ScienceDaily News Release: Working On The 'Porsche Of Its Time': New Model For Species Determination Offered

ScienceDaily News Release: Working On The 'Porsche Of Its Time': New Model For Species Determination Offered: "Novak investigated a Postosuchus specimen excavated from the Coelophysis Quarry of Ghost Ranch, New Mexico, a famous locality teeming with dinosaur fossils (mainly Coelophysis bauri) as a result of a mass death. While comparing the Ghost Ranch specimen with the two specimens of Postosuchus kirkpatricki from Texas, she noticed some differences in the bones. Not sure whether these differences were numerous enough or skeletally important enough to make the Ghost Ranch Postosuchus a new species, she decided to do a little more research before making a final decision.

Because the Rauisuchian fossil record is generally sparse, Novak instead dove into the dinosaurian fossil record in attempts to quantify the amount of skeletal difference historically regarded as valid to erect a new species within the same genera. She analyzed 28 genera containing 68 species from both the saurischian (lizard-hipped) and ornithischian (bird-hipped) orders. Using the fact that the skeleton of a dinosaur generally contains approximately 338 different bones, she catalogued the number of differences as well as where the differences were found on the skeleton. Calculations indicated that, on average, two species of dinosaur that are members of the same genera varied from each other by just 2.2 percent. Translation of the percentage into an actual number results in an average of just three skeletal differences out of the total 338 bones in the body. Amazingly, 58 percent of these differences occurred in the skull alone.

'This is a lot less variation than I'd expected,' said Novak, whose advisor is Josh Smith, Ph.D., Washington University assistant professor of earth and planetary sciences. 'As a concept, this is not statistically perfect. But I think it's something taxonomists can consider if they are in doubt over classifying something. It's a kind of benchmark with historical validity.'"