The Great “Toroceratops” Debate

Hey there every peoples!

I have really let this blog go. And I’m not happy about it. I seem to have a small cadre of followers and I feel I let you down. School and family issues have kept me tied up. Plus I have been working on this post on and off since January. It’s the longest one I have written yet. I hope it’s worth it!

What is it about dinosaurs? I often think that the reason I stick with fossil mammals over dinosaurs (at least in an official capacity. I still want to dig them up!) is because mammals simply make more sense. With dinosaurs there’s just all this fierce debate fueled by speculation and taxonomy is a complete mass. I pretty feel that if I get my Museum enterprise underway that I’ll just find dinosaurs and let other, more qualified people study them. So then I’ve find myself in a paradox, being a casual observer of dinosaurs and yet writing a lengthy in-depth essay covering one of the most hotly contended facets of dinosaur paleontology in recent times: the Toroceratops debate.

What is the Toroceratops debate? The debate surrounds the claim that instead of being a valid species, the late Cretaceous horned dinosaur Torosaurus was really just the final growth stage of Triceratops. Now there was a lot of outcry when this was first announced, mainly because of the people’s misconceptions that Triceratops was being sunk, not Torosaurus. But beyond the public scope in the academic arena (as well as the blogosphere) the discussion has raged over the validity of the idea. I remember when I first heard about the idea I didn’t really think much of it. At the timed I figured it was the most plausible of Horner’s wild and crazy synonymys. Back then it seemed to at least make more sense than his proposed pachycephalosaur ontogeny. But now after two years and really looking at both sides I now think that now “Triceratops is Torosaurus” is just as hard to swallow. nNow make sure you don’t have anything else on your plate because this is going to be a long one.

We must first start in the beginning, even before the original “Toroceratops” paper. For the better part of the last decade the intensely studied fauna of the latest cretaceous Hell Creek formation has been undergoing a second mass extinction. What I mean by that is many of the formation’s iconic dinosaurs have been sunk, synonomized (mostly by Horner and his colleagues) with other species. The first to go was Nanoyrannus. now believed to be a juvenile T-rex (and I will admit, it’s looking pretty solid). Next up on the list were Stygimoloch and Dracorex. These two dinosaurs, known as pachycephalosaurs or boneheads, were believed to represent two distinct taxa, closely related o the contemporary Pachycephalosaurus. In 2007 Jack Horner and Mark Goodwin proposed that these three dinosaurs are actually just growth stages of the same animal. I still find it a little hard to believe because in all pretty much the whole animal kingdom juveniles are not more richly ornamented than adults. The hadrosaur Anatotitan got  reclassified as a fully mature Edmontosaurus. At that rate it was only a matter of time before the horned dinosaurs would get thrown on the chopping block. For at least a decade Jack Horner and the Museum of the Rockies have been scouring the Hell Creek formation collecting every fossil they came across. This has given them an enormous hoard of fossils that allow them to really test ideas about dinosaur biology, ecology and behavior. One of the revelations to come out of all this material with the idea that Triceratops went through a rather weird ontogeny as it grew up. Babies were born with small brow horns and little spikes around the edge of their frill called epiossifications and small, straight horns. As the dinosaur grew its horns curved upward and it developed a short triangular nose stud. As they continue to grow the horns start to slowly droop back forward into the standard Triceratops position, the nasal stud grew into a long forward pointing horn,  and the epiossifications were absorbed into the bony frill. It’s certainly an extreme pattern, especially when compared with other horn dinosaurs. But the abundance of the fossils in the study made it clear that Triceratops had one hell of a growth curve. Just how weird is the point of contention in the current debate.

tri ontogeny 1

The growth series for Triceratops, as proposed by Horner and Goodwin. From Horner and Goodwin, 2006.

tri ontogeny 2

The absorption of epiossifications as Triceratops grew. From Horner and Goodwin, 2006.

The reason Horner and Scannella’s paper has caused such fevered discussion is because it proposes that late in life Triceratops went through yet another radical change: expanding a short saddle-shaped frill into a broad flat plate, its horns straighten and took on a steeper angle from the orbit, and split its epiossifications. Triceratops is one of the few horn dinosaurs that has a solid neck frill. Everyone else has huge holes in their parietals (Just so that we are clear, the frill is made of 3 main bones: the parietal forms the center and top of the frill, while the two squamosals make up the sides. All animals have these bones. It’s just that in horned dinosaurs these bones were greatly expanded to form the frill), believed by many scientists to have helped reduce the weight of the frill. Now it’s being argued that Triceratops did have a fenestrated frill but that it only developed way late in its life. So is all this possible? Let’s have a look at the evidence.

The biggest component is the frill. Triceratops’ frill was short and rounded with a somewhat deep saddle shape. It also has no or barely present epiossifications. Compare this to Torosaurus, whose frill looks more like the typical chasmosaurine frill. It is much larger, giving Torosaurus one of the largest skulls of any land animal. It is flat with squared off edges. In the skulls that exhibit epiossifications, Torosaurus has twice as many as Triceratops. And most obvious of all are the two large fenestrae in the middle of the parietal. The backbone of Horner and Scannella’s argument is skull histology. Histology is the practice of dissecting a fossil, cutting into it to analyze bone microstructure. Using this technique they found what they claim to be evidence that the frill of Triceratops was thinning to open into the huge holes, or parietal fenestrae, that are characteristic of Torosaurus. Well does it?

tritoroparietals 1

The top left specimen is from a juvenile Triceratops. The top right comes from a sub-adult Triceratops. The bottom left is also a sub-adult Triceratops. The bottom right is a Torosaurus. From Scannella and Horner, 2010.

The one on the upper left possesses an odd depression that H&S claim is the beginning of a fenestrae. I’m not so sure, the bone looks pretty distorted, and how did they rule out pathology or more likely, an artifact of preservation?The lower left specimen has a more blatant depression; H&S claim the bone broke away during fossilization because it was so thin. The lower right seems like the best of the group. It’s a full on fenestrae with a little bit of bone on the edge. Good good, that’s a fine bit of evidence. Is it concrete? Maybe it is, maybe it isn’t. How do we know these aren’t immature Torosaurus? Maybe it had small fennestrae as a juvenile and here we see it filling out in young adult hood. H&S claim that histology shows the Triceratops frill thinning in just the right places corresponding with the fenestrae in Torosaurus (but more on that in a little bit). Plus, there was something about their proposed growth series that bugged me. It may be significant or it may just be me reading too much into it. Here is their proposed growth series:

tritoroparietals 2

The proposed ontogeny for Triceratops, from baby to full adult (in this case, Torosaurus). From Scannella and Horner, 2010.

Notice anything, particularly with specimens G through J? Here, let me highlight it for you:

tritoroparietals 2a

Highlighted version of the proposed ontogeny. From Scannella and Horner, 2010.

G’s frill has a very steep angle. H’s angle is more horizontal, like what we see in Torosaurus. Then suddenly I shoots right back up, before J goes back to the more relaxed angle. What was supposed to be going on there? I mean, the proposed growth changes aren’t extreme enough, now we’re adding in alternating frill angles? Going over that figure I kept hearing Homer Simpson in my head saying “Frill goes up. Frill goes down. Frill goes up. Frill goes down.” Or is this more of that infamous Triceratops variation? More on that later.

H&S also say that the histology revealed Triceratops horns to still be quite plastic (or rather metaplastic) so that it could in fact change into the Torosaurus horns. Here is a section of triceratops horn and frill under the microscope:

horn core histology 1

Horn core histology, beginning with a juvenile. From Scanella and Horner, 2010.

The loosely knit cells indicate that the bone is still growing. As we follow the progression, the bone becomes more and more tightly knit. The final two images are densely packed, which, according to H&S, means it is mature.

This all sounds pretty interesting. But the big question raised by all this was: is there an intermediate? H&S say they have found over 100 new Triceratops specimens since 1999, with a couple Torosaurus thrown in. Combine that with all the specimens in other museums and you have what sounds like a very good sample size (especially for a fossils species). Surely there must be at least one or even a few examples out there of a Triceratops morphing into a Torosaurus. Indeed H&S in their paper cite such a transitional fossil: Nedoceratops. Wait, what?

I’m heeeeeeeere! (took this at the Smithsonian waaaay back in 2007. Sadly was my only trip to the East Coast ever…)

Nedoceratops is, hands down, the most controversial horned dinosaur out there. It’s had 3 names slapped on it throughout its history. Scientists still can’t decide what the hell is up with all the holes in its frill. And like most fossils, it’s incomplete. So when H&S claim it’s really an intermediate between Triceratops and Torosaurus, it’s gonna raise a few eyebrows. But H&S play it straight, claiming that Nedoceratops has all the features of a maturing Triceratops caught in the act of changing into Torosaurus. And thus begins the next chapter in this epic debate.

In 2011, a year after the Toroceratops paper was published, Andrew Farke of the Raymond Alf Museum of Paleontology published a paper on the problematic skull. Except that Farke’s paper went in a different direction. He argues that Nedoceratops is indeed a valid genus, based on subtle features of the skull. He also included a rebuttal of the “Toroceratops” hypothesis. Things were about to get interesting.

Perhaps the most notorious feature of Nedoceratops is all the holes in its frill. There are 3 in total. So what? Most ceratopsians have holes in there frills. But Nedoceratops has holes in its squamosals, a very unusual feature. Plus it has only one small slit in its parietal. The holes in the squamosals were long thought to be wounds from jousting with rivals. However, Farke couldn’t find any evidence of trauma. The holes edges were smooth and the shape of them was irregular overall. The irregular shapes led Farke to conclude that they were not morphological traits and relegated them to some as yet unknown pathology. This left the hole in the parietal. It’s a small slit, but Farke was able to discern that this one was likely a morphological trait. But all ceratopsians with parietal fenestrae have two. Why doesn’t Nedoceratops? The best answer:

nedoceratops tech1

The skull of Nedoceratops. The gray portions indicate parts that are restored. From Farke, 2011.

So whether Nedoceratops had a second slit or none at all, we may never know. The rest of the skull, on the other hand, still holds clues as to the animal’s status. The nasal horn is very short and round, resembling a mound more than the classic Triceratops horn. The brow horns (the left horn is mostly reconstructed, but the right horn is mostly complete) are very erect, whereas the horns of Triceratops are angled forward. And the surface of the frill is quite vascularized and rugose in texture. Juvenile bone usually has a “pebbly” texture because it is still growing. The deep vascularization and little to no sign of cranial sutures led Farke to conclude that Nedoceratops was a mature adult and thus had no more growing to do. It seems Nedoceratops had finally gotten back on its feet.

Near the end of his paper Farke discusses the taxonomy of latest cretaceous ceratopsians in North America. He goes on to list some criticisms of the “Toroceratops” hypothesis. First, since he was able to determine that Nedoceratops is a distinct genus, “Toroceratops” loses its transitional stage. Moreover Farke notes that Triceratops has 5-6 epiossifications (the bumps around the edge of the frill) while Torosaurus has spaces for 10 to 12. He points out that in well sampled ceratopsids there is no adding of epiossifications during ontogeny (though Scanella would later claim it might be happening in Anchiceratops and Agujacertops. This was quite recent, so it obsiously hasn’t had time to be properly tested). The number of epiossifications stayed the same all through the growth trajectory, with maybe a variation of one. So if the “Toroceratops” debate is correct, it would mean Triceratops went through the very unusual process of doubling its epiossifications. Remember that Triceratops almost fully absorbed it’s epiossifications as it grew; so first it would resorb these bones and then double them. Weird. Another odd trend, if correct, is the addition of the parietal fenestrae. Again Farke points out that in ceratopsians for which we have a good growth series, the parietal fenestrae appear early in the growth trajectory. As noted earlier Triceratops is unique in having a solid frill. That it retained a solid frill for most of its life before putting a couple holes in it is highly unusual. Finally, whether the “pebbly” texture heralds the development of the fenestrae. Centrosaurus apertus has this texture on its squamosals and yet not one specimen has squamosal fenestrae. This suggests that the “pebbly” texture may not be a sure fire indicator of fenestration.

Farke had made his case. Not only did he provide evidence that Nedoceratops was a valid genus after all, but also made the first in-press challenge to “Toroceratops”. And like any good science article, it was questioned. But in this case, it was by H&S. They wasted no time tearing into the article. They published their rebuttal the very same year.

In it, they reinforce their claim that Nedoceratops is the long sought after transitional form. Before the year was out they published a paper reaffirming their original assessment that Nedoceratops is the transitional form and that Torosaurus remains sunk. They claim a minute depression in the back of the frill is in the right place to open into fenestrae:

nedoceratops tech 7

The slit on the parietal of Nedoceratops. The dotted line marks where H&S think a fenestra is forming. From Scannella and Horner, 2011.

The lack of a nasal horn? Nope, that’s apparently too variably to use as a taxonomic marker:

tritoro nasals 1

A selection of nasals from various horned dinosaurs. The first comes from a Triceratops in the Smithsonian. The second is Nedoceratops. The third is Triceratops, though it was once thought to be a separate species (Ugrosaurus). The last one is Torosaurus. From Scannella and Horner, 2011.

And now for the epiossifications. The fact that Torosaurus has twice as many as Triceratops has been  a major kink in the “Toroceratops” idea. H&S attempted to solve this by hypothesizing that Triceratops doubled the number of epiossifications as it grew (as if all the other late-in-life changes weren’t enough). Their reasoning? One is that the frill is heavily vascularised, with lots of blood flow being directed to the epiossification or the slots for them. They say this is to fuel the splitting of one epiossification into two.

Quite a tall claim. They back up this idea with MOR 2975. This is an epiossification that appears to have two peaks instead of one, perhaps indicating that it was splitting. That’s great and all, but it’s far from conclusive. Remember, these guys claim to have collected 100 new triceratops specimens. With such a huge sample size it is a little troubling that can’t find more than one example. Indeed, the fact they can’t find a transitional form other than the always dubious Nedoceratops doesn’t look good for them.

tritoro exoccipitols 2

Is this eppoccipitol in the process of splitting into two? From Scannella and Horner, 2011.

We now arrive to the most recent (see: a year ago) play in the great “Torosaurus” debate. In early 2012 Nick Longrich and Daniel Field published their paper on the issue. They did a survey of 30 Triceratops and 6 Torosaurus skulls. They were examined for a suite of morphological characteristics and came up with a simple regimen to assess synonomy. It consists of three parts: range (do they live in the same place), intermediates (where does it fit on the current understand of the growth trajectory? Does it neatly slide in or does it look out of place?), and age (how old is the animal? Do all specimens turn up younger or older than specimens of the other?). First off, they plotted the known ranges of Triceratops and Torosaurus. While most of their range overlaps, only triceratops is found in the very north, while only Torosaurus is found in the south. While there may be something going on, Longrich and Field concede that the overlap mostly likely passes the test.

tritoro map

The known ranges of Triceratops and Torosaurus. From Longrich and Field, 2012.

Next is the intermediate test. As stated for something to be an intermediate it needs to fit in its proposed slot in the growth series. Nedoceratops, they argue, doesn’t pass muster because many of the proposed transitional features don’t properly match the Torosaurus morph. This is especially true of the frill fenestration. The fenestrae of Torosaurus are on the parietal. But the fenestra (at least the outlines claimed by H&S) is small and shaped nothing like those in Torosaurus. They concluded based this and other aberrant features, contra Farke, that Nedoceratops is merely a sick Triceratops. But what about the depressions in Triceratops frills? Longrich and Field tried to find them and found that the depressions didn’t line up with Torosaurus. They straddle the parietal and squamosal, something not seen in any other ceratopsid. Lastly they  argue that in the over 100 years that fossils of these animals have been collected, including the 100 at the Museum of the Rockies, no clearly, indisputable transitional fossil has been found. Paleontology can always surprise us, but given the time and number of specimens found, it’s looking pretty bleak. Longrich and Field conclude that the synonomy argument fails this test.

tritoro comparison 3

Comparison of Triceratops (YPM 1823) with Torosaurus (ANSP 15192). Those depressions don’t look like they are in the right place. From Longrich and Field, 2012.

Finally is age, which for those very few who haven’t been following closely, is the cornerstone of the debate. But how do you age a specimen? H&S think histology is the only real way. Remember, they claim that Torosaurus has more mature looking bone than Triceratops, ergo it’s the adult stage. But Longrich and Field took a different approach. They tried to look at the anatomy of the skulls themselves. They examined specimens, quite a few of which were collected in the old days, for telltale signs of aging (Longrich said that because they weren’t cutting into the bone for histology they were able to examine a broader range of specimens). They examined bone texture along the surface of the skulls. According to their hypothesis, younger animals have a rough bone texture because they are still growing. Conversely, older animals have smoother texture because they are done adding bone. Secondly, they looked at skull suture fusion. Dinosaurs, like many other animal groups, were born with their skulls and limb bones (the ends of them rather) unfused. As the animal grows and matures, the bones fuse together and upon adulthood are fully melded together. While unable to give a precise age, the presence of unfused bones is usually an indicator that the animal wasn’t mature yet.  So Longrich and Field set about the task of examining these traits in Triceratops and Torosaurus. If all the specimens of Triceratops turned out to be younger than all the specimens of Torosaurus, then that would be pretty good evidence for synonomy. But if there was one Torosaurus that proved to be younger than a Triceratops, then that would be a good case for them being separate. A juvenile Torosaurus is often cited as being the deal breaker for the “Toroceratops” hypothesis. Did they find it?

Their study found that there are many fully mature Triceratops specimens out there (according to their criteria, at least). This rather complicates the picture. All Triceratops must still be somewhat young if they were to grow into Torosaurus. But mature Triceratops isn’t enough. There needs to be an immature Torosaurus. Well as part of their study Longrich and Field examined two historic specimens: YPM 1830 and YPM 1831. These two skulls have been on display at Yale’s Peabody Museum since the days of O.C. Marsh. Unfortunately, the skulls are partials and were heavily reconstructed with plaster, even covering up the original bone.* Andrew Farke had a chance to look at them up close and tried to discern what was real and what was not. And because he’s such an awesome guy, he posted his notes online for everyone to access:


YPM 1830, the holotype specimen for the now defunct Torosaurus “gladius”. White areas indicate what he discerned to be real fossil. From the personal notes of Andrew Farke.


YPM 1831, a pivotal part of the “Toroceratops” debate. White areas indicate what he discerned to be real fossil. From the personal notes of Andrew Farke.

*(This illustrates wonderfully why original fossils shouldn’t be restored. If you’re going to do a full reconstruction, make casts and use those. This leaves the original fossils intact and available for study, without having to try and decipher which is real and what isn’t. If only more people understood this *coughLoneStarcough*)

YPM 1831 is the real star here. First it’s enormous, something like 8 feet long! But that alone isn’t what makes it interesting. Longrich and Field found what looked like unfused margins in the skull. They also found the rough texture that suggests the animals was still growing. Because of its size, they classified it as a sub-adult, an animal that is nearing maturity. This is huge. It’s a legitimate challenge to the “Toroceratops” hypothesis since there must be no immature specimens of Torosaurus to make it work. Longrich and Field do caution, however, that this may need more work due to another specimen, this one at the Academy of Natural Sciences in Philadelphia. This Torosaurus specimen is almost half the size of YPM 1831 and yet its degree of skull fusion and bone texture suggest an adult. Does this mean that Torosaurus started out big and shrank as it matured, like that one species of frog? Of course not. It could simply be a runt. Or maybe it’s a female, like Peter Dodson once suggested, meaning Torosaurus experienced a good degree of sexual dimorphism. Nonetheless, Longrich and Field had made their challenge. Were things finally coming to a head?

immature toro 1

Front part of the skull of YPM 1831. According to Longrich and Field, the unfused bones in the nasal (as well as elsewhere) indicate an immature animal. From Longrich and Field, 2012.

toro immature 3

YPM 1831 compared to ANSP 15192. Note the striking difference in size. From Longrich and Field, 2012.

This is a debate held at Yale shortly after Longrich and Field’s paper had been published. It pits Nick Longrich against Jack Horner in a primeval battle for supremacy. The next leg of the debate is summed up well in the words of Horner in this video:

Ok, imma ‘bout to go on a little rant here. I talk about the debate and the responses to the Longrich and Field paper below. If you want to skip my rambling, go right ahead.

What really sticks in my craw is Horner’s callous attitude towards certain fossils. Where he says Yale should stuff all their specimens in storage and go get new ones. How can you say that? Though sad to say this isn’t the first instance of him expressing his opinion that certain fossils (ones without data) are worthless. The first I heard was in a debate at the London Natural History Museum. It was over whether T. rex was a predator or scavenger, but they talked about other parts of paleontology as well. At one point Horner said that “All that stuff at the American Museum? You can throw all that away.” Seriously? No wait, it gets better! The ever  prolific dinosaur writer Brian Switek interviewed Horner on his blog back in 2008 . The first commenter gave us this anecdote:

Horner: “First off, that quote is in reference to specimens taken from the field before the proper scientific studies have been completed.”

Last semester, Horner spoke to my museum studies class at Montana State University about his paleontology field work and museum work at the Museum of the Rockies. First thing, he held up a piece of bone and asked, “Anyone know what this is?” “Vertebrae,” “dinosaur bone,” “T. rex toe bone,” – various answers. Horner replied, “It’s nothing,” and proceeded to slam the bone onto a table extremely hard, breaking into pieces. “When this bone was found, nobody recorded its location or other relevant data. It’s a worthless piece of bone, because it is out of context.” Then he proceeded to stress the importance of documentation in museum work.

And then Traumador the Tyrannosaur chimed in:

Mike V-

Having fossil hunted in both the same region of Montana as Dr. Horner typical does his work, and further north in Alberta there are soooooooo many random disarticulated bones in areas that you can afford to smash some (especially those taken out of stratographic context), and still give three times as many to the local skool (though in Alberta due to the laws involved with fossil conservation it’d be easier and more LEGAL for a curator to smash a fossil than give it away to a non museum entity).

Also his point about documentation is totally founded. The Royal Tyrrell has BOXES and BOXES of useless chum bones that the public brings in. Without a location or associated material frankly hadrosaur vertabrae, rib, or digit bones are really worthless. Heck even Albertosaur teeth fall into the category of useless. They make great additions to the education departments collection of them (some 20+ of them when I left the museum), but beyond that if we don’t know where their coming from they teach us NOTHING. We can’t even safely say their Albertosaur in fact!

So fossils are not only eligible for mothballing and ditching, but apparently wanton destruction is now acceptable?  All fossils are important. Everyone is an irreplaceable record of a world that vanished long before our own. They all have something to tell us about the ancient earth. Now let’s be clear: I fully understand the importance of field data. I try to record as much as I can whenever I collect fossils. But I find it simply inexcusable to imply that just because a fossil doesn’t have that data it can be tossed aside or even destroyed. Especially if that fossil was collected a century ago when they had no way of knowing how important field data was. They can be used for education, display, or any number of things (use your imagination!). And they could still have scientific value. My buddy Eric Scott said that “some data is better than no data at all”. He then went on to explain that excluding certain things because they do not meet your standards means you’re artificially reducing your sample size. He then described and instance where a scientist only used dates obtained by only one method and then acted surprised when her results were skewed. Again, I fully acknowledge how important field data i. If your worried about people messing with that record, try to explain how they can rectify the issue. For example, on my first Manix trip, Eric Scott explained how he’d love to get a “public curation” plan going. The idea is that regular people can go look for fossils, and if they find something, they leave it in place and notify the museum. Or explain to people that there are free gps apps for smart phones. If they bring the fossils in they can at least record its position. But what you shouldn’t do is iply that certain fossils are worthless and expendable. You never know how people will interpret something. For fucks sake, look at the snafu with your original “Toroceratops” paper! I may not have been Horner’s biggest fan, but this attitude if his makes me respect him much less. I honestly cannot imagine how someone can think that fossils can be sorted into “good” and “bad” and the bad ones suffer for it. Again, what kind of message does that send to the public? And how would you feel if that huge collection you have built for over a decade were, in the future, considered worthless because you did some they didn’t approve of? Fossils are totally unique. They are pieces of our ancient heritage. Not curiosities that we can arbitrarily abandon because it may not meet their standard of worth.

The responses Horner used at that debate are the same ones he and Scanella brought up when the paper was published: “Those variables are too variable!” That’s right. They claim that the timing of skull suture fusion is highly variable. They claim to have specimens that are really small yet have fully fused skulls and very large specimens with unfused skulls (though in the video, Horner claims this is probably due to taphonomy, but whether that has to do with suture  fusion is yet to be demonstrated). They also dismiss bone texture, insisting that histology is the only sure fire way to age a specimen. But how reliable is it? In their paper, Longrich and Field point out that in both histology studies done by Horner and Scanella they only tested one specimen of Torosaurus. This is such a puny sample size and frankly to think one specimen represents the norm is shoddy work. But in the debate Longrich shows a sheep bone that has what is thought to be immature and mature textures in the same bone! In the paper, they refer to a study about how types of strain can create different looking bone microstructure. He rightly points out that histology could be useful, but it needs to be ironed out and demonstrated in modern animals before we can apply it to extinct animals with any kind of certainty.

And how does Horner respond? He just smiles and says “We have those.” “We’ve done that.” “We don’t just do that.” It’s maddening! But this has been a supreme complaint of everyone (at least on the Torosaurus side) in the debate. It feels like Horner and Scanella are using cheap tactics to deflect criticism. When Farke and Longrich & Field came up potential diagnostic features, they simply say those features are too variable. Then how do we divide taxa? How do we tell one animal from another when one of the animals apparently exhibits unparalleled individual variation? They maintain that to do so you need stratigraphy, taphonomy, ontongeny (through histology, of course), and geography to assess dinosaur taxonomy. (see, this is why field data is important, to test stuff like this) But are these late cretaceous dinosaurs so homogenous that we need all that just to be able to identify the damn thing? I know you’re not supposed to argue from personally incredulity (don’t worry, I’ll atone for it later), but I guess I’m just having trouble wrapping my head around the concept that identification rests on everything except the actually gross morphology of the animal.

And at the end of the debate Longrich blasts Horner for what has been another complaint from the Torosaurus camp. In this debate, and elsewhere, H&S claim they have the specimens that neatly prove their hypothesis and disproves everyone else. Except that no one else has seen these. In the debate Horner whines that it takes a long time. Yeah it does and it sucks, but this is scientific discourse. Other people need to be able to see it so they can assess for themselves. Also so they know you aren’t just pulling stuff out your ass. Maybe you really do have all those specimens that could lay the debate to rest. Great! Publish them! What’s holding you back? Hell as Ach noted in his rant, it deosn’t have to be super professional, it could just be a photo essay on the internet show the full range of variability you keep your 100 triceratops specimens possess. We’re not talking about building a sky scraper here, just photographing fossils. Hell, all you really need an enthusiast with a camera and a flickr account. Like me (right know I’m photographing specimens in the Raymond Alf Museum’s collection for their online database). But plain and simple citing unpublished specimens leaves the rest of us with nothing to go on but your word. This is like all those math tests I took growing up telling me to show my work. You guys need to show your work. I know I know, it takes a long time. Well until shit actually gets published don’t act surprised when people have trouble believing you about these phantom specimens.

One argument for the synonymy is that Triceratops is very common while Torosaurus is very rare. People argue that most animals don’t survive to adulthood, therefore we shouldn’t be surprised that we have very few adult Triceratops (Torosaurus). True, most animals don’t make it to maturity. But then why is Triceratops such an outlier? The majority of the fossil record presents the opposite scenario: we mostly find adults while juveniles are rare. Hell they’re often a big deal whenever they are found. Maybe there is some taphonomic bias at work in the fossil record and it would be extremely weird that Triceratops figured out how to beat it. Another lkitte quirk is just what Torosaurtus is supposed to be a mature form of. In the Yale debate Horner touches on a revelation that was produced by the Hell Creek Project. Apparently the two known species of Triceratops were not contemporary as was long thought. Instead, Triceratops horridus is the oldest, found in the lower part of the Hell Creek. Next comes an unnamed intermediate that’s in the middle. And finally we have Triceratops pororsus at the top of the formation, living in the very last moments of the Cretaceous. Which one of these is Torosaurus supposed to be the adult of? This part may take a while to answer. And Zach suggested that we may have more specimens of Torosaurus then we realize. He notes that several skulls (including the one crowning the famous skeletal mount at the American Museum of Natural History) bear many anatomical similarities to Torosaurus. Unfortunately, at least a couple of these specimens don’t have the crucial parietal. I wonder what might turn up if these fossils were subjected to histology (unlikely, since they are also historic specimens). Are these animals in fact Torosaurus? It will be very difficult to say without that critical middle bone (They were restored with solid frills because they were assumed to be Triceratops) but I think it is an interesting idea that out to be followed up on. Or not. After all they have no data, so they are just useless hunks of rock who have no business being in the scientific discussion…

So where does the debate go from here? Basically each side has laid down what it would take to convince them: either an unmistakable juvenile Torosaurus or an unambiguous transitional skull between Triceratops and Torosaurus. Both sides no doubt will continue to butt heads over the identity of these two stupendous animals. Meanwhile the debate will continue to be discussed by everyone else. In a recent paper Darren Naish and his colleagues recently wrote a paper arguing that the “species recognition” hypothesis (which Horner is also one of the biggest proponents of)  doesn’t really have the same explaining power as sexual selection. Part of the paper argues that the late life ontogenetic hypothesis is in contradiction to species recognition. For starters, if there was only one species of ceratopsian (as opposed to two or even three) living in late Cretaceous North America, who was Triceratops trying to tell itself apart from? More importantly is the question of mate selection. Fossils show that many dinosaur lineages were able to breed before reaching skeletal maturity. If the “Toroceratops” idea were correct, animals looking to breed would be able to choose from short-frilled, long-frilled, and everything in-between. Naish suggests “…the ontogenetic morphing hypothesis is in direct contradiction with the species recognition hypothesis since the latter requires that members of a species are enough alike that we can obviously identify the anatomical features that allowed them to distinguish conspecifics from heterospecifics.” An interesting idea; we’ll have to see how this might play out in the “Toroceratops” debate, if at all. It was just a passing mention in a paper dealing with sexual selection as the best explanation for weird structures in dinosaurs. I can’t wait to see what the heavy hitters in the discussion come up with!

The reconstructed skeleton of Yoshi’s Trike, a triceratops found in 2010. While the size of an elephant (no matter, an elephant would still kick his ass!), histology says he still had some growing to do. From Flickr user WernerG2011

And that pretty much sums up the great thing about this debate. It has for the most part taken place out in the open. The initial paper that started the whole mess is paywalled, sure. But every subsequent paper has been published in an open access journal, where anyone can see it. Plus exchanges have been published on the Walking with Dinosaurs site as well as the debate above, which the folks at Yale had the good sense and decency to put on the ever popular Youtube. And I think it wonderfully illustrates how science is supposed to work (ok, mostly, obviously not in the case of citing unpublished specimens). Too often I see people jump on a paper and act like it’s concrete in its assertions. They treat it as absolute instead of a stepping stone. Science is based on evidence and when new evidence comes to light we have to discard old ideas. It’s not a rigid process either. A discovery is made and then it’s basically up in the air until further testing either proves or disproves it. What would science be like if every paper was just accepted right off the bat and treated as absolute truth? It would be this stiff, boring rigmarole that shambled from one stepping stone to the next. Science is supposed to be a fluid and vibrant river of ideas, a place where hypotheses and theories race along the surface. Sometimes they hit a rock (opposing ideas or evidence) and sink but it still had a hell of a ride before then. Horner and Scanella put forward a very compelling idea, but it’s one that will require a lot of work to make it float. Their case is not as airtight as they make it out to be (then again, neither is Farkes, or Longrich & Fields, or anyone else’s). The river of science flows through parts unknown and it’s impossible to say where it will go. Supposedly Scanella is finally gonna publish that long awaited paper on Triceratops biostratigraphy this year. Personally I can’t wait to see what it says, not just about the “Toroceratops” debate but about Triceratops. And maybe in the future they’ll have more concrete evidence of their assertions. Or maybe their detractors will. Or maybe H&S won’t find any. Regardless where this ends up, I think this will go down as one of the greatest debates in the annals of paleontology. Until then, the river of science will keep on flowing.

Tis next time!

Addendum (or in this case, addendumb): It has been brought to my attention that i have it all wrong. In my effort to cover one of the most furious and dynamic debates in paleontology i seem to have completely misunderstood how science is supposed to work. First off, it takes time. A lot of time. Not everything gets published at once. So it was asinine of me to think that citing unpublished specimens in discussions outside of print was unprofessional. Scientists don’t have that much time and resources to be able to write peer-reviewed articles, so not everything is made known right away. Secondly is specimens. Thinking that all fossils are important is apparently misguided. Only ones that meet current academic standards are worth saving. Everything else can be ignored. I’d like to continue thinking that every fossil can serve a purpose and thus deserves to be saved, but i don’t have a degree or  a museum full of specimens, so obviously I’m in no position to decide what is important and what isn’t. Finally, i learned that the demon makes me an ill-informed asshole. I tried my damnedest to approach this academically. I didn’t just read books, websites, or blog posts. No, i spent a crap load of time pouring over the actual scientific articles, making sure i got everything right and and trying to emulate academic discussion. But it was all in vein. The demon got the better of me and in turn made me demonize a hard working and apparently good-natured scientist who is likely far more important to the science than I could ever hope to be. So when writing a post, it’s best to actually contact them and get their personal statement on such things so that you don’t end up talking trash about them or their work and make your yourself look like some mean-spirited amateur. If i can’t impartially describe and discuss peer-reviewed research then what hope for me is there? I was obviously way in over my head on this one. It was terribly foolish of me to think i could actually approach dinosaurs academically. Hell, i can barely hold my own with fossil mammals, so it’s my fault for not realizing it sooner.

Further addendum: Remember those squamosal fenestrae in Nedoceratops? Well i noticed them (most likely not the first) in the Torosaurus specimen at the Milwaukee Public Museum:

Torosaurus at the Milwaukee Public Museum. Unlike most other Torosaurus specimens, this one was found with some post cranial bones. From Flickr user AStrangerintheAlps.

Torosaurus at the Milwaukee Public Museum. Unlike most other Torosaurus specimens, this one was found with some post cranial bones. From Flickr user Idontlikeribena.

It’s hard to see the one on the left squamosal,  but it’s there. Both are irregular in shape. And they can’t be artifacts of preservation because if they were the museum would have filled them in (as can be seen in other parts of the restoration). I found no mention of these in the papers listed above. So we have two specimens that have squamosal fenestrae. So what do they mean? Is this just some pathology that has no bearing on the debate? Are they actual traits (or a species specific malady?) that suggest that Nedoceratops was a juvenile Torosaurus? Or could the Milwaukee specimen be the long lost second specimen of Nedoceratops? I don’t know. It’s just an interesting little fact, don’t know how that popped in there. But given how bad i did with the main post you should probably just ignore it.

New Addendum: The debate enters a new round. Maiorino et. al argue that they are separate species. They used 2d geometric morphometrics to analyze skulls of each animal. Basically, geometric morphometrics measures the skulls of an animal and then uses that data to plot potential growth trajectories (i think). Their results: that Totosaurus and Triceratops are distinct taxa. No matter what was measured, the two simply did not cluster together. Now they admit that this study was limited based on the small number of known Torosaurus skulls, but what we have now certainly makes the case that the two are indeed separate animals. No doubt the lumper camp will have a response to this and the debate will continue.

7 thoughts on “The Great “Toroceratops” Debate

  1. Oh man. I’m happy to see somebody feeling the same way I do about this hypothesis and the issues surrounding it! That was a great essay–very nicely gives the history of Toroceratops!

  2. Very interesting and informative essay! Thank you very much.
    I appreciate that you have a life. I am not interested in being inundated with frivolous postings. Quality!

  3. First off Doug, the paper is Scannella and Horner, not Horner and Scannella. It’s Scannella’s Ph.D. dissertation. Secondly, they’re working with a much larger dataset than Longrich and Field, so be patient and wait for the additional publications – there will be more. Doing histology, and doing science correctly takes time. Scannella’s also busy trying to finish his dissertation, teach, let alone graduate with his Ph.D.

    Another issue which has been glossed over here – and has been virtually ignored by others working on this subject – is that there is stratigraphic variation in Triceratops crania on top of ontogenetic variation, and that is currently being worked on at the moment as well. So, wait for the paper.

    Second to last point: Horner is mostly right about bones out of context. They are damn near useless if they lack contextual data. At some institutions there’s just no point in holding on to boxes of bone fragments from who-knows-where-or-when, so sacrificing one lousy piece of worthless bone to make a point is ‘spilt milk’, so to speak. His point about ignoring the AMNH and Yale Triceratops specimens factors into stratigraphic position – it’s an important piece of data, and without it, a specimen will be more or less irrelevant to the debate. But again, wait for the paper.

    Lastly, while you seem to make Scannella and Horner out to be these two borderline-evil fringe
    scientists, it might behoove you to actually email one of them (John is more likely to respond by email than Jack) and just ask one of them about the research, and if they could answer any questions about their 2010 paper or if they could comment on Farke’s or Longrich’s recent papers. John’s an amiable chap and quite likely to set you straight on all of this, and not too unlike Andy for that matter (John and Andy are two of the nicest dudes in dino paleo).

  4. How would one discriminate remodeled bone tissue caused by old age from ones caused by changes/variations in the distribution of stresses different regions of a bone is/are subjected to? What if the rates of bone remodeling and/or extent of remodeled bone is not always consistent within, between, and/or amongst the different skeletal elements within the same individual(with some bones having a juvenile microstructure while an immediately adjacent bone having a senescent microstructure)? What standards in bone remodeling do they use (presence/absence of certain bone tissues, relative amounts of certain bone tissues, and tissue integrity/density). Which of those possible standards is/are truly reliable? When do they apply in the ontogeny of the animal? They may come in any possible combinations (and some combinations may potentially give contradictory results. And what about performing biomolecular examinations on good-quality preserved soft tissues(I am always clamoring for biomolecular examinations. BTW, I read recently that they have found soft tissues in a 65-MYA Triceratops horridus brow horn)? Even Dr. Farke said that it would be very good if someone would perform them. Here is his reply when I asked him about Triceratops and Torosaurus biomolecular profiling


    “Also, I would like to get your opinion in regards to blog messages in several sites (including your own) about confirming or refuting the TAST(Torosaurus -as-senescent-Triceratops) hypothesis using biomolecular examinations in cases where sufficient amounts of good-quality soft tissues are preserved in ceratopid post-cranial bones.”

    Dr. Farke:

    “It’s possible given enough samples (assuming biomolecule recovery works for enough fossils), but I think we will need many, many more specimens than what we have right now. Plus, there needs to be a lot of background work on fossils from other formations. Triceratops and Torosaurus, even if separate taxa, are very closely related, and so the first step would be to
    generate a reliable standard. How much variation is there across ceratopsians as a whole? How much variation between species would we expect in things like collagen or keratin (the most likely candidates)? A necessary first step would be to demonstrate the method in extant animals
    (perhaps using samples in the current molecular databases). So, I think the use of biomolecules to answer the question is intriguing, but a long, long way off.”

  5. Pingback: SV-POW! endorses Triceratops | Sauropod Vertebra Picture of the Week

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s