Selenogonus narinoensis Stirton, 1947 (Tayassuidae, Cetartiodactyla, Mammalia): taxonomic status and paleobiogeographic implications

The species Selenogonus narinoensis was described by Stirton (1947) based on a single specimen which comes from sediments cropping out in the Cocha Verde locality, Nariño Department (Colombia), tentatively referred to the late Pliocene–Pleistocene (MGN 931; IGM p002118, Museo Geológico Nacional, Servicio Geológico Colombiano, Bogotá). However, morphological studies and comparative morphometric observations of the specimen suggest that (1) no diagnostic character supports the validity of the species Selenogonus narinoensis (here considered species inquirenda); (2) a combination of features (e.g., the mandibular condyle located behind the posterior edge of the vertical mandibular ramus, the angular process which projects laterally outwards, a bunolophodont crown morphology, a mesodont crown height, and a simple crown morphology of the third lobe of m3) indicates it belongs to the genus Platygonus; (3) this specimen corresponds to one of the largest South American peccaries; (4) taking into account certain anatomical characters as well as its morphometric range, this specimen is assigned to Platygonus cf. marplatensis. Even though the stratigraphic provenance of the specimen is still doubtful, it can be proposed that (1) it could be one of the most ancient records of tayassuids in South America, as would be expected given its geographical position, and (2) considering the new taxonomic proposal, this specimen represents the first record of Platygonus cf. marplatensis in Colombia and represents one of the northernmost South American records of the genus. This new interpretation would be of great relevance in the Great American Biotic Interchange due to its strategic geographical proximity to the Isthmus of Panama.

The oldest unequivocal records of fossil tayassuids in South America date back to the late Pliocene (ca. 4-3.3 Ma; see Prevosti et al., 2006;Gasparini, 2013) and were found in sediments cropping out in southeastern Buenos Aires Province, Argentina. In contrast, Frailey and Campbell (2012), followed by Prothero and Pollen (2013), support the hypothesis of an early arrival of the Tayassuidae in South America, earlier than late Miocene and well before most of the Great American Biotic Interchange (GABI) that had occurred by the late Pliocene (Cione et al., 2015;O'Dea et al., 2016). This is based on the supposed record of peccaries, among other mammals of Nearctic origin (e.g., gomphotheres, tapirs, dromomerycine ruminants), coming from the Madre de Dios Formation (late Miocene) outcrops in southeastern Peru (Frailey and Campbell, 2012;Prothero and Pollen, 2013;Prothero et al., 2014) and on the divergence times suggested by molecular analyses (Theimer and Keim, 1998;Gongora and Moran, 2005;Perry et al., 2017). However, the chronological information about the sediments containing these remains and their taxonomic assignment is not accurate (Perini et al., 2016;Parisi Dutra et al., 2017a;Gasparini et al., 2021). According to these authors, the supposedly extinct Miocene species are indistinguishable from modern tayassuids (Tayassu pecari and Dicotyles tajacu), and the stratigraphic provenance of the specimens is highly dubious, and the fossils are likely Quaternary in age.
In Colombia, very few remains of tayassuids have been registered (Stirton, 1947;Villarroel et al., 1989;Gasparini, 2013;Moreno Mancilla et al., 2019). One of them is the holotype and single specimen of Selenogonus narinoensis Stirton, 1947, which comes from sediments cropping out at the Cocha Verde locality, Nariño Department, tentatively referred to the late Pliocene-Pleistocene. However, the chronological information of the fossil-bearing sediments is not accurate. In turn, Menégaz and Ortiz Jaureguizar (1995), considering the hypotheses of Kraglievich (1959, pp. 233-234) and Reig (1981, p. 41), suggested a lower-to middle-Pliocene age for those fossil-bearing sediments. Fossil remains described as Platygonus sp. were found in sediments cropping out in the Villa de Leyva locality (Boyacá Department), but chronologic and stratigraphical information is still unknown . Other Tayassuidae remains were registered in Late Pleistocene sediments cropping out in the paleontological site named Los Hoyos located at the Tatacoa Desert (Huila Department) (Tayassu aff. tajacu sensu Villarroel et al., 1989).
The goals of this contribution are (1) to report the results of study of the fossil specimen from Nariño (Colombia); (2) to comment on the systematic validity of Selenogonus narinoensis Stirton, 1947;and (3) to discuss the paleobiogeographic and paleoecological significance of this peccary.

Measurements abbreviations
Hrmv: height of the vertical branch of the mandible, taken from the ventral part of the angular process to the depression between the coronoid process and the condyle; Hrmhm3: depth of horizontal ramus below m3; Lm3: maximum length of molar 3 in a parallel line to the sagittal plane; Am3: maximum width of molar 3 in a perpendicular line to the sagittal plane.  Cione et al., 2015); coastal cliff in Chapadmalal region and Miramar (General Pueyrredón and General Alvarado counties), in southeastern Buenos Aires Province, Argentina (Kraglievich, 1952;Reig, 1952;Quintana, 2002;Gasparini, 2004;Gasparini, 2013) (Fig. 2).

Remarks
Castellanos (1927) referred some teeth exhumed from the southeastern coast of Buenos Aires Province to Homo chapadmalensis. Vignati (1941) included these remains in the species Homo neogaeus Lehmann Nitsche, 1907, a species based on a fossil found in the Monte Hermoso cliff. Reig (1952) established the species Platygonus marplatensis based on a partial mandible found in the Chapadmalal Formation, Barranca de los Lobos, General Pueyrredón county, Buenos Aires Province. Kraglievich (1959) created the genus Argyrohyus with the species Argyrohyus chapadmalensis and included in the synonym Homo chapadmalensis Castellanos, 1927; Homo neogaeus sensu Vignati, 1941; and Platygonus marplatensis Reig, 1952. Menégaz andOrtiz Jaureguizar (1995) suggested that A. chapadmalensis could be a synonym of Platygonus marplatensis.
Concerning the specific reassignment, it is worth mentioning that Rusconi (1930) included the species Listriodon chapadmalensis Ameghino, 1908, in the genus Platygonus. This generates a case of secondary homonymy (they are homonyms as a consequence of a new combination and not as a consequence of origin), and it must be solved by priority. Thus, the species of Ameghino, 1908, may continue to be called Platygonus chapadmalensis, while for that of Castellanos, 1927, a replacement must be considered. Therefore, the oldest name is Platygonus marplatensis Reig, 1952(see Wetzel, 1977Gasparini, 2007Gasparini, , 2013Parisi Dutra et al., 2017c).

Description
This mandible corresponds to a large specimen. It has a shallow but sharply outlined masseteric fossa and a deep internal temporal fossa. The angular process projects forward to the middle part of m3, and its lower edge is slightly projecting laterally outwards. The lower border of the horizontal ramus is sharply hooked (accentuated) distally. The mandibular condyle is located behind the posterior edge of the vertical mandibular ramus. The coronoid process ascends behind m3.
The lower molar 3 only preserves the distal pair of main cusps (hypoconid and entoconid) and the third lobe, which has a well-developed single cusp, and a pair of smaller accessory cusps both located labially and lingually. Despite its incompleteness, a mesodont and bunolophodont dental morphology is observed.

Measurements
See Table 1.

Taxonomic comments and comparisons
Stirton (1947) erected the genus and species Selenogonus narinoensis for this specimen based on some features that basically coincide with those of the Platygonus genus. Mainly these features are Large, evidently about equal in size to Platygonus uquiensis Rusconi; masseteric fossa shallow but sharply outlined; internal temporal fossa deep; [. . . ] ascending ramus relatively and actually higher than in other peccaries; angle below lower border of horizontal ramus, sharply hooked anteriorly; [. . . ] m3 hypso-brachyodont; evidently more lophiodont than in Platygonus. (Stirton, 1947, p. 322) However, these features do not differ from those described for Platygonus, given the diversity observed within the genus (Gasparini, 2007;De los Reyes et al., 2014). In addition, the following morphological features support the inclusion of this specimen within the genus Platygonus: mandibular condyle located behind the posterior edge of the vertical mandibular ramus, angular process projected laterally outwards, bunolophodont crown morphology, The length of m3 is partial because it only preserves the distal pair of main cusps (hypoconid and entoconid) and the third lobe. The development of mesodont and bunolophodont cheek teeth in Platygonus differs from the brachyodont and bunodont morphology observed in Tayassu pecari, Dicotyles tajacu, Parachoerus carlesi, and Brasiliochoerus stenocephalus; from the mesodont and bunodont cheek teeth typical of Catagonus metropolitanus and Catagonus bonaerensis; and from the mesodont crown height and bunodont with high crowns (referred to as "zygodont" by some authors (Gasparini, 2007;Prothero and Grenader, 2012;Gasparini et al., 2013)) observed in Parachoerus wagneri.
The sharp outline of the masseteric fossa is remarkable in the specimen studied. A similar condition was observed in Platygonus marplatensis, Platygonus chapadmalensis, Platygonus sp. (MMP 1212), Brasiliochoerus stenocephalus among South American species, and Platygonus compressus among North American species.
The lower margin of the horizontal ramus in the specimen studied is also singular. A sharply hooked (accentuated) distal condition was observed only in the holotype of Platygonus marplatensis (MMP-S Type 25; see Fig. 1c). However, this character is not present in the other specimens assigned to this species (MACN 5420 type synonym, MMP-S 200, MMP-S 188, MMP-S 199, MMP-S 674, MACN 19725, MACN 19726) due to their partial preservation.
The incomplete state of preservation of the material prevents us from obtaining absolute measurements. However, taking into account its relative size, this specimen is within the proportions observed in species of the genus Platy- gonus (see Table 1). It is worth mentioning that the vertical mandibular ramus of this specimen is one of the largest among South American species, only comparable to P. marplatensis, P. chapadmalensis, and P. scagliai and to the largest specimens of North American Platygonus (e.g., Platygonus cumberlandensis Gidley, 1920).
The specimen under study has no diagnostic characters that support the species Selenogonus narinoensis as valid and identifiable (species inquirenda). The combination of features suggests it pertains to the genus Platygonus. As a consequence, based on the reflected taxonomic problems and taking into account certain anatomical characters (e.g., the position of the mandibular condyle, the outline and laterally outward projection of the angular process, the crown height and morphology of the tooth, and the distal condition of the lower margin of the horizontal ramus) as well as its morphometric range, this specimen is considered Platygonus cf. marplatensis.

Paleobiogeographical aspects
The Tayassuidae represent one of the first North American immigrant mammals as well as the first ungulates that entered South America during the GABI (Gasparini, 2013;Cione et al., 2015;O'Dea et al., 2016).
According to phylogenetic analyses as well as chronological and geographical evidence, Platygonus represents a Tayassuidae lineage that originated in North America and corresponds to the first group of peccaries that entered South America during the late Pliocene (Prevosti et al., 2006;Gas-parini and Ubilla, 2011;Gasparini, 2013;De los Reyes et al., 2014;Parisi Dutra et al., 2017c). The greatest specific diversity of Platygonus is recorded during the late Pliocene-earliest Pleistocene in South America, mainly in Argentina (represented by P. marplatensis, P. chapadmalensis, P. scagliai, P. kraglievichi, and Platygonus sp.) and Uruguay (Platygonus sp.) Gasparini, 2013, and bibliographies cited therein). If the stratigraphic provenance of the fossil specimen found in Nariño (Colombia) is confirmed, it could be one of the most ancient records of tayassuids on this continent, as would be expected given its geographical position. In addition, considering the new taxonomic proposal, this specimen represents the first record of Platygonus cf. marplatensis in Colombia. At the same time, it is remarkable that this Colombian record, together with the one registered in Villa de Leyva, represents the northernmost South American records of the genus Platygonus. This could have had great relevance in the Great American Biotic Interchange due to its strategic geographical proximity to the Isthmus of Panama.
In the Early to Middle Pleistocene, the taxonomic diversity and abundance of records of Platygonus are notably reduced (P. cinctus) and the genus Catagonus (C. metropolitanus) appears reliably for the first time in the paleontological record and only in Argentina (Gasparini, 2013).
The greatest specific diversity and abundance of tayassuids is documented during the Middle Pleistocene to Late Pleistocene-Early Holocene in South America (e.g., Argentina,Brazil,Uruguay,Bolivia,Peru,Colombia,and 72 G. M. Gasparini et al.: S. narinoensis: taxonomic status and paleobiogeography Venezuela) represented by Catagonus, Parachoerus, Brasiliochoerus, Tayassu, and Dicotyles (Gasparini, 2013;Parisi Dutra et al., 2017b). In the Middle Pleistocene the genera Brasiliochoerus, Tayassu, and Dicotyles and probably the species Parachoerus carlesi and C. bonaerensis appear for the first time in the South American fossil record (Gasparini, 2013). In the Late Pleistocene-Early Holocene, P. wagneri is registered for the first time (e.g., Uruguay) and P. carlesi and C. bonaerensis are reliably recorded (e.g., Argentina and Uruguay) on this continent. At the Pleistocene-Holocene boundary (about 10 000 years ago) all megamammals and most of the large mammals became extinct in South America in a remarkable extinction event. The tayassuids were affected by this extinction, and only 25 % of their taxonomic diversity has survived. Currently, peccaries are represented by only three species (Tayassu pecari, Dicotyles tajacu, and Parachoerus wagneri) (Gasparini, 2013;Parisi Dutra et al., 2017c;Acosta et al., 2020).

Ecological and anatomical considerations
The peccaries of the genus Platygonus have certain morphological features that allow some inferences about their habits: a great development of nasal sinuses and chambers, orbits located in a superior-posterior position and behind upper molar 3 due to elongation of the rostrum, a possession of a distinct basicranial flexure, a laterally outward projection of the angular process of the jaw (providing a greater surface for insertion of the lateral deep masseter muscle), a reduction of the lateral digits in the limbs, a mesodont crown height, and a bunolophodont crown morphology, among others. Therefore, this set of anatomical characters indicates that these large-sized peccaries have diurnal habits, a herbivorous diet, and probably a foraging habit and lived in dry and relatively open environments (Guilday et al., 1971;Wetzel, 1977;Menégaz and Ortiz Jaureguizar, 1995;Gasparini, 2007;. Paleoecological studies (e.g., dental microwear and isotopic analysis) on North American Platygonus have shown a C 3 -browser to mixed-feeder diet and probably a C 4 -grass diet, under special conditions (Feranec and MacFadden, 2000;Feranec, 2007;Schmidt, 2008).
The faunal changes that have occurred since the late Pliocene could have been strongly influenced by climate (Cione et al., 2015, and references cited therein). The open and arid environments of great latitudinal extent developed during glacial cycles allowed the dispersion of Platygonus, Catagonus, Brasiliochoerus, and Parachoerus.
Based on paleontological records, as well as certain anatomical features (e.g., crown teeth morphology, limb development) linked to diet and life habits, together with body mass, it can be inferred that the species of Catagonus, Brasiliochoerus, and Parachoerus have replaced those of Platygonus since the Middle Pleistocene, probably as a consequence of the reduction of the open environments for which Platygonus species would be more specialized. This predom-inantly arid or semi-arid and cold period alternating with brief more humid and warmer pulses would have allowed the late expansion of the species of Tayassu and Dicotyles.

Taxonomic final remarks
Morphological studies and comparative morphometric observations of the Colombian specimen suggest that (1) no diagnostic character supports the validity of the species Selenogonus narinoensis (here considered species inquirenda); (2) a combination of features (e.g., the mandibular condyle located behind the posterior edge of the vertical mandibular ramus, the angular process that projects laterally outwards, a bunolophodont crown morphology, a mesodont crown height, and a simple crown morphology of the third lobe of m3) indicates it belongs to the genus Platygonus; (3) this specimen corresponds to one of the largest South American peccaries; and (4) taking into account certain anatomical characters as well as its morphometric range, this specimen is assigned to Platygonus cf. marplatensis.

Paleobiogeographical remarks
Even though the stratigraphic provenance of the specimen found in Nariño (Colombia) is still doubtful (late Pliocene or Pleistocene?), it can be proposed that (1) it could be one of the most ancient records of tayassuids in South America, as would be expected given its geographical position, and (2) considering the new taxonomic proposal, this specimen represents the first record of Platygonus cf. marplatensis in Colombia and, together with the one registered in Villa de Leyva, establishes the northernmost South American records of the genus. This new interpretation would be of great relevance in the Great American Biotic Interchange due to its strategic geographical proximity to the Isthmus of Panama.
Author contributions. GMG conducted the analysis and wrote the manuscript with contributions from all co-authors. OFMM prepared the figures. OFMM and JLC were in charge of the logistics to study the specimen (MGN 931, IGM p002118) housed at the Museo Geológico Nacional (MGN), Servicio Geológico Colombiano (Bogotá, Colombia).