Tuesday, July 26, 2011

Epichirostenotes and how it changes coding

Sullivan et al. (2011) just named two new caenagnathids, Ojoraptorsaurus boerei for a partial pubis and Epichirostenotes curriei for ROM 43250, a specimen from the Horseshoe Canyon Formation described as Chirostenotes by Sues (1997).  Jaime Headden wrote a great post on the situation here and I agree with his assessment.  I'm extremely doubtful Ojoraptorsaurus can be diagnosed, given the large amount of interspecific variation in most theropods and the low number of comparable caenagnathids.  But that's not too important considering it's just a partial pubis.  My main concern here is a more practical one about Epichirostenotes.

Ischia previously referred to Chirostenotes (from Sues, 1997).  A- RTMP 79.20.1, which is still placed in Chirostenotes pergracilis. B- ROM 43250, now the holotype of Epichirostenotes curriei.  The only other shared elements are ilial fragments and a sacrum for which no differences have been noted.

How should we now code Chirostenotes?  The genus has been used in many analyses, including the Theropod Working Group matrix, which I've been correcting for the description of a new paravian.  In that matrix, it is apparently based on a combination of the Chirostenotes, Macrophalangia and Caenagnathus holotypes, as well as partial skeletons RTMP 79.20.1 (Currie and Russell, 1988) and ROM 43250.  It wouldn't be too important, except ROM 43250 is our only described source for coding Chirostenotes' maxilla, braincase, cervicals, dorsals, caudals and pubis.  So if we leave it out, we're losing a lot of information.  And without a more complete caenagnathid to code, and few overlapping elements, there's a large chance it won't group with Chirostenotes, so could harm oviraptorosaur topology (like coding Caenagnathus separately did for Senter, 2007).  But combining it with Chirostenotes could be misleading once Elmisaurus and Hagryphus are added, since nothing argues ROM 43250 isn't closer to these latter genera.  While ROM 43250 isn't comparable to either, a second specimen from the same formation (metatarsal II CMN 9570, unmentioned by Sullivan et al.) is distinctive from Elmisaurus due to its lack of fusion and straight distal end.  These are plesiomorphies though, so do not argue strongly for its referral to Chirostenotes.  And the metatarsal isn't comparable to ROM 43250 anyway, so any referral to Epichirostenotes is based purely on size and provenence.  But nothing argues the Horseshoe Canyon specimens aren't Chirostenotes either, since so far the apomorphies of Chirostenotes are only known from the mandible (assuming Caenagnathus collinsi is properly synonymized) and other parts of the metatarsus.  The minor differences between the ischia of Epichirostenotes and RTMP 79.20.1 are just like those found between individuals of Tyrannosaurus rex, Microraptor zhaoianus and other species.  The distinction boils down to size (which could easily be ontogenetic) and provenence.  Now this doesn't make Epichirostenotes a nomen dubium, because the braincase and maxilla are certainly distinct from comparable named theropods, it's just that no other named caenagnathid can be compared for the most part. 

Tyrannosaurus rex ischia to show interspecific variation.  Left- FMNH PR2081 left and right (from Brochu, 2003); Right top- CM 9380 (from Osborn, 1906); Right bottom- AMNH 5027 (from Osborn, 1916).

Note how the Tyrannosaurus ischial variation mirrors that between Chirostenotes and Epichirostenotes.  The latter's ischial diagnosis is-
1. "Ischium long and expanded posteriorly; broad behind obturator process."  There is no known comparative length difference, since Epichirostenotes doesn't preserve an ilium and Chirostenotes doesn't preserve a pubis.  The posterior expansion and breadth posterior to the obturator process refers to the greater depth in Epichirostenotes.  As seen most distinctly between the upper right and lower left pictures above.
2. "obturator process expanded and triangular." This refers to the greater depth of the obturator process and non-rectangular tip in Epichirostenotes.  As seen between the two left pictures above.
3. "having a prominent anterior proximal hook-like projection." This refers to the pubic peduncle being longer and more ventrally projecting in Epichirostenotes, enclosing the obturator notch more.  Note the two right pictures above also differ in the length and angle of the ventral portion of the pubic peduncle.
Also notice the other differences within Tyrannosaurus.  The lower right pic has a narrower body proximal to the obturator process.  It also has a longer and more rounded proximodorsal process.  The two right pics have curved shafts compared to those on the left.

So it's a very awkward situation.  I'd bet curriei really is a distinct species from pergracilis, if only because that seems to be true for other theropods compared between those formations (though the published evidence for this is usually as tenuous as the current case).  But the known remains don't justify that anatomically, and whether curriei is more closely related to Chirostenotes than Elmisaurus or Hagryphus is basically unknowable.  If it is closer to pergracilis, then I could see only the most extreme splitter keeping them as separate species based on known remains, let alone genera.  Maybe CM 78000 and 78001 will help sort things out, and it would have been prudent to wait until Sues and Lamanna finish describing them before dealing with the taxonomy of ROM 43250.  I think for now I'll just call ROM 43250 Chirostenotes? curriei and keep the Chirostenotes OTU intact that way.  There's no evidence against it at least, even though it's not necessarily true.  And it's not unheard of for other taxa in the matrix.  The Struthiomimus OTU has always included both the Dinosaur Park S. altus and the unnamed Horseshoe Canyon species represented by AMNH 5257 and exceptionally complete RTMP 90.26.1.  If someone were to name the latter "Epistruthiomimus unnecessari" based on Longrich's comment the metacarpus is more slender, would we have to break up our Struthiomimus OTU?  Note that like the caenagnathid situation, the described differences between species are few and of debatable significance, while there has been no study showing they form a monophyletic group based on apomorphies.  So the facts are the same, but the new genus for ROM 43250 just complicates matters.


  1. I agree: until the Triebold and the Tyrrel new caenagnathids are described, in my analysis Epichirostenotes is part of the Chirostenotes OTU.

    1. Hi Andrea,

      The Triebold caenagnathid has been named Anzu wyliei (Lamanna et al. 2014). As we all know, Anzu is the only North American caenagnathid based on complete skeletons. In any case, since Anzu only confirms that Caenagnathus and Chirostenotes are in the same clade, and we now know that a faunal turnover occurred in the lower Horseshoe Canyon Formation for ornithischians, Sullivan and his colleagues made the right decision to remove ROM 43250 from Chirostenotes by virtue of being a few million years younger, and Longrich and others have also recognized Epichirostenotes from Chirostenotes proper.

      Lamanna MC, Sues H-D, Schachner ER, Lyson TR (2014) A New Large-Bodied Oviraptorosaurian Theropod Dinosaur from the Latest Cretaceous of Western North America. PLoS ONE 9(3): e92022. doi:10.1371/journal.pone.0092022

    2. First, the Anzu specimens can hardly be called complete skeletons. Both of the CM specimens are partial skeletons that combine to cover most areas of the anatomy, though the skull is still partially known and the feet fragmentary.

      Second, Epichirostenotes still isn't morphologically distinct from Chirostenotes. Everything I said in the post is still true after the publication of Anzu. Is it the "right decision" to name new species based on stratigraphy instead of morphology? I think the majority of dinosaur workers would say no. The general philosophy used in modern paleontology is that taxa have to be distinct morphologically from other taxa, not just separated by stratigraphy or geography. We'd be just as justified naming e.g. the Antarctic tetanurine distal tibia (Molnar et al., 1996) because it's widely separated stratigraphically from any similar taxon. It's near certainly not referrable to a known species based on that alone, yet no one advocates naming it because morphological autapomorphies have yet to be suggested. So the reality is that dinosaur workers view taxa as valid when they're morphologically diagnostic, not just when they're likely to be new species on other grounds. Not that this means dinosaur workers will ignore Epichirostenotes, because the reality is also that they generally follow the status quo.

  2. I think Sullivan et al. (2011) definitely have the right idea, but the material they are basing their claims on is highly insufficient.

  3. I didn't even get into the whole biostratigraphy angle that Sullivan, Jasinowski and van Tomme went into. That's one of the strongest reasons, I think, that they placed ROM 43250 into a new taxon. It's a strongly-slanted argument designed to push the concept of early Maastrichtian taxa as different from late Campanian taxa, given there is a "faunal turnover" between the two formations. Ojoraptorsaurus-rex-titan-mimus has more grounding, and even then ... there's the issue of comparability.