Improving the decision making in an agent-based modelling framework for land-use dynamics through ecosystem services changes monitoring

Through the coupling of models with scenarios for land use change and regional climate change, possible feedback mechanisms between both could be discovered. In order to model plausible future land-use change, a variety of methodologies exist. Amongst these, spatially explicit agent-based models (ABM) are proven to be useful for studying land-use dynamics (e.g.Bert et al. 2011). The APORIA framework developed by Murray-Rust et al. (2011) is a conceptual agent-based model for analysing socio-ecological systems in which farmers’ behaviours and practices are a crucial part. Farmers’ decision making influences the land use, and thus the quality and the quantity of Ecosystem Services (ES) a landscape can provide to people. Monitoring ES within an APORIA simulated landscape helps understanding the relationships between land management practices and levels of natural capital. Developing on these findings, we explore how spatial data concerning ES can be used in the decision making process of the agents by using the sensitiveness of agents to different ES. This sensitivity should be equally based on the economic value of the ES, the biophysical value of the ES and of the value given by people to the ES (the social value)(Fontaine et al. 2013). For the biophysical value, the ABM will be linked to dynamic vegetation models which are coupled to regional climate change models. The case study being the whole country of Belgium, three major agent classes will be used: farmers, forest managers and urbanizers. Farmers are sensitive about how landscape dynamics affect food production, which involves a variety of ES linked to a broad range of spatial characteristics such as soil erosion, soil compactness or soil fertility. Urbanizers are expected to be more sensitive to changes in land-use affecting cultural ES such as recreation and aesthetics or flood protection. Forest managers will be more interested in supporting ES that favour wood production. The sensitivity of these “human functional types” (Rounsevell et al. 2014), together with other preferences and characteristics, allows for the differentiation of a broad spectrum of agents with different choices for future land use. These choices and the outcome of these choices, can then again influence the agents in a new decision making phase. Though it’s still work in progress, we expect that resulting future land use change scenarios from the ABM are more plausible, allowing to better inform e.g. management policies.