Early South American Squamates
Crown squamates are known from the Jurassic (Evans, 2003; Longrich et al., 2012a) and their stem is thought to extend deep into the Triassic (Gauthier et al., 2012). Currently, however, Jurassic squamate remains are unknown in South America. Early studies (Huene, 1931; Casamiquela, 1962, 1975; Evans, 2003) had referred different fossils from the Jurassic and Cretaceous of Argentina to Squamata; nonetheless, their identity as squamates has been challenged by others (Estes, 1983; Albino, 2007).
The South American Cretaceous record of squamates is diverse and abundant, and includes both marine and continental taxa. The earliest irrefutable lizards recovered in South America are from the late Lower Cretaceous of Brazil (Evans and Yabumoto, 1998; Simões, 2012). This Aptian–Albian lizard fauna from the Araripe Basin is constituted by the very well-preserved Olindalacerta brasilensis (Evans and Yabumoto, 1998) and Tijubina pontei (Simões, 2012). Evans and Yabumoto (1998) conducted a preliminary phylogenetic analysis to allocate Olindalacerta among squamates and, although it forms a sister group to Scleroglossa with Eichstaettisaurus (late Jurassic of Germany), it also shows characters shared with the iguanids (i.e., deep pleurodont implantation and long posterodorsal process of the dentary, Evans and Yabumoto, 1998). Likewise, the phylogenetic position of Tijubina is very unstable. It was originally classified as a Teiidae (Bonfim-Júnior and Marques, 1997), later excluded from Squamata and referred to Rhyncocephalia (Martill and Frey, 1998), and once more included in Squamata, this time close to the root of the clade (Bonfim-Júnior and Avilla, 2002). More recently, Simões (2012) regarded Tijubina as a basal scleroglossan, more derived than Olindalacerta and outside Autarchoglossa. To date no definitive phylogenetic analysis of Squamata has included these Aptian–Albian lizards so it is unclear whether or not they are stem Scleroglossa from Gondwana. A further complication is that these fossils are from skeletally immature specimens (Evans and Yabumoto, 1998; Simões, 2012), an issue that is problematic in character coding while building a character matrix for phylogenetic analyses.
The Upper Cretaceous lizard fossil record of South America is composed of isolated fragments from Argentina and Brazil. The earliest of these records is of a frontal from the Cenomanian–Turonian of Patagonia described by Apesteguía et al. (2005). These authors consider that the fused and very constricted frontal (i.e., hour-glass shape) with pronounced ornamentation is of an iguanian (possibly Iguanidae). However, Daza et al. (2012) believe that the character combination listed by Apesteguía et al. (2005) is not exclusive to any lizard group. Independently, and among extant and extinct lizards of Argentina, we only observed the simultaneous presence of these characters in Iguanidae. Thus, we consider valid the tentative assignation given by Apesteguía et al. (2005). Among Iguanidae, Apesteguía et al. (2005) suggested similarities in the dermal sculpturing with the extant tropidurine* Liolaemus. However, the frontal exhibits well-developed supraorbital flanges (Fig. 2 of Apesteguía et al., 2005), that, according to Smith (2009), are present in Polychrotinae* and Corytophaninae. Also, the dermal sculpturing morphology is similar to that of some Polychrotinae* (personal observations). These notes suggest possible Polychrotinae* affinities of this fossil, rather than to Tropidurinae* (sensu Smith, 2009). The frontal described by Apesteguía et al. (2005) extends the South American presence of Iguania by 30 Ma, to the Cenomanian–Turonian, from the earlier described Pristiguana (Maastrichtian, see below).
Figure 2. Some Mesozoic and Paleogene squamate taxa cited in the text. A–B, dentary of Scincomorpha in labial (A) and lingual (B views); C–E, trunk vertebra of Alamitophis argentinus in posterior (C), lateral (D) and ventral (E) views; F–H, trunk vertebra of Madtsoia bai in posterior (F), lateral (G) and anterior (H) views; I–K, trunk vertebra of Chubutophis grandis in anterior (I), lateral (J) and dorsal (K) views; pz, parazygantral foramen.
Download figure to PowerPoint
Also from the Upper Cretaceous, but from sediments of the Turonian–Santonian of Brazil, Nava and Martinelli (2011) recognized a small isolated maxilla which was used to erect the species Brasiliguana prudentis. This taxon is considered a member of Iguanidae (sensu Gauthier et al., 2012) based on the presence of a weakly inclined anterior margin of the maxillary nasal process and the pleurodont tooth implantation (Nava and Martinelli, 2011). The maxilla of Brasiliguana does not present a strong/large palatine process (Fig. 2 of Nava and Martinelli, 2011), a condition shared with Corytophaninae, Polychrotinae*, Iguaninae, and Hoplocercinae (Smith, 2009). From the same Brazilian levels as Brasiliguana, Candeiro et al. (2009) described an indeterminate non-Serpentes squamate represented by a small and articulated portion of an axial skeleton bearing dorsal vertebrae. Unfortunately no further systematic discussion is possible.
Apart from these Upper Cretaceous records, a poorly preserved dentary from the early Campanian of northern Patagonia (Fig. 2) has been tentatively assigned to the Scincomorpha (sensu Gauthier et al., 2012) by Brizuela and Albino (2011). Not considering the lizards from the Araripe basin of Brazil mentioned above, whose relationships are unclear, this sole dentary is the only evidence of a member of the Scleroglossa in the South American Cretaceous. Within this large clade the Campanian dentary shows affinities with the Scincoidea (sensu Gauthier et al., 2012), and, along with the probable Cordylidae Konkasaurus from the Maastrichtian of Madagascar (Krause et al., 2003), suggests that scincoidea-like lizards would have had Gondwanan distribution (Brizuela and Albino, 2011).
The latest Cretaceous lizard comes from the Maastrichtian of Brazil. Estes and Price (1973) erected Pristiguana brasiliensis from cranial and appendicular skeletal remains, that they considered of an extinct crown Iguanidae. Though the relationships of Pristiguana to other iguanians are unclear, Estes and Price (1973) noticed some similarities with Enyalioides, Hoplocercus and Morunasaurus (e.g., large splenial). They also stated that some characters of the fossil (i.e., tooth morphology, open Meckelian canal, and frontal proportions) are similar to those of Teiidae. Borsuk-Białynicka and Moody (1984) considered that the arguments for the Iguanidae assignation are not convincing and that Pristiguana could be assigned to the Teiidae with equally convincing counter arguments. More recently, Daza et al. (2012) conducted a phylogenetic analysis of Squamata, focused on the Iguania, where they incorporated Pristiguana. In this analysis Pristiguana forms a clade with Huehuecuetzpalli (early Cretaceous of Mexico) sister to Iguania. Therefore, according to Daza et al. (2012), Pristiguana is not a member of the crown Iguanidae as argued Estes and Price (1973). Current phylogenetic position of Pristiguana is poorly supported, and needs to be reconsidered. The effect of the low number of scored characters (8.8%, Daza et al., 2012) needs to be evaluated critically; it is also possible that direct examination of the specimen could improve our understanding of the phylogenetic position of this fossil species. Nevertheless, without the referring of Pristiguana into Iguanidae, the Upper Cretaceous iguanids mentioned above (Apesteguía et al., 2005; Nava and Martinelli, 2011), along with Campanian Cnephasaurus and Maastrichtian Pariguana from North America (Gao and Fox, 1996; Longrich et al., 2012a), would indicate an important distribution of the Iguanidae in the Americas prior to the end of the Cretaceous.
Overall the South American Mesozoic lizard record is incomplete. During the Lower Cretaceous the lizard fossils are exceptionally well preserved, but restricted to the Araripe fauna of Brazil, with most-likely basal forms of unclear phylogenetic relationships. The situation in the Upper Cretaceous is inverted with poorly preserved fossils that are widely dispersed. The Scleroglossa Mesozoic record is very poor, limited to a tentatively referred single dentary from the early Campanian of Patagonia.
Altogether, present data suggests that the South-American Mesozoic lizards were probably more diverse and abundant than previously thought when the existing known data were restricted to few taxa (e.g., Evans, 2003). Taxa with well-supported systematic assignations (e.g., Nava and Martinelli, 2011; Brizuela and Albino, 2011) correspond to the two main clades, Iguania and Scleroglossa (sensu Gauthier et al., 2012), suggesting that Mesozoic lizard diversity is not properly illustrated by the existing fossil record and that more taxa will eventually appear.
With respect to Serpentes, remains of this clade are common and diverse in Cretaceous deposits from South America, especially in Patagonia. The record includes some of the most primitive forms of terrestrial snakes, the oldest of which is Najash rionegrina from the early Upper Cretaceous (Cenomanian–Turonian) of Patagonia (Apesteguía and Zaher, 2006). This species was described based on cranial elements and a nearly complete and articulated postcranial skeleton that included a sacrum and almost complete hind limbs (Apesteguía and Zaher, 2006; Zaher et al., 2009). A phylogenetic analysis of these fossil materials indicates Najash to be the most basal snake, lying outside the clade consisting of all other snakes (Apesteguía and Zaher, 2006; Zaher et al., 2009). Recently, Palci et al. (2013) modified the diagnosis of this species in the light of relevant considerations that cast doubt on the attribution of type and referred specimens. According to the reevaluation of the fossil material, the subsequent phylogenetic analyses rejected the hypothesis that N. rionegrina occupies a position as the most basal snake. Depending on the outgroup, Najash is placed (1) in a position basal to all living snakes, but more derived than the fossil forms Pachyrhachis, Eupodophis, and Haasiophis, or (2) as the most basal representative of a clade of fossil snakes that is the sister group of living snakes, or (3) as the most basal representative of a clade of fossil snakes that is located between the Scolecophidia and the Alethinophidia (Palci et al., 2013). Najash would have been a snake with body size around two meters long (Zaher et al., 2009). This size is inconsistent with a subterranean lifestyle for this snake suggested by Apesteguía and Zaher (2006). The morphology of the vertebrae of Najash allows only for the exclusion of extreme ecological habits, thus, Najash would have been a semi-fossorial, semi-aquatic, or surface living form of snake (Albino, 2011a; Palci et al., 2013).
Another snake from the South American Upper Cretaceous is Dinilysia patagonica, originally described by Smith-Woodward (1901) based on a specimen composed by a skull and associated fragments of vertebral column. This species is now the best characterized Cretaceous snake from several fragmentary cranial and postcranial skeletons recovered from various sites in Patagonia, from the Santonian to the early Campanian (Albino, 2007, 2011a). First, Smith-Woodward (1901) noticed Dinilysia has resemblances with aniliids and booids. Then, Romer (1956) classified Dinilysia in its own family, the Dinilysiidae, which is nested within the Booidea. Estes et al. (1970) conducted a detailed study of the skull and considered this snake to be closely related to aniliids and booids but with a greater number of lizard-like characteristics than those present in modern snakes. Rage (1977) estimated that Dinilysia represents the sister group of macrostomatan snakes, whereas Hecht (1982) supported the relationships of Dinilysia with aniliids and booids on the basis of the vertebral morphology. Rieppel (1979) regarded Dinilysia as the sister group of the Alethinophidia. More recently, Dinilysia has been included in several phylogenetic analyses (Caldwell, 1999; Rieppel and Zaher, 2000; Scanlon and Lee, 2000; Tchernov et al., 2000; Lee and Scanlon, 2002; Scanlon, 2006; Conrad, 2008; Zaher and Scanferla, 2011; Gauthier et al., 2012; Scanferla et al., 2013). All these recent hypotheses show Dinilysia as a relatively basal snake appearing as the sister group of the Alethinophidia or as the sister group to all extant snakes (Scolecophidia + Alethinophidia). Dinilysia was a medium-sized snake of around two meters long, with a relatively large head, and large, dorsally exposed orbits that presumably exploited semi-aquatic or semi-fossorial environments (Albino and Caldwell, 2003; Albino, 2007).
The alternative phylogenetic positions of Najash and Dinilysia demonstrate that non-fossorial snakes, with bodies greater than one meter long, likely consumed prey of diverse shapes and sizes, and appear to have developed early in snake phylogeny (pre-macrostomatan) (Albino and Caldwell, 2003; Albino, 2007, 2011a). Thus, they do not provide specific support to the presumption of a subterranean origin of snakes as opposed to the hypothesis of an aquatic origin for this group of squamates, but, together with the record of madtsoiids (see below), indicate that the earliest diversification of terrestrial snakes (i.e., not strictly aquatic nor subterranean) could have occurred in Gondwana (Albino, 2011a).
The record of South American Cretaceous snakes is enhanced with the occurrence of some extinct genera belonging to the Madtsoiidae. These were found in several sites from the Campanian–Maastrichtian of Patagonia (Albino, 1986, 1994, 1996a, 2000, 2007, 2011a; Martinelli and Forasiepi, 2004). The madtsoiids are interpreted as probably a monophyletic group (Scanlon, 1993, 2005; Albino, 1996a) that includes small, medium, and large-sized surface living forms. Recently, Mohabey et al. (2011) considered the presence of a large parazygantral foramen located within a fossa as the unique feature diagnosing Madtsoiidae, and they recognized this character as restricted to the Cretaceous–Paleogene large-bodied taxa Madtsoia, the African Gigantophis, and, possibly, the Australian Wonambi and Yurlunggur. However, the smaller Patagonian madtsoiid genera have exactly the same feature on their vertebrae (Albino, 1986, 1994), a large and deep parazygantral foramen adjacent to the zygantrum (Fig. 2). Mohabey et al. (2011) have not detailed arguments to refuse homology; then, this character seems to be a strong apomorphy uniting madtsoiids. Madtsoiids are described to comprise at least 10 genera (LaDuke et al., 2010), five of which have representatives in South American deposits from the Late Campanian–Early Maastrichtian to Eocene (Simpson, 1933; Hoffstetter, 1959; Albino, 1986, 1993, 1994, 1996a, 2000, 2007; Rage, 1998; Martinelli and Forasiepi, 2004). The phylogenetic information that they provide is limited because the remains consist mostly of isolated and fragmented vertebrae, although they are frequent in continental deposits of Cretaceous age. The genera of madtsoiids currently recognized in the Cretaceous of Patagonia include Alamitophis (with two species, A. argentinus and A. elongatus), Patagoniophis (P. parvus), and Rionegrophis (R. madtsoioides) (Albino, 1986, 1994). Alamitophis and Patagoniophis are also recorded from the early Eocene of Australia (Scanlon, 1993, 2005). The early Eocene of Australia also provided a rib fragment referred as cf. Madtsoia sp. (Scanlon, 2005), which is a genus of madtsoiid well documented in Paleogene deposits of Patagonia (see below). Thus, the chronological and geographical distribution of these three madtsoiid genera suggests a biogeographical continuity between Australia and southern South America extending across Antarctica between the late Cretaceous and the early Eocene (Scanlon, 1993, 2005; Albino, 2000, 2007). Although phylogenetic position of madtsoiids is debated, the distribution of madtsoiid genera during the Cretaceous and Paleogene in all major Gondwanan landmasses, except Antarctica (LaDuke et al., 2010) suggests they may have been a product of the earliest Gondwanan diversification of terrestrial snakes as well as for Najash and Dinilysia.
Another South American Cretaceous snake is the probable “anilioid” Australophis anilioides from the Late Campanian–Early Maastrichtian (Gómez et al., 2008). “Anilioidea” is likely a paraphyletic group of basal alethinophidians, prior to the divergence of macrostomatan (sensu Lee and Scanlon, 2002; Gauthier et al., 2012). Australophis would be closer in morphology to early Cenozoic Hoffstetterella from Brazil and extant South American Anilius than to any other snake. This would suggest an early divergent lineage (Gómez et al., 2008). This snake is small, less than one meter in length, likely with fossorial habits similar to the remaining “anilioids.” Its prey was likely restricted to those smaller than its head diameter.
There are two additional Mesozoic squamates from the Campanian–Maastrichian of South America with uncertain phylogenetic affinities. One of them is a dentary that Gómez (2011) found similar to dentaries of small madtsoiids and to those referred by Zaher et al. (2009) to the basal legged snake Najash. However, the dentaries referred to Najash by Zaher et al. (2009) have been recently removed from this taxon by Palci et al. (2013). The Campanian–Maastrichtian dentary was found associated with several isolated snake vertebrae of madtsoiids and it likely belongs to one of the already known madtsoiid taxa (Gómez, 2011). The other remain is an isolated vertebra distinct from Najash, Dinilysia and madtsoiids, and that lacks morphological affinities with any extant group (Albino, 2000). This specimen has been proposed to belong to Serpentes incertae sedis, although some plesiomorphic features were also recognized (Albino, 2000, 2007). Scanlon and Hocknull (2008) stated that this vertebra may not represent a snake but it rather resembles a specimen of cf. Coniasaurus sp. of Australia, an aquatic dolichosaur related to mosasaurs and snakes. This is probably the right conclusion about this vertebra (Albino, 2011a).
As a whole, the Mesozoic record of squamates suggests that lizards and snakes differ in their diversity and abundance in the Patagonian region through time. This disparity could be attributed, at least in part, to a biased fossil record because lizards are scarce in the Patagonian Cretaceous, but more abundant in other South American sites of the same age. This would demonstrate that they were usual components of the Mesozoic South American herpetofauna. At least two major clades of lizards that survive at the present are also found in the Mesozoic record: Iguania and Scincomorpha, whereas snake findings mostly include extinct and/or relatively basal representatives (Najash, Dinilysia, madtsoiids, and “anilioids”), most of them without well established relationships with the extant major clades Scolecophidia and Alethinophidia. The updated Mesozoic record demonstrates that South America was richer in both lizards and snakes than previously accepted, and that at least for snakes the southern continents played an important role in the early evolution of the group.
The Poor Squamate Record of the South American Paleocene
Based on published data, the South American Paleogene lizard record appears to be scarce in comparison to that of the Mesozoic, whereas snakes show an important Paleogene diversity but mostly concentrated in the Eocene. Nevertheless, unpublished observations provide evidence of a notable diversity of both Paleogene lizards and snakes in South America, including various genera and species of diverse clades. Patagonian deposits contrast with those of lower latitudes in having a relatively lower richness of fossils of lizards. It is not currently know whether the differences between these squamate faunas reflect the lack of lizards in southernmost Paleogene environments or poor sampling of small fossils in Patagonian Paleogene deposits.
Paleocene faunas of Patagonia, such as the well-known Peligran SALMA (Fig. 1), have not provided, thus far, remains of squamates, although lizards and snakes are well represented in other South American regions with Paleocene deposits (de Muizon et al., 1983; Rage, 1991; Head et al., 2009; Scanferla et al., 2013).
The early Paleocene of Tiupampa in Bolivia (Tiupampan SALMA, Fig. 1) has yielded several unnamed lizards and snakes. de Muizon et al. (1983) first noted lizard remains from the early Paleocene of Bolivia which was later listed by Rage (1991). This author described a dentary that has some characters that resemble those of the Teiidae, but since teeth would be deeply pleurodont, the author considered the specimen as of a probable Iguanidae. Several vertebrae were associated to this dentary but they were dismissed by Rage (1991) for systematic purposes. Five other tooth bearing fragments were also recovered from the Tiupampan of Bolivia and considered from different lizard taxa that cannot be allocated within Iguanidae, but without other diagnostic characteristics (Rage, 1991). If confirmed, the systematic placement of the Paleocene dentary to Iguanidae given by Rage (1991), it would be the earliest Iguanidae in the South-American Cenozoic and consistent with the Mesozoic record indicating an ancient presence of Iguanidae on the continent. Apart from the lizards, the Tiupampan fauna of Bolivia has produced snake vertebrae of diverse groups, including an indeterminate “anilioid,” at least two new unnamed genera of small to medium-sized boids, a madtsoiid or boid, and a tropidophiid, all of them indeterminate at lower taxonomic levels (de Muizon et al., 1983; Rage, 1991). This rich snake assemblage demonstrates that boids and tropidophiids were present in South America at least since the early Paleocene. More recently, a new genus and species of snake based on a partial skull was described from these sediments (Kataria anisodonta) representing the oldest macrostomatan skull recovered (Scanferla et al., 2013). The fossil mostly includes snout bones of an articulated skull. Its phylogenetic relationships show it would be a derived macrostomatan, basal to the Caenophidia+Tropidophiidae (Scanferla et al., 2013). This finding, together with other evidences, suggests that the origin and early diversification of derived macrostomatans may have taken place in southern continents during the early Cenozoic (Scanferla et al., 2013).
Apart from the records from Bolivia, an exceptionally large-bodied boine snake was described from the middle–late Paleocene of Colombia: Titanoboa cerrejonensis, based on articulated and isolated precloacal vertebrae from at least 28 individuals (Head et al., 2009). Vertebrae of Titanoboa are the largest recovered to date for any extant or fossil snake. The estimates of body size for Titanoboa greatly exceed those for the largest verifiable body lengths for extant Python and Eunectes (Head et al., 2009). It would likely have had an anaconda-like ecology (semi-aquatic), with a body length of about 13 m, and a mass of about 1,135 kg, suggesting that this snake may have required a minimum mean annual temperature of 30–34°C to survive (Head et al., 2009). Phylogenetic relationships of Titanoboa among Boinae were not tested, although Head et al. (2009) united Titanoboa with the extant Boa constrictor based on the presence of paracotylar fossae and foramina. Also, the concave anterior border of the zygosphene in dorsal view of Titanoboa vertebrae is a peculiar feature shared with Boa constrictor but not other extant boine genera (Albino, 2011b). Both of these characters together with a zygosphene narrower than the cotyle (wider in Boa) only appear in combination in two Boinae from the Eocene of Patagonia: Chubutophis grandis from the Casamayoran, and an indeterminate genus from the Mustersan, suggesting closer relationships of Colombian Titanoboa with these two Patagonian snakes than to extant Boa (Albino, 2012).
Finally, Rage (1981) mentioned the presence of a snake of the genus Coniophis from Laguna Umayo, Peru (latest Paleocene–earliest Eocene), but the material has not yet been described.
Latest Paleontological Records
South American Pleistocene and Holocene deposits containing squamate specimens are distributed in many countries, but fossils from Argentina are notable in quantity and diversity.
Extant genera of Teiidae and Iguanidae are the main components of the Pleistocene lizard fauna of South America, but Gekkonidae and Anguidae are also recorded. In general, the locality records are consistent with the recent ranges of the respective taxa. As in the beginning of the Neogene, Tupinambis was extensively distributed during the Pleistocene in Argentina, Bolivia, Brazil, and Uruguay (Rusconi, 1937; Hoffstetter, 1963, 1968; Estes, 1983; Ubilla and Perea, 1999; Hsiou 2007). But, unlike the Miocene record of the genus, many of these specimens have not been described, and in some cases the material could not be found (e.g., Rusconi, 1937). Hsiou (2007) described the extinct species Tupinambis uruguaianensis from Brazil, though its validity has been recently questioned (Scanferla et al., 2009; Brizuela, 2010). The Teiinae “Ameiva” and Dicrodon have also been mentioned for the Pleistocene of Brazil and Ecuador (Hoffstetter, 1970; Estes, 1983), but the remains have still not been described. The mention of Dicrodon is interesting since it is a basal Teiinae thought to have originated during the Eocene–Oligocene by vicariance (Guiliano et al., 2007). Teiinae and Tupinambinae remains have also been described from Argentina and several Brazilian Pleistocene–Holocene sites and they were mostly determined at levels of genera and species. Camolez and Zaher (2010) described many specimens from Brazil that they assigned to Tupinambis, “Ameiva” ameiva, and “Cnemidophorus” ocellifer. Albino (2005) described “Cnemidophorus” sp. from Argentina, an occurrence outside the current range of the genus. The Holocene Teiidae record is similar to that of the Pleistocene, with many references but few descriptions and all within their recent distributional limits. There are mentions of Tupinambinae (Tupinambis sp., Tupinambis teguixin) and Teiinae (“Ameiva” sp., “Cnemidophorus” ocellifer, Teius oculatus) in Brazil (de Queiroz, 2004; Camolez and Zaher, 2010; Hsiou et al., 2012), whereas all Holocene references from Argentina are of Tupinambis sp. (de la Fuente, 1999; Brizuela, 2010). Thus, it is possible to note that Teiidae have become more diverse since the end of the Miocene, with many extant genera of both subfamilies.
Regarding Iguanidae, they were also found within their recent geographical limits, but unlike Teiidae, most cited fossils have been described. Hoffstetter (1970) mentioned Iguana sp. in the Pleistocene of Ecuador. Late Pleistocene of southern South America has provided Iguanidae from Chile (Liolaemus cf. lineomaculatus) and Argentina (Leiosaurus bellii, Liolaemus sp.) (Van Denvender, 1977; Núñez et al., 2005; Agnolin and Jofré, 2011). Moreover, Camolez and Zaher (2010) described iguanids of the subfamilies Polychrotinae* (cf. Enyalius, Polychrus sp.) and Tropidurinae* (Tropidurus sp.) from several Holocene sites of Brazil, whereas Albino (2005) described a very rich and diverse association of lizards among which at least two species of Liolaemus were identified (L. darwinii and L. multimaculatus).
Some lizard families have a very poor fossil record, this is the case of Anguidae and Gekkonidae, both part of the extant South-American lizard fauna. The only described South American Anguidae (Diploglossus cf. D. fasciatus) is from the Holocene of Brazil (Camolez and Zaher, 2010). Gekkonidae (Homonota) are thus far only known from the late Pleistocene (Agnolin and Jofré, 2011) and late Pleistocene–Holocene (Albino, 2005) of Argentina. Homonota is a semiarid adapted gecko that is part of two fossil squamate associations that draw attention because of their diversity and, because both fossil associations are at lower latitudes than comparable Recent associations. Both associations are from central Argentina. The oldest, from the late Pleistocene, is composed by Homonota, Anops kingii and Liolaemus. The other association is from the late Pleistocene–Holocene, and it includes Teiidae (“Cnemidophorus” sp.), Tropidurinae* Iguanidae (Liolaemus darwinii, Liolaemus multimaculatus, Liolaemus sp.) and Gekkonidae (Homonota sp.). In both cases the ensembles are indicative of more arid and cooler environments during the latest Pleistocene of central Argentina.
Regarding Amphisbaenia, a short portion of an articulated column of an amphisbaenid, originally considered as a lizard (Rusconi, 1937), was recovered from the mid Pleistocene of Argentina. Later examination of the fossil determined it corresponds to an extinct species of Amphisbaenidae (Amphisbaena marelli, Torres and Montero, 1998), although the validity of this species has recently been questioned (Scanferla et al., 2009). The late Pleistocene record of Amphisbaenia in South America is geographically dispersed. From Brazil, Gans and Montero (1998) described two extinct species of Amphisbaenidae on the basis of cranial remains: Amphisbaena braestrupi and Amphisbaena laurenti. Along with these taxa, Gans and Montero (1998) also described an Amphisbaenia incertae sedis, represented by cranial and postcranial elements. However, part of these remains could be of Amphisbaena laurenti (Gans and Montero, 1998). Cranial remains of Amphisbaenidae have also been recovered in the late Pleistocene of Argentina and assigned to Anops kingii and Amphisbaena heterozonata (Scanferla et al., 2006; Agnolin and Jofré, 2011). Likewise, Hoffstetter (1968) mentioned two vertebrae from the late Pleistocene of Bolivia. One of them is incomplete but the other is similar to those of Leposternon, although Hoffstetter (1968) did not provide the characters supporting this assertion. Last, Camolez and Zaher (2010) described an Amphisbaenidae indet. and Leposternon sp. from the Holocene of Brazil. At present, and regarding Leposternon as an Amphisbaenidae (contra Kearny, 2003), only this family is represented in the South-American Neogene.
With respect to snakes, most Pleistocene remains could be assigned to living taxa. The record includes Boidae, “Colubridae,” Viperidae and Elapidae. In Argentina, boids are represented by Boa constrictor from the Lujanian age (Late Pleistocene, Fig. 1) (Albino and Carlini, 2008). The specimens described in this article represent the first snake record for the Lujanian and provide the minimum age for this species of boid snake. These fossils are outside of the range of the modern distribution of the species, because Boa constrictor is not found today in the Mesopotamian region of Argentina (Entre Ríos, Corrientes and Misiones provinces). This species occupies environments of the semiarid “Chaco” biome in provinces west to the Paraná river (Catamarca, Córdoba, Chaco, Formosa, Jujuy, La Rioja, Mendoza, Salta, San Juan, San Luis, Santa Fé, Santiago del Estero, and Tucumán). The paleoclimatic interpretation based on mammals indicates that, at least during a large part of the Pleistocene, the Mesopotamian region would have had a more humid and warmer climate, with a marked influence of Brazilian fauna that would have continued even during the last interglacial period (Carlini et al., 2004). The break between the Mesopotamian and the Brazilian fauna subsequent to the Late Pleistocene, which occurred as a result of the changes in the main rivers that delimit this region (Paraná and Uruguay), would have established the insularity conditions of Mesopotamia, leading to the extinction of some Brazilian elements. The disappearance of B. constrictor from this area may be attributed to this isolation (Albino and Carlini, 2008).
“Colubrids” are well represented in the Miocene and Pliocene of South America through indeterminate forms (see above), but fossils of extant genera are just described for the Pleistocene. Several extant “colubrid” genera have been recorded from the Lower–Middle Pleistocene of Argentina, including the pseudoboines Clelia sp. and Boiruna cf. B. maculata (Scanferla, 2006), and the xenodontines Lystrophis and Philodryas (Scanferla, Cenizo and de los Reyes, 2005; Scanferla, Agnolin and Boglino, 2009). Clelia (C. rustica) and Philodryas (P. patagoniensis) have also been documented from the Upper Pleistocene–Holocene (Albino, 1999, 2001). Among them, records of Clelia, Philodryas and Lystrophis are from regions where they are distributed today (Albino, 1999, 2001; Scanferla et al., 2005). The record of Boiruna in the Buenos Aires province is outside the current geographical range of the species of the genus, indicating warmer environmental conditions at the end of the Ensenadan age (Fig. 1) than today (Scanferla et al., 2009).
Within viperids, records from Argentina increase during the Pleistocene–Holocene (Albino, 1995, 1999; Albino et al., 2002; Scanferla and Nenda, 2005). They include the presence of the extant genus Bothrops with the species B. alternatus in the Upper Pleistocene–Holocene (Albino, 1999, 2001). A possible earlier record of the genus comes from the Lower–Middle Pleistocene in northeastern Buenos Aires province (Scanferla and Nenda, 2005). This last fossil indicates that during more humid periods, Bothrops lived in areas where they do not currently inhabit, probably because these areas are now too arid and urbanized (Scanferla and Nenda, 2005; Albino and Montalvo, 2006).
Snake fauna from the Late Pleistocene to Recent found in Brazil has been recently studied by Camolez and Zaher (2010). It includes the four genera of boids that occur presently in this country: Boa (B. constrictor), Corallus (cf. C. hortulanus), Epicrates (E. cenchria), and Eunectes (E. murinus and Eunectes sp.). “Colubrids” are represented by Mastigodryas cf. M. bifossatus, Chironius sp., Pseutes cf. sulphureus, Tantilla sp., Helicops gr. leopardinus, and cf. Philodryas. Some vertebrae were referred by these authors to the viperids Bothrops sp. and Crotalus durissus. Finally, cranial and vertebral remains were identified as the species Micrurus corallinus and Micrurus sp., representing the first record of elapids in South America. In addition to the Brazilian record, the late Pleistocene of southwestern Brazilian Amazonia has yielded a vertebra of an indeterminate viperid (Hsiou and Albino, 2010).
Apart from the diverse Pleistocene squamate fauna of Argentina and Brazil, Porta (1965) observed the presence of the “colubrid” Synophis aff. S. bicolor in the Late Pleistocene of Colombia, whereas indeterminate viperids come from the Upper Pleistocene of Venezuela (Head et al., 2006). The reports of boid and viperid remains in the upper Pleistocene of Bolivia (Hoffstetter, 1968) cannot be confirmed, because the material has not been described or illustrated.
This summarized record of snakes suggests that some elements of the South American snake fauna have been well established since the Pleistocene.