An innovative modeling approach of linking land use patterns with soil antibiotic contamination in peri-urban areas.

Due to the intensive use and continuous release, high and persistent concentrations of antibiotics are found in soils worldwide. This severe contamination elevates the risks associated with antibiotic exposure and resistance for soil ecosystems and human health. Estimating antibiotic concentrations in soils is a complex and important challenge because the limited information is available on antibiotic use and emission and the high exposure risk to human health occurred in peri-urban areas. In this study, soil antibiotic contamination was linked with land use patterns in a data-scarce peri-urban area in four different seasons, and we established a modeling framework based on land use to estimate spatially explicit distribution of antibiotics in soils. The soil antibiotic concentration was found to be substantially affected by surrounding land use patterns in buffer zones with a radius of 350 m. Agricultural land was the main source of antibiotics entering the soil. Notably, road networks also had considerable impacts on antibiotic residues in soils. Then, a statistical model was developed in describing the linkage between land use patterns and soil antibiotic concentration. Model evaluation suggested that the proposed model successfully simulated the variation of antibiotics in soil with good statistical performance (R2 > 0.7). Finally, the model was extrapolated to investigate detailed distribution of antibiotics in soils. Clear spatial and seasonal dynamics can be found in soil antibiotic concentration. To our knowledge, this was the first attempt to adopt a model focusing on land use pattern to estimate the spatially explicit distribution of antibiotics in soils. Despite of some uncertainties, the research provides a land-use-based modeling approach as a reference for preventing and controlling soil antibiotic contamination in the future.