Semi-Aquatic Psittacosaurs

As I've been reading on Facebook, many of you readers have not received your copy of the excellent, though largely mythical, New Perspectives on Horned Dinosaurs, a book that has been promised since those mighty creatures went extinct, but has been delayed an abnormal--and some might say hilarious--number of times. It was just published last month, and I recently got my copy. As I paged through its crisp, volumous contents, I stopped when I hit a particular paper that intrigued me.

Ford, T. L. & Martin, L. D. (2010). A Semi-Aquatic Life Habit for Psittacosaurus.

The fact that Larry Martin's name was attached to the paper instantly sent multiple red flags up in my mind. His moniker is the kind of warning label one usually associates with "Dougal Dixon" and "Alan Fedducia" (who, I'm told, can't even get ornithology right). Still, I tried to repress my angst and read on, determined to see this theory through. Larry can't get bird origins right to save his life, but maybe he's on the ball when it comes to ceratopsians. Maybe they'll make a slam-dunk case.

Readers, they do not.

The authors draw on six features of Psittacosaurus to make the case for a semi-aquatic lifestyle. First thing's first, though: I should point out that a modern analogue doesn't really exist, and the authors don't point to one. They never say something like, "Psittacosaurs were Mesozoic hippos!" or "Psittacosaurus lived like a crocodile!" No, instead, their vision of everyone's favorite parrot lizard is a polytomy of various semi (or fully) aquatic modern animals as the authors struggle to come up with "semi-aquatic" reasons for Psittacosaurus' anatomy. Let's dive right in, shall we?

The first point the authors make is that Psittacosaurus is often found lying on its belly, hindlimbs akimbo in a "sprawling" position, and sometimes hyperflexed. This is apparently evidence of a semi-aquatic lifestyle. That seems like a non-starter to me, though. I can name plenty of modern animals with sprawling limbs that are NOT semi-aquatic, and even some fossil animals with sprawling limbs that have never been considered semi-aquatic. It's also worth noting that many modern semi-aquatic animals have parasagittal postures. The authors also do a poor job of explaining how an offset femur head (which Psittacosaurus apparently has) equals semi-aquatic lifestyle. They suggest it has something to do with a "swimmer's kick," because from what I gather, no animal can swim without splaying its hindlimbs and using a scissor, or breast-stroke, kick. Frogs do, after all. Crocodiles don't (they swim with their tail). Birds don't. Mammals don't. But one branch of Lissamphibia does.

So, according to Ford & Martin, the animal most closely resembling the alleged swimming mode of Psittacosaurus is a non-amniote. Real good. It's not the only amphibian analogue the authors will make.

The femur thing is also inconsistent because in Figure 23.3 of their paper (on page 332), they show a rousing series of genasaur femora. The first two are psittacosaurs. Figure A (P. xinjiangensis) does seem to have an offset femur head. Figure B (P. sibiricus) really doesn't. Figure F (P. xinjiangensis) is a picture of P. xinjiangensis' femur abducted to a comical degree, without any consideration for muscle and cartilage. It's worth noting that plenty of non-aquatic animals with parasagittal hindlimbs have somewhat offset femur heads. I have a sheep femur with an offset head. Tyrannosaurus rex has a somewhat offset head. Animals with truly sprawling postures have very offset femur heads. No living animal can move between a completely sprawling and completely parasagittal posture. But I guess Psittacosaurus could.

Let's move on the foot. The authors suggest that the foot was broad, and that large attachment scars existed on the shafts of metatarsals 1-4, suggesting that the foot was used for "more than just walking." Perhaps running, or jumping, or simply being active. Where, exactly, can I find a rubrik telling me how much muscle is required for walking, and how much is excessive? Strong feet do not necessarily equal a "swimming kick." Also, again I say, very few habitually semi-aquatic animals use a scissor kick. Frogs use that same motion for jumping. Maybe Psittacosaurus was also an excellent leaper, and all that padding and muscle was used for shock absorbtion. Also, you don't need a lot of muscle in the foot itself for swimming. You need a lot of muscle in the part of the leg that provides propulsion--the thigh. Look at moose. Moose are perfectly capable swimmers (go figure). They have hooves. But they do have enormous thigh muscles.

How about the forelimb? According to the authors, Psittacosaurus couldn't pronate or supinate, so the palms faced medially, sort of like theropods. It had a tiny little hand with three main digits and a vestigal Digit IV. The proportions of the fingers bring to mind basal theropods Eoraptor and Herrerasaurus, but shorter and stockier and probably stiffer. Ford & Martin suggest that the fingers were webbed, and that flexion of the first digit (which is very small) may have folded the web during the return stroke. So now we're talking about a doggy-paddle. A very bad doggy-paddle, because the palms face medially. Not even frogs doggy-paddle. So you've got a frog kick combined with a horrible forelimb doggy-paddle.

And then, the best part: "...the manus of psittacosaurs may have been held together by thickened skin (e.g., sea turtles)."

They even handily show a picture of a sea turtle's FLIPPER on page 334, compared to a psittacosaur paw. They look NOTHING ALIKE. Their other examples look even less like Psittacosaurus: a whale, a sea lion, and a penguin. Did you guys even look at your other figures? Furthermore, the authors muse that the limited range of motion of the forelimb could not be used for digging or food gathering, so they must have been used for swimming. Interestingly, they cite a study by Phil Senter who played around with psittacosaur limbs to figure this out. He's come to similar conclusions about dromaeosaur arms, so maybe dromaeosaurs used their feathered arms for swimming, too and that, below all those feathers was a meat-encased flipper.
So now we've got an animal who uses a frog kick and a doggy-paddle with sea turtle FLIPPERS.

Now, they do raise interesting point: many psittacosaur specimens are found with gastroliths. Why use gastroliths for breaking down food when you're already doing that with your teeth self-sharpening, leaf-slicing teeth? Perhaps the gastroliths were used for ballast, as they are in crocodiles. It is unusual that Psittacosaurus used gastroliths--the only other dinosaurs with gastroliths (to my knowledge) are sauropods and ornithomimids, neither of whom chewed their food, so gastroliths would help with digestion in this case. It's also possible that Psittacosaurus ate a wide variety of foodstuffs, some of which were not sufficiently broken down by chewing alone, and so needed further gastrolith processing. They could have also been used to achieve negative bouyancy while submerged.

Next, we move to the tail. The authors consider the tail to be ""long"" (their word is actually in quotes, as if admitting that, no, it's not really all that long). They also argue that the tail is quite deep. It is not. Hadrosaur tails are deep. Stegosaur tails are deep. Psittacosaur tails are not. The authors similarly note: "The neural spines are proportionately tall in all species and are particularly tall in P. sinensis. In P. mongoliensis and P. sinensis distal neural spines are flattened side-to-side, and fan-shaped.... Thus the tail may have been laterally compressed, which would help in swimming as in some modern lizards.... (or crocodiles)" Help me out here, folks: what's the ossified tendon situation in psittacosaurs? I'm not sure myself. But here's what I do know: it's just as likely that Psittacosaurus used its tail for swimming as any other dinosaur with a tail unhindered by ossified tendons.

This is after spending the previous paragraph comparing the tail to that of a crocodile. So, just so we're all keeping track, we've got a frog-kicking, flipper-handed doggy-paddler with the deep tail of a crocodile but the compressed tail of a lizard. Clearly, Psittacosaurus was the ultimate semi-aquatic vertebrate.

What about the nose and orbit? They are "dorsally high" and favorably compared to those of crocodilians, hippos, and capybaras. What about the skin? It's thick...and strong! So it probably strengthened the limbs and tail for swimming. You need thick skin to swim! Just ask any amphibian! Or lizard! Or semi-aquatic mammal like the capybara! You know what kinds of animals DO have thick skin? Fully aquatic animals. Ichthyosaurs and whales.

Finally, we get to the mean 'n' potatos: the one anatomical structure of Psittacosaurus I hoped they'd comment on: the tail bristles. Famously known from one Chinese specimen (SMF R 4970), the bristles have been compared to the Stage 1 protofeathers of coelurosaur theropods and porcupine quills. They seemingly adorn only the front half of the tail. There are about 100 long, gently curving quills that are deeply attached in the skin. The usual suggestion is that they were used for defense or display. Ford & Martin take it to the next level:

"We suggest that these bristles may have supported a caudal fin that was somewhat analogous to the caudal fin in modern amphibians, such as the Hellbender...the Mexican axolotl, salamanders, and tadpoles... In contrast to the flexible collagenous tails of amphibians, however, the realtively large size of psittacosaurs may have selected for a stiffer structure in order to support the large tail and enable its use in swimming."


So, instead of just doing what every OTHER amniote with a sail does (elongate the neural spines...even Platyhystrix figured that one out), Psittacosaurus had to do its own thing, evolving a complicated quill-like integument that was prone to bending, then covering those quills (which were packed together, not in a nice neat line down the spine) with a thick dermal fin. What, no ventral tail-sail? The amphibian analogy only goes so far, huh? But Psittacosaurus was only willing to go halfway--only halfway down the tail, and didn't bother to evolve solid quills--just hollow tubes. No wonder it went extinct! Lazy fuckin' dinosaur.

Finally, the authors suggest that psittacosaurs "may have fed in lakes or rivers, perhaps crawling in the mud in search of aquatic plants...However, a variety of forelimb to hindlimb relative lengths suggest that some psittacosaurs were likely more terrestrial than others...." Nice save, guys. Their final comparison is with a beaver, who I guess lives in the same kind of environment that psittacosaurs are found.


So here we have a frog-kicking, flipper-handed doggy-paddler with the tail of both a crocodile and a lizard, the skin of an ichthyosaur, the tail of a salamander, and the environmental preference of a beaver. No other animal has evolved so many different, sometimes contradictory, strategies for semi-aquatic life. Psittacosaurus really wanted to get it right. Clearly, had its reign not been cut short at the end of the Mesozoic, we might well see this creature swimming the post-Cretaceous seas:

Look, all kidding aside, the problem(s) with Ford & Martin's idea is that almost every argument they make is an example of false conclusion. This is the same kind of argument you see Horner making in regards to Tyrannosaurus rex being a scavenger. "It's got tiny little arm" does NOT mean it had to, therefore, be a scavenger. Plenty of hunting animals don't use their arms to hunt. "It couldn't run fast" does NOT mean it had to scavenge because it's prey was running slower than it was. By the same token, having dorsally high eyes does NOT mean you spend a lot of time underwater. Crocodiles and hippos actually have somewhat telescopic eyes (that is, the eyes are above the skull table). This is not the case in Psittacosaurus, and in fact the eyes of many dinosaurs are proportionately as high or higher on the skull than Psittacosaurus. Broad feet does not necessarily imply a semi-aquatic lifestyle, either. Plenty of animals without broad feet swim (dogs, hippos, moose) and plenty of dinosaurs had strong, broad feet and are not thought of as semi-aquatic (tyrannosaurs, duckbills, ankylosaurs).

The cynic in me thinks that the impetus for this paper went something like this:

Martin: "God, oh god, they can't be protofeathers! Maybe they...uh...supported some kind of skin-like structure. Like a fin! Yes, a fin! That would mean Psittacosaurus has to be semi-aquatic or something. That'll never fly. Or will it...?"

Ford: "Hey, it's got broad feet. Maybe they were webbed?"

Martin: "Look! Ancient studies on the creature support that speculation! Let's go with it!"

Just to be clear, I'm suggesting (half-jokingly) that this paper exists because Larry Martin is so hellbent to disprove the dino-bird connection. If Psittacosaurus' quills are actually structural supports for an amphibian tail, then there's no WAY they could be homologous with protofeathers, so maybe actual protofeathers are...collagen fibers after all? I don't know where to go with this. In the end, this paper just doesn't pass muster. It's a real shame that it's side-by-side with Nick Longrich's much better-researched paper about potential nocturnality and burrowing habits for Protoceratops. The man makes a good case through actual comparisons. In fact, the pedal anatomy of Psittacosaurus is a lot closer to Protoceratops than Castor, so I wouldn't be surprised if Psittacosaurus were a burrower, too. That lifestyle would ALSO explain the taxon's bad habit of being constantly buried.

Anyway, my opinion is that it's a poorly-researched bit of speculation on the part of the authors, and is nowhere NEAR a slam-dunk. Just my opinion, of course. Maybe you readers out there in Readerland have something different to say about it. I'll be moving on to the Horner & Scannella Toroceratops paper next...they actually make a pretty convincing case.