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

Re: Tucson Tales II.

Paul Janke writes:
 > Tucson Tales II. "Birds of a feather"
 > One of these topics is actually an old topic that has been revived
 > in recent years, the bird-dinosaur connection.  ...
 > The conventional wisdom of today is simple. Birds _are_ dinosaurs.

Well, using the cladists definition that all descendants of a group
are members of the group, yes.  But others accept the same descent
but retain the classification of birds as a seperate class.

 > Protoavis has drawn a steady barrage of fire ever since Sankar
 > Chattergee first proposed it as the earliest bird. The debate has
 > been an enlightening one in many ways, as the participants have
 > had a difficult time staying purely scientific in their exchanges.
 > Some have delivered scathing attacks on Chattergees field
 > methodology, promptness of publication, and poor science.

There is some justification for this.  Dr. Chatterjee has a history
of odd phylogenetic proposals - especially ones that ignore time
constraints.  He once proposed that the tyrannosaurs evolved from
a group of Triassic thecodonts he discovered called poposaurids,
while all remaining dinosaurs descended from a different group
of thecodonts!?!?  The idea that a group of thecodonts could have
survived without fossil record for 100+ million years is, to
say the least, unlikely.

Having read the first paert of his article on Protoavis, I
have concluded two things:

        1. His placement of Protoavis as a bird ancestor is
        unjustified by the evidence.   (One might just as well
        call it a crocodilian ancestor, or even an unusual

        2. Assuming the remains represent one organism, its
        anatomy is consistant with it being a ceratosaurian
        theropod dinosaur, rather than a thecodont.
        (That is how I classify it in my taxonomy file).
 > The other camp (currently the minority view) is that birds are
 > more closely related to the crocodilians. This view is held by
 > logist Larry Martin (U of Kansas). He believes that the true
 > stock to modern birds is to be found in the thecodontia, and
 > as described by Chattergee generally fits within the scope of his
 > theory.

Yep, Dr. Martin has long held that opinion.  I remember him telling
me years ago that the inner ears of birds were inconsistant with
dinosaurian ancestry.  I have never seen him publish anything on
that idea though.  And I think that more recent evidence shows
greater variability in the inner ears of theropod dinosaurs than
was known then.  (This was circa 10 years ago).

 > One morning at the Quality Inn (Tucson) I went to the lobby and
 > up some informational booklets. One of them was called the "Big
 > Trader". While flipping through it I noticed an article entitled
 > "Scientist and Son find Ostrich Dinosaur". It turns out that the
 > scientist is none other than Sankar Chattergee!
 > They claim to have found:
 >      - a toothless therapod they refer to as an ostrich dinosaur
 >        (I assume it's an ornithomimid).
 >      - they named it Shuvosaurus inexpectatus after Sankar's son
 >        Shuvo, and described it as a seed and nut eating therapod.
 >      - it's claimed to be the oldest known "ostrich dinosaur",
 >        dating to the triassic at around 225 mya.
 >   [ALERT] This predates other ornithomimid fossils by ~140 MILLION

Uh-huh!  Another of Chatterjee's temporally implausible phylogenies.
It is more likely to be a convergent form - just like ornithomimids
and ostriches are convergent forms.  At 225 MYBP, I suspect it
is not even a dinosaur (that being about date of the earliest
known dinosaurs).

 > Does anyone else see a pattern to Chattergees work? First, he finds
 > a fossil that has features one would expect from a cretaceous bird
 > (I refer to Protoavis), but it comes from triassic mudstones near
 > Lubbock and thus pushes back the date of the earliest bird by around
 > 70 million years. Now, he finds an ornithomimid which predates any
 > other by 140 million years in those same mudstones.

And he found a Triassic thecodont, Poposaurus, that he claimed
had features otherwise only found in the Cretaceous tyrannosaurs.

 >The simplest
 >explanation for this is that he is actually finding a typical
 > faunal assemblage and somehow misdating the fossils.

Unfortunately, that is unlikely, since Poposaurus truly *is*
a thecodont, not a dinosaur, and is closely related to Ornithosuchus.
[The fragmentary remains sometimes considered a dinosaur called
Teratosaurus are now also placed in the family Poposauridae].

He is just repeatedly finding covergences and assuming they are
synapomorphies.  It is a simple fact that *functional* characters
are prone to convergence (as, for instance, the ostrich-like
anatomy).  Dr. Chatterjee fails to correct for this problem
and tries to make all character matches into synapomorphies,
even when they cannot be so.
 > So what are the implications _IF_ this hypothesis is correct?

That a typically Triassic thecodont, Poposaurus, lived
in the Cretaceous.
 > a) Sankar Chattergee
 >      - is correct that Protoavis is a bird displaying somewhat
 >        advanced avian features.

This is not true, as far as my reading of his article has
gone.  All of the characters it shares with birds are also
shared with either crocodiles or advanced thecodonts.  So,
unless Dr. Martin is correct and birds are derived from crocodilians,
there is no reason to treat these characters as synapomorphies
between Protoavis and true birds.  And if Dr. Martin is correct,
it is really more not particularly advanced as a bird, since it
lacks any truly unique avian characters.  (Other than Dr.
Chatterjee's *extremely* doubtful reconstruction of wings -
the actual anatomy of the fore-limbs is uncertain, as the
bones are too broken up to allow unambiguous reconstruction).

[Note: the sharing of similarities with crocodilians is from Dr.
Chatterjee's *own* words].

 >      - may be incorrect that Protoavis is not a bird fossil
 >        but represents several different non-bird animals.

The specimens do not really demand an avian interpretation.
The ankle is certainly similar in form to theropod dinosaurs.
And the skull is too fragmented to allow sure identification.
It *could* be a theropod dinosaur, or it could be a thecodont,
or it could even be an early crocodilian (there were terrestrial,
cursorial crocodilians in the Triassic, so it is not that far
out of line).

Thus, it could be several animals very easily, as far as
that goes.

Or, it could be an early theropod dinosaur with anatomical
convergences to early crocodilians.  (This assumes the
convergent features were functional, and thus subject to
similar selection pressures).

 >        Of all the criticisms about Chattergee and Protoavis, none
 >        of the cladists have suggested misdating as an explanation.

I suspect because fissure fill and such like are not compatible
with the geology of the locaion it was found in.

 >        They seemed determined to show it wasn't a bird no matter
 >        what it looked like,

Well, the actual bones do not particularly *look* like a bird
until after Dr. Chatterjee is done reconstructing them.  Everything
is in little itty-bitty pieces, so it doesn't look like much of
anything at all.  So, what it looks like depends on how you
put the pieces together.

 > [comment] Although he could accept a triassic bird, a full-flighted
 >        triassic bird might not have been within his expectations.
 >        In fairness, he actually didn't totally rule out flight,
 >        he just didn't see sufficient evidence to convice him that
 >        the feather "nodes" reported by Chattergee were real.

Hmm, I haven't gotten that far in the paper.

 >   Under pressure, he gets high marks for recognizing the
 >   truly avian characters and agreeing with the bird classification.

What truly avian characters?  All of the 'avian' characters
are shared with at least one other group besides birds.

[With the possible exception of the feather nodes - since
only flight feathers are attached to the bone].

 > Here's a few questions to stimulate some discussion:
 >      1) Which classes share a closer common ancestor, birds-dinos
 >          or birds-crocodilia?

Birds do admittedly share a number of characters with both
dinosaurs and crocodilians (and the crocodile-like thecodonts,
the pseudopsuchians).  The question, then, is which set of
similarities is more reliable for determining phylogeny.

Features that are directly adaptive to some specific lifestyle
are *not* reliable characters.  Complex derived features are
the most reliable.

Now, crocodiles and psuedosuchians have a very speialized,
derived, ankel joint, called an crocodilian ankle.  Dinosaurs
and birds do *not* have this type of ankle, and most specialist
do not think that the simple hinge joint (a meso-tarsal ankle)
they do have could be derived from the crocodilian type.  Instead
they see one of two alternatives as being more likely:

        - deriviation from the 'crocodile-reverse' ankle of the
        ornithosuchian thecodonts.

        - direct derivation from the simple primitive ankle of
        the proterosuchian thecodonts.

In either case, birds would be closer to dinosaurs than
to crocoddiles.

Dr. Martin's inner ear diffrences, *if* they are real, and not
based on the lack of adequate small theropod fossils, would be
much stronger counter-evidence than anything in Dr. Chatterjee's
paper (a least as far as I have gotten in it).
 > [This may seem to have already been decided in favor of the
 > but don't bet the farm before checking out Larry Martin's work.]

*Never* bet the farm on reconstructed phylogenies, they are always
 >      2)At what approximate point in _geological_ history does the
 >          ancestor share a common ancestor with the dinosauria?
 > [Some of the cladists have suggested cretaceous, as the dinosaurs
 > birdlike in the cladistic analyses are of cretaceous age, but what
 > a fully flight-feathered archaeopteryx in the jurassic and modern
 > analysis suggesting that a proto-bird ancestor most likely existed
 > earlier than archie?

Hmm, I am not aware of this last, but I have never been too
comfortable with the assumptions of rate uniformity necessary to
make date estimates from DNA differences.

 > Don't rule out a thecodont as the dino-bird close
 > common ancestor just yet.]

Nope, but an ornithosuchian or lagosuchian seems more likely,
at present,than a pseudosuchian.
 >      3) How many of the bird-dino anatomical similarities
 >         are attributable to convergence vs inherited from a common
 >         ancestor?
 > [The cladists tend to lean heavy toward the common ancestor but
 > bipedalism and the occupation of same niches may complicate things.]

Exactly.  Actually, that is Dr. Chatterjee's problem.  He fails
to account for convergence due to similarity of niches.
(For instance his Poposaurus=>Tyrannosaurus proposal, based
on convergences for the large carnivore niche - both were the
top carnivore of their time).

Most cladists *do* try to take this into account.
[By the way, Dr. Chatterjee is a cladist himself, just
not a very good one].

swf@elsegundoca.ncr.com         sarima@netcom.com

The peace of God be with you.