There are a number of threats to the biodiversity of Southeast Asia. This global biodiversity hotspot has experienced surges in commercial poaching, particularly in mainland Southeast Asia, extirpating large mammals like tigers and rhinoceros from prime habitat in some locations. In maritime Southeast Asia, large-scale habitat degradation and loss has been driven by selective logging and industrial commodity plantations. With many ongoing initiatives to improve land-use management and protect biodiversity led by regional governments, it is useful to measure the impacts of these initiatives on biodiversity in a quantifiable way. The first aim of this dissertation was to examine the impacts of forest management at multiple scales on one particular species of high conservation value, while the second aim was to test how spatial modelling can improve protection for biodiversity. The location for this study was Sabah, a Malaysian state on the island of Borneo. The overall distribution of sun bears Helarctos malayanus across the forested landscapes of Sabah was previously not quantified. We used the large amount of sun bear data available from the many camera trap studies across the state and applied occupancy models to assess and predict the current and future suitability of habitats for the species (chapter 2). We found that all large forested areas of Sabah were predicted to be suitable, with almost half of the area predicted to be suitable within forests where logging and plantation development is occurring or will occur in the future. Forests of higher aboveground carbon density (i.e. more intact forests) were predicted to be more suitable than forests of lower aboveground carbon density. We conclude that future conversions of forests into plantations will have an impact on sun bears. In addition, enough resources should continue to be channeled towards anti-poaching activities to protect the large forested landscapes where bears currently live. Sabah’s sustainable forest management, in particular reduced impact logging, has been practiced for almost 20 years. In two particular sites certified by the Forest Stewardship Council, Deramakot and FMU 17A, we used sun bear data from two successive survey periods four years apart with occupancy models to investigate how sun bear habitat suitability has changed (chapter 3). Our findings revealed that overall sun bear habitat suitability increased, but not significantly, over a 4 year period throughout Deramakot and FMU 17A. There were noticeable reductions only in areas undergoing active logging (Deramakot) and one area of tree plantation (FMU 17A). This shows that sustainable forest management can provide suitable habitat for sun bears over time, as long as protection from poaching is in place. Understanding how different poaching offenders use the landscape can help managers of protected and conserved areas plan anti-poaching operations. We used a three year anti-poaching patrol dataset in an occupancy framework to identify areas of high “use” by two distinct offender types, gaharu collectors and wildlife hunters, in Deramakot and FMU 17A (chapter 4). We found that our model predictions of offender “hotspots” overlapped with areas of known poaching activity. We highlight the importance of site-level ranger knowledge in anti-poaching planning and the usefulness of applying occupancy-based predictions in policing vast forested landscapes.
