Tetralophodon atticus ( Wagner, 1857 )

Geraads, Denis, Spassov, Nikolaï, Hristova, Latinka, Markov, Georgi N. & Tzankov, Tzanko, 2011, Upper Miocene mammals from Strumyani, South-Western Bulgaria, Geodiversitas 33 (3), pp. 451-484 : 455-458

? Tetralophodon atticus ( Wagner, 1857)

Four deciduous teeth at various stages of preservation demonstrate the presence of bunodont tetralophodont elephantoids in Str-2. Their measurements are given in Table 1.

FM-2187 is a relatively well-preserved DP2 sin. ( Fig. 3A ; the second loph is damaged on the posttrite side). The tooth is bilophodont, with strongly developed anterior and posterior cingula. The first loph consists of a large paracone and a very weak protocone fused with the paracone. The second loph is wider than the first, and is separated from it by a narrow interloph that, despite the presence of weak crests on the posterior wall of the paracone and the anterior of the hypocone, remains open. The valley between the second loph and the welldeveloped posterior cingulum is blocked by a cusp situated behind the median sulcus.

FM-2798 ( Fig. 3C ) is a DP3 dext. Although parts of the crown are missing, those that are preserved provide a fairly good idea of the tooth size and shape.It is trilophodont, with a strong anterior cingulum,a part of which is preserved on the pretrite side, and a posterior cingulum closely pressed to the third loph.

FM-2797is a DP3 dext.posterior fragment.Compared to FM-2798, the posterior cingulum is more extended buccally, and more clearly separated from the posterior loph,lacking the multicuspid connection to the pretrite semiloph observable in FM-2798.The pre- and posttrite halves of the posterior loph are of equal height, aligned and separated from the penultimate loph by a broad interloph. A strong cingulum is present on the linguo-distal part of the crown.

FM-2799 ( Fig. 3B ) is a right dp2 bearing little to no trace of use. The crown is mostly intact, with a fragment on the lingual side (between the first and second posttrite cusps) missing.The tooth is bilophodont, with a posterior cingulum clearly separated from the second lophid. The anterior cingulum is strongly reduced. The first loph, which is nearly twice as high as the second one, consists of a strongly developed protoconid and a low metaconid closely attached to it. On the second lophid, the entoconid and the hypoconid are clearly separated from each other, and the interlophid is free on the pretrite side. As far as can be judged from the preserved part of the second posttrite, the interlophid was blocked on the posttrite side, with a weak cross-contact between the first pretrite and the second posttrite cusps. On both the pretrite and the posttrite sides,the cusps of the second lophid are connected with the posterior cingulum by an additional posterior cuspid on each side.

Morphologically, as well as metrically ( Fig. 4 ), the DP2 from Str-2 is very similar to those in the two DP2-DP3 tooth rows of Tetralophodon atticus from Pikermi described by Wagner (1857; type specimen of T. atticus ) and Gaudry(1862 -1867).It differs from the DP2s of the earlier T. longirostris (Kaup, 1832) in the strongly reduced protocone and from the later Anancus arvernensis (Croizet & Jobert, 1828) in its larger size (see Metz-Muller 1996, 2000). Presence of a well-developed posterior cingulum in FM-2187 sets it apart from DP2s of A. kenyensis (MacInnes, 1942) from the Middle Awash described by Saegusa & Haile-Selassie (2009). The Turolian species T. atticus is known from various localities in Europe and west Asia, including Bulgaria ( Markov 2008 ) and attribution of FM-2187 to that taxon seems plausible. It must be noted, however, that morphology of DP2 (and several other teeth, including most of the deciduous premolars) is not known for “ Mastodon ” grandincisivus Schlesinger, 1917, an amebelodontid (see Tassy 1985, 1999, 2005) that not only developed tetralophodont intermediary teeth in parallel with tetralophodont gomphotheres but also has a stratigraphical and geographical distribution largely overlapping with those of T. atticus . Thus, except for the two specimens from Pikermi mentioned above, the status of all available material from Strumyani as well as the comparable finds from the Turolian of Europe and west Asia listed below, is somewhat ambiguous, as they might potentially belong to T. atticus or “ M.” grandincisivus.

Comparable material includes deciduous teeth from the locality Taraklia in Moldova referred by Khomenko (1914) to Tetralophodon longirostris and identified as T. atticus by Tassy (1985), followed by Metz-Muller (1996, 2000). Additional finds from Taraklia described by Riabinin (1929), and teeth from Grebeniki ( Ukraine) published by Burchak-Abramovich (1940), seem to belong to T. atticus too ( Markov 2008 ). At Hadjidimovo in Bulgaria, T. atticus seems to have co-occurred with “ M.” grandincisivus ( Markov 2004 , 2008). There is also a DP2-DP4 tooth row from Pikermi referred by Marinos & Symeonidis (1974) to Ch. pentelici (Gaudry & Lartet, 1856) and identified as a tetralophodont elephantoid (i.e. either T. atticus or “ M.” grandincisivus) by Markov(2004) ; unfortunately, no measurements are provided by the authors.

Metrically, the DP3 from Str-2, FM-2798, clearly surpasses the teeth of T. longirostris from Eppelsheim and Laaerberg, falling within the variation range of material from several Turolian localities in Europe and west Asia ( Fig. 5). In addition to Pikermi, Taraklia, Grebeniki and Hadjidimovo, these include AkkaşdaĞI in Turkey ( Tassy 2005), Cimişlia in Moldova (material described by Simionescu & Barbu [1939] as T. longirostris and referred to T. atticus by Markov [2008]), Kalimantsi in Bulgaria ( Markov 2004 , 2008), and Maragheh (material published by Mecquenem [1924] as Ch. pentelici and identified as T. atticus by Markov [2008]). Again, except for the Pikermi specimens described by Wagner (1857) and Gaudry (1862 -1867), the rest of the material might be heterogeneous and include teeth of “ M.” grandincisivus. The alignment of the halves of the preserved loph in FM-2797 is a derived trait observed in T. atticus , yet the state of this character is unknown in “ M.” grandincisivus.

As for the lower dp2, FM-2799, it is readily distinguished from dp2s of Anancus arvernensis analyzed by Metz-Muller (1996, 2000), a dp2 from the Upper Pliocene of Ahl al Oughlam ( Morocco) referred by Geraads & Metz-Muller (1999) to Anancus cf. osiris Arambourg, 1946, and A. kenyensis dp2s from Middle Awash ( Ethiopia) described by Saegusa & Haile- Selassie (2009). Larger than all these anancine teeth, it further differs from A. kenyensis and A. arvernensis in the development of the posterior cingulum, and from the Ahl al Oughlam specimen in the incomplete fusion of the meta- and protoconid. On the other hand, the only positively identifiable dp2 of “ M.” grandincisivus (in the unpublished mandible from Hadjidimovo, NMNHAs HD-38029) has a meta- and paraconid that are clearly set apart from each other and do not differ in height as drastically as in FM-2799. Interestingly, the dp2 from Taraklia published by Khomenko (1914) and referred to T. atticus by Tassy (1985) and Metz-Muller (1996), has a meta- and paraconid less closely attached to each other than in the Strumyani dp2, and is metrically closer to the dp2 of “ M.” grandincisivus from Hadjidimovo than to FM-2799. On the other hand, a dp2 from Csákvár in Hungary (HGI V.11393: see Gasparik 2001: 70) is metrically close to the dp2s from Taraklia and Hadjidimovo (GM pers. obs. HGI, 2005) but quite similar to the Str-2 dp 2 in the closely attached meta- and paraconid.

Rather than casting doubt on the attribution of the Taraklia dp2 to T. atticus , this example aptly demonstrates the difficulties in identifying isolated deciduous teeth as either T. atticus or “ M.” grandincisivus, two poorly known, co-occurring taxa with homoplastic characters.

Str-1 yielded only a few fragmentary post-cranial proboscidean remains.

Order PERISSODACTYLA Owen, 1848

Figures

FIG. 3. — Fossil mammals from the Miocene of Strumyani (SW Bulgaria): A-C,?Tetralophodon atticus (Wagner,1857); A, DP2 FM-2027, occlusal view;B, dp2 FM-2799, lingual view,C, DP3 FM-2798,occlusal view; D, Dihoplus pikermiensis (Toula, 1906), dorsal view of symphysis FM-2470;E, Microstonyx major (Gervais,1848),maxilla FM-2801;F, G, Dorcatherium cf.puyhauberti Arambourg & Piveteau,1929, mandible FM-2741 in occlusal (F)and lateral (G) views;H, Palaeotragus rouenii Gaudry,1861,mandible FM-2778;I, Miotragocerus (Pikermicerus) gaudryi (Kretzoi,1941), mandible FM-2103; J, Cervidae gen.et sp. indet., mandible FM-2100. Scale bar: A-C,F-J, 7.5 cm; D, 30 cm; E, 10 cm.

FIG. 4. — Bivariate plots (L × W) of DP2 in tetralophodont elephantoids from Strumyani and other Late Miocene sites. Data: GM (own measurements MNHN, NHM, NHMW, NMNHAs) and sources quoted in the text. Abbreviation: T., Tetralophodon.