Implementation and territorial application of a biogeochemical model to study the impact of plant protection products on water resources in the Veneto region

The widespread use of plant protection products (PPPs) in agriculture poses serious threats to soil and water quality, particularly in regions with intensive cultivation and hydrological vulnerability, such as the Veneto region in northeastern Italy, and stresses the need to support sustainable management practices mitigating non-point source contamination. This study tested a newly implemented version of the biogeochemical model EPIC (Environmental Policy Integrated Climate) to simulate and map the environmental fate and potential contamination provided by some selected PPPs (glyphosate and its metabolite AMPA, S-metolachlor, terbuthylazine, fluopicolide, dimethomorph, nicosulfuron, and copper sulphate) across the Veneto region from 2017 to 2023. The aim was to identify areas of heightened susceptibility and to support the development of site-specific sustainable management practices. The simulation identified S-metolachlor and terbuthylazine as the compounds with the highest potential for surface runoff losses, especially in lowland areas dominated by maize cultivation and low-infiltration soils. Fluopicolide and dimethomorph, in contrast, showed the highest vertical transport potential in vineyard areas with permeable soils, reflecting their physicochemical properties and high GUS index values, despite overcoming the limitations of using a single index by incorporating site-specific findings tailored to pedoclimatic and management conditions. Glyphosate, although generally considered immobile, exhibited localized leaching under specific conditions due to the implementation of a specific code to simulate preferential flow pathways, that was not present in the original EPIC, providing consistent results with field observations. Risk maps were generated by combining modelled environmental concentrations with compound-specific toxicity thresholds, enabling the identification of specific areas where the risk of environmental contamination is more or less pronounced. Notably, the model’s outputs were consistent with recent regulatory actions, such as the bans on S-metolachlor and dimethomorph, thereby reinforcing the reliability of our modelling approach. Overall, the results demonstrated EPIC powerful and useful as regional-scale decision-support tool for assessing the environmental risk derived from using PPPs. While limitations remain (including the use of average input data and the absence of direct field validation) the model offers a robust framework for spatial risk assessment and prioritization, particularly in data-scarce contexts, as well as for supporting farmers and policymakers in evaluating the effectiveness of proposed mitigation strategies.