L. Michelle Gilley
Gregory G. Turner
W. Mark Ford
Corinne A. Diggins
2020
<p><span>Surveying for flying squirrels by using traditional techniques produces extremely low detection rates compared with ultrasonic acoustics. Within Pennsylvania, the northern flying squirrel subspecies </span><i>Glaucomys sabrinus macrotis</i><span> is state listed as endangered due to habitat loss and parasite-mediated competition by and hybridization with the southern flying squirrel </span><i>Glaucomys volans</i><span>. This subspecies is isolated from adjacent populations in West Virginia and New York and has experienced drastic population declines. The discovery and characterization of ultrasonic vocalizations of </span><i>G. s. macrotis</i><span> and </span><i>G. volans</i><span>, as well as successful field surveys with ultrasonic acoustic detectors in the southern Appalachian Mountains, highlight the potential use of this technique for determining the presence of </span><i>G. s. macrotis</i><span>. To confirm the feasibility of using this technique on declining populations of </span><i>G. s. macrotis</i><span> sympatric with </span><i>G. volans</i><span>, we conducted 108 nights of passive ultrasonic acoustic surveys for </span><i>G. s. macrotis</i><span> at six survey sites by using two detectors per survey site (</span><i>N</i><span> = 12 detectors) in June 2017. We considered sites high quality (“high”) or low quality (“low”) based on the number of physical capture records during the past 2 decades and the dominance of boreo-montane conifer tree species in the overstory. We detected </span><i>G. s. macrotis</i><span> at four study sites and </span><i>G. volans</i><span> at all six study sites. We found higher average probability of detection for </span><i>G. s. macrotis</i><span> in high vs. low sites (0.28 ± 0.06 [mean ± SE] and 0.09 ± 0.07, respectively), whereas probability of detection was similar for </span><i>G. volans</i><span> between high and low sites (0.13 ± 0.05 and 0.17 ± 0.05, respectively). We also found </span><i>G. s. macrotis</i><span> had lower latency of detection at high vs. low sites (2.7 ± 0.8 and 7.83 ± 1.5 nights, respectively) but G. volans did not vary in latency of detection between sites (5 ± 1.6 and 3.8 ± 1.5 nights, respectively). Our study shows acoustics can be successfully used to efficiently survey </span><i>G. s. macrotis</i><span> in Pennsylvania, where populations are small and monitoring these populations more effectively is critical to determining changes in persistence due to climate- and disease-induced factors.</span></p>
application/pdf
10.3996/JFWM-20-020
en
U.S. Fish and Wildlife Service
Ultrasonic Acoustic Surveys of State Endangered Northern Flying Squirrels in the Pocono Mountains, Pennsylvania
article