Hydrological Connectivity Enhances Fish Biodiversity in Amazonian Mining Ponds: Insights From <scp>eDNA</scp> and Traditional Sampling

Artisanal and small-scale gold mining (ASGM) expansion in the Madre de Dios region of the Peruvian Amazon has transformed primary forests into a novel wetland complex of thousands of abandoned mining ponds. Despite their ecological relevance, post-mining recovery of these systems remains understudied, particularly regarding fish biodiversity and recolonisation. In this study, we evaluate fish community richness and composition in mining ponds of different dimensions, years post abandonment, physicochemical properties and degree of pulse flood connectivity using traditional collection-based methods and environmental DNA (eDNA) with the 12S and COI markers. We compared these two methods of biodiversity inventory and contrasted results from ASGM waterbodies with those obtained from nearby pristine oxbow lakes. Overall, we registered more fish richness at all sites using eDNA versus traditional methods, especially with the 12S marker. We identified 14 and 13 unique genera using traditional methods and eDNA, respectively, with 40 genera detected by both approaches, evidencing their complementarity. Notably, we found that the degree of pulse flooding connectivity was the main predictor of species richness among the abandoned mining ponds (p-value < 0.05). We registered 11-22, 23-71 and 56 morphospecies in non-flooded mining ponds, pulse flooded mining ponds and nearby oxbow lakes, respectively. Furthermore, the fish community composition of mining ponds most influenced by pulse flooding was similar to that of pristine lakes. Our findings highlight the role of hydrological connectivity in ecological recovery within mining-impacted wetlands. Future restoration efforts should enhance aquatic connectivity to accelerate recovery in post-mining environments.