Disentangling the drivers of taxonomic and phylogenetic beta diversities in disturbed and undisturbed subtropical forests.

Jinliang Liu,Hong Qian,Chuping Wu,Jianhua Chen,Xinliang Wei,Yi Jin,Xiaofeng Jin,Jiajia Liu,Mingjian Yu,Shuquan Yu

doi:10.1038/srep35926

Abstract

Understanding the relative importance of dispersal limitation and environmental filtering processes in structuring the beta diversities of subtropical forests in human disturbed landscapes is still limited. Here we used taxonomic (TBD) and phylogenetic (PBD), including terminal PBD (PBDt) and basal PBD (PBDb), beta diversity indices to quantify the taxonomic and phylogenetic turnovers at different depths of evolutionary history in disturbed and undisturbed subtropical forests. Multiple linear regression model and distance-based redundancy analysis were used to disentangle the relative importance of environmental and spatial variables. Environmental variables were significantly correlated with TBD and PBDt metrics. Temperature and precipitation were major environmental drivers of beta diversity patterns, which explained 7–27% of the variance in TBD and PBDt, whereas the spatial variables independently explained less than 1% of the variation for all forests. The relative importance of environmental and spatial variables differed between disturbed and undisturbed forests (e.g., when Bray-Curtis was used as a beta diversity metric, environmental variable had a significant effect on beta diversity for disturbed forests but had no effect on undisturbed forests). We conclude that environmental filtering plays a more important role than geographical limitation and disturbance history in driving taxonomic and terminal phylogenetic beta diversity.

Highlights

Understanding the relative importance of dispersal limitation and environmental filtering processes in structuring the beta diversities of subtropical forests in human disturbed landscapes is still limited
Regardless of whether an analysis included both gymnosperms and angiosperms or only angiosperms, multiple linear regression showed that AgeDist and Geographical distance (GeoDist) had no influence on beta diversity of all forests, Environmental distance (EnvDist) significantly correlated with Taxonomic beta diversity (TBD) and PBDt metrics (P < 0.05 in both cases, except for the Jaccard metrics), but no variables were significantly correlated with PBDb metrics (Table 1; Table S4)
mean annual temperature (MAT) and annual precipitation (AP) were the significant environmental variables (Table 2; Table S5), and together explained 7–27% of the variance in TBD and PBDt (Figs 3 and S1), whereas the pure spatial variable explained less than 1% of the variance in TBD and PBDt of all forests combined (Figs 3 and S1)

Summary

Introduction

Understanding the relative importance of dispersal limitation and environmental filtering processes in structuring the beta diversities of subtropical forests in human disturbed landscapes is still limited. Unraveling the relative importance of environment and space on TBD and PBD patterns is critical to understanding the roles played by historical and current ecological processes in shaping the regional biodiversity[3,4,5,6] The impacts of these drivers on TBD along spatial and environmental gradients have been well documented[3,7,8,9], but TBD analysis alone cannot detect the effects of evolutionary processes on community assembly[10]. If turnover of species within clades is weaker than turnover deep within phylogeny, entire clades track environmental conditions; otherwise, if turnover of species within clades is greater than turnover at deeper levels, selective pressures promote divergence into habitats, and recently evolved species are likely to occupy different environmental regimes than their ancestors[16] Using both PBDt and PBDb measures might help us to understand whether the phylogenetic divergence between an array of sites has occurred recently or deep in the past. Different patterns of beta diversity in different regions may reflect different mechanisms of community assembly; for example, environmental filtering was often found to be more important in temperate forests than in tropical forests[3,13,22,23]

Objectives

Methods

Results

Conclusion