On a wing and an ankle attachment

Proof, if it were needed, that the Pterosaur.net team do actually work on pterosaurs and not just blab endlessly about them comes from todays new paper featureing Ross Elgin, myself and Dino Frey. Once more, this is an odd time to talk about it since what has actually turned up are the uncorrected proofs, but it is out there and being read, so now is the time to talk about it. (You can download and read it here, and a very old post of mine here might be a good primer if you don’t know your pterosaur wings too well).

This paper has it’s roots in Ross’ PhD work. In short, it’s hard to test and discuss the flight profile of a given wingshape for a model pterosaur when there is little agreement (at least in the literature) over how the wings are shaped. If you look around you can find nearly a dozen different ‘kinds’ of wing that have been produced based on various specimens of various taxa at various times. It seemed high time that someone synthesised all of the available data from every specimen and tried to tie down exactly what the evidence was and what it told us about pterosaur wings.

Various wingshapes and attachment positions for pterosaurs. From Elgin et al. in press.

The two main variable here are how deep (broad chorded) the wing is, and where it attaches to the animal’s body – at the hip, somewhere on the leg, or right down at the ankle. If you ask around, (and being firmly embedded in the pterosaur research community, I have) most people seem quite happy with the idea that it terminates at the ankle. But there is very little in the literature on the termination point and most of what there is actually favours a more proximal position. Why the ankle position?

Well, for the simple reason that most good (and that word is important as we shall see) specimens of pterosaurs, that preserve an intact (again, important) wing, show that it reached the ankle (or close to it). Some of this is directly observable. Some comes from UV photos. Some comes from admittedly less good specimens. But all of them seem to have the same pattern. Interestingly, this is also incredibly conservative in that we get this is anuroganthids, derived rhamphorhynchoids, basal pterodactyloids and azdarchoids. No matter which pterosaur phylogeny you prefer, it’s hard to ignore the possibility (and indeed most parsimonious explanation) that all pterosaurs likely had this attachement. At the very least, it should be the default assumption.

So why the change from a non-ankle position? Mostly becuase we just have so many good specimens to wrok from that have appeared in the last few years, when in the 70s and 80s we were limited to often fragmentary wings of Solnhofen material. Natually China and Brazil has a lot to do with this, but we do include new German material too and refer to older specimens with a new eye. New and clearer information makes all the difference.

Schematics of pterosaur wing preservation and attachements. From Elgin et al. in press.

So what about the caveats of ‘good’ and ‘intact’. This comes down to the taphonomy / preservation of the material. Pterosaur wings you see, despite their complex structures, are still essentially huge membranes (that were likely highly elastic) and thus unlike bones potentially subject to great distortion between death and discovery. You can see in a few at least that there are wrinkles and folds in the membrane showing that what is seen directly can’t be the natural shape of the thing when unfurles and in flight (it must have been tight to serve as an aerofoil). Secondly, you can see large variation in the way (and most importantly the size) of the wing in some species. One specimen of Rhamphorhynchus for example has a tiny slice of a wing membrane that appears to be intact (it’s not a loose piece) but has clearly shrunk or rolled up to a fraction of the size seen in other specimens (like the Dark Wing and Zittel Wing specimens).

Small-winged Rhamphorhynchus. From Elgin et al. in press.

Our conclusion is that such specimens with impossibly small wings have undergone post-mortem shrinkage. In essence, if you see a pterosaur wing on a specimen, it must be the minimum size of that wing in life. They can shrink after death, but not get bigger. If you have two specimens of a species with one small and one big wing, you should start with the bigger one as your minimum. This also affects the apparent attachement point of the wing, since if the wing was rather narrow close to the body, even a small big of shrinkage might make it appear thinner still and appear to pull it anteriorly (and this is likely as the proximal wing lacks the actinofibrils of the distal wing). Conveniently, when wings do move or dissassociate from the rest of the animal they tend to pull off the medial attachment first (as seen in the Zittel wing for example) and then disentigrate. In other words, an apparently thin wing attaching to the hip could be the result of post mortem changes from a bigger, deeper wing, but a broad wing reaching the ankle can’t have shrunk from a bigger one (nothing to attach to) or been an expanded small one (it would detatch and fall apart).

One thing we are limited on though is the ability to determine the chord of the wing. While we can rule out a hip or knee based attachment, attacheing to the ankle does not mean the wing has to be broad. Those people who have argued for and illustrated narrow winged pterosaurs (especially ornithocheirids and pteranodontids) can still be right. A sharp inturn of the wing close to the body is plausible and gives a narrow chord wing with a ‘broad’ ankle attachement.

So there you have it. Pterosaurs had an ankle attachement and (in many cases at least) a bread-ish wing, though that could be sbuject to great and unseen variation. Those in the field will not find this much of a surprise, or even (I hope) find it rather too their taste. Even if you thoroughly disagree with everything else here, what we have at leact achieved is a framework for discussion. It’s really hard to debate the merits of various interpretations when links like the ankle attachement are, in places, used synonymously with a broad chord when this is not the case.

By searating out the issues into several (linked) issues rather than one comprehensive one it becomes much easier to deal with each aspect in turn and use problematic fossils that can contribute to one problm even if not the other. Others might (and let’s face it, probably will) argue in print, but I expect that the image at the top of the different wing planforms will serve as a useful guide for many a long year.

Ross A. Elgin, David W.E. Hone, and Eberhard Frey. The extent of the pterosaur flight membrane. Acta Palaeontologica Polonica in press. Available online 14 Sep 2010.

Note: this is an uncorrected proof. There are some typos and other errors in there and some of this might change a little before formal publication. I don’t like the system, but not much I can do about it.