Those way cool baso-eutyrannosaur skull restorations

[Gregory Paul]

Taking off from Napoli's comments, I took another run at the skull restorations. Doing these things is something I enjoy -- I am notorious for upgrading old images to keep them in use, or replace them when necessary (did that with Scelidosaurus skeletal when Norman's excellent description came out). The results of this effort are most interesting. Especially for our friend the Nanotyrannus. 

In the Z&N way cool Extended Data Fig 3 they image both sides of BM. They are not exactly the same scale, the upper right side is about 1.033 larger than the lower left (using the total length of detached front portion of the cranium of both for the scale comparison) which is not bad. To be meticulous I copied them across the street at the FedEx (on their fancy new copiers) to exactly the same scale, 1/3, the size I did my pencil tracing at. 

As explained in the methods of the BR preprint, both sides are used to fill in the other, including the teeth. Because there are minor diffs between the two sides in the size and shapes of elements, some due to normal left/right asymmetry in the living animal, minor crushing, disarticulations, the final illustration is a compromise between both sides, not an exact tracing for each element (this is normal for skull restoration illustrations). Not that it matters much, the two sides being very alike. 

N says that I have the premax somehow wrong. Yes, it is detached and somewhat too low in both pics of the preserved fossil. Which I accounted for in the BR paper Fig. 1C by moving the PM up a little to get the lower margin in line with the that of the maxillae. Obvious thing to do. And the PM looks pretty much like that of Jane in the preprint Fig. 1, so not sure what the problem is. At the other end of the skull tweaked the quadrate some. 

Next the maxillae. The depth on both sides from the top of the nasals to the ventral maxilla margin is nearly identical on both sides. Nor can I get a significant diff between the depth of the two maxillae -- irregularities in the tooth margin make it hard to be precise. And since tracing is of course easy, the max in my restoration is already the correct depth, it ain't gonna get any deeper folks. I see no evidence of general dorso-ventral flattening of the nose, it's a way crazy low snout like alioramians (probably convergence between the continents). The anterior profile of the antorbital fossa is a wee little sharper acute on the right than the left. N thinks this is crushing and that might be right (Z&N used the right side of the rostrum in their joined together image in EDFig 3 so I figured they were good with it). I altered that depression profile on the restoration to the rounder version. Even so, it is still much more acute than that of Nano, Jane, Zuni, and is much more like Stygi which also shares with BM the very sharp low anterior max not seen in any of the others, and is not a Nanotyrannus. And there is the chinless dentary (but they are not the same taxon, teeth much too diff for one thing, and Stygivenator is a bad holotype to assign anything too). 

They look like siblings as per Fig. 2E&F in the preprint. Probably at tribe or subfamily level. 

That's it for the BM skull restoration. Nothing else needs changing from the preprint as far as we can see. The length/height ratio is the same as in the published photos, as is the length/height of the maxilla. Where exactly are the actual significant differences? Seriously, where are they? 

Next Nano. Tis a royal pain in the butt. Bashed up cranium is not all that better holotype than Stygi. And all the photos (incl my own) and scans (in Witmer & Ridgely) are so variable that I was going nutty because the length/depth ratios and relative element sizes are so bloody variable. And to be honest, I am not a big fan of Nano, the skull being kind of dorky. But science marches on. Then at https://sketchfab.com/3d-models/nanotyrannus-lancensis-young-t-rex-7b0967fa27674d959647868686b6717b I got it. The posterior skull is notoriously way splayed out ventrally which helps give it that pseudo Tyrannosaurus top view profile. When people position for what they think is a direct side view pic they naturally want the skull roof to be flat on perpendicular to the viewer, so they tend to line up the lacrimal tops. But that means the side of the temporal section is then actually strongly sloped down and out towards the viewer, making the cheek look significantly shallower than it is (such is true of Fig. 4C in Witmer & Ridgely). To do the side properly it has to be flat on to the viewer. So I did that by rotating the above scan until the left side was as tall as it gets in order to get the direct lateral plan view. Another big headache has always been the errantly short lateral temporal fenestra. Rotating the scan indicates it has been pushed antero-mediall, so I placed the quadrate-quadratojugal a little aft until I got a more normal position for the quadratojugal. The result is a very deep boxy skull, blunt snouted. No apparent evidence that the rostrum was longer than preserved (same applies to BM when its to sections are put back together). Have indicated the rostrum breakage cracks which are not a gap to make folks happy. As noted in the preprint caption, the lacrimal hornlet is absent so it is not shown here, other than the solid black profile bump. Used the mandible as scanned in Fig. 15C in W&R. Because the lower jaw is jammed way up into the skull it has not been appreciated how massive it is.   

Getting the two skulls to be much more similar does not look possible. Would require serious long axis stretching of Nano which very probably is beyond what is workable, and even then the boxy rostrum shape, big jugal and quadratojugal and such would remain. And/or adding a lot of height to the entire BM skull and mandible. 

BM and Nano are about the same size so ontogeny is not involved. No established tyrannosaur genus contains nearly as variation as seen between BM and Nano. Clunky Nano has a hefty subrectagular skull, deep max and antorbital fossa, big jugal, large orbit, taller quadrate and quadratojugal, and that way heavy, rather Tyranno like lower jaw. Quite diff from Jane too. Gracile, shallow BM has the dentary that goes on forever. And that cute little adorable quadratojugal. 

Now, when the methods section in the preprint says those who disagree with the restorations need to show why in detail, Franco and I do not mean mere textual comments are sufficient. Needed are measurements, perhaps overlays when dealing with illustration of specific unrestored elements photos of which are in previous publications. Note that because restored skulls almost always contain some adjustments due to differing right and left sides because it was that way in life or later distortion, and having to assemble a disarticulated skull, and the like, overlays of overall skulls are likely to be misleading. Even complete AMNH 5027 is significantly distorted with the right and left side markedly different, and the right squamosal and quadratojugal disarticulated. Tufts-Love may be the only TT-zone tyrannosaur skull that is in really good shape and does not need restoration, my illustration in Fig. 2C in the Mesozoic paper is a straightforward tracing of the on display Tufts-Love. 

We want feedback, including that we encourage others to restore these skulls and see if they come up with something significantly different. 

I have sent images of the revised skulls to the usual suspects, if any wish to see them let me know.