Context. With the ongoing biodiversity crisis and the continued loss of species, it becomes crucial to find practical solutions to monitor threatened animal populations for wildlife conservation and management. However, in practice, monitoring is especially challenging for elusive, rare, and wide-ranging species, where estimating abundance is often expensive and time-consuming. Alternatively, estimating occupancy (i.e. detection/non-detection data) may be less resource-intensive, while still providing useful information for monitoring population trends.
Aims. We aimed to describe a new field method, the random walk grid survey, to conduct a habitat selection study on elusive diurnal forest-dwelling primates. We explored how to improve occupancy estimates when detection probability is low and determined the minimal effort needed for reasonable estimates on the species habitat selection by using site-occupancy models.
Methods. We collected data to assess the northern pigtailed macaques’ (Macaca leonina) occupancy and detection probability using a random walk survey of degraded forest fragments in Sakaerat Biosphere Reserve in Northeast Thailand. We ran simulations to identify what is required for minimum survey efforts to obtain reasonable estimates of occupancy and detection probability on small or relatively large spatial scales, covering a small primate community in Southeast Asia.
Key results. Simulations showed that the probability of detecting macaques increased dramatically with an increased survey effort. However, compared with similar line-transect survey methods, the random walk grid survey was less time-consuming. Additionally, the occupancy and habitat selection estimates were similar to our knowledge of macaque distribution within the study area.
Conclusions. Our findings suggest that the new random walk grid survey method is effective to assess the elusive northern pigtailed macaques’ occupancy, and to provide reliable data on habitat selection where there is low macaque abundance and detection probability in a degraded forest fragment.
Implications. Our survey protocol could be used as a starting point to target high location occupancy to start habituation processes, but also for further intensive studies on primate behaviour and habitat use of primate communities. Finally, combining the random walk grid survey with automated recording devices (e.g. camera traps or passive acoustic surveys) could help improve occupancy and detection probability estimates for long-term monitoring programs and over large spatial scales.
