Mitigating pesticide mixture hazard in global surface waters through agricultural management

Abstract

Pesticides are extensively used to improve food production, resulting in ubiquitous pesticide contamination in surface waters due to agricultural runoff. However, pesticide mixture toxicity to global aquatic organisms is poorly understood due to their common co-occurrence and interactions, limiting mitigation strategies. Here, we develop machine learning models to generate spatially explicit maps of pesticide mixture hazards in global surface waters (5 arc-min resolution), utilizing geospatial environmental parameters and measured mixture toxicity of 311 pesticides detected at 2,338 surface water sites for a model invertebrate and alga. Our findings reveal hotspots of pesticide hazard primarily in Central China, South Asia, the USA, and East Africa, with agricultural activities identified as the primary driver (36%–42%). Furthermore, we determine optimal intervals for pesticide and manure applications to mitigate pesticide mixture hazards. Our findings provide a valuable foundation for decision-makers and farmers aiming to adopt sustainable agricultural practices and protect aquatic ecosystems in these hotspots.