Systematic revision of the species of Protypotherium (Notoungulata: Interatheriidae) from the Santa Cruz Formation (Early-Middle Miocene), Argentinian Patagonia: a new phylogenetic hypothesis for the Interatheriidae Author Fernández, Mercedes División Paleontología de Vertebrados, Museo Argentino de Ciencias Naturales ‘ Bernardino Rivadavia’, Avenida Ángel Gallardo 470, C 1405 DJR Ciudad Autónoma de Buenos Aires, Argentina & Laboratorio Anatomía y Evolución de los Vertebrados, Departamento de Ciencias Básicas, Universidad Nacional de Luján, Ruta 5 y Avenida Constitución 6700 Luján, Buenos Aires, Argentina & Consejo Nacional de Investigaciones Científicas y Técnicas (CONICET), Juan Domingo Perón 2158, C 1040 AAH, Buenos Aires, Argentina mechisfernandezpaleo@gmail.com Author Fernicola, Juan C. División Paleontología de Vertebrados, Museo Argentino de Ciencias Naturales ‘ Bernardino Rivadavia’, Avenida Ángel Gallardo 470, C 1405 DJR Ciudad Autónoma de Buenos Aires, Argentina & Laboratorio Anatomía y Evolución de los Vertebrados, Departamento de Ciencias Básicas, Universidad Nacional de Luján, Ruta 5 y Avenida Constitución 6700 Luján, Buenos Aires, Argentina & Consejo Nacional de Investigaciones Científicas y Técnicas (CONICET), Juan Domingo Perón 2158, C 1040 AAH, Buenos Aires, Argentina Author Cerdeño, Esperanza Consejo Nacional de Investigaciones Científicas y Técnicas (CONICET), Juan Domingo Perón 2158, C 1040 AAH, Buenos Aires, Argentina & Paleobiología y Paleoecología, Instituto de Nivología, Glaciología y Ciencias Ambientales, Centro Científico Tecnológico-CONICET-Mendoza, Avenida Ruiz Leal s / n, M 5500 Mendoza, Argentina text Zoological Journal of the Linnean Society 2023 2023-07-11 199 2 417 444 http://dx.doi.org/10.1093/zoolinnean/zlad043 journal article 272446 10.1093/zoolinnean/zlad043 81098796-8acf-4360-a625-ad76a281e99e 0024-4082 8426263 Protypotherium praerutilum Ameghino, 1887b ( Figs 2A–C, F, H, K, M , 3A, C–E , 4D–F , 5–7 , 8A ; Supporting Information, Table S2) 1887b Protypotherium praerutilum Ameghino , p. 15 (original description); Ameghino 1889 , pl. 14, figs 6–7, 8, 12–13. 1909 Patriarchus leptocephalus Ameghino, 1891b ; Sinclair, p. 40. 1909 Protypotherium convexidens Ameghino, 1891b ; Sinclair, p. 40. Protypotherium attenuatum Ameghino, 1887b ; regarded as synonym herein. Patriarchus rectus Ameghino, 1891b ; regarded as synonym herein. Patriarchus diastematus Ameghino, 1891b ; regarded as synonym herein. Patriarchus icochiloides Ameghino, 1894 ; regarded as synonym herein. Protypotherium diversidens Ameghino, 1891b ; regarded as synonym herein. Eudiastatus lingulatus Ameghino, 1891a ; regarded as synonym herein. Protypotherium lineare Ameghino, 1894 ; regarded as synonym herein. Protypotherium globosum Ameghino, 1891b ; regarded as synonym herein. Icochilus endiadys Roth, 1899 ; regarded as synonym herein. Holotype : MACN-A 1081 ( Fig. 4D ), incomplete skull with less I1–M3 and right I1–2 (alveoli)–I3–C–dP1(alveoli)–P2– M3, and MACN-A 1082 ( Fig. 3E ), incomplete mandible with complete dentition (see Fernández et al. 2018 ), Early–Middle Miocene , SCF, Santa Cruz Province ( Argentina ). Referred materials: Supporting Information, Table S1 . Extended diagnosis: Protypotherium praerutilum differs from Pr. australe by exhibiting a smaller size (~20%) and a proportionally shorter I1 compared to I2–3; from Pr. antiquum by the talonid of m3 with a shallower (and even absent) labial groove, and its mesial lobe with a concave distal region; from Pr. colloncurensis and Pr. columnifer by the talonid of p2–4 sub-triangular in outline; from Pr. colloncurensis by a descending process of the maxilla with a moderate development; from Pr. claudum by having p2–3 with a differentiated entoflexid, p4 with a transversely deeper, narrower and more distally placed ectoflexid, and m1–3 with deeper entoflexids; from Pr. compressidens by exhibiting wider cheek teeth, particularly P2–M3; from Pr. minutum and Pr. distinctum by less cement covering each tooth, the P2–4 with an entoflexus located in a more distal position and slightly less expanded, and M1–3 with straighter ectolophs; and from Pr. columnifer by the talonid of p4 narrower than the trigonid. Figure 8. Protypotherium praerutilum , in occlusal view. A, MACN-A 4044, incomplete palate. Protypotherium claudum , in occlusal view. B, MACN-A 551, right mandibular fragment with p2 (alveolus)–p3–m2; (C) MLP 73-VII-6-4 d, less mandibular fragment with p2 (broken)–m2 (trigonid); (D) MPEF-PV 1394, incomplete mandible with less i1–dp1–p2–p4, and right i1 and dp1–p2–m3. Protypotherium compressidens , in occlusal view. E, MACN-A 4029, less maxillary fragment with P3–M3; (F) MACN-A 4030, less mandibular fragment with p4 (talonid)–m3; (G) MACN-A 3867, right mandibular fragment with dp1–2–p3–4 (trigonid); (H) MACN-A 9716–9732?g, right mandibular fragment with m1 (talonid)–3; (I) MPM-PV 19761, right maxillary fragment with M1–2; (J) MPM-PV 19781, right mandibular fragment with p2–m1; (K) MPM-PV 19782, less maxillary fragment with P2–M3. Abbreviation: ’, indicates broken tooth. Scale bar = 10 mm. Geographic and stratigraphic provenance: Santa Cruz Province ( Argentina ), SCF, Early – Middle Miocene , Santacrucian SALMA ; Río Negro and Neuquén provinces ( Argentina ), Collón Curá Fm. , Middle Miocene , Colloncuran SALMA . Measurements: Supporting Information, Table S2. Remarks: Regarding the upper permanent dentition, Pr. praerutilum exhibits I1 (x L = 4.4 mm ; see Supporting Information, Table S2) with a greater mesio-distal length than that of I2–3 (I2, x L = 4.0 mm; I3, x L = 3.7 mm ; see Supporting Information, Table S2), but the differences between these lengths are less than those in Pr. australe (I1, x L = 5.8 mm ; I2, x L ≈ 4.7 mm ; I3, x L ≈ 4.8 mm ; see Supporting Information, Table S2) (see Fig. 7 ). Tauber (1996) mentioned the high degree of imbrication between I1–3 as a diagnostic characteristic of Pr. praerutilum , but we do not consider it diagnostic due to the following reasons. First, this is a variable characteristic within the analysed sample; in fact, some asymmetries concerning the degree of imbrication when comparing the less and right premaxilla have been identified in some specimens of this species (e.g. MACN-A 3920, AMNH 9565, MACN-A 9620; Fig. 4E–F ), as well as in Pr. australe (e.g. MACN-A 9623–9633 and MACN-A 9644; Figs 2E , 4G ). Second, the holotype of the species ( MACN-A 1081–1082; Figs 3E , 4D ; see Fernández et al. 2018 ) does not show this sort of imbrication; in fact, its degree of imbrication is low. Several species are here considered, for the first time, as junior synonyms of Pr. praerutilum . Traditionally, Pr. attenuatum and Pr. praerutilum were considered separate taxa. According to Sinclair’s (1909) proposal, supported by Tauber (1996) , Pr. attenuatum differs from Pr. praerutilum mainly by its smaller size. However, according to the measurements provided by Sinclair (1909), the three specimens of Pr. praerutilum (e.g. L M1 = 6.0– 6.6 mm ; W M1 = 4.5 mm ; L m1 = 5.7 mm ; W m1 = 3.0 mm) and two of Pr. attenuatum (e.g. L M1 = 6.0– 6.5 mm ; W M1 = 4.0– 4.6 mm ; L m1 = 5.5 mm ; W m1 = 3.0 mm) completely overlap. In order to resolve this issue, we analysed the dental dimensions of M1–2 and m1–2 of the Santacrucian specimens of Protypotherium (note that Pr. claudum was excluded due to the existence of morphological differences that allow the species characterization; see below). When comparing M1–2 ( Figs 5A , 6A ) and m1–2 ( Figs 5B , 6B ) dimensions, the scaưer diagrams show three different groups ( Figs 5 , 6 ), particularly more evident in M/m1, that are named asser the valid name of the species: (1) Morphotype A or Pr. australe , which includes the holotype of the species and the referred materials, including the type specimens of the considered junior synonyms of Pr. australe (see below for further justifications); (2) Morphotype B or Pr. praerutilum , which contains the holotype of the species, the syntypes of Pr. attenuatum , and the referred materials of Pr. praerutilum , including the type specimens of the junior synonyms (see below for further justifications); and (3) Morphotype C or Pr. compressidens that only includes the holotype and a few referred specimens herein considered. It is worth mentioning that in the case of the third species, the grouping regarding m1 is not as conspicuous as in the other two groups, due to the number of specimens assignable to Pr. compressidens (a minimum of three is required to establish the grouping). Nevertheless, the values do not fall within the other two groups ( Fig. 5B ), showing a differentiation in terms of length and width of m1. In the case of m2, this differentiation is less evident, because only the holotype does not fall within the other two groups, leaving the remaining specimens within the size range of Pr. praerutilum ( Fig. 6B ). Excluding Pr. compressidens from the analysis, due to the presence of morphological differences with Pr. australe and Pr. praerutilum (see below), Morphotypes A and B were statistically tested in order to prove if they were statistically different in terms of the dental dimensions. All the statistical details are shown in Supporting Information, Appendix S4, and summarized as follows. In general terms, the requirement of normal distribution was not met for the length and/or width of M/m1–2 of Pr. australe and Pr. praerutilum (Shapiro–Wilk’s test, P <.01) and all datasets show a non-homogeneous variance (Levene’s test, P <.05). Given the non-compliance of both necessary requirements to carry out a Student’s t -test, a non-parametric study was used, in this case a Mann–Whitney U Test, in order to compare the tooth dimensions between the two groups. All samples analysed with a Mann-Whitney U test (α = 5%) showed a significant difference between dental dimensions (M/m1–2) for Morphotypes A and B (U = 0, P <.05). Therefore, Pr. australe and Pr. praerutilum statistically differ based on the length and width of M/m1–2. P. In this framework, Pr.australe involves large individuals (L M1 ≥ 7.1 mm ; W M1 ≥ 4.15 mm ; L m1 ≥ 7.15 mm ; W m1 ≥ 3.5 mm ) ( Figs 5 , 6 ), whereas Pr. praerutilum represents small to medium individuals (L M1 ≤ 6.8 mm ; W M1 ≤ 5.0 mm; L m1 ≤ 6.8 mm ; W m1 ≤ 4.0 mm) ( Figs 5 , 6 ). Our analysis suggests no significant difference in the size of M/m1–2 between Pr. praerutilum and Pr. attenuatum . Besides size, another character traditionally used to differentiate Pr. praerutilum and Pr. attenuatum is the curvature of the cheek tooth series, but when comparing the curvature of the holotypes of both species, independent from size and taking into consideration that Pr. attenuatum involves a fragmented maxilla, this feature does not differ. We argue that the morphological differences between Pr. praerutilum and Pr. attenuatum can be explained by ontogenetic and/or intraspecific variation, and because our statistical analyses demonstrate that size is not a diagnostic feature in order to separate both species, we consider them synonyms. As the name priority is determined by the reviewer who takes notice of this and not by the page priority ( CINZ 2000), we select Pr. praerutilum as the senior synonym because its holotype is more complete (see Fernández et al. 2018 ). In contrast to Sinclair (1909), who proposed that Pr. lineare was the junior synonym of Pr. australe based on its size (followed by Tauber 1996 ), we consider the former as the junior synonym of Pr. praerutilum , because the size of its holotype ( MACN-A 4038–4039; see Fernández et al. 2018 ) falls within the size range of Pr. praerutilum (Supporting Information, Table S2). Icochilus endiadys , from the Collon Curá Fm. ( Neuquén Province ; Roth 1899 ), was recently included in Protypotherium by Vera et al. (2017) . These authors mentioned the presence of a reduced first upper premolar that was sub-circular in cross section as a diagnostic feature of Protypotherium endiadys and stated that this condition is seen in the holotype of the species MLP 12-2886 ( Fig. 2H ). Although the less dP1 (P1 according to Vera et al. 2017 ) of this holotype is broken at the base, it does show a sub-circular outline. However, the base of right dP1, albeit very badly preserved, appears sub-elliptical. There is possibly an internal asymmetry regarding the morphology of the alveolus of dP1, but the only completely preserved base of dP 1 in this specimen is sub-circular. We consider this assessment indeterminate and that the morphology of this tooth is not diagnostic for Pr. endiadys . Vera et al. (2017) also mentioned the reduction of dP1 and dp1 as diagnostic features of the species, but according to the measurements provided by these authors, these teeth do not differ from those observed in the other species of Protypotherium , considered here as valid, particularly Pr. australe and Pr. praerutilum , which have the largest samples (see Supporting Information, Table S2). Therefore, we consider this character indeterminate too. As mentioned above, the M3 of Pr. praerutilum presents a distally expanded metaloph whose length varies from poorly to highly developed, which is one of the characteristics mentioned by Vera et al. (2017) to re-diagnose Pr. endiadys ; however, due to its variability, we do not support this character as diagnostic. Therefore, in the absence of any other characteristics that differentiate Pr. endiadys from Pr. praerutilum , the former is proposed here as a junior synonym of the laưer. Finally, Patriarchus diastematus was treated as Typotheria incertae sedis by Sinclair (1909). Originally, it was differentiated from the remaining species of Patriarchus by the presence of a sub-cylindrical first upper premolar and a diastema between this tooth and P2 ( Ameghino 1891b ). However, the maxilla of the holotype (MACN-A 4044; Fig. 8A ; see Fernández et al. 2018 ) shows a sub-elliptic alveolus of dP1, as do the majority of the alveoli of dP1 of Protypotherium (but they can also be sub-circular). Additionally, the ‘diastema’ between dP1 and P2 is in fact a small space because dP1 is slightly obliquely implanted and lays on P2, as observed in almost all specimens assigned to Protypotherium , including the holotype of Pr. praerutilum (MACN-A 1081; Fig. 4D ; see Fernández et al. 2018 ). Therefore, both characteristics are discarded as diagnostic; as there is no other feature that can contrast Pa. diastematus from Pr. praerutilum , the former is considered a junior synonym of the laưer.