Note from Kevin Padian

[rowe@lepomis.psych.upenn.edu (Mickey Rowe)]

Kevin Padian just attempted to send a message to the list, but it was
rejected.  Kevin, it wasn't your fault -- it was due to a random
coincidence -- so please don't do anything differently next time.

If you wish to respond to Kevin directly, please reply-to:
kpadian@violet.berkeley.edu instead of to my address.

Oh, and one more thing.  Some of Richard Cowen's messages (re: display
hypothesis of the origin of flight) contained non-ascii characters,
and your mailer may have rejected them.  Please write to me or Richard
if you think you missed a message because of this problem (it should
be easy to tell since he numbered the parts).

--
Mickey Rowe     (rowe@lepomis.psych.upenn.edu)

Without further adieu, this is from Kevin:

---------------------
I have been hanging out enjoying the contributions on many topics, especially
pterosaurs and bird flight, but I haven't chimed in until now because I think
that non-professionals ought to be able to converse freely without a lot of
crosstalk from us ivory tower folk; having said that, I think that the input
from the academics so far has been really great and helpful (most subscribers
would probably agree that Tom Holtz's contributions are one of the best reasons
for Newt not to disband the USGS!).

I have learned a lot from the discussions about bird flight and its origins.
My old friend Richard Cowen has put in his two cents in a nice review of the
problem, and Tom Holtz (among others, no doubt, by the time you read this) has
added some good commentary.  I just want to add a couple of thoughts, if people
don't mind, and then i hope to shut up on it.

1.  There has been a lot of discussion in the past weeks about this problem
based on what one group or another of living birds has or does.  These are very
interesting features, and they often give us clues of new ways to start 
analyzing
problems.  But they are hardly ever anything more than circumstantial evidene,
and can often be misleading.  A related example:  most mammals today give live
birth and only the two monotremes lay eggs.  If you didn't think to look at the
outgroups to mammals, you might get the impression that the monotremes were just
"weird" and the primitive condition for mammals was to give live birth.  But we
know that this is a derived condition.
        In the same way, we have to be careful about statements related to the
conditions in living birds.  Pterylosis (feather tract patterns) in living birds
do not tell us what the ancestral pattern was like; nor does the sequence of
development of feather types in ontogeny.  Developmental patterns and timing,
like everything else, are subject to change and selection in evolution.  There
are many good examples of this, and some are given in Brian Hall's recent book
on homology (esp. the chapters on developmental patterns). So, while it really
is interesting to note pterylotic patterns in living birds, we cannot know that
developmental timings have not been altered in evolution (especially in this
highly selective organ system).  The best approach to the pterylosis problem,
I think, would be to establish phylogenetic patterns of pterylosis within living
birds, and then to deduce the primitive pattern for living birds if possible.
This would at least give you the most primitive pattern for crown-group birds,
and might be something to point toward more ancestral patterns.  But given that
crown-group birds may share a common origin as late as the Late Cretaceous, and
the origin of flight was no later than Late Jurassic (with a great many now-
extinct bird groups in between), inferences would be chancy.
        
2.  Richard's discussion of "arboreal vs. cursorial" flight scenarios is a good
one, as usual.  On the arboreal side, I really must confess that I cannot see a
single positive hard piece of evidence for arboreality, though I know that some
arguments are persuasive.  No study has ever been done (to my knowledge) that
demonstrates climbing ability in Archaeopteryx or its relatives.  Feduccia's
1993 SCIENCE paper claimed that Archy's claws were more like arboreal birds than
like ground birds.  He cited, but did not really take into account the findings
of, three previous studies of this by Ostrom (who concluded the opposite),
Yalden (who did not realize that many of the claw features, which he compared
to those of squirrels and other mammals, were inherited from theropods and were
already quite curved in the hand), and especially Peters and Gorgner (1992, in
the LA County Museum volume dedicated to Pierce Brodkorb).  The last study is
the most comprehensive as P&G measured many 3-D features of bird claws; their
conclusions also differ from Feduccia's.  Again, the importance is to take into
account phylogenetic legacy: you don't have to explain the presence of a feature
in an animal if it is already present in its ancestor.  It may, of course, use
the feature for new functions, but that needs to be established on independent
grounds.

3.  Which brings us to the Cowen & Lipps display scenario for feathers.  I agree
with Tom Holtz and Richard that it's a good idea, and display MAY have ben
important for early birds and pre-birds (as it is now); in fact I suspect it
was, given that theropods were generally active and highly visual creatures.
The problem is that, as with the obvious but unsupported importance of viruses
in evolution through time, we have no positive evidence of such behavior, colors
or soft structures in the fossil record, so our indirect inferences are even
more tenuous than usual.  And again, what living birds NOW do with display an
color cannot be used as evidence for what the proto-birds did.  For the benefit
of subscribers who aren't hammered daily by the hegemony of phylogenetic
systematics (opprobrious as it is at times), the neat methodological thing about
seeing the problem in this way (to me) is that cladograms really do help clarify
these kinds of situations (like anything else, the more data, the clearer).
Behavioral features, though, are tough, and for anyone who is interested in how
phylogenetic analyses of behavior have clarified our understanding of reptilian
behavioral evolution, I would suggest looking into Harry Greene's recent work.

4.  Understanding the origin of feathers, however, only goes so far to showing
how flight evolved.  Richard's discusion takes us very well up through the work
of Caple, Balda, and Willis, who showed how a running, leaping animal with a
small expansion of surface toward the end of long extended arms could increase
its stability and lift (but also drag) jumping into the air after insects.  This
transferred the prey-catching ability from the hands (Ostrom's original model)
to the mouth.  However, this didn't solve the problem of the flight stroke,
which is the defining feature of flight (see my review in Palaeontology, 1985,
for some other background stuff I won't bore you with here). Jacques Gauthier
and I showed that the predatory motion of the deinonychosaurs, whipping forward
the sideways-flexing wrist and the elbow, was virtually the same motion as the
flight stroke cinematically recorded in birds and bats (see the Archaeopteryx
Conference volume edited by Hecht et al. in 1985).  This, we think, is the 
missing piece of the puzzle in terms of evolving active flight.  And it's
interesting that only birds and deinonychosaurs (sensu latu -- Tom will have to
decide what's the right term to use these days -- we're working on it, folks)
have this sideways-flexing wrist.  And these are obviously terrestrial animals.
        I don't want to belabor a false dichotomy between arboreal and cursorial
models.  There has been a number of publications on intermediate ecological
scenarios that really obviate this time-worn conflict.  I think we've got the
mechanical stuff generally nailed down; the ecological questions are tougher,
but in the end it may not matter if Archy's grandpa could climb a conifer 
(after all, goats and kangaroos get into trees with ease ...), or if he just
ran down a slope or hopped over a fallen tree trunk.  Gliding and flying are
very different adaptations, and leave different morphological traces (again,
sorry to belabor, so will refer you to the 1985 Palaeontology paper).  Bats
have all the attributes of gliders; birds have virtually none (and those that
they have may well be secondary).  There are some neat questions remaining.
Jeremy Rayner, one of the best in the business, has long been worried that the
speed needed to reach the point where the power needed for flight is lowest
(about 7 m/s) may be too fast for a little theropod, and figures on living
lizards et al. suggest that this is really pushing the upper limit.  However,
this assumes that flight would have to begin at the most efficient speed, and
also that the U-shaped power curve is correct -- but the first assumption may
be highly selectionist, and the second has never been empirically confirmed in
living animals (andI have just learned from a colleague's experiments in 
progress
that it may well be wrong!). So there's a lot more to do in this.
        I wouldliketo stress, though, that what differentiates science from the
armchair speculation that we all like to do (andthat points us the way, quite
often, to great investigations), is that science relies on methods and on
testable ideas (or hypotheses).  It's important to the cultural basis of science
(changeable as it is)to agree on what methods will give the most fruitful
answers, and to pursue these.  (Currently in comparative biology we rely a lot
on cladistics, because we think that if you want to ask evolutionary questions,
you have to begin with a phylogeny [and continue to test it as you go]; 
otherwise
you are just spinning "what might have been" scenarios.)

This certainly doesn't exhaust the issue, but I will stop babbling and hope not
to exhaust your patience any more.  Richard, I am using your book in my course
this term (first time!) so we will see what the students think!

Kevin Padian
Museum of Paleontology
Univ. of Calif., Berkeley

PS:  If anyone would like the full citations for the works I refer to here,
say so.