FABIEN KNOLL 1, 2 INTRODUCTION

Transcription

1 Bull. Soc. géol. Fr., 2004, t. 175, n o 1, pp Review of the tetrapod fauna of the Lower Stormberg Group of the main Karoo Basin (southern Africa) : implication for the age of the Lower Elliot Formation FABIEN KNOLL 1, 2 Key words. Upper Triassic, Lower Jurassic, South Africa, Lesotho, Tetrapods, Correlations Abstract. On the basis of its tetrapod fauna, the Stormberg Group of southern Africa can be divided into at least two zones. The lowest one is composed of the Molteno and Lower Elliot Formations. It has yielded remains of poorly known stereospondyls, unpublished rauisuchians, enigmatic carnivorous dinosaurs, numerous prosauropods, and rare sauropods and cynodonts. The presence of an ornithischian and a chelonian is alleged, but that of a dicynodont is quite doubtful. The age of the Lower Elliot Formation has been the subject of much speculation. A parallel is drawn with European Upper Norian- Rhaetian sites in which plateosaurid prosauropods constitute the vast majority of macrovertebrate remains. In conclusion, the Lower Elliot Formation is correlated with the Apachean Land Vertebrate Faunachron, which suggests that it is not older than late Norian. Revue de la faune de tétrapodes du «groupe de Stormberg inférieur» du bassin principal du Karoo (Afrique australe) : implication pour l âge de la formation Elliot inférieure Mots clés. Trias supérieur, Jurassique inférieur, Afrique du Sud, Lesotho, Tétrapodes, Corrélations Résumé. Le «Groupe du Stormberg» d Afrique australe peut être divisé, sur la base de sa faune de tétrapodes, en au moins deux zones. La plus ancienne de celles-ci regroupe les formations de Molteno et d Elliot inférieur. Elle a livré des restes appartenant aux groupes Stereospondyli, Rauisuchia, Dinosauria (Prosauropoda, Sauropoda et dinosaures carnivores d affinités incertaines) et Cynodontia. La présence d un Ornithischia et d un Chelonia est fortement suspectée alors que celle d un Dicynodontia est fort douteuse. L âge de la formation d Elliot inférieure est discuté. La prédominance des ossements de prosauropodes dans la formation d Elliot inférieure rappelle des site du Norien supérieur-«rhétien» d Europe. La formation d Elliot inférieure est finalement corrélée avec le «Land Vertebrate Faunachron» Apachéen, ce qui suggère que sa genèse ne remonte pas au-delà du Norien supérieur. INTRODUCTION The Karoo Supergroup of the eponym basin ends by the Stormberg and Drakensberg Groups (fig. 1). The Lesotho Formation, main flood basalts of the Drakensbergs, is essentially made up by tholeitic flows of a maximum thickness of about 1650 m [Schmitz and Rooyani, 1987]. The sediments that lie between the Beaufort Group and the Drakensberg Group, were initially divided by Dunn [1878], on a lithological basis, into three stages, from bottom to top : Coal Measures, Red Beds, and Cave Sandstone. These three subdivisions of the Stormberg Group s. s. correspond to the formations that are now called Molteno, Elliot, and Clarens sandstone, respectively [South African Committee for Stratigraphy, 1980]. The exhortations by the South African Committee for Stratigraphy [1980] to forsake the term Stormberg have hardly been followed. It is widely used since the 19th century (cf. Visser [1984]) and will still be in this work, yet between quotation marks. The Molteno Formation has formed through six sedimentary cycles and appears in the field under various FIG. 1. Location of the main layers of the Stormberg (in black) and Drakenberg (in dark grey) Groups in Africa. FIG. 1. Situation des principaux affleurements des Groupes du Stormberg (en noir) et du Drakensberg (en gris sombre) en Afrique. 1 Département des Sciences de la Terre, Muséum des Sciences naturelles, 6 rue Marcel Proust, F Orléans, France. 2 Département Histoire de la Terre, UMR 5149 CNRS, Muséum national d Histoire naturelle, 8 rue Buffon, F-75005, Paris, France. Manuscrit déposé le 19 février 2003 ; accepté après révision le 15 juillet 2003.

2 74 F. KNOLL lithologies (coarse conglomerates, sandstones, siltstones, mudstones or even coals). Its thickness is highly variable, up to ca. 600 m [Anderson et al., 1998 ; Hancox, 1998 ; Lucas and Hancox, 2001]. It is especially famous for its megaflora, intensively studied for about thirty years (cf. Anderson and Anderson [1970] et sq. to Anderson et al. [1998]). The Elliot Formation is characterized by the presence of shales, red or purple mudstones, and red to white sandstones. Its thickness is also variable, up to ca. 500 m [Hancox, 2000 ; Lucas and Hancox, 2001]. The Elliot Formation is generally divided into two members (Lower Elliot and Upper Elliot), the separation of which corresponds to the boundary between the Euskelosaurus range zone and the Massospondylus range zone, in turn corresponding to the Triassic-Jurassic limit [Smith and Kitching, 1997 ; Knoll 2002a]. TABLE I. Tentative attribution of the ichnogenera mentioned or named by Ellenberger [1970, 1972] in the Lower Stormberg Group of the main Karoo basin (southern Africa) to non-ichnological higher level taxa. TABL. I. Essai d attribution des ichnogenres signalés ou établis par Ellenberger [1970, 1972] dans le groupe du «Stormberg inférieur» du bassin principal du Karoo (Afrique australe) à des taxons non ichnologiques plus inclusifs. THE TETRAPODS OF THE LOWER STORMBERG Problematics Our knowledge of the tetrapod fauna of the Stormberg is extremely confused because, more often than not, the fossils yielded are incomplete and undiagnostic [Kitching and Raath, 1984, p. 123 ; Welman, 1999, p. 227] and the stratigraphical subdivisions are occasionally difficult to distinguish in the field. Thus, from the palaeontological point of view, a profusion of nominal taxa has been coined on the basis of very incomplete remains, which renders problematical the systematics of vertebrate groups such as the Fabrosauridae [Knoll, 2002a], the Heterodontosauridae [Hopson, 1975], the Prosauropoda [Gauffre, 1996], and the Cynodontia [Lucas and Hunt, 1994], for instance. From the stratigraphical point of view, there are unfortunately no objective and non-local criteria to easily differentiate the formations and subdivisions of the Stormberg. Defining, for example, the (possibly diachronous) limit between the Molteno and Elliot formations on the basis of the absence of reptiles in the former (cf. Turner [1972]) would have as a consequence to exclude by definition the discovery of such animals in the Molteno Formation. In the same line of thought, the Euskelosaurus andmassospondylus zones should not be considered strictly equivalent to the Lower Elliot and Upper Elliot-Clarens, respectively. As a matter of fact, (1) Euskelosaurus is possibly absent at the base of the Lower Elliot Formation [Kitching and Raath, 1984, fig. 2], (2) Massospondylus remains have been mentioned from the Lower Elliot Formation (although probably incorrectly) [Kitching and Raath, 1984, p. 118], (3) Massospondylus might have appeared (at the latest) in Norian times according to Gauffre [1996, p ] (but this hypothesis, based on a cladistic analysis, is debatable), and (4) Massospondylus may be present only at the base of the Clarens Formation [Kitching and Raath, 1984, fig. 2]. However, the tetrapods from the Stormberg can be divided into at least two distinct faunas. These correspond on the one hand to the Lower Stormberg (Molteno + Lower Elliot Formations) and, on the other hand, to the Upper Stormberg (Upper Elliot s. l. + Clarens Formations). The Zone A/7 of Ellenberger [1970], Red Beds inférieurs c du Lesotho, has yielded a fauna of amphibians, lepidosauromorphs, ornithischians, small prosauropods, and therapsids. It significantly bears the mark of the Upper Elliot Formation and it is assigned here to this subdivision (as implicitly proposed by Kitching and Raath [1984]). It is impossible to assess how representative a particular ichnofauna is of the animal communities that existed in a given area. The limited knowledge of the latter compromises the identification of track makers, which remains frequently highly conjectural (tab. I). In addition, the taxonomy of late Triassic and early Jurassic vertebrate tracks in general and of the Stormberg ones in particular is in a confused state at the present time, with an excess of names that in many cases are junior synonyms of other ichnotaxa. The ichnological aspect will consequently not be comprehensively considered in the review of the Lower Stormberg tetrapods presented below. Stereospondyli Ellenberger and Ellenberger [1956, 1960], Ellenberger et al. [1964, 1967], Ellenberger [1970], Hopson [1980, 1984], Kitching and Raath [1984], Smith and Kitching [1997], Anderson et al. [1998], Galton and van Heerden [1998], Gow and Latimer [1999], and Hancox [2000] have mentioned the presence of large amphibian remains in the Molteno-Lower Elliot formations, but only Warren and Damiani [1999] have given a detailed study of three specimens from the Lower Elliot : two Chigutisauridae gen. et sp. indet. and a Stereospondyli gen. et sp. indet. This suggests that Batrachopodiscus [Ellenberger, 1970, 1972] could derive from chigutisaurid-like animals.

3 TETRAPOD FAUNA OF THE KAROO BASIN (S. AFRICA) 75 Chelonia Dijaquesopus is an ichnogenus of the Molteno Formation in the Lesotho capital that Ellenberger [1970, p. 344, 1972, p. 44] was inclined to consider as being derived from a primitive crocodilian or an archaic chelonian. Dijaquesopus recalls Episcopopus [Ellenberger, 1970, 1972, 1974], an ichnotaxon of probable chelonian affinity. No detailed description of Dijaquesopus has been published so far, though P. Ellenberger planned to do it [Ellenberger, 1972, p. 43] and the original track is still visible [Ambrose, 1993, p. 17]. The presence of the Chelonia at the time of genesis of the Molteno Formation in southern Africa is also suggested by crescentic traces visible near Maclear (Eastern Cape, South Africa), which may be those of a Proganochelyidae [Raath et al., 1990 ; Anderson et al., 1998]. Lepidosauromorpha The ichnogenus Lacertoidopus [Ellenberger, 1970, 1972] might document the presence of Lepidosauromorpha in the Lower Stormberg. Erythrosuchia It is widely accepted that the mandible and the cervical vertebra of Kraai River (Eastern Cape) described by Seeley [1894] alongside the main remains of Euskelosaurus at his disposal do not belong to this Plateosauridae, but to the Erythrosuchidae Erythrosuchus, for which Kraai River is the locus typicus [Broom, 1911 ; Huene, 1911]. This material comes from the Cynognathus zone, like all the material of this taxon known to date [Welman, 1995, p. 10]. Aetosauria Based on the South African Museum (SAM, Cape Town) catalogue, Kitching and Raath [1984, p. 118] mentioned the Stagonolepididae Aetosaurus in the Lower Elliot. Examination of the SAM collections has not permitted to confirm this report. The presence of the Aetosauria in the Molteno-Lower Elliot formations could be indicated by the presence of the ichnogenus Brachychirotherium (cf. Olsen and Galton [1984], Lockley and Hunt [1995], and Heckert and Lucas [2000a]). Nevertheless, this ichnogenus is sometimes credited to the Rauisuchia [Anderson et al., 1998 ; Olsen et al., 2002] or to the Pseudosuchia [Melchor and de Valais, 2002]. Moreover, Gand et al. [2000, p. 609] rejected the proposal of Olsen and Galton [1984] according to which Pseudotetrasauropus is a junior synonym of Brachychirotherium : it would rather be a junior synonym of Otozoum. The latter conclusion is also reached by Lockley and Meyer [2000] who assigned, like Gand et al. [2000] and most recent authors (cf. Gierlinski [1999, p. 233], Matsukawa et al. [2002], Olsen et al. [2002, p. 1307, tab. 1], and Rainforth [2000, 2002]), Otozoum to the Prosauropoda. Hunt et al. [1998, p. 523] suggested that Pseudotetrasauropus should not be synonymized with any other taxon. They think, however, that it is related to a Prosauropoda (see also, for instance, Hunt et al. [1993a, p. 201, 1993b], Lockley et al. [1992, 1993, 2000], Farlow and Lockley [1993], Lockley and Hunt [1995], and Melchor and de Valais [2002]). Finally, the results of the recent work of Rainforth [2000, 2001, 2002] confirm the distinction between Pseudotetrasauropus and Otozoum, but the animal responsible of the former ichnotaxon would be closer to some Crurotarsi, like the Rauisuchia, than to the Prosauropoda. In conclusion, the presence of Aetosauria in the Molteno-Lower Elliot appears very doubtful. Rauisuchia The presence of a large Rauisuchia in the Lower Elliot Formation is attested [Hopson, 1984 ; Kitching and Raath, 1984 ; Olsen and Galton, 1984 ; Smith and Kitching, 1997 ; Anderson et al., 1998 ; Galton and van Heerden, 1998]. Unfortunately, no detailed study has been published so far. Huene [1932, p ] coined the nominal taxon Basutodon ferox for a crenulated tooth from the Lower Elliot Formation of Lesotho, which he believed to be of theropodan affinity. He also presented an isolated tooth of the same kind from the Free State [Huene, 1932, p. 100]. Ellenberger and Ginsburg [1966, p. 445] and Ellenberger [1970, p ] mentioned similar teeth that they believed to be of carnivorous dinosaur affinity too. However, van Heerden [1979, p. 78] considered that it is more probable that a tooth such as the holotype of Basutodon ferox belongs to an erythrosuchian archosaur and he most reasonably considered Basutodon ferox a nomen dubium. Charig et al. [1965] and Cooper [1980] thought that the archosaurian serrated teeth of carnivorous-like morphology found from time to time in the Lower Elliot Formation along with prosauropod remains [Charig et al., 1965, p. 207 ; Ellenberger and Ginsburg, 1966, p. 445 ; Galton and van Heerden, 1998, p. 172] most probably belong to the latter. Kitching and Raath [1984, tab. 4] followed this hypothesis in classifying Basutodon amongst the Prosauropoda, but they suggested that it could actually be of rauisuchian nature [Kitching and Raath, 1984, p. 114]. The latter proposal has been accepted by most recent authors, notably Anderson et al. [1998]. The rather large, but finely crenulated, teeth yielded by the Lower Elliot Formation are definitely not prosauropodan in origin. However, it appears quite hazardous to refer them to a rauisuchian rather than to a theropod. Be that as it may, some ichnotaxa such as Sauropodopus [Ellenberger, 1970, 1972] are best considered of rauisuchian derivation (tab. I). Crocodylomorpha Ellenberger and Ellenberger [1960] and Ellenberger [1970, 1972] described a small track (fig. 2) that they believe to be a relative of Comptichnus from Connecticut [Hitchcock, 1865]. Haubold [1971, p. 60] considered this genus a synonym of Batrachopus, an opinion accepted notably by Olsen and Padian [1986, p. 261]. Haubold [1971, p. 61, 116, 1984, p. 170] thus mentioned the presence of Batrachopus in the Lower Stormberg. However, whereas the crocodylomorph affinity of Batrachopus is pretty well-substantiated [Olsen and Padian, 1986 ; Demathieu and Sciau, 1992, 1995], the real identity of the Lesotho track remains open to discussion, based on the works of Ellenberger and Ellenberger [1960] and Ellenberger [1970, 1972].

4 76 F. KNOLL It is possibly due to erroneous identification of rauisuchian remains that Ellenberger [1970, p ] mentioned carnivorous dinosaurs in the Molteno-Lower Elliot Formations. However, the presence of such animals in the Lower Elliot Formation is attested. In fact, Aliwalia rex [Galton, 1985] is considered a carnivorous dinosaur of enigmatic, not necessarily theropodan, affinity [Galton and van Heerden, 1998]. Moreover, Ray and Chinsamy [2002] described a possible isolated theropod tooth from the Lower Elliot Formation of Lesotho. The presence of the Theropoda in the Molteno-Lower Elliot Formations is further supported by tridactyl traces often assigned to the ichnogenus Grallator [Haubold, 1971, 1986 ; Olsen and Galton, 1984 ; Raath et al., 1990 ; Raath, 1996 ; Anderson et al., 1998 ; Galton and van Heerden, 1998 ; Gow and Latimer, 1999 ; Heckert and Lucas, 1999, 2000b ; Rainforth, 2001]. Such ichnogenera as Deuterotrisauropus [Ellenberger, 1970, 1972] are possibly junior synonyms of Grallator, as already proposed by Olsen and Galton [1984], whereas others, such as Qemetrisauropus and Prototrisauropus [Ellenberger 1970, 1972], evoke Eubrontes. Recently, Melchor et al. [2002] described footprints with striking avian characters (such as a reversed hallux) from the Norian of Argentina. They briefly compared their material with Trisauropodiscus [Ellenberger, 1970, 1972], which they considered as being based on poorly preserved specimens similar to Anomoepus. This assertion is hasty because (1) as shown by Ellenberger s [1972, Pl. 15] iconography, numerous Trisauropodiscus specimens were well preserved and (2) Anomoepus is not known in the Lower Stormberg [Knoll and Battail, 2001 ; Knoll, 2002a]. Trisauropodiscus (which is possibly synonym with Prototrisauropodiscus) is doubtless the most questioning ichnotaxon of the Lower Stormberg. Its avian appearance is conspicuous and Ellenberger [1978, 1993] suggested that it may represent a true Aves group (Lacunavia). This opinionissharedbylockleyet al. [1992] who justly noted if they [the Trisauropodiscus tracks] were found in post-jurassic strata, they would readily be accepted as avian in origin. Even so, Rainforth [1997, 2001] and Lockley and Rainforth [2002] have been recently more disposed to consider this kind of tracks as deriving from nonavian dinosaurs. Be that as it may, Trisauropodiscus is unlikely to have ornithischian affinity and the Argentinean bird-like footprints can confidently be considered as pertaining to this genus. FIG. 2. Footprint described by Ellenberger [1970, 1972] as Comptichnus moorosii, Lower Elliot Formation, south of Maphutseng (Mohale s Hoek district, Lesotho). Scale bar = 50 mm. FIG. 2. Empreinte décrite comme Comptichnus moorosii par Ellenberger [1970, 1972], Formation d Elliot inférieure, sud de Maphutseng (district de Mohale s Hoek, Lesotho). Barre d échelle = 50 mm. Theropoda Prosauropoda The bones of the Prosauropoda are the most common fossils of the Lower Elliot Formation. Thus, all identifiable specimens collected from the Lower Elliot Formation during recent intensive field works in western and southern Lesotho consist of large prosauropod bones [Knoll, 2002b]. In fact, no less than about 80 % of the specimens yielded by the Lower Elliot Formation are prosauropodan [Anderson et al., 1998 ; Hunt, 1991, p. 51 ; pers. unpubl. data]. At least two taxa seem to be present : Euskelosaurus (Plateosauridae) and Melanorosaurus (Melanorosauridae). The taxon name Euskelosaurus is used with quotation marks throughout this paper because it should be considered a nomen dubium, as shown by Gauffre [1996]. Interestingly enough, despite the unquestionable dominance of prosauropods in the skeletal record of the Lower Stormberg, no prosauropod track is positively recognized in the ichnotaxa coined or identified by Ellenberger [1970, 1972] in these strata (tab. I). In a preliminary revision of these taxa, Olsen and Galton [1984] suggested that only Tetrasauropus unguiferus could be regarded with a good degree of certainty as prosauropodan in origin. However, Lockley et al. [2001] recently demonstrated a sauropodan nature for this species. It is true that the formation of the vertebrate tracks is independent of the number of individuals of the living population [Demathieu, 1980] and that the Lower Stormberg skeletal record is visibly biased toward large forms. Even so, the possibility that not even one prosauropod track exists amongst the quite rich and various ichnological assemblages available to Ellenberger [1970, 1972] in the Lower Elliot appears irrational : this situation may rather reveal the present difficulties to determine prosauropod tracks as such. Sauropoda Ellenberger [1970, p. 345] noted the rise of the Sauropoda as soon as the age of the Molteno Formation. Interestingly enough in that respect, Euskelosaurus could be a true Sauropoda for Ellenberger and Ginsburg [1966, p. 446]. Later on, Ginsburg [1970, p. 1308] formally classified all the large sauropodomorphs from the Upper Triassic of southern Africa in the Sauropoda. This view has been largely discarded, but according to the phylogenetic defini-

5 TETRAPOD FAUNA OF THE KAROO BASIN (S. AFRICA) 77 Prosauropoda (NM QR 3314) on display in the National Museum (NM, Bloemfontein) [Welman, 1995, 1999, 2000]. The latter specimen is from the Euskelosaurus zone. SAM P K 8027 would come from a level five meters below that of NM QR 3314 according to R. Smith [pers. comm., 1999], who found it. However, according to D. Norman [pers. comm., 1999], stratigraphically it may be relatively low in the Elliot Fm but it is still likely to be chronologically early Jurassic in age. The opinions of R. Smith and D. Norman appear in complete disagreement. Pending the publication of SAM K 8027, no ornithischian specimen is attested from the Molteno-Lower Elliot Formations (contrary to Padian 1997, p. 484]). Nevertheless, at least one ichnogenus coined by Ellenberger [1970, 1972], namely Paratrisauropus, has convincing ornithischian affinities [Haubold, 1986 ; Thulborn, 1990]. Dicynodontia FIG. 3. Footprint described by Ellenberger [1970, 1972] as Deuterosauropodopus major, Lower Elliot Formation, south of Maphutseng (Mohale s Hoek district, Lesotho). Scale bar = 50 mm. FIG. 3. Empreinte décrite comme Deuterosauropodopus major par Ellenberger [1970, 1972], Formation d Elliot inférieure, sud de Maphutseng (district de Mohale s Hoek, Lesotho). Barre d échelle = 50 mm. tion of the Sauropoda [Sereno, 1998, tab. 4] and the results of recent cladistic analyses of Sauropodomorpha [Yates, 2001, 2003 ; Upchurch et al., 2002], at least the Blikanasauridae Blikanasaurus [Galton and van Heerden, 1985] can confidently be considered a primitive Sauropoda. Traces of Sauropoda have been suspected amongst the variety of large tracks from the Molteno-Lower Elliot Formations of Lesotho. Nevertheless, their determination remains controversial [Ellenberger, 1955, 1965, 1970, 1972 ; Ellenberger and Ellenberger, 1958, 1960 ; Charig et al., 1965 ; Ellenberger et al., 1970 ; Ginsburg, 1970 ; Haubold, 1971, 1984 ; Thulborn, 1990 ; Lockley et al., 2001] and the possibility that some of them are in reality traces of Rauisuchia (Sauropodopus, Pseudotetrasauropus) cannot be excluded. However, the interpretation of Tetrasauropus as sauropodan in origin has received sound arguments recently [Lockley et al., 2001] and this may be correct for Deuterosauropodopus (fig. 3) as well [Thulborn, 1990, p. 171], though this is somewhat more questionable. Ornithischia The Ornithischia of Kromme Spruit (Eastern Cape), listed by Kitching and Raath [1984, p. 118] in the Lower Elliot Formation, very probably comes from the Upper Elliot Formation, which has yielded there a magnificent skeleton of Heterodontosaurus [Santa Luca et al., 1976]. A skeleton of a small primitive Ornithischia (SAM P K 8027, currently studied by D. Norman) has been found at the locality (Damplaats, Free State, South Africa) that has yielded the very remarkable skeleton (including the skull) of the Benton [1994, p. 396] and Lucas [1994, p. 871] cited Jachaleria colorata [Bonaparte, 1970, p. 680] as the most recent Anomodontia (i.e. Dicynodontia and a handful of more basal forms). Yet Jachaleria comes from the basis of the Los Colorados Formation, possibly of late Carnian age [Abdala et al., 2001, fig. 10]. The possible presence of a large Dicynodontia, apparently a Kannemeyeriidae close to Kannemeyeria, in the Molteno-Lower Elliot Formations [Charig et al., 1965, p ; Ellenberger, 1970] thus bears a particular importance. According to Battail [1993, p. 34], the Lower Elliot Formation has not yielded any Kannemeyeriidae. Nonetheless, the ichnotaxon Pentasauropus [Ellenberger, 1970, 1972] is frequently interpreted as the footprint of a large Dicynodontia [Demathieu and Haubold, 1974 ; Hopson, 1984 ; Olsen and Galton, 1984 ; Lockley and Meyer, 2000]. Interestingly enough, Kannemeyeria is present in the Beaufort of Lesotho and the distinctions of the Beaufort-Molteno-Elliot limits of Ellenberger [1970] have been rightly challenged [Turner, 1972]. The lack of a rigorous survey of the stratigraphical origin of both bones and traces as well as the unconvincing dicynodontian affinity of the latter (that may be sauropodan as well) definitely prevent from considering confident the presence of Dicynodontia in the Stormberg. Cynodontia Stockley [1947] reported on the discovery of bones in the Molteno Formation. Among this material, Boonstra [1947] identified an ilium and an almost complete skull of a Cynodontia (SAM 11484, currently on display) that he attributed to the Cynognathidae Cynidiognathus. Brink [1955] drew attention to the complete absence of significant anatomical differences between this genus and Cynognathus. These two genera are now usually considered as synonyms [Carroll, 1988, p. 624 ; Battail, 1991, p. 42 ; A. Cruickshank, pers. comm., 2000]. Turner [1972] demonstrated that SAM does not come from the Molteno Formation, but from the Cynognathus zone of the Beaufort. According to Ellenberger [1970, p. 345, 1972, p. 85, 1975, p. 413] and Ellenberger et al. [1970, p. 151] the Cynodontia Chiniquodontidae Belesodon would be present in the Stormberg, more exactly in the Molteno Formation Bull. Soc. géol. Fr., 2004, no 1

6 78 F. KNOLL at Leribe and Likhoele (Lesotho). Unfortunately, these authors have given neither description, nor illustration of the material in question. Yet this genus is restricted to the Triassic of South America according to Sigogneau-Russell and Hahn [1994]. According to B. Battail [pers. comm., 2001], the specimen from Leribe would be the badly preserved skull of Cynodontia from the Lower Elliot Formation briefly described by Costedoat [1962] and considered by her as belonging to a genus close to Belesodon. Further preparation of this specimen has permitted to recognise that it is a large Traversodontidae, presumably Scalenodontoides [B. Battail, pers. comm., 2001]. The Cynodontia Traversodontidae Scalenodontoides [Crompton and Ellenberger, 1957 ; Hopson, 1984 ; Gow and Hancox, 1993] is, therefore, the only Therapsida whose presence in the Lower Elliot Formation is confident. At least one ichnogenus, Moltenotetrapodiscus [Ellenberger, 1970, 1972], is likely to be referred to a Cynodontia s. l. (tab. I). DISCUSSION AND CONCLUSION As remarked by Anderson et al. [1998] among others, the correlation of exclusively terrestrial Gondwanan deposits with the international standard reference based on northern marine sequences is naturally difficult. Shubin and Sues [1991] also note that the early Mesozoic continental deposits can rarely be dated with a precision greater than the stage. The late Triassic age assigned for a long time to the Stormberg Group and almost universally accepted [Romer, 1975] was actually not founded on ascertained correlation. It rather resulted from circular reasoning, based e.g., on the observation that no Prosauropoda is known in the European Jurassic (generally poor in skeletal remains of terrestrial animals [Olsen and Galton, 1977]) and the conclusion that this group manifestly disappeared at the end of the Triassic [Romer, 1956, 1966 ; Appleby et al., 1967]. The first rigorous correlation proposals are due to F. Ellenberger, P. Ellenberger, and L. Ginsburg [Ellenberger, 1965 ; Ginsburg, 1966 ; Ellenberger et al., 1967, 1970]. Analogies of the ichnofauna led these authors to correlate their Upper Molteno and the Lower Elliot with the Keuper, the Upper Elliot with the Rhaetian -Hettangian, and the Clarens with the Hettangian-Sinemurian. This Upper Molteno has yielded quite numerous vertebrate remains, notably prosauropods, cynodonts, and even amphibians [Ellenberger, 1970, 1972]. Its palaeontological contents is similar to that of the Lower Elliot [Ellenberger et al., 1964, p. 323] and it is to group with the latter. In fact, the Molteno Formation would not yield any vertebrate skeletal remains, except for some fishes [Anderson et al., 1998 ; Hancox, 2000]. Palaeobotanists H. and J. Anderson (Anderson and Anderson [1970] et sq. to Anderson et al. [1998]) have traditionally assigned a Carnian age to the Molteno Formation, which is rich in plants. Because the upper part of the Molteno Formation would be synchronous with the basis of the Lower Elliot Formation [Dingle et al., 1983 ; Turner, 1983 ; Visser, 1984 ; Cairncross et al., 1995] (but see Hancox [1998] and Lucas and Hancox [2001, p. 5]), the latter is finally considered as being of Carnian age by Anderson et al. [1998], whereas a late Norian/ Rhaetian age had been previously advocated by the same authors for the Elliot Formation. The Carnian dating of the Molteno Formation is, however, by no mean ascertained [H. Anderson, pers. comm., 2000]. In fact, the flora of the Molteno Formation appears comparable to that of the Upper Norian of Greenland [Wing et al., 1992, p. 340] and, most probably, only the most basal part of the Molteno could be Carnian in age [Hancox, 2000]. Recently, Lucas and Hancox [2001, p. 5] estimate the maximum age of the upper part of the Molteno Formation that bears dinosaur tracks [Raath et al., 1990 ; Raath, 1996] as late Carnian and they integrate the Molteno Formation as a whole in the Carnian in a figure they provide [Lucas and Hancox, 2001, fig. 1]. Olsen and Galton [1984] proposed, from the survey of the bony remains and the ichnology, a late Carnian-early Norian age for the Lower Elliot Formation. On ichnological bases, Haubold [1984, tab. 9, 1986, fig. 15.1] correlate the Lower Elliot with the Rhaetian. Gauffre [1993, p ] gave an early Carnian age to the Lower Elliot Formation. Its argumentation rests on : (1) the assumption that no traversodontid is known after the Carnian and (2) the possibility that the traversodontidsrauisuchians-rynchosaurs assemblage is older than the association phytosaurs-metoposaurs-suchians (cf. Sues and Olsen [1990]). Latter, Gauffre [1996, p ] somewhat revised his judgement and assigned a late Carnian age to the base of the Lower Elliot Formation. The arguments are essentially the following : (1) large traversodontids are not known after the Carnian and (2) the underlying Molteno Formation is dated as early Carnian. Both Gauffre s [1993, 1996] datings must be refuted because : (1) it is now established that traversodontids are known until the end of the Triassic [Benton, 1994, p. 396, fig ; Lucas and Hunt, 1994, p. 344, fig ; Cuny, 1995, p. 353 ; Battail, 2001, p. 34 ; Lucas and Hancox, 2001, p. 7], (2) negative evidence is not an evidence of non existence, especially because, as remarked by Gauffre [1996, p. 35] himself, the study of microvertebrate remains is more frequent in Europe (see e.g., Evans et al. [2000]) than in the countries from which the large traversodontids come from, and (3) the dating of the Molteno Formation is uncertain. Without new particular arguments, the Lower Elliot is assigned to the late Carnian-Norian s. l. as a whole [Olsen, 1980, fig. 14 ; Olsen et al., 1982, fig. 17 ; Olsen and Sues, 1986, fig. 25.1), to the late Carnian or to the early Norian [Paul, 1988, p. 250 ; Barrett, 2000, tab. 3.2 ; Galton, 2001a, p. 40], to the Carnian or to the early Norian [Ochev and Shishkin, 1989, p. 162], to the Norian [Battail, 1993, p. 34 ; Benton, 1993], to the early Norian (early Lacian) [Benton, 1994], to the Norian (sensu S.T.S (cf. Cuny [1993, p ])) [Lucas and Hunt, 1994, fig. 20.4], to the Carnian [Galton and van Heerden, 1998 ; Damiani, 2001, p. 444], and to the Rhaetian [Heckert and Lucas, 1999, 2000b]. It should be underlined here that, despite its recognition as a valid stage by the Subcommission on Triassic Stratigraphy [Visscher, 1992], the correlation and even the maintaining of the Rhaetian stage remain controversial (cf., e.g., Cuny [1993, p ], Fraser and Sues [1994, p. 130], Lucas and Hunt [1994, p. 335], and Hallam [2002, p ]).

7 TETRAPOD FAUNA OF THE KAROO BASIN (S. AFRICA) 79 FIG. 4. Plateosaurid femur in the famous site discovered by F. Ellenberger in 1955 at Maphutseng village (Mohale s Hoek district, Lesotho). FIG. 4. Fémur de Plateosauridae dans le célèbre gisement découvert par F. Ellenberger en 1955 dans le village de Maphutseng (district de Mohale s Hoek, Lesotho). In the field, the Lower Elliot Formation is essentially characterized by the presence (generally sparse) of bones of large Prosauropoda [pers. obs., 2000, 2001 ; fig. 4], which have not been considered as they deserve to be in the attempts of dating the Lower Elliot Formation. The dinosaur dominance became inquestionable in the latest late Triassic and the Prosauropoda were especially frequent in late Norian- Rhaetian time, over a large geographical scale (cf. Hunt et al. [1998]). In contrast, in the formations of Carnian age, the remains of Sauropodomorpha are generally uncommon (the Dicynodontia and, especially, the Rhynchosauria being then the dominant primary consumers [Langer et al., 2000, p. 3-4]) and are never represented by large species. It is the case, for instance, of the Alemoa Member of the Santa Maria Formation of Brazil, of late Carnian age [Langer et al., 1999]. The prosauropod remains of the Lower Elliot Formation are often assigned to Euskelosaurus [Kitching and Raath, 1984 ; Rubidge and Hancox, 2002, p. 58]. Admittedly, they are often isolated elements (such as vertebrae) that are not identifiable on the basis of diagnostic features. Be that as it may, it is of interest that Euskelosaurus appears to be a close relative of Plateosaurus [Gauffre, 1996]. The two taxa could even be sister-groups [Benton et al., 2000 ; Pisani et al., 2002], although this remains to be confirmed by detailed studies of new good specimens such as NM QR The genus Plateosaurus constitutes a remarkably common element of many Upper Norian sites in Europe and Greenland [Sander, 1992, 1999 ; Schoch and Wild, 1999 ; Galton, 2001b]. Its presence in Rhaetian strata is possible, but is not well substantiated by the fossil record. According to Cuny [1993, 1995, pers. comm., 2003], plateosaurids went extinct in Europe due to a shift of local conditions in relation with the Rhaetian transgression. In conclusion, a Norian s. l. age for the Lower Elliot Formation appears well supported. The prosauropod abundance being an argument of correlation with the Apachean Land Vertebrate Faunachron [Lucas, 1998, p. 375 ; FIG. 5. Stratigraphy of the Lower Stormberg Group (see text). Approximation of the initial boundaries (Ma) of Carnian, Norian, and Rhaetian stages are from Gradstein and Ogg [1996] whereas the limits of the Hettangian are derived from Pálfy et al. [2000]. FIG. 5 Stratigraphie du «Groupe du Stormberg inférieur» (voir texte). Les datations (en Ma) des limites inférieures du Carnien, du Norien et du «Rhétien» sont arrondies de celles fournies par Gradstein et Ogg [1996] alors que l âge du début de l Hettangien, comme celui de son terme, découle du travail de Pálfy et al. [2000]. Lucas, 1999, p. 510], the Lower Elliot Formation could be more precisely late Norian- Rhaetian in age (fig. 5). The ichnotaxon Pseudotetrasauropus, which is probably valid, could also support a Late Norian- Rhaetian age (cf. Lockley et al. [1992] ; Hunt et al. [1993a, p. 201, 1993b] and Lockley and Hunt [1995, fig. 3.27]). Cooper [1982] reached a similar conclusion about twenty years ago, but this remained relatively ignored. In fact, this author placed Bull. Soc. géol. Fr., 2004, no 1

8 80 F. KNOLL the base of the Elliot Formation in a Plateosaurus biostratigraphical zone (see also Hancox [2000]). The present conclusions are therefore in agreement with those of Lucas and Hancox [2001] who assign a Norian s. l. age to the Lower Elliot Formation, based on the domination of the fauna by the Prosauropoda Euskelosaurus (considered as a vicar of Plateosaurus), ichnological correlations, and the stratigraphical continuity with the Upper Elliot Formation of early Jurassic age. Acknowledgments. The present paper is derived from my unpublished Ph.D. dissertation [Knoll, 2002a] presented at the CNRS-UMR 8569 (Muséum national d Histoire naturelle, Paris). I gratefully acknowledge its director at that time, P. Janvier (Muséum national d Histoire naturelle, Paris) for his continuous help and interest and for having improved this paper, as did G. Cuny (Geologisk Museum, København), N.-E. Jalil (Université Cadi Ayyad, Marrakech), and D. Sigogneau-Russell (Muséum national d Histoire naturelle, Paris) as well. Two field campaigns in Lesotho and visits of European and southern African institutions in which Stormberg vertebrates are housed were made possible through financial supports provided by the CNRS, the Chancellerie des Universités de Paris, the Jurassic Foundation, and the European Commission (T.M.R. Programme). For access to the collections under their care, I thank A. Milner and S. Chapman (The Natural History Museum, London), R. Smith and S. Kaal (South African Museum, Cape Town), J. Welman (National Museum, Bloemfontein), S. Gill and P. Nyabela (Morija Museum and Archives, Morija), F. Thackeray and H. Fourie (Transvaal Museum, Pretoria), and M. Raath (University of the Witwatersrand, Johannesburg). Useful information has been provided by H. Anderson (National Botanical Institute, Pretoria), B. Battail (Muséum national d Histoire naturelle, Paris), A. Cruickshank (University of Leicester, Leicester), D. Norman (Sedgwick Museum, Cambridge U.K.), and R. Smith (South African Museum, Cape Town). The photographs presented in this paper have been reproduced by courtesy of the Documentation Centre of the Muséum des Sciences naturelles (Orléans). The excavations and the examinations of track sites in Lesotho have been made possible thanks to the permits kindly issued by the relevant authorities. References ABDALA F., RIBEIRO A.M. & SCHULTZ C.L. (2001). A rich cynodont fauna of Santa Cruz do Sul, Santa Maria Formation (Middle-Late Triassic), southern Brazil. N. Jb. Geol. Paläont. Mh., 11, AMBROSE D. (1993). Maseru : an illustrated history. Morija Museum & Archives, Morija, 256 p. ANDERSON H.M. & ANDERSON J.M. (1970). A preliminary review of the biostratigraphy of the uppermost Permian, Triassic and lowermost Jurassic of Gondwanaland. Palaeont. Afric., 13, suppl., ANDERSON J.M., ANDERSON H.M. & CRUICKSHANK A.R.I. (1998). Late Triassic ecosystems of the Molteno/Lower Elliot biome of southern Africa. Palaeont., 41, 3, APPLEBY R.M., CHARIG A.J., COX C.B., KERMACK K.A. & TARLO L.B.H. (1967). Reptilia. In : W.B. HARLAND, C.H.HOLLAND, M.R. HOUSE, N.F.HUGHES, A.B.REYNOLDS, M.J.S. RUDWICK, G.E. SATTERTHWAITE, L.B.H. TARLO &E.C.WILLEY, Eds., The fossil record. Geological Society of London, London, BARRETT P.M. (2000). Prosauropod dinosaurs and iguanas : speculations on the diets of extinct reptiles. In :H.-D.SUES, Ed., Evolution of herbivory in terrestrial vertebrates : perspectives from the fossil record. Cambridge University Press, Cambridge, BATTAIL B. (1991). Les Cynodontes (Reptilia, Therapsida) : une phylogénie. Bull. Mus. natl. Hist. nat., sect. C, 13, 1-2, BATTAIL B. (1993). On the biostratigraphy of Triassic therapsid-bearing formations. N. Mex. Mus. Nat. Hist. Sci. Bull., 3, BATTAIL B. (2001). A short review of studies on cynodonts. Asoc. Paleont. Argent., Publ. Espec., 7, BENTON M.J. (1993). Reptilia. In :M.J.BENTON, Ed., The Fossil Record 2. Chapman & Hall, London, BENTON M.J. (1994). Late Triassic to Middle Jurassic extinctions among continental tetrapods : testing the pattern. In :N.C.FRASER & H.-D. SUES, Eds., In the shadow of the dinosaurs : early Mesozoic tetrapods. Cambridge University Press, Cambridge, BENTON M.J., JUUL L., STORRS G.W. & GALTON P.M. (2000). Anatomy and systematics of the prosauropod dinosaur Thecodontosaurus antiquus from the Upper Triassic of southwest England. J. Vert. Paleont., 20, 1, BONAPARTE J.F. (1970). Annotated list of the South American Triassic tetrapods. In :S.H.HAUGHTON, Ed.,Proceedings and papers of the second Gondwana symposium. Council for Scientific and Industrial Research, Pretoria, BOONSTRA L.D. (1947). Notes on some Stormberg fossil bones from Basutoland. In :G.M.STOCKLEY, Ed., Report on the geology of Basutoland. The Authority of the Basutoland Government, Maseru, BRINK A.S. (1955). On the Cynognathidae. Palaeont. Afric., 3, BROOM R. (1911). On the dinosaurs of the Stormberg, South Africa. Ann. South Afric. Mus., 7, 4, CAIRNCROSS B., ANDERSON J.M. & ANDERSON H.M. (1995). Palaeoecology of the Triassic Molteno Formation, Karoo Basin, South Africa sedimentological and palaeontological evidence. South Afric. J. Geol., 98, 4, CARROLL R.L. (1988). Vertebrate paleontology and evolution. W.H. Freeman and company, New York, 698 p. CHARIG A.J., ATTRIDGE J. & CROMPTON A.W. (1965). On the origin of the sauropods and the classification of the Saurischia. Proc. Linn. Soc. London, 176, 2, COOPER M.J. (1980). The first record of the prosauropod dinosaur Euskelosaurus from Zimbabwe. Arnoldia Zimb., 9, 3,1-17. COOPER M.J. (1982). A Mid-Permian to earliest Jurassic tetrapod biostratigraphy and its significance. Arnoldia Zimb., 9, 7, COSTEDOAT D. (1962). Etude de quelques reptiles fossiles. Mém. D.E.S., Univ. Paris, 75 p. CROMPTON A.W. & ELLENBERGER F. (1957). On a new cynodont from the Molteno beds and the origin of the tritylodontids. Ann. South Afric. Mus., 44, 1,1-13. CUNY G. (1993). Evolution des faunes de vertébrés à la limite Trias-Jurassique en France et au Luxembourg : implications à l Europe occidentale. Mém. Sci. Terre, Univ. P. et M. Curie, 93.21, 235 p. CUNY G. (1995). French vertebrate faunas and the Triassic Jurassic boundary. Palaeogeogr., Palaeoclimatol., Palaeoecol., 119, 3-4, DAMIANI R.J. (2001). A systematic revision and phylogenetic analysis of Triassic mastodonsauroids (Temnospondyli : Stereospondyli). Zool. J. Linn. Soc., 133, 4, DEMATHIEU G.R. (1980). Comparaison des informations fournies par l ichnologie des Vertébrés et par la paléontologie ostéologique dans le domaine de la chronologie. Bull. Sci. Bourgogne, 34, DEMATHIEU G.R. & HAUBOLD H. (1974). Evolution und Lebensgemeinschaft terrestrischer Tetrapoden nach ihren Fährten in der Trias. Freiberger Forschh., C, 298, DEMATHIEU G.R. & SCIAU J. (1992). Des pistes de dinosaures et de crocodiliens dans les dolomies de l Hettangien du Causse du Larzac. C. R. Acad. Sci., Paris,sér.II,315, 12, DEMATHIEU G.R & SCIAU J. (1995). L ichnofaune hettangienne d Archosauriens de Sauclières, Aveyron, France. Bull. Soc. Hist. Nat. Autun, 151, DINGLE R.V., SIESSER W.G. & NEWTON A.R. (1983). Mesozoic and Tertiary geology of southern Africa. A.A. Balkema, Rotterdam, 375 p. DUNN E.J. (1878). Report on the Stormberg coalfields. Cape of Good Hope Parliament Rep., 1878, 4,

9 TETRAPOD FAUNA OF THE KAROO BASIN (S. AFRICA) 81 ELLENBERGER F. & ELLENBERGER P. (1956). Quelques précisions sur la série du Stormberg au Basutoland (Afrique du Sud). C. R. Acad. Sci., Paris, 242, 6, ELLENBERGER F. & ELLENBERGER P. (1958). Principaux types de pistes de vertébrés dans les couches du Stormberg au Basutoland (Afrique du Sud). C. R. somm. Séances Soc. géol. France, 1958, 3-4, ELLENBERGER F. & ELLENBERGER P. (1960). Sur une nouvelle dalle à pistes de vertébrés découverte au Basutoland (Afrique du Sud). C. R. somm. Séances Soc. géol. France, 1960, 9, ELLENBERGER F., ELLENBERGER P. & GINSBURG L. (1967). The appearance and evolution of dinosaurs in the Trias and Lias : a comparison between South African Upper Karoo and western Europe based on vertebrate footprints. In : International Symposium on Gondwana Geology. UNESCO, Mar del Plata, ELLENBERGER F., ELLENBERGER P. & GINSBURG L. (1970). Les dinosaures du Trias et du Lias en France et en Afrique du Sud, d après les pistes qu ils ont laissées. Bull. Soc. géol. Fr., XII, 1, ELLENBERGER F., ELLENBERGER P., FABRE J., GINSBURG L. & MENDREZ C. (1964). The Stormberg Series of Basutoland (South Africa). In :R.K.SUNDARAM, Ed.,Proceedings of the XXII International Geological Congress. XXII International Geological Congress, New Delhi, ELLENBERGER F. & GINSBURG L. (1966). Le gisement de dinosauriens triasiques de Maphutseng (Basutoland) et l origine des sauropodes. C. R. Acad. Sci., Paris, sér. D, 262, 4, ELLENBERGER P. (1955). Note préliminaire sur les pistes et les restes osseux de vertébrés du Basutoland (Afrique du Sud). C. R. Acad. Sci., Paris,240, 8, ELLENBERGER P. (1965). Découverte de pistes de vertébrés dans le Permien, le Trias et le Lias inférieur, aux abords de Toulon (Var) et d Anduze (Gard). C. R. Acad. Sci., Paris,260, 22, ELLENBERGER P. (1970). Les niveaux paléontologiques de première apparition des mammifères primordiaux en Afrique du Sud et leur ichnologie : établissement de zones stratigraphiques détaillées dans le Stormberg du Lesotho (Afrique du Sud) (Trias supérieur à Jurassique). In :S.H.HAUGHTON, Ed.,Proceedings and papers of the second Gondwana symposium. Council for Scientific and Industrial Research, Pretoria, ELLENBERGER P. (1972). Contribution à la classification des pistes de Vertébrés du Trias : les types du Stormberg d Afrique du Sud (I). Palaeovertebrata, Mém. extr., 1972, ELLENBERGER P. (1974). Contribution à la classification des pistes de vertébrés du Trias : les types du Stormberg d Afrique du Sud. Deuxième partie : le Stormberg supérieur. Palaeovertebrata, Mém. extr., 1974, ELLENBERGER P. (1975). L explosion démographique des petits quadrupèdes à allure de mammifères dans le Stormberg supérieur (Trias) d Afrique du Sud ; aperçu sur leur origine au Permien (France et Karroo). Colloq. Int. Cent. natl. Rech. sci., 218, ELLENBERGER P. (1978). L origine des oiseaux. Mém. Trav. Ecole prat. Hautes Etud., Inst. Montpellier, 4, 1, ELLENBERGER P. (1993). Cosesaurus aviceps : Vertébré aviforme du Trias moyen de Catalogne. Mém. Univ. Sci. Tech. Languedoc, 741 p. EVANS S.E., PRASAD G.V.R. & MANHAS B. (2000). Jurassic vertebrates from India : new microvertebrates from the Kota Formation of India. J. Vert. Paleont., 20, suppl. 3, 40A. FARLOW J.O. & LOCKLEY M.G. (1993). An osteometric approach to the identification of the makers of early Mesozoic tridactyl dinosaur footprints. N. Mex. Mus. Nat. Hist. Sci. Bull., 3, FRASER N.C. & SUES H.-D. (1994). Faunal assemblages. In :N.C.FRASER &H.-D.SUES, Eds., In the shadow of the dinosaurs : early Mesozoic tetrapods. Cambridge University Press, Cambridge, GALTON P.M. (1985). The poposaurid thecodontian Teratosaurus suevicus v. MEYER, plus referred specimens mostly based on prosauropod dinosaurs, from the Middle Stubensandstein (Upper Triassic) of Nordwürttemberg. Stuttgarter Beitr. Natkd., Ser. B (Geol. Paläont.), 116, GALTON P.M. (2001a). Prosauropod dinosaurs from the Upper Triassic of Germany. In : C. FUENTES VIDARTE & F. TORCIDA FERNÁNDEZ-BALDOR, Eds.,Actas de las I Jornadas Internacionales sobre Paleontología de Dinosaurios y su Entorno. Colectivo Arqueológico-Paleontológico de Salas, Salas de los Infantes, GALTON P.M. (2001b). The prosauropod dinosaur Plateosaurus MEYER, 1837 (Saurischia : Sauropodomorpha ; Upper Triassic). II. Notes on the referred species. Rev. Paléobiol., 20, 2, GALTON P.M. & VAN HEERDEN J. (1985). Partial hindlimb of Blikanasaurus cromptoni n. gen. and n. sp., representing a new family of prosauropod dinosaurs from the Upper Triassic of South Africa. Geobios, 18, 4, GALTON P.M. & VAN HEERDEN J. (1998). Anatomy of the prosauropod dinosaur Blikanasaurus cromptoni (Upper Triassic, South Africa), with notes on the other tetrapods from the lower Elliot Formation. Paläont. Z., 72, 1-2, GAND G., VIANEY-LIAUD M., DEMATHIEU G. & GARRIC J. (2000). Deux nouvelles traces de pas de dinosaures du Trias supérieur de la bordure cévenole (La Grand-Combe, Sud-Est de la France). Geobios, 33, 5, GAUFFRE F.-X. (1993). Biochronostratigraphy of the Lower Elliot Formation (southern Africa) and preliminary results on the Maphutseng dinosaur (Saurischia : Prosauropoda) from the same formation of Lesotho. N. Mex. Mus. Nat. Hist. Sci. Bull., 3, GAUFFRE F.-X. (1996). Phylogénie des dinosaures prosauropodes et étude d un prosauropode du Trias supérieur d Afrique australe. Thèse Dr., Mus. natl. Hist. Nat., 156 p. GIERLINSKI G. (1999). Tracks of a large thyreophoran dinosaur from the early Jurassic of Poland. Acta Palaeont. Pol., 44, 2, GINSBURG L. (1966). Sur les traces des dinosaures. Sci. Avenir, 233, GINSBURG L. (1970). Les reptiles fossiles. In : P.-P. GRASSÉ, Ed.,Traité de Zoologie : Anatomie, systématique, biologie. Masson et C ie,paris,14, 3, GOW C.E. & HANCOX P.J. (1993). First complete skull of the late Triassic Scalenodontoides (Reptilia Cynodontia) from southern Africa. N. Mex. Mus. Nat. Hist. Sci. Bull., 3, GOW C.E. & LATIMER E.M.. (1999). Preliminary report of dinosaur tracks in Qwa Qwa, South Africa. Palaeont. Afric., 35, GRADSTEIN F.M. & OGG J.G. (1996). A Phanerozoic time scale. Episodes, 19, 1-2,3-5. HALLAM A. (2002). How catastrophic was the end-triassic mass extinction? Lethaia, 35, 2, HANCOX P.J. (1998). A stratigraphic, sedimentological and palaeoenvironmental synthesis of the Beaufort-Molteno contact in the Karoo basin. Ph.D. diss., Univ. Witwatersrand, 520 p. HANCOX P.J. (2000). The Continental Triassic of South Africa. Zbl. Geol. Paläont., 1998, 11-12, HAUBOLD H. (1971). Ichnia Amphibiorum et Reptiliorum fossilium. G. Fisher, Stuttgart, 124 p. HAUBOLD H. (1984). Saurierfährten. A. Ziemsen, Wittenberg Lutherstadt, 231 p. HAUBOLD H. (1986). Archosaur footprints at the terrestrial Triassic-Jurassic transition. In :K.PADIAN, Ed., The beginning of the age of dinosaurs. Cambridge University Press, London, HECKERT A.B. & LUCAS S.G. (1999). Global correlation and chronology of Triassic theropods (Archosauria : Dinosauria). Albertiana, 23, HECKERT A.B. & LUCAS S.G. (2000a). Taxonomy, phylogeny, biostratigraphy, biochronology, paleobiogeography, and evolution of the late Triassic Aetosauria (Archosauria : Crurotarsi). Zbl. Geol. Paläont., 1998, 11-12, HECKERT A.B. & LUCAS S.G. (2000b). Global correlation of the Triassic theropod record. Gaia, 15, HITCHCOCK E. (1865). Supplement to the Ichnology of New England. W. White, Boston, 96 p. HOPSON J.A. (1975). On the generic separation of the ornithischian dinosaurs Lychorhinus and Heterodontosaurus from the Stormberg Series (Upper Triassic) of South Africa. South Afric. J. Sci., 71, 10, HOPSON J.A. (1980). Tooth function and replacement in early Mesozoic ornithischian dinosaurs : implications for aestivation. Lethaia, 13, 1, HOPSON J.A. (1984). Late Triassic traversodont cynodonts from Nova Scotia and southern Africa. Palaeont. Afric., 25,