Genetic sampling and an activity index indicate contrasting outcomes of lethal control for an invasive predator

ABSTRACT Invasive mammalian predators are implicated in the ongoing decline of a suite of fauna and continue to be a major cause of human–wildlife conflict globally. Lethal control of invasive predators is a common management strategy; however, the use of activity indices to measure management effectiveness is problematic. Non‐invasive genetic sampling may be a viable alternative approach to monitoring as individual animals can be identified, allowing for direct estimation of population density through newly developed spatially explicit capture–recapture techniques. Here we compare inferences derived from a basic activity index (number of scats per survey) and genetic sampling of scats within a before–after control–impact design to evaluate the effectiveness of a lethal control programme targeting red foxes (Vulpes vulpes) in south‐eastern Australia. The activity index was highly variable through time and suggested the baiting programme reduced fox activity on the treatment transect relative to changes on the non‐treatment transect. In contrast, genetic sampling and spatially explicit capture–recapture analysis suggested fox density varied little throughout the study, with any changes unable to be attributed to the baiting programme. Additionally, genetic sampling confirmed many individuals persisted through 7 months of baiting. These contrasting results may be partially explained by changes in scat detectability due to seasonal changes in behaviour and the disproportionate contribution of some individuals to scat counts. Our pre‐baiting density estimate of 0.28 foxes km2 (95% CI: 0.22–0.38) was lower than expected given the high productivity, abundant prey species and lack of larger predators in the study region. Our results highlight the need for cautious interpretation of activity indices and demonstrates the value of incorporating recent methodological and statistical advances when evaluating lethal control programmes.