Freshwater fish community assessment using eDNA metabarcoding vs. capture-based methods: differences in efficiency and resolution coupled to habitat and ecology.

Environmental DNA (eDNA) metabarcoding has revolutionized ecological and environmental research by describing communities without relying on direct observations, making it a powerful, non-invasive, and cost-effective tool in biodiversity monitoring. However, the adoption of eDNA as a standard protocol in long-term monitoring programs, which have traditionally relied on capture-based methods, presents challenges in terms of data comparability. Here, we compared freshwater fish communities assessed through eDNA metabarcoding and electrofishing, across 35 sampling sites in the lower Tagus River basin, Portugal. For the majority of species or species-groups analysed individually (13 out of 17), a significant correspondence was observed between electrofishing and eDNA metabarcoding detections. A weaker correspondence was found between the number of specimens captured by electrofishing with the number of eDNA metabarcoding reads, with seven out of 13 taxa showing significant relationships. Species richness estimates based on the two methods were very similar at the basin level. The methods yielded significantly different species compositions, although these differences were driven by samples collected in the Tagus main channel, which is wider and has higher flow rates than tributaries. Benthic and shoreline fish communities showed similar species composition in the two methods, but this was not the case for pelagic communities, probably due to the higher water turnover of the pelagic zone and electrofishing inefficiency. Our results highlight the high potential of eDNA metabarcoding as a complementary method to electrofishing for freshwater fish monitoring, though further validation is needed to assess biases related to site-specific hydrological conditions and the ecology of the target species.