Statewide Evaluations
Data were recorded on standardized forms at 149 bridges and 911 culverts within 39 study areas
throughout the state, and of those, daytime roosts were found in 23 bridges and 18 culverts. A total of
221 night roosts also were discovered.
Survey sites were selected statewide to include a suitable sampling of all major ecological regions (see Figure 1). Within these ecoregions, each site was selected by simply choosing a highway intersection and sampling all bridges and culverts for equal distances along both roads. Sampling distances varied according to the frequency of bridges and culverts. Study sites within each ecoregion were selected to best represent the major highway types.
A one-tailed t-test was used to evaluate roosting preferences, with daytime bat use as the dependent variable (see Table 1). Micro-habitat characteristics included all the variables associated directly with highway structures. These characteristics include: construction materials, design, and measurements of the overall structure, roost dimensions, substrates, and textures, whether they were open or closed above, and the substrates found beneath potential roosting areas. Macro-habitat characteristics included ecoregions and local land-use patterns.
As in the paired comparisons, the t-tests from statewide evaluations showed significant bat preferences for: 1) the largest concrete bridges and culverts; 2) bridges made of diaphragmed, prestressed concrete girders; 3) increasing structure height and lack of vegetation or other obstacles beneath; 4) crevices 12 inches (30 cm) or more in depth by 0.75 to 1 inch-wide (1.9 to 2.5 cm) with covered tops; 5) culverts more than 30 inches (76 cm) tall; 6) roughened surfaces; and 7) locations found in mixed forests and grasslands in south, central, and western Texas. Within these regions (6, 7, and 10 in Map 1), 76% of bridges that provided covered crevices of suitable dimensions were occupied (50% for 6, 82% for 7, and 100% for 10).
Map 1
Stars Indicate Statewide Evaluation Study Sites
| Region | Percent use* | Ecoregion |
| 1 | 5 | Piney Woods |
| 2 | 3 | Gulf Prairies and Marshes |
| 3 | 13 | Post Oak Savannah |
| 4 | 0 | Blackland Prairies |
| 5 | 0 | Cross Timbers and Marshes |
| 6 | 79 | South Texas Plains |
| 7 | 53 | Edwards Plateau |
| 8 | 22 | Rolling Plains |
| 9 | 3 | High Plains |
| 10 | 85 | Trans-Pecos Mountains and Basins |
Map 1 illustrates the frequency of bat use by ecoregion for concrete bridges and for concrete culverts more than 30 inches (76 cm) tall. It includes both day and night roost locations, because even night use is indicative of where bats most likely would use highway structures by day if suitable roost dimensions were incorporated. Bat use was found significantly more frequently in structures located in the Trans-Pecos Mountains and Basins, the South Texas Plains, and the Edwards Plateau.
Twelve species of bats were found roosting in bridges statewide (see Table 2). Although approximately 70% of observed use involved the Mexican free-tailed bat, highway structures also provided important nursery roosts and migratory stopover sites for other species that simply are far less numerous and form much smaller colonies. No differences in species preferences for roost dimensions were noted. However, species diversity increases with topographical relief and as sampling progressed west and south in the state.
As in the paired comparisons, the statewide evaluations study also found that colonies of up to several hundred bats would occupy crevices as little as four inches deep or those not covered at the top. Bats also sometimes roosted on open culvert walls, though they more frequently used especially long culverts such as those found beneath divided highways. Culverts containing wall or ceiling cavities were also used more often (the deeper the better). Additionally, Mexican free-tailed bats and cave myotis often occupied vertical drain pipes in bridges or culverts when later modifications had closed the tops. In Central Texas, 90% of covered drains were occupied by one or more cave myotis (Myotis velifer).
Concrete bridge substrate temperatures at roosts are strongly buffered from ambient extremes, and within the geographic range of this study, were relatively high. Four bridges, occupied by day-roosting Mexican free-tailed bats in the Austin area, were monitored for 24 hours on 29 June 1994, and the temperature varied by less than 1.9 F (1.1C) among all four. At the highest and lowest ambient air temperatures of the 24-hour cycle, the bridges averaged 4.5F (2.6 C) warmer in the evening and 8.5F (4.9C) degrees warmer at dawn (see Figure 1). This was a typical summer pattern. However, in winter, concrete roost surfaces often remained cooler than ambient temperatures.
Figure 1. Averaged ambient and bridge temperatures taken from four Central Texas bridges over a 24- hour cycle.
Several of the largest colonies of Mexican free-tailed bats reared young in bridges, while most smaller
groups contained bachelors only. Mexican free-tailed bats typically migrate south to Mexico for winter,
though approximately 10% remain in Texas bridges and emerge to feed on warm nights. Both Mexican
free-tails and cave myotis were seen courting and mating in bridges and culverts, the former in March,
the latter in October. In Webb county, a colony of 35,000 cave myotis reared young on the walls of a
large culvert (300 feet long (91 m) and 10 feet tall (3 m)). Nursery colonies of other bat species were not
verified, though most found in Texas highway structures have been documented to use such structures
for rearing young in other states. Mexican free-tailed bats and cave myotis also used bridges and culverts
as migratory stopover sites, as evidenced by fall and spring accumulation of guano deposits. Additional
species probably hibernated in bridges and culverts, but were difficult to detect.
Preference Verification Experiments
Observations from 969 crevice samples demonstrated that the bats preferred widths of 0.56 to 1.25
inches (1.4 to 3.1 cm) (see Figure 2). To verify width as a primary determinant of choice, two 48-inch
(122 cm) sections of a 2-inch wide (5.08 cm) crevice that never had been used were altered by inserting
plywood panels that reduced widths to 0.75 inch (1.9 cm). Bats moved into both previously unoccupied
sections within a month, but not into the unaltered adjacent areas.
Preferences for crevice height above ground were easily tested at the McNeil bridge, because of its
gradual concrete inclines at both ends, which provided heights of 1 to 25 feet (0.3 to 7.6 m) for each
crevice width available. Starting at the abutment (where the bridge rests on the ground) at each end of the
bridge, each crevice was marked off at 5-foot (1.524 m) intervals for 20 feet (6.096 m). Height increased
at approximately 3.2 inches (~8 cm) per linear foot so that within 20 feet of each abutment, the height
had risen to approximately 7 feet (2.1 m). The presence of roosting bats was recorded based on their
position within the 5-foot (6.096 m) intervals. Roosting bats were also recorded for the remainder of each
crevice, which continued to increase to about 25 feet (~8 m) above ground. The bats preferred the highest
available roosts (see Figure 3).
Retrofitting Experiments
Figure 4. Texas Bat-Abode assembly: Plywood end pieces should fit snugly in between the
girders to prevent birds such as pigeons and owls from perching on the abode. Plywood partitions
are 12 to 24 inches deep (30 to 60 cm) and lined with 0.125-inch (0.3 cm) nylon-mesh screening.
Wooden spacers provide appropriate crevice widths of 0.5 to 1.25 inches-wide (1.3 to 3.2 cm.)
End pieces, spacers, and partitions are attached to each other using wood screws. Abodes should
be 10-12 feet above ground.
In 1994 and 1995, six bridges in Travis, Kerr, Kendall, Dewitt, Lavaca, and Gonzales counties were
retrofitted with a total of 17 reinforced, molded concrete panels. Each panel was 36 inches long (91 cm),
24 inches tall (61 cm), and provided one roosting crevice 1.5 inches wide (3.8 cm) at the bottom opening,
tapering to 0.75 inch (1.9 cm) at the top. The panels were attached to vertical concrete surfaces (such as
bent caps), using 0.25-inch (0.6 cm) concrete anchor bolts. Several hundred Mexican free-tailed bats and
one cave myotis occupied the four panels installed in Kerr County by September of 1995, approximately
a year after installation. The remaining panels, most of which were installed in 1995, had not yet been
occupied by the summer of 1996. Investigation of Structures Not Included in Planned Experiments
Experiments to test bat preferences for crevice widths, construction material, and height above ground
were conducted at the McNeil bridge in Williamson County. This is a box-beam style bridge that
includes 28 parallel crevices 26 inches-deep (66 cm) and 700 feet-long (213 m). It is used by a nursery
colony of approximately 750,000 Mexican free-tailed bats. Potential roosting crevices varied from 0.25
to 2.3 inches (0.6 to 5.8 cm) wide and ranged from 1 to 25 feet (0.3 to 7.6 m) above ground, providing an
ideal opportunity to test bat preferences within a single roosting location. Choices according to crevice
width and height were noted, and some crevices were modified to test specific preferences throughout
the spring and summer of 1996.
Box-beam bridges, such as this one located over McNeil Road, provide excellent
opportunities to study bats' roosting preferences.
Figure 2. Crevice width experiments at the McNeil bridge
Figure 3. Crevice height experiments at the McNeil bridge
Four bridges in Travis, Atascosa, and Gillespie Counties were retrofitted with wooden structures, called
Texas Bat-Abodes (see Figure 4). Each abode provides 10 to 24 crevices with 0.75 to 1 inch-wide (1.9 to
2.5 cm) roosting crevices 12 to 36 inches-deep (30 to 91 cm) and roughly 48 inches-long (122 cm). Each
Bat-Abode was built to fit between prestressed concrete girders and steel I-beams of bridges. One of
these was occupied within a month, another within two months. The other two had not yet attracted bats
by the end of their first summer. The first occupied bat-abode was located in a steel I-beam bridge which
provided no suitable roosting crevices, but was located in good habitat where suitable bridge crevices are
typically used. The other occupied bat-abode was between prestressed concrete girders in a bridge where
200 bats already lived in nearby small crevices.
Our study indicates that Mexican free-tailed bats prefer crevices that are between 0.5 and 1.25
inches wide (1.3 to 3.2 cm).
Brian Keeley inspects a Texas Bat-Abode.
Seventy-one bat roosts in highway structures were found through local reports and through personal
searches during travel among study sites. Although these cannot be included in the planned comparisons
and surveys, they do provide substantial insight into the potential for highway structures to serve as
major bat roosts. In fact, despite examination of 1,060 structures during the statewide evaluation, none of
the state's largest Mexican free-tailed bat colonies, including those in the Congress Avenue or McNeil
bridges in Travis and Williamson Counties or the Salado Culvert in Bell County were discovered in the
context of the quantitative studies. Although box-beam bridges were not frequently encountered within
the study sites, they were preferred as day roosts over other bridge designs.
