The adaptive capacities of small ruminants as a lever to design agroecological farming systems

International audienceAs a sustainable food production may increasingly build on resilient, low-input farming systems, enhancing livestock reliance on natural processes is of growing interest. In this respect, small ruminants present various assets sincethey cover a broad diversity of breeds adapted to contrasting environments, including harsh and variable rangeland conditions where they evolved strong adaptive capacities. Yet, from an agroecological perspective, adaptive capacitiesremain poorly accounted for into current management and breeding strategies. Several traits reflecting aspects of resilience to undernutrition or to infections are amenable to selection and can thereby contribute to reduce the inputsneeded for production or to an integrated health management. However, the consideration of adaptive capacities in breeding schemes generally targets conventional production systems and has only little scope for their re-design.Then, a more rigorous assessment of adaptive capacities would require to examine trade-offs among resilience traits, between resilience traits and production traits, during lifetime and across environments – which is virtually outof reach using classic experimental means. Finally, few studies have started to explore how improving individual adaptive capacities or managing their diversity within the herd can affect farm resilience, even though this aspect iskey in agroecology. Given previous limitations, the use of systemic approaches integrating the mechanistic basis of adaptive capacities as well as their consequences on herd performances holds promise. Specifically, we illustrate atthe animal level how resource allocation models could contribute to an integrative assessment of adaptive capacities, including their genetic trade-offs with production traits. In turn, the use of such individual models in herd simulationshould allow identifying synergies between breeding and management strategies to promote herd resilience. Beyond, expanding this modelling at the farming system level may support the design of new agroecosystems, particularlythose enhancing crop-livestock integration