Modelling biodiversity change in agricultural landscape scenarios - A review and prospects for future research

International audience

Inglés. Increased intensity of agriculture and landscape homogenization are threatening biodiversity in landscapes. We reviewed 67 case studies addressing the impact of agriculture on biodiversity in model based scenario approaches and compared the information they provide on biodiversity, spatial characteristics, scenarios, and landscapes. We found an overall large diversity of approaches that we summarized statistically into six groups. "Biodiversity based agent based models", "Expert based exploration of land use change with GIS" and "Land use approaches of biodiversity with spatially explicit statistical model" are specialized biodiversity studies with high complexity in terms of biodiversity modelling with agent-based models or mechanistic models. On the other hand, "Bioeconomic modelling of policy impacts in favor of restoration of beneficial habitats", "Participatory simulation studies of landscape futures" and "Large scale multi criteria studies of innovative scenarios with optimization" do not consider species' behavior or landscape configuration, but do address a large range of socioeconomic and environmental issues. As a contribution to developing quantitative and policy-relevant biodiversity conservation studies in landscape, we present the advantages and disadvantages of each approach. We then suggest combining different approaches, particularly with the use of agent-based models and mechanistic models, integrating spatially explicit drivers of biodiversity change and the socio-economic context of farming in a participatory manner. We give recommendations on the inclusion of more taxa in future studies and collaboration between scientists from different disciplines to develop innovative solutions that can halt the biodiversity decline in agricultural landscapes.