Four legs good, two legs bad?

Ohhhhh boy. Right, best make sure I am fully prepared for this one:

Can opener? Check.

Industrial grade can of worms that looks like it houses a few thousand of the critters from Dune or Tremors? Check.

Most contentious area of pterosaur research? Check

Right, let’s get going then. My main aim here is to get across my opinion without upsetting or offending most of my colleagues. It could prove tricky. Pterosaur terrestrial locomotion has been back and forth numerous times between numerous people numerous times and while I might dare to venture that *most* pterosaur researchers are now more or less agreed on the basics, there are some exceptions and they include some well respected and knowledgeable researchers. I am not trying to play up the controversy (that is not needed) but I want to make the point that whatever I write here, if it were a review paper, would be disagreed with strongly by some very expert colleagues (and if fact I have a few problems with bits of it myself). While this kind of post necessitates I don’t go into great detail (and I don’t bother with references, though I do generally check things in the literature before writing them) I will try to cover the various hypotheses, interpretations and issues at stake, without getting bogged down into the minutae of pelvic anatomy or footprint identity. You will just have to take my word for it, read a few papers, or ask probing questions in the comments that make we write up all the details I am trying to avoid….

First off, some basics. Hopefully you will know that pterosaurs are split into two major clades – the ‘rhamphorhyncoids’ and the more derived pterodactyloids. Phylogeny aside, they have some notable differences in anatomy, both involving the bones and soft tissues and both of these changes are important in this discussion. Probably the biggest is the change in the uropatagium (the rear wing) as in the rhamphorhynchoids it links the two hind limbs together and does not do so in pterodactyloids (see the image below, courtesy Dave Unwin).pteros-rhamps Secondly, I am assuming that the main wing attaches to the ankle in everything I talk about, this may not be true (though I think it highly likely), but given the inferred flexibility and elasticity of the membrane, I do not think it would actually have any real effect on any of the possible interpretations of pterosaur terrestriality in any case. Thirdly, pterosaurs did not hang by their claws from branches or caves like bats. For a start in most the claw shape is wrong to provide any kind of real grip in this sense, and secondly the feet are the wrong shape, bats and other hanging animals like sloths have digits that are all the same length to get a good grip, and not put excessive pressure on a single digit, which we never see in pterosaurs.

Right, onto some specifics, there are three main issues that need to be dealt with here that kind of overlap and interlink making this generally a hard thing to review, but I’ll try my best:

1. Did rhamphorhyncoids and pterodactyloids differ in their ability on the ground?

2. Were pterosaurs bipedal or quadrupedal?

3. Were pterosaurs arboreal or terrestrial?

As you can see there is a potentially awkward overlap as we can be talking about bipedal, terrestrial rhamphorhyncoids, or quadrupedal, arboreal pterodactyloids, or any other combination, so the ‘terrestrial’ nature of pterosaurs is a complex issue. We need to establish who was doing what, where, and how in order to make a comprehensive case for pterosaur terrestrial locomotion.

First off, did the two clades differ? Well, certainly there are extensive anatomical details between the two and thus one would expect this to translate into different approaches to flight and life on the ground. The key issue here is the uropatagium – with the hindlimbs shackled together in rhamphorhyncoids, they are left with all of their limbs effectively joined together, a wing membrane from finger to ankle, the uropatagium linking the two legs, and then the other wing on the other side. This would obviously make walking tricky as, despite the elasticity of the wings (and presumably a fair amount in the uropatagium) everything has to move together. If you want to move an arm forward then you will have to bring one leg with it, and that leg will bring the other leg with it, and then the other arm. This might sound impossible, but it certainly isn’t, (just look at bats and colugos), but it does make this tricky, and it certainly limits your stride length. Pterodactyloids of course have no such problems and can move their hindlimbs freely from each other (though of course each is attached to the respective arm) because their uropatagium does not link the hindlimbs together. Of less importance is the build of the femur and pelvis, which in pterodactyloids allows a far greater freedom of movement of the femur, and specifically allows it to be brought under the body, which is impossible in rhamphorhyncoids – theirs rather sticks out to the side. So far so good, they *are* different, but it hasn’t really answered the question of did they differ on the ground, not just were they different. To do that we need to explore the second area – bipedality vs quadrupedality.

No pterosaur has (to my mind) been demonstrated to be capable of bipedal locomotion. Certainly most, if not all, pterodactyloids could *stand* bipedally (it would happen during take off and landing) and perhaps they could take a few tottering steps in the pose, but one would not call them bipeds. But let us deal with the rhamphohyncoids properly first. As mentioned elsewhere here, the exact origins of the pterosaurs are still rather confused (despite my best efforts and that of my colleagues) but and as such we cannot really say what the exact mechanical pattern of a likely ancestor was – a biped or quadruped, or even facultative biped (one that could move on tow legs in some situation, such as at high speeds). There are lots of different archosaurs and archosuromorphs floating around with various configurations, and multiple transitions to and from bipedality, and different levels of ‘uprightness’ (from sprawlers like lizards, to fully upright animals like birds) so even if we could definitely determine that the sister taxon to ancestral pterosaurs was a biped, it would not really be a big problem to imagine pterosaurs as quadrupeds (though it would have to be carefully considered and actually the current consensus suggests that this is actually the case). With the arms and legs bound together it is very hard to see how and rhamphorhyncoid could possibly have brought itself upright onto two legs and been able to take even the shortest of strides with the ankles linked together, and being massively top heavy. Added to that the impossibility of bringing the femur under the body leaving them with sprawling hindlegs, and they are left with widely splayed legs, shackled ankles, and a heavy front end (the tail is nothing like as thick or long as a theropod dinosaurs, so would not be a decent counterweight either) and we are left with rhamphorhyncoids being quadrupeds.

But that is hardly the end of the story, quadrupeds yes, but what kind? Well again, with the sprawling posture, they would not have been truly upright. Despite their likely position as ornithodirans, the constraints of flight have left the rhamphorhyncoids with a sprawling to semi-erect posture (though they might be considered ‘erect’ by descent among other considerations). It is an interesting (if unsurprising) fact that with so many modifications to the archosaurian bauplan dedicated to getting them into the air, that this should cause all kinds of issues with them on the ground. So with their ‘splayed’ profile, rhamphorhyncoids would be quadrupeds with well spread out fore and hindlimbs. So, onto the pterodactyloids…

As I mentioned above, pterodactyloids certainly entered a bipedal pose on occasion (if only as part of the air to ground transition of take-off and landing) but themselves were also quadrupeds. In addition to an anatomical arrangement that argues against bipedality (long neck and large head, small feet, reduced tail all of which would have made balancing very hard before we introduce the heavy wings!) we have convincing pterodactyloid tracks. These are clearly pterosaurian (based on the number, shape and orientation of digits in both hands and feet) and demonstrate irrefutably that pterodactyloids were quadrupeds. They were however, upright quadrupeds, with legs brought in under the body and arms close to doing so.


Pterodactylus and footprints

Independent corroboration of both postures comes from unpublished modelling work by Don Henderson (though bits of it are mentioned in the literature and the graphics make an appearance in Dave Unwin’s recent pterosaur book). One final point that also applies to rhamphorhyncoids is the orientation of the foramen magnum (the hole at the back of the skull for the spinal cord to pass through) and the orientation of the inner ear, both of which suggest the head was held in line with the spine. In an upright quadrupedal posture and especially in a bipedal one, the head would be pointing at the sky and the animal would not be able to see where it was going, which is not the case for pterodactyloids who would be looking slightly down the whole time, which makes sense if they are upright quadrupeds. So, we now have both groups as quadrupeds, though rather different ones – with rhamphorhyncoids sprawling, their limbs linked, and pterodactyloids as upright walkers, but were they terrestrial or arboreal?

The pterodactyloids first as they are easier to deal with. These are considered to be truly terrestrial, living on the ground and while one would expect the odd venture into the trees (look at some of the ‘terrestrial’ animals that can climb surprisingly well such as goats and crocodiles) from a few of them, this would be very much the exception and not the rule. We do have numerous footprints known for a variety of pterodactyloids and it is noticeable that their appearance in the fossil record occurs almost simultaneously with the presence of actual pterodactyloid specimens. One would also think that the average pterodactyloid would struggle in the trees – they are typically much bigger, often have large crests and the positions of the arms would make the wings stick out awkwardly, all of which cause various bits of the to tangle in any loose branches or leaves. Their fundamentally upright posture would not help either as it means their centre of gravity would be far from the branch or trunk they were trying to cling to in the first place. Finally, in the opposite situation to the rhamphorhyncoids, their foramen magnum and inner ear leave the head pointing down from the spine – fine for an upright quadruped, bad for a ‘flat’ climber.

Bearing all this in mind, it is easy to see why the rhamphorhyncoids are considered arboreal animals (or possibly climbing around on rock faces). They are small, flattened on the substrate with the sprawling posture, don’t generally have big crests, the wings would not stick out (as they would rather point backwards than up as in the pterodactyloids), and the head was flat in line with the body (as was the tail). While one must always be careful of absent evidence, again there are no footprints that can be ascribed to rhamphorhyncoids despite their existence for a long period of time, in large numbers, and across a huge swathe of the world. How were they able to knock around for tens of millions of years in large numbers and in numerous environments (many of which, like pterodactyloids, featured water and coasts ideal for leaving prints)? The obvious answer is that they were simply not coming down on the ground, but staying safe in the trees or perhaps in some cases on rock faces.  This also fits with their inferred locomotion as being ‘sprawlers’, they would probably have had far more trouble than pterodactyloids in taking off from the ground, and a launch from a raised surface would have been much faster and safer for them. Doubtless they would have come to the ground occasionally (to drink or lay eggs) but being a rare occurrence this is unlikely to have made it into the fossil record, though prints may one day appear.

Hopefully what appears here is a coherent picture of pterosaur terrestrial biology. Bones, joints, footprints, modelling, the fossil record and inferred ecology all fit together in a complimentary manner. Rhamphorhyncoids were crawling or scuttling arboreal sprawling quadrupeds, with pterodactyloids striding around on the ground as upright quadrupeds. This is not to suggest that either group was slow or ungainly in its environment, anyone who has seen a grounded bat will tell you that they are not half as slow as one might imagine, and I would have expected most pterodactyloids to be no slower than other ‘imperfect’ quadrupeds (like crocs say). Just because they do not conform to our picture of what a ‘good’ terrestrial mover looks like, does not make them poorly adapted.

One final note I will leave you with is the (apparent) controversy over the picture I have painted here. If you talk to most pterosaur researchers about pterosaur terrestriality they will pretty much you the same answer that I have produced here, but there are some vocal dissenters. I won’t go into their arguments here (this is long enough as it is, and most of it is well documented in the literature) but suffice to say, I am unconvinced and I think the balance of the evidence is massively in favour of the information laid out above, but don’t be surprised if you see contradictory ideas being expounded in some quarters.

17 Responses to “Four legs good, two legs bad?”

  1. 1 Mike Taylor 01/12/2008 at 6:10 pm

    Er, is Rhamphorhynchoidea a clade now? When did that happen?

  2. 2 David Hone 01/12/2008 at 10:49 pm

    No, it’s still a paraphyletic group. I *should* have put that in, but then if you had read the ‘basics’ post you would know that already. As it happens I have several posts on hand on definitions of paraphyly and monphyly etc. and the issues of using those terms in systematic discourse. Ohhh, I can feel the excitment building. Possibly. I *will* go back and add ‘ ‘s though.

  3. 3 Jerry D. Harris 02/12/2008 at 1:06 am

    My understanding at present is that, technically, paraphyletic (and polyphyletic) groups cannot be called “clades” — “clade” apparently inherently includes monophyly as part of its definition. Paraphyletic and polyphyletic entities are certainly still groups…just not clades. Can’t recall offhand where I got this from…???

  4. 4 Zach Miller 02/12/2008 at 7:33 am

    How about just “long-tailed pterosaurs?” That about covers the entire paraphyletic group, aside from anurognathids.

  5. 5 David Hone 02/12/2008 at 9:17 am

    Jerry, that is a new one on me, but I’d be interested to see where you got that from. It’s the kind of thing that *sounds* right and I’d happily adopt it. I thought that was true of ‘taxa’ but not ‘clade’.

    Zach, you’ll have to wait for the post on that very subject, but you have already spotted the obvious flaw and it works both ways.

  6. 6 Jerry D. Harris 02/12/2008 at 10:59 am

    Hmmmm…still don’t know what is the ultimate source for my understanding of the limitations of the word “clade”, but it’s also how the Bezerkely people use it (, for whatever that’s worth. Certainly, however, there are innumerable examples out there in the published literature to “paraphyletic clades” and the like, so I don’t know how widespread a more restricted definition is. I’m one of those people that fully believes that definitions shouldn’t be all that flexible in most cases or else communication becomes obfuscated, so I like having clade mean just one thing (other than “group”), but I’m willing to be convinced otherwise…!

  7. 7 David Hone 02/12/2008 at 12:18 pm

    I certainly see where you aere coming from, sadly what we want and what we get are often two different things. ‘Trees’ and ‘cladograms’ are now pretty much synonymous when they really shouldn’t be. My one point would be what word would you replace XXXX with in the sentence “as for the pterodactyloids and ‘rhamphorhynchoids’, both XXXX show different features”. Once could use groups becuiase of the paraphyly of the rhamphs, but that actually undemines the true clad-y-ness of the pterodactyloids. And then the reverse is true if you use clades for XXXX. And if you use something else like ‘sets of taxa’ it all gets a bit vague. I’m really not sure there is an obvious and more importantly *practical* solution. As long as *some* people thing clade can be paraphyletic or monophyletic, it doesn’t matter how specific soem people are, others won’t be, and won’t interpret the use of the word in that way. It’s fien if you only use clade for monophyletic groups, but if i don’t (and don’t make it clear) you may think I am referring to monophyletic groups when I’m not.


  8. 8 Nick Gardner 02/12/2008 at 5:33 pm

    Clades are monophyletic groups only. I believe you are thinking of ‘grades’.

  9. 9 David Hone 03/12/2008 at 1:07 pm

    The plot thickens! My two primary sources for cladistics contradict each other.

    Kitching et al.’s ‘Cladistics’ calls a clade “a monophyletic group”, whereas Waegele’s superb ‘Foundations of Phylogenetic Systematics’ says explicitly a “clade is not necessarily a monophylum”. Neither even has the word grade in the index. ‘Transformed cladistics, taxonomy and evolution’ by Scott-Ram considers grades to be points marked along an evolutionary branch (i.e. a major change but not necessarily marked by a splitting event).

    In short, there is no consensus at all on ‘clades’ and Nick I can’t find a defintion of ‘grade’ that fits your description.

  10. 10 Nathan Myers 03/12/2008 at 3:15 pm

    Oh, great. Now “clade” means the same as “group”, and we need another word for a “monophyletic clade”. No doubt that one will get corrupted, too, and we’ll need another after it.

  11. 11 David Hone 03/12/2008 at 3:42 pm

    What’s wrong with the word ‘monophyletic’? You just need to add it to a ‘clade’ as a qualifier.

  12. 12 Jerry D. Harris 05/12/2008 at 3:41 am

    Oh, great. Now “clade” means the same as “group”, and we need another word for a “monophyletic clade”. No doubt that one will get corrupted, too, and we’ll need another after it.

    That’s the beauty of clade = monophyletic group — then clade does not = group. The qualifier gives the term a more restricted meaning. If clade = group, then Dave is correct that we’d always need qualifiers (monophyletic clade, paraphyletic clade, etc.), and that’s just silly — makes the word “clade” kind of superfluous (other than it sounds silly for people to do a groupistic analysis). “Group,” or even “taxon” would suffice, and there would not have been a need to invent a new word. The more I’ve thought about it, the more I like “clade” as meaning “monophyletic group” specifically for this very reason.

  13. 13 David Hone 05/12/2008 at 3:19 pm

    Actually Jerry it does appear as iif I have misread the quote in Waegele, is seems he may only have been referring to a complex hypothetical sitaution where you are referring to a monophyletic group that is not one (e.g. such as referring to a contradictory study) and thus he does consider clades monophyletic. However, you are right that there are still plenty of people out there who are not aware of this (including myself previously it would seem) who use ‘clade’ pretty much as a synonym for group as I agree that one often sees the phrase ‘paraphyletic clade’ in the literature. As with ‘trees’ a general clean up appears needded, but I have no idea how one would go about it.

  14. 14 Jerry D. Harris 06/12/2008 at 6:16 am

    I’m in the midst of tracking down what apparently (according to Wikipedia, at any rate) the first use of the term “clade” (Huxley, 1958) as opposed to “grade” — I’m just curious to see how the originator of the term intended it (even though it’s prior to the advent of Hennigian cladistics!).

  15. 15 David Hone 06/12/2008 at 9:00 am

    That’s an excellent question Jerry – I’d be intrigued to know. I also wodner if Hennig himself used it, though sadly I don’t have a copy to hand of the original.

  16. 16 Jerry D. Harris 06/12/2008 at 11:17 am

    I also wodner if Hennig himself used it, though sadly I don’t have a copy to hand of the original.

    I don’t know either, but couldn’t say based on experience anyway as I don’t have a copy of Hennig. However, Wikipedia — take it or leave it; I don’t know who wrote the article ( says:

    Hennig’s major book, even the 1979 version, does not contain the term cladistics in the index. He referred to his own approach as phylogenetic systematics, as implied by the book’s title. A review paper by Dupuis observes that the term clade was introduced in 1958 by Julian Huxley, cladistic by Cain and Harrison in 1960, and cladist (for an adherent of Hennig’s school) by Mayr in 1965.

    If this is correct, then I guess that what many people today call “cladistics” came together piecemeal from the contributions of many people along the line, beginning even before Hennig’s magnum opus!

  1. 1 Crystal Palace pterosaurs « Dave Hone's Archosaur Musings Trackback on 12/10/2011 at 4:32 pm
Comments are currently closed.

@Dave_Hone on Twitter


Enter your email address to follow this blog and receive notifications of new posts by email.

Join 574 other followers

%d bloggers like this: