In this paper, we describe a previously unknown species of the glirid Eliomys from the Late Miocene and Early Pliocene Cabriel, Alcoy and Granada basins of southeastern Spain. Eliomys yevesi sp. nov. is characterized by its relative small size, narrow lingual wall and common presence of two centrolophs in the upper molars, and well-developed centrolophids in the lower molars. The new species is the probable ancestor of E. intermedius, which in turn represents the ancestor of the extant E. quercinus. According to its morphologic and biometric features, the origin of E. yevesi sp. nov. is likely to be found in some population of E. truci from the Late Miocene. Based on these affinities, we propose the lineage E. truci—E. yevesi sp. nov.—E. intermedius—E. quercinus, in which there is a trend towards the development of centrolophs, as well as the reduction of accessory crests.
External, genetic and some morphometric characters of the skull distinguish at least three extant species of the genus Eliomys (Filipucci et al. 1988a, b; Krystufek and Kraft 1997; Holden 2005), including the garden dormouse Eliomys quercinus (Linnaeus, 1766), the Asian dormouse E. melanurus Wagner, 1839 (type species) and the Maghreb garden dormouse E. munbyanus Pomel, 1856. Its biogeographic range comprises Western Europe, where it is widespread, as well as scattered populations across Central, Eastern, and Southeast Europe, Turkey, and the north of Africa. In addition, fossil representatives of the genus include E. assimilis Mayr, 1979 and E. reductus Mayr, 1979 from Germany, and E. lafarguei Aguilar, Michaux, and Lazzari, 2007 from France, all them from the Late Miocene, as well as E. truci Mein and Michaux, 1970 from the Late Miocene-Pliocene, and E. intermedius Friant, 1953 from the Pliocene and Pleistocene, both of Western Europe.
Eliomys truci, E. intermedius, and the extant E. quercinus have been proposed to form part of a single phylogenetic lineage, following first a cladogenetic pattern of evolution, based on the coexistence of E. truci and E. intermedius during the Early-Late Pliocene (Adrover 1986), before culminating in the anagenetic replacement of E. intermedius by E. quercinus during the Pleistocene (Chaline 1972; Castillo 1990). However, this picture is made more complex by the appearance of an additional, intermediate form (E. aff. intermedius) sharing features with both E. truci and E. intermedius, in the Mio-Pliocene of the Granada Basin of southern Spain (García- Alix et al. 2008a). Here, we provide new information on the evolutionary history of the genus by describing new Mio-Pliocene material from rich localities with the Cabriel (Venta del Moro and La Bullana) and Alcoy basins of southeastern Spain, which, together with E. aff. intermedius from the Granada Basin, we refer it to the new species Eliomys yevesi sp. nov.
Institutional abbreviations. —DEPUG, Departamento de Estratigrafía y Paleontología of the University of Granada, Spain; MGUV, Museum of Geology of the University of Valencia, Spain.
Other abbreviations. —L, length; W, width; see Table 1 for locality name abbreviations.
Material and methods
Nomenclature and measurement methods follow Daams (1981) and Freudenthal (2004). Measurements are in millimetres and were taken on a Leica MZ75 binocular microscope by means of displacement of a mechanical stage, connected to Sony Magnescale measuring equipment. Photographs were taken with a HITACHI 4800 scanning electron microscope at the Central Support Facility for Experimental Research (SCSIE) of the University of Valencia. The specimens from Venta del Moro, La Bulla and Alcoy are kept at the MGUV, and labelled according to locality, level and, where applicable, year of collection (Table 1).
To determine its affinities, we directly compared our material with fossil collections from Huétor Tájar, Tollo de Chiclana, Moreda, Bélmez-1, Cañada Castaños, Gorafe-2 and 5, Botardo-C, Negratín-1, Calicasas-3B, Purcal-4, 23, 24A, and 25, and Otura-1, housed at the DEPUG (see SOM, Supplementary Online Material available at http://app.pan.pl/SOM/app60-Mansino_etal_SOM.pdf). In addition to these direct comparisons, we took into account published data contained from the localities of Balaruc 2, Layna, Caravaca, Crevillente 1, 3, 5, and 6, Tortajada A and C, and Masada del Valle 2 (all in Weerd 1976), as well as Orrios 3, Sete, Arquillo 3, Villalba Alta, Hautimagne, Los Mansuetos, and Aljezar B (Adrover 1986). Finally, we performed a phylogenetic analysis of the genus using the implicit enumeration option of the phylogenetic software package TNT (Goloboff et al. 2003, 2008). All of the characters included in this analysis are listed in the Appendix 1.
Locality abbreviations. Square brackets denote modifiers indicating particular levels (e.g., C, D) or years of collection.
The material described here was recovered from lacustrine and palustrine sediments belonging to the Villatoya-Venta del Moro Formation (Robles 1970; Montoya et al. 2006) exposed at Venta del Moro (VVm) and La Bullana 2B (LB- 2B), both situated in the Cabriel Basin of southeastern Spain (Robles 1970; Morales 1984; Montoya et al. 2006), as well as the localities of Alcoi Forn 1'06 (AF-1'06), Alcoi Barranc Sud 3A (ABS-3A), Alcoi Cristian 0 (AC-0), and Alcoi 2 (AL2-C and AL2-D), all of them located in the west margin of the lacustrine Basin of Alcoy (southeastern Spain). At Venta del Moro, specimens were collected from several levels, previously referred to as VM-A, B, C, BC, and D by Montoya et al. (2006). The latter are here labeled as VVm-A, VVm-B, VVm-BC, VVm-C, and VVm-D, respectively, to distinguish the specimens housed at the University de Valencia from previously collected material from the same locality housed at the Museo Nacional de Ciencias Naturales, Madrid (Montoya et al. 2006).
The age of the fossiliferous levels of Venta del Moro, has recently been dated to 6.23 Ma, or latest Miocene (late Turolian-late Messinian), based on their faunal assemblage and magnetostratigraphic data (magnetochron C3An; Gibert et al. 2013). This age estimate is confirmed by the presence of Apocricetus alberti Freudenthal, Mein, and Martín Suárez, 1998, Ruscinomys schaubi Villalta and Crusafont, 1956, Paraethomys meini (Michaux, 1969) and Apodemus gudrunae van de Weerd, 1976 (Freudenthal et al. 1998; Sesé 2006a; García-Alix et al. 2008a). This faunal association is consistent with the Apodemus gudrunae Assemblage Zone in the Guadix Basin (Minwer-Barakat et al. 2012). The presence of Stephanomys dubari Aguilar, Michaux, Bachelet, Calvet, and Faillat, 1991, in the localities of Venta del Moro (Montoya et al. 2006) differs from late Turolian (late Messinian) sites in the Teruel Basin, such as the locality of Arquillo 4, which instead yields S. ramblensis van de Weerd, 1976 (Mein et al. 1990; Adrover et al. 1993).
Stephanomys dubari Aguilar, Michaux, Bachelet, Calvet and Faillat, 1991, Apodemus gorafensis Ruiz-Bustos, Sesé, Dabrio, Peña, and Padial, 1984, Occitanomys alcalai Adrover, Mein, and Moissenet, 1988, and Paraethomys meini (Michaux, 1969) typically appear together in the latest Turolian-earliest Ruscinian localities of southern Spain (late Messinian-early Zanclean; García-Alix et al. 2008b, c). These taxa are present at PUR-4, CLC-3B, AF-1'06, AF- 1'07, and LB-2B, but Occitanomys alcalai and Apodemus gorafensis are absent in ABS-3A, which has yielded few micromammal remains. By contrast, Apocricetus barrierei (Mein and Michaux, 1970), an indicator of early Ruscinian (Zanclean) age (Freudenthal et al. 1998; Sesé 2006a; García- Alix et al. 2008a) is present at ABS-3A, PUR-4, CLC-3B AF-1'07, and LB-2B, yet absent at AF-1'06, which has yielded over a hundred specimens. The localities of PUR-4 and CLC-3B are within the Apocricetus barrierei Assemblage Zone (earliest Ruscinian—earliest Zanclean) of the Granada Basin (García-Alix et al. 2008c).
Eliomys intermedius is a typically Ruscinian taxon (Sesé 2006a). The presence of Apodemus cf. gorafensis, Occitanomys cf. alcalai, Paraethomys meini, and Stephanomys cordii at AC-0 place this locality in the Early Ruscinian (MN14). The faunal lists of AL2-C and AL2-D, which include Trilophomys cf. castroi, Blancomys sp., Stephanomys cordii, Paraethomys meini, Paraethomys aff. abaigari, Ruscinomys cf. lasallei, and Apocricetus cf. angustidens suggest also an early Ruscinian age (Mansino et al. 2013). However, the presence of two species of Paraethomys, the more developed crests of Stephanomys cordii, and the presence of Apocricetus cf. angustidens instead of A. barrierei indicate a younger age for these localities than for AC-0.
Order Rodentia Bowdich, 1821
Family Gliridae Muirhead, 1819
Subfamily Dryomyinae de Bruijn, 1967
Genus Eliomys Wagner, 1840
Type species: Eliomys melanurus Wagner, 1840; extant, Sinai Peninsula, Egipt.
Eliomys yevesi sp. nov.
1953 Eliomys aff. intermedius; Friant 1953: pl. 4: 9–19.
2008 Eliomys aff. intermedius; García-Alix et al. 2008a: 24–25.
Etymology: Dedicated to José María Yeves and his family, owners of the “Hostal Ventamorino”, for their friendly hospitality during the many summer campaigns in Venta del Moro.
Holotype: Right M1, VVmA-25.
Type locality: Venta del Moro A, Cabriel Basin, Spain; coordinates: 39° 28′ 29.12″ N, 1° 20′ 40.38″ W.
Type horizon: Villatoya-Venta del Moro Formation, MN13, late Turolian (Late Miocene, 6.23 Ma).
Material. —Venta del Moro: 2 p4 (VVmA-42, 43), 12 m1, m2 (VVmA-1-11, 32), 6 m3 (VVmA-33–38), 1 dP4 (VVmA- 44), 3 P4 (VVmA-39–41), 12 M1, M2 (VVmA-16, 21–31), 8 M3 (VVmA-12–15, 17–20); 1 m1, m2 (VVmB-1). VVm- BC: 2 m1, m2 (VVmBC-1, 2), 1 m3 (VVmBC-5), 1 M1, M2 (VVmBC-3), 1 M3 (VVmBC-4); 1 m3 (VVmC-1); 2 M1, M2 (VVmD-1, 2). Alcoi Barranc Sud 3A, MN14: 1 m1, m2 (ABS3A-38). Alcoi Forn, MN13-MN14: 1 m1, m2 (AF1'06– 190), 1 m3 (AF1'06-187); 1 M1, M2 (AF1'07-36). Purcal 4, MN14: 1 p4 (PUR-4 871), 12 m1, m2 (PUR-4 872–883), 3 m3 (PUR-4 884–886), 1 dP4 (PUR-4 887), 1 P4 (PUR-4 888), 2 M1, M2 (PUR-4 889, 890), 3 M3 (PUR-4 891–893). Calicasas 3B, MN14: 1 m3 (CLC-3B 67), 2 M1, M2 (CLC- 3B 68, 69), 1 M3 (CLC-3B 70).
Diagnosis. —Size intermediate between Eliomys truci and the youngest E. intermedius; sub-quadrangular m1, m2, usually with a posterotropid and a well-developed centrolophid; no posterotropid on m3; common presence of both centrolophs in the upper molars; posteroloph and endoloph usually connected.
Differential diagnosis. —Eliomys yevesi differs from E. assimilis in the more quadrangular shape of m1, m2, less reduced m3, and the absence of any accessory crests in the upper molars; from E. reductus in its bigger size, more quadrangular shape of the lower molars, discontinuous endolophid, and common presence of posterotropid; and from E. lafarguei in its bigger size, long centrolophid on m1, m2, discontinuous endolophid, the presence of a centrolophid on many m3, and the common presence of postcentrolophs in the upper molars.
Eliomys yevesi ranges in size from the biggest specimens of E. truci to the smallest ones of E. intermedius. It differs from E. truci in more frequently showing centrolophs in the upper molars, as well as the presence of well-developed postcentrolophs, reaching half the width of the tooth or more, in some specimens (Fig. 2G). E. yevesi further differs from E. intermedius in the subquadrangular shape (Fig. 2B, C, N, R) and lower and narrower lingual wall of the upper molars (Fig. 2F, G). A long centrolophid connected to the metalophid is present in all but one m1, m2 of E. yevesi, whereas in E. intermedius it is short in 27 out of 106 specimens. In addition, the posterotropid is more frequent and usually longer in E. yevesi than in E. intermedius (27 out of 29 m1, m2 of E. yevesi; 80 out of 107 of E. intermedius). In the upper molars, the two centrolophs are usually slightly longer and occur somewhat more frequently in E. intermedius (97 out of 165, 58.78%, compared to 17 out of 36, 47.22%, in E. yevesi).
Eliomys yevesi differs from the extant E. quercinus in its smaller size, but resembles the latter in the absence of accessory crests in the upper molars and an anteroloph-protoloph connection, as well as the frequent presence of two centrolophs. Two subspecies of E. quercinus with distinct m1, m2 morphologies occur in the Iberian Peninsula: E. q. lusitanicus, which differs from E. yevesi in the absence of a centrolophid; and E. q. quercinus, which differs from E. yevesi in the absence of accessory crests in the lower molars, a continuous endolophid on m1, m2 and in having a small centrolophid.
Measurements. —See Table 2.
Description. —Material from Venta del Moro: p4: The occlusal outline of this tooth is subtriangular. The protoconid and the large anterolophid create a high, triangular anterior complex, which is separated from the metaconid by a narrow furrow. One specimen (VVmA-42; Fig. 2A) has a short centrolophid. The mesoconid and entoconid are connected by the mesolophid, and separated from the anterior complex by a deep valley. The mesoconid and hypoconid are separated. There is no posterotropid. The posterolophid is high and curved. There are two fused roots.
m1, m2: The occlusal outlines of these teeth are sub-quadrangular. The anterolophid is connected to the protoconid in 6 out of 14 specimens (4 out of 12 from VVm-A, the only specimen from VVm-B and 1 out of 2 from VVm-BC). One specimen (Fig. 2C) has a vestigial anterotropid. Most molars show a connection between the metalophid and the metaconid, whereas the metaconid and entoconid are separated. The centrolophid is usually long and sometimes connected to the metalophid. The posterotropid is long in 10 specimens (9 out of 12 from VVm-A, 1 out of 2 from VVm-BC), short in three (one specimen from VVm-A, the only specimen from VVm-B and 1 of 2 from VVm-BC) and absent in another two (both from VVm-A). The hypoconid is large. There are three roots.
m3: The occlusal outline of this tooth is sub-trapezoidal. The anterolophid is usually separated from the protoconid. The metalophid is sometimes connected to the metaconid. There are no accessory crests. Where present, the centrolophid is not connected to the metalophid. In 1 out of 13 m3, the mesolophid does not reach the entoconid (Fig. 2D). The posterolophid is continuous.
dP4: Known only from a single, extremely worn molar. No features of the occlusal surface can be observed.
P4: The occlusal outline of this tooth is triangular. The paracone and metacone are higher than the protocone. The anteroloph is short, low and connected, at a low level, to the paracone. The protoloph is clearly discontinuous in 1 out of 4 specimens, whereas the remainder shows just a constriction in the central part of this crest. Two specimens present a well-developed precentroloph, and a further one a postcentroloph, which is not connected to the metacone (Fig. 2E). The metaloph is high and continuous. The posteroloph is low and lingually connected to the protocone; discontinuous in two specimens, and not connected to the endoloph in another one (Fig. 2E). The roots are not preserved in any of the specimens.
M1, M2: The occlusal outlines of these teeth are trapezoidal or subrectangular. The anteroloph is separated from the paracone and the protoloph. The paracone and metacone are high and separated. The protoloph and metaloph are continuous, and occasionally sinuous. There is a well-developed precentroloph, which is not connected to the paracone in one of the specimens and reaches the metaloph in another one. Where present, the postcentroloph is short, except in VVmA- 29, which has a well-developed crest, and VVmA-25, in which both centrolophs are fused into a central crest (Fig. 2G). The posteroloph is connected to the endoloph. There are three roots.
M3: The occlusal outline of this tooth is trapezoidal. The anteroloph is connected to the protocone. In some specimens, the protoloph and metaloph are sinuous. Both centrolophs are present in 5 out of 8 specimens from VVm-A, and the single tooth from VVm-BC. Two of the specimens from VVm-A bear a long postcentroloph (Fig. 2H). Where present, the precentroloph is usually attached to the paracone, whereas the postcentroloph is connected to the metacone. The endoloph is continuous, except for the single specimen from VVm-BC. There are three roots.
Material from Alcoy: These specimens from the Alcoy Basin resemble the specimens from Venta del Moro, with the exception of having an anterolophid connected basally with the protoconid, and a metalophid that does not reach the metaconid (both on m1, m2). In the m1, m2 from ABS-3A (Fig. 2J), the centrolophid is not continuous. In the M1, M2 from AF-1'07 (Fig. 2L), the anteroloph and the paracone are connected basally.
Remarks. —Eliomys yevesi is a relatively small-sized species, being smaller than Plio-Pleistocene E. intermedius and E. quercinus, and only slightly larger than the Miocene representatives of the genus (E. lafarguei, E. reductus, and E. assimilis) (Fig. 3). The molars from Venta del Moro and AF- 1'07 fall within the range of variation of E. truci. However, the specimens from the early Ruscinian localities of PUR-4 and CLC-3B (Granada Basin), and AF-1'06 and ABS-3A (Alcoy Basin), are slightly larger, and intermediate between E. truci and E. intermedius (Fig. 3).
Morphologically, the present material resembles E. truci, but differs in the development of the centrolophs on M1, M2, while the presence of both centrolophs is rare in E. truci (5 out of a total of 41 specimens, 12.90%; two specimens from Concud 3 and one each from OTU-1, PUR-23, and Orrios 3), 4 out of 12 specimens E. yevesi from VVm-A and VVm-BC (33.33%), and all of the M1, M2 from VVm-D, AF-1'07, CLC-3B, and PUR-4 (except one) have both centrolophs. Moreover, when present, the postcentroloph is usually reduced in E. truci, whereas in some M1, M2 of E. yevesi from VVm-A it is very well developed, reaching half the width of the molar or more (Fig. 2G). Similarly, E. truci never bears two centrolophs on M3, whereas they do occur in 5 out of 8 molars (62.50%) of E. yevesi from VVm-A, as well as the only M3 from VVm-C and CLC-3B.
In the upper molars of E. intermedius, 79 out of 130 (60.77%) M1, M2 and 18 out of 35 (51.42%) M3 have two centrolophs (Weerd 1976; Adrover 1986; Castillo 1990; García-Alix et al. 2008a). Specimens from both the extant (Castillo 1990) and relatively young fossil (e.g., Casablanca B; García-Alix et al. 2008a) populations of E. quercinus generally have two centrolophs, with the anterior one often being longer. Together, these observations support a trend towards better-developed centrolophs within a lineage comprising E. truci—E. yevesi—E. intermedius—E. quercinus.
Besides the change in development of the centrolophs, Elio mys also shows a trend towards the reduction of the centrolophid and the accessory crests in the lower molars (Castillo 1990; García-Alix et al. 2008a). This observation is borne out by the present data, m1, m2 of E. yevesi resemble those of E. truci in having a long centrolophid and a well-developed posterotropid. In E. intermedius, the centrolophid is still present in 104 out of 106 (98.11%) m1, m2, but in 26 of these specimens (24.53%) it is less than half the width of the molar—with the exception of the material from Sète (n = 30), which always displays a long centrolophid. The centrolophid is even more reduced in E. quercinus, where it is short in 8 (61.54%) and absent in 5 (38.46%) out of the 13 studied specimens. A posterotropid is present in 29 out of 30 m1, m2 of E. truci (96.67%), 27 out of 29 m1, m2 of E. yevesi (93.10%), 80 out of 107 m1, m2 of E. intermedius (74.76%), and 1 out of 11 m1, m2 (9.09%) of E. quercinus (2 of the 13 specimens studied here are too damaged or worn to observe this character).
Stratigraphic and geographic range. —MN13-14, Late Miocene- Early Pliocene; Cabriel, Granada, and Alcoy basins of southeastern Spain.
Measurements (in mm) of the (pre-)molars of Eliomys yevesi sp. nov. from Venta del Moro and Alcoy, and of Eliomys intermedius from Alcoy and La Bullana. n, number of measurable specimens. Repository numbers of the measured specimens are available in Appendix 2.
Eliomys intermedius Friant, 1953
Material. —Alcoi Cristian, MN14: 1 m3 (AC0-88), 1 M1, M2 (AC0-87). Alcoi 2, MN14: 1 m1, m2 (AL2C-97), 1 m3 (AL2C-47); 1 m1, m2 (AL2D-290), 1 m3 (AL2D-356), 3 M1, M2 (AL2D-291-293). La Bullana, MN14: 4 m1, m2 (LB2B-6, LB2B-176, LB2B-242, LB2B-243), 1 m3 (LB2B- 179), 1 P4 (LB2B-178), 2 M1, M2 (LB2B-241, LB2B-253), 1 M3 (LB2B-177).
Measurements. —See Table 2.
Description. —m1, m2: The occlusal outline of these teeth is trapezoidal or subrectangular. The anterolophid is connected to the protoconid in three specimens, and unconnected in a further three. There is no anterotropid. The metalophid is connected to the metaconid. The centrolophid is long, reaching the metalophid in 3 out of 6 specimens. The metaconid and entoconid are separated. The posterotropid is well developed. The hypoconid is large.
m3: The occlusal outline of this tooth is sub-trapezoidal. The anterolophid and the protoconid are separated. There is no anterotropid. The metalophid is connected to the metaconid. There is a short centrolophid, not connected to the metalophid (Fig. 2S). The metaconid and entoconid are separated. There is no posterotropid.
P4: The occlusal outline of this tooth is subtriangular. There is no anteroloph, and the protoloph is discontinuous. The paracone and metacone are higher than the protocone. There is a well-developed precentroloph and a short postcentroloph. The endoloph is continuous.
M1, M2: The outline of these teeth are subrectangular or trapezoidal. The anteroloph and paracone are connected basally in AC0-87, but not connected in any of the other specimens. In some specimens (especially AC0-97), the protoloph and metaloph are distinctly sinuous. Both centrolophs are usually present. The posteroloph is connected to the endoloph. There are three roots.
M3: The occlusal outline of this molar is trapezoidal. The anteroloph is separated from both protoloph and paracone. There is neither an anterotrope nor a posterotrope. The paracone and metacone are high and separated. The protoloph and metaloph are continuous. Both centrolophs are present and well developed, with the postcentroloph being longer and interrupted (Fig. 2P). The endoloph and posteroloph are connected and continuous.
Remarks. —Eliomys intermedius from the Early Pliocene (early Ruscinian-early Zanclean; Mansino et al. 2013) localities of AL2-C and AL2-D is slightly bigger than specimens from the slightly older localities of AC-0 and LB-2B (Table 2), but all of the specimens from these sites fall within the range of E. intermedius from Sète, the type locality of the species (Adrover 1986). The specimens of E. intermedius from AC-0, AL2-C, AL2-D, and LB-2B have higher and thicker lingual walls than E. truci and E. yevesi, sinuous crests, a rounded occlusal outline, and two well-developed centrolophs in the upper molars (M1, M2 from AC-0 and AL2-D, and M3 from LB-2B). These features agree with E. intermedius.
Stratigraphic and geographic range. —MN 14, Early Pliocene- early Pleistocene (MQ2) of the Alcoy and Cabriel, and Calatayud-Teruel basins of Spain and the Aquitaine, Paris, and Sud-est basins of France. Other reported occurrences (Bruijn et al. 1970) are dubious.
Phylogenetic relationships of Eliomys. —Some previous authors (Nadachowski and Daoud 1995) have suggested a relationship of the extant Eliomys quercinus with E. reductus and E. assimilis, both from the German locality of Hammerschmiede (early Vallesian-Tortonian, Late Miocene). However, this is seemingly contradicted by the absence of a clear relationship between E. reductus, E. assimilis and the likely ancestor of E. quercinus, E. intermedius (Daams and de Bruijn 1995).
An alternative hypothesis groups E. reductus with E. lafarguei from Lo Fournas 6C and Lo Fournas 16M (Vallesian- Tortonian, Late Miocene), based on the presence of a continuous endolophid and a trend towards larger size and reduced postcentrolophs, which commonly occurs in glirids (Aguilar et al. 2007). This relationship is further supported by m1, m2 being longer than wide in these two taxa (albeit only slightly so in E. lafarguei) and showing a tendency towards a shortening of the centrolophid. The latter is isolated in the holotype of E. reductus, an m2, and shows a low connection with the metaconid in the m1, whereas in E. lafarguei the centrolophid is absent or greatly reduced on m2, and variably isolated or connected to the metaconid at a low level on m1. By contrast, in E. truci, E. yevesi, E. intermedius, and E. quercinus these teeth are wider than longer, and the centrolophid is usually connected to the metaconid.
Based on these observations, we agree with Aguilar et al. (2007) in recognizing two distinct lineages within Eliomys, comprising E. lafarguei-E. reductus, and E. truci-E. yevesi-E. intermedius-E. quercinus, respectively. The former lineage may also include Eliomys sp. A from the early Turolian of Tortajada A (Teruel, Spain; Weerd 1976), based on its possible relationship with E. lafarguei (Aguilar et al. 2007). On the other hand, details of the morphology of E. assimilis, which has an accessory crest on M1, M2 suggest that this species does not form part of either of the two main lineages. These observations are confirmed by the results of our cladistic analysis (Fig. 4, Table 3).
Evolutionary process. —The oldest record of Eliomys is E. truci from the Middle Miocene of Solera (late Aragonian- early Tortonian; Daams and Freudenthal 1988). However, the identity of this material has been questioned (García-Alix et al. 2008a), thus possibly reducing the earliest occurrence of the genus to E. truci from the Late Miocene (early Vallesian- Tortonian) of Pedregueras 2A (Daams and Freudenthal 1988). Several members of the genus co-occur during the Miocene (Fig. 5), which may imply cladogenetic evolution. Thus, Adrover (1986) suggested E. truci, including its ancestor Eliomys sp. I from Viveros de Pinos (Teruel Basin, north-central Spain; early Turolian-Tortonian), and Eliomys sp. II from Aljezar B (Teruel Basin, middle Turolian, Tortonian), to be descendants of an older, yet unknown species of the same genus. A similar logic applies to E. truci on the one hand, and E. yevesi-E. intermedius on the other. Both of these lineages co-existed during the Ruscinian (Zanclean) and the earliest Villanyan (Piacenzian), as is evident at Orrios-3, Sète, Sarrión, Villalba Alta (Adrover 1986), La Gloria 4, Escorihuela (Mein et al. 1990), Moreda-1A, 1B, Rambla Seca 1 (Castillo 1990), TCH-1B and TCH-13 (García-Alix et al 2008a).
Until the first occurrence of E. quercinus in the earliest Pleistocene locality of Almenara-Casablanca 1 (Agustí et al. 2011), the only representative of Eliomys during the latest Pliocene and early Pleistocene was E. intermedius (García- Alix et al. 2008a).
Cladistic data matrix for Eliomys, including lower (1 to 4) and upper (5 to 8) molar dental characters.
In contrast to the earlier process of cladogenesis, several authors have hypothesized that E. quercinus may have arisen from E. intermedius through a process of anagenetic evolution (Chaline 1972; Castillo 1990; García-Alix et al. 2008a), involving a reduction in size and a trend towards an unconnected anteroloph and protoloph (Castillo 1990). Extant E. quercinus comprises at least eight subspecies, two of which are found in the Iberian Peninsula (Moreno Garrido 1984), the smaller E. quercinus quercinus, which inhabits relatively open environments and is characterized by slightly sinuous crests, a continuous endolophid, a small centrolophid, and no accessory crests; and the larger E. quercinus lusitanicus, which lives in more closed environments, and is marked by more distinctly sinuous crests, accessory crests, and the absence of both the endolophid and the centrolophids. In the latter two features, E. quercinus lusitanicus resembles E. intermedius, which even led to the suggestion that the latter should be regarded as a subspecies of E. quercinus (Martín-Suárez 1988).
In summary, it therefore seems that the lineage including E. truci, E. yevesi, and E. intermedius, and the extant E. quercinus originated and diversified through cladogenesis during the Late Miocene, and Early Pliocene, but then switched to a process of anagenetic evolution during the Late Pliocene and Pleistocene.
Based on new fossil material from the Alcoy and Cabriel basins of southeastern Spain, as well as material from the Granada Basin previously referred to E. aff. intermedius, we erect the new species Eliomys yevesi sp. nov., similar in size or slightly larger than E. truci. E. yevesi generally resembles E. truci in terms of its morphology, but differs in having better and more frequently developed centrolophs in the upper molars, similar to E. intermedius. Based on its size and morphology, we consider the most probable origin of the new species a Late Miocene population of E. truci (Fig. 5), likely an ecomorphotype characterized by well-developed centrolophs in the upper molars, and both a long centrolophid and a well-developed posterotropid in the lower molars. E. yevesi thus likely forms part of a lineage comprising E. truci-E. yevesi-E. intermedius- E. quercinus (to the exclusion of other members of the genus, such as E. assimilis and E. reductus; Fig. 4), which is characterized by a trend towards the reduction of accessory crests in the lower molars, as well as the development of centrolophs. This arrangement is complemented by a second lineage including E. reductus and E. lafarguei, marked by an increase in size, a gradual reduction of the postcentroloph and centrolophid, and a trend towards a more quadrangular occlusal outline (lower length-width ratio).
We thank Elvira Martín-Suárez (Universidad de Granada, Spain) for her comments, Sacramento Moreno (Doñana Biological Station, Sevilla, Spain) for providing a picture of the occlusal surface of the molars of Eliomys quercinus quercinus, Héctor Botella (Universidad de València, Spain) for his help with the cladistic analysis, and Mattijhs Freudenthal (Universidad de Granada, Spain), an anonymous reviewer, and editor Felix G. Marx (University of Otago, New Zealand), for their useful comments and criticism. The field work at Venta del Moro was approved and supported, from 1995 onwards, by the Conselleria de Cultura of the Generalitat Valenciana, with the permission of the Ministerio de Fomento and the Administración de infraestructuras ferroviarias. This investigation was supported by projects GV06/304 and GVPRE/2008/320 of the Valencian Government, CGL2011-25754 of the Spanish Ministry of Economy and Competitiveness, and RNM 8011 of the Junta de Andalucía.
- R. Adrover 1986. Nuevas faunas de roedores en el Mio-Plioceno continental de la región de Teruel (España). Interés bioestratigráfico y paleoecológico. 423 pp. Ph.D. Thesis, Publicaciones del Instituto de Estudios Turolenses, Teruel. Google Scholar
- R. Adrover , P. Mein , and E. Moissenet 1993. Roedores de la transición Mio- Plioceno de la región de Teruel. Paleontologia i Evolució 26–27: 47–84. Google Scholar
- J.P. Aguilar , J. Michaux , and V. Lazzari 2007. Lo Fournas 16-M (Miocène supérieur) et Lo Fournas 16-P (Pliocène moyen), deux nouvelles localités karstiques à Baixas, Sud de la France. Partie II - Nouvelles espèces de rongeurs, listes fauniques et remarque sur l'utilisation biochronologique des faunes karstiques. Géologie de la France 1: 63–81. Google Scholar
- J.P. Aguilar , J. Michaux , J.J. Delanno , and J.L. Guendon 1993. A Late Pliocene rodent fauna from Alozaina (Málaga, Spain). Scripta Geologica 103: 1–22. Google Scholar
- J. Agustí , M. Freudenthal , J.I. Lacomba , E. Martín-Suárez , and C. Nageli 1990. Primeros micromamíferos del Pleistoceno superior de la cuenca de Mula (Murcia, España). Revista de la Sociedad Geológica de España 3: 289–343. Google Scholar
- J. Agustí , A. Santos-Cubedo , M. Furió , R. de Marfá, H.A. Blain , O. Oms , and P. Sevilla 2011. The late Neogene-early Quaternary small vertebrate succession from the Almenara-Casablanca karst complex (Castellón, Eastern Spain): Chronologic and paleoclimatic context. Quaternary International 243: 183–191. Google Scholar
- H. de Bruijn , M.R. Dawson , and P. Mein 1970. Upper Pliocene Rodentia, Lagomorpha and Insectivora (Mammalia) from the isle of Rhodes (Greece). I, II and III. Proceedings van de Koninklijke Nederlandse Akademie van Wetenschappen, Series B 73: 535–584. Google Scholar
- C. Castillo 1990. Paleocomunidades de Micromamíferos de los Yacimientos kársticos del Neógeno Superior de Andalucía Oriental. 255 pp. Unpublished Ph.D. Thesis, University of Granada, Granada. Google Scholar
- J. Chaline 1972. Les rongeurs du Pléistocène moyen et supérieur de France. 410 pp. Unpublished Ph.D. Thesis, Cahiers de Paléontologie, Centre Nationale de la Recherche Scientifique, París. Google Scholar
- G. Cuenca-Bescós , C. Laplana , J.I. Canudo , and J.L. Arsuaga 1997. Small mammals from Sima de los Huesos. Journal of Human Evolution 33: 175–190. Google Scholar
- R. Daams 1981. The dental pattern of the Dormice Dryomys, Myomimus, Microdyromys and Peridyromys. Utrech Micropaleontological Bulletins Special Publication 3: 1–115. Google Scholar
- R. Daams and H. de Bruĳn 1995. A classification of the Gliridae (Rodentia) on the basis of dental morphology. Hystrix 6: 3–50. Google Scholar
- R. Daams and M. Freudenthal 1988. Synopsis of the Dutch-Spanish collaboration program in the Aragonian type area, 1975–1986. In : M. Freudenthal (ed.), Biostratigraphy and Paleoecology of the Neogene Micromammalian Faunas from the Calatayud-Teruel Basin (Spain). Scripta Geologica Special Issue 1: 3–18. Google Scholar
- M. Freudenthal 2004. Gliridae (Rodentia, Mammalia) from the Eocene and Oligocene of the Sierra Palomera (Teruel, Spain). Treballs del Museu de Geologia de Barcelona 12: 97–173. Google Scholar
- M. Freudenthal , P. Mein , and E. Martín-Suárez 1998. Revision of Late Miocene and Pliocene Cricetinae (Rodentia, Mammalia) from Spain and France. Treballs del Museu de Geologia de Barcelona 7: 11–93. Google Scholar
- M. Friant 1953. Une faune du Quaternaire ancien en France méditerranéenne (Sète, Hérault). Annales de la Société Géologique du Nord 73: 161–170. Google Scholar
- A. García-Alix , R. Minwer-Barakat , E. Martín-Suárez , and M. Freudenthal 2008a. Cricetidae and Gliridae (Rodentia, Mammalia) from the Miocene and Pliocene of southern Spain. Scripta Geologica 136: 1–37. Google Scholar
- A. García-Alix , R. Minwer-Barakat , E. Martín-Suárez , and M. Freudenthal 2008b. Muridae (Rodentia, Mammalia) from the Mio-Pliocene boundary in the Granada basin (southern Spain). Biostratrigraphic and phylogenetic implications. Neues Jahrbuch für Geologie und Paläontologie Abhandlungen 248: 183–215. Google Scholar
- A. García-Alix , R. Minwer-Barakat , E. Martín-Suárez , J.M. Martín , and M. Freudenthal 2008c. Biostratigraphy and sedimentary evolution of Late Miocene and Pliocene continental deposits of the Granada Basin (southern Spain). Lethaia 41: 431–446. Google Scholar
- L. Gibert , G.R. Scott , P. Montoya , F.J. Ruiz-Sánchez , J. Morales , L. Luque , J. Abella , and M. Lería 2013. Evidence for an African-Iberian mam mal dispersal during the pre-evaporitic Messinian. Geology 41: 691–694. Google Scholar
- P.A. Goloboff , J.S. Farris , and K.C. Nixon 2008. TNT, a free program for phylogenetic analysis. Cladistics 24: 774–786. Google Scholar
- M.E. Holden 2005. Family Gliridae. In : D.E. Wilson and D.A. Reeder (eds.), Mammal Species of the World. A Taxonomic and Geographic Reference , 4, 819–841. The John Hopkins University Press, Baltimore. Google Scholar
- S. Mansino , N. Fierro , F.J. Ruiz-Sánchez , and P. Montoya 2013. The fossil rodent faunas of the localities Alcoy 2C and 2D (Alcoy Basin, Spain). Implications for dating the classical locality of Alcoy-Mina. Journal of Iberian Geology 39: 261–284. Google Scholar
- E. Martín-Suárez 1988. Sucesiones de micromamíferos en la depresión de Guadix Baza. 241 pp. Unpublished Ph.D. Thesis, University of Granada, Granada. Google Scholar
- P. Mein , E. Moissenet , and R. Adrover 1990. Biostratigraphie du Néogène Supérieur du bassin de Teruel. Paleontologia i Evolució 23: 121–139. Google Scholar
- R. Minwer-Barakat , A. García-Alix , J. Agustí , E. Martín-Suárez , and M. Freu denthal 2009. The micromammal fauna from Negratín-1 (Guadix basin, Southern Spain): new evidence of african-iberian mammal exchanges during the late Miocene. Journal of Paleontology 83: 854–879. Google Scholar
- R. Minwer-Barakat , A. García-Alix , E. Martín-Suárez , M. Freudenthal , and C. Viseras 2012. Micromammal biostratigraphy of the Upper Miocene to lowest Pleistocene continental deposits of the Guadix basin, southern Spain. Lethaia 45: 594–614. Google Scholar
- R. Minwer-Barakat , J. Madurell-Malapeira , D.M. Alba , J. Aurell-Garrido , S. de Esteban Trivigno , and S. Moyà-Solà 2011. Pleistocene rodents from the Torrent de Vallparadís section (Terrassa, Northeastern Spain) and biochronological implications. Journal of Vertebrate Paleontology 31 (4): 849–865. Google Scholar
- P. Montoya , J. Morales , F. Robles , J. Abella , J.V. Benavent , M.D. Marín , and F.J. Ruiz-Sánchez 2006. Las nuevas excavaciones (1995– 2006) en el yacimiento del Mioceno final de Venta del Moro, Valencia. Estudios Geológicos 62: 313–325. Google Scholar
- J. Morales 1984. Venta del Moro: su macrofauna de mamíferos y biostratigrafía continental del Mioceno terminal mediterráneo. 340 pp. Editorial Universidad Complutense, Madrid. Google Scholar
- S. Moreno Garrido 1984. Biometría, Biología y dinámica poblacional del lirón careto, Eliomys quercinus (L), en Doñana, sureste de España. 310 pp. Unpublished Ph.D. Thesis, Universidad de Granada, Granada. Google Scholar
- A. Nadachowski and A. Daoud 1995. Patterns of myoxid evolution in the Pliocene and Pleistocene of Europe. Hystrix 6: 141–149. Google Scholar
- F. Robles 1970. Estudio estratigráfico y paleontológico del Neógeno del río Júcar. 275 pp. Unpublished Ph.D. Thesis, Universitat de València, Valencia. Google Scholar
- C. Sesé 2006b. Micromamíferos (Rodentia, Insectivora, Lagomorpha y Chiroptera) de la Peña de Estebanvela (Segovia). In : S. Ripoll Cacho and F. Muñoz (eds.), La Peña de Estebanvela (Estebanvela-Ayllón, Segovia). Grupos Magdalenienses en el Sur del Duero, Memorias, Junta de Castilla y León. Arqueología en Castilla y León 17: 145–165. Google Scholar
- C. Sesé and P. Villa 2008. Micromammals (rodents and insectivores) from the early Late Pleistocene cave site of Bois Roche (Charente, France): Systematics and paleoclimatology. Geobios 41: 399–414. Google Scholar
- C. Sesé , J. Panera , S. Rubio-Jara , and A. Pérez-González 2011. Micromamíferos del Pleistoceno Medio y Pleistoceno Superior en el Valle del Jarama: yacimientos de Valdocarros y HAT (Madrid, España). Estudios Geológicos 67: 131–151. Google Scholar
- A. van de Weerd 1976. Rodent faunas of the Mio-Pliocene continental sediments of the Teruel-Alfambra region, Spain. Utrech Micropaleontological Bulletins, Special Publication 2: 1–217. Google Scholar
- A. Wagner 1840. Beschreibung einiger neuer Nager, welche auf der Reise des Herrn Hofraths v. Schubert gesammelt wurden, mit Bezugnahme auf einige andere verwandte Formen. Abhandlungen der Mathematisch- Physikalischen Classe der Königlich Bayerischen Akademie der Wissenschaften 3: 175–218. Google Scholar
Dental characters included in the cladistic analysis.
Lower molars: 1, Average value of length/width ratio of m1, m2: (0) always less than 1, molars are clearly wider than long; (1) variably above or below 1, molars are sub-quadrangular to sub-rectangular in occlusal outline; (2) always higher than 1, molars are clearly longer than wid. 2, Endolophid: (0) always discontinuous; (1) continuous in some specimens. 3, Anterotropid: (0) always absent; (1) present in some specimens. 4, Posterotropid: (0) always absent; (1) present in some specimens. 5, Centrolophid: (0) absent or short (less than half the width of the molar); (1) usually long (half the width of the molar or longer).
Upper molars: 6, Presence of both centrolophs on M1, M2: (0) absent; (1) rare (less than 30%); (2) common (30% or higher). 7, Presence of both centrolophs on M3: (0) absent; (1) rare (less than 30%); (2) common (30% or higher). 8, Presence of an accessory crest in the upper molars: (0) absent; (1) present in some specimens. 9, Presence of two accessory crests in some upper molars; (0) absent; (1) present in some specimens. 10, Size: (0) clearly smaller on average than E. truci, with some dental elements out of the size range of this species (see Fig. 3); (1) equal or largely the same as E. truci; (2) clearly bigger on average than E. truci, with some dental elements out of the size range of this species. See Table 3 for the resulting data matrix.
Repository numbers of the specimens listed in Table 1.
p4: VVmA-42 (MGUV-27256), VVmA-43 (MGUV-27257).
m1/m2: VVmA-1 to VVmA-11 (MGUV-27215 to MGUV-27225); VVmB-1 (MGUV-27405); VVmBC-1 (MGUV-27487) and VVmBC-2 (MGUV-27488); ABS3A-38 (MGUV-23529); AF1'06-190 (MGUV-28334); LB2B-176 and LB2B-242 (MGUV-28727 and MGUV-28793); AL2D-290 (MGUV- 28063); AL2C-97 (MGUV-28924).
m3: VVmA-33 to VVmA-38 (MGUV-27247 to MGUV-27252); VVmC-1 (MGUV-27487); VVmBC-5 (MGUV-27490); LB2B-179 (MGUV-28730); AF1'06-187 (MGUV-28331); AC0-88 (MGUV-28551); AL2C 97 (MGUV-28924); AL2D- 356 (MGUV-28129).
D4: VVmA-44 (MGUV-27258).
P4: VVmA-39 to VVmA-41 (MGUV-27253 to MGUV-27255); LB2B-178 (MGUV-28729).
M1/M2: VVmA-16 (MGUV-27230); VVmA-21 to VVmA-31 (MGUV-27235 to MGUV-27245); VVmBC-3; VVmD-1 and VVmD-2 (MGUV-27543 and MGUV-27544); AF1'07- 36 (MGUV-28427); AC0-87 (MGUV-28550); LB2B-241 (MGUV-28792); AL2D-291 (MGUV-28064); AL2D-293 (MGUV-28066).
M3: VVmA-12 to VVmA-15 (MGUV-27226 to MGUV-27229); VVmA-17 to VVmA-20 (MGUV-27231 to 27234); VVm- BC-4 (MGUV-27489, LB2B-177 (MGUV-28728).