When some creatures want to get a message across there's no substitute for opening the mouth and calling. But not all: a few animals manage to vocalise,even with closed mouths. Tobias Riede explains that some birds, frogs and even primates project their mating calls through inflated air spaces, but how these animals communicate without using their mouths to modulate sounds puzzled him. Thanks to Gabriël Beckers and Roderick Suthers in Bloomington, Indiana,Riede already knew that the sounds emitted from the syrinx of the dove, which inflates its throat as it coos, were acoustically filtered to remove overtones and produce a sound with an almost pure single tone. But which structures in the throat act as acoustic filters and how they function wasn't clear. Riede and his colleagues decided to get inside a dove's skin, to find out how they call-filter (p. 4025).

Watch a cooing dove, and the first thing you'll see is the bird's throat swell with air, but Riede needed to know which internal organs became inflated during the serenade. The only way to view the throat's internal workings was to persuade a bird to sing in front of an X-ray camera. Contacting William Blevins at Purdue University, Riede asked if he could use the veterinary college's state-of-the-art X-ray equipment to collect X-ray movies of a bird's performance.

Riede remembers that once the birds had settled down in front of the camera, getting them to coo was straightforward; they started as soon as they heard another male's voice or saw a nearby female. And the movies were spectacular; the team could clearly see an enormous air bubble forming in the upper cervical oesophagus. However, the crop region of the oesophagus never inflated, suggesting that it featured little in dove sound production.

Knowing which regions of the throat became inflated as the bird cooed, the team inserted tiny tubes into the bird's trachea and oesophagus to record the coo's spectrum as it progressed through the throat. Comparing spectra from both regions of the throat with the sound recorded outside the bird's body,Riede realised that the trachea first damped some of the syrinx's overtones,while the oesophagus seemed to filter out the remaining harmonics.

Combining these results with a mathematical model that Riede and Neville Fletcher had derived previously, the team suspects that the trachea and oesophagus work together as acoustic filters, stripping the overtones while amplifying the coo's main tone. Riede suspects that the trachea may behave as a tube resonator that is closed at one end and tuned to the main tone's frequency by a variable sized opening at the top. This partially damps the overtones, before the sound is transmitted to the inflated oesophagus where the oesophagus outer wall and skin resonate at the main tone's frequency,further damping the overtones and transmitting the pure call as doves coo their piece.

Riede, T., Beckers, G. J. L., Blevins, W. and Suthers, R. A.(
). Inflation of the esophagus and vocal tract filtering in ring doves.
J. Exp. Biol.