Linking understory species diversity, community-level traits and productivity in a Chinese boreal forest

Question The consensus has been growing over the past decade that functional traits and diversity are better to explain forest overstory diversity-productivity relationships (DPRs) than species diversity. Although the understory accounts for the majority of plant diversity in forests, it remains unclear how understory aboveground biomass production (UABP) is influenced by its species diversity and community-level functional traits. Location Great Xing'an Mountains of ortheastern China. Methods We quantified the effects of species richness, community aggregated traits (community weighted mean trait values, CWM) and functional diversity (functional dispersion, FDis) on UABP using structural equation modeling (SEM), which simultaneously accounted for the effects of overstory tree basal area, stand age, and soil fertility. Results In the full model, species richness had a negative direct, a positive indirect and no total effect on UABP. Furthermore, CWM and FDis, respectively, exhibited positive and no effect on UABP. Among the covariates, soil fertility, stand age, and overstory tree basal area had, respectively, positive, negative, and no effect on UABP. In the model without species richness, all trait variables had similar effects on UABP to those in the full model. In the richness-only model without traits, species richness, soil fertility and stand age had no effect on UABP. Conclusions Our results suggest that the selection effect largely determined understory DPRs due to the stronger effects of CWM on UABP than of FDis. Soil fertility exhibited the strongest influence on understory DPRs due to its parallel influences on traits, diversity, and productivity. The increase in resource availability induced by overstory tree litter-fall likely promoted soil fertility as the main driver of the understory DPRs. Stand age exhibited a negative effect on UABP, which may have contributed to the increases in shrub dominance and decreases in production due to limited resources.