Wednesday, September 14, 2016

Titanosaur Heads

Illustration by "Stocktrek Images, Inc."
Sauropods are among the most familiar of all dinosaurs—consider that pretty much everybody knows Brontosaurus or Brachiosaurus. But what may surprise you is that most sauropods are known from fairly scant remains—isolated vertebrae or pieces of the axial skeleton. Brontomerus was named for, essentially, an incomplete ilium and scapula. Argentinosaurus, one of the largest tetrapods that ever lived, was named for a series of vertebrae, ribs, and a fibula. But you know what’s most often missing from a sauropod skeleton? The head.

The articulation between the head and the neck in sauropods is surprisingly weak, meaning that the skull often pops off and floats away prior to fossilization. While there are certainly famous examples of sauropod skulls—Diplodocus, Brachiosaurus, and Nigersaurus all preserve wonderful skulls—“cranial materials are only known for less than one third of sauropod genera and even fewer are known from complete skulls” (Chure et al. 2010). This problem is especially prevalent in titanosaurs, the “last” group of sauropods that exploded in diversity during the Cretaceous period and pretty much took over the South America. It seems like every other week a new gigantic titanosaur is described. According to Rogers & Forster (2004), “Cranial material is known from only ten titanosaurs and is limited to isolated elements and fragmentary braincases.” Obviously, things have changed since 2004, and the point of this post is to highlight the few complete skulls described for titanosaurs.

From Rogers & Forster (2004)
Rogers & Forster (2004) were the first to describe a complete (though composite) titanosaur skull--that of Rapetosaurus krausei from Madagascar. Its head is superficially similar to diplodocids, and in fact its dentition mimics that of diplodocids as well. There are obvious differences, however: the extremely large antrobital fenestrae and squared-off surangular stick out. However, it is interesting to me that the skull is actually quite similar to diplodicids--this is a group of sauropods that evolved from Camarasaurus-like macronarians. The diplodicid style was apparently quite popular.

And it was taken to an extreme in rebbachisaurids like Nigersaurus.

From Wilson (2005)
The very next year saw Wilson’s 2005 description of the skull of Nemegtosaurus mongoliensis, a Mongolian (sigh) titanosaur, based on a single well-preserved skull and jawbone. It is quite different from Rapetosaurus. The antorbital fenestrae are small and triangular, the whole skull is more robust, and the premaxillae and nasals arch, giving the naris a roughly brachiosaur-esque look. 

Like Rapetosaurus, though, the teeth are peg-like and restricted to the front of the mouth. Despite their differences, Wilson united Nemegtosaurus and Rapetosaurus in a monophyletic Nemegtosauridae for reasons that will become clearer later in this post.

It would be six years before another titanosaur skull was described, and that would be by Zaher et al. (2011) in their description of Tapuiasaurus macedoi, an Aptian titanosaur from Brazil. Although squashed laterally, the skull is incredibly well-preserved and even includes the hyoid bones. The authors note that Tapuiasaurus is very similar to those other two titanosaur skulls and so places it in the Nemegtosauridae—meaning that the group now spans three continents and at least 30 million years. The skull would later be more fully described in 2016 by Wilson et al. This paper includes a beautiful skull reconstruction:

From Wilson et al. (2016)

It’s actually quite a bit different from Nemegtosaurus or Rapetosaurus. While there are similarities, the differences really stand out (especially the ridiculously-enlarged naris). With a new understanding of the skull in hand, the authors re-evaluated the phylogenetic position of Tapuiasaurus and found that it, along with Nemegtosaurus, occupy unstable positions due to how much material (especially cranial) is unknown for other titanosaurs! The authors eventually concluded that Tapuiasaurus occupies a very basal position among titanosaurs, not highly derived, and that this “classic” titanosaur skull morphology must have appeared very early in the group’s evolution.

2016 also saw the publication of the skull of Sarmientosaurus musacchioi by Martinez, et al. This beautiful skull was found in Patagonia. It is not disarticulated (as in Rapetosaurus), deformed (as in Nemegtosaurus) or crushed (as in Tapuiasaurus). Rather, it largely intact, though a high-level reconstruction would've been appreciated. Interestingly, the skull differs in a number of ways from the previously-described skulls. The dentition seems to be an intermediate morphology between that of more basal macronarians and more advanced titanosaurs. The authors compare the teeth to basal titanosauriform Abydosaurus. Sarmientosaurus’ enormous orbit is also comparable to that of Abydosaurus.

From Martinez et al. (2016)

The authors ran several phylogenetic analyses and, in all cases, Sarmientosaurus falls out as a basal titanosaur, more derived than Euhelopus and Malawisaurus but basal to Tapuiasaurus (regardless of where that taxon ends up), Nemegtosaurus, and Rapetosaurus. The authors write that “Sarmientosaurus retains a relatively plesiomorphic skull and wide-crowned dentition reminiscent of those of non-titanosaurian titanosauriforms such as Abydosaurus and Giraffatitian,” yet it lived alongside more derived titanosaurs with more traditional titanosaur (diplodocid-like) dentition.

From Chure et al. (2010)

The authors also note that none of their analyses recovered Rapetosaurus or Tapuiasaurus as members of Nemegtosauridae, which now only includes Nemegtosaurus and Quaesitosaurus. They suspect that previous analyses found that linkage only because Rapetosaurus, Tapuiasaurus, and Nemegtosaurus had skulls that could be scored. By adding Sarmientosaurus, many of those characters could be redistributed. If you only had skulls from Triceratops and Centrosaurus, and could only code for skull characters from those two genera, they'd of course always wind up together even though they're actually rather far apart.

From Martinez et al. (2016). Wish they'd called up Wilson et al. (2016) for Tapuiasaurus.
So what we see is that titanosaurs very quickly evolved toward the diplodocid condition early in their diversification. If there's something particularly adaptable to that morphology, it's not terribly surprising to see it cropping up in another group of sauropods. While diplodocines were restricted largely to North America, there were South American (Leinkupal) and African ("Tornieria") representation, and then you've got the mostly Gondwannan rebbachisaurids. Titanosaurs, for the most part, replaced them all, turning up in South America, Africa, Asia, probably Europe, Australia, and Texas during the Cretaceous. This was a hugely successful group of sauropods and I should probably spend more time getting to know them.


  1. Er... Nemegtosaurus' skull was described in 1971 by Nowinski, Wilson merely redescribed it in 2005. Note Zaher et al. (2011) is only six years after the latter, not sixteen. Similarly, Rapetosaurus' was initially described by Curry Rogers and Forster in 2001, before the in depth 2004 osteology. While not complete, I think Antarctosaurus' skull is important, as it sets up the whole pre-2000s titanosaur skull narrative of them being diplodocoids due to their peg-like teeth (Huene, 1929). Then the story of Nemegtosaurus and Quaesitosaurus being viewed as diplodocoids, even as late as Upchurch's (1999) cladistic analysis of the question (they weren't thought to be titanosaurs until Salgado and Calvo, 1997). Finally, Malawisaurus throws a wrench into the simple diplodocation narrative with its boxy skull (Gomani, 2005). Oh, and as I wrote on SV-POW, the Sarmientosaurus analysis is terrible- .

  2. Titanosaurs really were some of the weirdest and most fascinating dinosaurs, and it seems like every time we find a new skull (which, granted, is not all that often) it has some unexpected, weird stuff going on. I look forward to the day when we finally begin to get a good handle on the diversity and relationships of the clade, so we can appreciate them for a bit more than the frustratingly incomplete material we have right now.

    Antarctosaurus is really quite interesting--it has a superficially rebbachisaurid-like snout, to the point where it was doubted whether it was from a titanosaur at all (it is, though: Bonitasaura deserves mention as well, for its role in correctly returning Antarctosaurus to the titanosaurs.

    Actually, Mickey, Calvo proposed that Nemegtosaurus and Quaesitosaurus were titanosaurs back in 1994, and Upchurch didn't back down on diplodocoid nemegtosaurs until after 2004 (despite Rapetosaurus, oddly enough).

    Euhelopus and Malawisaurus, with their boxy camarasaur-style skulls, definitely raise some questions about the evolution of the titanosauriform skull, but I think we still need a much better understanding of titanosauriform phylogeny to know whether Malawisaurus' boxy snout is plesiomorphic or apomorphic within the clade, and whether Euhelopus is a titanosauriform at all (Carballido and Sander's matrix says it isn't).

    Maybe those questions would be more clearly answered by now if there were fewer analyses like those in the Sarmientosaurus paper, though >:c

    Calvo JO. 1994. Jaw mechanics in sauropod dinosaurs. Gaia Ecological Perspectives For Science And Society 10:183–193.

  3. Thanks for the comments, guys.

    Mickey, yeah, I know Rapetosaurus and Nemegtosaurus were described earlier but I just wanted to focus on the heads. Similar answer to John: those skulls are great, yeah, but they're not complete.

    This post came together kind of haphazardly--I wanted to get something up, so I didn't do my usual research. My goal really was to show off complete titanosaur skulls and how different they were. An older version of this post incorporated Abydosaurus more as well as the recent Euhelopus skull description in PLOS One. I may tinker with this post over the next week or so and I'll be sure to incorporate your comments. I've got three or four posts "in the cooker" and I felt like I had to get something up in the meantime.

  4. I can't edit my own comment?! Mickey, one other thing. The original version of this post had "1995" by accident on Wilson's redescription, thus the following error with sixteen years instead of six. Like I said, in a rush.

  5. John- Ah yes, I forgot about Bonitasaura. Whose description also reminds me of the Auca Mahuevo embryos with their basically complete skulls (Chiappe et al., 2000 and 2001; detailed description in Garcia et al., 2010).

    Interesting re: Calvo's 1994 paper. I don't have it, but now I notice I have his 1994 thesis which has a chapter on that, no doubt what the Gaia paper was based on. In all fairness to Upchurch, his 2004 Dinosauria 2 analysis didn't include Rapetosaurus in the analysis, I assume due to a long period between writing and publication for many of those chapters. And the text does indicate some uncertainty here.

    I didn't realize Euhelopus was outside Titanosauriformes in some recent analyses. Would also be interesting to see how Campylodoniscus compares to other titanosaurs now that we have more skulls. Could it be similar to Sarmientosaurus, with their broad teeth?

    Zach- Well, Nowinski did focus on the head, since that's all that's known of Nemegtosaurus. ;) 40 pages of cranial description... I can send you the paper if you need it...