A high-precision prediction for spatiotemporal distribution and risk assessment of antibiotics in an urban watershed using a hydrodynamic model

Abstract

A methodology for the high-precision prediction and risk assessment of antibiotics at the watershed scale was established. Antibiotic emission inventory and attenuation processes were integrated into the MIKE 11 model to predict the spatiotemporal distribution of norfloxacin, ofloxacin, enrofloxacin, erythromycin, roxithromycin, and sulfamethoxazole in the Nanfei River watershed, China. Considering the variations in antibiotic removal in sewage treatment plants, manure composting, and lagoon systems, the high, medium, and low removal efficiencies of selected antibiotics across China were obtained and used as the best, expected, and worst scenarios, respectively, to evaluate the uncertainty of antibiotic emissions. The predicted concentrations were comparable to antibiotic measurements after flow calibration. The prediction results showed that the highest concentration exposures were mainly concentrated in urban areas with a dense population. Flow variations controlled the temporal distribution characteristics of antibiotics via the dilution effect, and the concentrations of antibiotics in the dry season were 3.1 times higher than those in the wet season. The median concentrations of norfloxacin and erythromycin ranged from 111.36 ng/L to 592.33 ng/L and 106.63 ng/L to 563.01 ng/L, respectively, which both posed a high risk to cyanobacteria and a medium risk to spreading antibiotic resistance. Scenario analysis further demonstrated that high removal efficiencies of these antibiotics can considerably reduce the potential ecotoxicity risks and bacterial resistance selection. The developed methodology for predicting the distribution and risk of antibiotics was suitable for the risk assessment and control strategy of human- and livestock-sourced pollutants.