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

Re: Eunotosaurus and turtles

Thank you for the general insight on the subject.
Sorry to everybody for furnishing wrong/non-updated data on the
osteoderms, and the number of expanded dorsals of Eunotosaurus (I got
that information from Romer 1956).
Fortunately, today I was able to access to Cox 69' so as to continue
this a little less ignorant.

David Marjanovic wrote:

> *E.* has 10 dorsal ribs, of which all are broadened along at part of their
> length; the 10th is fused to the vertebra and apparently more or less
> conical.

You indicate that there are 10 dorsal ribs, and that all of these are
broadened, but then say that the tenth is apparently conical. In such
a case, we can only say that 9 of the ribs are expanded.

Given the official last dorsal presents its rib fused to the vertebra,
that there is just one sacral according to Cox 69', and that the
sacral ribs direct cranially in fig. 13 of Cox 69', it seems that this
"last dorsal" may actually represent a sacral (contra the also
possible guess of Cox 69', below I give my reasons based on attempting
to establish one-to one homologies).

> *Proganochelys*, however, officially has 10 dorsal ribs, of which
> the last 9 participate in the carapace; I think the so-called 8th cervical
> rib, which is much longer and taller than the other cervical ribs and lies
> behind the scapula (see below), should also be counted as a dorsal one.

Considering that so-called 8th cervical a dorsal is also what Romer
56' does. In any case, perhaps it is called 8th cervical in Gaffney
90' more to indicate its homology with the last cervical of most

Now this seems interesting: according to Cox 69', there is a vertebra,
called cervical by him, which seems to be just cranial to his "10
dorsals". In that vertebra the capitular and tubercular articular
surfaces fuse, as in the dorsals behind. In this it does not resemble
vertebrae cranial to the "8th cervical" of Proganochelys.

Now, if in Eunotosaurus the "tenth" dorsal is an ex-sacral, the
"cervical" may be the first of ten dorsals, and (conserving Gaffney's
90' names for vertebrae) the nine expanded ribs of Eunotosaurus would
correspond to the nine costal plates/ribs of Proganochelys. Romer 56'
states that the last dorsal in Recent turtles at least reaches the
carapace (may have it embryonarily began a dermal expansion later
fused to the last costal plate?).

> The
> first full-length dorsal rib is the 3rd one in *E.* but the officially 1st
> one in *P.* and apparently *Odontochelys* and the officially 2nd one in
> other turtles.

Do not have access to Nature. Is the dorsal series of Odontochelys
complete (better question: is the presacral series complete)? Because
if its possess 8 expanded ribs/costal plates, as in most turtles, but
the first of these is that of dorsal 1, at difference of later
turtles, then it seems likely that the 8 expanded ribs in Odontochelys
are not the same than in later turtles.

>> 2- These dorsal ribs are said in Romer 56' to articulate with the
>> vertebra via a single attachment, as in turtles (I suppose it would be
>> the capitulum, as it seems to be low in the centrum).

> All of these are correlated to restricted or lost mobility of the back: few
> longer instead of many shorter vertebrae, and less mobile ribs.

I think you are right in that the longer cervicals decrease mobility
at equal trunk lenght and shape of the articular surfaces, and the
same for the wide ribs.

However, I think the correlation is not so clear for the single
attachment point. I think that a single attachment point, everything
being equal, in principle permits more mobility than two attachment
points: two attachment points restricts movement to the plane
perpendicular to the line uniting both points of contact, whereas just
one permits motion on all planes passing through the articulation
point, and also more long-axis rotation.

Today I see that Cox 69' fig. 13 indicate the para- and diapophyseal
surfaces fused to give the single point. It can then be argued that
the way in which the single articulation emerges is by incorporating
into the joint areas that previously separated the para- and
diapophysis, but in Cox 69' pictures the single articular surface
indicates both just approached each other, so as not to imply a
notably larger area of articular surface between rib and vertebra than
in taxa where capitular and tubercular facets are separated. Finally,
Romer 56' reports the articulation permitted freedom of movement.

> Also, *E.* really does have broadened ribs. It's much like uncinate plates
> gone wild: there is a very low "keel" near the front edge of the plate, and
> the broadening extends all the way to the rib head. *Odontochelys* is
> different: the broad part does not extend all the way to the rib head, and
> the much more prominent "keel" is about in the middle of the plate,
> sometimes not parallel to its edges; this also seems to hold for
> *Proganochelys* (fig. 79 of Gaffney 1990).

> For part of their
> length, the ribs are vertical plates that stand out from the carapace, like
> in *Chinlechelys* (*), and especially the first and last few ribs lift off
> from the carapace completely for a smaller part of their length (the 9th and
> 10th actually for a greater absolute length than the 1st, which, see above,
> is not officially considered part of the carapace). I conclude that the
> costal and the neural plates meet, and the ribs and the vertebrae meet some
> distance under them; this is more easily compatible with the hypothesis that
> the costal plates are dermal bones than with the hypothesis that they are
> broadenings of the ribs.

I think that the presence of differences between the ribs of
Eunotosaurus and those of turtles do not impede homologating their
broadening. For instance, you are right in that the greatest
broadening in Eunotosaurus is caudal, and that involves largely the
uncinate process. However, there is also a cranial broadening on the
external surface (fig. 7g), and in this it resembles the turtle

I see it difficult to say that the broadened ribs of Eunotosaurs are
different from the state in turtles because in the former they are
just rib broadenings while in the latter they are osteoderms or
osteoderm homologs. I do not deny in turtles the costal plates may
correspond to what originally were derivatives of dermal ossification
centers. But I ask why cannot be the same the broadenings of

Indeed, Cox 69' shows evidence of external surface sculpturing that
suggests attachment to some epidermal horny structure. This indicates
at least the distal extreme of the rib was in contact with the dermis.
And I do not see why would not be the more proximal too. Given that
the trunk appears to have been relatively unmovable, and that the
neural spines are low, we can infer the epaxial musculature was
reduced, and therefore much of the external surface of the ribs might
have bee in contact with the dermis.

Romer 56' writes that in turtles, the muscle between the carapace and
the proximal extreme of the ribs is a neck muscle (not surprisingly,
the epaxial trunk muscles should have dissapeared once the bones they
moved one relative to the other fused). Thus, if the neck was not so
long as in turtles, it may even be that there was less muscle above
the proximal extreme of the ribs in Eunotosaurus than in turtles (pure
speculation, however, but so is thinking musculature covered the
dorsal surface of the proximal extreme of the widened ribs withouth
showing compelling evidence of muscular attachments).

Cox 69' says that the shoulder girdle laterally covered the first two
broadened ribs, and this may indicate that these expansions were
laterally overlapped by endochondral bone, and in principle dermal
bone should not be deep to endochondral bones (forgetting the palate
and lower jaw and the explanation of that case). But there are just
these two ribs the ones where the expansion is more restricted
proximally, the distal part being nearly cylindrical. According to
Cox's 69' restoration in fig. 13,  the shoulder girdle is mostly
ventral to these expansions. Thus, it may still hold that for these
ribs, that the expansion is related to contact with the dermis, and
the distal rod-like part is just endochondral.

> Everyone in
> choir: the shoulder girdle lies inside the carapace, but in front of the
> ribcage. It also lies in front of the ribcage in *Eunotosaurus*... and thus
> in front of what is functionally a carapace...)

I think in turtles the shoulder girdle is actually medial to some
ribs. It had to get relatively more posteriorly located, from a point
originally cranial to them, as indicates the fact you say that it is
lateral to the ribs of the "8th cervical" of Proganochelys. It's
simply that some ribs (for example, the second dorsal, contacting the
first costal plate) are so much laterally set that when the shoulder
girdle "retracts", there is no other result than the rib being lateral
to the girdle. In Eunotosaurus, the girdle is mainly cranioventral to
the ribcage, but also lateral to a small part of it (the first two

Finally, in reading Cox's 69' phylogenetic analysis he finds the next
derived characters shared also with turtles:

4- Reduction in number of the dorsals to 10/11: this is not sure, but
accepting the profile of the body, which much thins cranially in the
holotype according to the holotype,  it seems there were no many
dorsals cranial to the first preserved rib. This is much less than in
most other amniotes, including parieasaurs and procolophonids.

5- Shortness of the transverse process: correlated by Cox with the
reduced insertion of axial musculature, and general lack of motility
at the trunk.

6- Sharp cranial thinning of the thorax (compared with parieasaurs and

Altrough it may be that even all the features that are derived and
resemble turtles are adaptations a specific lifestyle, they do not
need to be correlated. They can, but it is not necessary. For example,
the ribs can widen and the dorsal vertebrae not be necessarily
elongate (against what I suggested in the previous post in this thread
and Cox 69' say somebody said before), as shown by Thrinaxodon. If
they can overlap each other (as to some degree occurs in the holotype
according to Cox's 69' fig. 7), they can also permit motility. The
transverse processes can be present and the muscles inserted on them
lacking. There may be other mechanisms that restrict motility, even
fussion, or fussion to the ribs, and if present could be considered