New paper out today: Aureliano et al. (2022) on vertebral internal structure in the earliest saurischians

December 9, 2022

Micro-computed tomography of the vertebrae of the basalmost sauropodomorph Buriolestes (CAPPA/UFSM 0035). (A) silhouette shows the position of the axial elements. Artist: Felipe Elias. (B), three-dimensional reconstruction of the articulated cervical vertebral series and the correspondent high-contrast density slices in (D–I). Diagenetic processes partially compromised the internal structures in these cervicals. (C), 3D reconstruction of the articulated anterior dorsal vertebrae and the correspondent high-contrast density slices in (J–M). Small circumferential chambers occur both ventrally in the dorsal centrum (J) and laterally in the neural arch pedicles (D). All images indicate apneumatic chaotic trabeculae architecture. Some of the latter develop into larger chambers in the centrum (E,J,K). Nutritional foramina are broader at the bottom of the neural canal in the posterior cervicals (F,G). All slices were taken from the approximate midshaft. Anterior views in (D–H,J,K). Lateral view in (L). Ventral view in (H,I,M). Anterior/posterior orientation was defined based on the axial position, not the anatomical plane. cc circumferential chamber, ccv chamber in the centrum, ctr chaotic trabecula, d diapophysis, ltr layered trabeculae, nc neural canal, nf nutritional foramen, s neural spine. Scale bar in (A) = 500 mm; in (B–M) = 10 mm. Computed tomography data processed with 3D Slicer version 4.10. Figures were generated with Adobe Photoshop CC version 22.5.1 X64. (Aureliano et al. 2022: fig. 4)

Here’s a nice early holiday present for me: 51 weeks after our first paper together, I’m on another one with Tito Aureliano and colleagues:

Aureliano, T., Ghilardi, A.M., Müller, R.T., Kerber, L., Pretto, F.A., Fernandes, M.A.,Ricardi-Branco, F., and Wedel, M.J. 2022. The absence of an invasive air sac system in the earliest dinosaurs suggests multiple origins of vertebral pneumaticity. Scientific Reports 12:20844. https://doi.org/10.1038/s41598-022-25067-8

As before, I’m in the “just happy to be here” last author position, and quite happy to be so, too. I’ve had a productive couple of years, mostly because my colleagues keep inviting me to write a little bit, usually about pneumaticity, in exchange for a junior authorship, and that’s actually a perfect fit for my bandwidth right now. That dynamic has let me work on some cool and varied projects that have broadened my experience in satisfying ways. But enough navel-gazing!

Also as before, Tito made a really nice video that explains our findings from the paper and puts them in their broader scientific context:

For a long time now I’ve been interested in the origin of postcranial skeletal pneumaticity (PSP) in dinosaurs and pterosaurs (e.g., Wedel 2006, 2007, 2009, Yates et al. 2012, Wedel and Taylor 2013) — or is that origins, plural? Tito and crew decided to take a swing at the problem by CT scanning presacral vertebrae from the early sauropodomorphs Buriolestes and Pampadromaeus, and the herrerasaurid Gnathovorax. (Off-topic: Gnathovorax, “jaw inclined to devour”, is such a badass name that I adopted it for an ancient blue dragon in my D&D campaign.) All three taxa have shallow fossae on the lateral sides of at least some of their presacral centra, and some neural arch laminae, so they seemed like good candidates in which to hunt for internal pneumatization.

I’ll cut right to the chase: none of three have internal pneumatic chambers in their vertebrae, so if there were pneumatic diverticula present, they weren’t leaving diagnostic traces. That’s not surprising, but it’s nice to know rather than to wonder. The underlying system of respiratory air sacs could have been present in the ancestral ornithodiran, and I strongly suspect that was the case, but invasive vertebral pneumatization evolved independently in pterosaurs, sauropodomorphs, and theropods.

Detail of the vertebrae and foramina of the basalmost sauropodomorph Buriolestes (CAPPA/UFSM-0035). Cervical (A–C), anterior (D–F) and posterior (G–I) dorsal vertebrae in right lateral view. Note that nutritional foramina are present throughout the axial skeleton (dark arrows). Anterior/posterior orientation was defined based on the axial position, not the anatomical plane. Scale bar = 5 mm. Figures were generated with Adobe Photoshop CC version 22.5.1 X64. (Aureliano et al. 2022: fig. 4).

Just because we didn’t find pneumaticity, doesn’t mean we didn’t find cool stuff. Buriolestes, Pampadromaeus, and Gnathovorax all have neurovascular foramina — small holes that transmitted blood vessels and nerves — on the lateral and ventral aspects of the centra. That’s also expected, but again nice to see, especially since we think these blood vessels provided the template for invasive vertebral pneumatization in more derived taxa.

The findings I’m most excited about have to do with the internal structure of the vertebrae. Some of the vertebrae have what we’re calling a pseudo-polycamerate architecture. The polycamerate vertebrae of sauropods like Apatosaurus have large pneumatic chambers that branch into successively smaller ones. Similarly, some of the vertebrae in these Triassic saurischians have large marrow chambers that connect to smaller trabecular spaces — hence the term ‘pseudo-polycamerate’. This pseudo-polycamerate architecture is present in Pampadromaeus, but not in the slightly older Buriolestes, which has a more chaotic internal organization of trabecular spaces. So even in the apneumatic vertebrae of these early saurischians, there seems to have been an evolutionary trajectory toward more hierarchially-structured internal morphology.

Micro-computed tomography of the vertebrae of the herrerasaurid Gnathovorax (CAPPA/UFSM-0009). (A) silhouette shows the position of the axial elements. Artist: Felipe Elias. (B) 3D reconstruction of the anterior cervical vertebra and the correspondent high-contrast density slices in (D-I). Diagenetic artifacts greatly compromised the internal structures. (C) 3D reconstruction of the articulated posterior cervical vertebrae and the correspondent high-contrast density slices in (J–O). Minerals infilled between trabecular vacancies generate reddish anomalies. All images indicate irregular, chaotic, apneumatic architecture. Note the apneumatic large chambers in the centrum (ccv) and the smaller circumferential chambers at the bottom (cc). All slices were taken from the approximate midshaft. Anterior views in (D,H,I). Right lateral view in (E,L,M). Ventral view in (F,G,J,K). cc circumferential chambers, ccv chamber in the centrum, ce centrum, ctr chaotic trabeculae, d diapophysis, dia diagenetic artifact, nc neural canal, nf nutritional foramen, poz postzygapophysis, prz prezygapophysis. Scale bar in (A) = 1000 mm; in (B–O) = 10 mm. Computed tomography data processed with 3D Slicer version 4.10. Figures were generated with Adobe Photoshop CC version 22.5.1 X64.

But wait, there’s more! We also found small circumferential chambers around the margins of the centra, and what we’re calling ‘layered trabeculae’ inside the articular ends of the centra. These apneumatic trabecular structures look a heck of a lot like the circumferential pneumatic chambers and radial camellae that we described last year in a dorsal vertebra of what would later be named Ibirania (Navarro et al. 2022), and which other authors had previously described in other titanosaurs (Woodward and Lehman 2009, Bandeira et al. 2013) — see this post.

So to quickly recap, in these Triassic saurischians we find external neurovascular foramina from the nerves and vessels that probably “piloted” the pneumatic diverticula (in Mike’s lovely phrasing from Taylor and Wedel 2021) to the vertebrae in more derived taxa, and internal structures that are resemble the arrangement of pneumatic camerae and camellae in later sauropods and theropods. We already suspected that pneumatic diverticula were following blood vessels to reach the vertebrae and produce external pneumatic features like fossae and foramina (see Taylor and Wedel 2021 for a much fuller development of this idea). The results from our scans of these Triassic taxa suggests the tantalizing possibility that pneumatic diverticula in later taxa were following the vascular networks inside the vertebrae as well. 

A morphological spectrum of vertebral structure, as I thought of it 15 years ago. The Triassic saurischians described in the new paper by Aureliano et al. 2022 would sit between Arizonasaurus and Barapasaurus. (Wedel 2007: text-fig. 8)

“Hold up”, I can hear you thinking. “You can’t just draw a straight line between the internal structure of the vertebrae in Pampadromaeus, on one hand, and Apatosaurus, or a friggin’ saltasaurine, on the other. They’re at the opposite ends of the sauropodomorph radiation, separated by a vast and stormy ocean of intermediate taxa with procamerate, camerate, and semicamellate vertebrae, things like Barapasaurus, Haplocanthosaurus, Camarasaurus, and Giraffatitan.” That’s true, and the vertebral internal structure in, say, Camarasaurus doesn’t look much like either Pampadromaeus or Ibirania — at least, in an adult Camarasaurus. What about a hatchling, which hasn’t had time to pneumatize yet? Heck, what about a baby Ibirania or Rapetosaurus or Alamosaurus? Nobody knows because nobody’s done that work. There aren’t a ton of pre-pneumatization baby neosauropod verts out there, but there are some. There’s an as-yet-unwritten dissertation, or three, to be written about the vascular internal structure of the vertebrae in baby neosauropods prior to pneumatization, and in adult vertebrae that don’t get pneumatized. If caudal 20 is the last pneumatic vertebra, what does the vascular internal structure look like in caudal 21?

Cervical vertebrae of Austroposeidon show multiple internal plates of bone separated by sheets of camellae. Bandeira et al. (2016) referred to those as ‘camellate rings’, Aureliano et al. (2021) called them ‘internal plates’, and in the new paper (Aureliano et al. 2022) we call similar structures in apneumatic vertebrae ‘layered trabeculae’. (Bandeira et al. 2016: fig. 12)

To me the key questions here are, first, why does the pneumatic internal structure of the vertebrae of titanosaurs like Ibirania — or Austroposeidon, shown just above in a figure from Bandeira et al. (2016) — look like the vascular internal structure of the vertebrae of basal sauropodomorphs like Pampadromaeus? Is that (1) a kind of parallelism or convergence; (2) a deep developmental program that builds vertebrae with sheets of bone separated by circumferential and radial spaces, whether those spaces are filled with marrow or air; (3) a fairly direct ‘recycling’ of those highly structured marrow spaces into pneumatic spaces during pneumatization; or (4) some other damn thing entirely? And second, why is the vertebral internal structure of intermediate critters like Haplocanthosaurus and Camarasaurus so different from that of both Ibirania and Pampadromaeus— do the pneumatic internal structures of those taxa reflect the pre-existing vascular pattern, or are they doing something completely different? That latter question in particular is unanswerable until we know what the apneumatic internal structure is like in Haplocanthosaurus and Camarasaurus, either pre-pneumatization (ontogenetically), or beyond pneumatization (serially), or ideally both. 

A Camarasaurus caudal with major blood vessels mapped on, based on common patterns in extant tetrapods. A list of the places where blood vessels enter the bone is also a list of places where sauropod vertebrae can possibly be pneumatized. We don’t think that’s a coincidence. From Mike’s and my presentation last December at the 3rd Palaeo Virtual Congress, and this post. (Wedel and Taylor 2021)

I was on the cusp of writing that the future of pneumaticity is vascular. That’s true, but incomplete. A big part of figuring out why pneumatic structures have certain morphologies is going to be tracing their development, not just the early ontogenetic stages of pneumatization, but the apneumatic morphologies that existed prior to pneumatization. BUT we’re also nowhere near done just doing the alpha-level descriptive work of documenting what pneumaticity looks like in most sauropods. I’ll have more to say about that in an upcoming post. But the upshot is that now we’re fighting a war on two fronts — we still need to do a ton of basic descriptive work on pneumaticity in most taxa, and also need to do a ton of basic descriptive work on vertebral vascularization, and maybe a third ton on the ontogenetic development of pneumaticity, which is likely the missing link between those first two tons.

I’m proud of the new paper, not least because it raises many, many more questions than it answers. So if you’re interested in working on pneumaticity, good, because there’s a mountain of work to be done. Come join us!

References

4 Responses to “New paper out today: Aureliano et al. (2022) on vertebral internal structure in the earliest saurischians”

  1. Mike Taylor Says:

    Fantastic question at the end of that post!

    One really basic clarification:

    “Some of the vertebrae in these Triassic saurischians have large marrow chambers that connect to smaller trabecular spaces — hence the term ‘pseudo-polycamerate’.”

    How can you tell that these are marrow chambers rather than pneumatic chambers?

  2. Matt Wedel Says:

    How can you tell that these are marrow chambers rather than pneumatic chambers?

    They lack the big pneumatic external foramina that always go along with truly polycamerate construction in later theropods and sauropods. The foramina on the vertebral centra in Pampadromaeus are tiny and irregular, just like those in Buriolestes, which doesn’t have the pseudo-polycamerate internal structure.

    This gets into the IMHO not-completely-solved problem of how we tell small pneumatic foramina from neurovascular foramina. A rough rule of thumb that seems pretty accurate is that pneumatic foramina are usually wider in diameter than the thickness of the bone that they are penetrating, whereas vascular foramina are usually smaller in diameter than the thickness of the cortical bone. To a first approximation, pneumatic diverticula make doorways, whereas blood vessels make tunnels.

    Another study that we need is for someone to do some quantitative analysis on 3D scans of known pneumatic and apneumatic vertebrae in extant animals, and turn these rather subjective, rather arm-wavy distinctions into actual science — or falsify them. The qualitative size-based distinctions really break down at small scales, like in animals pigeon-sized or smaller. It would be nice if they didn’t.

  3. Adam Yates Says:

    I am reminded of those large internal spaces in Baru vertebrae that I got excited about and contacted you over, all those years ago. I guess I can now call them pseudo-camerae.

  4. Matt Wedel Says:

    Yes, absolutely! I was pretty clueless about the internal structure of vertebrae outside of Saurischia back then. Butler et al. (2012: fig. 3) illustrated some impressively large spaces in the vertebrae of extant crocs, turtles, and lizards. When someone does a more complete survey of the internal structure of reptile vertebrae, we may well find that ‘pseudo-polycamerate’ architecture is widespread, even in non-dinos.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: