Earlier this year, Andy Farke took the opportunity of a remodel to assess the skull of that classic of classic dinosaurs, Nedoceratops hatcheri. Formally named by Richard Swan Lull (completing a monograph that first OC Marsh had begun but uncompleted by hid death, and then resumed by John Bell Hatcher until his death), he presumed it may have been a pathological specimen, probably Triceratops (one of its species, many of which abounded at the time), owning to large irregular rather than concise fenestrae in the parietals and one of the squamosals; but he relented on the idea of its apparently unique features and coined the name Diceratops hatcheri. The name would be later found to be ironic.
(We’ll leave aside the issue of Diceratops becoming Diceratus (Mateus, 2008) as well as Nedoceratops (Ukrainsky, 2007), with the later being named, funnily enough, by an entomologist (see Ukrainsky, 2009 and Farke, 2011 for discussion).)
As I discussed in part here, it later became apparent that the specimen was particularly similar to what is now called Triceratops (again, to one of its species) and to Torosaurus (to one of its species), but researchers disagreed as to its affinities. It labored in chaos at least until the advent of phylogenetic analysis was brought to bear on ceratopsians, and Forster (1996) found it to come out into a position immediately between the two Maastrichtian ceratopsids. This made three Maastrichtian “genera” of ceratopsid, and a plethora of species sunk, leaving only prorsus and horridus in Triceratops, hatcheri in Diceratops, and latus and gladius in Torosaurus. What was a horrible complex and profusion had been drastically simplified. This would be advanced much, much later by analyses of growth and variation. Scannella & Horner (2010) disagreed with the general premises offered by past arguments and asserted that growth could account for the differences between all specimens, including all putative apomorphies. This would knock the species count down somewhat, although it is not clear to what as the authors only provided argument for “generic” synonymy, leaving the question of species synonymy to a later date (as one might imagine, I threw something of a fit about this). While humorous and exasperatingly marred by the public outrage of “We’re losing Triceratops?!” the paleo community awaited discussion of these acts and further elaboration. For example, Scannella & Horner merely asserted that specimens called Torosaurus belong to specimens called Triceratops, but that the species complexes were yet unresolved; it would be necessary to establish the species assignment of the various holotypes, as one cannot simply assert a species type is not longer typic and only a genus can be synonymized.
Here’s a secret: I think Scannella & Horner have it right — to a point [n2]. There does appear to be a single consistent lineage of ceratopsid, and its name is probably Triceratops. The issue here, as it has always been, is that the complex of species and the argument that there are in fact distinct species within Triceratops (averred by Forster, 1996, and supported by everyone since) is made in connection to this statement, but they are not mutual. One can have a clade called Triceratops and a whole mess of viable species. This occurs today among terrestrial herbivores which are closely related, especially savannah antelopes in Africa, or “subspecies” of wolves, “species” or complexes of such of varanid lizard (almost all of them called Varanus, but some whom refer to them by their “subgenera,” names that could under any other circumstance be “genera” instead), and this is just talking about well known megafaunal cases. My take on this is to treat Triceratops as a clade, and as a entity no different than Ceratopsidae. It would be stupid, if not unintelligible, for me to refer Torosaurus to be a junior subjective synonym of Ceratopsidae, in the sense the authors have done to it for Triceratops, but that is what they’ve effectively done. The problem here, then is that the authors aver some external validity to the nature of the “genus,” and the “family;” they argue as though the ranks Genus and Familia are real, viable objects. This, for many, many reasons (many laid out here and somewhat tangentially here, where I approach the ambiguous and essentially meaningless use and practice of the term “nomen dubium“) just doesn’t jive.
So, Torosaurus is Triceratops. Phylogenentically, it seems that prorsus and horridus form a complex regardless of hatcheri, and that latus is outside this, serially. This has been somewhat of a problem because generally, juveniles of taxa should tend to clade together, and they tend to share more features due simply to lacking adult characteristics to differentiate them. This works when diagnostic features tend to show up only during adulthood, or late adulthood, as Scannella & Horner argue. In some cases, it can show up earlier in ontogeny, but the authors aver that while they can segregate younger adults stages of Maastrichtian ceratopsid into two species, they choose to emplace these morphologies into a single broader category, and relegate all other potential taxa to the broader category, preserving the lesser (species). This makes things difficult again, and where I’ve had discussions with no resolution on the outcome with several of the principles in this “debate.” The issue is merely waiting for future work. This is not aided by the fact that the discussion of these taxa is in the “generic” (pardon the pun) rather than in the substantive or “specific”: Scannella & Horner eschew use of the species names prorsus or horridus in favor of Triceratops at nearly every opportunity (“prorsus” appears in the 2011 paper twice, “horridus” five times; in the 2010 paper, this is two and two, respectively, a ridiculously low count when you are talking about synonymizing taxa).
Perhaps further work, or further whining from me, might compel some sense that when you synonymize taxa, or aver they are the same, one does so in concern to the species, not the “genus.” This is especially important as increasingly new diverse taxa show up which show bizarre or divergent adult morphology, juveniles or subadults are being granted new “generic” status without quantifying ontogenetic relevance of the characters, and so forth.
Farke (2011) attempted to resolve the complex problem offered by Scannella & Horner (2010), only further complicated by the addition of a fourth Maastrichtian ceratopsid “genus” to the “Triceratopsini” complex, Eotriceratops xerinsularis (Wu et al., 2007). The taxa multiplied before our eyes. This was not counting the eventual naming of a fifth (Ojoceratops fowleri Sullivan & Lucas, 2011 from the Ojo Alamo of Formation of New Mexico (roughly equivalent with the late Maastrichtian) and a sixth (Tatankaceratops sacrisonorum Ott & Larson, 2011 from the Hell Creek of the Dakotas) — all of which I note here. While it’s been noted that the limited material (the holotype is a squamosal) of the former, bizarrely similar to the squamosal of Nedoceratops hatcheri in some respects, is geographically sundered from the Hell Creek and equivalent up north, biogeographic mapping places taxa that occur in both, including Tyrannosaurus rex in the north and south, and Alamosaurus sanjuanensis, which occurs in equivalent late Maastrichtian strata in Utah as well as New Mexico and Texas. Scannella & Horner (2011) have fired back a salvo in a sharp, precise paper, one that clearly argues that each of Farke’s reiterated diagnostic features (following some of Forster’s 1996 work) are represented as variation among the complex of specimens referred to Triceratops (as different species).
I’m beginning to suspect that much of this indicates that the faunas of the late Maastrichtian are either strongly latitudinally segregated, or homogenous, and changing at the same rate stratigraphically. This is the only way to so far explain the effect of keeping some disparate morphologies distinct, and others together. This is not to say I think the current model is wrong, or that it is right, and I will say up front I actually have no opinion. I think particular biogeographic biases may be interfering with a solution to the problem. It may be possible that all of these late Maastrichtian taxa are synonyms of some level below “Triceratopsini.” If so, it makes a distinct “Triceratopsini” clade containing them redundant, and these Triceratops-like taxa are Triceratops. This would render variation limited, individual, and pathologic, rather than specific (or “generic”). A major influence on this discussion has been the radical ontogenetic change Scannella & Horner (2010) argued for. It seems particularly strange to me that the authors should argue that this change should occur in such a limited respect, confined to merely a few species of a single “genus,” or that related but dissimilar taxa would not also show this same change. If this ontogenetic transitional series were to be plausible for a diverse range of species, yet all generally related to one another, and where one could not reasonably tell the juveniles of two closely related by distinct species apart (as was the case for a long time without the aid of biogeographic and stratigraphic models in hadrosaurs), then it would be impossible to tell three, or even four apart. It may instead be more simplistic to afford all of the taxa a single species, and ignore the whole “species referral” mess to which I bemoan, or that there may ever be distinct “genera,” but then this merely underscores the very problem of “genera,” and the utterly subjective and unscientific concept thereof.
Scannella & Horner (2011) argue (pg. 7):
Until a clearly juvenile ‘Torosaurus’ is recovered – with backwards curving postorbital horn cores, delta-shaped frill epiossifications, elongate squamosals, and a fenestrated parietal – it appears more likely that either: a) juvenile ‘Torosaurus’ were largely indistinguishable from Triceratops and differences in morphology between these two taxa only became apparent later in ontogeny, or b) ‘Torosaurus’ and Triceratops are synonymous. We favor the latter hypothesis for reasons previously discussed in detail in [Scannella & Horner (2010)].
Personally, I do not think it wise to modify the apparent taxonomy of these ceratopsians, call other arguments into question, then claim that the burden of proof is on them to prove you wrong, rather than attempting to demonstrate beyond the possibility of (a) above (the status quo) that those who’ve argued for current model should be held until proven directly wrong. As I said at the beginning, I think Scannella & Horner have it right, but this is in the generic (again, pardon the pun). I would have preferred that the authors determined a test to their argument by showing that Torosaurus latus or Nedoceratops hatcheri cannot be specifically or “generically” distinct from any category with Triceratops. And this is simply because I think all Scannella & Horner have done so far is examined a substantial heterochronous ontogenetic trend in a clade of chasmosaurine ceratopsians. I also think they put the desire to revise the taxonomy (the cart) before they determined the systematic implications of their ontogenetic scheme (the horse). This is simply backwards. I wonder how Scannella & Horner would consider the option of renaming Triceratops prorus something like “Eutriceratops” prorsus. I have just as much a reason to split the species into “genera” as to lump disparate species into a single genus for no sake than similar ontogeny. If biogeography or biostratigraphy is a viable option for distinguishing species, then may I use a formational boundary to define taxa? if there is none, but there is morphological variation, could i synonymize? The subjectivity abounds, and that is the problem with this approach.
[n1] While we’re on the way to confessions, I think Giraffatitan is a perfectly good container for Brachiosaurus brancai, but I’m not quite sure the reasoning is sound enough to accept it over, say, any other generic renaming of any species.
Farke, A. A. 2011. Anatomy and taxonomic status of the chasmosaurine ceratopsid Nedoceratops hatcheri from the Upper Cretaceous Lance Formation of Wyoming, U. S. A. PLoS ONE 6(1):e16196.
Forster, C. A. 1996. Species resolution in Triceratops: cladistic and morphometric approaches. Journal of Vertebrate Paleontology 16:259–270.
Hatcher, J. B., Marsh, O. C., Lull, R. S. 1907. The Ceratopsia. United States Geological Survey, Monographs 49:1–300.
Mateus, O. 2008. Two ornithischian dinosaurs renamed: Microceratops Bohlin, 1953 and Diceratops Lull, 1905. Journal of Paleontology 82(3):423.
Ott, C. J. & Larson, P. L. 2010. A new, small ceratopsian dinosaur from the Latest Cretaceous Hell Creek Formation, northwest South Dakota, United States: A preliminary description. p203-219 in Ryan, Chinnery-Allgeier & Eberth (eds.) New Perspectives on Horned Dinosaurs: the Royal Tyrrell Museum Ceratopsian Symposium. (Indiana University Press, Bloomington.)
Scannella, J. B. & Horner, J. H. 2010. Torosaurus Marsh, 1891, is Triceratops Marsh, 1889 (Ceratopsidae: Chasmosaurinae): Synonymy through ontogeny. Journal of Vertebrate Paleontology 30(4):1157–1168.
Scannella, J. B. & Horner, J. H. 2011. ‘Nedoceratops‘: An example of a transitional morphology. PLoS ONE 6(12):e28705.
Sullivan, R. M. & Lucas, S. G. 2010. A new chasmosaurine (Ceratopsidae, Dinosauria) from the Upper Cretaceous Ojo Alamo Formation (Naashoibito Member), San Juan Basin, New Mexico. p169-180 in Ryan, Chinnery-Allgeier & Eberth (eds.) New Perspectives on Horned Dinosaurs: the Royal Tyrrell Museum Ceratopsian Symposium. (Indiana University Press, Bloomington.)
Ukrainsky, A. S. 2007. A new replacement name for Diceratops Lull, 1905 (Reptilia: Ornithischia: Ceratopsidae). Zoosystematica Rossica 16(2):292.
Ukrainsky, A. S. 2009. Synonymy of the genera Nedoceratops Ukrainsky, 2007 and Diceratus Mateus, 2008 (Reptilia: Ornithischia: Ceratopidae). Paleontological Journal 43(1):116.
Wu, X-C., Brinkman, D. B., Eberth, D. A. & Braman, D. R. 2007. A new ceratopsid dinosaur (Ornithischia) from the uppermost Horseshoe Canyon Formation (upper Maastrichtian), Alberta, Canada. Canadian Journal of Earth Science — Revue canadienne de Sciences de la Terre 44(9):1243-1265.