Image: Rich Penney, Sandia National Laboratories

FAMILY of Parasaurolophus dinosaurs, with their strange hornlike crests, grazes by a swamp in Cretaceous New Mexico.
Click here for links to download the cry of the Parasaurolophus.

It's 75 million years ago, and you are somewhere in what is now the San Juan Basin of New Mexico. At the edge of a misty swamp, a family of dinosaurs munches on the leaves of nearby plants. The adults are almost 25 feet long and have strange mouths resembling duckbills. Even more curious, large, hornlike crests rise from the tops of their skulls. The biggest dinosaur abruptly stops feeding and looks your way. He raises his crest into the air, takes a deep breath and suddenly the earth is shaken by a deep rumbling cry of warning.

The eerie sound you just heard (if your browser supports RealAudio) is not the work of a Hollywood audio engineer taking a wild stab at re-creating a saurian bellow for the Jurassic Park soundtrack. Instead it may very well come extremely close to reproducing the actual voice of a Parasaurolophus (PAIR-uh-SOAR-uh-LOW-fuss) dinosaur.

Scientists at Sandia National Laboratories and the New Mexico Museum of Natural History and Science in Albuquerque produced the low-frequency sound by analyzing computerized tomography (CT) scans of a fossilized Parasaurolophus skull and simulating, on powerful computers, the sounds its internal passages and cavities would produce if air were blown through them like a musical instrument.

Image: Randy Montoya, Sandia National Laboratories

DIGITAL PALEONTOLOGISTS Carl Diegert of Sandia (left) and Thomas E. Williamson of the New Mexico Museum of Natural History and Science simulate the call of the "trombone dinosaur."

Parasaurolophus is a member of a group of duck-billed dinosaurs known as hadrosaurs--large, plant-eating dinosaurs that lived during the Late Cretaceous (100 million to 65 million years ago). Some hadrosaurs, such as Parasaurolophus, bore large, hollow crests on their heads and are referred to as lambeosaurs. The fossil making this simulation possible was discovered in August 1995 by a crew headed by Thomas E. Williamson, curator of paleontology at the New Mexico Museum of Natural History and Science. The 4.5-foot-long skull is one of only five ever found, and it is the most complete, lacking only the segment below the eyes; the bony crest was extremely well preserved (see sidebar).

The precise function of the Parasaurolophus's crest has long been a puzzle to paleontologists. Some believe the feature served to help the dinosaurs recognize one another or as a display during mating, much like a peacock's plumage. The fact that the crest is riddled with long passageways that connect to the animal's respiratory system also suggested to researchers that it served as a "radiator" to shed excess heat. Others speculated that the crests were acoustic resonators, capable of producing distinctive sounds. The latter idea and the unique shape of the crests led researchers to nickname Parasaurolophus the "trombone dinosaur."

David Weishampel of Johns Hopkins University first analyzed the potential vocal abilities of Parasaurolophus. He constructed simple plastic models of the crest to demonstrate that they could amplify low-frequency sounds. But Williamson sought further proof, so he turned to Carl Diegert, a computer scientist at nearby Sandia National Laboratories. Diegert is part of a group that creates complex, three-dimensional computer models for simulating problems that cannot be subjected to real-world tests.

Diegert and Williamson obtained a CT scan that produced 350 cross sections of the fossil skull and crest at three-millimeter intervals. The images revealed that the crest had a much more complicated internal structure than anyone had imagined. "Not only are there more tubes than the simple, trombone like loops described in previous studies, but there are new chambers within the crest," Williamson says.

Next, the digitized cross sections were used to reconstruct an undistorted model of the crest (see sidebar), separating material that was originally bone from the sandstone and clay that filled the fossil's cavities. Once the scientists had determined the size and shape of the air passages, they were able to calculate the natural frequency of the sound waves the dinosaur pumped out. "A very complex shape such as the dinosaur crest takes lots and lots of numbers to describe in the computer," Diegert observes. "It's only recently that computers have become powerful enough to allow that to happen."

Then, the computer was instructed to simulate air blowing through the crest. The resulting sound probably approximates the noises that the dinosaur crest could produce fairly well. But it is hard to say for certain. The researchers did have to reconstruct some missing dinosaur parts, such as the beak and nostrils, and the soft tissues of the head and throat that were not fossilized--all of which may have altered the overall sound quality. In addition, it is not known whether the Parasaurolophus had vocal cords, and so the researchers simulated sounds passing through vocal cords and sounds that had not.

All in all, they were not disappointed when the call of the Parasaurolophus rumbled from the speakers. As predicted, this prehistoric basso profundo was equipped with a rich vibrato and could vary the pitch. The digital paleontologists believe Parasaurolophus probably had a voice that was distinctive enough to distinguish it from other individuals. "The sound may have been somewhat birdlike, and it's probably not unreasonable to think they did songs of some sort to call one another," Diegert says. "Fossil records of the large bones in the dinosaurs' ears compared with corresponding bones in human ears suggest they were able to hear lower frequencies than humans."

The new data advance Sandia's simulation technology and provide important clues into dinosaur life. "This may support lines of evidence that suggest that hadrosaur dinosaurs were capable of fairly complex social behavior," Williamson says. Moreover, this high-tech computer simulation may well find other uses that go beyond pure science. Parasaurolophus's arpeggios have been copyrighted, and moviemakers are said to be interested in licenses.

So when the next dinosaur thriller shakes the movie house in Dolby Surround Sound, we might hear something closer to a real hadrosaur bellow. Could the not so dulcet tones of Tyrannosaurus rex be next?