Therizinosaurs and Ornithomimosaurs

["Jaime A. Headden" <qilongia@yahoo.com>]

Ken Kinman (kinman@hotmail.com) wrote:

<As for therizinosaurs, if Mickey's upcoming analysis shows that
they might be paraphyletic (or even polyphyletic?), I suppose
cladistically defining such a single "clade" would be premature
at best. If they do form a clade (and obviously I think that's a
big "if"), I believe therizinosaurs will end up branching off
between ornithomimids and oviraptorosaurs (or even before both
of them---see below). Placing therizinosaurs and oviraptorosaurs
as sister groups just doesn't feel right to me (I tried
classifying them as sisters, and ended up going back to
splitting off therizinosaurs earlier in the tree.)>

  Why don't they? We can't leave this to an intuitive position =
"It just doesn't feel that way." That's not science. Anyways,
there have been several analyses, some inherently flawed, that
tested the therizinosauroid + oviraptorosaur relationship:

  Xu et al., 1999, using three OTU's for segnosaurs, including
*Beipiaosaurus*, which has several unambiguous maniraptoran and
maniraptoriform synapomoprhies; Sereno, 1999, which failed to
include characters known only in therizinosauroids and
oviraptorosaurs, and even coded some key characters of the manus
as unknown in Caenagnathidae (e.g metacarpal I/II relative
length) as unknown, and not coding others which suggest the
(Therizinosaur (Caudipteryx (Caenagnath + Ovi))) lineage. I have
over 25 synapomorphies for this group, which is ridiculously
high, of which very few (~3) are known in other taxa like
ornithomimosaurs, etc.. Sereno only uses one OTU, and also uses
Ornithomimidae instead of Ornithomimosauria, as the definition
of the latter includes *Pelecanimimus*. Russell and Dong, 1994,
which has a suite of problems in the coding.

  So: Sereno finds the following characters supporting his
Therizinosaurs + Ornithomimosaurs clade:

1) no lachrymal antorbital recess; 2) overlapping lachrymojugal
articulation; 3) planar or grooved prefrontal articulation w/
frontal; 4) no pneumatic excavations of the antorbital fossa on
the jugal; 5) mid-cervical ribs subequal to centrum length; 6)
no subnarial foramen; 7) rostral ramus of maxilla equal or
longer than 50% of the antorbital fossa; 8) invaginated ventral
margin of the antorbital fossa; 9) no nasal anteroventral
process; 10) slit-like internal mandibular fenestra; 11) no
surangular ridge; 12) metacarpal I 50% or larger than mcII;
13)metacarpal/phalangeal joints 15degrees or less above the
horizontal; 14) spout-shaped mandibular symphysis; 15) reduced
dental crown curvature; 16) waisted bases of the crowns.

1) This is plain wrong, for in fact, there is a lachrymal recess
   in *Erlikosaurus*, as well as *Gallimimus*.

2) In some theropods, the lachrymal sits atop the jugal instead
   on having a flange laying on the medial side of it, including
   oviraptorosaurs and birds, but it should be clear that most
   maniraptoriforms have the flange including troodontids,
   dromaeosaurids, ornithomimosaurs, as well as tyrannosaurs,
   allosaurs.

3) In most theropods the articulation of the prefrontal and
   frontal forms a sloted articulation; typical fusion or
   problems of identification obviate the identification in some
   theropods like Oviraptoridae, but this is still the
   plesiomorphic condition and doesn't support the group.

4) Does the antorbital fossa invade the jugal to the degree of
   forming a foramen into it? No, but neither in oviraptorids,
   *Caudipteryx*, or birds is this so, nor in troodontids.

5) This is based on relative function and mobility of the neck.
   Most theropods have long cervical ribs which traverse at
   least 1.5 comparative vertebral lengths. The cervical series
   is largely undescribed, but personal observation tells me
   that some oviraptorids did not have this condition. While
   *Beipiaosaurus* has robust cervical ribs proximally expanded,
   they are broken and so are uninformative as to whether the
   condition is inherent to Sereno's group or convergent.

6) This is really irrelevant: many theropods lack a subnarial
   foramen, which is the slot between maxilla and premaxilla,
   including troodontids, oviraptorids, etc.

7) I'm going to assume that Sereno is using the length of the
   maxilla rostral to the antorbital fossa for his rostral
   ramus, which I disagree with; but anyhoo, on the given usage
   and my assumption, this is really based on the extent of the
   antorbital fossa rostrally, which in dromaeosaurids can be
   either short, or long (*Sinornithosaurus* and troodontids
   have very long fossa, short pro-fossoral rami of the
   maxillae; the reverse in *Erlikosaurus*, but ornithomimids
   typically have long maxilla, long fossae, but long rami);
   this is really two characters.

8) Is there a trough-like ventral extent of the antorbital fossa
   in these taxa? There is in troodontids except *Byronosaurus*,
   but not in oviraptorids, *Chirostenotes*, but maybe there is
   one in *Caudipteryx*. There is one in dromaeosaurids, so the
   character is widespread.

9) Ridiculous, the nasal is broken on one side the present on
   the other in complectitude in *Erlikosaurus*, and there was
   in fact a process. Besides, oviraptorids also lack an
   anteroventral nasal process...

10) This isn't even present in ornithomimosaurs, unless Sereno
    is refering to the space between the splenial and ascending
    ramus of the prearticular, which is plesiomorphic for
    theropods, and absent in oviraptorids, caenagnathids and
    presumably in *Caenagnathasia*, *Microvenator*, and some
    various strap-like element in the lower jaw of *Caudipteryx*
    specimens (several of them) may represent both splenial and
    prearticular preserving the same condition in
    caenagnathoids.

11) There is no surangular ridge in oviraptorids, caenagnathids,
    and the condition may or may not be present in
    *Velociraptor* based on fracture in the type skull and bad
    photos in Barsbold and Osmolska, 1999, in _Acta Pal Pol._;
    there is also a slight ridge in *Dromeciomimus*, but that
    could be convergent.

12) This one is also present in caenagnathids as well, though
    *Caudipteryx* lacks it.

13) This one is so functional it's ridiculous.

14) This character refers to the down-curved mandibular ramus,
    as in oviraptorids, *Caudipteryx*, but pretty much lacking
    in caenagnathids though it may still be indicated in the
    form of the ventral margin of the jaw.

15) & 16) These are also known in *Caudipteryx* and birds,
   whereas oviraptorids don't even have teeth and are thus
   incomparable.

  I need to sit down and do a comparison of the other matrices,
but until then, Sereno's support is based on maybe two ambiguous
synapomorphies (5 and 12 above) that could pertain to a group
including segnosaurs and ornithomimosaurs, to the exclusion of
oviraptorosaurs and *Caudipteryx*. But those are shaky. It is
clear that the Oviraptorosauria went through a few reversals
after *Caudipteryx* split off, but these are tenuous, and I can
see an equally parsimonious relationship with *Caudipteryx* in a
trichotomy with segnosaurs and ovis, or outside of them, or
closer to ovis, based on several of these "reversals" or
apomorphies.... We need resolution from the fossil record.

=====
Jaime A. Headden

  Aaaaaaaaaaaaaaaaaaaahhhhhhhhhhhhhhhhhhr-gen-ti-na
  Where the Wind Comes Sweeping Down the Pampas!!!!

__________________________________________________
Do You Yahoo!?
Get personalized email addresses from Yahoo! Mail
http://personal.mail.yahoo.com/