With the decrease in fish diversity, the monitoring of fish has become a current research hotspot. Here, the fish diversity and spatial distribution along a salinity gradient (0.28‰–3.32‰) in Xinglinwan Reservoir, China, were investigated using eDNA and the TSM. With the combination of both methods, 12 species from 6 orders, 9 families, and 11 genera were observed. The invasive species Oreochromis niloticus had the highest relative abundance, accounting for 56.91% in the TSM and 89.80% in eDNA, respectively. eDNA detected 83.33% of the species detected using the TSM, and an additional six species (50.0% of the total species) were detected via only eDNA. Moreover, eDNA supplemented the unavailable abundance data from the TSM. The TSM revealed that the diversity and relative abundance of fish inhabiting seawater and brackish water increased with an increase in salinity, but this was not found using eDNA. Thus, our study highlights that Xinglinwan Reservoir has a low species richness and is suffering from the great harm of ecological invasion. Our findings also show that eDNA could be reliable for providing a comprehensive picture of fish, but discriminating the spatial heterogeneity of fish communities is a challenge for its widespread adoption in lentic brackish systems.

Non-native fish species introduced into new areas, especially when they develop into large populations, pose a threat to native fauna. Understanding the current status of the fish community and invasion risks of non-native fish are essential for invasive species control and diversity conservation. The community structure of fish and a risk assessment on 10 non-native fish species were systematically assessed in the Jiulong River Basin, China, in January, April, and July of 2024. The species richness, with 105 species, showed a notable decrease compared to the 124 species recorded in 1975, while the number of non-native species has increased from zero to ten. Furthermore, the non-native fish species, Coptodon zillii and Sarotherodon galilaeus, have become dominant species, with IRI values of 4038.43 and 1180.30, respectively. The AS-ISK established risk thresholds for BRA and BRA + CCA as 29.5 and 35.5, respectively, identifying 70% of the non-native fish species as high-risk species, including C. zillii, S. galilaeus, Oreochromis niloticus, Clarias batrachus, Hypostomus plecostomus, and Oreochromis aureus. This study indicates that the fish species richness in the Jiulong River Basin has declined, with C. zillii and S. galilaeus becoming dominant and posing high ecological risks to the native fish community. In addition, targeted fishing during the breeding season should be used to control the population of tilapia and restore fish diversity.

A total of 32 fish species were detected in Xinglin Bay using a combination of environmental DNA metabarcoding (eDNA) and traditional morphological survey methods (TSM), covering eight orders, fifteen families, and twenty-six genera. The dominant order was Perciformes, accounting for 43.75% of the total species. Among the identified species, there were ten non-native fish species. Compared with the TSM, the eDNA detected 13 additional fish species, including two additional non-native fish species—Gambusia affinis (Baird and Girard, 1853) and Micropterus salmoides (Lacepède, 1802). In addition, the relative abundance of fish from both methods revealed that tilapia was overwhelmingly dominant, accounting for 80.75% and 75.68%, respectively. Furthermore, the AS-ISK assessment revealed that all non-native fish species were classified as medium or high-risk, with five identified as high-risk species, four of which belong to tilapia. These findings demonstrated that tilapia are the dominant and high-risk invasive species in Xinglin Bay and should be prioritized for management. Population reduction through targeted harvesting of tilapia is recommended as the primary control strategy. Additionally, the study highlights the effectiveness of eDNA in monitoring fish community structure in brackish ecosystems.