Galerix exilis, DE BLAINVILLE, 1839

Galerix exilis seu Galerix symeonidisi

Text-fig. 6

S t u d i e d m a t e r i a l a n d m e a s u r e m e n t s. MWQ1/2001: one right M2 (Pal. 3016: 2.10×2.50×2.39); one damaged left M2 (Pal. 3017), two left m2 (Pal. 3018: 2.56×1.46×1.54; 3019: 2.72×1.73×1.77).

MWQ2/2003: one upper incisor (Pal. 3397: 1.02×0.73); one upper canine (Pal. 3534: 1.39×0.65); one fragment of upper canine (Pal. 3539); one P1 (Pal. 3530: 1.30×0.63); one P2 (Pal. 3399: 1.65×0.83); one fragment of P4 (Pal. 3510); one right P4 (Pal. 3503: 2.44×2.11×2.48); one right M1 (Pal. 3518: 2.30×2.73×2.87); two right M3 (Pal. 3520: 1.34×1.57×1.89; 3521: 1.29×1.57×1.87); one p2 (Pal. 3400: 1.35×0.83); one left p2 (Pal. 3545: 1.51×0.82); one p3 (Pal. 3523: 1.39×0.80); one fragment of p3 (Pal. 3535: -×0.97); one left p4 (Pal. 3541: 1.82×1.21); two left m1 (Pal. 3508: 2.70×1.64×1.85; Pal. 3509: 2.91×1.61×1.85); one fragment right m1 (Pal. 3401: -×-×1.81); five left m2 (Pal. 3402: -×1.45×1.49; 3403: 2.34×1.53×1.61; 3531: 2.46×1.64×1.67; 3532: -×1.54×-; 3533: -×1.68×-).

MCQ3/2005: one right p4 (Pal. 3908: 1.94×1.21); one left m2 (Pal. 3934: 2.41×1.47×1.60); one left M2 (Pal. 3933: 2.05×2.62×2.35).

D e s c r i p t i o n. Upper incisor. The monocuspid tooth has a large and posteriorly curved root. Upper canine. The tooth shows two roots. It is very narrow, and its main cusp is anterior. The posterior talon is slightly elongated.

P1. The tooth has two roots and small dimensions. The main cusp is in a central position. A posterior cingulum is present.

P2. The tooth shows two roots, the posterior one being thicker. It is more robust in shape than the P1. The cusp is conical. A thin ridge connects its tip to a barely visible posterior cingulum.

P4. The paracone is massive and occupies the labial half of the tooth. A parastyle is present, but it is not connected to the paracone or to the protocone. The latter is about the same size as the hypocone, and both cusps are isolated from each other. There is a crest running labially from the protocone that ends before reaching the paracone ( Text-fig. 6a). A cingulum runs from the base of the hypocone along the posterior side of the tooth. It ends below the posterolabial crest of the paracone.

e

M1. The protocone connects directly to the paracone through the preprotocrista; the protoconule is absent. The parastyle is a weak cuspule. A short and straight crest is present between the postparacrista and the premetacrista. The postmetacrista does not end in a clear metastyle. The metaconule is large ( Text-fig. 6b) and attached to the base of the metacone through the premetaconule crest. The metaconule is also connected by a low crest to the protocone-hypocone loph. The postmetacrista ends freely near the posterior cingulum.

M2. Pal. 3016 is quite worn ( Text-fig. 6c), Pal. 3017 is damaged ( Text-fig. 6d), whereas Pal. 3933 is well preserved ( Text-fig. 6e). Yet, like in the M1s, its morphology prevents an attribution to the species level. The overall morphology of M2 resembles that of M1. It differs in a more rounded outline in occlusal view, a shorter postmetacrista, a less posterolingual position of the hypocone, and overall by slightly smaller dimensions. There is also a thin crest connecting the metaconule to the protocone-hypocone loph in Pal. 3933, but not Pal 3016-3017 ( Text-fig. 6c–e).

M3. The tooth is small and triangular. It is characterized by the presence of three cusps (protocone, metacone, paracone) at the three corners of the tooth, delimitating a rather rounded and shallow basin. The M3 also shows a well-developed parastyle ( Text-fig. 6f–g). The metaloph is short. Only the anterior cingulum is present.

p2. The tooth is two-rooted. The cuspid is more conical than the one in P2. Moreover, it shows a small anterior bulge attached to the base ( Text-fig. 6h).

p3. The tooth shows a wider and more compact shape than the p2. The protoconid is placed in a central position. The paraconid is low but clearly defined ( Text-fig. 6i). There is a well-defined postcristid of the posterior lingual side of the tooth.

p4. The tooth is subtriangular. The paraconid is the most anterior cuspid and is isolated from the protoconid ( Text-fig. 6j–k). The protoconid is the highest cuspid. It is connected lingually to the metaconid, which is about the same height as the paraconid. The talonid is noticeably lower than the trigonid. There is only an incomplete posterior cristid, which ends before reaching the lingual side of the tooth.

m1. The first lower molars have a trigonid and a talonid of similar length. The protoconid has a triangular shape, while the metaconid is more conical. Both cuspids are faintly connected by a metalophid divided into two short cristids. The protoconid also connects to the paraconid, the most anterior cuspid. There is a clear notch anterior to the protoconid ( Text-fig. 6l). The talonid displays a high entoconid. It shows a low anterior cristid running lingually toward the trigonid, interrupted by a superficial notch ( Text-fig. 6m). The hypoconid is lower and connects to the trigonid through an oblique cristid. The postcristid connects both trigonid cuspids.

m2. The m2 differs from the m 1 in the anterior compression of the trigonid, resulting in a more curved paralophid, completely including the paraconid and a shorter length. Also, the postcingulid is more strongly reduced ( Text-fig. 6n–r).

R e m a r k s. The genus Galerix is the most common insectivore in Early Miocene fossil sites in Europe. Although its first records come from the earliest Miocene of Anatolia ( de Bruijn et al. 1992, van den Hoek Ostende 1992), it is first identified in Europe in MN 3 localities. To date, four species have been identified in the Early Miocene: Galerix remmerti VAN DEN HOEK OSTENDE, 2003 , G. aurelianensis ZIEGLER, 1990 , G. symeonidisi , and G. exilis . To this is also included “ Galerix ” kostakii DOUKAS et VAN DEN HOEK OSTENDE, 2006, although several authors include this taxon to the genus Parasorex VON MEYER, 1865 , due to the advanced state of its metaconule on M1–2 ( Doukas and van den Hoek Ostende 2006, Prieto et al. 2012, Zijlstra and Flynn 2015). G. remmerti is mostly found in Spanish assemblages ( van den Hoek Ostende and Furió 2005) and is considered to be closely related to the Central European G. aurelianensis ( Ziegler 1990) , based on the less derived morphology and the high-crested pattern on upper molars. The main difference between them is the slightly smaller size of G. remmerti (van den Hoek Ostende 2003) . Both species are larger than our specimens. Regarding G. symeonidisi and G. exilis , the main differences between them reside in the morphology of P3 ( Doukas 1986, van den Hoek Ostende and Doukas 2003). The assemblage from Mokrá-Quarry yielded two different P3 morphologies, one that shows a well-developed hypocone and attributed to G. symeonidisi , and another that lacks a developed hypocone, referred to G. exilis ( Ziegler 1983) . Despite G. exilis rarely possessing a hypocone on P3 ( Ziegler 1983), this cusp is weakly developed and the parastyle is still relatively prominent, which is not the case here. In addition, the morphometric variability of the assemblage is consistent with a mixture of G. symeonidisi and G. exilis , at least in MWQ2/2003. This mixture of two distinct Galerix species has been identified in late MN 4 and earliest MN 5 of central Europe thanks to the in-depth study of van den Hoek Ostende and Doukas (2003). Most recovered upper molars resemble G. exilis , based on size similarity, a reduced protoconule, and the frequent presence of protoconule-metaconule connection. Pal. 3517 ( Text-fig. 5g) constitutes the only exception. The small size, reduced parastyle, and thin posthypocrista more closely resemble Galerix symeonidisi . Although most of the Galerix assemblage cannot be attributed to the species level, the overall morphological characteristics suggest that G. exilis was more abundant than G. symeonidisi .