiu, Roln. LXXXIX, nr 546, 229-236 (in Polish with English summary). Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 402 Sapek B. (2008) Potassium to magnesium ratio in meadow vegetation and soil as an indicator of the environmental changes in grasslands. Woda-Środowisko-Obszary Wiejskie Falenty, T. 8, z. 2b, 139-151 (in Polish with English summary). Szewczyk W., Kasperczyk M. and Kacorzyk P. (2007) Effects of fertilization scheme on grassland production and water environment. Grassland Science in Europe 12, 359-362. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 403 Effects of previous cropping and establishment method on mineral concentration of whole-plant spring wheat Fychan R., Scott M.B., Davies J.W., Crotty F.V., Sanderson R. and Marley C.L. Institute of Biological, Environmental and Rural Sciences (IBERS), Aberystwyth University, Gogerddan, Ceredigion, Wales, SY23 3EE, United Kingdom. Corresponding author: rhun.fychan@aber.ac.uk Abstract An experiment tested the null hypothesis that previous cropping and establishment methods would alter the mineral concentration of whole-plant spring wheat (WPSW) (Triticum aestivum). Four replicate plots (12 × 7.5 m) of perennial ryegrass (Lolium perenne), chicory (Ch) (Cichorium intybus), red clover (RC) (Trifolium pratense) and white clover (WC) (Trifolium repens) forages were established as pure swards in June 2009 in a randomized block design. Forages were harvested 5 times per annum during 2010-2012 and forage removed. Forages were sprayed using a non-selective herbicide before each plot was split, and spring wheat was sown in April 2013 either after ploughing or by direct drilling. Whole-plant spring wheat was sampled on 18 June for chemical analysis. Results showed previous cropping affected WPSW P, Mn and Cu concentrations. Lower N, P, K, S and Cu concentrations were found in WPSW after ploughing when compared with direct drilling. Within direct-drilled plots, WPSW following RC had higher Ca concentrations compared to other previous forage treatments. Within ploughed plots, WPSW following RC had a lower Mg concentration than Ch and WC. Overall, findings showed previous crop and establishment method alter mineral composition of WPSW but results varied for each mineral. Keywords: spring wheat, chicory, minerals, trace elements, direct drill Introduction Some agricultural forages contain higher concentrations of minerals than ryegrass (Fisher and Baker, 1996) which may help to develop sustainable approaches to the management of these nutrients within agricultural systems. For example, chicory (Cichorium intybus) contains higher mineral and trace element concentrations than ryegrass (Marley et al., 2013a;b). There has been extensive work on the effects of previous legume crops on subsequent cereal crops but much of this research focused on N utilization. Cultivation is known to change the availability of nutrients within soil, with previous research showing differences amongst cultivation techniques in their effects (Canell et al., 1980). However, there has been little work into the effects of these previous crops on a subsequent cereal crop with regards to other essential minerals. Here, an experiment tested the null hypothesis that previous cropping and establishment methods would alter the mineral composition of spring wheat. Materials and methods Four replicate plots (12 × 7.5 m) of perennial ryegrass (PRG) (Lolium perenne) (cv. Premium), chicory (Ch) (Cichorium intybus) (cv. Puna II), red clover (RC) (Trifolium pratense) (cv. Merviot) and white clover (WC) (Trifolium repens) (cv. Aberdai) were established on 29 June 2009, in a randomized block design. PRG and Ch plots received inorganic N at 200 kg N ha-1 year-1. During the harvest years 1-3 (2010-12), plots were harvested on 5 occasions per annum and evaluated as described in Marley et al. (2013a;b). In February 2013, Gallup 360 herbicide (360g l-1 glyphosate; Barclay Ltd, Dublin, Ireland) was applied at 4 l ha-1. Soil was sampled per plot to a depth of 150 mm during March. Each plot was then split and allocated at random to one of two cultivation treatments. One half of each plot (3.75 m wide) was ploughed to a depth of 175 mm on 20 March and power-harrowed on 4 April (Ploughed). The other half plot Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 404 was left undisturbed and used for direct drilling (DD). On 5 April, spring wheat (Triticum aestivum) (cv. Tybalt) was sown on all plots using a Duncan Ecoseeder (Duncan Ag, Timaru, NZ) at a rate of 253 kg ha-1, calculated to sow 569 viable seeds m-2. All plots were flat rolled. Fertilizer was placed with the seed at sowing (49 kg N, 9 kg P2O5, 28kg K2O and 16kg SO3 ha1 ). Prilled lime was top dressed at 370 kg ha-1. On 21 May (wheat at growth stage (GS) 25), fertiliser was applied at 127 kg N, 22 kg P2O5, 72 kg K2O and 42 kg SO3 ha-1. On 18 June (wheat GS 32), samples of whole wheat plants were cut at 5 cm above ground level at 8 sites on each sub plot. Samples were oven dried, milled and submitted for chemical analysis by Inductively Coupled Plasma-Optical Emission Spectroscopy. N and S were determined by the Dumas Technique. Data were analysed by ANOVA as a split plot using Genstat® 11.1. Multiple comparisons were based on Bonferroni adjusted LSDs. Results and discussion Soil samples (data not shown) prior to wheat establishment showed no effect of previous cropping on pH or ammonium-N, P, Ca, Mg, B, Fe, Cu or Zn concentrations. Soil nitrate-N and K concentrations in PRG treatment were lower in other forage treatments (P < 0.001 and P < 0.05, respectively). Soil Mn concentration was higher where PRG was grown compared to other forages (P <0.05). Results showed previous crop affected whole-plant spring-wheat P, Mn and Cu concentrations (Table 1). Lower N, P, K, S and Cu concentrations were found in WPSW after ploughing, compared with DD. Within DD plots, WPSW following RC had higher Ca concentrations compared with other previous forage treatments. Within ploughed plots, WPSW following RC had a lower Mg concentration than did Ch and WC. The Mn concentration of WPSW following Ch was higher than following WC. Higher Cu concentrations were observed following PRG compared with RC and WC, and lower following RC than both PRG and Ch. Within establishment method, Ca concentration was higher in RC than all other previous forage treatments. Within the ploughed treatment, wheat following WC had a lower Mg concentration than Ch and WC. No differences were observed between treatments in Fe, Zn or B concentrations (means: 56.2, 44.18, 3.21 mg kg-1 DM,). Conclusions Overall, findings showed previous cropping and establishment method alter the mineral composition of WPSW but results varied for each mineral. Acknowledgements This work is funded through the Rural Development Plan for Wales 2007 – 2013, which is funded by the Welsh Government and the European Agricultural Fund for Rural Development References Marley C.L., Fychan R., Scott M.B., Davies J.W. and Sanderson R. (2013a) Yield, nitrogen and mineral content of chicory compared to perennial ryegrass, red clover or white clover over two harvest years. Grassland Science in Europe 18, 249-251. Marley C.L., Fychan R., Scott M.B., Davies J.W. and Sanderson R. (2013b) Trace element content of chicory compared with perennial ryegrass, red clover or white clover over two harvest years. Grassland Science in Europe 18, 252-253. Cannell R.Q., Ellis F.B., Christian D.G., Graham J.P. and Douglas J.T. (1980) The growth and yield of winter cereals after direct drilling, shallow cultivation and ploughing on non-calcareous clay soils, 1974–8. The Journal of Agricultural Science 94, 345-359. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 405 Table 1. Chemical composition (g/kg DM unless otherwise stated) of whole plant spring wheat established by ploughing or direct drill cultivation methods following pure swards of perennial ryegrass, chicory, red clover or white clover Previous Crop Nitrogen Phosphorus Potassium Sulphur s.e.m. Prob Establish. Perennial Ryegrass Chicory Red Clover White Clover Mean Rows Columns Previous Crop Establish. Method Prev.Est Conven. 35.47 37.25 33.12 35.32 35.29 1.734 1.377 0.065 <0.001 0.709 D. Drill 39.85 39.97 35.32 38.07 38.31 16.6df 12df Mean 37.66 38.61 34.22 36.70 Conven. 3.085 2.925 2.622 2.770 2.851 0.1329 0.1112 0.001 0.022 0.230 D. Drill 3.405 3.130 2.635 2.817 2.997 17.5df 12df Mean 3.245c 3.027bc 2.629a 2.794ab Conven. 29.30 32.20 31.63 29.28 30.60 1.996 1.520 0.460 <0.001 0.101 D. Drill 35.33 36.03 31.90 33.90 34.29 15.9df 12df Mean 32.31 34.11 31.76 31.59 Conven. 2.630 2.825 2.482 2.658 2.649 0.1397 0.1219 0.171 <0.001 0.500 D. Drill 3.065 2.995 2.815 3.015 2.972 18.3df 12df Mean 2.847 0.055 0.604 0.030 0.095 1.151 0.026 0.039 0.974 0.113 0.001 <0.001 0.197 2.910 2.649 3.260 D. Drill 3.427 a Mean 3.344 3.319 3.823 3.391 Conven. 1.805ab 1.917b 1.640a 2.062b 1.856 0.1114 0.0718 D. Drill 1.737 1.925 1.737 1.805 1.801 13.6df 12df Mean 1.771 1.921 1.689 1.934 Manganese Conven. 47.25 53.50 54.75 42.25 49.44 3.812 3.723 (mg/kg DM) D. Drill 50.50 54.75 46.25 46.50 49.50 20.3df 12df Mean 48.88ab 54.12b 50.50ab 44.38a Copper Conven. 7.575 7.525 6.275 6.775 7.037 0.3630 0.3104 (mg/kg DM) D. Drill 9.325 8.162 17.9df 12df Calcium Magnesium Conven. Mean 8.450 3.250 a 3.388 a 2.836 a 8.400 c 7.962 3.905 b 3.420a 3.459 0.1828 0.0775 3.740 b a 3.479 10.8df 12df 7.175 bc 6.725 3.363 7.750 a 7.262 ab Within rows, treatment values with differing lower case superscript differ significantly (P < 0.05) Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 406 Effect of soil amendment in the cultivation of selected grass species Sosnowski J., Jankowski K., Kolczarek R. and Wiśniewska-Kadźajan B. Siedlce University of Natural Sciences and Humanities, 08 -110 Siedlce, ul. B. Prusa 1, Poland. Corresponding author: laki@uph.edu.pl Abstract Cultivation of Dactylis glomerata and Festulolium braunii was carried out in the years 201213. The following experimental treatments were applied: K- control (without fertilization and soil amendment), NPK - mineral fertilization, UG - UGmax preparation, EU - Ekoużyźniacz preparation, HA - Humus Active preparation, NPK + UG, NPK + EU and NPK + HA. Preparations were applied as sprays in both study years for the first regrowth of grasses. The following mineral fertilizations were used: N - 150 kg ha-1, P - 80 kg P2O5 ha-1, K - 120 kg K2O ha-1. All treatments were applied three times per year. Characteristics such as dry matter yield of plants and content of fibre fractions like NDF, ADF and ADL. The results were evaluated statistically by analysis of variance. Differentiations were verified by Tukey's test at a significance level P≤ 0.05. Regardless of the mineral fertilizers, soil preparations resulted in an increase in yield of D. glomerata and F. braunii. The studies showed that, regardless of the combination of fertilizer, the average annual yields of D. glomerata were over 17% higher than the yields obtained from Festulolium braunii. Fibre fractions, like NDF, ADF and ADL in the analysed plant material were not statistically differentiated and the values obtained were typical for the studied grass species. Keywords: Humus Active UGax, mineral fertilization, grasses Introduction Soil fertilizer, due to the organic components such as humus and microorganisms, helps to improve the biological activity of the soil, increases the binding of free nitrogen from the air, and reduces erosion and losses of nutrients (Sosnowski and Jankowski, 2012). The formulations have also been attributed to the increased activation of the mineralization processes of organic components of soil (Sosnowski and Jankowski, 2013). The aim of the study was to determine the effect of three soil-improvement preparations, used against a background of NPK fertilization, on the yield and fibre fractions (NDF, ADF and ADL) in the dry matter of Dactylis glomerata and Festulolium braunii. Material and methods In the years 2012-13 cultivation of Dactylis glomerata L. cv. Sulino Bora and Festulolium braunii cv. was carried. Plot area was 6 m2. The following experimental factors were used: Kcontrol (without fertilization and soil fertilizer), NPK - mineral fertilization, UG - UGmax preparation, EU - Ekoużyźniacz preparation, HA - Humus Active preparation, NPK + UG , NPK and EU+ NPK+HA. Preparations were applied as sprays at dose rates of 0.9 litre ha-1 of preparation diluted in 350 litres of water. It was used in both years of the study for the first regrowth (phase of shooting). The composition of each formulation is shown in Table 1. Mineral fertilization rates were as follows: N-150 kg ha-1, P-80 kg P2O5, K-120 kg K2Oha-1. All fertilizations were applied three times. Evaluations were made of the dry matter yield of plants and fibre fractions (NDF, ADF and ADL). Determination of the fractions was performed by NIRS method on NIRFlex N -500 using ready-calibration for dried fodder of the INGOT company. The results were evaluated statistically by analysis of variance. Mean differentiation was verified by Tukey's test at a significance level of P ≤ 0.05. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 407 Table 1. The composition of fertilizers Macroelements [g kg-1] Preparation Microelements [mg kg-1] N P2O5 K2O Ca Mg Na Mn Fe Zn Cu UG 1.2 0.5 3.5 - 0.1 0.2 0.3 - - - EU 0.6 0.6 0.8 - - - - - - - HA 0.2 3.0 5.5 3.0 0.5 - 15 500 3 1 Organic components UG Lactic acid bacteria, photosynthetic bacteria, Azotobacter, Pseudomonas, yeasts, actinomycetes EU Endomycorrhizal fungi, bacteria, enzymes involved in the metabolism of earthworms Results and discussion Applied experimental factors: there were significant differences in the annual DM yields of crops (Table 2). Regardless of the species, on the treatments without fertilization, the highest yield (average 11.5 t DM ha-1) occurred on treatments with Active Humus preparation. In contrast, the use of mineral fertilizers contributed to a significant increase in yield (13.1 t DM ha-1) of plants supplied with UGmax (a combination of NPK + UG). It is worth noting that, regardless of the applied fertilizer combinations, the greater yields were obtained from D. glomerata: its average annual yields were over 11 t DM ha-1and were 17.8% higher than the average annual yields of Festulolium braunii. Table 2. Annual yield (t DM ha-1) of Dactylis glomerata and Festulolium braunii depending on NPK and fertilizer type (average for research years). Means in rows marked with the same small letters do not differ significantly; means in columns marked with the same capital letters do not differ significantly. NPK fertilization Species No NPK NPK Mean K0 UG EU HA NPK UG EU HA No NPK Dg 9.30 Ab 11.2 Aa 9.63 Ab 11.6 Ab 12.0 Ab 15.9 Aa 10.1 Ac 11.3 Bbc 11.6 Aa 11.2 Aa 11.4 A Fb 5.96 Bc 9.19 Bab 8.49 Bb 11.5 Aa 9.51 Bbc 10.2 Bb 8.74 Bc 13.5 Aa 9.04 Bb 10.3 Ba 9.67 B 7.63 10.2 9.06 11.5 10.8 13.1 9.44 12.4 10.3 108 c a b a b a b ab a a Mean NPK The data presented in Table 3 showed that, regardless of fertilization, in the cultivation of D. glomerata the highest production results were obtained using soil fertilizer UGmax. Table 3. Annual yield (t DM ha-1) Dactylis glomerata and Festulolium braunii depending on the applied fertilizer (average for research years. Means in rows marked with the same small letters do not differ significantly. means in columns marked with the same capital letters do not differ significantly. Fertilizer Species K UG EU HA Dg 10.6 Ab 13.6 Aa 9.87 Ab 11.5 Bab Fb 7.74 Bc 9.70 Bb 8.62 Bbc 12.5 Aa 9.17 b 11.7 ab 9.25 b Mean Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 408 The average annual yield on these treatments amounted 13.6 t DM ha-1 and was about 28% higher than the yields from the control. The use of Eco-fertilizer decreased average yields of D. glomerata in relation to the control by more than 7%. In turn, in Festulolium braunii, the highest efficiency was with the Humus Active preparation. Plants sprayed with this preparation contributed to yield increases from 7.74 (control) to 12.5 t DM ha-1 (for treatments receiving Humus Active). Factors used in the experiment did not result in any statistically significant differences in contents of NDF, ADF and ADL fractions (Table 4). Table 4. Content [% DM] of factions in dry matter of Dactylis glomerata and Festulolium braunii depending on the applied mineral fertilizers and soil fertilizer (average of from cuts and study years). Means in rows marked with the same small letters do not differ significantly; means in columns marked with the same capital letters do not differ significantly. NPK fertilization Species No NPK K0 UG EU NPK HA NPK UG EU Mean NPK HA Without NPK Neutral detergent fibre NDF Dg Fb Mean 52.7 Aa 54.4 Aa 57.2 Aa 50.1 Aa 52.4 Aa 54.1 Aa 48.6 54.1 53.6 52.3 Ab Aa Aa Aa 48.6 44.3 48.7 54.0 52.9 48.3 52.9 Ba Ba 52.7 Aa 52.1 Aa 51.7 A Aa Aa Aa Aa Aa 50.7 a 49.4 a 52.9 a 50.9 a 52.6 a 53.1 a 51.3 a 53.5 a 50.9 a 52.6 a 35.0 Aa 30.7 Aa 32.9 A 34.4 Aa 30.0 Aa 32.2 A 53.0 A 50.6 A Acid detergent fibre ADF Dg Fb Mean 35.4 33.8 35.5 35.3 29.7 30.8 30.9 31.3 Aa Aa Aa Aa Aa Aa Aa Aa 35.8 34.7 33.5 35.6 30.8 28.2 30.5 30.6 Aa Aa Aa Aa Aa Aa Aa Aa 35.7 a 34.3 a 34.5 a 35.5 a 30.3 a 29.5 a 30.7 a 31.0 a 34.7 a 30.4 a 4.13 Aa 3.92 4.42 Aa Aa Acid detergent lignin ADL Dg Fb Mean 4.25 3.54 3.65 4.22 4.43 4.50 4.61 Aa Aa Aa Aa Aa Aa Aa 3.79 4.35 4.27 4.45 4.45 4.47 4.64 4.45 4.22 4.50 Aa Aa Aa Aa Aa Aa Aa Aa Aa Aa 4.02 a 3.95 a 3.96 a 4.34 a 4.44 a 4.49 a 4.63 a 4.29 a 4.07 a 4.46 a 4.17 A 4.36 A In previous studies (Sosnowski and Jankowski 2012) there were also no differences in the proportion of fibre fractions in the dry matter of grasses grown in pure stands on arable land Conclusion Regardless of the fertilization, the use of soil improvement preparations resulted in an increase in yield of D. glomerata and Festulolium braunii. The best production of D. glomerata was obtained by using the soil fertilizer UGmax, but for Festulolium braunii this was with Humus Active preparation. The application of NPK resulted in a 29% increase in crop yield when sprayed with UGmax. The studies have shown that, regardless of the combination of fertilizer, the average annual yields of D. glomerata were over 17% higher than the yields obtained from Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 409 Festulolium braunii. Fibre fractions, like NDF, ADF and ADL in the analysed plant material were not statistically differentiated and the values obtained were typical for the studied grass species. References Sosnowski J. and Jankowski K. (2012) Effect of soil fertilizer UGmax and fertilization with nitrogen, phosphorus, and potassium on the energy and nutrition values of Lolium multiflorum. Acta Scientiarum Polonorum, s. Agricultura 11(3): 65 – 74. Sosnowski J. and Jankowski K. (2013) Effect of soil fertilizer UGMax on feed value of some grass species depending on their cuts. Acta Scientarum Polonorum 12(1) 35 -44. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 410 Milk production and profitability in relation to size of grassland farms Jankowski K., Sosnowski J., Wiśniewska-Kadżajan B. and Kolczarek R. Siedlce University of Natural Sciences and Humanities, Department of Grassland and Green Areas Creation, Siedlce, Poland. Corresponding author: laki@uph.edu.pl Abstract In recent years the most profitable activity in agriculture is considered to be with milk production. The effectiveness of milk production depends primarily on the direct costs and the obtained price. Therefore the aim of this work was to analyse the financial results of dairy farms in the eastern part of the Mazovia region. The research was completed in 2010. The owners of 36 farms were sent a questionnaire containing 18 questions. The whole population was devoted to six production groups, depending on the number of physical units of dairy cows. Based on survey data, an analysis of the profitability of milk production was done, and statistical analysis by calculating the Pearson linear correlation coefficient (r) and coefficient of determination (R2) was done. The study showed that the highest annual yield of milk from a cow at the level of 7500 kg, was obtained by farmers with an area of 19.7 ha, with the density of 1.82 BFU ha-1. The performed regression analysis showed significant positive correlation between the effectiveness index and the surface agricultural lands in the farms or density of agricultural lands. Milk production was profitable on farms that had suitable areas of grasslands. Keywords: grassland, milk production, farms, costs Introduction In Poland, milk production has had the largest share of commercial agricultural production over many years, and the dairy sector (production and processing) involves 25% of the industry workforce and produces 17% of total production of the agricultural industry. Therefore, the milk market is one of the basic agri-food markets in our country. In addition, it is worth noting, that in 2008, Polish exports of dairy products on to European markets increased by up 73.6% (Baer-Nawrocka and Kiryluk-Dryjska, 2010). It can therefore be assumed that the Polish dairy sector is gaining importance in western markets. The aim of the study was to analyse of the financial results of farms producing milk, taking into account the factors of production occurring within farm, and market conditions affecting the profitability of the business. Material and methods Research by direct interview was conducted in 2010 on 36 farms from the eastern Mazovia region. A questionnaire containing 18 questions was sent to farm owners. The obtained data were used to characterize the various research objectives, resulting in a breakdown of the analysed population into 6 groups, depending on the number of physical units of dairy cows: group A up to 10, B - from 11 to 15, C - from 16 to 20, D - from 21 to 25, E - from 26 to 30, F - over 30. Moreover, in the study population, farms from group A, having 10 dairy cows, dominated. In the other production groups there were from 5 to 6 dairy farms. On the basis of questionnaire data, an analysis taking into account the profitability of milk production has done with the following economic size: the cost of herd renewal, the cost of feed (e.g. feed purchased), own feed, own meadow hay; own corn silage, own mineral mixture and the cost of specialist. Profitability index as the value of effectiveness ratio was calculated for each frm and are presented as average from production group, according to the following formula (Kisiel et al., 1996): Wo = (Wp : Wk ) × 100%. where: Wo - value of effectiveness ratio, Wp - value of commodity production, Wk - value of direct costs. In addition, statistical analysis was performed by calculating the Pearson correlation coefficient (r) and coefficient of determination Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 411 (R2) between the effectiveness ratio and the following variables: density of GPU ha- 1 of arable lands, surface of arable lands in hectares, the average annual milk yield of cows in kg, the price of 1 litre of milk in PLZ. Results The obtained data show that in farms with a higher number of cows, milk yield was two times higher than on farms with up to 10 dairy cows. The highest value of this parameter (7500 kg milk per year per cow) was achieved in farms with an average of 28 dairy cows (group E), used during a period of 8 years, with an average of approximately 19 ha of arable land and about 14 hectares of meadows and pastures. The study also showed a significant positive correlation between the surface of agricultural land and the density of dairy cows and the profitability of milk production (Table 1). Table 1. Statistical analysis for relation between individual diagnostic features r R2 y = ax ± b Syx Syx% 0.624 0.39 y = 121.96 + 1.223x 275 3.7 0.758 0.59 y = 112.63 + 36.471x 1.86 2.4 0.989 0.98 y = 87.694 + 0.1205x 1.29 1.6 0.655 0.43 y = 12.092 + 103.48x 2.35 3.1 index of milk production profitability x surface of agricultural lands in studded farms index of milk production profitability x milk cows density on agricultural lands index of milk production profitability x mean annual yield index of milk production profitability x value of milk price r -correlation coefficient, R2 –determination coefficient Syx –standard deviation ; Syx%– relative estimate error The correlation coefficient for this relationship was 0.624 and 0.758 respectively. Furthermore, data suggest that the profitability of the analysed production was largely influenced by the average annual yield of milk per cow. This feature, up to 98 % (R2 = 98 %) determined the volatility of the effectiveness ratio achieved in this population farms. In addition, there was wide diversity in milk prices. The highest unit prices were negotiated farmers with more than 26 dairy cows - Groups E and F. On average, these farms were paid from 1.45 to 1.54 PLZ per litre. Economic analysis of farms producing milk plays an important role. The data in Table 2 show that the value of commodity production within the analysed groups varied widely and ranged from 4560 PLZ (in farms from group A) to 11935 PLZ (in farms from group E). A similar tendency was also analysed in the direct costs of production. Farms with fewer cows also had the lowest costs in their business (3474.70 zł - Group A), while for farms that had more than 25 cows the value of direct costs were higher by more than 1,300 zł. Average gross margin per cow, obtained by the average farm producing milk in the study area, was 2.148 PLZ. In contrast, in analysing the value of the surplus, which reached the farms from various production teams, great diversity can be seen. The highest value of surplus was obtained farmers having an average of 26-30 dairy cows (3810 PLZ - Group E), and in second place were the farmers that had more than 30 cows (2550.33 PLZ - Group F). In terms of milk production, the least income was earned by the farmers having up to 15 dairy cows (528.00 PLZ - Group B). A similar tendency was observed in terms of the breakdown of the profitability index of milk production. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 412 Table 2. Analysis of financial results in studied farms by production groups (mean from farms) Value of commodity production (receipts) Production groups [zł] Direct cost [zł] Direct surplus [zł] Index of production profitability milk [%] A 4066.60 3049.95 1016.65 133.33 B 2400.00 1872.00 528.00 128.20 C 3416.66 2425.36 991.30 140.87 D 3995.83 2797.08 1198.75 142.85 E 7700.00 3890.00 3810.00 197.94 F 5931.00 3380.67 2550.33 175.43 Mean 4585.01 2902.51 2148.68 157.96 Conclusion The highest milk yield per cow was at the level 7500 kg per year, which was reached by farmers with an area of arable land of 19.7 ha. Annual gross margin for agricultural farms producing milk per cow ranged from 528 to 3810 PLZ. There was a significant impact on this value of the surface area of arable land, cow density, the average annual milk yield per cow and the unit price of milk. The development of dairy farms and the economic benefits derived from this activity are influenced by many factors. The most important of them are efficiency, which most determines the profitability of milk production (98%) and quality of produced raw material, which affected the value of the obtained purchase prices. References Baer-Nawrocka A. and Kiryluk-Dryjska E. (2010) Wpływ likwidacji kwot mlecznych na sytuację produkcyjna i ekonomiczną producentów mleka w Unii Europejskiej. Wieś i Rolnictwo, Warszawa, 135-147. Kisiel R., Juchniewicz M. and Kucka E. (1996) Koszty i opłacalność produkcji mleka. Rocz. Nauk Zoot., nr 23, 2, 321-334. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 413 Meadow apophytes in segetal communities Skrzyczyńska J.1, Ługowska M.1, Skrajna T.1, Rzymowska Z.1, Jankowska J.2 and Sosnowski J.3 1 Department of Agricultural Ecology, 2 Department of Agrometeorology and Land Reclamation, Poland, 3 Department of Grassland and Green Areas Creation University of Natural Sciences and Humanities in Siedlce, Poland. Corresponding author: laki@uph.edu.pl Abstract The work analyses the share of meadow apophytes in the structures of segetal communities. Floristic studies were conducted from 1995 to 2011 in east-central Poland. A total of 85 species representing meadow apophytes were found in winter and spring cereals as well as tuber and root crops, very rare and rare species being the most numerous group. The group of frequent and common apophytes comprised 14 species. Frequency of occurrence and the coverage of these species are dependent on the agricultural technology. Keywords: meadow apophytes, semi-natural and anthropogenic habitats, adaptation, biodiversity Introduction One of symptoms of flora synanthropisation is invasion of native species from natural and seminatural habitats, such as meadow habitats, to disturbed habitats. The process does not markedly affect the qualitative changes of flora but it indicates the adaptive attributes that some species have. Due to changes that have recently taken place in the cropping profile, as well as methods and means of agricultural production, agrocenoses are becoming more and more floristically impoverished. At the same time, the compensation of ubiquitous species is increasing thus destroying structures of segetal communities (Stehlik et al., 2007, Storkey et al., 2011). Native species spreading from meadow habitats increase the floristic diversity of segetal communities (Skrajna et al., 2010; Dąbkowska and Sygulska, 2013). The objective of this work was to: indicate apophytes in the flora of segetal communities in east-central Poland; determine, using the frequency of occurrence, constancy of occurrence and the index of coverage, the role that meadow apophytes play in agrocenoses; indicate the meadow species that are best adapted to occur in segetal communities. Materials and methods Floristic studies were carried out from 1995 to 2011 in east-central Poland. The area is part the Central Mazovian Lowland and it covers over 20 thousand km2 (Kondracki, 2009). The dominant soils are podzols and, in valley bottoms, alluvial soils of various origins. Locally on denudating plains, chernozems developed on periglacial silty and clay landforms. Places located lower are surrounded by bog and post-bog soils. The present work is based on over 1200 floristic inventories which were performed in cereal, root and tuber crops as well as stubbles. Next, meadow apophytes, which established in the agrocenoses as secondary habitats, were determined in the flora of these communities. Apophyte occurrences in field habitats were characterized by describing the following: soil conditions, frequency and constancy of occurrence, and the index of coverage in various habitats where they were found. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 414 Results Agrocenoses in east-central Poland comprise 85 meadow apophytes which made the segetal flora in this area richer (Table 1). Analysis of the biological spectrum revealed that hemicryptophytes were the most numerous group (60 species), followed by geophytes (14 species) and therophytes (11 species). Meadow apophytes occurred in winter and spring cereals and non-cultivated stubble fields. They also established in dense stands of tuber and root crops. Some of the species remained in communities occupying edges of fields and field borders. Very rare and rare species formed the richest group (71) and included, e.g.: Bromus hordeaceus, Alopecurus geniculatus, Plantago media, Phleum pratense, Lathyrus pratensis, Rumex confertus, Knautia arvensis. This group was followed by common and frequent meadow apophytes (14), e.g.: Achillea millefolium, Artemisia vulgaris, Cerastium holosteoides, Plantago lanceolata, Equisetum arvense. Meadow apophytes were found in segetal communities established on various soil types and kinds (Table 1), for example: grey brown podzolic soils which formed from sands of various origins, silty soils, true and degraded charnozems, brown soils and alluvial soils. Table 1. Meadow apophytes in the agrocenoses of east-central Poland; common and frequent species Life S.n. Species Winter cereals Spring cereals Tuber crops Stubble-fild form 1. A,Bw,Dz,F Achillea millefolium L. H 2. Artemisia vulgaris L. H 3. Cerastium holosteoides H Fr.em.Hyl. 4. Equisetum arvense L. G A,Bw,Dz,F A, Bw,Dz,F A,Bw,Dz,F ps;pgl;pgm;gl.płz;płi psp;pgl;pgm;płz;płi ps;pgl;pgm;płz ps;pgl;pgm;płz;płi pH – 5.5-8.0 pH – 4.5-7.5 pH – 4.5-6.0 pH – 4.5-8.0 S=I-II;W=4-37 S=I;W=7-20 S=I-II ;W=3-30 S=II-V;W=7-195 A,Bw,Dz,F Bw,Dz,F A, Bw,Dz,F A,Bw,Dz,F ps;pgl;pgm;gl;płz psp;pgl;pgm;płz;gs ps;pgl;pgm;płz;płi ps;pgl;pgm;gsp.płz pH – 4.5-8.0 pH – 4.5-7.5 pH – 5.0-6.5 pH – 4.5-8.0 S=I-II;W=8-47 S=I;W=5 S=I-II;W=8-39 S=II-IV;W=9-61 A,Bw,Dz,F Bw,Dz,F Bw,Dz,F Bw,Dz,F psp;pgl;pgm;płz pgl;glp;pgm;płz pgl;pgm;gsp;płz psp;pgl;glp;pgm;płz pH – 5.5-7.5 pH – 4.5-7.5 pH – 6.0-7.5 pH – 4.5-7.5 S=II-IV;W=12-37 S=I;W=3-7 S=I;W=5-20 S=II-IV;W=20-100 A,Bw,Dz,F Bw,Dz,F A, Bw,Dz,F A,Bw,Dz,F ps;pgl;pgm;gl;płz psp;pgl;pgm;gs;płz ps,pgl;płz ps;pgl;pgm;gl;płz pH – 4.5-7.0 pH – 4.5-6.5 pH – 4.5-6.5 pH – 4.5-6.0 S=II-IV;W=31-160 S=II-IV;W=30-173 S=II-V ;W=50-417 S=II-V;W=30-305 5. Glechoma hederacea L. H 6. Melandrium album /Mill./ Garcke T Bw,F Bw,Dz,F A,Bw, F A,Bw,Dz,F pgl;płz pgl;płz ps;pgl;płz pgl;pgm;płz pH – 5.0-7.0 pH – 4.5-7.5 pH – 5.5-6.5 pH – 5,0-7,5 S=I;W=5 S=I;W=5 S=I-III;W=10-50 S=I-III;W=30-65 A,Bw,Dz,F Bw,Dz,F A, Bw,Dz,F A,Bw,Dz,F psp;pglp;gsp;płz psp;pgl;płz; ps;pgl;płz psp;pgl;pgm;płz pH – 5.0-7.0 pH – 4.5-7.5 pH – 5.0-6.5 pH – 5.0-7.0 S=II;W=5-20 S=I;W=16 S=II;W=5-20 S=II-IV;W=21-465 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 415 7. T Poa annua L. 8. Plantago lanceolata L. H 9. Potentilla anserina L. H 10. Ranunculus repens L. H 11. Stachys palustris L. G 12. Stellaria media T /L./ Vill. 13. Taraxacum sp. H F.H. Wigg. coll. 14. Trifolium repens L. H A,Bw,Dz,F Bw,Dz,F A, Bw,Dz,F A,Bw,Dz,F ps;pgl;glp;płz psp;pgl;płz ps;pgl;płz ps;pgl;pgm;płz pH – 4.5-7.5 pH – 4.5-7.5 pH – 5.0-6.5 pH – 4.5-7.5 S=I-II;W=3-10 S=I-II;W=4-25 S=I-IV;W=12-48 S=II-IV;W=7-67 A,Bw,Dz,F Bw, F ps;pgl;płz pgl;płz pH – 5.5-7.0 pH – 5.0-7.0 S=I-II;W=25 S=I;W=5 A, Bw,Dz,F A,Bw,Dz,F ps;pgl;pgm;płz ps;pgl;pgmp;płz pH – 4.5-6.0 pH – 4.5-7.5 S=I-III;W=5-45 S=I-IV;W=8-70 Bw,Dz,F Bw,Dz,F A, Bw,Dz,F Bw,Dz,F pgl;pgm;płz psp; płz pgl;płz pgl;pgm;płz pH – 4.5-7.0 pH – 6.0-7.0 pH – 5.0-6.5 pH – 4.5-7.0 S=II;W=5-25 S=I;W=5 S=I-II;W=7-47 S=I-III;W=5-50 Bw,Dz,F, Bw,F Bw,Dz,F Bw,Dz,F pgl;płz;płi pgl;płz pgl;płz pgl;płz; płi pH – 5.5-7.0 pH – 6.0-7.0 pH – 5.0-6.5 pH – 5.5-7.0 S=II-III;W=5-40 S=I;W=3-15 S=I-III;W=3-50 S=I-IV;W=15-65 Bw,Dz,F Bw,Dz,F Bw,Dz,F Bw,Dz,F pgl;pgm;płz;płi pgl:gl;płz pgl;płz pgl;pgm;gl;płz;płi pH – 5.0-7.5 pH – 4.5-7.5 pH – 5.5-7.0 pH – 4.5-7.5 S=II-IV;W=31-123 S=I-III;W=4-84 S=I-V;W=10-430 S=I-V;W=20-715 Bw,D,Dz,F Bw,Dz,F Bw,Dz,F Bw,Dz,F ps;pgl;pgm;płz pgl;pglp;pgm;płz ps;pgl;płz ps;pgl;pgm;płz pH – 5.0-7.0 pH – 4.5-7.5 pH – 4.5-6.5 pH – 4.5-7.5 S=I-IV;W=8-15 S=I-III;W=38-138 S=II-V;W=75-438 S=I-IV;W=20-1245 Bw,Dz,F Bw,Dz,F A,Bw,Dz,F A,Bw,Dz,F ps;pgl;pgm;płz ps;pgl;płz ps;pgl;płz ps;pgl;pgm;płz pH – 5.0-8.0 pH – 4.5-7.5 pH – 5.0-6.0 pH – 4.5-8.0 S=II-IV;W=8-47 S=I-II;W=3-25 S=I-IV;W=10-73 S=II-V;W=20-154 A,Bw,Dz,F Bw,Dz,F A,Bw,Dz,F A,Bw,Dz,F ps;pgl;pgm;płz ps; pgl;płz ps;pgl;płz ps;pgl;pgm;płz pH – 5.5-7.0 pH – 4.5-7.5 pH – 4.5-6.5 pH – 4.5-7.0 S=II-IV;W=13-87 S=I-II;W=17-35 S=I-II;W=7-43 S=I-IV;W=10-86 Very rare: Alchemilla monticola Opiz – H, Alopecurus geniculatus L. – G, Alopecurus pratensis L. – G, Anhriscus sylvestris (L.) Hoffm. – H, Arrhenantherum alatius (L.) P. Baeur. ex J. Presl & C. Presl – H, Bellis perennis L. – H, Briza media L. - H, Campanulla patula Griseb. – H, Carex hirta L. - G, Carum carvi L. – T, Crepis biennis L. – T, Festuca rubra L. s.s – H, Filipendula ulmaria (L.) Maxim. – H, Heracleum sibiricum L. – H, Heracleum sphondylium L. S.STR. – H, Hieracium umbellatum L. – H, Leontodon hispidus L. – H, Leucanthemum vulgare Lam. S. STR – H, Lichnis flos-cuculi L. – H, Luzula campestris /L./ DC. – H, Lythrum salicaria L. – H, Polygonum bistorta L. – G, Ranunculus acris L. S. STR. – H, Sanguisorba officinalis L. – H, Tragopogon pratensis L. S. STR – H, Rare: Agrostis capillaris L. – H, Bromus hordeaceus L. – H, Centaurea jacea L. – H, Crepis capillaris (L.) Wallr. – T, Equisetum palustre L. – G, Festuca pratensis Huds. – H, Gagea pratensis (Pers.) Dumort – G, Galium mollugo L. – G, Galium verum L. – G, Geranium pratense L. – H, Holcus lanatus L. – H, Hypochoeris radicata L. – H, Lathyrus pratensis L. – H, Lysimachia nummularia L. – G, Lysimachia vulgaris L. – G, Odontites serotina (Lam.) Rchb. – T, Pastinaca sativa L. s.s. – H, Phleum pratense L – H, Plantago media L. – H, Potentilla reptans L. – H, Prunella vulgaris L. – H, Ranunculus sardous Crantz – H, Rumex acetosa L. – H, Rumex confertus Willd. – H, Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 416 Trifolium campestre Schreb. – T, Trifolium dubium Sibth. – T, Trifolium hybridum L. – H, Veronica chamaedrys L. S. STR. – H, Veronica serpyllifolia L. – H, Dość rzadkie: Agrostis gigantea Roth. – H, Agrostis stolonifera L. – H, Centaurium pulchellum (Sw.) Druce – T, Dactylis glomerata L. – H, Equisetum sylvaticum L. – G, Hypericum perforatum L. – H, Knautia arvensis /L./ J.M. Coult. – H, Leontodon autumnalis L. – H, Lolium perenne L. – H, Lotus corniculatus L. – H, Poa pratensis L. – H, Poa trivialis L. – H, Rumex crispus L. – G, Sagina procumbens L.- H, Stellaria graminea L. – H, Trifolium pratense L. – T, Vicia cracca L. – H, Explanations: T-terophytes, H-hemicryptophytes, G-geophytes; S-constancy class; W-cover coefficient; Apodsolic soil; Bw- brown leached; F-alluvial, D- proper black earth, Dz-degraded black earth; ps-slightly loamy sand; pgl-light loamy sand; pglp-silty light loamy sand; pgmp-hefty dusty loamy sand; psp-silty slightly loamy sand; pgm-heavy loamy sand; płz-silt; płi- gl-light clay; gs-maen clay; glp- silt loam; gsp- mean clay silt Conclusions Meadow apophytes enrich the segetal flora of east-central Poland. Frequent or mass occurrence of these species is closely connected with conventional soil and crop plant cultivation. An introduction of new technologies into agriculture removes meadow apophytes from segetal communities. References Dąbkowska T. and Sygulska P. (2013) Variations in weed flora and the degree of its transformation in ecological and extensive conventional cereal crops in selected habitats of the Beskid Wyspowy mountains. Acta Agrobotanica 66 (2), 123-136. DOI:10.5586/aa.2013.029. Kondracki J. (2009) Regional geography of Poland. Polski PWN, Warszawa. Skrajna T., Skrzyczyńska J. and Ługowska M. (2010) The segetal flora of the Mazowiecki Landscape Park. Plant Breeding and Seed Science 61, 93-104. Stehlik I., Caspersen J.P., Wirth L. and Holderegger R. (2007) Floral free fall in the Swiss lowlands: environmental determinants of local plant extinction in a peri-urban landscape. Journal of Ecology 95, 735-744. Storkey J., Meyer S., Still K.S. and Leuschner C. (2011) The impact of agricultural intensification and land-use change on the European arable flora. Proceedings of the Royal Society B 279, 1421-1429. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 417 Effectiveness of grassland management and mechanical methods for the weed control of Colchicum autumnale in permanent meadows Peratoner G., Figl U., Florian C., Klotz C. and Gottardi S. Laimburg Research Centre for Agriculture and Forestry, 39040 Auer/Ora (BZ), Italy Corresponding author: Giovanni.Peratoner@provinz.bz.it Abstract Several measures are commonly suggested to control meadow saffron (Colchicum autumnale L.), a poisonous weed well adapted to extensive management, but clear experimental evidence is not available for all of them. For this reason, the effect of two mechanical methods (early cut, rolling with a spike roller) in combination with fertilization (present/absent) and oversowing (present/absent) was investigated in a four-year field trial at two mountain permanent meadows. An average efficacy of 15% in reducing plant density was found following three consecutive treatments with an early cut, while rolling had virtually no effect on it. An early cut resulted also in a rapid decrease over the years of plant fresh weight and the proportion of fertile plants of Colchicum. No effect was detected for fertilization and oversowing. In order to achieve a strong reduction of Colchicum density, an early cut needs to be applied over a long time. Keywords: Colchicum autumnale; cut; spike roller; oversowing; fertilization Introduction Meadow saffron (Colchicum autumnale L.) is a toxic perennial geophyte with an unusual phenology, with flowering taking place in autumn and a short photosynthetically active period finishing in the middle of summer; for this reason, extensive management with late mowing or grazing promotes population growth and can lead to detrimental effects for forage quality and marketability of hay (Jung et al., 2011, Winter et al., 2011). Increasing abundance of Colchicum autumnale has been also repeatedly observed in the mountain region of South Tyrol. These reports apply mostly to permanent meadows with an unusually late first cut, which is mainly related to logistic issues of the farm management. Several measures concerning grassland management and mechanical weed control are reported by codes of practice (see, e.g., Caputa, 1984), but clear experimental evidence for the effectiveness of some of them is only partly available in the literature. In order to get reliable information for some of these methods (early cut, mechanical damage by means of a spike roller, also in combination with fertilization), a four-year field experiment was conducted in South Tyrol between 2009 and 2012. The effect of oversowing, a popular measure to improve forage composition, was investigated as well. Materials and methods A four-year field experiment was conducted in two permanent mountain meadows with a moderate infestation of Colchicum autumnale at an altitude around 1500 m a.s.l. (Table 1). Table 1. Description of the experimental sites. Experimental site St. Veit Kameriot Geographic coordinates Altitude (m a.s.l.) Slope (%) Exposition Colchicumdensity at trial start (plants m-2) Colchicumfresh weight at trial start (g m-2) 46° 43' 20" N 12° 43' 20" E 1490 60 S 5.6 65.2 46° 42' 6" N 12° 9' 44" E 1460 15 W 8.6 37.2 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 418 Both meadows are yearly cut twice, with the first cut being performed quite late (after mid June). Four factors were investigated: (i) weed control method: cut at ground level of Colchicum-plants by means of electric scissors (mowing), rolling using a motor mower for steep slopes (PS15, Brielmaier, Friedrichshafen, D), equipped with two 0.5 m-wide spike rollers instead of wheels and a machine weight of 200 kg, and an untreated control. Mechanical treatments were applied between beginning and completion of capsule formation, roughly corresponding to end of May-beginning of June; (ii) fertilization with 20 m³ ha-1 of 2:1 waterdiluted slurry applied after mechanical weed control (present/absent); (iii) oversowing (present/absent) after mechanical weed control with 20 kg ha-1 of a seed mixture (12% weight Agrostis tenuis, 6% Trisetum flavescens, 12% Dactylis glomerata, 28% Festuca rubra, 17% Festuca pratensis, 5% Lolium perenne, 8% Phleum pratense, 12% Poa pratensis). After seeding, harrowing was simulated by manual raking of the plots. An orthogonal design was laid out as a randomized complete block design with three replicates and a plot size of 2 m² (1 × 2 m). In each plot, the number of Colchicum-shoots was counted before applying the weed control treatments. The fresh weight of the epigeal Colchicum-biomass and the number of generative plants with capsules were assessed in the mown plots. Treatment efficacy on shoot density was calculated with the methods of Henderson and Tilton (1955), which also takes in account the changes of plant density in the control plots to assess treatment efficacy. Negative efficacy values, corresponding to increases of plant density with respect to the control treatment, were allowed to occur up to -100%, while 12 implausible values were removed from the data set. Data were analysed by means of a mixed model assuming the weed control method (only for the effectiveness data), the fertilization, the oversowing and the year as well as their interactions to be fixed effects and accounting for serial correlations due to repeated measures of the factor year with the plot as a subject. The block effect, the site factor and the interactions of the latter with the fixed factors up to the second order were considered to be random effects. Multiple comparisons were performed by Sidak’s test. The share of generative plants with capsules was analysed by means of non-parametric tests (Friedman test and multiple comparisons by Wilcoxon-Wilcox) because of violation of the assumptions of normal distribution of residuals and variance homogeneity even after data transformation. P ≤ 0.05 was considered to be significant. Results and discussion Efficacy in reducing the density of Colchicum plants was found to be affected by the weed control method only (P = 0.012), whilst no effect of fertilization (P = 0.69) and oversowing (P = 0.60) and of their interactions with other factors was detected. More precisely, a marginal significance of the control method with the year was observed (P = 0.50), as the efficacy of an early cut was found to differ from that of the rolling treatment only in the last observation year (Table 2). Table 2. Efficacy (%) of the weed control methods over the investigation period. Means without letters in common within each year significantly differ from each other. Weed control method Year Average across all years 2009-2010 2010-2011 2011-2012 Early cut 16.5 a -2.8 a 31.0 a 14.9 a Spike roller 6.2 a 2.5 a -6.3 b 0.8 b On average after three treatment seasons, the use of a spike roller seemed to produce no relevant effects, while the early cut led to an efficacy of about 15%. These results support recent experimental evidence about the negative effect of an early cut on population growth of Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 419 Colchicum autumnale (Winter et al., 2014), although the timing of the cut in the present experiment was not strictly bound to a well defined phenological stage (about 25 cm plant height and capsules with their top being about 10 cm above ground), as suggested by Jung et al. (2012). The mowing treatment resulted also in a decrease over time of the fresh weight of Colchicum, expressed as epigeal biomass as well as the mean plant weight and led to a very low proportion of fertile plants with capsules (Table 3). Table 3. Changes over time of plant biomass and plant weight of Colchicum autumnale and of the proportion of generative plants with capsules by performing a yearly mowing treatment with an early cut. Variable # † Year 2009 2010 2011 2012 Colchicum biomass fresh weight (g m-2) # 26.6 a 17.1 b 4.2 c 4.0 c Colchicum plant fresh weight (g plant-1) # 6.7 a 3.6 b 1.0 d 1.6 c Proportion of plants with capsules (%) † 33.8 a 22.6 a 1.7 b 1.3 b Analysis with mixed model, logarithm-transformed data. Back-transformed means are shown Friedman-test and multiple comparisons by Wilcoxon-Wilcox-test Whilst the lack of an effect of oversowing is not surprising because of to the known aleatory effect of this practice (Huguenin-Elie et al., 2006) and of the only light disturbance to the existing vegetation provided by raking, the lack of evidence for an effect of the fertilization was less expected, as this advice is often mentioned in codes of practice in order to enhance competition on Colchicum by other forage species. Conclusions The results provide evidence of the effectiveness of an early cut on the control of Colchicum autumnale in permanent meadows. This treatment contributes to a rapid reduction in plant weight and in the proportion of fertile plants, but it needs to be applied over a long time to achieve a strong reduction of plant density. The effect of often recommended measures, such as fertilization or rolling with a spike roller, could not be confirmed by the outcome of the present experiment. References Caputa J. (1984) Les "mauvaises herbes" des prairies/Die Wiesenunkräuter, AMTRA, Nyon, F, 194 pp. Henderson C.F. and Tilton E.W. (1955) Tests with acaricides against the brow wheat mite. Journal of Economic Entomology 48, 157-161. Huguenin-Elie O., Stutz C.J, Lüscher A., Gago R. (2006) Wiesenverbesserung durch Übersaat. Agrarforschung 13, 424–429. Jung L.S., Eckstein R.L., Otte A. and Donath T.W. (2012) Above- and below-ground nutrient and alkaloid dynamics in Colchicum autumnale: optimal mowing dates for population control or low hay toxicity. Weed Research 52, 348-357. Jung L.S., Winter S., Eckstein R.L., Kriechbaum M., Karrer G., Welk E., Elsässer M., Donath T.W. and Otte A. (2011) Biological flora of Central Europe: Colchicum autumnale L. Perspectives in Plant Ecology, Evolution and Systematics 13, 227–244. Winter S., Jung L.S., Eckstein R.L., Otte A., Donath T.W. and Kriechbaum M. (2014) Control of the toxic plant Colchicum autumnale in semi-natural grasslands: effects of cutting treatments on demography and diversity. Journal of Applied Ecology DOI: 10.1111/1365-2664.12217. Winter S., Penker M., Kriechbaum M. (2011) Integrating farmers' knowledge on toxic plants and grassland management: a case study on Colchicum autumnale in Austria. Biodiversity and Conservation 20, 1763–1787. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 420 Issues regarding the genus Fusarium in permanent grassland Nedělník J.1, Palicová J.2, Hortová B.2 and Strejčková M.1 1 Agricultural Research, Ltd., Troubsko, Czech Republic, 2 Crop Research Institute Prague, Czech Republic Corresponding author: nedelnik@vupt.cz Abstract The objective of this study was to detect Fusarium spp. in plant material and the air of permanent grasslands. The number of Fusarium colony-forming units (CFU) was significantly higher in the green mass during August than in May. The occurrence of Fusarium spp. was greater in the green mass at Lukov than at Závišice. In the air above the grasslands, the amount of Fusarium CFU was higher in August, but the differences were statistically non-significant. Keywords: Fusarium, permanent grassland Introduction Permanent grasslands (hereafter PG), consisting of meadows and pastures, are part of the agricultural land resource along with arable land, hop fields, orchards and gardens. They have characteristic plant communities that are dominated by species of the Poaceae, often with significant representation of clover, and also with the presence of sedges, rushes, bulrushes and a number of other plants. In addition to their economic functions, PG have a great number of non-productive functions, such as in conserving the biological diversity and ecological stability of a region, protecting surface water and groundwater resources, protecting the soil, and others (Rychnovská, 2008). Fusarium spp. are among the most serious contaminants of PG (regarding haylage produced from them), as these have allergenic effects on people as well as animals. They include significant producers of mycotoxins, in particular trichothecenes (deoxynivalenol, nivalenol, T2 toxin), which can have immunosuppressant, oestrogenic and teratogenic effects. In addition to these effects, many Fusarium spp. cause a number of respiratory tract illnesses, such as chronic colds, asthma and other allergenic reactions. Among the most commonly occurring diseases in grasses are pink snow mould, the source of which is the species Microdochium nivale (also known as Fusarium nivale), and Fusarium ear blight, caused by the species F. culmorum, F. graminearum, F. poae and F. avenaceum. The pathogens survive in the forms of mycelia or conidia in infected plants, soil, and dead plant matter. Many species also create chlamydospores, which aid in their survival under adverse conditions (Cagaš and Macháč, 2005; Smiley et al., 2007). Fusarium spp. are found in abundance not only in plant material of PG, but also in the air. During operations to maintain PG, the particles, spores, and hyphae of Fusarium spp. are released into the air and can induce mycoses (external and internal), allergic reactions, chronic colds and asthma. Species from the genera Alternaria, Cladosporium and Penicillium can also be found in the air. In addition, many of these fungi are among the producers of mycotoxins, which can have carcinogenic and mutagenic effects (Cvetnić and Pepeljnajk, 1997; Kakde and Kakde, 2012). This work presents recent results from studies examining the presence of microscopic fungi in permanent grasslands with a focus on Fusarium spp. The presence of Fusarium spp. was monitored in plant material and in the air at two separate locations in the Czech Republic. Materials and methods During 2013, samples of green mass and air were taken twice (before the first and second cuttings, in May and August respectively) at two locations: Závišice (Nový Jičín District, 281 m a.s.l.) and Lukov (Zlín District, 334 m a.s.l.). A phytosociological survey was undertaken Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 421 during sample collection. The green mass was always collected into a sample mixed from 4 randomly selected places within the monitored PG and, at the same time, Petri dishes with potato dextrose agar (PDA) were exposed in order to capture spores from the air (one Petri dish per plot; exposure time 10 min). The plant material was cut into segments (2–5 mm), which were subsequently placed in Petri dishes containing PDA. Each sample was processed into six dishes with agar medium (10 segments per dish). At the same time, the dry matter of the plant mass in the sample was determined and the number of detected colony-forming units (CFU) calculated per gram of dry matter. In addition to the detailed quantitative evaluation as to the presence of Fusarium spp., other genera of microscopic fungi were determined morphologically in the samples. Isolated Fusarium spp. were preserved in a collection of cultures for subsequent species identification, which is still ongoing. The collected data were evaluated statistically using the Statistica 10 program. Analysis of variance (ANOVA) was run, along with Cochran’s and Bartlett’s tests to verify homogeneity of variance, and, in cases of non-homogeneity of variances, Box–Cox transformation was used to normalize the data. Results and discussion In studying the green mass, a greater number of Fusarium CFU were discovered in the second collection at both monitored locations, as shown in Table 1. Table 1. Presence of Fusarium fungi (CFU/g dry matter) in green mass, May and August, sampled at Závišice and Lukov May August Závišice Lukov Závišice Lukov 31 48 470 1328 The difference between the individual collections in May and in August was statistically significant (P = 0.04), as can be seen in Table 2. Table 2. Statistical evaluation of Fusarium fungi occurrence in green mass (ANOVA) Effect SS Degrees of freedom MS F P Intercept 65.42674 1 65.42674 4151,306 0.009880 Locality 0.19226 1 0.19226 12,199 0.177523 Term of sampling 5.54513 1 5.54513 224,937 0.042384 Error 0.01576 1 0.01576 There were fewer Fusarium spp. present in Závišice, although the difference was not statistically significant (P = 0.18). This difference was clearly due to the composition of the PG. While grasses constitute only 23% of the PG in Závišice, they are the dominant component (60%) of the grassland in Lukov. From the preliminary results, it is clear that the green mass contains the species F. acuminatum, F. crookwellense, F. oxysporum, F. poae, F. sporotrichioides and F. subglutinans. It is known from the professional literature that fungi of the genus Fusarium occur abundantly in Poaceae plants and that, in some cases, they may be active even in wintering grasslands. Inch and Gilbert (2003) recorded during their work a high percentage of Fusarium fungi in grasslands (up to 62%), with the most frequently occurring species being F. graminearum, F. oxysporum and F. sporotrichioides. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 422 In the study of the airborne mycoflora, the statistics show that the quantity of captured Fusarium CFU was independent of the location and time of collection. While at both monitored locations more Fusarium CFU / m2 of area were found at the August collection than at the May collection (Table 3), the differences were not statistically significant. Table 3. Presence of Fusarium fungi (CFU / m²) in air, May and August, sampled at Závišice and Lukov May August Závišice Lukov Závišice Lukov 2516 2359 2988 3145 The results of the analyses of the green mass and of the air are in correspondence with one another: in both cases the presence of Fusarium spp. was higher in August than in May. It is known from the literature that the highest concentrations of micromycete spores in air are found in summer and autumn. This fact remains true even during the kind of weather characterized by above-average precipitation that occurred during August at both locations. Preliminary species determination of Fusarium spp. confirmed the presence of F. avenaceum, F. culmorum, F. oxysporum and F. poae in the air. Ctvetnić and Pepeljnjak (1997) had undertaken a study of toxigenic species in the atmosphere and found, in addition to Fusarium spp., species from the genera Alternaria, Cladosporium and Penicillium. Similar airborne mycoflora, including Fusarium spp., were also found above mulched grasslands by Hortová et al. (2013). Members of the genera of microscopic fungi in the green mass and air of PG were quite similar at both monitored locations. In addition to Fusarium spp., the relative abundance of the class Zygomycetes; the genera Alternaria, Cladosporium and Penicillium; and, rather sporadically, the genera Epicoccum and Stemphylium were found. All the isolated genera of fungi are clinically significant species (Hoog et al., 2000). Acknowledgements This contribution was created with support from the Ministry of Agriculture of the Czech Republic, project NAZV QI111C016. References Cagaš B. andMacháč J. (2005) Protecting grasses against diseases, pests, weeds and abiotic damage, Kurent s.r.o., Vrbenská 127/93, 370 01 České Budějovice, Czech Republic, 96 pp. Cvetnić Z. and Pepeljnjak S. (1997) Distribution and mycotoxin-producing ability of some fungal isolates from the air. Atmospheric Environment 31, pp. 491-495. Hoog G.S., Guarro J., Gené J., Figueras M.J. (2000) Atlas of clinical fungi, 2nd ed., CBS-KNAW Fungal Biodiversity Centre, Utrecht, Uppsalalaan 8, The Netherlands, 1126 pp. Hortová B., Palicová J., Strejčková M., Cholastová T. and Nedělník J. (2013) Effect of mulching intensity on the quantity of microscopic fungi in the air. Úroda scientific supplement 61, pp. 46-49. (In Czech). Inch S. and Gilbert J. (2003) The incidence of Fusarium species recovered from inflorescences of wild grasses in southern Manitoba. Canadian Journal of Plant Pathology 25, pp. 379-383. Kakde U.B. and Kakde H.U. (2012) Incidence of post-harvest disease and airborne fungal spores in a vegetable market. Acta Botanica Croatica 71, pp. 147-157. Rychnovská M. (2008) Substances and energy in riverscape grasslands. In: Štěrba O., Kubíček F., Měkotová J., Bednář V., Řehořek V., Šarapatka B. and Rychnovská M. (eds) Riverscapes and their ecosystems, Palacký University Olomouc, Biskupské náměstí 1, 771 11 Olomouc, Czech Republic, pp. 222-228. Smiley R.W., Denoeden P.H. and Clarke B.B. (2005) Compendium of turfgrass diseases, 3rd ed., American Phytopathological Society, 3340 Pilot Knob Road, St. Paul, MN 55121, USA, 167 pp. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 423 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 424 Theme 3 ‘Novel uses of grassland, including bioenergy and biorefining’ Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 425 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 426 Theme 3 invited papers Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 427 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 428 Novel products from grassland (bioenergy & biorefinery) Thumm U., Raufer B. and Lewandowski I. University of Hohenheim, Department of Crop Science, Biobased Products and Energy Crops, 70593 Stuttgart, Germany. Corresponding author: ulrich.thumm@uni-hohenheim.de Abstract Permanent grassland can be classified into three types with regard to management intensity and productivity. High-yielding, intensively-managed, agriculturally-improved grasslands (type 1) provide biomass with qualities suitable for anaerobic fermentation and biorefinery. Its use in biogas plants is a well-established practice. Green biorefinery offers several options for biobased products, but is still in a pilot stage. Grassland biomass from semi-natural grasslands (type 2) or from landscape conservation areas (type 3) have higher lignin contents and requires pre-treatment before the fermentation or hydrolysis process can break down the cellulosic fibre. This strongly lignified biomass is also suitable for combustion or pyrolysis. However, for these conversion technologies the main problems are often unfavourably high mineral and ash contents. These novel pathways for grassland biomass can help to preserve the multifunctionality of grassland in the landscape, even without traditional livestock farming. However, the costs of harvesting, transporting, conservation and conversion of grassland biomass, especially from low-yielding areas, can be too high for a cost recovery without subsidies. Keywords: biomass quality, biogas, combustion, pyrolysis, enzymatic hydrolysis, biorefinery Introduction At present, most grassland biomass is used for dairy farming. However, during the last three decades there have been notable changes in grassland use. Enhanced animal performance with increased milk and meat production have changed ruminant diet composition. On the one hand, the demand for roughage has fallen due to a higher percentage of concentrates in feed and, on the other hand, the yield of grassland is rising due to intensification. At the same time, national demand for ruminant products has decreased. In reality, there is already a surplus of grassland in many developed countries. For a long time the only possibility to use grassland biomass for human demands was the conversion by animals to meat, milk and wool. Nowadays there are new options to use grassland biomass for energy and as biobased products. Therefore the use of grassland is becoming partly independent from livestock production. Energy generated from biomass plays a key role in the European Union’s current strategies to deal with climate change and to meet the increasing energy demands (Ericsson and Nilsson, 2006). Biomass is the most multifunctional form of renewable energy, partly due to the wide range of materials which are classified as such (Roddy, 2012). Estimates indicate a surplus grassland area of 9.2-14.9 x 106 ha in the European Union for the year 2020 (Prochnow et al., 2009b). According to various studies (Rösch et al., 2007; Hartmann et al., 2011) there is a grassland surplus of about 20% in Bavaria and Baden-Württemberg. In the EU the area of surplus grassland represents about 13-22% of permanent grassland. Thus grassland could provide a proportion of 16-19% of the energy crop potential and 6-7% of the total bioenergy potential without encroaching on land needed for animal feed (Prochnow et al., 2009b). Therefore, surplus grassland holds a remarkable bioenergy potential and biomass supply for energy production is regarded as one suitable way to make use of it. In the context of a rapidly developing bioeconomy there is an increasing demand for sustainably produced biomass not only for the renewable energy sector but also for the production of Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 429 biobased products. On account of its high biodiversity, its function in carbon and nitrogen storage and its ability to fulfill several ecosystem functions (Werling et al., 2014) and at the same time produce biomass for different utilization options (Rose, 2012), grassland has the potential to become an important resource for sustainable biomass supply. In this paper an overview is given of grassland types, categorized according to their potential for biomass production, and the alternative uses are investigated. Characterization of grassland types for alternative uses The characteristics and yield of grassland depend on site conditions and management intensity (Dierschke et al., 2002). During the last 50 years, most grassland areas with extensive grazing and mowing managements were converted into productive agriculturally-improved grasslands with high mowing frequencies (three to six cuttings) and stocking rates (Rose, 2012). The natural production of plant biomass reflects the broad range of geological, pedological and in particular the hydrological conditions in different grassland types (Dierschke et al., 2002). The enhanced understanding of soil and plant nutrition, plant physiology and cultivar improvement has led to a considerable increase in grassland productivity (Hopkins and Wilkins, 2006). Agricultural intensification has also led to higher fertilizer application, in particular nitrogen and phosphorus (Rose, 2012). High-production swards can have an annual yield of 10-–12 t DM ha-1 y-1 for grazed swards and 15-20 t DM ha-1y-1 for cut swards under good conditions. Low-production swards only yield 2-3 t DM ha-1 y-1 (Peeters, 2009). According to the definition of Huyghe et al. (2014) permanent grasslands in Europe can be classified into three main types: 1) agriculturally-improved permanent grasslands, 2) seminatural grasslands and 3) permanent grasslands no longer used for production (Figure 1). Agriculturally-improved grasslands on favourable site conditions are frequently defoliated and regularly fertilized. The management is optimized for high yields and high biomass quality. Semi-natural grasslands are lower-yielding permanent grassland areas dominated by naturally occurring grass communities. Site conditions are usually not sufficient for a more intensive grassland management. Grasslands no longer used for production are marginal areas unsuitable for agricultural production. They generally have a very low yield potential and are managed for conservation purposes. Grassland management not only determines the yield but also the quality of the harvested biomass (see Figure1). Different pathways for novel uses of grassland biomass have different requirements on biomass quality. Demands on biomass quality may change in future on account of the continued development of conversion technologies. The most relevant parameters are, among others, protein, lignin and ash content of the biomass. Generally, the nitrogen content increases with higher N fertilizer doses and early cutting. The lignin content is lower in intensively managed grassland (Gützloe et al., 2014). In addition to management intensity, the botanical composition of grassland has a relevant impact on biomass quality. Concentrations of N, S, K, Ca, Mg are lower in grasses than in forbs (Tonn et al., 2010). Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 430 Site conditions soil, climate cutting frequency Management nitrogen fertilization Agriculturallyimproved grasslands Semi-natural grasslands Grasslands no longer used for production (landscape conserv.) Conservation Biomass yield silage (hay) hay (silage) hay harvestable biomass protein Biomass characteristics minerals, ash lignin Potential use: biogas, biorefinery, biogas, combustion, pyrolysis, hydrolysis hydrolysis combustion, pyrolysis Figure 1. Characterization of permanent grassland types and pathways for biomass use Alternative uses for biomass from permanent grassland Figure 1 provides an overview of potential uses of grassland biomass including energetic and material applications. Grassland biomass can be processed to biofuels or bioenergy via biogas production, combustion, pyrolysis and gasification or enzymatic hydrolysis and subsequent fermentation to ethanol. For material uses or combined energetic and material uses, grassland biomass is processed through biorefinery. These options are described in the following sections. Biogas Today the conversion of grassland biomass by anaerobic fermentation to methane is a wellestablished technology. In the last decade political frameworks such as the German ‘Renewable Energy Law’ have supported the economic and technical progress. Maize is the most common biogas crop on account of its high yields and easy cultivation, storage and processing. In Germany the area used for maize increased by 53% in the last ten years, of which one-third is used for biogas production (FNR, 2013). The consequences are high prices of forage and rising costs for land rental in areas with biogas plants (Habermann and Breustedt, 2011). Prochnow et al. (2009b) described the potential of biomass from agriculturally-improved grasslands for methane production. Effective biomass conversion depends on the botanical composition of grassland, cutting period and feedstock-specific methane yields. Shorter cutting periods increase the specific methane yields of grassland biomass. The methane yield per area Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 431 is mainly determined by biomass yield. The influence of grass species on methane yields seems to be secondary. However high nitrogen content in grassland biomass can lead to difficulties in the control and stability of the anaerobic fermentation due to high ammonium concentrations in the reactor (Andrade and Weber, 2013). Therefore a high percentage of legumes in grassland biomass and protein-rich biomass from early cuts are inappropriate for fermentation. The timing of the first cut is relevant for the biogas yield. It should not be before the vegetation stage ‘ear emergence’. A first cut taken too early reduces the methane yield per hectare (Amon et al., 2007). Consequently the intensity of grassland management can be lower when producing feedstock for biogas plants than in dairy production. A lower cutting frequency also reduces biomass production costs (Messner et al., 2011). Energy and CO2 balances for power generation in a CHP plant using grassland biomass showed the highest net energy yield and CO2-equivalent reduction with two cuts per year (Gützloe et al., 2014). Low-quality biomass from semi-natural grassland or grass from landscape management can be reasonably used for fermentation if the first cut is up to late summer. Herrmann et al. (2013) found that methane yields from semi-natural grassland biomass decreased with later harvest from 309 lN kg-1 organic dry matter in May to 60 lN kg-1 in February of the following year. The economic feasibility of biogas production from landscape management grass is only given with low supply and investment costs, suitable concepts for biogas production and use, additional income from the sale of heat, and subsidies for land use (Blokhina et al., 2011). In recent years mechanical, thermal, chemical, and enzymatic pre-treatment techniques have been tested to improve the fermentation of material with higher lignocellulose content, such as straw or late-cut grassland biomass (Weiland, 2010; Michalska et al., 2012). The use of pretreated hay from semi-natural grasslands instead of silage as feedstock in biogas plants can offer new options for the storage and utilisation of grassland biomass. For example, Bauer et al. (2014) found a slight increase in the methane yield of hay after steam explosion treatment. The future development of these methods and subsequent assessment of their economic suitability and sustainability will reveal the opportunities that these upcoming techniques offer for the use of lignocellulosic grassland biomass for fermentation. Another approach is the IFBB technique (integrated generation of biogas and solid fuel from biomass) for the energetic use of ensilaged biomass from semi-natural grassland (Wachendorf et al., 2009; Hensgen et al., 2014). After a hydrothermal conditioning process, the silage is mechanically dehydrated. The press cake, the solid fibrous fraction, can be used for combustion and the press fluid, the fraction containing easily fermentable compounds, for biogas production. Biomass conversion by IFBB has been shown to be more economically profitable than mulching as a landscape management system for the preservation of biodiversity (Blumenstein et al., 2012). Fermentation of biomass produces digestates as residues. The fertilizer value of the digestate is directly linked to the feedstock used, since there are no nutrient losses during fermentation to biogas. Degradation of organic compounds during the fermentation process leads to an increased ammonium content and therefore a higher proportion of plant-available nitrogen in the digestates (Möller and Müller, 2012). Slurry from livestock farming is frequently used for co-digestion. Slurry has a notable energy potential and helps to stabilize the fermentation process in the biogas plant (Wall et al., 2013). Combustion Grassland biomass as a solid fuel for combustion is less favourable than biomass from perennial energy grasses and wood. It has higher contents of ash, nitrogen, sulphur, potassium and chlorine. These can lead to problems in the combustion process, such as corrosion or slagging, and to environmentally critical emissions, such as NOx, SO2, HCl or dioxin (Lewandowski and Kicherer, 1997). Therefore late-harvested, highly-lignified and low-ash biomass is more Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 432 suitable for combustion (Prochnow et al., 2009a; Iqbal and Lewandowski, 2014). Hay from late-harvested semi-natural grasslands and grasslands no longer used for production (landscape management) is more suitable for combustion than hay from agriculturally-improved grasslands due to its higher contents of lignin and lower contents of ash, potassium and chlorine. Thus extensive grassland management systems with one late cut and low fertilization are preferable when using grass as a solid biofuel. A life-cycle analysis showed that the combustion of biomass from semi-natural grassland was carbon negative and delivered a net energy gain even with low biomass yields (Tonn et al., 2009). However the critical factor is the nitrogen content as this is responsible for NOx emissions and N losses from the ecosystem. Unlike other plant nutrients, for N no recycling is possible to the soil by ash (Tonn et al., 2010). Hay is very voluminous and requires compaction before it is worth transporting. From this point of view small local heating plants are more suitable than transportation of the biomass to use in a larger plant. On the other hand, co-firing in power plants with the latest filter technology can be a good pathway to use low-quality biomass. A further option is the pelleting of hay to provide a marketable product (Pilz et al., 2013; Cherney and Verma, 2013). Mixed biomass pellets can help to solve grassspecific combustion problems (Nunes et al., 2014). Pyrolysis (gasification) Thermochemical conversion by pyrolysis converts up to 75% of the energy content of plant material to a bio-oil or bio-crude (Carpenter et al., 2014). These products can be further refined in a subsequent step into various base chemicals or a synthetic transportation fuel (FischerTropsch liquids). There are several innovative ways of synthesis gas production and utilization which are being developed and tested in various pilot plants (Rauch et al., 2014). In the bioliq® gasification process the lignocellulosic biomass is liquefied in regional plants to produce an energy-dense bio-slurry for transportation to a central gasification plant (Dahmen et al., 2012). Separating the process in two steps can open up possibilities for the economic processing of plant material such as grassland biomass which is not worth being transported over long distances. Relevant feedstock quality parameters are calorific value, moisture content, proportion of fixed carbon, ash content and the minerals nitrogen, sulfur and chlorine (Robbins et al., 2012). Therefore the biomass commonly used in pilot plants is wood, on account of its low ash content. Gasification technologies for low quality biomass with higher ash values are not yet well established and require further adaption (Dahmen et al., 2012). The quality demands on the feedstock are similar to those for combustion. Hence biomass with high lignin content from semi-natural grasslands and grasslands no longer used for production is most suitable for pyrolysis. Enzymatic hydrolysis After pre-treatment of the biomass (physical, chemical) to make the lignocellulosic material accessible for the hydrolysis process, the cellulose chains are broken down by cellulase enzymes into fermentable sugars (Brodeur et al., 2011). As such, the glucose yield is directly dependant on the cellulose content of the biomass (Tutt and Olt, 2011). The sugars can then be fermented to bioethanol in a subsequent step. There are several approaches for the adaption and optimization of these processes to reduce the energy demand and enzyme costs (Martín and Grossmann, 2012). Methods for the assessment of the suitability of lignocellulosic biomass are based on fermentation tests (Anderson et al., 2010). For grassland biomass there are no data available for specific quality demands. Until now more homogeneous feedstocks than grassland biomass are usually used due to their easier process optimization. Biorefinery The aim of biorefinery is the integrated production of chemicals, fibre, food and feed products together with energy and biofuel. Green biorefinery technologies are used to exploit fresh grass Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 433 or silage-based feedstock systems. The first pilot plants started around 15 years ago in Switzerland (Grass, 2004). There are currently several green biorefinery activities being pursued in the Netherlands, Denmark, Austria and Germany (Mandl, 2010). An overview of the potential use of grassland biomass in biorefineries was given by O’Keeffe (2010) for Ireland. To date green biorefinery projects are mostly in the pilot phase and not yet economically sustainable. However the increase in activities in the last decade points to the opportunities and potentials of this pathway (Cherubini, 2010). Biorefinery can be defined as ‘the sustainable processing of biomass into a spectrum of marketable products and energy” (IEA Bioenergy, 2009). In green biorefinery the primary step is the mechanical fractionation of the biomass by pressing. The green juice (fresh material) or brown juice (silage) extracted and press cake recovered are used for the further processing of various products and energy (Figure 2). Decomposition methods (enzymatic, fermentative, hydrolytic, thermal, chemical) are sometimes applied before fractionation. The freshly-pressed green juice contains several components including proteins, lipids, glycoproteins, lectins, sugars, amino acids, dyes, minerals and enzymes. In addition silage juice contains relatively high concentrations of lactic acid, which can be used for the production of plastics, and inorganic salts (Kamm et al., 2010). The press cake is a fibrous fraction, which can be used as raw material for products such as insulation material, peat substitution products in horticulture, bio-composites, pulp and paper, and thermoplastics. Grass Feedstock (fresh & silage) Mechanical fractionation / pressing Press Juice Press cake Fresh juice (FJ), silage juice (SJ) Feed product Biogas → protein concentrate (FJ) → amino acid concentrate (SJ) → CHP → power/heat → treatment → BioCNG or gas net Amino acid (AA) Fibre applications → high-grade amino acid mixtures (SJ) (nutrition supplements, bodycare products..) → insulation material → fibreboards → horticultural → pulp & paper…. Lactic acid (LA) Solid fuel → fermentation (FJ) → separation (SJ) → combustion Direct use of juice (FJ, SJ) Biofuels → fermentation medium → feed (co-substrate) → biogas → 2nd generation ethanol via hydrolysis of lignocelluloses → gasification → FT synthesis Figure 2. Possible biorefinery products from different feedstock fractions (Mandl, 2010) Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 434 One main challenge is the generation of pure chemical products such as ethanol, organic acids or amino acids by different separation technologies (Kamm et al., 2010). Residues remaining after refining processes can be used as feedstock for energy production (biogas, combustion) or as fertilizer. The quality demands on the raw material depend on the intended product. For example, in a biorefinery model producing protein feed for animals and insulation material, the most important quality parameters for grass feedstock are fibre and protein content (Grass, 2004). Biomass from semi-natural grasslands normally has a too low content of exploitable ingredients and is not suitable for conservation as silage. The low content of exploitable components in fresh biomass from semi-natural grasslands normally renders it unsuitable for biorefinery. Therefore it is to be expected that only biomass from agriculturally-improved grasslands can be used in biorefinery processes. The achievement of dedicated quality demands can require special management or botanical composition of the meadows. The economic situation of green biorefinery systems has been evaluated in some studies (O'Keeffe, 2010; Höltinger et al., 2014). The variable operating costs, which include the costs of feedstock fractioning and downstream processes, are most important for the economy of green biorefineries. One question which remains to be answered is how to optimize the input volume of biorefinery facilities. High capacity plants have lower processing costs per biomass unit, but high costs for feedstock logistics. Also the energy balance between consumption and production will be a determining factor in the success of a biorefinery (O’Keeffe et al., 2012). For sustainable and truly green final products the use of sustainable biomass is not enough; protection of the environment requires methods and techniques with minimized impact on the environment (Cherubini, 2010). Conclusions Permanent grassland can supply biomass for energetic and material uses. However, grassland is also the main source of feed for ruminants. Therefore the availability and cost of grassland biomass for novel uses depend on the economic benefits of grass-based livestock systems and landscape and agrarian structure. For grassland areas with conditions which make harvesting of biomass by machinery problematic, traditional pasture systems are more suitable. Different grassland types provide biomass of varying quality. Biomass from agriculturallyimproved grasslands is easily degradable and appropriate for fermentation in biogas plants or conversion to biobased products in biorefineries. Fermentation of grassland biomass is widespread in several European countries. New challenges are the adaption of the process for biomass with higher lignin contents from semi-natural grasslands. It is possible to use this biomass and also that from landscape conservation for combustion in adapted heating plants. New options for producing second-generation biofuel by pyrolysis and enzymatic hydrolysis are still in a developmental stage. There are promising perspectives for converting low-quality grassland biomass in advanced facilities in future. Further challenges are the integration of these new value chains in the landscape without the loss of other functions of grassland. References Amon T., Amon B., Kryvoruchko V., Machmüller A., Hopfner-Sixt K., Bodiroza V., Hrbek R., Friedel J., Pötsch E., Wagentristl H., Schreiner M. and Zollitsch W. (2007) Methane production through anaerobic digestion of various energy crops grown in sustainable crop rotations. Bioresource Technology 98, 3204-3212. Anderson W.F., Dien B.S., Jung H.-J.G., Vogel K.P. and Weimer P.J. (2010) Effects of forage quality and cell wall constituents of bermuda grass on biochemical conversion to ethanol. BioEnergy Research 3, 225-237. Andrade D. and Weber A. (2013) Biogas from grassland biomass: Recommendations for the optimization of the degradation process. Landtechnik 68, 339-343. Bauer A., Lizasoain J., Theuretzbacher F., Agger J. W., Rincón M., Menardo S., Saylor M.K., Enguídanos R., Nielsen P.J., Potthast A., Zweckmair T., Gronauer A. and Horn S.J. (2014) Steam explosion pretreatment for enhancing biogas production of late harvested hay. Bioresource Technology 166, 403-410. Blokhina Y.N., Prochnow A., Plöchl M., Luckhaus C. and Heiermann M. (2011) Concepts and profitability of biogas production from landscape management grass. Bioresource Technology 102, 2086-2092. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 435 Blumenstein B., Bühle L., Wachendorf M. and Möller D. (2012) Economic assessment of the integrated generation of solid fuel and biogas from biomass (IFBB) in comparison to different energy recovery, animal-based and nonrefining management systems. Bioresource Technology 119, 312-323. Brodeur G., Yau E., Badal K., Collier J., Ramachandran K.B. and Ramakrishnan S. (2011) Chemical and physicochemical pretreatment of lignocellulosic biomass: A review. Enzyme Research 2011, 1-17. Carpenter D., Westover T.L., Czernik S. and Jablonski W. (2014) Biomass feedstocks for renewable fuel production: a review of the impacts of feedstock and pretreatment on the yield and product distribution of fast pyrolysis bio-oils and vapors. Green Chemistry 16, 384. Cherney J.H. and Verma V.K. (2013) Grass pellet quality index: A tool to evaluate suitability of grass pellets for small scale combustion systems. Applied Energy 103, 679-684. Cherubini F. (2010) The biorefinery concept: Using biomass instead of oil for producing energy and chemicals. Energy Conversion and Management 51, 1412-1421. Dahmen N., Henrich E., Dinjus E. and Weirich F. (2012) The bioliq® bioslurry gasification process for the production of biosynfuels, organic chemicals, and energy. Energy, Sustainability and Society 2:3, 1-44. Dierschke H., Briemle G. and Kratochwil A. (2002) Kulturgrasland: Wiesen, Weiden und verwandte Staudenfluren. Ulmer, Stuttgart, Germany, 239 pp. Ericsson K. and Nilsson L.J. (2006) Assessment of the potential biomass supply in Europe using a resourcefocused approach. Biomass and Bioenergy 30, 1-15. FNR (2013) Maisanbau in Deutschland: Fachagentur Nachwachsende Rohstoffe e.V., Mediathek. Grass S. (2004) Utilisation of grass for production of fibres, protein and energy. OECD Publication Service, Paris, September 2004, 169-177. Gützloe A., Thumm U. and Lewandowski I. (2014) Influence of climate parameters and management of permanent grassland on biogas yield and GHG emission substitution potential. Biomass and Bioenergy 64, 175-189. Habermann H. and Breustedt G. (2011) Einfluss der Biogaserzeugung auf landwirtschaftliche Pachtpreise in Deutschland. German Journal of Agricultural Economics 60, 85-100. Hartmann S., Diepolder M. and Lichti F. (2011) Grünland als Biogassubstrat. Biogas Fachforum Bayern I-12/2011. Hensgen F., Bühle L., Donnison I., Heinsoo K. and Wachendorf M. (2014) Energetic conversion of European semi-natural grassland silages through the integrated generation of solid fuel and biogas from biomass: Energy yields and the fate of organic compounds. Bioresource Technology 154, 192-200. Herrmann C., Prochnow A., Heiermann M. and Idler C. (2013) Biomass from landscape management of grassland used for biogas production: effects of harvest date and silage additives on feedstock quality and methane yield. Grass and Forage Science, early view, DOI: 10.1111/gfs.12086. Höltinger S., Schmidt J., Schönhart M. and Schmid E. (2014) A spatially explicit techno-economic assessment of green biorefinery concepts. Biofuels, Bioproducts and Biorefining 8, 325-341. Hopkins A. and Wilkins R.J. (2006) Temperate grassland: key developments in the last century and future perspectives. Journal of Agricultural Science 144, 503-523. Huyghe C., De Vliegher A., van Gils B. and Peeters A. (2014) Grasslands and herbivore production in Europe and effects of common policies. Éditions Quae, Versailles Cedex, France, 288 pp. IEA BIOENERGY (2009) Biorefineries: adding value to the sustainable utilisation of biomass. http://www.ieabioenergy.com/wp-content/uploads/2013/10/Task-42-Booklet.pdf. Accessed 14 Mar. 2014. Iqbal Y. and Lewandowski I. (2014) Inter-annual variation in biomass combustion quality traits over five years in fifteen Miscanthus genotypes in south Germany. Fuel Processing Technology 121, 47-55. Kamm B., Gruber P.R. and Kamm M. (2010) Biorefineries - industrial processes and products: Status quo and future directions. Wiley-VCH, Weinheim, Germany. Lewandowski I. and Kicherer A. (1997) Combustion quality of biomass: practical relevance and experiments to modify the biomass quality of Miscanthus x giganteus. European Journal of Agronomy 6, 163-177. Mandl M.G. (2010) Status of green biorefining in Europe. Biofuels, Bioproducts and Biorefining 4, 268-274. Martín M. and Grossmann I.E. (2012) Energy optimization of bioethanol production via hydrolysis of switchgrass. AIChE Journal 58, 1538-1549. Messner J., Nussbaum H. and Elsaesser M. (2011) Specific utilization intensity of permanent grassland used as biogas-substrate. Grassland Science in Europe 16, 392-394. Michalska K., Miazek K., Krzystek L. and Ledakowicz S. (2012) Influence of pretreatment with Fenton’s reagent on biogas production and methane yield from lignocellulosic biomass. Bioresource Technology 119, 72-78. Möller K. and Müller T. (2012) Effects of anaerobic digestion on digestate nutrient availability and crop growth: A review. Engineering in Life Sciences 12, 242-257. Nunes L., Matias J. and Catalão J. (2014) Mixed biomass pellets for thermal energy production: A review of combustion models. Applied Energy 127, 135-140. O’Keeffe S., Schulte R.P.O., Sanders J.P.M. and Struik P.C. (2012) II. Economic assessment for first generation green biorefinery (GBR): Scenarios for an Irish GBR blueprint. Biomass and Bioenergy 41, 1-13. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 436 O'Keeffe S.M. (2010) Alternative use of grassland biomass for biorefinery in Ireland: A scoping study. PhD Thesis, Wageningen University, The Netherlands 202 pp. Peeters A. (2009) Importance, evolution, environmental impact and future challenges of grasslands and grasslandbased systems in Europe. Grassland Science 55, 113-125. Pilz A., Döhling F., Kirsten C., Weller N. and Zeng T. (2013) Pelletizing and energy related use of hay from landscape preservation. Landtechnik 68, 349-352. Prochnow A., Heiermann M., Plöchl M., Amon T. and Hobbs P.J. (2009a) Bioenergy from permanent grassland – A review: 2. Combustion. Bioresource Technology 100, 4945-4954. Prochnow A., Heiermann M., Plöchl M., Linke B., Idler C., Amon T. and Hobbs P.J. (2009b) Bioenergy from permanent grassland – A review: 1. Biogas. Bioresource Technology 100, 4931-4944. Rauch R., Hrbek J. and Hofbauer H. (2014) Biomass gasification for synthesis gas production and applications of the syngas. Wiley Interdisciplinary Reviews: Energy and Environment 3, 343-362. Robbins M.P., Evans G., Valentine J., Donnison I.S. and Allison G.G. (2012) New opportunities for the exploitation of energy crops by thermochemical conversion in Northern Europe and the UK. Progress in Energy and Combustion Science 38, 138-155. Roddy D.J. (2012) Biomass and Biofuels - Introduction. In: Sayigh, A. (ed) Comprehensive Renewable Energy. Volume 5, Elsevier. Rösch C., Raab K., Skarka J. and Stelzer V. (2007) Energie aus dem Grünland - eine nachhaltige Entwicklung? Wissenschaftliche Berichte FZKA 7333. Forschungszentrum Karlsruhe. Rose L. (2012) The influence of land-use intensification and biodiversity on grassland biomass, water use and plant functional traits. PhD thesis, University of Göttingen, Göttingen, 147 pp. Tonn B., Thumm U. and Claupein W. (2009) Life-cycle analysis of heat generation using biomass from seminatural grasslands in Central Europe. Grassland Science in Europe 14, 284-296. Tonn B., Thumm U. and Claupein W. (2010) Seminatural grassland biomass for combustion: influence of botanical composition, harvest date and site conditions on fuel composition. Grass and Forage Science 65, 383-397. Tutt M. and Olt J. (2011) Suitability of various plant species for bioethanol production. Agronomy Research 9 (Special Issue 1) 261-267. Wachendorf M., Richter F., Fricke T., Graß R. and Neff R. (2009) Utilization of semi-natural grassland through integrated generation of solid fuel and biogas from biomass. I. Effects of hydrothermal conditioning and mechanical dehydration on mass flows of organic and mineral plant compounds, and nutrient balances. Grass and Forage Science 64, 132-143. Wall D.M., O’Kiely P. and Murphy J.D. (2013) The potential for biomethane from grass and slurry to satisfy renewable energy targets. Bioresource Technology 149, 425-431. Weiland P. (2010) Biogas production: current state and perspectives. Applied Microbiology and Biotechnology 85, 849-860. Werling B.P., Dickson T.L., Isaacs R., Gaines H., Gratton C., Gross K.L., Liere H., Malmstrom C.M., Meehan T. D., Ruan L., Robertson B. A., Robertson G. P., Schmidt T. M., Schrotenboer A. C., Teal T. K., Wilson J. K. and Landis D.A. (2014) Perennial grasslands enhance biodiversity and multiple ecosystem services in bioenergy landscapes. Proceedings of the National Academy of Sciences 111, 1652-1657. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 437 Grasslands for forage and bioenergy use: traits and biotechnological implications Barth S.1, Jones M.2, Hodkinson T.2, Finnan J.1, Klaas M.1 and Wang Z.-Y.3 1 Crops Environment and Land Use Programme, Teagasc Crops Research Centre, Carlow, Ireland. 2 School of Natural Sciences, Trinity College Dublin, Ireland. 3 Forage Improvement Division, The Samuel Roberts Noble Foundation, Ardmore, Oklahoma, USA Corresponding author: Susanne.Barth@teagasc.ie Abstract Growing energy crops on marginal lands will avoid land use needed for food crops and use this unused marginal land which is unfavourable for food production. Perennial rhizomatous grasses (PerRhG) have the advantage that several of the species of interest for biomass and bioenergy production can be grown on marginal land types. However, it will be necessary to improve management practices and to select for plant genotypes which are able to tolerate the stressful conditions they are exposed to. This review looks at the current state of genomics and biotechnology improvements of PerRhG species and the morphological and physiological traits of perennial rhizomatous grasses that could potentially be modified to maximize biomass production on marginal land. The traits include aspects of crop phenology, canopy and leaf photosynthesis, biomass partitioning, nutrient and water use efficiency, cold and salt tolerance, and moisture and chemical content. It is proposed that newly developed biotechnological methods combined with high-throughput plant phenotyping offer opportunities to rapidly select new genotypes that could achieve economic yields on large areas of marginal land. Keywords: biomass production, perennial grasses, plant traits, environmental stress, biotechnology Introduction The development of forage and bioenergy grass species which have a suite of desirable physical and chemical traits while maximizing biomass yields is challenging. Achieving this will depend on identifying the fundamental constraints on productivity and addressing these constraints using modern genomic breeding tools to identify and exploit suitable crops and to optimize the management of these crops to maximize biomass production. Biomass production, water and nitrogen use, tolerance to biotic and abiotic stress and low soil fertility are critical factors in selection of ideal bioenergy feed stocks (Bressan et al., 2011; Xin and Wang, 2011). Ideally, these perennial grasses should be also doing well on marginal lands that are unfavourable for food production and that are currently unused or underutilized. The aim therefore is to identify and breed the most highly productive plant species that can be grown on the various types of marginal land and to optimize production practices. Marginal land is land that is of poor quality for agriculture and which yields poor economic returns for the farmer. The cultivation of grass on degraded or exhausted agricultural soils can serve to restore soil organic carbon and improve its physical properties (Potter et al., 1999). Crops growing on marginal land are subjected to a range of abiotic stresses, including shortage/excess of soil water, poor nutrient availability, salinity and high and low temperatures, which reduce their yield to below their potential. Because these stresses vary from area to area a diverse range of feedstocks suitable for different niches are necessary and feed into breeding objectives (Yuan et al., 2011). In the case of several field crops, such as maize and oilseed rape, they have undergone much genetic improvement, and significant increases in yields have been achieved in the last fifty years. This selective breeding has gone hand in hand with improved management practices and Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 438 in many cases the relative contributions of management and breeding have been about equal (Richards, 2000; 2004). However, in the case of several forages and second generation biofuel crops such as perennial rhizomatous grasses (PerRhGs) most attention until now has been on improved management. The candidate PerRhGs are either largely undomesticated and have not undergone the selective breeding that characterises our current major food crops, or they have been selected to maximize quality and productivity for grazing animals (Heaton et al., 2008; Vermerris, 2008). The first, and arguably the most important, management choice affecting biomass production on marginal land is the choice of species or variety to be used for biomass production. Irrespective of the physical stress which renders land marginal, biomass production can be maximized by utilizing the genetic variability that exists among and within species and choosing the species/genotype likely to give the highest yield under a specific physical stress. Conventional biotechnology to improve traits General principles In most breeding schemes (with exceptions, e.g. mutation breeding) at least one round of recombination is required to introduce novel genetic variation by rearrangement of alleles by meiotic recombination. This recombination step with succeeding evaluation and selection defines one cycle in breeding. One breeding cycle can take between two to six years to complete. For most grass breeding schemes formal grass breeding began only ~ 1900 in Europe and USA, and about 20 to 25 breeding cycles have been completed in 100 years. Compared to maize breeding, which often uses two to three seasons per year, in winter nurseries over 100 breeding cycles have been completed since modern breeding started and has led, in a selection experiment, to a 4-fold increase in oil content (Dudley 2007). Thus there is a huge untapped potential for breeding of forage and bioenergy grasses. This untapped potential could be much more utilised in conjunction with Marker Assisted Breeding (MAS) Strategies to breed by phenotypic selection and with molecular marker support at the same time. The objective of molecular markers in breeding programmes should be to describe the effect of the marker on phenotypic trait performance. Unfortunately the situation for many agriculturally important traits is far from trivial since many of these traits are controlled by quantitative trait loci (QTL) which can affect up to a few hundred genes. A strategic decision needs to be taken if selection is carried out with a few markers, which has been termed marker assisted selection, or if many markers at a time, which are spread out throughout the entire genome, are being selected, which is termed genomic selection (GS). Genomic selection follows a marker index score which sums up all individual marker effects to arrive at a total score. A prerequisite to implement any type of marker assisted selection is that advances in genomics of the species under question have been made. These advances include the generation of collections of different types of molecular markers, expressed sequences tags (ESTs), annotated transcriptomes, complete genomes or shut gun sequences of genomes. GS requires a substantial amount of markers fairly evenly spread and at high density across the genome. Examples of state of art in genomics for one perennial grass species: Miscanthus Some recent research has used gene expression analysis to understand phenotypic variation in Miscanthus using methods such as RNA-Seq. Chouvarine et al. (2012) used transcriptome sequencing of rhizome samples to generate an exome sequence database for Miscanthus complete with gene ontology functional annotations. Barling et al. (2013) also generated a comprehensive expressed sequence tag (EST) catalogue using RNA-Seq that was predicted to represent a high proportion of the Miscanthus transcriptome using comparisons to sorghum Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 439 gene models. They also analysed expression profiles in rhizomes characterized in the spring compared to the autumn to reveal biological pathways that exhibit altered regulation. Some candidate gene work has also been undertaken to understand variation in important ligninrelated genes (Suman et al., 2011). Other research has focused on generating genetic linkage maps of Miscanthus that are needed for several applications such as QTL analysis and MAS. High-resolution maps based on sequence markers allow the use of QTL accessible from other grass species through alignment based on syntenic relationships (Ma et al., 2012). However, such maps have only recently been produced. Higher resolution genetic maps of Miscanthus species based on DNA sequence markers have recently been generated using next-generation sequencing technology (Ma et al., 2012; Swaninathan et al., 2012). This has allowed for data transferability and several comparative genomic analyses. Association mapping (linkage disequilibrium mapping) is a method of mapping QTLs that takes advantage of historic linkage disequilibrium to link phenotypes to genotypes (Myles et al., 2009). The genome is sampled for markers (such as SNPs) and associations statistically detected between markers and a particular phenotype. Associations are independently verified to show that they 1) directly contribute to the trait of interest, or 2) are linked to a QTL that contributes to the trait of interest. Association mapping in the form of a genome-wide association study (GWAS) is an advance on standard association mapping and has been most widely applied to the study of human disease, cattle breeding and more recently to plants including Miscanthus (Slavov et al., 2013). Genetic modification (GM) to improve traits General principle GM crops are obtained by genetic engineering techniques. Genetic engineering offers an opportunity to generate unique or novel genetic variations that could not otherwise be obtained by conventional breeding (Wang and Ge, 2006). Furthermore, transgenesis has become an indispensible tool for understanding basic biological questions (Dixon et al., 2007). The generation of transgenic plants, coupled with selection, has led to the development of transgenic cultivars in several major cash crops, such as maize, soybean and cotton. These transgenic cultivars have been widely adopted in many parts of the world. Transgenic technology has also been used to improve forage and turf species (Wang and Ge, 2006) and, more recently, to improve biofuel production from potential bioenergy grasses. Genetic transformation methods for perennial rhizomatous grasses Although the first proven transgenic forage grass was obtained by direct gene transfer to protoplasts (Wang et al., 1992), this method is rarely used nowadays because protoplasts are difficult to handle and regenerate. The two main methods currently used to produce transgenic plants are: microprojectile bombardment (biolistics) and Agrobacterium-mediated transformation. Biolistics utilizes high-velocity gold or tungsten particles to deliver exogenous DNA into the plant cells for stable transformation. Agrobacterium tumefaciens is a soil bacterium that harbors a tumor-inducing (Ti) plasmid. For the purposes of genetic transformation, the T-DNA of the Ti plasmid is replaced with foreign genes intended for incorporation into the plant cells. Compared to biolistic transformation, Agrobacterium-mediated transformation allows stable integration of the transgene into the plant genome in a relatively lower copy number and generally leads to fewer rearrangements and more stable transgene expression over generations (Dai et al., 2001; Hu et al., 2003). Agrobacterium-mediated transformation was originally used for dicotyledonous plants. Starting from the late 1990s, Agrobacteria have been successfully used to transform grasses. To date, transgenic plants have been obtained for many forage, turf and bioenergy species, such as tall fescue, meadow fescue, red fescue, perennial ryegrass, Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 440 Italian ryegrass, creeping bentgrass, Kentucky bluegrass, orchardgrass, bahiagrass, zoysiagrass, switchgrass (Wang and Ge, 2006; Wang and Brummer, 2012). Example of genetic improvement of perennial grasses by transgenic technology Cell wall composition directly affects processing properties of lignocellulosic biomass. Due to the association of lignin with cellulose and hemicellulose, cell wall materials are largely recalcitrant to hydrolytic enzymes. To solve the cell wall recalcitrance problem, a direct and effective approach is to modify lignin content/composition by down-regulation of the enzymes involved in lignin biosynthesis (Hisano et al., 2009). In switchgrass, the down-regulation caffeic acid methyltransferase (COMT) led to reduced lignin content and altered lignin composition, improved forage quality, and up to a 38% increase in ethanol yield using conventional biomass fermentation processes (Fu et al., 2011). Furthermore, the transgenic lines required less severe pretreatment and less cellulase for equivalent ethanol yield compared to unmodified switchgrass (Fu et al., 2011). The transgenic COMT plants were transferred to the field. Under field conditions, transgenic switchgrass were phenotypically normal and continued to show reduced lignin content, increased sugar release and improved ethanol yield (Baxter et al., 2014). Biomass yield is an important but complex trait for both forage and bioenergy purposes. The identification and manipulation of major regulatory genes that govern the expression of a group of downstream genes provide an effective way to improve complex traits (Fu et al., 2012). In recent years, miRNAs have emerged as a prominent class of gene regulatory factors (Zhang et al., 2006). Plant miR156 is a family of small, non-coding, endogenous RNAs with a relatively high expression level in the juvenile phase of plants (Matts et al., 2010). A miR156b precursor was overexpressed in switchgrass (Fu et al., 2012). The degree of morphological alterations of the transgenic switchgrass depends on miR156 level. Relatively low levels of miR156 overexpression were sufficient to increase biomass yield while producing plants with normal flowering time. Moderate levels of miR156 led to improved biomass but the plants were nonflowering. These two groups of plants produced 58-101% more biomass yield compared with the nontransgenic control (Fu et al., 2012). The improvement in biomass yield was mainly because of the increase in tiller number (Fu et al., 2012). The non-flowering phenotype offers an effective approach for transgene containment in grasses. Regulatory problems/concerns A major limitation for deployment of transgenics is the complicated GMO regulatory processes (Wang and Brummer, 2012). Despite the wide adoption and the beneficial economic and environmental impacts of major transgenic crops (e.g. corn, soybean, cotton, canola), it has been extremely difficult to deregulate and commercialize forage, turf and bioenergy crops. These crops do not enter the food chain directly; their potential environmental or ecological impacts are the main focus of risk assessment. Most widely grown forage, turf and bioenergy species are highly self-incompatible and outcrossing. Pollen-mediated transgene flow is a major concern for such outcrossing species. The impacts of regulations on research and environmental studies of transgenic forage, turf and bioenergy crops have been summarized by Strauss et al. (2010) and Wang and Brummer (2012). Several biological containment measures have been developed or proposed to control transgene flow. Such measures include male sterility, seed sterility, maternal inheritance, delayed flowering or non-flowering (Wang and Brummer, 2012). In addition, the cisgenic or intragenic approach may provide a cost-effective way for genetic engineering and commercialization of grasses (Wang and Brummer, 2012). Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 441 Traits to be improved and targets for improvement Physiological and morphological traits Plants have evolved functional traits or features which reflect their ecological strategies in a particular environment. A plant functional trait is therefore any morphological, physiological or phenological feature, measureable in the plant at the cell to the whole organism level, which potentially affects its fitness (Perez-Harguindeguy et al., 2013). Phenology (the timing of developmental stages in the plant life cycle), leaf area dynamics, radiation interception and utilization, and crop growth and partitioning are major processes that determine the harvestable biomass of energy crops. These processes are regulated by morphological and physiological traits and maximising yield is about optimising these processes. Yield potential for crops grown on marginal land is reduced below the optimal by stresses and the aim here is to identify plant traits which either allow avoidance or increase tolerance of these stresses. Breeding new varieties which possess these traits depends on identifying measurable physiological traits which can be used as selection criteria. The aim is to select appropriate species and genotypes which are adapted to local soil and climatic conditions. Selection criteria need to be based on the triple goals of maximizing productivity, minimizing inputs and maximizing utilization for energy production. Some of the traits of particular interest in recently instigated breeding programmes are drought tolerance (Clifton-Brown et al., 2002), frost tolerance (Clifton-Brown and Lewandowski, 2000), maintenance of growth at low temperature (Farrell et al., 2006), chemical composition (Lewandowski et al., 2003), resistance to pests and diseases (CliftonBrown et al., 2008), altering plant architectural features such as dwarf structure and erect leaves (Zhu et al., 2010) and differences in photosynthetic capacity (Carver and Hocking, 2001). After Karp and Shield (2008) are three main challenges in achieving yield improvement. (1) There should be a reduction in the thermal threshold for growth of the canopy leaves which extends the growing season. Second, above-ground biomass should be increased without depleting the below-ground biomass so much that there are insufficient reserves available for the next years’ growth. Third, above-ground biomass should be increased without restricting growth due to excess water depletion and developing water stress. Traditional plant breeding, selection and hybridization techniques are slow and for some PerRhGs there is a limited availability of germplasm. Miscanthus x giganteus, for example, is a sterile triploid, which is normally propagated from rhizome pieces. There has, however, been some long-term conventional breeding of another C4 grass, switchgrass (Casler, 2012), which has produced large yield gains. In the future, new biotechnological routes may produce even greater improvements. Genetically modified (GM) energy crop species may be more acceptable to the public than are GM food crops, particularly in Europe (Koh and Ghazoul, 2008), but there are still concerns about the environmental impact of such plants including gene flow from nonnative to native plant relatives. Consequently non-GM biotechnologies may be more attractive. Initially it is likely that the use of molecular biology will focus on the use of molecular markers that can be used in the rapid screening of germplasm within the breeding population. However, linking these molecular markers to complex traits such as yield is difficult because yield is controlled by many genes. With the advent of cost-effective and rapid sequencing technologies, there is currently an expansion of knowledge of genes and their expression profiles in potential biofuel crops including PerRhGs (Ma et al., 2012; Swaminathan et al., 2012). In addition, there has been recent development of automated greenhouse systems for high-throughput plant phenotyping which allows the non-destructive screening of plants over a period of time by means of image acquisition techniques. This involves applying sophisticated image analysis algorithms (Hartmann et al., 2011) to extract several phenotypic parameters related to growth, yield and stress tolerance and as a consequence reduces the tedious and time consuming manual analysis of large numbers of phenotypes. Marker-assisted selection (MAS) is the process of Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 442 using DNA, biochemical or morphological markers as indirect selection criteria for selecting agriculturally important traits in crop breeding. The use of MAS for manipulating simple/qualitative traits is relatively straight forward but its use for the improvement of complex/polygenic traits, including plant tolerance/resistance to abiotic stresses is far more complicated. Quantitative trait loci (QTL) mapping and MAS methods will utilize high-density single nucleotide polymorphism (SNP) data including genic SNPs. A recent development for MAS is genome-wide selection that uses high-density genotypic SNP data to find haplotype blocks associated with traits. This approach has been used successfully in cattle and should offer high potential in forage and bioenergy grasses (Nakaya and Isobe, 2012). The utilization of these techniques will allow us to identify in PerRhGs the physiological and morphological traits that are most important in determining yield and stress tolerance of grasses grown on marginal land and to suggest how they might be used as selection criteria in screening and breeding programmes. Biomass production and partitioning of biomass to different parts of the plant can be viewed as three steps. In the first step the processes involved are canopy development and light interception. In the second step intercepted light energy is converted into biomass, and in the third step biomass is partitioned into different part of the plant (Zhu et al., 2010). The yield that a crop can achieve under optimal management practices and in the absence of biotic and abiotic stresses is dependent on the amount of incident radiation, the efficiency of interception of that radiation by the canopy, and the efficiency of photosynthetic conversion of light energy into dry matter (Jones, 2011). In the following sections we review the key morphological and physiological processes which determine these traits. Phenological stages of development Phenology describes the plant life cycle and it is linked to environmental drivers, particularly accumulated temperature measured as cumulative growing degree days (GDD) (Larcher, 2003). This measure of thermal period is computed as an average of daily maximum and minimum temperatures above the base temperature for growth of a particular species. The GDD are determined by seasonal and interannual variations in weather and the phenological stages are assessed in terms of GDD (Larcher, 2003). Grasses progress through a series of phenological stages during the year and to maximize production it may be necessary to select for optimal phasing of these. For example, extending the reproductive phase may result in greater stem extension, and therefore increased standing biomass. Extending crop duration is the simplest genetic way to increase total photosynthesis during the growing season leading to higher crop biomass and yield. This is because longer crop duration increases the solar radiation interception during the crop growth period. However, not only can the duration of growth be manipulated to increase biomass and yield, but also its timing. Providing there are no other limiting factors, full light interception should be achieved by the time that daily solar radiation has reached its maximum. Therefore the need is to manipulate phenology to better match periods of highest solar radiation. Timing may also be important to increase yield and biomass in relation to water supply. Therefore, if water is a major limiting factor it may be beneficial to maximise growth when temperatures are lower and vapour pressure deficit is low as this is when transpiration is lowest and water use efficiency is highest (Richards, 2004). Flowering time has a key role in determining the quantity and quality of the harvested biomass and considerable variation in flowering time among genotypes of some species has been demonstrated (Jensen et al., 2011). Early flowering shortens the effective length of the growing season, thus reducing biomass production but where flowering does not occur before the autumn frosts, the onset of senescence and remobilisation of nutrients appears less efficient in some species such as Miscanthus (Clifton-Brown et al., 2008). This can result in higher ash contents on combustion which are naturally associated with higher uptakes of elements such as Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 443 N, P and K. Late flowering has also been associated with higher over-winter losses of Miscanthus genotypes in the first year following planting (Clifton-Brown et al., 2008). Canopy and leaf photosynthesis Leaf area development and canopy light interception is dependent on the number of leaves that develop per unit ground area and the rate at which leaves expand. The appearance of leaves and their rate of expansion are driven largely by temperature. Significant variation in the temperature threshold for leaf appearance and subsequent leaf expansion has been demonstrated in Miscanthus genotypes (Clifton-Brown and Jones, 1997; Farrell et al., 2006). The effect of expansion at lower temperatures is to allow earlier development of the canopy in spring and increase the length of the growing season. This increases the total amount of radiation intercepted by the canopy during the growing season and has the potential to increase yield significantly if leaf photosynthetic activity is maintained. In a study comparing the canopy duration among a diverse collection of 244 Miscanthus genotypes, Robson et al. (2013) found that yield was positively correlated with canopy duration as were both early establishment and later senescence. The process of photosynthesis is pivotal to the production of biomass and there is considerable variation in photosynthetic rates among crops and genotypes within them, particularly at low temperature (Naidu and Long, 2004). For example, Miscanthus can maintain 80% higher photosynthetic light-use efficiency than maize when grown at 14/11oC (day/night) (Naidu et al., 2003). The photosynthetic conversion efficiency of light energy is the combined gross photosynthesis of all the leaves within the canopy, less all the plant respiratory losses. It might be assumed therefore that there is a correlation between maximum leaf photosynthesis and yield. Unfortunately, most research has failed to clearly demonstrate this correlation although it has been shown recently that increasing photosynthesis using elevated CO2 concentration does indeed result in higher yields under field conditions (Long et al., 2006). Under the right circumstances it would appear therefore that there is a yield advantage associated with selecting for higher leaf photosynthetic rates. However, the measurement of photosynthesis is plagued by problems of integration over the life cycle of the plant (Horton, 2000). Spot measurements vary with leaf age, leaf position, time of day, light intensity, plant health and development stage. It is therefore important to obtain an integrated assessment of photosynthesis of the whole plant rather than of single leaves. Biomass partitioning and assimilate distribution In PerRhGs used as energy crops virtually all of the above-ground material is harvested. However, the below-ground roots and rhizomes are important stores of carbon which enable the spring growth of shoots, and nutrients which are recycled throughout the growing season into the shoots. The dynamics of carbon movement between shoots and rhizomes and roots show a seasonal pattern (Heaton et al., 2009). Consequently, when photosynthate is in short supply, for example during rapid stem extension in the spring, the reserves of carbon in the rhizomes may be important to support continued growth. PerRhGs develop extensive root systems which, in effect, transfer atmospheric carbon into soil biomass on their senescence and death. In perennial systems where the soil is not ploughed this results in substantial sequestration of the carbon in the soil. A number of studies have shown that annual rates of soil carbon sequestration range between 1-10 t/ha/yr (Zimmerman et al., 2012). Variation in sequestration appears to be influenced by the depth of the roots as well as the soil characteristics. Nutrient use efficiency In PerRhGs, nutrient use efficiency (NUE) is strongly influenced by the annual recycling of nutrients from the above-ground biomass to the rhizome system. The nutrient that most Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 444 frequently limits growth of energy crops is nitrogen. NUE is normally expressed as the ratio of increase in plant biomass to increase in plant nitrogen over the growing season. In energy crops this approximates to the ratio of biomass to nitrogen content at the end of the growing season (Xu et al., 2012). Nitrogen use efficiency can be improved at three levels in perennial species. First, NUE is enhanced by increasing the amount of biomass produced per unit of nitrogen invested into the photosynthetic apparatus, which is different in C3 and C4 plants (Jones, 2011). Second, NUE can be enhanced by increasing the fraction of soil nutrients that are taken up by the plant (Lewandowski and Schmidt, 2006). Third, NUE is enhanced by increasing the fraction of nitrogen translocated out of the leaf canopy and stems during senescence; translocated nitrogen can then be stored in the rhizomes for use in the following year (Beale and Long, 1997). This recycling process has the effect of reducing, and in some cases eliminating the need to use nitrogenous fertilizers (Christian et al., 2008). The recycling occurs late in the growing season, during the senescence of the crop and is most effective when the crop is harvested in the late autumn or winter. Consequently, late senescence or stay-green characteristics of PerRhGs are clearly linked with higher nutrient use efficiencies. Limiting steps in plant nitrogen metabolism are different under high and low N levels. At high N inputs major variation in NUE is due to differences in N uptake but at low N variations in NUE are dependent on changes in N utilization and in the remobilization of N. For PerRhGs cultivated on marginal land it is therefore more likely that the most important traits to focus on will be those associated with high N utilization and remobilization capacity. Water use efficiency The water use efficiency (WUE) of a biomass crop is the yield of harvested product achieved from the water made available to the crop through precipitation and/or irrigation (Richards, 2004). It is known for several crop species that genotypes with higher WUE have higher concentrations of the stable isotope 13C (Richards, 2004). This is because the enzymatic processes of photosynthesis discriminate against 13C in the atmosphere and in genotypes with higher WUE this discrimination is diminished since a higher proportion of CO2 in cells is utilized and consequently more 13CO2 is fixed (Condon et al., 2004). Plants are analysed for variability in the relative ratio of 13C/12C expressed as  13C relative to a standard and the stable isotope ratios can be used for screening cultivars/genotypes for higher WUE. Drought and flooding tolerance Globally, drought-induced losses in crop yield exceed losses from all other causes (Jones, 1988, Richards, 2004). Drought tolerance is the ability of plant tissues to maintain metabolism at low water potentials brought about by water stress. Drought tolerance can be achieved by careful selection of genotypes, breeding new varieties using methods such as MAS and using management practices such as the exogenous application of hormones or osmoprotectants (Farooq et al., 2009). Tolerance is often associated with the accumulation of solutes within cells which has the effect of retaining cell turgor (Jones, 1988). Flooding tolerance is little studied although flooding is a widely distributed problem and may become increasingly important under climate change with the increase in frequency of extreme events. Flooded environments present challenges for biomass production as crop growth can be reduced in wet conditions when oxygen concentrations in the root zone are reduced. However, some species such as Phalaris arundinacea are well adapted to waterlogged conditions and Etherington (1984) found differences in flooding tolerance among genotypes of Dactylis glomerata. More recently, some PerRhGs have been shown to have the potential capacity to mitigate flooding. Macleod et al. (2013) showed that novel Festulolium cultivars (Lolium perenne x Festuca pratensis) can significantly reduce run-off during flooding compared to commercially used Lolium and Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 445 Festuca cultivars and this was shown to be due to intense initial root growth followed by rapid senescence at depth. Cold tolerance There are vast areas of marginal land in the cool-temperate climatic zone, particularly in Eurasia and North America, which could potentially be cultivated to meet a future demand for productive bioenergy crops. Biomass production from grasses cultivated in environments with low temperatures may be reduced by a combination of a shorter growing season and poor growth at low temperatures. Gudliefsson et al. (1986) found significant differences in cold hardiness and ice tolerance when ten pasture grasses were tested in controlled environments. Timothy (Phleum pratense) and Kentucky bluegrass (Poa pratensis) proved to be the most resilient to these stresses while orchardgrass (Dactylis glomerata) and reed canary grass (Phalaris arundinacea) proved least resilient to these stresses. Indigenous varieties of Lolium perenne were found to be more winter hardy compared to more exotic varieties even if they came from colder climates (Lorenzetti et al., 1971). Clifton-Brown and Lewandowski (2000) demonstrated cold tolerance variation between species and genotypes of Miscanthus. Thus, selection of genotypes with enhanced low temperature tolerance represents an effective strategy for maximising biomass production in environments affected by cold stress. Salt tolerance Salt stress is today in Europe mainly a problem in southern and south-eastern regions such as Hungary, Romania and Spain (Tóth et al., 2008), but the areas may increase with climate change. Extensive genetic diversity for salt tolerance exists in plant taxa from halophytes which are native flora of saline environments to glycophytes which are salt sensitive or hypersensitive. Most crops are glycophytes (Yokoi et al., 2002). However, there is considerable variation in salt tolerance both within and between species of glycophytes, including grasses. If rankings of salt tolerance in different publications are compared, the order of the species varies, as most rankings are based on one or a few varieties per species. Great variation, not only between species but also between varieties within each species, has been observed (Grare, 2010). Different management practices can be used to alleviate the effects of salt; however, the most practical and, arguably, the most cost-effective means of managing salt stress is the choice of salt tolerant genotypes. Acknowledgements This study was co-funded from the European Union FP7 ‘GrassMargins’ project, FP7 KBBE2011-5-289461. References Barling A., Swaminathan K., Mitros T., James B.T., Morris J., Ngamboma O., Hall M.C., Kirkpatrick J., Alabady M., Spence A.K., Hudson M.E., Rokhsar D.S. and Moose S.P. (2013) A detailed gene expression study of the Miscanthus genus reveals changes in the transcriptome associated with the rejuvenation of spring rhizomes. BMC Genomics 14: 864 doi: 10.1186/1471-2164-14-864 Baxter H., Mazarei M., Labbe N., Kline L., Cheng K., Windham M., Mann D., Fu C., Ziebell A., Robert R., Rodriguez M., Davis M., Mielenz J., Dixon R.A., Wang Z-Y. and Stewart C.N. Jr. (2014) Two-year field analysis of reduced recalcitrance transgenic switchgrass. Plant Biotechnology Journal doi: 10.1111/pbi.12195 Beale C.V. and Long S.P. (1997) Seasonal dynamics of nutrient accumulation and partitioning in the perennial C 4grasses Miscanthus x giganteus and Spartina cynosuroides. Biomass and Bioenergy 12, 419-428. Bressan R.A., Reddy M.P., Chung S.H., Yun D.J., Hardin L.S. and Bohnert H.J. (2011) Stress-adapted extremophiles provide energy without interference with food production. Food Security 3, 93-105. Carver P. and Hocking T.J. (2001) Photosynthetic responses of Miscanthus genotypes. In: Biomass and Energy Crops II. (Eds. Bullard M.J. et al.) Aspects of Applied Biology 65, 215-222. Casler M.D. (2012) Switchgrass breeding, genetics, and genomics. In: A. Monti (ed.) Switchgrass, green energy and technology. Springer-Verlag, London, pp. 29-53. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 446 Christian D.G., Richie A.B. and Yates N.E. (2008) Growth, yield and mineral content of Miscanthus x giganteus grown as a biofuel for 14 successive years. Industrial Crop Production 28, 320-327. Clifton-Brown J.C. and Jones M.B. (1997) The thermal response of leaf extension rate in genotypes of the C4 grass Miscanthus: an important factor in determining the potential productivity of different genotypes. Journal of Experimental Botany 48, 1573-1581. Clifton-Brown J. and Lewandowski I. (2000) Overwintering problems with newly established Miscanthus plantations can be overcome by identifying genotypes with improved rhizome cold tolerance. New Phytologist 148, 287-294. Clifton-Brown J., Lewandowski I., Bangerth F. and Jones M.B. (2002) Comparative responses to water stress in stay-green, rapid- and slow senescing genotypes of the biomass crop, Miscanthus. New Phytologist 154, 335-345. Clifton-Brown J., Chiang Y.C. and Hodkinson T.R. (2008) Miscanthus: genetic resources and breeding potential to enhance bioenergy production. In: Vermerris W (ed.), Genetic Improvement of Bioenergy Crops p295-308. Springer, New York. Condon A.G., Richards R.A., Rebeteke G.J. and Farquhar G.D. (2004) Breeding for high water use efficiency. Journal of Experimental Botany 55, 2247-2480. Chouvarine P., Cooksey A.M., McCarthy F.M., Ray D.A., Baldwin B.S., Burgess S.C. and Peterson D.G. (2012) Transcriptome-based differentiation of closely-related Miscanthus lines. PLOS ONE 7: e29850. Dai S., Zheng P., Marmey P., Zhang S., Tian W., Chen S., Beachy R.N. and Fauquet C. (2001) Comparative analysis of transgenic rice plants obtained by Agrobacterium-mediated transformation and particle bombardment. Molecular Breeding 7, 25-33. Dixon R.A., Bouton J.H., Narasimhamoorthy B., Saha M., Wang Z-Y. and May G.D. (2007) Beyond structural genomics for plant science. Advances in Agronomy 95,77-161. Dudley J.W. ( 2007) From Means to QTL: The Illinois long-term selection experiment as a case study in quantitative genetics. Crop Science 47( Supplement 3), 20-31. Etherington J.R. (1984) Relationship between morphological adaptation to grazing, carbon balance and waterlogging tolerance in clones of Dactylis glomerata L. New Phytologist 98, 647-658. Farrel A.D., Clifton-Brown J.C., Lewandowski I., Jones M.B. (2006) Genotypic variation in cold tolerance influences the yield of Miscanthus. Annals of Applied Biology 149, 337-345. Farooq M., Wahid A., Kobayashi D.F. and Basra S.M.A. (2009) Plant drought stress: mechanisms and management. In: Lichtfouse E., Navarrete M. et al. (eds) Sustainable Agriculture. Springer, The Netherlands. Grare M.A.R. (2010) Variability of salinity response in Miscanthus sinensis. MSc, Wagenigen University. The Netherlands. Gudliefsson B.E., Andrews C.J. and Bjornsson H. (1986) Cold hardiness and ice tolerance of pasture grasses grown and tested in controlled environments. Canadian Journal of Plant Science 66, 601-608. Fu C., Mielenz J.R., Xiao X., Ge Y., Hamilton C., Rodriguez M., Chen F., Foston M., Ragauskas A., Bouton J., Dixon R.A. and Wang Z-Y. (2011) Genetic manipulation of lignin reduces recalcitrance and improves ethanol production from switchgrass. Proceedings of the National Academy of Sciences of the United States of America 108, 3803-3808. Fu C., Sunkar R., Zhou C., Shen H., Zhang J.Y., Matts J., Wolf J., Mann D., Stewart C.N., Tang Y. and Wang ZY. (2012) Overexpression of miR156 in switchgrass (Panicum virgatum L.) results in various morphological alterations and leads to improved biomass production. Plant Biotechnology Journal 10, 443–452. Hartmann A., Czauderna T., Hoffmann R., Stein N. and Schreiber F. (2011) HTPheno: an image analysis pipeline for high throughput phenotyping. BMC Bioinformatics, 12, 148 doi:10.11.18/1471-2105-12-148. Heaton E.A., Dohleman F.G. and Long S.P. (2008) Meeting US biofuel goals with less land: the potential of Miscanthus. Global Change Biology 14, 2000-2014. Heaton E.A., Dohleman F.G. and Long S.P. (2009) Seasonal nitrogen dynamics of Miscanthus x giganteus and Panicum virgatum. GCB-Bioenergy 1, 297-307. Hisano H., Nandakumar R. and Wang Z-Y. (2009) Genetic modification of lignin biosynthesis for improved biofuel production. In Vitro Cellular and Developmental Biology – Plant 45, 306-313. Horton P. (2000) Prospects for crop improvement through genetic manipulation of photosynthesis: morphological and biochemical aspects of light capture. Journal of Experimental Botany 51, GMP Special Issue, 475-485. Hu Y., Jia W., Wang J., Zhang Y., Yang L. and Lin Z. (2005) Transgenic tall fescue containing the Agrobacterium tumefaciens ipt gene shows enhanced cold tolerance. Plant Cell Reports 23, 705-709. Jensen E., Farrar K., Thomas-Jones S., Hastings A., Donnison I. and Clifton-Brown J. (2011) Characterization of flowering time diversity in Miscanthus species. GCB-Bioenergy, 3, 387-400. Jones M.B. (1988) Water stress. In: Jones M.B. and Lazenby A. (eds.) The Grass Crop – the physiological basis of production, pp.205-236. Chapman and Hall, London. Jones M.B. (2011) C4 species as energy crops. In: Raghavenra A.S. and Sage R.F. (eds.) C4 photosynthesis and related CO2 concentrating mechanisms. Springer, Dordecht. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 447 Karp A. and Shield I. (2008) Bioenergy from plants and the sustainable yield challenge. New Phytologist 179, 1532. Koh L.P. and Ghazoul J. (2008) Biofuels, biodiversity and people: understanding the conflicts and finding opportunities. Biological Conservation 141, 2450-2460. Larcher W. (2003) Physiological plant ecology. (4th Edition) Springer-Verlag, Berlin. Lewandowski I., Clifton-Brown J.C., Andersson B., Basch G., Christian D.G., Jorgensen U., Jones M.B., Riche A.B., Schwartz K.U., Tayebi K., Tiexera F. (2003) Environment and harvest time affects the combustion qualities of Miscanthus genotypes. Agronomy Journal 95, 1274-1280. Lewandowski I. and Schmidt U. (2006) Nitrogen, energy and land use efficiencies of Miscanthus, reed canary grass and triticale as determined by the line boundary approach. Agriculture, Ecosystems and Environment 112, 335-346. Lorenzetti F., Tyler B.F., Cooper J.P. and Breese E.L. (1971) Cold tolerance and winter hardiness in Lolium perenne. 1. Development of screening techniques for cold tolerance and survey of geographical variation. Journal of Agricultural Science 76, 199-209. Long S.P., Zhu X-G., Naidu S.L. and Ort D.R. (2006) Can improvement in photosynthesis increase crop yields? Plant, Cell and Environment 29, 315-330. Ma X-F., Jensen E., Alexandrov N., Troukhan M., Zhang L., et al. (2012) High resolution genetic mapping by genome sequencing reveals genome duplication and tetraploid genetic structure of the diploid Miscanthus sinensis. PLOS ONE 7(3), e33821. Macleod C.J.A., Humphreys M.W., Whalley W.R., Turner L., Binley A., Watts C.W., Skot L., Joynes A., Hawkins S., King I.P., O'Donovan S., Haygarth P.M. (2013) A novel grass hybrid to reduce flood generation in temperate regions. Scientific Reports 3, DOI: 10.1038/srep 01683. Matts J., Jagadeeswaran G., Roe B.A. and Sunkar R. (2010) Identification of microRNAs and their targets in switchgrass, a model biofuel plant species. Journal of Plant Physiology 167, 896-904. Myles S., Peiffer J., Brown P.J., Ersoz E.S., Zhanga Z., Costicha D.E. and Buckler E.S. (2009) Association mapping: critical considerations shift from genotyping to experimental design. The Plant Cell 21, 2194–2202. Naidu S., Moose S.P., Al-Shoaibi A.K., Raines C.A. and Long S.P. (2003) Cold tolerance of C 4 photosynthesis in Miscanthus x giganteus: adaptation in the amounts and sequence of C4 photosynthetic enzymes. Plant Physiology 132, 1688-1697. Naidu S.L. and Long S.P. (2004) Potential mechanisms of low temperature tolerance of C 4 photosynthesis in Miscanthus x giganteus: an in vivo analysis. Planta 220, 145-155. Nakaya A. and Isobe S.N. (2012) Will genomic selection be a practical method for plant breeding? Annals of Botany 110, 1303-16. Perez-Harguindeguy N., Diaz S., Garnier E., Lavorel S., Poorter H., et al. (2013) New handbook for standardised measurement of plant functional traits worldwide. Australian Journal of Botany 61, 167-234. Potter K.N., Torbert H.A., Johnson H.B. and Tischler C.R. (1999) Carbon storage after long term grass establishment on degraded soils. Soil Science 164, 718-725. Richards R.A. (2000) Selectable traits increase crop photosynthesis and yield of grain crops. Journal of Experimental Botany, 51, GMP Special issue 447-458. Richards R.A. (2004) Physiological traits used in the breeding of new cultivars for water-scarce environments. In: New directions for a diverse planet. Proceedings of the 4th International Crop Science congress. 2004, Brisbane, Australia. Robson P.R.H., Farrar K., Gay A.P., Jensen E.F., Clifton-Brown J.C. and Donnison I.S. (2013) Variation in canopy duration in the perennial biofuel crop Miscanthus reveals complex associations with yield. Journal of Experimental Botany 64, 2373-2383. Slavov G.T., Nipper R., Robson P., Farrar K., Allison G.G., Bosch M., Clifton-Brown J.C., Donnison I.S. and Jensen E. (2014) Genome-wide association studies and prediction of 17 traits related to phenology, biomass and cell wall composition in the energy grass Miscanthus sinensis. New Phytologist 201: 1227–39. Strauss S.H., Kershen D.L., Bouton J.H., Redick T.P., Tan H. and Sedjo R.A. (2010) Far-reaching deleterious impacts of regulations on research and environmental studies of recombinant DNA-modified perennial biofuel crops in the United States. Bioscience 60, 729-741. Suman A., Ali K., Arro J., Parco A.S., Kimbeng C.A. and Baisakh N. (2011) Molecular diversity among members of the Saccharum complex assessed using TRAP markers based on lignin-related genes. BioEnergy Research 5, 197–205. Swaminathan K., Alabady M.S, Varala K, De Paoli E, Ho I, Rokhsar DS, Arumuganathan AK, Ming R, Green PJ, Meyers BC, Moose SP, Hudson ME (2012) A framework genetic map for Miscanthus sinensis from RNAseqbased markers shows recent tetraploidy. BMC Genomics 13, 142. Tóth G., Montanarella L. and Rusco E (2008) Threats to Soil Quality in Europe. European Commission, Joint Research Centre, Institute for Environment and Sustainability. Vermerris W. (2008) Genetic improvement of bioenergy crops. Springer, New York. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 448 Wang Z-Y. and Brummer E.C. (2012) Is genetic engineering ever going to take off in forage, turf and bioenergy crop breeding? Annals of Botany 110, 1317-25. Wang Z-Y. and Ge Y. (2006). Recent advances in genetic transformation of forage and turf grasses. In Vitro Cellular & Developmental Biology – Plant 42, 1-18 Wang Z-Y., Takamizo T., Iglesias V.A., Osusky M., Nagel J., Potrykus I. and Spangenberg G. (1992) Transgenic plants of tall fescue (Festuca arundinacea Schreb.) obtained by direct gene transfer to protoplasts. Nature Biotechnology 10, 691-696. Yokoi S., Bressan R.A. and Hasegawa P.M. (2002) Salt stress tolerance of plants. JIRCAS working report 23, 2533 http://www.researchgate.net,.accessed 11/2/2013 Yuan J.S., Wang X. and Stewart C.N. (2011) Biomass feedstock: diversity as a solution. Biofuels 2, 491-493. Xin Z. and Wang M.L. (2011) Sorghum as a versatile feedstock for bioenergy production. Biofuels 2, 577-588. Xu G., Fan X. and Miller A.J. (2012) Plant nitrogen assimilation and use efficiency. Annual Review of Plant Biology 63, 153-182. Zhang B., Pan X., Cobb G.P. and Anderson T.A. (2006) Plant microRNA: A small regulatory molecule with big impact. Developmental Biology 289, 3-16. Zhu X-G., Long S.P. and Ort D.R. (2010) Improving photosynthetic efficiency for greater yield. Annual Review of Plant Biology 61, 235-261. Zimmerman J., Dauber J. and Jones M.B. (2012) Soil carbon sequestration during the establishment phase of Miscanthus x giganteus: a regional scale study of commercial farms using C-13 natural abundance. GCB Bioenergy 4, 453-461. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 449 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 450 Theme 3 submitted papers Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 451 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 452 Bioenergy potential of meadows of Ukraine Petrychenko V.1, Kurhak V.2 and Rybak S.21 National Academy of Agrarian Sciences, Kyiv, Ukraine, 03022 2 NRC Institute of Agriculture of the NAAS, Chabany, Kyiv region, Ukraine, 08162 Corresponding author: kurgak_luki@ukr.net Abstract We present results of a study of the bioenergy potential of the Ukrainian meadows, an area of about 6.6 million ha plus 10 million ha of fallow lands. The average energy production of unimproved meadows is 2.2 Mg ha-1 of dry biomass or 30.6 J ha-1 of heating (gross) energy. Economically viable annual energy potential of these meadows is 4.22 million Mg or 12% of total biomass potential. Meadow degradation is caused by the overgrowth of thick-stemmed herbs, shrubs and low forest. The ways of improving their energy production are shown further. After including legumes into grass mixtures or adding N120 to sown grass stands, the energy production of sown meadows increased to 5.93-12.07 Mg ha-1 of dry biomass or to 107.4-213,1 J ha-1 of heating energy. That is 1.3-1.9 times more than from sown grass stand, and 2.4-3.5 times more compared to natural grass stand without improvement of Р60К90 under the same conditions. The sown legume-grass is the most productive stand, involving different varieties of Trifolium pratense L., and in the case of liming, Medicago sativa L. Keywords: bioenergy, biomass, meadows, potential, productivity, sown grass stand. Introduction In the process of energy dependence reduction of Ukraine, the development and use of renewable energy resources for biofuels, including plant biomass, is of great importance. Its part in total production of primary energy in Ukraine remains low, at only 1.24%. However, Ukraine has a large economically feasible annual potential of biomass available for energy use, estimated at 34 million Mg of reference fuel, which is 18% of total energy consumption (Geletukha et al., 2013). The Energy Strategy of Ukraine for the Period to 2030 (2006) suggests increasing the part of biomass in total energy consumption up to 10%. Economically viable annual energy potential of perennial herbaceous plant communities is estimated at 7 million Mg of reference fuel, which is 20% of total biomass potential, including 4.22 million Mg (or 12%) from meadows (Kumrgak et al., 2013). Therefore, in the context of stable development, the usage of grassland biomass for energy, which in Ukraine is about 6.6 million ha, demands deserves special attention. In view of the catastrophic reduction in livestock for fodder production, their biomass is hardly ever used. However, until recent time, current studies on energy potential of Ukrainian meadows and development of measures to improve their energy production have not been carried out. Materials and methods We have conducted a study on the energy potential of meadows according to generally accepted field and laboratory methods, using geobotanical, measuring and weighing, calculatingcomparative, chemical and mathematical-statistical methods. Content of heating or gross energy was calculated according to the chemical composition of dry biomass. In terms of heating energy, 1 Mg of reference fuel was compared to 1 Mg of coal. To assess the impact of species composition and variety assortment of legume-grass mixtures in energy production of sown grass stand, we conducted our research during 2011-2013, against the background of annual addition of Р60К90. We conducted research on dry, wet, and drained meadows (village of Lytvynivka, Vyshgorodskyi district, Kyiv region). The soil type we studied is soddy gley, sandy loam, in 0-15 cm layer; pH 4.8; K is 7.3 mg and P is 3.1 mg per 100 g of soil. In the Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 453 research we used new recognized varieties of perennial legumes and grasses of selection of the NSC ‘Institute of Agriculture NAAS’. Weather conditions during the years of study were very contrasting; summer days were dry and hot. The most favorable atmospheric moisture for grassland cenosis was in the year of 2012. Results and discussion The geo-botanical surveys generally showed that unimproved meadows with natural grass swards in their present condition are characterized by low energy production. This is due to the lack of effective measures for their improvement and use. The average energy efficiency of unimproved meadows is 2.2 Mg ha-1 of dry biomass or 30.6 J ha-1 of heating (gross) energy, including Polissia (2.7 Mg ha-1) of dry mass, forest-steppe (3.8) and steppe (1.7 Mg ha-1) (Petrychenko and Kurhak, 2013). The most productive are lowland meadows, and the least productive are upland meadows. Annually the total area of unimproved meadows in Ukraine accumulates approximately 14 million Mg of dry biomass, or 240 million J of heating energy or 7 million Mg of reference fuel, and with fallow lands 22 million Mg and 370 million J and 11 million Mg. respectively. Addition of mineral fertilizers can increase their productivity and increase heating energy output by 2-3 times. In contrast to the 1990s, the degradation of meadows in Ukraine is due to the lack of cattle that would otherwise feed on the grasslands. Meadows are overgrown with thick-stemmed herbs. Those which border on the forest are overgrown with low forest and bushes; after 5-6 years they are of little use for mowing for fodder, but their attraction for energy remains. Among trees and shrubs the most widespread are Salix triandra L., Alnus glutinosa (L.) Gaertn and Betula pendula Roth. Among grasslands are thick-stemmed plants (genera: Cirsium, Rumex, Stenactis, Solidago, Urtica) and in the south Ambrosia, etc.) which can successfully be used for energy purposes. Our studies on the drained flood meadows of Polissia have shown that the creation of sown meadow grass swards with the use of symbiotic nitrogen from perennial legumes, or from nitrogen fertilizers, increases their energy efficiency by several times. A study was conducted of the influence of species and varietal composition of grass components on the productivity of a sown clover-grass mixture. Analysis of the results showed that the output of 1 hectare of dry biomass on average for 2011-2013, against the backdrop of annual addition of Р60К90 to these sown legume-grass swards, ranged from 5.93 to 11.22 Mg, heating energy was from 107.4 to 199.1 J. The most productive appeared to be the sown grass stand involving Dactylis glomerata L. (cv. Kyiv Early and Natalka), Festuca pratensis Huds, (Rosynka and Siveryanka) and Bromopsis inermis (Leys.) Holub (Arsen and Helius) Adding N120 to the control sown grass swards involving Festuca pratensis (cv. Siveryanka) and Lolium perenne L. (cv. Leta) increased meadow productivity from 4.68 to 10.41 Mg ha-1 of dry mass or by 2.2 times. Natural grass swards without improvement on the same background Р60К90 were by 1.7-3.4 times less productive compared to sown grass and clover-grass swards. The second research on the same flood drained meadows when adding Р60К90 included a study on energy production of sown swards with the inclusion of legume-grass mixtures or adding different species and new varieties of perennial legumes to grass plants (Table 1). On these swards, on average for 2012-2013, the addition of Р60К90 obtained 8.32-12.07 Mg per ha of dry biomass, or 145.8-213.1 J of heating energy, which is 1.3-1.9 times greater than from sown grass swards, and by 2.4-3.5 times greater compared to natural grass swards without improvement. The most efficient were sown legume-grass swards with Trifolium pratense, cv. Polyanka or cv. Polisyanka and, on the limed soil, Medicago sativa cv. Olga or cv. Intensive 174. Liming of soil increased productivity of swards by 0.08-2.57 Mg ha-1 of dry mass. The best response to liming was shown by grass mixtures containing Medicago sativa, and the poorest response was from mixtures with Lotus ukrainicus Klok. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 454 Table 1. Energy production of sown legume-grass mixtures in 2012-2013. Kinds and varieties of grass mixtures standard quantity of seed rates, kg ha-1 Dry mass, Mg ha-1 Without liming Heating Reference energy, fuel, J h-1 Mg ha-1 Limed background Dry Heating Reference mass, energy, fuel, Mg ha-1 J ha-1 Mg ha-1 Тrifolium pratense Polyanka, 9 + grass 10.63 186.3 5.32 11.78 200.3 5.89 plants* Тrifolium pratense Polisyanka, 9 + grass 10.60 177.0 5.30 11.65 206.2 5.83 plants* Тrifolium pratense Polyanka, 4.5 + 11.32 199.6 5.66 12.07 213.1 6.04 Polisyanka, 4.5 + grass plants* Medicago sativa Olga, 10 + grass plants* 8.36 147.7 4.18 10.97 192.5 5.49 Medicago sativa Intensive 174, 10 + grass 8.44 150.5 4.22 11.01 195.4 5.51 plants * Medicago sativa Olga, 5 and Intensive 174, 8.72 155.5 4.36 11.14 197.1 5.57 5 + grass plants * Lotus ukrainicus Local, 5 + grass plants * 8.60 150.1 4.30 9.20 163.3 4.60 Lotus ukrainicus Local, 2.5 + Тrifolium 11.01 192.4 5.50 11.09 196.0 5.55 pratense Polyanka, 4.5 + grass plants * Тrifolium repens L. Sprynt, 5 + grass plants 8.32 145.8 4.16 8.11 142.7 4.06 * grass plants * 6.18 109.5 3.09 6.85 120.5 3.43 grass plants * + N120 10.97 192.2 5.49 11.59 196.6 5.80 Natural grass stand 3.41 56.5 1.71 3.57 62.1 1.79 НІР0/05 0.72 0.82 * plants: Bromopsis inermis Arsen, 10+ Festuca pratensis Siveryanka, 8 + Phleum pratense Argenta, 6 kg ha-1 According to our data, the heating energy content in 1 kg of dry biomass of different sown grassland swards hardly depended on the species and varietal composition of mixtures, and was equal to 17.51-18.16 MJ, which is equivalent to the energy value of by-products (straw) of perennial grasses and winter wheat. Conclusions The economically feasible energy potential of meadows is 4.22 million Mg of reference fuel or 12% of total energy consumption. The average energy production of unimproved Ukrainian meadows with natural grassland swards is 2.2 Mg ha-1 and that of flood meadows of Polissia is 3.4 Mg.ha-1 of dry biomass. Addition of legumes to grass mixtures, or of N120 to the sown grass swards, increases the energy efficiency of meadows by 1.7-3.5 times. References Heletukha G., Zhelezna T. and Oliinyk E. (2013) Prospects of heating energy production from biomass in Ukraine. Industrial Heat Engineering 4 (t. 35), 5-15. Energy strategy of Ukraine for the period up to 2030 (2006) Decree of the Cabinet of Ministers of Ukraine No 145 dated 15 March 2006, 10 p. Petrychenko V. and Kurgak V. (2013) Cultural hayfields and pastures of Ukraine. Agrarian science 432 pp. Kurgak V., Levkovskyi A., Yefremova G. and Leshchenko O. (2013) Bioenergy potential of perennial plant communities. Collection of scientific works of the Institite of bioenergy crops and sugar beet NAAS, edition 19, pp. 63-68. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 455 Permanent grassland for anaerobic digestion: a novel insight into management–methane yield relations Herrmann C.1, Heiermann M.1, Schmidt F.2 and Prochnow A.1 1 Leibniz Institute for Agricultural Engineering Potsdam-Bornim, Max-Eyth-Allee 100, 14469 Potsdam, Germany 2 Thuringian State Institute of Agriculture, Naumburger Str. 98, 07743 Jena, Germany Corresponding author: cherrmann@atb-potsdam.de Abstract Grassland biomass as feedstock for biogas production is increasingly considered as an alternative use for grassland. Therefore, the joint research project ‘GNUT-Biogas’ aims at the identification of optimal management intensities for seven productive grassland types in Germany. Emphasis was placed on the effects of management intensity on the feedstockspecific methane yield. Results show that methane yields of grassland biomass can vary widely and are considerably influenced by management options. Higher cutting frequencies and early cuts result in low lignin content and consequential high feedstock-specific methane yields. The management variant corresponding to the production of high quality fodder for dairy cattle resulted in highest biomass qualities for biogas production. Further aspects have to be considered for a final conclusion on optimal grassland management for anaerobic digestion. Keywords: grass silage, biogas production, management intensity, digestibility, acid detergent lignin Introduction Grassland biomass has increasingly been considered as feedstock for biogas production in recent years. Grass silage has been shown to be a capable feedstock for anaerobic digestion (Prochnow et al., 2009; McEniry and O’Kiely, 2013). However, grassland biomass can differ considerably in quality and suitability for biogas production. The most relevant parameter that characterizes quality of biomass for biogas production is the feedstock-specific methane yield. It is influenced by several factors such as vegetation type and composition, intensity of grassland management, mechanical pretreatment during harvesting, and ensiling and storage (Prochnow et al., 2009). The aim of the joint research project ‘GNUT-Biogas’ is to identify the optimal management intensity for seven productive permanent grassland types in Germany considering biomass yield and quality. The present paper focuses on the effects of management variants on biomass quality. Materials and methods Biomass of seven productive grassland types was obtained from different typical sites in Germany. On each site, four management variants were compared in a randomized block experiment with four replications. Management variants differed in cutting frequency, time of cutting and level of fertilization and correspond to the production of: (1) high quality fodder for dairy cattle (180 – 300 kg N ha-1; 4 - 5 cuts per year with the first 3 cuts harvested before mid-July); (2) high quality fodder for dairy cattle with reduced level of nitrogen fertilization (120 – 220 kg N ha-1; 3 - 4 cuts per year with the first 3 cuts harvested before mid-July); (3) biomass utilizing local resources (110 – 200 kg N ha-1; 3 - 4 cuts per year with the first 2 cuts harvested during transition from vegetative to generative stage of growth); and (4) biomass focusing on maintenance of characteristic species composition (120 – 220 kg N ha1 ; 3 - 4 cuts per year with the initial cut early at the onset of grazing and the second cut after flowering of dominant species). Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 456 A detailed description of grassland types, site conditions and management variants is given by Schmidt et al. (2012). Harvested biomasses were wilted, chopped and ensiled in 1-litre glass silos. After storage for a period of 90 days, silages were analysed for chemical composition and feedstock-specific methane yield as described in detail by Herrmann et al. (2011). Methane yield was analysed in triplicates per cut in batch anaerobic digestion tests under mesophilic conditions. Batch tests were conducted for at least 30 days, but were not interrupted until daily gas production was below 0.5 % of the total gas production for 3 consecutive days. Results and discussion Feedstock-specific methane yields of grassland biomasses varied widely from 255 to 402 lN kgODM-1. Table 1 shows feedstock-specific methane yields of three selected grassland types in two years of harvest. Table 1. Effect of management intensity on methane yield (lN kgODM-1) of 3 grassland types in 2 years of harvest. Data show mean (standard deviation) of 3 replicates. Management variant Cut 1 1st 2nd 3rd 4th 5th 1st 2nd 3rd 4th 1st 2nd 3rd 4th 1st 2nd 3rd 4th 2 3 4 Lolio-Cynosuretum, typical character Year 2011 390.7 (2.1) 378.2 (1.9) 357.7 (5.0) 332.9 (1.1) 361.4 (5.9) 360.7 (3.6) 365.3 (4.6) 343.8 (1.3) 310.4 (2.3) 349.6 (1.1) 348.5 (1.7) 338.8 (2.3) 342.1 (3.9) 402.0 (2.0) 333.3 (2.5) 316.0 (2.5) 344.1 (4.6) Year 2012 367.2 (1.8) 364.2 (1.2) 317.4 (3.1) 351.1 (3.0) 294.0 (2.9) 356.7 (4.5) 334.2 (6.6) 329.2 (1.1) 303.0 (3.6) 347.0 (5.4) 309.9 (1.8) 305.7 (1.2) 309.5 (3.3) 366.9 (2.8) 297.2 (0.8) 307.0 (0.9) 342.1 (3.2) Trisetetum without sylvaticum Year 2011 341.4 (2.5) 323.6 (7.2) 329.3 (1.9) 304.6 (3.7) flavescentis, Geranium Alopecuretum pratensis, periodically flooded Year 2012 349.8 (2.6) 295.0 (3.3) 306.5 (5.8) 318.1 (0.7) Year 2011 367.2 (2.8) 314.7 (1.3) 282.1 (4.1) 279.1 (2.6) Year 2012 308.6 (3.6) 297.9 (3.5) 272.5 (5.3) 284.5 (0.7) 352.9 (4.1) 297.9 (7.4) 323.3 (2.0) 317.2 (3.8) 341.3 (4.3) 294.3 (7.6) 316.0 (3.8) 348.9 (2.8) 301.0 (0.4) 293.1 (0.3) 304.9 (1.3) 336.8 (0.9) 303.9 (0.3) 300.2 (2.7) 358.1 (3.4) 308.3 (2.8) 282.3 (3.0) 278.6 (3.6) 303.3 (3.9) 269.8 (0.9) 275.1 (2.4) 302.2 (2.4) 289.0 (1.3) 284.6 (0.4) 287.5 (2.9) 312.4 (0.8) 278.2 (1.2) 255.1 (2.9) 356.0 (7.7) 293.1 (5.5) 327.1 (0.8) 358.2 (3.7) 274.6 (1.2) 312.2 (3.2) 279.3 (6.3) 286.1 (5.5) 293.0 (7.1) 284.6 (1.2) 293.2 (2.4) 271.1 (4.0) Highest feedstock-specific methane yields were achieved by the Lolio-Cynosuretum vegetation whereas Alopecuretum pratensis revealed lowest overall biomass quality (Table 1). Feedstockspecific methane yield was considerably influenced by grassland management. Early cut grassland biomass, which was the first cut of management variant (1) and variant (4), with exception of the Alopecuretum pratensis vegetation, showed highest feedstock-specific methane production. With ongoing maturity or days of growth until harvest feedstock-specific methane yield tended to decrease. This is in accordance to findings of previous studies that observed negative impacts of advancing harvest dates on methane production of common grass species (McEniry and O’Kiely, 2013). Lower feedstock-specific methane yields with advancing stages of maturity are attributed to an increase in less digestible cell wall fractions (Prochnow et al., 2009). A relation between acid detergent lignin (ADL) content, which varied between 17 and 76 g kgDM-1, and feedstock-specific methane yield was found (R2 = 0.551; data not shown). Comparison of variants indicate that grassland management for the production of high quality fodder for dairy cattle also results in highest quality for biogas production, i.e. highest mean feedstock-specific methane yields of grassland biomasses. N fertilization only marginally affected methane production. A mean increase in feedstock-specific methane yield of 1 % was observed for management variant (1) with higher N fertilization compared with variant (2) at Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 457 similar harvest conditions. In contrast, weather conditions within the growth period may considerably affect biomass quality. Feedstock-specific methane yields were higher in year 2011 (Table 1), which was characterized by high rainfall especially during June and July, whereas the year 2012 revealed lower precipitation compared with the long term average. Conclusion Biomasses from permanent grassland reveal large differences in feedstock-specific methane yield depending on grassland type, management intensity and time of cutting. Methane production is mainly affected by the lignin content of the biomasses. High cutting frequencies and early cuts lead to grassland biomass with low lignin content and high feedstock-specific methane yields. Grassland management corresponding to the production of high quality fodder for dairy cattle resulted in highest biomass qualities; however, further aspects such as biomass yield, maintenance of vegetation type, energetic and economic efficiency, and ecological performance have to be considered for final conclusions on optimal grassland management. Acknowledgements The underlying work of this contribution is carried out with financial support of the Federal Ministry of Food, Agriculture and Consumer Protection of Germany via the Agency of Renewable Resources (FKZ 22007509). References Herrmann C., Heiermann M. and Idler C. (2011) Effects of ensiling, silage additives and storage period on methane formation of biogas crops. Bioresource Technology 102, 5153-5161. McEniry J. and O’Kiely P. (2013) Anaerobic methane production from five common grassland species at sequential stages of maturity. Bioresource Technology 127 143-150. Prochnow A., Heiermann M., Plöchl M., Linke B., Idler C., Amon T. and Hobbs P. J. (2009) Bioenergy from permanent grassland - a review: 1. Biogas. Bioresource Technology 100 4931-4944. Schmidt F., Gödeke K. and Hochberg H. (2012) Optimisation of sustainable biomass supply from representative permanent grassland sits of Germany for biogas production. In: Proceedings of the 20th European Biomass Conference and Exhibition, ETA-Florence Renewable Energies, Florence, Italy, 143-145. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 458 Potential use of native Piptatherum miliaceum (L.) Coss. for forage production and bioenergy Porqueddu C., Sulas L., Re G.A., Sanna F., Franca A. and Melis R.A.M. CNR – ISPAAM, Institute for animal production systems in Mediterranean environment, Sassari, Italy Corresponding author: claudio.porqueddu@ispaam.cnr.it Abstract The use of dual-purpose plants for forage production and bioenergy in Mediterranean agricultural system may enhance their flexibility and increase the competitiveness of agroenterprises. Piptatherum miliaceum (L.) Coss. is a native species of Sardinian pastures. It has good palatability for ruminants in the early stage of growth, while the senescent stems and leaves may be used for bioenergy production. A preliminary characterization of Sardinian germplasm was performed in a field experiment. Eleven natural populations were compared, using tall fescue and cocksfoot varieties as references, for aboveground biomass characteristics and phenology. The dry matter yield of P. miliaceum ranged from 1254 to 1900 g plant-1, while the moisture content of its biomass ranged from 38 to 46%. The aboveground biomass was allocated mainly in tillers. The quality of biomass was slightly lower than that of the conventional bioenergy species Panicum virgatum L. Piptatherum miliaceum showed a growing season one month longer than tall fescue and cocksfoot. The long growing season, the high biomass production, the ability to survive to summer drought and the quality of biomass of P. miliaceum are promising traits for the use of this species for both forage and bioenergy production. Keywords: native perennial grasses, smilo grass, forage production, bioenergy Introduction The availability of plant species that increase the flexibility of agricultural systems and offer to farmers the opportunity to differentiate their income is one of the strategies that may increase the competitiveness of agro-enterprises. Piptatherum miliaceum (L.) Coss. (smilo grass) is a herbaceous native perennial grass in dry Mediterranean environments. It is common in natural pastures and it can be found in marginal environments such as semiarid mining zones (Conesa et al., 2006). It has good palatability for ruminants in an early stage of growth, but becomes unpalatable in the old stands (Baldoni et al., 1974). Its high biomass production offers the chance to use the stems and leaves for bioenergy production. To our knowledge, no bioagronomic characterization for bioenergy production of smilo grass has been performed to date. The main objective of this study was to perform a preliminary characterization of some Sardinian smilo grass accessions for phenology, biomass production and quality. Materials and methods The experiment was carried out in Sassari (40° 45' 15" N 8° 25' 13" E), altitude 24 m a.s.l., on an alluvial soil with pH 7.8. Average total annual rainfall for the site is 550 mm, mainly concentrated in autumn and spring. Seeds of smilo grass collected from wild populations growing in Sardinia (Italy) during summer 2011 were grown in pots from the end of summer up to April 2012, when seedlings were transplanted into the field. Plots (2 m × 1 m) of 11 accessions of smilo grass were established in a completely randomized design with three replicates. Plant distance was 0.5 × 0.5 m and 8 plants per accession were transplanted in each plot. Festuca arundinacea Screb. (tall fescue) cv. Fletcha and Dactylis glomerata L. (cocksfoot) cv. Jana were used as control species. No irrigation or herbicides were applied after transplanting. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 459 Phenology of plants was observed weekly to define the growing stages in the experimental field conditions. All plots were harvested in 2012 and 2013, once a year in July, when smilo grass leaves dried out. The measurements of biomass and its components were made on two plants per plot. With the aim of estimating biomass partitioning, the dry weights of leaves, tillers and spikes were measured after drying the separated samples in an oven at 60 °C to constant weight and water content. The analysis of biomass quality was done determining hemicellulose, cellulose, lignin and ash contents according to Van Soest et al. (1991). Data were analysed by ANOVA using Statgraphics Centurion software. Data were transformed when needed and homoscedasticity of data was tested by Bartlett’s test. Sample means were compared by Tukey’s HSD test. Results and discussion In the first year, plant yields were negligible, both in smilo grass and in conventional grass species. In the second year, tall fescue and smilo grass showed higher plant dry weights than cocksfoot (Table 1). Table 1. Dry matter yield (g plant -1), moisture content at harvest (%), biomass partitioning (% DW) and full flowering date of 11 smilo grass accessions (PM), 1 cocksfoot (DA) and tall fescue (FE) varieties measured in 2nd year. Biomass was cut on July 2013. Different letters indicate statistical differences at P<0.05 (Tukey test). PM13SEM Dry matter yield (g plant-1) 1360.3 ab Species Moisture Tiller mass content at ratio harvest (%) (% DW) 44.5 ab 58.2 a Leaf mass ratio (% DW) 22.6 bc Spike mass Full ratio flowering (% DW) date in 2013 19.3 a 14 Jun PM14PZS 1592.1 ab 38.4 bc 59.1 a 19.5 c 21.4 a 14 Jun PM15VNM 1676.4 ab 39.4 bc 62.6 a 20.5 c 16.9 a 28 Jun PM16SLR 1855.9 a 44.1 abc 61.2 a 20.3 c 18.5 a 7 Jun PM18FRT 1732.9 ab 39.8 abc 58.6 a 24.3 bc 17.1 a 7 Jun PM19CMC 1435.1 ab 44.7 ab 59.5 a 17.0 c 23.5 a 7 Jun PM20PMT 1323.0 ab 40.2 abc 56.2 a 16.6 c 27.2 a 7 Jun PM21DCM 1916.1 a 45.1 ab 56.3 a 19.3 c 24.4 a 7 Jun PM22STG 1254.7 ab 45.9 a 55.4 a 19.0 c 25.6 a 7 Jun PM23ZCN 1691.7 ab 45.1 ab 61.0 a 20.7 bc 18.3 a 7 Jun PM24MNR 1648.7 ab 43.1 abc 57.5 a 20.0 c 22.5 a 7 Jun DA29JA 263.3 c 36.8 c 44.2 b 37.7 a 18.1 a 3 may FE28FL 825.9 bc 41.3 abc 64.2 a 31.2 ab 4.6 b 3 may However, smilo grass yields ranged from 1254 to 1900 g plant-1 while tall fescue yield was about 826 g plant-1. The aboveground biomass was allocated mainly in tillers, both in conventional species and smilo grass, with F. arundinacea showing the highest tiller mass ratio (64%), followed by smilo grass (59%) and finally, by cocksfoot (44%); that was the only species with statistically significant different values in tiller mass ratio. The biomass partitioning resulted favourable for bioenergy uses, where a higher ratio of stems is required. In smilo grass, the relatively high moisture content at harvest was due to the tillers, as they also remained green during summer in drought conditions. After the first cut in July 2012, smilo grass showed a fast regrowth and in September some plants flowered again but did not produce seeds. Vegetative growth extended from the end of August 2012 to the end of April 2013 for the early-flowering accessions and up to the third decade of May for the late-flowering accession PM15VNM. Heading was a long phase, lasting for one month before the full-flowering stage. Full flowering occurred in the first decade of June in most accessions, whereas two of them flowered in the second decade and only one in the last decade of June-beginning of July. Smilo grass accessions Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 460 showed a growing season extended a month more than cocksfoot and tall fescue. These latter, in fact, flowered in the first decade of May. Regarding the biomass quality of smilo grass components (Table 2), all calculated parameters were consistent with those reported for the conventional bioenergy species Panicum virgatum L. (switchgrass) by McKendry (2002). The ash content of tillers, spikes and leaves was higher than reported for switchgrass. Table 2. Hemicellulose, cellulose, lignin and ash contents in leaves, stems and spikes of a smilo grass accession. Hemicellulose (%) 23.59 26.58 36.32 Leaves Stems Spikes Cellulose (%) 40.67 39.15 28.01 Lignin (%) Ash (%) 3.41 9.44 9.93 16.23 5.34 5.24 Conclusions The long growing season, the high biomass production of smilo grass, its ability to survive in summer drought conditions and the quality of biomass for bioenergy production are promising traits for the use of this species for both forage and bioenergy production. A follow-up of this research is needed to exploit the variability of agronomical and phenological traits found among and within accessions and to establish the better crop management to combine forage and bioenergy production. Acknowledgments This study is part of the OPTIMA Project supported by the FP7 Grant no. 289642 of the European Commission. The Authors wish to thank Ms. M. Sassu, Mr. D. Dettori, Mr. D. Nieddu, Mr. P. Saba and Dr. M. Cuccureddu for their technical help. References Baldoni R., Kokeny B. and Lovato A. (1974) Oryzopsis miliacea (L.) Benth. et Hook. ex Aschers et Schweinf. In Le piante foraggere. Reda, 86. Conesa H.M., Faz Á. and Arnaldos R. (2006) Heavy metal accumulation and tolerance in plants from mine tailings of the semiarid Cartagena-La Unión mining district (SE Spain). Science of the Total Environment 366,1–11 McKendry P. (2002). Energy production from biomass (part 1): overview of biomass. Bioresource Technology 83, 37–46. Van Soest P.J., Robertson J.B., and Lewis B.A. (1991). Methods for Dietary Fiber, Neutral Detergent Fiber, and Nonstarch Polysaccharides in Relation to Animal Nutrition. Journal of Dairy Science 74, 3583-3597. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 461 Second generation bioethanol production from Phalaris aquatica L. energy crop Pappas I.A.1,2 , Kipparisides C.2 and Koukoura Z.1 1 Range Ecology Laboratory, Department of Forestry and Natural Environment, Aristotle University of Thessaloniki, 54124, Thessaloniki, Greece. 2 Chemical Process and Energy Resources Institute, Centre for Research and Technology Hellas, 6th km Harilaou Thermi Road, 57001, Thessaloniki, Greece. Corresponding author: pappas@cperi.certh.gr Abstract Nowadays the inevitable deficit of fossil fuels, the increasing energy consumption and the security of global energy market are the driving forces to utilize renewable energy resources, such as biomass, for bioenergy, biofuels and bio products. One type is lignocellulosic biomass that can be found in abundance in perennial forage species. The purpose of this study was to investigate aboveground biomass production and lignocellulosic concentration stability of rainfed crop (5-years old) from native Mediterranean perennial grass Phalaris aquatica L. at a lowland region of Central Greece as potential dedicated energy crop, and to investigate the integrated biochemical conversion processes efficiency of lignocellulosic biomass into second generation bioethanol. The results indicated that Phalaris aquatica L. energy crop could serve as a stable second generation bioethanol feedstock at lowland areas of Central Greece after the establishment year due to high biomass yield and cell wall concentration. The results also revealed that under the integrated biochemical conversion pathway (dilute acid pre-treatment, enzymatic hydrolysis, fermentation) glucose monomer derived from cell wall carbohydrates could efficiently (> 80%) be converted to bioethanol. Keywords: perennial grasses, cellulosic biomass, pretreatment, bioconversion, bioethanol. Introduction The production of bioethanol has attracted worldwide attention as a strategy for combating global warming and improving European energy security. This has placed attention on the utilization of fermentable sugars derived from lignocellulose, the largest known renewable carbohydrate source for biofuel production (Jorgensen et al., 2007). Herbaceous species such as perennial grasses are promising lignocellulosic feedstocks due to their high yield potential, low establishment cost and higher carbohydrate content than annual species. However, plant cell wall recalcitrance necessitates the efficient treatment in order to release the fermentable sugar monomers. Dilute acid pretreatment followed by enzymatic hydrolysis is the typical process used to convert lignocellulosic biomass to fermentable sugars (Iranmahboob et al., 2002). The objectives of the present study were: a) to evaluate aboveground biomass yield and cell wall concentration stability of Phalaris aquatica L. energy crop for five successive years (2008 – 2012) in Central Greece, and b) to investigate the bioconversion efficiency into bioethanol through the integrated biochemical processes. Materials and methods Biomass samples of perennial grass (Phalaris aquatica L. - Harding Grass, HG) rainfed plantation (5-years old) in Central Greece, located at Pteleos site at 15 m altitude, were collected at the end of the growing season for five successive years (2008-2012). Biomass production was measured using 20 (1 × 1 m2) plots. Harvested biomass was dried at 60 oC for 48h, milled to a size of 1 mm and subjected to fibre analyses using the Van Soest method (Van Soest et al., 1991). Hemicellulose and cellulose concentrations were determined by the differences of (NDF-ADF) and (ADF-ADL) respectively. Air dried biomass samples at 10% (w/v) solid Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 462 loading were mixed at mild pretreatment conditions (1.5% (w/v) H2SO4, 120 oC, 45 min) in 50ml screw cap bottles in oil bath. Solid residues recovered by filtration were hydrolyzed with cellulases mixture (Celluclast 1.5L and Novozyme 188). Enzymatic hydrolysis was performed in 250ml flasks with 100 ml of 0.05 M sodium citrate buffer (pH = 4.8) at 50 oC in an incubator at 150 rpm agitation. Kinetic analysis was performed at different biomass loading (2, 4, 6 % w/v) with enzyme loading (30 FPU/g biomass) and ratio 1:2 for 72 h. Afterwards, the glucose found in the hydrolysate was fermented by the yeast Saccharomyces cerevisiae Sigma (Type II) in 250ml flasks, incubated at 30 oC, pH = 6.5 at 150 rpm for 34h. Glucose and ethanol concentration were analysed by HPLC. All experiments were performed in duplicate. One-way ANOVA was used to compare biomass production and cell wall concentration of HG at different years. Further differences were evaluated with the LSD post hoc test, at a level of significance of 0.05 (Kinnear and Gray, 2008). Results and discussion Aboveground biomass production (t DM ha-1) of Phalaris aquatica L. crop varied significantly between the years 2008 - 2012 (Table 1). Dry matter yield increased from the first season after establishment to the second growing season. In the second year, yield increased by 661% (13.8 t ha-1), in the third year it was further increased by 45% , and in the next two years the increase was more stable, at 5.5% and 6.2 % respectively. The average dry matter yield for the 5-year period was 15.8 t ha-1. The same trend in biomass yield during several growing seasons were reported for perennial grass species such as reed canary grass (Sahramma et al., 2003) and Mischanthus x giganteus (Clifton-Brown et al., 2007). Growth year had a significant effect on cell wall compositon of HG aboveground biomass (Table 1). Table 1. Mean biomass yield and cell wall concentration of Phalaris aquatica L. crop at Pteleos site for five successive years (2008-2012). Year Biomass yield (Mg ha-1 DM) NDF (g kg-1 DM) ADF (g kg-1 DM) ADL (g kg-1 DM) Cellulose (g kg-1 DM) Hemicellulose (g kg-1 DM) 261b 327c 2008 1.8c 666c 339c 78a 292a 431a 2009 13.8b 795a 503a 72a 278a 410a 2010 19.9a 752b 474a 64b 258b 431a 2011 21.0a 765b 507a 76a 277 409a 2012 22.3a 760b 482a 73a 279 396b Mean 15.8 748 468 72.5 6.2 19.3 s.e.d 3.8 21.7 31.1 2.4 a,b: Means followed by the same letter do not differ statistically significant (LSD Test, P<0.05) Total structural carbohydrate s (g kg-1 DM) 588c 723a 688b 689b 687b 675 22.8 NDF, ADF and cellulose were higher in the subsequent growth years after the establishment year, with the loss of hemicellulose during this period owing to the increased deposition of cellulose. Significant annual variation of cell wall concentration from other perennial grasses grown for biomass, have been also reported (Schmer et al., 2010; Allison et al., 2012). The chemical composition of feedstock has a major influence on the efficiency of bioethanol generation. Dilute sulphuric acid pretreatment at mild conditions had significant impact on Phalaris aquatica cell wall, as increased cellulose by 45.5% and decreased hemicellulose by 79.4% and lignin by 13.9%, respectively (Pappas, 2010). According to Zheng et al., (2009) promotion of cellulose content and demotion of hemicellulose and lignin content can facilitate the enzymatic hydrolysis in perennial grasses. The increase of biomass loading resulted in linear increase of glucose in the enzymatic hydrolysate (Figure 1), with the maximum concentration of 18.8 g/l found at 6 % biomass loading at 48h. After hydrolysate condensation, glucose concentration reached 22.6 g/l and was used as a carbon source for ethanol production by yeast Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 463 fermentation. The fermentation resulted in 84.3% yield (ethanol concentration equal to 5.49 g/l) based on the theoretical maximum yield and 0.50 g/l/h ethanol productivity. 22 20 18 Glucose (g/l) 16 14 12 10 8 6 4 Biomass 2% Biomass 4% Biomass 6% 2 0 0 10 20 30 40 50 60 70 80 Hydrolysis time (h) Figure 1. Kinetics of glucose concentration (g/l) at different biomass loading Conclusion The results indicated that Phalaris aquatica L. as an energy crop could serve as a secondgeneration bioethanol feedstock in lowland areas of Central Greece after the establishment year, due to high and stable biomass yield and cell wall concentration. Furthermore, moderate acid pretreatment followed by enzymatic hydrolysis is an effective biochemical method to release sugar monomers from Phalaris aquatica L. cell wall carbohydrates which can be efficiently fermented to bioethanol. References Allison G.G., Morris C., Lister S.J., Barraclough T., Yates N., Shield I. and Donnison I.S. (2012) Effect of nitrogen fertiliser application on cell wall composition in switchgrass and reed canary grass. Biomass and Βioenergy 40, 19-26. Clifton-Brown J.C., Brewer J. and Jones M.B. (2007) Carbon mitigation by the energy crop Miscanthus. Global Change Biology 13 (11), 2296–2307. Iranmahboob J., Nadim F. and Monemi S. (2002) Optimizing acid hydrolysis: a critical step for production of ethanol from mixed wood chips. Biomass and Bioenergy 22, 401-404. Kinnear P.R. and Gray C.D. (2008) SPSS 15 Made Simple. Psychology Press. Hove. Jorgensen H., Kristensen J.B. and Felby C. (2007) Enzymatic conversion of lignocellulose into fermentable sugars: challenges and opportunities. Biofuels, Bioproducts and Biorefining 1 (2), 119 -134. Pappas I.A. (2010) Assessment of range plants growth potential and their development for bioenergy production. Ph.D. Thesis, Faculty of Forestry and Natural Environment, Aristotle University of Thessaloniki, Greece,125 pp. Schmer M.R., Mitchell R.B., Vogel K.P., Schacht W.H. and Marx D.B. (2010) Spatial and Temporal Effects on Switchgrass Stands and Yield in the Great Plain. Bioenergy Research 3, 159-171. Sahramaa M., Hihamaki H. and Jauhiainen L. (2003) Variation in biomass related variables of reed canary grass. Agricultural and Food Science in Finland 12, 213–225. Zheng Y., Pan Z., Zhang R. and Wang D. (2009) Enzymatic saccharification of dilute acid pretreated saline crops for fermentable sugar production. Applied Energy 86 (11), 2459-2465. Van Soest P.J., Robertson J.B. and Lewis B.A. (1991) Methods for dietary fiber, neutral detergent fiber and nonstarch polysaccharides in relation to animal nutrition. Journal of Dairy Science 74, 3583 - 3597. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 464 Hydrothermal processing of rush (Juncus spp.) and bracken (Pteridium aquilinum) dominant biomass from semi-natural landscape management Corton J.1, Ross A.B. 2, Lea-Langton A.R. 2, Donnison I.S. 1 and Fraser M.D. 1 1 IBERS, Aberystwyth University, Gogerddan, Aberystwyth, Ceredigion, SY23 3EE, United Kingdom, 2 Energy and Resources Research Institute, Faculty of Engineering, University of Leeds, Leeds, LS2 9JT, United Kingdom Corresponding author: jcc@aber.ac.uk Abstract Semi-natural habitats that are dominated by rush (Juncus spp.) and bracken (Pteridium aquilinum) show an increase in floristic diversity following management through cutting and biomass removal. In this study the waste biomass created is utilized for energy and product production. Biomass samples from rush- and bracken-dominated vegetation were converted by hydrothermal carbonization and hydrothermal liquefaction to derive oil, hydrochar, gaseous and aqueous fractions. Prior to this conversion the biomass had been ensiled, washed with water, and mechanically dehydrating using a screw press. This process produced a press cake with a reduced mineral composition compared to the silage. The process method had a significant impact on the oil, hydrochar and gas yields; liquefaction provided a higher oil composition, and carbonization a higher hydrochar composition. Washing and pressing did not have a significant impact upon the product yields following conversion via carbonization or liquefaction, but did impact on mineral composition. The community type (rush or bracken) had a significant impact upon the oil yield with the rush-dominant biomass providing a higher oil yield. Keywords: hydrothermal, Juncus, bracken, hydrochar, oil Introduction The indigenous vegetation communities found in the UK uplands support ecosystems with international conservation significance. Cutting management may be implemented in order to preserve biodiversity on these areas with the resultant accumulation of waste biomass. This biomass could potentially be available for further processing for energy or product development (Corton et al., 2013). Two common habitats that benefit from a cutting and biomass-removal management system are rush- (Juncus spp) and bracken- (Pteridium aquilinum) dominated semi-natural landscapes. Bracken- and rush-dominant biomass has been investigated as a resource for energy and product conversion as part of the EU PROGRASS project (Wachendorf et al., 2009). The associated process involved ensiling the biomass, washing the feedstock with water and mechanically dehydrating it to produce a press cake. This procedure reduced the mineral composition of the biomass in order to improve its combustion fuel qualities. Hydrothermal conversion is a method of processing biomass in order to obtain products for energy generation, carbon sequestration or chemical production. Superheated water is created by heating water under pressure within a vessel; during hydrothermal conversion the vessel also contains biomass. Under these conditions the biomass is converted into a multiple product mix that consists of aqueous, oil, gaseous and hydrochar fractions. By conducting hydrothermal conversion at different temperatures, oils (favoured in hydrothermal liquefaction) or hydrochars (favoured in hydrothermal carbonization) can be increased at the expense of other fractions in the product mix. In this study, biomass from rush- and bracken-dominated habitats, in the form of silage and press cake, were subjected to hydrothermal carbonization and liquefaction. This paper summarizes the impact of hydrothermal conversion route (carbonization or liquefaction), Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 465 pre-treatment methods (silage or press cake) and community type (rush or bracken-dominant) upon the product yields. Materials and methods The biomass tested had been cut from replicated plots of each vegetation type located at the Pwllpeiran Upland Research Centre, Wales, and had been ensiled without pre-treatment in experimental silos. The experimental work was conducted at the laboratories of the Energy Research Institute at the University of Leeds. Conversion via hydrothermal carbonization and hydrothermal liquefaction was performed in a 75 ml capacity, unstirred, bomb-type batch reactor (Parr, USA). The temperature and pressure inside the reactor was monitored using a Parr 4836 controller. The reactor was hydrocharged with ~3 g of sample and 27 ml of distilled water; this was pre-mixed and added to the reactor as slurry prior to sealing and heating. To achieve hydrothermal carbonization the residence time was set at one hour and the reactor temperature was set at 250 °C with a pressure of 44 bar. To achieve hydrothermal liquefaction the reactor temperature was set at 350 °C with a pressure of 170 bar and the residence time was one hour. Separation using dichloromethane, water and filtering were employed. The residence time was measured from the point at which the reactor reaches 250 °C or 350 °C respectively. The heating rate of the reactor was approximately 10 °C per minute and the reactor was purged with nitrogen prior to processing. Separation procedures are detailed in Figure 1. GenStat® Release 13 was employed for discerning the levels of significance using an analysis of variance. Results and discussion The oil yield was lower when biomass derived from the bracken plots was employed compared to that from the rush plots, but the hydrochar yields were the same (Table 1). Table 1. The impact of vegetation community, hydrothermal conversion route and biomass processing upon the percentage composition (DM ash free) of the products produced in hydrothermal conversion of biomass. The products include oil, hydrochar, gas and an aqueous fraction; these are expressed as a percentage of the resultant product stream. Interactions have been assessed but are not included for this report. Product fraction Treatment Vegetation community Biomass processing Hydrothermal conversion process rus bra sed P value sil pc sed P value htc htl sed P value Oil 12 7 2.01 0.046 10 9 2.01 0.649 2 17 2.01 <.001 Hydrochar 43 44 9.21 1 42 45 4.53 0.445 59 28 4.53 <.001 Gas 14 9 2.19 0.121 11 12 2.19 0.431 8 15 2.19 0.018 Aqueous 31 40 5.15 0.106 38 34 5.15 0.407 31 40 5.15 0.112 Abbreviations: rus = rush; bra = bracken; sil = silage; pc = press cake; htc = hydrothermal carbonisation; htl = hydrothermal liquefaction; sed = standard error of difference. It is possible that mass transfer into the aqueous fraction instead of the oil fraction occurred during hydrothermal conversion of bracken-derived biomass. There was no significant impact of biomass processing upon the product yields. However, in the current work there was no attempt to measure or calculate reaction rate and the gaseous phase composition was not analysed; these are research opportunities. Considering the amount of time and energy expended in a washing and pressing pre-treatment regime, hydrothermal yields are not sufficiently impacted upon by that process to render it worthwhile, although this conclusion is Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 466 based on fraction yields and not gaseous compositions. If a low mineral composition product stream is desirable (e.g. for combustion), then washed and pressed feedstocks may be preferable. The hydrothermal conversion process had a highly significant impact upon the oil, hydrochar and gas yields, as would be expected. The hydrothermal conversion process did not impact upon the aqueous yield. There is consensus across the literature that hydrothermal liquefaction will produce more oil and gas compared to hydrothermal carbonization, and this is reflected in the current study. The stability of the aqueous yield in relation to the hydrothermal process renders the fraction a candidate for investigation as a potentially useful by-product in a bio-refinery system. The productivity of the aqueous fraction appeared to be a parameter on which downstream processes could depend. The utilization of the aqueous fraction for microbial culture could add a cultivation process into the bio-refinery model, as has been investigated by Nelson (2013). Char production through hydrothermal carbonization of the press cake (demineralized) would be best employed for combustion fuel production. If oil production is a priority then rushderived biomass would provide the highest yield. Hydrothermal liquefaction would be the process of choice for multiple-product streams, and hydrochar for sequestration would be best produced through carbonization of the bracken-dominated biomass. Acknowledgements This work was jointly funded by the Engineering and Physical Sciences Research Council (EPSRC) Supergen Consortium (EP/E039995) and the European Union Life + programme via the PROGRASS research consortium (LIFE07 ENV/D/0000222). References Corton J., Buhle L., Wachendorf M., Donnison I.S. and Fraser M.D. (2013) Bioenergy as a biodiversity management tool and the potential of a mixed species feedstock for bioenergy production in Wales. Bioresource Technology 129, 142-149. Nelson M., Zhu L., Thiel A., Wu Y., Guan M., Minty J., Wang H.Y. and Lin X.N. (2013) Microbial utilization of aqueous co-products from hydrothermal liquefaction of microalgae Nannochloropsis oculata. Bioresource Technology 136, 522-528. Wachendorf M., Richter F., Fricke T., Gray R. and Neff R. (2009) Utilization of semi natural grassland through integrated generation of solid fuel and biogas from biomass. I. Effects of hydrothermal conditioning and mechanical dehydration on mass flows of organic and mineral plant compounds, and nutrient balances. Grass and Forage Science 64, 132-143. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 467 The yield and variation of chemical composition of cocksfoot biomass after five years of digestate application Tilvikiene V., Kadziuliene Z., Dabkevicius Z., Šarūnaitė L., Šlepetys J., Pocienė L., Šlepetienė A. and Ceceviciene J. Institute of Agriculture, Lithuanian Research Centre for Agriculture and Forestry, Instituto al. 1, Akademija LT-58344, Kedainiai distr., Lithuania Corresponding author: vita@lzi.lt Abstract Biogas production is the optimal way to utilize organic materials or energy crops and produce bioenergy. In many countries the number of biogas plants is increasing. The growth of biogas production directly influences the generation of digestate. An experiment on cocksfoot (Dactylis glomerata L.) fertilized with mineral fertilizer and digestate was conducted to evaluate the yield and chemical composition of cocksfoot biomass after five years of digestate application. The average results of the experiment suggest that within five years of sward use higher rates of nitrogen present in the digestate increased the biomass yield. The swards fertilized with mineral fertilizer N360 produced the same biomass yield as those fertilized with N180. The chemical and structural biomass components varied due to the influence of type and rate of fertilizers. Keywords: cocksfoot, digestate, chemical composition Introduction In order to increase bioenergy production and ensure that renewables contribute to the total energy consumption, the number of biogas plants is increasing (Angelidaki et al., 2011). The main products of biogas production are methane, carbon dioxide and digestate. Biomethane may be used for generation of electricity, for direct injection into the natural gas grid, or as a fuel for vehicles (Raposo et al., 2012). The use of digestate faces some problems and is a subject of considerable debate. In digestate, organic nutrients are converted and mineralized to more soluble and biologically available forms for plants (Voća et al., 2005). The application of digestate as fertilizer for grasslands could be an effective way to utilize residues from biogas plants, as it can reduce the need of mineral fertilizer and increase biomass productivity (Alburquerque et al., 2012). The aim of this study was to evaluate the yield and chemical composition of cocksfoot biomass after five years of digestate application. Materials and methods Field and laboratory experiments were carried out in Central Lithuania (55° 24' N) on an Apicalcari - Endohypogleyic Cambisol, light loam. The experiment was laid out in a randomized block with eight treatments and four replicates. Cocksfoot cv. Amba (developed in Denmark) was used for the experiment. No fertilization (control), two levels of mineral nitrogen fertilization (N180 and N360) and five levels of nitrogen in digestate (N90, N180, N270, N360 and N450) were chosen for pure swards of cocksfoot. One half of the annual fertilizer rate was applied in early spring at the beginning of vegetation growth, the second application was made after the first cut. For each application, the quantity of digestate was calculated according to the concentration of total nitrogen in it. Total ash content was determined by ISO 3451. The concentrations of K, Na, Ca, Mg, P were estimated in sulphuric acid digestates. Concentrations of K, Na, Ca and Mg were quantified by a flame atomic absorption. P concentrations were quantified by a colouring reaction with ammonium. Total C, N and S concentrations were determined by a dry combustion method (ISO 16634). For the ADF, NDF and ADL the cell Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 468 wall detergent fractionation method (Faithfull, 2002) was used. The average chemical composition of digestate is presented in Table 1. Table 1. The average chemical composition of digestate. Rates mg kg-1 N-NO3 0.020 N-NH4 0.416 Rates g kg-1 in natural material N 5.7 P 1.67 K 1.84 DM 34 C 12 Ca 0.77 Mg 0.12 pHKCl 8.4 The swards were harvested three times per season. The first cut was taken at heading stage of cocksfoot. The differences between treatments were estimated using Fisher’s Multiple Range tests. Statistical inferences were made at the 0.05 significance level. Results and discussion The average biomass yield, Mg ha-1 Increasing intensity of cultivation of agricultural crops results in greater productivity of biomass (Kering et al., 2012). The results of the experiment suggest that cocksfoot fertilized with 180 kg ha-1 of nitrogen present in digestate produced a similar biomass yield to that of swards fertilized with the same rate of mineral fertilizer (Figure 1). These results are consistent with previous research, suggesting that the nutrients in digestate are already mineralized and thus can be used by plants effectively, but digestate does not surpass mineral fertilization in biomass yield formation (Fuchs et al., 2008). The increase of fertilizer rate influenced intensive biomass accumulation of swards fertilized with nitrogen present in the digestate. Different results were obtained for the swards fertilized with mineral fertilizer. There was no significant difference between the swards fertilized with N180 and N360. Fertilized with mineral fertilizer Fertilized with digestate 12 10 8 6 4 2 0 0 180 360 Nitrogen rate, kg ha-1 Figure 1. The average biomass yield of cocksfoot biomass within five years of sward cultivation Because of the chemical composition of cocksfoot biomass, conventional technology is chosen. Cocksfoot biomass could be used as fodder, as a substrate for biogas production or for combustion. In our research, in cocksfoot biomass the highest variation was found for the following chemical and structural components: nitrogen, sulphur, phosphorus, calcium, chlorine, neutral detergent fibre and acid detergent fibre. The variation of these elements was influenced not only by the type but also by the rate of nitrogen fertilizer. The swards applied with mineral fertilizer exhibited significant higher nitrogen and calcium concentrations compared to those fertilized with digestate (Table 2). An increase in fertilization rate decreased the concentration of ash in the biomass. The contents of neutral detergent fibre and acid detergent fibre were higher in the swards fertilized with digestate. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 469 Table 2. The chemical composition of cocksfoot biomass Fertilizer Control Rate N N0 Mineral fertilizer N180 Digestate N90 N360 N180 N270 N360 N450 LSD05 Rates g kg-1 in dry matter Ash N S C 106 18.7 2.2 424 101 30.3 1.4 424 96 38.4 1.1 421 109 19.4 1.9 423 108 20.4 1.9 421 104 19.8 1.7 427 105 22.0 1.9 427 100 23.5 1.9 433 5.7 0.300 0.310 P K Ca 4.4 16.5 9.9 5.5 15.7 8.2 5.5 4.8 4.5 4.4 4.1 3.9 0.900 15.6 17.3 16.8 17.1 17.0 16.3 0.061 7.9 7.9 7.1 5.9 5.3 5.5 0.093 Mg Na Cl 3.9 3.8 7.1 475 4.8 3.6 3.3 476 5.2 3.9 3.6 3.3 3.4 3.4 0.142 4.5 3.8 3.9 3.7 4.3 4.2 0.050 3.3 7.4 7.5 9.4 9.7 10.9 0.097 NDF 465 500 534 543 545 547 0.269 ADF 339 334 327 355 358 371 366 363 3.1 Lignin 74.8 73.1 80.8 76.1 73.6 75.7 72.9 76.9 1.8 Conclusion The use of mineral fertilizer and digestate significantly increased the dry matter yield of cocksfoot. The nitrogen fertilizer N180 present in digestate was as efficient on cocksfoot biomass yield as mineral fertilizer nitrogen. Fertilization with digestate at a rate of N360 significantly increased the biomass yield of cocksfoot, compared with the swards fertilized with mineral nitrogen. The chemical and structural biomass components varied due to the influence of type and rate of fertilizers. Acknowledgments This work was supported by the project 'Scientific validation of C3 and C4 herbaceous plants multi-functionality for innovative technologies: phyto-raw materials - bio-products – environmental effects' No. VP1-3.1-ŠMM-08-01-023. References Alburquerque J.A., Fuente C., Ferrer-Costa A., Carrasco L., Cegarra J., Abad M. and Bernal M.P. (2012) Assessment of the fertiliser potential of digestates from farm and agroindustrial residues. Biomass and Bioenergy 40, 181–189. Angelidaki I., Karakashev D., Batstone D.J., Plugge C.M. and Stams A.J.M. (2011) Biomethanation and its potential. Methods in Enzymology 494, 327-351. Faithfull N.T. (2002) Methods in agricultural chemical analysis: a practical handbook. Wallingford UK: CABI Publishing. Fuchs J. G., Baier U., Berner A., Mayer J. and Schleiss, K. Effects of digestate on the environment and on plant production - results of a research project. http://orgprints.org/17981/1/fuchs-etal-2008-jgf.pdf (English). Kering M. K., Butler T. J., Biermacher J. T. and Guretzky J. A. (2012) Biomass yield and nutrient removal rates of perennial grasses under nitrogen fertilization. Bioenergy Resources 5, 61–70. Raposo F., Dela Rubia M.A., Fernández-Cegrí V. and Borja R. (2012) Anaerobic digestion of solid organic substrates in batchmode: an overview relating to methane yields and experimental procedures. Renewable Sustainable Energy Review 16, 861–877. Voća N., Krička T., Ćosić T., Rupić V., Jukić Ž. and Kalambura S. (2005) Digested residue as a fertilizer after the mesophilic process of anaerobic digestion. Plant, Soil and Environment 51, 262-266. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 470 Evaluating sample preparation method effects on the specific methane yield of pre-and post-ensilage grass in an in vitro batch anaerobic digestion assay Nolan P.1,2, Doyle E.M.2 and O’Kiely P.1 1 Animal & Grassland Research and Innovation Centre, Teagasc, Grange, Dunsany, Co. Meath, Ireland, 2 School of Biology and Environmental Science, UCD, Belfield, Dublin 4, Ireland. Corresponding author: Pearl.Nolan@teagasc.ie Abstract The objective of this experiment was to compare the effects of two biomass preparation methods on the specific methane yields of herbages during an in vitro batch anaerobic digestion assay. Three contrasting pre- and post-ensilage perennial ryegrass samples were either thermally oven dried (40oC for 48 h) or non-dried (frozen to -196oC), prior to milling through a 1 mm pore sieve. These were subsequently evaluated for biomethane potential (BMP) using a small-scale in vitro batch digestion system. The three pre-ensilage herbages (H1, H2, H3) yielded 279, 249, 202 and 209, 211, 191 L CH4 kg-1 VSadded with oven dried and non-dried frozen sample preparation methods, respectively. The three post-ensilage herbages (H1, H2, H3) yielded 167, 214, 182 and 214, 214, 209 L CH4 kg-1 VSadded with oven dried and non-dried sample preparation, respectively. It is concluded that herbage samples should not be oven dried when assessing their methane potential with a small-scale in vitro batch digestion assay. Keywords: grass, silage, sample preparation, anaerobic digestion Introduction Grass biomass can be a cost-effective feedstock for anaerobic digestion due to its potential to produce a high yield of appropriate quality biomass (McEniry et al., 2011) that can be ensiled to provide year-round supply of renewable energy (Plöchl and Heiermann, 2006). The biomethane potential (BMP) test is a small-scale in vitro batch digestion method commonly utilized to estimate the methane output of pre- and post-ensilage feedstocks (Angelidaki et al., 2009). A major issue for small-scale in vitro tests (i.e. in 100 ml to 200 ml vessels) is the need to pre-process biomass. This permits the use of relatively small (<1 g) but representative subsamples of the biomass (Purcell et al., 2011). A common solution to creating a homogenous representative sub-sample is to oven dry the sample prior to milling through a 1 mm pore sieve. However, drying can reduce protein and increase fibre proportions of herbage (Alomar et al., 1999). This can subsequently affect the rate and extent of degradation during anaerobic digestion. Furthermore, post-ensilage feedstocks contain many volatile compounds such as volatile fatty acids (VFAs), lactic acid, ammonia and alcohols (McDonald et al., 1991). These compounds can be partially lost during oven drying (Kruger et al., 2011) thereby leading to an underestimation of the potential to produce methane. In contrast, comminuting non-dried samples (frozen to -196oC) may provide an alternative method to create a small representative sub-sample without the undesirable loss of important substrate. Thus, the aim of this experiment was to compare the relative effects of oven dried milled samples versus non-dried samples manually comminuted following freezing to -196oC, on cumulative methane production of preand post-ensilage herbage samples, using a small-scale in vitro batch digestion assay. Materials and methods Three perennial ryegrass (Lolium perenne L., cv. Greengold) herbages and their corresponding silages were selected from those reported by Navarro-Villa et al. (2012). The herbage preconditioning treatments aimed to provide contrasting silage fermentation conditions by using managements predisposing the silages to undergo an extensive lactic acid dominant Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 471 fermentation with minimal clostridial activity (H1), an extensive fermentation with clostridial activity (H2), and a restricted fermentation with moderate clostridial activity (H3). Both preand post-ensilage samples were either: (a) thermally dried in a forced air circulated oven (40oC for 48 h) and milled, or (b) non-dried, frozen to -196oC by submerging in liquid nitrogen and milled, with milling in both cases through 1 mm pore sieve. The CH4 yield of each sample was measured in duplicate 160 ml serum bottles over 35 days, according to the method described by McEniry and O’Kiely (2013). Briefly, bottles were incubated at 38oC with a 2inoculum:1substrate volatile solids (VS) ratio and a 10 g kg-1 organic loading rate. Headspace pressure and concentration of CH4 in the biogas were measured. Evaluation of data included corrections for (a) inert gas dilution on day 2, (b) control (inoculum) gas production and (c) standard temperature and pressure conditions. Duplicate analytical estimates were averaged and log transformed data values were analysed with the GLIMMIX procedure (SAS 9.3) using repeated measures as a blocked split-split-plot design. The main plot comprised herbage type, the sub-plot was herbage state (i.e. pre- or post-ensilage herbage) and the sub-sub-plot was the preparation method (i.e. dried or non-dried). Replicate blocks were accounted for in the main plot. Differences between the means were compared using Tukey’s adjustment for multiple comparisons. Results and discussion Whereas preparation method did not have an overall effect (P>0.05) on specific methane yield, there were interactions between preparation method and herbage state (P<0.01) which were of considerable importance (Table 1). Dried pre-ensilage herbage had a higher specific methane yield than non-dried pre-ensilage herbage for H1 (P<0.05), but the similar numerical trend for pre-ensilage herbage H2 and H3 were not significant (P>0.05). Thus, drying had an inconsistent effect on specific methane yield with pre-ensilage herbages. Whereas it could be proposed that mechanical milling dry pre-ensilage herbages through a 1 mm pore sieve produces more fine particles than manually milling frozen pre-ensilage herbage through a similar sieve, and this could increase methane production during digestion, this explanation is difficult to accept in the present study because (a) the effect was not consistent across the three pre-ensilage herbages, and (b) the effect was not evident with the corresponding silages. Drying post-ensilage herbage H1 reduced (P<0.01) specific methane yield, which was the opposite outcome to its corresponding pre-ensilage herbage. In contrast, the absence of an effect of drying on specific methane yield for post-ensilage herbages H2 and H3 (P>0.05) is similar to the outcomes with their corresponding pre-ensilage herbages. This suggests that there may have been a greater loss of volatile compounds during the drying of post-ensilage herbage from H1 than for post-ensilage herbages H2 and H3, with the loss of volatiles resulting in a reduction in the supply of digestible substrates for anaerobic digestion. Thus, oven drying samples did not have a consistent effect on specific methane yield across post-ensilage herbages H1 to H3. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 472 Table 1. Effect of preparation method on the specific methane yield (L CH4 kg-1 VSadded) of three pre-conditioned perennial ryegrasses and their corresponding silages. a Log transformed values (untransformed values in parentheses) b NS = not significant; *P<0.05; ***P<0.01 Herbage type (H) Herbage state (S) Preparation method (P) Specific methane yielda (L CH4 kg-1 VSadded) H1 H2 H3 Grass Non-dried 2.32 (209) Grass Dried 2.45 (279) Silage Non-dried 2.32 (214) Silage Dried 2.40 (167) Grass Non-dried 2.28 (211) Grass Dried 2.31 (249) Silage Non-dried 2.33 (214) Silage Dried 2.22 (214) Grass Non-dried 2.33 (191) Grass Dried 2.33 (202) Silage Non-dried 2.32 (209) Silage Dried 2.26 (182) Standard error of the mean Herbage type (H) 0.03 Herbage state (S) 0.02 Preparation method (P) 0.02 SxP 0.03 Levels of significanceb H * S *** P NS SxP *** Conclusion Oven drying did not have a consistent effect on the specific methane yield of herbages from herbages H1 to H3 and in the case of H1 where the effect was significant these effects were in opposing directions for pre- and post-ensilage herbage states. Thus it is recommended that herbage samples for specific methane yield assessments should not be oven dried when utilizing a small-scale in vitro batch digestion system. References Alomar D., Fuchslocher R. and Stockebrand S. (1999) Effects of oven- or freeze-drying on chemical composition and NIR spectra of pasture silage. Animal Feed Science and Technology 80, 309-319. Angelidaki I., Alves M., Bolzonella N., Borzacconi L., Campos J.L., Guwy A.J., Kalyuzhnyi S., Jenicek P. and van Lier J.B. (2009) Defining the biomethane potential (BMP) of solid organic wastes and energy crops: a proposed protocol for batch assays. Water Science and Technology 59, 927-934. Kreuger E., Achu Nges I. and Björnsson L. (2011) Ensiling of crops for biogas production: effects on methane yield and total solids determination. Biotechnology for Biofuels 4, 44-51. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 473 McDonald P., Henderson A.R. and Heron S.J.E. (1991) The Biochemistry of Silage, (2nd Ed.). Chalcombe Publications, Marlow, Bucks, UK. Markert B. (1995) sample preparation (cleaning, drying, homogenisation) for trace element analysis in plant matrices. Science of the Total Environment 176, 45-61. McEniry J., O’Kiely P. and Crosson P. (2011) The effect of feedstock cost on biofuel cost as exemplified by biomethane production from grass silage. Biofuels, Bioproducts and Biorefining 5, 670-682. McEniry J. and O’Kiely P. (2013) Anaerobic methane production from five common grassland species at sequential stages of maturity. Bioresource Technology 127, 143-150. Navarro-Villa A., O’Brien M., Lopez S., Boland T.M. and O’Kiely P. (2012) In vitro rumen methane output of grasses and grass silages differing in fermentation characteristics using the gas-production technique (GPT). Grass and Forage Science 68, 228-244. Plochl M. and Heiermann M. (2006) Biogas farming in central and Northern Europe: A strategy for developing countries? Agricultural Engineering International: the CIGR Ejournal. Invited Overview No 8. Vol. VIII. Purcell P.J., O’Brien M., Boland T.M. and O’Kiely P. (2011) In vitro rumen methane output of perennial ryegrass samples prepared by freeze drying or thermal drying (40 oC). Animal Feed Science and Technology 166-7, 175182. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 474 Theme 3 posters Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 475 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 476 Area-specific bioenergy potentials from European floodplain grasslands – the DANUBENERGY project Bühle L.1, Hensgen F.1, Goliński P.2 and Wachendorf M.1 1 Department of Grassland Science and Renewable Plant Resources, University of Kassel, Steinstrasse 19, 37213 Witzenhausen, Germany 2 Department of Grassland and Natural Landscape Sciences, Poznan University of Life Sciences, Dojazd 11, 60-632 Poznan, Poland. Corresponding author: buehle@uni-kassel.de Abstract Grasslands in floodplain and other wetland areas fulfil important functions considering flood protection and biodiversity conservation. The maintenance of these areas requires regular cutting and biomass removal. As the nutritive value is frequently low and the possibilities to produce high quality forage is limited due to reduced accessibility in spring, alternative options of use, such as energy production, are sought. This study investigated the energy yield potential of eight grassland sites in Central Europe in the framework of the DANUBENERGY project. Highly varying productivity between 2.6 and 11.8 t ha-1 a-1 was found. Highest net energy outputs can be expected by thermal use of the biomass. Keywords: floodplain areas, wetland, bioenergy, solid fuel Introduction Floodplain grasslands are frequently characterized by valuable plant composition and, furthermore, fulfil various ecological services (Verhoeven and Setter, 2010). Considering flood protection, they play an important role as riparian grasslands are able to reduce the risk of flooding by increased infiltration capacity. In addition, the year-round plant cover minimizes the soil erosion and losses of nutrients. However, the management of these grassland sites is comparatively challenging due to particular soil conditions and, in many cases, the low nutritive value of the vegetation. Hence, these areas get increasingly abandoned in many European regions. The project DANUBENERGY aims at developing and implementing an approach to produce solid fuel from biomass of riparian grasslands in Central Europe, taking into account technical, social and economic aspects (Wachendorf et al., 2009; Blumenstein et al., 2012). Eleven partners from European regions cooperate in finding solutions of site specific wetland management and grassland use for energy production. This paper investigated the biomass yield potential from eight grassland sites and assessed the potential of bioenergy production by comparing different options of energetic use. Materials and methods Eight grassland sites were investigated in 2013 in Italy, Slovakia, Slovenia, Austria, Czech Republic, Hungary, Germany and Poland. At each site, 3 subplots of 5 × 5 m were analysed for botanical composition (plant coverage) and the dry matter (DM) yield was analysed by weighing of biomass from 5 m2 and drying of a subsample at 105 °C for 48 h. Based on the dry matter yield, potential net energy outputs of heat and electricity were calculated by using conversion efficiency that was analysed by Bühle et al. (2012) for seminatural grassland biomass. Three option of energy recovery have been considered: (1) integrated generation of solid fuel and biogas from biomass (IFBB, as a stand-alone and an add-on system to an existing energy plant with surplus heat), (2) anaerobic digestion and (3) direct combustion. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 477 Results and discussion Dry matter yields of the selected grassland sites ranged from 2.6 to 11.8 t DM ha-1 a-1 (Table 1). They were particularly high for typical wetland plant vegetation dominated by Phragmites australis and Phalaris arundinacea. Medium productivity was obtained by grassland vegetation with less influence of water, dominated by Juncus effusus and Carex species. Grasslands with lowest water availability showed the smallest dry matter yields. Table 1. Dominant species, number of plant species, % coverages (cov.) of grasses, sedges/rushes, herbs and legumes and dry matter (DM) yield (± standard error) on selected sites of floodplain grasslands in Central Europe. Site location Dominant species Number of plant species Grasses cov. Sedges/ Rushes cov. Herbs cov. Legumes cov. 8 ±1 42.6 ± 12.6 49.8 ± 12.7 7.7 ± 2.0 - t ha-1a-1 11.8 ± 1.1 42 ±4 46.7 ± 9.9 1.2 ± 0.9 46.0 ± 9.3 6.1 ± 0.4 5.7 ± 1.0 17 ±0 2.9 ± 1.1 65.2 ± 5.8 31.9 ± 4.9 - 5.3 ± 0.5 16 ±1 64.6 ± 13.6 19.5 ± 11.0 15.9 ± 4.6 <0.1 ± <0.1 8.6 ± 4.0 41 ±8 20.8 ± 7.7 1.8 ± 1.6 34.3 ± 4.9 43.1 ± 10.2 2.8 ± 0.3 15 ±3 74.5 ± 8.7 - 10.4 ± 3.8 15.1 ± 5.3 2.6 ± 0.5 15 ±4 20.2 ± 8.9 87.0 ± 15.7 6.1 ± 6.2 1.7 ± 2.2 3.2 ± 0.6 30 ±8 39.3 ± 15.7 46.9 ± 20.0 12.6 ± 3.7 1.2 ± 1.2 11.0 ± 0.9 % Italy Slovakia Slovenia Austria Czech Republic Hungary Germany Poland Phragmites australis Phalaris arundinacea Typha latifolia Agrostis capillaris Holcus lanatus Achillea millefolium Juncus effusus Carex species Alisma plantago-aquatica Agrostis stolonifera Phragmites australis Bolboschoenus maritimus Medicago falcata Bromus erectus Centaurea scabiosa Festuca rubra Agropyron cristatum Achillea millefolium Carex disticha Phragmites australis Holcus lanatus Phalaris arundinacea Carex acutiformis Carex gracilis DM yield The heterogeneity of productivity was reflected by highly varying net energy outputs, as shown in Figure 1. Highest net output could be obtained by direct combustion of the biomass. However, field drying for hay production is frequently not possible on wetland areas and, furthermore, high content of mineral nutrients requires costly combustion technologies. If the harvest takes place at a late date in autumn, the direct combustion can be the appropriate way of energy conversion, for example for Phragmites-dominated biomasses. In case that direct combustion of the dry biomass is not possible, the IFBB process offers an almost equivalent possibility in terms of net outputs, in particular when it is combined with a traditional biogas plant with various synergy effects. Using the grassland biomass by anaerobic digestion led to reduced net gains, what can be explained mainly by low digestibility and incomplete use of the waste heat of biogas combustion. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 478 Site Site location Site location location Site location Site location Site location Italy Italy Italy Slovakia Slovakia Italy Slovakia Slovenia Italy Slovenia Slovakia Slovenia Austria Italy Slovakia Austria Slovenia Austria Czech Republic Slovakia Slovenia Czech Republic Austria Czech Republic Hungary Slovenia Austria Hungary Czech Republic Hungary Germany Austria Czech Republic Germany Hungary Germany Poland Czech Republic Hungary Germany Poland Poland Hungary 0 Germany Poland 0 0 Germany Poland 0 Poland 0 5 5 5 5 5 10 10 10 10 10 Direct combustion (heat) Anaerobic digestion(heat) (heat) Direct combustion Direct combustion IFBB add-on (heat)(heat) Anaerobic digestion (heat) Anaerobic digestion (heat) stand-alone (heat) IFBB combustion add-on (heat)(heat) Direct IFBB add-on (heat) (electricity) Anaerobic digestion stand-alone Anaerobic digestion(heat) (heat) IFBB stand-alone (heat) Anaerobic digestion (electricity) Direct combustion IFBB add-on (heat) (heat) Anaerobic digestion (electricity) Anaerobic digestion (heat) IFBB stand-alone (heat) IFBB add-on (heat)(heat) Direct combustion Anaerobic digestion (electricity) IFBB stand-alone (heat) Anaerobic digestion (heat) 15 20 IFBB 25digestion 30 35 Anaerobic add-on (heat) (electricity) 15 20 25 30 35 -1 -1 IFBB stand-alone (heat) MWh ha a20 15 25 30 35 -1 -1 Anaerobic digestion (electricity) MWh ha a 15 25 30 35 -1 20 -1 MWh ha a 15 ha-1a-120 MWh 25 30 35 Figure 1. Net energy output (heat and electricity) of the integrated generation of solid fuel and biogas from -1 -1 MWh ha a20 10 combustion 15 25 30 35 biomass, anaerobic0digestion5 and direct on selected sites of floodplain grasslands in Central Europe. Conclusion -1 -1 MWh ha a In order to maintain grasslands in floodplain areas and the related functional services, e.g. biodiversity conservation and floodplain protection, energy-efficient conversion technologies are required in regions with declining forage use. Productivity and net energy output of the conversion technologies considered in this study were highly varying. With the IFBB conversion as an add-on system to existing bioenergy plants with surplus heat, highest net outputs could be obtained. Acknowledgement The authors would like to thank the EU for co-financing of the DANUBENERGY project within the framework of the Central Europe Programme (4CE561P3). References Blumenstein B., Bühle L., Wachendorf M. and Möller D. (2012) Economic assessment of the integrated generation of solid fuel and biogas from biomass (IFBB) in comparison to different energy recovery, animal based and non-refining management systems. Bioresource Technology 119, 312-323. Bühle L., Hensgen F., Donnison I., Heinsoo K. and Wachendorf M. (2012) Life cycle assessment of the integrated generation of solid fuel and biogas from biomass (IFBB) in comparison to different energy recovery, animal-based and non-refining management systems. Bioresource Technology 111, 230-239. Verhoeven J.T.A. and Setter T.L. (2010) Agricultural use of wetlands: opportunities and limitations. Annals of Botany 105, 155-163. Wachendorf M., Richter F., Fricke T., Graß R. and Neff R. (2009) Utilization of semi-natural grassland through integrated generation of solid fuel and biogas from biomass. I. Effects of hydrothermal conditioning and mechanical dehydration on mass flows of organic and mineral plant compounds, and nutrient balances. Grass and Forage Science 64, 132-143. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 479 Permanent grasslands under different management as potential source of biomass for combustion in the Czech Republic Štýbnarová M., Mičová P., Karabcová H. and Látal O. Agrovýzkum Rapotín Ltd., Výzkumníků 267, 788 13 Vikýřovice, Czech Republic Corresponding author: marie.stybnarova@vuchs.cz Abstract The aim of this work was to evaluate different grassland managements from the viewpoint of utilization of biomass for combustion. A small-plot trial was established in the locality of Rapotín (Czech Republic) in 2003. It was managed during seven years with four levels of grassland utilization: intensive, medium intensive, low intensive and extensive. Four levels of fertilization were applied for each utilization treatment: nil-fertilization, P30K60, N90P30K60, and N180P30K60 (pure nutrients). Nitrogen (N) fertilization and decreasing intensity of utilization significantly increased dry matter (DM) yield up to 8.0 t ha-1. The content of crude protein also increased with N fertilization up to 176.7 g kg-1 DM (intensively utilized treatment). The ash concentration was significantly higher in the intensive and medium intensive treatments (106.4 g kg-1 DM on average) than in low intensive and extensive treatments (93.3 g kg-1 DM on average), whereas the influence of fertilization was not significant. Excessive forage from extensively managed grasslands, with one or two late cuts and low level of fertilization, could be regarded as a favourable biofuel. Keywords: grasslands, yield, utilization, fertilization, combustion Introduction In the Czech Republic the surplus of roughages in 2010 reached 995 000 t of dry matter (DM). This represents the forage for 212 000 livestock units (LU) for which there was no use, and is about one-third of forage production from permanent grasslands and an area of about 300 000 ha of grasslands in the Czech Republic (Kohoutek, 2012). A possible solution for dealing with surplus grassland forage could be its utilization as biomass for producing energy. In addition, appropriate grassland management for energy use can contribute to reduction of greenhouse gas emissions and maintain essential ecological grassland functions. Maximizing the use of renewable resources is one of the key points of EU energy policy. The use of grassland biomass is intensively discussed for bioethanol and biogas production, and it is also considered as an alternative feedstock for solid biofuel. Each form of energetic utilization requires specific characteristics of the grassland biomass, being highly variable and depending mainly on grassland management (Goedeke et al., 2008). DM yield and chemical composition can limit the suitability of grassland biomass as a fuel through environmentally damaging emissions (Tonn et al., 2012). Using grassland biomass for combustion is a subject of broad research and is established in practice. Firing herbaceous biomass requires various specific adaptations of the different combustion technologies. Frydrych and Andert (2013) were engaged in alternative utilization of biomass from species-rich meadows in the Czech Republic; whole square hay bales were utilized for energy by their combustion in the unique biomass boiler (STEP Trutnov Comp). The aim of this work was to evaluate the production potential of grasslands under different management from the viewpoint of biomass utilization for combustion. Materials and methods A small-plot experiment (plot size 12.5 m2) with permanent grasslands was established in Rapotín (Czech Republic) in 2003. The experimental locality was at 390 m a.s.l. During the 7year monitoring period, average rainfall was 668 mm and temperature was 7.2 °C. The Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 480 dominant species at the beginning of the trial were: Dactylis glomerata, Poa pratensis, Lolium perenne, Trifolium repens and Taraxacum sect. Ruderalia. Grasslands were managed with four levels of intensity of utilization: I1 = intensive (1st cut by 15 May, 4 cuts per year – cuts at 45day intervals); I2 = medium intensive (1st cut between 16 and 31 May, 3 cuts per year at 60-day intervals); I3 = low intensive (1st cut between 1 and 15 June, 2 cuts per year at 90-day intervals); and I4 = extensive (1st cut between 16 and 30 June, 1 or 2 cuts per year, second cut after 90 days). Each utilization treatment was divided to give four levels of fertilization: F0 = no fertilization, FPK= N0P30K60; FPKN90= N90P+30K60, FPKN180= N180+P30K60 (pure nutrients). The experiment was a split plot design with four replicate blocks. Total DM yield was measured for all plots in each of the harvest years. The samples collected from the plots were analyzed for the contents of crude protein (CP), ether extract (EE), crude fibre (CF) and ash (A) by the Weende analysis. Brutto energy (BE) was predicted by the equations of Sommer et al. (1994). Statistical analyses were performed using ANOVA (package Statistica 10) with multiple comparisons according to Tukey (P < 0.05). Results and discussion A significant effect (P < 0.05) of N180PK fertilization was observed, with fertilized treatments showing a higher mean value of DM yield (7.97 t ha-1) than unfertilized (4.81 t ha-1) (Table 1). Table 1. Grassland total DM yield, content of nutrients and energy by different management over seven harvest years. Data is derived from a total of particular cuts in each harvest year. Treatment DM yield [t ha-1] CP [g kg-1 DM] EE [g kg-1 DM] CF [g kg-1 DM] A [g kg-1 DM] BE [MJ kg-1 DM] I1F0 4.63 146.3 32.6 239.3 104.2 17.89 I2F0 4.70 127.5 30.0 258.1 103.1 17.78 I3F0 4.89 109.3 27.9 281.0 93.4 17.84 I4F0 5.02 102.7 25.3 303.8 94.0 17.79 I1FPK 4.63 143.2 33.5 242.7 110.7 17.78 I2FPK 4.84 127.1 32.1 260.1 109.2 17.72 I3FPK 5.17 107.3 28.1 285.6 96.1 17.80 I4FPK 5.21 104.2 26.5 301.1 97.3 17.76 I1FN90PK 6.90 157.1 33.4 234.5 106.2 17.93 I2FN90PK 6.70 141.0 32.8 256.3 108.3 17.83 I3FN90PK 7.31 120.2 27.3 296.3 94.5 17.92 I4FN90PK 7.53 118.9 26.9 307.5 97.4 17.88 I1FN180PK 7.96 176.7 34.1 235.5 103.9 18.11 I2FN180PK 7.83 159.5 32.3 259.6 105.3 18.00 I3FN180PK 8.03 142.6 28.4 295.0 87.7 18.20 I4FN180PK 8.08 128.0 27.1 306.0 86.1 18.14 ANOVA F-ratio Intensity of utilization 11.8** 34.4** 17.6** 36.2** 6.5** 1.20 Fertilization 589.1** 17.7** 0.7 0.1 1.3 10.2** Year 360.6** 7.3** 3.9** 0.8 7.3** 13.2** Intensity of utilization: I1 = intensive, I2 = medium intensive, I3 = low intensive, I4 = extensive. Fertilization: F0 = nil-fertilization, FPK= N0P30K60, F N90PK= N90+P30K60, F N180PK= N180+P30K60 (pure nutrients). * P < 0.05; ** P < 0.01 With decreasing intensity of grassland utilization DM yield significantly increased; this was most apparent in unfertilized treatments. These results are in line with Kohoutek (2012). The heating value of the sward is determined by a number of factors, whereas grassland management affects not only DM yield but also the chemical content of the biomass. Contents of ash and of several mineral elements can lead to problems during the combustion process Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 481 (Tonn et al., 2012). Within our study, the ash concentration in intensively and medium intensively utilized treatments differed significantly, with the higher ash concentration (106.4 g kg-1 DM on average) from those utilized 'low intensively' and extensively (93.3 g kg-1 DM on average). The reason for this may be that soil contamination was higher by harvesting the swards of lower height. Regarding the energy content we found significant effect of N 180PK fertilization, whereas the highest concentration of BE showed grasslands utilized at 'low intensively' and extensively by N180PK fertilization (18.20 MJ kg-1 DM and 18.14 MJ kg-1 DM, respectively). We should be also aware of concentration of crude protein (CP), which increased significantly with N fertilization up to 176.7 g kg-1 DM (intensively utilized treatment). Nitrogen contained in the vegetation has a major role during combustion as it is the source of undesired NOx emissions (Harndorf et al., 2009). Taking into consideration that N concentration in the solid fuel shows a logarithmic correlation with the NOx emissions (van Loo and Koppejan, 2008) it can be stated that the risk of environmental pollution increases by excessive usage of nitrogen fertilizers. Conclusion On the basis of our results about DM yield and chemical properties of harvested biomass we can conclude that the grasslands utilized extensively with one or two late cuts by lower doses of fertilizers were more favourable for combustion within our study. Further research of this issue is needed. Acknowledgement The work was supported by the institutional support for the long-term conceptual development of the research organization, Ministry of Agriculture Decision No. RO0313 from 28 February 2013 and by the project INGO No. LG 13019. References Frydrych J. and Andert D. (2013) Alternative utilization of biomass production from species-rich meadows for energetic purposes. (In Czech) [online]. 2013-05-27 [cit. 2013-10-15]. Available at: <http://biom.cz/cz/odborneclanky/alternativni-vyuziti-produkce-lucnich-porostu-s-vysokou-druhovou-diverzitou-pro-energeticke-ucely>. Goedeke K., Prochnow A., Heiermann M., Ploechl M. and Hochberg H. (2008) Prospects of grassland use for bioenergy resource. Aspects of Applied Biology 90, 19-25. Harndorf J., Fricke T., Weisser W.W., Weigelt A. and Wachendorf M. (2009) Combustion of grassland biomass: Effects of species richness and functional groups on energy parameters. Grassland Science in Europe 14, 517-9. Kohoutek A. (2012): Sustainable management of permanent grasslands in the agricultural system of the Czech Republic). In: Kohoutek A., Pozdíšek J. (Eds) Proceedings from international conference 'Sustainable systems of grassland management in the Czech Republic and their perspective', (5.11.2012, Rapotín), pp. 19-41. Loo van S. and Koppejan J. (2008) The handbook of biomass combustion and co-firing. Earthscan, London, Sterling, VA, USA, 442 p. Sommer A., Čerešňáková Z., Frydrych Z., Králík O., Králíková Z., Krása A., Pajtáš M., Petrikovič P., Pozdíšek J., Šimek M., Třináctý J., Vencl B. and Zeman L. (1994) Nutrient requirements and tables of nutrient value of ruminant feeds. 1st ed. Pohořelice ČZS VUVZ, CR, 198 p. (in Czech) Tonn B., Thumm U. and Claupein W. (2012) Suitability of semi-natural grassland biomass for combustion and the effect of quality optimization strategies. Grassland Science in Europe 17, 445-447. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 482 What is the biomethane production potential of the available grassland biomass resource in Ireland? McEniry J.1, O’Kiely P.1, Wall D.M.1,2 and Murphy J.D.2 1 Animal & Grassland Research and Innovation Centre, Teagasc, Grange, Dunsany, Co. Meath, Ireland; 2 Bioenergy and Biofuels Research Group, Environmental Research Institute, University College Cork, Ireland. Corresponding author: Padraig.OKiely@teagasc.ie Abstract Grassland dominates the Irish landscape and is used primarily for ruminant production. In an average year, grass growth is in excess of livestock requirements and the potential exists to enlarge this excess. Four scenarios (two levels of biomass in excess of ruminant requirements; two rates of capture of the biomass excess) were assessed to elucidate the potential contribution of the available grassland resource to Ireland’s 2020 renewable energy targets. Grassland can make a significant contribution to 2020 renewable energy targets in Ireland without compromising the ability of traditional agricultural systems to meet the needs of ruminant livestock, even if the latter undergo planned increases in numbers. The potential exists to provide up to 14.83, 27.43 and 25.24% of the electrical, thermal and transport energy requirement by 2020. These become more significant when double-crediting is applied, as allowed for biofuels produced from lignocellulose feedstocks in the European Union. Keywords: Grassland, biomass, anaerobic digestion, methane, renewable energy target Introduction Grassland is the dominant biomass resource in Ireland and underpins most ruminant production systems. Not only is grassland plentiful, with 92% of agricultural land under the crop, but annual yields of up to 16 tonnes total solids (TS)/ha can be achieved (O’Donovan et al., 2011). Grass biomethane has been proposed as part of a renewable energy solution for Ireland, with biogas from the anaerobic digestion of grass being upgraded to biomethane and utilized as a transport fuel or injected into the natural gas grid (Murphy and Power, 2009; Smyth et al., 2009). McEniry et al. (2013) recently reported that under current grassland management practices there is an estimated annual grassland resource of ca. 1.7 million tonnes TS available in excess of livestock requirements. Furthermore, increasing nitrogen (N) fertilizer input combined with increasing the grass utilization rate of cattle has the potential to significantly increase this resource to 12.2 million tonnes TS/annum. The objective of this study was to determine the potential contribution of this available grassland resource to Ireland’s 2020 renewable energy targets under specific scenario conditions. Materials and methods McEniry et al. (2013) calculated the annual grassland resource available in Ireland as the difference between current estimated grass supply and the grass requirement of the national cattle herd and sheep flock. Briefly, grass supply on various soil types was calculated as a function of annual inorganic N fertilizer input, while grass requirement was calculated based on the annual grass and grass silage TS intake requirements of each category of cattle and sheep. Using data derived from McEniry et al. (2013; Table 1), four contrasting scenarios (two levels of biomass in excess of ruminant requirements; two rates of capture of the biomass excess) were investigated in the current study to determine the potential contribution of this available grassland resource to Ireland’s 2020 renewable energy targets. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 483 The potential biomethane and energy yield of the available grassland resource was calculated (Tables 1 and 2). The volatile solids (VS) concentration of grass biomass was assumed to be 0.9 tonne VS/tonne TS, with a potential CH4 yield of 400 mn3 CH4/tonne VS (Nizami et al., 2012). The potential contribution of this available grassland resource to Ireland’s renewable energy targets is expressed as a percentage of the expected energy in transport (118 Petajoule (PJ)/a), and thermal (173 PJ/a) and electrical (112 PJ/a) energy demand in Ireland in 2020 (Clancy and Scheer, 2011). Table 1. Annual available grassland resource for anaerobic digestion (AD) and potential biomethane yield Scenario Grassland Resource Resource Resource Specific CH4 Biomethane resource captured for AD for AD yield yield (t TS/a)1 (%) (t TS/a) (t VS/a)2 (mn3 CH4/t VS)3 (million mn3/a) 1 1,669,648 10 166,964 150,268 400 60.11 2 1,669,648 30 500,894 450,805 400 180.32 3 12,203,800 10 1,220,380 1,098,342 400 439.34 4 12,203,800 30 3,661,140 3,295,026 400 1,318.01 1 Derived from McEniry et al. (2013); TS = total solids. 2 VS = volatile solids; 0.9 t VS/t TS. 3 Derived from Nizami et al. (2012). Table 2. Potential contribution of the available grassland resource to Ireland’s 2020 renewable energy targets. Contribution to renewable energy targets (% of energy demand) 2 Electrical energy Thermal energy Energy in transport (RES-E)3 (RES-H) (RES-T) 1 2.16 0.68 1.25 1.15 2 6.49 2.03 3.75 3.45 3 15.82 4.94 9.14 8.41 4 47.45 14.83 27.43 25.24 1 PJ = petajoule; energy value of biomethane = 36 megajoules/mn3. 2 Expected energy in transport and thermal and electrical energy demand in Ireland in 2020 is forecast to be 118, 173 and 112 PJ/a, respectively (Clancy and Scheer, 2011). 3Assume 35 % electrical efficiency. Scenario Biomethane yield (PJ/a) 1 Results and discussion Fossil fuels account for 94% of the energy usage in Ireland (Howley et al., 2012). In 2010, the Irish government set targets of 40, 12 and 10% renewable energy share in gross energy production for electrical (RES-E), thermal (RES-H) and transport energy (RES-T) by 2020, respectively (NREAP, 2010). Under current grassland management practices there is an estimated annual grassland resource of ca. 1.7 million tonne TS available in excess of livestock requirements (McEniry et al., 2013). Table 1 outlines the potential biomethane yield generated from the capture of 10 (Scenario 1) and 30% (Scenario 2) of this resource, respectively. This has the potential to contribute 0.68 and 2.03% of Ireland’s 2020 electrical energy requirements (Table 2). McEniry et al. (2013) also reported that there was considerable scope for increasing N fertilizer inputs to increase grassland productivity and that there was significant potential for improvement of on-farm grass utilization rates through greater adoption of currently advised grassland management programmes. Increasing N-fertilizer input combined with increasing the grass utilization rate of cattle has the potential to significantly increase the available grassland resource to 12.2 million tonne TS/a (Scenarios 3 and 4; Table 1). This has the potential to contribute up to 14.83% of Ireland’s 2020 electrical energy requirements (Scenario 4; Table 2). Grass biomethane has been proposed as the best energy crop for meeting the 2020 renewable transport energy target in Ireland (Korres et al., 2010). Utilization of 30% of the current available grassland resource (Scenario 2) for energy in transport has the potential to make a significant contribution to RES-T (Table 2), considering the EU Renewable Energy Directive Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 484 allows a double credit for biofuels produced from lignocellulosic feedstocks (i.e. 6.9% contribution to RES-T in 2020 for Scenario 2). Conclusion The available grassland biomass resource in Ireland can make a significant contribution to the 2020 renewable energy targets. Furthermore, considerable potential exists to make a significant (up to 14.83% renewable energy share; Scenario 4) contribution to RES-E or to surpass the targets set by RES-T (up to 16.82 and 50.48% renewable energy share for Scenarios 3 and 4, respectively), through changes to grassland management practices. Of particular significance is that this may be achieved without competing with traditional agricultural production systems. Acknowledgements Funding was provided under the National Development Plan, through the Research Stimulus Fund (#RSF 07 557), administered by the Department of Agriculture, Food and the Marine. References Clancy M. and Scheer J. (2011) Energy forecasts for Ireland to 2020. Sustainable Energy Association of Ireland. 74 pages. <http://www.seai.ie/Publications/Statistics_Publications/EPSSU_Publications/Energy_Forecasts_for_Ireland_for _2020_-2011_Report.pdf> Howley M., Dennehy E., Ó Gallachóir B. and Holland M. (2012) Energy in Ireland 1990-2011. Sustainable Energy Association of Ireland. 96 pages. < http://www.seai.ie/Publications/Statistics_Publications/Energy_in_Ireland/Energy_in_Ireland_1990__2011.pdf> Korres N.E., Singh A., Nizami A.S. and Murphy J.D. (2010) Is grass biomethane a sustainable transport biofuel? Biofuels, Bioproducts and Biorefining 4, 310-325. National Renewable Energy Action Plan (NREAP) – Ireland (2010) Submitted by the Department of Communications, Energy and Natural Resources to the EU Community under Article 4 of Directive 2009/28/EC. 165 pages < http://www.dcenr.gov.ie/NR/rdonlyres/C71495BB-DB3C-4FE9-A725-0C094FE19BCA/0/2010NREAP.pdf> Nizami A.S., Orozco A., Groom E., Dieterich B. and Murphy J.D. (2012) How much gas can we get from grass? Applied Energy 92, 783-790. McEniry J., Crosson P., Finneran E., McGee M., Keady T.W.J. and O’Kiely P. (2013) How much grassland biomass is available in Ireland in excess of livestock requirements? Irish Journal of Agriculture and Food Research 52, 67-80. Murphy J.D. and Power N.M. (2009) An argument for using biomethane generated from grass as a biofuel in Ireland. Biomass and Bioenergy 33, 504-512. O’Donovan M., Lewis E. and O’Kiely P. (2011) Requirements of future grass based ruminant production systems in Ireland. Irish Journal of Agricultural and Food Research 50, 1-21. Smyth B.M., Murphy J.D. and O'Brien C. (2009) What is the energy balance of grass biomethane in Ireland and other temperate northern European climates? Renewable and Sustainable Energy Reviews 13, 2349-2360. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 485 Evaluation of biomass yield of energy crops using waste products as fertilizers Rancane S., Gutmane I., Berzins P., Stesele V. and Dzene I. LLU Research Institute of Agriculture, Zemkopibas inst. 7, Skriveri, LV-5125, Latvia Corresponding author: sarmite.rancane@inbox.lv Abstract Biomass from perennial grasslands might be a valuable renewable resource. At the same time the successful management of large amounts of waste products is very important. Field experiments were carried out in the central part of Latvia (56° 42' N and 25° 08' E latitude) to estimate reed canary grass (Phalaris arundinacea L.), festulolium (×Festulolium pabulare) and galega (Galega orientalis Lam.) dry matter (DM) yield in two years of sward use. The fertilization treatments were control (no fertilization), sewage sludge, biogas digestate and wood ash. Doses of fertilization were calculated in order to provide similar amount of plantavailable potassium (K) from each fertilizer. Trial results showed a significant dry matter yield dependence on the grass species and the applied fertilizer. The significant dry matter yield increase on average for two years of use at two cutting regimes for grasses and galega provided using sewage sludge (1.59 t ha-1 and 1.41 t ha-1 respectively) and digestate (1.27 t ha-1 and 1.34 t ha-1 respectively). Keywords: fertilization, perennial grasses, biomass yield, by-products, renewable resources Introduction Grasses and forage legumes used for biofuel production can provide a number of ecosystem services. Grassland can fix atmospheric nitrogen, improve soil texture and increase biodiversity at the field and landscape level (Prade et al., 2013). Grasslands, compared to other crops for agro-fuel production, can be produced on marginal agricultural land; they do not require large amounts of fertilizers and pesticides (Kryzeviciene et al., 2008; Peeters, 2008). Biomass cultivation could become an alternative to those farmers in Latvia whose agricultural land is not suitable for cultivation of cereals. The use of renewable energy resources will reduce the dependence on imports of fossil fuels (Adamovics et al., 2011). Reed canary grass, festulolium and galega are well known as productive species suitable for biogas or fuel-pellet production (Rancane et al., 2013). There is an important potential to use different by-products for grassland fertilization. Returning waste products to agricultural land by the application of waste-based fertilizer products is way to solve the problem of disposal of waste products and promote nutrient recycling (Bougnom et al., 2012; Brod et al., 2012). The objective of the current research was to study the effect of different bio-energy products and municipal waste products (digestate, wood ash and sewage sludge) used as fertilizers, the choice of grass species, and the intensity of management on grass and legume productivity. Materials and methods Field trials were performed at the Research Institute of Agriculture in Skriveri on Phaeozems (soil pHKCl 6.1, plant available phosphorus (P2O5) – 277.1 mg kg-1, potassium (K2O) – 136.8 mg kg-1, organic matter content 23 g kg-1). The trials were sown in July 2011 without a cover crop, in randomized block design with four replications, and a 20 m2 harvested plot size. Three potential energy crop grass species were investigated for dry matter yield: reed canary grass, festulolium (tall fescue type) and galega. Two intensities of management were applied: 3 cuts per vegetation period and 1 cut in October. The following fertilization treatments were used: control (no fertilization), sewage sludge, biogas digestate and wood ash. Doses of fertilizers Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 486 were calculated in order to provide a similar amount of plant-available potassium (K) from digestate and wood ash (Table 1). Table 1. The nutrient content and applied amount of fertilizers Nutrient content in the fertilizers dry matter, g kg-1 Applied nutrient per year, kg ha-1 N P2O5 K2O N P2O5 K2O Control 0 0 0 0 0 0 Wood ash 0.73 14.7 88.2 1 20 120 Digestate 29.2 22.4 53.8 65 50 120 Sewage sludge 46.3 71.0 2.5 150 230 8 Fertilizer The low content of potassium in sewage sludge did not allow the 120 kg ha-1 input of K2O per year to be reached; therefore calculation of the applied sewage sludge dose was conducted according to the input of medium-low nitrogen (N) dose (N150). Dry matter yield were determined for two years of sward use (2012 and 2013). The experimental data were statistically analysed using by two-way analysis of variance with species and fertilizer as factors, and the difference among means was detected by LSD at the 0.05 probability level (Excel for Windows 2003). Results and discussion The average dry matter yields of the two grasses and one legume were considered as satisfactory in the first (6.76 t ha-1) and moderate in the second (5.48 t ha-1) production year. The lowest yield in the second year can be explained by meteorological conditions – very late spring and an insufficient amount of precipitation and unfavourable distribution of rainfall in summer Using a 3-cut cutting frequency, there was a significant DM yield increase, on average for the two years, provided by sewage sludge (1.65 t ha-1) and digestate (1.32 t ha-1) for all investigated species (Table 2). Table 2. Average dry matter yield for two years of sward use at frequent cutting, t ha -1 Species (FB) Fertilizer (FA) Mean Control Wood ash Sewage sludge Digestate Phalaris arundinacea 4.64 5.16 6.18 6.30 5.57 Festulolium 4.32 4.98 5.99 5.42 5.17 Galega orientalis 6.71 7.36 8.46 7.88 7.60 Mean 5.22 5.83 6.87 6.54 LSD0.05 for DM yield: FA= 0.89; FB= 0.77; FAB= 1.54 The highest average DM yield at the frequent cutting regime was given by galega, as this Nfixing legume had good regrowth after cuts. The average festulolium and reed canary grass DM yields did not differ significantly. The same effect of fertilization was observed with cutting once in a season – at crop senescence. Significant DM yield increases were provided sewage sludge (1.40 t ha-1) and digestate (1.27 t ha-1) (Table 3). Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 487 Table 3. Average dry matter yield for two years of sward use at autumn cutting, t ha -1 Species (FB) Fertilizer (FA) Mean Control Wood ash Sewage sludge Digestate Phalaris arundinacea 6.13 6.04 7.21 7.67 6.76 Festulolium 4.91 5.24 6.96 5.66 5.69 Galega orientalis 5.14 5.64 6.20 6.64 5.91 Mean 5.39 5.64 6.79 6.66 LSD0.05 for DM yield: FA= 0.82; FB= 0.71; FAB= 1.42 Analysis of variance showed that the influence of species factor on DM yields was significant. The highest average DM yield was given by reed canary grass under at the management with one late cut. By mowing galega once per season in autumn a relatively low DM yield was obtained, which can be explained by partial plant defoliation. The DM yield increase achieved by applying wood ash was not significant for either of the intensities of management. On average for two production years, intensity of management did not provide significant differences of DM yield. Conclusion The productivity of perennial grass biomass was dependent on the type of applied fertilizer and the species. On average for two production years, the highest dry matter yield increase was provided by fertilizing with sewage sludge. For galega, the highest average DM yield was produced using a 3-cut cutting frequency, but for reed canary grass the highest average DM yield was at the management with one late cut. Acknowledgements This work is supported by the ERDF within the Project Elaboration of models for establishment and management of multifunctional plantations of short rotation energy crops and deciduous trees No 2010/0268/2DP/2.1.1.1.0/10/APIA/VIAA/118. References Adamovics A., Dubrovskis V., Plume I. and Adamovica O. (2011) Biogass production from Galega orientalis Lam. and galega-grass biomass. Grassland Science in Europe 16, 416-418. Bougnom B.P. and Niederkofler C., Knapp B.A., Stimpfl E. and Insam H. (2012) Residues from renewable energy production: Their value for fertilizing pastures. Biomass and Bioenergy 39, 290-295. Brod E., Haraldsen T. K. and Breland T. A. (2012) Fertilization effects of organic waste resources and bottom wood ash: results from a pot experiment. Agricultural and Food Sciences 21(4), 332-347. Kryzeviciene A., Jasinskas A. and Gulbinas A. (2008) Perennial grasses as a source of bioenergy in Lithuania. Agronomy Research 6 (Special issue), 229-239. Peeters A. (2008) Challenges for grasslands, grassland-based systems and their production. potential in Europe. Grassland Science in Europe 13, 9-24. Prade T., Svensson S. E., Mattsson J.E., Carlsson G., Bjornsson L., Borjesson P. and Lantz M (2013) EU sustainability criteria for biofuels potentially restrict ley crop production on marginal land for use as biogas substrate. Grassland Science in Europe 18, 528-530. Rancane S., Gutmane I., Berzins P., Jansone B., Stesele V. and Jansons A. (2013) The effects of different fertilizers and cutting frequencies on yield in three energy crops. Grassland Science in Europe 18, 551 – 553. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 488 Utilization of reed canary grass in pellet production Platace R. and Adamovics A. Latvia University of Agriculture, Liela iela,2, Jelgava, Latvia Corresponding author: aleksandrs.adamovics@llu.lv Abstract Pellets are renewable and environmentally friendly biofuel. Perennial grasses may be one of the raw materials used for pellet production. In some countries they are the main source of bioenergy, since their longevity also ensures substantial and stable production of biomass under less favourable climatic conditions. The research aims at finding out the suitability of reed canary grass (RCG) for production of pellets by studying the influence of various factors (e.g., component proportions in pellet, fertilizers applied) on suitability for combustion of the pellets. The study also investigated effect that the amount of nitrogen fertilizers applied on crops had on the pellet quality and suitability for combustion. Analysis of the combustion suitability showed the highest reed canary grass biomass indicators for component proportion 1/3 (RCG/wood). Therefore, reed canary grass may be cultivated as an alternative energy crop that may be utilized for pellet production. Keywords: reed canary grass, pellets, combustion ability Introduction In some countries perennial grasses are the main source of bioenergy, since their longevity ensures substantial and stable production of biomass under less favourable climatic conditions. Grasses tend to occupy large share of the agricultural area and are essential for the agricultural production sector. Pellets are an environmentally friendly biofuel, as they may be produced from renewable sources. This fuel is produced from dried woodworking waste – sawdust, shavings, bark, twigs, branches etc., but pellets may also be made from various grasses, mixtures thereof, natural grass from meadows, or reed (Lewandowski et al., 2003; Adamovics et al., 2007; Adamovics et al., 2009; Strasil, 2012). The energy released by 1 kg of wood pellets equals that from 0.5 l of solid fuel. It has been found that substitution of heavy fuel oil with pellets results in a reduction of approximately 4.8 t CO2 released annually, while substitution of gas in 2.5 t less. The transport and storage of pellets are not hazardous to the environment. Ash produced from burning can usually be used as a source of fertilizers, as its content of chemical elements does not exceed required standards (Adamovics et al., 2007). RCG biomass is currently considered as one alternative raw material source that can be used for pellet production. This grass plant has characteristics of good persistence under local climatic conditions and has high biomass yield per hectare. The aims of this research were to determine the suitability of reed canary grass for production of pellets by studying the influence left by various factors (e.g., component proportions in pellet, fertilizers applied) on suitability for combustion of the pellets. Materials and methods The Research objects were reed canary grass (Phalaris arundinacea L.) and energy wood species: osier (Salix viminalis L.) and poplar (Populus tremula L.). The energy woods (osier (sp ‘Tordis’) and poplar were grown at the Vezaiciai agricultural research institute in Lithuania. Pellets were produced by mixing reed canary grass (variety ‘Marathon’) with osier or poplar in various dry matter proportions – 1/3, 1/1, or 3/1. The pellets were made from 100% natural ingredients – chopped wood (osier or poplar) and chopped RCG biomass. Each biomass sample was tested three times. Within the pellet manufacturing process, plant biomass was chopped and ground in the laboratory mill (ЭМ-ЗА УХЛ 4.2) and afterwards the powder acquired was formed into a pellet with the hand press (‘IKA WERKE’). Combustion ability in the samples Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 489 was measured with the help of a calorimeter (capsule ‘IKA C 5003’) based on LST CEN/TS 14918:2006 standard; lignin content was in compliance with the LVS EN ISO 13906: 2008, and ash content, with ISO 5984: 2002 standards. The trial data were processed using correlation, regression and variance analyses (ANOVA) and descriptive statistics with Microsoft Excel for Windows 2000 (Arhipova and Balina, 2006). The means are presented with their LSD test. Results and discussion Combustion ability is the main feature of fuel as it indicates the its efficiency (Friedl et al., 2005). Analysis of combustion ability of pellets made from reed canary grass and osier at various proportions depending on amount of nitrogen fertilizer applied showed the best combustion indicators for samples not treated with nitrogen. The best combustion ability was found in pellets made from reed canary grass and osier in combination of 1/3. Samples that were previously treated with nitrogen fertilizer indicated a combustion ability (energy value) of 18.57 MJ kg-1, whereas ones from RCG not treated with fertilizer were 18.69 MJ kg-1 (Figure 1). Figure 1. Combustion ability of reed canary grass pellets depending on amount of nitrogen fertilizer applied The highest combustion ability of pellets made from reed canary grass and poplar was in the proportion 1/3, and their energy value comprised 18.59 MJ kg-1 with nitrogen fertilizer, and 18.83 MJ kg-1 without fertilizer. Our research did not show that nitrogen fertilizers have a notable influence on combustion ability; therefore Figure 2 summarizes pellet combustion ability, average lignin and ash content as well as component proportions. Figure 2. Combustion ability, lignin and ash content in pellets made from biomass in various proportions Reed canary grass biomass has high ash content (5.59%), but it is lower in pellets made in proportion 1/3 (RCG/wood); moreover both osier and poplar indicated relatively lower ash contents of 3.43% and 4.18%, respectively. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 490 The lignin content in reed canary grass biomass accounts for 10.78%, whereas higher lignin content was observed in samples made from reed canary grass and wood in proportion 1/3: 14.97% with osier and 17.28% with poplar. The research showed that pellets made solely from reed canary grass biomass have lower combustion ability than pellets made from reed canary grass together with osier and poplar. The highest reed canary grass combustion ability indicators were recorded for combination 1/3: 18.71MJ kg-1 for pellets RCG/osier and 18.63MJ kg-1 for RCG/poplar pellets. Conclusions Reed canary grass is suitable for pellet production. Nevertheless, it is advisable to mix it in proportion of 1/3 with wood. Average combustion ability of pellets made from biomasses in various proportions varies between 18.53 MJ kg-1 in samples not treated with nitrogen, and 18.34 MJ kg-1 in samples on which nitrogen was applied. The highest combustion ability was recorded in pellets made from reed canary grass and wood in combination 1/3. Nitrogen fertilizers leave a slight influence on pellet quality and combustion ability. Characteristics and chemical composition of reed canary grass are similar to timber, but when burnt, pellets of this grass produce more ash, and therefore when producing pellets it should be mixed with sawdust or woodchips. Acknowledgements The research was supported by the grant of the Ministry of Agriculture of the Republic of Latvia, Agreement No. 2013/86. References Adamovics A., Agapovs J., Arsanica A., Danilevics A., Dizbite T., Dobele G., Dubrovskis V., Iesalnieks V., Jure M., Kronbergs E., Lazdina, D., Lazdins A., Teliseva G., Urbanovics I., Varika A., Vedernikovs N., Zandersons J. and Zurins A. (2007) Energetisko augu audzesana un izmantosana (Energy crops cultivation and usage), Valsts SIA Vides projekti, Riga, 190 pp. Adamovics A., Dubrovskis V., Plume I., Jansons A., Lazdina D. and Lazdins A. (2009) Biomasas izmantosanas ilgtspejibas kriteriju pielietosana un pasakumu izstrade (Biomass sustainability criteria and the application of measures), Valsts SIA Vides projekti, Riga, 172 pp. Arhipova I. and Balina S. (2006) Statistika ekonomika. Risinajumi ar SPSS un Microsoft Excel (Economics statistics. Solutions with SPSS and Microsoft Excel), Riga, Datorzinibu centrs, 352 pp. Friedl A., Padouvas E., Rotter H. and Varmuza K. (2005) Prediction of heating values of biomass fuel from elemental composition. Analytica Chimia Acta 544, 191-198. Lewandowski I., Scurlock J.M.O., Lindvall E. and Christou M. (2003) The development and current status of perennial rhizomatous grasses as energy crops in the US and Europe. Biomass and Bioenergy 25, 335-361. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 491 Can specific methane yield of perennial ryegrass be reliably predicted? Herrmann A., Techow A., Kluß C., Loges R. and Taube F. Grass and Forage Science/Organic Farming, Kiel University, Hermann-Rodewald-Strasse 9, 24118 Kiel, Germany Corresponding author: aherrmann@email.uni-kiel.de Abstract We investigated the potential of two approaches to predict the specific methane yield (SMY; lN CH4 (kg OM)-1) of perennial ryegrass. Multiple linear regression (MLR) was found inappropriate, explaining at best 50% of the variation in SMY. Near-infrared reflectance spectroscopy (NIRS) seems a more promising approach, as indicated by prediction errors of 11.7 and 15.9 (lN CH4 (kg OM)-1) and coefficients of determination of 0.82 and 0.67 for the calibration and validation, respectively. Keywords: Lolium perenne, specific methane yield, forage quality, multiple linear regression, NIRS Introduction Methane production by anaerobic digestion is discussed as an alternative for grassland not used any longer by livestock. Furthermore, ley-arable systems might reduce potential adverse environmental impact of biogas substrate production, which is often dominated by continuous maize. The evaluation of grassland methane yield, however, still requires the use of labor- and cost-intensive batch assays. The present study aimed to investigate whether SMY of perennial ryegrass can be predicted from forage quality traits or by NIRS. Materials and methods The study was based on data collected in a 2-year (2009-2010) grassland trial conducted on a heavy clay soil (40% clay, gleyic Fluvisol (calcaric), pH 7.2) close to the west coast of Schleswig-Holstein, northern Germany. Two Lolium perenne cultivars differing in heading date (Trend, 4n, mid-early; Twymax, 4n, late) were grown in a 3- and 4-cut system. Additional samples were obtained from an adjacent cropping system experiment (2009-2010), where cv. Trend was grown in a 4-cut system. In both trials, mineral N fertilizer was applied as calcium ammonium nitrate at a level of 360 kg N ha-1. The contents of N, neutral detergent fibre (NDF), acid detergent fibre (ADF), acid detergent lignin (ADL), crude fat (XL), water-soluble carbohydrates (WSC), crude ash (XA), as well as enzyme soluble organic matter (ESOM), digestibility of organic matter (DOM) and metabolizable energy content (ME) were estimated by NIRS (NIRS-System 5000 monochromator, Foss NIRSystems, USA; WinISI II, Infrasoft International, USA). Specific methane yield (lN CH4 (kg OM)-1) was determined by a batch assay. For this purpose, fresh crop samples were chopped to 2 cm length and subsequently stored at -20 °C. The samples were slowly defrosted before analysis. The batch test was conducted in two replicates using 2 L glass vessels filled with 1.9 L of inoculum and a quantity of 3 g crop substrate per litre inoculum. The inoculum consisted of a mixture (1:1) of sludge obtained from a laboratory fermenter and digested sludge from a regional sewage plant. Vessels were incubated at 38 °C and the batch assays were continued until biogas emission became negligible (< 1% per day), at least for 30 days. Each run included one sample of a pure inoculum and one microcrystalline cellulose sample. The methane amount was determined with a gas chromatograph (Varian 3800). For further data analysis, SMY was adjusted to norm conditions (273 K, 1013 mbar). In a first approach, a MLR model was applied (lme-function in the Rpackage nlme) to determine how accurate SMY can be estimated by forage quality traits. To this end, ADL, HCEL, CELL, XP, XL, XA and WSC served as predictors, by assuming year and replicate as random. Aggregating quality traits, e.g. energy concentration, were not Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 492 considered to avoid multicollinearity. In a second approach, the ability of NIRS for predicting SMY was analysed. From the pooled data sets of the grassland and cropping system experiments, a calibration subset of 110 out of 183 samples was selected to represent the whole spectral and chemical variability, using the H-value as selection criterion. The remaining 73 samples provided the validation subset. Results and discussion The SMY ranged between 283 and 422 lN CH4 (kg OM)-1. The analysis of the relationships among quality traits revealed plausible correlations among all quality traits (data not shown). Slightly negative but significant relationships were detected between SMY and N content, cell wall fractions, and crude fat content. In contrast, WSC, energy content, ESOM and DOM seem to result in increased SMY. These results are partly in contrast to previous research (Kandel et al., 2013a, b; McEniry and O’Kiely, 2013). Overall, none of the single quality traits explained the variation of SMY to a sufficiently high degree. Figure 1. Results of MLR, conducted separately for samples collected in the grassland (a) or cropping systems experiment (c). Regression functions obtained for (a) were validated with the cropping systems data set (b) and functions obtained for the cropping system experiment (c) were validated with the grassland data set (d). ADL: acid detergent lignin (% DM), CEL: cellulose (% DM), HCEL: hemicellulose (% DM), XA: crude ash (% DM), WSC: water-soluble carbohydrates (% DM), and XP: crude protein (% DM). The MLR was conducted (i) separately for the grassland experiment and the samples originating from the cropping systems trial for calibration and validation purposes, (ii) with additional differentiation by the different cuts, and (iii) for the pooled data set of both experiments. Altogether, none of the approaches led to a statistical model that could explain the variation in SMY by more than 50%. This is exemplified in Figure 1 for a separate analysis of both data subsets (grassland, cropping system). In agreement, Raju et al. (2011) and Kandel et al. (2013a) investigating SMY of grasses or plant parts did not find meaningful prediction functions explaining more than 40 to 57% of SMY variation. The difficulties associated with MLR models can be overcome when predicting SMY from NIRS spectra (Table 1). The calibration equation was acceptable, as indicated by an R2 of 0.81 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 493 and a low standard error of calibration (SEC) of 11.7 lN CH4 (kg OM)-1. When comparing the variation of SMY found in our data set with the results of the multiple linear regression approach and the precision of the batch assay (Heuwinkel et al., 2009) the calibration seems promising. Independent validation of SMY was classified as reliable, as indicated by a standard error of prediction (SEP) of 15.9 lN CH4 (kg OM)-1 and a R2 of 0.67. Similar to our study, Raju et al. (2011) and Kandel et al. (2013b) reported a better performance of NIRS than of models indirectly estimating grass SMY from forage quality traits. Apparently, SMY is affected by a large number of constituents and corresponding associations, which are reflected in NIRS spectra, but cannot be reliably quantified in simpler multiple regression approaches. Table 1. Calibration and validation statistics for the prediction of specific methane yield (l N CH4 (kg OM)-1) by near infrared reflectance spectroscopy. Range, mean and standard deviation (SD) refer to the laboratorydetermined values of the calibration and validation subsets. SEC: standard error of calibration, SECV: standard error of cross validation, SEP: standard error of prediction. N Range Mean SD Mathematical treatment 1,2,3,4 SEC SECV SEP R2 Calibration set 110 283.5-421.9 348.7 24.3 2,5,5,1 11.7 15.6 -0.81 Validation set 73 301.9-392.0 345.5 22.1 2,5,5,1 --15.9 0.67 1 number of derivate of the log (1/R) spectrum; 2 segment of the gap over which the derivative was calculated; 3 number of data points used during first smoothing of the spectrum; 4 number of data points used during second smoothing of the spectrum Conclusion NIRS seems a promising tool for predicting ryegrass SMY, as our study indicates. However, further improvement will be necessary before it can be used as a tool for screening breeding materials or for optimizing feedstock provision for biogas plants. This may, in part, be due to the fact that spectra were recorded for dried and ground samples, whereas in the batch assay, chopped, frozen samples were used. References Heuwinkel H., Aschmann A., Gerlach R. and Gronauer A. (2009) Die Genauigkeit der Gasmessung von Substraten mit der Batchmethode. In: Bavarian State Research Center for Agriculture (LfL) (ed), Proc. Biogas Science 2009, Band 1, pp 95-103. http://www.lfl.bayern.de/publikationen/daten/schriftenreihe/p_37628.pdf. Kandel T.P., Sutaryo S., Møller H.B., Jørgensen U. and Lærke P.E. (2013a) Chemical composition and methane yield of reed canary grass as influenced by harvesting time and harvest frequency. Bioresource Technology 130, 659-666. Kandel T.P., Gislum R., Jørgensen U. and Lærke P.E. (2013b) Prediction of biogas yield and its kinetics in reed canary grass using near infrared reflectance spectroscopy and chemometrics. Bioresource Technology 146, 282287. McEniry J. and O’Kiely P. (2013) Anaerobic methane production from five common grassland species at sequential stages of maturity. Bioresource Technology 127, 143-150. Raju C.S., Ward A.J., Nielsen L. and Moller H.B. (2011) Comparison of near infrared spectroscopy, neutral detergent fibre assay and in-vitro matter digestibility assay for rapid determination of the biochemical methane potential of meadow grasses. Bioresource Technology 102, 7835-7839. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 494 Phytoestrogen content in clover (Trifolium spp.) and in grass stands depending on treatment and storage Řepková J.1, Nedělník J.2, Krtková V.3, Schulzová V.3, Novotná H.3, Hajšlová J.3 and Jakešová H.4 1 Department of Experimental Biology, Masaryk University Brno, Czech Republic; 2 Agricultural Research, Ltd., Troubsko, Czech Republic; 3 Department of Food Analysis and Nutrition, Institute of Chemical Technology Prague, Czech Republic; 4 Red Clover and Grass Breeding, Hladké Životice, Czech Republic Corresponding author: repkova@sci.muni.cz Abstract Beneficial as well as negative effects of phytoestrogens (PEs) on foodstuffs and fodder of plant origin have been studied. In this work, PE levels were examined over 2 years in red clover (Trifolium pratense), zigzag clover (T. medium) and their hybrid, as well as in haylage from two locations in the Czech Republic produced from the first cutting of grass on perennial grass stands. A statistically significant difference in PE content was found between the two species of clover. Plants in the hybrid population contained a statistically inconclusive difference in PE content in comparison with T. pratense and a lesser PE content than T. medium. In high- and low-quality haylage samples, a significantly higher PE content was found in samples from the Závišice location (2011 and 2012 harvest), which may be due to the difference in individual species representations in the fodders at the given locations. Keywords: phytoestrogens, Trifolium pratense, Trifolium medium, UHPLC-MS/MS Introduction Phytoestrogens (PEs) are biologically active chemical compounds of plant origin which display effects similar to those of oestrogen sex hormones (Velíšek et al., 2009). In recent years, interest in these substances has grown with a view to their possibly beneficial but also possibly negative effects on humans and animals. The main and most significant sources of PEs are considered to be pulses (in particular soya beans (Glycine max) and fodder crops (red clover (Trifolium pratense) and alfalfa (Medicago sativa)) which are used as feedstuffs (Kuhnle et al., 2008). The PEs most represented in fodder crops are biochanin A and formononetin while those in pulses are daidzein, genistein and glycitein (Beck et al., 2005). In mammals, PEs are metabolized to products which generally display higher oestrogen activity than those original forms. The aim of this two-year study was to examine the PE profile of a new hybrid population, which was bred through interspecific hybridisation of red clover and zigzag clover (T. medium) (Jakešová et al., 2011). A further aim was to evaluate PE content in high- and low-quality haylage. Materials and methods Breeding material acquired by T. pratense x T. medium interspecific hybridization was tested for content of daidzein, genistein, formononetin and biochanin A in 2011 and 2012. The ‘Amos’ variety and T. medium were parental genotypes used for comparison. The haylages were produced from the first cutting of perennial grass stands at Lukov and Závišice in the Czech Republic during 2011 and 2012. Haylage was produced and stored in the form of plasticwrapped hay bales. Low-quality haylage was prepared by artificially damaging the protective plastic film covering the hay bales in several places with the tines of a pitchfork (simulating damage from branches) and by slicing it with a knife (simulating greater damage). Approximately once per week the damaged bale was sprinkled with 10 L of water (simulating Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 495 rainwater leaking into the damaged bale). Total PE content in low-quality haylage was examined and compared with high-quality haylage. To isolate PEs from the matrix, direct extraction (to determine free PEs) and acid-based hydrolysis (to determine total PEs) were used. To determine the PEs, an optimized and validated method of ultra-performance liquid chromatography in combination with triple quadrupole mass spectrometry (ULPC-MS/MS) was used. Separation was done on an Acquity BEH C18 (50 x 2.1 mm; 1.7 µm) analytical column. Results and discussion Differences were found in the PE profile and content among the individual tested clover materials. A statistically significant difference in PE content was found between the two species of clover (Figure 1). Figure 1. Comparison of phytoestrogen (PE) content in Trifolium pratense, T. medium and their hybrid (JEH) in the two-year experiment. * Statistically significant difference. Plants in the hybrid (JEH) population contained a statistically inconclusive difference in PE content in comparison with T. pratense and a lower PE content than did T. medium. A higher content of biochanin A was observed in T. medium than in the hybrid JEH plants, and in the latter there was, by contrast, a higher content of formononetin. Breeding material founded on T. pratense x T. medium hybrids is a rich source of genetic variability for T. pratense. That material’s distinct DNA content had been analysed by flow cytometry and cytology (Řepková et al., 2012) and was characterized by increased variability in morphological and agronomic traits (Jakešová et al., 2011). The newly cultivated plants with the JEH designation were used to breed the new variety ‘Pramedi’. The ‘Amos’ variety was used to stabilise the genomes of the hybrids by repeated open pollination with the JEH genotypes. The Czech Plant Variety Office granted rights for the new variety in 2013 (variety number TPM14855 and variety code 5082339). A comparison of total PE content in high-quality (HQH) and low-quality (LQH) haylages (each material was always analysed from the upper layer, middle layer and at the centre of the bale) showed a slightly higher PE content in HQH samples from Závišice in 2012. No significant difference was observed in PE content between HQH and LQH from Lukov in either harvest (2011 or 2012). A higher PE content was noticed in 2012 for both HQH and LQH samples in both studied areas (Figure 2). Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 496 3,0 Center of the package -1 PE (g.kg ) 2,5 The middle layer 2,0 The upper layer 1,5 1,0 0,5 0,0 2011 2012 2011 2012 2011 2012 2011 2012 LQH HQH LQH Závišice HQH Lukov locality Figure 2. Comparison of total PE content in high-quality (HQH) and low-quality (LQH) haylages from Závišice and Lukov in the two-year experiment. Comparison of PE content in individual layers of the bales demonstrates that the middle layer of the bale has the richest PE content in both HQH and LQH. Only in one instance was PE content higher in the upper layer of the bale, namely in the LQH sample from Závišice from 2012. The overall higher PE content in haylage from Závišice is primarily owing to the higher representation of red clover in the grass stand there (about 70%). Conclusions A statistically significant difference in phytoestrogen content was found between the individual species of clover. Plants in the hybrid JEH population contained a statistically inconclusive difference in phytoestrogen content in comparison with T. pratense. In high- and low-quality haylage samples, a significantly higher PE content was found in the sample from Závišice (2011 and 2012 harvests). Acknowledgements The authors thank the Ministry of Agriculture of the Czech Republic (grants QI111C016 and QI111A019) and the Ministry of Education, Youth and Sports of the Czech Republic (grants MSM 6046137305 and CZ.1.07/2.4.00/31.0155, specific research 21/2013) for financial support. References Beck V., Rohr U. and Jungbauer A. (2005) Phytoestrogens derived from red clover: An alternative to estrogen replacement therapy? Journal of Steroid Biochemistry and Molecular Biology 94, 499-518. Jakešová H., Řepková J., Hampel D., Čechová L. and Hofbauer J. (2011) Variation of morphological and agronomic traits in hybrids of Trifolium pratense x T. medium and a comparison with the parental species. Czech Journal of Genetics and Plant Breeding 47, 28-36. Kuhnle G.G. Dell’Aquila C., Aspinall S.M., Runswick S.A., Mulligan A.A. and Bingham S.A. (2008) Phytoestrogen content of foods of animal origin: Dairy products, eggs, meat, fish, and seafood. Journal of Agricultural and Food Chemistry 56, 10099-10104. Řepková J., Simandlová J., Jakešová H., Nedělník J., Soldánová M., Hajšlová J. and Schulzová V. (2012) Vztah rodičovských genomů u mezidruhových hybridů Trifolium pratense x Trifolium medium [Intergenomic relationships in interspecific hybrids of Trifolium pratense x T. medium]. Úroda scientific supplement 60, 27-30. Velíšek J and Hajšlová J. (2009) Chemie potravin [Chemistry of Foodstuffs], 3rd ed., OSSIS, Kloktoská 115/4, Tábor, 390 01, Czech Republic, 620 and 644 pp. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 497 Demand for K and P in reed canary grass (Phalaris arundinacea) during the harvest years Palmborg C., Lindvall E. and Gustavsson A.-M. Department of Agricultural Research for Northern Sweden, Swedish University of Agricultural Sciences, SE-90183 Umeå, Sweden Corresponding author: Cecilia.Palmborg@slu.se Abstract Reed canary grass (RCG) is a tall perennial grass that can be used as fuel in boilers if harvested in a delayed harvest system. To increase the crop's ability to compete with forest fuels, there is a need for cropping cost reductions. We did fertilization experiments during two harvest years at two sites in Sweden, one in the north and one in the south, to determine if P and K fertilization can be reduced and to determine the effect of recycling of RCG ash as a fertilizer. All treatments were given 100 kg N per ha. Both sites were moderately low in plant available P and K. There was no increase in RCG yield by P, K or RCG ash fertilization. However, significant differences in the P and K concentrations in the crop and in plant available P and K in the soil were observed. We conclude that even in relatively P and K poor soils, P and K fertilization of RCG can be omitted or replaced with recycling of RCG ash without any negative short-term effects on yields. Keywords: Phalaris arundinacea, yield, PK fertilization, ash fertilization Introduction Tall perennial grasses have been identified as potential second-generation energy crops. They can be used as fuel in electricity and heat production or as raw material for ethanol and gas production. A system with cutting in late autumn, storage in windrows in the field during winter, and harvest in spring after aeration of the windrows (delayed harvest) was developed in Sweden in the early 2000s (Larsson et al., 2006). The costs for growing of RCG are lower than for most other crops since the fertilization costs are low and, if the crop is successfully established, it can be sustained over 15 years (Pahkala, 2007) so for each harvest year the establishment costs are low. Despite low growing costs, however, the total production costs are still too high for RCG to compete with fuels that are by-products from forestry in Sweden. Thus the costs need to be reduced, and one way can be to improve the fertilizer recommendations for P and K and also recycle nutrients in the ash from combustion of RCG to avoid ash disposal costs. RCG has a good capacity to relocate N, P and K from the shoot to the rhizomes during late autumn and reuse it spring growth the following year (Xiong et al., 2009) and it also is well known for efficient nutrient uptake. Thus it might be possible to reduce the P and K fertilization rates after the first two years when the rhizome system is established. The aim of this study is to determine the short-term requirement (2 harvest years) of P and K for RCG in a delayed harvest system on soils with low P status. We also aim to determine if this requirement can be satisfied by fertilization with RCG ash. Materials and methods Field experiments were performed at two RCG fields at farms, selected due to low concentrations of plant available P and K in both top soil and sub soil. The site in southern Sweden was Runtorp (56° 36' N 15° 58' E) close to Kalmar. The soil is a sandy loam. The RCG was sown in June 2011 and fertilized with 20 t ha-1 biogas residue prior to sowing. The experiment started in May 2012 at the start of the first production year for the crop. The site in northern Sweden, Röbäcksmyran (63° 48' N 20° 9' E) close to Umeå, is a drained former shallow peatland with the depth of peat approximately equal to the depth of the top soil (20-30 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 498 cm). The soil is humus-rich loamy sand with sandy subsoil. This field was sown in June 2007 and was fertilized with horse manure prior to sowing. In the following years it was only fertilized with 80 kg N ha-1, not with P or K. The start of this experiment was in June 2012. The experiments had a randomized block design with four replicate blocks. Each block had seven plots (3 × 9 m) with different treatments. All treatments received a basic fertilization of 100 kg N ha-1 as Axan NS 27-4. In addition, in the first year, RCG ash (3000 kg ha-1 with 34 kg P and 104 kg K ha-1) was spread one treatment. The ash was non-hardened (i.e. not treated by moistening before storage). The other treatments were two levels of P (34 and 17 kg ha-1), two levels of K (100 and 50 kg ha-1), a control fertilized with the recommended level of P and K for forage grass (20 kg P and 130 kg K kg ha-1), and a control without P and K fertilization. The second year, the ash and P treatments received only N, while the other treatments were the same as in the first year. All fertilizations were made at the start of the growing seasons. Soil samples were taken from the topsoil (0-20 cm) and the subsoil (40-60 cm) using an auger with 3.5 mm diameter for the top soil and 2.9 mm diameter for the subsoil. Five samples from each level and block were pooled to a composite sample at the start of the experiment. After harvest 2013, all plots in the unfertilized control treatment, the ash treatment and the PK fertilized control were sampled in the same way to one composite sample per plot. Plantavailable soil nutrients were estimated by extraction with 0.01 mol l-1 ammonium lactate and 0.40 mol l-1 acetic acid (AL method, Swedish standard, SS 028310). The plots were harvested each year with Haldrup harvesters in November in the south and in October in the north. The harvested plots were 1.5 × 6 m in the middle of the plots. The dried samples were milled on a hammer mill with 1 mm sieve before chemical analysis. The mineral concentrations of the milled crop 2013 were analysed after digestion with HNO3 using an Induced Coupled Plasma. The results were analysed by balanced design ANOVA with treatment as fixed effect and block as random effect using NCSS 8 (Hintze, 2012). Each experiment was analysed separately. The soil measurements for top soil and sub soil were analysed separately. Results and discussion Yields did not differ significantly between the treatments at any of the sites. At the southern site, Runstorp, the mean yield in November was 10.0 t ha-1 in 2012 and 9.7 t ha-1 in 2013. At the northern site, Röbäcksmyran yields were lower: The mean yield in October 2012 was 6.1 t ha-1 in 2012 and 7.5 t ha-1 in 2013. The yields were similar to other Swedish studies with harvest in late autumn (Landström et al., 1996). The summer of 2012 was a comparatively rainy summer in southern Sweden and this should have been advantageous for RCG. Its growth is often more limited by water than by N availability (Kätterer and Andren, 1999). Mineral concentrations in the autumn harvest differed more between the treatments at Röbäcksmyran than in Runstorp. At both sites the full PK fertilization caused Ca and Mg concentrations to be lowered (data not shown). In the southern site, Runstorp, the P and K concentrations were low and there were no significant differences between the fertilization treatments. In the northern site, Röbäcksmyran, the full PK fertilization treatment showed higher P and K concentrations in the harvested biomass than the control plots (Figure 1a and 1b). Biomass from the southern site, harvested in November, was all brown, while biomass from the northern site, harvested in October, was still green to some extent. This was probably the reason for the lower K content in Runstorp. Any differences in the K and P content between treatments might have been lost during relocation of nutrients from the shoots to the rhizomes in late autumn in the southern site. Thus we will need to analyse samples from green biomass in the summer to determine if there are any differences in plant uptake of nutrients between the treatments. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 499 b a Röbäcksmyran Runtorp g K kg DW-1 g P kg DW-1 Röbäcksmyran 1.3 1.2 1.1 1 0.9 0.8 0.7 0 Runtorp 10.0 9.0 8.0 7.0 6.0 5.0 4.0 3.0 20 40 -1 added kg P ha 0 100 200 300 added kg K ha-1 Figure 1. P and K concentrations in harvested RCG biomass in late autumn after 2 experimental years Plant available nutrients in soil also showed differences between the treatments in the topsoil, but not in the subsoil (Table 1 subsoil not shown). KAL in the topsoil from both sites after full PK fertilization was higher than the unfertilized control and also higher than at the start of the experiment. PAL in both the ash fertilization and the full PK fertilization treatments were higher than the control in the northern site, Röbäcksmyran. It also had increased since the start in the fertilized treatments in Röbäcksmyran and in all treatments in Runtorp. Table 1. Plant available nutrients (Ammonium lactate extraction) and pH in topsoil (0-20 cm) at experiment start (2012) and after two growing seasons (2013). Means±SE for four replicate samples per block (*= significant P<0.05 difference to the control) Site Year Treatment CaAL KAL MgAL PAL pH mg kg* soil-1 Röbäcksmyran Runtorp 2012 Before start 998 ±36 64 ±4 61 ±5 28 ±2 5.4 ±0.1 2013 RCG Ash + N 838 ±51 43 ±3 46 ±5 35 ±2 * 5.5 ±0.1 2013 NPK fertilization 833 ±32 88 ±12 * 41 ±2 36 ±1 * 5.4 ±0 2013 N fertilized control 755 ±52 34 ±4 35 ±4 31 ±1 5.4 ±0 2012 Before start 1223 ±47 88 ±7 46 ±4 37 ±2 6.4 ±0 2013 RCG Ash + N 1348 ±91 92 ±2 44 ±5 * 51 ±5 6.3 ±0.1 2013 NPK fertilization 1328 ±90 101 ±2 * 41 ±4 50 ±4 6.2 ±0 2013 N fertilized control 1303 ±86 71 ±2 37 ±5 48 ±5 6.3 ±0.1 MgAL after RCG ash fertilization was higher than the control in the southern site, Runtorp. The increases in plant available P and K after full PK fertilization show that the PK demand of delayed-harvest reed canary grass is considerably lower than for forage grasses. Also, the lack of significant yield increases after any of the NPK-fertilized treatments compared to the Nfertilized control, show that the short term demand for P and K in RCG can be satisfied with soil uptake on these loamy soils. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 500 Conclusions Reed canary grass in a delayed harvest system has a lower demand for P and K than forage grasses. In soils with moderately low P and K availability (Class II in the Swedish system), the short term (two years) demand during the harvest years can be covered by soil supply. However, in the long term, P and K in soil can be depleted and recycling of reed canary grass ash to the fields can be one way to reduce that depletion. Acknowledgement This research was financed by the Swedish Energy Board. References Hintze J. (2012) NCSS 8, NCSS, LLC, Kaysville, Utah USA. Kätterer T. and Andren O. (1999) Growth dynamics of reed canarygrass (Phalaris arundinacea L.) and its allocation of biomass and nitrogen below ground in a field receiving daily irrigation and fertilisation. Nutrient Cycling in Agroecosystems 54, 21-29. Landström S., Lomakka L. and Anderson S. (1996) Harvest in spring improves yield and quality of reed canary grass as a bioenergy crop. Biomass & Bioenergy 11, 333-341. Larsson S., Örberg H., Kalén G. and Thyrel M. (2006) Rörflen som energigröda. BTK-rapport. Umeå, BTK: 44. Pahkala K. (2007) Reed canary grass cultivation for large scale energy production in Finland. In (ed.) NJF Seminar 405 Production and utilization of crops for energy. (Vilnius, Lithuania), 52-55. Xiong S., Landstrom S. and Olsson R. (2009) Delayed harvest of reed canary grass translocates more nutrients in rhizomes. Acta Agricultura Scandinavica. Section B, Plant Soil Science 59, 306-316. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 501 Organic seed production of yellow oat grass – preliminary results Macháč R. OSEVA Development and Research Ltd., Zubří, Czech Republic. Corresponding author: machac@oseva.cz Abstract Seed production of mixed crops of yellow oat-grass (Trisetum flavescens L. P. Beauv.) and two annual legumes was compared in an organic cropping system with or without using organic manure. The best seed yield was achieved in the cropping system with organic manure, treated by weed harrowing and using black medic as the companion crop (mean of two seed harvest years - 203 kg/ha). This yield is comparable to the yield on conventional control. In a comparison of the two companion legumes, black medic was slightly better than clustered birdsfoot-trefoil. However, the greatest influence on seed yield was from fertilization. Keywords: Trisetum flavescens, seed, organic farming Introduction Yellow oat grass (Trisetum flavescens L. P. Beauv.) is a valuable forage grass species for organic grassland systems. It is able to produce satisfactory yields of high-quality forage under organic conditions. Yellow oat grass is preferred by animals on pastures, and also gives fine hay with good content of digestible nutrients. According to the Council Regulation (EEC), seed and vegetative plant material for organic farming should be organic, produced without use of inorganic fertilizers or pesticides. However, sufficient nutrition, mainly of nitrogen, is necessary for achieving favourable seed yield of grasses. In Denmark and Norway, there were attempts to grow grass seed together with legumes (Deleuran and Boelt, 2000; Solberg et al., 2007), predominantly using red and white clovers. However, based on Czech findings (Macháč and Cagaš, 2005) red clover and white clover often suppress the grasses, and so they recommended using black medic as a companion legume. This study was designed to identify which legume species and type of treatment are the best for growing yellow oat-grass for seed on organic farms. Materials and methods In 2012-13, a small-plot field trial of organic seed production of yellow oat-grass was conducted at the Grassland Research Station at Zubri. There was no use of any inorganic fertilizers or pesticides on the trial area for the three years before sowing the trial. Yellow oat-grass was undersown in 2010 into spring wheat. In the trial, three experimental factors were combined: (1) companion legume (none, black medic, clustered birdsfoot-trefoil); (2) method of nutrition (conventional control, only by nitrogen using legumes, organic manure); (3) treatment (none, weedy-harrowing – once or twice a year). The sowing rate of yellow oat-grass cv. Roznovsky was 11 kg/ha. Plot size was 10 m2, with 5 rows with 25 cm row spacing. Legumes were sown across the rows of yellow oat-grass. The sowing rate of black medic (Medicago lupulina L.) cv. Ekola was 15 kg/ha and clustered birdsfoot-trefoil (Lotus ornithopodioides L.) cv. Junak was 12 kg/ha. On the conventional control inorganic fertilizers were applied (70 kg/ha of nitrogen, 50 kg/ha P+K), and on the plots with organic manure slurry was applied (supplying 70 kg/ha of nitrogen). Applications of both types of fertilization were made in the first half of April. Weedharrowing was carried out using a plot weeder-harrow (width 1.5 m, work speed 12 km/hod). In first ley year the mixtures were harvested for forage. Direct combining (plot combine Wintersteiger Elite) of the yellow oat-grass for seed was carried out in the second a third ley years. Harvested seed was cleaned on laboratory screen machines in order to meet purity standards. The data were processed by ANOVA. To evaluate the significance of the differences between means, Tukey test with 5% significance level was used. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 502 Results and discussion In the first seed-harvest year (2012) the highest seed yield (273 kg/ha) was achieved from the variant with black medic as the companion crop, with organic manure and a single treatment with the weeder-harrow. In contrast, the lowest seed yield (179 kg/ha) was determined at variants without fertilization, with clustered birdsfoot-trefoil as the companion legume and a single treatment by the weeder-harrow. Generally, the seed yield of yellow oat-grass was in very high in the first harvest year and exceeded the average yield in the Czech Republic. In a comparison of experimental factors, statistically significant differences were found for type of treatment, when variants without weed-harrow treatment significantly overyielded the variants treated by the weeder-harrow. Major differences were to be observed for method of fertilization. The highest seed yield was achieved on variants with organic fertilization, which overyielded the conventional control as well as variants without fertilization (Table 1). Table 1. Effect of trial factors on seed yield (kg/ha) of yellow oat-grass. Trial Factor companion legume none clustered trefoil black medic fertilization Yield 225 Tukey rel. (%) a 100 Second seed year Yield Tukey rel. (%) 142 a 100 127 b 89 128 0.002 b 90 100 122 c 100 b 85 128 bc 105 ab 94 ab 110 conventional ANOVA 0.563 225 134 0.876 b 100 142 a 100 only legumes 171 c 76 113 b 79 organic manure 262 a 116 142 <0.001 a 100 birdsfoot- 201 a 89 217 a 97 ANOVA 0.404 233 a harrowing 1x 199 harrowing 2x 220 none treatment First seed year Variant ANOVA <0.001 In second harvest year, the highest seed yield (150 kg/ha) was achieved from the conventional control. In contrast, the lowest yield (94 kg/ha) was observed in variants without harrowing by the weeder-harrow. Significant differences were found for all the trial factors. For type of treatment, all the variants with harrowing overyielded (in contrast to results in previous year) the untreated variants. The highest seed yield was achieved from the variant with double weeder-harrowing. These variants significantly overyielded the variants without harrowing. It is possible to explain that weed-harrowing has a positive effect not only on weed removal of the seed crops, but also on removal of waste material and aeration of the sward, which have more relevance in older swards. In the comparison of types of fertilization, significant differences were found among variants with organic or inorganic fertilization and unmanured variants. The highest yields were observed for the conventional control, organically manured plots gave yields only slightly lower, but variants without fertilization gave yields about 21% lower. Conclusion Results of two harvest years brought new findings and information about the possibilities for organic seed production of yellow oat-grass. The unsubstitutable role of nutrition of the grass stand for the achievement satisfactory seed production was confirmed. Organic manure is able to be compensate for conventional fertilization. In the comparison of companion legumes, the Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 503 use of black medic was better than clustered birdsfoot-trefoil. Determination of common conclusions will be possible after third, final cycle of field trial in the year 2014. References Deleuran L. C. and Boelt B. (2000) Utilization of forage cuts in organic grass seed production. Grassland Science in Europe 5, 552–554. Macháč R. and Cagaš B. (2005) Black Medic – a beneficial companion crop for use in organic grass production. In: F.P O'Mara et al. (eds.) Proceedings of the XX International Grassland Congress: Offered papers, UCD Dublin, Ireland, 26 June – 2 July, p. 393. Wageningen: Wageningen Academic Press. Solberg S.O., Bysveen K. and Aamild T.S. (2007) Subterranean clover (Trifolium subterraneum) as a companion crop during establishment of organic seed crops of timothy. Proceedings of the Sixth International Herbage Seed Conference. Gjenestad, Norway. Bioforsk Fokus 2(12), 165–167. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 504 Theme 4 ‘Livestock production’ Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 505 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 506 Theme 4 invited papers Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 507 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 508 Quality and authenticity of grassland products Moloney A.P.1, Monahan F.J.2 and Schmidt O.2 1 Teagasc, Animal and Grassland Research and Innovation Centre, Dunsany, Co. Meath, Ireland, 2 School of Agriculture, Food Science and Veterinary Medicine, University College Dublin, Belfield, Dublin 4, Ireland. Corresponding author: aidan.moloney@teagasc.ie Abstract Grassland is an important component of the sustainability of ruminant production in many parts of Europe and products from animals on grassland are increasingly valued by the consumer. This paper considers the fatty acid composition, shelf-life and sensory characteristics of meat and milk products of grassland. Animal-based products of grassland have ‘added value’ among both food producers and consumers because of their perceived healthiness and environmental acceptability. This added-value carries with it an onus to be able to trace and authenticate the food products derived from grassland. A range of techniques has been used to gather data with the potential to discriminate between food products of grassland production and other production systems. Chromatographic and spectroscopic methods, along with mass spectrometry, have been widely used to quantify fatty acids, volatile compounds, carotenoids, tocopherols and stable isotope ratios, and to obtain fingerprint data capable, following multivariate statistical analysis, of discriminating between production systems. Among the challenges to the discrimination process and ultimately to the authentication and traceability of grassland products are the seasonal and geographic variation in the composition of grassland feedstuffs consumed by animals and the difficulty of detecting the consumption of non-grass feedstuffs in a grassland production system. Keywords: grassland, meat, milk, quality, authentication, traceability Introduction Grazed pasture is often a cost-effective feed option for producers. In recent years, pasture-based systems have come to be regarded as more environmentally and animal-welfare friendly alternatives to intensive/feedlot systems of production. There is also growing interest by consumers in the safety, healthiness and quality of their food, its origin, and in the methods by which it is produced. For this review, grassland products are defined primarily as meat and dairy products from fresh pasture. However, while animals may graze predominantly pasture at certain times of the year, e.g. in late spring and in summer, they consume non-grass feeds or conserved forage at other times. Products of conserved forage will therefore be mentioned as relevant to specific issues under discussion. The influence of forage feeding on the fatty acid composition, processing characteristics and sensory quality of milk and meat has been regularly reviewed (e.g. Coulon and Priolo, 2002; Scollan et al., 2005, 2014; Dewhurst et al., 2006). The focus with respect to quality of grassland products will be on recent information on this topic. The Codex Alimentarius Commission defines traceability as 'the ability to follow the movement of a food through specified stages of production, processing and distribution' (WHO/FAO, 2007). Thus, traceability requires a record of the various steps in the journey of a food from its site of production to consumption and 'each link requires keeping records of preceding and succeeding links' (TRACE, 2010). Because traceability systems depend on the maintenance of records, paper or computer based, they are open to error. Authentication, defined as 'the process by which a food is verified as complying with its label description' (Dennis, 1998), is therefore necessary to support traceability systems and to Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 509 prove beyond doubt that a particular food product is as is stated on the product label. This paper will focus on the authentication methodologies that underpin such an 'authenticity based traceability system' (TRACE, 2010) for grassland products; i.e. it will focus on authentication of grassland products, and not on traceability per se, recognizing that authentication is an essential validation tool for any traceability system. Quality of meat from grassland It is recommended that total fat, saturated fatty acids (SFA), omega-6 polyunsaturated fatty acids (PUFA), omega-3 PUFA and trans fatty acids should contribute <15-30%, <10%, <5-8%, <12% and <1% of total energy intake in human diets, respectively (WHO, 2003). Meat, fish, fish oils and eggs are important sources of omega-3 PUFA for humans, while beef and other ruminant products such as milk are dietary sources of conjugated linoleic acid (CLA; Scollan et al., 2014). The dominant CLA in pasture-fed beef is the cis-9, trans-11 isomer, which has being identified as processing a range of health-promoting biological properties including antitumoral and anticarcinogenic activities (De la Torrre et al., 2006). The influence of dietary forage on the fatty acid composition of beef has been recently reviewed (Daley et al., 2010; Morgan et al., 2012, Shingfield et al., 2013, Scollan et al., 2014). The findings of the large number of studies now available are generally consistent. Thus, feeding fresh grass compared to concentrates, results in higher concentrations of n-3 PUFA and CLA in muscle lipids. In addition, feeding forage compared to concentrates during the finishing period is frequently associated with a decrease in the concentration of SFA and an increase in the concentration of monounsaturated fatty acids in muscle (Shingfield et al., 2013). With regard to the type of forage, the fatty acid composition of muscle from cattle that grazed alfalfa, pearl millet or a mixed pasture of bluegrass, orchardgrass, tall fescue and white clover before slaughter was largely similar but the concentration of linolenic acid (18:3 n-3) was highest for steers grazing alfalfa (Duckett et al., 2013). In contrast, Dierking et al. (2010) observed no difference in the fatty acid composition of muscle from steers that grazed tall fescue, tall fescue-red clover rich pasture or alfalfa before slaughter. Similarly, Schmidt et al. (2013) observed no difference in the fatty acid composition of muscle from steers that grazed bermuda grass or alfalfa before slaughter, whereas Chiofalo et al. (2010) reported a higher 18:3 n-3 in muscle from lambs that grazed subterranean clover compared to those that grazed Italian ryegrass. There is increasing interest in cattle production from botanically diverse pastures but there is a paucity of information on the fatty acid composition of beef produced from such pastures. A general tendency for an increase in n-3 PUFA and total PUFA in intramuscular fat is observed. For a comprehensive review of this topic the reader is referred to Lourenço et al. (2008) and Moloney et al. (2008). The bright red colour of beef and lamb is important for consumers when making purchasing decisions. Priolo et al. (2001) concluded from their review of 35 studies that muscle from grazing cattle is darker. In contrast, Muir et al. (1998) considered 15 publications where beef from cattle fed forage-based diets was compared with beef from cattle fed cereal-based diets but, in which, animals had been finished to comparable slaughter weights or fat-cover levels. They concluded that there were no consistent effects of these feeds on fresh meat colour. Forage feeding leads to beef carcasses with more yellow fat than those from grain-fed animals, through the ingestion, absorption and deposition of carotenoids (mainly β-carotene and lutein). Since n-3 PUFA are more susceptible to oxidation than n-6 PUFA (Mahecha et al., 2010), grassfed beef should be more susceptible to oxidation which is considered the major cause of meat quality deterioration affecting colour, flavour, and nutritional value (Li and Liu, 2012). However, a consistent finding in the literature is that green forages contribute antioxidants such as α-tocopherol and β-carotene to the meat, which stabilize the fatty acids, increase self-life and Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 510 make the meat more desirable when compared with concentrate feeding (Descalzo and Sancho, 2008). Tenderness is the most important sensory influence on the acceptability of meat, but when tenderness is increased, flavour and juiciness increase in relative importance. When possible confounding influences are removed such as comparing animals at similar weights or fat cover there is little evidence for a consistent difference in tenderness or juiciness between grass-fed and grain-fed beef (Muir et al., 1998). The flavour of red meats which develops during cooking derives from the Maillard reaction between amino acids and reducing sugars and the thermal degradation of lipid. Diets which alter the fatty acid composition of the lipid fraction of meat could also alter the amount and type of volatiles produced and hence its aroma and flavour. In a comparison of beef from grass- and concentrate-finished animals, the concentrate-fed beef had higher concentrations of linoleic acid (18: 2 n-6) and on cooking produced seven compounds at over three times the level found in the grass-fed beef, which had much higher concentrations of 18:3 n-3 and produced a higher amount of only one compound, 1-phytene, a derivative of chlorophyll ingested with grass (Elmore et al., 2004). Sensory and flavour volatile analysis by Raes et al. (2003) showed that grass-fed animals had higher flavour intensity with higher contents of low molecular weight unsaturated aldehydes derived from oxidation of long chain PUFA than grain-fed animals. Vasta and Priolo (2006) reviewed the impact of diet on meat volatiles in ruminants. Animals grazing grass had more of the flavour usually associated with beef meat than the grain-fed cattle (Richardson et al., 2004) whereas in the studies of Tansawat et al. (2013), pasture feeding produced more 'barney, greasy and gamey' flavour than grain-fed beef. Scaglia et al. (2012) found no difference in flavour in muscle from cattle finished on alfalfa or tall fescue pastures while in the study of Schmidt et al. (2013), a greater proportion of consumers preferred beef from cattle that grazed alfalfa compared to beef from cattle that grazed bermudagrass. Other compounds, not derived from fatty acids, may also contribute to flavour. Skatole has been found in high concentrations in the fat of cattle and sheep that had been fed on grass (Young et al., 2003). Schreurs et al. (2008) suggest that the differences in 'pastoral' flavour observed with different forages may reflect their varying propensity to form skatole during rumen fermentation. They suggest a role of condensed tannins-rich plants in ameliorating the development of undesirable flavours in fat of animals consuming white clover or lucerne. The effect of forage type on flavour appears to be more intensive in lamb and is better documented than for beef. In general, meat from sheep that consumed white clover or lucerne alone has been reported to give a more intense, unacceptable 'sharp' and 'sickly' flavour than meat from grass-fed sheep. In more recent studies, different varieties of legume have been examined in the context of the sensory characteristics of lamb (see Schreurs et al. (2008) and Phelan et al. (2014). Increasing the n-3 PUFA intake of beef animals through feeding oilseeds can improve the concentration and proportions of n-3 PUFA in the meat but can also produce an oxidative (and sensory) challenge during retail display. This has been rectified by feeding supplementary αtocopherol acetate with concentrate diets (see Scollan et al., 2014). Quality of milk and milk products from grassland The influence of dietary forage on the fatty acid composition of milk has been reviewed (e.g. Dewhurst et al., 2006; Lourenco et al., 2008). It is well established that the proportions of CLA and n-3 PUFA in milk increase as grass replaces non-grass constituents in the diet of ruminants (Couvreur et al., 2006; Dewhurst et al., 2006, Shingfield et al., 2013). In ewes, Ostrovsky et al. (2009) found three times more CLA and twice as much 18:3 n-3 in the milk fat of grazing ewes compared to that of ewes fed a total mixed ration. The CLA concentration decreased two fold when the 18:3 n-3 content in pasture herbage decreased in mid-summer compared to May and Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 511 September. Within a grassland production system, milk fat composition varies with stage of maturity of the grass and its botanical composition and between years for similar animals (Butler et al., 2011). Several studies have shown elevated levels of 18:3 n-3 in milk and dairy products from cattle grazing Alpine pastures (Dewhurst et al., 2006). The Alpine meadows include legumes that are rich in condensed tannins, as well as red clover, which may have reduced rumen biohydrogenation and thereby increased the levels of 18:3 n-3 in milk. From a statistical analysis of available data, Lourenco et al. (2008) concluded that milk from cows grazing botanically diverse pasture had increased milk 18:3 n-3 and PUFA concentrations, whereas milk SFA concentrations were in most cases decreased. Kalber et al. (2011) reported that berseem clover inclusion in the ration of cows also increases the proportion of 18:3 n-3 in milk, while Larsen et al. (2010) found a higher 18:3 n-3 concentration in milk from cows that grazed pastures of perennial ryegrass mixed with white clover, compared to milk from cows that grazed pastures of perennial ryegrass mixed with red clover or lucerne, which did not differ. Cows grazing birdsfoot trefoil (Turner et al., 2005) and sheep grazing Sulla or burr medic (Addis et al., 2005) produced milk containing higher levels of 18:3 n-3 than cows grazing nontanniniferous herbages or sheep grazing other grasses. It appears that this effect is related to reduced biohydrogenation of 18:3 n-3 in the rumen as a consequence of the action of tannins. Petersen et al. (2011) noted an increase in 18:3 n-3 when cows were zero grazed a mixture of sown herbs (chicory, plantain, birdsfoot trefoil, white melilot and others) in comparison with zero-grazing white clover-perennial ryegrass. It is not clear which species and mechanisms were involved in this effect. Dewhurst et al. (2006) reviewed the effects of clover silages on fatty acids in milk. In comparison with milk from cows fed grass silages, both red clover and white clover silages led to highly significant increases in the proportion of 18:3 n-3. Van Dorland et al. (2008) included red clover silage or white clover silage at 40% of forage dry matter and increased the 18:3 n-3 proportion from 0.9% of milk fatty acids (grass silage control) to 1.04% (red clover silage) and 1.14% (white clover silage). In a meta-analysis of 8 published studies, Steinshamn (2010) statistically confirmed the above findings and reported an average increase in milk 18:3 n-3 proportion from 0.53 to 0.91% due to feeding red clover silage compared to grass clover silage. They found no statistical difference between white clover and red clover silage. As with meat, the fatty acid composition of milk can also influence the shelf-life and processing characteristics whereby milk with a high PUFA concentration is more susceptible to oxidation and therefore has a shorter shelf-life. Thus, milk from cows fed on red clover silage compared to grass silage contained more 18:2 n-6 and 18:3 n-3 which resulted in increased oxidative deterioration of milk (Al-Mabruk et al., 2004). The latter could be corrected by including supplemental vitamin E, an anti-oxidant, in the rations of the cows. Similarly, Havemose et al. (2006) reported a lower concentration of lipid hydroperoxides in milk from cows fed hay compared to those fed grass-clover silage, which reflected the higher 18: 3 n-3 concentration in the latter milk. In contrast, Adler et al. (2013) observed a lower concentration of hydroperoxides in milk from cows that grazed a grass-red clover pasture compared to a grass-white clover pasture despite no difference in 18:3 n-3 concentration. The authors suggested that this reflected differences in the concentration of vitamin E in the milk. The sensory quality of dairy products can be influenced by animal diet (reviewed by Coulon et al., 2004; Martin et al., 2005a). The fatty acid composition of milk may play a role in flavour; e.g., oxidized milk and milk products are characterized by metallic, cardboard or stale flavours and production of oxidized flavour at 8 days post-sampling was positively correlated with levels of 18:2n-6, 18:3n-3 and total PUFA in milk fat (Timmons et al., 2001). Croissant et al. (2007) reported that milk from cows fed a conventional ration was characterized by 'a sweet feed/malty flavour, a greater sweet aromatic flavor, and a sweet taste, but no grassy or mothball flavours' compared to pasture-based milk. Moorby et al. (2009) reported that milk from cows fed red Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 512 clover silage was rated 'more boiled' and was whiter and thinner textured compared to milk from cows fed grass silage. Larsen et al. (2013) reported that replacement of maize silage with lucerne silage resulted in milk that was more yellow and had 'less creamy flavour and less stale aroma'. Couvreur et al. (2006) observed that rancidity of butter (flavour and odour) decreased when the proportion of fresh grass in the diet increased, which the authors ascribed to a decrease in the proportion of 14:0 in milk fat. The relationship between the nature of dietary forage and the sensory characteristics of milk and milk products has been examined in France in the context of dairy products with a protected designation of origin (PDO) and their 'terroir'. Martin et al. (2005a) in their review indicate that differences in the sensory characteristics of cheese made from milk produced in the valleys or in the mountains are related in part to the presence of legumes in the pastures grazed by the cows (see also Moloney et al., 2008). Claps et al. (2009) reported that cheese made from goats that grazed Medicago sativa had less intense taste but was similar in acceptability to cheese made from milk from goats that grazed Lolium perenne. As with meat, other dietary compounds may also contribute to flavour. Terpenes have aromatic properties and abound in certain aromatic dicotyledons species found in diversified meadows (Mariaca et al., 1997). These molecules are found in higher concentrations in cheese when the animals are fed dicotyledon-rich natural grass forage (Viallon et al., 2000). However, it appears that the increase in terpene concentration in cheese is generally not sufficiently large to exert any marked effect on flavour (Bugaud et al., 2001). Food authentication techniques Strategies to authenticate grassland products have focussed firstly on the measurement of components that directly reflect the diets consumed (see above). Secondly, a ‘fingerprint’ approach can be taken whereby spectroscopic techniques are used to determine differences in the optical properties of foods derived from different production systems. Transcriptomic and proteomic techniques are now being also being evaluated in this regard (Hocquette et al., 2009; Shibata et al., 2009). Several recent reviews describe the methodologies now available for application to authentication of meat and milk (e.g Karoui and Baerdemaeker, 2007; Luykx and Van Ruth, 2008; Sun, 2008; Monahan et al., 2010, Capuano et al., 2013). Fatty acid data In a recent study (Roehrle, 2014), a discrimination model of muscle fatty acid data permitted differentiation of beef from animals raised on grass, a barley-based concentrate or on grassconcentrate combinations over a 12-month period with a correct classification of 92.9%. The mis-classified samples related to beef from animals raised on pasture for 12 months prior to slaughter being classified as 'beef from animals fed grass silage for 6 months followed by grass at pasture for 6 months.' Effectively, however, both groups could be considered grass-fed, since the silage is ensiled grass; therefore the mis-classification is not of major significance from the perspective of grass-fed beef authentication, and pooling these groups together gave 100%correct classification of beef according to diet. A similar approach has been used to distinguish organic milk from conventional milk (Molkentin and Giesemann, 2007) and upland from lowland milk (Engel et al., 2007). Povola et al. (2012) using a similar approach showed promising separation of cheese made from milk of cows that grazed Trifolium alpinum or Festuca nigrescens. Vetter and Schroeder (2010) attributed higher levels of phytanic acid, and its degradation product pristanic acid, in organic dairy products compared to conventionally-produced dairy products, to the predominant use of grass-based feedstuffs in organic production. These authors set a target value of at least 200 mg phytanic acid/100 g lipid for the verification of grass-fed, organic dairy products. However, this assumes that all conventional production is 'less' grassbased and uses diets that are sufficiently different from the organic diets. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 513 A potential limitation to the use of fatty acid profile as an indicator of grassland production is that non-grass sources of fatty acids could give 18:3 n-3, 18:2 n-6 or CLA contents in meat and milk similar to those derived from grass (Shingfield et al., 2013). Volatile compounds Among the meat volatile components influenced by diet are branched chain fatty acids, lactones, aldehydes, phenolic compounds, indoles, 2,3-octanedione, terpenes and sulphur compounds (Vasta and Priolo, 2006). Some compounds are directly incorporated into tissues from the dietary components while others are generated during cooking of meat fat (see above). In a study of beef from animals raised at pasture, on concentrates or on grass silage-pastureconcentrate combinations, skatole, 3-undecanone, cuminic alcohol and 2 methyl-1-butanol were identified as compounds that allowed discrimination between beef from animals animals fed pasture or concentrates or combinations thereof (Vasta et al., 2011). Germacrene D, a terpenoid, was a marker of grass feeding. Priolo et al. (2004) identified four terpenes in ovine fat, from a total of 33 terpenes detected, which permitted discrimination of lamb from sheep raised and finished on pasture from that of sheep raised on concentrate or concentrate-pasture combinations. Serrano et al. (2011) reported complete separation of suckling bulls supplemented with hay, cut green herbage or grazed pasture by measuring 2,3-octanedione, skatole and terpenes in adipose tissue. The terpenoid content and profile of pasture herbage depends on plant species, stage of growth and grazing management (Mariaca et al., 1997; Tornambé et al., 2006) so it is therefore not easy to conclude that terpenes generally, or indeed specific terpenes, are higher or lower in one production system compared to another. In milk and cheese, Martin et al. (2005b) reviewed a number of studies on the discrimination of milk and cheese from cows fed different diets. Analysis of terpenes was used to discriminate between milk from two regions of France in both summer and winter with geographical discrimination attributed to botanical differences in the forages. Terpene transfer from forages to milk was shown to be fast, as early as the first milking after consumption, and terpenes were transferred into cheese with minor alteration. Stable isotopes analysis Camin et al. (2007) demonstrated the potential for C, N, H and S analysis to discriminate between lamb sourced in different parts of Europe. The influence of grassland production was clearly evident, with lamb samples from island sources on the western seaboard of Europe (Ireland and the Orkney islands) clustering separately, mainly on the basis of lower 13C/12C ratios, from samples originating in mainland Europe, where supplemental cereal or maize-based inputs may be fed for periods of the year. Bahar et al. (2008) found differences in the stable isotope ratios (13C/12C, 15N/14N) of organic and conventional beef sourced from retail outlets. The extent of differences was seasonal with pronounced differences in stable isotope ratios between the two beef sources in the months after winter feeding and similar stable isotope ratios after summer grazing. However, recent studies have shown that it is possible to discriminate between meat from animals fed different C3 plant sources despite the relatively close 13C/12C ratios of the diets (e.g. Osorio et al., 2011) Rossmann et al. (2000) highlighted the potential of the stable isotope technique (C, N, O and S) when combined with measurements of other markers including fatty acids, carotenoids and trace elements for the discrimination of butter from different regions. In cow’s milk, Renou et al. (2004) demonstrated differences in 18O/16O between milk of animals raised in two regions of France (Brittany vs. the Massif Central). They also showed differences between milk in the Massif Central produced in spring from pasture, in winter from grass silage and in winter from hay, but in Brittany there was no difference between milk produced in winter from maize silage and in winter from hay. While differences between sites may reflect differences in the drinking water 18O/16O the contribution of water from feedstuffs must be considered, which can either Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 514 eliminate or accentuate inter-site differences. Bontempo et al. (2012) concluded that stable isotope ratios of H, C, N and O of milk and cheese are linked to 'the terroir, in particular to the type of vegetation and the environment'. Optical properties and carotenoids Prache and co-workers have examined the application of reflectance spectroscopy in the visible region (450-510 nm) to discriminate between lamb production systems (reviewed by Prache, 2009). They advocate measurement of carotenoids in both adipose tissue and blood to lower the likelihood of mis-classification. Issues that may limit the usefulness of this approach include the relatively slow rate of change of carotenoids in tissue following a change in diet and the variation in carotenoid concentration in pasture. In addition, since depletion of carotenoids in adipose tissue, after a change to a low carotenoid diet, occurs due to a dilution of existing adipose tissue by new adipose tissue; in mature animals carotenoid measurement in adipose tissue may not be an appropriate indicator of diet. In agreement with the work of Prache and co-workers, (Röhrle et al., 2010a) showed contrasting reflectance spectra (400-700 nm) for subcutaneous adipose tissue from animals fed pasture (P) vs. a barley-based concentrate (C) for a 12-month period. Furthermore, subcutaneous adipose tissue from a group fed silage for 6 months followed by pasture for 6 months (SiP) was distinguishable at slaughter from that of the group that fed on pasture for 12 months, indicating an effect of a diet consumed 6 months earlier on adipose tissue reflectance at slaughter. However, a group fed silage for 6 months followed by a 50:50 (DM basis) pasture-concentrate mixture for 6 months (SiPC) was not distinguishable from the SiP group, undermining somewhat this methodology as a means of diet discrimination. Noziere et al. (2006) demonstrated the complexities associated with attempting to relate carotenoid content of milk to production system. While recognizing that the carotenoid content of fresh grasses is higher than that of conserved forage and concentrates, seasonal variation in fresh grass carotenoids affected by the stage of growth, is also a contributory factor. Practical factors such as the pooling of milk and its bulk storage prior to processing pose a significant challenge in terms of authentication of milk or the processed dairy products derived from it. Reflectance measurements have also been applied to milk in an attempt to distinguish between grass vs. hay and concentrate feeding (in individual cows) – this was possible provided there was a least a 36-day interval between time of diet switch from the low carotenoid (concentrate, hay) to the high concentrate (pasture) diet (Noziere et al., 2006). Near and mid infra-red spectroscopy (NIRS and MIRS, respectively) has also shown promise as a tool for authenticating the products of grassland. Dian et al. (2008) examined the ability of spectroscopy between 400 and 2500 nm and reported that the longer wavelength range produced models of higher discriminant ability producing correct classification rates of 97.7%. Valenti et al. (2013) concluded that MIRS was better than NIRS when discriminating between milk from cows offered hay or pasture-based diets. Gori et al. (2012) further showed that Fourier transform infra-red spectroscopy coupled with artificial neural networks could distinguish the dietary regime of butters produced in the Parmigiano Reggiano cheese region in Italy. Vitamin E stereoisomers Analysis of stereoisomeric forms of -tocopherol in animal tissues can give information about whether animals received vitamin E from natural or synthetic sources. In muscle from grassfed beef cattle the RRR stereoisomer dominated, while in concentrate-fed animals and in beef of unknown dietary background stereoisomers of synthetic origin were evident (Röhrle et al., 2010b). Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 515 Functional genomics Cassar-Malek et al. (2009), in a comparison of outdoor pasture vs. indoor concentrate feeding of Charolais cattle, found Selenoprotein W to be under-expressed in pasture-fed animals and proposed it as a putative gene marker of the grassland system. Duckett et al. (2009) studied expression of genes involved in lipogenesis in muscle and found up-regulation of stearoyl-CoA desaturase, fatty acid synthase and Spot-14 and down regulation of signal transducer and activator of transcription-5 (STAT5) in the subcutaneous fat of grazing steers finished on a high-concentrate diet compared with a pasture only diet. Shibata et al. (2009) showed that differential expression of muscle proteins occurred during the fattening period in concentratefed vs grazed cattle. To obtain useful information from the data collected by fingerprint methods or where multiple variables are measured, multivariate statistical procedures, often termed chemometrics, are required, e.g. Karoui and De Baerdemaeker (2007) for dairy products. Chemometric methods may be applied to complete datasets or after a variable reduction procedure has been applied; in the case of spectral data in particular, the raw data may also be pre-treated mathematically to reduce or remove interferences caused by physical factors related to the sample. Conclusions Modern consumers have a greater interest in the environment, animal welfare and the origin and method of production of their food than heretofore. This is reflected in growing preference for food products of pasture-based systems of production. A substantial body of information is now available on the differences in composition and sensory properties of products from pasture-based and concentrate based systems of production. For meat, current research seems to be focussed on the relative effects of pasture species and varieties. For milk, there appears to be less focus on grassland in this regard, presumably because, in general, grassland represents a smaller proportion of diet of the lactating compared to the growing animal. Emerging data indicate that milk and meat produced from botanically diverse pastures have higher concentrations of fatty acids and anti-oxidants which are considered to be of benefit to human health. Much progress has been made in recent years in advancing our capabilities in the area of food authentication. Challenges clearly remain in applying authentication methodologies to products of grassland because potential markers are influenced by the complexities of pre-slaughter diets available to animals and of the production systems themselves. Novel approaches are required to overcome these hurdles. These are likely to involve, initially, the use of multiple methodologies to measure multiple markers in multiple tissues and advanced chemometric techniques. Rather than developing authentication methods for broad categories of animalderived foods, such as ‘grass-fed’, ‘organic’, ‘free range’ or ‘extensive’ production, a more effective approach may be to develop authentication methodologies for products produced locally and to specific restricted feeding regimes. Animal products produced to a particular feed ‘recipe’ in a specific location are more likely to hold a unique marker fingerprint to differentiate them from other products and thus underpin an authenticity-based traceability system associated with their production. From a grassland perspective, simplification of the production process so that high-value products are produced to a recipe which can be easily validated by authentication methodologies, but which cannot be mimicked by fraudulent production practices, may be the most promising way forward. References Addis M., Cabiddu A., Pinna G., Decandia M., Piredda G., Pirsi A. and Molle G. (2005) Milk and cheese fatty acid composition in sheep fed Mediterranean forages with reference to conjugated linoleic acid cis-9, trans-11. Journal of Dairy Science 88, 3443- 3454. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 516 Adle S.A., Dah A.V., Jense S.K., Thuen E., Gustavsson A.M. and Steinsham H. (2013) Fatty acid composition, fat-soluble vitamin concentrations and oxidative stability in bovine milk produced on two pastures with different botanical composition. Livestock Science 154, 93-102. Al-Mabuk R.M., Bec N.F.G. and Dewhurst R.J. (2004) Effects of silage species and supplemental vitamin E on the oxidative stability of milk. Journal of Dairy Science 87, 406-412. Bahar B., Schmidt O., Moloney A.P., Scrimgeour C.M., Begley I.S. and Monahan F.J. (2008) Seasonal variation in the C, N and S stable isotope composition of retail organic and conventional Irish beef. Food Chemistry 106, Bontempo L., Lombardi G., Paoletti R., Ziller L. and Camin F. (2012) H, C, N and O stable isotope characteristics of alpine forage, milk and cheese. International Dairy Journal 23, 99-104. Bugaud C., Buchin S. Hauwuy A. and Coulon J.-B. (2001) Relationships between flavour and chemical composition of Abondance cheese derived from different types of pasture. Lait 81, 757-774. Butler G., Stergiadis S., Seal C., Eyre M. and Leifert C. (2011) Fat composition if organic and retail milk in northeast England. Journal of Dairy Science 94, 24-36. Camin F., Bontempo L., Heinrich K., Horacek M., Kelly S.D., Schlicht C., Thomas F., Monahan F.J., Hoogewerff J. and Rossmann A. (2007) Multi-element (H,C,N,S) stable isotope characteristics of lamb meat from different European regions. Analytical and Bioanalytical Chemistry 389, 309-320. Capuano E., Boerrigter-Eenling R., Van der Veer G. and Van Ruth S. (2013) Analytical authentication of organic products: overview of markers. Journal of the Science of Food and Agriculture 93, 12-28. Cassar-Malek I., Jurie C., Bernard C., Barnola I., Micol D. and Hocquette J.-F. (2009) Pasture-feeding of Charolais steers influences skeletal muscle metabolism and gene expression. Journal of Physiology and Pharmacology 60, Suppl 3, 83-90. Chiofalo B., Simonella S., Di Grigoli A., Liotta L., Frenda A.S., Lo Presti V., Bonanno A. and Chiofalo V. (2010) Chemical and acidic composition of longissimus dorsi muscle of Comisana lambs fed with Trifolium subterraneum and Lolium multiflorum. Small Ruminant Research 88, 89-96. Claps S., Sepe L., Caputo A.R., Di Trana A., Paladino F., Di Napoli M.A., Pizzillo M. and Fedel V. (2009) Influence of four single fresh forages on volatile organic compound (VOC) content and profile and sensory properties of goat Caciotta cheese. Italian Journal of Animal Science 8, 390-393. Coulon J.B. and Priolo A. (2002) Influence of forage feeding on the composition and organoleptic properties of meat and dairy products, basis for a ‘terroir’ effect. Grassland Science in Europe 7, 513-524. Coulon J.-B., Delacroix-Buchet A., Martin B., and Pirisi A. (2004) Relationships between ruminant management and sensory characteristics of cheeses: a review. Lait 84, 221-241. Couvreur S., Hurtaud C., Lopez C., Delaby L. and Peyraud J.L. (2006) The linear relationship between the proportion of fresh grass in the cow diet, milk fatty acid composition, and butter properties. Journal of Dairy Science 89, 1956-1969. Croissant A.E., Washburn S.P., Dean L.L. and Drake M.A. (2007) Chemical properties and consumer perception of fluid milk from conventional and pasture-based production systems. Journal of Dairy Science 90, 4942-4953. Daley C. A., Abbott A., Doyle P. S., Nader G. A. and Larson S. (2010) A review of fatty acid profiles and antioxidant content in grass-fed and grain-fed beef. Nutrition Journal 9, 10. De la Torre A., Debiton E., Juaneda P., Durand D., Chardigny J.M., Barthomeuf C., Bauchart D. and Gruffat D. (2006) Beef conjugated linoleic acid isomers reduce human cancer cell growth even when associated with other beef fatty acids. British Journal of Nutrition 95, 346-352. Dennis M.J. (1998) Recent developments in food authentication. Analyst 123, 151R–156R. Descalzo A. M. and Sancho A. M. (2008) A review of natural antioxidants and their effects on oxidative status, odor and quality of fresh beef produced in Argentina. Meat Science 79, 423-436. Dewhurst R.J., Shingfield K.J., Lee M.R.F. and Scollan N.D. (2006) Increasing the concentrations of beneficial polyunsaturated fatty acids in milk produced by dairy cows in high-forage systems. Animal Feed Science and Technology 131, 168–206. Dian P.H.M., Andueza D., Jestin M., Prado I.N. and Prache S. (2008) Comparison of visible and near infrared reflectance spectroscopy to discriminate between pasture-fed and concentrate-fed lamb carcasses. Meat Science 80, 1157-1164. Dierking R.M., Kallenbach R.L. and Grun I.U. (2010) Effect of forage species on fatty acid content and performance of pasture-finished steers. Meat Science 85, 597-605. Duckett S.K., Pratt S.L. and Pavan E. (2009) Corn oil or corn grain supplementation to steers grazing endophytefree tall fescue. II. Effects on subcutaneous fatty acid content and lipogenic gene expression. Journal of Animal Science 87, 1120-1128. Duckett S. K., Neel J. P., Lewis R. M., Fontenot J. P. and Clapham W. M.(2013) Effects of forage species or concentrate finishing on animal performance, carcass and meat quality. Journal of Animal Science 91, 1454-1467. Elmore J.S., Warren H.E., Mottram D.S., Scollan N.D., Enser M., Richardson R.I. and Wood J.D. (2004) A comparison of the aroma volatiles and fatty acid compositions of grilled beef muscle from Aberdeen Angus and Holstein-Friesian steers fed diets based on silage or concentrates. Meat Science 68, 27-33. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 517 Engel E., Ferlay A., Cornu A., Chilliard Y., Agabriel C., Bielicki G. and Martin B. (2007) Relevance of isotopic and molecular biomarkers for the authentiction of milk according to production zone and type of feeding to the cow. Agricultural and Food Chemistry 55, 9099-9108. Gori A., Cevoli C., Fabbri A., Caboni M.F. and Losi G. (2012) A rapid method to discriminate season of production and feeding regimen of butters based on infrared spectroscopy and artificial neural networks. Journal of Food Engineering 109, 525-530. Havemose M.S., Weisbjerg M.R., Bredie W.L.P., Poulsen H.D. and Nielsen J.H. (2006) Oxidative stability of milk influenced by fatty acids, antioxidants and copper derived from feed. Journal of Dairy Science 89, 1970-80. Hocquette J.-F., Cassar-Malek I., Bernard-Capel C. and Picard B. (2009) Functional genomics and new markers for beef production – mini review. Animal Science Papers and Reports 27, 273-280. Kalber T., Meier J.S., Kreuzer M. and Leiber F. (2011) Flowering catch crops used as forage plants for dairy cows: Influence on fatty acids and tocopherols in milk. Journal of Dairy Science 94, 1477-1489. Karoui R. and De Baerdemaeker J. (2007) A review of the analytical methods coupled with chemometric tools for the determination of the quality and identity of dairy products. Food Chemistry 102, 621-640. Larsen M.K., Kidmose U., Kristensen T., Beaumont P. and Mortensen G. (2013) Chemical composition and sensory quality of bovine milk as affected by type of forage and proportion of concentrate in the feed ration. Journal of the Science of Food and Agriculture 93, 93-99. Lourenço M., Van Ranst G., Vlaeminck B., De Smet S. and Fievez V. (2008) Influence of different dietary forages on the fatty acid composition of rumen digesta as well as ruminant meat and milk. Animal Feed Science and Technology 145, 418-437. Li Y. F. and Liu S. M. (2012) Reducing lipid peroxidation for improving colour stability of beef and lamb: onfarm considerations. Journal of the Science of Food and Agriculture 92, 719-726. Luykx D.M.A.M. and Van Ruth S.M. (2008) An overview of analytical methods for determining the geographical origin of food products. Food Chemistry 107, 897–911. Mahecha L., Dannenberger D., Nuernberg K., Nuernberg G., Hagemann E. and Martin J. (2010) Relationship between lipid peroxidation and antioxidant status in the muscle of German Holstein bulls fed n-3 and n-6 PUFAenriched diets. Journal of Agricultural and Food Chemistry 58, 8407-8413. Mariaca R.G., Berger T.F.H., Gauch R., Imhof M.I., Jeangros B. and Bosset J.O. (1997) Occurrence of volatile mono- and sesquiterpenoids in highland and lowland plant species as possible precursors for flavor compounds in milk and dairy products. Journal of Agricultural and Food Chemistry 45, 4423-4434. Martin B., Verdier-Metz I., Buchin S., Hurtaud C., and Coulon J.B. (2005a) How do the nature of forages and pasture diversity influence the sensory quality of dairy livestock products? Animal Science 81, 205-212. Martin B., Cornu A., Kondjoyan N., Ferlay A., Verdier-Metz I., Pradel P., Rock E., Chilliard Y., Coulon J.B. and Berdagué J.L. (2005b) Milk indicators for recognizing the types of forages eaten by dairy cows. In: Hocquette J.F. and Gigli S. (eds) Indicators of milk and beef quality, EAAP Publication No. 112, Wageningen Academic Publishers, The Netherlands, pp. 127-136. Molkentin J. and Giesemann A. (2007) Differentiation of organically and conventionally produced milk by stable isotope and fatty acid analysis. Analytical and Bioanalytical Chemistry 388, 297-305. Moloney A. P., Fievez V., Martin B., Nute G. R., and Richardson R. I. (2008) Botanically diverse forage-based rations for cattle: implications for product composition and quality and consumer health. Grassland Science in Europe 13, 361-374. Monahan F.J., Moloney A.P., Downey G. Dunne P.G., Schmidt O and Harrison S.M. (2010) Authenticity and traceability of grassland production and products. Grassland Science in Europe 15, 401-414. Moorby J.M., Lee M.R.F., Davies D.R., Kim E.J., Nute G.R., Ellis N.M. and Scollan N.D. (2009) Assessment of dietary ratios of red clover and grass silages on milk production and milk quality in dairy cows. Journal of Dairy Science 92, 1148-1160. Morgan S., Huws S. A. and Scollan N. D. (2012) Progress in forage-based strategies to improve the fatty acid composition of beef. Grassland Science in Europe 17, 295-307. Muir P.D., Deaker J.M. and Brown M.D. (1998) Effects of forage-and-grain-based feeding systems on beef quality: A review. New Zealand Journal of Agricultural Research 41, 623-635. Nozière P., Graulet B., Lucas A., Martin B.,Grolier P. and M. Doreau M. (2006) Carotenoids for ruminants: From forages to dairy products. Animal Feed Science and Technology 131, 418–450. Osorio M.T., Moloney A.P., Schmidt O. and Monahan F.J. (2011) Beef authentication and retrospective dietary verification using stable isotope ratio analysis of bovine muscle and tail hair. Journal of Agricultural and Food Chemistry 59, 3295–3305. Ostrovsky I., Pavlikova E., Blasko J., Gorova R., Kubinec R., Margetin M. and Sojak L. (2009) Variation in fatty acid composition of ewes' milk during continuous transition from dry winter to natural pasture diet. International Dairy Journal 19, 545-549. Petersen M.B., Søegaard K. and Jensen S.K. (2011) Herb feeding increases n-3 and n-6 fatty acids in cow milk. Livestock Science 141, 90-94. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 518 Phelan P., Moloney A.P., McGeough E. J., Humphreys J., Bertilsson J., O’Riordan E. and O’Kiely P. (2014) Forage legumes for grazing and conserving in ruminant production systems. Critical Reviews in Plant Sciences (in press). Povolo M. Pelizzola V., Lombardi G., Tava A and Contarini G. (2012) Hydrocarbon and fatty acid composition of cheese as affected by the pasture vegetation type. Agricultural and Food Chemistry 60, 299-308. Prache S. (2009) Diet authentication in sheep from composition of animal tissues and products. Revista Brasileira de Zootecnia 38, 362-370. (supl. especial) Priolo A., Cornu A., Prache S., Krogmann M., Kondjoyan N., Micol D. and Berdagué J.-L. (2004) Fat volatile tracers of grass feeding in sheep. Meat Science 66, 475-481. Priolo A., Micol D., and Agabriel J. (2001) Effects of grass feeding systems on ruminant meat colour and flavour: a review. Animal Research, 50, 185-200. Raes K., Balcaen A., Dirinck P., De Winne A., Claeys E., Demeyer D. and De Smet S. (2003) Meat quality, fatty acid composition and flavour analysis in Belgian retail beef. Meat Science 65, 1237-1246. Renou J.P., Deponge C., Gachon P., Bonnefoy J.-C, Coulon J.-B., Garel J.-P., Vérité R. and Ritz P. (2004) Characterization of animal products according to geographic origin and feeding diet using nuclear magnetic resonance and isotope ratio mass spectrometry: cow milk. Food Chemistry 85, 63-66. Richardson R.I., Nute G.R., Wood J.D., Scollan N.D. and Warren H.E (2004) Effects of breed, diet and age on shelf life, muscle vitamin E and eating quality of beef. Proceedings of the British Society of Animal Science p 84 Röhrle F.T., Moloney A.P., Osorio M.T., Luciano G., Priolo A., Caplan P. and Monahan F.J. (2010a) Carotenoids, colour and reflectance parameters in bovine adipose tissue to discriminate between feeding systems. Meat Science 88, 347-353. Röhrle F.T., Moloney A.P., Black A., Osorio M.T., Sweeney T., Schmidt O. and Monahan F.J. (2010b) alphaTocopherol stereoisomers in beef as an indicator of vitamin E supplementation in cattle diets. Food Chemistry 124, 935-940. Rohrle F. (2014) Authentication of grass-fed beef. PhD thesis (in preparation), University College Dublin. Rossmann A., Haberhauer G., Hölzl, S., Horn P., Pichlmayer F. and Voerkelius S. (2000) The potential of multielement stable isotope analysis for regional assignment of butter. European Food Research and Technology 211, 32-40. Scaglia G., Fontenot J.P., Sweckr W.S., Corl B.A., Duckett S.K., Boland H.T., Smith R. and Abaye A.O. (2012) Performance, carcass, and meat characteristics of beef steers finished on 2 different forages or on a highconcentrate diet. Professional Animal Scientist 28, 194-203. Schmidt J.R., Miller M.C., Andrae J.G., Ellis S.E. and Duckett S.K. (2013) Effect of summer forage species grazed during finishing on animal performance, carcass quality, and meat quality. Journal of Animal Science 91, 44514461. Schreurs N.M., Lane G.A., Tavendale M.H., Barry T.N. and McNabb W.C. (2008) Pastoral flavour in meat products from ruminants fed fresh forages and its amelioration by forage condensed tannins. Animal Feed Science and Technology 146, 193–221. Scollan N.D., Dewhurst R.J. Moloney A.P. and Murphy J.J. (2005) Improving the quality of products from grassland. In: McGilloway, D.A.(ed) Grassland: a global resource, Wageningen Academic Publishers, The Netherlands, pp. 41-56. Scollan N.D., Dannenberger D., Nuernberg K., Richardson I., MacKintosh S, Hocquette J.F. and Moloney A.P. (2014) Enhancing the nutritional and health value of beef lipids and their relationship with meat quality. Meat Science (in press). Serrano E., Cornu A., Kondjoyan N., Agabriel J. and Micol D. (2011) Tracability of grass feeding in beef: terpenes, 2,3-octanedione and skatole accumulation in adipose tissue of young bulls. Animal 5, 641-649. Shibata M., Matsumoto K., Oe M., Ohnishi-Kameyama M., Ojima K., Nakajima I., Muroya S and Chikuni K. (2009) Differential expression of the skeletal muscle proteome in grazed cattle. Journal of Animal Science 87, 2700–2708. Shingfield K. J., Bonnet M. and Scollan N. D. (2013) Recent developments in altering the fatty acid composition of ruminant-derived foods. Animal 7, 132-162. Steinshamn H. (2010) Effect of forage legumes on feed intake, milk production and milk quality – a review. Animal Science Papers and Reports 28, 195-206. Sun D.-W. (2008) Modern Techniques for Food Authentication. Academic Press, London, UK, 689pp. Timmons J.S., Weiss W.P., Palmquist D.L. and Harper W.J. (2001) Relationship among roasted soybeans, milk components, and spontaneous oxidized flavour of milk. Journal of Dairy Science 84, 2440-2449. Tornambé G., Cornu A., Pradel P., Kondjoyan N., Carnat A. P., Petit M. and Martin B. (2006) Changes in terpene content in milk from pasture-fed cows. Journal of Dairy Science 89, 2309–2319. Tansawat R., Maughan C. A. J., Ward R. E., Martini S. and Cornforth D. P. (2013) Chemical characterisation of pasture- and grain-fed beef related to meat quality and flavour attributes. International Journal of Food Science and Technology 48, 484-495. TRACE (2010). http://www.trace.eu.org. Tracing the Origin of Food. EU Framework 6 Programme. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 519 Turner S.-A., Waghorn G.C., Woodward S.L. and Thomson N.A. (2005) Condensed tannins in birdsfoot refoil (Lotus corniculatus) affect the detailed composition of milk from dairy cows. Proceedings of the New Zealand Society of Animal Production 65, 283-289. Valenti B. Martin B., Andueza D., Leroux C., Labonne C., Lahalle F., Larroque H., Brunschwig P., Lecomte C, Brochard M. and Ferlay A. (2013) Infrared spectroscopic methods for the discrimination of cow’s milk according to feeding system, cow breed and altitude of the dairy farm. International Dairy Journal 32, 26-32. Van Dorland H.A., Kreuzer M., Leuenberger H. and Wettstein H.-R. (2008) Comparative potential of white and red clover to modify the milk fatty acid profile of cows fed ryegrass-based diets from zero-grazing and silage systems. Journal of the Science of Food and Agriculture 88, 77-85. Vasta V. and Priolo A. (2006) Ruminant fat volatiles as affected by diet. A review. Meat Science 73, 218-228. Vasta V., Luciano G., Dimauro C., Röhrle F., Priolo A., Monahan F.J. and Moloney A.P. (2011) The volatile profile of longissimus dorsi muscle of heifers fed pasture, pasture silage or cereal concentrate: implication for dietary discrimination. Meat Science 87, 282-289. Vetter W. and Schroeder M. (2010) Concentrations of phytanic acid and pristanic acid are higher in organic than in conventional dairy products from the German market. Food Chemistry 119, 746-752. Viallon C., Martin B., Verdier-Metz I., Pradel P., Garel J.P., Coulon J.B. and Berdagué J.L. (2000) Transfer of monoterpenes and sesquiterpenes from forages into milk fat. Lait 80, 635-641. WHO/FAO (2007). World Health Organization – Food and Agriculture Organization of the United Nations. Codex Alimentarius. Food Import and Export Inspection and Certification Systems. 3rd edition. p 79. WHO (2003) Diet, nutrition and the prevention of chronic diseases. Report of a joint WHO/FAO expert consultation. WHO technical report series 916, Geneva. Young O., Lane G., Priolo A. and Frazer K. (20030 Pastoral and species flavour in lambs raised on pasture, lucerne or maize. Journal of the Science of Food and Agriculture 83, 93-104. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 520 Sustainable intensification of grass-based ruminant production Baumont R.1, Lewis E.2, Delaby L.3, Prache S.1 and Horan B.2 1 INRA, UMR1213 Herbivores, Theix, 63122 St Genès Champanelle, France 2 Teagasc, Animal & Grassland Research and Innovation Centre, Moorepark, Fermoy, Co. Cork, Ireland 3 INRA, Agro Campus Rennes, UMR1348 PEGASE, 35590 Saint-Gilles, France Corresponding author: rene.baumont@clermont.inra.fr Abstract The sustainable or ecological intensification of grass-based food production systems provides an opportunity to align the ever increasing global demand for food with the necessity to regreen ruminant production. The challenge for production scientists now is to find innovative ways to improve grass-based production processes to maximize resource use-efficiency based on improved management practices. The objective of this paper is, firstly, to outline the potential opportunities to enhance the yield and quality of grasslands for grazing and conserved forages paying particular attention to species diversity and legumes. Subsequently, the paper addresses the necessity to choose appropriate animals and management practices to improve productive and reproductive performance within such systems. Finally, the paper reports experimental results from dairy cow and sheep production systems that succeeded in combining high animal performance with low environmental impacts. Keywords: grass yield, quality, legumes, animal productivity, grazing, dairy cow, sheep Introduction Global food demand, climate change, urbanization and bio-fuel production will increase competition for agricultural land use between crop and herbivore production. Consequently, it is expected that ruminant production will have to be concentrated on non-arable lands (permanent grasslands) and on arable lands where it would be the most profitable system of production. European grass-based ruminant production systems face a threefold challenge: i) to meet the rapidly changing demand for food within a resource-constrained environment; ii) to do so in an environmentally and socially sustainable manner for consumers and producers; and iii) to ensure that the products produced meet the highest standards of quality and nutritional value for increasingly discerning consumers. Producing more food from the same land area, while reducing environmental impacts, requires what has been referred to as ‘sustainable intensification’ (Pretty, 1997) or ‘ecological intensification’ (Griffon, 2013) of agricultural production. This is based on new innovative blueprints of production based on increased herbage production and quality, and improved utilization under grazing. The aim of the present contribution is to propose a framework of sustainable intensification in grass-based ruminant production. The definition and challenges of sustainable intensification in grass-based ruminant production will first be outlined. Then specific aspects, including grass production and quality, the type of animals and management practices will be addressed. Sustainable intensification: definition and challenges Since the 1970s, and the growing environmental concern over industrial agriculture and livestock production, there has been a consensus in society and the scientific community on the necessity to re-green agriculture. However, the best approach to achieve re-greening is still a matter for much debate among production scientists, legislators and other public stakeholders, and numerous concepts can be found in the literature (Griffon, 2013): 'sustainable agriculture', 'conservation agriculture', 'agro-ecology', 'organic farming', 'high nature-value agriculture', 'ecological or sustainable intensified agriculture' etc. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 521 Ecological intensification of agricultural systems was recently defined by Hochman et al. (2013) as producing more food per unit of resource used, while minimizing the impact on the environment. Griffon (2013) stated that ecological intensification contrasts with chemical and energy use for crop production and with the use of feed and drugs for animal production. Thus, ecological intensification relies more heavily on the use of natural resources and the functionality (or ecosystem services) they provide and it places less emphasis on the use of external inputs. The challenge for such systems is to improve the efficiency of natural resource use in order to increase food production from existing farmland while minimizing pressure on the environment. The literature on ecological intensification and agro-ecology in animal production is scarce. Recently, Dumont et al. (2013) proposed the development of ecology-based alternatives for animal production by adopting management practices to improve animal health, decreasing the inputs needed for production, decreasing pollution by optimizing the metabolic functioning of farming systems and by adapting management practices to enhance biological diversity and strengthen the resilience within animal production systems. Grass-based systems have been shown to be beneficial to the environment (Jankowska-Huflejt, 2006; Peyraud et al., 2010) and to be economically successful by reducing total costs of food production (Dillon et al., 2005). However, systems based on increased mobilization of services provided by natural resources (low supplementary feed input or organic systems) have to cope with more variability in relation to plants, animals, climate and bio-aggressors. Such systems therefore require plants and animals that are robust and easy to manage. The production performance of grassland is dependent on the yield of utilizable energy and protein in the grass grown. This means that increasing the yield and stability of high quality grass growth is imperative to ensure the robustness of the overall system. The animal required for efficient grassland-based production systems must be robust, autonomous and ‘easy care’, and capable of high levels of performance from a predominantly grazed pasture diet. Finally, the management of such systems has to maximize the utilization of a renewable low cost resource without adverse effects on the environment over the long term. Achieving and maintaining optimum soil fertility is a prerequisite for high productivity grassland production systems. The topic of soil fertility is, however, outside the direct scope of this paper. In order to maximize nutrient-use efficiency within grass-based production systems, and to minimize environmental impacts, nutrient recycling must be improved by closely matching nutrient supply to grassland demand according to the climatic conditions. Combining yield and quality of grasslands for grazing and conserved forages Forage production and forage quality must be increased without increasing the negative effects on the environment. In recent years research efforts have moved from improving the production and quality of single-species swards through breeding and fertilization to focus on the role of diversity and thus multispecies sown and permanent swards. The role of species diversity in sown and permanent swards In a recent review of the literature, Huyghe et al. (2012) showed that a positive relationship between species diversity in sown swards and biomass production is frequently found in controlled environments. More so than the number of species, the functional diversity (type of grasses, legumes, forbs) has a positive effect on forage yield (Kirwan et al., 2007), and the highest yields are often achieved with intermediate species diversity. The relationship between species diversity and herbage feed value has received little attention. In an experiment reported by Huyghe et al. (2008) in which species diversity ranged from one to eight species, the negative relationship between forage yield and in vitro digestibility was not affected by the number of species nor by the number of functional groups. A notable advantage Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 522 of complex swards combining grasses and legumes was the more stable chemical composition across the year that allows more flexibility in sward utilization for grazing and harvesting. The functional composition is also a key feature which explains variation in productivity and quality and their temporal patterns in permanent grasslands (Duru et al., 2010). In a recent study in France on a set of 190 permanent grasslands representative of most pedo-climatic conditions from Atlantic to Alpine areas (Michaud et al., 2014), feed value, both in early and late spring, was positively related to the proportion of legume species in the sward. A higher stability of forage quality in spring was related to high proportions of forbs and conservative grasses (Figure 1). 2nd Axis : 16.57% 0.8 Legumes 0.4 0 - 0.4 Feed value late spring Potential feed value (early spring) Summer yield Competive species Spring yield Forbs Feed value stability 1st Axis : 26.23% Conservative species Figure 1. Principal component analysis of herbage yield, feed value and functional composition on a set of 190 permanent grasslands in France (from the data of Michaud et al., 2014) Grasses - 0.8 - 0.8 - 0.4 0 0.4 0.8 Species diversity may have a positive effect on voluntary food intake, as diversity offers a choice of herbage species and previous studies have reported that choice significantly increases intake at pasture in sheep (Cortes et al., 2006) and indoors with conserved forages (Ginane et al., 2002 with heifers, Bruinenberg et al., 2003 with cows). It is not clear whether the variation in intake is a consequence of the diversity per se or of the choice situation, even though the experiments of Cortes et al. (2006) support the hypothesis of a true effect of diversity. In contrast, Soder et al. (2006) found no effect of species diversity on dairy cow intake, suggesting that the animal type could interact with the species diversity effect. The specific role of legume species Introducing legume species into conserved and grazed forages gives many advantages relating to feed value, animal performance and environmental impact. White clover has a high digestibility and a high energy value; this is attributed to its low fibre concentration which reflects the absence of structural components such as stems and sheaths (Ayres et al., 1998). A particular advantage of white clover is the reduced rate of decline in digestibility in the mid-season compared to perennial ryegrass (Ulyatt, 1970). Furthermore, white clover supports increased voluntary intake (Ribeiro Filho et al., 2003). Increased production performance as a result of increasing the sward white clover proportion has been observed in dairy cows (Dewhurst et al., 2003), beef steers (Thomas et al., 1981) and sheep (Orr et al., 1990). The results depend on the proportion of white clover in the sward. Egan et al. (2013) found that cows grazing a grass-clover sward (21.6% clover content) had higher milk yield and milk solids yield than cows grazing a grass-only sward, whereas Enriquez-Hidalgo et al. (2012) found no difference in milk yield and milk solids yield when the average sward clover content was only 13%. When expressed on a proportion of gross energy intake or unit-intake basis, methane emissions are often lower for forage legume-fed animals than grass-fed animals (McCaughey et al., 1999; Waghorn et al., 2002). Beauchemin et al. (2008) proposed that this was due to the lower fibre content, higher dry matter intake, increased passage rate and presence of condensed tannins Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 523 (CTs) in certain legume species such as Lotus corniculatus or Onobrychis viciifolia. Although the effect of CTs on methane emissions has been found in several studies in vitro (e.g. Niderkorn et al., 2011), it remains uncertain in vivo. Condensed tannins are also important in reducing the degradation of forage proteins in the rumen, without reducing the amount of microbial protein which is synthesized, and in deactivating internal parasites (Min et al., 2003). These features have positive implications for nitrogen excretion and the use of anthelmintic drugs respectively. Promising results were obtained with Onobrychisv viciifolia, with similar voluntary intake and digestibility as lucerne, but lower protein degradation in the rumen resulting in lower protein excretion in urine (Theodoridou et al., 2010). Another interesting secondary compound is polyphenol oxidase, which is present in red clover and which was shown to inhibit proteolysis and lipolysis (Lee et al., 2004). These properties could improve silage conservation, as CTs also do (Theodoridou et al., 2012), and limit fatty acid hydrogenation in the rumen. A challenge for the future is to find positive associative effects between plants, relying on synergies between the different bioactive compounds (Niderkorn and Baumont, 2009; Aufrère et al., 2012): e.g., could CTs from one plant bind with proteins from another plant? Recently, synergies between cocksfoot silage and red clover silage, and between ryegrass and chicory, were observed for DM and NDF intake and eating rate, with 50:50 being the optimal proportions. For the cocksfoot silage-red clover silage association, the synergistic effect was also observed on daily digestible organic matter intake (Niderkorn et al., 2014). Quality vs. quantity: biomass accumulation and stage of maturity As biomass is accumulating with plant growth, sward quality in terms of net energy, protein content and potential voluntary intake is decreasing as a result of plant maturation. This is well documented for single-species swards of grasses and legumes, e.g. in the feed value tables used in France (Baumont et al., 2007). A less mature plant contains a lower proportion of true stem and dead material and a greater proportion of leaf which is lower in fibre and highly digestible (Curran et al., 2010; Beecher et al., 2013). As the plant enters the reproductive stage, leaf proportion decreases and stem proportion increases, with negative effects on sward digestibility, crude protein concentration and voluntary intake. Thus, harvesting and grazing management have to deal with the trade-off between forage quantity and quality. In spring, for conservation, the forage should be cut at the beginning of grass heading to maximize net energy and protein harvested per ha. The decrease in feed value between the beginning and end of heading will necessitate 2.5 kg/day more concentrate for a dairy cow to produce 30 kg of milk. At grazing, increased frequency of defoliation results in high quality but a decrease in net herbage accumulation whereas infrequent defoliation leads to greater herbage production, but decreased grass feed value (Hoogendorn et al., 1992; Beecher et al., submitted). High biomass yield (kg DM ha-1) at grazing will limit animal performance through digestive constraints (low intake of poorly digestible matter), but high quality swards (low biomass yield) can also limit animal performance through behavioural constraints if the time required to graze the required quantity of grass is too great (Baumont et al., 2004). The effect of biomass yield at grazing may also vary over the grazing season. Tunon (2013) found no effect of biomass yield up to 2,300 kg DM ha-1 (>4 cm) from April to July, but observed a reduction in milk fat plus protein yield compared to 1,000 and 1,500 kg DM ha-1 swards from July to October. McEvoy et al. (2010) found similar results. This agrees with work by Beecher et al. (submitted) who observed no difference in OM digestibility between swards with a biomass <1500 kg DM ha-1 and >2000 kg DM ha-1 in spring. In summer and autumn however, increasing biomass yield resulted in a significant decrease in OM digestibility and in digestible OM intake (Figure 2). Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 524 Digestible OM intake (g OM/W 0.75 ) 60 55 y = -0.0029x + 53.3 R² = 0.32 50 45 40 Spring 35 Figure 2. Effect of pre-grazing biomass on digestible dry matter intake measured in sheep (data from Beecher et al., submitted) Summer 30 y = -0.0048x + 56.4 R² = 0.74 25 20 0 1000 2000 3000 4000 5000 Pre-grazing biomass (kg DM /ha) 6000 Increasing grass quality via the use of less-mature grass also has positive implications for product quality (Hoogendorn et al., 1992) and environmental measures. For example, biomass yield had no effect on enteric methane emissions in spring, but in summer, grazing swards of high biomass resulted in higher emissions (Wims et al., 2010). Finally, to manage the trade-off between quantity and quality, grazing consistently very low biomass yield swards (<1,200 kg DM ha-1) should be avoided during the main grazing season as this can depress pasture regrowth rate. Diversified swards in which species have different growth and maturation rates and the use of late-heading cultivars could help manage the tradeoff between quantity and quality by smoothing the biomass accumulation and the associated decline in feed value. Choosing the appropriate animal for sustainable grass-based production As only ~10% of the world’s milk production comes from grazing systems, the majority of global ruminant livestock have not been selected for grazing systems. The long running scientific debate on the importance of genotype × environment interactions has been refuelled in recent years as the interest in grass-based systems in Europe has increased. Until recently, most experimental results have indicated little or no importance of such interactions (Holmes, 1995); however, increasingly diverse genotypes and/or production environments have increased the likelihood of such interactions (Falconer, 1990). There is now strong evidence to show that the animals that are genetically best suited to non-grazing systems, are not suited to grazing systems (Delaby et al., 2010). Animals for grass-based systems Successful grazing systems require animals capable of achieving large intakes of forage relative to their genetic potential for production so that they can achieve their nutritional requirements almost entirely from grazing, with some conserved forage. As animal intakes at grazing are reduced relative to confinement systems, efficient animals within grazing systems have moderated feed requirements which are consistent with the feed supply capability of grazing. Consequently, such autonomous animals at grazing can achieve high milk production (and composition), retain optimal reproductive capacity and maintain adequate body reserves to avoid ill health within a restricted feed environment (Delaby et al., 2009; Cutullic et al., 2011). Additionally in the context of sustainable intensification, the necessity for more animals to be managed by individual farmers requires more robust autonomous livestock requiring less individual managerial assistance. Animals for grazing systems must also be able to graze effectively and to walk long distances, abilities that are not required in confinement systems. The use of alternative breeds or crossbreeding to satisfy the specificities of an animal suited to the grass-based system is now being considered by farmers in many countries. Dual purpose or cross breed cows seem more flexible and better adapted to grazing and have improved health, Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 525 milk value, reproductive performance, feed efficiency and beef value. Recent studies have examined the suitability of alternative cow breeds for grass-based milk production systems in both France (Delaby et al., 2014) and Ireland (Coleman et al., 2009; Prendiville et al., 2009). In France, an experiment run at the INRA experimental farm Le Pin-au-Haras (Normandy) evaluated the ability of different types of dairy cow to produce and to reproduce in response to two contrasting feeding strategies in a compact calving context (Table 1). Table 1. Milk and reproductive performance of dairy cows according to breed and feeding strategy (from the INRA experiment in Le Pin-au-Haras; n=380 lactations) Breed Holstein Feeding strategy High (1) Low Normande (2) High (1) Low (2) Milk yield (44 weeks - kg) 8515 6022 6332 4798 Fat content (g/kg) 38.0 39.5 41.0 41.8 Protein content (g/kg) 32.1 31.0 34.9 33.3 Milk Solids (kg) 587 418 466 351 Body Condition Score change (pts) -1.00 -1.25 -0.60 -0.90 First insemination success (%) 28 20 41 38 Recalving (%) 59 44 71 68 (1) The High feeding strategy had maize silage, grass silage, dehydrated alfalfa and concentrates in the indoor diet, a higher stocking rate, and supplementation with maize silage, grass silage and concentrate during the grazing season. (2) The Low feeding strategy had grass silage and haylage in the indoor diet, a lower stocking rate, and supplementation with grass silage during the grazing season. The high reactivity of the milk production in Holstein cows, as well as high body condition score loss and poor reproduction performance, makes the Holstein cow incompatible with the herbage system with no concentrate input and compact spring calving. In contrast, the Normande, a dual-purpose breed, appears less sensitive and better adapted to low input systems based on the maximization of grassland use for milk production. In Ireland, Prendiville et al. (2009) compared the biological efficiency of three genotypes (Jersey, Holstein-Friesian and Jersey × Holstein-Friesian) within grass-based systems. They reported higher milk production efficiencies among Jersey × Holstein-Friesian cattle compared to purebred Holstein-Friesian cattle. In comparison with purebred Holstein-Friesian dairy cattle, Jersey × Holstein-Friesian cattle achieved improved reproductive performance, a greater intake per kg of body weight at grazing and consequently, required 12% less grass to produce 1 kg of milk fat plus protein. Differences in intake and varying measures of feed conversion efficiency between dairy cows of various pure breeds have also been reported previously (for review, see Grainger and Goddard, 2004). Animal management for grass-based systems Grass-based systems of milk production require compact calving in spring to match feed supply and herd demand. This is based on achieving high rates of pregnancy within a short period of time following the start of breeding. Calving date is an important determinant of milk production and feed utilization in grass-based systems, through its impact on the alignment of feed demand with supply. Altering the mean calving date of the herd may have a role in reducing the reliance of grass-based farm systems on purchased feeds particularly at higher stocking rates. Both Dillon et al. (1995) and McCarthy et al. (2013) observed that delaying calving until March achieved a better alignment of dairy herd requirements and grass growth within Irish grass-based milk production studies, increased milk production from grazed grass, reduced the Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 526 requirement for purchased supplements and achieved a greater efficiency of energy utilization particularly at higher stocking rates. In sheep production, high ewe productivity (number of lambs produced per ewe per year × mean lamb carcass weight) is a key indicator to ensure economic sustainability. In organic sheep production systems, where hormonal treatments are prohibited and the price of concentrate feed increases the pressure on production costs, increasing reproductive capacity could be used to increase ewe productivity without undue penalty for other technical or economic performance targets. Increasing ewe productivity by increasing ewe reproductive capacity was evaluated on a rustic breed (Limousine) over a four-year period. In the study, one lambing per ewe per year (1/year) (with 50% lambings in spring and 50% in autumn) was compared to three lambings over two years (3in2) (Benoit et al., 2009). Ewe productivity in 1/year (1.51) was slightly lower than in 3in2 (1.61), but with a lower between-year variability, lower lamb mortality and parasitism level and lower concentrate feed consumption per ewe. Lamb carcass conformation, fatness and fat colour were not different between systems, but carcass weight and subcutaneous dorsal fat firmness were lower in 3in2 lambs than in 1/year. Intensification in an organic sheep system through increased reproduction rhythm therefore did not lead to better animal performance nor economic results and proved riskier, more variable and more difficult to manage, and thus less sustainable. The less intensive system (1/year) was highly efficient from the animal perspective and highly food self-sufficient (Benoit et al., 2009). Improving the efficiency of resource use in grazing systems Grazed grass is the cheapest feed source (Finneran et al., 2012) and commonly comprises 0.60 to 0.90 of ruminant animal diets within grass-based systems in Europe. Therefore, the production and utilization of increased quantities of higher quality grazed grass, coupled with the close alignment of grass production and animal requirements, has the potential to increase overall system productivity and contribute significantly to the sustainable intensification of agricultural production. Further research is therefore needed to develop management practices and technologies that will facilitate increases in milk and meat output from sustainable grassbased systems. High stocking rates in grass-based dairy systems can be compatible with high environmental performance Stocking rate (SR), traditionally defined as the number of animals per unit area of land used during a defined period of time, is widely acknowledged as the main driver of productivity from grazing systems and this applies across all grazing species (Rattray, 1987; Hoden et al., 1991; Baudracco et al., 2010; Crosson and McGee; 2011) due to its dominant effect on animal demand and hence pasture use. Increasing SR is usually associated with an increase in grazing severity (i.e., low post grazing residual sward height) and many studies have attributed the increased productivity of higher SR systems to an improvement in herbage utilization (McMeekan and Walsh, 1963). Penno (1999) suggested that the ideal stocking rate should balance the dual objectives of generous feeding to achieve high levels of production efficiency per animal and underfeeding to achieve high levels of pasture utilization to meet the overall objective of optimizing farm efficiency and profitability, while accounting for year-to-year variability in climate and grassland productivity. On that basis, intensified systems require grazing management practices that maximize pasture production and quality, which, in combination with increased stocking density, will result in increased overall system productivity (McEvoy et al., 2009; Curran et al., 2010; McCarthy et al., 2013). A recent review of SR experiments reported a 0.20 increase in milk production per ha arising from a 1 cow ha-1 increase in SR, where no extra supplement was fed as SR increased (McCarthy et al., 2011). In addition to the economic and animal welfare benefits associated with grazing, grass-based ruminant livestock production systems provide an environmentally sustainable food production Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 527 model. In comparison with cropping, grassland is an important biological filter which reduces nutrient and chemical run off and supports biodiversity and carbon storage. Compared with arable land, grassland is associated with a better conservation of soil against erosion, and reduced runoff and leaching of nutrients into surface and ground water (Briemle and Elsasser, 1997; Jankowska-Huflejt, 2006). Grassland also acts as an important carbon sink for GHG emissions, due to its high organic matter content relative to arable land (Leip et al., 2010). Notwithstanding these benefits, the efficiency of Nitrogen (N) use within grass-based systems is variable and can potentially result in nutrient loss to water resources by leaching. To comply with the Irish obligations pursuant to the EU Nitrate Directive (91/676/EEC), a long term study of the factors influencing nitrate loss beneath intensive dairy production systems in vulnerable soil types was undertaken at Curtins Farm, Teagasc Moorepark, in the south of Ireland over a 10-year period from 2001 to 2011. The Curtins Farm soil type is representative of the highest risk soils to nitrate leaching in Ireland. On the 48 ha site, cow numbers increased from 108 in 2001 to 138 in 2011, based on grazing management practices that increased grass growth and utilization which resulted in an increase in milk production from the site (Table 2). Table 2. The effect of farm system characteristics on the biological efficiency of grass-based milk production at Curtins Farm (Teagasc, Ireland) Year 2003 2005 2007 2009 2011 Stocking rate (cows/ha) 2.44 2.63 2.67 2.88 2.88 Grazing season (days) 293 295 306 287 285 Chemical N inputs (kg/ha) 289 331 313 248 249 Concentrate (kg/cow) 716 636 590 288 430 Milk volume (‘000 L/ha) 15.6 15.5 14.6 14.4 15.3 Nitrate (NO3-N mg/l) 11.1 13.3 12.4 9.7 6.6 Best nutrient management practices were used on the farm to increase slurry-use efficiency and reduce fertilizer N application to the levels stipulated by legislation. Based upon detailed fieldscale knowledge of soil capacity for nutrient retention, improved timing and rate of organic fertilizer application, reduced reliance on chemical fertilizer and the adoption of minimum-till cultivation reseeding, the result of this strategy was a consistent improvement in N-use efficiency and a decline in nitrate concentrations in groundwater at the site during the study period (Table 2). In future, the sustainable intensification of animal production from pasture necessitates that management practices on the farm must be increasingly tailored to achieve excellent nutrient management outcomes in addition to productivity improvement. The results of this study indicated that intensive dairy production systems based on improved nutrient management and agronomic practices can quickly improve groundwater quality and lead to high water quality standards even on highly vulnerable free draining soils. Combining high animal productivity with high feed self-sufficiency in grassland-based organic sheep production systems The aim of organic farming is to establish and maintain soil-plant-animal interdependence and to create a sustainable agro-ecosystem based on local resources. Feed self-sufficiency and particularly forage self-sufficiency is therefore one of the fundamentals of organic farming. To ensure sustainability in organic sheep production, the aim of one of the experiments conducted in the INRA Redon experimental site with a rustic breed (Limousine) and seminatural grasslands with a stocking rate of 0.8 livestock unit/ha, was to optimize both ewe Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 528 productivity and self-sufficiency. This optimization was managed by i) fitting the lambing distribution to the seasonal dynamics of the vegetation resources, with 65% of the lambings in spring and 35% in autumn and adapting stocking rate to resource potential, ii) fattening lambs at pasture together with controlling parasitism level, iii) ensuring the provision of young forage of high quality for animals with high requirements (with increased use of legumes in swards), iv) practicing winter grazing for animals with low requirements, and v) sowing mixtures of cereals (triticale, barley and oats) and peas to increase feed self-sufficiency, with the aim of producing 40% to 50% of concentrate feed requirements (Prache et al., 2011). In this way, feed self-sufficiency reached 95% in the last 3 years of the experiment. Moreover, combining a low reliance on bought-in concentrate feed with no mineral fertilization led to a very low use of non-renewable energy (51.0 MJ/kg lamb carcass, estimated using a Life Cycle Assessment approach; Pottier et al., 2009) and low net greenhouse gases emissions (11.1 kg eq-CO2/kg carcass; Prache et al., 2011). Conclusion Sustainable intensification of grass-based ruminant production must ensure that the increased global demand for food is met in an economically and environmentally sustainable manner, producing a product which is acceptable to increasingly discernible consumers from a quality, social and ethical perspective. In this paper we have given examples showing that sustainable intensification of grass-based ruminant production allows both economic and environmental performance improvement at the farm level. At a broader level, other environmental issues such as carbon storage, biodiversity and landscape, and cultural issues must also be considered. The key point which must be addressed is improving the efficiency of the animal-grass dynamic with fewer inputs and in a long-term sustainable manner. Further research should address the trade-off between quantity and quality in grass production, the resilience of the resource and its persistency over time. More systemic research on the animal-grass dynamic is needed to select and manage animals for grass and grass for animals. Future research should also address animal self-sufficiency, animals which are easy to manage and which are resistant to climate variation and parasites. In terms of management, sustainable intensification means changing from a position where the emphasis is on controlling all the management parameters to a position focused on compromising with risks and searching for equilibrium. Finally we should not forget that the production system is managed by the farmer. Helping farmers with management decisions through appropriate tools that combine grass, animal and system issues is a challenge for the greater development of grass-based systems. References Aufrère J., Dudilieu M., Poncet C. and Baumont R. (2013) Mixing sainfoin and lucerne to improve the feed value of legumes fed to sheep by the effect of condensed tannins. Animal 7, 82-92. Ayres J.F., Nandra K.S. and Turner A.D. (1998) A study of the nutritive value of white clover in relation to different stages of phenological maturity in the primary growth phase in spring. Grass and Forage Science 53, 250–259. Baudracco B.J., Lopez-Villalobos N., Holmes C.W. and Macdonald K.A. (2010) Effects of stocking rate, supplementation, genotype and their interactions on grazing dairy systems: a review. New Zealand Journal of Agricultural Research 53, 109-133. Baumont R., Cohen-Salmon D., Prache S. and Sauvant D. (2004) A mechanistic model of intake and grazing behaviour integrating sward architecture and animal decisions. Animal Feed Science and Technology 112, 5-28. Baumont R., Dulphy J.P., Sauvant D., Meschy F., Aufrère J. and Peyraud J.L. (2007) Chapitre 8. Valeur alimentaire des fourrages et des matières premières : tables et prévision. In Alimentation des bovins, ovins et caprins, Tables INRA 2007, Editions Quae, Versailles, pp 149-179. Beauchemin K.A., Kreuzer M., O’Mara F. and McAllister T.A. (2008) Nutritional management for enteric methane abatement: A review. Australian Journal of Experimental Agriculture 48, 21–27. Beecher M., Baumont R., O’Donovan M., Boland T.M., Aufrère J., Fleming C., Galvin N. and Lewis E. (2014) Influence of pre-grazing herbage mass on voluntary intake and in vivo digestibility of perennial ryegrass in sheep. Animal Submitted. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 529 Beecher M., Hennessy D., Boland T.M., McEvoy M., O’Donovan M. and Lewis E. (2013) The variation in morphology of perennial ryegrass cultivars throughout the grazing season and effects on organic matter digestibility. Grass and Forage Science doi: 10.1111/gfs.12081 Benoit M, Tournadre H., Dulphy J.P., Prache S. and Cabaret J. (2009) Is intensification of ewe reproduction rhythm sustainable in an organically managed sheep production system? A four-year interdisciplinary study. Animal 3, 753-763. Briemle G. and Elsässer M. (1997) The functions of the Grünlandes. In: Reports on Agriculture, Volume, 75, 73 pp. Bruinenberg M.H., Valk H. and Struik P.C. (2003) Voluntary intake and in vivo digestibility of forages from seminatural grasslands in dairy cows. Netherlands Journal of Agricultural Science 51, 219-235. Coleman J., Pierce K. M., Berry D. P., Brennan A. and Horan B. (2009) The influence of genetic selection and feed system on the reproductive performance of spring-calving dairy cows within future pasture-based production systems. Journal of Dairy Science 92, 5258-5269. Cortes C., Damasceno J.C., Jamot J. and Prache S. (2006) Ewes increase their intake when offered a choice of herbage species at pasture. Animal Science 82, 183-191. Crosson P. and McGee M. (2011) Suckler beef production in Ireland: challenges and opportunities. Teagasc National Beef Conference, Cillin Hill, Kilkenny, 5 April, 2011. 11 pages. Curran J., Delaby L., Kennedy E., Murphy J.P., Boland T.M. and O'Donovan M. (2010) Sward characteristics, grass dry matter intake and milk production performance are affected by pre-grazing herbage mass and pasture allowance. Livestock Science 127, 144–154. Cutullic E., Delaby L., Gallard Y. and Disenhaus C. (2011) Dairy cows’ reproductive response to feeding level differs according to the reproductive stage and the breed. Animal, 5,731-740. Delaby L., Faverdin P., Michel G., Disenhaus C. and Peyraud J. L. (2009) Effect of different feeding strategies on lactation performance of Holstein and Normande dairy cows. Animal 3, 891-905. Delaby L., Horan B., O’Donovan M., Gallard Y. and Peyraud J.L. (2010) Are high genetic merit dairy cows compatible with low input grazing systems? Grassland Science in Europe 15, 928-930. Delaby L., Hennessy D., Buckley F. (2014) Animal choice for grass based systems. Grassland Science in Europe 19, this volume. Dewhurst R.J., Fisher W.J., Tweed J.K.S. and Wilkins R.J. (2003) Comparison of grass and legume silages for milk production. 1. Production responses with different levels of concentrate. Journal of Dairy Science 86, 2598– 2611. Dillon P., Crosse S., Stakelum G. and Flynn F. (1995) The effect of calving date and stocking rate on the performance of spring-calving dairy cows. Grass and Forage Science 50, 286-299. Dillon P., Roche J.R., Shalloo L. and Horan B. (2005) Optimising financial return from grazing in temperate pastures. In: Utilization of grazed grass in temperate animal systems, Proceedings of a satellite workshop of the XXth International Grassland Congress, Cork, Ireland (ed. J.J. Murphy). Wageningen Academic Publishers, The Netherlands, pages 131–148. Dumont B., Fortun-Lamothe L., Jouven M., Thomas M. and Tichit M. (2013) Prospects from agroecology and industrial ecology for animal production in the 21st century. Animal 7, 1028-1043. Duru M., Cruz P. and Theau J.P. (2010) A simplified method for characterising agronomic services provided by species-rich grasslands. Crop and Pasture Science 61, 420-433. Egan M.J., Lynch M.B. and Hennessy D. (2013) The influence of white clover inclusion in perennial ryegrass swards on milk and herbage production in a high N fertilizer system. In: Agricultural Research Forum, Tullamore, Ireland, 11-12th March 2013, page 81. Enriquez-Hidalgo D., Gilliland T.J., O’Donovan M., Elliott C. and Hennessy D. (2012) Effect of white clover inclusion in perennial ryegrass swards on herbage and milk production under frequent tight grazing conditions. In: Agricultural Research Forum, Tullamore, Ireland, 12-13th March, page 96. Falconer D.S. (1990) Selection in different environments: effects on environmental sensitivity (reaction norm) and on mean performance. Genetic Research 56, 57–70. Finneran E., Crosson P., O'Kiely P., Shalloo L., Forristal D. and Wallace M. (2012) Stochastic modelling of the yield and input price risk affecting home produced ruminant feed cost. Journal of Agricultural Science 150, 123139. Ginane C., Baumont R., Lassalas J. and Petit M. (2002) Feeding behaviour and intake of heifers fed on hays of various quality, offered alone or in a choice situation. Animal Research 51, 177-188. Grainger C. and Goddard M.E.( 2004) A review of the effects of dairy breed on feed conversion efficiency- An opportunity lost? In: “Animal Production in Australia” Pages 77-80. Proceedings of the 25th Bienal Conf. Austral. Soc. Anim. Prod., University of Melbourne, Victoria. CSIRO, Victoria, Australia. Griffon M. (2013) Qu’est-ce que l’agriculture écologiquement intensive? Editions Quae, Versailles, France, 220 p. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 530 Hochman Z., Carberry P.S., Robertson M.J., Gaydon D.S., Bell L.W. and McIntosh P.C. (2013) Prospects for ecological intensification of Australian agriculture. European Journal of Agronomy 44, 109-123. Hoden A., Peyraud J.L., Muller A., Delaby L. and Faverdin P. (1991) Simplified rotational grazing management of dairy cows: effects of rates of stocking and concentrate. Journal of Agricultural Science 116, 417-428. Holmes,C. W. (1995) Genoype x environment interactions in dairy cattle: A New Zealand perspective. In: T.L.J. Lawrence, F.J. Gordon and A. Carson (eds.) Breeding and Feeding the High Genetic Merit Dairy Cow. Occasional Publication No 19, British Society of Animal Science, pp. 51-58. Hoogendorn C.J., Holmes C.W. and. Chu A.C.P (1992) Some effects of herbage composition, as influenced by previous grazing management, on milk production by cows grazing on ryegrass/white clover pastures. 2. Milk production in late spring/summer: effects of grazing intensity during the preceding spring period. Grass and Forage Science 47, 316–325. Huyghe C., Baumont R. and Isselstein J. (2008) Plant diversity in grasslands and feed quality. Grassland Science in Europe 13, 375-386. Huyghe C., Litrico I. and Surault F. (2012) Agronomic value and provisioning services of multi-species swards. Grassland Science in Europe 17, 35-46. Jankowska-Huflejt H. (2006) The function of permanent grasslands resources protection. Journal of Water and Land Development 10, 55–65. Kirwan L., Luscher A., Sebastia M.T., Finn J.A., Collins R.P., Porqueddu C., Helgadottir A., Baadshaug O.H., Brophy C., Coran C., Dalmannsdottir S., Delgado I., Elgersma A., Fothergill M., Frankow-Lindberg B.E., Golinski P., Grieu P., Gustavsson A.M., Hoglind M., Huguenin-Elie O., Iliadis C., Jorgensen M., Kadziuliene Z., Karyotis T., Lunnan T., Malengier M., Maltoni S., Meyer V., Nyfeler D., Nykanen-Kurki P., Parente J., Smit H.J., Thumm U. and Connolly J. (2007) Evenness drives consistent diversity effects in intensive grassland systems across 28 European sites. Journal of Ecology 95, 530-539. Lee M.R.F., Winters A.L., Scollan N.D., Dewhurst R.J., Theodorou M.K. and Minchin F.R. (2004) Plant-mediated lipolysis and proteolysis in red clover with different polyphenol oxidase activities. Journal of the Science of Food and Agriculture 84, 1639-1645. Leip A., Weiss F., Wassenaar T., Perez I., Fellmann T., Loudjani P., Tubiello F., Grandgirard D., Monni S. and Biala K. (2010) Evaluation of the livestock’s sector contribution to the EU greenhouse gas emissions (GGELS) – Final report. European Comission, Joint Research Centre, Ispra. McCarthy B., Delaby L., Pierce K.M., Journot F. and Horan B. (2011) Meta-analysis of the impact of stocking rate on the productivity of pasture-based milk production systems. Animal, 5, 784-794. McCarthy B., Pierce K.M., Delaby L., Brennan A., Fleming C. and Horan B. (2013) The effect of stocking rate and calving date on grass production, utilization and nutritive value of the sward during the grazing season Grass and Forage Science 68, 364-377. McCaughey W.P., Wittenberg K. and Corrigan D. (1999) Impact of pasture type on methane production by lactating beef cows. Canadian Journal of Animal Science 79, 221–226. McEvoy M., O’Donovan M., Kennedy E., Murphy J.P. Delaby L. and Boland T.M. (2009) effect of pre-grazing herbage mass and pasture allowance on the lactation performance of Holstein-Friesian dairy cows. Journal of Dairy Science 92, 414-422. McEvoy M., Delaby L., Murphy J.P., Boland T.M. and O'Donovan M. (2010) Effect of herbage mass and allowance on sward characteristics, milk production, intake and rumen volatile fatty acid concentration. Grass and Forage Science 65, 335-347. McMeekan C.P. and Walsh M.J. (1963) The inter-relationships of grazing method and stocking rate in the efficiency of pasture utilization by dairy cattle. Journal of Agricultural Science 61, 147-166. Michaud A., Plantureux S., Pottier E. and Baumont R. (2014) Links between functional composition, biomass production and forage quality in permanent grasslands over a broad range of conditions. Journal of Agricultural Science (in press). Min B.R., Barry T.N., Attwood G.T. and McNabb W.C. (2003) The effect of condensed tannins on the nutrition and health of ruminants fed fresh temperate forages: a review. Animal Feed Science and Technology, 106, 3-19. Niderkorn V. and Baumont R. (2009) Associative effects between forages on feed intake and digestion in ruminants. Animal 3, 951–960. Niderkorn V., Baumont R., Le Morvan A. and Macheboeuf D. (2011) Occurrence of associative effects between grasses and legumes in binary mixtures on in vitro rumen fermentation characteristics. Journal of Animal Science 89, 1138–1145. Niderkorn V., Martin C. and Baumont R. (2014) Associative effects between forage species on intake and digestive efficiency in sheep. Grassland Science in Europe 19, this volume. Orr R.J., Parsons A.J., Penning P.D. and Treacher T.T. (1990) Sward composition, animal performance and the potential production of grass/white clover swards continuously stocked with sheep. Grass and Forage Science 45, 325–336. Penno J. (1999) Stocking rate for optimum profit. South Island Dairy Event 1, 25-43. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 531 Peyraud J.L., Van den Pol-van Dasselaar A., Dillon P. and Delaby L. (2010) Producing milk from grazing to reconcile economic and environmental performances. Grassland Science in Europe 15, 865-879. Pottier E., Tournadre H., Benoit M. and Prache S. (2009) Maximizer la part du pâturage dans l’alimentation des ovins : intérêt pour l’autonomie alimentaire, l’environnement et la qualité des produits. Fourrages 199, 349-371. Prache S., Benoit M., Tournadre H., Cabaret J., Laignel G., Ballet J., Thomas Y., Hoste H., Pellicer M., Andueza D., Hostiou N., Giraud J.M. and Sepchat B. (2011) Plateforme INRA de recherches en production ovine allaitante AB : de l’étude de verrous techniques à la conception de systèmes d’élevage innovants. Rencontres Recherches Ruminants 18, 61-64. Prendiville R., Pierce K.M. and Buckley F. (2009) An evaluation of production efficiencies among lactating Holstein-Friesian, Jersey and Jersey × Holstein-Friesian cows at pasture. Journal of Dairy Science 92, 6176-6185. Pretty J.N. (1997) The sustainable intensification of agriculture. Natural Resources Forum 21, 247–256. Rattray P.V. (1987) Sheep production from managed grasslands. In: Snaydon, R.W. (ed.) Managed grasslands pp.113-122. Ribeiro Filho H.M.N., Delagarde R. and Peyraud J.L. (2003) Inclusion of white clover in strip-grazed perennial ryegrass swards: herbage intake and milk yield of dairy cows at different ages of sward regrowth. Animal Science 77, 499–510. Soder K.J., Sanderson M.A., Stack J.L. and Muller L.D. (2006) Intake and performance of lactating cows grazing diverse forage mixtures. Journal of Dairy Science 89, 2158-2167. Theodoridou K., Aufrère J., Andueza D., Pourrat J., Le Morvan A., Stringano E., Mueller-Harvey I. and Baumont R. (2010) Effects of condensed tannins in fresh sainfoin (Onobrychis viciifolia) on in vivo and in situ digestion in sheep. Animal Feed Science and Technology 160, 23-38. Theodoridou, K., Aufrère J., Andueza D., Pourrat J., Picard F., Le Morvan A. and Baumont R. (2012) Effect of condensed tannins in wrapped silage bales of sainfoin (Onobrychis viciifolia) on in vivo and in situ digestion in sheep. Animal 6, 245-253. Thomas C., Gibbs B.G. and Tayler J.C. (1981) Beef production for silage. 2. The performance of beef cattle given silages of either perennial ryegrass or red clover. Animal Production, 32, 149–153. Tunon G.E. (2013) Improving the use of perennial ryegrass swards for dairying in Ireland. PhD thesis. Massey University, Palmerston North, New Zealand Ulyatt M.J. (1970) Evaluation of pasture quality under New Zealand conditions. Proceedings of the New Zealand Grassland Association 32, 61–68. Waghorn G.C., Tavendale M.H. and Woodfield D.R. (2002) Methanogenesis from forages fed to sheep. Proceedings of the New Zealand Grassland Association 64, 167–171. Wims C.M., Deighton M.H., Lewis E., O’Loughlin B., Delaby L., Boland T.M. and O’Donovan M. (2010) Effect of pre-grazing herbage mass on methane production, dry matter intake and milk production of grazing dairy cows during the mid season period. Journal of Dairy Science 93, 4976–4985. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 532 Theme 4 submitted papers Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 533 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 534 Plant or animal needs - how to determine the optimal N intensity of grassland? Herrmann A., Techow A., Kluß C. and Taube F. Grass and Forage Science/Organic Farming, Kiel University, Hermann-Rodewald-Strasse 9, 24118 Kiel, Germany Corresponding author: aherrman@email.uni-kiel.de Abstract The impact of six plant- and animal-related indicators for deriving the optimal N intensity of grassland was studied, based on a 3-year, multi-site field trial. Crude protein (CP) content was a limiting factor for grass-only diets, whereas for mixed rations, an appropriate share of maize could balance even high CP of grass, allowing the exploitation its yield potential. Keywords: N fertilization, dry matter yield, N yield, crude protein, rumen N balance Introduction For high-yielding dairy cows dietary CP requirements are assumed to range between 14 and 18% (Pacheco and Waghorn, 2008). High CP content of grass silage may cause problems when paralleled by a high share of non-protein N, moderate energy concentration and if grass is fed as sole diet (Givens and Rulquin, 2004). If maize makes a significant portion of the diet the 18% CP threshold might be questioned and the N fertilizer recommendation might need reconsideration. The aim of the present study therefore was to derive the optimal N intensity of cut-only grassland when applying plant- and animal-related indictors. Materials and methods The study is based on a 2-factorial (site, N fertilization) randomized block experiment with 4 replicates conducted over three years (2009-2011) at five permanent grassland sites (sward age > 4 yrs, > 60% Lolium perenne) throughout Germany (Aulendorf, Baden-Württemberg; Eichhof, Hesse; Iden, Saxony-Anhalt; Riswick, North Rhine-Westphalia; Spitalhof, Bavaria). Nitrogen treatment comprised 6 levels (0 N-grass without clover; 0 N-white clover grass; 120, 240, 360, 480 kg N ha-1) applied as calcium ammonium nitrate in four dressings, i.e. in early spring and after each of the first three cuts. Six indicators were investigated to derive the optimal N input (mineral N + N from fixation): (i) three dry matter (DM) response functions (linearplus-plateau, quadratic-plus-plateau, exponential function; the latter with a marginal yield response of 10 kg DM (kg N)-1), (ii) the CP content to represent the animal need, assuming an optimal range of 14-18%, (iii) the N yield, and (iv) the N uptake efficiency integrating crop and environment needs, with a target of unity, corresponding to a N surplus of zero. N fixation was estimated according to Hoegh-Jensen et al. (2004) as a function of DM yield, clover proportion and soil type. Forage quality traits were estimated by NIRS and rumen N balance (RNB) was calculated according to GfE (1997). The relationship between N input and indicator values were investigated by first estimating function parameters separately for each siteyearreplicate combination and in a second step analysing the impact of site on a given parameter by an analysis of variance with site assumed fixed and year random. Results and discussion Maximum DM yield attained at the different sites was similar at 13-16 t DM ha-1 (Fig. 1). The tested DM response functions showed similar model fit (R² > 0.6). Thus no model could be favoured over any other. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 535 15 DM yield (Mg ha-1) e e q l 10 year 2009 2010 2011 5 Riswick l q e-function quadratic + plateau linear + plateau optimal Ninput (kg N ha-1) Spitalhof 15 e q l l e l q q 10 e 5 Eichhof Aulendorf 0 200 400 600 0 200 Iden 600 0 400 200 400 600 N input (kg N ha-1) Figure 1. DM response functions obtained for the different grassland sites. Depending on the function type, the optimal N input varied between 225 and 386 kg N ha-1, with the e-function tending to estimate lower N input than the two others (Table 1). Table 1. Optimal N input (kg N ha-1, mineral N and N from fixation), depending on the site and indicator applied. DM yield, linear-plus-plateau DM yield, e-function and marginal yield DM yield, quadratic-plus-plateau and marginal yield N yield, linear-plus-plateau N uptake efficiency CP content (18% threshold) Riswick 284 236 Spitalhof 386 274 Aulendorf 360 225 Eichhof 345 280 Iden 359 343 279 313 258 324 328 354 418 155 388 451 149 348 374 32 427 349 199 424 331 223 Nitrogen yield increased linearly before reaching a plateau at an N input of 350 to over 400 kg N ha-1. Correspondingly, the N uptake efficiency decreased almost linearly (data not presented). The N yield thus can be exploited only at a much higher N input than required for DM yield. Crude protein content showed a strong increase in N input and achieved weighted annual averages of up to 22%. The needs of the animal, as indicated by a CP content of 14-18%, thus could only be complied with when substantially reducing N input. This in turn would cause a considerable yield loss, e.g. 15-52% for the quadratic-plus-plateau function, and promote undesirable changes of sward composition. The DM yield potential was virtually exploited at an annual average CP content of 20%. According to GfE (1997), a RNB of 50 is tolerable in the total ration. For dairy cows with moderate dry matter intake (13 kg DM d-1) solely by grass silage with low CP content (grass silage 1, Table 2), a RNB of 50 could be complied with. The RNB threshold, however, will be exceeded substantially for grass silages with CP noticeably above 18% (grass silages 3 and 4, Table 2), resulting in increased N excretion and reduced N-use efficiency. The situation is entirely different if grass and maize silage are fed together, as exemplified in Table 2 for different grass/maize mixing ratios all achieving a RNB of zero. For grass silages with 16% CP a maize ratio of 30% is sufficient to balance RNB. With a higher maize share, as typical for many German dairy farms, grass silages with CP ≥ 20% can be compensated for. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 536 Table 2. Protein value of model grass silages with varying CP contents and grass/maize rations. ME: metabolizable energy (MJ ME (kg DM)-1), CP: crude protein (g (kg DM) -1), RUP: feed CP escaping degradation in the rumen, RUPCP: RUP share (%) of CP, uCP : utilizable CP (g (kg DM)-1), RNB: rumen N balance (g (kg DM)-1). uCP: sum of microbial CP and RUP; RNB: ruminal N balance, RNB = (CP-uCP)/6.25 Mixture Grass silage 1 Grass silage 2 Grass silage 3 Grass silage 4 Maize silage GS1:MS GS2:MS GS3:MS GS4:MS ME 10.5 10.5 10.5 10.5 11.0 CP 160 180 200 240 70 RUPCP 15 15 15 15 25 uCP 134 137 140 146 134 0.70 : 0.30 0.58 : 0.42 0.50 : 0.50 0.35 : 0.65 RNB 4.1 6.8 9.6 15.0 -9.8 -0.1 0.0 -0.1 0.3 The suitability of CP as indicator thus is depending on the ration. For high proportions of grass silage and low concentrate use, a threshold of 18% seems reasonable. When feeding diets with combinations of low grass but high proportions of maize and concentrates, higher grass CP contents can be easily compensated. When assuming such conditions for the tested sites, the optimal N fertilisation can be derived from one of the three DM response functions and the estimated N fixation. We used the quadratic-plus-plateau function since the functions did not differ statistically. With optimal N input ranging between 258 (Aulendorf) and 328 kg N ha-1 (Iden), and N fixation between 0 (Riswick) and 184 kg N ha-1 (Aulendorf), optimal N fertiliser input would range between 74 (Aulendorf) and 322 kg N ha-1 (Iden). Such grass swards require an efficient ensiling management in order to keep the protein quality on a high level. A crucial measure is a fast and extensive wilting, since protein degradation by plant proteases and microorganisms declines with increasing DM content (Edmunds et al., 2014). Conclusion The optimal N intensity of grassland varies substantially depending on the indicator applied. In terms of the availability of silage maize provided, high CP contents will cause little problems and the yield potential of grassland can be exploited. Thus, better use can be made of grassland as protein source. This needs refinement of fertilizer recommendations; in particular a greater differentiation is required with respect to the grassland use (grazing vs. cutting) and ration, as well as a better estimation of N fixation. Acknowledgements Data provision by the federal states’ extension services is gratefully acknowledged. References Edmunds B., Spiekers H., Südekum K.-H., Nussbaum H., Schwarz F.J. and Bennett R. (2014) Effect of extent and rate of wilting on nitrogen components of grass silage. Grass and Forage Science 69, 140-152. GfE, Gesellschaft für Ernährungsphysiologie (1997) Zum Proteinbedarf von Milchkühen und Aufzuchtrindern. Proceedings of the Society of Nutrition Physiology 6, 217-232. Givens D.I. and Rulquin H. (2004) Utilisation by ruminants of nitrogen compounds in silage-based diets. Animal Feed Science and Technology 114, 1-18. Høgh-Jensen H., Loges R., Jørgensen F.V., Vinther F.P. and Jensen E.S. (2004) An empirical model for quantification of symbiotic nitrogen fixation in grass-clover mixtures. Agricultural Systems 82, 181-194. Pacheco D. and Waghorn G.C. (2008) Dietary nitrogen – definitions, digestion, excretion and consequences of excess for grazing ruminants. Proceedings of the New Zealand Grassland Association 70, 107-116. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 537 Energy expenditure of two grazing Holstein cow strains Schori F.1, Thanner S.1,2, Görs S.3, Metges C.C.3, Bruckmaier R.M.2 and Dohme-Meier F.1 1 Agroscope, Institute of Livestock Science ILS, Tioleyre 4, 1725 Posieux 2 Veterinary Physiology, Vetsuisse Faculty, University of Bern, Bremgartenstr. 109a, 3001 Bern, Switzerland 3 Leibnitz Institute for Farm Animals Biology (FBN), Institute of Nutritional Physiology, 'Oskar Keller', Wilhelm-Stahl-Allee 2, 18196 Dummerstorf, Germany Corresponding author: fredy.schori@agroscope.admin.ch Abstract The study compared the energy expenditure (EE) of two Holstein cow strains in an organic grazing system. Twelve Swiss Holstein-Friesian (HCH) and 12 New Zealand Holstein-Friesian (HNZ) cows in the third stage of lactation were kept in a rotational grazing system without concentrate supplementation. Using the 13C bicarbonate dilution technique in combination with an automatic blood sampling system, EE based on measured CO2 production was determined during 6 h per d. The HCH cows were heavier and had a lower body condition score (BCS) compared to HNZ cows. Milk production and grass intake did not differ between the two cow strains, but HCH cows grazed for a longer time during the 6 h measurement period and performed more grazing mastication than the HNZ cows. No difference was found between the two cow strains with regard to EE per metabolic body weight (BW0.75), mainly due to a high between-animal variation. As efficiency and energy use are important in pasture-based dairy systems, the determining factors for EE and its variation, respectively, should be investigated in more detail in future studies. Keywords: Energy expenditure, pasture, dairy cows, Holstein-Friesian Introduction The economic benefit of pasture-based dairy systems requires the efficient use of pasture herbage and is also reasonable milk production per cow (Dillon et al., 2005). By limiting herbage allowance to improve the efficient use of pasture herbage per area, the herbage intake, and therefore the energy uptake per cow would be reduced (Delagarde et al., 2001). On the other hand, EE under grazing conditions is increased. For example, Bruinenberg et al. (2002) found that grass-fed dairy cows in the barn have a 10% higher metabolizable energy requirement for maintenance. Physical activity also influences EE, e.g. grazing cows had 21% higher EE compared to grass-fed cows kept indoors (Kaufmann et al., 2011). The energy requirements relative to maintenance may increase up to 50% depending on grazing conditions, including herbage availability (allowance and mass) and digestibility, distances walked (to the milking parlour and watering points), weather, topography, and interaction between these factors (CSIRO, 2007). In New Zealand, Holstein cows are bred for the specific needs of pasture-based, low-input dairy production, including selection for milk solids, lower body weight (BW), fertility, and longevity (Miglior et al., 2005). Compared to Swiss Holstein cows, the NZ Holstein cows had a lower BW, showed different BCS, ruminated longer, and tended to take more steps on the pasture (Schori and Münger, 2010). The objective of the current study was to determine EE based on CO2 production of HNZ cows and heavier, high-producing HCH cows in an organic, full-time grazing system without concentrate supplementation. To explain possible differences in EE between strains with differences in grazing behaviour or physical activity, these variables were recorded simultaneously. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 538 Materials and methods The experiment took place on the organic farm 'Ferme de l’Abbaye' in Sorens (824 m a.s.l., Switzerland) on the second half of September 2011. Twelve matched pairs of HCH and HNZ cows were formed according to the following criteria: number of lactations (2.6 ± 1.8), weeks in milk (28 ± 0.6 week), and age for primiparous cows (35 ± 3.9 month). All 24 experimental cows were separated from the rest of the lactating herd and received no hay or concentrate supplementation. Access to fresh water, minerals and common salt were provided. The experiment consisted of two consecutive experimental weeks. Cow pairs were equally divided between the 2 weeks so that each cow underwent a period of 7 d of data collection. Cows grazed from 08 to 14 h and from 18 to 4.30 h. The paddocks were long-established swards, predominantly of grasses, and were rotationally grazed based on sward height measured with an electronic rising plate meter (Jenquip, Feilding, NZ) for 1 to 3 d. The average pre-grazing sward height was 6.4 (±0.7) cm, and the average post-grazing sward surface height was 4.4 (±0.6) cm. During the experiment, the average temperature was 15.4 (min. 9.3, max. 18.1) C. Milk yield (7 d) and composition (3 d), BW (7 d), BCS (before and after the experiment) were measured during the respective experimental week. Individual dry matter intake (DMI) was estimated using the n-alkane double-indicator technique. Grazing and ruminating behaviour was recorded on 3 consecutive d using an automatic jaw-movement recorder with a pressure sensor (Datenlogger MSR145, MSR Electronics GmbH, Hengart, Switzerland). Physical activity, including time spent standing, lying and walking, and numbers of steps, were determined on 3 d using the IceTagTM pedometer (IceRobotics Ltd., Edinburgh, UK). The 13C bicarbonate dilution technique combined with an automatic blood sampling system was used to determine EE (Kaufmann et al., 2011). Between 07.45 and 13.45 h the CO2 production of one cow pair was measured. Data were analysed using a linear mixed model (SYSTAT 12). Results The HCH cows had a greater BW (615 vs. 567 kg, P = 0.01) but a lower BCS (2.54 vs. 2.84, P = 0.01) and showed the same BW losses (96 vs. 112 g/d, P = 0.91) compared to HNZ cows. Neither milk yield (18.8 vs. 17.5 kg, P = 0.31) nor energy corrected milk yield (ECM; 18.3 kg, P = 0.96) differed between the cow strains. The milk fat (4.01 vs. 4.49 %, P = 0.03) and protein (3.26 vs. 3.65, P = 0.001) content were lower for HCH than for HNZ cows. Similar amounts of DM (16.5 vs. 16.3 kg/d, P = 0.89) were consumed. The production efficiency measures, namely ECM produced per 100 kg of BW0.75 (14.8 vs. 15.7 kg, P = 0.25), ECM produced per kg of DMI (1.12 vs. 1.13, P = 0.85), and DMI per 100 kg of BW0.75 (13.3 vs. 14.0 kg, P = 0.41) were not affected by cow strain. During the 6 h of blood sampling for EE determination, HCH cows spent more time grazing (235 vs. 213 min., P < 0.001) and performed more grazing mastication (17514 vs. 15634, P = 0.001) than the HNZ cows. Ruminating behaviour did not differ between the two cow strains. Cow strain had no effect on time spent standing (280 vs. 281 min., P = 0.92) lying (80 vs. 79 min., P = 0.92) or on time spent walking (109 vs. 95 min., P = 0.34) and the number of steps (1186 vs. 1106, P = 0.43). Cow strain did not affect EE per kg BW0.75 over 6 h (309 vs. 273 kJ, P = 0.27). HCH were heavier, but EE per cow per d (152 vs. 127 MJ, P = 0.13) did not differ between cow strains. Discussion Energy expenditure measured with the 13C bicarbonate dilution method includes the metabolizable energy costs for maintenance (MEm, including energy for fasting metabolism and activity allowance, measured under thermoneutral conditions), as well as heat increment changes (heat losses, concrete differences between metabolizable energy and net energy for maintenance, production, and gestation). Although HNZ cows, in contrast to HCH cows, were selected indirectly for feed-conversion efficiency by considering BW in the breeding worth, Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 539 and thus intake capacity, as well as MEm in relation to the production of milk solids, equal ECM per kg DMI and EE per kg BW0.75 per d values were found. Possible reasons could be that, compared to earlier studies (Schori and Münger, 2010), an increasing similarity in the BW of the two cow strains was found. Furthermore, the variability of EE between animals was high (CV = 26%) and clearly higher compared to a CV of 17% as observed by Kaufmann et al. (2011). The larger variation may be explained by the use of different cow strains, the larger variation of the length of the path to the paddocks, and the topography of the pastures in the foothills. Under controlled activity conditions, Van Es (1961) found that the MEm of cows of the same breed and similar size may vary by as much as 8-10%. It is assumed that a major contributor to the variation in fasting EE is the genetically inherited amount of organ mass. Furthermore, Van Es (1961) discussed not only the between-animal variation but also analytical and physiological variation, as well as variation in the composition of rations as additional effects that can contribute to the variation in MEm. Moreover, under practical grazing conditions, variability is even greater compared to respiration chambers since, among others, herbage availability and quality, climate, and the physical activity of dairy cows also varies. Conclusion Similar production levels in late lactation, small differences in BW and physical activity of the two Holstein cow strains led to similar values of EE per kg BW0.75 in both strains. The high variability suggests that there is potential to improve the efficient use of consumed energy. Furthermore, the determining factors for EE of grazing cows should be investigated in more detail. References Bruinenberg M.H., van der Honing Y., Agnew R.E., Young F.J., van Vuuren A.M. and Valk H. (2002) Energy metabolism of dairy cows fed on grass. Livestock Production Science 75, 117-128. CSIRO (2007) Nutrient requirements of domesticated ruminants. CSIRO Publishing, Collingwood, AU. Delagarde R., Prache S., D’Hour P. and Petit M. (2001) Ingestion de l’herbe par les ruminants au pasturage. Fourrages 166, 189-212. Dillon P., Roche J. R., Shalloo L. and Horan B. (2005) Optimising financial returns from grazing in temperate pastures. In: Murphy J.J. (ed) Utilisation of Grazed Grass in Temperate Animal Systems. Proceedings Satellite Workshop, 20th International Grassland Congress, Cork, Ireland. Wageningen Academic Publishers, Wageningen, NL, pp. 131-147. Kaufmann L. D., Münger A., Rérat M., Junghans P., Görs S., Metges C. C. and Dohme-Meier F. (2011) Energy expenditure of gazing cows and cows fed grass indoors as determined by the 13C bicarbonate dilution technique using an automatic blood sampling system. Journal of Dairy Science 94, 1989-2000. Miglior F., Muir B. L. and Van Doormaal B. J. (2005) Selection indices in Holstein cattle of various countries. Journal of Dairy Science 88,1255-1263. Schori F. and Münger A. (2010) Grazing behaviour and intake of two Holstein cow types in a pasture-based production system. Grassland Science in Europe 15, 895-897. Van Es A. J. H. (1961) Between-animal variation in the amount of energy required for the maintenance of cows. Landbouwhogeschool Wageningen, NL, 116 pp. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 540 Effects of mechanically separated dairy cow slurry on grazing performance Henry C.A.1,2, Lee M.A.1, McConnell D.A.3, Wood B.L.2 and Roberts D.J.1 1 SRUC Dairy Research Centre, Crichton University Campus, Dumfries, United Kingdom; 2 Dumfries Campus, University of Glasgow, Crichton University Campus, Dumfries DG1 4ZL, United Kingdom; 3. DairyCo, Agriculture and Horticulture Development Board, Stoneleigh Park, Kenilworth, Warwickshire, CV8 2TL, United Kingdom. Corresponding author: Chris.Henry@sruc.ac.uk Abstract Inorganic fertilizers are widely used to fertilize grasslands in dairy systems. Increasing the nutrient-use efficiency of slurry applied as an alternative or supplementary fertilizer could reduce the volume of purchased fertilizer. This would reduce costs and improve security of future fertilizer supply since slurry is produced on-farm. Slurry separation could increase the potential for fertilizing grazed grassland with slurry. This experiment compared the effects on milk yields of fertilizing grazing swards with the liquid fraction of separated slurry, whole slurry or inorganic fertilizer. Three groups of 12 core cows were grazed for two 24-day rotations on 2-day paddocks between 18 and 24 days after fertilizer treatment application in a put-and-take design. Nutrient-use efficiency was significantly higher (P<0.05) from the liquid fraction of separated slurry (12.2 kg kgN-1) compared to whole slurry (8.0 kg kg N-1) application. This effect was most likely due to a combination of improved soil infiltration and reduced sward contamination. Keywords: slurry separation, grazing, dairy Introduction Increasing yields from grazed grasslands using sustainably sourced fertilizers represents a major opportunity if we are to sustainably maintain, or even increase levels of food production. Currently, slurry can be applied to grazed grasslands as a fertilizer, increasing the availability of essential plant nutrients (e.g. nitrates and ammonium) in the rhizosphere. However, inorganic fertilizers continue to be applied by farmers, since slurry contains a high proportion of dry organic matter which becomes assimilated into soils more slowly than inorganic fertilizers, resulting in lower grass yields (Møller et al., 2000). Separation of slurry into the liquid fraction prior to application may therefore benefit farmers by reducing dry matter (DM) content and increasing rates of soil infiltration; hence increasing herbage yields due to greater concentrations of nutrients being available to grass roots in solution (Møller et al., 2000). Reducing the DM content may also enhance the potential for using slurry on grazing ground due to the lower sward contamination of more liquid slurries (Rodhe, 2003), reducing the risk of sward rejection by grazing cattle. This is of interest given that cattle grazing performance can be adversely affected by slurry applications on grazing pasture (Gjestang et al., 1984). Previous research investigating the use of slurry on grazing ground has shown that applying slurry by shallow injection can improve the resultant grazing performance relative to splashplate application (Laws et al., 1996). Likewise, Dale et al. (2012) showed that inorganic fertilizer inputs can be reduced by replacing a portion with cattle slurry applied by trailing-shoe without adversely affecting dairy cow performance. This experiment compared the effects of applying separated slurry or unseparated slurry (without the addition of inorganic fertilizer) or inorganic fertilizers on grass yields, milk yields and nutrient-use efficiency. Materials and methods The experiment was conducted during the summer of 2013 at the SRUC Dairy Research Centre, Dumfries, Scotland (UK Ordnance Survey map ref NX981732). Three 1-ha paddocks were Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 541 established in each of four fields. Two of the fields were dominated by perennial ryegrass (Lolium perenne) and two were dominated by Italian ryegrass (Lolium multiflorum) (approx. coverage of dominant species greater than 95%). Each paddock in each field was then allocated one of three fertilizer treatments: whole (unseparated) slurry (W), liquid fraction of separated slurry (L) and ammonium nitrate fertilizer (F). Each treatment paddock was further subdivided into three 0.33 ha sub-paddocks with the aim of each sub-paddock providing two days grazing for twelve cows (with a herbage allowance of 15 kg DM cow-day-1). All fields were cut for silage in late May. Fields 1-3 were grazed initially on the silage aftermath and Field 4 was cut and baled before grazing. Dairy cow slurry was separated using a Sperrin dual cylinder separator and applied using a dribble bar approach at 24 m3 ha-1 (liquid) and 27.5 m3 ha-1 (whole). The different slurry rates, and an ammonium nitrate control, were calculated so that equal concentrations of available N (NO2- + NO3- + NH4+) were added to each treatment, according to standard values (Defra, 2010). Three groups of twelve mid- to late-lactation dairy cows were grazed on the treated pasture for two 24-day rotations, under a three-times-a-day milking regime, with 0.5 kg cow-1 of concentrates being fed at each milking. Additional cows of a similar yield and weight were added to the groups when required to provide a target herbage availability of 15 kg DM cow-day-1. Statistical analysis was carried out using R i386 3.0.2 (R Core Team, 2013). Differences between the composition of W, L and F were tested using student’s t-tests. Treatment differences for herbage yield, milk yield and nitrogen-use efficiency (yield / available N) were tested using Analysis of Variance tests (ANOVA) and Tukey’s Honestly Significant Difference (HSD) tests. Results and discussion Slurry separation decreased the mean DM content of the slurry from 52±0.8 g kg-1 for W to 28±0.3 g kg-1 for L (P<0.001), but also reduced the mean concentration of available N from 1.3±0.02 g kg-1 to 1.1±0.01 g kg-1, respectively (P<0.001), contradicting the standard values used to calculate the application rates (1.2 g kg-1 for W and 1.5 g kg-1 for L). The L fraction was therefore applied at a lower total rate over the experiment and had a lower concentration of available N, resulting in available N application rates of 68.9, 50.2 and 71.8 kg ha-1 for W, L and F, respectively. The stocking densities were 36.4±1.6 LU ha-1 for W, 36.4±1.4 for L and 45.8±1.8 LU ha-1 for F (P<0.001), showing that more grass was grown by F even though it was balanced with W and L by available N. Daily milk yields were not affected by treatment at the cow or area (stocking density x yield) level (P>0.05). This suggests that grazing has not been adversely affected by using slurry as a fertilizer, regardless of whether it has been separated or not. However, the nutrient-use efficiency (yield / available N applied) was significantly affected by treatment (Table 1). For every kg of available N applied, L produced 52% more milk than W. Greater productivity is most likely due to improved soil infiltration resulting in a greater concentration of nutrients available to roots in solution for L than W, coupled with lower sward contamination for L than W, in line with the findings of Rodhe (2003). The lower rate of available N application for L than W may account for some of the increase in efficiency, due to the non-linearity of N response curves. However, had the available N been equal between the treatments and the increased efficiency remained, there may have been a significant effect of separation on milk yield at the area level. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 542 Table 1. Mean daily milk and fat and protein yields. W = whole (unseparated) slurry, L = liquid fraction of separated slurry, F = ammonium nitrate fertilizer Daily milk yield (kg cow-1) Daily fat + protein yield (kg cow-1) Daily milk yield by area (kg ha-1) Daily milk yield by available N (kg kgN-1) NS, not significant; *P<0.05. W Mean 15.1 1.14 550 8.0a s.e. 0.2 0.02 8 0.1 L Mean 16.8 1.19 612 12.2b s.e. 1.0 0.07 35 0.7 F Mean 16.0 1.13 733 10.2ab sig. s.e. 1.5 0.10 66 0.9 NS NS NS * Conclusions The L treatment increased nutrient-use efficiency relative to W. This may be due to higher nutrient availability from L than W due to increased soil infiltration and smaller particle size. However, care needs to be taken in extrapolating these results since L was applied at a lower rate. Daily milk yield per cow was not affected by treatment; hence grazing performance does not appear to have been reduced by using either separated or whole slurry as a fertilizer relative to inorganic fertilizer. The results demonstrate the potential for growing grass for grazing from slurry alone, and suggest that separation may have a place in certain systems although more work is required to establish the economic implications. Acknowledgements This work was funded by DairyCo. References Dale A.J., Ferris C.P., Frost J.P., Mayne C.S. and Kilpatrick D.J. (2012) The effect of applying cattle slurry using the trailing-shoe technique on dairy cow and sward performance in a rotational grazing system. Grass and Forage Science 68, 138-150. Defra (2010) Fertiliser Manual (RB209) 8th Edition, https://www.gov.uk/government/publications/fertilisermanual-rb209 (accessed 12 December 2013). Gjestang K.E., Tjernshaugen O. and Tveitnes S. (1984) Grazing behaviour by heifers on pastures fertilized with different fertilizer categories. Applied Animal Behaviour Science 12, 33-41. Laws J.A., Rook A.J. and Pain B.F. (1996) Diet selection by cattle offered a choice between swards treated or untreated with slurry: effects of application method and time since application. Applied Animal Behaviour Science 48, 131-142. Møller H.B., Lund, I. and Sommer S.G. (2000) Solid-liquid separation of livestock slurry: efficiency and cost. Bioresource Technology 74, 223-229. R Core Team (2013) R: A language and environment for statistical computing. R Foundation for statistical Computing, Vienna, Austria. URL http://www.R-project.org/. Rodhe L. (2003) Methods for determining the presence of slurry on the crop and in the upper soil layer after application to grassland. Bioresource Technology 90, 81-88. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 543 Effect of growth stage on the phosphorus content of grass, and on phosphorus excretion on dairy farms Van Middelkoop J.C., Holshof G. and Plomp M. Wageningen UR Livestock Research, Lelystad, The Netherlands Corresponding author: jantine.vanmiddelkoop@wur.nl Abstract To protect the environment in the Netherlands, legal limits for P and N application in manure and mineral fertilizer have been established. Because phosphorus (P) in grass silage is an important source of P excretion on dairy farms in the Netherlands, decreasing the P content in grass can prevent manure export costs. The P content of grass decreases when it is cut at a later growth stage. However, feeding values also decrease, resulting in decreasing milk production or increasing use of manufactured feed. It is hard to predict the final effect of cutting grass at a later growth stage on P excretion at the farm level. Therefore we quantified the relation between growth stage and P content of grass and P excretion at the dairy farm level. The relation between P content of grass and growth stage was determined from the results of three grassland experiments. This relation was used in simulations with three farms in the farm model DairyWise. The P intake in roughages of dairy cows decreased when grass was cut at a later stage, but intake of manufactured feed increased. The largest decrease of manure export was reached at zero grazing (100% grass silage): 4.4 m3 ha-1 manure ( ≈ €66 ha-1). On all farms mowing costs decreased about €70 ha-1 to €115 ha-1. Keywords: phosphorus, grassland, P content, excretion, growth stage, simulation Introduction In regions with intensive animal husbandry, excessive use of phosphorus (P) and nitrogen (N) in the past has led to negative effects on the environment. To protect the environment in the Netherlands, legal limits for P and N application in manure and mineral fertilizer have been established, based on excretion standards. Dairy farmers have the possibility to prove lower excretion of their cattle by analysing and measuring roughage stocks. Because P in grass silage is an important source of P excretion, decreasing the P content in grass can prevent manure export costs. It has been shown that the P content of grass decreases when it is cut at a later growth stage (Wilson and McCarric, 1967; Fleming and Murphy, 1968; Whitehead, 2000). Feeding value, however, in terms of energy and protein content, also decreases. At the farm level, a result of cutting at a later growth stage could mean decreasing milk production, or increased use of manufactured feed. Because milk production and the feeding ration influence P excretion, it is hard to predict the final effect at the farm level of cutting grass at a later growth stage. The objective of this study is to quantify at the dairy-farm level, the relation between growth stage and P content of grass, and the effect on excretion, of cutting grass at a later growth stage than usual. Materials and methods Between 1999 and 2007 three grassland experiments, nine experimental years, took place (Table 1). The experiments were originally not designed for our study but provided useful data. In the experiments, dry matter (DM) yield and P content were determined every two weeks. The cuts were taken from new plots with the same treatments, in two replicates. The sward was dominated by Lolium perenne. Data were statistically analysed with Restricted Maximum Likelihood (ReML) technique (Harville, 1977), using Genstat (v.15). ReML fits a random and a systematic model to the data. The starting model included soil type, number of growing days, Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 544 N and P fertilization, starting day of growth and all interactions. Random model included experiment and year. Non-significant interactions (P≤0.05) were deleted. Table 1. Grassland experiments to quantify the relation between P content and growth stage, range of N and P fertilization, and dry matter yield per cut. Range per cut, kg ha-1 Experi ment Exp years Soil type Treatments No. of records N fertilization P2O5 fertilization DM yield 1 1999 2000 Sand, Clay Mineral P fertilization, +/– dairy manure, growth stage 1120 0-150 0-50 60-7700 2 2002 2003 Peat Mineral N growth stage fertilization, 1728 0-90 0-80 170-10000 3 2005 2006 2007 Peat Mineral N fertilization, groundwater level, growth stage 1495 0-120 0-45 65-10000 The developed statistical model was included in the farm model DairyWise (Schils et al., 2007) which simulates technical aspects and economics of dairy farms. Technical aspects are: grassland fertilization, growth, quality and use (grazing and mowing), consumption by the dairy herd and milk production. In the model the relation between feeding values of grass and growth stage is quantified. The economic results are calculated from costs and benefits. Using DairyWise, three model farms were calculated (Table 2), with cutting at two growth stages: normal (N: 3 t DM ha-1) and late (L: 4 t DM ha-1) cuts were taken for silage. Cuts for grazing and milk production were not varied. Table 2 Characteristics of three model farms Farm Farm area (ha) Grazing method Additional feeding during grazing period Feeding during housing period Milk production (t ha-1) 1 60.0 Day and night 0 100% grass silage 14.3 2 55.7 Day 8 kg silage 50% grass silage – 50% maize silage 15.4 3 60.0 zero --- 100% grass silage 14.3 DM/day maize Soil type: sand. Milk production per cow: 8600 kg yr -1 Results and conclusions In the statistical analysis of the grassland experiments it was proven that an effect of cutting grass at a later growth stage was a lower P content. The P content on a certain harvest date was higher if the N fertilization was higher. On peat the P content was highest, on clay lowest. Calculation with the model, using estimated factors, showed that in, e.g., a cut that started to grow at 1st March, on sand, fertilized with 120 kg N ha-1, the decrease of P content was about 0.028 g P kg-1 DM day-1. In the same situation, the decrease on peat was 0.035 g P kg-1 DM day-1 and on clay it was 0.014 g P kg-1 DM day-1. On sand fertilized with 80 kg N ha-1 the decrease was 0.025 g P kg-1 DM day-1. In calculations with the model Dairy Wise (Figure 1) results showed that the P intake of dairy cows that was obtained from roughages decreased on all farms when the grass was cut at a later growth stage. Part of this decrease, however, was compensated with an increased intake of P from manufactured feed. The largest decrease was reached on farm 3 (zero grazing, 100% grass silage): 1.2 kg P yr-1 cow-1. At the farm level, this was equal to a decrease of 4.4 m3 ha-1 manure export (≈ €88 ha-1). On farm 2 (50%-50% maize-grass in ration) the P excretion was already Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 545 low in the normal situation, and the effect of cutting at a later growth stage was small. On all farms, cutting at a later growth stage decreased the mowing costs by about €70 ha-1 on farms 1 and 3, and by €115 ha-1 on farm 2. Another effect of cutting at a later growth stage on farm 3 was that grazing pressure was higher as the swards required for the mowing cuts took more time to grow. Figure 1. Phosphorus intake of dairy cattle (kg P cow-1 year-1, milk production 8600 kg yr-1 cow-1) on three model farms at normal (N) and late (L) cuts. Acknowledgements This project was funded and supported by the Dutch Dairy board. References Fleming G.A. and Murphy W.E. (1968) Uptake of some major and trace elements by grasses as affected by season and stage of maturity. Journal of the British Grassland Society 23, 174-185. Harville D.A. (1977) Maximum likelihood approaches to variance component estimation and to related problems. Journal of the American Statistical Association 72, 320-338. Schils R.L.M., de Haan M.H.A., Hemmer J.G.A., van den Pol-van Dasselaar A., De Boer J.A., Evers A.G., Holshof G., van Middelkoop J.C. and Zom R.L.G. (2007) DairyWise, a whole-farm dairy model. Journal of Dairy Science 90, 5334-5346. Whitehead D.C. (2000) Nutrient elements in grassland; Soil-Plant-Animal relationships. Wallingford UK: CABI publishing. Wilson R.K. and McCarric R.B. (1967) A nutritional study of grass swards at progressive stage of maturity .I. Digestibility intake yield and chemical composition of dried grass harvested from swards of Irish perennial ryegrass, timothy and a mixed sward at 9 progressive stages of growth. Irish Journal of Agricultural Research 6, 267-279. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 546 Effects of concentrate levels on milk production and traffic of cows milked by a mobile automatic milking system on pasture Lessire F., Hornick J.L. and Dufrasne I. Animal Production Department, Faculty of Veterinary Medicine, University of Liège, Chemin de la Ferme, 6 B39. 4000 Liège, Belgium Corresponding author: flessire@ulg.ac.be Abstract Cows milked by an automatic milking system in pastures were assigned to two groups receiving different amounts of concentrates (2.1 kg vs 4.1 kg). The effect of level of concentrates on milk yield (MY) and returns to the robot was assessed. Level of concentrates had a positive influence on daily milk production over the grazing period, as cows of the low-concentrates group produced 21.43 ± 0.62 kg compared with 24.33 ± 0.62 kg in the high-concentrates group. However, this effect was modulated subsequently by grass quality and availability. Regarding daily voluntary returns to the robot, the high-concentrates group showed higher frequency (3.66 ± 0.05, compared with 3.22 ± 0.04 in the low concentrates group) demonstrating a positive impact of concentrate supply on cow traffic. Keywords: grazing, dairy cows, concentrate, automatic milking system, milk performance Introduction The use of an automatic milking system (AMS) implies there will be a stimulation of cows' traffic to the AMS. Several studies showed discrepancies about the effects on cows' traffic of the supply of concentrates distributed at milking; some of them finding no effect on milking frequency (Bach et al., 2007; Jago et al., 2007) while in more recent studies (Lyons et al., 2013) returns to the robot seem to be improved by supplementary feeding. The aim of this study was to analyse the influence of concentrates on milk yield and returns of grazing cows milked by an AMS in pasture. Material and methods The study was conducted from 1 May to 30 September 2013 at the Experimental Farm of Sart Tilman, University of Liège. Only those animals present from the beginning until the end of this period were taken into account. These were divided randomly into two groups each of which received a different level of concentrates. Fifteen cows, including 7 primiparous, days in milk (DIM): 97 ± 63 days, with an average lactation number (NL) 2.00 ± 1.25, were assigned to the low-concentrates (LC) group, while the high-concentrates group (HC) included 14 cows, of which 5 were primiparous (DIM = 94 ± 41 days; NL: 2.43 ±1.91). From 1 May to 30 October the cows’ diet was based on grass except for a variable amount of concentrate distributed in the AMS at milking. As grass availability decreased in September, a mixture of maize silage, dried beet pulp, alfafa pellets and straw, giving a total amount of 8 kg DM per day and per cow, was provided as a supplement from 11 September until the end of the study period. The cows were milked by an AMS Lely A3next® which was on a trailer in order that it could be moved onto pastures following the procedure described by Dufrasne et al. (2012). Transponders fixed on the HR-tag neck collar (SCR, Israel) were used in order to identify the cows and to register several parameters: milk yield (MY), number of milking, number of milking failures, and the amount of concentrate given and milking time. Twenty-four ha of pastures, which consisted mainly of perennial ryegrass and white clover, were divided into 15 plots ranging from 0.6 to 3.1 ha, and the maximum distance for the cows to walk to the robot of 700 m. Cows were assigned to different plots for day and night. The change from day to night plot was managed when cows came out the AMS, as they were driven by selection gates to their new allocation. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 547 Strip-grazing was achieved in order to provide cows with fresh grass every day. Animals stayed during 4.6 ± 1.5 days on each plot (min. = 1d; max.= 12 d). Grass height was measured using a herbometer (Jenquip®) when the cows came in and out, and the change from one parcel to the next one was decided on this basis. Grass cover was estimated by mowing a grass strip of 10 m length, and by weighing the mown sample (kg DM). Samples were analysed by NIRS to determine their nutritional value. Daily parameters (MY, number of milkings, kg concentrates per milking, nbr returns per cow calculated by summing the milkings, refusals and failures) were analysed over the grazing period. Influence of month, parity, lactation stage and their interactions was analysed using GLM procedure (SAS. 91). Results Grass mean heights at entrance were 9.3 ± 3.0 cm and 6.2 ± 1.0 cm at exit (Table 1). Grass digestibility declined over the grazing season (Table 2). Lignin content was highest in August, while the water soluble carbohydrate content was the lowest. This could be linked to weather conditions (dry with high temperatures). Table 1. Grass height at entrance and exit over the grazing period Grass height in (cm) Grass height out (cm) Difference (cm) May 16.3 6.8 9.5 June 10.3 5.8 4.5 July 12.8 6.7 6.1 August 11.0 5.9 5.1 September 9.9 5.6 4.3 Table 2. Evolution of grass nutritional values during grazing period (g kg DM -1) Month DM Crude protein Lignin Water soluble carbohydrate Digestibility (%) May 17.5 ± 2.3 162.0 ± 32.6 27.29 ± 3.20 186.6 ± 32.5 82.29 ± 4.15 June 17.0 ± 4.3 175.3 ± 40.1 26.88 ± 2.30 184.5 ± 42.4 83.63 ± 5.42 July 17.6 ± 4.1 180.5 ± 31.8 36.00 ± 4.60 141.3 ± 38.1 77.33 ± 5.89 August 20.9 ± 4.8 193.7 ± 44.5 43.29 ± 3.55 97.6 ± 26.4 79.71 ± 4.82 September 19.1 ± 6.2 206.1 ± 39.8 39.00 ± 2.55 126.9 ± 20.7 78.78 ± 5.04 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 548 Table 3. Effect of concentrates on milk yield and voluntary returns to the robot over the grazing period Month Milk yield (kg) mean ±SE Difference in MY per kg concentrate Nbr returns cow-1 d-1 Mean ±SE (kg kg-1) LC HC LC HC May 26.1 ± 0.7 26.3 ± 0.7 0.04 2.9 ± 0.1a 3.1 ± 0.1b June 21.5 ± 0.8a 24.3 ± 0.7b 1.19 3.4 ± 0.1a 3.8 ± 0.1b July 22.8 ± 0.7a 25.5 ± 0.7b 1.37 3.0 ± 0.1a 3.4 ± 0.1b August 19.7 ± 0.7a 24.0 ± 0.7b 2.43 3.4 ± 0.1a 4.0 ± 0.1b September 17.1 ± 0.7a 21.8 ± 0.7b 3.21 3.4 ± 0.1a 4.0 ± 0.1b LC and HC denote low concentrates and high concentrates, respectively. Means within a row with different letters superscripts differ (P<0.001). In both the LC and HC groups the MY decreased over the grazing time with the increase in number of days in milk (Table 3). In May, LC and HC cows produced the same amounts of milk. This absence of response to the distribution of concentrates can be explained by the high availability of good quality grass. Difference in milk production between the LC and HC groups increased from 2.69 kg in June to 4.66 kg in September. The impact of the level of concentrates was more pronounced when grass availability and/or its quality was poor, as occurred at the end of the grazing season or in August when weather conditions were unfavourable for grass growth and grazing. The distribution of the mixture 'maize silage, beet pulp and alfalfa pellets' did not prevent milk production from decreasing. On average, the HC group produced 2.9 kg more milk over the season, representing 1.5 kg of milk per kg concentrate. Returns to the robot were more frequent in the HC group, whatever the month. Conclusion The response of milk yield to concentrates at grazing can vary considerably during the grazing season. In this trial, milk yield was generally improved by higher supply of concentrates. When grass availability and quality decreased, the difference between the high-concentrates and lowconcentrates groups was more important. In this trial, the level of concentrates had a positive influence on the cows' traffic to the robot. References Bach A., Iglesias C., Calsamiglia S. and Devant M. (2007) Effect of amount of concentrate offered in automatic systems on milking frequency. feeding behaviour. and milk production of dairy cattle consuming high amounts of corn silage. Journal of Dairy Science 90, 5049-5055. Dufrasne I., Robaye V., Knapp E., Istasse L. and Hornick J.L. (2012) Effects of environmental factors on yield and milking number in dairy cows milked by an automatic system located in pasture. Grassland Science in Europe 17, 231-233. Jago J. G., Davis K.L., Copeman P.J., Ohnstad I. and Woolford M. (2007) Supplementary feeding at milking and minimum milking interval effects on cow traffic and milking performance in a pasture-based automatic milking system. Journal of Dairy Research 74, 492- 499. Lyons N.A., Kerrisk K.L. and Garcia S.C. (2013) Effect of pre-versus post-milking supplementation on traffic and performance of cows milked in a pasture-based automatic milking system. Journal of Dairy Science 96, 43974405. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 549 Relationship between fatty acid content and nutritive value of perennial ryegrass (Lolium perenne) Morgan S.A., Huws S.A., Tweed J.K.S., Hayes R.C. and Scollan N.D. Institute of Biological, Environmental and Rural Sciences, Aberystwyth University, Gogerddan, Aberystwyth, SY23 3EB, United Kingdom. Corresponding author: sgm8@aber.ac.uk Abstract There is growing interest in creating novel forages to address consumer demands, such as enhancing product quality and functionality and reducing environmental impacts of ruminant production. To address these, one area under investigation is the fatty acid content of forages. This paper focuses on the relationship between total fatty acid (TFA) content and other nutritional characteristics of perennial ryegrass. A study was set up involving twenty-four genotypes from two populations of perennial ryegrass. Genotype selection was based on historic soil plant analysis development (SPAD) data, with the intention of selecting a set of genotypes that best represent the variation in TFA content of each population. Plants were maintained in a polytunnel for three months prior to harvesting. Fatty acids were quantified using gas chromatography (GC) while CP and WSC were predicted using near-infrared reflectance spectroscopy (NIRS). Total fatty acid content was found to correlate positively with CP (R2 = 0.69, 94 d.f., P<0.001) and negatively with WSC (R2 = 0.14, 94 d.f., P<0.001). Keywords: Lolium perenne, fatty acids, water soluble carbohydrate, crude protein Introduction Plants are the primary source of n-3 fatty acids, both in marine and terrestrial ecosystems, due to their unique capability of de novo synthesis of 18:3n-3. This fatty acid is the building block for longer-chain n-3 series of essential fatty acids, via elongation and desaturation pathways (Barceló-Coblijn and Murphy, 2009). Although grass has a somewhat low total fat content, with the lipid content of leaf tissue varying between 3-10% of DM (Hawke, 1973), the majority of this comprises of the PUFA 18:3n-3 and 18:2n-6. Thus, it can contribute greatly, and in some situations wholly, to the total lipid intake of ruminant animals. Dietary lipid content and composition contributes towards 1) energy provision to the animal and 2) the fatty acid profile of ruminant products (i.e. meat/milk). With respect to improving the fatty acid profile of ruminant products, increased emphasis has been placed on the potential of using plant-based strategies of late (Morgan et al., 2012). Grass is a high PUFA feed source which is cheap, natural and environmentally sustainable, thus is a noteworthy strategy to improve fat composition of ruminant products (Dewhurst et al., 2003). Studies have highlighted a strong genetic basis to fatty acid composition (Dewhurst et al., 2001). Hegarty et al. (2013) have successfully identified regions of the genome which are associated with fatty acid composition of perennial ryegrass. However, large genotype × environment interactions have also been recognized (Dewhurst et al., 2003; Palladino et al., 2009). Materials and methods Four control genotypes were selected from an Aurora × AberMagic F1 mapping population and twenty experimental genotypes selected from the B674G intermediate-heading 13th generation breeding population. Genotypes were selected in order to provide the best possible representation of the variation in fatty acid content within each population, based on historic Soil Plant Analysis Development (SPAD) data which have been shown to correlate positively with fatty acid content (Morgan et al., 2013). Mature single ryegrass tillers were transplanted into 15-cm pots in potting compost during April 2012. Four replicate clones of each genotype Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 550 were used and were arranged in a randomized block design. Plants were maintained in a polytunnel, with actively reproductive heads cut back every two weeks in order to encourage tillering. Plants were harvested by hand in July 2012 to a cutting height of 5 cm. All plant material was collected and temporarily stored on ice during harvest then freeze-dried and stored at -20oC. Near infrared reflectance spectroscopy (NIRS) analysis was carried out as described by Lister and Dhanoa (1998) to estimate WSC and CP (= N × 6.25) content. Fatty acid content was determined via the procedure of Sukhija and Palmquist (1988) using Tricosanoic acid methyl ester (C23:0) as an internal standard (Sigma Aldrich Co, St. Louis, MO, USA). Fatty acid methyl esters (FAME) were separated and quantified using gas chromatography (GC-FID) system (CP-3800 with PAL Autosampler, Varian Inc, CA, USA) equipped with a CP-select 100 m × 0.25 mm chemically bonded for FAME column (Agilent technologies UK Ltd, Berkshire, England, UK). Peaks were identified using a 37 FAME standard (S37, Supelco, Poole, Dorset, UK) and quantified using the C23:0 internal standard. Data were analysed via Genstat (14th edition; VSN International Ltd, Hemel Hempstead, UK) using linear regression and correlation. Results and discussion Total fatty acid (TFA) content ranged from 14.5 to 34.0 g kg-1 DM with a mean value of 23.3 g kg-1 DM. These values are similar to those published by Dewhurst et al. (2001, 2002) and Van Ranst et al. (2009) for July harvests of perennial ryegrass. Crude protein content ranged from 6.9 to 20.1% DM and WSC content ranged from 7.9 to 34.6% DM, with mean values of 13.3% DM and 18.8% DM, respectively. A very strong, positive relationship was found between TFA and CP (R2 = 0.69, P<0.001), which is illustrated in Figure 1. Other studies which investigated the relationship between N or CP and fatty acid content of forages have also found positive relationships. Elgersma et al. (2005) found a very strong relationship between CP and C18:3n-3 (R2 = 0.90, P<0.001) with similar results for CP and TFA relationship, while Boufaied et al. (2003) reported a positive relationship between TFA and N content (R2 = 0.79, P<0.001). Crude protein and WSC are known to have an inverse relationship (Humphreys, 1989). Consequently, the relationship between TFA and WSC (illustrated in Figure 2) was also found to be negative (P <0.001); however it is worth highlighting that this was found to not be a strong relationship (R2=0.14). 30.0 17.5 WSC (% DM) CP (% DM) 20.0 15.0 12.5 10.0 25.0 20.0 15.0 10.0 7.5 5.0 5.0 10 15 20 25 Total fatty acids (g 30 kg-1 35 DM) Figure 1. Relationship between crude protein content (% DM) and total fatty acid content (g kg-1 DM) of perennial ryegrass 10 15 20 25 Total fatty acids (g 30 kg-1 35 DM) Figure 2. Relationship between water soluble carbohydrate content (% DM) and total fatty acid content (g kg-1 DM) of perennial ryegrass Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 551 Conclusion This study provides further testimony to the positive relationship between TFA and CP. Deeper exploration of this relationship is needed to provide insight into the mechanisms underpinning it. This study has also uncovered the relationship between TFA and WSC. Further research is needed to confirm this relationship and to assess if it is similar across a wider range of varieties/species and growth season/environmental conditions Acknowledgements This project is funded through KESS with additional support from Celtic Pride Ltd. KESS is part-funded by the European Social Fund (ESF) through the European Union's Convergence Programme (West Wales and the Valleys) administered by the Welsh Government. References Barceló-Coblijn G. and Murphy E.J. (2009) Alpha-linolenic acid and its conversion to longer chain n−3 fatty acids: Benefits for human health and a role in maintaining tissue n−3 fatty acid levels. Progress in Lipid Research 48, 355–374. Boufaïed H., Chouinard P.Y., Tremblay G.F., Petit H.V., Michaud R. and Bélanger G. (2003) Fatty acids in forages. I. Factors affecting concentrations. Canadian Journal of Animal Science 83, 501–511. Dewhurst R.J., Moorby J.M., Scollan N.D., Tweed J.K.S. and Humphreys M.O. (2002) Effects of a stay‐green trait on the concentrations and stability of fatty acids in perennial ryegrass. Grass and Forage Science 57, 360– 366. Dewhurst R.J., Scollan N.D., Lee M.R.F., Ougham H.J. and Humphreys M.O. (2003) Forage breeding and management to increase the beneficial fatty acid content of ruminant products. Proceedings of the Nutrition Society 62, 329–336. Dewhurst R.J., Scollan N.D., Youell S.J., Tweed J.K.S. and Humphreys M.O. (2001) Influence of species, cutting date and cutting interval on the fatty acid composition of grasses. Grass and Forage Science 56, 68–74. Elgersma A., Maudet P., Witkowska I.M. and Wever A.C. (2005) Effects of Nitrogen fertilisation and regrowth period on fatty acid concentrations in perennial ryegrass (Lolium perenne L.). Annals of Applied Biology 147, 145– 152. Hawke J.C. (1973) Lipids. In: Chemistry and Biochemistry of Herbage (eds G.W. Butlerand & R.W. Bailey), pp. 213–263. Acedemic Press, London. Hegarty M., Yadav R., Lee M., Armstead I., Sanderson R., Scollan N., Powell W. and Skøt L. (2013) Genotyping by RAD sequencing enables mapping of fatty acid composition traits in perennial ryegrass (Lolium perenne (L.)). Plant biotechnology journal 11, 572–581. Humphreys M.O. (1989) Water-soluble carbohydrates in perennial ryegrass breeding. Grass and Forage Science 44, 423–430. Lister S.J. and Dhanoa M.S. (1998) Comparison of calibration models for the prediction of forage quality traits using near infrared spectroscopy. Journal of Agricultural Science 131, 241–242. Morgan S., Huws S.A. and Scollan N.D. (2012) Progress in forage-based strategies to improve the fatty acid composition of beef. Grassland Science in Europe, 17, 295–307. Morgan S.A., Huws S.A., Tweed J.K.S., Hayes R.C. and Scollan N.D. (2013) Relationship between chlorophyll content (a + b) (SPAD value) and fatty acid composition of perennial ryegrass (Lolium perenne) – a preliminary study. In Innovation from animal science - a necessity not an option p. 146. Nottingham, UK. Palladino R.A., O’Donovan M., Kennedy E., Murphy J.J., Boland T.M. and Kenny D.A. (2009) Fatty acid composition and nutritive value of twelve cultivars of perennial ryegrass. Grass and Forage Science 64, 219–226. Van Ranst G., Fievez V., Vandewalle M., De Riek J. and Van Bockstaele E. (2009) Influence of herbage species, cultivar and cutting date on fatty acid composition of herbage and lipid metabolism during ensiling. Grass and Forage Science 64, 196–207. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 552 Can we use the fatty acid composition of bulk milk to authenticate the diet composition? Martin B.1,2, Coppa M.3, Chassaing C.1,2, Agabriel C.1,2, Borreani G.3, Barcarolo R.4, Baars T.5, Kusche D.6, Harstad O.M.7, Verbič J.8, Golecký J.9 and Ferlay A.1,2 1 INRA, UMR 1213 Herbivores, F-63122 Saint-Genès-Champanelle, France 2 Clermont Université, VetAgro Sup, UMR Herbivores, BP 10448, F-63000 Clermont-Ferrand, France 3 University of Turin, Department of Agricultural, Forest and Food Sciences, Via L. da Vinci 44, 10095, Grugliasco, Italy 4 Veneto Agricoltura Ist. Qualità e Tecnologie Agroalimentari, Via S. Germano 74, I-36016, Thiene (VI), Italy 5 Research Institute of Organic Agriculture (FiBL) Ackerstrasse, 5070 Frick, Switzerland 6 Kassel University, Faculty of Organic Agricultural Sciences, Nordbahnhofsstrasse 1, 37213 Witzenhausen, Germany 7 Norwegian University of Life Sciences, Dep. Animal and Aquacultural Sciences, Arboretveien 6, 1430 Ås, Norway 8 Agricultural Institute of Slovenia, Hacquetova 17, SI-1000 Ljubljana, Slovenia, 9 Plant Production Research Center (PPRC) – Grassland and Mountain Agriculture Research Institute (GMARI), Mladeznicka 36, 974 21 Banska Bystrica, Slovakia Corresponding author: bruno.martin@clermont.inra.fr Abstract The aim of this work was to predict cow diet composition from the fatty acid (FA) concentrations of bulk milk collected in ten different EU countries. A dataset comprising 1248 bulk cow-milk samples and corresponding information related to diet composition assessed by rapid interviews with farmers was used. The predictions based on FA for cow diet composition were excellent for authenticating the proportions in cow diets of fresh herbage (R2 = 0.81), good for hay, intermediate for maize silage and grass silage (R2 > 0.62) and poor for concentrates (R2 < 0.51). The prediction models we presented could offer a valuable tool to authenticate cow feeding. Keywords: milk fatty acid, feeding system, authentication Introduction In several recent studies, plant biomarkers (such as terpenes, polyphenols, carotenoids), stable isotopes, and milk fatty acid (FA) were proposed as candidate compounds in milk for the analytical authentication of animal diets. The milk FA composition was the more effective method to discriminate cow diets including low or high amounts of maize silage into the diet (Engel et al., 2007). Indeed, cow diet is known to be the most important driver for FA composition (Chilliard et al., 2007; Coppa et al., 2013). However, only few studies have tested the potential of bulk-milk FA profile to predict the composition of cow diets adopted by commercial farms and these studies were made at a large territorial scale (Engel et al., 2007). The aim of this work was to predict cow diet composition from FA concentrations of bulk milk collected in different EU countries. Materials and methods The FA profiles of 1248 bulk cow milk samples and their related production conditions were compiled from a selection of 20 published or unpublished studies carried out from 2000 to 2010 in ten EU countries: France, Germany, Italy, Norway, Slovakia, Slovenia, Czech Republic, Denmark, Sweden and the Netherlands. The detailed composition of this dataset is given by Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 553 Coppa et al. (2013). It included milk collected on commercial farms from a latitudinal range of 44° N to 60° N, from sea level to 2000 m altitude, from 13 different cow breeds and in all seasons. To develop a model of prediction of diet composition based on milk FA profile, a general linear model was applied using experiment as fixed factor and FA concentrations as covariates. Results and discussion The fresh herbage proportion in cow diets was reliably predicted by the model (R2 = 0.81). The FA explaining most of the model variability were iC17:0 and C18:2t11c15, both linearly and positively related to fresh herbage proportion in the cow diet. The C18:2t11c15 is an intermediate of ruminal biohydrogenation of dietary C18:3n-3, which is the main herbage FA (Chillard et al., 2007), and its increase in milk with increasing proportion of fresh herbage in the cow diet is well known (Couvreur et al., 2006; Chilliard et al., 2007). Vlaemink et al. (2006) showed a linear increase of iC17:0 with increasing of the major bio-hydrogenation intermediates, such as C18:2trans11cis15. In our model, fresh herbage proportion in cow diets also decreased linearly with C16:0, increased linearly with C17:1c9 and decreased cubically with C17:0 concentration in milk. An increase in proportion of fresh herbage in cow diets was previously associated with an increase in C17:0 in milk and with a decrease of C16:0 (Couvreur et al., 2006). The prediction model for maize silage proportion had a lower fit (R2 = 0.66) when compared to the fresh herbage model. Most of the model variability was explained by the C16:1c9+aiC17.0 to iC16:0 ratio, linear increase of concurred with increase in proportion of maize silage in cow diet. High values of this FA ratio were also associated by Engel et al. (2007) with a diet rich in maize. Most of the variability of the prediction model for hay proportion in cow diet (R2 > 0.74) was explained by milk iC14:0, C18:3n-3 concentrations and C18:2t11c15 to C18:1t10+t11 ratio. The positive quadratic relation between milk iC14:0 and hay proportion in the cow diet can be easily explained by the microbial origin of this FA, which is derived from ruminal cellulolytic bacteria (Vlaemink et al., 2006). The positive linear relation we found between C18:3n-3 concentration in milk and hay proportion in cow diet is in agreement with Chilliard et al., (2007). This trend agrees also with the negative linear relation we found in the ratio of C18:2t11c15 to C18:1t10+t11 (Chilliard et al., 2007). The model fit for grass silage proportion was similar to those of maize silage (R2 = 0.62). Grass silage proportion in cow diets decreased linearly with increasing milk isoC17:0 concentration. This trend is in agreement with findings of Vlaemink et al. (2006) which showed a decrease in this FA when grass silage was replaced by maize silage in cow diets. The C12:0 and C15:0 were negatively and positively quadratically related to grass silage, respectively. A decrease in C18:2c9t13 was linearly related to an increase in grass silage concentration, confirming the lower concentration of the intermediates of ruminal biohydrogenation of ingested C18:3n-3, when fresh herbage is substituted by grass silage (Chilliard et al., 2007). The lowest reliability of FA for the authentication of animal diets was found for the prediction of the proportion of concentrates in cow diets (R2 < 0.51). The FA explaining most of the model variability were iC14:0, C18:2n-6, and the C16:1c9+aiC17:0 to iC16:0 ratio. The concentrate proportion in cow diets decreased with linearly decreasing isoC14:0, and increased with quadratically increasing of the C18:2n-6 concentration in milk. These results are in agreement with the FA profile of milk derived from diets poor in forages (Chilliard et al., 2007). A linear decrease in C16:1c9+aiC17:0 to iC16:0 ratio was associated with an increase in concentrate proportion in cow diets in our model, in agreement with Vlaemink et al. (2006). Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 554 Table 1. Prediction models of proportions in cow diets of various feeds, based on milk FA composition. The FA are listed according to their decreasing weight in model determination (RMSE: root mean square error) Ingredients Fatty acid (% of DM diet) Fresh herbage iC17:0;C18:2t11c15;C17:1c9;C17:0^3;C16:0;C18:2n-6;C18:2t11c15^3; C17:1c9^3;C16:0^3;C18:2c9t13^2;C18:1t10+t11^3;iC16:0^3 n RMSE R2 693 15.6 0.81 Maize silage C16:1c9+aiC17:0;C18:3n-3;C20:5n-3;C18:3n-3^2;C20:4n-6;C15:0; 678 C18:3n-3^3;C15:0^2;C18:2n-6^3;C15:0^3; C18:2t11c15/C18:1t10+t11; C18:1c9+t13^3;C18:2c9t13^3;C18:2n-6 9.1 0.66 Hay iC14:0^2;C18:2t11c15/C18:1t10+t11;C18:3n-3;C18:3n-3^2;C16:0; C17:0^2;C15:0;C17:0;C18:2c9t13;C18:1t10+t11;C14:0^2;C20:4n-6; C14:0;C20:5n-3;C20:4n-6^3 683 11.2 0.75 Grass silage iC17:0;C12:0^2;iC17:0+C16:1t9^3;C15:0^2; C18:2c9t13;iC14:0; 659 C16:1c9+aiC17:0/iC16:0;C18:2t11c15/C18:1t10+t11;iC16:0;C18:2n6^3; iC16:0^3;C18:2n-6;C16:0;C16:0^3;C16:1c9+aiC17:0;C20:4n-6 iC14:0;C16:1c9+aiC17:0;C18:2n-6^3;C17:0;iC14:0^3;C18:1t10+t11; 883 C17:0^3;iC17:0+C16:1t9^2;C18:3n-3^2;C18:3n-3;C18:2c9t13^3 10.6 0.62 8.1 0.51 Concentrates Conclusion Our work provided original and reliable models to predict cow diet composition based on milk FA composition. A dataset composed of bulk milk collected in several European countries was used. These prediction models could offer a valuable tool to authenticate cow feeding. Acknowledgements This work was supported by the INRA-PHASE division that funded M. Coppa’s post-doctoral fellowship at the INRA-UMRH unit. References Chilliard Y., Glasser F., Ferlay A., Bernard L., Rouel J. and Doreau M. (2007) Diet, rumen biohydrogenation and nutritional quality of cow and goat milk fat. European Journal of Lipid Science and Technology 109, 828-855. Coppa M., Ferlay A., Chassaing C., Agabriel C., Glasser F., Chilliard Y., Borreani G., Barcarolo R., Baars T., Kusche D., Harstad O. M., Verbič J., Golecký J. and Martin B. (2013). Prediction of bulk milk fatty acid composition based on farming practices collected through on-farm surveys. Journal of Dairy Science 96, 4197– 4211. Couvreur S., Hurtaud C., Lopez C., Delaby L. and Peyraud J.L. (2006) The linear relationship between the proportion of fresh grass in the cow diet, milk fatty acid composition, and butter properties. Journal of Dairy Science 89,1956-1969. Engel E., Ferlay A., Cornu A., Chilliard Y, Agabriel C., Bielicki G. and Martin B. (2007) Relevance of isotopic and molecular biomarkers for the authentication of milk according to production zone and type of feeding. Journal of Agricultural and Food Chemistry 55, 9099-9108. Prache S., Cornu A., Berdagué J. L. and Priolo A. (2007) Traceability of animal feeding diet in the meat and milk of small ruminants. Small Ruminant Research 59, 157–168. Vlaemink B., Fievez V., Cabrita A. R. J., Fonseca A. J. M., and Dewhurst R. J. (2006) Factors affecting odd- and branched-chain fatty acids in milk: a review. Animal Feed Science and Technology 131, 389-417. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 555 Effect of dietary supplementation on milk production and milk composition of grazing dairy cows in late lactation Reid M.1,2, O’Donovan M.1, Lalor S.T.J.4, Bailey J.S.3, Elliott C.T.2, Watson C.J.3 and Lewis E.1 1 Teagasc, Animal & Grassland Research and Innovation Centre, Moorepark, Fermoy, Co. Cork, Ireland 2 Institute of Global Food Security, Queen’s University, Belfast, Northern Ireland, UK 3 Agri-Food and Bioscience Institute, Newforge Lane, Belfast, BT9 5PX, N. Ireland, UK 4 Teagasc, Crops, Environment and Land Use Programme, Johnstown Castle, Wexford, Ireland Corresponding author: Michael.Reid@teagasc.ie Abstract Grass silage is now recommended as an alternative to concentrate to supplement the grass-based diet of dairy cows in late lactation. The aim of this experiment was to establish the effect of supplementation on the milk production of grazing dairy cows in late lactation. Eighty-four spring-calving dairy cows were randomly allocated to one of six treatments: high grass-only allowance (HG), low grass-only allowance (LG), grass with low concentrate (GCL), grass with low grass-silage (GSL), grass with high concentrate (GCH) and grass with high grass-silage (GSH). Treatment GCH had a greater milk yield than all other treatments, and GCL had a greater milk yield than GSH, GSL, HG and LG. There was no difference between the milk yield of GSH, HG and LG; however, the milk yield of the GSL treatment was greater than on HG and LG. Milk solids yield was greater in the two concentrate-supplemented treatments than in the other four treatments, and LG had the lowest milk solids yield. Milk fat concentration was lower in the silage-supplemented treatments and in GCH than in HG. Milk protein concentration was significantly higher in the two grass-only treatments than in the four supplemented treatments. Concentrate supplementation in late lactation increased milk solids yield compared to grass-only or grass supplemented with silage. However, supplementation with either concentrate or silage negatively affected milk fat and protein concentration. Keywords: supplementation, nutrition, milk yield, milk composition, silage, concentrate Introduction Dairy farming in northern European temperate countries, such as Ireland, is based on springcalving and the efficient conversion of grazed grass into milk (Dillon et al., 1995). The objective of the system is to allow grazed grass to make up as large a proportion as possible of the diet of the lactating dairy cow (O’Donovan et al., 2004). Grass supply is often limited in the spring and autumn, and so at this time the predominantly grass-based diet is supplemented. Concentrate feeds are commonly offered in the late lactation period in order to meet the requirements for maintenance, growth, lactation and pregnancy. More recent recommendations are that supplementation could be in the form of grass silage. Grass silage is the next most important ruminant feedstuff after grazed grass (McEniry et al., 2011) and is a cheaper alternative to concentrate (Finneran et al., 2010). There are few comparisons in the literature of concentrate and grass silage as supplementary feeds in late lactation. Therefore, the aim of this experiment was to investigate the effects on milk production of offering grass silage or concentrate supplementation to grazing dairy cows in late lactation. Materials and methods Eighty-four spring-calving lactating dairy cows were blocked on milk yield, milk composition, parity, bodyweight, body condition score and EBI. From within block, the cows were randomly allocated to one of six treatments: high grass-only allowance (HG), low grass-only allowance Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 556 (LG), grass with low concentrate allowance (GCL), grass with low grass-silage allowance (GSL), grass with high concentrate allowance (GCH) and grass with high grass-silage allowance (GSH). Treatments HG and LG were allocated 17 and 14 kg DM grass per cow/day respectively. Treatments GCL and GSL were offered 14 kg DM grass per cow/day and 3 kg DM supplementation per cow/day. Treatments GCH and GSH were offered 11 kg DM grass per cow/day and 6 kg DM supplementation per cow/day. The average DMD of the silage was 697 g/kg and it was offered after morning milking. The concentrate offered was a 160 g/kg crude protein dairy concentrate and was offered in two equal amounts at morning and evening milking. Milk yield was recorded daily and milk composition was recorded weekly. Data were analysed using covariate analysis and the PROC GLM statement of SAS. Treatment, week of experiment and the appropriate covariate were included in the model. Results and discussion Table 1 shows the milk production performance of the study. Table 1. Effect of dietary supplementation on milk production and milk composition of grazing dairy cows in late lactation. 14 kg DM grass + 3 kg DM supplementary feed 11 kg DM grass + 6 kg DM supplementary feed Grass only 17 kg DM (HG) 14 kg DM (LG) Concentrate (GCL) Silage (GSL) Concentrate (GCH) Silage (GSH) S.E. P-Value Milk yield (kg/d) 11.9a 11.6a 14.8b 12.8c 16.0d 12.1ac 0.96 <0.001 Milk fat (g/kg) 57.9a 57.7ac 57.0ac 55.0bc 55.1bc 54.5b 1.30 <0.01 Milk protein (g/kg) 42.1a 41.6a 39.9b 40.1b 39.8b 39.7b 0.50 <0.001 Milk solids (kg/d) 1.16b 1.06a 1.40c 1.21b 1.54c 1.12b 0.030 <0.001 a-d Means within a row not sharing a common superscript differ significantly (P<0.05) Milk yield was significantly (P<0.001) affected by treatment. Treatment GCH had a greater milk yield than all other treatments, and GCL had a greater milk yield than GSH, GSL, HG and LG. There was no difference between the milk yield of GSH, HG and LG; however, the GSL treatment had a greater milk yield that HG and LG. The milk yields of HG and LG agree with the study of Reid et al. (2013). The increase in milk yield when late lactation grazing cows were supplemented also agrees with that study. Milk solids yield was greater in the two concentrate supplemented treatments than in the other four treatments, but did not differ between the two concentrate supplemented treatments. This indicates that there was no advantage in offering 6 compared to 3 kg DM concentrate in late lactation to grazing dairy cows. There was no difference in milk solids yield between HG, GSL and GSH, all of which were greater than LG. Milk fat concentration was significantly lower in the silage supplemented treatments and in GCH than in HG. The negative effect on milk fat concentration of silage supplementation in late lactation was previously measured (O’Brien et al., 1996) and the reasons for this should be explored in more detail. The negative effect of high concentrate supplementation on milk fat concentration was not unexpected as high levels of dietary concentrate can be associated with a reduction in rumen pH and consequent reduction in milk fat concentration (Abrahamse et al., 2008). There was a negative effect of supplementation on milk protein concentration, which agrees with previous research (Sutton et al., 1996) and which suggests a negative effect on milk processability (O’Brien et al., 1996). Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 557 Conclusion This experiment shows that in late lactation, supplementing a grass-based diet with concentrate increased milk solids yield compared to feeding grass only or grass supplemented with silage. Supplementation with either silage or concentrates negatively affected milk fat and protein concentration. The effects of supplementing grass-based diets in late lactation on milk processability should be measured. Acknowledgments Funding from the Department of Agriculture, Food and the Marine Stimulus Fund, Ref 11/sf/309. References Abrahamse P.A., Vlaeminck B., Tamminga S. and Dijkstra J. (2008) The effect of silage and concentrate type on intake behaviour, rumen function and milk production in dairy cows in early and late lactation. Journal of Dairy Science 91, 4778-4792 Dillon P., Crosse S., Stakelum G. and Flynn F. (1995) The effect of calving date and stocking rate on the performance of spring calving dairy cows. Grass and Forage Science 50, 286-299 Finneran E., Crosson P., O’Kiely P., Shalloo L., Forristal D. and Wallace M. (2010) Simulation modelling and the cost of producing and utilising feeds for ruminants on Irish dairy farms. Journal of Farm Management 14, 95-116 McEniry J., Forristal D. and O’Kiely P. (2011) Factors influencing the conservation and characteristics of baled and precision-chop grass silages. Irish Journal of Agricultural and Food Research 50, 175-188 O’Brien B., Crosse S. and Dillon P. (1996) Effects of offering a concentrate or silage supplement to grazing dairy cows in late lactation on animal performance and on milk processability. Irish Journal of Agricultural and Food Research 35, 113-125 O’Donovan M., Delaby L., Stakelum G. and Dillon P. (2004) Effect of autumn/spring nitrogen application date and level on dry matter production and nitrogen efficiency in perennial ryegrass swards. Irish Journal of Agricultural and Food Research 43, 31-41 Reid M., O’Donovan M., Bailey J.S., Elliott. C.T., Watson C.J., Murphy J.P., Coughlan. F. and Lewis E. (2013) The effect of offering different supplementary feeds to grazing dairy cows in late lactation on milk yield and milk composition. In: Agricultural Research Forum. Standard Printers, Tullamore, Ireland, pp. 109 Sutton J.D., Aston K., Beever D.E. and Dhanoa M.S. (1996) Milk production from grass silage diets: effects of high-protein concentrates for lactating heifers and cows on intake, milk production and milk nitrogen fractions. Animal Science 62, 207-215 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 558 Combining robotic milking and grazing Brocard V.1, Huneau T.2, Huchon J-C.2 and Dehedin M.2 1 Institut de l'Elevage, BP 85225, 35652 Le Rheu Cedex, France 2 Chambre d’agriculture 44, Ferme expérimentale de Derval, 44590 Derval, France Corresponding author: Valerie.Brocard@idele.fr Abstract In 2013 in France, nearly 3000 milk producers use an automatic milking system (AMS). It often leads to reduced grazing, for fear of a drop in milking frequency. A national project was set up to bring technical solutions to breeders who wish to combine AMS with grazing. During 5 years, the experimental farm of Derval tested several traffic options and herd management opportunities while grazing. The maize-silage silo was closed for 34 days (in 2012) and 56 days (2013). Holstein cows on a 100% grazed diet produced, on average, 27.5 kg of milk with 2.8 kg of concentrates. The feeding cost was one-third the winter ration. Each cow ingested up to 1.5 t of grazed grass per year. During the same programme, 20 French robotic farms were surveyed with various saturation levels, grass growth potential and traffic. This programme shows that grazing with an AMS remains possible as long as the farmer keeps motivation, a sufficient grazeable area, and implements the right traffic options. Keywords: AMS, grazing, milking frequency, dairy cows Introduction In France, the massive introduction after 2000 of AMS often led to reduced use of grazed grass in the diet of dairy cows (Jégou et al., 2007; Billon, 2009). The possible decrease in milking frequency which is often presented as an important factor for individual productivity often threatens the breeders. As they are also seeking to reduce working time, grazing does not always appear as a solution to reach this target, but keeping grazed grass in cow diets is advocated for its positive impacts on nutrition and health (Burow, 2011). It is also an efficient solution to the need to reduce production costs thanks to the low cost of per energy unit to produce milk from grazed grass compared with other feeds. The experimental farm of Derval (western France) tested different solutions to graze dairy cows milked with an AMS. The research programme had associated experiments in stations as well as pilot farms in order to broaden the list of solutions that can be used to reconcile AMS and grazing. Materials and methods The experiments lasted from 2009 to 2013. The following aimed at studying the impact of the share of grazed grass in the daily diet ("day grazing only", or "day and night grazing") on the performances of the cows and the robot, compared to the performances during the winter period when the cows are kept inside. The average number of cows (Holstein) milked by the robot (a Delaval one-stall VMS 2007) reached 72. The yearly production is ca. 9000 kg of milk /cow and there are ca. 150 milkings every day. The cow traffic is guided: the cowshed equipped with cubicles has a drafting gate only usable from the feeding area. After the robot, a second gate directs the cow towards grazing or towards an isolation box. Derval is located in a dry area (mean 600 mm rainfall /year). The grazable area reaches 28 ha of temporary grasslands of ryegrass-white clover. One large track (3.5 m) leads to the 3 paddocks of 10, 10 and 8 hectares. Grass management is a simplified rotational system. The maximal walking distance for the cows to reach the end of the furthest paddock is 800 m. One water trough is located at the entrance of the shed and a second just before the exit. No water is available in the fields. The buffer feed (maize silage) is adjusted to the amount of grass outside and the growth forecasts to maintain 10 days of grass ahead. It is delivered in the morning. When the cows get 8 kg DM per day, they do not have access to the trough as long as the whole herd is not inside the shed, Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 559 to ensure all the cows have the same access time to the buffer feed. The amounts of grass grazed were estimated through the difference between the calculated Intake capacity of the cows (INRA 2007, R Delagarde, pers. comm.) and the intakes of supplements. SAS ® software was used to analyse data with PROC MEANS, GLM and MIXED. For weather data, classification methods (PCA and HAC) were used. A number was dedicated to each feeding period: P1 (100% shed period), P2 (transition period) and P3 (100% grazing period.) Statistical analyses are based on 2011, 2012 and 2013. The international production cost method (IFCN) was used to assess the feeding cost and the margin over feeding cost. Results The access time to grazing allowed by the traffic management led to the valorization of more than 1 ton of grazed grass per cow per grazing season. In 2013 the maize clamp was closed for 56 days. The individual dairy production remained around 27.5 kg during the 100% grazed grass diet both in 2012 and 2013 (Table 1). Table 1. Dairy production, milking frequencies and concentrates per period. 2011 2012 2013 P1 61d P2 44d P3 11d P1 60d P2 74 d P3 32d P1 60 d P2 33d P3 56d Cow number 68 74 74 73 74 74 69 71 67 Lactation stage (d) 206 178 178 192 189 185 226 229 232 27.6 30.4 28.3 29.4 31.6 27.6 30.1 30.7 27.5 Diff to P1 / +2.8 +0.7 / +2.2 -1.8 / +0.6 -2.6 Concentrates kg c-1 d-1 4.8 3.9 2.7 3.9 3.9 2.3 5.6 4.6 3.3 Diff to P1 / -0.9 -2.1 / 0 -1.6 / -1.0 -2.3 1876 2253 2097 2130 2323 2044 2074 2182 1842 2.15 1.86 1.97 2.06 1.92 1.86 2.12 2.08 2.12 / -0.3 -0.2 / -0.14 -0.2 / -0.04 0 -1 Milk kg d c -1 Milk kg per AMS d -1 -1 Milking frequency c d Diff to P1 -1 During P3, concentrate (wheat) delivered is limited to 2.8 kg c-1 d-1 versus 4.8 average during P1. Adjusted means of the productions are 28.9 kg d-1 c-1 for P1, 30.3 for P2 and 27.5 for P3. An increase of 1.4 kg c-1 d-1 occurred during P2 (P< 0.0001). During P3, there was a decrease of 1.4 kg compared to the average production during the winter period P1 (P< 0.0001), partly due to a decrease by 1.6 to 2.3 kg c-1 d-1 of concentrates in 2012 and 2013. The average milking frequency ranges from 1.86 to 2.15 milkings c-1 d-1 in relation to the saturation level of the robot, though variation between P1 and P3 is only a decrease by -0.3 to -0.04 milkings c-1 d-1. The adjusted mean for milking frequency is 2.11 milkings c-1 d-1 in P1. This frequency decreases in P2 (1.96 milkings c-1 d-1) then reaches 1.99 milkings c-1 d-1 in P3. Thus, the effect of grazing leads to a decrease by 0.15 milkings c-1 d-1 (P<0.0001) in P2 and only by 0.11 milkings c-1 d-1 (P<0.0001) in P3. The weather influences the return of the cows from pasture (Lozach, 2011). Days with the lowest milking frequency (1.79 milkings c-1 d-1) are characterized by high wind speed with high humidity, with no effect of rainfall. In terms of working time, during P2, sorting cows takes ca. 5 min to check milking times on computer, and 10 min to sort cows inside the shed. This is done daily at ca. 8 am after delivering maize silage, so cows are brought to feed racks. During P2 and P3, the cows are fetched by the herdsman at ca.6 pm, 20-40 min. are required. In terms of economic impact, the feeding cost for 1000 L delivered decreases from 148 € in P1, to 43 € in P3. Thanks to grazing, the monthly margin over feeding cost (Figure 1, curve with dots) remains over 200 € per 1000 L whatever the seasonal effect of the milk prices. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 560 Figure 1: feeding cost and margin over feeding cost. Discussion The results show the same tendencies as on the 20 pilot farms followed during 3 years within the CASDAR research programme (Carles, 2013). On those farms the milking frequency decreased by 0.24 milkings c-1 d-1 with a drop by 1.7 kg of milk d-1 during the grazing period. The amount of grazed grass in Derval reaches the same level as on the 20 pilot farms (average of 1500 kg DM c-1 y-1; range 750 to 2600), and 1500 kg DM is exactly the average grazed grass intake by the average French dairy cow (Brunschwig, 2011). In terms of working time, two tasks are directly related to grazing: at 8 am the cows milked between midnight and 6 am are sorted out, and in the evening the herd has to be fetched from the field. Using two paddocks per 24 hours (day and night paddocks) could facilitate the separation by the farmer of cows already milked from those not milked (Oudshoorn, 2008) but this requires two daily interventions to empty the night- and day-paddocks. To avoid fetching the cows, a system with 3 paddocks per 24 h like in Ireland or New-Zealand (Fitzgerald, 2012, and Woolford, 2004) could also be implemented. Conclusion The Derval experiment with a robot at full capacity, together with results from pilot farms, clearly show that grazing (even 100% grazing with no silage) can be combined with robotic milking. The feeding cost was reduced to one third: this limited the negative effect of the fluctuations of the milk price in spring. The same tendency was shown in the pilot farms. The success of the system lies in the trust of the cows in terms of traffic. Cows have a strong capacity to adapt to any traffic solution as long as we give them enough time to do it. Acknowledgements Experiments were funded by France-CASDAR and EU FP7-SME-2012-314879 -. AUTOGRASSMILK References Billon P. (2009) Traite des vaches laitières, Ed France Agricole, p 501-506. Brunschwig P. (2011) Observatoire de l’alimentation des vaches laitières, CNIEL-Idele 38 p. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 561 Burow E. (2011) The effect of grazing on cow mortality. Preventive Veterinary Medicine 100, 237-241. Carles A. (2013) MFE Bordeaux Sci. Agro, IDELE, (in press). Fitzgerald S. (2012) Irish dairying planning for 2015, 116-118. Jégou V., Grasset M., Huneau T. and Séité Y. (2007) Cap’Elevage 19, 6. Oudshoorn F (2008) Mobile milking robot offers new grazing concept. Grassland Science in Europe13, 721-723 Spörndly E. (2004) Automatic milking and grazing—effects of distance to pasture and level of supplements on milk yield and cow behavior. Journal of Dairy Science 87, 1702-1712. Woolford M., Claycomb R., Jago J. et al. (2004) Automatic dairy farming in New Zealand using extensive grazing systems. In: A. Meijering et al. (eds) Automatic Milking: a better understanding, pp. 280-285. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 562 GPS tracking of Old Norwegian ewes on a coastal heathland-dominated island Lind V. and Bär A. Norwegian Institute for Agricultural and Environmental Research, Arctic Agriculture and Land Use division, N-8861 Tjøtta, Norway Corresponding author: vibeke.lind@bioforsk.no Abstract Coastal heathland is an endangered habitat type, which in Norway is maintained by grazing with Old Norwegian sheep. The breed is well adapted to graze the habitat type all year round. While there is excess of feed during the summer, the availability of winter feed can be a challenge, and in some years additional feed is necessary. In this paper, we evaluate the diet selection of Old Norwegian ewes grazing a coastal heathland-dominated island in Northern Norway during June/July, September and January/February. Twenty-eight ewes were equipped with GPS-collars and their position registered every six hours. Faeces samples from the ewes were collected during the three periods and analysed using micro-histology. In addition, the island was vegetation mapped. The three data sets allow us to suggest which species the ewes graze and how they utilize the pasture of the island. We conclude that during summer there is enough available feed for the 28 ewes, whereas during winter we expect the island to be over grazed. Keywords: native sheep, micro-histology, diet selection Introduction The Old Norwegian breed is the native breed of Norway and belongs to the group ‘Northern short-tailed breeds’. The breed is most common in coastal heathland habitats of Norway. The European coastal heathlands are considered an endangered habitat type (EU Habitat Directive 92/43/EEC) threatened by climate change, air pollution, land-use change and abandonment (Moen et al., 2006). There are special needs to protect the coastal heathland from becoming extinct as a habitat type. Grazing this habitat with Old Norwegian sheep has traditionally been one of the most important methods for maintenance. The sheep graze all year round and the challenge in this system is available winterfeed. Old Norwegian sheep are better adapted to the climate and poor vegetation during winter than domesticated breeds and their metabolism allows the animals to feed on woody forage like leaves, twigs and bark. In Norway, shelter is demanded for animals kept outdoors during winter and, in cases of feed shortage, the animals must have access to additional feed (Lovdata FOR-2005-02-18-160). There are few observations on natural diet selection during winter. Norderhaug and Thorvaldsen (2011) found that Calluna vulgaris and Carex sp are important species during the winter. In this paper, we evaluate the diet selection of Old Norwegian ewes grazing a coastal heathland-dominated island in Northern Norway during June/July, September and January/February. Materials and methods The island, Risvaer, is located in the Luroey community in Northern Norway (66.3oN, 12.6oE). The island, 48 ha, was vegetation mapped in 2011 following the classification system by Fremstad (1997). Seven permanent plots (1×1 m) were established for registration of species composition and their cover (%). Twenty-eight Old Norwegian ewes grazed the island from June 2012 until June 2013. The ewes were equipped with GPS-collars (Telespor®) logging their position every six hours. Kernel-Density maps were estimated on the points for each of three periods (June/July 2012, September 2012 and January/February 2013). Fresh faeces samples from 9, 6 and 10 ewes were collected in June, September and January, respectively, Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 563 for micro-histological analyses. The samples were analysed using the Garcia-Gonzalez method (Garcia-Gonzalez, 1984). Vegetation types and GPS-data from June/July, September and January/February (Figure 1) were compared on maps using ArcMap 10.1. Micro-histological analyses were compared against the maps to explain the different grazing patterns during the year. Results and discussion Risvaer is dominated by species-rich heathland vegetation. Usually, Calluna is the dominating species in coastal heathlands. However, under rich soil conditions, as in the southern part of the island, herbs and graminoides represent a considerable proportion at the expense of Calluna. In contrast, Empetrum nigrum, Vaccinium uliginosum and Calluna are dominating the heathland in the northern part of the island. Juniperus communis is also an important species at Risvaer. Abandoned, former cultivated grassland is the second-most important vegetation type dominated by graminoids such as Phleum pratensis, Agrostis spp. and Carex nigra, depending on variations in the soil moisture content. A shelter for the animals is placed in this area. We compared the GPS-data with the vegetation and the faeces samples. The GPS-data indicate where the ewes preferred to be during the three periods presented in this paper. Comparison of the three data sets allows us to suggest how the ewes utilize the pastures at Risvaer and which species they graze at different times of the year. Figure 1. Kernel-Density map showing position of ewes in (A) June/July 2012, (B) September 2012 and (C) January/February 2013. The darker spots indicate more ewes staying in this area. The map for June (Figure 1A) shows that the ewes stayed mainly on the grass-dominated area early in the summer, in order to find the first green shoots from Poaceae and herbs, which are high in nutrients. During summer the ewes spread over the island (Figure 1B) without any preference in type of vegetation; this is in contrast to winter (Figure 1C) when the ewes moved less and were found mainly in the heathland vegetation south at the island or around the shelter. The most abundant species found in the faeces analyses are presented in Table 1. The analyses do not take into account digestibility. Hence, differences in species portion can only be compared between different sample times but not within the same sample. However, analyses of faeces samples show that graminoids and herbs were important in June. In September, graminoids dominated the diet with a decline of Carex spp and herbs compared to June. In January, however, the diet changed considerably, and J. communis and Carex spp. seemed to be important. Juncus communis is a species that is rarely grazed due to its low nutritional value and high fibre content. Calluna, on the other hand, is considered an important species for Old Norwegian sheep during winter (Norderhaug and Thorvaldsen, 2011). Since this species is Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 564 scarcely represented at Risvaer, we suggest the intake of J. communis to be compensating for the lack of Calluna. Table 1. Average proportion (%) of micro-histological material in faeces samples for important species in June 2012, September 2012 and January 2013. Herbs Graminoids (total) Festuca spp. Deschampsia flexuosa Arrenatherum elatius Agrostis spp. Molinia caerulea Carex spp. Heather Calluna vulgaris Cypress Juniperus communis June 2012 (N = 9) 20.5 78.1 46.1 4.8 1.8 2.8 2.7 12.8 September 2012 (N = 6) 7.9 89.4 67.7 5.5 1.4 1.4 0.3 9.6 January 2013 (N = 10) 5.4 63.9 29.7 1.0 0.2 1.2 0.6 29.4 - 0.6 1.2 0.2 - 26.5 On the other hand, the preference of J. communis can indicate overgrazing during winter. In January, the high proportion of Carex spp. in the faeces samples is in accordance with Norderhaug and Thorvaldsen (2011) and Carex are considered important species during winter. Carex spp. were grazed in the heathland and not in the boggy areas of cultivated grassland where Carex spp are the dominating species. Possibly, these areas are too wet during winter, which is why the ewes prefer to graze the species in the drier heathland areas in the southern part of the island. References Fremstad E. (1997) Vegetasjonstyper i Norge [in Norwegian]. NINA-temahefte 12: 1-279. Garcia-Gonzales R. (1984) L’emploi des epidermes végétaux dans la determination du regime alimentaire de l’Isard dans les Pyrénées occidentales. Écologie des Milieux Montagnards et de Haute Altitude [in French]. Documents d’Écologie Pyrénéenne III-IV, 307-313. Moen A., Nilsen L.S., Aasmundsen A. and Oterholm, A.I. (2006) Woodland regeneration in a coastal heathland area in central Norway. Norwegian Journal of Geography 60, 277-294. Norderhaug A. and Thorvaldsen P. (2011) Variasjon i beitepreferanse gjennom året hos utegangersau på kystlynghei [in Norwegian]. In Brodin, J and Fog, M.O. (eds) Husdyrforsøksmøtet 2011, p. 369-372. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 565 Nutritive value of leaf fodder from the main woody species in Iceland Hejcman M.1, Hejcmanová P.1, Pavlů V.1, and Thorhallsdottir A.G.2 1 Czech University of Life Sciences, Kamýcká 129, CZ-165 21 Prague, Czech Republic 2 Agricultural University of Iceland, Hvanneyri, IS-311 Borgarnes, Iceland Corresponding author: hejcmanova@ftz.czu.cz Abstract In the past, leaf fodder from woody species played an important role as animal feed in Iceland. However, very limited information exists on the nutritive value of the main woody species used as fodder. The aim of our study was therefore to determine forage quality of leaves of Betula nana, B. pubescens, Salix lanata, S. phylicifolia and Sorbus aucuparia and to compare it with forage quality of a common native grass, Deschampsia cespitosa, and with an introduced grass, Alopecurus pratensis, used by contemporary Icelandic farmers for forage production. In late June 2013, we collected samples of all species at four localities in Iceland and determined concentrations of nitrogen, phosphorus, neutral and acid-detergent fibre and lignin and compared them with the optimum range required for sheep nutrition. Nutritive value of leaves of woody species was relatively high and browsing of their leaves and collection of their leaf fodder for winter feeding could satisfy sheep/cattle nutritional requirements for N and P. However, the high content of indigestible lignin, present in all the woody species, functions as a barrier to nutrient digestibility. Grasses were characterized by lower P. The forage quality of leaves of woody species increased in the order B. nana < B. pubescens < S. phylicifolia < S. aucuparia < S. lanata. Keywords: livestock feeding, forage quality, North Atlantic Isles Introduction In northern regions, human subsistence in the past was almost exclusively based on animal products. The colonization of the North Atlantic islands, including Iceland, by Norse settlers (AD 800 – 1000) was thus connected with the spread of livestock; namely cattle, sheep, horses and goats (Amorosi et al., 1997). Livestock browsing was, together with wood collection and charcoal production (Church et al., 2007), one of the reasons for forest degradation, because livestock diets were, in addition to grassland forage, based on leaves, generative organs, bark and the annual twigs of woody species (Gauthier et al., 2010). In Iceland, Norse colonization followed by year-round livestock grazing and browsing has been considered one of the main reasons for the decline of Betula pubescens forests at the beginning of the 20th century (Ólafsdóttir and Guðmundsson 2002). Although woody species played, and in some Nordic regions still play, an important role in livestock feeding, information on the nutritive value of leaves (i.e. nitrogen, phosphorus and fibre fractions) of common woody species has been missing (see though Sigvaldason (1967) for earlier estimates). The aim of this study was to determine forage quality of leaves of common woody species (Betula nana, B. pubescens, Salix lanata, S. phylicifolia and Sorbus aucuparia) in Iceland and to compare it the with forage quality of two grasses, both common in old hayfields in Iceland: the native Deschampsia cespitosa and the high-yielding grass Alopecurus pratensis, introduced from Europe (Helgadóttir et al., 2013; Kristinsson, 2013). Material and methods We collected leaf biomass of five common broad-leaved woody species (Betula nana, B. pubescens, Salix lanata, S. phylicifolia and Sorbus aucuparia) and two grass species (Deschampsia cespitosa and Alopecurus pratensis) at four localities in Iceland in late June 2013. A total 28 biomass samples were oven-dried at 60 °C for 48 hours, ground to powder and analysed for the concentration of nitrogen (N), phosphorus (P) and the content of neutral- (NDF) Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 566 and acid-detergent fibre (ADF) and acid-detergent lignin. The N concentration in the plant samples was determined using an automated analyser TruSpec (LECO Corporation, USA) and P concentration using ICP–OES (Varian VistaPro, Mulgrave, Vic., Australia) after mineralization using aqua regia of burnt samples in in a microwave oven at a temperature of 550 °C. Contents of NDF, ADF and ADL were determined by standard methods of AOAC (1984). All analyses were performed in an accredited national laboratory, Ekolab Žamberk (http:// www.ekolab.zamberk.cz). Data were tested by one-way ANOVA followed by post-hoc comparison using the post-hoc HSD Tukey’s tests to identify differences in concentrations of N, P, NDF, ADF and lignin contents among species. Results and discussion All values of concentrations of N, P, NDF, ADF and lignin are given in Table 1. Table 1. Concentration (means ± standard error of mean) of N, P, neutral detergent fibre (NDF), acid detergent fibre (ADF) and lignin in leaf biomass of studied species. Calculated by one-way ANOVA followed by Tukey post-hoc comparison test, significant differences (P<0.01) among species are indicated by different letters. Chemical properties of fodder from Icelandic grasslands (with dominant Alopecurus pratensis, Poa pratensis, Phleum pratense and Agrostis capillaris) follow Ragnarsson and Lindberg (2010) and Thorvaldsson et al. (1998) and optimum range for sheep and cattle follows Whitehead (1995). Species Betula nana Betula pubescens Salix lanata Salix phylicifolia Sorbus aucuparia Alopecurus pratensis Deschampsia caespitosa Grassland forage optimum range for sheep/cattle N (g kg-1) 24.1±1.0a 28.7±1.4 a 27.5±0.9 a 27.7±2.2 a 26.5±1.3 a 23.2±1.5 a 25.1±0.8 a 19-32 P (g kg-1) 2.6±0.2 a 3.1±0.2 ab 3.8±0.4 b 3.9±0.3 b 3.0±0.3 ab 2.5±0.2 a 2.0±0.2 a 1.9-2.4 NDF (g kg-1) 328±23 ab 315±16 ab 376±25 c 272±28 a 285±17 ab 631±22 d 614±12 d 500-550 ADF (g kg-1) 305±15 abc 294±16 abc 345±22 bd 243±15 a 269±12 ab 388±13 d 324±6 bcd 270-310 Lignin (g kg-1) 154±7 d 123±10 cd 96±6 bc 119±19 cd 95±3 bc 65±7 ab 38±2 a 30-50 19.2 - 25.6 2.3 - 3.7 330-450 190-300 max. 80 Concentrations of N were similar and concentrations of P, NDF, ADF and lignin were significantly different among species (Table 1). Nitrogen and phosphorus concentrations were also within the optimum range for nutrition of cattle in all analysed species, with the exception of the too-low P concentration in D. cespitosa and a slightly higher P concentration in S. phylicifolia. Optimum NDF content for sheep and cattle nutrition was recorded only in S. lanata, whereas optimum content of ADF was recorded in B. pubescens, S. phylicifolia and S. aucuparia. Content of lignin was substantially higher in woody species than in grasses. Relatively high lignin content in all woody species in comparison with grasses could be the most problematic for livestock metabolism, because digestibility of the biomass generally decreases with an increase in lignin content (Cherney et al., 1993). On the other hand, woody species offer considerable amounts of indispensable nutrients, particularly N and P. The nutritive value of leaves of the main woody species in Iceland was relatively high in comparison with the dominant broad-leaved woody species in Central Europe (Hejcmanová et al., 2013) and could satisfy livestock requirements. The forage quality of leaves of woody species increased in the order B. nana < B. pubescens < S. phylicifolia < S. aucuparia < S. lanata. Both Betula species are browsed by Icelandic sheep, but B. pubescens substantially more and mainly in the spring and early summer (Thorhallsdottir and Thorsteinsson, 1990) This pattern is in agreement with higher forage quality of B. pubescens than of B. nana and this probably explains why B. pubescens was harvested in the past for leaf fodder by Icelandic farmers while B. nana was rarely used (Gunnlaugsson, 1969). Very high P concentration was recorded in the leaves of both Salix species, probably due to lower temperatures and slower plant growth (Reich and Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 567 Oleksyn, 2004). We recorded that leaves of both Salix species were browsed by sheep in high quantity, much more than B. nana or B. pubescens, and that free-ranging sheep seem to be able to prevent regeneration of Salix shrubs in some Icelandic regions. Browsing of Salix species probably helps the animals to avoid P and Ca deficiency, especially high-milk-yielding ewes. Salix species are generally considered to be the best forage woody species in Nordic regions (Forbes et al., 2010; Myking et al., 2013). Conclusions The nutritive value of leaves of the main woody species in Iceland was relatively high and can satisfy livestock requirements for N and P. The most problematic for livestock metabolism would be the relatively high lignin content in all woody species, in comparison with grasses. Acknowledgements We gratefully acknowledge the valuable advice of Aslaug Helgadottir. The study was funded by grant from Czech University of Life Sciences Prague, CIGA 20114205. References Amorosi T., Buckland P., Dugmore A., Ingimundarson J.H. and McGovern T.H. (1997) Raiding the landscape: human impact in the Scandinavian North Atlantic. Human Ecology 25, 491–517. Cherney D.J.R., Cherney J.H. and Lucey R.F. (1993) In vitro digestion kinetics and quality of perennial grasses as influenced by forage maturity. Journal of Dairy Science 76, 790–797. Church M.J., Dugmore A.J., Mairs K.A., Millard A.R., Cook G.T., Sveinbjarnardóttir G., Ascough P.A. and Roucoux K.H. (2007) Charcoal production during the norse and early medieval periods in Eyjaflallahreppur, southern Iceland. Radiocarbon 49, 659–672. Gauthier E., Bichet V., Massa C., Petit C., Vanniere B. and Richard H. (2010) Pollen and non–pollen palynomorph evidence of medieval farming activities in southwestern Greenland. Vegetation History and Archeobotany 19, 427–438. Gunnlaugsson T. (1969) Notes on leaf fodder and leaf fodder collection (Um laufhey og laufheyskap) Ársrit Ræktunarfélags Norðurlands 66:76-84 (in Icelandic) Helgadóttir A., Eythórsdóttir E., Jóhannesson T. (2013) Agriculture in Iceland – A grassland based production. Grassland Science in Europe 18, 30–43. Kristinsson H. (2013) Flowering plants and ferns of Iceland. Mál og menning, Reykjavík. Myking T., Solgerg E.J., Austrheim G., Speed J.D.M., Bohler F., Astrup R. and Eriksen R. (2013) Browsing of sallow (Salix caprea L.) and rowan (Sorbus aucuparia L.) in the context of life history strategies: a literature review. European Journal of Forest Research 132, 399–409. Ólafsdóttir R. and Guðmundsson H. (2002) Holocene land degradation and climatic change in northeast Iceland. The Holocene 12, 159–167. Ragnarsson S. and Lindberg J.E. (2010) Nutritional value of mixed grass haylage in Icelandic horses. Livestock Science 131, 83–87. Reich P.B. and Oleksyn J. (2004) Global patterns of plant leaf N and P in relation to temperature and latitude. PNAS 101, 11001–11006. Sigvaldason J. (1967) What is leaf fodder? (Hvað er laufhey?) Ársrit Ræktunarfélags Norðurlands 64:79-82 Thorhallsdottir A.G. and Thorsteinsson I (1993) Behaviour and plant selection. Icel Agr Sci (Búvísindi) 7,59-77 Whitehead D.C. (1995) Grassland Nitrogen. Wallingford, CAB International, 397 pp. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 568 Sensory quality and authentication of lamb meat produced from legume-rich forages Devincenzi T.1,3, Prunier A.2, Nabinger C.3 and Prache S.1 1 Institut National de la Recherche Agronomique (INRA), UMR1213 Herbivores, F-63122 StGenès-Champanelle, 2 UMR1348 PEGASE, F-35590 Saint-Gilles, France. 3 Universidade Federal do Rio Grande do Sul (UFRGS), Av. Bento Gonçalves 7712, CEP 91501-970, Porto Alegre-RS, Brazil. Corresponding author: sophie.prache@clermont.inra.fr Abstract We investigated the dose-dependent response to dietary alfalfa:cockfoot proportion in grazing lambs of stable nitrogen isotope ratio in lamb meat and of fat skatole concentration and chop sensory attributes, and the ability of nitrogen isotope signature of the meat to authenticate meat produced from legume-rich diets. Four groups of nine male Romane lambs grazing a cocksfoot pasture were supplemented with different levels of fresh alfalfa forage to obtain four dietary proportions of alfalfa (0%, 25%, 50% and 75%) for 98 days on average before slaughter (groups U, L, M and H). The distribution of the δ15N values of the meat discriminated the U lambs from the H lambs, and gave a correct classification score of 85.3% comparing lambs that ate alfalfa with those that did not. Perirenal fat skatole concentration decreased linearly with the δ15N value of the muscle, indicating a linear increase with the dietary proportion of alfalfa. The intensity of ‘animal’ flavour increased from the lowest level of alfalfa onwards and did not increase further with increasing levels of alfalfa, suggesting that this sensory attribute reaches a plateau when perirenal fat skatole concentration is in the range 0.26–0.34 µg/g fat. Keywords: authentication, flavour, grazing, legumes, scatole, sheep Introduction Low-input and organic farming livestock systems embody features that consumers value, such as animal welfare, food healthiness and environmental acceptability. The ability to authenticate food products from these agro-ecological systems has therefore become an important challenge. Nitrogen (N) isotope signature has been proposed as a valuable tool to authenticate meat produced from organic beef (Bahar et al., 2008). The presence of forage legumes in organic systems is actually of major importance, because nitrogen-fixing plants improve pasture quality and reduce dependency on external inputs, and the dietary 15N/14N ratio is lower in legume-rich pastures due to the capacity of leguminous plants to fix atmospheric nitrogen. However, the occurrence of off-flavours and off-odours in the meat has been shown to increase in lambs grazing legume-rich pastures (Young et al., 2003). Flavours in lamb meat described as animal, pastoral and faecal have been related to the presence of skatole, which is an aromatic compound produced in rumen from the tryptophan amino acid. Its ruminal synthesis increases when the pasture is rich in forage species with high degradable protein content. This study therefore investigated i) the dose-dependent response to dietary alfalfa levels in grazing lambs of stable nitrogen isotope ratio in lamb meat and of fat skatole concentration and chop sensory attributes, and ii) the ability of nitrogen isotope signature of the meat to authenticate meat produced from legume-rich diets. Materials and methods Four groups of nine male Romane lambs grazing a cocksfoot pasture were supplemented with fresh alfalfa forage to obtain four dietary proportions of alfalfa (0%, 25%, 50% and 75%) for 98 days before (groups U, L, M and H). The mean daily voluntary dry matter intake was Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 569 estimated for each group at 78.1 g DM/BW0.75. Alfalfa was cut every morning and offered half at 9 a.m., and half at 4 p.m. after storage at 4 °C. We measured the 15N/14N ratio in the forages and in the longissimus thoracis et lumborum (LTL) muscle by isotope ratio mass spectrometry. We analyzed the data using variance analysis (GLM procedure of SAS) and linear discriminant analysis followed by a cross-validation procedure to classify the meat samples according to feeding treatments, using Minitab software v.13. Perirenal fat skatole concentration was measured by HPLC. Lamb chop flavour and odour was evaluated by experienced panellists. Six chops per lamb were grilled to an internal temperature of 75 °C and served to 12 panelists. Two pieces from the lean part and two pieces from the fat part were cut from each chop to provide a piece of each for each panellist. At each session, the panellists evaluated one lamb per treatment, with lambs from each treatment presented in randomized order. A preliminary session was performed using additional chops from one lamb with the highest skatole concentration and one lamb with the lowest skatole concentration, for the panellists to agree on common terms describing the perceptions related to the presence of skatole. These were ‘animal’ flavour and ‘animal’ odour. The data for lamb chop sensory evaluation underwent an ANOVA analysis using a mixed model, with treatment and panel session as fixed factors and panellist as a random factor, and using the Bonferroni test for pairwise comparisons. Full details are given in Devincenzi et al. (2014a and b). Results and discussion Mean daily alfalfa intake was 0, 272, 550 and 731 g DM, which corresponded to a mean dietary proportion of alfalfa of 25.8%, 48.7% and 62.4% for L, M and H lambs respectively. The δ15N value of the LTL muscle decreased linearly with the dietary proportion of alfalfa (PA, %) (P = 0.024), the regression equation being: δ15N value of the LTL muscle = 6.04 (±0.071) − 0.0107 (±0.00169) PA, where r2 = 0.95, RSD = 0.080, and n = 4. The distribution of the δ15N values of the meat discriminated all the U lambs from the H lambs, and gave a correct classification score of 85.3% comparing lambs that ate alfalfa with those that did not. Perirenal fat skatole concentration was higher for L and M lambs than for U lambs (Figure 1A). Surprisingly, perirenal fat skatole concentration in H lambs was lower than that of M lambs, and it was not statistically different from that of U lambs. As this result may be partly linked to the variability between individual animals in dietary choices, we used the δ15N value of the LTL muscle as an indicator of the dietary proportion of alfalfa at the individual level and explore the dose-dependent response further. Perirenal fat skatole concentration decreased linearly with δ15N value of the LTL muscle (P < 0.03, Figure 1C), supporting the hypothesis that the fat skatole concentration increases linearly with the dietary proportion of alfalfa. The intensity of ‘animal’ odour in the lean part of the chop and of ‘animal’ flavour in both the lean and fat parts of the chop increased from the lowest level of alfalfa supplementation onwards and did not increase further with increasing levels of alfalfa supplementation (Figure 1B). The outcome of this study therefore suggests that these sensory attributes may reach a plateau when perirenal fat skatole concentration is in the range 0.16-0.24 µg/g of liquid fat. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 570 (A) Perirenal fat skatole concentration (μg/g liquid fat) B b ab a (B) Treatment B A flavour Intensity of the 'animal' B B b b b A Treatment Lean part Fat part (C) Figure 1: (A) Mean value of perirenal fat skatole concentration according the dietary level of alfalfa in grazing lambs. Means with unlike superscripts differ (A, b: P < 0.05; a, b: P < 0.07). (B) Mean intensity of ‘animal’ flavour in the lean and fat parts of the chops according the dietary level of alfalfa in grazing lambs. For the lean and the fat parts, means with unlike superscripts differ (A, b: P < 0.05; A, B: P < 0.01). Bars represent standard error of the mean (SD/√n), where n is the number of lambs in each group. (C) Relationship between perirenal fat skatole concentration and δ15N value of the longissimus thoracis et lumborum muscle. White circles refer to unsupplemented lambs, black circles, triangles and squares refer to groups of lambs receiving a low, medium or high level of alfalfa supplementation respectively. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 571 Conclusions Perirenal fat skatole concentration was higher for lambs that consumed alfalfa than for those that consumed only cocksfoot, and it increased as soon as the dietary proportion of alfalfa reached 25%. The intensity of ‘animal’ odour in the lean part of the chop and of ‘animal’ flavour in both the lean and fat parts of the chop were increased from the lowest level of alfalfa supplementation onwards and did not increase further with increasing levels of alfalfa supplementation. The outcome of this study therefore suggests that these sensory attributes may reach a plateau when perirenal fat skatole concentration is in the range 0.16-0.24 µg/g of liquid fat. The distribution of the δ15N values of the LTL muscle discriminated all the U lambs from the H lambs, and gave a correct classification score of 85.3% comparing lambs that ate alfalfa with those that did not. These results may be of interest for the authentication of meat produced in low-input and organic production systems, in which leguminous plants are more widespread. References Bahar B., Schmidt O., Moloney A. P., Scrimgeour C.M., Begley I.S. and Monahan F.J. (2008) Seasonal variation in the C, N and S stable isotope composition of retail organic and conventional Irish beef. Food Chemistry 106, 1299–1305. Devincenzi T., Delfosse O., Andueza D., Nabinger C. and Prache S. (2014a) Dose-dependent response of nitrogen stable isotope ratio to proportion of legumes in diet to authenticate lamb meat produced from legume-rich diets. Food Chemistry 152, 456-461. Devincenzi T., Prunier A., Nabinger C. and Prache S. (2014b) Dose-dependent response of fat skatole concentration and chop flavor and odour attributes to dietary alfalfa levels in grazing lambs. Meat Science (submitted) Young O.A., Lane G.A., Priolo A. and Fraser K. (2003) Pastoral and species flavour in lambs raised on pasture, lucerne or maize. Journal of the Science Food and Agriculture 83, 93–104. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 572 Dynamics of dry matter intake in livestock production systems in the Netherlands Van den Pol-van Dasselaar A.1, Nolles J.E.2, Philipsen A.P.1 and Stienezen M.W.J.1 1 Wageningen UR Livestock Research, P.O. Box 65, 8200 AB Lelystad, the Netherlands, 2 CAH Vilentum, University of Applied Sciences, De Drieslag 4, 8251 JZ Dronten, the Netherlands Corresponding author: agnes.vandelpol@wur.nl Abstract Pasture-based dairy systems have several advantages. However, the number of grazing dairy cattle in the Netherlands is decreasing, partly as a result of scaling. Therefore, grazing research focuses on providing farmers, students and advisers with tools to optimize grazing whilst also improving efficiency. The aim of this study was i) to estimate the dry matter intake (DMI) from grazing on commercial dairy farms and the variation of this DMI throughout the season, ii) to estimate the associated feeding costs, and iii) to present this in a practical way. A DMIdashboard was developed to get day-to-day insight in DMI and feeding costs. It was tested on nine commercial dairy farms. Data show that there was a huge difference in DMI by grazing (800 to 1900 kg DM cow-1 yr-1). The between-farm variation in feeding costs was much larger than the within-year variation and is a good reflection of the differences in DMI by grazing between the different individual farms. By comparing results of individual farms in a network setting, the insight in the effect of grazing increased. This led to increased skills of farmers, students and advisers in optimizing grazing systems. Keywords: Dry matter intake (DMI), feeding costs, grazing, pasture, skills Introduction The number of grazing dairy cattle in the Netherlands has decreased in recent years, from 90% the total in in 2001 to 70% in 2012 (CBS, 2013). In 2012, the ‘Convenant Weidegang’ (‘Treaty Grazing’) was therefore initiated. The aim of the Treaty is to stabilize the percentage of farms that practise grazing. At the end of 2013 almost 60 parties in the Netherlands had signed this Treaty, including dairy farmers, dairy industry, feed industry, banks, accountant, semen industry, veterinarians, cheese sellers, retail, NGOs, nature conservation, government, education and science. As a result of the Treaty, many research activities have been initiated to optimize grazing and grassland management. In the last few years, the theme 'grazing' has also become more dominant in education. The interaction between education and science increasingly leads to participation of lecturers and students in scientific research on grazing, which in turn leads to an improved educational programme. Grazing is, on average, economically attractive (Van den Pol-van Dasselaar, 2014) and is preferred by society. However, dry matter yields are lower, nutrient losses are higher and often farmers experience grazing as difficult. This is especially true in the Netherlands with its mixed dairy systems that use relatively high levels of supplementation. Most farmers and advisers lack the necessary skills to manage grazing properly in such high-output systems. Support is needed and this support should be robust, simple and appealing (Van den Pol-van Dasselaar et al., 2012). The focus of grazing research in the Netherlands is to provide farmers, students and advisers with the tools to optimize grazing under Dutch conditions whilst also improving efficiency. The aim of this study was i) to estimate the dry matter intake (DMI) from grazing on commercial dairy farms and the variation of this DMI throughout the season, ii) to estimate the associated feeding costs, and iii) to present this in a practical way. Currently, these data are not available Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 573 on commercial dairy farms in the Netherlands. And if one does not measure, one cannot manage. Materials and methods In 2012 and 2013, participatory research on nine commercial dairy farms of the network ‘Dynamisch Weiden’ (Dynamic Grazing) was carried out to study the dynamics of grazing and to develop practical management tools. Researchers, advisers and dairy farmers together searched for suitable grazing systems. Farmers were encouraged to optimize their grazing system and to gain new knowledge and skills. The work was supported by students. The farms differed in scale, soil type, grazing intensity (cows ha-1, hours day-1, days on one paddock). Grazing and supplemental feeding was recorded in a grassland calendar. A DMI-dashboard was developed to get insight in day-to-day DMI and feeding costs. Total feed demand was estimated using characteristics of the herd, like number of cows, milk yield per cow, fat and protein content of the milk. The grass intake was calculated by subtracting the supplemental feed from the total feed demand. The feeding costs were calculated using the actual feed costs for supplemental feeding and estimated feed costs for home-grown forage. Results and discussion The dairy farmers in the network were eager to get insight in the results. They confirmed that for them the difficulty of grazing is the variation in milk production due to the unknown variation in grass growth, grass quality and DMI. A tool to get insight in more accurate grass intake was therefore considered to be very useful. An example of such a tool is the DMIdashboard, which has been developed in the project (Figure 1). Figure 1. The DMI-dashboard for an individual farm in the growing season 2012. The DMI-dashboards of the nine commercial dairy farms showed that the DMI from grazing varied between 800 and 1900 kg DM cow-1 yr-1. A recent Dutch study into the economics of grazing has shown that grazing is financially attractive if the cows eat sufficient amounts of fresh pasture grass ( > 600 kg DM cow-1 yr-1). If the intake of fresh grass is very low, grazing is not profitable (Van den Pol-van Dasselaar et al., 2014). The feeding costs were determined for the nine farms throughout the season. Figure 2 shows that the between-farm variation in feeding costs was much larger than the within-year variation. The between-farm variation Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 574 reflects the differences in DMI by grazing between the different individual farms. The dairy farmers in the study experienced these results as extremely valuable even though the costs of some individual feed components were only estimated. The within-year variation and the between-farm variation provided an opportunity for them to benchmark their individual farm results, which contributed to overall improvement of the grazing system. The effect of management options, e.g. ration changes, was immediately visible and operational management could be adjusted accordingly. Figure 2. Feeding costs in May, June, July and August 2013 for nine commercial dairy farms. Conclusion Today, many farmers in the Netherlands do not focus on grassland in their operational farm management. By comparing results of individual farms in a network setting, the insight in the effect of grazing increased. This has led to increased skills of farmers, students and advisers in optimizing grazing systems. Acknowledgements This project was funded by the Stichting Weidegang and the Ministry of Economic Affairs in the Netherlands. References CBS (2013) StatLine databank, http://statline.cbs.nl/. Van den Pol-van Dasselaar A., de Haan M.H.A., Holshof G. and Philipsen A.P. (2012) Requirements for Decision Support Tools for grazing. Grassland Science in Europe 17, 786-788. Van den Pol-van Dasselaar A, Philipsen A.P and de Haan M.H.A. (2014) Economics of grazing. Grassland Science in Europe 19 (these proceedings). Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 575 Cutting strategy of a five-cut system in different grassland mixtures Søegaard K. Department of Agroecology, Aarhus University, PO Box 50, 8830 Tjele, Denmark Corresponding author: Karen.Soegaard@agrsci.dk Abstract A stronger focus on the nutritive value throughout the growing season is needed to improve the utilization of annual grass-clover production. A field experiment tested five different cutting strategies with a total of five cuts per year and with different timings of the spring cut and duration of regrowth in three mixtures. The cutting strategies highly influenced seasonal growth and nutritive value and only slightly influenced the annual dry matter yield. Delaying the spring cut decreased the annual IVOMD and percentage of crude protein, whereas the annual percentage of NDF increased. Thus, the higher nutritive value in the shorter regrowth periods later in the season could not make up for the lower value in the late spring cut. The annual white clover content decreased and red clover content increased with a later spring cut. The results indicate that especially with mixtures containing red clover a planned cutting strategy is necessary. Keywords: cutting strategy, red clover, white clover, nutritive value, digestibility Introduction There is often a keen focus on optimizing the nutritive value of the spring cut and less focus on the rest of the growing season. However, the complete herbage production of the year is used on the farm for feeding different cattle groups with different requirements. The challenge is to optimize the nutritive value of that part of the seasonal production which is planned for feeding to high-yielding dairy cows. Herbage cuts are often co-ensiled in horizontal silos, which is one of the reasons why a planning tool for timing the cutting time, and which also indicates which cuts should be co-ensiled, could be useful. By comparing different feeding trials, Weisbjerg et al. (2011) showed that milk yield did not increase further when in vitro organic matter digestibility (IVOMD) of the grass-clover part of the feeding ration increased above 78%, and that the point was independent of stage in the growing season. This means that IVOMD can be used as a goal when planning the seasonal cutting strategy. Here results are presented from an experiment with different cutting times and durations of regrowth with a focus on the yearly mean feeding quality. Materials and methods Five different harvest strategies with five cuts per year (Figure 1) were tested on a sandy loam at Foulum in Denmark in three mixtures (26 kg seed ha-1), made up of either (1) 13% white clover (Trifolium repens L.) and 87% perennial ryegrass (Lolium perenne L.), (2) 11% red clover (Trifolium pratense L.), 7% white clover, 37% perennial ryegrass and 45% festulolium (Festulolium braunii K.A.) or (3) pure red clover only, cultivated for strategies 2 and 4. The swards were fertilized with 100 kg N ha-1 in spring and 60 kg N ha-1 after the second cut. There were four replicates. The grass-clover was undersown in spring barley (Hordeum vulgare) in 2011. Plots were harvested with a Haldrup plot harvester, nutritive value was determined by NIR, IVOMD calibrated to the method of Tilley and Terry, and botanical composition of dry matter (DM) was determined by hand separation of subsamples. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 576 Strat. May June a) 1 July 5 2 b) 3 b) 4 4 5 5 5 d) 5 September 5 c) 4 August 5 5 5 6 6 6 4 6 4 4 Oct. 5 5 4 Figure 1. Harvest strategy. Date of spring cut and length of regrowth period in number of weeks. Date for spring cut in 2012/2013: a) 14/23 May, b) 22/28 May, c) 29 May/4 June and, d) 6/11 June. Results and discussion 8 85 6 80 IVOMD t DM/ha The two harvest years differed weather-wise. The 2013-season was warmer and drier than 2012. However, the principal differences were comparable. Harvest time strongly influenced the seasonal profile of dry matter production and quality parameters (Figure 2). Early harvest in spring gave a more even seasonal production profile and the IVOMD was highly affected by both time of spring harvest and duration of regrowth. The strategies were planned to reach different goals. Strategy 3 aimed for an even nutritive value throughout the season, and therefore the third regrowth in late summer was short due to the normally higher temperature in this period, which gives a lower digestibility of organic matter. In contrast, in strategy 2 the third regrowth period was long and the durations of regrowth 1 and 2 were short, with the goal of optimizing the nutritive value before the warm period and then to produce herbage with a lower nutritive value in the warm period. For strategy 3 the goal was only partly reached, as IVOMD in the second regrowth was relatively low. Harvest time for this 5-week regrowth period was thus too late in the summer. 4 2 75 70 65 0 12345 Strategy 1 12345 Strategy 2 12345 Strategy 3 12345 12345 Strategy 4 Strategy 5 12345 12345 12345 12345 12345 Strategy 1 Strategy 2 Strategy 3 Strategy 4 Strategy 5 Figure 2. Dry matter yield (left) and digestibility of organic matter (right) per cut in mixture 2 (perennial ryegrass, festulolium, white and red clover). Mean of 2012 and 2013. Despite the very different cutting strategies, the annual yields were very similar (Table 1). The exception was strategy 4, with a lower annual yield due to poor growth in the last week before spring cut in both years. Including red clover and festulolium in mixture 2 increased the annual yield by an average of only 0.9 t DM ha-1. IVOMD was lower in mixture 2 than in mixture 1 and lowest in pure red clover (mixture 3). Red clover often has a lower IVOMD than grass (Kuoppala, 2010). Since red clover constituted a third of the herbage in mixture 2 (Table 1), this was the main reason for its lower IVOMD. Crude protein and NDF contents were nearly the same in mixtures 1 and 2. The annual content of white clover decreased with the later spring cut in mixtures 1 and 2, whereas the content of red clover in mixture 2 increased. In general, the yearly nutritive value decreased with a later spring cut. The yearly IVOMD and crude protein content decreased whilst NDF content increased. The percentage of annual yield harvested with an IVOMD higher than 78 is shown in Table 1. Late harvest of the spring cut decreased this proportion with a high IVOMD. Especially in mixture 2 was this significant, as Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 577 only a small part, 9-14%, had a high IVOMD with a late spring harvest. In pure red clover (mixture 3) only the autumn cut had a high IVOMD. The results indicate that the effect of cutting strategy is greater in mixtures with red clover. Under comparable conditions, the nutritive value is normally higher in the spring cut than later in the season. Our theory was that a later spring growth, which means a high yield in the period when the quality is potentially higher, together with an adapted cutting strategy for the rest of the growing season, which means a lower yield in the period with lower quality, could give a higher mean herbage quality overall. However, the results showed the opposite. Table 1. The annual dry matter yield and weighted mean of white and red clover content, IVOMD, crude protein and NDF, and the percentage of the harvested herbage with an IVOMD higher than 78. Mixture Strategy Yield Clover content IVOMD Crude protein (t DM/ha) (% of DM) (% of OM) (g/1000 g DM) White cl. 1 2 3 NDF IVOMD %>78 Red cl. 1 11.3a 37.2a 78.7a 190a 397b 61 2 11.6 a a a a c 54 3 11.3a 35.8a 79.3a 189a 4 10.4 b 29.8 b 76.5 b 184 a 5 11.6 a 27.4 b 77.2 b 159 b Lsd 0.4 5.7 1 12.5a 19.3a 2 12.ab 3 12.3 a 4 11.7b 12.7b 38.9a 72.9b 5 12.5 a b b b Lsd 0.7 2 36.6 79.0 196 381 386bc 76 426 a 43 434 a 45 1.3 14 12 31.8c 76.7a 200a 384a 44 18.7a 34.1bc 76.2a 199a 378a 48 a bc a a a 50 185b 432a 14 c a 9 18.6 10.4 3.5 32.9 35.1 3.2 76.3 73.6 195 167 384 421 1.2 10 12 11.1 74.2a 224 316b 8 4 10.2 71.0b 219 357a 2 Lsd ns 3.0 ns 24 Different superscripts within mixtures indicate a significant difference (P<0.05) Conclusion The results showed the potential to influence the nutritive value with minimal effect on the dry matter production. Especially for mixtures containing red clover, the strategy seemed to be essential for the yearly mean of nutritive value. The seasonal profile of nutritive value and production was, in general, strongly affected by the cutting strategy, and the results indicate opportunities for designing herbage quality to meet different end-purposes. To optimize the herbage production for its planned use on the farm, a decision support system is needed. To achieve this, well-defined goals of nutritive value dependent on the proportion of clover in the herbage are necessary. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 578 References Kuoppala K. (2010) Influence of harvesting strategy on nutrient supply and production of dairy cows consuming diets based on grass and red clover silage. PhD Dissertation, University of Helsinki, 114 pp. Weisbjerg M.R, Nielsen L. and Søegaard K. (2011) That’s the way the cutting strategy affects the feed intake and the milk production. Proceedings of Plantekongres, 45-46 (In Danish). Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 579 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 580 Theme 4 posters Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 581 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 582 Conserving high moisture spring field bean (Vicia faba L.) grains O’Kiely P.1, Stacey P.1 and Hackett R.2 1 Animal & Grassland Research and Innovation Centre, Teagasc, Grange, Dunsany, Co. Meath, Ireland; 2Crops Research Centre, Oak Park, Carlow, Ireland Corresponding author: padraig.okiely@teagasc.ie Abstract The high protein content of bean (Vicia faba L.) grains give them the potential to be used as a home-grown feedstuff that could replace soyabeans in ruminant diets. Crimping and additive treatment of beans, which were harvested at a high moisture content and then ensiled in laboratory silos, were assessed for their effects on aspects of conservation efficiency. Whole or crimped bean grains (751 g dry matter kg-1) were ensiled for 160 d either without additive or following the application of acid, urea, Lactobacillus buchneri or Lactobacillus plantarum plus Pediococcus pentosaceus-based additives. The grains in all treatments conserved successfully, undergoing limited fermentation and in-silo losses, and being relatively stable during feedout. Crimping stimulated a more extensive fermentation but increased in-silo losses and the susceptibility to aerobic losses at feedout. Each additive had its unique influence with no single additive improving all traits. Keywords: Vicia faba, ensile, crimp, additive, chemical composition, losses Introduction Field beans (Vicia faba L.; also known as faba, broad, tick or horse bean) are pulse legumes whose grains are included in cattle rations mainly for their high protein content, but whose available energy content is at least as good as cereal grains (McDonald et al., 2011). They are normally stored dry and, prior to feeding to ruminants, they are processed by cracking, rolling, coarse-grinding or steam-flaking. Grains with a moisture content higher than optimal for extended storage can be aerobically stored following artificial drying or treatment with agents such as propionic acid or ammonia that inhibit aerobic microbial activity. An option for beans of higher moisture content is to ensile them with or without physical processing pre-ensiling, and the opportunity exists to manipulate the ensilage process by treating the harvested grains with additives that restrict or enhance fermentation. These options have previously been explored for high-moisture barley grains (Stacey et al., 2011). This experiment assessed the effects of crimping and additive treatment of bean grains, harvested at a high moisture content and then ensiled, on their subsequent chemical composition, in-silo loss and aerobic stability characteristics. Materials and methods Spring field bean (cv. Scirocco) seed was sown at 150 kg ha-1 on 15 February at Oak Park in Carlow, Ireland, and received 185 kg of 0-7-30 fertilizer ha-1. It was combine-harvested on 2 September and representative samples were treated with the following additive treatments: (1) no additive (NA), (2) Graintona (FSL Bells Ltd., UK; acetic acid, isobutyric acid mixture) at 8 l t-1 (Acid), (3) NuGrain (Hydro Nutrition, Hydro Agri (UK) Ltd.; urea solution) at 50 l t-1 (Urea), (4) Biograin (Biotal Ltd., Wales; Lactobacillus buchneri) at 10 l t-1 (B1), and (5) Siloking (Agri-King, Inc., USA; Lactobacillus plantarum, Pediococcus pentosaceus) at 400 g t-1 (B2). For the B1 treatment, the DM concentration of grain was quickly assessed by microwave drying. The whole or rolled grain was then placed in a water-tight mixer with a quantity of water sufficient to reduce grain DM concentration to 550 g kg-1, and continuously mixed for up to 15 minutes. After removing unabsorbed water, the additive was applied at 8 l t-1. The B2 was applied as a dry formulation. All additives were intimately mixed with the grains Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 583 prior to ensiling. Approximately 4 kg grain dry matter (DM) were ensiled in each of triplicate laboratory silos per treatment, without compression within each silo. Silos were stored for 160 days at approximately 15oC. In-silo DM loss values were calculated from the change in weight of herbage DM between the start and end of ensilage (both dried at 98oC for 16 h). A subsample of each silage was subsequently stored aerobically in an insulated container at 20oC for 192 h, and the duration until its temperature rose >2oC above a reference ambient value (aerobic stability) and the accumulated temperature rise during the first 120 h exposure to air (aerobic deterioration) were calculated. Chemical composition, in-silo loss and aerobic stability data were statistically analysed using a general linear model that accounted for crimping, additive and the crimping x additive interaction. Results and discussion A grain fresh yield of 4886 kg ha-1 was recorded with a mean (s.d.) composition at ensiling of DM 751 (7.0) g kg-1, in vitro DM digestibility (DMD) 804 (9.1) g kg-1, ash 35 (0.3) g (kg DM) -1 , starch 335 (5.5) g (kg DM) -1, water-soluble carbohydrates 130 (4.4) g (kg DM) -1, crude protein 255 (2.1) g (kg DM) -1 and buffering capacity 209 (1.9) mEq (kg DM) -1. The highmoisture spring field-bean grains underwent relatively minor quantitative or qualitative losses during ensilage, and therefore conserved successfully. Crimping increased post-ensilage starch, lactic acid, acetic acid, lactic acid as a proportion of fermentation products (LA/FP), NH 3-N, in-silo loss (P<0.001) and aerobic deterioration (P<0.05) values; it reduced DM, pH (P<0.001) and aerobic stability (P<0.01) values, and it had no effect on DM digestibility (DMD), ash, crude protein, ethanol and WSC values (P>0.05) (Table 1). Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 584 Table 1. Spring field bean grain silage composition, recovery rate, and aerobic stability and deterioration characteristics Crimp (C) No No No No No Yes Yes Yes Yes Yes P value Additive (A) NA Aci d Ure a B1 B2 NA Acid Urea B1 B2 SEM1 C A CxA DM2 73 4 731 718 69 9 74 1 729 721 694 687 722 4.1 .188 <.001 <.001 DMD2 82 2 792 817 79 5 80 2 813 805 809 794 805 8.4 .720 .970 .061 Ash3 36 35 34 36 36 36 36 35 37 36 0.4 .866 .094 <.001 Starch3 32 8 308 297 31 8 31 5 340 323 319 343 340 7.0 .833 <.001 .007 C. protein3 29 1 288 336 29 5 28 6 289 286 328 295 288 2.5 .316 .226 6.1 5.9 8.8 5.6 6.1 5.9 5.9 8.9 4.9 5.2 0.05 <.001 <.001 <.001 2 2 6 13 4 7 5 5 23 15 0.6 <.001 <.001 <.001 Acetic acid3 1 2 4 3 1 2 2 4 8 3 0.3 <.001 <.001 <.001 Ethanol3 5 7 3 11 7 8 4 3 9 5 1.0 .075 .362 28 1 203 480 47 8 34 1 394 414 426 582 665 50.6 .018 <.001 .001 86 72 88 63 84 75 94 81 60 69 3.2 .189 <.001 <.001 2 2 26 7 2 3 3 24 11 5 0.4 <.001 <.001 <.001 In-silo loss 32 36 45 71 23 38 44 101 98 56 11.1 .001 <.001 .197 Aer.stab.7 5 10 10 9 5 4 3 7 10 3 1.5 .147 .009 .003 8 7 1 3 3 10 13 11 5 2 12 2. 8 .327 .040 .026 pH Lactic acid Lactic/FP 3 4, 2 WSC3 NH3-N 5 6 Aer. det. <.001 <.001 1 For C x A interaction; 2g kg-1; 3g (kg DM)-1; 4Fermentation products (lactic+acetic+propionic+butyric acids, and ethanol); 5g (kg N)-1; 6g DM (kg DM)-1; 7Days; 8 oC Propionic and butyric acids were at low concentrations or absent. Beans conserved without an additive underwent a restricted fermentation (12.9 g FP (kg DM) -1) with little proteolysis, and exhibited low in-silo losses. The Acid additive reduced pH without altering the extent or direction of fermentation. It caused a small decline in both starch and DMD values. The Urea treatment increased NH3-N, crude protein, acetic acid, pH and in-silo loss values, and reduced both starch and DM values. Both B1 and B2 reduced pH by increasing both lactic acid and LA/FP values, even though they both also increased NH3-N values. B1 additionally reduced DM, DMD and aerobic deterioration, and increased acetic acid, ethanol, in-silo losses and aerobic stability. B1 had a larger effect on pH, lactic acid and NH3-N values than B2. Crimping x additive interactions occurred, with crimping eliciting a larger pH decline in response to B1 and B2, and a smaller pH decline for Acid and Urea, than occurred for the NA treatment (P<0.05). This reflected a larger increase in lactic acid concentration in response to crimping for B1 and B2 than for the other treatments (P<0.05). The Lactobacillus buchneri in B1 increased the concentration of acetic acid, an effect most evident with crimped grain. Overall, these results with high-moisture bean grains agree with the findings of Stacey et al. (2011) and Stacey et al. (2005) relating to high moisture barley and wheat grains, respectively. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 585 Conclusions High-moisture spring field-bean grains were successfully conserved by ensilage, undergoing a limited fermentation and in-silo loss, and being relatively stable during aerobic feedout conditions. Under the circumstances of the evaluation, crimping stimulated a more extensive fermentation but increased in-silo losses and susceptibility to aerobic losses at feedout. Each additive had its unique influence on conservation characteristics and nutritive value, with no single additive being successful in its influence on all traits. Caution is required if extrapolating these results to what might happen on farms due to the greater challenges associated with rapidly achieving and maintaining anaerobiosis in farm silos. Acknowledgement Technical and farm staff at Teagasc Grange and Oak Park. References McDonald P., Edwards R.A., Greenhalgh J.F.D., Morgan C.A., Sinclair L.A. and Wilkinson R.G. (2011) Animal Nutrition. Seventh Edition. Harlow, Essex UK: Pearson Education Limited, 692pp. Stacey P., O’Kiely P., Rice B., Hackett R. and O’Mara F.P. (2005) Alternative conservation strategies for highmoisture wheat grains. Proceedings of the Agricultural Research Forum, Tullamore, Ireland, 14-15 March 2005, p 65. Stacey P., O’Kiely P., Hackett R. and O’Mara F.P. (2011) Conservation of high moisture barley grain. Proceedings of the Agricultural Research Forum, Tullamore, Ireland, 14-15 March, 2011, p96. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 586 Fava bean-rapeseed intercrop as a sustainable alternative to Italian ryegrass: production, forage quality and soil fertility evolution Jiménez J.D., Vicente F., Benaouda M., Soldado A. and Martínez-Fernández A. Servicio Regional de Investigación y Desarrollo Agroalimentario (SERIDA). P.O. Box 13; 33300 Villaviciosa (Asturias), Spain. Corresponding author: admartinez@serida.org Abstract The maize-Italian ryegrass forage rotation is very common among dairy farms due to its high productivity. However, this rotation is very demanding in terms of nitrogen fertilization, with negative effects on soil fertility. This has led the search for new and more sustainable forage sources, leading to a renewed interest in using cattle manure and slurry application to grow forage crops. Under the premise of keeping the maize as summer crop, an alternative winter intercrop (fava bean and rapeseed) to the Italian ryegrass was evaluated, in order to combine the N-fixing abilities of legumes with the ability of cruciferous species to mobilize soil nutrients. The results showed that the fava bean-rapeseed intercrop may be a sustainable alternative to Italian ryegrass as a winter crop, because there are differences in protein and energy performance between these forages. Furthermore, it use allows the reduction of inputs of synthetic fertilizers and herbicides, and has a positive effect on the balance of soil nutrients, especially by increasing the potassium content and maintaining the soil pH. Keywords: sustainability, forage, nutritive quality, organic fertilization, soil fertility Introduction The efficiency of chemical fertilizers used in maize cropping has become a major issue of concern, as the crop is often negatively associated with N-losses and impacts on surface and groundwater quality (Schröder et al. 2000). External inputs of nitrogen and phosphorus on farms should be reduced for environmental and economic reasons. As an alternative to chemical fertilizers, manure application to crop fields can recycle animal wastes and be a valuable soil nutrient resource. A number of studies have reported the on the benefits application of dairy manure on maize silage production (Butler et al. 2009). Furthermore, the production of forages today must be environmentally and ecologically sound, and aligned with public values. Fava beans (Vicia faba) through their fixing of atmospheric nitrogen, with high production, high protein content, and highly digestible and acceptable ensilability, are attractive for sustainable forage production (Martínez-Fernández et al., 2010). Rapeseed has a powerful and deep root system that mobilizes nutrients from deeper layers to the surface and has a 'herbicidal action' by reducing weed growth (Grundy et al., 1999) and maintaining soil fertility (Liebman and Davis, 2000). The aim of this study was to use a fava bean and rapeseed intercrop as an alternative to Italian ryegrass as a winter crop, in combination with the utilization of organic fertilization versus conventional management. The results of two consecutive agronomic years (Sept 2011-Sept 2013) are reported in this study. Materials and methods Two adjacent plots of 1.7 ha each were managed with either standard (SA) or alternative approach (AA). The annual fertilization of SA plot was with a basal fertilizer dressing of 60 kg N ha-1, 40 kg P2O5 ha-1 and 120 kg K2O ha-1 before sowing the winter crop (Lolium multiflorum Lam.), 70 kg N ha-1 applied as topdressing after the first spring cut for silage, and 125 kg N ha1 , 144 kg P2O5 ha-1 and 216 kg K2O ha-1 after second silage cut, before sowing the maize. When maize plants were 20 cm high, 75 kg N ha-1 was applied as a topdressing. The AA plot was fertilized with 32 m3 ha-1 of slurry and 36 t ha-1 of manure before sowing the associated winter Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 587 crop. Before the maize sowing, 33 t ha-1 of manure and 84 m3 ha-1 of slurry were applied. For the winter crops, the AA plot was sown with a fava bean-rapeseed intercrop (Vicia faba cv. Prothabon with Brassica napus cv. Fricola; HC) at seed rates of 150 and 8 kg ha-1 respectively. The SA plot was sown with 45 kg ha-1 of Italian ryegrass (Lolium multiflorum cv. Barextra; RI). As the summer crop, maize (Zea mays cv. Crazy) was sown in both plots, with a seeding density of approximately 90 000 plants ha-1. For weed control in maize, a selective herbicide was used in SA management at a dose of 4 L ha-1 at sowing. In AA management, the herbicide dose was reduced by half to evaluate the effect of weed control by the rapeseed. Before each harvest, samplings for yield and quality of the fresh forage were made. Fresh forage samples were dried at 60ºC for 24h for dry matter analysis (DM), then ground (0.75 mm) and analysed by NIRS for organic matter (OM), crude protein (CP), neutral detergent fibre (NDF), acid detergent fibre (ADF), organic matter digestibility and metabolizable energy. To evaluate the change in the soil profile by both managements, taking as reference the composition thereof in May 2011, the soil was sampled in both plots between harvests. Statistical analysis was performed as factorial model, with the management and year as main factors (SAS, 1999). The evolution of soil characteristics was evaluated by a model for repeated measures. Results and discussion Table 1 shows the forage production of two consecutives agronomic years in both managements for winter and summer crops. Table 1. Forage production (t DM ha-1) under two different managements in 2012 and 2013. Standard Alternative P= sem 2012 2013 2012 2013 1 Management Year Interaction Winter crop1 7.89 4.60 0.214 0.026 0.001 0.807 Summer crop 10.59 9.40 12.95 10.72 0.926 0.330 0.362 0.776 Weeds 1.42 1.17 0.157 0.050 0.185 0.174 Total rotation2 18.48 12.78 21.85 15.31 1.002 0.162 0.013 0.832 3.38 1.41 8.90 0.28 2 Standard Italian ryegrass; alternative: fava bean-rapeseed intercrop; excluding weeds. The 'alternative' winter crops had higher production in a single cut than the cumulated production of the two cuts of Italian ryegrass (6.75 vs. 5.63 t DM ha-1 respectively; P<0.05). The difference in forage production between years was a result of the differences in rainfall. During the 2012-13 winter season the rainfall was 1285 mm, three times the more normal rainfall of the 2011-12 winter (475 mm). There were no differences in maize production between treatments and years. The presence of weeds associated with maize showed significant differences between managements, demonstrating the herbicide effectiveness of the rapeseed crop. In both years, the AA forage had higher concentrations than SA forage of ash (100.7 vs. 78.2 g kg-1 DM), CP (159.6 vs. 90.6 g kg-1 DM), NDF (528.5 vs. 412.6 g kg-1 DM) and ADF (427.1 vs. 193.1 g kg-1 DM). The highest concentration of CP in the AA crop was due to the legume presence, while the highest proportions of NDF and ADF were the result of the high proportion of fibre of rapeseed, which induced to a lower OMD (53.2%) and energy concentration (9.1 MJ ME kg-1 DM) in the AA management than in SA management (80.5% and 11.9 MJ ME kg-1 DM respectively). No significant differences between the two types of management and years regard to maize were found. The pH value and the OM content in the soil ensure successful implementation and development of the crops in the AA management (Figure 1). However, the pH of soil in the SA management gradually declined from the beginning of the experiment. The content of available phosphorus indicates high levels of soil fertility, without significant differences between managements and years. However, the potassium level in sustainable management increased significantly during the study period. This Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 588 indicates the impact of organic fertilization as well as the extraction of K from the deeper soil layers by rapeseed. 400 9,0 350 8,0 7,0 300 6,0 mg kg -1 250 5,0 200 4,0 150 3,0 100 2,0 50 1,0 0,0 0 May 2011 October 2011 P Standard May 2012 P Alternative October 2012 K Standard May 2013 October 2013 K Alternative May 2011 October 2011 May 2012 October 2012 May 2013 MO (%) Standard MO (%) Alternative pH Standard pH Alternative October 2013 Figure 1. Changes in pH and concentrations of organic matter (OM), P and K of soil under two different managements. Conclusions Based on two years' results, the use of organic fertilization and the fava bean-rapeseed intercrop could be regarded as a sustainable alternative to the use of Italian ryegrass as a winter crop, and showed a substantial production of dry matter and protein. This intercropping system offers an opportunity to reduce the use of herbicides and chemical fertilizers, increase the K content and maintain a stable soil pH. Acknowledgements Work supported by Spanish Project INIA RTA2011-00112 co-financed with the European Union ERDF funds. José D. Jiménez is the recipient of an INIA (Instituto Nacional de Investigación y Tecnología Agraria y Alimentaria) Predoctoral Fellowship. References Butler T.J., Weindorf D.C., Han K.J. and Muir J.P. (2009) Dairy manure compost quality effects on corn silage and soil properties. Compost Science & Utilization 17, 18-24. Grundy A.C., Mead A. and Burnston S. (1999) Modelling the effect of cultivation on seed movement with application to the prediction of weed seeding emergence. Journal of Applied Ecology 36, 663-678. Liebman M. and Davis A.S. (2000) Integration of soil, crop and weed management in low-external-input farming systems. Weed Research 40, 27-47. Martínez Fernández A., Soldado A., Vicente F., Martínez A. and de la Roza Delgado B. (2010) Wilting and inoculation of Lactobacillus buchneri on intercropped triticale fava silage: effects on nutritive, fermentative and aerobic stability characteristics. Agricultural and Food Science 19, 302-312. SAS (1999) SAS (Statistical Analysis System) Institute, SAS/STATTM. User’s guide, SAS Institute, Inc. 10. Carry, NC, 378 pp. Schröder J.J., Neeteson J.J., Oenema O. and Struik P.C. (2000) Does the crop or the soil indicate how to save nitrogen in maize production? Reviewing the state of the art. Field Crops Research 66, 151-164. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 589 Fava bean-rapeseed and maize silages growing under organic fertilization as a sustainable alternative for dairy cow feeding Jiménez J.D., Martínez-Fernández A., González A., Soldado A., de la Roza-Delgado B. and Vicente F. Servicio Regional de Investigación y Desarrollo Agroalimentario (SERIDA). PO Box 13; 33300 Villaviciosa, Asturias, Spain. Corresponding author: fvicente@serida.org Abstract The possibility of fava bean-rapeseed and maize silages grown with organic fertilization, as an alternative to Italian ryegrass and maize silages grown with conventional fertilization, were evaluated for use in the diet of grazing dairy cows. A trial was performed with 10 lactating Holstein cows divided into two groups in a 2×2 Latin Square design. Two TMR-based on Italian ryegrass and maize silages, grown using chemical fertilization (SA treatment), or fava beanrapeseed and maize silages grown with organic fertilization (AA treatment) were compared. The results showed that there were no differences in the total dry matter intake and milk yield. However, the fatty acid profile was healthier in the AA treatment than in the SA treatment, with higher concentration of unsaturated fatty acids. The total concentrate intake was lower in the AA treatment than in the SA treatment: this treatment required 83.6 g of concentrate to produce one litre of milk, compared with 51.8 g of concentrate for the AA treatment. Using silages based on organically fertilized fava bean-rapeseed and maize in the diet of dairy cows can be an alternative to using Italian ryegrass and maize silages grown using chemical fertilization, and can reduce the required amount of concentrate in the TMR without affecting milk production. Keywords: dairy cow, sustainability, legumes, organic fertilization Introduction One factor in the current crisis affecting dairy profitability is the increasing prices of agricultural commodities. This has forced dairy farmers to reduce costs and improve the efficiency of using their own forage resources. It is leading to a renewed interest in the use of cattle manure and slurry to grow forage crops, and in searching for new and more sustainable forage sources. On most dairy farms the crop rotation of maize-Italian ryegrass is repeated continuously, which demands high amounts of nitrogen fertilization and has negative effects on the soil. Manure application to crop fields to supply organic fertilization can recycle animal wastes and can be a valuable soil nutrient resource. The benefit of dairy manure application can be attributed to the improvement of physical and chemical edaphic properties (Butler and Muir, 2006). An intercropping system of fava bean (Vicia faba L.) and rapeseed (Brassica napus L.) is presented as an alternative forage system, due to the N-fixing ability of the legume and the capability of the crucifer to mobilize soil nutrients (Jiménez et al., 2014). The aim of this work was to compare for grazing dairy cows, with minimal supply of concentrate in the diet, the feed intake, milk yield and milk composition of using fava bean-rapeseed and maize silages grown with organic fertilization as an alternative to using Italian ryegrass and maize silages grown with conventional fertilization. Materials and methods Two adjacent plots of 1.7 ha each were managed with standard (SA) or alternative approach (AA). Crops and fertilization details are described in the companion paper (Jiménez et al., 2014). Ten Holstein dairy cows of 2.4±0.2 lactations (mean±SD) and with 82±11 days in milk, and 31.3±1.62 L d-1 of milk in the previous week to the beginning of the experiment, were selected and randomly divided into two groups. Two isoenergetic (1.41 Mcal ENl kg-1DM) and Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 590 isoproteic (105.9 g CP kg-1DM) total mixed rations (TMR), SA and AA, were formulated according to NRC (2001). The SA TMR included ryegrass silage (36.9% dry matter basis), maize silage (35.3%) growing both under chemical fertilization, barley straw (16.2%) and concentrate (11.7%). The AA TMR included fava bean-rapeseed silage (43.5%), maize silage (32.5%) growing both under organic fertilization, barley straw (15.6%) and concentrate (8.3%). Both TMR were offered ad libitum and cows were moved after evening milking until the next morning milking to a fresh paddock of 1.0 ha. Cows with production higher than 30 L per day were supplemented with an extra concentrate offered during milking sessions. After an adaptation period of 14 days, the TMR intake and milk production were recorded daily during the assay period (5 days). Grass intake was estimated by the animal performance method (Macoon et al., 2003). TMR was sampled daily and these samples were pooled at the end of each assay period. Milk was sampled for 3 days in both milking sessions. Grass was sampled and body weight and body condition were recorded at the first and last day in each assay period. Both TMR, extra concentrate and grass were dried at 60ºC during 24h to dry matter (DM) analysis, ground (0.75 mm) and analysed by NIRS for organic matter (OM), crude protein (CP), neutral detergent fibre (NDF) and acid detergent fibre (ADF). The fat, protein and urea content of milk samples was analysed by MilkoScan FT6000, and milk fatty acids by gas-liquid chromatography as described Chouinard et al. (1999). Statistical analysis was performed in SAS (1999) using the MIXED procedure for a 2×2 Latin square design. Results and discussion The grass had a nutritional value of 152.7 g CP kg-1 DM and 1.46 Mcal kg-1 DM. The extra concentrate was of 187.4 g CP kg-1 DM and 1.85 Mcal kg-1 DM. Table 1 shows the DM intake in both treatments. Table 1. Dry matter intake (DMI; kg d-1) under two different feeding strategies, standard (SA) or alternative (AA). AA SA s.e. P TMR intake 2.71 8.78 0.532 0.001 Extra concentrate intake 1.12 1.24 0.185 0.503 Grass intake 17.22 17.95 4.331 0.867 Total DMI 21.05 27.89 4.410 0.142 The TMR intake was lower in the AA treatment than in SA treatment (2.7 vs. 8.8 kg DM d-1 respectively, P<0.001). The intakes of extra concentrate (1.1 kg DM d-1) and grass (17.6 kg DM d-1) were similar in both treatments. Therefore, the nutritive value of the whole experimental diets, taking into consideration all ingredients, was similar in both treatments (144.0 g CP kg-1 DM and 1.47 Mcal kg-1 DM). Although the total DM intake in the SA treatment was higher than in the AA treatment, there were no differences between them. Consequently, no differences in live weight or body condition were observed. There were no differences in milk production (26.5 kg d-1) between treatments (Table 2). According to our results, 83.6 g of total concentrate, including concentrate into the TMR and extra, was required to produce one kg of milk under the SA treatment, while the proposed alternative (AA) required 51.8 g kg-1 (P<0.001). The fat and protein contents of milk were not affected by treatments. However, the urea concentration in milk of dairy cows on the AA treatment was higher than for dairy cows feeding on SA treatment (306 vs. 214 mg kg-1 respectively, P<0.001). The unsaturated fatty acids concentration was higher in the AA treatment than the SA treatment, especially due to the higher concentration of oleic (16.88 vs. 12.96 g 100g FA-1), vaccenic (1.52 vs. 0.97 g 100g FA1 ) and linolenic (0.59 vs. 0.34 g 100g FA-1) acids (P<0.001) and CLA (0.59 vs. 0.37 g 100g FA1 ) and linoleic (0.95 vs. 0.66 g 100g FA-1) acid (P<0.05). Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 591 Table 2. Milk production and milk composition under two different feeding strategies, standard (SA) or alternative (AA). AA SA s.e. P Milk yield (kg d ) 26.09 26.96 2.297 0.709 Fat (%) 3.66 3.67 0.151 0.979 3.16 3.21 0.098 0.638 Urea (mg kg ) 306 214 20.0 0.001 Saturated fatty acid (g 100g FA-1) 76.91 82.41 0.876 0.001 23.09 17.59 0.789 0.001 2.08 1.37 0.159 0.001 -1 Protein (%) -1 Monounsaturated fatty acids (g 100g FA-1) -1 Polyunsaturated fatty acids (g 100g FA ) Conclusions The use in dairy cow diets of fava bean-rapeseed and maize silages grown with organic fertilization can be a sustainable alternative to silage of Italian ryegrass and maize grown using chemical fertilization. Their use could reduce the supply of concentrate in the TMR, without affecting milk production. In addition, the use of slurry and manure for forage fertilization could help to reduce the costs of milk production. Acknowledgements This work was supported by Spanish Project INIA RTA2011-00112 co-financed with the European Union ERDF funds. José D. Jiménez is the recipient of an INIA (Instituto Nacional de Investigación y Tecnología Agraria y Alimentaria) Predoctoral Fellowship. References Butler T.J. and Muir J.P. (2006) Dairy manure compost improves soil and increases tall wheatgrass yield. Agronomy Journal 98, 1090-1096. Chouinard P.Y., Corneau L., Barbano D.M., Metzger L.E. and Bauman D.E. (1999) Conjugated linoleic acids alter milk fatty acid composition and inhibit milk fat secretion in dairy cows. Journal of Nutrition 129, 1579-1584. Macoon B., Sollenberger L.E., Moore J.E., Staples C.R., Fike J.H. and Portier K.M. (2003) Comparison of three techniques for estimating the forage intake of lactating dairy cows on pasture. Journal of Animal Science 81, 23572366. SAS (1999) SAS (Statistical Analysis System) Institute, SAS/STATTM. User’s guide, SAS Institute, Inc. 10. Carry, NC, 378 pp. Jiménez J.D., Vicente F., Benaouda M., Soldado A. and Martínez-Fernández A. (2014) Fava bean-rapeseed intercrop as a sustainable alternative to Italian ryegrass: production, forage quality and soil fertility evolution. Grass Science in Europe Vol 19 (these proceedings). Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 592 Effect of harvest and ensiling on different protein fractions in three different legumes Wyss U., Girard M., Grosse Brinkhaus A., Arrigo Y., Dohme-Meier F. and Bee G. Agroscope - Institute for Livestock Sciences ILS, 1725 Posieux, Switzerland Corresponding author: ueli.wyss@agroscope.admin.ch Abstract Crude protein (CP) is an important nutrient in legumes. However, during harvest and ensiling proteolysis occurs which may affect the nutritive value of legumes. Thus, the objective of the present trial was to determine the relative amount of 5 feed nitrogen (N) fractions in the first and third cut of fresh, wilted and ensiled lucerne (Medicago sativa), sainfoin (Onobrychis viciifolia) and red clover (Trifolium pratense). Generally, all silages were of good fermentation quality as evidenced by the high ranking according to the Deutsche LandwirtschaftsGesellschaft (DLG). Regardless of the cuts, during wilting and ensiling the relative amount of the non-protein N (A), slowly degradable (B3) and undegradable fractions (C) increased, whereas that of the fast (B1) and variably degradable (B2) fractions decreased. Compared with red clover and sainfoin, the proportion of fraction A was greater and that of fractions B2 and B3 were lower in lucerne. These differences among legume species might be an effect of plant secondary compounds present in red clover and sainfoin. Keywords: legumes, silage, wilting, protein fractions Introduction During harvest and ensiling the crude protein (CP) fraction undergoes a degradation thereby altering the CP availability and overall nutritive value of legume forages. As a result of proteolysis and desmolysis, undesired substances such as biogenic amines are formed, which ultimately might impair animal health status. Licitra et al. (1996) proposed a fractionation method for CP. The non-protein nitrogen (NPN) was denoted as the A fraction, while the true protein was divided into B1 (buffer soluble CP = fast degradable), B2 (buffer insoluble CP = variable degradable), B3 (CP insoluble in neutral detergent = variable to slow degradable) and C (CP insoluble in acid detergent = indigestible) fractions based on decreasing solubility. As reviewed by Hoedtke et al. (2012) extrinsic and intrinsic factors influence the relative proportion of the different protein fractions. The aim of this study was therefore to monitor changes in the relative abundance of the five protein fractions in lucerne (LU), red clover (RC) and sainfoin (SF) just after cutting in the fresh state, after 1 d of wilting and after ensiling. This was repeated twice, in early summer and autumn (cuts 1 and 3, respectively). Materials and methods The LU, RC and SF were cultivated as pure swards in Posieux (altitude 650 m a.s.l.). In July and September 2012, fresh and wilted samples were collected at three different locations on the field on the day of cutting and 1 d after cutting, respectively. In addition, from the same location wilted legume samples were collected and ensiled for 86 and 95 d in 1.5 L laboratory silos. In the forage samples, the relative amount of the 5 fractions were analysed according to Licitra et al. (1996). In addition, in the silages, dry matter (DM), pH, NH3 content and fermentation products were determined. Data were analysed using analysis of variance with legume species (LU, RC, SF), cutting number (cut 1 and 3) and time of sample collection (fresh, wilted and ensiled) and the 2- and 3-way interactions as fixed factors (SYSTAT 13). Results and discussion After mowing, the DM content of the legumes ranged from 14 to 18%. After 1 d of wilting, the DM content of the first cut was greatest in LU (48%) followed by SF (39%) and RC (34%). For Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 593 the third cut, the LU and SF DM content was similar (32% for both) but lower compared to RC (36%). The DM content of the silages remained comparable in order and magnitude as the wilted LU, SF and RC and were 49, 37 and 33% in the first cut and 31, 31, and 34% in the third cut, respectively. Based on the DLG evaluation scheme (Staudacher and Schenkel, 2007) one LU batch of the first cut was evaluated with only 73 points whereas all other legume silages ranked between 96 and 100 points, implying a very good fermentation quality. The main reason for the lower score of the LU batch was a higher acetic acid content, but the quality could still be regarded as good. Expressed per total N, the proportion of ammonia N in the LU, RC and SF amounted to 8.4, 7.3 and 5.5% in the first cut and 10.9, 7.0 and 4.6% in third cut, respectively (P < 0.01). Regardless of cutting number, CP content of the fresh LU were similar (P = 1.00). By contrast, CP content of the RC and especially of the SF were greater in the third than the first cut (P < 0.01) (Figure 1a). Due to the fermentation activity and degradation of sugars, CP content was generally greater in the silages as in the fresh grass (Figure 1a). In Figure 1 b-f, the relative amounts of fraction A, B1, B2, B3 and C expressed as percentage of total CP of fresh, wilted and ensiled LU, RC and SF harvested in early summer (first cut) or autumn (third cut) are presented. The relative amount of NPN increased during the wilting and especially during the ensiling process (Figure 1b). The greatest changes were observed in LU where the relative amount of fraction A in the silages was 60% greater compared to RC and SF (P < 0.01). A possible explanation that this shift was most evident in LU might be the fact that condensed tannins in the SF or products of the polyphenol oxidase of the RC could have hindered proteolysis of CP, resulting in lower amount of fraction A. Similarly, Tabacco et al. (2006) showed that the addition of chestnut-tannin to lucerne reduced proteolysis and consequently the amount of NPN in silages. Regardless of cut number, the relative amount of fraction B1 and B2 decreased during the wilting and ensiling process (P < 0.01) (Figure 1c and d). The relative proportion of fraction B3 was greater in wilted than fresh LU and RC, but again lower in the silages (P < 0.01). In the third cut, the same was observed also for SF whereas in the first cut the relative amount increased from fresh to the ensiled samples (Figure 1e). In LU and RC the relative amount of fraction C was low in both cuts and only minimal changes were observed between fresh, wilted and ensiled samples (Figure 1f). In the first cut the relative amount of fraction C was greater than in the third cut and was greater in wilted than fresh SF with intermediate values for ensiled SF (Figure 1f). Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 594 b) a) c) d) e) f) Figure 1: Crude protein content (a) and relative amount (expressed in percentage of total crude protein) of fraction A (b), B1 (c), B2 (d), B3 (e) and C (f) in fresh, wilted and silages of lucerne, red clover and sainfoin harvested in early summer (first cut) and autumn (third cut). Values are least square means of 3 batches collected in the field at different locations. Conclusions The relative proportion of the CP fraction changed during the process of harvesting and conservation resulting in a marked shift towards a greater amount of NPN and a marked decrease of the fast and variable degradable true protein fraction. References Hoedtke S., Gabel M. and Zeyner A. (2010) Der Proteinabbau im Futter während der Silierung und Veränderungen in der Zusammensetzung der Rohproteinfraktion. Übers. Tierernährung 38, 157-179. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 595 Licitra G., Hernandez T.M. and Van Soest P.J. (1996) Standardization of procedures for nitrogen fractions of ruminant feeds. Animal Feed Science and Technology 57, 347-358. Staudacher W. and Schenkel H. (2007) Analytische Kenngrössen und Bewertung der Gärqualität von Silagen unter besonderer Berücksichtigung des DLG-Schlüssels. Übers. Tierernährung 35, 45-53. Tabacco E., Borreani G., Crovetto G.M., Galassi G., Colombo D. and Cavallarin L. (2006) Effect of chestnut tannin on fermentation quality, proteolysis and protein rumen degradability of alfalfa silage. Journal of Dairy Science 89, 4736-4746. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 596 Nutritive value evaluation of some grasses and legumes for ruminants Tomić Z.1, Bijelić Z.1, Mandić V.1, Simić A.2, Ruzić-Muslić D.1, Stanišić N.1 and Maksimović N.1 1 Institute for Animal Husbandry, Autoput 16, Zemun, Belgrade, 11080, R. Serbia 2 University of Belgrade, Faculty of Agriculture, Nemanjina 6, Zemun, Belgrade, 11080, R. Serbia Corresponding author: 52.zotom@gmail.com Abstract The experiment was conducted at the experimental field of the Institute for Animal Husbandry in the vicinity of Belgrade, in a randomized block design with four replications. The study included four grass species: tall fescue (Festuca arundinaceae Schreb.), perennial ryegrass (Lolium perenne L.), cocksfoot (Dactylis glomerata L.) and meadow fescue (Festuca pratensis L.) and two legumes: red clover (Trifolium pratense L.) and lucerne (Medicago sativa L.). The aim of this study was to compare the nutritional value of grasses and legumes that are commonly used in the Republic of Serbia for feeding of livestock. Lucerne had the highest content of CP (158.87 g kg-1), and ryegrass the highest ME content (9.83 MJ kg-1 DM). NEL content of the studied species was insufficient to meet the lactating requirements of cows, so that diets for cows should consist of forage and concentrated feed. Keywords: grass, legume, nutritive value Introduction Contemporary livestock production, especially the production of meat and milk from ruminants, is based on a combined diet containing roughage and concentrate. Lucerne hay is the most commonly used fodder in lowland regions due to its high productivity, resistance to high and low temperatures, uniform yield during the growing season and longevity. In the hilly and mountainous areas, in addition to lucerne hay, red clover hay and, in a small proportion, hay of grass-legume mixtures are used. Grass-only crops are very little used in the diet. In contrast to Serbia, in north-western Europe ryegrass has a more significant role in livestock production due to its high production and high nutritional value (Losche et al., 2008). Also, Lättemäe and Tamm (1997) report that grasses are the most important and basic fodder for dairy cows in Estonia. In their studies, it was suggested that grass-legume mixtures have higher contents of metabolic energy (of 10.6-11.00 MJ kg-1 DM) compared to that of grasses (10.310.6 MJ kg-1 DM), but the differences were not statistically significant. Thompson (2013) has concluded by comparing the nutritive value of tall fescue and cocksfoot in monoculture and in mixture, that the grasses in monoculture have similar nutritional value, but when mixed with lucerne, tall fescue shows better results. The aim of this study was to compare the nutritional value of certain grasses and legumes that are commonly used in Serbia in the diet of cattle, both in monoculture and in mixture. Materials and methods The trial was conducted at the experimental field of the Institute for Animal Husbandry in the vicinity of Belgrade (44° 49' 10" N 20° 18' 45" E) in a randomized block design with four replications. The main plot size was 5 m2. The study included four grasses: tall fescue (Festuca arundinacea Schreb.), perennial ryegrass (Lolium perenne L.), cocksfoot (Dactylis glomerata L.) and meadow fescue (Festuca pratensis L.) and two legumes: red clover (Trifolium pratense L.) and lucerne (Medicago sativa L.). Sowing was carried out in 2009, and the study was done in 2010. The soil on which the trial was established is the low carbonate chernozem, of favourable water and air regime. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 597 For this study samples of plant material from the first cut in four replications were used. The samples were dried at a temperature of 60 °C for 72 hours, and then milled and sifted through a 1 mm sieve. The crude protein content was determined using the Kjeldahl method, the content of crude fat according to AOAC method (1990), ash content by burning at 525 °C, and the content of NFE was calculated. For calculation of metabolic energy, and the net energy for lactation the following formulas were used (Nauman and Bassler, 1993; Baranauskas et al., 1998): ME=14.07+(0.0206xCf)+(0.0147xCF)-(0.0114xCP)±4.5%; NEL=9.10+(0.0098xCf)-(0.0109xCF)-(0.0073xCP), Cf- crude fat g kg-1; CF- crude fibre g kg-1;CP- crude protein g kg-1 The data were analysed by one way ANOVA (analysis of variance). The significance of the mean values was performed using the LSD test (StatSoft, Inc., 2007). Results and discussion Results of chemical analysis and the content of metabolizable energy (ME) and net energy for lactation (NEL) are given in Table 1. The content of crude protein (CP) was significantly higher in legumes compared to the grass species. Lucerne had the highest content of CP, 158.87 g kg1 . Among the examined grasses the highest content of CP was in cocksfoot, 126.25 g kg-1. Aufrere et al. (2008) also concluded that cocksfoot has the highest content of crude protein of all tested grasses. CF content also showed significant variation (P˂0.01) among the examined species. Cocksfoot had the highest CF content. No statistically significant differences in the content of CF were established between the species tested. Other studied parameters showed significant (P˂0.05) variation depending on the species. Ryegrass had the highest ME content of 9.83 MJ kg-1 DM, which was not significantly different from the ME content of red clover and meadow fescue. The lowest content of ME was recorded for lucerne (8.94 MJ kg-1 DM). Table 1. Chemical characteristics (g kg-1DM), level of metabolic energy and net energy for lactation (MJ kg-1 DM) of grass and legumes. (CP-crude protein; CF-crude fibre; Cf- crude fat; Nfe-nitrogen free extractives; MEmetabolic energy; NEL-net energy for lactation). Species Tall fescue Perennial ryegrass Cocksfoot Meadow fescue Red clover Lucerne Level of significance CP 93.65d 116.90cd 126.25bc 104.17cd 147.90ab 158.87a ** CF 284.42ab 237.40d 288.12a 278.02ab 241.00cd 262.32bc ** Cf 22.85 28.22 27.50 24.70 28.15 26.20 ns Ash 53.05c 73.52b 55.15c 70.10b 77.10b 89.58a ** Nfe 469.12a 450.70a 428.12a 443.45a 427.37a 375.57b * ME 9.29b 9.83a 8.96b 9.30ab 9.42ab 8.94b * NEL 5.54b 5.94a 5.30b 5.55b 5.67ab 5.33b * Fulkerson et al. (2007), in their study, also show results that lucerne is characterized by the lowest content of ME when compared to other tested grasses and red clover. Content of NEL was also significantly higher in perennial ryegrass (5.94 MJ kg-1 DM) in comparison with other grasses and lucerne. Our research confirms the results of Kohoutek et al. (2007) who studied the quality of individual varieties of perennial ryegrass, cocksfoot and meadow fescue, and show that the content of NEL in varieties of perennial ryegrass are slightly higher 5.73-5.85 MJ kg-1 DM, compared to varieties of cocksfoot (5.72-5.77) and meadow fescue (5.70-5.75). NEL requirements of dairy cows during the lactation period are different. In early lactation the NEL requirement amounts to 6.9-7.4 MJ kg-1 DM, in the middle of lactation 6.6-6.9 MJ kg-1 DM, and at the end of lactation 6.4-6.6 MJkg-1 DM (Jovanovic et al., 2000). Content of NEL, as the primary limiting factor in the production of milk, in the studied species of grasses and legumes, is inadequate to meet the needs of lactating cows. This deficiency in regard to NEL requirement of 6.4 to 7.4 MJ kg-1 DM should be compensated through supplementation with concentrated feed, especially at the beginning of the lactation period. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 598 Conclusion Based on the study of the quality of four grasses and two legumes for ruminant nutrition, as expected, a higher content of crude protein (CP) was obtained in legumes compared to grass species. Lucerne had the highest CP content of 158.87 g kg-1, while the perennial ryegrass had the highest ME content of 9.83 MJ kg-1 DM. NEL content was also significantly higher in the ryegrass (5.94 MJ kg-1 DM) compared to other grasses and lucerne. NEL content of the studied species was inadequate to meet the requirements of lactating cows, so the diets for cows should consist of forage and concentrated feed. Monocultures of the species tested, in regard to the quality, do not meet the needs of ruminants for nutrients and energy, so the combined cultivation of grass-legume mixtures is more useful. References Aufrere J., Carrere P., Dudilieu M., Baumont R. (2008) Estimation of nutritive value of grasses from semi-natural grasslands by biological, chemical and enzymatic methods. Grassland Science in Europe, 13, 426-428. Baranauskas S., Mikulioniene S., Kulpys J., Stankevicius R. (1998) Evaluation of energetic value of forage for dairy cows by means of Hohenheim′s test. Veterinary and zootechnics, 6(28), 62-65. Fulkerson W.J., Neal J.S., Clark C.F., Horadagoda A., Nandra K.S. and Barchia I. (2007) Nutritive value of forage species grown in the warm temperate climate of Australia for dairy cows: grasses and legumes. Livestock Science 107, 253–264. Jovanovic R., Dujic D. and Glamocic D. (2000) Ishrana domacih zivotinja, Drugo izmenjeno izdanje, StylosIzdavastvo, Novi Sad, Serbia, 701 pp. Kohoutek A., Carlier L., Zimkova M., Odstrčilova V., Nerušil P. and Komarek P. (2007) Yield, persistance and forage quality of some grass and legumes species under central European conditions. Grassland Science in Europe, 12, 90-93. Lattemae P. and Tamm U. (1997) Relations between yield and nutritive value of grass or grass legume mixtures at different cutting regimes. Agraateadus, 8, 66 – 80. Losche M., Salama H., Gierus M., Herrmann A., Voss P. and Taube F. (2008) Effect of plant mediated proteolysis on protein degradation among ten diploid perennial ryegrass (Lolium perenne L.) genotypes. Grassland Science in Europe, 13, 435-437. Nauman C. and Bassler R. (1993) Die chemische Untersuchung von Futtermitte ln Methodenbuch. Band III, Damstadt, VDLLIFA, 256 pp. Thompson D. (2013) Yield and nutritive value of irrigated tall fescue compared with orchard grass: in monocultures or mixed with alfalfa. Canadian Journal of Plant Science 93, 799-807. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 599 Forage quality in legumes and non-leguminous forbs Elgersma A.1, Søegaard K.2 and Jensen S.K.3 1 Independent scientist, P.O. Box 323, 6700 AH Wageningen, The Netherlands 2 Dept. of Agroecology and Environment, Aarhus Univ., 8830 Tjele, Denmark 3 Dept. of Animal Sciences, Aarhus Univ., 8830 Tjele, Denmark Corresponding author: anjo.elgersma@hotmail.com Abstract As data for non-leguminous broad-leaf grassland species are scarce, the aim of this study was to obtain novel information on forage quality components and yield for a number of nonleguminous forb and legume species compared to a grass-clover mixture. Four non-leguminous forb species, three legumes and a perennial ryegrass-white clover mixture were investigated in a cutting trial (randomized block) with four harvests (May-Oct) during 2009 and 2010. In pure stands, legumes outyielded non-leguminous forbs for protein content. Ribwort plantain and lucerne had higher ADF and lignin concentrations, but in other species ADF and lignin contents were equal to that of grass-clover. More research is needed to explore the potential value of non-traditional species in the forage-feed food chain. Keywords: herbs, forbs, legume, herbage quality Introduction Fresh herbage is an important natural source of nutrients in diets of ruminants and horses. Most studies on forages have been carried out with agronomical important grass species, such as perennial ryegrass, and with legume species such as white clover, but few studies investigated other grassland forage species. As data of dicotyledonous species grown in a sward are scarce, yield and quality in a number of non-leguminous forbs and legume species were compared with a grass-clover mixture to get an insight into species differences and seasonal variation. Materials and methods Pure stands with each of four non-leguminous forb species, i.e.: salad burnet (Sanguisorba minor), caraway (Carum carvi), chicory (Cichorium intybus) and ribwort plantain (Plantago lanceolata), and three legume species, i.e.: yellow sweet clover (Melilotus officinalis), lucerne (Medicago sativa) and birdsfoot trefoil (Lotus corniculatus) were sown, plus a commercial mixture (85% Lolium perenne + 15% Trifolium repens). In addition, chervil (Anthriscus cerefolium) was sown but only produced herbage in the first cut, and unreplicated plots were sown with borage (Borago officinalis) and viper’s bugloss (Echium vulgare); these species were excluded from statistical analyses. Net plot size was 1.5 m × 9 m. Swards were cut with a forage harvester on 29 May, 9 July, 21 August and 23 October 2009 and on 31 May, 13 July, 19 August and 21 October 2010. After cutting, samples of the harvested herbage were taken to determine DM content and quality parameters (ash, neutral detergent fibre (NDF), acid detergent fibre (ADF), and acid detergent lignin (ADL)) (Elgersma et al., 2014). The experimental design was a randomized complete block with two replications. There were eight ‘species’ (the seven forb species plus the mixture) and four harvests per year. Analysis of variance procedures were applied using the MIXED procedures of SAS. Yield and quality compounds were evaluated with a model that included fixed main effects of species, harvest date and their interaction. All tests of significance were made at the 0.05 level of probability. Results and discussion Average yields ranged from 3.9 to 15.4 t DM ha-1 year-1 and were lower in yellow sweet clover, salad burnet and caraway than in lucerne and the grass-clover mixture. Yields were lowest in the fourth harvest (P < 0.001). The seasonal growth pattern was very different in both years: in Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 600 2009, the first cut yielded most, whereas in 2010 the second cut was most productive, except in caraway. Annual yield was lower in 2010 than in 2009 for most species except caraway. Due to this variability there was no overall effect of year. For most species the DM yield of the first cut was much higher in 2009 than in 2010: e.g., that of grass-white clover yield was 5.3 and 2.4 t DM ha-1 and birdsfoot trefoil was 4.8 and 2.6 t DM ha-1, respectively (Elgersma et al., 2014). In pastures, species are mixed but here pure stands of forb species were studied. Caution should be taken when intrapolating yield data. The weather may provide an explanation for the fact that yields in the first harvest were much lower in 2010 than in 2009. First of all, the winter of 2009/2010 was severe and spring growth started late. The average temperature in April 2010 was only 6.5 oC whereas in April 2009 it was 9.4 oC. Also, in May 2010 the air temperatures were lower than in May 2009. Therefore, the effective primary growth period differed. This implies that the forage was probably less mature on 31 May 2010 than on 29 May 2009. Very high concentrations of ash (> 150 g kg –1 DM) were found in borage and viper’s bugloss, both of which have cuticular hairs (Figure 1a). These species had a very low hemicellulose (i.e., NDF – ADF) content, whereas that of chervil was very high (114 g kg –1 DM). Chervil also had a numerically high fatty acid concentration in the first cut (30 g kg –1 DM) compared to birdsfoot trefoil (28 g kg –1 DM in the first cut), which had a higher content (P < 0.01) than other species (Elgersma et al., 2013; Figure 1b). Among the replicated species, all parameters showed significant differences (P < 0.001) (Elgersma et al., 2014). Averaged across harvests and years, ribwort plantain and lucerne had the highest concentrations of NDF, ADF and ADL (Figure 1b). Birdsfoot trefoil had low NDF and ADF concentrations but a high ADL concentration and thus a high lignification of the cell wall, as well as a low ash content. The highest ash concentration (143 g kg –1 DM) was found in chicory. The crude protein concentration was highest in the three legume species and in the grass-clover mixture, and lowest in chicory and plantain. The concentration of ‘other compounds’ including water-soluble carbohydrates was significantly higher in salad burnet than in all other species; it was also higher in chicory and caraway than in the legume species and the mixture (Figure 1b). Among the replicated species, there were differences between harvests (P < 0.001) for all parameters. Species × cut interactions were not significant for yield and most quality compounds, except for concentrations of crude protein and ash (P < 0.05) (not shown). There were no effects of year, except for lignin content (P < 0.05). Large differences were found for ash and CP contents between harvests (P < 0.001). Ash concentrations were lower in cuts 1 and 2 than in cuts 3 and 4, and increased from the second cut onwards, whereas CP concentrations were higher in the first than in the second cut and then increased to become highest in cuts 3 and 4. There were large differences (P < 0.001) between harvests in the concentration of ‘other compounds’ including water-soluble carbohydrates, which were highest in the first cut and lowest in the third cut. The fibre contents (i.e., concentrations of NDF, ADF and ADL) differed between harvests (P < 0.001) and were lowest in the fourth harvest for all species. The concentrations of ADF and its components (hemicelluse and lignin) were higher in 2010 than in 2009 (P < 0.05). No other effects of year on forage quality compounds were found. parameters. Species × cut interactions were not significant for yield and most quality compounds, except for concentrations of crude protein and ash (P < 0.05) (not shown). There were no effects of year, except for lignin content (P < 0.05). Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 601 Yield composition (g kg DM 1) Yield composition (g kg DM 1) Figure 1. Concentrations of quality compounds in (left) chervil (Ch) in the first cut, borage (Bo) in cuts 2 and 3 in 1000 1000 800 Ash 800 600 Fatty acids Other compounds 600 400 Crude protein 400 200 Hemicellulose Cellulose 200 0 1 Ch 2 Bo 2009 2009 3 Bo 2009 1 Vb 2 Vb 3 Vb 4 Vb 2010 2010 2010 2010 Lignin 0 Sb Ca Chi Rp Ysc Lu Bt G/C 2009 and vipers’ bugloss (Vb) in cuts 1 to 4 in 2010, and (right) salad burnet (Sb), caraway (Ca), chicory (Ch), ribwort plantain (Rp), yellow sweet clover (Ysc), lucerne (Lu), birdsfoot trefoil (Bt) and grass-clover (G/C), averaged across 4 harvests and 2 years. (Modified from Elgersma et al., 2014.) Large differences were found for ash and CP contents between harvests (P < 0.001). Ash concentrations were lower in cuts 1 and 2 than in cuts 3 and 4, and increased from the second cut onwards, whereas CP concentrations were higher in the first than in the second cut and then increased to become highest in cuts 3 and 4. There were large differences (P < 0.001) between harvests in the concentration of ‘other compounds’ including water-soluble carbohydrates, which were highest in the first cut and lowest in the third cut. The fibre contents (i.e., concentrations of NDF, ADF and ADL) differed between harvests (P < 0.001) and were lowest in the fourth harvest for all species. The concentrations of ADF and its components (hemicelluse and lignin) were higher in 2010 than in 2009 (P < 0.05). No other effects of year on forage quality compounds were found. Lucerne, chicory, ribwort plantain and birdsfoot trefoil, in single-species stands, had similar DM yield to a perennial ryegrass-white clover mixture. The nutritional benefits of various forbs may encourage adoption of these species by farmers, but from a management viewpoint they must be balanced against the lack of persistence of most forbs in mixed swards under cutting / grazing. Conclusion Various forbs had a relatively high nutritive value (the legumes had a high protein content, salad burnet had a low NDF content and a high proportion of other compounds including watersoluble carbohydrates) and could enhance the nutritional profile of mixed-species pasture swards. References Elgersma A., Søegaard K. and Jensen S.K. (2013) Fatty acids, α-tocopherol, β-carotene and lutein contents in forage legumes, forbs and a grass-clover mixture. Journal of Agricultural and Food Chemistry 61, 11913-11920. Elgersma A., Søegaard K. and Jensen S.K. (2014 in press) Herbage dry matter production and forage quality of three legumes and four non-leguminous forbs in single-species stands. Grass and Forage Science (doi:10.1111/gfs12104). Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 602 Feed value of restrictedly and extensively fermented organic grass-clover silages from spring and summer growth Bakken A.K.1, Vaga M.3, Hetta M. 3, Randby Å.T.4 and Steinshamn H.2 1,2 Norwegian Institute for Agricultural and Environmental Research (Bioforsk), 1N-7512 Stjørdal and 2N-6630 Tingvoll, Norway 3 Swedish University of Agricultural Sciences, Department of Agricultural Research for Northern Sweden, Box 4097, S-904 03 Umeå, Sweden 4 Norwegian University of Life Sciences, Box 5003, N-1432 Ås, Norway Corresponding author: anne.kjersti.bakken@bioforsk.no Abstract The spring and the summer growth of an organic grass-clover sward were preserved as extensively and restrictedly fermented silages in laboratory silos.The aim was to develop and test the hypothesis that such crops contribute complementary energy and protein qualities that can be exploited in mixed rations. The summer growth, containing 76% red clover, contributed more, and more stable crude protein than the spring growth, which was dominated by grasses. Nevertheless, when preserved as silage, summer growth supplied less metabolizable protein and net energy lactation because of its lower digestibility. Lower feed value remains to be validated in feeding experiments, and the quality of regrowth silages may also be improved by more frequent or appropriate timing of harvests. Restricted fermentation obtained by application of formic acid improved energy and protein preservation. Keywords: digestibility, metabolizable protein, net energy lactation, red clover, soluble protein Introduction In grass-clover swards grown at high latitudes and with low external N supply, there are usually considerable differences in the content of legumes between spring and summer growths. Forage from the first cut is dominated by grass species and has a lower content of crude protein (CP) and an easily digestible carbohydrate fraction (Steinshamn and Thuen, 2008). In contrast, the yields from regrowths contain more legumes, more indigestible cell walls and more CP. At feeding, these qualitatively different forages might be utilized complementarily, according to the animal demands. To realize their potential alone or together and explore their complementarity, their respective values as protein and energy sources have to be determined. Knowing that energy preservation and prevention of protein degradation during wilting and ensiling are critical for the final feed value, we investigated gains and losses obtained by restricted versus extensive fermentation of the crops. The initial results presented here express feed value according to NorFor (Volden, 2011). Material and methods The plant material was harvested from a mixed crop of Phleum pratense, Festuca pratensis, Lolium perenne, L. boucheanum and Trifolium pratense. The spring growth was harvested at late stem-elongation of the dominant grass species, P. pratense, and the summer growth 614 d° afterwards (base temperature 0 °C). The content of T. pratense was 30% of DM in spring and 76% in summer growth. The herbage was wilted indoors for 24 hours before ensiling with different types of additives (4 ml kg FM-1) in evacuated and sealed polyethylene bags. The additives were: 1) water (Control treatment (C)), 2) formic acid (FA, 850 g kg-1), and 3) lactic acid bacteria (LAB) (Kofasil® Lac, Addcon Europe). The dosage of LAB corresponded to 105 cfu per g FM. Dried (60 °C) samples of herbage and silages were analysed for ash, CP, buffer soluble CP (sCP) (Hedquist and Uden, 2006), neutral detergent fibre (NDF) (Mertens et al., 2002) and Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 603 water-soluble carbohydrates (WSC) (Larsson and Bengtsson, 1983). Freshly frozen silage samples were analysed for pH, and content of organic acids and ethanol (Ericsson and André, 2010) and NH4-N. The oven DM contents of the silages were corrected for volatile losses according to NorFor. Concentrations of indigestible NDF (iNDF) were determined by a 288 h in situ incubation (Huhtanen et al., 1994). Organic matter digestibility (OMD) was calculated from iNDF and NDF concentrations (Huhtanen et al., 2013). Net energy lactation (NEL20), metabolizable protein (MP, calculated as amino acids absorbed in the small intestine (AAT20)) and protein balance in the rumen (PBV) were calculated according to NorFor at daily intake of 20 kg DM (Volden, 2011). The constituents in herbage were modelled using the procedure MIXED in SAS (SAS Institute Inc., 1999) with growth (spring or summer) and wilting as fixed factors and replicate (1-3) as random. For silages, the model included growth and silage additive as fixed factors and replicate (1-3) as random. Results and discussion The crop harvested from the spring growth was more digestible and contained less CP and more WSC than the summer growth, which was dominated by mature red clover (Table 1). All silages were well fermented as evaluated from their pH and the concentration of NH3-N. According to the content of organic acids, addition of LAB caused the most extensive fermentation, and FA the least. Protein solubility increased during wilting in the case of the spring growth, but not in summer growth (Table 1). Table 1. Organic matter digestibility (OMD) and content of dry matter (DM), ash, crude protein (CP), soluble CP (sCP), water soluble carbohydrates (WSC), neutral detergent fibre (NDF) and indigestible NDF (INDF) in a fresh and wilted grass-clover crop harvested in spring and summer growth. N=3. Herbage, DM growth (G) and wilt (W) g kg-1 Ash CP sCP WSC NDF iNDF OMD g kg DM-1 Spring growth, fresh 163 70 101 32 264 391 59 769 Spring growth, wilted 235 76 114 39 221 396 67 759 Summer growth, fresh 139 99 133 40 112 361 113 707 Summer growth, wilted 232 102 138 40 87 383 121 695 SEM 5.0 1.5 2.8 0.9 5.3 7.3 3.0 4.0 G, W G, W G, W G, W, G, W G G, W G, W Sign. effect of (P≤0.05) G×W The CP fraction was further solubilized during fermentation (Table 2), and the soluble proportion was higher in silages from spring growth than in silages from summer growth (P<0.001), and higher in extensively fermented (C, LAB) compared to restrictedly fermented (FA) silages (P<0.001). The restriction of silage fermentation caused by application of formic acid contributed positively to energy preservation and protein feed value (NEL20, MP), compared to treatments causing more extensive fermentation (Table 2). Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 604 Table 2. Organic matter digestibility (OMD) and content of dry matter (DM), ash, crude protein (CP), soluble CP (sCP),) metabolizable protein (MP), protein balance in the rumen (PBV) and net energy for lactation (NEL20) in wilted silages made from spring and summer growth of a grass-clover crop and according to type of additive. Silage, DM growth and additive g kg-1 Ash CP sCP OMD MP PBV g kg DM-1 NEL20 MJ kg DM-1 Spring growth (n=9) 246 84 124 72 760 80 7 5.9 Summer growth (n=9 237 111 155 70 699 67 56 5.2 SEM 4.9 1.0 3.5 1.9 3.0 0.4 3.4 0.03 Significance, P≤ NS 0.01 0.01 NS 0.001 0.001 0.01 0.001 Control (n=6) 243 99 142 77 731 69 41 5.5 Formic acid (n=6) 240 97 137 62 729 83 13 5.7 Lactic acid bacteria (n=6) 242 98 140 76 729 69 41 5.6 SEM 4.5 0.9 3.4 2.0 3.1 0.4 3.1 0.04 Significance, P≤ NS 0.05 NS 0.001 NS 0.001 0.001 0.05 Conclusions Fresh and wilted herbage from a summer growth, dominated by red clover, contributed more, and more stable CP than that of the spring growth, which was dominated by grasses. Preserved as silage, it still supplied less metabolizable protein because of its lower digestibility. Whether it has a potential as a complementary forage to silages from spring growth with a lower CP content needs to be evaluated in vivo. Further investigations will reveal if the quality of summer growth silages may be improved by more frequent or appropriate timing of harvests. Acknowledgements The study was financed by the Norwegian Agricultural Agreement Research Fund, Tine BA, the County governors and councils in Sør- and Nord-Trøndelag, and the Norwegian Agricultural Extension Service. References Ericson, B. and André, J. (2010) HPLC-applications for agricultural and animal science. Proceedings from the 1st Nordic Feed Science Conference, Uppsala, Sweden. Report 274, Swedish University of Agricultural Sciences, Dept. of Animal Nutrition and Management. pp. 23-26. Hedqvist, H. and Uden, P. (2006) Measurement of soluble protein degradation in the rumen. Animal Feed Science and Technology 126, 1-21. Huhtanen, P., Kaustell, K. and Jaakkola, S. (1994) The use of internal markers to predict total digestibility and duodenal flow of nutrients in cattle given six different diets. Animal Feed Science and Technology 48, 211-227. Huhtanen, P., Jaakkola, S. and Nousiainen, J. (2013) An overview of silage research in Finland: from ensiling innovation to advances in dairy cow feeding. Agricultural and Food science 22, 35-56 Larsson, K. and Bengtsson, S. (1983) Bestämning av lätt tillgängliga kolhydrater i växtmaterial. Metodbeskrivning 22. Uppsala: Statens Lantbrukskemiska Laboratorium, Sweden. Mertens, D.R. et al. (2002) Gravimetric determination of amylase-treated neutral detergent fiber in feeds with refluxing in beakers or crucibles: Collaborative study. Journal of AOAC International 85, 1217-1240. SAS institute Inc., 1999. SAS/STAT® User’s guide, Version 8, Cary, NC. 3884 pp. Steinshamn, H. and Thuen, E. (2008) White or red clover-grass silage in organic dairy milk production: Grassland productivity and milk production responses with different levels of concentrate. Livestock Science 119, 202–215. Volden, H. (2011) Feed calculations in NorFor. In: Volden, H. (ed). NorFor – The Nordic feed evaluation system – EAAP 130. Wagenigen Academic Publishers, The Netherlands, pp. 55-58. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 605 Feeding, mycological, and toxicological quality of haylage Nedelnik J.1, Strejckova M.1, Cholastova T.1, Both Z.3, Palicova J.2 and Hortova B.2 1 Agricultural Research, Ltd., Troubsko, Czech Republic, 2 Crop Research Institute Prague, Czech Republic, 3 OSEVA Development and Research Ltd., Czech Republic Corresponding author: nedelnik@vupt.cz Abstract The objective of this study was to detect Fusarium spp. and mycotoxins contents in plant material and haylage from permanent grasslands. Feeding quality was also evaluated. Quality and mycotoxin content were different in various level of haylage bales. Parameters of some upper and middle layers did not correspond to proper lactic acid preservation, and these samples were classified as harmful to health and not usable for feeding. Keywords: permanent grassland, haylage, Fusarium, quality, mycotoxins Introduction One of the factors which can negatively influence the quality of roughage relates to microorganisms which can colonize these materials. Mycobiota in haylage produced from various types of grass stands were examined during research focused on the presence of potential pathogenic and allergenic microorganisms, and in particular Fusarium spp. Two locations were chosen, with various types of grass stands used for producing haylage. Location 1 (ca 0.23 ha in total) lies in the foothills of the Hostýn Hills of eastern Moravia in the Zlín District (330–365 m a.s.l.). The grass stand at this location previously had been used to produce hay, and it currently serves for haylage production. Location 2 (ca 0.57 ha) lies in the southeast of the Czech Republic within the Nový Jičín District (281 m a.s.l.). The stand at this location was established in 2010 in an environmentally friendly farming system as a clover–grass mixture on arable land. In future, it will be used over the long term as permanent grassland. In the past, this plot was used as arable land. No intensification practices are applied to the stands at either location, and the haylage production is used within the same farming operations for feeding beef cattle (Charolais, Piedmontese and Limousin breeds). From a phytocoenological viewpoint, the chosen locations are markedly different. Location 1 has the character of a permanent grassland of the mountain meadow type, with the predominant species being members of the Poaceae (52%) and high representation of herbs (29%) and some clovers (12%). By contrast, the stand at Location 2 is characterized by a high proportion of red clover (75%) and annual ryegrass (23%); there are only scattered covering of other Poaceae species and the proportion of herbs is minimal, mainly weed species (Taraxacum sect. Ruderalia, Rumex spp., etc.) that are gradually replacing the dominant species. The objective of this study was to evaluate the feeding parameters and mycological and mycotoxicological quality of roughage production in an experiment involving simulated surface damage to wrapped hay bales. Materials and methods Green mass collection was undertaken at each location prior to the first and second cuttings. From each location, four representative partial samples were collected (random collection diagonally across the plot). Airborne microflora collection was undertaken while collecting the green mass. In the location of the green mass collection, in a square area of 0.5 m2, were placed diagonally opposite one another two Petri dishes with surface areas of 10 cm2 and containing a potato dextrose agar medium. Exposure time was 10 minutes. The technology used for the production of the haylage bales was different at the two monitored locations. The haylage was produced from the first mowing. The project’s designers Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 606 intentionally left the production of haylage fully in the hands of the specific agricultural company which was farming each monitored location, the objective being to have the quality of haylage subsequently used for the research as similar as possible to that produced in actual farming practice. At Location 1, the material was pressed and wrapped in bales with a high dry matter content in the cured fodder, which subjectively can be said to have resembled hay. By contrast, at Location 2, after mowing, the green mass was left to cure for only a short time (spreading it out and once turning it over) and then baled at a low dry matter content either that same evening or the following morning. From each monitored location, two bales of haylage were acquired for further analysis. The first was left in an undisturbed state as a control. The second bale had the plastic covering artificially broken in several places with the tines of a pitchfork (simulating damage from branches) or sliced with a knife (simulating greater mechanical damage). To simulate water leaking into the bale, at 10-day intervals water (10 L per bale) was poured over the bales. Individual samples of the haylage were collected at one time on a date approximately 90 d after the bales were wrapped. From each bale, 3 mixed samples were collected from three different layers: the edge of the bale, the middle layer of the bale, and the centre of the bale. From each sample, three partial samples were collected at random at the edge of the given layer. The minimum weight of one mixed sample was 2 kg. The presence of Fusarium spp. was morphologically evaluated in detail, and within the samples there were morphologically determined also other genera of microscopic fungi. The isolated Fusarium spp. were preserved in a cultures collection for subsequent species identification. Deoxynivalenol, T-2 toxin, zearalenone, fumonisin, and aflatoxin content were analysed using enzyme-linked immunosorbent assay (ELISA). Comprehensive feed analysis and sensory evaluation of the produced fodder were undertaken. Results and discussion Green mass. A greater number of Fusarium CFU was determined in the second collection at both locations. The difference between the collections in May and August was statistically significant (P = 0.04) and no effect of the location on the occurrence of detected fungi was found (P = 0.18). Even though fewer Fusarium spp. were present at Location 2, the reason clearly lies in the composition of the grasslands. While in this location grass species constitute only 23% of the stand, at Location 1 they are the dominant component (60%) of the stand. It is known from the scientific literature that fungi of the genus Fusarium occur abundantly in Poaceae plants and in some cases they may remain active even in wintering grass stands. Haylage samples. Fusarium spp. were found only in bales which had been artificially damaged, and only from the uppermost layers. The greatest numbers of CFU of this genus were found in samples taken from Location 2 (2391 CFU g−1 DM). At Location 1, the quantity of CFU was lower (111 CFU g−1 DM). Fusarium spp. were not found in the samples of high-quality haylage (Figure 1). The large difference between locations in the high-quality haylage can be explained in part by the extent of damage to the bales. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 607 3000 2391 CFU/g dry matter 2500 2000 1500 1000 500 0 0 0 0 0 111 0 0 0 0 0 lowqualityI. lowqualityII. lowqualityIII. highquality I. highquality II. highquality III. lowqualityI. lowqualityII. lowqualityIII. highquality I. highquality II. highquality III. 0 Zavisice Zubri Figure 1. Occurrence of Fusarium fungi (CFU / g dry matter) in haylage samples Qualitative parameters and concurrent sensory evaluation are summarized in Table 1. There was evidence of very low quality in several samples of haylage, mostly owing to the mechanical damage to the bales. Table 1. Feed analysis – selected parameters sample NL SNL Fibre (g/kg (g/kg (g/kg DM) DM) DM) ADF/NDF (%) Lactic acid (g/kg FM) Acetic Butter acid acid (g/kg (g/kg FM) FM) pH 1 128.8 81.14 326.27 63.09/38.13 17.3 1.8 1.5 5.5 2 159.65 100.58 310.74 59.40/35.83 20.4 2.9 1.0 5.1 3 146.79 92.48 299.25 55.41/36.05 19.6 3.9 1.3 5.2 4 239.1 155.42 210.72 45.8/48.65 1.5 0.3 0.0 8.4 5 137.18 86.42 382.27 60.74/47.89 5.3 5.8 0.6 8.7 6 139.61 87.96 326.39 61.11/41.20 20.6 5.9 1.0 5.6 7 97.98 48.99 290.83 59.62/36.95 12.9 4.1 1.6 5.5 8 116.93 80.68 297.68 61.15/37.81 18.7 4.7 0.7 4.9 9 99.95 49.97 292.89 62.23/41.49 13.6 4.7 0.5 5.2 10 289.39 188.1 118.86 55.32/53.09 1.0 1.0 0.0 8.4 11 157.56 99.26 347.31 67.39/55.73 1.3 5.9 0.0 9.2 12 106 53 271.63 53.05/41.69 9.2 7.3 0.0 4.9 The haylage from Location 2 (samples 1–3) was all evaluated as suitable for feeding, although the overall level of the fodder was rated as less than satisfactory. It is clear that as the fermentation process had progressed, the pH levels and acid content had developed accordingly. Nitrogen content, fibre, and other content parameters depended on the species composition of the grass stand. An identical source of green matter produced entirely different results when stored in a mechanically damaged bale (samples 4–6). In the upper layer of the bale (sample 4), the fermentation process and lactic acid preservation did not occur; the fodder was evaluated as unsatisfactory and even harmful to health and with a high content of macroelements and ash. The middle layer was conditionally usable for feeding, although even at a distance of 50 cm from the edge of the bale only partial lactic acid preservation had occurred. The centre of the bale was usable as feed. Had such bales been on the farm, at least 50% of their content would have been lost, because low-quality fodder cannot be used for feeding. Samples from Location Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 608 2 displayed similar characteristics. The undamaged bales (samples 7–9) show relatively standard parameters. By contrast, samples 10–12 again show parameters (pH levels and low lactic acid content) in their upper and middle layers that do not correspond to proper lactic acid preservation, and these samples were classified as harmful to health and not usable for feeding.Of the five mycotoxins detected in total (Table 2), only rarely in the 16 samples was an occurrence of at least one of them not observed. Table 2. Feed analysis – mycotoxin content sample DON (ppb) FUM (ppb) AFL (ppb) ZEA (ppb) T2 (ppb) 1 220 15 0.3 100 50 2 220 20 0.4 120 90 3 250 20 0.1 115 80 4 1500 85 0.0 50 50 5 500 50 0.4 80 20 6 720 40 0.4 75 70 7 115 0 0.1 30 25 8 120 0 0.0 40 70 9 180 0 0.1 45 75 10 1550 200 0.15 15 60 11 980 100 0.8 20 45 12 270 0 0.4 60 65 13 170 10 0.2 15 20 14 200 10 0.0 20 20 15 220 15 0,2 27 20 16 200 20 0.0 27 20 Green mass from the first and second cuttings (samples 13–16) was contaminated at low levels with all monitored mycotoxins. This testifies to the incidence in the grass stands of producers of these substances, and also to the stability of the system and to the relative harmlessness to health from the perspective of mycotoxins in the green forage. If not fully completed, the preservation process can markedly increase the level of mycotoxins. Pathogenic organisms are present even in preserved fodder and if the fermentation process is not quickly initiated in anaerobic conditions the production of mycotoxins continues. While deoxynivalenol (DON) content in the samples from well-wrapped bales was at a low level compared with that in the green mass (samples 1–3, 7–9), in the samples from damaged bales the content was many times greater and such fodder is harmful to the health of animals. A similar situation was detected for the amount of fumonisins (FUM). Such dependence was not recorded in the amount of aflatoxins, mainly because the levels were very low. Considering the character of the producers, which are Aspergillus spp., such dependence is also improbable. The T-2 toxin content is interesting, as it displays greater toxic activity than that for DON, so its presence in hundreds of parts per billion serves as a warning. There presently is ongoing debate (at EC level) on defining the hygienic limit for this toxin. Acknowledgements This contribution was created with support from the Ministry of Agriculture of the Czech Republic, project NAZV QI111C016. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 609 Mycotoxin and chemical characteristics of silages collected from horizontal silos on farms in Co. Meath, Ireland - a pilot study McElhinney C.1,2,3, Danaher M.3, Elliott C.2 and O'Kiely P.1 1 Teagasc Animal & Grassland Research and Innovation Centre, Grange, Dunsany, Co. Meath, Ireland, 2 Institute for Global Food Security, Queen’s University Belfast, Belfast, Northern Ireland, United Kingdom, 3 Food Safety Department, Teagasc Food Research Centre, Ashtown, Dublin 15, Ireland. Corresponding author: cormac.mcelhinney@teagasc.ie Abstract Mycotoxins are secondary fungal metabolites commonly found in silages and can cause a range of detrimental ailments in livestock including abortions, vomiting, lameness, immunosuppression, reduced intake or feed refusal, and reduced performance. The objectives of this study were to identify the mycotoxin challenge present in a selection of farm silages in Co. Meath, Ireland, and whether sampling position in the silo had an impact on chemical composition or mycotoxin profiles. Silages samples were collected from 2 silo locations (silage feed face vs. 3 m behind feed face) from one silo on each of 18 farms (15 grass and 3 maize). There was no difference (P>0.05) in pH, dry matter digestibility, lactic acid, acetic acid, watersoluble carbohydrates or crude protein concentrations between sampling locations. Mycotoxins (andrastin A, beauvericin, enniatins A1; B; and B1, mycophenolic acid, roquerfortine C and zearalenone) were detected in silages, with enniatin B being the most prevalent occurring in 12 out of 36 samples (mean 0.14 mg/kg DM). Twelve mycotoxins were below the limit of detection at the feed face and the remainder were below the EU threshold. There was no effect (P>0.05) of sampling location on mycotoxin concentration. Samples could be taken from either silo location without affecting conventional analysis characteristics (except DM concentration) or mycotoxin concentration value. Keywords: Mycotoxin, silage, fungal metabolites, sampling location Introduction Among the moulds identified in Irish silages, some, such as Fusarium and Penicillium, are toxigenic and thus can produce secondary metabolites named mycotoxins. The latter can induce a range of detrimental ailments in livestock including abortions, vomiting, lameness, immunosuppression, reduced intake or feed refusal, and reduced performance. In addition, consumption of animal products contaminated with mycotoxins can pose health risks to humans. Monitoring feed for mycotoxins is requested by the European Food Safety Authority and this study includes all mycotoxins that are regulated in Commission Directive (EC) No. 32/2002 and recommendation (EC) No. 576/2006. The objectives of this study were to identify the mycotoxin challenge present in a selection of farm silages in Co. Meath, Ireland, and whether the sampling position in the silo had an impact on chemical composition or mycotoxin profiles. This preliminary study was undertaken in order to identify optimal methodologies for a subsequent national survey. Materials and methods One horizontal silo was sampled on each of 18 farms (15 grass and 3 maize silages) during February and March 2012. Farms were randomly selected and were located within a 10 km radius of Teagasc, Grange, Co. Meath, Ireland (N 53.52, W 6.66). Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 610 Sampling Three perpendicular core samples were taken from a 3-m wide section at the vertical silage feed face due next for removal, using a motorized corer (length 65.0 cm; internal diameter 3.5 cm). Two full depth vertical core samples were also taken from 3 m behind the silage feed face at equal distances from the silo sides, using a manual corer (internal diameter 3.0 cm). Sampling devices were thoroughly cleaned between corings. Silage samples were stored (at -20ºC) prior to subsampling for conventional chemical and mycotoxin analyses. Conventional chemical analysis Silage dry matter (DM) concentration was determined by oven drying (85ºC for 16 h) and was corrected for loss of volatiles according to Porter and Murray (2001). Dried (40ºC; 48 h), milled (sieve with 1 mm apertures) silage samples were assayed for in vitro dry matter digestibility (DMD; Tilley and Terry, 1963), water soluble carbohydrates (WSC; automated anthrone method), ash (complete combustion in muffle furnace at 550ºC for 5 h) and crude protein (N*6.25; N determined by Dumas method on a Leco FP-428 nitrogen analyser). Aqueous extracts were used to determine pH (electrode), volatile fatty acids (VFA), ethanol (gas chromatography) and lactic acid (Boehringer method). Mycotoxin analysis Mycotoxin analysis was carried out using an inter-laboratory validated Ultra High Performance Liquid Chromatography tandem Mass Spectrometry (UHPLC/MS2) analytical method capable of detecting 20 mycotoxins in a single 16 minute run. This method includes regulated (aflatoxin B1 (AFB1), deoxynivelenol (DON), fumonisin (FUM) B1, B2, ochratoxin A (OTA), HT-2, T-2 and zearalenone (ZEA)) and unregulated aflatoxin B2, G1, and G2, andrastin A, beauvericin, enniatin (ENN) A1, A, B1 & B2, mycophenolic acid (MPA), roquerfortine (Roq) C, E) mycotoxins. The mycotoxin extraction procedure used a modified (0.1M HCl) QuEChERs platform with no clean-up step. Statistical analysis Conventional chemical analysis (Table 1) data were analysed using a paired t-test accounting for sampling location. Mycotoxin data (Table 2) were analysed using the Wilcoxon signed ranks test (values below the limit of detection (LOD) were censored (assigned 0.5 LOD)). Results and discussion There was no difference (P>0.05) in pH, DMD, lactic acid, acetic acid, WSC or crude protein concentrations between sampling locations (silage feed face vs. 3 m behind feed face). The silages sampled in this pilot study were well preserved (pH 3.85-3.89) and the proliferation of a lactic acid bacteria fermentation was evident. The higher (P<0.05) dry matter concentration at the silage feed-face than 3 m behind this may reflect (a) drying of the feed-face due to exposure to lower humidity ambient air, (b) drying of the feed face due to heat generated by respiration at the aerobic feed face and (c) moisture evaporating due to more evident frictional heating associated with the horizontal coring at the feed face than associated with vertical coring 3 m behind the feed face. Twelve mycotoxins (aflatoxin B1, B2, G1 and G2, DON, enniatin A1, FUM B1, B2, HT2, OTA, Roq. E, T2 toxin) were below the limit of detection at both the silage feed face and 3 m behind the feed face in all silos indicating that sampling location had no measurable impact on their concentrations. Eight mycotoxins (andrastin A, beauvericin, enniatins A1; B; and B1, MPA, Roq. C and zearalenone) were detected in the silages. In this study enniatin B was the most prevalent mycotoxin, occurring in 12 out of 36 samples (mean 0.14 mg/kg DM). The highest mycotoxin concentration was for MPA (1.41 mg/kg DM) and it was observed at the silage face. One EU regulated mycotoxin (zearalenone) was detected (0.076 mg/kg DM) at 4% of EU threshold. There was no effect (P>0.05) of sampling location on mycotoxin concentration. However, even though concentrations of individual mycotoxins did not give rise to concern, Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 611 caution needs to be exercised in concluding on the overall mycotoxin challenge since the effects of mixtures of mycotoxins can be more severe than the sum of their individual effects. Table 1. Summary statistics of the conventional chemical analysis carried out on horizontal farm silages collected in Co. Meath in 2012. (** denotes P <0.01) Silage face (n=18) 3 m behind face (n=18) Conventional analysis Dry matter (g/kg) Mean 257 SE 9.0 Mean 242 SE 9.4 P value ** pH DMD (g/kg) Lactic acid (g/kg DM) Acetic acid (g.kg DM) Propionic acid (g/kg DM) Butyric acid (g/kg DM) 3.85 705 104 27 2.9 3.4 0.111 13.1 11.6 2.5 0.82 1.09 3.89 702 98 27 2.4 5.0 0.086 10.3 8.5 1.7 0.47 1.65 NS NS NS NS NS NS Ethanol (g/kg DM) WSC (g/kg DM) Ash (g/kg DM) Crude protein (g/kg DM) Ammonia –N (g/kg N) 16 19 83 137 103 1.9 6.0 8.0 7.0 15.1 17 25 83 141 94 1.9 6.2 5.4 7.5 11.16 NS NS NS NS NS Table 2. Summary statistics of the mycotoxin analysis by UHPLCMS2 carried out on horizontal farm silages in Co. Meath in 2012. (LOD denotes Limit of Detection) Silage face (n=18) Mycotoxin Andrastin A (µg/kg DM) LOD 50 Beauvericin (µg/kg DM) Enniatin A1 (µg/kg DM) Enniatin B (µg/kg DM) Enniatin B1 (µg/kg DM) MPA (µg/kg DM) Roq. C (µg/kg DM) Zearalenone (µg/kg DM) 5 10 25 25 40 40 10 Max 863 25 204 66.3 1419 1194 - 3 m behind face (n=18) Mean 95 SE 45.1 n >LOD 1 Max 500 Mean 78.22 SE 28.07 n >LOD 2 11 58 27 117 115 - 0.8 13.6 2.3 77.0 64.4 - 0 1 5 1 1 2 0 21.8 23.1 255 80.9 167 500 76.6 5.93 10.73 71.11 28.11 57.73 76.82 13.70 0.930 0.730 18.82 3.11 9.73 27.31 3.70 1 1 7 1 3 2 1 Conclusion In this pilot survey, the challenge posed by individual mycotoxins for ruminants consuming silages was well below published guidelines. Samples could be taken from the silage feed face or from 3 m behind this without affecting conventional analysis characteristics (except DM concentration) or mycotoxin concentration values. References Porter M.G. and Murray R.S. (2001) The volatility of components of grass silage on oven drying and the interrelationship between dry-matter content estimated by different analytical methods. Grass and Forage Science 56, 405–411. Tilley J.M.A. and Terry R.A. (1963) A two-stage technique for the in vitro digestion of forage crops. Journal of the British Grassland Society 18, 104–111. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 612 Prediction of energy content of grass silages depending on grass and ensiling conditions Pickert J.1 and Weise G.2 1 ZALF, Eberswalder Str. 84, 15374 Müncheberg, Germany 2 PAGF, Gutshof 7, 14641 Paulinenaue, Germany Corresponding author: juergen.pickert@zalf.de Abstract A model that predicts energy concentration of grass silages was tested on 16 farms for the years 2002 and 2004. The model is based on grass energy concentration and the ensiling conditions. Estimates of ensiling conditions depend on ensiling capacity, silo compaction, silo temperature and use of silage additives. Factors that reduced energy concentration were determined and used to derive the expected energy content of grass silage. The model was able to predict the energy content of the grass silages with a difference of 0.1 MJ NEL kg DM-1 and a coefficient of variation of 2.22 and 1.09%, in 2002 and 2004, respectively. For planning silage supply for different cattle groups of a holding the model delivered quality data as early as possible. In the case of farms with several silos with different silage quality characteristics, silages can be stocked best according to the special requirements of the herds. By referring the analyses to areas where the grasses originate, farmers gain valuable information for grass sward management of a particular field or pasture. Keywords: model, sward management, feed planning Introduction In large agricultural farms with grassland of different sward compositions, the quality of the silages usually varies among the different silos. Knowing the grass quality, particularly the energy concentration, is a precondition for planning the effective use of the different silages according to the requirements of the different herds of a holding. In cases where there are of quality problems, a successful analysis is impossible without detailed reporting of the ensiling process. Normally the farmer gets forage quality data that are needed for planning from the forage quality analysis of the silages. Sampling takes place by drilling and taking cores or by sampling the silo opened for feeding. Particularly in large silos, sampling of drilled cores is inefficient and does not very often supply representative results. By sampling the opened silo, results are available just at the time of feeding. There is very often a great difference in time between ensiling and feeding. Particularly in larger silos, it is nearly impossible to refer the silage analyses results to certain fields or pastures. Materials and methods A model named ‘Normative Silokartei’ was developed particularly for large agricultural enterprises at the former Institut für Futterproduktion Paulinenaue (East Germany) in the 1980s. The model predicts dry matter (DM) losses and changes in different quality parameters during the ensiling process (Weise and Rambusch, 1983). The prediction is based on forage quality analysis, evaluation of the ensilability of the ensiled grass, and on the ensiling techniques by the farmer. The model was already introduced to farmers before 1989 (Knabe et al., 1986; Weise and Rambusch, 1988), but its use stopped due to the structural changes in the agricultural advisory system and, after 1989, the Northeast German farms themselves. As part of their actual grassland advisory activities the authors made use of the model in several farms since 2000 (Weise and Hertwig, 2011). In 2002 and 2004 the silage quality tool of the model was tested again under the actual practical conditions, in 5 and 9 farms respectively. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 613 In the model, the forage quality of silage is based on the analysis of the grass forage value at time of ensiling (Figure 1). Prediction of grass silage quality based on grass quality and ensiling conditions Silage quality = f (grass quality, ensiling conditions) Silage amount = f (grass amount, DM-losses) ensiling technique (performance, compaction, silo temperature) Quality decrease = f (ensiling conditions) grass quality = f (XP-, XF-, XAcontent; MJ NEL content) ensiling conditions = f (ensiling technique, ensilability) ensilability (species, DMcontent, additives) Figure 1. Structure of the model, part: prediction of silage energy content The sampling of the grass to be ensiled and of the feed silage was carried out by the farmer throughout the entire filling and feeding time of the silo. The data used for the original model parameterization were based on the chemical analysis of forage value of the grass and the grass silage from 23 large clamp silos based on the methods and the tables from the DDRFutterbewertungssystem (Beyer et al., 1971), which provided values for crude protein (XP), crude fibre (XF), crude ash (XA) and energy concentration (EC). In the 2002 and 2004 test, forage value was estimated by near infrared (NIR)-spectroscopy (VDLUFA, 2012). Therefore the EC is given as ‘Energetische Futtereinheit Rind’ (EFr) per kg DM in the results of the original model in 1983 and as ‘Mega Joule Net Energy Lactation’ (MJ NEL) per kg DM in the test of 2002 and 2004. The ‘ensiling conditions’ are characterized by the ‘ensilability’ of the grass and the ‘ensiling technique’, both described by three levels: ‘good’, ‘medium’ or ‘bad’. The ensilability mainly depends on the grass sward composition, the DM content at ensiling and the use of silage additives. For describing the ‘ensiling technique’ the silo filling und compaction process were evaluated. On the basis of a five level scale ‘ensiling conditions’ factors were derived to reduce grass EC ranging between 3 and 10 %. Results and discussion In the original 1983 model the silages contained 92% of the grass EC (Table 1). Table 1. Energy concentration of grass, prediction and silage (all years and farms) energy concentration 1 silage/grass, % grass prediction silage mean 1983 513 484 475 92 2002 6.3 6.0 6.1 96 2004 6.4 6.1 6.2 97 1 1983: EFr kg DM-1, 2002 and 2004: MJ NEL kg DM -1 year 2 mean 98 101 101 silage/prediction, % min max 91 109 100 105 100 103 CV 2 4.82 2.22 1.09 Coefficient of Variation With a coefficient of variation of 4.82%, the silage EC met the predicted EC on average by 98%. In the trials of 2002 and 2004, the silages had on average 96 and 97% of the EC of the grass respectively. The coefficient of variation was 2.22% in 2002 and 1.09% in 2004, while the difference between predicted and measured silage EC was 0.1 MJ NEL kg DM-1. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 614 Figure 2. Energy concentration of grass, prediction and silage (all farms 2004) Conclusions From a practical point of view, the model successfully predicted EC of the grass silages. The important information of energy content of the silage was available early enough for planning the feeding period. Different silos, with different grass silage energy contents, could be opened and fed to the herds according to the animal requirements. The places of origin for the different grass materials were available for the model. When the mineral content, i.e. P and K, of the samples are also available as well as the sward species composition, then the farmer has all the essential information for an effective sward management, such as fertilization, herbicide use or grass and legume reseeding (not reported). References Beyer M., Chudy A., Hoffmann B., Hoffmann L., Jentzsch W., Laube W., Nehring K. and Schiemann R. (1971) Das DDR-Futterbewertungssystem. VEB Deutscher Landwirtschaftsverlag, Berlin, DDR, 255 pp. Knabe O., Fechner M. and Weise G. (1986) Verfahren der Silageproduktion. VEB Deutscher Landwirtschaftsverlag, Berlin, DDR, 300 pp. VDLUFA (2012) The chemical analysis of feedstuffs. Method book III. incl. 8th supplement delivery. Verband Deutscher Landwirtschaftlicher Untersuchungs-undForschungsanstalten, Darmstadt, Germany. Weise G. and Hertwig F. (2011) Kontrollregime für die Bereitung von hochwertigen Silagen vom Grünland. In: Niedermoor und Sand - Grünlandstrategien in Nordostdeutschland. Deutsche Landwirtschaftsgesellschaft, Frankfurt/Main, Germany, pp. 19-23. Weise G. and Rambusch H. (1983). Methode zur EDV-gerechten Mengen- und Qualitätserfassung von Siliergut und Silagen unter Nutzung der Normativen Silokartei und eines Silierkataloges zur Voraussage der Silagemenge und Silagequalität. Research Report A 4, Institut für Futterproduktion, Paulinenaue, DDR, 114 pp. Weise G., Rambusch H. (1988) Produktionsüberwachung und Qualitätssicherung Normative Silokartei. Landwirtschaftsausstellung der DDR, Markkleeberg, DDR, 44 pp. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 615 Predicting organic matter digestibility by two enzymatic in vitro methods Beecher M.1,2, Baumont R.3, Aufrère J.3, Boland T.M.2, O’Donovan M.1, Galvin N.1, Fleming C.1 and Lewis E.1 1 Teagasc, Animal & Grassland Research and Innovation Centre, Moorepark, Fermoy, Co. Cork, Ireland 2 School of Agriculture and Food Science, University College Dublin, Belfield, Dublin 4, Ireland. 3 INRA, UMR 1213, Herbivores, 63122 St Genès Champanelle, France Corresponding author: Marion.Beecher@teagasc.ie Abstract In vitro methods are regularly used to predict in vivo grass organic matter digestibility (OMD). To ensure accurate comparisons can be made between laboratories repeatability between methods should be checked. The objective of this study was to compare two in vitro methods of OMD measurement: neutral detergent cellulase digestibility (NDCFD) and pepsin cellulase (PCD60) to in vivo OMD. Eighteen perennial ryegrass samples were used. The samples were subsequently milled through a 1-mm screen. The OMD of the samples oven-dried at 60oC was analysed using the PCDFD method and the corresponding freeze-dried samples were analysed using the NDC60 method. The methods gave similar OMD values. The results of the two methods were then compared to in vivo OMD data to assess which results aligned most closely with the in vivo model. The in vivo and NDCFD OMD were significantly different while the in vivo and PCD60 OMD tended to be different. This study highlights the importance of comparing in vitro and in vivo methods regularly to ensure accurate OMD predictions. Keywords: digestibility, Lolium perenne L., in vitro, in vivo Introduction Grass organic matter digestibility (OMD) is a measurement of grass quality and is a key determinant in estimating the energetic value of grass and thus the grass nutritive value. Accurate prediction of grass quality is essential to formulate diets and predict animal performance. The in vivo method is viewed as the ‘gold standard’ method for measuring OMD and is the method against which all other OMD methods are evaluated. The development of in vitro methods has decreased time, cost and the labour requirement for OMD measurement, while also providing accurate values (Givens and Deaville, 1999). Enzymatic methods such as the pepsin cellulase method (PCD; Aufrère and Michalet-Doreau, 1988) and neutral detergent cellulase method (NDC; Morgan, Stakelum and Dwyer, 1989) are commonly used as they dispel the need for rumen fluid by using cellulolytic enzymes to mimic digestion in the rumen. Care should be taken when comparing OMD results from different laboratories as sample preparation and the method of analysis may differ, thus altering the in vitro digestibility results and making comparisons difficult. The objective of this study was to compare two in vitro methods of OMD measurement: neutral detergent cellulase digestibility (NDC) and pepsin cellulase (PCD) to in vivo OMD. Materials and methods An in vivo study was conducted as a replicated 3×3 Latin square design, with three pre-grazing herbage mass treatments and three periods. A total of 18 perennial ryegrass samples were collected (one grass sample per treatment per period). Twelve wether sheep were individually housed and offered fresh grass twice-daily ad libitum (four sheep per treatment per period). Grass intake and faeces produced were recorded daily. A representative sample of grass offered and faeces voided from each sheep was collected daily during each period. In vivo OMD was determined for each treatment per period. The grass samples were dried in two ways: ovenGrassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 616 dried at 60oC in a Binder 720 drying oven and freeze-dried at -55oC (LS40+Chamber, MechaTech Systems Ltd., Bristol, UK). Following drying, the samples were milled through a 1-mm screen using a Cyclotech 1093 Sample Mill (Foss, DK-3400 Hillerød, Denmark). All samples were analysed for ash content by placing samples into a Gallenkamp muffle furnace size 3 (Thermo Fisher Scientific INC., Waltham, MA, USA) for 16 h at 500 °C. The crude protein (CP) concentration was analysed using a Leco N analyser (Leco FP-528; Leco Corporation, St., Joseph, MI, USA). The samples were analysed for acid detergent fibre (ADF) with an Ankom Fibre Analyser (Ankom Technology Corporation, NY, USA). The ADF values do not include ash. The grass samples oven-dried at 60oC were analysed for OMD using the in vitro PCD (PCD60) method of Aufrère and Michalet-Doreau (1988) as the PCD method was developed using oven-dried samples. The corresponding freeze-dried samples were analysed for OMD using the in vitro NDC method (NDC FD; Morgan et al., 1989; Fibertec™ Systems, FOSS, Ballymount, Dublin 12, Ireland) as the NDC was developed using freeze-dried samples. In vitro OMD was determined on all grass samples in triplicate. PROC GLM in SAS (2002) was used compare NDCFD to PC60 and to compare the in vitro methods to in vivo OMD. Method was included as a fixed effect in the model. The Tukey Kramer multiple range test was used for mean separation (P < 0.05). Results and discussion In vivo OMD was within the range reported for perennial ryegrass. The CP concentration of the grass samples ranged from 127 to 303 g kg-1 (mean = 216 ± 47.9 g kg-1). The ADF concentration ranged from 217 to 285 g kg-1 (mean = 256 ± 19.1 g kg-1). The ash content ranged from 61.7 to 71.9 (mean = 72 ± 8.2 g kg-1). The chemical composition of the grass was within the ranges reported for predominately perennial ryegrass grazing swards (O’Neill et al., 2013). Both in vitro OMD methods gave similar grass in vitro OMD values (P>0.05). The results of the two methods were then compared to in vivo OMD data to assess if the results aligned closely with the in vivo model (Figure 1). (a) 900 y = 0.581x + 330.68; R2 = 0.59; P >0.05 850 In vivo OMD g kg-1 In vivo OMD g kg-1 900 (b) 800 750 700 650 600 y = 0.7352x + 220.48; R² = 0.47; P<0.05 850 800 750 700 650 600 600 700 PCD60 g 800 kg-1 900 600 700 800 NDC FD g kg-1 900 Figure 1. The relationship between perennial ryegrass in vivo organic matter digestibility (OMD; g kg-1) and OMD predicted by two in vitro methods: (a) pepsin cellulase method (PCD60) using samples oven-dried at 60oC and (b) neutral detergent cellulase method (NDC FD) using freeze-dried samples A regression equation is necessary to relate in vitro OMD values to the in vivo method allowing for the in vivo OMD of future samples to be derived from the in vitro result. The regression equation currently used for the NDCFD method was developed in the 1980s using the in vivo Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 617 OMD of dairy cows and since then the equation has not be evaluated (Morgan et al., 1989). Since the 1980s, there have been significant improvements in grass breeding, grass management and animal genetics. These improvements may not be reflected in the regression equation being used to relate NDCFD OMD data to in vivo OMD. Regular validation of any method is essential to ensure its accuracy over time. Unlike the NDC, which had a once-off evaluation, the PCD60 method is regularly evaluated (Aufrère et al., 2007). Aufrère et al. (2007) developed plant species-specific equations to predict OMD and the equation used for perennial ryegrass also incorporates permanent pasture. The chemical composition of the grasses used by Aufrère et al. (2007) differed to the chemical composition of the grass used in the present study, which may explain why the PCD60 tended to predict different OMD compared to the in vivo OMD. In that study CP concentration ranged from 69 to 266 g kg-1 (mean = 150 g kg-1), ash content ranged from 55 to 193 g kg-1 (mean = 110 g kg-1) and crude fibre concentration ranged from 186 to 405 g kg-1 (mean = 263 g kg-1), which equates to a mean ADF concentration of 287 g kg1 . Despite this difference in chemical composition the PCD60 method was able to accurately predict in vivo OMD. Conclusion The PCD60 and NDCFD methods gave similar OMD results. The in vivo method provides the benchmark against which all other methods of measuring OMD are evaluated. Compared with in vivo OMD, the NDCFD method gave significantly different OMD results while the PCD method tended to give different OMD results. The PCD60 method is regularly evaluated and was able to predict in vivo OMD more accurately than the NDCFD method. This study highlights the importance of comparing in vitro and in vivo methods regularly to ensure accurate OMD predictions. References Aufrère J., Baumont R., Delaby L., Peccatte J.R., Andrieu J., Andrieu J.R. and Dulphy J.P. (2007) Laboratory prediction of forage digestibility by the pepsin-cellulase method. The renewed equations. INRA Productions. Animales 20, 129-136. Aufrère J. and Michalet-Doreau B. (1988) Comparison of methods for predicting digestibility of feeds. Animal Feed Science and Technology 20, 203-218. Givens D.I. and Deaville E.R. (1999) The current and future role of near infrared reflectance spectroscopy in animal nutrition: a review. Australian Journal of Agricultural Research 50, 1131-1145. Morgan D.J., Stakelum G. and Dwyer J. (1989) Modified Neutral Detergent Cellulase Digestibility procedure for use with the 'Fibertec' System. Irish Journal of Agricultural Research 28, 91-92. O'Neill B.F., Lewis E., O'Donovan M., Shalloo L., Mulligan F.J., Boland T.M. and Delagarde R. (2013) Evaluation of the GrazeIn model of grass dry-matter intake and milk production prediction for dairy cows in temperate grassbased production systems. 1–Sward characteristics and grazing management factors. Grass and Forage Science 68, 504-523. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 618 Carbon sequestration in silage maize as affected by N fertilization Herrmann A.1, Böttger F.1, Lausen P.2 and Taube F.1 1 Grass and Forage Science/Organic Farming, Kiel University, Hermann-Rodewald-Strasse 9, 24118 Kiel, Germany 2 Agricultural Chamber Schleswig-Holstein, Grüner Kamp 15-17, 24768 Rendsburg, Germany Corresponding author: aherrman@email.uni-kiel.de Abstract Changes in soil organic carbon (SOC) stocks of forage cropping systems have a large impact on their greenhouse gas balance. Analysis of a long-term silage maize experiment indicate a positive impact of N fertilization on SOC stocks. Observed differences among treatments were higher than values obtained from a carbon balance based on literature data. Keywords: soil organic carbon, maize, yield, N fertilization, carbon balance Introduction Reducing the carbon footprint (CF) along the production chain, especially in forage production, which contributes the major share of greenhouse gas emission, is regarded a key to reduce ‘Livestock's Long Shadow’. Knowledge of mechanisms governing C sequestration, which has a significant impact on the CF of forage crops, however, is still limited. The objectives of the current study were to investigate the impact of N fertilization on SOC stocks in silage maize, based on a long-term experiment. Materials and methods The study was based on a continuous silage maize experiment with four replicates, established in 1993 on a sandy sand soil (pH 4.8-5.3) in northern Germany. Nitrogen fertilization was varied in seven levels: control (0 kg N ha-1), three mineral N treatments (30, 70, 110 kg N ha-1 applied as calcium ammonium nitrate (CAN)), and three cattle slurry (CS) treatments (20, 30, 40 m³). All fertilized treatments received a banded N starter, 40 kg N ha-1. Further nutrients were applied according to good agricultural practice in order to avoid any nutrient deficiencies. After 18 years of varied N supply, samples for SOC determination were taken in autumn 2011 in three soil layers (0-30, 30-60 and 60-90 cm). Samples were dried, sieved (< 2 mm), and ground to fine powder. Total C and N contents were measured using a CN-analyser (Vario Max CN, Elementar Analysensysteme, Hanau, Germany). The organic C content was obtained as the difference of total C and carbonate content (Scheibler, DIN ISO 10693). Soil bulk density was determined at 10-15, 40-45, and 75-80 cm depth, applying a core method. The belowground data were supplemented by an analysis of dry matter (DM) yield. A one-factorial analysis of variance was conducted to investigate the effect of N fertilization on SOC stocks, using SAS 9.2 Proc mixed. The impact of N fertilization on maize yield was analysed by first fitting a three-parameter exponential function to each combination of N treatment and replicate and then conducting one-factorial analyses of variance for each parameter separately. Multiple comparisons of means were conducted by the Tukey-Kramer method. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 619 Results and discussion Over the 18-year experimental period an increase of DM 14 yield was found for all fertilized treatments, probably 12 due to residual N effects from 10 fertilizer and maize roots, while yield of the control did 8 not change over time (Figure 0N 1). Thus, the control 6 30-CAN 70-CAN exhibited a significantly 110-CAN 4 different response function to 20 m³ CS the remaining treatments. In 30 m³ CS 2 40 m³ CS addition, significant diffe0 rences were detected between 1990 1995 2000 2005 2010 2015 110-CAN and all slurry Figure 1. DM yield response functions as affected by N fertilization. treatments, as well as between 30-CAN and 110CAN. Soil bulk density tended to be lower for the 0-N and 70-CAN treatments, but no significant impact was found (not shown). Soil organic C stocks were highest in the top layer, but still considerable amounts were detected in the subsoil (Table 1). It is known from literature that often more than 50% of the SOC stock is found in deep soil horizons (Batjes, 1996), affected by root depth allocation, leaching, abiotic decay conditions, and bioturbation. SOC stocks were higher in the fertilized treatments compared with the control in the 0-30 cm layer, but not in the subsoil. A significant impact of N fertilisation became evident only in the lowest soil layer, where the 30 m³ CS treatment revealed up to 30 t ha-1 higher SOC content than 0-N, 70-CAN and 20 m³ CS. DM yield (t ha -1) 16 Table 1. Soil organic C stocks (t C ha-1) as affected by N fertilization treatment and soil depth. Soil depth 0-30 cm 30-60 cm 60-90 cm 0-N 78.5 52.0 20.0a 30-CAN 92.0 59.9 35.7ab Nitrogen treatment 70-CAN 110-CAN 20 m³ CS 89.7 85.3 91.9 49.0 49.0 47.1 a ab 23.0 27.7 21.9a 30 m³ CS 96.4 57.2 50.1b 40 m³ CS 99.2 56.0 28.1ab Although mostly not significant, we evaluated if the differences among treatments could be reproduced when applying a C balance approach. To this end, the C remaining in the soil from the C input in terms of slurry, maize roots and stubble was estimated based on literature values. In detail, we assumed:  a stubble-to-shoot ratio of 6% - 8%, with shoot DM yield taken from measurements  a root-to-shoot ratio of 0.2-0.28 (depending on N fertilization; Amos and Walters, 2006)  a stubble and root C content of 44%  a ratio of C-rhizodeposition to total root C input of 0.30-0.44 (depending on N fertilization; Amos and Walters, 2006)  a root turnover of 11% (Hullugalle et al., 2010)  a method-related root underestimation of 36% (Subedi et al., 2006)  31% of the C input (root, stubble) and 26% of the slurry-C to remain in soil (Flessa et al., 2000; Bertora et al., 2009) Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 620 ∆ soil C estimated (t C ha-1) The estimated C amounts remaining in soil for the different treatments ranged between 8 and 20 t C ha-1 and highly correlated with the corresponding measured soil organic C stocks (R = 0.87*) in the top soil layer, while the relationship was less close (R = 0.68*) for the 0-60 cm layer. When comparing the differences in calculated C remaining and the measured C stocks among the treatments, we found similar agreement for the topsoil and the 0-60 cm layer. However, the observed differences among treatments were considerably larger than the estimated C remaining in soil from slurry and maize C input, as indicated by an intercept and a slope significantly different from zero and unity, respectively (Figure 1). The assumptions taken in our C balance calculation thus considerably underestimate either the C input from maize and/or slurry or 30 overestimate their decomy = 1.48x + 0.44 position rates. The root-to25 shoot ratio is a key factor R2 = 0.66* 20 determining the C input. A ratio of 0.2 at silage maize 15 harvest was estimated in a meta-analysis by Amos and 10 Walters (2006). The 5 underlying data, however, revealed a large variation 0 and were based largely on 0 10 20 30 older studies. Recent results -1 by Qi et al. (2012) indicate ∆ soil C observed (t C ha ) that yield progress can Figure 1. Observed differences in soil C stocks among treatments of the 0-1 partly be attributed to an in60 cm layer (t C ha ) versus corresponding calculated carbon remaining creased root-to-shoot ratio. from slurry and maize C input over the 18-yr period, with treatment 40 m³ CS serving as reference. Conclusion Observed and calculated differences in SOC stocks among N treatments reveal further need for research with respect to soil C changes in silage maize production. Future work will therefore focus on the application of models to investigate the processes governing carbon fluxes in silage maize cultivation. References Amos B. and Walters D.T. (2006) Maize root biomass and net rhizodeposited carbon: an analysis of the literature. Soil Science Society of America Journal 70, 1489-1503. Batjes N.H. (1996) Total carbon and nitrogen in the soils of the world. European Journal of Soil Science 47, 151– 163. Bertora C., Zavattaro L., Sacco D., Monaco S. and Grignani C. (2009) Soil organic matter dynamics and losses in manure maize-based forage systems. European Journal of Agronomy 30, 177-186. Flessa H., Ludwig B., Heil B. and Merbach W. (2000) The origin of soil organic C, dissolved organic C and respiration in a long-term maize experiment in Halle, Germany, determined by 13C natural abundance. Journal of Plant Nutrition and Soil Science 163, 157-163. Hulugalle N.R., Weaver T.B. and Finlay L.A. (2010) Carbon inputs by irrigated corn roots to a Vertisol. Plant Root 4, 18-21. Qi W-Z., Liu H.-H., Liu P., Dong S.-T., Zhao B.-Q., So H.B., Li G., Liu W.-D., Zhang J.-W. and Zhao B. (2012) Morphological and physiological characteristics of corn (Zea mays L.) roots from cultivars with different yield potentials. European Journal of Agronomy 38, 54-63. Subedi K.D., Ma B.L., Liang B.C. (2006) New method to estimate root biomass in soil through root-derived carbon. Soil Biology and Biochemistry 38, 2212-2218. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 621 Accuracy of forage intake estimation with three different indirect prediction models Salas-Reyes I.G.1,2, Martínez-Fernández A.1, Morales-Almaráz E.1,3, Jiménez J.D.1, AlbarránPortillo B.2, de la Roza-Delgado, B.1 and Vicente F.1 1 Servicio Regional de Investigación y Desarrollo Agroalimentario (SERIDA). P.O. Box 13, 33300 Villaviciosa (Asturias), Spain. 2 Centro Universitario UAEM Temascaltepec, 51300 Temascaltepec (Edo de México), Mexico 3 Present Address: FMVZ-UAEM, 50090 Toluca (Edo de México), Mexico. Corresponding author: fernandovicentemainar@gmail.com Abstract The aim of the study was to compare the precision of the dry matter (DM) intake prediction among three different indirect methods. Observed DM intake from two trials carried out with two groups of 15 and 12 cows fed on TMR ad libitum containing maize silage mixed with Vicia faba silage or grass hay, under continuous stall conditions. Dry matter intake was measured by using a computerized system proposed by Bach et al. (2004). The results of measured DM intake were compared with three indirect methods for estimating DM intake: a) intake prediction equations of NRC (2001); b) the animal performance method (APM) proposed by Macoon et al. (2003), and c) the model of prediction of intake capacity (IC; Faverdin et al., 2011). The actual DM intake was 20.19 kg d-1 and the estimated intakes were 21.60, 18.30 and 24.07 kg DM d-1 for NRC, APM and IC methods respectively. Considering the standard error of prediction, the APM method performs better in relation to the actual intake. Using the APM model, the correlation between the estimated value and the actual value was better when the model is applied to dairy cows with a milk production higher than the average production (R2=0.85). The APM method has 95% prediction accuracy, and can be used as a non-invasive tool that affects animal´s DM intake. Keywords: dairy cow, feed intake, estimation methods Introduction Milk production based on grazing represents an opportunity to reduce production costs and improving farm sustainability. However, accounting for forage intake is one of the most difficult tasks. Several methods have been developed in order to estimate forage intake, and the use of markers like N-alkanes or Cr are among the most frequently used. However, they are laborious, time consuming and expensive; therefore, they are unviable in the current context of funding crisis in scientific research. Other tools allow us to make forage intake estimations indirectly, by using prediction equations based on animal´s performance, physiological status and diet quality. The aim of this study was to determine the precision of forage intake prediction by comparing different methods against actual forage dry matter (DM) intake, in order to validate estimates applied to grazing animals. Materials and methods Measurements were made of DM intake from two trials carried out with two groups of 15 and 12 cows fed on TMR ad libitum containing maize silage mixed with Vicia faba (trial 1) or grass hay (trial 2), under continuous stall conditions. During each assay, the DM intake of individual cows was automatically recorded by an electronic weighing system integrated to the scale pans using a computerized system (Bach et al., 2004) in three periods of seven days each. Milk yield was recorded and sampled daily at both milking times from each cow. All cows were weighed and body condition score recorded on the first and last day of each period in both assays after the morning milking. Nutritive values of both TMR were analysed by NIRS. Milk samples were Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 622 analysed by a Milko Scan FT6000. The details of these assays are described in MoralesAlmaráz et al. (2010). The results of measured DM intake were compared with three indirect methods for estimating DM intake: a) intake prediction equations of NRC (2001), based on the energy requirements of animals; b) the animal performance method (APM) proposed by Macoon et al. (2003), which includes net energy content of feeds, and dairy cow requirements for maintenance, lactation and physical motion, and c) the model of prediction of intake capacity (Faverdin et al., 2011), based on three factors that affect the ability of intake (IC): body weight, potential milk production modified by the metabolizable protein, and body condition. Comparison of means was performed using the Paired-Samples t-test procedure. Results and discussion The actual intake of dry matter was 20.19 kg d-1 and the estimated intake was 21.60, 18.30 and 24.07 kg DM d-1 for NRC, APM and IC methods respectively. APM methods underestimated DM intake in 5%, whereas IC and NRC overestimated in 26 and 13% respectively (Figure 1). Considering standard error, the APM performs better, being highly related to real estimations of DM intake, whereas IC is the less accurate, as this method estimates animal´s maximum DM intake. The higher accuracy of the APM method can be explained because is directly related to the animal´s performance, and only integrates individual intake per animal, but its disadvantage is that intake response is only for a defined period of time. 140.0 PERCENTAGE 120.0 100.0 80.0 Real 60.0 40.0 20.0 0.0 APM NRC IC GRAZING Figure 1. Accuracy in the estimation of dry matter intake (as percentage, actual intake was 100%). [APM: Macoon et al., 2003; NRC: NRC, 2001; IC: Faverdin et al., 2011]. Considering that APM is the more accurate prediction method, a correlation between this method and milk production was determined. For above-average milk yields (35 L d-1), the APM predictions are more accurate (R2=0.85), than for below-average yields (R2=0.56; Figure 2). Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 623 Milk production/ APM Dry matter intake (kg d-1) 24.00 y = 0.3334x + 4.6477 R² = 0.8476 22.00 20.00 y = 0.7826x - 7.293 R² = 0.5581 18.00 16.00 ˂ 35 Liters 14.00 ˃ 35 Liters 12.00 10.00 25.00 30.00 35.00 40.00 Milk production (L 45.00 50.00 d-1) Figure 2. Correlation between milk yields and APM forage intake prediction Conclusions The APM is the most accurate indirect method for dry matter estimation, with predictions up 95% related with the actual measured intake, and it can be used as a non-invasive tool without affecting the animals' forage intake. Acknowledgements This work supported by Projects FICYT PC06-006, INIA RTA2007-0058-C02 and INIA RTA2011-00112, co-financed with the European Union ERDF funds. Research internship of Ms Isela G.Salas-Reyes was financed by CONACYT-México. Mr. José D. Jiménez is the recipient of an INIA (Instituto Nacional de Investigación y Tecnología Agraria y Alimentaria) Predoctoral Fellowship. References Bach A., Iglesias C. and Busto I. (2004) Technical note: A computerized system for monitoring feeding behavior and individual feed intake of dairy cattle. Journal of Dairy Science 87, 4207-4209. Faverdin P., Baratte C., Delagarde R. and Peyraud J.L. (2011) GrazeIn: a model of herbage intake and milk production for grazing dairy cows. 1. Prediction of intake capacity, voluntary intake and milk production during lactation. Grass and Forage Science 66, 29-44. Macoon B., Sollenberger L.E., Moore J,E., Staples C.R., Fike J.H. and Portier K.M. (2003) Comparison of three techniques for estimating the forage intake of lactating dairy cows on pasture. Journal of Animal Science 81, 23572366. Morales-Almaráz E., Soldado A., González A., Martínez-Fernández A., Domínguez-Vara I., de la Roza-Delgado B. and Vicente F. (2010) Improving the fatty acid profile of dairy cow milk by combining grazing with feeding of total mixed ration. Journal of Dairy Research 77, 225-230. NRC (2001) Nutrient requirements of dairy cattle 7th rev. National Research Council, National Academy Press, Washington, DC, 381 pp. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 624 Compatibility of using TiO2 and the faecal near-infrared reflectance spectrometry for estimation of cattle intake Vandermeulen S.1,2, Decruyenaere V.1,3, Ramirez-Restrepo C.4 and Bindelle J.1 1 Gembloux Agro-Bio Tech, Animal Science Unit, University of Liège, 2 Passage des Déportés B-5030, Gembloux, Belgium 2 National Fund for Scientific Research, 5 Rue d’Egmont, B-1000 Brussels, Belgium 3 Walloon Agricultural Research Centre (CRA-W), Production and Sectors Department, 8 Rue de Liroux, B-5030 Gembloux, Belgium 4 CSIRO, Animal, Food and Health Sciences, Australian Tropical Sciences and Innovation Precinct, James Cook University, Building 145, James Cook Drive, 4811 Townsville, QLD, Australia Corresponding author: sophie.vandermeulen@ulg.ac.be Abstract Combining titanium dioxide (TiO2) as indigestible marker to faecal near-infrared reflectance spectrometry (F-NIRS) can be used to determine cattle feed intake and quality of ingested forage if F-NIRS spectra are not modified by the marker. This study aimed at determining the compatibility of TiO2 with F-NIRS. Three dry cows were fed a standard hay-based diet for three weeks supplemented with a daily dose of 0.1% (10g) TiO2 during the last two weeks of the experiment. Faeces samples were collected every day and analysed for TiO2 and F-NIRS. Results suggest that TiO2 did not interfere with F-NIRS analyses. The calculations of crude protein, NDF, ADL contents, as well as dry matter intake did not change over time with increasing TiO2 in the faeces (P > 0.05). Slight differences observed for other predicted parameters seemed to be independent from TiO2. The higher Mahalanobis distance (H) for chemical composition (H = 7.2) independent from TiO2 inclusion could indicate that faecal spectra did not correspond exactly to the prediction database. Although 0.1% incorporation of TiO2 seem not to interfere with F-NIRS measurements, caution must be taken with higher levels of TiO2 as nothing indicates that interference could not appear. Keywords: Ruminant, titanium dioxide, faecal near-infrared spectrometry, intake, diet chemical composition. Introduction Methods used to determine feed intake and quality of consumed forage in grazing cattle are time-consuming, expensive and sometimes controversial in respect to animal welfare as they include different techniques such as sward clipping techniques or oesophageal fistulated animals (Decruyenaere et al., 2009). A combination of an indigestible marker to faecal nearinfrared reflectance spectrometry (F-NIRS) can provide a useful alternative, providing that the marker fed daily to the animal does not interfere with F-NIRS spectra (Titgemeyer et al., 2001; Decruyenaere et al., 2012). As previous studies report that Cr2O3 is likely to interfere with NIRS calibration data (Decruyenaere et al., 2012), this study investigates if titanium dioxide (TiO2) used as indigestible marker was compatible with F-NIRS analysis. Materials and methods A three-week experiment was performed on three dry red-pied cows housed in free stalls in the Animal Science Unit of GxABT (Gembloux, Belgium). All cows received 7 kg d-1 of standard temperate hay and 2 kg d-1 of a mixed concentrate and had a free access to water. After an adaptation period of one week, 10 g of TiO2 mixed with 50 ml of molasses was distributed every day to each cow until the end of the experiment. The faeces were collected every day, dried at 60 °C, and ground to pass a 1-mm screen prior to TiO2 and F-NIRS analyses. TiO2 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 625 dosage in faeces was performed according Myers et al. (2004). The F-NIRS analyses were achieved as described by Decruyenaere et al. (2012). The chemical composition, dry matter intake (DMI) and in vivo organic matter digestibility (OMD) predicted from the F-NIRS database were compared daily along the entire experiment using the MIXED procedure of SAS 9.2 with the ‘cow×day’ as experimental unit. The correlation between these parameters and TiO2 content in the faeces was calculated with the CORR procedure of SAS 9.2. Results and discussion Figure 1 shows the evolution of TiO2 contents in the faeces before and during the daily incorporation of 10 g of TiO2 in the diet (day 6 being the first day of TiO2 distribution). During the adaptation period without TiO2, its faecal content was equal or close to zero, before increasing and then reaching a plateau towards the end of the experiment; so the TiO2 dosage did not face interference problems. Dietary TiO2 did not interfere with the F-NIRS analysis. Figure 1 Evolution of TiO2 contents (mg g-1) in the faeces of the three cows Prediction of crude protein (CP), neutral detergent fibre (NDF) and acid detergent lignin (ADL) contents as well as the DMI (Figure 2) did not change over time (P > 0.05; P = 0.0723 for NDF content as the lowest P-value). Despite some changes along days for the acid detergent fibre (ADF, P = 0.0381) content and the in vivo OMD (P = 0.0009) (Figure 2), the slightly different values seemed to appear independently before or after ingestion of titanium dioxide. The DMI (P = 0.4613; Figure 2), which seemed more fluctuating along the experiment, could be explained by individual differences of intake; for example, the 13th day of the experiment, the DMI of one cow reached 85.3 vs 47.5g kg-1 metabolic weight (MW) for that of another cow. The standardized Mahalanobis distance (H) which evaluates the correspondence between the faeces spectra and the F-NIRS database should ideally be lower than 3 for an accurate prediction (Shenk and Westerhaus, 1991). For OMD and DMI, the average distance H was below 3 for the DMI and the OMD (H = 2.8 and 2.87 respectively; two thirds of the samples being lower than 3) while it reached 7.2 for the chemical composition. This should probably not be due to TiO2 inclusion in the diet, but rather to a discrepancy between the samples and the calibration dataset. The results of correlation between the titanium dioxide content in the faeces and the parameters predicted by F-NIRS should be considered with caution. Indeed, most of parameters, as DMI or OMD, were not significatively correlated to the TiO2 content (P > 0.05 ; P = 0.1086 and r = 0.23206 for CP as the highest P-value and correlation coefficient (r)) but other parameters were significatively correlated to TiO2 content, such as total ash (TA), NDF and ADF content (P < 0.0001 and r = 0.58651 for TA). Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 626 Figure 2: Means and standard deviation of CP, NDF and ADF contents (g kg-1 of DM) of the diet and the DMI (g kg-1 MW d-1) and in vivo OMD (g kg-1) predicted by F-NIRS along the experiment Conclusion Feeding 10 g d-1 TiO2 as indigestible marker in cattle (0.1% incorporation level) did not interfere with F-NIRS prediction. The use of these results should be done with caution, as there is nothing that indicates that interference could not appear when higher levels of TiO2 are incorporated in the diets. Acknowledgments The authors are grateful to Walloon Agricultural Research Centre for the near-infrared reflectance spectrometry analyses. Sophie Vandermeulen, as a research fellow, acknowledges the support of the National Fund for Scientific Research (Brussels, Belgium). References Decruyenaere V., Froidmont E., Bartiaux-Thill N., Buldgen A., and Stilmant D. (2012). Faecal near-infrared reflectance spectrometry (NIRS) compared with other techniques for estimating the in vivo digestibility and dry matter intake of lactating grazing dairy cows. Animal Feed Science and Technology. 173(3), 220-234. Decruyenaere V., Buldgen A. and Stilmant D. (2009). Factors affecting intake by grazing ruminants and related quantification methods: a review. Biotechnologie, Agronomie, Société et Environnement. 13(4), 559-573. Myers W. D., Ludden P. A., Nayigihugu V. Hess and B. W. (2004). Technical Note: A procedure for the preparation and quantitative analysis of samples for titanium dioxide. Journal of Animal Science. 82(1), 179-183. Shenk, J.S. and Westerhaus, M.O. (1991). Population definition, sample selection, and calibration procedures for near infrared breflectance spectroscopy. Crop Science. 31(2), 469–474. Titgemeyer E.C., Armendariz C.K., Bindel D.J., Greenwood,R.H. and Löest C.A. (2001). Evaluation of titanium dioxide as a digestibility marker for cattle. Journal of Animal Science. 79 (4), 1059-1063. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 627 Dry matter intake and in vivo digestibility of four perennial ryegrass cultivars Garry B.1,2, O’Donovan M.1, Boland T.M.2 and Lewis E.1 1 Teagasc, Animal & Grassland Research and Innovation Centre, Moorepark, Fermoy, Co. Cork, Ireland 2 School of Agriculture and Food Science, University College Dublin, Belfield, Dublin 4, Ireland Corresponding author: Brian.Garry@teagasc.ie Abstract There is renewed interest in improving animal performance in grazing systems by using highly productive perennial ryegrass cultivars. Four cultivars (Glenroyal (GR), Delphin (DE), Tyrella (TY) and Astonenergy (AS)) were evaluated for in vivo dry matter intake (DMI) and dry matter digestibility (DMD) using Texel wether sheep housed in digestibility stalls. The structural profile and leaf, stem, and dead proportion of the sward were also measured. There was no effect of cultivar on DMI during two time periods (May/June (LS1) 1.53±0.069kg and Aug/Sept (LS2) 1.50±0.065 kg (P>0.05)). There was a cultivar × LS interaction for DMD. In LS1, TY had a greater DMD than DE (P<0.05). A reduction occurred in DMD from LS1 to LS2. This was associated with an increased amount of dead material in the sward in LS2. In LS2 there was no significant difference between cultivars in DMD but TY and GR had a greater proportion of dead material than DE and AS. These results indicate that there are small differences between perennial ryegrass cultivars in DMD but large effects of season on perennial ryegrass cultivar DMD. Keywords: In vivo digestibility, dry matter intake, perennial ryegrass, cultivar Introduction Dairy systems that maximize grass utilization are highly competitive and sustainable (Peyraud et al., 2010). However, fundamental to maintaining this competiveness is increasing efficiency within the system, by using new innovations and technology. Reseeding permanent pasture with perennial ryegrass (Lolium perenne L.) has been shown to increase farm profitability (Shalloo et al., 2011). Perennial ryegrass is the most commonly sown grass species in temperate climates as it has high growth potential and high nutrient value (Tas et al., 2005). Cultivars of perennial ryegrass vary in feeding value, due to differences in ploidy and heading date (Gowen et al., 2003) as well as sward structure (Wims et al., 2013). Tetraploid cultivars have higher in vitro organic matter digestibility (OMD) than diploid cultivars (Wims et al., 2013). There is a need to quantify any differences between cultivars in vivo, to be confident that in vitro differences are truly representative. The objective of this experiment was to determine the effect of four perennial ryegrass cultivars on dry matter intake (DMI) and dry matter digestibility (DMD) in sheep. Materials and methods Twelve Texel wether sheep were used to determine the DMI and in vivo DMD of four perennial ryegrass cultivars. The experiment was run as a 4×2 incomplete Latin square design. The four treatments (TRT) comprised four perennial ryegrass cultivars: Astonenergy (tetraploid, heading date 31 May) (AS), Delphin (tetraploid, heading date 1 June) (DE), Tyrella (diploid, heading date 3 June) (TY) and Glenroyal (diploid, heading date 3 June) (GR). The experiment was repeated twice: the first Latin square (LS1) was from 13 May to 7 June 2013 and the second Latin square (LS2) was from 19 August to 13 September 2013. All cultivars were sown in spring 2012 at a sowing rate of 32 kg/ha for diploids and 40 kg/ha for tetraploids. In April 2013, six experimental plots per cultivar were marked and each plot was designed to feed 3 sheep for Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 628 4 d. These plots were cut 3 to 4 weeks prior to feeding, to allow a pre-grazing herbage mass of approximately 1500 kg DM/ha at the time of feeding. The sheep were housed in individual stalls allowing for the total collection of urine and faeces. Body weight was a blocking factor at the start of each LS. Each LS had two periods and each period consisted of 12 days: 6 days adaptation and 6 days measurement (MP). Fresh grass was cut and chopped daily using a motor Etesia (Etesia UK Ltd., Warwick, UK). Sheep were fed ad libitum (110% of DMI) and grass DMI was recorded daily. Feeding times were 0900h and 1630h. During the MP, a representative sample of the grass offered and faeces voided by each sheep was collected daily. Daily grass and faeces samples were dried and then bulked to give one sample of each per cultivar per MP. On day-8 of each period a 40 g sample of each cultivar was separated into leaf, pseudostem, true stem and dead proportions >4 cm. The sward profile was measured from ground level prior to cutting once during each MP. The extended tiller height (ETH), pseudostem height (PH) and free leaf lamina (FLL) were measured on 100 tillers. Pre-grazing herbage mass was measured using Gardena hand shears (Accu 60, Gardena Int. GmbH, Ulm, Germany) and a 0.25 m2 quadrat, four times per cultivar during each period. Data for DMI, DMD and leaf, pseudostem, true stem and dead proportions >4 cm were analysed using PROC MIXED in SAS (2002). Treatment, period within LS, LS and the interaction between LS and treatment were included as fixed effects. Sheep was included as the random effect. Sward structural effects were analysed using PROC GLM and treatment, period within LS, LS and the interaction between LS and treatment were included as fixed effects. Results and discussion Results are presented in Table 1. Table 1. Dry matter intake (DMI), in vivo digestibility (DMD), leaf, stem, dead proportions, and sward profile of cultivars: Astonenergy (AS), Delphin (DE), Tyrella (TY) and Glenroyal (GR) TRT DMI (kg DM) DMD (g/kg) Leaf (%) Pseudostem (%) True stem % Dead (%) ETH (mm) PH (mm) FLL (mm) AS 1.44 837ab 62.9ac 21.0b 6.9a 9.3a 256.2 105.5 150.6 DE 1.53 796b 63.5abc 23.3a 5.5b 8.1a 243.6 91.2 153.0 TY 1.58 845a 63.3abc 24.4a 3.7c 8.7a 225.8 98.6 127.1 GR 1.58 826ab 65.2abc 21.0b 4.7cb 9.1a AS 1.57 734 c ab c DE 1.59 746c 68.1b TY 1.48 717c 64.3abc 1.38 713 c 61.6 c 15.5 SEM 0.115 15.4 1.45 TRT NS * LS NS LS× TRT NS LS1 LS2 GR 191.7 70.7 121.5 0 15.9 b 302.7 65.3 237.5 17.2c 0 14.9b 295.5 66.0 229.5 15.7c 0 19.1c 293.6 60.3 233.0 0 21.9 c 243.8 53.0 190.9 0.67 0.48 1.08 14.21 6.85 18.6 * *** ** *** * NS NS *** NS *** n/a *** * ** * ** *** *** n/a *** NS NS NS 67.9 15.9 c ETH-Extended tiller height, PH- pseudostem height, FLL, free leaf lamina, For DMD, there was a significant treatment × LS interaction (P<0.01). In LS1, TY had a significantly higher DMD than DE (P<0.05). The other two cultivars were intermediate Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 629 (P>0.05). In LS2, all four cultivars had similar DMD to each other (P>0.05). The difference in DMD between cultivars in LS1 could be due to differences in chemical composition (Wims et al., 2013) or to differences in sward structure (Beecher et al., 2013) which can be particularly evident during the reproductive phase. The reproductive phase coincided with LS1. In LS1 there was a higher amount of true stem in DE compared to TY, which agrees with the lower DMD found in DE. True stem has been shown to reduce the digestibility of swards (Beecher et al., 2013). The DMD of all cultivars decreased significantly from LS1 to LS2. The decrease in DMD in LS2 compared to LS1 may be attributed to a high soil moisture deficit in August and September 2013. This resulted in an increase in the dead proportion in the sward, which is the most indigestible morphological fraction in perennial ryegrass (Beecher et al., 2013). The dead proportion was higher in LS2 compared to LS1 (P<0.001) for each cultivar. This also agrees with (Hazard et al., 1998) who found that differences in digestibility between cultivars were associated with differences in the proportion of dead material in the top lamina. The PH was significantly higher in LS1 P<0.01), this shows that the stem was lengthening as it was becoming reproductive, and agrees with the higher proportion of true stem in LS1. This, combined with a lower ETH in LS1 (P<0.05), lead to a lower FLL content in the sward in LS1 than in LS2 (P<0.01). The lower FLL in LS1 did not however lead to lower DMD. The DMI for LS1 was 1.53±0.069 kg and for LS2 was 1.50 ±0.065 kg (P>0.05). There was no difference between cultivars for DMI during both LS. This differs from (Gowen et al., 2003) (O'Donovan and Delaby, 2005) who found that cows differed in their intake of cultivars, but in those studies the animals were grazing. In the present study the grass was cut and offered to the animals indoors. When cultivars were cut and fed to animals inside there was no difference in DMI (Tas et al., 2005), similar to the current study. Conclusion This study shows that cultivar has a small effect on sward DMD but season had a large effect on sward DMD. In order to improve the nutritive value of a sward, it is necessary to minimize the dead proportion of the sward. References Beecher M., Hennessy D., Boland T.M., McEvoy M., O'Donovan M. and Lewis E. (2013) The variation in morphology of perennial ryegrass cultivars throughout the grazing season and effects on organic matter digestibility.Grass and Forage Science:. doi: 10.1111/gfs.12081 Gowen,N., O'Donovan M, Casey I., Rath M., Delaby L. and Stakelum G. (2003) The effect of grass cultivars differing in heading date and ploidy on the performance and dry matter intake of spring calving dairy cows at pasture. Animal Research 52, 321-336. Hazard L., De Moraes A., Betin M. Traineau R. and Emile J.-C. (1998) Perennial ryegrass cultivar effects on intake of grazing sheep and feeding value. Annales de Zootechnie 47, 117-125. O'Donovan M. and Delaby L. (2005) A comparison of perennial ryegrass cultivars differing in heading date and grass ploidy with spring calving dairy cows grazed at two different stocking rates. Animal Research 54, 337-350. Peyraud J.L., Pol-van Dasselaar A.v.d., Dillon P. and Delaby L. (2010) Producing milk from grazing to reconcile economic and environmental performances. Grassland Science in Europe 15,163-164. Shalloo L., Creighton P. and O'Donovan M. (2011) The economics of reseeding on a dairy farm. Irish Journal of Agricultural and Food Research 50, 113-122. Tas B.M., Taweel H.Z, Smit H.J., Elgersma A., Dijkstra J. and Tamminga S. (2005) Effects of perennial ryegrass cultivars on intake, digestibility, and milk yield in dairy cows. Journal of Dairy Science 88, 3240-3248. Wims C.M., McEvoy M., Delaby L., Boland T.M., and O'Donovan M. (2013) Effect of perennial ryegrass (Lolium perenne L.) cultivars on the milk yield of grazing dairy cows. Animal 7, 410-421. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 630 Accurate monitoring of the rumination behaviour of cattle using IMU signals from a mobile device Andriamandroso A.L.H.1, Lebeau F.2 and Bindelle J.3 1 AgricultureIsLife platform, 2 Mechanics and construction unit, 3 Animal science unit, Gembloux Agro-Bio Tech, ULg, Belgium Corresponding author: alh.andriamandroso@ulg.ac.be Abstract Improving the monitoring of rumination in cattle could help in assessing their welfare status and their risk of acidosis. In this work, the monitoring of cattle behaviour was performed using the inertial measurement unit (IMU) present in smartphones mounted on the neck of cows. The processing of both time and frequency domains of the IMU signals was capable of detecting accurately the main behaviours (grazing, rumination and other) and highlight the characteristics of the rumination process. The algorithm for analysis of rumination was more accurate for grazing cattle than for silage-fed cattle in stables. Keywords: cattle, behaviour, rumination, signal processing, IMU Introduction All animal species display behaviours that indicate their physical, physiological and welfare status (Frost et al., 1997). For ruminants, grazing, ruminating and resting behaviours occupy more than 90% of the time budget of the animal on pasture (Kilgour, 2012). Rumination represents 5 to 9h/d for cattle (Vallentine, 2001). It is a cyclic process which completes the chewing of fibrous ingested feed after it has undergone anaerobic fermentation by microbes in the rumen. A cycle begins with the regurgitation of a rumino-reticular bolus followed by semicircular jaw movements and ends with the deglutition (Jarrige et al., 1995). Rumination routines are influenced by pasture quality, intake quantity especially in fibrous content (Jarrige et al., 1995; Fustini et al., 2011), physiological status, and the level of stress and anxiety of the animal (Soriani et al., 2012; Braun et al., 2013). In dairy cows, high-yielding individuals are fed high levels of concentrates leading to a low level of insoluble fibre intake. This low level of fibre puts them at risk of both acute and chronic acidosis, since it induces a decrease in the duration of the rumination in the daily cycle (De Vries et al., 2009). Characterizing rumination is therefore an interesting indicator of health and welfare in ruminants. Recent developments in sensor technology lead to their possible use for automated detection of domestic ruminants’ main behaviours, such as rumination or grazing using mainly GPS and accelerometers (Swain et al., 2008). The present work aimed at developing a white-box approach to detect differences in rumination patterns in order to enable its accurate monitoring based on signal processing of a mobile device’s IMU fixed on cows. Material and methods Two red-pied dry cows were fitted with an iPod Touch 4G or an iPhone 4S on their neck. Nine recording sessions were performed between September 2012 and November 2013: seven with the cows grazing a ryegrass-white clover pasture, and two in stables when cows were fed silagebased diets. Each session included: (1) 100Hz data acquisition from the 3-D accelerometer and the 3-D gyroscope of the IMU by means of Sensor Data software (Wavefrontlabs), and (2) simultaneous video recording of the cows to allow accurate observation of the behaviours and their decomposition as sets of movements of the head or the jaw. The dataset was divided in two independent sets, one for calibration (the 3 first fields’ data) and one for validation of the procedure (the 6 other data). Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 631 The data analysis included two steps and was performed using Matlab R2013a. The first task was to create an algorithm for the detection of the main behaviours: grazing vs. ruminating vs. other. This algorithm was based on criteria from the movements’ decomposition on 3Daccelerometer and gyroscope signals. The results of the classification were compared with the observed behaviours on the validation dataset to calculate the detection accuracies. The second part of the work focused on rumination. The duration and the number of bites were counted on the video files. The interesting IMU signal, chosen according to the most discriminative movement for the rumination process, was analysed on time and frequency domain using fast Fourier transform (FFT) and its inverse. A first filtration process was performed by choosing the most recurrent frequency between 1Hz and 6Hz to eliminate noise waves from the raw data to bring out the actual signal from rumination. Two algorithms were developed to characterize the deglutition and the mastication (number and duration of deglutition, number and duration of bites). The deglutition was described as a pause between two bouts of mastication (mobile standard deviation of a 2s sample of the selected signal<0.03rad/s at deglutition) while the mastication was known by its duration ranging between 15s and 60s. For the mastication bouts, a second filtration between 1 and 2Hz corresponding to normal frequency of bites during rumination was done. The number of mastication is counted with the number of zero crossing waves on the filtered signal. The foreseen results were compared with the observed data from the validation dataset. Results and discussion The detection accuracies for the grazing and rumination ranged between 90% and 100% on pasture and between 80% and 84% in the stable. The less-accurate detection in the stable is probably due to the different ration fed to the cows. The rumination process was analysed using the rotation rate signal along the x-axis of the mobile device which is aligned with the cows nose to tail axis. This signal shows the particular jaw movement of the cattle during rumination best, showing a discriminant peak between 1 and 2Hz on the frequency domain (Figure 1). This behaviour is characterized by succession of 32 to 48s of mastication and 2 to 4s of pause for deglutition. As shown in Fig 2, correct measurements were high for both duration and number of mastications in which over estimation is not respectively greater than 7s and 7 bites for fields’ data. For the number of mastications, a bite corresponds to four zero crossing waves. The average mastication rate equals to 1.06 ± 0.06 bites/s. For the stable sessions, the correct measurement yielded from the field data was much lower. The deglutition’s standard deviation is lower than 0.03rad/s. Figure 1. Time and frequency domain patterns for 10s mastication periods during rumination and grazing for field data along rotation rate on x-axis Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 632 Figure 2. Comparison between the observed and the measured duration and number of mastication for field and stable validation data The duration of mastication is also more scattered (between 15s and 40s). This situation is explained by the difference of the diets fed to the animals and requires further data processing to improve the algorithm. Conclusion The signals recorded from the IMU of iPhones and iPods offered an accurate detection of the behaviours of cattle. Deeper analysis about the rumination could measure its principal characteristics such as the duration and the number mastication but the approach requires further improvements to account for major changes in rumination patterns induced by differences between pasture and silage-based diets. References Braun U., Trösch L., Nydegger F. and Hässig M. (2013) Evaluation of eating and rumination behaviour in cows using a noseband pressure sensor. BMC Veterinary Research 9, 164. De Vries T.J., Beauchemin K.A., Dohme F. and Schwartzkopf-Genswein K.S. (2009) Repeated ruminal acidosis challenges in lactating dairy cows at high and low risk for developing acidosis: feeding, ruminating, lying behavior. Journal of Dairy Science 92, 5067-5078. Frost A.R., Schofield C.P., Beaulah S.A., Mottram T.T., Lines J.A. and Wathes C.M. (1997) A review of livestock monitoring and the need of integrated systems. Computers and Electronics in Agriculture 17, 139-159. Fustini M., Palmonari A., Bucchi E., Heinrichs A.J. and Formigoni A. (2011) Chewing and ruminating with various forage qualities in nonlactating dairy cows. The Professional Animal Scientist 27, 352-356. Jarrige R., Dulphy J.P., Faverdin P., Baumont R. and Demarquilly C. (1995) Activité d’ingestion et de rumination. In : Jarrige R., Ruckebusch Y., Demarquilly C., Farce M.H. and Journet M. (eds) Nutrition des ruminants domestiques:ingestion et digestion, INRA, Paris, France, pp 123-182. Kilgour R. (2012) In pursuit of ‘normal’: review of the behaviour of cattle at pasture. Applied Animal Behaviour Science 138, 1-11. Soriani N., Trevisi E. and Calamari L. (2012) Relationships between rumination time, metabolic conditions, and health status in dairy cows during the transition period. Journal of Animal Science 90, 4544-4554. Swain D.L., Bishop-Hurley G.J. and Wark T. (2008) Using high fix rate GPS data to determine the relationships between fix rate, prediction errors and patch selection. Ecological Modeling 212, 273-279. Vallentine J.F. (2001) Grazing Management. Burlington Academic Press, San Francisco, United States, 659 pp. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 633 Energy consumption and greenhouse gas emissions of DAIRYMAN farms in South-West Germany Jilg T., Herrmann K., Hummler T. and Elsaesser M. Landwirtschaftliches Zentrum (LAZBW), Atzenberger Weg 99, 88326 Aulendorf, Germany Corresponding author: Martin.Elsaesser@lazbw.bwl.de Abstract The EU Interreg IVb NWE project DAIRYMAN was established in 2009. Its aim was to enhance the sustainability of dairying in North-West Europe by improving the competitiveness and ecological performance of dairy farming. Therefore, farm information - economic, ecological and social - was collected from a pilot farm network of 127 dairy farms in 10 regions of North-West Europe from 2009-2011 in order to evaluate and compare farm performances. Fouteen farms in South-West Germany (Baden-Wuerttemberg) participated in this network. In addition to the data that were collected for the whole network, the LAZBW Aulendorf determined energy consumption and greenhouse gas emissions for these German farms in 2010 and 2011. This was made with the AgriClimateChange Tool (ACCT), a newly developed EXCEL-based tool that enables the calculation of energy inputs and output, energy efficiency, nitrogen balances and greenhouse gas emissions. This tool is suitable for indicating the correlations of influencing factors, and to clearly visualize the farmers’ efforts of adapting their farm management to climate-friendly systems. Keywords: nitrogen balance, greenhouse gases, energy consumption, dairy farm system Introduction DAIRYMAN was an EU Interreg IVb NWE project which was established in 2009 with 14 partners under the lead of the University of Wageningen (Aarts, 2012). The project ended in August 2013. The main objective of DAIRYMAN was the investigation of dairy farming systems with regard to ecological, economic and social performances. This was realized by establishing a pilot farm network of 127 dairy farms. In Baden-Wuerttemberg, the German Dairyman partner, there were 14 farms which participated in the network. At the beginning of the project special development plans were worked out for each farm, so that the initial farm values could be compared with the values reached at the end of the project. With the AgriClimateChange Tool (ACCT), developed by 4 partners of the LIFE-funded project Agriclimate Change (Solagro, Bodensee Stiftung, Región de Murcia, Communità Montana and Fundacíon Global Nature (Solagro, 2013)), there is now a tool available which can be used throughout the European Union in order to show energy consumption and greenhouse gas emissions at a farm scale. In 2010 and 2011 such measurements with ACCT were made on the 14 German dairy farms. In particular, the influence of concentrates used and of the nitrogen inputs on the global energy efficiency needed to be investigated. Material and methods As basic information the ACCT (Solagro, 2013) uses the direct energy input (fuel, electric energy, water) and the indirect energy input (concentrates, fertilizer, machines, buildings, agricultural pesticides, seeds, animals and other synthetic materials). Products leaving the farm (like milk, meat and crops) are used for calculating the energy output. The 14 selected farms were not representative of the average dairy farms in Baden-Wuerttemberg, but they stand for typical farms in the region with successful milk production. Farms from four typical milk regions and different climate conditions were examined: Swabian Alb, Black Forest, Oberschwaben and Allgaeu. The selected farm types were divided into dairy farms with grass Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 634 feeding, dairy farms with maize feeding and dairy farms combined with biogas production. Details are reported in Table 1. Table 1. Mean values, standard deviation, minimum and maximum values of selected farms for the attributes examined Attributes Unit Agricultural area Agricultural area (AA) ha ha ECM Milk per farm area µ ±s min max 123.9 92.6 56.5 31.4 54.6 45.5 265 159.9 kg cow-1 y-1 8757 932 6999 10150 -1 10106 3656 5226 17346 180.9 25 67 10 79.1 10 280.6 60 kg ha Livestock units (LU) Concentrate use in % of roughage % Fuel consumption GJ ha-1 7.6 2.3 4.2 11.2 -1 10.3 4.5 4.7 20,1 Feed purchase Fertilizer Input total -1 GJ ha GJ ha-1 GJ ha-1 8.5 5.5 37.1 4.5 2.7 10.3 3.6 0 20.3 18.8 11 55.4 Milk GJ ha-1 26.8 11.8 11.3 47.5 Meat GJ ha-1 2.8 1.3 1 5.7 Cultures GJ ha-1 35.2 45.1 0 133.7 64.9 38.4 22.7 147.5 Electricity per kg milk kJ kg Output total Global energy efficiency (GEE) t CO2 1.71 08 1 3.4 ha AA-1 y-1 9.8 3.3 5.8 15.7 t CO2 LU-1 y-1 6.3 1.2 4.5 9.3 1.07 0.1 0.8 1.2 80 44.9 -5 170.5 Energy efficiency dairy branch N balance kg ha -1 Results and discussion The energy consumption for fuel, fertilizer and feed purchase varied widely between the farms. Focusing on the global energy efficiency (GEE), which includes the branches milk, crop and biogas production, the differences between farms were remarkable. Dairy farms with biogas production had a 2 - 3 times higher GEE than dairy farms without biogas production. Only farms with biogas reached GEE values higher than 2.0. The energy efficiency for the dairy branch ranged between 0.8 and 1.2. It is remarkable that there is no close relation between GEE and nitrogen fertilization (Figure 1) and between GEE and the use of concentrates (Figure 2). Fertilizer used in energy input was only 15%, and it was obvious that the energy consumption from feed, electricity and fuel had more influence on the total energy input. Nevertheless, Figure 2 shows that increasing amounts of concentrates reduced the GEE. Regarding only the dairy branch of the farms, it seems that the amounts of concentrates are positively correlated to the energy efficiency (Figure 3). This may be due to a more efficient milk production with higher amount of concentrates. Energy efficiency is increasing until a milk performance of 4000 kg milk from roughage. More milk from roughage seems to have no further benefits in terms of energy efficiency. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 635 Global energy efficiency (GEE) Global energy efficiency vs N balance 4.0 3.5 3.0 y = 0.0008x + 1.6469 2.5 R² = 0.0018 2.0 1.5 1.0 0.5 0.0 -50 0 50 100 150 200 N-balance, kg Global energy efficiency (GEE) GEE vs amount of concentrates 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0 y = 8,1563x-0,538 R² = 0,4378 0 20 40 60 Concentrates in % of roughage Figures 1 and 2: Relation between global energy efficiency and N balance respectively portion of concentrates in roughage Energy efficience dairy sector Energy efficiency of dairy branch 1.4 1.2 1.0 0.8 0.6 y = -3E-07x2 + 0.0024x - 3.7601 R² = 0.5555 0.4 0.2 0.0 2500 3000 3500 4000 4500 5000 milk from roughage in kg cow-1 y-1 Figure 3: Energy efficiency of dairy sector related to milk performance from roughage Conclusions The dairy branches of the investigated farms differ in energy input, output and efficiency. This gives potential for improvements in optimizing the management of the farms on an individual basis, and it shows that intensive as well as extensive farms can be managed efficiently. The use of ACCT might be a helpful tool for farmers and extension services by showing strengths and weaknesses in terms of energy consumption of the individual farm. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 636 Acknowledgements The authors thank their partners in the DAIRYMAN project, especially Dr. Jacques Neeteson and Dr. Frans Aarts, Wageningen for the friendly collaboration and the EU Interreg-Programm IVb NWE and the Ministry of Laendlicher Raum and Verbraucherschutz in BadenWuerttemberg for the financial support. References Aarts F. (2012) DAIRYMAN for a more efficient use of resources by dairy farms. Grassland Science in Europe 17, 753-755. Solagro (2013) Manual to Agriclimatechange, www.agriclimatechange.eu Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 637 The DAIRYMAN-Sustainability-Index (DSI) as a tool for comparing dairy farms Elsaesser M., Herrmann K. and Jilg T. Agricultural Centre Baden-Wuerttemberg for cattle production, grassland management, dairy processing, wildlife research and fisheries (LAZBW), Aulendorf, Germany Corresponding author: Martin.Elsaesser@lazbw.bwl.de Abstract Sustainability of dairy farms is determined by a multiplicity of single indicators. A combination of these single factors established useful sustainability indicator systems, but Von Wiren-Lehr (2001) report the truth of sustainability is not possible even if complex models or time consuming measurements are used. Therefore the team of the Interreg IVb NWE project Dairyman chose a pragmatic way to install an indicator system for comparisons of so-called sustainability of dairy farm systems. The number of indicators for the calculation of the DSI was substantially reduced on 17 indicators. The DSI is not yet a fixed system, but it can be a proposal for an evaluation of sustainability. Keywords: sustainability, dairy farming, nutrient balances, farm economics Introduction Sustainability indicator systems have been established that combine a multiplicity of single indicators (Girardin, 2001; Belanger et al., 2012), for example KUL (Breitschuh and Eckert, 2006), RISE (HAFL, 2012) or MOTIFS (Meul et al., 2008). But how can the sustainable development of an individual farm be visualized and assessed? And why is it useful? A single characterization of 'sustainability' indicators does not give a good view of the whole-farm situation. More attractive, and of greater information value for farmers and advisors, is the combination of single factors in a so-called integrative view (Von Wiren-Lehr, 2001). An objective of the Interreg IVB NWE project DAIRYMAN (Aarts, 2012) was to create a management tool which is suitable for evaluating the development of sustainability of dairy farms as a combination of single indicators. This should visualize individual farm developments with the possibility to show differences in dairy systems in time and region. The DSI represents a holistic assessment of the DAIRYMAN pilot farms. It can help to show strengths and weaknesses of dairy systems. Materials and methods A network of 127 pilot dairy farms was installed in the Dairyman project in order to measure and observe processes in practical dairy farming in various countries of North-West Europe. For this purpose a high number of farm data in 2009-11 were collected. In a first step the chosen indicators for the DSI were selected by the project partner in Germany, after intensive discussions and with the use of a questionnaire answered by pilot farmers, farm advisors and teachers of agricultural schools. In a second step, the factors were further selected and discussed between all partners of DAIRYMAN in meetings (Elsaesser et al., 2013; Grignard et al., 2013). Based on the 'three-pillar model' of sustainability, it was decided that ecological, economic and social aspects would be treated equally, so that in each pillar a maximum of 100 points could be reached. All chosen factors were clearly defined and it was decided that they should be already gathered within the pilot farm network of all regions in order to reach an acceptable cost-benefit ratio. Biodiversity and soil erosion-susceptibility, although they are important attributes, could not yet be taken into account. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 638 Results Even though the task of the DSI was to harmonize the scoring values between all partners in the DAIRYMAN project, this objective could not be realized until now because single indicators of the DAIRYMAN partners are differently evaluated according to their importance in different regions. At first, this problem seemed unsolvable, but there are solution approaches (Larochelle et al., 2007; Meul et al., 2008; Belanger et al., 2012). The individual scores as a result of the discussions are summarized in Table 1. It is the task of the DSI user to discuss and interpret the gained results in a second step. Table 1. Indicator quantiles of DAIRYMAN farm data (2010) Indicator Economics (total) Income per 100 kg/milk (€) Income per family worker (€) Farm income / family labour unit (€) Dependency on subsidies Exposure to price fluctuations Ecology (total) N Balance kg/ha N Balance kg /1000 kg milk N Efficiency % P Balance kg/ha P Balance kg/1000 kg milk P Efficiency % Agro-environmental Payments (€) Greenhouse gas emissions Social aspects (total) Working conditions of farmers Education Social role/image Continuity of farm Score = 100% 16 34 32 Minimum 10%quantile 90%quantile Maximum -7.6 -69427 -69427 2.6 13323 18081 23.7 117467 109314 34.9 202916 188543 10 18 = 100% 15 11 13 11 8 10 10 22 = 100% 42 22 20 16 -3.3 0.4 0.2 0.5 1.4 1.0 7.2 1.5 17.1 3.9 11.8 -16.3 -4.6 19. 5 0.0 703.8 82.4 9.1 19.4 -4.6 -0.6 35. 9 0.0 932.3 268.0 34.3 47.5 17.9 3.0 157.9 122.6 1427.7 373.3 60.9 64.4 43.9 8.5 411.6 318.0 1816.9 These data were not worked out with quantiles. Scoring was made according to a multiple choice questionnaire (Elsaesser et al., 2013) Figure 1. Interregional comparisons of summarized scores of selected pilot farms Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 639 It was decided to take the quantile 10 and quantile 90 values of the complete 127-pilot-farm dataset as reference values for maximum and minimum scores. Pilot farms that are within the best 10% are awarded full marks for the particular indicator and farms within the worst 10% receive no points for the respective indicator. Points between these quantiles are calculated by linear regression. Multiplication of the measured values and the degree of target fulfillment give the score for each factor (Figure 1). Conclusion The DSI can be an integrative method in assessing developments of dairy farms. However, the aggregation of single factors has to be used carefully, because it is depending of a subjective scoring (Von Wirén-Lehr, 2001). It is no doubt, that the availability of a multi-annual data-set gives better results and a more solid analysis with lower sensitivity. The DSI offers a better insight on farm structures than comparisons of individual factors. Furthermore the DSI allows comparisons in dairy farm developments. Acknowledgements The authors would like to thank the European Regional Development Fund and the other financial partners of the project for their support as well as the pilot dairy farmers who were engaged in the collecting of data and lively discussions. References Aarts F. (2012) DAIRYMAN for a more efficient use of resources by dairy farms. Grassland Science in Europe 17, 753-755. Aarts F., Grignard A., Boonen J., De Haan M., Hennaert S., Oenema J., Lorinquer E., Foray S., Herrmann K., Elsaesser M., Castellan E.M. and Kohnen H. (2013) A practical manual to assess and improve farm performances. DAIRYMAN report, www.interregDAIRYMAN.eu Bélanger V., Vanasse A., Parent D., Allard G. and Pellerin D. (2012) Development of agri-environmental indicators to assess dairy farm sustainability in Quebec, Eastern Canada. Ecological Indicators 23, 421-430. Breitschuh G. and Eckert H. (2006) Kriteriensystem zur Analyse und Bewertung der Nachhaltigkeit landwirtschaftlicher Betriebe. KTBL-Workshop am 04. Mai 2006 in Osnabrück. Elsaesser M., Herrmann K. and Jilg T. (2013) The DAIRYMAN-Sustainability-Index (DSI) as a possible tool for the evaluation of sustainability of dairy farms in Northwest-Europe. Dairyman-Report, 3, www.interregdairyman.eu Girardin P. (2001) Französische Verfahren zur Bewertung von Umweltwirkungen landwirtschaftlicher Betriebe. ITADA - Forum 'Nachhaltige Landwirtschaft', Sissach (CH), 31 - 38, IfuL Müllheim. Grignard A., Bailey J., Boonen J., Stilmant D. and Hennaert S. (2013) Assessing the potential for improving the economic and environmental sustainability of dairy farming in North West Europe via farm diversification. Grassland Science in Europe 18, 96-98. HAFL (2012) Berner Fachhochschule, Hochschule für Agrar-, Forst- und Lebensmittelwissenschaften, RISE 2.0, Maßnahmenorientierte Nachhaltigkeitsanalyse der Agrarproduktion auf Betriebsebene. www.hafl.bfh.ch /fileabdmin/docs/Forschung/KompetenzenTeams/Nachhaltigkeitsevaluation/Rise/Was_ist_RISE.pdf. Larochelle D.L., Parent D.P., Allard G.A. and Pellerin D.P. (2007) Dairy farm sustainability: The economic component indicators. Journal of Animal Science 85, Suppl. 1, 330-331. Meul M., Van Passel S., Nevens F., Dessein J., Rogge E., Mulier A., Van Hauwermeiren A. (2008) MOTIFS: a monitoring tool for integrated farm sustainability. Agronomy for Sustainable Development 28, 321-332. Von Wiren-Lehr S. (2001) Sustainability in agriculture - an evaluation of principal goal-oriented concepts to close the gap between theory and practice. Agriculture, Ecosystems and Environment 84, 115-129. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 640 Dairy system sustainability in relation to access to grazing: a case study Decruyenaere V.1, Herremans S.1,2, Visser M.2, Grignard A.3, Jamar D.3, Hennart S.3, Campion M. 3 and Stilmant D. 3 1 Production and Sectors Department, Walloon Agricultural Research Centre – CRA-W, Rue de Liroux 8, B-5030 Gembloux, Belgium 2 Landscape Ecology and Plant Production Systems Unit, CP264/2 Université Libre de Bruxelles, 50 Avenue Franklin Roosevelt, 1050 Brussels, Belgium 3 Agriculture and Natural Environment Department, Walloon Agricultural Research Centre – CRA-W, Rue du Serpont 100, B-6800 Libramont, Belgium. Corresponding author: d.stilmant@cra.wallonie.be Abstract Extension of zero-grazing management scheme questions sustainability of dairy production. The aim of this work was to compare, during two seasons, the technico-economic and environmental performances of two experimental dairy herds with similar genetic potential. The first herd had full access to grazed grasslands during summer time while the second remained in the cowshed. Management scheme did not significantly affect dairy production, neither in quantity nor in quality. It only affected the distribution of the production across the year. From May till July, grazing cows produced, on average, significantly more milk than zerograzing ones (22.7 vs. 19.7 kg d-1 cow-1; P = 0.02) while the opposite was true from November till January (17.8 vs. 20.4 kg d-1 cow-1; P = 0.03). From the economic point of view these results underlined a huge increase in total production costs (feeding, bedding and reproduction costs), with an average increase of 30%, across both years, when shifting from a grazing to a zerograzing management scheme (22.8 ± 4.0 and 29.7 ± 4.5 € per 100 kg of standardized milk). From the environmental point of view, the zero-grazing system had better nitrogen balances with, on average and across both years, 112 kg N compared with 131 kg N ha-1. In term of climate footprint, the grazing system led to lower emissions of greenhouse gases than the zerograzing system due to the highest dependency of the zero-grazing system on fertilizer and feedstuff inputs. This result is of value whatever the functional unit mobilized, with 8550 ± 595 and 10700 ± 208 kgCO2eq ha-1 and 1140 ± 144 and 1350 ± 76 kgCO2eq ton of milk-1, for grazing and zero-grazing systems respectively. Based on these results, grazing-based systems appeared to be more sustainable than zero-grazing systems. Keywords: zero-grazing, GHG, N balance, dairy performance, DAIRYMAN Introduction As grasslands cover 50% of agricultural area in Wallonia, they play a major role in dairy production. Nevertheless, as in many European countries, zero-grazing systems are becoming more numerous. From a farmer's point of view, such management scheme allow improved control of diets, optimize grassland utilization, and achieve a higher milk production, higher labour efficiency, and lower nutrient losses (Meul et al., 2012). Grassland scarcity and/or accessibility, proportionally to herd size, or the use of a milking robot can also support the adoption of a zero-grazing management scheme (Arsenault et al., 2009). Nevertheless, such a scheme can increase dependency on external feedstuffs and feeding costs. The aim of this work was to compare, during two years, the technico-economic and environmental performances of two experimental dairy herds with similar genetic potential under contrasting management schemes: the first herd had full access to grazed grasslands during summer time while the second remained in the cowshed. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 641 Materials and methods Cows were allocated to the two herds in order to obtain balanced groups with similar parity distribution. Calving occurred all year long. Breeding phase, till first calving, was common to both herds. Heifers grazed as soon as the weather was favourable, as did dry cows of both herds. After calving, the heifers were introduced into their mother herd. Grazing management was based on a rotational scheme with residence time of 3.6 days, on average, and 37.7 days between two grazing periods. On average, 60 kg of mineral N were applied per ha and per year, to complement the slurry spreading. This fertilization scheme allows the development of a significant proportion of clover in the sward. System characteristics are presented in Table 1. Table 1. General characteristics of the two dairy systems compared Parameters Unit Zero grazing Grazing maximization 2010-11 2011-12 2010-11 2011-12 cows 25.5 27.5 22.5 22.5 kg - 642.6 - 641.8 Agricultural area for the herd ha 21.1 22.7 18.4 19.4 Grassland area ha 12.9 10.5 13.1 15.5 Grazed grasslands (GG) ha 4.8 4.1 8.0 8.7 Maize (other cereals) ha 7.7 (0.5) 7.9 (4.3) 3.3 (2.0) 2.9 (1.1.) Cereals ha 0.5 4.3 2.0 1.1 3.8 3.5 3.7 3.6 % 35 42 17 17 % 25 31 16 19 % 34 21 63 61 Average herd size Cow’s average weight Stocking rate in GG LU ha Maize silage in the dieta Concentrate in the diet a a Grass in the diet -1 a Annual average, on a DM basis. Concentrates are allocated on an individual basis, linked to the production potential of the cow at the rate of 1 kg of concentrate per 2.5 kg of milk over the 22 kg of daily production. Intake level, under grazing, was estimated through faeces analysis by NIRS (Decruyenaere et al. 2012). Animal performances were reflected by their annual production and the monthly profile of production of each herd. A simplified economic balance was calculated for both herds, excluding either subsidies and the following costs: water, electricity, taxes, insurance, administrative costs, rental costs, interest and depreciation. Finally, environmental pressure was quantified through N, P and K balances and quantification of emissions of Greenhouse Gases based on IPCC methodology (Tier 2) as defined by the DAIRYMAN project (Aarts et al., 2013). While animal performance characteristics led to statistical comparisons, the economic and environmental parameters, applied at the herd scale, allowed only descriptive approaches. Results and discussion Even though zero-grazing cows had, on average, a higher level of production (7868 (+ 8 %) vs. 7286 kg per cow) than grazing cows, the difference was not significant (P = 0.29). In the same way, the management scheme did not affect the monthly milk production per cow (P = 0.41) which turned out to be affected primarily by the season (P = 0.04) and this in a different way in each system (P season×system = 0.01) independently of the year (P = 0.60). Indeed, from May till July, grazing cows produced, on average, significantly more milk than zero-grazing cows (22.7 vs. 19.7 kg d-1 cow-1; P = 0.02) whereas the opposite was true from November till January Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 642 (17.8 vs. 20.4 kg d-1 cow-1; P = 0.03). The protein (3.4 g kg-1 of milk) and the fat (4.1 g kg-1 of milk) contents were not affected either by the management system (P= 0.14 and P = 0.28, respectively) or by the interaction between the management system and the season (P = 0.16 and P = 0.97, respectively). These results support earlier findings on the stability of production in zero-grazing systems, as grazing animals have to face resource variability, both in quantity and quality, across the season. Nevertheless, for animals with similar genetic potential, no significant production difference was highlighted on a yearly basis. From the economic point of view, the results underlined a huge increase in total production costs (feeding, bedding and reproduction costs), with an average increase of 30% across both years, when shifting from a grazing system to a zero-grazing management scheme (22.8 ± 4.0 and 29.7 ± 4.5 € per 100 kg of standardized milk). Due to its lower autonomy, the zero-grazing system also had higher N (+38 %), P (+ 33 %) and K (+ 35 %) inputs, which were compensated by higher exports, expressed per ha. This led to better mineral balances in zero-grazing systems with, on average across both years, 112 against 131 kg N ha-1; 6 against 19 kg P ha-1 and 98 against 104 kg K ha-1. Meul et al. (2012) underlined lower level of N surpluses in grazing than in zero-grazing commercial farms in Flanders, but our average values are below their lowest benchmark value of 136 kg N ha-1. In terms of climate footprint, the grazing system led to lower emissions of greenhouse gases than the zero-grazing system due to the highest dependency of the zero-grazing system on inputs of fertilizer and feedstuffs. This result is of value whatever the functional unit mobilized, with 8550 ± 595 and 10700 ± 208 kgCO2eq ha-1 and 1140 ± 144 and 1350 ± 76 kgCO2eq ton of milk-1, for grazing and zero-grazing systems respectively. Conclusions Based on these results, grazing-based systems appeared to be more sustainable than zerograzing systems and would allow for an improved buffering capacity against price volatility. Nevertheless, they require the availability of easily accessible grassland paddocks to support the herd size increase, as the production per hectare is higher in zero-grazing than in grazing systems (with 11600 vs. 9900 kg of milk ha-1). A good knowledge of herbage management is also necessary, in order to optimize grassland productivity. Acknowledgements The authors thank the European Fund for Regional Development and the Walloon area for funding of this research in the context of DAIRYMAN (INTERREG IV-B project). References Aarts F., Grignard A., Boonen J., de Haan M., Hennart S., Oenema J., Lorinquer E., Sylvain F., Herrmann K., Elsaesser M., Castellan E. and Kohnen H. (2013) A practical manual to assess and improve farm performances. 102 p. Arsenault N., Tyedmers P. and Fredeen A. (2009) Comparing the environmental impacts of pasture-based and confinement-based dairy systems in Nova Scotia (Canada) using life cycle assessment. International Journal of Agricultural Sustainability 7, 19–41. Decruyenaere V., Froidmont E., Bartiaux-Thill N., Bulgen A. and Stilmant D. (2012) Faecal near-infrared reflectance spectroscopy (NIRS) compared with other techniques for estimating the in vivo digestibility and dry matter intake of lactating grazing dairy cows. Animal Feed Science and Technology 173, 220–234. Meul M., Van Passel S., Fremaut D. and Haesaert G. (2012) Higher sustainability performance of intensive grazing versus zero-grazing dairy systems. Agronomy for Sustainable Development 32, 629-638. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 643 Beef productivity on the North Wyke Farm Platform in two baseline years Thompson J.B.1,2, Orr R.J.1, Dungait J.1, Murray P.J.1 and Lee M.R.F.1,2 1 Rothamsted Research North Wyke, Devon, EX20 2SB, United Kingdom 2 University of Bristol, School of Veterinary Sciences, Langford, BS40 5DU, United Kingdom Corresponding author: jt13651@bristol.ac.uk Abstract The North Wyke Farm Platform (NWPF) has been implemented to deepen the understanding of sustainable grassland management systems, including the impacts on livestock productivity. Two baseline years of data were collected (2011 and 2012) to establish the level of productivity of the existing permanent pasture on three hydrologically isolated farmlets. Cattle were put on the platform at weaning and remained until finishing. They were housed in winter and grazed on a continuous-stocking basis during the grazing period. Records of inputs, such as nitrogen fertilizer and farmyard manure, were taken along with regular monitoring of liveweight gain. Overall average daily gain (ADG) from weaning to selection for sale was 0.7 kg/day (± 0.01 SEM) in 2011 and 2012, and there were no significant differences between farmlets in either year. Silage DM yield was lower in 2012 due to heavy rainfall and low levels of solar radiation. The equal performance of cattle on the three farmlets, under existing permanent pasture, will allow future differences, when the new management systems are operational, to be accounted for by the effects of the new systems. Keywords: North Wyke farm platform, beef production, sustainable grassland, permanent pasture Introduction The Foresight Report (2011) highlighted the need for sustainable intensification of livestock production. However, there is a lack of understanding of the impact of grassland systems in terms of pollution and potential benefits, e.g. carbon sequestration. The North Wyke Farm Platform (NWFP) is a national capability developed to improve our understanding of intensive, pastoral ruminant livestock production systems with respect to their impact on the environment and value in contributing to global food demand and providing healthy nutritious human food (Murray et al., 2013). The NWFP provides a farm-scale facility with extensive instrumentation to collect core parameter data to assess the impacts and benefits of three different pastoral systems, each providing alternative approaches to potentially productive grassland farming. The first is a continuation of the existing permanent pasture improved with N-fertilizer application, the second is the introduction of N-fixing legumes and subsequent greatly reduced application of artificial N-fertilizer; the third is the introduction of innovative new species such as deep rooting Festulolium through planned reseeding. The effects of these systems on water, air and soil will be monitored. The productivity of the cattle and sheep on the platform is a vital consideration in terms of the inputs and outputs of the grassland system. Before introduction of the new management systems from spring 2013, baseline data were collected for two years to establish the level of beef production on the existing pasture on the NWFP, which will be critical to establish causative effects of the new imposed systems in future years. Materials and methods The farm platform comprises three farmlets, each approximately 21 ha. The platform has been instrumented to collect various core parameter data including analysis of water samples (collected via the French-drainage system installed around the perimeter of each field), soil conditions and atmospheric measurements. Detailed records of all inputs and outputs to the farm platform were maintained. N fertilizer was applied according to DEFRA RB209 (Defra, 2013) and NVZ guidelines, when conditions were suitable. Farmyard manure (FYM) was Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 644 applied following cutting. Cattle were grazed on a continuous stocking basis, maintaining sward surface height at approx. 6-9 cm (Hodgson et al., 1986). Animal performance was measured by regular weighing. Two years (2011 and 2012) of baseline data for beef performance were analysed to establish the level of production under the current management of permanent pasture. The existing pasture has been established across the platform for at least 10 years. Swards were surveyed for plant species abundance in summer 2013 and found to contain percentage covers (Rodwell, 1992) of: 64 Lolium perenne, 38 Agrostis stolonifera, 2 Holcus lanatus and 1 Alopecurus geniculatus as the main constituents. Results The farm platform carried 75 (28 heifers and 47 steers) and 81 cattle (38 heifers and 43 steers), 25 and 27 per farmlet in 2011 and 2012, respectively. There were also 15 younger calves in 2012; however, final data for these were not available at the time of writing. The cattle were weaned, housed and weighed in October. The average wean weight and age were 300 kg (± 5.44 SEM) and 180 d (± 4.00 SEM) in 2011 and 319 kg (± 5.31 SEM) and 212 d (± 7.8 SEM) in 2012. There were various breeds including: Aberdeen Angus, Simmental, Limousin, Hereford and Charollais crosses. The cattle were turned out on 13 April in both 2011 and 2012 with mean weight 391 kg (± 7.06 SEM) and 395 kg (± 5.43 SEM), respectively. Stocking rates on the whole NWFP were 1.3 and 1.7 cattle /ha but 5.8 and 5.9 cattle/ha on grazed areas, at time of grazing for 2011 and 2012, respectively. In 2011 the grazing period was 204 days; 30 cattle that were not finished at grass were housed on 3 November 2011 (mean weight 558 kg ±8.55 SEM). Table 1: Mean liveweight gain of beef cattle from weaning to finishing on the NWFP in two baseline years (ADG = average daily gain; FP = farm platform). 2011 Farmlet 1 2 3 P value FP Average ADG (kg) (1st winter, housed) 0.54 0.50 0.54 0.658 0.53 Farm SEM 0.039 0.033 0.032 Platform0.017 ADG (kg) (grazing period) 1.05 0.99 1.01 0.459 1.02 SEM 0.039 0.042 0.018 0.022 ADG (kg)(2nd winter, housed) 0.70 0.81 0.78 0.760 0.76 SEM 0.121 0.12 0.056 0.063 Overall ADG (kg) 0.72 0.70 0.73 0.394 0.72 SEM 0.017 0.017 0.018 0.010 2012 Farmlet 1 2 3 P value FP Average ADG (kg) (1st winter, housed) 0.46 0.39 0.46 0.283 0.44 Farm SEM 0.041 0.024 0.033 Platform0.019 ADG (kg) (grazing period) 0.93 0.83 0.91 0.155 0.89 SEM 0.032 0.027 0.032 0.018 ADG (kg) (2nd winter, housed) 1.35 1.33 1.21 0.712 1.30 SEM 0.162 0.106 0.061 0.068 Overall ADG (kg) 0.74 0.68 0.74 0.089 0.72 SEM 0.019 0.02 0.023 0.013 Farmlets: 1 = Planned reseeding, 2 = Improved permanent pasture, 3 = Increased legume utilization. The 2012 grazing season was 166 days; 48 cattle that were not finished at grass were housed on 3 October 2012 (average weight of 537 kg ± 7.89 SEM). Mean age at selection for sale was 591 d (± 9.0 SEM) and 617 d (± 10.39 SEM); mean number of days between weaning and selection was 401 d (± 6.6 SEM) and 379 d (± 6.15 SEM) for 2011 and 2012, respectively. There were no significant differences between farmlets for live weights, age or days to selection in either year. ADG for both years is shown in Table 1, with no significant differences between farmlets observed. Mean age at selection was significantly higher in 2012 than 2011 (P=0.016), number of days from weaning to selection approached significance and was higher in 2011 (P=0.055). Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 645 The total annual N fertilizer applied was 201 and 83 kg/fenced ha, with 11 and 8 kg/fenced ha of FYM N applied following cutting, for 2011 and 2012, respectively. In 2011 there was 63 ha cut for silage (1st & 2nd cut); average yield was 5.9 t DM/ha, a total of 369 t DM harvested with mean DM of 32.6%. In 2012 there were 39 ha cut for silage (1st & 2nd cut); average yield was 4.1 t DM/ha, a total of 157 t DM harvested with mean DM of 23.6 %. Discussion After April 2013 the three farmlets began to have different grassland management approaches as the new treatments are progressively rolled out. These base-year data demonstrate that there is no significant difference in performance of stock on each area under established permanent pasture, suggesting nutritional quality and herbage provision is equal across the platform and providing evidence that any future differences in performance observed will be the result of the different management systems. According to best-practice guidelines (EBLEX, 2008, 2005; http://www.eblex.org.uk/publications/), the target ADG at grass should be at least 0.8 kg/d. This was achieved on the farm platform in both years showing that the level of production is good in comparison with industry targets. The significant difference seen between years for age at selection for sale can be accounted for by the cattle being 1 month older at weaning in 2012 than 2011. Silage DM yield per hectare was lower in 2012, which was a result of reduced DM content of silage due to the significantly higher rainfall and low levels of sunshine for that year. N fertilizer application was less in 2012 in response to initial good spring growth and then poor weather conditions. Mean daily rainfall between 1 April and 30 October was 1.97 mm (± 0.250 SEM) and 3.42 mm (± 0.426 SEM) in 2011 and 2012, respectively (P>0.003). Stocking rates were higher in 2012; therefore the area conserved was an emergent property, which was reduced because it was determined by requirements for grazing stock. References DEFRA (2013) Fertiliser Manual (RB209) Version 8. https://www.gov.uk/government/publications/fertilisermanual-rb209 Foresight (2011) The Future of Food and Farming. Final Project Report. The Government Office for Science, London Hodgson J., Mackie C.K. and Parker J.W.G. (1986) Sward surface height for efficient grazing. Grass Farmer 24 p5-10 Murray P.J., Griffith B.A., Orr R. J., Shepherd A., Blackwell M.S.A., Hawkins J.M.B. and Peukert S. (2013) The North Wyke Farm Platform: A new UK national capability for research into sustainability of agricultural temperate grassland management. Paper presented at International Grasslands Congress, Sydney Australia, September 2013 Rodwell J.S. (1992) Grasslands and montane communities. British Plant Communities, Vol. 3. cambridge UK: Cambridge University Press. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 646 Milk production of sheep fed on preserved forage in winter and grazing in spring Stoycheva I., Kirilov A. and Simeonov M. Institute of Forage Crops, 5800 Pleven, Bulgaria Corresponding author: kirilovatanas@hotmail.com Abstract In Bulgaria there are 39 breeds of sheep reared primarily for milk. The purpose of this experiment was to test the effect of feeding rations based on maize silage, alfalfa hay and silage in the winter, and grazing on temporary and natural pasture in the spring, on the quantity and composition of sheep milk. Thirty sheep from the Pleven Blackface breed were used, divided into three groups. The experimental period of 84 days was divided into two sub-periods of 42 days. During the first sub-period the first group received maize silage, the second group alfalfa hay, and the third group alfalfa silage. In all rations, compound feed to cover the needs for 1.5 L milk was included. In the second sub-period, the first group grazed natural pasture, the third group grazed on temporary pasture of cocksfoot and sainfoin, and the second group remained on alfalfa hay. The sheep received 0.5 kg of grain maize during the grazing period. The highest daily milk yield was obtained by a feeding ration based on alfalfa hay, which was 14-26% higher compared to that from rations based on maize silage and alfalfa silage. The milk of sheep grazing on temporary pasture was 28.5% more than that of ewes grazing on natural pasture; this was probably due to the higher content of legumes in the temporary pasture. During the lactation period the quantity of milk decreased the least when grazing on temporary pasture. Keywords: sheep, feeding, milk, composition, pastures Introduction The number of sheep in Bulgaria until 1990 varied between 7 and 11 million, but in the period of democratic changes their number have decreased over five times. The development of sheep breeding is due to the presence of huge natural resources of meadows and pastures, which represent 28% of the country’s agricultural area. In Bulgaria, 39 breeds of sheep are reared and half of them are local breeds for milk. Sheep milk is processed primarily into cheese, which sells well on the local market and abroad. The production of more sheep milk and the economic prosperity of the farm are related not only to the improvement of the genetic potential of the sheep, but also to the improvement of the nutrition and the farming systems. The purpose of the experiment was to test the effect of feeding rations based on maize silage, alfalfa hay and alfalfa silage in the winter, and grazing on temporary and natural pasture in the spring, on the quantity and composition of milk from sheep. Materials and methods The experiment was carried out in 2013 at the Institute of Forage Crops, Pleven. For this purpose, 30 sheep from the Pleven Blackface breed were used, divided into 3 groups. The lambs were weaned 25 days after birth, and the sheep entered the experiment 35 days after lambing. The experimental period lasted 84 days and was divided into two equal sub-periods of 42 days (6 weeks). In the first sub-period, the first group received ad libitum maize silage, the second group received alfalfa hay, and the third group received alfalfa silage. The sheep in the three groups received compound feed of sunflower meal, rapeseed meal, maize grain, triticale and vitamin-mineral supplement to cover the needs of daily milk yield of 1.5 L, according to Todorov and Dardjonov (1995). In the second sub-period, the first group was moved from maize silage to grazing on natural pasture, and the third group went from alfalfa silage to grazing on temporary pasture. During grazing, both groups received 0.5 kg of grain maize per Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 647 sheep. The temporary pasture was of cocksfoot (Dactylis glomerata) and sainfoin (Onobrychis viciifolia). During the experimental period, the daily milk yield from each group was controlled and on two consecutive days of the week the individual milk yield per ewe was also controlled. The milk composition was determined by Milko Scan, model 133. Samples were taken of the natural and temporary pastures to determine the botanical and chemical composition. The dry matter (DM), crude protein (CP), crude fibre (CF), fat, ash and nitrogen-free extract substances (NFE) in the forages were determined. Results and discussion In the first sub-period of the experiment, the highest average daily milk yield was obtained from the second group of sheep, fed on a ration based on alfalfa hay (1.322 L; P> 0.05) (Table 3). When fed on ration based on alfalfa hay, the sheep gave 14-26% more milk compared to the group fed rations based on maize silage and alfalfa silage (P> 0.05). Between the first and the third groups, no significant differences in quantity of the milk were observed. In the second (grazing) sub-period the highest average daily milk yield (1.001 L) was observed in the group grazing on temporary pasture (P> 0.05). The sheep grazing on natural pasture and those which continued to feed on alfalfa hay, had lower milk yields than the group grazing on temporary pasture. The total milk in the second sub-period for the group grazing on temporary pasture was 28.5% and 25.8% higher than that of the sheep from the other two groups, i.e., grazing on natural pasture and alfalfa hay (P> 0.05) respectively. During the second sub-period compared with the first sub-period, the amount of milk in the third group (grazing on temporary pasture) decreased by 4.5% only, while the first group grazing on natural pasture and the second group fed on alfalfa hay the drop was 31.4% and 40%, respectively. The first sub-period of feeding and milking of the sheep coincides with the period after lambing, when the sheep give the most milk. Similar average daily milk yields in ewes of the same breed and in the same period of lactation were obtained by Kirilov et al. (1998) and after feeding on silage of alfalfa and peas (Kirilov and Simeonov, 2012). After the first two months of lambing the sheep milk yield decreases. This decrease is not significant in the group grazing on temporary pasture, which has high proportion of the legume component and a high content of crude protein (Tables 1 and 2). Table 1. Chemical composition of rations and forages, g kg-1 DM CP СF Fat Ash NFE Ration 1 gr. 148.98 178.37 21.33 53.02 598.30 Ration 2 gr. 140.85 216.10 18.50 57.10 567.45 Ration 3 gr. 146.82 218.53 24.55 63.47 546.63 Forages DM Temporary pasture 1-3 week 177.20±0.32 256.70±1.67 155.16±3.38 48.00±0.153 106.00±1.23 434.14±3.11 3-6 week 210.20±2.07 179.50±1.43 244.70±0.53 35.10±0.22 113.10±0.46 427.60±1.47 1-3 week 187.9±0.99 202.20±1.04 223.00±0.18 42.90±0.18 124.40±0.65 407.50±0.13 3-6 week 266.00±5.64 176.50±0.22 308.80±0.15 31.10±0.26 91.80±1.96 391.80±0.68 Natural pasture Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 648 Table 2.Yield and botanical composition (grass:legume:other) temporary (TP) and natural (NP) pastures 1 week 2 week 3 week 4 week 5 week 6 week 1497.3±108. 4 1637.2±77.7 3594.1±68.7 6480.3±69. 5 8503.5±157. 6 10012.3±187. 8 gr:leg:othe r 32:68:0 41:58:1 42:52:6 45:50:5 50:46:4 54:40:6 NP, kg ha-1 967.6±148.9 1000.1±107. 1 1177.0±134. 0 1315.6±81. 0 1923.0±109. 4 2125.6±170.2 gr:leg:othe r 95:0:5 92:4:4 87:7:6 83:10:7 80:11:9 79:12:9 TP, kg ha -1 As regards the composition of milk, no significant differences were detected between the indicators of fat and proteins between the groups during the first sub-period. In the second subperiod, an increased concentration of fat and protein in the milk is observed, compared to the first. This is likely due to the lower daily milk yield (Fuertes et al., 1998). The content of CP in temporary pasture at the beginning of grazing, 1-3 weeks, was high at an average of 256.7 g kg-1 DM, and decreased over the next three weeks to 179.5 g kg-1 DM, and the contents of CF increased from 155.2 g kg-1 DM and 244.7 g kg-1 DM (Table 1) (P> 0.05). During the same period, the contents of the CP in the natural pasture during the first 3 weeks, were lower than in the temporary pasture (P <0.05) and the content of CF was higher (P <0.05). A similar trend for low CP and high on CF in natural pasture was observed in weeks 3-6 compared to weeks 1-3 of the grazing period. Higher content of CP and low content of CF at the beginning of the grazing period (weeks 1-3) are regular; subsequently, with the advancing of the vegetation period, there is a decrease in the CP and increase of the CF. The high levels of CP in the temporary pasture compared to natural pasture is probably due to the higher proportion of the legume component, 40-68% compared with 0-12% in the natural pasture (Table 2). This is probably the reason for higher milk yield in the group grazing on temporary pasture, compared to the group grazing on natural pasture. Table 3. Milk production and composition. Different superscript letters in columns indicate differences are significant at P> 0.05 Group Daily milk yield, l Total milk per sheep, l Fat, % Protein, % Lactose, % Total Solids, % First sub-period 1 maize silage 1.155c±0.048 48.49b±0.22 5.15a±0.63 5.63ab±0.39 5.31c±0.25 16.50a±1.08 2 alfalfa hay 1.322d±0.040 55.55c±0.34 5.31a±0.98 5.83b±0.38 5.50d±0.26 17.12a±1.17 3 alfalfa silage 1.049bc±0.027 44.06b±0.19 5.22a±0.79 5.63ab±0.50 5.50d±0.17 16.83a±1.20 Second sub-period 1 nat. pasture 0.779a±0.058 32.71a±0.43 8.45c±0.91 5.42a±0.32 4.88a±0.14 19.35d±0.96 2 alfalfa hay 0.795a±0.068 33.41a±0.48 6.77b±0.60 6,15c±0.34 5.11b±0.41 18.63c±0.60 3 temp. pasture 1.001b±0.042 42.05b±0.29 6.27b±0.57 5.73b±0.40 5.31c±0.19 17.92b±0.85 Conclusions The sheep fed a ration based on alfalfa hay gave 14-26% more milk, compared to the milk yield of sheep fed on maize silage and alfalfa silage. Sheep grazing on temporary pasture with higher content of legumes and crude protein gave 28.5% more milk compared to those grazing on Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 649 natural pasture. During the course of the lactation period, the quantity of milk decreased the least when sheep were grazing on temporary pasture. References Fuertes, J. A., Gonzalo C., Carriedo J. A. and San Primitivo F. (1998) Parameters of test day milk yield and milk components for dairy ewes. Journal Dairy Science 8, 1300-1307. Kirilov A. and Simeonov M. (2012) Influence of soybean oil meal and brewer’s grains on the milk production of sheep. I. Ration composition and milk productivity. Journal of Mountain Agriculture in the Balkans15, 599-608. Kirilov, A., Zhelyazkov, T., Krachunov I., Carlier, L., Iliev T. Getov G. (1998) Effects of lucerne and peas (green mass and silage) on milk production of sheep. Zhivotnovadni nauki 4/1998, Vol.25, pp. 4-9 (Bulgarian with English summary). Todorov N. and Dardjonov T. (1995) Nutrient requirements of sheep and goats. Publishing House NIS-UZVM, Stara Zagora, p. 216. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 650 Evaluation of a home-grown crimped lupin and barley concentrate feed for finishing lambs Marley C.L., Jones H., Theobald, V., Sanderson R. and Fychan R. Institute of Biological, Environmental and Rural Sciences (IBERS), Aberystwyth University, Gogerddan, Aberystwyth, SY23 3EB, United Kingdom. Corresponding author: christina.marley@aber.ac.uk Abstract To investigate whether a home-grown lupin-barley concentrate feed could support similar levels of lamb productivity and carcass quality to a commercial lamb-finishing diet, two isonitrogenous concentrates were offered to castrated male Texel-cross lambs. Diets comprised barley straw plus either: home-grown concentrate, viz. crimped narrow-leafed lupin (Lupinus angustifolious cv. Boruta), crimped barley (Hordeum vulgare cv. Propino) and sheep mineral (705, 270 and 25 g kg-1 dry matter respectively); or, a pelleted, commercial lamb finisher (Control). Straw and each treatment-diet were offered ad libitum over 56 d to 4 replicate pens of 5 lambs, totalling 20 lambs per treatment. Group intakes were recorded over days 0-28 and lambs were weighed and condition scored at 7-d intervals throughout the experiment. From day 29 onwards, individual lambs were selected out for slaughter and carcass characteristics were recorded. Results showed no differences in the productivity, concentrate conversion efficiency, days to finish or carcass characteristics of lambs offered either the crimped lupin-barley concentrate or control diet. Overall, the findings of this study show that home-grown crimped lupin-barley concentrate diets can be used to finish lambs without any detrimental effects on productivity or carcass characteristics when compared to a commercial lamb finishing diet. Keywords: Lupin, Lupinus angustifolius, crimped grain, lamb, home-grown Introduction Livestock farmers worldwide are aiming to reduce reliance on imported and bought-in feedstuffs, which may be subject to world market price fluctuations and have a high environmental footprint. Oilseed rape and palm kernel cake and meal are typically used in commercial concentrate diets for sheep in the UK. Lupins (Lupinus; Leguminosae), although not traditionally grown as a field crop in the UK, yield grain with a potentially high nutritional value both in terms of protein content (compared with peas or beans) and oil content, which give them high energy value. Previous work has shown that both narrow-leaf and yellow lupins can be used within a pelleted concentrate for finishing lambs instead of either soya or compared to a control lamb finisher diet, without any adverse effects on lamb productivity or carcass killing-out percentages (Fychan et al., 2008). However, this approach requires the use of graindrying facilities which are not often available on livestock-based farms. An alternative approach to storing lupins is the use of crimping technology to preserve the grains and which overcomes the need for grain drying facilities and also allows for the crop to be harvested at an earlier date and stage of maturity. In crimping, grain is harvested moist (approx. 60 per cent dry matter content), passed through a crimper (to flatten and open the grains) and preserved using an additive before being ensiled. Here, we present the findings of a study to investigate the effects of incorporating crimped narrow-leafed lupins into the concentrate diets of finishing lambs, on lamb productivity and carcass characteristics when compared to a commercial lamb finisher diet. Material and methods Forty castrated male Texel-cross lambs were sourced from the same late-lambing flock for the experiment. The experiment involved a 14-d standardization period followed by a 14-d Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 651 adaptation period and 56-d measurement period (initial live weight 32 ± 2.8 kg). The two dietary treatments were a home-grown mixed ration comprising crimped narrow-leafed lupin, crimped barley and sheep minerals (705, 270 and 25 g kg-1 dry matter (DM) respectively) and control diet of commercial lamb finisher pellets. The home-grown ration was formulated to be isonitrogenous with the control concentrate. The lupin (Lupinus angustifolious cv. Boruta) and barley (Hordeum vulgare cv. Propino) for the experiment were established at the same site at Aberystwyth University on 16 April 2013 and their grain was harvested at a target DM content of 700 and 600 g kg-1, respectively. Moist barley grain was harvested on 8 August and moist lupin grain was harvested on 27 August 2013. Immediately after harvest, the grain was rolled through a Murska 350 S2 crimper (Aimo Kortteen Konepaja Oy, Ylivieska, Finland), and Crimpstore 2000S preservative (Kemira Oyj, Helsinki, Finland) was applied to the barley and lupins at the rate of 4 and 6 l t-1, respectively. Quantities of each feed were stored in sealed polythene bags within 500 kg tote bags to provide an amount that was sufficient for 7 days of feeding. During the standardization period, the lambs grazed a typical ryegrass-clover sward as a single group. At the start of the adaption period, lambs were allocated to treatment, and pen within treatment, on the basis of live weight and body condition score. The eight groups of five lambs were then housed in pens with sawdust bedding and dietary treatments and barley straw were introduced gradually. Fresh water was available at all times. Throughout the measurement period, barley straw and the treatment concentrate diets were offered ad libitum with a refusal margin of 0.10 to 0.15 d-1. Pen intakes were monitored over days 0-28 and lambs were weighed and condition scored at 7-d intervals throughout the experiment. Live-weight gain between days 0 and 28 was recorded using live weight recorded on two successive days at the start and end of this period. From day 29 onwards, lambs were selected-out for slaughter as they reached a target fat class of 3L and their live weight and carcass characteristics were recorded. Prior to slaughter, back fat and muscle depths were measured by ultrasound at the third lumbar vertebra. Killing-out ratio was calculated as the ratio of cold carcass and live weights. All data were analysed by ANOVA except days-to-finish-date, which were analysed by sign test, using Genstat® Version 14.2 (Payne et al., 2011). Pens were treated as the experimental unit in all analyses. Results and discussion The chemical analysis of the concentrate feeds are presented in Table 1. The data confirm visual observations that both crimped diets were well preserved, with similar dry matter (DM), nitrogen and ME values to previously reported data for lupins by Fraser et al. (2005). The results of the effects of each treatment diet on lamb productivity and carcass characteristics are presented in Table 2. The median time to finish was 29 d for the control diet and 32.5 d for the home-grown diet (P > 0.05). Results showed no differences in the productivity, concentrate conversion efficiency, days to finish or carcass quality characteristics of lambs offered either the crimped lupin-barley concentrate or the control diet. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 652 Table 1. Chemical analysis on concentrate feeds (g kg-1 DM unless otherwise stated) Dry matter (g kg-1) Crimped Barley Crimped Lupin Mean s.d. Mean s.d. 9.0 795 0.8 578 Control Mean 901 s.d. 16.3 Ash 19.9 0.10 35.2 0.60 87.9 9.72 Nitrogen 18.4 0.20 52.2 1.85 27.0 0.50 Oil 16.7 1.05 57.3 1.33 40.8 1.39 Crude fibre 72 8.4 215 10.5 158 6.0 879 13.8 950 1.5 765 4.3 NGCD ME (MJ kg-1 DM) 12.7 0.20 14.7 0.05 11.7 0.09 NCGD = Neutral cellulase gammanase digestibility; ME = Metabolizable energy Table 2. Lamb productivity and carcass characteristics in relation to dietary treatments Home-grown concentrate Concentrate intake# (kg DM d-1) 0.95 Live-weight gain# (LWG; g d-1) 110 # Concentrate conversion efficiency Control 1.31 159 s.e.d.$ 0.079 35.5 Prob 0.004 0.216 0.113 0.118 0.0337 0.885 3.34 3.42 0.079 0.309 (g LWG (g DM intake)-1) Condition score Weight empty (kg) 34.8 35.0 0.68 0.724 Cold carcass weight (kg) 17.0 17.5 0.59 0.390 0.0100 0.263 Killing-out ratio # 0.487 0.499 Muscle depth (mm) 25.86 26.95 0.530 0.086 Back-fat depth (mm) 4.7 4.5 0.31 0.438 = days 0 to 28; $ = 6 residual df for error Conclusions Overall, the findings of this study show that home-grown crimped lupin-barley concentrate diets can be used to finish lambs without any detrimental effects on productivity or carcass characteristics when compared to a commercial lamb finishing diet. Acknowledgements This work was funded through a Technology Strategy Board Project – LUKAA (Lupins in UK Agriculture and Aquaculture) a joint initiative co-funded by the Biotechnology and Biological Sciences Research Council and with industrial support from: Alltech, Alvan Blanch, Birchgrove Eggs, Ecomarine, Germinal Seeds, Kelvin Cave, NIAB TAG, PGRO, Soya UK, Wynnstay PLC and with the University of Plymouth. References Fychan A.R., Marley C.L., Lewis G.G., Davies D.R.W., Theobald V.J., Jones R. and Abberton M.T. (2008) Effects of feeding concentrate diets containing narrow-leaf lupin, yellow lupin or soya when compared with a control diet on the productivity of finishing lambs. Lupins for Health and Wealth. Proceedings of the 12th International Lupin Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 653 Conference, Fremantle, Western Australia, 14-18 September 2008 (Palta J.A. and Berger J. D. (eds.)) pp. 127130; International Lupin Association: Canterbury, NZ. Fraser M.D., Fychan R. and Jones R. (2005) Comparative yield and chemical composition of two varieties of narrow-leafed lupin (Lupinus angustifolius) when harvested as whole-crop, moist grain and dry grain. Animal Feed Science and Technology 120(1), 43-50. Payne R.W., Murray D.A., Harding S.A., Baird D.B. and Soutar D.M. (2011) Genstat® for WindowsTM (14th Edition). VSN International, Hemel Hempstead, United Kingdom, pp. 150. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 654 Optimal base temperature for computing growing degree-day sums to predict forage quality of mountain permanent meadow in South Tyrol Romano G.1, Schaumberger A.2, Piepho H.-P.3, Bodner A.1 and Peratoner G.1 1 Laimburg Research Centre for Agriculture and Forestry, 39040 Auer/Ora (BZ), Italy 2 Agricultural Research and Education Centre Raumberg-Gumpenstein, 8952 Irdning, Austria 3 University of Hohenheim, Bioinformatics Unit, Fruwirthstr. 23, 70599 Hohenheim, Germany Corresponding author: Giovanni.Peratoner@provinz.bz.it Abstract Knowledge of the appropriate base temperature (Tb) is needed for the analysis of meteorological effects on plant phenology using growing degree-days. However, several values of Tb have been used for mountain grassland. The main purpose of this research was to identify the optimal Tb for predicting changes in crude protein, acid detergent fibre and net energy for lactation of forage grown in mountain permanent meadows in South Tyrol (Italy). Forage samples were sequentially harvested weekly for a period of seven weeks starting from the pasture stage (15 cm growing height) at 202 environments (sites × years). To estimate the most appropriate Tb, growing degree day sum (GDDsum) were computed in the time interval between one week before pasture stage and the harvest time using a range of Tb between 0 °C and 5 °C at a 0.5 °C interval and using both observed temperatures from reference weather stations and interpolated temperature. Optimal Tb were estimated on the basis of the R² values of the regression equations of the different quality parameters on GDDsum and can be set at 0, 1.5 and 4.5 °C for the prediction of CP, NEL and ADF respectively. R² values exhibited in general little variation within the investigated range. Keywords: permanent meadows, mountain environment, forage quality, base temperature, growing degree-day sum Introduction Growing degree-days (GDD) are widely used in agriculture to describe the phenological development of plants. During the growth period, modification of the chemical composition has an impact on the nutritional value of the plants. Degree day sum have been used to describe variation in crude protein content and cell wall components at different phenological stages of alpine species in pastures (Bovolenta et al., 2008). Knowledge of the right base temperature (Tb) is crucial to compute reliable GDD. However, several values of Tb have been used by different authors for grassland. Parsons et al. (2006) predicted fibre content of legumes at low altitudes by using a Tb of 5 °C, while Bovolenta et al. (2008) used a Tb value of 0 °C for high altitude pastures in the Alps. More recently, in a study across permanent meadows at altitudes up to 1100 m in Austria, Schaumberger (2011) identified optimal Tb values of 3.0 °C and 2.0 °C for the beginning of flowering of Dactylis glomerata and the time of first cut, respectively. The main purpose of this research is to estimate the optimal Tb to be used for permanent meadows in South Tyrol (Italy) covering a wide altitude range (667 to 1593 m) and to be used for predicting the change of selected forage quality traits. Material and methods Forage samples were collected by weekly sequential sampling for a period of seven weeks starting from the pasture stage (15 cm growing height) at 175 environments (35 experimental sites × 5 years) from 2003 to 2007, at 20 environments (5 experimental sites × 4 years) from 2009 to 2012 and at 7 environments in 2013. Details concerning sampling design and forage analyses are described in Peratoner et al. (2010). For each experimental site, a reference weather station (RWS) was chosen among those of the measurement network of the Hydrographic Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 655 Office of the Province Bolzano/Bozen in terms of similar altitude, limited geographical distance and missing screen effect by hills. On a daily basis, the observed temperatures of the whole measurement network were spatially interpolated by georegression at a resolution of 100 m according to Schaumberger (2011), based on the correlation of observed mean monthly temperature and elevation combined with daily residuals. R² and intercept of the regression equations calculated by using interpolated temperatures at the experimental site on the observed temperatures of the RWS were used to assess the suitability of the RWS. A suitable RWS was found for only 20 of the 35 experimental sites. Growing degree days (GDD) were computed according to Schaumberger (2011). Growing degree-day sums (GDDsum) were calculated referring to the time interval between one week before the date at which pasture stage was attained and the harvest date, using a range of Tb between 0 °C and 5 °C at a 0.5 °C. GDDsum were computed both with the temperatures observed at the RWS and the interpolated temperatures at the experimental site. Data analysis was performed with a mixed model in SAS taking into account the serial correlation due to repeated measurements within each environment and the random effects experimental site, sampling area within the experimental site, year and the interaction year × experimental site. The relationships between the dependent variables (CP, NEL and ADF) and the GDDsum obtained with the different Tb were separately fitted by means of a second-degree polynomial. A first-order autoregressive covariance structure was used for modelling the serial correlation among repeated measures. CP data were square root-transformed to achieve normal distribution of the residuals and variance homogeneity. Nevertheless, visual assessment of residuals suggested mild heteroscedasticity, which will be explored in future work. A five-fold cross-validation (Hawkins et al. 2003) was performed for each model and the relationship between the resulting coefficient of determination and the Tb was fitted by a polynomial up to the third degree. Tb values corresponding to the stationary points of these equations (the maximum in case of the thirddegree polynomial) were regarded as optimal Tb values (Tbopt). Results and discussion Tbopt were found to vary mainly depending on the quality parameter investigated and to a lesser extent on the data set used for the analysis (Figure 1). 0.77 b) a) c) 0.75 R² 0.73 0.71 0.69 ADF 0.67 NEL CP Tbopt 0.65 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 Tb (°C) Tb (°C) Tb (°C) Figure 1. Polynomials fitting the determination coefficients of regression equations of different quality parameters on GDD sum depending on Tb and obtained based on three data sets: a) interpolated temperatures at 35 experimental sites, b) interpolated temperatures at the 20 experimental sites with an available RWS, c) observed temperatures of the 20 RWS. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 656 All in all, R² values ranged, across the quality parameters and the different data sets, in a quite narrow interval (between 0.664 and 0.757). Moreover, changes of R² of NEL, CP and ADF depending on Tb (from 0 °C to 5 °C) were very small in the investigated temperature range. Higher coefficients of determination were consistently achieved for CP and lower coefficients of determination were observed for ADF across the different data sets. Tb opt for CP was found to be located around the lowest end of the investigated temperature range, varying depending on the data set between -0.42 and 0.77 °C. The Tbopt for NEL was found close to it, varying between 1.2 and 1.7 °C, while Tbopt for ADF is placed at the other extreme of the temperature range, with estimated values between 3.4 and 6.2 °C (the latter is not shown in Figure 1c). For all quality parameters, the lowest Tbopt were found using GDDsum based on interpolated temperatures of all experimental sites. Differences of the estimated Tbopt might be due to imprecision in estimating temperatures through interpolation at certain sites. Conclusions Small changes in Tbopt for estimating changes in selected quality parameters depending on GDDsum were detected and can be specifically used for this task. For the prediction of CP, NEL and ADF, Tbopt of 0, 1.5 and 4.5 °C respectively seem to be adequate. On average, all Tbopt values lie within the range of Tb already known from the literature. Acknowledgements We wish to thank the Regional competitiveness and occupation EFRE 2007–2013 for financing the project webGRAS and the Hydrographic Office of the Province Bolzano/Bozen for providing temperature data. References Bovolenta S., Spanghero M., Dovier S., Orlandi D. and Clementel F. (2008) Chemical composition and net energy content of alpine pasture species during the grazing season. Animal Feed Science and Technology 146, 178-191. Hawkins D.M., Basak S.C. and Mills D. (2003) Assessing model fit by cross-validation. Journal of Chemical Information and Computer Science 43, 579–586. Parsons D., Cherney J.H. and Gauch H.G. (2006) Estimation of preharvest fiber content of mixed alfalfa–grass stands in New York. Agronomy Journal 98, 1081–1089. Peratoner G., Bodner A., Stimpfl E., Werth E., Schaumberger A. and Kasal A. (2010) A simple model for the estimation of protein content of first-cut meadow forage. Grassland Science in Europe 15, 539-541. Schaumberger A. (2011) Räumliche Modelle zur Vegetations- und Ertragsdynamik im Wirtschaftsgrünland, Dissertation, Technische Universität Graz, Institut für Geoinformation, 264 p. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 657 Added value chain of the dairy industry and its development in Central Switzerland Hofstetter P. Vocational Education and Training Centre for Nature and Nutrition (BBZN), CH-6170 Schuepfheim, Switzerland Corresponding author: Pius.Hofstetter@edulu.ch Abstract Since Switzerland (CH) is not a member of the European Community, it pursues its own agricultural policy, which imposes increased competition and an incremental approach to European conditions. In 2012, in Central-Switzerland (C-CH), 19% of the farms were in the hill region (HR) and 52% were in the mountain region (MR). The average agricultural area of the 4487 Central-Swiss dairy farms was 18 ha with a milk supply on contract of 123,500 kg/year and a productivity of 6260 kg of milk/ha. A full costs analysis showed that farms in the valley regions (VR) were more productive than farms in the HR and MR (78 kg vs. 68 kg vs. 50 kg of milk/h (working hour) respectively). The overhead and the internal overhead costs were higher in the HR and in the MR than in the VR. The cheese factories play an important role in added value, especially in remote valleys. A promising prospect under the challenging conditions (e.g. free cheese trade with the EU and the abandonment of milk quotas) is a higher efficiency of the added value chain, i.e. to reduce production costs and to produce niche products, including specialities (e.g. organic products) and brand merchandising. Introduction In Central-Switzerland (C-CH), situated in the northern foothills of the Alps, there are high quality grassland areas in the lowlands, as well as mountain areas of high natural value with particularly unfavourable topographical and climatic conditions. In 2008, 6.7% of the total working population (398,609 people) of C-CH was engaged in the primary sector (agriculture, forestry and fishery), 27.3% in the secondary sector (industry and handicraft) and 66.0% in the tertiary sector (services) (LUSTAT, 2012). There are significant differences between the six cantons, e.g. in Uri, 10.6% of the working population worked in the primary sector, mostly in the MR, but only 2.2% did so in Zug (in the majority of cases in the VR). Also, in Zug, 72.9% of the population is engaged in the tertiary sector. In C-CH, the dairy industry plays an important role. The aim of this study was to investigate the evolution of the dairy industry from 2000 to 2012, especially the full costs of dairy farms under different geo-economic conditions. Materials and methods Data on agriculture, dairy farms and factories, milk processing, consumption of cheese and food were taken from Bundesamt für Statistik (2014), the Swiss Federal Office for Agriculture (FOAG, 2014), the Swiss Milk Producer (SMP, 2014), the Central Switzerland Dairy Farmers’ association (ZMP, 2012) and the GOTTLIEB DUTTWEILER INSTITUTE (GDI, 2008). Full cost accounting was conducted based on the data by Haas and Höltschi (2013), and it was performed according to VOKO-Milch + Schweine by AGRIDEA (2014). Results and discussion Analysis of the farm size structure: Table 1 shows that from 2000 to 2012 the total of farms was reduced by 15.6% in C-CH, whereas they decreased all over Switzerland (CH) by nearly 20%. In C-CH, the share of farms in the HR and mountain region MR was about 15% higher compared to the Swiss average. The agricultural gross value added (i.e. production value minus preparatory efforts) decreased about 8% more in C-CH than in all Switzerland. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 658 Table 1. Farm size structure in Central-Switzerland in comparison to all Switzerland. Central-Switzerland (C-CH) Year 1 2000/01 2012/13 2000/01 2012/13 Farms (all production systems), number 10659 8998 70537 56575 Therefrom of full-time farms (>75%), % 73.7 72.5 69.8 71.1 Therefrom of organic farms, % 5.5 9.4 7 10.4 Therefrom in the hill region, % 18.6 18.7 13.7 14.3 Therefrom in the mountain region, % 51.2 51.6 41.5 41.3 Gross value added, Mio. (EUR) 492 334 3910 2975 Dairy farms, number 6984 4487 38082 24103 15.4 18.0 19.1 23.8 72351 123500 81691 135440 5113 6260 4278 5595 4720 5805 4994 5989 Area (ha) 2 Merchandised milk (kg) Merchandised milk/ha (kg) 2 Merchandised milk/cow (kg)2 1 Switzerland (CH) 2 In 2006, the former milk quota regulation of the Swiss Confederation was abolished. Data of C-CH from 2002/03. Dairy farms: From 2000/01 to 2012/13, in the C-CH, the milk quota (since 2006 by private regulation) rose by 70% to an average of 123,500 kg of milk/farm/year, which was 9% lower than the total Swiss average. The number of C-CH dairy farms decreased by 36%. In C-CH, the merchandised milk kg per cow and per ha increased, especially the area productivity (e.g. the 3459 ZMP milk producer had on average an area of productivity of 6925 kg/ha/year. This demonstrates the high production rate of C-CH dairy farms. Full costs accounting: The milk production per main forage area and the productivity of labour (kg of milk/h (working hour)) were very different in all regions (Table 2). More direct payments in the HR and MR could partly compensate for higher overhead (machinery, buildings and equipment) and internal overhead costs (wage entitlement) in these regions. The total agricultural income was significantly higher in the VR compared to the MR. In the MR, there was a higher return from by-products (calves or cull cows) compared to the VR and HR. Compared to the average of all production systems, full-time grazing systems in the VR, HR and MR showed a higher labour income (€/h) of 23%, 6% and 4% respectively. In the VR, organic farms showed a 39% higher labour income than usual farms (e.g. with an ecological performance certificate) and farms with hay (for cheese factories) conservation systems showed a 15% higher labour income than farms with silage. In the HR, organic farms (+25%) and farms with hay (+6%) generated a higher labour income. In the MR, organic farms had a higher labour income (+13%) while farms with hay conservation systems had nearly the same income (+1%). Milk processing and cheese factories: In 2012, within CH, 88% of the total amount of milk was processed (therefrom 48% went to cheese, 19% to butter, 9% to cream and 23 % to other products (milk preserves, yoghurt, etc.)). In C-CH the two largest creameries processed 44% (i.e. 1537 Mio. kg/year) of the total amount of merchandised milk. But the number of small cheese factory co-operatives (average milk processed of Mio. 2.28 kg/year) decreased from 139 in 2000 to 50 in 2012. In 2012, within C-CH, there were 26 Emmental cheese factories (75% fewer than in 2000) and 16 Sbrinz cheese factories (27% fewer than 2000), 2 Le Gruyère cheese factories (33% fewer than in 2000) and 8 semi-hard and soft cheese factories (11% fewer than in 2000). This immense decrease of hard cheese factories was due to changes in the eating and consumption behaviour of soft and semi-hard cheese, the abandonment or merger of Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 659 uncompetitive small units, and the farmers changing their conservation system from hay to silage or abandoning milk production in favour beef production. Limitations: The farms for the full costs analysis were not randomly selected. These data were recorded from farms whose managers were attending further vocational training in dairy farming. We suspect that these farms have a higher performance rate than average ones. Table 2. Full costs analysis of dairy farms (mean ± SD) in valley region (VR), hill region (HR) and mountain region (MR) in Switzerland (without group farming). Book keeping data from 2009 to 2012 VR HR MR Significance1 About half of these farms are in C-CH 174 dairy farms 68 dairy farms 105 dairy farms VR VR HR v v v Costs and income are per kg of milk Mean ±SD Mean ±SD Mean ±SD Agricultural area (ha) 31 15 25 12 25 14 ** ** Number of cows (cows/dairy farm) 35 16 29 13 22 10 * ** ** ** ** ** Merchandised milk (kg year-1) 257314 128378 203120 111513 125885 74028 HR MR MR Milk production/cow / year (kg) 7800 1074 7307 1110 6393 1117 ** ** ** Milk/main forage area (kg ha-1) 12800 3924 10601 3285 6666 3255 ** ** ** Labour productivity (kg of milk h1 ) 78 23 68 24 50 15 ** ** ** Direct costs (Cent kg-1) 2 -22 5.7 -22 5.4 -23 6.5 Therefrom concentrate (Cent kg-1) -9 3.2 -9 3.1 -9 3.9 Overhead costs (Cent kg-1) -32 8.2 -36 11.9 -46 15.0 ** ** ** Internal overhead costs (Cent kg -1) -26 11.5 -31 13.5 -51 23.0 * ** ** Full costs (Cent kg-1) -80 13.1 -88 22.7 -119 33.5 ** ** ** Milk price (Cent kg-1) 51 7.0 52 7.5 52 8.2 Direct payment (Cent kg-1) 17 7.0 23 7.1 39 15.2 ** ** ** Profit/loss (Cent kg-1) -12 11.2 -13 12.8 -28 22.5 ** ** Labour income h-1 (€) 11 8.1 13 5.8 11 5.4 Agriculture income dairy farming (€) 39014 28585 40914 27888 38516 26492 Therefrom by-products (€) 3 5559 5519 6738 5591 10967 9733 ** 63200 41412 58019 40185 48750 32715 ** Agricultural income total (€) 1 2 ** ** 3 * = P < 0.05%; ** = P < 0.01. Actual currency: 1 CHF = 0.818034 EUR. Calves and cull cow returns. Conclusions From 2000 to 2012, in C-CH, the number of dairy farms and their agricultural gross value added decreased. Therefore, jobs providing additional income are essential in such fringe areas. With regard to the high production costs of the dairy farms, small to medium sized farms, especially, are under considerable strain. The productivity of the dairy farms in all regions must improve due to rising competition, e.g. co-operation, particularly with respect to farm mechanisation and the use of buildings, becomes very important. With a rising demand for meat worldwide, the by-products will become increasingly important. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 660 In mountain regions, cheese factories maintain added value. If cheese factories can develop, apart from improving their efficiency, milk-based specialities and then merchandise them in the European marketplace, they may generate a higher-than-average milk price. Such a milk price would partially compensate for higher production costs in mountain areas. References AGRIDEA (2014) VOKO Milch + Schweine, Eschikon 28, CH-8315 Lindau. Bundesamt für Statistik (2014) Landwirtschaftliche Betriebszählungen und Betriebsstrukturen. http://www.bfs.admin.ch/bfs/portal/de/index/themen/07/03.html, CH-2010 Neuchatel. FOAG (2014) Auswertung der Daten über die Milchproduktion Milchjahr 2012/13. CH-3003 Bern. GDI (2008) European Food Trends Report. Bossart D. and Hauser M., CH-8803 Rüschlikon /ZH. Haas Th. and Höltschi M. (2013) Full costs accounting data. BBZN, CH-6276 Hohenrain. LUSTAT (2012) Statistik Luzern. Jahrbuch Kanton Luzern, 2012. Burgerstrasse 22, CH-6002 Luzern. SMP (2014) http://www.swissmilk.ch/de/produzenten/milchmarkt/zahlen-fakten-milchmarkt.html, Weststrasse 10, CH-3000 Bern 6. ZMP (2012) MVD 2012 ZMP. Central Switzerland Dairy Farmers’ association. CH-6002 Luzern. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 661 Economics of grazing Van den Pol-van Dasselaar A., Philipsen A.P. and de Haan M.H.A. Wageningen UR Livestock Research, P.O. Box 65, 8200 AB Lelystad, the Netherlands Corresponding author: agnes.vandenpol@wur.nl Abstract This study provides insight into the economics of grazing on modern Dutch dairy farms and shows how the yield from grazing can be improved. Model calculations show that grazing is financially attractive if the cows eat sufficient amounts of fresh pasture grass (> 600 kg DM cow-1 yr-1). If the intake of fresh grass is very low, grazing is less profitable than summer feeding. Statistical analysis of the actual financial results of Dutch commercial dairy farms confirmed the positive effect of grazing. The model calculations showed a larger financial benefit of grazing than actually achieved in practice. The differences were examined together with extension services. Feasible possibilities for increasing the benefits of grazing were studied. It was concluded that commercial dairy farms often do not take full advantage of grazing because their operational management is not optimally adapted to grazing and maximum fresh grass intake. Grazing strategies must be implemented consistently for optimal financial benefit. The intake of fresh grass must be high, but at least more than 600 kg DM cow1 yr-1. Keywords: economy, grazing, management Introduction In northwest Europe, grazing is a matter of public concern (e.g. Van den Pol-van Dasselaar et al., 2008). The main reasons for this are animal welfare, biodiversity and the positive image of grazing. Grazing has both positive and negative effects on the environment, the most obvious being nutrient loss. In general, grazing is economically attractive for farmers. However, an average situation will not be applicable to all farms. Certain conditions may be less favourable for grazing. Van den Pol-van Dasselaar et al. (2010) showed for the Netherlands that, for situations with automatic milking systems, large herds and high milk yields per cow, the farmer’s income remained the highest for grazing. However, in situations with more than 10 dairy cows ha-1 grazing surface, it was not profitable to practise grazing. There was a strong relationship between intake of grass in pasture, on a typical farm, and the difference in income between grazing and no grazing. The more grass the cows ate in the pasture, the larger the income profit from grazing compared to no-grazing and feeding roughage. However, some time has passed since this study. Since then, (environmental) policy has changed, farms developed and the corresponding increases in scale have continued. Farming systems have become increasingly intensive, and more and more farms have started to use automatic milking systems. This study therefore focuses on the economics of grazing on modern Dutch dairy farms and the possibilities to improve the financial benefits from grazing on these farms. Materials and methods Gross margins were used to calculate the economics of grazing in the Netherlands for the near future (2015-2020) using the model DairyWise (Schils et al., 2007). The DairyWise model is an empirical model that simulates technical, environmental and financial processes on a dairy farm. The central component is the FeedSupply model that balances the herd requirements, as generated by the DairyHerd model, and the supply of home-grown feeds, as generated by the crop models for grassland and maize silage. The GrassGrowth model predicts the daily rate of dry matter accumulation of grass, including several feed quality parameters. The final output is a farm plan describing all material and nutrient flows and the consequences on the environment and economy. As part of the calculation, the expected developments and trends with respect to Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 662 policy, increasing scale and automation were taken into account. Production intensities ranging from 15,000 to 30,000 kg milk ha-1 and average milk productions of 8,500 kg milk cow-1 yr-1 were used. Herd size varied between 70 and 280 cows. The benefits of grazing were primarily determined by the lower subcontracting costs, higher costs for roughage, lower costs for feed concentrate and lower costs for manure disposal. Next to modelling, data from real-life farms were used. The method data envelopment analysis (DEA) (Steeneveld et al., 2012) was used for statistical analysis of farm data collected by accounting firms and advisors. The results illustrate the actual financial results of approximately 10% of all Dutch commercial dairy farms in 2011. The study used six data sets of accounting firms and advisors (Countus, DLV, DMS, Flynth, LEI and PPP-Agro Advies). Results and discussion The results of the gross margins calculated with DairyWise are summarized in Figure 1. It shows that grazing is financially more attractive than summer feeding if the cows eat sufficient amounts of fresh pasture grass (> 600 kg DM cow-1 yr-1). If the intake of fresh grass is very low, grazing is not advantageous over summer feeding. In practice, the effect should be even more positive than shown in Figure 1, since the majority of the Dutch farmers currently receive a grazing premium of 0.5 euro for each 100 kg milk if they graze their dairy cattle for at least 120 days yr-1 and 6 h d-1. This additional income was not taken into account in Figure 1. When analysing the financial records of commercial dairy farms, large differences were found between farms regarding efficiency and gross operating profit. On average, grazing resulted in more efficient operational management and a higher gross margin. However, these positive results declined in relation to increasing farm size. In 2011 the transition point was, on average, a farm size of about 90 dairy cows. If grazing was combined with automatic milking, much of the financial advantage of grazing disappeared. In 2011, the majority of dairy farms did not have the option of receiving a grazing premium. Today, however, most dairy companies have implemented the grazing premium. The current grazing premium would have made the transition point move up to a farm size of approximately 130-140 cows. Unfortunately, the actual grass intake on the commercial dairy farms was not known. Therefore, it was not possible to relate the grass intake to the farm income. The category ‘grazing farms’ included both farms with very low grass intake and farms with full grazing. The results as shown in Figure 1 implicate that knowledge on the actual grass intake would have led to a more detailed insight in the economics of grazing on commercial dairy farms. In the model calculations, the financial benefits of grazing were larger than those actually achieved in practice in 2011. To understand these differences more clearly, the situation in practice was examined along with the feasible possibilities for increasing the benefits of grazing. This exercise was done together with extension services. It was concluded that farms which graze their livestock do not take full advantage because their operational management is not optimally adapted to grazing and optimal grass intake. The intake of fresh pasture grass by their cattle is inadequate. Dairy farmers can improve the financial yield from grazing with relatively simple measures, like providing less supplementary feed, and by starting the grazing early in the season and ending late in autumn. By using the land parcels closest to the barn for grazing, and growing silage maize further away, or providing access to pasture on the other side of the road, the fresh grass intake can possibly be increased even further. Finally, a grazing strategy must be implemented consistently for optimal financial benefit. To benefit financially, the intake of fresh grass must be sufficient, at least more than 600 kg DM cow-1 yr-1. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 663 Figure 1. Income with grazing minus income with summer feeding relative to the amount of fresh grass intake in kg dry matter (DM) cow-1 yr-1 for three soil types. Conclusion In many cases, grazing offers financial benefits, on larger farms and also on automated farms. However, this does not apply to every dairy farm. When transposing research results into practice, the context of the individual dairy farm must always be taken into account. The current study has provided possibilities for improving the financial yield on farms that choose to graze their cattle. The results show that the fresh grass intake of the cattle is crucial for financially beneficial grazing. Acknowledgements This study was funded by the Dutch ministry of Economic Affairs and the FP7 MultiSward project (Grant Agreement 244983). References Schils R.L.M., de Haan M.H.A., Hemmer J.G.A., van den Pol-van Dasselaar A., de Boer J.A., Evers A.G., Holshof G., van Middelkoop J.C. and Zom R.L.G. (2007) DairyWise, A Whole-Farm Dairy Model. Journal of Dairy Science 90, 5334-5346. Steeneveld W., Tauer L.W., Hogeveen H. and Oude Lansink A.G.J.M. (2012) Comparing technical efficiency of farms with an automatic milking system and a conventional milking system. Journal of Dairy Science 95, 73917398. Van den Pol-van Dasselaar A., Vellinga T.V., Johansen A. and Kennedy E. (2008) To graze or not to graze, that’s the question. Grassland Science in Europe 13, 706-716. Van den Pol-van Dasselaar A., De Haan M., Evers A. and Philipsen A.P. (2010) Simulation of the effect of grass intake on the farmer’s income. Grassland Science in Europe 15, 100-102. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 664 Does early spring grazing stimulate spring grass production? van Eekeren N.1, Rietberg P.1, Iepema G.2 and de Wit J.1 1 Louis Bolk Institute, Hoofdstraat 24, 3972 LA Driebergen, The Netherlands 2 Agricultural College van Hall Larenstein, Postbus 1528, 8901 BV Leeuwarden, The Netherlands Corresponding author: N.vanEekeren@Louisbolk.nl Abstract Early spring grazing may stimulate grass growth in spring by increasing grass root exudation, which enhances mineralization. The objective of this study was to investigate the net effect of early spring grazing on spring grass production. Therefore, we conducted two field experiments on production grasslands. Simulated early spring grazing negatively affected the yield of the first cut: compared to ungrazed plots on sandy soil and a tendency on clay soil, dry matter yield of early-grazed plots was reduced by 20% (sand, P=0.009) and 12% (clay, P=0.062), respectively. These differences were partly compensated in later cuts. Despite the negative effect on grass yield of the first cut, early grazing positively affected the crude protein (CP) content of the grass in the first cut on sandy soil, but only in the plots that had not been fertilized with CAN. The stimulating effect of early spring grazing on soil nutrient mineralization appears to be too small to compensate for the negative effects of early grazing on grass leaf area and photosynthesis capacity. Keywords: Grazing, root exudates, mineralization, spring production Introduction For dairy farmers the key question in spring is how to get the grass growing as soon as possible. The easiest method is to add nitrogen fertilizer, but due to stricter fertilizer regulations farmers have to depend more and more on soil nutrient mineralization. This process starts when the soil warms up in spring, and can be stimulated by mechanical soil aeration and lime additions. Furthermore, according to the experience of various dairy farmers, grass growth may also be stimulated by ‘early spring grazing’, a method in which grasslands are grazed in early spring for a short period of time. A possible explanation for this effect is that grazing leads to increased root exudation, which in turn triggers mineralization (Hamilton et al., 2008). However, it is not known whether this mechanism also works in spring when soil temperatures are still low. Furthermore, it is unclear whether the positive effects of grazing on mineralization could outweigh the negative effect of (early) grazing on photosynthetic leaf area and thus growth rate. Therefore, the objective of this study was to investigate the net effect of early spring grazing on spring grass production. Materials and methods In 2013 we conducted two field experiments on production grasslands: one on a shallow sandy soil in the south of the Netherlands, and one on a clay soil in the north. All plots in both experiments were fertilized with cattle slurry at 25 m3 ha-1 in mid-March. The experiment on sandy soil consisted of four treatments with six replicate plots: early spring grazing with, and without, the addition of artificial fertilizer; artificial fertilizer only; and a control. To this end, on 19 April half of the plots were mowed to ±4cm with a lawn mower to simulate early spring grazing. On the same day, half of the mowed and unmowed plots received 50 kg N ha-1 of calcium ammonium nitrate (CAN). The experiment on clay soil consisted of three treatments with five replicate plots: early spring grazing on 19 April (lawn mower treatment, see above); artificial fertilizer (50 kg N ha-1 of CAN); and a control. The effect of early spring grazing was measured in the first two cuts. The plots on sandy soil were cut on 19 May and 8 July; and on clay soil on 24 May and 2 July. Grass production of each plot was determined by mowing a Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 665 strip of 0.84 m × 5 m (sandy soil) or 1.50 m × 10 m (clay soil). Half of the yield from mowing the plots at t=0 to simulate early grazing was added to the yield of the first cut. After weighing the fresh biomass, a sub-sample was analysed for dry matter and crude protein (CP) content. Results were analysed for significance by ANOVA and Tukey’s test. Results Simulated early spring grazing negatively affected the yield of the first cut compared to ungrazed plots on sandy soil, and there was a tendency on clay soil: dry matter yield of earlygrazed plots was reduced by 20% (sand, P=0.009) and 12% (clay, P=0.062), respectively (Figure 1). These losses were partly compensated by the yields of later cuts: on sandy soil, total grass yield of the first and second cut of early-grazed plots was only 8% lower than of ungrazed plots; on clay soil it was only 5% lower. This ‘catch-up’ effect is likely to be due to higher regrowth rates after the first cut in the early-grazed plots, compared with the ungrazed plots where the first cut was heavier because no simulated grazing had taken place. Figure 1. The effect of early spring grazing on dry matter yield of the first cut (sandy soil: 19 May; clay soil: 24 May) in plots with and without artificial fertilizer addition. Note that the yield from early-grazed plots includes 50% of the yield from mowing these plots at t=0 to simulate early grazing. Slurry=cattle slurry (25 m3 ha-1, applied to all plots including the control), CAN (50 kg N ha-1). Error bars represent 2×standard deviation. Despite the negative effect on grass yield of the first cut, early grazing positively affected the CP content of the grass in the first cut on sandy soil, but only in the plots that had not been fertilized with CAN. On sandy soil, early grazing increased the CP content of the first cut by 12%, to 151 g kg dm-1 compared to 135 g kg dm-1 in ungrazed plots (Figure 2, P=0.001). On clay soil the difference of 6% was not significant: 144 g kg dm-1 in the early grazing plots compared to 136 g kg dm-1 in ungrazed plots. The results on sandy soil suggest that early spring grazing promotes feed quality. This can be considered a positive effect of early spring grazing, particularly if feed quality is an issue (for example, in the case of delayed cuts, and in organic dairy pastures with low clover density). Due to the higher relative CP content of early-grazed grassland, the total nitrogen yield of early-grazed versus ungrazed plots was not significant different, in both sandy and clay soil experiments. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 666 Figure 2. The effect of early spring grazing on CP content of the first cut (19 May) in plots with and without artificial fertilizer, in the sandy soil experiment. Slurry=cattle slurry (25 m3 ha-1, applied to all plots including the control), CAN (50 kg N ha-1). Error bars represent 2×standard deviation. Discussion Contrary to expectations, early spring grazing was not found to ‘kick-start’ grass production. It is possible that the timing of the first cut was too early to detect a positive effect but, even if that were the case, our results show that the effect of artificial fertilizer on dry matter yield and CP content is likely to be much greater than any effect of early grazing (Figures 1, 2). Thus, it appears that early spring grazing provides no real alternative to artificial fertilizer for encouraging early grass growth in spring. Other methods still worth exploring are mechanical soil aeration and liming, both of which should stimulate soil biotic activity, nutrient mineralization, and hence grass growth. Conclusions Early spring grazing leads to a higher crude protein content in grass on sandy soil. The stimulating effect of early spring grazing on soil nutrient mineralization appears to be too small to compensate the negative effects of early grazing on photosynthesis capacity. The negative effect of early grazing on dry matter yield of the first cut is compensated in later cuts. References Hamilton E.W., Frank D.A., Hinchey P.M. and Murray T.R. (2008) Defoliation induces root exudation and triggers positive rhizospheric feedbacks in a temperate grassland. Soil Biology & Biochemistry 40, 2865–2873. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 667 Use of milk fatty acid composition to authenticate cow diets Coppa M.1, Revello-Chion A.2, Giaccone D.2, Comino L.1, Tabacco E.1 and Borreani G.1 1 University of Turin, Department of Agricultural, Forest and Food Sciences, Via L. da Vinci 44, 10095, Grugliasco, Italy 2 Associazione Regionale Allevatori del Piemonte, Via Livorno 60, 10144, Turin, Italy Corresponding author: mauro.coppa@unito.it Abstract The aim of this work was to authenticate cow feeding system from fatty acid (FA) concentrations of cow milk. The milk samples and records of their related farming practices were collected on 156 commercial farms in the lowland area of the Piedmont Region (NorthWest Italy). Milk samples were analysed for FA composition by gas-chromatography. To define the main cow feeding systems, a hierarchical cluster analysis was performed on the proportion of different feeds in the cow diets. Samples were classified into two main groups, according to whether high or low forage-to-concentrate ratio (FC): HFC and LFC, representative of extensive and intensive farming systems. Each group was divided into two subgroups: the HFC into FP and FH, in which the main forage was given as pasture, or hay, respectively; the LFC into CR or CC, if concentrates were given as raw materials or as pre-formed commercial concentrate mix, respectively. Linear discriminant analysis (LDA) was performed on milk FA concentration to authenticate the FC groups, and correctly classified 95.5% of samples in crossvalidation. Promising results were found in the authentication of FP and FH subgroups (15.4% of error in cross-validation). LDA was not able to distinguish between CR and CC milk. Keywords: milk fatty acid, feeding system, authentication Introduction Several researches have been made to identify reliable authentication methods for dairy products (Prache et al., 2005). Among the different potential analytical tools, milk fatty acid (FA) composition was the more effective method to achieve precise information about cow feeding (Engel et al., 2007). Differences in milk FA composition according to cow diet have been shown by several authors (Chilliard et al., 2007; Coppa et al., 2013). However, only few studies have tested the potential of bulk-milk FA profile to authenticate the cow diets adopted by commercial farms. The aim of this work was to authenticate cow diets from FA concentrations of bulk milk collected on different commercial farms in North-West Italy. Materials and methods Bulk cow-milk samples and information on their related farming practices were collected from 156 commercial farms located in the lowland area of Piedmont Region, in North-West Italy, across spring and summer 2013. During each milk-sampling, cow diet composition was recorded through a detailed on-farm survey. Each milk sample was analysed for FA composition by gas-chromatography. To define the main cow feeding systems, a hierarchical cluster analysis was performed on survey data on the proportion of different feeds in the cow diet. Linear discriminant analysis (LDA) was performed on milk FA concentration to authenticate the groups derived from the cluster analysis. Results and discussion Milk samples were classified into two main groups by the cluster analysis, according to whther they had a high or low forage-to-concentrate ratio (FC): HFC and LFC. Each group was divided into two subgroups: the HFC into FP and FH, in which the main forage source was either pasture or hay, respectively; the LFC into CR or CC, in which the concentrates were given as raw materials or as pre-formed commercial concentrate mix, respectively. The milk yield was about Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 668 9.3 kg/cow per day for HFC, and 25.6 kg/cow per day for LFC, and 8.3, 10.2, 25.0 and 26.1 kg/cow per day for FP, FH, CR and CC, respectively. The average diet composition of HFC and LFC groups, and of FP, FH, CR and CC subgroups, are given in Table 1. The HFC feeding system was representative of extensive farming systems, with cow diets based on herbagederived forages, with high proportions of forage. The LFC feeding system was representative of intensive farming systems, which are based on cow diets of maize silage, with a high concentrate proportion. The FP and FH subsystems of HFC represented the seasonality of herbage-based forage production systems: grazing pasture when fresh herbage is available, and/or conserved forage (mainly hay) for the periods of the year in which fresh herbage is unavailable. The CR and CC subgroups of LFC represented different farming strategies, involving a cheaper, but more complex (CR), or a easier, but more expensive, (CC) concentrate supplementation and management (Borreani et al., 2013). Table 1. Average diet composition of the high (FC+) or low (FC-) forage-to-concentrate ratio groups, and of their two subgroups: main forages given as pasture (FP), or as hay (FH), and concentrates given as raw materials (CR) or as pre-formed commercial concentrate mix (CM). Feedings (% of diet DM1) Maize silage HFC 12.1 HFC LFC 38.6 LFC FP FH CR CC 6.5 17.2 38.6 38.6 Grass or legume silage 0.0 1.1 0.0 0.0 1.0 1.3 Hay 40.9 18.2 14.6 64.9 19.0 17.7 Fresh herbage 34.7 0.4 67.4 4.9 1.0 0.0 Soybean meal 0.2 4.6 0.4 0.1 9.7 0.6 Maize or barley flour 2.0 13.4 2.3 1.7 16.2 11.2 Commercial concentrate mix 6.6 17.0 4.1 8.9 4.2 27.4 Total forages 88.1 60.6 89.3 87.0 62.3 59.3 Total concentrates 11.9 39.4 10.7 13.0 37.7 40.7 The results of linear disciminant analysis based on milk FA composition between HFC and LFC and, within each group, between the subgroups, are given in Table 2. The LDA between HFC and LFC correctly classified the 95.5% of samples in cross-validation. The milk of HFC farms was characterized by high concentration of n-3 FA (C18:3n-3, C20:5n-3, C22:5n-3), conjugated linoleic acid (CLA)c9t11, CLAc9c11, C18:2t11c15, branched chain FA (BCFA), C17:0, C17:1c9, and C15:0. The LFC milk had higher concentrations of even-chain saturated FA (ECSFA), n-6 FA, C18:1t12, C18:2c9t12+c9t14, C18:1t6/7/8, C18:1t9, and C18:1t10. These FC+FA profiles are similar to those reported in the literature for milk derived from fresh herbage- or hay-based diets, and those of FP, to milk derived from low forage-based diet (Vlaemink et al., 2006; Chilliard et al., 2007).The results of the LDA performed between FP and FH groups were promising, especially considering the low number of samples, but were less reliable than those between the two main groups (15.4% of error in cross-validation). The milk of FP farms was characterized by higher concentrations of C18:1t11, C18:2t11c15, C18:1t9, C18:1t13, C18:1t5, C18:1c10, C18:2c9t13, CLAc9t11, CLAc9c11 and C18:0, compared to FH milk. The milk of the FH group showed high concentrations of ECSFA compared to FP milk. The effect of pasture feeding on milk FA composition is well known (Chilliard et al., 2007; Coppa et al., 2013) and in agreement with our results. The LDA was not able to distinguish between CR and CC milk: only 66.3% of samples were correctly classified in cross-validation. However, the CR milk had higher concentrations of BCFA, n-3 FA, C18:2t11c15 and C17:0, than the CC milk. The CC milk had higher concentrations of n-6 FA, and of several C18:1 isomers different from C18:1t11. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 669 Table 2. Results of linear discriminant analyses based on milk FA composition between the high (HFC) or low (LFC) forage-to-concentrate ratio groups, and within each group, between the two subgroups: main forages given ad pasture (FP), or as hay (FH), and concentrates given as raw materials (CR) or as pre-formed commercial concentrate mix (CM). Feeding system1 Number of milk samples HFC vs. LFC FP vs. FH CR vs. CC 157 65 92 Calibration Samples Correct misclassified classifications (%) 4 97.5 4 93.8 16 82.6 Cross-validation Samples Correct misclassified classifications (%) 7 95.5 10 84.6 31 66.3 Conclusion In the Piedmont Region, two main feeding systems were identified, corresponding to intensive and extensive farming systems. The milk FA composition was successfully used to authenticate milk derived from these two systems. The milk FA composition gave also promising results in the authentication of fresh herbage- or hay-based diets within the extensive systems. It was not, however, possible to authenticate the concentrate supplementation strategy within the extensive farming systems. Acknowledgements This research was funded through the 'Bando per la Ricerca Scientifica 2011 Fondazione CRC; Progetto Migliorlat: Miglioramento della qualità e dello sviluppo competitivo della filiera latte piemontese'. References Borreani G., Coppa M., Revello-Chion A., Comino L., Giaccone D., Ferlay A., and Tabacco E. (2013) Effect of different feeding strategies in intensive dairy farming systems on milk fatty acid profiles, and implications on feeding costs in Italy. Journal of Dairy Science 96, 6840-6855. Chilliard Y., Glasser F., Ferlay A., Bernard L., Rouel J., and Doreau M. (2007) Diet, rumen biohydrogenation and nutritional quality of cow and goat milk fat. European Journal of Lipid Science and Technology 109, 828-855. Coppa M., Ferlay A., Chassaing C. Agabriel C., Glasser F., Chilliard Y., Borreani G., Barcarolo R., Baars T., Kusche D., Harstad O.M., Verbič J., Golecký J. and Martin B. (2013) Prediction of bulk milk fatty acid composition based on farming practices collected through on-farm surveys. Journal of Dairy Science 96, 4197– 4211. Engel E., Ferlay A., Cornu A., Chilliard Y, Agabriel C., Bielicki G. and Martin B. (2007) Relevance of isotopic and molecular biomarkers for the authentication of milk according to production zone and type of feeding. Journal of Agricultural and Food Chemistry 55, 9099-9108. Prache S., Cornu A., Berdagué J.L. and Priolo A. (2007) Traceability of animal feeding diet in the meat and milk of small ruminants. Small Ruminant Research 59, 157–168. Vlaemink B., Fievez V., Cabrita A. R. J., Fonseca A. J. M., and Dewhurst R. J. (2006) Factors affecting odd- and branched-chain fatty acids in milk: a review. Animal Feed Science and Technology 131, 389-417. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 670 Potential lipid markers of plant species from grasslands to authenticate mountain dairy foods Barron L.J.R.1, Aldezabal A.2, Valdivielso I.1, Bustamante M.1, Amores G.1, Virto M.1, Ruiz de Gordoa J.C.1 and de Renobales M.1 1 Lactiker Research Group, Faculty of Pharmacy, University of the Basque Country (UPV/EHU), 01006 Vitoria-Gasteiz, Spain, 2 Department of Plant Biology and Ecology, Faculty of Science and Technology, UPV/EHU, 48940 Leioa, Spain Corresponding author: luisjavier.rbarron@ehu.es Abstract In different mountain areas of Europe milk and cheese are traditional foods, and their survival in the future will depend on the sustainability of the mountain grazing livestock. High quality milk and cheese production, together with adequate tools to protect and authenticate these traditional foods, are important issues to ensure the continuity of the mountain grazing systems. Plant lipids can be potential markers of products obtained from grass-fed animals due to the transfer of these compounds directly from pasture to foods or because they are precursors for other specific lipid compounds generated by the animal metabolism. In this study, fatty acids (FAs), tocopherols, carotenoids and terpenoids of the main plant species present in a grazing area of the Cantabrian Mountains were analysed. Some of the major potential lipid markers of dairy products obtained from mountain pasture-fed animals could be α-linolenic and linoleic acids, α-tocopherol and α-tocotrienol, lutein, β-carotene, α-pinene, β-ionone, β-thujene, αcopaene, γ-cadinene and β-cubebene. Keywords: mountain pasture, fatty acids, tocopherols, carotenoids, terpenoids Introduction Different types of lipids have been proposed as potential tracers for milk and cheese from grazing ruminants. Some of these compounds can come directly from pasture, whereas other compounds are derived from animal metabolism (Prache et al., 2005). The fatty acid (FA) composition of dairy products is primarily affected by the animals’ diet, and it has been reported that grazing livestock increases the content in the milk of some beneficial FAs, like vaccenic and rumenic acids. Dairy products from grass-fed ruminants are also enriched in polyunsaturated FAs (PUFAs) such as linoleic and linolenic, which are the major PUFAs in fresh grass (Cabbidu et al., 2005). Tocopherols (vitamin E) and carotenoids, such as lutein and β-carotene, can also be directly transferred from fresh grass into milk, but this depends on the animal species and other nutritional factors (Morand-Fehr et al., 2007). Animal metabolism will also affect the conversion of β-carotene to retinol (vitamin A), and therefore, its accumulation in milk will depend largely on the ruminant specie (Noziére et al., 2006). Terpenoids originate from the secondary metabolism of plants and they can be rapidly transferred into milk (Viallon et al., 2000). This paper summarises the composition of main FAs, tocopherols, carotenoids and terpenoids of the most abundant plant species found in Cantabrian mountain pastures. Materials and methods Thirteen major plant species were sampled in the second week of May and June (2012) in pastures of the Aralar Natural Park (42° 59' 48" N 2° 06' 51" W) in northern Spain. Grass samples were lyophilized before analysis. FAs were extracted and analysed by GC-FID according to a method adapted from Chávarri et al. (1997). Tocopherols and carotenoids were simultaneously extracted (Cardinault et al., 2008). Then, tocopherols were analysed by normalGrassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 671 phase HPLC with a fluorescence detector (Panfili et al., 1994), while carotenoids were determined by reverse-phase HPLC coupled with a diode-array detector (Kimura et al., 2007). Grass terpenoids were analysed by SPME-GC-MS using a method adapted from Abilleira et al. (2010). Mean values of the aforementioned lipid compounds were calculated for each of the botanical families present in the pastures. Student’s t test was applied to study significant differences (P ≤ 0.05) in the lipid composition of botanical families between May and June. Results and discussion Around thirty FAs were detected in the studied plant species and PUFAs represented the highest content in all the botanical families. α-Linolenic content was the largest in Poaceae (52%), Fabaceae (43%), Lamiaceae (39%) and Rosaceae (34%). On the other hand, linoleic content was the largest in Asteraceae (44%) and Juncaceae (38%) (Table 1). Table 1. Major potential lipid markers from the main botanical families collected in summer (May and June) in Cantabrian mountain pastures. Mean values are expressed as mg/100 of dry matter except for terpenoids (peak area values relative to that of an internal standard). Lipid compound Monocotyledon families Dicotyledon families Poaceaea Juncaceaeb Fabaceaea Asteraceaea Rosaceaea Lamiaceaec (s=4) (s=1) (s=4) (s=2) (s=1) (s=1) α-linolenic acid *507 156 526 *337 *232 315 linoleic acid 209 468 331 566 *195 230 α-tocopherol 2.09 0.637 2.64 1.94 *2.62 6.86 α-tocotrienol 0.152 nd 0.127 0.334 nd nd lutein 13.9 4.79 11.8 *7.57 *7.72 6.82 β-carotene *2.95 1.62 2.38 *1.09 *1.53 1.83 α-pinene 16.7 10.9 52.4 432 18.8 113·102 α-phellandrene 14.8 10.9 25.3 419 *19.5 448·101 1 β-thujene 15.0 11.2 30.5 129·10 *19.5 395·101 β-ionone 190 62.8 *128 *125 145 156 isoeugenol 27.0 18.6 68.9 117 40.3 723·101 α-copaene *124 *104 165 333 *89.0 249 β-cubebene *142 102 226 458 *50.3 399 β-caryophyllene *142 68.8 179 304 *107 172·101 δ-cadinene 38.3 23.4 72.5 101·101 *67.1 824 γ-cadinene 152 90.7 *162 861 *188 479·101 a Samples collected in May and June; bsamples collected only in May; csamples collected only in June; s, number of plant species collected; *statistically significant (P  0.05) differences between months. These differences could be mainly related to genetic factors, phenological stage of plants or climatological variables. In this respect, the most remarkable differences were the higher content of α-linolenic in June than in May for Poaceae, Rosaceae and Asteraceae families. Seven tocols were detected in the plant species but α-tocopherol was the largest (≥ 70%) in all botanical families. Among other minor tocols, α-tocotrienol was found to be higher than 5% in Poaceae, Fabaceae and Asteraceae (Table 1). For Rosaceae, only α-tocopherol content was higher in June than in May. Seven xanthophylls and four carotenes were identified in the plant species but many peaks with carotenoid spectra, depending on the plant species, remained unknown. Despite the differences observed in carotenoid composition, lutein was the major carotenoid in all botanical families, ranging from around 50 to 75% of the total identified carotenoids, while the content of β-carotene ranged from around 12% in Poaceae and Fabaceae to 26% in Juncaceae (Table 1). Higher content of lutein and β-carotene was found in June than in May when significant differences were observed in botanical families. More than seventy terpenoids were found in the plant species from the mountain pasture, most were monoterpenes (53%) followed by sesquiterpenes (43%) and chemically irregular terpenes (4%). Significant differences between botanical families were found in the terpenoid content Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 672 and Lamiaceae appeared to be the botanical family with the highest content, of both mono- and sesquiterpenes (Table 1). The most abundant terpenoids found in Lamiaceae, Poaceae and Asteraceae were α-pinene (21%), β-ionone (9%) and β-thujene (9%), respectively, whereas the most abundant terpenoids found in Rosaceae, Juncaceae and Fabaceae were γ-cadinene (9%), α-copaene (8%) and β-cubebene (8%), respectively. Changes in the content of terpenoid composition between May and June were dependant on botanical families and individual compounds. Conclusion The characteristic lipid profile of major plant species of Cantabrian mountain pastures could provide useful information to authenticate dairy products obtained from grazing animals. Some of the major potential lipid markers from plants could be α-linolenic and linoleic acids, together with α-tocopherol, α-tocotrienol, lutein, β-carotene, α-pinene, β-ionone, β-thujene, α-copaene, γ-cadinene and β-cubebene. References Abilleira E., de Renobales M., Nájera A.I., Virto M., Ruiz de Gordoa J.C., Pérez-Elortondo F.J., Albisu M. and Barron L.J.R. (2010) An accurate method for the analysis of terpenes in milk fat by headspace solid-phase microextraction coupled to gas chromatography-mass spectrometry. Food Chemistry 120, 1162-1169. Cabiddu A., Decandia M., Addis M., Piredda G., Pirisi A. and Molle G. (2005) Managing Mediterranean pastures in order to enhance the level of beneficial fatty acids in sheep milk. Small Ruminant Research 59, 169-180. Cardinault N., Lyan B., Boreau M., Chauveau B., Rock E. and Grolier P. (2008) Development of a method to determine carotenoid composition of fresh forages. Canadian Journal of Plant Science 88, 1057-1064. Chávarri F., Virto M., Martín C., Nájera A.I., Santisteban A., Barron L.J.R., de Renobales M. (1997) Determination of free fatty acids in cheese: comparison of two analytical methods. Journal of Dairy Research 64, 445-452. Kimura M., Kobori C.N., Rodríguez-Amaya D.B. and Nestel P. (2007) Screening and HPLC methods for carotenoids in sweetpotato, cassava and maize for plant breeding trials. Food Chemistry 100, 1734-1746. Morand-Fehr P., Fedele V., Decandia M. and Le Frileux Y. (2007) Influence of farming and feeding systems on composition and quality of goat and sheep milk. Small Ruminant Research 68, 20-34. Nozière P., Graulet B., Lucas A., Martin B., Grolier P. and Doreau M. (2006) Carotenoids for ruminants: From forages to dairy products. Animal Feed Science and Technology 131, 418-450. Panfili G., Manzi P. and Pizzoferrato L. (1994) High-performance liquid-chromatographic method for the simultaneous determination of tocopherols, carotenes, and retinol and its geometric isomers in Italian cheeses. Analyst 119, 1161-1165. Prache S., Cornu A., Berdagué J.L. and Priolo A. (2005) Traceability of animal feeding diet in the meat and milk of small ruminants. Small Ruminant Research 59, 157-168. Viallon C., Martin B., Verdier-Metz I., Pradel P., Garel J.P., Coulon J.B. and Berdagué J.L. (2000) Transfer of monoterpenes and sesquiterpenes from forages into milk fat. Lait 80, 635-641. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 673 Is phytanic acid a suitable marker for authentication of milk and dairy products from grass-fed cows or organic farming systems? Capuano E.1, Elgersma A.2, Tres A.1 and Ruth S.M. van1 1 RIKILT – Institute of Food Safety, Wageningen University and Research Centre, P.O Box 230, Wageningen, The Netherlands, 6700 AE 2 Independent Scientist, PO Box 323, Wageningen, The Netherlands, 6700 AH Corresponding author: anjo.elgersma@hotmail.com Abstract Cow milk samples were collected from herds of 30 Dutch farms and analysed by gas chromatography-mass spectrometry for phytanic acid (PHY) and its diastereomers SRR and RRR to test the hypothesis that PHY could be a suitable marker for authentification of milk from grass-fed cows. The samples differed in the proportion of fresh grass in the cows’ daily dry matter intake (0 to 94%). Grass was either fed indoors or grazed (during daytime, or day and night). Of the latter category, three farms had an organic and three a biodynamic farming system. PHY concentrations were not significantly higher in organic/biodynamic milk compared with conventional milk, nor were they correlated with the proportion or amount of fresh grass in the diet. The proportion of RRR in total PHY was positively correlated with the proportion of fresh grass in the diet. These results indicate that, in contrast to our hypothesis, PHY content is not a suitable indicators of pasture grazing or organic/biodynamic farming, while the proportion of diastereomers of PHY may be useful as such. Keywords: milk, authentication, grass-fed dairy, organic, phytanic acid, grazing, cow diet Introduction In the Netherlands, the label 'Weidemelk' ('pasture milk') was introduced in 2012 for milk from cows that are kept at pasture at least four months per year and 6 hours per day (Elgersma, 2012). Phytanic acid (PHY) is a branched-chain fatty acid, produced by bacteria from enzymatic degradation of chlorophyll in the rumen. PHY cannot be synthesized de novo by mammals, and thus its presence in milk and animal-derived products depends exclusively on feed, especially on green materials, i.e. grass and forage which contain chlorophyll. Since the proportion of grass in animal rations often is higher in organically raised cattle, PHY has been proposed as a marker for for pasture-fed milk and dairy products, as well as for organic dairy products (Vetter and Schroeder, 2010). We hypothesized this might be used in Dutch milk to authenticate 'weidemelk' and organic milk. Along with PHY content, the ratio between the two diastereomers of PHY, i.e. 3R,7R,11RPHY (RRR) and 3S,7R,11R-PHY (SRR), which form upon biohydrogenation of the phytol double bond, has been shown to largely vary according to animal diet and has also been proposed for the authentication of pasture grazing and organic feeding (Schroeder and Vetter, 2011; Baars et al., 2012). However, some recent findings indicate that total PHY content is not necessarily directly correlated with the intake of green feed, being linked in a more complex way to the whole diet (Che et al., 2013). Capuano et al. (2014) reported that PHY content was not a suitable indicator of pasture grazing or organic/biodynamic farming system. The aim of this investigation was to further explore the effect of fresh grass in cows’ diet and of organic/biodynamic farming systems on the share of RRR in total PHY in raw cow farm milk in the Netherlands and discuss implications for the verification of farm management system. Materials and methods Thirty farm-tank samples of raw milk were collected from 30 Dutch farms between 7 and 27 July 2011. Farm management and ration information of herds (amounts and proportions of feeds Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 674 in total DM intake) from the week of sampling was obtained by means of questionnaires and interviews. Six farms did not feed any fresh grass to the cows (group NG). The remaining 24 farms fed fresh grass in amounts ranging from 0.36 to 0.94 of total DM intake. Four of those farms fed cut grass indoors (group GI) whereas the remaining 20 farms practised grazing at pasture either during daytime only (8 farms, group Pd) or day and night grazing (12 farms). Of the latter group, there were six farms with a conventional farming system (group Pd+n), and three organic (Org) and three biodynamic farms (BD); jointly named OB. Milk fat was extracted and analysed by gas chromatography-mass spectrometry according to Capuano et al. (2014). Significance of differences was tested between milk samples from cows that had been (GI, Pd, Pd+n and OB) or had not been fed fresh grass (NG), and between organic/biodynamic (OB) and conventional (NG, GI, Pd and Pd+n) farming systems. Correlations were calculated between proportions of feed categories in total DM in cow diets, and the content of PHY and proportion of RRR in total PHY. Results and discussion The average PHY content of the milk was 1.5 mg g fat-1 with values ranging from 0.6 to 3.2 mg g fat–1 (Capuano et al., 2014). The distribution of PHY was significantly different only between milk from continuous stocking (Pd+n, highest) and fresh-cut grass indoors (GI, lowest, not shown). The concentration of PHY was not higher in milk from organic / biodynamic systems than in conventional milk, nor was it correlated with the relative or absolute amount of fresh grass or any other forage or feed type or combination of feed types in the daily ration (Capuano et al., 2014). These results did not confirm our hypothesis and contrast with findings reported by other authors. It has been repeatedly reported that fresh grass or grass-silage increase the PHY content in cows’ plasma and milk because of the higher level of phytol (derived from chlorophyll) in fresh grass compared with maize silage, hay or concentrate (e.g., Baars et al., 2012). However, the phytol content of the diet can vary among and within feeds. In grassland, the botanical composition can play a role, as well as sward structure: grazing cows select young leaves at the top of the canopy, whereas cut grass contains more older leaves and pseudostems that contain less chlorophyll (Elgersma et al., 2003). Che et al. (2013) found a negative correlation between the amount of pasture and the level of PHY in milk, but a positive correlation between the proportion of grazed legumes in DM intake and the share of RRR in total PHY, which ranged from 0.24 in a sample whith no legumes to 0.27 to 0.49 when the legume proportion ranged from 0.08 to 0.24 in the cow diet. In our study, the proportion of RRR in total PHY ranged from 0.14 to 0.70 (Figure 1) and was lower (P<0.05) in group NG (0.20) than in groups Pd+n (0.41), OB (0.45) and BD (0.50). Individual organic and biodynamic farms were rather similar, except one BD farm. RRR share data were non-normally distributed; the Spearman correlation coefficient between RRR share and the share of fresh grass was 0.803. The average RRR share in milk from grazing cows (0.39) was higher than from cows indoors (0.24). In line with this, the SRR/RRR-diastereomers ratio decreased as the amount of grass products in the diet increased, and it was numerically lower (P =0.057) in the organic/BD milk compared with the conventional milk samples in this study (Capuano et al., 2014). In another study (Capuano et al., 2013), based on triacylglycerol profile, the milk from cows that had fresh grass in the daily ration could be distinguished from milk from cows that had no fresh grass with sensitivity and specificity values >85%, but authentication of pasture grazing and of organic/biodynamic farming proved difficult during the grazing period. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 675 Share of RRR in total phytanic acid 0.8 y = 0.3x + 0.18 R2 = 0.59 P < 0.05 0.7 NG 0.6 GI 0.5 Pd 0.4 Pd+n 0.3 Org 0.2 BD 0.1 0.0 0.0 0.2 0.4 0.6 0.8 1.0 Share of fresh grass in dry matter intake Figure 1. Proportion of diasteromer 3R,7R,11R (RRR) in total phytanic acid in individual raw farm milk samples: NG, no fresh grass G; GI, cows indoors with cut fresh grass; Pd, daytime grazing; Pd+n, day + night grazing; OB, Pd+n in an organic (Org) or biodynamic (BD) farming system, in relation to the proportion of fresh grass in the cows’ diet. Conclusion Concentrations of PHY were not significantly correlated with the amount or proportion of fresh grass nor with the total grass-based forages in the animal diet. PHY cannot be used to authenticate pasture feeding or organic/biodynamic management systems. In contrast, the share of RRR in total PHY showed a good correlation with the proportion of fresh grass in the cow’s daily ration. PHY diastereomers might be used for the authentication of fresh grass feeding and/or pasture grazing, along with other biomarkers. References Baars T., Schröder M., Kusche D. and Vetter W. (2012) Phytanic acid content and SRR/RRR diastereomer ratio in milk from organic and conventional farms at low and high level of fodder input. Organic Agriculture 2, 13-21. Capuano E., Boerrigter-Eenling R., Elgersma A. and van Ruth S.M. (2013) Effect of fresh grass feeding, pasture grazing and organic/biodynamic farming on bovine milk triglyceride profile and implications for authentification. European Food Research and Technology (doi 10.1007/s00217-013-2137-0). Capuano E., Elgersma A., Tres A., van Ruth S.M. (2014) Phytanic and pristanic acid content in Dutch farm milk and implications for the verification of the farming management system. International Dairy Journal,35, 21-24. Che B.N., Kristensen T., Nebel C., Dalsgaard,T. K., Hellgren L. I., Young J. F. and Larsen M.K. (2013). Content and distribution of phytanic acid diastereomers in organic milk as affected by feed composition. Journal of Agricultural and Food Chemistry 61, 225-230. Elgersma A. (2012) New developments in The Netherlands: dairies reward grazing because of public perception. Grassland Science in Europe 17, 420-422. Elgersma A., Tamminga S. and Ellen G. (2003) Effect of grazing versus stall-feeding of cut grass on milk fatty acid composition of dairy cows. Grassland Science in Europe 8, 271-274. Schröder M. and Vetter W. (2011). GC/EI-MS determination of the diastereomer distribution of phytanic acid in food samples. Journal of the American Oil Chemists’ Society 88, 341−349. Vetter W. and Schroder M. (2010) Concentrations of phytanic acid and pristanic acid are higher in organic than in conventional dairy products from the German market. Food Chemistry 119, 746−752. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 676 Potential of fertilized grass clover swards to produce adequate herbage to support dairy cow milk production in high stocking rate grass based systems Egan M.J.1,2, Lynch M.B.2 and Hennessy D.1 1 Teagasc, Animal and Grassland Research Innovation Centre, Moorepark, Fermoy, Co. Cork, Ireland 2 UCD School of Agriculture and Food Science, University College Dublin, Belfield, Dublin, Ireland. Corresponding author: Michael.Egan@teagasc.ie Abstract White clover (Trifolium repens L.; WC) is beneficial in grass swards as it can fix atmospheric nitrogen (N) and make it available for grass growth. This experiment compared herbage and milk production from a perennial ryegrass (Lolium perenne L.; PRG) sward receiving 250 kg N/ha/yr (Gr250), a PRG-WC sward receiving 250 kg N/ha/yr (Cl250), and a PRG-WC sward receiving 150 kg N/ha/yr (Cl150) in an intensive grazing system. Forty-one dairy cows were allocated to graze each sward (n = 14). Sward WC content was similar in Cl150 and Cl250 (24.4 and 21.2%, respectively). Cows grazing Cl250 had higher (P<0.001) total milk yield than cows grazing Cl150 and Gr250 (6107, 5908 and 5757 kg/yr, respectively). Cows grazing Cl250 and Cl150 had higher (P<0.001) total milk solids yield compared to Gr250 (479, 476, 451 kg/yr, respectively). Increased milk production on the clover swards occurred after June. Treatment had no effect on pre-grazing herbage mass. It is concluded that including WC in grass swards can result in an increase in milk production. High N-fertilizer application (250 kg N/ha) did not reduce annual sward WC content compared to Cl150. Keywords: Trifolium repens L., nitrogen, dairy cow and milk production Introduction White clover (Trifolium repens L.; WC), a legume, is beneficial in grass swards as it can fix atmospheric N and make it available for grass growth. White clover and perennial ryegrass (Lolium perenne L.; PRG) have different temperature responses and different seasonal growth patterns (Davies, 1992). As a consequence of low WC growth rates before June/July in temperate regions of the northern hemisphere, the strategic use of nitrogen (N) fertilizer on grass WC swards in spring is commonly practised to increase herbage production relative to swards reliant solely on N fixation. Nitrogen fertilizer application can reduce sward WC content (Harris et al., 1996). Research has shown the benefit of WC over PRG for milk production, particularly in the second half of the year (July onwards) (Riberio Filho et al., 2003; Egan et al., 2013). The objective of the current study was to compare herbage production of and milk production from a grass-only sward receiving 250 kg N/ha with grass WC swards receiving 150 or 250 kg N/ha. Materials and methods A farm systems experiment was established at Teagasc, Animal and Grassland Research Innovation Centre, Moorepark, Fermoy, Co. Cork, Ireland. Forty-two spring-calving Holstein Friesian dairy cows (33 multiparous and 9 premiparous) were blocked on calving date, preexperimental milk yield and parity, and randomly allocated to one of the three treatments (n=14), a PRG-only sward receiving 250 kg N/ha/yr (Gr250), a PRG-WC sward receiving 250 kg N/ha/yr (Cl250), and a PRG-WC sward receiving 150 kg N/ha/yr (Cl150), from 17 February to 17 November 2013. All treatments were stocked at 2.74 LU/ha. Fertilizer N application was the same across treatments until May, after which N application to Cl150 was reduced. Herbage was allocated daily to achieve a target post-grazing sward height of 4 cm. Pre-grazing herbage Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 677 mass (>4 cm; HM) was determined twice-weekly, using an Etesia mower (Etesia UK Ltd., Warwick, UK). Pre- and post-grazing sward heights were measured daily using a rising plate meter (Jenquip, Feilding, New Zealand). Sward WC content for the clover treatments was quantified once in each paddock prior to each grazing as described by Egan et al. (2013). Milk yield was recorded daily and milk composition (fat and protein concentrations) was measured weekly. Milk solids (MS) yield was calculated as the sum of milk fat and protein yields. Data were analysed using PROC MIXED in SAS with terms for treatment, time (week or rotation) and the associated interaction. Fixed terms were treatment and week or rotation, and random terms were cow and paddock. Results and discussion There was a treatment × week interaction effect (P<0.001) on daily milk yield, MS yield and milk fat content (Figure 1). Figure 1. Effect of sward type on daily milk solids production for cows grazing a Grass only receiving 250 kg N/ha (Gr250) and grass clover swards receiving 150 kg N/ha and 250 kg N/ha (Cl150 and Cl250, respectively) for each week of experiment (17 February to 17 November). All treatments had similar milk yield until experimental week 12 (May) after which Cl250 had higher milk yield compared to the other two treatments until week 16. Milk yield was similar from week 16 to week 19 for all treatments. From week 20 (June) to the end of the experiment, milk yield was greater on Cl250 than on the other treatments; and from week 24 (July) milk yield was higher on Cl150 than GR250 and similar to Cl250. This trend in milk yield was similar to that reported by Riberio Filho et al. (2003) and Egan et al. (2013). A similar trend was seen for MS yield. The Cl150 treatment had lower MS yield from week 5 to 8; Gr250 had lower MS yield than both clover treatments from week 20 to 25; all treatments had similar MS yield in weeks 26 and 27, and Gr250 had lower MS yield than the two clover treatments from week 28 to 36, after which there was no significant difference between treatments. Daily milk fat content was similar on all treatments at the beginning of the experiment. The Cl150 treatment had higher daily milk fat content compared to the other treatments in weeks 9, 12, 14 and 15 and from week 20 (June) to week 35 (October). Treatment had no effect on daily milk protein. Treatment had an effect on cumulative milk yield (Table 1); Cl150 had lower (P<0.05) cumulative milk yield compared to Cl250 (5908 and 6107 kg milk/cow, respectively); there was no significant difference between Cl150 and Gr250 (5908 and 5757 kg milk/cow, respectively); and Cl250 had greater (P<0.001) cumulative milk yield than Gr250 (6107 and 5757 kg milk/cow, respectively). There was no significant difference in cumulative MS yield between clover treatments. The clover treatments had greater (P<0.001) cumulative MS yield than Gr250 (Table 1). Treatment had no significant effect on pre-grazing HM (Table 1). Sward WC content was affected by the treatment × week interaction (P<0.01). Sward WC content was similar between clover treatments from February to July, after which Cl150 had higher WC content for the remainder of the year, which coincided with the reduction in N fertilizer Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 678 application to Cl150. Ledgard and Steele (1992) found a similar effect of reduced N fertilizer application on sward WC content. There was no significant difference in the average annual sward WC content between Cl150 and Cl250 (Table 1). Rotation had a significant effect on sward WC content (P<0.001) which increased from 0.05 g/kg in February to a peak of 0.43 and 0.33 g/kg in Cl150 and Cl250, respectively, in August. Table 1. Daily and cumulative milk production and average pre grazing herbage mass on grass-only swards receiving 250 kg N/ha (Gr250) and grass-clover swards receiving 150 kg N/ha and 250 kg N/ha (Cl150 and Cl250, respectively) and average sward clover content on Cl150 and Cl250. ( 1S.E. = Standard Error; 2TRT = Treatment; 3 Week = Week of Experiment) Cl150 Cl250 Gr250 S.E.1 TRT TRT×Week Milk yield (kg/cow/d) 21.10 21.81 20.56 0.28 <0.001 <0.001 Milk solids (kg/cow/d) 1.70 1.71 1.61 0.02 <0.001 <0.001 Milk fat (g/kg) 4.60 4.41 4.39 0.42 <0.05 <0.001 Milk protein (g/kg) 3.59 3.59 3.58 0.04 NS NS Cumulative milk yield (kg/cow) 5908 6107 5757 78.40 <0.001 - Cumulative milk solids (kg/cow) 476 479 451 5.60 <0.001 - Pre-grazing herbage mass (kg DM/ha) 1400 1370 1575 90.54 NS 0.07 Clover content (%) 24.4 21.2 - 1.95 NS <0.01 Conclusions White clover had a positive effect on milk production regardless of N fertilizer application rate, especially in the second half of lactation. There was no effect of WC inclusion into grass swards on total herbage production. Reducing N fertilizer application in mid-summer increased sward WC content compared to the higher level of N fertilizer. Including WC in grass swards can result in an increase in milk production. Acknowledgements This experiment was funded through the Irish Farmers Dairy Levy Trust and the Teagasc Walsh Fellowship Scheme. References Davies A. (1992) White Clover. Biologist - Institute of Biology 39, 129-133. Egan M.J., Lynch M.B. and Hennessy D. (2013) The influence of white clover inclusion in perennial ryegrass swards on milk and herbage production in a high N fertilizer system. Ireland: Proceedings of the Agricultural Research Forum 81. Harris S.L., Clark D.A., Waugh C.D and Clarkson F.H. (1996). Nitrogen fertilizer effects on white clover in dairy pastures. Agronomy Society of New Zealand Special Publication No. 11 / Grassland Research and Practice Series No. 6. Ledgard S.F. and Steele K.W. (1992) Biological nitrogen fixation in mixed legume/grass pastures. Plant and Soil 141, 137-153. Riberio Filho H.M.N., Delagarde R. and Peyraud J.L. (2003) Inclusion of white clover in strip-grazed perennial ryegrass swards: herbage intake and milk yield of dairy cows at different ages of sward regrowth. Animal Science 77, 499-510. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 679 Feeding strategies and feed self-sufficiency of dairy farms in the lowland and mountain area of Western Switzerland Ineichen S.1, Piccand V.1, Chevalley S.1, Reidy B.1 and Cutullic E.2,3 1 School of Agricultural, Forest and Food Sciences HAFL, Bern University of Applied Sciences, Switzerland 2 INRA, UMR 1348 PEGASE, 35590 St-Gilles, France 3 Agrocampus-Ouest, UMR 1348 PEGASE, 35000 Rennes, France Corresponding author: beat.reidy@bfh.ch Abstract The feeding strategy and the feed self-sufficiency of 34 commercial dairy farms in the western part of Switzerland were investigated throughout the year 2010. Feed ration composition of farms from the lowlands with relatively high proportions of maize were compared to farms from the mountain area with relatively high proportions of herbage in the rations. Dairy feeds were categorized into herbage, whole-plant maize and concentrate. The proportions of dry matter (DM), net energy for lactation (NEL) and metabolizable protein (MP) sourced from herbage, maize and concentrates were calculated. Dairy farms located in the mountain area had higher proportions of herbage DM in the ration than farms located in the lowland (70% vs. 51 %), as well as herbage-sourced NEL (66 % vs. 47%) and herbage-sourced MP (68% vs. 49%). In contrast, lowland farms showed a higher proportion of maize DM in the ration (31% vs. 17 %). The degree of self-sufficiency of the total feed ration of mountain farms was shown to be larger for DM, NEL and especially MP. A general negative relationship between the proportion of maize DM in the ration and the dairy herd MP self-sufficiency could be observed. Keywords: feed self-sufficiency, feeding strategies, roughage, herbage Introduction Grassland-based milk production is of major importance in Switzerland. Traditionally, herbage provides the largest proportion of dairy feeds. Due to climatic and topographic restrictions, maize-based dairy systems are usually limited to the lowland area. Inclusion of higher proportions of maize in the rations consequently decreases the proportion of herbage, which in turn requires the purchase of protein rich concentrates. This feeding strategy increases environmental concerns regarding the importation of soy bean meal and the accumulation of excess nitrogen in dairy farms’ environment, decreasing the farms’ feed self-sufficiency. On the other hand, feeding strategies relying primarily on the utilization of herbage from grassland are considered environmentally sound and sustainably beneficial. However, with increasing specialization and intensification of agricultural production, grassland-based dairy farms tend to decrease in number in Switzerland. A recently established program “Grassland-based milk and meat production” of the Swiss Federal Office for Agriculture (Schweizerischer Bundesrat 2013), aims to promote the utilization of herbage from grassland and to counteract the tendency of a decreasing feed self-sufficiency of dairy farms. This preliminary field study was conducted to obtain a present overview about the differences in the composition of the feed rations and the degree of feed self-sufficiency of dairy farms in the lowland and mountain area of Western Switzerland. Material and methods Thirty-four dairy farms were randomly selected within the cantons of Neuchatel and Waadt (lowland: 22; mountain and hills: 12). The feed rations were investigated throughout the year of 2010. The dry matter (DM) consumption of every dairy herd was modelled on a monthly basis according to herd structure and cow characteristics (parity, live weight, lactation stage, Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 680 milk yield; Cutullic et al., 2012). The farmer indicated the proportions or fixed amounts of feeds fed to the dairy cows. As land use area, herd structure and milk production were known (extracted from national databases) the consistency between total consumption and available feeds for the herd, and between milk production and nutrients intake, was checked on-farm during the survey. Dairy feeds were categorized as herbage, maize (whole plant) and concentrate. Herbage and maize were additionally categorized as roughage. The relative proportions of DM, net energy for lactation (NEL) and metabolizable protein (MP) sourced from the different feed components were calculated. To estimate the feed self-sufficiency of the farms, feed components were divided into locally grown (origin < 50 km) or foreign (> 50 km) feeds. Results and discussion Lowland and mountain farms had similar area, herd size, total and per cow milk production (Table 1) and DMI per cow and day (Table 2). Table 1. Characteristics of dairy farms in the lowland (L) and mountain (M) area of western Switzerland (mean ± standard deviation): Farm size (ha) Cows per farm Milk production (t ECM produced per year) Milk production (kg ECM/cow) L 46 ± 30 28 ± 16 265 ± 162 25.8 ± 3.9 M 41 ± 20 22 ± 11 193 ± 98 23.4 ± 4.1 Table 2. Composition of feed rations of dairy farms in the lowland (L) and mountain (M) area of western Switzerland: Average proportions of dry matter (DM), net energy for lactation (NEL) and metabolizable protein (MP) from herbage, maize and concentrates. 1 r.s.e: residual standard error ; Significance 2 ***, *, +, n.s.: P < 0.001, 0.05, 0.10, not significant L M r.s.e1 Significance2 19.8 19.1 1.4 n.s. herbage 51 70 12 *** maize 31 17 10 *** roughage 82 86 7 n.s. concentrates 14 12 5 n.s. herbage 47 66 13 *** maize 31 17 10 *** roughage 79 83 8 n.s. concentrates 18 16 6 n.s. herbage 49 68 13 *** maize 22 12 7 *** roughage 71 80 1 * concentrates 26 20 8 + DM intake (kg) per cow and day: % DM from % NEL from % MP from Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 681 The composition of the feed ration, however, followed a clear pattern (Table 3): herbage DM proportion was significantly higher for mountain farms (70%) than for lowland farms (51%), resulting in higher proportions of NEL and MP from herbage. As expected, the proportion of maize DM was significantly higher in rations of lowland farms (31%) than of mountain farms (17%). Some of the mountain farms purchased maize cultivated in the lowland area, which explains the occurrence of maize in the feed ration of mountain farms. Inclusion of a higher proportion of maize DM in the rations of lowland farms significantly increased the contribution of maize NEL in the ration and to a lower extent of maize MP, as maize is a protein-poor forage. This was not compensated by an increased contribution of herbage MP, but by the use of dehydrated lucerne and protein-rich concentrates. Table 3. Degree of self-sufficiency of dairy farms in the lowland (L) and mountain (M) area of Western Switzerland for dry matter (DM), net energy lactation (NEL) and metabolizable protein (MP). 1 r.s.e: residual standard error ; Significance2 **, *, : P < 0.01, 0.05 L M r.s.e1 Significance2 DM 90 94 4 * NEL 88 93 5 * MP 80 88 8 ** Degree of self-sufficiency (%) Conclusions The feeding strategy of mountain farms, which focuses on a maximum utilization of herbage, results in a high feed self-sufficiency. In contrast, the feeding strategy of lowland farms, which include noticeable amounts of maize, results in a reduced feed self-sufficiency. However, in areas with frequent summer droughts, as in the case of the lowland area of western Switzerland, maize cultivation has also been shown to increase the resilience of a feeding system to drought (Mosimann et al., 2013). Acknowledgements The authors acknowledge the surveyed farmers, Prolait Fédération Laitière, Swissherdbook, Fédération suisse d'élevage Holstein, Braunvieh Schweiz, Office fédéral de l'agriculture and Banque de Données sur le Transfert des Animaux, for providing the required data. References Cutullic E., Chevalley S., Thomet P. and Piccand V. (2012) Etat des lieux sur l’affouragement des vaches laitières. Enquêtes sur les exploitations en lait de centrale de Prolait. Mosimann E., Deléglise C., Demenga M., Frund D., Sinaj S. and Charles R. (2013) Disponibilité en eau et production fourragère en zone de grandes cultures. Recherche Agronomique Suisse 4 (11–12), pp. 468 – 475. Schweizerischer Bundesrat (2013) Verordnung vom 23. Oktober 2013 über die Direktzahlungen an die Landwirtschaft (Direktzahlungsverordnung, DZV), art. 70 – 71. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 682 Weather effects and cattle behavioural characteristics Halasz A. and Nagy G. University of Debrecen Agricultural Centre, Department of Applied Economics and Rural Development, HU-4032 Debrecen, 138 Boszormenyi Rd., Hungary halasza@agr.unideb.hu Abstract The behavioural response of Hungarian Grey Cattle (HGC) to environmental factors, when managed under rangeland conditions, provides a good representation of the daily life of beef cattle in the absence of human interference. In addition to ethological observations, we collected zoometeorological information that affect cattle behaviour under natural conditions. We found a very complex interaction (neuro-endocrine complex), which utilizes different fields of science including meteorology, ethology, pharmacology, physics and agronomy. This paper’s aim is to present our research methods and results about environmental effects, in the context of animal behaviour. The programme has focused on animal behaviour, using GPS technology, but also attempted to collect zoometorological data in alternative ways (ion count, barometric measurements). The applied research methods include spatial data recorded with GPS collars, meteorological information (barometric pressure, wind direction, temperature, weather fronts) and field reports about regular behavioural attributes. Keywords: behaviour, cattle, GPS, zoometeorology, ion Introduction Zoometeorology is an interdisciplinary science, merging the principles of ethology and meteorology. The mutual effects of environmental factors are quite complex (Kovacs, 2010); therefore, they need to be examined in a broad context. Earlier zoometeorological observations (Malechek and Smith, 1975) pointed out that grazing cattle react in a variety of ways to different weather conditions. They have proven that the abiotic environment, animal physiology and behaviour are all correlated. Materials and methods The study area is 1191 ha of rangeland. There are two major parts: the North (688 ha) and South (503 ha). We applied the terminology of Czakó et al. (1985) and Kilgour (2012) to describe the animal behaviour and organize the behavioural traits into 3 main groups. The feed intake actions (grazing, rumination, drinking), sexual and calf-care actions (copulation, nursing) and social actions (fighting, play, moving). Two types of GPS receivers have been used (Snewi Trekbox, Bluetooth, GT-750 GPS data logger) to describe the animals' spatial position and calculate the speed, daily travel distance and the time spent standing. The loggers recorded animal daily route for 5 days. The positional data have been transformed to digital mapping. Animal behaviour has been observed periodically, every 20 minutes, with recordings of approximately 5 seconds for each. The most typical behaviour pattern was logged. We used binoculars to observe cattle. It was necessary to avoid bothering animals and keep out the flight-zone (average 50 m before cattle change behaviour). The typical behavioural patterns, (grazing, walking, fighting, calf care) were recorded with digital video camera. We also sub-categorized behavioural traits for statistical analyses (see Table 1). Meteorological data were collected from the national meteosurvey database and we have also made local measurements (air pressure, temperature, air ions). Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 683 Table 1. Categorization of Hungarian Grey cattle field behaviour. Ethological category Practical category Feed intake/Digestive/Nutrient cycle Feed intake Behavioural trait Grazing Grazing on the move Ruminating Drinking Excreting Walking Locomotion/Posture Lying Agonistic/Fight Social Grooming Social behaviour Watch Mounting/Mounted Sexual and Parental behaviour Sexual Suckling Results There is a significant correlation (r=0.491, P<0.01) between weather-front-induced barometric pressure and the length of the herd’s daily route. The front-free and cold-front weather systems create high air pressure (P≥1005 hPa at converted altitude) which results in calmer behaviour. The relaxed cattle spend more time feeding. Seeking fresh, nutritious grass is a natural, herbivore behaviour (Gere, 1977); therefore a non-stressed herd’s pasture-activity 8 of 10 of time is spent feeding. Results show that warm-weather front – low air pressure (P≤1005 hPa) – causes more stress because of the changing (dropping) air pressure effect on the parasympathetic nervous system (Kovács, 2010) and often combined with high temperature and humidity. The stressed animals gather, spend more time in shade at nearby water-source (riverbank) or continuously roaming on pasture and finally do not complete their average daily route (6-8 km; Haraszti, 1977). We have concluded that a connection exists between stable weather systems and increased feeding activity. The stable weather means high barometric pressure builds up (above 1005 hPa). These conditions occur during cold fronts and the time when no front is present. During such weather, cattle cover more distance to browse for grass (Table 2). Table 2. Distribution of behavioural traits during different weather system from 2010 to 2013 Front types No Front Warm front Cold front Feed intake 93% 47% 95% Sexual 2% 10% 0% Social 5% 43% 5% 100 % (368 pc) 100 % (423 pc) Behavioural trait Gross count of behavioural 100 % (1049 pc) traits Source: Own calculation; n=10 – marked cattle Discussion The network of physiological processes (serotonin reuptake response, thermoregulation) have created an amazingly flexible bio-system (Phelps, 2005), which eventually drives the cattle Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 684 behaviour. We suggest that the incoming weather systems change the air ion concentration, which results with serotonin neuro-transmitter stress. Throughout this process the animal becomes more frustrated (after the short period of serotonin indicated calm period) and spend less time at grazing and with comfort behaviour. The changing weather also has an effect on barometric pressure, which is easier to measure in time-lapse or real-time. Our results show that during calm weather/high pressure periods the herd behaves more balanced than in periods with low pressure/warm-fronts. Acknowledgement This work and its publication is supported by the TÁMOP-4.2.2/B-10/1-2010-0024 project. The project is co-financed by the European Union and European Social Fund. References Czakó J., Keszthelyi T. and Sántha T. (1985) Etológia Kislexikon (Ethological handbook), Natura kiadó, ISBN 963 233 113 3; 32. Gere Tibor (1977) Néhány tartástechnológiai tényező hatása a szarvasmarha viselkedésére (Technological factors’ effect on cattle behaviour). Különlenyomat az Agrártudományi Egyetem Gödöllő 1977. évi közleményeiből. Haraszti E. (1977) Az állat és a legelő (Livestock and pasture). 2. kiadás, Mezőgazdasági Kiadó, Budapest. ISBN63 230 196X, pp. 120-135. Kilgour R.J. (2012) In pursuit of 'normal': A review of the behaviour of cattle at pasture. Applied Animal Behaviour Science 138, 1-11. Kovács A. (2010) Agrometeorológiai és klimatológiai alapismeretek (Basics of Agrometeorology and climatology). 8. fejezet (chapter), Zoometeorológia (zoometeorology), 263-290 o. ISBN:978-963-286-598-0. Malechek J.C. and Smith B.M. (1975) Behavior of range cows in response to winter weather. Journal of Range Management 29, 9-12. Phelps J.M.D. (2005) From stress to genes, from mind to molecules. Chapter 9. http:// www.psycheducation.org/mechanism/9TrophicFactors.htm Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 685 Relationship between the composition of fresh grass-based diets and the excretion of dietary nitrogen from dairy cows Moorby J.M. IBERS, Aberystwyth University, Gogerddan, Aberystwyth, SY23 3EE, UK Corresponding author: jxm@aber.ac.uk Abstract In studies carried out in Aberystwyth, dry and lactating dairy cows were individually fed diets based on fresh grass or fresh grass-clover mixtures that differed in concentrations of crude protein (nitrogen (N) × 6.25) and water-soluble carbohydrates (WSC). The apparent partitioning of dietary N into milk, faeces, and urine was measured and the relationships between diet composition and N excretion were investigated. Relatively poor correlations between N intake and output of N in milk and faeces were found, but a strong split-line relationship between N intake and excretion of N in urine was detected (R2 = 0.79): up to an intake of 397 g N d-1, mean urine N output was 81 g d-1, and above this the slope of the relationship was 0.786 g g-1. Similarly, a strong negative split-line relationship was found between the whole diet WSC/N ratio and the apparent excretion of dietary N in urine. Below a ratio of 8.94, urine N/feed N = -0.0416 × diet WSC/N, and above this mean urine N/feed N was 0.199 g g-1 (R2 = 0.76). It is concluded that the ratio of diet WSC and N concentrations in fresh grass- and grass-clover-based diets could help predict the losses of N in urine. Keywords: grass, water-soluble carbohydrates, dairy cows, nitrogen excretion Introduction Significant but variable quantities of nitrogen (N) are excreted in the urine of dairy cows, where it contributes to pollutants such as nitrous oxide, ammonia and nitrates (McGechan and Topp, 2003). Leached losses of N from grazing pastures primarily originate from urine patches, and Li et al. (2012) suggest better knowledge of urinary-N deposition in urine would enable improved prediction of leached N losses from soil. While a key factor determining the excretion of N from dairy cows is N intake (Kebreab et al., 2001), the apparent partitioning of dietary N between productive purposes (i.e. milk protein) and excretion is influenced by diet composition and the effective use of N by rumen microbes. Increasing rates of N fertilization increases grass N concentration, but this does not necessarily lead to increased duodenal N flow in animals that eat the grass (Peyraud and Delaby, 2006). Surplus diet N is lost in urine following absorption as ammonia from the rumen when the rumen population has insufficient energy sources to capture it. Increased concentrations of grass water-soluble carbohydrates (WSC) have been shown to reduce the proportion of dietary N that is excreted in urine (Miller et al., 1999; Miller et al., 2001; Moorby et al., 2006), but at some point ammonia absorbed from the rumen will be minimal and further increasing grass WSC concentrations will have no additional benefit. This work investigated the relationships between the nutritional characteristics of fresh ryegrassbased diets and the excretion of N from dairy cows. Materials and methods A number of experiments have been carried out in Aberystwyth over the last 15 or so years in which individual dairy cows were offered zero-grazed diets based on ad libitum access to fresh grass or mixtures of fresh grass and white clover (Miller et al., 1999; Miller et al., 2001; Moorby et al., 2006; and J.M. Moorby, unpublished data). Diets were designed to create differences in the WSC and N concentrations of forage, and were supplemented with relatively small amounts (3-4 kg d-1 per cow) of concentrate feed. Data from 55 whole-body N partitioning measurements carried out to determine the apparent partitioning of feed N into milk, urine and faeces were Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 686 available. Briefly, intakes of fresh feeds were recorded and total collections of faeces and urine were carried out for 6 days. Concentrations of N in feeds, milk, faeces and urine were determined to allow calculation of daily inputs and outputs of N (g d-1), and apparent N partitioning was calculated by relative differences in intake and outputs (g output g-1 intake). Individual cow data were analysed by correlation analysis and split-line (‘broken stick’) regression using Genstat (15th Edition, VSN International Ltd, UK). Results and discussion Diet dry matter (DM) intakes and milk yields varied widely among the experiments (Table 1), and were largely dependent on the stage of lactation of the dairy cows used, which ranged from early to late lactation. Whole-diet concentrations of N and WSC varied by a factor of 2 across the experiments (whole diet crude protein ranged from 113 to 228 g kg-1 DM). Apparent outputs of diet N in milk and urine both varied by a factor of 3.6. There was a relatively poor correlation between diet N intake and outputs of N in milk (R 2 = 0.29) and faeces (R2 = 0.45). However, there was a significant (P < 0.001) split-line relationship between diet N intake and urine N output (adjusted R2 = 0.79; Figure 1). There appeared to be little relationship between N intake and urine N output up to an intake of 397 (s.e. 15.6) g N d1 with mean urine N output at 81 (s.e. 6.9) g d-1, but at higher N intakes the relationship was urine N output = 0.786 (s.e. 0.0775) × diet N intake (g d-1). A strong relationship between N intake and N excretion is well know (Kebreab et al., 2001), and in this case, above dietary crude protein intakes of approximately 2.5 kg d-1, more than 78% of dietary N was apparently excreted in urine. Table 1. Mean and standard deviation (SD) and range of selected parameters of individual dairy cow measurments (n = 55) in which the apparent partitioning of feed N was measured Mean SD Min Max Dry matter intake (kg d-1) 17.7 1.8 12.2 20.5 -1 Diet N (g kg DM) 24.9 5.2 18.0 36.5 Diet WSC (g kg-1 DM) 160 38 103 217 WSC intake (g d ) 2853 855 1297 4381 N intake (g d-1) 443 108 256 667 6.73 2.21 3.21 10.86 Milk yield (kg d ) 24.0 8.3 9.6 40.0 Milk N out (g d-1) 124 37 63 200 138 68 54 300 Milk N out/Feed N in (g g ) 0.29 0.07 0.12 0.43 Urine N out/Feed N in (g g-1) 0.30 0.09 0.14 0.50 -1 Diet WSC/N (g g-1) -1 -1 Urine N out (g d ) -1 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 687 Figure 1. Relationship between diet N intake and urine N output in individual cows fed diets based on fresh grass Figure 2. Relationship between whole diet ratio of WSC to N and apparent excretion of feed N in urine from dairy cows (n = 55). Dietary intake of WSC was poorly correlated with daily outputs of N in milk (R2 = 0.27), urine (R2 = 0.11) and faeces (R2 = 0.01). However, the ratio of WSC to N concentrations in the whole diet (g g-1) was negatively related to the apparent excretion of feed N in urine in a significant (P < 0.001) linear split-line relationship of urine N/feed N = -0.0416 (s.e. 0.0041) × diet WSC/N up to a diet WSC/N ratio of 8.94 (s.e. 0.457) g g-1. After the breakpoint, the apparent excretion of dietary N in urine was 0.199 (s.e. 0.0125) g g-1. The adjusted R2 (0.77) of this relationship was marginally better than the coefficient of determination of a standard linear regression (R2 = 0.76). A negative relationship between diet WSC/N and apparent urinary excretion of dietary N has been noted before (Edwards et al., 2007), using some of the current data. The current study builds on previous (and different) data and indicates that the negative relationship holds when including clovers in the diet of the cows. The split-line relationship also suggests that beyond a ratio of WSC to N concentrations in ryegrass-based diets of approximately 9 there is little additional benefit to increasing the concentration of WSC in terms of reducing the apparent excretion of dietary N in urine. Further work is required to investigate whether the relationship is maintained at higher ratios of WSC to N because the current dataset is relatively small at this end of the range. Conclusion In dairy cow diets based on grazing, grass WSC concentrations play an important role in helping rumen microbes capture grass N. The current results suggest that at typical grass crude protein concentrations between about 160 and 200 g kg-1 DM (about 26 to 32 g N kg-1 DM), mean WSC concentrations of between about 230 and 290 g kg-1 DM would minimize rates of urine N excretion in grazing dairy cows. As a target for ryegrass breeding programmes, this should be relatively easily achieved. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 688 Acknowledgements The author gratefully acknowledges the original funding sources for this work that includes the Defra, DairyCo., EBLEX, Germinal Holdings Ltd, and the European Community: EU Framework V project QLK-CT-2001-0498 (SweetGrass), and the Seventh Framework Programme under the grant agreement n° FP7-244983 (MultiSward). References Edwards G.R., Parsons A.J., Rasmussen S. and Bryant R.H. (2007) High sugar ryegrasses for livestock systems in New Zealand. Proceedings of the New Zealand Grassland Association 69, 161-171. Kebreab E., France J., Beever D.E. and Castillo A.R. (2001) Nitrogen pollution by dairy cows and its mitigation by dietary manipulation. Nutrient Cycling in Agroecosystems 60, 275-285. Li F.Y., Betteridge K., Cichota R., Hoogendoorn C.J. and Jolly B.H. (2012) Effects of nitrogen load variation in animal urination events on nitrogen leaching from grazed pasture. Agricultural Ecosystems & Environment 159, 81-89. McGechan M.B. and Topp C.F.E. (2004) Modelling environmental impacts of deposition of excreted nitrogen by grazing dairy cows. Agricultural Ecosystems & Environment 103, 149-164. Miller L.A., Moorby J.M., Davies D.R., Humphreys M.O., Scollan N.D., MacRae J.C. and Theodorou M.K. (2001) Increased concentration of water-soluble carbohydrate in perennial ryegrass (Lolium perenne L.): Milk production from late-lactation dairy cows. Grass and Forage Science 56, 383-394. Miller L.A., Theodorou M.K., MacRae J.C., Evans R.T., Adesogan A.T., Humphreys M.O., Scollan N.D. and Moorby J.M. (1999) Milk production and N partitioning responses in dairy cows offered perennial ryegrass selected for high water soluble carbohydrate concentrations. S. African Journal of Animal Science 29, 281-282. Moorby J.M., Evans R.T., Scollan N.D., MacRae J.C. and Theodorou M.K. (2006) Increased concentration of water-soluble carbohydrate in perennial ryegrass (Lolium perenne L.). Evaluation in dairy cows in early lactation. Grass and Forage Science 61, 52-59. Peyraud J.L. and Delaby L. (2006) Grassland management with emphasis on nitrogen flows. Chapter 6, pp 103123 in Fresh Herbage for Dairy Cattle. Eds. Elgersma A., Dijkstra J. and Tamminga S. Springer. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 689 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 690 Theme 5 ‘MultiSward’ Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 691 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 692 Theme 5 invited paper Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 693 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 694 Multi-species swards and multi scale strategies for multifunctional grassland-base ruminant production systems: An overview of the FP7MultiSward project Peyraud J.L.1, van den Pol–van Dasselaar A.2, Collins R.P.3, Huguenin-Elie O.4, Dillon P.5, Peeters A.6 1 INRA, UMR-1348, Joint Research Unit PEGASE, F-35590 St Gilles, France. 2 Wageningen UR Livestock Research, PO Box 65, 8200 AB Lelystad, the Netherlands. 3 Institute of Biological, Environmental and Rural Sciences, Aberystwyth University, UK 4 Agrosocpe, Institute for Sustainability Sciences, CH-8046 Zürich 5 Teagasc, Animal & Grassland Research and Innovation Centre, Moorepark, Fermoy, Co. Cork, Ireland 6 RHEA Research Centre, Rue Warichet 4 Box 202, 1435 Corbais, Belgium Corresponding author: jean-louis.peyraud@rennes.inra.fr Abstract MultiSward aimed to conceive, evaluate and promote sustainable ruminant production systems based on the use of grasslands with a high level of multi-functionality by the concerted use of diverse multi-species swards, diverse plant communities at farm and landscape levels, and diverse production systems. It was demonstrated that the use of sown multi-species swards combining legumes and grasses as well as shallow and deep-rooting species are as productive as fertilized grasses; they allow high levels of animal performance and these results are not conflicting with the delivery of a range of services. For permanent grassland, differentiated grassland management and climate variables influencing functional diversity criteria linked to ecosystem services were identified. Several innovations in grazing management were proposed to enhance the benefits of grassland-based systems. Different operational indicators sets easily implemented at farm and at landscape levels were developed and demonstrate the contribution of grassland-based systems and effects of management on the provision of food and environmental services including the maintenance of farmland biodiversity. Modelling of several policy options and price hypothesis showed that extensive grassland abandonment may be prevented by area payment supporting grassland use, whereas lower Common Agricultural Policy support payments and higher commodity prices may cause less control of land use. In conclusion, MultiSward provided a detailed EU-wide and multifunctional-orientated overview of grassland-based ruminant production systems thus opening the opportunity to define, on a new basis, the contribution of ruminant production on grassland to economic returns for farmers, biodiversity conservation and provision of ecosystem services. This will lead to a better consideration and acceptance of these ruminant production systems by EU citizens and farmers, and will contribute to secure optimal European grassland acreage. Keywords: multi-species swards, grazing, ruminants, competitiveness, public goods, public policy Introduction Grasslands are the main survival resource for about one billion people worldwide. In industrialised Europe grasslands cover some 39% of the agricultural area and form the basis of a strong ruminant livestock sector (Huyghe et al., 2014). Grasslands are not only a relatively cheap source of feed for ruminants, but are also increasingly being recognised for their contribution to the conservation of biodiversity, to the regulation of physical and chemical fluxes in ecosystems, to the mitigation of pollution and to the production of landscape amenity (De Vliegher et al., 2014). The relative importance of the multiple functions provided to society by grasslands varies depending on regional contexts and grassland type and are strongly Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 695 influenced by the type of management: extensive, species-rich and nutrient-poor systems result in less ‘provisioning’ services but are more effective at providing other types of services than intensive grassland-based systems. These functions are well recognized and appreciated by relevant stakeholders as shown by an on-line questionnaire developed during the MultiSward project (Van den Pol-van Dasselaar et al., 2014). Grassland acreage has been significantly reduced in Europe during the last thirty years (by approximately 15 M ha, i.e. 30%) in favour of the production of fodder maize and other annual crops ; even marginal grasslands tend to be abandoned, particularly in mountainous and Mediterranean areas. This was the consequence of farming systems intensification, specialization of production, decrease in cattle population, price support, premium systems and farm size growth (Huyghe et al., 2014) whereas in the same time the production of public goods in not (or poorly) remunerated by the market or the public policy. At the same time the simplicity of managing grass monocultures and the low price of mineral nitrogen have inhibited the use of legumes for forage production. EU Common Agricultural Policy (CAP 1992 with the premium to cereals) has helped shape the current decrease of Europe’s grassland areas (Peyraud et al., 2012). However, new opportunities recently appeared for grassland based systems. Rising global demand for meat and milk, European concerns about environment preservation and food quality and safety favour the increasing role of sustainable grassland-based ruminant systems in the future. Climate change mitigation policies could support the conversion of arable land to grassland for carbon sequestration. After the CAP reform in 2003 cross-compliance rule on the protection of permanent grassland area, rural development expenditures and less Favored Areas payments are a priori more favourable tools for the maintenance of the permanent grassland. In the same time, farming practices can have both positive and negative externalities and too little is known about how well the different grassland management systems and their localisations perform in delivering ecosystem services. This clearly stresses the necessity of developing comprehensive studies of the influence of different grassland management strategies in different local conditions on the positive and negative externalities of the production from field to landscape level. In this context, the main ambition of MultiSward is to conceive, evaluate and promote sustainable ruminant production systems based on the use of grasslands with a high level of multifunctionality in order to optimize economic efficiency, the provision of environmental goods and biodiversity preservation. MultiSward considered multi-species swards (MSS) because in fertile agrosystems MSS can reduce energy consumption by replacing highly energy demanding nitrogen (N) fertilizer by natural nitrogen fixation, whilst maintaining biomass production (Lüscher et al., 2014) and in relatively species-rich and nutrient-poor systems, the provision of ecosystems services is generally enhanced by species diversity. To achieve this ambition, the main objectives were to (i) identify and analyse the effects of socio-economic and policy scenarios on the future of grassland acreage and identify under which policy conditions sustainable grassland systems will become a viable option for farmers compared to crop-based systems; (ii) assess and optimize the performance of MSS to enhance the competitiveness of grassland based production systems and the provision of regulating and supporting services; (iii) design and evaluate innovations in grazing and animal management (including animal genetics) to assess the best way of combining high production efficiency, competitiveness and provision of services. MultiSward covered five main biogeographical regions (Atlantic, Continental, Alpine, Mediterranean, Boreal) and considered high inputs and low inputs systems, productive grassland and nutrient poor systems encountered across Europe thus covering a large array of ecosystem services. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 696 State of grassland and grassland based systems in Europe Evolution of grassland acreage in Europe MultiSward has provided a detailed view of grassland acreage and utilisation in Europe. Permanent grasslands (PG) cover over 57 million ha in the EU-27 (2007), temporary grasslands (TG) about 10 million ha. Together, they occupy about 39% of the European Utilized Agricultural Area (UAA). Grasslands still cover the largest agricultural area. Permanent grassland area is very important in Ireland (75% UAA), UK (58% UAA), Slovenia (58% UAA) and Austria (55% UAA) (Huyghe et al., 2014). In terms of number of hectares the United Kingdom (11 million ha), France (9.8 million ha), Germany (4.8 million ha), Italy (4.5 million ha) and Romania (4.5 million ha) represent 62% of the total permanent grassland area in EU27. These grasslands are the basis of the feeding of about 78 million Livestock Units (LU) of grazing livestock. They are managed by about 5.4 million holders, i.e. about 40% of all European farm managers. Among these farms managing permanent grasslands, 41% have an Economic Size Unit (ESU) lower than one (very small farm). The European grassland area has been significantly reduced during the last 30 years there were large differences in evolution trends between countries. Losses were very important in Belgium, France, Germany, Italy and The Netherlands while surfaces remained almost stable in Luxemburg, United Kingdom and Ireland. In the EU-6, these losses are estimated at about 30% and 7 million ha between 1967 and 2007 (Eurostat, 2009). Behind these mean figures, estimates vary according to the sources of information (Eurostat, FAOSTAT and national databases) and even within databases over time because the definition of grassland is variable according to the source, terms are often used in an imprecise and misleading way (for example terms like ‘meadows’ and ‘pastures’), the term ‘rough grazing’ does not represent all species-rich grassland types, Temporary grasslands are recorded as ‘Leguminous plants’ and ‘Temporary grass’ which induces doubt in the classification of grasslegume mixtures and changes in survey methods occurred over time in some countries (Greece, Italy, Portugal). This clearly obscures the vision of various stakeholders and does not allow taking into account all the diversity of grasslands by the Common Agricultural Policy (CAP). A better definition and classification of grassland terms should help to optimize the supports for grassland according to the services that they can provide and to secure grassland acreage in Europe with well targeted premiums. A comprehensive classification of grassland types was provided by an EGF / MultiSward group of experts coordinated by A Peeters. The proposal is described elsewhere (Peeters et al., 2014). It is a compromise between the practical aspects related to data collection and the level of precision that is necessary to reach the objectives described above and consist in (i) the classification of temporary grasslands into three categories: pure legume sowings, pure grass sowings and grass-legume mixtures; (ii) the classification of permanent grasslands into three categories: agriculturally-improved, natural and semi-natural, no longer used for production; and (iii) the introduction of a new category for grazed fallow land. Services provided by grassland area and grassland based production systems in various European regions To evaluate the impacts of ruminant production systems on the quality and the use of natural resources (air, water, soil, energy, biodiversity) and to assess a large range of ecosystem goods and services provided by grasslands and grassland-based systems an indicator system was developed and implemented. This system (MultiSward Indicator System - MIS), focuses primarily on provisioning services (production of food and feed), regulating services (atmosphere regulation with GHG and NH3 emissions; water quality protection: gross nitrogen and phosphorus balances, pesticide use), supporting services (biological N fixation), cultural services (landscape quality). The MIS is inspired by recent and effective systems, and Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 697 particularly by the 28 agri-environmental indicators of the European Commission (2006) calculated at country level. The scope of MIS is more restricted than the agri-environmental indicator system of the European Union but it is as much as possible compatible with this system. The MIS structure is based on the Driving force — Pressure — State — Impact — Response (DPSIR) framework of the European Environment Agency (EEA, 1999). Another characteristic of the MIS is that it includes two lists; one is calculated per farm type for a selection of regions, the second is calculated per region for the same selection (all farm types merged), while the EC indicator list is calculated at country level (all regions and farm types merged). The lists adapted by MIS to farm types and to regions levels include respectively 23 and 45 indicators. This conforms with EUROSTAT (2013) requirements that consider that regional data are of particular importance for agriculture and especially for the new risk indicators. MIS was applied in 16 NUTS 2 regions (Figure 1) corresponding to a large range of pedo-climatic, geographic and social conditions in Europe and to a large range of farming systems. Regions were chosen according to the typology of livestock regions described by Pflimlin et al. (2005) and EEA (2001). Figure 1. Livestock regions of Europe with the indication (circles) of regions studied. Grassland regions of the lowland (permanent grassland (PG) > 40% of agricultural area (AA) and less than 10% maize forage (MF) in main forage area (MFA)); Grassland and maize regions (PG/AA > 40% and MF/MFA < 10%); forage crop regions (PG/AA < 40% and MFA/AA > 50%); arable land and livestock (PG/AA < 40% and 20 < MAF/AA <50%); arable land and no livestock (PG/AA < 40% and MFA/AA < 20%). NUTS2 regions were Centre France (FR24), Brittany (FR52), Lower and Upper Normandy (FR25, FR23), West and East Wales (UKL1, UKL2), Midland-western and Southern-Eastern Ireland (IE01, IE02), Niederbayem (DE22), Zentralschweiz (CH06), Wielkopolskie voivodship (PL41), Trentino (ITD1), Alto-Adige (ITD2), Puglia (ITF4), Sardinia (ITG2). The utilisation of the MIS highlighted some contrasting evolutions between European regions and provided some original information on economic and environmental performances both at farm and regional levels. Although the permanent grassland area is still decreasing at the scale of Europe (especially in North West Europe), permanent grassland areas increase (+8% ITG2) or are relatively stable in ‘Permanent Grassland’ regions. Beef cattle farms are the best guarantor of these surfaces since dairy farms tend to use more green maize, when possible, temporary grasslands and even cereals in the diet of their productive animals. Permanent grassland areas decrease, sometimes quickly, in ‘Permanent Grassland & Maize’ and ‘Arable Land & Livestock’ regions (- 5 to -7% between 2000 and 2010 for FR252-253, FR23). That Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 698 shows that green maize exerts a powerful pressure on permanent grassland when cropping of this crop is possible and understood by farmers. In ‘Forage crops regions’, it seems very easy for farmers to convert temporary grasslands into green maize and other annual crops (cereals). Annual yields of agriculturally improved permanent grasslands were estimated (on the basis of expert knowledge) as ranging between 5 and 10 t DM ha-1. Annual yields of temporary grasslands were assessed at about 9 to 16 t DM ha-1. In Mediterranean regions both grassland type yields are lower. The high yields of temporary grassland swards can be partly explained by a good proportion of legumes. The MIS also provides information on the environmental performances of farming systems. At regional level, ammonia emissions (from 9 to 45 kg N-NH3ha-1) are best correlated with monogastric stocking rate, GHG emissions (0.8 to 4.7 t CO2 eq ha-1) with grazing livestock stocking rate (highest value for IE01, IE02, FR52, DE22 and DE21) and gross nitrogen balance (35 to 140 kg N ha-1) with total livestock stocking rate. Biological nitrogen fixation is low everywhere in permanent grasslands (< 25 kg N ha-1) showing that grassland productivity relies still on nitrogen fertilisation although some regions such as Wales (UKL2 = 70 kg ha -1) and Central Switzerland (CH06 = 52 kg ha-1) are however able to devote a much higher proportion of N to legumes in grassland swards. Biological nitrogen fixation is, however, much higher in temporary grassland (73 to 225 kg N ha-1) with the noticeable exception of Ireland (IE1 and IE2 = 2 kg ha-1). Incorporation of lucerne, red clover and other nitrogen fixing legumes is indeed common in temporary grassland mixtures. Average soil organic carbon (SOC) density in grassland soils is considerable, about twice the value of arable land in our sample and reaches 300 t CO2 eq. ha-1. Regarding global warming potential, gross nitrogen surplus and ammonia emissions, livestock farms are in average 2 to 5 times more polluting per surface unit than arable farms. However, this negative impact on the environment is lower than for ‘Specialist granivores’ that emit 10 to 100 times per surface unit more than arable farms. Moreover, grassland-based systems store carbon in soil organic matter which compensates GHG emissions and have also an important capacity to absorb nitrogen surplus in their SOC. Concerning biodiversity conservation, it is noticeable that specialised grazing livestock regions considered all together maintain and establish more landscape elements than arable land regions. The proportion of High Nature Value (HNV) farmland in the agricultural area is relatively high on average in Grassland regions but reflects mainly national policies. It. is high in Wales (46% AA) but low in areas where this proportion could be very high, like in the Italian Alps and Mediterranean areas as well as in Poland (8% AA). The number of ‘Natura 2000’ protected habitats in grassland is logically high in the Alps and in Poland (i.e, 7 to 10) but surprisingly low in the Mediterranean areas of Italy. Even in grassland regions, the trend of the populations of farmland birds is sharply decreasing although exceptions, like in Wales, show that an adequate management of landscape and favourable agri-environmental policies can revert the trend. ‘Specialists small ruminants’ (sheep, goats and other grazing livestock) are the best managers of biodiversity. They are the main managers of rough grazing areas and are characterized by a high level of agricultural biodiversity (livestock species diversity). They also maintain landscape elements very well. ‘Specialist dairying’ are much less efficient in terms of biodiversity conservation but they also maintain landscape elements that are important for wildlife. ‘Specialists monogastrics’, ‘Specialists vineyards’ and ‘Specialists arable crops’ are the worst managers of ecological infrastructures despite the fact that, in some regions, ‘Specialists arable crops’ can use and maintain rough grazing areas, probably with sheep flocks that constitute a marginal part of their income Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 699 Identification and analysis of the effects of socio-economic and policy scenarios on the future of grassland acreage After the radical reform of the CAP in 2003 the agri-environment measures have contributed to reduce the loss of permanent grassland (Peyraud et al., 2012), but much of the more intensive grassland areas will have remained outside of these schemes and may be used for cropping when commodity prices make this more profitable. MultiSward has identified some external drivers that can support grassland-based systems or at the opposite threaten them by assessing the likely effects of the revised pillar 1 CAP policies and how modified payment schedules could influence farmer’s decision to use land for growing grass or other crops in the case of 3 countries (Switzerland, Germany, Wales) (Hecht et al., 2014) using FARMIS model. Specifically, the analyses intended to quantify from current policy frameworks (baseline) alternative specific policies scenarios which could support grasslands and grassland-based systems under different market conditions or. The rationale for the choice of Germany, Wales and Switzerland is that the three countries differ considering the role and importance of grassland faming and the existing agricultural policies and direct payment system considered in the base year. The FARMIS agri-sector model (Offermann et al., 2005; Sanders et al., 2008) was used to model likely farmer behaviour with regard to grassland use and consequences on the basis of socio-economic outputs and environmental outputs (e.g. GHG emission, biodiversity, nutrient losses). Strong variations in input and output prices (50%) have a very significant effect on grassland area, the intensity of management of grassland systems and emissions to the environment, and are a key driver for the profitability of grassland systems. The high output price scenario results in farming intensification and overall production increases (milk by +29% in Germany, +29% in Wales, +51% in Switzerland; beef output by +17% in Germany, +43% in Switzerland; and sheep by +13% in Wales) and more fodder is grown on arable land; temporary grassland increases while extensive grassland use is strongly decreasing (-42% in Germany - especially in the intensive dairy production regions of Schleswig-Holstein, Lower Saxony and Bavaria, 55% in Wales). This scenario also leads to environmental problems. Increased input prices (fertilizers, energy, concentrates) in the baseline scenario led to a significant decrease in Farm Net Value Added in real terms according to countries and farm types. Dairy farms can maintain their income level, as farm and productivity growth and an increase in milk production can compensate for rising input costs while, in contrast, beef farms see their income declines which raises serious questions with respect to their economic sustainability in the long term. The considered policy scenarios had two common themes among countries (i) PREM1: greater payments to permanent grassland from 100 to 250€ ha-1 according to the country (in Germany with re-allocation of 15% budgets from the first pillar, in Wales and Switzerland with a “top up” payment); and ii) PREM2: greater payments to extensive grassland (i.e. +250€ ha-1), with reduced payments to arable land (-40€ ha-1 on the mean). The results for the area payment scenarios with and without transmission of payments from arable land illustrate the potential of such payments in maintaining grassland acreage and in further harmonising income levels between farm types, but this may not work for all areas. Whilst this appeared to work successfully in Germany and Switzerland, in Wales and other grassland-dominated areas, the potential budget transfer from arable areas is low, and therefore capacity for increased grassland payments is reduced. These payments have limited impacts on emissions except in Switzerland where support to permanent grasslands reduces Neutrophication through a reduction of arable land and temporary grassland. Regarding the scenarios in extensive grassland and on differentiated supports, there are some contrasting results. In Germany, the support of extensive grasslands is very positive for the extensive grassland area and for farmer’s income. In Switzerland, the best option is the differentiated support to grassland types (semi-natural > agriculturally improved permanent > temporary Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 700 grasslands). It has a very positive impact on the semi-natural grassland area, on farm income and positive but limited impact on N-eutrophication. Table 1. Impact of different market and policy scenarios on land use, production and income at the sector level in Switzerland (CH), Germany (GE) and Wales (W). Data expressed in % change to baseline scenario IP+50 CH GE OP+50 W CH GE PREM1 W CH GE PREM2 W CH GE W Arable land (ha) -2 -2 -3 3 0 0 -18 0 0 -40 0 -4 Permanent Grassland (ha) 0 -2 -7 0 0 2 0 0 2 7 0 -1 Intensive grassland (ha) 0 -10 -14 -2 16 25 21 0 1 46 -30 0 Extensive grassland 4 18 5 0 -42 -55 2 1 8 23 80 -5 Number of dairy cows -4 1 -9 51 29 29 0 0 0 14 -1 0 others ruminants -12 -19 -22 37 22 12 0 0 2 -4 -1 -2 Farm income (€ AWU-1) -10 -24 105 81 629 10 0 220 6 0 11 162 N balance (kg ha-1) -2 -7 8 7 -14 0 -22 -1 -1 CH4 (kg eqCO2 ha ) -2 -10 83 20 0 1 13 -1 “IP_+50”=Increase in input prices; “OP_+50”=Increase in output prices; “PREM1”=Increase in grassland payments; “PREM2”=Increase in support for extensively managed grasslands (+ transferred from arable land) Assessment and optimisation of the performances of MSS While the agronomic benefits of grass-legume mixtures over grass monocultures have been recognized for a long time, the simplicity of managing grass monocultures and the low price of nitrogen (N) fertilizer have in the past inhibited the use of MSS for forage production in many European countries (Peyraud et al., 2009). Field experiments have shown both higher and lower herbage yields from MSS than from monocultures, depending upon species composition, weather conditions, and management (Sanderson et al., 2004, 2005). In MSS, the ‘diversity effect’ has been defined as the excess of mixture performance over that expected from component species’ monoculture performances (Loreau, 1998). A highly species-rich system (> 30 species) may not necessarily meet farmers’ primary objective of producing high yields in productive and stable environments (Sanderson et al., 2004). It would clearly be useful in practical terms if the diversity effect did not rely on highly species-rich communities, but instead could be obtained with a mixture of a few species, well adapted to the appropriate environmental conditions. Plant diversity increases herbage production There has been a clear need for experimentation under realistic agronomic conditions to provide definitive evidence of the benefits of sward diversity to farmers in terms of primary (plant) and secondary (animal) production. A previous pan-European study (Kirwan et al., 2007) suggested that a major diversity benefit for yield in agricultural grasslands in temperate regions can be achieved with as few as four well-adapted plant species and these benefits occurred over a wide range of climatic conditions and nitrogen fertilization levels (Lüscher et al., 2008). In the MultiSward ‘Common Experiment’ (CE), agreed protocols were managed at seven sites and five countries during three years it was hypothesised that using grass/legume mixtures comprising a small number of strategically chosen species for forage production would be a viable option for achieving sustainable intensification of grassland-based agricultural production. Two grasses and two legumes were used. The plots were established based on 4 different ‘functional groups’ as follows (Collins et al., 2014): Nf = non-N fixing, Fi = N-fixing; Sr = shallow rooting; Dr = deep rooting. The latter contrast was included because enhanced nutrient uptake from a larger soil pool has been suggested to contribute to over yielding in mixtures (de Kroon, 2007; de Kroon et al., 2012). Mixtures of the following species were established: perennial ryegrass (Nf-Sr), Festuca arundinacea (Nf-Dr), white clover (Fi-Sr) and Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 701 red clover (Fi-Dr). In one treatment the second grass was replaced with chicory (Cichorium intybus) at some of the experimental sites (non-N fixing) or chicory was added as a fifth treatment. Chicory is a deep-rooted forb with a high nutritive value and is well-adapted to dry summers (Barry, 1998). The quantity of applied N was related to the expected biomass productivity (low = 12, high = 25 kg N t-1 expected DM). Results investigating primary production using cut swards (Collins et al., 2014; Figure 2) clearly showed that there was no detriment to DM yield in legume-based MSS compared to grass monocultures receiving high inputs of external nitrogen fertilizer. Indeed, in some instances MSS were more productive than the latter thus confirming previous results. Increased use of MSS therefore potentially represents a substantial economic and environmental saving when the various costs associated with the use of nitrogen fertilizer are considered. The trends shown by the results at some (AU-IBERS), but not all (Agroscope) sites indicating an increase in biomass production from one to two and from two to four species in the sward, as well as a predominant role played by the legumes, were in agreement with the results of the previous studies (Kirwan et al., 2007). However, over all time scales significant changes in the contribution of sown species to total biomass in mixtures occurred over time in MSS. This illustrates the complexity of species dynamics in MSS. Figure 2: Mean cumulative DM yield (sown + unsown species, kg ha -1) over 3 years in Aberyswyth (12 harvests) and Tänikon (18 harvests) in cut plots (PRG:perennial ryegrass). A novel aspect of the CE was the defoliation management (‘grazing vs. cutting’) comparison between biomass production in MSS and grass monocultures carried out at a subset of three sites (sheep in AU-IBERS; beef cattle in Agroscope-Tänikon; dairy cows in INRA-Rennes). There were no clear effects on sward yield of cutting versus grazing and no interactions between defoliation management and sward type with cattle and dairy cows. Grazing with sheep had no effect on grass-based swards but reduced sward yield in legume-based swards. Sheep grazing appeared to have a direct and detrimental effect on the legume component of MSS and this result support the hypothesis that selection of different species or sward types by grazing sheep can have a large effect on the yield, stability and composition of MSS, but that the identity of the grazing animal is the key determinant. Yarrow and Penning (1994) have previously found that the proportion of white clover in mixtures was lower when they were grazed by sheep than if the swards were harvested by cutting. Plant diversity increases animal performances and production of animal product on a per hectare basis Research on diversity-productivity relationships in sown grassland has generally focused on primary production, rather than performance of grazing animals. Interactions between different forage elements of the diet can occur and modify forage intake, metabolic processes in the rumen itself and emissions to the environment. MultiSward entails investigations of animal responses to complex swards to better understand the effects of interactions that can occur Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 702 between plants on digestion and intake. Experiments were conducted as part of the CE with ruminants fed indoors ad libitum to estimate the voluntary dry matter intake and in grazing situation. One experiment also considers permanent grassland. Indoors experiments with ad libitum feeding showed that voluntary DM intake was positively related to diet complexity in sheep, cattle and dairy cows. Three indoor feeding experiments were conducted with sheep fed ad libitum with mixtures of forages in controlled proportions (in %, 100:0; 75:25; 50:50; 25:75; 0:100). These experiments showed a positive effect on voluntary intake of dry matter between (i) silage of cocksfoot and red clover when increasing the proportion of red clover in the mixture, (ii) fresh rye grass and chicory when increasing the proportion of chicory and (iii) fresh ryegrass and white clover when increasing the proportion of white clover (Niderkorn et al., 2014). Synergy between some species in binary mixtures were observed (Experiments i and iii), optimal for the proportion 50:50, the percentage differences between the values measured for the plant combinations and the balanced median values from pure forages were +9.5% and +5.6% respectively. A cattle experiment (Morel et al., 2014) also suggested a greater voluntary intake for a mixture of four forages (ryegrass, chicory, white clover and red clover) compared to pure stand ryegrass (9.2 vs 8.8 kg DM day-1 respectively). These results agreed with previous data showing that a mixture of several forages could stimulate the motivation to eat and level of intake (Niderkorn and Baumont, 2009). Moreover, voluntary DM intake of legumes measured on sheep at maintenance is 10 to 15% greater than that of grasses of similar digestibility, and this is true whether legume forages are fed as silage, hay or fresh (INRA, 1989). Dewhurst et al (2003) also reported than silage DM intake is increased by 2 to 3 kg when cows are fed with red clover or white clover silage compared to ryegrass silage. Higher voluntary intake of legumes is attributed to both a lower resistance of legumes to chewing and a higher rate of particle breakdown, digestion and clearance from the rumen (Waghorn et al., 1989; Steg et al. 1994). Under grazing, the differences observed in forage yield and structure and management would affect the conclusion drawn from ad libitum fed animals. An experiment carried out over two years (13 rotations) managed at similar pasture allowance (22 kg DM per cow per day) clearly demonstrates advantage of MSS on a per cow basis for pasture DM intake and milk yield (RocaFernandez et al., 2014). The treatments were pure ryegrass, mixture of ryegrass and white and red clover, same mixture plus chicory and same mixture plus chicory and fescue. Pasture intake and milk yield were greater on the clover mixture than on pure grass (+1 kg per cow per day) and further increased by 0.5 kg on the more complex mixtures including chicory and fescue. In this experiment the total number of grazing days was hardly affected by the treatment, so that milk output per hectare was higher for MSS than for pure ryegrass swards (+ 1700 kg ha-1). Similar results were obtained for grazing cattle (Morel et al., 2014) when comparing ryegrass and the mixture of four species. Several short term trials have previously shown that increasing the content of white clover in pasture increases milk yield by 1 to 3 kg per cow per day when pastures were compared at the same herbage allowance (Philips and James, 1998; Ribeiro Filho et al., 2003). On low productive permanent grassland, a rotational grazing experiment was conducted in Germany over five years (2007-2011) using sheep and cattle in mono or mixed grazing of swards differing in diversity. Herbicides against dicotyledonous plants were used on plots in permanent grassland, resulting in a grass-dominated (gd) sward (6.9 ± 1.5 species m-2) compared to the untreated diverse (div) sward (10.3 ± 2.9 species m-2). Both diversity treatments were either grazed by sheep (S), or cattle (C). Grazing cattle were suckler cows and calves of the breed German Simmental. Ewes with lambs were Blackheaded and Leine sheep in comparable proportions of (C) or both (CS). Lamb production was slightly enhanced on the diverse swards and this effect was consistent over time, whereas calf performance was unaffected by sward type. Lamb growth was further improved by mixed grazing. These results Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 703 suggest that mixed, rotational grazing of cattle and sheep on phytodiverse agricultural swards is particularly appropriate to enhance lamb production (mixed grazing: + 17%; diverse swards: + 12%) without having any disadvantages for calf performance. Potential for optimising the environmental roles of grassland through the concerted use of plant species diversity at the field level Nutrient management is a main lever to improve the environmental performances of grasslandbased production (Huguenin-Elie et al., 2012). Notably, it is expected that reducing N fertilization while maintaining productivity thanks to the use of efficient, legume containing MSS could significantly modify the provision of regulating and supporting services from grassland-based ruminant production systems (Tilman et al., 2002; Finn et al., 2013). Different field experiments, including the CE, were performed to assess options for improving the environmental roles of grassland at the field level and to show that increased biomass productivity of grassland through the concerted use of legume-based MSS does not conflict with the delivery of a broad range of services. Considering N use efficiency and losses affecting water quality, the use of legume-based MSS showed clear advantages compared to fertilizer-based grass monocultures. Mineral N (Nmin including nitrate and ammonium N) was studied as large amounts poses a risk of pollution of ground water. Measurements of Nmin indicate that MSS combining both N-fixing and non-fixing species and including a deep-rooting species performed as well in term of residual mineral N in the soil (Nmin) as pure perennial ryegrass swards, and that they perform better in term of Nmin per unit of DM yield, showing a lower risk of N losses for a similar yield. The results further show that moderately fertilized monocultures of the shallow rooting perennial ryegrass swards do not guarantee a low Nmin content in the soil during winter, and that monocultures and mixtures with the deep-rooting red clover as N-fixing component perform better that those with the shallow rooting white clover (Figure 3). Three experiments (Wales, Ireland, Switzerland) considered the risk of nitrate leaching and showed that high-yielding forage production systems using symbiotic N2 fixation and moderate levels of N fertilization do not increase the risk of nitrate leaching compared to grass monoculture systems relying on heavy N fertilization. Under cutting, the potential of nitrate leaching always remains very low (less than 1.5 mg nitrate l -1 water in ceramic cups) and the grass-legume mixtures did not show any elevated risk as compared to pure grass swards. Under grazing, only white clover was used as an N fixing legume. Nitrate losses from the grass-legume grassland were similar to losses from the grassland receiving higher inputs of fertilizer N and were directly proportional to the amount of N cycling within the systems regardless of the original source of N. A slightly higher risk of nitrate leaching was observed at two of the four studied points in time in the Teagasc site. These results fit well with previous reported data (Loiseau et al 2001). Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 704 Figure 3: Effect of sward species composition on N min in the 0-40 cm profile. Average of 1 site in Switzerland, 1 site in Wales and 2 sites in Poland from the common experiments sampled in the early winter and in the early spring. Horizontal lines join treatments that are not significantly different. D: 4-species mixtures with the indicated species being dominant, B: binary mixtures, M: Monoculture, Centroid: equal stand mixture combining the 4 species in equal amounts on a seed weight basis, Lp: Lolium perenne, Ci: either Cichorium intybus (Poland and Switzerland) or Festuca arundinacea (Wales), Tp: Trifolium pratense, Tr: Trifolium repens. Considering the greenhouse gas (GHG) emissions impacting air quality, two experiments of MultiSward provided new information on N2O and CH4. Direct N2O emissions from symbiotic fixation are considered negligible (IPCC, 2006) and, taking into account the fertilization-linked N2O emissions (Nemecek et al., 2001), it seems reasonable to consider that the partial replacement of N fertilization with symbiotic N2 fixation from legumes is an effective way of lowering N2O emissions from productive grasslands. Direct N2O emission from symbiotic N2 fixation was found to be negligible, and considering the quantity of N2O emitted per unit of forage produced, grass-legume mixtures performed as well as or better than grass monocultures in the two experiments. For example in the Irish study, annual N2O emissions from unfertilized mown perennial ryegrass-white clover plots and unfertilized mown perennial ryegrass were similar (2.38 vs 2.45 kg ha-1 year-1). Under grazing there was an obvious trend of lower N2O emissions from ryegrass-white clover plots compared to highly fertilized ryegrass plots, where annual N2O emission were 15% higher (7.8 vs 6.4 kg N ha-1 year-1) thus confirming the compilation of Jensen et al (2012). Legumes can contribute to reducing ruminal methane production (Waghorn et al., 2006). Enteric methane emission per unit of feed intake was reduced in some MultiSward experiments, but not in all cases, depending on the forage and mixtures tested, and the animals. Enteric methane emission tended to be lower when pure legumes were fed indoors to sheep compared to pure grasses but the differences remained small (10%, Niderkorn et al., 2014) and were negligible when considering grass/white clover with 10 to 50% clover in the mixture. It is noteworthy that methane production linearly decreases when increasing the proportion of chicory in a mixture of ryegrass and chicory (-2% per 10% increase of chicory in the mixture). In grazing dairy cows, methane emission per unit of feed intake was lower for grass/white clover cows compared to grass-only cows (21.5 vs 24.5 g kg-1 DM intake respectively). These results can be attributed to the differences between grass and clover or chicory during the digestion process, due to the high digestibility of white clover and the low fibre content of chicory, and thus to a modified ruminal fermentation pattern toward propionate (which is a hydrogen carrier) combined with an increased passage rate of legume particles. Permanent grassland displays a strong potential for C sequestration (Soussana et al., 2010). However, uncertainties surrounding estimates are often larger than the sink itself. Long term data on C flux and soil organic carbon in permanent grasslands for two grazed long term observation sites located in central France were analysed comparing a control (1.1 LU ha-1 with 213 kg N ha-1 year-1 and an extensive system (0.6 LU ha-1, no fertilisation). The data confirm Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 705 the importance of multi species permanent sward for carbon storage (order of magnitude of 2 t ha-1 year-1) but the sink activity is very variable. The most intensive system led to higher annual sink activity in years of dry and warm growing seasons (i.e. 2003, 2005, 2008 and 2011), whereas in years with more ample seasonal rain events (i.e. 2004, 2006, 2007 and 2010) the extensively grazed paddock held a higher sink activity (Figure 4). When taking g CH4 and N2O emissions in the net GHG balance, the sink activity of these ecosystems remained but was lower (order of magnitude of 2 t ha-1 year-1) Figure 4. Cumulated net ecosystem exchange (NEE) measured from 2003 to 2011 for the extensively and intensively grazed paddock, with annual sums below. Concerning aspects of soil functioning, one commonly proposed mechanism to reach functional complementarity in MSS is belowground vertical niche differentiation between shallow-rooting and deep-rooting species (Berendse, 1981; von Felten and Schmid, 2008). Beside this potential vertical niche differentiation, the difference in phenology between species may lead to different nutrient demand in time and therefore to a reduced completion for nutrient in multi-species communities. Both would result in a more complete exploitation of the soil and its available resources leading to higher above-ground biomass production. The utilisation of the soil profile was assessed using tracers (Hoekstra et al., 2013) on temporary grasslands sown with a range of botanical compositions. The studies showed that the proportional nutrient uptake from the shallow soil layer was significantly higher for the two shallow-rooting species (perennial ryegrass and white clover) compared to the two deep-rooting species (chicory and red clover), resulting in niche complementarity between shallow and deep rooting species. This is consistent with the study on residual Nmin in the soil. Multi-species swards combining shallow- and deeprooting species therefore efficiently use the soil profile. On the other hand, some evidence for temporal niche differentiation in N uptake from fertilizer (using 15N) was found between perennial ryegrass (spring species) and red clover and chicory (summer species). Mixing species with different temporal patterns of nutrient uptake might therefore contribute to a high nutrient capture and consequent high biomass production in MSS. Concerning non-renewable energy consumption to produce forage, legume-based MSS allowing high productivity with little fertilizer could greatly improve the consumed/produced energy ratio of forage production. Using the LCA methodology to asses this question from the yield data of the CE, it was shown that the 4-species grass-legume mixtures performed better than the grass monoculture at all five sites, (2.87 and 2.10 MJ-eq kg-1 DM respectively for the mixtures and the grass monoculture). Per kg of produced forage, production with the MSS required between 67 to 84% of the energy required with the grass monoculture at the same level of fertilization. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 706 Grassland management as a lever to manage grassland biodiversity from the field to the landscape level The beneficial ecosystem services provided by grassland partly depend on biodiversity, and the choice of plant and animal taxa that were considered accounted for this aspect. Because recent advances in ecology have stressed the benefits of not only considering plant taxonomic classification but also their functional classification for the study of ecosystem functioning, MultiSward tested the relevance of the functional approach for understanding the effects of management and climate on grassland diversity. At the plot scale, the dependence of plant functional diversity criteria in permanent grasslands to management and climate was analysed using a large dataset from 439 permanent grasslands covering a large range of soil and climatic conditions and management gradient. A first data set contains the surveys of 140 permanent grasslands from all French regions apart from the Alps and Mediterranean area, a second one contains a survey of 70 permanent grasslands from the Vosges region and the third one consist in the survey of 229 permanent grasslands in the Swiss Alps. Plant species richness and the community-weighted value of Specific Leaf Area (SLA), Leaf Nitrogen Content (LNC) and of the onset of flowering (OFL) are the considered variables because they are involved in the delivery of many ecosystem functions and services such as forage quantity and quality, or pollination (Lavorel et al., 2011). Links between climatic and management variables on species richness were analysed using a regression tree approach to select the most important variables. More than 60% of the variance in species richness could be explained by the surveyed climatic and management variables. Plant species richness and the onset of flowering increased with altitude, while the community-weighted mean of SLA and LNC decreased with altitude. Both defoliation intensity and N inputs had positive effects on SLA and LNC which confirms the results of previous surveys (Lavorel et al., 2011). Defoliation intensity had a negative effect on the onset of flowering, which could be related to plant reproductive strategy, since an early onset of flowering would allow species to complete their reproductive cycle before being defoliated. The climatic variables generally influenced species richness more than management variables, from which total N input and intensity of defoliation were the most important variables and the type of utilisation (grazing vs mowing) had a weaker effect. The management variables that appeared to exert the strongest effect on plant species richness differed along the climatic gradient. The response of the number of species to management intensity was different between the grasslands from regions with a colder climate and the ones from regions with a warmer climate. Conditional effects of management on species richness were also observed by de Bello et al., (2006). The intensity of defoliation, which generally produces a negative effect of the number of species (Gaujour et al., 2012), appeared as a main lever of the within-plot species richness (alpha diversity) within the regions with a rather cold climate, but not within the regions with a warmer one. Increasing applications of mineral or organic fertilizer generally result in a drop in plant species richness (Schellberg et al., 1999). In our dataset fertilization appeared to be a factor strongly influencing species richness, but especially for the warmer grasslands with summer rainfalls of more than 200 mm. Grazing was found to be positive for the number of plant species per plots only in two regions. These results show that strategies targeting an increase in within-plot species richness by modifying management practices need to be developed at the scale of small regions. At the farm scale, the effect of heterogeneity among grasslands has nevertheless seldom been quantified, although it is widely recognized that habitat heterogeneity has a positive effect on biodiversity in agricultural landscapes (Benton et al., 2003). Grasslands managed differently within a farm can shelter different plant communities with to some extent different species, which could have a positive effect on the overall species richness. An analysis was based on a dataset consisting in the survey of permanent grasslands in the Swiss Alps (235 grasslandsGrindelwald region) with model farms having 20 different grassland plots and different Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 707 management strategies based on four management classes (intensive vs. extensive and low vs. high altitude) and additional dataset consisting in the survey of 69 grassland plots from 9 farms from Norway and 31 farms from French Jura. The results consistently show that most of the richness in plant species richness is due to the between-plot diversity (β-diversity; Figure 5). Figure 5: Average diversity contribution per plot of different management classes for the alpha, beta and gamma diversity (richness) at the farm level for the Grindelwald region and for a farm with 5 plots of each management classes. Dark grey: alpha contribution; light grey: beta contribution; total: gamma contribution; standard error shown for gamma contributions only. IH: intensive management, high altitude, EH: extensive high, IL: intensive low altitude: EL: extensive low As this β-diversity is large within all types of management of permanent grasslands, farms with many plots always have much larger species richness at the farm scale than the average αdiversity of their plots. Both a larger α- and a larger β-diversity was found within the sets of extensively managed plots than within the sets of more intensively managed ones. Heterogeneity between grasslands therefore proved very important for plant species richness at the farm level. Promoting heterogeneity thanks to a differentiated grassland management at the farm scale thus appears to be an important component of diversity conservation in productive ruminant production systems located in grassland dominated landscapes. Such a strategy should mainly consider supporting infrequently mown grasslands receiving no or little nutrients and situated in less favourable locations, as well as extensively grazed grassland. Considering fauna diversity at plot level, it still remains difficult to assess the effect of grassland management and plant diversity on the abundance and diversity of various taxa. MultiSward developed a set of indicators to evaluate the impacts of grassland plant diversity and management on the abundance and diversity of six animal taxa selected due to their contrasting biological requirements and key biological functions (e.g., pollination, pest control, soil fertility). The methodology combines multi-criteria decision trees to predict taxa diversity according to management practices, sward composition and plant functional traits with fuzzy partitioning, allowing assessment of different types of information (qualitative or quantitative, more or less accurate knowledge) and which makes possible a more precise assessment than the DEXI one, which only permits propositions having a value of truth or falsity. Decision trees aimed to predict taxa diversity (Plantureux et al., 2014). First results obtained on fauna biodiversity at plot level reveal the usefulness of including plant diversity and simple management inputs to improve the environmental evaluation of grassland-based systems. Plant species richness appeared to have a direct and positive effect on butterfly and moth diversity, and on grasshopper species richness. Bumblebee abundance was also positively related to legume abundance, and the abundance of erect growth-form plants was assumed to have a major effect on the diversity of web-building spiders. Finally, it is noteworthy that the abundance of grasses strongly influenced earthworm abundance, which suggests that earthworms might be the only group that would not take advantage of an increase in sward diversity. For all these Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 708 taxa grassland plant diversity and an adequate management can preserve a high level of biodiversity. Innovations in grazing management to increase competitiveness and environmental benefits of grassland based systems The challenge for farmers in the years ahead is to increase the competitiveness of their business through innovation, productivity gain and increased operational scale with revenue (at least for milk production) projected to fluctuate/fall and the cost of production increasing. Achieving an efficient use of grassland is a key issue in this context. Extending the grazing season to increase pasture utilisation Several experiments have shown there is considerable opportunity to extend the grazing season in early spring and late autumn in intensive dairy systems in the west part of Europe (Dillon et al., 1998; Peyraud et al., 2010). Extending the grazing season reduces the requirement for silage, purchased feedstuffs, housing, and slurry storage and spreading, thereby improving the economic returns to the producer from their ruminant production system. In continuation of this work, MultiSward tested this practice in the case of more extensive systems in central Europe, examining strategies that can be incorporated into grazing systems in autumn and spring with respect to concerns about nitrate leaching. As previously demonstrated in Western Europe, extending the length of the grazing season is also feasible in nutrient poor grassland in central Europe. In the western part of Poland, the extension of the grazing season for suckler cows until the end of the year is possible without adverse effects on animal welfare (Piatkowski et al, not published). Paddocks were grazed in August and then closed until grazing in late October, November or December by 5 cows (Angus and Angus×Limousin; BW 460 to 520 kg). Herbage intake decreased during the season. Climatic conditions in Central Europe fluctuate much more than in the Atlantic climatic zone. This is reflected in fluctuations in the yield of the pasture sward between years and months (from 1.7 to 3.5 t ha-1), its utilization rate (from 65 to 81%) and animal intake (from 5.6 to 11 kg DM cow-1 per day). Good availability of herbage and favourable grazing conditions occurred in November 2011 and the poorest conditions in November 2012. The sod damage in each paddock, as a consequence of suckler cows grazing was low with the exception of November 2010, when heavy rainfall occurred (133.8 mm). Restricted access time to pasture can be used as a strategy to increase the length of the grazing season while minimizing damage to pasture by poaching during periods of inclement weather. In spring, restricted access time to pasture has been shown to have no (Kennedy et al., 2009) or a slightly negative (Perez-Ramirez et al., 2008) effect on milk yield. An experiment conducted in Ireland has examined the effect of restricted access to pasture in the autumn of late-lactation spring-calving dairy cows and showed that that restricted access time to pasture can be implemented on dairy farms in autumn with no reduction in dairy cow production. Forty-eight cows were assigned to one of four treatments: full-time access to pasture, two 5-hour periods of access to pasture after a.m. and p.m. milking, two 3-hour periods of access to pasture after a.m. and p.m. milking, and alternating between full time and 2x3H access to pasture with no more than three continuous days on any one regime. Treatment had no effect on animal performances (milk solid yield (1.15 kg day-1) or herbage intake (15 kg day-1)). This was due to changes in the cows’ grazing behaviour as, when access time to pasture was restricted, the grazing intensity increased leading to higher intake per minute and per bite. Removal of herbage and poaching as a result of grazing in autumn and the low utilisation of urine and faeces during the late autumn due to low grass growth rates increases the possibility of leaching and runoff. This was addressed by two experiments showing that the risk of nitrate leaching is not severely increased by extending the grazing season length and can be brought under control. In Ireland, the effect of nine grazing season lengths on nitrate leaching to 1 m in Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 709 the soil was tested using ceramic cups. Ceramic cups were sampled on 56 occasions between 25 January 2007 and 15 June 2010. Cows were turned out to grass post calving on 1 February, 21 February or 15 March, and remained at grass fulltime until 10 October, 25 October or 10 November. The same stocking rate and fertilisation levels were applied (2.47 LU ha-1 and 250 kg N ha-1 year-1 respectively). Nitrate leached ha-1 was similar for the three autumn housing dates (152 kg N-NO3 ha-1). However, management of paddocks affected nitrate leached: the control treatment (no grazing or fertiliser N) had the lowest level (36 kg N-NO3 ha-1), the grazing-only management had the highest (181 kg N-NO3 ha-1), and paddocks that were grazed and had silage harvested from them had intermediate values (109 kg N-NO3 ha-1). In Belgium the effect of cutting in autumn was examined in more detail. The nitrate content in the 0-90 cm soil horizon was compared between plots either cut (one or two) or grazed from 1 September until the end of the growing season in nine pastures on three soil types (sand, sandy loam and clay) over three years. The evolution of soil nitrate content in the period 1 September – 15 November was not dependent on the grassland management, cutting instead of grazing having the same effect on the nitrate content in the soil profile. However there was high variability in soil nitrate content on 1 September within pastures intensively grazed during the growing season. Taking 2 cuts – end of August and middle of October - can decrease the nitrate Ncontent in the soil at the end of the growing season in comparison with grazing, but taking only one cut (in the middle of October) has less potential to do so (respectively 88, 71 and 50 kg NNO3 ha-1). The N-uptake by the grass in autumn, under a 1- and 2- cut regime was considerably different (144 vs 98 kg N ha-1 for 2 cuts and 1 cut respectively). Grazing management in uplands pastures designed to increase biodiversity For upland pastures one of the challenges is to manage a high level of biodiversity, while developing appropriate grazing management strategies can contribute to increased grassland biodiversity. An experiment evaluated the consequence of withdrawing animals during the main flowering period on the biodiversity. An ‘ecological rotation’ strategy, taking the animals away from one rotational subplot during the main flowering period to decrease the stocking density locally, thereby favouring flowering intensity, was compared with continuous grazing under cattle and sheep, and the effects on the abundance of bumblebees, butterflies and ground beetles in both grazing management systems was examined. Cattle grazed plots had a larger flowering cover than sheep grazed plots and the ecological rotation with sheep allowed better flowering cover than continuous grazing management; no difference was found in cattle grazed plots. Flowering, visiting insect density and species richness correlated positively with the flower cover of the plot. Butterflies and bumblebees benefitted from the ecological rotation management both in cattle and in sheep grazing but the benefit seemed weaker with sheep than with cattle grazing, sheep grazing, leading to lower butterfly and bumblebee abundance and lower species richness than cattle grazing. Managing the ecological rotation with cattle instead of sheep allowed an increase of 32% in the butterflies per transect, 28% in the butterfly species per transect, 61% in bumblebee abundance and 53% in bumblebee species richness. Appropriate animal for successful temperate grassland-based systems Apart from grassland management, the most profitable genotype or breed is a key factor to return the highest profit per unit of the most limiting input. For example it is now established that that cows selected solely on the basis of milk production have poorer fertility performances and are not well suited for grassland based systems (Dillon et al., 2003; Horan et al., 2005). The development of sustainable grassland-based ruminant systems clearly requires the most adapted animal genetics. In MultiSward, long term experiments were conducted to evaluate genotype x management interaction. Modelling data using the Moorepark Dairy Systems Model (MDSM) (Shalloo et al., 2004) from these long term grassland based dairy cow production systems allowed a more holistic evaluation of the systems, including production, some Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 710 economic and some environmental evaluation. On- and off-farm GHG emissions from dairy production were assessed using a cradle to farm-gate attributional LCA sub-model (O’Brien et al., 2012). A French experiment compared Normande (No), which is a dual purpose breed, with HolsteinFriesian (Ho) on High and Low input feed systems. Cows grazed from April to October. During lactation the High Input dairy cows consumed 1700, 2900 and 1450 kg DM cow-1 of conserved forage, grazing forage and concentrate, while the Low Input cow consumed 1600, 3600 and 90 kg DM cow-1 respectively. The response to feeding strategy was greater for Ho cows (-1325 kg milk, -168 kg milk solids) than for No cows (-117 kg milk solids) and the reproductive performance was highly altered for the Ho cows, with low gestation rate especially in the Low feeding group. Regardless of feeding strategy, no cows had lower BCS loss in early lactation and higher BCS at dry off than Ho cows. These results clearly showed the high reactivity of the milk production in Holstein cows to feeding level, which does not limit the body condition loss as well as the degradation of the reproduction performance, making the Holstein cow incompatible with low inputs systems especially when compact spring calving is required. On the contrary, the dual purpose breed appears more flexible and better adapted to low input systems based on the maximisation of grassland use for milk production. The Ho breed had the lowest carbon footprint of milk. However, the relative difference between Ho and No breeds was low (1.25 vs 1.32 kg eq CO2 t-1 milk respectively for Ho and No) and varied according to the allocation methods between milk and meat due to the value for surplus calves and culled cows. Briefly, regardless of feeding system, the Ho and No breeds are similarly profitable in a 2005-06 prices scenario but with the 2007-08 scenario (high cereal prices) the Ho breed was more profitable than No. This was largely due largely due to the fact that fewer cows and heifers are required for the same quota due to the higher milk yield of the breed and so less land is required for grass. The additional land for the Holstein breed is then converted to cereal crops and the higher the cereal price, the greater the benefit associated with intensification of the milk production system In Ireland, the biological efficiency of three genotypes (Jersey, Holstein-Friesian and Jersey × Holstein-Friesian) was compared across three grassland-based systems. The Holstein and JexHo animals were stocked at 2.5, 2.75 and 3.0 cows ha-1 while the Jersey animals were stocked 0.25 cows ha-1 higher for all treatments. Crossbred dairy cows are capable of production levels per cow at least similar to their Holstein-Friesian contemporaries on low cost systems (453 kg milk solids cow-1) but fertility and survival levels are markedly improved (e.g. six week in-calf rates were increased by over 10 percentage units with crossbreds). Jersey cow a lower yield (424 kg cow-1) but yield per hectare was not affected by the type of cows (1166 kg ha-1). In Ireland, the impending removal of EU milk quotas will result in land becoming the most limiting resource. Economic analysis demonstrates a substantial profit benefit per lactation with the F1 cows. The difference in performance equates to over €12,000 annually on a 40 ha farm. This result is primarily attributable to improvements in milk revenue and the large differences in reproductive efficiency/longevity observed with the crossbred herds although lower cull cow and male calf values, especially during periods of high beef value, negate the potentially larger benefits from this cross. The Je×Ho cross also had a lower carbon footprint of milk than the Ho breed (4-10%) The range of the difference varied according to the allocation methods between milk and meat due to the value for surplus calves and culled cows. Concerning meat production, a Polish experiment examined the efficiency of lamb production of four sheep breeds (White-headed meat sheep, Wielkopolska sheep, Romanov sheep, Blanc du Massif Central sheep) in continuous grazing conditions on lowland pasture. The results show that Romanov Wielkopolska sheep had similar live weight gain for the grazing season. It was mainly influenced by high live weight gain in May for Romanov whereas Wielkopolska sheep Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 711 had the highest live weight gain in summer when the pasture yield was greatest. Lamb live weight gain of white-headed meat sheep and Blanc du Massif Central sheep was lower. Conclusion MultiSward has provided science-based information and expertise that are of the utmost practical relevance for farmers and for EU agricultural policy for maintaining grassland acreage and developing competitive, productive and environmentally friendly livestock systems. Among the most promising outputs of the project, it is noteworthy that MultiSward has (i) provided a state-of-the-art review about the roles and utility of grassland and stakeholders expectations; (ii) provided useful information and pointed out the need for improving European statistics on grassland acreage and grassland term definition to design more efficient public policies for maintaining grassland acreage; (iii) demonstrated that margins of progression exist: clearly multispecies swards can contribute to more sustainable ruminant production systems; performance of grassland based system requires well suited breeds and appropriate grazing management strategies and (iv) provided adequate tools to assess the performances of production systems at different scales and for various territories. Future programmes developing integrated approaches with multi-scale and multicriteria approaches are still necessary to improve our knowledge and to propose innovations for more competitive and sustainable grassland based systems. Some key research questions remain to be solved: (i) strategies for maintaining a functional legume presence (in terms of biomass) over time under both cutting and grazing managements. The issue of decreasing sward legume content was particularly evident in the case of red clover, but it was also apparent in white clover; (ii) in the context of climate change, the development and testing of new plant production systems and new multispecies grasslands having fewer requirements for water and higher resilience to dryness is required; (iii) progress is still required to determine the most appropriate ruminant phenotype and appropriate indicator traits that reflect improved forage use efficiency and reduced ecological footprint, and to fully exploit the adaptive capacity of herbivores to make better use of grassland in marginal land (land on which the only thing that will grow is grassland). Beyond research, a key issue for the future of grassland is to convince farmers to continue to use grassland and to help them to progress technically and economically. It would be particularly interesting to build a European grassland network (based on a so-called multi-stakeholder approach) aimed at: informing farmers and all relevant stakeholders; identifying, sharing and adopting innovations; and proposing references and demonstrations to improve the performance of grassland based systems and increase farmer confidence in these systems. Acknowledgement This research was funded by the European Community's 7th Framework Programme under the grant agreement FP7-244983 (Multisward). References Barry T.N. (1998) The feeding value of chicory (Cichorium intybus) for ruminant livestock. Journal of Agricultural Science 131, 251-257. Benton T.G., Vickery J.A. and Wilson J.D. (2003) Farmland biodiversity: is habitat heterogeneity the key? Trends in Ecology and Evolution 18, 182-188. Berendse F. (1981) Competition between plant-populations with different rooting depths. 2. Pot experiments. Oecologia 48, 334-341. Collins R.P., Delagarde R. and Husse S. (2014) Biomass production in multispecies and grass monoculture swards under cutting and rotational grazing. Grassland Science in Europe 19 (this volume). de Bello F., Leps J. and Sebastia M.T. (2006) Variations in species and functional plant diversity along climatic and grazing gradients. Ecography 29, 801-810. de Kroon, H. (2007) How do roots interact? Science 318, 1562-1563. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 712 de Kroon, H., Hendriks, M., van Ruijven, J., Ravenek, J., Padilla, F. M., Jongejans, E., Visser, E. J. W. and Mommer, L. (2012) Root responses to nutrients and soil biota: drivers of species coexistence and ecosystem productivity. Journal of Ecology 100, 6-15. De Vliegher A., Van Gils B. and van den Pol-van Dasselaar A. (2014) Roles and utility of grasslands in Europe. Grassland Science in Europe, 15 (this volume). Proceedings of the 23th General Meeting of the European Grassland Federation, Germany, vol 13, 865-879. Dewhurst, R.J., Fisher, W.J., Tweed, J.K.S. and Wilkins, R.J. (2003) Comparison of grass and legume silages for milk production. 1. Production response with different levels of concentrate. Journal of Dairy Science 86, 25982611. Dillon P., Crosse S. and Roche J.R. (1998) The effect of grazing intensity in late summer/autumn on sward characteristics and milk production of spring-calving dairy cows. Irish Journal of Agricultural and Food Research 37, 1-15. Dillon P., Snijders S., Buckley F., Harris B., O’Connor P. and Mee J.F. (2003) A comparison of different dairy cow breeds on a seasonal grass-grazed system of milk production. 2. Reproduction and survival. Livestock Production Science 83, 35-42. EEA (1999) Environmental indicators: Typology and overview. European Environment Agency, Copenhagen, Technical report 25, 19 pp. EEA (2001) Biogeographical regions. Downloadable from: http://dataservice.eea.europa.eu/atlas/viewdata/viewpub.asp?id=155 European Commission (2006) Development of agri-environmental indicators for monitoring the integration of environmental concerns into the common agricultural policy. Commission Communication (COM final 0508/2006): 11 pp. Downloadable from: http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=CELEX:52006DC0508:EN:NOT European Commission (2013) Type of farm. Agriculture and Rural Development, Farm Accountancy Data Network. Downloadable from: http://ec.europa.eu/agriculture/rica/detailtf_en.cfm?TF=TF&tf_Version=13185 EUROSTAT (2009) Agricultural statistics edition 2010: main results 2007-2008: 126 pp. Finn J.A., Kirwan L., Connolly J., Sebastià M.T., Helgadottir A., Baadshaug O.H., Bélanger G. Black A., Brophy C., Collins R.P., Čop J., Dalmannsdóttir S., Delgado I., Elgersma A., Fothergill M., Frankow-Lindberg B.E., Ghesquiere A., Golinska B., Golinski P., Grieu P., Gustavsson A.-M., Höglind M., Huguenin-Elie O., Jørgensen M., Kadziuliene Z., Kurki P., Llurba R., Lunnan T., Porqueddu C., Suter M., Thumm U. and Lüscher A. (2013) Ecosystem function enhanced by combining four functional types of plant species in intensively managed grassland mixtures: a 3-year continental-scale field experiment. Journal of Applied Ecology 50, 365-375. Gaujour E., Amiaud B., Mignolet C., and Plantureux S. (2012) Factors and processes affecting plant biodiversity in permanent grasslands. A review. Agronomy for Sustainable Development 32, 133-160. Hoekstra N.J., Finn J. and Lüscher A. (2013) Nutrient uptake in soil niches affected by plant species and drought stress. Agricultural Research Forum 2013, Tullamore, Ireland. Jensen E.S., Peoples M.B., Boddey R.M., Gresshoff., P.M., Hauggaard-Nielsen H., Alves B.J.R. and Morrison M.J. (2012) Legumes for mitigation of climate change and the provision of feedstock for biofuels and biorefineries. A review. Agronomy for Sustainable Development 32, 329–364. Hecht J., Moakes S., Offermann F. and Peeters A. (2014) Threats and opportunities for European grassland areas under different market and policy scenarios. Grassland Science in Europe 19 (this volume). Horan B., Dillon P., Faverdin P., Delaby L., Buckley F. and Rath M. (2005) The interaction of strain of HolsteinFriesian cow and pasture based feed system for milk production, bodyweight and body condition score. Journal Dairy Science 88, 1231-1243. Husse S., Huguenin-Elie O., Buchmann N. and Lüscher A. (2014) Nitrogen capture in mixed swards benefits from temporal complementarity among species. Grassland Science in Europe 19 (this volume). Huyghe C., De Vliegher A., van Gils B and Peeters A. (2014) Grassland and herbivore production in Europe and effect of common policies; Ed Quae, 287 pp. Institut National de la Recherche Agronomique (1989) Ruminant Nutrition: Recommended allowances and feed tables. R. Jarrige Ed., INRA (Paris) and John Libbey, Eurotext (Paris-London-Rome) pp. 33-48. IPCC, (2006) IPCC Guidelines for National Greenhouse Gas Inventories. Volume 4: Agriculture, forestry and other land use. IGES, Kanagawa, Japan. Kennedy E., Garry B., Ganche E., O’Donovan M., Murphy J.P., and Hennessy D. (2014) Effects of restricting access time to pasture on late lactation dairy cow production. Grassland Science in Europe 19 (this volume). Kennedy E., McEvoy M., Murphy J.P., O'Donovan M. (2009) Effect of restricted access time to pasture on dairy cow milk production, grazing behavior, and dry matter intake. Journal of Dairy Science 92, 168-176. Kirwan, L., Lüscher, A., Sebastià, M.T., Finn, J.A., Collins, R.P., Porqueddu, C., Helgadóttir, Á., Baadshaug, O.H., Brophy, C., Coran, C., Dalmannsdóttir, S., Delgado, I., Elgersma, A., Fothergill, M., Frankow-Lindberg, B.E., Golinski, P., Grieu, P., Gustavsson, A.M., Höglind, M., Huguenin-Elie, O., Iliadis, C., Jørgensen, M., Kadziuliene, Z., Karyotis, T., Lunnan, T., Malengier, M., Maltoni, S., Meyer, V., Nyfeler, D., Nykanen-Kurki, P., Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 713 Parente, J., Smit, H.J., Thumm, U. and Connolly, J. (2007) Evenness drives consistent diversity effects in an intensive grassland system across 28 European sites. Journal of Ecology 95, 530-539. Lavorel, S., K., Grigulis, P., Lamarque, M.P., Colace, D., Garden, J., Girel, G., Pellet and R., Douzet (2011) Using plant functional traits to understand the landscape distribution of multiple ecosystem services. Journal of Ecology 99, 135-147. Loiseau P., Carrere P., Lafarge M., Delpy R. and Dublanchet J. (2001) Effect of soil-N and urine-N on nitrate leaching under pure grass, pure clover and mixed grass/clover swards. European Journal of Agronomy 14, 113121 Loreau M. (1998) Separating sampling and other effects in biodiversity experiments. Oikos 82, 600–602 Lüscher, A., Finn, J.A., Connolly, J., Sebastià, M.T., Collins, R.P, Fothergill, M., Porqueddu, C., Brophy, C., Huguenin-Elie, O., Kirwan, L., Nyfeler, D. and Helgadóttir, A. (2008) Benefits of sward diversity for agricultural grasslands. Biodiversity 9, 29-32 Lüscher A., Mueller-Harvey I., Soussana J. F., Rees R.M. and Peyraud J.L. (2014) Potential of legume-based grassland–livestock systems in Europe: a review. Grass and forage science. 69, 206–228 Morel I., Schmid E., Soney C., Aragon A. and Dufey P.A. (2014) Influence of ryegrass alone or blended with clover and chicory on feed intake and growth performance of steers. Grassland Science in Europe 19 (this volume). Nemecek T., Huguenin-Elie O., Dubois D., Gaillard G., Schaller B. and Chervet A. (2011) Life cycle assessment of Swiss farming systems: II. Extensive and intensive production. Agricultural Systems 104, 233-245. Nyfeler D., Huguenin-Elie O., Suter M., Frossard E.and Lüscher A. (2011) Grass-legume mixtures can yield more nitrogen than legume pure stands due to mutual stimulation of nitrogen uptake from symbiotic and non-symbiotic sources. Agriculture, Ecosystems and Environment 140, 155-163. Niderkorn V., Martin C., and Baumont R. (2014) Associative effects between forage species on intake and digestive efficiency in sheep. Grassland Science in Europe 19 (this volume). O'Brien, D., Shalloo L., Patton, J. , Buckley F., Grainger C. and Wallace M. (2012) Evaluation of the effect of accounting method, IPCC v. LCA, on grass-based and confinement dairy systems’ greenhouse gas emissions. Animal 6, 1512-1527. Peeters A., Beaufoy G., Canals R.M., De Vliegher A., Huyghe Ch., Isselstein J., Jones G., Kessler W., Kirilov A., Mosquera-Losada M.R., Nilsdotter-Linde N., Parente G., Peyraud J.-L., Pickert J., Plantureux S., Porqueddu C., Rataj D., Stypinski P., Tonn B., van den Pol – van Dasselaar A., Vintu V. and Wilkins R.J. (2014) Grassland term definitions and classifications adapted to the diversity of European grassland-based systems. Grassland Science in Europe 19 (this volume). Offermann F., Kleinhanß W., Hüttel S. and Küpker B. (2005) Assessing the 2003 CAP Reform Impacts on German Agriculture Using the Farm Group Model FARMIS. In: Arfini, F. (ed.) Modelling Agricultural Policies: State of the Art and New Challenges. Proceedings of the 89th European Seminar of the EAAE, Parma, Italy, February 35, 2005. Parma: Monte Universita Parma Editore: 546-564. Pérez-Ramírez E., Delagarde R., Delaby L. (2008) Herbage intake and behavioural adaption of grazing dairy cows by restricting time at pasture under two feeding conditions. Animal 2, 1384-1392. Peyraud J.L., Le Gall A., Lüscher A. (2009) Potential food production from forage legume-based-systems in Europe: an overview. Irish Journal of Agricultural and Food Research 48, 115-135. Peyraud J.L., Peeters A. and De Vliegher A. (2012) Place et atouts des prairies permanentes en France et en Europe. Fourrages 211, 195-204. Peyraud J.L., Van den Pol-van Dasselaar A., Dillon P., Delaby L., (2010) Producing milk from grazing to reconcile economic and environmental performances. Proceedings of the 23th General Meeting of the European Grassland Federation, Germany, vol 13, 865-879. Pflimlin A., Buczinski B. and Perrot C. (2005) Proposition de zonage pour préserver la diversité des systèmes d’élevage et des territoires européens. Fourrages 182, 311-329. Philips, C.J.C. and James, N.L. (1998) The effects of including white clover in perennial ryegrass pastures and the height of mixed pastures on the milk production, pasture selection and ingestive behaviour of dairy cows. Animal Science 67, 195-202. Plantureux S., Dumont B., Rossignol N., Taugourdeau S. and Huguenin-Elie O. (2014) An indicator-based tool to assess environmental impacts of multi-specific swards. Grassland Science in Europe 19 (this volume). Ribeiro-Filho, H.M.N., Delagarde, R. and Peyraud J.L. (2003) Inclusion of white clover in strip-grazed perennial ryegrass pastures: herbage intake and milk yield of dairy cows at different ages of pasture regrowth. Animal Science 77, 499-510. Roca-Fernández A.I., Peyraud J.L., Delaby L., Lassalas J. and Delagarde R. (2014) Interest of multi-species swards for pasture-based milk production systems. Grassland Science in Europe 19 (this volume). Sanders, J, Stolze and M, Offermann, F. (2008) The Swiss agricultural sector model CH-FARMIS. Agrarforschung 3, 138-143. Sanderson, M.A., Skinner, R.H., Barker, D.J., Edwards, G.R., Tracy, B.F. and Wedin, D.A. (2004) Plant species diversity and management of temperate forage and grazing land ecosystems. Crop Science 44, 1132-1144. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 714 Sanderson, M.A., Soder, K.J., Muller, L.D., Klement, K.D., Skinner, R.H., Goslee, S.C. (2005) Forage mixture productivity and botanical composition in pastures grazed by dairy cattle. Agronomy Journal 97, 1465–1471. Schellberg J., Möseler B.M., Kühbauch W. and Rademacher I.F. (1999) Long-term effects of fertilizer on soil nutrient concentration, yield, forage quality and floristic composition of a hay meadow in the Eifel mountains, Germany. Grass and Forage Science 54, 195-207. Shalloo, L., P. Dillon, M. Rath, and M. Wallace. (2004) Description and validation of the Moorepark Dairy System Model. J. Dairy Sci. 87, 1945-1959. Steg, A., Van Straalen, W.M., Hindle, V.A., Wensink, W.A., Dooper, F.M.H. and Schils, R.L.M. (1994) Rumen degradation and intestinal digestion of grass and clover at two maturity levels during the season in dairy cows. Grass and Forage Science 49, 378-390. van den Pol-van Dasselaar A., Goliński P., Hennessy D., Huyghe C., Parente G. and Peyraud J-L. (2014) Appreciation of the functions of grasslands by European stakeholders. Grassland Science in Europe 19 (this volume). Tilman D., Cassman K.G., Matson P.A., Naylor R. and Polasky S. (2002) Agricultural sustainability and intensive production practices. Nature 418, 671-677. von Felten S., Hector A., Buchmann N., Niklaus P.A., Schmid B. and Scherer-Lorenzen M. (2009) Belowground nitrogen partitioning in experimental grassland plant communities of varying species richness. Ecology 90, 13891399. Waghorn, G.C., Shelton, I.D. and Thomas, V.J. (1989) Particle breakdown and rumen digestion of fresh ryegrass (Lolium perenne L.) and Lucerne (Medicago sativa L.) fed to cows during a restricted feeding period. British Journal of Nutrition 61, 409-423. Waghorn, G.C., Tavendale, M.H. and Woodfield, D.R. (2002) Methanogenis from forages fed to sheep. Proceedings of the New Zealand Grassland Association 64, 159-165. Yarrow N.H. and Penning P.D. (1994) Managing grass-clover swards to produce different clover proportions. Grass and Forage Science 49, 496–501. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 715 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 716 Theme 5 submitted papers Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 717 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 718 Biomass production in multispecies and grass monoculture swards under cutting and rotational grazing Collins R.P.1, Delagarde R.2 and Husse S.3 1 Institute of Biological, Environmental and Rural Sciences, Aberystwyth University, UK 2 Institut National de la Recherche Agronomique, UMR1348 PEGASE, Saint-Gilles, France 3 Agroscope, Institute for Sustainability Sciences, CH-8046 Zürich, Switzerland Corresponding author: rpc@aber.ac.uk Abstract A Common Experiment (CE) was set up within the EU-FP7 project ‘Multisward’ across a subset of partner sites to analyse responses of multispecies swards (MSS) to grazing and cutting managements. Across sites and managements there was no detriment to yield in moderately fertilized legume-based MSS compared with perennial ryegrass monocultures receiving high inputs of external nitrogen fertilizer. The response of sward types to grazing depended on sward composition and the identity of the grazing animal species. Keywords: grazing, legumes, multispecies mixtures, perennial ryegrass, yield Introduction Strategically designed multispecies swards (MSS) could be a key element in improving the delivery of provisioning services from grassland-based production systems (Finn et al., 2013). A Common Experiment (CE) was set up within the EU-FP7 project ‘Multisward’ across a subset of three partner sites to analyse responses of MSS compared with highly fertilized perennial ryegrass (PRG) monocultures to grazing and cutting managements under temperate maritime (Aberystwyth (UK) and Rennes (FR)) and continental (Tänikon (CH)) environmental conditions. The CE imposed contrasting defoliation managements (grazing and cutting) typical of intensive production systems for 2-3 years on sward types differing in species number and composition. Differences in biomass production between sward types within each management were analysed. Within each sward type, grazed and cut plots received the same external applications of nitrogen fertilizer (N) and were defoliated at the same frequency to the same residual height, so that differences in sward responses to cutting and grazing managements could be attributed directly to the influence of the grazing animal. The effect of defoliation management on sward type was also analysed. Materials and methods Four forage species in common agronomic use in Europe were included in the CE: two grasses (perennial ryegrass - PRG; tall fescue – FA) and two legumes (white clover –WC; red clover – RC). The second grass was replaced with chicory (Ci) in Rennes and Tänikon. Over all sites the CE included MSS treatments (two legumes + PRG; two legumes + two grasses; two grasses; two legumes + two grasses + Ci) and monoculture swards of PRG, a subset of which received a ‘high’ application of N. All MSS received a ‘moderate’ application of N (150 kg N ha -1 yr-1 in Aberystwyth and Tänikon, and 70 kg N ha-1 yr-1 in Rennes). High-N PRG monocultures received 300 kg N ha-1 yr-1 in Aberystwyth, 350 kg N ha-1 yr-1 in Tänikon and 165 kg N ha-1 yr1 in Rennes. Biomass production was measured during 2011-2013 in Tänikon (18 harvests), and during 2012-2013 in Aberystwyth (12 harvests) and Rennes (13 grazing and 10 cutting harvests). Grazing management: Management was based on the concept of biomass accumulation and removal, as in rotational grazing systems. Different animals were used in the three sites: nonlactating ewes in Aberystwyth; dairy cows in Rennes; beef heifers in Tänikon. In each grazing interval, plots were grazed to a target sward height appropriate for the animal (5 cm in Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 719 Aberystwyth; 4 cm in Rennes; 6 cm in Tänikon), and the animals were then removed. The length of the regrowth period between grazing intervals was defined for each site and adjusted if necessary to take account of variation in forage production in response to climatic conditions. Biomass productivity in grazed plots was measured immediately prior to grazing by cutting the herbage in known areas within the plot to the target sward height. These samples were ovendried and weighed. Cutting management: Cut plots had the same species composition as grazed plots, with the same level of replication. In Tänikon these were fixed subplots within the grazed plots; in Aberystwyth they were randomized in a separate block beside the grazed plots; in Rennes cutting was only carried out on fixed subplots (high and moderate N) within PRG monocultures. Cut plots were mechanically defoliated, and biomass productivity was measured by subsampling the cut herbage, drying and weighing. All sward type treatments were randomized and replicated three times, except for high-N PRGmonoculture plots, which were replicated four times. Results for biomass production were analysed by appropriate ANOVA structures (Rennes and Tänikon) and by REML (Aberystwyth). Preliminary results from these analyses are presented here. Results and discussion To integrate the effects of time, biomass productivity was expressed as cumulative dry matter (DM) yield. Results for total yield (sown + unsown species) are shown in Table 1. Table 1. Cumulative total DM yields (sown + unsown species; kg ha -1) over 2/3 years in three sites of the Multisward CE. Sward types: 1M = PRG mono; 2 = 2 non-legumes; 3 = 2 legumes + PRG; 4 = 2 legumes + 2 nonlegumes; 5 = 2 legumes + 3 non-legumes; 1H = PRG mono receiving high N. (a) Aberystwyth (12 harvests) Sward type 1M 2 3 4 5 1H Mean CUT 15510 15683 27089 25073 - 17459 20163 GRAZED 12852 11944 19229 20865 - 17531 16484 Mean 14181 13814 23159 22969 1M 2 3 4 5 1H Mean CUT 17382 21769 24491 26678 - 22905 22645 GRAZED 20423 24363 28989 31651 - 29718 27029 Mean 18903 23066 26740 29164 Management 17495 (b) Tänikon (18 harvests) Sward type Management 26312 (c) Rennes (13 grazing harvests; 10 cutting harvests) Sward type 1M 1M 2 3 4 5 1H Mean - - - - 26352 25109 - 28590 29481 29011 - 28401 Management CUT 23865 GRAZED 22901 * Mean 23393 26520 Note: Two non-legume species in Aberystwyth = PRG and FA; in Zürich and Rennes = PRG and Ci. Sward type 5 = 2 legumes + PRG + Ci + FA. * Yield calculated from 10 harvests to allow direct comparison with cut treatment Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 720 Aberystwyth: There were significant effects of management and sward type (P<0.001), together with an interaction between them (P<0.047). Overall, cumulative yield was higher under cutting than grazing (20163 vs. 16484 kg ha-1). Sward types 3 and 4 were the highest-yielding swards, overall. The management × sward type interaction resulted from a large difference in response between legume-based and non-legume-based MSS to defoliation. Yields of the legume-based sward types 3 and 4 were significantly higher under cutting than under grazing, whereas yields of the grass-based treatments 1M, 1H and 2 were not affected by management. Tänikon: There was a significant effect of management averaged over sward type (P<0.008), and cumulative yield under grazing was higher than under cutting (27029 vs. 22645 kg ha-1). Averaged over managements, the effect of sward type was significant (P<0.006). Sward types 3, 4 and 1H were the highest yielding, and 1M was the lowest. There was no management × sward type interaction, so all sward types responded to defoliation in a similar way. Rennes: The experimental design at this site reduced the number of sward types under cutting. Within this management there was a significant difference between high and moderate N PRG monocultures (P<0.007), the former being more productive. There was no effect of defoliation management on cumulative yields of the 1M treatment. Under grazing there was no difference in the yields of the various sward types. The agronomic utility of MSS was tested here by comparing the yields of sward types under realistic defoliation managements. We observed different sward responses to defoliation management in different sites. However, there was a confounding effect of the use of different grazing animals at each site (although this aspect of the CE also added to its agronomic relevance). In Aberystwyth, grazing either had no effect on sward yield (grass-based swards), or reduced it (legume-based swards) compared with the cutting management. Sheep grazing appeared to have a direct and detrimental effect on the legume component of MSS. In Tänikon, cattle grazing had a positive effect on yield in all sward types and there was no interaction of defoliation management with sward type. Thus, the most productive swards under grazing would be also the most productive under cutting in this site. Rennes carried out only one cut vs. grazed sward-type comparison, in which there was no effect of defoliation management on yield. Taken together, these results support the hypothesis that selective grazing can have a large effect on the yield of MSS, and that the identity of the grazing animal is the key determinant. It is well known that species of grazing animal differ in their ability to select sward components. Yarrow and Penning (1994) found that the proportion of white clover in mixtures was lower when they were grazed by sheep than by cattle, both of which were lower than if the swards were harvested by cutting. Our results show that there was no detriment to yield in legume-based MSS compared with high-N PRG monocultures, and in some instances MSS were more productive. Increased use of MSS therefore potentially represents a substantial economic and environmental saving when the various costs associated with the use of nitrogen fertilizer are considered. Acknowledgements The research leading to these results received funding from the European Community's FP7 Programme under grant agreement no. 244983 (Multisward). References Finn J.A. Kirwan L., Connolly J., Sebastià M. T., Helgadottir A., Baadshaug O. H., Bélanger G., et al. (2013) Ecosystem function enhanced by combining four functional types of plant species in intensively managed grassland mixtures: a 3-year continental-scale field experiment. Journal of Applied Ecology 50, 365-375. Yarrow N.H. and Penning P.D. (1994) Managing grass-clover swards to produce different clover proportions. Grass and Forage Science 49, 496–501. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 721 Nitrogen capture in mixed swards benefits from temporal complementarity among species Husse S.1,2, Huguenin-Elie O.1, Buchmann N.2 and Lüscher A.1 1 Agrosocpe, Institute for Sustainability Sciences, CH-8046 Zürich 2 Institute of Agricultural Sciences, ETH Zürich, CH-8092 Zürich Corresponding author: sebastien.husse@agroscope.admin.ch Abstract Highly fertilized grass monocultures carry a risk of nitrogen (N) losses to the environment. Studies have shown that grass-legume swards help to reduce the need for fertilizer applications for similar levels of biomass production due to their access to nitrogen from the atmosphere. However, other functional traits, such as root architecture and differences of phenology in time, may influence N uptake. Our study aimed at assessing which combinations of functional traits allow large N acquisition and high biomass production. Four forage species with different belowground traits (N2-fixing/non N2-fixing; shallow/deep rooting) were sown as monocultures and as mixtures of two or four species. Recovery of fertilizer N in the harvested plant biomass was measured during one regrowth in spring and one in summer using 15N labelled fertilizer. Productivity and 15N recovery were correlated. The recovery of fertilizer N was not reduced by the presence of N2-fixing species grown in productive mixtures. No clear effect of rooting depth on the acquisition of fertilizer N was observed, although monocultures of shallow-rooting species performed poorly both in terms of yield and of fertilizer N recovery. In spring, 15N recovery was highest with mixtures containing Lolium perenne, whereas in summer it was highest with swards containing Trifolium pratense and/or Cichorium intybus. In an average of both seasons, combining L. perenne with one or both of these summer species allowed a large 15 N recovery and biomass production at each period of regrowth, which indicate benefits of temporal complementarity for N uptake. Keywords: 15N recovery, temporal complementarity, multi-species swards Introduction Intensive forage production based on grass monocultures and high nitrogen (N) fertilization carries a risk of N losses to the environment. Grass-legume mixtures have been suggested as an alternative to this system. Optimizing mixtures for high yields and complementarity in nutrient acquisition between species might increase efficiency of nutrient capture: Nyfeler et al. (2011) showed that the uptake of soil and fertilizer N can be higher in mixed swards combining 60% of grasses and 40% of legumes than in grass monocultures. Despite the high percentage of N derived from symbiotic N2 fixation, these grass-legume mixtures produced such a high yield that their demand on non-symbiotic N was large. Most studies on nutrient uptake in intensively managed mixed swards have focused on grasses and legumes. In addition to grass and legume species, Cichorium intybus was added in our experiment. This species is a non-N2 fixing species with a deep tap-root system, which differs from the grass root system. Such differences in root architecture might reduce competition for N within the soil profile, increasing the total N uptake of the community (Berendse, 1982). In addition, differences in phenology may induce asynchronous N demand between the species. This temporal complementarity between species may reduce competition for nutrients (Casper and Jackson, 1997). Our study aimed at determining which combination of plant traits (spatial and temporal complementarity) allows for high N acquisition from fertilizer and high biomass production in frequently defoliated multi-species swards. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 722 Material and methods Four replicates of eleven types of swards, based on four forage species differing in belowground traits, were sown in April 2011. These species were: Lolium perenne (non-N2 fixing and shallow rooting species), Cichorium intybus (non-N2 fixing and deep-rooting species), Trifolium repens (N2 fixing and shallow-rooting species) and Trifolium pratense (N2 fixing and deep-rooting species). The eleven types of swards included the monoculture of each forage species, all six combinations of two of the four species, and the four-species mixture with species sown in equal relative abundance. The swards were harvested six times per year and fertilized with 145 kg N ha-1 year-1. Two weeks before the harvests in May 2012 and July 2012, 0.64 L of a solution of 15N labelled fertilizer (ammonium nitrate; 0.03 g 15N m-2) was applied with a watering can on 0.64 m2. At harvest, a plant sample was cut at 6 cm in this subplot to measure 15N content in the plants. The 15N recovery (%) in the harvested biomass was calculated using the following 15 equation: 𝑁𝑝𝑙𝑎𝑛𝑡 − 15𝑁𝑛𝑎𝑡𝑢𝑟𝑎𝑙 15𝑁 𝑓𝑒𝑟𝑡𝑖𝑙𝑖𝑧𝑒𝑟 𝑥100 where 15𝑁𝑝𝑙𝑎𝑛𝑡 is the amount of 15N (g m-2) contained in the biomass harvested from the labelled area, 15𝑁𝑛𝑎𝑡𝑢𝑟𝑎𝑙 is the amount of 15N in plants harvested in an unlabelled subplot and 15 𝑁𝑓𝑒𝑟𝑡𝑖𝑙𝑖𝑧𝑒𝑟 is the amount of 15N applied with the fertilizer. An ANOVA was performed to compare the differences in 15N recovery and biomass production between the types of sward. Results and discussion Between 9 and 25% of the applied 15N were recovered in the plant biomass above 6 cm, two weeks after the application of the labelled fertilizer (Table 1). Table 1. 15N recovery in the harvested biomass at the end of the period of regrowth in May and July 2012 and annual yield in 2012. Each value corresponds to the mean of four replicates. Within a column, a common letter indicates no significant difference between types of swards. Within a line, 15N recovery did (*) or did not (ns) differ between May and July 2012. 15 Types of swards Root. depth N2 fixation non-N2 fixing Shallow (s) Lp Deep (d) Ci s and d Lp-Ci N2 fixing Shallow Tr Deep Tp s and d Tr-Tp Mix(fixation) Shallow Lp-Tr Deep Ci-Tp s and d Ci-Tr s and d Lp-Tp s and d Equal stand N recovery (%) May 2012 July 2012 Δ July-May mean Annual yield (g m-2) 16.4 abc 16.5 abc 24.4 a 11.2 ab 25.0 a 16.7 ab -4.8 ns 8.5 * -7.7 * 14.0 bc 20.8 a 20.6 ab 9.8 c 10.6 bc 10.3 bc 10.4 b 21.7 ab 19.1 ab 0.7 ns 11.0 * 8.8 * 10.1 c 16.1 abc 14.7 abc 669 d 1190 a 1225 a 21.0 8.9 14.9 20.7 20.7 13.7 22.4 21.1 16.8 19.4 -7.3 13.6 6.1 -3.9 -1.3 17.3 15.7 18.0 18.7 20.1 820 943 901 1077 1156 a c abc ab ab ab ab ab ab ab * * * ns ns ab abc ab ab ab 709 d 976 bc 929 bc cd bc bc ab a 15 N recovery was not generally smaller with the swards containing only N2 fixing species than with those containing only non-N2 fixing species (lower 15N concentration in the plants but larger yield; data not shown). In accordance, 15N recovery was not reduced by the presence of N2-fixing species in mixtures (P>0.05). No clear effect of rooting depth on the acquisition of fertilizer N (15N recovery) was observed. On average over both seasons, 15N recovery was generally lower with the monocultures of shallow-rooting species (L. perenne and T. repens) Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 723 than with the other swards. This could have been an effect of poorer growth, as biomass production was significantly lower in these two monocultures. In May, a higher 15N recovery was achieved in the L. perenne and C. intybus monocultures than in the T. repens monoculture. While 15N recovery did not significantly differ between seasons with the L. perenne monoculture, it significantly increased between May and July with the C. intybus monoculture. There was also a difference in the dynamics of 15N recovery between the two Trifolium species: in July, 15N recovery remained low in the T. repens monoculture, whereas it significantly increased between May and July in that of T. pratense (Table 1). Thus, with respect to the capture of fertilizer N, L. perenne may be considered a spring species and C. intybus and T. pratense summer species. This different temporal pattern between species might explain why mixed swards containing L. perenne allowed a larger 15N recovery than the others in May (swards with vs. without L. perenne: P<0.05) and that mixed swards containing T. pratense and/or C. intybus allowed a larger 15N recovery than swards without these species in July (swards with T. pratense/C. intybus vs. without: P<0.05). The lowest 15N recovery on average from both seasons was found in the T. repens monoculture, whereas the combination of spring and summer species led to a larger 15N recovery than other swards (spring-summer swards vs. other swards: P<0.05). Mixing species with different temporal pattern of fertilizer N capture, such as L. perenne (spring species) in combination with T. pratense and/or C. intybus (summer species), might reduce the competition for N during a given period of regrowth (Casper and Jackson, 1997) and lead to high total N uptake within the whole growing season. Recovery of fertilizer N was, nevertheless, equal in the monoculture of C. intybus as in the spring-summer species mixtures. As fertilizer acquisition and biomass production are correlated (May: P<0.05, R2=0.60; July: P<0.05, R2=0.51), swards combining spring and summer species conserved high fertilizer uptake and biomass production at each period of regrowth. That may contribute to stabilization of the productivity of swards within a year (Wayne Polley et al., 2007) and lead to a high annual biomass production in these swards (Table 1). Conclusion Recovery of fertilizer N was not reduced by the presence of N2-fixing species in multi-species swards. Mixtures with T. pratense achieved high yields and therefore had a high demand for nitrogen. No clear effect of rooting depth on the acquisition of fertilizer N was observed, although monocultures of shallow rooting species performed poorly both in terms of yield and fertilizer N recovery. Mixing species that are temporally complementary in terms of nitrogen uptake, such as L. perenne in combination with T. pratense and/or C. intybus, allowed for increased fertilizer uptake and substantial biomass production at each period of regrowth. Acknowledgements The research leading to these results has received funding from the European Community’s Seventh Framework Program (FP7/ 2007-2013) under the grant agreement FP7-244983 (Multisward). References Berendse F. (1982) Competition between plant-populations with different rooting depths. 3. field experiments. Oecologia 53, 50-55. Casper B.B. and Jackson R.B.(1997) Plant competition underground. Annual Review of Ecology and Systematics. 28, 545-570. Nyfeler D., Huguenin-Elie O., Suter M., Frossard E.and Lüscher A. (2011) Grass-legume mixtures can yield more nitrogen than legume pure stands due to mutual stimulation of nitrogen uptake from symbiotic and non-symbiotic sources. Agriculture, Ecosystems and Environment 140, 155-163. Wayne Polley H., Wilsey B.J., Derner J.D. (2007) Dominant species constrain effects of species diversity on temporal variability in biomass production of tallgrass prairie. Oikos 116, 2044-2052. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 724 Modelling DM growth of multi-species grassland plots in the Netherlands Holshof G. and van den Pol–van Dasselaar A. Wageningen UR Livestock Research, PO Box 65, 8200 AB Lelystad, the Netherlands. Corresponding author: gertjan.holshof@wur.nl. Abstract Natural and semi-natural grasslands show a great diversity of grass species and varieties, but the diversity is decreasing. Therefore it is of concern to preserve these areas. In the Netherlands farmers are subsidized to manage semi-natural grasslands in a proper way. A grass growth model was calibrated with a large set of data from experiments on intensively managed grasslands and semi-natural grasslands. This paper shows that the Dutch grass growth model as a component of the whole-farm model DairyWise is capable of predicting DM yield of seminatural grasslands and the DM yield of intensively managed grasslands. Keywords: DM yield, grass, grass growth, modelling, multi species, semi-natural grasslands. Introduction One of the objectives of the Multisward project is the development of tools to protect the diversity of grasslands all over Europe. In the Netherlands some semi-natural grasslands can be found, mainly in wet areas. Dutch farmers can get a so-called ‘green subsidy’ for extensively used grasslands. The yield of these grasslands (mainly used as grass silage from the first cut) will be used on dairy farms. To study the possibilities of using the yield on dairy farms, it is helpful to have information on grass production (in time, yield and quality). The Dutch wholefarm model DairyWise (Schils et al., 2007) contains a grass growth model, based on results from intensively managed grasslands (mainly perennial ryegrass (Lolium perenne )) and some data from semi-natural grasslands with different species. The question is, whether it was permitted to use a mixed set of data for calibration and, subsequently, to use the model for intensively managed grasslands as well as for semi-natural grasslands. The aim of this study is to test whether the grass growth model is capable of predicting DM yield of semi-natural grasslands. We hypothesize that multiple species of grass in semi-natural grasslands will show a different growth curve than grass in intensively managed and highly productive grasslands. Materials and methods With a large set of data, the Dutch grass growth model was calibrated in two steps: first the N yield was predicted and secondly the conversion of N yield to DM yield was predicted. Based on the data, an empirical N-uptake and grass growth curve was made using regression analysis and REML for the conversion (Holshof, in prep.). Input variables were soil type, the N supply of the soil, the N application, start date of the cut (day number) and number of growing days. The majority of data (16238 records) were collected on intensively managed grasslands with 85-98% perennial ryegrass (Lolium perenne). However, a part of this calibration set consisted of data from experiments in the 1980s on semi-natural grasslands (2060 records) as collected and described by Korevaar (1986). On 11 locations, about every seven days a fresh plot was mowed during the growing period of a cut. This resulted in six harvests per plot per cut. On an annual base, four cuts were harvested. The plots differed in N application (0 kg N ha-1 for the first cut, 0, 40 or 80 kg N for the second and third cut and 0, 20 and 40 kg N for the last cut). There were four replicates. All plots were mown and the DM yield and the nutrient value of the grass were measured. The locations were mainly wet peat soils (from 1979 to 1982, six experiments). Furthermore there was poor sandy soil (1980-1982, three experiments) and river clay soil (1980, 1982, two experiments). Table 1 provides an overview of the locations and their botanical composition, divided in sub-classes according to De Vries et al. (1942). Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 725 Table 1. Overview of the experiments on semi-natural grasslands and their botanical composition (% good, moderate and inferior grasses and % herbs; De Vries et al., 1942) used for calibration of the grass growth model. Experiment 1 2 3 4 5 6 7 8 9 10 11 Average Year 1979 1979 1980 1981 1981 1982 1980 1981 1982 1980 1982 Average Soil Peat Peat Peat Peat Peat Peat Sand Sand Sand river clay river clay Good 30 17 16 45 11 11 50 61 24 34 26 30 Moderate 47 41 32 7 19 33 12 2 5 41 60 27 Inferior 18 28 36 34 55 34 26 29 37 12 8 29 Herbs 5 14 16 14 15 22 12 8 34 13 6 14 The 30% value for 'good species' in Table 1 consists mainly of perennial ryegrass (8%), roughstalked meadowgrass (Poa trivialis,15%) and meadow fescue (Festuca pratensis, 3%). The moderate-value species (19%) are mainly creeping bent (Agrostis stolonifera,13%) and meadow foxtail (Alopecurus pratensis, 6%). The 28% of inferior grasses consist of velvet bent (Agrostis canina) and marsh foxtail (Alopecurus geniculatis, 14%). The herbs are mainly creeping buttercup (Ranunculus repens). The calibration of the prediction of the DM yield was done in several steps with a REML analysis. A factor ‘grassland type’ (semi-natural grasslands versus intensively managed grasslands) was added to the model and the significance of this factor on the DM yield was tested. Results and discussion The statistical analysis with REML showed no significant effect of the factor ‘grassland type’ on the prediction of the DM yield converted from the N yield (P = 0.75). Therefore, this factor was removed and the final model is made without this factor and used for predictions for both grassland types. In Figure 1, both the predicted DM yield by the calibrated model based on the total set of data and the measured DM yield are given. The dark dots represent the data of the semi-natural grasslands, and the light dots the data of the intensively managed grassland. The model slightly overestimates the DM yield above 6000 kg DM ha-1 cut-1. Harvests on farms, however, are usually less than 4500 kg DM ha-1. The N application of the semi-natural grasslands was between 0 and 200 kg N ha-1 yr-1. Higher N applications could affect the botanical composition in favour of higher appreciated grass species. Figure 1 shows that the model is suitable for semi-natural grasslands in a range of low N input (unfertilized first cut and low fertilization after the first cut). This is in accordance with results of Cowling and Lockyer (1965). The difference between semi-natural grasslands and intensively managed grasslands is not the DM growth pattern, but the feeding value and the digestibility. Korevaar (1986) showed that the digestibility of the grass from semi-natural grasslands is about 15-20% lower than the grass from intensively managed grasslands. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 726 Figure 1. Measured and predicted DM yield for intensively managed and semi-natural grasslands. Conclusions We conclude that the Dutch grass growth model is capable of predicting the DM yield of seminatural grasslands and also that of intensively managed grasslands. The prediction of the DM yield of semi-natural grasslands with this model, however, should be in a range of 0-200 kg N as annual application. The distinction between semi-natural grasslands and intensively managed grasslands could be based on a difference in N application and the digestibility of the grass rather than in a difference of the DM production. Acknowledgements The research leading to these results has received funding from the European Community's Seventh Framework Programme (FP7/ 2007-2013) under the grant agreement n° FP7-244983 (Multisward) and from the Dutch ministry of Economic Affairs (KB-12-006.04-003). References Cowling D.W. and Lockyer D.R. (1965) A comparison of the reaction of different grass species to fertilizer nitrogen and to growth in association with white clover. Journal of the British Grassland Society 20, 197-204. de Vries D.M., ‘t Hart L.M. and Kuijne A.A. (1942) A valuation of grassland based on the botanic composition. Landbouwkundig tijdschrift no. 54, pp. 245-265 (in Dutch). Korevaar H. (1986) Production and feeding value of grass from grassland with restrictions in use and in fertilization for nature conservation. Proefstation voor de Rundveehouderij, Schapenhouderij en Paardenhouderij (PR), Lelystad , Thesis (in Dutch). Schils R.L.M., de Haan M.H.A, Hemmer J.G.A., van den Pol-van Dasselaar A., de Boer J.A., Evers A.G., Holshof G., van Middelkoop J.C. and Zom R.L.G. (2007) DairyWise, a whole farm dairy model. Journal of Dairy Science 90, 5334-5346. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 727 Interest of multi-species swards for pasture-based milk production systems Roca-Fernández A.I., Peyraud J.L., Delaby L., Lassalas J. and Delagarde R. INRA-Agrocampus Ouest, UMR1348 PEGASE, F-35590 Saint-Gilles, France Corresponding author: Remy.Delagarde@rennes.inra.fr Abstract The objective of this 2-year study was to determine the potential of multi-species swards (MSS) for making secure forage availability in grazing dairy systems, and for increasing milk production on a per-ha basis. Four types of sward with increasing number of sward species (from pure perennial ryegrass to a mixture of five species including perennial ryegrass, two clovers, chicory and tall fescue) were compared with four block replicates. Treatments within blocks were simultaneously grazed by four homogeneous dairy herds at the same pasture allowance. Total grazing days per season or per year were unaffected by treatment, but milk output per ha was greater for grass-legume mixtures and MSS compared to pure perennial ryegrass swards. This was related to greater milk production per cow and per day. It is concluded that, under favourable weather conditions particularly in late spring and early summer, advantages of MSS on milk output on a per-ha basis are mainly due to an improved per-cow production rather than to increased pasture herbage production and grazing days/ ha. Keywords: dairy system, forage mixture, milk production, grazing, chicory Introduction Multi-species swards (MSS) with legumes could form the basis of sustainable pasture-based milk production systems. Benefits of ryegrass-clover mixtures over ryegrass monocultures include: their potential to supply greater forage yields, the replacement of mineral N fertilizer inputs by symbiotic N2 fixation by legumes, the reduction of total greenhouse gas emissions per kg of milk, the possibility of extending the pasture growth season, the greater nutritive value and the greater pasture DM intake and milk yield (Lüscher et al., 2013). However, ryegrasslegume mixtures are not well adapted to dry and hot weather conditions. To prevent negative effects of summer drought, several cool-season species may be included in MSS. For instance, chicory is a deep-rooted forb with a high nutritive value and is well-adapted to dry summers (Barry, 1998). Tall fescue is a drought-resistant grass which can show higher annual forage yield than perennial ryegrass due to better tolerance of dry soil conditions. Nevertheless, the effects of MSS on dairy cow performance (Soder et al., 2006) and milk production per ha at the system level are still relatively uncertain. The aim of this experiment was to compare four types of sown swards, differing in botanical composition, on seasonal and annual grazing days and milk production per ha when grazed by dairy cows. Materials and methods The experiment was carried out from September 2011 to August 2013, at the INRA farm of Méjusseaume (48.11° N 1.71° W; Le Rheu, France). The total grazing area (8.7 ha) was divided into 4 block replicates. Each block was divided into 4 paddocks randomly sown in Sept 2010 with either one of the four mixtures described in Table 1. It was hypothesized that the ability to produce forage in all seasons, together with improved tolerance of summer drought would increase with botanical complexity, i.e., from L to LTCF swards. Mineral nitrogen fertilization was similar between treatments (75 kg N/ha/year). Treatments were simultaneously grazed by four homogeneous herds of nine experimental autumn-calving Holstein-Friesian cows using a rotational strip-grazing system Pre-experimental reference periods were carried out at each season to randomize groups of cows. Grazing was organized by cycles and cows grazed nonexperimental pastures as one single herd between 2 cycles. Within a grazing rotation, the four blocks were grazed successively with the two following management rules: 1) same grazing Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 728 calendar (i.e., same dates) between treatments to avoid time lag, and 2) similar pasture allowance (22 kg DM/cow/day > 3 cm) between treatments that define a medium to high grazing pressure. To combine both rules, extra non-experimental cows were needed to adjust the grazing pressure within block and between treatments according to differences in pregrazing herbage mass. Cows grazed 20 h daily and received no feed supplements. Milk production was measured individually in two daily milkings. Pre-grazing pasture mass (> 3 cm), height (plate-meter), botanical composition (manual sorting) and chemical composition (oven drying) were determined before grazing at each block. Grazing days and milk output per ha were calculated considering the grazing calendar and both experimental and extra cows. Milk output per ha was calculated from the milk production of experimental cows and the grazing days per ha. Data were analysed by analysis of covariance. Orthogonal contrasts were used to test the effect of introducing legumes (contrast T: L vs. LT), the effect of MSS compared to single grass-legume mixtures (contrast M: LT vs. LTC/LTCF), and the effect of introducing tall fescue (contrast F: LTC vs. LTCF). Table 1. Sowing rate (kg/ha) of each species and objectives of the four sward type treatments. Treatment No. of species L LT LTC LTCF 1 3 4 5 Lolium perenne Aberstar 35 24 22 11 Trifolium repens Alice 3 3 3 Trifolium pratense Segur 3 3 3 Cicorium intybus Puna 2 1.5 1.5 Festuca arundinacea Callina 11 Treatment interest Control Legumes Deep-rooting forb Drought-resistant grass Results and discussion Pastures were grazed during 13 rotations in 2 years, with an average of 2 cycles in autumn, 2.5 cycles in spring and 2 cycles in summer for each year. Weather conditions were globally good for pasture growth in both years, with rainy springs and medium temperatures in early summer, enabling the maintenance of pasture growth until end of July, including in L swards. On average, legumes represented 20% of DM in LT, LTC and LTCF; chicory represented 30% of DM in both LTC and LTCF; and tall fescue represented 10% in LTCF. All sward types were of good quality (Table 2). Pre-grazing pasture mass and number of grazing days per ha were unaffected by treatments, either at Year level (Table 2) or any season (Figure 1). Table 2. Effect of MSS on pasture mass, sward chemical composition, ave cumulated grazing days/ha/year, average milk yield and milk output/ha/year of pastures grazed during two years. Treatment1 RSD2 Contrasts3 Variable L LT LTC LTCF T M F Pasture mass (kg DM/ha > 3 cm) 2485 2706 2607 2583 976.6 ns ns ns Pre-pasture height (cm, plate-meter) 12.6 13.7 15.8 15.2 4.23 ns * ns Pasture CP (g/kg DM) 187 190 189 200 23.3 ns ns * Pasture NDF (g/kg DM) 535 530 469 484 29.6 ns *** ns Pasture DM digestibility (g/kg) 752 734 776 758 2.5 ns ** ns Grazing days (/ha/year) 749 816 788 770 68.2 ns ns ns Milk yield (kg/cow/day) 17.1 18.1 18.4 18.8 1.24 *** * ns Milk output per year (kg/ha/year) 14020 16123 15579 15568 1432.0 ** ns ns 1 See Table 1; 2 Residual standard deviation; 3 Sig.: *** (P<0.001); ** (P<0.01); * (P<0.05); ns, not significant. This may be explained by the overall good pasture growth conditions which did not reduce the L growth rate, particularly in early summer. High levels of milk output per ha were reached in all treatments and high grazing pressure during the 2 years. Milk output per ha was greater Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 729 (P<0.01) on LT than on L, with no further increase between LT and the two MSS swards (Table 2). This increase between L and the 3 other treatments averaged 1737 kg milk/ha/year (+12%). The greater milk output/ha/year was mainly related to an increase in milk production per cow, which was greater (+1.0 kg/day, P<0.001) on LT than on L swards, as reported by Lüscher et al. (2013). A further increase in milk yield (+0.5 kg/day, P<0.05) was observed on MSS versus LT swards. This result can be explained by a slightly greater digestibility of pasture herbage (Table 2) and probably also to a greater DM intake considering the low NDF concentration and high voluntary intake previously reported for chicory (Barry, 1998). (b) 450 9000 4000 Autumn Spring Grazing season Autumn Spring LTC LTCF L LT LTC Summer LTCF L LT LTC LTCF L LT LTC 0 LTCF 0 L 1000 LT 2000 50 LTC 3000 100 LTCF 150 5000 L 200 6000 LT 250 7000 LTCF 300 8000 LT 350 LTC Milk output/ha (kg/ha) 400 Grazing days/ha (days/ha) 10000 L (a) Summer Grazing season Figure 1. Effect of MSS on (a) grazing days/ha per season and (b) milk output/ha per season and per year. Grazing season distribution: autumn (from mid-September to mid-December), spring (from mid-March to mid-June) and summer (from mid-June to mid-September). Conclusion Advantages of MSS on dairy cow production on a-per ha basis are mainly due to an improved per-cow production rather than to increased pasture production or grazing days per ha. Acknowledgements This research received funding from the European Community's Seventh Framework Programme under the grant agreement n° FP7-244983 (MultiSward). Financial support of Fundación Juana de Vega in the form of the first author’s post-doc fellowship is gratefully acknowledged. References Barry T.N. (1998) The feeding value of chicory (Cichorium intybus) for ruminant livestock. Journal of Agricultural Science 131, 251-257. Lüscher A., Mueller-Harvey I., Soussana J.F., Rees R.M. and Peyraud J.L. (2013) Potential of legume-based grassland-livestock systems in Europe. Grassland Science in Europe 18, 3-29. Soder K.J., Sanderson M.A., Stack J.L. and Muller L.D. (2006) Intake and performance of lactating cows grazing diverse forage mixtures. Journal of Dairy Science 89, 2158-2167. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 730 Influence of ryegrass alone or blended with clover and chicory on feed intake and growth performance of steers Morel I., Schmid E., Soney C., Aragon A. and Dufey P.-A. Agroscope, Institute of Livestock Science ILS, Tioleyre 4, 1725 Posieux, Switzerland Corresponding author: isabelle.morel@agroscope.admin.ch Abstract The purpose of the experiment was to compare the influence of a sward of four species (FS) with that of a perennial ryegrass monoculture (RG) on feed intake and growth of fattening steers. The FS sward was sown with a mixture of 50% perennial ryegrass, and 50% chicory, white clover and red clover (in a ratio of one-third each); in the case of both the FS and the RG sward, half was used for cutting (C) and half as grazed pasture (P). Over 8 cycles, each lasting 2 weeks, two groups, each consisting of 3 Angus (AN) and 3 Limousin (LM) steers, ingested either the FS mixture or the RG, and then the opposite, in turn, with a change of system (C and P) every other cycle. The intake, weighed in the stable (C) and estimated at pasture using the nalkane double-indicator technique, was significantly higher for the FS forage in the case of both C and P (P=0.022). In contrast, the daily growth was highest with the RG forage (P=0.036), pointing to better feed conversion efficiency, expressed in kg DM intake per kg average daily gain (ADG). A lower gut content and consequently a lower body weight at the end of the 14day periods of FS intake is the likely explanation for these unexpected results because the lower NDF content of the FS forage probably increased its transit rate. Keywords: multispecies, sward, pasture, cutting, beef steers, intake, growth Introduction Grassland systems can make a vital contribution to meeting the various challenges facing agriculture today, which include limiting the impact on the environment, preserving or increasing biodiversity, supporting sustainable development and preserving landscapes while continuing to produce high-quality food. Against this background, the use of forage areas in the form of grazed pasture responds to ethical needs (consumer wishes, closeness to nature), technical requirements (absence or restriction of the use of commercial nitrogen fertilizers) and economic needs (Peyraud and Baumont, 2002; Roinsard, 2011). Similarly, the plant species composition of swards may constitute an increasingly important lever in the future, particularly in terms of safeguarding forage systems against climate change. The importance of the complementarity of different species has been noted in many reports that have stressed, for example, the benefits of different root systems (Fustec et al., 2010), the improvement in nutritional value (Rodriguez et al., 2007; Niderkorn et al., 2008) and the better ingestibility (Roinsard, 2011), as well as the more stable yield (Fustec et al., 2008). Among the animals with the potential to optimize the use of these pastures, fattening cattle offer the greatest intake potential after dairy cows. Differences between breeds may be found, as illustrated by Dufey et al. (2002) in a comparison of 6 beef breeds. The objective of this experiment, carried out in the framework of the EU Multisward project (www.multisward.eu), was to compare two different swards (multi-species and monoculture), using two systems (pasture and stable) and steers of two different breeds (Angus and Limousin), in terms of growth performance. Materials and methods The experiment was performed on the experimental farm of Agroscope at Posieux (650 m altitude) from 27 April to 17 August 2012. A mixture of four species (treatment FS) comprising perennial ryegrass (Lolium perenne) cv. Alligator, white and red clovers (Trifolium repens and pratense) cv. Hebe and Dafila, and chicory (Cichorium intybus) cv. Puna II, was sown in August Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 731 2011 on a plot of 2 ha. The quantities sown were 17.5, 2.67, 2.74 and 1.09 kg/ha respectively, corresponding to ½, 1/6, 1/6 and 1/6 respectively of the usual quantity sown to grow a monoculture of the individual species. In parallel, a monoculture of perennial ryegrass (Lolium perenne) cv. Alligator (treatment RG) was sown on an adjacent plot of 2 ha at the rate of 35 kg/ha. Half of the 2 ha of each plot was intended for cutting and the other half as grazed pasture. To ensure that the sward established well, the first application of mineral fertilizer in the spring of the year of the experiment was doubled for the RG treatment to 54 kg N ha-1, compared to 27 kg for the FS treatment. This had been preceded by an application of 40 kg N in the form of organic fertilizer. Subsequently, during the period of the experiment, a total of 81 kg N was applied in three applications of 27 kg each to both of the swards, giving a total for the year of 148 kg N for the FS treatment and 175 kg N for the RG treatment. Six Angus (AN) and six Limousin (LM) steers produced from suckler cows, aged 14.2 months and weighing 447±33 kg, were divided into two groups each consisting of 3 AN and 3 LM cattle. During a total of eight consecutive two-week cycles, the two groups were kept either in a stable or at pasture, with a change of system every other cycle. One of the groups received the RG treatment and the other the FS treatment during the first cycle of a system; then the treatments were reversed for the second cycle. During the periods in the stable, forage was provided ad libitum and the daily forage intake was measured individually using feed containers mounted on electronic weighing machines (Insentec B.V., Marknesse, The Netherlands). The first 3 days of each cycle served as an adaptation period in both systems. The results of intake in the stable are based on the last 11 days of a cycle. In parallel, as well as during grazing periods, the intake of forage was estimated using the n-alkane double indicator technique. In each cycle, n-alkane HC32 was dosed orally in a gelatine capsule twice-daily during 11 days. Faeces were collected by rectal sampling twice-daily, from day 8 to day 11. During grazing, the same DM quantity of forage was offered each day to the two groups of animals. This quantity was calculated on the basis of biomass quantity measured each week on the test plots, taking account of the grass growth measured with an electronic rising plate meter (Jenquip, Feilding, New Zealand). During the first two and the last two cycles of grazing, the quantity offered was equal to 94% and 96%, respectively, of the average intake measured in the stable during the previous cycle. The animals were weighed before the start of each new cycle and at the end of the last cycle. The data were analysed using a General Linear Model (NCSS 2007, Dr. Jerry L. Hintze, Kaysville, Utah). Results and discussion Intake in the stable (ad libitum) was higher in the FS treatment than in the RG treatment (9.23±0.85 compared to 8.84±0.82 kg DM per animal per day) for an average of 4 cycles (P=0.002). The correlation with intake estimated by the HC31-HC32 n-alkane marker pair was more reliable than with the HC33-HC32 marker pair (R2=0.71 compared to 0.59). The analysis of intake over the whole experimental period (4 cycles in the stable and 4 cycles at pasture) was therefore based on the estimate obtained with the HC31-HC32 marker pair. With a difference of 0.4 kg of DM in favour of FS compared to the RG treatment, the significant effect measured in the stable was confirmed (8.28±1.70 compared to 7.88±1.55; P=0.022). These results support the findings of Baumont et al. (2008), Ginane et al. (2008) and Roinsard (2011), who observed an associative effect between different species such as grasses and legumes, which was positive in terms of ingestibility. In the stable the trend was for the DM intake of the LM cattle to be 7% (0.64 kg) lower than in the case of the AN (P=0.095). This difference was less marked over all the cycles combined, with an average intake of 7.91±1.45 kg DM d-1 for the LM breed compared to 8.26±1.78 for the AN breed. This trend is in agreement with the results of Dufey et al. (2002), who observed an 8% lower intake for LM cattle compared to AN cattle (n.s.), and with results published by Faverdin et al. (1997), who concluded that the LM breed has a 10% lower intake Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 732 capacity than the other breeds. With a gap of nearly 320 g ADG in favour of RG, there was a statistically significant difference between sward treatments in the stable cycles (1535±526 g/d for FS compared to 1852±594 g/d for RG; P=0.003). This disadvantage of the FS sward type was confirmed when results were averaged over the eight experiment cycles (stable and pasture), with a reduction in ADG of 21% compared with the RG sward (746±1094 g/d for FS compared to 943±1164 g/d for RG; P=0.036). No significant effect of cattle breed was observed. This lower animal growth for the FS treatment in spite of a higher intake indicates a much less-efficient feed conversion efficiency (6.94 kg DM per kg ADG for the FS treatment, cf. 5.29 kg DM per kg ADG for the RG treatment; P=0.001). A lower gut content and consequently a lower body weight at the end of the 14-day periods of FS intake is the most likely explanation for these unexpected results because the lower NDF content of the FS forage probably increased its gut transit rate. Conclusion Supplying a mixture of four forage species to cattle both in the stable and at pasture resulted in increased intake compared with a monoculture of perennial ryegrass. The results of daily weight gain and feed efficiency showed an apparent negative effect of the four-species mixture which could not be explained conclusively. Feeding trials without changing the type of forage would elucidate this issue and help to optimize the composition of the mixture with a view to improving the efficiency of forage utilization in cattle. Acknowledgement The research leading to these results has received funding from the EC Seventh Framework Programme (FP7/ 2007-2013) under the grant agreement n° FP7-244983 (Multisward). References Baumont R., Aufrère J., Niderkorn V., Andueza D., Surault F. and Peccatte J.-R. (2008) La diversité spécifique dans le fourrage: conséquences sur la valeur alimentaire. Fourrages 194,189-206. Dufey P.-A., Chambaz A., Morel I., and Chassot A. (2002). Performances d’engraissement de bœufs de six races à viande. Revue Suisse d'Agriculture 34, 117-124. Faverdin P., Agabriel J., Bocquier F. and Ingrand S. (1997) Maximiser l’ingestion de fourrages par les ruminants: maîtrise des facteurs liés aux animaux et à leur conduite. Rencontres Recherches Ruminants 4, 65-74. Fustec J., Coutard J.-P. and Gayraud P. (2008). Valeur agronomique de mélanges et d’associations conduits en agriculture biologique. Actes des Journées de l’AFPF. AFPF, Paris, pp. 49-58. Fustec J., Bernard F. and Corre-Hellou G. (2010). Contribution potentielle du lotier corniculé et du trèfle hybride à la productivité de prairies multispécifiques en sols limoneux. Fourrages 204, 247-253. Ginane C., Dumont B., Baumont R., Prache S., Fleurance G. and Farruggia A. (2008). Comprendre le comportement alimentaire des herbivores au pâturage: intérêts pour l’élevage et l’environnement. Rencontres Recherches Ruminants 15, pp. 315-322. Niderkorn V., Lemorvan A., Bergeault R., Papon Y., Baumont R. and Macheboeuf D. (2008). Etude in-vitro des interactions digestives entre graminées et légumineuses. Rencontres Recherches Ruminants 15, p. 279. Peyraud J.-L. and Baumont R. (2002). Qualité des fourrages: de la plante à la ration alimentaire. Fourrages 171, 241-251. Roinsard A. (2011). Les prairies multi-espèces: quelles pistes de recherche/expérimentation explorer? Institut technique de l’agriculture biologique [page consultée en janvier 2014] http://www.itab.asso.fr/downloads/autres-publi/rapport_prairies_roinsard.pdf Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 733 Associative effects between forage species on intake and digestive efficiency in sheep Niderkorn V., Martin C., and Baumont R. INRA, UMR1213 Herbivores, F-63122 Saint-Genès-Champanelle, France Corresponding author: vincent.niderkorn@clermont.inra.fr Abstract There is evidence that multi-species swards can increase biomass productivity and provide a number of ecosystem benefits. However, little is known regarding the possible interactions between forage species that can modulate positively or negatively the use of nutrients by ruminants. The objective of this study was to provide a better understanding of the associative effects between some forage species on intake and digestion parameters. Three sheep-feeding experiments were conducted according to a repeated Latin square design using models of simple forage mixtures under the form of fresh forage or silage, and during which, intake behaviour, DM digestibility and methane emissions were measured. Synergies between cocksfoot silage and red clover silage, and between ryegrass and chicory, were observed on DM intake and eating rate, with optima for the proportion 50:50. For the cocksfoot-red clover association, the synergy was also observed on daily intake of the digestible fraction that can reflect animal performances. No associative effect was observed on methane yield and the lowest emissions were observed for pure red clover and pure chicory. Keywords: associative effects, grass-legume mixtures, chicory, intake, digestion, methane Introduction Diverse pastures are considered as having the potential to better serve production and ecosystem services than species-poor grasslands. However, there is a need for an improved understanding of the utilization of complex grasslands by ruminants to optimize their management. This implies investigations of animal responses to multi-species swards and, in particular, a better understanding of the interactions that can occur between plants on digestion, intake and pollutant emissions as enteric methane. Indeed, the digestibility and feed intake of a combination of forages can differ from the balanced median values calculated from forages considered separately leading to synergistic or antagonistic effects instead of simple additivity (Niderkorn and Baumont, 2009). The objective of this study was to assess the associative effects between some common or less known (e.g. chicory) forage species on intake and digestive processes in sheep. Materials and methods Three sheep-feeding experiments were conducted at INRA Clermont-Ferrand-Theix (France) between 2010 and 2012. Models of simple forage mixtures were designed under the form of fresh forage or silage according common use for the tested species. The combinations tested were i) binary mixtures of silages of cocksfoot (Dactylis glomerata L., cv. Starly) and red clover (Trifolium pratense, cv. Diadem) in five controlled proportions (in % dry matter (DM), 100:0; 75:25; 50:50; 25:75; 0:100); ii) binary mixtures of fresh forages of perennial ryegrass (Lolium perenne L., cv. AberAvon) and white clover (Trifolium repens, cv. Merwi) in the same proportions; and iii) mixtures of fresh forages containing chicory (Cichorium intybus, cv. Puna II): 100% ryegrass, 50% ryegrass + 50% chicory, 100% chicory, 50% ryegrass + 25% white clover + 25% chicory. For each experiment, housed rumen-cannulated sheep were used in a repeated Latin-square design, 4×4 or 5×5 according to the number of treatments. Mixtures were prepared from five-weeks regrowth plants cultivated in pure swards. Each experimental period consisted of an adaptation week to diet followed by a measurement week. During the Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 734 measurement period, chemical composition of plants, intake kinetics and behaviour (Baumont et al., 2004), DM digestibility, and methane emissions using the SF6 tracer method according to the procedure described by Martin et al. (2008) were determined. Data were analysed using the MIXED procedure of SAS® v.9.1 software. Linear and quadratic contrasts were tested to highlight potential associative effects between species. Results and discussion Figure 1. Voluntary dry matter (DM) intake, DM digestibility and methane yield in sheep fed with different proportions of cocksfoot and red clover silages, ryegrass and white clover (fresh), and ryegrass and chicory (fresh). Full lines represent smoothed linear or quadratic responses, and dotted lines represent theoretical responses calculated from values obtained from pure forages. Significant positive quadratic effects were observed between silage of cocksfoot and silage of red clover (P < 0.001), and between fresh ryegrass and fresh chicory (P < 0.05) on voluntary DM intake indicating synergistic effects (Figure 1). The optimums were observed with the proportions 50-50, and the differences between the values measured for the plant combinations and the balanced median values from pure forages were +9.5% and +5.6% in voluntary DM intake for the mixtures cocksfoot-red clover and ryegrass-chicory, respectively. Adding a third species (chicory) did not improve intake and digestive efficiency of the ryegrass-white clover mixture (data not shown). These synergies did not seem to be due to a more efficient digestion, as positive quadratic effects were not observed on DM digestibility (P > 0.05). For the mixture cocksfoot-red clover, a quadratic effect was observed on daily eating rate (P = 0.008) suggesting a greater motivation to eat. For all the mixtures, very strong positive quadratic effects (P < 0.001) were observed on DM intake and eating rates during the main meals distributed in the morning and the afternoon, Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 735 indicating that the diversity in the ration stimulated intake in the short-term. A particularly relevant result was observed with the mixture cocksfoot-red clover as a synergy was also observed on daily intake of the digestible fraction that can be seen as an indicator of animal performances (Niderkorn et al., 2012). No associative effect was observed on methane yield (emissions in g/kg DM intake, P > 0.05). The lowest emissions were observed for pure red clover and pure chicory. Conclusion Taken together, our results indicate that synergy between some species in binary mixtures can be observed on voluntary intake in sheep, with an optimum for the proportion 50:50 without associative effects on methane emissions. Synergy seems to be rather due to a greater motivation to eat than to a more efficient digestion. Acknowledgements The research leading to these results has received funding from the European Community's Seventh Framework Programme (FP7/ 2007-2013) under the grant agreement no FP7-244983 (Multisward). References Baumont R., Chenost M. and Demarquilly C. (2004) Measurement of herbage intake and ingestive behaviour by housed animals. In: Penning P.D. (ed.) Herbage intake handbook, British Grassland Society, Reading, UK, pp. 121-150. Martin C., Rouel J., Jouany J.P., Doreau M. and Chilliard Y. (2008) Methane output and diet digestibility in response to feeding dairy cows crude linseed, extruded linseed, or linseed oil. Journal of Animal Science 86, 26422650. Niderkorn V. and Baumont R. (2009) Associative effects between forages on feed intake and digestion in ruminants. Animal 3, 951-960. Niderkorn V., Martin C., Rochette Y., Julien S. and Baumont R. (2012) Synergy between cocksfoot and red clover silages on voluntary intake and digestive processes in sheep. Grassland Science in Europe 17, 320-322. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 736 Effects of restricting access time to pasture on late lactation dairy cow production Kennedy E., Garry B., Ganche E., O’Donovan M., Murphy J.P., and Hennessy D. Teagasc, Animal & Grassland Research and Innovation Centre, Moorepark, Fermoy, Co. Cork, Ireland Corresponding author: Emer.Kennedy@teagasc.ie Abstract The objective of this study was to examine the effect of restricted access to pasture in the autumn on the milk production, grazing behaviour and dry matter intake of late-lactation springcalving dairy cows. Forty-eight cows were assigned to one of four treatments: full-time access to pasture (22H); two 5-hour periods of access to pasture after a.m. and p.m. milking (2×5H); two 3-hour periods of access to pasture after a.m. and p.m. milking (2×3H;) and alternating between full time and 2x3H access to pasture (2x3HV). Dry matter intake, measured during week 3, was similar for all treatments (15.1 kg/cow/day); consequently there were no differences in milk or milk solids yield (13.2 and 1.15 kg/cow/day, respectively) between treatments. This was due to changes in the cows grazing behaviour as, when access time to pasture was restricted, the grazing intensity increased leading to higher intake per minute and per bite. This indicates that restricted access time to pasture can be implemented on dairy farms in autumn with no reduction in dairy cow production. Keywords: restricted access, grazing, late lactation, dairy cows Introduction Including grazed herbage in the diet of lactating dairy cows in late autumn/winter maintains milk protein output due to the higher energy and crude protein concentrations of grass during this period compared with that of grass silage (Dillon et al., 1998). There is scope to increase the length of the grazing season on Irish dairy farms as currently only 241 grazing days are being achieved (Creighton et al., 2011). Restricted access time to pasture can be used as a tool to increase the length of the grazing season. In spring, restricted access time to pasture has been shown to have no effect on dairy cow production (Kennedy et al., 2009), whereas PérezRamírez et al. (2008) reported that restricting access time to pasture reduced milk yield and composition during the spring/early summer. Studies reporting the effects of restricted access to pasture during the autumn are limited. The objective was to examine the effect of restricted access to pasture in autumn on the milk production, grazing behaviour and dry matter intake (DMI) of late-lactation spring-calving dairy cows. Materials and methods Forty-eight (12 primiparous and 36 pluriparous) Holstein-Friesian dairy cows (mean calving date – 11 March, s.d. 23.1 days) were balanced on parity (2.6, s.d. 1.20), milk yield (13.3, s.d. 2.15 kg), body weight (501, s.d. 60.0 kg), milk fat (45.1, s.d. 4.61 g/kg), protein (36.7, s.d. 2.80 g/kg) and lactose concentrations (43.3, s.d. 1.41g/kg) in a randomized block design. From 26 September to 4 November 2012 cows were randomly assigned to one of four treatments: 22 hours (full time) access to pasture (22H; control); two 5-hour periods of access to pasture after a.m. and p.m. milking (2×5H); two 3-hour periods of access to pasture after a.m. and p.m. milking (2×3H); and alternating between full time and 2x3H access to pasture, with no more than 3 continuous days on any one regime, e.g. Monday – full-time access, Tuesday - 2x3H access, Wednesday - 2x3H access; Thursday – full-time access etc. (2×3HV). All treatments were offered fresh herbage daily at a daily herbage allowance of 17 kg DM/cow/day (>4 cm); no additional supplementation was offered. Treatment groups grazed as separate herds adjacent Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 737 to one another. Pre-and post-grazing sward heights were measured daily. Milk yield was recorded daily and milk composition, body condition score (BCS) and bodyweight (BW) were measured weekly. Dry matter intake was measured during the third week of the study using the n-alkane technique. Grazing behaviour was recorded using IGER grazing recorders. Grazing behaviour, DMI, milk yield and composition were analysed using covariate analysis in SAS, terms for parity, treatment, experimental week and the treatment×week interaction were included in the model. Results and discussion The pre-grazing herbage mass of swards offered to all treatments was similar (1544 kg DM/ha). The 2×3H had a significantly higher post-grazing sward height (4.6 cm; P<0.001) than all other treatments (4.2 cm). Similar to previous experiments (Kennedy et al., 2009) that have investigated the effects of restricted access to pasture, the lack of differences in cow production was due to the change in the cows grazing behaviour. Although cows in the 2x3H and 2x5H had a shorter grazing time they increased their intake per minute and per bite compared to the 22H cows (Table 1) to compensate for their restricted access to pasture. Consequently, there were no differences in grass DMI (15.1 kg/cow/day) or milk yield (13.2 kg/day), milk fat (48.2 g/kg), protein (39.0 g/kg) or lactose content (42.6 g/kg) or milk solids yield (1.15 kg/day). Similarly, there was no effect of treatment on end BW (483 kg) or BCS (2.66). Interestingly, when the 2x3HV were given full-time access to pasture their grazing behaviour was similar to the 22H cows, and when their pasture access time was restricted to two 3-hour periods they behaved like the 2x3H cows. Table 1. Effect of restricting pasture access time of late-lactation dairy cows on grass dry matter intake (GDMI), milk production and grazing behaviour. 22H 2×3HV 2×5H 2×3H S.E.D Sig GDMI (kg/day) 15.5 15.1 15.0 14.9 0.53 0.850 Milk yield (kg day) 13.2 13.3 13.7 12.6 0.55 0.197 MSY (kg day) 1.17 1.14 1.18 1.09 0.051 0.238 a b b c 32.7 0.001 2.66 0.001 0.051 0.001 Grazing time (min/day) 565 GDMI/min (g) 27.6a 33.9b 30.8a 41.3c GDMI/bite (g) a a a b 0.49 460 0.54 487 0.50 358 0.68 22H = 22 h access to pasture; 2 × 3 HV = Alternating between full time and two 3 h periods of access to pasture, with no more than 3 continuous days on any one regime; 2 × 5 H = Two 5 h periods of access to pasture; 2 × 3 H = Two 3 h periods of access time to pasture. Conclusion Late-lactation dairy cows can increase their intake per minute and intake per bite when access to pasture is restricted, resulting in no differences in grass DMI or milk production. There was no effect of alternating access time between 22H and 2×3H on milk production and DMI in the 2×3HV treatment. Therefore, it would be possible for restricted access time to pasture to be implemented on dairy farms in autumn with no reduction in dairy cow production. Acknowledgements The research leading to these results received funding from the European Community's Seventh Framework Programme under the grant agreement n° FP7-244983 (Multisward) and the Irish Dairy Levy Fund. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 738 References Dillon P., Crosse S. and Roche J.R. (1998) The effect of grazing intensity in late summer/autumn on sward characteristics and milk production of spring-calving dairy cows. Irish Journal of Agricultural and Food Research 37, 1-15. Creighton P., Kennedy E., Shalloo L., Boland T.M., O’ Donovan M. (2011) A survey analysis of grassland dairy farming in Ireland, investigating grassland management, technology adoption and sward renewal. Grass and Forage Science 66, 251-264. Kennedy E., McEvoy M., Murphy J.P., O'Donovan M. (2009) Effect of restricted access time to pasture on dairy cow milk production, grazing behavior, and dry matter intake. Journal of Dairy Science 92, 168-176. Pérez-Ramírez E., Delagarde R., Delaby L. (2008) Herbage intake and behavioural adaption of grazing dairy cows by restricting time at pasture under two feeding conditions. Animal 2, 1384-1392. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 739 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 740 Theme 5 special paper Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 741 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 742 Grassland term definitions and classifications adapted to the diversity of European grassland-based systems Peeters A.1, Beaufoy G., Canals R.M., De Vliegher A., Huyghe C., Isselstein J., Jones G., Kessler W., Kirilov A., Mosquera-Losada M.R., Nilsdotter-Linde N., Parente G., Peyraud J.L., Pickert J., Plantureux S., Porqueddu C., Rataj D., Stypinski P., Tonn B., van den Pol – van Dasselaar A., Vintu V. and Wilkins R.J. 1 RHEA Research Centre, Rue Warichet 4 Box 202, 1435 Corbais, Belgium Corresponding author: alain.peeters@rhea-environment.org Abstract Grasslands are represented in an unsatisfactory manner in agricultural statistics. The official definition of grasslands does not include vast areas of grazed shrubby and wooded ecosystems. Temporary grasslands are recorded as ‘Leguminous plants’ and ‘Temporary grass’ which induces doubt on the classification of grass-legume mixtures and often leads to the underestimation of temporary grassland areas. Some terms like ‘meadows’ and ‘pastures’ are often used in an imprecise and misleading way. The term ‘rough grazing’ does not include all types of natural and semi-natural grasslands, especially all types of rangelands, forest pastures and traditional hay meadows. It can thus not represent all species-rich grassland types. Improvements of the current situation are proposed. They consist mainly in: (i) better definitions of grassland terms including for pastures and meadows, permanent, agriculturallyimproved, semi-natural and natural grasslands; (ii) the classification of temporary grasslands in three categories: pure legume sowings, pure grass sowings and grass-legume mixtures; (iii) the classification of permanent grasslands in three categories: agriculturally-improved, natural and semi-natural, no longer used for production; and (iv) the introduction of a new category for grazed fallow land. The paper presents a comprehensive classification of fodder and grassland types in the agricultural area and a multilingual vocabulary. Keywords: Grassland term definition, agricultural statistical classification Introduction A Working Group on ‘Grassland Term Definition’ was set up during the 24th General Meeting of the European Grassland Federation (EGF) that took place on 3-7 June 2012 in Lublin (Poland). It gathered together 22 experts from 13 countries (Belgium, Bulgaria, France, Germany, Italy, Poland, Romania, Slovakia, Spain, Sweden, Switzerland, The Netherlands, United Kingdom). The group is thus representative of the diversity of thinking of European grassland researchers. The purpose of the creation of the Working Group was to support the European Union (EU) Institutions to enable better account to be taken of all the diversity of grasslands into the Common Agricultural Policy (CAP). The EGF adopted a resolution on the reform of the CAP during its Business Meeting. This resolution was sent to a wide group of decision makers of the EU. The participants at the EGF conference decided that a glossary would also be useful in addition to their proposals on grassland in the future CAP. It was thus decided to draft a vocabulary of grassland terms. After contacts with Eurostat, it appeared that this European Commission organization would also be interested in a better definition and classification of grassland terms. This classification could improve the present system of data collection and could lead to a better consideration of the importance and diversity of grasslands in European agricultural statistics. The system should be simple but at the same time be able to collect some new agri-environmental indicators on grasslands and grassland-based systems. This could be the basis of a better recognition of ecosystem goods and services that grasslands can provide. The present text is a trade-off Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 743 between the level of precision that is necessary to reach the objectives described above and the practical aspects related to data collection, in particular from farmers. The present text is largely inspired by the work of Allen et al. (2011) who defined many grassland terms at a global level. In this work, the Working Group adapted these definitions to European specificities. The text is restricted to agricultural grasslands; other types like recreational (e.g. lawns of sport fields) and ornamental grasslands are not considered. Grassland term definitions 1. Fodder areas: Part of the agricultural area that includes permanent grasslands, arable fodder crops and grazed fallow lands. 2. Arable fodder crops: Annual, biennial or perennial species sown on arable land for the production of forage and harvested as green material. They include temporary grasslands, green cereals (C3 species such as oats, barley, spelt, triticale, rye and C4 species such as maize and sorghum), green cereal-other crops mixtures, fodder roots, some Brassicaceae and Compositeae (e.g. sunflower) species. Additional remarks: Crops that are harvested as grain (cereal grain and pulses) and used for animal feeding are not classified in fodder crops. Cereals can represent a resource for mixed farming systems (livestock, grasslands and grain cereal production). This is traditionally the case in Mediterranean areas. Even when sown for grain production, their management can be flexible according to weather conditions prevailing during the growing season. For instance, cereal crops for grain production can be grazed in winter and then harvested for grain production, or grazed only when the predicted grain production will not cover the costs of mechanical harvesting. If cereals are harvested green, by grazing or harvested for silage as immature cereals, they should be defined as fodder crops. C3 cereals such as oats, barley and wheat can sometimes be mixed with other crop types like annual legumes (e.g. pea, vetch) and harvested green. These mixtures should also be considered as fodder crops. 3. Grasslands: Land devoted to the production of forage for harvest by grazing/browsing, cutting, or both, or used for other agricultural purposes such as renewable energy production. The vegetation can include grasses, grass-like plants, legumes and other forbs. Woody species may also be present. Grasslands can be temporary or permanent. Two management categories can be identified:   Meadows: grasslands that have been harvested predominantly by mowing over the last 5 years1 or since the establishment of the sward if it is less than 5 years old. Pastures: grasslands that have been harvested predominantly by grazing over the last 5 years2 or since the establishment of the sward if it is less than 5 years old. 4. Permanent grasslands: Grasslands used to grow grasses or other forage (self-seeded or sown and/or reseeded) and that have not been completely renewed after destruction by ploughing or spraying (herbicide) for ten years or longer. They can be agriculturally improved, semi-natural, natural or no longer used for production. European permanent grasslands can be dominated by:   one or several grass species; one or several grass species and one or several legume species; 1 Where there has been a recent change in the management strategy (more recently than 5 years), the new management type has to be taken into account. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 744     grasses, several forb species and possibly legume species; grass-like species and possibly forb species; shrubby zones (see ‘semi-natural grasslands’ for more information); grazed wooded areas (see ‘semi-natural grasslands’ for more information). Additional remarks: Long-term grasslands provide more ecosystem goods and services than short-term grasslands (e.g. carbon storage, biodiversity levels). In a previously cultivated soil a minimum duration of ten years is necessary in most situations to approach a level of soil organic carbon that is representative of long-term permanent grasslands. A period of ten years is also considered as a minimum for reaching soil biodiversity, and especially higher plant diversity, which is characteristic of long-term permanent grasslands for a given intensification level. The effects of cultivating and reseeding, however, can vary according to the region and type of grassland, and the acceptable frequency of cultivation also varies. For example, in Mediterranean areas, self-seeded permanent grasslands consisting mainly of annuals can be tilled (light harrowing, not deep ploughing) every few years without destroying floral biodiversity. Harrowing is a quite common practice to control scrub invasion, e.g. in Dehesas/Montados. In these conditions, permanent grasslands can be tilled more frequently than once every ten years. 5. Natural and semi-natural grasslands: Low-yielding permanent grasslands, dominated by indigenous, naturally occurring grass communities, other herbaceous species and, in some cases, shrubs and/or trees. These mown and/or grazed ecosystems have not been substantially modified by fertilization, liming, drainage, soil cultivation, herbicide use, introduction of exotic species and (over-)sowing. The occurrence of natural grasslands is not related to human activities, in contrast to the latter. Additional remarks: Occasional liming on acidic grasslands, or the application of very low amounts of organic fertilizers, if not combined with other ‘improvement’ techniques, are not considered to modify habitats substantially. If not associated with higher fertilization or stocking rate, drainage can transform wet semi-natural grassland into mesophilous semi-natural grassland. Although most semi-natural communities give low production, some of them, such as purple moor grass (Molinia caerulea) or tall sedge (Carex spp.) communities, can be quite productive. Natural vegetation types are communities where the vegetative cover is in dynamic balance with the abiotic and biotic (human species excluded) forces of its ecosystem. Semi-natural vegetation is not planted/sown by humans but is influenced by human actions such as grazing, cutting or burning. Previously cultivated areas that have been abandoned and where vegetation is regenerating may also evolve to semi-natural vegetation. In contrast with natural vegetation, semi-natural communities thus need regular anthropogenic disturbances to be maintained. Semi-natural grasslands are usually biodiverse. They include2: 2 These communities correspond for instance to the 1430, 21A0, 4010-4040, 5130, 52, some 53, 6210-6270, 62A062D0, 6310, 6410-6460, 6510-6530, 9070 NATURA 2000 Codes of Annex I of the Council Directive 92/43/EEC of 21 May 1992 on the conservation of natural habitats and of wild fauna and flora, the so-called ‘Habitats Directive’. They include also non NATURA 2000 habitats such as the Cynosurion, Bromion racemosi and Alopecurion. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 745       grazed (pastures) or mown (meadows) grasslands in the plain or low mountain areas including wet areas (riparian vegetation, valleys, flood areas) where grazing and mowing are usually combined in time and/or space; montane and sub-Alpine meadows and pastures; grazed steppes and dry pastures; land crossed during transhumance where the animals spend a part of the year (approximately 100 days) without returning to the holding in the evening; grazed wooded areas (agroforestry areas, Dehesa and Montado type for example). Forestland that produces, at least periodically, spontaneous native understorey vegetation that is grazed and where shrubs and trees are browsed is also considered as grazed semi-natural vegetation, including fire-break lines; grazed/browsed shrubby zones (e.g. heath, maquis, matorral, garrigue). Natural grasslands are also often biodiverse; they cover limited areas in Europe. They include for example3:       Alpine and boreal tundra grasslands (beyond the tree line); rupicolous pannonic grasslands for instance of Hungary; Macaronesian mesophile grasslands from the Atlantic islands (e.g. Azores); steppic grasslands for instance of Romania, Russia and Ukraine; Mediterranean xeric grasslands (e.g. main Mediterranean islands and Stipa grasslands in SE Spain); Grasslands developed on saline soils. ‘Agroforestry’ is the integration of woody perennials, crops and/or grasslands on an area of land. Trees may be single or in groups, inside parcels (silvoarable agroforestry, silvopastoralism, grazed or intercropped orchards) or on the boundaries (hedges, tree lines). Silvoarable systems are extensively used in Mediterranean areas; they include fodder crop rotation under the trees for feeding animals during shortage periods. Agroforestry systems are obtained by planting trees on agricultural land or by introducing agriculture in existing woodland (e.g. silvopasture). A silvopastoral system, like the Dehesa/Montado, has the chief aim of providing food for livestock while taking advantage of the presence of trees (for example for shade, shelter, milder microclimate, 'nutrient pump', strategic browsing and acorn grazing), whilst obtaining a secondary profit from trees in the mid/long term (from, for example, cork, wood, firewood). Silvopastoral systems also include those areas where the understorey of a forest is grazed. It combines grazing with tree production (wood, fruits, fodder) and the maintenance of the forest ecosystem. This system reduces fire risk by controlling inflammable understorey and preserves biodiversity through animal disturbance. In some high forest, the main productive use is wood, but many grazed forests have minimal value for timber or firewood, and grazing is really the most important productive use (e.g. in Mediterranean areas). There are different situations and the official classifications and statistics should recognize this. 3 Annex I of the Council Directive 92/43/EEC of 21 May 1992 on the conservation of natural habitats and of wild fauna and flora, the so-called ‘Habitats Directive’ recognizes 9 habitat types of natural grasslands (8 of them can be grazed: 6110, 6120, 6140-6190 NATURA 2000 Codes). The following habitats should also be considered as natural grazed communities: 1330, 1340, 1410-1430, 1630, 2130-2150, 21A0, 2230, 2240, 2330, 2340, 4060, 9050. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 746 6. Agriculturally improved permanent grasslands: Permanent grasslands on good or medium quality soils, used with more frequent defoliations, higher fertilization rates, higher stocking rates and producing higher yields than natural and semi-natural grasslands. Additional remarks: Agriculturally improved permanent grasslands can be dominated by:    one or several grass species; one or several grass species and one or several legume species; grasses, one or several forb species and possibly legume species. Agriculturally improved permanent grasslands are often classified, in terms of production, on the basis of the proportions of high-, medium- and low-productivity/quality grasses as well as on the proportion of legumes. 7. Permanent grasslands no longer used for production: Areas of permanent grasslands, regardless of the grassland type and the previous use, upon which the produced biomass is no longer used for agricultural production purposes, but which are maintained in good agricultural and environmental condition by appropriate measures. 8. Temporary grasslands: Grasslands sown with forage species that can be annual, biennial or perennial. They are sown on arable land and can be integrated in crop rotations or sown after another grassland vegetation. They are kept for a short period of time, from a couple of months to (usually) a few years. They can be established with pure sowings of legumes, pure sowings of grasses or grass-legume mixtures. Additional remarks: This category includes ‘Leguminous plants’ that are pure stands of leguminous plants or mixtures of predominantly leguminous plants mixed with grasses. Temporary grasslands can be grazed or harvested green as hay or silage. 9. Rangelands: Extensive, large-scale grazed grasslands. Rangelands can be fenced or not but they are usually not fenced, so a shepherd is often needed. Additional remark: Rangelands are dominated by grazed semi-natural vegetation. They may include natural and semi-natural grasslands, shrublands, steppes, tundras, alpine communities, marshes and the understorey of forestland. 10. Grazed fallow lands: Extensively grazed uncultivated land after a cropping episode. The duration of the fallow is typically between one and four years. The land is then cropped again. Additional remark: Fallow lands are very commonly grazed in Mediterranean areas for livestock feeding. They are also important for wildlife (e.g. breeding birds) and soil conservation. 11. Grazed common lands: Permanent grasslands where two or more persons have the right to let their animals grazing concurrently; in some cases these rights are not permanently vested in the same individuals but are allocated from time to time by a body with legal authority to do so. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 747 Additional remarks: Common lands are part of the utilized agricultural area. They can be private or public (state, parish, etc.). They are generally semi-natural, but not always; some common lands have been ‘improved’ by reseeding and fertilization. Rangeland is mostly common land, but not always; it can be in sole use4. Most common land is rangeland, but not always; it can consist of grassland, forest, horticultural or other land. Classification of grassland types into an agricultural statistics system Preamble In the classification system described below, three main ideas are introduced:  Permanent grasslands are described in three main categories: o Agriculturally improved permanent grasslands o Natural and semi-natural grasslands o Permanent grasslands no longer used for production  The existing Eurostat category ‘Fodder crops/Leguminous plants’ has been amalgamated with the category ‘Fodder crops/Temporary grass’ so creating a new category ‘Fodder crops/Temporary grasslands’  A new category is introduced for ‘Grazed fallow land’. Almost no surface data were collected in the past in Europe for the area of ‘Natural and seminatural grasslands’. They differ from ‘Agriculturally improved permanent grasslands’ since they are usually richer in biodiversity because of a lower intensification rate and less modification of their habitats. Statistical data on these natural and semi-natural grassland types could thus be an important biodiversity indicator. Two main types can be defined: ‘Pastures’ (including rangelands, rough grazing, forest pastures etc.) and ‘Traditional hay meadows’. Pastures can be managed in ‘Sole use’ or have the status of ‘Common land’. The two following arguments justify defining a category ‘Temporary grasslands’:   Grassland areas are often underestimated when they correspond to the category ‘Permanent grassland and meadow’ or even to the sum of the category ‘Permanent grassland and meadow’ and the category ‘Fodder crops/Temporary grass’. ‘Fodder crops/Leguminous plants’ are usually not included in grasslands since they are supposed to not include grass! The current category ‘Fodder crops/Leguminous plants’ is unclear. According to the understanding of farmers and national statistical services, they can include pure forage legumes or legume-grass mixtures. There is indeed no clear difference between the following covers: 100% lucerne, 90% lucerne with 10% grass, 80% lucerne with 20% grass, etc. Moreover, a pure sowing of lucerne can include after some time a variable proportion of spontaneously grown grasses. Regarding (red) clover-grass mixtures, the situation is even less clear since red clover is almost never sown in pure stand; it is almost always mixed with grass. Do they have to be classified in the category ‘Fodder crops/Leguminous plants’ or in ‘Fodder crops/Temporary grass’? For farmers, pure lucerne, lucerne-grass mixtures and red clover-grass mixtures are all fodder crops. They use one or the other according to soil characteristics, their experience and other factors. If red clover-grass mixtures are classified in ‘Fodder crops/Leguminous plants’ the problem is that after 2-3 years, they can include much more grass than clover! 4 The term ‘sole use’ is used for land that is not common. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 748 According to countries, legume-grass mixtures can thus be included in the category ‘Fodder crops/Leguminous plants’ or in the category ‘Fodder crops/Temporary grass’. All these problems are removed if the two categories ‘Fodder crops/Leguminous plants’ and ‘Fodder crops/Temporary grass’ are replaced by ‘Fodder crops/ Temporary grasslands’. In order to take into account the variable proportion of legumes in temporary grassland swards, a simple 3-level system is proposed: pure legume sowing; legume-grass mixtures; pure grass sowing. This system could be replaced in the future by a new one with more precise assessment of the proportion of legumes in the sward. That could be done in a later document for temporary and permanent grasslands. Classification based on all land use types of the Utilized Agricultural Area 1. Arable land 1.1. Fodder crops 1.1.1. Temporary grasslands 1.1.1.1. Pure legume sowing 1.1.1.2. Grass-legume mixtures 1.1.1.3. Pure grass sowing 1.1.2. Green cereals 1.1.2.1. Green oats, barley, spelt, triticale, rye and other C3 cereals 1.1.2.2. Green oats, barley, wheat and other C3 cereals mixed with other crop types like annual legumes (e.g. pea, vetch) 1.1.2.3. Green maize and sorghum 1.1.3. Fodder roots (including fodder beet) 1.1.4. Fodder brassicas 1.1.5. Fodder Compositeae: sunflower 1.2. Fallow lands 1.2.1. Grazed fallow lands 1.2.2. Non-grazed fallow lands 1.3. Other crop types 2. Permanent grasslands 2.1. Agriculturally improved permanent grasslands5 2.2. Natural and semi-natural grasslands 2.2.1. Pastures, including rangelands, rough grazing, forest pastures, etc. 2.2.1.1. Sole use 2.2.1.2. Common land 2.2.2. Traditional hay meadows 2.3. Permanent grasslands no longer used for production 3. Permanent crops 4. Other agricultural land such as kitchen gardens Multilingual vocabulary of grassland terms Fodder areas: Futterflächen (DE); Superficies forrajeras (ES); Surfaces fourragères (FR); Superfici foraggere (IT); Voederareaal (NL); Plochy krmovín (SK); Foderarealer (SE) Fodder crops: Futterpflanzen (DE); Cultivos forrajeros (ES); Cultures fourragères (FR); Colture foraggere (IT); Voedergewassen (NL); Krmoviny (SK); Fodergrödor (SE) Grasslands: Grünland (DE); Pastos (ES); Prairies (FR); Prati e pascoli (IT); Grasland (NL); Trávne porasty (SK); Gräsmarker (SE) 5 Almost always in sole use but occasionally common land Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 749 Meadows: Wiesen (DE); Pastos de siega (ES); Prairies de fauche (FR); Prati da sfalcio (IT); Maailand (NL); Lúky (SK); Slåtterängar (SE) Pastures: Weiden (DE); Pastos de pastoreo (ES); Pâtures, prairies pâturées (FR); Pascoli (IT); Weilanden (NL); Pasienky (SK); Betesmarker (SE) Permanent grasslands: Dauergrünland (DE); Pastos permanentes (ES); Prairies permanents, Surfaces toujours en herbe (FR); Prati e pascoli permanenti (IT); Blijvend grasland (NL); Trvalé trávne porasty (SK); Permanenta gräsmarker (SE) Agriculturally improved permanent grasslands: Landwirtschaftlich entwickeltes Dauergrünland (DE); Pastos mejorados (ES); Prairies permanentes améliorées (FR); Prati e pascoli permanenti migliorati (IT); Landbouwkundig verbeterd blijvend grasland (NL); Intenzifikované trvalé trávne porasty (SK); Förbättrade permanenta gräsmarker (SE) Natural and semi-natural grasslands: Natürliches und naturnahes Dauergrünland (DE); Pastos naturales y seminaturales (ES); Prairies naturelles et semi-naturelles (FR); Prati e pascoli naturali e semi-naturali (IT); Natuurlijk en half natuurlijk grasland (NL); Prírodné a poloprírodné trávne porasty (SK); Naturbetesmarker och hagmarksbeten (SE) Permanent grasslands no longer used for production: Aus der Produktion genommenes Dauergrünland (DE); Pastos permanentes no utilizados para la producción (ES); Prairies permanentes plus utilisées pour la production (FR); Prati e pascoli permanenti non più utilizzati per la produzione (IT); Blijvend grasland dat uit productie is genomen (NL); Neprodukčné trvalé trávne porasty (SK); Permanenta gräsmarker tagna ur produktion (SE) Rangelands: Rangelands, Hutungen (DE); Rangelands, Pastos de uso extensivo (ES); Parcours (FR); Rangelands, Spazi vasti a pascolo (IT); Rangelands, woeste gronden (NL); Rangelands, Extenzívne pasienky (SK); Rangelands, Extensiva betesmarker, Fjällbeten (SE) Temporary grasslands: Wechselgrünland (DE); Pastos sembrados temporales (ES); Prairies temporaires (FR); Prati e pascoli temporanei (IT); Tijdelijk grasland (NL); Dočasné trávne porasty (SK); Vall på åker (SE) Grazed fallow lands: Beweidete Brachen (DE); Barbechos y posíos (ES); Jachères pâturées (FR); Riposo pascolativo (IT); Braakliggende gronden die beweid worden (NL); Spásaný úhor (SK); Betad träda (SE) Grazed common lands: Gemeinschaftsweiden (DE); Pastos comunales (ES); Terrains communaux pâturés (FR); Terreni a pascolamento collettivo (IT); Gemeenschappelijke weidegronden (NL); Obecné pasienky (SK); Betad allmänning (SE) Acknowledgement This work would not have been possible without the support of the EU and particularly of the FP7 research project Multisward: ‘Multi-species swards and multi-scale strategies for multifunctional grassland-based ruminant production systems’ (Grant agreement Nr: FP7244983). It is a joint document of the EGF and Multisward. References Allen V.G., Batello C., Berretta E.J., Hodgson J., Kothmann M., Li X., McIvor J., Milne J., Morris C., Peeters A. and Sanderson M. (2011) An international terminology for grazing lands and grazing animals. Grass and Forage Science 66, 2-28. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 750 Theme 5 submitted papers Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 751 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 752 Roles and utility of grasslands in Europe De Vliegher A.1, Van Gils B.1 and van den Pol-van Dasselaar A.2 1 Institute for Agricultural and Fisheries Research (ILVO), Plant Sciences Unit, Crop Husbandry and Environment, BE-9820 Merelbeke, Belgium 2 Wageningen UR Livestock Research, P.O. Box 65, NL-8200 AB Lelystad, the Netherlands. Corresponding author: Alex.Devliegher@ilvo.vlaanderen.be Abstract This paper is a synthesis of the first deliverable of the MultiSward project 'Roles and utility of grasslands in Europe’ (De Vliegher and Van Gils, 2010). This report inventories the importance and spatial localization of grasslands and their multiple functions that benefit humans. In addition to the production of herbage for livestock, grasslands contribute to erosion prevention, biodiversity maintenance of flora and fauna, carbon sequestration, clean surface water and groundwater, and they provide an attractive environment for recreation and leisure activities. Keywords: animal production, biodiversity, grasslands , GHG, soil quality, spatial distribution Introduction Grasslands are the main survival resource for about one billion people worldwide. In industrialized Europe, permanent (33%) and temporary (6%) grasslands cover some 39% of the agricultural area and form the basis of a strong ruminant livestock sector. Next to this, grasslands perform a broad range of functions that benefit humans. In Europe, pressure on land use is high and it is important to establish the possibilities and constraints of combining the functions of grasslands. The first deliverable in the MultiSward project 'Roles and utility of grasslands in Europe' (De Vliegher and Van Gils, 2010) inventories the spatial localization of grasslands and determines the importance, roles and utility of grasslands in Europe. Grassland area and distribution The European grassland area has been significantly reduced during the last 30 years as a result of intensification of grassland and animal production, decrease in cattle population, use of concentrates and soybean in the ration, abandonment, and the effect of EU-policy (Huyghe et al., 2014). Nevertheless, grasslands still cover the largest proportion of the agricultural area. Permanent grassland is very important in Ireland (75% of UAA), UK (58% UAA), Slovenia (58% UAA) and Austria (55% UAA). In terms of number of hectares the United Kingdom (11 million ha), France (9.8 million ha), Germany (4.8 million ha), Italy (4.5 million ha) and Romania (4.5 million ha) represent 62% of the total permanent grassland area in EU-27. The percentage of UAA used as grassland varies considerably between countries and regions. Grassland productivity is affected by several factors: soil characteristics, climatic conditions particularly total and seasonal distribution of rainfall and temperature - altitude, latitude and management. Animal production The principal aim of most European grasslands is to support animal production for milk and meat. The grazing livestock (78.210 million LSU) in EU-27 is divided as follows: 82% are cattle, 13% are sheep and goats and 5% are equidae. In EU-27, 75% of the cows are dairy cattle and 25% are meat cattle. Dairy and meat sector varies strongly from region to region. The top 20% of dairy products is produced on 5% of the territory and the last 10% is produced on 24% of the EU-territory. Sheep and goats are mainly concentrated in the Mediterranean countries, the United Kingdom and Romania. Equines are more common in central and northern Europe. Grazing livestock density is an indicator of the intensity of grassland use and of the pressure of livestock farming on the environment. Manure produced by livestock contributes to greenhouse Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 753 gas emissions and nutrient leaching into water and air. Higher density means a higher amount of manure per ha UAA, which increases the risk of N-leaching. An excessively low livestock density increases the risk of land abandonment and increases the need for industrial fertilizers. Next to grazing livestock, farming practices also influence environmental impact. Organic farming has increased significantly in the period 2000-2011 to 4.1% UAA in EU-27. It is particularly present in regions with extensive livestock production systems based on permanent grasslands. This concerns mountainous and semi-mountainous regions in alpine areas and other parts of the EU. Strikingly, the significance of permanent grassland represents 33% of the UAA in EU-27, whereas it represents 47% of the whole organic area (Anonymous, 2013). Energy production and biorefinery The increasing cost of fossil fuels and environmental concerns about climate change also influence crop-based agro-fuel production and demand. Grassland and fodder area competes with arable land for first-generation bio-fuels like ethanol (maize, wheat, barley, sugar beet), bio-diesel (oilseed rape extraction) and methane (biogas maize). Combustion of grassland biomass is less favourable than other crops or residues like straw because of the NOx, SO2 and HCl emissions. Biorefinery is a concept using green biomass (pasture) as raw material to produce high value biochemicals from the liquid fraction and lower value products or energy generation from the grass-fibre fraction. The grass resource could be natural or cultivated grassland or verge grass that is not needed for traditional use (i.e. forage for herbivores). The general challenges in biomass processing are the transportation costs, the use of dry or wet products, the choice of a central or a mobile unit, and the choice between storage for a year-long period versus a campaign during the growing season. Soil quality and protection Grasslands act as a carbon sink. Several studies have shown a steady increase of soil organic carbon in grassland soils, where over time the carbon levels rise above those of arable soils. However, carbon losses also occur much faster after ploughing up the sward. This highlights the importance of conservation of grassland surfaces and sward longevity for climate-change mitigation. On the other hand, emissions of N2O from grassland soils and manure deposition and CH4 from grazing ruminants partially counterbalance the mitigating effects of carbon sequestration. Grasslands can also mitigate soil erosion and pollution. Grasslands provide a dense rooting system and a permanent soil cover. Ploughing grasslands is seen as one of the causes of increased erosion problems in some European regions. In general, pesticide use and risk of environmental pollution are much lower in grassland systems compared to annual (forage) crops. Nutrients and pollutants left on the grassland surface decompose quickly due to an intensive biological activity. Grasslands thus act as a biological filter for the migration of various chemicals towards the surface and groundwater systems. Grasslands and biodiversity One of the most important functions of (semi-natural) grasslands in Europe is supporting high biodiversity. Grasslands are crucial not only for a great variety of plant species but also for many species of farmland birds, butterflies, beetles, etc. Many species of grasslands are rarely found in other vegetation types. Through variations in management style, climatic and abiotic conditions, semi-natural grasslands show a great variety. Grassland plant communities in the EU are classified into seven main habitats according to EUNIS (2006): dry grasslands, mesic grasslands, seasonally wet and wet grasslands, alpine and subalpine grasslands, woodland fringes and clearings and tall forb stands, inland salt steppes, and sparsely wooded grasslands. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 754 Intensification of land use poses a threat to the botanical diversity in grassland swards, but so does land abandonment, related to the phenomenon of rural abandonment or abandonment of parcels that are of little agricultural value (e.g. on steep slopes or in marginal areas). Ceasing grassland management means that vegetative succession then progresses, with the encroachment of shrubs and other woody species leading to the disappearance of many typical grassland species. Soil biota play an important role in (grassland) ecosystem services and production, e.g. water regulation, nutrient supply. In arable fields the accent is more on bacterial communities whereas fungi fulfil a more important part in grassland soil ecosystems with increasing populations and genetic diversity with sward age. Last, grasslands contribute to an attractive landscape as they are perceived as a rather natural landscape feature and often preferred over other land use such as settlements or arable fields. Especially semi-natural grasslands tend to improve the 'naturalness' of a landscape as they show an increased colour and structure. For this reason, grassland areas are beneficial for tourism and outdoor recreation. Conclusion Grasslands always combine several functions but in different ratios depending on local situations. In addition to the production of herbage for livestock, grasslands contribute to the maintenance of biodiversity, sequester carbon into soil, clean surface and groundwater, prevent erosion and provide an attractive environment for recreation and leisure activities. Acknowledgements This research has received funding from the European Community's Seventh Framework Programme (FP7/ 2007-2013) under the grant agreement n° FP7-244983 (Multisward). References Anonymous (2013) Facts and figures of organic farming in the European Union, pp. 44 http://ec.europa.eu/agriculture/markets-and-prices/more-reports/pdf/organic-2013_en.pdf De Vliegher A. and Van Gils B. (2010) Report on roles and utility of grassland in Europe. Multisward, Collaborative Project Seventh Framework Programme, http://www.multisward.eu/multisward_eng/Outputdeliverables, pp 61. EUNIS (2006) EUNIS Habitat types: http://eunis.eea.europa.eu/habitats-code-browser.jsp?habCode=E#factsheet Huyghe C., De Vliegher A., Van Gils B. and Peeters A. (2014) Grasslands and herbivore production in Europe and effects of common policies. Les Editions Quae,, Centre INRA de Versailles France, pp. 300. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 755 An indicator-based tool to assess environmental impacts of multi-specific swards Plantureux S.1,2, Dumont B.3, Rossignol N.3, Taugourdeau S.1,2,4 and Huguenin-Elie O.4 1 Université de Lorraine, LAE, UMR 1121, Vandoeuvre-les-Nancy, 54500, France 2 INRA, LAE, UMR 1121, Vandoeuvre-les-Nancy, 54500, France 3 INRA UMR1213 Herbivores, 63122 Saint-Genès-Champanelle, France 4 Agroscope, Institute for Sustainability Sciences, 8046 Zürich, Switzerland Corresponding author: sylvain.plantureux@univ-lorraine.fr Abstract We developed a set of indicators to assess the impacts of grassland plant diversity and management on the abundance and diversity of six animal taxa (biodiversity indicators), on soil, water and air quality, and energy consumption (abiotic indicators). The methodology combines multi-criteria decision trees with fuzzy partitioning, allowing to deal with different types of information (qualitative or quantitative, more or less accurate knowledge). Biodiversity indicators were calculated from simple and easily accessible input variables (main management features and botanical composition), according to an analysis of the literature, and could be partly validated at plot scale. Abiotic indicators were calculated as model outputs at farm scale. Combining these two types of indicators allows assessment of the overall environmental footprint of grassland management practices and discussion of the benefits provided by multispecies swards. Here, we report major advances and obstacles closely linked to the available scientific and technical knowledge. Keywords: grassland, diversity, management, indicator Introduction For several decades grasslands have been gradually decreasing across Europe, and many remaining grasslands have become simplified to grass monocultures or simple grass-clover mixtures. On the other hand, research in ecology has shown the benefits of plant diversity for the functioning of grassland ecosystems. This stresses the need for increasing knowledge of the positive economic and environmental values of multi-species swards in agricultural systems. In agriculturally used grasslands, plant species richness varies from simple grass-clover mixtures to semi-natural grasslands that may contain up to 50 species per plot. A key objective of the European research project FP7 ‘MultiSward’ was to conceive, evaluate and promote sustainable ruminant production systems based on multi-species grasslands. One step to reach this objective was the elaboration and the validation of an operational evaluation tool (OET) based on a set of indicators sensitive to plant diversity and grassland management. Materials and methods The development of the OET was based on three sources: 1) expert consultation to decide key environmental impacts and the structure of the OET, 2) analysis of scientific literature to determine how plant diversity impacts on environmental outputs, and 3) real datasets used to either calibrate or validate the OET. Fifteen experts from five countries (CH, F, IRL, N, NL) were involved in the procedure. The OET consists in a decision tree where leaves reflect individual environmental impacts (e.g. nitrate in soil water or web-spider abundance) and branches or nodes reflect aggregated impacts (e.g. soil water quality or spider diversity). The calculation of basic (at leaf level) or aggregated (at node level) indicators results in a score between 0 and 10, (10 corresponds to an extremely favourable environmental impact as compared with other types of grassland management). Elaboration of the tree was performed with a qualitative multi-criteria decision modelling and support system called FisPro (Suárez Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 756 and Lutsko, 1999). It implements a decision tree with fuzzy partitioning model, which makes it possible to account for uncertainty in the decision boundaries between alternatives. Therefore, the nodes can present different kind of data-sources and data uncertainty. Results and discussion The complete evaluation tool is presented in Figure 1. Figure 1. Structure of the indicator-based tool for environmental assessment in MultiSward. Basic indicators in bold, aggregated indicators followed by a star. We initially aimed at calculating all indicators at plot and farm level, but literature analysis and expert consultation showed two major obstacles: 1) the lack of knowledge on the effects of sward plant diversity on 'abiotic impacts' (air, soil, water quality and energy) and 2) the lack of dataset to calibrate and validate biodiversity indicators at farm scale. Nevertheless, the OET in its current form calculates all biodiversity indicators at plot scale, taking account of the effects of sward diversity and management. Abiotic indicators can be calculated at farm scale, taking into account management factors but, very poorly, sward diversity. Indicator inputs are made of discrete or continuous management variables, soil features and botanical composition. Outputs (basic indicators) are then calculated either by models DairyWise (Schils et al., 2007) and Melodie (Chardon et al., 2012) for abiotic indicators, or by the newly developed decision tree for biodiversity indicators. As an example, Table 1 shows the impact of sward diversity and management on two biodiversity indicators calculated for 395 French grasslands across a wide range of pedoclimatic conditions, and housing between 1 and 69 plant species. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 757 Table 1. Biodiversity indicator (0 low – 10 high richness and abundance) for two taxa calculated on a set of 395 French grasslands. Sp = species richness per plot . TG or PG= Temporary or Permanent Grassland Management continuous grazing with medium stocking rate Rotational grazing with low stocking rate and low or high grass height after grazing period One cut and rotational grazing, short duration of grazing periods Plant diversity Spiders diversity and abundance anecic earthworms abundance TG 3.40 6.56 PG <20 sp 5.99 6.01 PG 20-35 sp 6.30 3.99 PG 36-50 sp 6.60 4.13 PG >50 sp 6.15 3.83 TG 5.88 5.35 PG <20 sp 5.76 3.41 PG 20-35 sp 6.15 4.29 PG 36-50 sp 6.32 3.96 PG >50 sp 6.82 4.85 TG 2.76 5.35 PG <20 sp 2.77 4.85 PG 20-35 sp 2.88 4.62 PG 36-50 sp 2.91 5.53 PG >50 sp N/A N/A A first run of validation with real data (indicator value vs biodiversity observations) and with expert opinions on outputs showed promising results. Nevertheless, the main restriction encountered during the development of the OET was the lack of knowledge, which in turns encourages further basic research on grassland functioning. Conclusion The OET developed in the MultiSward project is the first set of 'pressure indicators' (sensu DPSIR indicators typology of the European Environment Agency) sensitive to management and sward plant diversity. Indicators do not have to be compared to models outputs, as the main goal is not to predict a precise and real value but get a score which permits the right decision for the decision maker. The calculation of indicators is based on simple and easily accessible inputs, and its implementation in a free website (http://eflorasys.inpl-nancy.fr) is in progress. The current state of knowledge makes it difficult to fully calibrate and validate all abiotic and biodiversity indicators at plot and farm levels. First results obtained on fauna biodiversity at plot level reveal the usefulness of including plant diversity and simple management inputs to improve the environmental evaluation of grassland based systems. Acknowledgments The research leading to these results has received funding from the EC Seventh Framework Program (FP7/ 2007-2013) under the grant agreement n° FP7-244983 (Multisward). References Chardon, X. Rigolotte C., Baratte C., Espagnol S., et al. (2012) MELODIE: a whole-farm model to study the dynamics of nutrients in dairy and pig farms with crops. Animal 6, 1711-1721. Schils R.L.M, de Haan M.H.A., Hemmer J.G.A., et al. (2007) DairyWise, a whole-farm dairy model. Journal of Dairy Science 90, 5334-5346 Suárez A. and Lutsko J.F. (1999) Globally optimal fuzzy decision trees for classification and regression. IEEE Transactions on Pattern Analysis and Machine Intelligence 21, 1297-1311. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 758 Assessment of ecosystem services provided by grasslands and grasslandbased systems by indicators: a regional perspective Peeters A.1, Stolze M.2, Goliński P.3, Scimone M.4, Moakes S.5, Thorne F.6 and Plantureux S.7 1 Research Centre RHEA, Corbais, Belgium 2 Research Institute of Organic Agriculture (FiBL), Frick, Switzerland 3 Poznań University of Life Sciences (PULS), Department of Grassland and Natural Landscape Sciences, Poznan, Poland 4 University of Udine (UNIUD), Udine, Italy 5 Aberystwyth University, Institute of Biological, Environmental and Rural Sciences (AU-IBERS), Aberystwyth, Wales, United Kingdom 6 Irish Agriculture and Food Development Authority (TEAGASC), Agricultural Economics Department, Dublin, Ireland 7 University of Lorraine, Nancy, France Corresponding author: alain.peeters@rhea-environment.org Abstract The many ecosystem services that grasslands and grassland-based systems can provide should be better quantified in indicator systems. The MultiSward Indicator System (MIS) is inspired by the agri-environmental indicators of the European Commission (EC) calculated at country level. Its structure is based on the DPSIR framework. The MIS focuses on grassland-based ruminant systems. Its scope is thus more restricted than the agri-environmental indicator system of the EC but it tries to be compatible with this system. The work is based on data that are available within EC institutions. The MIS includes two lists: the first one is calculated per farm type for a selection of regions and includes 21 indicators, and the second one is calculated per region for the same selection (all farm types merged) and includes 45 indicators. The MIS should be considered as a pilot project. It should be extended in the future to all European Union regions. Keywords: indicator, ecosystem service, region, livestock, grassland-based systems Introduction Indicators are necessary for justifying, designing and assessing agricultural and environmental policies. There is an obvious lack of indicators for quantifying ecosystem goods and services in grassland-based systems. That is why an indicator set was developed at regional level in the FP7 project ‘MultiSward’. Its objective is: (i) to evaluate the impacts of ruminant stockbreeding systems on the quality and use of natural resources (air, water, soil, energy, biodiversity), and (ii) to assess a large range of ecosystem goods and services provided by grasslands and grassland-based systems. The Multisward Indicator System (MIS) aims to assess these impacts and the provision of these ecosystem goods and services for comparing: (i) ruminant production systems between them and with other farming systems like specialist field crops, within regions, (ii) regions with different proportions of permanent and temporary grasslands, green maize and arable crops in their agricultural area. Materials and methods At region level, the MIS is inspired by recent and effective systems, and particularly by the 28 agri-environmental indicators of the EC (European Commission, 2006) calculated at country level. Its structure is based on the DPSIR (Driving forces – Pressures – States – Impacts – Responses) framework of the European Environment Agency (EEA, 1999). The MIS focuses on grassland-based ruminant systems. Its scope is thus more restricted than the agrienvironmental indicator system of the European Union but it tries to be compatible with this Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 759 system. The work is based on data that are in their majority available within EC institutions to ensure that other actors could use this indicator system in the future. Table 1. List of indicators evaluated at region level. Responses Public policy 1. Share of agri-environment payments in gross farm income (%) 2. Agricultural areas under Natura 2000 (%AA) Technology and skills 3. Farmers’ training levels and use of environmental farm advisory services Market signal and attitudes 4. Area under organic farming Driving Input use forces 5. Energy use Land use changes 6. Evolution of the % of permanent grassland in the AA between 2000 and 2010 (%) Grassland and forage crop patterns 7. Fodder crops and grass (%AA) 8. Common land area (% PG) 9. Total permanent grassland and meadow (%AA) (=PG) 10. Temporary grass (%AA) 11. Temporary grass + leguminous plants (%AA) (=TG) 12. Total permanent and meadow + temporary grassland (%AA) (=MFA) 13. Total permanent and meadow + temporary grassland + leguminous plants (%AA) 14. Green maize (%AA) Livestock patterns 15. Stocking rate (grazing livestock) (LU/ha of fodder crop and grass) 16. Stocking rate (grazing livestock) (LU/ha of TG + PG) 17. Stocking rate (total livestock) (LU/ha of total AA) 18. Stocking rate (total livestock) (LU/ha of fodder crop and grass) 19. Stocking rate (total livestock) (LU/ha of PG+TG) (2010) Farm management 20. Grassland management (in %MFA) 21. Share of grassland grazed (%) Trends 22. Intensification / extensification 23. Grassland annual yield assessment (t DM/ha) per grassland type 24. Annual milk yield/cow (l/cow) 25. Specialisation 26. Risk of land abandonment Pressures Pollution and 27. Gross nitrogen balance (kg N/ha) (total surplus) Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 760 28. Risk of pollution by phosphorus (kg P2O5/ha) (total surplus) benefits 29. Pesticide use (pesticide use in grassland/pesticide use in arable land) (%) 30. Ammonia emissions (kg N-NH3/ha) 31. Greenhouse gas emissions (GWP, total CO2 eq.) (kg CO2 eq./ha) 32. CH4 emissions (kg CO2 eq./ha) 33. N2O emissions (kg CO2 eq./ha) Resource depletion 34. Annual soil erosion risk by water (t soil/ha) Benefits 35. High Nature Value grasslands (%AA) 36. Number of N2000 grassland habitats 37. Proportion of N2000 grassland habitats (%AA) State / Impact Planned / Agricultural biodiversity 38. Shannon diversity index of land use type 39. Shannon equitability index of land use type 40. Shannon diversity index of grazing livestock species expressed in LU 41. Shannon equitability index of grazing livestock species expressed in LU Functional biodiversity 42. Biological nitrogen fixation in grasslands (kg N/ha) Heritage biodiversity 43. Population trend of farmland birds Natural resources 44. SOC density in grassland soils (t CO2 eq./ha) Landscape 45. Landscape - state and diversity (tree lines, hedges, stone wall) (%holdings) Results and discussion The MIS includes two lists: the first one is calculated per farm type for a selection of regions, and the second one is calculated per region for the same selection (all farm types merged). The lists evaluated per farm type and per region include, respectively, 21 and 45 indicators. The MIS is quantified in four (Atlantic, Continental, Alpine, Mediterranean) of the five main European biogeographical regions and in 12 administrative regions selected according to a typology of livestock regions based on Pflimlin et al. (2005). The list of indicators evaluated at region level is presented in Table 1. Conclusion The MIS is developed and ready for use. It includes 45 indicators calculated at region level and 21 indicators calculated at farm type level within regions. It is able to compare individual NUTS2 regions, averages per NUTS2 regions and farm types. Indicators have proved to be sensitive. They correspond to a wide range of topics. The MultiSward and the EC indicator lists are not identical. They have their own specificities and are complementary. The MIS should be considered as a pilot project, but the indicator sets should be calculated in the future in all European Union regions. In the future, efforts should continue on data collection (e.g. on field Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 761 and by remote sensing), prediction (models) and development of missing indicators (water percolation; water use; water quality [nitrate, pesticide]; vegetation types of grasslands; proportion of legumes in temporary and permanent grasslands; grassland yields per vegetation type; grassland management). Acknowledgements This research was funded by the European Community's 7th Framework Programme under the grant agreement FP7-244983 (Multisward). The authors would like to thank EUROSTAT, DG Agriculture and Rural Development (DG AGRI) and the Joint Research Centre (JRC) for their collaboration. References EEA (1999) Environmental indicators: Typology and overview. European Environment Agency, Copenhagen, Technical report 25, 19 pp. European Commission (2006) Development of agri-environmental indicators for monitoring the integration of environmental concerns into the common agricultural policy. Commission Communication (COM final 0508/2006), 11 pp. Pflimlin A., Buczinski B. and Perrot C. (2005) Proposition de zonage pour préserver la diversité des systèmes d’élevage et des territoires européens. Fourrages 182, 311-329. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 762 Threats and opportunities for European grassland areas under different market and policy scenarios Hecht J.1, Moakes S.2, Offermann F.3 and Peeters A.4 1 Research Institute of Organic Agriculture (FiBL) –Department of Socio-economy 2 Aberystwyth University – Institute of Biological, Environmental and Rural Sciences 3 Thünen Institute –Institute of Farm Economics 4 Research Centre RHEA, Corbais, Belgium Corresponding author: judith.hecht@fibl.org Abstract Grasslands are an important landscape element in Europe, but have seen a reduction in area in the last 50 years. To assess future changes under varying price and policy scenarios the FARMIS agri-sector model was utilized to model likely farmer behaviour with regard to grassland use. The results indicate that the total area of grasslands and specifically extensive grassland can be improved if relevant output prices increase or support payments are transferred from arable to lowland/ hill pastoral farms. Keywords: grassland, scenarios, FARMIS, FADN, Switzerland, Germany, Wales Introduction Grasslands are an important landscape component in Germany (DE), and characterize the cultural landscapes of Wales (CY) and Switzerland (CH), covering 30%, 85% and 70% of agricultural land respectively. From an ecological point of view, grassland-based production systems have numerous advantages compared to pure arable or forage crop systems. They provide environmental benefits through capacity to reduce flooding, act as a carbon sink, support biodiversity and offer landscape amenity which can embody a range of values important to society (Bellarby et al., 2008; Dillon, 2011). Furthermore, areas under grasslands can be used exclusively for feeding ruminants, avoiding competition with humans for grains and pulses (Peeters, 2011). The question is how to maintain or increase grassland areas of Europe. Materials and methods The objective of this paper was to assess a range of price variations and policy scenarios utilizing the FARMIS agri-sector model to identify trends and effects. FARMIS is a comparative-static process-analytical programming model for farm groups (Offermann et al., 2005; Deppermann et al., 2013), in which production is differentiated for 27 crop and 15 livestock activities. The model specification is based upon information from the farm accountancy data networks (FADN), supplemented by data from farm management manuals where FADN lacks detail such as feed or fertilizer quantities. Key characteristics of FARMIS are: 1) The use of aggregation factors that allow for representation of the sectors’ income; 2) input-output coefficients which are consistent with the farm accounts; and 3) the use of a positive mathematical programming procedure to calibrate the model to the observed base year levels. Within this paper, four scenarios were analysed to assess the effects of selected changes in market or policy conditions on a grassland based farming systems. The two market scenarios assessed the effects of a 50% increase in the price of inputs (IP_+50) or outputs (OP_+50) relevant to grassland farming. The considered policy scenarios had two common themes, a) ‘+ grass payment’ considered a greater payment to permanent grassland, in DE with re-allocation of budgets, in CY and CH with a ‘top up’ payment, b) ‘+ ext. grass, - arable payment’ considered an extra payment to extensive grassland, with reduced payments to arable land. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 763 Results and discussion Impacts of these four scenarios on production, land use and income can be seen in Table 1. In general, increased input costs (output prices) caused a reduction (increase) in arable and intensive grassland areas but an increase (decrease) in extensively managed grassland areas. In some regions of Germany these decreases caused high rates of grassland abandonment (about 50% Schleswig Holstein and Lower Saxony) due to high output prices. In Switzerland a reduction of intensively managed grassland areas in favour of temporary grasslands was observed, but generally Swiss farmers reacted less intensively to price changes. Additional payments to grassland achieved small effects in Germany and Wales, but a much greater effect in Switzerland, linked to the higher ratio of support payments to production value of farm output. When a greater emphasis was placed upon payments for extensive grassland, at the expense of payments to arable land, a stronger impact was seen in Germany and Switzerland. In Wales, the area of arable land for reallocation of budget was minimal, causing lower payments to intensive grassland. Table 1. Impact of different market and policy scenarios on land use, production and income at the sector level in Switzerland (CH), Germany (DE) and Wales (CY) IP_+50 OP_+50 + grass payment CH DE CY CH DE CY CH DE % Change to Baseline Arable land Level of arable grass, leys Level of fodder maize -2 -3 -4 -2 -4 -5 -3 -5 -8 3 6 30 0 41 22 0 16 18 -18 -48 0 Permanent Grassland Level of intensive grassland Level of extensive grassland 0 0 4 -2 -10 18 -7 -14 5 0 -2 0 0 16 -42 2 25 -55 Number of dairy cows Number of suckler cows Number of beef cattle (CH), bulls (DE), ewes (CY) -4 1 1 -4 -9 -1 51 21 29 -2 -12 -19 -22 37 Production of beef (CH, DE), lamb (CY) Production of milk -7 -4 -12 1 -6 -9 + ext. grass, arable payment CY CH DE CY 0 -1 0 0 0 0 -40 -91 10 0 9 0 -4 -9 0 0 21 2 0 0 1 2 1 8 7 46 23 0 -30 80 -1 0 -5 29 20 0 0 0 0 0 2 14 11 -1 -11 0 -2 22 12 0 0 2 -4 -1 -2 43 51 17 29 13 29 0 0 0 0 1 0 6 15 -2 -1 -1 0 Production value total -2 -6 -8 63 Total Subsidies 0 -2 -7 3 Net added value at factor -8 -28 -97 81 prices per AWU Family Farm income + wages -10 -24 -162 105 per AWU Production ofillustration veal Source: Own based on FARMIS 2013 37 1 69 93 2 378 0 7 8 0 -1 0 1 91 156 7 -4 4 -1 1 3 -1 0 3 81 629 10 0 220 6 0 11 The economic effect of price-change scenarios illustrates how much input prices reduce output (through reduced intensity of farming) and how profitability is influenced negatively. For instance, the number of bulls in Germany, the number of ewes in Wales and the number of beef cattle in Switzerland decreased by -19% and -22% and -12% respectively. The increased output prices have the opposite effect and induce large increases in production value and profitability, though this increase in production has negative environmental effects through greenhouse gas emissions and nutrient balances. The economic effects of policy scenarios highlight how additional grassland payments led to small changes in economic values in Germany and partly Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 764 Switzerland. In Wales the scenario ‘+grass payment’ exceeds CAP budgets (+91%), had a massive effect on profitability (+156%), but only achieved minimal gains in extensive grassland areas (+8%). A support of extensive grassland (+ext. grass, - arable payment) achieved slight gains in profitability at sector level through transfer of income from the stronger arable sector to less profitable extensive grassland sector in all three countries. These policy scenarios had a minimal environmental effect in Germany and Wales, but in Switzerland increased numbers of ruminants caused increased greenhouse gas emissions but reduced nitrogen eutrophication. Conclusion In conclusion, it can be seen that input and output prices have a significant effect on the area and intensity of grassland systems. Use of support payments to encourage less-intensive grassland farming appears to achieve its goal when these payments are transferred from arable areas, but may be more difficult to achieve when there is a limited arable area to transfer budget from, as demonstrated by results in Wales. To prevent extensive grassland abandonment, an increasing proportion of support-payment budgets may need to be transferred from arable and lowland farms to upland areas. In general, policy effects at farm-type level can often vary due to varying intensities of grassland farming. Acknowledgement This research was funded by the European Community's 7th Framework Programme under the grant agreement FP7-244983 (Multisward). References Bellarby J., Foereid B., Hastings A. and Smith P. (2008) Cool Farming: Climate impacts of agriculture and mitigation potential. Campaigning for Sustainable Agriculture. Greenpeace. Deppermann A., Grethe H. and Offermann F. (2013) Distributional effects of CAP liberalisation on western German farm incomes: an ex-ante analysis. European Review of Agricultural Economics pp. 1–22. Advance Access published 20 Nov 2013. doi:10.1093/erae/jbt034. Dillon P. (2011) The characteristics of sustainable grass-based ruminant production systems are identified. Deliverable D4.1 of the EU-project MultiSward. Offermann F., Kleinhanß W., Hüttel S. and Küpker B. (2005) Assessing the 2003 CAP Reform Impacts on German Agriculture Using the Farm Group Model FARMIS. In: Arfini, F. (ed.) Modelling Agricultural Policies: State of the Art and New Challenges. Proceedings of the 89th European Seminar of the EAAE, Parma, Italy, February 35, 2005. Parma: Monte Universita Parma Editore: 546-564. Peeters A. (2011). Socio-economic and political driving forces. Deliverable D5.1 of the EU-project MultiSward. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 765 Appreciation of the functions of grasslands by European stakeholders van den Pol-van Dasselaar A.1, Goliński P.2, Hennessy D.3, Huyghe C.4, Parente G.5 and Peyraud J.-L.6 1 Wageningen UR Livestock Research, P.O. Box 65, NL-8200 AB Lelystad, the Netherlands, 2 Department of Grassland and Natural Landscape Sciences, Poznan University of Life Sciences, Dojazd 11, 60-632 Poznań, Poland, 3 Teagasc, Animal and Grassland Research and Innovation Centre, Moorepark, Fermoy, Co. Cork, Ireland, 4 INRA, Centre de recherche de Poitou-Charentes, Lusignan 86600, France, 5 Department of Agriculture and Environmental Science, University of Udine, Via delle Scienze 208, 33100 Udine, Italy, 6 INRA, UMR-1348, Joint Research Unit PEGASE, F-35590 St Gilles, France. Corresponding author: agnes.vandenpol@wur.nl Abstract The European project MultiSward aimed to increase the reliance of farmers on grasslands and on multi-species swards for competitive and sustainable ruminant production systems. Active participation of stakeholders was one of the key objectives of the project. The aim of the current study was to get an insight into the importance of grasslands for stakeholders in Europe. An online questionnaire on the functions of grasslands was developed in eight languages and 1959 valid responses were obtained. Belgium, France, Ireland, Italy, the Netherlands and Poland were the countries with the highest responses. All of the stakeholder groups that were identified as being important in the stakeholder analysis responded to the questionnaire. When asked about the importance of different aspects of sustainability, stakeholders, on average, valued economic aspects the highest, followed by ecological aspects and finally, social aspects. There were, however, differences between countries and stakeholder types. The results of the questionnaire show that individual functions of grasslands are highly recognized and appreciated by all relevant stakeholder groups. We conclude that the large European grassland area is considered by all stakeholders to be a valuable resource that is essential for economy, environment and people. Keywords: grasslands, multifunctionality, stakeholder, sustainability, questionnaire Introduction Grasslands, with their multifunctional roles, can provide a good basis for developing sustainable production systems in the long term (Peyraud et al., 2010). The project MultiSward (www.multisward.eu, 2010-2014) aimed to secure optimal acreage and utilization of grasslands in Europe, to highlight the benefits of grasslands and to conceive, evaluate and promote sustainable ruminant production systems, based on the use of grasslands with a high level of multi-functionality, to increase simultaneously the competitiveness of ruminant production systems and provide environmental goods and biodiversity preservation. During the last 40 years the European grassland area has significantly reduced, by 15 M ha in favour of the production of fodder maize and other annual crops (FAOSTAT, 2011). Even marginal grasslands tend to be abandoned, particularly in mountainous and Mediterranean areas, where they can be of crucial importance for preserving biodiversity, protecting soils against erosion and maintaining the local population density. The reduction has differed between countries. Losses were high in Belgium, France, Italy and the Netherlands while the grassland area remained almost stable in Luxembourg and the United Kingdom. In 2007, permanent grasslands covered over 57 million ha in the EU-27 and temporary grasslands about Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 766 10 million ha, which represents 33% and 6%, respectively, of the total utilized agricultural area (UAA) in the EU-27. In order to contribute to the overall objective of MultiSward, stakeholder requirements and expectations with respect to multi-functionality of grasslands within Europe should be known, because a better understanding of stakeholders’ perspective of grasslands leads to a better understanding of the importance of grasslands. Prior to the MultiSward project, the requirements and expectations of stakeholders with respect to the multi-functionality of grasslands in Europe were not known. Therefore, an active participation of stakeholders was one of the key objectives of the MultiSward project. An initial inventory was made of the requirements and expectations of stakeholders with respect to the multi-functionality of grasslands in Europe (Van den Pol-van Dasselaar et al., 2012 and 2013). The aim of the current study was to give new insights into the importance of grasslands for stakeholders in Europe. Materials and methods An international team of representatives from Ireland, the Netherlands, France, Italy and Poland was established representing Atlantic, Mountainous, Mediterranean and Continental regions. The work started with a stakeholder analysis (Pinxterhuis, 2011). The identification of stakeholders is an important first step in stakeholder consultation. Stakeholders are usually defined as those who either affect or are affected (e.g. Freeman, 1984). In the case of grasslands, this means that stakeholders are those who affect grasslands or are affected by grasslands. Both aspects were taken into account when prioritizing the stakeholders in the stakeholder analysis. A good stakeholder analysis is essential (Reeda et al., 2009), since only by understanding who has a stake in grasslands, can the appropriate stakeholders be effectively involved in the stakeholder consultation. The stakeholder analysis was undertaken to identify the people or institutions having a clear stake in the multi-functional use of grasslands, or being in the position to play an important role in the development and implementation of new management options for multi-species swards (e.g. can directly benefit, has political power, is executing governance, is economically dependent, etc.). The most important stakeholders were the traditional foursome of primary producer, policy maker, researcher and advisor. NGOs for nature conservation and for protection of the environment were also considered important, together with industry (mainly processing and seed industry) and education. Following the initial stakeholder analysis, the international stakeholder team undertook several studies, including national and international meetings. A questionnaire on the functions of grasslands was developed in eight languages: Polish, Dutch, Italian, French, English, German, Danish and Swedish, using SurveyMonkey (www.surveymonkey.com). The questionnaire included two main questions on the importance of grasslands in Europe. First, respondents were asked for their opinion on sustainability. This term covers economic, environmental and social issues (profit, planet, people). Respondents to the questionnaire were asked to divide 10 points across these three aspects of sustainability, giving most points to the one they considered the most important aspect (e.g., 4, 3, 3 if they considered that ecological and social aspects are of equal interest and that economy is slightly more important). Second, the respondents were asked to score 42 predefined functions of grasslands of grasslands for importance in their region (1 = not important; 5 = very important). These functions are examples of the ecosystem services that grasslands deliver. The concept of ecosystem services provides a good insight into the benefits that humankind gains from its interaction with natural resources, in this case with grasslands. The Millennium Ecosystem Assessment report (MEA, 2005) distinguishes four groups of ecosystem services: (i) provisioning services: products obtained from ecosystems, e.g. production of food, water, (ii) regulating services: benefits obtained from the regulation of ecosystem processes, e.g. control of climate and disease, (iii) cultural services: non-material benefits that people obtain from Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 767 ecosystems through spiritual enrichment, cognitive development, reflection, recreation, and aesthetic experiences, e.g. recreation and beauty of the landscape, and (iv) supporting services: ecosystem services that are necessary for the production of all other ecosystem services, e.g. nutrient cycles, crop pollination. Research partners of MultiSward actively distributed the questionnaire in Europe to stakeholders. Furthermore, several relevant associations with members from different stakeholder groups were approached, such as the national Grassland Societies in the respective countries. The questionnaire was available online from spring 2013 and closed at the end of 2013. The sustainability results were analysed using GenStat (VSN International, 2013). The observed points out of a total score of 10 have been treated as pseudo-binomial data, taking the variance to be proportional to binomial variance (McCullagh and Nelder, 1989). Differences between countries, stakeholder type, gender and age in preference of the respondents have been assessed by linear logistic regression analysis of the observed points using a logistic model with main effects. Main effects have been tested with approximate F-tests; differences between countries, stakeholder type, gender and age have been tested with approximate t-tests on all pairwise differences of fitted marginal means on the underlying logistic scale. Results and discussion At the time of closing the questionnaire, 1959 valid responses had been obtained for the question on sustainability aspects. The majority of respondents (1798) also provided answers to the question on the different functions of grasslands. The respondents originated from 27 different countries in Europe. There were six countries with more than 200 responses: France (21% of the total responses), Italy (17%), Ireland (13%), Poland (12%), Belgium (11%) and the Netherlands (11%). The remaining countries in the rest of Europe were grouped (15%). All the relevant stakeholder types described in Pinxterhuis (2011) responded to the questionnaire. Responses from researchers, advisers and farmers accounted for a high proportion of the total: 22%, 19% and 17% of the total responses, respectively. The contribution of policymakers was much lower (6%), but given the fact that there are obviously fewer policy makers and they are often less eager to respond, we were satisfied with this percentage. Other groups were students (16%), educators (6%), industry (5%), e.g. feed industry, dairy industry, seed industry, and finally NGOs (3%). The remaining group, which mainly consisted of people who identified themselves as consumers, press, in between jobs etc. was 6%. Some people identified themselves as belonging to two groups. In those cases, they were classified into the group which they mentioned first. With respect to age and gender, responses were obtained from all age categories. One-third of the respondents were female and two-thirds were male. The percentage of female respondents in the younger age groups was higher than the percentage of female respondents in the higher age groups. Finally, it was observed that the majority of the respondents had received a high level of education, as two-thirds of the respondents had attended university. It is to be expected that respondents in a number of stakeholder groups have a position that requires a relatively high level of education. The groups 'farmers', 'students' and the 'rest' group had a lower level of education. A further explanation might be that well educated people may be more willing to respond to a questionnaire. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 768 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Social Ecological Economic 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 100% 100% 90% 90% 80% 80% 70% 70% 60% Social 50% Ecological 40% Economic 30% Ecological Economic 60% Social 50% Ecological 40% Economic 30% 20% 20% 10% 10% 0% Social 0% <20 20-30 30-40 40-50 50-60 60+ Female Male Figure 1. Importance of economic, ecological and social aspects of sustainability for a) different countries, b) different stakeholder types, c) different age and d) different gender (total of economic, ecological and social aspects equals 10 for each group) (n=1959). When people were asked to divide 10 points over economic, ecological and social aspects of sustainability, on average, economy was valued the highest (3.7) followed by ecology (3.4) and social aspects (2.9). The differences were significant, but these means also show that all aspects of sustainability were considered to be important. The effect of country, stakeholder, age and gender is shown in Figure 1. Obviously, respondents only had 10 points to divide. This means that the effects on economic, ecological and social aspects are entangled. When a respondent, for instance, decides to give more points to social aspects, there will be fewer points left for the other two aspects. We therefore looked for pairwise significance. When analysing economic, ecological and social aspects, the effects of country and stakeholder type were significant (P<0.001). The effect of age and gender was less consistent; after having accounted for the remaining main effects of country and stakeholder type, the age and gender effect was often no longer significant. Italy showed the lowest ranking for economy, followed by Poland and France (Figure 1a). Belgium, Ireland and the Netherlands had a high ranking for economy. In accordance with this, Italy, France and to a lesser extent Poland, showed higher ranking for social aspects than the other countries. Ecological aspects were scored highest for Italy and Poland. Concerning the different stakeholder types (Figure 1b), farmers, industry and to a lesser extent advisers, showed the highest ranking for economy; the social aspects were valued the highest by NGOs and policy makers and lowest by industry. Ecological aspects were valued highest by education, research and students, and lowest by farmers. There was hardly any difference in the ranking of social aspects in relation to age (Figure 1c). It seems that economy is ranked a bit higher when people get older at the cost of ecological aspects. However, differences were not significant. Females ranked economy lower than males mainly to the benefit of ecological aspects (Figure 1d). Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 769 When people were asked to value different functions of grasslands, it was clearly shown that the different functions of grasslands are highly recognized and appreciated by all relevant stakeholder groups (see papers on appreciation of the functions of grassland by Belgian, Dutch, French, Irish, Italian and Polish stakeholders elsewhere in this volume, and Van den Pol-van Dasselaar et al. (2014) for a summary of all results). It is therefore important that future policies continue to support the conservation of grasslands. Scenarios with less grassland will lead to an overall decrease in total ecosystem services delivered, since grassland is the only land-use option which is capable of delivering that large a number of ecosystem services simultaneously. Conclusion MultiSward provided an insight into the appreciation of the different functions of grasslands in Europe. It clearly showed that the different functions of grasslands are highly recognized and appreciated by all relevant stakeholder groups. The large European grassland area appears to be essential for economy, environment and people. We conclude that all stakeholders consider grasslands to be a valuable resource in Europe. Maintaining or increasing the grassland area and thus securing the importance of the different functions and services of grasslands in Europe is a challenge for the coming years. It is, however, important since it will ensure the continuation of different ecosystem services being delivered simultaneously by multifunctional grasslands. Acknowledgements The research leading to these results has received funding from the European Community's Seventh Framework Programme under grant agreement n° FP7-244983 (Multisward). References FAOSTAT (2011) http://faostat.fao.org/. Food and Agriculture Organization of the United Nations, Rome, Italy. Freeman E.R. (1984) Strategic Management. A Stakeholder Approach. Boston: Pitman. 276 pp. McCullagh P. and Nelder J.A. (1989) Generalized linear models (second edition). Chapman and Hall, London. MEA (2005) Ecosystems and Human Well-being: Current State and Trends, Volume 1. 901 pp. Peyraud J.L., van den Pol-van Dasselaar A., Dillon P. and Delaby L. (2010) Producing milk from grazing to reconcile economic and environmental performances. Grassland Science in Europe 15, 865-879. Pinxterhuis J.B. (2011) Report on appreciation of the current and future functions of grasslands in Europe and identification of implementation gaps between today and future multi-functionalities, as seen by international stakeholders. Report MultiSward, Wageningen UR Livestock Research, Lelystad, the Netherlands Reed M.S., Graves A., Dandy N., Posthumus H., Hubacek K., Morris J., Prell C., Quinn C.H. and Stringer L.C. (2009) Who's in and why? A typology of stakeholder analysis methods for natural resource management. Journal of Environmental Management 90, 1933-1949. Van den Pol-van Dasselaar A., Goliński P., Hennessy D., Huyghe C., Parente G., Peyraud J.L. and Pinxterhuis J.B. (2012) Stakeholder’s requirements and expectations with respect to multi-functionality of grasslands in Europe. Grassland Science in Europe 17, 762-764. Van den Pol-van Dasselaar A., Goliński P., Hennessy D., Huyghe C., Parente G., Peyraud J.L. and Stienezen M.W.J. (2013) Appreciation of current and future functions of grassland by international stakeholders in Europe. Grassland Science in Europe 18, 219-221. Van den Pol-van Dasselaar A., P. Goliński, D. Hennessy, C. Huyghe, G. Parente and J.-L. Peyraud (2014) Évaluation des fonctions des prairies par les acteurs européens. Revue Forages (in press). VSN International (2013) GenStat for Windows 16th Edition. VSN International, Hemel Hempstead, UK. Web page: GenStat.co.uk Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 770 Theme 5 posters Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 771 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 772 Effect of grassland management in autumn on the mineral N content in soil De Vliegher A. and Vandecasteele B. Institute for Agricultural and Fisheries Research (ILVO), 9820 Merelbeke, Belgium Corresponding author: Alex.Devliegher@ilvo.vlaanderen.be Abstract In this experiment we studied whether cutting pastures in autumn affects the mineral N content in the soil profile. These results were compared with soil-N content after grazing. The intention was to evaluate whether this may reduce the risk of N leaching. Twenty-seven parcels with an intensive, mixed management were selected in 2010-2012. Two cutting frequencies were applied: a single cut in October (n=27) or a cut in September and in October (n=15). Cutting in October instead of grazing resulted in a significant decrease in NH4-N content in the 0-30 cm soil layer in October and November and a significant decrease in total mineral N content in the 0-30 cm soil layer in November. NO3-N, which is sensitive to leaching, was not significantly influenced in this period. Cutting twice versus once had no significant effects on NO3-N, NH4N, or mineral N content in the soil. Keywords: grazing, cutting, intensive management, mineral N content soil Introduction Dairy farming in Flanders, Belgium is intensive and farmers are allowed to apply 235 kg Navailable e.ha-1 year-1 on grassland on sandy soils (245 kg on non-sandy soils) under the terms of the Flemish Manure Decree. Included within this definition of fertilization are slurry, chemical fertilizers and dung and urine deposited during grazing. The level of nitrate-N residue in the soil (0-90 cm) at the end of the growing season is used as an indicator for the risk of nitrate-N leaching. The aim of the experiment was to study the effect of cutting in autumn in comparison with grazing to evaluate the effect on mineral N content (nitrate-N and ammonium-N) in the soil profile. Our intention was to evaluate whether the risk of N leaching could be reduced. Materials and methods The study was performed in Flanders on permanent grasslands dominated by perennial ryegrass (Lolium perenne L.). In this experiment, 27 parcels with mixed management from the beginning of the season until the end of August were selected on intensive dairy farms. Some pastures were followed during 2 or 3 years. Three pastures per soil texture – sand, sandy loam and claywere examined per year. No slurry or chemical fertilizers were applied later than 1 September. The pastures were mainly grazed by dairy cows in production - with supplementary feeding or by heifers, until October-November. Grazing time, animal type and numbers of animals grazing were recorded, and N input from deposition of dung and urine was calculated (VLM, see below). In 2010 we set up 3 enclosures per pasture (4 m x 8 m) to prevent grazing on that area for the rest of the growing season. In 2011 (9 parcels) and 2012 (6 parcels) the enclosures had a larger area (6 m x 8 m) and a strip of 1.4 m x 6 m was mown (pre-cut) within each enclosure because, at that time of year, these plots could generate a large biomass and cutting would probably stimulate grass growth and N uptake. Within the enclosures, one (2010) or two (2011, 2012) strips (1.4 m x 6 m = 8.4 m2) were mown in mid October. Dry matter yield of the grass was determined and herbage samples were analysed for N content to calculate the N export by the mown grass. Dry matter yield was not determined on the grazed area. At the end of the growing season, around 15 November, one (2010) or two (2011, 2012) strips were cut within the enclosures (8.4 m2) to estimate the amount of N present in the grass. Within and outside the enclosures, soil was sampled 3 times: at the beginning of the experiment, in mid October at the time of cutting, and in mid November. In 2011 and 2012, soil samples were taken within the enclosures on the two strips, i.e. with and without a pre-cut. The soil sample outside Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 773 the enclosures was taken near the enclosure to ensure that this soil had conditions at the beginning of the experiment similar to those within the enclosure. Samples were taken in 3 layers: 0-30 cm, 30-60 cm and 60-90 cm, with 6 drillings (drilling density on the mown strip: 7 drillings per 10 m2). The sampling methodology was the same for each pasture and each time of sampling. The soil samples were analysed for NO3-N and NH4-N according to ISO 142562:2005. The nitrogen input by the grazing cattle was calculated by the number of grazing days and excretion figures from VLM (http://www.vlm.be/landtuinbouwers/mestbank/dierlijkeproductie/Berekeningvandenettouitscheidingvanrunderen/Pages/default.aspx) Results and discussion Grazing cows excrete nitrogen on the pasture in a very heterogeneous way in space and time. The N efficiency is very low, especially during grazing in autumn. This makes it difficult to sample adequately and can help to explain the high variability in NO3-N and NH4-N concentrations in the soil between the 3 measurements per treatment, and per sampling period within a pasture. As a result, a pairwise t-test was used on the average of the 3 measurements per pasture in each sampling period to compare cutting without pre-cut with grazing (n=27) or to compare cutting with and without a pre-cut (n=15). Cutting versus grazing Please refer to Table 1 for relevant results. Table 1. Mineral N content in the soil in September (start of experiment), October and November under grazing and cutting (in the enclosures) on permanent grassland in Flanders Mineral N-content in the soil profile 0-90 cm NO3 -N kg.ha Period Treatment Parcels 0-30 Sept. Grazing (G) 27 25 1 Cut (C) 27 P value Oct. Grazing (G) 27 1 Cut (C) 27 P value Nov. Grazing (G) 27 1 Cut (C) 27 P value *: P< 0.05 -1 NH4 -N kg.ha-1 30-60 60-90 0-90 15 9 49 0-30 13 30-60 60-90 7 NO3 -N + NH4 -N kg.ha-1 0-90 0-30 4 24 38 30-60 60-90 0-90 22 13 73 23 16 9 48 13 5 5 23 36 21 14 71 ns ns ns ns ns ns ns ns ns ns ns ns 18 14 12 43 26 7 6 39 44 21 18 82 16 16 16 47 14 5 5 23 30 21 21 71 ns ns ns ns 0.024* ns ns 0.049* ns ns ns ns 22 16 12 50 21 5 4 31 43 21 16 81 21 31 22 19 72 ns ns ns 19 17 15 51 12 5 4 ns ns ns ns 0.034* ns ns 0.038* 0.029* ns: not significant In September, the NO3-N, NH4-N and the mineral-N concentration in the soil were the same (no significant differences) inside (cutting) and outside (grazing) the enclosures at the beginning of the experiment for each layer (0-30 cm, 30-60 cm, 60-90 cm, 0-90 cm). The total mineral N content in the 0-90 cm soil layer was on average 72 kg N ha-1 and was composed of 49 kg ha-1 NO3-N and 23 kg ha-1 NH4-N. In the middle of October, on the cutting date, the only significant difference (P<0.05) was for the NH4-N concentration in the 0-30 cm and 0-90 cm soil profiles. The NH4-N concentration in these soil layers was higher for grazing than for cutting. No significant differences in NO3-N concentration were found between grazing (on average 43 kg ha-1 in 0-90 cm soil) and cutting (on average 47 kg ha-1 in 0-90 cm soil) for any of the soil layers. In mid-November, at the very end of the grazing season and at the end of the official monitoring period for soil sampling according to the Nitrate Directive (91/676 EU) in Flanders, the only significant differences were found for the NH4-N content in the 0-30 cm and 0-90 cm soil layers and for the mineral-N content in the 0-30 soil layer. There were no significant differences in the NO3-N content in any of the soil layers between cutting and grazing. The average values in Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 774 the 0-90 cm layer were very close to the data at the beginning of the experiment (September): for grazing in September and November, 50 kg ha-1 and 49 kg ha-1, respectively; and for cutting in September and November, 51 kg ha-1 and 48 kg ha-1, respectively. For the total mineral-N content in the 0-90 cm layer there was a tendency towards higher N content for grazed pastures in November (81 kg N ha-1 versus 73 kg N ha-1 in September). A higher NO3-N concentration in the soil on the grazed area was expected because grazing intensity and consequently deposition of urine (N) on the pasture was high but very variable between the parcels, depending on stocking rate and number of grazing days. The average N deposition by the grazing cattle in the period September-November was 92 kg ha-1 (± 50 kg N ha-1). Export of N by grazing cattle was not estimated but the N export by cutting was, on average, 75 kg ha-1( ± 11 kg N ha-1). In the literature, a lower nitrate content in the soil (De Vliegher et al., 2004) or a lower mineral N content in the soil (Holshof and Willems, 2003) have been reported. A pre-cut or not at the beginning of September? There were no significant differences between a single cut in October and a cut at the beginning of September followed by a cut in October, for NO3-N, NH4-N and the mineral-N content in every soil layer in October, as well as in November. At the end of the growing season, the average NO3-N content and the mineral-N content in the 0-90 cm layer was 36 kg ha-1 and 57 kg ha-1 for one cut in October and 50 kg ha-1 and 71 kg ha-1 for a cut at the beginning of September followed by a cut in October. These differences were not significant, however. Cutting in September and October resulted, on average, in an N export of 99 kg ha-1 (± 48 kg ha-1) and by cutting only in October, the average N export was 74 kg ha-1( ± 24 kg ha-1). Cutting twice resulted in an extra N export of 24 kg N ha-1 (± 30 kg ha-1). Conclusion In this experiment, cutting instead of grazing in the period September – November resulted in a significant decrease in NH4-N content in the 0-30 cm soil layer in October and November and a significant decrease in total mineral-N content in the 0-30 cm soil layer in November. NO3N, which is sensitive to leaching, was not significantly affected in this period, which is in contrast with findings reported in the literature. Cutting twice had no significant effects on NO3N, NH4-N and mineral N content in the soil in comparison with a single cut. Acknowledgements The research received funding from the European Community's Seventh Framework Programme (FP7/ 2007-2013) under the grant agreement n° FP7-244983 (Multisward). References De Vliegher A., Grunert O. and Carlier L. (2003) Cutting or grazing in autumn: effect on grass yield, grass quality and soil nitrate content. Grassland Science in Europe 8, 157-159. Holshof G. and Willems J. (2004) The influence of earlier indoor confinement of livestock and the reduction of nitrogen fertilization on the amount of mineral N in the soil and nitrate concentration in the uppermost groundwater. WUR Animal Science Group, Praktijk Rapport Rundvee (in Dutch) 44, pp 51. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 775 Impact of plant diversity, with equal number of grass and legume species, on sward productivity and legume content under contrasted mowing management in a low input system Jamar D., Clement C., Seutin Y., Planchon V., Campion M. and Stilmant D. Walloon Agricultural Research Centre (CRA-W) - Agriculture and Natural Environment Department - Farming Systems, Territory and Information Technologies Unit, 100, B-6800 Libramont, Belgium. Corresponding author: d.jamar@cra.wallonie.be Abstract Studies have underlined the link between plant biodiversity and sward productivity. Higher production is associated with complementarities between the different functional groups involved. More especially, the occurrence of legumes seems to be a key point. In this context, this study aimed to validate the interest of improving species number in the sward on its productivity while controlling for the legume-occurrence effect, in a trial performed under organic farming and mowing schemes (3 and 4 cuts per year). To test the hypothesis, mixtures based on perennial ryegrass-white clover, cocksfoot-lucerne and Lolium hybridum–red clover, were diversified through the addition of grass-legume pairs until mixtures with six species of grass and six species of legumes were obtained. The performances of these mixtures (production in quantity and quality and plant composition) were followed during three full exploitation years. As reported previously by Sanderson (2010), our results underlined that there was no unique relationship between herbage yield and the complexity (number of species) of the mixture: everything is a function of the potential of the initial pair of species in the given soil-climate context and of its persistency. Legume content was also crucial, but did not, by itself, explain the performances recorded. Keywords: functional diversity, mixture persistence, ryegrass-white clover, cocksfoot-lucerne, Lolium hybridum-red clover Introduction Faced with increased input costs, farmers want to reduce fertilizers and use of external feedstuffs while optimizing internal resources. This evolution in management of farming systems is in response to greater market instability and also to social demand. In such a context and for herbivores production systems, grassland agro-ecosystem productivity, in terms of quantity, quality and stability are key points to insure technico-economical performances and resilience of the farming system. Previous studies have shown the link between plant biodiversity and sward productivity, in terms of quantity (e.g. Hector and Loreau, 2001) and stability (Tilman et al., 2006). Higher production is associated with complementarities between different functional groups involved, complementarities improving resources valorization. The occurrence of legumes seems to be the key point to improve and stabilized sward productivity (Sanderson, 2010). In this context, the aim of this study was to validate the interest of improving species number in the sward on its productivity while controlling for the legume-occurrence effect. This trial was performed under organic farming and mowing schemes. Materials and methods In order to test the impact on grassland production and plant diversity evolution of plant species diversity in mixed grass-legume (60%: 40% of viable seeds) and of management schemes, the following experimental scheme was set up, in a four complete blocks design, in a organic field, converted in 1998, in Libramont, Belgium (49° 55' N 5° 35' E). Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 776 Mixtures, based either on perennial ryegrass (Lolium perenne L.) - white clover (Trifolium repens L.), on cocksfoot (Dactylis glomerata L.) – lucerne (Medicago sativa L.) and on Lolium hybridum – red clover (Trifolium pratense L.), were increased in complexity through the addition of grass-legume pairs until mixtures with six species of grass and six species of legumes were obtained (Table 1). Two modalities with, respectively, only grass (12) or legume (10) species were also included in the experimental scheme. Table 1. Mixtures compared under contrasted cutting management schemes. Timothy (Phleum pratense L.); tall fescue (Festuca arundinacea Schreb.); meadow fescue (Festuca pratensis Huds.); smooth meadow-grass (Poa pratensis L.); red fescue (Festuca rubra L.); brome (Bromus sitchensis Trin); Egyptian clover (Trifolium alexandrinum L.); sainfoin (Onobrychis sativa Lamarck); yellow trefoil (Medicago lupulina L.); subterranean clover (Trifolium subterraneum L.); birdsfoot trefoil (Lotus corniculatus L.); alsike clover (Trifolium hybridum L.). Mixture basis Four species mixtures Eight species mixtures Perennial ryegrass Cocksfoot Lolium hybridum /White clover (PW2) /Lucerne (CL2) /Red clover (HR2) + Cocksfoot + Tall fescue + Perennial ryegrass /Red Clover (PW4) /Red clover (CL4) /White clover (HR4) + Tall fescue + Perennial ryegrass + Italian ryegrass & Timothy grass & Timothy grass & Tall fescue /Lucerne /White clover /Lucerne & & Yellow trefoil (PW8) & Yellow trefoil (CL8) Egyptian Clover (HR8) Twelve species mixtures +Smooth meadow-grass +Brome & Red fescue +Timothy grass & Meadow fescue /Bird’s-foot trefoil & Brome / Bird’s-foot trefoil & Subterranean clover /Yellow trefoil & Alsike (PW12) (CL12) & Sainfoin (HR12) The contrasted management schemes applied during three seasons consisted of two cutting schemes reflecting intensive (4 cuts per year – 4C) or extensive (3 cuts per year – 3C) practices (fertilization with 35 t ha-1 of composted beef cattle manure). Before each cut, proportions of grass and legumes species were quantified, on a weight basis, after hand sorting of a composite sample of the four replicates of the corresponding modality. Before the second cut, the proportions of each species in the mixture were also defined. Due to the difficulties of distinguishing between Italian and Hybrid ryegrasses, these different species were regrouped in an HI group. Individual plots, measuring 1.5 × 11 m, were harvested using a Haldrup plot harvester. Green forage was weighed and a sample was collected from each plot and dried to quantify dry matter yield and to determine its feeding value (cellulase digestibility). Data analysis (ANOVA including species richness and block parameters and multiple means comparison based on Student Newman Keuls method) were done using SAS 9.2. software. Results and discussion As expected, both the yield (10557±454 vs 9490±505 kg DM ha-1) and quality (66.7±1.4 vs 70.2±1.6 %), for 3C and 4C respectively, were affected by the number of cuts. Statistical analyses, performed independently for both management schemes, highlighted a significant to highly significant effect of the mixture basis species number × year interaction on DM yields (F3C(12,105) = 12.3; P < 0.001 and F4C(12,105) – 2.2; P = 0.01). Table 2 integrates the results of the analyses performed per mixture basis and year. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 777 These results show that once the binary mixture is able to exploit the production potential of this specific soil-climate-management situation (c. 12 t DM ha-1) there was no further increase due to additional species: this was the case for HR-based mixtures in 2010 and CL-based mixtures across the three years (Table 2). In contrast, mixture complexity improved the productivity of PW-based mixtures: less aggressive in the sowing year than the HR and CL binary mixtures, across the three years and from less long-lasting HR-based mixtures, already the second and third year of production (Table 2). These trends were equivalent for both cutting frequencies. Sward quality was affected more by mixture basis and cropping year than by species richness (Table 2). Table 2. Impact of species richness on DM yield (t ha -1) and sward digestibility (%), per year and mixture basis. Means sharing a letter are not significantly different (P > 0.05). PW-3C PW-4C CL-3C CL-4C Year Nbr Sp DM DIG DM DIG DM DIG DM DIG 2010 2 8.3b 78a 8.8b 81a 11.5a 66c 9.8a 2010 2010 4 8 11.7a 73bc 12.0a 71cd 10.8a 76b 10.6a 75b 11.8a 69cb 11.6a 71.2b 2010 12 12.4a 73bc 10.7a 77b 2010 12G 5.2c 76ab 5.5c 2010 10L 11.4a 69d 9.9ab 2011 2 7.5c 7.8c 2011 2011 4 8 9.1b 66b 9.8b 66bc 2011 12 10.5a 65bc 9.8a 2011 12G 5.5d 67b 5.1d 69ab 5.5d 67a 5.1d 2011 10L 9.3b 64c 8.8b 66b 9.3c 64b 2012 2 7.3d 68a 6.8b 72a 2012 2012 4 8 9.2b 62b 10.2a 63b 8.5a 8.5a 2012 12 10.4a 63b 2012 12G 8.1c 2012 10L HR-3C DIG DM DIG 73 13.1a 72b 10.8a 77 10.7a 10.5a 73 75 11.7b 73b 11.7b 69c 10.5a 10.1a 78 75 12.1a 70.3b 10.3a 76 12.1b 70c 10.0a 75 77b 5.2b 76a 5.5b 78 5.2c 76a 5.5b 77 74b 11.4a 69cb 9.9a 74 11.4b 69c 9.9a 74 71a 11.5a 64b 11.0a 65 9.3b 9.5b 69ab 8.8b 69ab 9.6a 67ab 10.3b 64b 10.9ab 65b 9.8b 10.6a 68 68 9.8b 68a 11.2a 64b 9.4b 69ab 10.4a 67ab 10.6ab 65b 10.4ab 68 11.1a 66ab 10.4a 66b 69 5.5c 67a 5.1c 69ab 8.8c 66 9.3b 64b 8.8b 66b 11.3a 59b 8.4ab 64 8.6c 66a 8.6a 69 66b 66b 9.9b 62ab 9.6b 61ab 8.0b 8.5ab 65 66 8.5c 66a 11.0ab 62b 7.8b 9.0a 69 66 8.6a 66b 10.4ab 61ab 8.6a 66 11.6a 62b 9.0a 66 63b 6.6b 67b 63a 6.6c 67 8.1c 63b 6.6c 67 10.4a 62b 8.7a 67b 10.4ab 62ab 8.7a 66 10.4b 62b 8.7a 66 72a 68ab 8.1c DM HR-4C 67a Even if legume content was significantly correlated to sward productivity (r = 0.6; N = 84), across the different mixtures and years, with an increase of 48 kg DM ha-1 year-1 per percentage increase in legume proportion, the classifications presented in the Table 2 were not modified by performing covariance analyses with the legume occurrence as co-variable (data not shown) confirming an interest improving species diversity in sward-mixture composition. This applies even without taking into account the legume effect, whether the mixture basis is not able to valorise fully the local productivity potential and/or whether it is not persistent enough. Nevertheless, including more than 8 species in the mixture appeared to be inefficient in the context of this trial. Conclusions As reported by Sanderson (2010), there was not a unique relationship between herbage yield and the complexity (number of species) of the mixture: everything is a function of the potential of the initial pair of species in the given soil-climate context and of its persistency. Legume content was also crucial but it did not, by itself, explain the performances recorded. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 778 Acknowledgements The European Fund for Regional Development and the Walloon area funded this research in the context of VETABIO (INTERREG IV-A program) and GESPERBIO projects. References Loreau M. and Hector A. (2001) Partitioning selection and complementarity in biodiversity experiments. Nature 412, 72-76. Sanderson M.A. (2010) Stability of production and plant species diversity in managed grasslands: a retrospective study. Basic and Applied Ecology 11, 216-224. Tilman D., Reich P.B. and Knops J. (2006) Biodiversity and ecosystem stability in a decade-long grassland experiment. Nature 441, 629-632. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 779 The effect of different fodder galega-grass mixtures and nitrogen fertilization on forage yield and chemical composition Meripõld H., Lättemäe P., Tamm U. and Tamm S. Estonian Crop Research Institute, EE-48309 Jõgeva, Estonia Corresponding author: heli.meripold@etki.ee Abstract Fodder galega (Galega orientalis Lam.) is a forage legume that has been grown in Estonia for approximately forty years. Pure galega is known to be a persistent and high-yielding crop rich in nutrients, in particular crude protein (CP). Galega is usually grown in a mixture with grass in order to optimize its nutrient concentration, increase dry matter (DM) yield and improve fermentation properties. There are certain grass species suitable for the mixture. In this study galega mixtures with timothy (cv.Tika), meadow fescue (cv. Arni) and bromegrass (cv. Lincoln) were under investigation in four successive years (2008-2012). Three cuts were carried out during vegetation in 2008, 2009, 2011, 2012 and two cuts in 2010. Nitrogen (N) fertilization rates were N0, N50, N100; this was applied in spring before the first and second cuts. Early season N applications to galega-grass swards can help N-deficiency in the spring. The total dry matter (DM) yield varied from 7.2 to 13.6 t ha. DM yield was dependent on the year, mixture and fertilization level. The CP concentration in the DM varied from 149−229 g kg-1. CP was dependent on the year, mixture and fertilization. High N-fertilization favoured grass growth and reduced the role of galega in the sward. The dry and warm summers favoured the galega growth in the years 2010-2011, but 2012 was rainy. Keywords: Fodder galega, goat’s rue, galega-grass mixtures, forage yield, fertilization. Introduction Along with other forage legumes, like lucerne and clovers, goat’s rue (fodder galega) has been grown in Estonia for almost forty years. Galega is very persistent with a high yielding ability. Results have shown that the yields can possibly be 8.5 to 10.5 tons of dry matter and 1.7 to 1.8 tons of crude protein (CP) per hectare, with CP concentration of 200-220 g kg-1 DM (Raig et al., 2001). The nutritive value is the highest when the 1st cut is made at shooting, budding or at the beginning of flowering (Raig et al., 2001). In order to connect the need for nitrogen fertilizer with biologically fixed nitrogen, it is optimal to grow galega in a mixture with grass. Of plant nutrients, nitrogen has the highest effect on yield and quality of forage crop. When choosing grasses for mixtures, the speed of species development, duration and the effect on nutritive value should be considered. Earlier results have shown that growing galega in mixtures with grasses improves the nutritive value and ensiling properties of the forage crop (Lättemäe et al., 2005; 2013). The aim of this investigation was to study the different galega-grass mixtures and N fertilization on DM yield and chemical composition of forage. Materials and methods The experimental field was established in 2003 in Saku Estonia (local latitude 57º 25/) and the data were collected from 2004. The data from four successive years (2008-2012) were recorded in this study. The trial plots were established on a typical soddy-calcareous soil where the agrochemical indicators were as follows: pHKCl 7.4 (ISO 10390); humus concentration Corg 4.1%; concentration of lactate soluble P and K were 97 and 166 mg kg-1 respectively. Three galega-grass mixtures were used. The galega variety Gale (Go) was sown in binary mixtures with meadow fescue cv. Arni (Fp) (10 kg seed ha-1), timothy cv. Tika (Pp) (6 kg ha-1) and bromegrass cv. Lincoln (Bi) (15 kg ha-1) respectively. The sowing rate of the seed of cv. Gale was 20 kg ha-1 in all mixtures. In order to increase competitiveness of grasses and yield of the Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 780 first cut, three N fertilization levels were used: N0, N50 and N 100 kg ha-1 (April, May I or II decade). The crop was cut by scythe, then weighed and samples were taken for analyses. The botanical composition of the crop was determined prior to sampling. A three-cut system was used during harvest and there were three replicates of the plots of each treatment. All statistical analyses were carried out by using the GLM procedure of SAS. Results and discussion The results indicate that galega-grass mixtures ensured high DM yield from since the trial field was established. In 2008-2012 the yields varied from 7.2 to 13.6 t ha-1 (Table 1). There were significant differences between the average yields at different N levels and mixtures. Table 1. The DM yield of fodder galega-grass mixtures in 2008-2012 Mixture 2008 2009 2010 2011 2012 N0 N50 N100 N0 N50 N100 N0 N50 N100 N0 N50 N100 N0 N50 N100 Go/Fp 11.6 10.7 11.6 7.6 7.2 7.8 10.4 10 9.4 12.7 13.3 12.4 11.9 11.5 10.3 Go/Pp 10.8 11.3 10.8 8.2 11.4 11.4 11.8 10.1 10.9 11.7 11.5 13 11.6 13.5 13.6 Go/Bi 11.3 12.4 10.9 8.1 10.1 11.4 9.8 9.7 10.7 12 11.4 11.7 10.9 11.0 11.9 Average 11.2 11.5 11.1 8 9.6 10.2 10.7 9.9 10.3 12.1 12 12.4 11.5 12.0 11.9 *LSD 0.05=0.39 *LSD 0.05 = 0.64 *LSD 0.05 = 0.45 *LSD 0.05 = 0.60 *LSD 0.05 =0.40 **LSD0.05=0.22 **LSD 0.05 =0.37 **LSD 0.05 =0.32 **LSD 0.05 =0.41 **LSD0.05=0.29 *- Least significant difference of N treatment; **- LSD0.05 of mixture treatment The average yield was higher in 2011 and variable, varying from 11.4 to 13.3 t ha-1. There were significant differences between Gale-Arni vs. Gale-Tika and Gale-Lincoln as well as Gale-Tika and Gale-Lincoln at N0 and N100 fertilization levels. Application N fertilizer changed the botanical composition of the sward. N fertilizer increased grasses and reduced the cv. Gale proportion in the pasture. The average cv. Gale proportion in the pasture was 59% in 2008. In the year 2009 the proportion of cv. Gale declined down to 27% due to frost damage during the second decade of May (temperature declined down to -3.9 degrees). When the two-cuts system was applied the Gale pasture recovered in 2010. The Gale proportion also declined considerably when fertilization increased (Figure 1.). The average Gale proportion is essential at N0 and N50 treatments. When it is higher the CP concentration increases in the crop (r=0.53; P<0.05). At fertilization level N0 and N50 the meadow fescue cv. Arni was less competitive. The highest competitiveness was shown by the bromegrass Lincoln at N100 fertilization in 2009. The nutritive value of mixtures is presented in Table 2. In general, the nutritive value of mixtures was mainly dependent on fertilization level. Lower CP and metabolizable energy (ME) concentrations were found in treatments when N fertilizer was not used. When the fertilization level increased, CP concentration and ME increased but NDF and ADF decreased. Galega usually has a faster rate of development than grasses. Therefore, the ADF and NDF concentrations were higher in treatments where the proportion of cv. Gale was higher and in Gale/Lincoln treatment, due to the higher fibre concentration of the bromegrass. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 781 90 'Gale' % in the mixture 80 70 60 2008 50 2009 40 2010 30 2011 20 2012 10 0 N0 N50 N100 N0 N50 Gale'/'Arni' N100 Gale'/'Tika' N0 N50 N100 Gale'/'Lincoln' Treatm ent Figure 1. The botanical composition of galega-grass mixture in the first cut in 2008-2012 Table 2. The nutritive value of the fodder galega-grass mixtures of first cut in 2008-2012 Mixture N CP NDF ADF ME fertilizer g kg-1 DM g kg-1 DM g kg-1 DM MJ kg-1 DM Gale/Arni N0 183 466 315 10.1 Gale/Arni N50 187 459 312 10.2 Gale/Arni N100 206 413 286 10.5 Gale/Tika N0 177 486 328 9.9 Gale/Tika N50 203 450 309 10.2 Gale/Tika N100 211 435 291 10.5 Gale/Lincoln N0 165 462 335 9.8 Gale/Lincoln N50 187 449 327 9.9 Gale/Lincoln N100 198 488 310 10.2 Conclusions The galega-grass mixtures maintained high yielding ability and nutritive value for many years. The nutritive value of mixtures was mainly dependent on fertilization. High N fertilization rate favoured grass growth but reduced the role of galega in the sward. The similar higher ME value was obtained in Gale-Arni and Gale-Tika mixtures. The ME concentration was lower in GaleLincoln mixture due to higher fibre concentration of bromegrass comparing to other grasses. On the basis of these results, fertilization rate of N50 should be recommended in order to avoid grasses being lost from the pasture and can help N deficiency in the spring. References Lättemäe P., Meripõld H., Tamm U. and Tamm S. (2013) The effect of different fodder galega-grass mixtures and nitrogen fertilization on forage yield and chemical composition. Grassland Science in Europe Vol. 18, 168-170. Lättemäe P., Meripõld H., Lääts A. and Kaldmäe H. (2005) The improvement of fodder galega silage quality by using galega-grass mixtures and additive. Grassland Science in Europe Vol. 10, 635-638. Raig H., Nõmmsalu H., Meripõld H. and Metlitskaja J. (2001) Fodder Galega monographia, ERIA, Saku, 141 pp. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 782 Grass only and grass-white clover (Trifolium repens L.) swards: herbage production and white clover performance Egan M.1,2, Enriquez-Hidalgo D.1,3, Gilliland T.3,4, Lynch M.B.2 and Hennessy D.1 1 Teagasc, Animal & Grassland Research and Innovation Centre, Moorepark, Fermoy, Co. Cork, Ireland 2 UCD School of Agriculture and Food Science, University College Dublin, Belfield, Dublin 4 3 Institute for Global Food Security, School of Biological Sciences, Queen's University, Belfast, Northern Ireland 4 Agri-Food and Biosciences Institute, Plant Testing Station, Crossnacreevy, Belfast, Northern Ireland. Corresponding author: Michael.Egan@teagasc.ie Abstract White clover (Trifolium repens L.; Wc) is the most important forage legume in temperate regions of the world. This experiment compared herbage production and tiller density in grass only (GO) and grass-white clover (GWc) swards in 2011 (year (yr) 1) and 2012 (yr 2). Sward Wc content and stolon mass in the GWc swards were quantified. Both swards received 260 kg N/ha/yr. There was no sward type effect on total herbage production in 2011 (14040 kg DM/ha). In 2012, the GWc swards produced more herbage than the GO swards, but this was not statistically different (14740 and 13580 kg DM/ha, respectively). White clover content was not statistically (P=0.12) different between yrs (2011 - 12.6%; 2012 - 21.8%). The perennial ryegrass (Lolium perenne L.) tiller density was greater in 2011 than in 2012 (P<0.001) and was greater for the GO than for the GWc swards (5952 and 4934, tillers/m2 P>0.001). Stolon mass was less in 2011 than 2012 (18.6 and 37.1 g DM/m2, respectively P>0.01), suggesting that WC becomes more established in the sward in the second production year. Including WC in grass swards can increase herbage mass in the second production year and tiller density declines as stolon mass increases. Keywords: Trifolium repens L., Lolium perenne L., white clover content, herbage mass, tiller density Introduction White clover (Trifolium repens L.; Wc) is the most important forage legume in temperate regions of the world (Frame and Newbould, 1986). When sown in mixed perennial ryegrass (Lolium perenne L.; PRG) Wc swards, Wc content is thought to stabilize at 20% of total DM production two years post-sowing when appropriately managed (Andrews et al., 2007). Sward Wc dominance is related to PRG tiller density; Wc content declines as tiller density increases beyond 5000 tiller/m2 (Brereton, 1995). The objective of this study was to investigate the effect of WC inclusion in a PRG sward on herbage production and tiller density during the first and second production years. Materials and methods In May 2010 swards of PRG-only (GO) and PRG-Wc (GWc) were sown at the Dairygold Research Farm, Teagasc, Moorepark, Fermoy, Co. Cork, Ireland. The GO swards were a 50:50 mixture of Astonenergy (tetraploid) and Tyrella (diploid) PRG cultivars sown at 37 kg/ha, and the GWc swards contained the same PRG mix and a 50:50 mixture of Chieftain and Crusader clover cultivars sown at 5 kg/ha. Swards were grazed by rotational stocking with lactating dairy cows from March to October in 2011 and 2012. There were 8 rotations (rot) in 2011 and 11 in 2012. Swards received 260 kg N/ha applied between February and mid-September. Swards were grazed once in late February of each year, before the start of the experiment. Pre-grazing Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 783 herbage mass (PGHM) (>4 cm) on a DM basis was determined twice weekly for each sward type (ST) by cutting three strips 1.2 m wide and a known length in the area due to be grazed next and dried at 90oC for 15hrs to determine DM. Sward Wc content was determined twice weekly; random herbage samples >4 cm were taken across the paddock and separated into grass and Wc fractions and dried at 40oC for 48 hours to determine the DM proportions. Forty-five turves (10 cm × 10 cm) were removed at random from each ST four times in 2011 and three times in 2012 to estimate PRG tiller density and Wc stolon mass (tiller/m2 and DM/m2, respectively). The PRG and other grass tillers (mainly, Poa annua L.) were separated and counted. The Wc stolons were removed from each turf, the roots and leaves removed, and then gently washed to remove excess soil; the stolons were then dried at 40 °C for 48 h and weighed to obtain kg DM/ha, as described by (Harris, 1994). Data were averaged as one value/paddock per rot for PGHM and sward Wc content and per collection or month (mo) for tiller density and stolon mass were analysed using PROC MIXED (SAS, 2005). The model included ST, rot or mo, yr, and the associated interactions that at least tended to be significant (P<0.1) as fixed effects; and the rot/mo within each yr was used as the repeated measure. Results and discussion The PGHM were greater in 2011 than 2012 (1700 and 1400 kg DM/ha SE=445, respectively; P<0.01), but were similar between the two treatments across the grazing season. There was no ST effect on total herbage production in 2011 (14040 kg DM/ha). In 2012, the GWc swards produced 1160 kg DM/ha more than the GO swards but were not statistically different (14740 and 13580 kg DM/ha, SE=701, respectively; P=0.44). The low sward-Wc content in 2011 is likely to have contributed to the lack of difference between the swards. Andrews et al. (2007) showed that HM is increased when sward-Wc content is greater than 20%, which could explain the slight increase in herbage production in 2012. The GWc sward-clover content was low in the first rot (8.8% and 8.4% in 2011 and 2012, respectively) and then increased as time progressed, similar to previous findings (Frame and Newbould, 1986; Brereton, 1995). Annual sward-Wc content was not statistically (P=0.12) different between the years (2011 - 12.6% and 2012 - 21.8%; SE=3.35). Although not statistically different, GWc swards had a higher annual clover content in 2012, suggesting that Wc was becoming more established in the sward during the second production year, which agrees with Frame and Newbould (1986) and Brock and Kane (2003). Following this trend, the clover DM yield tended (P=0.09) to be greater in 2012 than in 2011 (3170 and 1780 kg DM/ha, respectively, SE=445.8). The PRG tiller density was greater in 2011 than in 2012 (6110 and 4775 tillers/m2, respectively; P<0.01) and was greater for the GO than for the GWc swards (5952 and 4934 tillers/m2, respectively, P<0.01), which would agree with Brereton (1995) who reported that stolon mass would be decreased with a tiller density beyond 5000 tillers/m2. Harris (1994) also reported that as tiller density decreases, the stolon mass increases, similar to that seen in this experiment. Stolon mass was lower in 2011 than in 2012 (18.6 and 37.1 g DM/m2, SE=4.6, respectively, P<0.01). The lowest value for stolon mass content in the GWc swards was observed in February 2011 (4.9 g/m2) and stolon mass steady increased in 2011 and remained relatively constant in 2012. The maximum stolon mass value was obtained in October 2012 (47.6 g/m2; Figure 2). The greater stolon mass and the numerically greater sward-Wc content in 2012 (the second production yr) suggest that Wc stabilization in the mixed swards was on going in the first production yr, which would agree with Harris (1994) and Garwood (1969). Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 784 GWc 2011 15-Jun 15-Jul Month Figure 1. Sward white clover content for 2011 and 2012 PRG Tiller/m2 15-Mar 15-Apr GWc 2012 15-May Stolon 7500 GWc 15-Aug 15-Sep 15-Oct 60 GO 40 5000 20 2500 0 Nov-10 Feb-11 May-11 Aug-11 Nov-11 Feb-12 May-12 Aug-12 Stolon (g/m2) Clover content (DM %) 35 30 25 20 15 10 5 0 15-Feb 0 Nov-12 Figure 2. Grass tiller density and stolon mass in 2011 and 2012 in grass-only (GO) and grass-white clover (GWc) swards Conclusion When sward-Wc content was low in 2011 there was no difference in herbage production. As sward-Wc content increased in 2012, tiller density in the GWc swards decreased, herbage mass slightly increased and Wc DM yield increased. This shows that as Wc content becomes more established in the sward in the second year, herbage production and Wc DM yield can increase. Acknowledgements This work was supported by the Teagasc Walsh Fellowship Programme, the Dairy Levy Fund and the European Community's Seventh Framework Programme (FP7/2007-2013) under the grant agreement n° FP7-244983 (Multisward). References Andrews M., Scholefield D., Abberton M.T., McKenzie B.A., Hodge, S. and Raven J.A. (2007) Use of white clover as an alternative to nitrogen fertiliser for dairy pastures in Nitrate Vulnerable Zones in the UK: productivity, environmental impact and economic considerations. Annals of Applied Biology 151, 11-23. Brereton A.J. (1995) Regional and year to year variation in production. In: Jeffery D.W., Jones M.B. and McAdam J.H. (Eds) 1995. Irish grasslands - their biology and management. Dublin: Royal Irish Academy, pp. 12-22. Brock J.L. and Kane G.J. (2003) Variability in establishing white clover in pastures on farms. Proceedings of the New Zealand Grassland Association 65, 223-228. Frame J. and Newbould P. (1986) Agronomy of white clover. Advances in Agronomy 40, 1-88. Garwood E.A. (1969) Seasonal tiller populations of grass and grass/clover swards with and without irrigation. Journal of the British Grassland Society 24, 333-343. Harris S.L. (1994) White clover growth and morphology in dairy pasture in the Waikato region of northern New Zealand. New Zealand Journal of Agricultural Research 37, 487-494. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 785 The persistence of perennial ryegrass cultivars (Lolium perenne L.) in binary mixtures with white clover (Trifolium repens L.) under grazing Gregis B. and Reidy B. School of Agricultural, Forest and Food Sciences HAFL, Bern University of Applied Sciences, Switzerland Corresponding author: beat.reidy@bfh.ch Abstract The persistence of highly productive forage species in pastures is essential to maximize economic returns from grazing livestock. However, most forage cultivars are neither selected nor evaluated under grazing. To test the persistence of ryegrass (Lolium perenne L.) cultivars under grazing, a five-year plot trial was conducted on commercial dairy farms located in different climatic conditions in Switzerland. Plots were arranged in a randomized, complete block design with three replicates and sown in autumn 2007 and spring 2008 in binary mixtures with white clover (Trifolium repens L). The relative frequency of the perennial ryegrass cultivars was evaluated in 2008, 2009 and 2012 to determine their long-term tolerance to grazing. Significant interactions between both site and cultivar and site and years were found, revealing the importance of the site and its management for the performance of the cultivars. The persistence, as determined by the relative frequencies of the ryegrass cultivars tested, corresponded in the short-term (second and third years) to the rankings of the official variety trials but differed for the long-term observation (in the fifth year). This suggests the need for additional long-term observations under grazing conditions as an extension of the official variety recommendations for cost-efficient, pasture-based livestock production. Keywords: Lolium perenne, grazing tolerance, persistence, cultivar Introduction Its tolerance to grazing, its reproductive performance and its excellent nutritional value make perennial ryegrass (Lolium perenne L.) the principle grass plant for pasture-based dairy systems in temperate climates (Mosimann, 2002). Persistence is gaining an important role in minimizing costs through reducing the expenses for reseeding while maintaining good productivity of the grass sward. However, most forage cultivars are neither developed nor evaluated under grazing (Brummer and Moore, 2000). This is also the case for the official Swiss variety trials for perennial ryegrass conducted by Agroscope, the Swiss federal research institute for the agrofood industries, in which cultivars are evaluated under a cutting regime only. To simulate pasture conditions, plots for plant-density investigations were cut more frequently (Suter et al., 2012). Specific factors, which may affect the performance of a cultivar under grazing such as trampling, fouling, and tiller pull-up, are not present or may not be as prevalent, as in a clipping situation (Hopkins, 2005). Nevertheless, persistence (evaluated after three years) and the ability to build dense plant populations (stand-density) as an indirect indicator of grazing tolerance are evaluated. To assess tolerance to grazing, we tested seven official recommended cultivars under grazing conditions over a period of five years to compare the correspondence between the official substituted pasture parameters (persistence and stand-density) and the effective performance under grazing pressure. Materials and methods Seven officially tested and recommended perennial ryegrass cultivars (Arara, Salamandra, Elgon, Soraya, Alligator, Arvicola, Artesia) (Suter et al., 2012) were sown in field plots (3m × 7m) on five different dairy farms in Switzerland in autumn 2007 and spring 2008 in binary mixtures with white clover (Trifolium repens L.). The experimental sites covered a wide range Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 786 of climatic conditions and farm management practices (altitude (430 – 600 m a.s.l., precipitation (650 – 1100 mm/year), pasture type/pressure (continuous, rotational stocking), fertilizer use (80-190 kg N, 40-90 kg P, 100-150 kg K, 0-20 kg Mg/ha per year in the form of slurry and/or different mineral fertilizers) and sowing season (autumn 2007, spring 2008). The perennial ryegrass cultivars were sown at seed rates of 15 kg/ha (Arara) and 20 kg/ha (others) in binary mixtures with two white clover cultivars, differing in leaf size, and sown at a seed rate of 25 kg/ha and 15 kg/ha. At each site, the plots were repeated three times in a randomized complete block design. The relative frequency of the cultivars was measured in every plot in the autumn of 2008, 2009 and 2012 according to Daget and Poissonet (1969) by observing the plants at 50 points per plot at an equidistance of 10 cm. To analyse the effects of the cultivars, sites and time, a Brunner-Langer F2-LD-F1 model for longitudinal data was fitted using the nparLD package in R (R Core Team, 2013). Results and discussion Generally, relative frequencies of all cultivars within the experiment strongly declined in 2012 compared to 2008 and 2009 (Table 1). The development of the relative frequencies of the cultivars tested over the five years showed a statistically significant site × year (P < 0.001) and site × cultivar (P < 0.05) interaction. Both interactions indicate that the individual cultivars responded differently to the site-specific climatic conditions and/or the management practices, which underlines the importance of testing cultivars under a broad range of climatic and management conditions. The mean relative frequency per cultivar over the five years was highest for Arara and Arvicola (Table 1). This corresponds to the official variety results for stand-density, which also showed higher values for both cultivars (Suter et al., 2012). The official stand-density ranking for Arara as the only diploid cultivar differed connotatively from the ranking of the other cultivars. This may be related to the fact that diploid cultivars generally build more tillers than tetraploids (Laidlaw, 2004). Salamandra and Alligator were the cultivars with low stand densities in the official testing programme. For both cultivars, the lowest relative frequency after 5 years was measured in our experiment. This indicates that the more frequent cutting in the official tests was a reasonable simulation of the short-term grazing effect on given cultivar. Arara and Arvicola had the highest relative frequency after five years (Table 1). Table 1. Relative frequencies (as %) of seven perennial ryegrass cultivars over a period of five years at five different sites under grazing conditions. Cultivar Site Champvent Gampelen Hessigkofen Hohenrain Waldhof 08 09 12 08 09 12 08 09 12 08 09 12 08 09 12 Mean (s.d.) 08 09 12 Alligator 42 75 34 54 48 27 38 44 26 64 46 29 63 67 31 52 (12) 56 (14) 29 (3) Arara 42 81 34 78 59 63 39 49 32 73 61 28 65 81 57 59 (18) 66 (14) 43 (16) Soraya 57 74 35 61 53 48 42 45 29 63 50 43 68 60 39 58 (10) 56 (11) 39 (7) Artesia 43 73 33 64 57 34 36 44 32 69 58 31 63 69 49 55 (15) 60 (12) 36 (8) Arvicola 53 79 35 72 60 47 42 50 36 65 50 48 63 71 45 59 (12) 62 (13) 42 (6) Elgon 51.5 76 33 56 51 28 34 44 34 58 51 36.5 63 62 35 53 (11) 57 (12) 33 (3) Salamandra 49 58 30 51 53 27 36 36 29 58 50 34 64 69 42 52 (11) 53 (12) 32 (6) Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 787 Together with Soraya, these three cultivars showed the least reduction with respect to the relative frequency as compared to 2008 and 2009. We did not find any better persistence for Salamandra and Artesia, in contrast to the official persistence rankings. Alligator was the only cultivar for which we found the expected (according to the ranking of the official variety testing) lower persistence. The reason for this discrepancy may be explained by differences in the parameters investigated, the different impact of grazing versus cutting conditions or the varying observation period. Nevertheless, it emphasizes the need for long-term observations under grazing conditions to evaluate the persistence of cultivars under grazing conditions. Conclusion The results from this study suggest that not only the site, but also the climatic and management conditions are important determinants for the performance of perennial ryegrass cultivars. This implies that cultivars should be tested under a wide range of environmental conditions.Furthermore, although simulating pasture conditions through more frequent cutting, as in the official tests, can result in a fair estimation of the short-term grazing resistance of a cultivar, it is not sufficient for the measurement of long-term grazing persistence, as this study has shown. While evaluating cultivars for cost-efficient, pasture-based livestock production, additional long-term observations under grazing conditions could be a worthwhile extension to the official variety recommendations. References Brummer E.C., Moore K.J. (2000) Persistence of perennial cool-season grass and legume cultivars under continous grazing by beef cattle. Agronomy Journal 92, 466–471. Brunner E., Domhof S. and Langer F. (2002) Nonparametric analysis of longitudinal data in factorial experiments. Wiley, New York, 257 pp. Daget P. and Poissonnet J. (1969) Analyses phytologiques des prairies, applications agronomiques. Document No 48 CNRS-CEPE, Montpellier, 67 p. Hopkins A.A. (2005) Grazing tolerance of cool-season grasses planted as seeded sward plots and spaced plants. Crop Science 45, 1559 – 1564. Laidlaw A.S. (2004) Effect of heading date of perennial ryegrass cultivars on tillering and tiller development in spring and summer. Grass and Forage Science 59, 240–249. Mosimann E. (2002) Ray-grass anglais et trèfle blanc: quelles variétés pour la pâture continue? Revue suisse Agriculture 34 (5), 225–229. R Core Team (2013) R: a language and environment for statistical computing, R Foundation for Statistical Computing. Vienna, Austria. Suter D., Hirschi H., Frick R. and Aebi P. (2012) Englisches Raigras: 62 Sorten mussten sich bewähren. Agrarforschung Schweiz 3 (9), 414–421. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 788 Grass-only and grass-white clover (Trifolium repens L.) swards: dairy cow production Enriquez-Hidalgo D.1,2, Egan M.1,3, Gilliland T.2,4, Lynch M.B.3 and Hennessy D.1 1 Teagasc, Animal & Grassland Research and Innovation Centre, Moorepark, Fermoy, Co. Cork, Ireland; 2 Institute for Global Food Security, School of Biological Sciences, Queen's University, Belfast, Northern Ireland; 3 School of Agriculture and Food Science, University College Dublin, Dublin 4, Ireland; 4 Agri-Food and Biosciences Institute, Plant Testing Station, Crossnacreevy, Belfast, Northern Ireland Corresponding author: Daniel.Enriquez@teagasc.ie Abstract White clover (Trifolium repens L.; clover) can increase sustainability of grass-based dairy systems and has the potential to increase milk production. This experiment examined the seasonal effect of grass-white clover (GWc) and grass-only (GO) swards during the first (2011) and second (2012) production year on dairy cow production. Two groups of cows were allocated to graze each sward in 2011 (n=15) and 2012 (n=20). Swards were rotationally grazed. Clover content in GWc swards tended to be less in 2011 than 2012 (18.0 and 25.3 dry matter (DM)%, respectively, P=0.08). Cows grazing both swards had similar total milk and milk solid yields (2011: 3494 and 271; 2012: 4242 and 341 kg/yr, respectively). However, in 2012, GWc cows had slightly higher milk production than the GO cows from June onwards, a response to the greater sward clover content in those months. Sward type had little effect on milk fat, protein and lactose content, 4.52, 3.65 and 4.57%, respectively. It is concluded that the potential of clover to increase milk production depends on sward clover content. Keywords: Trifolium repens L., Lolium perenne L., mixed sward, dairy cow, milk production Introduction White clover (Trifolium repens L.; clover) can increase the sustainability of grass-based dairy systems (Frame and Newbould, 1986). Clover has better nutritional quality than perennial ryegrass (Lolium perenne L.; grass), and grass-clover swards (GWc) can increase dairy cow voluntary dry matter (DM) intake (DMI) compared to grass-only swards (GO), and cows grazing GWc can produce more milk (Harris et al., 1997). However, clover growth is dependent on temperature and therefore clover content in GWc swards varies throughout the year (yr). The objective of this experiment was to assess the seasonal effect of clover inclusion in grass swards during the first (2011) and second (2012) production years on dairy cow production. Materials and methods Grass-only and GWc swards were sown at Dairygold Research Farm, Teagasc, Moorepark, Fermoy, Co. Cork, Ireland in May 2010 and both received 260 kg N/ha/yr. Thirty/40 springcalving dairy cows were allocated to graze each sward type (ST) from April/March to October of 2011 and 2012, respectively. Swards were rotationally stocked and fresh herbage was offered daily (16 and 17 kg DM/cow/d in 2011 and 2012, respectively, at 4 cm above ground level) after morning milking, and 1 kg concentrate/cow/d in 2011. Concentrate supplementation was only offered in 2012 when feed demand exceeded grass supply. A herbage sample from the GWc sward was separated into grass and clover components to estimate sward clover content on a DM basis twice-weekly. Daily milk yield (MY) was recorded at each milking (7:30 h and 15:30 h; Dairymaster, Causeway, Co. Kerry, Ireland). Milk composition (fat (MFatc), protein (MProtc) and lactose (MLacc)) was determined weekly. Milk solids yield (MSY) was Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 789 calculated as the sum of milk fat and protein yields. Cow data were averaged as one value per cow/month (mo) and as one clover content value per paddock/mo and analysed using PROC MIXED (SAS, 2005). Cow data model included ST, mo, yr, and their interactions as fixed effects; parity, calving date, with the three-week pre-experimental milk data as covariates and the mo as a repeated measure. Interactions and covariates were removed from the model if they did not tend to be significant (P>0.1). Results and discussion There was a mo×yr interaction for the GWc sward clover content (Figure 1). Clover content (DM%) 40 30 20 10 0 Mar Apr May Jun Jul Aug Sep Oct Figure 1. Seasonal clover content in the GWc swards during the first (2011: dashed lines) and second (2012: solid lines) sward production year. Clover content followed the expected seasonal pattern, with low values in spring and maximum in late summer/early autumn (Frame and Newbould, 1986). Clover content was greater in July (P<0.05) and tended to be greater (P<0.1) in April, May and September in 2012 than in 2011. Annual clover content tended to be greater in 2012 than in 2011 (25.3 and 18.0% of DM, P=0.08). There was a ST×mo×yr interaction for all milk parameters analysed, except for MLacc which only had a ST×mo interaction (Figure 2). MY, MSY and MLacc decreased, but MFatc and MProtc increased across both years. Milk yield and MSY were greater in 2011 than in 2012 from May until August, and were greater for October in 2012. Clover inclusion had no effect on seasonal or cumulative MY or MSY (3494 and 271 kg/yr, respectively) in 2011. In 2012 MY tended (P=0.06) to be greater for GWc cows in June compared to the GO cows and was 19% greater (P<0.05) in August than for GO cows. The GWc cows also had greater (P<0.05) MSY than the GO cows in June, August and October in 2012. However, there was no ST effect on cumulative values (MY 4242 and MSY 341 kg/yr) in 2012. The year differences (P<0.001) in cumulative MY and MSY are related with the longer evaluation period in 2012. Previous work did not find differences in milk production between cows grazing either ST (Phillips et al., 2000; Schils et al., 2000). Such a lack of effect is related to the low clover content observed in the GWc swards. It has been proposed that milk production in mixed swards can be increased when clover content is greater than 30% (Harris et al., 1997). This also explains the MSY differences observed in the second half of 2012. Additionally, under this system cows had a restricted herbage allocation which did not allow the potential benefit in voluntary DMI that clover might have on cow production to manifest (Harris et al., 1997). MFatc was greater (P<0.05) in 2012 than in 2011 in April, May and July and August. MProtc tended (P=0.06) to be greater in April 2012 and also was greater (P<0.05) in May and June 2012 compared to those months in 2011. The GWc cows tended (P=0.06) to have greater MFatc than GO cows in October 2011. Clover inclusion increased the MLacc in September by 2% (4.37 and 4.46%, P<0.05) in both years. There was no significant ST effect but year tended (P=0.1) to affect total MFatc (2011: 4.38% and 2012: 4.67%) and total MLactc (2011: 4.60% and 2012: 4.55%), and had no effect on total MProtc (3.65%). Similarly, other research (Leach et al., 2000; Phillips et al., 2000) did not find a ST effect on milk composition. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 790 a) 25 b) MSY (kg/d) MY (kg/d) 1.8 20 15 1.2 0.9 10 d) c) 5.4 4.2 3.9 MFatc (%) MProtc (%) 1.5 3.6 3.3 3.0 Mar May Jul Sep 4.8 4.2 3.6 3.0 Mar May Jul Sep Figure 2. Seasonal effect of ST (GO: squares, GWc: circle) during the first (2011: dashed lines) and second (2012: solid lines) sward production year on dairy cow a) milk yield (MY), b) milk fat % (MFatc), c) milk solids yield (MSY) and d) milk protein % (MProtc). Conclusion White clover content in mixed swards was low during spring and greatest in late summer or autumn. White clover inclusion into grass swards had only minimal effect on milk production and composition in the first production year, but increased milk and milk-solids production in the second part of the second grazing season. It is concluded that the potential of white clover to increase milk production depends on sward clover content. Acknowledgements The Teagasc Walsh Fellowship Programme, Dairy Levy Fund and the European Community's Seventh Framework Programme (FP7/2007-2013, n° FP7-244983, Multisward). References Frame J. and Newbould P. (1986) Agronomy of white clover. Advances in Agronomy 40, 1-88. Harris S.L., Clark D.A., Auldist M. J., Waugh C.D. and Laboyrie P.G. (1997) Optimum white clover content for dairy pastures. Proceedings New Zealand Grassland Association 59, 29-33. Leach K.A., Bax J.A., Roberts D.J. and Thomas C. (2000) The establishment and performance of a dairy system based on perennial ryegrass-white clover swards compared with a system based on nitrogen fertilized grass. Biological Agriculture and Horticulture 17, 207-227. Phillips C. J. C. James N. L. and Nyallu H. M. (2000) The effects of forage supplements on the ingestive behaviour and production of dairy cows grazing ryegrass only or mixed ryegrass and white clover pastures. Animal Science 70, 555-559. Schils R.L.M., Boxem T. J., Sikkema K. and Andre G. (2000) The performance of a white clover based dairy system in comparison with a grass/fertiliser-N system. I. Botanical composition and sward utilisation. Netherlands Journal of Agricultural Science 48, 291-303. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 791 Effect of grass-only compared to grass-white clover swards on cow rumen function and methane emissions Enriquez-Hidalgo D.1,2, Lewis E.1, Gilliland T.2,3 and Hennessy D.1 1 Teagasc, Animal & Grassland Research and Innovation Centre, Moorepark, Fermoy, Ireland; 2 Institute for Global Food Security, School of Biological Sciences, Queen's University, Belfast, UK; 3 Agri-Food and Biosciences Institute, Plant Testing Station, Crossnacreevy, UK Corresponding author: Daniel.Enriquez@teagasc.ie Abstract A study was undertaken to identify the effect of white clover inclusion (Trifolium repens L.) in perennial ryegrass swards on dairy cow rumen characteristics and methane emissions. Rumen volatile fatty acids (VFAs), ammonia content and pH of cows grazing grass-only (GO) or grasswhite clover (GWc) swards during spring, summer and autumn were assessed. Methane emissions from 20 cows grazing each sward were assessed in autumn. The clover content of the GWc swards was 7.5, 8.8 and 30.9% in spring, summer and autumn, respectively. Clover inclusion influenced rumen characteristics (altered VFA composition, increased ammonia content and rumen pH) with increased effect as the season progressed due to greater sward clover content. Clover effects on rumen characteristics were greater in the afternoon than in the morning due to a greater clover inclusion in the diet during morning grazing. Despite the high sward clover content in autumn, clover inclusion only increased milk fat content; it had no effect on milk protein content and little effect on milk production. Clover inclusion reduced CH4 emissions per unit of intake but did not affect CH4 emissions per day or per unit product. Keywords: Trifolium repens L., Lolium perenne L., rumen function, methane emission Introduction The ability of white clover (Trifolium repens L.; clover) to fix N, combined with having a greater nutritional quality than perennial ryegrass (Lolium perenne L.; ryegrass), can contribute to improved sustainability of grass-based dairy systems (Frame and Newbold, 1986). Clover content in grass-clover swards (GWc) varies during the grazing season due to temperature and natural growth pattern. Grazing cattle show a partial preference for clover compared to ryegrass (Rutter et al., 2004). It is likely that seasonal clover content and the preference of cattle for clover might interact and alter rumen function and enteric methane emissions (e-CH4). The objective of this experiment was to identify the seasonal effect of including or not including clover in ryegrass swards on dairy cow rumen function and e-CH4. Materials and methods A grass-only (GO) and GWc sward were sown at Dairygold Research Farm, Teagasc, Moorepark, Fermoy, Co. Cork, Ireland in May 2010. In 2011, 30 spring-calving dairy cows were allocated to each sward type (ST; n=15) from April to October. Swards were rotationally stocked and fresh herbage offered daily (17 kg DM/cow/d) after morning milking. Pre-grazing herbage mass (HM) was estimated twice-weekly using a lawn mower (Etesia UK. Ltd., Warwick, UK). The GWc sward-clover content was estimated twice-weekly. Eight rumenfistulated dairy cows were arranged into four 2 (treatments) × 2 (14 d periods) Latin squares. This was repeated during SPR (16 May - 10 Jun), SUM (11 Jul - 5 Aug) and AUT (22 Aug 18 Sep). Rumen samples from these cows were taken after every milking (time of day: T; ~09:00 and ~16:00 h) on the 11th and 12th d of each period, and pH, volatile fatty acids (VFA) and ammonia content were measured. In AUT, five cows were added to each treatment (n=20) to estimate DM intake (DMI) and e-CH4 using the n-alkane (Mayes et al., 1986) and the SF6 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 792 (Zimmerman, 1993) techniques, respectively. Data were averaged as one value/cow/time point/period for the rumen sample data, and as one value/cow for e-CH4 data, and analysed using PROC MIX (SAS, 2005) with cow as a random factor. The e-CH4 data model included the effect of ST and the rumen data model included ST, period, T and the ST×T interaction. Results and discussion Pre-grazing HM was similar for each treatment in SPR and SUM (1600±65 and 1680±70 kg DM/ha, respectively), but was greater for GWc than GO in AUT (1880 and 1670±60 kg DM/ha, respectively). GWc sward-clover content was 7.5, 8.8 and 30.9% in SPR, SUM and AUT, respectively. There was a ST×T interaction for some rumen parameters (Table 1). Table 1. Effect of sward type (ST; GO: grass-only, GWc: grass-white clover), time of day and season on rumen volatile fatty acids (VFA) and pH of rumen-cannulated dairy cows. Spring (16 May to 10 Jun) Summer (11 Jul to 5 Aug) Autumn (22 Aug to 18 Sep) Time GO GWc sem ST ST×T GO GWc sem ST ST×T GO GWc sem ST ST×T Total VFA 09:00h 117 119 3.9 ns ns 109ab 104a 4.0 ns * 122 117 3.3 ns ns (mmol/L) 16:00h 137 146 3.9 115b 129c 4.0 158 158 3.3 Acetic acid 09:00h 63.3c 65.3d 0.78 ns *** 67.4 67.1 0.61 ns ns 66.2 67.2 0.45 * ns (%) 16:00h 58.2b 54.6a 0.78 61.8 62.6 0.61 59.3 60.0 0.45 Propionic acid 09:00h 20.2b 18.9a 0.60 ns * 16.2 16.7 0.58 ns ns 17.2 16.3 0.43 * ns (%) 16:00h 22.3c 22.5c 0.60 18.5 18.9 0.58 21.0 20.5 0.43 Butyric acid 09:00h 11.8a 11.3a 0.37 0.07 ** 12.7 12.4 0.46 * ns 13.0 12.7 0.28 ** ns (%) 16:00h 15.6b 17.7c 0.37 15.4 14.2 0.46 15.5 14.6 0.28 Valeric acid 09:00h 1.5a 1.4a 0.15 ns * 1.1 1.2 0.04 ns ns 1.3a 1.2a 0.05 ns * b c (%) 16:00h 1.8 2.1 0.15 1.5 1.5 0.04 1.6b 1.7b 0.05 Iso acids1 09:00h 3.2 3.2 0.15 ns ns 2.6 2.7 0.16 ns ns 2.3 2.5 0.14 *** ns (%) 16:00h 2.9 3.2 0.15 2.8 2.9 0.16 2.6 3.2 0.14 Lactic acid 09:00h 3.5 4.5 0.82 ns 0.07 4.9 4.9 0.81 ns ns 7.4 10.2 0.57 *** ns (mmol/L) 16:00h 3.6 2.61 0.82 4.7 4.24 0.84 3.4 4.72 0.57 Ammonia 09:00h 6.0 5.9 1.87 ns ns 11.2a 11.3a 1.11 * * 6.5a 9.1b 1.03 *** ** (mmol/L) 16:00h 13.2 16.1 1.87 . 17.6b 20.7c 1.11 20.3c 28.5d 1.03 Rumen pH 09:00h 5.86 5.91 0.075 ns ns 5.52 5.66 0.120 ns ns 5.29 5.37 0.069 ** ns 16:00h 5.30 5.21 0.075 5.99 6.13 0.120 5.81 6.06 0.069 1Isobutyric and isovaleric acids; Times with different superscript letters differ (P < 0.05). † = P < 0.10; * = P < 0.05; ** = P < 0.01; *** = P < 0.001. During SPR, clover inclusion increased acetic acid percentage (A%) in the morning, but lowered acetic A% in the afternoon. GWc cows had lower propionic A% in the morning than GO cows but similar in the afternoon. Clover inclusion increased butyric and valeric A% in the afternoon but had no effect in the morning. During SUM, clover inclusion increased total VFA and ammonia contents in the afternoon but not in the morning. There was a ST effect on butyric A% as GWc cows always had a lower butyric A% than GO cows. During AUT, clover inclusion increased ammonia concentration in morning and afternoon, but the difference between the ST was greater in the afternoon. There was a ST effect on a number of the rumen parameters in AUT. GWc cows had a greater acetic, lactic and iso A% (isobutyric plus isovaleric), but lower propionic and butyric A% than GO cows. The acetic and propionic A% responses to clover inclusion were similar to those reported by Ribeiro Filho et al. (2012) and were more evident when clover content was greatest (AUT). It was expected that clover inclusion would increase VFA; however, this was only evident in SUM afternoon samples. Iso-acids and ammonia contents responses may be related to the expected high dietary crude protein content of the GWc swards. Clover rumen fermentation effects were more evident in the afternoon, which may be related to GWc cows including clover in greater proportions in their diet in the morning grazing, as cows prefer clover compared to ryegrass (Rutter et al., 2004). Clover inclusion in swards had no effect on rumen pH during SPR and SUM but increased rumen pH in AUT, which was different to the findings of Ribeiro Filho et al. (2012). However, an in vitro study (Niderkorn Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 793 et al., 2011) found a higher pH when a binary mix of ryegrass and clover was incubated for 24 h, compared to ryegrass. Treatments had similar pre-grazing HM during the e-CH4 measurement week (1970±70 kg DM/ha) and GWc clover content was 24%. Clover inclusion had minimal effect on milk production or its composition (Table 2). Table 2. Effect of sward type (GO: grass-only, GWc: grass-white clover) on dairy cow performance and methane emissions during autumn (11 to 18 September). Cow production Methane emissions ratio (g CH4) GO GWc sem P-value GO GWc sem P-value g/cow/d 360 354 13.6 ns DMI (kg DM/cow/d) 15.0 16.5 0.6 0.07 g/kg of DMI 24.5 21.5 0.84 <0.05 Milk yield (kg/d) 14.1 14.0 0.64 ns g/kg of milk 26.5 26.0 1.14 ns MS yield (kg/d) 1.14 1.19 0.047 ns g/kg of MS 318 306 11.5 ns Fat (%) 4.38 4.72 0.097 <0.05 g/kg of milk fat 594 558 21.4 ns Protein (%) 3.76 3.87 0.059 ns g/kg of milk protein 690 680 27.4 ns DMI: dry matter intake; MS: milk solids. The lower milk fat content of the GO cows is in agreement with their lower rumen pH. There was no effect of treatment on e-CH4/cow/d. However, clover inclusion reduced the e-CH4 per kg DMI by 12%, comparable to the 9.6% reported by Lee et al. (2004). Similarly, those authors did not find differences in e-CH4 per unit output. Conclusion The mixture of ryegrass and clover influenced rumen characteristics with increased effect as the season progressed. The rumen fermentation effects of clover were more evident in the afternoon samples due a greater clover inclusion in the diet in the morning period. However, clover inclusion had little effect on milk production and only increased milk fat content. The presence of clover reduced e-CH4/kg DMI but did not affect the e-CH4 per unit output. Acknowledgements The Teagasc Walsh Fellowship Programme, Dairy Levy Fund and the European Community's Seventh Framework Programme (FP7/2007-2013, n° FP7-244983, Multisward). References Lee J.M., Woodward S.L., Waghorn G.C. and Clark D.A. (2004) Methane emissions by dairy cows fed increasing proportions of white clover (Trifolium repens) in pasture. Proceedings of the New Zealand Grassland Association 66, 151-155. Frame J. and Newbould P. (1986) Agronomy of white clover. Advances in Agronomy 40, 1-88. Mayes R.W., Lamb C.S. and Colgrove P.M. (1986) The use of dosed and herbage n-alkanes as markers for the determination of herbage intake. Journal of Agricultural Science 107, 161-170. Niderkorn V., Baumont R., Le Morvan A. and Macheboeuf D. (2011) Occurrence of associative effects between grasses and legumes in binary mixtures on in vitro rumen fermentation characteristics. Journal of Animal Science 89, 1138-1145. Ribeiro Filho H.M.N., Peyraud J.L. and Delagarde R. (2012) Foraging behavior and ruminal fermentation of dairy cows grazing ryegrass pasture alone or with white clover. Pesquisa Agropecuaria Brasileira 47, 458-465. Rutter S.M., Orr R.J., Yarrow N.H. and Champion R.A. (2004) Dietary preference of dairy cows grazing ryegrass and white clover. Journal of Dairy Science, 87, 1317-1324. Zimmerman P.R. (1993). U.S. Patent No. 5, 265,618. U. S. PaT. Office. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 794 Animal choice for grass-based systems Delaby L.1, Hennessy D.2, Gallard Y.3 and Buckley F.2 1 INRA, AgroCampus Ouest, UMR 1348, Physiologie, Environnement et Génétique pour l'Animal et les Systèmes d'Elevage, Saint-Gilles F-35590, France. 2 Teagasc, Animal and Grassland Research and Innovation Centre, Moorepark, Fermoy, Co. Cork. Ireland, 3 INRA, Experimental farm, UE 326, Borculo, Le Pin-au-Haras, F-61310 Exmes. Corresponding author: Deirdre.Hennessy@teagasc.ie Abstract Grass-based milk production systems require robust and ‘easy care’ cows capable of high levels of performance from grazed pasture. Cows intensively selected for milk yield are not well suited to seasonal grassland-based systems because of undesirable side effects on reproduction and survival. Dual-purpose or cross-breed cows should be more flexible and better adapted to grassland-based systems. The objective of the research reported in this paper was to compare the suitability of dairy cow breeds for pasture-based systems in France [Holstein (Ho) and Normande (No)] and in Ireland [Holstein-Friesian (HF) (and HF strains), Norwegian Red (NR) and HF×NR] in terms of milk production. In France, the high genetic-merit Ho cow is not suited for systems based on grassland and low concentrate input, whereas the No dual-purpose breed still performed well in these conditions and had higher fertility and milk value. In Ireland, crossing the HF with NR can increase overall animal performance by increasing herd health, fertility and milk value. Overall, the Ho/HF appears to be less flexible and poorly adapted to low-input systems based on the maximization of grassland use for milk production. Keywords: Dual purpose, crossbreeding, dairy cow, grass-based system, milk production Introduction The animal required for efficient grassland-based production systems must be robust and ‘easy care’, as well as being capable of high levels of performance from grazed pasture. Until recently most experimental results have indicated little or no importance of breed or strain by feedingsystem interaction. However, recent studies have shown large differences in performance (especially fertility and survival) and overall farm profitability between diverse breeds and strains of dairy cows (Dillon et al., 2007). Cows intensively selected for milk yield are not well suited to seasonal grassland-based systems because of undesirable side effects on reproduction and survival. Dual-purpose or cross-breed cows should be more flexible and better adapted to grassland-based systems. Dual-purpose dairy-breeds improve milk composition and increase beef value. Crossing the Holstein-Friesian (HF) with an alternative dairy breed sire can increase overall animal performances by increasing herd health, fertility and milk value. The objective of the research reported in this paper was to compare the suitability of dairy cow breeds for pasture-based systems in terms of milk production and fertility. Materials and methods An experiment was undertaken at the INRA experimental farm of Le Pin-au-Haras in the Northwest of France in Normandy (48.44ºN, 0.09ºE) comparing the Normande (No), a dual-purpose breed, with Holstein (Ho). From 2006, Ho (n=122) and No (n=112) dairy cows differing in genetic potential (evaluated within breed) either on milk potential (High Milk yield group – MY genetic group) or on milk fat and protein content potential (High Milk Solids content group – MSc genetic group), were allocated to two contrasting feeding strategies: the animal adapts itself to the feed available (Low) or the feed available is adapted in a way to satisfy the animal requirement and to allow it to express its genetic potential (High). The breeding period was 13 weeks and cows calved from January to March. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 795 In 2003, a farm participatory study was established in Ireland to enhance breeding value estimation for the Norwegian Red (NR) breed and NR × HF crossbreds. The study was a contemporary comparison design, whereby both parent breeds (NR and HF) and crossbreds (NR×HF) would be present on each farm to provide data relevant to breed and heterosis estimation. Semen from 10 proven NR AI sires was distributed to 55 commercial dairy herds to generate NR×HF crossbred females. In 2004, 393 purebred NR heifer calves sired by the same 10 proven NR AI sires used to generate the NR×HF animals were imported to Ireland. Animal performance data subsequently became available from 46 of these herds. All herds were milk recorded for five lactations and body weight and body condition score (BCS) were recorded. In order to augment the participatory data, herds containing both HF and NR genetics were identified from the national database. Because Ho and Friesian are considered different breeds within Irish genetic evaluations, and due to the intertwined nature of the two breeds within the cow population, it was considered appropriate to examine the relative breed and heterosis effect among the three breeds. Friesian genetics was further categorized as New Zealand Friesian (KF) and British (or European) Friesian (BF). Data from 2004 to 2010 was obtained for all herds. The production file used in routine genetic evaluations containing 305 d milk yields was provided by the Irish Cattle Breeding Federation (ICBF). Results and discussion France The Low feeding strategy resulted in a large reduction in total milk yield (-2000 kg/cow) and milk solids (MS) yield per cow (-70 kg and -72 kg of fat and protein/cow) (Table 1). Table 1. Impact of the breed (Holstein or Normande), the genetic group and the feeding strategy on the lactating and reproductive performance of the dairy cow. Breed Holstein Feeding strategy High MSc Normande Low MY MSc High MY MSc Effect (P<) Low MY Breed Feed strat. Genetic group (GG) MY MSc Milk yield (kg) 8841 8189 6368 5675 6608 6058 5010 4586 0.001 0.001 Fat content (g/kg) 35.8 40.2 36.9 42.1 39.6 42.5 40.1 43.4 0.001 0.001 Protein content (g/kg) 31.2 32.9 30.2 31.8 34.3 35.5 32.6 34 0.001 0.001 Milk Solids (fat +protein - kg) 587 587 421 414 476 457 356 346 0.001 0.001 Conceived (%) 84 61 86 82 0.01 0.01 Recalving (%) 59 44 71 68 0.01 0.01 The response to feeding strategy was greater (P<0.001) for Ho cows (-2500 kg milk yield and -168 kg MS) than for No cows (-1500 kg milk yield and -117 kg MS). The No cows produced milk with + 3.0 and + 2.3 g/kg for milk fat and protein content, respectively, in the High and Low feeding strategy. For both breeds, the High feeding strategy reduced milk fat content (1.5 and -0.8g/kg for Ho and No cows, respectively) and increased protein content (+1 and +1.6g/kg for Ho and No cows, respectively). The phenotypic expression of the genetic group for milk yield, fat and protein content were in agreement with the milk and MS content genetic index and did not have any interaction with feeding strategy. The reproductive performances were generally highly altered for the Ho cow with very low gestation rate especially in the Low feeding strategy. In contrast the No cow does not seem to be sensitive to feeding strategy. Ireland The Ho expressed a higher propensity for milk volume compared with the other breed groups investigated (Table 2). Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 796 Table 2. Breed and heterosis estimates for 305 d yields of milk, fat and protein. 1BF=British Friesian, KF=Kiwi Friesian, HO=Holstein, NR=Norwegian Red; 2SE=standard error Breed1 BF KF NR BF×HO KF×HO NR×HO Milk (kg) -597 -366 -504 122 249 161 6.3 27.6 26.0 4.7 18.8 24.8 Fat (kg) -20.6 3.3 -17.9 5.5 19.2 6.4 SE 0.25 1.08 1.02 0.18 0.74 0.97 Protein (kg) -18.3 -3.3 -14.7 4.4 11.4 6.3 SE 0.20 0.87 0.82 0.15 0.60 0.79 SE 2 Similarly, breed differences for fat and protein yields were lower for all three compared with the Ho. Although NR milk on average had higher fat and protein content compared to Ho, they produced lower 305-d yields of fat and protein. The data indicate that NR×HF dairy cows are capable of production levels per cow comparable to HF on low-cost systems, but fertility and survival levels are markedly improved, e.g. six-week in-calf rates were increased by over 10 percentage units with NR×HF. Body condition score, a trait genetically associated with differences in reproductive efficiency (Berry, 2003), was higher (P<0.001) for the NR at 3.03 compared with the HF at 2.85. That of the NR×HF (2.98) cows was similar to the Norwegian Red but higher (P<0.001) than the HF. Conclusion The most profitable breed is the one that returns the highest profit per unit of the most limiting input. In France, with contrasted feeding strategies, the Ho breed is much more sensitive compared to the No breed. The high genetic-merit Ho cow is not suited for systems based on grassland and no concentrate input, whereas the No dual-purpose breed still performed well in these conditions and had high fertility and milk value. In Irish grass-based systems crossing the Ho/HF with an alternative dairy breed sire such as NR can increase overall animal performance by increasing herd health, fertility and milk value. Overall, the Ho/HF appears to be less flexible and poorly adapted to low-input systems based on the maximization of grassland use for milk production. Acknowledgements The Research was funded through the European Community's Seventh Framework Programme (FP7/2007-2013, n° FP7-244983, ‘Multisward’). References Berry D.P., Buckley F., Dillon P., Evans R.D., Rath M. and Veerkamp R.F. (2003) Genetic correlations among body condition score, body weight, milk yield and fertility using random regression models. Journal of Dairy Science 86, 3704-3717. Dillon P., MacDonald K., Holmes C.W., Lopex-Villalobos N., Buckley F, Horan B. and Berry D. (2007) Cow genetics for temperate grazing systems. In: Australasian Dairy Science Sumposium 2007, Melbourne, 18-Oct2007, 152-184. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 797 Effect of sheep breed on lamb production from lowland pasture under continuous stocking Goliński P., Golińska B. and Biniaś J. Department of Grassland and Natural Landscape Sciences, Poznan University of Life Sciences (PULS), Dojazd 11, 60-632 Poznań, Poland Corresponding author: pgolinsk@up.poznan.pl Abstract The aim of this investigation was to evaluate the efficiency of lamb production depending on different breeds under grazing by continuous stocking on lowland pasture. The experiments were carried out in 2011-2013 on semi-natural lowland pasture. In the study, four different sheep breeds (4-4.5 month old) were included: 1) White-headed meat sheep, 2) Wielkopolska sheep, 3) Romanov sheep and 4) Blanc du Massif Central sheep. Forage from the pasture was the only feed the lambs received. It can be concluded that Wielkopolska sheep and white-headed meat sheep are suitable for efficient lamb production on lowland pasture in Poland. As a native breed the Romanov sheep is better for utilization of low quality pasture and can play an important role in environmental protection and management of the agricultural landscape, whereas Wielkopolska sheep prefer a good yielding sward under continuous stocking with intensive regrowth during favourable weather conditions for efficient lamb production. Blanc du Massif Central sheep did not appear well adapted for grazing conditions in Poland. Keywords: breed, continuous stocking, lamb production Introduction From the beginning of the last decade of the 20th century in many European countries there has been a decline in the sheep population. In Poland the total population of sheep in 2011 reached about 212.7 thousand heads and total ewes number about 143.8 thousand heads (Goliński, 2012). Sheep, as a grazing animal, are very important for maintaining the multi-functionality of grassland. Therefore, research work concerning the sheep characteristics and management to increase environmental benefits and competitiveness of grassland-based systems are necessary (Niżnikowski et al., 2010; Goliński, 2012). The aim of this study was to evaluate the efficiency of lamb production, for different breeds, under grazing by continuous stocking on lowland pastures. Materials and methods The pasture experiment was established in spring 2011 and was completed in October 2013. The experiment took place on semi-natural lowland pasture at the Brody Experimental Station of the Poznan University of Life Sciences (52º 26N, 16º 18E; 92.0 m a.s.l.; long-term mean annual rainfall 601 mm, and mean air temperature 8.3 °C) on a mineral soil, fertilized at the level 50 kg N ha-1, 40 kg P2O5 ha-1 and 60 kg K2O ha-1 applied in spring. In April 2011 the experimental area was fenced. Four paddocks (P1-P4) were prepared, each of 1000 m2, for continuous stocking of five lambs per different breed during the vegetative period. In the experiment the following sheep genotypes (4-4.5 month old) were included: P1) White-headed meat sheep, P2) Wielkopolska sheep, P3) Romanov sheep, P4) Blanc du Massif Central sheep. The grazing seasons in 2011, 2012 and 2013 lasted 160 days from 30 April to 7 October. Forage from the pasture was the only feed the lambs received. The animals were on the pasture during the day and night, and fresh water and minerals were available. The sward height was maintained at a constant level of 5-7 cm by changing the stocking rate. A rising plate meter was used to measure sward height. On each paddock, ungrazed plots of 3 m2 were established to determine sward botanical composition and pasture yield. Sward yield was influenced by Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 798 weather conditions during the grazing period (April to September, respectively) characterized by following amounts of monthly rainfall (mm) in 2011: 14.0, 34.0, 52.6, 175.4, 34.5, 46.0, and air temperature (monthly means, °C): 11.7, 14.1, 18.6, 17.9, 18.8, 15.3. In 2012 the rainfall reached following amounts: 22.0, 77.2, 163.0, 197.6, 60.1, 30.0 mm, and air temperatures (as monthly means) were: 8.8, 14.8, 16.0, 19.2, 18.7, 14.3 °C. In 2013 the rainfall reached following amounts: 15.4, 69.8, 125.3, 67.3, 51.5, 33.7 mm and monthly mean air temperatures: 8.0, 14.4, 17.3, 20.1, 19.1, 12.9 °C. The daily liveweight gain of the lambs was measured by systematically weighing animals at the end of each month (May-September). Sward botanical composition was evaluated using the point method of Levy and Madden (1933). On the ungrazed plots (1.5 m × 2 m) sward yield was determined each month. Tests of the main effects were performed by F-test. Means were separated by LSD and were declared different at P < 0.05. Results and discussion Botanical composition of the pasture sward was similar on each paddock and was dominated by grasses, particularly by Lolium perenne. In the sward, high proportions of Festuca rubra, Phleum pratense and Festuca arundinacea were also determined. Trifolium repens occupied only 1.4-1.8%. From the group of herbs Taraxacum officinale occurred in quantities of 1.32.2%. The total number of species in the pasture sward ranged from 24 to 25, depending on the paddock. It could be concluded that the botanical composition had no significant impact of feed value of the grazed sward by different lamb breeds. The DM yield of pasture showed significant variations between grazing months and years. Pasture yield was affected by weather conditions. The highest sward yield (1.8 t ha-1 DM) in 2011 was recorded in August, which can be attributed to very good weather conditions for the vegetation. In 2012 the highest sward DM yield was determined in May and reached, on average, 2.6 t ha-1 DM. In this period a large area of pasture was cut and the sward was conserved as hay. In the following months sward yield on the pasture was significantly lower. The lowest DM yield of pasture in the 3-year study period was determined in 2013. The results presented in Table 1 show that Romanov sheep breed had the highest daily liveweight gain during the grazing season in 2011. Table 1. Liveweight gain of lambs during the grazing period in 2011-2013 (g day-1) Sheep breed May White-headed meat sheep Wielkopolska sheep Romanov sheep Blanc du Massif Central sheep LSD0.05 135.8 84.2 197.9 113.7 - White-headed meat sheep Wielkopolska sheep Romanov sheep Blanc du Massif Central sheep LSD0.05 121.8 155.2 117.0 108.5 - White-headed meat sheep Wielkopolska sheep Romanov sheep Blanc du Massif Central sheep LSD0.05 64.0 92.2 82.2 75.7 - Period of lamb gains (month) June July August 2011 72.6 68.3 84.3 81.7 102.8 114.0 97.7 72.6 77.9 91.5 101.5 86.1 2012 102.9 116.1 93.4 82.3 106.5 101.4 74.0 97.5 94.0 68.6 101.6 83.6 2013 81.3 73.9 78.2 83.2 60.5 71.7 68.4 50.9 81.2 67.1 59.3 75.5 - September Means 73.6 81.6 73.2 28.0 - 86.9b 92.9ab 103.9a 84.2b 11.34 155.3 165.8 147.5 125.7 - 117.9a 122.2a 106.0b 97.6b 8.57 65.1 65.4 84.8 68.8 - 72.5 74.6 73.5 69.3 ns Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 799 In the next months the daily liveweight gain of this breed was lower. The lambs of the Wielkopolska sheep had lower liveweight gain compared to Romanov sheep, but the means for the grazing season between these breeds were not statistically significant. The lambs of Wielkopolska sheep had the highest liveweight gain in summer when the pasture yield was greatest. Lamb liveweight gain of the other two breeds, white-headed meat sheep and Blanc du Massif Central sheep, were significantly lower. In 2012 the best results for whole grazing season were for the Wielkopolska sheep. Slightly lower, but not significantly different, liveweight gain was observed for the white-headed meat sheep. Lambs of the Romanov sheep had lower liveweight gain, particularly in summer. In 2013 the liveweight gains of the four breeds did not differ significantly and were lower in comparison with previous years. It could be concluded the lambs of each breed showed in each month specific rhythm of pasture sward intake, which is dependent on the availability of fodder, its botanical and chemical composition and also on weather conditions. A good example is provided by the liveweight gains of lambs of all breeds in September: in 2011 this was the lowest in the whole season and in 2012 the highest. There are probably other conditions besides those evaluated in this study that influenced the differences in liveweight gain such occurrence of fungal diseases on the sward (Niżnikowski et al., 2010). Conclusion Wielkopolska sheep and Romanov sheep performed well. Wielkopolska sheep had liveweight gains that were not statistically different from Romanov in 2011 and 2013, and the highest liveweight gain in 2012. Romanov had the highest liveweight gain in 2011. Similar to Wielkopolska sheep is the white-headed meat sheep. These two breeds are suitable for efficient lamb production on lowland pasture in Poland. As a native breed the Romanov sheep is better for utilization of low quality pasture and can play an important role in environmental protection and management of agricultural landscapes, whereas Wielkopolska sheep prefer a good yielding sward and continuous stocking with intensive regrowth during favourable weather conditions for efficient lamb production. Blanc du Massif Central sheep did not appear well adapted for grazing conditions in Poland. Acknowledgements The research leading to these results received funding from the European Community's Seventh Framework Programme (FP7/ 2007-2013) under the grant agreement n° FP7-244983 (Multisward) and from Polish Ministry of Science and Higher Education (2162/7PR/2011/2). References Goliński P. (2012) Grazing in Poland. http://www.europeangrassland.org/fileadmin/media/ pdf/working_groups/Grazing_in_Poland._Piotr_Golinski_01.pdf. Niżnikowski R., Oprządek A., Strzelec E., Popielarczyk D. and Głowacz K. (2010) Comparison of reproduction level and body conformation of Suffolk and Charollais sheep breed in Poland. Annals of Warsaw University of Life Sciences – SGGW, Animal Science 47, 143-148. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 800 Appreciation of the functions of grassland by Belgian stakeholders De Vliegher A.1, Dufrasne I.2, Schellekens A.3, Peeters A.4, Van den Pol-van Dasselaar A.5 1 Institute for Agricultural and Fisheries Research (ILVO), BE-9820 Merelbeke, Belgium 2 ULg - FMV, Station Expérimentale Chemin de la Ferme 6, BE-4000, Liège, Belgium 3 Landbouwcentrum voor Voedergewassen vzw, Hooibeeksedijk 1, B-2440 Geel, Belgium 4 RHEA, Research Centre, Rue Warichet, 4 Box 202, BE - 1435 Corbais, Belgium 5 Wageningen UR Livestock Research, P.O. Box 65, 8200 AB Lelystad, the Netherlands Corresponding author: Alex.Devliegher@ilvo.vlaanderen.be Abstract The European project MultiSward studied the appreciation of different functions of grasslands by European stakeholders. This paper describes the importance of grasslands for stakeholders in Belgium. Belgium currently has 578 504 ha of grassland, which is 43.3% of the total agriculturally utilized area. Belgian stakeholders appreciate grasslands especially for feed protein delivery at farm level, as a source of high quality forage and low cost animal feed especially for dairy milk production under grazing conditions. The most appreciated ecological aspects are conservation of ecosystems and biodiversity, erosion control and beauty of the landscape. Keywords: grassland, stakeholders, Belgium Introduction Grasslands cover 43.3% (578 504 ha) of the agricultural area (AA) in Belgium: 38.0% is permanent grassland and 5.3% is temporary grassland. The area and the proportion of grasslands are higher in Wallonia (350 000 ha and 49% of AA) than in Flanders (228 400 ha and 37% of the AA) (Anonymous, 2014). The main forage production system in Flanders is based on regularly resown pastures and on annual forage crops (temporary pastures and silage maize). The system in Wallonia is based mainly on permanent pastures. Extensive grasslands are rare in Belgium. Intensive grasslands are dominated by Lolium perenne L. Compared with the European average, cattle intensification level is high to very high. The aim of the current study was to get an insight into the importance of grasslands for stakeholders in Belgium. Materials and methods An on-line questionnaire on functions of grasslands was distributed throughout Europe. This study provides results for Belgian stakeholders. A detailed description of the method can be found in Van den Pol-van Dasselaar et al. (2014, this volume). In Belgium, the questionnaire was distributed to members of the ILVO network, Landbouwcentrum voor Voedergewassen and the FP7-project Autograssmilk. In addition, it was spread via social media. Results and discussion When the questionnaire closed, 209 valid responses had been obtained. The majority of responses came from farmers (77 responses, or 37% of total response), followed by researchers (32 or 15%), advisers (30 or 14%), students (29 or 14%). policy makers (17 or 8%), industry (13 or 6%), education (6 or 3%) and NGOs (5 or 2%). The answers are grouped into four groups of ecosystem services: provisioning, regulating, supporting and cultural services. For the Belgian stakeholders the main provisioning service is to deliver high quality forage at a low cost for animal feed, mainly used for dairy milk production (Table 1). Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 801 Table 1. Importance of provisioning services of grasslands according to the respondents of the questionnaire (1 = not important; 5 = very important) Advice Education Farmers High quality forage Dairy cow milk production 4.2 4.2 4.0 3.7 4.4 4.3 Low cost animal feed 4.2 4.8 Nutritional quality of animal products for human consumption Beef meat production 3.8 Industry NGO Policy maker Research Students 4.1 4.2 4.0 4.6 3.9 4.1 4.3 4.5 4.0 4.0 4.4 4.1 4.8 3.9 4.4 3.7 4.5 3.8 3.5 3.4 3.9 4.1 3.7 4.1 3.8 3.7 3.9 4.4 3.9 4.0 3.8 Global food production 3.7 Region of origin of animal products 3.8 3.7 3.6 3.5 3.2 3.1 3.7 3.8 3.7 2.9 3.0 3.0 3.0 3.5 3.6 Honey production Sheep meat production 2.5 2.8 2.3 2.2 1.8 2.3 2.4 3.0 2.7 2.3 2.1 2.5 1.8 2.2 2.1 2.8 Biomass for energy production 2.1 2.3 1.7 2.0 1.8 2.1 1.9 2.5 Sheep milk production 2.2 2.2 2.0 2.2 2.0 2.0 1.8 2.4 Goat milk production 2.5 2.5 2.1 2.3 2.0 1.9 2.1 2.1 Wool production 2.1 1.7 1.7 1.8 1.6 1.9 1.7 2.3 Goat meat production 2.2 2.2 1.6 2.1 1.6 1.9 1.8 2.0 Production of plant fibre 1.8 2.0 1.5 1.9 1.4 1.6 1.6 2.6 Table 2. Importance of regulating services of grasslands according to the respondents of the questionnaire (1 = not important; 5 = very important) Advice Education Farmers Industry NGO Policy maker Research Students Biodiversity 3.7 Conservation of ecosystems quality 3.3 3.8 2.9 3.2 3.8 4.0 3.1 3.2 4.0 3.5 3.8 3.8 3.5 4.1 Water catchment 3.6 3.2 2.9 2.4 3.9 3.2 3.2 3.6 3.2 Erosion control 3.5 3.7 3.3 3.3 3.8 3.6 3.9 3.3 Carbon sequestration 3.7 3.5 Mitigating greenhouse gas emissions 3.5 3.5 3.3 3.3 3.0 3.9 4.1 3.4 2.7 3.2 2.4 3.1 3.5 3.2 Adaptation to climate change 3.1 Flood plains rivers 2.9 3.5 2.7 2.6 2.0 3.0 3.4 3.1 2.3 2.4 2.4 2.8 3.1 3.2 3.1 Pathogen control in cropping system Fire control 3.4 2.1 3.5 1.8 3.2 2.0 3.6 2.2 2.2 1.2 3.6 2.1 3.6 1.8 3.5 2.8 Avalanche control 1.8 1.5 1.5 1.2 2.4 1.6 1.5 1.8 Table 3. Importance of cultural services of grasslands according to the respondents of the questionnaire (1 = not important; 5 = very important) Advice Education Farmers Industry NGO Policy maker 3.5 4.1 4.0 3.3 3.4 4.2 4.7 3.8 3.8 4.6 Rural development 3.4 3.2 2.7 2.8 Maintaining population in rural areas 3.4 4.0 3.1 Cultural values 2.7 4.0 Tourism / recreation 3.0 3.2 Supporting horses for equestrian sport and recreation 2.2 2.2 Beauty of the landscape Positive perception production systems of animal Research Students 3.3 4.0 3.7 3.4 4.1 3.7 2.6 3.1 3.6 3.0 3.4 3.2 3.2 3.8 3.5 2.2 2.5 3.2 2.8 3.1 2.7 2.1 2.4 3.8 2.8 3.0 2.4 1.6 2.0 1.6 2.5 2.4 3.1 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 802 The most important regulating services are conservation of ecosystems, biodiversity and erosion control (Table 2). Belgian stakeholders establish the important role of grassland in the positive perception of animal production systems and beauty of the landscape. The score for tourism/recreation is rather low (Table 3). The most important supporting services are grazing and feed protein supply at farm level (Table 4). Table 4. Importance of supporting services of grasslands according to the respondents of the questionnaire (1 = not important; 5 = very important) Advice Education Farmers Industry NGO Grazing Animal health 4.3 3.9 4.5 4.2 4.2 4.2 3.6 3.9 4.0 3.4 Animal welfare 3.7 3.8 4.0 3.9 Conservation of soil structure and fertility in cropping systems 4.1 3.2 3.8 Feed protein supply at farm level 4.4 4.4 Competitiveness of farming systems 3.6 3.7 N fixation via legumes 3.7 Availability of water 3.4 Crop pollination 2.9 Policy maker Research Students 4.0 3.3 4.5 3.8 4.3 4.4 3.4 3.2 3.8 4.3 3.5 4.0 4.2 4.0 3.9 4.2 3.5 4.0 3.9 4.5 3.4 3.3 3.2 3.8 2.9 3.6 2.9 4.0 3.4 3.0 3.4 3.5 4.0 3.4 3.0 3.1 2.9 2.8 3.1 3.4 3.7 3.0 2.5 2.5 2.2 2.6 2.9 3.6 Conclusion Belgian stakeholders appreciate grassland especially for feed protein delivery at farm level, as a source of high quality forage and low cost animal feed especially for dairy milk production under grazing conditions. The most appreciated characteristics are conservation of ecosystems and biodiversity, erosion control and beauty of the landscape. Grasslands induce a positive perception of animal production systems and are considered as essential components of animal health and welfare. Acknowledgements This research has received funding from the European Community's Seventh Framework Programme (FP7/ 2007-2013) under the grant agreement n° FP7-244983 (Multisward) and was supported by Autograssmilk (grant agreement n° FP7-314879). References Anonymus (2014) http://economie.fgov.be/nl/modules/publications/statistiques/economie/downloads/ landbouw_-_landbouwgegevens_van_2012.jsp Van den Pol-van Dasselaar A., Goliński P., Hennessy D., Huyghe C., Parente G. and Peyraud J.L. (2014) Appreciation of the functions of grasslands by European stakeholders. Grassland Science in Europe 19 (this volume). Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 803 Appreciation of the functions of grassland by Dutch stakeholders Van den Pol-van Dasselaar A. and Stienezen M.W.J. Wageningen UR Livestock Research, P.O. Box 65, 8200 AB Lelystad, the Netherlands Corresponding author: agnes.vandenpol@wur.nl Abstract The European project MultiSward studied the appreciation of different functions of grasslands by European stakeholders. This paper describes the importance of grasslands for stakeholders in the Netherlands. There is currently 1 million ha of grassland in the Netherlands, which is 4045% of the total agriculturally utilized area. Dutch stakeholders appreciate the different functions of grasslands, especially high quality forage, dairy cow milk production, low cost animal feed and grazing. Functions that are less relevant for the Netherlands, like sheep and goat production, fire control and avalanche control, are also less appreciated. We conclude that stakeholders appreciate grasslands in the Netherlands as a valuable resource for many ecosystem services. Keywords: grassland, stakeholders, the Netherlands Introduction Grasslands cover almost 1 million ha in the Netherlands, which is 40-45% of the total agricultural area. The majority of these grasslands (75%) are permanent grasslands; 20% consists of temporary grasslands and 5% is natural grasslands (CBS, 2014). Grasslands are usually intensively managed and dominated by perennial ryegrass (Lolium perenne L.). The grasslands are used as feed for dairy cattle. The Dutch dairy sector is characterized by relatively high levels of supplementation, mainly silage maize and concentrates. Silage maize covers about 10% of the total agricultural area in the Netherlands (CBS, 2014) and is fully used for dairy feed. About 70% of the dairy cattle are grazing for at least part of the grazing season (CBS, 2014). The aim of the current study was to obtain an insight into the importance of grasslands for stakeholders in the Netherlands. Materials and methods An on-line questionnaire on the appreciation of functions of grasslands was distributed throughout Europe. A detailed description of the method can be found in Van den Pol-van Dasselaar et al. (2014, this volume). This paper describes the results for the Netherlands. In the Netherlands, the questionnaire was distributed to members of the Netherlands Society for Grassland and Fodder Crops and the Dutch farmers’ association LTO. In addition, it was distributed via social media, like LinkedIn. Results and discussion At the time of closing the questionnaire, 206 valid responses had been obtained. The majority of these responses came from farmers (90 responses, which equals 44% of total response), followed by advisers (34; 17%), researchers (30; 15%), industry (29; 14%), policy makers (8; 4%), education (8; 4%) and NGOs (5; 2%). Students were not included, since only 2 students responded. Tables 1 to 4 show the appreciation of the functions of grasslands by the different stakeholders. The functions are grouped into the four groups of ecosystem services: provisioning services, regulating services, supporting services and cultural services. Many provisioning services are highly appreciated by Dutch stakeholders but a number of them, such as sheep and goat production, are less relevant under Dutch conditions. This is also true for some regulating functions like fire control and avalanche control. Cultural services, such as the contribution of grasslands to the beauty of the landscape and the perception of animal production systems, are seen as important. With respect to supporting services, animal health Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 804 is valued highly, especially by farmers. Farmers and policy makers consider feed protein supply at farm-level also to be an important function of grasslands. Table 1. Importance of provisioning services of grasslands according to the respondents of the questionnaire (1 = not important; 5 = very important). High quality forage Dairy cow milk production Low cost animal feed Nutritional quality of animal products for human consumption Beef meat production Global food production Region of origin of animal products Honey production Sheep meat production Biomass for energy production Sheep milk production Goat milk production Wool production Goat meat production Production of plant fibre Advice Education Farmers Industry NGO Policy maker Research 4.4 4.6 4.4 3.9 4.0 4.3 3.8 3.6 4.8 4.5 4.6 3.9 4.6 4.6 4.3 4.0 4.0 4.2 3.2 3.2 4.0 4.6 3.9 3.8 4.5 4.7 4.1 3.9 2.9 3.6 3.5 2.4 2.1 1.8 2.1 2.3 1.8 1.8 1.8 2.8 3.8 3.4 2.8 2.6 2.3 2.4 2.1 2.0 1.6 1.8 3.0 4.0 2.9 2.0 2.1 1.5 1.7 1.8 1.7 1.4 1.5 3.7 4.2 3.6 2.3 2.9 2.3 2.6 2.8 2.2 2.3 2.2 2.8 2.6 3.6 2.8 3.2 2.0 3.2 2.8 2.8 2.6 1.8 2.4 3.5 3.5 2.6 1.9 1.8 1.5 1.8 2.1 1.3 2.3 3.3 3.7 3.1 2.1 2.3 2.1 2.0 2.0 2.0 1.7 1.9 Table 2. Importance of regulating services of grasslands according to the respondents of the questionnaire (1 = not important; 5 = very important). Biodiversity Conservation of ecosystems quality Water catchment Erosion control Carbon sequestration Mitigating greenhouse gas emissions Adaptation to climate change Flood plains rivers Pathogen control in cropping system Fire control Avalanche control Advice Education Farmers 3.6 3.8 3.3 2.6 3.3 3.3 3.5 3.0 3.5 1.9 1.4 4.1 4.0 2.8 2.9 2.8 3.8 2.7 3.6 3.3 1.8 1.8 3.2 3.3 2.4 2.3 2.9 2.9 2.8 2.3 3.1 1.7 1.3 Industry 3.3 3.4 3.2 3.0 3.2 2.9 2.8 3.3 3.4 2.2 1.6 NGO 4.4 4.6 2.8 3.5 4.0 4.0 3.4 3.4 3.2 2.4 2.0 Policy maker Research 3.5 3.9 3.3 2.4 3.9 3.9 3.6 4.6 3.3 1.6 1.1 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 805 3.7 3.4 3.2 2.5 3.5 3.2 3.2 3.2 3.2 1.8 1.7 Table 3. Importance of cultural services of grasslands according to the respondents of the questionnaire (1 = not important; 5 = very important). Beauty of the landscape Positive perception of production systems animal Rural development Maintaining population in rural areas Cultural values Tourism / recreation Supporting horses for equestrian sport and recreation Advice Education Farmers Industry NGO Policy maker Research 3.9 4.4 4.1 3.8 3.8 4.1 3.6 3.9 4.2 4.2 4.0 4.3 4.1 3.9 3.3 3.5 3.6 3.1 2.1 3.1 3.8 3.3 3.9 2.1 3.3 3.7 2.9 2.7 1.7 3.3 3.6 3.2 3.1 2.8 3.6 4.2 3.4 3.2 3.3 3.0 3.6 4.0 3.3 2.0 3.2 3.5 3.6 3.4 2.3 Table 4. Importance of supporting services of grasslands according to the respondents of the questionnaire (1 = not important; 5 = very important). Advice Education Farmers Industry NGO Policy maker Research Grazing Animal health Animal welfare Conservation of soil structure and fertility in cropping systems 4.5 4.0 3.9 4.1 3.8 3.9 3.9 4.1 4.2 4.4 4.0 4.0 4.3 4.1 4.0 3.9 4.0 3.8 4.2 3.6 4.8 3.9 4.1 4.0 4.3 4.0 3.7 4.0 Feed protein supply at farm level Competitiveness of farming systems N fixation via legumes Availability of water Crop pollination 4.4 3.5 3.4 3.4 2.6 3.9 3.3 4.1 3.8 2.7 4.7 3.7 3.2 3.5 2.4 4.3 3.6 3.4 3.6 2.6 3.8 2.8 3.4 3.8 3.6 4.8 3.9 3.9 3.4 2.8 4.4 3.9 3.3 3.3 2.3 Conclusion Dutch stakeholders appreciate the different functions of grasslands, especially high quality forage, dairy cow milk production and grazing. Functions that are less relevant for the Netherlands are also less appreciated. We conclude that stakeholders appreciate grasslands in the Netherlands as a valuable resource for many ecosystem services. Acknowledgements The research leading to these results has received funding from the European Community's Seventh Framework Programme (FP7/ 2007-2013) under the grant agreement n° FP7-244983 (Multisward) and n° FP7-314879 (Autograssmilk) and from the Dutch ministry of Economic Affairs (KB-12-006.04-003 and BO-22.04-005-002). References CBS (2014) StatLine databank, http://statline.cbs.nl/. Van den Pol-van Dasselaar A., Goliński P., Hennessy D., Huyghe C., Parente G. and Peyraud J.L. (2014) Appreciation of the functions of grasslands by European stakeholders. Grassland Science in Europe 19 (this volume). Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 806 Appreciation of the functions of grassland by French stakeholders Huyghe C.1, Peyraud J.-L.2, Brocard V.3, Van den Pol-van Dasselaar A.4 1 INRA, Poitou-Charentes Research Centre, 86600 Lusignan, France 2 INRA, UMR-1348, Joint Research Unit PEGASE, F-35590 St Gilles, France 3 Institut de l’Elevage, BP 85225, 35652 Le Rheu Cedex, France 4 Wageningen UR Livestock Research, P.O. Box 65, 8200 AB Lelystad, the Netherlands Corresponding author: Christian.huyghe@lusignan.inra.fr Abstract The European project MultiSward studied the appreciation of different functions of grasslands by European stakeholders. This paper describes the importance of grasslands for stakeholders in France. France currently has 11 million ha grasslands, which is 38% of the total agriculturally utilized area. French stakeholders especially appreciate the functions of production of forage quality, low production costs and production of protein as well as the suitability for grazing or biodiversity production. Functions like production of biomass for energy production or for fibre are less appreciated. We conclude that French stakeholders recognize the high potential of grasslands for developing animal production systems that are economically viable and environment-friendly. Keywords: grassland, stakeholders, France Introduction France has a very large acreage devoted to permanent and temporary grasslands, contributing to a total of 11 million ha (38% of the total agricultural utilized area). Annual forage crops are very important, especially silage maize (5.5% of the total agricultural utilized area). At present, most temporary grasslands are sown with mixtures of grasses and legumes producing large amounts of protein-rich feed. They also include a significant acreage of lucerne in some specific regions, grown as pure crops, although this acreage has declined since 1990. Permanent grasslands are abundant, especially in mountainous areas and hilly regions, where they are the main feed resources for the large herds of suckler-cows, but also in Normandy and eastern France for dairy herds where it is not possible to plough. Despite an important reduction in the last three decades, grasslands are a major component of most French landscapes. Dairy farming is important in lowland areas of the western part of France and also in the mountainous regions in the East (Franche-Comté and Alps) and in the Massif Central. In these regions, milk is processed to produce PDO cheese. High milk yield and animal performances and low production costs are key issues for the farmers, who are concerned by the work load and by the preservation of the environment. The aim of the current study was to get an insight into the appreciation of the functions of grasslands by stakeholders in France. Materials and methods This study provides results that were obtained via an on-line questionnaire on the functions of grasslands and which was distributed throughout Europe. A detailed description of the method can be found in Van den Pol-van Dasselaar et al. (2014, this volume). This paper describes results for French stakeholders. In France, the questionnaire was distributed to members of the French Grassland Society (AFPF) and promoted during conferences of this society. Results and discussion At the time of closing the questionnaire, 356 valid responses had been obtained from France. The majority of these responses came from advisers (142 responses, which equals 40% of total response) followed by research (116; 32%), industry (24; 7%), farmers (24; 7%), policy makers (21; 6%), education (17; 5%) and students (10; 3%). Three responses only were provided by Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 807 members of NGOs and their ratings will be little discussed. Tables 1 to 4 show the appreciation of the functions of grasslands by the different stakeholders. French stakeholders gave a wellbalanced appreciation across the various ecosystem services: provisioning services, regulating services, supporting services and cultural services. Table 1. Importance of provisioning services of grasslands according to the respondents of the questionnaire (1 = not important; 5 = very important). Advice Education Farmers Industry NGO Policy maker Research Students High quality forage 4.1 3.8 4.1 3.6 4.3 4.1 3.7 4.2 Dairy cow milk production Low cost animal feed 4.0 3.6 3.8 4.0 4.3 3.9 4.1 3.7 3.0 4.0 3.8 4.2 2.7 4.4 4.0 4.4 Nutritional quality of animal products for human consumption Beef meat production 3.7 3.5 4.0 3.2 4.3 4.4 3.7 3.9 4.2 3.8 4.0 4.0 3.7 4.3 3.7 3.6 Global food production 3.8 Region of origin of animal products 3.5 3.4 3.6 3.3 3.7 3.8 3.5 3.3 3.6 3.7 3.5 4.0 3.8 3.5 3.0 Honey production Sheep meat production 2.8 2.9 2.5 2.7 3.0 2.9 2.9 2.7 3.4 2.8 3.1 3.4 2.7 3.7 3.1 2.4 Biomass for energy production 1.9 2.1 1.7 2.3 2.3 1.8 2.0 2.2 Sheep milk production 2.4 2.9 2.1 2.9 2.0 2.2 2.4 2.2 Goat milk production 2.7 2.7 2.4 2.5 2.3 2.7 2.4 2.1 Wool production 1.8 2.0 1.8 2.2 2.0 1.8 1.7 2.0 Goat meat production 1.7 2.0 1.8 2.3 1.3 1.9 1.8 2.1 Production of plant fibre 1.6 1.6 1.7 2.1 1.7 1.6 1.8 1.4 Table 2. Importance of regulating services of grasslands according to the respondents of the questionnaire (1 = not important; 5 = very important). Advice Education Farmers Industry NGO Policy maker Research Students Biodiversity 4.1 Conservation of ecosystems quality 3.9 4.3 3.9 3.6 4.7 3.9 3.6 3.7 4.0 4.6 4.2 4.3 4.5 4.0 3.7 Water catchment Erosion control 4.0 4.1 3.7 3.9 3.6 3.5 3.7 3.7 4.0 3.9 3.9 3.2 2.7 3.8 3.6 3.8 Carbon sequestration 3.5 3.6 3.5 Mitigating greenhouse gas emissions 3.5 3.2 3.3 3.3 2.0 3.8 3.8 3.2 3.3 2.3 3.5 3.4 3.1 Adaptation to climate change 3.4 3.6 3.8 3.3 3.0 3.7 3.3 3.0 Flood plains rivers 2.9 Pathogen control in cropping system 3.1 3.3 3.4 3.0 3.3 3.5 3.0 2.3 3.0 3.1 3.0 2.0 3.0 3.0 3.4 Fire control Avalanche control 2.2 2.0 2.4 2.4 3.0 2.5 2.2 2.4 1.8 1.7 1.6 1.9 2.3 2.0 2.1 1.6 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 808 Table 3. Importance of cultural services of grasslands according to the respondents of the questionnaire (1 = not important; 5 = very important). Farmers Industry NGO Policy maker 3.9 4.2 3.2 4.7 4.2 3.9 4.0 3.8 Rural development 3.3 3.6 3.2 Maintaining population in rural areas 3.2 3.7 3.5 Cultural values 3.4 3.2 Tourism / recreation 3.2 Supporting horses for equestrian sport and recreation 2.2 Beauty of the landscape Positive perception production systems of animal Advice Education Research Students 4.0 4.3 3.9 3.8 4.0 4.4 4.0 3.4 3.2 3.7 3.9 3.4 3.6 3.0 3.7 4.1 3.4 3.3 3.0 2.8 4.3 3.9 3.4 2.8 3.6 3.2 2.8 4.0 3.8 3.3 2.5 2.1 2.1 2.3 3.0 2.4 2.3 2.1 Table 4. Importance of supporting services of grasslands according to the respondents of the questionnaire (1 = not important; 5 = very important). Advice Education Farmers Industry NGO Policy maker Grazing 4.6 Animal health 3.7 Animal welfare Conservation of soil structure and fertility in cropping systems Research Students 4.3 4.3 4.0 3.7 3.4 4.0 3.3 3.7 4.5 4.4 4.6 3.6 3.4 3.5 3.9 3.4 4.1 3.2 3.9 3.6 4.0 3.9 3.7 3.8 3.5 4.1 4.3 3.9 3.9 4.1 Feed protein supply at farm level 4.0 3.9 4.3 Competitiveness of farming systems 3.9 3.6 4.0 3.6 3.0 4.0 3.7 3.8 3.5 2.3 4.0 3.6 3.3 N fixation via legumes 4.0 3.9 Availability of water 3.2 2.8 4.0 3.8 2.3 3.7 3.7 4.3 3.4 3.3 2.7 3.4 3.3 3.1 Crop pollination 2.9 3.2 3.0 3.2 2.0 3.0 3.2 3.1 Conclusion French stakeholders provided a high number of responses to the on-line questionnaire. There were very subtle differences among groups of stakeholders, and this means that their views related to the most important functions are commonly shared and that a policy in favour of grasslands would probably be approved by all stakeholder groups. Provisioning and supporting services are considered as essential. But regulating services, especially biodiversity, and cultural services are also identified as important. Acknowledgements This research has received funding from the European Community's Seventh Framework Programme (FP7/ 2007-2013) under the grant agreement n° FP7-244983 (Multisward) and was supported by Autograssmilk (grant agreement n° FP7-314879). References Van den Pol-van Dasselaar A., Goliński P., Hennessy D., Huyghe C., Parente G. and Peyraud J.L. (2014) Appreciation of the functions of grasslands by European stakeholders. Grassland Science in Europe 19 (this volume). Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 809 Appreciation of the functions of grassland by Irish stakeholders Hennessy D.1 and Van den Pol-van Dasselaar A.2 1 Teagasc, Animal and Grassland Research and Innovation Centre, Moorepark, Fermoy, Cork, Ireland; 2 Wageningen UR Livestock Research, the Netherlands Corresponding author: Deirdre.Hennessy@teagasc.ie Abstract The European project MultiSward studied the appreciation of different functions of grasslands by European stakeholders. This paper describes the importance of grasslands for stakeholders in Ireland. Ireland currently has approximately 4.6 million ha of grassland, which is 90% of the total utilized agricultural area. Irish stakeholders consider grassland to be important for a range of functions and services including milk and meat production, forage production, animal health and welfare, perception of animal production systems and biodiversity. Functions like goat meat production, production of plant fibre, fire control and avalanche control are less appreciated. All stakeholder groups generally agreed on the importance of the functions evaluated. Keywords: grassland, stakeholders, Ireland Introduction Ireland has a grassland area of approximately 4.6 million ha of grassland (consisting of pasture, grass silage or hay, and rough grazing), which is 90% of the total utilizable agricultural area (O’Mara, 2008). Ruminant production systems (milk and meat) in Ireland are predominantly grass based and Ireland has a long grazing season (February/March to October/November, depending on soil type and rainfall). Grazed grass is the main feed source during the grazing season, and winter forage consists predominantly of grass silage. Grassland in Ireland is predominately permanent pasture and the proportion of agricultural land reseeded (from grass to grass) annually is low at approximately 2% (Shalloo et al., 2011). Perennial ryegrass is the most widely sown grass species. The aim of the current study was to get an insight into the importance of grasslands for stakeholders in Ireland. Materials and methods This study provides results for Irish stakeholders obtained via an on-line questionnaire on functions of grasslands. The questionnaire was distributed throughout Europe. A detailed description of the method can be found in Van den Pol-van Dasselaar et al. (2014, this volume). This paper describes the results for Ireland. In Ireland the questionnaire was distributed to members of the Irish Grassland Association, Teagasc (researchers, students, technicians, advisers), Universities, members of farming organizations, members of government departments and farmers. Results and discussion Two hundred and thirty-two valid responses were obtained. The majority of these responses came from advisers (67 responses; 29% of total), followed by researchers (66; 28%), farmers (37; 16%), students (26; 11%), education (23; 10%), policy makers (8; 3%), industry (3; 1%) and NGOs (2; <1%). Tables 1 to 4 show the appreciation of the functions of grasslands by the different stakeholders. The functions are grouped into the four groups of ecosystem services: provisioning services, regulating services, supporting services and cultural services. The provisioning services (Table 1) that Irish stakeholders consider to be most important include milk and meat (beef and sheep) production, the production of high quality forage, and nutritional quality of animal products for human consumption. Biodiversity and conservation Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 810 of quality of ecosystems are considered to be important regulating services (Table 2), and NGOs and policy makers consider these to be more important than do the other stakeholder categories. Fire and avalanche control are not very important in Ireland. Positive perception of animal production systems, beauty of the landscape, and maintaining population in rural areas are the most important cultural services of grassland (Table 3). Grazing, competitiveness of farming systems, and animal health and welfare are considered the most important supporting services (Table 4). In general, across the functions evaluated, there was good agreement between all stakeholder groups. Table 1. Importance of provisioning services of grasslands according to the respondents (by stakeholder category) of the questionnaire (1 = not important; 5 = very important). Advice Education Farmers High quality forage Dairy cow milk production 4.6 4.7 4.5 4.5 4.6 4.5 Low cost animal feed 4.6 4.4 Nutritional quality of animal products for human consumption Beef meat production 4.3 Industry NGO Policy maker Research Students 4.7 4.3 4.5 5.0 4.1 4.4 4.5 4.5 4.6 4.3 4.7 5.0 4.5 4.4 4.5 4.5 4.2 4.6 4.7 5.0 4.0 4.2 4.3 4.6 4.2 4.1 4.3 4.5 4.4 4.5 4.5 Global food production 4.2 Region of origin of animal products 3.5 4.0 4.1 4.0 4.5 4.1 4.1 4.2 3.5 3.8 4.3 3.5 3.4 3.3 Honey production 3.7 1.8 2.6 2.0 1.0 3.5 2.1 2.1 2.1 Sheep meat production 3.6 3.2 2.9 2.0 4.0 3.9 3.3 3.5 Biomass for energy production 1.8 2.2 2.0 1.7 2.5 1.6 2.2 2.7 Sheep milk production 1.8 1.7 1.7 2.0 3.5 2.1 1.8 2.1 Goat milk production 1.6 1.5 1.5 1.0 3.5 1.5 1.5 1.8 Wool production 2.3 2.3 2.1 1.3 3.5 2.4 2.0 2.5 Goat meat production 1.5 1.7 1.5 1.0 1.5 1.3 1.3 1.8 Production of plant fibre 1.6 1.8 1.7 1.0 3.5 1.4 1.6 2.3 Table 2. Importance of regulating services of grasslands according to the respondents (by stakeholder category) of the questionnaire (1 = not important; 5 = very important). Advice Biodiversity Conservation of ecosystems quality Water catchment 3.4 3.2 3.4 Education 3.9 3.9 3.8 Farmers 3.0 3.2 3.3 Industry 2.7 3.0 2.0 NGO 5.0 4.5 4.0 Policy maker 4.5 4.4 4.1 Erosion control 2.3 2.9 2.8 1.3 2.0 Carbon sequestration 3.0 3.1 2.8 2.3 4.0 Mitigating greenhouse gas emissions 3.3 3.1 2.7 2.3 Adaptation to climate change 3.1 3.2 3.4 3.0 Flood plains rivers 2.6 3.1 2.9 Pathogen control in cropping system Fire control 2.4 1.7 2.7 1.9 3.1 2.5 Avalanche control 1.1 1.5 1.6 Research 3.3 3.5 3.2 Students 3.4 3.4 3.4 3.3 2.8 2.7 3.9 3.3 3.3 4.0 4.1 3.2 3.2 3.0 3.6 3.1 3.5 1.3 3.5 3.5 2.8 2.9 2.3 1.0 4.0 1.5 3.0 1.6 2.5 1.7 2.8 2.2 1.0 1.0 1.0 1.2 2.0 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 811 Table 3. Importance of cultural services of grasslands according to the respondents (by stakeholder category) of the questionnaire (1 = not important; 5 = very important). Advice Education Farmers Industry NGO Policy maker 3.7 4.4 4.0 4.1 3.8 4.2 3.3 4.0 4.5 4.0 Rural development Maintaining population in rural areas 3.1 3.3 3.8 3.7 3.7 3.9 4.0 3.7 Cultural values 2.6 3.7 3.1 Tourism / recreation 2.8 3.5 2.7 Supporting horses for equestrian sport and recreation 2.4 2.5 2.4 Beauty of the landscape Positive perception production systems of animal Research Students 4.3 4.4 3.7 4.1 3.6 3.9 3.5 4.0 3.4 3.7 2.5 4.1 3.5 3.8 3.0 4.0 3.6 3.0 3.2 2.7 3.0 4.0 3.4 3.1 3.3 3.5 3.4 2.8 2.5 Table 4. Importance of supporting services of grasslands according to the respondents (by stakeholder category) of the questionnaire (1 = not important; 5 = very important). Advice Education Farmers Industry NGO Grazing 4.9 4.5 4.7 5.0 5.0 Animal health 4.3 4.3 4.6 5.0 4.5 Animal welfare 4.1 4.1 4.5 5.0 Conservation of soil structure and fertility in cropping systems 3.8 3.6 3.9 4.0 Feed protein supply at farm level 3.8 3.7 4.2 Competitiveness of farming systems 4.6 3.6 N fixation via legumes 2.9 3.4 Availability of water 2.8 Crop pollination 2.5 Policy maker Research Students 4.8 4.6 4.7 4.0 4.0 4.1 3.5 4.1 3.9 4.0 4.5 4.0 3.4 4.0 4.0 4.5 3.5 3.7 4.2 4.4 5.0 5.0 4.5 4.2 4.2 3.3 2.3 2.0 3.5 3.2 3.6 3.3 3.5 1.7 4.0 3.1 3.0 3.8 3.0 3.2 2.7 4.0 3.3 2.9 3.2 Conclusion A wide range of stakeholders responded to the survey in Ireland. Stakeholders consider grassland to be important for a range of functions and services including bovine milk production, beef and sheep meat production, forage production, animal health and welfare, perception of animal production systems and biodiversity. The least important functions considered by the stakeholders include goat meat production, production of plant fibre, fire control and avalanche control. All stakeholder groups generally agreed on the importance of the functions evaluated. Acknowledgements The research leading to these results has received funding from the European Community's Seventh Framework Programme (FP7/ 2007-2013) under the grant agreement n° FP7-244983 (Multisward) and from the Irish Farmers Dairy Levy Fund. References O'Mara F. (2008) Country Pasture/Forage Resource Profile - Ireland. http://www.fao.org/ag/AGP/AGPC/doc/Counprof/Ireland/Ireland.htm#climate Shalloo L., Creighton P. and O’Donovan M. (2011) The economics of reseeding on a dairy farm. Irish Journal of Agriculture and Food Research 50, 113-122. Van den Pol-van Dasselaar A., Goliński P., Hennessy D., Huyghe C., Parente G. and Peyraud J.L. (2014) Appreciation of the functions of grasslands by European stakeholders. Grassland Science in Europe 19 (these Proceedings). Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 812 Appreciation of the functions of grassland by Italian stakeholders Parente G.1 and Van den Pol-van Dasselaar A.2 1 University of Udine, Department of Agriculture and Environmental Science, Via delle Scienze 91, 33100 Udine, Italy 2 Wageningen UR Livestock Research, P.O. Box 65, 8200 AB Lelystad, the Netherlands. Corresponding author: agnes.vandenpol@wur.nl Abstract The European project MultiSward studied the appreciation of different functions of grasslands by European stakeholders. This paper describes the importance of grasslands for stakeholders in Italy. Italy currently has approximately 6.0 million ha of grassland, consisting of permanent grassland, pastures and temporary grassland, which is 47% of the total agriculturally utilized area (12.9 million ha). Italian stakeholders considered grassland to be important for a range of functions and services but they especially appreciate conservation of the quality of ecosystems and biodiversity such as beauty of the landscape and cultural values. Important functions like carbon sequestration and mitigating greenhouse gas emissions are less known and appreciated. All stakeholder groups generally agreed on the importance of the functions evaluated. Keywords: grassland, stakeholders, Italy Introduction Italy has a grassland area of approximately 6.0 million ha of grassland (consisting of permanent grassland, pasture and temporary grassland), which is 47% of the total utilizable agricultural area (12.9 million ha). Ruminant production systems are predominantly grass-based in the mountain areas and in southern Italy, and based on maize silage on the plains of northern Italy (e.g. Po valley). Cattle (5.7 million), sheep (6.6 million) and goats (about 0.8 million) are the most important types of animals reared. The most important grassland species cultivated are sainfoin in southern Italy and lucerne in central and northern Italy (INEA, 2012; ISTAT, 2010). The aim of the current study was to obtain an insight into the appreciation of the functions of grasslands by stakeholders in Italy. Materials and methods This study provides results that were obtained via an on-line questionnaire on the functions of grasslands and which was spread throughout Europe. A detailed description of the method can be found in Van den Pol-van Dasselaar et al. (2014, this volume). This paper describes the results for Italy. In Italy, the questionnaire was distributed and collected during national and international meetings held in different regions of Italy, to the Alpine Zootechnical Society (SooZooAlp), University of Udine (researchers, students, technicians, advisers), Universities, members of farming organizations, members of government departments and farmers. In addition, it was spread via social media. The majority of the answers to the questionnaires have been collected during the meetings and sent to the elaboration centre. Results and discussion At the time of closing the questionnaire, 241 valid responses were obtained from Italy. The majority of the 241 valid responses came from students (117 responses, which equals 49% of total responses), followed by NGO (30; 12%), education (24; 10%), policy makers (22; 9%), researchers (20; 8%), advice (14; 6%), farmers (11; 5%), and industry (3; 1%). Tables 1 to 4 show the appreciation of the functions of grasslands by the different stakeholders. The functions are grouped into the four groups of ecosystem services: provisioning services, regulating services, supporting services and cultural services. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 813 Table 1. Importance of provisioning services of grasslands according to the respondents of the questionnaire (1 = not important; 5 = very important). Advice Education Farmers Industry NGO Policy maker Research Students High quality forage 4.1 4.0 3.7 3.7 3.8 4.3 3.8 3.5 Dairy cow milk production 3.3 3.4 3.3 2.7 3.5 3.4 3.5 3.6 Low cost animal feed 3.6 3.1 2.9 2.0 2.7 4.1 3.5 3.0 Nutritional quality of animal products for human consumption Beef meat production 3.6 3.9 3.5 2.7 3.7 4.3 3.5 3.7 2.8 2.8 3.6 3.7 3.0 3.2 3.0 3.7 Global food production 2.7 3.1 3.4 3.0 2.8 3.3 2.8 3.4 Region of origin of animal products 3.5 3.7 3.5 3.0 3.5 3.9 3.8 3.8 Honey production 3.5 3.6 3.2 3.7 3.8 3.7 3.4 3.5 Sheep meat production 2.1 2.5 2.8 2.7 2.3 3.1 2.9 3.0 Biomass for energy production 2.4 2.2 2.7 3.0 2.1 2.1 2.0 3.2 Sheep milk production 1.9 2.5 2.3 2.7 2.4 3.7 2.9 2.8 Goat milk production 2.3 2.1 1.9 2.7 2.6 3.1 2.7 2.6 Wool production 1.5 2.0 1.8 2.0 2.0 2.2 2.2 2.5 Goat meat production 2.4 2.0 2.1 2.7 2.3 2.5 2.6 2.7 Production of plant fibre 2.4 2.1 2.2 2.0 1.8 1.7 2.0 2.9 Table 2. Importance of regulating services of grasslands according to the respondents of the questionnaire (1 = not important; 5 = very important). Advice Education Farmers Industry NGO Policy maker Research Students Biodiversity 4.6 4.7 3.7 3.0 4.5 4.7 4.1 3.9 Conservation of ecosystems quality Water catchment 4.7 4.7 4.1 3.0 4.3 3.7 3.3 3.3 4.5 4.4 4.0 4.0 3.5 3.6 3.3 3.4 Erosion control 4.1 3.9 3.3 Carbon sequestration 3.6 3.6 2.9 4.3 3.7 4.2 3.6 3.3 3.0 3.3 3.3 3.2 3.0 Mitigating greenhouse gas emissions 3.2 3.5 Adaptation to climate change 3.4 3.8 3.1 2.3 3.1 3.5 3.0 3.2 3.3 2.3 3.4 3.3 2.7 3.4 Flood plains rivers 3.9 Pathogen control in cropping system 3.3 3.7 3.5 3.3 3.2 3.0 2.9 3.5 3.1 2.7 2.7 3.2 3.1 2.5 3.4 Fire control Avalanche control 3.4 3.7 3.4 3.3 3.2 3.4 3.1 3.4 3.4 3.0 2.7 2.0 2.9 3.0 2.2 3.2 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 814 Table 3. Importance of cultural services of grasslands according to the respondents of the questionnaire (1 = not important; 5 = very important). Advice Education Farmers 4.6 3.7 4.8 3.4 4.3 2.8 4.0 3.3 4.4 3.4 Rural development 3.9 3.6 4.0 3.0 Maintaining population in rural areas 3.9 3.4 4.1 Cultural values 4.0 4.3 4.3 Tourism / recreation 3.8 4.0 Supporting horses for equestrian sport and recreation 1.8 1.9 Beauty of the landscape Positive perception of production systems animal Industry NGO Policy maker Research Students 4.5 4.0 4.0 3.7 4.0 3.4 3.6 3.9 3.4 3.4 2.7 3.4 4.2 3.2 3.3 2.7 4.0 4.4 3.7 3.7 3.6 3.3 4.0 4.0 3.4 3.6 1.8 2.5 2.1 2.6 1.9 2.3 Table 4. Importance of supporting services of grasslands according to the respondents of the questionnaire (1 = not important; 5 = very important). Advice Education Farmers Industry NGO Policy maker Research Students Grazing 3.6 3.9 3.2 3.7 3.7 4.6 4.2 3.7 Animal health 3.9 4.2 4.0 3.0 3.8 3.8 3.5 3.8 Animal welfare 3.6 4.1 3.7 3.3 4.0 3.8 3.7 3.7 Conservation of soil structure and fertility in cropping systems 4.0 3.9 3.7 3.7 3.9 4.1 3.6 3.6 Feed protein supply at farm level 2.8 2.6 2.7 2.0 2.4 3.5 2.6 2.9 Competitiveness of farming systems 3.8 2.9 3.3 2.7 3.2 3.7 2.9 3.2 N fixation via legumes 3.3 3.5 3.1 3.0 2.9 3.5 3.0 3.2 Availability of water 3.4 3.6 4.0 3.0 3.3 4.0 2.9 3.7 Crop pollination 3.6 4.0 3.3 2.7 3.3 3.5 2.9 3.6 Conclusion A wide range of stakeholders responded to the survey in Italy. Italian stakeholders considered grasslands to be important for a range of functions and services but especially appreciate conservation of the quality of ecosystems and biodiversity such as beauty of the landscape and cultural values. Important functions like carbon sequestration and mitigating greenhouse gas emissions are less known and appreciated. All stakeholder groups generally agreed on the importance of the functions evaluated. Acknowledgements The research leading to these results has received funding from the European Community's 7th Framework Programme (FP7/ 2007-2013) under the grant agreement n° FP7-244983 (Multisward). References INEA (Ministero delle Politiche Agricole, Alimentari e Forestali) (2012) Rapporto sullo stato dell’agricoltura. ISTAT (2010) www.istat.it/censimento-agricoltura/agricoltura-2010. Van den Pol-van Dasselaar A., Goliński P., Hennessy D., Huyghe C., Parente G. and Peyraud J.L. (2014) Appreciation of the functions of grasslands by European stakeholders. Grassland Science in Europe 19 (this volume). Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 815 Appreciation of the functions of grassland by Polish stakeholders Goliński P.1, Van den Pol-van Dasselaar A.2 and Golińska B.1 1 Department of Grassland and Natural Landscape Sciences, Poznan University of Life Sciences, Dojazd 11, 60-632 Poznan, Poland 2 Wageningen UR Livestock Research, P.O. Box 65, 8200 AB Lelystad, the Netherlands Corresponding author: agnes.vandenpol@wur.nl Abstract The European project MultiSward studied the appreciation of different functions of grasslands by European stakeholders. This paper describes the importance of grasslands for stakeholders in Poland. Poland currently has 3.29 million ha of grassland, which is 21.3% of the total agriculturally utilized area. Polish stakeholders consider grassland to be important for a range of functions and services including animal health, dairy cow milk production, nutritional quality of animal products for human consumption, low cost animal feed, beauty of the landscape and biodiversity. Functions like goat milk and meat production, wool production, production of plant fibre and avalanche control are less appreciated. Keywords: grassland, stakeholders, Poland Introduction Grasslands in Poland occupy a total area of 3.29 million hectares, which constitutes 21.3% of the total UAA or 12.6% of the entire area of the country. However, this area does not include leys established on arable land (temporary grasslands) covering 0.45 million ha, whose duration of utilization does not exceed 4-5 years (Goliński, 2014). Meadows represent 77% and pastures about 23% of the grassland area. The share of permanent pasture in Poland has decreased during the last 20 years by more than 50%. The number of cattle and sheep has also decreased. Several regional programmes have been initiated to stimulate economic development and preservation of cultural heritage. The peculiar characteristics of Polish grassland are their persistency, various conditions of their habitats, high floristic diversity, multifunctionality expressed in the predominance of mowing over grazing, moderate and low intensity of use, and also very important roles in the natural environment, culture and landscape (Warda and Kozłowski, 2012). The aim of the current study was to obtain an insight into the importance of grasslands for stakeholders in Poland. Materials and methods This study provides results for Polish stakeholders, which were obtained via an on-line questionnaire on the functions of grasslands that was spread throughout Europe. A detailed description of the method can be found in Van den Pol-van Dasselaar et al. (2014, this volume). In Poland, the questionnaire was distributed to members of the Polish Grassland Society (researchers, students, technicians, advisors), universities and institutions related to grassland, farming organizations, government departments, NGOs and farmers. Results and discussion At the time of closing the questionnaire, 204 valid responses had been obtained. The majority of these responses came from students (110 responses, which equals 54%), followed by researchers (42; 21%), farmers (20; 10%), advisers (15; 7%), education (9; 4%), industry (5; 2%) and NGOs (3; 2%). Policy makers were also included but no responses were received. Tables 1 to 4 show the appreciation of the functions of grasslands by the different stakeholders. The functions are grouped into the four groups of ecosystem services: provisioning services, regulating services, supporting services and cultural services. The provisioning services that Polish stakeholders consider to be most important include dairy cow milk production, Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 816 nutritional quality of animal products for human consumption and low cost animal feed. Biodiversity and conservation of the quality of ecosystems are considered to be important regulating services. Beauty of the landscape is the most important cultural service of grassland. Animal health and welfare are considered the most important supporting services. In general, across the functions evaluated, there was good agreement between all stakeholder groups. Table 1. Importance of provisioning services of grasslands according to the respondents of the questionnaire (1 = not important; 5 = very important). Advice Education Farmers Industry NGO Research Students High quality forage 4.1 3.9 4.2 4.4 3.0 4.0 3.3 Dairy cow milk production 4.5 3.9 4.2 4.2 4.0 3.8 3.5 Low cost animal feed 4.3 4.0 4.2 3.8 3.7 4.0 3.4 Nutritional quality of animal products for human consumption Beef meat production 4.1 4.0 3.6 4.6 4.0 3.9 3.8 3.5 3.3 3.9 2.8 2.7 3.2 2.9 Global food production 3.4 3.8 3.2 3.4 2.7 3.3 2.7 Region of origin of animal products 2.8 3.9 2.6 2.8 1.7 3.5 3.1 Honey production 3.1 2.8 3.4 2.6 2.7 3.0 3.1 Sheep meat production 2.0 2.3 1.4 1.6 2.0 2.2 2.0 Biomass for energy production 2.5 3.2 2.3 2.6 3.0 2.2 2.7 Sheep milk production 2.0 2.3 1.5 1.2 2.0 2.3 2.1 Goat milk production 1.8 2.4 1.6 1.2 2.0 1.9 1.9 Wool production 1.7 2.4 1.7 1.0 2.0 1.8 2.1 Goat meat production 1.8 1.9 1.5 1.4 2.0 1.7 1.9 Production of plant fibre 1.7 2.3 1.6 1.4 1.3 1.6 2.0 Table 2. Importance of regulating services of grasslands according to the respondents of the questionnaire (1 = not important; 5 = very important). Advice Education Farmers Industry NGO Research Students Biodiversity 3.6 4.0 4.2 3.4 3.7 4.4 3.9 Conservation of ecosystems quality 3.6 4.0 3.8 3.4 3.3 3.9 3.5 Water catchment 3.4 4.0 3.8 3.8 4.0 3.9 3.5 Erosion control 3.5 4.0 4.0 3.6 3.7 3.9 3.5 Carbon sequestration 2.5 3.1 2.7 2.2 2.3 3.3 2.7 Mitigating greenhouse gas emissions 3.2 3.7 3.4 3.8 3.3 2.9 3.2 Adaptation to climate change 2.8 4.2 3.0 3.6 3.7 3.2 3.2 Flood plains rivers 3.1 3.9 2.4 2.0 2.7 3.5 3.0 Pathogen control in cropping system 2.7 3.2 3.4 3.4 2.3 2.4 3.2 Fire control 2.7 3.3 2.7 2.4 2.0 2.4 3.3 Avalanche control 1.9 2.8 1.1 1.0 1.3 1.6 2.1 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 817 Table 3. Importance of cultural services of grasslands according to the respondents of the questionnaire (1 = not important; 5 = very important). Advice Education Farmers Industry NGO Research Students Beauty of the landscape 4.1 4.1 4.2 3.6 3.3 4.3 4.2 Positive perception of animal production systems 3.8 3.3 3.7 3.2 2.7 3.6 3.3 Rural development 3.1 3.9 3.3 3.6 2.3 3.2 3.5 Maintaining population in rural areas 2.7 3.8 2.6 2.4 1.7 2.6 3.0 Cultural values 2.5 3.9 3.0 2.8 2.7 3.1 3.6 Tourism / recreation 3.2 3.8 3.1 2.6 3.3 3.7 3.5 Supporting horses for equestrian sport and recreation 2.6 2.9 2.4 1.6 2.7 2.6 2.6 Table 4. Importance of supporting services of grasslands according to the respondents of the questionnaire (1 = not important; 5 = very important). Advice Education Farmers Industry NGO Research Students Grazing 3.8 3.8 3.2 3.6 3.3 3.8 3.3 Animal health 4.3 4.0 4.3 4.6 3.7 4.0 3.8 Animal welfare 4.1 3.4 4.0 3.8 4.0 4.0 3.7 Conservation of soil structure and fertility in cropping systems 3.7 4.0 3.5 3.4 2.3 3.8 3.7 Feed protein supply at farm level 4.4 4.0 4.0 4.0 2.3 3.6 3.3 Competitiveness of farming systems 3.1 3.0 3.2 3.6 3.0 3.0 2.9 N fixation via legumes 3.7 3.6 3.9 3.8 3.0 3.2 3.3 Availability of water 3.7 4.0 3.4 3.6 3.3 3.6 3.8 Crop pollination 3.1 3.0 3.3 2.8 3.0 3.2 3.2 Conclusion Polish stakeholders consider grassland to be important for a range of functions and services including animal health, dairy cow milk production, nutritional quality of animal products for human consumption, low cost animal feed, beauty of the landscape and biodiversity. The least important functions considered by the stakeholders include goat milk and meat production, wool production, production of plant fibre and avalanche control. All stakeholder groups generally agreed on the importance of the functions evaluated. Acknowledgements The research leading to these results has received funding from the European Community's Seventh Framework Programme (FP7/ 2007-2013) under the grant agreement n° FP7-244983 (Multisward) and from Polish Ministry of Science and Higher Education (2162/7PR/2011/2). References Goliński P. (2014) Grassland functions in ecosystems. In: Grzebisz W., Goliński P. and Potarzycki J. (eds) Fertilization of grasslands. Powszechne Wydawnictwo Rolnicze i Leśne, Warszawa, 9-36. (in Polish) Van den Pol-van Dasselaar A., Goliński P., Hennessy D., Huyghe C., Parente G. and Peyraud J.L. (2014) Appreciation of the functions of grasslands by European stakeholders. Grassland Science in Europe 19 (this volume). Warda M. and Kozłowski S (2012) Grassland – a Polish resource. Grassland Science in Europe 17, 3-16. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 818 Theme 6 ‘Approaches to forage crop improvement’ Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 819 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 820 Theme 6 submitted papers Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 821 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 822 Genomic characterization of survivor populations of red clover by GBS Ergon Å. and Rognli O.A. Norwegian University of Life Sciences, Dept. of Plant Science, P.O. Box 5003, N-1430 Ås, Norway Corresponding author: ashild.ergon@nmbu.no Abstract Survivor populations of red clover from different plots in a field experiment in southern Norway were genetically characterized using genotyping by sequencing (GBS), and compared with the original population and each other. Based on allele frequencies of single nucleotide polymorphisms (SNPs), pairwise genetic distances between populations were calculated for each SNP and across all SNPs as FST-values and Nei’s genetic distance, respectively. Across all SNPs, survivor populations were more different from each other than from the original population, indicating random selection or selection due to local variation in conditions within the site. Fifty-six SNPs were found to have been under selection in all four populations (FDR=0.016). These are candidate loci for persistence of red clover. Keywords: Trifolium pratense, genetic shifts, persistence, population genetics Introduction Red clover (Trifolium pratense L.) is a perennial outbreeding species and thus there is a considerable amount of genetic variation within cultivars. In a field, only a fraction of the seeded plants will survive the first few years due to competition and stress. Surviving plants are presumably genotypes that are best adapted to the local conditions, and may be utilized to identify traits and alleles useful for breeding. The aim of this study was to find out if it is possible to detect loci under selection within one generation of red clover using a GBS approach. Materials and methods The diploid red clover cultivar ‘Lea’ (Graminor, Norway) was included in a split-plot field experiment with pure stands and mixtures of grass and legume species (plot size 7.5 m2), sown in two replicates at Ås, Norway in June 2010. The plots were harvested either 3 or 5 times a year (3H and 5H) and leaves were sampled once or twice a year. In order to study changes in genetic composition, the original population and survivor populations were genotyped using GBS technology. DNA was extracted from leaves of 48 individual survivor plants sampled in October 2012 from both harvesting regimes in two replicate blocks (four survivor populations). In addition, DNA was extracted from 88 individuals of the original population seeded in the greenhouse. GBS library preparation and sequencing, as well as SNP calling, was done by the Institute for Genomic Diversity, Cornell University, according to Elshire et al. (2011). The enzyme ApeK1 was used for digestion of genomic DNA, and the GBS UNEAK analysis pipeline, an extension to the Java program TASSEL (Bradbury et al., 2007), was used to call SNPs from the sequenced GBS libraries. After removing SNPs with missingness >0.2 in one or more of the five populations, or minor allele frequency <0.05 in both the original population and at least one of the four survivor populations, 9203 SNPs remained for analysis. For all these SNPs allele frequencies were calculated for the original population and each of the four survivor populations separately. In order to study the overall differentiation between populations, the average Nei’s genetic distance, D (Nei 1972), across all SNPs was calculated for each population pair (Table 1). In order to identify SNP loci that had been under selection, pairwise FST-values (original population vs. each of the four survivor populations) were first calculated 𝑞 2 −𝑞 2 for each SNP as 𝑞(1−𝑞). Secondly, a chi-square test was used to identify SNPs with significant Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 823 FSTs at P<0.1, P<0.05 and P<0.01 levels, using the test statistic x2 = 2NFST, where 2N=the sum of genotyped gametes in the two populations (Hedrick 2011). Thirdly, SNPs with significant FST either in all four survivor populations, in both replicate 3H survivor populations or in both replicate 5H survivor populations, all relative to the original population, were identified. Corresponding estimates of the false discovery rate (FDR) were calculated for each P-level as 𝑙∗𝑃 𝑛 , where l=number of SNP loci tested (9203), n=the number of survivor populations tested against the original population (4), and d=the number of SNP loci identified as significant in all n population pairs. Pairwise FSTs between the 3H and 5H survivor populations within both replicates were also calculated and tested, but this did not result in any SNPs detected with an acceptable FDR. The SNP-containing sequences were blasted against sequences available in the Legume Information System (http://www.comparative-legumes.org/). 𝑑 Results and discussion Nei’s genetic distance between the original population and the survivor populations were significantly smaller than the distances between survivor populations (Table 1). Table 1. Nei’s genetic distance (D) between populations. D values were first averaged across 9203 SNPs for each population pair and then across 2-6 population pairs. Compared populations (number of pairwise comparisons) Original population and all survivor populations (4) Original population against both 3H survivor populations (2) Original population against both 5H survivor populations (2) All survivor populations against each other (6) 3H and 5H survivor populations against each other (4) Rep 1 and rep 2 survivor populations against each other (4) Average pairwise D ± S.E. 0.00351 ± 0.00014 0.00363 ± 0.00024 0.00339 ± 0.00019 0.00433 ± 0.00010 0.00443 ± 0.00009 0.00435 ± 0.00014 Also, the distance between 3H and 5H populations were no larger than the distance obtained when populations were grouped according to replicate block instead of harvesting regime. Together, this indicates that random selection or diversifying selection due to variation in local conditions in the field other than harvesting regime has occurred. If the original population has high genetic diversity and low linkage disequilibrium (typical of forage cultivars), it cannot be expected that selection acting on a relatively limited number of loci will affect average genetic distance measured across the genome. In order to identify such selection, genetic distance of individual loci must be considered. Indeed, some loci had been under selection in all four plots: 56 SNPs had significantly altered allele frequencies, measured as FST, in all four survivor populations relative to the original population (P<0.1 all survivor populations, corresponding FDR=0.016). Twenty of these were also significant at P<0.05 in all survivor populations, with a corresponding FDR<0.003. The use of several survivor populations makes it possible to use a less stringent significance level in the first screening of SNPs in individual populations. In addition, effects of random selection are removed. These 56 SNPs are likely selected by conditions that are common to all four plots and can be related to, e.g. establishment, competition, winter survival and the general environmental and management conditions. They may be associated with genes conferring improved survival through genetic linkage or population structure. Blastn hits were obtained for 13 of the 56 SNP sequences. Hundred SNPs had significant FSTs (P<0.01, FDR=0.028-0.046) in at least two survivor populations. The number of SNPs with significant FSTs in both 3H populations or in both 5H populations were not larger than the number of SNPs with significant FSTs when the survivor populations were grouped according to replicate blocks (data not shown). Thus, no loci under specific selection in response to harvesting regime could be detected. This may have been possible with a higher number of replicate populations, a higher number of sampled survivors per population, or more SNP loci and higher sequencing depth. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 824 Conclusions In conclusion, it was possible to identify loci that had been under natural selection within one generation in a red clover variety grown in a field experiment for two and a half years. This may be utilized in further identification of genomic regions, genes and alleles conferring persistence in red clover under various field conditions, which again can be utilized in breeding. In this experiment no specific selection in response to harvesting frequency could be detected this may be a matter of limiting sensitivity of the experimental setup. Acknowledgements This work has received funding from the European Community's Seventh Framework Programme (FP7/ 2007-2013) under grant agreement number FP7-244983 (Multisward). References Elshire R.J., Glaubitz J.C., Sun Q., Poland J.A., Kawamoto K., et al. (2011) A robust simple genotyping-bysequencing (GBS) approach for high diversity species. PLoS ONE 6(5): e19379.doi:10.1371/journal.pone.0019379. Bradbury P.J., Zhang Z., Kroon D.E., Casstevens T.M., Ramdoss Y. and Buckler E.S. (2007) TASSEL: Software for association mapping of complex traits in diverse samples. Bioinformatics 23: 2633–2635. Hedrick PW. (2011) Genetics of populations, 4th ed., Sudbury, USA: Jones and Bartlett Publishers, 675 pp. Nei M. (1972) Genetic distance between populations. American Naturalist 106, 283-292. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 825 Towards genomic selection in perennial ryegrass genetic improvement Skøt L.1, Grinberg N.F. 2, Lovatt A.1, Hegarty M.1, Macfarlane A.1, Blackmore T.1, Armstead I.1, King R.D.2 and Powell W.A.1 1 Institute of Biological, Environmental and Rural Sciences, Aberystwyth University, Gogerddan, Aberystwyth, SY23 3EE, United Kingdom, 2 Manchester Institute of Biotechnology, The University of Manchester, 131 Princess St, Manchester, M1 7DN, United Kingdom Corresponding author: lfs@aber.ac.uk Abstract Genomic selection (GS) for crop improvement makes use of genome-wide molecular marker information. GS has proven its value in animal breeding programmes, but its impact in plant breeding is just emerging. The objective of GS is to use genotype data to estimate and predict the breeding value of selection candidates. This can increase the speed of the breeding cycle, reduce the cost and effort of phenotyping, and achieve faster selection of candidates for crossing programmes and production of varieties. Such genomically estimated breeding values (GEBV) are based on models developed in a training population for which both phenotypic and genotypic data are available. The usefulness of GS is determined by the accuracy of GEBV. The size of the training population and the density of molecular marker coverage are key constraints for prediction accuracy. The perennial ryegrass recurrent selection breeding programme at IBERS is based on a relatively small founder population, and lends itself well to a GS approach. We will describe the integration of GS into the existing breeding programme, and present our initial results of prediction accuracy for various traits. Keywords: Lolium perenne, genomic selection, phenotype, genomically estimated breeding value. Introduction Perennial ryegrass (Lolium perenne L.) is one of the most important forage crops in temperate European livestock agriculture. Its genetic improvement is based on recurrent selection in populations or families with the aim of increasing the frequency of beneficial alleles or genes over successive generations. Recent developments in genomics technology have made genetic marker development much more cost effective. Genome-wide markers now has the potential to assist the breeding programmes by providing accurate information of estimated breeding values faster than phenotypic information. Genomic selection (GS) was first proposed by Meuwissen et al. (2001), and is now widely used in animal breeding programmes (Hayes et al., 2013). The potential of GS in plant breeding has now also been recognised (Heffner et al., 2009). The advantage of GS compared to traditional marker-assisted selection is that the effects of all markers are used simultaneously, irrespective of whether they are significant or not. Implementation of GS requires a reference or training population for which both phenotypic data and information of evenly distributed molecular markers are available. This information is used to generate prediction models to estimate the effects of each marker, and used to estimate the breeding value for plant material in test populations with genotypic data, but for which no phenotypic information is available. This is potentially advantageous for difficult or expensive to measure traits, and in perennial crops with long generation cycles. We are using the IBERS forage ryegrass breeding populations as templates for implementing GS. Here we describe the important features of this project, provide preliminary data on prediction models and their accuracy, and discuss future improvements. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 826 Materials and methods Phenotypic data for a range of traits were initially obtained from plot trials performed on halfsib progeny of mother plants from the breeding programme. Two breeding populations were used: the 13th generation of the intermediate flowering population, and the 5th generation of the late flowering population. This provided the first training population, and had a combined size of 159. An Illumina Infinium SNP-CHIP with single nucleotide polymorphisms from 2765 genetic loci was used to obtain the genotypic data from this population. In addition 100 motherplants from the 14th intermediate generation were also genotyped. They constituted the test population. We used a genomic best linear unbiased prediction model (GBLUP) to provide us with marker effects and GEBVs from the training population. Due to a relatively small size of the two populations, model fit was assessed by leave-one-out cross validation, in which prediction for each sample point was obtained by using a model fitted to the remainder of the data excluding that point. Accuracies were estimated by calculating the mean squared error of the cross-validated predictions and dividing it by the sample variance. This way one can compare performance of the BLUP model to that of the ‘base’ model that predicts every point with the sample mean. Results and discussion Figure 1. Principle component analysis of motherplants from two generations of the intermediate flowering and the late flowering populations. The analysis is based on marker data information. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 827 The late and intermediate flowering populations differ both at the genotypic and phenotypic levels. Principal component analysis separates Intermediate (F13 and F14) and Late (F5) genotypes into two clearly defined clusters (Figure 1). This suggests that the F13 data would be most useful in predicting the F14 breeding values. The traits possess large variability across years as well as across the two populations: some traits, like dry matter digestibility (DMD), are more or less homogenous across the two subpopulations, while most, like yield, vary considerably. Additionally some traits are highly correlated, e.g. DMD and water soluble carbohydrate (WSC) and yields in the two consecutive years. We analysed the two populations separately and combining the two into a larger pooled sample. Results for GBLUP are summarized in Table 1. The BLUP model performs rather poorly for some of the traits with no difference from the ‘base’ model and with little, if any, accuracy gain by using the pooled populations, but explains some of the variance in the others. Notably, it is the forage quality traits which are more homogenous across the two sub-populations that benefit from using the pooled sample. Table 1. Broad sense heritability (H2) and prediction accuracies for traits measured in the perennial ryegrass breeding populations. A low prediction value indicates a good model performance, a value close to 1 indicates no difference from ‘base’ model. gcscore – ground cover, DMD – dry matter digestibility, WSC – water soluble carbohydrates, N – nitrogen, DMyield2 – yield of 2nd cut, vegyld – yield of the other cuts, NDF_dig – neutral detergent fibre digestibility. Test Train gcscore_yr3 yield_yr1 yield_yr2 DMD WSC N DMyield2_yr1 DMyield2_yr2 vegyld_yr1 vegyld_yr2 NDF_dig Int F13 H2 ± SD 0.51±0.97 0.71±0.39 0.67±0.37 0.82±0.10 0.73±0.21 0.67±0.33 0.00±0.00 0.00±0.00 0.76±0.20 0.67±0.33 0.55±2.01 Int F13 – Prediction Int F13 Pooled 1.03 1.05 1.01 0.99 0.80 0.70 0.69 0.63 0.79 0.71 0.94 0.74 0.99 1.02 1.04 0.91 0.96 0.90 0.82 0.75 0.75 0.66 Late F5 H2 ± SD 0.29±9.37 0.00±0.00 0.57±0.60 0.39±1.28 0.61±0.52 0.51±0.52 0.36±1.10 NA 0.28±119 NA 0.58±0.58 Late F5 – Prediction Late F5 Pooled 0.93 0.92 1.02 1.04 1.03 0.96 0.96 0.92 0.86 0.82 0.85 0.81 1.07 1.03 0.89 0.86 0.86 0.87 1.04 1.03 0.88 0.85 The BLUP model is essentially a ridge regression, so no attribute selection is performed. With over 2700 markers this invariably causes overfitting and hence reduced prediction accuracy. Indeed, machine-learning techniques (regression trees, lasso regression) achieve slightly better results for some traits (e.g. vegyld) but make no difference in others. This suggests that larger data set (more markers as well as more plants) is required in order to fit more successful models. Nevertheless, GEBV of the motherplants in the Late-flowering 5th population were able to accurately predict three of four parents selected for variety production based on an overall merit phenotypic evaluation. Conclusions A larger training sample and more markers are required to fit models capable of uncovering links between phenotypic traits and genotypic data. We hope to achieve this by adding historic data as well as new data from existing generations. Moreover, a new SNP-CHIP will increase the number of markers to around 4000. Coupled with powerful machine learning techniques, capable of attribute selection as well as uncovering non-linear relationships between response and regressors, we hope to find markers important for each trait and thus achieve higher prediction accuracies. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 828 References Hayes B.J., Lewin B.J. and Goddard M.E. (2013) The future of livestock breeding: genomic selection for efficiency, reduced emissions intensity, and adaptation. Trends in Genetics 29, 206-214. Heffner E.L., Sorrels M.E. and Jannink J.L. (2009) Genomic selection for crop improvement. Crop Science 49, 112. Meuwissen T., Hayes B. and Goddard M.E. (2001) Prediction of total genetic value using genome-wide dense marker maps. Genetics 157, 1819-1829. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 829 Prospects for introducing genomic selection into forage grass breeding Fè D.1,2, Ashraf B.1, Byrne S.3, Czaban A.3, Roulund N.2, Lenk I.2, Asp T.3, Greve Pedersen M.2, Janss L.1, Jensen J.1 and Jensen C.S.2 1 Dept. of Molecular Biology and Genetics, Aarhus University, Blichers Allé 20, Postbox 50, 8830 Tjele, Denmark. 2 DLF-Trifolium A/S Research Division, Højerupvej 31, 4600 Store Heddinge, Denmark. 3 Dept. of Molecular Biology and Genetics, Aarhus University, Forsøgsvej 1, 4200 Slagelse, Denmark Corresponding author: csj@dlf.com Abstract During the last decade, Genomic Selection (GS) has led to significant genetic progress in animal breeding. Only recently has the potential of GS in plant breeding been discussed. This paper presents the first results of GS implementation in a ryegrass (Lolium perenne L.) breeding programme. Both phenotype and genotype information were collected from 990 F2 families produced within a standard breeding programme at DLF-Trifolium. Phenotype data were recorded as family means, while genotype data were expressed as family allele frequencies. Statistical analyses were performed using linear mixed models (GBLUP), and displayed the presence of a significant level of genetic variance in all traits that was almost fully explainable by the genomic relationship matrix. Preliminary cross-validation analyses on heading date and crown rust resistance showed reliabilities around 0.50. Further studies will be undertaken on model development and in the study of gene by environment interactions. Keywords: Genomic Selection, Lolium perenne, GBLUP, GxE Introduction Genomic Selection (GS) is a form of Marker Assisted Selection (MAS), which explains the total additive genetic variance and calculates the individual’s Genomic Estimated Breeding Value (GEBV) by simultaneously estimating the effects of all loci across the entire genome. GEBVs are calculated from genome wide markers of sufficient density to ensure all loci are in Linkage Disequilibrium (LD) with at least one marker. This has been most commonly achieved by genotyping single individuals with high-density hybridization-based arrays. GS is now well established in human medicine and animal breeding, mainly in cattle and pigs. In plants, it is expected to considerably speed up the genetic progress, resulting in a decrease of the progenies needed to obtain a certain level of gain, especially for complex traits with low heritability (Bernardo, 2008). However, so far it has been tested almost exclusively through simulations and its actual potential has yet to be evaluated in reality (Conaghan and Casler, 2011). The aim of this study was to explore the possibility for implementation of GS in perennial ryegrass breeding. This faces two major challenges: (i) phenotypes recorded on heterogeneous families, (ii) high heterozygosity, which makes genotypes calls difficult on hybridization-based arrays. This work shows the methodologies used to overcome such problems, together with the first preliminary results. Materials and methods Plant material: 990 F2 families, produced between 2000 and 2009 in a standard breeding programme run by DLF-Trifolium. Phenotypic data: all families were sown, in multiple plots, in seven locations across Europe in the production year, and in Bredeløkke (DK), in 2011. Plots were then farmed for two cropping seasons. Data on different agronomic traits have been collected: (i) forage yield (green [GMY] and dry matter yield [DMY]), measured in kg m-2, (ii) scored traits, measured with a scale from 9 to 1: heading date, aftermath heading, crown rust resistance, winter hardiness, density, spring Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 830 growth. All traits were scored once on each replicate, except for crown rust resistance, which was scored three times per year. Genotype data: Genotype-By-Sequencing (GBS) libraries were prepared from individually barcoded F2 families. Sequencing data were aligned to a draft assembly of the L. perenne genome and SNPs were identified within the population. Allele frequencies of the variant were calculated at each SNP position. Statistical models: preliminary analyses were conducted using the following Linear Mixed Model: y = X1t + Z1i + Z2p + Z3il + Z4pl + Z5c + e, where: Xi, Zi = design matrices for fixed and random effects; t = vector of trial within year and location; i = vector of breeding values; p = vector of parent populations (pps); il, pl = vectors of breeding values and pps within location; c = vector of environmental effects within trial; e = residuals. In the model for crown rust resistance, the effect “scoring time” and its interaction with the other variables were added. Variance components were estimated by Restricted Maximum Likelihood algorithm (REML) using the software DMU (Jensen et al., 1997; Madsen and Jensen, 2010). Heritabilities were expressed as ‘within’ and ‘between’ locations. Genomic information, based on family allele frequencies, was then implemented by GBLUP, which use the genomic relationship matrix (Gmatrix) as (co)variance structure of the breeding values. Preliminary analyses were performed for heading date and crown rust resistance. Prediction reliabilities were determined by crossvalidation, deleting one family at the time from the dataset. G-matrix: It summarizes the inheritance of multiple phenotypic traits. It was constructed in the following way (Legarra and Misztal, 2008; Vanranden, 2008; Forni et al., 2011): given a matrix S, comprised of allele frequency estimates, Xi being the allele frequencies of a certain family and M being the centred matrix M i  X ij  X i , with j  1,2,...990 , G-matrix is equal to: G= (M’M)/K. K is a scaling parameter (K = Average diagonal ( M M )). Different G-matrices were obtained by selecting SNP markers with different sequencing depth and used for analyses. Results and discussion Variance components: All traits showed a significant level of genetic variance (Table 1). Table 1. Heritabilities in the historical data (with SE) across locations (-A) and for one location (-L). GMY = green matter yield; DMY = dry matter yield. Character† GMY DMY Aftermath heading Winter hardiness Density Crown rust resistance Spring growth Heading date h2A h2L 0.20 0.026 0.65 0.022 0.57 0.029 0.30 0.030 0.34 0.046 0.16 0.049 0.17 0.072 0.26 0.033 ------------- 0.59 0.033 0.64 0.042 0.59 0.057 0.37 0.022 0.48 0.031 0.67 0.052 For heading date and spring growth, it was not possible to estimate heritabilities within location, due to lack of repeated plots. Heritabilities across locations ranged between 0.17 and 0.34, depending on the trait. Estimates at one location were always higher (up to 0.67), showing the presence of significant Gene × Environment (G×E) interactions. A remarkable exception was crown rust resistance. In this case, the larger level of variation occurred between scorings, rather than between locations. Results also showed correlations between first-cut yield, heading date and aftermath heading (around 0.30), with early heading plant showing highest yield in spring and increased stem production. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 831 GBLUP and cross-validation: The genomic relationship matrices were able to explain a high proportion of the genetic variance. Results from both simulated and real data indicate genomic variance to be overestimated at low sequencing depth. That is caused by inflated diagonals in the genomic relationship matrix based on sequence data and, consequentially leads to an underestimation of the remaining family effect. There is no impact of sequencing depth on the estimate of environmental variances. Preliminary results from cross-validation showed reliabilities of GEBVs (squared correlation between GEBV and true breeding value) around 0.50 for heading date and crown rust resistance. Conclusions The amount of genetic variance, the proportion of variance explained by the G-matrix and the first results from cross-validation appear promising for implementation of GS in the breeding programme. Further studies are needed to take account of low sequencing depth. Different models (Bayesian Lasso, Bayesian variable selection) will be tested on several traits, to identify potential genes of large effects and estimating the effect of G×E interactions. Results for crown rust resistance seem to show a potential for developing varieties, which perform well across locations. References Bernardo R. (2008) Molecular markers and selection for complex traits in plants: learning from the last 20 years. Crop Science 48, 1649–1664. Conaghan P. and Casler M.D. (2011) A theoretical and practical analysis of the optimum breeding system for perennial ryegrass. Irish Journal of Agricultural and Food Research 50, 47–63. Forni S., Aguilar I., Misztal I. (2011) Different genomic relationship matrices for single-step analysis using phenotypic, pedigree and genomic information. Genetics Selection Evolution 43, 1. Jensen J., Mantysaari E. A., Madsen P. and Thompson R. (1997) Residual maximum likelihood estimation of (co) variance components in multivariate mixed linear models using average information. Journal of the Indian Society of Agricultural Statistics 49, 215-236. Legarra A., Misztal I. (2008) Technical note: Computing strategies in genome-wide selection. Journal of Dairy Science 91, 360-366. Madsen P. and Jensen J. (2010) A users guide to DMU. A package for analysing multivariate mixed models, Version 6, Release 5.0. (http://dmu.agrsci.dk/dmuv6_guide.5.0.pdf ) VanRaden P.M. (2008) Efficient methods to compute genomic predictions. Journal of Dairy Science 91, 44144423. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 832 Population selection within perennial ryegrass cultivars under simulated grazing Cashman P.1,2, Gilliland T.J.2,3, O’Donovan M.1 and McEvoy M.1 1 Teagasc Animal & Grassland Research and Innovation Centre, Moorepark, Fermoy, Co. Cork, Ireland 2 Queen’s University Belfast, Belfast, UK 3 Agri-food Biosciences Institute, Plant testing Station, Crossnacreevy, Belfast, BT5 7QJ, UK Corresponding author: patrick.cashman@teagasc.ie Abstract Perennial ryegrass (Lolium perenne L.) sward populations have the potential to change as they age, either positively or negatively depending on the synergy with the environment where they are sown. Although previous studies have examined population selection, few have examined an array of perennial ryegrass genotypes. The objective of this study was to examine if sward populations became physiologically unique from that at sowing as a result of sward management. Genotypes from 12 cultivars managed under simulated grazing for 5 years (A) were compared to a control sown from breeders seed of the same cultivar (C), creating 24 accessions which were sown as spaced plants. Plants were subjected to a number of phenotypic measurements as described by the Distinctness, Uniformity and Stability (DUS) tests (UPOV, 2006). Control accessions were distinguishable from each other except for 2. Aged accessions where distinguishable from their C accession equivalents, except for 1. Mean date of inflorescence emergence was the most useful character in distinguishing C accessions, while natural height a mean date of inflorescence emergence was the most useful character in distinguishing between A and C accessions. Keywords: perennial ryegrass, population selection Introduction Perennial ryegrass cultivar populations can have large genetic variation within population (Guthridge et al., 2001). As sward genotypes age, the genetic variance of the remaining population will reduce from the initial sowing date (Charles, 1970). This creates the opportunity for the characteristics of the sward to change over time, potentially negating the positive characteristics of a newly established cultivar and may be influenced by both environmental and management factors. Charles (1970) found selection occurred for both a higher yielding population and a lower yielding population in two separate swards compared to the original breeders seed. In a more recent study, selection for later mean date of inflorescence emergence and smaller plant height within a grazed cultivar was noted (Hazard et al., 2006). Previous investigations into population selection have examined a small number of cultivars of perennial ryegrass. Due to the large variation between perennial ryegrass cultivars, it is unclear how cultivar choice will influence population change. The objective of this study was to examine if the accessions could be described as distinct ‘varieties’ and to determine if population change occurred across a wide range of perennial ryegrass cultivars. Material and methods Plant populations for twelve cultivars (4 diploids (D) and 8 tetraploids (T)) were established as spaced plants, with two treatments for each cultivar, creating 24 accessions. The cultivars and their heading dates were: Alto (D; 15 May), Arrow (D; 22 May), Bealey (T; 22 May), Dunloy (T; 8 June), Dunluce (T; 31 May), Glencar (T, 6 June), Greengold (T; 31 May), Lismore (T; 28 May), Malone (T; 22 May), Navan (T; 9 June), Portrush (D; 14 June), Tyrella (D; 8 June). The two treatments were: aged (A) and control (C). The A accessions were sown as plots in autumn Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 833 2006. Plots were arranged in a randomized block design with three replicates. Each plot was 5 m × 1.5 m (7.5 m2). The plots were mechanically defoliated using a motor agria to a height of 4 cm (Etesia UK Ltd., Warwick, UK) from 2007 to 2011, inclusive. The aged plots were defoliated 10 times per year for the duration of the experiment. In September 2011, 35 perennial ryegrass tillers were extracted from each of the plots and sown in multi-pot trays, creating a total of 105 collected tillers for each A accession. The location from which each tiller was extracted was evenly spread across the plot, while avoiding the boundary areas. At the same time, plants were established in multi-pot trays from seed. The C accessions were derived from seed of the original seed lines for each of the twelve cultivars. Plants were housed in a glasshouse over winter before been placed outside in the trays in spring 2012. Tillers were cut with a shears every 4 weeks and fertilized to maintain healthy plants and promote tillering. In August 2012, a ley area was ploughed and cultivated. Plants were sown in a randomized block design with eight replicates per accession and ten spaced plants per replicate. Each plant was spaced 0.75 m from each of the plants surrounding it. A row of non-experimental guard plants were sown around the perimeter of the experiment to negate boundary effects on plants included for measurement. Plants were subjected to a number of phenotypic measurements as described by the Distinctness, Uniformity and Stability (DUS) tests (UPOV, 2006). Full details of the 22 characters recorded are described in the DUS protocol, the subset listed below presents the most important characters within the current experiment. Plants were checked for inflorescences emergence every Monday, Wednesday and Friday until inflorescences emergence had been observed on all plants. Mean date of inflorescences emergence was recorded when 3 ears had visibly emerged beyond the ligule on an individual plant. On the mean date of inflorescence emergence, natural plant height at inflorescence emergence was recorded. Finally, inflorescence length was measured with a ruler 30 days after the mean date inflorescences emergence for each accession replicate. Analysis was conducted using DUST9, SUMM9 and ANAL9 modules of the DUST analysis system (Weaterup, 1998). Results and discussion In the current experiment all the control accessions were distinct from each other except for Dunluce and Greengold (Table 1). The large number of distinctions between the control accessions confirms that the experiment was successful in discrimination between accessions that are deemed to be distinct on the European common catalogue of perennial ryegrass varieties (European Commission). Mean date of inflorescences emergence and inflorescence length created the most distinctions between control accessions, separating 77% and 59% of the total pairs tested, respectively. In all genotypes except Navan, the C accessions were distinct from their A equivalents (Table 1). The greatest number of distinct characters was observed between the C and A accessions within the Glencar and Greengold genotypes with 14 of characters distinct for both genotypes. In contrast, between the C and A accessions within the Alto and Lismore genotypes, only 1 character was distinguishable. Natural height at inflorescence emergence created the greatest number of distinctions between C and A accessions within genotype, accounting for 58% of pairs separated. A reduction of 23% for natural height at inflorescence emergence was observed from the C to the A accessions. Mean date of inflorescence emergence was the only character that did not create any distinctions between C and A accessions within genotype. This indicates that perennial ryegrass genotypes change over time in a simulated grazed sward and genotypes can change independently of each other. These results show a distinct change in population but will require a second year of measurement in order to be confirmed. The change in cultivar population may partly explain a change in the trait expression of cultivar leys over time. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 834 Table 1. (1) Number of characters distinct between control genotypes, and (2) between control and aged accessions within genotypes Aged Portrush Navan Malone Lismore Greengold Glencar Dunluce Dunloy Bealey Arrow Alto Genotype Alto 1 Arrow 3 2 Bealey 5 6 Dunloy 17 17 14 Dunluce 17 16 13 1 Glencar 15 17 10 2 2 Greengold 15 14 12 1 0 2 Lismore 17 18 15 1 1 4 2 Malone 7 9 3 5 2 4 1 8 Navan 13 15 12 4 3 5 2 2 4 Portrush 12 12 15 8 9 11 7 9 9 6 Tyrella 6 9 11 12 13 10 7 15 6 9 5 10 12 14 14 1 3 0 7 4 3 Conclusion With one year of measurement in this experiment it was possible to distinguish between all but two of the control accessions. All genotypes except Navan changed over time; however, there was a marked difference in the extent of character change observed between genotypes tested, with a large amount of characters distinct between certain accession pairs compared to very few distinct between other accession pairs. Mean date of inflorescence emergence was the most useful character when discriminating between C accessions. The characters that were most powerful in distinguishing between C accessions were not the most useful in determining distinctness between the C and A accessions within genotype. Natural height at inflorescence emergence proved to be the most useful character in distinguishing between C and A genotypes. Population selection of perennial ryegrass cultivars has the potential to change the trait expression of cultivars over time. References Charles A.H. (1970) Ryegrass populations from intensively managed leys: I. Seedling and spaced plant characters. The Journal of Agricultural Science 75, 103-107. Guthridge K.M., Dupal M.P., Kölliker R., Jones E.S., Smith K.F. and Forster J.W. (2001) AFLP analysis of genetic diversity within and between populations of perennial ryegrass (Lolium perenne L.). Euphytica 122, 191-201. Hazard L., Betin M. and Molinari N. (2006) Correlated response in plant height and heading date to selection in perennial ryegrass populations. Agronomy Journal 98, 1384-1391. UPOV (2006) Gudelines for the conduct of test for Distintness, Uniformity and Stability. Retrieved 02/12/2013, from http://www.upov.int/edocs/tgdocs/en/tg004.pdf. Weaterup S.T.C. (1998) Distinctness, Uniformity and Stability trial (DUST) analysis system. User Manual., Department of Agriculture for Northern Ireland Biometrics Division, Belfast BT9 5PX. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 835 Genetic gain in yield of perennial ryegrass (Lolium perenne), Italian ryegrass (Lolium multiflorum Lam.) and hybrid ryegrass (Lolium x boucheanum Kunth) cultivars in Northern Ireland Recommended Lists 1972-2013 McDonagh J.1,3, McEvoy M.,1 O’Donovan M.1 and Gilliland T.J.2,3 1 Animal & Grassland Research and Innovation Centre, Teagasc, Moorepark, Co Cork 2 Agri-Food Biosciences Institute Plant testing Station, Crossnacreevy, Belfast, Co. Antrim 3 Department of Biological Science, Queen’s University Belfast, Belfast, Co Antrim. Corresponding author: Justin.McDonagh@teagasc.ie Abstract Ryegrass cultivars form the basis of grassland production in Ireland and the UK. In Ireland, perennial ryegrass (Lolium perenne) is the dominant ryegrass species followed by Italian ryegrass (Lolium multiflorum Lam.) and Hybrid ryegrass (Lolium x boucheanum Kunth). Such cultivars are selected for their high dry matter (DM) yield throughout the year. However, there is a knowledge gap regarding the genetic gain achieved in DM yields across ryegrass cultivars under Irish conditions in recent years. Data were obtained from testing carried out for the Recommended List in Northern Ireland on the DM yields of ryegrass species for 1973-2012 for perennial, 1972-2013 for Italian and 1974-2012 for hybrid ryegrass. Each cultivar was tested under a simulated grazing and conservation management for a minimum of three years. Cultivars were grazed by animals for each management in the first year. For the simulated grazing management, 320 kg N ha-1 were applied and there were seven harvests to a residual height of 3 cm. For the conservation management, 350 kg N ha-1 were applied and there were five harvests to a residual height of 6 cm over the following two years. Italian and hybrid ryegrass were subjected to a 3-cut conservation management receiving 425 kg N ha-1 Genetic gains in DM yield have been achieved in perennial (+0.43%/year), Italian (+0.37%/year) and hybrid ryegrasses (+0.26%/year). Keywords: Lolium perenne, Lolium multiflorum Lam., Lolium x boucheanum Kunth, genetic gain, DM yield Introduction Perennial ryegrass (Lolium perenne L.), Italian ryegrass (Lolium multiflorum Lam.) and hybrid ryegrass (Lolium x boucheanum Kunth) are the dominant forage grasses in north-western Europe (Wilkins and Humphreys, 2003). A major focus of ryegrass breeding has been to achieve high levels of annual herbage production. However, rates of genetic gain in ryegrass cultivars are not well documented. Previous estimates have reported gains in DM yield of 4-5% per decade in perennial ryegrass (PRG) since the 1970s in Europe (Wilkins and Humphreys, 2003) with similar rates of gain reported in New Zealand (Easton et al., 2002). It is known that cultivars can re-rank when managed under a simulated grazing management when compared to a conservation management (Wims et al., 2009). Therefore the rate of gain under both simulatedgrazing and conservation managements needs to be established. Furthermore, there is little information on the rate of genetic gains in the dry matter (DM) yield of hybrid and Italian ryegrass cultivars. The aim of this study is to evaluate the rate of genetic gain in the DM yield of perennial, Italian and hybrid ryegrass cultivars listed on the Northern Ireland recommended lists over the past 40 years. The data obtained from the Northern Ireland recommended list provide a unique opportunity to evaluate gains in DM yield, as all cultivars were managed under a consistent protocol, at the same site, and under low disease or environmental pressure. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 836 Materials and methods Data were obtained on the DM yield of perennial, Italian and hybrid ryegrass cultivars from the Crossnacreevy Plant Testing Station, Co. Down, for the period of 1973-2013. These data were primarily used to compile the Northern Ireland recommended lists of grass cultivars. The dataset comprised 205 (40 early, 88 intermediate and 77 late-heading) perennial ryegrass cultivars. The dataset also included 35 Italian and 27 hybrid ryegrass cultivars. All cultivars were tested for a minimum period of three years under the recommended list protocol (DARD, 2013). In the first year, cultivars were grazed with cattle and measurements were then taken over the next two years under both a simulated grazing management and a conservation management. Perennial ryegrass was subjected to a simulated grazing management comprising seven harvests to a residual height of 3 cm and applications of 320 kg N ha-1. The conservation management comprised five harvests to a residual height of 6 cm and applications of 350 kg N ha-1. Italian and hybrid ryegrass were subjected to a 3-cut conservation management receiving 425 kg N ha1 . Data were analysed using fitted constant analysis to create comparable over-years mean values for each variety. Results and discussion Results from the study are presented in Table 1. For perennial ryegrass higher rates of genetic gain have been achieved under conservation management (+0.51%/year). Longer re-growth intervals under conservation management may allow cultivars to express more of their genetic potential. Higher rates of genetic gain in DMY have been achieved in perennial (+0.43%/year) and Italian (+0.37%/year) compared to hybrid ryegrass (+0.26%/year). Perennial ryegrass is by far the most important of these three species in terms of seed sales, accounting for 95% of seed sales in Ireland (Culleton et al., 1992). This would suggest a larger breeding effort on perennial ryegrass cultivars by plant breeders resulting in greater gains in DM yield. Within perennial ryegrass there have been slightly higher rates of genetic gain in early- (+0.46%/year) and lateheading cultivars (+0.45%/year) in comparison to intermediate- (+0.39%/year) heading groups. Increase in late-heading performance maybe attributed to a greater uptake of late-heading cultivars on farms (Grogan and Gilliland, 2011). These results indicate that the rate of genetic gain in the DM yield of ryegrasses is significantly lower than has been achieved for maize (2.6%/year; Tollenaar, 1989) and wheat (1.0%/year; Calderini et al., 1995). This may be due to the greater difficulties encountered in breeding ryegrasses. For example, there is not much scope for altering the harvest index of grasses, whereas much of the improvement in the grain yield of cereal crops has been achieved by increasing the proportion of plant biomass allocated to grain (Wilkins and Humphreys, 2003). There is also genetic variation within a ryegrass population making selection for a specific trait more difficult. Furthermore, the higher seed sales of cereal crops allow greater financial investment in plant breeding. Conclusion This study demonstrated that gains in DM yield have been achieved over the last 40 years across ryegrass species with the exception of hybrid ryegrass. There have been greater gains achieved in conservation yields of ryegrass cultivars. However, gains achieved are not significantly greater in comparison with other estimates in Europe. This highlights that further research is required to investigate possible pathways to improve the rates of genetic gain in ryegrass breeding. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 837 Table1. The percentage of annual gains in dry matter yield achieved in perennial ryegrass (1973-2013), Italian ryegrass (1972-2013) and Hybrid ryegrass (1974-2012) for early-, intermediate- and late-heading varieties from the Northern Ireland Recommended List, under simulated grazing and conservation management. Grass species No. of Cultivars % gain per year Gain in total DM yield t ha-1 P- value Simulated yield 202 +0.35 +1.6 *** Conservation yield 199 +0.51 +2.8 *** Simulated yield 40 +0.44 +2.0 *** Conservation yield 39 +0.48 +2.7 *** Simulated yield 88 +0.26 +1.2 *** Conservation yield 86 +0.52 +2.9 *** Simulated yield 75 +0.4 +1.8 *** Conservation yield 73 +0.5 +2.8 *** 35 +0.37 +2.5 *** 27 +0.26 +1.0 NS and heading-date groups Perennial ryegrass Heading group Early Intermediate Late Italian ryegrass Conservation yield Hybrid ryegrass Conservation yield References Calderini D.F., Dreccer M.F. and Slafer G.A. (1995) Genetic improvement in wheat yield and associated traits: a re-examination of previous results and the latest trends. Plant Breeding 114, 108–112. Culleton N., Cullen T. and McCarthy V. (1992) The decline of the herbage seed production industry in Ireland. Irish Geography 25, 98–101. DARD (2013) Grass and clover recommended varieties for Northern Ireland 2013/14. Belfast: AFBI Easton H.S., Amyes J.M., Cameron N.E., Green R.B. Kerr G.A. Norriss M.G. and Grogan D. and Gilliland T.J. (2011) A review of perennial ryegrass evaluation in Ireland. Irish Journal of Agricultural Food and Research 50, 65-81. Stewart A.V. (2002) Pasture plant breeding in New Zealand: where to from here? Proceedings of the New Zealand Grassland Association 64, 173–179. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 838 Tollenaar M. (1989) Genetic improvements in grain yield of commercial maize hybrids grown in Tario from 1959 to 1988. Crop Science 29, 1365-1371. Wilkins P.W and Humphreys M.O. (2003) Progress in breeding perennial forages for temperate agriculture. Journal of Agricultural Science 140,129-150. Wims C., Kennedy E., Boland T. and O’Donovan M. (2009) Effect of evaluation protocol and cultivar on the classification of seasonal and total dry matter yield. Proceedings of Irish Grassland and Animal Production Association conference. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 839 Variation in the reproductive development of perennial ryegrass (Lolium perenne) cultivars Wims C.M., Lee J.M., Rossi L. and Chapman D.F. DairyNZ, Private Bag 3221, Hamilton 3240, New Zealand DairyNZ, PO Box 160, Lincoln University, Canterbury 7647, New Zealand Corresponding author: Cathal.Wims@dairynz.co.nz Abstract Plant breeding has manipulated the flowering of perennial ryegrass (Lolium perenne) by developing later heading cultivars. However, the impacts of breeding on the intensity and temporal distribution of flowering are not known. This study compared the reproductive development of 23 perennial ryegrass/endophyte combinations. Two replicate plots were closed from grazing and tillers were collected fortnightly over a 10-week period, commencing on 21 October 2013. Plant development stage was determined on a sub-sample of 30 individual tillers per replicate, according to the Moore et al. (1991) scale. The development of each cultivar was then calculated using the mean stage count formula of Moore et al. (1991). The rate and timing of reproductive development differed among cultivars. As expected, cultivars with mid-season heading dates matured earlier than cultivars classified as late- and very late heading. While the intensity of flowering was similar between maturity groups, the temporal distribution of reproductive development varied: the late- and very late-season maturing groups had lower proportions of reproductive tillers early in the season. Keywords: Lolium perenne, reproductive development, flowering behaviour Introduction The nutritive value and growth pattern of perennial ryegrass (Lolium perenne) in spring and summer is influenced by its flowering behaviour. Nutritive value declines as plants move from vegetative growth to reproductive growth in spring (September to November in the southern hemisphere), in part due to changes in plant morphology. Plant breeding has manipulated the flowering of perennial ryegrass by selecting for later-heading cultivars that maintain their leafiness for longer and retain better nutritive value in spring (Lee et al., 2012). There are now ryegrass cultivars available with a range of heading dates. For example, in New Zealand perennial ryegrass cultivar heading dates range from -17 to +25 days (with day ‘0’ typically 22 October). However, the impacts of breeding on the intensity and temporal distribution of flowering are not known. The objective of this study was to compare the reproductive development of 23 perennial ryegrass cultivar/endophyte combinations spanning four decades of plant breeding in New Zealand. Materials and methods The reproductive development of 23 perennial ryegrass cultivar/endophyte combinations was compared in a pasture field trial in the Waikato, New Zealand. Cultivars included mid- and lateheading diploids, and late- and very late-heading tetraploids with a 25-day range in nominal heading date. Based on their nominal heading date, cultivars were classified into one of three maturity groups: mid-season maturing (day 0 to +6), late-season maturing (day +7 to +21) and very late-season maturing (day +22 to +25). Two replicates plots were closed from grazing and from late spring to early summer (21 October to 16 December 2013), tillers were collected fortnightly over a 10-week period. Plant development stage was determined on a sub-sample of 30 randomly selected individual tillers per replicate, according to the Moore et al. (1991) scale. A numerical index was then applied allowing the development stage of each cultivar to be calculated which was expressed as an Adjusted Mean Stage Count value (Moore et al., 1991). Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 840 The proportion of vegetative, elongating and reproductive tillers was also determined on both a numerical and dry weight basis. Data were analysed using a mixed models approach to repeated measures analysis of variance (Proc Mixed, SAS 9.3) with maturity group, sampling event, and their interaction as fixed effects, followed by a Tukey’s test for pairwise comparisons. Results and discussion The pattern of reproductive development differed between maturity groups. As expected, cultivars with mid-season heading dates matured earlier, while cultivars classified as late- and very-late heading matured later in the season. Figure 1 presents the adjusted mean stage count for each maturity group through the experimental period. Adjusted Mean Stage Count 140 120 100 80 60 40 20 0 21-Oct 04-Nov 18-Nov Date 02-Dec 16-Dec Figure 1: The adjusted mean stage count (± standard error of the difference) of mid-season maturing ( ), lateseason maturing () and very late-season maturing () cultivars over five successive sampling events in late spring to early summer A maturity group by sampling date interaction (P<0.001) was observed as the rate of reproductive development differed between the maturity groups. Initially the rate of reproductive development was greater for the mid-season maturing cultivars, while during the final sampling period the rate of reproductive development was greater for the late- and very late-season maturing cultivars. At the final sampling event, the proportion of tillers that had entered reproductive development was greater (P<0.05) for the mid-season maturing cultivars compared to the late- and very late-season maturing cultivars (0.76 versus mean 0.65 respectively; Table 1). However, at the final sampling, the proportion of reproductive tillers on a mass basis was similar between maturity groups (mean = 0.88), which may be more meaningful as tillers accumulate more dry matter as they mature (Moore et al., 1991). This suggests that the intensity of flowering was similar for each maturity group, although an additional sampling period would have been valuable to confirm this. While from an animal performance point of view cultivars that have substantially less flowering are highly desirable, this result is perhaps not surprising, as cultivars must have sufficient seed yield to be commercially successful (Stewart and Hayes, 2011). Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 841 Table 1: Reproductive tillers in the mid-season maturing (Mid), late-season maturing (Late) and very late-season maturing (VLate) cultivars expressed as a proportion of total tiller number or total tiller mass in each sample. Number Mass Sampling Mid Late VLate P< Sampling Mid Late VLate P< 21 Oct 0.01 0.00 0.00 NS 21 Oct 0.01 0.00 0.01 NS 0.17 a 0.00 b 0.00 b 0.33 a 0.01 b b 0.001 0.39 a 0.07 b 0.01 b 0.63 a 0.14 b 0.03 c 0.001 0.77 a 0.58 b a 0.79 b 0.64 c 0.001 0.76 a 0.66 b 0.86 0.85 4 Nov 18 Nov 2 Dec 16 Dec 0.001 4 Nov 18 Nov c 0.001 2 Dec 0.93 b 0.05 16 Dec 0.93 0.43 0.63 0.001 0.00 Significance P< Significance P< Maturity 0.001 Maturity 0.001 Sampling 0.001 Sampling 0.001 0.001 Maturity × Sampling 0.001 Maturity × Sampling Not Significant; a,b,c NS Means within a row with different superscripts differ (P<0.05) The temporal distribution of reproductive development varied between the maturity groups; following the initial sampling on 21 October, where there were virtually no reproductive tillers present, a greater proportion (P<0.001) of reproductive tillers was recorded for the mid-season maturating cultivars between 4 November and 2 December on both a numerical and mass basis. Reproductive development in spring leads to a decline in pasture quality in part due to an increasing proportion of stem to green leaf. To lessen the effects of reproductive development on pasture quality, plant breeding has focused on delaying flowering in spring by developing later heading cultivars. Results from this study suggest that selection for later-heading cultivars indeed leads to fewer reproductive tillers during spring, which may be promising from an animal performance perspective. However, evaluating the impact of altered flowering behaviour on pasture nutritive value is difficult due to a lack of primary data. Further investigation to evaluate the impact of later heading cultivars on pasture nutritive value and animal performance in seasonal pasture-based livestock production systems in New Zealand would be valuable. Conclusions Selection for later-heading cultivars in New Zealand has altered the temporal distribution of reproductive tillers: late- and very late-season maturing cultivars maintain lower proportions of reproductive tillers during mid-spring. There was little evidence to suggest that plant breeding has altered the flowering intensity of perennial ryegrass cultivars. Acknowledgements This work was funded by New Zealand dairy farmers through DairyNZ Inc. References Lee J.M., Matthew C., Thom E.R. and Chapman D.F. (2012) Perennial ryegrass breeding in New Zealand: a dairy industry perspective. Crop and Pasture Science 63, 107-127. Moore K.J., Moser L.E., Vogel K.P., Waller S.S., Johnson B.E. and Pedersen J.F. (1991) Describing and quantifying growth stages of perennial forage grasses. Agronomy Journal 83, 1073-1077. Stewart A. and. Hayes R. (2011) Ryegrass breeding - balancing trait priorities. Irish Journal of Agricultural and Food Research 50, 31-46. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 842 Pasture profit index: updated economic values and inclusion of persistency McEvoy M.1, McHugh N.1, O’Donovan M.1, Grogan D.2 and Shalloo L.1 1 Teagasc, Animal & Grassland Research and Innovation Centre, Moorepark, Fermoy, Co. Cork, Ireland 2 Department of Agriculture, Food and the Marine, Backweston Farm, Leixlip, Co. Kildare, Ireland. Corresponding author: Mary.McEvoy@teagasc.ie Abstract A pasture profit index was developed to identify the economic merit of a perennial ryegrass (Lolium perenne L.) cultivar for a grass-based ruminant production system. The traits of importance were: spring, mid-season and autumn DM yield, 1st- and 2nd-cut silage DM yield, quality (per unit change in DMD/ kg DM) across the months of April to July (inclusive), and persistency (relative to a 10-year base). Persistency was quantified by determining the change in ground score (GS) across years. Each 1-unit decline in GS is associated with a 1683 kg loss in DM yield. The use of the grass economic index enables the identification of cultivars that will provide the greatest economic contribution to a ruminant grazing system. The index illustrates the strengths and weaknesses of individual cultivars and it is expected that it will encourage increased usage of the superior performing cultivars. Keywords: economic index, perennial ryegrass cultivar, DM yield, quality, persistency Introduction In Europe, since 1965, grass breeders have increased DM yield by approximately 0.5% per year (Chaves et al., 2009). Clear improvements in breeding cultivars with delayed heading, less secondary heading, increased persistency and improvements in grass digestibility have been less evident. McEvoy et al. (2011) introduced the concept of applying economic values to a set of traits that have been identified as the most economically important within grass-based production systems in order to determine the total economic merit of a cultivar. The key traits of importance in an Irish seasonal grass-based production system are: spring, mid-season and autumn DM yield, 1st- and 2nd-cut silage DM yield, grass quality from April to July inclusive, and grass cultivar persistency. When the index was applied to cultivar data, persistency was not included in the overall ranking indices reported by McEvoy et al. (2011) due to a lack of information on cultivar persistency. Monitoring the rate of change in GS and linking this to yield decline presents an opportunity to estimate the potential persistency of a cultivar over a period of time. Cultivar evaluation trials are generally conducted for a relatively short period of time (<3 years) whereas at farm level, cultivars are expected to persist for up to 10 years and often longer. This anomaly creates the requirement to allow long-term cultivar persistency to be predicted based on short-term trial performance. The objective of this study was revise the economic values associated with the key traits using up-to-date price and cost information, while including persistency in the total economic merit of a cultivar. Materials and methods Sixty-three perennial ryegrass cultivars (33 diploids and 30 tetraploids) were sown in experimental plots at 5 locations in 2010 throughout Ireland by the Department of Agriculture, Food and the Marine (DAFM). Site locations were Backweston, Co. Kildare (5º 22/N; 6º30/W); Raphoe, Co. Donegal (54º52/N; 7º36/W); Fermoy, Co. Cork (52º08/N; 8º17/W); Athenry, Co. Galway (53º18/N; 8º45/W) and Kildalton, Co. Kilkenny (52º21/N; 7º20/W). The experiment was a randomized complete block with three replicates of each cultivar at each site. A simulated grazing management (defoliations every 3 – 4 weeks from mid-March to mid-October) was implemented at 4 sites throughout 2011 and 2012. At the Kildalton site, a 2-cut silage Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 843 management was imposed, with the first defoliation in early April, followed by the 1st silage harvest in late May, and the 2nd silage harvest in late June. Three aftermath simulated grazings followed, with the final defoliation in mid-October. A total of 310 kg N/ha was applied per year. All plots were harvested to a residual height of approx. 4.5 cm. Harvested material was weighed and a subsample was dried at 80ºC in an oven for 16 hours for DM determination. The dried samples from the Backweston site were milled and analysed for DMD using NIRS technology. Seasonal production was defined as spring (all herbage harvested prior to April 10), mid-season (herbage harvested from April 11 to August 10) and autumn (herbage harvested from August 11 to the final harvest in November). Data from the 1st and 2nd silage harvests within the 2-cut silage management were used to determine 1st and 2nd-cut silage DM yields. Ground score was estimated visually in December 2011 and 2012 at the four sites where the simulated grazing management was imposed. The difference in ground score between the two years was used to determine GSΔ. The economic values for each trait were determined using the Moorepark Dairy Systems Model (MDSM; Shalloo et al., 2004) which simulated a physical change in each trait of interest independently. The difference between the farm net margin before and after the change was simulated was divided by the change in the trait of interest in order to determine the economic value for a unit change in each trait (McEvoy et al., 2011). Results and discussion To investigate the effect of an increase or decrease in cultivar performance within each trait, base values were necessary to predict the economic merit of each cultivar for each trait. The base level of DM production was calculated using the average level of DM production on commercial grassland farms in Ireland (9.1 t DM/ha; Shalloo, 2009). The base values 1.22 (spring), 6.02 (mid-season) and 1.86 t DM/ha (autumn), respectively. The associated economic values are: ±€0.16, ±€0.04 and ±€0.11/ha per year, for each 1 kg difference from the base value for spring, summer, and autumn DM yield, respectively. There are no quality data available on commercial swards; therefore, to determine the base level for quality the VCU dataset of 63 cultivars was used. The base value for quality (g/kg DMD) for each month was as follows: April (849.6), May (848.6), June (814.4) and July (810.2); the associated economic value is ±€0.001, ±€0.008, ±€0.010 and ±€0.009 per unit change in DMD on a g/kg DM basis, for April, May, June and July, respectively. Currently there is no silage DM yield data available on commercial swards. To determine the base level for 1st and 2nd cut silage DM yield the VCU dataset was used. The base value for 1st and 2nd cut silage DM yield was determined using the two years data on 63 cultivars at the Kildalton site, with all the cultivars and all replicates included in the generated values. The base value for 1st and 2nd cut silage DM yield was 3785 and 3692 kg DM/ha, respectively. The associated economic value was ±€0.041 and ±€0.028, for each 1 kg change in silage, relative to the base values, respectively. Current recommendations are to reseed at least 10% of the grassland area each year (Shalloo et al., 2011), with swards expected to last 10 years before regeneration. This indicates the standard sward should last for 10 years at farm-level, and this was the base level for persistency in the model. The total cost of reseeding (€672/ha; M. O’Donovan, unpublished data) divided across the expected 10-year sward persistency, equates to a depreciation cost of €67/ ha per year, but if a sward persists for a shorter period the cost of depreciation increases. Cultivars surviving 10 years or more had no negative value for the persistency trait. The economic values and the total economic merit for a subset of the 63 cultivars are presented in Table 1. Cultivar 1 had the highest PPI value (€217/ha per year); from the sub-indices it is clear that this cultivar had a high performance in all traits except 1st-cut silage (-€4). When selecting cultivars it is important that a farmer examines the sub-indices to identify which cultivar performs well in the traits required, e.g. if selecting a cultivar for a silage paddock, the Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 844 emphasis on 1st and 2nd-cut silage should be increased. Alternatively, if formulating a mixture, cultivars should be selected which complement each other’s strengths and weaknesses. The updated economic values reported in this paper have not resulted in any significant change in the rank order of the cultivars outlined below. Table 1. Total Economic merit of 10 cultivars and their performance in the sub-indices Cultivar Details Cultivar Ploidy 1 2 3 4 5 6 7 8 9 10 T D T T T T T T T D HD 28-May 29-May 05-Jun 20-May 23-May 5-Jun 01-Jun 06-Jun 04-Jun 23-May PPI Sub Indexes (€/ha per year) DM Yield Silage PPI Total Quality Longevity st €/ha per year Spring Summer Autumn 1 cut 2nd cut 217 60 51 54 -4 10 47 0 188 67 56 89 -2 -15 23 -29 166 34 57 62 -5 4 31 -17 165 76 41 35 17 -4 7 -7 147 39 55 22 31 -4 2 0 137 46 51 26 -18 7 26 0 129 38 52 21 7 1 10 0 99 4 58 22 -8 12 29 -17 94 14 47 6 3 0 24 0 67 61 40 28 16 -12 -20 -45 PPI Details Conclusions The use of the pasture profit index enables the identification of cultivars that will provide the greatest economic contribution to a ruminant grazing system. The total economic merit clearly identifies the strengths and weaknesses of individual cultivars and will heighten the end-use and commercial sales of superior performing cultivars. More importantly it will allow farmers to choose cultivars that are suitable for the individual requirements of their farm. References Chaves B., De Vliegher A., Van Waes J., Carlier L. and Marynissen B. (2009) Change in agronomic performance of Lolium perenne and Lolium multiflorum varieties in the past 40 years based on data from Belgian VCU trials. Plant Breeding 128, 680-690. McEvoy M., O’Donovan M. and Shalloo L. (2011) Development and application of an economic ranking index for perennial ryegrass cultivars. Journal of Dairy Science 94, 1627-1639. Shalloo L. (2009) Pushing the barriers on milk costs/ outputs. In: Teagasc National Dairy Conference, Mullingar and Killarney, Ireland. 18-19 Nov 2009, pp. 19-39. Teagasc, Carlow, Ireland. Shalloo L., Creighton P. and O’Donovan M. (2011) The economics of reseeding on a dairy farm. Irish Journal of Agricultural and Food Research 50, 113-122. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 845 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 846 Theme 6 posters Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 847 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 848 The effect of resistance to mildew infection on ruminal fermentation of Lolium perenne Claes J., Davies T.E., Rees Stevens P., Wilkinson T., Mur L.A.J. and Kingston-Smith A.H. Institute of Biological, Environmental and Rural Sciences, Aberystwyth University, Penglais, Aberystwyth, SY23 3FG, United Kingdom. Corresponding author: ahk@aber.ac.uk Abstract In the field, resistance to microbial pathogens increases crop yield. The work described here was designed to investigate whether the priming of endogenous anti-microbial defence responses associated with resistance mechanisms could affect subsequent utilization by the rumen micro-organisms during colonization and fermentation of the ingested fresh forage. The non-discriminatory approach Fourier Transform Infra-Red Spectroscopy (FTIR) was used to profile the entire metabolome generated during in vitro fermentation of Lolium perenne leaves which had been previously infected with an avirulent mildew, thus generating localized lesions of plant-cell death as part of the Hypersensitive Response defence reaction. Regardless of mildew exposure there was an effect of incubation time on metabolic profile, with separation becoming more obvious with increasing fermentation time. Within these individual timepoints, inoculation with mildew was also seen to affect the metabolic profile with the profiles of uninfected (control) leaves clustering away from infected leaves at later timepoints. We therefore propose that pre-exposure of grass to infection by avirulent mildew elicits plant defence responses that modify subsequent colonization by the rumen microbial population. The consequences of this for fermentation efficiency are currently being assessed. Keywords: Lolium perenne, pathogen resistance, forage quality Introduction In the field, plants are subject to exposure to microbial pathogens. Resistance to pathogens such as mildew involves a cell death response known as the Hypersensitive Response (HR). This forms a localized barrier of dead cells in the region of the developing spore preventing access of the spore to nutrients and hence preventing fungal development (Mur et al., 2008). We have previously demonstrated that plant tissues induce stress-related processes associated with cell death and proteolysis on exposure to rumen-like conditions (Kingston-Smith et al., 2013; 2012). These defence responses include alterations in metabolism which could affect quality of the ingested feed and provision of nutrients to the colonizing microbial population (Huws et al., 2013). Hence we have tested the hypothesis that the process and products of microbial fermentation in the rumen are affected by prior exposure of the fresh forage feed to biotic factors. We have assessed the effect of prior exposure of Lolium perenne to four doses of an avirulent strain of mildew (thus inducing HR) on the metabolic profile generated during fermentation to determine the implications of breeding for resistance on post-ingestion metabolism of fresh forage feeds. Materials and methods Perennial ryegrass (Lolium perenne, cv. AberDart) was grown for six weeks in compost. Six independent replicates were each inoculated with avirulent mildew (previously grown on oat) at approximately 0, 20, 50 and 100 conidia/ mm 2 by use of a settling tower, after which the plants were placed in a growth cabinet for 48 h to allow the infection to develop. For in vitro fermentation, Hungate tubes were prepared to contain 0.5 g FW of grass (which had been previously cut to ~ 1 cm lengths) to which a 10% rumen fluid inoculum was added under a CO2 stream before the tubes were sealed with butyl rubber stoppers. Tubes were placed at 39oC in Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 849 the dark until they were destructively sampled at 0, 2, 4, 6 and 24 h after inoculation. The remainder was fractionated into 'residue' (plant material plus attached, colonizing bacteria), 'pellet' (planktonic phase bacteria) and 'footprint (cell-free liquid phase) samples were for analysis by FTIR as described previously (Kingston-Smith et al., 2013). Principle component analysis of FTIR and metabolite data was performed with Pychem software (Jarvis et al., 2006). Results and discussion The main factor affecting the metabolic profile of the plant-microbial interactome formed during fermentation of the grass was time (Figure 1). This was consistent with previous FTIRbased analyses of fresh forage fermentation (Kingston-Smith et al., 2013). Discriminant function analysis (DFA) showed clear separation of 24 h samples away from all others, with evidence of clustering of 6 and 4 h samples. The samples from 0 and 2 h time points were more similar in composition. An effect of treatment on residue composition was also detected (Figure 1). Within each time cluster, uninfected grass (treatment A) separated from those samples which had been previously infected with mildew (treatments B, C & D), for which some evidence for a dose-dependent effect could be detected, especially in samples taken 6 and 24 h after inoculation with rumen fluid. Loadings plot revealed that the sources of these differences were in the regions of the spectra associated primarily with amide bond containing metabolites (Nakanishi, 1963). Figure 1. Discriminant Function Analysis (DFA) of FTIR spectra derived from residue samples collected after 0, 2, 4, 6, and 24 h of fermentation of L. perenne previously inoculated with avirulent mildew at approximately 0 (A), 20 (B), 50 (C) and 100 (D) conidia/ mm2. Arrow indicates direction of separation according to treatment. The DFA model is based on 6 principal components representing 99.6% of the total variation of the dataset. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 850 Figure 2. Loadings plot showing the main sources of variation in the FTIR spectra derived from residue samples collected after 0, 2, 4, 6, and 24 h of fermentation of L. perenne previously inoculated with avirulent mildew at approximately 0 (A), 20 (B), 50 (C) and 100 (D) conidia/ mm2. Regions associated with particular molecular groups are shaded in grey. The dotted line indicates a position where no contribution is made to the variation seen in Figure 1. Conclusion Pre-infection of forage grass with mildew was shown to affect the metabolic profile generated during fermentation of the forage feed by rumen micro-organisms. These differences could be a consequence of differential nutrient provision because of differences in the metabolome arising as a result of implementation of defence reactions. Alternatively, the pre-exposure of plant cells to a microbial pathogen, and induction of successful resistance mechanisms could cause systemic changes in the forage which alter niche provision during colonization and can perturb the functionality. Ongoing work will determine if these metabolic changes were associated with differential colonisation profiles and the implications of any changes on functionality and rumen efficiency. References Huws S.A., Mayorga O.L., Theodorou M.K., Onime L.A., Kim E.J., Cookson A.H., Newbold C.J. and KingstonSmith A.H. (2013) Successional colonization of perennial ryegrass by rumen bacterial. Letters in Applied Microbiology 56, 186-196. Jarvis R.M., Broadhurst D., Johnson H., O’Boyle N.M. and Goodacre R. (2006) PYCHEM: a multivariate analysis package for python. Bioinformatics 22, 2565-2566. Kingston-Smith A.H., Davies T.E., Edwards J.E., Gay A. and Mur L.A.J. (2012) Evidence for a role for foliar salicylic acid in regulating the rate of post-ingestive protein breakdown in ruminants and contributing to landscape pollution. Journal of Experimental Botany 63, 3243-3255. Kingston-Smith A.H., Davies T.E., Rees Stevens P. and Mur L.A.J. (2013) Comparative metabolite fingerprinting of the rumen system during colonisation of three forage grass (Lolium perenne L.) varieties. PlosOne 8; e82801 Mur L.A.J., Kenton P., Lloyd A.J., Ougham H. and Prats E. (2008) The hypersensitive response; the centenary is upon us but how much do we know? Journal of Experimental Botany 59, 501-520. Nakanishi K. (1963) Infrared absorption spectroscopy, practical. Journal of Pharmaceutical Science – USA 52, 716. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 851 Disease resistance in red clover (Trifolium pratense L.) to stem nematodes and Sclerotinia Lowe M., Kelly R., Skøt L. and Mizen K.A. Institute of Biological, Environmental and Rural Sciences, Aberystwyth University, UK Corresponding author: mjl@aber.ac.uk Abstract Two important pathogens of red clover in temperate areas are stem nematodes and the fungal disease Sclerotinia. The red clover breeding programme at IBERS, Aberystwyth is focusing on reducing the damage to plant persistence caused by these pathogens. Keywords: Trifolium pratense, Sclerotinia, stem nematode, DNA fingerprinting Introduction Red clover (Trifolium pratense L) is an increasingly important forage legume for sustainable grassland systems, producing high dry matter yields of quality forage. However, varieties of this species tend to lack persistence, particularly under grazing. New varieties, such as AberClaret and AberChianti, have demonstrated increased persistence (Marshall et al., 2014). In red clover, pests and diseases are major contributors to loss of persistency, in particular crown rot (Sclerotinia trifoliorum) and stem nematode (Ditylenchus dipsaci). Sclerotinia trifoliorum is a difficult fungal infection to control. When soil-borne it is easily transferred and very difficult to eradicate. Once the infection is in the field its presence can mean a long period of non-productivity for the farmer. Sclerotinia was one of the reasons for the decreasing use of red clover in the UK during the 1970s and 1980s. Coupled with a potentially high mutation rate when compared to stem nematode, it has potential for recurrent outbreaks in the UK, particularly as wetter, milder winters and drier springs increase in frequency. Stem nematode also has significant effects on red clover yield and persistence, and, although slightly slower to develop, symptoms include severe stunting and death of plants. A major breeding target, and the aim of this project, was to use AberClaret and AberChianti, as well as other elite material and natural populations, in a selection and crossing programme in order to generate red clover populations with improved resistance to these two pathogens. This has now progressed to a stage where multiple generations of material are available for back crossing within both populations. The use of a limited number of molecular markers has enabled us to follow the pedigree, and avoid issues with inbreeding and the use of agronomically inferior material. There is also potential to combine both populations at the F2 generation to create a potential third population resistant to both diseases. Methods and results The plant varieties and lines used were: Redhead (nematode-susceptible control), Milvus, Merviot, Formica, AberRuby, Aa4512, Aa4494 (now AberChianti) and Aa4495 (now AberClaret). Each group was duplicated and 120 plants of each variety were tested for both diseases. All plants were inoculated at approximately 8 weeks after germination by inserting the inoculum (liquid for stem nematode and an agar plug for Sclerotinia) into the axil containing the developing meristem. A period of high humidity was applied to secure the inoculum into the plant, followed by assessment over a period of time. The assessment period for Sclerotinia from the point of infection was six days, with infection scoring recorded on the second, fourth and sixth days. Each plant was then separated into individual pots and further assessed over several months for latent infection. Plants were scored on a 5-point system (0= uninfected; healthy 5= infected, dead). Stem nematode assessment was carried out over a six-to- eight-week Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 852 period, depending on the nematode inoculation population. Three scores were recorded at each assessment for the lamina, petiole and meristem at two weeks, four weeks and six weeks. All of the survivors from both tests were sampled for DNA fingerprinting with microsatellite markers. This allows us to track each plant’s pedigree as the project progressed. Plants were then poly-crossed within their appropriate disease groups, harvested and weighed. Initial selection for F1 generation testing included any plant setting seed. Measurements of average seed weight and seed weight distribution were used to select three seed groups within each variety (Low, Average and High seed yield), and selected varieties were planted in field trials to assess agronomic characteristics. The stem nematode programme is currently undergoing F2 screening. Using flower-head weights, seed weights and DNA fingerprinting we have been able to maintain genetic diversity from the original survivor plants. However, the original F1 Sclerotinia screen only provided 12 survivors, thus severely reducing the gene pool. Stem nematode resistance is more easily assessed as resistant or susceptible, whereas the Sclerotinia screen produces degrees of resistance and tolerance. Resistant material shows no infection, or a limited halo effect around the point of inoculation. Tolerant material can show some initial infection along the petiole, but most importantly not into the meristem. This allows the plant to survive and grow away from the infection, due to HSR (Hyper Sensitive Response). In the first screens, both resistant and tolerant plants were used within the poly-cross. We have now carried out a new disease screen using ecotype and landrace material held within our gene-bank with a reduced number of the original varieties together with new varieties added to maintain genetic diversity. In addition, we now only select resistant plants and leave these plants as long as possible to assess latent infection. As a result we now have 71 resistant plants from three screens, involving a total of over 2700 plants. The resistant material is now being poly-crossed to generate F1 seed. Discussion This work forms the basis of a breeding programme providing a wide range of resistant plant material which can be retested and introgressed into other elite material when required. DNA fingerprinting has allowed us to follow the progression of an individual plant’s heritage. The ability to trace both maternal and paternal lines of each plant has enabled us to maintain a broad genetic pool of material. The stem nematode programme has advanced furthest, and the marker information has already helped make breeding selection choices within the F2 population selected for screening. This will also be invaluable for the Sclerotinia population selections in the future. As Sclerotinia evolves, new resistant material can more easily be selected. Stem nematodes are naturally occurring within soil, and it is only when a population increases on a suitable susceptible host that problems occur. This may in part explain why we find a higher resistance rate compared to Sclerotinia (71 of 2940) (Table 1). Table 1. Number of tested and resistant plants in F0 and F1 populations of stem nematode. Control plants have been omitted. F0 F1 Number of plants Tested Resistant 1080 170 850 307 We intend to continue screening through to at least the F4 generation for each disease group and start a programme of back crossing and screening into elite lines with another field trial at the F3 stage. We also plan to combine both F2 generations in one poly-cross in order to attempt to identify resistant plants to both Sclerotinia and stem nematode. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 853 Acknowledgements This work is funded by TSB-BBSRC project no. TS/J002895/1 References Marshall A.H., Lowe M. and Vale J.E. (2014) Persistence of red clover (Trifolium pratense L.) varieties in mixed swards over four harvest years. Grassland Science in Europe 19 (this volume). Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 854 Selection of white clover (Trifolium repens L.) for improved phosphorus use efficiency Lloyd D.C., Vale J.E. and Marshall A.H. Institute of Biological, Environmental and Rural Sciences (IBERS), Aberystwyth University, Gogerddan, Aberystwyth, SY23 3EE, United Kingdom Corresponding author:dal35@aber.ac.uk Abstract The development of high yielding white clover (Trifolium repens L.) varieties with improved phosphorus use efficiency has potential environmental benefits as well as economic benefits to the grassland sector, challenged as it is to reduce the environmental consequences of production through sustainable intensification. Survivor plants were collected from a long term, low input trial and crossed to identify germplasm with enhanced phosphorus use efficiency (PUE). Several lines were identified that showed improved performance relative to control varieties under P-limitation. This also related to the performance of the companion ryegrass. Keywords: Trifolium repens, PUE, variety development Introduction Phosphorus (P) is an essential macronutrient of plants. It serves as a structural element in nucleic acids and phospholipids, and plays a central role in biochemical energy transfer. External application of phosphorus fertilizer, either through the use of mineral phosphate or livestock manures, has a marked positive effect on agricultural crop productivity. Cultivated legume species have long assumed to have a particularly high demand for P for nodulation (Marschner, 1997), however, the veracity of this has been challenged (Sprent, 1999). Most phosphorus destined for use in agriculture is derived from phosphate rock and extracted through mining. The world's major reserves of rock phosphate are in the control of relatively few countries, including Morocco, China and the US. Demand for P is increasing concomitantly with increased global demand for food, yet reserves of rock phosphate are being rapidly depleted and are expected to run out in the next 50 to 100 years (Cordell et al., 2009). Moreover, surplus soil phosphate is a significant contributor to the eutrophication of freshwater systems and over-use is discouraged (Lemercier et al., 2008). Indeed, the EU water framework directive seeks to address this problem. A demand thus exists for improved varieties of crop plants that are able to perform competitively under low P fertilization regimes, whether through conservation of use or through an enhanced acquisition or uptake (Vance et al. 2003, Vance 2011). Mass selection provides a simple yet effective method for genetic improvement of outbreeding crops (Acquaah, 2012). Selection of individual plants within a population is made on the basis of desirable characteristics or through the elimination of individuals with undesirable characteristics. The selected plants are intercrossed and the expectation is that the resulting progeny, assuming reasonable heritability, will be improved for the characteristics in question relative to the original population. In this study we report on the selection of survivor plants from a long-term, low-input trial as a method for selecting for phosphate use efficiency (PUE). Materials and Methods Survivor plants were collected from a long term, low input trial at Bronydd Mawr Upland Research Centre, Brecon, Wales. These comprised four populations of small leaved white clover, each of 40 survivor plants. The populations were individually polycrossed in a glasshouse using insect pollinators and the resulting F1 seed sown. The 50 most vigourous F1 plants from each population were taken and polycrossed again, producing four low P families Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 855 (LPF1-4). The F1 generation were also analysed for foliar P content and 50 plants with high foliar P content, taken randomly from the four populations, were polycrossed to give a "high foliar P" family (High FP). Similarly, 50 plants with low foliar P content, taken randomly from the four populations were polycrossed to give a "low foliar P" family (Low FP). These six lines were sown in March 2011, along with the control varieties AberAce and AberAtom (both small leaved white clover varieties), in replicated plots (six of each line or variety) in a field exhibiting low soil P (2.43 ± 0.14 ppm) at IBERS, Gogerddan near Aberystwyth, Ceredigion, Wales. Plot sizes were 5 m x 1.2 m. Sowing rates were 2.1 g of white clover seed and 10.8 g of the companion perennial ryegrass (Lolium perenne L.) seed (cv. Premium) per plot. Plots were managed as per National List trial guidelines i.e. plots were split in two with half of the plots left unfertilized with P while the other half were fertilized at a rate of 175 kg ha-1 P2O5 a-1. The plots were harvested with a Haldrup forage harvester at a cutting height of 5 cm. Harvests were taken three times in 2011, and twice each in 2012 and 2013 (reflecting poor growing conditions). Fresh weights were measured and grass and white clover content analysed on a 300 g subsample from each plot. Dry matter yields of each were calculated after drying the subsample in a forced draught oven at 80 °C for 18 hours. Results and discussion As expected, P treatment had a marked effect on DM yield, both of white clover and total DM yield. The two control varieties were broadly similar, with the untreated plots having 50-54 % of the white clover DM yield and 67-69 % of the total DM yield of the P treated plots. Yields were in line with those expected from plots of this nature, incorporating small leaved white clover and perennial ryegrass varieties. The low P families derived from survivor plants from the long term low input trial (LPF1-4) gave variable results. Of the four families, in terms of white clover DM yield, only LPF2 showed any improvement relative to AberAce, but was indistinguishable from the other control variety AberAtom. The remaining three families were either comparable or lower yielding than AberAce. Table 1: Mean annual dry matter yields (kg ha-1 a-1 dm ± se). Columns show yields without phosphate treatment (-P) and with phosphate treatment (+P) as well as the ratio of -P yield to +P yield. Results are presented as means of three harvest years’ data. AberAce AberAtom LPF1 LPF2 LPF3 LPF4 Low FP High FP Clover yield -P 846 ± 109 935 ± 83 540 ± 53 1154 ± 172 651 ± 98 648 ± 178 668 ± 138 523 ± 134 +P 1568 ± 133 1860 ± 275 1230 ± 201 1710 ± 211 1306 ± 187 1172 ± 128 1429 ± 114 1336 ± 206 Ratio 0.54 0.50 0.44 0.67 0.50 0.55 0.47 0.39 Total yield -P 5754 ± 375 5641 ± 244 4902 ± 38 7131 ± 124 7378 ± 204 6111 ± 203 5742 ± 156 5097 ± 122 +P 8527 ± 145 8211 ± 202 8067 ± 125 9272 ± 20 7019 ± 110 7832 ± 77 7312 ± 43 7297 ± 158 Ratio 0.67 0.69 0.61 0.77 1.05 0.78 0.79 0.70 Where there was significant improvement in the LPF families was in total DM yield. Both LPF2 and 3 yielded significantly better than both control varieties in the untreated plots, with LPF2 also out-yielding the control varieties in the P treated plots. LPF2, 3 and 4 also showed a lower overall response to P application than the control varieties, as seen in their higher -P DM yield/+P DM yield ratio. A similarly reduced response was seen in the low foliar P family (Low FP), although there was no measurable increase in total DM yield. The improvement in total DM yield in LPF2 and more generally the lower response to application of P in the LPF families indicates that PUE is a trait that can be selected for, and that collection of survivor plants from low-input systems is a useful means to identify Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 856 germplasm for this trait. The mechanism by which the LPF families perform better in P limited conditions appears to affect the companion grass as well as the white clover, given that much of the improvement in DM yield, particularly in LPF2, relates to grass DM yield. What this might be is currently unknown, but may relate to variation in the ability of white clover to acquire P from the rhizosphere, resulting in improved supply to both species. Alternatively it may be due to enhanced potential for nitrogen fixation in low P conditions. Acknowledgements The authors acknowledge funding from the Department of Environment, Food and Rural Affairs (DEFRA) through the Sustainable Livestock Link Programme. References Acquaah, G. (2012) Principles of Plant Genetics and Breeding. Second edition. John Wiley & Sons. Chichester, UK. Cordell, D., Drangert, J.-O. and White, S. (2009) The story of phosphorous: Global food security and food for thought. Global Environmental Change 19, 292-305. Lemercier, B., Gaudin, L., Walter, C., Aurousseau, P., Arrouays, D., Schvartz, C., Saby, N.P.A., Follain, S. and Abrassart, J. (2008) Soil phosphorus monitoring at the regional level by means of a soil test database. Soil Use and Management 24, 131-138. Marschner, H. (1997) Mineral Nutrition of Higher Plants. Second edition. Academic Press. London, UK. Sprent, J.I. (1999) Nitrogen fixation and growth of non-crop legume species in diverse environments. Perspectives in Plant Ecology, Evolution and Systematics 2, 149-162. Vance, C.P., Uhde-Stone, C. and Allan, D.L., (2003) Phosphorus acquisition and use: critical adaptations by plants for securing a nonrenewable resource. New Phytologist 157, 423–447. Vance, C.P. (2011) Phosphorus as a critical macronutrient. In: The Molecular and Physiological Basis of Nutrient Use Efficiency in Crops, Hawkesford M.J. and Barraclough P (eds), Wiley-Blackwell. Chichester, UK, 229-253. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 857 Selection of contrasting cold-tolerant white clover genotypes from twentyeight populations naturalized in southern Chile and Argentina Acuña H.1, Inostroza L.2 and Pino M.T.3 1 Universidad de Concepción, Facultad de Agronomía, Chillán, Chile, 2 Instituto de Investigaciones Agropecuarias INIA, CRI Quilamapu, Chillán, Chile, 3 Instituto de Investigaciones Agropecuarias INIA, CRI La Platina, Santiago, Chile. Corresponding author Hernán Acuña, email: gacunap@udec.cl Abstract A collection of 28 populations of naturalized white clover in Argentinean and Chilean Patagonia, and two white clover cultivars, was used to select two groups of contrasting cold tolerance genotypes: 96 cold-sensitive and 96 cold-tolerant. The objective was to form an association-mapping population. Sixty young plants of each population were cold stressed progressively at -2, -4, -6 and -8 ºC for 48 hours and the damaged plants recorded. The records were fitted to the Weibull distribution and the three most tolerant and the three most sensitive populations were selected using the LT50 value (Lethal Temperature for 50% of the population) to choose the contrasting genotypes. The white clover populations showed a broad genetic variability for cold tolerance, which allowed the selection of 192 genotypes with divergent coldtolerance and formation of the association-mapping populations. Keywords: Cold tolerance, naturalized populations, Weibull distribution, plant survival Introduction An important aspect of the interaction between white clover and grass in a grass-clover mixture is the lower clover growth rate at low temperatures (5-15 °C), which affects the clover competitiveness during early spring and late autumn, resulting in a low contribution of white clover to total yield. White clover is naturalized in Chile from the central regions to the extreme south, covering a broad range of soil types, altitude, climates and latitude (30 to 55° S). The forage species germplasm collection developed during the 1990s with material obtained from Argentinean and Chilean Patagonia (from 37 to 55° S), gathered 28 populations conserved at the INIA forage germplasm bank in Temuco. It is expected that these naturalized populations would have adapted to the specific environmental conditions, including cold environments, from where they were collected, after one or two centuries from their introduction to each particular region (Svenning et al., 1997). This collection, in addition to two white clover cultivars, was used for selecting the contrasting cold-tolerant genotypes using young plants subjected to progressive low temperatures in a frost chamber. The selection of contrasting germplasm would allow the identification of genotypes with a broad range of molecular and phenotypic patterns necessary for association mapping studies. The objective of this work was to develop an association-mapping population composed of 192 genotypes with divergent cold tolerance (96 sensitive and 96 tolerant) which would subsequently be subjected to molecular and phenotypic characterization under field conditions. Materials and methods During the summer of 1994, 1995 and 1996, twenty-eight white clover populations were collected as seeds in the sites described in Table1. The white clover cultivars included were Haifa and Weka. Sixty seeds were taken from each population and germinated in nursery containers with 128 small holes of 27 cm3 (3×3×3 cm) under greenhouse conditions. Sterile peat was used as the substrate. Periodically the 1800 plants (30 populations × 60 seeds) were watered with Hoagland’s medium nutrient solution (0.5 X) and one week after emergence were Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 858 inoculated with specific rhizobium. Forty days after germination the plants were transplanted into small pots of 216 cm3 (6×6×6 cm) and grown until the first stolons reached around 10 cm length. During July and August the plants were subjected to progressively lower temperatures (-2, -4, -6, and -8 ºC) in a cold chamber for 48 hours at each temperature. Due to the chamber capacity the plants were treated in three groups of 20 plants of each population. At the conclusion of each cold period the number of damaged plants (all leaves and the thinner stolons dead) was counted. The plant survival records and the environmental temperatures were fitted to a Weibull distribution (WD). The shape and scale parameters were estimated for each population using Statgraphics V15 Software. Also, the LT50 (Lethal Temperature for 50% of the population) was estimated from the survival function of WD. The three populations with the highest LT50 were considered the most tolerant and the three which presented the lowest LT50 were considered the most sensitive to cold. From these two groups of populations the 96 most cold tolerant genotypes and the 96 most cold-sensitive genotypes were selected as follows: all the genotypes without damage at -8 ºC from each population, plus plants damaged at -6ºC chosen at random to complete 32 plants when necessary (tolerant); and all the genotypes damaged at -2, -4 and -6 ºC plus genotypes damaged at -8 ºC chosen at random to complete 32 plants when necessary (sensitive). Table 1. Geographical information of collection sites of 28 white clover population and two cultivars. Altitude (m) CO Pop 94-14 S39 25 W72 11 94-18 S40 00 W72 33 nr 250 CL CL 94-38 94-41 94-19 S40 16 W72 39 94-22 S40 30 W72 14 220 nr CL CL 94-24 S40 23 W72 47 94-28 S40 48 W71 46 94-30 S40 40 W71 53 nr 820 960 94-33 S40 26 W71 36 94-35 S39 56 W71 38 94-36 S39 57 W71 40 Pop Georeference (hddd°m") Georeference (hddd°m") Altitude (m) CO Pop Georeference (hddd°m") Alt (m) CO S40 07 W71 39 S39 46 W71 37 1050 820 AR AR 94-62 S39 10 W71 15 94-65 S39 33 W70 57 990 120 AR AR 94-45 94-46 S39 54 W71 36 S40 00 W71 30 810 890 AR AR 95-20 S44 15 W71 50 95-36 S45 34 W72 06 nr nr CL CL CL AR AR 94-50 94-51 94-52 S39 37 W71 23 S39 37 W71 23 S39 35 W71 27 880 880 880 AR AR AR 95-71 S46 15 W71 52 95-77 S47 09 W72 21 96-33 S51 34 W72 31 nr nr nr CL CL CL 970 850 AR AR 94-54 94-55 S39 33 W71 25 S39 04 W71 12 890 1400 AR AR 96-64 S52 43 W71 03 Weka 0 CL AU 850 AR 94-58 S39 19 W71 03 1290 AR Haifa OL *CO: country of origin (CL, AR, AU and IL for Chile, Argentina, Australia and Israel, respectively) Results and discussion Controlled cooling has been described as an indirect and highly effective selection method for improving cold tolerance in perennial forage species (Waldron et al., 1998; Castonguay et al., 2009). In this work, successively cold-stressing young white clover plants at -2, -4, -6 and -8 °C allowed us to observe a high genetic variability for plant/tissue survival between the 30 populations. When the temperature decreased the percentage of damaged plants increased from 1.8% (-2ºC) to 70.0 % (-8ºC) (Figure 1a). The highest variations among populations were observed at -6 and -8 ºC, fluctuating in ranges of 1.7 - 76.7% and 23.3 - 100% of damaged plants, respectively (Figure 1a). The Haifa and Weka cultivars differed broadly in genetic backgrounds due to their origin. Haifa was developed in Israel in a Mediterranean environment, and therefore it shows low winter growth and high spring-summer growth due to its tolerance of thermic stress. On the other hand, Weka was developed in Australia and bred for improving its autumn–winter productivity. In spite of these differences, both cultivars showed similar performances under the conditions imposed in this experiment, where they presented a threshold of damage similar to the most sensitive naturalized populations (Figure 1b). The response of the white clover naturalized populations (survival) to the environmental temperature fall was fitted reasonably to the WD (D test=0.091; P>0.05). The shape parameter of the distribution curves, estimated for each population, varied broadly from 4.5 to 31.7. In general, the most cold-sensitive populations showed a low value of the shape parameter, which Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 859 means an abrupt decrease in survival when the cold increases (Figure 1b). The LT50 was estimated from the survival function of the WD (Figure 1b). Populations 94-22, 94-36 and 9454 presented the lowest values of LT50 (-7.2 °C), and populations 94-65, 94-19 and 94-30 showed the highest values of LT-50 (-8.4 °C). The cultivars Haifa and Weka showed the same value of LT-50 as the cold-sensitive populations. The effectiveness of the methodologies applied to white clover in this experiment is being evaluated under field conditions in three environments, with a gradient of cold associated with altitude (100 to 1500 meters above sea level). a) b) Figure 1. Percentage of damaged plants (a) and plant survival probability curves of 30 white clover populations subjected to a gradient of low temperatures (-2, -4, -6 and -8 °C). Bold lines are the most sensitive and most tolerant populations. Conclusions The white clover populations showed a broad genetic variability for cold tolerance, which allowed the selection of 192 genotypes with divergent cold-tolerance and formation of an association-mapping population. The Weibull curve and its parameters may be used as a selection criterion for cold-tolerant genotypes. Acknowledgements This work is part of the project N°1130340 funded by FONDECYT Chile. References Svenning M., Røsnes K. and Junttila O. (1997) Frost tolerance and biochemical changes during hardening and dehardening in contrasting white clover populations. Physiologia Plantarum 101, 31-37. Waldron B.L., Ehlke N.J., Vellekson D.J. and White D.B. (1998) Controlled freezing as an indirect selection method for field winter hardiness in turf-type perennial ryegrass. Crop Science 38, 811-816. Castonguay Y., Michaud R., Nadeau P. and Bertrand A. (2009) An indoor screening method for improvement of freezing tolerance in alfalfa. Crop Science 49, 809-818. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 860 Analysis of changes in population structure over time in components of multi-species swards Kelly R., Skøt L., Skøt K.P. and Collins R.P. Institute of Biological, Environmental and Rural Sciences, Aberystwyth University, Gogerddan, Aberystwyth, SY23 3EE, United Kingdom Corresponding author: rok@aber.ac.uk Abstract In multi-species swards, the genetic structure of populations of heterogeneous outbreeding species is likely to change over time, potentially affecting inter-species population dynamics. In this study, mixtures of outbreeding forage crops were established in swards, and periodically defoliated by cutting or grazing for two years. Changes in population structure varied from species to species. White clover showed significant genetic change over two years, while red clover did not. For perennial ryegrass, a sub group in the initial population exhibiting unique alleles was no longer present two years later. Key genes associated with stress tolerance and flowering time varied significantly with these changes in population structure. Keywords: population structure, allele frequency, red clover, white clover, perennial ryegrass Introduction Programmes of forage germplasm improvement usually result in synthetic varieties derived from intercrossing between small numbers of parental genotypes. However, many of the major components of temperate grasslands are outbreeding, heterozygous perennial species. This means that they contain considerable within-population genetic variation. Consequently, although a successful forage variety released in the EU must satisfy the requirement of distinctiveness, uniformity and stability (as well as value for cultivation and use), it will nevertheless be composed of genetically distinct individuals. Because of this, within-species variation in traits is an important aspect of grassland diversity. Given the high degree of genetic variation present in outbreeding forages, genetic change will inevitably occur in pastures over time, and this is likely to affect species interactions and dynamics. In sown pastures it is well known that the establishment year is characterized by high genotype mortality rates and a decrease in variability among the survivors (Charles, 1961). A Common Experiment (CE) was set up within the EU-FP7 project ‘Multisward’ across a subset of partner sites to analyse responses of multispecies swards (MSS), compared with highly fertilized perennial ryegrass (PRG) monocultures, to contrasting managements. The CE imposed grazing and cutting managements typical of intensive production systems for 2-3 years on sward types differing in species number and composition. Within each sward type, grazed and cut plots received the same external applications of nitrogen fertilizer (N) and were defoliated at the same frequency to the same residual height. In the Aberystwyth site we established CE plots containing mixtures of four species (two legumes and two grasses) and analysed the degree of genetic change occurring in populations over time. We compared allele frequencies in the starting population (i.e. before the defoliation managements were imposed), and populations sampled in plots exposed to six harvests per year for two years under cutting and grazing managements. The markers employed for genotyping comprised simple sequence repeats (SSR) that had previously shown good allelic variation, and single nucleotide polymorphisms (SNP) and SSR in known genes. Some of these genes are involved in the expression or biosynthesis of abscisic acid (ABA) and jasmonate, two plant hormones linked to stress tolerance and response. ABA plays important roles in environmental stress responses (Liming and Jian-Kang, 2003). Jasmonate is crucial to wounding response signalling: deficiency negatively affects a plant’s Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 861 tolerance of herbivore damage such as grazing (Howe and Jander, 2008). Other marker genes are linked to salt stress, desiccation or long day flowering time (Skøt et al., 2011) Any allelic differences could assist further characterization of these genes, as well as our understanding of the genetic basis of any changes in populations. Materials and methods Four forage species in common agronomic use in Europe were included in the CE in Aberystwyth: two grasses (perennial ryegrass - PRG; tall fescue – FA) and two legumes (white clover – WC; red clover – RC). Plots were sown in September 2010. The effect of defoliation management on genetic change in MSS was analysed over time in three of these species (PRG, WC and RC) in both cut and grazed plots of the four-species mixture. Leaf samples were collected from 45 genotypes of each species at ‘time zero’ (before defoliation treatments were imposed in summer 2011), and at the end of the experiment in October 2013. The WC populations were genetically screened using 345 AFLP markers. Samples of RC and PRG were each analysed using 15 SSR markers. Additionally, the samples were screened for predicted SNP and SSR in known genes (Table 1). The resulting data were analysed for allele frequency and population structure. Results and discussion The population structure results varied between species. A principal component (PCoA) analysis of the WC AFLP data showed that the populations from the two years were genetically distinct (Fig 1a). Analysis of molecular variance (AMOVA) showed that 97% could be attributed to within-population, and only 3% to among-population variation. Interestingly, four markers were identified as being outliers in terms of population structure differentiation, so are possibly located in a genomic region under selection. Pop 2011 Pop 2013 Coord. 1 Figure 1a: PCoA of AFLP data for white clover (WC), comparing the 2011 population with 2013. Principal Coordinates (PCoA) Coord. 2 Coord. 2 Principal Coordinates (PCoA) Pop 2011 Pop 2013 Coord. 1 Figure 1b: PCoA analysis of SSR data for perennial ryegrass (PRG), comparing the 2011 population with 2013. PCoA and population structure analysis of the 2011 PRG population showed the existence of a small but genetically distinct sub-population. This sub-population was not identified in 2013 (Figure 1b). Some of the genes in the 2011 sub-population had unique alleles. This was particularly notable for FT-LD1, a gene associated with flowering time (Skøt et al., 2011). The 2011 sub-population contained two alleles not present in the IBERS PRG breeding population. Their absence in the 2013 population could be due to selection pressure or genetic drift. RC showed a large number of SNP markers and good allelic variation between individuals. However, there was no significant difference in population or allele frequencies between the 2011 and 2013 populations. Population structure analysis suggested that no admixture had Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 862 occurred within the population. This suggests that the RC population remained relatively unchanged during the two-year period with few or no new additional plants from new seed. Table 1. Key genes showing SNP and SSR variation in red clover (RC) and perennial ryegrass (PRG). Species Gene Putative / known functions RC NCED PRG SAG29 RC & PRG Aquaporin Abscisic acid biosynthesis (environmental stress response) Senescence-associated salt-stress protein (Abscisic acid dependant pathway) Water absorption (link with salt stress) PRG SRS1 Salt-stress root protein PRG Jasmonate Induced Protein Defence, wounding-induced (relevance to grazing) RC & PRG Allene Oxide Synthase Catalyses biosynthesis of jasmonate PRG ABC1 Oxidative stress-related ABC1-like protein RC PCC13-62 Response to desiccation RC Nitrate Reductase Reduces nitrate to nitrite RC P5CS Proline biosynthesis RC Epimerase Dehydratase Sugar nucleotide metabolic process RC TpZIP Fatty Acid Desaturase (chlorophyll biosynthesis) PRG FT_LD1 Long day flowering time PRG PRO1 (isotig06929) Prostrate growth to erect growth patterns Acknowledgements The research leading to these results received funding from the European Community's FP7 Programme under grant agreement no. 244983 (Multisward). References Charles A.H. (1961) Differential survival of cultivars of Lolium, Dactylis and Phleum. Journal of the British Grassland Society 76, 69-75. Howe G.A. and Jander G. (2008) Plant immunity to insect herbivores. Annual Review of Plant Biology 59, 41-66. Liming X. and Jian-Kang Z. (2003) Regulation of abscisic acid biosynthesis. Plant Physiology 133, 29–36. Skøt L., Sanderson R., Thomas A., Skøt K.P., Thorogood D., Latypova G., Asp T. and Armstead I. (2011) Allelic variation in the perennial ryegrass FLOWERING LOCUS T gene is associated with changes in flowering time across a range of populations. Plant Physiology 155, 1013-1022. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 863 Temporal genetic shifts in mono- and bi-specific swards of perennial ryegrass and red clover Cnops G., Muylle H., De Vliegher A., Vleugels T. and Roldán-Ruiz I. Institute for Agricultural and Fisheries Research, Plant Sciences Unit, Caritasstraat 21, B-9092 Melle, Belgium Corresponding author: isabel.roldan-ruiz@ilvo.vlaanderen.be Abstract Understanding the temporal changes in species composition and within-species genetic diversity in mixed swards is important for the development of field management approaches that exploit the whole potential of the combined species. Here we present the changes in SSRlocus diversity over time in populations of perennial ryegrass and red clover. Standard commercial varieties of both species were sown either in monoculture or in bi-specific swards. Field plots were managed according to standard practice, with 3-5 cuts per year. SSR markers were used to determine the genetic composition of the original and the survivor populations. Genetic shifts over a period of three years were generally small. Over time, perennial ryegrass populations became differentiated from each other in a predictable way. The behaviour of red clover populations was more unpredictable, indicating that local and/or random selection effects played a more relevant role for this species. Keywords: Trifolium pratense, Lolium perenne, yield, persistence, genetic shifts Introduction The production of forage ‘on farm’ is an important aspect of sustainable livestock production. Red clover (Trifolium pratense L.) and perennial ryegrass (Lolium perenne L.) are two of the most used forage species in temperate regions. In Flanders, mixed swards of these crops can render annual dry matter yields of 13.0-18.9 ton/ha (De Vliegher, 2007) of forage with excellent nutritive properties in terms of crude protein. One of the main uncertainties associated with these bi-specific swards is the persistence of the individual species over time. In particular, red clover displays low persistence (Boller et al., 2010). Although perennial ryegrass is more persistent than red clover, probably only a subset of the genotypes will persist over several growth seasons, while others will disappear. The dynamics of both species in mixed swards depends on the properties of the varieties chosen and inter-specific interactions, but environmental and random effects also play a role in the determination of the genetic composition at a given moment in time. This can have an influence on the sward yield and its spread over the growth season. We have investigated the temporal genetic shifts experienced by bi-specific swards of these two species over a period of 3 years, and compared it to the dynamics of monocultures. Commercial varieties with contrasting properties expected to affect the performance of the plants in the sward were chosen for analysis. Materials and methods The perennial ryegrass varieties Merks (high tillering) and Meloni (low tillering), and the red clover varieties Crossway (creeping, highly branched) and Lemmon (erect, medium branched) (Cnops et al., 2010; Saracutu et al., 2010; VanMinnebruggen et al., 2014) were used (Table 1). Field plots (6 m x 1.4 m) were established in April 2011 at ILVO, Merelbeke (51° 00' N 3° 48' E, on sandy loam soil) in two replications (A and B). The plots were harvested three (2011), four (2012) or five (2013) times during the following seasons with a Haldrup forage harvester at a height of 6 cm. Ryegrass and clover weight proportions were determined in a subsample of 300-500 g. Dry matter yields were calculated after drying in a ventilated oven (Vötsch) at 75 °C for 48 hours. In the spring of 2011 (T1) and 2013 (T5) leaf samples were harvested for Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 864 SSR-analysis. In monocultures of red clover (M04), 40 leaf samples were taken at each sampling moment, in monoculture ryegrass (M01) 60 samples, and in the mixtures (M06 and M07) 40 ryegrass and 30 clover samples were taken. These proportions were determined according to the different number of seeds of each species used for the establishment of the plots (Table 1). Table 1. Initial composition of the plots. Mono- and bi-specific plots were sown according to agricultural practice (70% grass: 30% clover). Figures represent the percentage of each species at sowing. The estimated number of viable seeds per m2 is given in brackets. Plot M01 M04 M06 M07 Lolium perenne Merks Meloni 1.0 (1400) 0.35 (490) 0.7 (980) Trifolium pratense Crossway Lemmon 0.35 (490) 0.15 (105) 1.0 (700) 0.3 (210) 0.15 (105) SSR analysis in perennial ryegrass was carried out using a set of 12 primer pairs, amplified in two multiplex sets. For SSR analysis in red clover we used a set of 18 primer pairs, amplified in two multiplexes. FSTAT 2.9.3.2 was used to estimate the alleleic richeness (Ar) and pair-wise FST values; observed heterozygosities (1-Qintra) were calculated in GENEPOP 4.2; Pair-wise FST values were used to construct a UPGMA tree in STATISTICA 11. Results and discussion Immediately after establishment (T1) most bi-specific plots were dominated by L. perenne, with the exception of M06A, in which serious weed contamination occurred (e.g. weeds represented 40.2% of the dry weight at T1 in M06A). Three years later (T5) the opposite relation was observed, with all plots dominated by T. pratense. This was confirmed when all the cuts of 2013 were combined (results not shown), indicating a high persistence of the red clover variety(-ies) used in this experiment. Table 2. Genetic diversity at two sampling times. ‘Ar’: allelic richness, ‘1-Qintra’: observed heterozygosity. ‘%Lp’ and ‘%Tp’: dry weight contribution of each species to the spring cut. Time T1 T5 Plot M01A M01B M04A M04B M06A M06B M07A M07B M01A M01B M04A M04B M06A M06B M07A M07B Ar 4.78 4.78 5.75 5.50 4.57 4.88 4.93 5.12 5.80 5.18 4.54 4.81 Lolium perenne 1-Qintra % Lp 0.79 0.72 0.75 0.65 0.81 0.76 0.78 0.70 0.44 0.63 0.77 0.67 Ar Trifolium pratense 1-Qintra %Tp 14.1 66.8 61.9 58.0 7.04 6.85 7.04 6.66 7.39 7.83 0.63 0.35 0.43 0.47 0.51 0.61 45.7 25.3 21.6 28.7 47.2 39.3 42.1 26.8 7.48 7.15 7.13 7.28 7.03 6.90 0.60 0.54 0.50 0.55 0.54 0.48 52.8 60.7 57.9 73.2 In general terms, the allelic richness changed little over time (Table 2), as confirmed by nonsignificant t-test results obtained in comparisons of allelic richness at T1 and at T5 (L. perenne P=0.94; T. pratense P=0.89). In both species, the highest allelic richness at T1 was found in the plots in which two varieties were combined (M06 for L. perenne and M07 for T. pratense), Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 865 but the differences were not large. At T5 this was still the case only for ryegrass. In both species, t-tests demonstrated that heterozygosity levels were similar at both sampling moments (L. perenne P=0.14; T. pratense P=0.47), confirming that no large reductions in diversity had taken place over three years. Remarkable exceptions were L. perenne in M06A and T. pratense in M07B, in both cases plots in which two varieties of the corresponding species were combined. It might indicate selective loss of one of the varieties over time. This is probably not the most plausible explanation for ryegrass as, if this was the case, a differentiation between the M06A and M06B plots at T5 would be expected, but this was not observed (see Figure 1A and further in the text). In contrast, for red clover the M07A and M07B plots became clearly differentiated at T5 (see Figure 1B and further in the text). In general, pair-wise FST values were low and only in some cases significant (results not shown). To investigate the general patterns of genetic differentiation among plots and sampling moments, UPGMA trees were derived from the FST values (Figure 1). In perennial ryegrass, T1 and T5 were clearly differentiated, and plots with only Merks (M01 and M07) clustered separate from plots in which Merks and Meloni were combined (M06). The pattern was more complicated in red clover. In this case, the M07 plots at T1 (Crossway and Lemmon combined) were clearly different from others. At T5 M07A had become genetically similar to plots in with only Lemmon (M04 and M06), which could indicate that Crossway might have disappeared over time. Comparison of A and B in Figure 1 suggests a larger influence of random factors for red clover than for perennial ryegrass, with larger divergence over time between A and B plots of the same treatment in red clover. Figure 1. Genetic relationships at two sampling moments (T1 and T5), based on pair-wise FST values. A) L. perenne; B) T. pratense. Note: be aware of the breakpoints in the X-axes. References Boller B., Schubinger F. and Kölliker R. (2010) Red clover. In: Boller B., Posselt U. and Veronesi F. (eds) Fodder Crops and Amenity Grasses, handbook of Plant Breeding 5, Springer, The Netherlands, pp. 439-455. De Vliegher A. (2007) Potential of clover and lucerne on the dairy farm today. Demonstration project Sustainable Agriculture 2003-2006. Final report, pp. 1-8 (in Dutch). Saracutu O., Cnops G., Roldán-Ruiz I. and Rohde A. (2010). Phenotypic assesment of variability in tillering and early development in ryegrass (Lolium spp.). In: Huyghe C. (ed.) Sustainable Use of Genetic Diversity in Forage and Turf Breeding. Springer Netherlands, pp155-160. VanMinnebruggen A., Cnops G., Saracutu O., Goormachtig S., Van Bockstaele E., Roldán-Ruiz I. and Rohde A. (2014) Processes underlying branching differences in fodder crops. Euphytica 195, 301-313. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 866 Persistence of red clover (Trifolium pratense L.) varieties in mixed swards over four harvest years Marshall A.H., Lowe M. and Vale J.E. Institute of Biological, Environmental and Rural Sciences, Aberystwyth University, Gogerddan, Aberystwyth, SY23 3EE, United Kingdom Corresponding author:thm@aber.ac.uk Abstract Improving persistence of red clover (Trifolium pratense L.) varieties is an important target of the IBERS red clover breeding programme. Identification of the factors contributing to poor persistence and testing improved varieties in field trials are integral to the breeding of persistent red clover varieties. Red clover varieties and selection lines bred for greater persistence were grown in mixed swards with hybrid ryegrass or a mixture of hybrid ryegrass and perennial ryegrass over four harvest years. A significant difference in the DM yield of the red clover varieties was observed, with yield in harvest year 1 greater than in harvest year 4. The red clover varieties differed in the extent of this decline, due to differences in the persistence of red clover plants within swards. The implication of these results for the use of red clover in sustainable grassland systems is discussed. Keywords: Trifolium pratense, yield, persistence, variety development Introduction Sustainable livestock production systems are increasingly reliant on the production of high quality forage that can be produced ‘on farm’. The high protein content of red clover (Trifolium pratense L.) makes it an increasingly important forage legume in such systems producing high dry matter yields of good quality forage (Frame et al., 1997). Despite the merits of red clover, under a typical UK management of 3 conservation cuts per year and a late autumn grazing, red clover-based swards tend to persist for only 2 to 3 years after which dry matter yields decline. Red clover varieties that are high yielding for up to 4 harvest years would be advantageous. Identification of the factors contributing to the poor persistence of red clover within swards and applying this information to the development of red clover varieties that combine high forage yields with greater persistence is the primary aim of the IBERS red clover breeding programme (Marshall et al., 2012). In spaced plants crown diameter is the morphological characteristic most associated with plant mortality and this was used as a selection criteria to develop red clover varieties that were more persistent. A previous paper described the dry matter (DM) yield of these red clover varieties over three harvest years (Marshall et al., 2012). This paper includes results from a subsequent harvest year quantifying the yield of these new varieties and selection lines in comparison with current commercially available varieties. Materials and methods Field plots (5m × 1m) of 12 red clover varieties comprising control varieties and selection lines were sown at IBERS, Aberystwyth in summer 2008 on soil of the Rheidol series. The red clover varieties were sown at a seed rate of 7.5 kg/ha in mixed plots sown with hybrid ryegrass (Lolium boucheanum Kunth) cv. AberEcho sown at 35kg/ha or with a mixture of perennial ryegrass (Lolium perenne L.) cv. AberDart sown at 12.6 kg/ha and hybrid ryegrass cv. AberEcho sown at 22.4 kg/ha. All plots were harvested three times in each of the following four harvest years with a Haldrup forage harvester at a cutting height of 5 cm. Fresh weights were measured and grass and red clover content analysed on a 300 g subsample from each plot. Dry matter yields of each were calculated after drying the subsample in a forced draught oven at 80 °C for 24 hours. In spring of harvest year four two 0.25 m2 quadrats were placed at random in each plot Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 867 and the number of red clover plants within each quadrat recorded. The experiment was a split plot design with three replicate blocks with grass mixture as main plots and varieties as subplots. Results and discussion Red clover has an 18-19% crude protein content (Frame et al., 1997) and can be grown across the UK. However, it lacks persistence in mixed swards and yields tend to decline after the second harvest year. Increasing the persistence and yield of red clover is increasingly recognized as essential to capitalize on its high forage quality. Previous studies have shown differences in the DM yield of red clover varieties over three harvest years (Marshall et al., 2012) but also highlighted the decline in yield of some varieties after the second harvest year. Yields of all the varieties and selection lines within the experiment declined in the fourth harvest year compared with year 1 (Figure 1). Total yields were generally similar between years 1 and 4; however, the greatest difference was in the red clover DM yield (Figure 1), leading to a lower proportion of red clover in the swards. The red clover DM yield of the varieties ranged from 7.5 to 22.6 t ha-1 in year 1 and from 1.54 to 10.2 t ha-1 in year 4, while the highest total DM yield in each harvest year was 27.4 t ha-1 in year 1 and 21.01 t ha-1 in year 4 (Table 1). The varieties AberClaret, AberChianti and the selection line Aa4559 have been selected for improved persistence and yield. Although the yield of these varieties also declined, the difference in yield between years 1 and 4 was relatively small compared to the control varieties Milvus and Merviot. Total yield 30 40 A. Red clover 20 10 0 B. Red clover Year 4 yield (t DM ha-1) Year 2 yield (t DM ha-1) 40 Total yield 30 20 10 0 0 10 20 30 Year 1 yield (t DM ha-1) 40 0 10 20 30 Year 1 yield (t DM ha-1) 40 Figure 1. Red clover and total (red clover + grass) dry matter yield (t ha -1) of 12 red clover varieties and selection lines grown in mixed swards. A) Yields in year 1 and year 2; B) Yields in year 1 and year 4. The decline in DM yield is a consequence of a loss in red clover plants per unit area, which is greater in some of the less-persistent varieties (Table 1). Plant density in spring of year 4 was highest in mixtures containing AberChianti and Aa4559 (23.3), and lowest in mixtures containing Merviot (6.0). The plant density results also support the observation that field trials of three-years duration are insufficient to fully show differences in persistence between red clover varieties - increasing the duration of field trials beyond three harvest years is essential to provide good evidence of differences in persistence between varieties. This also has implications for official variety testing systems. The current UK red clover variety testing system that evaluates variety performance is carried out over 2 harvest years without grazing. Our data show that the major differences between varieties in persistence and DM yield were Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 868 not apparent until the third and fourth harvest year. This suggests that testing beyond two harvest years is necessary to identify appropriate varieties that are suitable for longer leys. Table 1. Red clover and total (red clover + grass) dry matter yield (t ha -1) of 12 red clover varieties and selection lines in harvest years 1 and 4. Data are derived from a total of 3 cuts in each harvest year. Variety Aa4557 Aa4559 Aa4560 Aa4561 Milvus AberChianti AberClaret Pavo Merviot Britta Vivi Amos s.e.d. Sign.(*** P<0.001) Year 1 Red clover 16.2 (8) 19.5 (5) 14.2 (10) 15.5 (9) 22.6 (1) 19.8 (4) 21.9 (2) 21.3 (3) 18.3 (6) 14.1 (11) 7.5 (12) 17.6 (7) 1.17 *** Total 21.5 (8) 24.9 (3) 19.1 (10) 19.8 (9) 26.4 (2) 24.3 (5) 27.4 (1) 24.5 (4) 22.4 (7) 18.5 (11) 13.6 (12) 23.2 (6) 0.86 *** Year 4 Red clover 3.54 (6) 9.93 (2) 3.22 (7) 2.59 (10) 6.27 (5) 7.54 (4) 10.19 (1) 7.62 (3) 1.98 (11) 2.99 (8) 1.54 (12) 2.76 (9) 0.89 *** Total 17.52 (9) 20.14 (3) 18.79 (6) 16.99 (12) 19.65 (5) 20.04 (4) 21.01 (1) 20.20 (2) 17.89 (8) 17.36 (10) 17.27 (11) 18.14 (7) 0.68 *** Plant density (plants 0.25m-2) 22.7 23.3 13.8 9.9 15.4 23.3 20.5 14.6 6.0 15.3 6.3 6.5 3.08 *** Conclusion Evidence from these field experiments over four harvest years shows that some red clover varieties maintain a high DM yield into the fourth harvest year and the yield decline between harvest years 1 and 4 is considerably less than in some of the commercially available varieties. The decline in yield is attributed to loss of plant numbers within the sward. The results have implications for the duration of red clover variety testing systems and their relevance to the use of red clover in grassland agriculture. References Frame J., Charlton J.F.L. and Laidlaw A.S. (1997) Red clover. In: Frame J., Charlton, J.F.L and Laidlaw, A.S. (eds) Temperate forage legumes. CAB International, Wallingford, UK, pp.181-224. Marshall A.H., Lowe M. and Vale J.E. (2012) Improved persistence of red clover (Trifolium pratense L.) varieties in mixed swards. Grassland Science in Europe 17, 73-75. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 869 Developing an optimal sampling strategy to assess the quality of perennial ryegrass varieties on a national variety evaluation scheme Burns G.A.1,2, O’Kiely P.1, Grogan D.3, Conaghan P.4 and Gilliland T.J.5 1 Animal & Grassland Research and Innovation Centre, Teagasc, Grange, Dunsany, Co. Meath, Ireland. 2 School of Biological Sciences, Queen’s University of Belfast, Co. Antrim, Northern Ireland. 3 Crop Evaluation and Certification, Department of Agriculture, Food and the Marine, Leixlip, Co. Kildare, Ireland. 4 Crops Research, Teagasc, Oakpark, Co. Carlow, Ireland. 5 Sustainable Agri-Food and Sciences Division, Agri-Food and Biosciences Institute, Crossnacreevy, Co. Down, Northern Ireland. Corresponding author: Gareth.Burns@afbini.gov.uk Keywords: perennial ryegrass, nutritive quality, sampling strategy, recommended list Abstract The Irish national variety evaluation scheme measures in vitro dry matter digestibility (DMD) and water soluble carbohydrate (WSC) concentration of perennial ryegrass (Lolium perenne L.) varieties. Evaluations are made over several years of sowing, each harvested in the two years post-sowing, in replicated trials which are cost- and resource-intensive. The aim was to identity opportunities to optimize resource allocation for assessing in vitro DMD and WSC. The DMD and WSC rankings of combinations of year of sowing, harvest years and replicate blocks at two conservation cuts were compared to ‘definitive’ conservation rankings. Initially, increasing years of sowing was the best strategy to reliably rank the quality of varieties; however, there was a decrease in the rate of gain for an additional year of sowing that resulted in using an additional harvest year or block as the next effective strategy. Overall, a sampling strategy would require a combination of years of sowing, harvest years and blocks to account for inherent variation, experimental error and genotype × environment (G × E) interactions. Introduction The Irish national evaluation programme for perennial ryegrass varieties assesses DMD and WSC under a combined grazing and conservation management. At present both DMD and WSC are reported as mean values from all harvests taken in the year. Burns et al. (2013) recommended reporting both DMD and WSC as separate conservation and grazing values, rather than one annual value, as an acceptable balance between providing accurate, relevant information to the end-user while maintaining an accessible format. For the purposes of evaluation schemes the relative performance (i.e. ranking) of varieties is more important than the absolute values as these are particular to the test site and conditions. Thus, rank changes are of primary concern to variety evaluation schemes as they provide an indication of the consistency of ranking (Conaghan et al., 2008). Evaluation of quality is carried out throughout the year to ensure that the conditions during the assessment of varieties are representative of the target (i.e. on-farm) conditions. Evaluation schemes are under a fixed annual reporting cycle with limited resources to achieve their aims. As these constraints limit the amount of data that can be utilized for making recommendations, important decisions need to be taken regarding the optimal allocation of resources. The objective was to assess the influence selection criteria (year of sowing, harvest year, and replicate blocks) had on the conservation harvests rankings of in vitro DMD and WSC. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 870 Material and methods Fifteen perennial ryegrass varieties (6 intermediate- and 9 late-heading) were sown in two maturity group trials based on the heading date of each variety, in each of three years (2001, 2005, 2006) in a randomized complete block design (4 blocks). Each plot was harvested in the two subsequent years following sowing under a 6-cut combined simulated grazing and conservation management. At the conservation cuts (cut 2 and 3; hereafter referred to as ‘Silage 1’ and ‘Silage 2’) a c. 300 g sub-sample was taken from each harvested plot and assessed for DMD and WSC using NIRS. The ‘definitive’ conservation value and rank for DMD and WSC was calculated as per Burns et al. (2013). To assess the effect of selection criteria (year of sowing, harvest years and block), a mean DMD or WSC value per variety was calculated for each combination of selection criteria. All permutations of each of the selection criteria were assessed; for example, one year of sowing with one harvest year for two blocks had each individual year of sowing (2001, 2005, 2006) assessed at each harvest year (one- or two-year old) for the six permutations of two blocks (1+2, 1+3, 1+4, 2+3, 2+4, 3+4). This resulted in a total of 18 different permutations of 1 year of sowing, 1 harvest year and 2 blocks. The rank order was calculated within each maturity group for both DMD and WSC at both ‘Silage 1’ and ‘Silage 2’, and Spearman rank correlations were carried out between each of these permutations with their respective ‘definitive’ conservation rank. The mean Spearman rank correlation of all permutations was calculated and the standard error of this average was calculated and pooled across maturity groups. This process was repeated for all combinations of selection criteria. Results and discussion Overall, many of the correlations were weak between combinations of selection criteria and the respective ‘definitive’ conservation rank for both DMD and WSC. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 871 Table 1. Mean Spearman rank correlation (r) of combinations of selection criteria correlated with ‘definitive’ conservation ranking for WSC and in vitro DMD Selection Criteria WSC In vitro DMD Years of Harvest Blocks Silage 1 Silage 2 Silage 1 Silage 2 sowing Years r SEM r SEM r SEM r SE M 1 1 1 0.411 0.06 0.29 0.07 0.18 0.07 0.35 0.07 7 1 4 1 1 2 0.350 0.05 0.32 0.05 0.47 0.04 0.28 0.05 9 4 8 1 1 3 0.519 0.05 0.45 0.02 0.72 0.02 0.44 0.09 9 1 4 1 1 4 0.598 0.08 0.52 0.12 0.62 0.12 0.45 0.14 8 3 8 1 2 1 0.501 0.08 0.34 0.08 0.71 0.06 0.42 0.10 7 1 6 1 2 2 0.544 0.05 0.50 0.05 0.75 0.03 0.51 0.07 9 6 9 1 2 3 0.601 0.05 0.55 0.06 0.78 0.02 0.63 0.07 4 6 3 1 2 4 0.587 0.08 0.58 0.09 0.81 0.04 0.61 0.13 6 5 0 2 1 1 0.511 0.04 0.34 0.05 0.68 0.04 0.43 0.05 0 5 8 2 1 2 0.586 0.03 0.47 0.03 0.75 0.01 0.50 0.04 7 4 8 2 1 3 0.630 0.03 0.58 0.03 0.78 0.01 0.57 0.04 1 9 0 2 1 4 0.676 0.03 0.61 0.05 0.81 0.02 0.57 0.07 9 4 6 2 2 1 0.596 0.07 0.46 0.06 0.76 0.04 0.57 0.09 3 7 3 2 2 2 0.646 0.04 0.59 0.04 0.82 0.02 0.69 0.05 9 6 7 2 2 3 0.706 0.04 0.66 0.03 0.84 0.01 0.73 0.05 4 8 1 2 2 4 0.693 0.06 0.71 0.05 0.84 0.04 0.82 0.06 9 8 1 3 1 1 0.603 0.07 0.38 0.07 0.71 0.04 0.52 0.08 9 6 6 3 1 2 0.650 0.05 0.54 0.04 0.79 0.02 0.64 0.04 3 8 2 3 1 3 0.756 0.04 0.65 0.03 0.82 0.02 0.70 0.05 5 9 2 3 1 4 0.730 0.09 0.72 0.05 0.84 0.03 0.73 0.07 3 3 8 3 2 1 0.632 0.09 0.53 0.08 0.83 0.03 0.61 0.19 6 2 4 3 2 2 0.726 0.06 0.65 0.05 0.84 0.02 0.75 0.06 1 0 3 3 2 3 0.771 0.04 0.71 0.07 0.84 0.02 0.77 0.06 9 9 4 3 2 4 0.771 * 0.72 * 0.88 * 0.87 * 3 3 6 Correlations were higher for ‘Silage 1’ than ‘Silage 2’ with two exceptions (Table 1).The many low correlations may reflect the 15 varieties being from elite germplasms, such that varietal differences in quality traits may be relatively small and the small scale of experimental error may be sufficient to permit a significant re-ranking. G × E interactions have been reported in Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 872 evaluation schemes (Conaghan et al., 2008) and weak correlations may suggest that evaluation has occurred over a wide range of environmental conditions that are representative of conditions on Irish farms. Another potential consideration is the relative maturity of varieties at harvest. For pragmatic reasons, all varieties within a maturity group are harvested on the same day. Therefore, there is the possibility that varieties were assessed at varying stage of maturity in each year, differentially affecting some varieties. Increasing the number of years of sowing from one to two increased the correlation with the ‘definitive’ rankings by 0.153 and 0.112 for DMD and WSC respectively. Adding an additional year of sowing increased the correlations further but at a lower rate of increase (0.065, 0.067 for DMD and WSC respectively). An additional harvest year increased the correlation with the definitive rankings for both DMD and WSC, but to a lesser extent than an additional year of sowing (0.1385, 0.079 for DMD and WSC respectively). For both WSC and DMD increasing the number of replicate blocks increased the correlation with the ‘definitive’ conservation rank, whereby the rate of increase was smaller for each additional block. The use of blocks as a form of replication is important as these provide an assessment of the variation within varieties and improves the confidence with which a statistical comparison between varieties can be made. Therefore, for any sampling strategy, two blocks would be a minimum requirement to allow statistical inferences between varieties to be made. Financial and time implications of choosing selection criteria must also be considered. For example, evaluating additional year of sowing or harvest years requires the same time-frame but an additional year of sowing is more costly than an additional harvest year due to the cost of the re-sowing process. Furthermore, on farm, grass swards have a longer life cycle than two years, and additional harvest years may be an important indicator of the long term performance of the sward. Another consideration when assessing several quality traits is the relationships between those traits. A loss in performance in one trait may be compensated for by an improvement in another. This would be of particular interest to grass breeders who aim to obtain improved quality traits without a negative trade-off in other traits. In conclusion, inherent variation, experimental error and G × E interactions influence the quality ranking of perennial ryegrass varieties. To account for these factors, a combination of years of sowing, harvest years and replicate blocks is required to provide a reliable and robust assessment. A sampling strategy would require a minimum of 2 blocks to allow for statistical comparisons between varieties. Initially an additional year of sowing would be the best strategy to improve the reliability of quality rankings; however, subsequently either an additional block or harvest year would improve reliability more than an additional year of sowing. References Burns G. A., O’Kiely P., Grogan D., Conaghan P. and Gilliland T. J. (2013) Is the ranking of perennial ryegrass varieties nutritive value at individual cuts representative of the mean annual ranking. BGS 11th Research Conference, 2-3 September 2013, Dumfries, Scotland. Conaghan P., Casler M.D., McGilloway D.A., O'Kiely P. and Dowley L.J. (2008) Genotype x environment interactions for herbage yield of perennial ryegrass sward plots in Ireland. Grass and Forage Science 63, 107120. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 873 Screening reveals opportunities for high sugar cultivars of Lolium perenne L. Suter D., Hofer D. and Lüscher A. Institute for Sustainability Sciences, Agroscope, CH-8046 Zurich, Switzerland Correspondence to: daniel.suter@agroscope.admin.ch Abstract Tetraploid cultivars of perennial ryegrass (Lolium perenne) contain higher levels of water soluble carbohydrates (WSC) than diploid cultivars, which is advantageous for animal nutrition, milk composition and the reduction of N-emissions. However, due to reduced tillering, their swards are not as dense as those of diploid cultivars and are therefore less suitable for grazing. In a field experiment, 63 cultivars of perennial ryegrass and one cultivar of xFestulolium loliaceum were screened for WSC content in autumn 2010. The tetraploid group exhibited higher (P < 0.001) herbage WSC contents (142 g kg–1 DM) than the diploid group (124 g kg–1 DM). Two diploid ‘high sugar grass’ cultivars (HSGs), especially bred for high WSC content, contained WSC (140 g kg–1 DM) comparable to the tetraploid group (P = 0.72). It is concluded that diploid HSGs can reach the levels of sugar content of tetraploid cultivars and, thus, provide both dense swards for grazing and the advantages of high WSC forage with respect to animal nutrition, milk production and the environment. Keywords: Lolium perenne, sugar, WSC, variety Introduction A high content of water soluble carbohydrates (WSC) in the forage can reduce N-losses caused by urinary excretion and increase milk-protein yield (Miller et al., 2001; Staerfl et al., 2013), as compared to low-WSC diets. In certain cases a high WSC content may also increase dry matter intake (Moorby et al., 2006). In perennial ryegrass (Lolium perenne L.) tetraploid cultivars are known to contain more WSC than diploid ones (Gilliland et al. 2002; Salama et al., 2012). However, for grazing, agronomy may favour diploid cultivars, because their higher number of tillers provides a denser sward (Swift et al., 1993), which in turn may reduce sward damage and forage spillage due to fouling, especially under wet conditions. Therefore, it is important to know if there are diploid cultivars that reach the same levels of WSC as tetraploids, in order to meet the requirements of agronomy and animal nutrition, as well as the environment. For this purpose, in a field experiment, 63 cultivars of perennial ryegrass and one cultivar of xFestulolium loliaceum have been tested for WSC content. Materials and methods Field plots (6 m × 1.5 m) of 63 cultivars of perennial ryegrass and one cultivar of xFestulolium loliaceum were sown at Zurich-Reckenholz (47° 26' N 8° 30' E, 440 m a.s.l., mean annual temperature: 9.4 °C, mean annual precipitation: 1031 mm) in spring 2009 at a rate of 22 kg ha– 1 , corrected for germination ability. Of the 63 perennial ryegrass cultivars, 26 were diploid and 36 were tetraploid. Additionally, two diploid cultivars known as ‘high-sugar’ grasses (HSGs) were included in the test. From each plot, 200 g samples of fresh herbage were taken at a 5 cm cutting height at dawn, noon and dusk in late October 2010. The samples were treated in a microwave oven in the field immediately after cutting in order to stop enzymatic activity and were subsequently dried at 55 °C for 48 h. The samples were ground with a cutting mill (RetschSM1, Retsch, Germany) using a 0.75 mm sieve. Extraction and subsequent analysis with anthrone (Fischer, 1998), modified according to Trethewey and Rolston (2009), was conducted after purification with chloroform (Bligh and Dyer, 1959). For this study, the dawn, noon and dusk WSC values, expressed as glucose equivalents, were pooled into a mean daily value. The experimental design was a Latin rectangle with four rows and four columns. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 874 Diploid, tetraploid and HSGs were compared by contrasts following the analysis of variance (ANOVA) using the R statistical software package (version 2.15.2). Results and discussion The tetraploid cultivars (Figure 1) exhibited markedly (P < 0.001) higher contents of WSC sugars (142 g kg–1 DM) than diploid cultivars (124 g kg–1 DM). The two HSGs, #32 and #18, showed the highest WSC of all non-tetraploid cultivars (140 g kg–1 DM). Their WSC contents were significantly (P < 0.01) higher than the mean content of the diploid cultivars but did not differ (P = 0.72) from the tetraploid group. These findings from 64 cultivars confirm results obtained in an investigation with 12 cultivars (Gilliland et al., 2002), in which the diploid HSG cultivars attained the WSC levels of tetraploid perennial ryegrass cultivars. Although distinct differences between ordinary diploid cultivars and HSGs were detected in our autumn screening, spring and summer harvests should also be assessed in order to get to an allencompassing overview. Figure 1. Content of water soluble carbohydrates (WSC) expressed as glucose equivalents (g kg–1 DM) in herbage of an autumn regrowth of 63 cultivars of Lolium perenne L. and one cultivar of xFestulolium loliaceum cut at 5 cm. Groups of common tetraploid, 4n (light grey) and diploid, 2n (dark grey), as well as ‘high sugar’ (HSG) diploid (black) cultivars. Bold circles indicate means. Interestingly, the diploid cultivars #44, #52 and #24, which are not specifically marketed as HSGs, had high contents of soluble sugar, similar to those of the two HSGs (P = 0.65), revealing a certain potential for selection of new high-sugar cultivars from common diploid material. In the variety trials for the Swiss list of recommended varieties (data not shown), which contained the cultivars of the experiment described above, both the diploid HSGs (ground cover of 72% at a 14 cm row spacing) and the ordinary diploid cultivars (ground cover of 75%) had significantly (both P < 0.001) denser swards than the tetraploid group (ground cover of 55%). The data also indicate that the ground cover of diploid HSGs does not differ (P = 0.49) from that of the ordinary diploid cultivars. Conclusion We conclude that diploid HSGs can reach the levels of WSC of tetraploid cultivars while providing sward characteristics similar to other diploid cultivars. Thus, they may offer both Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 875 good sward characteristics for grazing and the advantages of high-sugar forage regarding animal nutrition, milk-protein and the environment. References Bligh E.G. and Dyer W.J. (1959) A rapid method of total lipid extraction and purification. Canadian Journal of Biochemistry and Physiology 37, 911-917. Fischer B.U., Frehner M., Hebeisen T., Zanetti S., Stadelmann F., Lüscher A., Hartwig U.A., Hendrey G.R., Blum H. and Nösberger J. (1997) Source-sink relations in Lolium perenne L. as reflected by carbohydrate concentrations in leaves and pseudo-stems during regrowth in a free air carbon dioxide enrichment (FACE) experiment. Plant Cell and Environment 20, 945-952. Gilliland T.J., Barrett P.D., Mann R.L., Agnew R.E. and Fearon A.M. (2002) Canopy morphology and nutritional quality traits as potential grazing value indicators for Lolium perenne varieties. Journal of Agricultural Science 139, 257-273. Miller L.A., Moorby J.M., Davies D.R., Humphreys M.O., Scollan N.D., MacRae J.C. and Theodorou M.K. (2001) Increased concentration of water-soluble carbohydrate in perennial ryegrass (Lolium perenne L.): milk production from late-lactation dairy cows. Grass and Forage Science 56, 383-394. Moorby J.M, Evans R.T., Scollan N.D., MacRae J.C. and Theodorou M.K. (2006) Increased concentration of water-soluble carbohydrate in perennial ryegrass (Lolium perenne L.). Evaluation in dairy cows in early lactation. Grass and Forage Science 61, 52-59. Salama H., Lösche M., Herrmann A., Gierus M., Loges R., Feuerstein U., Ingwersen B., Stelling D., Luesink W. and Taube W. (2012) Limited genotype- and ploidy-related variation in the nutritive value of perennial ryegrass (Lolium perenne L.). Acta Agriculturae Scandinavica Section B-Soil & Plant Science 62, 23-34. Staerfl S.M., Amelchanka S.L., Kaelber T., Soliva C.R., Kreuzer M. and Zeitz J.O. (2013) Effect of feeding dried high-sugar ryegrass (‘AberMagic’) on methane and urinary nitrogen emissions of primiparous cows. Livestock Science 150, 293-301. Swift G., Vipond J.E., McClelland T.H. Cleland A.T., Milne J.A. and Hunter E.A. (1993) A comparison of diploid and tetraploid perennial ryegrass and tetraploid ryegrass white clover swards under continuous sheep stocking at controlled sward heights. 1. Sward characteristics. Grass and Forage Science 48, 279-289. Trethewey J.A.K and Rolston M.P. (2009) Carbohydrate dynamics during reproductive growth and seed yield limits in perennial ryegrass Field Crops Research 112, 182–188. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 876 The influence of autumn closing date and spring opening date on herbage production and quality in spring and throughout the growing season Lawrence D.1,2, O’Donovan M.1, Boland T.M.2 and Kennedy E.1 1 Animal & Grassland Research and Innovation Centre, Teagasc, Moorepark, Fermoy, Ireland; 2 School of Agriculture and Food Science, University College Dublin, Belfield, Dublin Corresponding author: David.Lawrence@teagasc.ie Abstract In perennial ryegrass swards, leaf senescence rate increases during the winter period, while leaf extension rate declines, resulting in a reduction in herbage mass and quality. The objective of this experiment was to examine the effect of 4 autumn closing dates and 2 spring opening dates on herbage yield and chemical composition over two full grazing seasons, with particular focus on the early spring period. Each day delay in closing date from 1 Oct to 15 Nov reduced opening herbage accumulation by 13.4 kg DM/ha. Each day delay in opening date from 1 Feb to 1 March resulted in 6.8 kg DM/ha of herbage accumulation. Opening plots in February had no effect on cumulative spring herbage production and resulted in herbage with a higher proportion of leaf and a higher concentration of crude protein and dry matter digestibility when compared to opening in March. There was no significant difference in total yearly herbage yield between the plots closed in October and plots closed in November. Keywords: perennial ryegrass, yield, quality, winter growth Introduction During the winter period, the growth of perennial ryegrass (PRG; Lolium perenne L.) is based on a reduction in leaf extension and an increase in leaf senescence rate (Hennessy et al., 2008). As a result, the net accumulation of grass during the winter is low, particularly as the sward reaches a ceiling in herbage yield (Ryan et al., 2010). The final autumn grazing date (closing date) has a large influence on the yield and quality of grass available for early spring grazing (Ryan et al., 2010). As closing date is delayed in the autumn, herbage accumulation is reduced in early spring by up to 15 kg DM/ha per day (O'Donovan et al., 2002). The initial spring grazing date (opening date) will also influence the yield and quality of herbage available in the first and subsequent rotations (O’Donovan and Delaby, 2008). The objective of this experiment was to examine the effect of four autumn closing dates and two spring opening dates on herbage yield and chemical composition over two full growing seasons Materials and methods The experiment was conducted at the Animal and Grassland Research and Innovation Centre, Teagasc, Moorepark, Fermoy, Co. Cork, Ireland (52° 16' N 8° 25' W), from 15 October 2011 for two years. A one-year-old PRG sward (0.79 PRG) was divided into 24 plots (3 m × 5 m). The study was a randomized block design with a 4 × 2 factorial arrangement of treatments and included three replicates. The experiment was repeated over two consecutive growing seasons. The treatment factors included four autumn closing dates (CD1: 1 October, CD2: 15 October, CD3: 1 November and CD4: 15 November) and two spring opening dates, (ODE: 1 February, ODL: 1 March). Cumulative spring herbage yield was the sum of herbage production from opening date until the end of April. Cumulative summer herbage yield was herbage production from May until late August. Autumn herbage accumulation was the herbage yield at closing date. Each treatment received 215 kg/ha of fertilizer N over the growing season. All plots were harvested with an Etesia rotary blade mower (Etesia UK Ltd., Warwick, UK) to a stubble height of 4 cm. All mown herbage from each plot was collected and weighed, and a 0.1 kg subsample was dried for 48 h at 40oC to determine dry matter (DM) percentage and to calculate yield in Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 877 terms of DM/ha. The dried sample was milled through a 1 mm sieve and analysed for dry matter digestibility (DMD) and crude protein (CP) content using near infra-red spectrometry (NIRS, Model 6500, FOSS-NIR System, 3400 Hillerød, Denmark) and the equation developed by Burns et al. (2010). Prior to harvesting at each closing and opening date, a sample of PRG tillers were collected by cutting them to ground level using a blade. A 40 g subsample was then cut at 4 cm above the base of the tiller (representative of the plot defoliation height), and the leaf blades, stem (total true stem and pseudostem) and dead components were sorted above the 4 cm stubble height. The separated fractions were dried for 16 h at 90oC for DM determination. Data were analysed using the mixed model procedure in SAS v9.3. Model terms included CLD, OD, CLD×OD, replicate and year. Results and discussion There was an interaction when DMD was examined (Table 1). Digestibility of herbage increased from CD1 to CD3 when plots were defoliated on ODE; however, when plots were defoliated on ODL there was no effect of CD on DMD. Table 2. The effect of closing date on total DM yield and sward morphology the following spring. OD CD 01 Feb 01Oct Open yield 1207Ax Total yield 9726 A DMD g/kg 808a 01 Mar Significance 15Oct 01Nov 15Nov 1Oct 15Oct 01Nov 15Nov S.E . 1023Bx 671Cx 1467Ay 1267By 883Cy 84.2 *** * N.S. 287. * 3 N.S N.S. B 805ac 2.0 *** *** * B 693Cx AB 10814 10230 10378 817b 816b AB 812ab A B 752Cy AB 10226 10439 10679 10291 795c 797c 804c A CLD OD INT 0.23Ax 0.24Ax 0.25Bx 0.25Bx 0.22Ay 0.21Ay 0.20By 0.22By 0.00 *** 2 *** N.S. Leaf >4cm 0.37Ax 0.44BCx 0.42Bx 0.44Cx 0.33Ay 0.37BCy 0.38By 0.41Cy 0.01 *** 6 *** N.S. Stem>4cm 0.21ABx 0.21ABx 0.18BCx 0.15Cx 0.28ABy0.26ABy 0.24BCy 0.23C 0.01 * 9 *** N.S. Dead>4cm 0.42A 0.40A 0.36B 0.38AB 0.36AB 0.03 *** 3 N.S N.S. CP 0.35B 0.40AB 0.41A CD = Closing date, OD = Opening date, INT = Interaction of OD×CD, yield = opening or total yearly yield in kg DM/ha; Leaf >4cm = proportion of leaf above 4 cm stubble height, SE = Standard error, xy = within a row means with different superscripts differ significantly for OD; ABC = means with significant difference for CD; abc = means with significant difference for OD×CD; N.S = not significant, * = P<0.05, ** = P<0.01, *** = P<0.001 Each day delay in closing date from CD1 to CD4 reduced opening herbage mass by 13.4 kg DM/ha, which is intermediate between the 11 kg DM/ha reduction reported by Carton et al. (1988) and 15 kg reduction reported by O’Donovan et al. (2002). As a result, plots which were closed in October had a greater cumulative spring herbage yield than plots which were closed in November (Figure 1). The effect of CD on cumulative herbage production during the spring period was largely due to the difference in herbage accumulation from CD to OD, as CD had no effect on herbage production in April. Plots which were opened in March accumulated an additional 194 kg DM yield, or 6.8 kg /ha per day more (P<0.05), than ODE plots; however, in April plots which were opened in February yielded 111 kg DM/ha more herbage than ODL plots (P<0.05); as a result OD had no significant effect on cumulative spring herbage production. There was no effect of CD or OD on summer herbage production. In the autumn period, each day delay in closing from CD1 to CD3 increased herbage availability by 17.4 kg Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 878 Cumulative Yield Kg DM/ha DM. There was no significant difference in autumn herbage yield between CD2 and CD4, after an additional 31 days of growth. This suggests a ceiling in autumn herbage accumulation was reached between CD2 and CD3, similar to that described by Hennessy et al. (2008). There was no significant difference in total herbage yield between plots which were closed in October and those closed in November. The concentration of CP in the spring was higher in herbage that was closed after CD2. Hennessy et al. (2008) reported an increase in CP concentration with later closing dates as a result of increased leaf and reduced stem and dead proportions in the sward. Similarly, in the present experiment the proportion of leaf was higher in swards which were defoliated after CD1 as the proportion of stem decreased. An increase in the proportion of leaf indicates an increase in sward quality (Beecher et al., 2013). There was a higher proportion of leaf and a lower proportion of stem upon opening at ODE than ODL. As a result, the concentration of CP and DMD was higher at ODE than ODL plots (P<0.001); however there was no residual effect of OD or CD on CP or DMD after opening. 7000 6000 5000 4000 3000 2000 1000 0 01-Oct 15-Oct 01-Nov 15-Nov Spring Summer Season Autumn Figure 1. The effect of closing date on herbage yield during the entire growing season Conclusion By delaying closing date from 1 October to 15 November there was a 13.4 kg DM/ha per day reduction in opening herbage accumulation. Closing swards on 15 October maximized spring DM yield but did not result in any reductions in total herbage yield. Grazing from 1 February provides highly digestible grass with a high crude protein concentration, and has no effect on the cumulative herbage production over the spring period. References Beecher M., Hennessy D., Boland T.M., McEvoy M., O’Donovan M. and Lewis E. (2013) The variation in morphology of perennial ryegrass cultivars throughout the grazing season and effects on organic matter digestibility. Grass and Forage Science (DOI: 10.1111/gfs.12081, Early view, in press 2014). Burns G.A., Gilliland T.J., McGilloway D.A., O'Donovan M., Lewis E., Blount N. and O'Kiely P. (2010) Using NIRS to predict composition characteristics of Lolium perenne L. cultivars. Advances in Animal Biosciences 1, 321-321. Carton O.T., Brereton A.J., O’Keeffe W.F. and Keane G.P. (1988) Effects of autumn closing date and grazing severity in a rotationally grazed sward during winter and spring: 1 Dry matter production. Irish Journal of Agriculture Science 21, 141-150. Hennessy D., O’Donovan M., French P. and Laidlaw A.S. (2008) Factors influencing tissue turnover during winter in perennial ryegrass dominated swards. Grass and Forage Science 63, 202-211. O’Donovan M., Dillon P., Reid P., Rath M. and Stakelum G. (2002) A note on the effects of herbage mass at closing and autumn closing date on spring grass supply on commercial dairy farms. Irish Journal of Agricultural Science 41, 265-269. Ryan W., Hennessy D., Murphy J.J. and Boland T.M. (2010) The effects of autumn closing date on sward leaf area index and herbage mass during the winter period. Grass and Forage Science 65, 200-211. SAS (2006). SAS Institute. Cary, NC, USA. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 879 Increasing protein yields from grassland by reseeding of legumes Elsaesser M.1, Engel S.1, Breunig J.2 and Thumm U.2 1 Landwirtschaftliches Zentrum (LAZBW), Atzenberger Weg 99, 88326 Aulendorf, Germany 2 Universität Hohenheim, Institut für Kulturpflanzenwissenschaften, 70593 Stuttgart, Germany Corresponding author: Martin.Elsaesser@lazbw.bwl.de Abstract According to the politically motivated protein initiative in Baden-Wuerttemberg, it was the objective of a field experiment in South Germany to increase protein yields of farm-grown roughage avoiding the use of genetically modified soya as a feed for dairy cows. Three different legumes were tested at two different locations. First, an intensively used grassland region of Oberschwaben, with high rainfall and the best growing conditions for grassland, and secondly, in the summer dry region Swabian Alb, with shallow soils and only marginal growing conditions. Lucerne (Medicago sativa), red clover (Trifolium pratense) and white clover (Trifolium repens) were reseeded in permanent grassland swards with different sowing rates and different sowing dates. First results show better establishment of legumes by early seeding and subsequently higher yields of dry matter. Red clover was more successful than white clover and reseeding of lucerne. Keywords: protein yield, permanent grassland, legumes, Medicago sativa, Trifolium pratense Introduction Consumers in Germany wish for sustainable food production without genetically modified consituents. For milk production, therefore, the use of imported soya seems to be limited in the future. It is possible that feed protein for milking cows can be produced as farm-grown roughage by better use of grasslands (Buchgraber, 2001). Increasing protein in permanent grassland can be maintained by, among other means, higher nitrogen fertilization, early cutting dates or increasing the percentage of legumes in the grassland swards. Moreover, higher proportions of legumes allow production of protein with lower support of fossil energy through using the symbiotic fixation of atmospheric nitrogen. The increase of farm-grown protein is a political target in Baden-Wuerttemberg, as in other states of Germany. Estimations show a potential of around 800 000 t protein from grassland and forage fields (Engel et al., 2013). But high percentages of legumes are not easy to maintain because the nitrogen content of manure from intensive dairy farming is too high for the growth conditions of legumes. Additionally, red clover and lucerne are mainly used as legumes for arable field cropping and do not tolerate the frequent mowing of intensive grasslands. Therefore, in 2012 the agricultural institute Baden-Wuerttemberg (LAZBW), together with the University of Hohenheim, establsihed two field trials at two different locations. The objective was to increase the percentage of red clover, white clover or lucerne in permanent grassland via reseeding. First, the rates of emergence of the legumes are observed. Secondly, yields and protein delivery from the experimental sites and, moreover, from 10 fields on practical farms are measured. Here the results of seed emergence and results from the first experimental year are reported. Materials and methods In two grassland regions of Baden-Wuerttemberg (Oberschwaben: intensive grass production with 5 cuts; 1000 mm rainfall, 670 m a.s.l.; and Swabian Alb: shallow calcareous soils; 650 mm rainfall, 850 m a.s.l.) two experiments on permanent grassland were established. In the split-plot-design (size of parcels 10m², 3 replications) the following treatments were compared (Table 1). As legume species, white clover (Trifolium repens), red clover (Trifolium pratense) and lucerne (alfalfa) (Medicago sativa) were reseeded on two dates (early: 19 June Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 880 at Oberschwaben and 4 July at Swabian Alb; and late: 23 August at Oberschwaben; 27 August at Sw.A.). Two seed rates were tested. In order to ensure optimum seed establishment conditions and better control of new seedlings, the permanent grassland swards were treated with a herbicide before seeding. Additionally, the grassland swards were opened by harrowing in order to allow soil contact of the seed. Legumes were seeded mechanically using the 'Vredo' slot seeder system. Table 1. Treatments of legume seed rates, species, varieties and sowing dates Treatments Factors Plants and Control: Control without reseeding seed rate WKL6: white clover 6 kg ha-1 WKL15: white clover 15 kg ha-1 RKL10: red clover 10 kg ha-1 (varieties: Milvus and Merula) RKL20: red clover 20 kg ha-1 LUZ10: lucerne 10 kg ha-1 (varieties: Daphne and Sanditi) LUZ20: lucerne 20 kg ha-1 Seed date Early: late: (varieties: Riesling and Merlyn) in region Oberschwaben 19.06.2012 in Jura (Swabian Alb) 04.07.2012 Oberschwaben 23.08.2012 Swabian Alb 27.08.2012 In order to control the effect of reseeding, the emergence of seedlings was observed 5 weeks after sowing and categorized with an evaluation scale of 1-5 (Table 2). Dry Matter (DM) and protein yields were investigated from the beginning of the second experimental year (2013). Table 2. Scoring scale for the evaluation of the reseeded legumes 1 no no seedlings visible 2 few few seedlings visible 3 medium various seedlings visible 4 strong scattered rows of seedlings visible 5 very strong various rows of seedlings likely to be visible Results An overview of seed emergence of the reseeded legumes is presented in Table 3. It is clear that the early sowing date was more successful at both locations. The late-sown seed was not well established, especially at the Swabian Alb location. This is in line with results reported from Black (et al., 2006) in New Zealand, where spring seeded ryegrass-white clover mixtures show better yields than later seed dates in the year. Grabber showed similar results for red clover and also mentioned the effects of seeding dates (Grabber, 2009). The comparisons of seed rates show better seed emergence with the higher seed rates. Red clover, especially, was well established and scored best, whereas white clover and lucerne had problems with establishment. The late seeding date of lucerne had an advantage only in the intensive grassland area of Oberschwaben. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 881 Table 3. Evaluation of reseeded legumes in field trials in Oberschwaben and Swabian Alb (scale from 5 = fully closed rows of legumes to 1 = no legumes visible (see Table 2)) Site and sowing date -1 Oberschw. Oberschw. Sw.Alb Sw.Alb Treatment Seed rate (kg ha ) early late early late 1 control 0 0e 0e 0 0b 2 w.clover 6 6 3.3 cd 3.3 d 3.7 c 0.7 b 3 w.clover 15 15 4.3 ab 4.0 c 4.3 b 1.7 a 4 r.clover 10 10 5.0 a 4.3 bc 4.0 bc 0.7 b 5 r.clover 20 20 5.0 a 5.0 a 5.0 a 1.7 a 6 lucerne 10 10 2.7 d 3.0 d 3.7 c 0.3 a 7 lucerne 20 20 3.7 bc 4.7 ab 3.7 c 1.0 b The percentages of legumes varied independently of the seed rate of establishment. At both the grass-rich swards in Oberschwaben, and the more herb-rich grassland in Swabian Alb, the early sowing dates mostly resulted in higher percentages of legumes in the first experimental year (Table 4). Proportions of red clover were higher than those of either lucerne or white clover. Table 4. Average proportion of legumes (%) of each treatment of the experiments in Oberschwaben and Swabian Alb – average of all growths in 2013 Site and sowing date Oberschw. Oberschw. Sw.Alb Sw.Alb Treatment Seed rate (kg ha-1) early late early late 1 control 0 0.97 c 1.2 b 8.3 e 5.0 b 2 w.clover 6 6 7.9 b 1.4 b 13.9 ce 3.3 b 3 w.clover 15 15 8.3 b 2.4 b 17.8 c 4.3 b 4 r.clover 10 10 17.4 a 4.3 ab 28.9 b 6.6 b 5 r.clover 20 20 20.3 a 9.1 a 37.8 a 12.8 a 6 lucerne 10 10 1.44 c 1.5 b 11.7 de 3.7 b 7 lucerne 20 20 1.9 c 2.9 b 14.4 cd 4.4. b Conclusion In the future it will be necessary to use the potential of legumes to produce protein-rich roughage in order avoid using high rates of mineral-N fertilization and thereby reduce N2O emissions. A profitable objective for modern grassland management is therefore to increase the amounts of legumes in order to get higher protein yields, higher DM-yields and better forage values. This leads at the same time to a reduction of nitrogen emissions and lower energy consumption. The first results, after introduction of national strategies, show that reseeding with legume seeds increases the percentages of legumes, and that red clover has better chances for establishment in permanent grassland than white clover or lucerne. Early seed sowing dates seem to work better than late seeding, and protein yields were higher in the early seeded treatments. Moreover, the establishment of lucerne, which is usually used only in arable field Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 882 cropping systems, showed positive effects. It is assumed that a frequent reseeding of red clover and lucerne can be also a way to increase protein yields of grassland. References Buchgraber K. (2001) Eiweißersatz aus dem Grünlandfutter. Arbeitsgemeinschaft landwirtschaftlicher Versuchsanstalten, Jahrestagung 2001 in Wolfpassing, 147-148. Engel S., Elsaesser M. and Thumm U. (2013) Protein vom Grünland - Potentiziale nutzen. Landinfo 1/2013, 914. Lüscher A., Mueller-Harvey I., Soussana J.F., Rees R.M. and Peyraud J.-L. (2013) Potential of legume-based grassland-livestock systems in Europe. Grassland Science in Europe 18, 3-29. Grabber J.H., (2009) Forage management effects on protein and fiber fractions, protein degradability, and dry matter yield of red clover conserved as silage. Animal Feed Science and Technology 154, 284-291. Black A.D., Moot D.J. and Lucas R.J. (2006) Spring and autumn establishment of Caucasian and white clovers with different sowing rates of perennial ryegrass. Grass and Forage Science 61, 430-441. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 883 Change in birdsfoot trefoil (Lotus corniculatus L.) nutritive value with stem elongation, flowering and pod formation Hunt S.R.1, Griggs T.C.2 and MacAdam J.W.1 1 Department of Plants, Soils and Climate, Utah State University, Logan, Utah, 84321 USA 2 Division of Plant and Soil Sciences, West Virginia University, Morgantown, WV 26506 USA Corresponding author: jennifer.macadam@usu.edu Abstract Birdsfoot trefoil (BFT) grazed in pure stands supports rapid gain and high milk production in ruminants compared with grass pastures, and retains high nutritive value when stockpiled. In this study, the change in forage nutritive value from 4 to 12 weeks of mid-summer regrowth was assessed. Rapid stem elongation occurred until 7.5 weeks, and new internodes were added at stem tips through 15 weeks. At 12 weeks of regrowth, crude protein (CP), acid detergent fibre (ADF) and amylase-treated neutral detergent fibre (aNDF) were 199, 292, and 348 g kg-1 dry matter (DM), respectively; values that are associated with ‘good’ quality lucerne hay. However, total digestible nutrients (TDN), an indicator of energy availability for ruminant production, was 620 g kg-1 DM at 12 weeks of regrowth, higher than TDN values reported for ‘premium’ lucerne hay. Keywords: Lotus corniculatus, nutritive value, ruminant production, total digestible nutrients Introduction Birdsfoot trefoil contains sufficient condensed tannin to precipitate excess plant protein in the rumen, preventing bloat and reducing ammonia synthesis from protein deamination and lowering urinary nitrogen. In the abomasum, pH-mediated protein release from BFT tannins allows digestion of plant protein and absorption of amino acids (Waghorn et al., 1987). High levels of ruminally undegradable protein result in higher liveweight gain on BFT than lucerne (Douglas et al., 1995); higher non-fibrous carbohydrate (NFC) in BFT compared with lucerne may also contribute to higher ruminant productivity (MacAdam and Griggs, 2013). Rotational stocking at the bloom stage is recommended for long-term productivity of BFT (Undersander et al., 1993). In this study, the nutritive value of BFT was determined at the first bloom, full bloom, pod formation and pod maturity stages of a single regrowth cycle. Materials and methods Ten individual plants of ‘Norcen’ BFT were grown in 11.4-L pots containing a medium of vermiculite, bark, peat moss, perlite and nutrients. Plants were started from seed on 3 April 2011 in a heated greenhouse with 54 mol m-2 d-1 supplemental lighting. Initial growth was cut to a 75-mm stubble on 7 June 2011 and plants were moved to an outdoor bench in Logan, UT (41.74° N 111.83° W) until autumn. After overwintering outside, the eight surviving plants were returned to the greenhouse in early April 2012. Initial growth was cut to 75 mm on 7 June 2012 and plants were again grown on an outside bench until autumn. In both years, plants were irrigated to field capacity daily with water containing 0.5 g L-1 of 210 g kg-1 N, 22 g kg-1 P and 166 g kg-1 K. In 2012 at 4, 6, 8, and 12 weeks of regrowth, on 4 and 18 July and 3 and 30 August, one-quarter of each plant was harvested, dried at 55 °C to constant weight, and ground to pass the 1 mm screen of a cutting mill. Forage nutritive value was determined using nearinfrared reflectance spectroscopy. A mixed grass-legume hay equation developed according to procedures of Shenk and Westerhaus (1991) from a calibration set containing multiple legume species including BFT was used to predict sample composition. Calibration data were obtained as ADF, acid detergent lignin, aNDF, and CP according to AOAC (Latimer, 2012); in vitro true DM digestibility (IVTDMD, 48-hr incubation) according to Goering and Van Soest Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 884 (1970); NDF digestibility (NDFD) was calculated from aNDF and IVTDMD, NFC (DM - (ash + CP + ether extract + aNDF), and TDN according to the 2001 NRC Dairy summative equation (NRC, 2001). Changes in BFT forage quality parameters over 12 weeks of regrowth were described with linear regression for lignin, or with linear regressions on ln (x)-transformed data for other variables, using Excel version 14.3.1 (Microsoft 2011, Seattle, WA USA). Significance of linear regressions was evaluated using StatPlus:mac LE version 2009 (AnalystSoft Inc., Alexandria, VA). Results and discussion Following removal of herbage to 75 mm, stem growth was initiated from axillary buds on the remaining stubble and stem internode number produced a sigmoid growth pattern (Fig. 1). Fig.1 Internode number was recorded for 10 BFT plants from 28 June through 21 September 2011. Drawings indicate periods of full flowering, pod formation and seed maturation at 6, 8 and 12 weeks of regrowth, respectively. The end of linear stem growth at 7.5 weeks coincided with the beginning of seed fill. During regrowth, CP, NDFD, TDN, NFC, and IVTDMD all declined while lignin and ADF increased, as would be expected (Table 1). Table 1. Forage nutritive value variables at 4, 6, 8 and 12 weeks of BFT regrowth (n = 8). P-values are for regressions; means separations are indicated by different letters within rows. Weeks of midsummer regrowth 4 Crude protein (g kg-1 DM) 6 8 12 P values 267 a 210 b 213 b 199 b < 0.01 Acid detergent fibre (g kg DM) 196 c 239 b 272 a 292 a < 0.01 Amylase-treated neutral detergent fibre (g kg-1 DM) 221 c 276 b 322 a 348 a < 0.01 457 a 370 b 333 c 338 c 0.01 34 d 39 c 46 b 54 a < 0.01 716 a 682 b 642 b 620 b < 0.01 437 a 451 a 408 b 388 c 0.03 879 a 826 b 785 c 769 c < 0.01 -1 -1 NDF digestibility, 48-hr (g kg NDF) Acid detergent lignin (g kg-1 DM) -1 Total digestible nutrients (g kg DM) Non-fibrous carbohydrates (g kg-1 DM) -1 In vitro true DM digestibility, 48 hr (g kg DM) Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 885 The forage nutritive value of oven-dried herbage will be higher than for the same material dried and baled in the field. However, when compared with the nutritional characteristics reported for lucerne hay by quality category (Robinson, 1998), values for TDN were significantly higher for BFT than for lucerne with comparable CP, ADF and NDF values. The components of TDN are NFC, CP, fatty acid concentration, and digestible NDF (NRC, 2001), and MacAdam and Griggs (2013) also reported higher NFC for BFT than lucerne for 2 years of field data. These data support the results of Collins (1982) who reported autumn average daily cattle gains of 1.1 to 1.4 kg on stockpiled BFT-Kentucky bluegrass (Poa pratensis L.) pastures. Conclusions High nutritive value is maintained in maturing BFT, including high NFC and TDN, which may contribute to the high liveweight gains of beef cattle grazing stockpiled BFT. References Collins M. (1982) Yield and quality of birdsfoot trefoil stockpiled for summer utilization. Agronomy Journal 74, 1036-1041. Douglas G.B., Wang Y., Waghorn G.C., Barry T.N., Purchas R.W., Foote A.G. and Wilson G.F. (1995) Liveweight gain and wool production of sheep grazing Lotus corniculatus and lucerne (Medicago sativa). New Zealand Journal of Agricultural Research 38, 95–104. Goering, H.K. and Van Soest P.J. (1970) Forage fiber analyses. Agric. Handbook No. 379. Agricultural Research Service, US Department of Agriculture, Washington, DC. Latimer, G.W., Jr., ed. (2012) Official methods of analysis of AOAC International, 19 th ed. AOAC International, Gaithersburg, MD. MacAdam J.W. and Griggs T.C. (2013) Irrigated birdsfoot trefoil variety trial: Forage nutritive value. Electronic Bulletin.AG/Forages/2013‐02pr. Utah Cooperative Extension Service, Logan. http://extension.usu.edu/htm/publications/publication=15329&custom=1 NRC (National Research Council) (2001) Nutrient Requirements of Dairy Cattle: Seventh Revised Ed. The National Academies Press, Washington, DC. Robinson P.H. (1998) What are dairy nutritionists looking for in alfalfa hay? Proceedings, 28th California Alfalfa Symposium, 3-4 December 1998, Reno, NV. UC Cooperative Extension, University of California, Davis. Shenk J.S. and Westerhaus M.O. (1991) Populations structuring of near infrared spectra and modified partial least squares regression. Crop Science 31, 1548–1555. Undersander D., Greub L., Leep R., Beuselinck P., Wedberg J., Smith D., Kelling K., Doll J., Cosgrove D., Grau C., Peterson S., Wipfli M. and English J. (1993) Birdsfoot trefoil for grazing and harvested forage. North Central Regional Extension Publication 474. Cooperative Extension Service, University of Wisconsin. Madison. Waghorn G.C., Ulyatt M.J., John A. and Fisher M.T. (1987) The effect of condensed tannins on the site of digestion of amino acids and other nutrients in sheep fed on Lotus corniculatus L. British Journal of Nutrition 58, 115-126. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 886 Studies on forage quality of weed species in subalpine meadows in the Southeastern Carpathians of Romania Ciopată A-C., Maruşca T., Oprea G., Mocanu V., Blaj V.A. and Haş E.C. Research-Development Institute for Grasslands Brasov, Romania Corresponding author: andreea.ciopata@yahoo.com Abstract Experiments were conducted in the Bucegi Mountains, at 1800 m altitude, in the sub-alpine level of mountain pine (Pinus mugo), on a degraded pasture of Nardus stricta in a proportion of 40-60%. After 18 years of pasture improvement through liming and chemical and organic fertilization, under cow grazing system, there was a very high increase in dry matter (DM) and milk production, and Nardus stricta declined to disappearance. In contrast, after improvements, some species from the category of so-called weed species have proliferated, such as: Deschampsia caespitosa (4-12%), Polygonum bistorta (10-15%) and Ligusticum mutellina (610%) for which the chemical analysis was carried out. In conclusion, as a result of improvement works, these species have a high forage value and are finally fully consumed by dairy cows. Over a period of 18 years, on invaded variants of these 'weed' species, on average, 4.70 t DM ha-1 and 3180 L of milk per ha were obtained during 85 grazing days, due to the forage quality of the grassy carpet. Keywords: grassland, Nardus stricta species, improvement, weed invasion, forage quality. Introduction The most common Carpathian Mountain pastures, with degraded grassy carpet, are invaded by Nardus stricta (Marusca et al., 2010). After their improvement by various methods, Nardus stricta species is replaced by other more valuable species, including some so-called weeds, which usually are considered worthless. In several cases it was found that the 'weeds' actually have high forage value because of their content in nitrogen, phosphorus, potassium, etc. and valuable organic substances (Vintu et al., 2009). In long-term experiments in the Bucegi Mountains, there was a proliferation of several species of so-called weeds on some of the most efficient variants, leading to improvement of degraded Nardus stricta grasslands, and where was obtained the largest production of DM and cow milk yield per hectare. The question is, whether to destroy these 'weeds' or to keep them? The answer to this question we need information on the feed value of these plants, the main objective of this paper. Materials and methods Experiments on methods of improving degraded subalpine Nardus stricta grasslands were carried out in the Bucegi Mountains, at 1800 m altitude, since 1995. On permanent grassland, where Nardus stricta has an initial contribution of 40-60%, liming was applied at 7.5 t ha-1 CaO, to correct soil acidity, and fertilizers (N150, P50, K50 kg ha-1), on three consecutive years, followed by dairy cows paddocking (1 cow / 6 sq m / 5 nights) from 6 to 6 years. In 2013, after 18 years of the first intervention with chemical fertilizers, and followed by organic fertilizers with paddocking method, green biomass samples were taken in June, July, August from species Nardus stricta, Deschampsia caespitosa, Polygonum bistorta, Ligusticum mutellina and also for Trifolium repens, as control species for the qualities of forage. The first two species were collected from unimproved and improved grassland, and the next three species from improved plots where they were better represented. Green samples were chopped and measurements were carried out by near infrared spectroscopy - NIRS. After drying and milling the samples were used to determine the total nitrogen, crude Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 887 protein (CP), by the Kjeldahl method, crude fibre (CF), the cell wall (NDF, ADF) by Van Soest method and the organic matter digestibility (OMD) coefficients in the NIRS. For the assessments, the average of the three samples collected at different stages of vegetation were taken into account, these not showing large differences in the results. Results and discussion Chemical composition and digestibility coefficients of organic matter of Nardus stricta and Deschampsia caespitosa species for unimproved and improved alternatives are shown in Table 1. Table 1. Influence of grassland improvement on the organic matter content of Deschampsia caespitosa and Nardus stricta species Variant improvement Species Nardus stricta Deschampsia caespitosa of Content, (%) CP CF ADF NDF OMD 1. Unimproved 11.9 39.7 42.0 67.1 47.4 2. Improved 12.9 36.8 39.6 63.7 50.7 %, 2 compared to 1 108 93 94 95 107 1. Unimproved 8.3 36.9 41.4 67.8 46.8 2. Improved 17.2 31.1 33.8 57.4 63.9 %, 2 compared to 1 207 84 82 85 136 For Nardus stricta, there was no major differences in improvement of the content of crude protein, as it increased by only 1% compared to the unimproved version. Crude fibre content was decreased on average by 3% in the improved version, as in the case of cell wall content (ADF, NDF). The same difference of 3% was maintained for digestibility coefficients of organic matter, recording low values between 47-50%, indicating a low nutritional value to both. Compared with Nardus stricta, for Deschampsia caespitosa a significant difference was recorded between the two versions, crude protein for the improved variant rising from a value of 8.3% to 17.2%. Crude fibre content decreased by 5% in the improved version. In the case of lignocelluloses (ADF), the values were lower by 9%. The same trend occurred for the NDF value, in this case a decline of 10% for the improved version. Regarding the organic matter digestibility coefficients, difference between the two variants was 17% in favour of the improved variant (63.9%). The chemical composition and organic matter digestibility coefficients of the three species of plant (Trifolium repens, Polygonum bistorta and Ligusticum mutellina) are shown in Table 2. From the nutritional point of view, these three plant species have recorded similar values for their chemical element contents and the concentration of the cell walls. The two species Polygonum bistorta and Ligusticum mutellina reached high levels of crude protein, with values between 24-25%, as compared to 28.7% for Trifolium repens. Note that all three species had low crude fibre content, about 18%. Their high nutritional value is confirmed by the low cellular constituent contents: ADF values between 22 and 25%, and NDF values between 30 and 32%. Organic matter digestibility coefficients, OMD, had high values. The highest value was obtained from Trifolium repens (72.9%) followed by the other two species with very similar values: 69.6% for Polygonum bistorta, and 69.8% for Ligusticum mutellina. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 888 Table 2. The chemical composition of some species from improved plots Content, (%) Species CP CF ADF NDF OMD Trifolium repens content 28.7 17.9 22.8 30.8 72.9 %, relatively 100 100 100 100 100 Polygonum bistrorta content 25.0 18.2 27.4 36.6 69.6 %, relatively 87 102 120 119 95 Ligusticum mutellina content 24.0 18.2 25.3 32.0 69.8 %, relatively 84 102 111 104 96 For all the three species, determinations of the contents of phosphorus and potassium were made; values are shown in Table 3. Table 3. The mineral element contents of some plant species of mountain improved grasslands Ash Phosphorus Potassium Species Value % Value % Value % Trifolium repens 11.3 100 0.393 100 2.85 100 Polygonum bistorta 10.2 90 0.494 126 2.93 103 Ligusticum mutellina 10.4 92 0.632 161 3.51 123 In terms of the mineral composition of the two species Polygonum bistorta and Ligusticum mutellina have higher levels of phosphorus and potassium compared with Trifolium repens. Ligusticum mutellina species, with the highest values of P (0.632%) and K (3.51%) is distinguished. Conclusion As a result of improving the degraded sub-alpine grasslands of the Southeast Carpathians, after 18 years of experimentation, Nardus stricta species had a lower response to the improvement factors (liming, fertilization) in comparison with Deschampsia caespitosa which is consumed by animals and having better forage quality in these conditions. These plants are consumed mainly in the second half of the grazing season. After improvement works, the invasive species, referred to until now as 'weed species', Polygonum bistorta and Ligusticum mutellina, can be considered good forage, with a quality close to that of Trifolium repens, taken as the control species. In the specific conditions of the Carpathian sub-alpine zone, on improved grasslands, the Deschampsia caespitosa, Polygonum bistorta and Ligusticum mutellina species are consumed by dairy cows and can be considered as having medium to high forage value. References Maruşca T. (coordonator) şi colab. (2010) Tratat de reconstrucţie ecologică a habitatelor de pajişti şi terenuri degradate montane, Editura Universităţii Transilvania din Braşov, România, ISBN 978-973-598-787-9 Vîntu V., Samuil C., Iacob T., Saghin Gh., Popovici I.C. and Trofin A., (2009) The influence of organic fertilizers on Nardus stricta L. grasslands in the Carpathian Mountains of Romania. Grassland Science in Europe 14, pp. 555-558. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 889 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 890 Index of Authors Abraham E.M., 154 Acuña H., 858 Adamovics A., 285, 489 af Geijersstam L., 342 Agabriel C., 553 Ahmed L., 112 Ahmed L.Q., 115, 122 Albarrán-Portillo B., 622 Aldezabal A., 671 Amores G., 671 Andriamandroso A.L.H., 631 Andrieu B., 112 Aragon A., 731 Arias G., 169 Armstead I., 826 Arrigo Y., 593 Asel A., 139 Ashraf B., 830 Asp T., 830 Audsley E., 61 Aufrère J.3, 616 Baars T., 553 Baert J., 172, 175 Bailey J.S., 556 Bakken A.K., 603 Balázsi Á., 294, 298 Baldissera T.C., 353, 356 Balshaw H., 270 Bannink A., 119 Bär A., 563 Barcarolo R., 553 Barre P., 112 Barro R.S., 353, 356 Barron L.J.R., 671 Barth S., 438 Bartley D., 97 Batista C., 288 Baumont R., 521, 616, 734 Beaufoy G., 743 Bee G., 593 Beecher M., 616 Bellocchi G., 97 Benaouda M., 587 Beňová D.1, 370 Berzins P., 486 Biegemann T., 125 Bijelić Z., 597 Bindelle J., 625, 631 Biniaś J., 798 Blackmore T., 826 Blaj V.A., 887 Bodner A., 655 Boland T.M., 616, 628, 877 Borreani G., 553, 668 Both Z., 606 Böttger F., 619 Breitsameter L., 103 Breunig J., 880 Brocard V., 559, 807 Brook A.J., 236 Bruckmaier R.M., 538 Buchmann N., 166, 722 Buckingham D.L., 236 Buckley F., 795 Bühle L., 477 Bullock J.M., 254 Burns G.A., 870 Bustamante M., 671 Byrne S., 830 Campion M., 376, 641, 776 Canals R.M., 347, 743 Capuano E., 674 Cardasol V., 397 Carpinelli S., 356 Carvalho P.C. de F., 353, 356 Cashman P., 833 Ceceviciene J., 468 Chadwick D., 61, 128 Chambers B.J., 270 Chapman D.F., 840 Chassaing C., 553 Chevalley S., 680 Chidovet S., 306 Chinea E., 288 Chodkiewicz A., 364 Cholastova T., 606 Christodoulou A.S., 145 Ciopata A.C., 397 Ciopată A-C., 887 Claes J., 849 Clarke A., 233 Clement C., 776 Cnops G., 864 Collins R.P., 695, 719, 861 Combes D., 112, 242 Comino L., 668 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 891 Conaghan P., 870 Copani V., 41 Coppa M., 553, 668 Corton J., 465 Cosentino S.L., 41 Cougnon M., 175 Crespi R., 169 Crespo D., 258 Critchley C.N.R., 233 Crotty F.V., 267, 404 Cruz P., 112 Cuitiño M.J., 169 Čunderlík J., 370 Cupina B., 312 Cutullic E., 680 Czaban A., 830 Dabkevicius Z., 468 Dąbrowska-Zielińska K., 142 Dalgaard T., 97 Danaher M., 610 Davies J.W., 404 Davies T.E., 849 de Haan M.H.A., 662 de la Roza-Delgado B., 590, 622 de Renobales M., 671 De Vliegher A., 29, 743, 753, 773, 801, 864 de Wit J., 665 Decau M.L., 112 Decruyenaere V., 625, 641 Dehedin M., 559 Del Prado A., 61 Delaby L., 521, 728, 795 Delagarde R., 719, 728 Devincenzi T., 569 Dillon P., 695 Dohme-Meier F., 538, 593 Domingos T., 258 Donnison I.S., 465 Dörsch P., 94 Doyle E.M., 471 Dufey P.-A., 731 Dufrasne I., 181, 547, 801 Dufrêne M., 376 Dumont B., 756 Dungait J., 644 Durand J.-L., 112, 115, 122 Dželetović Ž., 291 Dzene I., 486 Dzybov D.S., 394 Egan M., 783, 789 Egan M.J., 677 Ehrhardt F., 75 Eickhoff B., 166 Elgersma A., 600, 674 Elliott C., 610 Elliott C.T., 556 Elsaesser M., 634, 638, 880 Eludoyin A., 239 Ene T.A.1, 397 Engel S., 880 Enjalbert J., 112 Enriquez-Hidalgo D., 783, 789, 792 Ergon Å., 823 Eschen R., 236 Escobar-Gutiérrez A., 112, 242 Escobar-Gutiérrez A.J., 115, 122 Faverjon L., 242 Fè D., 830 Ferlay A., 553 Ferreiro-Domínguez N., 264, 336 Fiala K., 391 Figl U., 418 Finn J.A., 109, 166 Finnan J., 438 Fleming C., 616 Florian C., 418 Fort F., 112 Fourtie S., 122 Frak E., 112 Franca A., 459 Frankow-Lindberg B.E., 15, 339, 342 Fraser M.D., 251, 261, 362, 465 Furtado S., 169 Fychan R., 267, 404, 651 Gaisler J., 324, 327 Gallard Y., 795 Galvin N., 616 Ganche E., 737 García-Ciudad A., 184, 288 García-Criado B., 184, 288 Garry B., 628, 737 Gastal F., 112 Genever E., 309 Geoghegan A., 279 Georganta A., 154 Ghesquière M., 112 Giaccone D., 668 Gilliland T., 783, 789, 792 Gilliland T.J., 833, 836, 870 Giostri A., 353, 356 Giostri A.F., 353 Girard M., 593 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 892 Goldringer I., 112 Golecký J., 553 Golińska B., 798, 816 Goliński P., 29, 142, 477, 759, 766, 798, 816 González A., 590 Görs S., 538 Gottardi S., 418 Gregis B., 786 Greve Pedersen M., 830 Griffith B., 239 Griffith G.W., 248 Griffith V., 279 Griffiths J.B., 233 Griggs T.C., 884 Grignard A., 641 Grinberg N.F., 826 Grogan D., 843, 870 Grosse Brinkhaus A., 593 Grygierzec B., 400 Guera K.C.S., 356 Gustavsson A.-M., 498 Gutiérrez R., 347 Gutmane I., 486 Hackett R., 583 Häfliger P., 321 Hahn M.A., 321 Hajšlová J., 495 Halasz A., 683 Hald A.B., 318 Harris P., 239 Harstad O.M., 553 Haş E.C., 887 Haughey E., 166 Hautier L., 376 Hayes R.C., 550 Hazard L., 112 Heard M.S., 254 Hecht J., 763 Hegarty M., 826 Heiermann M., 456 Hejcman M., 324, 327, 566 Hejcmanová P., 566 Helander M., 184 Helgadóttir Á., 15 Hennart S., 641 Hennessy D., 677, 737, 766, 783, 789, 792, 795, 810 Henry C.A., 541 Hensgen F., 477 Herremans S., 641 Herrmann A., 492, 535, 619 Herrmann C., 456 Herrmann K., 634, 638 Hetta M., 603 Hodkinson T., 438 Hoekstra N.J., 109, 166 Hofer D., 109, 166, 874 Hoffmann R., 282 Hoffstätter-Müncheberg M., 106 Hofstetter P., 658 Holshof G., 245, 544, 725 Horan B., 521 Hornick J.L., 181, 547 Hortová B., 421, 606 Huchon J-C., 559 Huguenin-Elie O., 695, 722, 756 Hujerová R., 327 Hummler T., 634 Humphreys M.W., 215 Huneau T., 559 Hunt S.R., 884 Husse S., 719, 722 Hutchings N.J., 97 Huws S.A., 550 Huyghe C., 29, 330, 743, 766, 807 Hyland J.J., 128 Iepema G., 665 Ineichen S., 680 Inostroza L., 858 Isselstein J., 103, 106, 160, 199, 367, 385, 743 Jakešová H., 495 Jamar D., 641, 776 Jan Ten Hagen P., 91 Jančová Ľ., 370 Jankowska J., 414 Jankowski K., 333, 407, 411 Janss L., 830 Jensen C.S., 830 Jensen J., 830 Jensen S.K., 600 Jilg T., 634, 638 Jiménez J.D., 587, 590, 622 Jokubauskaite I., 187 Jones D.L., 128, 273 Jones G., 743 Jones H., 651 Jones M., 438 Jørgensen M., 142 Jouany C., 112 Jovanovic M., 312 Juaristi A., 347 Julier B., 330 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 893 Julier Koubaiti B., 112 Kadziuliene Z., 468 Kaiser M., 385 Kanianska R., 370 Karabcová H., 391, 480 Karachristos C.N., 145 Karakosta C., 191 Karatassiou M., 157 Kasperczyk M., 400 Kauer K., 315 Kayser M., 106, 199 Kelly R., 852, 861 Kennedy E., 737, 877 Keres I., 315 Kerr S., 309 Kessler W., 3, 743 Keszthelyi S., 282 Kigongo J., 160 King R.D., 826 Kingston-Smith A.H., 849 Kipling R.P., 97 Kipparisides C., 462 Kirilov A., 647, 743 Kizeková M., 370 Klaas M., 438 Klotz C., 418 Klumpp K., 75 Kluß C., 492, 535 Kolczarek R., 407, 411 Kosmala A., 151 Kostopoulou P., 157 Koukoura Z., 462 Krstic D., 312 Krtková V., 495 Kryszak A., 333 Kryszak J., 333 Küchenmeister F., 103 Küchenmeister K., 103 Kurhak V., 453 Kusche D., 553 Kyriazopoulos A.P., 154 Laidna T., 315 Lalor S.T.J., 556 Lassalas J., 728 Látal O., 480 Lättemäe P., 780 Lausen P., 619 Lawrence D., 877 Lazaridou A., 157 Lazaridou M., 154, 157 Lea-Langton A.R., 465 Lebeau F., 631 Lee J.M., 840 Lee M.A., 541 Lee M.R.F., 644 Lenk I., 830 Lessire F., 181, 547 Lewandowski I., 429 Lewis E., 521, 556, 616, 628, 792 Liaudanskiene I., 187 Lind V., 563 Lindvall E., 498 Linsler D., 385 Litrico I., 112 Lloyd D.C., 855 Loges R., 125, 492 Loit E., 315 López-Sánchez A., 382 Louarn G., 112, 115, 122, 242 Lovatt A., 826 Lowe M., 350, 852, 867 Ludvíková V., 324 Ludwig B., 385 Ługowska M., 414 Lüscher A., 109, 166, 321, 722, 874 Lynch M.B., 677, 783, 789 Maass B.L., 160 MacAdam J.W., 884 Macfarlane A., 826 Macháč R., 502 Maczey N., 236 Makovníková J., 370 Maksimović N., 597 Mălinaş A., 298 Mandić V., 597 Marichal M. de J., 169 Marković, J., 291 Marley C.L., 267, 404, 651 Marshall A.H., 350, 855, 867 Martin B., 553 Martin C., 734 Martínez-Fernández A., 587, 590, 622 Maruşca T., 887 Matthews J.M., 309 McConnell D., 309 McConnell D.A., 541 McDonagh J., 836 McElhinney C., 610 McEniry J., 483 McEvoy M., 833, 836, 843 McHugh N., 843 Meek W.R., 254 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 894 Melis R., 41 Melis R.A.M., 459 Meripõld H., 780 Merten M., 106 Metges C.C., 538 Meuriot F., 112 Mičová P., 391, 480 Migault V., 242 Miller A., 273 Misselbrook T., 61 Mizen K.A., 852 Moakes S., 759, 763 Mocanu V., 397, 887 Mølmann J., 142 Moloney A.P., 509 Monahan F.J., 509 Moorby J.M., 686 Morales-Almaráz E., 622 Morel I., 731 Morgan E., 178 Morgan E.R., 100 Morgan S.A., 550 Morvan Bertrand A., 112 Mosquera-Losada M.R., 61, 148, 264, 336, 743 Müller J., 359 Mur L.A.J., 849 Murphy J.D., 483 Murphy J.P., 737 Murray P.J., 644 Mutimura M., 160 Muylle H., 864 Nabinger C., 569 Nagy G., 683 Nedělník J., 421, 495, 606 Newell Price J.P., 270 Niderkorn V., 734 Nielsen A.L., 318 Nilsdotter-Linde N., 743 Ninane M., 376 Nissen T., 318 Noacco V., 239 Nolan P., 471 Nolles J.E., 573 Novotná H., 495 Nüsse A.M., 385 O’Donovan G., 215 O’Donovan M., 279, 556, 616, 628, 737, 833, 836, 843 O’Kiely P., 100, 178, 471, 483, 583, 610, 870 Offermann F., 763 Oprea G., 887 Orr R.J., 644 Østrem L., 15 Owen D., 248 Păcurar F., 294, 298 Pál-Fám F., 282 Palicová J., 421, 606 Palmborg C., 498 Papanastasis V.P., 191 Papaspyropoulos K.G., 145 Pappas I.A., 145, 191, 462 Pardeller M., 163 Parente G., 743, 766, 813 Parissi Z.M., 154 Parol A., 315 Paszkowski E., 151 Patakas A., 157 Patanè C., 41 Paul B.K., 160 Pavlů L., 251, 324, 327 Pavlů V., 251, 324, 327, 566 Pawłowicz I., 151 Peach W.J., 236 Peel S., 379 Peeters A., 695, 743, 759, 763, 801 Peratoner G., 163, 418, 655 Perlikowski D., 151 Petrychenko V., 453 Peyraud J.L., 695, 728, 743, 766, 807 Phelan P., 100, 178 Philipsen A.P., 573, 662 Piaggio L., 169 Piccand V., 680 Pickert J., 613, 743 Picon Cochard C., 112 Piepho H.-P., 655 Pino M.T., 858 Planchon V., 776 Plantureux S., 743, 756, 759 Platace R., 285, 489 Plomp M., 544 Pocienė L., 468 Poetsch E.M., 139 Pontes L. da S., 353, 356 Popovici C.I., 302 Porfírio-da-Silva V., 353 Porqueddu C., 41, 459, 743 Pottier J., 112 Powell W.A., 826 Poyda A., 125 Prache S., 521, 569 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 895 Pramsohler M., 163 Prins W.H., 3 Prochnow A., 456 Proença V., 258 Prudhomme M.P., 112 Prunier A., 569 Pywell R.F., 254 Raave H., 315 Rakić, V., 291 Ramezani K., 239 Ramirez-Restrepo C., 625 Rancane S., 486 Randby Å.T., 603 Rataj D., 743 Raufer B., 429 Re G.A., 459 Rebuffo M., 169 Rees Stevens P., 849 Reheul D., 175 Reid M., 556 Reidy B., 680, 786 Řepková J., 495 Resch R., 139 Revello-Chion A., 668 Rietberg P., 665 Rigueiro-Rodríguez A., 148, 264, 336 Rivedal S., 94 Roberts D.J., 541 Roca-Fernández A.I., 728 Rognli O.A., 823 Roig S., 382 Roldán-Ruiz I., 864 Romano G., 655 Rosa García R., 261 Rose H., 100, 178 Ross A.B., 465 Rossi L., 840 Rossignol N., 756 Rotar I., 294, 298 Roulund N., 830 Ruiz de Gordoa J.C., 671 Ruth S.M. van, 674 Ruzić-Muslić D., 597 Rybak S., 453 Rzymowska Z., 414 Saetnan E., 97 Salas-Reyes I.G., 622 Sampoux J.-P., 112, 122 Samuil C., 302, 306 San Emeterio L., 347 San Miguel A., 382 Sandars D., 61 Sanderson R., 267, 404, 651 Sanna F., 459 Santos B.R.C., 353 Šarūnaitė L., 468 Schaffner U., 321 Schäufele R., 163 Schaumberger A., 139, 655 Schellekens A., 801 Schmid E., 731 Schmidt F., 456 Schmidt O., 509 Schori F., 538 Schulzová V., 495 Scimone M., 759 Scollan N.D., 97, 550 Scordia D., 41 Scott M.B., 404 Scullion J., 267 Seither M., 373 Selge A., 315 Seppänen M.M., 15 Seutin Y., 776 Shalloo L., 279, 843 Shaw R., 273 Sheehy-Skeffington M., 215 Shepherd A., 239 Simeonov M., 647 Simić A., 291, 312, 597 Sizer-Coverdale E., 350 Skøt K.P., 861 Skøt L., 826, 852, 861 Skrajna T., 414 Skrzyczyńska J., 414 Skuodiene R., 388 Šlepetienė A., 187, 468 Šlepetys J., 187, 468 Søegaard K., 15, 576, 600 Soldado A., 587, 590 Soney C., 731 Sosnowski J., 407, 411, 414 Soussana J.-F., 75 Stacey P., 583 Stafin G., 353 Stanišić N., 597 Starodubtseva A.M., 394 Stavarache M., 302, 306 Steinshamn H., 603 Stesele V., 486 Stevens C.J., 254 Stienezen M.W.J., 573, 804 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 896 Stilmant D., 376, 641, 776 Stolze M., 759 Stoycheva I., 647 Strejčková M., 421, 606 Strychalska A., 333 Stukonis V., 187 Sturite I., 94 Štýbnarová M., 480 Stypiński P., 364, 743 Sulas L., 459 Supek Š., 324 Surault F., 330 Suter D., 874 Suter M., 109, 166 Sweers W., 359 Szewczyk W., 400 Tabacco E., 668 Taff G., 142 Tamm S., 780 Tamm U., 780 Tampere M., 315 Taube F., 125, 492, 535, 619 Taugourdeau S., 756 Techow A., 492, 535 Teixeira R.F.M., 258 Testa G., 41 Thanner S., 538 Thompson J.B., 644 Thorhallsdottir A.G., 566 Thorne F., 759 Thumm U., 429, 880 Tilvikiene V., 468 Tomchuk D., 388 Tomić Z., 597 Tonn B., 367, 743 Tres A., 674 Tuominen E., 239 Tweed J.K.S., 550 Vaga M., 603 Valada T., 258 Valdivielso I., 671 Vale J.E., 362, 855, 867 Van den Pol-van Dasselaar A., 61, 97, 119, 573, 662, 695, 725, 743, 753, 766, 801, 804, 807, 810, 813, 816 Van Eekeren N., 665 Van Gils B., 753 Van Middelkoop J.C., 245, 544 Van Schooten H., 91 Van Waes C., 172 Vandecasteele B., 773 Vandermeulen S., 625 Vázquez de Aldana B.R., 184 Verbič J., 553 Vicente F., 587, 590, 622 Vidican R., 294, 298 Viiralt R., 315 Vintu V., 302, 306, 743 Virto M., 671 Visser M., 641 Vleugels T., 864 Volaire F., 112 Vuckovic S., 312 Wachendorf M, 477 Wagner M., 254 Walker K.J., 254 Wall D.M., 483 Wang Z.-Y., 438 Watson C.J., 556 Weise G., 613 Wheeler K., 236 Wheelton P., 91 Wigley S., 91 Wilkins R.J., 743 Wilkinson J.M., 91 Wilkinson T., 849 Williams A.P., 128, 248, 273 Wims C.M., 840 Wiśniewska-Kadżajan B., 407, 411 Withers P.J.A., 248 Wood B.L.2, 541 Wrage-Moennig N., 339 Wrage-Mönnig N., 103, 106, 160, 367 Wyss U., 593 Zabalgogeazcoa I., 184 Zaka S., 112, 115 Živanović I., 291 Zwicke M., 112 Zwierzykowski W., 151 Zwierzykowski Z., 151 Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands 897 Heavy metals in urban green cuttings Piepenschneider M.1, De Moor S.2, Hensgen F.1, Michels E.2, Meers E.2 and Wachendorf M.1 1 Department of Grassland Science and Renewable Plant Resources, Kassel University, Steinstrasse 19, 37213 Witzenhausen, Germany 2 Department of Applied Analytical and Physical Chemistry, Ghent University, Coupure Links 653, 9000 Gent, Belgium Corresponding author: meike.piepenschneider@uni-kassel.de Abstract Urban grass cuttings are a potential source for energy production from biomass. However, concerns about heavy metal contamination, amongst others, have hampered the practical implementation in the past. At 10 sites within the city of Kassel, samples of roadside verge green cuttings were taken from a 2-cut and a 4-cut regime to analyse the heavy metal content (Cd, Cr, Cu, Mn, Pb, Zn). Elemental concentrations were measured with ICP-OES after microwave digestion. Cd and Pb could barely be detected. Cr contents were lower than in common agricultural grass and Cu as well as Zn contents were slightly higher. Mn contents meet the agricultural reference value. All elements fall below the limiting value of DIN EN 14961-6 for herbaceous biofuels (Mn is not considered by the DIN). Differences between 2and 4-cut regimes were not statistically significant. Heavy metal contamination of municipal green cuttings is of little significance regarding energy recovery. Considering implementation opportunities of suitable technology is suggested. Keywords: heavy metals, energy recovery, roadside verges, green cuttings Introduction Municipal grass cuttings are regularly mulched or composted, thereby stressing public budgets as well as the climate by the release of greenhouse gases. Utilization of green cuttings in energy recovery is possible with adapted technology but municipalities shy away from taking advantage of this opportunity because of concerns about the heavy metal content of the material and related legal uncertainties. Heavy metals occur naturally in soils, serving as micro-nutrients to organisms, but they can be enriched by human activity (e.g. traffic, industry) to potentially toxic levels (Cachada et al., 2012). In the city of Kassel, grass cuttings (main grass species: Lolium perenne L. and Festuca rubra L.) from roadside verges were investigated for their heavy metal content (Cd, Cr, Cu, Mn, Pb, Zn). Grass cuttings are used for anaerobic digestion as well as for solid biofuel production (Prochnow et al., 2009). As advancing stage of vegetation usually causes decreasing biogas yields due to an increasing content of crude fibre, we used a 4-cut management for estimating the biogas potential (Prochnow et al., 2009). For the IFBB technology, which is adapted to fibre-rich material (Hensgen et al., 2011), a 2-cut management is preferable due to economic and ecological reasons. In German legislation, there are no limiting values for heavy metals in material for energy recovery but there are limiting values regarding fuel specifications. Here, DIN EN 14961-6 is used as a reference, which specifies non-woody pellets for non-industrial use (DIN, 2012). Materials and methods In the city of Kassel, 10 dispersed sites were selected and divided into a 2-cut regime and a 4cut regime with an area of 40 m2 each. In 2013, samples were taken in triplicate for each site and cutting regime, and cut at 5 cm height using scissors. Cutting was done at 6-week intervals for the 4-cut regime and 12-week intervals for the 2-cut regime. Additionally, silage was produced from material from the 2-cut regimes. At a minimum of 6 weeks after the harvest date, the silage was processed with the IFBB-technique, with a tap water addition of ratio 4 to Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands A1 1 and a mashing temperature of 40°C for 15 minutes. The material was dewatered by a screw press of stainless steel containing Cr (standard machinery configuration). Samples of the press cake (PC) and raw material were dried at 65°C. An open microwave digestion was conducted with 3.5 mL HNO3 and 3.5 mL H2O2 added to 0.25 g of dried plant material. The mixture was allowed to react at room temperature overnight (minimum 12 hours) and placed in a microwave oven (1200 W, 0 psi), diluted until 25 mL, and filtered on an acid-resistant filter. The filtrate was analysed for trace elements with Inductively Coupled Plasma – Optical Emission Spectrometry (ICP-OES: Varian vista MPX, Varian Palo Alto, USA). For statistical analysis Mann-Whitney U test was used, applying R (A Language and Environment for Statistical Computing, R Core TeamVienna, Austria, 2013). Results and discussion All of the investigated heavy metals (Cd, Cr, Cu, Mn, Pb and Zn) were detected in the samples. However, the contents of Cd and Pb were lower than the detection limits (0.4/4 mg kg-1 DM, respectively) in 99% of samples and therefore they fall below the limiting values of DIN EN 14961-6 and were close to values found by other authors in uncontaminated grass (Cd: 0.27 mg kg-1 DM, Pb: 3.3 mg kg-1 DM (Kabata-Pendias, 2011)). Mean Cr concentrations were 0.66±0.18 mg kg-1 DM in the 2-cut regime and 0.67±0.09 mg kg-1 DM in the 4-cut regime (Figure 1). Figure 1. Contents of Cr, Cu, Mn and Zn in mg kg-1 DM with median (solid line in box) and arithmetic mean (full dot). Broken lines indicate element contents of agricultural grass (Kabata-Pendias, 2011; Lindström et al., 2013) and solid lines indicate DIN limiting values (not shown for Cr (limiting value: 50 mg kg-1 DM) and Mn (not considered by the DIN)). In 53 of 175 samples, the Cr content was below the detection limit and was therefore not considered in statistical analysis. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands A2 Both were below the contents found in common grass of 1.1 mg kg-1 DM (Kabata-Pendias, 2011) and about 100 times lower than the DIN limiting value of 50 mg kg-1 DM. The Cr content in the PC was higher (3±1.37 mg kg-1 DM) than the usual content of non-urban grasses but still below the DIN value. Similarly, concentrations of Cu and Zn in municipal grass samples of both cutting regimes, and of the PC, exceeded contents of uncontaminated grass. The concentrations of Cu were 7.45±1.31, 7.95±1.31 and 10.67±2.79 mg kg-1 DM in the 2-cut and 4-cut regime and the PC, respectively, compared to a common value of uncontaminated grass of 3.1 mg kg-1 DM (Lindström et al., 2013). Zn contents of 38.26±7.34, 46.59±21.54 and 28.65±10.31 mg kg-1 DM were determined in the 2-cut and 4-cut regime and the PC, respectively, while a natural content 17.7 mg kg-1 DM can be assumed (Lindström et al., 2013). Nevertheless, concentrations of Cu and Zn fall below the DIN limiting values of 20 and 100 mg kg-1 DM, respectively. Concentrations of Mn in the raw material are in the range of common agricultural grass, with the medians below and the means of 106.05±104.71 and 89.82±90.47 mg kg-1 DM in the 2-cut and 4-cut regime, respectively, above the value of 70 mg kg-1 DM (Kabata-Pendias, 2011). In the PC both values, median and mean (62.83±47.2 mg kg-1 DM), fall below this value. No significant differences in heavy metal concentrations of the treatments were found. Low contents of heavy metals in roadside material have also been observed for Cu, Pb and Zn in Italian ryegrass (Lolium multiflorum Lam.) in Finland (Yläranta, 1994). However, uptake of heavy metals is highly dependent on plant species and soil conditions. Conclusion The heavy metal content (Cd, Cr, Cu, Mn, Pb, Zn) of green cuttings from roadside grass in the city of Kassel is low, in general. Even considering a certain amount of contamination from the machinery, the contents of the investigated elements fall below the DIN limiting values. The results therefore provide no reason for not utilising the material in energy recovery systems. Thus, it is suggested that the economic and ecological conditions, which are necessary to successfully maintain green areas for energy recovery purposes, should be explored. Acknowledgements The authors are very thankful to the EU for co-financing the COMBINE project through the Interreg IV B regional development fund. References Cachada A., Pereira M.E., Ferreira da Silva E. and Duarte A. (2012) Sources of potentially toxic elements and organic pollutants in an urban area subjected to an industrial impact. Environmental Monitoring and Assessment 184, 15-32. DIN Deutsches Institut für Normung e. V. (2012) Solid biofuels - Fuel specifications and classes - Part 6: Nonwoody pellets for non-industrial use, DIN EN 14961-6, Beuth, Berlin. Hensgen F., Richter F. and Wachendorf M. (2011) Integrated generation of solid fuel and biogas from green cut material from landscape conservation and private households. Bioresource Technology 102, 10441–10450. Kabata-Pendias A. (2011) Trace elements in soils and plants, 4th. edn. CRC Press, Boca Raton, 520 pp. Lindström B.E.M., Frankow-Lindberg B.E., Dahlin A.S., Wivstad M. and Watson C.A. (2013) Micronutrient concentrations in common and novel forage species and varieties grown on two contrasting soils. Grass and Forage Science 68, 427–436. Prochnow A., Heiermann M., Plöchl M., Linke B., Idler C., Amon T. and Hobbs P. (2009) Bioenergy from permanent grassland – A review: 1. Biogas. Bioresource Technology, 100, 4931–4944. Yläranta T. (1994) Effect of road traffic on heavy metal concentrations of plants. Agricultural Science in Finland 4, 35–48. Grassland Science in Europe, Vol. 19 - EGF at 50: the Future of European Grasslands A3

RELATED PAPERS

RELATED TOPICS

Log In

Tree and pasture productivity in Pseudotsuga menziesii (Mirb.) Franco silvopastoral systems fertilized with sewage sludge

Tree and pasture productivity in Pseudotsuga menziesii (Mirb.) Franco silvopastoral systems fertilized with sewage sludge

Related Papers

RELATED PAPERS

RELATED TOPICS