What Did the Crests Sound Like? Reconstructing a Call From a Hollow Skull

The crest is built from the bones surrounding the nostrils, extended backwards and upwards into a hollow structure that in the best-known species forms a long curved tube projecting behind the head. The nasal passages run from the nostrils up into the crest, loop around inside it and return to the throat, which makes the airway several times longer than a direct route would be. The shape differs between species, with some forming a tube, some a broad plate and some a low solid ridge with no cavity at all, and those differences are the primary means of telling the species apart. Juveniles have small crests that grow disproportionately with age, which is a pattern that matters for interpreting what the structure was for.

Several functions have been proposed and the evidence supports some more than others:

• •Sound production, since a long tube resonates and would give the animal a distinctive low call
• •Species recognition, supported by the variation between closely related species being concentrated in the crest
• •Display, supported by the crest developing fully around the age of breeding
• •Improved sense of smell, since the extended passage carries more sensory tissue, which has some support
• •Heat exchange, warming and humidifying inhaled air, which the geometry would achieve incidentally
• •Snorkelling, proposed early and abandoned since the crest has no opening at the far end

The acoustic work proceeded by treating the passage as a resonating tube and calculating what frequencies it would support. Early attempts used physical models built from measurements. Later work used computed tomography to map the internal passage in three dimensions from an intact skull, built a digital model of the airway and computed its resonant frequencies, producing a prediction of a low sound in the range of a brass instrument of comparable length. The resulting audio has been widely reproduced. The reconstruction depends on assumptions that are stated in the papers and generally omitted afterwards, including how the soft tissue lining the passage behaved, whether the animal had a sound-producing structure in the throat and what the surrounding tissue did to the resonance, each of which would shift the result.

A useful comparison comes from the related animals whose crests are solid or absent, since they constrain what the hollow ones are for. Some duck-billed species carry a low bony ridge with no internal passage, some have a fleshy crest inferred from preserved skin impressions rather than from bone, and some have essentially nothing. Those animals were doing whatever they did without a resonating tube, which indicates that the tube is not necessary for anything essential and points towards a signalling role that different lineages solved differently. Soft-tissue crests are particularly interesting because they would leave almost no trace in most specimens, which raises the possibility that structures on the head were commoner than the bones suggest and that reconstructions showing bare skulls are systematically too plain.

A useful check comes from the other end of the system. The inner ear can be reconstructed from the same scans, and its dimensions constrain the range of frequencies an animal could hear, since the length of the cochlea relates to hearing range across living groups. Work on these animals indicates sensitivity to low frequencies, which matches the frequencies the crest would produce, and that correspondence between a proposed sound source and the hearing range of the animals that would receive the sound is the strongest evidence available for the acoustic interpretation. It is correspondence rather than proof, since a low-frequency hearing range has other explanations, and it is considerably better than an argument from the shape alone, which is why the ear work matters more than its lower profile suggests.

Nasal passages loop through a hollow crest, making the airway several times longer than a direct route, and modelling it as a resonating tube predicts a low sound. Variation between related species concentrated in the crest and full development at breeding age both point to signalling. Reconstructed inner ears indicate low-frequency hearing, which matches what the crest would produce.