|Rhinocerotidae (Mammalia) from the Late Miocene of Bulgaria
Denis Geraads, Paris, and Nikolaï Spassov, Sofia
With 3 plates, 3 figures and 7 tables.
Wir Beschreiben die Resten von Rhinocerotiden aus etwa zehn Obermiozän Lokalitäten in Bulgarien. Die wichtigsten Fundorte sind Kalimantsi, Hadjidimovo und derjenige kurz erfundene in Strumyani. Mindestens sieben Arten werden wegen Elementen vom Schädel repräsentiert. Ein Schädel mag zu Chilotherium kiliasi gehören, eine Art, für welche eine neue Untergattung, Eochilotherium, etabliert wurde. Einige Stücken gehören zu Ch. (Chilotherium). Man erreicht in Bulgarien der westlichen Grenze der Entdeckung dieser Gattung. Acerorhinus wird sehr oft in Kalimantsi gefunden, aber dieser Gattung soll wiedergearbeitet werden. Eine unerwartete Entdeckung von Brachypotherium in Ahmatovo ist die letzte in Europa. Die hornige Rhinos Ceratotherium und Dihoplus liegen beiden in zwei Lokalitäten. Diese taxonomische Mannigfaltigkeit bedeutet die Anwesenheit von verschiedenen Umgebungen, denn die Lagerungen ziemlich derselben Alters sind. Es entspricht der geographische Lage von Bulgarien, wo die Gebiete von asiatischen Chilotherium, von der pontische Acerorhinus und von der Balkan-Iraner Ceratotherium überschneiden. Dazu kommt die wahrscheinliche Einwanderung von Brachypotherium aus Afrika zu Europa.
We describe the remains of Rhinocerotidae from about 10 Late Miocene localities of Bulgaria, the main ones being Kalimantsi and Hadjidimovo, and the newly discovered site of Strumyani. At least seven species are represented, mostly by cranial elements. A skull is assigned to Chilotherium kiliasi, a species for which a new sub-genus, Eochilotherium, is established. Some fragmentary remains are referred to Ch. (Chilotherium), which reaches in Bulgaria the westernmost limit of its range. Acerorhinus is well represented at Kalimantsi, but this genus probably deserves revision. An unexpected discovery is that of Brachypotherium at Ahmatovo, its latest known occurrence in Europe. The horned rhinos, Ceratotherium and Dihoplus, co-occur at two localities. This relatively great diversity points to a variety of environments, as it is unlikely that the main fossil faunas are very different in age. It also reflects the geographic position of Bulgaria, where the ranges of the mainly Asiatic Chilotherium, of the Pontic Acerorhinus and of the Balkano-Iranian Ceratotherium overlap, in addition to a possible immigration of Brachypotherium from Africa into Europe.
Key-words: Rhinocerotidae, Mammalia, Late Miocene, Bulgaria, Balkans
Schlüsselwörter: Rhinocerotidae, Mammalia, Obermiozän, Bulgarien, Balkan
Addresses of the authors: Denis Geraads, CNRS UPR 2147, 44 rue de l'Amiral Mouchez, 75014 Paris, France
Nikolaï Spassov, National Museum of Natural History, Tsar Osvoboditel 1, 1000 Sofia, Bulgaria
Zusammenfassung - Abstract
Until now, the Miocene Rhinocerotidae of Bulgaria have been known only through a few outdated works, such as the early papers of Bakalov (1933, 1939), a brief description by Nikolov & Kovachev (1966) and the monograph of Bakalov & Nikolov (1962). Since then, more discoveries, and large-scale excavations conducted mainly in the areas of Hadjidimovo (Mesta River valley), Kalimantsi and recently at Strumyani (Struma River valley), all in south-western Bulgaria, have significantly improved the collections. A summary of our present knowledge of the main faunal associations is given in Spassov et al. (2006). Furthermore, several recent papers dealing with upper Miocene rhinos of the area (Heissig 1975, 1996, 1999; Kaya & Heissig 2001, Geraads 1988, 2005; Geraads & Koufos 1990; Fortelius et al. 2003; Giaourtsakis 2003; Giaourtsakis & Heissig 2004; Giaourtsakis et al. 2006) have shed new lights on the evolution of this family. The present paper is an attempt to revise all available material, stored in the National Museum of Natural History, Sofia (NMNHS), in its Assenovgrad Museum branch (NMNHA), and in the Department of Geology, University of Sofia (DGUS). It is compared with other Late Miocene to Recent rhinos, stored in the Natural History Museum, London (BMNH); Faculté des Sciences, Lyon (FSL); Hessisches Landesmuseum, Darmstadt (HLMD); National Museums of Kenya, Nairobi (KNM); Musée Géologique et Minéralogique, Lausanne (MGML); Macedonian Museum of Natural History, Skopje (MMNH); Muséum National d'Histoire Naturelle, Paris (MNHNP); Maden Tetkik ve Arama Museum, Ankara (MTA); Naturhistorisches Museum Basel (NHB); Naturhistorisches Museum, Wien (NHMW); National Museum of Ethiopia, Addis Ababa (NME); Paläontologisches Institut, Münster (PIUM); Senckenberg Museum, Frankfurt (SMF); Staatliches Museum für Naturkunde, Stuttgart (SMNS).
Fossil localities (Text-fig. 1)
Most of the material of Rhinocerotidae comes from the richest Bulgarian Late Miocene fossiliferous sites: Hadjidimovo and Kalimantsi. Detailed studies of the fauna from both these localities started recently (Kovachev 1988, 2001; Spassov & Ginsburg 1999; Geraads et al. 2001, 2003, 2005b, 2006; Kostopoulos et al. 2001; Spassov 2002; Hristova et al. 2002a & b; Hristova & Kovachev 2005; Koufos et al. 2003; Markov 2004a & b; Spassov & Geraads 2004; Spassov et al. 2006).
Hadjidimovo is situated in the Mesta river valley, East of the Pirin ridge, not far from the Greek border. Four fossiliferous sites are known, the most important of which being Hadjidimovo-1 (Hadjidimovo-Girizite), with light clayey sands belonging to the Nevrokop formation (Vatsev 1980). It yielded more than 20000 fossils (NMNHA) belonging to about 30 mammal species, including the rhinos described below, making it one of the richest Late Miocene sites of the Eastern Mediterranean and peri-Pontic areas. Although the Hadjidomovo site (unknown locality) was first mentioned by Nikolov (1973, 1985), this huge collection was mostly accumulated and cared for, thanks to the efforts of D. Kovachev, mainly between 1985 and 1998. Hadjidimovo-1 could be earlier than Pikermi, as indicated by the evolutionary stages of Mesopithecus (Koufos et al. 2003) Hipparion (Cremohipparion) mediterraneum (Hristova & Spassov 2005), Adcrocuta eximia (pers. obs. N.S.), and the presence of Tragoportax rugosifrons (Spassov & Geraads 2004). Spassov (2002) suggested that the locality could be dated to the MN11/MN12 boundary.
All materials from Hadjidimovo are stored in NMNHA.
The Late Miocene deposits around the Kalimantsi village, on the western side of the Pirin ridge, in the Middle Struma basin, cover a large area (ca. 6 km2) with more than 10 known fossiliferous spots, scattered over a vertical distance of about 100 m. The recent revision of their biochronology demonstrates that all Kalimantsi sites are of Turolian age. Kalimantsi-1 (the river bank), belonging to the newly established Gradishte genetic lithocomplex, is more likely to yield a fauna of early Turolian age. Hipparion cf. macedonicum is known only from this level. The other Kalimantsi faunas (referred to the Strumyani genetic lithocomplex) are considered to belong to the middle Turolian (Spassov et al. 2006), probably slightly older than the Pikermi fauna, according to the evolutionary stages of Hipparion (Cremohipparion) mediterraneum and H. (Hippotherium) brachypus (Hristova & Kovachev 2005). Gazella sp., H. (Cremohipparion) mediterraneum and H. (Hippotherium) brachypus dominate in these levels. Tragoportax cf. amalthea, Palaeoreas lindermayeri, Helladotherium duvernoyi, Bohlinia attica, Adcrocuta eximia, Mesopithecus pentelicus are also well represented (Spassov et al. 2006). Presently, at least 34 species are known from Kalimantsi (all localities combined). Some specimens of Rhinocerotidae from Kalimantsi were mentioned and illustrated by Bakalov & Nikolov (1962); they are stored in DGUS. New materials collected by D. Kovachev and co-workers, mostly from Kalimantsi-1, are stored in NMNHA.
The new rich locality Strumyani-2, near Sandanski in the Struma River valley, that we excavated in 2002-2003, has not been published in detail, but is likely to be of the same broad age (Turolian), probably early Turolian or first half of the middle Turolian, from the Hipparion assemblage (L.Hristova, pers. comm.).
Some other localities have yielded only a few fossils.
Central part of Western Bulgaria, Sofia region
A quarry near Staniantsi in the Godech region (N.-W. Bulgaria), near the Serbian border, yielded only a fragment of incisor. The fauna might be Turolian (Nikolov 1985; Spassov 2002; Markov 2004b). It yielded unpublished mandibular teeth similar to those of Tapirus balkanicus from Balsha (Spassov & Ginsburg 1999) that can be referred to the T. pannonicus - T. balkanicus group and also indicate an ante-Pliocene age for the locality.
Struma river basin, S-W Bulgaria
Some localities yielded fragmentary rhino remains, but very little is known about the geologic context. A rhino incisor comes from the coal-rich levels, at a depth of about 200 m, of the Oranovo quarry near Simitli (south of Blagoevgrad), eponym locality of the newly established Oranovo genetic lithocomplex, probably of early Upper Miocene age.
Two rhinoceros finds were discovered near Slatino, north of Blagoevgrad. The deposits of the Slatino genetic lithocomplex are also considered as belonging to the lower part of the Upper Miocene (Tzankov et al. 2005). Their precise provenance (most probably two different fossiliferous spots) is unknown. Their matrix is a grey-green gravellite clay concretion most probably washed by fluent waters. Such kind of clays, typical for the Slatino genetic lithocomplex (Spassov et al. 2006) crop out in several places around Slatino village.
At Kromidovo, not far from Kalimantsi, the fossils come from a spot in the village, but we could not locate it during our 2005 survey. The presence of Mesopithecus pentelicus (Koufos et al. 2003; Spassov et al. 2006), would point to an age close to MN12, but the measurements of the specimens do not quite match those of the Pikermi sample, and this age estimate is far from secure.
Maritsa and Tundja river basins, middle part of Southern Bulgaria
A number of Upper Miocene fossiliferous spots are known from the Cherkezitsa River valley near Plovdiv, between the villages Seltsi, Bogdanits and Ahmatovo (Nikolov & Kovachev 1966). All of them could be approximately contemporaneous. The richest mammal fauna is from the sand quarry of Ahmatovo, the most important locality in the Maritsa River basin (Bakalov & Nikolov 1962; Nikolov & Kovachev 1966). These authors reported Gomphotherium angustidens, but it is in fact Choerolophodon (Tassy 1983). Dragomanov et al. (1981) referred the fossiliferous spots of the quarry to three different stratigraphic levels, but our observations do not confirm this stratigraphic subdivision, and age differences must be minimal. The revised list of the proboscideans from Ahmatovo is: Deinotherium gigantissimum, “Mammut” cf. borsoni, Choerolophodon pentelici, Tetralophodon atticus, Anancus sp. (Markov 2004b). As the genus Anancus appears in Bulgaria and Europe probably in the middle Turolian but in post-Pikermi time (Spassov et al. 2006) and Choerolophodon is not definitely known in the late Turolian, the age of Ahmatovo could be the second half/the end of the middle Turolian (MN12). The Brachypotherium skull and isolated teeth of an old individual described below are from the Ahmatovo quarry itself, while the isolated P4 of the same genus is from the deposits of the river-bank of Tcherkezitsa between Seltsi and Ahmatovo, stratigraphically slightly lower, but probably not much older.
In 1988 some small bovid, small giraffid and rhinos remains were discovered in the yard of the coachbuilder in the town of Yambol (on Tundja river). The Neogene deposits along the Tundja River belong to the Elhovska formation (Kojumdjieva et al. 1984), the age of which ranges from the Maeotian to the Dacian-Romanian.
Order Perissodactyla Owen, 1848
Family Rhinocerotidae Gray, 1821
We must stress again that the systematics of fossil rhinos is strongly defective, mainly due to the lack of comparative descriptions and diagnoses. We urge future workers to follow the recommendations of Cifelli & Kielan-Javorowska (2005), and give differential diagnoses, of which we tried to provide some drafts, in the hope that these might serve as a basis.
Uppercase denotes upper teeth, lowercase denotes lower teeth. Measurements are in mm, and follow Guérin (1980). Measurements of tooth rows are at occlusal level, measurements of isolated teeth are maximum.
Genus Chilotherium Ringström, 1924
Type-species: Chilotherium anderssoni Ringström, 1924: 26. Late Miocene of China.
Diagnosis: Both sexes hornless. Straight nasals. Frontal region depressed, forming a shallow fossa. Parietal crests wide apart. Maxilla flat or slightly depressed, facial crest and nasal notch far from orbit. Premaxillae forming two thin vertical walls, lacking incisors. Lower jaw with widened symphysial part and large tusk-like i2, separated by a broad diastema. Permanent dental formula: I 0/1(2) C0/0 P 3/3 M3/3; deciduous dental formula DI 0/2 DC 0/0 DP 4/3(4). Antecrochet long. Body low and stout, belly low above ground. Limbs much shortened. Manus and pes tridactyl with short, diverging metapodials. Lateral metapodials obliquely directed backwards (mostly translated from Ringström 1924).
Sub-genus Eochilotherium nov. subgen.
Type-species: Aceratherium kiliasi Geraads & Koufos, 1990.
Diagnosis: supra-orbital process somewhat inflated, post-orbital process weak. Antecrochet moderate. Mandibular symphysis slightly broadened, incisors of plesiomorphic condition, close in structure to the i2 of Aceratherium / Acerorhinus (see below), not strongly enlarged, with subtriangular cross-section, with rounded border between the lateral and the ventral surfaces, which are subequal in width.
We reluctantly create this new sub-genus, because there are already too many names for European hornless rhinos, but it lacks several of the derived features which define the nominal sub-genus, C. (Chilotherium), which includes e.g., C. anderssoni from China, C. persiae from Maragha, and C. schlosseri from Samos. These apomorphic characters are: 1) lack of upper incisors; 2) enlarged flattened lower incisors, inserted wide apart in a broadened symphysis, with an upturned mesial (internal) edge - this is perhaps the most important apomorphic feature, unique to this group (see below for a more detailed description); 3) no supra-orbital process, and a well-marked postorbital one; 4) short premolars; 5) very long antecrochet; 6) brachypodial limbs. Most of these characters were already listed in the original diagnosis (Ringström 1924:26). In C. (Eochilotherium), in which we include C. (Eochilotherium) kiliasi and C. (E.) samium (see below), the premolars are also short, but (1) and (6) are unknown, the lower incisors are medium-sized, with plesiomorphic characters (Geraads & Koufos 1990, pl. 2, fig. 5), the symphysis is only moderately broadened, there is a strong supra-orbital process but no post-orbital one, and antecrochet is moderate. We agree with Heissig (1996, 1999) and Fortelius et al (2003) that these species can best be taken as primitive Chilotherium, but we prefer to keep them as taxonomically separate from the highly derived members of this group.
Chilotherium (Eochilotherium) cf. kiliasi Geraads & Koufos, 1990
Type locality: Pentalophos, Greece, late Vallesian (MN 10) ? (see Geraads & Koufos 1990: figs. 1-7; pl. 2, figs. 1-5; pl. 3, figs. 1 & 4; non pl. 3, figs 2, 3, 5).
Specific diagnosis: Nasal notch V-shaped, ante-orbital depression shallow. Zygomatic arches relatively weak. Upper incisors unknown. Cement present on premolars and molars. Antecrochet short. Long persisting di2, resulting in a hexaprotodont dentition even in old individuals. Diastema short (modified from Geraads & Koufos 1990 ).
Material: unnumbered partial skull, with M3 erupting, lacking the cranial vault and pre-dental portion, from Kromidovo, DGUS (Pl. 1, Figs. A-B).
Description: Only the ventral part of the skull is preserved. The zygomatic arches form an even curve, without sharp bend at the level of the glenoid fossa. The cranial basis is narrow. The premolars are small relative to the molars (measurements: table 1). All teeth have a long crochet, and a weak antecrochet, but small tubercles in the median valley, best expressed on M1, may represent its lingual extensions. The premolars have a moderate lingual cingulum, and the hypocone of P3 has a small mesial spur, but the median valley remains wide open. There is a distinct paracone fold, and the labial wall of the ectoloph is less flattened than in typical Chilotherium.
Comparison and discussion: We refer this specimen to C. (E.).cf. kiliasi on the basis of the close similarity in size and morphology of the teeth with those of a maxilla from Pentalophos (Geraads & Koufos 1990, fig. 2), but we are well aware that cheek-teeth alone are not a very sound basis for species identification. Geraads & Koufos (1990) described as A. kiliasi a sample of skulls and mandibles that Fortelius et al. (2003) showed to be heterogeneous. Indeed, the mandible PNT-142 has large premolars and is not anteriorly broadened, and must belong to a different species (see below). A. kiliasi, of which the best specimen from Pentalophos is PNT 12 (Geraads & Koufos 1990, pl. 3, fig. 3-4), was referred to Chilotherium by Heissig (1996, 1999) and Fortelius et al (2003), who also referred some material from Sinap to this species. They considered it to be close to C. samium (Weber, 1905), from Samos, a species that was excluded from Chilotherium by Geraads & Koufos (1990) on the basis of Weber's description. Since the type material of C. samium is lost, there will indeed always remain a doubt about the identity of the species, but sound hypotheses must be based upon actual specimens. Thus, our understanding of C. samium is based upon a skull from Samos, SMF-M3601, labelled as such by K. Heissig. Both species lack the apomorphic features of C. (Chilotherium) listed above, and are better referred to a separate sub-genus.
Fortelius et al. (2003) thought that C. samium is close to C. wimani Ringström, 1924, from China, but the published material of C. wimani does show the apomorphic features of the nominotypical subgenus, especially the huge lower incisors inserted on a much broadened symphysis (Ringström 1924, fig. 28 and pl. 8, fig. 1).
Sub-genus Chilotherium Ringström, 1924
Type-species: Chilotherium anderssoni Ringström, 1924. Late Miocene, China.
Diagnosis: Supra-orbital process may be present but weak; post-orbital process well-marked. Lower jaw with strongly widened symphysial part. Antecrochet very long. Lower incisors (i2) long and very large, with triangular cross-section. The flat narrow latero-ventral face (lateral in other rhinos) is slightly broadened towards the base, and separated from the medio-ventral one by a sharp edge. The large dorso-lateral and medio-ventral ones are strongly broadened towards the base; their common edge forms the cutting edge, kept sharp by attrition with the upper lip, and dorso-medially oriented.
Chilotherium cf. sarmaticum Korotkevich, 1958
Type locality: Berislav, Ukraine, late Vallesian or early Turolian.
Diagnosis: a small and less cursorial chilothere; parietal crests wider apart than in other chilotheres; nasals short; antecrochet moderate; metapodials less massive and limbs more shortened than in other species; talus usually with only two calcanear facets (simplified from Korotkevich 1958b and 1970).
Material: i2 from Oranovo; perhaps a fragment of incisor from the quarry near Staniantsi.
Description: The crown of the Oranovo incisor is slightly curved. The tooth is flattened, of regular width. The cross section is triangular, even at the base. The dorso-lateral surface lacks enamel, as in other Chilotherium s. str., and is slightly concave, abruptly becoming narrower towards the apex. It is separated from the slightly convex medio-ventral surface by a dorso-medial edge much sharpened by wear, but with a distinct prominence at its base. A third, ventro-lateral surface is flat or even slightly concave, and delimited by two clear keels from the other surfaces. It is of almost even width towards the base. This third surface is not so sharply delimited and becomes broader towards the base in the tusks of the larger species of Chilotherium. Measurements: max. basal width of the wear surface ~ 39 mm., crown length = 78 (restored ~ 86 mm.), max. root diameter = 35 mm., width of ventro-lateral surface ~12 mm.
Chilotherium sarmaticum was first described from Berislav, Ukraine (Korotkevich 1958a & b; 1970). Although not discussed in the synonymy lists of the chilotheres published by Heissig (1975, 1999), this species is smaller than the currently accepted taxa from the Turkey and Eastern Europe. The Oranovo incisor is also smaller than the large chilothere species from Grebeniki and Maragha (Spassov et al. 2006). Berislav has been referred to MN11a by Korotkevich (1988; Vallesian for this author), to the late Vallesian by De Bruijn et al. (1992), or to the early Turolian by Krakhmalnaya (1996). The morphology described by Korotkevich (1970) indicates a still less cursorial locomotory apparatus for Ch. sarmaticum than for other true chilotheres of the open spaces and the faunal analysis made by this author shows a more forested habitat in Berislav than in other localities of the same region such as Grebeniki and Novaya Emetovka. This would also be in good agreement with a Vallesian age. The Oranovo incisor is also virtually identical in shape and size with Ch. cf. sarmaticum described from Reghiu, Romania, a locality that was assigned to the beginning of MN11 (Stiuca 2003), but could be older. Oranovo could be of similar age. Codrea (1996) noted that it apparently never crossed the Carpathians. Oranovo marks together with Staniantsi and Kromidovo (Bulgaria), Pentalophos-1 (Greece) and Marievo region (Rep. of Macedonia: a molar fragment with preserved protoloph, coll. NMNHS) the western limit of the range of Chilotherium.
The tooth from Staniantsi is the apex of the left incisor, with a black to brown-grayish color, indicating that it comes from the coal levels of the quarry. It has all the characteristics mentioned in the diagnosis of Ch. (Chilotherium). The tip of the tusk (as in the Slatino i2) has a small abrasion facet indicating that it was also perhaps used for digging; from what remains of the tooth, it looks somewhat larger than the i2 of Ch. cf. sarmaticum from Oranovo; thus, we refer it only doubtfully to the same species.
Chilotherium cf. kowalevskii Pavlow, 1913
Type-locality: Grebeniki, Ukraine, probably first half of the Turolian.
Diagnosis: dorsal skull profile gently concave, zygomatic arch very robust, paroccipital process rather short; upper premolars large, with strong crochet, and long antecrochet connecting the metaloph, thus closing the medisinus almost up to the top of the crown; i1 shed late in life; limb bones stocky.
Material: FM 2700 unworn left P3 and FM 2699 right P4 (Pl. 1, Fig. C); FM 2698 fragment of left mandible with m1-m3; FM 2697 incomplete mandible with left i2 and left and right p2-m2 (coll. NMNHS), coachbuilder yard, Yambol.
Description: besides the slight difference in L/W proportions (P3: 34.2 x 42.6; P4: 38.5 x 45), the two upper teeth are remarkably alike and are certainly from the same individual. They are moderately hypsodont; the labial wall is rather flat, with only a small paracone fold which vanishes towards the base of the crown. There is a complete lingual cingulum around P3, and probably also around P4, but it is much higher anteriorly and posteriorly, where it closes a antero-posteriorly elongated deep post-fossette. The protoloph forms a regular curve from the paracone to the hypocone, completely closing lingually the medisinus up to the top of the crown on P3, whereas on P4 its lingual end (antecrochet) descends somewhat, to meet the hypocone lower down. The protocone is only apparent as a distinct entity in the lower half of the crown, a deep pinching separating it from the rest of the loph. Both teeth have strong crochet and crista, which probably met to close a medifossette after heavy wear.
The mandible FM 2697 is of a subadult specimen. It shows the broadened symphysis typical of Chilotherium. The right tusk (i2) is missing, and only about 1/3 of the left one is erupted. Its cross-section is clearly triangular, with well expressed edge between the ventro-lateral and ventro-medial faces. The paralophids of the cheek-teeth, including p2, are long, the anterior and posterior valleys are short, V-shaped. Measurements are given in Tables 2 & 3
The semimandible FM 2698 is from an old individual and lacks the anterior half. It has a thin corpus and relatively massive ramus, strongly developed and rounded mandibular angle, and large vascular incisure.
Comparisons: This morphology of the upper premolars (similar to the semimolariform pattern of Heissig, but probably secondarily acquired) is not common in the Eastern Mediterranean, and Heissig took it as characteristic of Chilotherium kowalevskii, a species first described from Grebeniki by Pavlov (1913). The mandibular teeth dimensions fit well those given by Korotkevich (1970) for the type locality, Grebeniki. Heissig (1996) listed it also from three Anatolian localities, and from Losengrad (Kirklareli) in European Turkey, from where Bakalov & Nikolov (1962) had reported an upper molar as Aceratherium angustifrons Andree, 1921, a Samos species that Heissig (1999) considered as synonymous with Ch. kowalevskii.
Pending detailed study of the material (especially the rich collection from Çorakyerler), we will accept Heissig's conception of the species, which is thus distributed along the Northern, Western and Southern parts of the peri-Pontic region.
Material: a right M3, coll. NMNHA – A1030, from Ahmatovo (Pl. 1, Fig. D).
Description and comparison: The tooth is in medium wear but was probably rather hypsodont. The labial wall is vertical, rather flat and with smooth relief, with weak paracone fold. There is a strong cingulum along the base of the protoloph base, plus a weak labial cingulum at the base of the ectoloph. The crochet is strong and curved mesially, the crista is vestigial. The antecrochet is characteristic of a chilothere in being very long, with a distal part curved lingually, descending abruptly down and reaching the lingual border of the tooth, leaving a narrow deep fossette between it and the protocone. The hypocone part is missing but was separated down to the bottom of the medifossette from the antecrochet. The protoloph is regularly curved. The protocone becomes abruptly wider towards its base; it is well separated from the rest of the protoloph by a strong double pinching. The tooth is large: ectoloph length = 54.4 mm (estimated restored length ~ 56 mm); width = 51 mm. It is larger than the teeth of Ch. sarmaticum from Berislav and of Ch. kowalevskii from Grebeniki (Korotkevich 1970). By its morphology and dimensions it represents a large and evolved species; for geographical reasons, the best choice would be Ch. schlosseri, known from Samos, but a specific identification would be harsh.
Genus Acerorhinus Kretzoi, 1942
Type-species: Acerorhinus zernowi (Borissiak, 1914), from Sebastopol (MN10). Diagnosis: post-orbital process usually small, enlarged supra-orbital process, nasal notch close to orbit, often with a strong facial crest, premolars long, upper teeth with strong crochet, lower teeth with a short paralophid. The symphysis is moderately widened; the lower incisors are large and tusk-like, with subtriangular cross-section, with smooth border between the lateral and the ventral surfaces, which are subequal in width, and with large dorsal surface, moderately enlarged towards the base, limited by sharp medial and lateral cutting edges.