ArticlePDF Available

Abstract and Figures

The discovery of the smallest Triceratops skull (UCMP 154452) provides a new ontogenetic end member for the earliest stage of ceratopsid (Centrosaurinae plus Chasmosaurinae) cranial development. The lack of co-ossification among the parietal, squamosals, postorbitals, quadratojugal arch, and the braincase preserves sutural contacts and bone surfaces that later become obscured in adults. The ability to document the early development and morphology of the horns and frill in Triceratops allows a reevaluation of their functional roles. UCMP 154452 shows that the cranial ornamentation of the frill and the postorbital horns were not restricted to adults, but began at an early age in this species. This evidence supports the hypothesis that the function of ceratopsid horns and frills was potentially important for visual communication and species recognition because in this young form it could not have functioned in sexual display. Although some features of UCMP 154452 anticipate or mimic the adult character states, some braincase characters recapitulate the juvenile and adult stages of more basal neoceratopsians.
Content may be subject to copyright.
Museum of Paleontology, University of California, Berkeley, CA 94720-4780,;
Department of Integrative Biology and Museum of Paleontology, University of California, Berkeley, California 94720-3140,;
Museum of the Rockies, Montana State University, Bozeman, Montana 59717-0040,
ABSTRACT —The discovery of the smallest Triceratops skull (UCMP 154452) provides a new ontogenetic end member
for the earliest stage of ceratopsid (Centrosaurinae plus Chasmosaurinae) cranial development. The lack of co-ossification
among the parietal, squamosals, postorbitals, quadratojugal arch, and the braincase preserves sutural contacts and bone
surfaces that later become obscured in adults. The ability to document the early development and morphology of the
horns and frill in Triceratops allows a reevaluation of their functional roles. UCMP 154452 shows that the cranial
ornamentation of the frill and the postorbital horns were not restricted to adults, but began at an early age in this species.
This evidence supports the hypothesis that the function of ceratopsid horns and frills was potentially important for visual
communication and species recognition because in this young form it could not have functioned in sexual display.
Although some features of UCMP 154452 anticipate or mimic the adult character states, some braincase characters
recapitulate the juvenile and adult stages of more basal neoceratopsians.
Triceratops is one of the most familiar genera of Late Creta-
ceous dinosaurs; it is recognized by its distinctive skull, with
three horns and massive frill made up of the parietal and paired
squamosals. Previous assessments of ontogeny in Triceratops are
based on an isolated juvenile postorbital horn (ca. 100 mm long)
described over 60 years ago by Brown and Schlaikjer (1940a)
from the Hell Creek Formation, Montana. Two supraorbital
horn cores (95 mm and 65 mm long) from the Frenchman For-
mation of Saskatchewan were described by Tokaryk (1997) but
could not be identified beyond Chasmosaurinae. Here we report
the discovery of the smallest Triceratops skull, UCMP 154452,
from the upper Hell Creek Formation (Maastrichtian), Garfield
County, Montana. This new skull is identified as Triceratops by
the presence of two 35-mm-long postorbital horns (outgrowths
of the postorbital bones) and a highly scalloped, unfenestrated
frill (Fig. 1).
This diminutive Triceratops skull is a mere 30 cm long and is
the smallest ceratopsid skull known. Like the young of many
other kinds of dinosaurs (Carpenter et al., 1994), UCMP 154452
has large orbits relative to skull size and a foreshortened face.
The next smallest Triceratops skull is of a subadult over four
times as long (Schlaikjer, 1935), and adult skulls are six to seven
times longer (Hatcher, 1907). UCMP 154452 brings the known
growth series of Triceratops to a new small extreme and shows
that cranial ornamentation in the frill and the postorbital horns
were not restricted to adult members, but began at an early age.
UCMP 154452 provides important information on the morphol-
ogy and development of the horns and frill in Triceratops and
allows a reevaluation of their functional significance. A compre-
hensive assessment of Triceratops ontogeny based on a very
complete cranial growth series in the collections of the MOR and
UCMP (Goodwin and Horner, 2001) will follow this study (Hor-
ner and Goodwin, pers. observ.).
UCMP 154452 was discovered in strata of the Hell Creek For-
mation exposed in a small badland area located just north of the
divide separating the drainages of Snow Creek, to the north, and
Hell Creek (UCMP locality V97006, Garfield County, Mon-
tana). The skull was preserved in a bed of essentially unstratified,
medium gray siltstone that weathers to light gray. Yellow, fer-
ruginous streaks and globules as well as fragmentary plant re-
mains occur throughout the sediment. Teeth and/or skeletal frag-
ments of Tyrannosaurus,Triceratops, and Meniscoessus cf. ro-
bustus were discovered in outcrops in the immediate vicinity of
the quarry at the same or slightly (ca. 2 m) higher stratigraphic
levels and document the latest Cretaceous age of the locality
(Lancian North American Land Mammal Age). The nearest ex-
posures of the contact of the Hell Creek and overlying Tullock
formations are approximately one mile (1.6 km) to the west and
2.5 miles (4.0 km) to the southeast. In both, the contact between
these formations is at an elevation of ca. 2860 feet (875.2 m). The
current elevation of V97006 is ca. 2770 feet (847.6 m). In the
region of V97006 the strata of these formations appear to be
essentially flat lying. The difference in current elevations of the
formational contact and the fossil locality, ca. 90 feet (27.5 m),
suggests that V97006 is within the upper third of the Hell Creek
Formation, which is approximately 300 feet (91.8 m) thick in the
valley of Hell Creek (see Wilson, 2004).
Institutional AbbreviationsMOR, Museum of the Rockies,
Bozeman, Montana; UCMP, University of California Museum
of Paleontology, Berkeley; USNM, United States National Mu-
seum, Washington, D.C.; UTEP, University of Texas at El Paso;
YPM, Yale Peabody Museum, Connecticut.
The individual cranial elements of UCMP 154452 share an
external bone texture that is striated and very porous, indicative
of fast-growing tissue (Sampson et al., 1997). All sutures are
patent and allow accurate articulation of this very young Tri-
ceratops skull. This early phase of cranial morphogenesis pre-
serves sutural contacts and bone surfaces that become hidden in
adults. The right side of the skull is more complete and the
following elements, from the right side unless noted otherwise,
are preserved: parietal, left and right squamosals, left and right
postorbitals, prefrontal, jugal, quadrate, quadratojugal, occipital
condyle, basioccipital, left and right exoccipitals, surangular, and
dentary. Less complete but identifiable fragments of the maxil-
Journal of Vertebrate Paleontology 26(1):103–112, March 2006
© 2006 by the Society of Vertebrate Paleontology
lary, left quadrate, left jugal, vertebral centra, ossified tendons,
and teeth were also found with the skull. Morphological descrip-
tions are based on the right side.
The parietal is nearly square(Figs. 2; 3B, E). It measures
124 mm in length along the midline and has a maximum width of
127 mm. The midline is ornamented by an undulating row of five
raised bony prominences. This feature was also described by
Dodson and Currie (1988) on a 210-mm-long parietal of the
previously smallest known ceratopsid, tentatively referred to
Monoclonius. Rostrally, in UCMP 154452, each prominence be-
comes progressively narrower along the midline, but remains
consistent in height, ca. 5 mm, above the surface of the parietal.
Caudally, the parietal is about 57 mm thick and thins rostrally
to less than 4 mm. This rostral thinning of the parietal is also
observed in adult Triceratops skulls (Dodson and Currie, 1990).
A prominent feature of UCMP 154452 is the scallopedcaudal
margin of the parietal. A series of three scallops on either side of
the parietal midline is bordered caudally by a central scallop.
These scallops continue onto the caudal margin of each squamo-
sal and provide a distinctive appearance to the frill. These scal-
lops are not separate ossifications but are formed by the parietal
and squamosal. Consequently, they are not homologous with the
epoccipitals that border the frill in subadult and adult Tricera-
tops, but merely mimic their shape. The dorsal and ventral sur-
faces of each scallop are smoother than the surrounding bone
and were likely covered by hard keratin before the epoccipitals
ossified. This scalloped edge becomes less pronounced and gent-
ly wavyin subadult Triceratops skulls when epoccipitals first
appear and ossify along the frill margin (Goodwin et al., 1997).
The left and right squamosals (Fig. 3A, C, D, F) are nearly
complete and are ca. 150 mm in maximum length. The squamosal
thins rostrally from 6.3 mm to 3.0 mm. It articulates rostrolater-
ally with the jugal by an overlapping sutural contact. It also
overlaps the caudal portion of the postorbital rostrodorsally. The
medial edge of the squamosal forms the border for the supra-
temporal fenestra. The squamosal and parietal contribute to the
frill along a fairly straight contact. The characteristic inward
bend of the adult squamosal is expressed in the squamosals of
UCMP 154452. This bend becomes greatly exaggerated in adult
Triceratops (Dodson and Currie, 1990; Dodson, 1993). The me-
dial edge of the squamosal curves slightly and does not appear to
overlap with the parietal rostrally as in subadult and adult skulls.
The caudal border of the squamosal has five distinct scallops. A
longitudinal series of raised prominences radiate rostrolaterally
onto the dorsal surface of the postorbitals. Ventrally, a bifurcat-
ing prominent ridge of bone serves as the articular surface for the
exoccipital and quadrate where these bones form a prominent
buttress beneath the frill (Fig. 4B).
The most distinctive feature of the left and right postorbitals is
the 35-mm-long postorbital horns (Fig. 5). The postorbital horns
FIGURE 1. A comparison of UCMP 154452, the smallest Triceratops
skull known, with an adult Triceratops skull in right lateral view, illus-
trates the dramatic changes in size, shape, and sutural contacts of cranial
elements that occur during ontogeny. A, restoration of UCMP 154452,
Triceratops. The frill is highly scalloped at this very young stage but is
minimally developed caudally and not fan-like compared with the adult
skull. Postorbital horn growth has already begun. The nasal region is
restored after the adult condition; the fenestra may not be present in
young individuals. B, YPM 1822, an adult Triceratops skull (1.75 m long;
modified from Romer, 1966). Abbreviations (restored bones drawn in
outline only): d, dentary; j, jugal; lac, lacrimal; m, maxilla; n, nasal; nh,
nasal horn; p, parietal; po, postorbital; poh, postorbital horn; pd, pre-
dentary; pfr, prefrontal; pm, premaxilla; q, quadrate; qj, quadratojugal; r,
rostral; sa, surangular; and sq, squamosal.
FIGURE 2. Parietal of UCMP 154452, Triceratops, in dorsal view.
Note the scalloped caudal margin and median row of bony ornamenta-
are oriented rostrally at ca. 10°and do not show the caudally
directed curvature of older juvenile and subadult Triceratops
postorbital horns (Goodwin et al., 1997). Indented vascular
grooves on the exterior surface of the postorbitals indicate that
they were covered by a keratinous sheath (Horner and Marshall,
2002). The grooves are deepest on the surface of the horns. A
row of raised prominences radiates caudolaterally and continues
onto the squamosals. The right postorbital confirms that the ven-
tral part of the cornual sinus at the basal region of the horncore
formed early (Fig. 5B). The rostral face of the postorbital horn
(Fig. 5C) has a prominent rugose sutural surface for the prefron-
tal. This roughened sutural surface covers nearly the entire ros-
tral surface of the postorbital horn. Laterally, the caudodorsal
region continues as a thin wedge of bone that articulates with the
squamosal by sliding beneath its rostral edge.
We identify a 39.6 mm long semi-lunate bone as the right
prefrontal (Fig. 6). It is ca. 7 mm thick. The medial and caudal
edges are dominated by a rugose sutural surface. Caudally, the
prefrontal thickens where it meets the rostral face of the post-
orbital. Three prominent foramina are on the anterodorsal sur-
face and a single central foramen penetrates the prefrontal. The
prefrontal forms the anterodorsal margin of the orbit and en-
ables the reconstruction of the front of the skull.
The jugal (Fig. 7A, B) forms the ventral border of the orbit
and the dorsal rim of the lateral temporal fenestra. This sutural
pattern of the jugal bordering the rim of the lateral temporal
fenestra dorsally is considered plesiomorphic by Forster (1996a:
261, character 1; also see Fig. 1). In the derived state, the squa-
mosal forms the dorsal rim of the lateral temporal fenestra and
does not extend across the top of the jugal in adult Triceratops.
In UCMP 154452, the squamosal remains excluded from most of
the dorsal margin of the lateral temporal fenestra by a caudally
directed jugal spur. Individual variation may be the cause of this
slight anatomical difference, but it could also be an expression of
the primitive condition retained in the adult Triceratops and at
this early stage of ontogeny. In all ceratopsids, the jugal expands
caudally and the squamosal enlarges caudoventrally. As a result,
the infratemporal fenestra is compressed and reduced in size.
The jugal-squamosal contact excludes the postorbital from the
infratemporal fenestra in adult Triceratops (Dodson and Currie
1990:600). This is also observed in UCMP 154452. A squamosal
process ventral to the temporal opening is undeveloped at this
early stage of ontogeny. Deep vesicle grooves are present around
the ventral rim of the orbit and become shallower on the remain-
ing jugal dorsally. A prominent feature of the jugal is the ven-
trally directed wedgeof bone that covers the anterolateral
surface of the quadratojugal along an overlapping sutural con-
FIGURE 3. The fill of UCMP 154452, Triceratops, in dorsal (AC) and ventral (DF) views. Right squamosal (A,F), parietal (B,E), and left
squamosal (C,D). The caudal margin of the frill is highly scalloped.
tact. Ventrally, the sutural surface of the jugal is slightly concave
and thinner (ca. 4 mm) compared to 6 mm in overall thickness.
The distal tip of the jugal is flared but does not show any evi-
dence for an epijugal, which evidently forms later in ontogeny.
The quadrate (Fig. 8) is 98 mm in maximum length dorsoven-
trally and is nearly totally excluded from the caudal margin of
the infratemporal fenestra by the overlapping quadratojugal.
The quadrate articulates with the quadratojugal along a vertical
axis laterally. This articulation is further supported by a spur of
bone that rises from the ventral sutural surface of the quadrate to
meet a gentle depression on the lower medial surface of the
quadratojugal. The quadrate expands transversely dorsally, a
precursor to the significant dorsal expansion in adult Triceratops
(Forster, 1996b). The flange of bone meets the underside of the
squamosal along a pronounced Vshaped ridge of bone (see Fig.
4B). It is overlain by the squamosal and underlain by the exoc-
cipital. This arrangement is also observed in adult skulls
(Hatcher, 1907; Ostrom and Wellnhofer, 1986). The surface for
articulation with the lower jaw lies at the rostral-most region of
the quadrate. It is robust and round, but does not form a trans-
verse articular surface or distinct double condyle, bisected by a
trough, for articulation with the mandible of adults (Hatcher,
1907; Ostrom and Wellnhofer, 1986). This ventral condyle is
rostral to the more caudally directed dorsal flange.
The quadratojugal (Fig. 8) is wedged between the quadrate
and the overlapping ventral jugal flange. The quadratojugal is
thick ventrally, thin dorsally, and wrapped around the caudal
portion of the quadrate. The medial ventral condyle is slightly
concave where it contacts the quadrate. Laterally, the quadrato-
jugal is marked by a ridge and faceted surface where the jugal
flange overlaps it. This arrangement is consistent with the adult
condition (Ostrom and Wellnhofer, 1986). All of the sutural sur-
faces are open and overlapping. The quadratojugal forms nearly
the entire ventral and caudal border of the lateral temporal fe-
nestra. In adult Triceratops, the squamosal forms this caudal
border of the infratemporal fenestra (Dodson and Currie, 1990).
In lateral view the quadratojugal is largely obscured by the jugal
in most adult skulls but not in UCMP 154452, particularly adja-
cent to the infratemporal opening.
Lateral Temporal Fenestra
The lateral temporal fenestra (Fig. 1) lies beneath and caudal
to the orbit. The opening is nearly oval and bordered by the jugal
FIGURE 4. Right squamosal of UCMP 154452, Triceratops, in dorsal
(A) and ventral (B) views. Note the well-developed transverse buttress
for articulation of the exoccipital and quadrate and support of the fill.
Abbreviations:b, buttress; ex, exoccipital articulation; q, quadrate
FIGURE 5. Postorbitals of UCMP 154452, Triceratops. Right postorbital in lateral (A,E), medial (B,F), and rostral (C,G) views; left postorbital
in lateral (D,H) view. Postorbital horns are present at this early stage of ontogeny. Indented vascular grooves cover their dorsal surface. Formation
of the corneal sinus is evident at the base of the horncore ventrally in (B). A rugose sutural surface for the prefrontal is preserved on the rostral face
of the right postorbital horn in (C). Abbreviations:pf, sutural surface for the prefrontal; sq, sutural surface for the right squamosal.
rostrally, the squamosal caudodorsally, and by the quadratojugal
The braincase of UCMP 154452 (Fig. 9) is well preserved and
reveals the substantial amount of ontogenetic transformation
that takes place into adulthood. The bones of the braincase are
unfused and are loosely coalesced by overlapping rugose or
tongue-and-groove sutures. Important differences with the adult
ceratopsid braincase in the arrangement and articulation of the
basioccipital, exoccipitals, and supraoccipital are noted below.
Occipital CondyleThe occipital condyle is 31.9 mm medio-
laterally and 28.9 mm vertically. It is nearly oval and unfused and
is formed by a one-third contribution from the basioccipital and
one-third from each ventral exoccipital (Fig. 9). This arrange-
ment is typical of all ceratopsids but is often obscured by fusion
in adult skulls (Lehman, 1989; Chinnery, 2004). The dorsal su-
tural surface is irregularly grooved where it accepts the exoccipi-
tal. The short pedunculate neck of the condyle is constricted
dorsoventrally and transversely.
BasioccipitalThe basioccipital is 48 mm long and 26.5 mm
wide. Caudally it is dominated by its contribution to the occipital
condyle. A median ridge divides the basioccipital. A distinctive
midline hourglass-shaped ridge and grooved sutural surface for
the alisphenoid are preserved. The rostrodorsal surface is rugose
where it articulates with the descending process of the exoccipi-
tal. Well-developed basioccipital tubera extend rostrolaterally.
The basioccipital is excluded from the foramen magnum by the
paired exoccipitals. A small portion of the left basisphenoid is
preserved in close contact along the rostroventral surface of the
basioccipital. The foramen magnum is 23.6 mm wide.
FIGURE 8. Right quadrate and quadratojugal of UCMP 154452, Tri-
ceratops, in caudal view. The quadratojugal overlaps the lateral quadrate
along a close-fitting, broad, sutural contact. Abbreviations:c, quadrate
condyle articulates with the surangular; j, sutural surface for the over-
lapping jugal spur; q, quadrate; qj, quadratojugal.
FIGURE 6. Right prefrontal of UCMP 154452, Triceratops. Right pre-
frontal in dorsal (A), ventral (B), lateral (C), and medial (D) views.
Postorbital and frontal sutural surfaces are heavily rugose. Abbrevia-
tions:f, sutural surface for frontal; po, sutural surface for postorbital.
FIGURE 7. Right jugal of UCMP 154452, Triceratops, in lateral (A)
and medial (B) views. Abbreviations:qj, sutural surface for the quadra-
tojugal; sq, sutural surface for the squamosal.
SupraoccipitalIn all adult ceratopsids, the supraoccipital is
excluded from the foramen magnum by the exoccipitals, which
unite above the foramen magnum (Hatcher, 1907; Dodson and
Currie, 1989; Forster, 1996b). UCMP 154452 does not share this
condition. At this very early stage of ontogeny, UCMP 154452
exhibits the primitive state in which the supraoccipital articulates
between the exoccipitals dorsomedially (see Fig. 9). This ar-
rangement is seen in the more basal Protoceratops andrewsi
(Brown and Schlaikjer, 1940; Dodson and Currie, 1989; Hailu
and Dodson, 2004). Contribution of the supraoccipital to the
foramen magnum also occurs in the neoceratopsian Leptocera-
tops gracilis (Sternberg, 1951) and Bagaceratops rozhdestvenskyi
(Maryanska and Osmólska, 1975). Chinnery and Weishampel
(1998) confirm that the paired exoccipitals meet above the fora-
men magnum and exclude the supraoccipital from any contribu-
tion to the foramen magnum in a juvenile braincase of Montan-
aceratops cerorhynchus (MOR 542). Fusion of the supraoccipital
with the exoccipitals obscures this arrangement in an isolated
braincase of M. cerorhynchus (AMNH 5244) described by Ma-
kovicky (2001). In UCMP 154452, the supraoccipital is posi-
tioned firmly between the exoccipitals and forms the dorsal roof
of the foramen magnum and the rostral endocranial cavity. Later
in ontogeny, the supraoccipital is pushed upward and becomes
completely underlain by the exoccipitals along a broad sutural
contact (Dodson et al., 2004) in adult Triceratops.
The caudal surface of the supraoccipital is abraded and broken
sagittally. A ridge of bone forms a midline crest. Small depres-
sions are present on either side of the crest. This area serves as
the attachment site for the epaxial musculature of the neck and
is significantly more developed in adult ceratopsids, where these
deep, paired depressions are separated by a thin vertical septum
of bone (Hatcher et al. 1907; Ostrom and Wellnhofer, 1986). The
ventral surface of the supraoccipital is rugose along the sutural
surface for articulation with the exoccipitals. The roof of the
braincase dominates the internal surface. The entrance and exit
foramina for the auditory nerve are preserved.
ExoccipitalThe exoccipitals extend as a wingof bone lat-
erally from the foramen magnum, forming a buttress that con-
tacts the ventral surface of the squamosal (see Fig. 4B). Accord-
ing to Ostrom and Wellnhofer (1986:123), this configuration pro-
vides major support for the entire frill along this junction of the
quadrate and squamosal contact. This arrangement remains con-
sistent into adulthood. The right exoccipital is more complete
than the left and preserves the relatively large exit foramina for
cranial nerves IXXI and XII with a septum of bone between the
foramina (Fig. 9).
The right surangular (Fig. 10) is 62.5 mm in length. It varies in
thickness from 3.5 mm at the most rostral edge where it meets
the dentary to 10.7 mm caudally along the articular surface for
the quadrate. Rostromedially, the surface is striated where it
articulates with the dentary. The mandibular foramen is pre-
served along the upper portion of the dorsal surface. The suran-
gular is laterally convex and slightly concave medially. The sur-
angular thickens and develops a flat curved shelf caudally. This
shelf is deflected laterally, flattened dorsally, and articulates with
the quadrate. Caudomedially, the surangular is sharply concave
where it meets the articular.
The dentary (Fig. 11) is 160 mm long and allows determination
of the maximum skull length and restoration of the skull in
Fig. 1. The coronoid process is robust, offset laterally, and curves
rostrally. Ostrom and Wellnhofer (1989) interpreted the sturdy
coronoid process as a critical attachment site for powerful ad-
ductor musculature. The dentary is straight, convex laterally and
concave medially except for the dental battery, which is nearly
vertical. The rostral edge is relatively thicker (6.510.4 mm)
where it meets the predentary bone. Caudally it is 4.4 mm thick.
Medially, the symphyseal surface is indicated by horizontal stria-
tions on the thickened bony facet. Vertically, the coronoid pro-
cess is offset about 30 degrees to the axis of the dentary. A strong
lateral ridge runs along the length of the dentary ventrolaterally.
Below this ridge, the ventral surface of the dentary is flattened
and striated where the splenial would lie longitudinally. A deep
adductor fossa is present caudomedially below the coronoid pro-
cess. Even at this young age, the fossa is relatively large and
sufficient as a major insertion site for the M. adductor posterior
(Ostrom and Wellnhofer, 1989). The Meckelian groove extends
ca. 48 mm along the rostromedial surface of the adductor fossa
from the caudal edge of the dental battery. A shallow longitu-
dinal groove lies along the ventral border of the fossa caudally,
indicating the place of attachment for the M. intramandibularis
(Ryan and Currie, 1998). The dental battery is 107 mm long and
FIGURE 10. Right surangular of UCMP 154452, Triceratops,inob-
lique view. Abbreviations:ar, articular articulation; d, dentary articula-
tion; mf, mandibular foramen; q, quadrate articulation.
FIGURE 9. Braincase of UCMP 154452, Triceratops, in occipital view.
In all ceratopsids, the supraoccipital is excluded from the foramen mag-
num as the exoccipitals unite above the foramen magnum, but not at this
early ontogenetic stage in UCMP 154452. Here, UCMP 154452 exhibits
the primitive state: the supraoccipital articulates between the exoccipitals
dorsomedially. The occipital condyle is formed by a one-third contribu-
tion from the basioccipital and one-third from each exoccipital. Exit
foramen for cranial nerves IXXI and XII are noted. Abbreviations:bo,
basioccipital; exo, exoccipital; fm, foramen magnum; oc, occipital con-
dyle; so, supraoccipital.
occupies a large extent of the dentary. At least 20 alveoli are
present along the length of the dental battery. No teeth are
preserved in the jaw. Laterally, the dentary bears two rows of
foramina. According to Lehman (1989), these foramina probably
communicate with the mandibular fossa, carrying branches of
the mandibular ramus of the trigeminal nerve, the mandibular
artery, and veins to the tissues of the cheek and predentary.
Rostrally, an upper third row of foramina is present.
A 55.3-mm-long left maxillary fragment was recovered with
the skull. Seven dental grooves are preserved on the interior
surface. No maxillary teeth are preserved in position. The dental
magazine is relatively delicate, and the bone becomes more ro-
bust along the caudodorsal margin.
Isolated leaf-shaped teeth were found associated with the
skull. They are double-rooted with a strong median ridge and a
lingual covering of enamel.
At least three isolated, fragmentary vertebrae were recovered
with the skull. Fragments of their centra reveal a very spongy
interior surrounded by a relatively thin periosteal exterior.
Toothed sutures remain open on the dorsal surface of the centra,
indicating that the neural spines are unfused. This is not un-
expected in such a small, young individual.
Ossified Tendons
Fragments of ossified tendons were closely associated with the
skull. The largest piece is 38.1 mm long and ca. 4 mm in diameter.
The medial surface is incompletely striated and the remaining
surface is smooth. One fragment is triangular and broad crani-
ally, like the adult tendon.
In adult Triceratops, the large, solid, saddle-shaped frill is 65
70% of the basal skull length (caudal surface of basioccipital
to tip of rostral bone; Forster, 1996b). In UCMP 154452, the frill
is only 48% of the estimated basal skull length. Although it
already has small postorbital horns and a solid, scalloped frill
that closely resembles the epoccipital-bordered adult skull,
UCMP 154452 had proportions very different from those of the
adult skull. The parietosquamosal frill is short and square,
whereas in adults it is elongate, fan-like, and more concave. Fea-
tures of the braincase in UCMP 154452 recall the adult condition
of more basal neoceratopsians (Brown and Schlaikjer, 1940b).
Here, the exoccipitals unite below but not above the foramen
magnum, allowing the supraoccipital to form the dorsal margin
of the foramen magnum and contribute to the roof of the brain-
case (Fig. 9). This condition is shared with Protoceratops, even as
adults (Brown and Schlaikjer, 1940b, Dodson and Currie, 1990;),
but is lost in adult Triceratops where the exoccipitals unite above
the foramen magnum, excluding the supraoccipital (Hatcher et
al., 1907; Brown and Schlaikjer, 1940b). In this sense, some ju-
venile features of Triceratops recapitulate a character state of
more basal neoceratopsians as might be expected. Evidence pre-
FIGURE 11. Right dentary of UCMP 154452, Triceratops, in lateral (A) and medial (B) views. The dental battery occupies a large area of the lower
jaw. Abbreviations:a, alveoli; af, adductor fossa; c, coronoid process; d, dentary; pd, predentary articulation; and sym, symphyseal surface.
sented by Gilmore (1917) and more recently by Lehman (1989)
suggests that this arrangement of the supraoccipital may be a
juvenile ceratopsid character. Gilmore (1917:fig. 11) determined
that the supraoccipital of the type of Brachyceratops montanensis
(USNM 7951) contributes to the formation of the foramen mag-
num in this immature ceratopsid. Lehman (1989:fig. 6B, C) ob-
served that the supraoccipital forms at least the caudal roof of
the endocranial cavity in an incomplete juvenile braincase re-
ferred to Chasmosaurus mariscalensis (UTEP P.37.7.068).
In contrast, certain juvenile features appear to anticipate the
structural condition of the adult Triceratops skull. First, although
no teeth were preserved in place in the dentary or the small
fragment of the maxillary, isolated teeth recovered from the sedi-
ment around the tiny skull share the crown pattern and double-
rooted form of adults (contra Carpenter, 1982). Second, the
lateral wings of the exoccipital expand from either side of the
occipital condyle into broad flanges that contact the ventral sur-
face of each squamosal. As in adults, this portion of the exoc-
cipital forms an expansive brace that provides primary support
for the overlying frill (Ostrom and Wellnhofer, 1986; Forster,
1996b). Third, a parasagittal row of low bosses ornaments the
superior surfaces of the postorbitals, squamosals, and the pari-
etal midline. The number of scallops on the caudal margin of the
squamosals and parietal equals the number of epoccipitals that
border the adult frill (Hatcher, 1907; Forster, 1996b). If the ep-
occipital-ornamented adult frill and postorbital horns served as
an important visual sign in species communication, then perhaps
the scalloped frill and horns of very young individuals served the
same function.
It has often been suggested that the ornamental skull features
of ceratopsids (horns and frills) reflect a role in mate competition
or species recognition (Forster and Sampson, 2002). Dimorphism
can be recognized in adequate population samples of taxa by a
divergence of biometric characters during ontogeny (Darwin,
1871). Among non-avian dinosaurs, sexual dimorphism has been
suggested in theropods (Colbert, 1990; Raath, 1990), hadrosau-
rids (Dodson, 1975; Hopson, 1975; Molnar, 1977), and ceratop-
sians (Ostrom and Wellnhofer, 1986; Dodson, 1996; Forster,
1996b) but in each case it is not extreme and has not been dem-
onstrated statistically. Only the basal ceratopsian Protoceratops
shows statistically significant dimorphism (Dodson, 1976), but it
is minor. Sexual dimorphism has been inferred for various cen-
trosaurines and chasmosaurines (Lehman, 1990; Forster, 1996b;
Sampson et al., 1997; Ryan et al., 2001), but again, this variation
has not been established statistically and has not been differen-
tiated from ontogenetic or within-normal-populational variation
(Padian et al., 2004). Dimorphism can be expressed early in on-
togeny, or as a late pulse that reflects maturity and agonistic
sexual behavior, usually among males, resulting in an extended
growth trajectory (Weckerly, 1998). This by itself does not con-
firm sexual display or associated mating behavior as the principal
function of low-level dimorphism; the morphology of horns and
frills may have served different functions at different times in an
individuals life. We suggest visual communication and species
recognition, perhaps involving complex signaling (Ord et al.,
2001) as an alternative but not exclusive function of these cranial
In general, dinosaurian cranial display features, such as horns,
spikes, and bony pads in ceratopsids (Forster, 1996a, Sampson
et al., 1997), crests on hadrosaurid skulls (Horner and Currie,
1994), and the frontoparietal domes of pachycephalosaurids
(Goodwin et al., 1998; Williamson and Carr, 2002; Goodwin and
Horner, 2004) did not appear until later stages of development.
Immature centrosaurines of different genera have similar horn-
core ontogenies (Sampson et al., 1997); adult features of horns
and frills appear only late in ontogeny, suggesting a function in
sexual display or species/mate recognition. Sexual dimorphism
has also been inferred for some chasmosaurines (Dodson, 1996),
but is not generally accepted for either chasmosaurines or cen-
trosaurines (Dodson et al., 2004). The new tiny Triceratops
shows that the normally late-developing features often associ-
ated (if dimorphic) with sexual display began to be expressed at
a very early age. This pattern appears likely for Chasmosaurus
(Lehman, 1989; 1990), though at relatively larger size and pre-
sumably later age. This clearly derived condition within chasmo-
saurines, given current knowledge of ceratopsian ontogeny and
phylogeny, suggests a heterochronic shift of the expression of
these characters. However, hypotheses of heterochrony can only
be tested by comparative ontogenies, which at present are insuf-
The basal neoceratopsian Protoceratops expresses some fea-
tures, such as a nasal boss and a vertical tilt to the frill, only late
in life, suggesting that the centrosaurine pattern may be primi-
tive. However, specimens referred to Zuniceratops (Wolfe and
Kirkland, 1998) appear to show long-developing horns that are
present in juveniles, like chasmosaurines but unlike centrosau-
rines. The phylogenetic placement of cf. Zuniceratops outside
Ceratopsidae (see Dodson et al., 2004) suggests that the function
of cranial structures in species recognition may have preceded
the divergence of centrosaurines and chasmosaurines. If so, then
the late-developing structures of centrosaurines would be de-
rived, heterochronically shifted features and could perhaps be
exaptively linked to sexual display (Lehman, 1990; Ryan et al.,
2001), if significant dimorphism can be established.
A function in sexual display (Farlow and Dodson, 1975; Mol-
nar, 1977; Sampson, 1997) or resisting predators (Colbert, 1948,
1961; Molnar, 1977) has long been the dominant model for cra-
nial ornamentation in dinosaurs, despite little evidence for sex-
ual dimorphism. However, it is difficult to support a hypothesis
of sexual display when the sexes show little or no evidence of
discrete morphological features apart from size (Darwin, 1871);
even so, hypotheses of a function in sexual display must be tested
by evidence beyond simple morphologic difference. Conversely,
species recognition is simply tested by the presence of low sexual
dimorphism with species-specific morphology that is apparent to
intra- and interspecific individuals (Vrba, 1984). Non-directional
morphologic trends in phylogeny and the presence of several
related sympatric or parapatric species are two tests of species
recognition as a factor that structures morphological diversity
(Padian et al., 2004). Ceratopsids pass these tests.
Low sexual dimorphism in ceratopsians supports our hypoth-
esis that the early ontogenetic expression of horn and frill mor-
phology in some ceratopsids reflects a visual cue for communi-
cation and species recognition; clearly these features appeared
well before sexual maturity in UCMP 154452. Extant African
bovids use an effective visual communication system that in-
volves horn morphology and body color (Vrba, 1984). These
explicit species differences have long histories of divergence and
sorting in bovid subclades (Vrba, 1984). Similar visual cues based
on horn and frill morphology may have stimulated greater spe-
cies diversity earlier in ceratopsian evolutionary history. Farlow
and Dodson (1974), Hopson (1975), and Sampson (1997) have
acknowledged the potential importance of species recognition in
dinosaur diversity.
The many forms of sexual dimorphism among birds are con-
ventionally split into body size and overall plumage-color attrib-
utable to melanins, carotenoids, and structural colors (Owens
and Hartley, 1998). Structural colors arise from the scattering of
ultraviolet light by collagen fibers. Prum et al. (1994, 1999) report
how skin color in the face and head is used by an assortment of
extant birds for visual communication. Deep vesicle grooves in
the horns and frill of Triceratops indicate that their skull was
covered with hard, keratinous skin (Horner and Marshall, 2002).
Paired with the visually dominating frill and horns, skin color
may have enhanced intraspecific communication. We propose
that species recognition is at least as plausible as sexual display in
explaining the diversity of horn and frill structures in Ceratop-
sidae (Centrosaurinae plus Chasmosaurinae) and their near rela-
tives, and that substantial sexual dimorphism has not yet been
established in ceratopsids.
Nearly all anatomical and behavioral studies of Triceratops
have been based on adult skulls. The development of larger,
more visible postorbital horns, a massive nasal horn, and the
ossification of epoccipitals along the frill margin may have sig-
naled sexual maturity and the onset of mating in adult Tricera-
tops. Functional analyses of these cranial features in adult Tri-
ceratops have often restricted their role to sexual display and
mating behavior, based on the presumption that these features
were not present in very young individuals and did not express
themselves until adulthood (Sampson et al., 1997; Dodson and
Currie, 1990; Sampson, 2001; but see Farke, 2004). To the con-
trary, UCMP 154452 demonstrates that in very young Tricera-
tops, these species-specific cranial characters began to be formed
early in lifeostensibly earlier than in centrosaurine ceratop-
sidsand may have been important for visual communication
and species recognition even at this early stage.
UCMP 154452 documents the youngest ontogenetic stage of
Triceratops and illuminates a great transformation in size, shape,
and rearrangement of cranial elements that occurred in the skull
during growth into adulthood. As the smallest ceratopsid skull
known, it provides a new end member for the youngest stage of
cranial ontogeny in Triceratops. The patent cranial sutures reveal
morphology that is often concealed in adult skulls. Juvenile fea-
tures of the braincase in UCMP 154452 recapitulate the primitive
character state of more basal neoceratopsians. On the other
hand, cranial ornamentation in the frill and the postorbital horns
of Triceratops were not restricted to adult members, but began at
an early age. The appearance of horns and a scalloped frill at this
small size and early age is support for the hypothesis that cranial
ornaments in ceratopsids were at least as important as a visual
organ for species communication as they may have been for
sexual display or agonistic encounters.
We thank Mr. Harley J. Garbani, who discovered the skull and
skillfully prepared the delicate specimen. We also thank Laura
Cunningham for her illustration of the skull; Karen Klitz for her
final illustration of the skull, pencil drawings of the individual
cranial elements, and figure layout; Jim Hendel and Patricia Hol-
royd for photographic assistance; Sterling Nesbitt and J. Howard
Hutchison for helpful discussion. Michael Holland and Jane Ma-
son undertook final restoration, molding and casting of the skull.
We thank Brenda Chinnery, Peter Dodson, Thomas Lehman,
and Scott Sampson for helpful comments on earlier versions of
this manuscript. We appreciate the editorial assistance of Nicho-
las Fraser, David Weishampel, and Mark Wilson. Financial sup-
port from the University of California Museum of Paleontology
is gratefully acknowledged. This is University of California Mu-
seum of Paleontology contribution no. 1883.
Brown, B., and E. M. Schlaikjer. 1940a. The origin of ceratopsian horn-
cores. American Museum Novitates 1065:17.
Brown, B., and E. M. Schlaikjer. 1940b. The structure and relationships
of Protoceratops. Annals of the New York Academy of Sciences
Carpenter, K. 1982. Baby dinosaurs from the Late Cretaceous Lance and
Hell Creek formations and a description of a new species of thero-
pod. Contributions to Geology, University of Wyoming 20:123134.
Carpenter, K., K. F. Hirsch, and J. R. Horner. 1994. Summary and pro-
spectus; pp. 366370 in K. Carpenter, K. F. Hirsch, and J. R. Horner
(eds.), Dinosaur Eggs and Babies. Cambridge University Press,
Cambridge, U.K.
Chinnery, B. J. 2004. Description of Prenoceratops pieganensis gen. et sp.
nov. (Dinosauria: Neoceratopsia) from the Two Medicine Forma-
tion of Montana. Journal of Vertebrate Paleontology 24:572590.
Chinnery, B. J., and D. B. Weishampel. 1998. Montanoceratops cerorhyn-
chus (Dinosauria: Ceratopsia) and relationships among basal neo-
ceratopsians. Journal of Vertebrate Paleontology 18:569585.
Colbert, E. H. 1990. Variation in Coelophysis bauri; pp. 8190 in
K. Carpenter and P. J. Currie (eds.), Dinosaur Systematics: Ap-
proaches and Perspectives. Cambridge University Press, Cambridge,
Darwin, C. 1871. The Descent of Man, and Selection in Relation to Sex.
2 volumes. John Murray, London, 423 + 475 pp.
Dodson, P. 1975. Taxonomic implications of relative growth in lambeo-
saurine hadrosaurs. Systematic Zoology 24:3754.
Dodson, P. 1976. Quantitative aspects of relative growth and sexual di-
morphism in Protoceratops. Journal of Paleontology 50:929940.
Dodson, P. 1993. Comparative craniology of the Ceratopsia. American
Journal of Science 293A:200234.
Dodson, P. 1996. The Horned Dinosaurs. A Natural History. Princeton
University Press, Princeton, New Jersey, 360 pp.
Dodson, P., and P. J. Currie. 1988. The smallest ceratopsid skullJudith
River Formation of Alberta. Canadian Journal of Earth Sciences
Dodson, P., and P. J. Currie. 1990. Neoceratopsia; pp. 593618 in D. B.
Weishampel, P. Dodson, and H. Osmólska (eds.), The Dinosauria.
University of California Press, Berkeley.
Dodson, P., C. A. Forster, and S. D. Sampson. 2004. Ceratopsidae; pp.
494513 in D. B. Weishampel, P. Dodson, and H. Osmólska (eds.),
The Dinosauria. University of California Press, Berkeley.
Farke, A. 2004. Horn use in Triceratops (Dinosauria: Ceratopsidae):
testing behavioral hypotheses using scale models. Palaeontologia
Electronica 7:110.
Farlow, J. O., and P. Dodson. 1975. The behavioral significance of frill
and horn morphology in ceratopsian dinosaurs. Evolution 29:
Forster, C. A. 1996a. Species resolution in Triceratops: cladistic and mor-
phometric approaches. Journal of Vertebrate Paleontology 16:
Forster, C. A. 1996b. New information on the skull of Triceratops. Jour-
nal of Vertebrate Paleontology 16:246258.
Gilmore, C. W. 1917. Brachyceratops, a ceratopsian dinosaur from the
Two Medicine Formation of Montana, with notes on associated rep-
tiles. United States Geological Survey Professional Paper 103:145.
Goodwin, M. B., and J. R. Horner. 2001. How Triceratops got its horns:
new information from a growth series on cranial morphology and
ontogeny. Journal of Vertebrate Paleontology 21(3, Supplement):
Goodwin, M. B., and J. R. Horner. 2004. Cranial histology of pachy-
cephalosaurs (Ornithischia: Marginocephalia) reveals transitory
structures inconsistent with head-butting behavior. Paleobiology 30:
Goodwin, M. B., E. A. Buchholtz, and R. E. Johnson. 1998. Cranial
anatomy and diagnosis of Stygimoloch spinifer (Ornithischia: Pachy-
cephalosauria) with comments on cranial display structures in ago-
nistic behavior. Journal of Vertebrate Paleontology 18:363375.
Goodwin, M. B., J. R. Horner, and W. A. Clemens. 1997. Morphological
variation and ontogeny in the skull of Triceratops. Journal of Ver-
tebrate Paleontology 17(3, Supplement):49A.
Hailu, Y., and P. Dodson. 2004. Basal Ceratopsia; pp. 478493 in D. B.
Weishampel, P. Dodson, and H. Osmólska (eds.), The Dinosauria.
University of California Press, Berkeley.
Hatcher, J. B., O. C. Marsh, and R. S. Lull. 1907. The Ceratopsia. U.S.
Geological Survey, Monograph 491-XXIX:1300.
Hopson, J. A. 1975. The evolution of cranial display structures in had-
rosaurian dinosaurs. Paleobiology 1:2143.
Horner, J. R., and P. J. Currie. 1994. Embryonic and neonatal morphol-
ogy and ontogeny of a new species of Hypacrosaurus (Ornithiscia,
Lambeosauridae) from Montana and Alberta; pp. 312336 in
K. Carpenter, K. F. Hirsch, and J. R. Horner (eds.), Dinosaur Eggs
and Babies. Cambridge University Press, Cambridge.
Horner, J. R., and C. Marshall. 2002. Keratinous covered dinosaur skulls.
Journal of Vertebrate Paleontology 22(3, Supplement):67A.
Lehman, T. M. 1989. Chasmosaurus mariscalensis, sp. nov., a new cera-
topsian dinosaur from Texas. Journal of Vertebrate Paleontology
Lehman, T. M. 1990. The ceratopsian subfamily Chasmosaurinae: sexual
dimorphism and systematics; pp. 211230 in K. Carpenter and P. J.
Currie (eds.), Dinosaur Systematics: Approaches and Perspectives.
Cambridge University Press, Cambridge.
Makovicky, P. J. 2001. A Montanoceratops cerorhynchus (Dinosauria:
Ceratopsia) braincase from the Horseshoe Canyon Formation of
Alberta; pp. 243262 in D. H. Tanke and K. Carpenter (eds.), Me-
sozoic Vertebrate Life. Indiana University Press, Bloomington.
Maryanska, T., and H. Osmólska. 1975. Protoceratopsidae (Dinosauria)
of Asia. Palaeontologica Polonica 33:133182.
Molnar, R. E. 1977. Analogies in the evolution of combat and display
structures in ornithopods and ungulates. Evolutionary Theory 3:
Ostrom, J. H., and P. Wellnhofer. 1986. The Munich specimen of Tri-
ceratops with a revision of the genus. Zitteliana 14:111158.
Owens, I. P. F., and I. R. Hartley. 1998. Sexual dimorphism in birds: why
are there so many different forms of dimorphism? Proceedings of
the Royal Society of London B 265:397407.
Padian, K., J. Horner, and J. Dhaliwal. 2004. Species recognition as the
principal cause of bizarre structures in dinosaurs. Journal of Verte-
brate Paleontology 24(3, Supplement):100A.
Prum, R. O., R. L. Morrison, and G. R. Ten Eyck. 1994. Structural color
production by constructive reflection from ordered collagen arrays
in a bird (Philepitta castanea: Eurylaimidae). Journal of Morphology
Prum, R. O., R. Torres, C. Kovach, S. Williamson, and S. M. Goodman.
1999. Coherent light scattering by nanostructured collagen arrays in
the caruncles of the Malagasy asities (Eurylaimidae: Aves). Journal
of Experimental Biology 202:35073522.
Raath, M. A. 1990. Morphological variation in small theropods and its
meaning in systematics: evidence from Syntarsus; pp. 91105 in
K. Carpenter and P. J. Currie (eds.), Dinosaur Systematics: Ap-
proaches and Perspectives. Cambridge University Press, Cambridge.
Romer, A. S. 1966. Vertebrate Paleontology. The University of Chicago
Press, Chicago, 468 pp.
Ryan, M. J., and P. J. Currie. 1998. First report of protoceratopsians
(Neoceratopsia) from the Late Cretaceous Judith River Group, Al-
berta, Canada. Canadian Journal of Earth Sciences 35:820826.
Ryan, M. J., Russell, A. P., Eberth, D. A., and P. J. Currie. 2001. The
taphonomy of a Centrosaurus (Ornithischia: Ceratopsidae) bone
bed from the Dinosaur Park Formation (Upper Campanian), Al-
berta, Canada, with comments on cranial ontogeny. Palaios 16:
Sampson, S. D. 1997. Dinosaur combat and courtship; pp. 383393 in
J. O. Farlow and M. K. Brett-Surman (eds.), The Complete Dino-
saur. Indiana University Press, Bloomington.
Sampson, S. D. 2001. Speculations on the socioecology of ceratopsid
dinosaurs (Ornithischia: Neoceratopsia); pp. 263278 in D. H.
Tanke and K. Carpenter (eds.), Mesozoic Vertebrate Life. Indiana
University Press, Bloomington.
Sampson, S. D., M. J. Ryan, and D. H. Tanke. 1997. Craniofacial ontog-
eny in centrosaurine dinosaurs (Ornithischia: Ceratopsidae): taxo-
nomic and behavioral implications. Zoological Journal of the Lin-
nean Society 121:293337.
Schlaikjer, E. M. 1935. The Torrington Member of the Lance Formation
and a study of a new Triceratops. Bulletin of the Museum of Com-
parative Zoology 76:2968.
Sternberg, C. M. 1951. Complete skeleton of Leptoceratops gracilis
Brown from the Upper Edmonton Member on Red Deer River,
Alberta. National Museum of Canada Bulletin, Annual Report
(19491950) 123:225255.
Tokaryk, T. T. 1997. First evidence of juvenile ceratopsians (Reptilia:
Ornithischia) from the Frenchman Formation (late Maastrichtian)
of Saskatchewan. Canadian Journal of Earth Sciences 34:14011404.
Vrba, E. S. 1984. Evolutionary pattern and process in the sister-group
Alcelaphini- Aepcerotini (Mammalia: Bovidae); pp. 6279 in
N. Eldredge and S. M. Stanley (eds.), Living Fossils. Springer-
Verlag, New York.
Weckerly, F. W. 1998. Sexual-size dimorphism: influence of mass and
mating systems in the most dimorphic mammals. Journal of Mam-
malogy 79:3352.
Williamson, T. E., and T. D. Carr. 2002. A new genus of derived pachy-
cephalosaurian from western North America. Journal of Vertebrate
Paleontology 22:779801.
Wilson, G. 2004. A quantitative assessment of mammalian change lead-
ing up to and across the CretaceousTertiary boundary in north-
eastern Montana. Unpublished Ph.D. dissertation, University of
California, Berkeley, California, 412 pp.
Wolfe, D. G., & Kirkland, J. I. 1998. Zuniceratops christopheri n. gen and
n. sp., a ceratopsian dinosaur from the Moreno Hill Formation (Cre-
taceous, Turonian) of west-central New Mexico; pp. 303317 in S. G.
Lucas, J. I. Kirkland, and J. W. Estep (eds.), Lower and Middle
Cretaceous Terrestrial Ecosystems. New Mexico Museum of Natu-
ral History and Science Bulletin 14.
Submitted 4 January 2005; accepted 13 June 2005.
... Bony horn development and ontogeny can be closely associated in diapsid reptiles. The horns and epiossifications often grow at differential rates on a single animal's skull, as in marginocephalian dinosaurs (e.g., Sampson, Ryan & Tanke, 1997;Goodwin et al., 2006) and phrynosomatid lizards (e.g., Powell, Russell & Ryan, 2002;Bergmann & Berk, 2012). It is plausible that weigeltisaurid horns also developed at different rates such that the lateral arcade of horns developed prior to the parietal horns, such that the condition in Coelurosauravus elivensis represents an earlier ontogenetic state than is represented by known specimens of Rautiania spp. ...
Full-text available
Background Weigeltisauridae is a clade of small-bodied diapsids characterized by a horned cranial frill, slender trunk and limbs, and a patagium supported by elongated bony rods. Partial skeletons and fragments are definitively known only from upper Permian (Lopingian) rocks in England, Germany, Madagascar and Russia. Despite these discoveries, there have been few detailed descriptions of weigeltisaurid skeletons, and the homologies of many skeletal elements—especially the rods supporting the patagium—remain the subject of controversy. Materials & Methods Here, we provide a detailed description of a nearly complete skeleton of Weigeltisaurus jaekeli from the upper Permian (Lopingian: Wuchiapingian) Kupferschiefer of Lower Saxony, Germany. Briefly addressed by past authors, the skeleton preserves a nearly complete skull, postcranial axial skeleton, appendicular skeleton, and patagial supports. Through comparisons with extant and fossil diapsids, we examine the hypotheses for the homologies of the patagial rods. To examine the phylogenetic position of Weigeltisauridae and characterize the morphology of the clade, we integrate the material and other weigeltisaurids into a parsimony-based phylogenetic analysis focused on Permo-Triassic non-saurian Diapsida and early Sauria (61 taxa, 339 characters). Results We recognize a number of intriguing anatomical features in the weigeltisaurid skeleton described here, including hollow horns on the post-temporal arch, lanceolate teeth in the posterior portion of the maxilla, the absence of a bony arch connecting the postorbital and squamosal bones, elongate and slender phalanges that resemble those of extant arboreal squamates, and patagial rods that are positioned superficial to the lateral one third of the gastral basket. Our phylogenetic study recovers a monophyletic Weigeltisauridae including Coelurosauravus elivensis , Weigeltisaurus jaekeli , and Rautiania spp. The clade is recovered as the sister taxon to Drepanosauromorpha outside of Sauria (=Lepidosauria + Archosauria). Conclusions Our anatomical observations and phylogenetic analysis show variety of plesiomorphic diapsid characters and apomorphies of Weigeltisauridae in the specimen described here. We corroborate the hypothesis that the patagial ossifications are dermal bones unrelated to the axial skeleton. The gliding apparatus of weigeltisaurids was constructed from dermal elements unknown in other known gliding diapsids. SMNK-PAL 2882 and other weigeltisaurid specimens highlight the high morphological disparity of Paleozoic diapsids already prior to their radiation in the early Mesozoic.
... Squamosal elongation -In juvenile chasmosaurines, the squamosal is anteroposteriorly short, similar to the condition in adult centrosaurine ceratopsids and more basal neoceratopsians (Lehman, 1990;Goodwin et al., 2006;Scannella & Horner, 2010;Mallon et al., 2014;Campbell et al., 2016). In chasmosaurines, the squamosal elongates through ontogeny, although the timing of the elongation varies phylogenetically (Lehman, 1990;Scannella & Horner, 2010). ...
Full-text available
Three new chasmosaurines from the Kirtland Formation (~75.0–73.4 Ma), New Mexico, form morphological and stratigraphic intermediates between Pentaceratops (~74.7–75 Ma, Fruitland Formation, New Mexico) and Anchiceratops (~72–71 Ma, Horseshoe Canyon Formation, Alberta). The new specimens exhibit gradual enclosure of the parietal embayment that characterizes Pentaceratops , providing support for the phylogenetic hypothesis that Pentaceratops and Anchiceratops are closely related. This stepwise change of morphologic characters observed in chasmosaurine taxa that do not overlap stratigraphically is supportive of evolution by anagenesis. Recently published hypotheses that place Pentaceratops and Anchiceratops into separate clades are not supported. This phylogenetic relationship demonstrates unrestricted movement of large-bodied taxa between hitherto purported northern and southern provinces in the late Campanian, weakening support for the hypothesis of extreme faunal provincialism in the Late Cretaceous Western Interior.
Full-text available
Our understanding of Late Cretaceous dinosaur ecosystems from North America has considerably improved through stable isotope analyses on fossil bones and teeth. Oxygen and carbon stable isotopic compositions of structurally-bound carbonate in these fossil apatites are commonly used to infer variations of ingested water and food sources, which are in turn related to environmental and climatic conditions. Incremental isotopic records potentially provide insights into seasonality and migratory behaviour. So far, these reconstructions are based on vertebrate remains from spatiotemporally diverse datasets. Here, we present oxygen (δ¹⁸O) and carbon (δ¹³C) isotopic records from a large, spatially and temporally well-constrained, Triceratops bonebed from the Upper Maastrichtian Lance Formation (eastern Wyoming. USA). These isotopic compositions allow to elucidate the palaeoecology of these large herbivores and their ecosystem in detail, as well as their habitat use, diet and possible migration. The δ¹⁸O signature from incrementally sampled Triceratops teeth reveal relatively low intra-tooth variation (average 1.3 ‰), comparable to contemporaneous dinosaur species as well as modern herbivorous mammals. Average δ¹³C values (−5.4 ‰) are somewhat higher than for modern C3 plant grazers, and hint towards complex interactions during carbon uptake by non-avian herbivorous dinosaurs. Calculated δ¹⁸O of drinking water (−14.8 ‰) combined with the local sedimentology of fine-grained siliciclastic deposits with high total organic and low carbonate contents strongly suggest a freshwater environment. Additionally, the combined average δ¹⁸O and δ¹³C Triceratops isotope signatures indicate a living environment intermediate between inland forests and coastal floodplains, expanding on earlier theories of ornithischian niche partitioning. Our robust dataset provides meaningful tests of habitat and palaeoecological hypotheses for Triceratops, and highlights the application of spatiotemporally well-constrained fossil remains.
Effective allometry research relies on appropriate size variables; however, two of the largest obstacles in subadult (ontogenetic) allometry research is small sample sizes and unknown dimensions. This study overcomes a barrier of ontogenetic allometry research by proposing alternative size variables that do not require additional calculations for use in subadult allometry research and retain general patterns among long bones when stature is used for size. Diaphyseal measurements, stature, and age were collected from computed tomography (CT) and full‐body radiographic images for a sample of subadults between birth and 13 years from the United States (U.S., n = 308) and South Africa (Z.A., n = 25). Nineteen alternative size variables were evaluated using reduced‐major‐axis regression to identify the closest one‐to‐one relationship to stature. The applicability across samples was then evaluated using the selected alternative size variables. Radius midshaft breadth (RMSB), femur midshaft breadth (FMSB), and the geometric mean of midshaft breadths (GM midshaft) yielded the closest isometric relationships to stature. Allometric relationships among long bones are maintained when using stature, FMSB, and GM midshaft as size variables for both the U.S. and Z.A. samples. A large, modern dataset facilitated an investigation into alternative size variables that can be used for single‐bone ontogenetic allometry. Generalizability of the model suggests FMSB and GM midshaft are persistent across populations. This methodology identifies alternative size variables appropriate for other allometry research and offers a robust approach even when historically relied upon size variables are unknown.
Dicraeosaurid sauropods are iconically characterized by the presence of elongate hemispinous processes in presacral vertebrae. These hemispinous processes can show an extreme degree of elongation, such as in the Argentinean forms Amargasaurus cazaui, Pilmatueia faundezi and Bajadasaurus pronuspinax. These hyperelongated hemispinous processes have been variably interpreted as a support structure for a padded crest/sail as a display, a bison‐like hump or as the internal osseous cores of cervical horns. With the purpose to test these hypotheses, here we analyze, for the first time, the external morphology, internal microanatomy and bone microstructure of the hemispinous processes from the holotype of Amargasaurus, in addition to a second dicraeosaurid indet. (also from the La Amarga Formatin; Lower Cretaceous, Argentina). Transverse thin‐sections sampled from the proximal, mid and distal portions of both cervical and dorsal hemispinous processes reveal that the cortical bone is formed by highly vascularized fibrolamellar bone interrupted with cyclical growth marks. Obliquely oriented Sharpey's fibres are mostly located in the medial and lateral portions of the cortex. Secondary remodelling is evidenced by the presence of abundant secondary osteons irregularly distributed within the cortex. Both anatomical and histological evidence does not support the presence of a keratinized sheath (i.e. horn) covering the hyperelongated hemispinous processes of Amargasaurus, and either, using a parsimonious criterium, in other dicraeosaurids with similar vertebral morphology. The spatial distribution and relative orientation of the Sharpey's fibres suggest the presence of an important system of interspinous ligaments that possibly connect successive hemispinous processes in Amargasaurus. These ligaments were distributed along the entirety of the hemispinous processes. The differential distribution of secondary osteons indicates that the cervical hemispinous processes of Amargasaurus were subjected to mechanical forces that generated higher compression strain on the anterior side of the elements. Current data support the hypothesis for the presence of a ‘cervical sail’ in Amargasaurus and other dicraeosaurids. Life restoration of Amargasaurus cazaui. Current data support the hypothesis for the presence of a ‘cervical sail’ in Amargasaurus and other dicraeosaurids. Illustration made by Gabriel Lio.
Ontogenetic niche shifts, the phenomenon whereby animals change their resource use with growth, were probably widespread in dinosaurs, but most studies of duck-billed dinosaur ontogeny have so far focused mainly on the development of the cranial ornamentation. Here, we quantify allometry of 13 ecomorphological variables of the skull and examine tooth microwear in a sample of North American hadrosaurids to better understand their ecological functioning with growth. Our results indicate that, consistent with the Jarman-Bell principle relating body size to fibre intake and feeding selectivity, juvenile hadrosaurids were relatively more selective than their adult counterparts and subsisted on softer, low-growing browse cropped using lateral rotations of the neck. Chewing movements of the jaw probably did not differ greatly between growth stages. Our findings invite further investigation relating to cranial ontogenetic allometry in hadrosauromorphs more broadly, and to the possible role of ontogenetic niche shifts in the size structuring of Late Cretaceous herbivore communities.
The cranial skeleton of the enigmatic gliding neodiapsid reptile Coelurosauravus elivensis (Lower Sakamena Formation, Lopingian, Southwestern Madagascar) is re-described in detail. All previously referred specimens are re-examined under both direct observations and Reflectance Transformation Imaging. Their exquisite preservation yields detailed three-dimensional information on the outline of individual bones and their osteological relationships, which are missing in the Laurasian remains. In contrast to previous studies, the ontogenetic maturity of all specimens is re-affirmed. Previously unidentified elements of the palate, braincase and mandible are described, and a novel reconstruction is proposed, including the first palatal reconstruction in a weigeltisaurid reptile. C. elivensis has the smallest skull of all weigeltisaurids and differs from other species in its facial ornamentation, parietosquamosal frill and larger anterior maxillary dentition. We also provide extensive comparisons with contemporaneous reptiles, possibly closely related taxa and more recent analogs, as well as a preliminary discussion of the functional anatomy of the peculiar cranial morphology of weigeltisaurids. The cranial skeleton is a truss construction with large orbits and temporal fenestrae. By analogy with extant chamaeleonids, the elongate parietosquamosal frill is associated with an increase in length and diameter of the temporal jaw adductors, resulting in an increased gape and/or bite force and speed. Additionally, the spikes and frills of weigeltisaurids most likely served as a display and defensive structure.
Full-text available
An associated incomplete skeleton of a ceratopsid dinosaur from the Campanian deposits of the Allison Member of the Menefee Formation in New Mexico, USA is described. Although it was originally described over two decades ago, newly prepared portions of the Menefee Formation skeleton and reinterpretations of previously known morphology, in addition to newly described specimens have provided new information on ceratopsids, and centrosaurines in particular. These new data allow for a thorough reassessment of the specimen and the erection of a new taxon: Menefeeceratops sealeyi gen. et sp. nov., potentially the oldest recognized member of Centrosaurinae. Menefeeceratops sealeyi is represented by diagnostic cranial and postcranial skeletal elements. The cranial elements include a portion of the left premaxilla, a nearly complete left postorbital horncore, a parietal fragment, the right and left squamosals, the left jugal, the predentary, and the left dentary. Postcranial material consists of two cervical vertebrae, eight dorsal vertebrae, a partial sacrum with six sacral vertebrae, 11 dorsal ribs, the distal left radius, proximal and distal portions of the left ulna, the left femur, and a left metatarsal II. The taxonomic validity of Menefeeceratops sealeyi is supported by a combination of several morphological characters. These include a lack of epiossifications on the lateroposterior edge of the parietal (shared with Machairoceratops), three epiossifications on the squamosal, and three smaller, secondary undulations as part of episquamosal locus S1. There are also two subequal embayments on the posterior free margin of the squamosal with the more dorsal embayment (between episquamosal loci 1 and 2) distinctly larger than the ventral (= lateroventral) one (between episquamosal loci 2 and 3), three ridges on the lateral (dorsolateral) surface of the squamosal, an elongate posterior portion of the squamosal, the presence of a shallow but distinct groove on the medial surface of the squamosal nearly paralleling the ventrolateral and ventroposterior edges, elongate postorbital (= supraorbital) horns that are anteriorly curved distally, and two elongate ridges on the lateral surface of the dentary that diverge anteriorly, creating a distinct anterior triangular fossa. Phylogenetic analysis of Menefeeceratops sealeyi places this new species as a basal centrosaurine, most closely related to Crittendenceratops krzyzanowskii, thus adding to the growing record of centrosaurines discovered in western North America. It thus provides new information about the diversity of morphologies throughout different species and the temporal and paleobiogeographic distribution of these animals throughout Laramidia during the Late Cretaceous. Its presence as one of the, if not the, oldest members of the Centrosaurinae also suggests centrosaurines originated in the southern portions of western North America and the southern Rocky Mountain region, and subsequently radiated north during the upper middle to late Campanian.
Full-text available
Morphology forms the most fundamental level of data in vertebrate palaeontology because it is through interpretations of morphology that taxa are identified, creating the basis for broad evolutionary and palaeobiological hypotheses. Assessing maturity is one of the most basic aspects of morphological interpretation and provides the means to study the evolution of ontogenetic changes, population structure and palaeoecology, life‐history strategies, and heterochrony along evolutionary lineages that would otherwise be lost to time. Saurian reptiles (the least‐inclusive clade containing Lepidosauria and Archosauria) have remained an incredibly diverse, numerous, and disparate clade through their ~260‐million‐year history. Because of the great disparity in this group, assessing maturity of saurian reptiles is difficult, fraught with methodological and terminological ambiguity. We compiled a novel database of literature, assembling >900 individual instances of saurian maturity assessment, to examine critically how saurian maturity has been diagnosed. We review the often inexact and inconsistent terminology used in saurian maturity assessment (e.g. ‘juvenile’, ‘mature’) and provide routes for better clarity and cross‐study coherence. We describe the various methods that have been used to assess maturity in every major saurian group, integrating data from both extant and extinct taxa to give a full account of the current state of the field and providing method‐specific pitfalls, best practices, and fruitful directions for future research. We recommend that a new standard subsection, ‘Ontogenetic Assessment’, be added to the Systematic Palaeontology portions of descriptive studies to provide explicit ontogenetic diagnoses with clear criteria. Because the utility of different ontogenetic criteria is highly subclade dependent among saurians, even for widely used methods (e.g. neurocentral suture fusion), we recommend that phylogenetic context, preferably in the form of a phylogenetic bracket, be used to justify the use of a maturity assessment method. Different methods should be used in conjunction as independent lines of evidence when assessing maturity, instead of an ontogenetic diagnosis resting entirely on a single criterion, which is common in the literature. Critically, there is a need for data from extant taxa with well‐represented growth series to be integrated with the fossil record to ground maturity assessments of extinct taxa in well‐constrained, empirically tested methods.
Full-text available
The fossil record provides compelling examples of heterochrony at macroevolutionary scales such as the peramorphic giant antlers of the Irish elk. Heterochrony has also been invoked in the evolution of the distinctive cranial frill of ceratopsian dinosaurs such as Triceratops. Although ceratopsian frills vary in size, shape, and ornamentation, quantitative analyses that would allow for testing hypotheses of heterochrony are lacking. Here, we use geometric morphometrics to examine frill shape variation across ceratopsian diversity and within four species preserving growth series. We then test whether the frill constitutes an evolvable module both across and within species, and compare growth trajectories of taxa with ontogenetic growth series to identify heterochronic processes. Evolution of the ceratopsian frill consisted primarily of progressive expansion of its caudal and caudolateral margins, with morphospace occupation following taxonomic groups. Although taphonomic distortion represents a complicating factor, our data support modularity both across and within species. Peramorphosis played an important role in frill evolution, with acceleration operating early in neoceratopsian evolution followed by progenesis in later diverging cornosaurian ceratopsians. Peramorphic evolution of the ceratopsian frill may have been facilitated by the decoupling of this structure from the jaw musculature, an inference that predicts an expansion of morphospace occupation and higher evolutionary rates among ceratopsids as indeed borne out by our data. However, denser sampling of the meager record of early‐diverging taxa is required to test this further. We find support for peramorphic evolution of the frill of ceratopsian dinosaurs, likely facilitated by the decoupling of this structure from jaw musculature.
Full-text available
Variation in the extent of sexual dimorphism among bird species is traditionally attributed to differences in social mating system. However, there are many different forms of dimorphism among birds, and not all of them show an obvious correlation with social mating system. For example, recent work has shown that many highly polygamous species are, in fact, monomorphic, whereas many putatively monogamous species are dimorphic. In this paper we break up sexual dimorphism into subcomponents and then use comparative analyses to examine the pattern of covariation between these subcomponents and various aspects of sexual, social, and parental behaviour. Our first finding is that size dimorphism and plumage-colour dimorphism do not show the same pattern of covariation. Differences in size dimorphism are associated with variation in social mating system and sex differences in parental care, whereas differences in plumage-colour dimorphism are associated with variation in the frequency of extra-bond paternity. These results suggest that size dimorphism is associated with the sort of intrasexual competition described by traditional classifications of social mating system, whereas plumage-colour dimorphism is associated with cryptic female choice. However, when we break up plumage-colour dimorphism according to whether it is due to melanins, carotenoids or structural colours, we find that each category of plumage-colour dimorphism shows a different pattern of covariation. The correlation between overall plumage-colour dimorphism and the rate of extra-bond paternity is due to structural colours, whereas melanin-based dimorphism is associated with sex differences in parental care. The former result is particularly interesting given that new work suggests structural colours are associated with active sexual displays and the reflection of ultraviolet light.
Advanced ceratopsians (family Ceratopsidae) are divisible into two groups, based primarily on the relative proportions and fenestration of their neck frills. Lambe's subfamily Eoceratopsinae is abandoned because of the placement of Eoceratops in synonymy with Chasmosaurus, and the inclusion of Triceratops within the "long-frilled' ceratopsids. Lambe's subfamilies Chasmosaurinae and Centrosaurinae should be revived, however, for the "long-frilled' and "short-frilled' ceratopsids, respectively. Species-level taxonomy of ceratopsids is complicated by pronounced individual and ontogenetic variability, and sexual dimorphism, in most species. A population sample of Chasmosaurus from Texas suggests that orientation of the supraorbital horncores is a useful criterion for separating sexual morphs. Based on this criterion, all chasmosaurine genera contain a species or group of species exhibiting the supposed "female' morph, and one exhibiting the supposed "male' morph. -from Author
A theory is presented that cranial crests of hadrosaurs were visual and acoustical display organs. Facial morphology and phylogeny of the Hadrosauridae and earlier theories of crest function are reviewed. The following hypothesis is presented: cranial crests, whether hollow or solid, served as visual signal structures, and hollow lambeosaur crests were also vocal resonators; all crests promoted successful matings within species, i.e. , they served as premating genetic isolating mechanisms. The following predictions are tested and found to support the hypothesis: (1) hadrosaurs had well-developed eyes and ears; (2) external features of crests varied independently of internal structure; (3) crest variations were species-specific and sexually-dimorphic; (4) crest distinctiveness correlates with species diversity; (5) crest size tended to increase through time. The circumnarial depression on the side of the face in hadrosaurines housed an inflatable diverticulum of the nasal passage which served as a visual display organ. Primitive hadrosaurs (kritosaurs) possessed a small nasal horn used as a butting weapon in intraspecific combat. Because the weapon was also used in intimidative displays, narial diverticula evolved to draw attention to it. In the kritosaur Brachylophosaurus fighting was modified to ritualized head-pushing using the flat nasal “shield”. Saurolophines expanded the diverticula on to the elongated nasal horn, converting the weapon to a dominance rank symbol. In non-crested edmontosaurs, enlarged diverticula assumed a vocalization function. Lambeosaurs created resonators by enclosing the diverticula in bone; they further enhanced the resonator function of the nose by forming elongated “organ pipes” in the premaxillae. This “pushed” the olfactory region above the eyes as a conspicuous dome which then was modified to form species-specific visual display organs.
The Ceratopsia or horned dinosaurs demonstrate great diversity of skull shape that makes them ideal subjects for morphometric investigation. Skulls of ceratopsians are subjected to landmark-based morphometric analysis using the robust mapping technique, RFTRA (resistant-fit theta-rho analysis). This technique elucidates morphological trends in skull shape in a series of pairwise comparisons. Morphological distance data among taxa from RFTRA comparisons are subjected to cluster analysis. The correlation of display characters with functional complexes suggests that no characters of the skull are genuinely trivial. Major morphological trends in the skull involve reorganization of the cheek region, caudal movement of the jugal, reduction of the quadratojugal, caudoventral expansion of squamosal caudal to the quadrate in centrosaurines, and further expansion of the squamosal at right angle to the previous growth axis in chasmosaurines. -from Author