地形与土壤因子是决定植物群落结构的主要因素，同时植物多样性在不同尺度上影响着景观中的流，植物、地形与土壤三者之间在不同尺度上密切联系，研究它们之间的关系对于了解生态系统功能过程非常重要。针对东灵山地区异质化景观现状，在小流域尺度上选取了76个20 m×20 m的样地进行调查，研究选取了16个指标，分别代表地形因子、植物群落与土壤性质。利用多元统计手段，对选取指标进行了主成分分析、聚类分析与典范相关分析。结果显示土壤、植被、地形因子三者相互影响：地形因子与土壤理化性质影响着植被的分布格局与群落结构，反过来群落因子也对土壤的理化因子产生较大的影响。土壤有机质含量是指示土壤肥力的一个敏感指标，3种不同肥力类型中，群落的层盖度与丰富度存在较大的差别，而决定其分布的地形特点也不相同。不同指标集团的典范相关分析表明，坡位主要影响乔木层盖度，海拔高度与全氮相关性密切，而乔木层物种丰富度与有机质含量关系密切。
Biodiversity pattern impacts on flow of landscapes on different scales, and contrarily landscape and soil characteristics influence plant community structure. Variation in soil resource levels is common and important to plants. At larger spatial scales, the resources available to plants change as the soil type changes, and this variation effects the distribution of plant species. At smaller spatial scales, soil resources continue to show considerable spatial heterogeneity. Small-scale heterogeneity has large impact on performance of individual plants, and hence, on the structure and dynamics of plant populations and communities. It is important to examine their close relationships to understand the process of ecosystems. Many recent studies mainly concentrate on factors influencing plant in grassland or shrub land.The Donglingshan Mountain is an extension of Xiaowutaishan Mountains and belongs to the broader Taihangshan Mountains. The zonal vegetation of this region is highly fragmented warm temperate-zone deciduous broad-leaved forest. Focused on the landscape heterogeneity of this area, we investigated 76 plots of 20 m×20 m on a small watershed scale and analyzed the detailed community structure, landform and soil features.Using multivariate statistical analysis techniques (Principle Component Analysis, Cluster Analysis, Discriminant Analysis and Canonical Correlation Analysis), we examined the complex relationships between the soil properties, plant biodiversity and landscape factors. Soil organic matter is a sensitive factor of soil fertility and the coverage and richness of plants have different features under different soil fertilities. Three community types were distinguished by the arbor species characteristics. Twenty species existed in all plots of type III, but type I and II included sixteen species. The dominant species of type I was maple which accounts for 19.7% of total species in terms of importance value. Second was Quercus liaotungensis (19.1%). Type II included Q. liaotungensis (34.8%) and Betula dahurica (17.5%). Type III was B. dahurica (30.8%) and maple (21.6%). The results also show that high fertility plots often exist on the north-facing and upper slopes. Canonical correlation analysis was used as large numbers of variables were involved in this study. The analysis elucidated the relationships of three different indices (i.e. topography, soil and plant). Slope positions were closely related with tree layer coverage, while elevation was positively correlated with total nitrogen. The richness of tree layer in biodiversity factors had the greatest influence on soil organic matter.
全 文 ：