Food web-based simulation for agroecology

Ecosystems are increasingly manipulated for agricultural and conservation goals. Ecosystem functions need to be sustained socially and ecologically. New frameworks must be built to simulate agrosystems based on ecological processes instead of external chemicals. Food web structures of agrosystems highly influence their agronomical performance and stability. Although it has been observed that living communities are ruling the performance of agroecosystems, these living communities are generally ignored by agronomists who focused mainly on abiotic factors. Indeed, agronomists usually focus on the soil-plant-atmosphere continuum. Now, ecological modellers can link food web models with soil-plant models to create innovative frameworks. Here, we advocate that food webs must be included in simulations of production and in studies of emerging properties. We emphasize the role of trophic chains in the regulation of pests. Emerging properties include aboveground and belowground interactions, pest control, and positive feedbacks on soil properties. We propose a conceptual structure for this framework and discuss how the structure of linked food web/cropping system models can account for the specific properties of agroecosystems. The proposed structure includes a process-based approach to link food webs with crop models. Such comprehensive models address the issue of trade-offs between ecosystem services, including regulation of crop pests by the ecosystem community, nutrient cycling, and crop production. (Résumé d'auteur