Exploring red fox (Vulpes vulpes) occupancy, detectability and statistical power to inform monitoring decisions at Booderee National Park

Context The difficulties of invasive predator management can be exacerbated by suboptimal monitoring data. Developing a cost-effective, operationally achievable and statistically robust monitoring program is therefore a key component of biodiversity conservation and management. Aims To explore options to better balance the survey effort, cost and statistical power of the red fox (Vulpes vulpes) monitoring program at Booderee National Park (BNP), south-eastern Australia. Methods We analysed existing camera-trap data over an 8-year period to estimate fox occupancy and detectability. We then conducted a power analysis to assess our ability to detect reductions in fox occupancy of varying magnitudes under six hypothetical monitoring scenarios with varying survey effort. We compared how statistical power varied among scenarios between seasons of high and low fox detectability (i.e. summer and spring respectively), and across three levels of initial occupancy (Ψ = 0.75, 0.50, 0.25). Annual costs were estimated for each hypothetical scenario as well as the current fox monitoring approach at BNP. Key results Fox occupancy at BNP increased between 2017 and 2022, while recent monitoring demonstrated seasonal variation in fox detectability. Additional monitoring sites could help maintain statistical power to detect a change in the fox population at BNP when fox occupancy is low. For a given number of sites, statistical power was similarly maintained by doubling the deployment length or conducting monitoring when detectability is highest without increasing deployment length, with the latter option avoiding increased data-processing costs. Conclusions To balance monitoring and operational needs at BNP we suggest a continuous camera-trap monitoring program, with permanently established posts. This approach could standardise camera orientation, reduce potential biases in the data and provide up-to-date operational information required by park managers. Implications Existing data can bolster our understanding of factors influencing detectability, and provide insight regarding monitoring options to better balance survey effort, cost and statistical power.