Abstract

Anthropogenic disturbances have become one of the primary causes of biodiversity decline in freshwater ecosystems. Beyond the well-documented loss of taxon richness in increasingly impacted ecosystems, our knowledge on how different facets of α and β diversity respond to human disturbances is still limited. Here, we examined the responses of taxonomic (TD), functional (FD) and phylogenetic (PD) α and β diversity of macroinvertebrate communities to human impact across 33 floodplain lakes surrounding the Yangtze River. We found that most pairwise correlations between TD and FD/PD were low and non-significant, whereas FD and PD metrics were instead positively and significantly correlated. All facets of α diversity decreased from weakly to strongly impacted lakes owing to the removal of sensitive species harboring unique evolutionary legacies and phenotypes. By contrast, the three facets of β diversity responded inconsistently to anthropogenic disturbance: while FDβ and PDβ showed significant impairment in moderately and strongly impacted lakes as a result of spatial homogenization, TDβ was lowest in weakly impacted lakes. The multiple facets of diversity also responded differently to the underlying environmental gradients, re-emphasizing that taxonomic, functional and phylogenetic diversities provide complementary information on community dynamics. However, the explanatory power of our machine learning and constrained ordination models was relatively low and suggests that unmeasured environmental features and stochastic processes may strongly contribute to macroinvertebrate communities in floodplain lakes suffering from variable levels of anthropogenic degradation. We finally suggested guidelines for effective conservation and restoration targets aimed at achieving healthier aquatic biotas in a context of increasing human impact across the ‘lakescape’ surrounding the Yangtze River, the most important being the control of nutrient inputs and increased spatial spillover effects to promote natural metasystem dynamics.