In order to evaluate the fishery resource conservation effects of the Haizhou Bay Marine Ranching and compare the monitoring and survey capabilities of environmental DNA （eDNA） technology and bottom trawling in complex habitats， simultaneous eDNA water sample collection （surface and bottom layers） and bottom trawl surveys were conducted in the artificial reef area and the control area of this sea area in spring （April） and autumn （October） 2024. Statistical analysis methods such as species composition analysis， diversity index calculation， Principal Coordinates Analysis （PCoA）， Similarity Percentage Analysis （SIMPER）， and Redundancy Analysis （RDA） were used to analyze the differences in fish community structure and their relationships with environmental factors. The results showed that： （1） The number of fish species detected by eDNA （28 in spring， 45 in autumn） was higher than that detected by bottom trawling （19 in spring， 22 in autumn）； （2） There was no significant difference in biological communities between the reef area and the control area in spring， while in autumn， the surface layer diversity in the reef area was significantly lower than that in the bottom layer， and the fish composition significantly differed between the two areas； （3） Water temperature and salinity were key environmental factors influencing the community distribution in autumn； （4） The two methods performed consistently in terms of community evenness， but eDNA exhibited primer bias and missed detections， while bottom trawling was limited by fishing efficiency and identification errors. The study indicates that eDNA outperforms bottom trawling in species detection efficiency， particularly in complex habitats， but cannot obtain biological samples； combining the two methods can compensate for their shortcomings and provide a more comprehensive understanding of fish community characteristics. This research provides a reference for future resource monitoring and ecological management of the Haizhou Bay marine ranching.