An agent-based model of groundwater over-exploitation in the Upper Guadiana, Spain

Irrigated agriculture is a main user of groundwater. Achieving a sustainable use of groundwater will often require agricultural land-use changes such as shifting to entirely different kinds of crops and/or technologies. Enhanced understanding of land-use change is hence required for developing policies for a sustainable water future. We use an agent-based model to investigate the history of irrigated agriculture in the Upper Guadiana Basin, Spain, in order to learn about the influence of farmers' characteristics on land-use change and associated groundwater over-use. A shift from vineyards and cereals to horticultural crops would provide a possibility for higher income with less water use. Such a shift cannot be observed historically. The model results suggest that risk aversion and path dependency are insufficient to explain this observation, and the organisational set-up of farms limiting the maximum labour force needs to be considered as additional explanatory factor. Furthermore, it is shown that different types of farms existing in the UGB can be expected to exhibit distinct responses to drivers of land-use change such as agricultural policies. It is concluded that a sound understanding of the social system making use of a resource is required to solve problems of resource over-use. This article demonstrates that agent-based models can be useful tools to enhance such an understanding even in situations of scarce and uncertain data that are often encountered when dealing with resource-use problems.