Journal of Vertebrate Paleontology

ISSN: 0272-4634 (Print) 1937-2809 (Online) Journal homepage: http://www.tandfonline.com/loi/ujvp20

Torosaurus Marsh, 1891, is Triceratops Marsh,

1889 (Ceratopsidae: Chasmosaurinae): synonymy
through ontogeny

John B. Scannella & John R. Horner

To cite this article: John B. Scannella & John R. Horner (2010) Torosaurus Marsh, 1891, is
Triceratops Marsh, 1889 (Ceratopsidae: Chasmosaurinae): synonymy through ontogeny,
Journal of Vertebrate Paleontology, 30:4, 1157-1168, DOI: 10.1080/02724634.2010.483632

To link to this article: http://dx.doi.org/10.1080/02724634.2010.483632

Published online: 13 Jul 2010.

Submit your article to this journal

Article views: 560

View related articles

Citing articles: 36 View citing articles

Full Terms & Conditions of access and use can be found at

http://www.tandfonline.com/action/journalInformation?journalCode=ujvp20

Download by: [Society of Vertebrate Paleontology ] Date: 22 August 2016, At: 12:03
Journal of Vertebrate Paleontology 30(4):1157–1168, July 2010
© 2010 by the Society of Vertebrate Paleontology

ARTICLE

TOROSAURUS MARSH, 1891, IS TRICERATOPS MARSH, 1889 (CERATOPSIDAE:

CHASMOSAURINAE): SYNONYMY THROUGH ONTOGENY

JOHN B. SCANNELLA* and JOHN R. HORNER

Museum of the Rockies, Montana State University, 600 West Kagy Boulevard, Bozeman, Montana 59717, U.S.A.,
[email protected]

ABSTRACT—Although they have been considered distinct genera for over a century, ontogenetic analyses reveal
that Triceratops and “Torosaurus” actually represent growth stages of a single genus. Major changes in cranial
morphology—including the opening of parietal fenestrae and the elongation of the squamosals—occur rapidly, very late in
Triceratops ontogeny and result in the characteristic ‘Torosaurus’ morphology. This report presents the results of a 10-year
field study of the dinosaurs of the Hell Creek Formation in Montana and is based on a collection of over 50 specimens of
Triceratops, including over 30 skulls, which have been amassed in that time, in addition to specimens from numerous other
North American museums. This large sample of individuals reveals the full ontogenetic spectrum of Triceratops. The syn-
onymy of Triceratops and ‘Torosaurus’ contributes to an unfolding view of extremely reduced dinosaur diversity just before
the end of the Mesozoic Era.

INTRODUCTION UND, Paleontology Collection, University of North Dakota,

Grand Forks; USNM, National Museum of Natural History,
Since their discovery in the late 1800s (Marsh, 1889, 1891),
Smithsonian Institution, Washington, D.C.; UWGM, University
Triceratops and ‘Torosaurus’ (Ceratopsidae; Chasmosaurinae)
of Wisconsin-Madison Geology Museum, Madison; YPM, Yale
have been regarded as the last of the horned dinosaurs. Tricer-
Peabody Museum, New Haven.
atops is characterized as having a three-horned face with a rather
short parietal-squamosal frill, and a solid, non-fenestrated pari-
etal (Ostrom and Wellnhofer, 1986; Forster, 1996). ‘Torosaurus’ MATERIALS AND METHODS
is distinguished from Triceratops solely on the basis of its ex-
panded, fenestrated parietal-squamosal cranial frill (Sullivan This analysis of Triceratops cranial morphology is, in part, a re-
et al., 2005; Farke, 2007). All other aspects of these animals’ sult of the Museum of the Rockies’ 10-year Hell Creek Project, an
remains are indistinguishable (Farke, 2007). Triceratops and attempt to reconstruct the ecosystems of the terminal Cretaceous
‘Torosaurus’ are known from the same geological unit, the Hell Hell Creek Formation of Montana and surrounding provinces.
Creek Formation of Montana and the Dakotas, and equiva- The discoveries that are resulting from the Hell Creek Project
lent Maastrichtian (latest Cretaceous) age sediments in adjacent (e.g., Horner and Padian, 2004; Horner and Goodwin, 2006, 2008,
states and provinces. 2009; Horner et al., 2009a; Scannella and Fowler, 2009) empha-
We report here on an analysis of changes in parietal- size the importance of large-scale field studies in which numerous
squamosal morphology in an ontogenetic sequence of Tricer- individuals are collected.
atops. ‘Torosaurus’ specimens were not excluded from this Triceratops is the most commonly recovered dinosaur in the
investigation and in all cases were found to exhibit ontogenetic Hell Creek Formation; 40% of dinosaur specimens collected dur-
markers of mature individuals, including anteriorly inclined ing the course of the Hell Creek Project, including over 30 skulls
postorbital horn cores, dorsoventrally flattened epoccipitals, and (Appendix 1) are referable to this genus (Horner, unpublished
fully developed osteohistologic features (Horner and Goodwin, data). Triceratops is represented by various ontogenetic stages,
2006). Furthermore, transitional features are noted in subadult ranging from specimens with skulls less than half a meter in
specimens of Triceratops that reveal the ontogenetic trajectory length to skulls that are over two meters in length (Horner and
of the genus, culminating in the frill morphology previously Goodwin, 2006). Most of these specimens have yet to be formally
considered diagnostic of ‘Torosaurus.’ described. ‘Torosaurus,’ on the other hand, is both rare and gen-
Institutional Abbreviations—AMNH, American Museum of erally larger: the smallest known skulls are around two meters in
Natural History, New York; ANSP, Academy of Natural Sci- length and the largest approach three meters in length.
ences, Philadelphia; BYU, Brigham Young University Earth Sci- The large collection of Triceratops cranial material amassed by
ence Museum, Provo; FMNH, Field Museum of Natural His- the MOR was supplemented with the examination of specimens
tory, Chicago; LACM, Natural History Museum of Los Ange- from several North American institutions, including the numer-
les County, Los Angeles; MCZ, Museum of Comparative Zo- ous type specimens at the National Museum of Natural History
ology, Harvard University, Cambridge; MOR, Museum of the and the Yale Peabody Museum. Specimens were placed within
Rockies, Bozeman; MPM, Milwaukee Public Museum, Milwau- the ontogenetic groups (baby, juvenile, subadult, adult) defined
kee; SMM, Science Museum of Minnesota, St. Paul; UCMP, by Horner and Goodwin (2006) based on the morphological cri-
University of California Museum of Paleontology, Berkeley; teria described therein. This study was not limited to complete
skulls; ontogenetic markers (such as shape of the postorbital horn
cores and epoccipitals) allow placement of partial specimens into
*Corresponding author. a growth series (Horner and Goodwin, 2006).

1157
1158 JOURNAL OF VERTEBRATE PALEONTOLOGY, VOL. 30, NO. 4, 2010

FIGURE 1. Development of thin regions of

the parietal throughout Triceratops ontogeny.
A, Ventral view of the parietal of a juvenile
Triceratops, MOR 2951; B, MOR 2924 (left
half of parietal). In subadult Triceratops, shal-
low depressions form on the ventral surface
of the parietal, adjacent to the lateral margins.
These depressions are initially undetectable on
the dorsal surface; C, MOR 2946 (left half of
parietal). As the parietal grows, the shallow
regions become isolated from the lateral mar-
gins. The parietal is still unperforated at this
stage. Dashed line indicates approximate ex-
tent of fenestra in ‘Torosaurus’; D, MOR 1122.
The thinning regions are resorbed late in on-
togeny, forming the characteristic fenestrae of
‘Torosaurus.’ Arrows indicate rim of thinning
region on the ventral surface. Scale bars equal
10 cm.

Measurements were taken with digital calipers, sliding calipers, frill continued to grow, the depressions became isolated from the
or a metric tape measure where appropriate. All measure- lateral margins of the parietal (MOR 2946, AMNH 970; Fig. 1C).
ments were taken at least twice and then averaged. Squamosal The parietal thins abruptly from approximately 2.5 to 3 cm or
length/width was plotted against squamosal length and an expo- more at its thickest point near the caudal edge to less than 0.5 cm
nential least squares regression analysis was performed using Mi- within the depressions (as exemplified by MOR 2946). These ar-
crosoft Excel. eas are often not fully preserved in specimens of Triceratops (e.g.,
Osteohistology specimens were processed following current AMNH 5116, MOR 004, MOR 2946, UND 3000), likely due to
techniques (Horner et al., 2009b). Specimens were embedded in the inherently fragile nature of the thin bone.
polyester resin, sectioned with a diamond powder disk on a pre- MOR 1122, the most complete skull of ‘Torosaurus,’ retains a
cision saw, ground on a lap wheel to achieve the desired optical shallow rim bordering the caudal margin of its left parietal fen-
contrast, and polished. estra (Fig. 1D). The parietal is 2.62 cm thick (measured 2 cm an-
terior to the caudal margin), intermediate in thickness between
RESULTS the characteristically thick parietal of Triceratops and the thin-
ner condition that is typical of ‘Torosaurus’ (1 to 2 cm). This al-
Ontogenetic Development of Parietal Fenestrae lowed for the retention of the ‘thinning rim’ on the ventral sur-
An ontogenetic series of Triceratops skulls reveals that late in face, which is otherwise eliminated by the overall thinning of the
ontogeny the ventral surface of the parietal developed thinning parietal. MOR 981, the only other ‘Torosaurus’ specimen known
regions, which we hypothesize are the precursors to fenestrae to preserve the entire border of a parietal fenestra (parietal thick-
(Fig. 1; Table 1). Expansion of parietal fenestrae through on- ness 1.9 cm), does not display a similar thickened rim on the ven-
togeny has been noted in previous studies of ceratopsians (Dod- tral surface. The maximum widths of the fenestrae of MOR 1122
son and Currie, 1988; Brown et al., 2009). Although Triceratops are 36.5 cm (left, with thickened rim present) and 49 cm (right).
has been diagnosed as possessing an extremely thick, solid pari- The sole preserved fenestra of MOR 981 has a maximum width
etal (Forster, 1996), this thickness is not uniform. The regions of of 53 cm.
the parietal that are fenestrated in other chasmosaurines became The osteohistology of the thinning region of the parietal of a
extremely thin (less than .5 cm thick) as Triceratops matured. young adult Triceratops (MOR 2946; Fig. 2) reveals that resorp-
Juvenile Triceratops possessed subtly thinner regions of the tion occurs on both the ventral and dorsal surfaces. Multiple ero-
parietal adjacent to the lateral margins, roughly 5–10 cm from sion surfaces are visible (Fig. 2D) indicating that the parietal is
the frill’s caudal edge. In larger subadult specimens, these thin extensively remodeled throughout ontogeny.
regions are expanded onto the ventral surface of the parietal and
form two prominent fossae (as seen in YPM 1823). These fos-
Squamosal Elongation
sae expanded into large, semi-circular depressions that extended
medially toward the parietal midline and rostrally toward the Concurrent with the development of parietal fenestrae is
supratemporal fenestrae (as seen in MOR 2924; Fig. 1B). As the the elongation of the squamosals (Figs. 3, 4). The smallest
 SCANNELLA AND HORNER—TOROSAURUS IS TRICERATOPS 1159

TABLE 1. Ontogenetic trends expressed in the parietal-squamosal frill of Triceratops.

Ontogenetic trends
Triceratops Growth stage Parietal length (cm) A B C D E F
UCMP 154452 Baby 12
MOR 2569 Juvenile 26 x
MOR 1199 Juvenile 38 x
MOR 2951 Juvenile 42 x
UCMP 136306 Juvenile 46 x
MOR 1110 Juvenile 49 x
MOR 1120 Subadult 71 np
YPM 1823 Subadult 64 x
MOR 2924 Subadult 68 x
MOR 2950 Subadult — x
MOR 2574 Subadult — x
MOR 1625 Young adult — x
LACM 59049 Young adult 83 x
AMNH 970 Young adult —∗ x
MOR 2946 Young adult — x
MOR 2702 Young adult — x∗∗ x
AMNH 5116 Young adult 83 x x
USNM 2412 Young adult 71 x x
UWGM 732 Young adult — x np
SMM P97.6.1 Adult 100 (est) x x
MOR 981 Adult 116 x x
YPM 1831 Adult 134 (est) x x
YPM 1830 Adult — x x
ANSP 15192 Adult 85 x x
MOR 1122 Adult 125 x x
MPM VP6841 Adult 107 (est) x x

est, estimate; np, not preserved. Ontogenetic trends: A, thinning on lateral margins of parietal; B, depressions along lateral margins on ventral side
of parietal; C, ventral depressions isolated from lateral margins of parietal; D, parietal exhibits textural change surrounding incipient fenestrae; E,
squamosals elongate; F, fenestrae open. ∗ AMNH970 not measured due to enclosure in display case. ∗∗ Lateral margins of MOR2702 are not preserved
and thus it is unknown if the incipient fenestrae are fully isolated from the margins.

known Triceratops skull (UCMP 154452) possesses extremely ual or stratigraphic variation (see comments on variation in
short squamosals which are reminiscent of those found in the Discussion).
‘short-frilled’ centrosaurine ceratopsids (Goodwin et al., 2006;
Dodson, 1996; Fig. 3A). In contrast to other chasmosaurines
The Ontogenetic Spectrum of Triceratops
in which the squamosals gradually elongate (Lehman, 1990),
Triceratops is paedomorphic and retains the juvenile morphol- The placement of specimens into an ontogenetic spectrum
ogy throughout most of its development (Fig. 3B–F). This reten- highlights transitional features. AMNH 5116 is a large skull
tion of relatively short, broad squamosals led to Triceratops being (200 cm skull length, measured from tip of the rostral to the
grouped with the centrosaurine ceratopsids for many years rather parietal-squamosal contact at caudal frill margin) with extremely
than in its rightful place among the ‘long-frilled’ chasmosaurines narrow squamosals and a fairly thin (2.3 cm, measured 2 cm an-
(Dodson, 1996; Lull, 1933; Sternberg, 1949). Only very late in terior to the caudal margin) parietal (Fig. 5A). The anteriorly
Triceratops ontogeny do the squamosals become elongate, pro- inclined postorbital horn cores of AMNH 5116 indicate matu-
ducing intermediates between the broad, juvenile condition and rity comparable to that seen in other ‘adult’ Triceratops speci-
the blade-like state found in ‘Torosaurus’ (Fig. 3G, H). These mens. This individual has been identified as Triceratops rather
intermediate squamosal morphologies correlate with advanced than ‘Torosaurus’ because its parietal is unfenestrated. However,
parietal thinning (see Table 1). A thickened squamosal bar forms close inspection reveals a pronounced sloping of the parietal,
along the contacts with the parietal in mature specimens (Colbert from thicker bone near the caudal margin to thinner bone in the
and Bump, 1947; Farke, 2007). regions that are fenestrated in ‘Torosaurus.’ Many ‘Torosaurus’
As seen in Figure 4, Triceratops squamosals eventually specimens retain a gradual sloping of the parietal in these regions,
adopted the morphology previously considered diagnostic of posterior to the fenestrae, which we hypothesize to be a remnant
‘Torosaurus.’ This is demonstrated by the good correlation of the transition from solid frill to perforation (MOR 1122, MPM
between squamosal length and elongation when an exponen- VP6841, SMM P97.6.1). The sloping of the parietal of AMNH
tial least squares regression is performed (R2 = .782). Size, 5116 coincides with a textural change from a striated texture to a
however, is not the most reliable indicator of maturity (see more rugose, pebbly surface texture within the areas where fenes-
Discussion). ‘Torosaurus’ specimen SMM P97.6.1 is notewor- trae would be present in ‘Torosaurus’ (Fig. 5C). A similar textural
thy in its possession of slightly wider squamosals than those transition has been noted to occur on the margins of developing
of other ‘Torosaurus’ specimens. This individual’s squamos- parietal fenestrae in centrosaurines (Brown et al., 2009) and rep-
als are wide rostrally and taper caudally (see Fig. 3J), which resents a transition from the characteristic solid parietal of Tricer-
is consistent with elongation from the immature morphology atops to the fenestrated condition found in mature individuals.
considered diagnostic of Triceratops. Interestingly, a specimen Striated frill texture corresponding to visible radial canals is
from the Frenchman Formation of Saskatchewan, which was observed in juvenile Triceratops (UCMP 154452, MOR 1199,
originally referred to ‘Torosaurus latus’ (Tokaryk, 1986; this MOR 2569, MOR 2951) as well as other juvenile ceratopsians
has since been contested by Sullivan et al., 2005), exhibits (Sampson et al., 1997; Brown et al., 2009). It is associated with
a similar squamosal morphology. This may represent individ- rapid growth (Francillon-Vieillot et al., 1990; Sampson et al.,
1160 JOURNAL OF VERTEBRATE PALEONTOLOGY, VOL. 30, NO. 4, 2010

FIGURE 2. Triceratops parietal fragment, MOR 2946, showing the histology of the ventral erosion surface. A, view of whole fragment. Ventral
surface faces up. Letters show placement of parts B, C, and D. Specimen is 6 cm long; B, ventral surface of the slope between the thickened and
thinned regions of the parietal. Scale bar equals 1 mm; C, erosion surface of the thinned region showing eroded secondary osteons. Scale bar equals
100 µm; D, erosion surface of thickened region showing an older erosion surface (arrows) with subsequent deposition. Scale bar equals 100 µm.

1997), the radial vascular canals being oriented in the direc- of Triceratops, ‘Triceratops hatcheri’ (Lull, 1933; Ostrom and
tion of expansion. Several ‘Torosaurus’ specimens have a sim- Wellnhofer, 1986). Suggested autapomorphies of this specimen
ilar surface texture on their frills, with the striations varying include the absence of a nasal horn and the presence of a small
in thickness from being fine to fairly broad, resembling what fenestra in the parietal (Forster, 1996). The apparent absence
Tumarkin-Deratzian (2009) refers to as a ‘grooved’ surface tex- of an epinasal in this specimen is unlikely to be an autapo-
ture. Histological examination of a parietal (MOR 981, a large morphy because similar nasal horn morphologies are found in
‘Torosaurus’) with striated surface texture reveals that it is as- Triceratops and ‘Torosaurus’ specimens (MOR 981, MOR 1122,
sociated with the expansion of the frill. The presence of a stri- UCMP 128561, USNM 2410). Alternatively, the epinasal could
ated surface texture on the parietal of AMNH 5116 and several have been lost in vivo or simply disarticulated from the fossil as
‘Torosaurus’ specimens (MOR 981, MPM VP6841, SMM P97.6.1, a result of taphonomic processes (Horner and Goodwin, 2008).
YPM 1831) that display clear ontogenetic markers of maturity, In addition, the specimen shows evidence of severe pathologies
including anteriorly inclined postorbital horn cores and flattened (Tanke and Farke, 2007). USNM 2412 possesses dorsoventrally
epocciptals, as well as less reliable ontogenetic markers (rela- flattened epoccipitals and subtly procurving postorbital horn
tively large size, apparent closure of cranial sutures, see Discus- cores—indicators of ontogenetic maturity. Its elongate squamos-
sion), suggests that the late expansion of the frill occurs rapidly, als and incipient parietal fenestra are exactly what would be pre-
concurrent with overall thinning. AMNH 5116 is a Triceratops dicted in a Triceratops that is transitional between the short, solid
that died just prior to attainment of the complete ‘Torosaurus’ subadult frill and the elongate, fenestrated condition found in
morphology. adults. We consider USNM 2412 an ontogenetically transitional
The problematic specimen, USNM 2412, has been placed in specimen of Triceratops.
its own genus ‘Nedoceratops’ (Ukrainsky, 2007, 2009); at times UWGM 732 is a specimen with anteriorly inclined postorbital
it has been the sole representative of ‘Diceratus’ (Mateus, 2008), horn cores and dorsoventrally flattened epoccipitals. It possesses
‘Diceratops’ (Lull, 1905; Hatcher et al., 1907), or its own species elongate squamosals (Fig. 3I) and a thin parietal (1.5 cm thick
 SCANNELLA AND HORNER—TOROSAURUS IS TRICERATOPS 1161

FIGURE 3. Ontogenetic elongation of the squamosals in Triceratops. Specimens arranged in hypothesized ontogenetic sequence: A, UCMP 154452;
B, MOR 2569; C, MOR 1199; D, MOR 1110; E, MOR 1120; F, MOR 004; G, MOR 2702; H, AMNH 5116; I, UWGM 732; J, ‘Torosaurus’ SMM
P97.6.1 (image reversed for comparison); K, ‘Torosaurus’ MOR 1122. Scale bars equal 10 cm.

FIGURE 4. Squamosal elongation in Tricer-

atops. Growth stages: baby: UCMP 154452; ju-
venile: LACM 149538, MOR 1110, MOR 1199,
MOR 2569, MOR 2927, MOR 2928, MOR
2929, MOR 2572, MOR 2951, UCMP 136306;
subadult: MOR 1120, MOR 2923, YPM1821,
YPM 1822, YPM 1823; young adult: AMNH
5116, FMNH P12003, LACM 59049, MOR
004, MOR 1625, MOR 2702, MOR 2942, SMM
P60.2.1, UCMP 113697, USNM 2412; UWGM
732∗ ; adult (‘Torosaurus’): ANSP 15192, MOR
1122, MPM VP6841, SMM P97.6.1, YPM 1830,
YPM 1831. Squamosal length measured from
parietal/squamosal contact at caudal frill mar-
gin to anterolateral most point of squamosal
margin. Squamosal width measured across
widest point of squamosal. ∗ Estimate.
1162 JOURNAL OF VERTEBRATE PALEONTOLOGY, VOL. 30, NO. 4, 2010

FIGURE 5. Comparison of AMNH 5116 to ‘Torosaurus latus’ specimens. A, AMNH 5116, lower jaw belongs to a separate individual (AMNH 5039).
Note anteriorly inclined postorbital horn cores and elongate squamosals. Scale bar equals 10 cm; B, YPM 1830, the type specimen of ‘Torosaurus latus.’
Scale bar equals 10 cm; C, close-up of the dorsal surface of the parietal of AMNH 5116 revealing the transition from striated surface texture over the
caudal end of the parietal to a pebbly surface texture in the region where the fenestrae will open. Scale bar equals 3 cm; D, ventral surface of the
parietal of ‘Torosaurus’ MOR 981, revealing striated surface texture. Scale bar equals 1 cm. Image of YPM 1830 © 2010 Peabody Museum of Natural
History, Yale University, used with permission.

measured 2 cm anterior to the caudal margin). However, the vealed to be immature; we will refer to this as the young adult
squamosals had yet to develop the thickened squamosal bar growth stage. ‘Torosaurus’ (e.g., MOR 981, MOR 1122, YPM
that is found in ‘Torosaurus latus’ specimens and it is unclear if 1830) represents the adult stage of Triceratops ontogeny.
the parietal was fenestrated at the time of death. We consider
UWGM 732 a young adult Triceratops. It is interesting to note Epiparietal and Episquamosal Variation
that the posteriorly curved postorbital horn core of a very small
juvenile Triceratops was collected from the same site as the The epoccipitals (epiparietals and episquamosals) of Tricer-
larger individual. atops undergo dramatic changes in morphology throughout on-
Inclusion of ‘Torosaurus’ specimens in the examination of togeny (Horner and Goodwin, 2006, 2008). All juvenile spec-
Triceratops cranial ontogeny highlights transitional morpholog- imens possess triangular epoccipitals. As individuals matured,
ical features that link these two taxa as ontogenetic stages of the epoccipitals became dorsoventrally flattened onto the frill
the same taxon. As such, the stages of Triceratops growth out- margin.
lined by Horner and Goodwin (2006) must be expanded upon Triceratops has previously been diagnosed as possessing an
in order to encompass the full ontogenetic spectrum of this genus unvarying configuration of five epiparietals, one of which con-
(Table 2). Individuals bearing anteriorly inclined postorbital horn sistently straddles the midline (Forster, 1996). This character
cores and dorsoventrally flattened epoccipitals, which have yet has been used to distinguish Triceratops from ‘Torosaurus’ and
to develop the complete ‘Torosaurus’ frill morphology (e.g., other chasmosaurines (Farke and Williamson, 2006; Farke, 2007).
AMNH 5116, MOR 004, MOR 1625, MOR 2702, USNM 2412) However, several Triceratops specimens possess more than five
and which were previously considered mature adults, are re- epiparietals (LACM 27428, USNM 2100; Horner and Goodwin,
 SCANNELLA AND HORNER—TOROSAURUS IS TRICERATOPS 1163

TABLE 2. Growth stages of Triceratops.

p.o. Horn orientation

Growth stage Example (curvature) Epoccipitals Squamosals Parietal
Baby UCMP 154452 Straight, no curvature np Short, broad Unfenestrated
Juvenile MOR 1199 Posterior Delta-shaped Short, broad Unfenestrated
MOR 1110
Subadult MOR 1120 Anterior tip of horn Delta-shaped Short, broad Unfenestrated
posterior
Young adult MOR 004 Anterior d-v Broad to elongate, Unfenestrated to
no squamosal bar small incipient
fenestrae
MOR 1625
MOR 2702
Adult MOR 981 Anterior d-v Elongate, squamosal Fenestrated
bar present
MOR 1122

Modified from Horner and Goodwin, 2006. np, not preserved; d-v, dorsoventrally compressed.

2008). USNM 1201 was described as possessing six epiparietals (MOR 1122, MPM VP6841) preserve seven episquamosals per
without evidence of one on the midline (Hatcher et al., 1907); squamosal.
however, this was an inference made from the morphology of
the parietal margin and only one epoccipital was found with the
specimen. MOR 2923 represents an individual with an unfen- Osteohistology: a Test of Ontogenetic Hypotheses
estrated parietal in which six epiparietals are clearly preserved, The osteohistology of the postorbital horn cores of Tricer-
none of which are found on the midline (Fig. 6). ‘Torosaurus’ atops reveal the expected tissue ontogeny seen in other bones
specimen MOR 981 preserves only the right half of its parietal, (Francillon-Vieillot et al., 1990; Chinsamy-Turan, 2005), in that
on which three epiparietals are clearly visible. Farke (2007) noted young Triceratops horns are composed of primary tissues,
the presence of five parietal marginal undulations, suggesting a whereas the horns of older individuals are composed of more
total count of 10 epiparietals for this individual. MOR 1122 (pre- mature tissues. As seen in Figure 7A–B, the posteriorly curved
viously assigned to ‘Torosaurus’) has 12 epiparietals and lacks a postorbital horn core of a small juvenile Triceratops (comparable
midline epiparietal as well. The number and position of epipari- in size to MOR 2569; see Fig. 3B) is very porous and composed
etals is variable within Triceratops and the greater number found exclusively of primary tissue. The postorbital horn core tissue of
in MOR 981 and MOR 1122 suggests that this number increases a much larger juvenile (orbital horns still arched backward) is
throughout ontogeny. slightly less porous and composed of tissues that are mostly pri-
The number of episquamosals is also variable. Triceratops has mary, but contain some secondary osteons.
classically been restored with four or five episquamosals per The postorbital horn cores of young adult Triceratops skulls,
squamosal in addition to one that bridges the parietal/squamosal approximately 2 m in length (Fig. 7E–H), have tissues that have
gap (YPM 1821, YPM 1823). However, MOR 1120 clearly been remodeled, but are not yet composed of multigenerational
preserves seven episquamosals on its right squamosal; six ‘Haversian’ canals. The tissue is still rather spongy, and fibro-
are preserved on the left squamosal. ‘Torosaurus’ specimens cytes are abundant and well ordered (Horner and Goodwin, 2009;

FIGURE 6. The parietal of Triceratops speci-

men MOR 2923 (skull slightly distorted from
postmortem crushing). Arrows indicate epi-
parietals. Dashed line represents the midline.
Scale bar equals 10 cm.
1164 JOURNAL OF VERTEBRATE PALEONTOLOGY, VOL. 30, NO. 4, 2010

FIGURE 7. Osteohistology of an ontogenetic

sequence of Triceratops postorbital horn cores.
A, UCMP 159233, juvenile postorbital horn
core (approximately 20 cm in length, arched
backward), transverse section, showing spongy
tissue with very high vascularization; B, close-
up of A showing that the spongy tissue is
entirely primary, indicative of its young age;
C, MOR-HYP-P31.Horn1-D2-2, large juvenile
postorbital horn core (approximately 50 cm
in length, arched backward), transverse sec-
tion showing spongy tissue with high vascu-
larization; D, close-up of C showing that the
spongy tissue is mostly composed of primary
tissues with some secondary osteons; E, MOR
1625.Horn1-1, young adult postorbital horn
(orbital horns incomplete), transverse section
showing compact ‘Haversian’ bone with nu-
merous open erosion rooms and fairly high
vascularization; F, close-up of E showing tis-
sue composed of loosely packed ‘Haversian’
systems with large erosion rooms; G, MOR
2702.Horn1-1, young adult postorbital horn
core (approximately 55 cm in length arched
forward), transverse section, showing dense
‘Haversian’ systems with fairly high vascular-
ization; H, close-up of G showing a tissue com-
posed of loosely packed ‘Haversian’ systems
with large erosion rooms; I, MOR 1122.Horn1-
2, adult postorbital horn core (approximately
60 cm in length, arched forward), transverse
section, showing dense tissue with moder-
ate vascularization; J, close-up of G show-
ing a tissue composed of dense, multigenera-
tional ‘Haversian’ canals, indicative of a ma-
ture individual.

Horner and Lamm, 2009). Squamosal morphology in these spec- as mature as individuals with the expanded, fenestrated parietal-
imens ranges from the broad morphology typical of what were squmosal frill morphology previously considered diagnostic of
previously considered adult Triceratops (MOR 1625) to fairly ‘Torosaurus.’
elongate (MOR 2702; Fig. 3G). At this point the parietal has pro-
nounced thinning regions and epoccipitals are dorsoventrally flat- DISCUSSION
tened.
The postorbital horn core of a mature Triceratops The radical changes in cranial morphology that occur through-
(‘Torosaurus’) is composed of multigenerational dense ‘Haver- out ceratopsid ontogeny (Sampson, 1995; Horner and Goodwin,
sian’ tissue, indicative of very mature bone (Fig. 7 I–J). This 2006) entail that an understanding of ontogenetic development is
histological ontogenetic sequence clearly demonstrates that critical to studies of their paleoecology and systematics. Tricer-
Triceratops that were previously considered to be adults are not atops is paedomorphic relative to other chasmosaurines in that it
 SCANNELLA AND HORNER—TOROSAURUS IS TRICERATOPS 1165

retains juvenile features (broad squamosals, unfenestrated pari-

etal) until extremely late in ontogeny. A large sample of speci-
mens reveals the full ontogenetic spectrum of Triceratops, with
intermediate morphologies linking individuals that could easily
be mistaken for distinct species without an understanding of the
degree of plasticity in the skulls of these animals.

Ontogenetic Variation
It has been suggested that ANSP 15192 is an immature
‘Torosaurus’ given its relatively small size (177 cm skull length;
Colbert and Bump, 1947). However, although it was originally
described as not possessing epoccipitals, the epiparietals and
episquamosals of ANSP 15192 are indeed preserved and exhibit
the mature morphology of being dorsoventrally flattened and
fused to the frill border (Fig. 8). In fact, all ‘Torosaurus latus’
specimens in which the epoccipitals are preserved exhibit this
mature morphology (ANSP 15192, MOR 981, MOR 1122, MPM
VP6841). No ‘Torosaurus’ specimen has posteriorly curving pos-
torbital horn cores, which would be indicative of immaturity
(Horner and Goodwin, 2006). To the extent that any ‘Torosaurus’
specimens have a degree of sinuosity to their postorbital horn
cores, it falls within the range of variation found in young adult
Triceratops (as seen in MOR 1604 and UCMP 113697). Inter-
estingly, the postorbital horn cores of several ‘Torosaurus’ speci-
mens show signs of pronounced resorption (ANSP 15192, MOR
981, MOR 1122). Remodeling of the postorbital horn cores was
still occurring in these mature animals.
Size alone is not a reliable indicator of ontogenetic stage and
there is a strong degree of variation in the timing of cranial
fusion in Triceratops (Horner and Goodwin, 2008). For example,
YPM 1822 is a small skull (154 cm skull length) in which the
cranial sutures appear to be fused. However, it is displayed
alongside the much larger skull YPM 1821 (187 cm skull length)
in which the cranial sutures are open. MOR 2952 is a recently
discovered large Triceratops in which the 27 cm long epinasal has
yet to fuse to the nasals. Similarly, the large ‘Torosaurus’ MOR
1122 (skull length 255 cm) was discovered with an associated
isolated premaxilla that was equivalent in size to its own and
yet unfused. BYU 12183 is an extremely large Triceratops skull
(estimated skull length 250 cm); unfortunately the areas of the
parietal, which may reveal thinning, incipient fenestrae, or tex-
tural transitions, are largely reconstructed. Ontogenetic markers
(horn core curvature, epoccipital morphology) can be used to FIGURE 8. The dorsoventrally flattened epoccipitals of ‘Torosaurus’
group individuals into growth stages, but in ideal situations a specimen ANSP 15192 reveal ontogenetic maturity. Scale bar equals
10 cm.
combination of factors (including osteohistology, bone texture,
sutural fusion, and size) should be considered.

Alternative Hypotheses
tors of maturity, such as anteriorly inclined postorbital horn cores
Alternative hypotheses to the synonymy of these genera may and/or dorsoventrally flattened epoccipitals.
include the suggestion that ‘Torosaurus’ does not conform to Ostrom and Wellnhofer (1990) suggested that the differences
the ontogenetic trends observed in Triceratops. This is unparsi- in frill morphology that distinguish ‘Torosaurus’ from Triceratops
monious given the close phylogenetic relationship proposed for may be the result of sexual dimorphism in a single genus. The
‘Torosaurus’ and Triceratops (Dodson et al., 2004). Posterior cur- predictable nature of parietal thinning throughout ontogeny in
vature of postorbital horn cores in immature individuals is ob- Triceratops, the osteohistological evidence for greater maturity
served in ceratopsians as primitive as Zuniceratops (Wolfe and in ‘Torosaurus’ specimens, and the lack of strong evidence for
Kirkland, 1998; Horner and Goodwin, 2006). sexual dimorphism in non-avian dinosaurs (Padian et al., 2004;
Another possibility is that ‘Torosaurus’ is a valid genus that Goodwin et al., 2006) argues against this proposal. Sexual dimor-
is indistinguishable from Triceratops until relatively late in on- phism accounting for differences in size cannot be ruled out at
togeny when its fenestrae develop. Given the geographic and this time.
stratigraphic overlap of these animals, ontogeny is more parsimo-
nious than erecting two distinct genera, particularly because tis-
Implications for Dinosaur Diversity
sue ontogeny reveals that ‘Torosaurus’ specimens are more ma-
ture than any Triceratops specimens, including those previously No centrosaurine material is known from the Hell Creek
considered adults. Furthermore, specimens that might otherwise Formation or adjacent equivalent-aged formations. It appears
be considered immature ‘Torosaurus’ but are here considered that only the chasmosaurine ceratopsids survived into the latest
transitional ontogenetic stages of Triceratops (e.g., AMNH 5116, Maastrichtian (Dodson, 1996). Of this group, Triceratops and
MCZ 1102, MOR 2702, USNM 2412) all bear ontogenetic indica- ‘Torosaurus’ were proposed to be the last of their lineages
1166 JOURNAL OF VERTEBRATE PALEONTOLOGY, VOL. 30, NO. 4, 2010

(Sternberg, 1949). The synonymy of these genera diminishes the Remarks—Given the amount of morphological change that oc-
diversity of ceratopsids in the Hell Creek Formation. Similar curs throughout the lifespan of an individual, the use of a single
synonymizations are occurring for several taxa, including tyran- type specimen for taxonomic purposes is problematic if the diag-
nosaurs (Carr, 1999) and pachycephalosaurs (Horner and Good- nosable characters are not limited to the adult ontogenetic stage.
win, 2009), in the latest Cretaceous of North America. These Hence, multiple ontogenetic stages are easily misinterpreted as
advances show that dinosaur diversity was more depleted than distinct taxa (Dodson, 1975). The revised diagnosis of Triceratops
traditionally thought well before the end of the Cretaceous presented here is based on characters found in a mature individ-
Period. ual.

SYSTEMATIC PALEONTOLOGY CONCLUSIONS

CERATOPSIA Marsh, 1888 Recognition of the full scope of Triceratops ontogeny empha-
NEOCERATOPSIA Sereno, 1986 sizes the radical degree of cranial morphological transformation
CERATOPSIDAE Marsh, 1888 that occurred throughout development. Postorbital horn cores
CHASMOSAURINAE Lambe, 1915 reoriented from being straight, to posteriorly directed, to an-
TRICERATOPS Marsh, 1889 teriorly directed correlated with the dorsoventral flattening of
(Fig. 5A, B) epiparietals and episquamosals throughout ontogeny (Horner
and Goodwin, 2006, 2008). The initially unfenestrated parietal-
Revised Diagnosis—Skull bears elongate postorbital horn
squamosal frill thickens throughout development, with some
cores (derived from fusion of the postorbitals and prefrontals
specimens reaching a thickness in excess of 6 cm (MOR 2969),
early in ontogeny) plus a single variable epinasal horn. Epinasal
before rapidly expanding, thinning, and ultimately adopting the
unites the rostral-nasal-premaxillae complex. The shape of the
fenestrated condition found in all other chasmosaurines. We hy-
frontal fontanelle is variable, though generally circular when
pothesize that these dramatic changes in cranial ornamentation
present; some specimens lack a frontal fontanelle as a result
functioned in intraspecific communication, signaling relative ma-
of closure of the frontals and parietal. Parietal-squamosal frill
turity.
relatively short, broad, and fan-like (compared to other chas-
It is telling that no confirmed juvenile ‘Torosaurus’ skulls have
mosaurine genera). Fenestrated and expanded in mature individ-
been reported. Incorporation of ‘Torosaurus’ into the spectrum
uals. Strong undulating midline parietal ridge throughout early
of Triceratops ontogeny explains why ‘Torosaurus’ is only known
ontogeny becomes a broad parietal bar in mature specimens; fen-
from a few mature individuals in the Hell Creek Formation,
estrae are ovate to nearly circular. Pronounced squamosal bars
whereas Triceratops is extremely abundant and represented by
alongside parietal-squamosal contacts in mature specimens. Epi-
several growth stages (Horner and Goodwin, 2006). Immature
parietals and episquamosals ornament posterior and lateral mar-
‘Torosaurus’ actually have been known for over a century but
gins of the parietal-squamosal frill. Epijugals on the jugal flange
have been called Triceratops. The fact that the majority of Tricer-
fuse the jugal-quadratojugal complex (modified from Ostrom and
atops specimens that have been collected since the initial descrip-
Wellnhofer, 1986, and Forster, 1996).
tion of the genus (Marsh, 1889) are not fully mature suggests that
Valid Species—Triceratops horridus Marsh, 1889 (type
either Triceratops mortality was fairly high before full maturity
species); holotype: YPM 1820 (Hatcher et al., 1907:pl. XXVI, figs.
was reached, or adults did not live in the same areas as immature
24, 25, 27); diagnosis same as for genus.
animals.
Triceratops prorsus Marsh, 1890; holotype: YPM 1822 (Forster
Ontogenetic analyses of dinosaur morphology highlight transi-
1996:fig. 9B); diagnosis as per Forster (1996).
tional features between specimens and suggest that much of the
Triceratops horridus and T. prorsus are stratigraphically sep-
variation previously attributed to taxonomic differences is actu-
arated (Scannella and Fowler, 2009). This finding and its evolu-
ally a product of developmental processes (Dodson, 1975). Stud-
tionary, taxonomic, and biogeographic implications will be ad-
ies focused on the dinosaur fauna of the Hell Creek Formation
dressed in detail in a future article.
are depleting latest Cretaceous dinosaur diversity through on-
Locality and Horizon (YPM 1820)—Section 2, T.36N., R.64W.
togenetic synonymies (Horner et al., 2007). As such, a trend of
Niobrara County, Wyoming, U.S.A. “About the middle of the
decreasing dinosaur diversity preceding the end of the Mesozoic
upper half of the Lance formation, Late Cretaceous” (Ostrom
Era is implied.
and Wellnhofer, 1986:156).
Synonyms—Torosaurus latus Marsh, 1891 (rediagnosed by
ACKNOWLEDGMENTS
Colbert and Bump, 1947; Sullivan et al., 2005; Farke, 2007); Ne-
doceratops hatcheri (Lull, 1905): Ukrainsky, 2007 (new combina- We are grateful to all of the institutions that permitted ac-
tion, Diceratops preoccupied) (= Diceratops hatcheri Lull, 1905 cess to specimens. Image of YPM 1830 in Figure 5 is copyright
[original description]; Diceratus hatcheri [Lull, 1905]; Mateus, Peabody Museum of Natural History, Yale University. We thank
2008 [new combination, junior subjective synonym]). the MOR field crews as well as Makoshika State Park and Lon
“Torosaurus” utahensis Gilmore, 1946—Chasmosaurine re- Bolick for discovery and collection of specimens. Thanks to C.
mains from the southern ‘Alamosaurus fauna’ (Lehman, 1987) Ancell, D. Barta, S. Brewer, T. Bridges, N. Carroll, B. Harmon,
of Utah, New Mexico, and Texas have been referred to J. Heuck, P. Hookey, J. Jette, and L. Roberts for preparation
“Torosaurus” utahensis; however, there is still considerable de- of specimens. E. Lamm performed histological preparation. We
bate by students of ceratopsids over how many of these (largely thank K. Baker, B. Boessenecker, C. Boyd, M. Carrano, P. Dod-
fragmentary) specimens are actually diagnosable to the genus son, A. Farke, D. Fowler, E. Freedman, M. Goodwin, L. Hall, J.
level (Sullivan et al., 2005; Hunt and Lehman, 2008). The syn- Hartman, M. Holland, F. Jackson, P. Makovicky, M. Pillet, K. Pa-
onymy of Triceratops and the type species of ‘Torosaurus’ implies dian, A. Poust, J. Stiegler, D. Varricchio, and H. Woodward for
that “Torosaurus” utahensis is either a species of Triceratops, or helpful comments, suggestions, and stimulating conversations.
a different genus; either Arrhinoceratops (Parks, 1925), to which H. Woodward produced illustrations. Comments by R. Holmes
it was initially referred by Gilmore (1946), or its own distinct and two anonymous reviewers improved an early version of the
genus. We favor the first hypothesis, but as this southern mate- manuscript. Specimen collection was funded primarily by Nathan
rial was not included in the present study, a further diagnosis of Myhrvold. The BLM, USFW, and CMR as well as the Twitchell,
“Torosaurus” utahensis is beyond the scope of this paper and will Taylor, and Holen families permitted access to lands. Research
be investigated in a following study. funding for J.B.S. was provided by the Theodore Roosevelt
 SCANNELLA AND HORNER—TOROSAURUS IS TRICERATOPS 1167

Memorial Fund of the American Museum of Natural History, the Horner, J. R., A.deRicqlès, K. Padian, and R. D. Scheetz. 2009b. Compar-
Doris O. and Samuel P. Welles Research Fund of the University ative long bone histology and growth of the “hypsilophodontid” di-
of California Museum of Paleontology, and the Sands brothers. nosaurs Orodromeus makelai, Dryosaurus altus, and Tenontosaurus
tillettii (Ornithischia: Eurornithopoda). Journal of Vertebrate Pale-
ontology 29:734–747.
Hunt, R. K., and T. M. Lehman. 2008. Attributes of the ceratopsian di-
LITERATURE CITED
nosaur Torosaurus, and new material from the Javelina Formation
Brown, C. M., A. P. Russell, and M. J. Ryan. 2009. Pattern and transition (Maastrichtian) of Texas. Journal of Paleontology 82:1127–1138.
of surfacial bone texture of the centrosaurine frill and their ontoge- Lambe, R. S. 1915. On Eoceratops canadensis, gen. nov., with remarks on
netic and taxonomic implications. Journal of Vertebrate Paleontol- other genera of Cretaceous horned dinosaurs. Geological Survey of
ogy 29:132–141. Canada Museum Bulletin 12:1–49.
Carr, T. D. 1999. Craniofacial ontogeny in tyrannosauridae (Dinosauria, Lehman, T. M. 1987. Late Maastrichtian paleoenvironments and dinosaur
Coelurosauria). Journal of Vertebrate Paleontology 19:497–520. biogeography in the western interior of North America. Palaeogeog-
Chinsamy-Turan, A. 2005. The Microstructure of Dinosaur Bone. Johns raphy, Palaeoclimatology, Palaeoecology 60:189–217.
Hopkins University Press, Baltimore, Maryland, 195 pp. Lehman, T. M. 1990. The ceratopsian subfamily chasmosaurinae: sexual
Colbert, E. H., and J. D. Bump. 1947. A skull of Torosaurus from South dimorphism and systematics; pp. 211–229 in K. Carpenter and P. J.
Dakota and a revision of the genus. Proceedings of the Academy of Currie (eds.), Dinosaur Systematics: Approaches and Perspectives.
Natural Sciences of Philadelphia 99:93–106. Cambridge University Press, New York.
Dodson, P. 1975. Taxonomic implications of relative growth in lam- Lull, R. 1905. Restoration of the horned dinosaur Diceratops. American
beosaurine hadrosaurs. Systematic Zoology 24:37–54. Journal of Science 20:420–422.
Dodson, P. 1996. The Horned Dinosaurs. Princeton University Press, Lull, R. S. 1933. A revision of the Ceratopsia or horned dinosaurs. Yale
Princeton, New Jersey, 346 pp. Peabody Museum of Natural History Memoirs 3:1–178.
Dodson, P., and P. J. Currie. 1988. The smallest ceratopsid skull-Judith Marsh, O. C. 1888. A new family of horned Dinosauria from the Creta-
River Formation of Alberta. Canadian Journal of Earth Sciences ceous. American Journal of Science, series 3:334–336.
25:926–930. Marsh, O. C. 1889. Notice of gigantic horned Dinosauria from the Creta-
Dodson, P., C. A. Forster, and S. D. Sampson. 2004. Ceratopsidae; ceous. American Journal of Science, series 3:173–175.
pp. 494–513 in D. B. Weishampel, P. Dodson, and H. Osmólska Marsh, O. C. 1890. Description of new dinosaurian reptiles. American
(eds.), The Dinosauria. University of California Press, Berkeley, Journal of Science, series 3:81–86.
California. Marsh, O. C. 1891. Notice of new vertebrate fossils. American Journal of
Farke, A. A. 2007. Cranial osteology and phylogenetic relationships of Science, series 3:265–269.
the chasmosaurine ceratopsid Torosaurus latus; pp. 235–257 in K. Mateus, O. 2008. Two ornithischian dinosaurs renamed: Microceratops
Carpenter (ed.), Horns and Beaks: Ceratopsian and Ornithopod Di- Bohlin, 1953 and Diceratops Lull, 1905. Journal of Paleontology
nosaurs. Indiana University Press, Bloomington and Indianapolis, 82:423.
Indiana. Ostrom, J. H., and P. Wellnhofer. 1986. The Munich specimen of Tricer-
Farke, A. A., and T. E. Williamson. 2006. A ceratopsid dinosaur parietal atops with a revision of the genus. Zitteliana 14:111–158.
from New Mexico and its implications for ceratopsid biogeography Ostrom, J. H., and P. Wellnhofer. 1990. Triceratops: an example of flawed
and systematics. Journal of Vertebrate Paleontology 26:1018–1020. systematics; pp. 245–254 in K. Carpenter and P. J. Currie (eds.), Di-
Forster, C. A. 1996. Species resolution in Triceratops: cladistic and nosaur Systematics: Approaches and Perspectives. Cambridge Uni-
morphometric approaches. Journal of Vertebrate Paleontology versity Press, New York.
16:259–270. Padian, K., J. R. Horner, and J. Dhaliwal. 2004. Species recognition as the
Francillon-Vieillot, H., V. de Buffrénil, F. J. Géraudie, J. Y. Meunier, principal cause of bizarre structures in dinosaurs. Journal of Verte-
L. Sire, L. Zylberberg, and A. de Ricqlès. 1990. Microstructure brate Paleontology 24:100A.
and mineralization of vertebrate skeletal tissues; pp. 471–548 in Parks, W. A. 1925. Arrhinoceratops brachyops, a new genus and species
J. G. Carter (ed.), Skeletal Biomineralization: Patterns, Processes of Ceratopsia from the Edmonton Formation of Alberta. University
and Evolutionary Trends. Van Nostrand Reinhold, New York. of Toronto Studies, Geological Series 19:5–15.
Gilmore, C. W. 1946. Reptilian fauna of the North Horn Formation of Sampson, S. D. 1995. Two new horned dinosaurs from the Upper Cre-
central Utah. U.S. Geological Survey Professional Paper 210C:1–52. taceous Two Medicine Formation of Montana; with a phylogenetic
Goodwin, M. B., W. A. Clemens, J. R. Horner, and K. Padian. 2006. The analysis of Centrosaurinae (Ornithischia: Ceratopsidae). Journal of
smallest known Triceratops skull: new observations on ceratopsid Vertebrate Paleontology 15:743–760.
cranial anatomy and ontogeny. Journal of Vertebrate Paleontology Sampson, S. D., M. J. Ryan, and D. H. Tanke. 1997. Craniofacial on-
26:103–112. togeny in centrosaurine dinosaurs (Ornithischia: Ceratopsidae): tax-
Hatcher, J. B., O. C. Marsh, and R. S. Lull. 1907. The Ceratopsia. US onomic and behavioral implications. Zoological Journal of the Lin-
Geological Survey Monograph 49:1–300. nean Society 121:293–337.
Horner, J. R., and M. B. Goodwin. 2006. Major cranial changes during Scannella, J. B., and D. W. Fowler. 2009. Anagenesis in Triceratops: evi-
Triceratops ontogeny. Proceedings of the Royal Society of London dence from a newly resolved stratigraphic framework for the Hell
B 273:2757–2761. Creek Formation; pp. 148–149 in 9th North American Paleonto-
Horner, J. R., and M. B. Goodwin. 2008. Ontogeny of cranial epi- logical Convention Abstracts. Cincinnati Museum Center Scientific
ossifications in Triceratops. Journal of Vertebrate Paleontology Contributions 3.
28:134–144. Sereno, P. C. 1986. Phylogeny of the bird-hipped dinosaurs (Order Or-
Horner, J. R., and M. B. Goodwin. 2009. Extreme cranial ontogeny in nithischia). National Geographic Society Research 2:234–256.
the Upper Cretaceous dinosaur Pachycephalosaurus. PLoS ONE Sternberg, C. M. 1949. The Edmonton fauna and description of a new
4:e7626. Triceratops from the Upper Edmonton member; phylogeny of the
Horner, J. R., and E. T. Lamm. 2009. Metaplasia provided dinosaur skulls Ceratopsidae. Annual Report of the National Museum of Canada,
extreme morphological plasticity during ontogeny. Journal of Ver- Bulletin 113:33–46.
tebrate Paleontology 29(3, Supplement):117A. Sullivan, R. M., A. C. Boere, and S. G. Lucas. 2005. Redescription of
Horner, J. R., and K. Padian. 2004. Age and growth dynamics of the ceratopsid dinosaur Torosaurus utahensis (Gilmore, 1946) and a
Tyrannosaurus rex. Proceedings of the Royal Society of London B revision of the genus. Journal of Paleontology 79:564–582.
271:1875–1880. Tanke, D. H, and A. A. Farke. 2007. Bone resorption, bone lesions and
Horner, J. R., M. B. Goodwin, and N. Myhrvold. 2009a. Hell Creek For- extra cranial fenestrae in ceratopsid dinosaurs: a preliminary assess-
mation dinosaur census reveals abundant Tyrannosaurus; p. 146 in ment; pp. 319–347 in K. Carpenter (ed.), Horns and Beaks: Ceratop-
9th North American Paleontological Convention Abstracts. Cincin- sian and Ornithopod Dinosaurs. Indiana University Press, Bloom-
nati Museum Center Scientific Contributions 3. ington and Indianapolis, Indiana.
Horner, J. R., M. B. Goodwin, and H. N. Woodward. 2007. Synonomy Tokaryk, T. T. 1986. Ceratopsian dinosaurs from the Frenchman For-
consequences of dinosaur cranial ontogeny. Journal of Vertebrate mation (Upper Cretaceous) of Saskatchewan. Canadian Field-
Paleontology 27(3, Supplement):92A. Naturalist 100:192–196.
1168 JOURNAL OF VERTEBRATE PALEONTOLOGY, VOL. 30, NO. 4, 2010

Tumarkin-Deratzian, A. R. 2009. Evaluation of long bone surface tex- Wolfe, D. G., and J. I. Kirkland. 1998. Zuniceratops christopheri n.
tures as ontogenetic indicators in centrosaurine ceratopsids. The gen. & n. sp., a ceratopsian dinosaur from the Moreno Hill For-
Anatomical Record 292:1485–1500. mation (Cretaceous, Turonian) of west-central New Mexico; pp.
Ukrainsky, A. S. 2007. A new replacement name for Diceratops 303–317 in S. G. Lucas, J. I. Kirkland, and J. W. Estep (eds.),
Lull, 1905 (Reptilia: Ornithischia: Ceratopsidae). Zoosystematica Lower and Middle Cretaceous Terrestrial Ecosystems, Volume 14.
Rossica 16:292. New Mexico Museum of Natural History and Science Bulletin,
Ukrainsky, A. S. 2009. Synonymy of the genera Nedoceratops Ukrainsky, Albuquerque.
2007 and Diceratus Mateus, 2008 (Reptilia: Ornithischia: Ceratopsi-
dae). Paleontologicheskii Zhurnal 1:108. Submitted May 8, 2009; accepted December 24, 2009.

APPENDIX 1. Sample of Triceratops cranial material collected by the MOR, the majority having been collected since 1999.

Specimen number Material Specimen number Material

MOR 004 Articulated skull MOR 2436 Partial disarticulated skull
MOR 335 Parietal MOR 2551 Partial skull
MOR 539 Partial skull MOR 2552 Partial skull
MOR 699 Disarticulated skull MOR 2569 Disarticulated skull
MOR 965 Partial skull MOR 2570 Partial skull
MOR 981 Articulated skull MOR 2572 Squamosal
MOR 989 Nasal horn, rostral MOR 2574 Disarticulated skull
MOR 1098 Postorbital horn core MOR 2576 Postorbital horn core
MOR 1110 Disarticulated skull MOR 2589 Postorbital horn core
MOR 1120 Disarticulated skull MOR 2590 Premaxilla
MOR 1122 Articulated skull MOR 2597 Partial skull
MOR 1199 Disarticulated skull MOR 2702 Partial skull
MOR 1604 Articulated skull MOR 2923 Articulated skull
MOR 1625 Partial skull MOR 2924 Disarticulated skull
MOR 2927 Squamosal, postorbital horn core MOR 2958 Postorbital horn core
MOR 2928 Squamosal, dentary MOR 2959 Partial skull
MOR 2929 Squamosal MOR 2969 Parietal
MOR 2937 Maxilla MOR 2970 Squamosal
MOR 2938 Partial disarticulated skull MOR 2971 Partial disarticulated skull
MOR 2942 Squamosal MOR 2972 Partial skull
MOR 2945 Dentary MOR 2975 Partial skull
MOR 2946 Parietal, squamosal MOR 2979 Partial skull
MOR 2950 Partial skull MOR 2980 Squamosal
MOR 2951 Disarticulated skull MOR 2982 Partial disarticulated skull
MOR 2952 Partial skull MOR 2984 Partial skull