Resilience of agricultural systems: biodiversity-based systems are stable, while intensified ones are resistant and high-yielding

Abstract

CONTEXTNew types of farming systems have emerged in response to societal demands to reduce environmental impacts and dependence on non-renewable resources. These agriculture models can be classified according to their relative use of anthropic inputs or biodiversity and associated ecosystem services as agricultural production factors. All these models are exposed to climate change, price volatility and other disturbances. However, little is known about their resilience levels and the factors that drive them. OBJECTIVEWe assessed the resilience of 30 highly contrasted cropping agroecosystems in north-eastern France which covered a broad gradient from simplified and intensified to complex and biodiversity-based agroecosystems. METHODSWe applied a quantitative analysis of the dynamics of three key performances over eight years: yield, gross margin and workload. An original combined set of adapted criteria − level, variability, trend and resistance − was used to assess their dynamics, i.e. their resilience. We use partial least squares regression, multiple regression trees and Pareto ranking to identify relationships among the characteristics and resilience performance of these systems. RESULTS AND CONCLUSIONSThe results showed that biodiversity-based agroecosystems, which provide a high level of ecosystem services to farmers or an increase in natural capital, tended to have stable performances. These systems also had a lower workload than intensified systems. Conversely, input-based systems, especially irrigated and tillage-intensive systems, had higher yields and gross margins, were resistant to major disturbances but were less stable. SIGNIFICANCEWhile our results show benefits of relying on biodiversity to stabilise system performances, we highlight a potential trade-off between performance level and stability, as well as possible solutions to overcome it. We also discuss the short-term strategy of anthropic input-based systems, which perform well, but consume large amounts of human and natural resources. We show potential advantages of systems that balance both ecosystem services and anthropic inputs to provide a strategy to fill the yield gap and attain resilience.