Tuesday, March 06, 2012

Toroceratops: Revengeance


The following blog post is about "Toroceratops." I've taken heat for my views on "Toroceratops" in the past, and I expect this time to be no different. You should know that it is EXTREMELY ranty and, at times, kind of pissed off. I am FULLY aware that my frothing guile may have caused me to misunderstand or misconstrue certain things. If that's the case, please correct me in the comments. This rant is not so much an actual, serious rebuttal of "Toroceratops" as much as it is my personal problems with how the discussion is being carried out.

THAT SAID, please continue.

As Quilong said over on his blog, the "Toroceratops" controversy continues, but in an excellent way: each new paper that comes out not only attempts to rebut the previous one, but also provides priceless new information on Triceratops and Torosaurus in the meantime. Perhaps the best part is that this back-and-forth is available (mostly) freely to the public thanks to that bastion of our Shiny Digital Future, PLoS One:

The initial paper, unfortunately published in JVP;
Andrew Farke's redescription of Nedoceratops;
Scannella & Horner's reassertion that Nedoceratops is a transitional Trike;
Longrich & Field's attempt to suss out age based on skull suture fusion.

But here's what bugs me: In both rebuttals to "Toroceratops" (by Farke and now Longrich & Field), Scannella and/or Horner come back and basically say "those variables you tested scientifically in an attempt to create some kind of growth rubrik for Triceratops? That shit's way too variable, yo. Can't be used. Not phylogenetically significant."

"Epioccipitals? Please. We've found Trike skulls with asymmetrical numbers of epioccipitals. There is an insane amount of variation there. Orbital horn core angle? Dude, are you kidding me? Nasal horn size? Well, Nedoceratops might be clear at one end of the spectrum, but let me tell you--there's a spectrum. I've seen it!

"Skull suture fusion? Allow me to break up the party: we've got Triceratops skulls that are from big adults who don't have all their skull bones fused up. And the opposite, too: small Triceratops skulls with fused skull sutures! See what I did there? I blew your effing mind. Tatankaceratops, baby. Think about it."

Now, look: I am fully ready to accept that there are some freaking wierdos out there, but you haven't shown your work. You have given me one transitional morph: Nedoceratops--one of the most controversial ceratopsid skulls in history. It has three names. Out of hundreds of Triceratops skulls, this is the only one you can point to that has parietal fenestrae? And even then, these particular fenestrae are in wierd places that don't match up with Torosaurus really at all. And it's got that irritating squamosal fenestra that just doesn't look healthy.

Please dig through your massive collection of Triceratops skulls (over 100, apparently, in Montana alone!) and pull out another contender. I know you've got one. Don't hold out on us.

Oh, and the epioccipital thing? You'd like to think that the reason Torosaurus has more epioccipitals than Triceratops is due to two factors: epioccipital count is apparently extremely variable in that basement full of Triceratops skulls you have (thus influencing how many the eventual Torosaurus morph would have); maybe--now work with me here--the epioccipitals in Triceratops split, like an amoeba, into two distinct epioccipitals. As evidence, you're point me toward...

The episquamosal of MOR 2975. The point has been worn down, which you folks suggest is because of "splitting." Yes, good. That's the most likely answer. Has epiocciptal splitting been demonstrated in any other ceratopsid? Hell, the fact that you can't find more than one potential example of epioccipital splitting--what with your baseball stadium filled with Triceratops skulls--is just a hair troubling.

In fact, remind me which ceratopsid currently known from a good growth series (like Pachyrhinosaurus, Centrosaurus, or Ajugaceratops) demonstrates such a spectacular morphological change late in life. Now Pachyrhinosaurus, man, he goes through one helluva puberty phase. But it happens surprisingly early, and at a constant rate. And it seems like lil' Ajugaceratops provides a damn good basis for adult Ajugaceratops. And in all three examples (Centrosaurus included), the juveniles have parietal fenestrae!

For Triceratops to transition into Torosaurus requires some pretty heavy special pleading. I'm not comfortable with that. I need more evidence. I need people to show their work. You've got a boatload of Triceratops specimens? Great. Publish some kind of photoessay, either in a print or online journal or, hell, Ye Olde Internet, showing me and everybody else the full goddamn range of Triceratops variability, which as you keep saying, is insane. Prove that there is not a single effing skull variable that cannot be explained away as either age-related, and therefore not phylogenetically informative, or individual variation, which must be staggeringly huge. Midline epioccipital? Nope. Number of epiossifications? Sorry. Horn size/angle? No dice. Basic things like timing of age-related characters? Not gonna happen.

Look, Pachyrhinosaurus lakustai displays some pretty widespread individual variation, too, but at least we know it's Pachyrhinosaurus lakustai! It's not like Einiosaurus has a mid-life crisis and transforms overnight (thus hiding all transitional evidence) into Pachyrhinosaurus. That's not really how this works. Pachyrhinosaurus is diagnosable across the spectrum of individual variation. And hey, Triceratops clearly is, too. The question is whether all that individual variation has anything to do with what Torosaurus looks like.

I mean, it's one thing to say that Triceratops occupies a wide range of individual morphological variation. I can buy that. But it's tough for me when you start saying that Triceratops exhibits an incredibly wide range of individual growth timing variation. I know, I know, you've got that warehouse full of Triceratops specimens. One might be a big individual with little skull suture fusion, and one might be a small individual with lots of skull fusion. So, in theory, ANSP 15192 could just be a Trike that hit its growth spurt way too early, and Tatankaceratops is the Trike equivalent of Benjamin Button. But here's where my problem is: given that Triceratops apparently ages as fast as it goddamn pleases and exhibits more variation than Varanus, how can we adequately test the "Toroceratops" hypothesis?

Is nothing sacred? What kind of rubrik can you use when there is no rubrik? Scannella continues to argue that bone histology and microstructure is the only real way to figure out who's who, but we've already seen that the growth dynamics of Triceratops are apparently not set in stone. All you can really tell is whether Triceratops (or Torosaurus) is still growing or not. Just because your Triceratops is still growing does NOT mean that Torosaurus is the obvious next step. Can we get some postcranial, long-bone histology done? If all the Torosaurus skeletons are older than Triceratops skeletons, then slap my mouth wide open--THAT is good evidence.

Or wait, maybe it's not. After all, Triceratops wasn't keeping a firm growth schedule. ANSP 15192 might just be a Triceratops that started its transition really early, while the Triceratops individuals who appear to be older than ANSP 15192 just decided they liked having short frills. It's hard for me to believe that Triceratops figured out how to avoid the age-related growth dynamics that shackle the rest of us.

You can't sit there and tell me that, out of all the ceratopsids known and studied, and in fact most animals in the world, Triceratops was unique in its growth timing and morphology. Tyrannosaurus rex has a wierd growth curve for a tyrannosaur (or, indeed, any big theropod) but guess what? It's consistent! You can age a tyrannosaur. You apparently can't age a Triceratops. There must be certain morphological characteristics that appear at certain age ranges. Hell, it's been demonstrated for Triceratops by Horner & Goodwin! Are we abandoning that research now? The full range of variability in Triceratops apparently wipes out the morphological characters that define each age class, so are we just fucked?

I need consistency. I cannot abide it when Farke or Longrich & Field come up with testable cranial characteristics and they are basically brushed aside with this "that variable is too variable" comment. Meanwhile, Torosaurus must be Triceratops. Because THAT variable is not up for debate. It's clear as mud.

And what about all the strange variations on Triceratops that have cropped up lately? Tatankaceratops, Ojoceratops, and Eotriceratops? Are they all just somewhere on the incredibly generous bell curve of individual variation on Triceratops? I mean, they probably are! Shit, you could probably find other chasmosaurine genera that fit in that range of variation. Let's get Arrhinoceratops in that line! Aside from the slightly squared-off frill, he doesn't look too horribly different. And there are probably plenty of Triceratops specimens with slightly squared-off frills.

My point is that there needs to be a testable rubrik for "Toroceratops" to work or even not work. You can't just say "Triceratops is really variable, therefore Torosaurus." There have to be established baseline growth trajectories. I will say this: I am NOT opposed to the "Toroceratops" hypothesis. If it's true, it's intruiging and, apparently, unique among ceratopsids. But there is SO much more than needs to be done, and I don't like how the conversation is going. I don't care how much data you have if you're not sharing it with the rest of the class.

That is all. Rant over. I have a PowerPoint to work on.


Craig Dylke said...

Well put. I'm not invested at all, but I find it very frustrating how the Horner camp continues to claim things with so few examples out of that huge collection of specimens...

So yes let's have that photoessay Montana boys!

Alessio said...

Agree with your rant... I was skeptical when this teory first came up and i'm still skeptical now, so, yeah, in Nature everything is possible, but i guess that, as for now, i can still look at Toro and Trike as two separate species.

Jaime A. Headden said...

I agree that we want to make our conclusions with the full dataset available. As I said in previous posts on my website (as "Qilong", not "Quilong" -- but don't worry about it) the data is being slowly revealed. The authors' primary conclusion was to set up the metaplastic work and provide the intial hypothesis: they are now set to respond to all comers, rather than shore up the primary data they are basing their conclusion on. Personally, I would prefer a monographic survey of the material in question, and NO ONE in this debate has done this (some of the material is "not available" for histological sectioning, which is a problem), but that requires even more, more time than we've got. I am more appreciative of the effort that folks like John Scannella have gone to to communicate constructively with critics of the results of the analysis, and hope for more communication in the future.

Nick said...

I agree with Jaime, work (good or bad) takes time. Have patience.

I didn't realise it until I started doing it myself how much time this kind of thing takes.

Nima said...

And there are some other points the Hornerites miss as well.

* No Torosaurus has ever been found in association with a Triceratops. Wouldn't older animals with more reabsorbed horns seek the protection of a herd of their younger stronger kin? Why would they ever leave? With all the Triceratops remains that Horner and Scannella have dug up, you'd think a Torosaurus would also appear among them in the same exact bone beds and sites at least ONCE, if they are truly the same genus.....

* No juvenile land vertebrate ever has larger postcranial bones than an adult (density is another issue entirely)

* Remodelling of "plastic" bone does not necessarily indicate a young animal or any ontogeny with another genus that also undergoes remodelling.

* The beaks of Torosaurus and Triceratops are radically different. No ceratopsian completely remodels its feeding equipment as it reaches adulthood.

* NO ceratopsian has ever been known to ADD epoccipitals as it ages. They all reabsorb or LOSE epoccipitals as they get older.

* The "Triceratops eurycephalus" skull represents a very old Triceratops individual with almost completely reabsorbed frill studs (even more completely reabsorbed than some Torosaurus specimens), and yet NO fenestration is present. The AMNH Triceratops also appears to lack fenestration despite heavy reabsorption of epoccipitals as typical of a mature ceratopsid.

* Torosaurus is known from Utah - no Triceratops there.

* Torosaurus referred juvenile remains are known from Big Bend, Texas. Again, no Triceratops there, and these juveniles do not appear to be Triceratops. Interestingly both Utah and Texas were also Alamosaurus' stomping grounds, though it never overlapped ranges with Triceratops...

*Torosaurus is so much rarer than Triceratops. Why, if they are the same? If a Triceratops survives to the 30-foot "adolescent" stage (really these are adults) then why is it so common in that form, only to become super rare if it morphed into Torosaurus as a mere 25-foot adult? Assuming Horner were right, why would so many of the 30-footers die off before having the chance to grow up and become 'Torosaurus'? Surely they were not more vulnerable to predation than Torosaurus. They had greater size, longer horns, and far higher numbers!

* Torosaurus is sometimes found in areas that had dry climates in the Maastrichtian, while Triceratops is entirely limited to wetter swampier areas (which were not very Alamosaurus-friendly due to the loose ground). Does anyone truly believe that ontogenetic stages of the very same land vertebrate genus would have different habitat needs/specializations?

Nima said...

The fact is that Triceratops and Torosaurus don't show any real continuity in ontogeny.

Notice that with both T. horridus and T. prorsus, the old adults reabsorb the epoccipitals and end up with orbital horns strongly curved forward. Whereas Torosaurus has nearly twice as many epoccipitals, and even on some large specimens like the MOR skull they are NOT fully reabsorbed, and the horns are more or less straight or even occasinally have a double curve like in adolescent Triceratops (the ANSP skull is like this).

So essentially we have Horner and company claiming that Torosaurus was the mature form of Triceratops, when in fact in terms of frill stud and horn ontogeny, most or all known Torosaurus skulls are immature animals! And their horn shape is more like the teenage Triceratops than the adult Triceratops!

Not only that, but most old adult Triceratops skulls (such as the Carnegie Museum mount) have a proportionally shorter frill than many immature Triceratops (such as the AMNH mount). So Triceratops did not increase the frill size relative to the rest of the skull as it matured, instead it made the frill more compact relative to overall skull size - this actually leads it AWAY from Torosaurus, and in another direction entirely.

Furthermore, the most immature T. latus skull, the ANSP specimen, is only half the size of the other skulls and ontogenically much younger than many Triceratops skulls, yet still has large fenestrae, a high epoccipital count, a low sloping condor-like beak, a low snout-to-beak length ratio, and other trademark features of Torosaurus that are NOT found in Triceratops.

The end conclusion: Torosaurus is Torosaurus, not Triceratops. No matter WHAT growth stage you look at. The odd nasal and frill morphology and young epoccipitals of the huge MOR skull also indicate there may be multiple species of Torosaurus BESIDES T. utahensis.