Coupling pesticide emission and toxicity characterization models for LCA: Application to open-field tomato production in Martinique

Abstract

The environmental evaluation of fruits and vegetables in life cycle assessment (LCA) requires a proper estimation of pesticide emissions and associated (eco-)toxicity impacts. In response, we developed an approach to consistently combine state-of-the-art emission inventory and impact assessment models for assessing human toxicity and freshwater ecotoxicity impacts from pesticide applications, and tested our approach in an LCA case study on pesticides applied to an open-field tomato produced in Martinique (French West Indies). Our results show that impact scores vary over several orders of magnitude, mainly as function of differences in pesticide properties and application time in relation to crop growth stage. Overall, impacts related to pesticide field emissions leading to exposure to pesticide residues in crop harvest are a main contributor to LCA performance results for tomato produced in Martinique, with fertilizer and packaging manufacturing as other dominating aspects. While the proposed approach is applicable to refine currently LCA methods for assessing pesticides, large uncertainties remain. These are mostly related to the parametrization of impact factors for tropical species. Based on our findings, we recommend using initial emission distribution fractions in combination with steady-state characterization factors for environmental emissions and with time-dependent characterization factors for pesticide residues in crop harvest in LCA, while further improving the use of secondary emission fractions to allow for better consideration of local field, soil and climate characteristics.