Discriminating the effects of local stressors from climatic factors and dispersal processes on multiple biodiversity dimensions of macroinvertebrate communities across subtropical drainage basins.

• We tested ecological drivers of biodiversity indices across subtropical rivers. • Local environmental stressors were more important than the other factor groups. • Taxonomic distinctness indices performed best in portraying human disturbances. • Species and functional diversity indices were also related to climatic and spatial factors. • Bioassessment should consider various factors causing variation in bio-indicators. Metacommunity ecology emphasizes that community structure and diversity are not only determined by local environmental conditions through environmental filtering, but also by dispersal-related processes, such as mass effects, dispersal limitation and patch dynamics. However, the roles of dispersal processes are typically ignored in bioassessment approaches. Here, we simultaneously explored the potential influences of four groups of factors: local stressors, climatic factors, within-basin spatial factors and basin identity in explaining variation in diversity indices of macroinvertebrate assemblages from seven subtropical tributary rivers. A total of 12 biodiversity indices based on species identities, functional traits and taxonomic relatedness were calculated and used in the subsequent statistical analysis. Our results showed that, although differing in their relative importance, the four explanatory factor groups all played important roles in explaining variation in biodiversity indices. Of the pure fractions, index variation was best explained by local environmental stressors, whereas the other three explanatory factor groups appeared less influential. Furthermore, diversity indices from species, functional and taxonomic dimensions responded distinctly to the focal ecological factors, and differed in their abilities to portray the effects of human disturbances on macroinvertebrate communities. Taxonomic distinctness indices performed best, with the highest amount of variation associated to local stressors and hardly any variation explained by other factors, implying that these indices are robust in portraying human disturbances in streams. However, species diversity and functional diversity indices were also affected by spatial processes and climatic factors, suggesting that these indices should be used with caution in bioassessment. We hence conclude that environmental assessment of riverine ecosystems should not rely entirely on the perspective of species sorting. In contrast, both roles of spatial processes and environmental variables related to human disturbances and climatic variation should be incorporated in management and conservation of riverine ecosystems. [ABSTRACT FROM AUTHOR]