Scaling Up

Let's turn our attention from hadrosaur skin to ankylosaur skin, a topic which has received surprisingly less attention in the published literature than I would have thought. I should qualify that statement, however, by saying that by 'ankylosaur skin' I mean ankylosaur skin impressions, because ankylosaur dermal elements are well known and the focus of many a paper – I refer of course to osteoderms, which form within the dermis of the skin and which give ankylosaurs their spiky and armoured appearance.

For a couple of years now I've been keeping notes about occurrences of skin impressions in ankylosaurs, which eventually lead to a paper by myself, Mike Burns, Phil Bell, and Phil Currie. We reviewed the morphology of scale patterns in the few specimens that preserve skin, and found that there were some intriguing differences in scalation between different ankylosaurs.

The holotype of Scolosaurus cutleri, NHMUK R5161, has the best preserved integument for any North American ankylosaur, and has loads  of scale impressions lying overtop of the in situ osteoderms. In Scolosaurus, the scales form rosettes around the osteoderms. The largest scales are generally found closest to the osteoderms, but some large scales are scattered in between the osteoderms as well. Underneath the scales, small ossicles (little osteoderms less than 1 cm in diameter, but usually only 2-4 mm wide) fill the spaces between the larger osteoderms.

Scolosaurus is hard to photograph well, sorry!

In contrast, a very unusual specimen (ROM 813) has a completely different morphology. This specimen includes unusual long, rectangular osteoderms that aren't present in NHMUK R5161. The scales are on average much smaller, don't form much of a rosette pattern around any of the osteoderms, and are more uniform in size overall. ROM 813 is a little bit difficult to interpret because it is partially disarticulated (which is also intriguing given that such large portions of the integument are intact), but our best guess for the preserved portions is shown here.

Another super cool thing about ROM 813 is that it preserves the epidermal covering of an osteoderm, and it is the only example of this in an ankylosaur that I know about. In the photo below, the smooth side of the osteoderm is the epidermal scale, and the rough side of the osteoderm is the true bony part of the osteoderm.

Moving over to Mongolia, a specimen referred to Tarchiagigantea lacks the small pavement of ossicles seen in the Albertan ankylosaurs, and the epidermal scales are huge and more rectangular. In the portion of the integument preserved, osteoderms are separated by only one row of scales.

There's enough overlapping material between these specimens to allow us to compare scale patterns among different ankylosaurs, and the differences support the hypothesis that these are different taxa. Unfortunately, right now we can't assign ROM 813 to any known ankylosaurid taxon from Alberta – this could represent the postcrania of Euoplocephalus tutus, or Dyoplosaurus acutosquameus, or (less likely) a new taxon of ankylosaurid from the Dinosaur Park Formation. I think it's safe to say that the differences between Scolosaurus and ROM 813 represent true taxonomic differences, a finding that is in line with previous work by Phil Bell on scalation differences between Saurolophus angustirostris and Saurolophus osborni.

Illustrations by Lida Xing and via PLOS ONE.

One more comment about ankylosaur skin: In 2010 I had the opportunity to study the holotype of Liaoningosaurus paradoxus, and very interesting little ankylosaur from the Liaoning Formation of China. The original authors described Liaoningosaurus as possessing a ventral plastron (bony shield, like that found in turtles), which would have been a highly unusual anatomical feature given that no other ankylosaurs possess a plastron. Having looked at this specimen, I think a better interpretation for the plastron is that this is a segment of skin impressions from the belly region – there didn't seem to be any bony texture around the edges of this area, and the pattern is more consistent with scales than any osteoderms in other ankylosaurs.

Belly scales for Liaoningosaurus. The scale bar is in millimetres.

Papers!

Arbour VM, Burns ME, Bell PR, Currie PJ. 2014. Epidermal and dermal integumentary structures of ankylosaurian dinosaurs. Journal of Morphology 275:39-50.

Arbour VM, Lech-Hernes NL, Guldberg TE, Hurum JH, Currie PJ. 2013. An ankylosaurid dinosaur from Mongolia with in situ armour and keratinous scale impressions. Acta Palaeontologica Polonica 58:55-64. Many thanks to Dr. Hurum for inviting me to help describe this specimen!

Bell PR. 2012. Standardized terminology and potential taxonomic utility for hadrosaurid skin impressions: a case study for Saurolophus from Canada and Mongolia. PLOS ONE 7:e31295.

Xu X, Wang X-L, You H-L. 2001. A juvenile ankylosaur from China. Naturwissenschaften 88:297-300.

Who-oplocephalus: The Fellowship of the Half Ring

Behold, NHMUK R5161: the extraordinary holotype of Scolosaurus cutleri. This is truly one of the most amazing dinosaur fossils that has been collected from Alberta, and is one of the best preserved ankylosaurs in the world. And the best part is that it is on display for everyone to enjoy in the galleries of the Natural History Museum in London.

(Many thanks to Angelica Torices for snapping this photo for me! NHMUK R5161 is beautiful to look at, but difficult to photograph well, and I'm afraid most of my photos from my visit in 2009, while useful to me, are not necessarily that nice to look at.)

Most of the time, when you're looking at a drawing or model of Euoplocephalus, what you're really mostly looking at is NHMUK R5161. This specimen preserves a large amount of intact skin, which means that the free-floating osteoderms are, for the most part, in their correct life positions. Osteoderms form in the dermis of the skin, and so osteoderms are usually found as isolated elements, or scattered around disarticulated or associated skeletons. Because NHMUK R5161 is so complete, it has formed the basis for most restorations of Euoplocephalus.

Yup, all of these are kind of Scolosaurus. Especially the pink one. Perhaps somewhat embarassingly, this does not represent my complete collection of ankylosaur 'scientific models'.

However, you may have noticed something important: Scolosaurus lacks a skull and tail club. So, how could I figure out if Scolosaurus was the same as Euoplocephalus, if I couldn't look at the patterns of cranial ornamentation? If you remember back to part 2 of this series, I said that the morphology of the first cervical half ring was useful for diagnosing Euoplocephalus. 

The first cervical half ring of NHMUK R5161 has some important differences when compared to CMN 0210, UALVP 31, or AMNH 5406. In NHMUK R5161, the two medial osteoderms (closest to the midline) are round and lack keels, instead having a centrally located bump. Even in specimens of Euoplocephalus that have relatively flat medial osteoderms on the cervical half rings, like AMNH 5404, the medial osteoderms always have a keel and are more oval than circular. This suggested that Scolosaurus was distinct from Euoplocephalus and Anodontosaurus, even though it didn't have a head or tail club. (I largely agree with Penkalski and Blows' (2013) assessment that Scolosaurus is a valid taxon, but for somewhat different reasons than what they present in their paper.)

This reconstruction of Scolosaurus was drawn by Alice Woodward based on NHMUK R5161, and I suspect that many subsequent images of Scolosaurus were inspired by Woodward's art.

(A quick note about tail clubs and Scolosaurus – many artists have reconstructed this animal with a short tail and spiked tail club. The tail of NHMUK R5161 is broken at about the midpoint of the tail, probably just in front of where the tail club would have started. The skin is kind of sloughed out in this area, which has led many to erroneously interpret Nopcsa's figures and drawings as showing the knob of bone at the end of the tail. The 'spikes' on the tail club are really just osteoderms present at about the midpoint of the tail – no ankylosaurs had spikes on their tail clubs.)

The collection of NHMUK R5161 is one of the more interesting stories of palaeontological collecting in Alberta. The specimen was discovered and primarily excavated by William Edward Cutler in 1914, working for the Calgary Syndicate for Prehistoric Research (an excellent name for an organization that is, sadly, defunct). During excavation, the ankylosaur block collapsed on Cutler, no doubt causing grievous injuries. One of the Sternbergs finished the excavation, and the specimen was shipped to London.

NHMUK R5161 is significant, beyond being so complete, because it may derive from the Oldman Formation of Dinosaur Provincial Park, rather than the Dinosaur Park Formation. The quarry is somewhere across the river from Happy Jack's, the University of Alberta's field camp since 2008. However, the precise locality of the quarry is somewhat up for debate, and Oldman Formation sediments crop out in this area. Next week I'll be heading to Dinosaur Provincial Park, and one of the goals is to check out some of the potential quarry locations with our crew from the UofA, as well as Darren Tanke from the Tyrrell Museum, who has been investigating the NHMUK R5161 quarry for some time. It's important to know whether or not Scolosaurus comes from the Oldman or Dinosaur Park formation, so we can know whether or not Scolosaurus lived at the same time as Euoplocephalus. We will also be on the lookout for any leftover material…like the skull and tail club.

Two Medicine Formation skulls, from Arbour and Currie (2013). Notice the long, curved squamosal horns.

For a long time I was bitterly disappointed that there was no skull known for Scolosaurus. Could some of the isolated skulls I was referring to Euoplocephalus instead belong to Scolosaurus? Or was the skull of Scolosaurus noticeably different from Euoplocephalus? I despaired that I wouldn't have an answer to that question unless we miraculously collected another Scolosaurus from Dinosaur Park during my thesis. In the meantime, I was trying to figure out the identities of "Euoplocephalus" specimens from the Two Medicine Formation of Montana, which seemed very different from the Albertan specimens. These skulls had much longer and pointier squamosal horns with a 'backswept' appearance compared to Euoplocephalus and Anodontosaurus. I thought perhaps that the Two Medicine ankylosaurids might represent a new species of ankylosaurid (as did others – Penkalski named Oohkotokia earlier this year based on MOR 433, which has this unusual squamosal horn morphology).

From left to right, the first cervical half rings of AMNH 5337, TMP 2001.42.9, USNM 7943, and NHMUK R5161. The top row shows the half rings in anterior or posterior view, and the bottom row shows the rings in dorsal view. NHMUK R5161 is in dorsal view. Modified from Arbour and Currie (2013).

One day I was flipping through my photos of the Two Medicine Formation specimens when I realized that the morphology of the first cervical half ring of TMP 2001.42.9, the only specimen that had a half ring and skull, matched that of NHMUK R5161. Both had flat medial osteoderms with a central bump, rather than a keel. And indeed, another isolated half ring from the Two Medicine formation had this morphology as well. Scolosaurus was present in Montana!

I'll just take a moment here to address Oohkotokia: while I agree with Penkalski (2013) that the Two Medicine Formation ankylosaurid differs from Euoplocephalus, I do not think it differs sufficiently from Scolosaurus for these to be considered separate species. Perhaps one day in the future we'll find another Scolosaurus from Dinosaur Park that has flat medial osteoderms on the cervical half ring, but a skull with different squamosal horns compared to the Two Medicine skulls. If that happens, then I think you could make the case that Oohkotokia is valid. Until then, Oohkotokia is a junior synonym of Scolosaurus.

With the referral of the Two Medicine ankylosaur material to Scolosaurus, Scolosaurus now had a head! TMP 2001.42.9 even has a tail club, so we know that Scolosaurus had a round tail club knob. With the in situ osteoderms and skin impressions of NHMUK R5161, Scolosaurus is now one of the best understood ankylosaurids in the world. Now, if we could just figure out for certain exactly where it was collected from…

Next time: wrapping up loose ends, and figuring out what it all means.




PAPERS!


Arbour VM, Currie PJ. 2013. Euoplocephalus tutus and the diversity of ankylosaurid dinosaurs from the Late Cretaceous of Alberta, Canada, and Montana, USA. PLOS ONE 8:e62421.

Nopcsa BF. 1928. Palaeontological notes on reptiles. V. On the skull of the Upper Cretaceous dinosaur Euoplocephalus. Geologica Hungarica, Series Palaeontologica 1:1-84.

Penkalski P, Blows WT. 2013. Scolosaurus cutleri (Ornithischia: Ankylosauria) from the Upper Cretaceous Dinosaur Park Formation of Alberta, Canada. Canadian Journal of Earth Sciences 50:171-182.

Penkalski P. 2013. A new ankylosaurid from the Late Cretaceous Two Medicine Formation of Montana, USA. Acta Palaeontologica Polonica, in press.

I like ankylosaur butts, and I cannot lie.

Yes, I seem to have this recurring fascination for the derrière of these dinosaurs. In 2009 I published a paper (with my fellow grad students Mike Burns and Robin Sissons) on Dyoplosaurus, in which we argued that the pelvis (specifically, the sacral ribs) is different than that of Euoplocephalus. I reconstructed the muscles of the tail and pelvis in my 2009 PLoS ONE paper on tail clubbing. Twice now I have spent significant portions of my summer preparing the pelvis of two different ankylosaur specimens. And now this month I have another paper (again with Mike Burns and my supervisor Phil Currie) on the ankylosaur pelvis, this time on the armour of the pelvic region.

The lovely behind of Aletopelta, from my visit to the San Diego Natural History Museum in 2009. It is an excellent museum, you should check it out!

It has been recognized for a long time now that some ankylosaurs were doing really weird things with the armour over their pelvis. In some taxa, the osteoderms fuse together to form a carapace-like shield over the hips. Previously, the presence or absence of this shield has been used in phylogenetic analyses to examine whether or not a third group of ankylosaurs, the Polacanthidae, is a valid taxon (in addition to the Nodosauridae and Ankylosauridae). However, it is not just the presence or absence of the pelvic shield, but the way that the pelvic shield is constructed, that may be important.

In this paper I and my coauthors propose a revised way of looking at the pelvic shield that breaks shields up into three categories: 1. fused, rosette pattern, 2. fused, uniform-sized polygons, and 3. not fused. “But Victoria, if the osteoderms aren’t fused, then why call it a pelvic shield?” you may ask. Well, after looking at the AMNH’s Sauropelta and the BMNH’s Euoplocephalus, I noticed that although the osteoderms of the pelvis weren’t coossified, there also weren’t any transverse bands segmenting the body in that region (see image below of BMNH R5161, the holotype of Scolosaurus discussed in this post). Although the osteoderms aren’t fused together, they still form a continuous shield over the pelvis.

The always exciting headless, clubless, BMNH R5161, modified from Nopsca's 1928 paper and as seen in Arbour et al. 2011.

More interesting, though, is what happens when we look at the other two categories – rosettes vs. uniform polygons – in a stratigraphic and geographic context. Rosettes are restricted to the Jurassic and Lower Cretaceous, whereas uniform polygons are primarily Upper Cretaceous. Uniform polygon shields are only found in North America (with the possible exception of Antarctopelta...which is from Antarctica). We did not run a phylogenetic analysis for this paper, but I plan to incorporate this data into subsequent analyses stemming from my PhD thesis on the phylogenetic relationships and biogeography of the Ankylosauridae. I will be very interested to see if any clades show up that reflect these stratigraphic and geographic patterns.


Here’s the paper! (behind a paywall, unfortunately...):

Arbour VM, Burns ME, Currie PJ. 2011. A review of pelvic shield morphology in ankylosaurs (Dinosauria: Ornithischia). Journal of Paleontology 85:298-302.
 
 
More papers about ankylosaur pelves!
 
Arbour VM, Burns ME, Sissons RL. 2009. A redescription of the ankylosaurid dinosaur Dyoplosaurus acutosquameus Parks, 1924 (Ornithischia: Ankylosauria) and a revision of the genus. Journal of Vertebrate Paleontology 29:1117–1135.
 
Arbour VM. 2009. Estimating impact forces of tail club strikes by ankylosaurid dinosaurs. PLoS ONE 4: e6738.