[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Subject Index][Author Index]

Re: flapping from gliding (long)

First of all, many thanks to Achut Reddy for providing some meaty material to
chew upon, unlike some of the stuff I've responded to recently here. My
replies are bracketed between hyphen rows for ease of reading. 

[I don't find this at all easier to follow.  I frequently reformat
 messages in order to make them look like clean versions of the
 standards most commonly used in USENET articles.  In fact I have to
 reformat almost all of George's messages because he's such a
 non-conformist.  However, I'll be danged if I'm going to pull this
 apart and put it back together.  If any of you have opinions on
 George's idea for what makes reading easy, please let me know.  -- MR ]

He writes on
96-10-10 17:13:55 EDT:

<<Dinogeorge writes:

>It is "conceptually easy" because the force of gravity pulls downward, not
>upward; the energy keeping a gliding creature aloft comes from its fall
>through the earth's gravitational field (in still air, that is). Arboreal
>other acronomic animals require mechanisms to keep themselves from injury
>from falls, and every arboreal theory of the origin of flight starts out
>this point. Cursorial flight-origin theories, however, invariably work
>_against_ gravity; there is no way to keep the animal aloft unless it
>has the very wings you're trying to get it to evolve.

The "force of gravity" is a red herring.  It pulls downward with the
same force on _everything_ regardless of whether you are cursorial or
arboreal, and the same force must be overcome in all cases.  BTW, cursorial
animals also have to protect themselves from falling injuries (remember
the recent paper about tripping being fatal for a running T. rex?)>>

The force of gravity is not a red herring at all; it is central to the thesis
that powered flight evolved from the trees down rather than from the ground
up. An acronomic animal "overcomes" gravity when it climbs into a tree or up
a cliff, a feat for which all it needs, perhaps, is small size, sharp claws
(unless it is a tree-climbing snake), and a bit of patience. For a large,
cursorial, bipedal animal to evolve into a powered flier >without at any time
in its evolutionary history climbing into the trees and becoming arboreal,<
and >without at any time becoming a gliding animal,< requires far too large a
suite of individually improbable evolutionary events, occurring in sequence,
to be even remotely possible. I'll elaborate below.

<<I believe the real argument you are trying to make is that gliders
already have the necessary equipment for powered flight, and that
cursorial animals cannot possibly have them since wings are of
no use unless you are already in the air.  Of course, neither one
is necessarily true.>>


First of all, in the business of scenario-building, >nothing< is >necessarily
true< or >necessarily false<. Every scenario has counterexamples, just as
every scenario is theoretically possible. What we have to do is examine the
scenarios to see which ones are more likely than others. It is theoretically
possible for a quadrupedal animal to evolve true wings on its back and use
them to fly, but just how likely is this to happen? Evolution occurs by
incremental, naturally selected improvements to pre-existing anatomical
structures. A quadruped will more likely change one or both sets of its
appendages into wings, rather than evolve a new set of appendages from its
back. Most gliding animals use body parts for gliding that are extremely
unlikely to evolve into appendages suitable for powered flight, so the
evolution of powered flight, >even from an already gliding animal,< is
unlikely. But for the forelimbs of a bipedal, cursorial animal to transform
into fully feathered wings suitable for powered flight, without first
undergoing a stage in which the wings are used for unpowered flight, is
magic, not evolution.

Consider what has to happen for, say, _Coelophysis_ to change into an
_Archaeopteryx_: (1) it has to evolve to approximately one-sixth its former
size and 1/200 of its former weight (not impossible, but less likely than
evolving into a form six times as large: Cope's "Rule"); (2) the hallux on
the foot has to increase in size and acquire appropriate musculature to for
use in perching--without interfering with the animal's cursorial ability (not
impossible, but far less likely than simply becoming even more reduced to
improve cursoriality, as it did in practically all known theropod lineages);
(3) it has to evolve feathers, or, if perchance it already had them, to
evolve flight feathers on its hands and all the specialized musculature that
moves and spreads these feathers (how likely is this in a non-flying
animal?); (4) it has to lose the fourth digit of the hand while
simultaneously lengthening the penultimate phalanges of the first three
digits (what kind of grasping function would this serve?); and (5) it has to
lose most of the chevrons in the tail (all known non-avian theropods retained
the chevrons; why would _Coelophysis_ be different?). It is possible to
contrive scenarios in which these things occur, just as it's possible to
morph Roseanne into Madonna. But how likely is it that such scenarios will
occur in real life, under conditions in the wild?

<<First, yes, gliders have something resembling a wing, but it takes
a lot more than a flap of skin to achieve true powered flight. 
Such as: a strong breastbone for flight muscle attachment,
shoulder support, great muscle coordination, high intelligence, 
high metabolism, lightweight construction, good vision, etc., etc.
A glider would have to evolve all of these things in much the
same way a cursorial animal would.  Except, the cursorial flight-origin
candidates already have a head start on most of these things.>>


No known theropod, not even _Archaeopteryx_, had a "strong breastbone" for
any reason, let alone "for flight muscle attachment." "High metabolism" is
too vague a feature to argue with, besides which, several decades of debate
have shown that we do >not< yet have a handle on dinosaur metabolism.
Likewise "high intelligence" and "shoulder support": too vague to argue with.
Insofar as some of the other characters in your list were sporadically
present in a few theropods, BCF asserts that they were there because
theropods >descended from< gliding and flying forms, so their presence in
vestigial form supports neither ground-up nor trees-down hypotheses.

Arboreal animals with a lifestyle of leaping among tree branches, using their
hind limbs for propulsion and their forelimbs to catch onto tree surfaces,
will >naturally< evolve excellent hand-eye coordination and vision, and
strong forelimb musculature and "shoulder support." It is >very easy< to
understand how they might develop "high intelligence" (whatever that means),
lightweight construction, oversized claws, retroverted hallux, elongated
hands and forelimbs, and so forth. These are all such clear, obvious, and
>most important< AVAILABLE improvements to an arboreal lifestyle that I find
it extremely difficult to see how anyone could argue otherwise. The initial
appearance of feathers might be a fluke, but once they appeared, they would
have been of obvious utility to arboreal animals for solving the Falling
Problem and trajectory control.

<<Second, there can be other uses for "proto-wings" (wing-like
structures that are not yet capable of fully powered flight).
For example:  insulation, using them as a net to catch prey (Ostrom),
sexual display, using them as "spoilers" for ground-effects
to aid in tight ground maneuvering, using them to gain additional
height in hopping or jumping, etc.  None of these require being
airborne before the benefits can be realized.>>


Again: why is it so difficult to envision the evolution of wings >for flight<
to begin with, instead of envisioning all kinds of unlikely prior uses for
"proto-wings"? How many extant flightless birds can you cite that presently
use their rudimentary wings for insulation? Prey nets? Gaining additional
height (more than a few inches) in hopping? If "proto-wings" had such uses,
then one must wonder why most flightless birds lose their wings almost
completely instead of having them reacquire some of their former functions.
Why is it that, when flightlessness evolves in a formerly volant lineage, the
wings so often vestigialize? Surely if wings had other, non-flight uses to
bipedal, cursorial animals, such uses would return when flightlessness

<<BTW, _Exocoetidae_ (flying fish) is another counter-example to
your gravity argument.  It has to overcome not only gravity,
but also the viscosity of the water in order to become airborne.
But I have never seen a fish climb up a tree first :-)>>


Flying fish also contradict almost every single feature you cite as being
necessary for the achievement of powered flight in a cursor: Flying fish do
not have a strong breastbone for flight muscle attachment, nor shoulder
support, nor great muscle coordination, nor high intelligence, nor high
metabolism, nor lightweight construction, nor good vision, etc., etc. At
least, certainly not the kind we see in birds. Also, flying fish >use< the
viscosity of water when they swim in it to attain the necessary speed to
"fly," so they are not acting >against< viscosity. And their flight is
primarily a glide, not powered flight as in birds, easily understandable as
an extension of the ability of many fish to leap out of the water. So even
flying fish have gliding in their lifestyle.

<<>So, I can't see how you could possibly think that evolution of flight from
>the ground up is "conceptually easier" than evolution of flight from the
>trees down.

Both are _evolutionarily_ possible.  Which one is "conceptually easier"
depends more on psychological predisposition than biological feasibility.>>


Not in this case, as I hope I have shown above. To repeat: feasibility is not
the same thing as likelihood.

<<>> ...Consider extant flightless birds. I believe all of them are good
>> runners.  If you believe that modern flightless birds are descended
>> from previously flighted forms, then you acknowledge that the
>> reverse transition has taken place.  Why then is the transition in
>> the other direction conceptually hard?

>Some transitions are simply irreversible, for all practical purposes.
> [rest of entropy argument deleted]

Of course I know that some transitions can theoretically be irreversible. 
I took thermodynamics in college too.  What I am claiming is that
the running-flight transition specifically _is_ reversible;
in fact, that both transitions have occurred at least once.
To make your irreversibility argument, you will have to show how the
specific changes in the genotype at the DNA level are irreversible.
And showing an increase of entropy in one direction is insufficient
since organisms can forcibly reduce entropy locally (at the expense
of increasing entropy even more globally).>>


I gave the example of the breaking egg simply to illustrate that not all
processes are reversible. It was the wording of your paragraph, which
suggests that you think that every process is reversible, that prompted me to
put it this way. But even the breaking egg could theoretically reassemble
itself and leap back onto the plate, given a nearly infinite amount of time
to wait for the right combination of molecular motions to occur by chance.

Now, according to you, the running-to-flight transformation is practically as
likely as the opposite flight-to-running transformation. This is contradicted
strongly by the dozens and dozens of times that flightlessness has occurred
among various lineage of birds. As far as we know, running-to-flight happened
at most once (I contend that it happened >never<). So already we have a
tremendous imbalance--dozens and dozens to at most one--favoring
flight-to-running over running-to-flight. This is hardly an easily reversible

Indeed, since flightless birds already have much of the gear needed for
powered flight, one would expect the reverse transformation--of a flightless
bird back to a flying bird--to take place >even more readily< than when birds
are supposed to have evolved flight the first time. After all, dinosaurs had
to evolve all the gear from scratch. So how many times has flight re-evolved
among flightless birds? I think the answer is >zero<.

<<>We do not know how pterosaurs evolved, just like we do not know how birds
>bats evolved. In fact, we do not yet know just how well pterosaurs were
>adapted for walking and running; as with birds, it is more likely that
>pterosaurs evolved from climbing forms, not running and jumping forms.

Then why do pterosaurs, especially primitive ones, show no signs
of being adapted for climbing?  A climbing pterosaur would be an
awkward sight indeed.  The structure of the feet was unsuited for
grasping tree limbs, and lacked the claws needed to dig in.  They
would therefore have had to rely on the three fingers of their hands
to do the grasping.  This would further be made more difficult by the
fact the hands are attached to the front of the wings, so the wings
would open up as hands stretched out to climb.  This means it would
be unable to keep its huge wings tightly folded as it climbed.  Further 
awkwardness in climbing would be caused by the extremely large heads
(up to half the body size).  Hopefully, the tree doesn't contain many
branches or leaves, or the hapless pterosaur would be unable to get
through them without catching its wings.  And of course it would do
no good to climb only part way up the tree and jump off, risking 
that space between the foliage would be insufficent to fully open
its wings.  No, it would have to climb clear to the very top of the
tree, in order to be sure to clear the whole forest canopy.  I don't
know about you, but I find this whole scenario pretty absurd.

On the other hand, there _is_ a lot of evidence for pterosaurs being
adapted for running.  In fact, very well adapted.  The shin bone/femur
ratio, often considered a reliable indicator of running speed is
nearly 2, putting it in the class of the very swift agile runners.
The articulation of the leg bones shows that it could stand fully erect.
See the wonderful exhibit at the Museum of the Rockies showing
a Quetzalcoatlus life model standing fully erect (thanks to Ellen
in the Museum lab for giving me a lot of info about this exhibit).
Not to mention all the pterosaur ancestors we can find are cursorial.

I believe this whole climbing pterosaur idea was originally motivated
by the (now obsolete) idea that pterosaurs were awkward fliers; once
grounded, the only way the poor thing could get airborne again would
be to climb up a tree, or find a cliff and jump off.

We now know pterosaurs to be supremely competent fliers.
Pterosaurs probably ran quickly on open ground (or water) to
reach takeoff speed, in much the same way the waters birds do


Unfortunately, >every single one< of your objections to the ability of
pterosaurs to climb, run, etc. is the subject of considerable debate among
pterosaur specialists right now. You have presented only one side of the
debate, and you have presented it as if the debate is over, in your favor. On
this list I'm frequently accused of presenting the BCF side of the evolution
of flight as fact rather than hypothesis, so it is interesting to see that
other people are also afflicted with this disease. I'm not about to start
debating pterosaurs here.

<<>_Scleromochlus_ could well have been a flightless form descended from
>climbing and gliding forms ancestral to pterosaurs.

Oh sure, it's possible.  Find me evidence of such ancestral forms and
then we'll talk.  For now, all we have to go on is _Scleromochlus_,
which was clearly non-arboreal.>>


It was "clearly" nothing of the kind. It is virtually impossible to tell from
a skeleton of an animal whether it was arboreal or not. Looking at the
skeletons of a rat and a squirrel, could you say which was clearly the
arboreal form and which the terrestrial form? Which features of the skeleton
of _Scleromochlus_ make it clear to you that it was not arboreal?

<<>It is entirely incorrect to state that "the further back you go in the
>pterosaur family tree, the better adapted they are for running." There is no
>evidence for this whatsoever, unless you go way back to the original
>quadrupedal forms when they first became arboreal.

The evidence is there, you just have to look for it.>>


What evidence is this? I've researched the pterosaur family tree somewhat,
and I find no forms "better adapted to running" among pterosaur ancestors.
Indeed, I find few if any known pterosaur ancestors at all. And some of the
earlier pterosaurs, such as _Sordes_, seem encumbered by uropatagial
membranes that would make it quite difficult to run--like running with your
feet tied together.