Source analysis of nitrogen pollution in basin by export coefficient modeling and microbial source tracking with mutual verification

Nitrogen pollution in rivers has long been a significant ecological and environmental concern, and research on nitrogen pollution source tracking serves as the foundation for pollution control, playing a crucial role in quantifying different pollution sources and formulating effective mitigation strategies. This study proposes a technical framework for pollution source resolution based on the export coefficient model and microbial source tracking model. Initially, key environmental factors and their spatiotemporal characteristics were analyzed to preliminarily identify potential nitrogen pollution sources, including wastewater treatment plants, stainless steel plants, electroplating factories, chemical plants, pig farms, poultry farms, rice fields, vegetable farms, and tea plantations. Subsequently, hydrochemical and microbial metagenomic analyses were conducted to further refine the identification of nitrogen pollution sources. NMDS analysis revealed significant differences in microbial community structures among different pollution sources, facilitating effective discrimination. Additionally, co-occurrence network analysis was employed to construct microbial fingerprint maps specific to each pollution source. Finally, a Bayesian community-wide non-culture microbial source tracking method (SourceTracker) was used for quantitative pollution source apportionment. The export coefficient model estimated that point-source nitrogen loads were primarily derived from domestic wastewater, whereas non-point source nitrogen loads predominantly originated from rural domestic wastewater and agricultural cultivation. By integrating the microbial source tracking model, the primary sources of nitrogen pollution were accurately identified. During the dry season, domestic wastewater (47.3%) was the dominant contributor, including wastewater treatment plants, rural domestic sewage, stainless steel plants, and electroplating factories, with fecal and agricultural sources mainly stemming from pig farms and rice fields. In contrast, during the wet season, agricultural cultivation (20.5%) and natural soil (27.8%) were the predominant contributors, encompassing rice fields, vegetable farms, and tea plantations. This source-tracking approach provides a valuable tool for guiding precise regional pollution control and is particularly applicable in complex pollution environments.