FORGOTTEN GIANTS, #1: Puertasaurus

Posted by Nima On Monday, March 8, 2010 32 comments

Ahoy there, dinosaur fans! After much anticipation and work, I am proud to announce the first installment of my new Sauropod reconstruction series:

Forgotten Giants will focus exclusively on titanosaurs and their closest basal relatives. And the first "episode" in this series is the largest of them all - Puertasaurus reuili. It's also apparently the widest and most massive, but one of the least understood - so let's get started. Here's the (not so) skinny on the NEW "biggest dinosaur".

A giant among giants

Discovered only in 2005, Puertasaurus is not just huge. It's not just gigantic. It's literally a record-smasher. Whereas other titanosaurs may have approached or reached the 100-foot (30m) mark, Puertasaurus, even by conservative estimates, easily exceeded it by at least a few meters. Though it's only known from four vertebrae, their huge size points to a creature large enough to possibly rival even the enigmatic Amphicoelias fragillimus.


Diagram of giant sauropod dorsal vertebrae in frontal view (taken from the Hairy Museum of Natural History)
 (a) The diplodocid Seismosaurus hallorum, dorsal 8 (after Herne and Lucas 2006)
(b) The diplodocid Amphicoelias fragillimus, dorsal 9/10? as reconstructed by Cope 1878,
(c) Amphicoelias fragillimus dorsal 9/10? as restored by Carpenter 2006, after A. altus 
(d) Amphicoelias altus dorsal 10? (Carpenter 2006, modified from Osborn & Mook 1921),
(e) Puertasaurus reuili dorsal 2 after Novas et al 2005, Figure 2). Scale bar equals 1 meter.

The dorsal from Puertasaurus is far wider than even the reconstructed dorsal of A. fragillimus - but far less tall. This was a squat, wide titanosaur, whereas A. fragillimus, despite its size, was a proportionally slim, deep-bodied diplodocid. And for human scale.....

Here's Dr. Fernando Novas, who described and named the new sauropod Puertasaurus reuili, after the two farmers, Puerta and Reuil, who first discovered the remains. This is just a taste of how big this animal was - the bone makes entry impossible through an otherwise decent-sized university office - and from what I've heard, Novas is a tall man, around 6 feet or so. As you can see, the dorsal vertebra had to be reassembled from its fragments, but most of it is there. Note the picture on the computer monitor - more on that later.

Here's the same dorsal (D2) from the front, viewed in an exhibit. The other remains known from Puertasaurus are a cervical vertebra (probably the ninth) and two caudals that have not been published (the reason for this is unknown, and it's a shame they weren't published in Novas' paper, as they would have yielded valuable information about how the tail should be restored).

A little note on this dinosaur's family tree - even the few bones found so far indicate that it's a member of the family Lognkosauria, an intermediate group of titanosaurs that evolved in the mid-cretaceous, and are more derived than the Andesauridae (such as Argentinosaurus) but more basal than the Saltasaurs and Antarctosaurs that replaced them toward the end of the Cretaceous. Lognkosaurs are not very well known (in fact, there are only three other confirmed genera in the family) but they were present in both South America and Africa, and all share a very robust build, huge cervical rib loops, massive neural spines, and very long necks that are either unusually wide, deep, or both. These are the Gothic cathedrals of the Titanosauria. What's interesting about Puertasaurus in particular, is that it appeared much later than the other Lognkosaurs. Most of them date from the mid-Cretaceous, in the Albian, Cenomanian, and Turonian epochs (roughly from 110 to 90 million years ago), but Puertasaurus stomped onto the scene very late, in the Maastrichtian epoch - the very end of the Late Cretaceous, 70-65 million years ago, a time when T. rex was already roaming Montana thousands of miles to the north, a time when practically every sauropod was a titanosaur, and only one (Alamosaurus) made its home in North America, a land that was then mostly swamps and inland seas - not too sauropod-friendly - but was gradually becoming drier and more hospitable to the giants. Alamosaurus itself was a saltasaur with South American ancestors, and probably migrated north across the land bridge that was rapidly forming between the two continents in the Late Cretaceous.

Now given that Puertasaurus is such a new dinosaur, it's understandable there's never been a truly high-fidelity illustration of it. In fact there have only been two complete life illustrations done in non-digital media, both by Gabriel Lio. Though why this animal is not as popular with both artists and the public as Argentinosaurus, despite almost certainly being larger, is a bit of a puzzle.

The first Gabriel Lio (?) painting of Puertasaurus shows a rather small-looking and short-necked sauropod dwarfed by what appears to be an oak tree on the left. Not only is the presence of a huge oak tree dubious for Cretaceous Argentina (where conifers, like those in the background, dominated the flora), but it also greatly under-emphasizes the size of this animal. It's not too far from those little running ornithopods, yet it looks strangely small relative to their scale. Even a modest-sized Apatosaurus should be bigger than this relative to those little buggers. What's more, the primitive "generic long-neck" head and the not-so-long neck combined with a fairly nondescript body and tail, make this guy look more like an a boring 50-foot cetiosaur than a record-breaking titanosaur that challenges A. fragillimus for sheer size.

Here's the second painting by Gabriel Lio, incorporated into a magazine diagram on the new discovery. It's the same image on Fernando Novas' computer screen in the earlier picture, and was published on many websites that announced the new discovery - ironic, since this illustration isn't all that scientific. Here we get a highly exaggerated, Marvel comic-like picture, with again an unusually short neck, Donald Duck-like jaws, a torso so short that the hands touch the feet in backstride, an inward-twisted left hand whose angle looks like a nightmare wrist sprain in the making, oversized hands and feet, long and dislocated toes, and bulging arms and legs built more for tackle football than for simple graviportal walking. Believe it or not, it actually does not take huge bulging muscles to move a sauropod-sized body at a slow pace. Ceratopsians on the other hand, had enormous muscle crests and scars on their short limbs, as do galloping rhinos today - moving fast requires proportionally more muscle than moving mass. Add to this the furious snarling facial expression (did sauropod even HAVE facial expressions? Even most herbivorous mammals today don't, despite having muscular faces), and you basically have a 100-ton monster with earthquake-inducing roid rage. Of course it could just as easily be a 6-ton wimp. It's got the proportions (however pumped up) of a short-necked dwarf sauropod.

In an attempt to draw a more believable Puertasaurus (and one that actually LOOKS like it has the proportions of a hundred-foot-plus sauropod), as well as one that had the recognizable features indicated by its bones, I started with the actual published material. As I think visually, I could see no essential connection between the actual bones and what Gabriel Lio has drawn (not that I doubt his actual talent as an artist in any way - but his art is heavily exaggerated and comic-ized, and much of it just isn't too terribly scientific). Thus I went purely by the bones and by those of Puertasaurus's better-known relatives. Extrapolating based on the length of the published ninth cervical vertebra and a general rule of 15 cervicals, most of roughly similar length, for titanosaurs, I ended up with a far longer neck than Gabriel Lio illustrated. One wonders if he actually went off of Novas' published figures and measurements, or just drew his fanciful sauropod 100% from imagination and called it Puertasaurus.

In making a truly accurate Puertasaurus restoration (as much as is possible given that there are only four bones found, and only two of them have been published), one of my aims was to include scale drawings of the described bones, dorsal D2 and cervical C9, as they would have appeared when complete and uncrushed. Novas et. al. included a smooth pencil drawing of the 'reconstructed' D2 but did not put in one for C9 - the only visual diagrams of C9 are a few line drawings and a set of four grainy photos of the incomplete vertebra, only partially reconstructed, with the rear of the neural arch and the proximal arches of the cervical ribs incomplete. D2 was easy to draw, but C9 posed a real challenge.The long portion of the cervical ribs was also absent, so I drew them as they may have looked based on the proportions of cervical ribs on other titanosaurs and titanosauriformes (primarily Rapetosaurus and Uberabatitan, but looking at a a few Giraffatitan pics didn't hurt).

This ended up being a triple view restoration, much like Gregory Paul does with skeletal diagrams. I mainly did multiple views of the body to give some idea of its likely 3D form. Though to my knowledge, no artist before has regularly done life restorations in multiple view. This may be the first example.

Late in the process of drawing it, I got the idea to include the largest known predatory dinosaur, Giganotosaurus carolinii (incidentally also from Argentina, but from an earlier time) for scale. Now there's a real idea of the vast scale of Puertasaurus, a sauropod so colossal that even the largest meat-eaters barely came up to its knee. And yes, T. rex was even smaller.

This first restoration was somewhat messy and smudged, but already it's possible to get a grasp of what this animal may well have looked like in life. The torso is incredibly wide, and I drew it that way for a reason - the dorsal vertebra D2 is freakishly wide - much wider than it is tall. Novas published a width measurement of 1.68m for this bone, and accounting for crushing, erosion, and edgewise damage, I arrived at a maximum restored width of just under 2m. In particular, the wing-like diapophyses are enlarged far beyond those of any other sauropod, even other wide-bodied titanosaurs. The expanded processes must have supported a very wide and massive rib cage, and then when you consider that D2 is only the second dorsal, and that the dorsals further back have much wider ribs in all sauropods, then it's not too hard to see that the maximum width was record-breaking. With multiple scaling attempts, I arrived at a fairly radical 7.5m width at the rib cage's widest point (which would probably be around dorsal 8 or 9).

Also note the extreme width of C9, especially the cervical ribs. The neck was proportionately squat rather than deep, and this probably culminated in a wide, flattened head (as well as potentially resulting in a very extreme vertical range of motion for the neck). In terms of proportions, it's already clear even based on these two published bones, that this is not just one of the biggest sauropods, but also one of the weirdest.

Next there was some pretty standard digital refining, cleanup, and color adjustment.

Then I labeled everything, including the described bones. And I also added the human figure for scale.

However this image had a few problems - the most obviously glaring one is the top view of the cervical C9. There neural spine is one thin ridge, and the neural arch below and around it looks flat. And suddenly I realized this was not accurate. I went back and consulted the original photos of this bone in Novas et. al. 2005.

Images of Puertasaurus cervical C9 from Novas et. al. 2005:
A: frontal view
B: left side view
C: top view
D: ventral (bottom) view

Notice that the neural spine as seen in (A) is quite wide and robust. It's either partially wrapped in plaster or heavily reconstructed, but the visible dark portion on the right of the neural spine in (A) indicates the extent of the spine's true width. From the top, in (C) the actual contours of the spine are very hard to discern due to all the white plaster. However, from the bony margins and shadows of the spine, it's very evident that this is a much wider structure than I had drawn. So I altered it and added shading to the image of C9, since a simple outline drawing would not show the gradual anterior slope of the spine or the pit in front of it. The result is a much more robust and complex design with both curves and angular features.

This was the version that made it to the ArtEvolved Sauropod gallery, and later was featured on SV-POW (kudos to the SV-POWsketeers for hosting the detailed and enlightened discussion that followed).

Yet I decided this restoration was still missing an important element - a frontal view of C9. I went back to the photos from paper for reference, and fixed the symmetry to remove the crushing (which is minimal but asymmetric). Then I reconstructed the missing portions (mainly the neural arch and the upper portions of the proximal arches of the cervical ribs, these are drawn less massive than in the line drawings in Novas I also drew a rough outline of the neck circumference, lightly muscled, as in real life the vertebra was internally pneumatic. I also added complete proximal arches in the other views on C9. This frontal view was drawn and scanned separately and re-sized to fit the scale of the Puertasaurus.

One more main change remained: the upper jaw line looked a bit too straight to belong to a late-cretaceous titanosaur. I did not want to make it as scooped out as that of Antarctosaurus or the saltasaurids, but more so that its previous primitive-looking shallow curve. The improvement is consistent with the general assumption that the heads of Lognkosaurians would have had a shape somewhere between basal titanosaurs/titanosauriforms and saltasaurs. Also I changed the circumference outline on the frontal view of C9 to show more room for the neck ligament on the neural spine. The final result is below.

The final product: 130 feet of pure sauropod awesomeness. The head alone would have been big enough to swallow something the size of a very large man. And the claws on its feet could take out even a Giganotosaurus-sized predator in one blow. And if the rib cage had been preserved, you could probably shove a couple of Hummers in there without too much trouble. It's not a beast most dinosaurs would ever want to mess with (or could, for that matter, even if they wanted to!)

* Epilogue: The paleo-artist known as "Rexisto" used my Puertasaurus as the basis for many of his revised mega-titanosaur silhouettes, mistakenly claiming that my restoration was world-renowned paleoartist Greg Paul's work! While I took this as a HUGE compliment, unfortunately it means that Rexisto didn't ask permission to use my work, but after I pointed this out to him, he apologized and I accepted his apology since it was just a misunderstanding due to language barriers, and because his silhouette work is non-commercial in nature. But in the future, people, please ASK me if you want to base your work on mine, in most cases I'll be totally cool with it as long as you ask me first - it's the same thing I do when I want to re-interpret another artist's work (as I did in the case of "remixing" John Conway's Chasmosaurus - see my finished version HERE), and I'd appreciate it if everyone else does the same. And if you just want to re-post my images on your blog, that's fine, but please be professional and give me proper credit. If your English is not so good, email me anyway, I'll deal with translating and replying to it. It's not that hard. Read my TERMS OF USE on the right sidebar, or even just the little blurb under the blog title. You have been warned.

UPDATE: Rexisto has made a correction on the Mesozoico forums to credit me as the creator of the work he based his silhouettes on. Here's the link. Thanx, Rexisto!

Peace out :) And by all means, share your thoughts on the Puertasaurus!