To evaluate the fish diversity of national aquatic germplasm resource reserves in Guangdong Province at the molecular level, we collected 550 fish samples from 12 reserves, which were initially identified morphologically as 69 species belonging to 53 genera, 16 families, and 9 orders; by performing DNA barcoding sequencing of the mitochondrial COI gene and combining phylogenetic tree, genetic distance, and haplotype network analyses, a DNA barcode reference database for fishes in the reserves was constructed, with results showing that the species identification success rate of this database reached 98.55% (68/69), confirming its reliability; species delimitation analysis reveals that 7 species including Cyprinus carpio, Cultrichthys erythropterus, and Hemibagrus macropterus formed 2–4 molecular operational taxonomic units (MOTUs) with large genetic distances between units, suggesting the potential existence of cryptic species diversity that may be related to factors such as drainage isolation, complex geological historical activities in South China, artificial release, and misclassification; in contrast, Culter recurviceps and C. alburnus, as well as Sarotherodon galilaeus and Oreochromis aureus, had small genetic distances and shared one MOTU, showing characteristics of recent divergence. In summary, we established a DNA barcode database for freshwater fishes in national aquatic germplasm resource reserves in Guangdong Province, uncovered cryptic diversity underestimated by morphological identification, and provided an important scientific basis for the subsequent protection and management of fish resources.

The Yuanjiang River is one of the most significant rivers in Yunnan, characterized by abundant fish resources; however, in recent years, human activities have led to a decline in these resources. We employed environmental DNA (eDNA) technology to investigate the fish species in the Yuanjiang River, providing baseline data for fish resource management and conservation. In October 2023, 16 sampling stations were established along the river to collect water samples, and fish diversity was analyzed using eDNA techniques. A total of 47 fish species (5 orders, 15 families, and 40 genera) were detected, including rare and endangered species such as Bagarius rutilus and Hemibagrus guttatus, as well as 8 invasive fish species. Alpha diversity analysis shows Shannon-Wiener index of 0.92–2.41, Simpson index of 0.44–0.88, Pielou evenness index of 0.35–0.60, and Margalef index of 0.30–1.16. Hierarchical cluster analysis and non-metric multidimensional scaling (NMDS) were applied to examine spatial distribution patterns of fish communities, showing distinct differences among stations that grouped into three clusters from downstream to upstream, with geographically proximate stations clustering together. Similarity percentage analysis (SIMPER) identified species contributing to inter-group differences, with Culter alburnus, Channa gachua, Cirrhinus molitorella, and Abbottina rivularis as the primary contributors. Redundancy analysis (RDA) assessed relationships between fish diversity and environmental factors, indicating that the Yuanjiang River fish community was primarily influenced by altitude, salinity, conductivity, total dissolved solids, redox potential, and total phosphorus. This study demonstrates that eDNA technology is effective for analyzing fish species composition and distribution in the Yuanjiang River and can serve as a crucial tool for biodiversity monitoring and conservation.