Abstract: To explore the potential application of environmental DNA (eDNA) technology in the early warning of invasive fish species in urban lakes, we investigated three representative lakes in the Pearl River Delta region, Xing Lake in Zhaoqing, Shunfengshan Lake in Foshan, and Dongshan Lake in Guangzhou. All of them are heavily impacted by invasive fish. Water samples were collected from Xing Lake and Shunfengshan Lake in March 2023 and from Dongshan Lake in November 2024. Using eDNA metabarcoding, we monitored the status of invasive fish, provided early-warning assessments, and evaluated the effectiveness of eDNA detection. A total of 59 fish taxa were detected, including 13 invasive species, with 7, 8, and 10 species being identified in Xing, Shunfengshan, and Dongshan lakes, respectively. The relative abundance of invasive fish reached 44.45%, 47.19%, and 35.58% in the three lakes. Coptodon zillii, Oreochromis niloticus, and Pterygoplichthys pardalis exhibited high overall relative abundances (14.01%, 14.05%, and 6.64%) and were dominant species at certain sampling sites. In Xing Lake, the relative sequence abundance of invasive fish obtained via eDNA showed a significant positive correlation with both the relative abundance and relative biomass derived from traditional methods (R=0.89 and 0.91 respectively, P< 0.01), demonstrating a high reliability of eDNA for quantitative detection of invasive fish distributions. This study provides a scientific basis for the prevention and ecological management of invasive fish in urban lakes and confirms the efficiency and potential of eDNA technology in invasive fish species monitoring.

Key words: eDNA, non-native fish, early detection, urban lake