Taxonomy Of The Shastasaurids From The Middle To Late Triassic Of Guizhou, China And The Phylogenetic Analysis Of Triassic Ichthyopterygians

Shastasauridae is one of most important ichthyosaur groups in the Middle to Late Triassic and has a wide paleogeographic distribution all over the world. It has been reported for over a century since Merriam(1895, 1902) first described the materials from the Upper Triassic of California, USA. However, compared with the well-preserved complete skeletons of ichthyosaur materials from the Jurassic and Cretaceous, Shastasauridae specimens were poorly preserved and almost no complete skeletons were found besides Besanosaurus before this century. Most of the genera and species of the family were erected based on the poor holotypes, most of which are partial postcranial skeletons, while this situation has caused seriously controversial on the taxonomy of the ingroup taxa. Maisch and Matzke(2000) pointed out that it it has served as a waste-basket for medium to large Middle to Upper Triassic, non-cymbospondylid ichthyosaurs. Recently abundant materials of ichthyosaurs have been found from the Triassic of South China and at least two genera are close related with Shastasauridae. Here we describe the holotype and referred materials of Guizhouichthyosaurus tangae and Guanlingsaurus liangae from the Upper Triassic of Guanling, Guizhou and revise the diagnosis and synonymy. A new phylogenetic analysis is conducted based on a reexamination of the global Triassic ichthyosaur specimens and revision of the character list. The phylogenetic positions of the newly published Chinese ichthyosaur taxa are discussed for the first time. We also measure the skull and vertebrae morphology of Guanlingsaurus liangae and some related specimens and discuss the functional morphology on tailbend and feeding pattern.Based on the reexamination on the holotype and referred materials, the diagnosis of Guizhouichthyosaurus tangae is revised as follows: over 6m in adult; about 63 presacral vertebrae; a long and shallow groove present anterior to the external naris; scapula fan-shaped, shaft absent proximally; presence of a pisiform distal to the ulnare and remarkably smaller than other distal carpals; the proximal phalangies of digit II much smaller than those of digit III. “Cymbospondylus asiaticus”, “Panjiangsaurus epicharis”, and “Shastasaurus tangae” are all junior synonymy of Guizhouichthyosaurus tangae and they are all from the same locality and have the same stratigraphic occurrence. The diagnosis of Guanlingsaurus liangae is revised as follows: large ichthyosaurs, over 8m in adult; shout snout, edentulous; skull appearing wide triangular in dorsal view; premaxilla without dorsal lamina and only forming the antero-ventral margin of the external naris; parietals without separating anteriorly as Y-shape and eliminating the frontal from the parietal foramen; parietal foramen located within parietals; postorbital entering the upper temporal fenestra; 77-80 presacral vertebrae, the largest number in ichthyosaurs; radius and ulna nearly round-shaped; fibula semicircular, contiguous margin slightly concave, posterior flange present; pubis foramen widely open as fossa.The measurements on the vertebrae of Guanlingsaurus liangae suggest that at least three wedge-shaped centra are present in the middle of the tail and form a small angle downward. This is the first report of a true tailbend in Triassic which was supposed to exist only in post-Triassic ichthyosaurs. However, the angle of the tailbend in Guanlingsaurus is relatively small(10°-15°), much smaller than that of a typical tailbend in Jurassic(60°), suggesting that the tailbend in Guanlingsaurus is still really primitive. The presence of a tailbend in Triassic ichthyosaurs indicates a great progress that ichthyosaurs have achieved in swimming ability and they probably have already invaded the open ocean at the Late Triassic.Comparison on the hyoid and skull morphology of Triassic ichthyosaurs, whales and sharks shows that the hyoid and skull shape of Triassic ichthyosaurs appear more similar to those of ram-feeding marine vertebrate and quite different from those of suction feeders. Also, ichthyosaurs don’t have ossified hyoid central corpus which is probably necessary for suction feeding behavior. Therefore, at least the four species involved in the analysis are rather ram feeders than suction feeders.Based on previous researches and the materials reported from China recently, we condult a phylogenetic analysis of Triassic ichthyosaurs including 34 taxa and 106 characters. 28 most parsimonius trees are obtained. The strict consensus tree confirms the monophyly of the clades Mixosauridae, Cymbospondylus, Shastassauridae and Toretocnemidae. Compared with previous research, the Shastasauridae is enlarged to include the new taxa from South China:(Besanosaurus(Guizhouichthyosaurus(“Callawayia”.wolonggangensis(Shastasaurus(Guanlingsaurus, Shonisaurus))))). The sistergroup of Guanlingsaurus and Shonisaurus form a monophyletic clade with Shastasaurus from North America. Guizhouichthyosaurus and “C. wolonggangensis” appear as the outgroup of this clade and Besanosaurus is the most primitive taxon of Shastasauridae. The phylogenetic analysis suggests that “C. wolonggangensis” and C. neoscapuris do not form a monophyletic clade. “C. wolonggangensis” does not show the diagnosis of the latter such as maxilla without dorsal lamina. Before more specimens are found, we can’t confirm the relationship between “C. wolonggangensis” and Guizhouichthyosaurus. Qianichthyosaurus zhoui from South China and Toretocnemus from North America form a monophyletic clade based on their synapomorphies on the postcranial morphology. However, one major difference between previous research and the current study is the reversal of position between Cymbospondylus and Mixosauridae. We believe that Cymbospondylus is closer to Shastasauridae as they both have large nasal exposure, well extended postfrontal medially and large number of presacral account. Mixosauridae is relatively primitive based on the skull morphology and many synapomorphies have been found between Mixosauridae and Chaohusaurus from the Early Triassic. Therefore, Mixosauridae is possibly closer to the Early Triassic type than the Late Triassic. That the Triassic ichthyosaur sequence shown by our phylogenetic analysis basically matches the stratigraphic occurrences, also indicates that our analysis is acceptable. The paleogeographic distribution of the Triassic ichthyosaurs indicates a close interrelationship between South China and the west Tethys during the Middle Triassic and between South China and North America during the Late Triassic. The stratigraphic and the paleogeographic occurrences together suggest a possible immigration pathway of the Shastasauridae from Europe to South China during the Middle Triassic and then to North America during the Late Triassic.