基于多主体模型的种植业面源污染控制模拟：化肥税-环境服务付费功效比较

构建了一个基于复杂适应系统理论和Swarm 平台的用于农村环境管理模拟的多主体模型. 针对我国种植业面源污染问题，对模型应用进行了实证研究：以控制化肥为主的污染性生产投入为重点，设计和模拟了实施化肥税和环境服务付费两套政策，通过模拟模型的人工村落中农户自适应行为对政策的响应，评估了政策的环境绩效. 研究表明，目前并不是实施化肥税的最佳时机，且仅依靠化学税政策难以达到理想的面源污染控制目标，若实施化肥税，宜优先考虑中等税率方案；环境服务付费政策应从分级设计补偿标准，合理确定环境目标，引入市场激励机制三方面入手降低补偿成本，并注意平衡环境安全和粮食安全的关系. 该多主体模型为我国农村环境管理决策支持提供了新的思路和方法.

Simulation on non-point pollution control from crop production based on a multi-agent model：Comparative research between fertilizer tax and payments for environmental services

This paper constructed a multi-agent model for rural environmental management simulation based on complex adaptive system theory and Swarm simulation platform, and focused on the control of agricultural production inputs. Using the multi-agent model, two different control measurements, including fertilizer tax and payments for environmental services, were designed. Their environmental performance was evaluated through model simulation of farmers' behavioral responses to these policies. The results indicate that it is proper to use the middle-rate fertilizer tax plan. However, it is not the best time to implement fertilizer tax. Moreover, fertilizer tax policy is not enough to reach a satisfied target. As for payments for environmental services, it should keep the balance between environmental protection and food security. In order to reduce the compensation cost, different payment standards should be set according to the input level. Reasonable environmental goals and market incentives should also be taken into consideration. The study shows that multi-agent model is an effective tool to provide decision-making support on rural environmental management.