Monday, May 26, 2014

Is Thecocoelurus an ornithomimosaur?

Allain et al. (2014) recently claimed Thecocoelurus, a small partial theropod cervical vertebra from Early Cretaceous England, is "morphologically identical" to an undescribed supposed ornithomimosaur from the Early Cretaceous of France (from here on called the Angeac taxon).  After reading Andrea Cau's post on the topic, I decided to do some in depth comparison.

Besides the Angeac taxon, Thecocoelurus has been compared to several taxa with surprisingly similar cervical vertebrae- caenagnathids (Naish et al., 2001; Naish and Martill, 2002); Falcarius (Kirkland et al., 2004; Zanno, 2010a); and noasaurids (Naish, 2011).  These all have elongate amphicoelous cervicals with low neural spines, anterior peduncular fossae, two pairs of pleurocoels and a transversely concave ventral surface defined by lateral ridges confluent with the parapophyses.  Which is most similar to Thecocoelurus?

Naish et al. (2001) and Naish and Martill (2002) both argue it is closer to oviraptorosaurs than to therizinosauroids based on the rounded pleurocoel and thin neural spine, but this is also the case in basal therizinosaurs (Falcarius, Jianchangosaurus) (as noted by Zanno, 2010a) and noasaurids. Naish (2011) on the other hand, felt "the idea that large caenagnathids were present in Western Europe during the Barremian is difficult to reconcile with what we know of oviraptorosaur biogeography and distribution", thus favored an abelisauroid identity (I suppose based on Genusaurus).  I'd argue basal oviraptorosaurs could and do have similar morphologies (e.g. the Early Cretaceous Similicaudipteryx) and that small theropod diversity in Cretaceous Europe is very poorly known. 

Allain et al. make two new comparative arguments for Thecocoelurus being closest to the Angeac taxon.
1. They distinguished both from Noasaurus based on their concave anterior central surface, but this is true in Masiakasaurus as well. It is also true of all coelurosaurs being compared.
2. They also paired Thecocoelurus with the Angeac taxon based on pneumatic foramina above the prezygapophyses which invade the neural arch. Yet these are present in cervicals 6-10 of Masiakasaurus (Allain et al. state they are "in a modified form" but don't elaborate), at least cervical 9 of Heyuannia, and in Conchoraptor and "Ingenia" (Lu, 2004; contra Allain et al.'s claim they are unknown in oviraptorosaurs). They are also only present on the left side of Thecocoelurus (pf? in figure below, lower left), further posterior than in at least Masiakasaurus and the Angeac taxon, perhaps suggesting breakage of a naturally hollow area or pneumatic asymmetry. The only preserved posterior cervical of Falcarius doesn't preserve this area, nor does the illustrated and best preserved cervical of Chirostenotes.

So let's compare!  Contra Naish and Martill, the specimen resembles posterior cervicals more than anterior ones, and indeed the supposedly identical Angeac vertebra matches the seventh or eighth of Harpymimus based on elongation, central articular surface orientation, prezygapophyseal length and orientation, etc. (contra Allain et al.).

Anterior half of posterior cervical vertebrae in (descending) right lateral, anterior, ventral and dorsal views.  From left to right- Thecocoelurus holotype (after Naish and Martill, 2002); Angeac taxon ANG 10/175 (after Allain et al., 2014); Falcarius UMNH VP 14657 (after Zanno, 2010b); Masiakasaurus FMNH PR 2481 (after Carrano et al., 2011); Chirostenotes or Epichirostenotes ROM 43250 (after Sues, 1997); Similicaudipteryx holotype (after He et al., 2008).

The elongate parapophyses resemble Falcarius, Chirostenotes and Similicaudipteryx more than Masiakasaurus or the Angeac taxon.
The anterior pleurocoels are placed in an obvious fossa, like Falcarius and the Angeac taxon, but unlike Masiakasaurus, Chirostenotes or Similicaudipteryx.
The infradiapophyseal fossa is developed as an elongate groove, as in Falcarius but unlike Masiakasaurus, the Angeac taxon, Chirostenotes or Similicaudipteryx.
The centrum is taller than wide (midline height / width minus parapophyses 133%) as in Falcarius (118%), but unlike the Angeac taxon (95%) and especially Chirostenotes (74%) and Masiakasaurus (64%).
The anterior peduncular fossae are well defined as in Chirostenotes and at least anterior Falcarius cervicals, but unlike Masiakasaurus or the Angeac taxon. They are however also well defined in anterior Masiakasaurus cervicals, so the condition in Falcarius isn't necessarily better than that genus or the Angeac taxon.
They are also placed far below the diagonal prezygapophyseal surface as in at least anterior Falcarius cervicals, but unlike the Angeac taxon, Chirostenotes or Masiakasaurus. The same could be said re: anterior Masiakasaurus cervicals.
The prespinal fossa is broad like Chirostenotes and Masiakasaurus but unlike the Angeac taxon.
It has anteroposteriorly broad exposure dorsally as in the Angeac taxon and at least anterior Falcarius cervicals, but not Masiakasaurus (including anterior cervicals of the latter).

Overall, Thecocoelurus is most similar to Falcarius and least similar to Masiakasaurus.  There are four good characters shared with Falcarius to the exclusion of the Angeac taxon, and three characters that are more similar to Chirostenotes than to the Angeac taxon, but two characters that are more similar to the Angeac taxon than to Chirostenotes.

Thecocoelurus holotype completed with the posterior of Falcarius (modified from Naish and Martill, 2002 and Zanno, 2010).  This results in a centrum length of 68 mm for Thecocoelurus, compared to Naish's (2011) estimate of 70-90 mm.

Falcarius does differ from Thecocoelurus in having a ventral median ridge on its centra, but this is an autapomorphy not seen in other therizinosaurs.  Besides this, no characters differ between the specimen except exact size and shape of pneumatic features, which themselves vary between right and left sides of Thecocoelurus.  Both are Barremian, and Thecocoelurus is 58% the size of the Falcarius individual that preserved the posterior cervical (though a growth series is known, where that individual falls is unreported).  Whether Thecocoelurus and Falcarius share derived characters to the exclusion of other therizinosaurs would require more study, but at the moment is seems most parsimonious to consider Thecocoelurus a basal therizinosaur and not closely related to the Angeac taxon.

References- Sues, 1997. On Chirostenotes, a Late Cretaceous oviraptorosaur (Dinosauria: Theropoda) from Western North America. Journal of Vertebrate Paleontology. 17(4), 698-716.

Naish, Hutt and Martill, 2001. Saurichian dinosaurs 2: Theropods. In Martill and Naish (Eds). Dinosaurs of the Isle of Wight. The Palaeontological Association. 242-309.

Naish and Martill, 2002. A reappraisal of Thecocoelurus daviesi (Dinosauria: Theropoda) from the Early Cretaceous of the Isle of Wight. Proceedings of the Geologists’ Association. 113, 23-30.

Kirkland, Zanno, DeBlieux, Smith and Sampson, 2004. A new, basal-most therizinosauroid (Theropoda: Maniraptora) from Utah demonstrates a Pan-Laurasian distribution for Early Cretaceous therizinosauroids. Journal of Vertebrate Paleontology. 24(3), 25-26.

Lu, 2004. Oviraptorid dinosaurs from Southern China. PhD Thesis. Southern Methodist University. 249 pp.

He, Wang and Zhou, 2008. A new genus and species of caudipterid dinosaur from the Lower Cretaceous Jiufotang Formation of Western Liaoning, China. Vertebrata PalAsiatica. 46(3), 178-189.

Zanno, 2010a. A taxonomic and phylogenetic re-evaluation of Therizinosauria (Dinosauria: Maniraptora). Journal of Systematic Palaeontology. 8(4), 503-543.

Zanno, 2010b. Osteology of Falcarius utahensis: Characterizing the anatomy of basal therizinosaurs. Zoological Journal of the Linnaean Society. 158, 196-230.

Carrano, Loewen and Sertich, 2011. New materials of Masiakasaurus knopfleri Sampson, Carrano, and Forster, 2001, and implications for the morphology of the Noasauridae (Theropoda: Ceratosauria). Smithsonian Contributions to Paleobiology. 95, 53 pp.

Naish, 2011. Theropod dinosaurs. In Batten (ed.). English Wealden Fossils. The Palaeontological Association. 526-559.

Allain, Vullo, Le Loeuff and Tournepiche, 2014. European ornithomimosaurs (Dinosauria, Theropoda): An undetected record. Geologica Acta. 12(2), in press.


  1. I've scored the Angeac taxon based on the vertebra, tibia, metatarsal and a few notes present in Allain et al. (2014), to test where it goes among theropods (just a very preliminary run, pending the full description). Details soon...

  2. I once wrote (and submitted) a manuscript where Thecocoelurus and Falcarius were directly compared (thanks to Jim Kirkland, I was actually able to put cervical vertebrae of the two side by side). While they were similar overall, there were also notable differences, and these led me to give up on the idea (then being seriously discussed by Kirkland and myself) that Thecocoelurus might be a therizinosaur. I need to dig out the manuscript.. I gave up on it after it was rejected. Currently aiming to revisit this whole mess on Tet Zoo anyway...

  3. Ok, I just dug out my old manuscript. It was rejected from a journal because they didn't want to run it, not because they found any problems with it, and I then gave up and moved on. I have lots of photos of Thecocoelurus and Falcarius side-by-side, with most of the stuff I regarded as significant pertaining to the anterior surface. Keeping in mind that this written in 2006 (and that the Falcarius vertebrae I was basing my comparison on don't look exactly like the one figured by Zanno), I dismissed a close affinity with Falcarius on the following basis...

    In Thecocoelurus, the interspinous ligament is about equal in width to the neural canal. This is the case in caenagnathids, but not in Falcarius where the ILF is about half neural canal width.

    In Thecocoelurus, pedicular/peduncular fossae flank the neural canal and are located ventral to the interspinous ligament fossa. Again, this is the case in caenagnathids but not in Falcarius where the pedicular fossae are entirely dorsal to the neural canal, and lateroventral to the interspinous ligament fossa.

    Thecocoelurus and caenagnathids also share a subrectangular anterior articulatory surface of the centrum, whereas the surface is distinctly rounded in Falcarius.

    In both Thecocoelurus and caenagnathids, ventrolateral ridges on the ventral surface of the centrum are subparallel for most of their length, but diverge laterally at the centrum’s anterior end. Ventrolateral ridges are present in F. utahensis, but in this taxon they are subparallel for their entire length and do not diverge anterolaterally to the same marked degree.

    There is now good reason to revamp and republished this manuscript. In time.

  4. Thanks for the detailed reply, Darren.

    Perhaps the major issue here is which vertebrae you were comparing. If Thecocoelurus is a posterior cervical, then we'd need to compare posterior cervicals of Falcarius too. Zanno, (2008, 2010) states "Only a single complete caudal cervical vertebra is known", which doesn't preserve the areas of your first two differences. But maybe you had less complete posterior cervicals that did preserve those areas. Or maybe you were comparing anterior cervicals.

    The interspinous ligament fossa is also narrow in the Angeac taxon, so if true in Falcarius as well, would be a way these differ from Thecocoelurus, caenagnathids and noasaurids.

    The pedicular fossae overlap both the neural canal and interspinous ligament fossae in the Angeac taxon, caenagnathids and noasaurids. They are ventral to the interspinous ligament fossa in Thecocoelurus, as you say. If they are dorsal to the neural canal in Falcarius, this would indeed be a difference between the taxa, but not one useful for phylogeny (both Thecocoelurus and Falcarius would be autapomorphic, albeit in different directions).

    The anterior articular surface shape could actually be known in the vertebra shown by Zanno, she just doesn't figure it. I'd say the Angeac taxon and Masiakasaurus are both rounded too. Unlike the previous two characters, this changes between vertebrae in Masiakasaurus, with the anterior cervicals more oval and the posterior ones having straighter edges.

    The angle and projection of anterolateral divergence differ so much between sides in Thecocoelurus that I question its systematic utility. This is especially true given the rugose parapophyseal surface of Falcarius that presumably contacted cartilage for the rib articulation. This is expected to be particularly variable and change with age/ossification.

    So while a good analysis would require more details on the specimens you used, your characters don't seem to affect how Thecocoelurus relates to the Angeac taxon at least. If your first three characters are true for Falcarius posterior cervicals, they would still be equal in number to my three shared with Falcarius but not Chirostenotes (anterior pleurocoels placed in obvious fossa; infradiapophyseal fossa developed as an elongate groove; centrum taller than wide). Then we have Andrea's observation the section of Thecocoelurus' vertebra looks camerate, which is unlike Chirostenotes, Falcarius or neoceratosaurs.

  5. Yeah... it's scary how many things there are to consider when dealing with issues of this sort - I didn't mean to imply that my comments provided a 'complete' analysis, only that the above case is the one I was making in 2006. On the one hand, I'm keen to see the argument about Thecocoelurus revised and tested in more complete and rigorous fashion... on the other I'm somewhat loathe to keep rehashing interpretations of the same fragmentary fossil. And no sign yet of any more complete material of whatever Thecocoelurus is. More comments coming at Tet Zoo.