Tropical forests are vital global reservoirs of biodiversity and carbon (C). Deforestation and degradation of these ecosystems greatly threatens their capacity to provide crucial ecosystem functions and services, by altering complex plant-soil interactions and biogeochemical cycles underpinned by soil microbes. Forest disturbance is accelerating in Southeast Asia, through widespread selective logging (SL) and forest conversion to oil palm plantation (OP). This has major implications for soil microbial communities and functions, although effects of tropical forest disturbance on belowground biodiversity and the resistance and resilience of soil microbial nutrient and C-cycling are unresolved. The potential to restore soil microbial communities and essential functions is also largely unknown. The aim of this thesis was to evaluate impacts of tropical forest modification (degradation, conversion and restoration) on soil microbial community attributes, and implications for ecosystem biogeochemical cycling. The rainforests of Borneo were used as a model study system, representing a hotspot of biodiversity and forest degradation. I conducted survey and experiments across land-use contrasts of old-growth (OG), SL and restored forest and OP. SL and OP significantly affected relative abundances of mycorrhizal fungal types, with implications for soil C storage. Mycelial productivity was not affected by SL but was negatively impacted by OP, indicating potential for SL forest rehabilitation but consequences for OP nutrient cycling and restoration. Logging gaps in SL forest significantly altered bacterial and fungal community structure, reducing mycorrhizal abundance and altering supply rates of key nutrients. Logging gaps may also supress microbial C cycling, implied by reduced soil respiration, although soil functioning appeared more resilient to drought than adjacent closed canopy forest. Prolonged increases in soil respiration in SL forest soils after rewetting highlight sensitivity of tropical forest to future climate change perturbations. Overall, soil microbial communities did not recover with restoration of SL forest, suggesting current management practices select for different microbial taxa which may impede rehabilitation of ecosystem functions. Overall findings demonstrate human modification of tropical forest affects key soil microbial groups, with potential consequences for biogeochemical cycling and atmospheric C feedbacks at the landscape scale. This highlights the need to incorporate evaluation of belowground communities and function in predicting impacts of land-use and climate change, as well as effective ecosystem rehabilitation for biodiversity conservation and provision of essential ecosystem services by these vital, hyperdiverse environments.