The ubiquity of the contamination was striking. Three compounds in particular, trimethoprim, chlortetracycline, and sulfamethoxazole, were detected in 100 percent of the samples analyzed, meaning that no single site in the entire river system was free of antibiotic residues. Concentrations of individual antibiotics ranged widely, from as little as 2.8 nanograms per liter for lomefloxacin to a maximum of 1235.8 nanograms per liter for erythromycin, a macrolide antibiotic commonly used in both human medicine and livestock production. While such concentrations are measured in parts per trillion and are far below therapeutic doses, chronic exposure at these levels is precisely the condition that scientists worry can drive the evolution and spread of antibiotic resistance genes in aquatic microbial communities.

Spatial patterns in the data told a clear story about how the pollution moves through the watershed. Antibiotic concentrations followed the order of midstream greater than downstream greater than upstream, suggesting that the middle reaches of the river system, where urban development and population density are highest, act as a major loading zone. Equally telling was the comparison between the main river channel and its tributaries: levels in the tributaries were generally higher than in the mainstream, indicating that the main stem of the Beiyun River is strongly influenced by inflows from the smaller streams draining its sub-catchments. In effect, the tributaries function as conveyor belts, collecting antibiotic residues from neighborhoods, hospitals, livestock facilities, and industrial sites, and delivering them into the larger river.

To move beyond simple detection and identify where the antibiotics actually come from, the research team employed a quantitative source-apportionment method known as principal component analysis with multiple linear regression, or PCA-MLR. This technique, originally developed for tracing air pollution sources, works by reducing a large dataset of measured pollutant concentrations into a small number of underlying statistical factors, each of which corresponds to a characteristic pollution signature. Regression analysis then estimates how much each factor contributes to the observed concentrations of individual compounds. In this study, the model resolved the antibiotic load in the Beiyun River system into six identifiable pollution sources plus one unknown source, providing a quantitative breakdown that single-source monitoring cannot achieve.

The results of the apportionment pointed to two dominant contributors. Mixed pollution source inputs, a category that aggregates diffuse urban runoff, agricultural activity, and other combined pathways, explained 35.4 percent of the variance in the data, making it the single largest source. Wastewater treatment plant discharges ranked a close second, accounting for 31.1 percent of the variance. The chemical fingerprints of these two sources were distinct. The mixed source contributed most heavily to the quinolone antibiotics difloxacin and ciprofloxacin, accounting for 72.1 percent and 71.4 percent of their respective loads. The wastewater treatment plant source, by contrast, dominated the sulfonamide class, contributing 74.8 percent of sulfacetamide, 73.9 percent of sulfachloropyridazine, 73.3 percent of sulfadimethoxine, and 70.9 percent of sulfadiazine.

The prominence of wastewater treatment plants as a source is a recurring theme in urban water pollution research, and the Beiyun findings add Beijing to a long list of cities where conventional treatment infrastructure falls short of removing pharmaceutical micropollutants. Standard activated-sludge treatment processes are designed to strip out organic matter, nitrogen, and phosphorus, but many antibiotics are either poorly biodegraded or are re-released back into the water column as parent compounds or metabolites during sludge handling. Sulfonamides in particular are known to pass through treatment barriers with relatively high efficiency, which is consistent with the strong association between these compounds and treatment plant discharges found in the new study. Upgrading plants with advanced oxidation processes, ozonation, or activated carbon polishing has been shown elsewhere to substantially reduce pharmaceutical loads, and the source apportionment here suggests such upgrades would deliver measurable benefits in this watershed.

On the question of ecological risk, the study’s multi-level assessment delivered a nuanced verdict. Overall, the ecological risks posed by antibiotics across the Beiyun River system remained at acceptable levels, a conclusion that may surprise readers given the universal detection of the compounds. Risk quotients, calculated by comparing measured environmental concentrations with predicted no-effect concentrations for sensitive aquatic species, stayed within thresholds regarded as tolerable for the system as a whole. However, the assessment flagged specific exceptions. Tetracyclines showed relatively high risk quotients, reflecting the combination of their persistent detection and the known sensitivity of algae and other primary producers to this drug class. Clarithromycin, a macrolide, exhibited low but non-negligible risks to certain aquatic species, consistent with prior toxicological work showing that macrolides can affect photosynthesis and growth in freshwater organisms at environmentally relevant concentrations.

The significance of these findings extends well beyond the boundaries of a single watershed. Antibiotic residues in surface water are widely recognized as a key driver of antimicrobial resistance, one of the most serious public health threats of the coming decades. When bacteria in rivers are exposed to sub-inhibitory concentrations of antibiotics over long periods, strains carrying resistance genes gain a selective advantage, and those genes can then move between species and environments through horizontal gene transfer. Urban rivers like the Beiyun, which flow through neighborhoods where millions of people live, work, and recreate, represent critical nodes in this process, linking clinical and agricultural antibiotic use to the broader environmental resistome. Understanding exactly which sources contribute which compounds is therefore a prerequisite for designing effective interventions.

The study also highlights the value of source apportionment as a policy tool. Knowing that mixed diffuse inputs and treatment plant effluents together account for roughly two-thirds of the antibiotic burden allows regulators to prioritize interventions, whether through upgrading treatment technology, controlling runoff from livestock and aquaculture operations, tightening prescription and disposal practices, or managing the hydrological connections between tributaries and the main stem. The finding that tributaries consistently carry higher concentrations than the mainstream suggests that targeted action on sub-catchment scale pollution could yield disproportionate benefits for the entire river system.

For Beijing, a megacity of more than twenty million people whose rivers have undergone decades of engineering and restoration, the new data provide a baseline against which future management efforts can be measured. The research was funded by the Natural Science Foundation of Jiangxi Province, the Jiangxi Provincial Department of Science and Technology, and the National Natural Science Foundation of China, and was conducted by a team led from Jiangxi Agricultural University together with collaborators in Beijing. As antibiotic use continues to grow worldwide and urbanization intensifies across Asia, Africa, and Latin America, the Beiyun River system offers a case study in what complete pharmaceutical contamination of an urban watershed looks like, and a demonstration that with careful chemical detective work, the sources of that contamination can be named, ranked, and ultimately addressed.

Subject of Research: Antibiotic contamination, sources, and ecological risks in an urbanized Beijing watershed

Article Title: Occurrence, source apportionment, and risk assessment of antibiotics in the Beiyun River system: a highly urbanized watershed in Beijing, China

Article References: Lou, Q., Zhao, X., Zhang, X., Liang, L., Han, L., Tu, W., & Chen, M. (2026). Occurrence, source apportionment, and risk assessment of antibiotics in the Beiyun River system: a highly urbanized watershed in Beijing, China. Environmental Monitoring and Assessment, 198(10), Article 1125. https://doi.org/10.1007/s10661-026-15933-7

Image Credits: AI Generated

DOI: 10.1007/s10661-026-15933-7

Keywords: antibiotics, Beiyun River, Beijing, water pollution, source apportionment, PCA-MLR, ecological risk, wastewater treatment, sulfonamides, tetracyclines, macrolides, quinolones