A modeling framework of a territorial socio-ecosystem to study the trajectories of change in agricultural phytosanitary practices

International audience

English. Despite the growing societal demands to reduce pesticide use, public policies struggle to reverse the current upward trend. Agroecology emerges as a promising solution, as it promotes ecological regulation and sustainable practices. The systemic nature of the agro-ecological transition requires the development of an interdisciplinary approach. In this respect, models integrating ecological, economic, and social aspects are valuable for understanding the dynamics of agroecosystems. The objective of this article is to investigate the trajectories of change in the pesticide practices of agricultural territories taking into account ecological and economic dynamics. We built a socio-ecological modeling framework of a socio-ecosystem at the territorial scale. This generic agent-based model is developed to simulate both farmers’ decision-making (based on its own profit and that of others, and its willingness to change) and disease dynamics (based on the inoculum, the probability of infection, the probability of disease escape, and the previous disease). We first used the model to simulate the impact of different economic or ecological territorial characteristics on the trajectories of agricultural practices in the territory. Second, we aimed to analyze the final states of the territories (at equilibrium) based on scenarios of variations in the economic (the cost of pesticides), ecological (the dispersal capacity of the disease), and agronomic (the ability to escape the disease of no-input farming) parameters. The final states of the territories were analyzed using four categories of indicators (farming practices, the profits, the number of infected fields and the use of fungicides). The study revealed strong threshold effects, non-linear effects and linear effects, on the number of farmers performing the different practices in the territory. These effects are highlighted respectively for the scenarios of increased cost of pesticides, increased disease escape of no-input farming and increased the disease dispersal. Our results highlighted the need to take into account combinations of levers and to study trajectories of change in order to promote sustainable agriculture. Finally, we discussed the possibility of using such models to guide public policies in favor of agroecology.