Abstract
- Top of page
- Abstract
- INTRODUCTION
- GEOLOGICAL SETTING
- AGE OF THE GUICHÓN FORMATION
- SYSTEMATIC PALAEONTOLOGY
- SKULL
- LOWER JAW
- DENTITION
- CERVICAL VERTEBRAE
- PHYLOGENETIC RELATIONSHIPS
- STATUS OF URUGUAYSUCHUS TERRAI
- CONCLUSIONS
- ACKNOWLEDGEMENTS
- REFERENCES
- Appendix
- Supporting Information
Uruguaysuchus is a small mesoeucrocodylian known from several partial skeletons and skulls from the Guichón Formation (middle Cretaceous, Uruguay). Several authors have pointed out derived similarities of this taxon with different basal notosuchian genera, highlighting its importance for mesoeucrocodylian phylogeny and biogeography. However, the holotype is only partially prepared and has not been available for study for many years. Thus, phylogenetic studies have included this form based on the original description, thereby resulting in a large amount of missing data in the character scorings of this taxon. Here, we describe a new specimen from the type locality consisting of a partial skull, lower jaw and cervical vertebrae which can be referred to U. aznarezi. The new specimen allows for the recognition and scoring of several characters previously unknown for this taxon, thus providing a more extensive diagnosis, as well as new information for understanding its phylogenetic relationships. These characters are congruent with the morphology present in basal notosuchians. The relationships of Uruguaysuchus are tested through a cladistic analysis using a recently published data set including the new information. The phylogenetic results differ from previous analyses, recovering this taxon as the sister group of the Araripesuchus clade. U. terrai is considered a juvenile individual of U. aznarezi.
© 2011 The Linnean Society of London, Zoological Journal of the Linnean Society, 2011, 163, S173–S198.
INTRODUCTION
- Top of page
- Abstract
- INTRODUCTION
- GEOLOGICAL SETTING
- AGE OF THE GUICHÓN FORMATION
- SYSTEMATIC PALAEONTOLOGY
- SKULL
- LOWER JAW
- DENTITION
- CERVICAL VERTEBRAE
- PHYLOGENETIC RELATIONSHIPS
- STATUS OF URUGUAYSUCHUS TERRAI
- CONCLUSIONS
- ACKNOWLEDGEMENTS
- REFERENCES
- Appendix
- Supporting Information
In 1932, several partial skulls and skeletons belonging to at least seven individuals of small, terrestrial crocodyliforms were discovered during the drilling of a water well in the town of Guichón, Paysandú province, Uruguay (see Fig. 1). The agronomist Jorge Aznárez sent the abundant remains to the Argentinean palaeontologist Lucas Kraglievich (1886–1932), who unfortunately died before he was able to study them. Kraglievich's wife sent them to another Argentinean palaeontologist, Carlos Rusconi (1898–1969), who published a detailed description (Rusconi, 1933), and recognized a new genus, Uruguaysuchus, and two new species, U. aznarezi and U. terrai. Since then, Uruguaysuchus has played a key role in the recognition of the infraorder Notosuchia (Gasparini, 1971), which is the most diverse group of Crocodyliformes from the Cretaceous of Gondwana and is characterized by the presence of heterodont dentition with a remarkable disparity in the tooth anatomy among its members. Moreover, Uruguaysuchus is still the best-represented Mesozoic vertebrate from Uruguay.
Unfortunately, the holotype of Uruguaysuchus aznarezi, a c. 1.2-m-long articulated adult individual (including a complete skull, mandible, right forelimb, 18 vertebrae, both ilia and partial right hindlimb, as well as a few associated osteoderms), is only partially prepared. More importantly, the type material is housed at a private collection (Colección Aznárez, Bella Vista, Maldonado province, Uruguay) and has not been available for study for many years. Unfortunately, all five paratypes are currently lost. The holotype of the second species described by Rusconi, U. terrai, is also housed in the Colección Aznárez. Although it has been heavily damaged during its transportation (as already noted by Rusconi, 1933), a photograph of the original specimen (a complete mandible and skull in articulation) does exist.
The first detailed study on the affinities of Uruguaysuchus was conducted by Gasparini (1971), after performing a first-hand study of the original material. Gasparini (1971) noted similarities between this taxon and Araripesuchus gomesii (Price, 1959) and clustered these two taxa in the family Uruguaysuchidae. Recent cladistic studies have included this form based on the published information (Rusconi, 1933), albeit with an abundance of missing character data (e.g. Buckley et al., 2000; Ortega et al., 2000). These studies have retrieved Uruguaysuchus in alternative positions, closely allied with other basal notosuchians such as Simosuchus (Buckley et al., 2000; Turner, 2006), Candidodon (Andrade & Bertini, 2008; Fiorelli & Calvo, 2008), Notosuchus (Jouve et al., 2006), or as the most basal member of Ziphosuchia (i.e. the clade including all notosuchians except Araripesuchus; Pol, 2003; Gasparini, Pol & Spalletti, 2006; Turner & Buckley, 2008). As noted by Gasparini et al. (1998), Uruguaysuchus can be considered as an ‘enigma in the evolutionary history of crocodiles’.
The new specimen of U. aznarezi described herein was found in the fossil vertebrate collection of the Facultad de Ciencias in Montevideo (Uruguay), and is probably the same specimen briefly mentioned but not described nor figured by Lambert (1940). It comes from the same locality that yielded the previously known Uruguaysuchus specimens. These remains include a partial skull, a lower jaw and three cervical vertebrae and allow for the recognition of several characters previously unknown for this taxon that shed light on the phylogenetic relationships of Uruguaysuchus.
Institutional abbreviations: CA, Colección Aznárez (Bella Vista, Uruguay); FC-DPV, Colección de Vertebrados Fósiles, Facultad de Ciencias (Montevideo, Uruguay); MPCA, Museo Paleontológico Carlos Ameghino (Cipolletti, Argentina); MPMAB, Museo Paleontológico Municipal Alejandro Berro (Mercedes, Uruguay); MUNHINA, Museo Nacional de Historia Natural y Antropología (Montevideo, Uruguay).
GEOLOGICAL SETTING
- Top of page
- Abstract
- INTRODUCTION
- GEOLOGICAL SETTING
- AGE OF THE GUICHÓN FORMATION
- SYSTEMATIC PALAEONTOLOGY
- SKULL
- LOWER JAW
- DENTITION
- CERVICAL VERTEBRAE
- PHYLOGENETIC RELATIONSHIPS
- STATUS OF URUGUAYSUCHUS TERRAI
- CONCLUSIONS
- ACKNOWLEDGEMENTS
- REFERENCES
- Appendix
- Supporting Information
The Guichón Formation (Bossi, 1966) crops out in Paysandú, Río Negro and Salto provinces, western Uruguay (Fig. 1). According to Goso & Perea (2004), it comprises mainly pink-greyish to reddish sandstones, which contain moderate to well-sorted, subrounded, fine to medium-sized grains in a pelitic matrix. These sandstones (which compositionally are feldspathic wackes) are either massive or may instead exhibit parallel lamination, cross-lamination and graded bedding. These lithologies were deposited in south-west-trending alluvial–fluvial systems comprising low-sinuosity channels traversing through sandy plains. Subordinate to the already mentioned sandstones are conglomeratic and pelitic lithologies, interpreted as channel-fill and overbank deposits, respectively. It is inferred that the Guichón Formation was deposited in warm, semi-arid climatic conditions (Goso, 1999; Goso & Perea, 2004).
The Guichón Formation unconformably overlies flood basalts of the Arapey Formation (radiometrically dated to 132 Ma – Hauterivian; Féraud et al., 1999), and is in turn unconformably overlain by sandstones, conglomerates and pelites of the Mercedes Formation (Campanian–Maastrichtian; Daners & Guerstein, 2004; Goso & Perea, 2004). Both the Arapey and the Mercedes Formations were formally defined by Bossi (1966). Although Bossi (1966) also defined the Asencio Formation, it must be noted that most authors have partly or totally included the Asencio Formation within the Mercedes Formation (Ford & Gancio, 1988; Goso, 1999; Bossi & Ferrando, 2001; Goso & Perea, 2004; see discussion in Martínez & Veroslavsky, 2004). The Guichón and Mercedes Formations are included in the Paysandú Group (Bossi & Navarro, 1991).
Deposition of the Paysandú Group began under endorrheic conditions after extrusion of the Neocomian flood basalts (an event regionally known as the Serra Geral magmatism), which in Uruguay extends over 25 000 km2, with a maximum thickness of 220 m. This sedimentary scenario continued in the Argentinean Mesopotamia, where the unit depocentre is located. In particular, the Guichón Formation can be partly correlated with the Argentinean Puerto Yeruá Formation (Goso, 1999; Goso & Perea, 2004). Traditionally, the Paysandú Group has been considered to represent the youngest deposits of the Norte Basin, although other authors consider that given its peculiarities the recognition of a separate basin is justified: the Litoral del Río Uruguay Basin, or Litoral Basin (Goso, 1999; Goso & Perea, 2004).
AGE OF THE GUICHÓN FORMATION
- Top of page
- Abstract
- INTRODUCTION
- GEOLOGICAL SETTING
- AGE OF THE GUICHÓN FORMATION
- SYSTEMATIC PALAEONTOLOGY
- SKULL
- LOWER JAW
- DENTITION
- CERVICAL VERTEBRAE
- PHYLOGENETIC RELATIONSHIPS
- STATUS OF URUGUAYSUCHUS TERRAI
- CONCLUSIONS
- ACKNOWLEDGEMENTS
- REFERENCES
- Appendix
- Supporting Information
In addition to the Uruguaysuchus material (see Rusconi, 1933), the Guichón Formation has yielded a few dinosaurian teeth described by Huene (1934) that are currently lost, including a single theropod tooth (perhaps the same that was tentatively identified as coming from a sebecosuchian by Mones, 1997) and two teeth with ornithischian affinities.
The age of this unit is not well established. It is only certain that it is younger than Hauterivian and older than Campanian. Palaeontological evidence was previously insufficient or incorrectly interpreted. According to Huene (1934) the theropod tooth was very similar to one described by Lambe (1902) as occurring closely associated to ornithomimid bones in Late Cretaceous deposits from Canada. Given that the edentulous nature of the ornithomimids was then unknown, Huene (1934) mistakenly referred the tooth from Guichón to the family Ornithomimidae. This was the main argument upon which the Guichón Formation was considered of Senonian age in the following decades. Bonaparte (1978) considered that the two ornithischian teeth belonged to the Iguanodontidae, which in his interpretation supported the same age. Soto & Cambiaso (2006) reviewed both Huene's and Bonaparte's early determinations, and concluded that the purported ornithomimid and iguanodontid material should be better regarded as belonging to indeterminate theropods and basal iguanodontians, respectively. This reinterpretation eliminated the evidence supporting a Late Cretaceous age for the Guichón Formation. Although the phylogenetic position of this taxon is not well established, most authors have retrieved it in a basal position within Notosuchia. In South America, these forms (e.g. Candidodon, Araripesuchus gomesii, A. patagonicus, A. buitreraensis) are usually recorded from the middle Cretaceous (Aptian–Cenomanian). In more recent Cretaceous deposits (e.g. Campanian–Maastrichtian) these forms are absent and the notosuchian fauna comprises derived notosuchians (e.g. Notosuchus, Sphagesaurus, Comahuesuchus, Mariliasuchus and baurusuchids). In other regions of Gondwana, however, basal notosuchians seem to have survived toward the end of the Cretaceous (e.g. Simosuchus and A. tsangatsangana from Madagascar).
On the other hand, the geological evidence, such as the discordance with the overlying Mercedes Formation (Campanian–Maastrichtian; Daners & Guerstein, 2004; Goso & Perea, 2004) and the facies similarity with the Migues Formation (Albian; Campos et al., 1997) from the southern Santa Lucía Basin, has prompted some authors (Goso, Perea & Perinotto, 1999; Goso & Perea, 2004) to suggest an Early Cretaceous age for this unit. It should also be noted that the putatively correlated Puerto Yeruá Formation (Argentina) is also difficult to constrain beyond a Cretaceous age (De Valais, Apesteguía & Udrizar Sauthier, 2003).
DENTITION
- Top of page
- Abstract
- INTRODUCTION
- GEOLOGICAL SETTING
- AGE OF THE GUICHÓN FORMATION
- SYSTEMATIC PALAEONTOLOGY
- SKULL
- LOWER JAW
- DENTITION
- CERVICAL VERTEBRAE
- PHYLOGENETIC RELATIONSHIPS
- STATUS OF URUGUAYSUCHUS TERRAI
- CONCLUSIONS
- ACKNOWLEDGEMENTS
- REFERENCES
- Appendix
- Supporting Information
The specimen of Uruguaysuchus described here has several important features preserved in its dentition, although none of the toothrows is completely preserved. FC-DPV 2320 has preserved three premaxillary alveoli (Fig. 1B), although the presence of an additional tooth socket is highly probable given that the anterior tip of the premaxilla has not been preserved. This is congruent with the four premaxillary alveoli described by Rusconi (1933) for this taxon and the generalized condition of four premaxillary teeth found in basal mesoeucrocodylians (including most notosuchians except for Sphagesaurus, Chimaerasuchus, and baurusuchids). All the premaxillary teeth (and premaxillary alveoli) are relatively small (Figs 2B, 3) and therefore Uruguaysuchus lacks an enlarged premaxillary caniniform, a feature present in all notosuchians (e.g. Libycosuchus, Candidodon, Mariliasuchus, Notosuchus; Andrade & Bertini, 2008) except for Araripesuchus and Simosuchus (Buckley et al., 2000).
The right maxilla bears 12 alveoli but due to incomplete preservation it is not clear if the left maxilla had 12 or 13 teeth (Fig. 1B). Thus, the entire upper toothrow would be composed by 16–17 teeth, a number that matches the toothcount of the lower dentition (17) (Fig. 8A). It must be noted that Rusconi (1933), based on the holotype of U. aznarezi, reported 13 teeth in the upper tooth row and 12 in the lower one; the lack of preparation of this specimen allows us to suppose that more teeth (or at least the alveoli) would have been present in the holotype. The maxillary tooth count of Uruguaysuchus is relatively high in comparison with most notosuchians that have seven or fewer maxillary teeth (e.g. Candidodon, Malawisuchus, Notosuchus, Mariliasuchus, Comahuesuchus, Sphagesaurus, baurusuchids). However, basal notosuchians such as Araripesuchus, Libycosuchus, and Simosuchus (Buckley et al., 2000; Pol & Apesteguía, 2005) have higher tooth counts as in Uruguaysuchus, resembling the generalized condition of basal mesoeucrocodylians.
The tooth morphology of Uruguaysuchus bears several autapomorphic characters that distinguish this taxon from all other known crocodyliforms. The dental series has a high degree of heterodonty, showing three main morphological types: incisiviform, caniniform, and post-caniniform teeth.
Incisiviforms are present in the anterior region of both the upper and the lower toothrows. In the upper toothrow this tooth type is present through the second maxillary position (Fig. 2). In the lower dentition, this morphology seems to be present in the first six or seven positions (Fig. 10A). Incisiviforms are small and conical teeth that are slightly recurved lingually (Figs 3, 10A). An isolated incisiviform shows that the crown is slightly inflated at its base but the crown-root is only slightly constricted (Fig. 12C). The enamel surface of this tooth is mostly smooth although a faintly developed wrinkling pattern can be observed at high magnification by scanning electron microscopy (Fig. 12D). The distal margin of this tooth bears a well-developed keel that extends from the tooth apex down to the bulbous region of the base of the crown (failing to reach the crown-root limit). This keel is slightly sinuous but lacks true denticles (non ziphodont sensuPrasad & de Lapparent de Broin, 2002; Fig. 12D).
The caniniform is a conical tooth that is approximately twice as large as the other teeth (Fig. 3) and is exclusively present in the upper toothrow, occupying in FC-DPV 2320 (as in U. terrai and a juvenile specimen of U. aznarezi, but differing from the holotype of U. aznarezi) the third maxillary alveolous (Fig. 12B). An enlarged maxillary caniniform located in the second or third maxillary alveolous is also present in several basal notosuchians (e.g. Araripesuchus, Anatosuchus, Malawisuchus; Gomani, 1997; Pol & Apesteguía, 2005) but is absent in more derived notosuchians (e.g. Notosuchus, Mariliasuchus, Sphagesaurus; Pol, 2003; Andrade & Bertini, 2008). The caniniform of Uruguaysuchus, however, differs from that of other notosuchians in the presence of four thick apicobasal carinae. The base of the crown of the caniniform tooth is bulbous and a more developed constriction is present between the crown and the root in comparison with that of the incisiviform teeth.
The post-caniniform teeth are located in the upper and lower toothrows. In the maxilla these elements are present from the fourth to the 12th (or 13th) position, and in the lower toothrow these elements are present posteriorly to the seventh alveoli. The post-caniniform teeth of Uruguaysuchus have several autapomorphic features. The crown of these teeth is markedly flattened buccolingually and has a circular outline when viewed in lateral view (Fig. 12A). The rounded profile of the tooth crown is basally limited by an extremely well-developed constriction between the crown and the root (Fig. 12), resembling the condition of Simosuchus. The crown of these teeth is characterized by the presence of a central apical cusp that has a variable degree of development along the toothrow (Fig. 12E, G). Two slightly developed grooves extend apicobasally on the lingual and buccal surfaces of the crown, diverging basally from the mesial and distal limits of the central cusp. The central cusp and its associated grooves are more developed in some of the post-caniniform teeth (Fig. 12E) than in others (Fig. 12G, H). In some teeth, the buccal or lingual surface of the crown bears additional apicobasal grooves located between the two major grooves mentioned above (Fig. 12G). Mesial and distal from this central cusp, the margins of the crown bear denticles that vary in their number, development, and extension along the tooth margins in different post-caniniform teeth.
Some teeth have either seven or eight denticles located on the distal and mesial margins of the crown that are well separated from the central cusp and that extend basally down to the midpoint of the crown, approximately at the level of the maximum mesiodistal expansion of the crown (Fig. 12E). However, in those post-caniniform teeth in which the central cusp is not as developed, there are only four mesial and distal denticles that are smaller, closely spaced between each other, and located closer to the central cusp (Fig. 12G, H). Furthermore, in these teeth, the mesial and distal denticles are restricted to the apical region of the crown, being well separated from the maximum point of mesiodistal extension of the crown (Fig. 12G). Therefore, in these teeth the central cusp and the denticles form a subhorizontally oriented occlusal margin (Fig. 12G, H).
The shape of the mesial and distal denticles is also interesting in terms of their similarity with those of other crocodyliforms. These denticles are well separated from each other by interdenticular slits created by a constriction of the enamel and dentine and therefore cannot be described as pseudoziphodont in which the ‘denticles’ are exclusively formed by enamel ridges and foldings (sensuPrasad & de Lapparent de Broin, 2002). Each denticle is buccolingually broad and tuberous in overall shape. This denticular morphology resembles that of Mariliasuchus (Andrade & Bertini, 2008) and Notosuchus (Lecuona & Pol, 2008) but contrasts with the buccolingually flattened denticles present in other ziphodont crocodyliforms (e.g. Sebecus, Dakosaurus; Legasa, Buscalioni & Gasparini, 1994; Pol & Gasparini, 2009). However, the denticles of Uruguaysuchus bear also a sharp cutting edge that extends along the mesial or distal edge of the crown (Fig. 12F). Interestingly, this sharp ridge extends over the interdenticular slits in addition to being present over each of the denticles (Fig. 12F). The presence of this sharp carina distinguishes the teeth of Uruguaysuchus from those of the derived notosuchians (e.g. Notosuchus, Mariliasuchus), in which the denticles are completely rounded and lack a sharp cutting edge. A similar sharp edge is, in turn, present in most crocodyliforms with ziphodont dentition (e.g. Sebecus, Dakosaurus). Thus, the denticles of Uruguaysuchus bear a unique combination of characters that can be interpreted as intermediate between the generalized ziphodont dentition and the condition of derived notosuchians (termed ziphomorph for Mariliasuchus by Andrade & Bertini, 2008).
The presence of denticulated tooth margins has also been reported for Araripesuchus wegeneri and is also present in the posterior lower teeth of A. gomesii (AMNH 24450). Its presence, however, cannot be confidently determined (or rejected) in A. patagonicus or A. buitreraensis. The unusual crown morphology of Simosuchus has also been compared with that of Uruguaysuchus (Buckley et al., 2000). The superficial similarities of post-caniniform teeth of Uruguaysuchus with distal teeth of Simosuchus led Soto (2005) to propose that the diet of the former genus included plant material (besides other animals), as postulated for Simosuchus by Buckley et al. (2000). However, the dentition of the Malagasy taxon is highly modified, as it has multicusped teeth in the entire toothrow and most crowns (except those of posteriormost teeth) have three triads of cusps separated by extremely deep notches rather than the central cusp and denticulated mesial and distal margins that characterize post-caniniform teeth of Uruguaysuchus.
CERVICAL VERTEBRAE
- Top of page
- Abstract
- INTRODUCTION
- GEOLOGICAL SETTING
- AGE OF THE GUICHÓN FORMATION
- SYSTEMATIC PALAEONTOLOGY
- SKULL
- LOWER JAW
- DENTITION
- CERVICAL VERTEBRAE
- PHYLOGENETIC RELATIONSHIPS
- STATUS OF URUGUAYSUCHUS TERRAI
- CONCLUSIONS
- ACKNOWLEDGEMENTS
- REFERENCES
- Appendix
- Supporting Information
The atlantal intercentrum, two articulated anterior cervical vertebrae, and several cervical ribs are the only postcranial remains associated with this specimen. The intercentrum is broader than long, with a large anterior articular facet for the occipital condyle and two posterior articular facets for the atlantal ribs (Fig. 13). The posterior rib facets project posterolaterally forming an angle of 45° with each other (Fig. 13). The central region between the articular processes of the atlas intercentrum is extremely short and broad, and has a broad concavity on both its dorsal and its ventral surfaces.
The anterior cervical vertebrae are probably the third and fourth elements of the cervical series (Fig. 14). The neural spines of these vertebrae are lost, although the preserved dorsal surface of the neural arches suggests their spine was anteroposteriorly short and located mostly over the posterior half of the dorsal surface of the neural arches. The neural arches are anteroposteriorly short and dorsoventrally high in comparison with those of neosuchian crocodyliforms but resemble the condition of Notosuchus (Pol, 2005). The prezygapophyses are short and dorsally recurved, so that their distal end is projected dorsally rather than anterodorsally (Fig. 14A), also resembling the condition of Notosuchus (Pol, 2005). The postzygapophyses are incompletely preserved and it cannot be determined if they had a suprapostzygapophyseal lamina connecting them to the neural spine. Anterior to the articular facet of the postzygapophyses, an anteroventrally oriented ridge extends on the dorsal region of the lateral surface of the neural arch, being more developed in the third cervical vertebra. The articular surfaces of the zygapophyses form an angle of approximately 45° with the sagittal plane. The diapophyses are robust and directed ventrolaterally (Fig. 14C). Their articular facets are ovoid with the major axis oriented anteroposteriorly. A deep depression is present below the diapophyses and above the parapophyses, extending over both the neural arch and the vertebral centrum. The cervical centra are low and long, being approximately twice as long as high. The parapophyses are located at the anterior margin of the centrum and are laterally projected (Fig. 14A, C). A thin longitudinal ridge extends posteriorly to the parapophyses along the vertebral centrum, delimiting ventrally the concave surface located between the parapophysis and the diapophysis. Below this ridge the centrum is also markedly concave and its ventral surface bears a longitudinal ridge that delimits the ventral extension of the concavity (Fig. 14B). The anterior end of this ridge is ventrally projected forming a moderately developed hypapophysis (Fig. 14A, B). The articular surfaces of the centrum are subcircular in outline and concave, and the neural canal above them is mediolaterally wider than dorsoventrally high (Fig. 14C).
The preserved cervical ribs (Fig. 14D) have a short and dorsoventrally high anterior process that laterally overlaps the posterior process of the preceding vertebra. Posteriorly, the rib has an elongated and low posterior process. The lateral surface of the posterior process bears a longitudinal keel and lacks the posterodorsally directed accessory process present in A. tsangatsangana and Mahajangasuchus (Turner, 2006).
PHYLOGENETIC RELATIONSHIPS
- Top of page
- Abstract
- INTRODUCTION
- GEOLOGICAL SETTING
- AGE OF THE GUICHÓN FORMATION
- SYSTEMATIC PALAEONTOLOGY
- SKULL
- LOWER JAW
- DENTITION
- CERVICAL VERTEBRAE
- PHYLOGENETIC RELATIONSHIPS
- STATUS OF URUGUAYSUCHUS TERRAI
- CONCLUSIONS
- ACKNOWLEDGEMENTS
- REFERENCES
- Appendix
- Supporting Information
Rusconi (1933) originally placed the genus Uruguaysuchus in the family Notosuchidae, but later Gasparini (1971) defined Uruguaysuchidae to cluster Uruguaysuchus with Araripesuchus, as a family of the infraorder Notosuchia. As noted above, after these studies several authors noted derived similarities of Uruguaysuchus either with notosuchians (Pol, 2003) or with Araripesuchus and neosuchians (Ortega et al., 2000). Recent cladistic studies retrieved this taxon as alternatively allied with several notosuchians, such as Notosuchus, Candidodon, and Simosuchus (e.g. Buckley et al., 2000; Pol & Apesteguía, 2005; Jouve et al., 2006; Andrade & Bertini, 2008). However, as noted above, these authors based their inferences on the information available from the literature given that the specimens were not available for study, which resulted in a large number of missing entries (or even erroneous scorings) for Uruguaysuchus in phylogenetic datasets.
The specimen described herein provides new information for more rigorously testing the phylogenetic affinities of Uruguaysuchus. This new information was used to complete the character scorings of Uruguaysuchus in a recently published dataset (Pol & Gasparini, 2009) comprising 257 characters scored across 59 taxa. In this previous study Uruguaysuchus had 57% of the cells scored with missing entries. The new specimen provided information to score an additional 59 characters that were previously unknown or uncertain for this taxon, reducing the amount of missing entries to 33% (see Supplementary Information, Appendices S1–4, for the full character list, data matrix, and complete list of unambiguous synapomorphies).
The updated data matrix was analysed in TNT (Goloboff, Farris & Nixon, 2008) using equally weighted parsimony analysis. A heuristic tree search was performed conducting 1000 replicates of Wagner trees (using random addition sequences) followed by TBR branch swapping (holding ten trees per replicate). The analysis yielded two most-parsimonious trees (MPTs) of 859 steps (CI = 0.365, RI = 0.710) that were found in 74% of the replicates. The strict consensus of the MPTs (Fig. 15) places Uruguaysuchus within Notosuchia as the sister taxon to the clade Araripesuchus. This result contrasts with more recent cladistic analyses (including that of Pol & Gasparini, 2009) that depicted Uruguaysuchus closer to more derived notosuchians. The clade of Uruguaysuchus and Araripesuchus forms a monophyletic Uruguaysuchidae, as originally conceived by Gasparini (1971). In our analysis Uruguaysuchidae is diagnosed by two unambiguous synapomorphies: surangular forming approximately one-third of the glenoid fossa (character 156.1) and dorsal surface of mandibular symphysis strongly concave and narrow, trough shaped (character 189.1). The latter character is so far only recorded in Uruguaysuchus (see Fig. 10B), A. gomesii (DGM 423-R), and in one specimen referred to Araripesuchus from the Cenomanian of Patagonia (MPCA-PV 236; see Pol & Apesteguía, 2005). All other basal mesoeucrocodylians lack this derived feature. The former synapomorphic feature is known in A. gomesii (AMNH 24450) and Uruguaysuchus (Fig. 11B) but the condition of other species of Araripesuchus is not known. A surangular participation in the mandibular glenoid facet, however, has also been reported to be present in some basal mesoeucrocodylians not included in this analysis (e.g. Hamadasuchus, Montealtosuchus, Mahajangasuchus, Sebecus) as well as in other groups of crocodyliforms (e.g. Protosuchus, dyrosaurids). This feature has evolved several times in Crocodyliformes and may have a broader distribution among basal mesoeucrocodylians than in the optimization obtained in our analysis. Further studies expanding the taxonomic sampling of basal mesoeucrocodylians are needed to assess a more thorough understanding on the evolution of this character. Finally, although this position is the most parsimonious explanation of the available data, the sister-group relationship of Araripesuchus with Uruguaysuchus is poorly supported; trees with one extra step depict the latter taxon as more closely related to ziphosuchians than to Araripesuchus (as in some of the previous phylogenetic studies; Pol, 2003; Gasparini et al., 2006; Turner & Buckley, 2008; Pol & Gasparini, 2009). Further data on Uruguaysuchus, such as a re-study of the more complete type specimen, will undoubtedly help to resolve this issue.
Despite these uncertainties, the notosuchian affinities of Uruguaysuchus are strongly supported by the available data (constrained analysis forcing this taxon to be placed outside Notosuchia requires at least 11 extra steps). Even when Uruguaysuchus and Araripesuchus are both forced to be placed outside Notosuchia (i.e. more closely related to peirosaurids and neosuchians), the most parsimonious topologies are six steps longer than in the unconstrained analysis.
Furthermore, the available information on Uruguaysuchus strongly supports a basal position for this taxon within Notosuchia. Constrained analyses forcing Uruguaysuchus to be more derived than Simosuchus (or even closely related to this taxon as suggested by Buckley et al., 2000) requires between six and ten extra steps, implying more extra steps as more derived positions are forced in the constrained searches. In summary, based on the current information Uruguaysuchus can be safely depicted as one of the most basal notosuchians, either as the sister group of Araripesuchus (as in the MPTs obtained here) or possibly as a basal taxon of the clade formed by Lybicosuchus and more derived forms (Ziphosuchia sensuOrtega et al., 2000; see Fig. 15).
STATUS OF URUGUAYSUCHUS TERRAI
- Top of page
- Abstract
- INTRODUCTION
- GEOLOGICAL SETTING
- AGE OF THE GUICHÓN FORMATION
- SYSTEMATIC PALAEONTOLOGY
- SKULL
- LOWER JAW
- DENTITION
- CERVICAL VERTEBRAE
- PHYLOGENETIC RELATIONSHIPS
- STATUS OF URUGUAYSUCHUS TERRAI
- CONCLUSIONS
- ACKNOWLEDGEMENTS
- REFERENCES
- Appendix
- Supporting Information
As already stated, Rusconi (1933) recognized a second species of Uruguaysuchus, U. terrai, based mainly on differences in the tooth count in the upper tooth row. Both Soto (2005) and Andrade & Bertini (2005) independently questioned the taxonomic validity of maintaining the distinction of U. terrai from U. aznarezi. U. terrai apparently possesses four maxillary teeth more than U. aznarezi: two incisiviforms, one caniniform, and ten post-caniniforms versus one incisiviform, one caniniform and seven post-caniniforms, respectively (Fig. 16). However, as already stated, it must be noted that the lack of preparation of the holotype of U. aznarezi does not allow an assessment of the total number of post-caniniform teeth (i.e. whether there are teeth beyond the ninth maxillary position).
Moreover, the maxilla of Rusconi's specimen no. 4 (Fig. 16), a juvenile U. aznarezi according to this author, exhibits two incisiforms and one caniniform (being the third maxillary tooth), the same condition as in U. terrai. Curiously, Rusconi (1933) did not explicitly mention this relevant point. Interestingly, his fig. 20 illustrates seven post-caniniform alveoli although only six are implied in the text.
Furthermore, several measurements of U. terrai are consistently smaller than those of the only adult individual of U. aznarezi (holotype), approaching those of juveniles of U. aznarezi.
As proposed by Soto (2005), minor differences in the dental formulae can be explained by intraspecific variation in tooth count, which is rather common among crocodyliforms, either fossil or extant (C. Brochu, com. pers., 2008). In particular, if it is true that U. terrai, U. aznarezi no. 4 and FC-DPV 2320 represent juvenile individuals, it may be necessary to invoke ontogenetic loss of tooth positions to explain the fact that the adult individual of U. aznarezi (holotype) has only one incisiviform maxillary tooth instead of two. Such a phenomenon has already been recorded in several crocodylian species (e.g. Crocodylus cataphractus, C. porosus, C. siamensis, and Tomistoma schlegelii; Mook, 1921; Wermuth, 1953; Iordansky, 1973) as well as in the tyrannosaurid theropods Tyrannosaurus rex and Albertosaurus libratus (Carr, 1999).
On the other hand, the hypothesis that U. aznarezi could bear more post-caniniform teeth than recognized by Rusconi (1933) received support when FC-DPV 2320 was prepared. Indeed, the specimen described herein (Fig. 16) showed the presence of at least 12 maxillary teeth (two incisiviforms, one caniniform, and at least nine post-caniniforms), reducing the apparent gap between the tooth count of U. aznarezi and U. terrai.
In conclusion, given that no real differences in the maxillary dentition exist, we regard U. terrai as a juvenile individual of U. aznarezi. Thus, as proposed by Soto (2005) and Andrade & Bertini (2005), U. terrai must be considered a junior synonym of the latter taxon.
CONCLUSIONS
- Top of page
- Abstract
- INTRODUCTION
- GEOLOGICAL SETTING
- AGE OF THE GUICHÓN FORMATION
- SYSTEMATIC PALAEONTOLOGY
- SKULL
- LOWER JAW
- DENTITION
- CERVICAL VERTEBRAE
- PHYLOGENETIC RELATIONSHIPS
- STATUS OF URUGUAYSUCHUS TERRAI
- CONCLUSIONS
- ACKNOWLEDGEMENTS
- REFERENCES
- Appendix
- Supporting Information
Uruguaysuchus from the middle Cretaceous of Uruguay has been shown to be a relevant taxon for mesoeucrocodylian phylogeny and biogeography (e.g. Buckley et al., 2000).
The new specimen described herein allows the recognition of several characters previously unknown for this taxon, thereby providing new insight for understanding its phylogenetic relationships.
According to the phylogenetic analysis presented herein, Uruguaysuchus is a basal notosuchian. Features such as a completely septated choanal opening, a well-developed anterior and posterior process on the ectopterygoid lateral end, a dentary failing to extend beneath external mandibular fenestra, the presence of a maxillary dental groove (absence of interalveolar septa), a slightly elongated glenoid facet that lacks a posterior ridge, and participation of the surangular in the glenoid facet conform a unique combination of characters present in basal notosuchians (e.g. Araripesuchus, Simosuchus), being part of the evidence that led us to postulate a similar position for Uruguaysuchus.
Moreover, Uruguaysuchus is closely related to Araripesuchus, both genera being recovered as sister taxa in the phylogenetic analysis, although it must be noted that this relationship is poorly supported. A similar result was also obtained by Turner & Buckley (2008) in one of the runs of the phylogenetic analysis they performed.
Overall, taking into account the plesiomorphic characters of Uruguaysuchus and the absence of basal notosuchians after the Cenomanian in different basins of South America (e.g. Neuquén and Bauru Basins), a middle Cretaceous age for the Guichón Formation seems a plausible hypothesis given the available evidence. Recently discovered remains of sauropod dinosaurs could shed more light on this topic (Perea et al., 2006; M. Soto & D. Perea, unpubl. data).
U. terrai probably represents a juvenile individual of U. aznarezi, and must be considered a junior synonym of the latter taxon. Minor differences in the dental formulae can be probably explained invoking intraspecific variation in tooth count.
The specimens from Guichón thus represent a monospecific assemblage. As suggested by Rusconi (1933), given that the individuals were concentrated in a small 2-m2 surface, and that they included articulated material (notably the holotype of U. aznarezi, but also some of the paratypes and FC-DPV 2320), they were probably buried due to the collapse of the walls of a burrow. The shovel-shaped snout would assist in excavating burrows, as suggested for Malawisuchus and Simosuchus (Gomani, 1997; Buckley et al., 2000), although in the absence of characters clearly correlated with burrowing this remains highly speculative.
Notosuchian crocodyliforms included carnivorous (e.g. sebecosuchians), omnivorous (e.g. Mariliasuchus), and presumably herbivorous (e.g. Chimaerasuchus, Simosuchus) forms. Uruguaysuchus probably had an omnivorous diet, given the disparate morphology between incisiviform and post-caniniform teeth.
Uruguaysuchus is a heterodont notosuchian, with incisiviform premaxillary, mesial dentary, and mesial maxillary teeth. The second (in adults) or third (in juveniles) maxillary tooth is an hypertrofied caniniform.
The post-caniniform, spatulate lateral teeth are highly apomorphic: they are strongly buccolingually compressed and subcircular in shape (in buccal or lingual view), with a pointed central cusp and minute denticles in a single row along the mesial and distal margins.
This particular morphology adds to the increasing variety of dentition and feeding styles currently recognized among notosuchian crocodyliforms (e.g. Pol, 2003).
Although less enigmatic than a decade ago, the basal notosuchian Uruguaysuchus is still raising questions: new and more challenging ones. We will be probably close to answer at least some of them provided that the holotype is properly prepared in the near future.