From chirps to complex melodies, the songs of birds can vary greatly, resembling a grand orchestra.
However, ovenbirds — stout, terrestrial warblers — possess a very limited selection of songs.
The male ovenbird sings only one tune: a quick melody often referred to as teacher-teacher-teacher.
While to the untrained ear, all ovenbirds may seem alike, each bird has its own unique interpretation of that song.
Currently, researchers are employing artificial intelligence to monitor the well-being of ovenbird populations based solely on their distinctive songs.
A study released this month in the journal Nature Methods utilized AI technology to recognize individual ovenbirds from recordings collected in Alberta and Pennsylvania.
A Singular Focus
The ovenbird’s “specific” song pattern made it an ideal candidate for this research, according to lead researcher Sam Lapp, a postdoctoral fellow at the University of Colorado.
“They are notably committed to performing one very particular song,” Lapp shared in a discussion with CBC News.
“Many birds develop a vocal signature over their lifetime, often integrating various songs into what we refer to as their repertoire.
“In contrast, the ovenbird possesses the unique ability to maintain one very stable song throughout its life, only performing a single version of it.”
Ovenbirds produce their main song in the spring, primarily to establish territory and attract mates, and this song remains consistent throughout their lifespan.
Under Lapp’s direction during his tenure at the Department of Biological Sciences at the University of Pittsburgh, the research team analyzed recordings from nearly 130 locations in Pennsylvania over four years.
They trained an AI model to detect unique features in the songs and group similar sounds together, based on the assumption that each bird returns to the same area each spring.
The team examined 9,072 hours of audio to identify 405 individual birds. They then validated the model with an additional set of recordings gathered in Alberta’s mixed wood and boreal forests during the 2016 spring breeding season.
Lapp noted that the model could accurately associate songs with individual birds with up to 96 percent precision.
This method is significantly less labor-intensive compared to leg banding, which involves capturing birds, carefully attaching markers to their legs, and then trying to catch them again the following spring.
The research idea originated from a partnership with biologists in Alberta, who were exploring methods to locate birds across extensive areas using a grid of microphones.
“Somewhat Amazing”
According to Lapp, human listeners cannot discern individual ovenbirds in real-time due to the rapid nature of their vocalizations.
However, by using spectrographs — graphical representations of sound — the unique songs of each bird become identifiable, Lapp explained.
“I began to wonder if my eyes could detect these patterns, could we train a machine to recognize them and automate individual identification?
“Even if you’re passionate about birds and focused on identifying species, it’s astonishing and remarkable that this depth of information exists that we’ve simply overlooked… yet it’s there if you pay close attention.”
Erin Bayne, a biology professor at the University of Alberta and a co-author of the study, has been monitoring ovenbirds for many years and is familiar with the challenges of yearly population assessments.
During his doctoral studies, Bayne dedicated countless hours to tracking banded birds, to the extent that he could distinguish individuals based on their unique vocal differences.
“What that bird is saying over there differs slightly from what that other bird is saying,” he remarked. “I could recognize each one.
“They exhibit remarkable courtesy while singing. It’s amusing, really. One bird will sing, and then the other will patiently wait for it to finish before immediately responding with its own song.”
Feathered Friends Communicate
Bayne stated that the findings indicate that bird communication is much more intricate than many people realize, and that exchanges are occurring, even if we don’t comprehend the dialogue.
Just as humans have unique speech patterns, researchers are discovering that certain animals also develop distinct voices of their own, Bayne noted.
“In many respects, it’s not surprising that individual birds or animals possess detectable differences. The more we investigate, the clearer it will become that this is indeed the case.”
While the model has its constraints, the study’s authors hope it will offer a less invasive method to assess the health of songbird populations simply by eavesdropping.
The AI system currently struggles to effectively track the varied calls of female ovenbirds or the basic sounds produced by young birds.
The present model would not be suitable for species like magpies and mockingbirds that frequently mimic new sounds or have expansive song repertoires.
It is best suited for birds that consistently sing an unchanging song throughout their lives — about 30 percent of songbird species — along with certain bat and owl species.
The researchers have made the program code available on open-source platforms, aiming to assist other scientists in monitoring the survival of various songbird populations or even different species.
Richard Hedley, a statistical ecologist from the Alberta Biodiversity Monitoring Institute who did not participate in the study, emphasized its remarkable potential.
While AI acoustic monitoring has been employed to track birds at the species level for some time, demonstrating its capability to differentiate between individual birds is a significant achievement, he stated.
There exists a vast amount of existing acoustic data that could be utilized in innovative ways, he added.
“Catching a single bird can take many hours. Therefore, the prospect of achieving this by merely placing a microphone in the bird’s territory is very encouraging.”
Tag: rewrite-pending