Heterodontosaurus

Norman, David B., Crompton, Alfred W., Butler, Richard J., Porro, Laura B. & Charig, Alan J., 2011, The Lower Jurassic ornithischian dinosaur Heterodontosaurus tucki Crompton & Charig, 1962: cranial anatomy, functional morphology, taxonomy, and relationships, Zoological Journal of the Linnean Society 163, pp. 182-276 : 218-224

HETERODONTOSAURUS

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figures, or formal description of the maxilla were provided, but the specimen was described and illustrated in their unpublished ms notes.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 (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 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 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 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 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 Zahnreihe - derivable 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.

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 ( SAM- PK-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 .

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-K 1334 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-PK- K1334 ( 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.

Figures

Figure 21. Heterodontosaurus tucki Crompton & Charig, 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.

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.

Figure 28. Heterodontosaurus tucki. SAM-PK-K10487 (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.)

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.

Figure 30. Heterodontosaurus tucki. SAM-PK-K1334 (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.

Figure 31. Heterodontosaurus tucki. SAM-PK-K1334 (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.

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.

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.

Figure 37. A, Geranosaurus atavus Broom, 1911. SAM- PK-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.

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).

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).