Abstract
Phylogenetically close and/or functionally similar species are generally assumed to compete more strongly than phylogenetically distant and functionally dissimilar species in a community. However, how coexisting species with different extents of variation in their phylogeny and functional traits determine community function such as aboveground biomass across forest strata remains an unresolved question. We hypothesize that phylogenetically close and functionally similar species of overstorey, but phylogenetically distant and functionally dissimilar species of understorey drive high aboveground biomass in structurally-complex subtropical forests. Multiple linear regression models were used to test the consequences of phylogenetic distances and functional traits of tree species, and environmental factors on aboveground biomass in a subtropical forest in Eastern China. Overstorey aboveground biomass was driven by functional evenness (beta = -0.21), phylogenetic species evenness (beta = -0.27) and phylogenetic diversity (beta = 0.31). Understorey aboveground biomass was driven by functional richness (beta = 0.25), functional dispersion (beta = -0.21), soil nutrients (beta = -0.17) and topography (beta = 0.36). Whole-community aboveground biomass was best predicted by functional divergence (beta = 0.36), functional dispersion (beta = -0.38), phylogenetic diversity (beta = 0.24) and soil nutrients (beta = -0.20). Our results suggest that understorey aboveground biomass is great for groups of phylogenetically distant species having high functional richness due to specific functional strategy shared by all the species. By contrast, high overstorey aboveground biomass is related with groups of phylogenetically close species having low functional trait diversity (i.e. high functional trait similarity) due to the evolutionary relatedness. The mechanism at the whole community level might result from the mixed effects of overstorey evolutionary relatedness and understorey functional trait diversity. This study highlights that disentangling the effects of evolutionary diversity and functional trait diversity across forest strata may be helpful for better understanding of ecological mechanisms for predicting aboveground biomass in a subtropical forest.