3.0 Results

3.1 Implementation and Field Survey Applications

Treatment and control applications were executed prior to 15 April and concluded by 15 August, the proposed deadlines for implementation and end of the 2023 and 2024 field season. All structures (excepting structures SFN 110029, 110017, 700066, 700043, 700141, 50-0206, 46-0013, 50-0236, 50-0280, and 42-0016) were surveyed on 12 monitoring count dates while treatments were applied and are detailed below.

3.1.1 Acoustic Deterrents

Acoustic treatment and control applications were completed 16 to 28 March 2023 in Florida, 3 to 11 April 2023 in Ohio, and 18 March to 12 April 2024 in California.

In Florida, control and treatment structures were adjacent structures where a multi-lane divided highway crossed a smaller road. Similarly, California structures at pair CA-01 were adjacent structures beneath highway 99. Thus, each treatment structure exhibited structure characteristics nearly identical to its corresponding control structure. Proximity of treatment and control structures likely influences bat movement between these structures.

However, Ohio and remaining California treatment and control structures were not collocated (i.e., possessed similar but not identical traits and were not geographically adjacent) thus pairwise comparisons were less meaningful. In all three states, counts of individuals varied in response to application of acoustic deterrent treatments and across time.

3.1.2 Physical Exclusion

Physical exclusion treatment and control applications were completed 1 to 24 March 2023 in Florida and 3 to 14 April 2023 in Ohio. In Florida, control and treatment structures were adjacent structures where a multi-lane divided highway crossed a smaller road. Thus, each treatment structure exhibited structure characteristics nearly identical to its corresponding control structure. Proximity of treatment and control structures likely influenced bat movement between Florida structures. Counts of individuals exhibited similar negative response patterns to application of physical exclusions treatments. Unlike Florida, Ohio treatment and control structures were not collocated (i.e., possessed similar but not identical traits and were not geographically adjacent) thus pairwise comparisons were less meaningful. Counts of individuals varied in response to application of physical exclusion treatments and over time but generally showed a greater reduction of bats observed in treatment versus Ohio control structures.

Summer storms during mid-June 2023 caused rainwater from the road deck to flow into excluded spaces of Florida structures and partially or completely wash-out exclusion materials. Because experimental structures are immediately adjacent control structures

(housing colonies of more than 1,000 bats), bats reoccupied experimental structures before exclusion failures were detected during monthly checks. In compliance with Florida Fish and Wildlife Conservation Commission policy, no exclusion repairs were made during the maternity season at Florida structures. Any exclusion material that fell from exclusion structures was removed from the site and structures were monitored throughout the monitoring period (15 August 2023) to document bat reoccupation of the structure. Physical exclusion structures in Ohio were regularly maintained throughout the 2023 field season to ensure exclusion efficacy.

3.2 2023 Bat Activity (Ohio and Florida)

The difference between initial and final survey counts varied considerably across Florida and Ohio structures. In 2023, more bats were present at initial surveys in Florida than in Ohio for both acoustic and physical treatments and controls. Acoustic and physical treatment structures demonstrated a mean reduction of bats after treatment application in both Florida and Ohio. In Florida the mean number of bats was reduced after installation of both acoustic deterrents and controls, but acoustic controls demonstrated a larger reduction in bats compared to deterrents. However, physical treatments in Florida resulted in a greater mean reduction of bats compared to controls. Few bats were present in Ohio bridges when treatments were installed. Low initial counts of bats in Ohio resulted in an overall lower mean difference in bat presence for both acoustic deterrents and physical exclusions installations compared to Florida. Over time, the number of bats in control bridges (both acoustic and physical) increased while the mean number of bats in bridges that received treatments decreased.

In summary, physical exclusion reduced bat activity in Florida and Ohio, while acoustic deterrents only showed a greater mean reduction of bats compared to controls in Ohio.

3.3 2024 Bat Activity (California)

The difference between initial and final survey counts varied considerably across California structures. In 2024, more bats were present at initial surveys in control structures than in treatment structures in California. The number of bats present at control structures during installation varied from zero bats to 319 bats. Bats continued to occupy all control structures where bats were observed prior to the study, although numbers varied. Notably, counting at CA-03C (SFN 50-0280) was precluded by wildfire following count event 9.

No bats were present when deterrents were applied in California; however, bats arrived at all but one structure during the study period (no bats noted roosting in or emerging from the structure; CA-01 treatment [SFN 46-0031R]). A maximum of 121 bats were observed emerging from treatment structure CA-06T (SFN 46-0056L); many bats observed were noted as roosting near acoustic deterrents and flying in the vicinity of deterrents upon emergence. Structure CA-03T (SFN 50-0206) trended toward increasing bat roosting and emergence activity up to count event 9 when wildfire moving through the area precluded further inspections; the same fire also affected CA-03C (SFN 50-0280). Unlike control structures, bats gradually occupied treatment structures, primarily reaching

the highest overall numbers during monthly monitoring periods, possibly indicating bats became acclimated to deterrents over time.

3.4 Analysis

3.4.1 Monitoring Counts

When examined across all three states, variables “treatment versus control” and “state” best predicted bat presence at structures as determined by the AICc. Treatment (regardless of acoustic or physical application), decreased bat presence when compared to control applications. The variable “state” likely represents an interplay among number of bats, species of bats, habitat beneath structure, and geographic location. State designation of “Florida” is associated with a larger reduction in number of bats between initial and final count events, as determined by the GLM. At installation, overall Florida bat numbers were much higher compared to occupancy in both California and Ohio structures. A greater variety of bat species dominated California structures, including Brazilian free-tailed bats, big brown bats, canyon bats, and Yuma Myotis.

Assessing California alone, all variables together (“treatment vs. control”, “species”, “structure length”, “structure width”, “service type”, and “structure type”) best predicted bat presence as determined by the AICc. However, the resulting GLM did not reveal any significant terms determining bat presence at structures. Thus, the number of bats present at a structure was predicted by a mix of variables as opposed to functional deterrent presence; however, because the resulting model was a better fit than an identical model without treatment suggests there is biological value to treatment applications.

Ascertaining which variable was attributable for variations in Florida and Ohio bat presence is not possible.

3.4.2 Acoustic Monitoring

By automatic classification, acoustic monitoring of structure flyways documented calls consistent with six species in Florida (big brown, eastern red, northern yellow, southeastern, tricolored, and Brazilian free-tailed bats), seven species in Ohio (big brown, eastern red, hoary, silver-haired, little brown, Indiana, and tricolored bats), and 15 species in California (pallid, Townsend’s big-eared, big brown, California mastiff, eastern/western red, hoary, silver-haired, western yellow, California, western small-footed, western long-eared, fringed, Yuma, canyon, and Brazilian free-tailed bats).

In 2023, detectors on Florida flyways consistently recorded greater numbers of acoustic files compared to those on Ohio flyways. General trends of bat acoustic activity along flyways demonstrated at least one spring surge of activity followed by decreased activity in both Florida and Ohio. Flyways predominantly demonstrated a resurgence of bat activity post-treatment, excepting structure pair FL-01 flyway, but resurgence magnitude and timing varied among flyways. Species richness patterns mirrored bat activity patterns.

In 2024, general trends of bat acoustic activity along flyways in California varied between and within treatment types. Flyways either demonstrated a resurgence of bat activity post-treatment, steady activity throughout the duration, or a steady decline of bat activity through summer 2024; resurgence magnitude and timing varied among flyways. Species richness patterns mirrored bat activity patterns. Bat presence in flyways was not always indicative of bat species present at structures. For example, the silver-haired bat, was recorded at the CA-01 treatment flyway, although the species was never observed roosting or emerging from structures during 2024 surveys, thus there was no evidence that activity along flyways was impacted by acoustic deterrents.

3.5 Financial Comparisons

3.5.1 Initial Installation

3.5.1.1 Acoustic Deterrents

Some costs are standardized across the three states such as individual acoustic deterrent costs ($495.00/acoustic deterrent at the time of analysis). Acoustic deterrents were remarkably weather resistant, but five acoustic deterrents were stolen or destroyed by floods, rain, and high winds. Powering each acoustic deterrent posed the greatest challenge. Although not applied during field studies, the preferred option is to use 120-volt power routed through a power inverter available from the manufacturer. Inverters cost $75.00 per unit and one unit can power up to three acoustic deterrents. Power inverters were not used as units are designed to plug into power outlets (AC power sources), unavailable for field deployment. D-cell battery packs used in Florida and Ohio were inexpensive ($10.00 including hardware such as wire connectors and tape). Each battery pack required eight batteries at a cost of $1.50 per battery ($12.00 total cost). A timer was wired into each battery box to ensure conservation of power; acoustic deterrents were activated by timers when bats were active ($20.00 total cost).

Costs associated with recurrent changing of D-batteries and frequent potential for acoustic deterrent failure prompted a switch to rechargeable lead acid battery packs. Battery packs cost approximately $100.00. Each battery pack was linked to a single acoustic deterrent. In Florida, battery packs, including timer, were used without support of a solar backup and required changing once per week. Solar panels used in Ohio and California cost $150.00 per unit with each unit powering a single lead acid battery pack and timer. Solar panels used in California eliminated the need for maintenance visits between emergence counts. Battery packs and electrical connections to the solar panel were contained in a plastic ammunition box costing approximately $15.00 each. The plastic ammunition boxes were lockable, water-tight, available at most sporting goods stores, and provided protection from vandals, wildlife, and rain.

When a structure’s sides did not accommodate fastening solar panels, structures were erected using 6 feet (1.8 m) of PVC pipe (approximately $15.00), four PVC elbows (approximately $6.00 each), and a forked metal post used to hold the PVC pipe in place (approximately $6.00). Each PVC setup can hold two solar panel setups. Other materials used for acoustic deterrents included low-cost items (e.g., plywood, spring-loaded binder

clips, and silicone adhesive) typically at $25.00 per structure. Each Ohio structure also required approximately 250 feet (76.2 m) of paracord, priced at $35.00 per roll.

Control structures in California used empty plastic water bottles spray-painted black. One can of spray paint cost approximately $7.00; three cans of spray paint were used to coat 22 sham deterrents. Thus, the cost for one sham deterrent was approximately $1.00.

Acoustic attachment methods differed across the three states and resulted in differences in installation hours. In Florida, acoustic deterrents were mounted along the side of the structure or were mounted in the center of each section of the structure. Acoustic deterrent installation averaged 35 staff hours per structure with a minimum of 19 hours at FL-02 and a maximum of 54.4 hours at FL-01.

In Ohio, acoustic deterrents were mounted to the bottom of each structure using a series of pulley ropes. Acoustic deterrent installation averaged 29 staff hours per structure with a minimum of 19.25 hours at OH-02 and a maximum of 35.25 hours at OH-01.

In California, multiple mounting methods were used depending on the structure’s physical properties. These methods included mounting deterrents along the side of the structure, fastening deterrents beneath weep holes or cracks, mounting deterrents to PVC pipes on the ground, or hanging deterrents on structure walls. Acoustic deterrent installation averaged 11.17 staff hours per structure with a minimum of 8 hours at CA-04 and a maximum of 15 hours at CA-02.

Structures in Florida were typically larger and more complicated than in Ohio. Because the bats were roosting in gaps formed by sections of the structure that sat on the pillars, biologists had to use ladders and install detectors in each section of the structure and along the sides. This required more staff-time and more deterrents per structure. For example, Florida structures required multiple deterrents per roosting area. Conversely, most Ohio structures were narrow, box beam structures. In these cases, a centrally located deterrent could cover the entire width of the structure. Crews in Ohio also became highly skilled at installing deterrents using the rope systems and traffic was low enough to allow biologists to safely work atop the deck.

In California, acoustic deterrent installation averaged 2.5 hours for control structures and 8.6 hours for treatment structures. Two treatment structures (CA-05T and CA-02T) required more complicated installation procedures with deterrents mounted beneath roosts using long PVC poles placed atop metal U-post stakes with anchor plates in the ground and thus took longer compared to installations at other treatment structures. In California, many structures were “smooth” and lacked sufficient area or components (e.g., slats) for attaching solar panels thus requiring more creative measures for deterrent application (e.g., use of PVC pipe to project deterrents into roosts, or pulley systems to raise deterrents to roosts). Short structures (i.e., standing height or lower) with weep holes required the least installation time as the BD-100/sham deterrents were easily applied using binder clips wedged into cracks and crevices. Taller structures required access via ladders or other methods (e.g., hanging deterrents from walls) to install deterrents.

3.5.1.2 Physical Exclusion

Materials required for physical exclusion were relatively inexpensive with a typical structure in Florida requiring approximately $100.00 in supplies; the absence of bats in structures at the time of installation reduced costs to $75.00 per structure.

As with acoustic deterrents in Florida, exclusion materials were focused on gaps in the deck above pillars where bats roost. Because bats were already roosting in crevices, foam tubes were cut to fit gap sections; spray foam and silicone sealant were then applied to keep the foam tubes in place. One-way excluder devices were also installed allowing bats to leave but not reenter the structure. Exclusion material installation averaged 31.5 staff hours per structure with a minimum of 27.5 hours at FL-04 and a maximum of 36.5 hours at FL-05.

In Ohio, physical exclusions were installed for all gaps at least 0.5 inches (1.27 cm) wide. Biologists visually confirmed bats were absent and used spray foam to seal each crack. Physical exclusion installation averaged 10.75 staff hours per structure with a minimum of 7.75 hours at OH-05 and a maximum of 16.5 hours at OH-04.

As noted above, Florida structures were typically larger and required ladders to access. Large numbers of bats were present in all sections of all structures. Thus, biologists had to exercise care to trim and insert foam backer rod into the gap before foam or silicone was used to seal the backer rod in place.

Conversely, the smaller structures in Ohio contained relatively few bats. In many cases, a biologist was able to verify that no bats were roosting in a crevice. When that occurred, expanding foam could be applied from the ground greatly reducing installation times.

3.5.2 Maintenance

3.5.2.1 Acoustic Deterrents

In 2023, most maintenance costs for acoustic deterrents were associated with replacing batteries. The D-cell system is only suitable for short-term deployments. Time to replace batteries was typically 30 staff-minutes per acoustic deterrent in Florida, primarily based on the need to use a ladder (requiring a climber and spotter) to access the battery pack. The pulley system used in Ohio was more efficient than ladders, allowing a single biologist to change batteries in 10 minutes. Ohio structures also required fewer acoustic deterrents because line-of-site existed across most of the structures and, thus, structures had fewer obstacles which block an acoustic deterrent’s ultrasonic acoustic field. As noted above, each battery pack required eight batteries at a cost of $1.50 per battery ($12.00 in cost). D-cell batteries required replacement every two to three days.

Lead acid batteries required a similar time to swap out, but solar panels in Ohio and California eliminated the cost as batteries did not require charging after deployment of solar panels and required less maintenance. Long-term deployment requires a skilled field crew four hours every six weeks to replace support ropes.

In 2024, costs for acoustic deterrent maintenance are associated with replacing entire setups. Following application of both control and treatment deterrents, no issues associated with setup maintenance following installation arose (e.g., solar panels operated, batteries remained charged, deterrents continued to operate). However, where solar panels and/or deterrents were stolen, a replacement set of solar panels, deterrents, and accessories was required, additional to costs associated with personnel hours to install the setup.

3.5.2.2 Physical Exclusion

Following strong storms, each Florida structure required approximately four staff hours to reinforce physical exclusions. Each visit required approximately $15.00 in supplies. Ohio structure maintenance required approximately one staff hour and $15.00 in supplies each month and was typically paired with monitoring efforts.

3.5.3 Monitoring

Monitoring cost for all three states were similar. A typical emergence count required three staff hours per structure. Adding acoustic detectors required an additional 15 minutes per site for deployment, but detectors also required recovery. Depending on application, an additional site visit was potentially warranted. Using automated analysis such as KPro reduces costs to approximately 15 minutes per night of study.