taxonID	type	description	language	source
6F447E05005A1F7CFCECFCCAFBB1FB5C.taxon	description	(FIGS 2 – 12; SUPPORTING INFORMATION, APPENDIX S 1)	en	Boscaini, Alberto, Gaudin, Timothy J, Mamani Quispe, Bernardino, Münch, Philippe, Antoine, Pierre-Olivier, Pujos, François (2019): New well-preserved craniodental remains of Simomylodon uccasamamensis (Xenarthra: Mylodontidae) from the Pliocene of the Bolivian Altiplano: phylogenetic, chronostratigraphic and palaeobiogeographical implications. Zoological Journal of the Linnean Society 185 (2): 459-486, DOI: 10.1093/zoolinnean/zly075, URL: https://academic.oup.com/zoolinnean/article/185/2/459/5187717
6F447E05005A1F7CFCECFCCAFBB1FB5C.taxon	description	The species Pleurolestodon dalenzae was erected by Saint-André et al. (2010) only a few pages before S. uccasamamensis. However, the ‘ page priority criterion’ has no formal standing in the International Code of Zoological Nomenclature (ICZN), and these species have to be considered equally old (see ICZN, 1999). In these cases, the order of priority is determined by the first reviser [i. e. the first author (s) to consider their synonymy; ICZN, 1999, article 24.2]. As first revisers, we therefore establish that Pleurolestodon dalenzae is a junior synonym of S. uccasamamensis, on the basis of the more abundant and well-determined material of the latter species in comparison with the former (Saint-André et al., 2010). Holotype MNHN-Bol V 11731 (ex GB 078; Fig. 2), anterior part of cranium without dentition (Saint-André et al., 2010). Measurements See the Supporting Information (Appendix S 5). Revised stratigraphic and geographical occurrence Latest Miocene – earliest Pliocene of Choquecota, early Pliocene of Pomata-Ayte, Casira and Inchasi (Oruro and Potosí Departments, Bolivia) and late Pliocene of Ayo Ayo – Viscachani (La Paz Department, Bolivia). Revised diagnosis Fossil sloth smaller in size than Glossotheridium chapadmalense, Paramylodon garbanii and Pleurolestodon acutidens, and roughly biometrically similar to Glossotheriopsis pascuali; long zygomatic processes of squamosal; wide braincase in relationship to the total cranial length and wide V-shaped palate; in ventral view, the medial palatal process of the maxilla is more extended mediolaterally than anteroposteriorly and the occipital condyles are well separated from the condyloid foramina, as in the Miocene species Pleurolestodon acutidens; the foramen magnum shows a detached notch located on its dorsal border, similar to that observed in Pleurolestodon and Mylodon; long and slender ascending process of the jugal, which strongly resembles that of Glossotheridium chapadmalense; the diastema between Cf 1 and Mf 1 is absent or extremely reduced and, in lateral view, Cf 1 presents an almost vertical wear facet similar to the condition in Pleurolestodon acutidens; Mf 2 and Mf 3 possess a marked lingual sulcus, comparable to Pleurolestodon acutidens and Paramylodon garbanii; cf 1 is bevelled, with well-developed mesial and distal wear surfaces; mf 3 is marked by a deep apicobasal sulcus on its lingual side, absent on the labial side and not covered by the ascending ramus of the mandible in lateral view (like Glossotheridium chapadmalense and unlike Pleurolestodon acutidens and Paramylodon garbanii); the symphyseal spout is anteriorly flat as in Glossotheridium chapadmalense and not rounded as in Pleurolestodon acutidens and Paramylodon garbanii. Paratype MNHN-Bol V 3321 (Fig. 3), maxillary and premaxillary fragments with left Mf 1 – Mf 3 and right Mf 2 – Mf 4 (Saint-André et al., 2010). Referred specimens See the Supporting Information (Appendix S 1). Description Cranium: The cranium of S. uccasamamensis appears elongated, with the cranial roof and cranial base roughly horizontal in lateral view. The nasal region is slightly depressed in relationship to the braincase. In dorsal view, MNHN-Bol V 3348 (Fig. 4 A), 3711 (Fig. 5 A) and 3726 (Fig. 6 A) are more slender than MNHN-Bol V 3717 (Fig. 7 A) and 3718 (Fig. 8 A) (see measurements in Supporting Information, Appendix S 5). The snout is elevated and widened anteriorly and relatively short; the braincase is wide in relationship to the total cranial length (Figs 5 A, 7 A, 8 A). In dorsal view, the nasals are narrow at the level of the antorbital constriction, and gradually broaden anteriorly and posteriorly (Figs 2 A, 4 A, 5 A, 6 A, 7 A, 8 A). The nasals are probably wider in MNHN-Bol V 3718 (Fig. 8 A) than in the other specimens as a result of diagenetic dorsoventral compression of this skull. A V-shaped, posteriorly pointed nasofrontal suture is present in all specimens examined. The anterior border of the nasals is strongly convex in MNHN-Bol V 3711 and 3726 (Figs 5 A, 6 A) and straighter in MNHN-Bol V 11731, 3717 and 3718 (Figs 2 A, 7 A, 8 A); in MNHN-Bol V 3348 (Fig. 4 A), the nasals are strongly convex but with a narrow, anteriorly projecting median process. The temporal fossa and the temporal lines are not clearly observable in all the specimens, owing to breakage and / or diagenetic deformation (Figs 6, 7). In general, the temporal lines appear straight and roughly parallel to one another, diverging anteriorly as they approach the postorbital processes of the frontal bones. The temporal lines curve laterally and ventrally at their posterior limit, extending parallel and slightly anterior to the nuchal crest (Figs 4, 5, 8). In lateral view, the anterior part of the cranium is dominated by the maxilla. This bone contacts the nasal and the frontal dorsally, the lacrimal in its middle part and the palatine and the alisphenoid posteriorly. The premaxillae are preserved in six specimens. They are V-shaped in MNHN-Bol V 3348, 3711 and 3726 (Figs 4 C, 5 C, 6 C), whereas they show a more arched profile (and even flatten anteriorly as they approach their midline junction) in MNHN-Bol V 11731, 3717 and 3718 (Figs 2 C, 7 C, 8 C). Their medial and lateral rami are nearly equivalent in size, with the lateral ramus slightly longer than the medial one, in MNHNBol V 3348, 3711 and 3726 (Figs 4 C, 5 C, 6 C), whereas the lateral ramus is substantially longer than the medial ramus in MNHN-Bol V 11731, 3717 and 3718 (Figs 2 C, 7 C, 8 C). The palate is rugose and strongly widened anteriorly, especially in MNHN-Bol V, 11731, 3321, 3717 and 3718 (Figs 2 C, 3 B, 7 C, 8 C); whereas it is narrower in MNHN-Bol V 3348, 3711 and 3726 (Figs 4 C, 5 C, 6 C; see also ‘ Morphometric analyses’ subsection below). In lateral view, the palate is concave at the level of Cf 1 – Mf 1 and convex posteriorly at the level of Mf 2 – Mf 4. The anterior palatal foramina (sensu De Iuliis et al., 2011) are always clearly delineated (Figs 2 C, 3 B, 4 C, 5 C, 6 C, 7 C, 8 C) and continue into distinct anterior grooves that extend medially to connect with the incisive foramina (not visible in ventral view). Also, enlarged postpalatal foramina (sensu Gaudin, 2011) are observable in all specimens (Figs 2 C, 3 B, 4 C, 5 C, 6 C, 7 C, 8 C). The sphenorbital fissure and the optic and sphenopalatine foramina are visible in MNHN-Bol V 3711, 3717 and 3718 (Figs 5, 7, 8). These foramina open into a common depression and are approximately aligned horizontally. The sphenorbital fissure is the posteriormost and largest of the three foramina mentioned above. The optic foramen is adjacent to the sphenorbital fissure and the sphenopalatine foramen is farther anterior, at the anterior margin of the common depression. Posteriorly and ventrally, the foramen ovale is located between the squamosal and the lateral plate of the pterygoid and opens onto the lateral wall of the cranium. In ventral view, the structures of the middle region of the basicranium are difficult to observe, owing to poor preservation and complete fusion of the sutures. The pterygoids are inflated at their base (Fig. 5). The descending laminae of the pterygoid are broad and deep, but only preserved fully in MNHN-Bol V 3711 and 3717 (Figs 5, 7). In lateral view, the lacrimal is more elongated anteroposteriorly than dorsoventrally and is pierced by a rounded lacrimal foramen. The orbital portion of the bone is larger than its facial portion. The jugal is firmly attached to the lacrimal and possesses ascending, descending and middle processes (Figs 5 E, 6 B, 7 B, 8 B), as is typical for sloths (Gaudin, 2004). The ascending process is the longest of the three. It is wide at its base, becoming narrower posteriorly, and ends as a rounded tip located at the level of the anteroposterior midpoint of the zygomatic process of the squamosal. The ascending and middle processes are strongly divergent in MNHN-Bol V 3717 and 3718 (Figs 7 B, 8 B) and more convergent in MNHN-Bol V 3711 and 3726 (Figs 5 E, 6 B). The ascending process of the jugal is marked by a weak postorbital process near its base (Figs 5 E, 6 B, 7 B, 8 B). The middle process of the jugal is triangular in shape and closely approaches the zygomatic process of the squamosal posteriorly. The descending process of the jugal bears a posteriorly or posteromedially extended hook in MHNH-BOL V 3717 and 3718 (Figs 7, 8) that is not present in MNHN-Bol V 3711 and 3726 (Figs 5, 6). The long zygomatic process of the squamosal is almost horizontal in lateral view and anterolaterally directed in dorsal view (Figs 4 – 8). In lateral view, the occipital is inclined anteriorly in MNHN-Bol V 3711 (Fig. 5 B), whereas it appears more vertical in MNHN-Bol V 3718 (Fig. 8 B). In posterior view (Fig. 5 D), its transverse breadth exceeds its dorsoventral height. It is marked by a slight median external occipital crest that does not continue dorsally into a sagittal crest and culminates ventrally in a well-marked notch on the posterior edge of the foramen magnum (Fig. 5 D). The nuchal crests are strongly developed and clearly visible in dorsal, lateral and posterior views (Figs 5, 8). The occipital condyles, in ventral view, are roughly triangular in shape, with a slightly concave medial edge, and are slightly more elongated mediolaterally than anteroposteriorly (Fig. 8 C). In lateral view, they appear more prominent in MNHN-Bol V 3711 (Fig. 5 B) than in MNHN-Bol V 3718 (Fig. 8 B), a difference that is probably related to the inclination of the occiput already discussed. Upper dentition: The upper tooth rows of S. uccasamamensis, composed of five teeth, are divergent anteriorly. The most mesial tooth is caniniform, whereas the remaining four are molariform (Figs 4 – 8). There is no diastema between Cf 1 and Mf 1 (Figs 2 – 8). Cf 1 is roughly semicircular in cross-section in most specimens: MNHN-Bol V 3711 and 3726 (Figs 5, 6), a shape that is also observable in the alveoli of MNHN-Bol V 11731 (Fig. 2). The straight side, corresponding ideally to the diameter of this semicircle, faces linguomesially, whereas the arched outline faces labiodistally. Variations of this shape are observable in the almost triangular caniniform of MNHN-Bol V 3717 (Fig. 7) and the nearly circular Cf 1 (or correspondent alveoli) of MNHN-Bol V 3718 and 3321 (Figs 3, 8). In all the specimens where it is preserved, the occlusal surface is almost vertical in lateral view and directed lingually and distally in occlusal view. Mf 1 is ovate in cross-section, elongated along the main axis of the tooth row. This tooth bears a lingual apicobasal sulcus in MNHN-Bol V 3718 (Fig. 8), absent in MNHN-Bol V 11371, 3321, 3711 and 3726 (Figs 2, 3, 5, 6). In MNHN-Bol V 3717 (Fig. 7), the lingual sulcus of Mf 1 is slightly marked on the left and absent on the right, indicative of the great variability of this character. Mf 1 bears a bevelled occlusal surface, with a mesial wear facet that is larger than the distal facet. Mf 2 and Mf 3 are bilobate and exhibit a deep lingual apicobasal sulcus (Figs 3 – 8). In occlusal view, these two teeth are roughly triangular in cross-section, with the orthogonal angle disposed mesiolingually (Figs 3 – 8). Mf 2 and Mf 3 vary in their occlusal outlines, but Mf 2 is generally longer mesiodistally than transversely, whereas in Mf 3 the transverse width is equal to or exceeds the mesiodistal length. The wear facet of Mf 2 is more pronounced distally than mesially, whereas in Mf 3 it is more pronounced in the central part of the tooth than in its labial and lingual extremities (e. g. MNHN-Bol V 3711 and 3717; Figs 5, 7). Mf 4 is T-shaped in occlusal view, with the distal lobe clearly narrower transversely than the mesial one. The last upper tooth presents both lingual and labial longitudinal sulci, the latter more pronounced than the former (e. g. MNHN-Bol V 11731, 3321, 3711 and 3717; Figs 2, 3, 5, 7). Mandible and lower dentition: The lower caniniform, like the upper one, is also generally semicircular in cross-section, with some exceptions represented by the triangular shape of MNHN-Bol V 3296 (Fig. 9 A) and the ovate shape in MNHN-Bol V 3711 and 3371 (Figs 5 F, 10 D). It appears as the highest tooth of the lower tooth row and is nearly equal in size to mf 1 (Figs 5 – 7, 9 – 12). The cf 1 is bevelled, with the mesial wear facet broader than the distal one. It also presents a slight lingual apicobasal sulcus in MNHN-Bol V 3726 (Fig. 6), 3296 (Fig. 9) and the juvenile mandibles (Fig. 12), a feature not observable in the other specimens (Figs 5, 7, 10, 11). A lingual apicobasal sulcus is a consistent feature of mf 1. This tooth is transversely wider mesially than distally and bears an oblique, distally inclined wear facet (Figs 5 – 7, 9 – 12). The irregular cross-section of mf 2 resembles a parallelogram (Figs 5 – 7, 9 – 12). The presence of longitudinal sulci on this tooth is variable among the observed specimens. In MNHN-Bol V 3296 (Fig. 9), sulci are present on all four sides, whereas in MNHN-Bol V 3371 (Fig. 10 D) they are almost absent. These two morphologies represent the extremes of the observable variation for mf 2. In occlusal view, mf 3 is strongly bilobate, with the mesial lobe wider transversely than the distal lobe (Figs 5 – 7, 9, 10, 12). The mesial lobe is extended mesiolabially, with an apicobasal sulcus that faces mesiolingually. The distal lobe of mf 3 is rounded distally. The two lobes are separated by a thin isthmus accompanied lingually by a deep, broad apicobasal sulcus, absent on the labial side (Figs 5 – 7, 9, 10, 12). The mandible of S. uccasamamensis is short and deep, with the ventral border of the horizontal ramus nearly horizontal in lateral view (Figs 5 – 7, 9, 10, 12). The tooth row is aligned in occlusal view, with the exception of the cf 1, which is slightly displaced laterally. The dorsoventral depth of the horizontal ramus of the mandible is constant along the tooth row, becoming narrower towards the symphyseal spout and deepening posteriorly at the base of the ascending ramus. The profile of the symphyseal spout in lateral view is irregular, with a strong convexity flanked by marked dorsal and ventral concavities (Figs 5 – 7, 9 – 11). In occlusal view, the symphysis is wider distally than proximally, with a visible constriction anterior to the caniniforms. The mandibular foramen is located on the medial side of the mandible, well posterior and slightly ventral to the base of mf 3. The posteroexternal opening of the mandibular canal (a characteristic feature of sloths among xenarthrans; see Gaudin, 2004; De Iuliis et al., 2011) faces laterally at the level of the posterior edge of the root of mf 3, well ventral to the dorsal edge of the horizontal ramus. The mandibular canal emerges anteriorly through the mental foramina situated on the anterolateral surface of the symphyseal spout. The mental foramina are highly variable in size and number (e. g. a single one in MNHN-Bol V 3296 and six in MNHN-Bol V 3298; Figs 9, 11 A – C), even in the two dentaries of a single individual (e. g. MNHN-Bol V 3717, 3371). On the ascending ramus, the angular, condyloid and coronoid processes are equally divergent (Fig. 9). The angular process is deeply concave medially, with a strongly convex ventral edge that is clearly demarcated from the ventral edge of the horizontal ramus (Fig. 9). The transverse width of the mandibular condyle is much greater than its anteroposterior length. It has a hooked aspect in dorsal view, and its lateral portion is oriented horizontally in posterior view, whereas the medial portion is downturned ventrally. The coronoid process is tall, with its posterior edge orthogonal to the main mandibular axis in lateral view (Figs 6 E, 9 B, 12 G). It is somewhat hooked posteriorly in MNHN-Bol V 3726 and 3359 (Figs 6 E, 12 G), but not in MNHN-Bol V 3296 (Fig. 9 B). However, a complete coronoid process belonging to an adult individual is lacking in the present sample and therefore this feature cannot be assessed properly. No important differences have been detected among the mandibular features of the specimens of S. uccasamamensis. In general, the mandibles MNHN-Bol V 3717, 3358 and 3298 (Figs 7, 10 A – C, 11 A – C) are more robust than the specimens MNHN-Bol V 3711, 3726, 3296, 3371 and 12518 (Figs 5, 6, 9, 10 D – F, 11 D – F). This greater robustness is exemplified by the two specimens MNHN-Bol V 3358 and 3371, illustrated in Fig. 10, with the former (Fig. 10 A – C) showing a more robust dentition and more anteriorly divergent, thicker and deeper horizontal rami than the latter (Fig. 10 D – F). These variations were treated by Anaya & MacFadden (1995) as possible indicators of different ontogenetic stages and / or the existence of sexual dimorphism.	en	Boscaini, Alberto, Gaudin, Timothy J, Mamani Quispe, Bernardino, Münch, Philippe, Antoine, Pierre-Olivier, Pujos, François (2019): New well-preserved craniodental remains of Simomylodon uccasamamensis (Xenarthra: Mylodontidae) from the Pliocene of the Bolivian Altiplano: phylogenetic, chronostratigraphic and palaeobiogeographical implications. Zoological Journal of the Linnean Society 185 (2): 459-486, DOI: 10.1093/zoolinnean/zly075, URL: https://academic.oup.com/zoolinnean/article/185/2/459/5187717
6F447E05005A1F7CFCECFCCAFBB1FB5C.taxon	description	Cranium and upper dentition: In S. uccasamamensis, the presence of a high braincase, a deep and anteriorly elevated snout, and an approximately horizontal cranial base in lateral view are typical features of Mylodontidae (Gaudin, 2004). The width of the braincase relative to total skull length is comparable to Pleurolestodon acutidens (FMNH P 14495; Rovereto, 1914) and is greater than Glossotheridium chapadmalense, Glossotherium robustum, Mylodon darwinii or Paramylodon harlani (Owen, 1842; Kraglievich, 1925, 1934; Stock, 1925; McAfee, 2009; Brandoni et al., 2010). Other features, above all in the rostral region, allow assignment of Simomylodon to the Mylodontinae. In particular, the snout is relatively short, widened anteriorly in dorsal view and depressed in lateral view (Figs 2, 4 – 8). The wide rostrum of S. uccasamamensis is also accompanied by an anterior enlargement of the nasals. The nasal becomes narrower posteriorly and widens again at the level of the nasofrontal suture (Figs 2, 4 – 8). This expansion is a recurrent feature in mylodontids, as is the great enlargement of the external nares (Gaudin, 2004). The characters observed on the rostrum are closely related to the morphology of the palate. Indeed, S. uccasamamensis shows a V-shaped palate (Figs 2 – 8), comparable with all the mylodontines and lestodontines (Gaudin, 2004) except Mylodon, which has secondarily lost Cf 1 s, and consequently, reduced anterior palatal width (Kraglievich, 1934; Brandoni et al., 2010). The medial anterior palatal processes of the maxilla are projected farther anteriorly than the lateral ones (Figs 2 – 8). This characteristic is also observed in Glossotheridium, Glossotherium, Paramylodon, Pleurolestodon and Mylodon (Owen, 1842; Rovereto, 1914; Kraglievich, 1925, 1934; Stock, 1925; McAfee, 2009; Brandoni et al., 2010). The extension of the medial palatal processes of the maxilla is moderate anteroposteriorly, but the processes are broad mediolaterally (Figs 2 – 8). In this respect, S. uccasamamensis resembles more Pleurolestodon acutidens (FMNH P 14495) than Glossotheridium chapadmalense (Kraglievich, 1925). On the lateral cranial wall, the lacrimal is wide, with its orbital portion larger than its facial portion. The lacrimal is pierced by a small lacrimal foramen (Figs 4 B, 7 B), the diminutive size of this opening being a feature of all mylodontines and lestodontines (Gaudin, 2004). All members of Mylodontidae are characterized by complex jugals with distinct ascending, descending and middle processes. The middle process is elongated and triangular, and the descending process is hooked posteriorly (Gaudin, 2004). All these features have been observed in the specimens attributed to Simomylodon (Figs 5 E, 6 B, 7 B, 8 B). Moreover, this taxon presents a weak postorbital process of the zygomatic arch, as in Glossotheridium, Paramylodon, Pleurolestodon and the lestodontines (Kraglievich, 1925; Stock, 1925; Webb, 1989; Gaudin, 2004; McAfee, 2009). This is situated at the base of the ascending process, which is long and slender (Figs 5 E, 6 B, 7 B, 8 B), resembling that of Glossotheridium, Glossotherium, Paramylodon and Mylodon (Kraglievich, 1925, 1934; Stock, 1925; McAfee, 2009; Brandoni et al., 2010). In contrast, Pleurolestodon, Lestodon and Thinobadistes have shorter and more robust ascending processes of the jugal (Rovereto, 1914; Webb, 1989; Bargo et al., 2006). Finally, the ascending process of S. uccasamamensis is nearly horizontal, as in most mylodontids (Figs 5 E, 6 B, 7 B, 8 B; Gaudin, 2004). The length of the zygomatic process of the squamosal is peculiar in Simomylodon because it is the longest (relative to the total cranial length) ever observed among mylodontids. Its almost horizontal orientation and its broad and flattened tip (Figs 4 – 8), however, are common features of mylodontines and lestodontines (Gaudin, 2004). In dorsal view, the frontals and parietals are anteroposteriorly and mediolaterally flattened and the sagittal crest is absent (Figs 4 – 8). This morphology is present in all mylodontines and lestodontines except Lestodon (Gaudin, 2004). The presence of a flat temporal fossa, delimited by non-connecting temporal lines, is shared with Pleurolestodon, Glossotherium, Paramylodon and Mylodon (Owen, 1842; Rovereto, 1914; Stock, 1925; Kraglievich, 1934; McAfee, 2009; Brandoni et al., 2010). The strongly developed nuchal crest of constant width and aligned with the posterior surface of the occiput (this latter showing a median external occipital crest connecting the nuchal crest to the dorsal edge of the foramen magnum) are features of Simomylodon shared with all mylodontids (Gaudin, 2004). Simomylodon also exhibits a detached notch on the dorsal border of the foramen magnum, comparable with that observed in Pleurolestodon (FMNH P 14495) and Mylodon (Kraglievich, 1934). In posterior view, the foramen magnum is limited laterally by roughly triangular occipital condyles. These lie at the level of the dentition in lateral view (Figs 5 B, 8 B). Both are common conditions in Mylodontidae (Gaudin, 2004). As in all mylodontines and lestodontines, the occipital condyles extend posteriorly to the posteriormost edge of the foramen magnum in ventral view (Figs 5 C, 8 C). The condyles are widely separated from one another, in this respect resembling Glossotherium, Pleurolestodon and Thinobadistes; also, they are mediolaterally elongated in ventral view to an extent that is comparable with Mylodon, Pleurolestodon and the lestodontines Thinobadistes and Lestodon (Owen, 1842; Rovereto, 1914; Stock, 1925; Kraglievich, 1934; Webb, 1989; McAfee, 2009; Brandoni et al., 2010). In ventral view, the occipital condyles are well separated from the condyloid foramina (Figs 5 C, 8 C), as in Pleurolestodon (FMNH P 14495). The upper dentition of S. uccasamamensis is similar to the other representatives of Mylodontidae, with five teeth on each side aligned in two divergent tooth rows (Figs 2 – 8; Gaudin, 2004). Exceptions to this pattern are represented by Scelidotheriinae, which have parallel tooth rows (McDonald, 1987), and Octomylodon and Mylodon, which exhibit anterior tooth loss (Scillato-Yané, 1977; Brandoni et al., 2010). The Cf 1 – Mf 1 diastema is absent or extremely reduced (Figs 5 C, 8 C), as in most mylodontids. Exceptions to this mylodontid pattern are represented by Mylodon and Octomylodon (Scillato-Yané, 1977; Brandoni et al., 2010), in which Cf 1 is secondarily lost, and Lestodon, which shows a derived and extremely elongated diastema (Czerwonogora & Fariña, 2013).	en	Boscaini, Alberto, Gaudin, Timothy J, Mamani Quispe, Bernardino, Münch, Philippe, Antoine, Pierre-Olivier, Pujos, François (2019): New well-preserved craniodental remains of Simomylodon uccasamamensis (Xenarthra: Mylodontidae) from the Pliocene of the Bolivian Altiplano: phylogenetic, chronostratigraphic and palaeobiogeographical implications. Zoological Journal of the Linnean Society 185 (2): 459-486, DOI: 10.1093/zoolinnean/zly075, URL: https://academic.oup.com/zoolinnean/article/185/2/459/5187717
6F447E05005A1F7CFCECFCCAFBB1FB5C.taxon	description	Mandible and lower dentition: In occlusal view, cf 1 is equivalent in size to mf 1 (Figs 5 – 7, 9 – 12), and mf 3 is the largest lower tooth, an invariant trait of Mylodontidae also known to occur in Bradypodidae (Gaudin, 2004). The cf 1 is roughly semicircular in cross-section in most specimens, exceptionally displaying ovate or triangular cross-sections (Figs 5 – 7, 9 – 12). The latter two conformations are typical among Mylodontinae and Lestodontini, respectively (Gaudin, 2004). The caniniform of Simomylodon has a bevelled occlusal surface (Figs 5 – 7, 9, 10, 12), resembling that of Pleurolestodon acutidens (Rovereto, 1914) and Paramylodon garbanii (Robertson, 1976). The cf 1 is bevelled also in Glossotheridium chapadmalense, but both wear facets are well developed in Simomylodon and Pleurolestodon, whereas the distal facet is extremely reduced in Glossotheridium chapadmalense (Kraglievich, 1925). Whereas cf 1 of Paramylodon garbanii projects strongly mesially and labially (Robertson, 1976), that of S. uccasamamensis and Glossotheridium chapadmalense is implanted vertically (Kraglievich, 1925). The irregularly lobate mf 1 and mf 2 (Figs 5 – 7, 9 – 12) are very similar to those of Pleurolestodon acutidens, Glossotheridium chapadmalense and Paramylodon garbanii, but in Simomylodon they are significantly smaller in size. Moreover, Paramylodon garbanii shows deeper apicobasal sulci on the lingual and distal sides of mf 2 (Robertson, 1976), whereas Glossotheridium chapadmalense has a more elongated and almost straight-walled mf 2 (Kraglievich, 1925). The elongated and bilobate mf 3 is a recurrent feature in Mylodontinae and Lestodontini (Gaudin, 2004). As already noted, mf 3 has asymmetrically developed lingual and labial apicobasal sulci in S. uccasamamensis, with the former markedly deeper than the latter (Figs 5 – 7, 9, 10, 12). Among late Miocene – Pliocene Mylodontinae, the pattern of S. uccasamamensis is very similar to that of Glossotheridium chapadmalense, but differs from that of Pleurolestodon acutidens (FMNH P 14495, 14521), in which both labial and lingual apicobasal sulci are well developed, and Paramylodon garbanii (UF 10922), which displays an extra bulge on the labial side of mf 3. The mandible of Simomylodon presents some typical mylodontid features, such as the straight horizontal ventral edge of the horizontal ramus in lateral view, and a condyle located at the same level as the tooth row (Figs 5 – 7, 9, 10, 12) (Gaudin, 2004; Saint-André et al., 2010). In general, the mandible of Simomylodon is smaller than that of all the other late Miocene – Pliocene mylodontids (i. e. Pleurolestodon acutidens, Paramylodon garbanii and Glossotheridium chapadmalense; Rovereto, 1914; Kraglievich, 1925; Robertson, 1976). The articular condyle of Simomylodon is convex and medially hooked in dorsal view, as in many Mylodontidae, in contrast to the condyle of lestodontines, which is extended both laterally and medially (Gaudin, 2004). The angular process is the posteriormost process of the mandible (Fig. 9), another common mylodontid trait. Among Mylodontidae, only Nematherium and Octomylodon show an equally posterior extension of the condyloid process (Gaudin, 2004). In medial view, the mandible of Simomylodon shows a detached oblique ridge that extends from the anteroventral edge of the angular process towards the root of the last tooth (Fig. 9 D). This last condition is shared with Pleurolestodon, Glossotherium, Paramylodon (Gaudin, 2004) and specimen MACN Pv 8675 of Glossotheridium chapadmalense. In lateral view, the ascending ramus does not cover mf 3 (Figs 6 E, 9 B, 10 E, 12 D, G), resembling the condition observed in Glossotheridium chapadmalense and Paramylodon garbanii (Kraglievich, 1925; Robertson, 1976). This partial coverage of mf 3 is observed in some other mylodontid genera, such as Octodontotherium, Pseudoprepotherium, Mylodon and Pleurolestodon (Gaudin, 2004). The anterior edge of the mandibular spout is broad and flat in occlusal view, as it is in Glossotherium, Glossotheridium and Lestodon (Owen, 1842; Kraglievich, 1925; McAfee, 2009; Czerwonogora & Fariña, 2013) (Figs 5 – 7, 9, 10). Other mylodontine genera, such as Mylodon, Paramylodon and Pleurolestodon, possess anteriorly rounded mandibular spouts (Stock, 1925; Kraglievich, 1934; McAfee, 2009; Brandoni et al., 2010). Four mandibles can be ascribed to juvenile individuals of S. uccasamamensis, based on their reduced size and lack of wear on the lower dentition (Fig. 12). These remains already display the main diagnostic features that have been found in the adults (e. g. a straight ventral margin, the absence of a diastema between cf 1 and mf 1, and the extreme reduction of the labial apicobasal sulcus on mf 3; Fig. 12). All these features are in strong contrast to those observed in the juvenile mandibular fragment described by Oliva & Brandoni (2012) and tentatively attributed to a ‘ mylodontid cf. Simomylodon ’ (Huayquerian SALMA, Buenos Aires Province, Argentina). The present data suggest that the specimen described by Oliva & Brandoni (2012) does not belong to Simomylodon, but rather is more compatible with the genus Pleurolestodon, also recognized in the Huayquerian SALMA of Argentina (Rovereto, 1914). However, given that no juvenile mandibular remains are known for that taxon, we prefer to consider the latter specimen as Mylodontinae indet. MORPHOMETRIC ANALYSES The results of PCAs among the Neogene Mylodontinae are depicted in Figures 13 and 14, showing the two distinct modules: cranium and upper dentition (Fig. 13) vs. mandible and lower dentition (Fig. 14). We followed this approach in order to overcome the problem of the paucity of the data, thus maximizing the number of specimens that could be included in the analyses (for further details, see Material and Methods and Supporting Information, Appendix S 5). In the cranial dataset (Fig. 13), principal component (PC) 1 explains 51.80 % of the variance and, given that all the variables have positive loadings, it reflects body size. Size is lower on the left side and higher on the right. Principal component 2 (Fig. 13 A) explains 17.42 % of the variance. Positive values on this axis reflect skulls that have slender palates and thin rostra relative to total skull lengths, whereas negative values represent skulls with relatively wider palates and rostra. Finally, PC 3 (Fig. 13 B) explains 12.77 % of the total variance. Positive values for PC 3 are associated with robust dentitions and a long and deep snout relative to total skull lengths, whereas negative values are correlated with a reduced dental series and a short and slender snout (in relationship to total length). The Miocene – Pliocene sloths are well segregated along PC 1 (Fig. 13), with S. uccasamamensis and Glossotheriopsis pascuali as the smallest taxa, and Glossotheridium chapadmalense as the largest taxon in the dataset. Pleurolestodon acutidens and Paramylodon garbanii occupy intermediate positions (Fig. 13). On PC 2, the extreme morphologies are represented by Glossotheriopsis pascuali in the positive range and Paramylodon garbanii in the negative range (Fig. 13 A). However, these morphologies must be treated cautiously, because the result may be affected by the lack of total skull length measurements for both species, given that neither is represented by complete skulls. Simomylodon shows important variation along PC 2, whereas Pleurolestodon acutidens and Glossotheridium chapadmalense do not overlap (Fig. 13 A). This means that Glossotheridium chapadmalense exhibits a wider palate and rostrum relative to total skull length than is the case for Pleurolestodon acutidens. On PC 3, S. uccasamamensis still shows high variation, including Glossotheriopsis pascuali from southern Argentina in its morphometric range (Fig. 13 B). The most extreme morphologies are represented by Pleurolestodon acutidens (the highest values) and Glossotheridium chapadmalense (the lowest values). These two taxa, together with Paramylodon garbanii, partly overlap along PC 3 (Fig. 13 B). In Fig. 13, the S. uccasamamensis specimen MNHNBol V 3348 (Fig. 3) is represented by a red triangle. This cranium was previously attributed to the species Pleurolestodon dalenzae by Saint-André et al. (2010). The present dataset shows that this specimen falls far outside the morphometric range of the genus Pleurolestodon, but well within the range of variation for Simomylodon (Fig. 13). Likewise, a second PCA (Fig. 14) was performed on the variables of the mandible and lower dentition, yielding the same general pattern as the cranial analysis. Principal component 1, which explains 57.96 % of the total variance, is again a representation of size, and Simomylodon occupies the lowest positions on the left side of the graph (Fig. 14). Glossotheridium chapadmalense shows the largest mandibular values, whereas Pleurolestodon acutidens and Paramylodon garbanii are recovered in intermediate positions. Principal component 2 explains 19.42 % of the variance. Higher values are correlated with a long dental series and a deep mandibular ramus at the level of the dentition, relative to total mandibular length, whereas lower values correspond to a shorter dental series and less robust mandible. Finally, PC 3 explains 9.45 % of the variance and reflects mandibles with a long horizontal ramus and anteroposteriorly narrow ascending ramus (positive values) vs. mandibles displaying a shorter horizontal ramus and a more anteroposteriorly enlarged ascending ramus (negative values). On both PC 1 and PC 2 (Fig. 14 A), Simomylodon shows the greatest range of variation, probably attributable to the inclusion of juvenile individuals in the dataset (Fig. 12). These specimens are retrieved in the far bottom-left portions of the graph depicting PC 1 vs. PC 2 (Fig. 14 A), corresponding to the lowest values for both principal components. This means that they are the smallest specimens in the dataset (as expected), but they also possess a dental series that is reduced in length and horizontal rami that are of moderate depth relative to total mandibular length. Glossotheridium chapadmalense shows the highest variation on PC 3 (Fig. 14 B), an effect that is probably related to the incompleteness of the dataset for this taxon (no complete mandibles are known). As before, the red triangles indicate the S. uccasamamensis specimens that were formerly assigned to another taxon. These are the specimens MNHN-Bol V 3358 (Fig. 10 A – C), 3371 (Fig. 10 D – F) and 3359 (Fig. 12 G – I) that Anaya & MacFadden (1995) assigned to Glossotheridium chapadmalense. The more extensive data of the present analysis support their inclusion in the genus Simomylodon instead. PHYLOGENETIC ANALYSIS The phylogenetic analysis recovered a single most parsimonious tree (tree length: 755 steps, consistency index = 0.662, retention index = 0.927), with a topology compatible to that of the consensus tree from the analysis by Gaudin (2004). In our dataset, Simomylodon is deeply nested within Mylodontinae, more precisely as the sister taxon of the monospecific genus Pleurolestodon (Fig. 15). The node uniting the latter two taxa is well supported, with bootstrap and jackknife values of 53 and 70, respectively. These values are even greater than those supporting Mylodontinae (Supporting Information, Appendix S 7). However, and in accordance with the previous study of Gaudin (2004), other groups are better supported, such as Lestodontini, Mylodontidae, Scelidotheriinae and Folivora (Supporting Information, Appendix S 7). The unambiguous synapomorphies that link Simomylodon and Pleurolestodon include: the Cf 1 placed at the edge of the premaxilla (Gaudin, 2004: character 21, 1 → 0), the relatively wide braincase (Gaudin, 2004: character 82, 2 → 3) and the pronounced separation of the occipital condyles from the hypoglossal foramina (Gaudin, 2004: character 194, 1 → 2). The close morphological affinity of Simomylodon and Pleurolestodon, suggested by the present phylogenetic analysis, is probably the cause of the taxonomic misunderstanding of Saint-André et al. (2010), who assigned a gracile specimen of Simomylodon to a new species of Pleurolestodon. Autapomophies of S. uccasamamensis, as retrieved by the present phylogenetic analyses are as follows: an intermediate shape of the coronoid process (Gaudin, 2004: character 47, 2 → 1), a posterodorsal inclination of the mandibular condyloid process in lateral view (Gaudin, 2004: character 52, 1 → 0), a short mandibular symphysis (Gaudin, 2004: character 62, 2 → 1), a weak buccinator fossa of the maxilla (Gaudin, 2004: character 106, 0 → 1), an ascending process of the jugal longer than the descending process (Gaudin, 2004: character 151, 0 → 1), a very elongate zygomatic process of the squamosal (Gaudin, 2004: character 168, 2 → 3), an enlarged condyloid foramen (Gaudin, 2004: character 187, 1 → 2) and a narrow and fairly deep mastoid depression (Gaudin, 1995: character 34, 1 → 0).	en	Boscaini, Alberto, Gaudin, Timothy J, Mamani Quispe, Bernardino, Münch, Philippe, Antoine, Pierre-Olivier, Pujos, François (2019): New well-preserved craniodental remains of Simomylodon uccasamamensis (Xenarthra: Mylodontidae) from the Pliocene of the Bolivian Altiplano: phylogenetic, chronostratigraphic and palaeobiogeographical implications. Zoological Journal of the Linnean Society 185 (2): 459-486, DOI: 10.1093/zoolinnean/zly075, URL: https://academic.oup.com/zoolinnean/article/185/2/459/5187717
