Crete

Sen, Sevket, 2017, A Review Of The Pleistocene Dwarfed Elephants From The Aegean Islands, And Their Paleogeographic Context, Fossil Imprint 73 (1 - 2), pp. 76-92 : 78-79

Crete

Dwarf elephants are best known from Crete , with three species being distinguished: Mammuthus creticus ( BATE, 1907) , Palaeoloxodon creutzburgi ( KUSS, 1965) and P. chaniensis ( SYMEONIDIS, THEODOROU et GIANNOPOULOS, 2000) (see below).

Bate (1907) described a small species of elephant, that she named Elephas creticus , from the Early-Middle Pleistocene site of Cape Melekas in northwestern Crete . The type material consists of “nine imperfect molars and a few fragments, among which are a portion of an incisor and the dorsal half of a vertebra” (p. 239). Bate (1907: 243) noted that “except with regard to the immense difference in size the characteristics of the molars (from Cape Melekas), more especially in the lowness of the crowns, appear to resemble more closely those of Elephas meridionalis than any other of the larger Elephants of the Mediterranean region”. At the same site, P. Y. Sondaar discovered a lower molar together with a lower jaw of the murid Kritimys kiridus (BATE, 1942) ( Mol et al. 1996, Van der Geer and Lyras 2011). Herridge and Lister (2012) reported additional specimens from this site. All these authors reconsidered the systematic status of this dwarf elephant and confirmed its affinities with Mammuthus meridionalis (NESTI, 1825). M. creticus has molars with a reduced number of lamellae, low crown, wide occlusal surface, broad cement areas, and simple enamel bands. Poulakakis et al. (2006) studied ancient DNA (aDNA) from a rib bone fragment from Cape Melekas which evidenced the affinities of this small elephant with Mammuthus BROOKES, 1828 . However, Binladen et al. (2007) and Orlando et al. (2007) considered that the 43 bp sequence used by Poulakakis et al. (2006) was not the most appropriate for determining the generic status of the Cretan elephant. Lastly, Herridge and Lister (2012) studied the Bate’s collection housed at the Natural History Museum in London, plus some new fossils, and confirmed the morphological evidence suggesting some ancestral relationship between M. creticus and the Early Pleistocene M. meridionalis or the Late Pliocene M. rumanus (ŞTEFĂNESCU, 1924). These authors estimated that M. creticus may have had an average body mass of about 310 kg and shoulder height of 1.13 m (see also Poulakakis et al. 2002, Lomolino et al. 2013). M. creticus was found associated with the murid Kritimys kiridus, and they apparently reached Crete during the first migration wave of land mammals during the Early Pleistocene ( Mayhew 1977, 1996), i.e. ca. 0.9 My ago ( Reese et al. 1996, Sondaar and Van der Geer 2005, Van der Geer et al. 2010, Van der Geer and Lyras 2011). However, Stuart and Lister (2012) claimed that this date is not accurate and needs to be checked using new dating methods. According to Dermitzakis and De Vos (1987: 400), M. creticus “became extinct during higher sea level stand of an interglacial, when Crete supposedly became subdivided into different smaller islands, not large enough to support elephant populations”. This period could have been the Günz-Mindel interglacial which lasted between ca. 0.75–1.0 Ma.

The second elephant species from Crete is P. creutzburgi ( KUSS, 1965) ( Text-fig. 2). Its type locality Kalo Chorafi in northern Crete yielded three cheek teeth (P2, P3 and m1; Text-fig. 2c) and some tooth and bone fragments ( Kuss 1965). Poulakakis et al. (2002) enumerated 22 localities in Crete , all supposedly of latest Middle or Late Pleistocene age, that yielded remains of this palaeoloxodontine elephant. The material from most localities is fragmentary and not adequately studied. The best fossils referred to this species from Crete are, in addition to the type material from Kalo Chorafi cave, a damaged mandible of an old individual, an atlas and a femur from Krida Avlaki ( Simonelli 1908, Palombo and Petronio 1989), a damaged mandible with m2 and m3 from Kharoumes Bay ( Bate 1907), an m3 from Simonelli Cave ( Text-fig. 2e; Kotsakis 1980) and an m2 from Katharo plain ( Poulakakis et al. 2002), all referred to P. creutzburgi by the last authors. This is a large sized form having palaeoloxodontine characters in its dentition: straight or gently curved tusks, narrow lamellae separated by cement areas that are narrower than the plates, gently folded enamel, anterior and/or posterior enamel bulges on the plates. Using the humerus length (790 mm) and the femur length (950 mm), Lomolino et al. (2013) estimated the body mass of this species as about 3,000 kg, i.e., half the weight of the European straight-tusked elephant P. antiquus (FALCONER et CAUTLEY, 1847), from which it probably derived.

Belluomini and Delitala (1983) dated an elephant femur from the Simonelli Cave in Crete to 49,000 ± 20% BP using the Amino Acid Racemization method. This femur was determined as belonging to P. creutzburgi (Palombo and Petronio 1989; see also Reese et al. 1996). No other chronological data are available for the localities that yielded P. creutzburgi, which are tentatively correlated to the Late Pleistocene.

Several authors mentioned the presence of a third species of elephant in Crete which would have been larger than P. creutzburgi from the Kalo Chorafi Cave and about 20% smaller than the mainland species P. antiquus. This form was initially referred to P. antiquus or P. cf. antiquus ( Kuss 1973, Dermitzakis and Sondaar 1979, Symeonidis and Theodorou 1982, Dermitzakis and De Vos 1987, Mol et al. 1996). Symeonidis et al. (2000) recorded 41 well-preserved limb bones from a submerged cave close to Vamos in northwestern Crete , and they erected a new species Elephas chaniensis based on this material (see also Symeonidis et al. 2001, Poulakakis et al. 2002). They also referred to this species a well-preserved complete lower jaw bearing the m3s from Coumbes Cave near Rethymnon ( Text-fig. 2b) that Symeonidis and Theodorou (1982) initially identified as “a new endemic Pleistocene subspecies of P. antiquus ”. Symeonidis et al. (2000: 106) diagnosed this species as “Endemic elephant of relatively large dimensions about 20% smaller than continental P. antiquus ”.

The cheek teeth are the most diagnostic elements for species distinction. As it can be seen in Tables 1 and 2, the number of teeth referred to P. creutzburgi and E. chaniensis is still too limited to discriminate between the two different palaeoloxodontine species in Crete . In all these specimens, the plates are narrow, the interplate cement areas are narrower than the plates, the central part of the plates is a little larger (mesio-distal) than the lateral parts, the plates bear anterior and/or posterior enamel bulges in their central part, the enamel is gently folded but much less so than in the teeth of P. antiquus in a similar stage of wear, the slightly abraded plates are composed of one wide central loop and two or more lateral ellipsoid or rounded rings, and the lamellar frequency is quite similar. Consequently, Poulakakis et al. (2002) suggested the synonymy of the large elephant from Vamos Cave and Coumbes Cave with P. creutzburgi. Fieldwork carried out during the last two decades did not provide any additional fossil material, except for “a part of deciduous molar of a dwarf elephantid” from the Koutalas Cave (Chania) that Iliopoulos et al. (2010: 5) referred to Elephas sp. The systematic status of this large elephant from Crete was not discussed in later studies dealing with the Crete elephants, such as by Van der Geer and Lyras (2011), Herridge and Lister (2012), Lomolino et al. (2013) and Van der Geer et al. (2014, 2016). Consequently, we suggest to maintain the species distinction of “ E. chaniensis ” pending a systematic review of the Late Pleistocene elephants of Crete .

In summary, at least two, probably three species of elephants inhabited Crete during the Pleistocene. Taking into consideration their systematic affinities and their degree of dwarfism, it appears that elephants reached Crete in three successive waves of migration, some together with cervids, hippos and murids. The paleobiogeographic context of these dispersal events will be discussed in the next chapter.