Cascading effects of elevation, soil moisture and soil nutrients on plant traits and ecosystem multi-functioning in Poyang Lake wetland, China

Trait-based approaches have been used to explore ecosystem level consequences of local changes in biodiversity along environmental gradients. Using a novel response–effect trait framework, we seek to identify and explain the mechanisms governing the cascading effects of environment on plant traits and ecosystem multi-functioning in the large Poyang Lake wetland, China. We sampled the above-ground biomass of the plant communities in 129,100 m2 plots and estimated the community weighted mean (functional identity, CWM) and distribution (functional diversity, FD) of seven plant morphological and phenological traits. The response–effect trait model (as showed by generalized multilevel path models) explained 97% and 51% of the variation in community productivity and soil carbon, respectively, and indicated that CWM morphometric (-morph) and CWM phenotypic (-pheno) traits promoted community productivity. Soil nutrient, followed by soil moisture exerted dominant and direct control on soil carbon. Elevation as well as soil moisture and nutrients had indirect effects on community productivity and soil carbon either via CWM or FD traits. Our study indicates that mass ratio processes (as reflected by CWM) modified the variations in community productivity and that niche complementary processes (as reflected by FD) explained some of the variations in soil carbon, though the relationships were weak. We conclude that a trait-based approach is useful in predicting the cascading effects of environmental factors on productivity, while the tested traits (that did not include root traits) mediated only a few environmental effects on soil carbon.