The effect of oil palm‐dominated landscapes on the home range and distribution of a generalist species, the Asian water monitor

The Asian water monitor lizard, Varanus salvator, is one of the largest predators in Southeast Asia which persists in human‐dominated landscapes and, as such, is a suitable model to understand the behavioral plasticity of generalists in anthropogenic landscapes. We used Local Convex Hull with adaptive algorithm to estimate the home range size of 14 GPS‐tagged individuals, followed by a MAXENT approach and community prey composition to understand the habitat preferences within the landscape. We estimated larger home ranges in forest than in oil palm plantations, as well as a larger diversity and abundance of mammals. Core home ranges were always linked to water bodies. However, the use of underproductive oil palm, freshwater swamp forest, and degraded forest by monitor lizards were higher than other kind of vegetation. This suitable habitat is proportionally larger in forest (73.7%) than in oil palm plantations (39.6%). Generalized estimation equation models showed that, while full home range size was negatively associated with the abundance of mammals, core areas depicted a positive association with mammal abundance, as well as with the proportion of suitable habitat within the home range. Besides having smaller home ranges in oil palm plantations, our findings suggest that limited suitable habitat availability forces the Asian water monitor lizard's population to establish only one or very few core areas. Contrastingly, under the protection of forest, they have more core areas, widely dispersed within larger home ranges. We conclude that regardless the plasticity of the species, human‐dominated landscapes are altering natural patterns of home range establishment in the monitor lizard's population, creating a potential ecological trap where conditions may not remain favorable for them in the long run. A deeper understanding of the ecological implications on the species and the prey community is advisable.