Re: pterosaur femora sprawl

[Michael Habib <mhabib5@jhmi.edu>]

> "Why" questions are difficult to answer. But the dual use of pterosaur 
> forelimbs was secondary, as demonstrated by Sharovipteryx and 
> Longisquama.

Assuming those are the outgroups, then yes, it could be that quad gait 
is secondary in pterosaurs.  However, it would be a secondary gait that 
shows up right at the basal node.

> The dual use of bird forelimbs (think Archaeopteryx and Hoatzin) was 
> secondary as demonstrated by Coelophysis and kin.

Hoatzins are derived neornithine birds, so they're not really 
informative for the basal braches of Aves (though they are very 
interesting functionally).  Archaeopteryx doesn't actually tell us the 
ordering of the forelimb functional characters - dual use in feeding 
and climbing may have predated use in flight.  Or, the forelimbs might 
have been feeding/flying implements first, and climbing was rather 
unimportant (or became useful later).  We don't really know.  In all 
likelihood, the forelimbs of basal birds were, as you suggest, 
secondarily dual-use as least in terms of flight.  However, birds may 
be the exception rather than the rule.

> Until I see an example of the kinematics of quadrupedal launching from 
> a horizontal surface, I can't envision it. Not when pterosaur 
> ancestors and the first pterosaurs were incapable of touching the 
> ground with their forelimbs (while balancing glenoids over toes) -- 
> and they had big thighs with a pelvis 1/3 the torso length.

I don't know of any pterosaurs that were incapable of touching the 
ground with their forelimbs.  Not a single pterosaur specimen has 
longer hindlimbs than the walking portion of the forelimb to the best 
of my knowledge.  Even if we prefer the walking position that you're 
describing (and I prefer a slight different one, as do many others), 
and bring the forelimbs up off the ground in a few basal taxa, the 
animals can still rock forward to reach the ground for quad launch.  
Sure, they have good sized thighs - they still have bigger 
chests/forelimbs.

As for being able to envision the launch kinematic, that's really just 
an issue with visualization, and is one reason that I am happy to be an 
illustrator on the side (and happen to know other illustrators superior 
to myself).  So, I'll work up the images as time permits.  That said, 
remember that similar "envisioning" issues could be just as easily 
raised for bipedal launch.  Bipedal launching is not the "default 
state", nor the null model.  There is no null model for launch in 
vertebrates - we have to examine the mechanics of the animals in detail 
to determine which launch mechanisms are most likely.  The fact that 
bipedal launch is the standard in the literature is simple a by-product 
of thinking that pterosaur should operate like birds, when pterosaurs 
are, in reality, much more unique.  I have a hard time "envisioning" 
bipedal launch because many pterosaur species can't get off the ground 
that way without special conditions.

> Certainly Anhanguera had a different launch mechanism, with the 
> largest forelimbs and smallest hind limbs and feet in pterosaurland. 
> But you're bringing up a _very_derived taxon. This is a taxon that, 
> like an albatross, rarely landed.

Nope, not really.  The specific kinematic is a bit different, sure, but 
it's still a quad launch.  Incidentally, while I used Anhanguera as my 
example, similar ratios are found in azhdarchids and dsungeripterids, 
as well as pteranodontids.  Thus, it's basal to at least most 
pterodactyloids.  I'm working on rhamph data now, and the trend holds - 
the ratios are not as extreme in the smaller bodied animals, just as 
one would expect, but they still don't follow an avian trend.

> Not sure (can't envision) that a quadrupedal launch would work here 
> either. A stiff breeze might have been all that was necessary. And for 
> that: a bipedal platform.

Again, why is a bipedal launch easier to "envision"?  If it's just a 
matter of what we're used to seeing, then that's not really a good test 
of the model, is it?  And yes, gusts are handy - quad launchers can use 
gusts, too.  Being reliant on special wind conditions all the time, 
however, is a poor way to make a living.  You seem to associate bipedal 
platforms with some kind of launch advantage.  In reality, that is the 
primary cost of bipedal launch systems.  Birds are much more 
constrained in many ways, especially with regards to size, than 
pterosaurs ever were.  However, birds get one big advantage: bipedal 
platforms provide decoupled locomotor modules, so limb use disparity 
can be very extensive and functional diversity is very plastic.  Thus, 
birds have produced things like loons and grebes, that dive with the 
hindlimbs, as well as long-legged cursors like secretary birds, all the 
way to amphibious flyers like puffins.  There is huge range of 
functional  diversity in birds: divers, perchers, clingers, runners, 
soarers and hypercarnivores.  Pterosaurs, by contrast, were more like 
"variations on a theme".

> The beachcombing taxa, yes. No prints yet for the soarers, skimmers, 
> insect-eaters, basal forms.

The supposed beachcombing track-ways could just as easily be from 
soarers and skimmers.  They walk around on beaches as well.  The fact 
is that a range of pterosaurs from small to large definitely walked 
around on the ground quadrupedally.  That means that an individual 
pterosaur would have to shift gait to a bipedal stance to launch using 
the hindlimbs only, at the cost of losing most of its launch power 
potential, and gaining virtually nothing in return.

> False paradigm based on beachcoming taxa: those with short fingers, 
> small unguals, plantigrade pedes, small fifth toes. Not basal. Not the 
> majority.

But long fingers, large unguals, and a more digitigrade stance would 
hardly negatively affect quad launching.  And, quite frankly, given the 
pterodactyloid diversity, we could consider it the majority.

> Why do you take the odd conviction that quad gaits were basal for the 
> group when the best candidates for pterosaur outgroups were bipedal 
> (Sharovipteryx, basal dinosaurs, Scleromochlus)? That's an irony that 
> you quad guys are overlooking.

I'm not overlooking it, I'm arguing that, if the ancestry was indeed 
bipedal, then pterosaurs were secondarily quadrupedal.  It's either 
basal to the group or a secondary adaptation - the fact that the 
outgroup might be bipedal does not turn pterosaurs into bipeds any more 
than the possibility of basal dinosauromorphs being quadrupeds makes 
basal dinosaurs have to be quadrupedal.  We know that the most basal 
dinosaurs on record were bipeds - if the outgroup turns out to have 
been quadrupedal, then that just means there's a secondary gait 
acquisition near the root.  Same thing for pterosaurs.

> Never discount phylogenetics. It's the key you're trying to ignore 
> because it doesn't fit your paradigm. Shake off all prejudices and let 
> the phylogenetic evidence guide you. It's real. It's not made up by 
> engineers.

I never do discount phylogenetics.  In fact, my primary training is in 
comparative methods.  However, some questions simply are not answered 
phylogenetically.  In particular, phlylogenetic brackets don't predict 
derived states - they can give you ancestral reconstructions to 
polarize what *is* derived, but you seem to be arguing that we can 
reconstruct pterosaur locomotion using the state at the root node.  
That is not an effective methodology in this particular case.

I should also add that the phylogenies jive just fine with the 
"paradigm" I am proposing: pterosaurs could be secondarily quadrupedal 
animals that evolved from a bipedal ancestry (in which case the 
hindlimb was expanded very little, but the forelimb was expanded a lot) 
or they may have inherited quad gaits from their ancestors, in which 
case the hindlimbs were expanded slightly more, and the forelimbs were 
expanded slightly less (relative to the outgroup).  Those working on 
the trees (like yourself) hold the keys to answering those questions of 
sequence and ancestry.  That's not the same as reconstructing the 
specific, derived mechanical abilities of the ingroup, however, which 
is a different sort of problem.

Oh, and the jab at engineers really isn't appropriate - they aren't 
making up the mechanical relationships.  Colin Pennycuick once noted 
(in one of his excellent books on flight) that many biologists think of 
aerodynamics and mechanics as very theoretical.  In reality, they are 
constrained by known rules of physics, and thus are much *less* 
uncertain than things like paleoecological inference.  The biological 
conclusions are not as constrained or certain, but the mechanics are 
quantitative.  Pterosaurs were simply better built for quad launch than 
bipedal launch - that doesn't make it certain that all pterosaurs used 
quad launches, but I see no reason at present to assume a bipedal 
launch for any them (except as an occasional, facultative dynamic used 
by smaller species).

> I agree that mechanics can give answers, but workers who have provided 
> us with robo-pterosaurs have not included options exploring other 
> possibilities (and in the rare instances where that has happened, the 
> options were sabotaged with poor mirroring of nature).

Well, I'm not trying to build any robots; I just calculated the load 
potentials for the skeletal spars.  I've layered on some more 
qualitative analysis of things like mechanical bracing and muscle 
fractions.  In all cases, quad launch is supported.

> Just as in the pteroid orientation question, whatever you come up with 
> has to work (albeit modified) for all pterosaurs.

Well, technically speaking, a proposed dynamic doesn't have to work for 
them all (there can be diversity in behavior).  However, in this case 
the model does work for all known taxa (at least so far).

>  I brought up short-legged, long-handed Nyctosaurus because it would 
> seem it was operating with oversize "crutches" that would not have 
> been able to be angled at a posterior vector for a quad launch that 
> sent the pterosaur forward.

They can be; it's not in a passively stable position, of course, but 
that's fine.  The quad launch dynamic includes a phase of pre-load, 
with the animal in a transient 'unbalanced' phase (just as many animals 
are when leaping).  Think of the early phase as more of a vault, and 
less of a posterior push.  A launching pterosaur would launch by first 
pushing onto the forelimbs, essentially vaulting over them into 
preload, then unloading the forelimbs before the vault was completed.

> Silver bullets: Pedal digit V. Sternal complex. Aktinofibrils. 
> Antorbital fenestra without a fossa. Large manual digit IV, short V. 
> Extended ilium. Prepubis. Chevrons parallel to centra. All these are 
> found nowhere else (except some characters are found in Proterosuchus, 
> easily convergent).

Very interesting; I look forward to seeing the full datasets and 
topology.  Incidentally, if they are easily convergent in at least one 
group, then I wouldn't call them "silver bullets".  Compelling, 
perhaps, but not a slam dunk.  Still, sounds like a neat character list.

> Finding a single tree in a cladogram with 25,000 matrix boxes 
> indicates that lots of possibilities were looked at and only one 
> solution was indicated.

True, but it doesn't say that the single tree recovered is any more 
correct than the 500 MPT's that someone else retrieves.  After all, you 
can't possibly have *the* tree, because we don't have all the taxa (and 
probably never will).

> BTW, convergence occurs in 90% of my characters and often more than 
> twice.

Very interesting - what characters are most often convergent?

> Thank goodness for computers.

No kidding; I'll second that.

Cheers,

--Mike


Michael Habib, M.S.
PhD. Candidate
Center for Functional Anatomy and Evolution
Johns Hopkins School of Medicine
1830 E. Monument Street
Baltimore, MD 21205
(443) 280 0181
habib@jhmi.edu