Thursday, June 18, 2009

Dr. Mark Witton weighs in on Pterosaur wings

It seems our recsent discussions on Pterosaur reconstructions, in particular wing attachments, has attracted the attention of at least one expert, Dr. Mark Witton. (To catch up on these discussions, they can are in the comment sections of these posts here and here).

For those of you who have never heard of Mark Witton, he is an English Palaeontologist who specializes in Pterosaurs. He has done a lot of collaberative work with internet biology/palaeontology guru Darren Naish, but also made several discoveries on his own.

Of more interest given our palaeo-art spin here at ART Evolved, Dr. Witton is an active (and very talented, I might add) palaeo artist. You can see his work here on his flickr page (including a Pterosaur gallery). The lovely picture above being a more outlandish example of his work (I choose it as I'm under the impression the young man bending over to retrieve his hat is a self portrait of Dr. Witton).

Given his expertise in both the scientific field of Pterosaurs and also in how to reconstruct them, he wanted to add his opinion to the discussion. As he lacked a google account Dr. Witton emailed our member Zach, who himself was unable to post the email. So I have been entrusted with this task.

So without further ado Dr. Mark Witton's comment:

Hi Zach,

I was just checking out some bits online and came across the ART Evolved website. Seeing that you're doing pterosaurs next, I checked out what was being said and noticed that there were some discussions about membrane attachments. Seeing as the concensus between your commenters was that there was no concensus, I think you need to reconsider your ideas.

There are at least 3 reconstructions that are regularly discussed, but only one is worth any salt. For this reason, all my papers (at least, unless I've totally forgotten something) say that the ankle-based attachment is best supported.

Look, here's a quote from Witton and Naish (2008):

"...evidence from anurognathids, campylognathoidids, rhamphorhynchids,> ctenochasmatoids and non-azhdarchid azhdarchoids [86]–[91] indicates that ankle-attached wing configurations are more accurate."

In a nutshell: there is _no_ support for a supports a hip attachment, one specimen may show a knee attachment (but it's ambiguous at best), whereas specimens of Eudimorphodon, Anurognathus, Jeholopterus, Rhamphorhynchus, Sordes, Beipopterus and a tapejarid (at least, there may be some I've forgotten) all give either hints of an ankle attachment or show it quite convincingly.

Hence, if you draw any other type of brachiopatagial attachment on your pterosaurs, you're ignoring the wealth of evidence in it's favour.

Darren Naish gave more detail on this a while back, and not much has changed since then: The question of membrane shape has is different, nowadays: did they taper in close to the body before hitting the ankle, or extend in a broader fashion?

That's a considerably harder question to answer. I did try to put this on the blog, but I couldn't post it as I don't have a Google account. In any case, it may be an idea to pass it onto your chums so everyone has the same data. Anyway, I'm looking forward to seeing the results, and I might have something new for my own site up soon. Ish.



I apologize to both Zach and Dr. Witton if I garbled the formatting here a bit. The version I received was a HUGE mess, due to gmail replacing spaces and enters with "<"s.


Zach said...

Thanks, Craig. Looks great over here.

Peter Bond said...

Awesome guys. Thanks so much to Dr. Witton (and Zach and Craig) for coming to our aid and giving evidence towards an ankle-based membrane attachment. I look forward to reading Darren Naish's article for more detail.

Only two weeks until the Pterosaur Gallery opens! Keep working on your submissions and send them in to: Can't wait to see them all!

Nima said...
This comment has been removed by the author.
Nima said...
This comment has been removed by the author.
Nima said...

Hey thanks Craig. It's good we finally got input from a PhD scientist. Even if I don't quite agree with his view...

I definitely admire Dr. Witton's talent as an artist. There are very few PhD's who are also good artists.

But (and I say this with all respect for Witton and his research) to say that only the ankle-attachment thesis is "worth any salt" seems a MASSIVE over-exaggeration. The main problem with pterosaur wings is that pterosaur fossils themselves are often badly squashed in positions where the wings are not in a lifelike position.

I'll clarify what I mean by this: Jeholopterus is a prime example of very ambiguous evidence.

Here is a photo of the type specimen:

The one wing that IS slightly open, gives no clue to where the membrane attached. It just fades and mixes with seemingly unrelated objects. If we were to interpret every brown stain around the bones as a skin membrane, then there is a HUGE wart-like lump on the shin that makes no anatomical sense (it's probably an impression of displaced flesh). This fossil in fact shows a very narrow wing membrane, and a bunch of fuzzy stuff on the ankles which merely LOOKS like it's attached to the wing due to the knee being folded up.

AT MOST we can say it has tufts of fuzz on its ankles and individual leg membranes. I have not seen most of the other genera specimens, but I would assume they follow a similar pattern. Of course smaller rhamphorhynchoids like Sordes MAY have had bat wings but only as an adaptation to an arboreal lifestyle. I highly doubt such a thing could work for tapejarids and azdarchids.

So far every line of "evidence" for bat wings that I have ever seen, can EASILY be interpreted other ways, and if we take every dark stain as a wing membrane, that's not really reliable because there are other "membrane stains" not related to the wings that don't make any sense if you take them all to be genuine structures of the live animal (as people like David Peters do). Now I'm sure such speculation isn't even close to Witton's professional methods, but it is just about the only way one can state with absolute certainty that Jeholopterus had ankle-attached wings.

With other pterosaurs (especially larger ones) the same sorts of caveats apply.

When corpses get crushed, some of the tissue gets squashed in weird positions where it's hard to distinguish a genuine wing membrane from a piece of displaced skin. I'm skeptical of ankle attachment for this very reason, and because it is physiologically maladaptive for most species.

(Ironically, I quoted Witton and Naish in an earlier comment somewhere, where they said that there is no absolute consensus on the wing membrane structure - so it seems very strange to me that Witton now says only one view is "worth any salt".)

Of course I am not a PhD, and as an artist I'm all in favor of taking into account the latest research (and most of Witton's is actually quite nice), but I am not in favor of accepting any one theory without first questioning the science behind it.

I've seen a lot of artists who appear to merely take dictation from a paleontologist and end up producing less-than-accurate art (i.e. practically any dinosaur book with outdated art). I want to move away from that trend and evaluate the science for myself.

BTW, this is not meant to belittle or put down any of my fellow crew members, so don't take it personally.

A lot of times in the past, fossil evidence has been interpreted literally (and wrongly) without assessing the feasible physiology of the animal in question... and so far the bat-winged pterosaur model isn't exactly walking or quacking like a duck.

So while Dr. Witton's art is very high-quality, I respectfully disagree with a lot of the science behind it.

Rachael said...

It's all very confusing but I have to say to a certain extent I agree with Nima. Not that I've researched this subject in any great depth at all but from what i have seen there doesnt seem to be any clear evidence towards any theory regarding the wing membrane. It's just the nature of fossils. I've been researching Dimorphodon and I'd be happy if anybody could let me know if there is any conclusion on this little fella as the only fossils I've seen have been a jumble of bones.
Consequently I've steered away from attaching the wing to the leg at all as I've come to the conclusion he was a bit of a climber as his feet and hands seem comparartively large. I think an ankle attachment would hinder it.
It's all supposition though...

Mark said...

Well, now look what you've made me do: I've gone and got myself all Googled up. Much easier than I thought.

Anyway, Nima: thanks for the nice words about my art: really appreciated. You're also clearly very talented and I've enjoyed seeing your work around the site here. However, and this is a really big, massive however, you need to reconsider not only this question on pterosaur restoration but, if you're really into the science behind these pictures, your scientific methodology. It's obviously correct to question things, but I wonder if your approach is a little flawed. Here's why:

Firstly, by your own admittance, you've not seen many pterosaur fossils indicating ankle attachments. Is that really blurry image of Jeholopterus the only one you've seen? Basing all your arguments on this one specimen, and not a particularly good image of the specimen at that, isn't scientific at all. You should assess all the evidence, in this case all the relevant pterosaur specimens, and then make up your mind. To do this, you need to attack the primary literature: the authors of these papers will tell you why they've interpreted fossilised wing tissue as just that and not, say, iron leaching caused by weathering. In many cases, the actual specimen preserves details that convincingly reveals why membrane remains are distinct from other fabrics, but they just don't photograph or illustrate well.

Next up: if you're so convinced that the interpretation of these specimens are bogus, you need to provide an explanation why. Specifically, you need to generate a more parsimonious explanation as to why all these fossils - at least eight or nine of them - seem to indicate that there is an ankle based attachment. Are they all flukes of preservation? You may argue that for one or two, but eight or nine fossils from different stratigraphic ages, geographical regions and presrvational conditions? And each animal preserved in a different posture? What alternative interpretations can you provide, and what evidence do you have to support them?

Saying that 'I would assume they [pterosaur fossils with preserved membranes] follow a similar pattern' is totally unscientific. The whole point of science is testing ideas: you keep assumptions to a minimum, and you certainly don't assume the outcome of the hypothesis you're trying to test. If you're going to do that, why undertake the investigation in the first place?

In Witton and Naish (2008), we did say that the wing membrane attachment was controversial, but we also clearly stated that ankle-based attachments are probably the way to go based on recent evidence. We're not alone in this regard: Dino Frey, David Unwin, Ross Elign, David Hone, Chris Bennett, David Martill and Matt Wilkinson are just some authors that have recently published papers condoning ankle-based attachments, and only one or two recent papers of dissent have been published to counter the idea. What's more, these anti-ankle attachment papers aren't really that solid: I've chatted informally about them with several colleagues and agreed that much of the science doesn't stand up, and even said so in print. My point is that, without wishing to be rude, do you really think that looking at a few images on the internet is a substitute for studying multiple genuine specimens in the manner that these authors have? That's not to say that your opinions aren't valid, but it would really be in your interests to get to know the specimens and literature on any topic before dismissing the hard work of numerous individuals. To do so implies that we're all a bit thick and, while I can't vouch for myself, I can state that the individuals in that list are anything but simple.

Mark said...

[sorry, this has run onto two comments. Who knew these things had a character limit?]

Picking on wing membrane specific issues now: as you pick on Sordes, I think it bears stating that no-one has yet provided a convincing alternative to ankle-attachment in that specimen. The preservation really is that good: even knee- or hip-attachment championess have conceded this one. Why wouldn't the same thing work for tapejarids and azhdarchids? And have you told the Crato Formation tapejarid with a wing membrane preserved tapering to it's ankle (see Darren's blog for that one - link in the post. Both of us have seen this specimen and can vouch for it's interpretation).

I guess ankle-based membranes are maladaptive for vampire bats too? Doesn't seem to stop them running around. Why can't ankle-attached pterosaur wings be highly elasticated medially and disappear in a similar way when the animals are grounded? Indeed, several people have commented on the possibility of this in the literature.

Just what, specifically, is wrong with the 'bat-wing' pterosaur configuration? You say it's flawed, but never say why.

And I suppose that will do. Apologies if that's a bit rambly and preachy, but I just don't agree that you're being as scientific as you're claiming to be. Dave Hone recently discussed something similar on his blog:

and it may make for good reading.


Mark Witton

Glendon Mellow said...

Works for me.

Thanks, Mark.

Nima said...

>>"Well, now look what you've made me do: I've gone and got myself all Googled up. Much easier than I thought."<<

Thanks very much! In all seriousness, I never would have guessed in my wildest dreams that I could have dragged a PhD paleontologist onto ArtEvolved merely by challenging his theory. Just beware of the temptation of becoming a Googly-eyed addict like most bloggers out there (I myself am sadly not completely immune!)

As for the criticism - that's good, I don't mind the preachiness and in fact I like open debate... BTW I've seen FAR worse. (I'd invite you to read some of my debates with members of BAND, if only I could find the right file in my mess of a hard drive and post it on my blog...)

But a few clarifications I'd like to make.

1. I don't take this personally, but far be it from me to label anyone as thick. What's consensus to all the top experts today may well be rejected by the experts years down the line. Almost everything that Lull, Gilmore, Barmum Brown, Edwin Colbert, and many experienced professors and curators of the last century published about dinosaur posture and functional anatomy has either been rejected by the field or highly modified since their time (the prevailing pterosaur image of that time, incidentally, was bat-winged). And I doubt anyone belittles any of these highly accomplished scientific giants as "thick" despite their many errors.

2. This is related to the first point - evidence can often be misconstrued even by the most knowledgeable professional experts and the best information of the time. Case in point, the hadrosaur mummies in the AMNH. Many paleontologists concluded the impressions of flattened skin around the fingers indicated a sort of web, and for decades artists drew aquatic, web-footed Duckbills without question. Then Bakker pointed out that the "web" was totally the wrong shape for a web, that hadrosaur tails were stiff and poorly adapted for swimming, and that the web was actually squashed skin that was once a a mitten that bound the digits immovably together. Brown and all the old-school experts were merely incorrect in their judgment, not lacking in intelligence. And these mummies are still among the best-preserved dinosaur specimens ever found.

3. As for actual fossil remains that prove ankle attachment - It's true that I don't personally have access to specimens in museum vaults, but I have seen several on display and I only include internet links here for brevity. Would these be some of the ones you're referring to?

SORDES: the bones are jumbled and that dark stain around the tail stretches well outside the ankles - is that really a membrane?

RHAMPHORHYNCHUS: I think this was posted before by Peter Bond, but here it is again - there's no trace of wings attaching to ankles (in any case the legs are jumbled up and not in any sort of natural position)

PTERODACTYLUS: No definite trace of wing membranes anywhere near then ankles. This is a very pristine fossil and not at all jumbled, yet no ankle membranes - just flat stone.


I saw this on Darren Naish's blog... this fossil fragment has caused quite a stir, but I'm curious as to why only a dark stain at the EDGE of the alleged membrane is present, and there appears to be nothing between it and the body except flat stone... are there any fibers?

Nima said...

Finally the main reason I don't agree with the bat-winged model is that all confirmed fossils that preserve the wing membranes well, show distinct fibers of connective tissue within the wings, which would have stiffened the membrane and made it tough enough to withstand wind on a large scale (which big Azdarchoids certainly would have needed). The bat-wing model works fine for bats because they are small and don't normally fly very high or over long distances - and they AVOID windy weather because their wings are so fragile. They don't have the same reinforcing fibers.

My main objection is that if a pterosaur had ankle-attached membranes, they would make walking of any sort - bipedal OR quadrupedal - very clumsy and cumbersome. While I like your art a lot, I have a hard time imagining pterosaurs folding their wings and walking on all fours as in your depictions, UNLESS the wings were free of the legs. Otherwise, I fear the fiber-stiffened membrane either would prevent the necessary change in posture from flight mode to crawling mode, or would simply rip from being twisted so far when the 5th digit bends backward and up. T

I have read many papers that mention these fibers, and also some fossils clearly show them. The fibers' stiffness would also severely hinder the movement of the legs. But if the wing fibers are either insignificant in terms of stiffness, or not real, by all means inform me.

Vampire bat's membranes are not stiffened by fibers so they could theoretically run around, but they can only crawl on a surface to my knowledge, and are horribly clumsy on the ground. And their wing's folding style is totally different from that of pterosaurs.

My lesser objection to bat winged pterosaurs is this: In the old days of Marsh and company, the consensus was, Pterosaurs had skin membranes not feathers, therefore they were built more like bats than birds, bats have ankle-attachment so pterosaurs probably did too. This just seems like the exact same assumption that you said is unscientific.

It's true that bats are the living animals most similar to pterosaurs, but I don't think the similarity in wing structure goes anywhere beyond the superficial.

I know there are many better fossils today that can be interpreted as having ankle-attached wing membranes, but for reasons I pointed to above, I am skeptical about many of these fossils, not just Jeholopterus. If you have images of the genera I mentioned above with higher resolution, by all means post a link. Regretfully, the pterosaur fossil images that ARE on the internet are often wanting for detail.

Thanks for slogging through all this. Being primarily an artist rather than a hands-on researcher, I sometimes find it frustrating that some artists (primarily commercial nature artists who don't specialize in paleo-art) make basic errors in prehistoric illustration that make it very hard for us to be taken seriously as actually having an educated opinion regarding paleontology.

I questioned your conclusions, not because I'm some high-horse detractor with an ax to grind, but because I'm honestly interested in seeing what the evidence can actually prove. And as far as pterosaur fossils go - I've seen a lot of opposition to free-legged pterosaurs but not a lot of details on how the often squashed imprints can be validly interpreted. And I have a nagging apprehension that the "webbed duckbill hand" fallacy may also be rearing its head again in some circles, this time in pterosaur form.

Weapon of Mass Imagination said...

Nima- Though I'm following what you're saying, and appreciate it for the sake of debate.

However I am noticing you have a very "it must be new or bust" attitude towards research and knowledge.

Yes it is true that MANY of the established conventions of early palaeontologists have proven wrong, but at the same time many of them have held up true through the ages.

You can't even hope to talk about or research Dinosaurs that were described by the likes of Barnum Brown or Charile Sternberg without first reading THESE authors! Sure they got stuff wrong, but at the same time some of their work was top notch, even by today's standards.

From what I've seen, typically if an idea has not only passed the hurtle of the pre and THEN post Renaissance eras, but more to the point has been endorsed by both pre and post workers it tends to have something behind it. Otherwise it'll be dogged by conterversy.

So far the kind of conterversey you'd expect hasn't been presented. Rather just the typical few odd guys out on a subject... Who you can find arguing against any safely established "fact" (the bird/dino link for example)

Weapon of Mass Imagination said...

Part 2-

Another thing I'm noticing is your fighting with negative evidence. Which might call into question the ankle attachment, but it doesn't disprove it at the same time.

One problem with using this arguement strategy, is it is easy to have turn right back around on you.

I find these strengthening fibres interesting. One it makes sense. It explains how Pterosaurs were keeping their wings strong without their other fingers holding up the membrane.

However just because they had these fibres in PARTS of their wing doesn't mean it has to be in the whole thing...

Which is where I find you using them as evidence against ankle attachements flawed. If these fibres were all throughout their wing, won't you expect the wing to be tougher and stay attached to the knee or hip since they could afford the these attachments to be stiffer.

Likewise at the least wouldn't we have a consistent case of these fibres found right up to the side of the body... Which I'm assuming hasn't been the case, or you'd have quoted it?

To me the fact that the wing's attachments don't reliably perserve indicate that they were weaker, and thus more flexible at the attachment point. Supporting what Dr. Witton has been saying about the ankle attachments.

The fact we don't find the membranes or fibres attaching to the ankles (or anywhere for that matter, as this arguement is showing) is stronger proof of the membrane attaching where strength and ridgedness were not advantageous.

Rather they had to be weaker so as to allow for functionality. Which makes the most sense if flexibility were the purpose.

Rachael said...

Mark, I'm very glad you decided to join the debate. I just want to ask, as a person who has studied this in some depth, would you agree that there is always room for error when deciphering such ambiguous data as fossils? You've obviously spent some years researching palaeontology so it interests me that you should be so sure of the 'facts'.I'm asking because evidence from pictures of wing membranes( and I confess that's the only resource available to me)can be easily be misconstrued. In the 1882 fossil of Rhamphorhyncus the wing membrane appears to be attached to the leg. This was indeed how Zittel depicted it in his restorations. But what if it only appeared to be attached to the leg but was actually attached behind the leg? I simplify this arguement greatly -and use but one example - and I'm not for one minute suggesting that this has never occured to you and your fellow researchers - but how can anybody be so sure of anything in this field?

Nima said...

>>"However I am noticing you have a very "it must be new or bust" attitude towards research and knowledge."<<

Well Craig, that's not entirely accurate about me, but it's quite a funny slogan so I'll make sure to remember it!

In fact there are plenty of new paleo-theories that I don't agree with, though they're not really pterosaur-related so I won't waste time arguing about them here.

My standard for a theory isn't that it be new, but rather that it fit the facts of both the fossils AND a logical understanding of modern zoology better than the alternative theories. To quote Bakker, "I only champion heresies if they fit the facts better than orthodoxy". And the field has produced several new theories that are as bad if not worse than many of the old ones.

I think you may be on to something though, with the idea that the fibers didn't occupy the entire area of the wing. Now despite the "negative evidence", lets suppose that the ankle-attachment hypothesis was true, and that it actually worked without damaging the wings in walking mode.

The question still comes up as to WHY the wings would evolve this way and not, say, in a tail-attachment configuration that freed up the legs.

The main evolutionary argument for ankle-attachment is that it maximizes wing surface area and improves steering (since the legs can affect banking and rolling like ailerons on an aircraft).

But attaching wings to the base of the tail (or, in short-tailed pterodactyloids, the entire tail) would actually make steering easier, since the legs would have their own independent membranes that could be used for controlling air flow - far more efficiently than if they were bound to the wings. Ankle attachment has certain parallels to delta-wing supersonic aircraft, where the entire wing surface merges with the rear of the plane to minimize stress on the airframe at high speeds and simplify turning (although such a design implies necessary sacrifices in quick maneuverability). Of course, subsonic planes rarely have any need for such a radical design and keel the wings and tail separate to ensure maximum maneuverability. Think of pterosaur legs as having aerial functions similar to the tail of a sub-sonic plane or glider - stabilization, slowing down, and steering.

Pterosaurs had no need for a delta-wing or limb-merging sacrifice of maneuverability since supersonic speed is unthinkable for them, and I can't think of any sort of stress that would simply yank the legs off if they were free from the wings. Though I may be accused of getting far too speculative here, it simply makes sense that animals would not waste a capability they needed (maneuverability in open winds) in order to evolve another one that was impossible and unnecessary.

Nima said...

I think pterosaurs evolved not merely to maximize wing surface area (though this certainly was important) but also to balance it with great maneuverability. It's somewhat pointless for a giant Azdarchid with a 40-foot wingspan to be both a clumsy aeronaut and a clumsy walker due to restrained legs, and be tossed at the mercy of the thermals rather than skilfully riding them with free legs for steering.

Bats don't face this dilemma because:

1. they don't fly very high or far outside of caves and other stagnant airspaces - thus they don't need stiffening fibers and ankle attachment does not affect maneuverability much for them.

2. their wing structure has much more flexibility than that of pterosaurs, and allows them to curve their wings inward to use the wings themselves as steering surfaces.

3. many bats use their ankle-attached wing to catch insects. Some also do this with the caudal flap between their legs. Such actions require a very supple spine and ribcage, and ornithodirans - especially pterosaurs - are well known for having a very STIFF and rigid backbone, especially in the torso.

As a final note, I'm don't really follow the "typical odd guys out" argument. There ARE credible experts who endorse the free-legged model. I wouldn't lump Kevin Padian with the BAND wackos who dany the dinosaur-bird link. He's trained many of the new generation of PhDs including Matt Wedel. And Mike Habib isn't some selective evidence-rigging hack. And of course Greg Paul has always endorsed good science in paleo-restoration.

Peter Bond said...

This debate is fascinating! I'm glad I asked the question in the first place!

I do have one more question: Why is it that talented and well-known artists such as John Conway and Greg Paul resurrect their pterosaurs with hip-attached wing membranes, when there seems to be (as Mark described) such an extensive amount of evidence supporting ankle-attached membranes?

It just doesn't make much sense.

Nima said...

I don't know, Peter, but I have a very strong hunch the reasoning behind Conway's and Paul's "radical" restorations is twofold. You may already be familiar with some of this stuff, but I'll post it in full for the benefit of EVERYONE reading this:

1) First, Paul and Conway appear to consider the evidence of ankle-attachment a LOT weaker than Mark Witton does (with the links I provided above, it's not hard to see why - the evidence does indeed seem scant).

Even the Crato Tapejarid specimen, which is currently among the best "proof" for ankle attachment, has some serious problems. For one thing, the "membrane stain" actually keeps on going past the ankle (if the fossil were complete we'd have a better indication of what this stain actually is), and it seems to continue down as if there is a ridiculously long tail (longer than the tibia anyway) anchoring it.

Now Witton and Naish are firmly convinced this is an Azdarchoid of some kind, probably a Tapejarid. In fact I have no objection to this, as they are the experts here and classifying the thing is primarily a skeletal matter.

The problem is that tapejarids DIDN'T have a long tail - their tails were tiny. So the "membrane stain" could just extend on forever into nothingness and have no connection or attachment to the animal at all - it's also a bit strange that only the edge of the "membrane" got preserved and the inner part of it is just plain blank stone like the rest of the slab.

I don't even see fibers near the LEADING edge of the wing, assuming Craig is correct in his hypothesis that only part of the wing contained fibers (which is still entirely possible). This fossil looks like no imprint of the membrane itself may not have been preserved. Of course this is only one fragment, the rest of the fossil (if it can be recovered) may give better clues.

2) Second, artists like Paul and Conway have a different view of pterosaur biology and structure than some others. I have a copy of his article "The Many Myths, Some Old, Some New, of Dinosaurology" and one of the myths he dispels is that Pterosaurs HAD to be either bat-winged shufflers (the traditional view) OR extreme bipeds (Kevin Padian's view).

Paul argues that neither one is correct, and that Pterosaurs were bipeds when taking off but that the bigger pterodactyloids were quadrupeds when walking. BUT to allow both stances (and a transition between the two) a wing membrane would have to be light and efficient, in terms of NOT hampering the legs.

A big pterodactyloid would need to get a good running start for takeoff, but running would be almost impossible with the legs restricted by wing membranes. One thing I notice when looking at Witton's art is that while there's a lot of emphasis on the probably lifestyle and feeding habits of pterosaurs, there seems to be NOTHING in his portfolio that actually shows the really big ones taking off!

I know "Walking with Dinosaurs" managed to show a takeoff with their bat-winged Ornithocheirus, but that's where the movie magic and computer manipulation takes over from science. The thing was horribly clumsy on the ground and could never realistically take off on its own. Of course I don't think the model was realistic to begin with.

Nima said...

In the final analysis, regardless of how you interpret the fossil evidence, the MAIN issue of wing attachment location is a tradeoff of surface area for ease of movement/maneuverability (which I don't think anyone has seriously addressed here yet).

Ankle attachment has ONE advantage: it gives the wings maximum surface area and thus (theoretically) maximum lift.

But it has one huge DISADVANTAGE as well - it restricts the movement of the legs, making steering difficult and the initial takeoff almost impossible (unless your pterosaur is cave-dwelling or arboreal or hangs on to cliffs every minute that it's not flying - which is VERY unlikely for big Azdarchoids and their relatives). Furthermore, if the wings were attached to the tail (not the hips OR ankles) this would allow a large surface area without hindering leg mobility.

Sure it wouldn't be as much surface area as with the ankle attachment, but large soaring birds like condors and the extinct Argentavis don't need so much surface area, and so it's doubtful big pterosaurs (which already had much longer wings) did either. Imagine Azdarchoids as parallel to big soaring birds rather than cave-dwelling bats. It's a far more realistic comparison in lifestyle, and helps explain a lot of things that the bat-winged model doesn't.

CONCLUSION: people like Paul and Conway probably choose the free-legged model over the bat winged model because it's far more efficient for takeoff as well as walking on the ground. A hypothetical bat-winged pterosaur can not both run and flap its wings at the same time - which is a DISASTROUS limitation for self-powered takeoff from land. Furthermore, these guys likely didn't NEED the excess surface area that ankle-attachment would have provided.

So Conway endorses the hip attachment model, but Paul's model is actually a bit different, he uses the tail as the attachment point, so the membranes actually go above the hips (which provides a bit more surface area than Conway's model, and also explains why pterodactyloids didn't lose their little tails entirely). Paul's model thus eliminates a good bit of the tradeoff dilemma between wing surface area and leg mobility. The rest can be best resolved by - well - Argentavis.

Both Conway and Paul accept the possibility of independent leg membranes for steering or additional lift. I prefer Paul's model because it's more elegant and intuitive (some fossils do appear to show a tail-attached wing membrane that, due to folded-up knees and overlapping leg membranes, can be misinterpreted as ankle-attached wings - see Jeholopterus) but I don't have any serious issues with Conway's model either.

Another competing model is Mike Habib's, which is essentially a very VERY sharply angled ankle attachment with extremely thin leg margins for the wing membrane - more mobile than Witton's version, but with less surface area.

* Note (for those that don't already know), Greg Paul has also been a published veteran of peer-reviewed literature on paleontology for many years and his contributions to the field should not be ignored - sadly I see many people dismissing his theories and labeling him as merely an artist because he isn't a PhD - but his research is peer-reviewed and should be judged mainly on its content, not just his degrees.

Nima said...

Clarification: "The Many Myths, Some Old, Some New, of Dinosaurology" is Paul's article, not Conway's.

Dinorider d'Andoandor said...

Excellent!! I´m more confused than what I had thought but I like this debate! I'm looking forward to seeing more!

Christopher said...

Ok the extreme length of some of these posts is going to make commenting on every thing said a bit difficult but here goes nothing.

Some things Ive noticed, the "strengthening fibers" of which you speak, nobody calls them that. They are called actinofibrills. Though strengthening is usually their suggested role, at least one other theory to their use has been put forward. In either case they are never found inbord of the elbow so even in an ankle attachment, they would not hinder ground movement at all. In other words, there is no reason to consider pterosaurs clumsy or cumbersome. Vampire bats are a little awkward on the ground not because of an ankle attachment but because their legs are evolved to hang from trees and caves, and are not at all like pterosaur hind limbs. The membrane of pterosaurs was highly elasticized and in this region, between the elbow and body wall down to the ankle attachment, and would have sinched in tight to the body and leg when not under tension from the flexed wing finger. And then the leg which although not completely parisagital (they arent in dinosaurs either btw) is free to move unimpeded. This explanation has been used to explain why the two specimens of Pterodactylus that show wing membrane, would seem to suggest a knee attachment.

At the vary least Paul's usage (cause it's not his theory btw, he just incorporated it in to his illustrations because it was the latest research at the time) of a tail based attachment is flawed. Its more parsimonious to conclude that all pterodactyloids at least, had a knee based attachment rather than a tail based atachment because of these two specimens. The Vienna specimen ( and the "I dont remember its moniker" specimen ( Both show the exact same preservation of the right wing membrane, with the membrane seeming to anchor itself on the femur just above the knee. This has been interpreted in several ways. One, that it represents a genuine knee attachment. Or, two, that the wing passes behind the limb and is rooted at the ankle, with the membrane otherwise obscured by the leg and the sediments.

Another thing, an ankle attachment would not hinder a bipedal launch at all, simply because they did not bipedally. They launched quadrupedally like vampire bats, but with a little more initial push from their legs in addition.

This is starting to get get longer than I anticipated, so to sum every thing up: you are not wrong, but your not the least bit right either. I for one am not entirely convinced that these fossils all show what we are told they do. I disagree strongly with Mark's saying that the authors tell you why they interpret certain stains certain ways. Ive have read the majority of the literature and the fact is that they more often than not, say nothing of what sparks in their brains. I do think a lot of that sort of brain picking goes on at svp an such, which is unfortunate cause we dont get to benefit from a lot of it. However, the fact remains that there was at the very least some form of leg attachment and a tail attachment is definitely not supported by any form of evidence. Also, these traces really don't photograph well, and the scientists have had their faces mere centimeters from the fossils in question so I think we should give them the benefit of the doubt. At least for now.

FlyingDaggers said...

Wow, this is quite the conversation! I have not had a chance to read everything said in detail, yet, but a few comments:

1) Pterosaur wings did not have consistent material properties throughout. We know that the inboard wing was more elastic, and this would allow stretching during walking and launch (see below on the latter). This elasticity means that the inboard wing would have reduced lift potential, but that has little impact because inboard vorticity is weak, anyway.

2) A super-broad wing has some potential structural problems, but a narrow wing with a sharp turn to the ankle would not. Interestingly, placing the turn at the most well-supported point, based on known fossils, also puts the angle behind the elbow nacelle, which would then be expected to become an advantageous vortex shedding point.

3) Bit of a shameless plug here: no evidence for a running launch in any pterosaurs, and no evidence for a bipedal launch, either. On the other hand, there is a great deal of evidence for a quadrupedal launch (the first chunk of which showed up in my Zitteliana manuscript: see Habib, 2008).

4) All flying animals use the wings to turn; bats fly well without fibers imbedded in the wing membrane because they have multiple spars running through the wing. Bats are also capable of high altitude flight, so the ankle attachment does not hamper them in that regard. Pterosaurs could also camber effectively, in part by using the propatagium. If the wing attached to the ankle (likely), then trailing edge shape could be affected by hind limb position.



Nima said...
This comment has been removed by the author.
Nima said...

Wow, Mike Habib! Welcome to ArtEvolved!

Our crowd of guests just keeps getting more interesting!

While I personally prefer to illustrate pterosaurs with Greg Paul's wing model, I certainly accept that yours is a very viable possibility. Props to both you and John Conway for keeping the field diverse and lively.

But one question - what exactly is a "quadrupedal takeoff" ? Does the creature jump up and flap? Or dive off a cliff? And what is the basic evidence to support it?

I know there isn't DIRECT evidence of running bipedal takeoffs, but for large pterodactyloids I can't imagine an alternative... they were just so front-heavy. They could run with the body tilted up as per Padian, and flap to get airborne... but how would they flap in a quadrupedal posture?


Also a note to Christopher - nothing personal, but you seem to post comments with an aggressive certainty as if you have some sort of inside information we lack... if so, by all means share it. Mainly I see a lot of absolute assumptions with very little explanation or proof (the absence of fibers near the legs does NOT imply a fiber-less membrane - or any sort of membrane - attached to the legs). Alternatively, you can make a blog of your own to explain your ideas in more detail and familiarize us with your scientific background.

Perhaps you may also want to re-examine the confrontational tone of your criticism. I'm well aware that the 'strengthening fibers' are called actinofibrils. You're perfectly free to use either term - one's physically descriptive, the other is more formal. Let's try to keep the debate civil here instead of playing jargon police with patronizing statements like "nobody calls them that".

Now putting all that aside, welcome to ArtEvolved!

Weapon of Mass Imagination said...

Nima and Christopher-

To start off with Nima, I didn't read any comments so far as aggresive towards you, but rather pointed towards your arguments, which I think is a key distinction. It is bound to happen with these sorts of debate...

Though a friendly reminder to all to keep it, well, friendly (not that I've noticed anyone getting too heated.)

As for Nima's request to you Christopher to not come down on people for not using techno jargon, I agree with him completely.

I remind everyone (as the administrator) this is a art site, and thus many of us here are not properly scientifically trained, and wouldn't understand terms like actinofibrils without a 'dumbed' down explaination like "strengthening fibres". (Which I will add was a great description Nima! I followed what you were talking about instantly)

At the same time I didn't interpret it as a totally hostile statement, but it does have a hint of elitism too it. Lines like "nobody calls them that" are not needed. Please use neutral language like "those fibres are called *insert* in the literature" or by workers etc.

I understand that with contentious issues like these feelings can run high, but everyone keep it civil please. Also remember that our natural internet reading filters tend to go for negative and confrontation. Keep this in mind before responding while you're reading.

FlyingDaggers said...

Great questions about launching!

There is a great animation of the proposed sequence now available, pulled together by a Masters student in Art as Applied to Medicine here at Hopkins (I was her advisor last year for the project). However, she plans to unveil it at AMI, so I will discuss it with her before I post the URL here.

In the meantime, some text description will have to do. The quad launch hypothesis suggests that the initial takeoff leap comes from all four limbs - the animal vaults over the forelimbs by pushing with the hindlimbs, then pushes powerfully with the forelimbs (which provide most of the power). This launch sequence uses the powerful forelimbs and pectoral musculature for most of the push, which greatly improves launch performance. Once the animal is airborne, it begins to flap (upstroke immediately follows the push from the ground).

This may seem odd, but vampire bats quad launch, and birds use a "push first, flap second" sequence, as well: it's just so fast we cannot see it without videography. Earls, Rayner, Tobalske and others have demonstrated that the vast majority of the launch force in a bird comes from the hindlimbs. The wings only provide 10 to 15% of the launch force in the majority of taxa. Even hummingbirds provide 59% of the launch force from the hindlimbs.

Running launch turns out not to be a size-related system; running launchers are nearly all aquatic, or descended from aquatic taxa (based on both morphological and molecular phylogenies). It is a way of dealing with a bipedal launch from a compliant surface (i.e. water). Giant pterosaurs do not have any of the structural traits associated with running launch, and we have little reason to reconstruct a running takeoff for giant pterosaurs. They would not gain anything, and probably are not capable of it, anyway.

In terms of evidence, here is some of the most compelling:

-Pterosaurs have a reversed limb strength scaling trend, as compared to birds. Their humerus scales in a manner very similar to the femur of birds (gets proportionally stronger with size). By contrast, the femur becomes more gracile in large pterosaurs.

-The femora of giant pterosaurs would break under a bipedal launch (Quetz sp, for example, can only take about a body weight per hindlimb). Humeral strength, however, is more than sufficient for taking most of the launch force.

-The quad launch model predicts and explains the larger maximum size of pterosaurs, relative to birds.

-Orientation of major trabecular bracing in the dp crest of azhdarchid pterosaurs is consistent with a quad launch, but would be unexpected under flapping loads, alone.

-Pterosaurs are very front-heavy, especially the large species (you pointed this out, as well)

-All pterosaurs I have so far examined appear to possess a lock and release catapult mechanism for flexor digitorum longus - this would increase quad launch power, but have practically no function otherwise.

-Making pterosaurs launch bipedally requires a change in gait: this means that they must switch to a bipedal pose, which it dubious to begin with, in order to launch, even though that switch would actually negatively affect takeoff performance.

-Quad launch allows even the largest pterosaurs to take off from a standstill, with no cliffs or special gusts. This would be expected for things like azhdarchids, which seem to have spent large amounts of time on the ground in floodplains. Bipedal models would ground them while feeding, requiring a multi-mile trek to higher ground to take off.

-Quad launch would place all known pterosaurs at or above stall speed from the leap alone, removing the requirement for a ring-shedding, slow-speed gait. This is important, because most pterosaurs (animals like anurognathids excepted) had a morphology that made them relatively committed to a continuous vortex gait, as best I can tell.

There's more, but that is a start. Hope it helps!



Nima said...

Wow Mike, that's quite a lot of description. I can almost imagine a Quetz doing a quad takeoff now, though it still looks a bit odd in my mind... can't wait to see the animation video, hope she's ok with showing it.

I always knew big pterosaurs were very front-heavy but I had no idea quite how much strength was in the chest and arms. Good news is, now it's well known they were strong enough to flap and didn't have to rely on cliffs and thermals for every takeoff.

Though with little guys like Dimorphodon I have a hunch they could still run bipedally.

Rachael said...

Flying Daggers, Mike-

Hope you're still following this debate - I'm a bit late responding!

Looking forward to more exclusive info on this if we are able to see it.

Is what you described a kind of 'leap frog' motion into take off?