关键词: 累积经验分布函数, 模型拟合, 中性理论模型, 生态位模型, 纯统计模型, 物种多度分布

Abstract

Species abundance distribution (SAD) delineates abundance of all species sampled within a community. As one major stepping stone in understanding the community, the generation mechanisms of SAD have attracted much attention. Evergreen and deciduous plants are two types of species with distinct phenological traits and growth strategies. They widely coexist in evergreen broad-leaved forests (EBLFs). Compared to deciduous plants, evergreen species have slightly lower species richness but substantially higher abundance and basal area in the 20 ha EBLF plot at Tiantong. This study independently analyzing their SAD characteristics provided a new perspective on the understanding of species diversity maintenance in EBLFs. Therefore, in order to compare SADs and determine reasons for differences, an empirical cumulative distribution function (ECDF) was utilized to describe the SADs of evergreen and deciduous trees in Tiantong plot. A Kolmogorov-Smirnov test (K-S test) was employed to detect the significance of these differences. Additionally, three types of models, including statistic model (log-normal model and log-series model), niche model (broken-stick model and niche preemption model) and neutral theory model (metacommunity zero-sum multinomial distribution model and Volkov model), were used to fit the SAD of each lifeform. The K-S test and AIC values were applied to test the goodness of fit for each model. We found that the differences in SAD between the two life forms were not significant based on the results of the K-S test. Among the three types of models, the neutral theory model was the best fitting model, and the niche model was the poorest fit. Thus we conclude that evergreen and deciduous trees had similar SAD patterns, although they differed in species richness and abundance. However, the model fitting results were found to be a necessary but insufficient condition in understanding the maintenance mechanism of biodiversity. Hence we may only preliminarily conclude that neutral processes had a major effect on the generation of biodiversity patterns of both evergreen and deciduous trees, whereas the possible contributions made by other processes, such as niche differentiations, could not be excluded and measured by this method.

Key words: empirical cumulative distribution function, model fitting, neutral theory model, niche model, purely statistical model, species abundance distribution