4.0 Discussion

4.1 Acoustic Deterrents

BD-100 acoustic deterrents did not completely drive bats from structures or prevent bats from roosting in structures. However, closer examination of data revealed a trend toward either more intense bat use of control structures, initial deterrence of bats, or suppression of larger numbers of bats. Monitoring count analysis captured a relative reduction of Brazilian free-tailed bat activity between treatment and control structures in Florida. Notably the treatment structure at site FL-01 remained consistent with an estimate of 3,600 bats prior to acoustic deterrent deactivation while the control structure eventually contained an order of magnitude more bats. Similarly, acoustic deterrents at site FL-02 appeared to temporarily drive bats from the treatment structure with only a gradual increase in number of bats prior to acoustic deterrent deactivation.

In Ohio, acoustic deterrents apparently reduced bat use, particularly in structure pair OH-03. Notably, Ohio data also offer evidence of an apparent migration pulse of big brown bats through the state’s structures in early May 2023.

In California, the treatment structure at site CA-03 treatment contained a maximum of 53 bats; however, due to wildfire, treatment and control structures CA-03 were not sampled after count event 9. Bats trended toward increasing numbers at the site during subsequent checks, and the trend may have continued if structures remained accessible for sampling.

AICc results for the model pooling all states (Ohio, Florida, and California) found treatment versus control served as a better predictor of activity than the type of treatment (acoustic deterrent or physical exclusion). Treatment, regardless of acoustic or physical application, reduced overall bat numbers. However, overall results were potentially influenced by the large numbers of bats in both treatment and control structures in Florida. The California-only GLM analysis revealed acoustic treatment reduced bat numbers, although the result was not significant. Smaller population numbers and differences between structures in Ohio and California potentially confounded observable variation between treatment and control.

Observation of potential migratory pulses of big brown bats in Ohio and Brazilian free-tailed bats at structure CA-05 control raises a bat- and structure-related issue not often addressed under current protocols. Studies for this treatment were specifically designed to follow any regulatory deadlines in states where each study occurred. Such dates are intended to allow addition of exclusions or deterrents before bats return to the site and form maternity colonies. If sites are used only during migration, then exclusion or deterrent efforts should logically occur before migration. Although the role of structures as migratory stopovers is not well studied, several cases exist including rare species using structures as migratory stopovers (Sparks et al. 2019). For the current study, both approaches (physical exclusion and deterrents) were designed around installation of

devices prior to the maternity season. Documentation of big brown bats using structures in Ohio into late fall indicate a possibility of some Ohio structures serving as hibernacula, as indicated by earlier observations of bats chattering in a Kansas bridge in late winter (Sparks et al. 2011). Similarly, late season arrival of Brazilian free-tailed bats at structure CA-05 treatment on 14 August 2024 potentially represents a migrator pulse. Alternatively, a homeless camp was established beneath CA-05 treatment between the time it was reconnoitered in fall 2023 and the initiation of studies in 2024. The camp remained in place until count event 12, when bats were recorded. Thus, bat presence at structure CA-05 is potentially a result of bats returning following departure of people camping under the structure. Regardless of approach, the possibility of migratory bat presence requires consideration when developing a plan to deter or exclude bats.

Acoustic deterrent presence did not prohibit long-term bat use of flyways. Surges in bat activity indicated natural fluctuations except structure pair FL-01 exhibiting a dramatic decline in bat activity following treatment that did not recover. Bats potentially did not use the serviced roadway as the main flyway (where the detector was installed) and initial high numbers possibly indicate disturbance caused by acoustic deterrent activation or natural fluctuations in bat activity post-hibernation (i.e., spring staging). FL-01 offered alternative flyways including a more direct path to a neighboring wetland and complex of retention ponds directly west of the structures.

Field observations during acoustic deterrent installation in Florida indicated bats were noticeably aware of device activation; constant vocalizations of the colony silenced for a few moments before resuming with increased agitation. Changes in vocalization indicated individuals in the colony positioned proximate the acoustic deterrent indeed perceived the signal. Ascertaining the distance of bats responding to acoustic deterrents was not possible. Ohio field observations indicated varying responses by bats roosting under acoustic treatment structures.

In California, species richness was similar between treatment and control structures. However, total activity (total number of acoustic call files) overall was greater at control structures compared to treatment, possibly indicating either a) historically greater numbers of bats were roosting in control structures compared to treatment structures, or b) treatment application of BD-100 deterrents reduced overall bat activity.

Implementation of studies were designed to avoid times when bats were present thus, observing how bats responded to initial deterrent installation was not possible in California. Following installation, observers did not visually or audibly observe any bat responses to BD-100 deterrents. No variations in emergence behavior were observed between bats emerging from treatment or control structures.

4.2 Physical Exclusions

Compared to control structures, physical exclusion treatment structures demonstrated exclusion efficacy supported by no or low numbers of bats after treatment. Monitoring account analysis captured a relative reduction of Brazilian free-tailed bat activity between treatment and control structures in Florida. A greater reduction of observed emerging bats

was recorded in response to physical exclusion application compared to acoustic deterrent treatments. All physical exclusion structure observations in Florida decreased to zero individuals by the end of the weekly monitoring period until treatment failure.

Compared to acoustic deterrent structures, overall observed numbers of bats were lower at Ohio physical exclusion structures. Ohio responses were similar to Florida; however, patterns of bat response to treatment versus control were less dramatic in Ohio than Florida structures because fewer bats were present. Based on GLM results, the difference of big brown bats before and after treatment in Ohio was less influential than presence of Brazilian free-tailed bats in Florida.

4.3 State/Species Differences Discussion

Early project objectives included comparing physical exclusion and acoustic deterrent effectiveness across multiple regions and species necessitating a need to work with variously designed structures containing different bat species.

For the current study, ESI selected structures in Ohio, Florida, and California. Florida structures were all represented by steel I-beam construction, exclusively associated with multi-lane divided highways crossing roads that handle lower traffic volumes and used by large (2,000+) colonies of Brazilian free-tailed bats. Structure pairs were thus nearly identical, sharing the same landscape and used by the same bat species. Discounting a bat’s ability to readily move between structures, such similar structures offered a perfect opportunity for statistical comparison and easier exclusion installation as all structure roosting areas were accessible via extension ladders.

Ohio’s study structures were single structures crossing streams, primarily comprising box beam designs carrying very low traffic volumes (allowing safe access to the deck), and used by smaller numbers of big brown bats, compared to numbers of Brazilian free-tailed bats in Florida structures. Several Ohio structures excluded from study contained protected species including Indiana, little brown, northern long-eared, and/or tricolored bats.

California studies involved various structure types, including culverts, box beams, T-beams, and stringer/multibeam structures. Most California structures were over waterways (although not always flowing with water), with only one structure (CA-02 treatment) over a roadway. Species found in structures varied widely; structures were used by Yuma Myotis, Brazilian free-tailed bats, canyon bats, and big brown bats. Although some structures had high traffic volume overhead (e.g., CA-02C), foot traffic or road traffic (CA-02 treatment) was low.

Substantial differences between Florida and Ohio structures, and their resident bats confound interpretation. Thus, California structures selected for study present a mix of different bat species and services beneath the structures (roadway or waterway). Subsequent statistical analyses following California fieldwork attempted to improve differentiation of variable effects on bat presence, although results from the analysis did not reveal a definite effect of BD-100 acoustic deterrents.

On box beam structures, joints between sections represent primary roosting areas and effectively run the length of the structure. Physical exclusion requires accessing the underside of the structure. Conversely, a rope and pully system is used to suspend acoustic deterrents from structures of any size provided the structure deck is accessible. Thus, most acoustic deterrents in Ohio were installed without use of adhesives such as caulk.

In Florida and Ohio, GLM results indicate big brown bats are more tolerant of acoustic deterrent devices, although potentially an artifact of low count size and limited sample size. In multiple cases, big brown bats were observed roosting within feet of active acoustic deterrents. Similarly, big brown bats did not noticeably respond to acoustic deterrent activation; observations potentially representing a difference in the way acoustic deterrents were perceived by different species. In Florida, some acoustic deterrents were modified to “aim” sound into the spaces used by bats from the side while others were set to “obscure” the roosting area from below. Based on structure design variations, acoustic deterrents on Ohio structures only obscured roosting areas from below. Because big brown bats are notoriously human-associated (Brack et al. 2010), the species is potentially more tolerant of disturbances including acoustic deterrents.

In California, results were mixed as different species occupied structures; canyon bats were the dominant species present at CA-02, CA-03, and CA-04 treatment structures. Numbers of canyon bats emerging at CA-02 initially increased by count event 5, then stabilized to between 10 and 20 bats throughout the study period. At CA-03, numbers of canyon bats increased from zero at count event 1 to 53 by count event 9. At CA-04 treatment, numbers of bats remained low throughout the study period, ranging from 0 to 4 bats. Structures serving as roost locations for these bats comprised T-beam (CA-02T and CA-04T) and box beam (CA-03T). California GLM results indicated only the Yuma Myotis, canyon bat, and Brazilian free-tailed bat did not respond to treatment or control measures.

4.4 Time and Pattern of Emergence

A point of interest involves observation of large numbers of bats in Florida remaining in roosts well after dark (later than typical emergence); notably including all structures with active acoustic deterrents and control structures for both acoustic deterrents and physical exclusion. In structure pairs with acoustic deterrents, fewer bats were consistently observed emerging from the experimental structures compared to control structures despite estimates of similar numbers of bats present. The maximum number of bats observed is derived from a visual estimate, considered an index as opposed to a strict count. Smaller exit counts possibly result from fewer bats in the experimental structures and/or acoustic deterrents potentially affecting bat departure time.

Similarly, moderate to large numbers of bats sometimes remained in roosts after dark in California, limiting the value of emergence counts to estimate populations. At structure CA-04 control, accurate roost counts were difficult because bats tended to occupy cracks above a large concrete pillar during the study. However, during counts, bats emerged after complete darkness and counts were concluded. Similarly, roosting bats were noted

at CA-01 control during count event 11; however, bats never emerged prior to dark. Unlike Florida, observers in California did not note large quantities of bats remaining in the roost at most treatment sites, indicating BD-100 deterrents alone did not influence bat emergence or time of emergence. The noted exception is CA-06 treatment; bats at this structure often remained in roosts (particularly count events 1 through 4), although the trend appeared less persistent as monitoring events continued.

Importantly, the maximum number of bats observed is derived from a visual estimate, considered an index as opposed to a strict count. Smaller exit counts possibly result from observer presence at structures, and because smaller numbers of bats tended to occupy treatment structures.

Acoustic deterrents are designed to jam bat echolocation calls thus, emergence is potentially deferred because bats are less able to detect potential predators outside the roost. Active deterrents possibly prevent bats from “scanning” for predators before exiting. Nocturnal bat behavior is well-established as having an antipredator component (Sparks et al. 2000, Arndt et al. 2018). The predator-detection hypothesis predicts bats are likely to delay emergence.

Possibly, bats are simply unable to scan for obstacles or consider acoustic deterrent installers and structure monitors as potential predators. Presumably, disturbance at a roost would delay emergence at both treatment and control structures shortly following treatment, as would the inability to scan for obstacles at structures with active acoustic deterrents, for example at structure CA-06 treatment. However deterrent effectiveness apparently waned toward the study’s end as bat emergence patterns remained primarily unchanged at CA-06 treatment (excepting count event 11).

No behavioral differences were documented by field biologists comparing behavioral differences between bats emerging from treatment versus control structures. Additionally, bats did not emerge from experimental structures with acoustic deterrents later than from associated control structures.