麻豆传媒

GERALD BORGIA has spent 20 years spying on the spectacular wooing techniques of male bowerbirds. He has watched them construct intricate bowers, and adorn these with decorations gathered from the forest. He has marvelled at their flawless vocal impressions of laughing kookaburras and honeyeaters, cats鈥� meows and car engines. But even Borgia was left slack-jawed by what he caught on tape during one field trip to the remote highlands of Papua New Guinea.

Poring over hundreds of hours of footage back in his lab at the University of Maryland in College Park, Borgia came across a male streaked bowerbird belting out an extraordinary soundscape. It started with dogs barking and people chatting as they traipsed through the forest slashing at foliage, and continued with the sounds of machete strikes and a falling tree, complete with the rustle and shake of leaves and the great crash of it hitting the ground. 鈥淚 played it back again and again because I couldn鈥檛 believe it,鈥� says Borgia. 鈥淭his bird was mimicking an entire audio scene with the most astonishing accuracy. I was amazed鈥� shocked. It鈥檚 by far the most impressive piece of vocal mimicry I鈥檝e ever heard.鈥�

(Image: Tom Flower)

The streaked bowerbird may be a virtuoso, but it is not the only bird with such talents. The branches of the avian family tree occupied by perching birds, including songbirds, are bristling with expert impersonators. These range from familiar species such as the northern mockingbird 鈥� whose Latin name translates as 鈥渕any-tongued mimic鈥� 鈥� to exotic ones like the superb lyrebird that put in a star turn in David Attenborough鈥檚 The Life of Birds, with its impressions of an old-style camera shutter and a chainsaw.

These flabbergasting feats have been celebrated for centuries, but only now have scientists begun to figure out why birds do impressions. Their discoveries are overhauling the idea that avian impersonations are a useless by-product of the vocal learning process. Instead, they suggest that mimicry can pay off in a variety of ingenious ways. In other words, the ability to mimic can be an evolutionary adaptation that increases an individual鈥檚 chances of passing on its genes.

Seductive routines

Until a few years ago, there was precious little evidence that pirating sounds was profitable. Some people believed that birds mistakenly copy other species when learning the calls of their own kind. Another idea was that mimicry was an efficient way to develop a large repertoire of songs, something which helps males win mates. 鈥淏ut that doesn鈥檛 fully explain why so much vocal mimicry is so accurate,鈥� says Laura Kelley at the University of Cambridge. 鈥淔or mimicry to be as faithful as we know it is, you might expect some selection pressure for accuracy.鈥�

In 2007, that there is. The male satin bowerbird is a talented mimic with lustrous blue-black plumage that struts its stuff in the tropical forests of Australia. Having watched hundreds of these seductive shows, Borgia wanted to find out whether females judge males by the accuracy of their impersonations. He recorded 123 courtship displays at Wallaby Creek in New South Wales, during which each male mimicked five different bird species, including the kookaburra and Lewin鈥檚 honeyeater. Then he compared visual read-outs of these songs to those of the birds being imitated. Sure enough, females preferred males who produced highly accurate impersonations. In fact, the accuracy and range of vocal mimicry were more closely correlated with mating success than bower structure or number of decorations.

Female satin bowerbirds are suckers for a good mimic (Image: Konrad Wothe/Minden Pictures)

鈥淭he female is basically saying, 鈥榶ou did the kookaburra better than your rivals, so I鈥檓 picking you鈥�,鈥� says Anastasia Dalziell at Cornell University in Ithaca, New York, who has studied vocal mimicry in superb lyrebirds. This suggests that mimetic accuracy provides females with signals about a male鈥檚 quality, although it鈥檚 hard to know exactly how this works. It might be that learning and copying the songs of other species is a cognitive challenge, so the best mimics are also the smartest birds and therefore the most desirable mates. Or it might simply indicate maturity, with older males having had longer to perfect their imitations.

So being a good mimic can increase a male鈥檚 ability to win a mate. But that is not the only advantage: many birds produce copycat calls when attacked, and a clutch of studies has now demonstrated that doing so helps some to escape predators.

The greater racket-tailed drongo, which hangs out in mixed-species flocks across Asia, mimics alarm calls produced by its neighbours when attacked. Last year, Eben Goodale, now at Guangxi University in Nanning, China, wanted to see how this affects the behaviour of those neighbours. Working in Sri Lanka, he recorded drongos鈥� copycat alarm calls and played them back to orange-billed warblers and ashy-headed laughing-thrushes. , and they responded more vigorously to imitations of their calls than to the drongos鈥� own alarm calls. , and there is far less mayhem when it uses its own alarm call.

So it appears that mimics can trick other species into helping them mob an attacker 鈥� and that could be the difference between life and death. 鈥淚t鈥檚 likely to be a safety-in-numbers thing,鈥� says Branislav Igic at the University of Akron in Ohio. 鈥淭he more prey there are flying around, the less likely the predator is to single you out. Mobbing also produces a lot of sound and movement, and that may confuse the predator.鈥�

. When Igic confronted brown thornbills with a taxidermy owl placed on the ground, they made calls produced by other bird species in response to terrestrial threats such as snakes or perched avian predators. When confronted by a gliding fibreglass sparrowhawk, however, these diminutive songbirds mimicked aerial-specific alarm calls. Intriguingly, they also produce these calls when disturbed in their nests by humans or captured in mist nests. Because there is no actual aerial threat in these cases, Igic suspects that this may be an attempt to deceive the assailant. 鈥淭hey might trick terrestrial predators into thinking there is a flying predator around, and in doing so deter the predator from the nest,鈥� he says.

For now, the idea remains untested. But there is no doubt that some birds do use mimicry to deceive. Consider the cuckoo, a renowned parasite that disguises its eggs to trick a host species into raising them as its own. In Australia, Horsefield鈥檚 bronze cuckoo chicks take this deception a step further by impersonating calls made by nestlings of its primary host, the superb fairy wren. The chicks are born with this ability, a trait selected by evolution over countless generations. Amazingly, though, when Naomi Langmore at the Australian National University in Canberra transferred chicks into nests of the buff-rumped thornbill, . It seems that Horsefield鈥檚 bronze cuckoo chicks can fine-tune their impersonations through trial and error to arrive at a call that elicits the greatest response from at least two different host species.

Fork-tailed drongos cry 鈥渉awk鈥�, then steal food from other birds and meerkats (Image: Tom Flower)

Perhaps the craftiest of all avian impressionists, however, is the fork-tailed drongo. 鈥淭hey鈥檙e demonic-looking birds with beady red eyes and hooked beaks,鈥� says Tom Flower at the University of Cape Town, who has spent years tracking wild drongos in South Africa鈥檚 Kalahari desert. They are very gregarious, associating with meerkats and southern pied babblers, and often warning their allies of approaching predators. But they are also excellent impressionists, and will gladly trick their allies in order to steal food.

Flower used playback experiments to show that the drongos鈥� impressions of alarm calls made by another local bird, the cape glossy starling, successfully trick babblers and meerkats. Even if there was no predator to be seen, animals of both species fled the scene when they heard the faked calls. : these drongos get about a quarter of their food by tricking their neighbours. It might seem as if the meerkats and babblers are easily duped, but it makes sense to pay attention to the drongos鈥� alarm calls. After all, they only lose a meal by responding to a false alarm, but if they ignore a genuine warning, they pay with their lives.

. When Flower played the same type of alarm call three times in succession, he found that babblers were slower to flee. However, it seems the drongos are wise to this. Over the course of hundreds of hours of following them, Flower noticed something: 鈥淲hen targets stopped responding to one type of alarm call, they switched their calls,鈥� he says. To find out whether this was a deliberate strategy, he recorded drongo calls on 151 separate occasions during which they made repeated attempts to steal food from the same individual. He found that drongos almost always swapped their alarms, especially after failed attempts 鈥� a ploy that increased their chances of stealing a snack.

鈥淚f drongos can attend to what their targets are doing and then change their tactics accordingly, it suggests they鈥檙e extremely flexible in their deceptive behaviour,鈥� says Flower. So, how clever are they?

It has even been suggested that using vocal mimicry for tactical deception indicates that the mimic understands what鈥檚 going on inside the minds of its targets. This ability is called theory of mind, and is often thought to be unique to humans. The idea that birds have it might sound like a stretch. But in the past decade, certain birds have proved to be far more intelligent than you might think (see 鈥�Who鈥檚 a clever bird?鈥�). These include western scrub jays, which apparently understand what their peers are thinking.

鈥淯sing vocal mimicry for deception could be a sign that the mimic knows what other birds are thinking鈥�

Not so bird-brained

Scrub jays like to stash their food, hiding tasty morsels out of sight of fellow jays that might try to steal them. Nicky Clayton at the University of Cambridge and Nathan Emery, now at Queen Mary University of London, found that if a bird was observed cacheing food by another, it would privately find a new hiding place later. Such re-cacheing did not happen in the absence of an observer and was confined to birds that had been thieves in the past. That led Emery and Clayton to argue that this behaviour relies on a form of mental attribution: simulating the point of view of a potential thief. 鈥淭hat鈥檚 the strongest evidence for theory of mind in any [non-human] animal, bird or otherwise,鈥� says Clayton.

Are fork-tailed drongos equally intelligent? They are potential candidates, says Clayton. 鈥淚f you鈥檙e looking for a place to find theory of mind, you would start with big-brained animals that seem to exist in an ecology where this ability would be advantageous. But first you would need to rule out simpler explanations.鈥� Indeed, Flower has his doubts. 鈥淚n humans, this sort of behaviour requires very sophisticated cognition, but that鈥檚 not necessarily the case for a drongo,鈥� he says. 鈥淢y impression is that they鈥檙e likely to be trial-and-error learners.鈥� He suspects they follow simple rules that amount to a 鈥渨in-stay, lose-shift鈥� strategy. But he is no killjoy: 鈥淎ctually, I think it鈥檚 amazing what these birds can accomplish in terms of complex behaviour with such a simple mechanism.鈥�

Either way, the debate about whether avian mimicry can increase a bird鈥檚 evolutionary fitness is over. Birds can undoubtedly benefit from their impersonations. The challenge now is to figure out how they evolved the mechanisms that allow them to do so. 鈥淲hat we鈥檙e trying to do is show why particular mechanisms are selectively advantageous; why it pays for drongos to use trial-and-error learning and why humans use theory of mind, for example,鈥� says Flower. If they can answer those sorts of questions, then understanding mimicry could even help unlock the mysteries of how thinking evolved.

This article appeared in print under the headline 鈥淭he impersonators鈥�