Designing a robust survey and monitoring program: A simulation approach to evaluate the power of various sampling designs to detect changes in possum population trends

Monitoring of fauna is an essential component of adaptive conservation management, and it is critical that managers can effectively detect changes in species trends between two time points. Sampling arrays should be designed to incorporate the species' ecology and the assumptions of the target analysis framework. Here we evaluate the strength of various spotlighting sampling designs to detect changes in ngwayir (western ringtail possum; Pseudocheirus occidentalis ) population trends between surveys, using the Upper Warren Region of Western Australia as a case study. We simulated various population scenarios, conducted virtual spotlighting surveys, and then built distance sampling models. This allowed us to compare how close the resulting estimates were to the “true” simulated population using a suite of parameters quantifying accuracy, precision, and power. At the local scale, 240–360 km of effort was suitable to monitor populations in medium‐ and high‐density habitat. Low‐density areas could not be effectively sampled using this framework. At the regional scale, the unique population distribution in the Upper Warren introduces considerable uncertainty and detecting subtle population changes is challenging. We recommend an adaptive approach to regional monitoring, whereby effort is primarily concentrated in the current hotspot. These findings can directly inform the design of programs to monitor ngwayir in the Upper Warren and beyond. The simulation framework utilized here could also be applied to evaluate designs for other species or regions of interest, ultimately ensuring sufficient power to accurately and precisely monitor fauna population trends.