Evaluation of camera placement for detection of free‐ranging carnivores; implications for assessing population changes

1. Introduced carnivores are often cryptic, making it difficult to quantify their presence in ecosystems, and assess how this varies in relation to management interventions. Survey design should thus seek to improve detectability and maximize statistical power to ensure sound inference regarding carnivore population trends. Roads may facilitate carnivore movements, possibly leading to high detectability. Therefore, targeting roads may improve inferences about carnivore populations. 2. We assessed our ability to monitor feral cats Felis catus and red foxes Vulpes vulpes on‐ and off‐road, with explicit consideration of the location of monitoring sites on our ability to detect population changes. We also assessed whether there was evidence of spatial or temporal interaction between these species that might influence their road‐use. 3. Surveys were conducted in a conservation reserve in south‐eastern Australia between 2016 and 2018. At each of 30 sites, we deployed two motion‐sensor cameras, one on‐road, and the other off‐road. Using occupancy models, we estimated cat and fox occupancy and detectability, and conducted a power analysis to assess our ability to detect declines in occupancy under three monitoring regimes (efforts targeted equally on‐ and off‐road, efforts targeted entirely off‐road and efforts targeted entirely on‐road). 4. On average, on‐road detectability was seven times higher for cats and three times higher for foxes. Targeting survey effort on‐road yielded the greatest power for detecting declines in both species, but our ability to detect smaller declines decreased with decreasing initial occupancy probability. No level of decline was detectable for cats when survey efforts were targeted off‐road, while only large declines (>50%) were detectable for foxes (assuming high initial occupancy probabilities). We found little evidence of spatial or temporal segregation, suggesting limited avoidance or suppression between the two species within this landscape. 5. Our results suggest that targeting monitoring on roads may be an effective approach for detecting declines in introduced carnivore populations, particularly following management intervention (e.g. lethal control), and in the face of resource limitations. We provide a framework that can help assist land managers to make informed decisions, which balance monitoring efforts and resource constraints with sufficient statistical power to assess management objectives.