phylogenetic position of Achillobator, Yandangornis, etc.

["Mickey Mortimer" <mickey_mortimer@email.msn.com>]

    With all of the recent discussion of Yandangornis' phylogenetic
position, I thought I'd give my two cents.  How convenient I just finished
my preliminary analysis of coelurosaurs with PAUP.  This analysis includes
274 characters and 41 taxa, including Allosaurus as an outgroup.  The
characters are taken from many papers (Holtz 1994, Sereno 1999, Makovicky
and Sues 1998, Russell and Dong 1994, Xi et al. 1999, Sampson et al. 1998,
etc.) and my own observation.  My old (accidentally deleted) analysis had
330 characters, so I expect to add more to this one soon.  In addition, I
have to check the numerous new papers I got recently, so this is only a
preliminary run.
    I ran a heuristic analysis and the results are 10000 equally
parsimonious trees of 715 steps each.  These have an amazingly high ;) CI of
.3832, a homoplasy index of .6168 and a retention index of .6352.  I will
describe the strict consensus tree here.
    The most basal coelurosaur is Deltadromeus, followed by the
tyrannosaurs.  I used Dryptosaurus, Stokesosaurus, Siamotyrannus, Aublysodon
and Tyrannosaurinae as separate OTU's.  The only resolution in the clade is
that the last two are sister groups.  A surprising result is that
Shanshanosaurus comes out as the sister group to all coelurosaurs besides
Deltadromeus and tyrannosaurs.  A good description would be helpful, of
course, in verifying this.  Next is a polytomy between Gasosaurus,
Ornitholestes, Nedcolbertia and remaining coelurosaurs.  Then a polytomy
between compsognathids, Coelurus, Scipionyx and maniraptoriforms.
Maniraptoriforms splits off into ornithomimids and the maniraptorans, which
divide into the standard segnosaur-oviraptorosaur and dromaeosaur-bird
groups.  The structure of the first of these is (Segnosauria
(Protarchaeopteryx (Caudipteryx (Microvenator (Caenagnathidae +
Oviraptoridae))))).  In the next group (the paravians, technically),
Bagaraatan branches off first, then alvarezsaurids (Alvarezsaurus,
Patagonykus and mononykines were all separate OTU's), then troodontids, then
dromaeosaurids and avialans.  Recently, the monophyly of dromaeosaurs has
been questioned on this list, based primarily on Achillobator.  I included
Sinornithosaurus, Dromaeosaurus, Achillobator, Deinonychus and Velociraptor
as separate OTU's to test this hypothesis.  It turned out to be
monophyletic, with Sinornithosaurus being outside of the rest, as suggested
by Xi et al. (no further resolution occured).  After the dromaeosaurs comes
a polytomy of Rahonavis, Unenlagia, Archaeornithoides and avians.  Note that
the new dromaeosaur synapomorphies of Unenlagia mentioned in the new
Velociraptor paper were not included (they will be later), although the
paper was used to update Velociraptor and some other taxa in my analysis.
Avians form the sequence (Archaeopteryx (Avimimus (Yandangornis
(Confuciusornithidae (Enantiornithines (Patagopteryx + Ornithurae)))))).
Note that this position of Yandangornis largely agrees with what Nick
Longrich has been saying.
    Of course, without the actual data that my analysis is based on, it may
not mean much to many of you.  I'm sure all of you theropod workers out
there have many characters in your databases that I lack in mine (especially
Tom).  However, this does at least give a partial answer to several
questions that have been debated recently (What are Shanshanosaurus,
Bagaraatan, Yandangornis, Achillobator, Avimimus, etc), and some that
haven't but probably should (What is Archaeornithoides? Does no one else
care?!).

Mickey Mortimer