Parrots’ unique voiceprints may help recognize birds in flocks, no matter what they’re saying

Parrots are exceptionally good talkers. They learn new sounds throughout their lives and accumulate an almost unlimited vocal repertoire. At the same time, parrots emit calls so that members of the flock can recognize them individually. This raises the question of how parrot calls can be uniquely identified, given their wide variety.

A study of the northern parrot conducted by the Max Planck Institute for Animal Behavior and the Barcelona Museum of Natural Sciences may provide an answer. Like humans, individuals have unique vocal tones known as voiceprints. This discovery in wild parrots raises the possibility that voiceprints also exist in other species with flexible voices, such as dolphins and bats.

“It makes sense that the cockatoo would have a rudimentary voiceprint,” says Max Planck lead author Simeon Smeele. on paper was announced on Royal Society Open Science. “This is an elegant solution for birds that dynamically change their calls, but still needs to be recognized in very noisy flocks.”

Although humans have a complex and flexible vocal repertoire, we can still recognize each other through voice alone. This is because humans have voiceprints. Our vocal tract leaves a unique signature in the tone of every voice we speak.

Other social animals also use vocal cues to be recognized. For example, in birds, bats, and dolphins, individuals have unique “signature calls” that allow them to be identified by group members. However, the signature call encodes the identity in her only one call type. To date, little evidence exists that animals have unique signatures that underlie all calls produced by individuals. In other words, very few animals are known to have voiceprints.

Smeele, a postdoctoral researcher at the Max Planck Institute for Animal Behavior, was surprised by this. He studies how parrots use their superior vocal abilities to socialize in large groups. Like humans, parrots use their tongues and mouths to modulate their calls. That means “their grunts and shrieks sound much more human than the pretty whistles of songbirds,” he says.

Also, like humans, parrots live in large groups with fluid members. “There could be dozens of birds singing at the same time,” he says. “We need a way to track which people are making what sounds.”

Smeele suspected that parrots, with their correct anatomy and need to navigate complex social lives, may also have evolved voiceprints. To find out, he traveled to Barcelona, ​​home to the largest population of individually marked parrots in the wild. The northern parrot is an invasive species that flocks to the city’s parks in flocks of hundreds.

A monitoring program run by the Barcelona Museum of Natural Sciences has been tracking parrots for 20 years and has so far identified 3,000 individuals, a boon for research into smealets and their vocal recognition.

Smeele and his colleagues were equipped with shotgun microphones and recorded the calls of hundreds of individuals, collecting more than 5,000 calls in total. This was the largest study of individually tagged parrots in the wild to date. Importantly, Smele re-recorded the same person over his two years, revealing how stable the calls were over time.

They then used a series of models to detect how recognizable individuals were within each of the species’ five main call types. Surprisingly, they found that there was great variation in the so-called “contact calls” that birds use to announce their identity. This overturned the long-held hypothesis that contact calls contain stable individual signals and suggested that parrots use something else for individual recognition.

To test whether voiceprints played a role, Smeele turned to machine learning models widely used in human speech recognition. This model uses the properties of the voice to detect the identity of the speaker. They trained the model to recognize individual parrot calls, which were categorized as tonal “tonality.”

Once the model was trained on an individual, we tested whether the model could detect the same individual from a different set of calls that were classified as “grunts” in the sound. The model was able to do this three times better than expected by chance, providing evidence that false lovebirds have voiceprints and, according to Smeele, “individuals may be able to recognize each other no matter what they say.” There is.”

The authors caution that the evidence is still preliminary. “Before we talk about real voiceprints, we need to make sure that the model can reproduce this result when trained with more data from more individuals, and that birds can also recognize this timbre of vocalizations.” says Smere.

The team in Barcelona will complement future experiments and analyzes with ecological studies and equip the parrots with GPS devices to determine the extent to which individuals overlap in wandering areas.

“This could provide insight into the species’ amazing ability to distinguish calls from different individuals,” says Juan Carlos Senar from the Barcelona Museum of Science and Nature.

And finding that parrots have real voiceprints could answer the question of how parrots can be vocal and sociable at the same time, Smere said. Its meaning goes beyond just the parrot. “We hope this discovery will encourage further research to uncover the voiceprints of other social animals that can flexibly change their vocalizations, such as dolphins and bats,” he says.