Deciphering the occurrence, distribution, and source apportionment of antibiotics in the Ningxia section of the Yellow River

Abstract The pervasive release of antibiotics into aquatic environments necessitates a thorough understanding of their fate. This study systematically investigated the occurrence, sources, and controlling factors of antibiotic pollution in the Ningxia section of the Yellow River. Across four major cities, 44 target antibiotics were monitored. Fluoroquinolones and macrolides were the predominant groups, with tylosin, enrofloxacin, and ciprofloxacin as the most abundant compounds. Concentrations exhibited significant spatiotemporal variation, peaking in the dry season (up to 150.97 ng/L) and clustering near intensive livestock areas and wastewater outfalls. Source apportionment via PCA-MLR quantified major contributions from combined medical/aquaculture and livestock sources (50.8%) and livestock farming alone (29.8%). Structural Equation Modeling identified livestock activity as the principal spatial driver, with significant positive paths to NH₄⁺-N and altitude, collectively explaining 68.2% of concentration variance. Furthermore, a stable predictive model (adj. R² = 0.833) was established, highlighting swine density as a key positive predictor and residential land proportion as a major negative predictor. This work concludes that antibiotic contamination is primarily anthropogenic, driven by livestock production with seasonal shifts between point and non-point pathways. The integrated methodology and mechanistic insights offer a valuable framework for developing targeted mitigation strategies in similar semi-arid river basins.