Trachytherus spegazzinianus

Billet, Guillaume, Muizon, Christian De & Quispe, Bernardino Mamani, 2008, Late Oligocene mesotheriids (Mammalia, Notoungulata) from Salla and Lacayani (Bolivia): implications for basal mesotheriid phylogeny and distribution, Zoological Journal of the Linnean Society 152 (1), pp. 153-200 : 175-178

TRACHYTHERUS SPEGAZZINIANUS

The holotype described and figured by Ameghino (1889a, b) ( Fig. 16A ) was supposedly lost ( Reguero, 1999; A. Kramarz, pers. comm.) but, in fact, the specimen is still present in the MACN collection. The confusion arose because Ameghino represented the holotype of Trachytherus spegazzinianus as a complete rostrum with right and left tooth rows. In its previous original description of this species ( Ameghino, 1889a), he also mentioned a nearly complete palate. Remarkably, analysis of the figure shows that the two sides are fully symmetric in all their parts and especially in the numerous enamel folds on the lingual edge of the mesial lobe of M1 and in the breakage of M3. MACN 235 is identical to the right half of the palate represented by Ameghino. Undoubtedly, this is the specimen on which Ameghino’s description is based ( Simpson & Minoprio, 1949, also figured this specimen). It is clear that Ameghino reconstituted the rostrum by reproduction of the left side symmetrical to the right side. However, specimen MLP 52-XI-2-1 probably represents the left part of an originally complete fossil palate, which has been sectioned in a sagittal plane. It is probable that MLP 52-XI-2-1 was part of a complete palate with MACN A235, which also presents traces of section in the corresponding region. Ameghino might have sectioned the palate that was originally in the MLP collection, and brought with him the right part in the MACN collection (A. Kramarz & M. Bond, pers. comm.). Later, he described and figured the palate of T. spegazzinianus from a symmetrical reproduction of MACN A235.

Podestá (1899) created a new genus and species Ameghinotherium (now Trachytherus ) curuzucuatiense from a partial skull collected at Curuzú Cuatiá in the Corrientes Province, Argentina. Bond et al. (1998) identified a mandible fragment from the same locality, which they refer to the same species. These authors considered the presence of a conspicuous lingual notch on the i2 of this mandible as a major difference with T. spegazzinianus . Reguero (1993, 1999) considered T. curuzucuatiense as congeneric with Trachytherus and maintained specific distinction with other Trachytherus species (i.e. T. spegazzinianus and T? mendocensis) on the basis of more hypsodont molars. The holotype of this species (MACN 6867) is an incomplete skull without anterior part and with abraded ventral surface providing peculiar cheek teeth sections ( Fig. 17A ). As a consequence, it is very difficult to distinguish much detail on this specimen. Nevertheless, it exhibits a strong labio-lingual width on molars and a deep lingual sulcus on M3, which resembles the configuration observed in Trachytherus spegazzinianus , and which markedly differs from that described in T. alloxus . However, the size of the specimen, too large to correspond to either T? mendocensis or T. subandinus , does not provide any evidence to distinguish it from T. spegazzinianus or T. alloxus . Trachytherus curuzucuatiense is here tentatively regarded as a junior synonym of T. spegazzinianus because the difference in hypsodonty evocated by Reguero (1993) has not been confirmed by our observations. Furthermore, the lingual notch mentioned by Bond et al. (1998) in T. curuzucuatiense is also present in T. spegazzinianus especially on unworn (or slightly worn) incisors (e.g. the specimen mentioned below as MACN A52-490). This notch (or groove) is also present in T. alloxus .

Reguero & Castro (2004) erected a new genus and species Anatrachytherus soriai for a single partial mandible MACN A52-490 collected in Cabeza Blanca, Chubut Province, Argentina, previously referred by Ameghino (1897) to Trachytherus spegazzinianus . Characters on which Anatrachytherus soriai is based are: shortened mandibular body, i1–2 less procumbent, more complete inferior dental formula with i/2, c/1, p/4, m/3,* short diastema between i2 and canine, and m3 shorter than m2. [*Note that no canines have been observed by us on this specimen.] Given the ontogenetic patterns in T. alloxus and taking into account that MACN A52-490 is restored in some parts, these characters do not support a taxonomic distinction. The mandible is that of a young adult with erupting p2 and m3, and very little worn i2, p3–4 ( Fig. 17B, C ). Thus, the erupting condition of m3 La Plata ( Argentina) very probably also belong to this species. They do not differ from Trachytherus spegazzinianus except, possibly, in their size. Actually, the most complete La Plata specimens are included in the inferior part of the size variation range observed here for Trachytherus spegazzinianus , and therefore are also in the superior part of the Trachytherus alloxus size variation range. All the specimens here assigned to T. spegazzinianus have been either studied by photo or directly observed. It is highly probable that unseen specimens, such as the trachytheriines of the AMNH collections, should, at least in part, be referred to this species.

The Bolivian locality of Lacayani has yielded remains of basal mesotheriids, which have been referred to Trachytherus spegazzinianus ( Hoffstetter et al., 1971; Marshall & Sempéré, 1991; Reguero & Castro, 2004). However, these fossils have never been studied in detail. Furthermore, recent field work by one of us (BMQ) has provided a significant number of new specimens. These specimens have been considered in the light of the present redefinition of T. spegazzinianus and are confidently assigned to this species. They are of the same size, and bear parallelogram-like upper premolars, strong lingual enamel infoldings on the protoloph of M1, a clearly trilobate M3, rounded lobes and deep lingual sulcus on upper molars ( Fig. 18 ), a combination of characters currently unknown in other species of Trachytherus . results in a short length compared with m2 but, with increasing wear, this would not have been the case. This is confirmed from a lateral view showing the increasing length of m3 from the top to the base of that tooth ( Fig. 17C ). Furthermore, the more complete dental formula is likely to be also linked to the young age of this specimen, which thus retains teeth that are absent in older individuals. Finally, the restoration of this specimen, especially in the inferior part of the symphyseal region, precludes definitive conclusions concerning the size of the mandibular body or the procumbence of the incisors. For these reasons, specimen MACN A52-490 is referred here to a young adult of Trachytherus spegazzinianus .

The numerous specimens we have from El Pajarito, Chubut province, Argentina, housed in the Museo de

BASAL MESOTHERIID SYSTEMATICS

The present study of a new species of Trachytherus and the further study and definition of T. spega z zinianus provide interesting new data and characters relevant to the systematics of basal mesotheriids, which, in fact, correspond to the trachytheriines. A phylogenetic analysis has been carried out on this group on eight taxa and 18 characters (one multistate, ordered) to attempt to generate new hypotheses regarding their relationships (see character list in Appendix 5; see data matrix in Appendix 6). The polarization of the characters is based on the outgroup comparison criterion. The outgroup used are the non-mesotheriid typotherians Oldfieldthomasia and Archaeohyrax ( Notopithecus could have been also used as outgroup with the same results). The analysis of the data matrix was performed using PAUP (version 4.0b10, Swofford, 1998) with the exact ‘Branch and Bound’ algorithm.

The parsimony analysis resulted in eight most parsimonious trees of 21 steps (CI = 0.905, RI = 0.923). The strict consensus (26 steps, CI = 0.731, RI = 0.731, Fig. 19A ) consists of an unresolved polytomy for basal mesotheriids and mesotheriines mainly because of the problematic taxa T. subandinus and T? mendocensis (see Discussion below on T? mendocensis). Many characters are unknown for these species, and the known characters provide contradictory information (see below).

The majority-rule consensus (21 steps, CI = 0.905, RI = 0.923, Fig. 19B ) groups Trachytherus spegazzinianus with mesotheriines, indicating that it occurs in 75% of the most equally parsimonious cladograms. This most frequent grouping is based on the fact that T. spegazzinianus presents all derived features shared by at least a basal mesotheriid and mesotheriines: upper premolars parallelogram-shaped (also in T. alloxus ), lingual extremity of the protoloph rounded when getting close to contacting the metaloph (also in T? mendocensis and T. subandinus ), deep and highly persistent lingual sulcus on upper molars (unknown in T? mendocensis and T. subandinus ), M3 trilobed and trapezoidal with lingually convergent mesial and distal faces (also in T? mendocensis, unknown in T. subandinus ), and large size (also in T. alloxus ). The Adams consensus (23 steps, CI = 0.826, RI = 0.846, Fig. 19C ) also places T. spegazzinianus in the same position as this information is contained in all the intersecting set of taxa common to the most parsimonious cladograms.

The majority-rule consensus places T. alloxus as the sister taxa of all other mesotheriids (in 63% of cases). The presence of a rounded lingual extremity of the protoloph when close to contact with the metaloph (char. 10) and the presence of a well-trilobed M3, trapezoidal with lingually convergent mesial and distal faces (char. 12) in T? mendocensis, T. subandinus (char. 12 unknown), T. spegazzinianus and mesotheriines, and the absence of these in T. alloxus are arguments in favour of the basal placement of T. alloxus . However, the great size (char. 13) and the parallelogram-shaped premolars (char. 9) shared by T. alloxus , T. spegazzinianus and mesotheriines, to the exclusion of T? mendocensis and T. subandinus , argue to the contrary.

The contradictory information provided by characters 9, 10, 12 and 13 implies homoplasy and this varies per the cladograms considered. More needs to be known about these characters to understand their variation better and to guage their pertinence.

In spite of this poorly informative result, we have established a list of key synapomorphies of the mesotheriids on the basis of the cladistics results and the observations made in this paper and from Croft et al. (2004), Flynn et al. (2005) and Reguero & Castro (2004). This list has been used to diagnose the Mesotheriidae (see above). Key synapomorpies include characters 1–8. Characters 10 and 12 are also synapomorphies of the mesotheriids according to the strict consensus (these characters are reversed in T. alloxus ) but not to the majority-rule consensus. Except characters 4 and 5 (labial fossettes disappearing long before isolation of the central fossette on upper molars; trilobed upper molars when little worn, with a rounded median lobe), all others characters are not observable in T? mendocensis and T. subandinus . Thus, future analyses might not place some of these apomorphies for Mesotheriidae as a whole but for less inclusive groups. Additional material for T? mendocensis and T. subandinus is needed.

Previous analyses of mesotheriids did not include all basal mesotheriids. Croft et al. (2004) and Flynn et al. (2005) essentially focused on the systematics of mesotheriines rather than that of basal mesotheriids. The cladistic analysis of Reguero & Castro (2004), of basal mesotheriids, did not include Trachytherus subandinus and the new species from Salla, but considered Anatrachytherus soriai and T. curuzucuatiense , which are regarded here as junior synonyms of T. spegazzinianus . Their analysis also championed the paraphyly of the genus Trachytherus . However, their results differ from ours in two aspects: they concluded that T? mendocensis was the most basal mesotheriid and that the sister taxon of mesotheriines was a clade formed by T. spegazzinianus and T. curuzucuatiense . Our study also differs from that of Reguero & Castro (2004), in the introduction of new characters in the analysis.

Our analysis suggests that the genus Trachytherus may be paraphyletic. If this is the case, Trachytherus should be restricted to the type species ( T. spegazzinianus ) and new genera should be created for the three other species ( T. alloxus , T. subandinus and T? mendocensis). Although this solution would be preferable from a strict phylogenetic point of view, our current knowledge is too weak to ensure that no special relationship actually exists between any representatives of basal mesotheriids. Thus, we suggest not splitting the systematics of the basal mesotheriids (i.e. the genus Trachytherus ) for the time being, until additional material is available. For the sake of convenience, the genus Trachytherus (including four species) is still used herein, although we strongly suggest that it may be paraphyletic. It is to be hoped that future studies, if additional material is known, will seek for this.

Examination of juvenile specimens of Trachytherus alloxus provides new relevant data for phylogenetic reconstructions. The typical typotherian face pattern of fossettes on unworn upper molars and deciduous premolars emphasizes the phylogenetic position of mesotheriids nested among typotherians. A distal cingulum is also present on upper deciduous premolars and molars of T. alloxus . This study also confirms, as hypothesized by Patterson (1934b), that in mesotheriids the median lobe is formed by the union of a long crochet, running mesially from the metaloph to the second crista. The unworn deciduous lower premolars and molars show that trigonid and talonid were originally separated. Their first connection is lingual and it leads to a deep labial sulcus at later wear stages. This sulcus is permanent and characteristic of mesotheriines.

Figures

Figure 16. Upper dentition of Trachytherus spegazzinianus. A, holotype, MACN A235, partially restored with right and left I1, P2–M3 (after Ameghino, 1889b). B, partial palate, with left I1, P2–M2, MLP 52-XI-2-1. C, upper right (reversed) dP3–4, M1, MACN A52-566. Scale bar = 1 cm.

Figure 17. A, holotype of T. curuzucuatiense (Podestá, 1889) = T. spegazzinianus, MACN 6867. B and C, holotype of Anatrachytherus soriai Reguero & Castro, 2004 = T. spegazzinianus, MACN A52-490; B, occlusal view; C, lateral view. Scale bar = 1 cm.

Figure 18. Trachytherus spegazzinianus from Lacayani. A, left M1, LAC 14. B, right (reversed) M2–3, MNHN-BOL-V 010725. C, right dentary with p3–4, m1 (or p4, m1–2), LAC 19. D, left maxillary with dP3–4, LAC 4. Scale bar = 1 cm.

Figure 19. Phylogenetic relationship between basal mesotheriids. A, strict consensus (26 steps, CI = 0.731, RI = 0.731; Bremer indices: Mesotheriidae = 2, Mesotheriinae = 1). B, majority-rule consensus, with percentages of nodes with less than 100% indicated (21 steps, CI = 0.905, RI = 0.923). C, Adams consensus (23 steps, CI = 0.826, RI = 0.846).