New Domes, New Interpretations

Thanks to Bob Sullivan, many of his co-authored publications from the New Mexico Museum of Natural History and Science (NMNHS) Bulletin are now available online (which you can find here). I talked about one of them, naming new oviraptorosaurs Ojoraptorsaurus boerei and Epichirostenotes curriei (Sullivan, Jasinski & van Tomme, 2011). The new papers recently made available include the descriptions of two new pachycephalosaurs, and comments on others. First, a frontoparietal from the Bayn Shire Formation (= “Baynshire Suite(Formation)”) of southern Mongolia is named Amtocephale gobiensis (Watabe, Tsogtbaatar & Sullivan, 2011). Second, a new species of Stegoceras, Stegoceras novomexicanum, is named from several specimens from the Fruitland and Kirtland Formations in the San Juan Basin of New Mexico (USA) (Jasinski & Sullivan, 2011).

Stegoceratine pachycephalosaur ontogeny. A, MPC-D 100/1203, holotype of Amtocephale gobiensis Watabe et al., 2011; B, NHMNS P-33983, holotype of Stegoceras novomexicanum Jasinski & Sullivan, 2011; C, AMNH 5450, holotype of (Ornatotholus browni Wall & Galton, 1979), referred to Steoceras validum (Schott et al., 2011); D, ontogenetic series of Stegoceras validum, after Schott et al., 2011, CMN 515 is the holotype; E, ontogenetic series of Colepiocephale lambei (Sternberg, 1945), after Schott et al. (2009) CMN 8818 is the holotype.

Domes of Mongolia

The easiest paper to read is on the earlier pachycephalosaur, Amtocephale gobiensis[n1], from Watabe Mahito, Tsogtbaatar Khishigjaw and Robert Sullivan (2011), in which a frontoparietal dome of the ostensibly oldest pachycephalosaur yet known has been recovered from southern Mongolia, far to the east of typical dinosaur localities near the Nemegt Valley. The dome is relatively small and dorsally sculptured, a feature notable in other small pachycephalosaur domes. This may become relevant later. It derives from the Bayn Shire Formation[n2], dating roughly to the Santonian or Campanian of the early Late Cretaceous.

The diagnosis for the new taxon (which I will simply call Amtocephale gobiensis, even though the authors use the convention of separating the genus and species in their systematic section) follows thus for the type and only specimen, MPC-D 100/1203 [Watabe et al., 2011, pg.490 (remarks is brackets “[]” are mine)]:

“A small pachycephalosaurid that differs from Prenocephale brevis and all other pachycephalosaurids in having the following combination of characters: [1] deep supratemporal fossae; [2] broad (craniocaudally) prefrontal sutural surfaces with the nasal prefrontal and anterior supraorbitals surfaces lying in a single plane; [3] short parietal (parietal length to the frontoparietal length is 2.44); [4] posterior medial extension of the parietal sharply downturned; [5] posterior medial extension of the parietal wide relative to the maximum width of the frontoparietal; [6] and supratemporal fenestrae absent.”

Several of these features are associated with age of the specimen, including closure of the supratemporal fenestra and depth of some features, as they are associated with expansion of the dome dorsally, as in Stegoceras validum (Schott et al., 2011). Despite this, relative proportions and size indicate one of the smallest pachycephalosaurs with some unique features.

A decent portion of the paper is bent to assessing proportions of the dome and the morphological structure. The rest is devoted to assessing the land vertebrate age correlation of the Bayn Shire Formation and a systematic analysis using cladistics, which is the reason I’m discussing this paper in the first place. However, as a decent amount of effort has been put into assessing the age of the non-marine-corellated Bayn Shire Formation, it is notable that the authors choose the “youngest” age of 83.5Ma (98.5-83.5Ma) for the formation (early Santonian), based on Hicks et al. (1999), whereas a younger date of 80.6Ma (93.5-80.6Ma) was provided by Ogg et al. (2004). The faunal composition is broadly considered similar to the Milk River Formation (dated 84.5-83.5Ma; Payenberg et al., 2002), and thus may be the same age. Ogg et al.’s date would place the formation almost directly beneath the Djadokhta Formation, while that of Hicks et al. would displace the two further apart, or possibly pull the Djadokhta into older sediments. This impacts the relative correlation of the Djadokhta, and I am surprised the authors did not discuss this, but perhaps this is in the works.

Watabe et al. present a variation of the phylogenetic analysis of Sereno (2000), as modified by further authors (e.g., Butler & Sullivan, 2009; Schott et al., 2010). That last analysis is a work that described a new specimen referrable to Colepiocephale lambei (Sullivan, 2003), raised on the basis of a unique loss of the posterior parietosquamosal shelf, between supratemporal fossa. Arguing that the supratemporal fenestra had completely closed, and broad and elongate surpatemproal fossae on the undersurface of the dome indicated large areas of attachment for muscles formerly inserting through the fenestra, Sullivan determined that the parietal was reduced to the small, highly rugose surface that gave Colepiocephale it’s name of “knuckle head”[n3]. Schott et al. (2009) argued that, based on new material including UCMP 130048, the skull is simply less well-preserved than presumed, as in fact resembles other stegoceratine pachycephalosaurs in often being recovered without the parietosquamosal shelf; the rugose surface is an abraded surface reflecting loss of material, and not ornamentation. Schott et al. continue to note that Sullivan (2003), following other authors, failed to note the lack of preservation, rather than the preservation of its lack, of the shelf, in addition to the actual preservation of supratemporal fenestra, which are clearly indicated in UCMP 140048. [Schott et al., 2009:pg.781]:

“The broken bone surface for the posteromedial extension of the parietal can be seen clearly on the posterior and ventral surfaces of the parietal in TMP 92.88.01, TMP 70.02.01, and UCMP 130048, and this region is broken and abraded in the holotype (Figs. 3–7).”

As I noted when discussing Schott et al.’s recent paper (2011) on Stegoceras validum cranial ontogeny, the shape and extent of the dome includes closure of the supratemporal fenestrae, relative “sloping” of the parietosquamosal shelf, relative thickness of the parietal, and a host of other features including the sculpturing of the caudal margin of the cranium in this area. Much of the ontogeny in pachycephalosaurs requires looking at a host of other features, including the roof of the endocranium, neural foramina when present, etc. Additional features set lambei apart from other pachycephalosaurs, including the shape of the rostral lobs of the dome, while proportions of the endocranium and sutures set Amtocephale gobiensis apart, and affirm its diagnostic validity.

Yet More Stegoceras Species

Once, there were many species corralled into Stegoceras. Over the years, they’ve been slowly winnowed out as type species of new “genera,” or synonymized with other taxa in some form or other.

Steve Jasinski & Robert Sullivan (2011) name a new species of Stegoceras, Stegoceras novomexicanum (literally, “New Mexican” or “derivative of New Mexico”). This is based on a partial array of frontoparietal domes and fragments of very young animals in which the sutures are so open they are separate. The holotype is NMMNH P-33983, and is somewhat smaller than the holotype of Stegoceras validum (CMN 515). The doming of the skull is nearly identical to that of the holotype, despite it’s smaller size, suggesting it is developmentally similar at a smaller size, and this may be considered a useful split between the two. [Jasinski & Sullivan, 2011:pg.202 (remarks is brackets “[]” are mine)]:

“Differs from Stegoceras validum […] in possessing the following features: [1] posteromedial extension of the parietal reduced and sub-rectangular; [2] squamosal sutural surface contacts of the posteromedial extension of the parietal roughly parallel; [3] supratemporal fenestrae more medial and enlarged; [4] gracile and small adult size[.]”

Many of these features may be related to age, as the holotype specimen has open supratemporal fenestrae, a partial dome and a large parietosquamosal shelf, an apparent “closed” suture between the left parietal and frontal which hasn’t been obliterated, and distinct unobliterated cranial sutures on the dorsal endocranial surface. Though Jasinski & Sullivan argue that their concern is the taxon is mature, they note that some features of maturity are based not on the holotype but referred specimens which, despite being larger, exhibit similar features. They note that smooth dorsal surfaces, a “capped” layer of bone in the dome, and partial frontal fusion, indicate maturity. Using a referred specimen (a paratype, SMP VP-2555), Jasinksi & Sullivan (2011) perceive four distinct zones of bone, labeled “h1” through “h4,” with three zones of cancellous bone as h1-h3, and the fourth referring to nearly avascular bone.

Sections through pachycephalosaur crania and vascular zones. A, SMP VP-2555, paratype of Stegoceras novomexicanum Jasinksi & Sullivan, 2011; B, ROM 53555, Stegoceras validum after Schott et al., 2011; C, closeup of HCRT section near frontoparietal suture of ROM 53555 showing three vascular zones (Z3 is the most avascular); D, relabeled SMP VP-2555, showing the same zones in C, after Jasinksi & Sullivan, 2011.

Schott et al. (2011) distinguished three zones of vascularity, relating to the proportion of void space to bone in HCRT. The relative proportion of vascularity decreases with age, as the doming increases with larger degrees of avascular bone forming the periphery of the dome. This bone, certainly avascular metaplastic bone (Horner & Goodwin, 2009), is present even in juvenile pachycephalosaurs (Schott et al., 2011), for example AMNH 5450 (holotype of Ornatotholus browni (Wall & Galton, 1974), referred to Stegoceras validum). Its presence cannot be considered an argument for maturity, but rather its thickness. As Schott et al. (2011) show, even half-domed specimens such as CMN 515 and a specimen with similar degree of morphological development of the dome and parietal fenestrae (as in, NMMNH P-33983) possess three distinct degrees of vascularization, but are separable by degrees of fusion and proportion of vascular “void zones.” Comparison of these specimens to the paradigm offered by Schott et al. (2011) may resolve more certainly the age of the specimens, and their potential taxonomic utility. Until then, size alone and degree of development is distinct between species validum and novomexicanum, and this may be enough to affirm their nature.

Phylogeny in Pachycephalosauria Respected

Phylogenetic analyses of Pachycephalosauria (abbreviated). A, Schott et al., 2009; B, Longrich et al., 2010; C, Watabe et al., 2011.

Watabe et al. (2011) do not reassess the analyses of Williamson & Carr (2002) or Sullivan (2003), which are themselves extrapolations from Sereno (1999, 2000). However, Schott et al. (2009) do, and recoding the Colepiocephale lambei specimens in both analyses produce a pair of cladograms in which lambei is placed in a polytomy with Stegoceras validum and the clade leading to (Prenocephale prenes + Tylocephale gilmorei) and that leading to Pachycephalosaurus wyomingensis (in which Sullivan had originally recovered Colepiocephale lambei), alongside Hanssuesia sternbergi. Longrich et al. (2010) represents an anlysis of 20 taxa and 88 characters, nearly twice the number of characters in either Williamson & Carr (2002: 22/55), Sullivan (2003: 17/49) and Watabe et al. (2011: 18/48). In this analysis, Colepiocephale lambei groups with Stegoceras lambei, rather than Prenocephale prenes.

A revised systematic analysis produced by Watabe et al. results in two large changes, as noted in the figures. Firstly, Amtocephale gobiensis is within a clade formed by all species of Prenocephale as argued by Sullivan (2003, 2006), along with Tylocephale gilmorei (although that latter position has always been there). Secondly, Prenocephale prenes is the sister taxon to all other taxa included in this grouping (brevis, edmontonense, goodwini, gilmorei and gobiensis), which renders the clade Prenocephale sensu Sullivan (2003) polyphyletic. Thirdly, Colepiocephale lambei is placed alongside the Pachycephalosaurus side of the tree, rather than the Prenocephale side. This is achieved by Watabe et al. using Sereno (2000), removing three characters (relating to preacetabular ala length of the ilium, frontoparietal thickness, and doming extent, the latter being based on supposed similarities with pachycephalosaurines); eleven characters are added from Sullivan (2003: n=9), themelves (n=1), and another paper I’ll get to in a moment (n=1). Watabe et al. specifically chose to limit their rescoring in the case of Colepiocephale lambei (Schott et al., 2009, who recoded 14 characters) based on reservations, but based on character preservation, only one state based on a known was accepted, “34. Parietal-squamosal position relative to occiput: dorsal (0); posterodorsal (1)”. Because this state is related to the relative extent of a possible parietosquamosal shelf, and that Watabe et al. (2011) describe the posteroventral slope of the parietal in accord with that Prenocephale prenes and Amtocephale gobiensis (pg.483), I can only assume that the authors reject the conclusion of Schott et al. (2009) in reconstructing a parietosquamosal shelf which is simply missing in the specimen.

Codings for Colepiocephale lambei in various analyses. Blue represents codings removed from Sullivan (2003) for Watabe et al. (2011); Green represents coding differences between Sullivan (2003) and Schott et al. (2009); Yellow represents codings added by Watabe et al. (2011).

[Ideally, above, I would have also included Williamson & Carr (2002) and further analysis in differences with Longrich et al. (2010). This will happen farther down the road.]

I thus think it is unlikely that, given the preservation involved, that the skull of MPC-D 100/1203, holotype of Amtocephale gobiensis Watabe et al. (2011),  lacks a parietosquamosal shelf naturally or that the supratemporal fenestrae have closed. As the latter is an ontogenetic character, and the relative closure of fenestra occurs well into cranial ontogeny and after obliteration of cranial sutures, the skull may likely be that of a subadult animal, and that further larger, and older specimens will be recovered that support this hypothesis, and include a parietosquamosal shelf.

Similarly, as in Longrich et al. (2010), I fear it is untenable to continue to suppress usage of Sphaerotholus as a valid taxon when it can be supportive of phylogenetic relationships. Not only is it supportive of the characters of goodwini, edmontonense (including buccholzae Williamson & Carr, 2002) and breve which Sullivan supports, several analyses raise doubt as to the consistency of a clade that includes them and Prenocephale prenes, but excludes other “genera.” The argument that subsuming the genus can support biostratigraphy is irrelevant when it is possible that the species themselves may be useful in this regard. That Tylocephale gilmorei and apparently Amtocephale gobiensis can fall into this complex (Watabe et al., 2011), that Prenocephale prenes may align with a strictly Asian clade (Longrich et al., 2010), or that the species may also be closer to Pachycephalosaurus wyomingensis and polytomize Sullivan’s use of Prenocephale doubly affirms that continued reference of these species as “Prenocephale [species]” is simply not useful. Thus I argue that Sphaerotholus should be used for goodwini (type species) over that of Prenocephale as their container (or “genus,” if you will). The allocation of breve and edmontonense can revert back to Stego0ceras, although the bulk of data appears that species breve, bucholtzae and edmontonense (to which the second taxon was referred by Sullivan, 2003) can be contained monophyletically within Sphaerotholus. This is even more considerate of potential new information that may affirm more if not each species be usefully separated from historical Stegoceras as a new genus-species couplet.

[n1] “Amto” follows the convention for the Bayn Shire ankylosaur, Amtosaurus, in deriving from Amtay, an older transcription for Amtgai (Amtay, Amtgay). This convention doesn’t follow Benton (2000), where the derivative should be “amtgo-“.
[n2] Similarly to [n1], “Baynshire” (n=3) (often also Bayn Shiree, Baynshireh, Bayanshiree, etc.) and “Bayn Shireh” (n=5) has been used, but Bayn Shire is prefered in Benton (2000). The authors thus inconsistently use different nomenclature for a stable name, “Bayn Shire,” although omitting the space and capital in the first set of usages can be considered “equivalent.” It might be favorable to think that the nomenclature was being standardized in the manuscript, but several instances were missed, and that the name in the abstract and title was the intended spelling (as it would be the most significant to the casual reader). If so, they can be excused an editorial oversight. It should be further noted that the term “Baynshirenskaya Svita” is not equivalent to “Bayn Shire Formation.” [Benton, 2000:xviii]:

“Svitas, on the other hand, are largely lithostratigraphic unites (Art.VII.5), given a locality name that is close to their characteristic exposure. The definition of a svita incorporates a mix of field lithological observations and biostratigraphic assumpions: ‘In distinguishing a new svita, one ought without fail to establish at least an approximate, sufficiently proves correlation of it with subdivisions of the unified [international] scale’ (Interdepartmental Statrigraphic Committee, 1959, p.34). The requirement for ‘approximate correlation’ has been deleted in the 1992 code, and the emphasis is on mappable lithological features[.] […] The new understanding of the ‘Svita’, according to the 1992 code, allows for a closer equation with the ‘Formation’, since the emphasis is on local lithological and mapping criteria. […] For the present, though, until they are explicitly revised and redefined, classical svitas cannot be assumed to be purely lithostratigraphic units.”

In otherwords, use of the term “svita” does not simply get replaced with “formation” as if they were the same thing. An additional issue lies in the use of the convention for indicating a svita, whereby the name is appelled with the Russian genetive -skaya, followed by Svita: Baruungoyotskaya Svita. In this case, this is part of the linguistic convention, and exists alongside the Barungoyot Formation, not in lieu of it. Simply replacing Svita here with Formation does not and should not result in Baruungoyotskaya Formation.
[n3] The “knuckle” of Greek χωλέπο– (kolepo-), derived from χωλός (kolos, halt, as in “limping gait”) and carries a sense of “abrupt abbreviation.”

Benton, M. J. 2000. Conventions in Russian and Mongolian palaeontological literature. pp.xvi-xxxii in Benton, Shishkin, Unwin & Kurochkin (eds.) The Age of Dinosaurs in Russia and Mongolia. Cambridge University Press (Cambridge, United Kingdom).
Butler, R. J. & Sullivan, R. M. 2009. The phylogenetic position of the ornithischian dinosaur Stenopelix valdensis from the Lower Cretaceous of Germany and the early fossil record of Pachycephalosauria. Acta Palaeontologica Polonica 54(1):21-34.
Hicks, J. F., Brinkman, D. L., Nichols, D. J. & Watabe M. 1999. Paleomagnetic and palynologic analyses of Albian to Santonian strata at Bayn Shireh, Burkhant, and Khuren Dukh, eastern Gobi Desert, Mongolia. Cretaceous Research 20:829-850.
Horner, J. R. & Goodwin, M. B. 2009. Extreme cranial ontogeny in the Upper Cretaceous dinosaur Pachycephalosaurus. PLoS ONE 4(10): e7626.
Jasinksi, S. E. & Sullivan, R. M. 2011. Re-evaluation of pachycephalosaurids from the Fruitland-Kirtland Transition (Kirtlandian, Late Campanian), San Juan Basin, New Mexico, with a description of a new species of Stegoceras and a reassessment of Texacephale langstoni. in Sullivan et al. (eds.) The Fossil Record, volume 3. New Mexico Museum of Natural History and Science, Bulletin 53:202-215.
Longrich, N. R., Sankey, J. & Tanke, D. 2010. Texacephale langstoni, a new genus of pachycephalosaurid (Dinosauria: Ornithischia) from the upper Campanian Aguja Formation, southern Texas, USA. Cretaceous Research 31:274–284.
Ogg, J. G., Agterberg, F. P. & Gradstein, F. M. 2004. The Cretaceous Period. pp.344-383 in Gradstein, Ogg, Smith & Ogg (eds.) A Geologic Time Scale. Cambridge University Press (Cambridge, United Kingdom).
Payenberg, T. H. D., Braman, D. R., Davis, D. W. & Miall, A. D. 2002. Litho- and chronstratigraphic relationships of the Santonian-Campanian Milk River Formation in southern Alberta and Eagle Formation in Montana utilizing stratigraphy, U-Pb geochronology, and palynology. Canadian Journal of Earth Science — Revue canadienne de sciences de la Terre 39:1553-1577.
Schott, R. K., Evans, D. c., Goodwin, M. B., Horner, J. R., Brown, M. & Longrich, N. R. 2011. Cranial ontogeny in Stegoceras validum (Dinosauria: Pachycephalosauria): A quantitative model of pachycephalosaur dome growth and variation. PLoS ONE 6(6):e21092.
Schott, R. K., Evans, D. C., Williamson, T. E., Carr, T. D. & Goodwin, M. B. 2009. The anatomy and systematics of Colepiocephale lambei (Dinosauria: Pachycephalosauridae). Journal of Vertebrate Paleontology 29(4):771-786.
Sereno, P. C. 1999. The evolution of dinosaurs. Science 284:2137–2147.
Sereno, P. C. 2000. The fossil record, systematics and evolution of pachycephalosaurs and ceraptosians from Asia. pp.480-516 in Benton, Shishkin, Unwin & Kurochkin (eds.) The Age of Dinosaurs in Russia and Mongolia. Cambridge University Press (Cambridge, United Kingdom).
Sternberg, C. M. 1945. Pachycephalosauridae proposed for dome-headed dinosaurs, Stegoceras lambei, n. sp., described. Journal of Paleontology 19(3): 534–538.
Sullivan, R. M. 2003. Revision of the dinosaur Stegoceras Lambe (Ornithischia, Pachycephalosauridae). Journal of Vertebrate Paleontology 23(1):181-207.
Sullivan, R. M. 2006. A taxonomic review of the Pachycephalosauridae (Dinosauria: Ornithischia). New Mexico Museum of Natural History and Science, Bulletin 35:347-365.
Wall, W. P. & Galton, P. M. 1979. Notes on pachycephalosaurid dinosaurs (Reptilia: Ornithischia) from North America, with comments on their status as ornithopods. Canadian Journal of Earth Science — Revue canadienne de science de la Terre 16:1176–1186.
Watabe M., Tsogtbaatar K. & Sullivan, R. M. 2011. A new pachycephalosaurid from the Baynshire Formation (Cenomanian-Late Santonian), Gobi Desert, Mongolia. in Sullivan et al. (eds.) The Fossil Record, volume 3. New Mexico Museum of Natural History and Science, Bulletin 53:489-497.
Williamson, T. E. & Carr, T. D. 2002. A new genus of derived pachycephalosaurian from western North America. Journal of Vertebrate Paleontology 22(4):779-801.