You can predict how rabbits run by looking at their skulls (using this one weird trick!)
March 17, 2015
I have a new paper out today in PeerJ: “Ecological correlates to cranial morphology in leporids (Mammalia, Lagomorpha)”, with coauthors Brian Kraatz, Emma Sherratt, and Nick Bumacod. Get it free here.
I know, I know, I have fallen from grace. First Aquilops, now rabbits. And, and…skulls! I know what you’re thinking: that maybe I’m not just experimenting with the non-vertebrae of non-sauropods anymore – maybe I have an actual problem. But I don’t. I can quit anytime! You’ll see.
Actually rabbits are the freakiest of all mammals and their skulls are wicked cool. They have double incisors, with the second set right behind the first, hence the name Duplicidentata for rabbits and their close relatives. They have weird fenestrations in their maxillae (pretty much all taxa) and parietal and occipital bones (some more than others) – I’ll come back to that in a bit. And, as we discuss in our new paper, you can tell something about how a rabbit runs by looking at its skull. I thought it would be fun to relate how we figured that out, and why.
A long time ago in a graduate seminar far, far away…
1950: DuBrul, Laskin, and Moss
I met Brian Kraatz at Berkeley, where he and I were part of the cohort of students that came into the Integrative Biology Department in the fall of 2001 (faithful readers may remember Brian from his work tracking oliphaunts from, gosh, three years ago already). We took a lot of classes together, including a seminar by Marvalee Wake on evolutionary morphology. I’m pretty sure that seminar was the first time I’d actually read DuBrul and Laskin (1961), “Preadaptive potentialities of the mammalian skull: an experiment in growth and form”, or as I think of it, “How to turn a rat skull into a pika skull for fun and profit.”
Pikas (Ochotonidae) are the sister group to rabbits (Leporidae) and together these groups make up crown Lagomorpha. If you’re not familiar with pikas, Brian describes them as starting with bunny rabbits and then making them even cuter and cuddlier. Seriously, go do an image search for ‘pika’ and try not to die of cute overload.
Pikas are interesting because in many ways their skulls are intermediate between those of rodents, especially rats, and rabbits. This is maybe not surprising since rodents are the sister group to lagomorphs and are united with them in the clade Glires. E. Lloyd DuBrul was all over this rat-pika-rabbit thing back in the mid-twentieth century. Here’s an illustration from DuBrul (1950: plate 2; labels added by me):
So DuBrul knew from pikas and in particular he had the idea that you could maybe just tweak a rat skull – say, by knocking out the basicranial sutures in a baby rat to limit the growth of the skull base – and produce a gently domed skull like that of a pika. That’s what DuBrul and Laskin (1961) is all about. They did that experiment and here are their results (DuBrul and Laskin (1961: plate 3). Normal rat skull on the right, and dotted in the bottom diagram; experimental “pika-morph” rat skull on the left, and solidly outlined below.
What’s going on here morphogenetically is that the facial skeleton is getting tilted down and away from the back end of the skull. DuBrul was hip to that, too – here’s a relevant image from his 1950 paper (plate 4; labels added by me):
The common reference point against which these skulls are registered is the cranial base (the floor of the braincase just forward of the foramen magnum). Again, the pika is a pretty good intermediate between the rat and a ‘normal’ rabbit, and the dang-near-dog-sized Flemish Giant rabbit takes the lagomorph face-tilting thing to its extreme. (‘Flemish Giant rabbit’ is another entertaining image search that I will leave you as homework.)
Turns out there’s another way you can get rat skulls with different geometries: you can cut off their legs and make them walk on two feet. In an experiment that you might have trouble getting past an Institutional Animal Care and Use Committee today, Moss (1961) lopped off the forefeet or hindfeet in two experimental batches of rats, to see what effect this would have on their skulls. I’ll let Moss speak for himself on this one (Moss, 1961: pp. 301-303, emphasis in the original):
Circumnatal amputation of the forelimbs has successfully produced what are in essence “bipedal rats,” i.e., rats whose habitual mode of kinetic and static posture is permanently altered. […] The animals never became bipedal in the exact sense; that is, they never walked erect on two limbs at all times. […] Nevertheless, bipedal posture and motion were more frequently observed than in controls. […]
Animals whose hind limbs were removed represented another picture. They most certainly did not walk about on their intact forelimbs. Neither did they seem able to use their hind limb stumps as satisfactory substitutes. Their gait was not uniform and seemed to consist in a series of short pushes or hops. The most noticeable thing about them was, among other things, apparent accentuation of their cervical vertebral curvature. The sum of these changes was an upward rotation of the skull.
He wasn’t kidding: when the two groups of bipedal rats grew up, their facial skeletons were tilted relative to the control group, but in different directions (Moss, 1961: fig 3; ‘fore’ and ‘hind’ refer to which limbs the animals had left to locomote with):
Brian and I read Moss back at Berkeley, too. In fact, we were minor Moss junkies. If you’re interested in how living forms come into being, you owe it to yourself to read Moss (1968), “A theoretical analysis of the functional matrix”.
The upshot of all of this is that although neither Brian nor I had done anything with our deep (and, okay, deeply weird) knowledge of how to experimentally jack up rat skulls by the time we graduated from Berkeley, we were also primed to be thinking about how skulls attain their shapes – especially the skulls of rodents and rabbits.
2009: American Museum of Natural History
I went to the AMNH in February, 2009, to visit Brian, who was on a postdoc there at the time, and to spend one day looking at sauropods with Mike, who was over from England for a conference. What Brian and I planned to work on was the fenestration of rabbit skulls, because I’m always interested in the strategic loss of bone from skeletal structures. We spent probably half a day talking about that, and I filled a whole page in my notebook with related noodlings:
But as the sketch on the right shows, it didn’t take us long to figure out that there was something even more interesting to do with rabbit skulls. Brian had a whole shedload of rabbit skulls from different taxa sitting on his desk, and we noticed pretty quickly that one of the primary ways they varied was in the tilt of the facial skeleton relative to the back of the skull. Here’s the very next page of my notes from that trip:
The skull up top belongs to Caprolagus, the Hispid hare, which I tend to think of as the “bulldozer hare”. Seriously, it looks like a tank. It doesn’t bound or even hop, it scrambles. Here, stare into the abyss:
That rabbit will cut you, man. And just look at how flat its skull is. Even in life Caprolagus looks more rodent-y than rabbit-y. Or, more precisely, more Ochotona-y.
At the the other extreme are taxa like Bunolagus and Pronolagus, which really push the “I’m going to cute you to death by dint of my incredible bunnosity” thing:
As Brian and I started going through skulls of as many extant rabbits as we could, we noticed that the flatter-skulled taxa, with less pronounced facial tilt, tended to be the stolid, foursquare scramblers like Caprolagus, whereas the speed demons tended to have more strongly tilted skulls. It also seemed like the latter group were achieving that pronounced facial tilt by changing the geometry of the occipital region of the skull. Look back up at the red quadrilaterals I drew on the Caprolagus and Bunolagus skulls in my notebook – those mark the basioccipital ventrally and the dorsal exposure of the supraoccipital. Perhaps unsurprisingly, supraoccipital length is not the whole story; it turns out that some face-tilters get that way by having longer or more strongly arched parietals, BUT it remains true that if you find a rabbit skull with a long dorsal exposure of the supraoccipital, it will also have pronounced facial tilt.
ANYWAY, by my last night in New York, Brian and I decided that the best way to attack this would be to go down to the basement and stay up most of the night drinking beer and measuring rabbit skulls. We then tried to correlate the various measurements and angles with information on the locomotor and burrowing habits of each species. That was a big job, and after a couple of years with little forward progress (to be fair, Brian was moving across the country and taking his first tenure-track job in this interval, and I was helping birth a sauropod) we brought in Brian’s graduate student, Nick Bumacod, to do most of it. Later on the three of us were forced to acknowledge that we knew enough statistics to get ourselves into trouble but not enough to get back out. Brian had taken a geometric morphometrics course for which Emma Sherratt was a TA, and he started bugging her for help with the stats. Emma has been involved in writing new software packages for R, and we realized that the paper would be a lot stronger if we just brought her on as an author and gave her free rein with the data. Along the way Brian and Nick were giving presentations on the project everywhere from the local Western Area Vert Paleo meeting to the World Lagomorph Conference in Vienna. I got my name on four abstracts along the way, which I think is record abstract-to-paper ratio for me (especially considering that 90% of my effort on the paper was invested in a single evening in 2009 over a couple of six-packs).
But enough navel-gazing, what did we find?
2015: Rabbit skulls reveal their mode of locomotion
Our results, which you can read for free, support the hunch that Brian and I had back in 2009: slow-moving rabbits that locomote by scrambling or scampering instead of hopping tend to have less facial tilt, and faster-moving saltatorial (hopping) and cursorial (leaping and bounding) rabbits have more facial tilt. Interestingly, facial tilt does not distinguish the saltators from the cursors. So the break here is between scrambling and any kind of hopping or leaping, but not between hoppers and leapers.
Why would that be so? We don’t know for sure yet, but our top hypothesis is that if you’re moving fast, it pays to see the ground in front of you more clearly, and getting your nose down out of the way probably helps with that. This is pretty similar to the hypothesis that tyrannosaurs have pinched nasals for better binocular vision (Stevens, 2006). Rabbits are prey animals and probably can’t afford to point their eyes forward, and they may need wide nasal airways as air intakes while they’re sprinting. Tilting the nose down may be the next best thing.
Some circumstantial support for this comes from the Caviidae, the family of South American rodents that includes guinea pigs, cavies, maras, and capybaras. Here’s another plate from DuBrul (1950: plate 6) contrasting the flatter skull of the guinea pig (Cavia porcellus, top) with the decidedly arched skull of the mara or Patagonian hare (Dolichotis magellanica, bottom). Compare the mara skull to the sectioned rabbit skull in the other DuBrul plate, above – there aren’t a lot of obvious characters to separate the two (beyond the lack of double incisors in the mara).
Despite being commonly referred to as ‘hares’ and looking a lot like short-eared rabbits, maras are rodents that evolved their rabbit-like form independently. The acquisition of pronounced facial tilt in two separate lineages of small fast-moving herbivorous mammals is further evidence for the influence of locomotor mode on skull form. Irritatingly, I think we neglected to mention the guinea pig : mara :: pika : rabbit correspondence in the paper. Oh well, it wasn’t our novel observation, and we did cite DuBrul (1950).
We found lots of other interesting things, too. The PCA plots we produced from our data separate the living rabbits in unexpected ways. The length of the diastema (the toothless portion of the upper jaw) and the diameter of the auditory bulla seem to be particularly important. Diastema length isn’t too hard to figure out – most of the face-tilters have long diastemas, and the flat-heads tend to have short ones. We have no idea what bulla diameter means yet. I mean, obviously something to do with hearing, but we don’t have any ecological variables in our analysis to address that because we didn’t see it coming. So there’s a chunk of new science waiting to be done there.
Speaking of new science, or at least a relatively new thing in science, we published the full peer-review history alongside the paper, just as Mike and I did back in 2013 and as Mike did with his stand-alone paper last December. More than 80% of PeerJ authors elect to publish the peer review histories for their papers. I can’t wait until it’s 100%. PeerJ reviews are citeable – each one gets a DOI and instructions on how to cite it – and I’m tired of having my effort as a peer reviewer used once and then thrown away forever.
If you’ve been reading this whole post with gritted teeth, wondering why we were using linear measurements instead of geometric morphometrics, chillax. Brian and Emma are on that. They’ve been CT scanning the skulls of as many extant rabbits as possible and plotting landmarks for 3D morphometrics – if you were at SVP last fall, you may have seen their talk (Kraatz and Sherratt, 2014). So stay tuned for what will soon be a new ongoing series, Rabbit Skulls: The Next Generation. (Update: pilot episode here.)
I probably won’t be on that voyage. I’ve had fun getting acquainted with a completely different part of the tree of life, but there are an awful lot of shards of excellence – busted-up sauropod vertebrae, that is – crying out for my attention, and I need to stop neglecting them. I’m done with rabbit skulls, I promise. I’m going clean. (Wish me luck!)
- DuBrul, E. L. (1950). Posture, locomotion and the skull in Lagomorpha. American Journal of Anatomy, 87(2), 277-313.
- DuBrul, E. L., & Laskin, D. M. (1961). Preadaptive potentialities of the mammalian skull: an experiment in growth and form. American Journal of Anatomy, 109(2), 117-132.
- Kraatz, B., and Sherratt, E. (2014). Evolution, ecology, and modularity of the lagomorph skull. Journal of Vertebrate Paleontology, 35(3, Supplement), 162A.
- Kraatz, B.P., Sherratt, E., Bumacod, N., and Wedel, M.J. 2015. Ecological correlates to cranial morphology in leporids (Mammalia, Lagomorpha). PeerJ, 3:e844. https://dx.doi.org/10.7717/peerj.844
- Moss, M. L. (1961). Rotation of the otic capsule in bipedal rats. American Journal of Physical Anthropology, 19(3), 301-307.
- Moss, M. L. (1968). A theoretical analysis of the functional matrix. Acta Biotheoretica, 18(1), 195-202.
- Stevens, K. A. (2006). Binocular vision in theropod dinosaurs. Journal of Vertebrate Paleontology, 26(2), 321-330.