Divergence From Temperate Models: Pollution Dominance and Stochastic Assembly in a Continental Freshwater System...

Current understanding of how multiple stressors shape freshwater biodiversity at continental scales relies heavily on evidence from postindustrial temperate regions, potentially misrepresenting the dynamics of ecosystems facing rapid development and extreme environmental heterogeneity. Here, we present a standardized, continental-scale field survey of benthic macroinvertebrates across 502 sites in China to test the universality of established macroecological patterns. Contrary to the expectation that physical habitat and climatic gradients primarily drive community structure, we found that organic pollution, with a focus on oxygen-demand assays (COD and BOD), emerged as the single most influential variable, explaining the largest proportion of variation in taxonomic, functional, and phylogenetic diversity. Together with total nitrogen and total phosphorus loading, this chemical pressure largely overrode the explanatory contribution of land-use and climatic drivers, with stressor interactions characterized by asymmetric dominance rather than the synergistic amplification commonly predicted. Furthermore, despite steep environmental gradients that theoretically favor deterministic sorting, community assembly was overwhelmingly dominated by stochastic processes ( $\sim$ 86%, based on overall pairwise comparisons across all sites). We suggest this pattern is consistent with a "stochastic trap" hypothesis, potentially arising from high-frequency anthropogenic disturbances that weaken trait-environment matching. These findings indicate that biodiversity frameworks derived from stabilized, temperate systems may not generalize to the compressed modernity of developing regions. We propose that global conservation strategies should consider prioritizing the reduction of chemical bottlenecks, which may act as critical physiological constraints on freshwater resilience in the Anthropocene.