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

Re: Jouve's and Tacquet's Ctenochasma sp.

Coding restored, citation format clarified, broken links corrected:

----- Original Message -----
From: "david peters" <davidrpeters@earthlink.net>
To: <dinosaur@usc.edu>; <jouveste@mnhn.fr>
Sent: Friday, October 22, 2004 4:11 PM
Subject: Jouve's and Tacquet's Ctenochasma sp.

> Stéphane Jouve asked that I post this letter to the dinosaur list.
> With all honors, I do so now without comment, other two preliminary notes:
> 1) a phylogenetic analysis of every decent specimen in the inventory of
the Pterosauria reveals that that the Pterodactyloidea is paraphyletic, with
separate origins for the Ctenochasmatoidea, etc. and the Azhdarchidae, etc.
from various Dorygnathus specimens; and the Cycnorhamphidae +
Ornithocheiridiade, etc. and Pterodactylidae + Germanodactylidae etc. etc.
from various Scaphognathus (themselves derived from basal Dorygnathus). Thus
comparing brain sizes in Pteranodon and Tapejara to Ctenochasma crosses many
more taxon lines than comparisons to Dorygnathus (Parapsicephalus) purdoni
and the micro ctenochasmatids.  The latest single tree can be seen at
> www.pterosaurinfo.com/familytree.html

Sorry to say this, but your analysis reveals nothing of that sort. (I'll
send you the details privately.)

- Many of your characters are ontogeny- and/or size-dependent;
- many of your specimens should first be checked for their ontogenetic
age -- otherwise you'll have to make _extremely_ sure that there are no
ontogeny- and/or size-related characters in your matrix;
- lots of your characters are partly dependent on each other, so you end up
weighting certain characters (for example, state 159(0) is identical to
160(0) and 161(0)!);
- several characters can only be coded for many, sometimes all taxa if we
believe your tracings -- I can comment at most one of them, but I'm
confident that the skull of the *Cosesaurus* that is being born consists of
two unassociated lumps of rock (of which one is roughly triangular, vaguely
like the tip of a snout) and must therefore be classified with the face and
the pyramids on Mars;
- and NONE of your many multistate characters is ORDERED!!!

The last of these problems may well be the biggest, worse than all your
tracings. You have _many_ characters with states like "large -- small -- 
absent" or "< 1.1 -- 1.1-1.5 -- 1.5-2.0 -- 2.0-2.6 -- > 2.6"; it is obvious
that an evolutionary change from state 0 to state 2 must pass through state
1, so it must be counted as two steps, not as one. This means that the
length of your tree is bogus, and I'm not exaggerating. It ALSO means that
the possession of an intermediate state will be seen as an apomorphy
compared to the possession of more derived states, instead of correctly as a
plesiomorphy with respect to those. To illustrate this, imagine the
following data matrix (taxa left, states of the one & only character right):

A   0
B   0
C   0
D   1
E   1
F   1
G   2
H   2
I   2

Unrooted tree if the character is unordered:

|  |--B
|  `--C
|  |--E
|  `--F

If the outgroup has 0, this becomes the following rooted tree:

  |  |--E
  |  `--F

Unrooted tree if the character is ordered:

|  |--B
|  `--C

If the outgroup has 0, this becomes the following rooted tree:


I think you see the huge difference that ordering makes. BTW, some of your
characters could need more complex stepmatrices.

> And 2) questions regarding identification of bones need to go through two
steps: labeling the specimen and creating a valid reconstruction. Problems
arising with misidentification (and they are legion in the literature)
usually can be discovered during the reconstruction process. And if I'm at
fault here, then I'll make corrections.

Problem is... without seeing the specimens, in many cases you'll not be able
to test if your or anyone's interpretations are correct. We all know how
sutures can magically appear and disappear on even the best available
photos... not to mention descreened photos, which fraudulently claim to
possess information that they lack.

> > > Please take the following notes and criticisms kindly. I have learned
> > > much in my cladistic study of all the pterosaurs and not very much of
> > > what I have learned is in the literature yet.

Sorry to be blunt... you have seemingly learned rather little on how to code
characters in realistic ways, respectively you haven't given it much

> > > > > Fig. 4. The "brain" in the Saint-Dizier specimen is much too
> > > >
> > > > NO!! The brain is much smaller than the external bones of the skull
(see text!!).
> > >
> > > This needs to be reexamined. It cannot be compared to Pterandon and
Tapejara which are not related, but must be compared to sister taxa,
including Parapsicephalus and P. elegans.
> >
> > Why can they not be compared to Pteranodon or Tapejara??? They are also
Pterodactyloidea!!!! No ? P elegans is juvenile (see text) and
Pareapsicephalus is not a pterodactyloidea (I am not sure on that). The
condition of the skull in Pterodactyloidea is only known in Pteranodon and
Tapejara (the true shape of brain!!)

I agree. That *"P." elegans* is juvenile is a very important point: relative
brain size is strongly ontogeny-dependent. It decreases _throughout_ the
growth of just about any animal that has a brain. This means that the head,
including the braincase, grows faster than the brain! Please be also a bit
more careful with "not related".

> > > Fig. 9. The uneven ventral margin of the dentary in a
> > > specimen labeled, Pterodactylus "kochi" B St 1878 VI 1 [No. 13 of
> > > Wellnhofer 1970] noted as an example of a fossilization artifact, is
> > > not. an artifact. Instead this character is a genuine feature and the
> > > genesis of the deep mandible found in some Nyctosaurus and Pteranodon,
> > > which are sister taxa. (See pterosaurinfo.com > taxa > family tree.

Why, why don't you explain _why_ it isn't an artifact???

> > Your critisism are interesting, but you seem to not take into account a
lot of things. First: have you seen the specimen??


> > If you don?t, how can you provide a reconstruction????

By trusting the photo... the resolution of which isn't breathtaking.

> > (have you a problem with Chris Bennett ????

Yes... :-}