天然次生林中水曲柳种子库的空间格局与过程

THE SPATIAL PATTERN AND PROCESS OF SEED BANK OF FRAXINUS MANDSHURICA IN SECONDARY HARDWOOD FORESTS

研究土壤种子库的空间格局与过程是认识林木更新的基础。应用地统计学方法定量分析了东北林业大学帽儿山实验林场天然次生林中土壤种子库水曲柳 (Fraxinusmandshurica)种子的统计特征、空间格局、尺度和格局的变化过程。结果表明 ,落种后水曲柳种子库的平均密度有很大变幅 (5 .8～ 10 5个·m-2 ) ,平均年削减率为 91.6 %～ 92 .7%,5～ 8月的削减率高达 98%以上。变异函数分析及理论模型拟合显示 ,一年之内种子库的空间格局强度及尺度变化非常明显。种子雨可形成较强的异质性格局 ,空间自相关尺度较大 ,为 36 .6～ 38.1m ,自相关引起的空间变异占 5 7%～ 72 %,随机变异占 2 8%～ 43%。翌年 5～ 8月 ,自相关尺度减小为 2 .3～ 3.3m ,但空间结构方差增大 ,自相关引起的空间变异高达 87%以上 ,由随机因素引起的空间变异只占 13%以下。空间局部插值 (Kriging)分析显示 ,种子库空间格局动态变化强烈。落种一年之后 ,种子库格局强度降低、空间尺度减小、空间格局呈破碎化。在种子库格局变化过程中 ,小尺度上的土壤真菌腐坏、动物摄食和微生境干扰等起重要作用。

Seed bank should strongly influence stand regeneration and thus affect the whole forest dynamics. Understanding its spatial pattern and process is fundamental to understanding its important role in the ecological process. We surveyed soil seed bank of Manchurian ash (Fraxinus mandshurica) during 1999 and 2000 in two 0.09 hm 2 plots (each 30 m×30 m) in secondary hardwood forests, on Maoershan Experiment Station of Northeast Forestry University, north-eastern China. Seed bank samples, by using geostatistical techniques and regular grid-based spatial sample design, were collected from soil cores (10 cm diameter, 6.7 cm deep) which sampled at the intersection points of grid plots (2 m×2 m grid, 225 sample points and 0.5 m×0.5 m grid, 90 sample points) for each plot. The spatial pattern intensities, scales, and dynamics of seed bank were assessed by semivariogram modeling and kriging maps, which revealed considerable spatiotemporal heterogeneity of seed bank. The mean value for seed bank density varied significantly at different plots and among seasons (5.8-105 seeds·m -2 ). The annual seed loss rate in soil seed bank was about 91.6%- 92.7% in a year, and over 98% during next May and August. According to isotropic semivariogram theoretical model parameters, the stronger spatial pattern (Sill is higher, between 8.52-9.53) and the broader autocorrelated scale (36.6-38.1 m) were found at the end of seed rain season. In those variations, spatially structured variance accounted for 57%-72%, and 28%-43% randomly. However, pattern intensities became weak (Sill became low, only 0.06-0.49) and scales became smaller (2.3-3.3 m) during the next growing season after seed dispersing, but with higher proportion of spatially structured variance (>87%) and smaller randomly variance (<13%). Fractal dimension analysis indicated that there was no obviously anisotropic variation for spatial pattern of seed bank in research sites. Kriging maps also showed that there were the strongly patterns and processes, from larger patterns with clearly gradient, higher density seeds, changing to fragmental and smaller patterns. The soil microbes, animals and microhabitats were the important factors in patterning of seed bank. Those results demonstrated that the initial patterns of seed bank were governed primarily by seed dispersal, lately were depended on the spatial heterogeneity at small scales.