Schmitt Walter, Auteri Domenica, Bastiansen Finn, Ebeling Markus, Liu Chun, Luttik Robert, Mastitsky Sergey, Nacci Diane, Topping Chris, Wang Magnus
Bayer CropScience AG, Environmental Safety, Monheim, Germany.
European Food and Safety Agency, Pesticides Unit, Parma, Italy.
Integr Environ Assess Manag. 2016 Jan;12(1):46-57. doi: 10.1002/ieam.1640. Epub 2015 Jun 3.
This article presents a case study demonstrating the application of 3 individual-based, spatially explicit population models (IBMs, also known as agent-based models) in ecological risk assessments to predict long-term effects of a pesticide to populations of small mammals. The 3 IBMs each used a hypothetical fungicide (FungicideX) in different scenarios: spraying in cereals (common vole, Microtus arvalis), spraying in orchards (field vole, Microtus agrestis), and cereal seed treatment (wood mouse, Apodemus sylvaticus). Each scenario used existing model landscapes, which differed greatly in size and structural complexity. The toxicological profile of FungicideX was defined so that the deterministic long-term first tier risk assessment would result in high risk to small mammals, thus providing the opportunity to use the IBMs for risk assessment refinement (i.e., higher tier risk assessment). Despite differing internal model design and scenarios, results indicated in all 3 cases low population sensitivity unless FungicideX was applied at very high (×10) rates. Recovery from local population impacts was generally fast. Only when patch extinctions occured in simulations of intentionally high acute toxic effects, recovery periods, then determined by recolonization, were of any concern. Conclusions include recommendations for the most important input considerations, including the selection of exposure levels, duration of simulations, statistically robust number of replicates, and endpoints to report. However, further investigation and agreement are needed to develop recommendations for landscape attributes such as size, structure, and crop rotation to define appropriate regulatory risk assessment scenarios. Overall, the application of IBMs provides multiple advantages to higher tier ecological risk assessments for small mammals, including consistent and transparent direct links to specific protection goals, and the consideration of more realistic scenarios.
本文介绍了一个案例研究，展示了3个基于个体、空间明确的种群模型（IBM，也称为基于主体的模型）在生态风险评估中的应用，以预测农药对小型哺乳动物种群的长期影响。这3个IBM在不同场景中分别使用了一种假设的杀菌剂（杀菌剂X）：在谷物中喷洒（普通田鼠，草原田鼠）、在果园中喷洒（田鼠，东方田鼠）以及谷物种子处理（林姬鼠，小林姬鼠）。每个场景都使用了现有的模型景观，其大小和结构复杂性差异很大。定义了杀菌剂X的毒理学特征，以便确定性的长期一级风险评估会得出对小型哺乳动物的高风险，从而提供了使用IBM进行风险评估细化（即更高层级风险评估）的机会。尽管内部模型设计和场景不同，但结果表明，在所有3种情况下，除非以非常高（×10）的剂量施用杀菌剂X，否则种群敏感性较低。局部种群影响后的恢复通常很快。只有在故意高急性毒性效应的模拟中发生斑块灭绝时，由重新定殖决定的恢复期才值得关注。结论包括对最重要的输入考虑因素的建议，包括暴露水平的选择、模拟持续时间、具有统计稳健性的重复次数以及要报告的终点。然而，需要进一步研究并达成一致，以制定关于景观属性（如大小、结构和作物轮作）的建议，以定义适当的监管风险评估场景。总体而言，IBM的应用为小型哺乳动物的更高层级生态风险评估提供了多个优势，包括与特定保护目标的一致且透明的直接联系，以及对更现实场景的考虑。
