What Kind of Sound Did a Duck-Billed Dinosaur Make?

Using a nearly intact fossil skull from a beast that roamed the earth 75 million years ago, scientists here are trying to re-create the sounds that a duck-billed dinosaur might have made.

It's the kind of work that sent high technology, such as computer simulations, bumping up against ancient history. The Parasaurolophus lived in northwestern New Mexico. A skull of one of the creatures was dug up last August near Farmington, about 180 miles northwest of Albuquerque.

Based on preliminary calculations, scientists believe they have pinned down what was likely the primary frequency echoing through Parasaurolophus's distinctive 4-1/2-foot crest, says Tom Williamson, a paleontologist with the New Mexico Museum of Natural History and Science.

And eventually, through computer models, researchers hope to re-create the voice of the Parasaurolophus. The great beast likely communicated across its swampy environs in a low frequency - below 10 hertz, or 10 vibrations a second - a pitch too low for humans to hear, Mr. Williamson says.

But Michael Brett-Surman, a dinosaur specialist with the Smithsonian Institution, says that while the skull may be able to indicate only low-frequency sounds, the Parasaurolophus likely also made sounds that would be audible to humans. Higher-frequency sounds probably were needed to communicate with the dinosaurs' young, he says. Elephants also have such wide-ranging sound frequencies, including some inaudible to humans.

"These long, low sounds are called 'infra-sounds,' and they're excellent for long-distance, out-of-sight communication. In elephants I think it carries five miles," Mr. Brett-Surman says.

THE skull, which is largely intact but distorted by fossilization, was dug up by Williamson and Robert Sullivan of The State Museum of Pennsylvania.

The skull's nasal passages loop back and down, much like the brass tubing of a trombone, inside a backward-sweeping crest.

Scientists are working to reconstruct a computer model of the skull because simulations could replace physical tests that would be invasive.

"The benefit is we don't have to destroy the specimen. We don't have to cut it in half," Williamson says. The reconstruction is still incomplete, he says.

"We don't have the very tip of the snout with the nostrils ... so we have to reconstruct the shape of that [foot-long] area, what that would be," he says. "We assume sound would be produced in the throat, below the skull."

Williamson and Carl Deigert, a Sandia National Laboratories scientist, hope to eventually use supercomputer simulations - "digital paleontology" - to re-create the voice.