Tuesday, May 17, 2011

Tyrannosaur Tooth Count

Making the rounds right now in the media is a story about a newly described, well-preserved baby Tarbosaurus bataar that helps shed some light on the way tyrannosaurs grow, as well as touches on lingering controversies. Plenty of other blogs have already covered this, so here's a link to the backstory from Brian Switek at Dinosaur Tracking.

Interestingly, the baby Tarb has 15 teeth in the lower jaw, the same number as adult T. bataar. There has been controversy over whether or not tyrannosaurs reduced their number of teeth as they grew, particularly when it comes to the controversial taxon Nanotyrannus lancensis. Nano is known from two specimens (one is nicknamed "Jane") that, depending who you talk to, might really be simply juvenile specimens of the contemporary Tyrannosaurus rex. The differences cited to separate the two boil down to differences in the braincase (certain braincase changes were demonstrated in the new juvenile Tarbosaurus as well), and the number of teeth. Adult T. rex are usually said to have only about 12 teeth in the dentary, while specimens of N. lancensis have a whopping 17. The new juvenile Tarb suggests that in at least some tyrannosaurs, the tooth count is not drastically reduced during growth from juvenile to adult. However, as the authors caution, this same pattern may not necessarily hold true for other tyrannosaurs, even very close relatives.

And, the same pattern does not hold true for the very closely related T. rex. Also making the blog rounds these last few days has been this video of Jack Horner's talk at TEDx in Vancouver (thanks to David Orr at Love in the Time of Chasmosaurs for posting the video link!).

Here Horner gives the basics of his theory that dinosaurs are oversplit, not in the subjective taxonomic sense, but in the more objective biological sense that specimens that could be shown to belong to one species actually represent juveniles of other species. You've all heard the details before, but towards the end he shows a slide (reproduced above) that is pretty damning to the crowd who support the validity of N. lancensis. In fact, adult specimens of T. rex show a very wide ranging tooth count, and it even appears to correspond with relative size (and presumably growth stage. If anything, the number of teeth seen in N. lancensis specimens are only one or two teeth outside the range of variation for T. rex proper, a minor variant that can almost certainly be attributed to ontogeny, and not some cryptic species of giant tyrannosaur lurking in the Lancian faunas that has so far only been identified by two juvenile specimens, while the very common T. rex is known from no juveniles at all.

Anybody know the tooth count for the "Tinker" specimen, currently held in a private collection?


  1. Did we watch the same video...? You came to the exact opposite conclusion i.e. should have said: Here Horner gives the basics of his theory that dinosaurs are overSPLIT... specimens that WERE THOUGHT to belong to the DIFFERENT species actually represent juveniles of THE SAME species.

    Other than that, nice pics of the jaw.

  2. I don't think the slide indicates what Horner thinks it does. At least, that's the slide. It doesn't contradict Tsuihiji et al., for example, in arguing there is a taxonomic split in allometry (both positive and negative) of the dentary along with decreasing tooth count in taxa other than Tarbosaurus bataar, while the massive jump from 17 to 12 is not shown relative to size. I can give it that Horner is showing a non-sclaed series of dentaries which vary in alveolar count, and the demonstration is very good ... but only for his sample size, which he assumes to be only Tyrannosaurus rex. When you have an artificial bin like this being used to form a metric for other taxa, I think you start failing to prove your point in the general, even if you're right in the specific.

  3. Not long ago I've watched a recent video from "Dinosaur George" where he said something along the lines of someone doing research that involved a close exam of a new (I guess) juvenile T. rex specmimen and a comparison with Nanotyrannus, and that the authors could already tell that they were definitely different species.

  4. @Reprobus Whoops! you're right, I meant to say overSPLIT there. Revised that sentence.

    @qilong So you're saying it's more parsimonious to assume the "Samson" specimen and possibly MOR 1125 represent N. lancensis, rather than admit that the N. lancensis holotype falls within or close to the range of variation for T. rex? Essentially that there really are two similarly sized tyrannosaur taxa in the Hell Creek and that they're differentiated by tooth count alone. How would one disprove such a hypothesis? Horner's lumper hypothesis could be disproved by, as chemolithoautotrophic pointed out, discovery or identification of a juvenile tyrannosaur morph that is closer to T. rex than to N. lancensis. The splitter hypothesis is untestable as far as I can see, short of an adult, rex morph sitting on a nest containing eggs with lancensis morph juveniles. Dismissal of any potential intermediates between the two morphs as "well, we don't know which one that is" doesn't address either hypothesis. How can we ever know which species these intermediate morphs are if we ignore parsimony and assume diversity a priori?

  5. Matt, what I am saying is that here you have two studies, one which purports that these Asian Maas tyrannosaurids likely all represent one taxon, and another study that purports that all NA Maas tyrannosaurids represent one taxon, and they both differ on several particular details. Both may be correct, but when one insists the other study should conform to its own particulars on ontogeny, then I start having conniptions. There is no third study that supports the conclusions of either, save the ontogenetic studies of Carr et al., which was used to support the late-juvenility of Raptorex but which, surprisingly enough, shows it still presents unique features at a stage older than MPC-D 107/7 when in other taxa (we only have albertosaurins and bataar to go with) these features are lost. I am not taking sides here, although I must respect that Carr has, so far, the better-based of the two. This also isn't about Horner-bashing; I just think he's wrong to assume his complex is whole a priori.