Humans are damaging the natural world at an alarming rate. Urbanisation, climate change and deforestation are affecting ecosystems on which our society relies for important services, such as crop pollination by insects and pest control by natural predators. Decisions such as where to build a highway or dam, or which residential areas to expand, must take into account the risk to ecosystems to adequately calculate the potential cost of these decisions. Mathematical models inform decision-making: they formalise assumptions, factor in data and quantify the outcomes of alternative scenarios. Decision-makers and scientists need better, actionable models to inform policies involving ecosystem services. One hurdle is that two potentially useful types of model are being developed in isolation from one another. Biogeographers are developing models to predict how changes in the environment (e.g., temperature increases) affect the geographical distribution of individual species. Meanwhile, community ecologists are developing models to predict how species behaviour and interaction patterns change when habitats are modified, and how ecosystem services are thus affected. These two types of model tell us how environmental change affects where species are, and how species interactions affect ecosystem services. If combined, they would relate environmental changes to the provision of ecosystem services. In this synthesis project, I am proposing a new framework that links these two types of model using cutting-edge techniques in mathematics and computation. I will be able to directly trace and investigate the feedbacks between human decision-making (e.g., converting natural vegetation to agriculture) and the provision of ecosystem services arising from species interactions (e.g., insect pollination and biological pest control), and, using field data, identify the relative importance of different socioeconomic and environmental drivers (e.g., household income and deforestation) and types of species (e.g., specialist or generalist, rare or common). The framework will inform policies that safeguard the benefits of the natural world for future generations.