A six-year quest is finally complete! Almost all known sauropod necks are incomplete and distorted

January 24, 2022

Today finally sees the publication of a paper (Taylor 2022) that’s been longer in gestation than most (although, yes, all right, not as long as the Archbishop). I guess the first seeds were sown almost a full decade ago when I posted How long was the neck of Diplodocus? in May 2011, but it was submitted as a preprint in 2015. Since then it’s taken far longer than it should have done to get it across the line, and it is primarily with a feeling of relief that I see the paper now published.

Taylor (2022: figure 4). W. H. Reed’s diagram of Quarry C near Camp Carnegie on Sheep Creek, in Albany County, Wyoming. The coloured bones belong to CM 84, the holotype of Diplodocus carnegii; other bones belong to other individuals, chiefly of Brontosaurus, Camarasaurus and Stegosaurus. Modified (cropped and coloured) from Hatcher (1901: plate I). Cervical vertebrae are purple (and greatly simplified in outline by Reed), dorsals are red, the sacrum is orange, caudals are yellow, limb girdle elements are blue, and limb bones are green.

In this quarry map for the Carnegie Diplodocus, does it seem to you that the vertebrae of the neck (in purple) are drawn unconvincingly, compared with the fairly detailed drawings of the dorsals? Does that suggest that maybe Reed — who drew this diagram years after the excavation was complete — didn’t really remember how the neck was laid out? How well does the textual description of the skeleton in situ match this map? These are the kinds of questions I was asking myself as I started thinking about what has become the paper published today.

In some ways it’s a really simple paper, pretty much summarised by its title: almost all known sauropod necks are incomplete and distorted. It started out as a formalised version of three posts on this blog (How long was the neck of Diplodocus?, Measuring the elongation of vertebrae and The Field Museum’s photo-archives tumblr, featuring: airbrushing dorsals), but somewhere along the line the tale grew in the telling and it’s ended up as 35 pages of goodness. In the process of review it acquired a lot of new material, including: a discussion of how to locate the cevicodorsal junction (summary: it’s complicated); a couple of ways to numerically quantify the degree of distortion along a neck; and a brief discussion of retrodeformation (summary: it’s complicated).

Head and neck from Janensch’s (1950b: plate VI) skeletal reconstruction of Giraffatitan brancai (= “Brachiosaurusbrancai of his usage) mounted specimen based on MB.R.2181 (formerly HMN SII). The parts of the head and neck that were lost to damage are greyed out, including the first two cervicals and the neural arches and spines of all cervicals after C8. Oh, and the head.

I hope this paper will be of use, especially to people coming into the field with the same unrealistic assumptions I had back in the early 2000s. Back then, I had in mind a project to determine the thickness of intervertebral cartilage in the neck of Diplodocus by measuring the radii of curvature of the condyles and cotyles of successive vertebrae — an idea that distortion makes unrealistic. I took the DinoMorph work at face value — something that seems incredible to me knowing what know now. The paper that came out today is basically the one I wish I’d been able to read in 2000 (but updated!)

By the way, when I was fine-tooth-combing the proof PDF a few days ago, I was delighted to be reminded that I got the phrase “rigidly defined areas of doubt and uncertainty” into the paper — a reference of course, to the words of the philosopher Vroomfondel in The Hitch-Hiker’s Guide to the Galaxy. I’ll file this alongside the Monty Python reference in my history-of-sauropod-research book chapter and the Star Wars paraphrase that opens a computer-science paper I lead-authored in 2005.

References

26 Responses to “A six-year quest is finally complete! Almost all known sauropod necks are incomplete and distorted”

  1. llewelly Says:

    congratulations on your remarkably complete documentation of our incomplete knowledge about sauropod necks

    The state of neck knowledge is a lot worse than I thought it was.

  2. LeeB Says:

    Very nice paper.
    I suspect the tails of sauropods are also equally incomplete; especially at the distal ends.

  3. llewelly Says:

    With respect to tails, I have the impression – and I am not at all sure this is correct – that caudals tend to be less pneumatic than cervicals.

    This would imply individual caudal verts would generally be better preserved than cervical verts of comparable size. And that might imply the articulated caudal sequences with the highest vert counts would have higher vert counts than the articulated cervical sequences with the highest vert counts.

    But I think sauropod tails are generally estimated to have more vertebrae than their necks? So even if there’s somewhat better preservation, it’s probably not enough to result in a higher proportion of complete tails – especially not when the often very small verts near the tip of the tail are considered.

    It does make me more skeptical of the estimates of sauropod tail vertebrae counts in general, and especially of the estimates of 75 or 80 sometimes given for Diplodocus.

    With all this uncertainity about both necks and tails, I am also more skeptical of length estimates for sauropods.

  4. Mike Taylor Says:

    Yes, llewelly, it certainly is the case that tails are significantly less pneumatic than necks — and individual caudal vertebrae are smaller and simpler than individual cervicals. For both reasons, they will be less prone to distortion.

    I’m not at all sure that it follows that “articulated caudal sequences with the highest vert counts would have higher vert counts than the articulated cervical sequences with the highest vert counts”, though. That is a matter of taphonomic disarticulation rather than distortion. A neck can be complete but badly distorted/damaged — as for example in the case of the Mamenchisaurus hochuanensis holotype. Tails are less likely have damaged vertebrae but if anything are more likely to be incomplete, as the small vertebrae at the tip are easily lost.

  5. Tom2015 Says:

    What about Xinjiangtitan? :-) Anyway, great work, congratulations!

  6. Mike Taylor Says:

    @Tom2015, Xinjiangtitan is included in the Catalogue of complete necks section of the paper.

  7. llewelly Says:

    Scavenging is only mentioned briefly in your paper, but I’m wondering if it might play a larger role; are sauropod cervicals pneumatic enough to be scavenged by animals that aren’t usually thought to scavenge bone?

    (I thought of this because I’ve read that porcupines, deer, and some moths scavenge deer antler. Probably not a particularly good comparison, but it did make me wonder if there might some surprising scavengers of sauropod neck vertebrae.)

  8. Mike Taylor Says:

    That’s a really interesting thought. I’ve heard this idea about generally herbivorous animals scavenging discarded antlers, and never really given it the thought I ought to have. Do they also scavenge regular bones? If not, why not? What’s the difference?

  9. llewelly Says:

    what I seem to recall – and I’m definitely not an expert on the topic – is that it’s specifically deer antler that deer are mentioned as scavenging; I don’t recall seeing mention of deer scavenging other bone.

    As for porcupines, I also don’t recall mention of porcupines scavenging bone other than deer antler. Also, I should specify, these are North American porcupines – not the Eurasian or African porcupines, from which they are separated by I think about 40+ million years (and which evolved their quills independently)

    As for the difference between deer antler and normal bone – I don’t know, but having handled deer antler, they always seem quite a bit lighter than other bones, and antler are mostly cancelous bone, I think. I am sorry that’s not much to go on – you need a real expert, rather than me.

  10. Marja Erwin Says:

    I wonder if that’s just because carnivores take the carcasses…?

  11. Marco Says:

    Congratulation for your pubblication!
    Can I ask about the neck of Rapetosaurus?
    The juvenile seems pretty complete and the hi-fi skeletal recostruction (both Hartman and Paul) shown a higher number of cervical vertebrae than the 14/15 usually assumed for Titanosaurs (based on Futalognkosaurus I Guess).
    Thank you!

  12. Mike Taylor Says:

    Thnks, Marco.

    Yes, it seems from various reconstructions that Rapetosaurus likely had 16 or 17 cervicals. But to my knowledge there is no single well-preserved neck. Curry Rogers (2009:figure 1) is a skeletal reconstruction of FMNH PR 2209, which I assume is the best individual, and it is shown as missing C3, C6–9 (all but some cervical ribs), and the neural arches and spines of C14–16 are all missing.

  13. kris michael Says:

    Porcupines are known to collect and stockpile bones. Other animals that particularly enjoy bones include cattle, tortoises, bears, giraffes, birds, etc etc. Cattle with a phosphorous deficiency are known to contract botulism from eating their decayed relatives. Wouldn’t be surprised if sauropods behaved in a similar fashion.

  14. Mike Taylor Says:

    I’m not so sure about that, kris: cows are excellent chewers, and well equipped to crunch bone. Sauropods would not have been able to do nearly as well. Unless they could just swallow some small bones whole and digest them slowly, I guess.

  15. Marco Says:

    Thank you!
    Maybe this picture Made me think the neck was more complete than how it is in reality https://www.sciencebuzz.org/sites/all/files_static/ootm/archive/images/2001-09_2.jpg

  16. Matt Wedel Says:

    SO much interesting stuff in this comment section already!

    I suspect that we have so many reports of antler scavenging for a few reasons:
    1. A deer can drop many sets of antlers over its life, but only one skeleton, and although the skeleton might contain as much or more bone than the cumulative load of antlers, the skeleton can only get dropped once, whereas the antlers can be scattered all over the place, leading to a higher “hit rate” for both scavengers and people encountering obviously-scavenged bone.
    2. When the antlers are dropped they are clear of soft tissue, so they won’t attract scavengers looking for meat, or get putrid, so they may represent a better source of minerals than a skeleton trapped inside a groady, rotting carcass — even though antler has a lower mineral content than most other types of bone (see next comment).
    3. Other than the shafts of the long bones, most skeletal elements have thin cortices, and if those are damaged by gnawing, the whole element is at greater risk of falling apart from weathering or further scavenging. In contrast, antlers have fairly thick cortices, so they are more likely to survive the gnawing process, and there are a lot of antlers around to get gnawed and then get found (point 1).
    4. On the “getting found” part: antlers have complex, distinctive, eye-catching shapes, they’re generally clean and dry, and they’re decorative, so I suspect that people out in the wilderness spot and pick up antlers more often than other bones, further increasing the discovery rate compared to other elements.

    For all these reasons, my guess is that antler scavenging is so well-reported both because it happens a lot (natural supply side), and because the gnawed antlers are so likely to be found and the gnaw marks identified (human sampling side), even if animals aren’t turning their noses up at other bones. As archaeologists will attest, rodents gnaw pretty much any bones they can get their teeth on, including humans.

  17. Mike Taylor Says:

    I love that this comment stream is 100% antler scavenging, 0% sauropod necks :-)

  18. Matt Wedel Says:

    llewelly wrote:

    As for the difference between deer antler and normal bone – I don’t know, but having handled deer antler, they always seem quite a bit lighter than other bones, and antler are mostly cancelous bone, I think.

    Interesting observation, and it lines up with my experience. Which is weird, because my perception was that antlers look quite dense in cross-section, and a quick image search for ‘antler cross-section’ confirmed that. So I did a little digging. Paramio et al. (2012) got a mean whole-antler density of 1.1 g/mL by both 3D modeling and the water displacement method. In contrast, Chen et al. (2009: table 1) report most samples between 1.7 and 2.0 g/mL, with a couple of outliers at 0.83 and 0.91 g/mL.

    What’s going on here — why the wildly discordant measurements? I strongly suspect that it has to do with whether the volume measured was that of the whole element, or only the bony tissue. Non-pneumatic bones are filled with marrow in life, which contributes to their weight, but when the bones are cleaned and dried the marrow is greatly reduced, often to near zero. The marrow spaces in dry bones are filled with air, which has the effect of artificially pneumatizing them. Compact bone has a density between 1.7 and 2.0 g/mL, so it would make sense if the high numbers reported by Chen et al. were for the density of the bone material only, and the low numbers in Chen et al. and Paramio et al. were for the whole antler, including the air spaces that replaced the marrow. Paramio et al. did dunk their antlers, but they mentioned that “floatability” was a problem, and indeed it would be difficult for water to displace all the air in an antler during a quick dunk.

    But even the bony tissue of antlers does seem to be oddly un-dense. Chen et al. (2009: table 2) found that even the compact bone in their elk antler samples only had a density of 1.7, and that “antler was found to have the lowest mineral content and consequently the lowest elastic modulus” in a broad sample of amniote bones (p. 695).

    So antler does seem to be genuinely un-dense. Even if antlers have a lower mineral content than other bones, that lower density may make antlers an attractive scavenging target — especially combined with the plethora of antlers lying around, conveniently separated from the bodies that built them.

    References

    Chen, P.Y., Stokes, A.G. and McKittrick, J., 2009. Comparison of the structure and mechanical properties of bovine femur bone and antler of the North American elk (Cervus elaphus canadensis). Acta Biomaterialia, 5(2), pp.693-706.

    Paramio, M.A.R., Muñoz-Cobo, J., Moro, J., Gutierrez, R., Oya, A., Tellado, S. and Azorit, C., 2012. Assessing red deer antler density with a hydrostatic method versus a new parametric volume-modelling technique using 3D-CAD. Animal Production Science, 52(8), pp.750-755.

  19. Mike Taylor Says:

    “So antler does seem to be genuinely un-dense”

    Yes, it’s always bothered be that “dense” seems to have no satisfactory antonym. Perhaps “sparse” is the least bad option.

  20. Matt Wedel Says:

    Mike wrote:

    I’m not so sure about that, kris: cows are excellent chewers, and well equipped to crunch bone. Sauropods would not have been able to do nearly as well. Unless they could just swallow some small bones whole and digest them slowly, I guess.

    …Except that sauropods were extremely well-equipped to render all but the largest bones into bite-size chunks by the simple expedient of stomping on them. I’ll bet it happened all the time: big, fast-growing animals with massive calcium demands, stuck with tiny heads and wimpy jaws on one hand, and multi-ton stomp force on the other.

  21. Mike Taylor Says:

    Maaybe. But it’s not like sauropods wore boots. I would expect the bottoms of their feet be soft and pliable, like elephants’ feet.

  22. Matt Wedel Says:

    llewelly wrote:

    But I think sauropod tails are generally estimated to have more vertebrae than their necks? So even if there’s somewhat better preservation, it’s probably not enough to result in a higher proportion of complete tails – especially not when the often very small verts near the tip of the tail are considered.

    This point is discussed at length in Hone (2012) and revisited in Hone et al. (2021). The upshot is that you are correct: we have *very* few complete dinosaur tails.

    It does make me more skeptical of the estimates of sauropod tail vertebrae counts in general, and especially of the estimates of 75 or 80 sometimes given for Diplodocus.

    Oh, some diplodocids definitely had more than 80 caudal vertebrae. (Or are you possibly implying that Diplodocus had even more caudals than 75-80? I can’t tell if you are arguing that the count is an underestimate or an overestimate.) Here’s Gilmore (1936: p. 204) on the caudal count in Apatosaurus:

    In completing the tail of the mounted skeleton [of CM 3018, the holotype of A. louisae], nine additional artificial elements were added to the tip making a total of 73 vertebrae. That the complete series is in excess of that number is now clearly shown by specimen No. 3378 C.M., see Pl. XXVIII, of which the complete series formed a part of an articulated vertebral column that extended from the atlas to the very tip of the tail. There are 82 caudal vertebrae in this series, as now numbered, although an elongated element near the tip may represent a coalesced pair in which event the total would be 83.

    (Incidentally, Mike, CM 3378 would seem to be relevant to your new paper! Have we ever seen or even talked about that specimen? I hadn’t grokked its existence until writing this comment sent me back to Gilmore.)

    Turning to Diplodocus specifically, Holland (1906: pp. 253-254) made a pretty good case for there being more than 73 vertebrae in the tail. He also went on to point out that some extant lizards have up to 100 caudal vertebrae.

    Heck, even Gilmore’s (1925) juvenile Camarasaurus has a complete tail (missing two centra, but with those neural arches present and articulated) with 53 vertebrae, and Cam doesn’t have a particularly long tail or a whiplash.

    References

    Gilmore, Charles W. 1925. A nearly complete articulated skeleton of Camarasaurus, a saurischian dinosaur from the Dinosaur National Monument, Utah. Memoirs of the Carnegie Museum 10:347-384.

    Gilmore CW. 1936. Osteology of Apatosaurus, with special reference to specimens in the Carnegie Museum. Memoirs of the Carnegie Museum 11:175-271.

    Holland, W.J. 1906. The osteology of Diplodocus Marsh. Memoirs of the Carnegie Museum. 2: 225–264.

    Hone DWE. 2012. Variation in the tail length of non-avian dinosaurs. Journal of Vertebrate Paleontology 32:1082-1089.

    Hone DWE, Persons WS, Le Comber SC. 2021. New data on tail lengths and variation along the caudal series in the non-avialan dinosaurs. PeerJ 9:e10721 https://doi.org/10.7717/peerj.10721

  23. Matt Wedel Says:

    Maaybe. But it’s not like sauropods wore boots. I would expect the bottoms of their feet be soft and pliable, like elephants’ feet.

    Something can be pliable and still apply enough pressure to break a bone. Maybe sauropods did it by stomping, maybe they did it by just letting a few tons of weight down with steady pressure through a comparatively small foot. Either way, if you’re trying to argue that sauropods were incapable of breaking bones by stepping on them, I wish you luck.

  24. llewelly Says:

    Matt, thank you so much for your detailed replies. I didn’t know whether to think 75-80 caudals for a diplodocid too low or too high – just, not very confident either way.

    I didn’t realize Gilmore’s baby Cam had 53 caudals – I guess that reveals I haven’t read enought about it.

  25. Matt Wedel Says:

    Happy to help! I had some of the sauropod caudal stuff loaded in RAM from talking with Dave Hone about his dinosaur tail projects over the past several years. I had forgotten that the baby Cam had so many caudals. Just eyeballing it I might have guessed like 40. I guess those little ones at the end add up fast.


  26. […] Last time, we looked briefly at my new paper Almost all known sauropod necks are incomplete and distorted (Taylor 2022). As hinted at in that post, this paper had a difficult and protracted genesis. I thought it might be interesting to watch the story of a published paper through its various stages of prehistory and history. Strap in, this is a long one — but hopefully of interest, especially to people who are just coming into academia and wonder how this stuff works in practice. […]


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: