A Comparison of Survey Methods for Documenting Presence of Myotis leibii (Eastern Small-Footed Bats) at Roosting Areas in Western Virginia

Abstract

Many aspects of foraging and roosting habitat of Myotis leibii (Eastern Small-Footed Bat), an emergent rock roosting-obligate, are poorly described. Previous comparisons of effectiveness of acoustic sampling and mist-net captures have not included Eastern Small-Footed Bat. Habitat requirements of this species differ from congeners in the region, and it is unclear whether survey protocols developed for other species are applicable. Using data from three overlapping studies at two sampling sites in western Virginia’s central Appalachian Mountains, detection probabilities were examined for three survey methods (acoustic surveys with automated identification of calls, visual searches of rock crevices, and mist-netting) for use in the development of “best practices” for future surveys and monitoring. Observer effects were investigated using an expanded version of visual search data. Results suggested that acoustic surveys with automated call identification are not effective for documenting presence of Eastern Small-Footed Bats on talus slopes (basal detection rate of 0%) even when the species is known to be present. The broadband, high frequency echolocation calls emitted by Eastern Small-Footed Bat may be prone to attenuation by virtue of their high frequencies, and these factors, along with signal reflection, lower echolocation rates or possible misidentification to other bat species over talus slopes may all have contributed to poor acoustic survey success. Visual searches and mist-netting of emergent rock had basal detection probabilities of 91% and 75%, respectively. Success of visual searches varied among observers, but * Corresponding author: jhuth@VT.edu Virginia Journal of Science, Vol. 66, No. 4, 2015 http://digitalcommons.odu.edu/vjs/vol66/iss4 414 VIRGINIA JOURNAL OF SCIENCE detection probability improved with practice. Additionally, visual searches were considerably more economical than mist-netting. INTRODUCTION There has been an estimated mortality of more than 6 million bats in the genus Myotis in White-Nose Syndrome (WNS) affected areas (Blehert et al. 2009; Ford et al. 2011; Francl et al. 2011; Minnis and Lindner 2013; Puechmaille et al. 2011). This disease has continued to spread across the Northeast into the Appalachians, Midwest and mid-South (Francl et al. 2012), and now is present throughout much of the eastern United States and Canada (U.S. Fish & Wildlife Service 2016a). Undoubtedly, this increased geographic footprint has led to higher overall mortality than original estimates. Biologists have long relied on capture methods such as mist-netting near roosts or water sources and along flyways to document presence of bats (Kunz et al. 2009). Declines in bat populations due to WNS have made previous standard capture methods largely ineffective for some bat species of conservation concern in WNS-impacted areas (Coleman et al. 2014; Ford et al. 2011). As early as 1994, long before the WNS emergence, the U.S. Geological Survey (USGS) acknowledged a need to resolve questions about bat population status, recognizing that data available from state and federal agencies were insufficient to provide population estimates and assess trends, thereby recommending new sampling strategies (Loeb et al. 2015). Threats of additional population declines and regional extirpation of some bat species from WNS have heightened the need to effectively monitor long-term trends in population status, distribution, and structure of species assemblages within both WNS and presumed future WNS-impacted areas. The distribution, use of hibernacula, and foraging and roosting habits during the maternity season by Myotis leibii (Eastern Small-Footed Bat) were poorly documented prior to WNS, compared to its congeners (Krutzsch 1966; Best and Jennings 1997; Chapman 2007; Johnson et al. 2011). In Virginia, lack of targeted survey efforts and research has led to considerable variability in conclusions about the species’ conservation status; including designations as locally abundant in western Virginia (Dalton 1987), uncommon in Virginia (Webster et al. 2003), and greatest conservation need, Tier I Virginia Wildlife Action Plan (Virginia Department of Game and Inland Fisheries 2016). Moreover, reports of declines in population sizes associated with WNS vary among bat species (Hayes 2012). It has been difficult to precisely document declines for Eastern Small-Footed Bats because they often hibernate alone, in small groups, and often in obscure locations opposed to aggregative hibernators such as Myotis lucifugus (Little Brown Bats) and Myotis sodalis (Indiana Bats; Veilleux 2007:Turner et al. 2011; Francl et al. 2012). In 2013, the U.S. Fish & Wildlife Service (USFWS) was petitioned to consider listing Eastern Small-Footed Bat as threatened or endangered under the Endangered Species Act (U.S. Fish & Wildlife Service 2014). After reviewing the available scientific information, USFWS (U.S. Fish & Wildlife Service 2013) determined that listing the Eastern Small-Footed Bat was not warranted; however, numerous data gaps were noted that need to be addressed to better understand Eastern Small-Footed Bat ecology and true conservation status. Virginia Journal of Science, Vol. 66, No. 4, 2015 http://digitalcommons.odu.edu/vjs/vol66/iss4 SURVEY METHODS FOR Myotis leibii 415 For most Myotis in WNS-impacted areas, acoustic monitoring has emerged as an increasingly-used method to detect presence. Acoustic monitoring requires less effort and mitigates the higher costs, low detection probabilities, and potential false negatives from surveying with mist-nets (Coleman et al. 2014). Accordingly, USFWS now allows acoustic surveys to document presence or presumed absence of the endangered Indiana Bat (Niver et al. 2014) and is currently developing similar guidelines for the threatened Myotis septentrionalis (Northern Long-Eared Bat; Mike Armstrong, U.S Fish & Wildlife Service, personal communication). Although mist-netting allows gathering of information on sex ratios, body condition, and reproductive condition (Kunz et al. 2009), acoustic detectors are an attractive alternative sampling tool because they are relatively simple to operate and can collect large amounts of data for extended periods (Morris et al. 2011). Acoustic detectors also are capable of sampling a much larger area than nets (O’Farrell and Gannon 1999), and detection should be less sensitive to abundance, adding to the technique’s utility. Even prior to WNS, a combination of sampling methods had been proposed as the most effective monitoring strategy, as this maximized information collected and leveraged the strengths of each method (O’Farrell and Gannon 1999; Patriquin et al. 2003; Flaquer et al. 2007; Robbins et al. 2008). Although acoustic monitoring is effective for many species, a post-WNS study on bat detection probabilities in northwestern New York using opportunistic capture and acoustic methods found that Eastern Small-Footed Bats had substantially lower detection probabilities than other species in that area (Coleman et al. 2014). Because Coleman et al. (2014) focused on Indiana and Little Brown Bats’ foraging habitats, the efficacy of acoustic surveys in habitats more likely to be used by Eastern Small-Footed Bats (i.e., emergent rock formations and nearby 1 and 2 order streams) largely is unknown. To address the lack of comparisons of detection methods within Eastern SmallFooted Bat roosting areas in the central Appalachians and to aide in the development of “best practices” for future surveys and monitoring, a post-hoc comparison of detection probabilities of three survey methods was performed: acoustic surveys with automated identification of calls, visual searching for roosts on emergent rock formations, and mist-netting at sites where Eastern Small-Footed Bats were known to occur. Secondary benefits of each survey method also were considered. MATERIALS AND METHODS This post-hoc study used Eastern Small-Footed Bat detection data collected during three separate studies from sites in Virginia where Eastern Small-Footed Bats were known to occur. To maximize comparability, the original datasets were reduced to two local sites utilized by all three studies and where Eastern Small-Footed Bats previously had been detected (Moosman et al. 2015). The study sites were post-Pleistocene colluvial fields (talus slopes) in western Virginia. Sites differed in their specific geology and physical setting. Site one, Devil’s Marbleyard (hereafter DMY), is a 3.0 ha field of large Antietam quartzite boulders located in the George Washington and Jefferson National Forest in Rockbridge County (37.581332°N, 79.471420°W, datum WGS 84). The DMY is surrounded by a mixed deciduous forest predominated by Quercus prinus L. (Chestnut Oak), Quercus rubra L. (Northern Red Oak), Quercus coccinea (Scarlet oak), Pinus virginiana (Virginia Pine), and Acer rubrum L. (Red Maple) (Mengak and Castleberry, 2008). Site two is a 3.34 ha talus slope of smaller Virginia Journal of Science, Vol. 66, No. 4, 2015 http://digitalcommons.odu.edu/vjs/vol66/iss4 416 VIRGINIA JOURNAL OF SCIENCE scree composed of quartzite with some larger boulders located within the Sherando Lake’s Recreation Area (hereafter Sherando) of the George Washington and Jefferson National Forest in Augusta County (37.929370°N, -79.004356°W, datum WGS 84). Sherando is surrounded by a mixed deciduous forest similar to that surrounding DMY. As a capture baseline, mist-net data were collected during June 2009 and July 2014 (Moosman et al. 2015), and visual search and acoustic data were collected between June and August 2014. Mist-nets were deployed with 38-mm mesh in two manners. Two 12-m-long x 3-m-high nets end to end directly on the talus slope were deployed at DMY because the location lacked corridors conventionally considered suitable for surveys with mist-nets. Mist-nets were placed perpendicular to the forest edge extending toward the center of t