Comparison between eDNA and traditional morphological methods for fish diversity monitoring in rivers

Environmental DNA (eDNA) is increasingly employed for biodiversity monitoring, yet its accuracy and consistency relative to conventional fish surveys remain under debate. Here, we evaluated fish assemblages at 14 sites across six major rivers in Liaoning Province, China, using four primer sets (12 S rRNA, COI, MiFish, and a high-variation MiFish) coupled with high-throughput sequencing, and compared results with capture-based surveys. eDNA detected 211 species across 17 orders, 71 families, and 146 genera-substantially exceeding traditional methods-but failed to identify nine species recorded morphologically and could not provide ecological attributes such as body size or population structure. Species richness and composition varied between tributary and mainstem sites, reflecting hydrological and habitat gradients. Detection efficiency also differed among primers, underscoring the importance of multi-primer strategies. Overall, eDNA provides a sensitive and efficient complement to conventional surveys, but its limitations in taxonomic resolution and ecological inference highlight the need for integration of molecular and conventional morphological methods to achieve comprehensive fish diversity assessments.