As any serious birder will tell you, bird songs and calls are often the best – and sometimes the only – way to tell bird species apart in the field. In the central Great Plains, for example, Eastern and Western Meadowlarks (Sturnella magna and S. neglecta) look nearly identical. But as soon as they sing or call, they can be told apart easily.

Because male birds often use song to attract females, biologists have historically thought that natural selection should favor divergent songs in species that occur together (i.e., sympatric species). Males with songs that differ substantially from those of closely related sympatric species would be less likely to pair accidentally with females of another species. Since hybrid offspring often suffer fitness costs relative to offspring whose parents belong to the same species, males with the most unambiguous songs would, on average, have the highest-quality offspring. Therefore, we expect closely related, sympatric bird species to evolve quite different songs – like the meadowlarks.

But this is not always the pattern, and one fascinating exception was described recently from the Amazon basin: a pair of Amazonian antbird species in the genus Hypocnemis. These birds are being studied by evolutionary biologists Joseph Tobias and Nathalie Seddon at Oxford University’s Edward Grey Institute of Field Ornithology.

In a paper published last year in the journal Evolution, Tobias and Seddon describe geographic variation in the songs, calls, and plumage of the sympatric Peruvian Warbling Antbird (Hypocnemis peruviana) and Yellow-breasted Warbling Antbird (Hypocnemis subflava). Since these common names are a bit of a mouthful, I’ll stick with the Latin: peruviana and subflava.

In the late 1960s, ecologist Martin Cody proposed an evolutionary mechanism by which competing species might converge in color or song. Many birds are territorial – that is, they defend an area against other individuals of their own species. Some species also treat members of other species as territorial rivals. Why would they do this? Cody argued that this “interspecific territoriality” would occur for the same basic reason that territoriality made sense within species… because exclusive access to a small space is often better than non-exclusive access to a large space.

Imagine that you’re a bird provisioning your young. The ravenous chicks are in a nest in the center of your territory, so in order to feed them, you have to forage in your territory and transport food back to the nest. Now, suppose you live alongside another bird of your species. It feeds its chicks exactly the same kind of food that you feed yours. If you both forage in the same large area, you will each have access to half of the total available food. If instead you divide the large space in half, and each restrict your foraging to your own half (i.e., your “territory”) you also have access to half of the food. Same difference, right?

Not quite. In the second scenario, because your territory is smaller, the distance traveled to and from the nest is reduced. Moving around the territory can be costly – it requires time and energy, and it exposes you to predators. So, assuming that negotiating the territory boundaries with your rival is less costly than traveling around a large shared space, the territorial strategy is better for everyone than the space-sharing strategy.

Now, suppose your rival doesn’t feed its young exactly the same things that you do (e.g., your neighbor belongs to a different, but closely related species). In this case, the payoff of sharing a larger territory is greater than before. Why? Because some of the food that your rival consumes is food that you wouldn’t be using anyway. Since the payoff of the space-sharing strategy is higher when your neighbor belongs to another species, the alternative strategy – interspecific territoriality – is less common than intraspecific, or within-species, territoriality.

Nevertheless, if the costs of traveling around the territory are high, and the costs of negotiating the territory boundaries are relatively low, then defending exclusive “interspecific territories” may still be the best solution for all parties – especially if the two species use fairly similar resources. Cody argued that in such situations, natural selection would drive convergent evolution in songs or other traits used to recognize competitors. Recognition across species boundaries, he reasoned, would enable effective territorial communication. In 2009, we published a basic computer model that demonstrated that Cody’s verbal predictions could be replicated in a simulated environment (Grether et al. 2009).

Unfortunately, empirical evidence for convergent evolution driven by this mechanism – as opposed to mimicry or another well understood evolutionary process – is not especially strong. But Tobias and Seddon’s antbirds provide some new and compelling evidence for this process.

Antbirds (Thamnophilidae), unlike many songbirds, do not seem to learn their songs; this makes them good subjects for studying the evolution of song. Occupying partially overlapping ranges in Amazonia, peruviana and subflava tend to occupy slightly different habitats, but they can nevertheless be found side by side where their ranges overlap. Each species responds aggressively to the others’ songs, and even experienced ornithologists consider their songs to be virtually indistinguishable.

Using dozens of recordings of each species, Tobias and Seddon showed that the songs of peruviana and subflava overlapped in each of 20 measured acoustic parameters. Even a statistical model designed to tell the two species apart, using all 20 parameters, got the species wrong almost 20% of the time. Some vocal traits were more similar between sympatric peruviana and subflava in locations where they both occurred than in populations where only one species was present, suggesting that their vocal similarity may be related to their interactions where they co-occur.

But more compelling than this geographic variation in song was the contrasting pattern observed in a non-territorial vocalization and plumage color. These traits were highly divergent between peruviana and subflava, which is unsurprising since the species last shared a common ancestor an estimated 3.4 million years ago. Yet despite the substantial divergence in these traits that are not involved in territorial interactions, the territorial songs were quite similar between these competing species.

In Hypocnemis antbirds, like many tropical birds, males and females both participate in territory defense, which occurs year-round. Remarkably, exactly the same pattern of variation was found in female traits as in males – the two species differed dramatically in courtship vocalizations and color, but had extremely similar territorial songs.

This study cannot distinguish between convergence per se versus a failure to diverge (there is no way to know what the songs of these species sounded like in the distant past). Nevertheless, even a failure to diverge over 3.4 million years of evolutionary history is striking, especially in light of the substantial divergence in the vocal and visual traits that are not involved in territorial communication.

Tobias and Seddon have documented one of the most convincing examples yet of what we have termed “agonistic character displacement” or ACD – evolution that’s driven by aggressive interactions between sympatric species. Several members of the at UCLA are studying ACD in a variety of species (including me, with my work on anoles). We hope to learn how widespread this process is, and how important it has been in creating and maintaining diversity in animal signals and other traits.

Literature cited:

Tobias, J. A. and N. Seddon. 2009. Signal design and perception in Hypocnemis antbirds: Evidence for convergent evolution via social selection. Evolution 63(12):3168-3189.

Grether, G. F. , N. Losin, C. N. Anderson, and K. Okamoto. 2009. The role of interspecific interference competition in character displacement and the evolution of competitor recognition. Biological Reviews 84(4):617-635.