Exploring path-dependencies and spatial variability in landscape scale scenarios for ecosystem services assessments

Ecosystem services (ES) of landscapes evolve with landscapes themselves and with changes in their characteristics or uses due to human or natural causes. Land use changes may have immediate impacts upon the delivery of services in the short term (e.g. for food production), or longer term (e.g. sequestration of carbon by woodland due to a time lag of tree growth). The spatial location of land use changes also impacts on ES; e.g. woodlands might serve different roles for flood regulation depending on their location, through stabilising slopes against erosion or slowing water flow in flood plains. The integration of high temporal and spatial resolution data with scenarios of land use change provides a dynamic platform with which to assess ES [Rounsevell et al. 2006]. Studies often only consider snapshots in time and spatially limited scenarios of land use change [Millennium Ecosystem Assessment 2006, Nelson et al. 2009, UK National Ecosystem Assessment 2011]. This paper proposes a simulation framework and tools for exploring spatiotemporal dependencies of land use change and ecosystems services. The LandSFACTS toolkit [Castellazzi et al. 2010a, Castellazzi et al. 2010b] allows exploration of stochastic realisations of land use changes within socio-economically defined scenarios. Land use changes are then dynamically assessed for impacts on ES such as carbon sequestration, and biodiversity. The approach can be used to explore impacts of path-dependencies and spatial variability of land uses change upon ES. To exemplify our approach, a case study of woodland expansion in Scotland is used. Key findings show the sensitivity of certain spatial outcomes on key ES, such as the support of biodiversity through habitat networks. Conclusions are drawn with respect to the potential of such tools for informing strategic and regional planning, taking account of potential impacts between different ES.