Divisadero Largo Formation

Locality: Camp 3, Mina Atala, vicinity of Cerro Divisadero Largo, ca. 8 km west Mendoza city (Fig. 1D). From this site, the team collected and numbered the specimens 20-58M to 46-58M (Table 1). The lower levels of the Divisadero Largo Formation (i.e., zone with anhydrite) were dated in 55.02 ± 0.82 myr (Paleocene–Eocene; Mescua et al., 2017), which indicates a maximum age. However, it should be note that the bearing levels of the Divisadero Largo fauna are located higher from the dated level.

Remarks: Testudines are represented by fragmentary remains of the carapace, such as MCZ VP 101618 (edge of carapace; Fig. 2A), MCZ VP 101619 (large part of carapace, much crouched and broken), and MCZ VP 101620 (portion of carapace; Fig. 2B). These specimens add to other testudine records from the Divisadero Largo Formation (Minoprio, 1947; Pascual and de la Fuente, 1993).

Referred material: MCZ VPM-7406, fragments of mandible; MCZ VPM-7408, parts of both mandibular rami with left p4–m2 (Fig. 2C); MCZ VPM-7409, fragment of maxillary with left P3–M3 (Fig. 2D); MCZ VPM-7410, fragment of maxillary with left M1–2 (Fig. 2E); MCZ VPM-7411, fragment of mandible with left m2–3 (Fig. 2F); MCZ VPM-7413, fragment of mandible with p1–m2; MCZ VPM-7414, incomplete right mandible; MCZ VPM-7417, fragment of maxillary with right P1–4; MCZ VPM-7420, two fragments of maxillary. See more details in Simpson et al. (1962) and López (2008).

Referred material: MCZ VPM-7415, numerous small fragments of upper and lower jaws with identifiable parts of right P4–M1 (Fig. 2G). See more details in Simpson et al. (1962) and López (2008).

Referred material: MCZ VPM-7412, fragment of maxilla with left M1–3 (the M3 is broken and preserves only the paraloph; Fig. 2H); MCZ VPM-7423, fragments of cranium and postcranium.

Remarks: Specimen MCZ VPM-7412 was included in the referred material of A. atalaensis by Simpson et al. (1962: 271) as coming from the type locality (vicinity of Cerro Divisadero Largo). According to the information in Romer's field notebook (1958), this specimen comes from the ‘Casamayor’ level, Unit 2.

Many years after, in his revision of the fauna from Divisadero Largo, López (2008) maintained MCZ VPM-7412 as A. atalaensis but had doubts about the taxon of MCZ VPM-7423.

Referred material: MCZ VPM-7418, incomplete mandible with right p3–m1 and left c, and p1–3 (Fig. 2I); MCZ VPM-7422, fragment of mandible with right p3–p4?, distal end of femur, and right astragalus (Fig. 2J–L).

Remarks: These specimens were not published by Simpson et al. (1962), while MCZ VPM-7422 was referred with doubts to Xenostephanus chiotti by López (2008).

Lower teeth MCZ VPM-7422 have considerably eroded occlusal surfaces, but some features resemble Oldfieldthomasia, like how the p3 is longer than the p4, and the metaconid lingually curves and directs distally. The astragalus MCZ VPM-7422 is characterized by having a symmetrical and high trochlea with parallel but inclined lateral and medial sides (∼36° from the vertical axis), a short neck relative to the trochlea, and a well-developed dorsal crest. In proximal view, a foramen seems to be present into a long sulcus, but sediment does not permit detailed observation. In posterior view, a wide and deep inarticular sulcus separates the ectal facet from the sustentacular facet. The ectal facet is a rectangular and concave surface inclined at ∼50° from the vertical axis. The sustentacular facet is wide, convex, and well-delimited from the head by a deep sulcus. It is a bit larger (L = 9.9 mm) than that of Notopithecus (Vera, 2012) and differs from it and Allalmeia atalaensis (Lorente et al., 2014) in having a symmetrical and more inclined trochlea, and a shorter and wider neck (Fig. 2L). In turn, its features resemble Colbertia (Bergqvist et al., 2007), excepting a well-protruding tibial protuberance among other differences in the latter species. The morphologies indicated for this material do not allow us to conclude a taxonomic assignment, so we prefer to consider that it corresponds to a Notoungulata indet.

Referred material: MCZ VPM-7416, fragment of lower jaw with part of p4, and badly preserved m1 and m2 (Fig. 2M); MCZ VPM-7407, maxillary fragment with right M2–3 (Fig. 2N), the M2 preserves only the lingual half and the M3 lacks the ectoloph; Table 1.

Remarks: both specimens were included in the referred material of Phoradiadius divortiensis by Simpson et al. (1962), who also figured the M3 of MCZ VPM-7407, and López (2008). According to the field notebook, these specimens, including 20-58M to 25-58M (not figured), come from the lower Buff Layer, Minoprio's ‘Musters,’ Unit 3.

Other Fossils of the Romer Collection from the Divisadero Largo Formation

Other reports from the Divisadero Largo Formation correspond to the following: specimens 29-58M (ramus fragment with two molars) and 33-58M (ramus fragment with m2–3) attributed to ‘Notoungulata indet.’; and specimen 24-58M (part of skull or mandibles) listed as ‘Groeberia?’. However, these specimens could not be located in the collection, as well as in the cases of 27-58M (maxillary fragment with two molars), 36-58M (ribs and unidentified bones), 41-58M (incomplete ramus with teeth), 42-58M (fragment of ramus with teeth), and 43-58M (fragment of maxilla with broken teeth; Table 1). Additionally, MCZ VPM-7421 (miscellaneous), MCZ VPM-7419 (fragments of maxillary and mandible), and MCZ VP 101621 (skeletal bones) are too fragmentary to be identified.

Aisol Formation

Locality: Camp 1, north of Mina Zitro (Fig. 1B).

Remarks: according to Romer's unpublished field notebook (1958), 15 specimens were collected in this locality, numbered as 1-58M to 15-58M. The fossils of this locality pertained to the Aisol Formation (Burdigalian, early Miocene; Garrido et al., 2014).

Referred material: MCZ VP 101597, fragment of right mandible with four teeth (mf1 to mf4; Fig. 3A–B); Table 1.

Dimensions: mf1 (L = 9.0 mm, W = 5.8 mm), mf2 (L = 12.6 mm, W = 6.9 mm), mf3 (L = 16.8 mm, W = 8.7 mm), mf4 (L = 18.8 mm, W = 7.8 mm), molariform series length (mf1–mf4, 59.5 mm), and maximum mandibular depth at the last tooth level (46.7 mm). There is a miniscule and shallow alveolus in front of the mf1 (Fig. 3A), which probably belongs to the cavity for a vestigial incisor (Scott, 1903; González Ruiz et al., 2020 and its respective references). The posteroventral margin of the symphysis extends to the widest part of the first lobe of the mf4. The mf1 is obliquely oriented (anterolingual–posterolabial) with respect to the mf2–mf4, which are anteroposteriorly aligned. It has a subtriangular occlusal outline, whose longer side coincides with the labial face; the labial and linguodistal faces are convex, while the mesiolingual face is concave. The mf2 presents a subtriangular outline. This molariform differs in occlusal shape with the mf1 by having a distolabial end conspicuously extended and a concave distal face. This peculiar extension confers to the mf2 an asymmetrical shape (Fig. 3A). The mf3 clearly differs from the mf2 in having trilobed labial and lingual faces; that is, the anterior, middle, and posterior lobes limited by two labial and two lingual slight grooves. The mf4 is a trilobed tooth as mf3 but differs from it in having deeper labial and lingual grooves, a better-developed labial half-middle lobe, and larger anterior and posterior lobes.

Remarks: The features observed in MCZ VP 101597 here described (e.g., mf1 and mf2 are subcircular or subelliptic in transversal section, and mf3 with slight indication of lobation) characterize to the propalaehoplophorine glyptodonts (Carlini et al., 2008). Propalaehoplophorinae represents one of the early-divergent Glyptodontidae groups (late Oligocene–middle Miocene), whose genera were initially recognized in Argentinian Patagonia (Carlini et al., 2008). During the Santacrucian SALMA (late early Miocene), the Propalaehoplophorinae showed great diversification, evidenced in the recognition of five genera and eight species: Propalaehoplophorus Ameghino, 1887b (P. australis and P. minor), Cochlops Ameghino, 1889 (C. muricatus and C. debilis), Eucinepeltus Ameghino, 1891 (E. complicatus and E. petestatus), Metopotoxus Ameghino, 1898 (M. anceps), and Asterostemma Ameghino, 1889 (A. depressa; see Scott, 1903). From these five genera, confidently assigned mandibular fragments correspond to Propalaehoplophorus and Eucinepeltus, while references to mandibular specimens of Cochlops, Eucinepeltus, and Metopotoxus were not clearly assessed, published, lost, or even more their validity questioned (see González Ruiz et al., 2020). Other early Miocene Glyptodontidae records correspond to Parapropalaehoplophorus septentrionalis from the Chucal Formation (northern Chile; Croft et al. 2007).

Comparing with glyptodonts of similar age to that considered for Aisol Formation, tooth dimensions of specimen MCZ VP 101597 are larger than in Propalaehoplophorus australis, P. minus, Eucinepeltus petesatus from the Santa Cruz Formation (González Ruiz, 2010; González Ruiz et al., 2020), and Parapropalaehoplophorus septentrionalis from the Chucal Formation (northern Chile, early Miocene; Croft et al., 2007). Its horizontal ramus height and the morphology of the mf2 are comparable to that of Pa. septentrionalis (Croft et al., 2007: fig. 6). However, MCZ VP 101597 distinguishes from Pa. septentrionalis in having the first three molariforms different in occlusal shape between them (mf1 trapezoidal, mf2 subtriangular, and mf3 is labially and lingually trilobed), while Pa. septentrionalis has quite similar subtriangular mf1–mf3 as Eucinepeltus (González Ruiz et al., 2020: fig. 9). In turn, the trilobed configuration of mf3 in MCZ VP 101597 (Fig. 3A) is comparable to the same tooth of Propalaehoplophorus.

In sum, features that closely relate MCZ VP 101597 to Propalaehoplophorus are a mandibular predental zone markedly everted, the horizontal ramus with a convex ventral margin (Fig. 3B), mf1 different from mf2, and a trilobed mf3. However, the trapezoidal morphology of the mf1 and the distolabially extended mf2 of MCZ VP 101597 also resembles Parapropalaehoplophorus. This suggests that MCZ VP 101597 could represent a new species of Propalaehoplophorus or even a different taxon. However, considering the lack of clearly assessed and published references to mandibular specimens of other Propalaeohoplophorinae genera (e.g., Cochlops, Eucinepeltus, and Metopotoxus) the taxonomic attribution of MCZ VP 101597 is cf. Propalaehoplophorus.

Following a previous report in the Aisol Formation by Soria (1983), Forasiepi et al. (2011) indicated the presence of three glyptodontids: Propalaeohoplophorinae gen. et sp. indet. (isolated osteoderms from different individuals, MACN–PV 18608, MACN–PV 18612) and Hoplophorinae gen. et sp. indet (a small fragment of caudal tube, MACN–PV 15074; probably belonging to Lomaphorini). Unfortunately, on account of the fragmentation of the remains we cannot refine the taxonomic identification.

Referred material: MCZ VP 101606 comprises two osteoderms—MCZ VP 101606a, fixed osteoderm (Fig. 3C); and MCZ VP 101606b, moveable osteoderm (Fig. 3D; see Table 1).

Remarks: following the standard terminology for describing cingulate osteoderms, the figures are raised areas of the osteoderm; sulci are the grooves that delineate figures; and piliferous pits are round indentations on the surface of the osteoderm that have been interpreted as being associated with hairs (see Croft et al., 2007 and references therein).

MCZ VP 101606a (L = 12.65 mm; W = 7.92 mm; Fig. 3C) has six peripheral figures surrounding the principal figure, which is elongated and convex. The two anterior figures are pentagonal in contour with a plane to slightly convex surface, while the lateral figures (two in each side) are rectangular and externally convex. Piliferous pits are observed in the posterior side. MCZ VP 101606b (L = 14.18 mm; W = 9.28 mm; Fig. 3D) displays three longitudinal figures similar-sized and convex: two lateral figures, each with two faint transverse grooves, and a principal figure separated from the lateral figures from shallow grooves. Piliferous pits are placed in these grooves between the lateral and principal figures, and in the posterior side. The features described in both specimens are characteristic of the osteoderms of the pelvic shield and moveable band of Stenotatus.

Stenotatus is a genus typical from Patagonia (Argentina), including five species (S. patagonicus, S. hesternus, S. ornatus, S. centralis, S. planus) from the late Oligocene to middle Miocene (Scillato and Carlini, 1998). Outside of Patagonia, Stenotatus sp. nov. was recorded in the early Miocene Chucal fauna in northern Chile (Croft et al., 2007). Stenotatus sp. was also recorded in the Chinches bearing level, oldest fauna of the Chinches Formation (San Juan Province, Argentina; Santacrucian SALMA; see López et al., 2011; fig. 4b). López et al. (2011) related the specimen from San Juan to Stenotatus sp. nov. described by Croft et al. (2007) and with Stenotatus cf. S. patagonicus described by Kramarz et al. (2010) for the early Miocene of Gran Barranca (Chubut, Argentina).

Comparatively, osteoderms of MCZ VP 101606 from the Aisol Formation seem to be larger than the osteoderms from the Chucal (Croft et al., 2007) and Chinches faunas (López et al., 2011). They also differ from S. planus (Cerro Boleadoras Formation, Santa Cruz Province; Scillato and Carlini, 1998; fig. 1d) in having a more stylized principal figure and a wider groove separating it from the peripheral figures in MCZ VP 101606a, and a plane and narrower anterior side in MCZ VP 101606b.

With the available evidence (only two osteoderms), we opt to identify both specimens as Stenotatus sp. This report constitutes the first mention of a Dasypodidae Eutatini Stenotatus for the Aisol Formation, increasing the diversity of the Cingulata previously recorded for this outcrop (Forasiepi et al., 2011; Garrido et al., 2014).

Referred material: MCZ VP 101594, left m3 (Fig. 3E–F); MCZ VP 101596, fragment of mandible with a right m3 (Fig. 3G–H); Table 1.

Remarks: The m3 of MCZ VP 101594 (Fig. 3E–F) is characterized by having a vertical groove in the center of the lingual face of the talonid, which extends over the entire length of the crown. This groove confers a bilobed shape to the lingual border of the talonid (Fig. 3F). The m3 MCZ VP 101596 (Fig. 3G–H) has broken the lingual face of the trigonid and there is a fracture in the middle of the talonid. This specimen is probably associated with the m3 MCZ VP 101594 because of the similar size of both teeth (Table 2) and the shared morphology, such as a continuous layer of enamel and the presence of a lingual sulcus on the talonid, although it is less evident in the molar MCZ VP 101596 because of a fracture in this area (Fig. 3H).

Among Mesotheriidae from the Aisol Formation, Eutrachytherus modestus Roth, 1899, was reported based on specimen MNHSR-PV 1152 (Garrido et al., 2014). These authors also reported that the specimen MNHSR-PV 154 from the overlying La Huertita Formation could correspond to Pseudotypotherium exiguum (Ameghino, 1887a) or P. subinsigne (Rovereto, 1914). The La Huertita Formation lacks absolute dating but was referred to Montehermosan–Chapadmalalan SALMA (Zanclean–Piacenzian; Garrido et al., 2014). All previous mentions correspond to the upper dentition, which cannot be compared with the specimens here described from the Romer collection (lower teeth of MCZ VP 101594 and MCZ VP 101596). It should be noted that Eutrachytherus Ameghino, 1897 is a nomen illegit pro Trachytherus Ameghino, 1889 (Mones, 1986) and no other mention concerning the species E. modestus was performed, although it was also described on the basis of the upper dentition (Roth, 1899).

More recently, a new taxon of Mesotheriinae, Rusconitherium mendocense Cerdeño, Vera, and Combina, 2018, was described from the fauna of the Mariño Formation (Santacrucian SALMA; Mendoza Province). These authors established general similarities between the specimen from the Aisol Formation (MNHSR-PV 1152) and R. mendocense, inferring that the former probably belongs to the same species, although they emphasized that comparison with lower teeth of mesotherines from the Aisol Formation is necessary to check their hypothesis. Specimens MCZ VP 101594 and MCZ VP 101596 share with R. mendocense a lingual sulcus on the talonid of the m3, but they differ from it in having larger dimensions and a continuous layer of enamel. On the contrary, in R. mendocense there is a posterolingual interruption of the enamel in the m3, which is shared with specimen Ar-96-49 attributed to a Mesotheriinae gen. et sp. nov. from Las Hornillas level of the Chinches Formation, San Juan Province (López et al., 2011).

In turn, the shape of the m3 of MCZ VP 101594 and MCZ VP 101596, with an elongated talonid, a pointed posterior end, the continuous layer of enamel, and a pronounced lingual sulcus in the talonid are shared with both species of Altitypotherium (A. paucidens and A. chucalensis) and Eotypotherium chico from the Chucal Formation (Chile; Croft et al., 2004). However, dimensions of the m3 MCZ VP 101594 and MCZ VP 101596 (Table 2) are larger than E. chico, smaller than A. paucidens, and comparable to those of A. chucalensis (Croft et al., 2004; tabs. 2, 4, 5), which also shares similarities with MHNSR-PV 1152 (Garrido et al., 2014).

Considering the scarce material available, we cannot be certain about its taxonomic identity below subfamily level. However, the morphology and dimensions of MCZ VP 101594 and MCZ VP 101596 allows establishing affinities with Altitypotherium sp. from the Chucal Formation (Chile), supporting the early Miocene age (Santacrucian SALMA) attributed to the fossil-bearing bed of the Aisol Formation (Forasiepi et al., 2015).

Specimens MCZ VP 101594 and MCZ VP 101596 constitute the first record of mesotheriids from the Mina Zitro locality and the first lower dentition reported to this group from the Aisol Formation.

Referred material: MCZ VP 101591, two fragmentary cervical vertebrae (Fig. 3I–J); MCZ VP 101592, proximal epiphysis of tibia (Fig. 3K) and a patella (Fig. 3L); MCZ VP 101593, phalanx I (Fig. 3M) and fragments of two distal metapodials (Fig. 3N–O).

Remarks: It should be noted that an inconsistency was detected regarding the provenance and identification of the skeletal elements that compose MCZ VP 101593. According to the field notebook, it was listed as a ‘Macraucheniid?’ coming from the Aisol Formation, but there is a label with the bones indicating that it is a Litoptern from the ‘Huayquerías de San Carlos’ (>200 km further north from the Aisol Formation; Fig. 1A, C). We consider that this label was incorrectly associated with MCZ VP 101593, and it probably belongs instead to the specimen MCZ VP 100410 listed as a litoptern from the Huayquerías Formation (see below).

Vertebrae of MCZ VP 101591 (Fig. 3I–J) do not preserve processes, but they have a long centrum, which is typical of litoptern macraucheniids. The fragment of tibia and patella MCZ VP 101592 (Fig. 3K, L) are poorly preserved, but they resemble those figured for Theosodon (Scott, 1910; pl. XX, figs. 4, 5).

The phalanx of MCZ VP 101593 (L = 41.67 mm; Fig. 3M) is symmetrical with a funnel profile. The proximal epiphysis is wider lateromedially (W = 21.6 mm) than the distal epiphysis (W = 18.36 mm). Based on its dimensions and ratios, it could belong to the digit III; it is slightly larger than the same bone of Theosodon (Scott, 1910). The articular surface of the distal epiphysis is antero–posteriorly extended, but more expanded over the posterior face. Both metapodials MCZ VP 101593 (Fig. 3N–O) have a dorsoventrally, well-developed median keel on the distal trochlea. The great extent of the keel is characteristic of noncentral metapodials (Mp II and IV) in many macraucheniids (Shockey, 1999).

Postcranial bones of Macraucheniidae for the Aisol Formation were already reported by Soria (1983) and Forasiepi et al. (2011), who suggested these probably correspond to Theosodon, the most common genus of Macraucheniidae Cramaucheniinae for the early Miocene. However, the bones here described are not enough to perform a confident generic determination.

Other Fossil Remains of the Romer Collection from the Aisol Formation

Other groups reported from the Aisol Formation at the MCZ collection correspond to: Aves, MCZ VP 101611 (proximal end of ulna); mammal megatheres, MCZ VP 101612 (two phalanges), MCZ VP 101614 (distal end of radius), and MCZ VP 101615 (partial humerus, ulna, and fragment of undetermined bones); a rodent indet., MCZ VP 101610 (isolated tooth; Table 1); and a toxodontid, MCZ VP 101613 (axis and six fragments of vertebrae), but its fragmentary condition does not allow the authors to identify it properly. Finally, an ‘astrapothere?’ (a distal end of humerus) and an invertebrate gastropod (field number 5-58M) were also reported in Romer’ unpublished field notebook (1958), but unfortunately, they could not be found in the collection (Table 1).

Mariño Formation

Locality: Camp 4, Agua de las Avispas, Cacheuta Basin (Fig. 1E).

Remarks: According to Romer's unpublished field notebook (1958), only two specimens were collected from the Mariño Formation (17.96 ± 0.2 and 17.4 ± 1.3 myr, Burdigalian; Buelow et al., 2018) at this locality, 47-58M and 48-58M (Table 1).

Referred material: MCZ VP 101622, part of a skull (posterior braincase; Fig. 3A). It was found approximately 20 m above cliff formed by ‘Areniscas Entrecruzadas,’ near end of NE spur of Cuchilla del Toscal, not far SW of division between Río Seco de Cacheuta and Arroyo Seco de las Minas (Fig. 1E).

Remarks: Concerning specimen 48-58M, it was not found at the MCZ collection. According to the field notebook, this set of fossils includes various isolated gastropod shells referred to Corbicula elchaensis Rusconi, 1949. This material was discovered in a white lens 1 m thick, on top of cliff formed by ‘Areniscas Entrecruzadas,’ approximately 800 m SW from specimen MCZ VP 101622.

Tunuyán Formation

Locality: Baños de la Salada area, 15 km East of San Carlos, Mendoza (Fig. 1C).

Remarks: This area is situated at the Huayquerías de San Carlos, West from the Meseta Sinclinal del Guadal (Synclinal Plateau; Fig. 1C). In this site, the team collected four specimens, 16-58M to 19-58M (Table 1) from the Tunuyán Formation (≤5.84 Ma for the lower levels; latest late Miocene–early Pliocene; Garrido et al., 2017). It is important to note that the explorers indicated that their specimens were exhumed from the same area where the procyonid Chapalmalania altaefrontis Kraglievich and de Olázabal 1959 was found; that is, in the upper section of the unit (Chiesa et al., 2019 and its respective references).

Referred material: MCZ VP 101595, a right mandible with four broken molars (Fig. 4B–C); Table 1.

Remarks: The specimen MCZ VP 101595 (Fig. 4B–C) is at least 190 mm in length, although it is an incomplete and restored mandible. Height of the mandibular bone at the level of the last tooth is approximately 69.4 mm. The length of the third molariform is approximately 9.45 mm and the length of the first molariform is 9.0 mm. This specimen was catalogued as belonging to Pronothrotherium Ameghino, 1907, a genus of Nothrotheriinae (Megatheriidae). This reference could be related to the mention of a nothrotheriine sloth similar to Pronothrotherium, but from a more primitive aspect, in the Cacheuta Basin at 30 km to the southwest from Divisadero Largo (Simpson et al., 1962: 281). However, the size of MCZ VP 101595 clearly surpasses those reported for P. typicum Ameghino, 1907, P. subtypicum Rovereto, 1914, P. mirabilis (Kraglievich, 1925), and P. figueirasi Perea, 1988 (but see Gaudin et al., 2020, and Discussion). On the contrary, the size and features of MCZ VP 101595 (e.g., squared molariforms aligned in a continuous series and separated by equidistant spaces and the oblique orientation of the transverse crests) correspond to megatheriines (Fig. 4B). Furthermore, the mesiodistal compression observed in the molariforms, inclination of the molariform ridges with respect to the sagittal plane, and relatively shallow dentary (Fig. 4C) are known in Megathericulus Ameghino, 1904, and Pyramiodontotherium Rovereto, 1914 (De Iuliis et al., 2004). The former genus, Megathericulus, was recorded from middle Miocene faunas (see Brandoni et al., 2020), while Pyramiodontotherium is a Megatheriinae known in the late Miocene to early Pliocene faunas (De Iuliis et al., 2008). These authors recognized three species of Pyramiodontotherium—P. bergi (Moreno and Mercerat, 1891) and P. brevirostrum Carlini, Brandoni, Scillato-Yané and Pujos, 2002, both from Bajo de Andalhuala (Catamarca Province), and P. scillatoyanei De Iuliis, Ré and Vizcaíno, 2004, which comes from the lower member of the Toro Negro Formation (La Rioja Province, dated 6.87–4.95, late Miocene–early Pliocene; Amidón et al., 2016). Other reports refer to Pyramiodontotherium sp. from the Ituzaingó Formation (Entre Ríos Province, Brandoni and Carlini, 2009, and references therein).

Before this report, megatheriids were not known in the Tunuyán Formation. Thus, the mandible MCZ VP 101595 here identified as Pyramiodontotherium sp. constitutes the first report of a Megatheriinae, which increases the list of xenarthran taxa previously identified in the Tunuyán Formation (see Pascual and de la Fuente, 1993; Chiesa et al., 2019).

Referred material: MCZ VP 100410. At present, it is a set with four bone fragments with the same number, here referred as MCZ VP 100410a–d (Fig. 4D–G); Table 1.

Remarks: MCZ VP 100410a (Fig. 4D) is the occipital region of a skull. It is tall and displays the general features described for Huayqueriana cf. H. cristata Rovereto, 1914 from the Huayquerías Formation (Forasiepi et al., 2016: 34; fig. 18) in having a vertical ridge (i.e., external occipital crest) descending from the nuchal crest, separating two lateral depressions that are laterally bordered by occipital protuberances, and a circular foramen magnum. However, in Huayqueriana cf. H. cristata the occipital condyles are dorsoventrally compressed, differing from the longer and convex condyles in MCZ VP 100410a.

The other fragments of bones correspond to the following: MCZ VP 100410b, articular head of a humerus (Fig. 4E); MCZ VP 100410c, proximal end of a tibia (Fig. 4F); and MCZ VP 100410d, a fragment of proximal radius (Fig. 4G). They are probably associated with the occipital MCZ VP 100410a. Unfortunately, the fragmentary condition of these remains does not allow us to assert on their taxonomy beyond subfamily level.

Other Fossils of the Romer Collection from the Tunuyán Formation

Other groups reported for the Tunuyán Formation correspond to the following: a megathere, MCZ VP 101617 (several fragments of skull?) and a rodent catalogued as Ctenomyidae (field number 18-58M, fragment of molar). This latter specimen, however, was not found in the collection (Table 1).

La Pilona Formation

Locality: 1 km approximately north of Puesto Agua del Cajón, Mendoza Province (Fig. 1E). From this unit (Serravalian–Tortonian age, middle-late Miocene), the expedition collected specimens with the field numbers 49-58M to 52-58M (Table 1).

Referred material: MCZ VP 101598, right half of a badly preserved skull with P3 in eruption and P4–M2 (Fig. 4H); Table 1.

Remarks: according to the field notebook, this specimen was referred to a Pachyrukhinae. However, the features of the teeth correspond to the Hegetotheriinae morphotype (see below).

Specimen MCZ VP 101598 is a juvenile individual evidenced by the incompletely erupted P3 and P4. Although the teeth are severely fractured (Fig. 4H), approximate measurements of the P4 are L = 6.9 mm and W = 3.9 mm. There is a large antorbital foramen above the P4–M1 level (diameter = 4.6 mm). The most anterior border of the zygomatic arch starts just behind the antorbital foramen. Size and morphology of MCZ VP 101598 resembles that of the Hemihegetotherium morphotype because of several features, such as imbricated upper check teeth with mesiolabially elongated parastyle and convex lingual face, trapezoidal-outlined M1 and M2, and the M2 wider and shorter than M1.

Hemihegetotherium comprises four widely distributed species in the middle to late Miocene faunas of Argentina and Bolivia: H. achataleptum Rovereto, 1914; H. torresi (Cabrera and Kraglievich, 1931); H. tantillum Vera, 2019; and H. trilobus Croft and Anaya, 2006 (more details see Vera, 2019 and its respective references). Specifically, in the Mendoza Province, the species H. achataleptum was described from the Huayquerías Formation based on a skull and a fragment of mandible (MACN PV 8491; Rovereto, 1914). Remains of Hemihegetotherium were also reported from the Río de los Pozos Formation (late Miocene–Pliocene; Yrigoyen, 1993) in the Tupungato area, approximately 100 km south from Cacheuta Basin (Cabrera, 1937; Yrigoyen, 1993). However, no collection numbers were provided, and subsequent studies did not refer to these specimens or provide descriptions or illustrations.

Considering that MCZ VP 101598 is a juvenile individual, and its teeth are severely damaged, we can establish comparisons with some of the Hemihegetotherium species. MCZ VP 101598 differs from H. achataleptum (MACN-PV 8491: L P4 = 8.4 mm; L M1 = 15 mm) and H. trilobus (Croft and Anaya, 2006) in its smaller size and having stylized P4 and M1 (subtriangular shaped), with shorter lingual faces and convex distal faces. In turn, MCZ VP 101598 is morphologically similar and closer in size to H. torresi (MLP 76-VI-12-25; see Vera, 2019; tab. 1) from the Arroyo Chasicó Formation (Buenos Aires Province), with which it shares subtriangular P3–4 with developed parastyle and paracone folds, a P4 more like P3 than M1, and a M2 quite differentiable from the M1.

Before this work, fossil vertebrate reports were unknown from the Pilona Formation in the Cacheuta area. The identification of the Hegetotheriidae Hemihegetotherium is consistent with the middle to late Miocene biochron of this genus. Moreover, we consider that MCZ VP 101598 has more affinities with H. torresi from the Chasicó Formation (Buenos Aires Province; 9.43–8.7 myr; Zárate et al., 2007) than to H. achataleptum from the Huayquerías Formation (5.84 ± 0.41 myr; Garrido et al., 2017). This is in concordance with the Serravalian–Tortonian dating provided for the Pilona Formation (12 to 8.4 myr; Irigoyen et al., 2000; Buelow et al., 2018).

Referred material: MCZ VP 101607, a left mandible with all teeth covered with sediment.

Remarks: According to the field notebook, it belongs to Pachyrukhinae. Despite being covered, we agree with the determination, but this specimen cannot be identified with certainty because of its condition.

Referred material: The set MCZ VP 101698 is composed of several specimens probably associated with the same individual. To describe all the specimens in this set, we use letters to differentiate each one.

Specimen MCZ VP 101698a is a fragment of the mandible with left p3–m2 (m2 is broken). A thick layer of cement covers all the teeth. The enamel is absent in the mesial face of p3 and the distolingual end of p3–m1. The p4 is more like the molars than the premolars (i.e., molariform); that is, the talonid is larger than the trigonid and like that of m1. This latter feature differentiates this specimen from Tremacyllus from the Huayquerías Formation (Vera and Ercoli, 2018). According to its size (Table 2), it corresponds to the genus Paedotherium. It is larger than P. typicum and the specimen MLP 66-IV-23-8 (= MCZ 19313, a specimen from the Las Vertientes, Mar del Plata), although MCZ VP 101698a has a smaller p3 than MCZ 19313. Specimen MCZ VP 101698b is a fragment of mandible with right m1–3 (the talonid of m3 is broken), which may be related to the fragment MCZ VP 101698a (Table 2).

Specimen MCZ VP 101698c is a fragment of the mandible with right i1–2 and p2–3, and left i1 and p2. The i2 is a reduced tooth in comparison to the i1, having a tear-shaped outline. The p2 is labially trilobed, differing from P. typicum (Ameghino, 1887a) and resembling P. minor Cabrera, 1937 (Cerdeño and Bond, 1998; Vera and Ercoli, 2018).

Specimen MCZ VP 101698d is a left m3, lingually broken and probably related to the specimen MCZ VP 101698a (Table 2). The MCZ VP 101698e is a right I1 (L = 8.2 mm; W = 1.4 mm). The MCZ VP 101698f is a left M1 or M2 (L = 4.7 mm; W = 3.0 mm) with cement on the lingual face.

Other Fossils of the Romer Collection from the La Pilona Formation

Rodentia indet.: Specimen MCZ VP 101623 is an isolated tooth, which was attributed to Pachyrukhinae, but its morphology is typical of a rodent (Fig. 4I).

San Juan Province (Fig. 1A, F)

Huachipampa Formation

Locality: Camp 6 at the YPF encampment, Sierra de Mogna, Huachipampa Basin, East from Tucunuco, Jáchal Department, San Juan Province (Fig. 1F). In this site, the team collected and numbered the specimens 53-58M to 70-58M.

Remarks: a few years after the MCZ–MACN Expedition in the area of Sierra de Mogna, one of the expedition members, Rosendo Pascual, published a new genus of Rodentia Cardiatheriinae Xenocardia diversidens Pascual and Bondesio (1963) from the Huachipampa Formation (Tortonian, late Miocene). This species was based on a specimen (MLP 57-XII-23-5) collected by Dr. H. de la Mota (YPF) and deposited at the MLP. In this work, Pascual and Bondesio (1963: 44) pointed out that the fossils from the Huachipampa Formation were originally determined by Dr. Bryan Patterson (MCZ) and R. Pascual (MLP), who established a correlation with the Arroyo Chasicó Formation and the Estratos de los Llanos (San Luis Province) and assigned a Chasicoan age to the fauna of the Huachipampa Formation. This agrees with the information detailed in the field notebook.

Referred material: MCZ VP 101626 (part of carapace, Fig. 4J) and MCZ VP 101629 (osteoderms, postcranial bones; Fig. 4K).

Referred material: MCZ VP 101599, a nearly complete mandible preserving left p4–m2 and right p4–m3 (Fig. 5A–B) and associated left humerus (Fig. 5C–E); MCZ VP 101603, fragment of maxilla with badly preserved right P3–M2? (Fig. 5F); MCZ VP 101604, fragment of maxilla with left P3–M2 and a fragment of mandible with unidentifiable teeth (Fig. 5G–H); MCZ VP 101601, fragment of right mandible with a broken p4 and the trigonid of the m1 (Fig. 5I); MCZ VP 101605, a left humerus without proximal epiphysis (according to the field notebook, this set also includes a facial region of skull and other bones, but these elements are presently lost); MCZ VP 101602, a left humerus (Fig. 5J–K), proximal fragment of right ulna (Fig. 5L–M), and proximal fragment of right radius (Fig. 5N–O).

Dentition: Among the tooth remains of mesotheres from the Huachipampa Formation, mandible MCZ VP 101599 is the best-preserved specimen (Fig. 5A–D). The morphology and size of its lower teeth, especially the occlusal shape of p4 have a mesially and labially pointed trigonid, resemble specimen MLP 12-1678 of Pseudotypotherium insigne (Ameghino, 1887a) from the Monte Hermoso Formation. The specimen MCZ VP 101601 (Fig. 5I) is a bit larger than MCZ VP 101599 (Table 2), although both could be related to the same taxon.

The upper teeth are represented by several fragments of multiple maxillae (Fig. 5F–G), but, unfortunately, the damaged teeth limit our comparisons. Despite that, the general morphology and dimensions are comparable to MACN Pv 8469, the type specimen of Pseudotypotherium subinsigne (Rovereto, 1914) from the Tunuyán Formation (Mendoza). Furthermore, MCZ VP 101604 and MCZ VP 101603 were collected at the same place and present the same state of preservation, which means both fragments may represent the opposite sides of the same palate.

Postcranial bones: The left humerus MCZ VP 101599 (Fig. 5C–D) is almost complete, lacking only its proximal epiphysis. It is a gracile bone, with a slender diaphysis and a minimally developed deltoid crest. Its distal end (W = 55.4 mm; AP = 33.0 mm; W trochlea = 43.7 mm) is wide and high relative to the narrow epiphysis, having a convex anterior aspect and a conspicuously concave posterior aspect. The trochlea is a wide and low articular surface, which is not as concave as that of Plesiotypotherium (Cerdeño et al., 2012: fig. 4). MCZ VP 101599 is characterized by a shallowly projected medial epicondyle and the epicondylar crest, which is well-developed forward of the bone rather than laterally, forming the pronounced concavity on the posterior aspect of the epiphysis. The humerus MCZ VP 101602 preserves the proximal epiphysis, which is mostly broken, but a posteriorly well-projected head is observed (Fig. 5K). The humerus of MCZ VP 101605 has a partially broken diaphysis and a large and well-marked medial trochlear crest, but it is similar to the humeri of MCZ VP 101599 and MCZ VP 101602.

The features observed in these humeri from the Huachipampa Formation (MCZ VP 101599, MCZ VP 101602, MCZ VP 101605) contrast with a more projected medial epicondyle and a laterally well-extended crest in other mesotheres, such as Plesiotypotherium (Cerdeño et al., 2012: fig. 4). Published data of mesotheres are available only for Trachytherus, Plesiotypotherium and Mesotherium, in which the humeri are robust, with a sigmoid shaft, massive and laterally extended deltopectoral crest, a more projected medial epicondyle and entepicondyle, and a laterally well-extended crest (Fernández-Monescillo et al., 2017). These features are clearly different from the humeri here attributed to Pseudotypotherium based on the associated dentition. In some respects, humeri of Pseudotypotherium sp. from the Huachipampa Formation are similar to Plesiotypotherium achirense Villarroel, 1974, sharing a large supratrochlear foramen and slender shaft.

Among the ulnae of mesotheres figured by Fernández-Monescillo et al. (2017; fig. 7), that of MCZ VP 101602 (Fig. 5L–M) is comparable to specimen MACN 7972 attributed to Pseudotypotherium sp. These specimens share a similar olecranon profile, both in medial and anterior aspects; in medial view, the distal part of the trochlear notch is anteriorly more projected than the proximal part; and a similar curvature of the posterior side of the olecranon. In this sense, this similitude between the ulnae confirms our attribution of MCZ VP 101602 to Pseudotypotherium.

The radius of MCZ VP 101602 (Fig. 5N–O) only preserves its proximal part. The epiphysis is anteroposteriorly flattened, having a lateromedial maximum diameter and a kidney-shaped contour. In proximal view, the articular facets form a sigmoidal and barely concave–convex surface. Published radii of Pseudotypotherium are unknown. Comparatively, MCZ VP 101602 shows a similar aspect to the radius of Plesiotypotherium (Fernández-Monescillo et al., 2017; fig. 8).

Based on our comparisons, the mesothere here described is certainly attributed to the genus Pseudotypotherium. The upper teeth (MCZ VP 101604 and MCZ VP 101603) are comparable in overall morphology to P. subinsigne from the Tunuyán Formation (Mendoza). This could mean the mesothere from San Juan represents the same species, a hypothesis also supported by the geographic proximity. However, the state of preservation of the upper teeth of the mesothere from Huachipampa Formation does not allow a suitable comparison with the holotype of P. subinsigne (MACN Pv 8469). Additionally, the lower dentition of this latter species is unknown until the present, impeding comparisons with MCZ VP 101599.

Previous reports of Mesotheriidae from San Juan Province correspond to Typotheriopsis chasicoensis Cabrera and Kraglievich, 1931, and T. ?silveyrai Cabrera, 1937, from the Lomas de las Tapias Formation (late Miocene; Contreras and Baraldo, 2011; Contreras et al. 2019). The former species, T. chasicoensis, was originally described in the Arroyo Chasicó Formation, Buenos Aires Province (Chasicoan SALMA), while T. silveyrai is known in the Huayquerian SALMA (Late Miocene, but extending to early Pliocene), coexisting in some localities with the genus Pseudotypotherium (Cerdeño, 2018 and bibliography cited there). Pseudotypotherium, in turn, was reported from the Pliocene (Montehermosan and Chapadmalalan faunas) from Buenos Aires (P. maendrum) and Mendoza (P. subinsigne) provinces (Cerdeño, 2018); in San Luis Province, Río Quinto Formation (P. subinsigne, P. maendrum, and Pseudotypotherium sp.; Chiesa et al., 2019 and its respective references), La Rioja Province, Salicas Formation (Pseudotypotherium sp.; Tauber, 2005), and La Pampa Province, Cerro Azul Formation (P. subinsigne; Sostillo et al., 2021).

Referred material: MCZ VP 101600, several fragments that allow us to reconstruct the right and left series of the same mandible: the symphysis with right i1–2 and left i1, right p3–m1, and left p3–m1 and isolated m3 (Fig. 6A–D); Table 1.

Remarks: The m1 of MCZ VP 101600 has a well-worn occlusal surface. A thick layer of cement covers the labial face of the teeth. Dimensions of MCZ VP 101600 are smaller than MCZ VP 101698 (Table 2), which was referred to Paedotherium. On the contrary, the size and morphology are like Tremacyllus impressus, which was identified in the Huayquerías and Tunuyán formations from Mendoza Province (Vera and Ercoli, 2018). However, the material is not enough to identify it at the species level, which is why we opt to identify it as Tremacyllus sp.

Reffered material: MCZ VP 101628, fragment of skull and mandible (Fig. 6E-F) .

Comparatively, teeth of specimen MCZ VP 101628 look similar to those of Xenocardia diversidens (Hydrochoeridae, Cardiatheriinae), also from the Huachipampa Formation (Pascual and Bondesio, 1963), but more comparisons are needed for a taxonomic identification.

Reffered material: 54-58M (a partial skull; lost); 67-58M (cranium; lost), and MCZ VPM-4739 (jaw; Table 1).

Other Fossils of the Romer Collection from the Huachipampa Formation

Other groups correspond to the following: bird? MCZ VP 101627 (Fig. 6H); mammal megathere MCZ 100411 (left ramus), and toxodontids MCZ VP 101624 and field number 62-58M (portion of right ramus with broken m2–3). However, these specimens are too fragmentary to be identified or are not found in the collection.