Maximizing mating opportunities: higher autumn swarming activity in male versus female Myotis bats

Abstract

Many animal taxa exhibit intersexual differences in sociality and resource selection that can result in variation in energy allocation budgets. Asymmetry of reproductive energetics between sexes can lead to intersexual variation in behavior to maximize lifetime reproductive success. Temperate bats show marked intersexual differences during summer when sexual segregation occurs. Although some intersexual differences have been shown in activities during the autumn mating period, they have not been examined in the context of fitness consequences for each sex. We studied autumn swarming activity of little brown and northern myotis bats (Myotis lucifugus and M. septentrionalis) in Nova Scotia, Canada, to test predictions of the hypothesis that intersexual variation in swarming behaviors occurs to maximize fitness. We conducted capture–mark–recapture surveys at swarming sites to characterize the nature and extent of intersexual variation in swarming activities. Relative to females, males occurred in disproportionally large numbers; had longer swarming seasons overlapping the female swarming season; and accounted for a disproportionately large number of recaptures at swarming sites suggesting that they had returned more frequently. No movements among swarming sites, which ranged in pairwise distance from 27.9 to 98.9 km, were detected for either species. Activity at swarming sites was highest early in the season for both species. As predicted, males engaged more frequently in swarming activities than females which likely reflect males maximizing opportunities for mating. Although their activities overlap during this period, differences suggest sex-specific activity budgets and it is likely that for each sex, individuals reconcile energetic constraints differently to maximize fitness.

Chiroptera, hibernacula, life history strategy, movements, Myotis, population dynamics, sex-specific, swarming

For many animal taxa, intersexual differences can occur in degree of sociality and resource selection strategies and these differences can lead to significant differences in activity budgets. For example, gray seals (Halichoerus grypus) have been shown to have striking differences in home range locations before and after the breeding season (Breed et al. 2006). Intersexual differences are largely thought to reflect the asymmetry of reproductive energetics and parental investment (i.e., reproductive cost) that lead to different maximization strategies for individual expected lifetime reproductive success (Trivers 1972). In mammals, sexual segregation in spatial organization or resource use in the nonbreeding season is common for species where the sexes live in separate groups or as solitary individuals (Conradt 1998; Ruckstuhl and Neuhaus 2000; Main 2008). During the breeding season, the degree of sexual segregation decreases to varying extents across taxa. However, many of these same factors continue to act differentially on the sexes in promoting individual fitness while still facilitating courtship and mating behaviors. Regardless of where individuals are in the seasonal and reproductive cycles, they must continue to strategically allocate time to specific activities to optimize the balance between fitness costs and benefits.

Male mammals maximize fitness by securing many mating opportunities (Bateman 1948; Andersson 1994), potentially at the expense of other activities such as foraging (Miquelle 1990; Alberts et al. 1996). This strategy may be possible due to physiological mechanisms such as metabolic compensation where males reduce foraging time by reducing their metabolic resting rate and thus can lower their energetic costs (Becker et al. 2013). Female mammals are physiologically limited in the number of offspring they can produce, so they maximize fitness by investing more energy into fewer offspring and do not need to secure as many mating opportunities as males (Andersson 1994). Thus, during the breeding season, mating strategies and activities may differ for males and females which have been shown in a diversity of taxa including rodents (Michener 1998), ungulates (Tettamanti and Viblanc 2014), and pinnipeds (McCann 1983).

Male and female temperate bats show sexual segregation during the summer (Senior et al. 2005), where intersexual differences are well documented in foraging activity (Wilkinson and Barclay 1997; Kerth and Morf 2004; Dietz and Kalko 2007), roost selection (Broders and Forbes 2004; Barclay and Kurta 2007), and use of torpor (Grinevitch et al. 1995); the latter 2 being tightly linked to microclimate preferences (Willis 2006; Boyles 2007). Females incur higher energetic costs during the reproductive period (Kurta and Kunz 1987; Kurta et al. 1990; Mclean and Speakman 2000) which occurs during the temperate spring and summer (Racey and Entwistle 2000). During this same period, males incur lower energetic costs, relative to reproductive females, associated primarily with their own self-maintenance and for spermatogenesis (Wimsatt 1969; Racey and Entwistle 2000). During the autumn, many hibernating temperate bats migrate from summering areas to winter areas, mate, and deposit fat stores for hibernation; several species form large mixed-sex aggregations of individuals within which they engage in swarming activities (Parsons and Jones 2003; Rivers et al. 2005; Furmankiewicz and Altringham 2007). During swarming, bats congregate at underground sites prior to hibernation and engage in vocalizations, chasing and mating behaviors (Thomas et al. 1979). It is thought to be the primary mating event for many species and can occur in autumn and spring (Kerth and Morf 2004; Veith et al. 2004; Rivers et al. 2005; Furmankiewicz et al. 2013). Visits to swarming sites are highly variable and may occur on an hourly or nightly basis but are not well characterized (Fenton 1969; Humphrey and Cope 1976; Rivers et al. 2006; Furmankiewicz 2008). Swarming bats may also gather or exchange information regarding suitability of hibernation sites, knowledge of migration routes, or orient young-of-the-year to such sites (Davis 1964; Fenton 1969; Parsons et al. 2003b).

Compared to the summer, little is known about the intersexual variation in behavior and resource use of bats during the swarming season. Literature documenting resource use (e.g., roosts) is limited (although see Furmankiewicz 2008; Parsons and Jones 2003), likely owing to the difficulty in tracking animals during the migratory period where they roost away from swarming sites. However, physiological studies quantifying the energetics of fat storage prior to hibernation characterize intersexual differences in the timing patterns of fat deposition which suggests the possibility of similar differences in activities undertaken during the swarming period (Kunz et al. 1998; Ingersoll et al. 2010). Lastly, a large observed male bias during swarming (e.g., Cope and Humphrey 1977; Thomas et al. 1979; Kerth et al. 2003; Piksa 2008) has led some to hypothesize intersexual variation in the seasonal timing of use and time spent at swarming sites is due to behavioral differences among the sexes related to mating (Rivers et al. 2006; Glover and Altringham 2008). These studies collectively suggest energetic constraints may explain intersexual variation in activities during swarming. Understanding these intersexual differences may therefore provide insight into optimal fitness strategies of each sex that may in turn lead to insights into population dynamics.

In this study, we conducted capture–mark–recapture surveys at swarming sites to establish if intersexual differences in visits to swarming sites, as a proxy measure of swarming activity, were present in 2 temperate insectivorous bats species. The little brown myotis (Myotis lucifugus) and the northern myotis (M. septentrionalis) are widely distributed temperate insectivorous species of North America. Both species make regional migrations from summering areas to winter hibernacula and are known to swarm during the autumn (Fenton and Barclay 1980; Caceres and Barclay 2000). In the summer, M. septentrionalis is a forest specialist species where it forages within the forest and typically roosts in trees (Jung et al. 2004; Henderson and Broders 2008). M. lucifugus is a more generalist species, roosting in buildings and trees and foraging in or along more open areas such as ponds, wetlands, and forest gap/edge margins (Anthony and Kunz 1977; Fenton and Barclay 1980; Broders and Forbes 2004). Although more is known about M. lucifugus compared to M. septentrionalis, including swarming activities, the similarity of the general life history characteristics related to reproductive and seasonal cycles supports similar expected intersexual differences in both species.

We tested the hypothesis that sexual differences in energy budgets would lead to different strategies in the frequency and timing of activities during swarming (Rivers et al. 2006; Glover and Altringham 2008). Specifically, males having spent the summer mainly for self-maintenance would allocate more energy to mating activities during swarming to secure as many mating opportunities as possible. In contrast, reproductive females having spent the summer rearing young may not need as many copulations to maximize fitness and should allocate more activities to rebuild their own energy stores in preparing for hibernation and less to mating activities. We test the predictions that for bats captured at swarming sites 1) there would be a male bias resulting from male bats spending more time at swarming sites and 2) the swarming season would be longer for males than females. In recaptures of bats, we predicted 3a) a higher proportion of male recaptures than females and 3b) individual males to be recaptured, on average, more often than females because to maximize copulations males should stay longer at swarming sites or visit more frequently than females. Lastly, 4) a greater proportion of male recaptures would be at the site of initial capture compared to females as males should have a higher swarming site fidelity using fewer swarming sites to allow them to visit them more frequently.

Materials and Methods

Capture and tagging.

Bats were captured at 6 swarming sites in Nova Scotia, Canada (Fig. 1; Table 1), during the autumn and spring seasons of 2008–2011 using harp traps (Austbat Research Equipment, Lower Plenty, Victoria, Australia) or mist nets (Avinet, Dryden, New York). Sites were separated by distances ranging from 27.9 to 98.9 km. Captures during the spring seasons were done to increase the number of tagged individuals for the study but were not included in analysis. Individuals were identified to species and sex with age (young-of-the-year; or adult) determined by examining the degree of ossification and shape of the epiphyseal growth plates of the metacarpals (Anthony 1988). Depending on the nightly capture numbers, we tagged all or a subset of captures with permanent, passively integrated transponder tags (PIT-tags; Trovan ID 100, EIDAP Inc., Sherwood Park, Alberta, Canada) following methods described in Burns and Broders (2015). All bats were released prior to sunrise with a mean total handling time from capture to release of 42 (range: 10–180) min. Methods for the capture and handling of bats followed American Society of Mammalogists guidelines (Sikes et al. 2011), were approved by the Saint Mary’s Animal Care Committee, and were under yearly reviewed permits from Nova Scotia Department of Natural Resources.

Table 1.

Number of in-hand identified, adult Myotis lucifugus and M. septentrionalis bats captured at 6 swarming sites in Nova Scotia, Canada (2009–2011) by location.

Site	MAP code	Sample nights	M. lucifugus			M. septentrionalis
			Males	Females	Total	Males	Females	Total
Cave of the Bats	CVB	6	25	6	31	41	21	62
Cheverie Cave	CHC	6	38	25	63	14	17	31
Hayes Cave	HAY	11	105	112	217	26	19	45
Lake Charlotte Mine	LKCH	5	37	26	63	24	7	31
Lear Mine	LEAR	9	65	46	111	31	17	48
Rawdon Mine	RAW	26	187	53	240	109	61	170
Total		63	457	268	725	245	142	387

Fig. 1.

Locations of swarming sites (caves and abandoned mines) surveyed to assess intersexual differences in swarming activity of bats, Nova Scotia, Canada (2008–2011). Site codes are listed in Table 1.

An emergent fungal pathogen, Pseudogymnoascus destructans, which causes white-nose syndrome (WNS), has resulted in both study species suffering dramatic recent declines in their populations in eastern North America (Blehert et al. 2009; Turner et al. 2011; Minnis and Lindner 2013). These 2 species were recently listed as endangered under the Canadian Species at Risk Act (SARA). Therefore, we used the most up-to-date precautionary WNS decontamination protocols provided by the United States Fish and Wildlife Service to try and minimize potential spread of fungal spores via our capture and handling methods (United States Fish and Wildlife Service 2015). In the late winter of 2010/2011, WNS was detected in Nova Scotia and therefore to reduce the chance of transmission from our work, we reduced the number of active trapping and tagging sessions in 2011 with no spring tagging session occurring.

Recaptures of bats were assessed during the autumn swarming period (2009–2011) via 3 methods. First, in all 3 years, we conducted active trapping sessions at swarming sites (autumn only) to actively hand-scan all captured bats for PIT-tags. Second, in 2010, we set up harp traps with PIT-tag antenna fitted in holes we cut in the sides of the harp trap bags (PIT-harp trap). This facilitated bats being captured and passively scanned for a tag as they escaped out through the holes housing the antenna. These modified PIT-harp traps were left out over multiple nights (6–24) at secure sites to passively scan bats; those left out > 1 week were checked minimally on a weekly basis. Prior to deploying the PIT-harp traps without personnel present, we conducted trials where we observed the behavior of bats in the traps and confirmed that captured bats found the holes quickly and escaped. The extended deployment of these traps had the potential to influence bat behavior at the site (e.g., avoidance). However, we examined capture records for the year before (2009), during (2010), and the year after (2011) their deployment and general seasonal trends were similar in all 3 years (e.g., peak in capture numbers during the same 2-week period and similar capture totals). This suggests that bats did not avoid the site due to the presence of the PIT-harp trap. Third, in 2011, we installed PIT-tag antenna in temporary mesh gates constructed at 4 swarming site entrances to passively scan bats as they entered underground sites. The mesh allowed air flow into the sites to minimize any effect of the gates on hibernacula microclimates. Therefore, recaptures encompass PIT-tagged bats detected via one of the 3 methods. We discontinued using the passive PIT-harp traps in 2011 because they had the potential to be a vector for the spread of WNS. Active trapping and scanning was conducted at sites in autumn 2011 where gates could not be constructed with a few sessions still occurring at gated sites.

Analyses.

We excluded young-of-the-year from our analyses because they may represent another intraspecific group with distinct activity patterns. Therefore, our data do not represent total captures at each site but rather a subset composed of in-hand sex-identified adults. We used G-tests to examine differences in swarming activity measures between male and female bats because of their additive properties (Sokal and Rohlf 1995; Macdonald 2009). For all tests, significance was considered at α = 0.05.

To determine if there was male bias in bats using swarming sites (prediction 1), we compared the proportions of adult males and adult females captured during autumn swarming, for each species, using G-tests at 3 levels. We first tested if the overall proportion of males was greater than that of females across all sites, for each year, using a replicated goodness of fit G-test. Second, we tested the proportions of each sex at the site level, across all years, using the same procedure as above. Lastly, we conducted an unplanned G-test of the homogeneity of replicates, using the simultaneous test procedure, to examine if proportions of each sex captured at swarming sites changed during the swarming season. Here, we classified the swarming season into 8, week-long periods for the capture data that began on 11 August (the earliest survey date of all years). Sites were not sampled on the same night every year, or with equal frequency owing to variability in weather among years and to meet other concurrent study objectives. Therefore, we pooled sampling nights across sites and years for each week-long period with the sum of captures of each sex during each week-long period used for analysis. The mean number of sampling nights included in each period was 7.8 (range: 4–13). Sample weeks were ordered from highest to lowest proportions of males observed. Varying sets of weeks, starting from the largest and from the smallest proportion of males observed, were examined in sequence for homogeneity in the magnitude of the observed proportions until heterogeneous sets were identified. These heterogeneous sets are indicative of significant changes in the proportions of each sex.

Small sample sizes resulting from a limited number of swarming seasons and the necessary exclusion of the reduced 2011 capture season (due to the appearance of WNS) precluded statistical assessment of season length between the sexes for prediction 2. However, we qualitatively present data trends as supplementary data to inform our research question. First, we restricted the data to 1 site, Rawdon, which was sampled in every week of the 8-week-long swarming period in 2009 and 2010. We calculated the minimum season length for each sex by considering the number of nights between the first and last identified capture of each sex at this site for each year. As a 2nd measure of season length at the site, we used recapture data from 2010 when we had a PIT-harp trap deployed for most of the season at the site from 14 August to 19 October and concurrently actively trapped bats on 9 nights during this period. In 2011, we had continuous nightly PIT-tag gate sampling of tagged individuals entering or exiting the site from 28 April through to 8 November 2011. We first excluded any individuals from the data set that were detected using the site in June or July as we considered these individuals as “local summer residents” of the surrounding swarming area. We then calculated the minimum season length of recaptured, “autumn transient” bats as the number of nights between the first and last recapture detection of each sex between the dates of 1 August and 31 October 2011.

To test if the recapture rates for males were greater than that of females (predictions 3a and 3b), we compared the total number of individual males recaptured to that of females using only those individuals that were adults at the time of recapture. Recaptures were classified on a per night basis where individuals were detected ≥ 1 time. Since there were many instances of only 1 sex recaptured per site for a given season, we pooled the data over sites and years testing total male and female recaptures. We used a 2-sample equality of proportions test (Sokal and Rohlf 1995) to determine if the proportion of tagged males that were recaptured was greater than the proportion of tagged females that were recaptured. We also examined recaptures for 2011 at Rawdon at the individual level where we summed the total number of recaptures detected per adult and compared the number recaptures per individual for males and females using a Mann–Whitney U-test (Sokal and Rohlf 1995). For visualization of the data, we classified them into 4 categories of recapture histories: 1, 2, 3, and ≥ 4 recaptures because there were few recapture histories that exceeded 4 recapture events for females of both species.

As a 2nd assessment of weekly seasonal swarming activity patterns derived from capture data, we compared the changes in the proportion of male and female captures over the season from pooled capture data to that of recapture data collected in 2011 from Rawdon. We reduced the Rawdon 2011 recapture data set to encompass 11 August to 5 October, for the same weekly intervals, and summed the total bats recaptured of each sex in each week. Finally, no recaptures of males or females tagged at our swarming sites were detected using different swarming sites precluding any analysis for prediction 4.

Results

In total, from 2008 to 2011 (spring and autumn seasons), we tagged 865 M. lucifugus (220 females, 645 males; Supporting Information S1) and 482 M. septentrionalis (167 females, 315 males). No inter-swarming site movements were detected as all recaptures were at the site of capture. Over the 3 autumn swarming seasons, we captured and identified in-hand 725 adult M. lucifugus (268 females, 457 males) and 387 adult M. septentrionalis (142 females, 245 males) at the 6 swarming sites (Table 1). The overall proportions of adult males and adult females captured during autumn swarming had a large male bias (prediction 1). In examining by each season (year), there was a male bias observed for both M. lucifugus and M. septentrionalis (G_T; Table 2); a similar male bias was detected when data were examined by site (Table 2). Despite the overall male bias, for M. lucifugus, the magnitude of male bias differed among years and among sites as shown by the heterogeneity G-test (G_H; Table 2). In M. septentrionalis, a similar magnitude of male bias was detected at all sites and in each year as shown by the nonsignificant heterogeneity G-test (G_H; Table 2).

Table 2.

G-test statistics and significance of heterogeneity G (G_H), pooled G (G_P), and total G (G_T) to test for male bias in adult captures at swarming sites in Nova Scotia for Myotis lucifugus (MYLU) and M. septentrionalis (MYSE), 2009–2011. Calculations were performed by pooling the data and testing for site differences and yearly seasonal differences.

	By site						By year
	MYLU	d.f.	P	MYSE	d.f.	P	MYLU	d.f.	P	MYSE	d.f.	P
GH	50.1	5	< 0.001	7.9	5	0.159	16.5	2	< 0.001	2.3	2	0.325
GP	49.8	1	< 0.001	55.5	1	< 0.001	49.8	1	< 0.001	27.7	1	< 0.001
GT	99.9	6	< 0.001	63.4	6	< 0.001	66.3	3	< 0.001	30.0	3	< 0.001

Variability in the degree of male bias observed was found for different weeks in the swarming season in M. lucifugus. The simultaneous test procedure indicated that weeks 2, 7, and 8 had a higher proportion of females relative to week 1 which had a larger degree of male bias (Fig. 2). From the recapture data collected in 2011, we detected more females in weeks 2, 3, and 7 supporting general trends of the capture data (Fig. 3; Table 3). Notably, total recaptures of both male and female M. lucifugus were dramatically reduced during weeks 2 and 3 despite this being the period of high captures from capture surveys. For M. septentrionalis from capture data, weeks 2, 3, 4, and 6 had a higher proportion of females compared to the latest period of the swarming season in week 8. This reflects that over the 3 swarming seasons sampled, no adult female M. septentrionalis was ever captured and identified during week 8. Recapture data for M. septentrionalis (2011) showed a similar seasonal pattern to that characterized by capture data with more females detected during weeks 2 through 5 and then declining to a larger degree of male bias in weeks 7 and 8. Trends from data collected at Rawdon suggest that male bats, of both species, may have had longer swarming season lengths than females (prediction 2; Table 4).

Table 3.

G-test statistics and significance of heterogeneity G (G_H), pooled wG (G_P), and total G (G_T) to test for male bias in adult recaptures at swarming sites in Nova Scotia for Myotis lucifugus (MYLU) and M. septentrionalis (MYSE), 2009–2011.

Species	GH	d.f.	P	GP	d.f.	P	GT	d.f.	P
MYLU	3.4	3	0.341	110.9	1	< 0.001	114.3	4	< 0.001
MYSE	2.7	3	0.447	28.2	1	< 0.001	30.9	4	< 0.001

Table 4.

Swarming season length of each sex, for Myotis lucifugus and M. septentrionalis at Rawdon, Nova Scotia, 2009–2011. Data were the total in-hand identified captures from 2009 to 2010 and recapture data of PIT-tagged bats from 2010 to 2011.

Year	Data type	Sampling period length (# nights)	M. lucifugus		M. septentrionalis
			Males	Females	Males	Females
2009	In-hand captures	53	53	48	38	27
2010	In-hand captures	53	51	48	51	37
2010	PIT recapturesa	67	58	36	66	26
2011	PIT recapturesb	92	79	65	75	57

^aRecaptures from active and passive methods.

^bRecaptures from passive methods only.

Fig. 2.

Proportion of total identified in-hand captures that were adult males and adult females and sex ratio during each week-long period of the autumn swarming season at 6 sites in Nova Scotia (2009–2011) for A) Myotis lucifugus (MYLU) and B) M. septentrionalis (MYSE). Number of sampling nights per week period is indicated in brackets where each week encompassed 7 days starting on 11 August of each year.

Fig. 3.

Proportion of total adult males and females recaptured during each week-long period of the autumn swarming season at Rawdon, Nova Scotia (2011), for A) Myotis lucifugus (MYLU) and B) M. septentrionalis (MYSE). Each week encompassed 7 days starting on 11 August through to 5 October.

For M. lucifugus, a significantly higher proportion of recaptured males were detected than females (prediction 3a: 17% male versus 6.3% female recaptures; χ² = 14.1, d.f. = 1, P < 0.001). The large male-biased pattern was the same for M. septentrionalis (18% male versus 12% female recaptures; χ² = 3.3, d.f. = 1, P = 0.034). Individual male M. lucifugus recapture histories ranged from detections of 1 to 19 times and females, 1 to 4 times. Individual male M. septentrionalis recapture histories ranged from 1 to 22 recaptures and females, 1 to 11 recaptures. Although the range of the number of recaptures for individual females was smaller compared to males (Fig. 4), the difference was not statistically significant for M. lucifugus (U = 702, P = 0.813) or M. septentrionalis (U = 403.5, P = 0.079).

Fig. 4.

Number of individual male and female bats, in each recapture history category of the total number of recapture detections observed for A) Myotis lucifugus (MYLU) and B) M. septentrionalis (MYSE).

Discussion

In line with our predictions, we found that male bats had higher autumn swarming activity compared to female bats of both study species. A male bias was found in captures across all sites and in all swarming seasons. Previous studies of M. lucifugus detected this male bias during swarming (Fenton 1969; Humphrey and Cope 1976; Schowalter 1980) and male bias has been shown in many European swarming species (Kerth et al. 2003; Rivers et al. 2006; Furmankiewicz 2008; Glover and Altringham 2008; Piksa 2008). Taken together, these data support the assertion that a large male bias is suggestive of intersexual variation in strategies to maximize fitness during this specific season rather than reflecting population-level sex ratios. For our work, we made the necessary assumption that the sex ratio in the regional population is equal or female biased compared to our swarming results. With no available data on sex ratios from summering or overwintering sites, this assumption remains untested. However, since similar biases have been observed in multiple species that face similar seasonal constraints during swarming, these data collectively support a strong signal of swarming male bias. Although there are differences in the timing of use of swarming sites among species as noted in our own and other studies (e.g., Schowalter 1980; Parsons et al. 2003a; Glover and Altringham 2008), a general large male bias is detected regardless of when the peak of swarming activity of each species is.

Despite the overall male bias, differences in the degree of male bias can be found at varying temporal scales for swarming bats. For example, Piksa (2008) documented subtle differences in the nightly timing of swarming M. mystacinus at a high elevation site where there was typically an influx of females late in the evening compared to early captures. We did not sample continuously throughout each night to be able to assess if similar variation occurred at our study sites. Within a swarming season, we found female activity peaked early to mid-swarming season for M. septentrionalis and M. lucifugus. We suggest this reflects the different activities of each sex during this period. Females appear to visit less often, and temporally concentrate their activities, possibly for mating, compared to males which visited more often to maximize potential copulations. Females, having reared young in the summer, may spend the majority of their time away from swarming sites allocating more time to foraging to rebuild depleted energy stores for hibernation and reproduction in the following spring (Jonasson and Willis 2011). Mid-season peaks in female visits to swarming sites have been found in other swarming species in Europe (Glover and Altringham 2008; Piksa 2008). The short swarming season for females, relative to males, supports the contention that females make fewer visits to swarming sites than males. Further, the swarming season of males encompasses the entire season for females, likely to maximize potential for copulations.

In M. lucifugus, a 2nd smaller peak of female activity occurred at swarming sites late in the season. We propose that female M. lucifugus may initially show up in the 1st wave in late August at the site to mate as has been found in other areas (Thomas et al. 1979; McGuire et al. 2009), possibly to assess the site for suitability for hibernation and then leave. The 2nd peak may then represent females returning to a site for immergence into hibernation. As a forest specialist, female M. septentrionalis may use autumn forest roosting and foraging resources in the surrounding area of the swarming site such that there is less of a gap between mating and/or site assessment and actual immergence at the site for hibernation. Therefore, their activity at swarming sites appears more continuous. Detailed tracking studies on females of each species would be required to test this hypothesis.

If female bats maintain distinct pulses of high swarming activity year after year, then males can potentially cue in on these female activity peaks to maximize copulations when more females are available. In examining recapture records where we had continuous scanning coverage in 2011, we found that for M. septentrionalis, the levels of activity by tagged males and females closely matched the seasonal pattern of activity from the capture data. Although not an independent data set, this congruence suggests female M. septentrionalis may concentrate their swarming activity at approximately the same time each year and that at least some males appear to track this. For M. lucifugus, the pattern is partially discordant between the 2 data sets where the 2nd peak of females is present near the end of the season, but notably, a large peak in female and male recaptures corresponding to the female capture peak was not detected.

We speculate that this is a period of high transiency by M. lucifugus since captures at swarming sites remain high during this period which shows that bats are still swarming during this time. However, these captures appear to be dominated by transient individuals rather than more local bats to the site and thus this may reflect a period of migratory or dispersal movements among swarming sites. The recapture of a few males and females during this time suggests at least some individuals show a degree of swarming site fidelity. There are several observations of M. lucifugus roosting in atypical locations and structures following summer colony breakup (early to mid-swarming—Davis and Hitchcock 1965; Schowalter 1980; Riskin and Pybus 1998). Further, a recent analysis of banding records found that M. lucifugus captured during swarming had the highest movement rates of all individuals studied (summer, winter, or swarming—Norquay et al. 2013). Together, these studies suggest this period is one of high movement and transitioning by individuals.

Given the maximum longevity records of the species to be 18.5 (M. septentrionalis—Caceres and Barclay 2000) and 34 years (M. lucifugus—Davis and Hitchcock 1995), it may be that some males learn to exploit this temporal peak in female abundance. This could include older more experienced males or those that roost near females in the summer that can track their movements to swarming areas. We speculate that this could have potential consequences for male individual reproductive success. Although the mating system for M. lucifugus (Thomas et al. 1979) and possibly for M. septentrionalis is characterized as promiscuous, some males could have higher reproductive success if they match their activity to when more females are available and are thus able to secure more copulations. Reproductive skew in the number of offspring sired by males or male lineages has been shown for M. lucifugus where mating during hibernation (Watt and Fenton 1995) or cryptic female choice (Wilkinson and McCracken 2003) has been posited to potentially explain this pattern. Some males may also achieve copulations away from swarming sites similar to that found in swarming M. daubentonii and M. myotis (Zahn and Dippel 1997; Angell et al. 2013) which could also explain this skew. Along with our work, these studies suggest the possibility of different temporal, spatial, or social aggregating mechanisms acting on female bats which permit different male mating strategies to occur as individual males maximize their own fitness.

The 2nd component of our study using recapture records of tagged bats further supports the prediction that males had greater swarming activity than females. Although the inherent male bias in captures resulted in us tagging more males, after correcting for this, our total recaptures of females were still lower than that of males. This finding mirrors that of swarming M. nattereri and M. daubentonii in the United Kingdom where fewer females were recaptured compared to males (Parsons et al. 2003a; Rivers et al. 2006). At the individual level, we did not find differences in the recapture histories of males and females. However, we believe this is a result of limitations in tracking individuals continuously, with equal sample effort throughout our study, and from the generally low recapture rate of tagged individuals. As found with other swarming tagging studies (Humphrey and Cope 1976; Rivers et al. 2006; Norquay et al. 2013), our overall recapture success was low where even among those that were recaptured, many were only recaptured once during the entire study. We believe this may reflect the high degree of mobility of these species during this time or possibly that we sampled only a small portion of the larger population. Although we did not detect any inter-swarming site movements, we were not able to monitor all 6 sites continuously for the 3 seasons and there are many other known and potential swarming sites in Nova Scotia that we did not monitor (Moseley 2007; Randall and Broders 2014). Since our study was restricted to 3 years of monitoring relative to the long life span of individual bats, age-related factors determining the tendency or frequency of movements of individuals among sites over a lifetime may also play an important role in the swarming dynamics of these species.

Capture and tagging data have inherent biases due to higher likelihood of capturing the most mobile or easily trapped individuals (Biro and Dingemanse 2008) which may have influenced our study. Also, our recapture survey effort differed by the method used (capture versus passive detection) and varied throughout the study which may also have impacted detection of recaptures. However, we believe the general concordance of the metrics via both methods, despite the inherent limitations and biases, support a clear signal of males exhibiting higher swarming activities compared to females. Since other studies of swarming species show similar trends (e.g., Fenton 1969; Kerth et al. 2003; Rivers et al. 2006; Glover and Altringham 2008; Piksa 2008), we believe these results collectively demonstrate that male and female bats do show intersexual differences in swarming activities.

In conclusion, we have shown that intersexual differences in activities occur for 2 temperate swarming species of bats in the timing and frequency of swarming site visitations. During autumn swarming, males are more abundant at sites and appear to spend more time individually and collectively over the season visiting swarming sites compared to females. Although their activities overlap during this period, male activity may be strongly determined by female activity. The differences in activities may further suggest that sexual segregation may occur in day roosting or foraging areas used despite eventual meeting at swarming sites for mating and other activities. Since little is known of resources bats use (foraging and roosting) and movements they make (e.g., routes used and frequency of) during this time period, future work characterizing these aspects may provide additional insights into the intersexual differences among swarming bats.

Supporting Information

The Supporting Information documents are linked to this manuscript and are available at Journal of Mammalogy online (jmammal.oxfordjournals.org). The materials consist of data provided by the author that are published to benefit the reader. The posted materials are not copyedited. The contents of all supporting data are the sole responsibility of the authors. Questions or messages regarding errors should be addressed to the author.

Supporting Information S1.—Number of individual Myotis lucifugus and M. septentrionalis tagged and later recaptured, by sex (M = males, F = females) at 6 swarming sites in Nova Scotia (2008–2011).

Acknowledgments

We are grateful to the Weatherby family who allowed us access to their land for sampling. We thank L. Lawrence, L. Farrow, R. Hearn, A. Park, A. Lowe, J. Randall, Z. Czenze, and A. Burns for providing excellent assistance in the field. Funding was provided by the Canadian Wildlife Federation, Nova Scotia Habitat Conservation Fund (contributions from hunters and trappers), Nova Scotia Species at Risk Conservation Fund, Bat Conservation International, the American Society of Mammalogists, Patrick F. Lett Graduate Student Assistance Bursary to LEB, and Natural Sciences and Engineering Research Council (NSERC) grants to LEB (NSERC-PGS-D) and HGB (Discovery Grant). We thank M. Leonard, H. Whitehead, T. Frasier, and G. Jones, and an anonymous reviewer for helpful comments that greatly improved this manuscript.

Author notes

Associate Editor was Richard D. Stevens.