Beta diversity is central to understanding the variation of ecological communities in space and time. Many measures of beta diversity have been proposed, and most of them are based on pairwise comparison. In reality, however, the majority of community studies contain at least three sites that need to compare with each other. In a multiple-site study, pairwise-based beta diversity measures often lead to the loss of ecological information among multiple comparisons. Moreover, pairwise-based beta diversity measures cannot avoid the co-variance between pairwise comparisons which increase the likelihood of inaccuracy. In this study, we introduced a new multisite measure for calculating beta diversity in ecological communities. We first defined the multiple-site beta diversity, and then derived a statistical model that describes the spatial patterns of multiple-site beta diversity. We then did a simulation using four common species abundance distributions(SADs). Finally, we applied our model to calculating the beta diversity of a plant community and a microbial community.
Results/Conclusions
We found that increasing the number of sub-sampling units led to an increase in beta diversity in a concaved down fashion. Such a pattern might result from conspecific spatial aggregation. We derived a mathematical equation to describe this pattern based on a negative binomial distribution. Our model fit well to the real datasets of a large scale plant community (BCI) and a soil microbial community. Our work suggested that the multi-site beta diversity measure could offer an alternative, integrated approach in describing changes and dynamics of ecological communities.