Crop-livestock integration and agroecological performances of mixed farming systems

Agricultural systems are expected to face new challenges: producing more and better in a changing world.Finally, farming systems are expected to meet agroecological goals: productivity by producing more; efficiencyby producing with less; resilience by adapting them; and self-sufficiency to be less dependent. Nowadays, agriculture is supported by a majority of familial small-scale mixed farming systems (MFS), which produce halfof the world food. These MFS are characterized by the diversity of their agricultural productions and especiallyby crop-livestock integration (CLI) which corresponds to interactions and complementarities between crop andlivestock systems. By developing effective CLI, MFS would be relevant to meet agroecological goals. Even ifmany studies deal with such interest, very few of them take into consideration its complexity. In fact, CLI is discussed either very broadly or too specifically. In this work, we offer to illustrate the complexity of CLI and thehypothetical agroecological performances associated with this complexity. We propose a theoreticalconceptualization of CLI situations as corresponding flows network configurations. CLI is function of threecomponents: the diversity of productions, CLI flows, i.e. flows between these productions, and flows betweenthe farming system and its environment, i.e. inflows and outflows. Similarly, four agroecological performancesare conceptualized in function of these three components relations: productivity, efficiency, self-sufficiency andresilience. Then, we analyze the performances of four CLI configurations, from specialized and nonintegratedMFS to diversified and integrated MFS. Finally, we try to understand the complex relationships between CLIand the agroecological performances. Diversification of MFS is a step toward more agroecological farmingsystems. However, fully agroecological systems will be achieved through both diversification and CLI of MFS.Finally, this conceptualization of CLI, based on flows network characterization, allows modeling various configurations of MFS for further analysis. CLI is in fact the whole set of relationships between crop and livestock subsystems but also between MFS and its environment.