Recent evaluations of the Baño Nuevo-1 collections have led to the identification of an Arctotherium sp. (Carnivora, Ursidae, Tremarctinae) tooth fragment from the upper sections of Layer 5 (Late Pleistocene), associated to Macrauchenia sp., Equidae, Felidae, Lama guanicoe, Camelidae and Mylodontidae remains (mainly dermal bones). On the immediately overlaying Layer 4B, Diabolotherium cf. nordenskioldi, Lama guanicoe, Macrauchenia sp., Equidae, Camelidae, Felidae and Mylodontidae remains have also been recorded (López, 2009; Bostelmann, et al., 2011; López & Mena, 2011).

This sample (one specimen) opens new issues for the osseous record of the final Pleistocene at the site, as another possible taphonomic agent has to be considered, while discussing the faunal remains founded on rockselters from this time period at the Chilean central Patagonia in general. Beyond description and taxonomic identification, it also opens new grounds for paleoecological studies, such as those already developed in southern Patagonia (Martín, 2008; 2013; Martín & San Román, 2010).

The finding of an Arctotherium remain at Baño Nuevo-1 has been barely mentioned in previous studies (see López & Mena, 2011). Given its importance, however, this brief note aims to provide valuable information on its stratigraphical setting and chronological context and a detailed description as well as a preliminary discussion on the origins of the fossil.

Baño Nuevo-1 is located ca. 80 km, NE of Coyhaique (45° 17 S-71° 32′W) (Fig. 1). The cave is located on an Aptian volcanic complex, locally known as “Cerro Grande del Campo Seis” (Large Hill on Section 6). With a depth of 20 m and an average width of 4 m, the cave nowadays has a relatively restricted access due to the entry of coluvial debris fallen through a lateral shaft (Velásquez & Mena, 2006). The stratigraphic record has been previously presented in several articles (see Mena et al., 2000, 2003; Núñez et al., 2005; Mena & Stafford, 2006; Velásquez & Mena, 2006; Mena, 2009; López & Mena, 2011; Trejo & Jackson 1998).

The tooth specimen presented, was recovered from Layer 5, characterized by clay and organic sand, dated between ca. 13.500 and 11.200 cal ¹⁴C B.P. (see Fig. 2). Just like Layer 4A, this layer presents abundant fossil remains and organic sediments that could be assigned to Mylodontidae dung (Mena & Stafford, 2006; Núñez et al., 2005). This layer overlies Layer 6, composed by fine microlaminated sands deposited by a large Pleistocene proglacial lake devoid of any faunal remains (Mena & Stafford, 2006; Núñez et al., 2005).

The material corresponds to a left I₃. Currently, it is stored at the Facultad de Ciencias Sociales, Universidad de Chile, under the code Baño Nuevo-1/Unidad 5C/Capa 5/N°F171. The specimen was compared with collections of Arctotherium vetustum at the Museo Argentino de Ciencias Naturales Bernardino Rivadavia, Lycalopex culpaeus, Otaria sp. and Puma concolor from the Laboratorio de Zooarqueología of the Universidad de Chile, and Panthera onca from a private collection. Since part of the cusp is broken, mesio-distal and lingual-labial measurements had to be taken only at the crown base. The occlusal surface was observed through Scanning Electron Microscopy (SEM) (Zeiss, model EVO MA10) at the Departamento de Antropología of the Universidad de Chile.

Institutional abbreviations: MACN, Museo Argentino de Ciencias Naturales Bernardino Rivadavia.

Carnivora Bowdich, 1821

Ursidae Gray, 1825

Tremarctinae Merriam and Stock, 1925

Arctotherium Burmeister, 1879

Arctotherium sp.

Material: Baño Nuevo-1/Unidad 5C/Capa 5/N° F171, an isolated I₃ (Fig. 3A, B, C and D).

Geographical occurrence: Baño Nuevo-1 cave, Coyhaique, Central Patagonia, Chile. UTM: 0301821 E-4981638 N (Fig. 1A-B).

Stratigraphical occurrence: Layer 5, Late Pleistocene dated between ca. 13.500 and 11.200 cal ¹⁴C B.P. (Fig. 2).

Nowadays the Tremarctinae Subfamily can be found exclusively on the Americas (Soibelzon, 2004). Four genera are recognized: Plionarctos Frick, 1926 and Arctodus Leidy, 1854 in North America, Arctotherium Burmeister, 1879 in South America and Tremarctos Gervais, 1855 with an extinct species from North America and a living one in South America (Soibelzon, 2004).

According to the latest reviews (Soibelzon, 2004; Soibelzon et al., 2005) the South American extinct bears have been assigned to the genera Arctotherium, being recorded in Venezuela, Brazil, Bolivia, Uruguay, Argentina and Chile from the Ensenadan (Upper Pliocene to Middle Pleistocene) to the Lujanian (Upper Pleistocene-Early Holocene). Five species are recognized within this genera: Arctotherium vetustum Ameghino, 1885 from Bonaerian levels (Middle Pleistocene) at the Argentine provinces of Buenos Aires and Entre Ríos and one possible finding from Brazil; Arctotherium wingei Ameghino, 1902 from Brazilian and Venezuelan Lujanian levels and Pleistocene deposits in Bolivia; Arctotherium bonariense (Gervais 1848–1852) from Bonaerian and Lujanian levels of the Buenos Aires province (Argentina); Arctotherium angustidens Gervais & Ameghino, 1880 from Ensenadan levels at the same province and at Tarija, Bolivia; and finally Arctotherium tarijense Ameghino, 1902 from Bonaerian and Lujanian levels at this Bolivian locality, Argentine provinces of Santa Fe and Buenos Aires, Uruguay and at the XII Region of Magallanes, Chile (Soibelzon, 2004).

In Chile, the record of Ursidae is restricted to central and southern Patagonia. It is composed by a molar from the Pilauco site (40° S) assigned to Ursidae (Pino et al., 2013). Another finding corresponds to a femur from Cueva del Milodón (51°S), originally assigned to Arctotherium sp. and later on reassigned to Pararctotherium pamparun by Oliver-Schneider (1935). At Cueva de Los Chingues (52° S) an I² was recorded. Originally assigned to Pararctotherium sp. by Prevosti et al. (2003), it was later identified as Arctotherium tarijense by Soibelzon (2004) and has recently been reassigned to Arctotherium sp. by Prevosti & Martín (2013). Finally, there must be mentioned another femur fragment collected from the surface of Cueva del Puma (52° S) and identified as Arctotherium tarijense with a date of 10.345±75 B.P. (Martín et al., 2004, Martín, 2013). Thus, the evidence now reported from Baño Nuevo-1 fills the intermediate gap between the areas from where it was known before.

The Arctotherium sp. remain from Layer 5 at Baño Nuevo-1 cave is associated to Mylodontinae, Felidae (aff. Panthera onca mesembrina), Equidae, Lama guanicoe and Macrauchenia sp. remains with no traces of human presence (López & Mena, 2011). Paleoenvironmental studies based on pollen records from neighboring areas (High Cisnes River, De Porras et al., 2012; High Simpson river, Markgraf et al., 2007; Middle Chacabuco river, Villa-Martínez et al., 2011) reveal an open steppe environment.

One of the questions triggered by the finding of Arctotherium at Baño Nuevo-1 is related to the taphonomy of the fossil assemblage at the cave. Studies about the role of Arctotherium as a potential agent on the formation and transformation of bone assemblages are almost nonexistent, unlike the Old World Ursidae situation, whose taphonomic effects have been subject of intensive research (Gargett, 1996; D'Errico et al., 1998; Stiner et al., 1998; Wolverton, 2001; Quilès et al., 2006; Rabal-Garcés & Cuenca-Bescós, 2009; Arilla et al., 2014). One of the few American studies, specifically geared to the case of Arctotherium angustidens, reveals a high proportion of broken teeth, as likely result of chewing on hard materials such as bones (Figueirido & Soibelzon, 2010). On the other hand, studies by Haynes (1983) have revealed that bears (Ursus arctos and Ursus americanus in particular) leave traces like those made by rodents, that is short and parallel with punctures related to parallel grooves on the crest of long bones. Andrews and Fernández-Jalvo (1997) found punctures of up to 10.4 mm in diameter, attributed to the action of bears. These marks were recorded both on the shaft and articular surfaces, just like those findings done by D'Errico and collaborators (1998). The only puncture recorded on remains from Layer 5 at Baño Nuevo-1 was found on a Lama sp. patella and both its shape and size are consistent with those produced by canids.

However, with this finding, Arctotherium joins the Felidae (aff. Panthera onca mesembrina) and Dusicyon avus previously known from the site as potential agents in the formation of the bone assemblages from the layers of the Final Pleistocene at the site (Trejo & Jackson, 1998). As said before, the wear patterns and dental pathologies observed in different samples indicate an omnivorous diet with non-negligible meat consumption, including bones (Prevosti & Vizcaíno, 2006; Figueirido & Soibelzon, 2010; Martín, 2013). In fact, most of the known Arctotherium angustidens remains have broken teeth's, most likely due to chewing on hard bones (Soibelzon et al., 2009; Figueirido & Soibelzon, 2010).

In the occlusal surface of the I₃ of Baño Nuevo-1, part of the cusp is fragmented and it presents a caries that could be interpreted as a result of eating carbohydrate-rich foods, such as fruit or honey (Ferigolo, 1992; Soibelzon & Prevosti, 2007; Soibelzon et al., 2014). Though a microwear analysis is not applicable to South American giant short-faced bear, because in species with omnivore diet the interpretation of hard plant versus bone consumption is problematic (Soibelzon et al., 2014:1241), we performed a SEM observation to see microwear that could reveal chewing use after cusp breakage. We looked the labial and lingual faces of the occlusal surface, finding striations and pits that point to a chewing action on the lingual face, while no such traits were observed on the labial face (Fig. 4A–B). Such evidences points to a postdepositional fracture of a portion of the cusp, or an immediate loss of it together with the complete incisive from its alveolar cavity.

Given the lack of direct evidence of damage due to bear action, the origin of Arctotherium at Baño Nuevo-1 has to be questioned. Soibelzon and collaborators (2009) point to two possible scenarios for the presence of bears in caves: the use for hibernating, and the sporadic search of carrion inside the caves. Hibernation has been mentioned in other works (Martín, 2013:28), but it is not clear if it can be applied to Tremarctinae. Accordingly, the sporadic use of Baño Nuevo-1 cave for both food and shelter purposes, is the most plausible scenario, taking into account the number and sort of evidence recovered, along with no presence of taphonomic traces attributable to Arctotherium. In any case, this very low density of Arctotherium material is consistent with the record from other sites in Patagonia, such as Milodón, Los Chinges or El Puma caves, as it is the almost null evidence of taphonomic traces left by this taxon (Martín, 2008:364).

The record of Arctotherium at Baño Nuevo-1 raises several questions that must be dealt with in the future, such as the competition with other animals to occupy the cave or its role in the formation and transformation of the deposits.