Pneumatic Air Sacs in Dinosaurs

Following all that work I’ve been doing on anatomy in oviraptorids, it should not come as a surprise that I am looking for ways to effectively represent this amassed data in digestible chunks. I don’t always want to write novellas discussing a topic, like a huge book review. Today, a small look at the various pneumatic structures found in theropod dinosaurs, especially oviraptorids.Everything but Skin and Bones

The pneumatic system of any given animal is comprised mostly of the lung. Indeed, all the little pockets that arise comes from expansions of the primary bronchus, a tissue that surrounds the lung itself, and it is this tissue that forms little sacs, the “pneumatic diverticulae,” that ends up through the body, and inside the bones, of various animals; the cranial sinuses, however, arise from the nasopharyngeal passage within the skull, rather than directly from the lung. Crocodilians have them, birds certainly have them, and so do pterosaurs. It stands to reason that dinosaurs, in all their diversity, have them, though for the moment ornithischian dinosaurs only seem to exhibit the paranasal sinus system (1, in red above). Most archosaurs, including pterosaurs, seem to start out with primarily only cranial sinuses, then demonstrate expansion of the lungs into the dorsal vertebrae (4, in dark yellow above), shoulder, and then neck. The latter two come from the clavicular air sac (3, 3a and 3b, in orange above). The belly and ribs and internal spaces of the body cavity are pneumatized from the thoracic air sacs (5 and 6, in shades of green above), while the hips and hindlimbs, sacrum, and tail seem to be pneumatized from the abdominal air sac (7, 7a and 7b, in teal above). The lung, shown in yellow, is the only gas-exchanging portion of the system, and the trachea (shown in grey), is a non-collapsible tube for bringing air to and from the lungs.

It’s probably not the same in every animal. Indeed, birds, pterosaurs and sauropods seem to have developed their degrees of pneumatization independently from one another, but given the lack of any ancestor each shares with as extensive a pneumatic system (crocs lack anything as progressive as seen above), it is likely that only some of thoracic air sacs were present plesiomorphically, to then be elaborated upon by each clade separately. While birds and pterosaurs both have brachial air sacs, pneumatization of the forelimb in pterosaurs doesn’t seem to occur in basal clades, and some lineages of pterosaur may have elaborated these structures in the forelimb and wing membrane differently from one another, especially as the presence and position of pneumatic foramina differ amongst pterodactyloids where the tissues appear extensively.

There is reasonable certainty that the basic theropod air sac plan (shown above) follows the same Bauplan as does that of birds, and that birds merely elaborated upon this plan from their ancestors. The presence of most pneumatic systems in archosaurs is a tricky thing, based largely on inference; however, it is possible to directly determine the presence of an air sac from the presence of pneumatically hollow bones (pterosaurs, and saurischian dinosaurs including birds), pneumatic foramina (structures not apparently serving a neurovascular purpose, seen in crocs, some archosauromorphan “pre-crocs,” pterosaurs, and saurischian dinosaurs including birds), or the structures that form on bone surfaces from diverticulae pushing against them, most dramatically those of the vertebral centra in sauropods.

Passing from the front of the body (head) to the rear, but not in evolutionary diverging order or basal system, pneumatic systems invading the bones of archosaurs are:

1. Paranasal Sinus System (PNS)

This system is shown in Red.

Starting with the antorbital air sac, this structure invades the bones surrounding it (maxilla, lacrimal, nasal, palatine, ectopterygoid, pterygoid), and from there into adjacent bones (premaxilla from maxilla and possibly the nasal, frontal from nasal, and probably the squamosal). This system is responsible for the crazy-straw nasal passages and expanded snout of ankylosaurians, and probably the gigantic narial aperture in ceratopsians. The PNS is also present in mammals, and surrounds the orbit and forms a form of “crumple zone.”

2. Paratympanic Sinus System (PTS)

This system is shown in Blue.

Posterior to the PNS, and closer to the exit of the nasopharyngeal passage through the bony choanae (or “internal nares”), the paratympanic system surrounds the inner ear and from there invades the bones surrounding the braincase. This is typically where they stay, although branches of them form the interorbital air sacs, sit between muscles in the adductor chamber in birds, and invade the quadrate and, from there, into the articular of the mandible; from the articular, the PTS expands into the mandible. The PTS is also responsible for pneumatic hyoid bones, especially the marvelous thing (from Darren Naish’s Tetrapod Zoology blog) that exists in the howler monkeys (Alouatta).

3. Clavicular Air Sac (CLAS)

This system is shown in Orange.

The first branch of the CLAS is the clavicular air sac itself, which unique amongst the other branches of the systems arising from the lung itself is a median structure, although the structure develops from left and right distinct air sacs that fuse. This air sac lies behind the shoulder girdle and between it, and when it presses against the sternum can form a pocket within it. Pneumatic furculae are the result of this air sac. In cases where the accessory clavicular apparatus of the shoulder, as the interclavicle and clavicles, are fused to the sternum, this air sac invades them too (as only so far in pterosaurs). This structure pneumatizes the shoulder, and consequently also the humerus, forming the brachial air sac.

4. Cervical Air Sac (CAS)

This system is shown in Magenta.

On either side of the lungs, this pair of structures run between the shoulders and parallel to the vertebral column and between the central and the cervical ribs. The CAS is responsible for some fairly intricate pneumatic structures of the neck, including the awesome ribs of abelisaurid theropods, and of course the staggeringly complex structures of the neck in sauropods, from thin laminae and cavernous chambers to simple fossae but amazingly tiny and complex “spongy” bone. The pattern of pneumatic foramina this air sac creates in cervical vertebrae, and the divisions of fossae formed by sub-diverticulae, can be diagnostic among saurischian dinosaurs.

5. Dorsal Air Sacs (DAS)

This system is shown in Dark Yellow.

It might be problematic to call this a “system,” as the DAS actually derives from multiple different pockets formed independently from the lung and each invades the adjacent vertebra, and communicates through the spinal cord and the inter vertebral diverticulae, as evidences from median foramina and pneumatic structures in birds and sauropods. It might be best to characterize this system, the first noncranial system to invade the skeleton, as a reticulate diverticulum, or “pneumatic net.”

6. Anterior Thoracic Air Sac (ATAS)

This system is shown in Green.

This system arises from the ventral region of the anterior lung and pneumatizes the gastralia and ribs, as well as the abdominal muscles and is the primary pneumatic system involved in pelvic-assisted gastric breathing in crocs, birds, and probably dinosaurs.

7. Posterior Thoracic Air Sac (PTAS)

This system is shown in Dark Green.

This system arises caudal, as its name may suggest, to the ATAS, and is responsible for pneumatizing the internal organs of the animal; this structure leaves no osteological evidence for its presence, but is followed by the AAS, and consequently if there is skeletal evidence for the AAS and for the ATAS, it can be inferred that there is a PTAS.

8. Abdominal Air Sac (AAS)

This system is shown in Teal.

This system pneumatizes the sacrum, adjacent dorsal vertebrae that have become fused to it (forming a synsacrum), and the caudal vertebrae; it also pneumatizes the pelvis and the femur.

In birds, distal limb elements and subdermal structures are pneumatized by subcutaneous air sacs, but as these develop from their adjacent proximal limb-based air sac (branchial from the CLAS, femoral from the AAS) they thus can be grouped under those headings. These structures pneumatize their respective elements (meta-, epi- and autopodial elements; integument; feathers in birds) externally, forming small pockets in the bone within which a foramen penetrates the bone. Similar to pterosaur wing structures, distal limb pockets, such as in the astragalus and calcaneum of theropodan dinosaurs, arise from these subsidiary, external diverticulae.

This entry was posted in Art, Paleobiology, Paleontology, Reconstruction, Terminology and tagged , , , , , . Bookmark the permalink.

16 Responses to Pneumatic Air Sacs in Dinosaurs

  1. Andrea Cau says:

    The absence of osteological correlates of postcranial pneumatic systems in ornithischian bones may not mean they lacked postcranial air sacs at all, since they’re bracketed by taxa showing evidence of these features (pterosaurs and saurischians). It may just indicate ornithischians had these air sacs but the latter did not invade the bones.

    • If postcranial skeletal pneumaticity were ancestral for all clades, then yes, ornithischians should have it. But they don’t. This actually seems to suggest that not only do ornithischians lack PSP, but that basal saurischians may not have it, either, nor that basal theropods/sauropodomorphans had it, just as basal pterosaurs seem to lack it (no pneumatophores, no large fossae, etc.). Dinosaurs and other ornithodirans seem to have their heads straight when it comes to cranial pneumaticity, but that seems all they have. Ornithischians don’t even seem to have that much in the way of the PTS, just the PNS, and in that it’s mostly the NPP that has become elaborate, though obviously sending off diverticulae into the snout and around the naris. The question really is: why reduce the cranial pneuamticity from your ancestral condition — if indeed it was ancestral? — and I think there’s something in it about diet, but cannot test this at the moment (most mammals with a similar inferred diet seem to have extensive cranial pneumaticity, at the least, even the diving ones).

      • Andrea Cau says:

        Presence of air sacs is not presence of PSP. Note that: 1) in some extant bird lineages, air sacs do not extensively invade the postcranial skeleton; 2) in birds, skeletal pneumatisation expands along ontogeny, following a pattern that recalls both sauropod and theropod expansion among the two lineages. 1) indicates that absence of PSP does not always mean absence of air sacs, 2) means that at least the last common ancestor of saurischians had already evolved some form of air sacs, the latter expanding convergently along derived theropods and derived sauropods. Combining 1), 2) and the presence of air sacs in pterosaurs and saurischians, I think it is more parsimonious to consider air sacs as an ornithodiran synapomorphy, and the EXTENSION of the air sacs into the bones (=PSP) as a parallelism among Theropoda, Pterosauria and Sauropoda, with the three lineages simply expanding their sacs from the ancestral condition present in basal dinosauromorphs (Silesaurids show incipient presacral pneumatisation, partially filling the gab between saurischians and pterosaurs), basal theropods, some basal sauropodomorphs (but not all: Thecodontosaurus shows incipient cervical pneumatisation) and ornithischians. Thus, I think ornithischians did not reduce their PSP, just retained the ancestral condition of all ornithodirans: presence of sacs but absence of PSP.

        • It is difficult to determine the presence of a particular air sac without skeletal invasion by it. While one might make an inference that certain air sacs were present, without osteological correlates, these are at best a second level inference, but without certain phylogenetic information it is a third level inference. Ornithischians lack any and all sorts of PSP, so to assume they had it is to be reasonable, but highly speculative. We have some evidence for PSP in some stem crocs (Butler, Barrett and Gower, 2012) but it is a second level inference itself. Basal pterosaurs do not show evidence of PSP (Bonde & Christiansen, 2003), making it more difficult to that the conclusion of archosaurian PSP is effectively “foregone.” I’ll note that in some cases, postcranial air sacs that do NOT invade bones can be detected when there is evidence for soft-tissue preservation such as the post-brachial “wedge” in pterosaurs. Problematically, this suggests either 1) that postcranial air sacs are present, extensive, but do not invade bone, but approach it, or 2) that development of certain air sacs are unique to clades, but they can be convergently acquired.

          • Andrea Cau says:

            The arguments in Butler et al. (2012) cited by you agreed with my interpretation: postcranial air sacs as an ornithodiran synapomorphy, PSP derived from this basal condition and developed indipendently among three ornithodiran lineages. I find less parsimonious an indipendent origin of air sacs three times and then each ornithischian lineage developing extensive pneumatisation. My scenario: 4 steps (1 common origin of air sacs + 3 developments of PSP [in derived pterosaurs, in derived sauropodomorphs, in derived theropods]); alternative scenario: 6 steps (pterosaur origin of air sacs + 1 develpment of PSP + sauropod origin of air sacs + 1 development of PSP + theropod origin of air sacs + 1 development of PSP).

          • I also agree with Butler et al. on this point; but that wasn’t MY point. I am saying that when it comes to making statements of presence or absence, you can only make inferences, and just that, without explicit data regarding these structures: It is ambiguous, as they state, whether laminae or fossae are directly related ONLY to pneumatic diverticulae. Without better data, the presence/absence cannot merely be stated to be known.

          • Andrea Cau says:

            I meant ‘ornithodiran’ not ‘ornithischian’…

  2. Throat-warbler Mangrove says:

    Err… Just a suggestion here. You might want to run coloured diagrammes past someone with long-wave sensitive cone cells first. The CLAS and DAS are both showing up as green to me, it sort of threw me off when I made my first quick skim through the article and wasn’t paying attention to the numbers.

    • I have this debate with Andrea, too. I use the color wheel in PS to determine the colors because I have difficulty figuring it out at a glance. That’s why I put in numbers, and wrote the words out, trying to circumvent the obvious issues, like colorblindness or partial CB. I see the same hues across multiple values, and this makes colors tricky for me.

  3. For a devil’s advocate view of air sacs and what they might or might not mean for archosaur respiratory evolution, I recommend a thorough read through Colleen Farmer’s papers, most w/info here:!publications . If you read them closely (and the PeerJ paper that the SV-POW blog covers in the 1st link above), they call into question much of the recent “consensus” on even air sac morphology/distribution. I’m not sure what to think of those criticisms myself; I haven’t studied air sacs and skeletal anatomy super duper carefully firsthand in enough diversity of living animals to consider myself an expert. But Colleen has, and I think her criticisms need to be considered openly and head on. If she’s right on some major points, there could be a lot of building of houses of cards in the area of air sac evolution. How strong is the anatomical evidence and where are the weaknesses? I think everyone should be asking themselves that question if they are working in this area. Colleen and collaborator Emma Schachner have already pointed out some areas where the physiological/ventilatory importance of air sacs may have been incorrectly assumed (e.g. end of our PeerJ paper), and I suspect they might be right on that. But if pneumaticity is not even that reliable an indicator of air sac size/location/presence etc., as some of Colleen’s earlier papers suggested, then there is a bigger problem at hand.

    • Excellent point, John. This is acerbated by the fact that we only have birds to go one for what air sacs pneumatize what, though these are consistent amongst birds. I found myself concerned over this, but figured a primar on what were generally dinosaurs with some factors such as pneumatic shoulder elements would allow me to speculate broadly on dinosaurs being ancestral to the avian condition and all that entails. But yes, it is wise not to paint too broadly a stroke with too narrow a brush as this.

  4. Pingback: A Look Back at the Bite Stuff, 2013 Edition | The Bite Stuff

  5. Wow!… your articles are more than interesting! I especially enjoyed the one about lips too! :)

    I collect and buy fossils as a hobby and took some photos of my split and polished Allosaurus vertebra (that shows pneumatic diverticulae evidence). See it in one of my old blogs at:

    I’ll continue to read and enjoy your blogs. Additionally, i followed you! Keep up the amazing work! :)

    • I am glad you found this helpful! Future work will increase knowledge of pneumatic systems in various archosaurs, and hopefully help us understand the hows and whys of dinosaurs with bones made of air.

  6. ridwan says:

    i wonder why would you expect antorbital fenestra houses an air sac? so i guess the external side is covered by muscles and skin. What’s the function of the air sac in antorbital fenestra? do you think ornithomimus has this too in its antorbital fenestra? reply to

    • We expect this to house an antorbital air sac because of the presence of direct evidence that this fenestra is associated with pneumatic tissues, including surface textures, presence of internal diverticulae that branch off from it and evidence from birds and some crocs, which retain these fenestrae. We thus have a good strong case that in extinct dinosaurs, the large open antorbital fenestra is associated with the antorbital sinus system. Larry Witmer has an excellent run down of this in his 1990 and 1995 papers, available here and made open access (scroll down to near the bottom).

      In addition, the presence of the lateral antorbital fossa implies this airsac expanded onto the outside of the skull bones

Leave a Reply

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

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

Facebook photo

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

Connecting to %s