The Enigma of Psittacosaurus

["Steve Brusatte" <dinoland@lycos.com>]

Hey all,
I've been away for a few days (working-coverning the Western Open for the local 
newspaper).  I've returned to find several additional messages, mostly from 
Mickey, regarding Psittacosaurus.  Where to begin?? :-)

Mickey wrote:

>Psittacosaurus mazongshanensis is diagnosed in Xu (1997) by- elongate snout
(ventral surface of mandible formed mostly be dentary); deep buccal recess
>on maxilla; posterior maxillary process projects ventrally; sharp right
angle between ventral and medial margins of jugal; lateral process on dorsal
>rim of predentary; more than fourteen denticles on maxillary crowns; fossa
ventral to exoccipital-opisthotic contact.  Additional characters supposedly
>distinct from other Psittacosaurus species are- large size; mandibles
>Y-shaped ventrally; lateral ridge of postorbital weaker; deeply cupped
articular surface for quadrate on mandible.  I think Xu makes a good case
>for this species being valid.  It has characters 2, 4, 6, 9 and 11 that
diagnose Psittacosaurus, in the order listed by Steve.  It lacks the
>upturned lateral margin of the prefrontal of P. mongoliensis; has an
external mandibular fenestra unlike P. sinensis (and P. youngi, if it's
>valid); the snout is more elongate than other species, opposite of P.
meileyingensis; lacks the anteriorly flat jugal horn of P. xinjiangensis;
>the external mandibular fenestra and antorbital fossa distinguish it from P.
neimongoliensis; it lacks P. ordosensis' maxillary fontanel.  In conclusion,
>I see no reason to support making this taxopn a nomen dubium as Sereno
(2001) suggests.

Thanks for his diagnosis.  After reading this I find it very, very odd that 
Sereno dismissed this species in but a single paragraph in his 2001 paper.  He 
states that it possesses "no clear diagnostic features that are absent in other 
psittacosaur species," and refers it to _Psittacosaurus_ incertae sedis.  Very 
odd.  I have not seen even photos or illustrations of the type specimen, but of 
the above characters, the posterior maxillary process and elongate snout with 
dentary contribution seem like valid characters to me.  The others seem 
reasonable, but I cannot verify nor deny.

>Russell and Zhao (1996) include 13 characters in their Psittacosaurus
   phylogenetic analysis-
   1. skull rectangular with long antorbital region (0); intermediate (1);
>rounded with short antorbital region (2).

A potential valid character.  There certainly seems to be significant 
differences between both the shape of the skull and length of the snout in 
different species.  _P. meileyingensis_ has a rounded skull with a snout that 
makes up 27% of the skull length, and other species do show a rectangular and 
intermediate dichotomy, IMHO.

>2. antorbital fossa very shallow or absent.

Maybe a better way of doing this would be an ordered character, as Alessandro 
Marisa mentioned a few days ago.  Something like:
Antorbital fossa present (0), present as a shallow or "secondary" depression a 
la Sereno (1), or absent (2).  With outgroups that all possess antorbital 
fossae, its absence would definitely be a derived character.  This is one of 
the few characters, IMHO, that doesn't suffer from an unknown polarity.

>3. ten or less maxillary and dentary teeth.

Bad character, in my opinion.  The work of Coombs and Sereno has shown that 
tooth count is extremely variable throughout ontogeny.  It might have been best 
if they left this out, as there appears to be no set pattern in psittacosaur 
tooth count.

>4. lateral margin of prefrontal strongly upturned.

I thought this was an apomorphy of _P. mongoliensis_.  Is this shared by any 
other psittacosaur species?

>5. postorbital region broad

Hmmm...another possible bad character.  The breadth (is that the word??) of the 
postorbital region might be functionally correlated with the size of the snout. 
 More work probably needs to be done on this.  Also, the paucity of material in 
some species makes this somewhat difficult to document. 

>6. horizontal ridge strongly developed on postorbital.

Don't know enough about this character, but, based on illustrations, the ridge 
appears to be more strongly developed on _P. mongoliensis_ and _P. youngi_ than 
on _P. sinensis_ and _P. meileyingensis_.  But, this could be an artifact of 
either incomplete material or poor illustration...or both.

>7. anterior squamosal process extends to anterior margin of supratemporal
   fenestra.

Once again, I don't know enough about this character except for the 
illustrations I have seen.  It appears as if _P. youngi_ has a squamosal 
process that extends to about the midpoint of the supratemporal fenestra, while 
the process of _P. mongoliensis_ extends to the anterior margin.  It's hard to 
tell in _P. sinensis_ and _P. meileyingensis_.

>8. laterally projecting jugal horn.

A good character that has been recognized since Sereno's paper in _The 
Dinosauria_.  The polarity is known, which makes it valuable in deducing 
psittacosaurid relationships.

>9. posterior quadrate margin deeply sulcate.

Illustrations of all species show a sulcate quadrate margin, unless I am 
misinterpreting the character. I would have to see the original material to 
even begin to study this character in more detail.

>10. external mandibular fenestra absent.

IMHO, another good character that has much potential.  I've already discussed 
this one in previous posts, so I won't go into detail on it here.

>11. ventrolateral dentary ridge.

Too hard to tell from illustrations.

>12. posteroventrally angled primary ridge of maxillary teeth.

A tooth character that might be valid, but, once again, too hard to discern 
from the papers I've read.

>13. metatarsal I >65% of metatarsal III.

Ah, a postcranial character.  I like this.  I'll have to look into this one 
better.

>I think 5 is useless, as it's just an apomorphy of P. mongoliensis.  Also, I
   would recommend quantifying several characters, but I have coelurosaur
>characters to worry about, someone else can refine psittacosaur phylogeny.
   :-)

Maybe it's an apomorphy of _P. mongoliensis_, but other species appear to have 
very similarly broad postorbital regions.  As I stated above, I think it is 
perhaps useless because it might be correlated with the first character in 
Russell and Zhao's list.  And, as for coelurosaurs, why are you spending your 
time with *theropods*? :-) (apologies to the theropod lovers, which make up 
about the entire list)

>P. mazongshanensis codes the following way-
                     111
           1234567890123
   outgroup0000000000000
   P.mongo 0101001000111
   P.sinen 1010111111000
   P.meile 211001101011?
   P.xinji ?11??0110??1?
   P.neimo 1010110001011
   P.ordos 101?1??100101
   P.mazon 00?010?00011?

Thank you so very much.  I'll look at this much better when I can actually sit 
down and think (not likely anytime soon :-))

   Running this through PAUP yields 3 most parsimonious trees, with the
   following topology-
   |-mazongshanensis
   `-+-mongoliensis
     `-+-meileyingensis
       |-xinjiangensis
       `-+-neimongoliensis
         |-ordosensis
         `-sinensis
   So if you have some urge to split up Psittacosaurus, I recommend making P.
   mazongshanensis a new genus, though I don't think it would be very 
   useful.

Well, as I've said before, it's not quite the splitting into different genera 
that's useful, but recognizing possible interrelationships within the current 
genus _Psittacosaurus_.  If it is (ever) possible to recognize two or three 
distinct clades, naming a new genus (while we're still using the term genus) 
might clear up a lot of confusion.  I mean, Sereno's suggestions aside, there 
may very well be 10 or 12 valid species of _Psittacosaurus_.  If we can group 
some of these together, recognizing separate genera might enable us to remember 
all of these species :-)

>I should have coded P. xinjiangensis as "0" for character 11 (ventrolateral
   dentary ridge) based on Brinkman et al. (2001).  This destroys all structure
>in the consensus tree.  Even in the old tree, the clades had terrible
   bootstraps (19% for everything but mazongshanensis; 32% for everything but
>mongoliensis and mazongshanensis; 22% for sinensis + ordosensis +
   neimongoliensis).  Clearly more characters are needed to determine the
>interrelationships of Psittacosaurus species.

Oh crap.  Well, that's no good.  More characters are certainly needed.  I'm 
surprised, for instance, that Russell and Zhao didn't include any characters 
relating to the dentary flange and the jugal-lacrimal-maxilla-premaxilla joint. 
 Also, the subtraction of "bad" characters, such as those relating to tooth 
count and the postorbital region might also elucidate some interrelationships.  

We also have to remember that, when Russell and Zhao conducted their analysis, 
_Liaoceratops_ was still in the ground and _Chaoyangsaurus_ was just an 
oft-talked about, but never described mystery genus.  The basal position of 
these two taxa might (I'm optimistic at this point) help determine the polarity 
of a few additional characters.  

Now, for a brief foray into heterodontosaurids...

>The question is "has there been a solid cladistic analysis that has shown
   heterodontosaurids are ornithopods?".  I think the only ornithischian
>phylogenetic analyses with this conclusion have been Sereno's, which did not
   test a marginocephalian position for heterodontosaurids, as they left out 
potential >synapomorphies.

Ah, good point.  I'm certainly not as well versed in the literature as most on 
this list, but am very familiar with Sereno's analysis.  That has likely led to 
my confusion.

>Sereno (1986) gave four synapomorphies linking heterodontosaurids with
   ornithopods...
>4. premaxilla contacts lacrimal.
   Absent in Agilisaurus, "Agilisaurus" multidens, Jeholosaurus Hypsilophodon,
>Tenontosaurus and Parksosaurus, but present in Psittacosaurus.

Absent, as far as I can tell, in _Stegoceras_.  Seems to be a possible derived 
character of the Marginocephalia that was lost in both pachycephalosaurs (at 
least most forms) and neoceratopsians.  But, I'm unsure of the state in 
_Chaoyangsaurus_.  Or, it could be that it is a primitive character of the 
Ornithischia (or a less inclusive group) that was modified in pachycephalosaurs 
and independently in neoceratopsians.  That might be more likely.

And, for one final point to end this unbelievably long post, awhile back it was 
mentioned on list that _Hypsilophodon_, _Iguanodon_, and _Stegoceras_ have 
coronoids, along with the neoceratopsians _Protoceratops_ and 
_Montanaceratops_, while _Chaoyangsaurus_ and _Psittacosaurus_ lack them.  How 
much potential does this have as a possible psittacosaurid + _Chaoyangsaurus_ 
synapomorphy.  What other possible synapomorphies link the two?

Steve

---
***************************************************************
Steve Brusatte-DINO LAND PALEONTOLOGY
SITE: http://www.geocities.com/stegob
ONLINE CLUB: http://clubs.yahoo.com/clubs/thedinolanddinosaurdigsite
WEBRING: http://www.geocities.com/stegob/dlwr.html
INTERNATIONAL LANGUAGE SITE: http://www.geocities.com/stegob/international.html
****************************************************************


_____________________________________________________
Supercharge your e-mail with a 25MB Inbox, POP3 Access, No Ads
and NoTaglines --> LYCOS MAIL PLUS.
http://www.mail.lycos.com/brandPage.shtml?pageId=plus