Osborn Was Right

I am not going to go into too much detail with this post.

Following recent discussion on the applicability of blogs as distributors of information, I am going to try a tactic whereby I outline an argument I’ve been cultivating and testing for nearly two decades. For me, this is my only “real” pet project, my personal hypothesis. What follows is an outline with a small historical thesis, so it’s being told as a research propsal. This is something I do intend to publish on — consider this a long-form abstract.

In it, I argue that Oviraptor philoceratops, and indeed most oviraptorids, are adapted, specialized carnivores. Recent work has supported the conclusion that they were instead herbivores and/or omnivores, from the work of Dr. Lindsay Zanno and colleagues; this work provides “outs” in that specialization is often hard to find in the fossil record, and especially when diet among extant animals can be extremely variable. Moreover, despite several seeming “herbivorous” adaptations, such as lack of teeth, this should not itself result in herbivory: oviraptorosaurs lack the expanded, wide gut of an herbivore, retaining predatory features of the manus and even pes suggesting they engaged in live prey handling. There was, it seems, no direct evidence for herbivory in oviraptorids.

In 1924, Henry Fairfield Osborn supplied the world with this bizarre animal and it’s monicker — the egg thief, fond as it was of the eggs of Protoceratops andrewsi — arguing that it was, in fact, a consumer of eggs. In that work, he opined that the association of the first oviraptorid described ever to a nest it was assumed to belong to Protoceratops andrewsi, and the seeming absence of teeth and a “ventral prong” on the roof of the mouth, should have clearly determined it was an eater of eggs, a consumer of shells, and depredator on the future lives of the Gobi’s “sheep.”

AMNH 6517, holotype and only known skeletal specimen of Oviraptor philoceratops Osborn, 1924. From Norell, Gaffney & Dingus, 1992.

No matter that Osborn put forward Velociraptor mongoliensis; Tarbosaurus bataar would not be discovered for 30 years: The Gobi Desert had its first villain. Depicted in countless art books and in some recent film as a dastardly destroyer of nests, Oviraptor was that most insidious of creeps: the sneak. Blink, and you could miss him.

In the 1990s, the American Museum of Natural History returned to the Gobi Desert, on the heels of the Russians (who’d trained the Mongolians in paleontology, and helped them grow into the force they are today), the Polish (headed by some of the only prominent women paleontologists in the entire world), and the Chinese (who traded Mongolia with the Russians before Mongolian independance), who were also joined by the Canadians, who also later partnered with the Japanese in current expeditions. The Americans, who were there first when China had power, returned, and among their first discoveries located nests and nests of dinosaurs … only this time without a ceratopsian on top. Names like “Big Momma” and “Big Auntie” are now useful monickers. James Gurney used the name “Ovinutrix” as an effective replacement for the name “Oviraptor,” seeing as how it was not so much the predator as the nurse, the mother. Osborn’s specimen, as discovered by Walter Granger, was a delicate flower.

[I personally think “Ovinutrix gurneyi” is a fine name for a new oviraptorid, and no this is not the name I would use for MPC-D 100/42.]

It wasn’t the only word on the subject, though. In the 1970s and continuing into the 1980s, Mongolia’s own Barsbold Rinchen offered a contrary opinion on the diet of oviraptorids, owing as it did to complete skulls the Polish and Mongolians has uncovered, later bearing the names “Ingenia” yanshini, or Rinchenia mongoliensis, or Conchoraptor gracilis. These make up the classic Oviraptoriteers, with Oviraptor philoceratops as D’Artangan. They are the basis of mechanical analysis Barsbold used to propose an alternate food source, molluscs. Rather than eggs, they seem “overbuilt” for that, and would rather have been suited to crushing clamshell.

I have been somewhat skeptical on these conclusions on a variety of grounds, and though my education has limited my resources, I had spent that time developing work on determining what, exactly, it was that oviraptorids ate. I looked at development of edentulousness in terrestrial and marine tetrapods, development of durophagy, and specifically development of egg-eating. As it turns out, some factors do not align themselves well with an herbivorous or even molluscivorous oviraptorid, or at least for the majority or the largest ones. Several features of the skulls of oviraptorids are of supreme importance to this debate:

1. The presence of a ventral projection from each maxilla to the midline of the skull;
2. Absence of a broad and mostly flat platform expanding across the jaw or along it;
3. Absence of a broad platform on the mandible, or indeed any broad feature of the mandibular margin.

I’ve investigated the myology of the skull, focusing on the jaw adductors and discriminating jaw protractors and retractors to attempt to resolve jaw attitude during the bite stroke. I’ve also looked at and surveyed the way extant durophages and ovophages eat. When it comes to ovophages, however, I am sadly disappointed in that virtually all extreme or exclusive ovophages are snakes, and of these, oral processing is limited; it is present in some species of Oligodon (kukri snakes), and the novae dental morphology is interesting to examine, but in other ovophagous snakes it is quite distinct, and occurs almost exclusively post-orally. Durophages suited to consuming hard-shelled prey also focus on broadened dentition supported by labiolingually expanded dental platforms, and so lizards like blue-tongued skinks (or shinglebacks) Tiliqua sp., the caiman lizard Dracaena sp., the Varanus species exanthematicus or niloticus/ornatus, or the globidensine mosasauroids Globidens sp. are important in noting that consumption of hard-shelled prey often occurs on one side of the jaw at a time, rather than through both. This is possible due to weak kinesis of the skulls in all of these taxa, allowing each side to distort relevant to the stresses of the side applying pressure. It is a model that applies for Globidens, a pet project of mine as well but in smaller doses. But it suggests to me that oviraptorids are doing something different.

Indeed, a major component of the study was to determine what exclusive egg-eaters were, if any, and how they got that way. That question has a plausible answer, and it links us back to the American Museum’s second suite of expeditions to Mongolia. This was largely built on the discovery of perinate troodontid skulls in a nest of Citipati osmolskae. Although many possible explanations exist, including random association or predation on the oviraptorid eggs by a troodontid, the first hypothesis might be the most sound: oviraptorid predation on troodontids. There are carnivores, then there are carno-vores: predation upon carnivores, specifically. Likely, though, it was opportune feeding. One of the big complaints about egg-eating in oviraptorids is seasonality, and energetics. Snakes manage this by being cold-blooded, but also where Dasypeltis sp. and similar egg-eaters have slower metabolisms than less picky consumers. They make efficient use of a particularly picky diet. Alan de Quieroz and Javier Rodríguez-Robles found that predation of egg-laying species neatly bracketed predation on eggs of that species, regardless of whether the predator was a consumer of birds, or of squamates; they would prefer the eggs of the species they otherwise interact with. This has the added import that if an animal were a consumer of eggs, as a dinosaur, it might favor consumption of that egg layer first. And this pulls us back to oviraptorids as carnivores.

It is further important to note that I am fairly well-aware of the problem of exclusive egg-consumption related to potential seasonality of egg sources. First, I must be able to determine availability of eggs as a resource relative to time, and then the energetic value of the eggs in question (which would likely use bird eggs such as those of ostrich or other large-egg layers as a model, and go down from there  — and try to exclude domestic fowl, for I hope obvious reasons).

How they might have eaten also became a part of the research. For one, they cannot necessarily bite on just one side of the jaw: The hemimandibles are fixed relative to one another, and while they aren’t fused there is a tight interdigitation which keeps them in relative position. The jaw joint has brought parallels with birds in some fashion, but especially turtles and dicynodonts. This meant more work exploring diet in those groups, and jaw mechanics. It is likely that the jaw could precisely occlude on one side or the other, depending on the ability to shift articulation of the jaws; however, a distinct and prominent intercotylar ridge of the articular seemingly prevents this, as well as producing distinct distal quadrate condyles, and thus that the jaw was likely enforced in articulation without much “yaw” in the orientation. And the real hurdle.

Wishing as I did the energy to pursue things as I did when I first started on this project so long ago, and the knowledge of anatomy I now possess back then, I likely would have had substantive work already prepared. See, what I wanted then was to test my model mechanically. There were three ways I could do this: I could apply Finite Element Analysis, but this requires the software and the model, for which I would have preferred a scanned three-dimensional skull of an actual oviraptorid; I could use simple beam analysis, but this doesn’t help when testing different food stuffs; and I could use box-frame analysis, but that’s simply a larger version of beam analysis, and when you have a working group that is largely considered akinetic, pointless. I could also construct a mechanical frame in person, and fashion it out of a resin or steel duplicate of an oviraptorid skull, and have it do all the biting for me. This one is a fun one, and a practical one which allows me to do many other things which might not be entirely “scientific,” but still quite enjoyable. Like Anne Schulp’s Carinodens model, I could use it and a pressure gauge to determine the force required under a variety of materials and derive potential efficiency. Being “overbuilt” is not a problem, actually, as it permits me to provide myself an upper limit: it is obvious that the animal might be able to overcompensate its bite, especially if it retains the bite force from less soft materials. The difficulty is in having “underbuilt” jaws for a particular food: If the jaws cannot process bone, or any reasonable thickness of snail or clamshell, then there is no reason I should expect it to feed on those animals. Building the thing would have to com first, then I could feed it clams, oysters, and even eggs. The cost is beyond me at the moment, though.

But what I really want is an FEA analysis and the grounding to work on real oviraptorid skulls, to refine muscle placement and morphology. This led to figuring out how the soft-tissue of the skull might work out, and that’s why you’ve seen occasional images like this:or this:

These represent stages along the way to reconstructing a skull model which becomes the basis of further tests that affirm my premise:

Oviraptorids ate eggs, likely as a massive component of their diet, and because of this Osborn was right — on at least that premise, and the form of the name.

Part of the research has developed into reconstruction of rhamphotheca, taphonomy of tissue loss and association of cranial tissues from only bone. I’ve focused on effects around the oral margin, but have not ignored other elements of the skull. So, I’ve attempted to expand my comprehensiveness on knowledge of cranial and jaw morphology and diet in archosaurs, and indeed tetrapods, but I’m not smart enough to simply know this stuff. I lack the library to crank this stuff out or the laboratory to do practical analysis so easily. And I lack the travel to get to the places where these skulls, these oh so valuable skulls are located, so that I may examine them … at length. (I wouldn’t be wasting my time on just them, however, as these places also have specimens with actual teeth I may also desire to examine, down to the enamel microstructure and arrays of denticles.) So … it takes time.

Argument in a [Nut]Shell

So here’s what I think is going on:

Oviraptorids in general are not generalist carnivores — they are specialist carnivores, and this is reflected in their very derived jaws; they are durophages in the broad sense, in that they possess a broadly akinetic skull and crushing platforms in part of the jaw; that they were consumers of eggs, considering both that unlike typical molluscivores or seed-eaters, the floor of the oral cavity does not form a broad, hard palate-like platform, and that rather than a crushing platform of the upper jaw, it possessed what appear to be piercing structures; the jaws are wicked strong, and the jaw adductors are ridiculously huge, far more relative to jaw length than in any other dinosaur, but also that the jaws retained efficiency from the front to the back, rather than developing precision or sectoralized biting regions — implying that biting was strictly orthal (up-and-down) and the object being “bitten” was present along the jaw rather than in a section of it; high degrees of mobility of the jaws suggest that intraoral processing is advanced beyond typical dinosaurian jaws, but was largely propalinal (fore-aft) and orthal, with limited transversal (side-to-side) movement.

Thanks for listening.

Barsbold R. 1977. Кинетизы и особенности строений челюстного аппарата у овирапторов (Theropoda, Saurischia) (Kineticism and peculiarity of the jawl apparatus of oviraptorids (Theropoda, Saurischia)). Трудй – Совместная Совестко-Монгольской Палеотологыческая Зкспедитсия — Joint Soviet-Mongolian Paleontological Expedition, Transactions 4:37-47. (in Russian w/ English summary)
Cracraft, J. 1971. Caenagnathiformes: Cretaceous birds convergent in jaw mechanism to dicynodont reptiles. Journal of Paleontology 45:805-809.
de Queiroz, A. & Rodríguez-Robles, J. A. 2006. Historical contigency and animal diets: The origins of egg eating in snakes. The American Naturalist 167(5):682-692.
Gurney, J. 1992. Dinotopia: A Land Apart from Time. (Harper Collins, New York City.)
Norell, M. A., Clark, J. M., Chiappe, L. M. & D. Dashzeveg 1995. A nesting dinosaur. Nature 378:774–776.
Osborn, H. F. 1924. Three new Theropoda, Protoceratops Zone, central Mongolia. American Museum Novitates 144:1-12.
Schulp, A. s. 2005. Feeding the mechanical mosasaur: What did Carinodens eat? Netherlands Journal of Geosciences – Geologie en Nijnbouw 84(3):345-357.
Zanno, L. E. & Makovicky, P. J. 2010. Herbivorous ecomorphology and specialization patterns in theropod dinosaur evolution. Proceedings of the National Academy of Sciences, Philadelphia 108:232-237. [PDF]