How fat was Camarasaurus?

January 16, 2011

For reasons that will soon become apparent (yes, that’s a teaser), Matt and I wanted to figure out how heavy Camarasaurus was.  This is the story of how I almost completely badgered up part of that problem.  I am publishing it as a cautionary tale because I am very secure and don’t mind everyone knowing that I’m an idiot.

Those who paid close attention to my recent paper on Brachiosaurus and Giraffatitan will remember that when I estimated their mass using Graphic Double Integration (Taylor 2009: 802-804) I listed separately the volumes of the head, neck, forelimbs, hindlimbs, torso and tail of each taxon.  In Giraffatitan, the torso accounted for 71% of the total volume (20588 of 29171 litres), and in Brachiosaurus, 74% (26469 of 35860 litres), so it’s apparent that torso volume hugely dominates that of the whole animal.  In the giant balloon-model Giraffatitan of Gunga et al.’s (1995, 1999) estimates, the torso accounted for 74% of volume (55120 of 74420 litres) so even though their fleshing out of the skeleton was morbidly obese, the relative importance of the torso came out roughly the same.  Finally, Gunga et al’.s (2008) revised, less bloated, model of the same Giraffatitan had the torso contributing 68% of volume (32400 of 47600 litres).  So far as I know, these are all of the published accounts that give the volumes of separate parts of a sauropod body, but if there are any more, please tell me in the comments!   (Odd that they should all be for brachiosaurids.)

3D “slim” version of reconstruction of the “Brachiosaurus” brancai mounted and exhibited at the Museum of Natural History in Berlin (Germany).  A. Side view, upper panel; B. top view, lower panel.  The cross in the figure of upper panel indicates the calculated center of gravity.  (Gunga et al. 2008: figure 2)

So it’s evident that, in brachiosaurs at least, the torso accounts for about 70% total body volume, and therefore for about that much of the total mass.  (The distribution of penumaticity means that it’s denser than the neck and less dense than the limbs, so that its density is probably reasonably close to the average of the whole animal.)

Now here’s the problem.  How fat is the sauropod?  Look at the top-view of Giraffatitan in the Gunga et al. figure above: it’s easy to imagine that the torso could be say 20% narrower from side to side, or 20% broader.  Those changes to breadth would affect volume in direct proportion, which would mean (if the torso is 70% of the whole animal) a change in total body volume of 14% either way.  Significant stuff.

So what do we know about the torso breadth in sauropods?  It obviously dependant primarily on the orientation of the ribs and their articulation to the dorsal vertebrae.  And what do we know about that?


Well, OK, I am over-simplifying a little.  It’s been mentioned in passing in a few papers, but it’s never been discussed in any detail in a published paper that I know of.  (There’s a Masters thesis out there that starts to grapple with the subject, but I don’t know whether I should talk about that while it’s still being prepared for publication, so I won’t say anything more.)  The most important published contribution is more than a century old — Holland’s (1910) smackdown of Tornier’s and Hay’s comical Diplodocus postures, which included the following cross-sections of the torsos of several animals at the seventh dorsal vertebra:

(This figure previously appeared on SV-POW! in Matt’s post, Sauropods were tacos, not corn dogs, which as far as I am aware is the only existing non-technical treatment of sauropod torso-shape.)

Holland unfortunately did not discuss the torso shape that he illustrated, merely asserting it.  Presumably it is based on the mounted skeleton of the Diplodocus carnegii holotype CM 84, which is at the Carnegie Museum in Pittsburgh, where Holland was based.  I have no reason to doubt it; just noting that it wasn’t discussed.

All right then — what about Camarasaurus?  I think it’s fair to say that it’s generally considered to be fairly rotund among sauropods, as for example this skeletal reconstruction by Greg Paul shows:

Camarasaurus lentus skeletal reconstruction, in dorsal and right lateral views. (Paul 2010:197)

Measuring off the height and width of the torso at the seventh dorsal vertebra, using GIMP, I find that they are 341 and 292 pixels respectively, so that the eccentricity is 341/292 = 1.17.  This compares with 1760/916 = 1.92 for Holland’s Diplodocus above, so if both figures are accurate, then Camarasaurus is much fatter than Diplodocus.

But is Paul’s Camarasaurus ribcage right?  To answer that, I went back to my all-time favourite sauropod paper, Osborn and Mook’s (1921) epic descriptive monograph of Camarasaurus (and Cope’s other sauropods).  I knew that this awesomely comprehensive piece of work would include plates illustrating the ribs; and in fact there are four plates that each illustrate a complete set of dorsal ribs (although the associations are doubtful).  Here they all are:

Left dorsal ribs of Camarasaurus (Osborn and Mook 1921:pl. LXXVIII)

Left dorsal ribs of Camarasaurus (Osborn and Mook 1921:pl. LXXIX)

Left dorsal ribs of Camarasaurus (Osborn and Mook 1921:pl. LXXX)

Left dorsal ribs of Camarasaurus (Osborn and Mook 1921:pl. LXXXI)

But hang on a minute — what do you get if you articulate these ribs with the dorsal vertebrae?  Osborn and Mook also provided four plates of sequences of dorsal vertebrae, and the best D7 of the four they illustrate is probably the one from plate  LXX.  And of the four 7th ribs illustrated above, the best preserved is from plate LXXIX.  So I GIMPed them together, rotated the ribs to fit as best I could and …

What on earth?!

I spent a bit of time last night feeling everything from revulsion to excitement about this bizarre vertebra-and-rib combination.  Until I happened to look again Osborn and Mook — earlier on, in the body of the paper, in the section about the ribs.  And here’s what I saw:

(Note that this is the vertebra and ribs at D4, not D7; but that’s close enough that there’s no way there could be a transition across three vertebrae like the change between this and the horrible sight that I presented above.)

What’s going on here?  In the plates above, the ribs do not curve inwards as in this cross-section: they are mostly straight, and in many case seem to curve negatively — away from the torso.  So why do O&M draw the ribs in this position that looks perfectly reasonable?

And figure 70, a few pages earlier, makes things even weirder: it clearly shows a pair of ribs curving medially, as you’d expect them to:

So why do these ribs look so totally different from those in the plates above?

I’ll give you a moment to think about that before I tell you the answer.

Seriously, think about it for yourself.  While you’re turning it over in your mind, here is a picture of the beautiful Lego kit #10198, the Blockade Runner from the original Star Wars movie.  (I deeply admire the photography here: clear as a bell.)

OK, welcome back.

Got it?  I bet most of you have.

The answer was right there in figure 71:

Osborn and Mook 1921:fig. 71. Left rib of Camarasaurus supremus Cope. Rib 4 (Amer. Mus. Cope Coll. No. 5761/R-A-24). (A) direct external view when placed as in position in the body; (B) direct anterior when placed as in position in the body. Capit. capitulum; Sh. shaft; Tub. tuberculum. Reconstructed view, portion in outline.

Osborn and Mook 1921:fig. 71. Left rib of Camarasaurus supremus Cope. Rib 4 (Amer. Mus. Cope Coll. No. 5761/R-A-24). (A) direct external view when placed as in position in the body; (B) direct anterior when placed as in position in the body. Capit. capitulum; Sh. shaft; Tub. tuberculum. Reconstructed view, portion in outline.

And, my word, isn’t it embarrassingly obvious once you see it?  I’d been blithely assuming that the ribs in O&M’s plates were illustrated in anterior view, with the capitula (which articulate with the parapophyses) located more medially, as well as more ventrally, than the tubercula (which articulate with the diapophyses).  But no: as in fact the captions of the plates state perfectly clearly — if I’d only had the wits to read them — the ribs are shown in “external” (i.e. lateral) view.  Although it’s true that the capitula in life would indeed have been more medially positioned than the tubercula, it’s also true that they were more anteriorly positioned, and that’s what the plates show at the rib heads.  And the curvature that I’d been stupidly interpreting as outward, away from the midline, is in fact posteriorly directed: the ribs are “swept back”.  The ventral portions of the ribs also curve medially, away from the viewer and into the page … but of course you can’t see that in the plates.

The important truth — and if you take away nothing else from this post, take this — is that I am dumb bones are complex three-dimensional objects, and it’s impossible to fully understand their shape from single-view illustrations.  It’s for this reason that I make an effort, when I can, to illustrate complex bones from all cardinal directions — in particular, with the Archbishop bones, as for example “Cervical S” in the Brachiosaurus coracoid post.

Because ribs, in particular, are such complex shapes — because their curvature is so unpredictable, and because their articulation with the dorsal vertebrae is via two points which are located differently on successive vertebrae, and because this articulation still allows a degree of freedom of movement — orthogonal views, even from all cardinal directions, are of limited value.  Compositing figures will give misleading results … as demonstrated above.  PhotoShop is no more use here.  Fly, you fools!

Paradoxically, our best source of information on the shapes of saurpod torsos is: mounted skeletons.  I say “paradoxically” because we’ve all grown used to the idea that mounts are not much use to us as scientists, and are really there only as objects of awe.  As Brian Curtice once said, “A mounted skeleton is not science.  It’s art.  Its purpose is to entertain the public, not to be a scientifically accurate specimen”.  In many respects, that’s true — especially in skeletons like that of the “Brontosaurus” holotype, YPM 1980, where the bones are restored with, and in some cases encased in, plaster so you can’t tell what’s what.  But until digital scanning and modelling make some big steps forward, actual mounted skeletons are the best reference we have for the complex articulations of ribs.

Giraffatitan brancai paralectotype HMN SII, composite mounted skeleton, torso in left posteroventrolateral view (photograph by Mike Taylor)

And I finish this very long (sorry!) post with yet another note of caution.  Ribs are long and thin and very prone to damage and distortion.  It’s rare to find complete sauropod ribs (look closely at the O&M plates above for evidence), but even when we do, we shouldn’t be quick to assume that the shape in which they are preserved is necessarily the same as the shape they had in life.  (If you doubt this, take another look at rib #6 in the third of the four O&M plates above.)  And as if that weren’t enough to discourage us, we should also remember that the vertebra-rib joints would have involved a lot of cartilage, and we don’t know its extent or shape.

So bearing in mind the complicated 3D shape of ribs and of dorsal vertebrae, the tendency for both to distort during and after fossilisation, and the complex and imperfectly known nature of the joints between them, I think that maybe I wasn’t too far wrong earlier when I said that what we know about sauropod torso shape is: nothing.

It’s a sobering thought.


About these ads

17 Responses to “How fat was Camarasaurus?”

  1. Heinrich Mallison Says:

    What can I say – there is nothing more scientific than a editable 3D mounted skeleton based on high resolution scans of the real bones of a complete individual.

    but Mike, shame on you! Using 2D data….. that MUST screw with your brain! ;)

    and you just wait: one day I will digitally remount the Berlin Giraffatitan perfectly, and you’ll turn green from envy ;)

  2. [...] This post was mentioned on Twitter by Bec Crew and JP – Research Lab. JP – Research Lab said: How fat was Camarasaurus?: For reasons that will soon become apparent (yes, that’s a teaser), Matt and I wanted … [...]

  3. Mike Taylor Says:

    Heinrich, am I right in thinking that the digital data already exists to model the Berlin skeleton? Did Gunga et al. do the necessary scanning?

    But that still leaves us with the problems of distortion, degree of freedom, and unknown articular cartilage.

  4. Vertebrat Says:

    The distribution of penumaticity means that it’s denser than the limbs and less dense than the neck

    Other way round, right?

    [Yes, the other way round. Thanks for spotting this, now fixed -- Mike.]

  5. 2D => 3D = bad times. But you are some writer to turn that technical info into something readable. Thank god for pop culture.

  6. Heinrich Mallison Says:

    Mike, there is a laser scan of the old mount at relatively high res, and a less complete one of the new mount, but no data on separate bones except for what I mechanically digitized while the animal was dismounted. :(

    Yes, we will have lots of uncertainties left, but (see Kentrosaurus and Plateosaurus) those can be overcome at least in part by using articulated finds, comparison with extant animals, and patient 3D articulation.

  7. Mike,

    Great discussion. You caught yourself, and were it not for me, you’d be your own worst critic. (I kid; seriously, I kid.)

    As an artist who’s reconstructed skeletons in the fashion seen above, as well as in cross-section, and attempting to recover a view of the animal in a volumetric rather than flat way, this question of volume becomes very important, and one that is heavily overlooked in various analyses.

    This is why a simple flat skeleton as from Liaoning tends to confuse people in reconstructions, because apart from their 2D arrangement, they are also flattened and distorted, and this also affects how they are viewed by artists — and scientists — attempting to reconstruct life models due to these perceptions (as well as make arbitrary modifications without a substantive level of restraint involved).

  8. Asier Says:

    Dear all,

    It is really difficult to know how fat was any sauropod, but we can compare to an extant analog, the elephants. Today and extinct elephants had similar body structure as sauropods, with long legs like pillars and huge abdomens.

    I study the proboscidea an I can say that is probably that Camarasaurus like in proboscideans, should had the ribs from the spines almost horizontal from the spina before the pelvis bone, while, at the shoulders the ribs almost pointing downwords, because the scapulas, but elephants has/had the scapulas higher than Camarasaurus, so the later should had broader chest, and also the abdomen should had been a bit wider than pelvis breadth.

  9. Well, I’m glad someone has finally shed the light on this largely ignored. It has often been the most exasperating thing I have come across when trying to estimate the mass of sauropods. The fact that nearly no new sauropods that have been described in the past 20 years (and even earlier) adequately figure the dorsal ribs (when known) makes it almost complete guess work when trying to do a dorsal view skeletal restoration for a sauropod.

    I hope that when that Master’s thesis is published, some dorsal views of the resulting torso shaped are shown.

    And Mike, I’m still waiting for your critique of my Giraffatitan GDI….

  10. [...] I don’t know if the ribcage articulation is perfect but it looks reasonable, and as we saw last time that is a major consideration. Since Heinrich built the digital skeleton in digital space, he knows [...]

  11. Adam Polanski Says:

    Wait a minute! The Berlin skeleton of Brachiosaurus is the PARAlectotype? When and why did the paratroopers sail in?

  12. Mike Taylor Says:

    Adam Polanski asked: “Wait a minute! The Berlin skeleton of Brachiosaurus is the PARAlectotype? When and why did the paratroopers sail in?”

    Darn, I’d hoped no-one would pick up on that :-) I made a dumb nomenclatural mistake in the 2009 paper (where I nominated SII as the lectotype). There’s a short correction in press at SVP which credits the person who pointed this out to me. I’ll post about it (briefly since it’s not very exciting) when it comes out.

  13. Adam Polanski Says:

    Thank you Mike. It’s not often I get a chance to nitpick like this. ;)

    Hugs, and love your work!

  14. [...] paralectotype specimen of Giraffatitan brancai.  (Butchering a wallaby, photographing big bones, How fat was Camarasaurus, and baby giraffe neck, in case you were wondering.) Giraffatitan brancai paralectotype HMN SII in [...]

  15. [...] more so than in most other tetrapods (see this and this, and for a more pessimistic take, this). This is something that is easy to get wrong; we are used to seeing round mammalian torsos and a [...]

  16. […] jumped the gun a bit in asking How fat was Camarasaurus? a couple of years ago, or indeed How fat was Brontosaurus? last year. As always, we should have […]

  17. […] The first is that a lot of sauropods had distinctly non-round body cross-sections (Diplodocus, Camarasaurus). So assuming circular cross-sections might inflate the body well beyond its likely volume. Second […]

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 )

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


Get every new post delivered to your Inbox.

Join 391 other followers

%d bloggers like this: