Re: questions about the Odontochelys study

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

>  I actually find the bizarre preservation curious, the presence of a 
> nearly fully formed plastron and unique ribs. All this tell me that dorsal 
> osteoderms would have been present but were lost. They might simply not 
> fuse to the ribs in this taxon, especially since neurals are apparently 
> present. This seems if anything counter to the original presentation of 
> dorsally bare but ventrally intact armor.

As I mentioned before, I think the biggest dorsal osteoderms -- the costal 
plates, each of which covers a rib -- _are_ present and fused to the ribs, 
rather than the ribs being broadened. (Broadened ribs occur in 
*Eunotosaurus* and *Pumiliopareia* and look different.) The difference to 
later turtles is that the costal plates are not sutured to each other and 
were therefore able to disarticulate (each together with a rib). As you 
mention, the neurals are there and apparently made the back as immobile as 
if an ordinary carapace had been present. What is really missing are the 
peripherals, the nuchal and the pygal.

I also mentioned that Li et al. believe (on arguments from development) that 
there are no costal plates in any turtle and instead the ribs are broadened, 
and that I disagree based on what their own paper says about the development 
of the neurals.

BTW, I find it utterly unsurprising that the plastron develops first in 
turtle ontogeny. That's because the _parts_ of the plastron are all 
evolutionarily older than the carapace: the plastron consists of the 
clavicles, the interclavicle, and the remaining few pairs of gastralia. 
There's nothing new in the plastron; at least the dermal shoulder girdle 
(clavicles & interclavicle) develops early, and once it's there, part of the 
plastron is there. The carapace, on the other hand, contains neomorphic 
osteoderms, so I'd expect it to develop later than the plastron.

(The plastron of those cyamodontoid placodonts that have one, on the other 
hand, lies outside the gastralia and the dermal shoulder girdle and consists 
of neomorphic osteoderms, just like the carapace of the same animals.)

While I am at it, I should mention that gastralia are homologous to the 
ventral scales of non-amniotes/non-diadectomorphs. "Scales" as in "fish 
scales". Transitional morphologies are common in temno- and especially 
lepospondyls. (Most members of both taxa also retain the dorsal scales.)