CONFUSION sparked by differences in the bone structure between juvenile and adult specimens of the same species of dinosaur is a growth area. Last month, Thomas Carr told Walking With Dinosaurs about the huge differences in body shape between young and mature tyrannosaurs that led to some specimens being wrongly identified as different species.

Meanwhile, for the past three years, the growth patterns of Triceratops and Torosaurus have fuelled a debate over whether these two types of horned dinosaur are actually one and the same.    

In 2010, John Scannella and Jack Horner of the Museum of the Rockies published a paper suggesting that Torosaurus was the mature form of Triceratops. A third dinosaur, Nedoceratops, was proposed as a potential intermediate form between the two.

Scannella and Horner's hypothesis has been rebuffed by some scientists. In 2011, Andrew Farke of the Raymond M. Alf Museum of Paleontology published a paper reinforcing the existence of Nedoceratops as a distinct genus.

Last year, Nicholas Longrich and Daniel Field of Yale university published the results of a study of the skulls of 35 Triceratops and Torosaurus specimens and concluded that among them were young Torosaurus individuals as well as old Triceratops specimens. 

Who is right? Over the past week, Andy Farke and John Scannella have been discussing the evidence in an exchange of emails...

From John Scannella:
Hi Andy,
You and I are both interested in the lives and evolution of the horned dinosaurs, but we disagree over the question of whether or not Torosaurus was a distinct genus of dinosaur or a growth stage of Triceratops. 

Given recent discoveries about developmental change in dinosaurs, I don't think we can dismiss the possibility that the fossils of Torosaurus and Triceratops actually represent growth stages of the same animal. 

Just a few years ago, Jack Horner and Mark Goodwin described the radical changes that the skull of Triceratops underwent during growth. The brow horns completely changed shape as the animal matured: they curved backwards in juveniles and curved forwards in more mature individuals. 

Also, the epiossifications (spikes) bordering the frill at the back of the skull transformed from little sharp triangles into big, flat lumps. And it's not just Triceratops: the skulls of other dinosaurs, including Pachycephalosaurus, Tyrannosaurus, and many of the duck-billed dinosaurs have been found to undergo dramatic shape changes as they grew up. 

In the past, growth stages of these dinosaurs were mistakenly thought to represent different species. The primary difference between Triceratops and Torosaurus is that the cranial frill of Torosaurus was much larger and thinner than that of Triceratops, it had two holes through the central portion of it, and a greater number of epiossifications around its border.

This could indicate differences between two closely related species, but Torosaurus was discovered and named before we knew that dinosaur skulls were capable of major developmental changes.

Given that all known specimens of Torosaurus have forward curving brow horns and flattened epiossifications, it is possible that Torosaurus specimens actually represent the fully mature skull morphology of Triceratops.
John

From Andy Farke:
Hi John,
There's no doubt to me that many dinosaurs changed their skull shapes as they grew up from hatchlings to old adults. The evidence for this is undeniable in most cases (particularly the changes early in the life of Triceratops). In fact, I would even go so far as to say that we agree on 100% of the facts and about 95% of the interpretation. 

Of course, the main sticking point is on the relationship between Torosaurus and Triceratops.

Although the hypothesis that they're different growth stages of the same animal is intriguing, I personally am not yet convinced.

One of the biggest sticking points in my mind concerns the epiparietals - the epiossifications around the central piece of the frill. Triceratops has only 5 or 6 of these epiparietals, but Torosaurus has 10 or 12.

If Torosaurus is the older form of Triceratops, then it would have doubled the number of epiparietals as it aged. But, for other horned dinosaurs with good growth series from ‘baby’ to adult, there is no evidence for such an increase.

Juvenile Centrosaurus have the same number of epiparietals as adult Centrosaurus. I don't see, at least yet, good evidence for Triceratops/Torosaurus being an exception. Of course, I would change my mind if we had a good transitional specimen that showed the epiparietals in the process of splitting.

Another problem concerns juvenile or young adult versions of Torosaurus. We wouldn't expect to see these at all if Torosaurus was the old adult form of Triceratops. But I hypothesise that there is a ‘young’ Torosaurus already known - a specimen collected back in the 1890s by Yale University. 

This skull, which is undeniably Torosaurus, has all the features we'd expect to see in a young animal. The sutures in the braincase are open (just like in young humans), the extra bony nubs on the skull are not ossified, and the frill has a fairly smooth surface texture (generally typical of young horned dinosaurs). 

This does not seem consistent with Torosaurus being an aged Triceratops - in fact, I think it best supports them being two different animals altogether.
Andy

From John Scannella:
Hi Andy,
You raise very good points. The difference in the number of epiparietals between Triceratops and Torosaurus could indicate that these were two distinct dinosaurs. But I don't think that should necessarily be the null hypothesis.

 Other chasmosaurine ceratopsids, such as Agujaceratops and Anchiceratops, appear to have increased their number of frill epiossifications as they grew up. The number and configuration of epiossifications in Triceratops varies between many specimens (and in some cases even between the right and left sides of a single skull).

Also, as you've noted in the past, two specimens of Torosaurus at the Museum of the Rockies have different numbers of epiparietals. One of the skulls has surface textures on its frill similar to those on the Yale specimen, and it also has not fused all of its epiossifications to its frill. When Jack and I examined thin slices of this skull under a microscope, we found structures consistent with a very mature individual - despite the surface texture and unfused epiossifications.

One of the great things about Triceratops is that there are hundreds of specimens, many of which were collected recently.

As we study this tremendous data set we see a lot of variation in size, fusion of cranial sutures, the number of epiossifications, and the expression of surface textures.

We also find transitional morphologies which bridge the gap between the typically short, thick frill of Triceratops and the expanded, thin frill of Torosaurus. 

As Triceratops matured, the frill became increasingly thinner in the regions where there are holes in Torosaurus. By examining the microstructure of these regions you can actually see evidence of bone resorption and that the rest of the frill was expanding at the time the animal died. 

If this process had continued, a Triceratops with a thick solid frill would have eventually had a much thinner, expanded frill with holes in it - just like Torosaurus. The skull at the Smithsonian referred to as Nedoceratops has a small hole in the region of its parietal where Triceratops has no hole and where there is a big hole in Torosaurus.

Nedoceratops could be a distinct genus of horned dinosaur that has a tiny hole in its parietal, or it could be a Triceratops (or a Torosaurus) that was in the process of developing larger holes in its frill.
John

From Andy Farke
Hi John,
It is interesting about the possibility of Agujaceratops and Anchiceratops adding epiossifications (including epiparietals) to the frill as they got bigger. I'm not yet certain if this represented growth changes, or just differences between individuals.

There certainly is some variation between these specimens, but it is only by one or two, rather than a difference of five or six. It is still interesting to note, though, that the smallest Triceratops have just about the same number of bumps on the edge of their frills for these epiossifications as the largest unequivocal Triceratops (varying by only one or two on each side of each bone). Thus, it seems a little odd to me that they don't really add many from the young juvenile to young adult stage, and all of a sudden add a whole bunch in the transformation from young adult (Triceratops-like) to old adult (Torosaurus-like).

I am very intrigued by your comments on how surface bone texture and skull fusion don't necessarily give consistent apparent ages for animals (ie. animals that look young in terms of fusion are old in terms of microscopic anatomy). 

This is a pretty cool line of work, and I would like to see a broader sample applied here (and perhaps add limb bones into the mix, which can be more tightly and unambiguously age-constrained). 

Here's hoping that someone, someday, can cut up the Yale specimens! That would certainly put my mind at ease on this point.

It is awesome to witness so many new fossils of Triceratops (and Torosaurus) with solid geological and biological data. Even though I am not personally convinced that Torosaurus is ‘just’ the old form of Triceratops, I am firmly optimistic that the new data you (and others) are collecting will help to resolve the issue. 

For my part, a good and complete frill that is unambiguously transitioning between Triceratops and Torosaurus skull shapes would be really helpful (the ‘Nedoceratops’ skull is nice, but it has some oddities that are maybe due to injury, and its geological data are thin).

Or, an unambiguously juvenile Torosaurus skull (at least one we can all agree is juvenile) could sway things the other way. No matter what, the burst of new information generated as a result of this discussion is changing the way we look at Triceratops.

I will admit that even if I end up being wrong about Torosaurus (and that's certainly possible), it would be with complete awe at how incredibly cool (and surprising) horned dinosaurs are. I know we both agree absolutely on that point!
Andy

- John Scannella is a PhD student at the Department of Earth Sciences, Montana State University, and the Museum of the Rockies. His research focuses on the growth and evolution of Triceratops.

- Dr. Andy Farke is Augustyn Family Curator of Paleontology at the Raymond M. Alf Museum of Paleontology in Claremont, California. He studies the evolution and behaviour of horned dinosaurs, as well as other research into Cretaceous ecosystems.

- Illustration of Torosaurus: Nobu Tamura