Genesis of an instant palaeo-art classic

March 3, 2011

After a couple of relatively hardcore posts on ilial osteology, we though it would be good to look at something lighter this time.  If you’re interested in dinosaurs, or indeed alive, you will hardly have been able to avoid seeing Francisco Gascó’s glorious life restoration of Brontomerus.  Here it is again, in case you’ve been in a coma:

As well as being Figure 12 of the paper (Taylor et al. 2011), it’s popped up absolutely everywhere in media coverage: among many others, it was used by the BBC, Guardian, Telegraph and Independent in the UK; by USA Today, Fox News and National Geographic in the USA; by Spiegel in Germany; and by SVT (the state-funded national TV station) in Sweden.  There’s no question that this image contributed hugely to selling the paper to the secular media.  It’s probably responsible for 80% of all the coverage our work got, and I’m confident that it’s going to quickly become one of those images that everyone recognises, like the tyrannosaur/styracosaur fight on the cover of The Dinosaur Heresies and indeed Charles Knight’s classic swamp-bound Brontosaurus.

So it was a huge win for us, and it’s worth looking at how it came about.

Back in 2004, Matt gave a talk at the SVP annual meeting, entitled Skeletal pneumaticity in saurischian dinosaurs and its implications for mass estimates.  The material in this talk became a chapter in the Wilson and Curry Rogers edited volume of sauropod papers (Wedel 2005).  Some time in 2006, Matt heard about the Paleonturology competition, which is all about making palaeontology accessible for teenagers, especially in Spain: anyone who’d had a paper published in 2005 was invited to submit it, and the judges would choose the one that seemed most amenable to being rewritten in a compelling way for non-specialists.  Matt’s 2005 paper won the 2006 competition, and the rewritten version of that paper was translated into Spanish and published in a very nice booklet which I am pleased to have a copy of (Wedel 2007).  In 2007, Matt was invited to the Fundación Dinópolis in Teruel, Spain, to receive his award, launch the booklet, and act as a judge for the 2007 competition.

The relevance of all this now finally becomes apparent: while in Teruel, Matt met Francisco “Paco” Gascó, and looked through some of his portfolio of palaeo-art — including, for example, this rearing Camarasaurus:

So Matt had Paco in mind as a promising palaeo-artist.  Then towards the end of May 2008, when we were readying the Brontomerus paper for submission (not to Acta Palaontologica Polonica — to a different journal, which didn’t take it) I wrote to Matt saying:

I’m attaching a tentative skeletal reconstruction that I did.  […]  Now I’m thinking: should we approach a palaeoartist to see if we can get a life restoration done in time for the launch?  The eponymous thunder thighs hardly make an impact in the skeleton, after all.

And Matt quickly replied suggesting that Paco could be the person to do it.  He sent a few samples, including the Camarasaurus above, and I was sold.   A couple of days later, Matt suggested the idea to Paco, he was up for it, and so we were all systems go.

At that time, Paco (who is now getting towards the back end of his Ph.D) was a humble recent graduate, which was great because it meant that Matt and I got to boss him about — something that he accepted with enormous good grace as we went through a sequence of some 44 images on the way to the pair that we ended up with.

By this time, Matt and I had already realised that we wanted the artwork to show kicking.  So we started out by asking Paco to mock up three very rough sketches of how he thought a Brontomerus-kicking-a-predator scene might be composed.  Here is one of the three — pretty representative:

Although this scenario is pretty sweet, it’s not really what we wanted as it shows Brontomerus kicking backwards like a horse, rather than forwards like a footballer.  (That’s a soccer player, for those of you in the USA.)  So Matt offered this concept sketch:

(It’s well worth clicking through and seeing the details.)

It’s interesting to see how much of the final image was already in place even in that very early sketch: the basic pose of the adult sauropod, the juvenile behind, the theropod getting its arse kicked — even if at this stage it was a juvenile Acrocanthosaurus rather than a mature Utahraptor.

With this reference in the back of his mind, Paco started work on a 3D model of the sauropod that would be the core of the composition.  He was quickly able to show us a first draft that had all the pieces in place to look convincing at least as a generalised sauropod:

Already at this stage, I was pushing for the uniquely Brontomerus-like aspects of the anatomy to be made more apparent, so I sent back this modified and uglified version of the image to give a sense of where we wanted to be heading:

(You might want to open this image and the previous one in two tabs and switch back and forth between them.)

The purpose of the modified version, of course, was to show how high the ilium would sit on the torso and how it brings forward the anterior margin of the leg muscles.  We wanted the thigh to be much more thundrous!  This was part of a merciless campaign of anatomical criticism of many, many aspects of the in-progress restoration — for example, the cross-sectional shape of the neck, which at this point had a crest on top and flat sides.  (You’ll notice that in the final version, the neck has the distinctive subtriangular cross-section that is produced by the ventrolateral excursion of the cervical ribs in sauropods.)

While that stream of refinement was going on, Paca was starting to skin the model.  Here’s the first version we saw that had skin texture:

Seeing this was an exciting moment in the progress of the project.  It was the first time that the artwork started to look like actual art, and the sauropod to look like an actual sauropod.  We knew then that we were on the way to somewhere good.

That feeling intensified the first time Paco showed us the complete cast of our little drama: mother, baby, and evil raptor, all skinned and showing rather fetching stripes which survived in more or less this form through to the final version:

By this stage, most of the anatomical problems are getting ironed out: the flat-sided shape of the neck has gone, replaced by a broader and wrinklier ventral aspect; and the ilium was higher on the torso, with the shape of the dorsal margin more closely reflecting that of the Brontomerus holotype ilium.

Note the care that Paco took with the juvenile: it’s not just a scaled-down copy of the adult, but proportioned subtly differently in a way that reflects what we know of sauropod ontogeny: the limbs grow isometrically, but the neck is positively allometric, so that Baby Bronto’s neck is noticably shorter in proportion.  At this stage, the baby is too big — more than half Mama’s size, whereas the sizes of the elements from the Hotel Mesa quarry suggested that he should be closer to a third of her size.

At the same time that Paco was working on the details of the models, we were all still batting around composition ideas, trying to find the best way to put our three actors together.  This version of Paco’s was similar in concept to Matt’s earlier sketch, but different in a lot of details: the baby is running away rather than sheltering, the theropod is rather bigger than before, and has morphed from an acrocanthosaur to a raptor; and it’s upside-down in an attempt to show that it’s not in control of the situation:

I wasn’t convinced by this version, because the theropod seems to have been spun 180 degrees on the spot as well as kicked upside-down: I felt that he needed to be in a posture that more naturally emerges from having been facing Mama when he was kicked, so I ‘shopped Paco’s sketch into this version:

As well as turning the baby around (something that didn’t really help). I flipped the raptor and tried, clumsily, to convey that Mama had broken its neck.  Of course, that didn’t really work, because the extension at the base of the neck is habitual for most tetrapods anyway, but it at least gave us a sense of the direction we wanted to go in.

OK, so back to the model.  Paco had sent us a simple lateral-view render of Mama alone, as well as the group shown above, so that we could more easily critique its anatomy in isolation.  Here is that simple render, followed by the vandalism I did on it to show changes that we still wanted.  (See what I mean about Paco being patient?)

As you can see  (and as you’ll see more clearly if you flip back and forth between the two images), I was asking for two changes.  The simpler was that I wanted to see the distinctive profile of the Brontomerus scapula showing through the skin.  The more interesting is in the profile of the tail.  It’s been shown in many sauropods that there is a distinctive upwards kink at the base of the tail, so that the dorsal profile of the body does not progress smoothly from hips to tail, and I wanted to see that in Brontomerus.  At the same time, the tail needed to have more flesh on it and the ischium should have been producing a visible bulge in the ventral margin behind the hips.

The next version addressed these points (though the scapula outline was not yet right):

But picky as I am I still wasn’t satisfied…

I made a few changes here — again, in a hacky way using the GIMP, with the result not in any way intended as in improvement in itself, but as a sketch of how the model could be improved.  I shifted the tail up a little, smoothed the dorsal profile so that there was no longer a sort of dip at the base of the tail, and smoothed out the rear margin of the top of the thigh, so that there was no longer a “buttock”, but a hint of caudofemoralis musculature connecting the tail-base with the thigh.

Once Paco had made the necessary changes to the model, the next render looked superb — and very recognisable as the basis of the now-ubiquitous final version:

At this point, work on the main model was essentially complete, and Matt and I were both really happy with the result.  For people who’ve spent as much time gazing at the Brontomerus ilium and scapula as we have, this is very obviously Brontomerus and not just a generic sauropod.  Now it was time to put the model together with the composition ideas we’d been playing with:

We went through several versions of this, mostly varying in the posture of the theropod, but this is the one that led to the final piece.  For the first time, we were all happy with Baby Bronto in this one, too: he’s about the right size, and has a sort of skittering look to him, as though he wants to be elsewhere but doesn’t want to leave Mama.  (Am I anthopomorphising?  Very well; I contain multitudes.)  It’s a bit too close to the adult, though, so we can’t quite see its shape.  This was fixed in the next version, which also contained a backdrop for the first time:

Now we’re really getting somewhere.  You’ll notice that the raptor’s head is bent further back this time, hopefully conveying that its neck is broken.  But because I was really keen on getting it across that the raptor is DOWN and it’s NOT getting up again, I once again vandalised Paco’s work, this time with buckets of blood:

What I wanted to convey was: if this raptor wasn’t already secondarily flightless, it is now.  Still, I admit that the amount of blood, and the vividness of its colour, are a little over the top.  So in the final version, Paco took some of the blood back out, and toned it down to a more realistic colour.  The other important difference is that the raptor was moved a bit closer to the sauropod — not because that’s necessarily a better composition, but because we expected newspapers and other media outlets to crop the image mercilessly, and we wanted to give them best chance of keeping all the key element in frame when they did.

And so we arrive at the final version, as it appeared in the paper:

The very last thing we did was ask Paco for a second render of the same scene, so that media outlets would have a choice of artwork and wouldn’t all need to use the exact same image.  That was doable because all three dinosaurs, with their skin-textures, were built as a 3D model, which can be viewed from any angle.  But producing a finished artwork from this is not trivial: once an angle is chosen and the animals rendered, there is still a lot of post-production work to be done in putting in the background, the blood, the dust and so on.  So we didn’t ask for a complete array of 128 of these — just the one addition.  After reviewing a few candidate exported renders, we settled on one from a more anterolateral perspective, and Paco worked his magic to yield this alternative take:

I hardly have words to tell you how much I love this.  Several times, looking at it, I’ve found myself laughing out loud at how comprehensively the theropod is getting owned.  It’s OVER for that would-be predator.  It’s DONE.  The only question is whether Mama is going to put it out of its misery by stomping it flat, or whether it’ll be left to bleed out.  Either way, it picked on the wrong victim for dinner.

Part of what I love so much about this is that Brontomerus looks like an animal, not like a monster.  It works anatomically, feels like something that lives and breathes … and, indeed, kicks.

Let me close by clearly stating that 99% of all the Awesome here is the work of Paco — a talented and hardworking guy, who made Matt’s vision come to life.  My own input was basically restricted to whining.  I hope we’ll be seeing this image for many years to come, and that plenty more of Paco’s pieces make it out into the wide world where they belong.

For more of Paco’s stuff, please see his blog, El Pakozoico (TARDIS alert!), and his deviantART page.


60 Responses to “Genesis of an instant palaeo-art classic”

  1. Albertonykus Says:

    Really enjoyed this post!

  2. dmaas Says:

    Fantastic. Thanks for the walk-through.

  3. kattato Garu Says:

    Hi Mike, I really enjoyed the post which is a great example of artist and scientist working creatively together. But I must say my first reaction to the picture when I saw it was, whoa, that looks very unlikely. Pretty much every quadrepedal animal which uses kicks as a weapon, does so forwards with the front legs and backwards with the back legs. Wouldn’t the back leg have to swing outwards pretty far for such a rear kick to be an effective habitual defense? Which would be rather destabilising in such a vast animal. Or, would require the whole body to curve rather too much (I think?) as in the ‘top view’ in the pencil sketches. It would be a pretty darn slow and silly theropod that would get tangled underneath a sauropod. I can’t think of a very good reason for the massive protractor muscles either, but I think it seems more likely to have to do either with locomotion over hilly terrain, with nest building (scraping?) or possible some kind of intraspecific contest, than with kicking theropods in the teeth. I’m not saying it’s impossible…

  4. Stu Pond Says:

    Superb post about a superb image. I have to say, I rather liked the vivid blood rather than the toned down version.

  5. dmaas Says:

    May I request an additional post about your press work? The press package you put together goes a long way towards a “best-practices” for outreach pr, love the fact sheet and videos. How much direct journalist contact was there? What were the main multiplicator venues which got copied by others? How would you rate the content that got picked up on and your influence thereof?

  6. dmaas Says:

    oh… and thoughts to Garu’s comments on plausibility… I like the posed defense reaction. If the theropod approached from behind, he’d be no less silly, as there’s a ton of fly-swatter in the form of the tail. And he’s a lone attacker, so I figure he must be desperate anyway.

    I’ve often imagined why those toe claws persisted, and while digging, ground-scratching seems more likely, there’s nothing against using what you;ve got for more than one purpose.

  7. Mike Taylor Says:

    Thanks to all for the kind comments. As Paco, Matt and I were working on the artwork (across several years!) I always had in mind that the process might make a good case-study, and I’m delighted that people are finding it enlightening as well as interesting.

    Garu, we’ve been around the plausibility of the kick a few times, but to summarise: yes, it’s possible that the enlarged protractors were doing something else and we acknowledge that in the paper; we didn’t feel we could ask Paco to do three or four pieces illustrating all the possibilities we came up with (though it would have been awesome if we could!). The forward kick is plausible in part because the enlarged preactetabular blade anchors abductor muscles (which draw the leg laterally away from the midline of the body) as well as the protractors that draw it forward. So we envisage the kick being directed somewhat anterolaterally rather than directly forward. This isn’t really apparent in the 2D rendering, but if you look at the animation at you can get a sense of it.

    David, we might do a post about the press work, but the problem with such posts is that they tend to come across as “Oh, look at us, we’re so awesome”. And since we’ve caught some flak before for overhyping our work, we want to tread reasonably carefully.

    On the use of sauropod toe-claws for more than one purpose: absolutely — one of the things we tried to get across in some of the Brontomerus publicity is the importance of exaptation, that features which evolve for one purpose so often end up being used for another. Unfortunately, and perhaps predictably, that part of the message often got edited away, but it does survive in places — for example, in the last four paragraphs of this article in the Telegraph.

  8. Mathew Wedel Says:

    May I request an additional post about your press work? The press package you put together goes a long way towards a “best-practices” for outreach pr, love the fact sheet and videos.

    Thanks for the kind words. There’s not a ton to say, but three things come to mind. The first is that we always make it as easy as we can for people who want to find the paper and extras to do so. This is pretty easy and it can be free to the authors, if you use a free blog from WordPress, Blogger, etc., to host everything (as Mark Witton and Darren Naish did for their azhdarchid paleobiology paper). For our joint papers, we put everything on Mike’s website. For my standalones, the extras go here on SV-POW!

    The second is that even doing all that, there is simply no predicting whether a given story will take off or not. Our neck posture paper from 2009 got loads of media coverage in Europe and almost none in the States, and that was with an identical joint press release covering both areas. Don’t get me wrong, we’re gratified when our work gets popular attention, but it’s not something one should get hung up on. In the end, we want to a paper we can be proud of, and popular coverage is gravy.

    The third is that having some actual content in the press release helps. I say that because even though we had a bunch of additional information online, the initial wave of publicity drew almost exclusively from the press releases (we did two, one for Europe through UCL and a joint OMNH/WesternU release for North America). Clare Ryan at UCL, Linda Coldwell at OMNH, and Jeff Keating at WesternU all worked very hard on the press release with us; we batted the releases back and forth with them almost as many times as we swapped suggestions with Paco for the artwork. In other words, we didn’t try to write it ourselves, nor did we leave it to someone else; we had the good fortune to have smart, capable collaborators, and we worked with them instead of for them or over them.

    To answer your other direct questions:

    How much direct journalist contact was there?

    It varied, a lot. Mike and I both went out of our way to be available to journalists for the first 2-3 days after the embargo lifted. That resulted in some cool coverage, but we also saw some great pieces that we had no direct involvement with.

    What were the main multiplicator venues which got copied by others?

    The BBC was the One Outlet to Rule Them All, I think. Their online story got over 100,000 hits on the first day, and a lot of subsequent coverage seemed to point back that way. UCL’s video on YouTube was also quite popular.

    How would you rate the content that got picked up on and your influence thereof?

    I’m sorry, I’m not exactly sure if this is what you mean, but most of what we saw in the media came directly from us in one way or another. There was one glaring exception that we will blog about shortly (hint: it’s the one that opened with Brontomerus “terrorizing its pray [sic]”), but in general we were quite happy with how things played out.

  9. BJN Says:

    I think the stiff-legged kick pose and flying raptor image is dramatic but really incredible (in the not in least bit credible sense of the word). To land that kind of blow, the sauropod would have had to do a big backswing of that long leg pendulum. Is it remotely credible that a lone, agile raptor would fail to notice that motion?

    I’ve seen many illustrations that impress the media but that are just plain silly when you look at them objectively. This is one.

  10. Mike Taylor Says:

    Well, giraffes successfully kick lions — presumably the lions don’t see it coming.

  11. Zach Miller Says:

    Conceptually, I love the picture. Take that, ya stupid raptor! And I have absolutely no problem believing that a big sauropod could make a quick swing like that with its leg. Big modern animals move shockingly quickly all the time.

    I do have trouble with the idea of showing a life restoration of an animal known from so few remains. I understand it’s a camarasaur, and it’s probably not going to look too different from Camarasaurus, but it’s just a pet peeve of mine. I had the same complaint when everybody was drawing Masiakasaurs’ head based on a very wierd dentary.

    But. This was a wonderful post. I love reading about the collaboration between science and art, and the beautiful results of that union.

  12. Matt Wedel Says:

    I understand it’s a camarasaur, and it’s probably not going to look too different from Camarasaurus

    Whoa! Brontomerus is not a camarasaur. It’s a camarasauromorph, meaning that it’s somewhere in the big clade containing Camarasaurus, brachiosaurs, and titanosaurs. We described it as a camarasauromorph because that was the most inclusive clade that captured all of its positions in the array of most parsimonious trees thrown up by the phylogenetic analysis. But missing data plays a big part in that. The overall gestalt of the bones suggests that it’s a somphospondyl (in the more exclusive clade that includes Euhelopus and titanosaurs but not camarasaurs or brachiosaurs), and the 50% majority rule tree backs that up.

    Admittedly, this only strengthens your point, which is that we don’t know what the animal looked like. IMHO, Brian Switek said it best: “I certainly can’t blame the authors for coming up with a hypothetical restoration rather than showing a juvenile ilium floating in the air.”

  13. Zach Miller Says:

    Bah! Camarasauromorph! I’ve got to get my sauropod taxonomy down. One of my big blind spots, you know. I totally agree with Brian on this point. :-)

  14. Mike Taylor Says:

    This leaves me wondering whether we would have been wiser to describe Brontomerus as a Macronarian. It’s less precise, but less susceptible to misunderstanding.

  15. Mickey Rowe Says:

    Great post, Mike! I just wonder if you’ve ever considered seeing a therapist to help you get past all of the pent up anger that you have for theropods.

  16. Mike Taylor Says:

    Anger, Mickey? No, no, you quite mistake me. Theropoda is absolutely my second favourite top-level eusaurischian clade!

  17. Andy Says:

    I am particularly interested in Matt’s comment above, related to the press release. At risk of going off-topic (hmm, maybe this is a good post for my own blog), what factors do you consider in deciding whether or not to run with a press release? I know that you guys typically do blog roll-outs for your papers (as most of us in the blogging world do – and should do), but was there anything particular about Brontomerus that made you sit back and say, “Aha, we must do a news release!”?

  18. Mike Taylor Says:


    I sense an impending post about the whole publicity gig … even though we’d not planned to do one. I won’t tell the whole story now (although Matt has told much of it already), but in a nutshell my attitude is: if you’ve spent a couple of years working on a paper, it’s worth investing another couple of weeks to let the world know about it.

  19. Andy Says:

    Cool – I look forward to any impending posts! This is an issue I’ve been giving a lot of thought to lately in relation to some of my own work. . .what papers are worth publicizing? What papers are best left to “just” the blogosphere? Are we as paleontologists putting out too many press releases, or not enough?

  20. Nathan Myers Says:

    I wonder whether the nerve-signal transit time between eye/head and the leg muscles would allow for any sort of aim. I’ve always thought of sauropod hindquarters as working like the back end of a tractor-trailer rig (“articulated lorry”, for you Brits) that follows along more or less automatically. The brain’s role might be analogous to that of JPL directing Spirit across the Martian tundra.

    So, mark me down in favor of it using its big leg muscles to dig up tubers.

  21. Michael Richmond Says:

    I have enjoyed your description of the genesis of this picture, and I have a pleasantly visceral reaction to the parent defending its child, but I share BJN’s concern: the motions stretch credulity.

    I teach physics classes and try to give my students simple problems like the one this picture presents: how fast must the theropod be moving as it flies through the air? Where will it land? I used the chest height of the adult sauropod to set the scale: 2 meters vertically corresponds to 79 pixels in the final image. I assumed that the theropod left the ground at the current location of the saropod’s left hind foot, and that the theropod’s original center of mass was at the height of the toes on that foot. I drew an admittedly arbitrary trajectory for the theropod which reaches a peak slightly to the right of its current position; that is, it assumes that the theropod is just starting to fall down at the moment of the picture.

    I’ll spare you the details, but the end result was the the original velocity the theropod must have had as it started its unconscious motion through the air was about 7.7 meters per second, and the inert body will strike the ground about 7 meters from the point of impact, just outside the picture frame to the left.

    I find it hard to believe that a creature the size of the theropod — its mass is what, 400 kg or so? — could be thrown a distance of 7 meters by a kick. The leg muscles of a sauropod must be immensely strong, but I don’t think they are designed for the very quick motions which would be necessary to toss the theropod into the air so violently.

    Giraffes have long, thin legs which can impart high speeds to their feet as they kick. Elephants and sauropods do not, to my (small) understanding. If there are examples of elephants kicking substantial items long distances, I’d be happy to learn of them and be corrected.

    So, bottom line, this picture reminds me a bit of a Hollywood action movie in which the hero doesn’t just punch the villain so that he falls backwards onto the floor, but punches him so hard that he rises into the air, flies backwards into a window, crashes through the window, and flies gracefully 10 feet into the shrubbery below. It’s satisfying to see, but not really physically plausible.

  22. Mike Taylor Says:

    Michael, your approach is interesting — we should run the numbers properly. Whether the theropod can be thrown 7 meters is surely not the question, so much as whether a speed of 7.7 m/s can be imparted to it. Intuitively, this doesn’t seem unreasonable to me: it’s not at all unusual for footballers to kick the ball faster than 30 m/s, and we can easily imaging the foot of Brontomerus moving equally fast. Of course, it would not in kicking have imparted all of the leg’s momentum; but given that I would guess that the lower leg of Brontomerus massed about the same as the whole theropod, then imparting one quarter of the momentum should be enough to give us the 7.7 m/s we need.

    So the appropriate analogy would not be Sarah-Michelle Gellar punching David Boreanaz across the room, but David Beckham kicking a smallish dog into the goal from a range of maybe six yards. I’m pretty confident I could do that (but I won’t be seeking grant funding for confirmatory experiments).

    Finally: of course you are right that elephants don’t kick. But Brontomerus was not like an elephant — that really is our whole point! The osteological evidence suggests a much more athletic animal: have you looked at the pathetic ilia of an elephant recently? :-) (I foresee a new article … “Your elephant’s ilia are minuscule”)

  23. Gareth Dyke Says:

    I am intruiged guys: What was the journal that didn’t want the paper the first time around? Memoirs of the Utah Geological Survey … ? ( ;)

  24. Mike Taylor Says:

    Hi, Gareth. The journal in question was JVP. Our paper ran aground of a philosophical difference: one of the reviewers just didn’t feel it was acceptable to name a new taxon based on fragmentary remains, whereas we felt and still feel that diagnosability is much more important than mere volume of bone. The reviewer would have been happy enough with the paper if we’d withdrawn the new name, but my perspective is that not naming something that is clearly distinct and diagnosable is actively misleading.

    We could have argued it through with the editor, I guess. But instead — partly because I’d had another bad experience with JVP shortly before that — our feeling was that we might as well go to a different journal with a fresh slate. Also, in the time the manuscript had been in review, my feelings about open access had hardened, and both Matt and I were increasingly reluctant to give our work to a journal that would lock it up and hand the keys to Taylor and Francis.

    In the end, as frustrating as it was at the time, I can look back now and be glad that we got that bad review at JVP — yes, it held the paper up for the best part of a year, but it means that the work is freely available to the world, and that the figures are reproduced in colour. Also, the Acta Pal. Pol. guys are just good to work with — for example, they fixed a precise online publication date well ahead of time, so that we could make the publicity arrangements.

  25. Matt Wedel Says:

    Nathan Myers wrote:

    I wonder whether the nerve-signal transit time between eye/head and the leg muscles would allow for any sort of aim. I’ve always thought of sauropod hindquarters as working like the back end of a tractor-trailer rig (“articulated lorry”, for you Brits) that follows along more or less automatically. The brain’s role might be analogous to that of JPL directing Spirit across the Martian tundra.

    Nerve conduction velocities vary widely, but the impulses controlling voluntary muscle are typically very fast, from 50 to over 100 m/s. Your suggestion about the brain’s role is sneaking up on the truth. For most day-to-day activities like walking, central pattern generators in the spinal cord are sufficient to keep the body moving. The role of the brain is to provide balance and coordination and to ‘step in’ when needed–like when you start placing your feet carefully as you cross rough ground. Or indeed when you kick something. The point in Paco’s life restoration is that ‘Momma’ is not going to much trouble here, just lengthening her normal step cycle to give an insignificant theropod the terminal boot.

    Michael Richmond wrote:

    I find it hard to believe that a creature the size of the theropod — its mass is what, 400 kg or so? — could be thrown a distance of 7 meters by a kick.

    Interesting analysis. We can probably improve it by using more realistic numbers. If the Utahraptor shown here massed 400 kg (882 lbs in 1 g, or the size of a grizzly bear), I’ll eat my hat. There are reported specimens that big, but this one is roughly human-sized, and had air sacs and a pneumatic skeleton. I would guess that 150-200 kg would be closer to the mark, for this individual (hopefully if I’m way off, someone who knows more about theropods will chime in).

    I did a quick-n-dirty GDI on the Camarasaurus/Brontomerus from our skeletal inventory, just to have something to tell the media. I got a volume of 234 L for the hindlimb from the knee down (I didn’t do the thigh separately from the body). Given that thighs are quite a bit bulkier than calves, it doesn’t seem unreasonable that the entire moving mass of the hindlimb was something like 750 kg.

    Assuming perfect conservation of momentum, a 750 kg hindlimb would only need to swing at 2 m/s to impart a velocity of 7.7 m/s to a 200 kg, which seems positively leisurely compared to some motions I’ve seen from equally large elephants.

    Further thoughts?

  26. Mike Taylor Says:

    Dude, that can’t be right. However much momentum it has, an object moving at speed v can’t impart a speed greater than v to something it hits. In the case where a heavy thing hits a light thing, the way momentum is conserved is that the heavy thing retains most of its speed (and hence, of course, momentum), giving up only as much as is needed to bring the light thing up to the same speed as itself.

  27. Matt Wedel Says:

    Crap. See, this is why I shouldn’t do math in public.

    Okay, you’re the maths guy. If the stupid theropod only masses half of 400 kg, surely it doesn’t still have to be flying at the same velocity. So what velocity is it flying at, and is that velocity attainable by a swinging sauropod hindlimb?

  28. Mike Taylor Says:

    My understanding (and I hope Michael Richmond will correct me if I’m mistaken) is that in order for Brontomerus to kick the raptor at 7.7 m/s, two things need to happen. First, the leg itself needs to be moving that fast (actually a little faster); and secondly, it needs to be heavy enough that, even with imperfect transmission of momentum, it can pass enough of its momentum to the raptor to get it moving that fast. The first half of this is not a problem: if footballers regularly exceed foot velocities of 30 m/s, then Brontomerus can certainly do a third of that. The second half, based on your mass estimates of 234 kg for the lower limb and 200 kg for the raptor, suggests it might be possible, but does require pretty efficient transmission. But how much of the upper leg’s momentum can also be transferred? I don’t know — I guess it’s to do with the stiffness of the knee.

    Where’s John Hutchinson when we need him? :-)

  29. kattato Garu Says:

    “However much momentum it has, an object moving at speed v can’t impart a speed greater than v to something it hits.” Sure it can. That’s what conservation of momentum is all about. A heavier object striking a lighter object in an elastic collision imparts a much higher velocity to the lighter object. I doubt a theropod would have been as rigid as a billiard ball but it certainly wasn’t a bag of mush so a big slow *heavy* swinging club of sauropod foot could…in theory… send it flying. I’ll let cleverer men do the maths on the trajectory of the theropod…

  30. Mike Taylor Says:

    Garu, that can’t be right. Imagine two perfectly rigid billiard balls — one hollow, the other solid, so that the solid one weighs ten times as much as the hollow one. Let the solid ball roll towards the hollow one at 1 m/s. When they strike, the hollow ball does not fly off at 10 m/s.

    The Wikipedia article on momentum is more complex than I care to invest in at the moment, but its equations of momentum exchange in one dimension do — contra my earlier assertion — allow for a light object to move off more quickly than a heavy object striking it. But only twice the speed (and that only when the light object’s mass is negligible).

  31. kattato Garu Says:

    Hi, you’re putting words in my mouth – I said “a much higher velocity”… e.g. for a heavy object travelling at v1=1 striking a stationary light object (v2=0), the highest possible velocity for the light object tends towards 2 (mass of the light object is negligeable comared to heavier object). If the 2 are travelling towards each other at the same velocity (+1 and -1), the lighter object speeds off at 3 times the velocity of the heavier one (that’s the best v1/v2 ratio you can get). Assuming perfect conservation of kinetic energy. I don’t know how Matt came up with his figure but you were right to raise your eyebrows – a 750 kg leg would need to strike a perfectly rigid theropod at ca 4.9 m.s-1 to send it flying at 7.7 m.s-1.
    Hmm. have you got any elephant foot velocities handy? :)

  32. Matt Wedel Says:

    I don’t know how Matt came up with his figure but you were right to raise your eyebrows – a 750 kg leg would need to strike a perfectly rigid theropod at ca 4.9 m.s-1 to send it flying at 7.7 m.s-1.

    I’m sorry, I don’t understand what is eyebrow-raising about that. Are you saying that it’s improbable or very probable? Also, I thought I had made my (possibly faulty) chain of inference clear. If not, please let me know what part wasn’t. I’m really interested in solving this now.

    Hmm. have you got any elephant foot velocities handy? :)

    Well, elephants can amble at up to 40 kph, which is 11 m/s. I’m not sure what leg velocity that comes out to, but the legs have to swing faster than the whole-animal velocity during each stride. For an elephant going flat-out at 11 m/s, it does not seem unreasonable that the leg might be able to swing at 22 m/s or more. And Momma Bronto had better mechanical advantage for swinging her hindlimb forward.

    Of course the theropod is not a perfect solid, nor will conservation of momentum be perfect. But it doesn’t have to be–Momma’s foot might have met the theropod when her leg was vertical, which would give her leg plenty of time and space to accelerate the theropod to what I’m going to call fail-to-escape velocity.

    Where’s John Hutchinson when you need him?

  33. Mike Taylor Says:

    “Hmm. have you got any elephant foot velocities handy? :)”

    Yep. John Hutchinson (who I keep hoping will chime in on this subject!) had a very approachable two-page paper in Nature in 2003 entitled “Are fast-moving elephants really running?” which contains some useful numbers. (You can download a copy from John’s publication list). John and his co-author showed from their own videos that the Indian elephant Elephas maximas “can move at surprisingly high speeds of up to 6.8 m/s (25 km/h)”, and “they used duty factors as low as 0.37”. That means that their feet were stationary on the ground, in the power stroke, for 0.37 of the time, and moving forward, in the recovery stroke, for the other 0.63 of the time. So the average speed of the foot in the recovery stroke is 6.8/0.63 = 10.79 m/s — more than twice the speed you say would be sufficient to kick the theropod at 7.7 m/s.

    Now bear in mind that (A) this is the average speed throughout the whole of the recovery stroke, which includes the time needed to accelerate the leg at the start of recovery and decelerate it to prepare for the power stroke, so that the maximum speed was probably significantly higher than that; (B) the ilial area for attaching protractors in sauropods generally is much larger than in elephants; and (C) the area is much greater again in Brontomerus. Putting all those things together, I don’t think it would be at all unreasonable to imagine the foot moving at 20 m/s or more.

    So the more I think about this, the more I think that the life restoration is actually perfectly credible.

  34. Michael Richmond Says:

    This conversation is unfortunately being oversimplified to the point that it isn’t very useful, but for perfectly understandable reasons.

    The main problem is that, while we _can_ use freshman physics to compute the distance that a limp corpse will travel through the air (hence the estimate of 7.7 m/s for initial velocity is pretty secure, modulo the small uncertainties in estimating scale from the picture), we _can’t_ use freshman physics to make accurate estimates of the interaction of two large, extended pieces o’ flesh. The big problem is time. If the leg met the theropod for only a teeny-tiny fraction of a second — say, the same interval during which two billiard balls remain in contact on a pool table — then we might try to apply the conservation of momentum in the simple manner shown in previous comments. Alas, the simple equation one might find in a textbook (you’ll have to imagine the vector signs over the velocities)

    m1*v1_i + m2*v2_i = m1*v1_f + m2*v2_f

    isn’t quite right; it ignores the influence of external forces on the two bodies during the period of contact. “External forces” here means any forces which might be pushing either the leg or the theropod’s body while they are slamming into each other.

    Now, in this case, there’s a very significant external force: the muscles in the sauropod’s leg will be pulling the leg forward during the time that the leg touches the theropod. In fact, I suspect that the contact will last for a long enough time that anyone watching it in real time would perceive the interaction to be more of an extended push than an instantaneous rebound — more of two large American football players colliding than a ping-pong ball bouncing off the table. One of the reasons the contact would last so long is the sheer size of the theropod: as the foot of our heroine makes contact with, say, the gut of the nasty fellow, it will push the gut backwards; but, at the same time, the head and limbs of the nasty fellow will for a short time continue to move forward due to their inertia. It’s only when the flesh and tendons and bones which are being move backward finally pull on the legs and heads and tail that the rest of the theropod will slow down and reverse its motion.

    During this entire time, the sauropod’s muscles will be exerting forces on the leg, continuing to accelerate it forward.

    Sorry — that was a bit long-winded. The important point is that one must account for the influence of those forces exerted by the muscles to figure out the final velocity of the dead theropod, so it’s a really, really complicated problem.

    One way to approach it would be to construct a three-dimensional finite-element model of the theropod and of the sauropod’s leg, and run them into each other while moving the leg under the influence of some assumed muscular force. Ugh.

    It might be more reasonable to look for footage of real examples of contact between extent large creatures — moose and wolves or lions and rhinos or elephants, for example.

    Or, if there are any volunteers, someone could run up to an elephant and start poking it in the leg with a sharp pointy stick. :-)

    I hope no one will take me to task if point out that this conversation can work only if everyone works together. It’s clear that some of us know a whole lot about skeletons, and some of us know know a whole lot about animal behavior, and some of us know a whole lot about physics … but (apparently) none of us knows everything about everything.

  35. kattato Garu Says:

    This reminds me of the physics joke that starts “Let us consider the cow as a perfect sphere of radius r”… Matt I wasn’t calling your chain of inference into account – if anyone knows the mass of a sauropod’s leg, it must be you. But Mike was right to call you out on the calculus i.e. it’s impossible for an object travelling at 2 m/s to impart a velocity of nearly 8m/s to another object (even a perfectly rigid, spherical theropod – Fifasaurus?) However, this was clearly an intuitive conclusion (hey who remembers all that boring high school physics anyway). On the other hand, I’m quite impressed with the elephant leg speeds if they’re correct. That’s some fast moving herbivore flesh there. Maybe Brontomerus could also score a goal from 60 yards out…

  36. Matt Wedel Says:

    kattato Garu:

    Okay, cool. I wasn’t saying that I hadn’t screwed up, I just wanted to make sure that it was clear where I had screwed up! :-)

    Michael Richmond:

    I hope no one will take me to task if point out that this conversation can work only if everyone works together. It’s clear that some of us know a whole lot about skeletons, and some of us know know a whole lot about animal behavior, and some of us know a whole lot about physics … but (apparently) none of us knows everything about everything.

    Not only will you not be taken to task, if you’re not careful you might find yourself press-ganged into writing a paper. You’re absolutely right, this is all interesting stuff and I suspect that just about everyone at the table wishes he knew as much as everyone else. Certainly the case for me.

    I echo your “ugh” on the FEA solution. I have a feeling we’d just be piling up assumptions until the result was not even wrong. I still have the hope, possibly forlorn, that this might yield to plain old brutal arithmetic, at least to the low level of confidence I’d be happy with (and would find most honest given all of the unknowns).

  37. Nathan Myers Says:

    Are you going to advance the hypothesis that the driver for overmusculature in B. is increasing capacity to loft raptors?

    I’m wondering about plausible drivers. Digging is one. What are the others?

  38. Where am I when I need me? I do wonder that sometimes. Usually just goofing off, that’s where. Most recently, in the near-Arctic of Boden, Sweden where I enjoyed a 100% sauropod-free holiday.

    I confess I don’t know about the 6000kg kicks 200kg 7m thing. It’s messy. Leg segments/muscles can transmit energies in funny ways that a “let’s treat the sauropod leg [and the raptor body] as a single rigid segment” won’t capture. It’s a very dynamic activity and so would be computationally non-trivial, so I don’t find back-of-the-envelope calculations much more than qualitatively heuristic (at best). I don’t know that it’s such a pressing research question that it deserves that level of detail though. Perhaps a qualitative approach would be sufficient.

    More importantly to me, we don’t know anything about how living animals kick, in a biomechanical sense; there are essentially no studies of this behavior in non-humans (but a lot of kinesiology studies of humans, esp. for sport biomechanics). So to me it seems to be very wild speculation to get quantitative about sauropod Jackie Chan capacities.

    That being said, I don’t doubt that at least one time in the past, a sauropod did manage to kick a theropod and said theropod was none too happy about the experience as a result.

    I also can’t think of many legged tetrapods that won’t/can’t kick. If they can, why wouldn’t they when it was useful to them? And if they can walk (i.e. bend their legs), they probably can muster a kick.

    I’ve been almost kicked in the head by far too many elephants (yes they certainly do kick!! especially when tickled), rhinos, etc. Certainly my (currently) faster reflexes/nerve scaling constraints have helped a lot there, much as they would for smaller theropods.

    I’ve had the “biomechanics of animal kicking” question come to me a lot in the last ~5 years. Maybe someone (ideally more invulnerable than me) needs to study it experimentally to understand the basic “rules” underlying kicking. There is a groundswell of interest in the subject of the biomechanics of aggression (e.g. Dr David Carrier, now on sabbatical at the RVC).

    I think that pretty much exhausts what I can say on the topic. Sorry.

    TL;DR: maybe.

  39. Mike Taylor Says:

    Thanks, John. Shame that we can’t make more progress with this, but if the work hasn’t been done on extant animals then this is necessarily all guesswork.

    Nathan, in many interviews we’ve said that we thought the most likely evolutionary driver for the enlarged protractors of Brontomerus was competition for mates — sexual selection via combat. But of course once such a feature has evolved it will inevitable be exapted for use in other scenarios … such as RAPTOR ATTACK!!!1!

  40. […] to say, one of the things I love most about Paco’s Brontomerus artwork is that it’s a rare and welcome example of the much neglected Sauropods Stomping Theropods […]

  41. Warren B. Says:

    I don’t have anything grand or revelatory to contribute (in fact, the maths in this comments section makes me want to curl into the foetal position), but on Nathan’s comment:

    “I wonder whether the nerve-signal transit time between eye/head and the leg muscles would allow for any sort of aim.”

    I don’t think most sensible theropods would want to risk strenuous tests of the hypothesis.

  42. […] 先々週話題になった「キックが得意な恐竜」ですが、論文著者らのブログで、あの印象的な復元イラストの制作過程が紹介されています。 Genesis of an instant palaeo-art classic […]

  43. Nathan Myers Says:

    Of the theropods I have experience with, chickens (er) seem pretty easy to kick.

    But a half-second round trip doesn’t seem to lend itself to fine control. B. can only loft you if it can predict where you’ll be a second later, and kick there. A second can be a very long time. If you only need to kick the side of a barn, OK. (Comparisons of lovelorn sauropods to barns are left as an exercise for the reader.)

  44. Matt Wedel Says:

    But a half-second round trip doesn’t seem to lend itself to fine control. B. can only loft you if it can predict where you’ll be a second later, and kick there.

    Whatchoo talkin’ ’bout, Willis? Using the skeletal atlas, I estimate a brain-hip distance of maybe 6 meters in Momma. At a leisurely 60 m/s nerve conduction velocity–like the impulses running down my arms as I type this–the hip muscles would start to fire 1/10 of a second after the signal left the brain. If sauropod nerves were selected for fast conduction, the conduction velocity might have been more like 120 m/s, which would cut the transit time down to 1/20 of a second.

    There’s another way that slow sauropods could have kicked faster theropods. It’s the reverse of Full Metal Jacket:

    Q: How can you kick defenseless theropods?
    A: Easy, you just lead ’em a little more. Ain’t the Cretaceous hell?

  45. Nathan Myers Says:

    I won’t argue with your dimensions, but I do insist that the operative number is at least 2 round-trip periods, unless that eNORmous brain is astonishingly good at matrix arithmetic (B.‘s, not yours; I know yours is) and spatial transformations. The rest of us make do with linear approximations and feedback, but feedback requires round-trips.

  46. kattato Garu Says:

    “One way to approach it would be to construct a three-dimensional finite-element model of the theropod and of the sauropod’s leg, and run them into each other while moving the leg under the influence of some assumed muscular force.” It would be much easier (and more fun) to rig up a pig’s carcass and a singing ‘brontoleg’ mockup. It’s the type of experiment that some genius at Animal Planet will do with great panache. I can see a whole suite of messy dino-experiments: how fast does a triceratops have to run to skewer a T.rex? Could an ankylosaur’s club smash an American Footballer’s head in, helmet and all? Could Allosaurus have bitten Pamela Anderson in half? Empirical evidence is urgently needed for all of these questions.

  47. […] Vertebra Picture of the Week: Needless to say, one of the things I love most about Paco’s Brontomerus artwork is that it’s a rare and welcome example of the much neglected Sauropods Stomping Theropods school […]

  48. […] art is by Mauricio Antόn and we’re super happy with it; as before we had the opportunity to go back and forth a lot and arrive at a finished piece that shows […]

  49. BJ Nicholls Says:

    Here’s a video of an elephant kicking a soccer ball, FWIW.

    It appears to me that the elephant prefers to kick with its front legs, likely because they’re more accurate – being closer to the field of vision. The “reconstruction” of Brontomerus is clearly to provide some entertaining application for the large thigh muscle. And unlike an elephant, whatever head I can imagine for Brontomerus doesn’t have much stereo vision to guide an accurate kick.

  50. Mike Taylor Says:

    Wow, that elephant is good!

  51. […] we really wanted the artist who’d done the Brontomerus work to be credited: Paco Gasco did a fantastic job, and deserved to be known for […]

  52. […] to exhibit some more Brontomerus artwork.  Once more, as with National Geographic and indeed the original life restoration in the paper, Matt and I had the opportunity to work with the artist, feeding back on an initial […]

  53. […] That’s what the preacetabular blade of the ilium is for: to anchor thigh muscles (discussed here, and also nicely illustrated here). Unless the animal had some kind of wasting disease, there was […]

  54. […] the fourth or fifth time ever.  For yet another, I met honorary SV-POW!er Ranger Vanessa Graff and Brontomerus artist Francisco “Paco” Gasco for the first time.  And it’s always good to spend time with people like biomechanics wizard […]

  55. […] is, btw, the guy who made the cool Brontomerus illustration for Mike, Matt & Rich. Great […]

  56. […] are always thrilled when we have the opportunity to show how awesome paleoart came into being (like this and this), and this case is no exception. Best now if I just get out of the way, so — over to […]

  57. […] care about. I am fairly certain that, just as Brontomerus coasted to global fame on the strength of Paco Gasco’s dynamite life restoration, whatever attention Aquilops gets will be due in large part to Brian Engh’s detailed and […]

  58. […] between paleontologists and artists when it comes to reconstructing and restoring extinct animals (example 1, example 2). Until now, I’ve always been the guy making suggestions about the art, and asking […]

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 )

Connecting to %s

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

%d bloggers like this: