Blackwell Science, LtdOxford, UKBOJBotanical Journal of the Linnean Society0024-4074The Linnean Society of London, 2003? 2003 142? 8391 Original Article MORPHOLOGY OF PINUS ULIGINOSA K. BORATYŃSKA, A. BORATY SKI and A. LEWANDOWSKI Botanical Journal of the Linnean Society, 2003, 142, 83–91. With 4 figures Morphology of Pinus uliginosa (Pinaceae) needles from populations exposed to and isolated from the direct influence of Pinus sylvestris KRYSTYNA BORATY N¢ SKA*, ADAM BORATY N¢ SKI and ANDRZEJ LEWANDOWSKI Polish Academy of Sciences, Institute of Dendrology, Parkowa 5, 62-035 Kornik, Poland Received June 2002; accepted for publication December 2002 Eleven needle traits of two contrasting populations of Pinus uliginosa (P. mugo complex), one located in a Pinus sylvestris forest, the other isolated from the influence of this species, were compared to quantify the degree of hybridization. Statistically significant differences between the studied populations were found, but both of them differed by a similar degree from P. sylvestris. Surprisingly, the P. uliginosa population surrounded by the P. sylvestris forest, which was expected to consist of individuals with intermediate traits, appeared to be more closely related to P. mugo than those isolated from the direct influence of P. sylvestris. Large differences in the distances between the analysed pair of populations of P. uliginosa and P. sylvestris have not influenced the morphological traits of the species. Gene flow from P. sylvestris to P. uliginosa is possible, but is likely to be small and does not differ compared with populations of P. uliginosa. The decline observed during the last three decades in both populations of the species has not influenced their needle morphological traits. © 2003 The Linnean Society of London, Botanical Journal of the Linnean Society, 2003, 142, 83–91. ADDITIONAL KEYWORDS: biometric analysis – gene flow – plant conservation – Poland – Sudety mountains – variability. The relationships among Pinus uliginosa Neum. ex Wimm (= P. pumilio Haenke) and other pine taxa of the section Sylvestres, namely, P. mugo Turra, P. sylvestris L. and P. uncinata Ramond ex DC., have been the subject of several taxonomic studies in the last few decades. In terms of morphology, P. uliginosa is intermediate between P. mugo and P. sylvestris (Staszkiewicz 1985, 1993a,b) and/or between P. mugo and P uncinata (Staszkiewicz & Tyszkiewicz, 1972; Krzakowa, Naganowska & Bobowicz, 1984; Christensen, 1978a,b; Siedlewska & Prus-G l¢owacki, 1994, 1995). P. uliginosa is lately regarded as a stabilized hybrid taxon resulting from ancient crosspollination of the two pairs of species mention above (see Christensen, 1987a for a recent taxonomic revision, and Siedlewska & Prus-G l¢owacki, 1995, Christensen & Dar, 1997; Lauranson-Broyer, Krzakowa & Lebreton, 1997; Prus-G l¢owacki, Bujas & Ratynska, 1998; Lewandowski, Boraty n¢ ski & Mejnartowicz, 2000). *Corresponding author. E-mail: borkrys@man.poznan.pl It has been proposed that Pinus uliginosa and other taxa from the P. mugo complex (see Christensen, 1987a) are subject to genetic erosion because of an inflow of genes from P. sylvestris. Hybridization is expected especially when populations of these taxa grow near each other (see Christensen, 1987a for detailed literature review and Marcet, 1967; Staszkiewicz & Tyszkiewicz, 1969, 1972; Szweykowski, 1969; Szweykowski & Bobowicz, 1983; Staszkiewicz, 1985; Bobowicz, 1990; Staszkiewicz, 1993a,b, 1994; Siedlewska & G l¢owacki, 1994, 1995). The hybridization intensity among pine species has been found to be weak (Christensen, 1987b; Filppula, Szmidt & Savolainen, 1992; Christensen & Dar, 1997), however this does not exclude the possibility of gene flow, especially from populations comprising numerous individuals towards populations of restricted numbers of individuals. Testing this hypothesis using P. uliginosa from two populations was the main aim of the present study. Pinus mugo and P. sylvestris were used and comparisons were made between the three taxa. © 2003 The Linnean Society of London, Botanical Journal of the Linnean Society, 2003, 142, 83–91 83 84 K. BORATY N¢ SKA ET AL. Over the last three decades a large reduction in the number of individuals in populations of Pinus uliginosa in south-western Poland has been observed (Boraty n¢ ski, 1994; Danielewicz & Zieli n¢ ski, 2000; Gol¢ a˛b, 1999). Because the examined populations of this species from We˛ gliniec and Batorów were earlier characterized biometrically (Szweykowski, 1969; Staszkiewicz & Tyszkiewicz, 1969, 1972), it was also our aim to determine if the decline of these populations affected the average values of their needle traits. MATERIAL AND METHODS Two populations of Pinus uliginosa, one from We˛ gliniec and one from Batorów were used for the present study. The first population has been surrounded by P. sylvestris forest for perhaps thousands of years, while the second is much more isolated from direct contact with P. sylvestris for at least the last few generations. Consequently, the population of P. uliginosa at We˛ gliniec has been subject to gene flow from adjacent P. sylvestris for a much longer period of time and in the greater intensity than the population at Batorów. If gene flow has occurred between the species, this influence was expected to be visible in the needle traits of P. uliginosa. Phenological observations conducted in 1999–2001 in We˛ gliniec and Batorów confirmed the possibility of cross-pollination between P. uliginosa and P. sylvestris. Plant material for the study was collected in both above mentioned populations of Pinus uliginosa. Comparative material came from one population of P. mugo and two populations of P. sylvestris, one from the forest surrounding the reserve in We˛ gliniec and the other from Mt Szczeliniec Wielki (Table 1). Each population was represented by at least 30 individuals, and ten 2-year-old dwarf shoots (20 needles) were collected from each individual. The analysis involved 11 measured traits and four traits derived from calculations (Table 2). Needle Table 1. Location of studied populations of Pinus uliginosa, Pinus sylvestris and Pinus mugo Number of of sample Taxon Location Acronym Longitude E Latitude N Number of individuals 1 Pinus uliginosa P. uliginosa W 51∞17¢50≤ 15∞14¢20≤ 52 2 Pinus uliginosa P. uliginosa B 50∞27¢36≤ 16∞15¢25≤ 50 3 Pinus sylvestris P. sylvestris W 51∞17¢50≤ 15∞14¢20≤ 34 4 Pinus sylvestris P. sylvestris Sz 50∞28¢55≤ 16∞16¢46≤ 30 5 Pinus mugo Reserve near We˛gliniec (Lower Silesia Forest) Large Batorów Peatland (Sto l¢ owe Mts.) Forest near We˛gliniec (Lower Silesia Forest) Szczeliniec Wielki Mt. (Sto l¢ owe Mts.) Równia below S¢ nieżka (Karkonosze Mts.) P. mugo R 50∞44¢45≤ 16∞42¢50≤ 30 Table 2. Needle traits analysed No. Trait Precision 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Needle length Number of stomatal rows on convex (abaxial) side of needle Number of stomatal rows on flat (adaxial) side of needle Number of stomata on 2-mm long section of needle, on convex (abaxial) side Number of stomata on 2-mm long section of needle, on flat (adaxial) side Number of resin canals Needle width Needle thickness Distance between vascular bundles Height (thickness) of epidermal cells Width of epidermal cells Marcet’s coefficient (= traits 9 ¥ 7/8) Stomatal rows ratio (= traits 2/3) Needle thickness/width ratio (= traits 8/7) Cell of epidermis width/thickness ratio (= traits 11/10) 1 mm 1 mm 1 mm 1 mm 1 mm 1 mm © 2003 The Linnean Society of London, Botanical Journal of the Linnean Society, 2003, 142, 83–91 85 MORPHOLOGY OF PINUS ULIGINOSA length was measured on fresh material, the other traits were measured on material preserved in 70% alcohol. Anatomical traits were analysed on semi-permanent preparations of sections taken from the middle part of the needle (for details see Boraty n¢ ska & Bobowicz, 2000, 2001; Boraty n¢ ska & Pashkevich, 2001). The data were analysed statistically by Statistica PL for Windows software on the basis of means for individuals. The arithmetic means, standard deviations and coefficients of variation were calculated and analysed. The correlation between traits was examined. Differences between Pinus uliginosa populations were analysed with the Student’s t-test for unpaired data, and among all studied populations by Tukey’s T-test with a separate assessment of variance (post hoc comparison of means, analysis of variance). A discriminant analysis was performed and the position of samples was examined along the first two or three discriminant variables to show the intra- and interpopulation variation, as well as intra- and interspecific variation (Marek, 1989; Morrison, 1990; Zar, 1999). The populations and taxa were also compared using the graphical method of Jentys-Szaferowa (1959). RESULTS VARIABILITY OF P. ULIGINOSA The analysed populations of Pinus uliginosa were significantly different with respect to most traits. The largest differences were observed in the ratio of sto- matal rows (trait 13), height of epidermal cells (trait 10), needle length (trait 1) and in the ratio of epidermal cell width/thickness (trait 15). Differences were nonsignificant only for three traits: number of stomata on convex side of needle (trait 4), width of epidermal cells (trait 11) and needle thickness/width ratio (trait 14) (Table 3). The majority of the analysed needle traits were significantly correlated (Tables 4 and 5). The strongest correlations within P. uluginosa from both the Batorów and We˛ gliniec populations were found between needle width (trait 7) and thickness (trait 8) (Pearson’s coefficients of correlation r = 0.87 and r = 0.86, respectively). The numbers of stomatal rows on the two sides of the needle were also correlated (r = 0.79 in both populations). The population from We˛ gliniec appeared to be more variable than the population from Batorów, with coefficients of variation higher for most traits (Fig. 1). The number of resin canals (trait 6), distance between vascular bundles (trait 9) and Marcet’s coefficient (trait 12) were the most variable traits in both populations. The needle thickness/width ratio (trait 14) had the lowest coefficients of variation, 4.71% for the We˛ gliniec population and 4.03% for the Batorów population. In addition, the variability of the stomatal rows ratio (trait 13) in needles from We˛ gliniec was also very low (4.56%), whereas the coefficient of variation in this trait was significantly greater (about 10%) in needles from Batorów. Table 3. Needle trait differences between Pinus uliginosa populations analysed by Student’s t-test for unpaired data Arithmetical means Character We˛gliniec 1. Needle length 47.54 2. Number of stomatal rows on the convex (abaxial) side of 9.51 needle 3. Number of stomatal rows on the flat (adaxial) side of needle 7.45 4. Number of stomata on a 2-mm long section of needle, on 18.90 convex (abaxial) side 5. Number of stomata on a 2-mm long section of needle, on flat 17.85 (adaxial) side 6. Number of resin canals 5.11 7. Needle width 1337.99 8. Needle thickness 795.45 9. Distance between vascular bundles 122.06 10. Height (thickness) of epidermal cells 36.55 11. Width of epidermal cells 15.68 12. Marcet’s coefficient (= traits 9 ¥ 7/8) 206.95 13. Stomatal rows ratio (= traits 2/3) 1.06 14. Needle thickness/width ratio (= traits 8/7) 0.597 15. Cell of epidermis width/thickness ratio (= traits 11/10) 0.432 Batorów t d.f. P 32.58 8.55 12.009*** 3.628*** 100 100 0.000 0.000 5.73 19.39 8.385*** - 1.777 100 100 0.000 0.079 19.83 - 7.013*** 100 0.000 3.21 1189.24 715.93 77.65 28.85 15.32 129.97 1.52 0.604 0.538 4.711*** 6.953*** 6.592*** 6.700*** 14.559*** 1.763 6.674*** - 21.462*** - 1.325 - 12.658*** 100 100 100 100 100 100 100 100 100 100 0.000 0.000 0.000 0.000 0.000 0.081 0.000 0.000 0.188 0.000 © 2003 The Linnean Society of London, Botanical Journal of the Linnean Society, 2003, 142, 83–91 86 K. BORATY N¢ SKA ET AL. Table 4. Correlation coefficients between traits of Pinus uliginosa from We˛gliniec. Trait numbers as in Table 2 2 3 4 5 6 7 8 9 10 11 Traits 0.33* 0.42** -0.01 0.01 0.04 0.42** 0.44** 0.20 0.11 -0.07 1 0.79** 0.21 0.34* -0.11 0.59** 0.62** 0.19 0.48** 0.15 2 0.20 0.27 0.14 0.59** 0.53** 0.14 0.20 0.02 3 0.86** 0.08 0.08 0.15 -0.03 0.21 0.11 4 0.07 0.14 0.18 -0.02 0.23 0.02 5 0.29* 0.25 0.28* -0.14 -0.26 6 0.86** 0.50** 0.29* -0.16 7 0.42** 0.30* 0.00 0.19 0.02 8 9 0.47** 10 *P = 0.05; **P = 0.01 Table 5. Correlation coefficients between traits of Pinus uliginosa from Batorów. Trait numbers as in Table 2 2 3 4 5 6 7 8 9 10 11 Traits 0.03 0.22 -0.08 -0.11 0.23 0.40** 0.29* 0.35* 0.10 -0.09 1 0.79** 0.07 0.05 0.09 0.62** 0.59** 0.51** 0.28 0.13 0.17 0.02 0.36* 0.69** 0.60** 0.57** 0.18 0.02 2 3 0.32* 0.25 0.12 0.18 0.17 -0.03 0.06 4 0.13 0.13 -0.01 0.01 0.11 0.08 5 0.45** 0.40** 0.38** 0.13 0.16 0.87** 0.63** 0.38** 0.12 0.49** 0.47** 0.22 0.01 0.00 6 7 8 9 0.40** 10 Figure 1. Coefficients of variation of needle traits in analysed populations of Pinus spp. (acronyms as in Table 1). © 2003 The Linnean Society of London, Botanical Journal of the Linnean Society, 2003, 142, 83–91 MORPHOLOGY OF PINUS ULIGINOSA The analysis of the discriminant function for populations of P. uliginosa from We˛ gliniec and Batorów showed that the intrapopulational variation within both populations was relatively low in the space of the first three canonical variables (Figs 2, 3). In the sample from the We˛ gliniec population, trees 11, 22 and 7 appeared the most distinct. Upon analysis of the measurement results, it was noticed that needles of these individuals were longer, wider and thicker than those of all other individuals. In addition, tree 11 had needles with an extremely large distance between vascular bundles (268 mm) and a very high Marcet’s coefficient (448 mm). In the sample from the Batorów population, individuals were differentiated to a much lesser degree than those from We˛ gliniec. 87 Table 6. Tukey’s T-test of differences between five populations of studied Pinus species COMPARISON OF PINUS ULIGINOSA, P. SYLVESTRIS AND P. MUGO Comparisons between the studied populations of all three taxa were analysed by Tukey’s T-test (Table 6). The width of epidermal cells (trait 11) is the only trait that did not differ among the three taxa and five populations studied, whereas the other 14 traits differed among the populations. Pinus uliginosa and P. sylvestris from We˛ gliniec had similar needle lengths (trait 1) and numbers of stomatal rows on both sides of the needle (traits 2 and 3), but differed significantly in all other traits. The largest differences between the two populations were in the numbers of stomata on both sides of the needle U3 (16%) 7 22 U3 (15%) ) (3 1 U Figure 2. Result of discriminant analysis based on 11 needle traits of Pinus uliginosa from We˛gliniec plotted along the three first canonical variables U1, U2 and U3, which accounted for 71% of the total variation. 9% %) ) (1 (22 32% 2 U2 U 4% ) 11 ( U1 Figure 3. Result of discriminant analysis based on 11 measured needle traits of Pinus uliginosa from Batorów plotted along the three first canonical variables U1, U2 and U3, which accounted for 67% of the total variation. © 2003 The Linnean Society of London, Botanical Journal of the Linnean Society, 2003, 142, 83–91 88 K. BORATY N¢ SKA ET AL. (traits 4 and 5), number of resin canals (trait 6), distance between vascular bundles (trait 9) and thickness of epidermal cells (trait 10). Pinus uliginosa from We˛ gliniec and P. sylvestris from Szczeliniec differed markedly in all analysed traits except for the width of epidermal cells (trait 11). Pinus uliginosa had significantly longer needles, thicker epidermal cells (trait 10), fewer stomata on both sides of the needle (traits 4 and 5), fewer resin canals (trait 6) and a shorter distance between vascular bundles (trait 9). Pinus mugo from the Karkonosze Mts and P. uliginosa from We˛ gliniec did not differ in numbers of stomata on both sides of the needle (traits 4 and 5) or in numbers of resin canals (trait 6). The other traits differed significantly among these populations. By contrast, P. mugo and P. uliginosa from Batorów were not statistically different with respect to distance between vascular bundles (trait 9) and Marcet’s coefficient (trait 12), while all other traits significantly differentiate these populations (Table 6). Pinus uliginosa from Batorów had smaller needles (shorter, narrower and thinner, traits 1, 7 and 8, respectively), fewer stomatal rows (traits 2 and 3), and more stomata per 2-mm-long section of the needle, especially on the flat side of the needle (trait 5). Moreover, it had fewer resin canals (trait 6) and a lower mean thickness of epidermal cells (trait 10). Pinus uliginosa from Batorów and P. sylvestris from We˛ gliniec differed significantly in all traits except the number of stomatal rows on the convex side of the needle (trait 2). P. sylvestris had much longer needles (trait 1), more stomata on both sides of the needle (traits 4 and 5), a longer distance between vascular bundles (trait 9) and three times more resin canals (trait 6) than P. uliginosa. Pinus uliginosa from Batorów differed significantly from P. sylvestris from Szczeliniec in all traits except needle thickness (trait 8). The two populations differed the most strongly in the number of resin canals (trait 6), distance between vascular bundles (trait 9), and Marcet’s coefficient (trait 12). Values of those traits were two to three times or even four times higher in P. sylvestris than in P. uliginosa. The lowest coefficients of variation in all studied populations were detected in needle thickness/width ratio (trait 14), particularly in P. mugo (Fig. 1). Low variation was observed in the width of epidermal cells (trait 11) in both populations of P. sylvestris, and in stomatal rows ratio (trait 13) in P. uliginosa from We˛ gliniec. Coefficients of variation of the two traits did not exceed 5%. The number of resin canals (trait 6) was the most variable in all populations. The coefficients of variation were highest in Pinus uliginosa (50% and 39% in We˛ gliniec and Batorów, respectively) lowest in P. mugo and intermediate in P. sylvestris. Other highly variable traits included the distance between vascular bundles (trait 9) and Marcet’s coefficient (trait 12). They were markedly less variable in P. sylvestris and P. mugo than in P. uliginosa, especially in P. sylvestris from Szczeliniec. (Fig. 1). The proportions among the studied traits from all compared populations of Pinus uliginosa and P. mugo (Jentys-Szaferowa, 1959) were similar and reflect the relatedness of these species. More substantial differences were noticed in traits such as needle length (trait 1), thickness of epidermal cells (trait 10), ratio of stomatal rows (trait 13), needle thickness and width (traits 8 and 7) and number of stomatal rows on both sides of the needle (traits 2 and 3). With respect to the width of epidermal cells (trait 11) and numbers of stomata, especially on the convex side of the needle (trait 4), the studied populations were the most comparable. The last traits can be considered as characteristic for all compared taxa. Discriminant analysis showed that the ratio of stomatal rows (trait 13), needle length (trait 1) and thickness of epidermal cells (trait 10) together has the greatest discriminant power. The studied populations were not discriminated at a statistically significant level by needle thickness (P = 0.05) and by needle width and Marcet’s coefficient (P = 0.01). All other traits discriminated the analysed taxa and populations at somewhat lower, but nevertheless highly statistically significant levels (Table 7). The two populations of Pinus sylvestris were closely related and clearly distinct from all other studied populations based on the plane of the first two discriminant variables U1 and U2. They were separated by the first variable U1 (Fig. 4). Both populations of P. uliginosa were intermediate between P. sylvestris and P. mugo, but closer to P. mugo in the space described by the first canonical variable U 1. Surprisingly, P. uliginosa from Batorów was somewhat closer to the P. sylvestris group than P. uliginosa from We˛ gliniec (variable U1), which was much more exposed to potential hybridization with P. sylvestris. The variable U2 separated the Batorów population of Pinus uliginosa from the other populations. Pinus mugo from the Karkonosze Mts and P. uliginosa from We˛ gliniec were the most closely related taxa. This was even more conspicuous in the plot of the distribution of canonical variables of the individuals than the mean values for the populations (Fig. 4), where those two populations overlap, whereas P. uliginosa from Batorów is clearly distinct. Squares of Mahalanobis distances between centroids of the analysed populations indicated that the closest were the two populations of P. sylvestris. © 2003 The Linnean Society of London, Botanical Journal of the Linnean Society, 2003, 142, 83–91 MORPHOLOGY OF PINUS ULIGINOSA 89 Table 7. Discriminant power testing for needle characters of Pinus uliginosa, Pinus mugo and Pinus sylvestris Trait F statistic P value 1. Needle length 2. Number of stomatal rows on the convex (abaxial) side of needle 3. Number of stomatal rows on the flat (adaxial) side of needle 4. Number of stomata on a 2-mm long section of needle, on convex (abaxial) side 5. Number of stomata on a 2-mm long section of needle, on flat (adaxial) side 6. Number of resin canals 7. Needle width 8. Needle thickness 9. Distance between vascular bundles 10. Thickness of epidermal cells 11. Width of epidermal cells 12. Marcet¢c coefficient (= traits 9 ¥ 7/8) 13. Stomatal rows ratio (= traits 2/3) 14. Needle thickness/width ratio (= traits 8/7) 15. Epidermal cells width/thickness ratio (traits 11/10) 28.203 12.586 11.489 4.572 5.723 7.576 2.811 2.004 2.846 26.488 17.668 3.168 53.437 4.415 17.704 0.000 0.000 0.000 0.002 0.000 0.000 0.027 0.096 0.026 0.000 0.000 0.015 0.000 0.002 0.000 10 P.uliginosa W. 8 P.uliginosa B. P.sylvestris W. 6 P.sylvestris Sz. P.mugo R. U2 (20%) 4 2 0 -2 -4 -6 -8 -12 -10 -8 -6 -4 -2 0 2 4 6 8 10 U1 (64%) Figure 4. Results of discriminant analysis based on 11 needle traits for five populations of Pinus spp. (acronyms as in Table 1) on the plane of the first two discriminant variables: U1 and U2, which accounted for 84% of the total variation. The most distant was P. mugo in relation to the two populations of P. sylvestris. Pinus uliginosa from Batorów was clearly distinct from the others, while P. uliginosa from We˛ gliniec was closest to P. mugo. DISCUSSION In previous studies, Pinus uliginosa from Batorów was more closely related to P. mugo than to P. sylvestris (Szweykowski, 1969; Staszkiewicz & Tyszkiewicz, © 2003 The Linnean Society of London, Botanical Journal of the Linnean Society, 2003, 142, 83–91 90 K. BORATY N¢ SKA ET AL. 1972; Siedlewska & Prus-G l¢owacki, 1995; PrusGl¢owacki et al., 1998), whereas P. uliginosa from We˛ gliniec was more closely related to P. sylvestris, especially in such traits as the number of stomata, Marcet’s coefficient and number of resin canals (Staszkiewicz & Tyszkiewicz, 1972). Results of our biometric analysis indicated that needles of Pinus uliginosa and P. sylvestris from We˛ gliniec were similar only with respect to the number of stomatal rows on both sides of the needle, whereas in all traits mentioned by Staszkiewicz & Tyszkiewicz (1972) the species differed significantly (P £ 0.05). However, P. uliginosa from Batorów and P. sylvestris from We˛ gliniec also had a similar number of stomatal rows on the convex side of the needle. Data of Staszkiewicz & Tyszkiewicz (1972) concerning the number of resin canals, Marcet’s coefficient and numbers of stomata are generally similar to the present results. Small but statistically nonsignificant decreases in the number of resin canals as well as small changes in numbers of stomata were noticed in both populations of P. uliginosa. With respect to the numbers of stomata on both sides of the needle and number of resin canals, Pinus uliginosa from We˛ gliniec was most similar to P. mugo from the Karkonosze Mts. The great similarity of this population of P. uliginosa to P. mugo (from the Tatra Mts.) is also visible in data of Staszkiewicz & Tyszkiewicz (1969). An exception is Marcet’s coefficient of P. uliginosa from We˛ gliniec, as it is most comparable to that of hybrids between P. sylvestris and P. mugo from the Nowy Targ Basin (Staszkiewicz & Tyszkiewicz, 1969, 1972). Pinus uliginosa from Batorów and P. mugo from the Karkonosze Mts were identical in terms of distance between vascular bundles and Marcet’s coefficient. This is also observed in comparisons with P. mugo from the Tatra Mts (Staszkiewicz & Tyszkiewicz, 1969). Despite their similarities, the two populations of Pinus uliginosa were surprisingly different in many morphological and anatomical traits. The two analysed populations were different in all traits except for the number of stomata on the convex side of the needle, needle thickness/width ratio, and width of epidermal cells (although none of the studied populations differed significantly in the last trait). The two populations of P. uliginosa also differ in several enzymatic systems (Lewandowski et al., 2002). The differences between analysed populations may be due to their long isolation and possibly contrasting origins. It is rather unlikely that the differences result from the direct influence of P. sylvestris. The population of P. uliginosa from We˛ gliniec was not more closely related to P. sylvestris than the population of P. uliginosa from Batorów with respect to all studied traits. Thus, although genetic erosion of the population of P. uliginosa at We˛ gliniec as a result of pollination by P. sylvestris is possible, it has not been confirmed. Phenological observations of these pine species revealed only slight differences in periods of development of macro- and microstrobili of P. uliginosa and P. sylvestris, so cross-pollination between the two species cannot be excluded. Our results as well as those of isozyme assays of P. uliginosa seeds from We˛ gliniec indicate that effective hybridization between the two species is not a likely occurrence. CONCLUSIONS 1 The population of Pinus uliginosa from We˛ gliniec, although surrounded by a P. sylvestris forest, does not show considerable affinity to that species and does not seem to be under its marked influence. 2 The population of P. uliginosa from We˛ gliniec is much more variable than those from Batorów. 3 Both populations of P. uliginosa are more closely related to P. mugo than to P. sylvestris. 4 In most analyses the two populations of P. uliginosa and single population of P. mugo are close to each other and form one group, whereas the two populations of P. sylvestris form another distinct group. 5 Pinus uliginosa from We˛ gliniec is more closely related to P. mugo than P. uliginosa from Batorów. 6 Needle traits in both populations of P. uliginosa did not change significantly during the last 30 years, although the numbers of individuals in both populations have been greatly reduced. 7 Populations of P. uliginosa from We˛ gliniec and Batorów differ from each other significantly and both should not only be passively preserved but also actively conserved. ACKNOWLEDGEMENTS We are grateful to Mark G. Tjoecker (Department of Forest Science, Texas A & M University) for having corrected the English language of a previous version of the manuscript. The study was partly sponsored by the Polish Committee for Scientific Research, grant no. 6P04G 060 16. REFERENCES Bobowicz MA. 1990. The hybrids of Pinus mugo Turra ¥ Pinus sylvestris L. from ‘Bór na Czerwonem’ reservation in the Nowy Targ Valley. University of Adam Mickiewicz, Poznan¢ , Series Biologia 40: 1–284. Boraty n¢ ska K, Bobowicz MA. 2000. Variability of Pinus uncinata Ramond ex DC. as expressed in needle traits. Dendrobiology 45: 7–16. © 2003 The Linnean Society of London, Botanical Journal of the Linnean Society, 2003, 142, 83–91 MORPHOLOGY OF PINUS ULIGINOSA Boraty n¢ ska K, Bobowicz MA. 2001. Pinus uncinata Ramond taxonomy based on needle characters. Plant Systematics and Evolution 227: 183–194. Boraty n¢ ska K, Pashkevich NA. 2001. Variability in needle traits of Pinus mugo Turra in the Ukrainian Carpathians. Acta Societatis Botanicorum Poloniae 70: 181–186. Boraty n¢ ski A. 1994. Protected and rare trees and shrubs from the Polish part of Sudety Mts. and its foothills. 7. Pinus mugo Turra and P. uliginosa Neumann. Arboretum Kórnickie 39: 63–85. Christensen KI. 1987a. Taxonomic revision of the Pinus mugo complex and P. ¥ rhaetica (P. mugo ¥ P. sylvestris) (Pinaceae). Nordic Journal of Botany 7: 383–408. Christensen KI. 1987b. A morphometric study of the Pinus mugo Turra complex and its natural hybridization with P. sylvestris L. (Pinaceae). Feddes Repertorium 98: 623–635. Christensen KI, Dar GH. 1997. A morphometric analysis of spontaneous and artificial hybrids of Pinus mugo ¥ sylvestris (Pinaceae). Nordic Journal of Botany 17 (1): 77–86. Danielewicz W, Zieli n¢ ski J. 2000. Ochrona sosny b l¢ otnej Pinus uliginosa A. Neumann na terenie Borów Dolno s¢l a˛skich. [Protection of the longleaf pine Pinus uliginosa A. Neumann in the Low Silesian Pinewood area]. Przegl a̧d Przyrodniczy 11 (2–3): 113–124. Filppula S, Szmidt AE, Savolainen O. 1992. Genetic comparison between Pinus sylvestris and P. mugo using isozymes and chloroplast DNA. Nordic Journal of Botany 12 (4): 381– 386. Go¢l a˛b Z. 1999. Sosna bl¢ otna (Pinus uliginosa Neumann) na Wielkim Torfowisku Batorowskim w Górach Sto l¢ owych. [Pinus uliginosa Neumann in Wielkie Torfowisko Batorowskie (The Great Batorowskie Peatbog) in the Stolowe Mountains]. Szczeliniec 3: 41–48. Jentys-Szaferowa J. 1959. A graphical method of comparing the shapes of plants. Nauka Polska 7 (3): 79–110. [In Polish]. Krzakowa M, Naganowska B, Bobowicz MA. 1984. Investigations on taxonomic status of Pinus uliginosa Neuman. Bulletin de la Société des Amis des Sciences et des Lettres de Pozna n¢, Série D, Sciences Biologiques 24: 87–96. Lauranson-Broyer J, Krzakowa M, Lebreton P. 1997. Reconnaissance chimiosystematique et biometrique du pin de tourbiere Pinus uliginosa (Neumann). Comptes Rendus de l’Academie des Sciences, Serie III, Sciences de la Vie 320: 557–565. Lewandowski A, Boraty n¢ ski A, Mejnartowicz L. 2000. Allozyme investigations on the genetic differentiation between closely related pines – Pinus sylvestris, P. mugo, P. uncinata and P. uliginosa (Pinaceae). Plant Systematics and Evolution 221: 15–24. Lewandowski A, Samo c¢ko J, Boraty n¢ ska K, Boraty n¢ ski A. 2002. Genetic differences between two Polish populations of Pinus uliginosa compared to P. sylvestris and P. mugo. Dendrobiology 48: 51–57. Marcet E. 1967. Über den Nachweiss spontaner Hybriden von Pinus mugo Turra und Pinus sylvestris L. auf Grund von 91 Nadelmerkmalen. Berichte der Schweitzerische Botanische Geselschaft 77: 314–361. Marek T. 1989. Analiza skupien¢ w badaniach empirycznych [The clustering analyses in empirical investigations]. Warszawa: Polish Scientific Publishers, 1–47. Morrison DF. 1990. Multivariate statistical methods. Warszawa: Polish Scientific Publishers, 1–585. [Polish edition]. Prus-G¢l owacki W, Bujas E, Raty n¢ ska H. 1998. Taxonomic position of Pinus uliginosa Neumann as related to other taxa of Pinus mugo complex. Acta Societatis Botanicorum Poloniae 67: 1–6. Siedlewska A, Prus-G¢l owacki W. 1994. Allozyme variability of putative hybrid swarm population (Pinus mugo Turra ¥ P. sylvestris L.) from Topielisko peat-bog near Zieleniec. Genetica Polonica 35 (4): 285–302. Siedlewska A, Prus-G¢l owacki W. 1995. Genetic structure and taxonomic position of Pinus uliginosa Neumann population from Wielkie Torfowisko Batorowskie in Stolowe Mts. (locus classicus). Acta Societatis Botanicorum Poloniae 64: 1– 8. Staszkiewicz J. 1985. Kilka uwag o sos¢nie b l¢ otnej Pinus uliginosa. [Some remarcks on the pine Pinus uliginosa]. Chro w¢ my Przyrode˛ Ojczysta˛ 41 (5): 56–61. Staszkiewicz J. 1993a. Pinus ¥ rhaetica Brügger. In: Zarzycki K, Ka z¢mierczakowa R, eds. Polish Plant Red Data Book. Kraków: Polish Academy of Sciences, W. Szafer Institute of Botany and Institute of Nature Conservation, 38–39. Staszkiewicz J. 1993b. Variability of Pinus mugo ¥ P. sylvestris (Pinaceae) hybrid swarm in the Tisovnica nature reserve (Slovakia). Polish Botanical Studies 5: 33–41. Staszkiewicz J. 1994. Differentiation of populations of Pinus ¥ rhaetica (Pinaceae) from the ‘Mezi Bormi’ Nature Reserve of Slovakia. Fragmenta Floristica et Geobotanica, Series Polonica 1: 223–233. [In Polish]. Staszkiewicz J, Tyszkiewicz M. 1969. Les hybrids natureles de Pinus mugo Turra et Pinus sylvestris L. dans la Valleé de Nowy Targ. Fragmenta Floristica et Geobotanica 15: 187– 212. [In Polish]. Staszkiewicz J, Tyszkiewicz M. 1972. Variability of the natural hybrids of Pinus sylvestris L. ¥ Pinus mugo Turra (= P. ¥ rotundata Link) in South-western Poland and in some selected localities of Bohemia and Moravia. Fragmenta Floristica et Geobotanica 18: 173–191. [In Polish]. Szweykowski J. 1969. The variability of Pinus mugo Turra in Poland. Bulletin de la Societé des Amis des Sciences et des Lettres de Poznan¢ , Série D, Sciences Biologiques 10: 37–54. Szweykowski J, Bobowicz MA. 1983. Variation in Pinus sylvestris L., Pinus mugo Turra and putative hybrid populations in Central Europe. I. The position of one-year-old conelets. Bulletin de la Société des Amis des Sciences et des Lettres de Poznan¢ , Série D, Sciences Biologiques 22: 43–50. Zar JH. 1999. Biostatistical analysis. New Jersey: Prentice Hall, 1–662. © 2003 The Linnean Society of London, Botanical Journal of the Linnean Society, 2003, 142, 83–91