A PRIMITIVE LATE TRIASSIC `ICTIDOSAUR' FROM RIO GRANDE DO SUL, BRAZIL by
J O S EÂ F . B O N A P A R T E , J O R G E F E R I G O L O
and A N A
MARIA RIBEIRO
ABSTRACT. A primitive `ictidosaur' from lower Norian beds of southern Brazil, Riograndia guaibensis gen. et sp. nov., represented by a fragmentary skull and a lower jaw bearing a complete dentition, shows a more generalized morphology than Chaliminia from the Upper Triassic of Argentina and Pachygenelus and Diarthrognathus from the Lower Jurassic of South Africa, Canada and Greenland. The frontal bone borders the orbit, and ventrally contacts the dorsal process of the palatine. The secondary bony palate extends back to the last postcanine. I1 and i2 are reduced, whereas I2-3 and i1 are hypertrophied. Both PC 1±7 and pc 1±7 have blade-like crowns without cingula and with 5±9 small sharp cuspules. The upper postcanine crowns are semicircular in labial view with the cuspules around their margins. The lower postcanine crowns are asymmetrical with most of the cuspules dorsodistally distributed. The possible origin of this peculiar dentition is interpreted as the retention of the juvenile dentition of ancestors. The hypothesis that Riograndia guaibensis and the so-called `ictidosaurs' might have been derived from gomphodont cynodonts is presented. KEY WORDS:
Late Triassic, Ictidosaur, Brazil.
R E C E N T ®eld work in the Triassic of Rio Grande do Sul State, southern Brazil, organized by the Museu de CieÃncias Naturais (FundacËaÄo ZoobotaÃnica do RS), sponsored by the ProÂ-GuaõÂba Project, and Conselho Nacional de Desenvolvimento Cientõ®co e TecnoloÂgico (CNPq), led to the discovery of a very rich fossil locality yielding `ictidosaurians' and tiny cynodonts. The locality is 8 km west of CandelaÂria City (Text-®g. 1), a classic Triassic region known from the work of Huene (1935, 1942), Price (1947), and Barberena et al. (1985), among others. The fossiliferous level is in the upper portion of the Santa Maria Formation (Faccini 1989; Caturrita Formation of Andreis et al. 1980). Other outcrops of the Caturrita Formation bear the kannemeyeriid dicynodont Jachaleria (ArauÂjo and Gonzaga 1978), the rhynchosaur Scaphonyx (Azevedo 1982), and the traversodontid Exaeretodon (Barberena et al. 1985). Although better chronological information is needed for these levels, the presence of Exaeretodon and Scaphonyx suggests faunal relationships with the better dated Ischigualasto Formation of Argentina (Bonaparte 1973, 1997), whereas the dicynodont Jachaleria correlates better with the lower section of Los Colorados Formation in Argentina. The three genera recorded suggest a Late Triassic age, older than the upper Los Colorados Formation, which also yields rauisuchids, aeÈtosaurs, ornithosuchids, sphenosuchians, protosuchians, prosauropods, tritylodontids and `ictidosaurs' (Bonaparte 1972), as well as more recently discovered turtles (Rougier et al. 1995) and theropods (Arcucci and Coria 1997). Here we describe and discuss a new primitive `ictidosaur' which suggests that the transition from cynodonts to mammals involved many complex anatomical processes that are not yet well understood (Kemp 1982). Another paper in preparation deals with the new small cynodonts from the same locality, and a third with a revision of the previously known material of therioherpetids, an advanced family of Cynodontia. Material. The material belongs to the Fossil Vertebrate Collection of the Museu de CieÃncias Naturais, FundacËaÄo ZoobotaÃnica do Rio Grande do Sul, Porto Alegre, Brazil; abbreviation MCN-PV. [Palaeontology, Vol. 44, Part 4, 2001, pp. 623±635]
q The Palaeontological Association
624
PALAEONTOLOGY, VOLUME 44
SYSTEMATIC PALAEONTOLOGY
Infraorder CYNODONTIA Family RIOGRANDIDAE fam. nov. Diagnosis. As for the only known species. Genus RIOGRANDIA gen. nov. Type species. Riograndia guaibensis sp. nov. Derivation of name. With reference to the State of Rio Grande do Sul, Brazil, where it was collected.
Riograndia guaibensis sp. nov. Text-®gures 2±11 Derivation of name. With reference to the GuaõÂba Basin where the material was collected. Holotype. MCN-PV 2264, anterior portion of skull, from the tip of snout to the fronto-parietal contact, with complete dentition. Hypodigm. In addition to the holotype: MCN-PV 2265, an almost complete left lower jaw with complete dentition lacking the postdentary bones and MCN-PV 2271, fragment of the middle portion of right lower jaw, with three postcanines. Locality. The exposure is 8 km west of CandelaÂria City (Text-®g. 1), in a cutting on the road (BR 287) to Santa Maria, Rio Grande do Sul State, Brazil.
Diagnosis. Advanced cynodonts with the following associated derived characters: (1) absence of postfrontal; (2) weak anterior portion of the zygomatic arch; (3) dorsoventrally deep lacrimal; (4) secondary bony palate extending posterior to the last postcanine; (5) contact of ventral process of frontal with dorsal process of palatine; (6) reduced upper incisor 1 and hypertrophied lower incisor 1; (7) last upper and lower postcanine crowns imbricated, teeth implanted at an angle to the long axis of the palate; (8) postcanines blade-like with 5±9 subequal small, sharp cuspules regularly distributed on the almost semicircular border of the crown in the upper postcanine crown and in the posterodorsal border of the lower ones. The only autapomorphy is 8, which might be a character reversal. Although some derived characters are present in other advanced cynodonts, the association of the indicated characters appears to be unique to Riograndia. Brief description Skull. The holotype (MCN-PV 2264, Text-®gs 2±4, 7±9) is preserved from the tip of the snout to the anterior portion of parietals. What is preserved of the jugal indicates a weak zygomatic arch. The lacrimal, which appears to be fused to the prefrontal, is elongate, anteroposteriorly narrow, and has a pronounced intraorbital portion. The frontal borders the orbit, and its intraorbital process is in contact with the dorsal process of the palatine. The parietal extends forward, laterally covering the posterior portion of the frontal. The postorbital bone is absent, as is the postorbital bar. A reduced
BONAPARTE ET AL.: TRIASSIC `ICTIDOSAUR'
TEXT-FIG.
625
1. Map showing location of the fossil site.
septomaxilla is present, and the internarial bar is complete. The posterior opening of the infraorbital canal is bordered by the lacrimal and the dorsal process of the palatine. An infraorbital foramen is well de®ned on the maxilla below the anterior border of the orbit, and located at the level of postcanine 4. Two other foramina are located above the ®rst and second postcanines. The secondary bony palate extends back a short distance posterior to the level of the last postcanine (Text-®g. 4). The palatal suture between the premaxilla and maxilla cannot be observed. The postcanine tooth row is well set in from the lateral border of the maxilla, which shows a pronounced maxillary bulge, a feature commonly seen in gomphodont cynodonts. Lower jaw. The almost complete left lower jaw (MCN-PV 2265; Text-®gs 5, 10) is a rather robust construction, not very different from that of Pachygenelus (Crompton and Luo 1993, ®gs 4±11) with a high and thick symphysial region. The symphysis is rather large, inclined forwards, without any indication of fusion. The horizontal ramus is uniformly high and thick, with a distinct medial angular process. The coronoid process is high. The articular process is well de®ned, forming the most posterior point of the jaw; although elongate, it shows no transverse expansion for a condyle. The postdentary groove is rather large, transversely concave and extends forwards to the level of the penultimate postcanine. From there, it continues only as a narrow slit, the Meckelian canal, which reaches and expands on the symphysial plane. Dentition. The upper dentition comprises three procumbent incisors, a modest canine, and seven blade-like postcanines. Incisor 1 is very small; incisor 2 is the largest, oval in cross-section, procumbent, with a long wear facet on its lingual surface. The enamel is only present on the buccal surface, including the mesial/distal borders. Incisor 3 is rather similar in morphology to the previous one but smaller. Wear has affected its entire lingual surface.
626
PALAEONTOLOGY, VOLUME 44
2. Left lateral view of the holotype of Riograndia guaibensis (MCN-PV 2264). FR, frontal; IFR.F, infraorbital foramen; J, jugal; L, lacrimal; MX, maxilla; N, nasal; PMX, premaxilla; SMX, septomaxilla. Scale bar represents 10 mm.
TEXT-FIG.
3. Dorsolateral view of the holotype of Riograndia guaibensis (MCN-PV 2264). FR, frontal; IFR.C, infraorbital canal; J, jugal; L, lacrimal; LF, lacrimal foramina; MX, maxilla; N, nasal; P, parietal; PL, dorsal process of palatine; PMX, premaxilla; SMX, septomaxilla. Scale bar represents 10 mm.
TEXT-FIG.
The canine is separated from the incisors by a distinct diastema. Its crown is smaller than those of incisors 2 and 3. Its base is strong, lingually ¯at and buccally bulbous. In buccal view it is conical, recurved, with a sharp point. Its mesial border is sharp, keeled and smooth, whereas the distal border bears three accessory cuspules. After a short diastema, the postcanine series is closely packed together. All postcanines show the same crown structure: they are blade-like, with 6±9 aligned cuspules, each of them bordered by a shallow groove on the buccal side, without cingula. One of the cuspules, usually the central one, is a bit larger than the others (Text-®g. 2). The number of cuspules and other characters for each tooth are as follows: Pc1: 2 mesial, 1 `central' larger, and 3 distal cuspules, rather symmetrical in buccal view, slightly convex mesiodistally as well as dorsoventrally; Pc2: 3 mesial, 1 `central' larger, 4 distal cuspules, crown wider than the root, bucally convex; Pc3: 3 mesial, 1 `central' larger (broken), 4 distal cuspules, mesiodistally convex in buccal view; Pc4: 3 mesial
BONAPARTE ET AL.: TRIASSIC `ICTIDOSAUR'
627
4. Palatal view of the holotype of Riograndia guaibensis (MCN-PV 2264). Regular stipple corresponds to wear surfaces. IFR.F, infraorbital foramen; MX, maxilla; PLT, palatine; PMX, premaxilla. Scale bar represents 10 mm.
TEXT-FIG.
cuspules, `central' cuspule not de®ned by size, and 4 distal cuspules, crown mesiodistally convex; Pc6: ? mesial, 1 `central' larger (which lies slightly behind the midpoint of the crown) and more than 2 distal cuspules; Pc7: ? mesial, 1 `central' larger and 4 distal cuspules. The imbrication angle of the upper postcanines increases posteriorly, from about 5 degrees on postcanine 1±30
TEXT-FIG. 5. Medial view of the left lower jaw of Riograndia guaibensis (MCN-PV 2265), with complete dentition. AP,
angular process; GDL, groove dental lamina; MC, Meckel canal; PAP, prearticular process; PDG, Posdental groove. Scale bar represents 10 mm.
628
TEXT-FIG.
PALAEONTOLOGY, VOLUME 44
6. Tentative position of the snout (right side) and lower jaw (left side) of Riograndia guaibensis before occlusion of the postcanines.
degrees in postcanine 7. Probably an adjusted shearing occlusion was present because the anterolingual border of postcanines 5±7 shows strong lingual wear, suggesting alternating occlusion. The lower dentition is reasonably well preserved in the left dentary (MCN-PV 2265; Text-®gs 5, 10). Incisor 1 is hypertrophied in thickness as well as in length, and is procumbent. It shows two elongate basins separated by a central smooth ridge on its lingual surface (Text-®g. 5).
TEXT-FIG.
7. Stereophotograph of the holotype of Riograndia guaibensis (MCN-PV 2264), lateral view. Scale bar represents 10 mm.
BONAPARTE ET AL.: TRIASSIC `ICTIDOSAUR'
TEXT-FIG.
629
8. Stereophotograph of the holotype of Riograndia guaibensis (MCN-PV 2264), palatal view. Scale bar represents 10 mm.
Lingual wear is probably not from occlusion with upper incisors but only from food wear. The crown has enamel preserved only on its buccal side and on both medial and distal borders, which seems to be its original distribution. In a specimen of Pachygenelus (kindly shown to JFB by Prof. A. W. Crompton in Harvard University) the lingual `wear' is a morphological feature present when the new incisor erupts. Incisor 2 is much smaller and vestigial in character.
TEXT-FIG.
9. Stereophotograph of the holotype of Riograndia guaibensis (MCN-PV 2264), dorsal view. Scale bar represents 10 mm.
630
TEXT-FIG.
PALAEONTOLOGY, VOLUME 44
10. Stereophotograph of the lower jaw of Riograndia guaibensis (MCN-PV 2265), medial view. Scale bar represents 10 mm.
Incisor 3 is smaller than incisor 1, less procumbent, with two wear areas on its lingual surface. In both, the distribution of the enamel is as in incisor 1. The canine is small, nearly vertical in position, with strong wear on its distolingual side, possibly by occlusion with the upper canine. The postcanines are closely packed together, of rather uniform size, and separated from the canine by a short diastema. The lower teeth are also imbricated but to a much lesser degree than the uppers. In lateral view the distal border of each tooth covers the mesial border of the following tooth. The last postcanine is completely covered in lateral view by the coronoid process. All postcanines are blade-like, lingually convex, without any evidence of a lingual cingulum. Postcanines have 5±9 cuspules that are not very different in size, the main one being slightly larger and separated from the following one by a shallow notch. The crown is usually larger than the root. There is an incomplete subdivision of the roots in the last postcanines, suggesting an incipient bifurcation of the roots. The number and position of cuspules is as follows: pc1: 3 mesial, 1 central larger, and 2 distal; pc2: 2 mesial, 1 central larger, and 2 distal; pc3: 2 mesial, 1 central larger (broken), and 3 distal; pc4: 3 mesial, 1 central larger, and 5 distal; pc5: 3 mesial, 1 central larger, and 4 distal; pc6: 3 mesial, 1 central larger, and 5 distal; pc7: 3 mesial, 1 central larger, and 3 distal. COMPARISONS
The 12 specimens of Riograndia guaibensis collected (only three are described in this paper) suggest that the holotype and the lower jaw described here probably correspond to sub-adult specimens. The size of most of the specimens suggests a skull length of approximately 35 mm, except for one specimen, which is about half this length. The sub-adult condition is inferred because the three last postcanines are mesiodistally imbricate. In some specimens that are broken, and the roots of the postcanines exposed, there are no indications of any
BONAPARTE ET AL.: TRIASSIC `ICTIDOSAUR'
TEXT-FIG.
631
11. Lower postcanines in lateral view of Riograndia guaibensis (MCN-PV 2271), showing the cuspules and roots. Scale bar represents 10 mm.
kind of replacement, either by gomphodont postcanines as is the case in Andescynodon (GonÄi 1986; GonÄi and Goin 1987), or a frequent replacement as in a specimen of a very derived pachygenelid (Shapiro and Jenkins in press) from the Upper Triassic of Greenland. We interpret the vast majority of the specimens of Riograndia as adult or sub-adult, with a low rate of tooth replacement. The postcanine dentition of Riograndia is unlike that in other cynodonts and represents an interesting theme for discussion. Here we consider Riograndia as a possible member of the so-called `ictidosaurs' (Broom 1912) because of the characters of the incisors and canine. They are currently considered cynodonts of the family Tritheledontidae (Crompton 1958; Gow 1980; Hopson and Barghusen 1986; Shubin et al. 1991). This family appears to be valid only for Tritheledon, which has upper postcanines of transversely enlarged outline, implanted at 90 degrees to the axial plane of the palate, but not for Pachygenelus (Gow 1980) and Chaliminia which show a different morphology of the upper postcanines. Details of the crown of the upper postcanines of Tritheledon are almost unknown (Gow 1980). However, Prof. J. Hopson sent us ®ne unpublished drawings of the upper postcanines of the holotype of Tritheledon riconoi, and his interpretation is that there is a morphocline in the postcanine morphology of Pachygenelus, Diarthrognathus and Tritheledon. In opinion of one of us (JFB), this interpretation is far from well demonstrated. According to the available evidence, more reasonable would be to propose a family name based on Pachygenelus, including Chaliminia, and possibly Diarthrognathus, all of which are reasonably well represented (Bonaparte 1980). Comparisons of Riograndia appear to be meaningful only with Chaliminia from the Upper Triassic of Argentina (Bonaparte 1980); Pachygenelus (Watson 1913; Gow 1980; Shubin et al. 1991), from the Lower Jurassic of South Africa and Canada; and Diarthrognathus (Crompton 1958; Gow 1980) from the Lower Jurassic of South Africa. Comparison of the dentition of Riograndia with those of Chaliminia, Pachygenelus and Diarthrognathus indicates that the upper postcanines show strong differences, without any lingual process or buccal cingulum in the uppers, nor lingual process and cingulum in the lowers. In addition, the number of cuspules in the latter is much higher, and the main ones are only slightly larger than the others (see Text-®g. 10). The only other Late Triassic `ictidosaur' known to date, Chaliminia musteloides, has the upper postcanines of the same general morphotype as Pachygenelus (Bonaparte 1980, ®g. 5), and very different from those of Riograndia.
632
PALAEONTOLOGY, VOLUME 44
Comparison of the lower postcanines shows some similarities to the African genera, although Riograndia appears to be more generalized. The number of cuspules on each postcanine is higher than in Pachygenelus and Diarthrognathus, and the main cuspule is hardly de®ned as such. The anterior accessory cuspules in Riograndia are not present in the African genera. The dental morphotype of Riograndia provides an opportunity to discuss the possible origin and evolution of the postcanines in this advanced cynodont. Their transversely compressed, multicusped postcanines are different from the morphotype present in galesaurids, cynognathids, chiniquodontids and probainognathids (Text-®g. 13). In all of these cynodonts the postcanines bear three or four buccal cusps that are aligned mesiodistally, and sometimes a buccal cingulum on the uppers as well as a lingual one on the lowers. The possibility that a gradual transformation of this morphotype gave rise to the Brazilian `ictidosaur' dentition is remote. The juvenile dentition of the traversodontid Andescynodon mendozensis, recorded in two specimens with skull and lower jaws (GonÄi 1986, pl. 1, ®gs 1±3; GonÄi and Goin 1987) is rather similar to that of Riograndia as far as the last postcanines are concerned. In the upper jaw, they are transversely compressed with three or four cusps of similar size. In the lower jaw, they are also transversely compressed, but with seven cusps of similar size, except for the highest one that is followed mesially by two cusps and distally by four, bearing a small lingual cingulum without de®ned cusps. It is not easy to understand the meaning of such similarities. However, we hypothesize that the juvenile dentition recorded in Andescynodon was more widely distributed among gomphodont cynodonts, and it was ®xed as a paedomorphic character in adult individuals of an unknown ancestral group, giving rise to the `ictidosaurs'. The dental morphotype of Riograndia appears suitable to give rise to the more derived morphotype present in Chaliminia (Late Norian) and Pachygenelus (Early Jurassic). The main cuspule on the upper and lower postcanines of Riograndia is usually barely the largest one. On the upper postcanines of Chaliminia and Pachygenelus we hypothesize that the central cuspule became larger, but only one anterior and one posterior accessory cusp were retained. The buccal cingulum was developed at a later stage. In the lower postcanines, the main cuspule possibly increased in size and the anterior accessory cuspules were lost. The lingual expansion of the upper postcanines of Pachygenelus appears to be ancestrally connected with their imbricating implantation in the tooth row, a character also recorded in the upper and lower postcanines of Riograndia (Text-®gs 7±8, 10).
DISCUSSION
As far as we know the morphotype of the postcanine dentition present in Riograndia is unknown in cynodonts, except in part in two juvenile specimens of the traversodontid genus Andescynodon (GonÄi 1986; GonÄi and Goin 1987) discussed above, in a juvenile lower jaw of Massetognathus sp. in the collection of the Museum of Comparative Zoology, Harvard University (pers. obs.), and in Boreogomphodon (Sues and Olsen 1990; Sues et al. 1994). Here we consider Riograndia as an advanced non-mammalian cynodont, but with a more generalized dentition than any of the genera of that group already recorded. The known `ictidosaurs' from the Lower Jurassic of South Africa and Canada show strong development of the lingual cingula on the lower and presence of labial cingula on the upper postcanines, characters absent in Riograndia. The derived characters supporting its possible relationships with known `ictidosaurs' except Tritheledon (Broom 1912; Gow 1980) are: (1) dorsoventrally deep lacrimal; (2) reduced upper incisor 1 (convergence with Tritylodontidae); (3) hypertrophied lower incisor 1 (convergence with Tritylodontidae); (4) imbricating implantation of the last postcanines in the maxilla and dentary; (5) one large infraorbital foramen and two medium-sized foramina on the maxilla; (6) grooved roots on the last three postcanines (also present in Therioherpeton); (7) jaw articulation with the skull higher than the line of alveoli (also present in Pachygenelus). Primitive characters of Riograndia, which are no longer present in Pachygenelus, Chaliminia and Diarthrognathus are: (1) postcanines with poorly differentiated main cuspule and several secondary ones that are more numerous posterior to the main cuspule than anterior to it; (2) upper and lower postcanine
BONAPARTE ET AL.: TRIASSIC `ICTIDOSAUR'
633
12. Upper and lower postcanines in lateral view of `ictidosaurs' for comparison with those of Riograndia guaibensis. A, Riograndia guaibensis; B, Chaliminia; C, Pachygenelus; D, Diarthrognathus; E, Tritheledon.
TEXT-FIG.
crowns transversely narrow; (3) absence of cingula in postcanines; (4) articular process of dentary without articular condyle. It is generally accepted that the `ictidosaurs' are represented by the family Tritheledontidae (Crompton 1958; Gow 1980; Hopson and Barghusen 1986; Shubin et al. 1991; Crompton and Luo 1993; Luo 1994). However, Bonaparte (1980) argued that Pachygenelus and Chaliminia merit a separate family, Pachygenelidae, because their upper postcanines are radically different from those of Tritheledon. Shubin et al. (1991) erroneously stated that Chaliminia was referred to the Tritheledontidae by Bonaparte (1980). Chaliminia, with signi®cant anatomical characters, was placed along with Pachygenelus in the family Pachygenelidae (Bonaparte 1980), and Therioherpeton (Bonaparte and Barberena 1975), also wrongly cited by Shubin et al. (1991), was referred to the family Therioherpetidae. Riograndia probably represents a new family of advanced cynodonts, Riograndidae nov., characterized by the derived characters 1±7, associated with the more generalized characters 1±4 listed above. The possible origin of the postcanines of Riograndia appears to be the result of heterochronic processes. In Text-®gures 12 and 13, upper and lower teeth of some representative cynodonts are shown for comparison with those of Riograndia. We fail to ®nd a possible morphological relationship of Riograndia with any of them, except with juvenile traversodontids (Andescynodon, Massetognathus and Boreogomphodon) cited above. The similarities of these deciduous postcanines of Early and Late Triassic traversodontids to those of Riograndia strongly suggest that the inferred permanent dentition of the Brazilian genus may correspond to the retention of the juvenile dentition present in an unknown gomphodont cynodont, according to the available evidence (GonÄi 1986; Sues and Olsen 1990). But it might also have been an insectivorous or carnivorous cynodont, perhaps belonging to the chiniquodontid group. We tentatively favour a gomphodont ancestry because of the inset position of the upper dentition, the procumbent and reduced size of some incisors, the presence of one lower postcanine medially to the coronoid process of the lower jaw, the dorsoventrally thick horizontal ramus of the lower jaw, and the thick angular process.
TEXT-FIG. 13. Upper and lower postcanines of selected cynodonts for comparison with those of Riograndia guaibensis. A, Procynosuchus; B, Thrinaxodon; C, Cynognathus; D, Probelesodon; E, Probainognathus; F, Andescynodon (juvenile); G, Riograndia guaibensis.
634
PALAEONTOLOGY, VOLUME 44
The presence of interpterygoid vacuities in Chaliminia (Bonaparte 1980, ®g. 4), Diarthrognathus (Crompton 1958, ®g. 6) and the tritylodontid Kayentatherium (Sues 1986) is an anatomical character also recorded in the Middle Triassic advanced cynodont Probelesodon (see Bonaparte 1980, ®g. 8). Presence of the vacuities in `ictidosaurs' might be not an indication of relationships with the Therocephalia as suggested by Crompton (1958), but may be a primitive and juvenile character retained, such as in the case of the postcanines. From a more general, and consequently less documented view, of the evolutionary processes involved in the transition from advanced cynodonts to primitive mammals, heterochronic processes probably played an important role. Acknowledgements. We are indebted to the FundacËaÄo ZoobotaÃnica do Rio Grande do Sul, Projeto ProÂ-GuaõÂba for ®nancial support and to the CNPq; to Guillermo Rougier and Hans-Dieter Sues for critical reading of the ®rst manuscript; to James Hopson for important suggestions and for showing us unpublished ®gures; and to Eduardo Borsato for collecting the holotype and other signi®cant specimens. REFERENCES and MONTARDO, D. K. 1980. O Grupo RosaÂrio do Sul (TriaÂssico) no Rio Grande do Sul. Anais, 31 Congresso Brasileiro de Geologia, 2 (Sociedade Brasileira de Geologia, CamboriuÂ, Santa Catarina), 659±673.  JO, D. C. and GONZAGA, T. D. 1980. Uma nova espe  cie de Jachaleria (Therapsida-Dicynodontia) do TriaÂssico do ARAU Brasil. Actas, II Congreso Argentino de PaleontologõÂa y BioestratigrafõÂa y I Congreso Latinoamericano de PaleontologõÂa, 1 (AsociacioÂn PaleontoloÂgica Argentina, Buenos Aires), 159±174.  n Los ARCUCCI, A. B. and CORIA, R. 1997. Primer registro de Theropoda (Dinosauria ± Saurischia) de la Formacio Colorados (Triasico Superior, La Rioja, Argentina). XIII Jornadas Argentinas de PaleontologõÂa de Vertebrados, La Rioja. ResuÂmenes, Ameghiniana 34 (AsociacioÂn PaleontoloÂgica Argentina, Buenos Aires), 531.  ssico do Rio Grande do AZEVEDO, S. A. K. 1982. Scaphonyx sulcognathus (sp. nov.) um novo rincossaurõÂdeo do Neotria Sul, Brasil. MSc dissertation, Universidade Federal do Rio Grande do Sul, Curso de PoÂs-GraduacaÄo em GeocieÃncias, Porto Alegre, Brasil, 86 pp.  JO, D. C., LAVINA, E. L. and AZEVEDO, S. A. K. 1985. O estado atual do conhecimento sobre os BARBERENA, M. C., ARAU tetraÂpodes permianos e triaÂssicos do Brasil Meridional. In VIII Congresso Brasileiro de Paleontologia (1983), Rio de Janeiro. ColetaÃnea de Trabalhos PaleontoloÂgicos, SeÂrie Geologia, SecËaÄo Paleontologia e Estratigra®a, BrasõÂlia, 27 (2), 21±28.  podos del sector superior de la FormacioÂn Los Colorados, La Rioja, Argentina (TriaÂsico BONAPARTE, J. F. 1971. Los tetra Superior). Opera Lilloana, 26, 1±183. ÐÐ 1973. Edades/reptil para el TriaÂsico de Argentina y Brasil. Actas, V Congreso GeoloÂgico Argentino (AsociacioÂn GeoloÂgica Argentina, Buenos Aires), 3, 93±129. ÐÐ 1980. El primer ictidosaurio (Reptilia-Therapsida) de AmeÂrica del Sur, Chaliminia musteloides, del TriaÂsico Superior de La Rioja, Argentina. Actas, II Congreso Argentino de PaleontologõÂa y BioestratigrafõÂa y I Congreso Latinoamericano de PaleontologõÂa, 1 (AsociacioÂn PaleontoloÂgica Argentina, Buenos Aires), 123±133. ÐÐ 1997. El TriaÂsico de San Juan-La Rioja, Argentina y sus dinosaurios. Museo Argentino de Ciencias Naturales, Buenos Aires, 190 pp. ÐÐ and BARBERENA, M. C. 1975. A possible mammalian ancestor from the Middle Triassic of Brazil (TherapsidaCynodontia). Journal of Paleontology, 49, 931±936. BROOM, R. 1912. On a new type of cynodont from the Stormberg. Annals of the South African Museum, 7, 334±336. CROMPTON, A. W. 1958. The cranial morphology of a new genus and species of ictidosauran. Proceedings of the Zoological Society of London, 130, 183±216. ÐÐ and LUO, Z. 1993. Relationships of the Liassic mammals Sinoconodon, Morganucodon oehleri and Dinnetherium. 30±44. In SZALAY, F. S., NOVACEK M. J. and MCKENNA, M. C. (eds). Mammal phylogeny. Mesozoic differentiation, multituberculates, monotremes, early therians, and marsupials. Springer-Verlag, New York.  ssico do Rio Grande do Sul. Uma anaÂlise sob o ponto de vista das sequÈeÃncias FACCINI, U. F. 1989. O Permo±Tria deposicionais. DissertacËaÄo de Mestrado em GeocieÃncias, Universidade Federal do Rio Grande do Sul, 121 pp. Ä I, R. 1986. Reemplazo de dientes postcaninos en Andescynodon mendozensis Bonaparte (Cynodontia, TraversoGON dontidae). Actas, IV Congreso Argentino de PaleontologõÂa y Bioestratigra®a, 2 (Mendoza), 7±14. ÐÐ and GOIN, F. J. 1987. El origen de los poscaninos gonfodontes de Andescynodon mendozensis Bonaparte (Cynodontia, Traversodontidae). Ameghiniana, 24, 235±239. ANDREIS, R. R., BOSSI, G. E.
BONAPARTE ET AL.: TRIASSIC `ICTIDOSAUR'
635
1980. The dentitions of the Tritheledontidae (Therapsida: Cynodontia). Proceedings of the Royal Society of London, Series B, 208, 461±481. HOPSON, J. A. and BARGHUSEN, H. 1986. An analysis of therapsid relationships. 83±106. In HOTTON, N., MCLEAN, P. D., ROTH, J. J. and ROTH, E. C. (eds). The ecology and biology of mammal-like reptiles. Smithsonian Institution Press, Washington DC. È damerikanischen Gondwanalandes. C. M. Beck, MuÈnchen, 332 pp. HUENE, F. VON 1935±1942. Die fossilen Reptilien des su KEMP, T. S. 1982. Mammal-like reptiles and the origin of mammals. Academic Press, London, 363 pp. LUO, Z. 1994. Sister-group relationships of mammals and transformation of diagnostic mammalian characters. 98±128. In FRASER, N. C. and SUES, H.-D. (eds). In the shadow of the dinosaurs. Cambridge University Press, Cambridge. Â ssico do Rio Grande do Sul. DivisaÄo de Geologia e Mineralogia, PRICE, L. I. 1947. Um procolofonõÂdeo do Tria Departamento Nacional da Producao Mineral, Boletim, 122, 26 pp. ROUGIER, G. W, FUENTE, M. S. and ARCUCCI, A. B. 1995. Late Triassic turtles from South America. Science, 268, 855±858. SHAPIRO, M. D. and JENKINS, F. A. in press. A new cynodont from the Greenlandic Triassic: dental succession and multiple roots. In: A symposium in honor of Fuzz Crompton. Harvard University Press, Cambridge. SHUBIN, N. H., CROMPTON, A. W, SUES, H. D. and OLSEN, P. E. 1991. New fossil evidence on the sister-group of mammals and early Mesozoic faunal distributions. Science, 251, 1063±1065. SUES, H.-D. 1986. The skull and dentition of two tritylodontid synapsids from the Lower Jurassic of western North America. Bulletin of the Museum of Comparative Zoology, 151, 217±268. ÐÐ and OLSEN, P. E. 1990. Triassic vertebrates of Gondwanan aspect from the Richmond Basin of Virginia. Science, 249, 1020±1023. ÐÐ ÐÐ and KROEHLER, P. A. 1994. Small tetrapods from the Upper Triassic of the Richmond basin (Newark Supergroup), Virginia. 161±170. In FRASER, N. C. and SUES, H.-D. (eds). In the shadow of the dinosaurs. Early Mesozoic tetrapods. Cambridge University Press, Cambridge. WATSON, D. M. S. 1913. On a new cynodont from the Stormberg. Geological Magazine, 10, 145±148. GOW, C. E.
JOSEÂ F. BONAPARTE
Museu de CieÃncias Naturais FundacËaÄo ZoobotaÃnica do Rio Grande do Sul Current address Museo Argentino de Ciencias Naturales Dept. of Paleontology Buenos Aires, Argentina JORGE FERIGOLO ANA MARIA RIBEIRO
Typescript received 20 March 2000 Revised typescript received 8 September 2000
Museu de CieÃncias Naturais FundacËaÄo ZoobotaÃnica do Rio Grande do Sul Dr Salvador FrancËa, 1427 Jardim BotaÃnico 90.690-000 Porto Alegre RS, Brazil e-mail
[email protected] [email protected]