How Animals Develop Regional Accents

The song of the hermit thrush sounds a bit like an orchestra warming up: the experimental trill of a flautist, a violinist testing his vibrato on a few high notes, all just beginning to suggest a larger melody. This undulating song can be heard throughout North America in the summer. But moving east to west, a keen listener might notice that the bird’s repertoire changes: lower notes introduce each song, like a viola has been added to the mix, and the trills that follow are shorter.

For biologists, this change is the anthem of evolution.

One million years ago, the hermit thrush population was separated, likely by a large glacier that cleaved North America in half. A recent study shows that this separation has had an audible effect: today, the eastern and western hermit thrush populations sing their songs differently.

As in many songbirds, male hermit thrush are the singers of the species, learning their songs as chicks from other adult males—fathers, uncles, brothers—around their nest. But that learning is imperfect, and every new singer introduces small differences to the song, the way changes appear in a passed phrase during a game of Telephone.

With the eastern and western populations separated, those differences couldn’t be shared, explains lead author Sean Roach. The hermit thrush songs that we hear today have diverged—as though they have passed through two concurrent and very long games of Telephone.

“If there are sufficiently large song differences between populations, it can get to the point where they don’t recognize what they’re hearing as being from the same species,” Roach says.

Above is the song of an eastern-type hermit thrush, recorded in New Hampshire. Listen for the wide range of introductory note frequencies used.

Above, the song of a western-type hermit thrush, recorded in Baja California. Listen for longer introductory notes in a more limited frequency, with shorter phrases post-introduction.

Over time, the accumulation of differences can create a feedback loop: as two groups of a species begin to sound more and more unlike, they are less likely to recognize each other as potential mates. This makes the groups less likely to share both genes and songs, pushing their songs to become even more dissimilar.

In this way, sounds can be both a signal and a driver of evolution, making vocalizations rich territory for biologists looking to spot evolution in action.

On the islands of Hawaii, roughly four new species appear in the cricket genus Laupala every million years—an explosion, by evolutionary standards, and the highest documented for arthropods, which usually see new species more on the order of 0.16 per million years. The driving force behind this evolutionary outbreak? Males in different species sing their pulsing courtship song at a different rate, so researchers theorize that the songs have something to do with it—that over time, female preference for different songs has pushed the group to continually diverge.

Differences in sounds can also help species to share without out-competing each other. Like many family members, five bat species from the European genus Myotis share a lot—they look physically similar, live in the same areas, and have similar strategies for hunting. Yet each species of bat has a slightly different echolocation signal, allowing each species to target a slightly different prey.

Below are the differing echolocation calls of three species of the Myotis bat genus, time-expanded to be audible to human ears. These sounds were recorded by Dr. Stuart Parsons at the University of Bristol.

In studying the hermit thrush, Roach saw that the division in song structure matched with genetic and physical differences between the two groups. Geologic data, as well as evidence of similar splits in other birds, then made it possible for him to point to the glacial event as the cause of this split and the resulting vocal differences. Yet in Hawaii’s crickets, Europe’s bats, and many other vocal species, it can be extremely difficult to tell what came first—whether vocalizations are a cause or a symptom of speciation.

This issue is especially apparent in studies of another well-known animal singer: whales.

“It’s kind of like the chicken and egg: is the song separating [whales] into discrete units, or is it a symptom of being discrete, a thing that evolved after they became separate?” says Erin Oleson, lead scientist in the Cetacean Research Program at the National Oceanographic and Atmospheric Administration.

For many species, we still can’t say. Acoustic data from whales can be highly variable, changing even over the course of a season as animals learn new songs from each other.

In humpback whales, songs—sung by males to attract mates—are exchanged seasonally, with new trends migrating slowly between populations. The song below was recorded by Dr. Ellen Garland and other researchers off of Eastern Australia in 2009. Because they change constantly, humpback songs don’t give the same ancient evolutionary cues as other animals’—but they do hint at a complex and rich social structure that spans ocean basins.

Below is a humpback song from the next year, recorded by the same researchers off New Caledonia (an island roughly 1000 miles east of Australia). Note the theme between 0:22 and 0:40, which can be found in the 2009 Australian song between 0:20 and 0:40. One theory is that these songs might be shared at their feeding grounds in Antarctica.

Additionally, you need genetic data to show a change in tune has an evolutionary basis, which can be difficult to come by—though that doesn’t mean some aren’t trying. Amy Van Cise, a PhD candidate at the Scripps Institution of Oceanography, has been analyzing archived genetic data, specimens, and call recordings for short-finned pilot whales living on opposite sides of the Pacific ocean.

“I was really interested in how things like social structure and acoustic structure can be linked with genetic structure, and how those things might affect evolution on a longer term scale,” Van Cise says.

Her work shows that the two groups, known as the Shiho (eastern) and Naisa (western) type, are genetically, physically and acoustically distinct, to the point that they could be classified as separate sub-species. With more study, Van Cise hopes to pin down exactly when the two groups split, in order to learn if their differing songs came before or after the division.

The “canary-like” song of the short-finned pilot whale. The differences between the eastern and western sub-species can’t be caught by a human ear; Van Cise used an audio processing program in order to analyze them. (Courtesy of Amy Van Cise)

Perhaps one of the most exciting results of analyzing animal sounds is catching this division in action.

“Genetics is the common way of IDing species, but it’s a very long-term process,” says Oleson. “Acoustics is much more plastic. We can detect changes on a much shorter time scale than you could with genetics or morphology, which could take centuries to propagate.”

Using audio analysis programs on hundreds of recordings, Roach may have observed this process within a sub-group of hermit thrush in western Canada. Birds at higher altitudes in the Canadian rockies had songs distinct from those at lower altitudes. This could be a reaction to their environment: low-altitude birds need to sing at lower frequencies to be heard through dense trees, but with less forest around them, high-altitude birds can sing much higher.

For Roach, these fine-grained differences have much larger implications than birding wonkery. In a time where habitat and biodiversity is dwindling worldwide, he sees it as vital to understand the diversity between species in order to maintain it. But another big reason he continues this work is a personal one.

“Most birdsong research involves getting up at four, four-thirty in the morning, which is a time most people hear and cringe and look disturbed,” he says. “But it’s wonderful to be out there. There’s nobody else out there. It’s just me and the birds singing.”