GSA Annual Meeting in Indianapolis, Indiana, USA - 2018

Paper No. 213-13
Presentation Time: 5:00 PM

BAYESIAN CHARACTERIZATION OF THE LATE TRIASSIC ADAMANIAN-REVUELTIAN FAUNAL TURNOVER, PETRIFIED FOREST NATIONAL PARK, AZ, USA


HAYES, Reilly1, PUGGIONI, Gavino2 and FASTOVSKY, David E.1, (1)Department of Geosciences, University of Rhodes Island, 9 East Alumni Ave, Kingston, RI 02881, (2)Department of Computer Science and Statistics, University of Rhode Island, 9 Greenhouse Rd, Kingston, RI 02881

The Alvarez et al. (1980) proposal that attributed the Cretaceous-Paleogene mass extinction to an asteroid impact sparked interest in the role of impacts in shaping Earth’s biota. However, any plausible extinction mechanism must be consistent with the patterns of extinction. Identifying such patterns, however, is difficult in stratigraphically complex terrestrial settings, especially with sporadically-sampled vertebrate taxa. In the Late Triassic Chinle Formation of Petrified Forest National Park (PEFO), AZ, USA, however, stratigraphic, geochronological, and fossil data coincide to permit quantitative reconstruction of the Adamanian-Revueltian (A-R) faunal turnover. This event has been recently tied to the 215.5 Ma Manicouagan impact of Quebec, Canada; however, warming and aridification after the mid-Norian collapse of a mega-monsoonal system in western equatorial Pangea have also been linked to this event. Here, we use the Bayesian method of Alroy (2014) to quantitatively reconstruct the A-R extinction and subsequent diversification to determine if the patterns of extinction are concordant with those expected from an asteroid impact.

We utilize 11 high-precision sequential CA-TIMS U-Pb dates through the thickness of the Chinle Formation in PEFO, coupled with a robust lithostratigraphy, to carry out this analysis. Our approach draws upon the stratigraphic distribution of fossil occurrences to compute extinction probabilities in sequential time intervals, each conditioned on those recovered from the previous time interval, allowing us to construct 95% credible intervals constraining the moment of an extinction for each taxon. These intervals then are used to assess the synchronicity of faunal originations and extinctions in PEFO with each other, the 215.5 Ma Manicouagan impact, and the mid-Norian climactic shift. We also determine whether the A-R boundary, variously considered a regional and global biostratigraphic datum, is reasonably construed as a single event. Preliminary results suggest that although several extinctions precede the Manicouagan impact, for many synchronicity with the event cannot be ruled out.