Can you hear me?Bioacoustics Has Its Moment, But Technology Still Needs Adjustments

When you think of Sonoma County, you think of picturesque valleys and vineyards. But the region is also home to a rich and incredible biodiversity. Soundscapes to Landscapes, the county’s biodiversity monitoring initiative, aims to document just that.

Over the past five years, the initiative has collected an enormous amount of sound data from 1,300 acoustic recorder locations throughout the county, here in California’s wine regions from mid-spring to late summer. The project, run by Sonoma State University, conservation NGO Point Blue Conservation Science, and several other partners, armed citizen volunteers with recorders and worked with private landowners to collect audio. and processed and classified using artificial intelligence technology.

In a Mongabay video interview, Leonardo Salas, quantitative ecologist at Point Blue Conservation Science, said: “Based on that, we can characterize the entire environment.”

This methodology is effective for monitoring ecosystem change and studying wildlife patterns. Before the 2017 California wildfires, Soundscapes to Landscapes installed audio recorders in parks. Examining post-fire data, the team detected a “dominance” of lazuli bunting (Passerina Amoena), a type of songbird that had never been seen or heard in the park before the fire. Initially, citizen scientists monitoring the park thought it was an AI model error. But then he speculated that songbirds may have favored burned areas and flew in after the fires, helping the team understand how the fires changed the park’s ecosystem.

Audio data has been used for decades to monitor, study and protect wildlife. In recent years, bioacoustics has gained prominence as a non-invasive method for studying wildlife. Not only can it be used to study entire landscapes and detect species, as Salas’ team is doing, but it can also be used to understand animal behavior and communication patterns.

Audio recorders can collect large amounts of data, making them more efficient than traditional camera traps and remote tracking methods.Research published in journals Ecology and Evolutionary Methods In 2020, passive acoustic monitoring was found to be a “powerful tool for species monitoring” to detect chimpanzees in the wild (pan troglodytes) is 5x faster than the visual method in Tanzania.Another study published in the journal ecological indicator in 2019 compared acoustic recorders with camera traps and found the advantage of the former to be “a superior detection area 100 to 7,000 times wider than camera traps.”

However, larger coverage areas mean more data to analyze, and healthy data analysis takes effort. Technological innovations such as artificial intelligence and machine learning have made the process easier. But conservationists say the technology still has a long way to go before it can process audio data faster and easier.

Salas says the AI ​​models used in Soundscapes to Landscapes often expose these technical gaps. In the past, models have mistaken the sound of a motorcycle engine for the chirping of pigeons and the chatter of a little girl for the chirping of quail. “There is tremendous capacity to monitor wildlife using voice data, but the technology is not there yet,” he says. “My concern is [whether] It could happen fast enough so that we can start keeping track of how the Earth is changing. ”

“I am skeptical of AI without human ground truth,” says Darren Proppe, who has spent years using audio data to study songbirds in Texas. Human intervention, he says, is necessary not only to spot errors, but also to raise larger issues that automated analysis cannot infer.

“If you’re just looking for birds, mountain lions, or insects, vocalization will tell you,” Proppe, director of the Wild Basin Creative Research Center at St. Edward’s University in Texas, told Mongabay. in a video interview. “But the bigger question is what is missing?

Accessibility to inexpensive real-time monitoring and data transfer is another concern for processing bioacoustic data.

It’s a problem that Daniela Hedwig knows all too well. As director of her project Elephant Listening at Cornell University, she and her team have spent years listening to and recording elephants in the forests of Africa (Loxodonta cyclotis) roaming the rainforests of Central Africa. As a keystone species, elephants play an important role in maintaining and shaping forest structures. The data collected by the project will be passed on to governments, who can use them to locate conservation activities. The project also collects data to help track poaching activity by detecting gunshots in audio. However, coupled with the inefficiency of auto-detectors, the inability to do real-time monitoring makes the process slow and cumbersome.

Data is collected from recorders every four months. Hedwig’s team then spends almost three weeks examining and analyzing the audio. Audio often reaches 8 terabytes, and about 1,100 hours of 4K-quality video of her streaming on Netflix. “The reason is that detectors aren’t perfect. We have to go through each detection and look at it to determine if it was actually a gunshot,” she said. She speaks in her Mongabay video interview.

According to Hedwig, overcoming these challenges along with the inclusion of real-time monitoring will take bioacoustic technology a step further. She says she’s optimistic given how much interest the field has received recently.

“Imagine an anti-poaching unit sitting in a control room, and they can get information about poachers in real time and say, ‘We need to send people to catch them,'” Hedwig said. says Mr. “It’s going to be a big game changer.”

Quote:

Crunchant, A., Borchers, D., Kühl, H., and Piel, A. (2020). Hearing and Seeing: Do Camera Traps and Acoustic Sensors Detect Wild Chimpanzees More Efficiently in Open Habitats? Ecology and Evolutionary Methods, 11(4), 542-552. Doi:

Enari, H., Enari, HS, Okuda, K., Maruyama, T., & Okuda, KN (2019). Evaluating the efficiency of passive acoustic monitoring compared to camera traps in detecting deer and primates. ecological indicators, 98, 753-762. doi:10.1016/j.ecolind.2018.11.062

Related audio from Mongabay’s podcast: Ana Bellalami, research analyst for the Elephant Listening Project, describes the role of forest elephants as a keystone species for tropical forest survival, and plays some recordings of elephant behavior and vocalizations that inform the project’s work. increase. Listen here: