Trait‐Based Biomonitoring Using eDNA Metabarcoding to Assess Anthropogenic Disturbances on Freshwater Fish Communities

Various anthropogenic disturbances affect the succession of aquatic habitats along dendritic river networks. Bioindicator taxa, such as fish, can be used to assess the effects of these disturbances on habitat quality. Environmental DNA (eDNA) metabarcoding offers a novel approach to complement traditional sampling and analysis of bioindicator taxa. Here, we apply a trait‐based biomonitoring framework, focusing on fish tolerance to pollution, to assess habitat quality and fragmentation within two watersheds in southern Québec (Canada). We sampled 193 sites within the dendritic networks of the Châteauguay and St. François watersheds and estimated fish community tolerance indices on the basis of 12S metabarcoding. We found a significant correlation between the fish community tolerance index and environmental factors such as subwatershed land use, precipitation and elevation. We also found that river fragmentation caused by dams affected fish assemblages and native fish movement but also prevented the spread of the non‐native common carp. Finally, we applied random‐forest modelling to predict the tolerance of fish communities to disturbances in unsampled areas, providing a broader understanding of habitat quality within catchments. Our research highlights how eDNA metabarcoding for large‐scale biomonitoring and river fragmentation studies provides a cost‐effective and non‐invasive method for assessing fish biodiversity and riverine ecosystem health.