Learn to be sceptical by seeing how the sausage is made

September 16, 2022

Years ago, when I was young and stupid, I used to read papers containing phylogenetic analyses and think, “Oh, right, I see now, Euhelopus is not a mamenchisaurid after all, it’s a titanosauriform”. In other words, I believed the result that the computer spat out. Some time after that, I learned how to use PAUP* and run my own phylogenetic analysis and realised how vague and uncertain such result are, and how easily changed by tweaking a few parameters.

These days good papers that present phylogenetic analysis are very careful to frame the results as the tentative hypotheses that they are. (Except when they’re in Glam Mags, of course: there’s no space for that kind of nuance in those venues.)

It’s common now for careful work to present multiple different and contradictory phylogenetic hypotheses, arrived at by different methods or based on different matrices. For just one example, see how Upchurch et al.’s (2015) redescription of Haestasaurus (= “Pelorosaurus“) becklesii presents that animal as a camarasaurid (figure 15, arrived at by modifying the matrix of Carballido at al 2011), as a very basal macronarian (figure 16, arrived at by modifying the continuous-and-discrete-characters matrix of Mannion et al. 2013), and as a basal titanosaur (figure 17, arrived at by modifying the discrete-characters-only matrix from the same paper). This is careful and courageous reporting, shunning the potential headline “World’s oldest titanosaur!” in favour of doing the work right.) [1]

But the thing that really makes you understand how fragile phylogenetic analyses are is running one yourself. There’s no substitute for getting your hands dirty and seeing how the sausage is made.

And I was reminded of this principle today, in a completely different context, by a tweet from Alex Holcombe:

Some of us lost our trust in science, and in peer review, in a journal club. There we saw how many problems a bunch of ECRs notice in the average article published in a fancy journal.

Alex relays (with permission) this anecdote from an anonymous student in his Good Science, Bad Science class :

In the introduction of the article, the authors lay forth four very specific predictions that, upon fulfillment, would support their hypothesis. In the journal club, one participant actually joked that it read very much as though the authors ran the analysis, derived these four key findings, and then copy-pasted them in to the introduction as though they were thought of a priori. I’m not an expert in this field and I don’t intend to insinuate that anything untoward was done in the paper, but I remember several participants agreeing that the introduction and general framework of the paper indeed felt very “HARKed“.

Here’s the problem: as the original tweet points out, this is about “problems a bunch of ECRs notice in the average article published in a fancy journal”. These are articles that have made it through the peer-review gauntlet and reached the promised land of publication. Yet still these foundational problems persist. In other words, peer-review did not resolve them.

I’m most certainly not suggesting that the peer-review filter should become even more obstructive than it is now. For my money it’s already swung way too far in that direction.

But I am suggesting we should all remain sceptical of peer-reviewed articles, just as we rightly are of preprints. Peer-review ain’t nuthin’ … but it ain’t much. We know from experiment that the chance of an article passing peer review is made up of one third article quality, one third how nice the reviewer is and one third totally random noise. More recently we found that papers with a prestigious author’s name attached are far more likely to be accepted, irrespective of the content (Huber et al. 2022).

Huber et al. 2022, figure 1.

We need to get away from a mystical or superstitious view of peer-review as a divine seal of approval. We need to push back against wise-sounding pronouncements such as “Good reporting would have noted that the paper has not yet been peer-reviewed” as though this one bit of information is worth much.

Yeah, I said it.

References

Notes

  1. Although I am on the authorship of Upchurch et al. (2015), I can take none of the credit the the comprehensiveness and honesty of the phylogenetics section: all of that is Paul and Phil’s work.

 

Posted by Mike Taylor
Filed in Did I just say that out loud?, Haestasaurus, Peer review
1 Comment »

Ask SV-POW!: Could you wield a club made from a dinosaur femur?

April 11, 2018

Norwescon 41 Guests of Honor: Ken Liu, Galen Dara, and, er, me. Mike would like to remind you that you can get your own ‘Kylo Stabbed First’ t-shirt here.

The week before last I was fortunate to be the Science Guest of Honor at Norwescon 41 in Seattle (as threatened back when). I had a fantastic time. I got to give talks on binocular stargazing and the sizes of the largest sauropods and whales (ahem), participate on panels on alien biology and creature drawing, and meet a ton of cool people, including my fellow Guests of Honor, multiple-award-winning author Ken Liu and multiple-award-winning artist Galen Dara, both of whom turned out to be humble, easygoing, regular folks (if frighteningly talented).

I also had a lot of great conversations with folks who were attending the con, which is exactly what I wanted. One of the most interesting was a hallway conversation with a fellow DM named Shawn Connor. He had a great question for me, which I liked so much I wanted to answer it here on the blog. Here’s his question, copied with permission from a follow-up email:

I run tabletop RPGs, and in my current game one of the characters is a caveman type who naturally grew up hunting dinosaurs. As one does. His weapon is a dinosaur bone, customized and used as a club. I have attached the picture that he came up with [below]. Now understanding the picture is obviously not of a real dinosaur bone – it’s probably a chicken bone or a cow bone or something – let’s assume for the sake of this exercise that it is and that it is four feet long stem to stern. Given that, two questions: discounting the extra bling attached how heavy would such a bone be, and what kind of dinosaur could it have come from?

I’m going to answer those questions out of order. Advance warning: this will be a loooong post that will go down several rabbit holes that are likely of more intense interest to me, personally, than to anyone else on the planet. Read on at your own risk.

Whose femur is in the image?

First, Shawn is correct in noting that the femur in the image provided by his player is not a dinosaur femur. The prominent trochanters and spherical head offset on a narrow neck clearly make it a mammal femur, and if it’s four feet long, it could only have come from an elephant or an indricothere. Or a giant humanoid, I suppose, which is what the anatomy of the bone in the image most closely resembles. (It also appears to be foreshortened to make the distal end look bigger, or deliberately distorted to enhance the clubby-ness.)

Mounted elephant at the Museum of Osteology in Oklahoma City, with Tyler Hunt for scale.

But let’s play along and assume it’s from a non-human mammal. How big? Back in 2016 I was fortunate to get to measure most of the mounted large mammal skeletons at the Museum of Osteology in Oklahoma City, along with Tyler Hunt, then a University of Oklahoma undergrad and now finishing up his MS thesis under my mentor, Rich Cifelli.* The mounted elephant at the Museum of Osteology has a shoulder height of 254 cm (8 ft, 4 in) and a femur length of 102 cm (3 ft, 4 in). Assuming isometric scaling, a world record elephant with a shoulder height of 366 cm (12 ft) would have a femur length of 147 cm (4 ft, 10 in). So a four-foot (122 cm) femur would belong to an elephant roughly in the middle of that range, about ten feet (3 m) tall at the shoulder. That’s the size of the big bull elephant mounted at the Field Museum in Chicago.

The big mounted bull elephant at the Field Museum is 10 feet tall at the shoulder and weighed 6 tons in life. Note Mike for scale on the lower right. He and the elephant are about equidistant from the camera, so he should make a roughly accurate scale bar. Photo from our visit in 2005!

* Two further notes: first, I have roughly a zillion awesome photos from that 2016 visit to the Museum of Osteology, both of the specimens and of Tyler and me measuring them – not having posted them yet is one of the things I was whingeing about in the post that kicked off our return-to-weekly-posting thing this year. And second, I owe a belated and public thanks to the folks at the Museum of Osteology for accommodating Tyler and me. They helped us with ladders and so on and basically gave us free rein to play with collect data from their mounted skeletons, which was incredibly generous and helpful, and fortunately reflects the pro-research and pro-researcher attitude of most museums.

Which dinos had four-foot femora?

As for what kind of dinosaur a four-foot femur could have come from, we can rapidly narrow it down to a handful of clades: sauropods, ornithopods, theropods, and stegosaurs.

  • Sauropods. The longest complete femora of Patagotitan are 238 cm (7 ft, 10 in; Carballido et al. 2017), and an incomplete femur of Argentinosaurus has an estimated complete length of 250 cm (8 ft, 2 in; Mazzetta et al. 2004). So a four-foot femur would not be from a particularly large sauropod – something about elephant-sized, as you might expect from the elephant comparison above. Our old friend Haplocanthosaurus will fit the bill, as we’ll see in a bit.
  • Ornithopods. Femora of 172 cm (5 ft, 8 in) are known for the hadrosaurs Shantungosaurus (Hone et al. 2014) and Huaxiaosaurus (Zhao and Li 2009), and Zhao et al. (2007) reported a 170 cm (5 ft, 7 in) femur for Zhuchengosaurus (Huaxiaosaurus and Zhuchengosaurus may be junior synonyms of Shantungosaurus). But those are all monsters, well over 10 metric tons in estimated mass. So a four-foot femur would be from a large but not insanely large hadrosaur.

Mmmmmm…suffering. OM NOM NOM NOM!!

  • Theropods. Among the largest theropods, the holotype of Giganotosaurus has a femur length of 143 cm (4 ft, 8 in; Coria and Salgado 1995), and ‘Sue’ the T. rex (a.k.a. FMNH PR2081) has a right femur 132 cm long (4 ft, 4 in; Brochu 2003). So a four-foot femur from a theropod would definitely be from one of the monsters. The femur of Saurophaganax was 113.5 cm long (Chure 1995), just under four feet, which I only note as an excuse to use the above photo, which I adore.
  • Stegosaurs. I don’t know the longest femur that has been recovered from a stegosaur, but getting in the ballpark is easy. NHMUK PV R36730 has a femur 87 cm long, and the whole animal was approximately 6 m long (Maidment et al. 2015). Partial bits and bobs of the largest stegosaurs suggest animals about 9 m long, implying a femur length of about 130 cm (4 ft, 3 in), or just over the line.

I think that’s it. I don’t know of any ceratopsians or ankylosaurs with femora long enough to qualify – I assume someone will let me know in the comments if I’ve forgotten any.

How much would a four-foot femur weigh?

There are a couple of ways to get to the answer here. One is to use Graphic Double Integration, which is explained in this post.

Limb bones are not solid – in terrestrial tetrapods there is virtually always a marrow cavity of some sort, and in marine tetrapods the limb bones tend to be cancellous all the way through. Estimating the mass of a limb bone is a lot like estimating the mass of a pneumatic bone: figure out the cross-sectional areas of the cortex and marrow cavity (or air space if the bone is pneumatic), multiply by the length of the element to get volumes, and multiply those volumes by the density of the materials to get masses. I piled up all the relevant numbers and formulas in Tutorial 24, a move that has frequently made me grateful to my former self (instead of cussing his lazy ass, which is my more usual attitude toward Past Matt).

Currey and Alexander (1985: fig. 1)

Sauropod limb bones are pretty darned dense, with extremely thick cortices and smallish marrow spaces that are not actually hollow (tubular) but are instead filled with trabecular bone. My gut feeling is that even a four-foot sauropod femur would be almost too heavy to lift, let alone wield as a club, so in the coming calculations I will err in the direction of underestimating the mass, to give our hypothetical caveman the best possible chance of realizing his dream.

Some of the proportionally thinnest cortices I’ve seen in sauropod limb bones are those of the macronarian Haestasaurus becklesii NHMUK R1870, which Mike conveniently showed in cross-section in this post. I could look up the actual dimensions of the bones (in Upchurch et al 2015: table 1 – they passed the MYDD test, as expected), but for these calculations I don’t need them. All I need are relative areas, for which pixels are good enough.

First, I took Mike’s photo into GIMP and drew two diameters across each bone, one maximum diameter and a second at right angles. Then I drew tick marks about where I think the boundaries lie between the cortex and the trabecular marrow cavity. Next, I used those lines as guides to determine the outer diameters (D) and inner diameters (d) in pixels, as noted in the image.

For the radius, on the left, the mean diameters are D = 891 and d = 648. I could divide those by 2 to get radii and then plug them into the formula for the area of a circle, etc., but there’s an easier way still. For a tubular bone, the proportional area of the inner circle or ellipse is equal to k^2, where k = r/R. Or d/D. (See Wedel 2005 and Tutorial 24 for the derivation of that.) For the Haestasaurus radius (the bone, not the geometric dimension), d/D = 0.727, and that number squared is 0.529. So the marrow cavity occupies 53% of the cross-sectional area, and the cortex occupies the other 47%.

For the ulna, on the right, the mean diameters are D = 896 and d = 606, d/D = 0.676, and that number squared is 0.457. So in this element, the marrow cavity occupies 46% of the cross-sectional area, and the cortex occupies the other 54%.

(For this quick-and-dirty calculation, I am going to ignore the fact that limb bones are more complex than tubes and that their cross-sectional properties change along their lengths – what I am doing here is closer to Fermi estimation than to anything I would publish. And we’ll ground-truth it before the end anyway.)

Left: rat humerus, right: mole humerus. The mole humerus spits upon my simple geometric models, with extreme prejudice. From this post.

You can see from the photo (the Haestasaurus photo, not the mole photo) that neither bone has a completely hollow marrow cavity – both marrow cavities are filled with trabecular bone. By cutting out good-looking chunks in GIMP and thresholding them, I estimate that these trabecular areas are about 30% bone and 70% marrow (actual marrow space with no bone tissue) by cross-sectional area. According to Currey and Alexader (1985: 455), the specific gravities of fatty marrow and bone tissue are 0.93 and 2.1, respectively. The density of the trabecular area is then (0.3*2.1)+(0.7*0.93) =  1.28 kg/L, or about one quarter more dense than water.

But that’s just the trabecular area, which accounts for about one half of the cross-sectional area of each bone. The other half is cortex, which is probably close to 2.1 kg/L throughout. The estimated whole-element densities are then:

Radius: (0.53*1.28)+(0.47*2.1) = 1.67 kg/L

Ulna: (0.46*1.28)+(0.54*2.1) = 1.72 kg/L

Do those numbers pass the sniff test? Well, any skeletal elements that are composed of bone tissue (SG = 2.1) and marrow (SG = 0.93) are constrained to have densities somewhere between those extremes (some animals beat this by building parts of their skeletons out of [bone tissue + air] instead of [bone tissue + marrow]). We know that sauropod limb bones tend to have thick cortices and small marrow cavities, and that the marrow cavities are themselves a combination of trabecular bone and actual marrow space, so we’d expect the overall density to be closer to the 2.1 kg/L end of the scale than the 0.93 kg/L end. And our rough estimates of ~1.7 kg/L fall about where we’d expect.

Femur of Haplocanthosaurus priscus, CM 572, modified from Hatcher (1903: fig. 14).

To convert to masses, we need to know volumes. We can use Haplocanthosaurus here – the femur of the holotype of H. priscus, CM 572, is 1275 mm long (Hatcher 1903), which is just a hair over four feet (4 ft, 2.2 in to be exact). The midshaft width is 207 mm, and the proximal and distal max widths are 353 and 309 mm, respectively. I could do a for-real GDI, but I’m lazy and approximate numbers are good enough here. Just eyeballing it, the width of the femur is about the same over most of its length, so I’m guessing the average width is about 23 cm. The average width:length ratio for the femora of non-titanosaur sauropods is 3:2 (Wilson and Carrano 1999: table 1), which would give an anteroposterior diameter of about 15 cm and an average diameter over the whole length of 19 cm. The volume would then be the average cross-section area, 3.14*9.5*9.5, multiplied by the length, 128 cm, or 36,273 cm^3, or 36.3 L. Multiplied by the ~1.7 kg/L density we estimated above, that gives an estimated mass of 62 kg, or about 137 lbs. A femur that was exactly four feet long would be a little lighter – 86.6% as massive, to be exact, or 53.4 kg (118 lbs).

I know that the PCs in RPGs are supposed to be heroes, but that seems a little extreme.

But wait! Bones dry out and they lose mass as they do so. Lawes and Gilbert (1859) reported that the dry weight of bones of healthy sheep and cattle was only 74% of the wet mass. Cows and sheep have thinner bone cortices than sauropods or elephants, but it doesn’t seem unreasonable that a dry sauropod femur might only weigh 80% as much as a fresh one. That gets us down to 43 kg – about 95 lbs – which is still well beyond what anyone is probably going to be wielding, even if they’re Conan the Cimmerian.

Picture is unrelated.

I mentioned at the top of this section that there are a couple of ways to get here. The second way is to simply see what actual elephant femora weigh, and then scale up to dinosaur size. According to Tefera (2012: table 1), a 110-cm elephant femur has a mass of 21.5 kg (47 lbs). I reckon that’s a dry mass, since the femur in question had sat in a shed for 50 years before being weighed (Tefera 2012: p. 17). Assuming isometry, a four-foot (122 cm) elephant femur would have a dry mass of 29.4 kg (65 lbs). That’s a lot lighter than the estimated mass of the sauropod femur – can we explain the discrepancy?

 

Femora of a horse, a cow, and an elephant (from left to right in each set), from Tefera (2012: plate 1).

I think so. Elephant femora are more slender than Haplocanthosaurus femora. Tefera (2012) reported a circumference of 44 cm for a 110-cm elephant femur. Scaling up from 110 cm to 122 cm would increase that femur circumference to 49 cm, implying a mean diameter of 15.6 cm, compared to 19 cm for the Haplo femur. That might not seem like a big difference, but it means a cross-sectional area only 2/3 as great, and hence a volume about 2/3 that of a sauropod femur of the same length. And that lines up almost eerily well with our estimated masses of 29 and 43 kg (ratio 2:3) for the four-foot elephant and sauropod femora.

A Better Weapon?

Could our hypothetical caveman do better by choosing a different dinosaur’s femur? Doubtful – the femora of ‘Sue’ are roughly the same length as the Haplo femur mentioned above, and have similar cross-sectional dimensions. Hadrosaur and stegosaur femora don’t look any better. Even if the theropod femur was somewhat lighter because of thinner cortices, how are you going to effectively grip and wield something 15-19 cm in diameter?

I note that the largest axes and sledgehammers sold by Forestry Suppliers, Inc., are about 3 feet long. Could we get our large-animal-femur-based-clubs into the realm of believability by shrinking them to 3 feet instead of 4? Possibly – 0.75 to the third power is 0.42. That brings the elephant femur club down to 12.3 kg (27 lbs) and the sauropod femur club down to 18 kg (40 lbs), only 2-3 times the mass of the largest commonly-available sledgehammers. I sure as heck wouldn’t want to lug such a thing around, much less swing it, but I can just about imagine a mighty hero doing so.

Yes, there were longer historical weapons. Among swing-able weapons (as opposed to spears, etc.), Scottish claymores could be more than four feet long, but crucially they were quite light compared to the clubs we’ve been discussing, maxing out under 3 kg, at least according to Wikipedia.

T. rex FMNH PR2081 right fibula in lateral (top) and medial (bottom) views. Scale is 30 cm. From Brochu (2003: fig. 97).

If one is looking for a good dinosaur bone to wield as a club, may I suggest the fibula of a large theropod? The right (non-pathologic) fibula of ‘Sue’ is 103 cm long (3 ft, 4.5 in), has a max shaft diameter just under 3 inches – so it could plausibly be held by (large) human hands, and it probably massed something like 8-9 kg (17-20 lbs) in life, based on some quick-and-dirty calculations like those I did above. The proximal end is even expanded like the head of a war club. The length and mass are both in the realm of possibility for large, fit, non-supernaturally-boosted humans. Half-orc barbarians will love them.

And that’s my ‘expert’ recommendation as a dice-slinging paleontologist. Thanks for reading – you have Conan-level stamina if you got this far – and thanks to Shawn for letting me use his question to freewheel on some of my favorite geeky topics.

References

Posted by Matt Wedel
Filed in elephant, fame, femur, fibula, goofy, Haestasaurus, Haplocanthosaurus, mass estimates, math, mole, People we like, radius, Saurophaganax, stinkin' appendicular elements, stinkin' mammals, stinkin' SV-POW!sketeers, stinkin' theropods, Tyrannosaurus, ulna
17 Comments »

Correction: there is no support for Haestasaurus as a derived titanosaur

June 6, 2015

In my blog-post announcing Haestasaurus as the new generic name for the misassigned species “Pelorosaurus” becklesii, I briefly surveyed the three phylogenetic analyses in the paper. Of the third — the one based on the Mannion et al. (2013) Lusotitan matrix using both discrete and continuous characters — I wrote that it …

… recovers Haestasaurus as a titanosaur — as sister to Diamantinasaurus and then Malawisaurus, making it a lithostrotian well down inside Titanosauria.

My mistake! I was working from the result of an earlier version of that analysis. In the final version included in the paper, things are rather different:

Fig 17. Strict consensus tree (LCDM). A strict consensus tree based on the 17 most parsimonious trees generated by analysis of the Mannion et al. [18] LCDM with the revised scores for Haestasaurus and the addition of six new characters. GC values (multiplied by 100) are shown in square brackets for all nodes where these values are greater than 0. Abbreviations: Brc, Brachiosauridae; Dd, Diplodocoidea. N.B. the tree topology shown here means that the clades defined by Brachiosaurus+Saltasaurus (Titanosauriformes) and Andesaurus+Saltasaurus (Titanosauria) are identical. See main text for details.

Upchurch et al. (2015: Fig 17). Strict consensus tree (LCDM).
A strict consensus tree based on the 17 most parsimonious trees generated by analysis of the Mannion et al. [18] LCDM with the revised scores for Haestasaurus and the addition of six new characters. GC values (multiplied by 100) are shown in square brackets for all nodes where these values are greater than 0. Abbreviations: Brc, Brachiosauridae; Dd, Diplodocoidea. N.B. the tree topology shown here means that the clades defined by Brachiosaurus+Saltasaurus (Titanosauriformes) and Andesaurus+Saltasaurus (Titanosauria) are identical. See main text for details.

As you can see, Haestasaurus is indeed a titanosaur in this analysis — but not a derived one at all. In fact, it’s part of the most basal clade of titanosaurs, along with Janenschia and Dongbeititan. In this tree, we have a really nice, big Brachiosauridae, containing 19 OTUs split fairly evenly between two subclades.

[Side-note: Upchurch et al. (2015) uses phylogenetic definitions that I’m not crazy about. I prefer the arrangement that I followed in my brachiosaur paper (Taylor 2009), in which Titanosauriformes = Brachiosauridae + Titanosauria is a node-stem triplet. Hopefully, some time soon, the wretched PhyloCode will finally be implemented, and we’ll be in a position to nail down a single set of definitions for the whole community to use.]

Anyway, the upshot of all this is that all three phylogenetic analyses in the paper return Haestasaurus as a pretty basal macronarian, and on the balance of evidence it’s likely not a titanosaur after all. (That’s why the name “Haestatitan“, which was in some earlier drafts of the paper, was changed to Haestasaurus. Kind of a shame, given how mundane -saurus names are, but probably the wisest course of action.)

What is the takeaway lesson from this? It’s not just “Haestasaurus is not a derived titanosaur”. It’s that all our phylogenetic hypotheses are just that — hypotheses. Papers that publish only a single cladogram are always at risk of being misinterpreted as conveying much more certainty than they really do, and Paul and Phil are to be commended for including the whole messy story in this paper. The position of Haestasaurus shifts around far too easily for us to have a strong sense of what it is, and it’s good that the paper makes that clear.

(It also makes me glad that way back in Taylor and Naish (2007), I and Darren didn’t give a more precise position of Xenoposeidon than that it’s probably some kind of neosauropod. And even that is not something I would put money on.)

References

Posted by Mike Taylor
Filed in Haestasaurus
7 Comments »

Looks like “Pelorosaurus” becklesii is actually Haestasaurus becklesii!

June 3, 2015

Well, who knew? There I was posting images of “Pelorosaurus” becklesi‘s humerus, radius and ulna, and skin impression. There I was saying that this beast is due a proper description, and warrants its own generic name. And what should come out today but a new paper by Paul Upchurch, Phil Mannion and, oh yes, me, which does exactly that.

Screen Shot 2015-06-03 at 19.05.12

The headline news is the long-overdue establishment of a new genus name for this species — something that we’ve known was needed at least since Upchurch’s (1993) dissertation. Paul and Phil came up with the name Haestasaurus, from “Haesta”, the name of the putative pre-Roman chieftain whose people apparently settled the area of Hastings and gave the town its name. It’s nice that I can finally stop typing the scare-quotes around the no-longer-relevant old genus name “Pelorosaurus“!

Upchurch et al. 2015: figure 2. Left humerus of Haestasaurus becklesii (NHMUK R1870). A, anterior view; B, posterior view; Abbreviations: af, anconeal fossa; dp, deltopectoral crest; hh, humeral head; ltf, lateral triceps fossa; mtf, medial triceps fossa.

Upchurch et al. (2015: figure 2). Left humerus of Haestasaurus becklesii (NHMUK R1870). A, anterior view; B, posterior view; Abbreviations: af, anconeal fossa; dp, deltopectoral crest; hh, humeral head; ltf, lateral triceps fossa; mtf, medial triceps fossa.

(As you can see, the photography is rather better than in my own illustrations, which I made independently some years ago.)

Of course Paul has had an eye on this work, on and off, since the early 1990s. Then in the late 2000s, when I was working on Xenoposeidon and other Wealden sauropods, I started work independently on a redescription — which of course is why I prepared the figures that have appeared in the last few posts. But that work petered out as I started working more on other specimens and on the problems of the sauropod neck. More recently, Paul and Phil hunkered down and got the nitty-gritty descriptive work done.

Once they had a complete draft manuscript, they very graciously invited me onto the authorship — not something they had to do, but they chose to based on my previous interest in the specimen. My contribution was minor: I provided two of the illustrations, tidied up the early versions of several others, and did an editing pass on the text.

Upchurch et al. (2015: figure 1). Map showing England and Wales, with boundaries for English counties. The magnified inset shows the Isle of Wight and East and West Sussex in more detail, marking the positions of selected major towns/cities and the fossil localities mentioned in the main text. Based on

Upchurch et al. (2015: figure 1). Map showing England and Wales, with boundaries for English counties. The magnified inset shows the Isle of Wight and East and West Sussex in more detail, marking the positions of selected major towns/cities and the fossil localities mentioned in the main text. Based on “English ceremonial counties 1998” by Dr. Greg, http://en.wikipedia.org/wiki/File:English_ceremonial_counties_1998.svg. CC By-SA 3.0.

(This map is one of the two illustrations that I provided; the other is the multi-view photograph of the Pelorosaurus conbeari humerus.)

I’m grateful to Paul and Phil, both for inviting me onto this project, and for taking into account my strong preference for an open-access venue. It’s largely because of the latter that the paper now appears in PLOS ONE, where the glorious colour illustrations appear at full resolution and may be re-used for any purpose subject to attribution.

So: what actually is Haestasaurus? Is it the early titanosaur that we’ve all been assuming? The unexciting answer is: we don’t really know. Our paper contains three phylogenetic hypotheses (all of them Paul and Phil’s work, I can’t take any credit). These results are from adding Haestasaurus to the Carballido and Sander (2014) matrix, to the Mannion et al. (2013) standard discrete matrix and to the Mannion et al. (2013) continuous-and-discrete matrix. Only the last of these recovers Haestasaurus as a titanosaur — as sister to Diamantinasaurus and then Malawisaurus, making it a lithostrotian well down inside Titanosauria.

Both both of the other analyses find Haestasaurus as a very basal macronarian — outside of Titanosauriformes. Here is the result of the analysis based on Carballido and Sander’s Europasaurus matrix:

Upchurch et al. (2105: figure 15). Strict consensus tree (CSM). A strict consensus tree based on the 28 most parsimonious trees generated by analysis of the Carballido and Sander [19] data matrix with the addition of Haestasaurus and six new characters (Tendaguria excluded a priori). GC values (multiplied by 100) are shown in square brackets for all nodes where these values are greater than 0. The monophyletic Diplodocoidea has been collapsed to a single branch in order to reduce figure size. Abbreviation: Brc, Brachiosauridae.

Upchurch et al. (2105: figure 15). Strict consensus tree (CSM). A strict consensus tree based on the 28 most parsimonious trees generated by analysis of the Carballido and Sander [19] data matrix with the addition of Haestasaurus and six new characters (Tendaguria excluded a priori). GC values (multiplied by 100) are shown in square brackets for all nodes where these values are greater than 0. The monophyletic Diplodocoidea has been collapsed to a single branch in order to reduce figure size. Abbreviation: Brc, Brachiosauridae.

As you can see, Haestasaurus is here a camarasaurid, making it (along with Camarasaurus itself) the most basal of all macronarians. In the second analysis — the one using discrete characters only from Mannion et al.’s Lusotitan paper — Haestasaurus is again in the most basal macronarian clade, but this time as sister to Janenschia and then Tehuelchesaurus. (In this topology, Camarasaurus is the next most basal macronarian after that three-taxon clade.)

So it looks like Haestasaurus is either a very basal macronarian or a pretty derived titanosaur. We don’t know which.

But, hey, at least it has a proper name now!

Acknowledgements

It’s Matt’s birthday today. I’d like to dedicate a sauropod to him, but I don’t have the authority to do that. So instead, I dedicate this blog-post to him, and declare it the Mathew J. Wedel Memorial Blog Post.

References

Posted by Mike Taylor
Filed in Haestasaurus, humerus, papers by SV-POW!sketeers, stinkin' appendicular elements, timely, Wealden
11 Comments »

The skin impression of “Pelorosaurus” becklesii

June 2, 2015

We’ve seen the humerus of the Wealden-supergroup putative titanosaur “Pelorosaurus” becklesi. We’ve seen the bones of the forearm, the radius and ulna. That’s it for bony remains: no other bones have been found.

But there is one other fossil that’s part of the same specimen: this skin impression:

Skin impression of

Skin impression of “Pelorosaurus” becklesii holotype NHMUK R1868. (Note that the other elements of this specimen are all catalogued as R1870.)

As you can see, the body — or at least this part of the body — was covered with roughly hexagonal tessellating (non-overlapping) scales, of about 1-2 cm diameter. But what part of the body is it from? The initial — extremely brief — description of this specimen, by Mantell (1852:143) hardly mentions the skin impression at all. All it says is:

A portion of the scaly cuirass which covered the limbs and is composed of hexagonal plates, was exhibited.

Still, this does at least suggest that the skin impression was from a limb — hardly surprising given then the left forelimb was the only part of the skeleton recovered. Upchurch et al. (2004:295) were more specific:

This skin impression was found in close association with the elbow region of the forelimb of Pelorosaurus becklesii.

But I don’t know whether this assertion is based on something written earlier, or is just a surmise.

Assuming that the skin impression is indeed from the elbow, and putting it all together, here’s what we know of “Pelorosaurus” becklesii:

Schematic of

Schematic of “Pelorosaurus” becklesi holotype NHMUK R1870, showing the preserved humerus, radius, ulna, and skin impression of the elbow. Based on Scott Hartman’s skeletal reconstruction of Alamosaurus. This is not a skeletal reconstruction of “Pelorosaurus” becklesi.

It’s not much, but it’s enough to be diagnostic.

But what actually is this beast? A titanosaur, as often assumed? A more basal macronarian? Something else entirely? Who can tell? Someone really ought to get onto that.

References

  • Mantell, Gideon A. 1852. On the structure of the Iguanodon, and on the fauna and flora of the Wealden Formation. Notices of the proceedings at the meetings of the members of the Royal Institution, with abstracts of the discourses delivered at the evening meetings 1:141-146.
  • Upchurch, Paul, Paul M. Barrett and Peter Dodson. 2004. Sauropoda. pp. 259-322 in D. B. Weishampel, P. Dodson and H. Osmólska (eds.), The Dinosauria, 2nd edition. University of California Press, Berkeley and Los Angeles. 861 pp.
Posted by Mike Taylor
Filed in Haestasaurus, teaser, timely, Wealden
6 Comments »

What about the forearm of “Pelorosaurus” becklesii?

June 2, 2015

Yesterday, we looked at (mostly) the humerus of the Wealden sauropod “Pelorosaurus” becklesii, which you will recall is known from humerus, radius, ulna and a skin impression, and — whatever it might be — is certainly not a species of Pelorosaurus.

Now let’s look at the radius and ulna.

Left forearm of

Left forearm of “Pelorosaurus” becklesii holotype NHMUK R1870, articulated, in anterior view, with proximal to the left: radius in front, ulna behind.

They fit together pretty neatly: the proximal part of the radius is a rounded triangular shape, and it slots into the triangular gap between the anteromedial and anterolateral processes of the proximal part of the ulna.

Left forearm of “Pelrosaurus” becklesii holotype NHMUK R1870 in proximal view, with anterior to the right. The arms of the ulna enclose the radius.

Left forearm of “Pelorosaurus” becklesii holotype NHMUK R1870 in proximal view, with anterior to the right. The “arms” of the ulna enclose the radius.

Let’s take a closer look at the ulna:

Left ulna of

Left ulna of “Pelorosaurus” becklesii holotype NHMUK R1870. Top row: proximal view, with anterior to the bottom. Middle row, from left to right: medial, anterior, lateral and posterior views. Bottom row: distal view, with anterior to top.

And the radius:

Left radius of

Left radius of “Pelorosaurus” becklesii holotype NHMUK R1870. Top row: proximal view, with anterior to the bottom. Middle row, from left to right: medial, anterior, lateral and posterior views. Bottom row: distal view, with anterior to top.

As you can see, it’s pretty well preserved: there’s no evidence of significant crushing in any of the bones, and the 3d shape is apparent.

In short, it’s a really sweet specimen. Someone really ought to get around to describing it properly, and giving it the new generic name that it clearly warrants.

Posted by Mike Taylor
Filed in Haestasaurus, radius, stinkin' appendicular elements, teaser, timely, ulna, Wealden
15 Comments »

Whatever happened to “Pelorosaurus” becklesii?

June 1, 2015

It’s an oddity that in eight years of SV-POW!, we’ve never written about one of the best of all the Wealden-formation sauropod specimens: the forelimb and associated skin impression NHMUK R1870 that is known as “Pelorosaurus” becklesii.

Let’s fix that. Here is all the bony material (i.e. everything except the skin patch) in a photo taken in the basement of the Natural History Museum back in 2007:

Left forelimb material of

Left forelimb material of “Pelorosaurus” becklesii holotype NHMUK R1870. Left: humerus, in posterior view. Right, from top to bottom: ulna in anterior view; radius in anterior view. Yes, I should have turned the humerus over before taking this photo. What can I tell you? I was young and stupid then.

As you can see, the two lower-limb bones were broken back then (though I believe they have since been repaired), but the breaks are very clean, and it’s actually quite interesting to see inside the bones:

Breakage in bones of the lower left forelimb of

Breakage in bones of the lower left forelimb of “Pelorosaurus” becklesii holotype NHMUK R1870. Left: proximal part of radius in distal view. Right: proximal part of ulna in distal view.

I wish I knew enough about mineralisation to comment intelligently on what we can see there. If anyone has thoughts, do leave them in the comments.

We can look in more detail at those lower-limb bones in a subsequent post, but for now, here’s the humerus:

“Pelorosaurus” becklesii holotype NHMUK R1870, left humerus. Top row: proximal view, with anterior to the bottom. Middle row, from left to right: medial, anterior, lateral and posterior views. Bottom row: distal view, with anterior to the top.

As you can see it’s in really nice shape, and pretty distinctive. Way back in my 2007 Progressive Palaeo talk (Taylor 2007), I coded up the humerus (alone, without the other elements) in the Harris-based phylogenetic matrix that I’ve used repeatedly in other projects. It came out as the sister taxon to the titanosaur Malawisaurus (which in that matrix comes out fairly basal within Titanosauria): in fact, it could hardly do anything else, since the coding was exactly the same as that of Malawisaurus.

And indeed it’s been pretty widely accepted that “P.” becklesii is a titanosaur — one of the earliest known, and the only name-bearing one from the Wealden Supergroup, unless you count the extremely indeterminate Iuticosaurus, which predictably enough is based on a single eroded partial mid-caudal centrum. Still, the titanosaurian identity of “P.” becklesii has never been convincingly demonstrated — only inferred by non-cladistic means.

“Pelorosaurus” becklesii holotype NHMUK R1870, left humerus in anterodistal view (anterior to the left).

So why the quotes around the genus name “Pelorosaurus“? Because it’s long been recognised that, whatever this specimen might be, it ain’t Pelorosaurus, which is based on the “Cetiosaurus” brevis caudals and a much more slender humerus.

Here’s that humerus, so you can see how different it is from that of “Pelorosaurus” becklesii:

Right humerus of Pelorosaurus conybeari holotype NHMUK 28626. Top row: distal view, anterior to bottom. Middle row, left to right: lateral, anterior and medial views. Bottom row: distal, anterior to top. Missed parts reconstructed from the humerus of Giraffatitan brancai (Janensch 1961: Beilage A)

Right humerus of Pelorosaurus conybeari holotype NHMUK 28626. Top row: distal view, with anterior to bottom. Middle row, left to right: lateral, anterior and medial views. Bottom row: distal view, with anterior to top. Missing parts reconstructed from the humerus of Giraffatitan brancai (Janensch 1961: Beilage A)

Paul Upchurch recognised the generic distinctness of “Pelorosaurus” becklesii way back in his (1993) dissertation. But because of Cambridge University’s policy of only making copies of dissertations available for £65, that work is effectively unknown. (Perhaps we should all chip in a fiver, buy a copy and “liberate” it. Or maybe 22 years on, Paul would rather leave it in obscurity and let his reputation continue to rest on his impressive body of later work.)

What has happened to this specimen in the last 22 years? Very little has been published about it. It got a mention in the systematic review of sauropods in Dinosauria II (Upchurch et al. 2004), but the only mention that is more than in passing, as far as I’m aware, is that of see Upchurch’s first published (1995) phylogenetic analysis. From page 380:

The only reliable Lower Cretaceous titanosaurid material, apart from Malawisaurus, comes from Europe, especially England. The earliest of these forms may be represented by the forelimb of ‘Pelorosaurus becklesii‘ (Mantell 1852) from the Valanginian of Sussex. This specimen was considered to be Sauropoda incertae sedis by McIntosh (1990b). However, a skin impression shows polygonal plates of a similar shape and size to those found in Saltasaurus (Bonaparte & Powell 1980). The ulna and radius are robust and the ulna bears the typical concavity on its anteromedial proximal process. Upchurch (1993) therefore argued that this form should be provisionally included within the Titanosauridae.

[Update: as Darren points out in the comment below, Upchurch et al. (2011) figure the specimen in colour and devote three pages to it. They leave it as Titanosauria, and “refrain from naming a new taxon until more comparative data are available” (p. 501).]

Given my interest in the Wealden, it’s surprising that we’ve never blogged about “Pelorosaurus” becklesii before, but it’s true: I’ve mentioned it three times in comments, but never in a post. It’s good to finally fix that!

Next time: the radius and ulna.

References

  • Janensch, Werner. 1961. Die Gliedmaszen und Gliedmaszengurtel der Sauropoden der Tendaguru-Schichten. Palaeontographica (Suppl. 7) 3:177-235.
  • Taylor, Michael P. 2007. Diversity of sauropod dinosaurs from the Lower Cretaceous Wealden Supergroup of southern England. p. 23 in Graeme T. Lloyd (ed.), Progressive Palaeontology 2007, Thursday 12th-Saturday 14th April, Department of Earth Sciences, University of Bristol. 38 pp.
  • Upchurch, Paul. 1993. The Anatomy, Phylogeny and Systematics of Sauropod Dinosaurs. Ph.D dissertation, University of Cambridge, UK. 489 pages.
  • Upchurch, Paul. 1995. The evolutionary history of sauropod dinosaurs. Philosophical Transactions of the Royal Society of London Series B, 349:365-390.
  • Upchurch, Paul, Paul M. Barrett and Peter Dodson. 2004. Sauropoda. pp. 259-322 in D. B. Weishampel, P. Dodson and H. Osmólska (eds.), The Dinosauria, 2nd edition. University of California Press, Berkeley and Los Angeles. 861 pages.
  • Upchurch, Paul, Philip D. Mannion and Paul M. Barrett. 2011. Sauropod dinosaurs. pp. 476-525 in: Batten, David J. (ed.), English Wealden Fossils. The Palaeontological Association (London).
Posted by Mike Taylor
Filed in Haestasaurus, humerus, stinkin' appendicular elements, teaser, timely, Wealden
8 Comments »