Analysis of mathematical modelling approaches to capture human behaviour dynamics in agricultural pest and disease systems

The integration of dynamic human behaviour into epidemiological models aims to improve the effectiveness of pest and disease management in crops and livestock, crucial for sustainable food and agriculture systems, especially amid current challenges such as climate change, globalisation and chemical use. This structured scoping review focuses on how dynamic human behaviour is integrated into mathematical models for pest and disease control, using mathematical frameworks such as game theory and agent-based modelling, providing a comprehensive analysis of literature from the last decade. To identify and assess relevant studies, an extensive and systematic search for literature was conducted using the Web of Science database, followed by a manual screening process. Most studies focused on stochastic and spatially explicit types of models that capture the complexity of agricultural pest and disease systems. Various control strategies, including chemical control, cultural methods, and integrated pest management, are identified, highlighting the presence of context-specific approaches in epidemiological models, considering diversity of agricultural systems and heterogeneity in farmer behaviour. The review also identifies gaps in current research, such as the limited focus on dyadic behaviours (e.g. farmer-to-advisor interactions), limited interdisciplinary collaboration, and reliance on secondary data sources. By addressing these gaps, a more comprehensive and practical understanding of the dynamics within agricultural systems, particularly in relation to human behaviour can be achieved. Modellers are encouraged to foster interdisciplinary work with social scientists and collection of primary social data to better capture human behaviour for modelling agricultural pest and disease systems. • Models yield more realistic insights when including collective behaviour. • Decision-making overshadows other behavioural factors, e.g. willingness. • Limited interdisciplinary input from social sciences hinders model insights. • Diverse delayed control strategies require context-specific modelling. • Spatial stochastic models better fit agricultural pest and disease trends.