Abstract Pollution is a widespread threat to freshwater biodiversity, yet its impacts are typically evaluated/conceptualized as local processes, overlooking the role of dispersal in shaping landscape outcomes. Dispersal can propagate local impacts to the metacommunity by promoting pollutant‐tolerant species or attenuate impacts by a rescue effect from less impacted communities. Scaling up the local effects of pollutants on regional biodiversity requires understanding these effects. We considered two mechanisms linking pollution to species performance: an abundance effect that reduces all species abundances and a filtering effect with a trait‐based impact of the pollutant that promotes tolerant species. The consequences of these local effects on regional biodiversity were evaluated using spatially explicit metacommunity models parametrized for the Maldonado stream basin (Uruguay) and considering three functional species groups with contrasting dispersal capacities. We found that landscape connectivity can buffer or amplify biodiversity loss depending on the dominant mechanism. The decline in diversity due to a reduction in total community abundance is attenuated by connectivity, due to a rescue effect of dispersal. However, when pollutants act through species‐specific tolerance differences, connectivity promotes the dispersal and local dominance of tolerant species, producing a widespread decline in biodiversity. In addition, abrupt spatial changes in biodiversity, not expected from pollutant concentration, are frequently observed. Consequently, the distance to the pollutant source or its concentration can be a poor predictor of its impacts. Synthesis and applications : The results highlight key features for the assessment of pollutants risks to biodiversity. Landscape connectivity, species dispersal capacities, trait‐mediated shifts in competition, and reduction in community abundance determine the final effect of pollutant and biodiversity. Explicitly considering these features might be crucial in biomonitoring and risk assessment, which should not be limited to experimental bioassays or pollutant concentration in the environment. These results advance the role of metacommunity principles in both impact assessment and restoration planning.
León et al. (2026) studied this question.