Ecological qualities and resources in coasts are threatened by various human activities, such as pollution and fishery. Impact evaluation of environmental stressors over a wide coastal stretch has been limited due to lack of efficient and standardizable biodiversity monitoring and assessment tools. Integrating environmental DNA (eDNA) and ecological traits, a holistic approach was developed to assess the impact of pollution and aquaculture on fish biodiversity in Chinese coastal areas. Taking the Yalujiang Estuary (YLJK) from the Yellow Sea and the Nan'ao Island Area (NAO) from the South China Sea as cases, the performance of the eDNA biomonitoring workflow was validated. First, the eDNA results of 22 sampling sites reached more than 85% of the asymptotes of species or ASVs in each area. A total of 115 fish species in both areas were detected and NAO was 1.8 times richer than YLJK using eDNA and the fish eDNA composition was consistent with the historical data. eDNA recovered distinct variations of fish sequence, taxonomic and functional diversity, and the corresponding trends following the offshore distance between the two areas. Fish sequence diversity was decreased primarily by estuarine pollution factors (chemical oxygen demand and zinc) in the YLJK. Compared with no breeding areas, lower fish sequence diversity was in breeding areas in the NAO. By integrating ecological traits, the eDNA approach offers promising opportunities for future fish biodiversity monitoring and assessment in national and global coastal environments.