Abstract
Noninvasive survey methods, such as camera trapping and acoustic monitoring, offer valuable tools for large‐scale wildlife surveys because of their efficacy, cost efficiency, and minimal disturbance to animals. We evaluated the use of camera traps and acoustic detectors, including low‐cost AudioMoth detectors, for surveying northern and southern flying squirrels Glaucomys sabrinus and G. volans , respectively, across the state of New York, USA. We manually classified acoustic data to species and used morphological data collected from camera trap images to predict species identity using a random forest classifier. We evaluated the efficacy of each technique and estimated the probability of detection using a multi‐method occupancy framework and latency to detection. Both methods offered a high probability of detection and low latency to detection, but we were unable to reliably differentiate between the two species from camera traps in most cases (73% of detection events). Further, while acoustic detectors and camera traps had similar detection rates during spring, the performance of acoustics when using primarily low‐cost detectors declined in summer, while camera trap performance remained stable. Integrating both methods may enhance large‐scale survey efforts, providing robust data for conservation strategies. We recommend further optimization of these techniques, such as extending camera deployment durations, shortening camera trigger delays, and improving acoustic detector sensitivity to improve accuracy of species identification and overall survey effectiveness.