613 A kashiwa oak (Quercus dentata) tree-ring width chronology from northern coastal Hokkaido, Japan1 Rosanne D. D’Arrigo, David K. Yamaguchi, Gregory C. Wiles, Gordon C. Jacoby, Akira Osawa, and David M. Lawrence Abstract: Few paleoclimatic records exist for Hokkaido, the northernmost, coldest, and least environmentally disturbed of Japan’s main islands. Here, we present a chronology for kashiwa oak (Quercus dentata Thunb.) at Lake Saroma on the north central coast of Hokkaido. This record (1710–1993) was compared with meteorological data from Abashiri, the longest existing station (1899–1989) near the site. Growth correlates positively with current June–July temperatures, but negatively with prior-year August temperatures. Growth also correlates positively with prior November, December, and March temperatures and with prior February precipitation. Results suggest that these oaks grow best during warm, early summers or after warm, snowy winters. The lowest growth year is 1784, following the “year without a summer” elsewhere in Japan. This record contributes to a growing tree-ring network for the North Pacific rim, including sites in Japan, Kamchatka, Korea, Alaska, and Canada. It also adds a new species to those considered useful for dendroclimatology. Résumé : Il existe peu de données paléoclimatologiques pour l’île d’Hokkaido, la plus septentrionale, la plus froide et la moins perturbée, du point de vue environnemental, parmi les plus importantes l’îles japonaises. Cette étude présente un chronologie pour le chêne dentée (Quercus dentata Thunb.) près du lac Saroma qui est situé sur la côte nord dans la partie centrale de l’île d’Hokkaido. Ces données (1710–1993) furent comparées aux données météorologiques provenant d’Abashiri, la plus vieille station météorologique (1899–1989) à proximité du site. La croissance est corrélée positivement avec les températures des mois de juin et juillet de l’année en cours et négativement avec les températures du mois d’août de l’année précédente. La croissance est également positivement corrélée avec les températures des mois de novembre, décembre et mars de l’année précédente et avec les précipitations du mois de février de l’année précédente. Les résultats suggèrent que ces chênes croissent de façon optimale lorsque le début de l’été est chaud ou suite à un hiver doux et avec beaucoup de neige. La croissance a été la plus faible en 1784, l’année après celle qu’on a qualifiée d’année sans été ailleurs au Japon. Ces données s’ajoutent à un réseau de cernes annuels en pleine expansion autour du Pacifique Nord, incluant des sites au Japon, au Kamchatka, en Corée, en Alaska et au Canada. Elles ajoutent une nouvelle espèce d’arbres à celles qu’on considère comme utiles en dendroclimatologie. [Traduit par la Rédaction] Introduction Tree rings have become increasingly valuable in disclosing the long-term dynamics of climate in many areas of the globe (e.g., Cook et al. 1996; Lara and Villalba 1993). Such studies are particularly important in regions such as the North Pacific where instrumental records are short and paleoclimatic records Received March 28, 1996. Accepted December 17, 1996. R.D. D’Arrigo,2 G.C. Wiles, G.C. Jacoby, and D.M. Lawrence.3 Tree-Ring Laboratory, Lamont–Doherty Earth Observatory, Palisades, NY 10964-8000, U.S.A. D.K. Yamaguchi.4 Forestry and Forest Products Research Institute, 7 Hitsujigaoka, Toyohira, Sapporo 062, Japan. A. Osawa. Faculty of Intercultural Communication, Environmental Studies Laboratory, Ryokoku University, Seta-Ohe, Shiga 520-21, Japan. 1 2 3 4 Lamont–Doherty Earth Observatory Publication No. 5654. Author to whom all correspondence should be addressed. Present address: 26 Columbus Drive, Tenafly, NJ 07670, U.S.A. Present address: Xylometric, 5020 27th Avenue South, Seattle, WA 98108, U.S.A. Can. J. For. Res. 27: 613–617 (1997) are limited (e.g., Hughes 1991). Lately, coverage across this vast region has improved with reports of tree-ring records from coastal Alaska and Canada, Kamchatka, and Korea (e.g., Briffa et al. 1992; Park 1994; Wiles et al. 1996; Gostev et al. 1996). However, a major gap in present understanding is the shortage of tree-ring records from Japan. Relatively little has been published on pre-20th century climates in Japan, especially outside Honshu, Japan’s densely populated main island. Dendroclimatic research on Honshu before the late 1980s was largely exploratory. Sizhong et al. (1982) reviewed some early studies, mostly analyses of single trees. Kojo (1987) investigated the climatic response of Japanese cedar (Sugi, Cryptomeria japonica (L.f.) D. Don) at Ashiu, western Honshu. He found that moisture stress during current spring–summer and prior summer–autumn limits radial growth. A reconstruction of winter temperatures since A.D. 1177 was developed from Hinoki cypress (Chamaecyparis obtusa (Sieb. et Zucc.) Endl.) at Mount Ontake, central Honshu (Sweda 1994). This report describes a tree-ring record from Hokkaido, the northernmost main island of Japan (Fig. 1). Hokkaido is perhaps the most promising of the principal islands for sampling due to its late settlement by modern Japanese, and thus lower © 1997 NRC Canada 614 Fig. 1. Main islands of northern Japan showing localities and currents mentioned in text. Inset: map of Hokkaido. The triangle shows the Lake Saroma tree-ring site, a few metres above sea level. Can. J. For. Res. Vol. 27, 1997 Fig. 2. Lake Saroma oak ring-width chronology for 1710–1993 and sample depth in number of cores. The arrow marks 1784, the year of least growth. was positively linked to warm, dry summers. Summers of the 1780s were found to be drier and 1830s summers wetter than 20th century averages. Density of Yezo spruce (Picea jezoensis (Sieb. et Zucc.) Carr.) in Tomakomai Experimental Forest (100–200 m, Fig. 1) shows positive responses to current June–July and prior December–January temperatures and current August–September precipitation (Kobayashi et al. 1995). Both Saghalin and Yezo spruce are subarctic–subalpine species (Japan Forest Technical Association 1964; Horikawa 1972; Shidei 1974). Climate of Hokkaido and northern Japan Monsoonal reversal of wind direction dominates the climate of the northwestern Pacific. In winter, the Siberian High and Aleutian Low pressure systems drive cold, dry northwesterly to northerly winds across the region (Fukui 1977). In summer the circulation reverses, and southerly to southeasterly winds from the Hawaiian High carry warm, moist air masses of monsoonal air northward (Martyn 1992). The Abashiri station illustrates this monsoonal climate with temperatures remaining below freezing from December–March. Summers are wet, with highest precipitation in September. Abashiri monthly mean temperatures and total precipitation are generally only weakly intercorrelated. Hokkaido’s climate is also influenced by ocean currents (Fig. 1), its proximity to the sea, and orographic effects. The Tsushima Current warms and moistens the cold, dry winter air of the Siberian High over the Japan Sea, contributing snowfall to Hokkaido. The Sakhalin Current brings cold water and sea ice to northeast Hokkaido in winter, resulting in freezing coastal waters during January–March (e.g., Tabata 1972a, 1972b). Central north–south mountain ranges limit winter precipitation in eastern Hokkaido, and resulting clear skies contribute to severe cold there. Tree-ring data degree of disturbance. Its northern location provides a sufficiently cold climate to limit annual tree growth. In targeting Hokkaido, we build on the construction of a density record dating to the 1700s for Saghalin spruce (Picea glehnii (Fr. Schm.) Mast.) at Teshio Experimental Forest (300 m) and its use in reconstructing Asahikawa August precipitation (Fig. 1; Yasue et al. 1994, 1995). This density record Paired cores were collected from mature Quercus dentata Thunb., a deciduous hardwood named “kashiwa” oak (Horikawa 1972; Ohwi 1984). The trees are on the western sand spit separating Lake Saroma (a tidal lagoon) from the sea (Fig. 1). This area falls under the coastal forest classification of Shidei (1974), characterized by strong winds and high evapotranspiration rates. Kashiwa oak are typically tall (up to 15 m) canopy trees, with a straight, single-trunk growth form © 1997 NRC Canada 615 Notes Fig. 3. Correlations between monthly climate variables at Abashiri (43 m) and Lake Saroma oak growth during 1891–1989. All correlations are for April preceding the year of growth to September ending each growth season. One and two asterisks are significant at the 95 and 99% levels, respectively, in two-tailed tests of prewhitened (Box et al. 1994) climate and tree-ring data. Vertical lines divide approximate prior, dormant, and current physiologically active seasons. size drops to nine cores from seven trees. Lowest growth periods occur in the early 1780s and around 1860, with aboveaverage growth intervals about 1795, 1825, and 1850. Climate analyses The chronology was correlated with monthly temperature and precipitation data from nearby Abashiri (Figs. 1 and 3). It correlates positively with June–July temperatures of the current growing season, but negatively with prior August temperatures. Growth also correlates positively with temperatures during November–December and March. The only significant correlation with precipitation is a positive one with prior February precipitation. All the above-mentioned correlations are significant at the 95% level in two-tailed tests; November, March, and June temperatures are significant at the 99% level. Correlations are based on prewhitened tree-ring and climate data for 1891–1989. Discussion and trunks to 60 cm diameter. In Hokkaido, this species flowers in June and its acorns ripen in October (Japan Forest Technical Association 1964). Kashiwa occurs in Japan, Korea, China, Taiwan, and Mongolia (Horikawa 1972). It is one of the most prevalent trees below 750 m in northern Japan, near the northern end of its distribution (Horikawa 1972). The fixed dunes at the site are typical of the nutrient-poor sandy soil in which kashiwa oak commonly grows (Japan Forest Technical Association 1964; Numata 1974). Cores were dated and processed using conventional techniques (Fritts 1976; Cook and Kairiukstis 1990). Successful cross dating is indicated by a series intercorrelation of 0.602 and average mean sensitivity of 0.254. Ring-width measurements were standardized by fitting smoothing splines with a frequency-response cutoff of two thirds the length of each series. This was appropriate due to nonsynchronous zones of suppression and release, probably related to competition and disturbance (e.g., Graybill et al. 1982; Cook and Briffa 1990). The chronology dates from 1710 to 1993 and consists of 31 cores from 19 trees (Fig. 2). Although several cores date back to the 1690s, we truncated the series at 1710 when the sample Results suggest that optimal growth depends on warm, early summers and warm, wet winters. The apparent importance of warm temperatures during the first growing season months (June–July) is consistent with the cool Hokkaido climate and location of Lake Saroma towards the northern limit of the species. Current August temperatures may be less important because by then the ring is largely formed. Also, the principal limiting factor late in the season probably shifts from cool temperatures to inadequate precipitation during a time of active transpiration. Significant positive correlations with dormant season conditions are harder to explain. Warm Marches may contribute to longer growing seasons. The trees might also benefit from early spring break-up of sea ice, which would facilitate aerial transport of sea nutrients to trees, as has been found to be beneficial to growth elsewhere (Gorham 1961; Art et al. 1974). November–December temperatures may contribute to carryover effects even in deciduous species (Kim and Siccama 1987). Cold November–Decembers might be harmful to roots because ground snow cover is still thin. Roots are not as cold hardy as stems, and can be injured or killed by cold that does not harm the aboveground tree (Kramer and Kozlowski 1979). Greater winter root damage occurs in sandy or dry soils, as at the Lake Saroma site. February snowfall might play a role in the above interpretation by insulating roots from low winter air temperatures. Abashiri snow depth during this coldest and driest month is highly variable, and can be shallow (range 14–106 cm during 1892–1980, mean 53 cm, standard deviation 21 cm; National Weather Bureau of Japan 1982). Soil temperature and frost heave data from Mombetsu (1966–1967, 1968–1969) suggest that local soils remain frozen from mid-December to late April (to >60 cm depth in February and March). They are also subject to frost heave which can shear roots, especially from January to March (Kinoshita et al. 1967, 1969). Snowy or mild winters with less root damage would aid growth during the ensuing season. Climate–growth linkages for dormant-season months have been documented in hardwoods elsewhere. Oak ring widths © 1997 NRC Canada 616 were positively related to prior October temperatures in Britain (Hughes et al. 1978) and winter temperatures in Germany and Austria (Eckstein and Frisse 1982). American beech (Fagus grandifolia Ehrh.) have a significant positive relationship with previous-year September and November temperatures, although other species indicate weaker to negative linkages (Kim and Siccama 1987). Alaskan balsam poplar (Populus balsamifera L.) show positive correlations with prior September–November temperatures from the 1920s to 1950s, after which warming reduced temperature stress (Jacoby et al. 1996). Thus, dormantseason climate – tree growth relationships similar to those at Lake Saroma appear widespread. Low growth in the early 1780s at Lake Saroma (Fig. 3) coincides with other records indicating cold conditions in central Japan at this time (e.g., Mikami 1988). The lowest growth year in the record is 1784, following 1783, the “year without a summer” elsewhere in Japan (Mikami and Tsukamura 1992). This episode may partially reflect atmospheric cooling caused by major volcanic eruptions at Laki, Iceland, and Asama, Japan (Wood 1992; Bradley and Jones 1992). Summary We have described a kashiwa oak tree-ring width chronology, one of few such records to be developed for Hokkaido, the least environmentally disturbed main island of Japan. The trees appear to grow best during long, warm growing seasons preceded by warm, wet winters. This record builds on existing tree-ring coverage and efforts to reconstruct the pre-20th century climate of the data-sparse North Pacific region. It also adds a new species to those known to be useful for dendroclimatology. Acknowledgments We thank the Yuubetsu national forest staff for permission to sample the site. We thank Y. Begin, O. Kobayashi, W. Park, and K. Yasue for comments that improved the manuscript. This project was funded by NOAA Office of Global Programs grant NA56GPO235. Fieldwork was sponsored by Hokkaido Television Broadcasting, Sapporo. Yamaguchi was supported by a Science and Technology Agency of Japan fellowship. References Art, H.W., Bormann, F.H., Voigt, G.K., and Woodwell, G.M. 1974. Barrier island forest ecosystem: the role of meteorological inputs. Science (Washington, D.C.), 184: 60–62. Box, G.E.P., Jenkins, J.M., and Reinsel, G.C. 1994. Time series analysis, forecasting and control. 3rd ed. Prentice Hall, Englewood Cliffs, N.J. Bradley, R.S., and Jones, P.D. 1992. Records of explosive volcanic eruptions over the last 500 years. In Climate since A.D. 1500. Edited by R.S. Bradley and P.D. Jones. Routledge, London. pp. 606–622. Briffa, K.R., Jones, P.D., and Schweingruber, F.H. 1992. Tree-ring reconstructions of summer temperature patterns across western North America since 1600. J. Clim. 5: 735–754. Cook, E.R., and Briffa, K.R. 1990. Data analysis. In Methods of dendrochronology, applications in the environmental sciences. Edited by E.R. Cook and L.A. Kairiukstis. Kluwer Academic Publishers, Boston, Mass. pp. 97–162. Can. J. For. Res. Vol. 27, 1997 Cook, E.R., and Kairiukstis, L.A. (Editors). 1990. Methods of dendrochronology, applications in the environmental sciences. Kluwer Academic Publishers, Boston, Mass. Cook, E.R., Buckley, B.M., and D’Arrigo, R.D. 1996. Interdecadal climate variability in the Southern Hemisphere: evidence from Tasmanian tree rings over the past three millenia. In Climate variations and forcing mechanisms of the last 2,000 years, 3–7 Oct. 1994, Il Ciocco, Italy. Edited by P.D. Jones, R.S. Bradley, and J. Jouzel. NATO ASI Ser. I41: 109–124. Eckstein, D., and Frisse, E. 1982. The influence of temperature and precipitation on vessel area and ring width of oak and beech. In Climate from tree rings. Edited by M.K. Hughes, P.M. Kelly, J.R. Pilcher, and V.C. LaMarche, Jr. Cambridge University Press, Cambridge, England. p. 12. Fritts, H.C. 1976. Tree rings and climate. Academic Press, New York. Fukui, E. (Editor). 1977. The climate of Japan. Elsevier Science Publishers B.V., Amsterdam, Netherlands. Gorham, E. 1961. Factors influencing supply of major ions to inland waters, with special reference to the atmosphere. Bull. Geol. Soc. Am. 72: 795–840. Gostev, M., Wiles, G., D’Arrigo, R., Jacoby, G., and Khomentovsky, P. 1996. Early summer temperatures since 1670 AD reconstructed based on a Siberian larch tree-ring width chronology. Can. J. For. Res. 26: 2048–2052. Graybill, D.A., Hughes, M.K., Aniol, R.W., and Schmidt, B. 1982. Chronology development and analysis. In Climate from tree rings. Edited by M.K. Hughes, P.M. Kelly, J.R. Pilcher, and V.C. LaMarche, Jr. Cambridge University Press, Cambridge, England. pp. 21–31. Horikawa, Y. 1972. Atlas of the Japanese flora, an introduction to plant sociology of east Asia. Gakken Co., Tokyo. Hughes, M.K. 1991. The tree-ring record. In Global changes of the past. Edited by R.S. Bradley. University Corporation for Atmospheric Research, Boulder, Colo. pp. 117–137. Hughes, M.K., Gray, B., Pilcher, J., Baillie, M., and Leggett, P. 1978. Climatic signals in British Isles tree-ring chronologies. Nature (London), 272: 605–606. Jacoby, G.C., Wiles, G.C., and D’Arrigo, R.D. 1996. Alaskan dendroclimatic variations for the past 300 years along a north–south transect. Radiocarbon, 235–248. Japan Forest Technical Association (Editor). 1964. Illustrated important forest trees of Japan. Chikyu Shuppan Co., Tokyo. Kim, E., and Siccama, T.G. 1987. The influence of temperature and soil moisture on the radial growth of northern hardwood tree species at Hubbard Brook Experimental Forest, New Hampshire, USA. In Proceedings, International Symposium on Ecological Aspects of Tree-Ring Analysis. Edited by G.C. Jacoby and J.W. Hornbeck. U.S. Dep. Energy Publ. CONF-8608144-26-37. pp. 26–37. Kinoshita, S., Suzuki, Y., Horiguchi, K., Tanuma, K., and Aota, M. 1967. Frost heave in Mombetsu (1966–1967). Low Temp. Sci. (Univ. Hokkaido) Ser. A, 25: 229–232. [In Japanese.] Kinoshita, S., Suzuki, Y., Horiguchi, K., Tanuma, K., Aota, M., and Ono, T. 1969. Frost heave in Kitami and Mombetsu (1968–1969). Low Temp. Sci. Ser. A, 27: 279–293. [In Japanese, English summary.] Kobayashi, O., Fukazawa, K., and Ohtani, J. 1995. Densitometric analysis on tree rings of non-vital conifers growing in Hokkaido, Japan. In Tree rings, from the past to the future. Proceedings, International Workshop on Asian and Pacific Dendrochronology. Forestry and Forest Products Research Institute, Tsukuba, Japan, Scientific Meeting Report 1. pp. 146–150. Kojo, Y. 1987. A dendrochronological study of Cryptomeria japonica in Japan. Tree-Ring Bull. 47: 1–21. Kramer, P.J., and Kozlowski, T.T. 1979. Physiology of woody plants. Academic Press, New York. Lara, A., and Villalba, R. 1993. A 3620-year temperature record from © 1997 NRC Canada Notes Fitzroya cupressoides tree rings in southern South America. Science (Washington, D.C.), 260: 1104–1106. Martyn, D. 1992. Climate of the world. Elsevier Science Publishing Co., Inc., New York, and Kodansha Ltd., Tokyo. Mikami, T. 1988. Climatic reconstruction in historical times based on weather records. Geogr. Rev. Jpn. 61B: 14–22. Mikami, T., and Tsukamura, Y. 1992. The climate of Japan in 1816 as compared with an extremely cool summer climate in 1783. In The year without a summer? World climate in 1816. Edited by C.R. Harington. Canadian Museum of Nature, Ottawa, Ont. pp. 462–476. National Weather Bureau of Japan. 1982. Climate of Hokkaido. Hokkaido Branch, Sapporo. [In Japanese.] Numata, M. (Editor). 1974. The flora and vegetation of Japan. Elsevier Science Publishing Co., Inc., New York. Ohwi, J. 1984. Flora of Japan. Edited by F.G. Meyer and E.H.Walker. Smithsonian Institution, Washington, D.C. Park, W.-K. 1994. Tree-ring networks for detecting global warming in Korea. J. Korean For. Energy, 14: 80–87. Shidei, T. 1974. Forest vegetation zones. In The flora and vegetation of Japan. Edited by M. Numata. Elsevier Science Publishing Co., Inc., New York. pp. 87–124. Sizhong, Z., Xiangding, W., and Zhenyao, L. 1982. Status of dendroclimatology: Asia. In Climate from tree rings. Edited by M.K. Hughes, P.M. Kelly, J.R. Pilcher, and V.C. LaMarche, Jr. Cambridge University Press, Cambridge, England. pp. 155–158. Sweda, T. 1994. Tree-ring reconstruction of the last sub-millennial climate of central Japan. Terr. Atmos. Ocean. Sci. 5: 431–442. 617 Tabata, S. 1972a. Radar network for drift ice observation in Hokkaido. Sea Ice Conference Proceedings, Reykjavik. pp. 67–71. [Reprinted in Collected reports of Sea Ice Research Laboratory. Vol. 1. Sea Ice Research Lab, The Institute of Low Temperature Science. 1973. Hokkaido University, Hokkaido, Japan.] Tabata, S. 1972b. Observations of deformation and movement of ice field with the sea ice radar network. Sea Ice Conference Proceedings, Reykjavik. pp. 72–79. [Reprinted in Collected reports of Sea Ice Research Laboratory. Vol. 1. Sea Ice Research Lab, The Institute of Low Temperature Science. 1973. Hokkaido University, Hokkaido, Japan.] Wiles, G.C., D’Arrigo, R.D., and Jacoby, G.C. 1996. Temperature changes along the Gulf of Alaska and the Pacific Northwest coast modeled from coastal tree rings. Can. J. For. Res. 26: 474–481. Wood, C.A. 1992. Climatic effects of the 1783 Laki eruption. In The year without a summer? World climate in 1816. Edited by C.R. Harington. Canadian Museum of Nature, Ottawa, Ont. pp. 58–77. Yasue, K., Funada, R., Noda, M., and Fukazawa, K. 1994. Dendroclimatological study of Picea glehnii growing in the Teshio Experimental Forest of Hokkaido University. Res. Bull. Hokkaido Univ. For. 51: 243–266. [In Japanese, English abstract.] Yasue, K., Funada, R., Fukazawa, K., and Ohtani, J. 1995. Reconstruction of summer temperature and precipitation back to A.D. 1700 using tree-ring maximum density of Picea glehnii. IGBPPAGES/PEP-II Symposium on Paleoclimate and Environmental Variability in Austral–Asian Transect During the Past 2000 Years, 28 Nov. – 1 Dec. 1995, Nagoya University, Nagoya, Japan. p. 33. [Abstract.] © 1997 NRC Canada