Arquivos do Museu Nacional, Rio de Janeiro, v.66, n.l, p.63-82, jan./mar.2008 ISSN 0365-4508 MORPHOLOGY OF THE DENTAL CARINAE IN MARILIASUCHUS AMARALI (CROCODYLOMORPHA, NOTOSUCHIA) AND THE PATTERN OF TOOTH SERRATION AMONG BASAL MESOEUCROCODYLIA 1 (With 7 figures) MARCO BRANDALISE DE ANDRADE 2’3 REINALDO J. BERTINI 2>4 ABSTRACT. Carinated teeth are common in Mesoeucrocodylia, and the occurrence of denticles over the carinae is related to high predacious species, often referred as ziphodont. This characteristic is broadly recognized as homoplastic. Carinae morphology is cryptic, difficult to be studied under common techniques, and Scanning Electronic Microscopy (SEM) allows the access to detailed information, offering a higher degree of confidence. Previous SEM study allowed the recognition of true/false ziphodont patterns, according to the morphology of the denticles, but such studies on gondwanan mesoeucrocodyles are uncommon. Mariliasuchus amarali is an Upper Cretaceous notosuchian mesoeucrocodyle from South America (Bauru Group, Brazil), with carinated teeth and specialized dentition. Its geological and biochronological distribution are reappraised. SEM study of two teeth shows carinae composed of isolated tuberous anisomorphic true denticles, supporting previous study. Enamel ornamentation does not develop over the carinae, and fabric becomes anastomosed in middle and posterior teeth. Carinae only occur in posterior molariform teeth, related to food processing. Morphological variability of Mariliasuchus is commented, focusing on dentition. Overall characteristics, molariform morphology and wear planes support a non-predacious habit for Mariliasuchus. Mariliasuchus pattern could not be related to true/false ziphodont patterns, either by morphology or function, and is defined as ziphomorph. Ziphomorph pattern is evaluated within the range of mesoeucrocodyles. The detailed study of homoplastic characteristics, such as dental carinae, may provide useful apomorphic information for cladistic analysis. Key words: Tooth morphology. Crocodylomorpha. Notosuchia. Cretaceous. Ziphomorphy. RESUMO. Morfologia das carenas dentárias em Mariliasuchus amarali (Crocodylomorpha, Notosuchia) e a variação no padrão de carena em dentes de Mesoeucrocodylia basais. Dentes carenados são comuns em Mesoeucrocodylia, e a ocorrência de dentículos sobre a carena está relacionada a espécies altamente predatórias, frequentemente referidas como zifodontes. Esta característica é amplamente reconhecida como homoplástica. A morfologia da carena é críptica, difícil de ser estudada através de técnicas comuns, e Microscopia Eletrônica de Varredura (MEV) permite acesso a informações detalhadas, oferecendo um grau maior de confiança. Estudos anteriores em MEV permitiram o reconhecimento de padrões zifodontes verdadeiro/falso, de acordo com a morfologia dos dentículos, porém este tipo de estudo em mesoeucrocodilos gondwânicos é incomum. Mariliasuchus amarali é um mesoeucrocodilo gondwânico do Cretáceo Superior da América do Sul (Grupo Bauru, Brasil), com dentes carenados e dentição especializada. Suas distribuições geológica e biocronológica são reavaliadas. Estudos em MEV de dois dentes mostraram que carenas são compostas por dentículos verdadeiros, tuberosos e anisomorfos, suportando estudo anterior. Ornamentação não se desenvolve sobre a carena, e o padrão se torna anastomosado em dentes médios e posteriores. Carenas ocorrem apenas em dentes molariformes, relacionados ao processamento do alimento. A variabilidade morfológica de Mariliasuchus é comentada, com foco em dentição. Características gerais, morfologia dos molariformes e a presença de planos de desgaste suportam um hábito não predatório para Mariliasuchus. O padrão de carenas de Mariliasuchus 1 Submitted on September 14, 2006. Accepted on February 19, 2008. This paper was a contribution to the II Congresso Latino-americano de Paleontologia de Vertebrados, held in August, 2005, in Rio de Janeiro (RJ, Brazil). 2 Universidade Estadual Paulista, Instituto de Geociências e Ciências Exatas, Departamento de Geologia Aplicada. Campus Rio Claro, Caixa Postal 178, 13506-900, Rio Claro, SP, Brazil. MBA support by MSc Scholarship (2003-2005) from Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES), Brazil. 3 University of Bristol, Faculty of Sciences, Department of Earth Sciences. BS8 1RJ, Bristol, Avon, England, United Kingdom. E-mails: [email protected]; [email protected]. Financial support by the Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq - Grant n° 200381/2006-8), Brazil. 4 E-mail: [email protected]. 64 M.B.ANDRADE & RJ.BERTINI não pôde ser relacionado aos padrões zifodontes verdadeiro/falso, tanto por morfologia quanto por função, sendo aqui definido como zifomorfo. O padrão zifomorfo é avaliado dentro do espectro dos Mesoeucrocodylia. O estudo detalhado de características homoplásticas, como o carenamento de dentes, pode fornecer informações apomórficas úteis para análises cladísticas. Palavras-chaves: Morfologia dentária. Crocodylomorpha. Notosuchia. Cretáceo. Zifomorfia. INTRODUCTION Clemens etal, 2003; Pol, 2003; Sereno etal, 2003; Turner & Calvo, 2005; Turner, 2006; Zaher et al, 2006). The morphology of the carinae, present in Features regarding dentition are widely used in evolutionary studies, including crocodylomorphs several species, is of particular interest. Ortega et (e.g., Woodward, 1896; Rusconi, 1933; Colbert, 1946; al (2000) defined the Ziphosuchia as a group of Price, 1950; Berg, 1966; Kuhn, 1968; Edmund, 1969; Mesoeucrocodylia comprised by Notosuchus, Langston, 1956, 1975; Gasparini, 1971, 1972; Libycosuchus, and Sebecosuchia, which should have the ziphodont dentition, defined by the carinae Buffetaut, 1976, 1979, 1982; Benton & Clark, 1988; morphology. Nevertheless, there is not much Carvalho & Campos, 1988; Clark et al., 1989; Bonaparte, 1991; Buffetaut & Marshall, 1991; Ortega agreement on this characterization. As Turner etal, 1993, 2000; Carvalho, 1994; Clark, 1994; Wu (2006) pointed out, for long time the use of & Sues, 1996; Wu etal, 1995; Gomani, 1997; Carvalho ziphodont dentition is considered to be of limited & Bertini, 1999; Buckley etal, 2000; Riff & Kellner, value as phylogenetic information (Langston, 1956; 2001; Prasad & Broin, 2002; Clemens et al, 2003; Berg, 1966; Hecht & Archer, 1977; Turner & Calvo, Pol, 2003; Sereno etal, 2003; Turner & Calvo, 2005; 2005; Zaher et al, 2006). Turner, 2006; Zaher et al, 2006). From general Although used in previous studies (Langston, 1956; aspects [e.g., arrangement between dental series) Berg, 1966), the classical ziphodont dentition to very specific morphological features (e.g., (Langston, 1975) is defined as crocodylomorph teeth morphology of the carinae), information proved to with morphology similar to equivalents observed be both useful and controversial to phylogenetic and in carnivorous dinosaurs. The concept is based on paleoecologic aspects. Crocodylomorph teeth have characteristics such as general tooth shape, apex a wide range of morphological variation, including morphology and presence of carinae. Ziphodont number and arrangement of cusps, inclination and carinae are typically serrated and formed by orientation of the apex, overall shape in lateral view, isolated denticles. This idea was posteriorly compression of the crown, compression of the root modified by Prasad & Broin (2002), restricting the and presence of cingulus, base-to-apex definition to the composition of the dental carinae, ornamentation, amongothers (Price, 1950; Carvalho, which allowed: a) some morphological variability 1994; Wu et al, 1995; Wu & Sues, 1996; Gomani, in dental series and specimens; b) the recognition 1997; Buckley et al, 2000; Riff & Kellner, 2001; of other crocodylomorphs as ziphodont species Nobre & Carvalho, 2002; Vasconcellos & Carvalho, (Fig.l). Examples of ziphodont crocodylomorphs, 2005; Elias, 2006; Turner, 2006; Zaher etal, 2006). by this definition, include Iberosuchus, Sebecus, The variations include convergences with Pristichampsus, Hamadasuchus, and cf. mammalian dentition (Carvalho & Campos, 1988; Araripesuchus wegeneri. Clark etal, 1989; Bonaparte, 1991; Carvalho, 1994; Prasad & Broin (2002) also described another Wu & Sues, 1996; Wu et al, 1995; Gomani, 1997), pattern, defined as false-ziphodont dentition, with a similar nomenclature (incisiforms, which is attributed to mesoeucrocodylians, such caniniforms, and molariforms) referring to as Asiatosuchus, Trematochampsa, Sarcosuchus, specialized teeth. and Sphagesaurus. False-ziphodont teeth are The term “ziphodont” have long been applied to characterized by the presence of crenulations, Mesoeucrocodylia, including several genera from composed by the extension of the enamel ridges a broad range of families. Characters related to the over the carina. These ridges are often irregular, ziphodont dentition are included (explicitly or not) creating an anastomosing fabric over the labial as part of several works in phylogenetics (e.g., and lingual teeth surface. When this fabric Benton & Clark, 1988; Clark et al, 1989; Clark, reaches out up to the mesial and distai borders, it 1994; Wu & Sues, 1996; Wu et al, 1995; Gomani, modifies the morphology of the carinae, which 1997; Buckley et al, 2000; Ortega et al, 2000; usually have a continuous and uniform structure. Arq. Mus. Nac., Rio de Janeiro, v.66, n.l, p.63-82, jan./mar.2008 MORPHOLOGY OF THE DENTAL CARINAE IN MARI LI A S U CHUS AMARAL! 65 The resulting surface becomes crenulated, giving the include detailed descriptions and images from false impression, under observation by simple optical dinosaur teeth, but most of them are almost totally resources, that the carina is composed by several dedicated to Laurasian theropods (Farlow, 1987; isolated denticles (Prasad & Broin, 2002). Thispattern Currie et al, 1990; Farlow et al, 1991; Fiorillo & seems to be analogous to the true ziphodont Currie, 1994; Rauhut & Werner, 1995; Buscalioni et morphology, but as Prasad & Broin (2002) point out, al, 1996; Franco-Rosas, 2000). In the other hand, its structure is completely different (Fig.2). Prasad & there are few publications dedicated to the dental Broin (2002) stress that the identification of patterns morphology in crocodylomorphs, with the help of is especially difficult without sufficiently magnified SEM (e.g., Carvalho, 1994; Legasa et al., 1994; Prasad views, and the use of Scanning Electronic Microscopy & Broin, 2002; Andrade, 2005; Elias, 2006), and (SEM) can prove to be a valuable tool. information about Gondwanan taxa is very limited. The morphological description of the carina as to While this kind of information may be significant two basic types, ziphodont and false-ziphodont, for evolutionary studies to crocodylomorphs, there seems to be limited when the wide range of is still a huge lack of knowledge regarding the morphology types is taken into consideration. In fact, descriptions of teeth from South-American taxa. the nature of the denticles and their distribution Among the South-American mesoeucrocodyles, the over the crown, seems to be much wider. Also, Brazilian Maríliasuchus amarali Carvalho & Bertini, several basal Mesoeucrocodylia were heterodont, 1999, from the Campanian of the Bauru Group and morphologic variation can be expected along (Araçatuba/Adamantina formations) is well known the series. Thus, teeth morphological variation in from several specimens (Carvalho & Bertini, 1999; crocodylomophs should not be represented solely Andrade, 2005; Vasconcellos & Carvalho, 2005, by “theropod-like” and “false-theropod-like” 2006; Zaher et al, 2006). Tooth morphology was morphologies. studied by Zaher et al (2006), under common Furthermore, there seems to be a sample bias optical techniques, describing the serrations as regarding information from Scanning Electronic “ composed of a series ofround tubercles, instead of Microscopy (SEM). Several scientific contributions sharp denticlespresent in ziphodont crocodiliforms”. Fig. 1- Ziphodont crocodylomorphs, showing major features of the true ziphodont pattern: A) Sebecus icaeorhinus skull (above), with detail of the carina from MNHN (P) VIV-69, Sebecus sp. (below); B) cf. Araripesuchus wegeneh, GDF 700, holotype (above), with detail of its maxillary teeth bearing carinae, composed of true denticles (below). Scale bars = O.lmm (A); lOmm (B). (A - adapted from Colbert, 1946 and Prasad & Broin, 2002; B - adapted from Ortega et al, 2000 and Turner, 2006). Arq. Mus. Nac., Rio de Janeiro, v.66, n.l, p.63-82, jan./mar.2008 66 M.B.ANDRADE & RJ.BERTINI Fig.2- False-ziphodonty in Asiatosuchus: A) general aspect of MNHN (P) AG-20, caniniform tooth; B) apex of the tooth MNHN (P) BR-15230, showing superficial ornamentation; C) detail of the carina of the tooth MNHN (P) BR-15230, showing ornamentation composed by enamel ridges that develop over the carina, resembling denticles of ziphosuchian Mesoeucrocodylia. Note that such condition is very difficult to identify without Scanning Electronic Microscopy. Scale bars = 10mm (A); 0.5mm (B-C). (Adapted from Prasad & Broin, 2002). Here we study teeth from Mariliasuchus amarali enamel ridge; FMP, maxillo-palatinae fenestra; under Scanning Electronic Microscopy, review the FSO, suborbital fenestra; laf, labial face; lif, lingual information provided by Zaher et al. (2006) and face; ma, maxillary tooth; Mx, maxilla; Pal, palatine; compare this particular morphology to the typical Pmx, premaxilla; ro, tooth root; Sp, splenial. ziphodont dentition. Functional aspects of Mariliasuchus are explored, to further demonstrate Material that this morphology is truly diverse from the ziphodont pattern. Mariliasuchus amarali is a Notosuchia (sensu Gasparini, 1971) and most probably a Notosuchidae (Carvalho & Bertini, 1999; Andrade, 2005; Fiorelli MATERIAL AND METHODS & Calvo, 2005; contra Carvalho et al, 2004; Zaher et al, 2006), as Notosuchus terrestris Woodward, 1896. Remains come from several outcrops, at the Abbreviations vicinities of the Marília City (Nava, 2004), and are Institutional. DES, Department of Earth Sciences, currently housed by several institutions, including University of Bristol, Bristol, United Kingdom; GDF, MUZUSP, MN, UFRJ, and URC (Andrade, 2005; MNHN (P) AG, MNHN (P) BR, MNHN (P) VIV, Muséum Vasconcellos & Carvalho, 2005, 2006; Zaher et al, National d’Histoire Naturelle, Paris, France; IGCE- 2006). It is agreed that Mariliasuchus comes from UNESP, Departamento de Geociências e Ciências the Late Cretaceous of Bauru Group, in the Exatas, Universidade Estadual Paulista, Rio Claro, vicinities of Marília City (Carvalho & Bertini, 1999, Brazil; MEF, Museo Paleontologico Egidio Feruglio, 2000; Andrade, 2005; Vasconcellos & Carvalho, Trelew, Argentina; MN, Museu Nacional, UFRJ, Rio 2005, 2006; Zaher etal, 2006). de Janeiro, Brazil; MUZUSP, MZSP-PV, Museu de We studied two well-preserved isolated teeth from Zoologia, Universidade de São Paulo, São Paulo, Mariliasuchus amarali under Scanning Electronic Brazil; UFRJ, Universidade Federal do Rio de Microscopy. They were both found in close Janeiro, Rio de Janeiro, Brazil; URC, Museu de association to well-preserved and partially Paleontologia e Estratigrafia “Prof. Dr. Paulo Milton articulated M. amarali cranial and post-cranial Barbosa Landim”, Universidade Estadual Paulista, remains (URC R*67, URC R«68, URC R*69). It is Rio Claro, Brazil. not certain if the teeth come from either one of those Anatomical. c, hypertrophied caniniform tooth; cr, specimens or from a fourth individual. Furthermore, tooth crown; de, carina denticle; Den, dentary; er, they could not have come from URC R*67, as this Arq. Mus. Nac., Rio de Janeiro, v.66, n.l, p.63-82, jan./mar.2008 MORPHOLOGY OF THE DENTAL CARINAE IN MARILIASUCHUS AMARALI 67 specimen has a complete dental series preserved. Zaher etal. (2006), describing the geologic settings of The isolated teeth were respectively identified as URC Maríliasuchus, refers to a single locality for all R*74 (caniniform) and URC R*75 (molariform) by specimens, at the left margin of the “(.. .)Agua Formosa comparison with URC R»67 and URC R»68. All these creek (coordinates 22°20’28”S and 49°56’46”W), 10 specimens, including the teeth, came from the type- km south from the urban area ofMarília (...)” (Zaher et locality of the Rio do Peixe outcrop. The specimens al, 2006; p.2, lst column, 2nd §). In the same paper, of MN 6298-V and MN 6756-V were also studied for the authors provided locality and horizon as “(...) a further comparison. MN 6298-V is composed of a road cut at the left margin of the Peixe River, 18 km partial skull, without the mandible, while MN 6756- from the city of Marília, (...) from the upper part of the V is composed of a well-preserved set of skull and Adamantina Formation, Bauru Group” (Zaher et al, mandible. This last specimen shows lateral 2006). Differences of distance are clearly due to the compression (Zaher et aí, 2006). In Zaher etal (2006; way they were obtained, as 10km is the distance in a p.7, 2nd column, lines 8-15), the Identification of straight line, taken from maps, and 18km can be these specimens is changed, as MN 6298-V is understood as the distance taken using main roads identified as MN 6756-V and vice versa. necessary to access the outcrop. The locality itself is well known as Rio do Peixe outcrop from previous Geological Settings works (Carvalho & Bertini, 1999, 2000; Andrade, 2005) and there is no question as to which river is related A bibliographic review of Maríliasuchus shows some the outcrop. The Peixe River spring is located differences of interpretation on the origin of the northeastern to the GPS location provided by Zaher specimens. Carvalho & Bertini (2000), Vasconcellos et al. (2006), closer to Garça City. From its spring, & Carvalho (2005), Candeiro & Martinelli (2006), the Peixe River flows to the western, passing through and Zaher etal (2006) considered that the remains the Maríliasuchus locality and continuing West- came from the Adamantina Formation. Andrade Northwestern to the Parana River, without changing (2005) and Vasconcellos & Carvalho (2006) itsname [e.g., Batezelli, 1998). Further disagreement described them as originated from the Araçatuba/ comes from the collection of Maríliasuchus. Most Adamantina formations. Divergences may be papers refer to the same Rio do Peixe outcrop, but partially explained because of the different referring to one or few specimens (Carvalho & Bertini, definitions of the Araçatuba sedimentary unit. 1999, 2000; Andrade, 2005; Vasconcellos & Carvalho, These sediments have been usually considered as 2005, 2006). Zaher et al (2006) declare that all specimens came from the same location, which is a the base of the Adamantina Formation (as in Kellner broad definition, as location’ could define ‘outcrop’, & Campos, 1999; Dias Brito et al, 2001; Candeiro & Martinelli, 2006). Barcelos (1984) referred this but also ‘the vicinities of Marília City’. Nava (2004), geological unit as Member Araçatuba. Its original on the other hand, clearly States that Maríliasuchus remains have been found in at least four sites in the definition as Araçatuba Formation (Zaine etal., 1980) same region, and many specimens have been was most recently modified (Batezelli, 1998, 2003; recovered from these outcrops. It is possible that Batezelli etal, 1999, 2003; Fernandes etal, 2003), extending the area of occurrence and lithologic Maríliasuchus specimens were collected in other outcrops, but unfortunately, localities and specimens column. Although Carvalho & Bertini (1999, 2000) and Vasconcellos & Carvalho (2005) use the traditional were not individually identified by Nava (2004), preventing further discussion. Nevertheless, holotype definition (Araçatuba as a lithofacies of the and URC specimens came from the type locality, Adamantina Formation), it should be noticed that vicinal road that gives access to Fazenda Doreto, specimens are always preserved in close association Marília Municipality, 10km from the municipal with pelitic sediments (Carvalho & Bertini, 1999, 2000). Vasconcellos & Carvalho (2006) considered headquarters, as described by Carvalho & Bertini (1999). No other locality has been officially identified. difficulties in the determination of the units and limits, assuming Araçatuba/ Adamantina Formation Some divergences regard the provenance of the for the UFRJ specimens. Nobre & Carvalho (2006) materiais in the lithologic column. The Rio do Peixe directly address the problem and State that outcrop includes only the Araçatuba and the Adamantina sediments on the margins of the Peixe Adamantina formations. The limits of these River, at the base of the Rio do Peixe outcrop, are the sedimentary units are not clearly defined, as the same as the Araçatuba Formation, as defined by Araçatuba Formation broadly interbeds with the Batezelli etal (1999) and Fernandes etal (2003). Adamantina Formation [e.g., Batezelli, 1998, 2003). Arq. Mus. Nac., Rio de Janeiro, v.66, n.l, p.63-82, jan./mar.2008 M.B.ANDRADE & RJ.BERTINI At least the holotype, the UFRJ specimens, and the 1999, 2000; Andrade, 2005; Vasconcellos & Carvalho, URC specimens were recovered from a horizon close 2005, 2006; Nobre & Carvalho, 2006). to the bottom of the lithologycal column (Carvalho & Further debate also exists on the age of the Upper Bertini, 1999, 2000; Vasconcellos & Carvalho, 2006; Cretaceous deposits from the Bauru Group. Dias- Nobre & Carvalho, 2006), where there is a significative Brito et al. (2001) argues for a Turonian- contribution of siltic matrix over sandstone Maastrichtian age for the Bauru Group, with a (Araçatuba Formation sensu Batezelli, 1998; Batezelli Campanian depositional hiatus, indicating an early et al, 2003). As discussed previously, most studies age for the Araçatuba Formation, possibly Turonian. agree that sediments at the base of the Rio do Peixe The proposal by Dias-Brito et al (2001) is widely outcrop, where Mariliasuchus is originated, represents adopted (Vasconcellos & Carvalho, 2005, 2006; Nobre the contact between the Araçatuba and Adamantina & Carvalho, 2006; Zaher etal, 2006). Nevertheless, formations, thus close to the bottom of the the existence of several gradational contacts between Adamantina Column (Carvalho & Bertini, 1999, 2000; the Adamantina and Marília formations (Batezelli, Andrade, 2005; Vasconcellos & Carvalho, 2005, 2006; 1998, 2003; Batezelli etal, 1999, 2003), recognized Nobre & Carvalho, 2006). A different statement is by Zaher et al (2006), implies that a Campanian provided by Zaher et al (2006), which consider the depositional hiatus is unlikely to occur. Zaher et al fácies association as representative of the upper part (2006) considers a Campanian to Maastrichtian age of the Adamantina Formation, close to the contact of for Mariliasuchus, although accepting a modified the Marília Formation (Zaher et al, 2006). The version of the model proposed by Dias-Brito et al specimens are assigned in fact to four horizons (Zaher (2001), and considering the lithologic column from et al, 2006) in the columnar section of the referred the type-locality as representative of the upper outcrop, each one showing a different lithology. These Adamantina section. are always rich in fine grained sediments, where Correlations based on charophytes, ostracods, and brown/dark-brown shale interclasts are usually vertebrates (Gobbo-Rodrigues et al, 2000a, 2000b, associated, and also a metric mudstone layer (Zaher 2000c; Gobbo-Rodrigues, 2001; Santucci & Bertini, et al, 2006). This description matches the upper 2001) indicate that the Araçatuba Formation was section of the Araçatuba Formation (sensu Batezelli, most probably Campanian (Fig.3), rather than 1998), and its intergrading contact with the Turonian. Although the age attributed for Adamantina Formation. Mariliasuchus is similar for Zaher et al (2006) Although disagreement is present in the (Campanian-Maastrichtian), both models represent bibliography, a conservative approach is here different interpretations of the data available. preferred. URC specimens carne from the same locality and horizon provided for the holotype, and RESULTS AND DISCUSSION possibly for several other specimens, on the margins of the Peixe River, Rio do Peixe outcrop. The sediments associated with these specimens Description of the Material have been referred to as the Adamantina URC R*74 shows a caniniform morphology (Fig.4), Formation (Carvalho & Bertini, 1999, 2000; slightly curved, the apex not acute. URC R*75 is a Vasconcellos & Carvalho, 2005), and several typical molariform (Fig.5) although not particularly studies (Batezelli, 2003; Batezelli et al, 1999, well-developed. In both elements, there is no 2003; Nobre & Carvalho; 2006) recognized the constriction between crown and root, though same sediments as the gradational contact differences of color and surface allowed the between the Araçatuba Formation sensu Batezelli, recognition of the actual boundaries. 1998. Type-horizon is therefore considered as the Araçatuba/Adamantina formations, rather than URC R»74 is small and could have been positioned to the upper Adamantina column. As the as an anterior premaxilary tooth, but not the Araçatuba and Adamantina formations are hipertrophyed caniniform. It is comparable in size considered to be (at least) partially synchronic and general morphology to the regular premaxilary (Batezelli, 1998, 2003; Batezelli etal, 1999, 2003; caniniforms of URC R*67. The crown is lightly Fernandes et al, 2003), the occurrence of the same curved, with a circular cross-section and no lateral species in both sedimentary units is likely. In this compression. There was no evident difference context, we understand that there is no between the lingual and labial surfaces. This tooth disagreement with most studies (Carvalho & Bertini, does not show any kind of serration, either in the Arq. Mus. Nac., Rio de Janeiro, v.66, n.l, p.63-82, jan./mar.2008 MORPHOLOGY OF THE DENTAL CARINAE IN MARILIASUCHUS AMARALI 69 Fig.3- Mariliasuchus amarali and its geographical range: A) general aspect of the skull from URC R*67; B) artistic reconstruction of Mariliasuchus; C) map showing the geographical distribution of the sediments from the Bauru Group; D) lithologic column for the State of São Paulo, showing type-locality of holotype, UFRJ and URC specimens. Bar: 10mm (A). (B - illustration by Felipe A. Elias; C - modified from Fernandes & Coimbra, 1996; D - adapted from Batezelli et al, 2003). mesial or the distai surfaces. It rather had a a maxillary tooth, or one of the posterior smooth irregular surface, where base-to-apex mandibular teeth. Based on the morphology and ridges develop. The ridges are proportionally low comparison to URC R«68, it is more likely that the and wide, are present through most of the crown specimen represents the fifth left mandibular tooth. length, and probably represent enamel The crown is lanceolated in lateral view, but short ornamentation. The ridges do not progress to the and with a blunt apex. The lingual and labial apex, which seems to be a natural characteristic, surfaces are different, with a “D-shaped” cross- as there is no indication that they were worn out section. The lingual surface is not as convex as the or suffered physical erosion. The very apex is the labial surface. Considered this interpretation, neither round, nor acute. It seems to have been serrations developed preferentially on the mesial worn out in a single, though irregular, plane. surface, while the distai surface shown a smoother URC R»75 is also small, and could have been either area and denticles were not so easily characterized. Arq. Mus. Nac., Rio de Janeiro, v.66, n.l, p.63-82, jan./mar.2008 70 M.B.ANDRADE & RJ.BERTINI Fig.4- Labial view of the caniniform tooth URC R«74, from Mariliasuchus amarali, observed in scanning electronic microscopy: A) general aspect, showing the absence of carinae and the presence of ornamentation composed by base-to-apex enamel ridges; B) detail of the tooth surface, showing the ridges. Arq. Mus. Nac., Rio de Janeiro, v.66, n.l, p.63-82, jan./mar.2008 MORPHOLOGY OF THE DENTAL CARINAE IN MARILIASUCHUS AMARALI 71 Fig.5- molariform tooth URC R»75, from Mariliasuchus amaráli, observed in scanning electronic microscopy: A) general aspect from the molariform tooth in lingual view, showing the light ornamentation over the surface and the denticles at the border; B) detail of the denticles from the mesial border, with a very distinctive tuberous profile. Note the anastomosed pattern composed by the enamel ridges present over the labial and lingual faces of the crown. Scale bar = 0.25mm (B). Arq. Mus. Nac., Rio de Janeiro, v.66, n.l, p.63-82, jan./mar.2008 72 M.B.ANDRADE & RJ.BERTINI Each carina is formed by a collection of rhomboidal also preserved in the right hypertrophied caniniform, denticles, undefined in shape (anisomorphic), with but it is more labial than mesial. This feature is not subcircular cross-section. They are tuberous, with exclusive from URC specimens and is figured for an irregular aspect. Furthermore, no additional MZSP-PV-50 (Zaher etal, 2006). In fact, Vasconcellos structures could be observed over the denticles, or & Carvalho (2005) also report wear surfaces in UFRJ between th em (Fig.5), as in Sebecus denticles (Fig. 1). DG-105-R e UFRJ DG-106-R. Furthermore, Zaher et URC R»75 also has an ornamentation pattern quite al (2006) describe extensive wear facets on the lingual surfaces of some second to fourth maxillaiy and sixth evident on its surface, with ridges developing from to eighth mandibulary teeth of MZSP-PV-50 and base to apex, but in an anastomosed pattern. This MZSP-PV-51. Extensive lingual worn surfaces can ornamentation does not extend over the carinae also be seen in three MN 6756-V maxillary denticles, as would be expected for a false-ziphodont. molariforms, and at least in one of MN 6298-V. In These ridges are irregular and anastomosed. MN 6756-V mandible, the sixth pair of molariforms Observation of the dental series of URC R»67 and show apical-labial wear surfaces. URC R*68 shows that this pattern progress from the anterior to the posterior teeth in a particular Another aspect of Mariliasuchus deserving attention way. On the anteriormost teeth these crests or ridges is that molariform teeth can show a certain degree of are bigger and longer, occurring in smaller numbers, paramesial rotation, resulting into a slightly obliqúe while in posterior teeth a greater number of ridges implantation, as observed by several authors is present, and the anastomosis is more evident. (Andrade, 2005; Vasconcellos & Carvalho, 2005, 2006; Zaher et al, 2006). The distai carina is Although URC R*67 and URC R*68 could not be positioned coincident with the sagittal plane of the studied under SEM, observation under common skull. This can be observed both in the maxilla and optical resources can be included, especially mandible. In URC R»68 this is more evident in three regarding the carinae and wear surfaces. In URC R*68 of the most developed right maxillary molariforms, the maxilla and the dentary are not bound together, and also from the sixth to the eighth right and teeth can be examined in several positions, which mandibular molariforms. As previously reported, is particularly important. The dental carinae are most this particular disposition can also be seen in MZSP likely situated on both mesial and distai surfaces, PV-50 (Zaher et al, 2006), on two maxillaiy pairs, for most molariforms, but are present in all and MN 6298-V and MZSP PV-51 (Zaher etal, 2006), molariforms, without exception. Nevertheless, part for three maxillary pairs. At least in the mandible of the dental series of URC R»68 had wear surfaces from MZSP PV-50 (Zaher et al, 2006), MZSP PV-51 where the serrations should have developed, and it (Zaher et al, 2006) and MN 6298-V, there is a slight was impossible to positively identify the presence of degree of rotation in the fifth to the eighth teeth. The denticles. Abrasion surfaces are plane, pattern is more evident in URC R»68, and also in a anteroposteriorly elongated and positioned over either variable degree and not in all the same mandibulary the mesial or the distai border of the molariform teeth, teeth for the other specimens, but it is present. but not on both surfaces of the same tooth. These planes can be especially seen on the sixth and seventh mandibular molariforms, and the opposing maxillary Carinae and Teeth from Mariliasuchus amarali and the teeth. In mandibulary molariforms, the worn planes CONCEPT OF ZlPHOMORPH DeNTITION are present only on the mesial surface, inclined The morphology observed in these isolated teeth of anteriorly and labially. In the opposing maxillaiy Mariliasuchus amarali shows clearly the presence of teeth, these surfaces are present on the distai surface, true denticles constituting a serrated border, on the facing posteriorly and lingually (Fig. 6). The upper and molariform tooth observed. These structures are lower wear surfaces match each other, and the coherent with the description provided by Zaher et complete set (maxilla, premaxilla and mandible) were al (2006) for teeth of other specimens, although in found in occlusion, in close association (Fig.7). their descriptions they preferred to consider these Worn areas have also been found in hypertrophied structures as tubercles. Observations using SEM caniniforms of both URC R»67 and URC R»68. In URC allowed to clearly State that the ornamentation does R»67 there is an eroded plane on the left caniniform not participate in the composition of the carina and mesial crown surface. The worn plane is positioned the denticles are real and individualized structures. on the tip of the crown, developing over the mesial This excludes completely the possibility of these surfaces of the teeth. In URC R»68 this worn plane is teeth as to be characterized as false-ziphodont teeth. Arq. Mus. Nac., Rio de Janeiro, v.66, n.l, p.63-82, jan./mar.2008