Pierlot Frédéric, Marks-Perreau Jonathan, Soulé Emma, Keichinger Olivier, Bedos Carole, Prevost Laetitia, Van Dijk Paul, Bockstaller Christian
LAE, Université de Lorraine, INRAE, F54000 Nancy, France; Chambre Régionale d'Agriculture Grand Est, Laxou, France.
Arvalis, Institut du végétal, France.
Sci Total Environ. 2023 Apr 20;870:161000. doi: 10.1016/j.scitotenv.2022.161000. Epub 2023 Jan 20.
Stakeholders involved in actions to reduce the use and the impacts on the environment or human health of pesticides need operational tools to assess crop protection strategies in regard to these impacts. I-Phy3 brings together all improvements introduced since the first version of the indicator to better meet user's needs and requirements of integrating processes. I-Phy3 was deeply modified to ensure its predictive quality. I-Phy 3 is structured in three levels of aggregation in form of hierarchical fuzzy decision trees designed with the CONTRA method. At the 1st level, five basic subindicators assess the risk of contamination (RC) for the different transfer pathways involved in surface water, ground water and atmosphere contamination: leaching, runoff, drainage, drift, volatilization. At the 2nd level, RC subindicators are aggregated with a toxicity variable (human or aquatic) in a risk indicator. At the 3rd level, the global indicator I-Phy3 results from the aggregation of three risk indicators for groundwater, surface waters and air. I-Phy3 yielded better validation results than its previous versions. This effort to assess the predictive quality of the indicator should be pursued and completed by a feasibility and utility test by end-users. A subindicator on risk of soil contamination is a gap which remains to fill.
参与减少农药使用及其对环境或人类健康影响行动的利益相关者需要实用工具,以便从这些影响的角度评估作物保护策略。I-Phy3整合了自该指标第一版发布以来引入的所有改进措施,以更好地满足用户需求和整合流程的要求。I-Phy3经过深度修改以确保其预测质量。I-Phy 3采用CONTRA方法设计的分层模糊决策树形式,分为三个汇总级别。在第一级,五个基本子指标评估地表水、地下水和大气污染所涉及的不同传输途径的污染风险(RC):淋溶、径流、排水、漂移、挥发。在第二级,RC子指标与毒性变量(对人类或水生生物)汇总为一个风险指标。在第三级,全球指标I-Phy3由地下水、地表水和空气的三个风险指标汇总得出。I-Phy3的验证结果优于其先前版本。应通过终端用户进行可行性和实用性测试,继续并完成评估该指标预测质量的工作。土壤污染风险子指标是一个有待填补的空白。