The Snout and Anterior Dentition of Early Miocene Anthracotheres (Artiodactyla, Mammalia) from Europe: Implications for Taxonomy, Systematics and Phylogeny

The literature on the Early Miocene, Old World anthracotheres currently classified in the genus Brachyodus, reveals that diverse opinions have been published concerning the interpretation of their anterior teeth, both deciduous and permanent. The European fossil record encompassing the osteology of the snout, and the odontology of the incisors and canines in situ in jaws of these large artiodactyls is poor, with descriptions of only four or five mandibular symphyses that contain teeth and only three premaxillae, of which only one specimen of each contains all the teeth, at least on one side. Furthermore, three of the described mandibles were of juveniles with some unerupted teeth, rendering it difficult to determine whether the partly preserved empty alveoli originally housed deciduous or permanent teeth. In contrast, there are many isolated anterior teeth in museums and private collections. There is also the question of sexual dimorphism which has been difficult to address due to the scarcity of teeth in situ in the jaws, although males and females of some taxa were likely to have been dimorphic and bimodal in some tooth positions (I1/, i/2) but were not so clearly different in other elements of the dentition, especially the molars and premolars.

The present article focuses on several undescribed anthracothere mandibular symphyses and premaxillae from France, Egypt, Kenya and Uganda that throw some light on the matter and it includes a reinterpretation of previously described specimens. However, because so few incisors and canines have been found in situ in jaws, it does not resolve all the issues, especially those concerning the deciduous dentition, because no specimens have been recovered with the anterior milk teeth in situ. It is reasonably clear though, that some early forms previously attributed to Brachyodus possessed three permanent incisors and a canine in the lower jaws, whereas later forms of the genus generally lacked the canines and often the lower central incisors (either suppressing their development entirely or shedding them during ontogeny). Reduction of the quantity of teeth in the incisor row is also likely to have occurred in the upper jaws, with two of the available premaxillae showing the presence of three incisor alveoli, whereas two other specimens show only two teeth in situ.

The sample of mandibular symphyses of large anthracotheres from the Early Miocene of Moghara, Egypt, fall into five morphological categories (Pickford & Abdel Gawad, 2024), but the sample from the French Faluns and the Tagus Basin, Portugal, comprise only two main forms, one of which is fused early in ontogeny and is elongated, referred to Brachyodus, the other shorter and remaining unfused into the sub-adult phase of ontogeny and even when fused in adults, continues to show a suture between the two mandibles. The former group, attributed to Brachyodus, exhibits a marked degree of dimorphism in the tusk-like second lower incisors, with females having more gracile teeth than males. The latter group exhibits weaker dimorphism in the i/2s and is attributed to a new genus, Chitenaymeryx.

A major difference between Chitenaymeryx and Brachyodus is the presence of lower canines in the former genus, but their absence in the latter one. There are also important differences in dental morphology in these two genera.

A revision of the large anthracotheres from Moghara, Egypt, was recently undertaken because studies have revealed that some of the material previously identified as Brachyodus depereti Fourtau, 1918, belongs to Jaggermeryx africanus (Andrews, 1899) (ex naida, Miller et al. 2014), as does most of the material previously attributed to Brachyodus africanus. Other specimens belong to Aegyptomeryx Pickford & Abdel Gawad (2024) the biggest anthracotheres known from the deposits. The species Afromeryx grex Miller et al. 2014, is a junior synonym of Mogharameryx mogharensis (Pickford, 1991). The species Jaggermeryx palustris Miller et al. 2014 was transferred to a new genus Masrimeryx Pickford & Abdel Gawad (2024). There are thus five species of large anthracotheres at Moghara, which are accompanied by four small species: Nov. gen. plus Sivameryx moneyi (Fourtau, 1918,) Afromeryx zelteni Pickford, 1991, and Sivameryx africanus (Andrews, 1914). Some details are provided in Pickford & Abdel Gawad (2024).

This monograph attempts to answer questions regarding the significance of the differences observed in the anterior dentitions of anthracotheres, focussing on the European material. A major conclusion is that there are at least four taxa of anthracotheres in the French Faluns and the Sables de l’Orléanais, three species of which are attributed to Brachyodus, and one to Chitenaymeryx nov. gen. Two new species of Brachyodus are diagnosed.

This monograph looks briefly into the systematic relationships between subgroups of anthracotheres and it also delves into the relationship between anthracotheres and hippopotamids.

Key words: Anthracotheriidae, Europe, Africa, Early Miocene, dentition, snout, taxonomy, systematics.

Introduction 8
Materials and methods  9
Institution Abbreviations 10
Abbreviations of tooth position and measurements  10
Dental nomenclature 10
Pre-Darwinian context 13
Post-Darwinian studies 13

Systematic Palaeontology
Genus Brachyodus Depéret, 1895  35
Species Brachyodus nancrayensis nov.  35
Species Brachyodus onoideus (Gervais 1859)  42
Species Brachyodus pontigneensis nov.  54
Subfamily Bothriodontinae Scott, 1940  56
Genus Chitenaymeryx nov.  56
Species Chitenaymeryx intermedius (Mayet, 1908)  56
Isolated anterior teeth of anthracotheres from the Early Miocene of France  58
Tusk-like teeth of Brachyodus  58

Descriptions and morphometric analyses of incisiform anthracothere teeth  59
Upper second incisors  76
Upper third incisors  88
Lower central incisors  93
Lower second incisors  112
Lower third incisors  143
Canines  150
Lower first premolars  158
Tooth of doubtful meristic position  163

General discussion 164
Diversity  166
Distribution  167
Palaeoecology  167
Implications of the revision of the Moghara anthracotheres  167
Comparisons with East African Early Miocene anthracotheres  167
Phylogeny and classification  169
Possible precursors of Brachyodus 169
Possible precursors of Chitenaymeryx  173
Relationships of Brachyodus and other anthracotheres to Hippopotamidae  173

Conclusions  175

Acknowledgements 175

References  176

Reports on, and interpretations of, the dental formula of the early Miocene, Old World anthracothere genus, Brachyodus, have varied from author to author (Depéret 1895; Roman, 1907; Mayet, 1908, 1909; Stehlin, 1910, 1925; Daxner-Hoeck, 1971; Cabard et al. 1980; Dineur, 1982; Hellmund, 1991; Ginsburg et al. 2000; Ginsburg & Chevrier, 2005; Pickford & MacLaren, 2022; Pickford, 2022a, 2022b). For some investigators, the large tusk-like tooth in the antero-lateral corner of the mandibular symphysis is the second incisor (Cabard et al. 1980; Hellmund, 1991; Pickford & MacLaren, 2022), while for others it was considered to be the third incisor (Dineur, 1982; Ginsburg & Chevrier, 2005) and for even others, when found isolated, it was thought to be an upper canine (Depéret, 1895; Mayet, 1908; Daxner-Hoeck, 1971).

Difficulties have also been encountered in the identification of the meristic position of isolated upper incisors (Hellmund, 1991) and canines of anthracotheres, with at least five instances of lower tusk-like incisors of Brachyodus being incorrectly identified as upper teeth (Depéret, 1895; Mayet, 1908, pl. 7, fig. 2; Stehlin, 1925; Daxner-Hoeck, 1971, pl. 1, fig. 9; Cabard et al. 1980, fig. 8). There is even discussion about the meristic position of teeth in situ in the premaxilla of the genus (Ginsburg et al. 2000).

Mis-attribution of these teeth naturally affected metric analyses, because the mesio-distal length and labio-lingual breadth measurements were transposed in the tables of measurements. For example, if a tusk-like tooth is considered to be a lower canine because it has a wear facet on its distal edge, as in carnivores for example, the same tooth has a lingual wear facet if it is correctly identified as an upper central incisor of Brachyodus.

Some of the maxillae previously attributed to Brachyodus have an alveolus for the upper canine at the anterior end of the maxilla, separated from the cheek teeth by a diastema. However, none of the maxillae preserves this tooth in situ. Of great interest is the observation that a maxilla (CGM M94-90) of Brachyodus depereti, housed in the Cairo Geological Museum (Pickford & Abdel Gawad, 2024) has an alveolus for an upper canine. However, none of the available specimens attributed to Rusingameryx possessed this tooth.

The anthracothere species in the French Faluns and the Sables de l’Orléanais, in which the mandibular symphysis is unfused or incompletely fused, has lower second incisors that are not markedly tusk-like, unlike species of the genera Brachyodus and Rusingameryx, in which these two teeth are tusk-like. It is surmised that the upper central incisors of this anthracothere were also not markedly tusk-like.

As a consequence, several questions arise: are the differences that exist between the anterior dentitions of anthracotheres of specific, generic or subfamilial significance? Are the differences in morphology and dimensions linked solely to dimorphism and/or to bimodality? Are the differences in dental formulae simply cases of individual variation linked to ontogenetic stages, with older individuals shedding teeth that they possessed when they were young? All four scenarios are present in the literature. The morphological diversity among the incisors suggests that there are likely two genera represented in the European collections, one of which shows some resemblances to Bothriogenys and Elomeryx (herein attributed to the new genus Chitenaymeryx) the other being Brachyodus.

Although the relationship between anthracotheres and hippopotamids is not the main focus of this monograph, some remarks are made about the subject, because the anterior teeth of the two groups indicate that their relationship within the Suiformes is remote, the differences suggesting a distinction at least at the superfamily rank. In this paper, the hippos are classified within Suoidea, and Brachyodus within Anthracotherioidea, unlike the preference of Scherler et al. (2019) in which the anthracotheres are classified in Hippopotamoidea (sensu Gentry & Hooker, 1988). One of the few derived (apomorphic) characters shared between the two groups is the suiform morphology of the talus, but even with this bone, there are significant morphological differences between hippos and anthracotheres (Pickford, 2008). In the two groups, the rest of the bony skeleton is markedly divergent (Pickford, 2015), as are the teeth, especially the anterior ones, the focus of this contribution.

Apart from that, there are major differences between the anterior dentitions of Anthracotheriinae sensu stricto (i. e. Anthracotherium, Paenanthracotherium, Microbunodon, Nabotherium and close relatives) (Teller, 1886; Scherler et al. 2019; Hünermann, 1964; Sileem et al. 2016) and several (but not all) genera currently included in the Bothriodontinae (Bothriogenys, Brachyodus, Rusingameryx and close relatives comprising one subgroup; Libycosaurus, Sivameryx and Afromeryx comprising another; while some genera such as Epirigenys are of poorly understood affinities) (Pickford, 2022b). Indeed, the differences suggest that, as currently understood, the Bothriodontinae contains a heterogeneous assemblage of genera that are better classified in separate subfamilies.

In the Anthracotheriinae and in some genera of Bothriodontinae sensu lato (Libycosaurus, Sivameryx, Afromeryx) the tusk-like teeth are canines (Hellmund, 1991) with the usual occlusal relationships (the upper canine is positioned slightly distal to the lower canine and thus its mesial edge occludes with the distal edge of the lower canine) (Pickford, 2005). In contrast, in some genera currently included in Bothriodontinae (Brachyodus, Rusingameryx) the tusk-like anterior teeth are incisors, in which the posterior surface of the enlarged upper central incisor occludes with the anterior edge of the enlarged lower second incisor. The canines in these latter taxa are either reduced in dimensions and are incisiform in morphology or, in some taxa they are suppressed. These morphofunctional differences pose questions regarding the currently accepted taxonomies and systematics of the Anthracotherioidea (or Hippopotamoidea of some authors) (Scherler et al. 2019). As currently classified by these authors, the Bothriodontinae is paraphyletic, and the higher level ranks, Anthracotheriidae and Hippopotamoidea, are polyphyletic.

The aim of this contribution is thus to examine the fossil anthracotheres from the Faluns de La Touraine and Anjou and the Sables de l’Orléanais with a view to clarifying some of the uncertainties concerning the anterior dentition. However, inadequacies of preservation leave room for doubt concerning many of the isolated teeth, especially the upper third incisors. The presence or otherwise of canines in these anthracotheres remains to be solidly demonstrated, but currently available samples suggest that the lower canines had been suppressed in Brachyodus onoideus, but that «Brachyodus » intermedius (herein called Chitenaymeryx intermedius) retained the upper and lower canines.

It is becoming clearer that some specimens from the French and Portuguese Early Miocene sedimentary deposits, hitherto attributed to Brachyodus, possessed three incisors and a canine in the lower and upper jaws, but that later forms suppressed one or more teeth, usually the lower central incisors and the lower canine in the mandible, and the third incisor (Dineur, 1982) or the second incisor (Ginsburg et al. 2000) in the premaxilla and the canine in the mandible. The differences are great enough that it is no longer taxonomically realistic to attribute the species onoideus and intermedius to the same genus. Not only does the presence-absence of lower canines represent a major difference between the two taxa, but there are other differences, especially to do with snout length, mandibular symphysis length, fusion or otherwise of the mandibular symphysis, length of diastemata, and the form of the anterior margin of the mandible between the anterior edges of the second incisors: curved or straight.

The anterior deciduous dentition of anthracotheres from the French Faluns is still poorly represented among the available fossils, so it is not possible to deduce the correct juvenile dental formula, although it is clear that a few isolated deciduous incisors are present in the collections (Ginsburg & Chevrier, 2005). The most complete juvenile specimen available, from Horta das Tripas, Portugal (Roman, 1907) preserves alveoli for four deciduous teeth on either side of the mandibular symphysis, which is unfused, and thus does not belong to Brachyodus, but to the another genus, herein named Chitenaymeryx.

It is not possible to resolve the main taxonomic issues concerning Brachyodus by studying only the fossils from the French Faluns and the Sables de l’Orléanais. Because the Early Miocene strata at Moghara, Egypt, share at least one taxon (possibly two) with the French deposits, it was necessary to revise the large anthracotheres from Moghara as an integral component of the overall study. A separate paper devoted to this effect has been published (Pickford & Abdel Gawad, 2024). The anthracotheres from the Eggenburg deposits in Austria, and the sediments of the Tagus Basin, Portugal, are included in this paper when necessary, as are fossils from Uganda and Kenya.

Prior to the descriptions of the fossil material, a detailed historical overview of the various scientific issues relevant to understanding the diverse interpretations of the osteology and morphology of the anterior teeth of these anthracotheres is presented. It will include Anthracotheriidae verii (i. e. true anthracotheres originally identified as such, as well as specimens that were initially mis-attributed to non-anthracotherian mammals) and Anthracotheriidae spurii (i. e. specimens that have, at one time or another, been identified as anthracotheres, but which belong to other mammals). This section will be followed by the descriptions of the fossils from Europe, which will be discussed in systematic order.

To complicate matters, a few bibliographic references were incorrect when first published and the errors tended to propagate through the literature, and on occasion even mutated into increased errors. An example is provided by references to the presence of anthracotheres at Eggenburg, Austria. First published by M. Neumayr in 1888, the date of publication was given as 1883 by Depéret (1895) and by 1908, not only had the wrong date been propogated, but also the author’s name had changed to Neumeyr (Mayet, 1908). This class of errors is annoying, but is not serious from a scientific point of view. It does however, reveal weaknesses in the research protocols of scientists who seemingly failed to read the orginal papers concerning the topic upon which they were working.