So what were apatosaurs doing with their crazy necks?

September 14, 2015

We’ve noted that the Taylor et al. SVPCA abstract and talk slides are up now up as part of the SVPCA 2015 PeerJ Collection, so anyone who’s interested has probably taken a look already to see what it was about. (As an aside, I am delighted to see that two more abstracts have been added to the collection since I wrote about it.)

It was my privilege to present a talk on our hypothesis that the distinctive and bizarre toblerone-shaped necks of apatosaurs were an adaptation for intraspecific combat. This talk was based on an in-progress manuscript that Matt is lead-authoring. Also on board is the third SV-POW!sketeer, the silent partner, Darren Naish; and artist/ethologist Brian Engh.

Here is our case, briefly summarised from five key slides. First, let’s take a look at what is distinctive in the morphology of apatosaur cervicals:

Screen Shot 2015-09-12 at 11.22.26

Here I’m using Brontosaurus, which is among the more extreme apatosaurs, but the same features are seen developed to nearly the same extent in Apatosaurus louisae, the best-known apatosaur, and to some extent in all apatosaurs.

Now we’ll look at the four key features separately.

Screen Shot 2015-09-12 at 11.22.57

First, the cervicals ribs of sauropods (and other saurischians, including birds) anchored the longus colli ventralis and flexor colli lateralis muscles — ventral muscles whose job is to pull the neck downwards. By shifting the attachments points of these muscles downwards, apatosaurs enabled them to work with improved mechanical advantage — that is, to bring more force to bear.

Screen Shot 2015-09-12 at 11.23.06

Second, by redirecting the diapophyses and parapophyses ventrally, and making them much more robust than in other sauropods, apatosaurs structured their neck skeletons to better resist ventral impacts.

Screen Shot 2015-09-12 at 11.23.15

Third, because the low-hanging cervical ribs created an inverted “V” shape below the centrum, they formed a protective cradle for the vulnerable soft-tissue that is otherwise exposed on the ventral aspect of the neck: trachea, oesophagus, major blood vessels. In apatosaurus, all of these would have been safely wrapped in layers of connective tissue and bubble-wrap-like pneumatic diverticula. The presence of diverticula ventral to the vertebral centrum is not speculative – most neosauropods have fossae on the ventral surfaces of their cervical centra, and apatosaurines tend to have foramina that connect to internal chambers as well (see Lovelace et al. 2007: fig. 4, which is reproduced in this post).

Screen Shot 2015-09-12 at 11.23.22

Fourth, most if not all apatosaurs have distinctive ventrally directed club-like processes on the front of their cervical ribs. (It’s hard to tell with Apatosaurus ajax, because the best cervical vertebra of that species is so very reconstructed.) How did these appear in life? It’s difficult to be sure. They might have appeared as a low boss; or, as with rhinoceros horns, they might even have carried keratinous spikes.

Putting it all together, we have an animal whose neck can be brought downwards with great force; whose neck was mechanically capable of resisting impacts on its ventral aspect; whose vulnerable ventral-side soft-tissue was well protected; and which probably had prominent clubs or spikes all along the ventral aspect of the neck. And all of this was accomplished at the cost of making the neck a lot heavier than it would have been otherwise. Off the cuff, it seems likely that the cervical series alone would have massed twice as much in apatosaurines as in diplodocines of the same neck length.

Doubling the mass of the neck is a very peculiar thing for a sauropod lineage to do – by the Late Jurassic, sauropods were the leading edge of an evolutionary trend to lengthen and lighten the neck that had been running for almost 100 million years, through basal ornithodirans, basal dinosauromorphs, basal saurischians, basal sauropodomorphs, and basal sauropods. Whatever the selective pressures that led apatosaurines to evolve such robust and heavy necks, they must have been compelling.

The possibility that apatosaurs were pushing or crashing their necks ventrally in some form of combat accounts for all of the weird morphology documented above, and we know that sexual selection is powerful force that underlies a lot of bizarre structures in extant animals, and probably in extinct ornithodirans as well (see Hone et al. 2012, Hone and Naish 2013).

What form of combat, exactly? There are various possibilities, which we’ll discuss another time. But I’ll leave you with Brian Engh’s beautiful illustration of one possible form of combat: a powerful impact of one neck brought down onto the dorsal aspect of another.

ApatoNeckSmashRoughWeb

We’re aware that this proposal is necessarily somewhat speculative. But we’re just not able to see any other explanation for the distinctive apatosaur neck. Even if we’re wrong about the ventrolateral processes on the cervical ribs supporting bosses or spikes, the first three points remain true, and given how they fly in the face of sauropods’ long history of making their necks lighter, they fairly cry out for explanation. If anyone has other proposals, we’ll be happy to hear them.

References

  • Hone, D. W., Naish, D., & Cuthill, I. C. (2012). Does mutual sexual selection explain the evolution of head crests in pterosaurs and dinosaurs?. Lethaia 45(2):139-156.
  • Hone, D. W. E., & Naish, D. (2013). The ‘species recognition hypothesis’ does not explain the presence and evolution of exaggerated structures in non‐avialan dinosaurs. Journal of Zoology 290(3):172-180.
  • Lovelace, D. M., Hartman, S. A., & Wahl, W. R. (2007). Morphology of a specimen of Supersaurus (Dinosauria, Sauropoda) from the Morrison Formation of Wyoming, and a re-evaluation of diplodocid phylogeny. Arquivos do Museu Nacional, Rio de Janeiro 65(4):527-544.
Advertisements

38 Responses to “So what were apatosaurs doing with their crazy necks?”


  1. The hypothesis about the possible spike-like process in the ventral region of the neck is really interesting! But are there any traits on the bone surface of the club-like processes to infer that?

  2. Mike Taylor Says:

    An excellent question. I defer to Matt, who has looked at an order of magnitude more apatosaur cervical ribs than I have.


  3. And, in such a similar scenario, it would be nice to know what was the role of the diverticula. I mean, in hitting behaviour, the risk to squeeze something that stays volume-costant, is.. well, a risk. Or, on the other hand, the diverticula might had been “pillows” to absorb impacts…

  4. Mike Taylor Says:

    I’d like to know what, if anything, is known of the mechanical role of diverticula in cushioning blows in birds. Again. Matt or Darren will know more than I do on this. But I don’t imagine the ventral trough of apatosaur necks filled with gigantic, fragile, inflatable balloon-like diverticula! It seems much more likely that they had many small diverticula, like those that Matt found in rhea necks, all intermingled with tough connective tissue.

    (One of the benefits of doing dissections is that you realise how much more stuff there is in there than the nice, clean muscles and organs that textbooks illustrate. I always find myself running into a whole mess of super-tough fascia which is really hard to dissect out, even with a good, sharp scalpel.)


  5. Pfft, everybody wants to be an ankylosaur these days. ;)

    But more seriously: I often use giraffe neck battles as the closest extant analogue I can think of to illustrate ankylosaur tail clubbing – giraffes are just clubbing each other with their heads rather than their tails. This is interesting stuff, and in particular the idea that there may have been epidermal structures like bosses or spikes laying atop the cervical rib knobs is intriguing! Looking forward to seeing where you go next with this!


  6. Since part of your argument relies on identification of the tumescent features of the cervical ribs as “bosses” have you considered looking into the microstructure of these features in any detail to confirm the argument in light of work from Hieronymus and Witmer on how these structures support rhino-like bosses (or “spikes”) in mammals, birds and other sauropsidan, etc? This would seem to be an interesting line of investigation in order to seek disconfirmation of your hypothesis.

    Secondly, after making claims about the function of loading and attachment into the cervical ribs, can you dismiss any other soft tissue attachment site for the bosses before then proceeding to agonistic structures, a la giraffe head bosses? Looking into loading regimes and strain dissipation in the cervical rib under various strain positions would likely reveal where, as in ceratopsian frills, regions of least strain would appear. If the bosses occur in areas of least strain, they are likely not bearing tendon or ligament loads, but probably being used agnostically. If, however, they are in areas of strain – when considering the effects of muscles around them – could that disconfirm the notion of agonistic knobs?

    I may be overly concerned on this point, but these seem practical, and important tests of the hypothesis.

  7. Kevin McDunn Says:

    What? Nonsense. Since their necks are shaped vaguely like upside-down alligator tails, they obviously used them to swim through the water. Backwards.

  8. ncmncm Says:

    Are any birds known to fight / have fought by neck clubbing? Some extinct bird species were big enough to pack a wallop. I expect if there were, you would have mentioned it. I guess there aren’t many apatosaurine verts to look at, but do any show evidence of (maybe healed) cracks? (I imagine breaks would be fatal.)

    It all makes me wonder about tree-climbing behavior, most probably in juveniles. But trees had no proper branches to hook a neck over, at the time, did they?

    Clobbering seems less plausible than some equivalent to arm-wrestling, but then who would guessed about giraffes? Do giraffe verts show evidence of trauma?


  9. Very cool. I don’t necessarily disagree with the fighting interpretation (in fact I quite like it) but for the sake of argument I can come up with at least one alternate hypothesis. Although I haven’t yet made the sauropod muscle cross-section post we discussed previously, the short version is that I disagree with Schwartz et al 2007 and agree with Tsuihiji, 2004 and others that there was substantial connective tissue and muscle above the top of the neural spines in diplodocid. As my skeletal silhouettes indicate, I think this reaches its zenith in apatosaurines (this will have to go in a future post…or paper). So the modifications to the ventral portion of the neck could just be a response to having to overcome greater amounts of elastic and contractile tissue dorsally.

    Of course these hypotheses aren’t mutually exclusive – apatosaurines like Brontosaurus that fought with their necks might have wanted to be able to move their necks with force upwards as well as down. And of course there are competing interpretations of dorsal neck soft-tissues in these critters. But at least it’s a plausible alternative and/or complementary hypothesis.

  10. Matt Wedel Says:

    Thanks, everyone, for your interest and your comments. I’ll tackle ’em one by one.

    Filippo and Mike raised the question of whether diverticula could absorb impacts. I don’t know if it’s ever been directly tested. You’d be amazed at how little we know about bird diverticula. But there is some strong circumstantial evidence. Pelicans make their living diving into the water after fish, and they have a jacket of subcutaneous diverticula that apparently looks like bubble wrap (and having seen the network of diverticula in a rhea neck, I’m inclined to believe it). The subcutaneous diverticula of pelicans may be more important for insulation than for shock absorption, but the fact remains that they are subject to some level of impact when pelicans hit the water (admittedly, probably much less than two apatosaurs smacking necks).

    Another relevant point: when Kent Sanders and I were messing around with bird necks in the late 90s, he found that if you bent a bird’s neck into different poses, the air in one system of diverticula would flow from the compressed side to the extended side. So in the case of a short, sharp shock the diverticula could potentially function as airbags, which do deflate but slowly enough to cushion the impact. Maybe.

    Anyway, this brings up a distinction that we need to make clear in the paper: we’re not arguing that diverticula were the primary cushion between the necks as they impacted. Rather, we know that the trachea, esophagus, and big blood vessels (carotid arteries and jugular veins) ran in the ventral ‘trough’ formed by the cervical ribs, which also held diverticula, and those diverticula could have protected the cervical viscera from damage.

  11. Matt Wedel Says:

    Next point: whether the cervical ribs functioned as clubs, and if so, whether they supported any soft tissues like keratinous bosses or spikes. Honestly, this is the most speculative and least important part of the whole story. As I described in this post, we don’t even know if the cervical ribs on apatosaurs would have projected outside of the muscular envelope – although I think they may have, given where and how the muscles were probably running – see Mike’s mechanical advantage slide (point #1) in the post above. Even if they did, the club hypothesis is really a side dish here. To me, the first three points are more compelling and more interesting – why did apatosaurines reverse such a long and apparently powerful evolutionary trend to make their necks fat and heavy? There are a lot of specific features of their cervical vertebrae that are rather neatly tied together if they were pushing against each other with their necks, and as far as I know no-one has proposed an alternative hypothesis that accounts for all of the morphology as neatly (Scott, I haven’t neglected your suggestion – I’ll discuss it further in another comment).

    As far as texture on the ‘clubs’ that would indicate attached soft tissue, I think it’s a wash. The OMNH specimen is too sanded, and I haven’t seen the relevant CM or YPM specimens up close enough, BUT given how much subtle restoration and lacquer those specimens include, I’m not optimistic. Histology to test the club hypothesis is a non-starter – we have zero apatosaur necks with good cervical ribs that aren’t parts of holotypes, and the case for cutting them up is not very compelling.

    I’m not sure that histology would be a good discriminator in this case anyway. The alternative hypothesis for the function of the knobs, which Kent and I proposed back in 2002, is that they anchored neck muscles. Whether they anchored neck muscles or some gnarly soft tissue for combat, the internal histology is probably the same – lots of collaginous Sharpey’s fibers running through the bone. The only surprising result would be a dearth of Sharpey’s fibers, which would pretty well falsify the muscle attachment hypothesis but not the club hypothesis, since the clubs don’t require attached soft tissue to function. The cervical ribs could be fully incorporated into the muscular envelope and still be useful pillar for exerting force ventrally, whether the skin over them was modified or not.

  12. Matt Wedel Says:

    Jaime raised some solid biomechanical points, especially this bit:

    If the bosses occur in areas of least strain, they are likely not bearing tendon or ligament loads, but probably being used agnostically. If, however, they are in areas of strain – when considering the effects of muscles around them – could that disconfirm the notion of agonistic knobs?

    Yeah, could do. The problem is that I wouldn’t even know where to start with that. Biomechanical studies on vertebrae are horribly complicated because vertebrae are horribly complicated – they don’t just have a few muscles and ligaments attached, they’re essentially embedded in a matrix of muscles and ligaments, not all of which leave diagnostic traces on the bones. As measure of how difficult this stuff is, it’s taken about three decades of global cooperation to get to the point where we have a reasonable description of what goes on with respiratory rib movements in dogs and humans. And these are the two most commonly dissected extant mammals!

    The problem with trying this sort of analysis on sauropods is that we just don’t know enough. I can’t remember if it was in a paper or a blog post, but Mike once made the point that we don’t even know how much soft tissue to put around sauropod necks – which would influence the mass of each segment, which would tell you something about the forces the muscles had to bear, which would influence the kind of strain analysis you’re talking about.

    So it’s not that I think the study you propose wouldn’t be awesome. I really would. But given how little we know right now about some of the basic anatomy – like whether the intervertebral joints were cartilaginous or synovial – I think we’d just be piling assumptions on assumptions, with no real way to know whether our model approximated reality.

    And even if we somehow succeeded (and had some other line of evidence to check so we’d know that we’d succeeded), that would only test the club bit, which is really a sidekick hypothesis here. The ventrolateral processes only became interesting in retrospect, after we realized how well the rest of the cervical morphology seemed to be adapted to exerting force ventrally. I’m not that wound up about whether they functioned as clubs or not – if not, the rest of the package still coheres and requires an explanation.

  13. Matt Wedel Says:

    Nathan asked three interesting questions:

    Are any birds known to fight / have fought by neck clubbing? Some extinct bird species were big enough to pack a wallop.

    None that I know of, although of course we should double-check with Darren, living repository of tetrapod weirdness. Among extant birds, the big ones tend to either beat you with their wings, or kick the crap out of you, or both. Apatosaurs may have been driven to neck-pushing from a lack of any other appendage to push forward with. (Side thought, which just occurred to me: I wonder if tail clubs in shunosaurs et al. might have evolved at least partly for intraspecific combat. Hard to imagine that at least some sauropods didn’t beat each other with their tails at least some of the time.)

    I guess there aren’t many apatosaurine verts to look at, but do any show evidence of (maybe healed) cracks? (I imagine breaks would be fatal.)

    Nope. I’ve looked, never seen any. But I’m not surprised, for reasons I’ll explain momentarily.

    Do giraffe verts show evidence of trauma?

    I know that there are giraffes out there with neck injuries from combat but they are not common. I’ve looked at lots of giraffe skeletons and never seen any. In the whole world we probably have something like half a dozen reasonably complete apatosaur necks (Yale, AMNH, Carnegie, Chicago, U Wyoming, Tokyo…where else?). The chances that one of them would have a broken and healed injury from neck fighting, even if apatosaurs engaged in neck fight all the time, are vanishingly remote.

    Here’s something to think about: for a lot of critters, it’s easier to disarticulate adjacent vertebrae, at least partly, than it is to break them. So even if apatosaurs fought with their necks a lot, injured animals may have been more likely to die of spinal injuries than to suffer fractures that would be bad enough to show up on the skeleton (i.e., not green-stick or stress fractures that might get completely remodeled away in time), but not bad enough to kill them.

  14. Matt Wedel Says:

    Finally, Scott wrote:

    the short version is that I disagree with Schwartz et al 2007 and agree with Tsuihiji, 2004 and others that there was substantial connective tissue and muscle above the top of the neural spines in diplodocid. As my skeletal silhouettes indicate, I think this reaches its zenith in apatosaurines (this will have to go in a future post…or paper). So the modifications to the ventral portion of the neck could just be a response to having to overcome greater amounts of elastic and contractile tissue dorsally.

    Interesting. Are you basing that on the width of the neural spine clefts, or the ligament scars at the bases of the clefts, or the morphology of the neural spines themselves, or something else? I’m not disagreeing with you, just curious about what evidence you find compelling.

    It’s an interesting idea and I have a few counterpoints to raise. First, as you allude to, having a lot of tension members dorsally isn’t mutually exclusive with the combat hypothesis. And in a way, saying that all the ventral gloop was only there to balance the dorsal gloop is still unsatisfying, because it doesn’t explain what apatosaurines were doing that required all that extra gloop in the first place. In short, it explains one package of weird morphology with another package of weird morphology, without tying either package to a function – which the combat hypothesis does.

    Second thing, which I think is more of a problem: apatosaurine cervical ribs aren’t just a little bit bigger than those of other sauropods, they’re GROSSLY immense. If they existed just to balance epaxial muscle and tendon forces, where’s the corresponding gross expansion of the neural arch and spines? Except for maaaaybe A. louisae, most apatosaurines don’t seem to have neural spines that are any larger than those of other diplodocids – and several, like the Chicago and Tokyo specimens, have pretty small neural spines.

    Third thing: I can see displacing the cervical ribs a long way out from the centrum to give them more leverage in your force-balancing scenario, but why make them so robust? Camarasaurs have cervical ribs that are out pretty far laterally, but the ribs themselves are still slender, as are the diapophyseal and parapophyseal rami that support them. So we know that sauropods didn’t have to make their cervical ribs and supporting bony struts massive just because they were a long way out from the centra. (I hadn’t thought about that until I was writing this, so hey, new thought!)

    It’s late and that’s some stuff pulled off the top of my head. I’m not trying to tear down your idea – far from it! Hoping instead that we can bash through some of these points and all emerge a little more knowledgeable. So please fire back when you get the time – not just Scott, but everyone who has commented so far, as well as those of you who have questions or comments but are so far holding fire.

  15. Mike Taylor Says:

    Matt noted, in passing:

    I can’t remember if it was in a paper or a blog post, but Mike once made the point that we don’t even know how much soft tissue to put around sauropod necks.

    You may be thinking of the section in Sauropods were corn-on-the-cob, not shish kebabs where I showed the neck cross-sections of Paul (1997:fig. 4) and Schwarz et al. (2007:fig. 7a) — which to my knowledge are still the only published sauropod neck-cross sections. They are wildly, wildly different.

    Honestly, the issue is not so much that we don’t know how much soft tissue sauropods had around their cervical vertebrae, it’s that no-one has yet even attempted to analyse it, beyond the very qualitative discussions in Schwarz et al. and Taylor and Wedel (2013a) in the “Extent of soft-tissue relative to size of vertebrae” section.

    There’s a project there waiting for someone with the time, inclination and comparative anatomy expertise.

  16. Mike Taylor Says:

    Matt wrote:

    As far as texture on the ‘clubs’ that would indicate attached soft tissue, I think it’s a wash. The OMNH specimen is too sanded, and I haven’t seen the relevant CM or YPM specimens up close enough, BUT given how much subtle restoration and lacquer those specimens include, I’m not optimistic.

    Unfortunately, I strongly agree. Check out our old post The Field Museum’s photo-archives tumblr, featuring: airbrushing dorsals for a frightening example of how “restored” old sauropod specimens are. And of course see Barbour’s (1890) classic rant about Marsh’s fictionalised fossils.


  17. Thanks Matt. There’s a lot of good stuff in your post, so let me try to unpack most of my thoughts. Regarding ligament height, I think Tsuihiji simply makes a more compelling case for osteological correlates. Schwartz et al.’s entire basis for the rejection of an avian-style supraspinal ligament (“ligamentum nuchae” of Tsuihiji) comes down to this paragraph:

    “As in extant crocodylians (Frey 1988; Salisbury 2001), the heights of the neural spines in the neck of Diplodocidae gradually increase from cranial to caudal (Appendix 2, Fig. 7). In contrast, the cervical neural spines of extant birds are high in the cranial and caudal neck region, but very low in the middle cervical region (Appendix 2). For the neck of Diplodocidae, we therefore infer a configuration of the
    supraspinal ligament similar to that of extant crocodylians, connecting the apices of each neural spine with each other.”

    I don’t agree that the change in height is the best way to determine the soft-tissue configuration here, as in several other ways the neural spines of sauropods are more like birds than crocodilians (including of course that some birds actually have bifurcated neural spines). Second, I think they misunderstand (or simply don’t consider) the correlation between posture, vertebral column curvature, and height of neural spines. I think the pattern of neural spine height in extant birds is related to the extreme flexion of avian necks; as a result the mechanical leverage (and thus desirable height of a neural spine or depth of the supraspinal ligament) in birds will not be identical to the more moderately curved necks of sauropods. Also note that while diplodocids see a general increase in neural spine height down the cervical series (superficially like crocs), in apatosaurines (those with the most pronounced system of neural spine bifurcation) the combination of dorsal column curvature and neural spine increase in the anterior dorsals is very different from crocs, and is suggestive of a supraspinal ligament system that combines an avian-style highly- organization with some functional aspects that parallel ungulates, with anteriorly-facing neural spines of the back giving rise to the base of the supraspinal ligament in a similar way as the withers of ungulates.

    I could (at the expense of many words) make a similar argument about the medial axial neck musculature, where Schwarzt et al admit that EPB suggests there should be such a system (present in crocs and birds…page 182 if you’re looking) mostly because they are using a too-straight version of the vertebral column (cf Hatcher, Parrish & Stevens, etc.) and because they can’t imagine how Amargasaurus would have evolved its neural spines if other diplodocids had extensive dorsal axial musculature on the neck (I think this is pretty easy to do since necks like that of Dicraeosaurus were obvious intermediates on the way to Amargasaurus, but that’d be another 1000 words).

    There’s more (and some of it I’d rather hold closer to my vest publicly, so shoot me an email if you want to go into the nitty gritty), but TL;DR for neck soft tissue is that I find an updated and completed version of Tsuhiji’s reconstruction in figure 5b to be more likely that the Schwartz et al., reconstruction for dorsal axial soft tissues.

    I’ll just post this and then address the other question I think…


  18. OK, so starting from the interpretations I outlined in the previous post (TL:DR – avian-like deep supraspinal ligaments as well as deeper dorsal axial neck muscles in diplodocids; especially in apatosaurines, where the curvature of the back and neck combined with the height and orientation of the dorsal neural spines suggests they have deeper soft-tissue over their necks posteriorly than other diplodocines).

    Matt, you said:

    “First, as you allude to, having a lot of tension members dorsally isn’t mutually exclusive with the combat hypothesis. And in a way, saying that all the ventral gloop was only there to balance the dorsal gloop is still unsatisfying, because it doesn’t explain what apatosaurines were doing that required all that extra gloop in the first place. In short, it explains one package of weird morphology with another package of weird morphology, without tying either package to a function – which the combat hypothesis does.”

    I continue to stipulate that my comments aren’t mutually exclusive with combat, though it could potentially change the form that combat takes. But I could also hypothesize that it reflects differences in food gathering strategy. Ungulates vary the length of their necks and the thickness of their nuchal ligament system quite a lot (even within the same genus), and it is presumed that they do so based on the energetics of feeding (I made a presentation on similar arguments for hadrosaurs, FWIW). Optimizing efficiency via muscles and/or elastic ligaments depends on what sort of feeding behaviors are most common, e.g. do you feed mainly at one height or do you need to move your neck through a wide vertical range. And if you often feed at one height, are you routinely keeping your neck in a more neutral position, or are you holding it where elastic ligaments have to be opposed. So it’s not hard to imagine that differences in neck thickness and length in diplodocids reflects some sort of niche partitioning related to feeding envelope; analogous to the differences we see in ungulates, but supersized for sauropods.

    Matt: “Second thing, which I think is more of a problem: apatosaurine cervical ribs aren’t just a little bit bigger than those of other sauropods, they’re GROSSLY immense. If they existed just to balance epaxial muscle and tendon forces, where’s the corresponding gross expansion of the neural arch and spines? Except for maaaaybe A. louisae, most apatosaurines don’t seem to have neural spines that are any larger than those of other diplodocids – and several, like the Chicago and Tokyo specimens, have pretty small neural spines.”

    The simple answer here is that the nuchal ligament and muscular expansion is driven by the change in curvature and dorsal neural height (and to a degree the expansion of the bifurcated region of neural spines), not by the cervical neural spine size itself.

    Matt: “Third thing: I can see displacing the cervical ribs a long way out from the centrum to give them more leverage in your force-balancing scenario, but why make them so robust? Camarasaurs have cervical ribs that are out pretty far laterally, but the ribs themselves are still slender, as are the diapophyseal and parapophyseal rami that support them.”

    I could arm-wave that there would be larger forces exerted on the cervical ribs (since I think there’s more and better-leveraged soft tissue opposing them in apatosaurines than in camarasaurs), so it could be a solution for avoiding mechanical failure. But without calculating the forces involved and such this is pretty speculative (and also not exclusive of fighting). But my point isn’t that you’re wrong, but that alternative hypotheses can be constructed, so the key is whether fighting better explains the features you mustered in defense of fighting.

    The odd curved morphology of the cervical ribs is what I find most compelling about your argument (that and the awesome illustrations it should spawn!). Short a finite element analysis that shows apatosaurine cervical shape is an ideal way to oppose plausible neck forces generated with the soft-tissue regime I infer, I agree that it is calling out for some other explanation, and a series of bony cudgels seems as plausible as any other interpretation, and better than most.

    Phew. For not really disagreeing that was a lot of writing. Back to working on my prospectus now…

  19. Mike Taylor Says:

    “I think the curvature of the back and neck in apatosaurines combined with the height and orientation of the dorsal neural spines in apatosaurines suggests they have much deeper necks posteriorly than other diplodocines.”

    Huh, wait, what? Than other diplodocines? Is this one of the typos you mentioned? But anyway, the curvature of the back and neck and the height and orientation of the dorsal neural spines is pretty much the same in apatosaurines as in diplodocines, isn’t it? The differences lie elsewhere.

  20. Matt Wedel Says:

    Thanks, Scott, for the thoughtful and detailed responses. I don’t really disagree with anything you wrote. I agree that using neural spine height to predict muscular or ligamentous anatomy seems…ill-founded. One important difference between crocs and mammals on one hand, and birds and sauropods on the other, is that the necks of crocs and mammals are short enough for individual muscles to span the whole distance from the torso to the head, which except for the biventer cervicis just doesn’t happen in birds, and probably did not happen in sauropods, either. So anytime I see an argument about sauropod necks that relies on croc necks, which have a whole other category of muscles that sauropods probably lacked, my antennae go up.

    Anyway, all useful points – thanks for contributing. No worries on the need to keep some things close to the vest for now. I’m grateful to you for sharing as much as you have. I’ll be in touch a little closer to our submission point to sound you out on how, or if, to raise some of these points in the manuscript.


  21. No, that one’s actually not a type Mike. Excluding Supersaurus (due to some uncertainty in both axial curvature and phylogeny) I reconstruct apatosaurine dorsal series as being more strongly curved, and the dorsal neural spines increasing more in height (relative to those of the cervicodorsal transition). I also find that diplodocine necks seem to flex more sharply over a smaller number of posterior cervicals, while apatosaurine necks curve up in a more “sweeping” curve over a greater number of posterior cervicals. The result is a significant difference in the mechanical leverage available to the two groups via expansion of nuchal ligaments and dorsal axial musculature, which you can see reflected in my skeletal drawings. The range where that expansion occurs lines up fairly closely with the range of bifurcated neural spines in the two groups as well. All of this holds true even if we pull both groups’ necks into a more S-shaped curve as you would probably prefer.

  22. Mike Taylor Says:

    “No, that one’s actually not a type Mike.”

    Or a typo? :-)


  23. Thanks Matt. There’s a lot of good stuff in your post, so let me try to unpack most of my thoughts. Regarding ligament height, I think Tsuihiji simply makes a more compelling case for dorsal soft tissue inference. Schwartz et al.’s entire basis for the rejection of an avian-style supraspinal ligament (“ligamentum nuchae” of Tsuihiji) seem to come down to this paragraph:

    “As in extant crocodylians (Frey 1988; Salisbury 2001), the heights of the neural spines in the neck of Diplodocidae gradually increase from cranial to caudal (Appendix 2, Fig. 7). In contrast, the cervical neural spines of extant birds are high in the cranial and caudal neck region, but very low in the middle cervical region (Appendix 2). For the neck of Diplodocidae, we therefore infer a configuration of the supraspinal ligament similar to that of extant crocodylians, connecting the apices of each neural spine with each other.”

    I don’t agree that the change in height is the best way to determine the soft-tissue configuration here, as in several other ways the neural spines of sauropods are more like birds than crocodilians (including of course that some birds actually have bifurcated neural spines). Second, I think they misunderstand (or simply don’t consider) the correlation between posture, vertebral column curvature, and the location of changes in neural spine height. I think the pattern of neural spine height in extant birds is related to the extreme flexion of avian necks; as a result we shouldn’t expect the mechanical leverage (and thus desirable height of a neural spine or depth of the supraspinal ligament) in birds to be identical to the more moderately curved necks of sauropods. Also note that while diplodocids see a general increase in neural spine height down the cervical series (superficially like crocs), in apatosaurines (those with the most pronounced system of neural spine bifurcation) the combination of dorsal column curvature and neural spine height increase in the dorsals is very different from crocs, and is suggestive of a supraspinal ligament system that combines an avian-style organization with functional aspects that parallel ungulates. Specifically I think the anteriorly-facing neural spines of the back that are significantly taller than the proceeding neural spines (due to both height increase and back curvature) give rise to the base of the supraspinal ligament in a manner analogous to the withers of ungulates.

    I would make similar arguments about the medial axial neck musculature, where Schwarzt et al admit that EPB suggests there should be such a system (present in crocs and birds…page 182 if you’re looking) mostly because they are using a too-straight version of the vertebral column (cf. Hatcher, Parrish & Stevens, etc.) and because they can’t imagine how Amargasaurus would have evolved its neural spines if other diplodocids had extensive dorsal axial musculature on the neck (I think this is pretty easy to do since necks like that of Dicraeosaurus were obvious intermediates on the way to Amargasaurus, but that’d be another 500 words).

    There’s more (and some of it I’d rather hold closer to my vest publicly, so shoot me an email if you want to go into the nitty gritty), but TL;DR for neck soft tissue is that I find an updated and completed version of Tsuhiji’s reconstruction in figure 5b to be more likely that the Schwartz et al., reconstruction for dorsal axial soft tissues (that’s also why I view their neck cross-sections as squares with triangles on top, rather than strict Toblerones).

    I’ll just post this and then address the other question I think…

  24. Mike Taylor Says:

    “The simple answer here is that the nuchal ligament and muscular expansion is driven by the change in curvature and dorsal neural height (and to a degree the expansion of the bifurcated region of neural spines), not by the cervical neural spine size itself.”

    If increased epaxial musculature can be achieved simply by raising the muscle origins without needing to make the neural spines bigger, why wouldn’t corresponding increased hypaxial musculature similarly be achieved by lowering the muscle origins (deeper torso) without making the cervical ribs bigger?

    The main issue about apatosaurs is: why the dorsal/ventral asymmetry? Why is the ventral part of the neck skeleton hypertrophied and not the dorsal part?


  25. Mike Taylor wrote:

    “If increased epaxial musculature can be achieved simply by raising the muscle origins without needing to make the neural spines bigger, why wouldn’t corresponding increased hypaxial musculature similarly be achieved by lowering the muscle origins (deeper torso) without making the cervical ribs bigger?

    The main issue about apatosaurs is: why the dorsal/ventral asymmetry? Why is the ventral part of the neck skeleton hypertrophied and not the dorsal part?”

    Because the base of the neck curves up, not down. With the neck curving away from your point of origin ventrally there isn’t significant leverage to be gained by moving the muscle origins out, which means you’re pretty much stuck with having to lengthen the lever arm (i.e. making the cervical ribs longer so the insertions are further from point of rotation). Monkeying around with origin depth ventrally might also risk interfering with locomotor modules (which are irritatingly occupying some of the same space) but simple neck geometry and leverage are the most immediate culprits.


  26. […] putting together our thoughts on how apatosaurs used their necks, we were motivated by genuine curiosity — which in Matt’s and my case, at least, goes […]

  27. LeeB Says:

    Shock absorption by air sacks has been considered in gannets which plunge dive the same way brown pelicans do.

    There is a paper on this here: http://www.seabirdgroup.org.uk/journals/seabird_21/SEABIRD%2021%20(2008)%20Daoust%20et%20al.64-76.pdf

    It might be interesting to compare the air sac system of Brown pelicans with that of other non-plunge diving pelicans, and with gannets and boobies and the unrelated terns which also plunge dive.

    LeeB.


  28. […] the morning of Tuesday 1st December, on SVPCA day 1, I gave my talk about apatosaur neck combat. In one of the afternoon sessions, I sat next to Bob Nicholls, and found myself thinking how […]

  29. Duane Nash Says:

    A little late to this but what about the suggestion that those big necks helped to barrel through thick vegetation or even topple trees? Of course bashing conspecifics and theropods also works… And to add a new idea how about the neck adding traction and stability to cover rough terrain. I know it sounds weird but elephants will often use their trunk as a bit of 5th limb when clambering up banks. I don’t think we always have to imagine apatosaurs in wide open spaces. Dense forests full of snags and uneven terrain could have have necessitated creative use of the neck and tail to push through such habitat with even the neck pushing off the ground to add stability at times.

  30. Matt Wedel Says:

    If the robustness of the neck were only or mostly for bashing through vegetation, I’d expect to see similar adaptations in other sauropods. There should be more of a spectrum. As it is, no other sauropods really come close. Camarasaurs have wide cervical ribs but they aren’t any more robust than what you see in brachiosaurs, etc.

    To me, the key thing is that apatosaurs necks are so solidly built, which so much more bony tissue than other sauropods, that they must have been much heavier. That flies in the face of a looong evoluionary trend, stretching back through basal sauropodomorphs, saurischians, dinosaurs, and dinosauromorphs, of making the neck longer and lighter. Structures that bizarre occasionally have purely feeding or locomotor functions, but usually they have something to do with sexual display or combat.

    Two other things to note: why use the neck, which even in apatosaurines was a lot less massive than the body, to smash aside vegetation, instead of just walking forward, threading the neck between the trees, and pushing them down with the limbs and torso? Also, if the neck was massive to push aside vegetation, I’d think that it would be better adapted for lateral than ventral excursions – the morphology of apatosaur vertebrae suggests the opposite.

  31. Duane Nash Says:

    Good points we should expect to see more of a spectrum if brush clearing/tree toppling/difficult terrain leveraging was the prime driver. I think you are right that sexo-social drives was the main reason for apatosaurine necks.

    But this does beg the question why did no other sauropods evolve neck clobbering anatomy like apatosaurines? Well there is Amargasaurus, various tail clubbed sauropods, and Agustinia (whatever that guy really looks like who knows?).

    Any indication that some sauropods were adapted for lateral blows with their necks?

    Furthermore maybe some titanosaurs moved away from combat and into more intricate flexibility displays. I recall something about intertwining of the neck and tail in certain species? That would be amazing.

  32. Matt Wedel Says:

    But this does beg the question why did no other sauropods evolve neck clobbering anatomy like apatosaurines?

    That is a fantastic question, and it bugs the hell out of me. Usually when one lineage goes in for something like this, related lineages that have the same basic gear at least take some steps in the same direction. Take headgear in artiodactyls – horns in bovids, horns with sheddable sheaths in pronghorns, antlers in deer, ossicones in giraffids, etc. And how many clades of birds have evolved long pretty tail feathers?

    So whether apatosaurines were using their necks for combat or something else, it is weird that other sauropods don’t really get that close in terms of cervical rib morphology. Camarasaurus and Isisaurus both have wide cervical ribs, but not robust ones – and they’re both macronarians. The other diplodocoids all have slender cervical ribs, on short parapophyses. I assume that somewhere out there is the transitional apatosaurine with mid-sized cervical ribs. But that still leaves the question of why no other known sauropods went for apatosauresque cervical ribs. I’m tempted to say that it’s because selection for long, light necks was apparently pretty strong in sauropods, and fat, heavy necks reverse that. But then again, selection for flight in volant birds is strong, but peacocks, lyrebirds, long-tailed widowbirds, and plenty of others presumably pay a price for their tails in flight performance. So who knows.

    Furthermore maybe some titanosaurs moved away from combat and into more intricate flexibility displays. I recall something about intertwining of the neck and tail in certain species? That would be amazing.

    Yeah, that’s an interesting suggestion. A lot of titanosaurs have wide, flat, cervical zygapophyses for good lateral flexibility, and the weird placement of the zygs in saltasaurines is something I’ve seen elsewhere in only in owls, which have crazy flexible necks. And titanosaur tails are all over the place with funky articulations. So maybe they were lovers, not fighters. :-)

  33. brnngh!!! Says:

    While we don’t see many examples of neck clobbering looking adaptations in sauropods, we do see many examples of what look like display structures on sauropod necks. It’s possible that in many taxa visual signaling was the primary means of settling territorial/sexual dominance related disputes & in those taxa tails, manus claws and regular old sauropod-strength necks sufficed when visual signaling gave way to occasional ritualized combat. Perhaps in the super-charged sauropod-thick Morrison ecosystem you just had so much diversity that there was an oddball group that really got into the dangerous business of neck-clobbery.

    My prediction is that as more sauropod rich ecologies come to light we’ll see lots of other weird neck adaptations, & perhaps even other non-apatosaurine taxa with what appear to be battle-necks. The fact of the matter is there just aren’t that many really complete sauropods.

  34. Mike Taylor Says:

    That’s right, Brian — in particular we have the crazy-tall necks of dicraeosaurids, taken to extremes in Amargasaurus, and the surprisingly broad neck of Barosaurus — and, independently, of Puertasaurus. There is a lot of weird neck morphology going on in different sauropods, and the idea that they were all basically tubes (as illustrated in every drawing/painting before about 1980) couldnt be more wrong.

    And, importantly:

    “The fact of the matter is there just aren’t that many really complete sauropods.”

    This is so, so true. (I have an in-progress short manuscript that makes essentially the exact same point as the linked blog-post, just so I have something to cite when I make this point in other papers.) Basically, we’re working with a few crushed, distorted and disarticulated vertebrae, and that’s just for the good necks.


  35. […] If we accept that the distinctive ventral projections of the gigantic and ventrally displaced cervical ribs of apatosaurs were likely the base of some form of soft-tissue rugosity — such as keratinous horns like those of rhinos — then does it follow that those necks were used in combat as we suggested? […]


  36. […] in the process of putting together artwork to illustrate our apatosaur neck combat hypothesis, Brian tried out a whole bunch of outlandish concepts. Here are two that he showed us, but which […]

  37. Ictonyx Says:

    I know I’m late to this but I have a question.

    Mike Taylor mentions in the post that an apatosaurine cervical series might weight about twice as much as a diplodocine one of similar length, and explains that the energetic expenditure involved begs for an explanation. I find this pretty convincing, but would find it more convincing if the rest of the apatosaurine’s body was more in line with a diplodocine.

    Morrison diplodocid skeletals (e.g. by Scott Hartman and Greg Paul) show Apatosaurus to be a much more robust animal than Diplodocus or Barosaurus. Besides the neck, the torso is very deep and broad and the limb bones very chunky. As far as I know, mass estimates for Apatosaurus are often close to twice those for Diplodocus.

    So my question is, is it clear to people working on sauropods that the robusticity of apatosaurine necks is extreme enough, relative to the overall robusticity of their bodies, to require a separate explanation? Or could it just be that their necks are twice as big as a diplodicine’s because THEY are twice as big as a diplodocine?

    Note: I am an interested layperson not a palaeontologist; this is a genuine question, I have no pre-conceived ideas about the answer.

  38. Mike Taylor Says:

    You’re right that apatosaurs are robust all over, although that trend is taken to its extreme in the neck. In fact, one of the first statements about sauropods that ever struck me strongly enough to make a note of it was this, from Matt Wedel, back before I was in contact with him: “Apatosaurus sort of looks like a pro wrestler when most of the other sauropods tend to look like ballerinas”. That was in an interview with the Prehistoric Planet web-site, which sadly they seem to have lost. There’s a little bit more of that interview on an old and incomplete page of mine.

    I think that when you look at skeletons of apatosaurines and diplodocines together, the difference in robustness of the neck is pretty evidently more extreme than what is seen elsewhere. Check out this photo of the old Carnegie mounts, and this newer one of the same individuals in their newer poses. Note in particular that the tail of Apato is less substantial than that of Diplo.


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 )

Google+ photo

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

Connecting to %s

%d bloggers like this: