Peters is wrong (surprise) about Herrerasaurus and a new idea about Austroraptor

It's like shooting fish in a barrel, I know, but I felt the need to correct Peters' most recent Photoshop manipulation disaster, since in encroaches on Theropoda.  Peters has a blog which mostly concerns itself with taxa outside my area of expertise, but in this case he tackles Herrerasaurus.  Taking a photo of skull PVSJ 407, he reinterprets it as having a narrow premaxilla subnarial process and elongate antorbital fossa extending to the premaillary-maxillary fenestra (making the latter a promaxillary fenestra I suppose).  Yet that skull's actually featured on Digimorph, where we can use the CT scan to test Peters' idea without the skull in hand.

Picture from Peters' blog, of Herrerasaurus skull PVSJ 407 (taken from Alcober and Martinez 2010 I think) with his interpretation of snout elements on top.  Note his antorbital fossa in dark green.

As you can see, Peters thinks the premaxilla-maxillary suture is a crack in the maxilla.

Herrerasaurus skull PVSJ 407 ct scan taken from Franzosa (2001 online).

Here's the CT scan of the skull, and notice two things.  First is that now that we only have topography to work with, Peters' antorbital fossa disappears.  The maxilla simply isn't depressed there.  Second is that the anterior edge of his supposed maxilla is actually a crack that runs through the premaxilla.  How can we tell it's definitely a crack?  It continues downward through a tooth!  I've highlighted a portion of it in the comparison pic below.

Comparison of two pictures above, with part of crack Peters interprets as a suture highlighted in red.

Just more evidence not to trust Peters' morphological interpretations or his phylogenies based on them.



But that doesn't mean professionals always interpret things correctly.  Austroraptor is a supposed unenlagiine paravian which preserves one pneumatic opening in its maxillary antorbital fossa.  Novas et al. (2009) and everyone else has interpreted it as a maxillary fenestra, which would mean Austroraptor would have lost its promaxillary fenestra like derived troodontids. However, it seems more likely to me to be a promaxillary fenestra.  First note that it is of similar shape, size and position as the promaxillary fenestrae of basal paravians such as Archaeopteryx, Sinusonasus and Velociraptor- small, lenticular, along anterodorsal border of antorbital fossa.  Maxillary fenestrae are usually placed more posteriorly, away from the anterior antorbital fossa edge and close to the antorbital fenestra.  The exception is derived troodontids, but their maxillary fenestrae are highly elongate and placed along the ventral edge of the antorbital fossa.  Second, remember that Austroraptor is odd in having a posteriorly extensive postantral wall (labeled pan in the figure below).  As a bit of explanation, the antorbital fossa surface we usually see on the maxilla is the outside wall of the maxillary antrum.  There's an inside wall running alongside it that sometimes peeks out a bit and is responsible for taxa like Scipionyx that look like they have a maxillary fossa, since the inside wall can back the maxillary fenestra.  This inside wall is called the postantral wall where it sticks out behind the lateral wall.  But another way to think of it is that the lateral wall is missing its back end, and what would be in the back end of that lateral wall if it were there?  A maxillary fenestra.  So I hypothesize Austroraptor reduced the ossification of its posterior antorbital fossa wall, merging the maxillary fenestra with the antorbital fenestra, but keeping an ancestral promaxillary fenestra morphology.  I think this hypothesis is more parsimonious than losing the promaxillary fenestra, moving the maxillary fenestra into the same position as the promaxillary fenestra, then elongating the postantral wall.

Austroraptor maxilla (center, from Novas et al., 2009) compared to dromaeosaurid maxilla (left, from Norell et al., 2006) and mostly troodontid maxillae (right, from Senter et al., 2010).  Promaxillary fenestrae circled in blue, maxillary fenestra circled in green.  For Austroraptor, I've circled the part of the postantral wall in dotted green where I think the maxillary fenestra ancestrally was before the lateral wall was reduced.

I've often said I think that Peters is right insofar as most published cladistic analyses of reptiles suck, in both taxon and character count.  Yet his own analysis is worse because it's based on imaginary morphologies.  The same is true of his constant refrain to test others' interpretations, since scientists do sometimes get things wrong.  But this must be done intelligently.

References- Franzosa, 2001 online. Herrerasaurus ischigualastensis (On-line), Digital Morphology. http://digimorph.org/specimens/Herrerasaurus_ischigualastensis/

Norell, Clark, Turner, Makovicky, Barsbold and Rowe, 2006. A new dromaeosaurid theropod from Ukhaa Tolgod (Omnogov, Mongolia). American Museum Novitates. 3545, 51 pp.

Novas, Pol, Canale, Porfiri and Calvo, 2009. A bizarre Cretaceous theropod dinosaur from Patagonia and the evolution of Gondwanan dromaeosaurids. Proceedings of the Royal Society B. 276(1659), 1101-1107.

Alcober and Martinez, 2010. A new herrerasaurid (Dinosauria, Saurischia) from the Upper Triassic Ischigualasto Formation of northwestern Argentina. ZooKeys. 63, 55-81.

Senter, Kirkland, Bird and Bartlett, 2010. A new troodontid theropod dinosaur from the Lower Cretaceous of Utah. PLoS ONE. 5(12), e14329.