Nathan Myhrvold cautions against drawing unsupported conclusions from dinosaur growth data
BELLEVUE, WASH. — (Dec. 16, 2013) — Improved statistical techniques for evaluating dinosaur growth data could change the way paleontologists estimate dinosaur growth rates and other quantitative and qualitative growth traits. For years, some paleontologists have analyzed data on growth rates using methods involving subjective criteria such as assumptions about maximum size and nature of growth. But a new study by Nathan Myhrvold, recently published in PLOS One, found the field should reconsider some basic assumptions.
“Insufficient attention has been given to problematic issues in estimating the ages and masses that dinosaurs achieved during their lives,” Myhrvold explained. Myhrvold uses new methods to critically examine and re-analyze previous growth rate studies with increased scientific rigor and objectivity.
Specifically, the new statistical analysis:
Makes improvements in the choice of variables used to model dinosaur growth;
Uses a statistical criterion that ensures selection of the best-fitting growth model from among a large collection in an objective and reproducible way; and
Employs statistical resampling techniques to estimate the magnitude of error in the analysis.
The examination of previous growth-rate studies revealed that some previously reported results cannot be replicated using the methods originally reported, and that results obtained from the improved methods are in many cases different in both the quantitative rates and the qualitative nature of dinosaur growth. Astonishingly, the examination of 11 species contained within two studies reported in Nature (Nature, 2001 & Nature, 2004) and one in The Anatomical Record revealed that each of the 11 species included multiple serious errors in a key result or graph. The study also revealed that problematic results from these papers were quoted in additional papers in Science and Trends in Ecology and Evolution.
The new results, which demonstrated much better fits to the data, revise the growth rates downward in some cases, including for Tyrannosaurus – by a factor of about two. Despite this, Myhrvold said his results “do not change the broad conclusion that dinosaurs were relatively fast-growing creatures.” The study also found that mature specimens seem to be missing or underrepresented in many data sets, and it cautioned against drawing conclusions about growth traits that occur far beyond the age of the oldest known specimens until more data has been analyzed.
Another potentially interesting finding involves the Allosaurus. While speculative at this point, the study found that growth data for what was thought to be a single dinosaur species, Allosaurus fragilis, may actually contain specimens of two distinct species, including one that is a fast-growing and likely gigantic, previously unidentified Allosaur. While other interpretations are possible, it is worth further study. If confirmed, it would represent the first discovery of a new species through data modeling, and demonstrates one potential value of using growth curves to study dinosaurs.
# # #
About Nathan Myhrvold
Nathan P. Myhrvold, Ph.D., is chief executive officer and founder of Intellectual Ventures, a company focused on the funding, creation and commercialization of invention. An avid inventor, Dr. Myhrvold holds several hundred patents with several hundred patents pending. His research has been published in scientific journals including Science, Nature, Paleobiology, Environmental Research Letters, Journal of Vertebrate Paleontology and the Physical Review and he has contributed articles to publications including Harvard Business Review, The Wall Street Journal, Fortune, Time and National Geographic Traveler. For more on Dr. Myhrvold, please visit his website: http://www.nathanmyhrvold.com.