To investigate the pollution characteristics of ARGs in an aquaculture pond, this study utilized metagenomic techniques to detect and analyze the presence of ARGs and their interactions with microbial communities and environmental variables. The results identified 21 categories of ARGs, comprising 1,092 subtypes, with pond sediments serving as the primary reservoir of ARGs. The antibiotic resistance genes macB, tetA(58), and nov exhibited the highest relative abundance, with multidrug resistance and efflux pumps being the dominant ARG types and resistance mechanisms. There was a significant difference in microbial community composition between pond water and sediments (P < 0.05). Several genera within the dominant phyla Proteobacteria, Cyanobacteria, and Actinobacteria showed significant positive correlations with various ARGs, indicating the potential for horizontal transfer of ARGs among these bacterial taxa. Microbial communities further facilitated the persistence of ARGs within the system and their spread into the water column. Additionally, ammonium nitrogen, nitrite nitrogen, and nitrate nitrogen were identified as key environmental factors influencing the distribution of ARGs and the composition of microbial communities in the aquaculture ponds. This study establishes potential baseline levels of ARGs in eco-recirculating aquaculture systems and provides quantitative data indicating a high risk of ARG dissemination in effluent waters. The findings offer valuable insights for optimizing aquaculture systems, reducing ARG pollution, and protecting the aquaculture environment.