A new primitive elephantid from Turkey

Quaternary International 445 (2017) 43e49

Contents lists available at ScienceDirect

Quaternary International journal homepage: www.elsevier.com/locate/quaint

A new primitive elephantid from Turkey Ebru Albayrak ü, Tabiat Tarihi Muzesi, 06800 Ankara, Turkiye Maden Tetkik ve Arama Genel Müdürlüg

a r t i c l e i n f o

a b s t r a c t

Article history: Available online 6 October 2016

lı In this study, a complete mandible and a tusk of an elephantid which were found in Büyükyag (Kırıkkale, Turkey) are identified and discussed. The mandible bears the left m1 and both left and right m2 and m3. Molars of this specimen have very primitive features, e.g. low plate number and lamellar frequency, and thick enamel, that make it similar to Elephas planifrons of the Siwaliks and early Mammuthus species, e.g. M. rumanus of Europe. According to previous studies, the locality is dated to the early Late Pliocene, i.e. MN16a. As Turkey occupies a very important position for the migration of Elephantidae lı may be a key to understanding the between Africa, Asia and Europe, the material from Büyükyag evolution of early elephantids out of Africa. © 2016 Elsevier Ltd and INQUA. All rights reserved.

Keywords: Turkey Elephant Elephas planifrons Mammuthus rumanus Mandible

1. Introduction Turkey occupies a key position between Africa, Asia and Europe, and some important elephantid specimens of different evolutionary stages have been found and studied there. Falconer (1857) described the species Elephas armeniacus on the basis of molar teeth from Erzurum, although the status of this species is now considered problematic (Adam, 1988; Dayan, 1989; Lister, 1996). S¸enyürek (1960, 1961) identified Archidiskodon planifrons, A. meridionalis and Elephas trogontherii from different localities of Turkey, but these records are in need of revision. Becker-Platen and Sickenberg (1968) and Sickenberg et al. (1975) identified Mammuthus (Archidiskodon) meridionalis, Palaeoloxodon sp. ex gr. Palaeoloxodon antiquus and A. planifrons from the Villafranchian €g üto €nü, as well as the gomphothere Anacus locality of Yukarıso arvernensis. Albayrak and Lister (2012) identified and discussed several species of elephantids from different ages and localities. They identified primitive M. meridionalis specimens which may provide evidence of transition from M. rumanus to M. meridionalis. M. trogontherii was identified from two localities and was metrically typical for the species. Lastly, they described E. maximus remains from Gavur Lake Swamp which were dated to around 3500 BP according to radiocarbon analysis. Boulbes et al. (2014) identified an incomplete M3 of ‘Archidiskodon’ (¼Mammuthus)

Abbreviations: MTA, The General Directorate of Mineral Research and Exploration; NHM, Natural History Museum, London. E-mail address: [email protected]. http://dx.doi.org/10.1016/j.quaint.2016.09.029 1040-6182/© 2016 Elsevier Ltd and INQUA. All rights reserved.

meridionalis meridionalis from the Denizli Basin which has features similar to the type specimens from Upper Valdarno, Italy. There are uncertainties and questions about the evolution and the distribution of early elephantids, especially on the genus Mammuthus. And also disagreements concerning the genus affiliation and the species of early elephants are still continuing (Lister, 1996; Lister and Sher, 2001; Lister and van Essen, 2003; Lister et al., 2005; Markov, 2012; Baygusheva and Titov, 2012; Baigusheva et al., 2016). Even with the small number of studies of elephantids from Turkey, it can be seen that there are important remains that can help solve problems concerning the evolution and migration of the group. In this context although the systematic position of fossil elephants is mainly determined by the features of skull structure and morphometric data of last molars (Maglio, 1973; Baigusheva et al., 2016) the mandibular specimen described here is a potential key to understanding the migration and early evolution of elephantids outside Africa. In the present study, the new remains from Kale Tepe-3 are discussed. Because of the primitive features of the mandible the definite identification is problematic.

2. Locality The locality, Kale Tepe-3, is in the southwestern part of the Çankırı-Çorum Basin (ÇÇB) (Fig. 1). ÇÇB, is one of the largest basins of Central Anatolia to have developed Oligocene to Quaternary terrestrial environments (Karadenizli, 2011). Kale Tepe-3 stratiı Formation. graphically fits within the upper levels of the Akkas¸dag This formation was first defined by Kazancı et al. (2005) and is composed of massive mudstones, gravelly sandstones, bedded

44

E. Albayrak / Quaternary International 445 (2017) 43e49

Fig. 1. The locality Kale Tepe-3. Map adapted from Google Earth.

limestones, laminated claystones and tuffs. The mammalian fauna ı tuffs is dated to MN12, i.e. latest Late Miocene from the Akkas¸dag (S¸en et al., 1998; Kazancı et al., 1999; S¸en, 2005). The upper levels of the formation yielded a small mammalian fauna at Balıs¸eyh, a locality close to Kale Tepe-3, dated to MN15, i.e. Early Pliocene, which places the formation in the Late Miocene-Early Pliocene (Karadenizli, 2011). 3. Materials and methods lı, Kırıkkale, a tusk In 2010, during road construction in Büyükyag and a humerus of a proboscidean were found in the new locality called Kale Tepe-3, followed in 2011 by a complete mandible. As these remains were unearthed during road construction, unfortunately some more remains were destroyed by workers unintentionally. Because the humerus is still in plaster jacket, it is not discussed in this paper. As the tusk and the humerus were found very close to the mandible in the very same level and no other remains were found at the same side of the road, it is considered that these remains belong to the same individual. In addition to elephantid remains, at the opposite side of the road at the same level, other remains such as fragments of an egg, fragments of turtle shell (most probably Mauremys sp., Evangelos Vlachos, 2015, pers. comm.), a carnivore canine, a cervid mandible, and other mammalian tooth fragments were found. These remains will be dealt with separately. Terminology follows Maglio (1973) for molar teeth and Van der Merwe et al. (1995) for the mandible. As all the tooth specimens in this study are lower molars, they are indicated by lower-case letters. Measurements of teeth were taken according to Beden (1979), Maglio (1973) and Lister (2012) and those of the mandible according to Todd (1997). A small plate at the front or back of a tooth that does not extend fully to the crown base but instead attaches to the adjacent plate is indicated with “x”. As the teeth are still buried in the jaw, lamellar frequency was taken near the top of the preserved crown, which may slightly underestimate the average value because of plate convergence (Maglio, 1973). And the height could not be measured because of the position of the teeth in the mandible. The X-ray image was taken in the museum with a portable X-ray machine. 4. Description 4.1. Mandible The mandible is almost complete, lacking the tip of the symphysis (Fig. 2). The symphysis is broken rostrally, but it is clear that it has a long, rostroventrally-oriented symphyseal process. Behind the posterior border of the symphysis, the ventral border of the

corpus is slightly concave. The corpus becomes dorsoventrally deeper toward the front of the mandible and is deepest at the level of anteriormost tooth, which is m1 in this specimen. The coronoid process extends rostrally to the level of the second plate of m3 (Fig. 2a). The dorsal border of the coronoid process is not far from the level of caput and there is only a slight concavity between them. On the lateral side of the ramus, the masseteric fossa is well defined and quite deep, while on the medial side, the pterygoid fossa is gently concave. The mandibular foramen is very large, elliptical in shape, and close to the posterior border of the mandibular ramus. It is a little higher than the level of the occlusal plane. There are two principal lateral mental foramina, as well as accessory foramina (Fig. 2a). The posterior mental foramen (PMF) is located behind the posterior border of the symphysis, in the middle of the corpus, below the first plate of m2. The anterior mental foramen (AMF) is located in front of the posterior border of the symphysis. The number of accessory foramina differs between the left and right halves of the jaw. On the left side there is one foramen, placed just at the level of the posterior border of the symphysis. On the right side there are two accessory foramina, placed between the anterior and posterior foramina, close to the anterior one. The accessory foramina are close to the dorsal border of the symphysis. There is no coronoid foramen or medial mental foramen. Measurements of the mandible are given in Table 1.

Table 1 Measurements of the mandible from Kale Tepe-3 (MTA Natural History Museum, Ankara no.71KT01). Measurements in round brackets are approximate, because of the broken parts. All measurements are in millimetres. Width of the whole mandible, anterior to toothrow 168.0 Width of the whole mandible at coronoid process 550.0 Length of the toothrow Left (350,0) Right (340.0) Corpus thickness (medio-lateral width) at anterior of first molar in toothrow Left 98.6 Right 89.5 Corpus thickness at the base of coronoid process Left 192.9 Right 187.4 Height of corpus at anterior alveolar border of first molar in toothrow Left 230.0 Right 243.0 Height of mandible at coronoid process Left 345.0 Right 335.0 Antero-posterior width of corpus at coronoid process Left 335.0 Right 345.0 Maximum length of mandible (710.0)


45

Fig. 2. The mandible (MTA Natural History Museum, Ankara, no. 71KT01) and tusk (MTA Natural History Museum, Ankara, no. 71KT02) of the elephantid from Kale Tepe-3. a. Lateral view of the mandible from the right side; b. Rostral view of the mandible; c. Occlusal view of the tooth row of the mandible; d. Lateral view of the tusk. Scale is 10 cm.

4.2. Teeth On the left side of the mandible three teeth are exposed at the occlusal plane, while two teeth are exposed on the right side (Fig. 2c). As the rearmost tooth is still partly unerupted in the jaw, an X-ray image of the mandible from the left side was taken to determine the form of this tooth. According to this image, the posterior tooth has 10 plates and anterior and posterior talonids in total, anterior talonid and 5 of the plates exposed on the occlusal plane, 5 buried in the jaw (Fig. 3). In the X-ray image, the tooth tapers strongly at the back in side view, as in a typical m3. Together, this suggests that the tooth is m3 with 10 plates. The posterior tooth extends far toward the back end of the jaw, but this can be seen in other elephantid mandibles, e.g. North American Mammuthus jaws figured by Osborn (1942, p. 1006, Figs. 892 and 893). Thus, in the Kale Tepe-3 specimen the three teeth exposed on the mandible are a small remnant of m1 in front, m2 in the middle, and m3 at the back. However, the m1 is seen only on the left side (Fig. 2c); the one on the right side was probably lost after the animal's death. The measurements of the teeth are given in Table 2. Table 2 Measurements of the teeth of 71KT01. Measurements in round brackets are considered not to represent original values due to wear. Measurements in square brackets are estimates of original values. Specimen

Number of plates

Length

Width

Lamellar frequency

Enamel thickness

m2 m2 m3 m3

∞3x [6e7] ∞4 [6e7] x6-[x10x] x5-[x10x]

147.8 (154.4) e e

95.5 95.8 95.2 96.0

4.39 4.12 3.85 3.69

4.18 4.07 e e

(L) (R) (L) (R)

All measurements are in milimetres.

All plates of m2 are at an advanced stage of wear, especially the anterior ones which are almost flat (Fig. 2c). Although the left m2 preserves ∞3x plates (three plates behind a dentine platform) and the right one ∞4x, according to the length of the crown the original number of plates is suggested to be about 6e7. Lamellar frequencies are very low (4.39 for the left, 4.12 for the right m2). Because of the advanced stage of wear, original crown height cannot be measured. The enamel pattern can be seen clearly in the last three plates; is the bands are very thick and their edges little-wrinkled, but they are intensively folded, especially at the center of the plate. Anterior and posterior expansions of the enamel are significant and the plates are in contact at that point. In the first plate from the back of the tooth, anterior to the posterior talonid, the enamel is more wrinkled, especially on the posterior side of the enamel loop.

Both left and right m3s are at a very early stage of wear, with the first three plates still divided into small enamel rings (Fig. 2c). According to the X-ray image, the m3 has 10 plates and anterior and posterior talonids. Because of the early stage of wear, it is not possible to see the enamel loop morphology. In the second and third plates, the enamel rings are equal, but with further wear in the first plate they have fused into unequal loops. In addition, accessory conules are seen between the plates. On the right m3, these accessory conules are seen between the second and third plates and the third and fourth plates. In the first plate, according to the shape of the first enamel ring from the medial side, there may also be an extra enamel ring which has joined with the anterior side of the enamel due to wear. The accessory conule between the second and third plates is situated medially and closer to the anterior side of the third plate. Similarly, the conule between the third and the fourth plates is situated at the center and closer to the anterior side of the fourth plate. On the left m3, there is only one accessory conule between the first and the second plates, on the anterior side of the second one. There are no posterior accessory conules. The tusk is almost complete, lacking the tip and the basal part (Fig. 2d). At the proximal end, the pulp cavity is visible. The length along the outside curve is 205 cm and the proximal circumference is 56 cm. It is not twisted as in Mammuthus but has a slight curve as seen in Elephas.

5. Comparison and discussion Distinction between early members of Elephas and Mammuthus is difficult and problematic, especially based on molar and mandible morphology. In cranial morphology they differ significantly (Maglio, 1973), but unfortunately, there are no cranial parts from Kale Tepe-3. In measurements such as plate number, enamel thickness and lamellar frequency of m2 and plate number of m3, the Kale Tepe3 specimen invites comparison with Elephas planifrons from Asia, early mammoths such as M. rumanus from Europe (MN16), and specimens from Bethlehem (Table 3). The number of plates of 71KT01's m3 fits with Lister and van Essen (2003) and Markov's (2012) count of M. rumanus (8e10). It would also fit Maglio's (1973) count for Elephas planifrons (8e13), but according to personal observation in NHM collection and Rabinovich and Lister (in this volume), Maglio's data includes worn specimens and the range of plate number for E. planifrons is between 11 and 15.

46


Table 3 Comparative measurements for second and third lower molars of primitive elephantids from Turkey, Bethlehem, E. planifrons from Asia and M. rumanus from Europe. aAlbayrak ret and Mayet (1923), fMaglio (1973), gRabinovich and Lister (in this and Lister (2012), bMarkov (2012), cRadulescu and Samson (1995), dLister and van Essen (2003), eDepe volume), hpers. obs. (in NHM collection). Measurements in round brackets are considered not to represent original values due to wear. Specimen Kale Tepe-3 m2 (L) m2 (R) m3 (L) m3 (R) €g üto € nü Yukarıso €g üto € nü)a m3 (1853, Yukarıso €g üto € nü)a m3 (2428, Yukarıso M. rumanus m3 (RRMH 594, Bossilkovtsi)b m3 (Tuluces¸ti)c m3 (Red Crag, Rendlesham)d m3 (Red Crag, Pontier)e m3 (Bethlehem, M18582)b E. planifrons m3f m2f m3g,h

Number of plates

Lamellar frequency

Hypsodonty index

Enamel thickness

(6e7) (6e7) x10x x10x

4.39 4.12 3.85 3.69

e e e e

4.18 4.07 e e

x11x 8x

3.86 4.77

(1.13) (1.13)

3.34 2.85

x10x e 10 10 x10x

3.75 3.0 (crown base) 4.93 4 3.75

e e 1.2 e e

4 e 3 4.0e5.0 4

8e13 9 11e15

3.2e5.8 3.7e5.4 e

0.83e1.15 1.07e1.41 0.97

2.2e5.2 3.0e4.0 e

The enamel pattern of 71KT01 seems somewhat different from that of Bethlehem specimen, M18582, and similar to those of E. planifrons from the Siwaliks. In the Bethlehem molars, the enamel bands are unfolded or only coarsely undulating. In the less worn plates the enamel rings are bigger and unequal (NHM collection, pers. obs.; see also Rabinovich and Lister, in this volume). In E. planifrons specimens from the Siwalik (Falconer and Cautley, 1845; pl. 8 Fig. 2, pl. 11 Figs. 2 and 6) conserved in the NHM collection (M18582, M36736, M3073, M3090, M3074, pers. obs.), the enamel is more wrinkled than Bethlehem specimens, especially on the median part of the plate. And in less worn plates of m3 of E. planifrons, the enamel rings are small and almost equal, as seen in m3s of 71KT01. In the m3s of 71KT01, anterior accessory conules are significant, especially in lamellae 2e4 of the right m3. In early Elephas, Sanders et al. (2010) describes E. ekorensis with anterior and larger posterior accessory conules throughout much of the crown of the molar plates. In E. planifrons from Siwalik collection (in NHM) although some specimens have clear posterior accessory conules, e.g. the m3s of M3073 from the same collection, anterior accessory conules are more common (pers. obs. in NHM collection; Rabinovich and Lister, in this volume). In early Mammuthus and Bethlehem specimens this feature is more consistent. In the m3s of the mandibles (M18582, M18586) and in M3 M18524 from Bethlehem collection, the posterior accessory conules are clear (pers. obs. in NHM collection; Rabinovich and Lister, in this volume). Also in M. subplanifrons posterior accessory conules predominate (Sanders, tes¸ti (Romania) 2007) and in the M. rumanus molars from Cerna (Lister and van Essen, 2003), the Red Crag (England) (Lister and van Essen, 2004), Montopoli (Italy) (Palombo and Ferretti, 2005) and Bossilovtsi (Bulgaria) (Markov and Spassov, 2003) posterior accessory conules are prominent. Accordingly, the molars of Kale Tepe-3 mandible show resemblance to primitive Elephas, such as E. planifrons with the presence of anterior accessory conules in the m3s. Another primitive elephantid finding from Europe was discussed by Kostopoulos and Koulidou (2015). The specimen is a right maxilla with the last two molars. According to primitive features of the molars and maxilla, the specimen was ascribed to a primitive Eurasian mammoth species, more advanced than the “Hadar type” of mammoth from Ethiopia and less than Pleistocene M. meridionalis from Eurasia (Kostopoulos and Koulidou, 2015). This specimen is close to Kale Tepe-3 specimen in respect to age. But the Greek specimen is closer to Mammuthus according to its molar, and

especially cranial, features. Albayrak and Lister (2012) discussed the primitive molars with low plate number and lamellar frequency and thick enamel from €g üto €nü locality in Turkey. Although the wear patterns the Yukarıso and metrics of these specimens were similar to both E. planifrons and M. meridionalis, higher hypsodonty index suggested the identification as primitive M. meridionalis, i.e. an early mammoth, morphologically intermediate between typical M. rumanus and M. meridionalis. By contrast, the lower plate number of 71KT01 m3s preclude its attribution to that species. And also according to enamel thickness of m3, 71KT01 seems more primitive than €g üto €nü specimens (Table 3). Yukarıso The tusk of 71KT01 is not twisted significantly, but has a slight curve. In Elephas species, the tusk is gently curved and not twisted. On the other hand, in known Mammuthus species, the tusk is always both curved and spirally twisted (Maglio, 1973). So according to the tusk morphology, the Kale Tepe-3 specimen resembles Elephas. The mandible of Kale Tepe-3 has some primitive features discussed in Ferretti and Debruyne (2011). It does not have a linguoid process (LPR) on the mandibular foramen. The large and medially protruding spine-shaped LPR on the anterior border of the mandibular foramen has been observed in specimens of Loxodonta africana, E. maximus and M. primigenius. But data on this feature are not sufficient to draw any definitive conclusions about its systematic value (Ferretti and Debruyne, 2011). In primitive representatives of Loxodonta (L. adaurora) and Elephas (E. nawataensis), and also in M. subplanifrons from Langebaanweg, there are two lateral mental foramina (Maglio and Ricca, 1978; Tassy, 2003; Ferretti and Debruyne, 2011). In more advanced representatives of elephantids, e.g. M. meridionalis, M. primigenius and E. maximus, the number of lateral mental foramina typically ranges from two to three, although some individuals show as many as seven (Debruyne, 2003; Ferretti and Debruyne, 2011). With two lateral mental foramina, 71KT01 resembles primitive specimens, but this is not a definite diagnostic feature. The medial mental foramen (MMF) is a feature that has been observed in proboscideans only amongst modern-type elephantines (Ferretti and Debruyne, 2011). The single known mandibles of M. subplanifrons and M. rumanus, respectively, do not have MMF. In Mammuthus species, MMF has been observed only in M. meridionalis, M. trogontherii, M. primigenius, and M. columbi. In primitive species of Elephas, such as E. ekorensis and E. planifrons, MMF remains unknown; it has been observed only in the extant


47

In 71KT01, the corpus of the mandible becomes dorsoventrally deeper toward the front of the mandible and is deepest at the level of the anteriormost tooth. The same feature is seen in the E. planifrons specimens from the Siwaliks in Falconer and Cautley (1845, pl. 8 Fig. 2a) and in NHM collection (M36736, M3090, pers. obs.). In the Bethlehem mandible, by contrast, the corpus is not deeper toward the front of the mandible. As primitive species of Elephas and Mammuthus have similar features in their molar teeth and mandibles, it is not always possible to distinguish them using these parts, leading to unresolved questions on the early evolution of these genera. It is very obvious that 71KT01 is a primitive elephantid with low plate number and lamellar frequency, and thick enamel, similar to E. planifrons and M. rumanus. To sum up the comparisons: - The low plate number of m3s (10) of 71KT01 is similar to those of M. rumanus (8e10) and close to E. planifrons (11e15). - Small and almost equal enamel rings in less worn plates of m3s of 71KT01 is similar to those of E. planifrons. In Bethlehem specimens, the enamel rings in less worn plates are bigger and unequal. - In the m3s of 71KT01 anterior accessory conules are clear. In E. planifrons, they are more common whereas in M. rumanus and Bethlehem specimens posterior conules are prominent. - Although distinction between early representatives of Elephas and Mammuthus based the morphology of mandibles is not straightforward, the slight concavity of the ventral margin of the mandible of 71KT01 and dorsoventrally deeper corpus toward the front of the mandible are the features similar to those of E. planifrons. - Untwisted tusk 71KT02 resembles Elephas rather that Mammuthus. Although the tusks of M. rumanus are unknown, in Mammuthus the tusk is spirally twisted and curved. As the tusk and the mandible of Kale Tepe-3 most probably belong to the same individual, as they were found very close to each other in the field without any other remains closer, the shape of the tusk strongly supports the tendency to favour Elephas. Eventually, the mandible from Kale Tepe-3 can be attributed to E. planifrons from Siwaliks rather than primitive Mammuthus type in metric and morphological features. Furthermore, if these specimens belong to the same individual it is plausible to attribute them to primitive Elephas such as E. planifrons. But even without the Elephas type tusk, 71KT01 is more similar to E. planifrons. Fig. 3. Mandible of the elephantid from Kale-Tepe 3; a, Occlusal view of the left tooth row; b, Lateral view of the mandible from the left side; c, X-ray image of the mandible from the left side, numbers and x indicate the numbers of the plates and talonid. Scale is 10 cm.

species E. maximus with lower incidence than Mammuthus, being about 30 per cent (Ferretti and Debruyne, 2011). The absence of MMF in 71KT01 is a feature usually seen in primitive specimens. In sum, while primitive elephantids have different features in their mandibular foramina than advanced species, there are no definite features to distinguish primitive species of Elephas and Mammuthus as they share similar primitive characteristics (Ferretti and Debruyne, 2011). Comparing the mandible of 71KT01 with those of other primitive elephantids, the ventral margin of the corpus is only slightly concave, as in the mandible of Elephas planifrons from the Siwalik Hills in Falconer and Cautley (1845, pl. 8 Fig. 2a, pl. 11, Fig. 3) in NHM collection (M36736, M3090, pers. obs.) and M. rumanus from Bossilkovtsi. In Bethlehem specimen, M18582, however, the concavity in the ventral margin of the corpus is more pronounced.

6. Conclusion Maglio (1973) suggested that E. planifrons was a strictly Asiatic form and put the European specimens referred to this species to early stages of M. meridionalis. As the Anatolian part of Turkey is in Asia, it is plausible to expect Asiatic forms of elephantids there. Maglio also followed Hooijer (1958) in referring the Bethlehem material to E. cf. planifrons. If this is correct, it could fill the morphological and geographical gap between E. ekorensis of Africa and E. planifrons of Asia (Maglio, 1973). But there are still discussions about the identification of Bethlehem specimens (Markov, 2012); according to Rabinovich and Lister (in this volume), the Bethlehem elephants are more primitive than either of these Elephas species. The Kale Tepe-3 mandible is obviously very primitive by Eurasian standards, with low plate number and lamellar frequency, thick enamel and some primitive features of the mandibular foramina. Although molars of primitive elephantid species are very difficult to distinguish, the molar teeth of 71KT01 seems more like E. planifrons of Asia in its molar measurements and features, as well

48


as in the form of its tusk and mandible. As the tusk and the mandible presumably belong to the same individual, the Elephas shape of the tusk makes this identification very reasonable. Together with the primitive elephantids from Bethlehem (Rabinovich and Lister, in this volume) and Tsotylion, Greece (Kostopoulos and Koulidou, 2015), the remains from Kale Tepe-3 represent the most primitive elephantid out of Africa. Kostopoulos and Koulidou (2015) ascribe the Tsotylion maxilla to a primitive Eurasian mammoth species, more advanced than the “Hadar type” specimen and less than Mammuthus africanavus from North Africa or Pleistocene M. meridionalis from Eurasia. In Bethlehem specimens the situation is more complex. The generic attribution of these specimens are still under debate. The cooccurrence of two different taxa, primitive Mammuthus and Elephas, at Bethlehem is one of the interpretations (Rabinovich and Lister, in this volume). Hopefully, with the discovery of more primitive elephantid remains, the relations between all these primitive specimens will be able to be enlightened. There are many debates and questions concerning early elephantids. New discoveries of primitive specimens may help to solve these problems. Because of its key position between Africa, Asia and Europe, Turkey has potential importance for elucidating the history of early Eurasian elephantids. The primitive mandible from Kale Tepe-3 is a significant piece in this puzzle, a step forward in understanding the migration and early evolution of elephants. If 71KT01 is E. planifrons or its immediate ancestor, this means that this is the most westerly specimen of this lineage. But if it is a primitive representative of Mammuthus, then it would equally be an important specimen to understand the early evolution of mammoths outside of Africa. Acknowledgements I am very grateful to Adrian Lister for sharing his precious comments and criticisms about these specimens, Georgi Markov for sharing the photos of Bossilkovtsi mandible and for valuable discussions in the VI International Conference on Mammoths and their Relatives with Dimitris Kostopoulos and Evangelos Vlachos for his comments on turtle's remains. I would like to thank S¸evket S¸en, € Pierre-Olivier Antoine, Maeva Orliac, Cengiz Oztürk, Cenk Karaca and Mustafa Koç for their great efforts in the field, Yusuf S¸en for all his help in taking the X-ray image and Volkan Sevinç for his great efforts in restoration. The field and laboratory studies were supported by MTA project no. 2011-08-16-01 and the collection study in NHM received support from the Synthesys Project http://www. synthesys.info/ which is financed by European Community Research Infrastructure Action under the FP7 “Capacities” Program (GB-TAF-5352). The manuscript could not be improved without the helpful comments from Adrian Lister, an anonymous reviewer and the guest editor, Athanassios Athanassiou. References €ne Elefanten-Funde im Umland von Erzurum in Adam, K.D., 1988. Über Pleistoza Ostanatolian, Ein Beitrag zur Namengebung von Elephas armeniacus und Elephas trogontherii. Stuttgarter Beitr€ age zur Naturkunde, Serie B (Geologie und €ontologie) 146, 1e89. Pala Albayrak, E., Lister, A.M., 2012. Dental remains of fossil elephants from Turkey. Quaternary International 276e277, 198e211. Baigusheva, V.S., Titov, V.V., Foronova, I.V., 2016. Teeth of early generations of Early Pleistocene elephants (Mammalia, Elephantidae) from Sinyaya Balka/Bogatyri site (Sea of Azov Region, Russia). Quaternary International 420, 306e318. Baygusheva, V., Titov, V., 2012. The evolution of eastern meridionaloid elephants' dental characteristics. Quaternary International 255, 206e216. Becker-Platen, J.D., Sickenberg, O., 1968. Die unterpleistoz€ anan Kiese von Eskis¸ehir (Anatolien) und ihre S€ augetierfauna. Mitteilungen aus dem Geologischen Institut der Technischen Universit€ at Hannover 8, 7e20. Beden, M., 1979. Les Elephants (Loxodonta et Elephas) d’Afrique Orientale:

matique, Phyloge nie, Inte re ^t Biochronologique. Unpublished The se DocSyste des Sciences des L'Universite de Poitiers. torat des Sciences. Faculte res du Boulbes, N., Mayda, S., Titov, V.V., Alçiçek, M.C., 2014. Les grand mammife rieur des travertins du Bassin de Denizli (Sud-Ouest AnaVillafranchien supe tolie, Turquie). L’Antropologie 118, 44e73. €ne Elefanten-Funde im Umland von Erzurum in Dayan, E., 1989. Über Pleistoza Ostanatolian, Ein Beitrag zur Fundgeschichte des Elephas trogontherii im Becken von Pasinler. Stuttgarter Beitr€ age zur Naturkunde, Serie B (Geologie und Pal€ aontologie) 147, 1e21. renciation morphologique et mole culaire des Elephantinae Debruyne, R., 2003. Diffe um National d’Histoire Naturelle, (Mammalia, Proboscidea). Ph.D. Thesis. Muse Paris. ret, C., Mayet, L., 1923. Les e le phants Plioce nes. Part II. Annales de l’Universite Depe de Lyon, n.s. I 42 (II), 89e213. Falconer, H., 1857. On the species of Mastodon and elephant occurring in the fossil state in Great Britain, Part I: Mastodon. Quarterly Journal of the Geological Society of London 13, 307e360. Falconer, H., Cautley, P.T., 1845. Fauna Antiqua Sivalensis, Being the Fossil Zoology of the Sewalik Hills in the North of India. London. Smith, Elder and Co, London. Ferretti, M.P., Debruyne, R., 2011. Anatomy and phylogenetic value of the mandibular and coronoid canals and their associated foramina in proboscideans (Mammalia). Zoological Journal of the Linnean Society 161, 391e413. Hooijer, D.A., 1958. An early Pleistocene mammalian fauna from Bethlehem. Bulletin of the British Museum of Natural History, London, Geology 3 (8), 267e292. Karadenizli, L., 2011. Oligocene to Pliocene palaeogeographic evolution of the Çankırı-Çorum Basin, central Anatolia, Turkey. Sedimentary Geology 237, 1e29. lu, G., de Bonis, L., Bouvrain, G., Araz, H., Varol, B., Kazancı, N., S¸en, S¸., Seyitog Karadenizli, L., 1999. Geology of a new late Miocene mammal locality in Central Anatolia, Turkey. Comptes Rendus de l’Academie des Sciences - Series IIA Earth and Planetary Science 329 (7), 503e510. lu, G., S¸en, S¸., Alçiçek, M.C., Varol, B., Saraç, G., Kazancı, N., Karadenizli, L., Seyitog Hakyemez, Y., 2005. Stratigraphy and sedimentology of Neogene mammal ı area, Turkey. Geodiversitas 27 (4), 527e551. bearing deposits in the Akkas¸dag Kostopoulos, D.S., Koulidou, I., 2015. An early mammoth maxilla from northwestern Greece. Quaternary International 379, 155e163. Lister, A.M., 1996. Evolution and taxonomy of Eurasian mammoths. In: Shoshani, J., Tassy, P. (Eds.), The Proboscidea: Evolution and Palaeoecology of Elephants and Their Relatives. Oxford University Press, Oxford, pp. 203e213. Lister, A.M., 2012. Quantitative analysis of mammoth remains from Lynford, Norfolk, England. In: Boismier, W.A., Gamble, C., Coward, F. (Eds.), Neanderthals Among Mammoths, Excavations at Lynford Quarry. English Heritage, Norfolk, pp. 205e214. Lister, A.M., Sher, A.V., 2001. The origin and evolution of the woolly mammoth. Science 294, 1094e1097. nescu), the earliest Lister, A.M., van Essen, 2003. Mammuthus rumanus (S¸tefa mammoth in Europe. In: Petulescu, A., Stiuc a, E. (Eds.), Advances in Vertebrate ” Institute of Speleology, Bucharest, Paleontology “Hen to Panta”. “Emil Racovit¸a pp. 46e52. Lister, A., van Essen, H., 2004. The earliest mammoths in Europe. In: 18th International Senckenberg Conference: Late Neogene and Quaternary Biodiversity and Evolution: Regional Developments and Interregional Correlations, Weimar, Germany, vol. 2004, pp. 152e154. Lister, A.M., Sher, A.V., van Essen, H., Wei, G., 2005. The pattern and process of mammoth evolution in Eurasia. Quaternary International 126e128, 49e64. Maglio, V.J., 1973. Origin and evolution of the Elephantidae. Transactions of the American Philosophical Society 63 (3), 1e149. Maglio, V.J., Ricca, A.B., 1978. Dental and skeletal morphology of the earliest elephants. Verhandelingen der Koninklijke Nederlandse Akademie van Wetenschappen, Afd. Natuurkunde 29, 1e51. Markov, G.N., 2012. Mammuthus rumanus, early mammoths, and migration out of Africa: some interrelated problems. Quaternary International 276e277, 23e26. Markov, G.N., Spassov, N., 2003. Primitive mammoths from Northeast Bulgaria in the context of the earliest mammoth migrations in Europe. Advances in Vertebrate Paleontology “Hen to Panta” 53e58. Osborn, H.F., 1942. Proboscidea: a monograph of the discovery, evolution, migration and extinction of the mastodonts and elephants of the world. In: Stegodontoidea, Elephantoidea, vol. II. The American Museum Press, New York. Palombo, M.R., Ferretti, M., 2005. Elephant fossil record from Italy: knowledge, problems, and perspectives. Quaternary International 126e128, 107e136. Rabinovich, R., Lister, A.M. (in this volume), The earliest elephants out of Africa: taxonomy and taphonomy of proboscidean remains from Bethlehem, Quaternary International. Radulescu, C., Samson, P.M., 1995. Review of the Villafranchian s.s. faunas of Romania. Il. Quaternario 8, 377e382. Sanders, W.J., 2007. Taxonomic review of fossil Proboscidea (Mammalia) from Langebaanweg, South Africa. Transactions of the Royal Society of South Africa 62 (1), 1e16. Sanders, W.J., Gheerbrant, E., Harris, J., Saegusa, H., Delmer, C., 2010. Proboscidea. In: Werdelin, L., Sanders, W.J. (Eds.), Cenozoic Mammals of Africa. University of California Press, Berkeley, pp. 161e251. Sickenberg, O., Becker-Platen, J., Benda, L., Berg, D., Engesser, B., Gaziry, W., Heissig, K., Hünermann, K.A., Sondaar, P.Y., Schmidt-Kittler, N., Staesche, U., €heren Jungtertia €rs und Steffens, P., Tobien, H., 1975. Die Gliederung des ho Altquart€ ars in der Türkei nach Vertebraten und ihre Bedeutung für die

E. Albayrak / Quaternary International 445 (2017) 43e49 internationale NeogenStratigraphie. Geologisches Jahrbuch, Reihe B 15, 1e167. ı, Late Miocene of Central AnaGeology, Mammals and Environments at Akkas¸dag tolia. In: S¸en, S¸. (Ed.), Geodiversitas 27 (4). lu, G., Karadenizli, L., Kazancı, N., Varol, B., Araz, H., 1998. Mammal S¸en, S¸., Seyitog biochronology of Neogene deposits and its correlation with lithostratigraphy in Çankırı-Çorum Basin, Central Anatolia-Turkey. Eclogae Geologicae Helvetiae 91, 307e320. S¸enyürek, M., 1960. Ankara Üniversitesi’nde muhafaza edilen fosil fil kalıntılarına dair bir not. Türk Tarih Kurumu Belleten XXIV 96, 693e695. an. Türk Tarih S¸enyürek, M., 1961. The molar of an Archidiskodon from Akdog

49

Kurumu Belleten XXV 99, 339e353. Tassy, P., 2003. Elephantoidea from Lothagam. In: Leakey, M.G., Harris, J.M. (Eds.), Lothagam: the Dawn of the Humanity in Eastern Africa. Columbia University Press, New York, pp. 331e358. Todd, N., 1997. Comparison of Variation in Cranial and Dental Morphology of Elephas recki and the Extant Elephants: Implications for Phylogeny and Macroevolution. Ph.D. Thesis. George Washington University, Washington, DC. Van der Merwe, N.J., Bezuidenhout, A.J., Seegers, C.D., 1995. The skull and mandible of the African elephant (Loxodonta africana). Onderstepoort Journal of Veterinary Research 62, 245e260.