Tuesday, October 5, 2010

That supposedly indeterminate Coelophysis type material

The Kayentavenator study encouraged me to clean up and flesh out the coelophysoid portion of my website.  In doing so, I found something interesting.

By now, everyone knows the history of Coelophysis.  Cope based the genus on a collection of disarticulated elements found in 1881, but in 1947 the famous Ghost Ranch* bonebed of hundreds of specimens was found, which Colbert referred to Coelophysis.  Since then, the Ghost Ranch specimens have formed the basis for our understanding of Coelophysis bauri.  Hunt and Lucas (1991) stated Cope's material is indeterminate and created the new taxon Rioarribasaurus colberti for the Ghost Ranch taxon.  Colbert et al. (1992) petitioned the ICZN to make a Ghost Ranch specimen the neotype of Coelophysis, which was accepted by the ICZN in 1996.

* Why is this now being called the Whitaker quarry, by the way?  It seems pointless to rename such a well known locality.

Now Cope's Coelophysis material was referred to three species, C. longicollis, C. bauri and C. willistoni.  Hay (1930) designated bauri as the type species without comment.  After a flawed attempt by Welles (1984), Colbert (1989 for C. bauri) and Hunt and Lucas (1991 for C. longicollis and C. willistoni) were the first to actually designate lectotype specimens from Cope's syntypes.  C. longicollis' lectotype is the cervical vertebra AMNH 2701.  Hunt and Lucas state-

"The lectotype of C. longicollis is not diagnostic beyond the level of a primitive theropod, and we consider the taxon a nomen dubium. In so concluding, we emphasize that whichever of COPE'S syntypes of C. longicollis is chosen, it a nomen dubium, because none of the syntypes are diagnostic."

Lectotype of Coelophysis longicollis (AMNH 2701)- anterior cervical vertebra in lateral (a), ventral (b) and anterior (c) views (after Huene, 1915).  Note anterior pedicular foramen to left of neural canal in c.

Just as with most declared nomina dubia, a brief statement without any comparison is deemed sufficient to make a taxon indeterminate.  Yet anyone who saw AMNH 2701 could note it's certainly not just a primitive theropod.  If nothing else, its elongation is only seen in derived coelophysoids.  Beyond that though, today I noticed it has anterior pedicular foramina.  These are so far only known in Coelophysis bauri, and are absent in "Megapnosaurus" kayentakatae and the Shake-n-Bake coelophysid for instance.  Tykoski (2005) considered them autapomorphic for C. bauri, though he noted the condition in Megapnosaurus is unknown due to the brief description.

Interestingly, C. longicollis was based on more specimens than C. bauri, which were more complete than those of the latter species, was larger, and was described first in Cope's paper.  It could have easily ended up being the type species of Coelophysis if not for Hay's seemingly arbitrary decision.  And in that case, the type of Coelophysis would have been diagnostic after all, making the neotype designation unecessary.  As it is, Coelophysis and its type species bauri are both based on a complete Ghost Ranch skeleton now, so it doesn't even matter if the bauri lectotype (an incomplete sacrum and ilial fragment) are diagnostic.  But longicollis can now be made a synonym of bauri at least with evidence besides the assumption all of Cope's material belonged to one taxon.


  1. Why is this now being called the Whitaker quarry, by the way? It seems pointless to rename such a well known locality.

    Because there are multiple quarries within the rather expansive grounds of Ghost Ranch.

  2. Why is this now being called the Whitaker quarry, by the way? It seems pointless to rename such a well known locality.

    Think about it for a second. The answer is in your first paragraph.

  3. Jerry, In 1991 there was only one fossil quarry at Ghost Ranch. The Canjilon Quarry is on Forest Service property adjacent to Ghost Ranch. Alex Downs