Re: Tyrannosauroids and dromaeosaurs

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

> Yeah.  I just added Anserimimus (codable for 24 characters), and it
changed
> the topology a bit.  Oddly, it's most parsimoniously next to
Compsognathus.

%*)

> I changed the "double-headed quadrate" character into two characters-
> "quadrate contacts squamosal and prootic or opisthotic" and "proximal
heads
> of quadrate widely separated".  Did you know troodontids have the former?

No...

> I wonder why Maryanska and Osmolska (1997) made
> such a big deal about caenagnathoids having it.

Because they have the latter... don't they?

> I deleted the "quadrate not sutured to quadratojugal" character [...]

Good idea.

> I added a "dentary anterior to emf short compared to mandibular height
above
> emf" character, which combines Microvenator and oviraptorids in Makovicky
> and Sues (1998).

Ah!

> Well, now I get-
> |--Carnosauria (inc. Sinraptor, Allosaurus, Monolophosaurus)
> `--+--+--Bagaraatan
>    |  |--Dryptosaurus
>    |  |--Iliosuchus
>    |  |--Fukuiraptor
>    |  |--Stokesosaurus
>    |  `--+--Tyrannosauridae
>    |     `--+--"Chilantaisaurus" maortuensis
>    |        `--+--Eotyrannus
>    |           `--Aviatyrannis
>    `--other coelurosaurs
> Anserimimus fixed it. :-)

Oh, _wow_...

> Nah, segnosaurs are just drawn more basally quite often.  Sometimes as
> arctometatarsalians.  Now Enigmosauria exists again, though Beipiaosaurus
> and Erliansaurus form a clade in the basal Coelurosauria and
Therizinosaurus
> is directly basal to avepectorans.  Stupid segnosaurs, it's so hard
keeping
> them together.  Damn the ghost of Ken Kinman! ;-)

:-D Cool. Well, the pelvis etc. will fix them, I predict.

> As for why avimimids are more derived than Caudipteryx-
> - [...].

Quite a lot, 7 or 8 characters... impressive. Would've thought one or two.
:-)

> And Avimimidae is supported by-
> - premaxillae fused anteriorly.
> - paraquadrate foramen absent.

Sounds good...

> > > `--+--Protarchaeopteryx
> >
> > [...]
>
> Now it's a basal enigmosaur, sister to Segnosauria + Oviraptorosauria.

:-)

> [...] now everything's back to normal in that area,
> except that troodontids are avialan.

Er...
+--dromies
`--+--Archie
   `--+--troodontids
      `--usual birds
?

> I hope this post shows the importance
> of including all relevent taxa in an analysis.

Indeed!!!

Wait.
What does the strict consensus look like? :-) Maybe *Anserimimus* just
reduced the polytomies in some parts of the tree, created new ones in
others, and thus changed the 90 % majrule tree without changing the strict
consensus all that much.

> > Why not *M. gui* and *C.* as sistergroups?
>
> - sternum length (anterior median tip to posterior median tip) width
> (narrowest width across costal margins, or greatest width for those taxa
> without costal margins) ratio > 170.
> - [...].

Cool. But in that character, you're measuring two completely different
widths depending on whether there are costal margins or not.

> New 90% majority rule-
> [...]
>                      `--+--Deinonychus
>                         `--Unenlagia

*Archaeornithomimus* makes *Unenlagia* jump... this sounds like the strict
consensus is pretty unresolved. Is it? :-)

> > >     `--+--+--Noguerornis
> > >         |  |--Boluochia
> > >         |  |--Neuquenornis
> > >         |  |--Gobipteryx
> > >         |  `--Enantiornis
> >
> > How did you get the position of *Boluochia* so well resolved?
>
> I coded it. :-)  For 16 characters.  Is a polytomy with four other taxa
> really well resolved?

Oh yes, considering how it came out in my analyses: in the basal polytomy of
Enantiornithes. OK, the strict consensus with all taxa didn't even have
Enantiornithes, but with dust like *Alexornis* in the matrix that's not
surprising.

> > >   |--+--Ambiortus
> > >   |  `--Apsaravis
> >
> > How come?
>
> - [...].

Quite interesting...

> My analysis doesn't like having Patagopteryx as the most basal
> ornithuromorph.

In mine, it jumped "above" the yanornithids after reweighting, and otherwise
it stayed the 2nd most basal euornithine after *Vorona*.

> Or Songlingornis as a yanornithid.

I always got it somewhere higher up, with *Ambiortus*, *Apsaravis* and
*Liaoningornis*.

> Or yanornithids being monophyletic at all.

Never got them that way -- but I didn't include the sternal characters that
support their monophyly, unlike you!

> Eh, much like the Incisivosaurus vs. Caudipteryx situation, I think people
> see a trend they like and arrange taxa on that (rather Olshevskian).  They
> think everything has to be developing towards ornithomimids.  So they
judge
> a taxon's placement based on tooth reduction and manual digit I
elongation.

Very keen observation!!! :-)