Most people don’t think of turtles as being exceptionally chatty—or even making sounds at all. But research published today in Nature Communications reveals that at least 50 turtle species vocalize—and that several other types of cold-blooded vertebrates previously assumed to be silent do so, too. The finding has broader implications because of the evolutionary history of the species studied. The fact that these supposedly silent species all use sounds to communicate allowed researchers to trace vocalizations back to a common vertebrate ancestor that lived 407 million years ago.

Prior to the study, “most of these species were considered to be mute,” says lead author Gabriel Jorgewich Cohen, a doctoral candidate in evolutionary biology at the University of Zurich. According to the new findings, though, “the sounds that turtles are making have the same evolutionary origin as our own vocal communication.”

Animal communication can take a variety of forms, including visual, tactile and chemical cues. But acoustic signals are probably the most widespread means of conversing among vertebrates. Acoustic communication may include nonvocally produced sounds—a rattlesnake’s rattle, a rabbit’s thump or a cricket’s chirp, for example—or vocal ones.

Vocalizations are especially intriguing to scientists, given their importance across the animal kingdom. Sound leaves no trace in the fossil record, however, so researchers seeking to understand the evolutionary origins of vocalizations have to extrapolate backward using data from current species.

In 2020 scientists compared phylogenies of around 1,800 vocal and nonvocal species and estimated that acoustic communication arose roughly 100 million to 200 million years ago in association with nocturnal activity. The 2020 paper also indicated that this form of communication arose repeatedly and independently in most major vocalizing vertebrate groups, including birds, frogs and mammals—the opposite of the conclusion of Jorgewich Cohen and his colleagues’ new work, which suggests that this behavior did not emerge separately and instead traces back to a common ancestor. Turtles were not categorized as vocal in the earlier study—something that the new paper indicates was a misclassification and that appears to alter the conclusions made from the 2020 analysis.

Tuning in to the secret vocal life of turtles required Jorgewich Cohen to visit wildlife institutions in five countries, where he made recordings of 53 different species. In addition to 50 species of mostly-aquatic turtles, he also examined lungfish, caecilians (a type of limbless amphibian) and tuatara, an evolutionarily unique, lizardlike species endemic to New Zealand.

Jorgewich Cohen collected at least 24 hours of video and acoustic recordings for each species. And in an attempt to capture the breadth of social situations the animals might face, he recorded them in both isolation and various groupings: females only, males only, mixed sex couples and individuals of different ages. After collecting the recordings, Jorgewich Cohen undertook the painstaking task of sifting through more than 1,000 hours of audio.

He found that every species he recorded produced some sort of sound, including quacks, bleeps, grunts, crackles, toots and snorts. Many had a repertoire of different noises that changed depending on the social situation. The South American wood turtle, for example, produced an incredible 30 different vocalizations. Some species were very vocal, whereas others produced “just one to two sounds every 10 hours,” Jorgewich Cohen says. For almost all of them, the research represents the first time their vocalizations are known to have been recorded.

Listen to some chatty South American river turtles (Podocnemis expansa) out for a group swim. Credit: Camila Ferrara

Based on the evolutionary history of the turtle species included in the new study, the researchers calculated that vocal communication must have appeared 200 million or more years earlier than previously thought. The findings also imply that this behavior did not arise independently in different animal groups but was instead conserved over time, dating back to a common vertebrate ancestor that lived at least 407 million years ago.

Tecumseh Fitch, an evolutionary biologist at the University of Vienna, who reviewed the new study but was not directly involved in the work, says it is “an important contribution, both because vocalizations from many important species are analyzed for the first time and because they lead to a convincing argument” that this behavior was present in early vertebrate ancestors.

“Reconstructing the evolution of behaviors is always a daunting task, and acoustic communication even more,” adds Darcy Kelley, a neurobiologist at Columbia University, who was not involved in the work. One reason the paper is exciting, she says, is because it “gives us a lot more species to study in order to understand which parts of the neural circuitry that supports vocal communication are ancient and conserved across long evolutionary periods and which are newer.”

John Wiens, an evolutionary biologist at University of Arizona and co-author of the 2020 study, points out, however, that Jorgewich Cohen and his colleagues did not prove that the turtles and other species they recorded “are actually using those sounds to communicate with each other.” That, he says, “seems like a big omission.”

“A major focus of their paper seemed to be on reanalyzing our data and coming to a different conclusion,” Wiens adds.

Jorgewich Cohen acknowledges that follow-up studies will have to be conducted on individual species to explore their full repertoire of sounds and confirm those sounds’ meanings.

Ultimately, the new paper is “a great example that sometimes you do have to actually go and look—or rather listen—rather than accept a standard in a particular field,” adds Irene Ballagh, a zoologist at the University of British Columbia, who was not involved in the work.

The new findings struck a personal note for Ballagh, who grew up in New Zealand. She recalls hearing about tuatara vocalizations from her mother and others who were familiar with those animals. Yet throughout her career, she encountered “quite definitive statements” by much of the scientific establishment that tuatara do not vocalize. The researchers who made those assumptions were all based in the Northern Hemisphere, she points out, and “were not ever thinking to ask local people studying wild populations” what they might have observed.

“I would really love to see more people following up with more work connecting local and Indigenous sources of knowledge about the potential vocalizations for species groups that are still listed as an ‘absence of data’ in this paper,” Ballagh says. “I think the data might already be out there in some form if we just start to think more carefully about who we should be listening to.”