Witmer estimates that T. rex's head would have been 18 percent heavier if not for the air spaces in its skull.
This savings may have allowed T. rex to pack more muscle onto its head, which possibly strengthened its bite and allowed it to tackle bigger prey.
(Explore an interactive of bizarre-looking dinosaurs.)
The nasal airways in the ankylosaurs, however, were surprisingly convoluted. It was as if "crazy straws" had been rammed up the creatures' snouts, Witmer said.
These winding airways were often located next to large blood vessels.
"Whenever we see that, it raises the possibility that we're looking at heat transfer," Witmer said.
This setup would have allowed hot blood circulating through the creatures' heads to dump excess heat into the airways, helping to cool their brains and the rest of their bodies.
The transferred heat also could have warmed up air the dinosaurs breathed, making gas exchange in the lungs easier.
In addition, the twisty nasal passages may have acted as resonating chambers for sounds.
The two ankylosaur species examined had slightly different airways, so their voices would have been subtly different, Witmer said.
The research could provide new clues about how dinosaurs achieved flight.
Some of the new study's research subjects were theropods, the group of dinosaurs from which modern birds are descended.
(Related: "T. Rex Protein "Confirms" Bird-Dinosaur Link" [April 24, 2008].)
"Very often people have thought that birds have hollow bones because they fly, but it could be the other way around," Witmer said.
"They could have evolved hollow bones for other reasons, and that gave them the lower body mass necessary to take to the air."
Hans-Dieter Sues is a dinosaur expert at the Smithsonian National Museum of Natural History in Washington, D.C., who also did not participate in the research.
Witmer "certainly makes a strong case for paranasal sinuses [air-filled spaces within the skull] reducing the weight of the skull in certain dinosaurs," Sues said.
Sues cautioned, however, that "such functional hypotheses are difficult to test even in living species, including our own."
SOURCES AND RELATED WEB SITES