Materials & Methods

Study area.–The Fresno Creek Bridge is one of five bridges along Texas State Hwy. 170 between Lajitas and Redford, within or adjacent to the southern boundary of BBRSP. The other four bridges, which cross Topado Canyon and Contrabando, Panther, and Madera creeks, are not structurally conducive for T. b. mexicana roosting, and no sign of T. b. mexicana was observed at those four bridges during preliminary examination of these structures.

The Fresno Creek Bridge (29.297033 N, 103.848182 W; 734 m) was originally built in 1982 and has a total length of 50.3 m, with the only vertical supports located at the center of the bridge. For each half of the bridge, eight abutted pre-cast concrete box beam girders traverse from each ledge of Fresno Canyon to the center vertical supports. Among the eight concrete girders, seven open vertical slots created during construction cross the bridge span and are used as roosting sites by bats (Fig. 1). Based on 14 random measurements, slots varied from 1.9 to 2.2 cm wide, with the outer two slots being the narrowest at 1.9 cm. Depths varied at arbitrary points from 14.6 to 27.3 cm due to flowed concrete poured during construction, although the majority of measurements ranged from 23.0 to 27.3 cm. Space between the bridge bottom and guano covered canyon ledges was ca. 1 m in height. Because the shear canyon walls have a ca. 20 m drop to the creek bed, safe access was restricted to only ca. 5–7.5 m at each end of the bridge. Habitat at the level of the bridge is typical Chihuahuan Desert scrub dominated by creosote-bush (Larrea tridentata) with a mosaic of other plants including mesquite (Prosopis glandulosa), ocotillo (Fouquieria splendens), acaia (Acaia sp.), agarito (Berberis trifoliolata), and Opuntia cacti (Yancey 1997) with sparse patches of seepwillow (Baccharis sp.) in the desert riparian environment in the creek bed near the bridge (Fig. 2). The intermittent spring fed drainage flows southward into the nearby Rio Grande, ca. 1.8 km south of the bridge.

View Figure

• Download as Powerpoint
• Download Figure

View Figure

• Download as Powerpoint
• Download Figure

Bat capture and observations.–Investigation of T. b. mexicana roosting in the bridge included a general visual and auditory assessment of bat activity during each visit (e.g., flight activity, bats moving in slots versus torpidity, presence of vocalizations, etc.). For 10 dates examined, bats were always present however there was no sampling on two dates because bats were not safely accessible. Search for individuals from slots under the bridge was accomplished by supine visual examination. Bats were removed from vertical slots using two soft-wood paddles to persuade each individual bat to crawl down to the slot opening where bats were captured by hand. No mist netting was done because bats left the roost at the inaccessible center of the bridge. Individual bats were examined for age, sex, and reproductive condition, and their handling followed guidelines in Gannon et al. (2007).

Each T. b. mexicana was determined to be an adult or juvenile (considered herein as fledgling or young of the year) based on the degree of ossification of the metacarpal-phalangeal finger joints of the forelimbs and shape of finger joints after the cartilaginous plates have become fully mineralized (Anthony 1988). Finger joints follow a continual progression from evenly tapered closed joints to the adult condition of fully knobby or bulbous joints; experienced qualitative assessment can readily determine young of a season easily through the following winter and has been used to identify bats to one year of age (Anthony 1988). Juvenile bats in this study are considered to be maternity colony fledglings of the current or previous summer. Location of each bat examined was recorded in terms of the end of bridge (west or east) and specific slot number (numbered 1–7 north to south). Selected voucher specimens (skin, skull, tissues) were collected and deposited in the Natural Science Research Laboratory of the Museum, Texas Tech University. All other T. b. mexicana individuals were released immediately at the capture site after data were collected.

Results & Discussion

The Fresno Creek Bridge was examined 10 times from 3 January 2016 to 28 July 2017. Numbers of T. b. mexicana were conservatively estimated to be between 2,000–4,000 for dates when an exodus was observed and photographed, whereas on other dates bat density was estimated based on distant visual examination of slot fill and vocalizations. During eight of 10 visits, some bats roosted in slots at the ends of the bridge, allowing access for sampling of individuals. On two visits, although bats were present in substantial numbers (estimated from slot observation toward center of bridge and vocalizations), no individuals roosted near the ends of the bridge and capture by hand was not possible. A total of 153 T. b. mexicana was sampled; nine adult females, 80 adult males, 28 juvenile females, and 36 juvenile males (Table 1). The sample on 17 April 2017 included a mixture of older adults with fully bulbous metacarpal-phalangeal finger joints and a large number of presumptive juveniles of the previous summer, which had highly variable metacarpal-phalangeal finger joints that were near the adult condition. All 30 individuals on this date were considered adults (Table 1). Where 20 or more bats were sampled, specific locations of each bat capture in the bridge by sex and age are presented in Table 2. On 17 April 2017, two male Yuma Myotis (Myotis yumanensis) were captured, both in separate T. b. mexicana clusters. Two male T. b. mexicana individuals from 3 January 2016 (TK195516, TK195517) and one M. yumanensis (TK195528) were retained as voucher specimens.

Table 1 Results of surveys for Tadarida brasiliensis mexicana at the Fresno Creek Bridge, Big Bend Ranch State Park, Presidio, Co., Texas, and the 24 h temperature ranges (°C) at the airport in Presidio, Texas. Activity is defined as being in torpor or active (bats were either vocal, moving within slots, or in flight). N = total number of bats captured, sexed, and aged for the sampling date; ns = none sampled, bats present but not accessible.

View Fullscreen

Table 2 Distribution of Tadarida brasiliensis mexicana by sex and age within the slots at the accessible ends of the Fresno Creek Bridge, Big Bend Ranch State Park, Presidio Co., Texas. Only samples of 20 or more are included. Slots are numbered from north to south.

View Fullscreen

Flights of bats at the bridge were viewed on four dates and are here seasonally compared for sunset activity. On 19 March 2016, bats were congregated toward the central two-thirds of the bridge and were vocal and active in slots. After sunset some bats were seen flying under the bridge shifting from slot to slot but were hesitant to fly away from the bridge, suggesting late winter daily temperatures (8.3–19.5° C) were not favorable for a feeding exodus. On 24 and then 26 June 2016 (23.9–37.8° C), summer evening flights began just before sunset with the densest exodus occurring about 7–10 min after sunset. For a winter date of 31 Jan 2017 (−2.8–13.9° C), several bats were noted flying under the bridge at 1844 h and flying continued until ca. 1915 h with no flight away from the underside of the bridge. The majority of bats were moving and calling but remained in their roosting slots at the center of the bridge.

Although the inability to sample across the full length of the bridge was unfavorable to fully understanding the year-round colony of T. b. mexicana using the Fresno Creek Bridge at BBRSP, the bridge consisted of a summer maternity colony and was used by nonreproductive adult males during spring and summer (Table 1). Although non-volant juveniles were not observed in the slots at the ends of the bridge, use of the bridge by T. b. mexicana as a maternity colony is evidenced by the presence of volant juveniles over two seasons.

Juvenile bats were found in June and September 2016 and February 2017, and juveniles were again found for another maternity season in July 2017 (Table 1). Overall totals of juvenile sexes were somewhat similar at 36 males and 28 females, however proportions of juveniles are skewed for two larger samples (18 September 2016 and 5 February 2017). These two samples suggest that juveniles may be sexually segregating among slots, at least to some degree (Table 2). A similar pattern of roost use was found in a highway overpass in central Texas where banded T. b. mexicana individuals were found to segregate by both sex and age (Sgro & Wilkins 2003).

No adult T. b. mexicana individuals were collected during the winter months for both years surveyed at the Fresno Creek Bridge, only juveniles were sampled for January and February. First encounters of adults for each year were 26 June 2016 and 17 April 2017. Overall patterns of sampling suggest that adult bats are absent by mid-to-late summer with at least some young of the year remaining at the bridge through the winter (Table 1). Juveniles of the summer being left at a maternity roost to overwinter is a novel behavior for T. b. mexicana. Mass departure of adults during summer could be due to either the natural timing of southward migration or a response to overcrowding at the bridge after fledglings have become volant. Ammerman et al. (2012) noted that after late July at cave colonies, fledglings predominate in feeding flights and tended to reside at their birth place until the onset of cool weather in October and November. In the lower Rio Grande Valley of New Mexico, overwintering T. brasiliensis individuals were found at 12 bridges over two years with their numbers progressively decreasing from October to winter lows during December, January, February, and March (Table 3 in Geluso and Mink 2009), however this study did not discriminate ages during winter. Other studies from central Texas, which documented T. b. mexicana remaining at the roost during winter, also did not differentiate between adults and young of the year (Spenrath & LaVal 1974; Keeley & Keeley 2004; Scales & Wilkins 2007; Mink 2012; Weaver et al. 2015).

During this study, 89 adult T. b. mexicana individuals were examined, 80 (90%) were males and nine were females (10%). These highly skewed proportions are most likely due to adult males segregating to the ends of the bridge and consequently are more readily sampled, while females are concentrated in slots at the center of the bridge. Male T. b. mexicana have been found to be more faithful to their roost and geographic area than females (Scales &Wilkins 2007) and males at a highway overpass roost in central Texas were found to dominate a single section of an overpass (Sgro & Wilkins 2003). The nine adult females were sampled only during spring and summer (April to July). Based on patterns of bridge use by adult bats, the inference is made that adult females use the central (inaccessible) part of the bridge as a maternity roost with adult males relegated to ends of the bridge. Two maternity period sampling dates in 2017 reinforce this pattern of adult distribution at the bridge. On 13 May, 32 T. b. mexicana individuals were removed, 31 adult males and one adult female; on 28 July, 26 bats included 24 adult males, no adult females, and two volant juveniles (Table 1). Furthermore, the central region of the bridge would be more beneficial as a maternity area, as it would not be accessible to local terrestrial predators (e.g., a large, 161 cm, Trans-Pecos Rat Snake, Bogertophis subocularis, was observed under the west end of the bridge just before a June sunset). During late spring and early summer, adult females and their juveniles would have fidelity to the center of the bridge, whereas adult males would be segregated to the ends of the bridge. Sgro & Wilkins (2003) found a similar pattern in a highway overpass in central Texas where 70% of banded T. b. mexicana segregated by both sex and age, and each group had site-specific fidelity within the large roost. Whether non-reproductive spring and summer adult males are transients from other roosts or spring migrants that have remained at a lower latitude is not determinable from sampling in this study.

Compared to natural caves, bridges potentially provide more favorable roosting sites for T. brasiliensis. Allen et al. (2011) found significantly higher cortisol levels in cave-roosting T. brasiliensis than individuals roosting in bridges, and those authors concluded that bridge-roosting individuals are less stressed than cave-roosting conspecifics. Maternity colonies at bridge roosts may also have favorable conditions for raising pups as supported by having an average heavier birth weight and faster growth than pups from cave roosts (Allen et al. 2010). Additionally, bridges warm faster than caves, especially during the colder, late fall season, which can be beneficial to later fall migrants (Ammerman et al. 2012). A warm bridge in late fall can lead to extended occupancy by individuals in a T. b. mexicana colony, which may procrastinate migration through the entire winter season. This is most likely the process that has occurred with the Fresno Creek Bridge T. b. mexicana juveniles, at least for the two winter seasons of this study. Bats were observed in torpor on 3 January (5.0–8.9°C) and 4 January 2016 (6.7–10.6°C) when prior and current weather conditions were overcast and rainy. Conversely, on the coldest day that bats were observed during this study on 31 January 2017 (−2.8–13.9°C), many individuals were actively flying under the bridge with most bats vocal and moving within slots. Although cold, conditions on and before this date were dry and sunny, which allowed the bridge to accumulate solar energy. Also, individual bats that were in torpor would be able to readily awaken to an active state with an adequate supply of solar radiation. A prevalence of sunny days during winter in the vicinity of the Fresno Creek Bridge would make the bridge a suitable winter roost in a region where historically T. b. mexicana was documented to migrate to warmer latitudes in Mexico (Schmidly 1977). The Fresno Creek Bridge would be able to satisfy basic physiological needs for a winter colony with its close proximity to food resources along the nearby Rio Grande and adequate thermal roosting environment, which is maintained by accumulating solar radiation from the black-top road surface.

Temporal shifts and long-term changes in migratory behavior of the subspecies T. b. mexicana across its range in Texas may be associated with consequences of a changing climate. Distributions of numerous species of Texas bats have been changing over historic times (Yancey & Jones 2006), probably in response to overall warming of climate as a result of rising CO₂ levels in the atmosphere (Scheel et al. 1996). This is a general theme hypothesized for many taxa of bats occurring in the New World (LaVal 2004; Popa-Lisseanu & Voigt 2009).