The phylogeny of Tetanurae (Dinosauria: Theropoda)

Abstract

Tetanuran theropods represent the majority of Mesozoic predatory dinosaur diversity and the lineage leading to extant Aves. Thus their history is relevant to understanding the evolution of dinosaur diversity, Mesozoic terrestrial ecosystems, and modern birds. Previously, the fragmentary and poorly sampled fossil record of basal (non-coelurosaur) tetanurans led to uncertainties regarding their basic interrelationships. This in turn prevented determining the relationships of many incompletely known taxa that nonetheless document a global radiation spanning more than 120 million years. We undertook an exhaustive examination of all basal tetanurans and all existing character data, taking advantage of recent discoveries and adding new morphological, temporal and geographic data. Our cladistic analysis of 61 taxa achieved significantly improved phylogenetic resolution. These results position several ‘stem’ taxa basal to a succession of monophyletic clades (Megalosauroidea, Allosauroidea and Coelurosauria). Megalosauroids include nearly 20 taxa arrayed amongst a basalmost clade (Piatnitzkysauridae, fam. nov.) and the sister taxa Spinosauridae and Megalosauridae; the latter includes two subfamilies, Megalosaurinae and Afrovenatorinae subfam. nov. Allosauroidea contains a diverse Metriacanthosauridae (= Sinraptoridae), Neovenatoridae, Carcharodontosauridae and a reduced Allosauridae. Finally, we assessed more than 40 fragmentary forms and hundreds of additional reported tetanuran occurrences. Tetanuran evolution was characterized by repeated acquisitions of giant body size and at least two general skull forms, but few variations in locomotor morphology. Despite parallel diversification of multiple lineages, there is evidence for a succession of ‘dominant’ clades. Tetanurae first appeared by the Early Jurassic and was globally distributed by the Middle Jurassic. Several major clades appeared prior to the breakup of Pangaea; as such their absence in specific regions, and at later times, must be due to poor sampling, dispersal failure and/or regional extinction. Finally, we outline a general perspective on Mesozoic terrestrial biogeography that should apply to most clades that appeared before the Late Jurassic.