Birds have a variety of alarm calls that warn other members of the flock about impending danger. But for some birds, the very act of taking off is enough to sound the alarm. Mae Hingee and Robert Magrath from the Australian National University have found that crested pigeons have modified wing feathers that produce distinct whistles when the birds take off quickly and steeply. That's exactly the sort of flight that they undertake when they're alarmed, and other pigeons treat the resulting whistles as cues to take to the skies themselves.
Crested pigeons are comical-looking birds that are widespread across Australia. Foraging in groups gives them some measure of protection from predators, but the birds have no alarm calls of any sort. They do, however, have whistling wings. Thanks to an eighth primary feather that is much narrower than its neighbours, the wings produce loud, almost metallic noises in flight. These are so conspicuous that birders can use them to identify the species. They have even earned the bird the nickname of "whistle-winged pigeons".
To use wingbeats as alarm signals, birds have to differentiate the flaps of panic from those of general flight. Certainly, startled birds take off faster and at a steeper angle and by recording the sounds of flying pigeons, Hingee and Magrath showed that these small changes produce audibly different sounds. The whistles were are and have a more rapid tempo when the birds take off in alarm, compared to regular flight. Based on these characteristics, the duo could correctly classify the vast majority of alarm whistles.
And it appears the pigeons can too. When they played back alarm whistles to flocks of foraging pigeons, the birds took off immediately around 70% of the time. The sounds of normal fight, by comparison, never triggered a mass ascension. It was the combination of volume and tempo that was important. Making normal whistles louder didn't make them better at triggering fleeing, although making alarm whistles softer reduced the panic the caused.
The playback experiments clearly show that the pigeons take off when they hear others doing so - they don't need to actually see others flying away. Wing whistles may warn other species of birds too, although that will need to be tested.
The fact that the pigeon's eighth primary feather seems to have been modified to make these noises suggest that they're not an accidental by-product of flight but an adaptive signal. However, it may originally have evolved to signal something other than alarm, such as strength or quality as a mate. It could even allow flocks of flying pigeons to keep track of one another, or to stay in formation.
Regardless of these other uses, the whistles clearly also provide an easy way of rapidly and reliably telling others about threats. The whistles even allow the pigeon to tell its peers about the urgency of the threat. The greater the peril, the more desperately an individual will try to flee from it, and the louder their wing whistles will sound.
Using noise to sound an alarm has many advantage - it provides a warning even to birds who might have their heads down or who can't see the individual that first spots the danger. But the whistles even have benefits over more familiar alarm calls. It's unlikely to be used dishonestly - while birds could conceivably make fake alarm calls to scare others away from valuable resources or mates, a crested pigeon couldn't fake its warning whistle without spending the energy it takes to fly off strongly. And in warning others, an individual doesn't need to draw attention to itself; the very act of fleeing to safety mobilises the rest of the group.
Reference: Proc Roy Soc B doi:10.1098/rspb.2009.1110
More on bird calls:
Fascinating! As an Aussie I see little top-knots (their other nickname) all over the place. I've heard their take-off whistles countless times, however I always assumed the whistles were vocal. The reason for that is that it made sense to me that a small exhalation on a wingbeat would be a plausible explanation for tweeting and taking off simultaneously. Of course I never tested this. The modified feather is a much more interesting answer than my assumption! Thanks.
For people not familiar with the sound, or who need reminding, the ABC's report contains audio: http://www.abc.net.au/science/articles/2009/09/02/2674184.htm (link to mp3 under "related stories").
I was feeding some of these birds at a park when they took flight, and to me it seemed that the whistle was most pronounced on the up-stroke of the wing. I have not had a chance since to confirm that observation.
These birds are becoming very common here in the suburbs of Melbourne, Australia. Once they were restricted to the inland but land clearing has enabled them to expand their domain and to increase in such numbers that I wonder if they will out-compete the introduced Spotted Turtle Dove, which itself forced out native pigeons and doves.
As Hank says, sometimes they are referred to as Topknot pigeons. This could be confusing as there is another Australian pigeon with the same name, Lopholaimus antarcticus.