Exploring trade-offs in agro-ecological landscapes: Using a multi-objective land-use allocation model to support agroforestry research

Abstract

Finding the optimal land allocation for providing ecosystem services, conserving biodiversity and maintaining rural livelihoods is a key challenge of agricultural management and land-use planning. Agroforestry has been widely discussed as a sustainable land-use solution and as one strategy to improve the provision of multiple ecological and economic functions in agricultural landscapes. In this study, we use the backdrop of agroforestry research to evaluate a method from the multi-criteria decision analysis toolbox: robust multi-objective optimization. The key feature of this modelling approach is its capacity to integrate uncertain ecological and socio-economic data. We illustrate the optimization model with a case study from eastern Panama, showing how the model can bring together scientific and practical knowledge to provide potentially desirable landscape compositions from the perspective of farmers, a public perspective, and a compromise solution. Example results of our case study show how to assess whether agroforestry is a desirable component in a landscape composition to satisfy multiple objectives of different interest groups. Furthermore, we use the model to demonstrate how different objectives influence the optimal area share and type of agroforestry. Due to its parsimonious nature, the model could be used as a starting point of an interactive co-learning process with decision-makers, researchers and other stakeholders. The model, however, is not yet suitable for an exact prediction of future land-use dynamics, for questions of spatially explicit land-use configuration, studies going beyond the regional scale or for socio-economic interactions of agents. Therefore, we outline future research needs and recommendations for other types of models or hybrid approaches.