Tropical forests support over two-thirds of the world's terrestrial biodiversity. However, between 35% and 50% of tropical forests have already been degraded, and the rate of deforestation continues to increase. Secondary forests, plantations and other human-modified habitats now dominate tropical landscapes, leading to concerns that human degradation of these landscapes will elevate greenhouse gas emissions and jeopardise ecosystem services at local, regional and global scales. The area of protected forests is unlikely to increase greatly in the future, so the persistence of tropical biodiversity and the important biogeochemical cycles and ecosystem services associated with it will depend to a large extent on the way we treat the wider tropical landscape. The Human Modified Tropical Forests programme seeks to 'significantly improve our understanding of the links between biodiversity and biogeochemical cycles in tropical forests' through 'integrated observations and modelling linked to gradients in forest modification'. To contribute towards this goal our consortium will use surveys along a modification gradient within the SAFE landscape in Sabah (Malaysian Borneo) to detect patterns, combined with manipulative field experiments to gain a mechanistic understanding of biodiversity-function linkages. We will assess links between above- and belowground components of tropical biodiversity and investigate the extent to which different elements of biodiversity (e.g. species of conservation concern) are associated with measures of ecosystem function (decomposition processes and biogeochemical cycles). We will then upscale from the experimental sites to the landscape-scale to generate spatial layers of ecosystem function, biodiversity, and greenhouse gas fluxes to inform policy scenario modeling. Our work will thus (1) characterise soil microbial function and measuring associated biogeochemical fluxes; (2) Experimentally test the links between aboveground biodiversity and soil function; (3) Build and add to existing datasets for bird and mammals, and explore correlations between ecosystem functioning and the distribution of species of conservation concern; and (4) Explore policy scenarios for optimising biodiversity and function protection.