The State of Bats in Conservation Planning for the National Wildlife Refuge System, With Recommendations

The United States is home to just 47 of the more than 1,100 species of bats in the world; these bat species encompass 11% of the nation's mammalian fauna (IUCN 2011). However, the relatively limited diversity of temperate bat species in the United States belies their ecological and economic importance. Besides their intrinsic value, insectivorous North American bats are often unappreciated predators of agricultural pests; one recent model estimates the value of services provided by bats to agriculture in the contiguous 48 states is roughly US$22.9 billion/y (Boyles et al. 2011). Several North American bat species now face an unprecedented array of threats, including emerging diseases and anthropogenic stressors.

Presently, 12 bat species and subspecies are protected or are candidates for protection as endangered or threatened species under the U.S. Endangered Species Act (ESA 1973, as amended); candidate species are those species whose status has been reviewed by the U.S. Fish and Wildlife Service (USFWS) and they have been found to warrant protection under the ESA, but their addition to the endangered species list is precluded by higher listing priorities. In addition to the already protected and candidate species, the USFWS recently determined that ESA protection may be warranted for the northern long-eared bat Myotis septentrionalis and the eastern small-footed bat Myotis leibii (USFWS 2011b). The USFWS is currently completing a status review for these two species to determine whether they warrant ESA protection, as well as a status review for a third cave-obligate bat species, the little brown bat Myotis lucifugus.

Although many of the threats faced by the numerous imperiled bat species occur on private lands as a result of factors over which land management agencies have little or no control, the U.S. Federal Government has the ability to minimize other threats and promote the long-term viability of these species. This end can be achieved through careful and proactive management of the >600 million acres of public lands administered by the U.S. Forest Service, the National Park Service, the Bureau of Land Management, the Department of Defense, and the USFWS. The primary land management program of the USFWS, the National Wildlife Refuge System (NWRS), is unique among federal land management agencies because of its central mandate to manage for animal and plant resources. In the NWRS Improvement Act (Pub. L. No. 105-57), the NWRS was charged with a mission of administering “a national network of lands and waters for the conservation, management, and where appropriate, restoration of … wildlife … and their habitats within the United States ….” This mission underlies all Refuges, although many have additional purposes defined in the laws or executive orders that authorized the creation of individual Refuges. This review examines the extent to which the NWRS is meeting the challenge of conserving bats and their habitats, and it offers recommendations regarding future management of these bat species.

Even before undertaking a systematic review, it was apparent that there was a significant range of intensity focused on bat conservation in the NWRS. Some Refuges specifically and actively manage for bat habitat, whereas others have bat habitat as a consequence of management for other species. In the first round of CCPs, some took the perspective that “although there are little specific data regarding population sizes, the interspersion of woodlands, wetlands, abandoned fields, agricultural lands, creeks, ponds and river make it reasonable to assume that these species [of bats and other nongame mammals] are faring well” (USFWS 2008). Although this statement was perhaps justifiable at the time it was written, the growing and novel combinations of threats facing bats mean that its stated assumption may no longer be valid in many locations. Many of these challenges are not new; indeed, threats such as habitat loss are not unique to bats. However, aspects of bat biology, such as differing habitat requirements throughout the year for hibernation or migration, make them particularly vulnerable. For example, conservation biologists have long recognized that habitat loss is a significant risk to populations across taxa (Fahrig 1997) and that habitat loss or degradation was an important factor in the endangered listing of the Indiana bat Myotis sodalis in 1967 (USFWS 2007a). But for species such as the Indiana bat, preventing habitat loss is not as simple as protecting a set number of forested acres. The recovery plan for that species identifies changes in hibernaculum temperature due to commercial use of caves; flooding of hibernacula due to dam construction; collapse or sealing of mines; and loss of woodlands used for maternity colonies, summer foraging habitat, and migration habitat all as habitat-related factors that led to the species' decline (USFWS 2007a).

As predators, insectivorous bats are exposed to toxins and environmental contaminants that have accumulated in their prey. Heavy metals such as lead and mercury have been shown to bioaccumulate in bat tissues (Hickey et al. 2001; Walker et al. 2007) as do insecticides such as dieldrin and other organochlorine pesticides (O'Shea et al. 2001). Bats that forage at sites with heavy mercury contamination accumulate 10–40 times more mercury in their tissues (Nam et al. 2012), although this exposure does not appear to trigger increases in hormones that are indicative of environmental stress in other taxa (Wada et al. 2010). The effects of chronic sublethal exposure to these environmental contaminants are poorly understood, but organochlorine pesticide use has been implicated in dramatic local population declines by the Mexican free-tailed bat Tadarida brasiliensis (Clark 2001) and has been suggested as a contributing factor to susceptibility to pathogens in the eastern United States (Kannan et al. 2010).

In spite of their reputation for being a clean, renewable source of energy, wind farms have been shown to kill substantial numbers of bats (Arnett et al. 2008; Smallwood 2013). This mortality does not appear to be the result of random collisions, because species such as big brown bat Eptesicus fuscus have been found to be the most frequently captured in the vicinity of wind farms, but carcasses of that species are rarely recovered, whereas seldom-observed species such as the hoary bat Lasiurus cinereus suffer the highest mortality (Johnson et al. 2004). Wind energy facilities appear to cause disproportionately high mortality in migratory bats (Arnett et al. 2008); one study in southern Minnesota found that 89% of fatalities were migratory tree bats of the genus Lasiurus (Johnson et al. 2003). High mortality in nonmigratory populations of Mexican free-tailed bat also has been found when turbines were located close to maternity colonies (Piorkowski and O'Connell 2010). Similar recent findings have been reported in nonmigratory big brown and little brown bats (Grodsky et al. 2012). The reasons for these nonrandom collisions are not entirely clear, although research suggests that bats are drawn to prominent landmarks on the landscape (Cryan and Brown 2007), possibly due to insect aggregations around turbines (Rydell et al. 2010) or as part of their mating behavior (Cryan 2008).

Perhaps the most significant threat facing North American hibernating bats is white nose syndrome (WNS), a recently discovered and aptly named emerging infectious disease. This disease is responsible for the dramatic decline of several hibernating bat species in the northeastern United States (Blehert et al. 2009), and it may pose a considerable threat to hibernating bats throughout North America. Since first being documented in New York state in 2006, the pathogen associated with WNS has been detected in 21 states as of August, 2012. Recently, the USFWS and other partners estimated that at least 5.7 million to 6.7 million bats have died as a result of WNS (USFWS 2012). Another recent report estimated an 88% decline in the number of hibernating bats at 42 sites from New York, Pennsylvania, Vermont, Virginia, and West Virginia (Turner et al. 2011). Furthermore, WNS has been forecast to cause the regional extirpation of the little brown bat, formerly one of the most common bats in North America, in the eastern portion of its range (Frick et al. 2010).

The impact of WNS varies by species; however, seven North American hibernating bat species (little brown bat, Indiana bat, northern long-eared bat, eastern small-footed bat, big brown bat, tricolored bat Perimyotis subflavus, and gray bat Myotis grisescens) are known to be affected by WNS. The fungus that causes WNS has been detected on two additional species: southeastern bat Myotis austroriparius and cave bat Myotis velifer. The disease is caused by the newly described psychrophilic (i.e., cold-loving) fungus, Pseudoascocarpus destructans (P.d.), that grows on exposed tissues of hibernating bats (Gargas et al. 2009; Lorch et al. 2011; Warnecke et al. 2012).

It appears that the disease impacts the bats by disrupting their physiology, possibly by causing the bats to arouse more frequently from normal torpor cycles as they become dehydrated (Cryan et al. 2010; Warnecke et al. 2012). There is also evidence of an overly aggressive and damaging immune response exhibited by infected bats after they awake from torpor (Meteyer et al. 2012). The pathogen, P.d., has been detected in Europe, where bats do not appear to suffer mass mortality (Puechmaille et al. 2010, 2011a; Šimonovičov et al. 2011). This difference has led to speculation that European bats have coevolved with the fungus and that it is an invasive species in North America, although pathogenicity and relationship of European P.d. to North American P.d. is still an ongoing area of research (Wibbelt et al. 2010; Puechmaille et al. 2011b; Warnecke et al. 2012). The high mortality at infected North American hibernacula, or winter roosts used specifically for hibernation, is manifested in dramatic declines of summer activity (Dzal et al. 2011; Ford et al. 2011) and potential effects on the timing and success of reproduction in some species (Francl et al. 2011), implying that it is substantially impacting their populations.

The NWRS Improvement Act (Pub. L. No. 105-57) required, among other things, that the USFWS develop a CCP for each Refuge. It is in the CCP that goals, objectives, and strategies for wildlife and visitor services are articulated and vetted, and it serves as the guiding management document for a Refuge. These CCPs are meant to be revised no less frequently than every 15 y. Because the earliest CCPs are due for revision, and novel and compounding threats are placing formerly widespread bats in peril, now is the time for the USFWS to ensure that the NWRS is meeting its obligations to bats under the Refuge Act and making a concerted effort to prevent the endangerment of additional bat species.

To that end, we conducted a review of how bats are addressed in CCPs throughout the NWRS. We assessed the quality and consistency of current CCPs with regard to bat management and identified particular geographic areas most in need of greater attention to bats in Refuge planning. Based on these analyses, we offer general recommendations for improvements to CCPs and to bat management in the NWRS.

Nearly all electronically available CCPs from each USFWS region, except Alaska, were reviewed. In total, we reviewed 226 final CCPs, 11 draft CCPs, seven Comprehensive Management Plans (CMPs) that were completed just before the 1997 mandate for CCP planning and that serve as stand-in CCPs for those Refuges, and three Conceptual Management Plans or Interim CCPs that guide the management of those new Refuges until a CCP can be completed. Hereafter, these different categories of plans are referred to generically as CCPs. A list of the CCPs reviewed is included Table S1 (Supplemental Material). To facilitate time-effective review, keyword searches of these CCPs were conducted for the terms bat, Chiroptera, Myotis, and Lasiurus. In all instances where a CCP included discussion of Myotis or Lasiurus, there was also mention of “bat,” so no further searches were conducted by other generic names. Ten CCPs that did not mention bats and were unlikely to provide habitat for bats (e.g., Shell Keys National Wildlife Refuge [NWR] that consists solely of shell-derived islets with little vegetation) were excluded from further analysis. We qualitatively assessed the degree to which bat conservation and management was addressed in the CCP and assigned a score according to the criteria in Table 1.

As a means of exploring correlates of planning emphasis on bats, the general linear model procedure in the statistical package R (version 2.1.5.0) was used; the default settings for the general linear model were used in the analysis. The USFWS administrative region; year the CCP was approved; and the presence or absence of federally endangered, threatened, or candidate bat species were analyzed as covariates with CCP score. Statistical significance was assessed with an α of 0.05.

The need for strong bat emphasis in a CCP is not the same for all Refuges, and the need depends on geographically distributed characteristics that can be summarized using spatial analyses. To determine the relative importance of effective conservation planning for bats at each Refuge, we used a simple GIS analysis to generate an index of need for each Refuge. The index is based on bat species richness; habitat type, as generalized by ecoregion; and geological characteristics, specifically the existence of karst formations, or geologic formations shaped by the dissolution of bedrocks such as limestone or dolomite and often containing caves or sinkholes. By comparing CCP scores to this index, we were able to assess the overall adequacy of each CCP for its Refuge and to identify the units most urgently in need of improvement. Units with low CCP scores and high need indices were identified as having the most “improvement potential.” In addition, we identified the Refuges most at risk from two specific threats: the spread of WNS and wind power development.

Using data sets with full coverage of the lower 48 states, we identified the following characteristics found within each Refuge: bat species occurrence according to range maps (England 2003), Environmental Protection Agency (EPA) Level III Ecoregions (EPA 2011), surface karst geology classifications (Davies et al. 1984), wind power potential according to a low-resolution wind map (NREL 2011), and the most current county-level WNS status at time of writing (Butchkoski 2012). Each Refuge was treated as the sum of its land units, such that characteristics encountered on any parcel were included in the description of the entire Refuge. Because these data are not extant for the Refuges of Guam and Hawaii, those Refuges were excluded from this analysis.

Bat species occurrence was used as the primary indicator of bat diversity and of the need for consideration of bats in CCPs. Range maps for 45 North American bat species (England 2003) were used to generate a list of species potentially occurring on each Refuge. Refuges made up of many dispersed parcels may have higher than expected bat diversity and may include unanticipated species if the Refuge includes satellite locations that are distant from the unit's core.

Ecoregion classifications provided a coarse description of both the ecological and geographic setting of each Refuge. The use of Level III Ecoregions allowed us to generalize about land characteristics such as topography, vegetation, land use, water cover, and wind power potential in a broad but simple and systematic way. These generalizations may be informative in determining the importance of particular areas to bat conservation, or the specific suite of threats likely to be encountered at a particular Refuge.

The karst classification map (Davies et al. 1984) is a full-coverage data set that we used to identify the likely location of caves or other geologic formations providing potential roosts for cave-dwelling bats. Hibernacula and other roost structures are critical to the 25 North American bat species that hibernate, and they may support winter or year-round occupancy for many species. Although karst topography does not necessarily indicate suitable habitat or present use by bats, Davies' karst assay (Davies et al. 1984) used a consistent method to identify areas likely to include caves or other potential roost structures required by cave-dwelling bats.

White nose syndrome arose as the most imminent threat facing susceptible cave-obligate bats within the zone of WNS infection after many initial CCPs were written. We use county-level WNS status classifications for each Refuge (Butchkoski 2012). This disease is likely to be a leading conservation concern for any Refuge in a county where the impacts of WNS have been documented.

To determine which CCPs are most in need of additional attention to bats, we calculated their improvement potential (IP), based on the results of the GIS analysis and their CCP scores. First, data regarding species richness (i.e., number of species present), threatened and endangered species presence, and karst topography were used to make a broad estimate of each Refuge's overall value to bats. Then, CCP scores were subtracted from this estimate to produce an IP metric. Refuges with both high value to bats and inadequate CCPs (low CCP scores) have the highest IP. Specifically, we used the following formula:

where IP is a Refuge's improvement potential metric; S is number of species occurring on the Refuge; E is 1 if a threatened or endangered bat species occurs on the Refuge, and 0 if not; K is 1 if karst formations exist on the Refuge, and 0 if not; and C is CCP score for the Refuge.

Weights were scaled to produce a range of IP metrics centered around 0, with Refuges most in need of improvement having high positive values and Refuges with the little room for improvement having negative values. The relative weighting of variables we used (e.g., 0.25 for each bat species vs. 2 for threatened or endangered bat species) was determined subjectively, with the goal of emphasizing the importance of endangered species presence and wintering habitat, while also allowing species richness to contribute substantially to IP. The resulting values are not intended to summarize the overall adequacy of each CCP (e.g., a “−6” does not indicate perfection); rather they are intended to rank Refuges and identify those Refuges with the greatest need for improvement. We did examine the use of more complex metrics, including such variables as genus richness and specific karst classifications as well as various weighting schemes with different valuations of richness, endangered species, and karst, but we found that the resulting IP rankings were effectively the same as the simple points–based formula above. For full base data used to calculate IP metric, as well as all IP scores, see Table S2 (Supplemental Material).

White nose syndrome and wind power are recently arisen threats whose future impacts are largely unknown. As such, these factors were not included in the calculation of the IP metric, but they were used to focus analysis on high-risk groups of Refuges. Separate rankings were compiled using the IP metric among Refuges with suspected or confirmed WNS, as well as those with high wind power potential. The WNS status of each Refuge was identified by converting a county-level WNS emergence map (Butchkoski 2012) into a GIS layer and overlaying Refuge boundaries. Please note that as WNS continues to spread to new areas, the data presented here will soon be outdated. Wind power potential was estimated as high, medium, or low per ecoregion, based on speed category coverage within the ecoregion, using a National Renewable Energy Laboratory 80-m onshore average wind speed map (NREL 2011). Our wind power categorizations do not necessarily identify areas most suitable for wind energy development, because wind farm siting includes a suite of additional factors not considered here, but they do identify Refuges located in regions with the potential for development based on wind resource.

There was inconsistent attention to bats in the first round of comprehensive conservation planning (Table 2); almost 14% of Refuge CCPs failed to mention bats at all. In 39% of CCPs, the sole mention of bats was on the Refuge species list; and of those, barely a quarter indicated that surveys had confirmed the use of the Refuge by those species. The objectives for Refuge management included strategies directly or indirectly related to bats in 19% of CCPs. For raw CCP scores, please refer to Table S1 (Supplemental Material). The USFWS administrative region was not a significant predictor of CCP score (t  =  −1.117; P  =  0.2653). There was a positive, significant relationship between the year of CCP approval and the CCP score for bat emphasis (t  =  3.678; P < 0.0001). This relationship is not likely to be related to the designation of additional species of conservation concern over time, because all of the federally protected bat species were listed before CCP began in the late 1990s. There have been two bat species classified as candidates for protection under the ESA since that time: subspecies of Pacific sheath-tailed bats Emballonura semicaudata semicaudata and Emballonura semicaudata rotensis, both of which are not known to occur on any Refuges and therefore are not included in any CCPs. In addition, the Florida bonneted bat Eumops floridanus has recently been proposed for listing as endangered; it was mentioned in one CCP while it was still classified as a candidate species (J. N. Ding Darling NWR; USFWS 2010a).

Beyond a general trend of insufficient knowledge of, and planning for, bats in the NWRS, our review found higher but inconsistent focus even on ESA-protected bat species. The documented or likely presence of federally protected bat species is correlated with higher CCP scores (t  =  6.362; P < 0.0001), with a mean score of 4.61 for Refuges with ESA-protected bat species compared with 2.19 for Refuges without ESA-protected bat species. In the NWRS Improvement Act (Pub. L. No. 105-57), Congress stated that the NWRS “serves a pivotal role in the conservation of … endangered and threatened species, and the habitats on which they depend.” As a federal agency, the USFWS is required to comply with Section 7 of the ESA, and its actions would be required to minimize or mitigate effects on endangered species. Thus, it is logical that CCPs would, at a minimum, include measures that would minimize adverse effects to endangered bat species. This pattern is not universal. For example, the James Campbell NWR CCP states that the endangered Ōpe'ape'a or Hawaiian hoary bat Lasiurus cinereus semotus “may use the refuge for foraging areas … They have been documented near the refuge.” (USFWS 2011c). However, there were no explicit mentions of surveys to confirm the presence or absence of the species, no mention of management of habitat for that species, and the included intraservice Section 7 consultations do not even include the species on the list of those considered for the actions under consultation. In contrast, Guam NWR, in the same USFWS administrative region, has identified the endangered Mariana fruit bat Pteropus mariannus mariannus as a focal species. That Refuge is researching roost and forage locations of the species, and it has included specific objectives regarding protection and restoration of the species' preferred limestone forest and support for reintroduction of captive-bred bats (USFWS 2009b). Guam and James Campbell NWRs are both part of the Hawaiian and Pacific Island NWR Complex, both are known or potential habitat for endangered bat species, and their CCPs were only produced 2 y apart; and yet, they prioritize bat conservation completely differently. This disparity could stem from the difference in known habitat use, but at a minimum it would be prudent for Refuges within the range of an endangered species to acknowledge and analyze their potential contribution to that species, and if that is unknown, include research into the unknown role of its lands for it.

We also noted varying focus on bat species with state conservation status or on bat species identified as species of greatest conservation need in state Comprehensive Wildlife Conservation Strategies, that states use to identify conservation goals to qualify for the federal State Wildlife Grant Program. In Section 5.a.4.E and M, the NWRS Improvement Act (Pub. L. No. 105-57) requires that the USFWS shall “ensure effective coordination, interaction, and cooperation with … fish and wildlife agencies of the States in which the units of the System are located” and “ensure timely and effective cooperation and collaboration with … State fish and wildlife agencies during the course of acquiring and managing Refuges.” These mandates require, at a minimum, that the service participate in the “place-based collaboration” concept of cooperative federalism (Fischman 2005), wherein Refuge planning efforts should incorporate the conservation objectives of the states so long as they do not conflict with the purpose of the Refuge or federal law. An exceptional example of this is the CCP for Bogue Chitto NWR in Louisiana, a state that has no federally listed bat species but may have habitat suitable for the state concern species Rafinesque's big-eared bat Corynorhinus rafinesquii, Seminole bat Lasiurus seminolus, and northern yellow bat Lasiurus intermedius. For its nongame species objective, that Refuge included several specific strategies related to these and other bat species, including occupancy assessment, restrictions on infrastructure alteration that could adversely affect bats, and specific pine and bottomland forest management strategies to provide roost trees for bats (USFWS 2011a). This is an appropriate approach to designing objectives for Refuges that contain species with a state conservation status.

Although the results of our spatial analysis indicated a need for consideration of bats in conservation planning for many Refuges, this need also varied greatly among Refuges and geographical regions (Figure 1). Species counts were distributed unevenly, and diversity hot spots were a prominent feature in several areas, but especially in western and southwestern USFWS Regions 1, 2, and 8. Four Refuges (Leslie Canyon, San Bernardino, San Diego Bay, and Seal Beach) overlapped, with more than 20 species ranges, and an additional eight ranges overlapped with at least 15 species (for complete results, see Table S2, Supplemental Material). Most of the high-diversity Refuges were found to occur in ecoregions in southern and central California and southwestern border regions. The spatial analysis also identified variation in the amount of karst topography among regions and Refuges. Karst features were found on at least part of 56 Refuges, and although these features were more common in certain geographic areas, they were not found to be heavily prevalent among any regional group of Refuges. Although the analysis revealed major differences among CCPs in their level of attention to bats, no clear spatial or regional patterns regarding CCP scores emerged from the spatial analysis. However, when calculating improvement potential, high diversity in the west and southwest caused IP metric values to be particularly high for Refuges in those areas.

The IP metric was useful in allowing us to rank Refuge CCPs according to their need for additional attention to bats relative to one another; however, it does not provide an absolute measure of adequacy. The mean IP score was 0, and positive IP values indicated CCPs with the most room for improvement. Twenty-five Refuges had an IP of ≥2 (Table 3). The top-ranking Refuges were largely determined by species diversity, with San Bernardino and Leslie Canyon topping the list, despite its moderate CCP score. Refuges in California and the Southwest are well represented among the high-IP CCPs owing mainly to the level of species diversity in those areas; however, Refuges from all USFWS administrative regions are among those with substantial room for improvement.

Counties with suspected or confirmed infections of P.d. are currently limited to the Northeast, Midwest, and Southeast. Several Refuges in WNS-positive counties also have low CCP scores (Table 4). Four Refuges have both a WNS status and karst features. Although three of these four Refuges also have high-scoring CCPs, the Upper Mississippi River National Wildlife and Fish Refuge CCP has no mention of bats and thus may be particularly vulnerable to the threat of spreading WNS. Eleven Refuges were found to have at least some land in counties with either confirmed or suspected infections of P.d. Table 4 lists confirmed cases first, then suspected cases, and is ordered by the IP metric within those groups. Many of these Refuges have recently written, high-scoring CCPs, but most CCPs will require at least some updates to address this growing threat. Relatively low species diversity results in low IP metrics for many of these Refuges, but WNS impacts or threats to these units may be an overriding reason to prioritize bat conservation, especially measures to protect hibernacula and prevent possible anthropogenic transmission.

Wind power potential was found to be high at 37 of the 217 Refuges examined in the spatial analysis, medium at 69 Refuges, and low at 111 Refuges. As would be expected, high wind values are most prevalent for Refuges in the midwestern portions of USFWS Regions 2, 3, and 6. Many of the high-wind Refuges also have low CCP scores and resultantly high improvement potential. The 25 Refuges with both high wind power potential and positive IP metrics are shown in Table 5. Although species diversity is relatively low among this group, all of the high-wind Refuges are within the range of at least two of the three at-risk species of North American migratory tree-roosting bats (hoary bat, eastern red bat, and silver-haired bat Lasionicteris noctivagtans), and most are within the range of all three species. Fatality patterns observed at wind energy facilities have indicated that these three species may be disproportionately at risk from wind energy development, especially during late summer and early fall (Arnett et al. 2008).

We recommend that Refuge managers, planners, and cooperating agencies involved in Refuge planning should, first and foremost, undertake a thorough preplanning process before revising a CCP; the process should include reevaluating the attention given to different taxa. Given the continued push for wind energy development and its mostly unknown effects on the already poorly studied migratory tree bats, as well as the ongoing expansion of the P.d. fungus, it is likely that most Refuges are at least within the range of one or more species of concern. Refuges should explore their roles, if any, in the life cycles of the diversity of taxa, including bats. Among the roles a Refuge could have are the presence of hibernacula, maternity colonies, migratory roosts, and foraging habitats. Determining the relative importance of a particular life-cycle component for a given species at a particular Refuge is outside the scope of this paper and would likely require expert opinion, literature surveys, and field research; the absence of these data at the start of planning would naturally lend itself to the addition of research goals and objectives in the CCP. Once a planning team feels that they have or could have a role to play in maintaining sustainable populations of bat species, they may need to develop additional objectives regarding research, management, land protection, and conservation partnerships.

Baseline inventories are lacking on many Refuges. “Currently, no information exists concerning bats on the Refuge. Mist netting studies will be necessary to determine the species composition and status of bats on the Refuge. Initiation of these studies will depend on funding levels” (USFWS 1999). This quote is promising in that it indicates that the Refuge (Deep Fork NWR) recognizes a real need for baseline data, but it believes pragmatic concerns may preclude its collection in the near term. Rather than deferring needed research, we recommend that baseline inventories and ongoing monitoring be conducted using acoustic surveys. Acoustic surveys tend to account for more species than mist netting (O'Farrell and Gannon 1999). Also, because monitoring stations can be automated, they require far less personnel than thorough trapping surveys. Identification of recorded calls to species, once onerous and requiring potentially subjective expert opinion, has become more straightforward with development of objective techniques (e.g., Britzke et al. 2011) and software capable of automated species identification such as Sonobat 3 (www.sonobat.com), BCID (www.batcallid.com), and EchoClass (http://www.fws.gov/midwest/Endangered/mammals/inba/inbasummersurveyguidance.html). It is important to note that some such surveys have been undertaken since the development of the reviewed CCPs: for example, the USFWS Inventory and Monitoring Initiative completed its first summer of Refuge bat surveys in 2012 at 13 Refuges in the Pacific Northwest (Barnett 2012). Whether it is through partnerships between individual Refuges and external researchers, or the adoption of bat surveys as a national Inventory and Monitoring priority, it is critical that, at a minimum, Refuges are aware of what and how bat species of concern use their lands so that they can base informed management decisions on well-written strategies.

Beyond simple presence or absence or activity surveys, it is important for a Refuge to understand what the bats are actually doing there so that common management practices such as building renovation and prescribed fire do not harm local populations. This understanding requires, for example, the identification of hibernacula, important roosts for migratory tree bats, and maternity roosts. If the Refuge suspects that important roosts may be present but they have insufficient knowledge to inform management, the CCP could include an explicit objective related to research needs. For example, D'Arbonne NWR's CCP states in Species of Special Concern Objective 4, “Conduct a research project to determine roost habits, reproductive success, and wintering roost locations of Rafineseque's big-eared bats and southeastern Myotis bats on the refuge” (USFWS 2006a).

Funding and expertise to conduct research may not be present within a given Refuge, but this is not a reason to avoid filling vital knowledge gaps. When determining how to gather needed data, Refuge biologists and managers should think outside the boundaries of their Refuge. They should leverage the expertise present in universities, agency partners, and nongovernmental organizations that may be accessible at little or no cost if the data generated have other uses (e.g., if they are part of a broader experiment or a regional monitoring program). One such strategy was included in the threatened and endangered species objective for Patoka River NWR and Management Area: “Every five years cooperate/contract with university coop unit/ES endangered species specialist to determine status of Indiana bats on the Refuge” (USFWS 2008). Wallkill NWR takes a similar approach in their CCP, calling for collaboration with Great Swamp NWR to recruit students to conduct research on bat ecology on the Refuge (USFWS 2009c). Farallon NWR goes one step further by recognizing the importance of the NWRS in facilitating basic scientific research, even though it does not harbor any threatened or endangered bat species. That Refuge, with participation from researchers from Point Reyes Bird Observatory and U.S. Geological Survey, is participating in active, long-term monitoring of bat migration through the Farallon Islands to help elucidate the natural history of poorly understood migratory species (USFWS 2009a).

In evaluating the roles that a Refuge can play in bat conservation, and analyzing the potential effects of other refuge management actions on bat species, managers should recognize that these species each have unique life histories.

Although species such as the little brown bat and big brown bat are well known for forming summer colonies in artificial structures, other species often or exclusively use trees, including federally protected bat species such as the Hawaiian hoary bat, Mariana fruit bat, Indiana bat, and the ESA-petitioned northern long-eared bat. Roost selection is nonrandom in tree-roosting species (Foster and Kurta 1999; Limpert et al. 2007; Perry and Thill 2007; Hein et al. 2009). Factors influencing roost tree selection include characteristics of the trees themselves (e.g., diameter at breast height, height, species), vegetative structure, and landscape context (e.g., proximity to riparian areas), among others. A thorough review of tree bat ecology is beyond the scope of this paper, but it is incumbent upon those involved in comprehensive conservation planning to understand the natural history of tree bats on their Refuges and to develop and implement management strategies that encourage their persistence. Examples of habitat management targeted to tree-roosting bats include the following:

• “No suitable trees will be removed between April 30 and October 1 [when females are rearing pups]. Suitable trees include any species greater than 9″ diameter at breast height. Exceptions include … suitable trees for which a competent wildlife biologist determines via exit survey that no bats (any species) are present.” —Shiawassee NWR (USFWS 2001)

• “Encourage colonization of Indiana bats on Refuge Complex land through forest restoration (day roost and nursery habitat) … throughout the life of this plan … Ensure that 20 percent of tree species (big nut and shell bark hickories) used in future forest restoration contribute to meeting the needs of Indiana bats.” —Illinois River National Wildlife and Fish Refuge Complex (USFWS 2004)

• “[A range of possible management alternatives] would be expected to positively benefit the [hoary bat], primarily due to actions promoting late successional characteristics in forested habitats (considered best for roosting), the actions to restore riparian habitats, and the maintenance of some fields as openings.” —Little Pend Oreille NWR (USFWS 2000a)

Most species of non-Lasiurus bats in the United States form summer colonies, often with disproportionately high representation of females. If these maternity roosts are found on the Refuge, managers should endeavor to ensure their persistence. Female bats can exhibit fidelity to the maternity colonies where they congregate to birth and rear pups (Arnold 2007; Dixon 2011) and can vacate an area entirely if they are disturbed or excluded from their roosts (Neilson and Fenton 1994). Lactating females have limited home range sizes due to energy constraints (Henry et al. 2002), so even if other potential roosts are present elsewhere on or near a Refuge, they may or may not provide the density of resources that resulted in the success of an existing colony. Therefore, if exclusion of bats is necessary for reasons of public health or to ensure the integrity of a historic building, efforts should be made to maintain the roost near the original location. If done carefully by providing a range of alternative artificial roosts appropriate for the conditions of the site, it is possible to maintain colonies after they have been excluded from a structure (Brittingham and Williams 2000). Regardless of the solution chosen, exclusion should not occur until all the young of a colony have fledged so that nonvolant individuals are not trapped inside the structure. Examples of strategies for structure-roosting bat conservation include the following:

• “Refuge structures/facilities planned for closure or removal should be surveyed for use as a bat roost site before closure/removal.” —Felsenthal and Overlow NWRs (USFWS 2010b)

• “Continue protection of abandoned houses used by Rafinesque's big eared bats and conduct research as warranted regarding movement patterns and special needs (e.g., frequency and distribution). —St. Catherine Creek NWR (USFWS 2006b)

Cave-hibernating and summer cave-roosting bats are conspicuously vulnerable. Many of these species form large aggregations, sometimes into the hundreds of thousands, and they have long been subject to vandalism, disturbance by spelunkers, direct persecution, and now WNS. All three bat species currently under review for ESA protection are obligate cave or mine hibernators. The destruction of a colony in a single cave or mine can have regionally significant population repercussions. For example, a comprehensive survey of bat hibernacula in Minnesota found that roughly 99% of the state's little brown bats were hibernating in just two abandoned mines (Nordquist 2000). The NWRS has made great efforts to protect hibernacula on their lands; indeed, Refuges such as Key Cave, Sauta Cave, Fern Cave, and Wheeler NWRs were established specifically to protect colonies of Indiana bat and gray bat. The next role for the NWRS is to actively participate in research and recovery efforts for struggling cave bats, particularly in the eastern United States. Great Bay NWR, for example, is proposing to experiment with modified, abandoned Army bunkers on the Refuge to serve as hibernacula and possibly refugia from WNS because, unlike natural caves, they can be easily disinfected after the bats have departed in the spring.

The policy of the USFWS is to integrate land acquisition planning efforts into CCP preparation whenever possible (USFWS 2000b), although the development of a land protection plan can precede the development of a CCP where the acquisition would not be part of an existing Refuge. To ensure the long-term viability of North American bats, particularly cave bats imperiled by WNS, the NWRS should actively seek to protect privately held caves and abandoned mines. Our review did not indicate that any existing Refuges were studying the acquisition of hibernacula under their current CCP, although Fern Cave NWR was studying the acquisition of lands that contained access to its eponymous cave to protect their endangered gray bat population (USFWS 2007b). It is likely that hibernaculum protection could be accomplished in most cases through the use of conservation easements, although it may be appropriate to use acquisition in fee-title in some circumstances when hibernacula are adjacent to blocks of contiguous summer habitat. Given the limited footprint of cave and mine entrances, the protection of these sites is unlikely to be expensive or controversial and could potentially protect hibernacula that contribute to summer populations over hundreds of square miles.

It would be an overstatement to suggest that Refuges can maintain range-wide populations of most species of bats, many of which are migratory and have substantially different habitat requirements throughout the year. However, even for widespread and common species, Refuges may provide a critical resource. Telemetry work conducted at Rocky Mountain Arsenal NWR indicated that big brown bats trapped at the Refuge were commuting an average of 10 times their normal nightly flight distance from day roosts in Denver, Colorado, to forage at the Refuge (Everette et al. 2001). None of the bats roosted on the Refuge, so the ability of the Refuge to manage for that species is partially dependent upon actions taken by private landowners. Similarly, Wallkill NWR discovered the presence of endangered Indiana bat through mist net surveys in 2008, and the Refuge was already thought to provide quality foraging habitat for the species. However, the nearest known maternity colonies and hibernacula are off-Refuge (USFWS 2009c). For both Wallkill and Rocky Mountain Arsenal NWRs, a logical approach to conserving these bats on-Refuge would include bats as part of an energetic interpretation and environmental education program. This could include specific CCP or Visitor Services Plan strategies for engaging people who have bat roosts in their homes or on their land. Habitat improvement or management on these important private lands could be facilitated by the USFWS Partners for Fish and Wildlife Program (http://www.fws.gov/partners/) that provides funding for such projects to interested landowners.

Some strategies for mitigating the impacts of wind energy development adjacent to or in the vicinity of Refuges may be made possible by early engagement with wind power developers. Adjustments to wind farm layouts are most feasible early in the planning process, and Refuge personnel may be in a position to identify sensitive or high-risk areas. In addition, wind energy companies may be willing to cooperate in mitigation efforts such as curtailment during low wind speeds (Arnett et al. 2013), or in data collection and monitoring projects (Roppe et al. 2012) that may be beneficial to the Refuge.

Another situation that requires that Refuges work outside of their boundaries to conserve the species within them is when the conservation challenges for that species occur or are anticipated to occur in arenas wherein partner agencies or other USFWS programs have primary jurisdiction. For example, in Nantucket NWR's Draft CCP, they discuss the fact that a proposed offshore wind project could impact Refuge resources including migratory tree bats but that the Refuge's authority ends at the low waterline. They propose to work with the USFWS Endangered Species field office to review wind energy proposals and to recommend studies to fill data gaps. Although this discussion was omitted from the final CCP (USFWS 2013), working with internal and external agency partners to understand and mitigate risks is an appropriate approach to managing wildlife resources that are seldom limited to Refuges.

Bats, like other conservation topics that have been reviewed in first-generation CCPs (e.g., biological management issues [Meretsky and Fischman 2012] and climate change [Fischman et al. 2012]), ranged from completely ignored to a primary focus of CCPs across the NWRS. We acknowledge that even the best-written CCP will not result in effective conservation delivery; implementation requires funding and willingness from managers to allocate it, when necessary, for active bat management or research. However, in a time of increasing vulnerability and as-yet unknown population impacts for many bat species, it is the responsibility of planners, wildlife Refuge managers, and biologists to take a hard look at the role Refuges play in the life histories of bat species and to take the steps that they can to ensure that these species continue to play their vital ecological role. All Refuges, but particularly those with poor IP metrics, should assess how their CCPs, and management in general, could be improved to address this responsibility. Bats are not the only taxa facing conservation challenges, and the NWRS is not the only large system of conservation lands. It is our hope that this approach can serve as a template for analyzing planning for other underrepresented species groups on public and private conservation lands.

Please note: The Journal of Fish and Wildlife Management is not responsible for the content or functionality of any supplemental material. Queries should be directed to the corresponding author for the article.

Table S1. National Wildlife Refuge System management plans reviewed. CCP  =  Comprehensive Conservation Plan; CMP  =  Comprehensive Management Plan; Con. MP  =  Conceptual Management Plan; Int. CCP  =  Interim CCP; CCP score  =  qualitative measure of bat emphasis in CCP.

Found at DOI: 10.3996/122012-JFWM-106.S1 (21 KB XLSX).

Table S2. Full results of CCP Scoring and GIS analysis. NWR  =  National Wildlife Refuge; USFWS Region  =  U.S. Fish and Wildlife Service administrative region; CCP  =  Comprehensive Conservation Plan; CMP  =  Comprehensive Management Plan (predecessor to current CCPs); Conceptual MP  =  Conceptual Management Plan (interim plan for new Refuges before completion of a full CCP); ESA species  =  species protected as threatened or endangered under U.S. Endangered Species Act; Species  =  number of bat species occurring on Refuge, according to range maps; CCP score  =  qualitative measure of bat emphasis in CCP; MTB  =  number of migratory tree bats (red bat Lasiurus borealis, hoary bat, and silver-haired bat) occurring on Refuge, according to range maps; Ecoregions  =  number of EPA Level III Ecoregions within the Refuge; Wind  =  wind power potential, generalized by region with scores of high, medium, or low; Karst  =  existence of karst topography within the Refuge; WNS  =  white nose syndrome status and year the status was declared, with C  =  confirmed and S  =  suspected; IP  =  improvement potential, calculated from spatial characteristics and CCP score for the Refuge. For a complete description of how the IP metric was calculated, please see Methods, Spatial Analysis.

Found at DOI: 10.3996/122012-JFWM-106.S2 (145 KB XLSX).

Reference S1. Davies, WE, Simpson JH, Ohlmacher GC, Kirk WE, Newton EG. 1984. Engineering aspects of karst. U.S. Geological Survey, Reston, Virginia.

Found at DOI: http://dx.doi.org/10.3996/122012-JFWM-106.S3; also available at http://pubs.usgs.gov/of/2004/1352/ (9.6 MB PDF).

Reference S2. USFWS. 2013. Nantucket National Wildlife Refuge Comprehensive Conservation Plan. U.S. Fish and Wildlife Service, Nantucket, Massachusetts.

Found at DOI: http://dx.doi.org/10.3996/122012-JFWM-106.S4; also available at http://digitalmedia.fws.gov/cdm/singleitem/collection/document/id/1603/rec/9 (4.9 PDF).

Reference S3. Environmental Protection Agency. 2011. Level III Ecoregions of the continental United States. National Health and Environmental Effects Research Laboratory, Washington, D.C.

Found at DOI: http://dx.doi.org/10.3996/122012-JFWM-106.S5 (2.7 MB PDF).

Reference S4. National Renewable Energy Lab. 2011. United States - Annual average wind speed at 80 m. National Renewable Energy Lab, Golden, Colorado.

Found at DOI: http://dx.doi.org/10.3996/122012-JFWM-106.S6 (4.1 MB PDF).

Reference S5. Butchkoski, CM. 2012. White nose syndrome (WNS) occurence by county/district, 08/08/2012. Pennsylvania Game Commission.

Found at DOI: http://dx.doi.org/10.3996/122012-JFWM-106.S7; also available at http://whitenosesyndrome.org/resources/map (416 KB PDF).

Reference S6. Barnett, J. 2012. USFWS Region 1, eastside refuges acoustic bat inventory. Pages in Andrusiak L, Hogan B, editors. Western Bat Working Group News, Fall 2012. Western Bat Working Group.

Found at DOI: http://dx.doi.org/10.3996/122012-JFWM-106.S8; also available at http://static.whitenosesyndrome.org/sites/default/files/resource/wbwg_newsfall_2012.pdf (2.5 MB PDF).

Reference S7. USFWS. 2012. North American bat death toll exceeds 5.5 million from white nose syndrome. News release. Office of Communications.

Found at DOI: http://dx.doi.org/10.3996/122012-JFWM-106.S9; also available at http://www.batcon.org/pdfs/USFWS_WNS_Mortality_2012_NR_FINAL.pdf (109 KB PDF).

Reference S8. USFWS. 2000b. Service Manual 602 FW 1: National Wildlife Refuge System Planning. U.S. Fish and Wildlife Service, Arlington, Virginia.

Found at DOI: http://dx.doi.org/10.3996/122012-JFWM-106.S10; also available at http://www.fws.gov/policy/602fw1.html (161 KB PDF).

Reference S9. USFWS. 2007a. Indiana bat (Myotis sodalis) draft recovery plan: first revision. U.S. Fish and Wildlife Service, Fort Snelling, Minnesota.

Found at DOI: http://dx.doi.org/10.3996/122012-JFWM-106.S11; also available at http://ecos.fws.gov/docs/recovery_plan/070416.pdf (1.4 MB PDF).

Reference S10. USFWS. 2011b. Endangered and Threatened Wildlife and Plants; 90-Day Finding on a Petition to List the Eastern Small-Footed Bat and the Northern Long-Eared Bat as Threatened or Endangered. U.S. Fish and Wildlife Service, Washington, D.C.

Found at DOI: http://dx.doi.org/10.3996/122012-JFWM-106.S12; also available at http://ecos.fws.gov/speciesProfile/profile/speciesProfile.action?spcode=A0AB (191 KB PDF).

Reference S11. USFWS. 1999. Deep Fork National Wildlife Refuge Comprehensive Conservation Plan. U.S. Fish and Wildlife Service, Albuquerque, New Mexico.

Found at DOI: http://dx.doi.org/10.3996/122012-JFWM-106.S13; also available at http://digitalmedia.fws.gov/cdm/ref/collection/document/id/400 (8.6 MB PDF).

Reference S12. USFWS. 2000a. Final Comprehensive Conservation Plan and Environmental Impact Statement for the Little Pend Oreille National Wildlife Refuge. U.S. Fish and Wildlife Service, Sherwood, Oregon.

Found at DOI: http://dx.doi.org/10.3996/122012-JFWM-106.S14; also available at http://digitalmedia.fws.gov/cdm/compoundobject/collection/document/id/739/rec/20 (4.8 MB PDF).

Reference S13. USFWS. 2001. Shiawassee National Wildlife Refuge Comprehensive Conservation Plan. U.S. Fish and Wildlife Service, Bloomington, Minnesota.

Found at DOI: http://dx.doi.org/10.3996/122012-JFWM-106.S15; also available at http://digitalmedia.fws.gov/cdm/singleitem/collection/document/id/547/rec/18 (3.0 MB PDF).

Reference S14. USFWS. 2004. Illinois River National Wildlife and Fish Refuges Complex Comprehensive Conservation Plan. U.S. Fish and Wildlife Service, Bloomington, Minnesota.

Found at DOI: http://dx.doi.org/10.3996/122012-JFWM-106.S16; also available at http://digitalmedia.fws.gov/cdm/ref/collection/document/id/727 (4.1 MB PDF).

Reference S15. USFWS. 2006a. D'Arbonne National Wildlife Refuge Comprehensive Conservation Plan. U.S. Fish and Wildlife Service, Farmerville, Louisiana.

Found at DOI: http://dx.doi.org/10.3996/122012-JFWM-106.S17 (5.9; also available at http://digitalmedia.fws.gov/cdm/ref/collection/document/id/399 MB PDF).

Reference S16. USFWS. 2006b. St. Catherine Creek National Wildlife Refuge Comprehensive Conservation Plan. U.S. Fish and Wildlfie Service, Atlanta, Georgia.

Found at DOI: http://dx.doi.org/10.3996/122012-JFWM-106.S18; also available at http://digitalmedia.fws.gov/cdm/singleitem/collection/document/id/548/rec/19 (5.4 MB PDF).

Reference S17. USFWS. 2007b. Wheeler National Wildlife Refuge Complex Comprehensive Conservation Plan and Environmental Assessment. U.S. Fish and Wildlife Service, Atlanta, Georgia.

Found at DOI: http://dx.doi.org/10.3996/122012-JFWM-106.S19; also available at http://digitalmedia.fws.gov/cdm/ref/collection/document/id/588 (9.4 MB PDF).

Reference S18. USFWS. 2008. Patoka River National Wildlife Refuge Comprehensive Conservation Plan. U.S. Fish and Wildlife Service, Bloomington, Minnesota.

Found at DOI: http://dx.doi.org/10.3996/122012-JFWM-106.S20; also available at http://digitalmedia.fws.gov/cdm/singleitem/collection/document/id/506/rec/1 (5.9 MB PDF).

Reference S19. USFWS. 2009a. Farallon National Wildlife Refuge Final Comprehensive Conservation Plan. U.S. Fish and Wildlife Service, Newark, California.

Found at DOI: http://dx.doi.org/10.3996/122012-JFWM-106.S21; also available at http://www.fws.gov/cno/refuges/farallon/ (8.7 MB PDF).

Reference S20. USFWS. 2009b. Guam National Wildlife Refuge Comprehensive Conservation Plan. U.S. Fish and Wildlife Service., Honolulu, Hawaii.

Found at DOI: http://dx.doi.org/10.3996/122012-JFWM-106.S22; also available at http://digitalmedia.fws.gov/cdm/singleitem/collection/document/id/424/rec/16 (8.0 MB PDF).

Reference S21. USFWS. 2009c. Wallkill River National Wildlife Refuge. U.S. Fish and Wildlife Service, Hadley, Massachusetts.

Found at DOI: http://dx.doi.org/10.3996/122012-JFWM-106.S23; also available at http://digitalmedia.fws.gov/cdm/singleitem/collection/document/id/585/rec/4 (3.4 MB PDF).

Reference S22. USFWS. 2010a. J. N. Ding Darling National Wildlife Refuge Comprehensive Conservation Plan. U.S. Fish and Wildlife Service, Atlanta, Georgia.

Found at DOI: http://dx.doi.org/10.3996/122012-JFWM-106.S24; also available at http://digitalmedia.fws.gov/utils/getfile/collection/document/id/403/filename/404.pdf (12.7 MB PDF).

Reference S23. USFWS. 2010b. Felsenthal and Overflow National Wildlife Refuges Comprehensive Conservation Plan. U.S. Fish and Wildlife Service, Atlanta, Georgia.

Found at DOI: http://dx.doi.org/10.3996/122012-JFWM-106.S25; also available at http://www.fws.gov/southeast/planning/PDFdocuments/FelsenthalOverflowFinal/FelsenthalOverflowFinalCCP.pdf (31.4 MB PDF).

Reference S24. USFWS. 2011c. James Campbell National Wildlife Refuge Comprehensive Conservation Plan. U.S. Fish and Wildlife Service, Honolulu, Hawaii.

Found at DOI: http://dx.doi.org/10.3996/122012-JFWM-106.S26; also available at http://www.fws.gov/uploadedFiles/Region_1/NWRS/Zone_1/Oahu_Complex/James_Campbell/Documents/James%20Campbell%20NWR%20CCP%20(final)%2012-01-11.pdf (7.8 MB PDF).

We thank M. Artmann, S. Selbo, and USFWS Region 6 Division of Refuge Planning for review and input on this manuscript, as well as R. Fischman, an anonymous reviewer, and the editors of the Journal of Fish and Wildlife Management for critical comments and recommendations that substantially improved the document. This paper received support from USFWS Region 6 and USFWS Region 3.

Any use of trade, product, or firm names is for descriptive purposes only and does not imply endorsement by the U.S. Government.