Further review of the tyrannosaurids

Dr. Thomas Holtz

I'm completely confused about Albertosaurus, Gorgosaurus, and other members of the Tyrannosauridae family. Which animals would have been contemporaries of T rex? What's the family tree look like now?

First off, there are some early forms (Late Jurassic Stokesosaurus, Early Cretaceous Siamotyrannus, early Late Cretaceous Itemirus and teeth in Utah) which may represent tyrannosaurid ancestors.

Within the diagnostic Tyrannosauridae proper, there are a mess of primitive forms and a clade of very closely related forms. The primitive forms ("aublysodontsines" or "shanshanosaurines") may or may not form a monophyletic group: their monophyly is supported solely by lack of serrations on their premaxillary teeth. (Given that the immediate outgroup to tyrannosaurids ALSO lack serrations on their premaxillary teeth, there is some question as to the polarity of this character state). These primitive forms include:

Alectrosaurus olseni: early Campanian (earlier studies placed it earlier in the Late Cretaceous), eastern Asia.

Shanshanosaurus huoyanshanensis: early Maastrichtian, eastern Asia.

The "Aublysodon" complex: mostly teeth in North America, ranging from the Cenomanian to the latest Maastrichtian. Aublysodon molnari (aka "Stygivenator") is a diagnostic skeletal-based form from the late Maastricthian.

So, dependent on the character polarity, the above form either an unresolved polytomy with Tyrannosaurinae, or a monophyletic Aublysodontinae with unresolved internal structure.

The Tyrannosaurinae: a very well supported clade. In my latest analysis, Alioramus branched off first from this clade, although some have suggested it is much more derived.

Alioramus remotus: early Maastrichtian, eastern Asia.

Above the level of Alioramus, the structure is dependent on whether or not several small forms ("Gorgosaurus sternbergi", "Maleevosaurus novojilovi", "Nanotyrannus lancensis") are valid taxa or not.

If a valid taxon, "Gorgosaurus" sternbergi is the next to branch off from the other Tyrannosaurinae.

"Gorgosaurus" sternbergi: late Campanian, western North America.

If the character states for G. sternbergi are included in G. libratus, then things follow pretty much as below:

Depending on the character polarity of two ectopterygoid features (I'll be spending a lot of time with ectopterygoids this year, I fear), there may (or may not) be a monophyletic bi-specific "Albertosaurus". If these states are considered derived, then Albertosaurus sarcophagus and Gorgosaurus libratus (even if it includes G. sternbergi) form a monophyletic pair.

Gorgosaurus libratus: late Campanian, western North America.

Albertosaurus sarcophagus: early Maastrichtian, western North America.

If these states are considered primitive, then these forms and Daspletosaurus form an unresolved polytomy with respect to the Tyrannosaurus complex (see below).

Daspletosaurus falls out in most trees as the sister taxon to all remaining tyrannosaurines.

Daspletosaurus torosus: late Campanian (?to early Maastricthian, although this may be a new species and/or genus), western North America.

The remaining structure is contingent on whether "Maleevosaurus" and "Nanotyrannus" are considered valid taxa. If so, they form a basal polytomy with respect to the big boys.

Maleevosaurus novojilovi: early Maastrichtian, eastern Asia.

Nanotyrannus lancensis: late Maastricthian, western North America.

Note that the features which unite the final forms to the exclusion of these smaller guys are ADULT features. Therefore, there remains the possibility that either or both of the above are actually just juveniles of the Asian and North American Tyrannosaurus populations, respectively.

Okay, here's the interesting part, and something I will be spending a lot of professional time on the next few years (dependent on grant money).

Having examined the skull of Tyrannosaurus bataar, there is nothing preserved in this specimen which forbids it from being EITHER a specimen of Tyrannosaurus rex OR a specimen of "Tarbosaurus" efremovi. It is lacking those few portions of the skull by which rex and efremovi can be distinguished.

Therefore, if you code them as three different taxa, there is a trichotomy at this level: rex + bataar + efremovi. Alternatively, you can code bataar and efremovi as a single taxon, and you find rex + bataar. Adding "Gorgosaurus" lancinator doesn't do much to add to the situation: it (at best) tends to unite bataar and efremovi. In any case, there really isn't any good support for separating Tyrannosaurus and Tarbosaurus at a generic level: I'd almost feel confident enough to follow Jim Farlow's suggestion (in the 1993 paper on theropod energetics) and just call them all T. rex (although there are a few features by which they can be distinguished).

Tyrannosaurus rex: late Maastrichtian, western North America.

Tyrannosaurus bataar: early Maastricthian, eastern Asia.

Tyrannosaurus efremovi (?= T. bataar): early Maastrichtian, eastern Asia.

Tyannosaurus lancinator (probably = T. bataar, = T. efremovi): early Maastrichtian, eastern Asia.

Of course, the addition of new taxa (YES!!), new characters, new character states in previously poorly known basal forms (YES!!), and so forth may alter the tree topologies discussed here.