Tyrannosauridae

Christopher A. Brochu, 2003, Osteology of Tyrannosaurus rex: insights from a nearly complete skeleton and high-resolution computed tomographic analysis of the skull, Journal of Vertebrate Paleontology 22, pp. 1-138 : 2-3

SYSTEMATIC REVIEW OF TYRANNOSAURIDAE

Tyrannosauridae is an undoubtedly monophyletic assemblage of theropods. It is best represented in the Late Cretaceous of western North America and eastern Asia, with possible occurrences elsewhere in space and time ( Leidy, 1856; Marsh, 1892; Cope, 1892; Lambe, 1904, 1914b, 1917; Maleev, 1955, 1974; Gilmore, 1933; Kurzanov, 1976b; Lawson, 1976; Dong, 1977, 1979; Carpenter, 1982; Schwimmer et al, 1993; Lehman and Carpenter, 1990; Currie et al., 1990; Carpenter, 1990; Carr and Williamson, 2000; Hutt et al., 2001; Schwimmer and Kiernan, 2001; Carr et al., in press). Members of the group share several derived character states, including D-shaped incisiform premaxillary teeth; a squamosal recess; a quadratojugal-squamosal flange bisecting the infratemporal fenestra; an enlarged surangular foramen; a reduced retroarticular process; contact between the lacrymal and postorbital over the orbit; a functionally didactyl hand; an enlarged lesser trochanter on the femur; and an arctometatarsalian pes (see Holtz, 2001a for review). These are among the most popular dinosaurs in the public and were surely among the dominant predators of their ecosystems.

Given the popularity of tyrannosaurids in both academic and lay circles, one would expect Tyrannosauridae to be one of the best-characterized dinosaurian groups. In fact, there is considerable debate concerning virtually any aspect of tyrannosaurid systematics, from group membership and diagnosis to the position of Tyrannosauridae within Theropoda. Small tyrannosaurs might be dwarf taxa or young examples of other largebodied species (Carpenter, 1992; Carr, 1999; Currie and Dong, 2001a). Many names are based on fragmentary remains, and whether all are properly diagnosable is an open question. The group seems to bounce around in various phylogenetic analyses relative to other tetanuran groups. Indeed, even the meaning of Tyrannosauridae itself is controversial and problematic.

Unambiguous skeletal remains of tyrannosaurids are restricted to the latest Cretaceous. Alectrosaurus olseni Gilmore, 1933 was formerly thought to be of Turonian or Cenomanian age, but more recent work indicates a Campanian age for the Iren Dabasu beds where A. olseni was found (Eberth and Currie, 1993). The better-known Asian and North American taxa also derive from Campanian or Maastrichtian units ( Russell, 1970; Carpenter, 1992; Holtz, 2001a).

A partial skeleton from the Barremian of the Isle of Wight—Eotyrannus lengi— was described as a possible tyrannosauroid by Hutt et al. (2001). It shares several features, including fused nasals and very tyrannosaurid-like premaxillary dentition, with other tyrannosaurids. If alliance with tyrannosaurs is upheld, it would extend the known range of the lineage to the Barremian and its geographic range to Europe.

Isolated teeth have been used to extend the stratigraphic range of Tyrannosauridae into pre-Campanian units ( Eaton et al., 1997; Cifelli et al., 1999; Manabe, 1999; Lucas et al., 2000; Carr and Williamson, 2000). Indeed, the first tyrannosaurid remains described in the scientific literature were isolated teeth ( Leidy, 1856; Marsh, 1892). But although isolated tyrannosaurid teeth may be biogeographically and biostratigraphically informative, it is unclear how useful they are in diagnosing particular tyrannosaurid taxa, especially in light of suspected ontogenetic changes in tyrannosaurid dentition ( Carr, 1999; Carr and Williamson, 2000).

Other pre-Campanian nondental remains possibly related to Tyrannosauridae include the Jurassic Stokesosaurus from North America ( Madsen, 1974; Chure and Madsen, 1998); Itemirus , known from an isolated braincase from the Cretaceous of Asia (Kurzanov, 1976b); and Siamotyrannus , an Early Cretaceous form from Thailand based on hip and hindlimb elements (Buffetaut et al., 1997). Available material is incomplete enough to render phylogenetic analysis problematic (Holtz, 2001a) and comparisons with FMNH PR2081 difficult.

Although the focus of this study is the description of a single specimen, a review of tyrannosaurid taxonomy is relevant because other tyrannosaurs form the core of the comparative basis used here. This section focuses on alpha-level taxonomy, reviewing taxa currently thought to pertain to Tyrannosauridae in the literature, with particular emphasis to those examined for this study.

Definition of Tyrannosauridae

It seems odd that this would be a problem—after all, everyone knows what a tyrannosaur is. But as we move from a typologically-oriented nomenclatural system to one centered on phylogeny, we seek a meaning of “ Tyrannosauridae ” that ties the name to an evolutionary phenomenon rather than a collection of characters that imply some sort of “ideal” tyrant dinosaur. Unfortunately, the assemblage of animals to fall under the aegis of Tyrannosauridae is controversial.

Tyrannosauridae sensu Sereno (1998) is a stem-based group name including tyrannosauroids closer to Tyrannosaurus than to Alectrosaurus, Aublysodon , or Nanotyrannus . This would mean that Alectrosaurus and Nanotyrannus would not be tyrannosaurids, as they would be in most previous taxonomic treatments of the group (e.g., Russell, 1970; Bakker et al., 1988; Molnar et al., 1990; Carpenter, 1992; Holtz, 1994), but would instead be part of the more-inclusive Tyrannosauroidea. Tyrannosaurinae , in this system, is also a stem-based group name, this time including tyrannosaurids closer to Tyrannosaurus than to Albertosaurus , Daspletosaurus , or Gorgosaurus . No phylogenetic analysis accompanied these definitions, but the definitions imply a set of relationships congruent with that of Bakker et al. (1988), in which Nanotyrannus is a very basal member of the tyrannosaur group.

In this case, Tyrannosauridae and Tyrannosaurinae both rely on Nanotyrannus and Aublysodon as specifiers. Nanotyrannus is very likely based on an immature specimen of Tyrannosaurus rex ( Rhozhdestvensky, 1965; Carr, 1999). I agree with Carr and Williamson (2000, 2001) that the “dental tyrannosaur” Aublysodon is a nomen dubium. As currently defined, Tyrannosauridae sensu Sereno, 1998 is redundant with Tyrannosaurus rex Osborn, 1905 . Indeed, if we could work out the reticulation within T. rex , Tyrannosauridae could even conceivably refer to a subset of T. rex , because the types of Nanotyrannus and Tyrannosaurus presumably shared a common ancestor after the lineage as a whole arose. This would also render Tyrannosauridae and Tyrannosaurinae synonymous, since Tyrannosaurinae is based on Tyrannosauridae and the only tyrannosaurid would be Tyrannosaurus rex . Both names would refer to the same clade.

Holtz (2001a) defined Tyrannosauridae as a node-based group comprising the last common ancestor of Tyrannosaurus and Aublysodon and all of its descendents. Node-based groups allow more precise consolidation of a group’s diagnosis, and this particular definition avoids problems associated with Nanotyrannus , but it still uses Aublysodon as a specifier for Tyrannosauridae . If Aublysodon is undiagnosable, the definition of Tyrannosauridae is rendered ambiguous. Tyrannosaurinae is the stem-based group name including Tyrannosaurus and tyrannosaurids closer to it than to Aublysodon , but if Aublysodon is invalid, Tyrannosaurinae and Tyrannosauridae would again be redundant.

Stem-based group names are most appropriate when the membership of a group is unclear (Sereno, 1999). Some potential tyrannosaurs are controversial, but there is a core group of taxa that are universally understood to belong to this group, and a node-based group name is most appropriate. Using more than two names in the definition can circumvent the problem of Aublysodon's potential invalidity; status changes in one specifier will not change the meaning of the group. I thus amend Holtz’s definition as follows: Tyrannosauridae is the last common ancestor of Tyrannosaurus rex , Tarbosaurus bataar , Albertosaurus sarcophagus , Gorgosaurus libratus, Alectrosaurus olseni , Alioramus remotus , and Daspletosaurus torosus and all of its descendents.