New paper: Hart et al. 2025 on neural canal ridges in crocs

New paper out:

Hart, W.J., Atterholt, J., and Wedel, M.J. 2025. First occurrences of neural canal ridges in Crocodylia. Acta Palaeontologica Polonica 70(4): 749–753.

This one started last autumn. On October 2, 2024, I got an email from William Jude Hart, then an undergraduate at Hofstra University. At the time he was preparing to present a poster at the upcoming 2024 SVP meeting, on a caudal vertebra of a large extinct croc, Thecachampsa. Thecachampsa was a tomistomine gavialoid, closely related to the extant False Gharial, Tomistoma, which in turn is a member of Gavialidae and therefore a lot less “false” than we used to think. Thecachampsa lived on the east coast of North America in the Miocene, and it was bigger and scarier than any croc alive today. William had seen Atterholt et al. (2024), my paper with Jessie and a gang of other folks on neural canal ridges (NCRs) in non-avian dinosaurs — which we interpreted as bony spinal cord supports, following Skutschas and Baleeva (2012; see this post and this one). He had noted similar structures in his Thecachampsa caudal, and he offered to send photos.

I was interested to see photos of the Thecachampsa vert, but I was trying to moderate my excitement. A lot of crocs have bilobed neural canals, shaped like a snowman or a numeral 8, with a larger lower passage for the spinal cord and its associated meninges, and a smaller upper passage for the large supraspinal vein (a character shared with many birds — see Atterholt et al. 2025 and this post). The two passages are often divided by longitudinal bony ridges, and these can mimic bony spinal cord supports. Criteria exist to distinguish the two, as we discussed in Atterholt et al. (2024), but it’s not always super clear-cut. I wondered if the structures in the Thecachampsa vert would just be elaborate ridges between the neural and vascular compartments.

As the photo at the top of the post demonstrates, I should have had more faith in William’s perspicacity as a morphologist. When he sent the photos, my jaw hit the floor. These are the thinnest, spikiest, least ambiguous bony spinal cord supports I’ve seen in any amniote, extinct or extant. They’re up there with the rose-thorn-esque bony spikes in tuna vertebrae (see photos in this post).

I’m pretty confident that these have nothing to do with separating the supraspinal vein from the spinal cord. For one thing, this Thecachampsa vert does not have a bilobed neural canal. On the contrary, rather than having a longitudinal ridge on the side of the canal, the Thecachampsa caudal has a very subtle transverse ridge running up each side of the canal — highlighted in blue in the photo above — on which the neural canal ridge sits like a fairly abrupt summit. The photos above are closeups of the right NCR, but the same is true on the left, as you can see in the photo at the top of the post. That morphology looks a lot more consistent with bony spinal cord supports than with physically demarcating the canal into upper and lower halves.

Also, the Thecachampsa spikes are at about the midpoint of the neural canal, where we tended to find the NCRs in sauropods and other dinosaurs. That makes perfect sense if the bony spinal cord supports are remnants of embryonic myosepta, as hypothesized by Skutschas and Baleeva (2012; see this post for more discussion). The neural arch pedicles form within those myosepta, so if the bony spinal cord supports are also myoseptal remnants, they should be located near the craniocaudal midpoint of each neural arch pedicle, which is just another way of saying “about halfway down the neural canal”. Et voila, so they are, in both Thecachampsa and non-avian dinosaurs.

(Why not just check in extant crocs and see what soft tissues are tethered to these things? We’re working on that. Even if we’re wrong about NCRs being bony spinal cord supports, they’re bony somethings, presumably related to interesting soft tissues, and with anatomical and phylogenetic distributions that are very far from being fully mapped.)

Armed with the knowledge that NCRs were present in crocs, I drove out to Hemet to visit the Western Science Center. Andrew McDonald has been digging in the Menefee Formation of New Mexico for years, unearthing cool critters like the tyrannosaur Dynamoterror, the armored Invictarx, the hadrosaur Ornatops (which you’ve seen here before), some turtles (McDonald and Wolfe 2018, McDonald et al. 2018, 2021, Adrian et al. 2025) — and, oh yeah, the gigantic and terrifying Cretaceous croc Deinosuchus (Mohler et al. 2021). Thanks to the kind offices of Andrew and Alton Dooley, a good friend and Haplocanthosaurus partner in crime, I got my mitts on the Menefee Deinosuchus caudals. Two of the three vertebrae that I examined had an NCR preserved on at least one side. The third vertebra, by far the most complete externally, ironically had the worst-preserved neural canal. But the others were enough.

We had all the ammo for this paper about a year ago. William had the original discovery, the nicer specimen, and everything he needed to publish on his own, but he kindly invited me to contribute. We put together — well, William put together, with about 95% of the work — a presentation for the 5th Palaeontological Virtual Congress this spring. We looped in Jessie Atterholt for the paper, and she made a lot of improvements. And here we are.

(Incidentally, I created the silhouettes for Figure 1 myself, mostly tracing public domain images but drawing a few on my own. Why not use PhyloPic? Partly my own cussed persnickettiness, and partly because properly crediting 17 people was going to be cumbersome in such a short paper. I should make the originals available for everyone else — watch this space.)

Why do we find NCRs in some taxa but not others? Some animals are prevented from developing them: sharks don’t have a way to ossify their ligament attachments, and the denticulate ligaments of mammals don’t anchor to bone (see this post for more). Also, I suspect that NCRs are like the ossified traces of most muscle, tendon, and ligament attachments, in that they can be present but are not always present, even when the muscle, tendon, or ligament is. But that’s just kicking the can down the road — why do we see prominent NCRs in certain groups, and in certain regions of the vertebral column? We advance a hypothesis in the new paper (p. 752):

NCRs are prevalent in clades with laterally undulating locomotion (e.g., Teleosti; Skutchas and Baleev 2012), tail-driven femur retraction (e.g., Dinosauria; Atterholt et al. 2024), or both (e.g., Urodela; Wake and Lawson 1972), and absent in clades that have more rigid torsos, an absence of tail-driven femur retraction, or both, such as Anura, Aves, and Mammalia (Fig. 1A). This apparent distribution is consistent with the hypothesis that NCRs anchor the spinal cord against lateral undulatory motion.

That’s our best guess right now, but a LOT of work remains to be done. We hint at three fronts in the paper:

1. Discovery

NCRs are turning up all over the place. When we published the first NCR paper last year, they were known in salamanders but not in other lissamphibians. Now they’ve been documented in caecilians, by Santos et al. (2025), which we were able to cite in the new paper. The pace of discovery is rapid, but there is a lot of ground not yet covered. At this point, the number of non-sauropod archosaurs with published NCRs is very small — one individual each of Thecachampsa, Deinosuchus, Allosaurus, Ceratosaurus, Stegosaurus, and an indeterminate hadrosaur — but very suggestive, because Archosauria is a big, diverse clade. Not to mention all the other vertebrates. Someone is going to the be the first to document NCRs in, gosh, all the other things. Tyrannosaurs, anyone?

In particular, you may be thinking that it’s all very well for NCRs to be present in these giant extinct crocs, but what about mortal extant crocs? Stay tuned — we’re working on that, with William leading the charge. He’s a grad student now, pursuing his Master’s at East Tennessee State, and I’m confident you’ll be hearing a lot more about his work in the future.

2. Investigating soft tissues

We think the NCRs in crocs and dinos are bony spinal cord supports, but it would be very nice to have that confirmed via dissection. Also, where do the denticulate ligaments attach in vertebrae with bilobed neural canals? Could some of the longitudinal ridges in croc verts be doing double duty, dividing the vascular and neural compartments and anchoring denticulate ligaments at the same time? These are open questions, which are about one dead alligator away from being answered.

Also, as mentioned above, if the NCRs of crocs aren’t bony spinal cord supports, what the heck are they?

3. Biomechanical testing

Assuming NCRs are bony spinal cord supports, is lateral movement of the vertebral column the primary driver in their formation, just one factor among many, or a complete red herring? This is the kind of thing that could easily lend itself to logistically intensive approaches like 3D scanning and modeling, but might also get solved by just, like, pulling on things to see what happens (e.g., Baumel 1985).

Conclusion

If you want to get in on this, it’s a pretty straightforward gig: find some vertebrae, peer in the neural canals, document what you find, tell the world. If you don’t find NCRs you might find pneumatic cavities or blood vessel tracks or some totally new thing to add to the neural canal zoo. There are whole big clades of vertebrates about which we know basically nothing, neural-canal-wise, and opportunities for new discoveries abound — as our new paper shows. Come play.

References

• Adrian, B., Smith, H.F. and McDonald, A.T. 2025. A revised turtle assemblage from the Upper Cretaceous Menefee Formation (New Mexico, North America) with evolutionary and paleobiostratigraphic implications. PeerJ 13: e19340.
• Atterholt, J., Wedel, M.J., Tykoski, R., Fiorillo, A.R., Holwerda, F., Nalley, T.K., Lepore, T., and Yasmer, J. 2024. Neural canal ridges: a novel osteological correlate of postcranial neuroanatomy in dinosaurs. The Anatomical Record, 1-20. https://doi.org/10.1002/ar.25558
• Atterholt, Jessie; Burton, M. Grace; Wedel, Mathew J.; Benito, Juan; Fricano, Ellen; and Field, Daniel J. 2025. Osteological correlates of the respiratory and vascular systems in the neural canals of Mesozoic ornithurines Ichthyornis and Janavis. The Anatomical Record. http://doi.org/10.1002/ar.70070.
• Baumel, J.J. 1985. Suspensory ligaments of nerves: an adaptation for protection of the avian spinal cord. Anatomia, Histologia, Embryologia 14(1): 1-5.
• Hart, W.J., Atterholt, J., and Wedel, M.J. 2025. First occurrences of neural canal ridges in Crocodylia. Acta Palaeontologica Polonica 70(4): 749–753.
• McDonald, A.T. and Wolfe, D.G. 2018. A new nodosaurid ankylosaur (Dinosauria: Thyreophora) from the Upper Cretaceous Menefee Formation of New Mexico. PeerJ 6: e5435.
• McDonald, A.T., Wolfe, D.G. and Dooley Jr, A.C. 2018. A new tyrannosaurid (Dinosauria: Theropoda) from the Upper Cretaceous Menefee Formation of New Mexico. PeerJ 6: e5749.
• McDonald, A.T., Wolfe, D.G., Fowler, E.A.F. and Gates, T.A., 2021. A new brachylophosaurin (Dinosauria: Hadrosauridae) from the Upper Cretaceous Menefee Formation of New Mexico. PeerJ, 9, p.e11084.
• Mohler, B.F., McDonald, A.T. and Wolfe, D.G. 2021. First remains of the enormous alligatoroid Deinosuchus from the Upper Cretaceous Menefee formation, New Mexico. PeerJ 9: e11302.
• Santos, R.O., Wilkinson, M., and Zaher, H. 2025. An overview of the postcranial osteology of caecilians (Gymnophiona, Lissamphibia). The Anatomical Record, 1–26, https://doi.org/10.1002/ar.70000.
• Skutschas, P. P., & Baleeva, N. V. 2012. The spinal cord supports of vertebrae in the crown-group salamanders (Caudata, Urodela). Journal of Morphology 273(9): 1031–1041.
• Wake, D. B., & Lawson, R. 1973. Developmental and adult morphology of the vertebral column in the plethodontid salamander Eurycea bislineata, with comments on vertebral evolution in the Amphibia. Journal of Morphology 139(3): 251–299.

Source: https://svpow.com/2025/12/12/new-paper-hart-et-al-2025-on-neural-canal-ridges-in-crocs/