PALEVO 1631-0683 Elsevier S1631-0683(13)00080-8 10.1016/j.crpv.2013.05.008 Research article General Palaeontology, Systematics and Evolution (Vertebrate Palaeontology) General palaeontology, systematics and evolution (Vertebrate palaeontology) Tiny teeth of consequence: Vestigial antemolars provide key to Early Miocene soricid taxonomy (Eulipotyphla: Soricidae) Petites dents d’importance : les antémolaires vestigiales fournissent la clé de la taxonomie des soricidés (Eulipotyphla : Soricidae) du début du Miocène Klietmann Johannes johannes.klietmann@yahoo.de a Nagel Doris a Rummel Michael b van den Hoek Ostende Lars W. c University Vienna, Department of Palaeontology, Althanstraße 14, 1090 Vienna, Austria University Vienna Department of Palaeontology Althanstraße 14 Vienna 1090 Austria Naturmuseum Augsburg, Augusta Arcaden, Ludwigstraße 2, 86152 Augsburg, Germany Naturmuseum Augsburg Augusta Arcaden Ludwigstraße 2 Augsburg 86152 Germany Naturalis, PO Box 9517, 2300 RA Leiden, The Netherlands Naturalis, PO Box 9517 RA Leiden 2300 The Netherlands 12 5 S1631-0683(13)X0006-5 257 267 © 2013 Académie des sciences. Published by Elsevier B.V. All rights reserved. 2013 Académie des sciences Full (PDF)

In the soricid fauna of the Early Miocene Petersbuch 28 fissure fill from Bavaria, Germany, specimens of Miosorex desnoyersianus (Lartet, 1851) and Paenelimnoecus micromorphus (Doben-Florin, 1964) retain a tiny penultimate antemolar. This antemolar was previously unknown for P. micromorphus, because the tiny antemolar has not been preserved until now. Miosorex desnoyersianus retains the rudimentary tooth; its possible function was to keep the p4 in upright position. The younger species of Paenelimnoecus have lost this vestigial tooth. Based on the alveoli, Miosorex pusilliformis is interpreted as a junior synonym of M. desnoyersianus, and the diagnoses of Miosorex desnoyersianus and Paenelimnoecus micromorphus are emended.

Dans la faune des soricidés du remplissage de fissure bavarois Petersbuch 28, Allemagne, Miocène inférieur, des spécimens de Miosorex desnoyersianus (Lartet, 1851) et Paenelimnoecus micromorphus (Doben-Florin, 1964) conservent une pénultième antémolaire (a2) minuscule. Cette antémolaire était auparavant inconnue chez P. micromorphus. Le nombre différent d’antémolaires n’est pas considéré comme suggérant une espèce nouvelle, mais indique plutôt que la minuscule antémolaire a été conservée pour la première fois. Miosorex desnoyersianus conserve la dent rudimentaire ; sa fonction était peut-être de maintenir la position verticale de la p4. Les espèces les plus récentes de Paenelimnoecus ont perdu cette dent vestigiale. En nous fondant sur les alvéoles, nous interprétons Miosorex pusilliformis comme un synonyme junior de M. desnoyersianus, et les diagnoses de Miosorex desnoyersianus et Paenelimnoecus pusilliformis sont émendées.

 Miosorex, Paenelimnoecus, Antemolar reduction, Germany, Miocene, Soricidae Miosorex, Paenelimnoecus, Réduction des antémolaires, Allemagne, Miocène, Soricidae presented Presented by: Philippe Taquet Introduction

The number of antemolars is a key character in the taxonomy of shrews (e.g., Hugueney and Maridet, 2011, Ziegler, 1989 and Ziegler, 2003). In spite of their importance, their provenance is not fully understood, hence the term “antemolars”. It is unknown whether they are incisors, canines, premolars or several of the above – and how many of each kind of tooth types are represented. The most posterior lower antemolar is commonly considered the p4, and some recent authors have restricted the antemolar count to the teeth anterior to it (Hugueney et al., 2012 and Rofes and Cuenca-Bescós, 2009); in older descriptions, the p4 is counted among the antemolars (Baudelot, 1972, Doben-Florin, 1964 and Ziegler, 1989). In the upper dentition, the P4 is much larger and more complex morphologically than the antemolars and is therefore not counted as an antemolar, even in older descriptions (Baudelot, 1972, Doben-Florin, 1964, Hugueney et al., 2012, Rofes and Cuenca-Bescós, 2009 and Ziegler, 1989). This assignment agrees with studies of living species (Choate, 1968).

For the other antemolars, the situation is less obvious. Different interpretations were given by Merriam (1895), Ärnbäck-Christie-Linde (1912), Kindahl (1960) and James (1963). Although Repenning (1967) disagreed with the previous designations, James’ (1963) dental formula of I1-I2-I3-C-P2-P3 for the upper antemolars and i3–c for the lower antemolars of Sorex Linnaeus, 1758 and Blarina Gray, 1838 is widely accepted for recent soricids (Choate, 1968 and Hillson, 2005). The number of antemolars can differ in living and extinct species, and no embryological studies can be conducted on the latter. Therefore, the antemolars cannot be determined as incisors, canines or premolars, except for the upper and lower P4 (Hugueney et al., 2012 and Rofes and Cuenca-Bescós, 2009).

The molars of Soricidae are notoriously conservative in their morphology, and therefore have limited use in distinguishing taxa. In order to identify fossil shrews confidently, the molars, the anterior dentition and the condyle are needed. Unfortunately, complete anterior dentitions are rare. Of course, the number of antemolars can also be determined on the basis of an alveolar row. But still, complete reference material is needed for determining the sizes and numbers of roots of the antemolars.

Many specimens from the Early Miocene Petersbuch 28 fissure fill of Bavaria show excellent preservation (Rosina and Rummel, 2012). A surprising feature is the presence of a tiny lower antemolar (the a2) in Miosorex and Paenelimnoecus. This element had previously not been noted in the latter genus. In this paper, we describe the a2 and determine the consequences of this discovery for the taxonomy of the two genera. The teeth are small and featureless enough to pass as sediment grains, so we conducted a chemical analysis to make sure that they are really teeth. They are too radiant for bone or dentin and too rounded for a broken bit of enamel. The number of upper antemolars does not correspond to the number of lower antemolars (Doben-Florin, 1964 and Ziegler, 1989). Furthermore, vestigial antemolars have never been found in the upper dentition. The mandibles are more commonly found than maxillary remains, so the mandibular teeth are better known and often more important for shrew taxonomy (Baudelot, 1972, Doben-Florin, 1964, Engesser, 2009 and Ziegler, 1989). Therefore, we confine ourselves to the lower antemolars.

Nomenclatural history of <italic>Miosorex</italic> and <italic>Paenelimnoecus</italic>

The first description of a species now belonging to the genus Miosorex concerned a small shrew with three antemolars described by Lartet (1851) under the name of Sorex desnoyersianus; it was placed within Miosorex by Baudelot (1972). The next member of the genus was described by Depéret (1892), who named a quite similar but larger “Sorex pusillus race grivensis”, after the French site of La Grive. The use of the generic name Sorex for these fossil shrews was found to be erroneous by Kretzoi (1959), who baptized the genus Miosorex, raising Depéret's “race” to a distinct species. Some years later, Doben-Florin (1964), in her extensive work on Miocene shrews of Wintershof-West, added another species to this group, but still used the generic assignment of Sorex, calling the new species “Sorex pusilliformis” and described it as being unique in having only two mandibular antemolars. Doben-Florin would have preferred the name “Sorex pusillus”, but this name had already been given to another soricid by von Meyer (1846), who gave neither a sufficient description nor good drawings of the species. Kretzoi (1959) had renamed “Sorex pusillus” Oligosorex meyeri, because the name Sorex pusillus was preoccupied by a species named by Gmelin (1774). Von Meyer's (1846) “Sorex pusillus” is now considered a nomen dubium (Storch, 1988). Oligosorex meyeri is therefore invalid as well. The species pusilliformis was included in Miosorex by Ziegler (1989).

In Paenelimnoecus, a comparable situation is found. The oldest description of one of its species was given by Kormos (1934) under the name of “Pachyura pannonica”, which was eventually included within Paenelimnoecus by Reumer (1984). The next of these tiny shrews to be described was P. micromorphus, first named “Limnoecus micromorphus” (Doben-Florin, 1964) for the similarities to the American genus Limnoecus Stirton, 1930. Bachmayer and Wilson (1970) described “Petenyiella repenningi”, later included in Paenelimnoecus by Reumer (1984). The taxon Paenelimnoecus had been erected by Baudelot (1972), along with the new species P. crouzeli. Gibert Clols (1975) erected the species “Limnoecus truyolsi”, which was reallocated to the genus Paenelimnoecus by Van den Hoek Ostende et al. (2009). The latest members described are P. obtusus Storch, 1995 and P. chinensis Jin and Kawamura, 1997. None of the previous studies on these taxa ever mentioned a number of lower antemolars other than two – even though Jin and Kawamura (1997) showed drawings of P. chinensis with only one antemolar, the p4. Table 1 gives an overview over the species discussed here.

 Material and methods

One hundred and forty-five mandibles of Miosorex desnoyersianus and 48 mandibles of Paenelimnoecus micromorphus from the Early Miocene locality Petersbuch 28, Bavaria, Germany, were examined. They will be stored at the Naturmuseum Augsburg (NMA) and the Collection Michael Rummel (CMR), Weißenburg in Bayern (Bavaria). The Miosorex specimens have the collection numbers NMA 2012-105/2058–110/2058, CMR-P/28-1781, 1796, 1822, 1824, 1825, 1828, 1829, 1831, 1834, 1837, 1838, 1892, 1905, 1917, 1919, 1924, 1935, 1950, 1972, 1976, 1988, 2114, 2166, 2649, 2668, 2676, 2684, 2703, 2734, 2757, 2800, 2809, 2858, 2870, 2877, 2878, 2880, 2881, 2883–2887, 2889–2894, 2896–2901, 2903–2906, 2908, 2909, 2911, 2912, 2914–2918, 2920, 2921, 2923, 2924, 2926, 2929, 2932, 2933, 2936–2939, 2941, 2946–2948, 2950–2953, 2955, 2957, 2958, 2960–2962, 2964–2971, 2973, 2979–2987, 2990, 2991, 2994–2998, 3001, 3002, 3004, 3007–3010, 3014, 3015, 3017–3019, 3022–3024, 3026–3030, 3032, 3034–3038, 3041, 3043–3045, 3047, 3050, 3051, 3053–3058, 3060–3065, 3067, 3068, 3070, 3071, 3074, 3077, 3078, 3081, 3082, 3085–3087, 3089–3091, 3094, 3096, 3099, 3110, 3119, 3144, 3145, 3159. The Paenelimnoecus specimens have the numbers NMA 2012-111/2058–118/2058, CMR-P/28-1840, 2869, 3084, 3098, 3100, 3102, 3103, 3111, 2859, 3107, 3109, 3112–3118, 3120–3124, 3126–3135, 3137–3143, 3146–3157.

The nomenclature of the morphological features follows Reumer (1984). The descriptions and measurements of the teeth are given by Klietmann (2013) and are under preparation for publication. The chemical analyses of the newly found rudimentary teeth of Miosorex and Paenelimnoecus were done by Hugh Rice and Johannes Klietmann on the SEM in low vacuum, using an energy dispersive X-ray spectrometry (EDS) system, the EDAX Genesis Apex 4 X-ray Microanalysis System of Metek, which identifies the various elements through the X-rays caused by the electrons hitting the object. The machine further counts how often the signal from each searched element was found. In order to have comparable results, each probing was done for 100 s.

The MN-Units as used herein were extensively discussed in Van der Meulen et al. (2011).

 Systematic palaeontology

Class: MAMMALIA Linnaeus, 1758

Order: EULIPOTYPHLA Waddell et al., 1999

Family: SORICIDAE Fischer, 1814

Subfamily: CROCIDOSORICINAE Reumer, 1987

Genus Miosorex Kretzoi, 1959

Type species: Sorex grivensis Depéret, 1892

Original diagnosis (Kretzoi, 1959): (..) Miosorex n.g. (holotype: Sorex pusillus race grivensis Depéret, 1892) which retains but two unicuspids [= antemolars] like Myosorex.

Three species have been assigned to the genus: Miosorex desnoyersianus (Lartet, 1851), M. grivensis (Depéret, 1892) and M. pusilliformis (Doben-Florin, 1964).

Miosorex desnoyersianus (Lartet, 1851)

Fig. 1A

Holotypus: none given.

Neotypus: right Mandible with p4-m3, Sa 1201, Muséum national d’Histoire naturelle, Paris.

Locus typicus: Sansan, France (Gers). Middle Miocene, MN 6.

Time range: Early Miocene (MN 3) (Ziegler, 2006a) to Middle Miocene (MN 6) (Engesser, 2009).

Emended diagnosis: small crocidosoricine with three antemolars, the second of which (a2) is extremely small. The p4 has two roots. The distal part of the first antemolar and the p4 ridges are high and resemble small accessory cuspules. On m1 and m2, the hypolophid is distinctly separated from the entoconid and forms a clear entostylid. The buccal cingulid is weak; the lingual cingulid is wide, but wears no ridge. The oblique cristid of the m1 ends about equally far from the buccal tooth border as in the m2. The m3 has a talonid with only one cusp. The foramen mentale is below the p4. The lower incisor is long and tricuspulate, the posterior border of its crown extends only a little farther backward on the buccal than on the lingual side. The upper and lower facets of the condyle are not divided; the processus condylaris is emarginated on the buccal side (crocidurine shape). The hypocone of the M1 and M2 is developed as a weak cusp or a ridge, the hypoconal flange of the M1 is extensive, the anterolingual corner is rounded, and the posterior cingulum of P4, M1 and M2 is mostly uninterrupted.

Description: the single-rooted a1 is almost triangular with a concave posterior margin. The anterior margin reaches up where the tooth covers the incisor. The cusp is in the tooth's anterior third. The cusp's ridges form a Y, having a rounded anterior ridge and two posterior ridges. Both posterior ridges end short of the cingulid. A large valley is enclosed between them. The cingulid encircles the whole tooth except the small most anterior part. It is most pronounced at the posterior corners. The tiny a2 is completely overlain by the p4 on all specimens; it is by far wider than long and is very low and bulbous. Because the p4 covers the a2, no measurements of the a2 could be taken. The tooth seems to point in an anterobuccal direction; therefore, the highest point is at the anterior end, but it does not form a clear tip, because it is in contact with the p4 over its complete length. Two tiny ridges or bulges seem to run in lingual and buccal direction, respectively, in the larger specimens (CMR-P/28-2898, CMR-P/28-2909, CMR-P/28-2960). The buccal part of the tooth is larger and may reach further buccal than the anterior part of the p4. The root is small and reaches posteriorly below the tip of the p4.

The p4 has two roots, the posterior being large; the small anterior root was never seen in the material described herein. The anterior margin reaches upward, where the tooth contacts the previous antemolar. The only cusp is Y-shaped and bends a bit lingually. The lower anterior margin forms a small concavity for the contact to the a2.

Subfamily: ALLOSORICINAE Fejfar, 1966

Genus Paenelimnoecus Baudelot, 1972

Type species: Paenelimnoecus crouzeli Baudelot, 1972

Diagnosis (translated from Baudelot, 1972): it is a Limnoecine and is characterized by the loss of the entoconid at the molars and its reduced size; on the m1, the buccal valley opens well above the cingulid and the metaconid is very close to the protoconid; the talonid of the m3 is reduced to a cutting crest and only a weak rest of the lingually open basin is left; the teeth were coloured.

The genus currently includes Paenelimnoecus micromorphus (Doben-Florin, 1964), P. pannonicus (Kormos, 1934), P. repenningi (Bachmayer and Wilson, 1970), P. crouzeli Baudelot, 1972, P. truyolsi (Gibert Clols, 1975), P. obtusus Storch, 1995 and P. chinensis Jin and Kawamura, 1997.

Paenelimnoecus micromorphus (Doben-Florin, 1964)

Figs. 1B–C2 and 2

Holotypus: right Mandible with p4, m1, m2 and m3 in situ, posterior part broken off. Bayerische Staatssammlung für Paläontologie und historische Geologie, Munich, 1937 II 10472.

Locus typicus: Wintershof-West, Germany. Early Miocene, MN 3.

Time range: Early Miocene (MN 3–MN 4);?Middle Miocene (MN 5) (Ziegler, 2006b).

Emended diagnosis: very small shrew with two clearly visible and one vestigial lower antemolars, an elongated incisor and a much-reduced entoconid and entocristid on the m1 and m2. The m2 is usually longer than the m1. The talonid of the m3 bears one cusp. Contrary to all other species of the genus, P. micromorphus has a lower p4 with two posterior ridges and a minute penultimate antemolar.

Description: the single-rooted a1 is tear-shaped with a straight posterior margin. The cusp is a large, three-sided pyramid bending slightly in a posterior direction. Its ridges form a Y consisting of the rounded anterior and two posterior ridges. Both posterior ridges end at about half the length of the tooth and border a shallow valley. The cingulid encompasses the whole tooth except for the most anterior part. It is largest in the posterobuccal corner and small on the lingual side. In only one mandible (NMA 2012-114/2058; Fig. 1B), the tiny a2 is large enough to be visible in occlusal view; in this specimen, the a2 is almost elliptical with a concave posterior side; it is clearly wider than long and bears no ridges, only a wide, but weak bulge alongside the main axis. It is the only specimen big enough to be measured; its length is 0.13 mm and the width is 0.23 mm. Like the other antemolars, it extends anterodorsally. In the other specimens, the a2 is smaller than 0.1 mm; it cannot be measured correctly, because it is reduced to a tiny chip between the other antemolars, completely covered by the p4. It still appears to be wider than long.

The p4 has one root and a Y-shaped wear facet. The anterior margin reaches upward where the tooth overlays the previous antemolar. In buccal view, the anterior margin follows the margin of the a2, creating a pronounced concavity, and forms a small protrusion in anterior direction above the a2.

 Comments on the material from Petersbuch 28

Antemolars were quite rarely preserved in the mandibles. Of the 145 mandibles of Miosorex desnoyersianus found in Petersbuch 28, only 26 still include antemolars at all. Only eight of these 26 specimens retain other antemolars than just the p4 (Klietmann, 2013). The small additional antemolar was found in five of these eight specimens. On one of the remaining three mandibles with antemolars, all antemolars were damaged and the a2 was not found. The other two mandibles might just have lost the tiny a2 through taphonomy. None of the tiny a2 is measurable, but a considerable difference in size is visible; in one specimen, the tooth is clearly visible and projects in a buccal direction clearly out of the p4's cover. In all the other specimens, it is just a tiny bump between the antemolars. Independent of the number of antemolars, if any of them is tiny; it is the penultimate one (Klietmann, 2013). In Paenelimnoecus micromorphus, only thirteen out of 48 mandibles include antemolars. In the other specimens, the antemolars are missing. In seven of the thirteen specimens containing antemolars, both the a1 and p4 are preserved. The additional antemolar is preserved in four of the seven jaws; one of the three others has antemolars covered by sinter and the two remaining mandibles have damaged antemolars, so the tiny tooth is probably lost. The largest a2 (NMA 2012-114/2058; Fig. 1B) can be seen in occlusal view, because it is too large to fit low between the other teeth. The other three specimens of the a2 are extremely tiny and nearly completely featureless.

The small structure could theoretically be a sand grain; the shiny surface forbids any determination as bone or dentin. In order to test whether it was not just a grain from the matrix, a scan on the EDS System was performed and showed a quite similar chemical composition of the smallest tooth of Paenelimnoecus micromorphus when compared to the p4 next to it (Fig. 2). The results for Miosorex desnoyersianus and the unusually large a2 of P. micromorphus are given in the supplementary material. In comparison to the results for rodent teeth obtained by García-Alix et al. (2013), the high amount of oxygen and the extra elements like sodium, magnesium, silicium and aluminium demonstrate the specimen to be diagenetically altered. Still, the results of the structure under question and the neighbouring p4 are quite similar towards each other, apart from including more sediment. The sediment consists predominantly of aluminium, silicium and oxygen. Judging from the similarity of the chemical composition, the tiny tooth consists of hydroxyapatite that was diagenetically altered exactly like the p4. Therefore, the structure most probably belongs to the dentition. A misidentification of the small tooth as sediment grain can be excluded. Bone or dentin cannot be confused with enamel, because they are not as radiant and translucent under a light microscope; therefore, the structures cannot consist of bone or dentin. An enamel particle is a highly improbable explanation for the observed structure. First, enamel particles of comparable sizes and shapes would have been caught in exactly the same spot in multiple mandibles and nowhere else. Second, broken enamel would not be rounded and similarly smooth on all visible surfaces. Enamel particles could only be rounded under circumstances that would abrade the enamel of the neighbouring teeth considerably, but no sign of such an abrasion is visible in the material.

 Discussion Recognizing the a2 in other Miocene soricid material

This structure between a1 and p4 was not recognised as a tooth immediately. Rather, it was thought to be a sediment particle. This is excluded by the analysis using the X-ray Microanalysis System, as the small teeth show the same chemical composition as their respective neighbouring p4. In the element counts of the small antemolar (Fig. 2B) compared to the element counts of the neighbouring p4 (Fig. 2A), aluminium and silica are more common. This is due to problems of cleaning the tiny teeth between the larger ones, so more sediment remained on their surface; also, there are fewer surfaces to choose really clean spots for the analyses. The element counts of Miosorex desnoyersianus (Supplementary Figs. S1 and S2) and Paenelimnoecus micromorphus (Fig. 2, Supplementary Figs. S3 and S4) are also comparable, so the pattern of P. micromorphus corroborates the presence of an additional tooth, the a2.

Another possible explanation of the structure is that it formed part of a neighbouring tooth. In specimen NMA 2012-114/2058 of Paenelimnoecus micromorphus, the tooth was large enough to be recognised as a tooth of its own. Upon removing the a2, its large root became visible. Undamaged a1 and p4 do not have an enlargement that could be mistaken for the a2, so the enamel structure under discussion cannot be part of a neighbouring tooth.

Additionally, in these specimens, there is still a groove for the respective tooth between the a1 and the p4. Therefore, it can safely be concluded to be a tooth of its own, a tiny a2.

The relative size of this tooth is very small, but this is not unusual in soricids; a tiny penultimate antemolar of comparable size ratio is found in Lartetium petersbuchense (Ziegler, 1989) and is also present in Miosorex grivensis (Baudelot, 1972 and Viret and Zapfe, 1951) and in the genus Oligosorex (Van den Hoek Ostende, 2001).

The number of lower antemolars is of great importance for the determination of soricids (e.g., Reumer, 1984, Van den Hoek Ostende, 2001, Ziegler, 1989, Ziegler, 1999, Ziegler, 2003, Ziegler, 2006a, Ziegler, 2006b and Ziegler, 2009). The vestigial tooth of Miosorex desnoyersianus is easily lost in the taphonomic process; it is still unknown from Sansan, the locus typicus, where only the number of alveoli is known (Engesser, 2009). Doben-Florin (1964) used the number of antemolars to differentiate between Miosorex desnoyersianus and M. pusilliformis. This view was followed by Ziegler (1989), who described M. pusilliformis from different German sites. It is clear from his extensive discussion that recognizing M. pusilliformis as a separate species led to a taxonomic puzzle. Presumably, Miosorex pusilliformis had fewer antemolars than M. desnoyersianus and should therefore be considered to display a more derived dental formula. Yet, M. pusilliformis appeared in the older localities, with a mixed population of the two Miosorex species in the fissure Petersbuch 2. Thus, Ziegler (1989) had to assume the replacement of a species with a fairly derived dental formula by a virtually indistinguishable form with a more primitive dental formula.

This problem resulted from the lack of an exhaustive description of the alveoli at the time; the alveoli were not discussed in detail by Baudelot (1972). Engesser (2009) finally rectified this situation by giving a detailed and explained picture of Miosorex desnoyersianus alveoli from Sansan (Fig. 3C). Based on the rareness of the a2 even in the exceptionally rich material of Petersbuch 28 and the minute size of the p4's anterior root, the taxonomic situation of Miosorex “pusilliformis” had to be evaluated. The direct comparison of the pictures of M. “pusilliformis” given by Doben-Florin (1964) and Ziegler (1989) to the drawing of Engesser (2009) shows completely identical structures of the alveoli (Fig. 3). The presumed difference in the number of alveoli was the only reason for erecting a new species, so M. “pusilliformis” is according to the new data a junior synonym of M. desnoyersianus.

Consequently, the question arises how the antemolar could have escaped attention up to now. Doben-Florin (1964) had only one specimen of M. “pusilliformis” with both large antemolars still in place; in a photograph of this mandible, the region of the accessory antemolar appears to be covered by sinter. Also, it could have been lost without leaving a visible gap. All other specimens lacked one or both antemolars. Since it was already known that the tiny third antemolar of Miosorex grivensis is always represented by its alveolus (Viret and Zapfe, 1951), both large alveoli covered by the p4 were seen as belonging to this tooth (Doben-Florin, 1964). The small groove on the bony ridge below the p4 was obviously not interpreted as being an alveolus, so the specimens of Wintershof-West seemed to have only two antemolars. The two roots of the p4 were described by Doben-Florin (1964). Ziegler (1989) determined Miosorex “pusilliformis” and “M. desnoyersianus” based mostly on the number of alveoli, since most mandibles had already lost the antemolars. The new found synonymy of the two species indicates that the specimens determined as Miosorex “pusilliformis” belong in fact to M. desnoyersianus; the specimens determined as “M. desnoyersianus” in Petersbuch 2 by Ziegler (1989) could either belong to a different species or be specimens with an unusually large anterior alveolus of the p4. This question can only be answered by a revision of the original material.

Even the specimens of Petersbuch 28 that have lost the a2 in course of the taphonomic process do not have a clear gap between the other teeth. Therefore, only the larger sample size provided the opportunity to actually find this tooth. Sigé et al. (1997) did not give the number of specimens or of teeth from Bouzigues 1 and 2. Stephan-Hartl (1972) worked on the entire fauna of Frankfurt-Nord Bassin; in the single shrew specimen, the anterior antemolars were lost.

In Paenelimnoecus micromorphus, the preservation situation is comparable. Doben-Florin (1964) had no specimens with both other antemolars in situ and described the p4 as having two roots. Ziegler (1989) also mentioned the presence of a two-rooted p4 in contrast to the one-rooted p4 in the younger species. The finding of the vestigial tooth shows that the p4 is one-rooted in P. micromorphus, too. The smallest shrew within the Early Miocene faunas tends to be present in relatively small numbers (Doben-Florin, 1964, Klietmann, 2013 and Ziegler, 1989). Given that the tiny mandibles are usually damaged, it is nearly impossible to find a tooth of this size. In fact, except for the unusually large one, the tooth was only recognized by examining SEM photos of the specimens.

The shrew faunas of the Czech Republic's sites, as listed by Fejfar and Sabol (2005), still await description and the determinations are therefore preliminary; presence or absence of the new found tooth has still to be assessed.

Of course, following this reasoning, we must accept that within one genus, the number of antemolars is variable and the loss of an antemolar is convergent in several taxa. This is acceptable as long as it concerns teeth in a state of reduction just prior to complete loss. In Heterosorex neumayrianus, a heterosoricid shrew, the number of antemolars is used for determining the subspecies (Doben-Florin, 1964 and Ziegler and Fahlbusch, 1986), so variability can be expected.

The last question arising from the new finds is the old question of “Why?”. A tooth of this extremely small size had surely lost function in processing food, because it barely came into occlusal contact with the food or the upper dentition. Nonetheless, it was retained within Miosorex desnoyersianus from MN 3 in Wintershof-West to MN 6 of Sansan (Doben-Florin, 1964, Engesser, 2009 and Klietmann, 2013). One possible interpretation is that the tiny tooth served somehow as an additional column, taking over the role of an additional anterior root, bearing part of the pressure exercised on the p4; this is improbable because a direct contact with the other antemolar could facilitate this better. The easiest explanation is that the small size of this tooth is part of a trend to reduce the number of antemolars, a common trend in soricids, until it did not “get in the way” anymore; after having reached this point, a further reduction would only result in a negligible saving of enamel and dentin, which was probably not important enough to cause further reduction. The tooth position could be conserved at little energetic cost – even if it did not have any function in the dentition. If we follow Furió et al. (2011) and consider Oligosorex to be closely related or even congeneric to Lartetium, this tooth position may have been retained over a long time in a single lineage, as Oligosorex lasted from the Upper Oligocene (MP 28) to the Early Miocene (MN 2) (Van den Hoek Ostende, 2001); Lartetium is present in the Middle Miocene (MN 4 to MN 6).

The most probable function is that of a wedge, holding the p4 in its position (Van den Hoek Ostende, 2001); in contrast to the first explanation of an additional column, this does not mean it is necessary to bear the pressure, but it helps to keep the correct position of the p4. This would at least decrease the speed of the further reduction, as any reduction would have to be balanced by some change within the other antemolars to keep the p4 exactly upright. In Paenelimnoecus the complete reduction of the vestigial tooth happened quite early, as only the oldest species still retained this tooth position. Small changes of the adjacent teeth made the minute tooth expendable. In his figure 2, Reumer (1992) showed that the space occupied by the vestigial a2 in P. micromorphus is filled by an extension of the lower margin of the p4 in P. crouzeli. Therefore, the loss of the vestigial tooth had to be compensated by the p4 in order to keep the p4 in upright position. The minuscule tooth still played a role within the dentition of P. micromorphus. A similar function may be assumed in Miosorex desnoyersianus and other shrews still retaining minute penultimate antemolars.

 Trends in soricid antemolar dentition

The earliest crocidosoricine genus known is Srinitium Hugueney, 1976. It had five lower antemolars anterior to the p4, more than all other known genera (Hugueney, 1976 and Ziegler, 1998). In another Oligocene genus, Ulmensia Ziegler, 1989, the number of antemolars is reduced to three, not counting the p4 (Ziegler, 1989). Reducing the number of antemolars remained a trend within the soricids and occurred repeatedly as parallel evolution within the Crocidosoricinae and the Allosoricinae. We included Paenelimnoecus Baudelot, 1972 in the Allosoricinae; the taxonomic state of the genus is still debated (Hugueney et al., 2012, Prieto and Van Dam, 2012 and Van den Hoek Ostende et al., 2009). Within the Crocidosoricinae, most genera retained three lower antemolars, the last being vestigial (Fig. 4). Soricella discrepans Doben-Florin, 1964 and Florinia stehlini (Doben-Florin, 1964) reached the state of having two large antemolars with no vestigial antemolar. Both species most probably reached this state as a parallelism. Given the basal position of Soricella discrepans in the cladogram of Hugueney and Maridet (2011), it must have evolved the reduced number of antemolars independently of most other crocidosoricines.

In Miosorex Kretzoi, 1959, the third antemolar was lost and the second antemolar became vestigial. Therefore, it was supposed to be the next relative to the Myosoricini/-inae by Furió et al. (2007). In the analysis of Hugueney and Maridet (2011), no species of Miosorex shows affinities with Myosorex Gray, 1838. Furthermore, the reduced number of antemolars might have evolved independently, possibly even as a parallelism within Miosorex.

We could demonstrate that the vestigial second antemolar was present in the oldest species of Paenelimnoecus, although it was lost in the younger ones. In the recent genera of soricids, there are no vestigial lower antemolars left (Dannelid, 1998), except in Myosorex (Dannelid, 1998 and Furió et al., 2007). According to Stanley et al. (2005), vestigial lower antemolars are sometimes present in the African genus Congosorex Heim de Balsac and Lovett, 1956, but always absent in C. phillipsorum Stanley et al., 2005. Soricinae and Crocidurinae reduced their dentitions in parallel, because their respective sister clades retain a higher number of antemolars (Fig. 4).

The upper antemolars are independent of their lower counterparts in their evolutionary trends. For example, Oligosorex Kretzoi, 1959 and Lartetium Ziegler, 1989, have three lower antemolars anterior to the p4, but Lartetium has four upper antemolars and Oligosorex has five (Hugueney et al., 2012). Soricella discrepans has two lower antemolars anterior to the p4 and four upper antemolars (Doben-Florin, 1964), whereas Miosorex desnoyersianus (Lartet, 1851) has a further reduced second lower antemolar, but five upper antemolars. Likewise, the recent genus Sorex Linnaeus, 1758 retains five upper antemolars of slightly variable sizes, but only one lower antemolar (Moska et al., 2008). The numbers of upper and lower antemolars of more taxa are given in Fig. 4.

The evolutionary reductions of the upper and lower antemolar number are thus independent of each other and of systematic position, occurring repeatedly as parallelisms within the Soricidae. For Blarina brevicauda, antemolar crowding was found to be the reason for the aberrant loss of individual antemolars (Choate, 1968). The presence of vestigial antemolars should also be taken into account when considering the identification of the antemolars. Usually, the premolars are reduced from anterior to posterior direction (Hillson, 2005). If the antemolar row cannot contain four premolars, the first and then the second premolars are interpreted as being lost during evolution (Hillson, 2005 and James, 1963). However, the various evolutionary states found in the Crocidosoricinae and in Paenelimnoecus indicate a different trend. Whenever there is a vestigial antemolar, it is the last one anterior to the p4 – independent of the actual number of antemolars. There is one antemolar anterior to the vestigial one in Miosorex desnoyersianus, M. grivensis (Depéret, 1892) and Paenelimnoecus micromorphus (Doben-Florin, 1964) and two in most other crocidosoricines. Soricella discrepans and Florinia stehlini do not retain any vestigial antemolars (Doben-Florin, 1964 and Ziegler, 1989). The trend of reducing the premolars is therefore different in shrews than in most other mammals. The penultimate premolar – no matter which one it is – becomes tiny and is subsequently lost within the various lineages. Otherwise, we would have to assume that the vestigial premolar remained present, whereas a larger one within the antemolar row was lost completely in quite a short time. The premolars are therefore reduced in the anterior direction. The p4 is never reduced; probably its function is crucial to the mastication process.

In the upper antemolar row, the last antemolar is usually smaller than the others (Dannelid, 1998 and Moska et al., 2008). The upper antemolar row itself is formed by quite uniform teeth that act as a functional unit. It is not likely that one tooth within the antemolar row would be lost completely and the others would fill the gap. The last upper antemolar is more commonly lost than most other antemolars (Feldhaber and Stober, 1993 and Jogahara et al., 2007). The antemolar interpreted as the third incisor or as the canine is lost more often in aberrant dentitions than the others (Feldhaber and Stober, 1993 and Jogahara et al., 2007). Nonetheless, the state of aberrant dentition cannot be used to determine evolutionary trends, because most cases of tooth loss are postnatal and probably due to periodontal disease (Jogahara et al., 2008). More likely, the functional unit of the antemolar row is reduced at one of its ends, most probably at the posterior end, where the large P4 almost covers the small last antemolar. Like the lower antemolars, the upper antemolar row is reduced in anterior direction. Therefore, the first premolar lost is the upper or lower P3. Any hypothesis that the antemolars include the P3/p3, but not the P1/p1, can therefore be rejected on grounds of the most probable evolutionary trend.

 Conclusions

Miosorex desnoyersianus (Lartet, 1851) and Paenelimnoecus micromorphus (Doben-Florin, 1964) were found in Petersbuch 28 (Germany, Early Miocene, MN 3/4); both species exhibit a tiny penultimate antemolar that has never been described for Paenelimnoecus micromorphus until now. Based on the comparison of pictures of the alveoli of Miosorex desnoyersianus from Sansan (France, Middle Miocene, MN 6) and M. “pusilliformis” of Wintershof-West (Germany, Early Miocene, MN 3) and Petersbuch 2 (Germany, Early Miocene, MN 4), M. “pusilliformis” must be regarded as a junior synonym of M. desnoyersianus. The usually poor preservation of the small soricids and the extremely small size of P. micromorphus tooth, which is hardly visible using anything but an electron microscope, has made it very difficult to spot this feature. The younger species of Paenelimnoecus lack these small teeth. The function of the tooth was to keep the p4 in upright position, without adding to the length of the antemolars. In Paenelimnoecus, the tooth was lost in the younger species, but functionally replaced by an enlargement of the p4 covering the respective space.

A compilation of the known crocidosoricine and allosoricine genera and the recent Soricinae, Crocidurinae and Myosoricinae shows that reduction of antemolars was a common parallelism in shrews, occurring independently in multiple lineages. Based on the small sizes of the upper and lower last antemolars in most shrews, the reduction of the premolars is interpreted to start at the upper and lower P3 and commence in an anterior direction, contrary to the trends in most other mammalian taxa.

 Acknowledgements

We want to thank Gerhard Storch (Forschungsinstitut Senckenberg, Frankfurt am Main, Germany) for the identification of most species.

Also, we are grateful to Reinhard Ziegler (Staatliches Museum für Naturkunde Stuttgart, Germany) for helpful advice. We also thank Gertrud Rößner (Bayerische Staatssammlung für Paläontologie und Geologie, Munich, Germany) for granting access to comparative material from Munich. We are greatly indebted towards Christa Hofmann (Department for Palaeontology, University Vienna, Austria) and Hugh Rice (Department for Geodynamics and Sedimentology, University Vienna, Austria) for granting access to and helping with the SEM and the EDAX X-ray Microanalysis System. We thank the reviewers and the editor for corrections, remarks and suggestions helping to improve the article and to broaden its significance.

 Supplementary data

Ärnbäck-Christie-Linde A. Der Bau der Soriciden und ihre Beziehungen zu anderen Säugetieren Gegenbauers morphologisches Jahrbuch 44 1912 201–295 Bachmayer F. Wilson R.W. Die Fauna der altpliozänen Höhlen- und Spaltenfüllung bei Kohfidisch Burgenland (Österreich) Annalen des Naturhistorischen Museums Wien 74 1970 533–587 Baudelot, S., 1972. Étude des Chiroptères, Insectivores et Rongeurs du Miocène de Sansan (Gers). PhD Thesis, Université Paul Sabatier de Toulouse, 364 p. Choate J.R. Dental Abnormalities in the Short-Tailed Shrew Blarina brevicauda J. Mammal. 49 2 1968 251–258 Dannelid E. Dental adaptation in shrews Wójcik J.M. Wolsan M. Evolution of Shrews 1998 Mammal Research Institute, Polish Academy of Sciences Bialowieza 156–174 Depéret C. La faune des mammifères de la Grive-Saint-Alban (Isère) Archive du Muséum d’Histoire naturelle de Lyon 5 1892 1–93 Doben-Florin U. Die Spitzmäuse aus dem Altburdigalium von Wintershof-West bei Eichstätt in Bayern Abhandlungen der Bayerischen Akademie der Wissenschaften, mathematisch-naturwissenschaftliche Klasse (Neue Folge) 117 1964 1–82 Engesser B. The Insectivores (Mammalia) from Sansan (Middle Miocene, south-western France) Schweizerische Paläontologische Abhandlungen 128 2009 1–91 Fejfar O. Die plio-pleistozänen Wirbeltierfaunen von Hajnáĉka und Ivanovce (Slowakei), CSSR. V. Allosorex stenodus n.g. n.sp. aus Ivanovce A Neues Jahrbuch für Geologie und Paläontologie, Abhandlungen 123 1966 221–248 Fejfar O. Sabol M. Czech Republic and Slovak Republic Van den Hoek Ostende L.W. Doukas C.S. Reumer J.W.F. The Fossil Record of the Eurasian Neogene Insectivores. (Erinaceomorpha, Soricomorpha, Mammalia), Part I Scripta Geologica Special Issue 5 2005 51–60 Feldhaber G.A. Stober T.L. Dental anomalies in five species of North American shrews Mammalia 57 1 1993 115–122 Fischer G. Zoognosia tabulis synopticis illustrata, in usum praelectionum academiae imperialis medico-chirurgicae Mosquensis edita. Volumen tertium. Quadrupedum reliquorum, cetorum et montrymatum descriptionem continens 1814 Typis Nicolai Sergeidis Vsevoloszky Moscow 732 p Furió M. Santos-Cubedo A. Minwer-Barakat R. Agustí J. Evolutionary History of the African Soricid Myosorex (Insectivora, Mammalia) out of Africa J. Vert. Paleontol. 27 4 2007 1018–1032 Furió M. Casanovas-Vilar I. Van den Hoek Ostende L.W. Predictable structure of Miocene insectivore (Lipotyphla) faunas in western Europe along a latitudinal gradient Palaeogeogr. Palaeoclimatol. Palaeoecol. 304 2011 219–229 García-Alix A. Minwer-Barakat R. Martín Suárez E. Freudenthal M. Delgado Huertas A. Cinnabar mineralization in fossil small mammal remains as a consequence of diagenetic processes Lethaia 46 1 2013 1–6 Gibert Clols J. New Insectivores of the Miocene of Spain I and II Proceedings of the Koninklijke Nederlandse Akademie van Wetenschappen Serie B 78 1975 108–133 Gmelin S.G. Reise durch Russland zur Untersuchung der drey Natur-Reiche. Dritter Theil. Reise durch das nördliche Persien, in den Jahren 1770. 1771. bis im April 1772 1774 Imperial Academy of Sciences St. Petersburg 514 Gray J.E. Revision of the genus Sorex, Lin Proceedings of the Zoological Society of London 1837 1838 123–126 Heim de Balsac H. Lovett J.C. Évolution et phylogénie des Soricidès africains Mammalia 20 1956 140–167 Hillson S. Teeth 2005 Cambridge University Press New York/Cambridge 373 p Hugueney M. Un stade primitive dans l’évolution des Soricinae (Mammalia, Insectivora) : Srinitium marteli nov. Gen., nov. Sp. de l’Oligocène moyen de Saint-Martin-de-Castillon (Vaucluse) C. R. Acad. Sci. Paris, Série D 282 1976 981–984 Hugueney M. Maridet O. Early Miocene Soricids (Insectivora Mammalia) from Limagne (Central France): new systematic comparisons, updated biostratigraphic data and evolutionary implications Geobios 44 2011 225–236 Hugueney M. Mein P. Maridet O. Revision and new data on the Early and Middle Miocene soricids (Soricomorpha, Mammalia) from central and south-eastern France Swiss J. Palaeontol. 131 2012 23–49 James G.T. Paleontology and nonmarine stratigraphy of the Cuyama Valley Badlands, California. Part 1. Geology, faunal interpretations, and systematic descriptions of Chiroptera, Insectivora, and Rodentia University of California Publications in Geological Sciences 45 1963 1–154 Jin C.Z. Kawamura Y. A new species of the extinct shrew Paenelimnoecus from the Pliocene of Yinan, Shandong Province, northern China Transactions and Proceedings of the Palaeontological Society of Japan (New Series) 1 1997 67–75 Jogahara T. Koyasu K. Oda S. Kawai T. Hanamura H. Quest for the cause of oligodontia in Suncus murinus (Soricomorpha Soricidae): Morphological re-examination Archives of Oral Biology 52 2007 836–843 Jogahara T. Oda S. Kawai T. Hanamura H. Koyasu K. Numerical variation of teeth in the wild house musk shrew Suncus murinus captured from Nagasaki Japan Archives of Oral Biology 53 7 2008 617–621 Kerbis Peterhans J. Stanley W.T. Hutterer R. Demos T.C. Agwanda B. A new species of Surdisorex Thomas, 1906 (Mammalia Soricidae) from western Kenya Bonner zoologische Beiträge 56 3 1906 175–183 Kindahl M. Some aspects of the tooth development in Soricidae Acta Odontologica Scandinavica 17 1960 203–237 Klietmann, J., 2013. Systematic and ecological analysis of Marsupialia and Eulipotyphla from Petersbuch 28 (Germany, Lower Miocene). Unpublished PhD. thesis, University of Vienna. Kormos T. Neue Insektenfresser, Fledermäuse und Nager aus dem Oberpliozän der Villànyer-Gegend Foldtani Közlony 64 1934 296–321 Kretzoi M. New names for Soricid and Arvicolid homonyms Vertebrata Hungarica 1 2 1959 247–248 Lartet E. Notice sur la colline de Sansan Annuaire du Département du Gers 1851 1851 1–42 Linnaeus C.N. Systema Naturae per Regna Tria Naturae, secundum Classes, Ordines, Genera, Species, cum Characteribus, Diferentiis, Synonymis, Locis Tomis I 1758 Laurentii Salvii, Holmiae Stockholm 823 p Lopatin A. Early Miocene small mammals from the North Aral Region (Kazakhstan) with special reference to their biostratigraphic significance Paleontological Journal 38 Supplement 3 2004 217–323 Merriam C.H. Revision of the shrews of the American genera Blarina and Notiosorex North American Fauna 10 1895 5–34 and 102–107 Moska M. Laskowska M. Kosowska B. Strzala T. Marszalek-Kruk B. Variation of the common shrew (Sorex araneus L.) dentition Zoologica Poloniae 53 1–4 2008 49–56 Prieto J. Van Dam J.A. Primitive Anourosoricini and Allosoricinae from the Miocene of Germany Geobios 45 6 2012 581–589 Repenning C.A. Subfamilies and genera of the Soricidae United States Geological Survey Professional Paper 565 1967 pp. iv and 1–74 Reumer J.W.F. Ruscinian and Early Pleistocene Soricidae (Insectivora Mammalia) from Tegelen (The Netherlands) and Hungary Scripta Geologica 73 1984 1–173 Reumer J.W.F. Redefinition of the Soricidae and the Heterosoricidae (Insectivora Mammalia) with the description of the Crocidosoricinae, a new subfamiliy of Soricidae Revue de Paléobiologie 6 2 1987 189–192 Reumer J.W.F. The taxonomical position of the genus Paenelimnoecus Baudelot, 1972 (Mammalia: Soricidae): a resurrection of the subfamily Allosoricinae J. Vert. Paleontol. 12 1 1992 103–106 Rofes J. Cuenca-Bescós G. A new genus of red-toothed shrew (Mammalia, Soricidae) from the Early Pleistocene of Gran Dolina (Atapuerca, Burgos Spain), and a phylogenetic approach to the Eurasiatic Soricinae Zool. J. Linn. Soc. 155 2009 904–925 Rosina V. Rummel M. The bats (Chiroptera Mammalia) from the Early Miocene of Petersbuch (Bavaria, southern Germany) Geobios 45 5 2012 463–478 Sigé B. Crochet J.Y. Sudre J. Aguilar J.P. Escarguel G. Nouveaux sites d’âges variés dans les remplissages karstiques du Miocène inférieur de Bouzigues (Hérault, Sud de la France). Partie I : sites et faunes 1 (insectivores, chiroptères, artiodactyles) Geobios 20 1997 477–483 Stanley W.T. Rogers M.A. Hutterer R. A new species of Congosorex from the eastern Arc Mountains Tanzania, with significant biogeographical implications J. Zool. 265 2005 269–280 Stephan-Hartl R. Die altmiozäne Säugetierfauna des Nordbassin und der Niederräder Schleusenkammer (Frankfurt/M., Hessen) und ihre stratigraphische Stellung Abhandlungen des hessischen Landesamts Bodenforschung 64 1972 1–97 Stirton R.A. A new genus of Soricidae from the Barstow Miocene of California. University of California Publications Bull. Dpt. Geol. Sci. 19 8 1930 217–228 Storch G. Insectivora (Mammalia) aus dem Kalktertiär (Oberoligozän–Untermiozän) des Mainzer Beckens Geol. Jahrbuch A 110 1988 337–343 Storch G. The Neogene mammalian faunas of Ertemte and Harr Obo in Inner Mongolia (Nei Mongol), China.–11. Soricidae (Insectivora) Senckenbergiana lethaea 75 1/2 1995 221–251 Van Dam J.A. Van den Hoek Ostende L.W. Reumer J.W.F. A new short-snouted shrew from the Miocene of Spain Geobios 44 2011 299–307 Van den Hoek Ostende L.W. Insectivore Faunas from the Lower Miocene of Anatolia, Part 6: Crocidosoricinae Scripta Geologica 122 2001 47–81 Van den Hoek Ostende L.W. Insectivores (Erinaceomorpha, Soricomorpha Mammalia) from the Ramblian of the daroca-Calamocha area López-Martínez N. Peláez-Campomanes P. Hernández Fernández M. En torno a fósiles de mamíferos: datacíon, evolución y paleoambiente 1 2003 Coloquios de Paléontología, Volumen extraordinario 281–310 Van den Hoek Ostende L.W. Furió M. Garcia-Paredes I. New data on Paenelimnoecus from the Middle Miocene of Spain support the shrew subfamily Allosoricinae Acta Palaeontologica Polonica 54 1 2009 159–164 Van der Meulen A. García-Paredes I. Álvarez-Sierra M.Á. Van den Hoek Ostende L.W. Horkijk K. Oliver A. López-Guerrero P. Hernández-Ballarín V. Peláez-Campomanes P. Biostratigraphy or biochronology? Lessons from the Early and Middle Miocene small Mammal Events in Europe Geobios 44 2011 309–321 Viret J. Zapfe H. Sur quelques soricidés miocènes Eclogae Geologicae Helvetiae 44 1951 411–426 von Meyer H. Mittheilungen an Professor Bronn. Neues Jahrbuch für Mineralogie, Geognosie, Geologie und Petrefaktenkunde 1846 1846 462–476 Waddell P.J. Okada N. Hasegawa M. Towards Resolving the Interordinal Relationships of Placental Mammals Systematic Biology 48 1 1999 1–5 Ziegler R. Heterosoricidae und Soricidae (Insectivora Mammalia) aus dem Oberoligozän und Untermiozän Süddeutschlands Stuttgarter Beiträge zur Naturkunde B 154 1989 1–73 Ziegler R. Marsupialia und Insectivora (Mammalia) aus den oberoligozänen Spaltenfüllungen Herrlingen 8 und Herrlingen 9 bei Ulm (Baden-Württemberg) Senckenbergiana lethaea 77 1998 101–143 Ziegler R. Order Insectivora Rössner G.E. Heissig K. The Miocene land mammals of Europe 1999 Dr. Friedrich Pfeil Munich 53–74 Ziegler R. Shrews (Soricidae, Mammalia) from the Middle Miocene karstic fissure fillings of Petersbuch near Eichstätt Southern Franconian Alb (Bavaria) Paläontologische Zeitschrift 77 2 2003 303–322 Ziegler R. Insectivores (Lipotyphla) and Bats (Chiroptera) from the Late Miocene of Austria Annalen des Naturhistorischen Museums in Wien 107A 2006 93–196 Ziegler R. Miocene Insectivores from Austria and Germany–an overview Van den Hoek Ostende L.W. Nagel D. Harzhauser M. Festschrift für Univ-Doz. Dr. Gudrun Daxner-Höck gewidmet aus Anlass ihres 65 30 2006 Geburtstages. Beiträge zur Paläontologie 481–494 Ziegler R. Soricids (Soricidae Mammalia) from Early Oligocene fissure fillings in South Germany – and a phylogenetic analysis of the Heterosoricinae Palaeodiversity 2 2009 321–342 Ziegler R. Fahlbusch V. Kleinsäuger-Faunen aus der basalen Oberen Süßwasser-Molasse Niederbayerns Zitteliana 14 1986 2–80

SEM photographs of the vestigial teeth of Miosorex desnoyersianus (Lartet, 1851) and Paenelimnoecus micromorphus (Doben-Florin, 1964) from Petersbuch 28, Germany, Early Miocene, MN 3/4. Miosorex desnoyersianus. A. NMA 2012-110/2058: left a1, a2, p4, m1, lingual view. Paenelimnoecus micromorphus: B. NMA 2012-114/2058: right a1, a2, p4, m1; largest a2, lingual view. C1. NMA 2012-118/2058: right i, a1, a2, p4, m1, buccal view. C2. NMA 2012-118/2058: right i, a1, a2, p4, m1, lingual view.

Photos MEB des dents vestigiales de Miosorex desnoyersianus (Lartet, 1851) et Paenelimnoecus micromorphus (Doben-Florin, 1964) de Petersbuch 28 (Allemagne, Miocène inférieur, MN 3/4). Miosorex desnoyersianus. A. NMA 2012-110/2058 : a1, a2, p4, m1 gauches, vue linguale. Paenelimnoecus micromorphus : B. NMA 2012-114/2058 : a1, a2, p4, m1droites; a2 plus grande, vue linguale. C1. NMA 2012-118/2058 : i, a1, a2, p4, m1 droites, vue buccale. C2. NMA 2012-118/2058 : i, a1, a2, p4, m1 droites, vue linguale.

Element counts of teeth of the Paenelimnoecus micromorphus (Doben-Florin, 1964) specimen NMA 2012-118/2058 from Petersbuch 28, Germany, Early Miocene, MN 3/4. A. Element count of the p4. B. Element count of the minute a2.

Comptage des éléments des dents de Paenelimnoecus micromorphus (Doben-Florin, 1964) spécimen NMA 2012-118/2058 de Petersbuch 28, Allemagne, Miocène inférieur, MN 3/4. A. Comptage des éléments de la p4. B. Comptage des éléments de la minuscule a2.

Comparison of the alveoli of Miosorex “pusilliformis” from Winterhof-West (A) and Stubersheim 3 (B), both from Germany, Early Miocene, MN 3, to the alveoli of Miosorex desnoyersianus (Lartet, 1851) from Sansan (C), France, Middle Miocene, MN 6. Buccal (1) and occlusal (2) view. A after Doben-Florin (1964), B after Ziegler (1989), C after Engesser (2009). Not to size.

Comparaison des alvéoles de Miosorex “pusilliformis” de Wintershof-West (A) et Stubersheim 3 (B), tous les deux d’Allemagne, Miocène inférieur, MN 3, avec les alvéoles de Miosorex desnoyersianus (Lartet, 1851) de Sansan (C), France, Miocène moyen, MN 6. Vue buccale (1) et occlusale (2). A d’après Doben-Florin (1964), B d’après Ziegler (1989), C d’après Engesser (2009). Échelle non respectée.

Stratigraphic and systematic overview over known Crocidosoricinae and descendants of the subfamily. Solid lines indicate the stratigraphic distributions. The numbers of upper and lower antemolars (not counting the P4/p4) are given after each taxon. (+) and (–) indicate if the last antemolar is vestigial. The cladogram follows Hugueney and Maridet (2011), the stratigraphic data are based on Furió et al. (2007). If no different result was found in the cladogram of Hugueney and Maridet (2011), monophyly of higher taxa was assumed. Further information are from Doben-Florin, 1964, Fejfar, 1966, Hugueney, 1976, Hugueney et al., 2012, Kerbis Peterhans et al., 2009, Lopatin, 2004, Reumer, 1992, Stanley et al., 2005, Van Dam et al., 2011, Van den Hoek Ostende, 2003, Viret and Zapfe, 1951, Ziegler, 1989 and Ziegler, 1998.

Vue d’ensemble stratigraphique et systématique des Crocidosoricines connus et des descendants de la sous-famille. Les traits pleins indiquent les distributions stratigraphiques. Les nombres du antémolaires supérieures et inférieures (sans compter la P4/p4) sont donnés pour chaque taxon. (+) et (–) indiquent si la dernière antémolaire est vestigiale ou non. Le cladogramme suit Hugueney et Maridet (2011), les données stratigraphiques sont basées sur Furió et al. (2007). Si aucun résultat différent n’a été trouvé dans le cladogramme de Hugueney et Maridet (2011), une monophylie des taxons supérieurs a été supposée. Des informations supplémentaires proviennent de Doben-Florin (1964) ; Fejfar (1966) ; Hugueney (1976) ; Hugueney et al. (2012) ; Kerbis Peterhans et al. (2009) ; Lopatin (2004) ; Reumer (1992) ; Stanley et al. (2005) ; Van Dam et al. (2011) ; Van den Hoek Ostende (2003) ; Viret and Zapfe (1951) ; Ziegler, 1989 and Ziegler, 1998.

List of species included in Miosorex Kretzoi, 1959 and Paenelimnoecus Baudelot, 1972, with their original names.

Liste d’espèces comprises dans Miosorex Kretzoi, 1959 et Paenelimnoecus Baudelot, 1972, avec leurs noms d’origine.

Genus Original name Current name Authority Miosorex Kretzoi, 1959 Sorex desnoyersianus Lartet, 1851 Miosorex desnoyersianus (Lartet, 1851) Baudelot (1972) Sorex pusillus race grivensis Depéret, 1892 Miosorex grivensis (Depéret, 1892) Kretzoi (1959) Sorex pusilliformis Doben-Florin, 1964 Miosorex desnoyersianus (Lartet, 1851) This paper Unknown Sorex pusillus von Meyer, 1846 = Oligosorex meyeri Kretzoi, 1959 [Nomen nudum, species invalid] Storch (1988) Paenelimnoecus Baudelot, 1972 Pachyura pannonica Kormos, 1934 Paenelimnoecus pannonicus (Kormos, 1934) Reumer (1984) Limnoecus micromorphus Doben-Florin, 1964 Paenelimnoecus micromorphus (Doben-Florin, 1964) Ziegler (1989) Petenyiella repenningi Bachmayer and Wilson, 1970 Paenelimnoecus repenningi (Bachmayer and Wilson, 1970) Reumer (1984) Paenelimnoecus crouzeli Baudelot, 1972 Paenelimnoecus crouzeli Baudelot, 1972 Baudelot (1972) Limnoecus truyolsi Gibert Clols, 1975 Paenelimnoecus truyolsi (Gibert Clols, 1975) Van den Hoek Ostende et al. (2009) Paenelimnoecus obtusus Storch, 1995 Paenelimnoecus chinensis Jin and Kawamura, 1997