Species Associations And Their Formation Mechanisms In An EvergreenBroadleaved Forest

The interspecific relationship was considered to be one of the essential characteristics of plant communities. It also was an important factor to determine community construction and maintain the biodiversity of community. Both inter- and intraspecific interactions will affect the species coexistence of a community at local space. For each species, the spatial distribution of individuals not only depended on the local habitat, but also was influenced by neighboring individuals. In this study, we used data from a fully mapped 20-ha subtropical evergreen broad-leaved forest plot in Tiantong to conduct a community-wide assessment of species association and its formation mechanism. Specifically, the study analysed species association based on homogeneious Poisson null model, species-habitat association based on Poisson cluster null model, species interaction based on heterogeneious Poisson null model and how this interaction change with ontogeny.The main results and conclusions are as follows.1. The results of species association which based on homogeneious Poisson null model showed that most of species association patterns are random at small scales (0-1 Om), but more than 70% species pairs showed positive or negative association at large scale (> 30m). Species-habitat association test showed that about 90% of the species were positively or negatively associated with at least one of seven habitats, indicating that species generally was impacted by habitat heterogeneity in Tiantong forest plot. In addition, at large scale (>30m), species with same habitat association usually have significant positive interspecific association (nearly 90%), and species with distinct habitat association mainly have significant negative interspecific association, suggesting that habitat heterogeneity affected species association at large scale.2. We performed a total of 34×33=1122 bivariate point pattern analysis for all pairs of the 34 species. The heterogeneious Poisson null model was used to control habitat heterogeneity. The goodness-of-fit test detected significant associatioins for 9.4% species pairs. Repulsion occurred more frequently than attraction at small scale, and pair frequencies peaked at 0-2 m scale.32 of 34 species were positively or negatively interacted with at least one of other specie, and a species usually have no significant interaction with most other species. To find out whether the significanc of our results were dependent on the number of stems of the species pair or the univariate species spatial structure of the component paterns, we calculated for all 1122 species pairs the Spearman rank correlation between the rank of goodness-of-fit test and the number of stems of species pairs, and the value of the pair-correlation function at some scales. The results show that the interactions between species only have weak correlation with species abundance and aggregation (correlation coefficient<0.2 and< 0.11, respectively).3. According to tree size, total 24 species were divided into three stages (small trees, middle trees and large trees). For each stage, we analyzed bivariate species interactions between species specific stage and heterogeneious large tree with the spatial point pattern methods under heterogeneous Poisson process null model.The results of the goodness-of-fit test at scale 0-20m show that the proportion of species pairs which have significant interaction reduced with ontogeny (from 23.9% to 14.1% to 12.9%).In conclusion, the study show that:the effect of habitat filtering is the main reason of species association at large scale; In small scale, although interspecific interactions will change with ontogeny, the proportion of significant interaction is low. Therefore, the interspecific interactions have little influence on species association pattern, in keeping with the general rusults in the species rich forest community.