Optimizing future cropland allocation in a biodiverse savanna by integrating agricultural benefits and ecological costs

Abstract Context Producing sufficient food to meet a growing population while minimizing the ecological impacts of agricultural expansion is a major global challenge, particularly in biodiverse regions where cropland development threatens ecological integrity. Tanzania exemplifies this tension, as efforts to meet food security goals often conflict with the conservation of its rich ecosystems. Objectives We aimed to optimize the spatial allocation of future cropland to balance agricultural productivity with key ecological objectives. Additionally, we evaluated alternative strategies for meeting future food demands, such as increasing cropland usage intensity and expanding high-yield crops, to reduce pressure for further cropland expansion. Methods We developed a spatially explicit trade-off model that linearly aggregates multiple land use objectives using flexibly assigned weights. The model integrates crop yield potential (for maize, paddy rice, sorghum, cassava, and common beans), travel time to markets, and ecological costs related to biodiversity, carbon sequestration, and landscape connectivity. We applied the model to Tanzania to identify optimal areas for cropland expansion under a range of decision-making solutions and agricultural development scenarios. Results Our analysis revealed that incorporating more decision-making factors, even with modest weights, yields greater overall benefits than emphasizing fewer objectives. Compared to a yield-only strategy, a hybrid strategy that equally weighs agricultural and ecological priorities reduced travel time to markets by 25.4%, biodiversity loss by 1.4%, carbon loss by 0.8%, and connectivity loss by 27.5%, while requiring only 2.6% more land. Additionally, increasing cropland area usage intensity and expanding the cultivation of high-yield crops can effectively boost food production, with the potential to double the current production using existing cropland alone. Conclusions Our findings highlight the potential of spatially optimized, multi-objective planning to reconcile food production with ecological conservation at the landscape scale. By explicitly quantifying trade-offs and synergies among competing objectives, this approach offers a flexible and transferable framework for sustainable landscape planning in other ecologically sensitive agricultural frontiers.