The Lower Jurassic ornithischian dinosaur Heterodontosaurus tucki

Transcription

Zoological Journal of the Linnean Society, 2011, 163, 182–276. With 41 figures
The Lower Jurassic ornithischian dinosaur
Heterodontosaurus tucki Crompton & Charig, 1962:
cranial anatomy, functional morphology, taxonomy,
and relationships
zoj_697
182..276
DAVID B. NORMAN FLS1*, ALFRED W. CROMPTON2, RICHARD J. BUTLER1,3,
LAURA B. PORRO1,4 and ALAN J. CHARIG†
1
The Sedgwick Museum, Department of Earth Sciences, University of Cambridge, Cambridge CB2
3EQ, UK
2
Museum of Comparative Zoology, Harvard University, Cambridge, Massachusetts, USA
3
Bayerische Staatssammlung für Paläontologie und Geologie, Richard-Wagner-Str. 10, 80333 Munich,
Germany
4
Department of Organismal Biology and Anatomy, University of Chicago, IL 60637, USA
Received 23 April 2010; revised 25 August 2010; accepted for publication 27 August 2010
The cranial anatomy of the Lower Jurassic ornithischian dinosaur Heterodontosaurus tucki Crompton & Charig,
1962 is described in detail for the first time on the basis of two principal specimens: the holotype (SAM-PK-K337)
and referred skull (SAM-PK-K1332). In addition several other specimens that have a bearing on the interpretation
of the anatomy and biology of Heterodontosaurus are described. The skull and lower jaw of Heterodontosaurus are
compact and robust but perhaps most notable for the heterodont dentition that merited the generic name. Details
of the cranial anatomy are revealed and show that the skull is unexpectedly specialized in such an early
representative of the Ornithischia, including: the closely packed, hypsodont crowns and ‘warping’ of the occlusal
surfaces (created by progressive variation in the angulation of wear on successive crowns) seen in the cheek
dentition; the unusual sutural relationships between the bones along the dorsal edge of the lower jaw; the very
narrow, deeply vaulted palate and associated structures on the side wall of the braincase; and the indications of
cranial pneumatism (more commonly seen in basal archosaurs and saurischian dinosaurs). Evidence for tooth
replacement (which has long been recognized, despite frequent statements to the contrary) is suggestive of an
episodic, rather than continuous, style of tooth replacement that is, yet again, unusual in diapsids generally and
particularly so amongst ornithischian dinosaurs. Cranial musculature has been reconstructed and seems to
conform to that typically seen in diapsids, with the exception of the encroachment of M. adductor mandibulae
externus superficialis across the lateral surface of the temporal region and external surface of the lower jaw.
Indications, taken from the unusual shape of the occlusal surfaces of the cheek dentition and jaw musculature, are
suggestive of a novel form of jaw action in this dinosaur. The taxonomy of currently known late Karoo-aged
heterodontosaurids from southern Africa is reviewed. Although complicated by the inadequate nature of much of
the known material, it is concluded that two taxa may be readily recognized: H. tucki and Abrictosaurus consors.
At least one additional taxon is recognized within the taxa presently named Lanasaurus and Lycorhinus; however,
both remain taxonomically problematic and their status needs to be further tested and may only be resolved by
future discoveries. The only other named taxon, Geranosaurus atavus, represents an invalid name. The recognition
of at least four distinct taxa indicates that the heterodontosaurids were speciose within the late Karoo ecosystem.
The systematics of Heterodontosaurus and its congeners has been analysed, using a restricted sample of taxa. A
basal (nongenasaurian) position within Ornithischia is re-affirmed. There are at least four competing hypotheses
*Corresponding author. E-mail: [email protected]
†
Alan Jack Charig died 15 July 1997. This paper has its origin in notes, illustrations, and photographs, in the possession of
A. J. C., all of which were passed to D. B. N.
182
© 2011 The Linnean Society of London, Zoological Journal of the Linnean Society, 2011, 163, 182–276
HETERODONTOSAURUS: CRANIAL ANATOMY
183
concerning the phylogenetic placement of the Heterodontosauridae, so the evidence in support of the various
hypotheses is reviewed in some detail. At present the best-supported hypothesis is the one which places
Heterodontosauridae in a basal (non-genasaurian) position; however, the evidence is not fully conclusive and
further information is still needed in respect of the anatomy of proximate outgroups, as well as more complete
anatomical details for other heterodontosaurids. Heterodontosaurids were not such rare components of the late
Karoo ecosystem as previously thought; evidence also suggests that from a phylogenetic perspective they occupied
a potentially crucial position during the earliest phases of ornithischian dinosaur evolution.
© 2011 The Linnean Society of London, Zoological Journal of the Linnean Society, 2011, 163, 182–276.
doi: 10.1111/j.1096-3642.2011.00697.x
ADDITIONAL KEYWORDS: Upper Elliot and Clarens Formations – Dinosauria – Sinemurian/Pliensbachian
– Ornithischia – phylogenetics – systematics.
INTRODUCTION
The Elliot and Clarens Formations (Upper Triassic–
Lower Jurassic) are exposed in the upper part of the
stratigraphical succession within the Karoo Basin of
southern Africa and have proved crucial to understanding the diversity of basal ornithischian dinosaurs (Norman, Witmer & Weishampel, 2004a;
Butler, Porro & Norman, 2008a). Early work was
hindered by the fragmentary nature of the discoveries: Robert Broom (1911) described the extremely
fragmentary remains of Geranosaurus atavus from
the Clarens Formation [Hettangian-Sinemurian
(Olsen & Galton, 1984) – or possibly PliensbachianToarcian (Yates, Hancox & Rubidge, 2004) – see
Systematic Palaeontology below, the date is
more probably Pliensbachian]; Sidney Haughton
(1924) described a broken jaw fragment and teeth
of Lycorhinus angustidens, which was originally
regarded as an ictidosaur (cynodont synapsid).
However, it was not until Crompton & Charig (1962)
provided a preliminary description of a newly discovered, crushed but nearly complete skull (SAM-PKK337) that the affinities of the two previously named
taxa became clear. The new taxon, named Heterodontosaurus tucki, came from the Clarens Formation and
demonstrated the presence of a distinct group of ornithischians with a strikingly mammal-like (heterodont) dentition, notable for the presence of a pair of
enlarged, caniniform premaxillary and dentary ‘tusks’
positioned in front of a row of heavily worn and
therefore bluntly truncated cheek teeth (Figs 1, 2, 4,
5 and also Appendices 3–6). Crompton & Charig
(1962) recognized the heterodontosaurid affinities of
Geranosaurus and Ly. angustidens and Kuhn (1966)
proposed that all three taxa be assigned to the family
Heterodontosauridae.
New and important ornithischian material, collected mostly from the Lower Jurassic of South Africa,
was described in the years that followed, notably
the gracile, non-heterodontosaurid ornithischians
‘Fabrosaurus’/Lesothosaurus (Ginsburg, 1964; Thulborn, 1970a, 1972, 1992; Galton, 1978; Sereno, 1991a;
Knoll, 2002a, b), Stormbergia (Butler, 2005), and the
Late Triassic Eocursor (Butler, Smith & Norman,
2007; Butler, 2010). However, heterodontosaurids
continued to be discovered, including two new taxa:
Abrictosaurus consors and Lanasaurus scalpridens
(Thulborn, 1970b, 1974; Gow, 1975, 1990; Hopson,
1975, 1980). The discovery of a an almost complete
skull and associated articulated skeleton of another
individual attributable to the genus Heterodontosaurus was reported by Santa Luca, Crompton & Charig
(1976) and its postcranial skeleton was described in
detail by Santa Luca (1980). A partial juvenile skull
was described and referred to Heterodontosaurus by
Butler et al. (2008a).
All currently known heterodontosaurid material
from southern Africa has been listed by Butler
(2010); however, heterodontosaurids are known from
other geographical areas and time periods: the Late
Triassic of Argentina (Báez & Marsicano, 1998, 2001);
the Early and Late Jurassic of the USA (Attridge,
Crompton & Jenkins, 1985; Galton, 2002, 2007;
Butler et al., 2010); the late Middle-Late Jurassic
of China (Zheng et al., 2009); and the earliest
Cretaceous of England (Barrett, 1999; Norman &
Barrett, 2002). However, with some exceptions (e.g.
Zheng et al., 2009) these remains are generally very
incomplete and most material has yet to be fully
described.
Many aspects of heterodontosaurid taxonomy, systematics, and palaeobiology remain controversial. The
fragmentary nature of much of the Elliot and Clarens
Formation material has generated unstable taxonomy
(e.g. Thulborn, 1970b, 1974, 1978; Charig & Crompton, 1974; Hopson, 1975, 1980; Gow, 1990;
Weishampel & Witmer, 1990; Norman et al., 2004a;
Xu et al., 2006; Butler et al., 2008a; Butler, Upchurch
& Norman, 2008b). The phylogenetic position of Heterodontosauridae within Ornithischia is poorly
resolved (see Butler et al., 2008b); and there is much
184
D. B. NORMAN ET AL.
Figure 1. Heterodontosaurus tucki Crompton & Charig, 1962. SAM-PK-K337 (holotype). A, right lateral view drawn from
the perspective in which the skull roof is aligned to the horizontal plane. B, annotated outline drawing of the same (even
grey tone represents adhering matrix and/or plaster infill). See also Appendix 3. Abbreviations: An, angular; aof,
antorbital fossa; apf, anterior premaxillary foramen; Ar, articular; D, dentary; emf, external mandibular fenestra; itf,
infratemporal fenestra; J, jugal; jb, jugal boss; jp, ventrolateral jugal process; Mx, maxilla; mxr, lateral maxillary ridge;
N, nasal; nf, narial fossa; nsul, internasal sulcus; paf, posterior antorbital fenestra; Pd, predentary; Pf, prefrontal; Pmx,
premaxilla; Po, postorbital; por, postorbital ridge; Ppb, palpebral; Q, quadrate; qf, quadrate (paraquadratic) foramen; Qj,
quadratojugal; Sa, surangular; sc, sagittal crest of the parietal; Sq, squamosal; sqr, squamosal ridge.
debate over several aspects of heterodontosaurid
palaeobiology: jaw action and feeding mechanisms
(Weishampel, 1984; Norman & Weishampel, 1985,
1991; Crompton & Attridge, 1986; Barrett, 1998,
2000; Porro, 2007, 2009; L. B. Porro, unpubl. data);
sexual dimorphism (Thulborn, 1974; Hopson, 1975;
Butler et al., 2008a); ontogeny and life history
strategies (Thulborn, 1978; Hopson, 1980; Butler
et al., 2008a). Clearly heterodontosaurid ornithischians are still enigmatic, in a number of respects.
Although the postcranial anatomy of H. tucki has
been described (Santa Luca, 1980), the long-promised
185
Figure 2. Heterodontosaurus tucki Crompton & Charig, 1962. SAM-PK-K337 (holotype). A, left lateral view drawn from
slightly ventrolateral perspective. B, annotated outline drawing of the same (even grey tone represents matrix and/or
plaster infill). See also Appendix 3. Abbreviations: af, internal mandibular adductor fossa; An, angular; Ar, articular; aof,
antorbital fossa; Bo, basioccipital; Bs, basisphenoid; bsf, basisphenoid flange; bst, basisphenoid tuber; Co, coronoid; D,
dentary; fm, foramen magnum (internal wall of right side) imf, internal mandibular fenestra; lbpt, left basipterygoid
process; Ls, laterosphenoid; Mx, maxilla; N, nasal; nf, narial fossa; Os, orbitosphenoid; ovc, occipital vascular canal; Pa,
parietal; Part, prearticular; Pd, predentary; pocc, paroccipital; Pro/Op, proötic-opisthotic; Prs –?presphenoid; Psp,
parasphenoid; Pt, pterygoid (right); ptf, pterygoid flange; ptmr, pterygoid medial ridge; Q, quadrate; qf, quadrate
(paraquadratic) foramen; Qj, quadratojugal; qpt, quadrate wing of the pterygoid; rbpta, right basipterygoid articular facet;
S, supraoccipital; Sp, splenial; Sq, squamosal (broken fragment); V, trigeminal fossa (cranial nerve 5).
186
D. B. NORMAN ET AL.
account of the cranial morphology (e.g. Crompton &
Charig, 1962; Charig & Crompton, 1974) did not
materialize. This paper fills that lacuna by providing
a description of the cranial osteology of the species; it
also considers the taxonomy of currently known
southern African heterodontosaurs, some of the
biological attributes that are linked to unique aspects
of its overall morphology (see also Porro,
2009 and unpubl. data) as well as re-addressing
the systematics and phylogenetics of this genus
and associated heterodontosaurids by reference to a
restricted selection of ornithischian taxa (see Butler
et al. 2008b for a more complete systematic analysis).
Two specimens: the holotype SAM-PK-K337
(Figs 1–3, 7, Appendix 3) and the referred specimen
SAM-PK-K1332 (Figs 4–7, 16–18, Appendices 4–6)
provide the main sources of information; however,
supplementary information from additional specimens has been introduced where it is considered to be
appropriate.
INSTITUTIONAL
ABBREVIATIONS
BP, Bernard Price Institute for Palaeontological
Research, Johannesburg; NHMUK, The Natural
History Museum, London; NM, Nazionale Museum,
Bloemfontein, South Africa; SAM, Iziko South African
Museum, Cape Town.
Figure 3. Heterodontosaurus tucki Crompton & Charig, 1962. SAM-PK-K337 (holotype). A, dorsal view. B,
ventral view. Even grey tone represents matrix and/or plaster infill. Abbreviations: Ar, articular; Bo, basioccipital;
bpt, ‘footplate’ of basipterygoid; Bs, basisphenoid; bsf, basisphenoid flanges; bst, basisphenoid tubers; D,
dentary; imf, internal mandibular foramen; jb, jugal boss; Mx, maxilla; N, nasal; nsul, internasal sulcus;
Pa, parietal; Pd, predentary; Pf, prefrontal; Pmx, premaxilla; Po, postorbital; Ppb, palpebral; Pt, pterygoid;
pft, pterygoid flange; pocc, paroccipital; Q, quadrate; Qj, quadratojugal; Sa, surangular; Sp, splenial; Sq,
squamosal.
SYSTEMATIC PALAEONTOLOGY
DINOSAURIA OWEN, 1842
ORNITHISCHIA SEELEY, 1887
FAMILY: HETERODONTOSAURIDAE KUHN, 1966
GENUS: HETERODONTOSAURUS CROMPTON &
CHARIG, 1962
Generic diagnosis: As for species (below)
Type
1962
species:
H.
tucki
Crompton
&
Charig,
Synonymy
Lycorhinus tucki (Crompton & Charig) Thulborn,
1970a: 244.
Lycorhinus tucki (Crompton & Charig) Thulborn,
1974: 161.
Generic and specific characteristics
General: Basal ornithischian dinosaur, known
stratigraphical range: Sinemurian/Pliensbachian.
Cranial (*indicates autapomorphy – see also discussion in Phylogenetic Relationships section): Deep
buccal emargination is formed by a strongly dorsoventrally compressed and transversely expanded maxillary ridge, which forms the ventral margin of the
external antorbital fenestra and is thickened along its
lateral margin*; antorbital fossa extends posteriorly
to form a channel on the external surface of the
jugal*; quadratojugal forms a thin wing that overlaps
the entire external surface of the quadrate (contacting
the squamosal dorsally and terminating ventrally just
above the articular condyle) and contacts the jugal via
a narrow bridge of bone*; quadratojugal has a constricted scarf suture with the jugal*; narrow and
obliquely orientated ventral jugal projection closely
aligned against the lateral surface of the lower jaw*;
prominent laterally expanded ‘boss’ on the jugal*;
sharply defined curved ridge on the external surface
of the postorbital that is continuous with a similar
ledge on the dorsolateral margin of the squamosal*;
remnants of intracranial pneumatism preserved as
pits on the paroccipital process and quadrate, and as
sinuses on the jugal boss and anteromedial process of
the maxilla*; narrow and deep pterygoid flanges lie
close to the medial surface of the lower jaw (forming
a slot-like guide with the ventral process of the
jugal)*; paroccipital wings perforated by a discrete
vascular/neural canal*; basisphenoid flanges are
large, oblique and extend medial to the pterygoids
and enclose narrow fossae on either side of the
ventral midline of the braincase; surangular develops
two finger-like rami that form much of the dorsal
margin of the coronoid eminence anterior to the jaw
joint*; elongate, slot-shaped surangular foramen*;
187
broad depression on the lateral surface of the
angular*.
Dentition (*indicates autapomorphy): Premaxillary
and dentary caniniforms have fine, blunt, serrations
(six per mm) running down their posterior margins;
premaxillary caniniform lacks serrations along its
anterior edge; dentary caniniform has widely spaced,
rounded denticulations running down the upper
portion of its anterior edge*; columnar maxillary and
dentary teeth have crowns that are only slightly
expanded either anteroposteriorly or transversely
above the root (the ‘cingulum’ and ‘neck’ at the crownroot junction are completely absent)*; labial surface of
maxillary crowns possess three prominent ridges that
separate equal-sized, clearly defined excavated
regions*; lingual surface of dentary crowns display a
mesially offset principal ridge and crown margins
that create subequal adjacent crown areas*; extensive
wear facets on the upper and lower dentitions display
a warp because successive teeth are worn at differing
angles*.
Postcranial characters (*indicates autapomorphy –
see also discussion in Phylogenetic Relationships
section): (derived from Santa Luca, 1980 – with additions and modifications) axial vertebral column: 21
vertebrae (9 cervical, 12 dorsal)*, sacrum: 6 fused
vertebrae*, caudal vertebrae: 34+; prominent epipophyses present on anterior cervical postzygapophyses*, ossified tendons distributed across the neural
spines of dorsal and sacral vertebrae only; scapular
blade narrow and elongate with expanded distal
(extrascapular) portion; humerus with a large deltopectoral crest and large entepicondyle*; humerus
lacks a posterior (olecranon) fossa; ulna with prominent olecranon; manus length more than 40% of the
combined length of humerus and radius; nine carpal
bones; manus digits 1–3 parallel, digits 4–5 reduced in
size and divergent; penultimate phalanges of digits 2
and 3 more elongated than the proximal phalanges;
extensor pits present on the dorsal surface of distal end
metacarpals and phalanges; manual unguals strongly
recurved, and with prominent flexor tubercles. Ilium,
with a narrow vertical facet on the ischial peduncle
that resembles an avian antitrochanter*; prepubic
process short and deep, postpubis as long as ischium;
obturator process absent; ischial shaft marked by an
elongate lateral ridge that is drawn out to form a
prominent lateral shelf along the mid-section of the
shaft*; femoral greater and anterior trochanters not
separated by a cleft; transverse axis of distal femoral
articular surface obliquely orientated; fibula reduced
and fused to tibia distally*; astragalus and calcaneum
fused*; astragalocalcaneum fused to the distal ends of
tibia and fibula*; three distal tarsals present but fused
to proximal ends of their metatarsals*; metatarsals
1–4 fused together*.
188
D. B. NORMAN ET AL.
189
Figure 4. Heterodontosaurus tucki Crompton & Charig, 1962. SAM-PK-K1332 (referred specimen). A, right lateral view
with lower jaw articulated. B, outline annotated drawing of the same, see also Appendices 4A and 5C. Abbreviations: aaf,
anterior antorbital fenestra; An, angular; aof, antorbital fossa; apf, anterior premaxillary foramen; bc, basioccipital condyle;
Bo, basioccipital; bst, basisphenoid tuber; D, dentary; dia, diastema; emf, external mandibular fenestra; F, frontal; f-po,
frontal-postorbital suture; J, jugal; jb- jugal boss; La, lacrimal; lPt, left pterygoid; lPpb, left palpebral; lQ, left quadrate; Ls,
laterosphenoid; Mx, maxilla; mxr, lateral maxillary ridge; N, nasal; nf, narial fossa; olf, channels on ventral surface of
frontals for the olfactory bulbs; orb, recesses forming the dorsal surface of the eye socket; Pa, parietal; paf, posterior
antorbital fenestra; Pd, predentary; Pf, prefrontal; Pmx, premaxilla; Po, postorbital; pocc, paroccipital; por, postorbital ridge;
Ppb, palpebral (right); Pro/Op, prootic-postorbital; Psp, parasphenoid; Pt, pterygoid (right); Q, quadrate (right); Qj,
quadratojugal; Sa, surangular; sc, sagittal crest (of the parietal); Sq, squamosal.
䉳
Holotype
SAM-PK-K337 – Iziko South African Museum, Cape
Town (Figs 1–3, 7, 20–22 and Appendix 3); see also
Crompton & Charig, 1962: fig. 1; Charig & Crompton, 1974: figs 10, 11; Galton, 1986: fig. 16.6r,s; Báez
& Marsicano, 2001: fig. 5C). Nearly complete skull
and lower jaw embedded in greyish-yellow sandstone. The surface of the skull is encrusted by an
adherent, and very tough, layer of haematite that
has been removed partially using a small diamond
saw. Most of the preparation was carried out by
Arthur E. Rixon, formerly in charge of the Palaeontological Laboratory of the British Museum
(Natural History); one of the authors (A. W. C.) continued preparation as far as seemed prudent at the
time. Postcranial remains were also listed by
Crompton & Charig (1962: 1075) but the whereabouts of this potentially extremely important material is currently unknown.
During fossilization the specimen has been compressed laterally. The skull roof slopes steeply
toward the right (at an angle of about 35° from the
horizontal, the latter being taken as normal to the
sagittal plane – see Fig. 7). Structures on the right
side have been displaced ventrad and a little backwards with respect to those in the mid-line. The
skull is nevertheless reasonably well preserved on
the right side (Fig. 1, Appendix 3A). A short section
of the dentary ramus, just behind the dentary–
predentary contact, is missing. On the left side the
superficial bones of the skull have largely been
eroded away, revealing parts of the palate and
braincase; the only trace of the left lower jaw is the
anterior tip of the dentary and adjacent predentary
(Fig. 2, Appendix 3B).
Provenance: 1890 m above sea-level on the mountain
behind Tyindini trading store, Herschel District,
Eastern Cape Province, Republic of South Africa
(30°32′S, 27°32′E; Kitching & Raath, 1984: table 1).
Discovered by A. W. C. during the 1961/1962 joint
British/South African expedition to the Upper Triassic
outcrops in South Africa and Basutoland (= Lesotho).
Stratigraphical occurrence: Clarens Formation
(formerly ‘Cave Sandstone’, Stormberg Series):
Hettangian-Sinemurian (Olsen & Galton, 1984),
Pliensbachian-Toarcian (Yates et al., 2004). The most
probable age for the Clarens Formation: latest
Sinemurian-Pliensbachian (Jourdan et al., 2005,
2007, 2008).
Referred specimens
SAM-PK-K1332
Articulated skull, lower jaw (Figs 4–7, 16–18, 23–27
and Appendices 4–6; see also Santa Luca et al., 1976:
fig. 1; Santa Luca, 1980: figs 1, 2; Weishampel &
Witmer, 1990: fig. 23.4; Norman et al., 2004c:
fig. 18.10A), and postcranial skeleton (Santa Luca
et al., 1976: fig. 1; Santa Luca, 1980: figs 1, 3–22).
Preparation of this skull was undertaken by Mrs Ione
Rudner at the Iziko South African Museum, Cape
Town. A considerable number of minor breaks (probably having their origin in post-mortem deformation,
as well as some caused by the mechanical preparation
and by later mishandling) have been repaired by
adhesive; additionally, a moderate-to-thick layer of
adhesive-consolidant has been applied to the surfaces
of bones and teeth, which obscures finer anatomical
details (consolidant on the occlusal surfaces of the
teeth has since been removed by L. B. P.). A detailed
photographic record (including stereo-pair images)
made at the time of the original preparation for A. W.
C. has been archived at the Sedgwick Museum, Cambridge, UK.
The specimen was preserved in red sandstone. The
skull (as with the holotype) has suffered lateral crushing and displacement; however, it is more complete
and its preservation overall is better than that of the
holotype. Unlike the holotype, the skull was not
covered in a layer of haematite.
Provenance: At an altitude of about 1770 m on the
northern slopes of Kromspruit (alternative spelling
Krommespruit) Mountain, Voyizane (= ‘Voisana’;
30°34′S, 27°26′E; Kitching & Raath, 1984: table 1), Herschel District, Eastern Cape Province, South Africa.
190
D. B. NORMAN ET AL.
Figure 5. Heterodontosaurus tucki Crompton & Charig, 1962. SAM-PK-K1332 (referred specimen). A, left lateral view of
the skull, lower jaw removed. B, outline annotated drawing of the same, see also Appendix 4B. Abbreviations: aaf, anterior
antorbital fenestra; aof, antorbital fossa; bc, basioccipital condyle; J, jugal; jb, jugal boss; jp, ventrolateral jugal process;
La, lacrimal; Mx, maxilla; mxr, lateral maxillary ridge; N, nasal; nf, narial fossa; nsul, internasal sulcus; Pa, parietal; paf,
posterior antorbital fenestra; Pf, prefrontal; Po, postorbital; por, postorbital ridge; pocc, paroccipital; Ppb, palpebral; Psp,
parasphenoid; Pt, pterygoid; ptf, pterygoid flange; ptq, pterygoid wing of the quadrate; Q, quadrate; qf, quadrate
(paraquadratic) foramen; qpt, quadrate wing of the pterygoid; rJ, right jugal; rPo, right postorbital; Sq, squamosal; sqr,
squamosal ridge.
191
Figure 6. Heterodontosaurus tucki Crompton & Charig, 1962. SAM-PK-K1332 (referred specimen). A, dorsal view of the
skull. B, ventral view of the skull, see also Appendices 4C,D and 5A,B. Abbreviations: Bo, basioccipital; bpt, basipterygoid;
Bs, basisphenoid; bsf, basisphenoid flange; dia, diastema; F, frontal; f-po, frontal-postorbital suture; J, jugal; jb, jugal boss;
N, nasal; nf, narial fossa; nsul, internasal sulcus; Mx, maxilla; P, parietal; Pf, prefrontal; Pmx, premaxilla; Po, postorbital;
pocc, paroccipital; Ppb, palpebral; Pt, pterygoid; ptq, pterygoid wing of the quadrate; Q, quadrate; Qj, quadratojugal; qpt,
quadrate wing of the pterygoid; Sq, squamosal.
Discovered by A. W. C. on the 1966–67 joint South
African Museum, Yale University, British Museum
(Natural History), University of London expedition to
the ‘Red Beds’ of South Africa and southern Lesotho
(Attridge & Charig, 1967; Crompton, 1968).
Stratigraphical occurrence: From the upper part of
the Elliot Formation (formerly ‘upper Red Beds’),
Stormberg Series; probably of Early Jurassic age
(stratigraphically lower than the horizon at which the
holotype was found): Hettangian-Sinemurian (Olsen
& Galton, 1984) or Pliensbachian-Toarcian (Yates
et al., 2004). The most probable age for the upper part
of the Elliot Formation is upper Sinemurian (R. Irmis
pers. comm. 2010).
SAM-PK-K1334
Partial left maxilla containing seven erupted crowns
and three replacement crowns, with attached fragments of the jugal and lacrimal bones (Figs 30–33).
Provenance: Site 18a, Voisana, Kromspruit 9 Farm,
Herschel District, Eastern Cape Province, South
Africa (30°34′S, 27°26′E; Kitching & Raath, 1984).
Collected on the 1966–67 expedition of the South
African Museum, Yale University, British Museum
(Natural History) and University of London. Reports
and letters exchanged between A. J. C. and A. W. C.
suggest that it was recovered from the same locality
(Voyizane/Voisana) in Herschel District that yielded
the complete, articulated heterodontosaur skeleton
(SAM-PK-K1332) and several other specimens.
Stratigraphical occurrence: Upper Elliot Formation
(as above).
SAM-PK-K10487
Partial skull (Figs 28, 29; see also Butler et al., 2008a:
figs 1–3, 4B, 5) of a demonstrably ontogenetically
immature specimen.
192
D. B. NORMAN ET AL.
Figure 7. Heterodontosaurus tucki Crompton & Charig, 1962. Occipital views. A, SAM-PK-K337 (holotype). B, SAMPK-K1332 (referred specimen). Sheared surfaces that cut obliquely across the posterior left portion of the holotype skull
– SAM-PK-K337 – are indicated by even grey tone, as is the foramen magnum, which is partially plugged with matrix.
Abbreviations: Bo, basioccipital (condyle visible in B); bpt, basipterygoid process; bst, basisphenoid tuber; Ex, exoccipital;
jb, jugal boss; lbpt, left basipterygoid process; nc, nuchal crest; Op, wing of opisthotic (A, left opisthotic sheared off at its
base against the remainder of the braincase); ovc, occipital vascular canal; Pa, parietal; ?pn, possible pneumatic opening
in quadrate; Po, postorbital; pof, post-temporal fenestra; pocc, paroccipital process; ptf, pterygoid flange; Q, quadrate;
qf, quadrate (paraquadratic) foramen; Qj, quadratojugal; rPt, right pterygoid; S, supraoccipital; Sq, squamosal.
Provenance: Kromspruit area, Herschel District,
Eastern Cape Province, South Africa.
Stratigraphical occurrence: Probably upper Elliot
Formation.
Specimens provisionally referred to
Heterodontosaurus sp. pending further study
NMQR 1788: Partial skull, undetermined stratigraphical horizon, Tushielaw, Barkly East, Eastern Cape Province, South Africa (L. B. Porro et al., in press).
SAM-PK-K10488: Partial lower jaw with evidence of
replacement crowns, and fragments of upper jaw.
Upper Elliot Formation, Eastern Cape Province.
(L. B. Porro, unpubl. data).
DESCRIPTIVE ANATOMY
Explanatory note on anatomical orientation
The skull of Heterodontosaurus is elongate and laterally compressed as shown in the full reconstructions
(Figs 8–14). The general shape and horizontal pose of
193
Figure 8. Heterodontosaurus tucki Crompton & Charig, 1962. A. Skull reconstruction in left lateral view. B. Annotated
outline. Abbreviations: aof, antorbital fossa; An, angular; Ar, articular; Bo, basioccipital; Bs, basisphenoid; Co, coronoid;
D, dentary; F, frontal; J, jugal; jb, jugal boss; jp, ventrolateral jugal process; La, lacrimal; Ls, laterosphenoid; Mx, maxilla;
N, nasal; nf, narial fossa; Pa, parietal; paf, posterior antorbital fenestra; Pd, predentary; Pf, prefrontal; Pmx, premaxilla;
Po, postorbital; Ppb, palpebral; Pro/Os, prootic-opisthotic; Psp, parasphenoid; ptq, pterygoid wing of the quadrate; Q,
quadrate; qf, quadrate (paraquadratic) foramen; Qj, quadratojugal; qpt, quadrate wing of the pterygoid; Sa, surangular;
Sq, squamosal.
194
D. B. NORMAN ET AL.
Figure 9. Heterodontosaurus tucki Crompton & Charig, 1962. Reconstruction of the skull in left lateral view. Bones of
the upper jaw and suspensorium have been highlighted for clarity. For full list of abbreviations see end of paper.
the skull, and the absence of teeth from the anterior
end of both jaws, create a simple orientational framework. As a consequence traditional anatomical
descriptors such as anterior, posterior, dorsal, ventral,
medial, and lateral are, without exception, unambiguous and more obscure (and confusing) terminology
(Harris, 2004) can be avoided. With respect to tooth
morphology, the general conventions are followed:
mesial (toward the jaw symphysis), distal (toward the
jaw articulation), lingual (toward the tongue), labial
(toward the presumed location of lips), apical (toward
the occlusal edge of the crown), and thecal (toward
the floor of the tooth socket).
GENERAL
DESCRIPTION OF THE SKULL
The skull is small, its greatest length (from snout to
posterolateral tip of opisthotic) being 108 mm in the
holotype (SAM-PK-K337, Appendix 3) and 121 mm in
the referred specimen (SAM-PK-K1332, Appendix 4).
Laterally (Fig. 8) the skull forms a low triangular
wedge; the highest point is at the posterior end of the
sagittal crest, from which the upper border descends
gradually before curving abruptly toward the anterior
tip; a narial fossa depresses the lateral surface of the
premaxilla (Figs 1, 2, 4, 5, 8, nf). The upper portion of
the posterolateral border of the skull is convex and
terminates in a blunt, ventrally deflected, hook
formed by the squamosal and paroccipital wing of the
opisthotic (Figs 4, 5, 8, 15, pocc); beneath the paroccipital ‘hook’ the posterior edge of the skull is offset
anteriorly by the pillar-like quadrate, which descends
beyond the level of the maxillary tooth row (Figs 1, 5,
9). The lower border of the skull (Fig. 9) is interrupted
anteriorly by a diastema between the maxilla and
premaxilla (Fig. 4, dia); behind the maxillary dentition there is another deeply arched recess between
the jugal process (Figs 1B, 8, jp) and the quadrate.
Details within the lateral face of the skull include the
large and almost circular orbit, the upper portion of
which is partially and obliquely bisected by the long,
tapering palpebral (Ppb). Anteroventral to the orbit
there is a well-defined, triangular external antorbital
fenestra framing an extensive fossa (aof), the inner
wall of which is itself perforated by two subsidiary
fenestrae (Figs 1, paf; 4, aaf); the ventral edge of the
fenestra is formed by a prominent shelf or maxillary
ridge (mxr), which projects laterally from the body of
the maxilla and marks the dorsal margin of the very
prominent buccal (cheek) recess. A prominent somewhat irregular ‘boss’ (Figs 1B, 4B, 5B, jb) on the
lateral surface of the jugal is found at the posterior
end of the maxillary ridge (mxr); the position of the
jugal boss on the body of the jugal differs slightly
195
Figure 10. Heterodontosaurus tucki Crompton & Charig, 1962. Attempted reconstruction of the skull in left lateral view
with the bones of the snout sectioned along a sagittal plane. The left suspensorial elements have been removed to reveal
the anatomy of the braincase elements and palate bones. Note: contrary to appearance the left basipterygoid process (lbpt)
is not in articulation with the right basal articulation (rba). For full list of abbreviations see end of paper.
between that on the holotype and referred skull
(compare Figs 1, 4, 5) and this is regarded, in the
absence of sufficient comparative material, as an
intraspecific variation. Beneath this feature the jugal
forms a posteroventrally tapering blade (jp) that,
along with the prominent pterygoid flange, appear to
have formed a narrow slot that would have guided the
motion of the lower jaw. Behind the orbit the
infratemporal fenestra (Fig. 1B, itf) is large and ovoid
in outline, tapering anteroventrally and with its longaxis orientated obliquely. A prominent ridge (sqr, por)
runs forward from the posterodorsal corner of the
skull formed by the paroccipital wing, along the
dorsal margin of the interfenestral bar before curving
downward along the edge of the postorbital bar; this
ridge encloses a depression that frames the infratemporal fenestra (a less developed but similar ridge has
been observed on the postorbital of the basal iguanodontian Zalmoxes – Weishampel et al., 2003).
In dorsal aspect (Figs 3, 6, 12) the anterior half of
the outline of the skull is narrow and triangular,
tapering toward the anterior tip. There are three
embayments along the dorsal profile: one near the tip
of the snout as a consequence of the swollen buttresses for the roots of the premaxillary caniniform
teeth; the upper margin of the orbit is strongly emar-
ginated (and notably rugose); and the intertemporal
bars appear to be slightly bowed medially. The
supratemporal fenestrae (stf) are parasagittally
aligned ellipsoids separated by the prominent (parietal) sagittal crest (sc). The anterior skull roof is
formed by the frontals and is slightly concave, with a
slight midline ridge continuous with the sagittal
crest. Further anteriorly, the internasal suture sinks
along the midline to generate a distinct median
furrow (the nasal sulcus – Figs 1, 3, 5, 12, nsul, see
also Appendix 4C,D). At the anterior tip, the lateral
margins of the nasals roll inward to occlude the nasal
sulcus and form a fused median prong that overhangs
the narial fossa.
In ventral aspect the palate (Figs 6B, 13 – full
reconstruction) is narrow and tapers toward the premaxillary beak. There is a narrow premaxillary
palatal roof that projects posteriorly to form a ledge
that accommodates a median anterior extension of
the paired vomers (V); the latter form a narrow keellike septum that marks the inner wall of the nasal
passages. The palatines (Pal), as restored, back the
vomers and form an arched plate beneath the anterior portion of each orbital cavity; computed tomography (CT) scans of SAM-PK-K1332 suggest that the
dorsal medial edges of the palatines are firmly
196
D. B. NORMAN ET AL.
Figure 11. Heterodontosaurus tucki Crompton & Charig, 1962. Attempted reconstruction of the skull and the articulated
right lower jaw in left lateral view, revealed following sagittal sectioning and removal of the right vomer. Note: the
contours of the floor of the braincase and its internal walls are conjectural; the thickness and anterior extent of
the supraoccipital are unknown and have been illustrated very conservatively; the anatomy of the contacts between the
vomer, palatine, pterygoid and maxilla in the anterior-central part of the snout are conjectural. For full list of
abbreviations see end of paper.
sutured (L. B. P., pers. observ.). As restored, the
ectopterygoids (Ec) and pterygoids (Pt) contribute to
the prominent and robust pterygoid flanges; the
flanges resemble those of crocodilians in that they
project ventrally in close proximity to the medial
surface of the lower jaw (see Figs 2, 11) although they
do not form a medially fused ‘apron’ as in crocodilians.
Posterodorsal to these flanges each pterygoid expands
(wing-like) dorsoventrally and overlaps the equally
deep pterygoid wing of the quadrate to form an
oblique, near vertical, sheet that forms the inner wall
of the temporal adductor muscle chamber. Dorsal to
the pterygoid flanges there is a robust shelf area that
projects medially and appears to be rugose (Fig. 2); it
is uncertain (because of distortion in SAM-PK-K1332)
whether the pterygoids where sutured along the
entirety of the midline. The robust shelf supports the
raised articular facets of the basal articulation, which
receive the oblique, anteroventrally directed basipterygoid processes; the posterolateral margins of
the latter flare dorsally and laterally as thin flanges
that screen the lateral surface of the portion of the
braincase beneath the trigeminal fossa. The anteromedial edges of the basipterygoid processes coalesce
to form the base for the long, narrow basisphenoid/
parasphenoid rostrum.
The occiput (Figs 7, 14) is broad and deep, and
formed by massive and oblique paroccipital wings
that project obliquely (posterolaterally) from the
midline; these are capped by massive squamosals
that, in turn, flank the narrow occipital condyle/
basioccipital, supraoccipital and parietals. A midline
nuchal crest (nc) bisects the supraoccipital; it arises
from a thickened lip of the supraoccipital that forms
the dorsal rim of the foramen magnum. The dorsal
portion of the supraoccipital locks into the poster-
197
Figure 12. Heterodontosaurus tucki Crompton & Charig, 1962. Attempted reconstruction of the skull in dorsal view. For
full list of abbreviations see end of paper.
Figure 13. Heterodontosaurus tucki Crompton & Charig, 1962. Attempted reconstruction of the skull in ventral view
(lower jaw removed). Note: the anatomical configuration of the anterior palate is conjectural; the proximity of the medial
edges of the pterygoids is uncertain, although they seem unlikely to have been sutured in place, but may have been closer
together than shown here, and linked by connective tissue, in life; the thin vertical sheets of bone that form the medial
wall of the adductor chamber (pterygoid and quadrate wings, and the much-expanded flange on the lateral surface of the
basisphenoid) appear to have been quite close. For full list of abbreviations see end of paper.
oventral edge of the parietal. The occipital plate is
almost entirely framed by a thickened edge that runs
from the lateral region of the foramen magnum until
it merges with the posterior edge of the parietal. The
paroccipital wings are perforated by discrete foramina
(ovc) and the lower corner of the ‘hook’ is indented
with what might represent a pneumatic opening (?pn
– CT scans lack clear resolution in this area and do
not immediately confirm a connection between this
pit and an adjacent sinus). The post-temporal fenestrae (ptf) are reduced to narrow channels at the
junction between the supraoccipital, squamosal, and
opisthotic. The quadrates were held against the
anterolateral surface of the squamosals and paroccipital wings by ligamentous attachment judged by the
rugosity of the adjacent otoccipital surface (A. W. C.
198
D. B. NORMAN ET AL.
Figure 14. Heterodontosaurus tucki Crompton & Charig,
1962. Attempted reconstruction of the skull in occipital
view. The post-temporal fenestra (pof) appears to be
reduced to a narrow channel that enters the body of the
neurocranium between the parietal, supraoccipital and
(perhaps the squamosal more posteriorly). The smaller,
but discrete vascular opening on the paroccipital wing is
assumed to have served for passage of parts of the cranial
vascular system (Romer, 1956). The indented regions (?pn)
on the lower lateral corner of the paroccipital wing, and
the other on the rear surface of the quadrate shaft, medial
to the quadrate (paraquadratic) foramen, hint at the presence of cranial pneumatism (Witmer, 1997). For full list of
abbreviations see end of paper.
photographic archive), although the convex and
smooth head of the quadrate fits into a smoothly
arched depression on the ventral surface of the squamosal. The main body of the quadrate is a curved and
vertically grooved pillar; the anterior side is deep and
broadly grooved so that its medial and lateral edges
support the pterygoid and quadratojugal sheets; posteriorly the shaft is notched laterally by the quadrate
(paraquadratic) foramen and medial to this the
surface is recessed and may also house a remnant
pneumatic opening (?pn) – although, again, there is
1962. A, reconstruction of the neurocranium in left lateral
view (with surrounding bones indicated as transparent).
B, outline based on (A) above with annotations to identify
specific structural features. Reconstruction based on the
skulls of SAM-PK-K337 and K1332 supplemented by
stereo photographs prepared for A. W. C. in anticipation of
the earlier planned publication. Notes: the suture between
the lateral walls of the braincase and the floor is unclear;
the sutures between the laterosphenoid, proötic and
opisthotic are not discernible; the extent of the dorsal
lappet that appears to project from the dorsomedial edge
of the ?proötic is uncertain (it remains a possibility that
the ‘lappet’ represents an anterior extension of the
supraoccipital); the details of the anatomy of the basisphenoid flange in the proximity of the trigeminal fossa are
unclear – even in the original stereo pairs prepared for
A. W. C. For full list of abbreviations see end of paper.
199
Figure 16. Heterodontosaurus tucki Crompton & Charig, 1962. SAM-PK-K1332 (referred specimen). Left lower jaw in
(A) lateral view (B) annotated outline drawing (see Appendix 6A). For full list of abbreviations see end of paper.
no CT evidence of an adjacent sinus in the fabric of
the quadrate. The remainder of the quadrate shaft
extends ventrally and curves slightly anteriorly to
terminate at the transversely expanded articular
condyles; the latter are slightly bicondylar and orientated such that the axis of rotation implied from their
structure is medioventral and twisted anteromedially
(Fig. 13).
The lower jaw (Figs 1, 4, 16–19, Appendices 5C,D,
6A-D) tapers anteriorly and the symphysial region is
capped by a triangular predentary. The principal component of the lower jaw (the dentary) is conspicuously
robust; its upper border curves gently, and then more
steeply, to the coronoid eminence that is positioned
(when the jaw is in articulation – Fig. 8) beneath the
posterior portion of the orbit. A crescentic, bevelled
ledge on the outer surface of the jaw, beneath the
tooth row mirrors the ‘cheek’ recess of the maxillajugal (Fig. 8). The coronoid eminence marks the posterior limit of the tooth-bearing part of the jaw; the
postdentary bones are notably less robust. A large,
depressed area on the outer surface of the angular –
bounded ventrally by a curved ledge, the angular
ridge (Fig. 16, anr) and dorsally by the lower anterior
surangular ramus – is perforated by a small, oval
external mandibular fenestra (emf), adjacent to the
dentary suture. The upper margin of the postdentary
part of the lower jaw has the form of a narrow
finger-like process formed from the surangular; this
curves at first quite gently and then more steeply as
it approaches the articular glenoid. A surangular
‘foramen’ (Figs 16–19, sfor) – although structurally
this is a long narrow slot between the two forwardly
projecting surangular rami – extends forward to
contact the dentary and is not fully visible in lateral
view and communicates with the adductor fossa
(Fig. 17B, af). There is a robust, horizontally
orientated, retroarticular process formed by the surangular, articular and prearticular. Medially, the
mandible has a well-developed adductor fossa and a
small, circular Meckelian foramen (internal mandibular fenestra – imf) between the splenial and prearticular, anterior to the adductor fossa.
The extreme tip of the premaxillary upper jaw is
toothless and slightly pendulous; behind this edentulous region there follow, in succession, two small
incisiform teeth, a larger caniniform tooth, a
diastema, and then a series of 11 closely packed and
heavily worn maxillary ‘cheek’ teeth that are graded
in size: smaller anteriorly, largest in the middle and
again smaller posteriorly (e.g. Figs 8, 21A). The maxillary teeth are heavily coated by enamel labially and
the crowns have a highly distinctive shield-like shape.
In the lower jaw the predentary is entirely toothless
and, immediately behind the predentary, the first
dentary tooth is caniniform and projects into the
premaxilla-maxilla diastema (Figs 4, 13 – dia); this
tooth did not occlude with the opposing premaxillary
200
D. B. NORMAN ET AL.
Figure 17. Heterodontosaurus tucki Crompton & Charig, 1962. SAM-PK-K1332 (referred specimen). Left lower jaw in
(A) medial view (B) annotated outline drawing, see also Appendix 6B. For full list of abbreviations see end of paper.
Detailed comparative comments relating to skull
morphology in this species are located in the discussions to be found in sections (below) on the Taxonomy
of South African heterodontosaurids and the Phylogenetic Analysis.
OSTEOLOGY:
DERMAL BONES OF THE FACIAL REGION
AND SKULL TABLE
1962. SAM-PK-K1332 (referred specimen). Right lower
jaw (anterior portion only) in (A) lateral and (B) medial
views, see also Appendix 5C,D. For full list of abbreviations see end of paper.
caniniform. The dentary dentition matches that of the
maxilla in general characteristics (except for the
reversal of the thickly enamelled surfaces which lie
lingually on the crowns and are similarly (but not
identically) sculpted into shield-like surfaces).
The tooth-bearing portion of the premaxilla (Figs 1–8,
Pmx) is offset ventrally relative to the maxillary tooth
row and is sutured to the remainder of the snout by
a conjoined posteromedian ‘process’ that is wedged
between the anterior extremities of the maxillae (and
median vomers – see Fig. 13); laterally and posterodorsally, a long tapering blade (the posterolateral
premaxillary process) extends between the nasals
and maxillae before contacting the lacrimal and
terminating close to the prefrontal. CT scans of
SAM-PK-K1332 indicate that the maxilla fits into a
slot anteriorly and that this is formed along the
butt-jointed sutural surface between the premaxillae,
but that this suture becomes bevelled (scarf-like)
posteriorly.
The anterior border of the premaxilla extends dorsally as a tapering, median, finger-shaped process
(the tip of which appears to be broken in both the type
and referred specimens – Figs 1, 4, 5) that may not
201
Figure 19. Heterodontosaurus tucki Crompton & Charig, 1962. Attempted reconstruction of the left lower jaw in
(A) lateral view and (B) medial view. For full list of abbreviations see end of paper.
create a complete bridge above the external nares by
contacting the anterior tip of the conjoined nasals as
reconstructed in Figure 8 – the apparent gap in this
region must have been short and was, in all probability, bridged by connective tissue. The lower edge of
the anterior border is thickened and its surface
rugose for the attachment of a small keratinous beak
(rhamphotheca) and curves posteroventrally creating
a slightly pendulous ‘droop’ relative to the alveolar
margin immediately behind; the premaxillary margin
seems likely to have been scalloped slightly where it
forms the lateral walls of the alveoli for the three
premaxillary teeth, although this area is broken in
both the holotype and referred skulls. Immediately
behind the caniniform premaxillary tooth the body of
the premaxilla contracts sharply medially to create a
rim (continued on the body of the maxilla) that
defines the margin of a pronounced embayment
(diastemal fossa – dia), which accommodates the
dentary caniniform when the jaws were adducted
(Fig. 8). The medial wall of the diastema is formed by
the premaxilla and maxilla, which meet along a
sinuous suture; the premaxilla forms a cuff around
the oblique, anteromedially directed, anterior maxillary process (Fig. 13: see also Butler et al., 2008a).
The lateral wall of the premaxilla, above the caniniform, is swollen laterally because it accommodates
the massive root of the underlying tooth and forms
the equivalent of a cranial buttress that angles dorsally and posteriorly (following the curvature of the
root) into the base of the posterolateral premaxillary
process and adjacent thickened lateral nasal shoulder. The posterolateral process tapers toward its
wedge-like termination (Figs 1, 4, 5) between nasal
and lacrimal but whether it abuts the anterior tip of
the prefrontal, as reconstructed in Figure 8, is conjectural. The suture between the premaxilla and nasal is
straight and forms a notch at the posterodorsal corner
of the external naris. The external margin of the naris
is ovoid and backed (except posterodorsally) by an
oval depression (the narial fossa – nf) on the body of
the premaxilla. Near the anteroventral corner of the
fossa, a small anterior premaxillary foramen (apf)
runs into the body of the premaxilla. CT scans (Butler
et al., 2008a) reveal that this foramen connects via a
canal in the body of the premaxilla with a foramen on
the internal surface of the posterolateral premaxillary
process (Figs 10, 11, ppf). The palatal roof formed by
the premaxillae is narrow and vaulted – more so in
the midline where there is a distinct channel leading
202
D. B. NORMAN ET AL.
forward into a premaxillary recess (pr – filled in the
fossil, but visible in the original specimen in the
photographic archive) that extends anterodorsally
into the roof of the premaxillary vault (Fig. 24A). The
premaxillary recess resembles a structure that was
described in Hypsilophodon (Galton, 1974: figs 4–6,
cav).
The nasals are fused along the midline [these two
bones meet at a butt joint in immature specimens,
e.g. SAM-PK-K10487 (Fig. 28) – see also Butler et al.,
2008a – and form a medially indented plate that roofs
the anterior snout (Figs 3, 5B, 6, 12, N, nsul)]. The
nasals extend posteriorly to meet the frontals above
the orbit at a V-shaped suture, with the apex formed
by the fused frontals directed anteriorly (Fig. 12);
they overlap the frontals in a scarf suture (Fig. 10),
with some evidence of ridging. The posterior tip of
each nasal is pointed and forms a wedge between the
prefrontal and frontal, adjacent to the orbital rim
(Fig. 12). In lateral view the nasal runs forward along
the margin of the prefrontal to meet the dorsomedial
edge of the posterolateral premaxillary process
(Fig. 8) and in this region its edge appears to be
thickened, forming modest ridges or ‘shoulders’
bounding the lateral margins of the dorsal surface of
the snout. The nasal-premaxilla suture is bevelled
and orientated ventromedially (ridges and grooves
are exposed along this suture on the left nasal of the
holotype in which this suture is dislocated). Further
anteriorly, the thickened margins seemingly roll
medially to create a median sulcus (nsul) that is
pinched out at the extreme anterior end where the
conjoined nasals form a narrow decurved finger-like
process above the external nares; as mentioned
earlier, this process fails to meet the median dorsal
process formed by the premaxillae in all currently
known specimens, and does not form an osseous internarial bridge as reconstructed in the contemporary,
and commensurate, ornithischian Lesothosaurus
(Sereno, 1991a: fig. 12).
The frontals are fused along the midline (a suture
that is butt-jointed in immature specimens – SAMPK-K10487 – Fig. 28) and form a tabular structure
that projects forward as a wedge between the nasals
(Figs 8, 12,F). The frontonasal suture is continuous
with the prefrontal-frontal suture that angles posterolaterally to the orbital margin. In dorsal view
the frontal portion of the orbital margin is shallowly
embayed and the margin itself is sharp and
irregularly puckered (a feature commonly seen in
ornithischians and associated with a connective
tissue connection, across the incomplete supraorbital
fenestra, to the adjacent palpebral – Fig. 12; Maidment & Porro, 2010). The frontals are sutured in the
midline to form a plate, which is slightly concave
transversely when viewed dorsally, and bisected by
the slightly raised midline sutural ridge (see Fig. 12).
The frontals widen posteriorly, above the orbit, to the
point where they contact the postorbitals; the suture
with the latter curves, in an undulating line, posteriorly and crosses a bevelled edge that marks the
anterior margin of the supratemporal fenestra and its
adjacent fossa, beyond which it curves smoothly medially to contact the fused parietals in a prominently
marked transverse interdigitating suture; this
feature is complete obscured in the holotype because
of a combination of breakage, missing elements and
crushing that causes the right postorbital to overlap
the adjacent frontal. The ventral surface of the frontals (visible through the orbit in Fig. 4) is marked by
an hour-glass-shaped, vaulted structure (marking a
constriction that separated the midline olfactory lobes
from the laterally positioned roof to each orbital
cavity) that is contained on either side by ridges (to
which were attached connective tissue sheets forming
the interorbital septae) that continue along the
ventral surface of each prefrontal and also extend
posteriorly on the postorbitals. A concave facet on the
ventral frontal-postorbital suture marks the contact
with the anterolateral tip of the laterosphenoid. CT
scans of an immature specimen (Butler et al., 2008a)
indicate that within the thicker parts of the frontal
plate there is trabecular cortex.
The parietals (Figs 2–5, 8, 12, 14, 15, Pa) are fused
along the midline of the roof of the braincase and
form the curved inner walls of the supratemporal
fenestrae (Fig. 12). There is a well-marked transverse
suture with the frontals anteriorly, whereas ventrally
the parietals contact the lateral walls of the braincase
formed successively by the laterosphenoids, proötics,
‘otoccipitals’ (fused opisthotics and exoccipitals), and
supraoccipital. The parietals are flattened transversely across the area where they meet the frontals;
however, further posteriorly their surface is constricted transversely to create a narrow, elevated sagittal crest. Posteriorly, the parietals form a rim that is
thickened and incised in the midline where they overhang the supraoccipital (the latter being ‘locked’ into
the ventral surface of the parietals by means of a
median ascending process as revealed by CT scanning
and as predicted by comparative embryology: Romer,
1956). The upper edge of parietals flares laterally to
border the posteromedial part of the supratemporal
fenestrae and sutures abruptly against medially projecting ‘wings’ from the flanking squamosals. The
posterior rim of the skull table is formed by the
continuous thickened edge of the parietals, squamosals, and paroccipital wings that, in effect, frame the
entire occiput (Figs 7, 14). The holotype (SAM-PKK337) and referred specimen (SAM-PK-K1332)
exhibit traces of a triangular fenestra (see Figs 8, 15,
vpar) that lies in the suture between the proötic and
parietal, and adjacent to the position where it would
be expected that the temporal opening of the channel
from the post-temporal foramen/fissure would lie; this
fenestra in all probability marks the opening for
venous drainage from the occiput and temporal region
into the medial cerebral vein (Norman, 1980, 1986;
Fig. 15, vpar).
The prefrontal (Figs 1, 4, 5, 8, 12, 28A, Pf) is a
curved and tapering bone that forms the sharp and
rugose anterodorsal rim of the orbit. Medially, the
bone becomes thicker and is overlapped by the frontal
and nasal along a curved but slightly irregular
margin, which terminates (in all probability – see
Fig. 5) anteriorly against the tip of the premaxilla
and lies dorsal to the lacrimal; the latter is a scarf
suture, orientated dorsolaterally and increasing in
width and depth toward the orbit margin. Adjacent to
the orbital margin the prefrontal is drawn out into a
short ventral extension that appears to support the
orbital portion of the lacrimal. The lower corner of the
prefrontal immediately in front of the orbit has a
shallow facet (that extends on to the adjacent body of
the lacrimal) for the articulation of the palpebral; the
latter has a medially directed tab that articulates
with the orbit margin.
The palpebral (Figs 1, 3–5, 8, 12, Ppb) articulates
with the prefrontal and lacrimal by means of a
depressed facet. The more robust proximal portion is
expanded and closely moulded to the surfaces of the
bones to which it was (ligamentously?) attached and
a medially directed tab wraps around the orbital
margin and, doubtless, helped to stabilize the articulation of this bone. The remainder of the palpebral is
a curved, tapering rod that extends obliquely (chordlike) across the upper portion of the orbit; its curvature follows the expected contour of the eyeball itself.
The rod-like portion ends in a bluntly rounded point
close to (but not touching) the posterior edge of the
orbit. The edge of the palpebral that lies adjacent to
the rugose margin of the orbit (formed by the prefrontal, frontal, and postorbital) is also rugose; it is
probable that the slot between the palpebral and the
osseus orbital margin (Fig. 12) was spanned by a
sheet of connective tissue (Maidment & Porro, 2010).
The postorbital (Figs 1, 3–13, Po) is triradiate and
forms a major part of the boundaries of both temporal
fenestrae and the orbit. The postorbital is most
massive where it forms the posterodorsal portion of
the orbit margin. Medially, the main body of the
postorbital meets the frontal plate at a deep, undulating, butt-jointed suture (Fig. 12). Posteriorly, a substantial bar of bone extends horizontally toward, and
overlaps in a substantial scarf-joint, a shorter but
similar bar projecting from the body of the squamosal,
thereby forming the intertemporal bar. The medial
edge of the intertemporal bar bears a distinct ridge
203
marking the lateral margin of the supratemporal
fenestra (stf); the lateral edge of this bar is also
marked by a prominent ledge (Figs 1, 4, 5, por) that
extends in a shallow arch from the posterolateral
edge of the paroccipital wing along the squamosal
(sqr) toward the orbital margin before sweeping ventrally along the posterior edge of the orbital cavity
and eventually fading out as it approaches the
postorbital-jugal suture (Fig. 8). This distinctive
structure forms the anterior margin of a shallow
recess that frames the upper and anterior portions of
the infratemporal fenestra. The ventral portion of the
postorbital is robust and curves anteriorly toward its
distal end as it forms an extensive scarf-jointed
suture with the ascending process of the jugal. On its
medial surface the postorbital also bears a ridge that
extends dorsally to meet the underside of the skull
roof, adjacent to a concave pocket or facet that marks
the contact between the ventral postorbital-frontal
surfaces and the condyloid distal end of the anterolateral process of the laterosphenoid.
The squamosal (Figs 1, 3–15, Sq) is a complex bone
locked in position in the dorsolateral corner of the
skull, connecting the skull table and associated braincase to the temporal framework, facial skeleton, and
palate. As described above, it contacts the posterolateral edge of the parietal at the back of the supratemporal fenestra, and also sends a robust bar anteriorly
to contact the postorbital; the squamosal is most
robust at the junction between these two processes,
where it forms an anteroventrally directed cuff and
smooth socket for the head of the quadrate and contacts the closely associated dorsal extension of the
quadratojugal. The cuff-like, anterodorsally directed
process that projects from this socket caps and wraps
itself around the anterodorsal portion of the quadrate
and merges (posterodorsally) with the main body of
the squamosal, which itself is smooth and inset
beneath a rim (sqr). The medial portion of the squamosal (adjacent to the cotylus) is sutured against the
oblique lateral wall of the paroccipital process
(Figs 10, 15, sqs) and wraps over the dorsal margin of
this latter process forming the upper quadrant of the
occiput in posterior aspect (Fig. 14) and may just
make contact with the lateral corner to the diminutive post-temporal ‘fenestra’ (Figs 14, 15, pof). The
postcotylus portion of the squamosal appears to wrap
itself around the posterior portion of the quadrate
head and forms a short, oblique flange that lies
against the paroccipital wing, but does not extend
ventrally to support the quadrate shaft at a level
lower than the quadrate cotylus (Figs 8, 9).
The maxilla (Figs 1–6, 8–13, Mx) is triangular in
outline and comprises a robust ventral portion that
accommodates the dentition and, above this, medial
and lateral laminae as well as a medially offset ante-
204
D. B. NORMAN ET AL.
rior process. The tooth-bearing portion has room for
12 alveoli in the holotype (SAM-PK-K337) and 11 in
the referred specimen (SAM-PK-K1332); a lower
count is present in the juvenile SAM-PK-K10487
(Butler et al., 2008a). The maxilla forms a thickened
border to the anterodorsal and ventral portions of the
external antorbital fenestra and the posterior half of
the diastema; it contacts the premaxilla anteriorly
and dorsally, the lacrimal and jugal posterodorsally,
the vomer and palatine medially (as reconstructed
– Figs 10, 11, 13), and the ectopterygoid posteriorly.
The anterior maxillary process contributes to the
posterior half of the diastema and wedges inside the
medial posterior process of the premaxilla (shown
schematically in Figs 10, 11, 13). The lateral surface
of the anterior process of the maxilla is fractured in
SAM-PK-K10487 (Butler et al., 2008a) revealing a
cavity, which has been interpreted as possibly pneumatic (Fig. 28, ?pn). The anterior processes of the
maxillae meet in the midline (mxs) and probably
receive the rostral tips of the fused vomers (as seen
in the basal ornithopod Hypsilophodon, see Galton,
1974 – as illustrated schematically in Figs 10, 11).
The main tooth-bearing portion of the maxilla flares
laterally as a prominent horizontal shelf: the maxillary ridge (mxr, or ‘supralveolar lamina’ of Witmer,
1997), beneath which the teeth are inset medially
creating the dorsal half of the ‘cheek’ (buccal emargination). In occlusal view the tooth row is slightly
bowed (convex laterally). An anterior extension of the
maxillary ridge (mxr – Figs 4, 5) frames the anterior
part of the antorbital fenestra as it curves dorsally
and then sharply posterodorsally, lateral to the
anteromedial maxillary process; this extension contacts the posterolateral premaxillary process and
then tapers to contact the anterior tip of the lacrimal.
Posteriorly, the maxillary ridge continues on to the
jugal. A thin medial lamina of the maxilla forms a
substantial part of the internal wall of the antorbital
fossa (aof). The antorbital fossa is unusual when
compared to ornithischians more generally in that it
develops a narrow posterior extension into the body of
the jugal and associated jugal boss (jb). The inner
wall of the antorbital fossa is penetrated by two
clearly visible fenestrae: an anterior maxillary fenestra (Fig. 6B, aaf), an elliptical opening that occupies
an anterodorsal location; and a slightly larger posterior antorbital fenestra located along the obliquely
orientated suture between lacrimal and maxilla
(Fig. 6B, paf). A smaller additional opening (identified
in CT scans of heterodontosaur skulls by L. B. P.)
located in the anteroventral corner of the antorbital
fossa is comparable in location to the promaxillary
fenestra of theropods and birds (Witmer, 1997;
Tykoski & Rowe, 2004); if genuinely homologous, this
latter opening may support the report of pneumati-
zation in the maxilla of Heterodontosaurus (Butler
et al., 2008a). The medial surface of the maxilla
expands dorsal to the alveolar margin and forms a
robust medial maxillary shelf, visible in ventral view
of the palate in SAM-PK-K1332; the medial maxillary
shelves do not contact each other medially to form a
full secondary palate.
The lacrimal (Figs 1, 4, 5, 8, 10, 11, La) is superficially strap-like but strongly deflected at the junction
between its horizontal (anterior) portion and the
portion contributing to the orbital margin. However,
its shape is more complex than a simple strap because
it comprises both lateral and medial parts: it forms
the dorsal and posterior margins of the external
antorbital fenestra as well as the posterior portion of
the wall of the antorbital fossa, and the anteroventral
margin of the orbit. The anterior portion tapers to a
point and contacts the premaxilla and prefrontal; it is
overlain laterally by the proximal portion of the
palpebral. The external and orbital surfaces of the
lacrimal both bear trough-like indentations (Figs 4, 8)
and the foramen for the lacrimal canal is located in
the orbital indentation. The medial edge of the orbital
exposure of the lacrimal is sharp and probably supported connective tissues that separated the orbital
and nasal cavities. A thin lamina of the lacrimal
appears to be sutured to the medial lamina of the
maxilla and forms the posterodorsal half of the posterior antorbital fenestra (paf).
The jugal (Figs 1, 4–6, 8–13, J) is complex, comprising discrete anterior, dorsal, posterior, and ventral
processes, as well as an unusual jugal ‘boss’ (jb). The
jugal forms the ventral margin of the orbit, and
accommodates a channel for the posterior extension of
the antorbital fenestra; it also contributes to the
postorbital bar and lower temporal bar as well as
forming a brace-like structure (the jugal process, jp)
lateral to the lower jaw. The jugal is sutured to the
lacrimal anteriorly and there appears to be an extensive anteromedial suture against the maxilla; the
extensive scarf suture to the postorbital dorsally is
clear, as is that for the quadratojugal posteriorly; the
presumed suture with the ectopterygoid has been
partially restored based on SAM-PK-K1332.
The external surface of the anterior process is excavated by a channel-like extension of the antorbital
fossa and this (gradually subsiding) channel divides
robust dorsal and ventral processes that contact the
lacrimal and maxilla, respectively. The portion of the
jugal that abuts the maxilla forms a continuation of
the everted maxillary ridge (mxr). The dorsal process
swings upward as it forms the orbital margin to
contact the postorbital at an extensive scarf suture.
The posterior process is transversely thin and dorsoventrally expanded, and contacts the quadratojugal
in a comparatively narrow, scarf suture, forming a
relatively narrow lower temporal bar; however, the
contact between these bones is complex and strong,
combining extensive overlap with fine interdigitations. The ventral process (jp) of the jugal is posteroventrally directed and is thickened along its anterior
edge and transversely flattened at its tip, which is
bluntly rounded. This process is particularly striking
because it descends to a level slightly above that of
the, equally prominent and ventrally deflected, pterygoid flange; these two bones thus form the equivalent
of a narrow and deep slot or ‘guide’ that evidently
constrained the motion of the lower jaw.
A stout ‘boss’ (jb) projects from the body of the jugal.
CT scans (L. B. P.) reveal a small sinus at its centre
and the punctate appearance of the boss suggests
external openings associated with this cavity. A small
ventrolateral opening has been observed on both
jugals of SAM-PK-K1332, which may represent the
openings to the cavity of the jugal boss. It should also
be noted that the osteological extension of the antorbital fossa fades out in close proximity to the base of
the jugal ‘boss’ and there may have been soft-tissue
links between these structures, especially if the antorbital fossa is a manifestation of cranial pneumatism
(Witmer, 1997).
The quadratojugal (Figs 1, 4–9, 13, 14, Qj) is exceptionally extensive when compared to most known
ornithischian dinosaurs (and in some respects more
closely resembles that of some saurischian dinosaurs).
It forms a thin sheet of bone that was presumably
tightly (ligamentously) applied to much of the lateral
surface of the quadrate. In outline the quadratojugal
forms the posteroventral margin of the infratemporal
fenestra (itf), contacting the jugal anteriorly and the
squamosal posterodorsally, thereby excluding the
quadrate from participation in the border of the itf;
its anterior process is narrow and transversely compressed, yet forms a tight suture with the jugal. The
dorsal process forms a sheet that overlaps the main
shaft of the quadrate and contacts the prequadratic
‘cuff’ formed by the squamosal (Fig. 8) at an oblique
suture. The posterior edge of the quadratojugal lies
against the anterior margin of the dorsal half of the
quadrate shaft (so that the latter is visible in lateral
view); however, ventrally the quadratojugal overlaps
the lateral surface of the quadrate adjacent to the
(paraquadratic) foramen (qf) and adjacent vertical
recesses (as a consequence the quadrate foramen is
almost totally obscured in lateral view – see Fig. 8).
The contact between the quadratojugal and quadrate
terminates just above the edge of the quadrate articular condyle; from this point the quadratojugal curves
dorsally and anteriorly to contact the jugal, creating a
sizeable embayment between the quadrate articular
condyle and the ventral process (jp) of the jugal
(Fig. 9).
205
The quadrate (Figs 1–11, 13, 14, Q) comprises a
tall, bent and axially twisted, hemicylindrical pillar,
the anterior surface of which forms an open vertical
trough. The dorsal portion of the quadrate shaft
bends posterodorsally and terminates in a rounded
articular head, which forms a ball-and-socket joint
with the smooth cotylus on the ventral surface of the
squamosal (visible in the disarticulated, partially prepared skull SAM-PK-K1332 – photographic archive).
Ventral to the quadrate head a thin sheet of bone
projects from the anteromedial edge of the shaft
forming a pterygoid wing of the quadrate (Figs 5, 8,
13, ptq) that has an extensive overlapping suture
with the quadrate wing of the pterygoid (qpt);
together these form a deep and transversely thin
sheet that lines the medial wall of the adductor
chamber (physically separating the adductor musculature from the cranial nerves and vascular supply
associated with the lateral wall of the braincase). The
anterolateral edge of the quadrate shaft is bevelled
for attachment of the quadratojugal, and dorsally
there is a short, sinuous, flange-like wing that
projects anterolaterally for a short distance to meet
the upper portion of the quadratojugal.
The exceptional height of the quadrate accounts for
the marked ventral offset of the jaw articulation. At
the jaw articulation the shaft is transversely
expanded, producing a larger and more robust medial
articular condyle and a ventrally offset, lateral
condyle; the articular surfaces of these condyles when
viewed ventrally appear to be twisted anteromedially
(Figs 6, 13). Above the condylar region the shaft of the
quadrate contracts transversely into a curved and
tapering shaft. Laterally the quadrate shaft presents
a large, essentially rectangular, surface for attachment of the quadratojugal. At approximately midheight the posterolateral margin of the quadrate
shaft is notched creating a laterally compressed horizontal channel that is fully enclosed laterally by the
quadratojugal; this represents the paraquadratic
foramen (Figs 1, 5, 7, 8, 14, qf), which opens into the
adductor muscle chamber anteriorly. Dorsal and
medial to the paraquadratic foramen, in a shallow
vertical groove on the posterior edge of the quadrate
shaft, there is a further small depression (Figs 7, 14,
?pn); the function of this is unclear, but it might be
again a remnant of cranial pneumatism – as suggested in the maxilla and jugal boss. Above these
features the lateral edge of the quadrate develops a
sinuous flange that follows the suture to the posterior
margin of the quadratojugal; this strongly resembles
the condition in marginocephalians (Figs 1, 4, 5, 8).
OSTEOLOGY:
NEUROCRANIUM, OCCIPUT, AND PALATE
The laterosphenoid (Figs 2, 4, 10, 11, 15, Ls) is a
curved plate that, with its counterpart, form the
206
D. B. NORMAN ET AL.
lateral walls of the cavity occupied in life by the
cerebral lobes; it simultaneously lines the anteromedial portion of the adductor chamber and appears to
be butt-jointed to the parietal (Pa) dorsally. The anterior end tapers and curves laterally to form a rounded
boss that fits into a recess on the ventral surface of
the frontal/postorbital suture (Fig. 10 – shaded). A
short spur, directed posteroventrally from the body of
the laterosphenoid contacts the basisphenoid (Bs);
above this spur the laterosphenoid is emarginated
and appears to contact the proötic portion of the
lateral braincase wall along an oblique fracture in
SAM-PK-K1332 that meets the parietal dorsally
(Figs 4, 5 – this position is consistent with the
approximate line of suture between these bones in
most ornithischians). In the holotype (SAM-PK-K337)
the internal wall of the adductor chamber is smoothly
concave in this area with no clear indication of a
suture. The proötic (Pro) lies immediately behind the
laterosphenoid and forms much of the visible remainder of the inner wall of the adductor chamber and
(typically in diapsids) overlaps and is closely (in this
case invisibly – see ‘Pro/Op’ Figs 2, 4) sutured to the
opisthotic. In SAM-PK-K1332 the proötic and
basisphenoid surround an elongate fenestra (Fig. 15,
B, V); this area (probably exaggerated by fracturing)
represents the trigeminal fossa (associated with the
principal ganglion of cranial nerve V and its associated nerve tracts V1–3). Further posteriorly, an oblique
buttress separates the trigeminal fossa from a recess
that probably represents the combined exits for
branches of the facial nerve (facialis – cranial nerve
VII) followed by an auditory recess – all of which
would have been surrounded by the components of
the lateral wall of the braincase that form the otic
capsule. The opisthotic (Op) typically, encloses the
majority of the otic capsule (although this latter
structure can invade adjacent bones, i.e. the proötic
and supraoccipital); it forms the bulk of the posterolateral braincase wall and is firmly sutured to the
underlying basisphenoid. There is no trace of a suture
on the posterior sidewall of the braincase that might
betray the extent of the exoccipital, although the
suture would be expected to lie in the pillar-like
region separating the foramina tentatively identified
as associated with the passage of the vagus (cranial
nerve X) and the jugular vein (jug) and the more
posteriorly located foramina for cranial nerves XI and
XII. As noted above, the lower border of the opisthotic
is penetrated by a number of foramina/fenestrae associated with cranial nerves (IX–X); the vascular/
lymphatic drainage of the occipitonuchal region, the
adductor chamber and endocranial cavity; and the
recess associated with the middle ear cavity.
Posterior to the proposed location of the auditory
recess (Fig. 15B, fo) additional foramina exit the
braincase; these undoubtedly represent a combination
of posterior cranial nerve passages as well as those
associated with venous drainage of the endocranium
and the lymphatic system (Romer, 1956; Galton,
1974; Norman, 1980, 1986), but their identity cannot
be established with certainty. Tentative identifications of the apertures are suggested (Fig. 15B) based
on the general configuration seen in several other
ornithischian dinosaur crania (Brown & Schlaikjer,
1940; Ostrom, 1961; Galton, 1974; Evans, Ridgely &
Witmer, 2009). The exoccipital (Ex) normally occupies
the posterior portion of the lateral wall of the braincase and is sutured to the basioccipital ventrally, but
in this instance its sutural relationships are lost
through fusion. The supraoccipital is located dorsal to
the exoccipital pillars on either side of the foramen
magnum (see Figs 7, 14) and shows a pronounced
dorsal lip (even though this general area is transversely deformed in both the holotype and referred
skulls – Fig. 7) that demarcates the supraoccipital
relative to the exoccipitals.
The
proötic-‘otoccipital’
(opisthotic-exoccipital)
complex anchors massive paroccipital wings that are
formed principally by the opisthotics (Figs 4, 5, 10, 15,
pocc), which project posterolaterally from the area
occupied by the otic capsule; they are distinctive in
occipital aspect (Fig. 14) in that each wing is pierced
by a centrally located foramen (Figs 2, 7, 14, ovc); this
is here interpreted as a remnant of the cranioquadrate passage carrying the main trunk of the facial
nerve, orbitotemporal artery, and lateral head vein –
cf. Romer, 1956: 142) rather than representing a
remnant of the post-temporal fenestra as more commonly supposed (Weishampel & Witmer, 1990); there
is also evidence of a small dimple-like depression
(?pn) near the ventrolateral corner of the paroccipital
wing on both sides in SAM-PK-1332. In position and
form this feature is most reminiscent of the small
depressions located on the posterior face of the quadrate, mediodorsal to the quadrate foramen (described
above and similarly labelled:?pn). Whether this structure represents a remnant trace of cranial pneumatism is uncertain (there is no evidence for an
associated sinus); however, the possibility is intriguing when linked to similar structural features
observed on the quadrate shaft, jugal boss (referred to
above), and in the maxilla (SAM-PK-K10487 –
Figs 28, 29 – Butler et al., 2008a).
The supraoccipital (Figs 7, 11, 14,S) is a tall, broad
and principally triangular element with a sharp,
median, nuchal crest that rises steadily from its
ventral border (Fig. 14). Excavations on either side of
the nuchal crest provide attachment sites for neck
musculature (as does the entire surface of the occipital plate). An anterodorsally inclined, tapering,
ascending process of the supraoccipital projects into
the posteroventral surface of the parietal on the
occipital plate (Fig. 11 – schematic sagittal section),
with the posterior edge of the lateral wings of the
posterior margin of the parietal folded down on either
side of the supraoccipital. The ventral margin of the
supraoccipital develops a slight lip and is arched to
form the dorsal edge of the foramen magnum and, on
either side of this foramen, elongate pedicels of the
exoccipitals (Ex) form its lateral walls; these latter
contact the basioccipital ventrally. The lateral
margins of the supraoccipital, ventral to the parietal
contact seem to form the margins of a slot-like
opening (pof), which is interpreted as representing
the diapsid post-temporal fenestra that runs forward
along the parietal/opisthotic/supraoccipital suture
toward a cleft on the upper part of the sidewall of the
braincase (Fig. 15, vpar). The lower lateral part of the
supraoccipital is locked against the medial sides of
the oblique, paroccipital wings.
The basioccipital (Bo) is a, comparatively, short
component at the rear of the floor of the braincase. In
occipital view the basioccipital is dominated by the
U-shaped outline of the occipital condyle; the latter is
transversely flattened (Fig. 13, bc), but dorsoventrally
strongly convex and thus rather roller-like and the
articular surface extends anteriorly in the ventral
midline toward the basisphenoid. Dorsally and laterally the basioccipital contacts the base of the exoccipital pillars and there is a small, posteriorly
projecting lip that marks the dorsolateral edge of the
condylar surface (and might be interpretable as foot
formed by the exoccipital in the absence of sutural
evidence). The anterior part of the basioccipital contracts and is locked into the rear of the basisphenoid
between the prominent basal tubera as suggested in
Figure 11.
The basisphenoid (Bs) is by far the most extensive
bone in the floor of the braincase. In addition to the
contacts already described, it is fused to the parasphenoid (Psp) and ?presphenoid (Ps; Fig. 11), and
articulates with the pterygoids anteroventrally
(Figs 10, 11, 13, 14). In ventral view (Fig. 14) the
basisphenoid presents a flared posterior margin
formed by the rugose, lip-shaped, posteriorly facing,
basal tubera (Figs 10, 11, 13, 14, bst); the rims of
these tubera are thickened and contract anteriorly
into the subcylindrical body of the basisphenoid,
which has a shallow ventral sulcus; the anterior
margin of this sulcus forms a raised rim beyond
which there is a narrow, sagittally grooved median
extension that projects anteriorly into the base of the
(parasphenoidal) cultriform process and divides anteriorly into pedicels from which develop the basipterygoid processes (Fig. 13, bpt). Laterodorsal to the
basipterygoid pedicels, their lateral edges (in the area
directly beneath the trigeminal fossa) are expanded to
207
form thin, deep flanges (bsf) with slightly rugose,
curved posterodorsal margins (see Fig. 10); each
flange laterally overlies the diagonal bulge on the
sidewall of the braincase that separates the trigeminal fossa from the facial nerve/auditory recess
complex. The basisphenoid flange (bsf) curves medially beneath the trigeminal area on the side wall of
the braincase and merges anteriorly with the anterior
‘cultriform’ process of the parasphenoid (forming deep
basipterygoid recesses reminiscent of those seen in
theropods such as Coelophysis. The posterior part of
the ventral edge of the cultriform process divides
along the midline and each side merges with the
anteromedial margin of each basipterygoid process.
The basipterygoid processes are fairly stout projections that end in anteroventrally and slightly divergent articular pads, which contact separate articular
facets that form as discrete raised platforms above
the medial pterygoid shelves (Figs 13, 14). The anterior margins of each basipterygoid process curve posteriorly and medially from their articular expansions,
meet in the midline, and swing abruptly anteriorly to
form a deep, tapering blade (cultriform process) –
presumably V-shaped in cross-section – representing
the basisphenoid and its enveloping parasphenoid;
there is no hint of a suture between these two bones
and the cultriform process projects horizontally
between the orbital cavities almost to contact the
palatines (seen through the orbital cavity); this structure undoubtedly supported connective tissue sheets
dorsally, associated with the interorbital septum. At
the posterior dorsal end of the parasphenoid there is
a median plate of bone that probably represents the
presphenoid/cultriform process (Fig. 11, Ps).
The palate (Figs 2, 10, 11, 13) is deeply vaulted and
relatively narrow (even after compensating for considerable post-mortem transverse compression in
both of the principal skulls) – whether the pterygoids
were sutured together medially along their entire
length (anterior to the basal articulation) remains
uncertain. The extreme anterior end of the palate
comprises a solid and arched roof formed by the fused
premaxillae that are in turn firmly sutured to (and
ensheath) the anterior maxillary processes. The maxillary portion of the palate is partly enclosed by the
medial maxillary shelves, which curve medially to
form a partial dorsolateral roof to the oral palate,
while the midline vomers divide the left and right
nasal passages. The palatines are reconstructed as
curved plates that separate the orbital cavities from
the nasal passages at the posterior end of the maxillary palate. The pharyngeal palate is roofed by the
pterygoids, which form inclined plates bounded laterally by the pterygoid flanges; the central plate of each
pterygoid also supports the raised articular facets for
the basal articulation (Fig. 10, rba).
208
D. B. NORMAN ET AL.
The vomers (V) are reconstructed (Figs 10, 13) as
elongate, transversely compressed, sheet-like bones
fused to form a keeled structure in the midline; they
are shown anchored between the anteromedial maxillary processes anteriorly (based on the configuration
described in Hypsilophodon – Galton, 1974) and
extend posteriorly as far as the palatines (whether
they make contact with the latter is unclear and
contact with the anterior processes of the pterygoids
is uncertain). The vomers form thin septae that separate the nasal passages.
Despite being obscured by matrix and crushing, the
palatines (Figs 10, 11, 13, Pal) have been visualized
from CT scans; they form thin, arched plates that
seem to meet in the midline thus forming a vaulted
roof to the posterior maxillary palate. A thin posterior
process lies medial to, but seems not to contact the
anterior pterygoid ramus (this region is however
broken and the contact has probably been lost). Ventrolateral processes are firmly sutured to the medial
surface of the medial maxillary shelf. The palatines
expand laterally but taper anteriorly to lie adjacent to
the posterior tips of the vomers. CT scans suggest
that the dorsomedial edges of the palatines were
sutured together; certainly their juxtaposition in a
laterally compressed and obliquely sheared skull
(SAM-PK-K1332) is highly suggestive of a degree of
fusion. There are no distinct fossae on the dorsal
surfaces of the palatines as in Lesothosaurus (Witmer,
1995).
The pterygoids (Figs 2, 6, 9, 11, 13, Pt) have a
complex structure. Each comprises a tapering anterior process, a dorsoventrally deep and transversely
compressed quadrate ramus (qpt), and a robust pterygoid flange (ptf); all these processes are anchored to a
central plate that is twisted but also provides an
oblique platform that supports the articular facets for
the basal articulation. The anterior process is poorly
preserved but is sutured laterally with the ectopterygoid but contact with the maxilla is uncertain. It is
unfortunate that the extent of its anterior contacts
with the palatine and vomer are uncertain. The quadrate ramus is a very thin, sheet-like, wing of bone
that is closely pressed against the pterygoid wing of
the quadrate (ptq). The proximal portion of the quadrate ramus expands medially (ptmr) to form thickened shelves that face posterodorsally and support
raised articular facets (Figs 10, 14, rba) for the basal
articulation that are lodged against the medial
surface of the quadrate wing; these latter facets are,
as preserved, quite close to the midline. The pterygoid
flange (ptf) is unusually prominent (and, particularly
in lateral aspect, appears reminiscent of that seen in
crocodilians); it is well exposed in the holotype (SAMPK-K337 – Fig. 2, ptf) but in the referred skull these
were broken off and lay adjacent to the main body of
each pterygoid. The flange is finger-like, has a triangular cross-section, and is drawn out into a posteroventrally directed process whose tip is at the level of
the jaw articulation (Fig. 11). A pronounced recess
extends up the posterior surface of the pterygoid
flange on to the dorsal surface of the body of the
pterygoid. The lateral surface of the pterygoid flange
lies in close proximity to the medial surface of the
lower jaw and is both thickened and appears to be
smooth, suggesting that it may have served to guide
the movement of the lower jaw. The ectopterygoid was
sutured to the lateral surface of the base of the
pterygoid flange.
The holotype (Fig. 2) has only the right pterygoid in
place, the left side of the skull having been sheared
away. The medial edge of the central pterygoid plate
has the potential to be bound, at least by connective
tissue, to its neighbour. The referred skull (Fig. 6)
shows both left and right pterygoids in closely
appressed position, albeit slightly displaced, in a skull
that has clearly been subject to both lateral compression and some dorsoventral shearing. It seems probable that the pterygoids were quite closely associated
in the articulated and undistorted skull – this would
have been necessary simply because the basipterygoid
processes are relatively close together in the midline.
The medial edges of the pterygoids may have been
bound fairly substantially by connective tissue, rather
than having been fused. This general configuration
reinforces the idea (supported by the apparent
midline connection between the palatines) that the
pterygoids provided a brace between the posterior
ends of the upper jaw that tended to counteract
lateral displacement of the upper jaw during occlusion with the lower jaw.
Nota bene. The ventral reconstruction of the skull
(Fig. 13) is conjectural and partly idealized, in that it
shows the pterygoids separated by a narrow interpterygoid vacuity; this is regarded as a conservative
reconstruction that allows view of the immediately
overlying parts the cranium. The pterygoids (anterior
to the basal articulation) were probably sutured in
the midline.
The ectopterygoid (Figs 10, 11, 13, Ec) is reconstructed (partly speculatively – Fig. 6B) as a relatively short, strap-like bone that spans the adjacent
surfaces of the jugal and maxilla before curving medially and ventrally to form a solid contact with the
pterygoid close to the base of the pterygoid flange.
This bone ties together firmly the bones at the rear of
the palate against the maxillary unit.
OSTEOLOGY:
LOWER JAW
Although the lower jaw of the holotype (SAM-PKK337 – Fig. 1) is incomplete, and partly obscured by
irremovable matrix, the referred specimen (SAM-PK1332 – Figs 4, 16–18) has both lower jaws fully prepared out and in a fair state of repair, the left lower
jaw being more complete than the right.
The lower jaw rami are tied together via a triangular, block-like toothless predentary. The toothbearing (dentary) part of each jaw ramus is
particularly robust, with the dentition inset medially
and followed by an elevated coronoid eminence;
behind this region the postdentary bones support a
ventrally offset jaw-joint but, apart from the area
surrounding the jaw articulation, the intervening
postdentary bones are comparatively sheath-like and
thin-walled (Fig. 19).
The predentary (Figs 1, 3B, 4, 18, 19, Pd) is triangular in dorsal view and rather similar in lateral
aspect. The occlusal margin is concave medially and
raised relative to the midline and when viewed from
above the predentary is narrow and scoop-shaped.
The external surface of the predentary is slightly
rugose and pitted by small foramina and, no doubt,
supported a keratinous beak (rhamphotheca) in life
that occluded against a similar (although probably
less extensive) beak on the anterior part of the premaxilla as well as with the medial surfaces of the
premaxillary teeth. The posterior sutural border is
slightly sinuous and fits snugly against the blunt,
similarly sinuous and inflated anterior ends of each
dentary: there is a short dorsolateral flange and
median ventral flange that helped to stabilize the
suture [the predentary-dentary suture, which is
exposed in SAM-PK-K1332 and SAM-PK-K10487:
Figs 28, 29, cannot be described as ‘spheroidal’ –
Weishampel, 1984, as previously noted by Barrett
(1998) and Porro (2009)]. Although the flanges
referred to above might represent precursors (of the
more complex predentary-dentary sutural structures
seen in more derived ornithischians), the predentary
exhibits neither pronounced dorsolateral processes
that run obliquely along the anterodorsal edge of
each dentary, nor the flap-like posteriorly directed
medioventral process that underlies the dentary symphysis (cf. Galton, 1974; Norman, 1980; Sereno,
1991a).
The dentary is a robust ellipsoid bar that deepens
posteriorly; it bears a longitudinal medial groove
(Meckel’s groove – mg) and is partially sheathed by
the postdentary bones. There are 11 alveoli (including that of the enlarged caniniform) in SAM-PKK1332. At its anterior end each dentary contacts the
predentary anterolaterally and the opposite dentary
medially. As described above, the anterior tip is
inflated and blunt (this is principally a consequence
of the requirement for accommodation space for the
root of the large dentary caniniform). The sutural
surface bears a deep, medial projection (Fig. 28A,
209
pds) that wedges into the posterior surface of the
predentary and also forms the dorsal part of the
dentary symphysis medially (Figs 17–19, sy); the
lateral surface of this projection bears a prominent foramen that is visible on SAM-PK-K10487
(Fig. 28A, sf) and also apparent in CT scans of SAMPK-K1332 – Porro, 2009). Immediately posterior to
the predentary-dentary suture, below the level of the
foramen just described, another prominent foramen
exits on to the surface of the dentary (Figs 18, 28,
29, adf). The Meckelian canal (mg), which is wide
and opens into the adductor fossa posteriorly, contracts markedly below the position of the eighth
dentary tooth.
The dentary tooth row is strongly inset along the
dorsal border, creating a pronounced ‘cheek’ recess,
which is accentuated by a ridge that follows a curved
course along the lateral surface of the dentary and
sweeps up toward the coronoid eminence (Figs 18A,
19A). The alveoli (and dentition) follow the dorsal
edge of the dentary toward the base of the coronoid
eminence and describe, as they do so, a shallow longitudinal curve that is concave laterally (this is
opposed by a maxillary dentition that is bowed in the
opposite sense: being straight to slightly convex laterally). The posterior margins of the dentary flare
outward to envelope and articulate with the postdentary bones (the suture being indented by the presence
of the external mandibular fenestra – Figs 1, 4, 18,
19, emf). The medial side of the dentary is more
strongly notched; however, this is masked by the
prearticular, coronoid, and the sheet-like splenial,
which extends far forward and may have approached
the dentary symphysis (Figs 18B, 19B). The dentarydentary symphysis (Figs 17–19, sy) is represented by
a deep, flattened, and rugose buttress-like structure
that is positioned just anterior to the thickening of
the anterior dentary ramus associated with the alveolus for the large caniniform; this large sutural surface
extends from the ventral margin of the ramus nearly
to the upper margin and clearly formed a very strong
suture that not only stabilized the dentary-dentary
contact, but formed a firm point for anchorage of the
predentary.
The splenial (Figs 17–19, Sp) is a thin plate
applied to the medial surface of the lower jaw. It
contacts the medial wall of the dentary extensively
and extends posteriorly to contact the coronoid and
prearticular; it overlies the posterior portion of the
Meckelian groove (mg); the anterior limit of the
splenial and the point at which Meckel’s canal
opens on to the medial surface of the dentary is
below the fifth dentary tooth (L. B. P. – CT scans of
SAM-KP-K1332). The posterior edge of the splenial
bifurcates to contact the coronoid and prearticular,
and forms the anterior margin of an internal man-
210
D. B. NORMAN ET AL.
dibular fenestra (Figs 17–19, imf). A ventral sleeve
formed by the splenial encloses the prearticular and
angular posteriorly.
The coronoid (Figs 17–19, Co) is a strap of bone
applied to the medial surface of the dentary and the
dorsal edge of the splenial; it forms the summit of the
coronoid eminence. Its dorsal border is rugose, indicating tendinous attachment for some of the principal
jaw adductor muscles. The coronoid forms a parapet
along the medial side of the dentary adjacent to the
alveoli and extends ventrally to contact the splenial
posteriorly but separates from the latter further anteriorly and appears to form a splint that tapers to a
blunt tip – its anterior limit is uncertain (see Figs 17,
18B) and has been reconstructed in this instance
lying medial to the noncaniniform dentition [by reference to the approximately contemporary ornithischians Lesothosaurus (Sereno, 1991a) and Scelidosaurus – (D. B. Norman, unpubl. data)].
The surangular (Figs 16, 17, 19, Sa) has an unusually complex and unique structure. It is a sinuous,
strap-like element capping the other postdentary
bones and forming the upper margin of the posterior
lower jaw; it connects the coronoid eminence to the
articular glenoid and is sutured to the dentary anteriorly, the angular ventrally, and the prearticular
and articular posteriorly. The upper margin of the
surangular is formed by a finger-like, cylindrical
ramus that arises from the transversely thickened
body of the surangular immediately in front of the
articular glenoid. The dorsal surangular ramus
arches anteriorly before contacting the dentary just
posterior to the highest part of the coronoid eminence (unfortunately this area is damaged, perhaps
reflecting a generally weak sutural connection, in all
known specimens). The dorsal ramus is separated by
a long, narrow cleft from a similar but slightly
longer, narrower, and ventrolaterally positioned
ventral surangular ramus; at the posterior end of
this cleft is a discrete elliptical surangular foramen
(Fig. 19, sfor) that opens directly into the adductor
fossa medially (the foramen is obscured in lateral
aspect by a shoulder of the ventral surangular
ramus). The ventral surangular ramus contacts the
dentary at an irregular, butt-jointed suture between
the coronoid eminence and the external mandibular
fenestra. The ventral ramus lies along the dorsal
edge of the angular; however, the mid-section of this
contact, immediately above the external mandibular
depression (Fig. 16A) shows no obvious sutural line
(even allowing for the consolidant) and was presumably fused; posteriorly a clear crease marks a more
definite contact. In medial view the external wall of
the adductor fossa is composed of surangular dorsally and angular ventrally, but it would appear that
the angular overlaps the surangular mediodorsally
and a medial splint of the dentary appears to lie
medial to the angular and lower ramus of the surangular. Posteriorly, the surangular forms the ventrolateral wall of the glenoid (to which it contributes
a lipped edge the anterior margin of which can be
seen in lateral profile). The small foramen (sfor) lies
a short distance anterior to the anterior rim of the
glenoid. Beyond the articular glenoid the body of the
surangular continues posteriorly as a tapering
lateral wall of the retroarticular process, which is
deeply excavated dorsally for the attachment of m.
depressor mandibulae (MDM) and laterally for the
insertion of m. pterygoideus posterior (MPTP;
Figs 34–36). The posterolateral wall of the surangular thickens transversely and curves medially to
contact the prearticular immediately in front of the
articular at the back of the adductor fossa.
The angular (Figs 16, 17, 19, An) forms the majority
of the external mandibular depression and forms all
but the anterior margin of the external mandibular
fenestra (emf). Its dorsal portion is a laterally concave
sheet of bone that contacts the ventral surangular
ramus, forms the lateral wall of the adductor fossa
and a curved ridge, which defines the lower margin of
the depression. Ventrally, the angular is overlapped
by a posterior extension of the dentary and forms a
slightly dorsally bowed and thickened continuation of
the ventral margin of the lower jaw (probably
marking a pulley-like channel for part of the pterygoideus musculature). Ventrally the angular forms a
trough (flooring the adductor fossa) and meets the
prearticular and wedges against the splenial medially; it also has a long suture with the surangular
where it curves beneath the region of the articular
glenoid.
The prearticular (Figs 17–19, Part) is a slender
splint of bone that runs along the posteromedial
surface of the lower jaw. It extends forward and
bifurcates as it forms the posterior margin of the
internal mandibular fenestra (imf) with the splenial
(whether it contacts the dentary medially in this
region cannot be ascertained, but might be expected).
In this region its ventral edge is overlapped by a
posterior sheet of the splenial medially and this
suture extends posteriorly leaving the thin vertical
wall of the prearticular to curve ventrolaterally to
form the floor of the adductor fossa; it curves gradually dorsally and laterally to meet the main body of
the surangular in a transverse buttress immediately
in front of the glenoid (and marking the posterior
corner of the adductor fossa). Further posteriorly the
prearticular forms a tapering sheet that lies medial to
(and encloses, along with the surangular) the articular and forms part of the medioventral wall of the
retroarticular process. In the area beneath the anterior margin of the articular glenoid, on the ventral
211
surface of the lower jaw, the prearticular forms a
sutural contact with the angular.
The articular (Figs 16, 17, 19, Ar) is irregular but
basically lozenge-like in dorsal view, held in a troughlike slot between the surangular laterally, the angular
ventrally, and prearticular medially. It forms a transverse, semi-cylindrical, glenoid fossa for articulation
against the quadrate, and its axis tilted laterally
when the jaw is held so that the caniniform is vertical; the axis of the articular hinge is also orientated
anteromedial-posterolateral rather than being strictly
transverse. A median ridge separates lateral and
medial concavities for the condyles on the distal end
of the quadrate. Both condylar facets are slightly
longer (anteroposteriorly) than the condyles that they
accommodate. Raised ridges or ‘lips’ formed by the
surangular (laterally) and articular (medially) indicate the margins of the jaw joint. As the lateral
quadrate condyle extends ventral to the central
portion of the jaw articulation, it would have resisted
torsion (rotation about the long axis of the jaw caused
by medial deflection at the dentition) of the lower jaw.
The posterior portion of the articular forms part of the
retroarticular process.
TOOTH
MORPHOLOGY
General description
Subtle heterodonty is common in living diapsids
(Edmund, 1969; Throckmorton, 1976; Kieser et al.,
1993) and has also been noted in a variety of dinosaurs: sauropodomorphs (Barrett, 2000); theropods
(Paul, 1988); and ornithischians (Thulborn, 1970b;
Maryańska & Osmólska, 1974; Norman et al., 2004a).
However, the dentition in Heterodontosaurus is
strongly heterodont (Figs 1, 4, 5, 8), to a degree that
is usually associated with synapsid mammals (Broom,
1911; Haughton, 1924); this feature is one of the
primary characteristics of this group of early ornithischians, as reflected in the choice of its generic name
(Crompton & Charig, 1962). The upper dentition is
distributed along the premaxilla and maxilla,
whereas the lower jaw has its teeth restricted to the
dentary alone.
The premaxilla has three teeth that are similar
in shape but of varying size; the first two are
comparatively small, slightly recurved, cone-like
crowns that increase in size posteriorly, whereas the
third is a greatly enlarged, slightly recurved, and
laterally compressed caniniform anchored by a deep
root that creates a distinctly buttress-like swelling on
the posterolateral surface of the premaxilla (see
Figs 1, 2, 4, 5). The full array of premaxillary teeth is
only visible in the left lateral view of the holotype
skull (Figs 2, 20 – inset). These teeth are clustered
posteriorly on the premaxilla, leaving an edentulous
1962. SAM-PK-K337 (holotype). Anterior dentition preserved on the right side of the skull (based on a photograph taken during early preparation – prior to the
complete removal of matrix from the lateral surface of the
dentary caniniform. Note, the first premaxillary caniniform is not preserved on this side of the skull; the first
premaxillary incisiform is only preserved on the left side of
the skull. Inset diagram: sketch of the profiles of the left
premaxillary dentition of SAM-PK-K337 (prepared by
A. W. C./A. J. C.). For full list of abbreviations see end of
paper.
anterior tip that was covered by a keratinous beak
(rhamphotheca) in life that may well have partially
enveloped the two incisiforms and the base of the
caniniform (judged by the rugose nature of the entire
lateral and lower margin of the premaxilla – see
Fig. 20). The premaxillary crowns and beak appear to
have occluded against the edentulous lateral margin
of the predentary (see Figs 1, 8), which would also
have been sheathed in a keratinous beak.
The maxillary dentition, which is separated from
the premaxillary dentition by a pronounced embayment or diastema (dia) that accommodated the tip of
the dentary caniniform, comprises a chisel-edged,
closely packed, ‘battery’ of 11 or 12 columnar teeth
(Figs 1, 4, 5, 8, 21, 22). The worn edge of the battery
truncates each crown obliquely, so that this surface
212
D. B. NORMAN ET AL.
1962. SAM-PK-K337 (holotype). A, the right maxillary
dentition (M.2–M.12) in lateral view (see Fig. 22 for details
of M.1–M.4); B, occlusal plan view of crowns M.8–M.11 to
demonstrate the minor imbrication between adjacent
crowns.
1962. SAM-PK-K337 (holotype). Anterior maxillary teeth
(M.1–M.4) of the right maxilla, as preserved. The presence
of M.1 is inferred from the ghost-like impression preserved
in the matrix. For full list of abbreviations see end of
paper.
faces mainly medioventrally and individual teeth,
whose crowns are thickly enamelled laterally, are
bowed gently medially along their vertical axes. Note:
the replacement crown rep ‘M.7’ revealed in the alveolar bone of SAM-PK-K1334 (Fig. 33C) indicates the
probable absence of enamel on the medial surface of
maxillary crowns. Radiography reveals seven preserved maxillary teeth in the presumed juvenile
(SAM-PK-K10487, Figs 28, 29 – Butler et al., 2008a)
and seven tooth positions are also preserved in the
incomplete maxilla (SAM-PK-K1334, Figs 30–33 –
which also provides the first unambiguous evidence of
tooth replacement in H. tucki).
Dentary teeth (Figs 18, 25, 26) include an enlarged,
gently recurved caniniform anteriorly separated by a
short diastema from a closely packed battery of
‘cheek’ crowns that oppose those borne on the maxilla.
SAM-PK-K1332 (Figs 16–19, 25, 26) has a complete
dentary tooth count of 11 (including the caniniform),
although this number is exceeded in a more recently
recognized, larger specimen (NMQR 1788 – R. J.
Butler, unpubl. data), which has a complete dentary
tooth count of 12. In the maxillary battery the worn
surface displays an enamel-dentine interface that is
flush and a worn surface that is essentially planar (or
very slightly concave posteriorly) and medioventrally
inclined; in contrast, the anterior and posterior ends
of the dentary battery exhibit similarly flush-planar
wear facets (that are steeply inclined dorsolaterally),
whereas the mid-section of the battery displays facets
that are oblique but quite clearly concave (Figs 25–
27) and appear to bear, as a consequence, a slightly
lipped or stepped lower edge (Fig. 27, lip).
As mentioned above, the upper and lower batteries
are markedly inset, creating the ‘cheek’ recess that is
typical of ornithischians (Lull & Wright, 1942; Romer,
1956; Galton, 1973a). The ‘cheek’ recess is far more
prominent than that seen in the contemporary
ornithischian Lesothosaurus (Sereno, 1991a) or
the contemporary heterodontosaurid Abrictosaurus
(NHMUK RU B54). The batteries oppose each other,
at an oblique occlusal plane which, contrary to
received wisdom (Charig & Crompton, 1974; Thulborn, 1978; Norman & Weishampel, 1985, 1991;
Crompton & Attridge, 1986; Galton, 1986;
Weishampel & Norman, 1989; Barrett, 1998; Norman
et al., 2004a), do not form completely continuous
shearing blades (as argued by Hopson, 1980).
Implantation is thecodont and although caniniform
roots taper slightly toward the crown, CT scans confirmed that cheek tooth roots are elongate, parallelsided, and penetrate deep into the bodies of the
premaxilla, maxilla, and dentary (A. W. C. – archive
photographic X-ray plates; Butler et al., 2008a – see
also SAM-KP-K1334 – Figs 30–33). Tooth roots have a
large pulp cavity initially and these become progressively occluded once the crowns are fully erupted;
thus continuous growth (linked to a high-wear
chewing regime) did not occur. Pulp cavities, which
seem to be large during initial phases of growth
of the crown-root complex (SAM-PK-K1334)
become constricted by deposition of dentine on the
anterior and posterior surfaces of the pulp cavity
walls so that the pulp cavity becomes progressively
narrower and transversely orientated once teeth
have become permanently implanted – in SAM-PK-
K10487: Butler et al. (2008a: fig. 5) the cavity
becomes completely occluded in fully-erupted and
functioning teeth.
The upper tooth rows diverge posteriorly (Fig. 13)
and, overall, this geometry is mirrored in the lower
jaw (Porro, 2009). In occlusal view the dentary dentition is concave laterally, whereas the maxillary dentition appears to be essentially straight, or slightly
convex (judged by the alignment of lateral margins of
successive alveoli in SAM-PK-K1332).
Detailed comments
The premaxillary dentition is dominated by the large
caniniform tooth, anterior to which are two smaller
incisiforms; these are best seen in the holotype skull
(SAM-PK-K337 – Figs 1, 2, 20).
The incisiforms are graded in size, the first being
smaller than the second. There is little evidence of
any constriction between root and crown at the level
of the alveolar border although each crown is most
swollen at its base before tapering to a bluntly
pointed apex; each crown has a recurved profile, with
the mesial edge having a greater curvature than the
distal margin (there appears to be no evidence of the
‘lingual shelf ’ reported by Weishampel & Witmer,
1990). In cross-section each crown is mesiodistally
ovoid, rounded mesially, and slightly tapering distally,
reflecting mild transverse compression of the posterior half. The surfaces of these teeth are neither
ridged nor serrated and appear to be evenly coated
with enamel (Figs 8, 20).
The premaxillary caniniform (pc) is preserved on
the right side of the holotype (SAM-PK-K337 –
Figs 1, 20) but is incomplete because its tip is
obliquely eroded, whereas on the left side the caniniform is sheared off and exposed in longitudinal
section (Fig. 2). The massive subcylindrical root is
contained within the heavily buttressed region of the
premaxilla (between the narial fossa and diastema –
see Figs 1, 4) whereas the crown is transversely compressed, blade-like, and recurved, and reminiscent of
that seen in a wide range of carnivorous archosaurs;
it is slightly flattened transversely, tapers apically,
and its mesial edge is convex, whereas its distal edge
is slightly concave toward the tip (Figs 8, 20). The
distal edge is keeled and bears small square-edged
serrations (approximately six per mm) that wrap
around the keeled edge. The caniniforms in SAMPK-K1332 are not well preserved, and are coated in
consolidant; nevertheless the apical portion of the
keeled edge exhibits a cluster of very similar small
serrations. No equivalently serrated keel can be seen
on the anterior edge of this crown in either specimen. Crompton & Attridge (1986) described an
abrasion facet on the medial surface of the caniniform of SAM-PK-K1332, which they suggested
213
indicates occlusion with the lateral margin of the
predentary beak (note also the discussion relating to
a wear facet on the dentary caniniform below); otherwise these caniniforms show no obvious signs of
wear.
Opposing caniniforms do not appear to have
occluded naturally during jaw closure; their relative
positioning (upper caniniform biting anterior to the
lower) is similar to that seen in other archosaurs with
caniniforms. This pattern is the reverse of the pattern
typically exhibited by synapsids with enlarged
canines: the upper canine biting behind the lower one
and sometimes forming an occlusal facet with the
lower canine.
The maxillary dentition is well preserved in both
the holotype and referred skull and comprises a row
of 11 (SAM-PK-K1332) closely packed columnar teeth
that form a ‘battery’ (Figs 21–24). The elongate rootcrown junctions of maxillary teeth are pressed
together (almost stockade-like) with only a slight gap
between each near the alveolar border; the shieldshaped labial surface of each crown expands toward
the occlusal margin, a feature that is visually accentuated by the curved, thickened margins of the crown.
The maxillary crowns are not expanded either mesiodistally or labiolingually above the root (they lack a
‘neck’ and ‘cingulum’ as typically seen in ornithischians). The crowns are almost all (except for those at
the anterior and posterior ends of the row) abruptly
truncated by a lingually inclined wear surface that
appears to be continuous across adjacent teeth;
however, individual facets produced by predominantly
orthal occlusion are discernible especially at the posterior end of the dentition (Fig. 27 – note also discussion in Hopson, 1980). The crown of each tooth curves
medially along its vertical axis and the occlusal surfaces form angles of: 20–88° with the horizontal axis
(Porro, 2009); as a consequence of incremental change
in wear angulation along the occlusal surface it displays a distinct ‘warp’ along its length. The relative
degree of angulation depends upon the position of
each tooth within the dentition: steeply inclined
facets are found at the anterior end and more oblique
ones in the middle–posterior end of the battery. The
relative size and proportions and degree of wear (and
to some extent degree of eruption) of the maxillary
crowns varies along the dentition: at the anterior end
of the series teeth M.1–M.4 (Fig. 22) are generally
smaller and narrower but steadily increase in overall
size and height posteriorly (Fig. 21); crowns M.5–M.9
are essentially the same size, whereas the posterior
members M.11–M.12 are slightly smaller [this
‘cadence’ is a very common feature in ornithischian
dinosaurs (Norman, 1980; Sereno, 1991a) as well as
diapsids more generally]. The extreme anterior maxillary crowns appear to lean toward the diastema
214
D. B. NORMAN ET AL.
Figure 23. Heterodontosaurus tucki Crompton & Charig, 1962. SAM-PK-K1332 (referred specimen). A, photograph of the
left maxillary dentition as preserved in lateral view. B, photograph of the right maxillary dentition in lateral view.
1962. SAM-PK-K1332 (referred specimen). The upper dentition in: A, occlusal view based on original specimen as
preserved; B, partial reconstruction of the lateral view of
the maxillary dentition alone, with the isolated M.1 added
in position; C, reconstruction of the medial view of M.1–
M.3. For full list of abbreviations see end of paper.
(Figs 21, 22), and exhibit lighter wear than the more
posterior dentition.
The anterior maxillary dentition is sufficiently distinct that it merits detailed description. M.1 (provided that this is not an artefact of preservation) is
almost completely eroded away in the holotype
(SAM-PK-K337 – Fig. 22, ?M.1) making it difficult
to interpret anything except its vague outline.
However, this outline does resemble the first left
maxillary tooth of SAM-PK-K1332 [text retrieved
from the original draft manuscript (ms) by A. W. C.
and A. J. C., and the original specimen is preserved
as a stereo image on file – A. W. C. photographic
archive]. M.1, described by A. W. C. and A. J. C.
(ms) as an isolated tooth recovered from the matrix
beside the skull (Figs 24B, C), is unworn and conforms more closely in appearance to that of ‘typical’
basal ornithischians (Sereno, 1991a). Its root is circular in section at the alveolar border and tapers as
it approaches the crown-root junction. This tooth
does not show a discrete ‘neck’ that separates root
from crown as in Lesothosaurus. The crown is, in
contrast to the root, laterally compressed, diamondshaped and more or less symmetrical in profile
(Fig. 24B, C). The apical denticle dominates the
profile of the crown and the mesial and distal
margins of the crown bear three blunt, rounded
accessory denticles; the mesial and distal edges terminate at a denticle and are thickened thecally to
form a buttress, which extends at a shallow angle
toward the root, thereby framing the lower half of
the diamond-shaped crown. On both lingual and
labial surfaces of the crown, ridges extend thecally
and have the appearance of buttresses supporting
each accessory denticle. In SAM-PK-K1332 the
second and third right maxillary teeth are dislodged
from their alveoli and lie between the maxillae
(Fig. 24A), whereas the remainder on the right side
are in place.
The second maxillary tooth (M.2 – damaged, but
visible in lateral aspect in SAM-PK-K337 – Fig. 22) in
SAM-PK-K1332 (Fig. 24B, C) bears a planar and
steeply inclined wear facet (~88° – Porro, 2009) and
there is a slight constriction at the base of the crown
as it merges with the cylindrical root; ridging (buttressing) associated with the marginal denticles is
again seen labially and lingually, but the crown is
considerably taller and narrower in its overall proportions compared to the first [and more similar in
general form to succeeding ones in having a more
pronounced principal (median) ridge that supports
the apical cusp, and better defined flanking ridges
that create the characteristic shield-like labial crown
surface].
The crown of the third maxillary tooth (M.3 –
Figs 22–24) is sculpted, especially labially (described
below) and its apex is truncated by wear to a greater
extent than M.1 or M.2. The wear facet (observable in
SAM-PK-K1332) is planoconvex and far less steeply
inclined (~66° – Porro, 2009) than that of M.2.
Whereas the roots of maxillary teeth appear to be
straight, cylindrical, and well exposed beyond the
alveolar margin (Figs 21–24), each crown is transversely compressed and the mesial and distal margins
are thickened so that, when viewed across the
occlusal surface, each is approximately rectangular
(Figs 21B, 24). The more posterior teeth are less
strongly transversely compressed, so their crosssections shift from rectangular to square (compare
M.10–M.8, Fig. 21B).
The labial surfaces of the crowns of M.3–M.10/11 are
coated with enamel and truncated abruptly by a sharp,
notched edge (Fig. 21B). The labial face of the crown is
dominated by a shield-shaped depression, bounded by
thickened mesial and distal margins and bisected by a
prominent median principal ridge (p.r.) that originates
as a swollen eminence at the base of the depressed area
and becomes progressively more elevated and nar-
215
rower as it approaches the occlusal margin (Figs 21–
24). This highly distinctive ‘shield-like’ structure to the
enamelled labial surface of the main cheek tooth
crowns is an autapomorphy of Heterodontosaurus. The
labial (cutting) edge of the occlusal surface displays
a characteristic W-shape when viewed occlusally
(Figs 21B, 24A). Some maxillary crowns bear faint
accessory ridges (a.r.), which run more or less parallel
to the principal ridge (Figs 21–24). The lingual surfaces of the maxillary teeth are not at all well exposed
in either the holotype or referred skull, but there does
seem to have been a slight central apicothecal ridge
flanked by thickened edges (as can be deduced from
various occlusal views – Figs 21B, 24). The last maxillary tooth (M.12) is obliquely truncated in the holotype (Fig. 21); in the referred skull the last preserved
maxillary crown (M.11) preserved exhibits a similarly
oblique profile and also retains some large cusps along
its apicodistal edge. Its labial surface is poorly preserved but seems to have been dominated by a median
vertical crease (a feature that can also be seen to be
developing in M.10 – Fig. 23A) rather than a median
principal ridge (Fig. 23B).
Viewed occlusally, the crowns also show slight evidence of ‘imbrication’, despite their stockade-like
arrangement (Fig. 21B) the labiodistal edge overlaps
the adjacent (labiomesial) edge of each succeeding
crown, as commonly reported in ornithischians
(Norman, 1980, 1984a, b; Weishampel, 1984; Sereno,
1991a). Judged by the relative exposure of roots,
crowns, and their degree of wear, the maxillary teeth
in both the holotype and referred specimens did not
grow simultaneously (contra Thulborn, 1978) but
rather emerged in pulses or ‘waves’ akin to the spatial
patterns of eruption termed Zahnreihen (Edmund,
1960; Osborn, 1975; Hopson, 1980). In very general
terms, M.3–M.5; M.6–M.8, and M.9–M.11 form a
series of ‘clusters of eruption triplets’ in which toothwear and degree of eruption increase posteriorly (see
also Hopson, 1980).
The dentary caniniform (dc) is positioned adjacent to
the predentary-dentary suture and resembles the premaxillary caniniform in overall shape, although the
crown is taller than its opposite. In both specimen
s it is more flattened transversely and slightly
procumbent. The right caniniform is preserved
intact in the holotype, is sharply pointed, and fits
tightly into the apex of the diastema (Fig. 1); in contrast, the left is missing, along with its associated
lower jaw (Fig. 2). SAM-PK-K1332 retains both caniniforms (Figs 4, 5, 16–18); however, they are in relatively
poor condition as a consequence of post-mortem
damage, mechanical preparation, and a thick consolidant coating.
Note. An apparent wear facet can be seen on the
lateral tip of the right dentary caniniform of SAM-
216
D. B. NORMAN ET AL.
PK-K1332 (Figs 18A, 25A). If this is not an artefact
then it is puzzling because there is no obvious
caniniform occlusion that might have generated
such a facet. One possibility is that this facet
reflects some malocclusion with the opposing caniniform. Other (much less likely) proposed uses of the
caniniforms for digging and/or some type of agonistic behaviour might be considered as potential
means of accounting for such a facet (see review in
Norman et al., 2004c), but the absence of more
general indications of wear and retention of marginal denticles seems to obviate digging/rooting as
the cause.
The curvature of the mesial edge of the dentary
caniniform exceeds that of the distal margin and the
overall recurvature of the crown is similar to that of
the premaxillary caniniform. In contrast to the
upper caniniform, both mesial and distal edges of
the crown bear serrations. The mesial ‘serrations’
are relatively widely spaced enamel bulges (giving
the impression of their being ‘degenerate’ versions
of the tightly spaced serrations seen on the
distal edge). The distal serrations are similar in
form to those seen in the upper caniniform with a
density of about six per mm and being square in
profile.
The post-caniniform dentition is neither well preserved nor properly visible in the holotype (SAM-PKK337 – Figs 1, 2) because the left jaw ramus is
missing and the right ramus only exposes teeth that
have been damaged during removal of tightly adhering matrix. In SAM-PK-K1332 the lower jaws are
preserved and separated from the skull (Figs 16–18,
25–27); however, fine surface details are somewhat
obscured by a layer of consolidant. The right dentary
teeth of SAM-PK-K1332 (Figs 18, 25–27) display the
clearest details and appear to show a reverse pattern
of imbrication compared to the maxillary battery (the
more prominently thickened mesial margins of each
posterior crown overlapping the distal margin of the
preceding crown – Fig. 27, compare with Fig. 21B).
The pattern of ridging-and-grooving on the medial
surface of the crowns, although broadly similar to
that seen on the labial surface of the maxillary dentition is less pronounced and the positioning of the
principal ridge differs in being offset mesially. The
enamel coating appears to be correspondingly
restricted to the lingual side of the crown. The labial
surface of each dentary tooth is columnar with a
median apicothecal furrow – the mesial ‘column’ is
consistently more pronounced than the distal
(Figs 25–27).
There is some variation in the morphology of the
post-caniniform dentition at the anterior end of the
row, as was the case with the maxillary dentition.
The first post-caniniform (De.2) tooth is dimi-
nutive (as also seen in other heterodontosaurids), is
separated from the dentary caniniform by a gap,
and resembles M.1. It does not reach the general
occlusal plane and is consequently unworn (Figs 25,
26); its lingual surface is dominated by a ridge
extending down from an apical cusp that is
displaced mesially and slightly offsets the diamondshaped profile of the crown. There is a small cusp
anterior to the main cusp and two more behind
it and on both medial and lateral surfaces,
low accessory ridges converge toward the base of
the crown from these cusps. The distribution of
enamel on this crown cannot be assessed, but it
seems probable that it was coated both labially and
lingually.
The remaining dentary crowns (De.3–De.11 –
Figs 25, 26) are truncated by obliquely inclined wear
facets that face apicolabially and are variously orientated (ranging between 38–68° from the horizontal
plane – Porro, 2009). A broadly similar ‘warp’ along
the adjacent occlusal surfaces is seen to that observed
in the maxillary ‘battery’: anterior and posterior teeth
exhibit more steeply inclined wear surfaces (~68°),
whereas the middle teeth are more obliquely worn
(~40°). The worn surfaces resemble those of the
maxilla, although individual occlusal facets are more
readily discernible (Hopson, 1980 – Fig. 27) and differ
from maxillary crowns in being more consistently
scoop-shaped (concave) across the middle of the series
(Figs 25–27). Some wear facets also exhibit a discrete
lip (Fig. 27) that marks the lower edge of the occlusal
facet and the limit of overlap with the opposing maxillary crown – this feature is not found in the maxillary dentition. The labial surface of the dentary
crowns merge with the roots and are essentially
columnar and closely packed; although, rather than
being uniformly cylindrical, each bears a shallow,
median apicothecal furrow flanked anteriorly and
posteriorly by thickenings so that the crown appears
in cross-section (at the occlusal surface) as a ‘pinched’
rectangle.
The enamelled lingual surface of each crown
(Figs 18, 25, 26) bears curved ridges running down
from the mesial and distal margins that meet,
enclosing
a
generally
shield-shaped
surface
(although the fine detail of this topography varies
along the tooth row). The principal ridge (pr) is
offset mesially on the lingual surface (compared
to the centrally located equivalent principal ridge
on maxillary crowns) and divides the shield-like
area asymmetrically into two elongate, shallow
depressions adorned to a varying extent by fainter
parallel accessory ridges that undoubtedly extend as
buttresses from individual (now obliterated) denticles on the crown margin in unworn crowns (see
supporting evidence in Fig. 30). The tenth crown is
217
Figure 25. Heterodontosaurus tucki Crompton & Charig, 1962. SAM-PK-K1332 (referred specimen). Dentary dentition.
A, right dentition in lateral view (top) and medial view (bottom). B, left dentition in lateral view (top) and medial view
(bottom).
218
D. B. NORMAN ET AL.
1962. SAM-PK-K1332 (referred specimen). Line drawings
of the right dental battery (De.2–De.11). A, lateral view; B,
medial view; C, occlusal view of De.2–De.4.
damaged, but its medial surface shows traces of
three or four accessory ridges within a trough-like
depression (Figs 25, 26).
The last dentary tooth (De.11) is small, asymmetrical, and partially sunk into the base of the
coronoid eminence; its occlusal surface is aligned
with those of the preceding teeth. It bears a wear
facet laterally and does not seem to exhibit a strong
primary ridge on its medial face, but rather has a
set of low accessory denticle ridges that diverge
(rather than converge) toward the sloping occlusal
margin.
Dentary crowns (De.3–De.6) exhibit some evidence
of individual wear facets created by opposing teeth,
rather than a continuous cutting surface (as first
recognized by Hopson, 1980). De.7–De.9 (see Fig. 27)
show some evidence of double wear facets; each tooth
exhibits wear facets that are confluent with those on
adjacent crowns. The posterior occlusal surface of
De.8 is raised into a lip level with the anterior wear
facet of De.9. De.9 and De.10 both feature a faint
ridge that crosses the occlusal surface obliquely and
defines the edges of discrete anterior and posterior
wear facets.
Tooth wear patterns in the dentary battery are
less consistently indicative of replacement sequences
than in the maxilla. De.4–De.6 represent one
sequence that displays successive and increasing
levels of wear, consistent with the triplet pattern
seen in the maxillary dentition; however, the other
dentary cheek teeth, although they display varying
1962. SAM-PK-K1332 (referred specimen). Line drawings
based on scanning electron micrograph images of right
dentary teeth: De.7–De.9. A, worn surfaces of the teeth in
oblique dorsolateral aspect showing the lipped lower edges
of the wear facets and the minor imbrication between
crowns (the overlap pattern is the reverse of that seen in
the maxillary dentition – see Fig. 21B). B, close-up of De.9
to show details of the wear facet and the lower ‘lip’ at the
base of the wear facet. For full list of abbreviations see end
of paper.
degrees of eruption and wear, do not form such
obvious groupings.
EVIDENCE OF TOOTH REPLACEMENT
HETERODONTOSAURUS
IN
Apart from the logical expectation of tooth replacement associated with skull enlargement during
growth, the staggered patterns of eruption and correspondingly differential levels of wear within the
dental batteries in upper and lower jaws in the two
adult skulls of H. tucki suggest that tooth growth and
replacement was an integral part of its life-history
(Hopson, 1980). Furthermore, tooth replacement is
the only mechanism that would allow tooth size to
increase while maintaining the distinct morphology of
the cheek teeth. What is remarkable is that none of
the cranial remains of Heterodontosaurus described
so far (even the apparent juvenile specimen – Butler
et al., 2008a) show any evidence of active in situ tooth
replacement. This surprising condition has provoked
a number of speculations concerning the life history
and mode of jaw action in this genus (summarized in
Norman et al., 2004c).
SAM-PK-K1334 (Figs 30–33). This specimen was
first mentioned (without a specimen number) by
Thulborn (1970b: 243), who identified it as Heterodontosaurus and briefly characterized its dentition.
Charig & Crompton (1974: 172) suggested that the
information provided by Thulborn (1970b) relating to
this specimen was inaccurate and stated that the
maxilla differed from the holotype of H. tucki in
several respects (although they did not specify or
discuss these features) and noted the presence of
unerupted replacement teeth. Charig & Crompton
(1974: 185) alluded to this specimen as ‘. . . the incomplete maxilla of what appears to be another heterodontosaurid from the Stormberg Series with functional
teeth possessing the typical characters of the family
but also with two unerupted replacing teeth and other
evidence of replacement’. They neither described nor
figured this material, which was later referred to by
Hopson (1980: 103), who repeated the quote. No specimen number, locality information, figures, or formal
description of the maxilla were provided, but the
specimen was described and illustrated in their
unpublished ms notes. The specimen has, since their
drafting of the original ms, been further prepared to
expose one of the more obvious replacement crowns
and CT scanned (Figs 31–33).
General description
SAM-PK-K1334 comprises the posterior part of the
left maxilla, the incomplete and eroded anterior
ramus of the left jugal, and the incomplete left
lacrimal. Segmented CT scans (Figs 30, 31, 33), in
which the sediment encasing the specimen has been
digitally stripped away show an additional transversely compressed fragment of bone dorsal to the
maxillary shelf (fr). This fragment may represent
either a fragment of the palate, or a portion of the
medial lamina of the maxilla that contributed to the
wall of the antorbital fossa; the fragment does not
contact the lacrimal. The maxilla contains seven fully
erupted teeth (Figs 30–33: ‘M.1’–‘M.7’): six of which
are reasonably well-preserved, erupted teeth with
worn crowns and a broken fragment of a seventh
tooth (‘M.1’) anteriorly. Above the tooth row, the
lateral surface of the maxilla is broken along the
ventral edge of the external antorbital fenestra
(Fig. 30A, br). The preserved parts of the lacrimal
219
(La) and jugal (J) formed the anteroventral margin of
the orbit (Fig. 30A, orb). Crowns are chisel-shaped
in profile and broadest at the ventrolateral occlusal
edge (where they are truncated by wear lingually
and somewhat damaged laterally) and taper gently
toward the root (Figs 30–33). The crowns are tightly
packed, with anterior and posterior edges of adjacent
crowns contacting one another at the occlusal surface.
The crowns although imperfectly preserved show no
obvious imbrication and their labial surfaces exhibit
the shield-like pattern of ridges and grooves very
similar to the pattern described in both the holotype
and referred specimens (SAM-PK-K337, K1332). As in
these latter examples, there is sporadic development
of accessory ridging: in ‘M.7’ a single accessory ridge
is present between the mesial and principal ridges,
and two accessory ridges are present between the
principal and distal ridges. A single accessory ridge
can also be identified between the primary and posterior ridges of ‘M.6’, and between the primary and
anterior ridges of ‘M.4’. Accessory ridges were
undoubtedly present in other crowns but have been
obliterated by wear/damage to the crown surfaces.
The lingual surfaces of erupted crowns are dominated
by wear facets, but there is a broad and rounded
median ridge, separated from mesial and distal ridges
by shallow grooves (‘M.5’ and ‘M.6’ – Fig. 32A). The
mesial and distal ridges, and the grooves that separate them from the median ridge, are less strongly
developed on the lingual surface than the labial.
Crowns ‘M.4’ and ‘M.6’ are not fully erupted; this
contrasts with other specimens of Heterodontosaurus
(SAM-PK-K337, K1332) in which the crown bases are
fully visible in lateral view along the entire length of
the tooth row (Figs 21, 23). Crown size decreases
anterior to ‘M.4’. In occlusal view, crowns ‘M.1’–‘M.4’
form a closely packed array that is slightly offset from
a triplet of closely grouped crowns ‘M.5’–‘M.7’. Large,
high angle (at about 70–80° degrees to the horizontal)
planar wear facets are present on all well-preserved
crowns and form an apparently continuous surface
across adjacent crowns. Towards the anterior end of
the tooth row (‘M.2’), the wear facet is relatively close
to the alveolar margin.
Evidence of tooth replacement
In a carefully argued paper Hopson (1980) demonstrated that apparently ontogenetically mature specimens of H. tucki (SAM-PK-K337, K1332) with heavily
worn dentitions still exhibited remnants of differential tooth eruption and tooth wear that could be
explained only by phases of active tooth replacement;
this was also the conclusion of Butler et al. (2008a) on
the basis of an ontogenetically immature specimen
(SAM-PK-K10487 – Figs 28, 29). The specimen
220
D. B. NORMAN ET AL.
Figure 29. Heterodontosaurus tucki. SAM-PK-K10487
(referred specimen). Partial skull of a juvenile heterodontosaur as preserved in: A, oblique left lateral aspect to
show details of the maxillary dentition; B, ventral view.
For full list of abbreviations see end of paper.
(referred specimen). Partial skull of a juvenile heterodontosaur as preserved in: A, left lateral aspect; B, right
lateral aspect. For full list of abbreviations see end of
paper. (See also Butler et al., 2008b.)
described here provides the first unambiguous evidence of tooth replacement in this genus.
A replacement crown (rep ‘M.2’) is clearly visible
medial and dorsal to erupted crown ‘M.2’ (Figs 30B,
31B, 32); this crown has been exposed by removal of
the medial surface of the maxilla (Fig. 32A, B). The
crown, which was only partially mineralized, exhibits
some damage, although primary and accessory ridges
are present. Four denticles are present along the
distal margin (between the most distal denticle and
the apex) and are supported by weak accessory ridges
that extend on to the lingual crown surface (Fig. 32).
Similar accessory ridges are also present mesial to
the principal ridge. These denticles and accessory
ridges would have been quickly obliterated by high
rates of wear, as is evident in the fully erupted
teeth.
Just dorsal to the alveolar margin, on the medial
surface of the maxilla there is a shallow, trough-like,
linear feature that appears to represent the groove for
the vascular and neural supply to the dental lamina
(Figs 31, 32, gr). The surface of the maxilla beneath
the groove is depressed relative to the general medial
surface, and its texture is more ‘spongy’ compared
with that above the groove. [Similar bone textures
are also observed along the medial alveolar margin of
the posterior portion of the dentary of a small
fragmentary heterodontosaur skull, which also shows
evidence of tooth replacement (SAM-PK-K10488) – L.
B. Porro, unpubl. data.] Small pits or notches
(Edmund, 1960: ‘special foramina’) may be present
adjacent to this groove (Fig. 32A) above ‘M.3’ and
‘M.4’ (although these might also be simply a reflection
of erosional and/or preparation damage); however,
those associated with ‘M.5’ and ‘M.7’ appear to be
genuine, and a similar pit was present dorsal to
crown ‘M.2’, before mechanical preparation was
undertaken (preserved in the archive of documents
relating to Heterodontosaurus at the Sedgwick
Museum). The tip of a replacement crown (with apical
and mesial/distal cusps) can be seen within the pit
(referred specimen). Posterior portion of the left maxilla.
Segmented image prepared from computed tomography
scans of the original. A, lateral aspect. Bones displayed as
grey, erupted crowns shown in yellow. B, semi-transparent
image of the maxilla to show the extent and arrangement
of individual teeth in the maxilla. Red areas indicate the
presence of replacement crowns embedded within the body
of the maxilla. For full list of abbreviations see end of
paper.
above crown ‘M.5’ (Fig. 32A, rep) and is better visualized in the segmented CT-based image (Fig. 31B,
rep ‘M.5’); a third replacement crown (rep ‘M.7’) – not
visible externally – has also been visualized using a
segmented CT image of the maxilla (above crown
‘M.7’ – see Fig. 31B).
Replacement teeth lack mineralized roots and are
triangular in lateral outline with a clear apex
marking the median principal ridge on the labial
surface. The principal ridge is flanked by thickened
margins and separated by deep troughs, as in other
specimens of H. tucki. Thus, the original shape of the
221
(referred specimen). Posterior portion of the left maxilla.
Segmented image prepared from computed tomography
scans of the original. A, medial aspect. Bones in grey,
erupted crowns shown in yellow, replacement crowns
shown in red. B, semi-transparent image to show the
position and extent of erupted and replacement teeth
within the maxilla. For full list of abbreviations see end of
paper.
unworn ‘cheek’ teeth of Heterodontosaurus is triangular, resembling the shape of other basal ornithischian
teeth; it is heavy tooth wear that produces the distinctively truncated ‘chisel-edge’ seen in the functional dentition.
The cheek teeth vary in their degree of eruption:
crown ‘M.2’ is completely erupted and erosion of its
root (Figs 30B, 33A) shows that it was in the process
of being replaced. The degree of eruption decreases
posteriorly in crowns ‘M.3’ and ‘M.4’ (Fig. 30A), suggesting that these three teeth form a ‘replacement
triplet’. More posteriorly, crowns ‘M.5’ and ‘M.7’ are
more completely erupted and are also in the process
of being replaced, whereas crown ‘M.6’ is considerably
222
D. B. NORMAN ET AL.
Figure 32. Heterodontosaurus tucki. SAM-PK-K1334 (referred specimen). Medial view of the alveolar wall and adjacent
bone and crowns. A, line drawing of the dentition showing the details of the groove (gr) that connects ‘special foramina’
(‘pit’) above the alveolar margin, the apical tip of replacement crown ‘M.5’ and the revealed crown of ‘M.2’; the anterior
half of this area is quite badly eroded so bone surface details are unclear. B, magnified detail of the replacement crown
above ‘M.2’. For full list of abbreviations see end of paper.
less erupted and has a substantial hollow root that
extends close to the upper surface of the maxilla
(Fig. 31B). CT data (Figs 30–33) show that the roots
of teeth are elongate and tubular, with parallel anterior and posterior margins for most of their length.
The roots of the maxillary teeth penetrate deep into
the maxilla and are visible (in CT scans) protruding
from the dorsal surface of the maxillary shelf
(Fig. 31A, rt ‘M.4’); this condition is also seen in the
presumed juvenile specimen (SAM-PK-K10487). The
roots are inclined slightly distally toward their thecal
bases. In anterior or posterior view the lateral surface
of the root is convex (Fig. 33B), whereas the medial
surface is relatively straight; thus the surfaces converge apically (to form the laterally compressed
crown) and basally. The roots are hollow with
extensive pulp-cavities. In those teeth undergoing
replacement the roots are in the process of being
resorbed medially.
Tooth replacement patterns in Heterodontosaurus
Edmund (1960) provided the first comprehensive study
of tooth replacement in nonmammalian amniotes. He
described patterns of tooth eruption that sweep
through alternating tooth positions along the jaw and
related this to an ontogenetic scheme involving pulses
of tooth-growing activity that moved posteriorly along
the jaw. Each pulse was termed a Zahnreihe, a term
that originates in the work of Woerderman (1919); the
spacing between adjacent Zahnreihe (in terms of
numbers of tooth positions) is referred to as its
Z-spacing. DeMar (1972) demonstrated that the apparent direction of replacement waves correlates with
Z-spacing. A Z-spacing > 2.0 generates an anteriorly
directed replacement wave and a Z-spacing of
approximately 3.0 has been demonstrated for the
heterodontosaurid Lanasaurus by Hopson (1980),
based upon the pattern of eruption of teeth in sequential triplets within which the teeth become older
posteriorly (Hopson, 1980: fig. 5). A similar Zahnreihederivable pattern of consecutively emergent crowns is
seen in SAM-PK-K1334 (Figs 30, 33B): heavily worn
functional crown ‘M.2’ (which is in the process of being
replaced), the less extensively worn functional crown
‘M.3’, functional crown ‘M.4’ (which has not completely
erupted above the alveolar margin), followed by the
(normally invisible) unerupted replacement crown
above ‘M.5’ (Figs 31B, 33). However, it should be noted
that the replacement crown above ‘M.7’ does not
conform to the Zahnreihe pattern.
223
Figure 33. Heterodontosaurus tucki. SAM-PK-K1334 (referred specimen). Isolated teeth, extracted from the original
computed tomography image. A, medial/lingual aspect to show the long hollow roots to the fully erupted crowns and the
position of the replacement crowns on the medial side of the roots of functional teeth. B, oblique view of the same to show
the curvature of the roots and crowns, the angulation of the wear facets and the relative positions of the replacement
crowns. For full list of abbreviations see end of paper.
The developmental significance of Zahnreihe has
been challenged consistently (Osborn, 1970, 1971,
1975 see also Fastnacht, 2008). However, the term
has some utility as a descriptor of geometry and order
within tooth rows, and for this reason more than any
other it has been used widely in the palaeontological
literature. It is noteworthy that SAM-PK-K1334 is
the first South African heterodontosaurid to be
described with ‘special foramina’ and exhibits tooth
replacement (see also the Morrison Formation heterodontosaurid Fruitadens Butler et al., 2010). This
observation adds some support to the idea that these
foramina are linked to replacement (Edmund, 1957),
although the actual functional relationship between
the foramina and tooth eruption requires further
investigation.
Note on the taxonomic identity of SAM-PK-K1334
SAM-PKK1334 closely resembles the holotype (SAMPK-K337) and referred specimens (SAM-PK-K1332,
SAM-PK-K10487) of H. tucki in possessing a number
of features that were discussed by Butler et al.
(2008a) as potential autapomorphies of this taxon:
1. Columnar maxillary teeth lack anteroposterior or
mediolateral expansion above the root (i.e. the
typical ornithischian ‘neck’ and ‘cingulum’ are
absent).
2. The lateral surface of the maxillary crowns has a
‘shield-shaped’ structure enclosed within prominent, curved anterior and posterior ridges, which
are bisected by a primary ridge that separates
flute-shaped recessed areas.
3. Maxillary teeth are closely packed and form
an inclined occlusal blade with small gaps
between the teeth present only near the alveolar
margin.
4. Maxillary teeth are transversely expanded relative
to their anteroposterior length and exhibit heavier
wear and planar single wear facets compared
with examples of Lycorhinus, Abrictosaurus, or
Lanasaurus.
224
D. B. NORMAN ET AL.
In addition to possessing autapomorphies of
Heterodontosaurus, SAM-PK-K1334 can be distinguished from other known southern African
heterodontosaurids:
1. The maxillary teeth of A. consors (NHMUK RU
B54; Thulborn, 1974: figs 3, 39A, B) lack prominent median, anterior, and posterior ridges, are
less closely packed, and less heavily worn than
SAM-PK-K1334.
2. The holotype of Ly. angustidens (SAM-PK-3606 –
Fig. 37A) is a dentary only and cannot be directly
compared to SAM-PK-K1334.
3. Other South African heterodontosaurid specimens
(see Taxonomic review below) include the holotype
specimen of La. scalpridens (BP/1/4244 –
Fig. 39C), NHMUK RU A100 (Thulborn, 1970b –
Fig. 38) and BP/1/5253 (Gow, 1990). All of these
specimens possess maxillary teeth that differ from
SAM-PK-K1334 in possessing a prominent basal
‘cingulum’, a less well-developed primary ridge,
and a well-developed ridge along the posterior
margin of the crown that is substantially better
developed than the equivalent ridge on the anterior margin; those that have teeth also exhibit
double, and clearly angled, wear facets, rather
than apparently flush occlusal surfaces.
SAM-PK-K1334 cannot therefore be referred to
Abrictosaurus, Lycorhinus, or Lanasaurus.
Several features distinguish SAM-PK-K1334 from
other specimens of Heterodontosaurus:
1. Active tooth replacement (replacement crowns,
groove for dental lamina punctuated by ‘special
foramina’ and demonstrable presence of replacement crowns).
2. High-angle wear facets (between 70–80° to the
horizontal) – in other specimens of Heterodontosaurus the facets are more variably inclined along
the dentition (ranging between 30–80° to the horizontal) depending upon their position within the
battery. The roughly equivalent tooth positions in
the holotype and referred skulls have lower-angle
facets.
3. The wear facets extend close to the alveolar
margin.
The major difference between SAM-PK-K1334 and
the holotype and referred specimens of H. tucki is
the evidence of active tooth replacement, and it
seems probable that characters 2 and 3 above are
correlated and simply a consequence of the
advanced ontogenetic age of the functional dentition:
wear angulation and extent may well reflect the
absence of well-developed roots and relative mineralization (enamel : dentine) of individual crowns.
Hopson (1975, 1980) suggested that the absence of
tooth replacement in the holotype and referred
skulls of H. tucki reflected their ontogenetic maturity, and that immature individuals of H. tucki probably replaced their teeth continuously. If Hopson’s
hypothesis is correct then SAM-PK-K1334 might
represent a juvenile specimen of H. tucki. However,
SAM-PK-K1334 is close in size to the holotype specimen (SAM-PK-K337) of H. tucki, in which there is
no evidence of active tooth replacement. Furthermore, CT scans of a smaller, probable juvenile individual of H. tucki (SAM-PK-K10487) show
no evidence for tooth replacement at an earlier ontogenetic stage than that represented by SAM-PKK1334 (Butler et al., 2008a).
It should be noted that the difference in absolute
tooth size between the ‘immature’ (SAM-PK-K10487)
and ‘mature’ (SAM-PK-K337, SAM-PK-K1332) individuals indicates that replacement must have
occurred. Butler et al. (2008a) speculated that tooth
replacement in H. tucki was episodic rather than
continuous during growth; if so, SAM-PK-K1334
might represent an individual of this species that
died during one of these replacement events. Alternatively, SAM-PK-K1334 could be a second, closely
related species of Heterodontosaurus with a different
ontogenetic trajectory. With regard to the other differences, ontogenetic maturation in the functioning
dentition (with the central portion of the occlusal
surface becoming increasingly warped as the teeth
lock into position in the dental battery and become
functionally adapted to jaw action) may explain the
comparative steepness and planar nature of the
occlusal surfaces in this specimen, as well as the
minimal emergence of the crowns from the alveoli.
The limited samples (both taxonomic and ontogenetic) of southern African heterodontosaurids do not
permit conclusive resolution of these anatomical
inconsistencies; nevertheless, on the current information reference of SAM-PK-K1334 to H. tucki
seems justified.
SAM-PK-K1334 provides the first unequivocal
example of active tooth replacement in heterodontosaurids; moreover, the evidence of ‘waves’ of tooth
replacement indicates that the dentition was not
replaced as a single unit (confirming the conclusions
of Hopson, 1980). Although these observations do
not support the hypothesis that the entire heterodontosaur dentition was replaced en masse during
seasonally induced periods of aestivation (Thulborn,
1974, 1978). Nevertheless, the evidence that heterodontosaurs indulged in sporadic episodes of rapid
tooth replacement that generated a stable, clearly
hypsodont (high-wear adapted) dentition linked to an
unusual and complex jaw mechanism (Porro, 2009)
hints at an extremely interesting set of interactions
between the ontogeny, functional biology of feeding,
and ecology of these ornithischians within the Early
Jurassic Karoo environment.
CRANIAL MYOLOGY RECONSTRUCTED:
HETERODONTOSAURUS JAW FUNCTION SUMMARIZED
Myological reconstructions
Preservation of the holotype and referred skulls of
H. tucki (SAM-PK-K337, K1332) although far from
perfect is sufficient to determine the distribution and
orientation of the principal jaw-closing muscles.
Reconstruction of the adductor musculature of Heterodontosaurus is required in order to understand the
general function of the jaws and teeth during feeding
(Porro, 2009). Additionally, these myological reconstructions will be used in biomechanical analyses
(including finite element analysis) to estimate forces
generated by the adductor muscles, and associated
reaction forces at the teeth and jaw joints (Porro,
2009; L. B. Porro, unpubl. data).
The reconstruction of soft tissues (such as muscles)
in fossil vertebrates is inevitably speculative, and the
risk of amplification of errors when larger-scale interpretations (function, mode of life, palaeoecology) are
based on inaccurate reconstructions is significant. The
attachment sites of unpreserved muscles in fossil taxa
can be identified by distinct marks (osteological correlates – Witmer, 1995) left by muscles on bone surfaces.
Muscles attach either directly to bone or via tendons
and aponeuroses. Direct (fleshy) attachments (in
which the muscle fibres merge with the periosteum)
rarely produce marks whereas tendinous or aponeurotic attachments are characterized by ossification
at the connective tissue–bone interface that produces
either a smooth surface or ‘scarring’, respectively
(Bryant & Seymour, 1990). Bryant & Seymour (1990)
showed from dissection that muscles often have larger
areas of attachment than those indicated by osteological correlates, implying that estimates of muscle size/
force from such features should be treated cautiously.
Nevertheless, osteological correlates do provide evidence for major muscles (Nicholls & Russell, 1985) and
reconstructions of cranial musculature have been
attempted for a wide range of ornithischian dinosaurs:
Lesothosaurus (Thulborn, 1971a), ankylosaurs (Haas,
1969), Hypsilophodon (Galton, 1974), hadrosaurs
(Ostrom, 1961; Rybczynski et al., 2008), pachycephalosaurs (Maryańska & Osmólska, 1974; Sues & Galton,
1987), and ceratopians (Haas, 1955; Ostrom, 1964;
Holliday, 2009; Sereno, Xijin & Lin, 2010). Furthermore, Holliday & Witmer (2007) demonstrated that
despite radical changes in skull morphology, the
adductor muscle topology of living diapsids is surprisingly conservative.
Adopting some measure of phylogenetic constraint
as proposed by Bryant & Russell (1992) and Witmer
225
(1995) constrains speculation and permits a level of
confidence to be assigned to soft tissue reconstructions in extinct taxa. In both these methodologies,
reliance was placed upon the distribution and form of
soft tissues and their osteological correlates in phylogenetically related living taxa. If anatomical features
are equivocal or the fossil taxon exhibits a morphological novelty, then extrapolation may be required in
the Bryant & Russell (1992) approach (form-function
relationships, biomechanical principles). Witmer’s
(1995) extant phylogenetic bracket (EPB) offers the
potential to assign a level of confidence to soft tissue
reconstructions. This approach requires the identification of osteological correlates and corresponding
soft tissues in living taxa that phylogenetically
‘bracket’ the fossil in question; the presence of these
osteological correlates in the extinct taxon strongly
suggests the presence of corresponding soft tissues.
Soft tissue reconstructions linked to any autapomorphic anatomical features present in the fossil are
assigned a lower level of confidence. Thus, following
Witmer (1995), each muscle is assigned a level of
inference if: the corresponding osteological correlate
is present in both extant bracket taxa and the fossil
taxon (Level I); one extant bracket taxon and the
fossil taxon (Level II); or only the fossil taxon (Level
III). If no osteological correlates are present amongst
the taxa despite the presence of the muscle, the level
of inference is assigned a ‘prime’ designation (I’, II’,
III’). Sites of cranial origin and mandibular insertion
can be assigned different levels of inferences.
Although attachment sites and osteological correlates
for each muscle in extant bracketing taxa are briefly
described here, Holliday (2009) should be consulted
for detailed information.
Ornithischian dinosaurs such as Heterodontosaurus
are phylogenetically bracketed by a large array of
fossil taxa that separate them from extant members
of the clade: crocodilians and birds; furthermore, both
living outgroups have highly derived (and distinct)
crania and jaw musculature. Bryant & Seymour
(1990) refer to this particular problem in ornithischians and suggest using a range of more remotely
related extant diapsid taxa when reconstructing softtissues in these animals. Numerous studies detailing
the cranial myology of crocodilians (Iordansky, 1964,
1973; Schumacher, 1973; van Drongelen & Dullemeijer, 1982; Busbey, 1989; Cleuren & De Vree, 2000;
Rayfield, 2001; Holliday & Witmer, 2007), birds
(George & Berger, 1966; Vanden Berge & Zweers,
1993; Holliday & Witmer, 2007), lepidosaurs, and
Sphenodon (Iordansky, 1970; Haas, 1973; Throckmorton, 1976; Gorniak, Rosenberg & Gans, 1982; Smith,
1982) were consulted in order to reconstruct the jaw
adductor muscles of Heterodontosaurus. Additionally,
the head of a Caiman was dissected to identify muscle
226
D. B. NORMAN ET AL.
attachment sites and osteological correlates, and the
dry skulls of various crocodilian and lepidosaur
species, and Sphenodon, were examined for osteological correlates.
The principal jaw muscles of Heterodontosaurus
tucki
Jaw muscles are divided according to their position
relative to the trigeminal (cranial nerve V – Luther,
1914; Lakjer, 1926) resulting in four groups:
M. adductor mandibulae externus superficialis
(MAMES), M. adductor mandibulae posterior
(MAMP – which includes the pterygoideus MPT),
pseudotemporalis (MPST), and M. constrictor internus dorsalis (MCID).
M. adductor mandibulae externus
superficialis (MAMES)
Origin (Level 1): MAMES originates on the ventral
surfaces of the quadratojugal and quadrate of crocodilians; the lateral surface of the squamosal in birds;
and the ventral surface of the upper temporal bar in
lepidosaurs. Many dinosaurs exhibit a longitudinal
ridge that dorsally bounds a depression on the ventrolateral surface of the upper temporal bar; this is
widely considered the origin of mAMES. In Heterodontosaurus, a prominent ridge frames the upper and
anterior parts of the infratemporal fenestra (Fig. 34 –
MAMES). It extends from the posterolateral margin
of the orbit dorsally along the upper temporal bar and
across the lateral edge of the squamosal; a smooth
depressed area is enclosed by this ridge and indicates
the origin of MAMES.
A faint oblique ridge on the upper temporal bar
divides the presumed origin of MAMES into anterior and posterior portions. The main body of
MAMES is generally associated with the upper temporal bar and squamosal (Fig. 34A). The excavated
posterolateral surface of the ventral process of the
postorbital appears to have been the origin for a
separate (but topologically related) extra portion of
MAMES (referred to here as MAMESX – Level III
inference). The only other ornithischian that bears a
similar postorbital structure is the unusual ornithopod Zalmoxes (Weishampel et al., 2003: figs 2,
8A).
The medial (deep) extent of this MAMES is uncertain but is taken here as the ventral edge of the upper
temporal bar.
Insertion (Level I): This muscle attaches to the dorsolateral margin of the surangular, posterior to the
coronoid process, in crocodilians and lepidosaurs;
and on the lateral surface of the lower jaw in birds.
The dorsolateral surface of the surangular is the
Figure 34. Heterodontosaurus. Attempted reconstruction
of some of the principal adductor musculature. A, M.
adductor mandibulae externus superficialis (MAMES)
showing a subdivision into portions associated primarily
with the postorbital and squamosal (MAMESX in the
text), respectively. B, M. depressor mandibulae (MDM).
Dark shading represents visible musculature, light
shading indicates the path taken by muscle covered by
superficial bones.
most likely insertion for MAMES in non-avian dinosaurs (Holliday, 2009). The lateral surface of the
lower jaw in Heterodontosaurus has a complex
surface with two potential areas for insertion of
MAMES/MAMESX: the surangular is highly modified as a pair of arched struts that lie along the
dorsal edge of the lower jaw; the lower of these two
struts is partially fused to the angular and offers
one area of insertion (which would conform with the
majority of reconstructions seen in archosaurs and
dinosaurs). However, the lateral surface of the
angular is concave and edged, ventrally, by a distinct thickened rim (see Fig. 34B). This more robust
portion of the lower jaw is regarded as the most
probable area of an aponeurosis for this important
muscle – the ventral surangular strut is here
regarded as a more probable area of insertion of
MAMEM (see below).
227
medialis (MAMEM)
Origin (Level I): MAMEM originates from tendinous
insertions on the ventral surface of the quadrate in
crocodilians; in noncrocodilian diapsids (which have
vertically orientated quadrates, as does Heterodontosaurus) MAMEM originates on the medial surface of
the upper temporal bar and posterolateral surface of
the braincase; such an origin for MAMEM has been
proposed for many dinosaurs as well. In Heterodontosaurus the medial surface of the upper temporal bar
bears a longitudinal ridge that marks the dorsal and
medial (deep) extent of MAMEM (Fig. 34B). MAMEM
extended anteriorly to the frontal-postorbital suture,
where the anterior margin of the supratemporal
fenestra is represented by a bevelled edge (Fig. 12).
The posterior extent of MAMEM cannot be determined, but is assumed to have reached the
squamosal-parietal suture. Its lateral (superficial)
extent is taken as the ventral edge of the upper
temporal bar.
Insertion (Level I’): MAMEM inserts onto the dorsal
surface of the surangular (medial to the insertion for MAMES) in crocodilians and to the
coronoid region in birds. In all extant diapsids, the
insertion of MAMEM is aponeurotic (Holliday, 2009).
In most dinosaurs, MAMEM is reconstructed as
occupying a space from the coronoid process to the
jaw joint that is intermediate in position between
the insertions of MAMES (laterally) and MAMEP
(medially).
MAMEM probably inserted along the dorsal
margin of the ventral ramus of the surangular
(Fig. 34B) of Heterodontosaurus, which features a
smooth surface.
profundus (MAMEP)
Origin (Level I): MAMEP originates from the
supratemporal fossa (including the laterosphenoid,
parietal, and squamosal) of crocodilian and birds; it
originates from the posterior supratemporal fossa
(prootic, parietal) and post-temporal fenestra in
lizards. Typically this muscle is reconstructed
originating on the medial surface of the supratemporal fenestra and extending onto the sagittal crest
(when present) in dinosaurs (Ostrom, 1961; Galton,
1974), including Heterodontosaurus (Fig. 35A). The
relationship between MAMEM and MAMEP cannot
be determined in the known specimens of Heterodontosaurus; the boundary between these muscles is
taken as the parietal-squamosal contact. In contrast,
a faint vertical swelling on the lateral surface of the
braincase (laterosphenoid, parietal) above the
of some of the principal adductor musculature. A, M.
adductor mandibulae externus posterior (MAMEP) attaching to a presumed bodenaponeurosis on the apex of the
coronoid bone; M. adductor mandibulae posterior (MAMP)
restored in its most probable position. B, M. pseudotemporalis (MPS) shown inserting on the presumed coronoid
bodenaponeurosis. M. pterygoideus posterior (MPTP)
originating on the posteroventral surface of the pterygoid,
and shown inserting on the lateral surface of the surangular beneath the jaw joint. Dark shading represents
visible musculature, light shading indicates the path
taken by muscle covered by superficial bones.
trigeminal foramen in Heterodontosaurus is taken as
the boundary between MAMEP and mPST; such as
swelling has been noted in other dinosaurs (Ostrom,
1961; Thulborn, 1971a; Galton, 1974; Rayfield, 2001;
Holliday, 2009).
Insertion (Level I): MAMEP attaches (via aponeurosis) to the coronoid process in all extant diapsids; as
a result, the insertion site is characteristically rugose.
MAMEP is usually reconstructed as inserting on the
coronoid process or surangular, medial to the insertion of MAMEM. MAMEP is presumed to have
inserted, probably via an aponeurosis, along the dorsomedial margin of the lower jaw, primarily the
228
D. B. NORMAN ET AL.
dorsal surface of the coronoid, in Heterodontosaurus.
The posterodorsal portion of the coronoid is exceptionally rugose in Heterodontosaurus and this probably
reflects the insertion of a anteroposteriorly restricted
and yet strong bodenaponeurosis (Fig. 35A). The adjacent upper limb of the surangular articulates at an
apparently partly mobile junction with the dentary
immediately posterior to the coronoid; this area would
be an unlikely one for the attachment of significant
jaw adductors.
M. adductor mandibulae posterior (MAMP)
Origin (Level I): MAMP originates on the anterior
surface of the shaft of the quadrate in all extant
diapsids, leaving a series of fossae and crests; it is the
most phylogenetically/anatomically consistent muscle
in the adductor chamber (Holliday, 2009). The anterior surface of the quadrate is deeply incised in Heterodontosaurus, and bordered medially and laterally
by sharp edges; this area was undoubtedly the origin
of MAMP in Heterodontosaurus (Fig. 35B). Although
Sues & Galton (1987) suggested the pterygoidquadrate wing as a possible site of origin for MAMP,
this wing is very thin in Heterodontosaurus and is an
unlikely attachment site for large jaw adductor
muscles.
Insertion (Level I): MAMP inserts within the mandibular fossa (and on to Meckel’s cartilage) in
crocodilians, birds, and lizards, and is typically reconstructed as inserting into the mandibular fossa of
dinosaurs. The most probable area of insertion of
MAMP would have been the floor of the adductor
fossa in Heterodontosaurus, where the jaw is thickest
and reinforced by the sutures and overlapping sheets
of the splenial, angular, dentary, and prearticular.
M. pseudotemporalis (MPS)
Origin (Level I): Living diapsids exhibit two portions
of MPS. The superficial portion originates on the
laterosphenoid; the deep portion originates on the
laterosphenoid of crocodilians and birds, and on the
epipterygoid of lizards. The epipterygoid is absent in
Heterodontosaurus as is the case in most dinosaurs
[although retained in basal sauropodomorphs (P.M.
Barrett, 2010, pers. comm.) and the basal thyreophoran Scelidosaurus (D. B. Norman, unpubl. data)], the
origin of the deep portion is unclear, and it is possible
that this portion was lost. Thus, MPS is treated as a
single muscle in this study. The anterior limit of MPS
is marked by the bevelled edge of the supratemporal
fossa on the frontal anteriorly; its lateral extent is
taken as the frontal-postorbital suture. Dorsally, MPS
extended onto the anterior half of the sagittal crest
(Fig. 35B; the posterior portion of the crest was occu-
pied by MAMEP). A lateral swelling of the braincase
marks the boundary between MPS (anteriorly) and
MAMEP (posteriorly).
Insertion (Level I): MPS inserts onto the medial
surface of the coronoid process in lizards; in crocodilians and birds, it attaches to the medial coronoid and
mandibular fossa via tendons (in crocodilians, this
tendon includes the sesamoid cartilago transiliens).
The cartilago transiliens may leave a small facet on
the medial surface of the lower jaw; its inferred
presence in dinosaurs is a Level II’ inference. Instead,
Holliday (2009) suggested that MPS attached to the
coronoid process of many ornithischians, including
ornithopods and ceratopsians; in these animals, the
coronoid is well developed and heavily striated. Likewise, it seems probable that MPS inserted on the
coronoid process of Heterodontosaurus which, as
noted above, is heavily scarred for such attachment.
Crocodilians and some birds possess a muscle
extending into the Meckelian canal referred to as M.
intramandibularis. Holliday & Witmer (2007) identified M. intramandibularis as a part of MPS. CT scans
show the Meckelian canal stretching anteriorly to the
level of the fifth dentary tooth in Heterodontosaurus.
The canal is large posteriorly but tapers abruptly
below the eighth dentary tooth. An anterior extension
of MPS may have occupied the posterior end of the
Meckelian canal (Level I).
M. pterygoideus anterior (MPTA)
Origin (Level I): This muscle has a variable distribution within diapsids: extensively developed across the
dorsal surface of the palate, interorbital septum and
braincase in crocodilians; absent in most lizards;
present in birds and Sphenodon, in which it inserts
on the dorsal aspect of the palate). The anterior
extent of MPTA on the dorsal palate can be difficult to
distinguish from the smooth, excavated surfaces left
by the nasal passages/air sinuses (Witmer, 1997).
Internal structural constraints (including a thick
lacrimal-prefrontal ‘preorbital pillar’ and a narrow,
highly vaulted anterior palate) limit the origin of this
muscle in Heterodontosaurus to the dorsal surfaces of
the ectopterygoid and pterygoid. The posterior surface
of the pterygoid flange is depressed; this depression
continues anteriorly on to the dorsal surface of the
main body of the pterygoid and ectopterygoid. Thus,
the origin of MPTA appears to have been anteroposteriorly long and mediolaterally narrow. It should be
noted that this muscle may have had its functionality
subsumed within that of MPTP (Fig. 36).
Insertion (Level I): MPTA inserts onto the medial
surface of the articular and retroarticular process
(ventral to the jaw joint) of extant diapsids. Its inser-
229
saurus a shallow arch on the ventral surface of the
splenial-angular may indicate the general path of
MPTP around the lower jaw, and the retroarticular
process has a smooth excavation on its lateral surface
suggestive of a likely site for insertion of this muscle.
MPTP may have also inserted on the thickened
ventral margin of the angular.
M. depressor mandibulae (MDM)
of some of the principal adductor musculature. M. pterygoideus posterior (MPTP) showing its presumed origin on
the pterygoid and pterygoid flange. ?M. pterygoideus anterior (MPTA) shown as a slip associated with the MPTP
and inserting on the posteromedial surface of the prearticular. Deeper shading indicates the supposed shared
origin of the two muscle masses on the pterygoid.
Origin (Level II): In noncrocodilian diapsids this
MDM originates from the dorsal surface of the occiput
(squamosal, post-temporal bar, parietal margin, and
the dorsal, posterior edge of the quadrate and fascia
of the neck); in crocodilians, MDM originates from the
posterior margin of the squamosal and paroccipital
process. The massive and plate-like paroccipital ‘processes’, especially the thickened and rugose lateral
margins, offer potential areas for the origin of this
muscle in Heterodontosaurus (Fig. 34A).
tion in Heterodontosaurus is well supported by a
rough, somewhat excavated surface on the medial
surface of the prearticular beneath the jaw joint
(Fig. 36).
Insertion (Level I): MDM inserts onto the dorsal
surface of the retroarticular process in all extant
diapsids. In Heterodontosaurus, there is a clearly
defined excavation on the dorsal surface of the retroarticular process (posterior to the jaw joint) as
shown in Figure 34A, bounded by strong ridges, that
marks the insertion of MDM.
M. pterygoideus posterior (MPTP)
M. constrictor internus dorsalis group (MCID)
Origin (Level I): MPTP attaches to the posterior edge
of the pterygoid flange in crocodilians and birds, and
occasionally on the ventral surface of the pterygoid
in birds; it is believed to have originated from the
posterior and lateral margins of the pterygoid in
most dinosaurs. The pterygoid flange is well developed in Heterodontosaurus (Fig. 11) and its thickened and rugose posteromedial margin provides an
appropriate area for an aponeurotic origin for MPTP
(Figs 2, 36). The body of MPTP or a subdivision may
have originated in the pocket-like recess of the
pterygoid ventral to the shelf that supports the
pedicels of the basal articulation in the same
manner as described in hadrosaurs by Ostrom
(1961). Adjacent areas of the maxilla and pterygoquadrate wing have also been suggested as areas of
origins for this muscle in other ornithischians
(Thulborn, 1971a; Sues & Galton, 1987) but this is
unlikely in Heterodontosaurus.
Origin: This group of muscles includes the M. levator
pterygoideus (MLPT), M. protractor pterygoideus
(MPPT), and M. levator bulbi (MLB). These muscles
in diapsids lie between the neurocranium and palate,
and are thought to be involved in cranial kinesis
(Holliday & Witmer, 2008).
MPPT originates ventral to the trigeminal opening
and dorsal to the basipterygoid processes on the anterolateral surface of the basisphenoid; it inserts on the
pterygoid, close to its articulation with the quadrate.
MLPT originates on the laterosphenoid anterodorsal to the trigeminal fossa and inserts on the pterygoid dorsal to the basal articulation.
MLB originates on the parasphenoid and membranous anterior wall of the braincase and inserts on the
eyelid (Ostrom, 1961; Weishampel, 1984) – the latter
only rarely leaves osteological traces.
Osteological traces indicating the existence of
MPPT and MLPT have been identified in Hypsilophodon, Iguanodon, and hadrosaurs (Galton, 1974;
Weishampel, 1984; Norman, 1977, 1984b; Norman &
Weishampel, 1985); nevertheless Weishampel (1984)
in a broad-ranging review of ornithischian cranial
muscular anatomy was unable to identify origins for
either of these portions of MCID in Heterodontosaurus. A narrow ridge-depression exists above the
Insertion (Level I): MPTP wraps around the ventral
edge of the posterior end of the lower jaw to insert,
via tendons, on the lateral surface of the jaw and
retroarticular process (Figs 35B, 36) in most crocodilians, birds, and lizards. Such an insertion has also
been postulated for most dinosaurs. In Heterodonto-
230
D. B. NORMAN ET AL.
trigeminal fossa (SAM-PK-K1332 – Fig. 15, mcid)
and rugosities are present on the adjacent lateral
surface of the basisphenoid, immediately beneath
the trigeminal fossa. These latter features correlate
with the presence of remnants of MPPT and MLPT.
Insertions for such muscles (if they existed) cannot
be reliably identified on the pterygoid, although the
medial edge of the pterygoid of the holotype (Figs 2,
15, ptmr) is notably rugose, as if for connective
tissue attachment.
Functional interpretations of jaw action
in Heterodontosaurus
Jaw action in Heterodontosaurus has been the subject
of sporadic and inconclusive debate (Thulborn, 1974,
1978; Hopson, 1980; Weishampel, 1984; Crompton
& Attridge, 1986; Norman & Weishampel, 1991;
Barrett, 1998; Porro, 2007). An unusual suite of features (including an apparently akinetic cranium, a
potentially kinetic lower jaw, a heterodont mammallike dentition, a small keratinous beak, a reduced
anterior dentition including enlarged caniniforms,
and heavily worn, hypsodont cheek teeth exhibiting
oblique wear facets) as well as the small size of the
specimen, have impeded attempts at a generally
accepted explanation for jaw action.
Although the unusually mammal-like nature of
the dentition of Heterodontosaurus was first recognized by Crompton & Charig (1962), the first functional speculations were provided by Thulborn
(1971a, 1974, 1978) and focused on the ‘continuous’
occlusal surfaces of the cheek teeth. Thulborn proposed that such occlusal surfaces permitted (and
reflected) anteroposterior sliding (propaliny); furthermore, he suggested that this had consequences both
for the mode of tooth replacement and general
biology of heterodontosaurids. Hopson (1980), in
critically reviewing several aspects of Thulborn’s
work, provided evidence based on tooth structure
and wear (in both Heterodontosaurus and related
heterodontosaurids), that jaw action was essentially
orthal but also involved some lateral-to-medial
translation of the lower dentition against the upper
dentition during occlusion in a remarkably mammallike power-stroke. Weishampel (1984) offered a wideranging analysis of the functional and mechanical
implications of tooth structure and cranial form on
jaw action in ornithopod dinosaurs. In the case of
Heterodontosaurus, oblique tooth wear confirmed
Hopson’s description of a primarily orthal jaw action
with a mediolateral power stroke between the upper
and lower jaws. Yet because the skull was compact
and akinetic, and had a typically diapsid isognathic
jaw frame, translation of the lower jaw against the
cranium was impossible. This functional impasse
was overcome by invoking a jaw mechanism that
was unique to Heterodontosaurus: passive long-axis
rotation of the tooth-bearing dentaries against the
predentary during occlusion, made possible by a
novel form of symphysial flexibility: ‘. . . the
predentary-dentary joint is similar in form to a spheroidal joint.’ (Weishampel, 1984: 47).
Crompton & Attridge (1986) re-visited the problems associated with jaw action in Heterodontosaurus in some detail. They observed that the
premaxilla-predentary beak would have, in effect,
immovably clamped the rostrum when the jaw was
closed; they also claimed that the wear facets along
the dentition became transversely wider on the
posterior teeth (Crompton & Attridge, 1986:
fig. 17.11C). Having observed that in most herbivorous mammals the tip of the lower jaw describes an
elliptical path (in the transverse plane) during a
chewing cycle (to allow for medial translation of the
lower jaw during occlusion and subsequent repositioning ready for the next bite), it seemed clear that
heterodontosaurids were incapable of mammalianstyle jaw excursions because of the restrictions
placed on motion by the rostral beak and the (typically diapsid) isognathic jaw frame. The rigidity of
the skull roofing bones suggested that mediolateral
motion between teeth in opposing jaws must involve
displacement of the lower jaw [as Weishampel
(1984) had deduced]. Weishampel’s model of longaxis rotation of the lower jaws was rejected because:
(1) the wear facets on the teeth were stated to be
planar, and (2) the quadrate-articular jaw joint was
transversely expanded (rather than spheroidal). It
was accepted that the predentary-dentary joint was
of a ‘ball-and-socket’ type that allowed mobility and
on that basis it was proposed that the dentaries
inversely ‘wishboned’ (flexed medially) against the
predentary during occlusion. This motion would, in
their view, not have been restricted by the transverse expansion of the quadrate-articular joint, and
provided an explanation for their observation (and
illustration) of an increased width of wear facets
toward the posterior of the dental battery. Barrett
(1998) in reviewing herbivory in ornithischians
noted that in heterodontosaurids the theories that
invoked some form of mandibular rotation were contradicted by the evidence and favoured propalinal
jaw action.
What emerges from these attempts to reconcile
anatomy with putative models of jaw function is contradictory. For clarification the following observations
can be confirmed:
1. Jaw motion was almost entirely orthal (although
as Porro, 2009, has observed, microwear on the
worn surfaces of the teeth shows some smallscale palinal motion – partially supporting
2.
3.
4.
5.
6.
7.
8.
9.
the interpretations of Thulborn, 1971a and
Barrett, 1998); orthal motion is confirmed by the
shape of individual wear facets and the flush interface between the enamel and dentine on these
facets.
The structure of the rostral beak would have prevented lateral excursions of the anterior end of the
lower jaw during occlusion.
Lateral excursions of the lower jaw and long-axis
(medial) rotation of the dentaries during occlusion were prohibited by the tight fit between
the lower caniniforms and the medial walls of
the diastema (Barrett, 1998). Additionally, the
elongate pterygoid flange and ventral process
of the jugal formed a ‘slot’ that guided jaw
closure and prevented lower jaw translation and
rotation.
The predentary-dentary-dentary joints are not
spheroidal (‘ball-and-socket’) but complex, dorsoventrally elongate, and intimate (Barrett, 1998;
Butler et al., 2008a); symphyseal morphology
inhibited long-axis rotation; however, inverse
‘wishboning’ could have been possible (Porro,
2009).
The wear facets on the ‘batteries’, although
apparently ‘continuous’ show a steady change in
angulation and degree of wear (Porro, 2009) along
the length of the dentition such that the entire
surface is warped. The broadest part of the worn
surface (in occlusal view) is the middle section,
which corresponds to the position of the largest
teeth.
The cheek teeth, although they do not grow continuously throughout life, exhibit structural
adaptations in terms of the height and complexity of the crown that are consistent with hypsodonty (as more commonly seen in selenodont
mammals).
The extreme anterior and (less clearly) posterior
cheek teeth tend to have high-angle (steep)
planar wear facets.
The central maxillary teeth exhibit low-angle,
planar wear facets.
In contrast to the maxillary dentition, the central
dentary teeth exhibit low-angle and concave (and
lipped) wear facets.
The absence of significant rotational mobility at the
predentary-dentary suture, which underpinned the
most plausible explanations of jaw action provided by
Weishampel (1984) and Crompton & Attridge (1986),
prompted reinvestigation of the problem using a wide
range of biomechanical and engineering design techniques (Porro, 2009); these have produced a thorough
reassessment of the structural and functional
integration of the skull, lower jaw, and the teeth to
231
provide a rational explanation of the jaw action in
Heterodontosaurus (Porro, 2009; L. B. Porro, unpubl.
data).
General observations on tooth function
Caniniform teeth: The absence of significant levels of
wear and the lack of occlusal relationships between
upper and lower caniniforms is not consistent with
their use for cropping or rooting for vegetation. It is
also unlikely that they were used in intraspecific
displays because they are present in all known heterodontosaurid specimens (excepting Abrictosaurus),
including juveniles. Heterodontosaur caniniforms
resemble those of peccaries in that they are vertical
and the lower canines fit into a prominent diastema;
these canines act as ‘occlusal guides’ during chewing
and inhibit the lateral jaw excursions typically
observed during mammalian chewing cycles (Herring,
1972). Their vertical orientation also indicates that
such teeth are not adapted for delivering lateral
blows (as in various tragulids and cervids, e.g. Muntiacus). The general form of heterodontosaurid caniniforms also suggests that they were unlikely to have
been used for subduing large, active prey; indeed the
caniniforms most closely resemble those of animals
that use such teeth for capturing small prey (L. B.
Porro, unpubl. data).
In comparison with other early ornithischians, Heterodontosaurus has a relatively long forelimb and
large manus (Santa Luca, 1980: fig. 23); the manus is
adapted for grasping, rather than digging or weightbearing, judged by their elongate penultimate phalanges and trenchant claws (see discussion, Santa
Luca, 1980: 198). The manus more closely resembles
that of early theropods such as Eoraptor and Herrerasaurus. The combination of cursorial hindlimbs,
raptorial forelimbs, and puncturing caniniforms is
suggestive of an ability to catch and consume small
prey items. Barrett (2000) proposed basal sauropodomorphs, heterodontosaurids, and Lesothosaurus as potential omnivores (in the same way that
herbivorous iguanas are known to be opportunistic
predators, nest raiders, or carrion feeders). Given the
semi-arid and seasonal environmental setting
suggested for the Elliot Formation (Smith, 1990) facultative herbivory would represent a selective advantage during periods of time when vegetation was
scarce.
Dental batteries: Although most herbivorous dinosaurs used rapid tooth replacement to counter high
levels of abrasion, the frequency of tooth replacement was apparently reduced in Heterodontosaurus
in favour of the development of tooth-tooth occlusion, complex jaw movements, and increased tooth
durability. A general assumption (Norman &
232
D. B. NORMAN ET AL.
Weishampel, 1991; Norman et al., 2004c) has been
that smaller herbivorous dinosaurs would have been
selective browsers that took softer and more nutritious items, whereas larger taxa could feed on
tougher, less nutritious vegetation. Consideration of
Heterodontosaurus within the context of the
herbivores of the Early Jurassic of South Africa,
calls the wisdom of such generalizations into
question.
Within the Elliot Formation herbivores exhibit a
range of cranial and dental morphologies. The
small, lightly built Lesothosaurus has a comparatively lightly built skull, slender lower jaw, variably
worn, leaf-shaped teeth that were continually
replaced, and conventional orthal jaw action
(Crompton
&
Attridge,
1986;
Norman
&
Weishampel, 1991; Sereno, 1991a). Contemporary
prosauropods display teeth that are broadly similar,
although they show little evidence of wear; however,
abundant gastroliths often associated with their
skeletal remains suggest that these dinosaurs were
capable of breaking down tough vegetation (Crompton & Attridge, 1986; Norman & Weishampel, 1991;
Galton & Upchurch, 2004a). Heterodontosaurus
exhibits a range of cranial specializations: rigid
cranium, deep and robust lower jaws with pronounced ‘cheek’ recesses, close packing of the ‘cheek’
dentition to form a battery with an apparently contiguous set of obliquely angled wear facets that correlate with a transverse power stroke, asymmetric
enamel distribution on tooth crowns, complex
enamel ridging on individual teeth, and a
measure of hypsodonty that reflects both increased
rates of tooth wear and arrested rates of tooth
replacement.
These features correlate with a stronger bite, more
efficient grinding, and repetitive chewing of abrasive
foods (and correspondingly reduced chewing efficiency
for non-abrasive foods – Rensberger, 1973, 1975).
Smaller, cursorially adapted, basal ornithischians
show no evidence (to date) of accommodating gastroliths in their gut (although they are known in small
ceratopians – Norman & Weishampel, 1991); however,
amongst heterodontosaurids a suite of cranial and
dental specializations permitted the consumption of
tough vegetation and may well have reflected niche
partitioning amongst Early Jurassic herbivores. Later
ornithopods and ceratopians developed tooth batteries
and complex jaw movements to cope with tough
vegetation; however, higher bite forces (in absolute
terms) in these larger animals made it unnecessary
for them to evolve individually complex teeth or
precise occlusion because they relied on constant
tooth replacement and a structural and functional
integration of tooth crowns and roots to maintain
their dental batteries.
THE TAXONOMY OF SOUTHERN AFRICAN
HETERODONTOSAURIDS
Five monospecific genera have been named from the
upper Elliot and Clarens Formations of southern
Africa: Geranosaurus, Heterodontosaurus, Abrictosaurus, Lycorhinus, and Lanasaurus. However, the taxonomy is poorly resolved (Thulborn, 1970b, 1974,
1978; Charig & Crompton, 1974; Gow, 1975, 1990;
Hopson, 1975, 1980; Weishampel & Witmer, 1990;
Norman et al., 2004c; Butler et al., 2008a). For
example, Geranosaurus is considered a nomen
dubium (Weishampel & Witmer, 1990; Norman et al.,
2004c) and Lanasaurus has been considered to
be a junior subjective synonym of Lycorhinus (Gow,
1990; Weishampel & Witmer, 1990; Norman et al.,
2004c).
Three valid taxa (H. tucki, A. consors, and
Ly.
angustidens)
are
currently
recognized
(Weishampel & Witmer, 1990; Norman et al., 2004c;
Butler et al., 2008b) although the assignment of individual specimens to each taxon remains confused. For
example, NHMUK RU A100 was described by Thulborn (1970b) as an individual of Ly. angustidens;
Charig & Crompton (1974) criticized this referral and
considered NHMUK RU A100 to represent a new
heterodontosaurid genus, whereas Hopson (1975) provisionally referred NHMUK RU A100 to Abrictosaurus. Hopson’s assignment was followed by the
majority of subsequent authors (Weishampel &
Witmer, 1990; Norman et al., 2004c); however, Gow
(1990) reverted to the assignment originally advocated by Thulborn (1970b). This review of southern
African heterodontosaurid taxonomy is provisional –
it should be noted that for more adequate assessment,
further preparation of several specimens and the collection of more material is required.
GERANOSAURUS
ATAVUS
BROOM, 1911
Diagnosis: The type and only specimen of Geranosaurus atavus is poorly preserved and fragmentary, and
is not diagnosable on the basis of unique characters,
or on the basis of a unique combination of characters,
and is regarded as a nomen dubium; the holotype and
referred specimens are considered Heterodontosauridae indet.
Holotype: SAM-PK-1871 (Fig. 37A) a heavily eroded
skull currently preserved in two parts, the first
embedded in a block of sandstone includes a predentary and rostral ends of both dentaries (Broom, 1911:
plate 17, fig. 24), whereas the second consists of a
partial maxillary tooth row with crowns heavily
damaged.
233
Referred specimen: SAM-PK-1857, a poorly preserved
partial hindlimb, referred to Geranosaurus by Broom.
Broom (1911: 306) also mentioned ‘a number of very
imperfect vertebrae’. The whereabouts of all these
specimens is unknown.
Horizon and locality: Holotype and referred specimens were collected from the Clarens Formation
(‘Cave Sandstone’ Early Jurassic), road-cutting near
the summit of Barkly Pass (31°27′S, 27°51′E; Kitching & Raath, 1984: table 1), Farm Tulloch, Elliot,
Eastern Cape Province, South Africa.
Figure 37. A, Geranosaurus atavus Broom, 1911. SAMPK-1871 (holotype). Upper image: dorsal view of the fragmentary (and subsequently damaged) remains of the
anterior part of the dentaries and predentary; lower
image, the extremely fragmentary right maxilla (with
traces of the premaxilla), showing the broken crowns and
roots of cheek teeth of a heterodontosaurid from Eastern
Cape Province. Matrix on dentary block signified using
irregular tone. B, Lycorhinus angustidens Haughton, 1924.
SAM-PK-3606 (holotype). Partial left dentary, based on the
original illustration by Haughton (re-drawn by Hopson,
1975: fig. 1a). The specimen can no longer be found (apart
from the dentary crown) and is now represented only by a
moulded impression of the original. C, L. angustidens
Haughton, 1924. Lateral view of the left dentary dentition
(from Hopson, 1975: fig. 1B) retrieved from an impression
of the dentition left on the sandstone matrix preserved
with the original specimen.
Discussion: The holotype specimen of Geranosaurus is
poorly preserved (Fig. 37A). The crowns of all of the
dentary teeth are missing and the crowns of the
maxillary teeth are broken immediately above their
bases. The dentaries are highly fractured. The specimen has evidently been damaged since its original
description by Broom (1911): the enlarged dentary
caniniform is now missing and the maxillary teeth
also appear to have been damaged because Broom
(1911: 307) described them as having ‘. . . flat, chiselshaped crowns with the outer face feebly ridged’;
these features can no longer be confirmed.
The presence of a predentary that is wedge-like
(lacking a well-defined median ventral process) and a
dentary caniniform, and the apparent absence of
replacement foramina, confirm the heterodontosaurid
affinities first recognized by Crompton & Charig
(1962). Crompton & Charig suggested that Geranosaurus differed from H. tucki in lacking a large
diastema between the premaxilla and maxilla and the
absence of an inset maxillary tooth row. Hopson
(1980) suggested that the maxillary tooth row was
inset, but poor preservation means the latter character is difficult to establish with any confidence. Geranosaurus is distinct from Abrictosaurus in possessing
an enlarged dentary caniniform. However, neither
autapomorphies nor a unique combination of characters must be recognized and so, taxonomically, Geranosaurus atavus can be considered to be a nomen
dubium (Thulborn, 1974, 1978; Hopson, 1980;
Weishampel & Witmer, 1990; Norman et al., 2004c).
LYCORHINUS
ANGUSTIDENS
HAUGHTON, 1924
Revised diagnosis: Lycorhinus angustidens is retained
provisionally as a valid taxon on the basis of the
following character combination: dentary caniniform
present; diastema between caniniform and first postcaniniform tooth is short, equal in length to the
anteroposterior length of the first postcaniniform
crown; postcaniniform crowns are closely packed and
adjacent crowns contact one another (with some
overlap); lateral and medial surfaces of the postca-
234
D. B. NORMAN ET AL.
niniform crowns possess a distinct basal swelling
(‘cingulum’) and are therefore expanded mediolaterally above the root; postcaniniform crowns are
expanded anteroposteriorly above the root and therefore possess a ‘neck’.
Holotype: SAM-PK-3606, an incomplete left dentary
with caniniform and 11 postcaniniform crowns.
(Fig. 37B, C; Haughton, 1924: fig. 8; Broom, 1932:
fig. 104i; Charig & Crompton, 1974: figs 8, 9; Hopson,
1975: figs 1, 2, 3e; Hopson, 1980: fig. 3; Galton, 1986:
fig. 16.6q; Gow, 1990: figs 2, 3, 7; Weishampel &
Witmer, 1990: fig. 23.2c; Smith, 1997: fig. 3e; Norman
et al., 2004c: fig. 18.2c).
Holotype horizon and locality: Upper Elliot Formation
at Paballong (30°26′S, 28°31′E; Kitching & Raath,
1984: table 1), near Mount Fletcher, Eastern Cape
Province, South Africa. Broom (1932) mistakenly suggested that the holotype specimen was collected from
Witskop near Burgersdorp.
Discussion: Haughton (1924) named Ly. angustidens
on the basis of a partial left dentary exposed in lateral
view (SAM-PK-3606), identifying it as a cynodont.
Broom (1932) considered Lycorhinus a possible ictidosaur (Trithelodontidae), but also noted similarities
between the postcaniniform dentition and that of
dinosaurs. By the time of Broom’s publication the
specimen had already been damaged, and most of the
specimen (except for part of the crown of the caniniform tooth) was subsequently lost. Crompton &
Charig (1962) first noted the heterodontosaurid affinities of SAM-PK-3606 on the basis of similarities to
the dentition of H. tucki.
The incomplete preservation of SAM-PK-3606
(Fig. 37B, C) makes assessing the validity of Lycorhinus difficult, especially given the brief description
provided by Haughton (1924). More recent descriptions of the dentary and dentition have been based
upon silicone moulds of the original specimen and
casts of the remaining impressions (Hopson, 1975,
1980; Gow, 1990). Although recognizing autapomorphies is difficult given the nature of the holotype
material, Lycorhinus should be retained provisionally
as a distinct taxonomic entity on the basis of a unique
character combination, absent in other heterodontosaurids (Hopson, 1975).
Comparison with A. consors (Fig. 39A, B – discussed below) is difficult, because in Lycorhinus the
dentary teeth have been best described in medial
view, whereas the holotype of Abrictosaurus
(NHMUK RU B54) only exposes the dentition laterally. One striking difference is the absence of a
dentary caniniform in Abrictosaurus (cf. Thulborn,
1974 – as Lycorhinus consors), which is likely to be a
feature of genuine taxonomic significance (as opposed
to ontogenetic or sexually dimorphic significance, as
originally suggested, see below).
Heterodontosaurus tucki differs from Lycorhinus in
that the dentary cheek teeth do not substantially
overlap one another (although minor imbrication is
present – see Figs 25, 27); have crowns that are not
expanded labiolingually or mesiodistally above the
root (i.e. a distinct ‘cingulum’ and a distinct ‘neck’
are both absent); have a nearly continuous wear
surface developed across all teeth; and have a primary
ridge on the medial surface that is offset anteriorly
(Hopson, 1975) with secondary ridges commonly developed on the medial surface posterior to the primary
ridge. Comparisons of Lycorhinus with La. scalpridens
(Fig. 39C) are not possible because the holotypes are
lower and upper jaws, respectively.
NHMUK RU A100
This specimen is represented by an incomplete and
disarticulated skull (Fig. 38), and was collected from
the same locality as SAM-PK-3606 (the holotype
of Ly. angustidens). It was referred to the genus
Lycorhinus by Thulborn (1970b) who suggested that
NHMUK RU A100 and SAM-PK-3606 might represent different parts of a single individual; this interpretation is not credible: Thulborn (1970b: 236)
acknowledged the substantial hiatus (~40 years)
between the collection of SAM-PK-3606 and the collection of NHMUK RU A100 in 1960–1961 and this
was further reinforced by the observations of Charig
& Crompton (1974: 176) on the nature of the Red Bed
exposure in this area and the annual rates of erosion.
Furthermore, Thulborn (1970b: fig. 5) misidentified
the dentary of NHMUK RU A100 as the right
dentary; it in fact appears to be the left dentary
exposed in medial view (R. J. B., pers. observ.), the
same element as SAM-PK-3606. Charig & Crompton
(1974) discussed the possibility that NHMUK RU
A100 might represent more than a single individual;
however, we agree with their conclusion that this is
unlikely given that the specimen contains no duplicate elements.
A number of characters of NHMUK RU A100 has
been reported to differ from SAM-PK-3606 (Charig
& Crompton, 1974: 179; Hopson, 1975), and on this
basis it has been suggested that NHMUK RU A100
may represent either a distinct but unnamed genus
(Charig & Crompton, 1974) or a second individual
of A. consors (Hopson, 1975, 1980; Weishampel
& Witmer, 1990; Norman et al., 2004c). Butler
et al. (2008b) treated NHMUK RU A100 as a
distinct taxon for phylogenetic analysis because of
this uncertainty. The characters by which NHMUK
RU A100 is suggested to differ from SAM-PK-3606
are:
235
Figure 38. NHMUK RU A100 (BMNH A100). Unnamed heterodontosaur remains based on a partial skull that was
originally referred to the genus Lycorhinus by Thulborn (1970b) – composite image manipulated to create the effect of a
partial anterior skull viewed from left side. A, left premaxilla in medial view (reversed). B, left maxilla in lateral view.
C, maxillary dentition in medial view (reversed). D, left dentary in medial view (reversed). Illustrations derived from
Charig & Crompton (1974: figs 4–7).
1. The dentary caniniform of NHMUK RU A100
serrated on its anterior margin only (Thulborn,
1970b), whereas both mesial and distal margins
are serrated in SAM-PK-3606 (Hopson, 1975).
2. The base of the dentary caniniform is more
slender in SAM-PK-3606 than in NHMUK RU
A100 (Charig & Crompton, 1974).
3. The dentary is deeper beneath the caniniform in
SAM-PK-3606 when compared to NHMUK RU
4. The length of the diastema between the caniniform and the first postcaniniform dentary tooth
is greater in NHMUK RU A100 than in SAMPK-3606 (Charig & Crompton, 1974).
5. The tip of the caniniform is bevelled anteromedially in SAM-PK-3606 but not in NHMUK RU
6. Postcaniniform dentary crowns are inclined
slightly mesially [anteriorly] in SAM-PK-3606
but not in NHMUK RU A100 (Charig & Crompton, 1974).
7. Dentary crowns are closely packed with the
cusps of adjacent crowns contacting one another
in SAM-PK-3606 but not in NHMUK RU A100
(Charig & Crompton, 1974).
8. Dentary crowns are heavily worn in SAM-PK3606 but not in NHMUK RU A100 (Charig &
Crompton, 1974).
9. Dentary crowns asymmetrical in labial view in
SAM-PK-3606 but symmetrical in NHMUK RU
10. Bases of the dentary crowns are notably more
strongly swollen in SAM-PK-3606 than in
NHMUK RU A100 (Hopson, 1975).
Character 1 cannot be assessed for NHMUK RU
A100 (contra Thulborn, 1970b) because the posterior
margin of the caniniform is largely covered by sediment. The exposed apicothecal height of the caniniform of NHMUK RU A100 is approximately 3.1 times
the mesiodistal length of its base (Thulborn, 1970b);
the corresponding ratio as preserved is ~ 2.8 in
SAM-PK-3606 (Thulborn, 1970b) but the tip of the
caniniform is missing. There is therefore little difference in Character 2 between NHMUK RU A100 and
SAM-PK-3606. Adequate assessment of Character 3
is precluded by the incomplete exposure of the
dentary of NHMUK RU A100 and the incomplete
ventral margin of the dentary of SAM-PK-3606
(Hopson, 1975: fig. 1a). Character 4 does represent a
236
D. B. NORMAN ET AL.
valid difference: in NHMUK RU A100 the postcaniniform diastema is equivalent in mesiodistal length to
the mesiodistal length of the first two postcaniniform
crowns, and is ~70% of the length of the base of the
caniniform. By contrast, the diastema is significantly
shorter in SAM-PK-3606, being equivalent only to the
mesiodistal length of the first postcaniniform crown,
and ~25% of the length of the base of the caniniform.
Character 5 cannot be adequately assessed for
NHMUK RU A100 because the dentary is only visible
in medial view (contra Thulborn, 1970b) and the tip of
the caniniform is, in any case, inadequately exposed.
Character 6 does not appear to be valid: a slight
anterior inclination of several of the postcaniniform
crowns (notably crowns 4 and 5) is evident in
NHMUK RU A100, although not in the drawings of
NHMUK RU A100 provided by Thulborn (1970b:
fig. 5). Character 7 is a valid distinction: the exposed
postcaniniform crowns of NHMUK RU A100 are comparatively broadly spaced and the margins of adjacent crowns do not contact or overlap one another,
unlike the tightly packed dentition of SAM-PK-3606.
Character 8 is difficult to assess adequately for
NHMUK RU A100 because so few of the dentary
crowns are adequately exposed, and those that are
exposed are exposed in medial view (wear facets on
dentary crowns occur on the lateral surfaces).
However, the preservation of denticles along the
anterior and posterior margins of the anterior
dentary crowns of NHMUK RU A100 does suggest
that the crowns have received little wear – this
may therefore represent a valid difference from
SAM-PK-3606 in which the crowns are heavily worn
and cusps are rarely preserved (see Fig. 37C). Characters 9 and 10 cannot be assessed adequately
because the dentary teeth of NHMUK RU A100 are
exposed only in medial view, whereas those of SAMPK-3606 are best documented in lateral view
(Hopson, 1975; Gow, 1990). However, it is clear that in
medial view, the dentary teeth of NHMUK RU A100
are not symmetrical: the primary ridge is offset
towards the anterior margin, so Character 9 is
invalid. Although the anterior dentary crowns of
NHMUK RU A100 are well differentiated from the
root and expanded anteroposteriorly above, this is
also true of the anterior dentary crowns of Lycorhinus
(Hopson, 1975: fig. 1B), so Character 10 seems to be
invalid.
Three valid distinctions between NHMUK RU A100
and SAM-PK-3606 remain: the difference in the relative length of the diastema (Character 4); the difference in the packing of the postcaniniform teeth
(Character 7); and differences in the degree of dental
wear (Character 8). The last character is inadmissible
because it is a developmental and functional variable
between individuals. NHMUK RU A100 and SAM-
PK-3606 are commensurate so it seems unlikely that
the first two of these distinctions can be ascribed to
ontogenetic stage differences. The degree of intraspecific variation in the length of the diastema in heterodontosaurids is unknown: in Heterodontosaurus
the diastema is only adequately preserved in the
referred specimen SAM-PK-K1332. Heterodontosaurus shows little ontogenetic or individual variation in
the packing of cheek teeth.
It is impossible to evaluate the significance of the
anatomical differences between NHMUK RU A100
and SAM-PK-3606 because the material is too imperfect to establish whether they represent individual
variations, species-specific features or components of
higher-level taxonomic separation. In view of these
differences, and in view of the inadequate nature of
the holotype specimen of Ly. angustidens (SAM-PK3606), we feel that referral of NHMUK RU A100 to
Ly. angustidens is not justified at this stage.
Note. NHMUK RU A100 will be referred tentatively
to the taxon La. scalpridens Gow, 1975 (see below).
ABRICTOSAURUS
= LYCORHINUS
CONSORS
CONSORS
(THULBORN, 1974)
THULBORN, 1974
Revised diagnosis: Heterodontosaurid that lacks
enlarged caniniform premaxillary and dentary teeth;
anterior end of dentary ramus dorsoventrally deeper
than mid-section; 12 maxillary and 14 dentary
crowns; crowns are generally leaf-shaped in profile,
but show little development of a median (primary)
ridge and show only weak development of anterior
and posterior ridges; prominent laterally projecting
maxillary shelf is absent, and there is, consequently,
no well-marked cheek recess;
Holotype: NHMUK RU B54 (formerly UCL B54),
partial skull and incomplete postcranial skeleton
(Fig. 39A, B; Thulborn, 1974: figs 2–4, Hopson, 1975:
fig. 3c, d, Galton, 1986: fig. 16.6m, Smith, 1997:
fig. 3c, d).
Holotype horizon and locality: Upper Elliot Formation, from a streamside exposure at the village of
Noosi (30°03′S, 28°32′E; Kitching & Raath, 1984:
table 1), 5.1 miles east of Whitehill, southern Lesotho.
Discussion: Thulborn (1974) described a new species
of heterodontosaurid based upon a single specimen
(NHMUK RU B54 – Fig. 39A, B) and referred this
new species to the genus Lycorhinus as Lycorhinus
consors. He distinguished it from previously named
heterodontosaurids on the basis of several craniodental characters, notably the absence of caniniform
teeth. The postcranial material of NHMUK RU B54
was not described. There are a number of errors in
Figure 39. A, Abrictosaurus consors (Thulborn, 1974).
NHMUK RU B54 (holotype). A partial skull (illustrated in
abstract form – and based on the reconstruction in Thulborn, 1978). Shading represents known elements of the
skull and dentition. The specimen includes postcranial
elements that (along with the skull) are in need of further
preparation and re-description. B, Abrictosaurus consors
(Thulborn, 1974). NHMUK RU B54 (holotype). The maxillary and dentary dentitions, as drawn by Thulborn (1974:
fig. 3). C, Lanasaurus scalpridens Gow, 1975. BP/1/4244
(holotype). The maxillary dentition in medial aspect
[reversed for comparison with (B)] showing the almost
complete dentition (crown 12 obliquely truncated) and
with the individually angled wear facets accentuated by
cross-hatching (redrawn from Hopson, 1980: fig. 2).
Thulborn’s description, some of which are highlighted
below, and a full redescription is required, although
this requires further preparation of the specimen.
Thulborn used the new species name as a means of
highlighting his suspicion that the lack of caniniform
teeth indicated that NHMUK RU B54 represented a
female individual.
237
Hopson (1975) provided a detailed account of the
anatomical differences between the holotypes of Ly.
angustidens and Ly. consors, and erected the new
genus Abrictosaurus for the latter species. Hopson
also provisionally referred NHMUK RU A100,
described by Thulborn (1970b), to A. consors. With the
exception of Thulborn (1978), later authors have
accepted the generic distinctiveness of A. consors.
Most subsequent authors, with the exceptions of
Thulborn (1978), Gow (1990), and Butler et al.
(2008a), have uncritically accepted the referral of
NHMUK RU A100 to Abrictosaurus (Galton, 1986;
Weishampel & Witmer, 1990; Smith, 1997; Norman
et al., 2004c). Weishampel & Witmer (1990) and
Norman et al. (2004c) regarded the type specimen
(NHMUK RU B54) of Abrictosaurus as juvenile
and/or female and NHMUK RU A100 as adult and/or
male, and thus reaffirmed assumptions about the
probability of sexual dimorphism as a factor to be
considered in the taxonomy of some heterodontosaurs. Norman (1985) suggested that the holotype
specimen of Abrictosaurus might represent a female
individual of Heterodontosaurus.
The ontogenetic stage of NHMUK RU B54 is
undetermined: the neurocentral sutures of the vertebrae are not sufficiently well exposed to assess the
presence/absence of fusion and histological analysis
has not yet been carried out. The estimated anteroposterior length of the orbit in NHMUK RU B54 is
estimated at 24 mm, 73% of the estimated preorbital
length (33 mm) and 34% of the estimated total skull
length (70 mm). In SAM-PK-K1332 (the referred
skull of H. tucki) the corresponding ratios are 73
and 32%, whereas in SAM-PK-K337 (the holotype of
H. tucki) the corresponding ratios are 69 and 31%.
In contrast, the immature skull of H. tucki (SAMPK-K10487, Fig. 28A) the orbit is approximately
90% of preorbital length. The proportions of the
skull of NHMUK RU B54 are therefore more similar
to those of presumed adult individuals of H. tucki
than juveniles.
Although there have been several unpublished suggestions that NHMUK RU B54 contains elements
from more than a single individual (M. Evans, A.
Yates, unpublished notes associated with NHMUK
RU B54) this observation cannot be confirmed. This
suggestion results from a block containing two
scapulae from the syntype specimen (NHMUK RU
B17 – Thulborn, 1972: fig. 2) of Lesothosaurus diagnosticus having been mistakenly catalogued under
the specimen number NHMUK RU B54 (A. Yates,
pers. comm. 2007, R. J. B., pers. observ.). However
there is no clear evidence for duplication of elements
in the remaining material and appears to represent
a single individual, pending preparation and
redescription.
238
D. B. NORMAN ET AL.
Although seemingly entrenched in the literature,
the referral of NHMUK RU A100 to Abrictosaurus
cannot be supported. An enlarged and serrated premaxillary caniniform tooth is present in NHMUK
RU A100, the apicobasal height of which significantly exceeds the height of both the maxillary dentition and the preceding premaxillary dentition. By
contrast,
the two preserved premaxillary incisiform teeth of
NHMUK RU B54 [note that, contra Thulborn (1974),
a possible cross-section of the root of a third premaxillary crown appears to be present just anterior
to those described and illustrated in NHMUK RU
B54 – see Fig. 39A] are subequal in size and do not
exceed the apicobasal height of the maxillary dentition. NHMUK RU B54 lacks the prominent maxillary ridge that, in NHMUK RU A100, forms the
ventral margin of the external antorbital fenestra
and the dorsal margin of a well-developed cheek
recess. In contrast, the tooth row of NHMUK RU
B54 is positioned marginally so there is no wellmarked cheek recess. The maxilla is proportionally
deeper below the antorbital fossa in NHMUK RU
B54 and numerous foramina are present on its
lateral surface; such foramina have not been recognized in an examination of NHMUK RU A100 (R. J.
B., pers. observ.). The maxillary teeth of NHMUK
RU A100 possess a well-developed swelling (‘cingulum’) on their lateral surfaces, an inflated, apicobasally extending, median swelling (in an equivalent
position to, but considerably less well developed
than, the primary ridge of Heterodontosaurus); and a
well-developed ridge along the posterior margin of
the crown that is substantially better developed than
the equivalent ridge on the anterior margin. By contrast, the maxillary teeth of Abrictosaurus (NHMUK
RU B54) lack both a basal swelling (cingulum) and a
median swelling or ridge; they are also less strongly
expanded anteroposteriorly above the root and more
closely packed than in NHMUK RU A100. Although
anterior and posterior ridges are evident on at least
some of the crowns of Abrictosaurus, these ridges are
very subtle and weak, and the posterior ridge is not
better developed than the anterior ridge. The differences between the maxillary dentitions of NHMUK
RU A100 and NHMUK RU B54 are striking. This
observation contrasts with that of Hopson (1975:304)
who stated (based largely on comparisons of the
dentary teeth: see his fig. 3) that the dentition of
NHMUK RU A100 was ‘not distinct enough from
that of the type of ‘L.’ consors [NHMUK RU B54] to
merit its specific separation’.
Differences between NHMUK RU A100 and
NHMUK RU B54 are also evident in the mandible. The
anterior end of the dentary of NHMUK RU B54 is
dorsoventrally expanded relative to more caudal parts
of the element (see Fig. 39A); in contrast, the anterior
end of the dentary tapers in dorsoventral height in
NHMUK RU A100. Perhaps more significantly, there is
an enlarged and serrated caniniform tooth at the
anterior end of the tooth row in NHMUK RU A100
which is, of course, absent in NHMUK RU B54.
Further comparison between the dentary teeth
of NHMUK RU A100 and NHMUK RU B54 is
limited because the teeth of NHMUK RU B54 are
exposed in lateral aspect only, whereas those of
NHMUK RU A100 are exposed in medial view (contra
Thulborn, 1970b).
Characters that distinguish Abrictosaurus from
Lycorhinus and Heterodontosaurus are discussed
above. Abrictosaurus can further be distinguished
from the holotype specimen of Lanasaurus (BP/1/4244
– Fig. 39C, discussion follows) by the same characters
of the maxillary dentition that distinguish Abrictosaurus from NHMUK RU A100.
Suggestions that the absence of caniniform teeth
might indicate that Abrictosaurus represents an
early ontogenetic stage or gender differences within
another Stormberg taxon are unlikely for several
reasons:
1. Differences between Abrictosaurus and other
Stormberg taxa are numerous and substantial and
are not simply limited to the presence/absence of
caniniforms;
2. It has not been demonstrated that NHMUK RU
B54 is a juvenile individual;
3. Evidence from a juvenile specimen of H. tucki
(SAM-PK-K10487, Figs 28, 29) indicates that the
dentary and premaxillary caniniforms were
present at a comparatively early developmental
stage in this taxon and are unlikely to have been
late ontogenetic or secondary sexual characters;
4. Eleven Stormberg heterodontosaurid specimens
of which we are aware have the anterior end of
the dentary preserved. Ten of these (SAM-PK1871, 3606, K337, K1332, K10487, K10488;
NHMUK RU A100, RU C68, RU C69; NMQR
1788) have evidence of an enlarged caniniform;
only in NHMUK RU B54 is the caniniform
missing. Although larger sample sizes would be
informative, a Chi-squared test of goodness-of-fit
indicates that the observed ratio differs significantly from the predicted 50 : 50 ratio expected if
caniniforms are sexually dimorphic (c2 = 5.82;
d.f. = 1; P = 0.016).
On this (admittedly limited) evidence it is considered that A. consors represents a valid species and we
reject the possibility that NHMUK RU B54 is either
a juvenile or female individual of one of the currently
recognized species. No additional specimens of Abrictosaurus have yet been recognized.
LANASAURUS
SCALPRIDENS
GOW, 1975
Revised diagnosis: Maxillary tooth row that is
strongly bowed inwards along its length; maxillary
teeth in labial view possess a posterior ridge that is
significantly more strongly developed than is the
anterior ridge.
Holotype: BP/1/4244, left maxilla [Fig. 39C; Gow,
1975: figs 1, 2, pl. 1, Hopson, 1980: fig. 2, Galton,
1986: fig. 16.6o–p, Gow, 1990: fig. 6, Weishampel &
Witmer, 1990: fig. 23.2a (as Ly. angustidens), Norman
et al., 2004c: fig. 18.2a (as Ly. angustidens)].
Holotype horizon and locality: Upper Elliot Formation, ‘Buck Camp’, Golden Gate Highlands National
Park (28°30′S, 28°37′E; Kitching & Raath, 1984:
table 1), Free State Province, South Africa.
Referred specimens: BP/1/5253, partial left maxilla,
unspecified stratigraphical level, Bamboeskloof
Farm (30°45′S, 27°12′E; Gow, 1990), Lady Grey,
Eastern Cape Province, South Africa (Gow, 1990:
figs 1, 3–5).
NHMUK RU A100 (formerly UCL A100), partial
skull, upper Elliot Formation at Paballong (30°26′S,
28°31′E; Kitching & Raath, 1984: table 1), near
Mount Fletcher, Herschel district, Eastern Cape Province, South Africa [Thulborn, 1970b: figs 1–5; Charig
& Crompton, 1974: figs 4–7; Hopson, 1975: fig. 3a–b;
Galton, 1986: fig 16.6n (as Abrictosaurus consors);
Weishampel & Witmer, 1990: fig 23.2b (as A. consors);
Smith, 1997: fig 3a, b (as A. consors); Norman et al.,
2004c: fig 18.2b (as A. consors)].
Discussion: Gow (1975) named La. scalpridens on the
basis of the holotype maxilla, diagnosing it on the
basis of its dental morphology, the presence of a
maxillary tooth row that bows medially (toward the
midline) along its length, and the absence of the
accessory openings within the antorbital fossa identified in Heterodontosaurus. Gow (1975) did, however,
speculate that La. scalpridens might be represent the
same taxon as A. consors. Thulborn (1978) considered
La. scalpridens to be highly atypical for heterodontosaurids, and even questioned its heterodontosaurid
affinities. Hopson (1980:94) first suggested that
Lanasaurus might be synonymous with Ly. angustidens and was supported by Gow (1990), Weishampel
& Witmer (1990), Norman et al. (2004c), and Butler
et al. (2008b).
Gow (1990) did not evaluate the characters that
supported his tentative reference to Abrictosaurus; it
is however evident that they included general similarities in dental wear patterns. The assignment of
Lanasaurus to Lycorhinus is not compelling and
239
cannot be proven using any currently available specimens. One obvious problem with any comparison
between Lanasaurus and Lycorhinus is their nonoverlapping holotypes, respectively: BP/1/4244 is a
maxilla and SAM-PK-3606 is a dentary.
It is considered preferable to provisionally retain
La. scalpridens as a distinct and diagnosable taxon.
Two potential autapomorphies of La. scalpridens are
here recognized:
1. The presence of a maxillary tooth row that is
strongly bowed inwards along its length (Gow,
1975); and
2. Maxillary teeth in lateral view possess a posterior
ridge that is significantly more strongly developed
than the anterior ridge.
Although Gow (1990) reported an inwardly arched
tooth row in the holotype of Ly. angustidens (SAMPK-3606) it has not been possible to confirm this
[based upon examination of a latex cast of SAM-PK3606 (NHMUK R8180)]. Characters that distinguish
Lanasaurus from Heterodontosaurus and Abrictosaurus have been discussed above.
BP/1/5253
A second specimen, the partial left maxilla (BP/1/
5253) described by Gow (1990) and assigned by him to
Ly. angustidens is also referred to La. scalpridens. As
discussed above for the holotype of La. scalpridens,
BP/1/5253 cannot be referred to Ly. angustidens with
any confidence because it does not overlap anatomically with the holotype specimen of the latter taxon.
However, BP/1/5253 does show close resemblances to
BP/1/4244, the holotype of Lanasaurus, as discussed
by Gow (1990). Both maxillae share the possible autapomorphies of having a tooth row that is strongly
bowed inwards along its length and teeth with posterior ridges on their lateral surfaces that are more
pronounced than the anterior ridges. These maxillae
share other similarities, such as the anterodorsal
margin of the antorbital fossa not being sharply
defined, as occurs in Heterodontosaurus and Abrictosaurus. Few meaningful differences occur between
BP/1/5253 and BP/1/4244: the tooth row of the former
is marginally more strongly curved inwards along its
length and foramina cannot be recognized within the
buccal emargination; both of these differences may
reflect post-mortem effects rather than genuine
anatomy. In light of the presence of shared potential
autapomorphies, and only minor differences, referral
of BP/1/5253 to Lanasaurus appears justified.
NHMUK RU A100
This controversial specimen is referred, provisionally,
to La. scalpridens. The specimen has been referred to
both Ly. angustidens (Thulborn, 1970b, 1974, 1978;
240
D. B. NORMAN ET AL.
Gow, 1990) and A. consors (Hopson, 1975;
Weishampel & Witmer, 1990); however, neither can be
supported (discussion above). Several right maxillary
teeth of NHMUK RU A100 are well exposed in lateral
view (Fig. 38 – and Thulborn, 1970b: fig. 3). These
have a similar general morphology and degree of
packing to Lanasaurus and they share, with Lanasaurus, the presence of posterior ridges on the lateral
surfaces of the crowns that are more pronounced than
the anterior ridges. As preserved, the right tooth row
of NHMUK RU A100 is bowed inwards along its
length, and curves outwards markedly at its anterior
end. The curvature of the tooth row, particularly at
the caudal end, is not as pronounced as in BP/1/4244,
but this may be a post-mortem artefact.
In view of the fact that NHMUK RU A100 possesses one autapomorphy of La. scalpridens, it is
referred tentatively to this latter taxon. If this proves
correct (by further preparation and study of NHMUK
RU A100) it may demonstrate that Lanasaurus and
Lycorhinus are distinct taxa, because the dentary of
NHMUK RU A100 differs in several respects from
that of Ly. angustidens (discussion above).
PHYLOGENETIC RELATIONSHIPS
Unequivocal ornithischian affinities of heterodontosaurids (based primarily upon the anatomy of Heterodontosaurus) are not in dispute. They include the
following anatomical characters:
1. An edentulous predentary bone at the dentary
symphysis.
2. A toothless, rugose anterior tip to the premaxilla.
3. A palpebral bone that articulates at the anterior
orbital margin between the lacrimal and prefrontal and forms a tapering, curved rod that projects
obliquely across the dorsal part of the orbital
cavity.
4. Medially inset maxillary and dentary dentitions.
5. Six sacral vertebrae.
6. Ossified tendons preserved lateral to the neural
spines of the dorsal and sacral vertebrae.
7. A prominent, elongate, and curved preacetabular
process on the ilium.
8. The principal ramus of the pubis is long, narrow,
and rod-like, and lies parallel to the ischium (the
opisthopubic condition); it also terminates at the
distal end of the ischial shaft; and there is a short,
blunt anterior pubic process.
SYSTEMATIC
POSITION OF THE HETERODONTOSAURIDS
A restricted phylogenetic analysis of Heterodontosaurus, derived from the data compiled by Butler et al.
(2008b) is presented below as a simplified framework
for the more detailed discussion of the issues that
surround consideration of ornithischian phylogenetics
and evolution. The taxa or operational taxonomic
units (OTUs) included in this analysis are a combination of supraspecific taxa and terminal taxa. As
described by Butler et al. (2008b) the supraspecific
taxa chosen are widely recognized as monophyletic;
they are coded as single OTUs and comprise:
Supraspecific taxa
• Ankylopollexia
• Pachycephalosauridae
• Psittacosauridae
Other ingroup taxa are represented by the presently recognized range of diagnosable heterodontosaurids as well as some well-preserved, approximately contemporary (Early–Middle Jurassic) and
structurally distinctive ornithischians; the exception
being Hypsilophodon (Early Cretaceous), which benefits from being anatomically generalized and particularly well described in the literature (Galton, 1974) –
see comments in Butler et al. (2008b: 6–12).
Heterodontosaurids
• Heterodontosaurus tucki Crompton & Charig, 1962
• Abrictosaurus consors (Thulborn, 1974)
• Unnamed taxon: NHMUK RU A100 (tentatively
referred to La. scalpridens, see above)
Representative ornithischian taxa
• Emausaurus ernsti Haubold, 1990
• Scelidosaurus harrisonii Owen, 1861
• Agilisaurus louderbacki Peng, 1990
• Hexinlusaurus multidens (He and Cai, 1984)
• Hypsilophodon foxii Huxley, 1869
• Lesothosaurus diagnosticus (Galton, 1978)
Outgroup taxa comprise a basal archosaur that is
phylogenetically distant from the Dinosauria (Sereno,
1991b) and a basal saurischian dinosaur (Novas,
1993; Sereno, 1993), both of which are known from
good material and good quality descriptions.
Outgroups
• Euparkeria capensis Broom, 1913
• Herrerasaurus ischigualastensis Reig, 1963
ANALYSIS
AND TOPOLOGY
The complete matrix (Appendix 1) comprised 14 taxa,
coded for 221 characters (Appendix 2) sourced from
Butler et al. (2008b). The data matrix was analysed
using the search options in PAUP*4.0b10 (Swofford,
2002).
The full data matrix was run under the ‘branchand-bound’ search option in PAUP*, with all charac-
Figure 40. Topology of the single most parsimonious tree
generated by a branch-and-bound search of the data
matrix using PAUP*4 in Appendix 1 (tree length 309
steps). N.B. ‘BMNH A100’ has recently been formally
renumbered NHMUK RU A100.
ters of equal weight and unordered, topological
constraints were not applied, trees were unrooted,
and character-state optimization was run under the
‘ACCTRAN’ option.
The analysis revealed that 125 characters were
parsimony informative, 65 characters were uninformative in the sense that their distribution did not
affect the underlying topology, and 31 characters
remained unchanged throughout the analysis (see
Appendix 2). A single most parsimonious tree of 309
steps (min: 207, max: 449) was generated (Fig. 40)
with reasonably high indications of statistical support
(consistency index: 0.67, retention index: 0.58,
homoplasy index: 0.33).
The general topology of the tree agrees, as should
be expected, with that of Butler et al. (2008b: figs 3,
4). The key elements within the topology repeat the
observation, noted earlier, that heterodontosaurids
represent the sister-group to all the other ornithischians. The latter clade is very inappropriately (given
the condition seen in heterodontosaurids) named
Genasauria (‘reptiles with cheeks’). Successive clades
within Genasauria are: a clade recognized as
Thyreophora (Lesothosaurus, Emausaurus, and Scelidosaurus) followed by two stem-lineage neornithischians (Agilisaurus and Hexinlusaurus) that are
successive outgroup taxa to the clade Cerapoda
241
Figure 41. The single most parsimonious tree from Figure 40, subjected to minor branch swapping (using MacClade) within the clade Cerapoda [Hypsilophodon,
Ankylopollexia, Pachycephalosauridae, and Psittacosauridae]; this part of the tree is not the principal focus of the
formal analysis. Tree length: 310 steps. N.B. ‘BMNH A100’
has recently been formally renumbered NHMUK RU
A100.
(comprising in this analysis Hypsilophodon,
Ankylopollexia, Psittacosauridae, and Pachycephalosauridae). Figure 41 represents a more conventional
topology than Figure 40 in that branch swapping
within the clade Cerapoda has been undertaken to
realign membership of the clades Ornithopoda (Hypsilophodon and Ankylopollexia) and Marginocephalia
(Psittacosauridae and Pachycephalosauridae); this
minor manipulation produces a tree that is entirely
conformable with the topology seen in Butler et al.
(2008b) and is only a single step longer (310 steps)
than the single most parsimonious tree illustrated in
Figure 40.
A REVIEW AND ASSESSMENT OF
PREVIOUS HYPOTHESES
The position of heterodontosaurids remains an
outstanding problem in ornithischian phylogeny.
Precladistic analyses referred heterodontosaurids
almost universally to the suborder Ornithopoda (e.g.
Crompton & Charig, 1962; Steel, 1969; Thulborn,
1971b; Galton, 1972), and even to the family Hypsilophodontidae (Thulborn, 1971b; Galton, 1972); this
242
D. B. NORMAN ET AL.
reflected the precladistic usage of Ornithopoda and
Hypsilophodontidae as ‘waste-basket’ taxa for bipedal
ornithischians (Romer, 1956). Santa Luca (1980)
noted that bipedality could not be used as a sufficient
character to group ornithopods together, and that
previous conceptions of Ornithopoda were paraphyletic; he therefore removed heterodontosaurids from
Ornithopoda and considered their phylogenetic
position uncertain; however, Santa Luca did also note
a number of similarities between Heterodontosaurus
and ceratopsians.
Following the work of Santa Luca (1980), several
authors (e.g. Maryańska & Osmólska, 1984; Cooper,
1985) positioned heterodontosaurids as the sister
group to Marginocephalia. However, Maryańska &
Osmólska (1985) suggested Heterodontosauridae
formed the sister clade to a grouping of ornithopods,
pachycephalosaurs, and ceratopsians, a clade
currently referred to as Cerapoda. Norman (1984a)
suggested two alternative phylogenetic positions
for Heterodontosauridae: either as the sister group
to Cerapoda (which also included ‘fabrosaurs’
in his analysis), or as the sister group of ‘Fabrosauridae’.
Sereno (1984, 1986) and Gauthier (1986) argued
that Heterodontosauridae form a basal clade of Ornithopoda. This proposal gained widespread acceptance
and was reinforced by its adoption in the first edition
of The Dinosauria (Weishampel & Witmer, 1990).
As a consequence, Heterodontosauridae (and in
particular H. tucki) began to be used as an outgroup
for phylogenetic analyses of euornithopods (e.g.
Weishampel & Heinrich, 1992); this phylogenetic
position was further supported by the analysis of
Sereno (1999).
A few early studies proposed that heterodontosaurids represent very basal ornithischians. Bakker &
Galton (1974) noted the strong similarities between
the manus of saurischians and that of Heterodontosaurus; based upon this character evidence they considered heterodontosaurids to represent a basal
branch of the ornithischian tree (Bakker & Galton,
1974: fig. 4), a view discussed further by Olsen &
Baird (1986: fig. 6.16).
With the descriptions of new, Chinese, basal ceratopsian taxa (Zhao, Cheng & Xu, 1999; You, Xu &
Wang, 2003; Xu et al., 2006) recent authors have
revived the idea of heterodontosaurids as the sister
group to Marginocephalia. Zhao et al. (1999) noted
similarities between Chaoyangsaurus (a basal ceratopsian) and Heterodontosaurus, but did not
incorporate these similarities into a phylogenetic
analysis. You et al. (2003) analysed the position of
Heterodontosaurus relative to ornithopods, ceratopsians, and pachycephalosaurs using a small cladistic
matrix. They concluded that Heterodontosaurus
formed the sister group to Marginocephalia and
referred to this clade as Neornithopoda (a name
originally coined by Cooper, 1985). Norman et al.
(2004c) reviewed the phylogeny of basal Ornithopoda and noted the unstable position of heterodontosaurids. Although the full phylogenetic analysis
carried out by these authors supported an ornithopod identity, Heterodontosauridae also grouped as
the sister taxon to Marginocephalia when analyses
were carried out using reduced data sets (following
the removal of unstable taxa). Xu et al. (2006)
carried out a large analysis that placed heterodontosaurids as the sister taxon to Marginocephalia,
and named the clade containing heterodontosaurids
and Marginocephalia as Heterodontosauriformes
(this clade name appears to be a junior synonym of
Neornithopoda).
Butler (2005) positioned heterodontosaurids as
the sister taxon of Cerapoda, reviving the hypothesis
of Maryańska & Osmólska (1985). Subsequent
expanded versions of this cladistic data matrix
have suggested that heterodontosaurids might in fact
be one of the most basal of all ornithischian radiations, being positioned basal to Genasauria (Butler
et al., 2007, 2008b, 2010; Witmer, 2009; Zheng et al.,
2009).
In summary, there is no long-term consensus concerning the position of the clade Heterodontosauridae
within Ornithischia. Four alternative phylogenetic
positions have been proposed: (1) as basal ornithopods; (2) as the sister group to Marginocephalia; (3) as
the sister group to Ornithopoda + Marginocephalia
(= non-cerapodan ornithischians); (4) the most basal
major ornithischian radiation (= non-genasaurian
ornithischians). The character evidence that has been
marshalled in support of each of these positions is
discussed below.
HETERODONTOSAURIDS
AS BASAL ORNITHOPODS
Heterodontosaurids have been commonly considered
to represent basal members of Ornithopoda (e.g.
Crompton & Charig, 1962; Thulborn, 1970b, 1971b;
Galton, 1972; Santa Luca et al., 1976; Gauthier,
1986; Sereno, 1986, 1999; Weishampel & Witmer,
1990; Norman et al., 2004c). However, relatively
little anatomical evidence has been advanced in
support of this view, and the strength of the available evidence has never been rigorously assessed.
The strongest case in favour of an ornithopod identity has been made by Sereno (1986, 1999) and
Weishampel (1990). These authors have suggested a
number of characters that might link heterodontosaurids with ornithopods:
1. Presence of a contact between the premaxilla and
lacrimal (Sereno, 1986, 1999; Weishampel, 1990).
243
cerapodans (e.g. Makovicky, 2001). In most ornithopods (e.g. Zephyrosaurus, Sues, 1980; Dryosaurus,
Galton, 1983) the distal end of the paroccipital
process is unexpanded dorsally but has a strong
ventral process, giving it a pendent form. Distally
pendent processes are also well developed in
pachycephalosaurs (e.g. Prenocephale, ZPAL MgD-I/
104; Goyocephale, Perle, Maryańska & Osmólska,
1982: pl. 42, fig. 1). Amongst Heterodontosauridae,
paroccipital processes are preserved only in Heterodontosaurus and are unusually deep relative to
their length. They appear to be expanded ventrally
at their distal end (Fig. 14; contra Butler et al.,
2008b). The similarities between the condition in
heterodontosaurids and ornithopods may provide
some evidence for the ornithopod hypothesis,
although, as noted above, similarly pendent processes also occur in pachycephalosaurs and thus
appear to be either more widely distributed or
homoplastic within Ornithischia.
Premaxilla–lacrimal contact on the external surface
of the skull is absent in ornithischian outgroups, in
basal ornithischians such as Lesothosaurus (Sereno,
1991a) and Agilisaurus (Peng, 1992; Barrett, Butler
& Knoll, 2005), in thyreophorans (e.g. Galton &
Upchurch, 2004b; Norman, Witmer & Weishampel,
2004b), and pachycephalosaurs (e.g. Sereno, 2000). In
these taxa a contact is present between the nasal and
the maxilla. Contact between the premaxilla and lacrimal on the external surface of the skull, preventing
maxilla–nasal contact, occurs in Heterodontosaurus
and in many ornithopods such as Dryosaurus and
Camptosaurus (Norman, 2004). In the basal ornithopod Jeholosaurus, a contact is present, but is ontogenetically variable and is broader in older individuals
(Barrett & Han, 2009). A contact is absent in most
basal ornithopods (Weishampel & Heinrich, 1992;
contra Sereno, 1986, 1999) including Hypsilophodon
(Galton, 1974), Parksosaurus (Galton, 1973b), Orodromeus (Scheetz, 1999), Zalmoxes (Weishampel
et al., 2003), Talenkauen (Novas, Cambiaso & Ambrosio, 2004), Tenontosaurus (Norman, 2004), and
Thescelosaurus (Boyd et al., 2009). A contact may also
be absent in the heterodontosaurid Abrictosaurus
(NHMUK RU B54), although the relevant portion of
the skull is poorly exposed and requires additional
preparation.
By contrast, a premaxilla–lacrimal contact is
present in the basal ceratopsians Psittacosauridae
(Sereno, 1990) and Liaoceratops (Xu et al., 2002). Poor
preservation in this area means that the presence/
absence of a contact cannot be ascertained in Yinlong
or Archaeoceratops (R. J. B., pers. observ.). However,
it is plausible that the presence of a contact could
represent the ancestral ceratopsian condition.
The distribution of this character within Ornithischia is complex, and requires further study. It is not
clear that the premaxilla/lacrimal contact was universally present in heterodontosaurids, the contact was
absent in many basal ornithopods and pachycephalosaurs, but occurs in basal ceratopsians (and is subsequently lost later in the evolutionary history of the
clade). This character as currently understood therefore provides little support to link heterodontosaurids
with ornithopods.
3. Jaw articulation offset ventral to the maxillary
tooth row (Sereno, 1986, 1999; Weishampel, 1990).
The jaw articulation is markedly ventrally offset in
Heterodontosaurus (Fig. 8) and ornithopods (e.g. Hypsilophodon, Galton, 1974); in these taxa the quadrate
condyles are level with the midpoint of the dorsoventral depth of the dentary when the skull and mandible are in articulation. However, a well-developed
and comparable ventral offset is also present in the
basal ornithischians Lesothosaurus (Norman et al.,
2004a: fig. 14.1) and Emausaurus (Haubold, 1990:
fig. 2), stegosaurs (e.g. Sereno & Dong, 1992: fig. 6),
and some pachycephalosaurs (e.g. Prenocephale,
ZPAL
MgD-I/104;
Maryańska,
Chapman
&
Weishampel, 2004). Indeed, the only ornithischians
that appear to lack a ventrally offset jaw articulation
are ceratopsians (e.g. Xu et al., 2006: fig. 2), and the
possession of a jaw articulation that is in line with the
maxillary tooth row may represent a synapomorphy
of Ceratopsia. This character thus provides little
support for the proposed Heterodontosauridae
+ Ornithopoda clade. Moreover, the jaw articulation
is, at present, only well preserved in Heterodontosaurus amongst heterodontosaurids.
2. ‘Crescentic’ or ‘pendent’ paroccipital process
(Sereno,
1986,
1999;
Weishampel,
1990).
Weishampel & Witmer (1990: 496) described this
character as: ‘paroccipital processes strongly
pendent laterally’. In ornithischian outgroups and
basal ornithischians the paroccipital processes
extend laterally and are slightly expanded at their
distal end (e.g. Lesothosaurus, Sereno, 1991a); this
weak expansion occurs both dorsally and ventrally.
This condition is retained in ceratopsians, amongst
4. Premaxillary tooth row offset ventral to the maxillary tooth row (Sereno, 1986, 1999; Weishampel,
1990). In the majority of ornithischians the maxillary
and premaxillary tooth rows are situated at the same
level. In heterodontosaurids (Abrictosaurus, NHMUK
RU B54; Heterodontosaurus, Fig. 8) and some ornithopods (e.g. Hypsilophodon, Galton, 1974) the premaxilla is ventrally offset relative to the maxillary
tooth row. In many basal ornithopods the ventral
offset is however absent (Changchunsaurus, Jin et al.,
244
D. B. NORMAN ET AL.
2010; Jeholosaurus, Barrett & Han, 2009, Orodromeus, MOR 1141; Thescelosaurus, Boyd et al.,
2009: fig. 4; Zephyrosaurus, Sues, 1980: fig. 16), and is
only pronounced in derived forms (e.g. Iguanodon
bernissartensis, Norman, 1980). A ventrally offset premaxilla is also present in pachycephalosaurs (e.g.
Prenocephale, ZPAL MgD-I/104; Goyocephale, Perle
et al., 1982: pl. 42, fig. 5), and the offset is comparable
to that seen in some basal ornithopods and in
heterodontosaurids. In view of the absence of this
character in many basal ornithopods and its presence
in some pachycephalosaurs, this character provides little support for a Heterodontosauridae +
Ornithopoda clade.
5. Thickening of the enamel on one side of the cheek
teeth, and the curvature of maxillary and dentary
teeth towards one another (Crompton & Attridge,
1986). Sereno (1984, 1986) demonstrated that both of
these characters have wider distributions within
Ornithischia, and are present in heterodontosaurids,
ornithopods, ceratopsians, and pachycephalosaurs.
In addition, asymmetrical enamel appears to be
absent in some heterodontosaurids (Echinodon,
Norman & Barrett, 2002), basal ceratopsians (Zhao
et al., 1999), and some basal ornithopods (Norman
et al., 2004c) and so the derived state (asymmetrical
enamel) may have arisen numerous times within
Ornithischia.
6. Reduction or loss of the cingulum (Weishampel &
Witmer, 1990). ‘Cingulum’ in this sense is used to
describe a mediolateral expansion of the base of the
crown above the root (see discussion in Irmis et al.,
2007). A ‘cingulum’ is actually present in many basal
ornithopods (Changchunsaurus, Jin et al., 2010; Jeholosaurus, Barrett & Han, 2009; Orodromeus, Scheetz,
1999; Zephyrosaurus, Sues, 1980) and some heterodontosaurids (e.g. Fruitadens, Butler et al., 2010). The
‘cingulum’ in these taxa is comparable in development to basal ornithischians such as Lesothosaurus
(Sereno, 1991a). Thus, the losses of the cingulum in
some heterodontosaurids and in ornithopods are
likely to be convergent.
One character that might link heterodontosaurids
with some basal ornithopods is the presence of a jugal
boss. A boss is present on the jugal of Heterodontosaurus and the basal ornithopods Changchunsaurus
(Jin et al., 2010), Orodromeus (Scheetz, 1999), and
Zephyrosaurus (Sues, 1980). The boss is very similar
in size and form in Heterodontosaurus and the latter
two ornithopods (the boss of Changchunsaurus is
smaller and covered with a nodular ornamentation:
Jin et al., 2010), suggesting that it could represent a
shared derived feature. However, all other known
ornithopods, including basal forms such as Hypsilophodon, lack such a boss on the jugal.
Summary: Most of the character states that have
been suggested to support a sister-group relationship
between Heterodontosauridae and Ornithopoda
appear to be absent in basal ornithopods, basal
heterodontosaurids, or both, or have wider distributions within Ornithischia. Evidence supporting a
Heterodontosauridae + Ornithopoda clade is therefore
weak.
HETERODONTOSAURIDS AS THE SISTER-TAXON
TO MARGINOCEPHALIA
The following characters have been suggested
by previous authors in favour of the clade
Heterodontosauridae + Marginocephalia:
1. Preorbital skull length 40–50% of the total length of
the skull (You et al., 2003). In basal ornithischians
(e.g. Lesothosaurus, Norman et al., 2004a) the preorbital length of the skull is typically approximately
50% of the total skull length. By contrast, in the basal
ceratopsians Chaoyangsaurus (Zhao et al., 1999),
Psittacosauridae (Sereno, 1987, 1990) and Yinlong
(Xu et al., 2006) the preorbital region of the skull is
shortened to less than 40% of the total skull length,
and this has been cited as a synapomorphy of Psittacosauridae (e.g. Sereno, 1990). You et al. (2003) suggested that some shortening of the preorbital region
of the skull is also present in heterodontosaurids
and basal marginocephalians. In Heterodontosaurus
(SAM-PK-K337, K1332; Fig. 8) the preorbital region
accounts for about 45% of preorbital skull length,
and similar values occur in the basal ceratopsians
Archaeoceratops (IVPP V11114, You & Dodson, 2003)
and Liaoceratops (IVPP V12738; Xu et al., 2002).
However, this represents only a very minor shortening of 3–5% when compared to other basal ornithischians. Moreover, the preorbital region of the skull
does not appear to be shortened in all pachycephalosaurs (e.g. Stegoceras, Maryańska et al., 2004:
fig. 21.2). This character therefore provides only weak
support for the proposed Heterodontosauridae +
Marginocephalia clade.
2. Presence of a contact between the premaxilla and
lacrimal (Xu et al., 2006). Although typically cited as
a character linking heterodontosaurids with ornithopods (see above), Xu et al. (2006) optimised this character as a synapomorphy of Heterodontosauridae +
Marginocephalia. As discussed above, this character
has a complex distribution within Ornithischia, and is
present in some basal ornithopods and absent in
pachycephalosaurs.
3. Midline contact between maxillae excludes premaxillae from the margins of the internal nares (Butler
et al., 2008b). In ornithischian outgroups and in most
ornithischians the footplate of the vomer makes
contact anteriorly with the posterior end of the premaxillary palate (e.g. Hypsilophodon, Galton, 1974).
By contrast, in marginocephalians median contact
between the maxillae prevents premaxilla–vomer
contact and excludes the premaxillae from the margin
of the internal nares (e.g. Maryańska & Osmólska,
1974: fig. 1A3, C3), and this has often been suggested
to be a marginocephalian synapomorphy (e.g. Sereno,
1999, 2000). Butler et al. (2008b) noted that the anterior processes of the maxillae meet along the midline
in Heterodontosaurus, and may exclude the premaxillae from the margins of the internal nares, and
suggested that this might be homologous with the
condition in marginocephalians. However, as noted by
Butler et al. (2008a), the condition in Heterodontosaurus may differ from that of marginocephalians in that
the vomers may still contact the premaxillae (Fig. 13),
and a similar condition to that seen in marginocephalians is now known in the basal ornithopod Jeholosaurus (Barrett & Han, 2009). Moreover, palatal
morphology is generally exceptionally poorly known
within Ornithischia. For these reasons, Butler et al.
(2008a) noted that this character provides only weak
support for a Heterodontosauridae + Marginocephalia
clade.
4. Jugal lateral expansion, or ‘horn’ or ‘flange’
(Cooper, 1985; Olshevsky, 1991; Weishampel & Heinrich, 1992; You et al., 2003; Xu et al., 2006). There do
not seem to be close similarities between the jugal
‘boss’ of Heterodontosaurus and the jugal ‘horn’ or
‘flange’ of ceratopsians. In Heterodontosaurus, the
boss is a discrete posterolaterally directed projection
located at the base of the jugal–postorbital bar. It is
not associated with any lateral expansion of the jugal,
nor is it supported by a descending ridge from the
jugal–postorbital bar. The jugal is not sufficiently well
preserved in other heterodontosaurids to indicate
whether a boss is present or not. The jugal horn of
most basal ceratopsians (e.g. Psittacosauridae:
Sereno, 1987, 1990) is located on the ventral margin
of the jugal, and slightly posterior to the jugal–
postorbital bar. The horn is associated with a broad
lateral expansion and thickening of the jugal, and is
supported by a ridge that descends from the jugal–
postorbital bar. This ridge divides the lateral surface
of the jugal into anterior and posterior surfaces. A
similar, although very subtle, ridge divides the lateral
surface of the jugal in some pachycephalosaurs
(Sereno, 1987) and in the most basal known ceratopsian Yinlong (RJB pers. obs.). In Yinlong a low
ridge subdivides the lateral surface of the jugal
topographically into two surfaces: an anterior surface
that faces laterally and slightly anteriorly, and a
245
posterior surface that faces laterally and slightly
posteriorly. This low ridge curves posteriorly and
runs horizontally along the posterior process of the
jugal, terminating near to the posterior extend of
the jugal.
In light of the morphological differences between
the jugal horn of marginocephalians and the boss
seen in Heterodontosaurus (and additionally in the
basal ornithopods Changchunsaurus, Orodromeus
and Zephyrosaurus, see above), it seems likely that
the features have been independently acquired.
Because of the detailed differences in morphology, the
‘jugal boss’ of Heterodontosaurus and the ‘jugal horn’
of ceratopsians were coded as separate characters in
the phylogenetic analysis of Butler et al. (2008b).
5. Elongated supratemporal fenestra, more than 25%
of basal skull length (Xu et al., 2006). The supratemporal fenestrae of Heterodontosaurus are just over
25% of skull length (Fig. 12); however, this is only
marginally larger than the supratemporal fenestrae
of ornithischians such as Hypsilophodon (Galton,
1974) and Lesothosaurus (Sereno, 1991a) in which the
supratemporal fenestrae are 20–25% of the length of
the skull. Indeed, the condition in Heterodontosaurus
appears more similar to these taxa than to the basal
ceratopsian Yinlong in which the fenestrae are nearly
45% of skull length (Xu et al., 2006). When present in
pachycephalosaurs (possibly only in juvenile individuals), the supratemporal fenestrae do not appear particularly elongate (e.g. Perle et al., 1982). This
character therefore provides little support for
the proposed Heterodontosauridae + Marginocephalia
clade.
6. Squamosal process of postorbital subequal to or
longer than the jugal process (Xu et al., 2006). Xu
et al. (2006) noted that the squamosal process of the
postorbital of Yinlong is long, forming the entire
dorsal margin of the infratemporal fenestra, and their
phylogenetic analysis optimised the presence of an
elongate squamosal process as a synapomorphy of
Heterodontosauridae + Marginocephalia. Heterodontosaurus possesses a moderately long squamosal
process (Fig. 8) that forms more than half of the
dorsal margin of the infratemporal fenestra; however
similarly elongate squamosal processes to those of
Heterodontosaurus are widespread within ornithischians, including Changchunsaurus (Jin et al.,
2010), Hypsilophodon (Galton, 1974), Lesothosaurus
(Sereno, 1991a), and Orodromeus (Scheetz, 1999). As
a result, this character does not support the proposed
Heterodontosauridae + Marginocephalia clade.
7. Infratemporal fenestra large, subequal to or larger
than the orbit (Xu et al., 2006). The infratemporal
246
D. B. NORMAN ET AL.
fenestra of Yinlong and psittacosaurids (but not other
basal ceratopsians) is greatly enlarged relative to the
condition in other ornithischians, and greatly exceeds
the orbit in size. The infratemporal fenestra of Heterodontosaurus is relatively large relative to the conditions in basal ornithischians (e.g. Lesothosaurus,
Sereno, 1991a) and ornithopods (e.g. Changchunsaurus, Jin et al., 2010), but it does not approach the
dimensions relative to the orbit of the fenestrae of
Yinlong and psittacosaurids. Moreover, the infratemporal fenestra is invariably small in pachycephalosaurs (Sereno, 2000). Therefore, this character
provides relatively ambiguous support for the
proposed
Heterodontosauridae + Marginocephalia
clade.
8. ‘Fewer than five premaxillary teeth’ or ‘1–3 premaxillary teeth’ (You et al., 2003; Xu et al., 2006). In
heterodontosaurids there are invariably three premaxillary teeth [Heterodontosaurus; Abrictosaurus,
NHMUK RU B54, contra Thulborn (1974); Fruitadens, Butler et al. (2010)]. Three premaxillary teeth
are also present in pachycephalosaurs (Maryańska
et al., 2004), while in basal ceratopsians the number
is variable [premaxillary teeth absent in psittacosaurids; one premaxillary tooth in Xuanhuaceratops
(Zhao et al., 2006); two teeth present in Chaoyangsaurus (Zhao et al., 1999); three teeth present in
Liaoceratops (Xu et al., 2002), Archaeoceratops (You &
Dodson, 2003) and Yinlong (Xu et al., 2006)]. In other
ornithischians the number of premaxillary teeth is
variable (and in multiple clades they are lost
completely), but is generally five or greater in basal
ornithischians (e.g. Lesothosaurus, Sereno, 1991a;
Hypsilophodon, Galton, 1974). A premaxillary tooth
count of 1–3 is only known in heterodontosaurids and
marginocephalians, and this character may therefore
provide support for this clade.
9. ‘Peg-like’ or
‘fang-like’ premaxillary
teeth
(Olshevsky, 1991). In most heterodontosaurids (e.g.
Abrictosaurus, NHMUK RU B54; Heterodontosaurus;
NHMUK RU A100) the crowns of the premaxillary
teeth are unexpanded above the root and increase in
size posteriorly, with the third crown being considerably larger than the maxillary crowns. This condition
is not universally present however: in Fruitadens
haagarorum the premaxillary crown are low, triangular, subequal in size, and not larger than the maxillary teeth (Butler et al., 2010). Similar to Fruitadens,
most ornithischians have premaxillary teeth with
crowns that are expanded mediolaterally and
anteroposteriorly above the root (e.g. Lesothosaurus,
Sereno, 1991a: fig. 6C; Huayangosaurus, Sereno &
Dong, 1992; Agilisaurus, ZDM T6011; Hypsilophodon,
Galton, 1974).
Basal ceratopsians have been occasionally described
as having ‘peg-like’ premaxillary teeth, supposedly
similar to those of most heterodontosaurids, and their
teeth are generally rather straight in lateral view and
lack well-developed recurvature. In some taxa there is
indeed relatively little distinction between the crown
and root (e.g. Archaeoceratops, IVPP V11114; Protoceratops); however, in other, more basal, taxa the root is
distinctly compressed beneath the crown (e.g. Yinlong,
IVPP V14530; Liaoceratops, IVPP V12738). In most
basal ceratopsians the premaxillary crowns are apicobasally short, and neither increase in size posteriorly
nor exceed the size of the maxillary dentition. The one
exception is Yinlong, in which the premaxillary crowns
are greatly enlarged relative to the maxillary crowns
(Xu et al., 2006). However, in Yinlong the largest crown
appears to be the second, not the third, and the crowns
are strongly expanded above the root and lack recurvature (IVPP V14530).
Pachycephalosaurs have also been described as
having ‘caniniform’ teeth, and this condition occurs
most clearly in those taxa with recurved premaxillary
teeth and only a very weak constriction of the root
beneath the crown (Goyocephale, Perle et al., 1982: pl.
42, figs 6, 7; Prenocephale, ZPAL MgD-I/104). By contrast, in Stegoceras validum this expansion is well
developed (e.g. Sues & Galton, 1987: fig. 4A, B). The
crowns are not enlarged relative to the maxillary
teeth.
Considerable variation is present in the premaxillary tooth morphology of heterodontosaurids, ceratopsians and pachycephalosaurs, and it is not clear that
the basal condition for each of these clades is to have
‘peg-like’ teeth. As a result this character fails to
provide support for the proposed Heterodontosauridae + Marginocephalia clade.
10. Arched diastema between the premaxilla and
maxilla, into which a caniniform dentary tooth fitted
(Olshevsky, 1991). A caniniform tooth is present in the
anterior dentary of some (although not all) heterodontosaurids (e.g. Heterodontosaurus; NHMUK RU
A100; Fruitadens, Butler et al., 2010), while an
arched diastema is present even in those heterodontosaurids that apparently lack a caniniform tooth
(e.g. Abrictosaurus, NHMUK RU B54). An arched
diastema with a pit to receive a dentary caniniform is
known in the pachycephalosaurs Goyocephale (Perle
et al., 1982) and Prenocephale (Maryańska & Osmólska, 1974), and an enlarged caniniform tooth is preserved in Goyocephale (Perle et al., 1982; the
mandible of Prenocephale is unknown). An enlarged
caniniform dentary tooth and arched diastema are
absent in the pachycephalosaur Stegoceras; in this
taxon the mesial dentary teeth are moderately
recurved, but are not caniniform and are subequal in
height to the more posterior dentary teeth (Sues &
Galton, 1987). The similarities between the caniniform and diastema/pit seen in heterodontosaurids
and that of Goyocephale and Prenocephale are
marked; however, the plesiomorphic state for
Pachycephalosauria remains unclear.
A diastema between the premaxillary and maxillary teeth is absent in the earliest and most basal
ceratopsian Yinlong downsi (Xu et al., 2006), and also
appears to be absent in Chaoyangsaurus (IGCAGS
V371), while the diastema in other ceratopsians is not
arched (e.g. You & Dodson, 2004). Furthermore a
caniniform dentary tooth is absent in all known ceratopsian taxa. It is also worth noting that a somewhat
similar feature to the arched and recessed diastema
of heterodontosaurids is present, and has been
described as a ‘subnarial notch’, in early theropods
(e.g. Nesbitt et al., 2009) and some other archosaurs
(e.g. Gauthier, 1986).
In summary, this character provides ambiguous
support for a link between heterodontosaurids and
marginocephalians because of its absence in all ceratopsians and the uncertain reconstruction of the
plesiomorphic condition for Pachycephalosauria.
11. Subcylindrical cheek teeth with planar wear surfaces (Cooper, 1985), and
12. Denticles restricted to upper third of crowns of
cheek teeth (Zhao et al., 1999). Both of these characters relating to the form of the cheek dentition are
problematic because they underestimate the variability of the dental morphology in heterodontosaurids
and basal marginocephalians, as well as more widely
within Ornithischia. There are indeed strong similarities between the dentition of the heterodontosaurids
Abrictosaurus (NHM RU B54), Echinodon (Norman &
Barrett, 2002), Lycorhinus (Hopson, 1975), NHM RU
A100 (Thulborn, 1970b) and Lanasaurus (Gow, 1975)
and the dentition of the early ceratopsians Chaoyangsaurus (IGCAGS V371) and Yinlong (IVPP V14530).
In all of these taxa the crowns are ‘chisel-shaped’,
with denticles restricted to the apical third, welldeveloped mesial and distal ridges, a weak basal
cingulum, and no primary ridge. However, similar
crowns are also present in the basal neornithischians
Agilisaurus and Hexinlusaurus (Barrett et al., 2005:
fig. 2) suggesting that this could represent a more
widespread morphology. In addition, although this
dental morphology is common within Heterodontosauridae, it is absent in Fruitadens (Butler et al.,
2010), which has low, triangular, Lesothosaurus-like
crowns, and which might represent the plesiomorphic
condition for the heterodontosaurid clade. The condition in Heterodontosaurus differs from that of basal
ceratopsians in that the crowns are much more
247
closely packed, lack cingula, and possess welldeveloped primary ridges. Pachycephalosaurs typically have crowns that are relatively low and
triangular and not chisel-shaped (e.g. Goyocephale,
Perle et al., 1982; Stegoceras, Sues & Galton, 1987;
Wannanosaurus, Butler & Zhao, 2009).
Although there are strong similarities between the
cheek teeth of basal ceratopsians and many heterodontosaurids, these characters are not present in all
heterodontosaurids or in pachycephalosaurs and are
not present in Fruitadens; moreover, these characters
are present in ornithischian taxa (Agilisaurus, Hexinlusaurus) that are referable to neither Marginocephalia nor Heterodontosauridae. As a result, these
characters provide weak support for the proposed
13. Maxillary teeth closely packed, with spaces
between them eliminated (Xu et al., 2006). Although
the maxillary teeth of Heterodontosaurus are
extremely closely packed, forming a dental battery,
this is not true of all heterodontosaurids. Gaps remain
between the crowns, particularly at their bases, in
other heterodontosaurids, including Abrictosaurus
(Thulborn, 1974), Echinodon (Norman & Barrett,
2002), Fruitadens (Butler et al., 2010), and Tianyulong
(Zheng et al., 2009). Likewise, gaps between adjacent
crowns remain in pachycephalosaurs and basal ceratopsians including Yinlong (IVPP V14530). This character does not therefore support a link between
heterodontosaurids and marginocephalians.
14. Akinetic skull (Olshevsky, 1991). Olshevsky
(1991) noted that heterodontosaurids lack a kinetic
skull, which is seen in most ornithopods. However,
the lack of cranial kinesis likely represents the
plesiomorphic ornithischian state (present in basal
ornithischians, thyreophorans, heterodontosaurids,
ceratopsians, and pachycephalosaurs) and cannot
be used to link heterodontosaurids with marginocephalians.
15. 21–22 presacral vertebrae (Cooper, 1985). Most
ornithischians have at least 24 presacral vertebrae,
consisting of 9 or more cervicals, and 15 or more
dorsals, with the numbers of both tending to increase
within Ornithischia. By contrast, only 21 presacral
vertebrae are present in Heterodontosaurus (consisting of 9 cervicals and 12 dorsals: Santa Luca, 1980),
21 are present in Psittacosauridae (Sereno, 1987),
and 22 are present in basal neoceratopsians (You &
Dodson, 2004). The number of presacral vertebrae is
unknown in pachycephalosaurs. Presacral number
might represent a synapomorphy linking heterodontosaurids with Marginocephalia; however assessment
of this character is compromised by missing data for
pachycephalosaurs.
248
D. B. NORMAN ET AL.
16. Asymmetric manus (Olshevsky, 1991). Olshevsky
(1991) suggested that heterodontosaurids and ceratopsians (in particular, Psittacosauridae) shared a
manus in which digits IV and V are reduced.
However, reduction of the outer digits does not appear
to have occurred to a greater extent in Heterodontosaurus than in other basal ornithischians. For
example, in Lesothosaurus metacarpal IV is 71% of
the length of metacarpal III (Sereno, 1991a), while
metacarpal IV of Heterodontosaurus is only moderately shorter, being 68% of the length of metacarpal
III (Santa Luca, 1980). The relative length of metacarpal V is nearly identical in Lesothosaurus and Heterodontosaurus (Santa Luca, 1980; Sereno, 1991a).
Therefore, an ‘asymmetric manus’ is an ornithischian plesiomorphy and fails to support a
17. Postacetabular process of ilium subequal in depth
to the preacetabular process (Xu et al., 2006). The
postacetabular process of the ilium of Heterodontosaurus is dorsoventrally very shallow, and differs
from the ilium of basal ornithischians such as
Lesothosaurus (Thulborn, 1972; Sereno, 1991a) in
that the brevis shelf is either reduced and horizontally directed or absent (SAM-PK-K1332). A
dorsoventrally shallow postacetabular process is
present in Yinlong (Xu et al., 2006: fig. S2) and some
basal ceratopsians (e.g. Archaeoceratops, You &
Dodson, 2003), although this is not as shallow
as in Heterodontosaurus. This character may
provide some support for a Heterodontosauridae +
Marginocephalia clade, although this is partially
contradicted by the presence of a deep postacetabular process in pachycephalosaurs (Maryańska &
Osmólska, 1974).
18. Prominent eversion of dorsal margin of postacetabular process (Xu et al., 2006). Xu et al. (2006)
noted that the dorsal margin of the postacetabular
process ‘flares laterally’. It is unclear exactly what
Xu et al. (2006) mean by this – the photographs of
the ilium (Xu et al., 2006: fig. S2) indicate only that
the dorsal margin of the ilium is thickened relative
to the rest of the body of the element. Such slight
thickening of the dorsal margin is relatively
common in ornithischians, although most pronounced in many marginocephalians (Butler & Sullivan, 2009). In Heterodontosaurus the dorsal
margin of the postacetabular process is also slightly
thickened (SAM-PK-K1332), and at the posterior
end of the process this thickened dorsal margin
expands in transverse width (Santa Luca, 1980:
fig. 18). This posterior transverse expansion of the
dorsal margin is not present in Yinlong, as far as
can be ascertained, and so the dorsal margins of the
postacetabular processes of Yinlong and Heterodontosaurus do not appear more similar to one another
than to other ornithischians.
In addition to these 18 characters, Xu et al. (2006)
noted several other similarities between heterodontosaurids and marginocephalians in their text. They
noted that the jugal contributes significantly to the
posterior margin of the antorbital fossa in Heterodontosaurus and Yinlong. A contribution by the jugal to the
margin of the antorbital fossa is, however, plesiomorphic for Ornithischia (and indeed Dinosauria),
being present in Lesothosaurus (Sereno, 1991a),
thyreophorans, and Agilisaurus (Barrett et al., 2005).
Xu et al. (2006) suggested that the occurrence of the
paraquadratic foramen within the quadrate was
shared by Heterodontosaurus and Yinlong; however, as
described in this paper the paraquadratic foramen is
positioned between the quadrate and quadratojugal in
Heterodontosaurus, as in other ornithischians (e.g.
Lesothosaurus, Sereno, 1991a). Xu et al. (2006) suggested that a ridge was present on the lateral surface
of the postorbital of Yinlong, and that this was a
similar condition to that seen in Heterodontosaurus;
however, examination of the holotype specimen of
Yinlong (IVPP V14530) reveals that the published
drawings (Xu et al., 2006: fig. 2) are somewhat misleading in this regard. There is no ridge on the lateral
surface of the postorbital of Yinlong; instead, the
posterior margin of the postorbital is thickened
and slightly raised relative to the rest of the bone.
This feature is distinct in position and morphology
from the much more prominent ridge seen in
Heterodontosaurus. Finally, Xu et al. (2006) cited the
enlarged premaxillary teeth of Yinlong as a character
shared with heterodontosaurids (see above for discussion of the similarities and differences between the
premaxillary teeth of heterodontosaurids and marginocephalians).
Summary. Although a large number of characters
have been proposed to support a Heterodontosauridae + Marginocephalia clade, many of these are
either problematic, have wider distributions amongst
ornithischians, or are only present in some heterodontosaurids and/or one of the marginocephalian
subclades.
The best evidence in support of a Heterodontosauridae + Marginocephalia clade appears to be provided at present by the reduced number of premaxillary teeth seen in heterodontosaurids and all
marginocephalians, the dentary caniniform and
associated arched diastema seen in heterodontosaurids and some pachycephalosaurs (but absent in ceratopsians), and the reduction in the number of
presacral vertebrae seen in Heterodontosaurus and
ceratopsians (unknown in pachycephalosaurs). A rigorous assessment of most of these characters is also
hampered by the absence of basal (pre-Late Cretaceous) pachycephalosaurs in the known fossil
record.
EVIDENCE
FOR A BASAL (NON-CERAPODAN OR
NON-GENASAURIAN) POSITION OF
HETERODONTOSAURIDS WITHIN
ORNITHISCHIA
Character evidence for a more basal (non-cerapodan
or non-genasaurian) position of heterodontosaurids
within Ornithischia is based upon the retention
of plesiomorphies, rather than the possession of
synapomorphies. The major problem has been
ascertaining the plesiomorphic character states for
ornithischians – the analysis of Butler et al. (2008b)
used the well-known and well-described saurischian
Herrerasaurus as the most proximate outgroup for
Ornithischia, assuming that it represented a very
basal saurischian morphology, as suggested by
Langer & Benton (2006). However, recent analyses
place Herrerasaurus within Theropoda (Nesbitt et al.,
2009), and this may affect the choice of outgroup taxa
for Ornithischia in future analyses.
A range of characters suggestive of a relatively
basal position for heterodontosaurids within ornithischian phylogeny are assessed below:
1. Accessory openings present within the antorbital
fossa (Butler et al., 2008b). In Heterodontosaurus
there are two anteriorly placed fenestrae within the
antorbital fossa, in addition to the internal antorbital
fenestra. Such additional fenestrae are generally
unknown in other ornithischians, with the only exception being Hypsilophodon (Galton, 1974). Butler et al.
(2008a) documented possible pneumatization of the
anterior process of the maxilla of the Heterodontosaurus specimens SAM-PK-K1332 and SAM-PK-K10487,
and suggested that this pneumatization occurred via
the accessory fenestrae. They discussed the possibility that one or both of these fenestrae could be
homologous with topologically similar features, the
maxillary and promaxillary fenestrae (e.g. Rauhut,
2003; Sereno, 2007), observed amongst theropods and
Herrerasaurus, with the implication that their presence in Heterodontosaurus would represent the retention of a basal dinosaur plesiomorphy. As noted by
Butler et al. (2008a), problems with this hypothesis
include the uncertain distribution of accessory openings within Heterodontosauridae and the uncertainty
regarding the plesiomorphic condition within Saurischia. At present therefore this character provides
little support for the proposed basal position of heterodontosaurids within Ornithischia, but requires
further investigation.
2. Presence of a contact between the squamosal and
the quadratojugal (Butler et al., 2008b). In ornithis-
249
chian outgroups (e.g. Herrerasaurus, Sereno & Novas,
1993: fig 7A; sauropodomorphs, Sereno & Novas,
1993: fig 10B) the ascending process of the quadratojugal contacts a long descending process of the squamosal. This contact is retained in Heterodontosaurus
and a few basal ornithischians, including Lesothosaurus (Sereno, 1991a: fig 12A) and Scelidosaurus
(NHMUK R1111), but it is lost in almost all other
ornithischians. The retention of this character in Heterodontosaurus may therefore support a noncerapodan phylogenetic position, as suggested by
Butler et al. (2008b).
3. Absence of a spout-shaped mandibular symphysis
(Butler et al., 2008b). Generally in ornithischians the
ventral surface of the mesial end of the dentary is
turned strongly inwards to meet its opposing
member; as a result, in dorsal view the symphysis is
somewhat ‘spout’-shaped (as opposed to the V-shaped
symphysis of outgroups), and this character state has
often been considered an ornithischian synapomorphy
(e.g. Sereno, 1999). Norman & Barrett (2002) noted
that a spout-shaped symphysis is absent in heterodontosaurids, but present in almost all other ornithischians, and suggested that the absence of this
character was a synapomorphy of Heterodontosauridae. Butler et al. (2008b) reinterpreted the apparent
absence of a spout-shaped symphysis as a retained
plesiomorphy supporting a basal position for heterodontosaurids. We note here, however, that this character is rather poorly defined at present, and
additional study of the variation in the morphology of
the dentary symphysis within Ornithischia is
required before this character can be considered
to provide strong support for a basal position of
heterodontosaurids.
4. Alveolar foramina absent medial to maxillary and
dentary tooth rows (Butler et al., 2008b). The presence
of a regular series of foramina positioned on the
inside of the jaws, near the base of the teeth, was
discussed by Edmund (1957), who noted that such
‘special foramina’ (= alveolar foramina) are widespread within ornithischians, being present in
ceratopsians, ornithopods, pachycephalosaurs, stegosaurs, and ankylosaurs. Later work has additionally
demonstrated the presence of special foramina
in basal thyreophorans (Colbert, 1981; Haubold,
1990; Barrett, 2001) and basal ornithischians (e.g.
Lesothosaurus, Sereno, 1991a). However, until
recently, alveolar foramina were not known in any
heterodontosaurid taxon. This led Thulborn (1974,
1978) to suggest that normal patterns of tooth
replacement were suppressed in heterodontosaurids,
and that the teeth were replaced as a unit during
periods of aestivation (see Hopson, 1980; Butler et al.,
250
D. B. NORMAN ET AL.
2008a). Norman & Barrett (2002) interpreted the
absence of special foramina as a synapomorphy of
heterodontosaurids, and Butler et al. (2008b) suggested that this absence might instead be plesiomorphic for Ornithischia, and thus support a basal
position for Heterodontosauridae. Subsequent work
has demonstrated the presence of replacement teeth
in the heterodontosaurids Heterodontosaurus (this
paper) and Fruitadens (Butler et al., 2010). When
present, the replacement teeth are invariably associated with foramina on the medial surface of the
maxilla that are likely to be homologous with the
alveolar foramina of other ornithischians. The low
number and sporadic appearance of such foramina
(presumably because of relatively low replacement
rates) remains a distinctive feature of heterodontosaurids within Ornithischia. In addition, alveolar
foramina somewhat similar to those of ornithischians
are known in saurischians (e.g. Galton, 1984; Welles,
1984) and are formed at bases of the junctions
between adjacent interdental plates. Further study of
this character is required, but at present it does not
appear to support a basal position for heterodontosaurids within Ornithischia.
5. Presence of the external mandibular fenestra
(Butler et al., 2008b). In ornithischian outgroups a
large external mandibular fenestra is present posterior to the dentary tooth row, situated on the intersection between the dentary, surangular, and
angular, and this opening is retained in heterodontosaurids (Heterodontosaurus; Abrictosaurus, R. J.
B., pers. observ. of NHMUK RU B54, contra Thulborn, 1974) and some basal ornithischians and
thyreophorans
(Emausaurus,
Haubold,
1990;
Lesothosaurus, Sereno, 1991a; stegosaurs, Sereno &
Dong, 1992). By contrast, the external mandibular
fenestra is absent in most other ornithischians, and
Butler et al. (2008b) suggested that the retention of
this feature in heterodontosaurids might support a
non-cerapodan position for the clade. However, we
note that a small external mandibular fenestra is in
fact known in the basal ceratopsian Yinlong (Xu
et al., 2006) and some species of Psittacosaurus (e.g.
Sereno et al., 1988). Therefore, the distribution of
this character within Ornithischia is likely to be
complicated, involving multiple secondary losses, and
provides little support in isolation for a basal position for heterodontosaurids.
6. Premaxillary crowns not expanded mesiodistally or
apicobasally above the root (Butler et al., 2008b). In
ornithischian outgroups, the crowns of premaxillary
teeth are almost completely confluent with the roots,
and there is no anteroposterior or mediolateral expansion of the crown above the root. The premaxillary
crowns are generally either subcircular and spike-
like, with only a few weak serrations (e.g. Euparkeria,
Ewer, 1965; basal sauropodomorphs, Yates, 2003;
basal theropods, Tykoski & Rowe, 2004), or transversely compressed (e.g. Herrerasaurus, Sereno &
Novas, 1993). In these features the premaxillary
teeth of ornithischian outgroups resemble those of
heterodontosaurids. By contrast, the premaxillary
teeth of most other ornithischians (although see discussion for marginocephalians, above) have crowns
that are at least weakly expanded mediolaterally and
anteroposteriorly above the root. Butler et al. (2008b)
therefore interpreted the morphology of the premaxillary teeth of heterodontosaurids as a retained basal
dinosaur plesiomorphy. However, the recognition of
expanded premaxillary crowns in the heterodontosaurid Fruitadens (see above; Butler et al., 2010) has
complicated assessment of this character, and this
may not provide support for a basal position.
7. Pronounced epipophyses on anterior cervicals
(Butler et al., 2008b). The development of epipophyses
on anterior cervicals has often been considered (e.g.
Gauthier, 1986) to represent a saurischian synapomorphy; however, as noted by Langer & Benton
(2006), remnants of epipophyses are likely to be
present in many ornithischians as simple unexpanded
ridges on top of the postzygapophyses (e.g. Lesothosaurus, Sereno, 1991a; Tenontosaurus, Forster, 1990:
fig. 1). These epipophyses are much better developed
in Heterodontosaurus (SAM-PK-K1332) than in any
other ornithischian, and in the cervical three they
actually extend beyond the posterior margin of the
postzygapophysis. This morphology is most similar to
the condition in Herrerasaurus (Sereno & Novas,
1993), some sauropodomorphs and most theropods
(Langer & Benton, 2006), and may therefore represent a basal dinosaur plesiomorphy that has been
retained in heterodontosaurids and lost in other ornithischians, as suggested by Butler et al. (2008b).
8. Manus length (measured along digit 2 or 3, whichever is longest) is more than 40% of the combined
length of the humerus and radius, or, longest manual
phalanx more than 15% of the length of the humerus
(Butler et al., 2007; Butler et al., 2008b). The manus
is almost completely unknown in immediate dinosaurian outgroups and is poorly known in many ornithischians. The manus is relatively short (20–35% of the
combined length of the humerus and radius) in most
ornithischians for which it is known (e.g. Lesothosaurus, Sereno, 1991a; Hypsilophodon, Galton, 1974).
Gauthier (1986) suggested that an enlarged manus
(more than 40% of the length of the humerus and
radius) is synapomorphic for saurischians (see discussion in Langer & Benton, 2006). An enlarged manus
(56% of the combined length of the humerus and
251
radius) is also present in Heterodontosaurus (Santa
Luca, 1980). Although Thulborn (1974) described the
length of the manus of Abrictosaurus as ‘diminutive’,
it is not possible to determine the exact length as the
manus is incomplete (NHMUK RU B54). The
enlarged manus of Heterodontosaurus has generally
been considered independently derived from that of
saurischians; however, it could alternatively represent the retention of a basal dinosaurian plesiomorphy as suggested by Butler et al. (2008b). Butler et al.
(2007) formulated this character in terms of the size
of the manual phalanges relative to the humerus, in
order to incorporate data indicating that the manus of
the basal ornithischian Eocursor was also proportionally large.
their respective proximal phalanges (Gauthier, 1986;
Novas, 1996; Rauhut, 2003; Langer, 2004; Langer &
Benton, 2006). This may represent the basal saurischian condition, although this is ambiguous, as this
character state is absent in known basal sauropodomorphs and apparently Eoraptor (Langer &
Benton, 2006), and the condition is unknown in most
dinosaurian outgroups. Elongate penultimate manual
phalanges are also present in Heterodontosaurus
(Santa Luca, 1980) and Eocursor (Butler et al., 2007;
Butler, in press). The presence of this character in
Heterodontosaurus may represent a retained plesiomorphy, lost in most other ornithischians, rather than
convergence as previously assumed (Langer &
Benton, 2006).
9. Metacarpals with block-like proximal ends (Butler
et al., 2008b). Sereno (1986) suggested that the presence of metacarpals with ‘blocklike proximal ends’
was a synapomorphy of Heterodontosauria (= Heterodontosauridae). Sereno was referring to the fact that
in Heterodontosaurus and Abrictosaurus the proximal
and medial/lateral surfaces of the metacarpals meet
each other at approximately 90 degrees (e.g. Santa
Luca, 1980), rather than along a gently rounded
margin as in most other ornithischians. A similar
condition is present in at least some basal saurischians (e.g. Sereno, 1993) and thus Butler et al.
(2008b) suggested that the condition in heterodontosaurids might represent a retained basal dinosaur
plesiomorphy supporting non-genasaurian affinities
for heterodontosaurids and subsequently lost in other
ornithischians.
12. Manus unguals strongly recurved with prominent
flexor tubercle (Butler et al., 2008b). Well-formed
ventral tubercles for the insertion of the flexor
tendons on the manual phalanges are present in
saurischians (e.g. Rauhut, 2003; Galton & Upchurch,
2004a: fig. 12.8; Nesbitt et al., 2009) and in Heterodontosaurus (Santa Luca, 1980: fig. 11), but are absent
in all other ornithischians. This may therefore represent a basal dinosaur character that has been
retained by heterodontosaurids but lost by other
ornithischians.
10. Extensor pits on the dorsal surface of the distal end
of metacarpals and manual phalanges (Butler et al.,
2008b). As discussed by Novas (1993) and Langer &
Benton (2006), extensor depressions for phalangeal
hyperextension are present on the metacarpals and
phalanges of Heterodontosaurus (Santa Luca, 1980),
Herrerasaurus, theropods, and sauropodomorphs
(although weakly developed), but are absent in most
other ornithischians with the exception of Eocursor, in
which extensor depressions are weakly developed on at
least some manual phalanges (Butler et al., 2007;
Butler, in press), and Scutellosaurus, in which a weak
extensor depression is present on the distal end of one
metacarpal (Colbert, 1981; Langer & Benton, 2006).
This may therefore represent a basal dinosaur character that has been retained by heterodontosaurids but
lost in most other ornithischians.
11. Length of the penultimate phalanx of the second
and third fingers (Butler et al., 2008b). In theropods
and Herrerasaurus the penultimate phalanges of the
second and third digits are subequal to or longer than
13. Pubic peduncle of ilium large, not reduced in
size (Butler et al., 2008b). In ornithischian outgroups
and a number of basal ornithischians (e.g. Lesothosaurus, Sereno, 1991a; Scelidosaurus, NHMUK
R1111; Stormbergia, Butler, 2005), including Abrictosaurus (NHMUK RU B54) and Heterodontosaurus
(Santa Luca, 1980: fig. 18), the pubic peduncle of the
ilium is large, elongate, and robust, and generally
extends ventrally at least as far, or further, than
does the ischial peduncle. By contrast, in other ornithischians the pubic peduncle is reduced in size and
relatively short when compared to the ischial
peduncle (e.g. Agilisaurus, Peng, 1992; Hypsilophodon, Galton, 1974). Sereno (1986, 1999) suggested
that the derived state for this character supported
the clade Genasauria to the exclusion of Lesothosaurus, whereas Butler (2005) noted that this
character had a wider distribution and Butler et al.
(2008b) suggested that its retention in heterodontosaurids might support non-cerapodan affinities for
the clade.
14. Prepubic process is short, not anteriorly elongated (Butler et al., 2008b). A prepubic process is
present in all ornithischians for which the condition
is known, but is short and stub-like in basal ornithischians (Lesothosaurus, Sereno, 1991a; Stormbergia, Butler, 2005; Scelidosaurus, NHMUK R1111),
252
D. B. NORMAN ET AL.
including Heterodontosaurus (Santa Luca, 1980:
fig. 18). By contrast, the prepubic process is
extended anteriorly into an elongate process in
Yinlong (Xu et al., 2006), psittacosaurids (Sereno,
1987, 1990), the basal neoceratopsian Archaeoceratops (IVPP V11114), pachycephalosaurs (Maryańska & Osmólska, 1974), Agilisaurus (ZDM T6011),
Hexinlusaurus multidens (ZDM T6001; He & Cai,
1984), and most ornithopods (e.g. Hypsilophodon,
Galton, 1974). As a result, Butler et al. (2008b) suggested that the presence of a short prepubic presence in heterodontosaurids might support noncerapodan affinities for the clade.
15. Fossa trochanteris does not form a distinct constriction separating the femoral head and greater
trochanter (Butler et al., 2008b). In basal saurischians (e.g. Herrerasaurus, Novas, 1993: fig. 7; Saturnalia,
Langer,
2003:
fig. 4),
and
basal
ornithischians
(Lesothosaurus,
Sereno,
1991a;
Stormbergia, Butler, 2005; Agilisaurus, Peng, 1992;
Scutellosaurus, MNA P1.175), including Heterodontosaurus (SAM-PK-K1332), the femoral head and
the greater trochanter are continuous with one
another in anterior and posterior views. In proximal
view, a weak groove extends from the anteromedial
corner of the femoral head posterolaterally across
the proximal surface of the femur, dividing the
surface into anterolateral and posteromedial areas;
this groove has been identified as the fossa trochanteris (Langer, 2003).
As noted by Langer (2003), in many more derived
ornithischians
(e.g.
ornithopods,
ceratopsians,
pachycephalosaurs) the fossa trochanteris is highly
modified into a well-developed constriction that separates the elevated greater trochanter from the
inturned medial part of the head. Butler et al. (2008b)
therefore suggested that the absence of this constriction in heterodontosaurids might support noncerapodan affinities for the clade.
In addition to the characters discussed above, there
are a number of additional similarities between heterodontosaurids and some basal theropods that are
worthy of further investigation but have yet to be
incorporated into ornithischian phylogenetic analyses. These include the ‘subnarial notch’ and basisphenoid recess (e.g. Nesbitt et al., 2009), the morphology
of the humerus (which in Heterodontosaurus
resembles Herrerasaurus in possessing a distinct
medial tubercle proximally and a well-developed facet
on the entepicondyle distally; Santa Luca, 1980;
Sereno, 1993), and some features of the proximal
tarsals (see Butler et al., 2010).
In general, characters supporting a nongenasaurian position for heterodontosaurids are
concentrated in the hindlimb (although other poten-
tially important characters have been recognized in
the skull and axial skeleton), and it is possible that
the analyses of Butler et al. (2007), Butler et al.
(2008b) have overweighted a suite of functionally
integrated features. This concern is somewhat counterbalanced by the apparent presence of some of
these features (e.g. elongate distal phalanges, presence of extensor pits on distal ends of phalanges) in
the incomplete manus of the early ornithischian
Eocursor (Butler et al., 2007; Butler, in press).
Assessment of the distribution of most of these characters is also complicated by the uncertainties surrounding basal saurischian phylogeny and the
uncertain scorings for many nondinosaurian dinosauromorphs (and thus the correct outgroup condition for Ornithischia), as well as by missing data for
most heterodontosaurid taxa. Therefore, although
there are striking similarities between the anatomy
of Heterodontosaurus and some basal saurischians,
further work is required to determine whether a
basal position for Heterodontosauridae within Ornithischia is more plausible than the alternatives.
SUMMARY
Heterodontosaurus tucki is a highly characteristic
Early Jurassic ornithischian, whose remains include
a partly crushed holotype skull (SAM-PK-K337) along
with an almost complete skull and its associated
postcranial skeleton (SAM-PK-K1332). More complete
(although not entirely definitive) description of most
of the presently available cranial material assigned to
this taxon is presented here and reveals a range of
anatomical features that had been either unclear or
unknown, given the very limited description of the
craniology that had been available hitherto. The skull
of Heterodontosaurus is compact and, for its size,
surprisingly robust. The differentiation of its dentition is its most distinctive feature: it combines anterior teeth (incisiforms and caniniforms) that are
reminiscent, in their detailed structure and proportions, of those seen in basal carnivorous archosaurs;
these are associated with remarkable maxillary and
dental ‘batteries’ of hypsodont teeth that become worn
down to form oblique, contiguous, and warped cutting
blades. Although the cheek teeth form an apparently
uniform stockade-like array there is also subtle
differentiation in form between upper and lower teeth
as well as within each battery. The dentary and
maxillary batteries are inset markedly from the
lateral surface of the skull, creating a pronounced
‘cheek’ recess (the ‘genasaurian’ condition) that is
found universally within the ornithischian clade. The
cheek teeth have long roots that are firmly embedded
in their alveoli and appear to form stable arrays.
Material long-known, but first described here,
appears to provide evidence that the batteries were
replaced episodically, rather than continuously
throughout the animal’s ontogeny.
The mandible is highly unusual by virtue of the
structure of the postdentary bones. The surangular
develops two curved, strap-like, anterior rami, separated by a narrow cleft, that lie above a shallow
depression on the lateral surface of the jaw, formed by
the angular. The upper ramus of the surangular
appears to be loosely sutured to the dorsal portion of
the dentary immediately behind the coronoid eminence; this remnant of the archosaurian intramandibular joint permitted flexure within the dorsal part
of the lower jaw that was linked to a highly unusual
jaw action.
The palatal vault is very narrow and deep, and the
basal articulation sits anteroventral to the braincase,
which is unusual in ornithischians generally. A
central plate, formed by the pterygoids at the posterior end of the palate suggests that the pterygoids
were closely opposed in the midline and may have
had a limited – connective tissue-spanned – interpterygoid vacuity. The ventrolateral wall of the
braincase is partly overlain by a thin, sheet-like
basisphenoid flange that extends ventrally to enclose
a narrow slot parasaggital to the midline axis of the
main body of the basisphenoid.
The occiput is dominated by the broad and deep
paroccipital processes and is surrounded by a
thickened frame-like structure that, dorsally, is
reminiscent of the parietosquamosal shelf of basal
marginocephalians. The post-temporal fenestra has
been identified as piercing the latter as a small,
discrete, foramen. Positionally this foramen is
considered more likely to be the archosaurian cranioquadrate passage (connected to the otic capsule);
the post-temporal fenestra is regarded as being
reduced to a narrow passage located at the cleft
between the supraoccipital, squamosal, and paroccipital and communicating with a fissure further
anteriorly on the dorsolateral surface of the braincase. There is some evidence for cranial pneumatism
in the maxilla and the jugal boss; additionally
small ‘dimples’ in the paroccipital and posterior
shaft of the quadrate may be indicative of similar
pneumatism.
Jaw musculature of this dinosaur can be described
in broad detail on the basis of cranial morphology and
local osteological indicators. The reconstructed musculature appears to be typical of that seen in modern
diapsids, with the exception of the remarkable
curtain of superficial adductor muscles that drape the
lateral surface of the skull between the postorbitalsquamosal bar to the rim of the angular on the lower
jaw, which would impart torsional forces upon the
lower jaw.
253
Taxonomic review of the heterodontosaur material
recovered and described from southern Africa suggests that there are currently at least four valid
heterodontosaurids: H. tucki, Ly. angustidens, A.
consors and La. scalpridens. Additionally, NHMUK
RU A100 (provisionally referred to La. scalpridens)
and A. consors are in need of further preparation and
analysis; further discoveries are also currently being
worked on and are likely to increase still further the
diversity of heterodontosaurs known in the Early
Jurassic. Previously named taxa have proved either
to be undiagnosable or referable to the above-named
taxa.
The systematics and phylogenetics of Heterodontosaurus were reviewed using a restricted range of taxa.
The clade Heterodontosauridae appears to be basal
within the Ornithischia as the sister group to the
Genasauria (all other ornithischians). The phylogenetic position of the Heterodontosauridae and the
evidence that has been used to support a range of
positions within the overall phylogeny of the Ornithischia in past decades were reviewed. The characters that have been used to support opposing views
have been assessed and although a basal (nongenasaurian) position is favoured by the evidence
available it is by no means unambiguous; a number of
issues relating to the polarity of characters in immediate outgroups as well as missing data within
heterodontosaurids more generally need further
attention.
Heterodontosaurs, rather than being comparative
rarities (as they had seemed when first discovered)
were clearly significantly speciose in the Early
Jurassic of southern Africa and must have been an
important, and curiously specialized (anatomically,
physiologically, and behaviourally) component of the
ecological community at this time. It is quite clear
that these dinosaurs still have a lot to tell palaeobiologists about the tempo and mode of early phases of
ornithischian evolution.
ACKNOWLEDGEMENTS
The late Dr T. A. Barry and Dr Michael Cluver,
formerly, respectively, Director and Assistant Director
of the [Iziko] South African Museum, Cape Town,
permitted the preliminary study of the two original
skulls of H. tucki in their care by A. W. C./A. J. C.; this
period of work was subsequently augmented by loans
of these skulls and related ornithischian material
through Sheena Kaal and Roger Smith (Iziko South
African Museum to D. B. N.). The holotype of H. tucki
(SAM-PK-K337) was originally prepared by the late
Mr Arthur Rixon of the Natural History Museum,
London, with some additional work undertaken by A.
W. C. The referred skull (SAM-PK-K1332) was
254
D. B. NORMAN ET AL.
prepared by Ms Ione Rudner at the South African
Museum. The National Science Foundation provided
funds for the 1966–67 expedition to southern Africa,
in which A. W. C. and A. J. C. participated, and during
which the second skull and skeleton of Heterodontosaurus (SAM-PK-K1332), the maxillary fragment
(SAM-PK-K1334) and the additional skull (NHMUK
R2501 attributed to Lesothosaurus diagnosticus) were
found. The authors are particularly indebted to Mrs
Sheena Kaal, Dr Roger Smith, and Dr Gillian King
(currently or formerly of the Iziko South African
Museum, Cape Town) for loan of material and generous access to collections in their charge during several
visits by the authors. For access to specimens in their
care we are also grateful to A. C. Milner, P. M. Barrett
and S. D. Chapman (all of the Natural History
Museum); E. Butler (Nazionale Museum, Bloemfontein); M. Raath (Bernard Price Institute, University
of the Witwatersrand, Johannesburg) and C. Schaff
(MCZ, Harvard University). The authors are grateful
to Dr Paul Barrett (NHM) and Dr Randall Irmis
(University of Utah) for detailed comments upon the
submitted manuscript and any errors that remain
herein must be the responsibility of the corresponding
author. A partial manuscript on the cranial anatomy
of H. tucki was prepared jointly by A. W. C. and A. J.
C. during the period 1966~1986; the unfinished
manuscript was passed to D. B. N. by Mark Charig,
under the terms of the will of Dr Alan Jack Charig
FLS (following his death in 1997), with the wish for it
to be published. Dr A. W. (Fuzz) Crompton very kindly
and generously passed his original set of prepared
illustrations as well as an extremely valuable archive
of photographs (prepared for him in Cape Town and
at the MCZ, Harvard University) that documented
stages during the original preparation of the referred
skull (SAM-PK-K1332). The Royal Society and NRF
(formerly the FRD) of South Africa funded research
visits and an expedition to the Karoo of South Africa
by D. B. N. The Natural Environment Research
Council, UK (NERC), and the Gates Foundation &
Cambridge Commonwealth Trust supported doctorate
research studentships for R. J. B. and L. B. P., respectively, under the supervision of D. B. N. in the Department of Earth Sciences (Cambridge). Figures 20
(inset), 21, 24, 26, and 30 were prepared by Mr L.
Mesoly but have been edited where necessary by
D. B. N.; the segmented CT scans of the maxilla
showing active tooth replacement were prepared
by L. B. P.; the illustrations that appear in
Figure 38A–D were prepared for A. J. C. by Marilyn
Holloway of the Natural History Museum; Figures 23,
25 and Appendices 4–6 were prepared from photographs taken by Dudley Simons (Dept Earth Sciences, Cambridge); all other figures were prepared
and finished by D. B. N.
REFERENCES
Attridge J, Charig AJ. 1967. Crisis in evolution: the Stormberg Series. Science Journal 1967: 48–54.
Attridge J, Crompton AW, Jenkins FA. 1985. The southern African Liassic prosauropod Massospondylus discovered
in North America. Journal of Vertebrate Paleontology 5:
128–132.
Báez AM, Marsicano CA. 1998. A heterodontosaurian ornithischian in the Upper Triassic of Patagonia? Journal of
African Earth Science 27A: 14–15.
Báez AM, Marsicano CA. 2001. A heterodontosaurid ornithischian dinosaur from the Upper Triassic of Patagonia.
Ameghiniana 38: 271–279.
Bakker RT, Galton PM. 1974. Dinosaur monophyly
and a new class of vertebrates. Nature 248: 168–
172.
Barrett PM. 1998. Herbivory in the non-avian Dinosauria.
PhD thesis, University of Cambridge, UK.
Barrett PM. 1999. A reassessment of the enigmatic ornithischian Echinodon becklesii. Journal of Vertebrate Paleontology 19 (Abstracts): 31A.
Barrett PM. 2000. Prosauropod dinosaurs and iguanas:
speculations on the diets of extinct reptiles. In: Sues H-D,
ed. Evolution of herbivory in terrestrial vertebrates: perspectives from the fossil record. Cambridge: Cambridge University Press, 42–78.
Barrett PM. 2001. Tooth wear and possible jaw action of
Scelidosaurus harrisonii Owen and a review of feeding
mechanisms in other thyreophoran dinosaurs. In: Carpenter
K, ed. The armored dinosaurs. Bloomington: Indiana University Press, 25–52.
Barrett PM, Han F-L. 2009. Cranial anatomy of
Jeholosaurus shangyuanensis (Dinosauria: Ornithischia)
from the Early Cretaceous of China. Zootaxa 2072: 31–
55.
Barrett PM, Butler RJ, Knoll F. 2005. Small-bodied ornithischian dinosaurs from the Middle Jurassic of Sichuan,
China. Journal of Vertebrate Paleontology 25: 823–
834.
Boyd CA, Brown CM, Scheetz RD, Clarke JA. 2009.
Taxonomic revision of the basal neornithischian taxa
Thescelosaurus and Bugenasaura. Journal of Vertebrate
Paleontology 29: 758–770.
Broom R. 1911. On the dinosaurs of the Stormberg,
South Africa. Annals of the South African Museum 7: 291–
308.
Broom R. 1913. Note on Mesuchus browni, Watson, and on a
new South African pseudosuchian (Euparkeria capensis).
Records of the Albany Museum 2: 394–396.
Broom R. 1932. The mammal-like reptiles of South Africa
and the origin of mammals. Witherby: London.
Brown B, Schlaikjer EM. 1940. The structure and relationships of Protoceratops. Annals of the New York Academy of
Science 40: 133–266.
Bryant HN, Russell AP. 1992. The role of phylogenetic
analysis in the inference of unpreserved attributes of extinct
taxa. Philosophical Transactions of the Royal Society of
London, Series B 337: 405–418.
Bryant HN, Seymour KL. 1990. Observations and
comments on the reliability of muscle reconstruction
in fossil vertebrates. Journal of Morphology 206: 109–
117.
Busbey AB. 1989. Form and function of the feeding apparatus of Alligator mississipiensis. Journal of Morphology 202:
99–127.
Butler RJ. 2005. The ‘fabrosaurid’ ornithischian dinosaurs
of the Upper Elliot Formation (Lower Jurassic) of South
Africa. Zoological Journal of the Linnean Society 145: 175–
218.
Butler RJ. 2010. On the anatomy of the basal ornithischian
dinosaur Eocursor parvus from the lower Elliot Formation
(Late Triassic) of South Africa. Zoological Journal of the
Linnean Society 160: 648–864.
Butler RJ, Zhao Q. 2009. The small-bodied ornithischian
dinosaurs Micropachycephalosaurus hongtuyanensis and
Wannanosaurus yansiensis from the Late Cretaceous of
China. Cretaceous Research 30: 63–77.
Butler RJ, Sullivan RM. 2009. The phylogenetic position of
the ornithischian dinosaur Stenopelix valdensis from the
Lower Cretaceous of Germany: implications for the early
fossil record of Pachycephalosauria. Acta Palaeontologica
Polonica 54: 21–34.
Butler RJ, Smith RMH, Norman DB. 2007. A primitive
ornithischian dinosaur from the Late Triassic of South
Africa, and the early evolution and diversification of Ornithischia. Proceedings of the Royal Society of London B 274:
2041–2046.
Butler RJ, Porro LB, Norman DB. 2008a. A juvenile skull
of the primitive ornithischian dinosaur Heterodontosaurus
tucki from the ‘Stormberg’ of southern Africa. Journal of
Vertebrate Paleontology 28: 702–711.
Butler RJ, Upchurch P, Norman DB. 2008b. The phylogeny of the ornithischian dinosaurs. Journal of Systematic
Palaeontology 6: 1–40.
Butler RJ, Galton PM, Porro LB, Chiappe LM, Henderson DM, Erickson GM. 2010. Lower limits of
ornithischian body size inferred from a new Upper
Jurassic
heterodontosaurid
from
North
America.
Proceedings of the Royal Society of London B 277: 375–
381.
Charig AJ, Crompton AW. 1974. The alleged synonymy of
Lycorhinus and Heterodontosaurus. Annals of the South
African Museum 64: 167–189.
Cleuren J, De Vree F. 2000. Feeding in crocodilians. In:
Schwenk K, ed. Feeding: form, function and evolution
in terrestrial vertebrates. San Diego: Academic Press, 337–
358.
Colbert EH. 1981. A primitive ornithischian dinosaur from
the Kayenta Formation of Arizona. Bulletin of the Museum
of Northern Arizona 53: 1–61.
Cooper MR. 1985. A revision of the ornithischian dinosaur
Kangnasaurus coetzeei Haughton, with a classification of
the Ornithischia. Annals of the South African Museum 95:
281–317.
Crompton AW. 1968. In search of the ‘insignificant’. Discovery. Yale Peabody Museum 3: 23–32.
255
Crompton AW, Attridge J. 1986. Masticatory apparatus of
the larger herbivores during the Late Triassic and Early
Jurassic times. In: Padian K, ed. The beginning of the Age
of Dinosaurs. New York: Cambridge University Press, 223–
236.
Crompton AW, Charig AJ. 1962. A new ornithischian
from the Upper Triassic of South Africa. Nature 196: 1074–
1077.
DeMar R. 1972. Evolutionary implications of Zahnreihen.
Evolution 26: 435–450.
Edmund AG. 1957. On the special foramina in the jaws of
many ornithischian dinosaurs. Contributions to the Life
Sciences Division of the Royal Ontario Museum, Canada 48:
1–14.
Edmund AG. 1960. Tooth replacement phenomena in the
lower vertebrates. Contributions to the Life Sciences Division of the Royal Ontario Museum, Canada 52: 1–190.
Edmund AG. 1969. Dentition. In: Gans C, ed. Biology
of the Reptilia. London & New York: Academic Press, 117–
200.
Evans DC, Ridgely R, Witmer LM. 2009. Endocranial
anatomy of lambeosaurine hadrosaurids (Dinosauria:
Ornithischia): a sensorineural perspective on cranial
crest function. The Anatomical Record 292: 1315–
1337.
Ewer RF. 1965. The anatomy of the thecodont reptile
Euparkeria capensis Broom. Philosophical Transactions of the Royal Society of London, Series B 248: 379–
435.
Fastnacht M. 2008. Tooth replacement pattern in
Coloborhynchus robustus (Pterosauria) from the Lower
Cretaceous of Brazil. Journal of Morphology 269: 332–
348.
Forster CA. 1990. The postcranial skeleton of the ornithopod
dinosaur Tenontosaurus tilletti. Journal of Vertebrate Paleontology 10: 273–294.
Galton PM. 1972. Classification and evolution of ornithopod
dinosaurs. Nature 239: 464–466.
Galton PM. 1973a. The cheeks of ornithischian dinosaurs.
Lethaia 6: 67–89.
Galton PM. 1973b. Redescription of the skull and mandible
of Parksosaurus from the Late Cretaceous with comments
on the family Hypsilophodontidae (Ornithischia). Life
Science Contributions, Royal Ontario Museum 89: 1–
21.
Galton PM. 1974. The ornithischian dinosaur Hypsilophodon
from the Wealden of the Isle of Wight. Bulletin of the British
Museum (Natural History) Geology 25: 1–152.
Galton PM. 1978. Fabrosauridae, the basal family of ornithischian dinosaurs (Reptilia: Ornithopoda). Paläontologische Zeitschrift 52: 138–159.
Galton PM. 1983. The cranial anatomy of Dryosaurus, a
hypsilophodontid dinosaur from the Upper Jurassic of
North America and East Africa, with a review of hypsilophodontids from the Upper Jurassic of North America. Geologica et Palaeontologica 17: 207–243.
Galton PM. 1984. Cranial anatomy of the prosauropod dinosaur Plateosaurus from the Knollenmergel (Middle Keuper,
256
D. B. NORMAN ET AL.
Upper Triassic) of Germany. I. Two complete skulls from
Trossingen/Württemburg with comments on the diet. Geologica et Palaeontologica 18: 139–171.
Galton PM. 1986. Herbivorous adaptations of Late Triassic
and Early Jurassic dinosaurs. In: Padian K, ed. The beginning of the Age of Dinosaurs: faunal change across the
Triassic-Jurassic boundary. Cambridge: Cambridge University Press, 203–221.
Galton PM. 2002. New material of the ornithischian (?heterodontosaurid) dinosaur Echinodon (Early Cretaceous,
southern England) from the Late Jurassic of Fruita near
Grand Junction, Colorado, USA. Journal of Vertebrate Paleontology 22 (Abstracts): 55A.
Galton PM. 2007. Teeth of ornithischian dinosaurs (mostly
Ornithopoda) form the Morrison Formation (Upper Jurassic) of the western United States. In: Carpenter K, ed.
Horns and beaks: ceratopsian and ornithopod dinosaurs.
Bloomington: Indiana University Press, 17–47.
Galton PM, Upchurch P. 2004a. Prosauropoda. In:
Weishampel DB, Dodson P, Osmólska H, eds. The Dinosauria, 2nd edn. Berkeley: University of California Press, 232–
258.
Galton PM, Upchurch P. 2004b. Stegosauria. In:
Weishampel DB, Dodson P, Osmólska H, eds. The Dinosauria, 2nd edn. Berkeley: University of California Press, 343–
362.
Gauthier J. 1986. Saurischian monophyly and the origin of
birds. Memoirs of the California Academy of Sciences 8:
1–55.
George JC, Berger AJ. 1966. Avian myology. London: Academic Press.
Ginsburg L. 1964. Découverte d’un scélidosaurien (dinosaure
ornithischien) dans le Trias supérieur du Basutoland.
Comptes Rendus hebdomadaires de l’Academie des Sciences,
Paris 258: 2366–2368.
Gorniak GC, Rosenberg HI, Gans C. 1982. Mastication in
the tuatara Sphenodon punctatus (Reptilia: Rhynchocephalia): structure and activity of the motor system. Journal of
Morphology 171: 321–353.
Gow CE. 1975. A new heterodontosaurid from the Red Beds
of South Africa showing clear evidence of tooth replacement.
Zoological Journal of the Linnean Society (London) 57:
335–339.
Gow CE. 1990. A tooth-bearing maxilla referable to Lycorhinus angustidens Haughton, 1924 (Dinosauria: Ornithischia). Annals of the South African Museum 99: 367–380.
Haas G. 1955. The jaw musculature of Protoceratops and in
other ceratopsians. American Museum Novitates 1729: 1–24.
Haas G. 1969. On the jaw muscles of ankylosaurs. American
Museum Novitates 2399: 1–11.
Haas G. 1973. Muscles of the jaws and associated structures
in the Rhychocephalia and Squamata. In: Gans C, ed.
Biology of the Reptilia, Vol. 3. London: Academic Press,
285–490.
Harris JD. 2004. Confusing dinosaurs with mammals: tetrapod phylogenetics and anatomical terminology in the world
of homology. The Anatomical Record Part A 281A: 1240–
1246.
Haubold H. 1990. Ein neuer Dinosaurier (Ornithischia,
Thyreophora) aus dem unteren Jura des nördlichen Mitteleuropa. Revue de Paléobiologie 9: 149–177.
Haughton SH. 1924. The fauna and stratigraphy of the
Stormberg Series. Annals of the South African Museum 12:
323–497.
He XL, Cai K. 1984. [The Middle Jurassic dinosaurian fauna
from Dashanpu, Zigong, Sichuan, Vol. 1, The ornithopod
dinosaurs]. Chengdu: Sichuan Scientific and Technological
Publishing House. [In Chinese with English summary].
Herring SW. 1972. Sutures – a tool in functional cranial
analysis. Acta Anatomica 83: 222–247.
Holliday CM. 2009. New insights into dinosaur jaw muscle
anatomy. The Anatomical Record 292: 1246–1265.
Holliday CM, Witmer LM. 2007. Archosaur adductor
chamber evolution: integration of musculoskeletal and topological criteria in jaw muscle homology. Journal of Morphology 268: 457–484.
Holliday CM, Witmer LM. 2008. Cranial kinesis in dinosaurs: intracranial joints, protractor muscles, and their significance for cranial evolution and function in diapsids.
Journal of Vertebrate Paleontology 28: 1073–1088.
Hopson JA. 1975. On the generic separation of the ornithischian dinosaurs Lycorhinus and Heterodontosaurus from
the Stormberg Series (Upper Triassic) of South Africa.
South African Journal of Science 71: 302–305.
Hopson JA. 1980. Tooth function and replacement in early
Mesozoic ornithischian dinosaurs: implications for aestivation. Lethaia 13: 93–105.
Huxley TH. 1869. On Hypsilophodon a new genus of Dinosauria. Abstracts of the Proceedings of the Geological Society,
London 204: 3–4.
Iordansky NN. 1964. The jaw muscles of crocodiles and some
related structures of the crocodilian skull. Anatomischer
Anzeiger 115: 256–280.
Iordansky NN. 1970. Structure and biomechanical analysis
of functions of the jaw muscles in the lizards. Anatomischer
Anzeiger 127: 282–413.
Iordansky NN. 1973. The skull of Crocodilia. In: Gans C, ed.
The biology of the Reptilia, Vol. 4. London: Academic Press,
201–262.
Irmis RB, Parker WG, Nesbitt SJ, Liu J. 2007. Early
ornithischian dinosaurs: the Triassic record. Historical
Biology 19: 3–22.
Jin L, Chen J, Zan S, Butler RJ, Godefroit P. 2010. Cranial
anatomy of the small ornithischian dinosaur Changchunsaurus parvus from the Quantou Formation (Cretaceous:
Aptian–Cenomanian) of Jilin Province, northeastern China.
Journal of Vertebrate Paleontology 30: 196–214.
Jourdan F, Féraud G, Bertrand H, Kampunzu AB, Tshoso
G, Watkeys MK. 2005. Karoo large igneous province:
brevity, origin, and relation to mass extinction questioned by
new 40Ar/39Ar age data. Geology 33: 745–748.
Jourdan F, Féraud G, Bertrand H, Watkeys MK. 2007.
From flood basalts to the inception of oceanization: example
from the 40Ar/39Ar high-resolution picture of the Karoo
large igneous province. Geochemistry, Geophysics, Geosystems 8: 1–20.
Jourdan F, Féraud G, Bertrand H, Watkeys MK, Renne
PR. 2008. The 40Ar/39Ar ages of the sill complex of
the Karoo large igneous province: implications for
climate change. Geochemistry, Geophysics, Geosystems 9:
1–20.
Kieser JA, Klapsidis C, Law L, Marion M. 1993. Heterodonty and patterns of tooth replacement in Crocodylus
niloticus. Journal of Morphology 218: 195–201.
Kitching JW, Raath MA. 1984. Fossils of the Elliot and
Clarens Formations (Karoo Sequence) of the Northeastern
Cape, Orange Free State and Lesotho, and a suggested
biozonation based on tetrapods. Palaeontologia Africana 25:
111–125.
Knoll F. 2002a. New skull of Lesothosaurus (Dinosauria:
Ornithischia) form the Upper Elliot Formation (Lower
Jurassic) of southern Africa. Géobios 35: 595–603.
Knoll F. 2002b. Nearly complete skull of Lesothosaurus
(Dinosauria: Ornithischia) from the Upper Elliot Formation
(Lower Jurassic: Hettangian) of Lesotho. Journal of Vertebrate Paleontology 22: 238–243.
Kuhn O. 1966. Die Reptilien, System und Stammesgeschichte.
Oeben: Krailling bei Munchen.
Lakjer T. 1926. Studien über die Trigeminus-versorgte Kaumuskulatur der Sauropsiden. Reitzel: Copenhagen.
Langer MC. 2003. The pelvic and hind limb anatomy of the
stem-sauropodomorph Saturnalia tupiniquim (Late Triassic, Brazil). PaleoBios 23: 1–40.
Langer MC. 2004. Basal Saurischia. In: Weishampel
DB, Dodson P, Osmólska H, eds. The Dinosauria, 2nd edn.
Berkeley: University of California Press, 25–46.
Langer MC, Benton MJ. 2006. Early dinosaurs: a phylogenetic study. Journal of Systematic Palaeontology 4: 309–
358.
Lull RS, Wright NE. 1942. Hadrosaurian dinosaurs of North
America. Geological Society of America. Special Paper:
1–242.
Luther A. 1914. Uber die vom N. Trigeminus versorgte
Muskulatur der Amphibien. Acta Societas Scientiarum
Fennica 44: 1–151.
Madison DR, Madison WP. 2003. MacClade: analysis of
phylogeny and character evolution, Version 4.06. Sunderland, MA: Sinauer Associates Inc.
Maidment SCR, Porro LB. 2010. Homology of the palpebral
and origin of the supraorbital ossifications in ornithischian
dinosaurs. Lethaia 43: 95–111.
Makovicky PJ. 2001. A Montanoceratops cerorhynchus
(Dinosauria: Ceratopsia) braincase from the Horseshoe
Canyon Formation of Alberta. In: Tanke DH, Carpenter K,
eds. Mesozoic vertebrate life. Bloomington: Indiana University Press, 243–262.
Maryańska T, Osmólska H. 1974. Results of the
Polish-Mongolian palaeontological expeditions. Part V.
Pachycephalosauria, a new suborder of ornithischian dinosaurs. Palaeontologia Polonica 30: 45–102.
Maryańska T, Osmólska H. 1984. Phylogenetic classification of ornithischian dinosaurs. In: Bogdanov NA, ed.
Abstracts of 27th international geological congress. Moscow:
Nauka Press, 286–287.
257
Maryańska T, Osmólska H. 1985. On ornithischian phylogeny. Acta Palaeontologica Polonica 30: 137–150.
Maryańska T, Chapman RE, Weishampel DB. 2004.
Pachycephalosauria. In: Weishampel DB, Dodson P, Osmólska H, eds. The Dinosauria, 2nd edn. Berkeley: University
of California Press, 464–477.
Nesbitt SJ, Smith ND, Irmis RB, Turner AH, Downs A,
Norell MA. 2009. A complete skeleton of a Late Triassic
saurischian and the early evolution of dinosaurs. Science
326: 1530–1533.
Nicholls EL, Russell AP. 1985. Structure and function of
the pectoral girdle of Struthiomimus altus (Theropoda:
Ornithomimidae). Palaeontology 28: 643–677.
Norman DB. 1977. On the anatomy of the ornithischian
dinosaur Iguanodon. Unpublished PhD dissertation, King’s
College London, UK.
Norman DB. 1980. On the ornithischian dinosaur Iguanodon
bernissartensis from Belgium. Mémoires de l’Institut Royal
des Sciences Naturelles de Belgique 178: 1–105.
Norman DB. 1984a. A systematic reappraisal of the reptile
order Ornithischia. In: Reif W-E, Westphal F, eds. Proceedings of the third symposium on Mesozoic terrestrial ecosystems. Tübingen: Attempto Verlag, 157–162.
Norman DB. 1984b. On the cranial morphology and evolution of ornithopod dinosaurs. Symposia of the Zoological
Society of London 52: 521–547.
Norman DB. 1985. The illustrated encyclopedia of dinosaurs.
London: Salamander Books.
Norman DB. 1986. On the anatomy of Iguanodon atherfieldensis (Ornithischia: Ornithopoda). Bulletin de l’Institut
Royal des Sciences Naturelles de Belgique 56: 281–372.
Norman DB. 2004. Basal Iguanodontia. In: Weishampel DB,
Dodson P, Osmólska H, eds. The Dinosauria, 2nd edn.
Norman DB, Barrett PM. 2002. Ornithischian dinosaurs
from the Lower Cretaceous (Berriasian) of England. Special
Papers in Palaeontology, Palaeontological Association 68:
161–189.
Norman DB, Weishampel DB. 1985. Ornithopod feeding
mechanisms: their bearing on the evolution of herbivory.
American Naturalist 126: 151–164.
Norman DB, Weishampel DB. 1991. Feeding mechanisms in
some small herbivorous dinosaurs: processes and
patterns. In: Rayner JMV, Wootton RJ, eds. Biomechanics
in evolution. Cambridge: Cambridge University Press, 161–
181.
Norman DB, Witmer LM, Weishampel DB. 2004a. Basal
Ornithischia. In: Weishampel DB, Dodson P, Osmólska H,
eds. The Dinosauria, 2nd edn. Berkeley: University of California Press, 325–334.
Norman DB, Witmer LM, Weishampel DB. 2004b. Basal
Thyreophora. In: Weishampel DB, Dodson P, Osmólska H,
eds. The Dinosauria, 2nd edn. Berkeley: University of California Press, 335–342.
Norman DB, Sues H-D, Witmer LM, Coria RA. 2004c.
Basal Ornithopoda. In: Weishampel DB, Dodson P, Osmólska H, eds. The Dinosauria, 2nd edn. Berkeley: University
of California Press, 393–412.
258
D. B. NORMAN ET AL.
Novas FE. 1993. New information on the systematics and
postcranial skeleton of Herrerasaurus ischigualastensis
(Theropoda: Herrerasauridae) from the Ischigualasto Formation (Upper Triassic) of Argentina. Journal of Vertebrate
Novas FE. 1996. Dinosaur monophyly. Journal of Vertebrate
Novas FE, Cambiaso AV, Ambrosio A. 2004. A new basal
iguanodontian (Dinosauria, Ornithischia) from the Upper
Cretaceous of Patagonia. Ameghiniana 41: 75–82.
Olsen PE, Baird D. 1986. The ichnogenus Atreipus and its
significance for Triassic biostratigraphy. In: Padian K, ed.
The beginning of the Age of Dinosaurs. Cambridge: Cambridge University Press, 61–87.
Olsen PE, Galton PM. 1984. A review of the reptile and
amphibian assemblages from the Stormberg of southern
Africa, with special emphasis on the footprints and the age
of the Stormberg. Palaeontologia Africana 25: 87–110.
Olshevsky G. 1991. A revision of the Parainfraclass Archosauria Cope, 1869, excluding the advanced Crocodylia,
Mesozoic Meanderings, no 2. San Diego: Publications
Requiring Research.
Osborn JW. 1970. New approach to Zahnreihen. Nature 225:
343–346.
Osborn JW. 1971. The ontogeny of tooth succession in
Lacerta vivipara Jacquin (1787). Proceedings of the Royal
Society of London B 179: 261–289.
Osborn JW. 1975. Tooth replacement: efficiency, patterns and
evolution. Evolution 29: 180–186.
Ostrom JH. 1961. Cranial morphology of the hadrosaurian
dinosaurs of North America. Bulletin of the American
Museum of Natural History 122: 33–186.
Ostrom JH. 1964. A functional analysis of jaw mechanics in
the dinosaur Triceratops. Postilla 88: 1–35.
Owen R. 1842. Report on British Fossil Reptiles. Part 2.
Report of the British Association for the Advancement of
Science (Plymouth) XI: 60–204.
Owen R. 1861. Monograph of the Fossil Reptilia of the Liassic
Formations. Part 1. A monograph of a fossil dinosaur (Scelidosaurus harrisonii Owen) of the Lower Lias. Palaeontographical Society Monographs XII: 1–14.
Paul GS. 1988. Predatory dinosaurs of the world. New York:
Simon and Schuster.
Peng G. 1990. [A new species of small ornithopod from
Zigong, Sichuan]. Journal of the Zigong Dinosaur Museum
2: 19–27. [In Chinese].
Peng G. 1992. [Jurassic Ornithopod Agilisaurus louderbacki
(Ornithopoda: Fabrosauridae) from Zigong, Sichuan,
China]. Vertebrata PalAsiatica 30: 39–51. [In Chinese with
English summary].
Perle A, Maryańska T, Osmólska H. 1982. Goyocephale
lattimorei gen. et sp. n., a new flat-headed pachycephalosaur (Ornithischia, Dinosauria) from the Upper Cretaceous of Mongolia. Acta Palaeontologica Polonica 27: 115–
127.
Porro LB. 2007. Feeding and jaw mechanisms in Heterodontosaurus tucki using finite element analysis. Journal of
Vertebrate Paleontology 27: 131A.
Porro LB. 2009. Cranial biomechanics in the early dinosaur
Heterodontosaurus. Unpublished PhD thesis, University of
Cambridge, UK.
Porro LB, Butler RJ, Barrett PM, Moore-Fay S, Abel RL
(in press). New heterodontosaurid specimens from
the Lower Jurassic of southern Africa and the early
ornithischian dinosaur radiation. Earth and Environmental Science Transactions of the Royal Society of
Edinburgh.
Rauhut OWM. 2003. The interrelationships and evolution of
basal theropod dinosaurs. Special Papers in Palaeontology
69: 1–215.
Rayfield EJ. 2001. Cranial form and function in a large
theropod dinosaur: a study using Finite Element Analysis.
Unpublished PhD thesis, University of Cambridge, UK.
Reig OA. 1963. La presencia de dinosaurios saurisquios en
los ‘Estrados de Ischigualasto’ (Mesotriásico superior) de las
Provincias de San Juan y La Rioja (Republica Argentina).
Ameghiniana 3: 3–20.
Rensberger JM. 1973. An occlusion model for mastication
and dental wear in some herbivorous mammals. Journal of
Rensberger JM. 1975. Function in the cheek tooth evolution
of some hypsodont geomyoid rodents. Journal of Paleontology 49: 10–22.
Romer AS. 1956. Osteology of the reptiles. Chicago: University of Chicago Press.
Rybczynski N, Tirabasso A, Bloskie P, Cuthbertson R,
Holliday CM. 2008. A three-dimensional animation
model of Edmontosaurus (Hadrosauridae) for testing
chewing hypotheses. Palaeontologia Electronica 11: 14.
Santa Luca AP. 1980. The postcranial skeleton of Heterodontosaurus tucki (Reptilia: Ornithischia) from the Stormberg of South Africa. Annals of the South African Museum
79: 159–211.
Santa Luca AP, Crompton AW, Charig AJ. 1976. A complete skeleton of the Late Triassic ornithischian Heterodontosaurus tucki. Nature 264: 324–328.
Scheetz RD. 1999. Osteology of Orodromeus makelai and the
phylogeny of basal ornithopod dinosaurs. PhD dissertation,
Montana State University, Bozeman, USA.
Schumacher GH. 1973. The head muscles and hyolaryngeal
skeleton of turtles and crocodilians. In: Gans C, ed. The
biology of the Reptilia, Vol. 4. London: Academic Press,
101–199.
Seeley HG. 1887. On the classification of the fossil animals
commonly named Dinosauria. Proceedings of the Royal
Society of London 43: 165–171.
Sereno PC. 1984. The phylogeny of the Ornithischia: a reappraisal. In: Reif W-E and Westphal F, eds. Proceedings of the
Third Symposium on Mesozoic Terrestrial Ecosystems
Tübingen: Attempto Verlag, 219–226.
Sereno PC. 1986. Phylogeny of the bird-hipped dinosaurs.
National Geographic Research 2: 234–256.
Sereno PC. 1987. The ornithischian dinosaur Psittacosaurus
from the Lower Cretaceous of Asia and the relationships of
the Ceratopsia. PhD dissertation, Columbia University,
New York, USA.
Sereno PC. 1990. Psittacosauridae. In: Weishampel
DB, Dodson P, Osmólska H, eds. The Dinosauria, 1st edn.
Sereno PC. 1991a. Lesothosaurus, ‘Fabrosaurids’ and the
early evolution of Ornithischia. Journal of Vertebrate Paleontology 11: 234–256.
Sereno PC. 1991b. Basal archosaurs: phylogenetic relationships and functional implications. Journal of Vertebrate
Paleontology 11 (Memoir 2): 53.
Sereno PC. 1993. The pectoral girdle and forelimb of the
basal theropod Herrerasaurus ischigualastensis. Journal of
Sereno PC. 1999. The evolution of dinosaurs. Science 284:
2137–2147.
Sereno PC. 2000. The fossil record, systematics and evolution of pachycephalosaurs and ceratopsians from Asia. In:
Benton MJ, Shishkin MA, Unwin DM, Kurochkin EN, eds.
The Age of Dinosaurs in Russia and Mongolia. Cambridge:
Cambridge University Press, 480–516.
Sereno PC. 2007. The phylogenetic relationships of early
dinosaurs: a comparative report. Historical Biology 19: 145–
155.
Sereno PC, Dong Z-M. 1992. The skull of the basal stegosaur Huayangosaurus taibaii and a cladistic diagnosis of
Stegosauria. Journal of Vertebrate Paleontology 12: 318–
343.
Sereno PC, Novas FE. 1993. The skull and neck of the basal
theropod Herrerasaurus ischigualastensis. Journal of Vertebrate Paleontology 13: 451–476.
Sereno PC, Chao S, Cheng Z, Rao C. 1988. Psittacosaurus meileyingensis (Ornithischia; Ceratopsia) a new
psittacosaur from the Lower Cretaceous of Northeastern
China. Journal of Vertebrate Paleontology 8: 366–377.
Sereno PC, Xijin Z, Lin T. 2010. A new psittacosaur from
Iner Mongolia and the parrot-like structure and function of
the psittacosaur skull. Proceedings of the Royal Society of
London B 277: 199–209.
Smith JB. 1997. Heterodontosauridae. In: Currie PJ, Padian
K, eds. Encyclopaedia of dinosaurs. San Diego: Academic
Press, 317–320.
Smith KK. 1982. The electromyographic study of the function
of the jaw adducting muscles in Varanus exanthematicus (Varanidae). Journal of Morphology 173: 137–
158.
Smith RMH. 1990. A review of stratigraphy and sedimentary
environments of the Karoo Basin of South Africa. Journal of
African Earth Science 10: 117–137.
Steel R. 1969. Ornithischia. Handbuch der Paläoherpetologie,
Vol. 15. Stuttgart: Gustav Fischer Verlag.
Sues H-D. 1980. Anatomy and relationships of a new hypsilophodontid dinosaur from the Lower Cretaceous of North
America. Palaeontographica, Abteilung A 169: 51–72.
Sues H-D, Galton PM. 1987. Anatomy and classification of
the North American Pachycephalosauria (Dinosauria: Ornithischia). Palaeontographica, Abteilung A 198: 1–40.
Swofford DL. 2002. PAUP* (phylogenetic analysis using parsimony (*and other methods), Version 4.0b10. Sunderland,
MA: Sinauer Associates Inc.
259
Throckmorton GS. 1976. Oral food processing in two herbivorous lizards, Iguana iguana (Iguanidae) and Uromastix
aegyptius (Agamidae). Journal of Morphology 48: 363–
390.
Thulborn RA. 1970a. The skull of Fabrosaurus australis, a
Triassic ornithischian dinosaur. Palaeontology 13: 414–
432.
Thulborn RA. 1970b. The systematic position of the Triassic
ornithischian dinosaur Lycorhinus angustidens. Zoological
Journal of the Linnean Society (London) 49: 235–245.
Thulborn RA. 1971a. Tooth wear and jaw action in the
Triassic ornithischian dinosaur Fabrosaurus. Journal of
Zoology 164: 165–179.
Thulborn RA. 1971b. Origin and evolution of ornithischian
dinosaurs. Nature 234: 75–78.
Thulborn RA. 1972. The postcranial skeleton of the Triassic
ornithischian dinosaur Fabrosaurus australis. Palaeontology 15: 29–60.
Thulborn RA. 1974. A new heterodontosaurid dinosaur (Reptilia: Ornithischia) from the Upper Triassic Red Beds of
Lesotho. Zoological Journal of the Linnean Society (London)
55: 151–175.
Thulborn RA. 1978. Aestivation among ornithopod dinosaurs
of the African Trias. Lethaia 11: 185–198.
Thulborn RA. 1992. Taxonomic characters of Fabrosaurus
australis, an ornithischian dinosaur from the Lower Jurassic of southern Africa. Geobios 25: 283–292.
Tykoski RS, Rowe T. 2004. Ceratosauria. In: Weishampel
DB, Dodson P, Osmolska H, eds. The Dinosauria, 2nd edn.
Berkeley: California University Press, 47–70.
Van Drongelen W, Dullemeijer P. 1982. The feeding apparatus of Caiman crocodilus: a functional-morphological
study. Anatomischer Anzeiger 151: 337–366.
Vanden Berge JC, Zweers GA. 1993. Myologia. In: Baumel
JJ, King AS, Breazile JE, Evans HE, Vanden Berge JC, eds.
Avaian anatomy: Nomina anatomica Avium. Cambridge,
MA: Nuttall Ornithological Club, 189–247.
Weishampel DB. 1984. The evolution of jaw mechanisms in
ornithopod dinosaurs. Advances in Anatomy, Embryology
and Cell Biology 87: 1–110.
Weishampel DB. 1990. Ornithopoda. In: Weishampel DB,
Dodson P, Osmólska H, eds. The dinosauria, 1st edn.
Weishampel DB, Heinrich RE. 1992. Systematics of Hypsilophodontidae and basal Iguanodontia (Dinosauria: Ornithopoda). Historical Biology 6: 159–184.
Weishampel DB, Norman DB. 1989. Vertebrate herbivory
in the Mesozoic: jaws, plants and evolutionary metrics. In:
Farlow JO, ed. Paleobiology of dinosaurs. Boulder: Geological Society of America, Special Paper, 87–100.
Weishampel DB, Witmer LM. 1990. Heterodontosauridae.
In: Weishampel DB, Dodson P, Osmolska H, eds. The Dinosauria. Berkeley: University of California Press, 486–
497.
Weishampel DB, Jianu CM, Csiki Z, Norman DB. 2003.
Osteology and phylogeny of Zalmoxes (n.g.), an unusual
euornithopod dinosaur from the latest Cretaceous of
Romania. Journal of Systematic Palaeontology 1: 65–123.
260
D. B. NORMAN ET AL.
Welles SP. 1984. Dilophosaurus wetherilli (Dinosauria,
Theropoda), osteology and comparisons. Palaeontographica,
Abteilung A 185: 85–180.
Witmer LM. 1995. The extant phylogenetic bracket and the
importance of reconstructing soft tissues in fossils. In: Thomason JJ, ed. Functional morphology in vertebrate paleontology. New York: Cambridge University Press, 19–33.
Witmer LM. 1997. The evolution of the antorbital cavity of
archosaurs: a study in soft-tissue reconstruction in the fossil
record with an analysis of the function of pneumaticity.
Journal of Vertebrate Paleontology 17: 73.
Witmer LM. 2009. Dinosaurs: fuzzy origins for feathers.
Nature 458: 295–295.
Woerderman MW. 1919. Beitrage zur Entwichlungsgeschichte von Zahnen und Gebiss der Reptiien I-III. Archiv
für mikroskopische Anatomie und Entwicklungsgeschichte
92: 105–244.
Xu X, Makovicky PJ, Wang XL, Norell MA, You H-L.
2002. A ceratopsian dinosaur from China and the early
evolution of Ceratopsia. Nature 416: 314–317.
Xu X, Forster CA, Clark JM, Mo J. 2006. A basal ceratopsian with transitional freatures from the Late Jurassic of
northwestern China. Proceedings of the Royal Society of
London B 273: 2135–2140.
Yates AM. 2003. A new species of the primitive dinosaur
Thecodontosaurus (Saurischia: Sauropodomorpha) and its
implications for the systematics of early dinosaurs. Journal
of Systematic Palaeontology 1: 1–42.
Yates AM, Hancox PJ, Rubidge BS. 2004. First record of a
sauropod dinosaur from the upper Elliot Formation (Early
Jurassic) of South Africa. South African Journal of Science
100: 504–506.
You H-L, Dodson P. 2003. Redescription of neoceratopsian dinosaur Archaeoceratops and early evolution of
Neoceratopsia. Acta Palaeontologica Polonica 48: 261–272.
You H-L, Dodson P. 2004. Basal Ceratopsia. In: Weishampel
DB, Dodson P, Osmólska H, eds. The Dinosauria, 2nd
edn. Berkeley: University of California Press, 478–493.
You H-L, Xu X, Wang X-L. 2003. A new genus of Psittacosauridae (Dinosauria: Ornithopoda) and the Origin and
Early Evolution of Marginocephalian Dinosaurs. Acta Geologica Sinica 77: 15–20.
Zhao X-J, Cheng Z-W, Xu X. 1999. The earliest ceratopsian
from Tuchengzi Formation of Liaoning, China. Journal of
Zhao X-J, Cheng Z, Xu X, Makovicky PJ. 2006. A
new ceratopsian from the Upper Jurassic Houcheng
Formation of Hebei, China. Acta Geologica Sinica 80: 467–
473.
Zheng X-T, You H-L, Xu X, Dong Z-M. 2009. An Early
Cretaceous heterodontosaurid dinosaur with integumentary
structures. Nature 458: 333–336.
APPENDIX 1
DATA MATRIX
The character-score matrix (generated using MacClade 4, Madison & Madison, 2003) used in these analyses.
Taxon number reference (see listing below)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
–
–
–
–
–
–
–
–
–
–
–
–
–
–
Euparkeria
Herrerasaurus
Abrictosaurus
Heterodontosaurus
NHMUK RU A100 (APPEARS IN FIGURES 40 & 41 AS ‘BMNH A100’)
Lesothosaurus
Emausaurus
Hypsilophodon
Scelidosaurus
Ankylopollexia
Psittacosauridae
Pachycephalosauridae
Agilisaurus
Hexinlusaurus
261
APPENDIX 1 Continued
Taxon numbers:
11111
12345678901234
Abbreviated characters (see Appendix 2)
Skull props
Skull length
Rostral bone
Rostral keeled
Rostral vent lat pr
Premaxillary teeth
Pmx post lat pr
Oral margin of pmx
Pmx oral margin ht
Pmx foramen
Pmx palate arching
Pmx dor pr-nasal
Fossa at pmx-ma
Pmx-Mx diastema
Diastema arched
Posnof ventral margin of narial fossa
Posn of ventral margin of ext naris
External naris size
Longit internarial cleft
Int antorb fenestrasize
Ext antorb fenestra present
Ext antorb fenestra shape
Additional openings at ant of aof
Maxilla ant lat boss cont pmx
Mx accessory ant process
Mx buccal recess
Prominent ridge above buccal recess
Slot in Mx for lacrimal
Accessory orbital ossifications
Form of palpebral – free, bound in
Palpebral shape in dorsal view
Palpebral number
Palpebral length across orbit
Jugal excluded from aof
Props of ant jugal ramus
Skull width across jugals
Position of max width of skull
Jugal/epijugal ridge
Epijugal abs/pres
Jugal boss abs/pres
Nodular ornament of jugal
Jugal-Porb bar width
Jugal-Port joint
Jugal form of postorb process
Jugal-Sq contact above itf
Jugal post ramus forked
Jugal post ramus detail
Jugal-QJ contact
Postorbital margin (orb projection)
Postorbital T-shaped or not
Porb-Par contact abs – broad
QJ-Sq contact
0000?00000100?
0000?0?0000000
00000000001000
----------0------------0--001111?111111?
0001?000011000
0000000001000?
0011?001?1010?
0011?1?1?0111?
00?1111111011?
00?1?00000000?
0011100100000?
001110010--10?
--111--0--10-?
01000000?0100?
01000000?0100?
0000?000?10000
01?1???1000011
00111111111111
0000000000?100
0100000011?-00
00?1?001000-00
00?00000000???
0000000001000?
00011111111111
00011000001100
000001000000??
0011?111111111
--00?000100100
--00?010-01-10
--00?000200110
--00?000-00-10
0000?00100?-01
00?0?010102000
00000000001000
--?-------0--00?00000001000
00?0?000000000
00?1?000000000
00?00000000100
00?00000100100
00?0?000000100
00?00000000000
00?00000000000
01?0?110101000
00?0?000010000
00?0?000000000
00?0?000000000
00?0?000000-00
00?0?000000100
00?0?001011100
262
D. B. NORMAN ET AL.
Taxon numbers:
11111
12345678901234
QJ shape
QJ ventral margin close to jaw jnt
QJ orientated laterally vs postlat
QJ transverse width
Oval fossa on pt ramus of Q
Q lateral wing pres/abs
Q shaft convex in lat vw or straight
Paraq fenestra small vs large
Paraq fen/notch orientation postlat vs lat
Paraq fen position quad vs qj
Q articular surfaces (eq, med, lat)
Frontal proportions
Stf open vs closed
Stf proportions (elong vs ovoid)
Parietal septum
Parietosquam shelf abs/pres
Parietosquam frill
Composition of parsquam shelf
Postorb-Sq bar shape
Postorb-Sq tubercles
Tubercles on Sq enlarged
Frontoparietal thickness
Paroccipital processes shape plate vs hook
Paroccipital pr shape
Posttemp fossa position
Supraocciptial in fm
Bocc in fm
Bsph shape: elongate vs short
Prootic-Bsph sheet
Basal tubera shape
Bpt procs orientation
Pmx-vomer contact
Deep palatal keel in midline
Pterygovomerine keel props
Pt-Mx contact
Pt-Quad rami project ventrally
Cortical remodelling of skull bones
Predentary: abs/pres
Predent size
Predent rostrum: broad vs pointed
Predent oral margin smooth/dent
Predentary tip upturned
Predent ventral process (shape)
Predent ventral process (pres/red)
Dent symphysis: v or spout
Dent tooth row str – downturned ant
Dors vent dent margins converge or not
Ventral flange on dentary
Coronoid process prominence
Ant dorsal edge of cor proc formed by
Coronoid position rel to dentition
Ext mand fenestra
00?-?111111100
00?0?000010000
0000?000000000
00?0?000000000
00?0?0?0000000
00?0?000000000
00?0?000000000
00?0?001000000
00?0?001010000
00?0?001000000
11?2?0?010201?
00?1?001000000
00?00000000100
00?00000000-00
00?00000000100
00?0?000001100
--?-?-----00---?-?-----10-0?0?0000001000
00?0?000000100
00?0?000000100
00?00000000100
00?0?0?1010111
00?0?0?000000?
11?1?0?0?10???
00?0?0?001000?
00?0?00000000?
0??0?00010000?
00?1?0?000010?
00?0?00000010?
00?0?0011200??
0??0???0?011??
00?0?0?010000?
--?-?-?-0----?
00?0?0?000100?
00?0?0?000010?
00000010100000
0011?1?1?1111?
--11?00101100?
--01?1?1?01???
--00?0?0?10?0?
--00?0?0?00?0?
----?0?0?10?0?
--11?0?0?00?0?
00000111111011
00000010100000
00100000110000
00000000001000
0011?001011111
0011?111111111
0000?000011000
0000?00111011?
263
Taxon numbers:
11111
12345678901234
Small fenestra on surang-dent joint
Ridge on lateral surf surangular
Retro artic proc
Nodular ornament on dentary
Level of jaw joint
Mandibular osteoderm
Pmx teeth pres/abs
Pmx tooth number
Pmx tooth shape
Pmx tooth size change
Mx – Dent tooth shape
Mx-Dent marginal denticles
Enamel coating on mx – dent crowns
Apico-basal ridging
Ridging confluent with denticles
Prominent ridge on lateral Mx crowns
Prominent ridge on medial Dent crowns
Position of primary ridges symm/offset
Expansion of crowns
Heterodonty
Peg-like tooth located ant of dent
Special foramina pres/abs (medially)
Curvature of max/dent teeth
Imbricate pattern of crowns
Crown expanded above root
Max tooth size in rows
Close packing of teeth
Fusion of intercent-axis
Epipophyses present/abs
Cervical artic surfaces
Cervical number
Cervical artic at zygs
Dorsal number
Sacral number
Sacrum artic with pubis
Sacral ribs elongate posteriorly
Ant caudal centra props
Caudal n. spines
Elongate tail
Chevron shape
Sternal seg of dorsal ribs ossified
Gastralia
Clavicles ossified
Sternal plate shape
Humerus-Scap proportions
Scapula blade props
Scap acromion shape
Scap blade shape
Humerus length
Dpc development
Humerus shaft form str vs bowed
Manus length
0000?00101000?
00?0?11010000?
0000?00000000?
0000?000000100
00?1?001010000
00000000100000
0000000001100?
223330111--31?
000001111--01?
000110000--00?
00111111121111
00111111111111
00?1?001011?00
00010001011001
---0---0-010-0
00010000011000
00?1?001011000
--?0?--0-00--11010000010000
00011000000?2?
00001000000?00
11?01000000000
00111111111011
00111111111111
01011111111111
00111111111111
00000001010000
00???0?0000?00
10?0?0?1011?11
00?0?0?0010?00
02?1???1000?11
00???0?0000100
11?0???1220?11
0023?2?3203322
00???0?10?1?00
0000???0000100
00?0???0000100
00?0???001??00
00?0???0000?00
00?0???0000000
00?0???1000000
???1???1111111
1??0???0001?00
0??????1?112??
0?00?0?0000011
01?1?0?0000100
1111?1?1101111
1100?0?0000000
0000?1?0000110
0000?0?0000100
0000?0?0000100
?1?1?0?0?00??0
264
D. B. NORMAN ET AL.
Taxon numbers:
11111
12345678901234
Metacarpals block like prox ends?
Mc 1 & 5 shorter
Penult phal of manus short or long
Manus digit 3: phal count
Length of prox phals of digits 2–4
Extensor pits present
Manus unguals recurved and narrow
Acetab closed vs perforate?
Preacetab process shape
Preacetab proc length
Preacetab lateral deflected
Dors margin of peacet and ilium shape
Preacet expansion
Dorsal margin of ilium
Subtriangular process of ilium
Subtriangular process shape
Brevis fossa/shelf
Length of postacet process
medioventr acetab flange on ilium
Supracetabular crest
Ischial peduncle of ilium
Pubic peduncle
Pubic process of ischium
Ischium shaft shape
Ischial shaft X sect
Ischial shaft expansion
Groove on dorsal margin of ischium
Tab-shaped obt process
Ischial symphysis
Pubis orientation
Shaft of pubis X section
Shaft of post pubis length
Red of postpubic shaft
Body of pubis size
Body of pubis massive
Prepubic process abs/pres
Prepubic process: shape
Prepubic process: length
Prepubic rel to ant process of ilium
Extent of pubic symphsis
Femur shape in med/lat view
Femoral head shape
Ant trochanter morphology
Level of top edge of ant trochanter
Fourth trochanter shape
Fourth trochanter position
Anterior inter cond groove
Post int cond groove
Lat condyle of distal femur
Distal tibia shape
Fibular facet shape
Calcaneum proximal surface shape
0111?000?00??0
0000?010?00??0
?1?1?0?0?00??0
?0?0???0?10??0
?000?0?0?10??0
0111?0?0?00??0
?1?1?0?0?00??0
11?0?0?0000000
0011?1?1111111
00?1?1?1111100
00?0?0?0000100
0000?0?0100100
0000?0?0000100
0000?0?0000000
0000?0?0000100
----?-?----1--011?0?1011101
2111?1?2012211
00?1?0?1011100
10?1?0?1011101
0000?0?1011100
0000?0?1011111
00?0?0?0010100
00?0?0?0000100
01?0?0?0010000
00?0?0?0010000
00?0?1?0000010
00?0?0?1010011
00???0?1111101
00?1?1?1111111
00?1?1?1111111
00?0?0?0011100
--?-?-?--011-00?0?0?0000100
00?0?0?0100000
00?1?1?1111111
--?0?0?1001101
--?0?0?1011111
--?0?0?1010101
00?1?1?1111111
0000?0?0010000
00?0?0?1011100
0122?2?3233322
-0?1?0?1011100
0122?2?2222122
0000?0?0110000
0000?0?0011?00
00?0?0?0010000
00?0?0?0010?00
0011?1?1111?11
00???1?1111?11
00???1?1111?11
265
Taxon numbers:
11111
12345678901234
Medial distal tarsal
Metatarsal arrangement
Digit 1 development
Pes digit 4 phalangeal count
Metatarsal 5 length
Metatarsal 5: digits?
Pedal ungual shape
Epaxial OTs
Hypaxial OTs
OT arrangement
Parasagittal osteoderms
Lateral row of keeled osteoderms
U-shaped cervical/pectoral collars
00?0?0?1000?01
0000?0?0000?00
0011?1?0000?11
0000???0000?00
0011?1?1111?11
0011?1?1111?11
0000?010100?00
00?1?1?1111111
00?0?0?1000100
--?0?0?1020100
1000?010100000
?000?010100000
0000?0?0100000
APPENDIX 2
CHARACTER
LISTING
After: Butler et al. (2008b). In this analysis of 221
characters, 125 were informative, 65 were uninformative (autapomorphic), and 31 were constant (had no
influence on topology).
Bold identifies the ‘analytically informative’
characters.
Italics identify the ‘constant’ characters.
1. Skull proportions: 0, preorbital skull length
more than 45% of basal skull length; 1, preorbital length less than 40% of basal skull
length.
2. Skull length (rostrum–quadrate): 0, 15%
or less of body length; 1, 20–30% of body
length.
3. Neomorphic rostral bone, anterior to premaxilla:
0, absent; 1, present.
4. Rostral bone, anteriorly keeled and ventrally
pointed: 0, absent; 1, present.
5. Rostral bone, ventrolateral processes: 0, rudimentary; 1, well developed.
6. Premaxilla, edentulous anterior region: 0,
absent, first premaxillary tooth is positioned adjacent to the symphysis; 1,
present, first premaxillary tooth is inset
the width of one or more crowns.
7. Premaxilla,
posterolateral
process,
length: 0, does not contact lacrimal; 1, contacts the lacrimal, excludes maxilla–nasal
contact.
8. Oral margin of the premaxilla: 0, narial portion
of the body of the premaxilla slopes steeply from
the external naris to the oral margin; 1, ventral
premaxilla flares laterally to form a partial floor
of the narial fossa.
9. Position of the ventral (oral) margin of the
premaxilla: 0, level with the maxillary
tooth row; 1, deflected ventral to maxillary
tooth row.
10. Premaxillary foramen: 0, absent; 1,
present.
11. Premaxillary palate: 0, strongly arched,
forming a deep, concave palate; 1, horizontal or only gently arched.
12. Overlap of the dorsal process of the premaxilla
onto the nasal: 0, present; 1, absent.
13. Fossa-like depression positioned on the
premaxilla– maxilla boundary: 0, absent; 1,
present.
14. Premaxilla–maxilla diastema: 0, absent,
maxillary teeth continue to anterior end of
maxilla; 1, present, substantial diastema of
at least one crown’s length between maxillary and premaxillary teeth.
15. Form of diastema; 0, flat; 1, arched ‘subnarial gap’ between the premaxilla and
maxilla.
16. Narial fossa surrounding external nares on
lateral surface of premaxilla, position of
ventral margin of fossa relative to the
ventral margin of the premaxilla: 0, closely
approaches the ventral margin of the
266
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
D. B. NORMAN ET AL.
premaxilla; 1, separated by a broad, flat
margin from the ventral margin of the premaxilla.
External nares, position of the ventral
margin: 0, below the ventral margin of the
orbits; 1, above the ventral margin of the
orbits.
External naris size: 0, small, entirely overlies
the premaxilla; 1, enlarged, extends posteriorly
to overlie the maxilla.
Deep elliptic fossa present along sutural
line of the nasals: 0, absent; 1, present.
Internal antorbital fenestra size: 0, large,
generally at least 15% of the skull length; 1,
very much reduced, less than 10% of skull
length, or absent.
External antorbital fenestra: 0, present; 1,
absent.
External antorbital fenestra, shape: 0, triangular; 1, oval or circular.
Additional opening(s) anteriorly within
the antorbital fossa: 0, absent; 1, present.
Maxilla, prominent anterolateral boss articulates with the medial premaxilla: 0, absent; 1,
present.
Maxilla, accessory anterior process: 0, absent; 1,
present.
Maxilla, buccal emargination: 0, absent; 1,
present.
Eminence on the rim of the buccal emargination of the maxilla near the junction
with the jugal: 0, absent; 1, present.
Slot in maxilla for lacrimal: 0, absent; 1,
present.
Accessory ossification(s) in the orbit
(palpebral/supraorbital): 0, absent; 1,
present.
Palpebral/supraorbital: 0, free, projects
into orbit from contact with lacrimal/
prefrontal; 1, incorporated into orbital
margin.
Palpebral, shape in dorsal view: 0, rodshaped; 1, plate-like with wide base.
Palpebral/supraorbital, number: 0, one; 1,
two; 2, three.
Free palpebral, length, relative to anteroposterior width of orbit: 0, does not traverse
entire width of orbit; 1, traverses entire width of
orbit.
Exclusion of the jugal from the posteroventral margin of the external antorbital
fenestra by lacrimal–maxilla contact: 0,
absent; 1, present.
Anterior ramus of jugal, proportions: 0,
deeper than wide, but not as deep as the
posterior ramus of the jugal; 1, wider than
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
49.
50.
51.
52.
53.
54.
55.
deep; 2, deeper than the posterior ramus of
the jugal.
Widening of the skull across the jugals, chord
from frontal orbital margin to extremity of jugal
is more than minimum interorbital width: 0,
absent; 1, present, skull has a triangular shape
in dorsal view.
Position of maximum widening of the skull: 0,
beneath the jugal–postorbital bar; 1, posteriorly,
beneath the infratemporal fenestra.
Jugal (or jugal–epijugal) ridge dividing the
lateral surface of the jugal into two planes: 0,
absent; 1, present.
Epijugal: 0, absent; 1, present.
Jugal boss: 0, absent; 1, present.
Node-like ornamentation on jugal, mostly on, or
ventral to, the jugal–postorbital bar: 0, absent;
1, present.
Jugal–postorbital bar, width broader than
laterotemporal fenestra: 0, absent; 1,
present.
Jugal–postorbital joint: 0, elongate scarf joint; 1,
short butt joint.
Jugal, form of postorbital process: 0, not
expanded dorsally; 1, dorsal portion of postorbital process is expanded posteriorly.
Jugal–squamosal contact above infratemporal
fenestra: 0, absent; 1, present.
Jugal posterior ramus, forked: 0, absent; 1,
present.
Jugal, posterior ramus: 0, forms anterior and
ventral margin of infratemporal fenestra; 1,
forms part of posterior margin, expands towards
squamosal.
Jugal–quadratojugal contact: 0, overlapping; 1,
tongue-and-groove.
Postorbital, orbital margin: 0, relatively smooth
curve; 1, prominent and distinct projection into
orbit.
Postorbital: 0, T-shaped; 1, triangular and
plate-like.
Postorbital–parietal contact: 0, absent, or very
narrow; 1, broad.
Contact between dorsal process of quadratojugal and descending process of the
squamosal: 0, present; 1, absent.
Quadratojugal, shape: 0, L-shaped, with
elongate anterior process; 1, subrectangular with long axis vertical, short, deep
anterior process.
Quadratojugal, ventral margin: 0, approaches
the mandibular condyle of the quadrate; 1, well
removed from the mandibular condyle of the
quadrate.
Quadratojugal, orientation: 0, faces laterally; 1,
faces posterolaterally.
56. Quadratojugal, transverse width: 0, mediolaterally flattened; 1, transversely expanded and triangular in coronal section.
57. Prominent oval fossa on pterygoid ramus of
quadrate: 0, absent; 1, present.
58. Quadrate lateral ramus: 0, present; 1, absent.
59. Quadrate shaft: 0, anteriorly convex in lateral
view; 1, reduced in anteroposterior width and
straight in lateral view.
60. Paraquadratic foramen or notch, size: 0, absent
or small, opens between quadratojugal and
quadrate; 1, large.
61. Paraquadratic foramen, orientation: 0, posterolateral aspect of quadrate shaft; 1,
lateral aspect of quadrate or quadratojugal.
62. Paraquadratic foramen, position: 0, on quadratequadratojugal boundary; 1, located within
quadratojugal.
63. Quadrate mandibular articulation: 0, quadrate condyles subequal in size; 1, medial
condyle is larger than lateral condyle;
2, lateral condyle is larger than medial.
64. Paired frontals: 0, short and broad; 1,
narrow and elongate (more than twice as
long as wide).
65. Supratemporal fenestrae: 0, open; 1, closed.
66. Supratemporal fenestrae, anteroposteriorly elongated: 0, absent, fenestrae are subcircular to oval
in shape; 1, present.
67. Parietal septum, form: 0, narrow and smooth; 1,
broad and rugose.
68. Parietosquamosal shelf: 0, absent; 1,
present.
69. Parietosquamosal shelf, extended posteriorly as
distinct frill: 0, absent; 1, present.
70. Composition of the posterior margin of the parietosquamosal shelf: 0, parietal contributes only a
small portion to the posterior margin; 1, parietal
makes up at least 50% of the posterior margin.
71. Postorbital–squamosal bar: 0, bar-shaped; 1,
broad, flattened.
72. Postorbital–squamosal tubercle row: 0, absent; 1,
present.
73. Enlarged tubercle row on the posterior squamosal: 0, absent; 1, present.
74. Frontal and parietal dorsoventral thickness: 0,
thin; 1, thick.
75. Paroccipital processes: 0, extend laterally
and are slightly expanded distally; 1, distal
end pendent and ventrally extending.
76. Paroccipital processes, proportions: 0, short and
deep (height ⱖ 1/2 length); 1, elongate and
narrow.
77. Post-temporal foramen/fossa, position: 0,
totally enclosed with the paroccipital
process; 1, forms a notch in the dorsal
78.
79.
80.
81.
82.
83.
84.
85.
86.
87.
88.
89.
90.
91.
92.
93.
94.
95.
96.
97.
98.
267
margin of the paroccipital process,
enclosed dorsally by the squamosal.
Supraoccipital, contribution to dorsal margin of
foramen magnum: 0, forms entire dorsal margin
of foramen magnum; 1, exoccipital with medial
process that restricts the contribution of the
supraoccipital.
Basioccipital, contribution to the border of the
foramen magnum: 0, present; 1, absent, excluded
by exoccipitals.
Basisphenoid: 0, longer than, or subequal
in length to, basioccipital; 1, shorter than
basioccipital.
Prootic–basisphenoid plate: 0, absent; 1,
present.
Basal tubera, shape: 0, knob-shaped; 1, plateshaped.
Basipterygoid processes, orientation: 0,
anteroventral; 1, ventral; 2, posteroventral.
Premaxilla–vomeral contact: 0, present; 1,
absent, excluded by midline contact
between maxillae.
Dorsoventrally deep (deeper than 50% of snout
depth) median palatal keel formed of the
vomers, pterygoids, and palatines: 0, absent; 1,
present.
Pterygovomerine keel, length: 0, less than 50% of
palate length; 1, more than 50% of palate length.
Pterygoid–maxilla contact at posterior end of
tooth row: 0, absent; 1, present.
Pterygoquadrate rami, posterior projection of
ventral margin: 0, weak; 1, pronounced.
Cortical remodelling of surface of skull
dermal bone: 0, absent; 1, present.
Predentary: 0, absent; 1, present.
Predentary size: 0, short, posterior premaxillary teeth oppose anterior dentary teeth;
1, roughly equal in length to the premaxilla,
premaxillary teeth only oppose predentary.
Predentary, anterior end in dorsal view: 0,
rounded; 1, pointed.
Predentary, oral margin: 0, relatively smooth; 1,
denticulate.
Tip of predentary in lateral view: 0, does not
project above the main body of predentary;
1, strongly upturned relative to main body of
predentary.
Predentary, ventral process: 0, single; 1,
bilobate.
Predentary, ventral process: 0, present, well
developed; 1, very reduced or absent.
Dentary symphysis: 0, V-shaped; 1, spout
shaped.
Dentary tooth row (and edentulous anterior portion) in lateral view: 0, straight; 1,
anterior end downturned.
268
D. B. NORMAN ET AL.
99. Dorsal and ventral margins of the dentary:
0, converge anteriorly; 1, subparallel.
100. Ventral flange on dentary: 0, absent; 1, present.
101. Coronoid process: 0, absent or weak, posterodorsally oblique, depth of mandible at
coronoid is less than 140% depth of mandible beneath tooth row; 1, well developed,
distinctly elevated, depth of mandible at
coronoid is more than 180% depth of mandible beneath tooth row.
102. Anterodorsal margin of coronoid process
formed by posterodorsal process of
dentary: 0, absent; 1, present.
103. Coronoid process, position: 0, posterior to
dentition; 1, lateral to dentition.
104. External mandibular fenestra, situated on
dentary-surangular-angular boundary: 0,
present; 1, absent.
105. Small fenestra positioned dorsally on the
surangular-dentary joint: 0, absent; 1,
present.
106. Ridge or process on lateral surface of surangular, anterior to jaw suture: 0, absent; 1,
present, strong anteroposteriorly extended
ridge; 2, present, dorsally directed fingerlike process.
107. Retroarticular process: 0, elongate; 1, rudimentary or absent.
108. Node-like ornamentation of the dentary and
angular: 0, absent; 1, present.
109. Level of jaw joint: 0, level with tooth row, or
weakly depressed ventrally; 1, strongly
depressed ventrally, more than 40% of the
height of the quadrate is below the level of
the maxilla.
110. Mandibular osteoderm: 0, absent; 1, present.
111. Premaxillary teeth: 0, present; 1, absent,
premaxilla edentulous.
112. Premaxillary teeth, number: 0, six; 1, five; 2,
four; 3, three; 4, two.
113. Premaxillary teeth, crown expanded above
root: 0, crown is unexpanded mesiodistally
above root, no distinction between root and
crown is observable; 1, crown is at least
moderately expanded above root.
114. Premaxillary teeth increase in size posteriorly: 0, absent, all premaxillary teeth
subequal in size; 1, present, posterior premaxillary teeth are significantly larger in
size than anterior teeth.
115. Maxillary and dentary crowns, shape: 0,
apicobasally tall and blade-like; 1, apicobasally short and subtriangular; 2, diamondshaped.
116. Maxillary/dentary teeth, marginal ornamentations: 0, fine serrations set at right
117.
118.
119.
120.
121.
122.
123.
124.
125.
126.
127.
128.
129.
130.
131.
132.
angles to the margin of the tooth; 1, coarse
serrations (denticles) angle upwards at 45°
from the margin of the tooth.
Enamel on maxillary/dentary teeth: 0, symmetrical; 1, asymmetrical.
Apicobasally
extending
ridges
on
maxillary/dentary teeth: 0, absent; 1,
present.
Apicobasally extending ridges on lingual/labial
surfaces of maxillary/dentary crowns confluent
with marginal denticles: 0, absent; 1, present.
Prominent primary ridge on labial
side of maxillary teeth: 0, absent; 1,
present.
Prominent primary ridge on lingual side of
dentary teeth: 0, absent; 1, present.
Position of maxillary/dentary primary ridge: 0,
centre of the crown surface, giving the crown a
relatively symmetrical shape in lingual/labial
view; 1, offset, giving crown asymmetrical appearance.
At least moderately developed labiolingual
expansion of crown (‘cingulum’) on
maxillary/dentary teeth: 0, present; 1,
absent.
Heterodont dentary dentition: 0, no substantial heterodonty is present in dentary
dentition; 1, single, enlarged, caniform
anterior dentary tooth, crown is not mesiodistally expanded above root; 2, anterior
dentary teeth are strongly recurved and
caniform, but have crowns expanded mesiodistally above their roots and are not
enlarged relative to other dentary teeth.
Peg-like tooth located anteriorly within dentary,
lacks denticles, strongly reduced in size: 0,
absent; 1, present.
Alveolar foramina (‘special foramina’)
medial to maxillary/dentary tooth rows: 0,
present; 1, absent.
Recurvature in maxillary and dentary
teeth: 0, present; 1, absent.
Overlap of adjacent crowns in maxillary
and dentary teeth: 0, absent; 1, present.
Crown is mesiodistally expanded above
root in cheek teeth: 0, absent; 1, present.
Position of maximum apicobasal crown
height in dentary/maxillary tooth rows: 0,
anterior portion of tooth row; 1, central
portion of tooth rows.
Close-packing and quicker replacement
eliminates spaces between alveolar border
and crowns of adjacent functional teeth: 0,
absent; 1, present.
Fusion between the intercentum of the atlas and
the neural arches: 0, absent; 1, present.
133. Epipophyses on anterior (postaxial) cervicals: 0, present; 1, absent.
134. Cervicals 4–9, form of central surfaces: 0, amphicoelous; 1, at least slightly opisthocoelous.
135. Cervical number: 0, seven/eight; 1, nine; 2,
ten or more.
136. Articulation between the zygapophyses of dorsal
vertebrae: 0, flat; 1, tongue-and-groove.
137. Dorsals, number: 0, 12–13; 1, 15; 2. 16 or
more.
138. Sacrals, number: 0, two; 1, three; 2, four/five;
3, six or more.
139. Sacrum, accessory articulation with pubis:
140. Posterior sacral ribs are considerably longer than
anterior sacral ribs: 0, absent; 1, present.
141. Anterior caudal vertebrae, length of transverse
processes relative to neural spine height: 0, subequal; 1, longer than neural spine.
142. Proximal caudal neural spines: 0, height the
same or up to 50% taller than the centrum; 1,
more than 50% taller than the centrum.
143. Elongate tail (59 or more caudals): 0, absent; 1,
present.
144. Chevron shape: 0, rod-shaped, often with slight
distal expansion; 1, strongly asymmetrically
expanded distally, width greater than length in
midcaudals.
145. Sternal segments of the anterior dorsal ribs: 0,
unossified; 1, ossified.
146. Gastralia: 0, present; 1, absent.
147. Ossified clavicles: 0, absent; 1, present.
148. Sternal plates, shape: 0, absent; 1, kidneyshaped; 2, shafted or hatchet-shaped (rod-like
posterolateral process, expanded anterior end).
149. Proportions of humerus and scapula: 0,
scapula longer or subequal to the humerus;
1, humerus substantially longer than the
scapula.
150. Scapula blade, length relative to minimum
width: 0, relatively short and broad, length
is five to eight times minimum width; 1,
elongate and strap-like, length is at least
nine times the minimum width.
151. Scapula acromion shape: 0, weakly developed or
absent; 1, well developed, spine-like.
152. Scapula, blade-shape: 0, strongly expanded
distally; 1, weakly expanded, near parallelsided.
153. Humeral length: 0, more than 60% of
femoral length; 1, less than 60% of femoral
length.
154. Deltopectoral crest development: 0, well developed, projects anteriorly as a distinct flange; 1,
rudimentary, is at most a thickening on the
anterolateral margin of the humerus.
269
155. Humeral shaft form, in anterior or posterior
view: 0, relatively straight; 1, strongly bowed
laterally along length.
156. Manual length (measured along digit ii or
iii, whichever is longest) as a percentage of
the combined length of the humerus and
radius: 0, less than 40%; 1, more than 40%.
157. Metacarpals with block-like proximal ends:
158. Metacarpals 1 and 5: 0, substantially shorter in
length than metacarpal 3; 1, subequal in length
to metacarpal 3.
159. Penultimate phalanx of the second and
third fingers: 0, shorter than first phalanx;
1, longer than the first phalanx.
160. Manual digit 3, number of phalanges: 0, four; 1,
three or fewer.
161. Manual digits 2–4: 0, first phalanx relatively
short compared to second phalanx; 1, first
phalanx more than twice the length of the second
phalanx.
162. Extensor pits on the dorsal surface of the
distal end of metacarpals and manual phalanges: 0, absent or poorly developed; 1,
deep, well developed.
163. Manual unguals strongly recurved with
prominent flexor tubercle: 0, absent; 1,
present.
164. Acetabulum: 0, at least a small perforation;
1, completely closed.
165. Preacetabular process, shape/length: 0,
short, tab-shaped, distal end is posterior to
pubic peduncle; 1, elongate, strap-shaped,
distal end is anterior to pubic peduncle.
166. Preacetabular process, length: 0, less than
50% of the length of the ilium; 1, more than
50% of the length of the ilium.
167. Preacetabular process, lateral deflection: 0,
10–20° from midline; 1, more than 30°.
168. Dorsal margin of preacetabular process
and dorsal margin of ilium above acetabulum: 0, narrow, not transversely expanded;
1, dorsal margin is transversely expanded
to form a narrow shelf.
169. In dorsal view preacetabular process of the ilium
expands mediolaterally towards its distal end: 0,
absent; 1, present.
170. Dorsal margin of the ilium in lateral view: 0,
relatively straight or slightly convex; 1, sinuous,
postacetabular process is strongly upturned.
171. Subtriangular process extending medially from
the dorsal margin of the iliac blade: 0, absent; 1,
present.
172. Subtriangular process, form and position: 0,
short and tab-like, above acetabulum; 1, elongate
and flange-like, on postacetabular process.
270
D. B. NORMAN ET AL.
173. Brevis shelf and fossa: 0, fossa faces ventrolaterally and shelf is near vertical and
visible in lateral view along entire length,
creating a deep postacetabular portion;
1, fossa faces ventrally and posterior of
shelf portion cannot be seen in lateral
view.
174. Length of the postacetabular process as a
percentage of the total length of the ilium:
0, 20% or less; 1, 25–35%; 2, more than 35%.
175. Medioventral acetabular flange of ilium,
partially closes the acetabulum: 0, present; 1, absent.
176. Supra-acetabular ‘crest’ or ‘flange’: 0,
present; 1, absent.
177. Ischial peduncle of the ilium: 0, projects
ventrally; 1, broadly swollen, projects ventrolaterally.
178. Pubic peduncle of ilium: 0, large, elongate,
robust; 1, reduced in size, shorter in
length than ischial peduncle.
179. Pubic process of ischium, shape: 0, transversely compressed; 1, dorsoventrally
compressed.
180. Ischium, shape of shaft: 0, relatively straight;
1, gently curved along length.
181. Ischial shaft, cross-section: 0, compressed
mediolaterally; 1, subcircular and bar-like.
182. Ischial shaft: 0, expands weakly, or is parallelsided, distally; 1, distally expanded into a distinct ‘foot’; 2, tapers distally.
183. Groove on the dorsal margin of the
ischium: 0, absent; 1, present.
184. Tab-shaped obturator process on ischium:
185. Ischial symphysis, length: 0, ischium
forms a median symphysis with the opposing blade along at least 50% of its length;
1, ischial symphysis present distally only.
186. Pubis, orientation: 0, anteroventral; 1,
rotated posteroventrally to lie alongside
the ischium (opisthopubic).
187. Shaft of pubis (postpubis), shape in
cross-section: 0, blade-shaped; 1, rodshaped.
188. Shaft of pubis (postpubis), length: 0,
approximately equal in length to the
ischium; 1, reduced, extends for half or
less the length of the ischium.
189. Reduction of postpubic shaft: 0, postpubic shaft
extends for around half the length of ischium;
1, postpubic shaft is very short or absent.
190. Body of pubis, size: 0, relatively large, makes
substantial contribution to the margin of the
acetabulum; 1, reduced in size, rudimentary,
nearly excluded from the acetabulum.
191. Body of the pubis, massive and dorsolaterally
rotated so that obturator foramen is obscured
in lateral view: 0, absent; 1, present.
192. Prepubic process: 0, absent; 1, present.
193. Prepubic process: 0, compressed mediolaterally,
dorsoventral
height
exceeds
mediolateral width; 1, rod-like, mediolateral width exceeds dorsoventral height.
194. Prepubic process, length: 0, stub-like and
poorly developed, extends only a short
distance anterior to the pubic peduncle of
the ilium; 1, elongated into distinct anterior process.
195. Prepubic process, extends beyond distal
end of preacetabular process of ilium: 0,
absent; 1, present.
196. Extent of pubic symphysis: 0, elongate; 1,
restricted to distal end of pubic blade, or
absent.
197. Femoral shape in medial/lateral view: 0, bowed
anteriorly along length; 1, straight.
198. Femoral head: 0, confluent with greater
trochanter, fossa trochanteris is groovelike; 1, fossa trochanteris is modified into
distinct constriction separating head and
greater trochanter.
199. ‘Anterior’ or ‘lesser’ trochanter, morphology: 0, absent; 1, trochanteric shelf ending
in a small, pointed, spike; 2, broadened,
prominent, ‘wing’- or ‘blade’-shaped, subequal in anteroposterior width to greater
trochanter; 3, reduced anteroposterior
width, closely appressed to the expanded
greater trochanter.
200. Level of most proximal point of anterior
trochanter relative to level of proximal
femoral head: 0, anterior trochanter is
positioned distally on the shaft, and
separated from ‘dorsolateral’ trochanter/
greater trochanter by deep notch visible
in medial view; 1, anterior trochanter
positioned proximally, approaches level
of proximal surface of femoral head,
closely appressed to ‘dorsolateral’/greater
trochanter (no notch visible in medial
view).
201. Fourth trochanter of femur, shape: 0, low
eminence, or absent; 1, prominent ridge; 2,
pendent.
202. Fourth trochanter, position: 0, located
entirely on proximal half of femur; 1,
positioned at midlength, or distal to
midlength.
203. Anterior (extensor) intercondylar groove
on distal end of femur: 0, absent; 1,
present.
204. Posterior (flexor) intercondylar groove of the
femur: 0, fully open; 1, medial condyle inflated
laterally, partially covers opening of flexor
groove.
205. Lateral condyle of distal femur, position and
size in ventral view: 0, positioned relatively
laterally, and slightly narrower in width
than the medial condyle; 1, strongly inset
medially, reduced in width relative to medial
condyle.
206. Distal tibia: 0, subquadrate, posterolateral
process is not substantially developed; 1,
elongate posterolateral process, backs
fibula.
207. Fibular facet on the lateral margin of the
proximal surface of the astragalus: 0,
large; 1, reduced to small articulation.
208. Calcaneum, proximal surface: 0, facet for
tibia absent; 1, well-developed facet for
tibia present.
209. Medial distal tarsal: 0, articulates distally
with metatarsal 3 only; 1, articulates distally with metatarsals 2 and 3.
210. Metatarsal arrangement: 0, compact, closely
appressed to one another along 50–70% of their
length, spread distally; 1, contact each other
only at proximal ends, spread strongly outwards distally.
211. Digit 1: 0, metatarsal 1 robust and well
developed, distal end of phalanx 1–1
projects beyond the distal end of metatarsal 2; 1, metatarsal 1 reduced and
212.
213.
214.
215.
216.
217.
218.
219.
220.
221.
271
proximally splint-like, end of phalanx 1–1
does not extend beyond the end of metatarsal 2; 2, metatarsal 1 reduced to a vestigial splint or absent, does not bear
digits.
Pedal digit 4 phalangeal number: 0, five; 1, four
or fewer.
Metatarsal 5, length: 0, more than 50%
of metatarsal 3; 1, less than 25% of
metatarsal 3.
Metatarsal 5: 0, bears digits; 1, lacks
digits.
Pedal unguals, shape: 0, tapering, narrow
pointed, claw-like; 1, wide, blunt, hooflike.
Epaxial ossified tendons present along
vertebral column: 0, absent; 1, present.
Ossified hypaxial tendons, present on
caudal vertebrae: 0, absent; 1, present.
Ossified tendons, arrangement: 0, longitudinally arranged; 1, basket-like arrangement of fusiform tendons in caudal region;
2, double-layered lattice.
Parasagittal row of dermal osteoderms on
the dorsum of the body: 0, absent; 1,
present.
Lateral row of keeled dermal osteoderms
on the dorsum of the body: 0, absent; 1,
present.
U-shaped cervical/pectoral collars composed of
contiguous keeled osteoderms: 0, absent; 1,
present.
272
D. B. NORMAN ET AL.
APPENDIX 3
Heterodontosaurus tucki Crompton and Charig, 1962. The holotype skull SAM-PK-K337. A. Right lateral view.
B. Left lateral view. Mechanical preparation of this skull has been made more difficult because of the extremely
adherent layer of haematite impregnated mudstone matrix (reddish-stained) in which the skull was embedded.
Scale bar is 10 mm.
273
APPENDIX 4
Heterodontosaurus tucki Crompton and Charig, 1962. Referred specimen SAM-PK-K1332. The major portion of
the skull that has been fully prepared. A. Right lateral view. B. Left lateral view. C. Dorsal view with the main
body of the skull held vertically. D. Dorsal view with the skull roof held horizontally (the skull frame having
been displaced to the left during burial/compaction. Scale bar is 10 mm.
274
D. B. NORMAN ET AL.
APPENDIX 5
Heterodontosaurus tucki Crompton and Charig, 1962. Referred specimen SAM-PK-K1332. Skull in: A. Ventral
view. B. Ventral view to show the floor of the braincase and posterior palate. C. Right lower jaw (partial) in
lateral view (with the distal end of the quadrate/quadratojugal in articulation). D. Right lower jaw in medial
view. Scale bar 10 mm.
275
APPENDIX 6
Heterodontosaurus tucki Crompton and Charig, 1962. Referred specimen SAM-PK-K1332. Lower jaw. A. Left
lower jaw in lateral view. B. Left lower jaw in medial view. C. Left lower jaw in dorsal view. D. Right lower jaw
in dorsal view (including the distal end of the quadrate-quadratojugal in articulation and the predentary
capping the distal end of the dentary. Scale bar 10 mm.
276
D. B. NORMAN ET AL.
ABBREVIATIONS USED IN
FIGURES 9 ONWARDS
aaf, anterior antorbital fenestra; adf, anterior dentary
foramen; af, internal mandibular adductor fossa; An,
angular; aof, antorbital fossa; apf, anterior premaxillary foramen; Ar, articular; a.r, accessory ridge; bc,
basioccipital condyle; Bo, basioccipital; Bs, basisphenoid; bsf, basisphenoid flange; bst, basisphenoid
tuber; ca, caniniform tooth; car, recess marking probable position of entry of carotid artery into braincase;
ch, choana (internal opening of the narial passage);
Co, coronoid; D, dentary; dc, dentary caniniform
tooth; De., dentary tooth (numbered); dia, diastema;
Ec, ectopterygoid; emf, external mandibular fenestra;
Ex, exoccipital; F, frontal; fm, foramen magnum; fo,
fenestra ovalis; f-po, frontal-postorbital suture; gl,
articular glenoid; gr, groove associated with the
vascular/neural supply to the dental lamina; hpc,
hollow pulp cavity; i1, i2, first and second premaxillary incisiform teeth; imf, internal mandibular fenestra; itf, infratemporal fenestra; J, jugal; jb, jugal boss;
jp, ventrolateral jugal process; jug, jugular vein
(foramen of); La, lacrimal (lachrymal); lap, medial
lappet of the prootic; lbpt, left basipterygoid process;
lPpb, left palpebral; lPt, left pterygoid; lQ, left quadrate; Ls, laterosphenoid; M., maxillary tooth (numbered); mg, Meckel’s groove; Mx, maxilla; mxr, lateral
maxillary ridge; mxs, sagittal maxillary suture; N,
nasal; nc, nuchal crest on supraoccipital; nf, narial
fossa; ns, sagittal nasal suture; nsul, median nasal
sulcus; olf, vaulting on roof of frontals for olfactory
bulbs; Op, opisthotic; orb, recesses on ventral surface
of the frontal and postorbital, forming the roof to the
eye socket; Os, orbitosphenoid; ovc, occipital vascular
canal; Pa, parietal; Pal, palatine (inferred); Part,
prearticular; paf, posterior antorbital fenestra; pc,
premaxillary caniniform; Pd, predentary; pds,
predentary-dentary suture; pd-ds, lateral facet on the
dentary for a thin predentary tongue; Pf, prefrontal;
pf, palatal foramen; pit, pit in dental groove; Pmx,
premaxilla; pmxs, sagittal premaxillary suture; ?pn,
possible pneumatic opening; Po, postorbital; pocc,
paroccipital wing; pof, post-temporal fenestra; por,
postorbital ridge; pos, sutural surface for medial
portion of the postorbital; Ppb, palpebral; ppf, posterior premaxillary foramen; p.r, primary ridge; Pro/Op,
proötic-opisthotic (fused); Prs, presphenoid (inferred);
Psp, parasphenoid; Pt, pterygoid; ptf, pterygoid
flange; ptmr, medial ridge on the pterygoid; ptq,
pterygoid wing of the quadrate; Q, quadrate; qf, quadrate (paraquadratic) foramen; Qj, quadratojugal; qpt,
quadrate wing of the pterygoid; rba, right basal
articulation; rbpta, right basipterygoid articular facet;
rep, replacement crown; rJ, right jugal; rPo, right
postorbital; rPt, right pterygoid; rt, tooth root base
(emergent on dorsal surface of maxilla); S, supraoccipital; Sa, surangular; sc, sagittal crest (of the parietal); sf, symphyseal foramen; sfor, surangular
foramen; Sp, splenial; Sq, squamosal; sqs, squamosal
suture; sqr, squamosal ridge; stf, supratemporal
fenestra; sy, dentary symphysis; V, vomer (partly
inferred); vpar, position of parietal fissure; vs, sagittal
vomer suture (inferred); w.facet, individual wear
facet.
Roman numerals: V, trigeminal fossa (cranial nerve
5); VII, foramina for branches of cranial nerve 7;
IX–XII, foramina for cranial nerves 9–12.

The Lower Jurassic ornithischian dinosaur Heterodontosaurus tucki

Transcription

Similar documents

Solution for the functional analysis of dental occlusion

Ancistargis hamata

Uberabasuchus terrificus sp. nov., a New Crocodylomorpha from the

pdf - Portal de docentes FCT/UNL

they

Jerusalem Crickets - Orange County Mosquito and Vector Control

Autopsyfiles.org - Anna Nicole Smith Autopsy Report

Michale article - Montana Myofunctional Therapy