Modeling farmer decision-making using an agent-based model for studying best management practice adoption rates: A typological approach

Nutrient pollution is a significant water quality issue causing algal blooms and hypoxia in watersheds, impacting local aquatic life, industries, and the health and safety of people. The Mississippi watershed and the northern Gulf of Mexico are examples where this is a particular problem. Agriculture is a major non-point source contributor to this problem. The implementation of best management practices (BMPs) by farmers could greatly reduce the amount of nutrient runoff into the surrounding watershed. For this to be effective, however, farmers must implement these practices. Understanding the factors that farmers consider when making decisions and the value system that farmers use to analyze potential practices is a necessary but challenging step in reducing nutrient pollution. A computer model, NitroShed, was created to model the decision-making process of farmers that results in the adoption of BMPs. NitroShed was developed using Python and the package Mesa to create an agent-based model for studying complex decision-making. This was done by using a decision-making algorithm to simulate the farmer considering investments in environmental infrastructure and selection of management practices. The model includes a typology based on the factors farmers contemplate when making decisions. These typologies are Business, Conventional, Environmental, Innovator, and Supplemental. Each group of farmers is unique in the way they value economics, social pressure, environmental protection, risk aversion, and innovation in agricultural methods. NitroShed was then validated through modeling adoption rates in an Iowa watershed. Finally, the model was used to test policy and financial incentives to influence adoption rates of desired practices. Results of the model show that each farmer type reacts differently to social and financial incentives for adopting BMPs. As farmer demographics change in the future, adoption rates may naturally increase or could be influenced by raising cost share assistance or implementing a fine. The use of the model in expanded research is possible due to a modular design and thorough documentation. The NitroShed model is applicable for further study of farmer decision-making and interfacing with hydrological and geographic modeling software.

Name thesis:degree_name

M.S.

Level thesis:degree_level

Thesis

Discipline thesis:degree_discipline

Agricultural & Biological Engr

Grantor

University of Illinois at Urbana-Champaign

Year dc:date

2020

Author dc:creator

• Zeman, Kendra R.

Contributors dc:contributor

• Rodriguez, Luis F.
• Bhattarai, Rabin
• Marshall, Anna-Maria

• Agent-Based Modeling, Farmer Decision-Making, Typology, Best Management Practices, Nutrient Pollution, Farmer Behavior, Farmer Adoption Rates, Object-Oriented Design

Statement dc:rights

• Copyright 2019 Kendra Zeman

Language dc:language

en

Handle dc:identifier

http://hdl.handle.net/2142/106273

OAI identifier oai:identifier

oai:www.ideals.illinois.edu:2142/106273