Although ecologists generally recognize that environmental and spatial processes jointly determine the beta diversity of plant communities, the results regarding the relative importance of the two processes are largely incomparable and inconsistent. This may be due to the differences in species and taxonomic groups, evolutionary processes, or scale (i.e. extent, grain and spatial lag) of analysis in different studies. Using data of community composition, topographic and spatial variables from a subtropical broad-leaved forest in a large-scale stem-mapping plot in east China, we systematically investigate the relative contributions of environmental and spatial processes on structuring beta diversity at different spatial scales.

Results/Conclusions

Our results show that (1) the relative contributions of environmental and spatial processes are highly scale-dependent, and the combination of different lag, extent and grain in ecological studies can produce very different results. Our work in part explains recurrent results of previous studies; (2) despite of the inconsistent results at different scales, topographic and spatial variables dominate beta diversity only at particular extents, grain sizes and lags in this study. These results suggest that future studies gauging the importance of environmental processes against the strength of spatial community dynamics processes should use more sophisticated sampling designs to include multiple scales, and detect whether the included environmental and spatial variables dominate the community structure.