Increasing food production while maintaining the same level of ecosystem services is a great challenge of this century, given the higher request of proteins from an increasing population and the need of a sustainable use of resources. Such a challenge could be addressed at different spatial levels, i.e., the spatial boundaries within which food production should be increased with no-loss of other ecosystem services can be more or less extended. In this work, we investigated how the conflict between food production and other ecosystem services can be addressed at different spatial levels. For doing this, we calibrated ecological production functions for predicting animal production, crop production, carbon sequestration, and timber growth starting from land cover and land use management variables. We then ran scenarios optimizing animal production on the French territory posing constraints of no-loss on other ecosystem services. The scenarios differed by the spatial level at which the no-loss constraints were posed: at the department level (NUTS 3, the smaller level), at the regional level (NUTS 2, medium level), and at the national level (largest level). Our findings showed, on the one hand, that constraints posed at larger spatial levels allowed higher values of animal production than constraints posed at smaller spatial levels. On the other hand, constraints posed at smaller spatial levels allowed a homogeneous distribution of ecosystem services in the territory, whereas with constraints posed at the larger level, ecosystem services were more heterogeneously distributed, leading to social inequalities. Our study implies that extending the spatial level can indeed help solving conflicts between ecosystem services, but the price to pay is social inequality. Research is needed to find the best spatial level into which defining the conflict between ecosystem services, finding the best tradeoff between food production and social equality.