Agriculture is a complex social-ecological system with numerous interactions and feedbacks between policies, markets, farm management, landscapes, and ecosystems. Because of these interconnections, policy changes, societal trends, and environmental crises can have widespread knock-on effects that threaten the stability of the entire system. Agent-based models have become a valuable tool used for studying agricultural systems and providing policy advice. However, they often only consider one or few aspects of the complete social-ecological system. Here, we review 50 agent-based models and analyse which aspects of agricultural systems they include. We find that there has been significant work done in the last decade, both in monodisciplinary and interdisciplinary models. There is a particularly robust tradition of using agent-based models for economic impact analyses of policy changes. Many models also study environmental impacts of agriculture. However, ecological and biodiversity-oriented models continue to be largely disconnected from the rest of the agricultural modelling literature. Based on our review, we provide recommendations for future research in ecological, socio-economic, and social-ecological modelling of agriculture. Areas of possible improvement include simulating farm management and landscape dynamics in ecological models, risk management in economic models, and bidirectional human-nature interactions in social-ecological models. Building on these recommendations, we develop a concept for an integrated model that could be used to study the impacts of agricultural policy on both farms and biodiversity.
Vedder et al. (Fri,) studied this question.