Southeast Asia supports the greatest diversity of felids globally, but this felid diversity is likely to be threatened by the severe forest loss and degradation that is occurring in the region. The response of felids to disturbances appears to differ depending on their ecology. For example, the largely terrestrial and nocturnal leopard cat (Prionailurus bengalensis) thrives near forest edges and in oil palm plantations where it hunts rodents (Muridae) at night, thereby avoiding human activity peaks. Conversely, we hypothesized that the sympatric and similar-sized marbled cat (Pardofelis marmorata) would respond negatively to edges and relatively open oil palm plantations as they are more arboreal than leopard cats, rely on tree connectivity for hunting, and are diurnal so have less potential to temporally avoid humans. Using new and previously published camera-trapping studies from Southeast Asia to test habitat associations at multiple spatial scales using zero-inflated Poission GLMMS and hierarchical occupancy modelling, we examined the habitat associated and diel behaviour of the marbled cat in relation to human disturbances. We found that marbled cats were positively associated with large intact forests and, in contrast to leopard cats, negatively associated with oil palm plantations. Furthermore we found preliminary evidence suggesting marbled cats may adapt their diel behaviour to become more crepuscular in degraded forests, likely shifting their activity to avoid humans. These findings suggest that the marbled cat's IUCN Red List conservation status should be revaluated and potentially upgraded from Near Threatened to Vulnerable, matching other forest-dependent felids in the region. We posit our findings may be generalizable and that semi-arboreal and diurnal felids could face greater threats from habitat degradation than their terrestrial and nocturnal relatives.

Marbled cat occurrence data was collated through new camera trapping in Southeast Asia as well as a literature review to source previously published marbled cat occurrence records. This data was used to create a marbled cat detection history matrix. The marbled cat detection history data was analysed alongside covariate data sourced from GIS layers.