Abstract
Adult male big brown bats (Eptesicus fuscus) sometimes occur within maternity roosts. We investigated male attendance at big brown bat maternity roosts in a Colorado study area that shows a pattern of sexual segregation by elevation. We tested multiple predictions of three nonmutually exclusive hypotheses to explain male attendance patterns: 1) sex-specific differences in energetic strategies of males and females are maintained at the lower elevation; 2) natal philopatry of inexperienced young males accounts for most attendants; 3) males gain a reproductive advantage for late summer mating at maternity roosts. We tested predictions based on captures of bats at emergence, automated monitoring of annual, seasonal, and daily roost attendance by known-age bats tagged with passive integrated transponders, and anatomical evidence for mating. Findings were most consistent with the first two hypotheses. Adult males accounted for just 3.1% of 8,192 captures of bats at 285 evening emergences from 46 roosts during 2001–2005. Daily attendance rates of males during each active season (0.10 detections/day at age 1 year and 0.20 detections/day at ≥ 2 years old) were lower than in females (0.34 at age 1 year and 0.45 at ≥ 2 years old). Only 92 of 299 males tagged as volant juveniles were detected as adults at five maternity roosts monitored 2002–2005, far fewer than female returns in every age category. We detected known-age adult males almost exclusively at their natal roosts and most males that returned (91 of 92) in years after tagging as juveniles were first detected as 1-year-olds; 20 of 21 individuals that returned at 2–4 years of age were previously detected as 1-year-olds. One-year-old males were re-encountered at annual rates 5–16 times higher than 2- to 4-year-old males, and 1-year-old males and females had lower daily attendance rates than older bats. The male reproductive advantage hypothesis was not well supported. None of 80 females examined in late summer had evidence of insemination, and not all males showed distended scrota. Daily attendance rates of tagged adult males (n = 155) and females (n = 788) were lowest during late summer, suggesting that little reproductive advantage was accrued by males utilizing maternity roosts. Attendance of male big brown bats at maternity roosts at our study area is consistent with the sex-specific energetic strategies and natal philopatry hypotheses, and mating probably occurs at higher elevation autumn roosts and hibernacula.
The behavior and ecology of male insectivorous bats at roosts in temperate zones have not been well studied in comparison with females (Weller et al. 2009; Alston et al. 2019), despite the demographic importance of males for conservation. One notable exception, however, has been the finding that geographic and habitat segregation of the sexes occurs in multiple species of insectivorous bats in various temperate regions of the world (McGuire and Boyle 2013). Examples of sexual segregation are widespread in group-living terrestrial and marine vertebrates, and can be the result of a variety of selective pressures (Ruckstuhl and Neuhaus 2002, 2005; Ruckstuhl 2007; Wearmouth and Sims 2008). In insectivorous bats, its most obvious form is the disproportionate occurrence of adult males and nonreproductive females at higher elevations, contrasting with the occurrence of reproductive females in maternity roosts at lower elevations (reviewed in Weller et al. 2009; McGuire and Boyle 2013).
Sexual segregation by elevation in temperate-zone insectivorous bats is thought to be driven primarily by the energetics of reproduction: females occupy maternity roosts at warm lower elevation sites because warmer temperatures and shared body heat facilitate rapid growth and development of young, and because abundance of prey is presumably higher at lower elevations (McGuire and Boyle 2013). In contrast, males roosting solitarily or in small groups at higher elevations can use heterothermy and lower their body temperatures to achieve energy savings. Males can adjust metabolically to presumed lower food availability at higher elevations by reduced energy expenditure, and risk of predation may be decreased by roosting apart from colonies and being less active at night (Weller et al. 2009; McGuire and Boyle 2013). In some species of bats, males may be actively excluded from lower elevation sites by females (Weller et al. 2009), but in other species, a small proportion may roost with females at lower or intermediate elevations and gain an advantage in mating and reproductive success (Senior et al. 2005; Angell et al. 2013).
In temperate western North America, seasonal elevational segregation of the sexes in areas with mountainous topography has been noted for at least 17 species of insectivorous bats ranging from Mexico to Canada, including the big brown bat, Eptesicus fuscus (McGuire and Boyle 2013). Elevational segregation of big brown bats is widespread and includes locations as diverse as Texas (Easterla 1973); British Columbia (Fenton et al. 1980); South Dakota (Cryan et al. 2000); Washington (Baker and Lacki 2004); and Colorado (Neubaum et al. 2006; O’Shea et al. 2011a). Although adult male big brown bats also use separate roosts as solitary individuals or in small groups in eastern and central North America (Mills et al. 1975; Willis et al. 2006; Perry and Thill 2008), observers have sometimes noted the presence of adult males at big brown bat maternity roosts during summer in these regions, including areas with and without significant topographic relief. Adult male big brown bats have been reported to use maternity roosts mostly during late summer in Maryland (Christian 1956), Ohio (Mills et al. 1975), and Kentucky (Davis et al. 1968), but also have been found earlier in summer at maternity roosts in Indiana, Illinois, and Ontario (Brigham and Fenton 1986; Vonhof et al. 2006, 2008).
Details regarding presence and attendance patterns of males using maternity roosts have not been well-quantified in any region and the possible functions of this behavior have not been explored. Past research, however, has revealed avenues for further investigation. The use of maternity roosts by males may be consistent with the differing energetic strategies of the sexes thought to be a driver of elevational segregation (Weller al. 2009; McGuire and Boyle 2013). At some maternity roosts, the use of microclimates by adult males and females within the same structure differs based on their separate energetic strategies (Hamilton and Barclay 1994; Grinevitch et al. 1995). Unpublished observations cited by Hamilton and Barclay (1994) and Grinevitch et al. (1995) further suggest that some adult male big brown bats captured at maternity roosts in buildings at Medicine Hat, Alberta, were captured at roosts where they were first marked when juveniles, and genetic analysis has revealed that some males found at maternity roosts in Illinois and Indiana are either sons or fathers of females using the same roosts (Vonhof et al. 2008). Unspecified reproductive activity is thought to account for the late summer presence of adult males at a maternity roost of big brown bats in Kentucky (Davis et al. 1968), and Vonhof et al. (2006) speculated that the summer occurrence of both sexes at maternity roosts in Indiana and Illinois may be linked with potential future mating opportunities. Late summer mating at maternity roosts occurs in Daubenton’s bat (Myotis daubentonii) in Germany (Encarnação 2012; Encarnação and Reiners 2012) and at intermediate elevations in England (Senior et al. 2005; Angell et al. 2013). Other examples of late summer mating in Old World temperate-zone bats include M. myotis (Zahn and Dippel 1997) and two species of Pipistrellus (Gerell-Lundberg and Gerell 1994). In North America, late summer mating has been documented in M. lucifugus in Ontario (Thomas et al. 1979) and Myotis bats in northern New Mexico (O’Farrell and Studier 1973); late summer mating readiness in male lasiurine bats is known from several locations across the continent (Cryan et al. 2012).
These findings suggested three nonmutually exclusive hypotheses and their predictions to explain the presence of adult males at maternity roosts of big brown bats at our study area in Colorado: 1) the sex-specific differences in energetic strategies hypothesis; 2) the natal philopatry of inexperienced young males hypothesis; and 3) the male reproductive advantage hypothesis.
If the sex-specific energetic strategies hypothesis holds for male attendance patterns at our study area, we predict that: a) adult males emerging at maternity roosts would be captured in low proportions compared to other sex and age groups; and b) most males tagged as juveniles at maternity roosts would not return, and adult male annual return rates would be lower than annual returns of females. We also predicted that c) adult males that use maternity roosts at lower elevations would use these roosts less frequently on a daily basis than adult females, and d) less frequently in warmest months with peak maternity activity. We had four predictions if the natal philopatry of inexperienced young males hypothesis was the case: a) 1-year-old males at our study area would return mostly to the roost where they were born; b) natal philopatry would wane with greater maturity such that males older than 1 year would return at lower annual rates; c) higher daily attendance rates of 1-year-old males using maternity roosts would occur primarily early in the year compared with females; d) daily attendance rates of 1-year-old bats would be higher than in older males. The male reproductive advantage hypothesis predicts: a) big brown bats in late summer prior to autumn migration will show anatomical evidence for mating or mating readiness; and b) that daily attendance rates of adult males at maternity roosts would be higher late in the summer than at other times.
Materials and Methods
Study area and study population
Sampling took place during active seasons of 2001–2005 in and near Fort Collins, Larimer County, Colorado, United States, at the northern end of the Colorado Front Range Urban Corridor. Detailed study area descriptions are available elsewhere (e.g., Neubaum et al. 2006, 2007b; O’Shea et al. 2011a). Areas to the east of Fort Collins are sparsely developed rural prairies devoted primarily to farming and ranching. The area to the west of Fort Collins is mountainous, with coniferous forest and numerous rock outcrops and cliffs. Elevation is 1,525 m and climate is temperate and semiarid. Mean ± SD daily active season temperatures (2002–2005) at Fort Collins (Colorado Climate Center 2020), by 30-day Julian date (JD) periods used in this study and corresponding stage of female reproductive phenology were: Period 1 (JD 125–154, 5 May to 3 June) 15.0 ± 3.9°C, pregnancy and maternity colony formation; Period 2 (JD 155–184, 4 June to 3 July) 18.9 ± 4.1°C, lactation; Period 3 (JD 185 –214, 4 July to 2 August) 22.3 ± 3.1°C, lactation and volancy of young; Period 4 (JD 215–244, 3 August to 1 September) 19.8 ± 2.6°C, postlactation and disbanding of colonies; Period 5 (JD 245–274, 2 September to 1 October) 16.4 ± 2.3°C, dispersal to hibernacula.
Big brown bats are by far the dominant species of bat in and around the study area, where they appear to roost exclusively in buildings; colonies in many buildings are routinely excluded by occupants every few years (O’Shea et al. 2011a). Buildings used as maternity roosts by bats have higher and larger exit points, warmer internal roost temperatures, and shorter distances to similar buildings, than randomly selected buildings (Neubaum et al. 2007b). Roosts are not limited to attics, but often are in confined spaces between walls or crawl spaces inaccessible for observation by researchers (O’Shea et al. 2011a). Females switch roosts among nearby buildings during summer based on daily variations in maximum temperatures (Ellison et al. 2007a). Maternity colony sizes typically range from 20 to 50 bats, with a maximum of 219 counted at one roost during emergence (O’Shea et al. 2011a). Both eastern and western mitochondrial DNA genotypes of the species (Turmelle et al. 2011) are found within the same roosts in Fort Collins (Neubaum et al. 2007a). Detailed demographic analyses of female big brown bat population dynamics in Fort Collins indicate a population growing during the period of study (O’Shea et al. 2010, 2011b). The bulk of the population makes regional migrations to hibernacula in inconspicuous rock crevices at higher elevations during late summer and early autumn, but a few bats may remain in buildings over winter (Neubaum et al. 2006; O’Shea et al. 2011a). Mist-netting records at foraging and drinking areas, as well as public health records of bats submitted for rabies diagnoses, show a continued presence but lower proportion of adult males at the study area during summer compared with a greater proportion at higher elevations (Neubaum et al. 2006; O’Shea et al. 2011a).
Bat capture, handling, and tagging
We captured bats as they emerged from maternity roosts at dusk, using mist nets, harp traps, funnel traps, and handheld nets. Bats were transported to the laboratory, tagged, sampled for disease studies (e.g., Dominguez et al. 2007; George et al. 2011; O’Shea et al. 2014), and released at the roost on the same night. We assessed age (volant juvenile or adult) based on epiphyseal closure criteria in Anthony (1988) and sex and reproduction status following Racey (1988). Tagged bats were marked individually by subdermal insertion of uniquely numbered passive integrated transponders (PIT tags; AVID, Norco, California), as described by Wimsatt et al. (2005). We deployed hoop-style PIT tag readers over the openings of roosts to monitor and record presumed entrance and exit dates and times of tagged bats automatically without need for capture (O’Shea et al. 2004; Wimsatt et al. 2005; Ellison et al. 2007b). PIT tag reader records were maintained on an internal Structured Query Language (SQL) database and later transformed to a SAS (SAS Institute Inc. 2016) database for analysis. We categorized bats by known age (1–4 years old) based on the date of PIT tagging as juveniles, or as ≥ 1, ≥ 2, ≥ 3, ≥ 4, or ≥ 5 years old based on the year after first tagging as adults. We also captured male bats in mist nets set over water in Fort Collins and in the nearby mountains to assess reproductive condition (see below). All bat capture, tagging, and sampling, procedures were approved by the Institutional Animal Care and Use Committees of the U.S. Geological Survey and Colorado State University and conformed to past and present standards of the American Society of Mammalogists (Animal Care and Use Committee 1998; Sikes et al. 2016).
Numbers and proportions of adult males captured at maternity roosts
We provide the pooled cumulative numbers of big brown bats by sex and age class captured during evening emergence from all maternity roosts sampled over the entire study area over all years of study to evaluate the prediction of a lower proportional abundance of adult males under the sex-specific energetic strategies hypothesis. This cumulative measure includes unmarked bats and is comparable to typical results based on mist-netting captures of bats in many past studies that bear on sexual segregation (reviewed in McGuire and Boyle 2013). Variation among roosts in levels of sampling effort, seasonal timing of effort, colony size, access to emergence points, roost persistence, and our ability to determine precise roosting locations within buildings and their associated covariates was too great to allow more sophisticated modeling of differences in proportions or numbers of each sex among roosts.
Annual return rates and natal philopatry
We computed annual return rates to investigate roost fidelity of adult males compared to females under the predictions of the sex-specific energetic strategies hypothesis that most males tagged as juveniles at maternity roosts will not return, and that adult male annual return rates will be lower than annual returns of females. We assumed the roosts where volant juvenile bats were first tagged at emergence were their natal roosts and used annual return rates to evaluate the prediction under the natal philopatry of inexperienced young males hypothesis that 1-year-old males at our study area would return mostly to the roost where they were born. We were unable to use standard Cormack–Jolly–Seber (CJS) models to estimate annual male capture and survival probabilities for comparison with previously published estimates for females (O’Shea et al. 2010, 2011b) because males most likely violated the CJS model assumptions (e.g., permanent versus temporary emigration and independence of fates—Williams et al. 2002; O’Shea et al. 2004). Instead we calculated simple annual return rates of males to maternity roosts based on PIT reader detections at any time during the active season each year. We also calculated annual presence–absence rates of females for comparison with males. We calculated annual return rates for bats of known ages ≤ 4 years old at five roosts that were monitored during every year of the study (2001–2005), as well as returns of bats of unknown age but ≥ 1 year old first tagged as adults. We calculated annual return rates of the latter males at 12 roosts monitored with tag readers for 1–4 years following the year of tagging.
Daily and seasonal attendance rates of returning bats
We used daily and seasonal attendance rates to test predictions of each hypothesis. For the sex-specific energetic strategies hypothesis we tested the predictions that adult males would attend maternity roosts less frequently than adult females on a daily basis, and less frequently in warmest months with peak maternity activity. Under the natal philopatry of inexperienced young males hypothesis we tested the predictions that daily attendance rates of 1-year-old males would occur primarily early in the active season, and that their daily attendance rates would be higher than those of older males. For the male reproductive advantage hypothesis, we tested the prediction that daily attendance rates of adult males at maternity roosts would be higher late in the summer than at other times.
We determined individual daily attendance rates for those adult males and females known to use a subset of nine maternity roosts monitored for the full duration of each active season for 1–4 years after the year of tagging. Daily attendance was calculated as the number of days with detections of individual males and females by readers each active season as a proportion of the number of days readers were known to be operable that season. Bats often switched roosts among nearby buildings (Ellison et al. 2007a) that were not all monitored, could use minor openings that lacked PIT tag readers at some roosts, and readers sometimes malfunctioned temporarily. We estimated mean daily attendance rates in two ways: 1) over the full active season for each year, and 2) for all 30-day periods from 5 May to 7 October over all years combined. We calculated daily attendance rates for male and female bats that were 1 year old or ≥ 2 years old (based on years after the year of first tagging as volant juveniles or as unknown-age adults).
Evidence for late summer mating
We examined both females and males for anatomic evidence of mating or mating readiness in late summer prior to autumn migration, as predicted by the male reproductive advantage hypothesis. We sampled female bats that were captured and released in late summer 2004 using lavage and histological examination. We flushed the vaginas of live bats with 0.2 ml of phosphate-buffered saline solution using a plastic pipette tip and collected the wash in a 1.5-ml microcentrifuge tube. The flushed fluid was agitated and examined for sperm by an experienced andrologist the following day using phase contrast microscopy. We also prepared sections of uteri from a smaller sample of adult females captured in Fort Collins in August 2005. Bats were euthanized and uteri excised and fixed in 10% buffered formalin. Standard hematoxylin and eosin-stained sections of uteri were examined (40× objective) with a Nikon Eclipse E800 light microscope to detect sperm in the lumen or in contact with the endometrium. Reference sperm was obtained from the dissection of testes of a 2-year-old male. Assessment of reproductive condition of males was limited to examination for distended scrotal sacs in live males after 1 August each year. As with other temperate-zone vespertilionids (Miller 1939), spermiogenesis in adult big brown bats studied in Maryland typically begins in early summer and ends in early autumn when testes retrogress, with viable sperm stored in the enlarged caudae epididymes by September for subsequent mating (Christian 1956). We assumed that during late summer distended scrotal sacs reflected reproductively active males (with either active testes or engorged caudae epididymes) and nondistended sacs indicated nonreproductive males.
Statistical analyses
We present simple summary statistics for annual and daily roost attendance rates. We estimated attendance rate means and their 95% confidence intervals (CIs) with an intercept-only logistic regression model using SAS Proc Genmod (SAS Institute Inc. 2016), where the number of days the reader was in operation was a weighting variable (Agresti 1990). For pairwise comparisons (e.g., annual return rates for males versus females) we did not compute a test statistic and corresponding P-value. Rather, we examined whether the two CIs overlapped (following Johnson 1999; Anderson et al. 2001); in cases where they did not overlap, estimates differed with a P-value much less than 0.05 (Payton et al. 2003).
Results
Sex-specific energetic strategies hypothesis
Sex and age composition of bats captured at emergence supported the prediction of a low proportional abundance of adult males at maternity roosts compared to other sex and age groups. We captured 8,192 bats (including repeat captures and unmarked bats) during 285 evening emergences from 46 roosts in 2001–2005. We recorded 255 captures of adult males at 23 of the 46 roosts. Adult males comprised 3.1% (CI = 2.8–3.5%) of the 8,192 captures of adult and juvenile bats of both sexes, and 4.4% of 5,836 captures (CI = 3.9–4.9%) of adults.
Most males tagged as juveniles at maternity roosts were not detected returning to these roosts, as predicted under this hypothesis. We tagged 299 juvenile males at the five roosts monitored every year and 207 (69.2%; CI = 63.6–74.4%) were never detected in subsequent years. The overall mean annual return rate for the 92 returning known-age males 1–4 years old (weighted by the number of years of resampling) at the five roosts was 0.144 (CI = 0.121–0.170) compared to 0.489 (CI = 0.456–0.552) for 308 known-age adult females. Within every age group, females had higher annual re-encounter rates than males, with nonoverlapping CIs between sexes (Fig. 1). Mean return rates of same-aged females and males varied from 2.2 times higher in 1-year-old females to 17.4 times higher in 3-year-old females (Fig. 1).
Annual return rates (and 95% CI) during 2002–2005 for known-age male and female big brown bats (Eptesicus fuscus) first tagged as volant juveniles at five maternity roosts, 2001–2004, Fort Collins, Colorado, United States. Sample sizes of original tagged cohorts are in parentheses above each bar. Height of each bar represents the proportion of that sample for each age and sex class.
The prediction that males would show lower daily attendance rates than females under the sex-specific energetic strategies hypothesis also was supported (Fig. 2). We compiled 180 active season records of daily attendance of 140 adult males (some individuals encountered in multiple years) at nine maternity roosts during 2002–2005 (113 individual males at 1 year old and 42 individuals at ≥ 2 years old, including 15 males recorded in both age groups). We also compiled 1,625 active season records of daily attendance at the same roosts for all years combined of 788 adult females (218 returning at 1 year old and 570 at ≥ 2 years old, including 100 individuals recorded in both age groups). Daily attendance at roosts for 1-year-old males was lower than for males ≥ 2 years old, and both age groups of males attended at lower rates than females of either age group in every year (Fig. 2). One-year-old females also attended at lower daily rates than older females (Fig. 2). Daily attendance rates varied by year across each sex and age group, but the relative ranking of each group was the same in every year (Fig. 2). For all years combined, 1-year-old males attended at a daily rate of 0.105 (CI = 0.100–0.110), whereas males ≥ 2 years old attended at about twice that rate (0.204; CI = 0.196–0.214). Females ≥ 2 years old attended at a higher daily rate (0.446; CI = 0.443–0.448) than 1-year-old females (0.338; CI = 0.333–0.344) for all years combined, and both age groups of females attended at higher daily rates than males.
Daily attendance rates (and 95% CI) each year for male and female big brown bats (Eptesicus fuscus) known to return to nine maternity roosts at 1 year of age and at known minimum ages ≥ 2 years old, 2002–2005, Fort Collins, Colorado, United States.
Daily attendance rates varied seasonally by sex and age based on 30-day time periods for all years combined (Fig. 3) and the daily attendance rates supported the prediction that adult males would use maternity roosts less frequently in warmest months with peak maternity activity under the sex-specific energetic strategies hypothesis. Males attended at much lower daily rates than females for the first three time periods (corresponding to early May through early August), and 1-year-old males attended at lower rates than males ≥ 2 years old during the first four time periods (through late August; Fig. 3). Similarly, 1-year-old females attended at lower rates than females ≥ 2 years old during the first three time periods (Fig. 3). During the fourth period (postlactation) attendance of older males increased compared to the prior three periods, and remained higher than rates of 1-year-old males; daily attendance rates of females during the fourth period dropped in comparison to the prior three periods and did not differ between 1-year-olds and older females. During the fifth period daily attendance of all sex and age classes was at its lowest, although attendance rates of 1-year-old males overlapped with Period 3 (Fig. 3).
Daily attendance rates (and 95% CI) by 30-day periods (all years combined) of male and female big brown bats (Eptesicus fuscus) known to return to nine maternity roosts at 1 year of age and at known minimum ages ≥ 2 years old, Fort Collins, Colorado, United States, 2002–2005. Approximate phenology of female reproduction is: Period 1 (5 May to 3 June) pregnancy and maternity colony formation; Period 2 (4 June to 3 July) lactation; Period 3 (4 July to 2 August) lactation and volancy of young; Period 4 (3 August to 1 September) postlactation and disbanding of colonies; Period 5 (2 September to 1 October) dispersal to hibernacula.
Natal philopatry of inexperienced young males hypothesis
The prediction that 1-year-old males at our study area would return mostly to the roost where they were born was supported. We tagged 475 juvenile males at the 12 maternity roosts monitored for natal philopatry in 1–4 subsequent years. Most (342; 72.0%, CI = 67.7–76.0%) did not return. However, nearly all (132 of 133; 99.2%, CI = 95.9–100.0%) males tagged as juveniles that were detected as adults at these 12 maternity roosts were found only at the roosts where they were originally tagged, including three that also used alternate roosts favored by the natal colony during roost-switching events. The one exception was a male detected as a 1-year-old at the natal roost and then again as a 3-year-old on one date at a roost not known to be used by its natal colony. We did not monitor 14 additional roosts where males were tagged as juveniles, but none of the 131 males tagged at these unmonitored roosts were detected at any of the roosts that were monitored during 1–4 active seasons after tagging. This observation is consistent with likely low dispersal to non-natal roosts. In addition to known-age males first tagged as juveniles, we also tagged 37 adult males of unknown ages at monitored sites during the first year of study (2001) and searched for their presence 4 years later (2005). Six of these older males were detected at ≥ 5 years old, all at maternity roosts where first tagged in 2001, showing continued roost fidelity.
The prediction that males ≥ 2 years old would return to natal roosts at much lower annual rates than 1-year-old males was supported (Fig. 1). Nearly all male bats (91 of 92; 98.9%, CI = 93.2–99.9%) that we documented returning to the five maternity roosts monitored every year were first detected as 1-year-olds, and 20 of the 21 individuals that returned at 2–4 years of age had been previously detected at these roosts as 1-year-olds. Markedly lower numbers and proportions of males were detected as 2- to 4-year-olds at these five roosts: 1-year-old males were re-encountered at annual rates 5–16 times higher than 2- to 4-year-old males (Fig. 1). Overall annual return rates of 52 adult males of unknown age tagged from 2001 to 2004 averaged 0.305 (CI = 0.239–0.379) at the five roosts during 2002–2005, with CIs showing overlap among years (Table 1). This rate was nearly identical to that of known-age 1-year-old males (0.304; CI = 0.254–0.359; Fig. 1), suggesting the bulk of these returns also were 1 year old or younger at the time of tagging (epiphyseal closures are more difficult to diagnose late in the juvenile year). The overall annual return rate of 570 unknown-age adult females over the same period at the same roosts was more than twice the male rate at 0.630 (CI = 0.607–0.652) and was about twice the male rate each year (Table 1).
Mean and 95% CI annual return rates by year of adult male and female big brown bats (Eptesicus fuscus) tagged at unknown ages (but ≥ 1 year old) each year during 2002–2004 at five maternity roosts monitored with passive integrated transponder (PIT) tag readers each summer, 2001–2005, Fort Collins, Colorado, United States.
| Year | Adult females | Adult males | ||
|---|---|---|---|---|
| (95% CI) | N | (95% CI) | N | |
| 2002 | 0.840 (0.791–0.879) | 268 | 0.417 (0.241–0.617) | 24 |
| 2003 | 0.648 (0.601–0.692) | 152 | 0.289 (0.168–0.451) | 14 |
| 2004 | 0.620 (0.578–0.660) | 109 | 0.320 (0.206–0.460) | 12 |
| 2005 | 0.528 (0.487–0.569) | 41 | 0.250 (0.151–0.384) | 2 |
| Overall | 0.630 (0.607–0.652) | 570 | 0.305 (0.239–0.379) | 52 |
| Year | Adult females | Adult males | ||
|---|---|---|---|---|
| (95% CI) | N | (95% CI) | N | |
| 2002 | 0.840 (0.791–0.879) | 268 | 0.417 (0.241–0.617) | 24 |
| 2003 | 0.648 (0.601–0.692) | 152 | 0.289 (0.168–0.451) | 14 |
| 2004 | 0.620 (0.578–0.660) | 109 | 0.320 (0.206–0.460) | 12 |
| 2005 | 0.528 (0.487–0.569) | 41 | 0.250 (0.151–0.384) | 2 |
| Overall | 0.630 (0.607–0.652) | 570 | 0.305 (0.239–0.379) | 52 |
Mean and 95% CI annual return rates by year of adult male and female big brown bats (Eptesicus fuscus) tagged at unknown ages (but ≥ 1 year old) each year during 2002–2004 at five maternity roosts monitored with passive integrated transponder (PIT) tag readers each summer, 2001–2005, Fort Collins, Colorado, United States.
| Year | Adult females | Adult males | ||
|---|---|---|---|---|
| (95% CI) | N | (95% CI) | N | |
| 2002 | 0.840 (0.791–0.879) | 268 | 0.417 (0.241–0.617) | 24 |
| 2003 | 0.648 (0.601–0.692) | 152 | 0.289 (0.168–0.451) | 14 |
| 2004 | 0.620 (0.578–0.660) | 109 | 0.320 (0.206–0.460) | 12 |
| 2005 | 0.528 (0.487–0.569) | 41 | 0.250 (0.151–0.384) | 2 |
| Overall | 0.630 (0.607–0.652) | 570 | 0.305 (0.239–0.379) | 52 |
| Year | Adult females | Adult males | ||
|---|---|---|---|---|
| (95% CI) | N | (95% CI) | N | |
| 2002 | 0.840 (0.791–0.879) | 268 | 0.417 (0.241–0.617) | 24 |
| 2003 | 0.648 (0.601–0.692) | 152 | 0.289 (0.168–0.451) | 14 |
| 2004 | 0.620 (0.578–0.660) | 109 | 0.320 (0.206–0.460) | 12 |
| 2005 | 0.528 (0.487–0.569) | 41 | 0.250 (0.151–0.384) | 2 |
| Overall | 0.630 (0.607–0.652) | 570 | 0.305 (0.239–0.379) | 52 |
Daily attendance rates provided mixed support for two predictions of the natal philopatry of inexperienced young males hypothesis. The prediction that daily attendance rates would show 1-year-old males using maternity roosts primarily early in the year compared with females was supported. On a seasonal basis, 1-year-old males used maternity roosts at higher daily rates during the first two periods of the season than during Periods 3 and 5, but CIs overlapped slightly with Period 4 (Fig. 3). However, the prediction that daily attendance rates of 1-year-old males would be higher than those of older males was not met: 1-year-old males attended less frequently than the other three age groups in all periods except the fifth (Fig. 3). As noted under the sex-specific energetic strategies hypothesis, daily attendance at roosts for 1-year-old males was lower than for males ≥ 2 years old, but both male age groups attended at lower rates than females of each age group in every year (Fig. 2).
Male reproductive advantage hypothesis
We found no anatomical evidence in support of this hypothesis in females. No sperm was observed in microscopic examination of vaginal lavage fluids from 65 females captured between 12 August and 16 September 2004. None of the histological sections of uteri from 15 females (10 were > 2 years old) taken 5–17 August 2005 was positive for sperm. We assessed reproductive condition of 115 adult males captured in Fort Collins between 1 August and 20 September 2001–2005: 46 (40.0%; CI = 31.1–49.6%) had distended scrota and 69 (60.0%; CI = 50.4–68.9%) did not. Fifty adult males captured at higher elevations west of Fort Collins also were examined in August and September: 22 (44.0%; CI = 30.3 – 58.7%) had distended scrota and 28 (56.0%; CI = 41.4 – 69.7%) did not. We further assessed a subset of 23 known-age adult males captured at maternity roosts from 1 August to 14 September 2002–2005 for engorged scrotal sacs. None of eight 1-year-old, three of six 2-year-olds, and seven of nine males known to be ≥ 3 years of age had distended scrota.
We predicted that daily attendance rates of adult males using maternity roosts at our study area would be higher late in the summer than at other times. During the fourth period, males ≥ 2 years old attended at a greater rate than earlier in summer and greater than 1-year-old males (Fig. 3), seemingly consistent with this prediction. However, both age groups of females declined and converged in daily attendance rates during this fourth period, attending at daily rates with CI’s overlapping or similar to rates of older males but higher than 1-year-old males (Fig. 3). By the fifth and final period (September) older male daily attendance had declined to a rate similar to 1-year-old males, and daily attendance of females of both age groups was at their lowest rates of the year and slightly lower than daily rates of males (Fig. 3), suggesting fewer mating opportunities at maternity roosts in late summer–early autumn than earlier in the year.
Discussion
We found support for most predictions of the first two of three nonmutually exclusive hypotheses for patterns of male attendance at maternity roosts of big brown bats. Adult males used maternity roosts in our study area in patterns that differed from those of adult females in ways consistent with the sex-specific energetic strategies hypothesis and with prior observations of sexual segregation by elevation (Neubaum et al. 2006; O’Shea et al. 2011a). Adult males were captured during emergence at maternity roosts in much lower numbers than adult females, most males did not return to maternity roosts after their juvenile year, and tagged males that used maternity roosts did so at lower daily attendance rates than tagged females for most of the active season. Consistent with the natal philopatry hypothesis, most males detected at maternity roosts as adults were born at those roosts in the previous year. Older males also used natal roosts but at much lower annual return rates than 1-year-old males or adult females, and in differing daily and seasonal patterns of attendance. In contrast, the two predictions of the male reproductive advantage hypothesis were not met.
Although a small proportion of adult males did not segregate by elevation but used maternity roosts at the study area, patterns of usage were consistent with the sex-specific energetic strategies hypothesis. Sex ratios of adults captured at emergence were decidedly skewed in favor of females, whereas those of volant juveniles at the study area are even (678 juvenile males and 684 juvenile females—O’Shea et al. 2011a). The change in sex ratio from juveniles to adults suggests that many males do not return to the lower elevation area from their winter hibernacula for extended periods. Lower annual return rates of adult males of both known and unknown ages to maternity roosts in comparison with females also are concordant with other findings of sexual segregation by elevation in this region (Neubaum et al. 2006; O’Shea et al. 2011a). However, it is noteworthy that the proportion of adult males taken in captures during emergence at maternity roosts (3.1%, 255 of 8,192 captures; CI = 2.8–3.5%) was much lower than the proportion taken in mist nets over water later in the night in the urban area (23.2%, 80 of 346 captures; CI = 19.0–27.8%—O’Shea et al. 2011a). The lower proportions of males taken during emergence at maternity roosts than over water suggest that some males continue to use the lower elevation habitat for foraging but may roost at locations other than maternity roosts, a speculation supported by observations of solitary roosting male big brown bats at our study area (O’Shea et al. 2011a; Castle et al. 2015) and elsewhere (Mills et al 1975; Perry and Thill 2008).
We did not observe the behavior of bats within roosts. However, roosting solitarily or away from aggregating females within the same structure is consistent with the sex-specific energetic strategies hypothesis: such males would be more likely to achieve lower body temperatures for energetic savings through daily torpor compared to females and young that cluster in warm parts of maternity roosts. Although we were unable to test this as a prediction of the sex-specific energetic strategies hypothesis in our study, observations of male big brown bats at maternity roosts in other areas indicate this may be the case. In Alberta, males make up a smaller proportion of adult big brown bats using maternity roosts compared to females (Hamilton and Barclay 1994; Grinevitch et al. 1995), choose sites within maternity roosts that differ in microclimates from areas chosen by reproductive females (Hamilton and Barclay 1994), and use these sites to achieve deeper torpor than reproductive females (Hamilton and Barclay 1994; Grinevitch et al. 1995). Adult males at maternity roosts of big brown bats in Kentucky also roost solitarily away from the main groups of females (Davis et al. 1968). In addition to energetic benefits, roosting away from females could dampen the likelihood of disease exposure and transmission among bats by reducing social contacts (Plowright et al. 2015; Webber et al. 2016), as supported by the lower seroprevalence of rabies virus-neutralizing antibodies in adult males in our study area compared to females (Bowen et al. 2013; O’Shea et al. 2014).
Differences between sex and age groups in daily attendance patterns at maternity roosts met predictions of the sex-specific energetic strategies hypothesis. Daily attendance rates of males at maternity roosts always were lower than those of females throughout the active season, and consistent with the above suggestion of the likely use of alternate sites elsewhere that may provide energetic advantages to males. The higher daily attendance rates of females at maternity roosts declined during the last two 30-day periods of the active season (through September), consistent with their likely use of alternative roosting sites where greater energetic savings can be accrued following the demands of pregnancy and lactation. Use of different roosts with differing microclimates at separate phases of the reproductive cycle based on disparate energetic strategies of the sexes is known in other temperate-zone bats. In Daubenton’s bat, for example, females switch to roosts that allow deeper torpor after the maternity period, whereas some males switch to warmer roosts to increase spermatogenesis and access to mates (Lučan and Hanák 2011). Adult male big brown bats in Alberta are less likely than females to leave the roosts to forage on cool nights (Grinevitch et al. 1995), travel farther to foraging areas that were presumed to be less favorable (Wilkinson and Barclay 1997), and choose alternate roost sites at other locations on some days (Hamilton and Barclay 1994). Similar behavior would explain the lower daily attendance rates of adult males compared to females at our study area. In addition, intermittent movements of males to higher elevations in the nearby mountains also could explain lower daily attendance rates. In our study area the greater proportion of adult males netted over water at mountain elevations above 2,000 m included sites as close as 10 km to Fort Collins, within individual nightly foraging ranges of big brown bats in this region (O’Shea et al. 2011a; Castle et al. 2015) and in some other areas (Hamilton and Barclay 1994; Arbuthnott and Brigham 2007).
Annual return rates of known-age individuals show that most males found at maternity roosts at our study area were 1-year-old males returning to their natal roosts, as predicted by the natal philopatry of inexperienced young males hypothesis. Unpublished observations cited by Hamilton and Barclay (1994) suggest that most males at maternity roosts in their Alberta study area also were 1-year-old males. However, Grinevitch et al. (1995) suggested that some males at the Alberta maternity roosts were older than 1 year. We found that all returning known-aged tagged adult males ≥ 2 years old also were detected only at the roosts where first captured as juveniles, and that adult males of unknown ages but ≥ 5 years old showed philopatry to the same roosts where originally tagged as adults 4 years earlier. Although some adult males continued to use maternity roosts at older ages, our observations support the prediction that use of maternity roosts by adult males would wane with increasing age. On an annual basis, males ≥ 2 years old at the five maternity roosts monitored over five consecutive years were detected at a lower frequency in years subsequent to their first adult summer. They also were detected much less frequently than adult females of the same ages.
Daily attendance rates of 1-year-old males at maternity roosts were lower than those of adult females. However, they also were lower than daily attendance rates of males ≥ 2 years old. This was contrary to our prediction that daily attendance in males would be highest in 1-year-olds and wane at older ages. But, the monthly distribution of daily attendance rates by sex and age categories followed the prediction that daily attendance rates of 1-year-old males would be higher earlier in the active season than later, but with a slight rise in the penultimate (fourth) period (Fig. 3). Except for the final (fifth) period, daily attendance of 1-year-old males was always lower than daily attendance of older males and all females. The tendency for a reduction in daily attendance rates of 1-year-old males after the first two periods of the active season could be a result of young males being actively excluded from maternity roosts (Weller et al. 2009) at the peak of lactation and early volancy of young. Alternatively, as 1-year-old males mature they may tend to roost away from the colony at the time of highest activity without active exclusion. We posit that perhaps some 1-year-old males have not fully matured in their roosting and foraging behavior in their year of birth (particularly those born later in the summer) and continue to associate with their mothers the next summer after hibernation until the next cohort of young is produced at mid-summer. We have no data that bear directly on these possibilities. Variation in the long-term postweaning development of landscape-level foraging and roosting patterns of bats in areas of sexual segregation is not well known and is deserving of future study.
Although previous studies have suggested that adult males at maternity roosts of big brown bats may gain some unspecified reproductive advantage (Davis et al. 1968; Vonhof et al. 2006), predictions of the male reproductive advantage hypothesis that involved late summer mating were not supported by our data. We found no microscopic evidence for late summer to early autumn mating in samples of females examined for the presence of sperm. Annual return rates of known-age individuals showed that most males found at maternity roosts were 1-year-old males returning to their natal roosts; most 1-year-old males at the study area did not show distended scrotal sacs in late summer to early autumn and about 60% of unknown-age adult males were in similar condition, indicating likely absence of mating capability in many males at that time of year. The decline in daily attendance of females at maternity roosts during the last two 30-day periods of the active season (through September) suggests that fewer mating opportunities for either sex were available at maternity roosts in late summer. Males and females of both the 1-year-old and ≥ 2-year-old age groups were detected at maternity roosts at their lowest daily rates during the final period, when mating readiness would seem more likely (Wimsatt 1944; Christian 1956). The increase in daily attendance of males ≥ 2 years old in the penultimate period seems anomalous and perhaps could indicate some degree of mate seeking, but it also may be a response to less crowded conditions and cooler microclimates in the roost, providing an energetic benefit not available in preceding months. Alternatively, ≥ 2-year-old males may have sought warmer microclimates within maternity roosts during this period to enhance spermatogenesis (Lučan and Hanák 2011).
Experimental evidence from studies of big brown bats taken at Pennsylvania caves documented that most mating takes place somewhat later in autumn, either prior to or at the beginning of hibernation (Wimsatt 1944). This has been verified in the field in big brown bats in Kansas (Phillips 1966), where peak size of testes and caudae epididymis occurs in August. This timing of mating also may be true in our study population. Anecdotal evidence has shown that copulation of big brown bats can occur later in winter (Mumford 1958), but winter mating is thought to be less frequent than autumn or early winter copulations (Wimsatt 1944). Little is known about mating behavior or mate selection in wild big brown bats. Individuals have unique qualities to their echolocation pulses (Masters et al. 1995; Burnett et al. 2001) and in the laboratory, captive males show distinct differences in echolocation pulses during the mating season that correlate with number of copulations (Grilliot et al. 2014). Perhaps reproductively active male big brown bats (≥ 2 years old) at maternity roosts gain some future mating advantage through individual recognition and familiarity to females that join them for copulation at hibernacula.
We conclude that attendance of male big brown bats at maternity roosts at our study area does not contradict the sex-specific energetic strategies hypothesis thought to be a driver of sexual segregation by elevation. Unlike females, most of the males born at these maternity roosts do not return in subsequent years, but probably contribute to the higher proportion of adult males reported at higher elevations along the Colorado Front Range. Those adult males that return to lower elevation sites are mostly 1-year-olds that come back to the natal roost, with a few of these returning at ≥ 2 years old in subsequent years as well. These males have an alternative strategy for energy savings other than movement to higher elevations. As demonstrated by studies in Alberta (Hamilton and Barclay 1994; Grinevitch et al. 1995), adult males that are found at maternity roosts in our area probably favor microclimates within roosts that provide more energetic savings than the microclimates used by reproductive females in the same structures. These males also have lower daily attendance rates that indicate they roost elsewhere more often and most likely under different microclimates than reproductive females. Nonreproductive females of many species of temperate-zone insectivorous bats segregate by elevation similar to males (McGuire and Boyle 2013). About a third of the 1-year-old females that use maternity roosts at our area are nonreproductive (O’Shea et al. 2010). One-year-old females in the present study also occupied roosts at lower daily rates than older females and thus may have an energetic strategy similar to males. Adult females at our study area had high CJS capture and survival probabilities and high breeding probabilities after age 1 year that resulted in the positive population growth rates documented during the period of study (O’Shea et al. 2011b). Positive population growth rates may indicate that carrying capacity relative to foraging has not been reached, allowing some adult males and nonreproductive females to continue to use the area for foraging and roosting rather than move to higher elevations. We did not detect evidence for late summer breeding at maternity roosts: male reproductive condition was similar to males at higher elevations and daily attendance of both sexes at maternity roosts was low at this time of year. This suggests that most mating may take place at higher elevation autumn roosts and hibernacula common to individuals that use both elevational strategies in summer, perhaps providing a wider range of choices for mate selection.
Acknowledgments
We thank P. Cryan and two anonymous reviewers for comments on earlier drafts of the manuscript. This project was part of a larger study supported by the U.S. Geological Survey and a grant from the National Science Foundation Ecology of Infectious Diseases Program to Colorado State University (EF-0094959). Constructive input to the study was provided by D. R. Anderson, L. E. Ellison, R. Reich, C. E. Rupprecht, and J. A. Wimsatt. We thank S. Almon, J. Ammon, L. Ansell, T. Barnes, J. Boland, L. Bonewell, M. Carson, K. Castle, S. Cooper, T. Dawes, L. Ellison, D. Emptage, L. Galvin, D. Grossblat, M. Hayes, B. Iannone, E. Kennedy, R. Kerscher, J. LaPlante, H. Lookingbill, G. Nance, S. Neils, C. Newby, V. Price, C. Reynolds, S. Smith, L. Taraba, J. Tharp, T. Torcoletti, and M. Vrabely, for help with data entry and sampling bats in the field and laboratory. Use of trade, product, or firm, names is for descriptive purposes only and does not imply endorsement by the United States government. Bats were captured under authority of scientific collecting licenses issued by the Colorado Division of Wildlife.
