Thread by @TomHoltzPaleo

Since apparently I will get no rest from colleagues and other online friends until I comment, a 🧵 about the newly published (but available as a preprint for months) paper on Nanotyrannus 1/
mdpi.com/2813-6284/2/1/1

Background: there are series of (relatively) small specimens of tyrannosaurid from the Hell Creek Formation (photos below) which have been proposed by some as being juveniles of Tyrannosaurus rex, and by others as a distinct taxon Nanotyrannus lancensis 2/

For shorthand, I'll call the hypothesis that Nanotyrannus is a juvenile Tyrannosaurus hypothesis N=T, and the hypothesis it is distinct NnotT.

The new paper summarizes previous evidence, and includes new lines, supporting NnotT. 3/

Some of the things I like about the paper:
* A nice history of the issue for those unfamiliar with it
* New potential Nano specimens are incorporated into a study
* Attempts to summarize the morphological and morphometric distinctiveness of Nano 4/

Before I delve into the details, I should point out the basic philosophy (as I see it) behind the arguments of this paper vs. my own approach. I am not saying either is right or wrong; aspects of both approaches are a part of alpha taxonomy (assignment of specimens to species)

Longrich and Saitta (L&S) agree with everyone that there is an interesting pattern here: that the classic Nano specimens are from individuals not fully grown, and that there are no definite Tyr specimens in that size range. (They have a discussion of this at the end.) 6/

So broadly speaking, L&S are using a likelihood approach: is it likely that animals with the morphology of Nano would be able to transform into ones with the Tyr form in ontogeny? They argue the answer is "no", and thus NnotT 7/

Take, for instance, tooth count vs. tooth row length (their Fig. 18). They show there is no evidence of decreasing tooth counts in the Nanos, and that there are no intermediate forms between N and T. 8/

There is, however, another approach: the falsificationist approach. I have argued this one for literal decades on this particular subject. That is, we have two sympatric forms, one only known from juveniles and another from only adults. 9/

Given this, the null hypothesis should be N=T. It is falsifiable with the discovery of either a) a fully adult Nano that is demonstrably not Tyr, or a juvenile Tyr of the same age class as Nano that is demonstrably not Nano. We all agree such specimens are currently unknown. 10/

Okay, so let's take a look at some of the arguments in L&S. They argue that (most) other tyrannosaurid-bearing formations have at least two tyrannosaurid genera present. That is indeed true (although the Horseshoe Canyon doesn't, as they note.) 11/

But is this really a "rule"? Do ecosystems follow "rules"? If so, why do Hell Creek and other Lancian formations break other dinosaur "rules". Unlike earlier Laramidian faunas, they lack centrosaurines, parasaurolophinins, corythosaurinins, short-snouted hadrosaurids, etc. 12/

Anyway you look at it, the Hell Creek is different from earlier Laramidian faunas. (This has been noted since the 1980s by Bakker.)

More salient are the morphological changes. So next up, those. But first, let's put in a Jane picture for fun. 13/

I'm not going to deal with all of these: that would be a whole response paper, which I wouldn't do as just something online. But people wanted my thoughts so here you go:

L&S do have the most complete list of differences between N & T every documented, so this is helpful 14/

Several of the differences have to do with degree of inflation of pneumatic chambers in the facial and palatal region, and resultant changes of the external shape of these bones. Carr and others have documented how these changes do occur in other tyrannosaurid ontogenies. 15/

Similarly, they point out that the maxillae of Nano is more pointed with a low depth-to-length ratio, while those in Tyr are blunt with a higher such ratio. And they note that in Tarbosaurus there is no such low ratio in younger individuals (their Fig. 20) 16/

Sorry, here's the figure! 16.5/

However, their illustration for PIN 552-2 (second from the bottom) seems to show a deeper maxilla than the original illustration. Indeed, Rozhdestvensky (1965) specifically mentioned this. (Fig. is from Maleev 1955, reprinted in his 1974 monograph)

(Another figure note: in Fig. 21 the thumb claw of Sue (FMNH PR 2081) is shown as complete, when in fact the distal end preserved. It is still somewhat smaller than Petey's, though.) (pic from Fig. 87 in Brochu 2003) 18/

It is worth noting that other well-sampled tyrannosaurids show some of the changes we would expect in the N=T hypothesis, only not to the same extreme as the endpoint isn't as extreme. Here's Gorgosaurus growth, from Voris et al. (2021) 19/

It is worth noting that we have "Jane-equivalents" for a few other tyrannosaurid taxa: PIN 552-2 for Tarbosaurus, the aforementioned young Gorgosaurus, etc. 20/

At this point, I do want to remind the readers still with us that none of what I have said disproves NnotT!! It is still a perfectly reasonable hypothesis. The question is whether one finds the particular likelihood methods in the paper convincing or not. 21/

One of the newer contributions here is Saitta's growth curves (some from his earlier work, some from Woodward, others new here). This histological information is definitely important for addressing the growth question. Here's Figs 27-28 22/

That said, it would be interesting to plot the data with the "inferred age at time of death" rather than only "observed LAGs", since we know old LAGs get eaten up by growth of the medullary cavity (Fig. from Woodward-Ballard et al. 2013) 23/

One thing to consider in all this, though, is that the detailed osetological studies of two of the critical specimens (Jane at the Burpee and Dueling Dino specimen at NCSU) are still not published. They may yield very important data on these issues. 24/

Wrapping up (for now): this paper does gather most of the arguments in one place, and L&S have clearly stated their case. Others working in this field can evaluate it, and see how these arguments stand. Science doesn't work by "slam dunk" papers. 25/

So, no, this isn't resolved (at least not for some of us). But that is okay. That's the nature of taxonomy, and science in general.

So I'm closing (for now) with a @amatorydino picture of EITHER some Nanos checking out a Tyr kill, or kid Tyrs waiting for their share. 26/