Pterosaur wings clarification (quite long)

[Unwin <h0662eka@rz.hu-berlin.de>]

Those who have been following the discussion of pterosaur wings and related 
topics may welcome clarification of some of the issues.

Bird wings, bird-wing model, bat wings, bat-wing model. 

These phrases are often used as short hand to refer to reconstructions in which 
the wing membranes are free from (birds) or attached to (bats) the hind limb. 
One can immediately see the problem. It sort of works at a very general level, 
but not in detail. Reconstructions of a narrow wing, with brachiopatagium free 
of the hind limbs, or a deep wing, with the brachiopatagium attached to the 
leg, and also with a uropatagium, clearly do not match the wings of birds, or 
bats when considered in detail (e.g. birds have a discontinuous flight surface 
of feathers, pterosaurs a continuous flight surface of membrane) and almost 
certainly did not function in exactly the same way. Numerous authors have made 
this point on many occasions, but we all continue to use, or rather abuse, 
these terms. Perhaps 'coupled' pertaining to a brachiopatagium attached to the 
hind limbs and 'uncoupled', pertaining to a free hind limb might be better?. 


In thrall to the bat wing paradigm.

It has been claimed that some if not most workers on pterosaurs, past and 
present, grew up with the so called bat wing model (see above) and remain 
permanently in thrall to it. A close reading of the literature shows that this 
is not true. Ever since pterosaurs were first discovered there has been a wide 
variety of opinion regarding wing structure with workers supporting both 
coupled and uncoupled models, and lots of variations on these. Moreover, early 
palaeobiologists like Abel, Broili, Arthaber, Wiman and others were clearly 
aware of these issues and gave detailed descriptions of fossil evidence and 
reasoned arguments based on functional and mechanical ideas to support their 
reconstructions: coupled and uncoupled. Certainly, in the last 20 years no one 
can have approached a fossil specimen without being acutely aware of and 
sensitive to the issues involved. Consequently, so far as I can tell, all 
recent accounts of wing shape and extent are genuine attempts to first descr!
ibe what's there and then to accont for it in terms of existing models. 


Were pterosaurs 'coupled' or 'uncoupled'?

All the wing membrane specimens I have seen in the last twenty years either 
show the brachiopatagium attaching to the hind limb, or its disposition in the 
region near the body is ambiguous. If one wanted to demonstrate that fore and 
hind limbs were uncoupled there are two problems to be solved:

First, one would need to demonstrate that all those specimens that apparently 
show attachment of the brachiopatagium to the leg have been misinterpreted. 
Peters, for example, has attempted this (1995, Wing shape in pterosaurs [reply 
to Unwin & Bakhurina 1994]. Nature 372, 315­316.). The difficulty with this and 
other attempts to explain away fossils such as Sordes is that they always lapse 
into special pleading: certain parts of the wings must have moved from one 
place and come to lie in another location that just happens to exactly match 
the position of the hind limb and, astonishingly, exactly the same happens on 
the other side of the body as well, even though for some reason these 
displacements don't affect other part of the wing membranes. OK, maybe there is 
a one in a million chance that this might have happened (Pratchett readers will 
know that this makes it inevitable that it will happen) once, but to have this 
happen repeatedly in different taxa from different locali!
ties with different styles of preervation? Not even special pleading has that 
kind of extensibility. 

Different wing shapes also present another problem for the uncoupled model. It 
is conceivable that a deep wing could be obscured by the leg, or get folded 
over, or shrink, or contract, so that it looks a bit like a narrow wing. But 
its much harder to go the other way. How can we get a narrow wing to look like 
a deep wing, especially since according to recent contributors to this debate 
the chord is 'Not variable. Sorry. The actinofibrils will not permit it'. 

Second, supporters (the supporter?) of the uncoupled model should present some 
clear evidence to show the brachiopatagium running to the body, and not 
attaching to, interfering with, or generally associated with the hind limb. All 
the specimens I have inspected over the last 20 years fall in the latter 
categories or are ambiguous. So, where is the fossil evidence for a narrow 
wing? Is anybody, anywhere willing or able to cite a single specimen in support 
of this idea? 


The Zittel wing

The rear edge of the cheiropatagium has been cleaned up at some stage by 
someone running a knife, or other sharp implement along it. So, the problem is 
how much was lost? One, two, five, ten millimetres? I suspect on the basis of 
looking at it for a few days using a microscope and comparing it with lots of 
other Rhamphorhynchus specimens that not much is missing. But I'm not sure, so 
I wouldn't want to use this specimen to argue for brachiopatagium shape in 
Rhamphorhynchus, especially as its an isolated wing. Why was it cleaned up in 
the first place? Obvious - because good looking fossils fetch more money, and 
thats what 19th century dealers worried about, not what we might happen to 
think 125 years later. This has happened to lots of specimens, and not just 
those from Solnhofen (Holzmaden ichthyosaurs for example), as I am sure readers 
of this list are only too aware. 


Dead wings.

The point being made here really is a very simple one. Dead wings should not be 
taken at face value (if they can then there is no question that Sordes and 
other taxa show the coupled condition). In some Messel bats (e.g. in the 1992 
Messel book edited by Schall and Ziegler, fig. 263) the wings are well 
preserved. Moreover, in this and other specimens it is possible to work out the 
basic arrangement of the membranes. Tellingly, for example, they show quite 
clearly that there was a uropatagium and that the cheiropatagium attached to 
the hind limbs. But, none of the fossils I have seen show what I presume was 
the true shape of the cheiropatagium. The chord, for example, is always much 
shorter than it was in real life. So, in this respect, the dead wings of Messel 
bats provide us with some important lessons regarding our understanding of 
pterosaur wings. 


The pterosaur track record. 

Pterosaur tracks are now known from more than 30 localities ranging from the 
late Middle Triassic to the Late Cretaceous, with concentrations in the Late 
Jurassic and Early Cretaceous. Thousands of prints have been found, especially 
in Crayssac (Late Jurassic, France) and in some of the new Early Cretaceous 
Spanish sites (should be published in the Buffetaut vol. in 2003). Most of 
these tracks appear to be made by pterodactyloids although some are claimed to 
be 'rhamphorhynchoid' tracks. All tracks found so far were made by pterosaurs 
proceeding in a quadrupedal plantigrade fashion. There are some bipedal tracks 
consisting of hand only prints which might be pterosaurs showing off (hey look 
at me, no feet) or more probably an under print phenomenon. However, so far as 
I am aware, there are no trackways showing pterosaur proceeding on their hind 
feet alone. 

As my colleagues and I discuss in our recent paper (Hwang et al. 2002), while 
tracks show many general features in common, the morphotypes recognised so far: 
Pteraichnus, 'the other Crayssac track type', Purbeckopus and Haenamichnus all 
show small, but distinct differences from one another. Peters (2000) has 
suggested that all pterosaur tracks were made by members of the 
Ctenochasmatidae. But, since there are some clear temporal and morphological 
incongruencies between some of the tracks and ctenochasmatids (e.g., 
Haenamichnus is Late Cretaceous and was made by a giant pterosaur, while 
ctenochasmatids are unknown from this interval and do not appear to have 
achieved more than 3-4m in wingspan) this idea seems extremely unlikely. Latest 
studies suggest that Haenamichnus was probably made by azhdarchids, while some 
of the Late Jurassic/Early Cretaceous tracks were possibly made by 
ctenochasmatoids and perhaps also by dsungaripteroids. Purbeckopus might be 
ornithocheiroid, thou!
gh I really have doubts about thi. 

Taking all the evidence together (skeletal anatomy, wing membrane disposition, 
computer based models, the track record) there is no evidence to suggest that 
any pterosaur habitually preceded in bipedal fashion. Certainly, they may have 
been able to stand for brief periods on their feet alone, but this was probably 
highly unstable irrespective of how one orients the spinal column, and they 
either fell over, took off, or resorted to all fours. What I find particularly 
compelling is the consistency between the track record and other studies based 
on skeletal anatomy, computer models and so on. Might I add that Sarah 
Sangster, who recently finished her PhD on Dimorphodon, also came to the same 
conclusions as other recent studies on this pterosaur (e.g. Clark et al. 1998, 
Unwin 1988) that it was a quadrupedal plantigrade. The idea that some groups 
were habitually bipedal, but never left any tracks, is fun to speculate on but 
doesn't need to be taken seriously until some real evid!
ence turns up. 

I'm done.

Dave 

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
Institut fur Palaontologie, MUSEUM FUR NATURKUNDE 
Zentralinstitut der Humboldt-Universitat zu Berlin
Invalidenstrasse 43, D-10115 Berlin, GERMANY

Email: david.unwin@rz.hu-berlin.de

Telephone numbers:
0049 30 2093 8577 (office)
0049 30 2093 8862 (department secretary)
0049 30 2093 8868 (fax)
xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx