In the summers it spends in Europe, the great reed warbler fills every hour of daylight with song. At the edges of reed beds, males sing from the tops of the tallest grasses to attract mates, and before sunset, when other birds fall quiet, you can hear their rhythmic calls half a mile away. Singers with a larger repertoire of sounds can snag four or five females in a season, while bad singers stay lonely.

Then, like many other songbirds, the great reed warbler packs up and migrates to sub-Saharan Africa for the winter. With breeding season over, you would think that warblers would spend their African sojourn biding their time, fattening up for next year. But growing up in South Africa, Claire Spottiswoode remembers hearing great reed warblers sing.

That puzzled her. An avid birdwatcher as a child, Spottiswoode, now at the University of Cambridge, recalls listening to the warblers as they ran through sets of staccato syllables on African mornings. “They always reminded me of a deep-throated jazz musician,” she says. “It’s all very deep and loud and constantly changing.” As her interest in birds grew into a research career, she wondered: why were great reed warblers singing in Africa at all?

While singing attracts mates in Europe, it makes much less sense over the winter in Africa. For one, it burns calories. Second, it tempts predators. Those costs might be worth it if it meant attracting female attention, but that’s not happening in winter.

In the past, ornithologists have explained away winter singing in many different species as territorial posturing. Or as an accidental byproduct of the birds building up testosterone to prepare for the breeding season to come. Now, a new study  led by Spottiswoode’s former grad student Marjorie Sorensen argues for a much cuter explanation: The birds are rehearsing for next summer.

Sloshing through tall grasses in Zambia’s wetlands, Sorensen’s team tested the three competing theories. If the birds were singing to protect their territories, playing back a pre-recorded song would provoke aggression, singers wouldn’t want to share the same space, and the songs would be short missives like the ones warblers use to warn away invaders in Europe. None of these predictions came true.

Rejecting the hormone theory was even simpler. If the birds were singing because they were jacked up on testosterone, singers caught in a mist net would show higher testosterone levels than the roughly half of males who didn’t sing—but once again, they didn’t.

Next, they tested the “rehearsal” hypothesis. By pointing a long-distance microphone at singing birds, Sorensen recorded their calls, graphed them, and found that they looked more like summer breeding songs than short territorial warnings. But with one crucial difference.

Though the African songs are just as complex, they follow a slower tempo, and they stretch on and on. “They don’t really have breaks between their songs,” Sorensen says. “They’re just singing, singing, singing.” The songs also seem to meander mid-verse, with the birds switching more rapidly from sound to sound.

Sorensen thinks that might be because they are working through material that isn’t yet crowd-ready, instead of trotting out old standards. In the midst of cutthroat competition on the breeding grounds, birds tend to repeat complex phrases insistently, as many as four of five times. That doesn’t happen in Africa. “If a female is there, and you know you have a very complex syllable, you want to make sure that she hears that syllable,” Sorensen says. But when no one’s listening, flipping through different sounds might be a better way to learn.

Reckoning that African singing might be more important to birds that value song when picking mates, Sorensen’s team also combed through the literature for mentions of winter singing in 57 other migratory birds. They found that two types of birds were the most enthusiastic winter vocalists: those in which males sing more complex songs, and those in which males are drab, and need to use songs instead of colorful plumage to lure females. In both cases, sex comes from songs, and better songs, Sorenson and Spottiswoode argue, come from practice.

Together, the low testosterone measurements, the fact that the birds don’t quarrel over territory, and Sorensen’s analysis of the songs make a persuasive argument that the birds are training themselves, says Dennis Hasselquist at Lund University, who has spent over three decades studying great reed warblers at the height of summer in Sweden. “It’s a logical explanation. But no one has really had any data, so I think this is a great step forward in that respect,” he says.

Listening for reed warbler calls in Zambia’s grasslands. (Marjorie Sorensen)

What would really seal the deal, he says, is to follow individual birds back and forth from Africa and Europe to watch how their sound evolves. It’s a big ask with current tracking devices, though: Spottiswoode calls it “pretty damn near impossible.” But as we wait, Hasselquist thinks that Sorensen’s paper will inspire parallel studies of how practice improves song.

“If you had talked to people experienced in wintering warblers they would probably have been more inclined to go for the other two hypotheses,” he says. “So there will be more people willing to test this idea in other species now.”

Robert Montgomerie, who studies sexual selection in birds at Queen’s University in Ontario, agrees. “I think their study will probably open up this field,” he says. “It’s so provocative and interesting that others will go out and look for it.”

And what makes the work remarkable, Montgomerie thinks, is that it came from a simple observation of nature.

“Occasionally I’m walking to work and I hear a robin singing in the middle of winter,” he says. “You’d never think: why are they doing that?”

Related Video

Transforming ordinary footage of birds and insects into ethereal illustrations of movement