Quantifying farmer decision-making in an agent-based model

Nutrient pollution is a major issue in the Mississippi watershed and the Gulf of Mexico. Agriculture in the Midwest is attributed as a major non-point source contributor to this problem. The implementation of best management practices by farmers could greatly reduce the amount of nutrients released from farms into surrounding watersheds. Determining what motivations farmers need in order to implement these practices is a necessary but challenging step in reducing nutrient pollution. A computer model, NitroShed, was created to model the farmer decision-making process and how policy influences could impact adoption rates of best management practices. The model was developed using the agent-based model package Mesa in Python. The model includes a farmer decision-making algorithm to simulate the behavior of farmers considering investments in environmental infrastructure and management practices. The model includes a farmer typology based on the factors farmers consider when they are making decisions. The identified typologies were: Business Oriented, Environmentally Oriented, Innovators, Traditionalists, and Supplementalists. Each group of farmers is unique in the way they consider factors such as risk, social expectations, economics, innovation, and the environment. The model was then run under varying conditions to test adoption rates based on policy changes and financial influences. This will help policy-makers and extension services determine the most effective action plan in increasing farmer adoption of best management practices.