Re: Protoavis?

[Jonathon Woolf <jwoolf@erinet.com>]

Jonathan R. Wagner wrote:
>         Okay, time out. Let's all catch our breaths and compose ourselves...

Fair enough.  Let's do (and it might also be a good idea for both of us
to forget certain even less amicable comments made in private e-mails).

> For my part, I apologize for not understanding.
> 

Accepted.  In return, I apologize for over-reacting.  

>         b) Your post was not taken as a personal attack so much as another
> instance which might serve to perpetuate misconceptions concerning a subject
> which has already been so misconstrued as to represent a blemish on the
> reputation of vertebrate paleontology, not to mention on the records of the
> participants. Any sensitivity with regards to this issue may be traced back
> to devisive nature of the questions involved. I would suggest that you try
> not to take the reaction personally, and in future perhaps tread more
> cautiously on this ground.

Understood.  However, I think you might have answered in a way that
didn't sound _quite_ so patronizing.  I really am willing to listen to
sound arguments logically presented. <g>

> >Would it be too much to ask that you remember that not everyone on this
> >list is a degreed paleontologist, and not everyone is interested in
> >getting the jargon exactly right?
>         My point was that the jargon is jargon for a reason. Hard as it may
> be for you to believe, scientists do not sit around making up jargon to
> confuse the layfolk. Jargon is used to aid in precision of thought and
> communication. 

All true, and I'm not arguing with the fact that jargon is sometimes
necessary.  But when jargon is (or seems to be) used as a put-down, it's
a mis-use of the language and the concepts behind it.  In this case, I
didn't really think the jargon was necessary.  There's a distinct
subgroup of Dinosauria commonly referred to as "birds," and I _think_
everybody knows more or less what that group includes.  

> >Be that as it may, if we had only one partial fossil of a serval, and no
> >other fossils of any animal in the Felidae,nor of any potential ancestor
> for >it, [...]
>         As an aside, I should note here that the presence or absence of a
> "potential ancestor", however that status may be determined, is usually
> irrelevant for the purposes of determining relationships.

It's not necessary, I agree, but it's often useful.  Establishing the
exact evolutionary relationships between birds and other theropod
dinosaurs would be somewhat more difficult if we didn't have
_Archaeopteryx_.

> >we couldn't say a whole lot about exactly how the
> >serval was related to other carnivores.  Unique skull, unique teeth
> >pattern, unique legs and feet . . .
>         I have never played with carnivores all that much, so I couldn't
> say, but somehow I doubt the situation would be as bad as you suggest.

I think it might be.  If all we had was one serval fossil, then all the
apomorphies that define Felidae would be unique to that single
specimen.  Thirty-two teeth instead of the standard forty-four;
rudimentary molars specialized as carnassials; a short, blunt skull with
huge eye orbits; an extremely flexible backbone; retractile claws; and
so on.  Add to that the features that make the serval a most unusual
cat, like the hyperextended ankles, and yes, I think you have an animal
whose relationships to other carnivores would be difficult to deduce.  

> Anyone with experience in the field of felid phylogeny (a little
> alitteration always alerts the auditory apparatus) care to comment?

<g>

>         If we cannot tell the relationships of an animal from one skeleton,
> we are indeed persuing the largest scientific dead-end since neptunism. 

Not necessarily.  I'm not trying to cast aspersions on the entire
science of comparative anatomy.  I'm simply saying that _Protoavis_ may
be an anomaly that standard methods can't deal with.  It may occupy the
same sort of place among the known Dinosauria that the platypus occupies
in the known Mammalia -- we know it's related to all the other mammals,
but doggone if anyone can figure out how, because it's a confusing
mixture of primitive and specialized traits, and there's almost no
fossil record for monotremes.

> >> Is the presence of birds
> >> delimitied by the oldest occurance of feather preservation?
> >
> >Given how similar some theropod dinosaurs are to primitive birds, in my
> >amateur's opinion, the answer is yes.
>         Unless _Archaeopteryx_ is the ancestor of all later birds
> (unprovable), then birds HAD TO exist *before* _Archaeopteryx_.

Correct.  Perhaps I should say that the _provable_ presence of birds is
delimited by the presence of feathers.  In fact, since feathers or
protofeathers appear to be shared by at least some non-avialan
dinosaurs, I might go so far as to say that birds should be defined by
the specific presence of _flight_ feathers.  Nonflying descendants of
flying birds could be included under the normal rules of taxonomy, but
at the root of the avian family tree, only the first animal with flight
feathers and its descendants could be called birds.

>         What does the similarity of the two groups have to do with it? In
> order to succinctly hypothesize that a specimen is a "bird" (Avialae,
> Gauthier 1986), one must simply present a well supported phylogenetic
> hypothesis which suggests that it is a decendant of the most recent common
> ancestor of _Archaeopteryx_ and _Corvus_ (happy Pete?). Chatterjee believes
> he has done this. No feathers involved.

But if it comes seventy million years before that ancestor lived, then
it can't very well be descended from that ancestor.  Can it?

> >For Late Jurassic and earlier
> >fossils, you cannot assume you have a bird based only on skeletal
> >characteristics.
>         This is simply wrong.

On reflection, I have to agree.  A complete skeleton showing all the
uniquely avian apomorphies -- wishbone, carpo-metacarpus,
tarso-metatarsus, pygostyle, etc. -- would be classifiable as a bird no
matter what strata it was found in.  Does _Protoavis_ show such a suite
of characteristics?

> >Remember, if it didn't have feathers _Archaeopteryx_
> >would be classified as just a small, odd theropod.  Only the feathers
> >mark it definitively as a bird.
>         However, were an animal outside of the archie+modern birds clade to
> be discovered with flight feathers, then later analysis would suggest it as
> a bird, even if it had no feathers. 

Um, I'm afraid I can't make any sense of this as written.  Could you
rephrase it, please?

> The feathers *diagnosed* it as a bird,
> but birds can be *diagnosed* without feathers. See _Ichthyornis_ and
> _Hesperornis_.

_Ichthyornbis_ and _Hesperornis_ are Cretaceous, and they show a number
of avian apomorphies.  It's no great trick to diagnose a bird when you
already know that birds were around before then.  

>         The "flip-flopping" you see [in _Mononykus_] are the result of 
> homoplaisy,
> evolutionary reversals and/or convergences, which are present in basal
> maniraptorform theropods. This is a normal problem in systematics, and has
> nothing to do with the presence or absence of feathers (at least, it 
> shouldn't).

Is this not also a possibility with _Protoavis_?

> >I want to see a smoking gun, preferably with fingerprints on it.
>         Then you will be waiting a long time. As it is, paleontology
> sometimes has to stretch the limits of interpretation beyond what should
> make even a relatively "liberal" scientist squeemish. We have been
> outstandingly lucky to get the evidence for bird origins that we have. In
> many other areas, we have not seen, nor are we likely ever to see, "smoking
> guns" of the type you describe.

In many other areas, we don't need it.  In this particular case, I want
to see it simply because this case is so unusual.  The idea of a
Triassic bird more advanced than _Archaeopteryx_ runs counter to
literally everything else I know about the fossil history of both
dinosaurs and birds.  It's as if somebody found a single fossil of a
cynodont therapsid whose pelvis looked like it might have been adapted
for viviparity, and based on that _one_ specimen, this someone announced
that Mammalia must be polyphyletic, with Eutheria descending from his
new therapsid and _all_ known Jurassic and Early Cretaceous mammals (all
of which are probable egg-layers) coming from another line of
therapsids.  That's a major rewrite of mammalian phylogeny.  Likewise,
_Protoavis_ as a bird is a major rewrite of archosaurian phylogeny.  I
want to see something more than subjective interpretations before I
accept that.  If I'm being unreasonable, fine.  Doesn't bother me none. 
(I do that a lot, as you may have noticed. <g>)  I'd rather play "hard
to convince" and be wrong than jump to embrace a new idea and find out
later that I'm wrong.  

In fact, I have a better example than that.  I know of a geneticist who
has done a bit of work with recovered RNA from fossil bone.  His work is
ongoing and hasn't been published yet, but he says his analysis shows
that mammals and ornithischian dinosaurs are closer to each other than
either is to birds, and that in fact birds diverged from other amniotes
in the Permian.  He doesn't pay any attention at all to the fossil
evidence that says otherwise.  

To my eyes, there isn't a great deal of difference between what he's
saying and what the _Protoavis_ advocates are saying.  He says, "to heck
with the morphological evidence; the genes say this, and the genes are
never wrong."  _Protoavis's_ supporters seem to be saying, "to blazes
with the timeline; phylogenetic analysis says this, and phylogenetic
analysis is never wrong."  My view is that the hypothesis has to fit
_all_ the evidence, or else I'm going to cast a very skeptical eye upon
it.  Throwing away stratigraphic evidence because it makes the problem
more complicated is IMHO an extremely unscientific approach.  In this
case, it seems to me that the hypothesis that best fits _all_ the
evidence, including the timeline, is that _Protoavis_ is either a
chimaera or an anomaly, and the true lineage of modern birds starts with
_Archaeopteryx_ or a very close ancestor of _Archaeoptryx_.  

> >In the
> >case of _Protoavis_, that means unmistakable evidence of feathers, and
> >preferably flight feathers.
>         Like I said before, feathers may not be only a "bird" trait.

But having definite feathers, especially flight feathers, makes the
"bird" classification a lot less open to question.

>         However, most terrestrial vertebrate workers whose methods are
> reproducable and consistant (i.e. cladistic) seem to prefer to leave
> stratigraphy out of their analyses. Why? Well, because the terrestrial
> record is so poor that it biases the data. Is this a rational approach? 

So these people assume that the record is biased, and incorporate that
assumption into their analyses?  Seems a mightily dangerous approach to
me.  In some cases the record most certainly is biased, but in other
cases it may or may not be, and in still other cases it almost certainly
isn't.  In this case, the absence of anything that even _might_ be a
bird between _Protoavis_ and _Archaeopteryx_ is IMHO too great an
anomaly to dismiss so cavalierly.

-- JSW