EN
Tropical forests are among the most complex ecosystems on Earth. The high alpha-diversity of tropical forests has been amply documented, however, beta diversity, equally important for us to understand the mechanism of biodiversity maintenance, has seldom been studied. The main current hypotheses about the origin of beta diversity are as follows: (1) species composition is uniform over large areas; (2) species composition fluctuates in a random, auto-correlated way emphasizing spatially limited dispersal; (3) species composition is related to environmental conditions. Testing these hypotheses is important to understand ecosystem function and to manage ecosystems effectively. In this study, we quantified the relative influence of environmental variation and spatial distance on the beta diversity of woody plant functional groups in a tropical forest of Hainan Island, China. Floristic and environmental data were collected from 135 grid-sample plots. We classified woody plant functional groups based on frequency, growth forms and successional status, respectively. To see whether environmental variation and spatial distance influence beta diversity, the simple and partial Mantel tests, in conjunction with linear and spline regression models were performed. The results showed that environmental heterogeneity and spatial distance were the primary determinants of pair-wise species composition differences between plots in the study area. Common, tree and climax species were more sensitive than the others to changes in environmental heterogeneity and spatial distance. Except for rare, pioneer and shrub species, the others had apparently negative relationship between spatial distance and similarity. The distance decay of similarity was determined by joint influences of spatial distance and spatially-structured environmental variables. Spatial distance had the clearest effect on beta diversity at distances <15 km and with the increase in spatial distance, beta diversity was gradually controlled by environmental heterogeneity. This study provides further support for neutral theory and environmental control hypothesis as opposed to uniformly distributed hypothesis.