Every so often, new data gets thrown out there that requires us artists to change how we present the scientist’s work. Recently, I’ve been sharply reminded that even when scientists mean well, some things just get mucked up. One of the best examples in the Zamyn Khond oviraptorosaur, commonly called Citipati sp., which is based on a well-known, but rather obscure specimen, MPC-D 100/42 (formerly, IGM or GI or GIN or GI (SPS) 100/42 — so many names!). Previously, this material has been poorly illustrated, either just by its skull (Barsbold, 1981) or with selected elements shown (Barsbold, 1983; Barsbold et al., 1990; Osmólska et al., 2004), and always illustrated by pen, or just a photo of its skull.
Photographs of the rest of the material have never been published, and photos of the material that do exist tend to be proprietary, so I can’t even show you what I know. But there’s a lot there, missing some of the tail, half of the dorsal vertebrae, some portions of one whole arm and a section of one leg. I did a skeletal reconstruction once, was dissatisfied, did it again, and while I am still dissatisfied with it, have had little opportunity to modify it. This is because, unfortunately, lack of detailed (or any) scalebars of the various elements in the photos produced make it very difficult to show how large a thing is, meaning we are stuck with the scale bars in the figures that are produced, as well as measurements given by the authors in their text, if any.
Of course, not all this is peachy. This doesn’t tell you about important and often distinct allometric differences when scaling certain elements to one another. As Fanti et al. (2012) wrote [pg.7]:
The estimates based on shaft widths suggest that the ratio of radius length to humerus length [in Nemegtomaia barsboldi specimen, MPC-D 107/15] is 0.95, which is close to the same ratios in Citipati osmolskae (MPC-D 100/979), and Conchoraptor gracilis (MPC-D 110/21). In contrast, the ratio is less than 0.80 in Ingenia (MPC-D 100/32, 100/33, 110/03) and Rinchenia (MPC-D 100/32-1).
One cannot expect that the ratios of even closely-related taxa to be perfectly constrained, although it is a great thing to keep in mind when estimating, as it keeps the estimator within a range of certainty; having a given potential range means that estimates may be compared empirically, rather than just saying a measure is “so-so” or “roughly” a given value. Empirical estimation fails when the given measurements and the figures can show widely different results, creating a sense of distrust in figures’ veracity for artists. Matt Wedel and I ran into this problem back in the early 2000′s — as Matt says, when we were young and ignorant (I’m paraphrasing his exact words) — shortly after Titanosaurus colberti (Jain & Bandyopadhyay, 1997) was described [n1], which resulted in a crazy-weird reconstruction from both of us.
This underscores the problem with either estimating from constraining proportions, or taking a few measurements given and a few scale bars, and going from there. Unless the measurements are given, are accurate, and can be verified in some manner (instead of using line or Photoshop drawn bars, use a GSA or SVP or other institutional scalebar strip with your specimen photo, or a microscope-given scale for smaller objects), these are not useful for the reader or artist.
For the Zamyn Khond oviraptorid (which, as I may have noted, I hope would be the only taxon I would name) measurements therefore suck. We got a taste of some measurements from Mickey Mortimer’s Theropod Database, which includes measurements of the skull and femur, while Fanti et al.’s recent paper has increased the given measurements, but shown disparity. You see, direct measurement from the given illustrations above have produced a measurement of 175mm for both myself and Mickey for the skull, done as a line draw from the posterior margin of the distal quadrate to the very ventral tip of the premaxillary “beak,” the BSL (basal skull length). Fanti et al., however, while confirming Mickey’s measurement of 305mm for the femur, give the skull at 180mm. While I would use this as an opportunity to deride the methods by which we measure certain things, and aim for a better future when we are illustrating the axes by which we measure precisely, it shows that there is some doubt to the process indeed for measurements even when attempting to be explicit. But this is largely beside the point.
I mucked up. I erred, was wrong, and must now correct my process. You see, for a while I’ve been touting a skeletal reconstruction as close to accurate as I thought I could get, using photos of the skeleton I had available and using the figures to provide measurements. Well, as you can see below, there are problems with this method:
Scaling down to the animal’s actual skull length of 180mm results in a skeletal 20% smaller; the animal is seemingly distinctly smaller than I had estimated it to be, where otherwise it was one of the largest oviraptorids out there. Of course, not even treating the skull at 175mm will fix this, it just gets smaller, about 1% further from the original size. If I treat the femur as correctly scaled at 305mm, the animal shrinks to 40% smaller than my original estimate! Yowza! I think the reason this looks so different is that the rear and forward sections of the skeleton are incorrectly scaled. The femur is shorter, the tibio-tarsus longer, and the legs otherwise smaller in comparison to the rest of the body, which still needs to shrink. Correctly scaling this so that the head, legs and arms are all at the right scale will require an update, where I will completely re-do the skeleton, fixing what I can. It may take a little time, however. What it will do, of course, if give everyone a much better, clearer idea of the body of this animal, given the few of us who have seen decent photos of the material, or have had the chance to see the material in person (I rely on one who has done both). The only thing better would be to be there in person, with a pair of calipers and a camera, and hopefully a way to remove the individual bones for measurement. Ideally, I would just digitally scan each individual element, where no estimation whatsoever would ever be necessary. That’s not possible, so we’re going to retry this … again. Stay tuned.
What this also means, of course, is that MPC-D 100/42 is significantly smaller than Citipati osmolskae, based on MPC-D 100/978 or 100/979, and specimens like MPC-D 100/1004 suggest larger individuals. This also means that oviraptorids probably never got that much larger than, say, the Triebold oviraptorosaur, and caenagnathids would have been generally larger. But more on that when I get around to fixing the skeletal reconstruction.
[n1] Titanosaurus colberti was later rebranded Isisaurus colberti by Wilson & Upchurch , to separate it from the arguably “highly ambiguous” Titanosaurus species complex and the argued non-diagnostic nature of the Titanosaurus indicus type caudal vertebrae.
Barsbold R. 1981. Беззубйе жыщнйе динозаврй Монголий (Edentulous carnivorous dinosaurs of Mongolia). Трудй – Совместная Совестко-Монгольской Палеотологыческая Зкспедитсия — Joint Soviet-Mongolian Paleontological Expedition, Transactions 15: 28-39, 124. (in Russian, w/ English summary)
Barsbold R. 1983. Жыщнйе динозаврй мела Монголий (Carnivorous dinosaurs from the Cretaceous of Mongolia). Трудй – Совместная Совестко-Монгольской Палеотологыческая Зкспедитсия — Joint Soviet-Mongolian Paleontological Expedition, Transactions 19:1-120. (in Russian, w/ English summary)
Barsbold R., Maryanska, T & Osmólska, H. 1990. Oviraptorosauria. pp.249-258 in Weishampel, Dodson & Osmólska (eds.) The Dinosauria. University of California Press (Berkeley).
Fanti, F., Currie, P. J. & Badamgarav D. 2012. New specimens of Nemegtomaia from the Baruungoyot and Nemegt Formations (Late Cretaceous) of Mongolia. PLoS ONE 7(2):e31330.
Jain, S. L. & Bandyopadhyay, S. 1997. New titanosaurid (Dinosauria: Sauropoda) from the Late Cretaceous of central India. Journal of Vertebrate Paleontology 17(1):114–136.
Osmólska, H., Currie, P. J. & Barsbold R. 2004. Oviraptorosauria. pg.165-183 in Weishampel, Dodson & Osmólska (eds.) The Dinosauria, Second Edition. University of California Press (Berkeley).
Wilson, J. A. & Upchurch, P.2003. A revision of Titanosaurus Lydekker (Dinosauria– Sauropoda), the first dinosaur genus with a ‘Gondwanan’ distribution. Journal of Systematic Palaeontology 1(1):125–160.
This work is licensed under a Creative Commons Attribution-ShareAlike 3.0 Unported License (I own only the final image, and the right half of the third; Barsbold Rinchen owns the first, Karol Sabath the second, and Matt Wedel the left of the third).