[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Subject Index][Author Index]
Re: Coelurosaur analysis update
Jaime Headden wrote-
> <For what? Bagaraatan in Tyrannosauroidea, or tyrannosauroids being so
> basal? Both occur in almost every run of my analysis.>
>
> And the maniraptoriform, maniraptoran or more avian features features
> that Osmólska noted in *Bagaraatan* as well as presented otherwise here,
> are apparently absent, I gather?
Of course. The nine characters you listed are all hindlimb features, so
they have yet to be included.
> These indlude the raised femoral head,
> fused greater and anterior trochanters, an anterior femoral crest below
> the anterior trochanters, femoral caput dorsal to trochanteric crest
> formation of a tibiotarsus with fustion between distal tibia and fibula
> with both the astragalus and calcaneum, and fusion of the calcaneum to the
> astragalus, the fibula less than 1/10th the distal tibial width above the
> distal epiphysis, cnemial crest projects laterally and not cranially,
> absence of a rounded fossa on the medial fibular head.
The first and fourth seem identical (my character 241) and are found in
tyrannosauroids anyway. The others are my characters 244, 247, 254, 257,
262, 263 and 283. As I just finished coding 191, I have a ways to go until
I get to them.
> Access into the
> foramen for the chorda tympani on the medial and caudal surface of the
> articular, noted by Mortimer earlier, is also found in *Tyrannosaurus*, as
> well as caenagnathids, so this is not exclusively tyrannosaurid;
> similarly, the "large surangular foramen" is not as in tyrannosaurids, but
> is relatively far smaller and apparantly either paired, with a column of
> bone dividing the aperture into two, or the rostral surangular foramen is
> conjointed into it, likely as both seem to form at the margins of two
> separate fossae; either way, this condition appears to be less than
> homologous with that of tyrannosaurs. The tibia and femur in comparison
> are virtually identical to troodontids, including the precence of a large
> lateral "cnemial" crest (not homologus with the avian crest, which arises
> from the later surface of the cranial crest, cranial to the incisura
> tibialis versus caudal to it, as in *Bagaraatan* and troodontids,
> mononykids, and smaller dromaeosaurids; in tyrannosaurids, this is a
> smaller, cranially oriented crest, as in *Deltadromeus* and
> *Dryptosaurus*, while that in maniraptorans and *Bagraatan*, it is larger
> and projects laterally. As in maniraptorans, the tibial shaft is strongly
> bowed laterally, with both medial and lateral margins forming distinct
> arcs, but this is not the case in tyrannosaurids, which show a shallower
> medial arc and a nearly absent lateral arc (though this may relate to
> size). The medial fibular fossa in tyrannosaurs is absent in Baga. The
> caudal edge of the distal tibia expands caudally, but not in tyrannosaurs,
> resembling the condition in troodontids and mononykids. The proximal iliac
> peduncle of the pubis shows a lateral fossa on the caudal end that appears
> to have serves the embayment for a split pubic peduncle of the ilium, as
> in some maniraptorans, but not in any non-maniraptoran. As in
> maniraptoriforms, the ventral margin of the postacetabular ala is
> ventrally curved, but this is absent in tyrannosaurids.
We all know a character shouldn't be rejected just because it has a
consistancy index of less than one. I'm not Sereno, my characters can't be
organized to have only evolved once in each clade. The relative size and
division of Bagaraatan's posterior surangular foramen is of little
consequence to its homology- the relative size varies within Tyrannosauridae
and could be size-related, while tyrannosaurids also have a smaller foramen
anterodorsal to the main one (Carr, 1996). And yes, the hindlimb has some
derived characters. I could go and list off primitive characters of the
tail that aren't found in eumaniraptorans like troodontids, but it's all
useless until we can see which convergences are deemed most parsimonious by
PAUP.
> While I do not disblieve that *Bagaraatan* is a tyrannosauroid, as even
> Holtz appears to be finding this, I would like to know his this would
> possibly affect, if included, the majority of non-maniraptoriform
> features?
I can only assume Holtz included the characters you listed. He does have a
huge matrix, and those characters have been recognized for some time now.
> <Including Yandangornis changes the Aves section to-
>
> |--+--Scansoriopterygidae
> | `--+--Rahonavis
> | `--+--Archaeopteryx
> | `--Wellnhoferia
> `--+--Shenzhouraptor
> `--+--Jixiangornis
> `--+--Yandangornis
> `--+--Sapeornis
> |--Confuciusornis
> |--Changchengornis
> `--+--Protopteryx
> |--Jibeinia
> `--Ornithothoraces >
>
> So what characters link *Shenzhouraptor* and *Jixiangornis*, and what
> about my earlier argument of possibly testing *Shenzhouraptor* monophyly
> by using *Jeholornis* as a separate taxon?
It's hard to say now, as the addition of more characters since last time has
changed the tree-
|--+--Scansoriopterygidae
| `--+--Archaeopteryx
| `--Wellnhoferia
`--+--Rahonavis
`--+--Shenzhouraptor
`--+--Jixiangornis
`--+--Sapeornis
`--Pygostylia
I keep Shenzhouraptor and Jeholornis as separate OTU's, I just also have a
"Shenzhouraptor + Jeholornis" OTU I've been using.
> <No, the antarticular is not a character in my analysis.>
>
> Ah, good. Or not ... who knows. Perhaps including it may change things.
I doubt it. It's only known in Allosaurus and Bagaraatan, and the former is
in the outgroup. So there would be no chance of them ever being sister taxa
in my analysis.
> <Note Holtz (2000) also found Bagaraatan to be a basal coelurosaur, so
> this position is not "never found".>
>
> Which was Osmólska's general conclusion.
Actually, Osmolska (1996) preferred a tentative placement as the sister
group to Avetheropoda/Neotetanurae.
> <Though the hindlimb has some derived characters, the tail is primitive as
> well.>
>
> Except for the extremely thin walls and strongly defined ventral keel,
> as in mononykids.
Perhaps. The centra are said to be narrow ventrally, but not necessarily
keeled (the more distal ones certainly aren't). Shuvuuia only has keels on
the proximal couple caudals anyway, which may not be preserved in
Bagaraatan. Thin-walled caudals are known in basal taxa too (eg. Coelurus)
and are possibly size-related.
> <While the data is premature in the sense it hasn't been "completed" yet,
> calling it misleading is, well... misleading. It's not like 186 cranial,
> axial and pectoral characters are inherently worse than 186 characters
> from the entire skeleton.>
>
> Selective portions of a skeleton may evolve at different rates and, as
> alluded to by Sampson for ceratopsids, may be almost entirely absent
> depending on the region selected for; that one can say it cannot be as
> worse than if one choses portions from all over the skeleton would be like
> only analyzing the cranial characters from Holtz' GAIA analysis. This
> provides a distinct result than the whole thing, and in retrospect, would
> serve to wait for a completed anatomical selection prior to running the
> matrix. It is likely that even 10 pedal characters would change the matrix
> significantly because some taxa are united almost solely on pedal
> autapomorphies. Which is why I favor relativity among the skeleton, rather
> than completeness in a single section.
But while chances are good some clades will be diagnosed largely on
characters from one part of the skeleton, this isn't necessarily true.
There are probably also clades diagnosed by characters from several areas of
the body. It's all very random, and no mix of skeletal areas sampled is any
more likely to generate "correct" trees, given the same amount of
characters. Analyzing only the cranial part of Holtz's analysis would
probably be equivalent in "accuracy" to analyaing the same number of
characters from all over. I'm not presenting my results as being conclusive
anyway, they are largely so others can follow my progress, and see how the
results change, as well as to generate discussion.
> `--+--+--Caudipteryx zoui
> | `--+--Nomingia
> | `--+--Caudipteryx sp. nov.
> | `--Avimimus
> `--+--Incisivosaurus
> [ `--other oviraptorosaurs ]
>
> The avimimid-caudipterygid clade is diagnosed by-
> 1 - maxillary teeth absent (unknown in Nomingia and Avimimus)
> 2 - dorsal premaxillary process extends past antorbital fossa (unknown in
> Nomingia and Avimimus)
> 3 - posterodorsal dentary process >8% of mandibular length (unknown in
> Nomingia and Avimimus)
> 4 - dentary toothless (unknown in Nomingia)
> 5 - number of caudal vertebrae less than 24 (unknown in Avimimus)
> 6 - most elongate caudal prezygopophyses over 40% of centrum (unknown in
> Avimimus)
> 7 - caudofemoral ratio less than 1.9 (unknown in Avimimus)
> 8 - posterior processes of distal chevrons formed from a new
> posteroventral process (unknown in Avimimus)
> 9 - anterior edge of acromion process does not contact coracoid (unknown
> in Nomingia and Avimimus)
>
> Sadly, of the 9 characters used, 8 are coded as absent in *Avimimus*,
> leaving support down to character 4, which is found in all other
> oviraptorosaurs save *Incisivosaurus*; the relative region of the dentary,
> mostly missing in *Avimimus*, may be contingent between the two, as both
> lack symphyseal dentition, and so I am not sure any of this means much
> given the data. Character "8" is unelaborative, and I think the shape of
> the chevrons from progression from *Allosaurus* to *Deinonychus* indicate
> there is no real "new" process. However, if Mickey could explain this
> process...?
Hey, I didn't design this topology. It's the most parsimonious with the
current data. In fact, only with the old data. The current data supports
Caudipteryx as basal within Oviraptorosauria, Avimimus next, then
Incisivosaurus, Protarchaeopteryx, Microvenator, Nomingia and caenagnathoids
in a polytomy.
Good point regarding dentary toothlessness in Avimimus, I'll revise the
character and codings.
There is indeed a new process in Caudipteryx compared to Allosaurus.
Allosaurus just expands and sweeps the distal end posteriorly to form
elongate chevrons, while in Caudipteryx, the main body of the chevron bends
anteroventrally and a new process (formed by the dorsoventral flattening of
distal chevrons) is responsible for the posteroventral process.
> <Caudipteryx sp. nov. and Avimimus are sister taxa based on-
> - dorsal pleurocoels absent (tentatively scored for C. sp. nov. based on
> Zhou et al. 2000)>
>
> As I beleive the anterior dorsal in *Avimimus* do have small
> pleurocoels, and the condition in *Caudipteryx* so far unknown or
> uncertain, I think this character should be re-evaluated, perhaps split
> into regional conditions with variation in size of pleurocoely.
Nope. Avimimus is completely without dorsal pleurocoels (Makovicky, 1995).
Caudipteryx's dorsals were said to "lack deep pleurocoels" by Zhou et al.
(2000). My dorsal pleurocoel character is split up into multiple parts
(four states), but I only listed the applicable state above.
> However, I am also curious about the characters that do not permit
> *Incisivosaurus* from being a more basal oviraptorosaur, or rather, which
> relate it closer to caenagnathoids than is *Caudipteryx*? Based on the
> tree provided, the ancestor would either have had to re-evolve teeth, or
> the two clades reduced their dentition convergently. Has *Microvenator*
> been included in any analysis, perchance?
Surely losing teeth multiple times isn't unlikely, given Shuvosaurus,
ornithomimids, confuciusornithids, Gobipteryx, neornithines, etc..
Microvenator is in my analysis, and was listed in my original post. In my
current trees, the clade containing Incisivosaurus and caenagnathoids, but
not Caudipteryx and Avimimus, is diagnosed by-
- subnarial process of premaxilla contacts nasal.
- nasal recesses (based on Holtz's coding for Caudipteryx).
- palate mostly ventral to maxilla and jugal (based on shape of palatal
bones in Caudipteryx).
- naris completely above maxillary fenestra (optimization due to DELTRAN).
- intramandibular joint reduced.
- splenial extends to mandibular symphysis.
- jaw articulation of mandible extremely convex dorsally.
- dorsal pleurocoels present in all dorsal centra (based on
Protarchaeopteryx and Nomingia that are in a clade with Incisivosaurus in
the tree I'm describing).
- pleurocoels in proximal caudal vertebrae (again, based on Nomingia).
- number of caudal vertebrae with transverse processes more than sixteen
(Nomingia again).
- radial-metatarsal ratio >80 (Protarchaeopteryx this time).
I also noticed the plesiomorphic triangular splenial has been transformed in
Incisivosaurus, caenagnathids and oviraptorids, but not Caudipteryx. But
this character isn't included yet.
Mickey Mortimer