Re: questions about the Odontochelys study

[David Marjanovic <david.marjanovic@gmx.at>]

> Confirming the sequence data, which characters can we look to as
> synapomorphies between crocs (or any other archosaur) and turtles?

There is a long, though small, morphological tradition among neontologists
of considering turtles archosauromorphs. I think there are braincase
characters.

Aside from neontology, the aetosaurs have been suggested as being good
candidates if the turtles are indeed closer to crocs than to birds. AFAIK
some have one elongate armor plate above each rib -- just like turtles (and
unlike even the most amazing cyamodontoid placodonts).

> <<    Was wondering which species Rieppel used when he linked turtles
> to lepidosaurs or squamates and sauropterygians?
>
> Mostly *Proganochelys*, IIRC.  >>
>
> Which species on the non-turtle side of the equation?

I don't have the papers in question here, I can check them out tomorrow if
you like. Probably he used suprageneric OTUs: Placodontia and
"Eosauropterygia" (what other people call Sauropterygia). It must be said,
though, that Rieppel knows both of these clades very well and is not
unlikely to have correctly reconstructed the ancestral state of each.

> <<(Which, incidentally, is not trivial to interpret. Kordikova 2002
> claims to have discovered postparietals and tabulars in the skull,
> both never seen before and both supposed to be absent.) >>
>
> Is Gaffney's interpretation of the skull gospel?

Is there any other, except for Kordikova's?

> And if so, which outgroup did Gaffney use to decide which skull bones
> were 'as is' and which were the product of fusion?

Wrong question. Like most people except Carroll, he called single bones by a
single name and didn't care whether any independent bone was lost by fusion,
by ontogenetic resorption, or by never starting ossification in the first
place. Any single bone of an adult might be a fusion product; without a
growth series or a _very_ detailed phylogeny, this is more or less
impossible to tell. Kordikova thinks she has seen _sutures_ between the
parietals, postparietals and tabulars that other people have overlooked.

The fossils are not in an ideal state. The ventral halves of the sutures
between quadrate, quadratojugal and jugal have never been clear to anyone;
you'll see stippled lines in all reconstructions that aren't simplified.

> Is there enough time for the evolution of the carapace in each of  these
> groups? Can we eliminate some possibilities due to lack of time?

Nobody knows, within several orders of magnitude, how long it would
minimally need to take, nor how long it actually did take. Sure, it can't
have happened overnight, but we have tens of millions of years at our
disposal under each possibility.

To reconstruct the speed of morphological evolution, you need a _very_
detailed fossil record (a very high stratigraphic resolution and a very high
number of fossils) and a _very_ detailed phylogeny. We lack both. It might
work with Miocene horses (and of course people have tried). It does not work
with Triassic turtles.