An ultrasonic census: Counting bats under urban bridges

As twilight falls on summer evenings, enormous clouds of bats emerge from their roosts beneath Tucson’s “bat bridges.” The nightly spectacle attracts bat fans from across the Southwest. In the late summer of 2023, that included researchers from the School of Natural Resources and the Environment (SNRE), who used it to test out a new way of calculating urban bat populations using acoustic monitoring.

“One of the basic questions any wildlife manager or conservationist needs to answer is, ‘How many of these animals are there, and is that number changing?’” said Julia Laurenzi, an SNRE alumna and lead author of the study. “Traditional methods like visual counts can vary in accuracy. I started thinking there must be some better way to get population counts for these groups.”

So over the course of six evenings, Laurenzi packed up her video camera and microphone and headed out to two of Tucson’s bat bridges to test whether acoustic monitoring provided an accurate count of the Mexican free-tailed bat populations who use those bridges as their summer home. The study, published in the June 2025 issue of The Wildlife Society Bulletin, showed that not only did it work well for urban bat populations, but also that it has great potential for other types of wildlife monitoring in the future.

A little background

Image

Bats play crucial roles in their ecosystems by pollinating plants, dispersing seeds, or – in the case of Tucson’s Mexican free-tailed bats – by controlling insect populations. According to the U.S. Fish and Wildlife Service, recent studies estimate that insectivorous bats eat enough pests to save American agricultural producers more than $3 billion per year in crop damage and pesticide costs.

Unfortunately, many North American bat populations are in decline due to habitat loss, wind energy installations, climate change and a widespread fungal disease called white-nose syndrome. Conservation efforts are underway across the continent, making it essential that researchers and land managers have accurate methods of measuring abundance.

Mexican free-tailed bats – the most common bat species in the Southwest – are faring better than most, due in part to their ability to live near humans. They’re common in Tucson from April to October, when they congregate in large groups during the summer maternity season. The city’s expansion bridges provide near-ideal roosting conditions, housing tens of thousands of bats at a time.

Laurenzi said that’s why they made such good subjects for her research.

“I’m interested in urban wildlife and urban bat colonies,” she explained. “I wanted to test whether acoustic monitoring could provide accurate counts an urban setting, where there are very different conditions than you find in a natural landscape.”

Clicks versus squeaks

According to Laurenzi, traditional survey methods for bat populations involve visual counts of bats as they sleep or as they emerge from their roosts. Getting those counts can be difficult, often requires expensive equipment, and frequently results in significant over- and underestimates.

“With a visual count, you’re out there with the kind of clicker you might use at a sports arena, clicking each time you count a certain number of bats,” she said. “It’s easy to accidentally double- and triple-count the same bats. Also, since bats emerge before and after sunset, you may need thermal imaging or night-vision equipment. Otherwise, you can only count the bats you see before the sun goes down.”

Her frustrations with existing methods led Laurenzi to a set of acoustic monitoring methods used to count bats as they emerged from their roosts in lava tube caves in New Mexico. Researchers there were able to establish a reliable relationship between the number of emerging bats and the volume of their ultrasonic vocalizations – the more bats, the louder the squeaks.

Laurenzi wondered whether those methods might be adapted to work for urban roosts like Tucson’s bat bridges.

“I had a theory it would work, but there are some major differences between those two environments,” she said. “When bats emerge from bridges, they’re not necessarily emerging from the same small opening and heading in the same direction. Bridges are much longer, so as bats emerge, they tend to shift back and forth, and that movement over a microphone might affect volume, making it harder to get an accurate count.”

Mics, camera, action

Laurenzi tested her theory by visiting the Campbell Avenue bridge and the 22nd Street bridge on several evenings in September and October of 2023. At each bridge, she set up a microphone capable of catching the bats’ high-pitched vocalizations, as well as a video camera to record them as they took flight. Then, her team used computer software to go through the video footage and audio recordings second-by-second, counting bats and measuring volume.

Image

“We averaged the volume per second, and we visually counted the bats that were visible in that frame at the same time,” she said. “It turned out to be a fairly linear relationship between the number of bats and the volume that they produce as they emerge.”

The team’s analysis showed that the model trained at one bridge successfully predicted the number of counted bats at the other bridge, and vice versa, using acoustic data alone.

“It was pretty gratifying to see the relationship hold true across each bridge,” Laurenzi said. “It suggests that this is a robust methodology across different types of roosts.”

Improving research by reducing barriers

Laurenzi explained that part of the appeal of acoustic monitoring is that presents fewer obstacles than other methods of measuring animal populations. She noted that acoustic data can also be collected by nonexperts, and it can even be automated, which would eliminate the need for a researcher to be physically present.

“I wanted to use something that would be accessible to more people,” she said. “The method we used was relatively low-cost, and it’s something that a single person or small group of people could do.”

She hopes that her team’s findings will encourage further research into acoustic monitoring and its use in wildlife management.

“There’s a lot of promise that this could become a standardized method of measuring abundance,” she said. “Our evidence supports the idea that these methods work and could potentially be adapted for different species and habitat types in the future.”