Agent-based models and mixed crop-ruminant livestock systems: advances, gaps, and implications for decision-making

Agent-based (AB) models have emerged as powerful tools for the simulation of agricultural systems, enabling researchers and policymakers to explore complex interactions among biophysical processes, human decision-making, and socio-ecological dynamics. In mixed crop-livestock systems, AB models hold particular promise for evaluating alternative management strategies, improving resource use, and assessing trade-offs between economic performance and environmental outcomes. However, the application of AB models to these mixed systems remains limited, with significant gaps in representing behavioural diversity, feedback mechanisms, and structural complexity. This review synthesizes the current state of AB model applications in agriculture and identifies key gaps in their development and use for mixed crop-livestock systems. Using the Rural Futures Multi-Agent Simulation (RF-MAS) model as a case study, we highlight the strengths and limitations of current modelling approaches. We argue that future agent-based models must incorporate more realistic agent behaviours, system-level interactions, and context-specific constraints to better support decision-making in these integrated systems.