To preserve both biodiversity and agricultural production, two land management strategies have been proposed: land sparing (setting apart natural ecosystems while intensifying agriculture) and land sharing (based on agroecosystems that can support both food production and biodiversity conservation). While the scientific community has been debating which of these strategies is best, the debate about optimal land-use strategies for biodiversity conservation and food production is mired in large part because of the influence of the context in which each study is conducted. Because the study context greatly influences its empirical resolution, it provides little guidance for generalization. We addressed this challenge using a theoretical modeling approach that allows for better generalization through its ability to modify the context. We were interested in how study context factors influence biodiversity and landscape productivity, and thus, the effectiveness of different landscape management strategies. Our model was based on a controller-pest interaction to focus on ecosystem services important for food production. The context factors we assessed included landscape composition, species parameters, ecosystem services, impact of pest on crops and the planning timescale.
Results/Conclusions
Our results confirmed the importance of accounting for context factors when evaluating landscape management effectiveness. The productivity difference between intensive agriculture and agroecosystems, usually in favor of intensive practices, can be attenuated or reversed depending on context factors. For example, agroecosystem productivity increases with a high dispersal capacity of the controller or a high value of the ecosystem service it provides. Alternatively, the productivity advantage of intensive agriculture decreases with an increasing economic impact of the pest on crops. The effects induced by species parameters, ecosystem services and pest impact on crops can intensify depending on landscape composition, thereby modifying the productivity ratio between intensive agriculture and agroecosystems. We applied our model to a landscape of cocoa agroforestry, pastures, and tropical forests to illustrate the diversity of optimal solutions for landscape management depending on the context, even when context factors differ slightly. In all, our results indicate that landscape context is of utmost importance for the best strategy to support both biodiversity and food production. Our modeling approach also showed that a mixed strategy, which we call “the land-blending strategy” might provide optimal outcomes for the trade-off between agricultural production and biodiversity conservation while providing flexibility for change and uncertainty.
