Re: pterosaur femora sprawl

[Michael Habib <mhabib5@jhmi.edu>]

> Since most bats typically neither run nor walk, perhaps they (vampires 
> secondarily excepted) basically have no gait.

Interesting thought.  I'd say that they have a climbing gait, though (I 
forgot to include that before in the definition of possible gaits, my 
fault).  Pteropodids, for example, are essentially quadrupedal 
suspensory climbers, and that should count.  But yes, bats are odd.  
That's why they're fun to work on!

> No competing studies include civets and Ptilocercus. That's why 
> they're all running into brick walls. What if blood tests differ? 
> Well, let's put it this way, we'll follow all the clues and use 
> parsimony. DNA tests are contradictory with regard to reptile 
> relations. So, dna tests are not perfect.

Absolutely, molecular trees have their own problems.  I'm just 
cautioning against firm statements where they are not yet warranted - 
you originally stated that civets and bats are close relatives, as if 
it were a certain relationship.  What you meant, of course, is that 
your tree places them there, specifically.  You might be right on 
target, but I'd also suggest being more open to the uncertainty of the 
situation.  That's just par the course for phylogenetics.

> If you're using your hands pressed against the substrate to locomote, 
> your hands have no opportunity to become oversized, foldable flying 
> organs.

Why is that?  After all, the rules of macroevolutionary trends are not 
the same as individual daily-use.  Seeing as how bats and pterosaurs 
have/had dual-use forelimbs, how can we be certain that the limbs 
didn't evolve that way in the first place?

> > True, a small pterosaur could facultatively launch bipedally.  
> > However,
> > I see no reason to expect that this was their normal mode of launch.
>
> If primitive pterosaurs were small and launched bipedally, that, by 
> definition is the normal mode. Anything deviating from that norm is 
> derived.

Ah, but I said that they could *facultatively* launch bipedally.  That 
is, quad launch was probably the norm for small, primitive pterosaurs, 
but they were not limited to using only a quad launch simply because 
their required launch speeds were not very high.  I was referring to 
the "normal" mode of launch in the sense of day to day activity.  That 
is, a given small-bodied, early pterosaur would normally launch with 
all four limbs, but might sometimes use only the hindlimbs if need be.

> Please, Mike, this I have to see. I will even animate it, once I have 
> the stages involved as I have animated walking and landings before. 
> I'm fascinated by the possibility and eager to explore your theories, 
> but if I see a reason it can't or has trouble working, I'll let you 
> know.

I'm working on it; don't want to do a sloppy job, after all.  Feedback 
always appreciated.

> Pterosaurs have big thighs, great for launches. But vampires use their 
> forelimbs, so let's see which model is more parsimonious. Don't keep 
> me in the dark.

Yes pterosaurs have reasonably large thigh muscles, but they have much 
larger muscles associated with the forelimb and pectoral girdle.  The 
bone strengths are also vastly different in the forelimb and hindlimb, 
which further supports this trend.  The ratio of bone strength between 
the humerus and femur (strength in bending, that is) for Anhanguera is 
56 times the same ratio for an albatross.  If pterosaurs were bipedal 
launchers, like birds, then the ratio should actually be smaller in 
Anhanguera (given it is larger than an albatross).  Remember, the 
hindlimbs are still involved in the quad launch, they just generate 
less power than the forelimbs, and the forelimbs leave the ground last.

Also, vampires have reasonably large thighs and strong femora relative 
to other bats.  Their femora are not the strongest among bats, which 
indicates that hindlimb strength is not the only critical factor in 
terrestrial mobility for bats (see Riskin), but several studies 
(including my own dataset) show that vampires actually have elevated 
strength in the femora as well as their humeri.  However, the humeri 
are still much stronger than the femora - pterosaurs match this trend 
(they have good, strong hindlimbs, but much stronger forelimbs).

In terms of parsimony, quad launching does not require pterosaurs to 
change gaits for launch - we know they walked quadrupedally, so 
switching to a bipedal launch dynamic is actually less parimonious.  
Phylogenetically, it's still uncertain - there might have been an early 
shift to quad walking (and, as Jim and I suggest, quad launching), 
assuming the ancestors were bipeds, or quad gaits might be basal for 
the group.  Either way, it's a single, fairly simple shift on the tree, 
so phylogenetic parsimony is probably not the best way to judge the 
likelihood of the proposed launch system.  It's really a mechanics 
problem - we can detect what the animal's were capable of (and their 
likely locomotor behavior) using structural mechanics.  Of course, we 
need the phylogenies to determine things like timing and convergence.

> > Nyctosaurus too?

I suspect that Nyctosaurus quad launched, yes, though with rather less 
power and speed than critters like azhdarchids (which had power-house 
builds).

> Please send some drawings whenever you can. Jim has had the same 
> request for awhile. And what would it take to get you to be convinced 
> that there are no known closer relatives to pterosaurs than 
> Sharovipteryx and kin?

There will be illustrations made available.  In terms of pterosaur 
sister-taxa, I'd be convinced by the typical means - if the proposed 
sister group (Sharovipteryx and allies) repeatedly falls out next to 
pterosaurs within the analyses of various authors using a variety of 
character sets.  Quantitative support systems also help, of course.  
And, of course, the characters in question can make a difference, but 
this can be very group specific - synapomorphies like feathers can be 
unusually compelling, for example.  Hard to say what this would be for 
pterosaurs, but it's a secondary issue (I don't require any sort of 
silver bullet, nor expect one).

> Bats? or Pterosaurs? In either case I've used heuristic and bootstrap 
> algorithms with high 90s and 80s except where taxa are known by skulls 
> only or without skulls. I employ various decay analyses, deleting 
> taxa, characters, randomly, non randomly all with confirming results.

Sounds great; I look forward to the upcoming publication(s).  Out of 
curiosity, what are your strongest and weakest clades in terms of 
bootstrap and decay analyses?

>  Unlike prior workers, I get a single tree. One in which sisters look 
> alike. Speaking of which, no prior pterosaur analyses discuss the 
> issues you raise.

Finding a single tree is convenient, but actually doesn't say much 
about accuracy one way or the other.  Sisters should "look alike" from 
a character standpoint by definition - I assume you mean in a broad, 
qualitative sense.  That's interesting, as it suggests low levels of 
predicted convergence, but it also doesn't say much about accuracy, per 
se.  It could mean, for example, that some homoplasy is being missed.


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