Sarahsaurus Aurifontanalis from the Early Jurassic Kayenta Formation
Total Page:16
File Type:pdf, Size:1020Kb
RESEARCH ARTICLE Anatomy and systematics of the sauropodomorph Sarahsaurus aurifontanalis from the Early Jurassic Kayenta Formation 1,2 1 Adam D. MarshID *, Timothy B. Rowe 1 The Jackson School of Geosciences, University of Texas at Austin, Austin, Texas, United States of America, 2 Division of Science and Resource Management, Petrified Forest National Park, Arizona, United States of America * [email protected] a1111111111 a1111111111 a1111111111 Abstract a1111111111 a1111111111 Sarahsaurus aurifontanalis, from the Kayenta Formation of Arizona, is one of only three sauropodomorph dinosaurs known from the Early Jurassic of North America. It joins Anchi- saurus polyzelus, from the older Portland Formation of the Hartford Basin, and Seitaad reussi, from the younger Navajo Sandstone of Utah, in representing the oldest North Ameri- OPEN ACCESS can sauropodomorphs. If it is true that sauropodomorphs were absent from North America Citation: Marsh AD, Rowe TB (2018) Anatomy and during the Late Triassic, the relationship among these three dinosaurs offers a test of the systematics of the sauropodomorph Sarahsaurus mechanisms that drove recovery in North American biodiversity following the end-Triassic aurifontanalis from the Early Jurassic Kayenta extinction event. Here we provide the first thorough description of Sarahsaurus aurifontana- Formation. PLoS ONE 13(10): e0204007. https:// doi.org/10.1371/journal.pone.0204007 lis based on completed preparation and computed tomographic imaging of the holotype and referred specimens. With new anatomical data, our phylogenetic analysis supports the con- Editor: William Oki Wong, Indiana University Bloomington, UNITED STATES clusion that Sarahsaurus aurifontanalis is nested within the primarily Gondwanan clade Massospondylidae, while agreeing with previous analyses that the three North American Received: December 4, 2017 sauropodomorphs do not themselves form an exclusive clade. A revised diagnosis and Accepted: September 1, 2018 more thorough understanding of the anatomy of Sarahsaurus aurifontanalis support the Published: October 10, 2018 view that independent dispersal events were at least partly responsible for the recovery in Copyright: This is an open access article, free of all North American vertebrate diversity following a major extinction event. copyright, and may be freely reproduced, distributed, transmitted, modified, built upon, or otherwise used by anyone for any lawful purpose. The work is made available under the Creative Commons CC0 public domain dedication. Data Availability Statement: Locality information Introduction and the fossils of TMM 43646-2 are available at the Univeristy of Texas Vertebrate Paleontology Lab in In the late 19th Century O. C. Marsh of the Yale Peabody Museum described the first remains Austin, TX (Accession number: A2000.43646). of North American early sauropodomorph dinosaurs, from the Early Jurassic Portland Forma- Linear measurements and phylogenetic character- tion of Massachusetts and Connecticut [1±6]. Several names were coined for these specimens, taxon matrices for all analyses in the paper are but most recent authors consider all this material to represent a single taxon, and most (but included as supplemental information. Digital videos of CT renderings and the CT data not all) recognize Anchisaurus (Ammosaurus) polyzelus as its valid name [7±10]. We follow themselves are available at The University of Texas this convention below. High-Resolution X-Ray Computed Tomography Several decades would pass before additional early sauropodomorph specimens were found Facility (www.ctlab.geo.utexas.edu). in North America, and nearly a century before informative new specimens were recovered. PLOS ONE | https://doi.org/10.1371/journal.pone.0204007 October 10, 2018 1 / 108 Anatomy and systematics of Sarahsaurus aurifontanalis Funding: This research was funded by the Jackson The next sauropodmorph to be described was a fragmentary postcranial skeleton collected by School of Geosciences and the National Science Lionel F. Brady of the Museum of Northern Arizona. It was found in northern Arizona in the Foundation (NSF EAR 1258878, EAR 1561622, EAR 1160721, EAR 0948842, and IIS-0531767) to uppermost formation of the Glen Canyon Group, the Early Jurassic Navajo Sandstone. Brady TBR. Andrew Milner and the St. George Dinosaur referred this material to Anchisaurus (Ammosaurus)[11, 12], believing this taxon to be a thero- Discovery Site at Johnson Farm funded Brian pod dinosaur. Galton [7, 13] later correctly interpreted Anchisaurus (Ammosaurus) as an early Engh's art of the Kayenta fauna. The funders had sauropodomorph, to which he referred Brady's material plus a second fragmentary specimen no role in study design, data collection and from the Navajo Sandstone of Arizona that was collected by Charles L. Camp for the Univer- analysis, decision to publish, or preparation of the sity of California Museum of Paleontology. Yates [8] subsequently argued that the two Navajo manuscript. Sandstone specimens represented a single taxon and that it was distinct from Anchisaurus Competing interests: The authors have declared (Ammosaurus). He thought it was possibly related to Massospondylus carinatus [14] from that no competing interests exist. southern Africa, but he deemed the material too incomplete for a definitive referral. Abbreviations: BP, Evolutionary Studies Institute Between 1976 and 1978, a joint expedition from the Museum of Northern Arizona and (formerly the Bernard Price Institiute for Harvard's Museum of Comparative Zoology (MCZ) explored another part of the Glen Canyon Paleontological Research), Johannesburg, South Group in northern Arizona, the so-called `SiltyFacies' of the Early Jurassic Kayenta Formation, Africa; BSP, Bayerische Staatssammlung fuÈr PalaÈontologie und Historische Geologie, MuÈnchen, which underlies and interfingers with the Navajo Sandstone. They collected a single broken Germany; LV, Lufeng Dinosaur Museum, Jingshan, sauropodomorph skull and associated postcranial fragments from the base of a geographic fea- Yunnan Province, China; MCP, Museu de Ciêncian ture known as `RockHead.' Owing to the questionable taxonomic identity (below), we infor- e Tecnologia PUCS, Porto Alegre, Brazil; MCZ, maly designate this as the `RockHead specimen.' The specimen was reposited in the MCZ Museum of Comparative Zoology, Harvard (MCZ 8893). University, Cambridge, Massachusetts; SMNS, Staatliches Museum fuÈr Naturkunde, Stuttgart, In the initial description [15], the Rock Head specimen was referred to Massospondylus sp., Germany; TMM, Vertebrate Paleontology a well-known taxon represented by numerous relatively complete skeletons from multiple Laboratory, University of Texas at Austin, Austin, localities in southern Africa [14]. Its identification as Massospondylus sp. was taken as evidence Texas. further supporting the concept that a cosmopolitan dinosaur fauna of low taxonomic diversity occupied Pangaea in the Late Triassic and Early Jurassic [9, 10, 16±18]. Subsequent workers expressed doubts about its referral to Massospondylus sp., and it became known as `the unnamed Kayenta prosauropod' (e.g., [19] p. 27, [20]). Uncertainty about its identity also raised some of the first doubts to be cast over the prevailing view of cosmopolitan dinosaur faunas of low taxonomic diversity across Pangaea in the Late Triassic and Early Jurassic. In 2006, the sauropodomorph ªFendusaurus eldoniº from the Early Jurassic McCoy Brook Formation of Nova Scotia was named on a very poorly preserved specimen and given prelimi- nary description in an unpublished dissertation [21]. A formal description and diagnosis has yet to be published. In 2010, preliminary descriptions of two new early sauropodomorphs from the Glen Can- yon Group of the Colorado Plateau were published. The first was Seitaad ruessi, described by Sertich and Loewen [22] based on a single, partial skeleton lacking the skull. It was the third sauropodomorph specimen reported from the Navajo Sandstone; it was collected in southern Utah by the Utah Museum of Natural History [22, 23]. Seitaad ruessi was postulated to have affinities either to plateosaurid or to massospondylid sauropodomorphs, but its incomplete- ness left a measure of phylogenetic uncertainty that was compounded by more general uncer- tainty and poor phylogenetic resolution for this early history of the sauropodomorph lineage [24, 25]. The two specimens previously reported from the Navajo Sandstone were found to compare favorably to Seitaad ruessi, but they were considered too fragmentary for a confident referral [22]. Next to be described was Sarahsaurus aurifontanalis, based on two partial skeletons [25]. Both individuals were collected from a single quarry in the Silty Facies of the Kayenta Forma- tion during a collaborative survey by The University of Texas Vertebrate Paleontology Labora- tory and the Navajo Nation EcoScouts that took place between 1997±2000. In the initial diagnosis and description of Sarahsaurus aurifontanalis, the holotype was designated as the more mature and complete of the two skeletons and the only one that included parts of a PLOS ONE | https://doi.org/10.1371/journal.pone.0204007 October 10, 2018 2 / 108 Anatomy and systematics of Sarahsaurus aurifontanalis disarticuated skull; the less-mature and less-complete skeleton was designated the paratype [25]. In this initial description, the Rock Head skull was provisionally referred to Sarahsaurus aurifontanalis. Below, we present a more detailed description of the