Monday 21 March 2022

Torosaurus in Canada.

The Ceratopsids, an iconic group of Dinosaurs which were an important part of the faunas of Late Cretaceous North America and Asia, were large, quapruped herbivores with distinctive neck frills and facial horns over their eyes and noses. One of the largest Ceratopsids was Torosaurus, known from the End Cretaceous of North America, which is thought to have weighed up to 6500 kg, and achieved skull lengths of about 3 m. However, Torosaurus, is similar to Triceratops, another large End Cretaceous Ceratopsid, which has led to an active debate in recent years, as to whether the two genera are in fact the same, and, therefore, whether or not Torosaurus is a valid name at all (in taxonimy, where a group has been named twice, the older name is considered to have priority). This argument derives from the fact that all known specimens of Torosaurus are as large as, or larger than, the largest specimens of Triceratops, while the horns and crest of Torosaurus are essentially a more exagerated variant in the pattern seen in Triceratops, which might imply that Torosaurus fossils represent more mature specimens of Triceratops, rather than a separate species.

To date, there have been two reports of Torosauraus specimens being found in Canada, both of fragments of frills that have not fully been described. The first, EM P16.1, was recovered from the Frenchman Formation of southern Saskatchewan in 1947 by Charles Morey, upon whose land it was found, and local amateur palaeontologist Harold ‘Corky’ Jones. This specimen was briefly described and illustrated by palaeontologist Tim Tokaryk in 1986, who attributed the specimen to Torosauraus, but this diagnosis has since been disputed, and it has been suggested that the specimen might in fact belong to Arrhinoceratops, which can be hard to differentiate from Torosaurus. The second specimen, UALVP 1646, was excavated from the lower Scollard Formation in the Red Deer River Valley of southern Alberta by Charles Stelck in 1964, and has been mentioned as an example of Torosaurus in a number of publications, but never formally described or illustrated.

In a paper published in the Zoological Journal of the Linnean Society on 1 March 2022, Jordan Mallon of the Beaty Centre for Species Discovery and Palaeobiology Section at the Canadian Museum of Nature, and the Ottawa-Carleton Geoscience Centre and Department of Earth Sciences at Carleton University, Robert Holmes, also of the Beaty Centre for Species Discovery and Palaeobiology Section at the Canadian Museum of Nature, and of the Department of Biological Sciences at the University of Alberta, Emily Bamforth of the T. rex Discovery Centre at the Royal Saskatchewan Museum, and Dirk Shuman of Fibics Incorporated, formally describe specimens EM P16.1 and UALVP 1646, and discuss their attribution to Torosaurus and the implications of this to the ongoing Torosaurus/Triceratops debate.

The site where specimen EM P16.1 was collected is no longer accessible, but is believed to be an outcrop of the upper-half of Frenchman Formation along the Frenchman River, making it uppermost Maastrichtian in origin (i.e. from the very end of the Cretaceous). Shortly after it's discovery, Harold Jones wrote a letter to Charles Sternberg of the Geological Survey of Canada in which he described, in addition to the crest fragment, the associated discovery of two large bones whicj he believed to be the Animal's femurs, which were in good condition and each measured 46" in length and 9" in diameter at the thinnest part (117 cm long and 23 cm in diameter). The subsequent fate of these bones has not been recorded, however, the Eastend Historical Museum, which houses EM P16.1, also hosts some partial, poorly preserved limb material bearing the field number B9, which is also the field number assigned to EM P16.1, suggesting that these are the femurs mentioned by Jones. Much of Jones' collection is known to have been damaged in 1952, when the basement of an old schoolhouse in which it was housed was flooded, so the disparity in condition between Jones' description and that of the Eastend Historical Museum material is not surprising.

Femoral material pertaining to Torosaurus cf. Torosaurus latus (EM P16.1). (A) Proximal portion of right femur, medial view; (B) distal portion of left femur, cranial view; (C) proximal portion of left femur, cranial view; (D) proximal diaphysis of left femur, transverse view; (E) proximal portion of left femur, caudal view. Red arrows indicate where left femur was sampled for osteohistology. Mallon et al. (2022).

The second specimen, UALVP 1646, was uncovered in a quarry at a quarry in the Red Deer River Valley in 1964, with its current assignment to the lower Scollard Formation being based upon a description that it was found approximately 8 m below the Nevis coal seam.

Locality information pertinent to Canadian Torosaurus specimens. (A) Map showing locations of EM P16.1 and UALVP 1646 (stars). (B) stratigraphic context of EM P16.1 and UALVP 1646, with distribution of Hell Creek Formation Torosaurus for context. Note that the precise placement of EM P16.1 within the Frenchman Formation is unknown. The precise location of MPM VP6841 within the upper Hell Creek Formation is likewise uncertain. Abbreviations: Fm, Formation; K-Pg, Cretaceous-Palaeogene boundary; Mtn, Mountain. Mallon et al. (2022).

Mallon et al. identified the three remaining fragments of limb bone associated with specimen EM P16.1 as the proximal portion of the right femur, the distal end of the left femur missing the condyles, and the proximal metaphysis and diaphysis of the left femur (excluding the fourth trochanter, which is missing). This third piece appeared to be the least damaged, although it shows signs of having been hastily repaired in places with plaster and shellac, and from this specimen Mallon et al. took a sample from the caudolateral margin of the metaphysis and caudal left femoral diaphysis for osteohistological sampling.

The frill of specimen EM P16.1 was reconstructed with plaster and clay shortly after being extracted, and previous descriptions of the specimen have been based upon that reconstruction, even though it was known to be inaccurate. The specimen was re-prepared by experts from the Royal Saskatchewan Museum in the early 2000s, and Mallon et al. present a new description of the specimen based upon that preparation.

The frill of EM P16.1 is mostly complete, although the left squamosal is heavily reconstructed with plaster, and the median parietal bar is missing entirely. The frill is of moderate dimensions, being somewhat larger than that of Arrhinoceratops brachyops, but smaller than the largest Torosaurus frills, particularly in the transverse dimensions. The frill is nearly square in outline, being just 1.1 times wider than long, as in the holotype of Arrhinoceratops brachyops. The frill is notably more triangular in the holotype of Torosaurus latus (YPM 1830, excavated from the Lance Formation of southern Wyoming by John Bell Hatcher in 1891), which is 1.79 times wider caudally than rostrally (although the parietal of this specimen is heavily reconstructed). The width of the transverse parietal bar of EM P16.1 is among the smallest known for Torosaurus, and most nearly approximates that of ANSP 15192 (which is 1225 mm). The frill is flat in the transverse plane, not saddle-shaped as in Triceratops, although this flattening may be exaggerated by taphonomic compression.

Parietosquamosal frill of Torosaurus cf. Torosaurus latus (EM P16.1.). (A) Dorsal surface; (B) interpretive reconstruction of dorsal surface; (C) schematic transverse cross-section of squamosal near caudal edge, based on sketch by Harold Jones; (D) schematic of frill, showing paired fossae and channels on ventral surface of squamosals, based on sketch by H. Jones. Abbreviations: p, parietal; pf, parietal fenestra; sq, squamosal; sqaf, accessory fenestra of squamosal; sqb, squamosal bar. Mallon et al. (2022).

Specimen UALVP 1646 consists of a partial parietal, including portions of the median and caudal bars. As preserved, the maximum dimensions of the specimen are 513 mm rostrocaudally by 568 mm transversely. The parietal is weakly bowed dorsally, indicating that the frill was originally saddle-shaped, albeit not to the extent seen in Triceratops. The partial midline parietal bar is thickened (20 mm dorsoventrally) and bears three sagittally aligned bumps dorsally, similar to those reported in other Torosaurus. The anterior-most of these is most prominent; the remaining two become progressively broader and lower toward the caudal margin of the parietal. Portions of the finished margin of the left parietal fenestra are preserved. The fenestra seems to have been small and probably had a maximum diameter of not much more than 15 cm. The bone surrounding the fenestra is thin (8 mm), thickening outwardly in all directions. The caudal margin of the parietal is 21–25 mm thick (compared to 10–19 mm thick in YPM 1831) and lacks a median emargination. The parietal bears three epiparietals co-ossified to the frill. The interstitial sutures are visible. These epiparietals are interpreted as the left and right P1s and the left P2. There is no epiossification straddling the midline.

Parietal of Torosaurus cf. Torosaurus latus (UALVP 1646). (A) Dorsal surface; (B) interpretive reconstruction of dorsal surface; (C) ventral surface; (D) interpretive reconstruction ventral surface; (E) detail of left parietal fenestra. Arrows in (B) and (D) indicate epiparietal loci. Arrows in (E) indicate unbroken margin of parietal fenestra. Abbreviations: mb, median bump; pf, parietal fenestra; pfo, parietal fossa. Mallon et al. (2022).

Triceratops is known to have ranged into Canada, but the question of whether-or-not the closely related Torosaurus also did so has until now remained open; two specimens found in Canada have previously been referred to Torosaurus, but only one of these has previously been (briefly) described, and both have had their reference to the genus questioned. Mallon et al.'s redescription of EM P16.1 rules out the exclusion of that specimen from Torosaurus, on the basis that all of the features used to reject that hypothesis were ether artefacts of the way in which the specimen had been preserved, or were also present in the holotype of Torosaurus, if not in every specimen attributed to that taxon. Specimen UALVP 1646 is less complete than specimen EM P16.1, but again has no features which preclude its inclusion within the genus Torosaurus. Mallon et al. note that specimens assigned to Torodsaurus do seem to be somewhat variable in some features, and note that Triceratops, which is closely related, is considered to have undergone some species turnover during the 200 000 years represented by the Hell Creek Formation, and it would not be unreasonable to expect a similar level of turnover in Torosaurus.

Mallon et al. also note that the frill of specimen EM P16.1 is considerably smaller than seen in many other Torosaurus specimens, although the associated femora were of comparable size to the largest known Triceratops specimens. They note that size is not always a reliable indicator of maturity in Dinosaurs, but that EM P16.1 shows a number of morphological and osteological features which lead them to conclude that it was not a mature specimen. The existence of a sub-adult Torosaurus specimen of comparable size to the largest Triceratops specimens leads them to conclude that it is unlikely that Triceratops represents an immature form of Torosaurus.

See also...

Online courses in Palaeontology. 

Follow Sciency Thoughts on Facebook.

Follow Sciency Thoughts on Twitter.