Multivariate relationship between tree diversity and aboveground biomass across tree strata in a subtropical evergreen broad-leaved forest

Abstract

The correlation between primary producer diversity and the productivity (or biomass) of ecosystems is one of the most important and broadly studied relationship in ecology. Ecologists have discussed this relationship from two fundamentally different perspectives. Historically, productivity has been viewed as driver of species diversity. Recently, many studies have demonstrated that diversity can also control, rather than responds to, the production of biomass. These contrasting points of view have led to the debate about whether species diversity is the cause or the consequence of community productivity. Multivariate productivity-diversity hypothesis has been put forward to reconciling this debate. This hypothesis state that: (i) the environmental factors are the direct driver of species that can coexist within an area; (ii) the biomass of the area is directly influenced by the environmental factors that limit production; (iii) the environmental factors indirectly influenced biomass of the area via influencing the species number to coexist within the area that affects how efficiently environmental resources are converted into biomass. To date, empirical support for this hypothesis is scarce, especially for structurally complex terrestrial ecosystems. In this study, we modified multivariate productivity- diversity hypothesis to accommodate complex vertical structure of forest ecosystems, and used structural equation modeling and data from a large evergreen broad-leaved forest dynamic plot (24 hm2 in area) in subtropical China to understand the causal relationships among environmental factors, species richness, dominance and aboveground biomass. Trees were grouped into tow functional groups (overstorey and understorey) in order to improve the ability to detecting the diversity effect. The final model explained 30% and 58% of the variation in aboveground biomass of overstorey and understorey, respectively. Dominance was the most important factor in explaining the variation of aboveground biomass of overstorey, while both species richness and environment factors were the most important factors in explaining the variation of aboveground biomass of understorey. Consistent with the predictions of multivariate productivity-diversity hypothesis, (i) aboveground biomass were directly influenced by the environmental factors in both overstorey and understorey; (ii) environmental factors directly influence tree dominance and species richness of overstorey, but only directly influence species richness of understorey; (iii) environmental factors indirectly influence aboveground biomass of overstorey by influencing dominant, while indirectly influence aboveground biomass of understorey by influencing species richness. Interestingly, species richness was only important to explaining variation in aboveground biomass of understorey. This may be attributed to the low light condition in understorey, which may alleviative competition among trees of understorey. To our knowledge, this study provides the first empirical evidence supporting the multivariate productivity-diversity hypothesis in forest ecosystem thus contributes to extending the generality of this framework. Our result also highlights both dominant species and higher species richness is required to maximizing forest biomass.