Thursday, February 18, 2016

Sauropod Thoughts

So this month I've been concentrating on adding sauropods to the Database, going in reverse chronological order.  I find sauropods interesting because unlike theropods, their phylogenetics are almost completely unresolved.  I mean sure there's a basic outline and diplodocoids are pretty well researched, but there are huge swaths of cetiosaur-grade and titanosaur(iform/oid) grade taxa that can go almost anywhere between Vulcanodon and Neosauropoda on the one hand, and between Camarasaurus and Eutitanosauria on the other.  What follows are some thoughts on taxa or papers describing them...

Zby is based on mostly a forelimb that was originally assigned to Turiasaurus in a SVP abstract (Mateus, 2009), but later described as a new genus (Mateus et al., 2014).  But the seven characters listed by the latter authors as distinguishing the taxa are only determinable in their holotypes, not the referred specimens of Turiasaurus.  Given their stratigraphic and geographic proximity and the variation seen within theropod species, it seems plausible Mateus was right in 2009.

Holotype of Zby atlanticus (ML 368) (after Mateus et al., 2014).

I didn't realize how sad the descriptive situation is for Galeamopus. Sure the type skull was described almost a century ago, and the other skulls have featured in discussions of Diplodocus crania, but the postcrania of the type or the new Howe Quarry specimen (SMA 0011) have never been described.  We need not only a description of these, but also a paper describing all the supposed Diplodocus skulls, since according to Tschopp et al. (2015), no skulls can actually be referred to that genus.  Thus the Diplodocus OTUs in every other analysis are composites.

Despite having four generally stellar coauthors, the description of Abydosaurus (Chure et al., 2010) is... well... bad.  Even though it has seven supplementary documents, one of which is a spreadsheet of several hundred sauropodomorph tooth elongation indices, there's no measurement table for anything except two cervicals.  The most we get for the skull is "The four known skulls of Abydosaurus are nearly the same size, measuring approximately 0.5 m long and half as tall posteriorly."  You could have used less words and given us precise data for all four specimens.  And the table with the cervical measurements includes data for seven other sauropods, all but two of which are from previously published literature anyway.  Why spend space on data already out there instead of publishing new data?  The referred postcrania don't have repository numbers listed either.  Finally, the phylogenetic analysis goes out of its way to delete data from Wilson's original version, removing all but two non-neosauropods, combining genera to make four suprageneric OTUs, deleting characters that don't vary due to those removals, etc..  All this does is make it more tedious for future workers to add Abydosaurus to the base version of Wilson's matrix that tens of other papers use, and it's not like Wilson's 234 character, 27 taxon matrix is unwieldy for any 2010 computer to handle.  They just spent a lot of time and effort to make the analysis worse.

Proximal caudal vertebra of Normanniasaurus genceyi (MHNH-2013.2.1) (after Le Loeuff et al., 2013). Scale = 10 cm.

Normanniasaurus (Le Loeuff et al., 2013) is one of the most fragmentary taxa I've seen erected in the past few decades (the holotype consists of two presacral prezygapophyses, three fused sacral centra, an incomplete proximal caudal vertebra, an incomplete mid caudal vertebra, partial scapula, two ilial fragments, ischia [one partial, one fragmentary], a femoral fragment and a fibular fragment).  This is quite ironic considering Le Loeuff's famous 1993 paper declaring most European titanosaurs to be indeterminate (Macrurosaurus, Hypselosaurus, Aepisaurus, "Titanosaurus" lydekkeri).  Most of the supposedly diagnostic characters listed by the authors are either primitive for titanosauriforms (presacral vertebrae with hyposphene-hypantrum articulation; middle caudal vertebrae amphicoelous; ilium with anterolaterally expanded blade) or classic characters diagnosing larger clades within it (internal texture of presacral vertebrae camellate; proximal caudal vertebrae deeply procoelous; middle caudal vertebrae with anteriorly placed neural arch).  I'm very doubtful whether the remaining few caudal characters will prove diagnostic if anyone looks into them.  When added to Mannion et al.'s (2013) matrix, Normanniasaurus resolves as a eutitanosaur closer to Saltasaurus than opisthocoelicaudiines. 

Finally, Mannion and Otero (2012) did a superb job redescribing Argyrosaurus.  The taxon is only known from a forelimb, though apparently the entire skeleton was originally there but did not survive excavation.  This is my favorite kind of paper- redescribing a taxon known for decades in a modern context.  Mannion's one of my favorite sauropod authors, having redescribed Mongolosaurus (2010), "Brachiosaurus" atalaiensis (2013) and B? nougaredi (2013).  Mannion and Otero didn't include Argyrosaurus in a phylogenetic analysis, saying it has some plesiomorphic characters despite sharing many characters with derived titanosaurs.  I coded Argyrosaurus in Mannion et al.'s macronarian matrix, and I have to say the characters are well formed, so that coding is objective.  I'm not a sauropod expert, so I don't have the background comprehension of sauropod characters, but Mannion makes them easy to code.  Argyrosaurus ends up as a nemegtosaurid (sister to Mongolosaurus and Rapetosaurus), so despite Mannion et al.'s lack of many lithostrotians, the genus seems to be a member of the clade.  If it is a nemegtosaurid, Argyrosauridae Bonaparte 1987 has priority over Nemegtosauridae Upchurch 1995.  Interesting.  One critique of the paper is that it doesn't specify the horizon and locality of referred Argyrosaurus specimens (so I had to refer back to Powell 2003 and Bonaparte and Gasparini 1979), and proposes these are nomina dubia without detailed examination or comparison.

Will the next post get back to theropods? Depends on the papers published in the near future.  Come on, O'Connor, I dare ya. ;)

References- Mateus, 2009. The sauropod dinosaur Turiasaurus riodevensis in the Late Jurassic of Portugal. Journal of Vertebrate Paleontology. 29(3), 144A.

Chure, Britt, Whitlock and Wilson, 2010. First complete sauropod dinosaur skull from the Cretaceous of the Americas and the evolution of sauropod dentition. Naturwissenschaften. 97(4), 379-391.

Mannion and Otero, 2012. A reappraisal of the Late Cretaceous Argentinean sauropod dinosaur Argyrosaurus superbus, with a description of a new titanosaur genus. Journal of Vertebrate Paleontology. 32(3), 614-638.

Le Loeuff, Suteethorn and Buffetaut, 2013. A new sauropod dinosaur from the Albian of Le Havre (Normandy, France). Oryctos. 10, 23-30.

Mannion, Upchurch, Barnes and Mateus, 2013. Osteology of the Late Jurassic Portuguese sauropod dinosaur Lusotitan atalaiensis (Macronaria) and the evolutionary history of basal titanosauriforms. Zoological Journal of the Linnean Society. 168(1), 98-206.

Mateus, Mannion and Upchurch, 2014. Zby atlanticus, a new turiasaurian sauropod (Dinosauria, Eusauropoda) from the Late Jurassic of Portugal. Journal of Vertebrate Paleontology. 34(3), 618-634.

Tschopp, Mateus and Benson, 2015. A specimen-level phylogenetic analysis and taxonomic revision of Diplodocidae (Dinosauria, Sauropoda). PeerJ. 3:e857.


  1. According to the review history on PeerJ, a description of SMA 0011 was originally part of the paper that named Galeamopus, but was cut due to length. I can't speak for the authors, but I assume that that paper is forthcoming.

    Somphospondyl taxonomy is 'fun,' isn't it?

  2. Cut for length in a purely electronic journal?! Thanks for giving us a paper with less information. :| Honestly, some of the advice sounds more like advertising than science- "some of the more pertinent and important details of the research may be lost to the reader in a sea of text, tables, and figures. In my opinion, aspects of the paper (particularly apomorphy lists) could be moved to supplementary information or presented more succinctly than given here..." The very concept of supplementary information in the digital age is nonsensical outside of the tabloids, with the exceptions of NEXUS/TNT files, spreadsheets, 3D figures and other information that is inefficient for pdfs. Interested scientists will dig through huge papers to get the info they want. No one told Welles he should separate his 154 page description of Dilophosaurus from his 22 page comparison with other theropods, or Walker he should split his 53 page description of Ornithosuchus from his 24 page theropod discussion. Ah well, thanks to PeerJ's format, the original draft including the description of SMA 0011 is downloadable, so in a sense we have its description anyway (though lacking measurement tables). It's not like peer review would notably change it. Funny that Tschopp et al. were going to describe it as a new species of Galeamopus, whose name and diagnosis are redacted in the draft. I think they oversplit taxa as it is, but that's a topic for another day.

    As for somphospondyl taxonomy, the main issue is that so many clades are defined using Saltasaurus. A Sereno-ism just like how he plagued ornithischian definitions with Parasaurolophus. Also, because there are a billion Cretaceous basal somphospondyls and Late Cretaceous South American titanosaurs being described every year, I'm sure we have a ton of synonymies to look forward to once someone critically reviews them.

  3. Yeah. It's 300 pages long; nobody but the most determined are going to read the whole thing in the first place so it does seem like a strange omission.

    Looking at the description of Normanniasaurus myself, it seems to have a pretty weird combination of characters, such as hyposphene-hypantrum articulations and non-procoelous mid-caudals (unlike derived titanosaurs) and strongly inclined caudal centra (like aeolosaurs), so despite the poor quality of the type specimen, I am cautiously optimistic about its diagnosability.