Capture and Reproductive Trends in Summer Bat Communities in West Virginia: Assessing the Impact of White-Nose Syndrome

In the northeastern and Mid-Atlantic United States, white-nose syndrome (WNS) has severely impacted most bat species that hibernate in caves (Turner et al. 2011), such as members of the genus Myotis, the tri-colored bat Perimyotis subflavus, and to a lesser extent, the big brown bat Eptesicus fuscus; all have confirmed individuals with the fungus, Geomyces destructans, on their nonfurred membranes (Gargas et al. 2009; Meteyer et al. 2009) and documented mortality. G. destructans is believed to be the causative agent of WNS and the mortality of >1 million bats (Blehert et al. 2009). Modeling efforts based on WNS-associated mortalities predict regional extinction of the little brown bat in as little as 16 y (Frick et al. 2010). Whether WNS can cause extirpation of other myotines in the region is unknown but cannot be discounted.

In West Virginia, WNS was first detected in several caves in Pendleton County in the eastern portion of the state in January 2009 and in Hellhole Cave in January 2010. The latter discovery was particularly worrisome to managers because Hellhole is the winter hibernaculum (i.e., overwintering site) for >6,000 endangered Virginia big-eared bats Corynorhinus townsendii virginianus, >12,000 endangered Indiana bats Myotis sodalis, and >150,000 little brown bats M. lucifugus (Stihler and Brack 1992; Johnson and Strickland 2004; Szymanski et al. 2009). Given known migratory pathways and the presumed pathway of transmission of WNS from the Northeast to the Mid-Atlantic, it was believed that WNS was confined to the Ridge and Valley province of West Virginia in summer 2009 (Szymanski et al. 2009). By winter 2009–2010, WNS transmission likely extended across the state after entering the Appalachian Plateau province by spreading west of the Allegheny Front and south from hibernacula in Pennsylvania (Britzke et al. 2009; Szymanski et al. 2009). Indeed, winter 2010–2011 surveys detected WNS in hibernacula west of the Allegheny Front in the Greenbrier Limestone formations. This included Big Springs Cave (Tucker County), the state's largest Indiana bat hibernaculum on public lands (C. Stihler, West Virginia Division of Natural Resources, personal communication).

Researchers are attempting to understand the short-term effects of WNS on surviving bats because postemergence health of bats that survive winter hibernation might remain compromised. Some bats have visible wing-membrane damage (Blehert et al. 2009; Reichard and Kunz 2009), but it is unclear whether these lesions directly manifest themselves on additional aspects of physiological well-being or impact reproductive success (Cryan et al. 2010).

One method to assess effects of WNS on bat fitness is to compare pre- and post-WNS capture rates by species and within-season temporal trends in captures of females that are pregnant or lactating, and captures of volant juveniles later in summer. We hypothesized that post-WNS capture success of cave-hibernating bats would decline, especially within the genus Myotis. In WNS-affected bats, we predicted that females in 2010 would exhibit delayed or shorter reproductive cycles, as documented by a greater proportion of pregnant females captured later in the summer. Reproduction would be delayed when energy demands exceed availability and daily torpor is used to bridge this gap and fetal development is slowed (Racey 1973). If this delay is common, the period over which a population's parturition and lactation occurs would be compressed, shortening the reproductive cycle. Little brown bats born earlier in the summer (late May) have a significantly higher probability of surviving their first year than young born later in the summer (mid-July; Frick et al. 2009). In addition, though we perceive it as less likely, if WNS-caused energy constraints persist through the summer, pups might be weaned earlier. Thus, we predicted that captures in 2010 would reveal a shorter period of lactation (beginning later and perhaps ending earlier) as compared to pre-WNS capture trends. Finally, severe energy constraints might result in fetus resorption or abortion, or an inability to care for offspring. We expected higher reproductive failure to reveal itself in a lower relative proportion of juvenile captures. Herein, we compare trends in capture and reproductive rates and timing of reproductive events for bats in years before WNS was recognized in West Virginia (1997–2008; hereafter, “pre-WNS”), and after (2010; hereafter, “post-WNS”). Although summer 2009 data were available, we did not use them because WNS was not yet conclusively statewide in occurrence.

We compiled data from 37 counties in West Virginia: 31 counties surveyed prior to detection of WNS in the state (1997–2008; 3,577 net-nights) and eight counties surveyed in 2010 (892 net-nights). Across all years, all data were collected from 15 May to 15 August of survey years, and included netting data from published research studies (Gates and Johnson 2006; Schirmacher et al. 2007; Johnson et al. 2009, 2010), and unpublished regulatory compliance surveys for the Indiana bat that are routinely required to demonstrate compliance with the Endangered Species Act (ESA 1973, as amended). These surveys are conducted any time a federally funded and/or federally permitted project (such as construction of pipelines, roads, or coal mines) involve the removal of trees that are used by Indiana bats for summer roosts. All surveys for the endangered Indiana bat were completed following the U.S. Fish and Wildlife Service (USFWS) protocol (USFWS 2007). The 2010 data included May surveys from southern Green County, Pennsylvania (immediately adjacent to West Virginia) because compliance work cannot begin in West Virginia until 1 June. All 2010 records and the majority (≥92%) of historical records followed standard Indiana bat surveying protocol of two nets set for two nights per site for 5 h each (USFWS 2007). Because not all pre-WNS surveys used exactly two nets per night (8.0% of sites), we standardized efforts to a per-net-night basis. Also, because on a limited number of nights (4.3% of sites), researchers did not survey for 5 consecutive h, we counted those net-nights as a proportion of a full 5-h net-night (e.g., one net set for 4 h was calculated as 4/5 or 0.8 net-nights). If net set-ups or hours surveyed were not explicitly documented, we did not include these records in captures per-net-night analyses.

Species, sex, reproductive condition (pregnancy determined by abdominal palpitation and lactation determine by condition of mammary glands [Haarsma 2008]), and age (juvenile or adult based on degree of epiphyseal ossification [Anthony 1988]) were recorded for each captured bat. If age or reproductive condition (in females) was not clearly defined, we removed these records from sex- or age-specific analyses.

In SYSTAT 10, we used a two-sample t-test (SPSS, Inc., 2000), to statistically examine pre-WNS vs. post-WNS capture rate (captures per net-night) by species. One sample was considered one night's survey at one site. Because proportional reproductive data are not commonly analyzed statistically (Kurta and Teramino 1992) and no clear method is utilized when the factor of time is added, additional analyses were limited to descriptions of visual trends. Therefore, in SAS 9.2, we employed localized regression (PROC LOESS; SAS Institute Inc., Cary, North Carolina) to examine species-specific trends in reproductive condition and the proportion of juveniles captured each night between pre- and post-WNS periods. Localized regression is a nonparametric line-fitting procedure that is appropriate in visualizing trends without forcing a fit to standard linear, exponential, or other parametric forms. Because we lacked precognition of patterns in these data, we believe this regression with few assumptions and limited statistical metrics was an appropriate exploratory choice.

We compared capture rates by species for 11,831 bats pre-WNS and 1,310 bats post-WNS where we could calculate captures per net-night by species (Table 1; Tables S1–S3, Supplemental Material, http://dx.doi.org/10.3996/062011-JFWM-039.S1). We included approximately 100 bats that escaped before identification in our calculations of total capture rate. Post-WNS capture rates for six species declined significantly: the northern long-eared Myotis septentrionalis (22.9% of historical rate; t  =  13.086, P < 0.001), little brown (20.1%; t  =  12.262, P < 0.001), tri-colored (22.9%; t  =  10.113, P < 0.001), Indiana (10.8%; t  =  2.837, P  =  0.005), and eastern small-footed M. leibii (16%; t  =  5.254, P < 0.001; Table 1) bat. Capture rates for hoary bats Lasiurus cinereus also declined (t  =  4.177, P < 0.001), with captures rates approximately 37.0% of the historical capture rate (Table 1), which is surprising given that the species is thought to be minimally impacted by WNS. Capture rates for big brown bats were 18.8% higher in 2010 but not significantly different between pre- and post-WNS (t  =  1.361, P  =  0.174). Eastern red Lasiurus borealis bat capture rates did not change between the two time periods (post-WNS rates 97.5% of historical rates; t  =  0.247, P  =  0.805; Table 1). We did not compare capture rates for three species captured only in pre-WNS surveys: Virginia big-eared, evening Nycticeius humeralis, and silver-haired Lasionycteris noctivagans (Table 2) bats.

We documented 11,083 bats pre-WNS and 1,231 post-WNS that we could examine for age and reproductive trends. To ensure sufficient capture comparisons pre- and post-WNS, we focused our analyses on the four most commonly captured species both pre- and post-WNS: the northern long-eared, little brown, big brown, and eastern red bat. Pre-WNS pregnancy rates for little brown bats peaked at approximately Julian date (JD)  =  162 (11 June), with 57% of females pregnant (Figure 1A). Post-WNS, peak pregnancy was earlier in the season (JD  =  150, 30 May) with pregnancy rates of 66% (Figure 1A). A similar pre-WNS pattern was observed for the northern long-eared bat, with a peak pregnancy rate of 80% at JD  =  158 (7 June; Figure 1B). Post-WNS, the best-fit line displayed nearly 100% pregnancy early in the season (JD  =  140, 20 May) and a sharp decline thereafter (Figure 1B). By JD  =  158, pregnancy rates of northern long-eared bats were half the historical rate (Figure 1B). Big brown bat pregnancy rates were similar pre- and post-WNS, displaying less dramatic declines. In both time periods, pregnancy rates of big brown bats peaked at approximately 60% at the beginning of the survey period, with steady declines through JD  =  180 (29 June), when pregnancy rates were near zero (Figure 1C). Finally, eastern red bats showed weak patterns pre-WNS, with pregnancy rates reaching just 30% at the beginning of the survey period (Figure 1D). Post-WNS, pregnancy rates of eastern red bats peaked at 80% by JD  =  145 (25 May) and cessation of pregnancy was reached by JD  =  165 (14 June; Figure 1D).

In examining lactation rates for the four species evaluated for pregnancy, pre-WNS rates for the little brown bat remained at approximately 60% for approximately 30 d (JD  =  170–200, 19 June–19 July; Figure 2A). Post-WNS, lactation rates peaked at 42% at JD  =  180, with a markedly shorter period of lactation (Figure 2A). Post-WNS, lactation peaks pre-and post-WNS were both approximately 80% for the northern long-eared bat, but the peak date shifted from JD  =  195 (14 July) pre-WNS to JD  =  178 (27 June) in 2010 (Figure 2B). Lactation rates of big brown bats peaked at 82% with a narrow peak at JD  =  180. Pre-WNS, the peak lactation rates of big brown bats (65%) remained constant for approximately 25 d (JD  =  170–195, 19 June–14 July; Figure 2C). Peak lactation rates for eastern red bats shifted to earlier in the season (JD  =  175–180, 24 June–29 June pre-WNS as compared to JD  =  170 in 2010), but peaks for both time blocks were approximately 60% (Figure 2D).

The proportion of juveniles captured peaked at the end of the survey period (JD  =  227) for little brown bats (Figure 3A). Pre-WNS, the proportion of juvenile little brown bats was 60%, but just 20% in 2010 (Figure 3A). Similarly, proportion of juvenile northern long-eared bats declined by more than half, with juvenile captures peaking at JD  =  217 (5 August, 40% pre-WNS), and JD  =  227 (18%) in 2010 (Figure 3B). No difference in the proportion of juveniles was noted for the big brown bat pre- vs. post-WNS, with captures peaking at 35% (JD  =  220, 8 August; Figure 3C). Captures of juvenile eastern red bats also peaked at the end of the survey period (JD  =  227; 55%) for the pre-WNS data set (Figure 3D). In 2010, proportion of juvenile eastern red bats remained at 45% for approximately 20 d (JD  =  195–215, 14 July–3 August; Figure 3D).

In light of these findings, we also examined the differences in relative proportions of species between our two data sets. We documented 11,734 bats pre-WNS and 1,304 post-WNS that were identified to species (Table 2). Before discovery of WNS in West Virginia, the community composition was dominated by northern long-eared (41.1% of all captures) and little brown (25.3%) bats. Big brown (11.5%) and eastern red (11.1%) bats also were commonly collected. Eight additional species comprised the remaining 11% of the catch (Table 2). In 2010, the same four species again were most common, but the distribution of individuals was markedly different. Big brown (33.2%) and eastern red (25.5%) bats comprised the majority of captures in 2010, with fewer northern long-eared (24.1%) and little brown (11.9%) bats. Four additional species comprised the remaining 5.3% of captures (Table 2).

In West Virginia, a significant decline in capture success and changes in reproductive parameters of the myotids and tri-colored bats suggest that WNS effects are profound. An increase in captures of big brown bats, coupled with the stability in captures of eastern red bats, may signal a shift in bat community composition and dominance from a community formerly dominated by little brown and northern long-eared bats. Ford et al. (2011) posited that big brown response could be an ecological release, perhaps due to less auditory crowding in foraging areas once dominated by little brown bats, although significant changes in big brown bat numbers rather than increases in local foraging activity is speculative. Low capture rates of adults of these two myotids may be accentuated by low recruitment in 2010. These data from West Virginia for little brown bats are consistent with regional extinction models presented by Frick et al. (2010) for the northeastern United States.

While it is important to remember these observations are based on 1 y of data, this recruitment level is below those needed for annual replacement even before considering overwinter mortality. Although weather in a topographically complex area such as West Virginia is highly variable, July–August 2010 was near normal for precipitation. It was slightly above average for temperature across all divisions within the state (NOAA National Climatic Data Center, http://www.ncdc.noass.gov/temp-and-precip/). These conditions should not have greatly influenced bat foraging activity or biased susceptibility of netting efforts as might other weather conditions (e.g., wetter and cooler).

Increased reproductive success of big brown bats may be counterintuitive for a species that can be found hibernating in the same WNS-positive caves as the myotids (e.g., Blehert et al. 2009). However, a number of factors may make this species less susceptible to effects of WNS. First, most big brown bats likely overwinter in alternative hibernacula (attics, hollow trees, rock crevices; Whitaker and Gummer 1992, 2000; Neubaum et al. 2006), and raises the possibility that cave-hibernating individuals may travel to alternate sites during the hibernation period. Their tolerance for warmer hibernacula (up to 20°C; Whitaker and Gummer 1992, 2000) suggests that exposure to WNS may be far more limited than at large communal hibernacula. In addition, they routinely hibernate at warmer temperatures (3–20°C; [x¯  =  10°C] in Indiana [Whitaker and Gummer 1992] and 9.5 ± 1.5 in Ohio [Brack 2007]) that may be above the optimal range for G. destructans (Blehert et al. 2009). Typically, only a few individual big brown bats are found in each cave, and they are often solitary or clustered in small groups during hibernation (Whitaker and Gummer 1992; Brack 2007), which may play a role in reducing the transfer of spores (Cryan et al. 2010). Moreover, their greater volume : surface area ratio suggests that these bats have a higher threshold for starvation, and are capable of losing relatively more body mass during the winter without mortality compared to smaller bodied myotids (Davis 1970). The ecological phenomenon of disease-mediated, competitive release is well-documented for a variety of species (Bradley and Altizer 2006).

While we found an increase in big brown bat captures post-WNS, our juvenile big brown bat numbers do not suggest increased recruitment. Ford et al.'s (2011) research in New York suggests that there has been a shift in habitat use; post-WNS, big brown bats are more frequently documented where they were once suppressed by little brown bat activity—and that that could be hinting at release. It is not unexpected to postulate increased survival with decreased competition; indeed, Tuttle (1976) found that juvenile gray bats Myotis grisescens that traveled farther to feed gained weight more slowly and experienced greater mortality. On the other hand, because there are few actual net or cave counts that definitively document more big brown bats, we presume the bats are merely foraging where they previously could not. The extra benefit isn't translated into higher productivity because big brown bat mortality, albeit lower than myotids, is still on the rise from WNS. This is a hypothesis that researchers should continue to investigate.

One caveat about our big brown bat findings is that part of the increased capture rate of this species may be related to researcher and regulatory biases. Much of the data presented herein were collected while conducting surveys for the endangered Indiana bat. During these surveys, biologists are likely to give preference to removing a myotid (i.e., potential Indiana bat) from the net over a big brown bat and, thus, big brown bats are more likely to escape. Following WNS, multiple captures occur less often and possibly big brown bats escape less often.

Our study provides a preliminary examination of the West Virginia bat community immediately after detection of WNS. The dramatic declines of myotids provide substantive evidence that these species increasingly are in peril over large portions of their range. We acknowledge that our before-and-after data sets are noisy and imperfect but robust in size. Ideally, we would have revisited the same sites pre- and post-WNS, but the reality of regulatory compliance studies is that our grouping can only be by intra-state and/or region. Additionally, in 2009, the USFWS ruled that compliance surveys could not begin in West Virginia until 1 June. Therefore, researchers can no longer rely on data gathered solely from compliance surveys if they want to examine trends pre-June 1 in this state. Our data set serves as a baseline. In future summers, researchers can track whether our documented low captures continue.

It is unknown whether these reproductive trends are short-lived or lasting shifts. We acknowledge that species with low sample sizes (e.g., hoary and eastern small-footed bats) may have inherently higher variation and bias. However, consistently negative trends within our four most common species cannot be ignored. Continued monitoring of bat communities in West Virginia and other areas of the northeastern and mid-Atlantic United States are the basis for documenting long-term reproductive trends for both common and rare species. Research (Kannan et al. 2010) and modeling (Frick et al. 2010) of bat populations in the northeast, where WNS has been documented in hibernacula since 2006 or 2007, do not project a positive scenario for several bat species once considered common in West Virginia.

Potential declines in some bat populations as a result of WNS cannot be taken in isolation. It is part of a larger group of anthropogenic and ecological threats. Additive and cumulative effects come from landscape alterations to roosting and foraging habitat from surface mining (Townsend et al. 2009), intensive forest management (Menzel et al. 2002; Owen et al. 2004), and wind-energy development (Arnett et al. 2008). For example, the Mountaineer Wind Project in Tucker County, West Virginia recorded some of the highest bat mortality rates in the eastern United States among wind-energy production sites (Kerns and Kerlinger 2004). In addition to high mortality of migratory bats such as eastern red (41% of documented casualties) and hoary (18.9%) bats, the Mountaineer site also had high mortality of little brown (18.3%) and tri-colored (18.3%) bats, with >2,000 deaths (all species combined) in 2003 (Kerns and Kerlinger 2004; Arnett 2005; Arnett et al. 2008). It is possible that the decline in capture rates for hoary bats (a species not known to be susceptible to WNS) is related to high mortality at area wind farms. The cumulative effects of anthropogenic forces and WNS may result in regional extirpation of several species. At present, USFWS is conducting status reviews for eastern small-footed and northern long-eared bats to determine if they warrant listing as threatened or endangered under the Endangered Species Act as proposed by a petition submitted by the Center for Biological Diversity (2010). Similarly, a petition for emergency listing (expediting the listing process) of the little brown bat (Kunz and Reichard 2010) has precipitated a 90-d status review (a 90-d review of evidence submitted to determine if petitioning is warranted).

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. Captures per trap-night for 2,249 netting events in West Virginia before (pre-2009) or since (2010) detection of white-nose syndrome in the state. Provided are the date of survey and captures per net-night for each bat species (see Table S3 for codes) and for all captures (Total CNN).

Table S2. Age (adult vs. juvenile), sex, and reproductive condition (NR  =  nonreproductive, PREG  =  pregnant, LAC  =  lactating, PL  =  postlactating, SCR  =  scrotal, or UNK  =  examined but unknown) for 12,314 bats examined from 1997 to 2010 during summer captures in West Virginia. Provided are the year and Julian date of survey, and the bat species (see Table S3 for codes).

Table S3. Bat scientific names and four- or five-letter codes utilized in Tables S1 and S2.

All found at DOI: http://dx.doi.org/10.3996/062011-JFWM-039.S1 (491 KB XLSX).

We thank Environmental Solutions and Innovations, Inc. team leaders and field assistants for assistance with data collection. We are grateful to additional data providers: J. Chenger (Bat Conservation and Management), J. Johnson (University of Maryland - Appalachian Labs), B. Sargent (WVDNR), M. Schirmacher (Bat Conservation International), K. Tyrell (BHE Environmental), S. Jones (USFWS), C. Johnson (USDA Forest Service), J. D. Wilhide (Compliance Monitoring Labs, Inc.). We also thank the Subject Editor and reviewers for their comments on previous versions of this manuscript.

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