Agent-based model for assessing ex-ante impacts of farm management innovations on performance of GL/DC-based smallholder systems - Design framework and theoretical parameterization

Abstract

Viable management and policy options for sustaining smallholder farming systems in grain legume and dry cereal production regions need special attention. Although a great deal of knowledge on ways to efficiently agronomic measures exists, too few studies seek to understand how agricultural policy, financial services, farming technologies, local capabilities interactively affect smallholders’ decision about farming system management. As a methodological opportunity, multi-agent system (MAS) or agent-based model (ABM) has been recently recognized as a promising approach for explaining complex human-environment interactions in agroecosystems. This report presents the concept, framework and theoretical parameterization of a MAS/ABM for the typical coupled community-landscape system that can be used for ex-ante assessment of long-term impacts of management and policy options on soil fertility, food productivity and profitability of smallholder agroecosystems in different geographic regions. The goal is to provide insights into appropriate strategies for promoting the viability of smallholder agricultural livelihoods over the long term.