Simulation of Regulation Policies for Fertilizer and Pesticide Reduction in Arable Land Based on Farmers’ Behavior—Using Jiangxi Province as an Example

Abstract

A multi-agent model for the simulation of arable land management based on the complex adaptive system theory and a Swarm platform was constructed. An empirical application of the model was carried out to investigate the pollution of arable land in Jiangxi Province. Two sets of policies—a fertilizer tax and an ecological compensation scheme—were designed and simulated, and the analysis focused on the control of polluting inputs, mainly chemical fertilizers and pesticides. The environmental effects of each policy were evaluated by simulating farmers’ self-adaptive behaviours in response to the policy in the artificial village of the model. The results showed the following: (1) Both the fertilizer tax policy and the ecological compensation policy somewhat alleviated the negative impact of input factors, such as fertilizers and pesticides, on arable land; (2) if the fertilizer tax policy is implemented, the medium tax rate scheme should be given priority—the effect does not necessarily improve as the tax rate increases, and a high-tax policy will threaten food security in the long term; and (3) if an ecological compensation policy is implemented, high-government-compensation scenarios are better than low-government-compensation scenarios, and the differential-government-compensation scenario is better than the equal-government-compensation scenario, and the differential-government-compensation scenario can lighten the burden on the government.