[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Subject Index][Author Index]
Re: Pterosaur take-off movie on the NG site
David Peters wrote:
Actually, no fossil shows the inboard wing clearly at all;
Not true. See Peters 2002. Unfortunately, no one else since then has
shown any different.
Actually, a *lot* of people have shown differently in the literature.
Just for starters: Wild 1993; Unwin and Bakurhina 1994; Wang et al.
2002; Lü 2002; Bakhurina and Unwin 2003; Frey et al. 2003 and Bennett
Now, you may disagree with their arguments, and that's fine, but to
argue that the situation is clear and obvious is a bigger stretch
altogether. If you are so confident that your interpretation is
correct, while the above authors are confident in their
interpretations, then at the very least, the mesopatagial structure is
not obvious. It must be, at minimum, contentious.
the ankle/tibia attachment is a plausible reconstruction that
several authors currently support.
IMHO twisting evidence to achieve an old paradigm. Several authors
also support the weird pteroid placement and plantigrady for all
pterosaurs. Evidence works, popularity without evidence doesn't have
the same effect.
There is no need to draw ad hominem or political straws here. I'm
merely stating that evidence has been presented in the literature for
a broader inboard wing. Evidence is also very strong for a narrow
actinopatagium, and from a flight mechanics standpoint, that is rather
more important. The presence/absence of a hind limb attachment for
the wing mostly matters for certain aspects of anchoring and inboard
flutter control (i.e. how flutter is reduced - there are ways to do so
under either attachment reconstruction). Furthermore, keep in mind
that the original complaint was that Raul used a hind limb attachment
for the wing. Given that many authors (most, in fact) support this
model, it is completely reasonable for a paleoartist, searching the
literature, to use such an attachment. The fact that your manuscript
argues a different model need not change his mind (although I would be
happy to see someone illustrate your version, too).
Where it goes from there is a matter of debate, because the more
elastic mesopatagium is generally a mess (or missing) in most
specimens, even those that have a nice outboard wing preservation.
List a specimen or two and I will show you that it's not a mess.
Well, to really do this, we would need to be in the same room, with
the specimen. This was the case in Munich, which actually convinced
me of a hind limb attachment (of some sort) in many specimens (though
it by no means suggests that this was the case for all taxa). It also
demonstrated just how difficult it actually is to follow the inboard
wing on even the best specimens. I know that you can supply photo
outlines that seem suggestive of a more narrow wing. I do not find
them conclusive, in and of themselves. Some of Helmut Tischlinger's
UV photo work really seems to indicate that there is some mesopatagium
lateral to the tibia in some of the specimens. I find his work to be
somewhat more convincing, at this time.
Other positions can also be hypothesized, and also work. We don't
know what the "ordinary" hindlimb posture is for any large
pterodactyloid. The more vertical posture you prefer is quite
plausible; it is not the only plausible posture, however, nor does
it actually affect the overall launch sequence: if the starting
posture is more vertical, then the animal can simply shift weight
forward into a more horizontal posture first. This adds a small
amount of time to the overall sequence. It is not of major
Which is why it stays in the "minor" category. If big pteros were
different than little pteros than there should be some morphological
marker that shows up in a cladogram to flag this change. I don't see
it. Can you show it?
Why should there be some single "marker" in this manner? Stance and
gait are the result of several different proportions. Furthermore,
these proportions need not split according to size. If you wish to
map gait and stance on a cladogram, then the traits of interest would
be things like limb length, the relative lengths of each element in
the limbs, robustness of limb elements, and the breadth, length, and
depth of the thorax and abdomen. And yes, any one of us could show
that these features differ between major clades. Reconstructing the
effect on gait is something altogether, though I think it should be
fairly obvious that something like an azhdarchid didn't walk and run
the same way as an ornithocheirid.
They were probably there, yes - but we worked directly from the
AMNH 22555 specimen, which doesn't have them.
It also doesn't have several other bones, which you added. You
really need to add the prepubes.
I didn't add them - they had been estimated by the folks at the AMNH
already. They didn't bother to estimate the prepubes. I didn't add
them because it really isn't relevant to launch. Maybe some other time.
We don't know what the gait was in Anhanguera. There are multiple
possible gaits. One of them is the near-bipedal gait you
describe. A more thoroughly quad gait is also plausible, and is
what we used. Contrary to your assertion, it is possible for the
forelimbs to provide walking power in Anhanguera, albeit not a
great deal of power because of their position. I agree it was
likely airborne most of the time, and I would not expect it to be
capable of a canter or gallop, as I suspect that most azhdarchids
We can eliminate certain gaits because they don't provide propulsion.
Perhaps - have you done a biomechanical power and propulsion analysis
to constrain the potential ground-walking gaits? I've done some, but
not enough yet for publication.
Two ornithocheirid tracks are known:
1. Purbeckopus pentadactylus (DORCM G 6664; Delair, 1963; Wright et
al. 1997) an early Cretaceous ichnite.
2. TATE 0049-06 (Southwell and Conneley, 1997) a late Jurassic
I can send you pdfs if you don't already have them. Neither has the
single deep indention you describe. Just enough room for three
They might not be ornithocheirids, actually, but putting that aside:
note that they are not "hollowed" at the base of the fingers. I said
explicitly that it would not be a single, deep indentation. Instead,
the mcIV/phIV joint impression is apparent as the middle of the manus
track. The bases of the "free" fingers are confluent with the sides
of MCIV. Their bases cannot touch the substrate. Therefore, the
"thing in the middle" is not the base of the free fingers. It is the
MCIV/PhIV impression. It need not be especially deep. I think you
presume that more details of weight support are evident in tracks than
is usually the case. Sometimes a weight shift can be seen, but the
tracks need to be especially well preserved, and substantial weight
shifting needs to occur. Tracks are often counter-intuitive,
especially if the substrate was deep and soft (see Steve Gatesy's work
with avian tracks). There is a deepening of the manus impression in
some East Coast tracks (as yet unpublished, so I won't say more on the
identity), in what appears to be a launch sequence. That's not
surprising, as it would amount to about 2 or 3 times body weight,
which is a substantial increase and can be see in the ichnite with
Okay, let's say your resistance and recoil hypothesis worked on
hard, unyielding surfaces. Would it work on soft mud or wet sand at
all scales for all pterosaurs? Or would the substrate slip? And
along with it, the wing finger under tension?
Good questions. It would work on most substrates for most species.
As such, there would be some taxa and/or some surfaces that would be
harder to launch from than others. Again, the catapult mechanism is
not required, it just helps.
Even presuming, for a moment, that this is the case, I can say with
confidence that digits 1-3 did not take much of the weight not
accommodated by the hind limbs, as these small digits are not
structurally capable of doing so.
Show me some pterosaur tracks in which digits I-III are -not- taking
the weight. They -are- structural capable. Ichnites show it.
Not true; see above. You cannot glean as much about weight support
from the tracks as you might think, though you can see a fair bit.
Again, if MCIV/PhIV was not contacting the ground, then the tracks
would have a more pronounced concave crescent. This applies to
essentially any trackway at all. But just for starters, let's look at
the very nice trackway just published by Mazin et al. (the possible
landing track). Look at figure 1b: note that the manus prints are all
*convex*, not concave. If you were correct, the tracks would lack any
impression in the area of MCPhIV joint, but there is a very nice
impression of it in each track. Keep in mind that digits 1-3 are
rooted, as it were, to MCIV. Thus, the area in the "middle" of each
track where the fingers seem to come together must be MCIV. It is
leaving an impression, and a rather large one, actually. As for
structural capacity, a quantitative analysis of bending strength is
more informative. My quick calculation indicates failure for the
small digits in a large azhdarchid trying to stand on them alone. Of
course, this is just illustrative, because what would actually happen
is that the animal would simply sink down onto the MCPhIV joint, given
that the fingers are not single rigid structures.
I also note that articulating many specimens, especially large
pterodactyloids, into a walking posture forces the MCPhIV joint to
touch the substrate. Personally, I have been unable to find any
articulation in which digits 1-3 touch the ground and the MCPhIV joint
does not. They all touch together, except in things like Nyctosaurus,
which just seems to walk on the hind limbs and MCPhIV.
It's also a coincidence that the pteroid fits into the cup of the
preaxial carpal and look at what trouble that has brought to those
who move it there in reconstructions. Mike, I am following your
logic. In the end though, there's too much trouble for me to accept
this "click beetle"-like launch mechanism.
I don't see any trouble. If you wish to refute it, however, then I am
all ears. However, you will need to supply stronger evidence.
Preferably a quantitative, biomechanical analysis of structural load
capacities. If nothing else, you would need to demonstrate that an
ornithocheirid cannot place the MCPhIV joint on the ground. I argue
that it did so nearly all the time, and that the free fingers in some
taxa could barely touch the substrate. I make this argument based on
finger proportions and trackways. However, even if my assertion is
incorrect in this regard, the catapult storage mechanism is still
plausible and consistent with morphology so long as the tendon *can*
be compressed, even if it was not locked during walking.
I appreciate that the numbers are working on one side and not the
other. Not sure what the solution is if not gooney-bird- or pelican-
Well, I've given a plausible alternative. Actually, there is really
nothing in any pterosaur suggestive of a gooney-bird or pelican-like
launch. We have no reason to assume such a gait, and it turns out to
be impossible for all large pterodactyloids. In fact, all of the
morphological traits we would presume to exist in a bipedal, running-
launching species are not only absent in large pterodactyloids, they
trend in the *opposite* direction of what we see in water-launching
birds. The femora in large pterosaurs are proportional more gracile
than in small species, not more robust. The feet in many of the
largest pterosaurs are small, not large. The proximal forelimb is
robust, not gracile. The list goes on for a while.
First of all, not all animals leap with the position you suggest.
In fact, vampire bats, the living quad launchers, don't do so. But
more to the point:
Figure 3 of The Journal of Experimental Biology 200, 3003–3012
(1997) THE DYNAMICS OF FLIGHT-INITIATING JUMPS IN THE COMMON VAMPIRE
BAT DESMODUS ROTUNDUS shows this. Figure 2 stops short of becoming
airborne. And the thumb tip is the last element to leave the
ground. I also note that the humerus is more than half the length of
the ulna and at maximum flexion the humerus and ulna are at 90º to
one another in front view. This is where the leverage is in
Desmodus. In some pterosaurs there are similar proportions. In
others the humerus is much smaller and therein lies trouble.
Yes, the limbs extend and the thumb is last to leave the ground, but
the leap is not frog-like, either, which was the gist of my earlier
comment. In any case, much of the leverage in Desmodus actually comes
from the thumb. There is some from the elbow and wrist, too, of
course. While the longer humerus does increase the excursion length,
it also limits the maximum stress that the humerus can take. The
shorter humerus and antebrachium of pterosaurs gives them much greater
power limits about the elbow. The excursion length of the limb is
then increased by the long MCIV (relative to the bat). So, in both
cases, there is a shorter, stouter element that sees a lot of stress,
and some longer elements that help add excursion length. It's just
that which elements are which differ between the bat and the
pterosaurs. So once again, no trouble.
1) the wing finger *is* forcefully extended - because of its shape,
only the proximal end (the flexor tubercle itself, essentially)
actually hits the ground. The rest of the finger does not "lay
flat", as it were.
That doesn't show in the animation. If something is indeed
forcefully extended it needs to be extended to the limits of its
joint mobility. Not held back.
I already explained this: the limit of extension, in this case, is
when the tubercle contacts the substrate. Incidentally, it is also
not true that anything forcefully extended becomes ballistic in nature
and goes to the limit of joint mobility. Actually, it almost never
happens - antagonists usually fire even in powerful motions. For
example, a person throwing a kick does not (generally) fully lock
their knee. For more technical treatments, look at the myological
literature - there are a lot of nice electromyograph papers out there.
2) the long flexors for the fingers also cross the wrist. Much of
the stored power goes into mechanically extending (anatomically
flexing) the wrist.
You say that, but the animation doesn't show any flexing of the
wrist prior to launch. Only extension after launch.
Well, if it extends, then there is the power. We show a bit of
flexing prior to launch, but not a great deal, as we have it starting
in a highly flexed position.
I'm looking forward to seeing the evidence for implantation of digit
IV in the substrate. Shoot it my way when you can. That's key. If
you can't provide it, there may indeed be problems.
See above. Again, there are multiple ways this can work, which are
all feasible. By contrast, I have yet to see any positive evidence
for a biped launch, and I can supply a lot of negative evidence. The
femur snapping in large species is a pretty good one, just for starters.
Best regards, and congratulations on all the great PR.
Thanks very much; I appreciate it.
Assistant Professor of Biology
Woodland Road, Pittsburgh PA 15232
Buhl Hall, Room 226A