Quaternary International xxx (2011) 1e12 Contents lists available at ScienceDirect Quaternary International journal homepage: www.elsevier.com/locate/quaint Phylogeny and biogeography of fossil and extant Microtus (Terricola) (Mammalia, Rodentia) of Sicily and the southern Italian peninsula based on current dentalmorphological data Daria Petruso a, Elisa Locatelli b, *, Giovanni Surdi a, Chiara Dalla Valle b, Federico Masini a, Benedetto Sala b a b Department of Geology and Geodesy, Via Archirafi 22, 90123, Palermo, Italy Department of Biology and Evolution, C.so Ercole I d’Este 32, 44100 Ferrara, Italy a r t i c l e i n f o a b s t r a c t Article history: Available online xxx The fossil record of the Savi vole, Microtus (Terricola) savii, is analyzed in terms of morphological and morphometrical variability of the first lower molar, in order to reconstruct phylogenetic relationships between insular and mainland populations and the dispersal events. The present work gives a contribution to better understand the phylogenetic history of this taxon in Sicily and Southern Italy during the interval Middle Pleistocene e Holocene, in an attempt to reconstruct the relationships between insular and continental voles and to clarify some paleobiogeographical aspects. The morphometrical data have been acquired by traditional measurements of the first lower molar and analyzed by dispersal diagrams and Principal Components Analysis. The fossil Sicilian samples have a greater size variation than extant ones. Two different morphological groups have been identified and named on the basis of the similarity with European or Italian populations: a “subterraneomorph” one (characterized by a tighter symmetric anterior cap and longer anteroconid), similar to European M. (T.) subterraneus species, and a “savimorph” one (more confluent and asymmetric anterior cap and shorter anteroconid), occasionally with the accentuation of morphological characters of M. (T.) savii. The results, in particular the differences between Pleistocene and Holocene Sicilian populations and the similarities with the mainland ones, suggest that the Savi vole dispersed at least twice in Sicily. One colonization took place likely during the cold stage MIS 6 (recorded at Isolidda 3) with dispersal events, made possible by the sea level drop and the connection with the mainland, and a second one (documented at San Teodoro cave and other Holocene assemblages) during MIS 4 (by dispersal events) or MIS 3 (by accidental transit or limited faunal exchanges). Ó 2011 Elsevier Ltd and INQUA. All rights reserved. Keywords: Microtus (Terricola) savii Sicily Pleistocene Holocene Insularism Paleobiogeography 1. Introduction Microtus (Terricola) savii is widespread all over Italy, with the exception of the northeast and Sardinia, but many questions about its origin and evolution are still unsolved. According to Tougard et al. (2008), M. (T.) savii originated in Italy and it has close relationships with M. (T.) multiplex. Four subspecies are recognized nowadays: Microtus (Terricola) savii savii (de Sélys Longchamps, 1838), in Northern and Central Italy; Microtus (Terricola) savii nebrodensis (Minà Palumbo, 1868), in Sicily; Microtus (Terricola) savii tolfetanus (Contoli, 2003) in Latium; and Microtus (Terricola) * Corresponding author. E-mail addresses: elisa.locatelli@unife.it, elisal40@hotmail.com (E. Locatelli). savii niethammericus (Contoli, 2003) in Basilicata, Abruzzo and Apulia. The occurrence of a distinct species Microtus (Terricola) brachycercus (von Lehmann, 1961), closely related to the Savi vole, is recognized in the Calabria peninsula, based on karyological features (Galleni et al., 1992, 1994, 1998; Galleni, 1995; Wilson and Reeder, 2005). The phylogenetic distance of the Northern and Central Italian populations of M. (T.) savii from M. (T.).brachycercus has been confirmed by molecular analyses by Jaarola et al. (2004) and Castiglia et al. (2008). The latter authors found also a greater affinity of the Calabria species with populations referred to M. (T.) savii niethammericus by Contoli (2003). These results lead Contoli and Nappi (2009) to place the subspecies “niethammericus” within the species M. (T.) brachycercus. The morphological distinction between the Calabrian voles (namely from Aspromonte) 1040-6182/$ e see front matter Ó 2011 Elsevier Ltd and INQUA. All rights reserved. doi:10.1016/j.quaint.2011.03.013 Please cite this article in press as: Petruso, D., et al., Phylogeny and biogeography of fossil and extant Microtus (Terricola) (Mammalia, Rodentia) of Sicily and the southern Italian peninsula based on current dentalmorphological data, Quaternary International (2011), doi:10.1016/ j.quaint.2011.03.013 2 D. Petruso et al. / Quaternary International xxx (2011) 1e12 and Southern Italian, particularly Sicilian, populations, was first noted by De Luca (1998) and illustrated by Petruso (2003) based on measurements and morphological indexes of the first lower molar (M1). On the other hand, “small differences” between the Calabrian vole and vole samples from Northern, Central and Southern Italy were detected by Nappi et al. (2005) using Canonical Discriminant Analysis on a set of linear measurement of the M1. These authors concluded that if the Calabrian vole is proved to be a distinct species from the Savi vole, separation must have been achieved recently. Multivariate morphological analysis carried on first lower molars by Piras et al. (2010), based on geometrical morphometric methods, apparently confirms the morphological distance of Calabrian populations from some samples of M (T.) savii from Southern Italy and two samples from Central and Northern Italy. Nappi et al. (2006), working with the methodology of Nappi et al. (2005) on some 50 samples distributed all over Italy and Sicily (but not Calabria), detected a morphological differentiation between north-central and southern populations of Italy. The authors infer the existence of a latitudinal gradient (a morphocline) based on the analogies between the intraspecific differences within the Italian populations and the interspecific differences within the Western European group - between the northern multiplex-subterraneus group (sensu Chaline, 1972) and the southern Microtus (Terricola) duodecimcostatus (de Sélys Longchamps, 1839). Conversely, Piras et al. (2010) deny the significance of differences between central and southern populations and claim the influence of a “more complex climatic effect” and the importance of phylogeographic relationships in modeling the variability among Microtus (Terricola) populations and species. The same authors in a previous paper (Piras et al., 2009) had already sustained some correlations between the morphological variation of the anterior portion of the first lower molar and the climatic fluctuations as expressed by the d18O climatic proxy. Even though the extant vole populations from Sicily were referred to M. (T.) savii nebrodensis in the critical revision of Contoli (2003), diverging opinions are expressed in the literature concerning their actual phylogenetic relationships, taxonomy and origin. Castiglia et al. (2008) identified a noteworthy distance in Cytochrome b haplotypes (7%) between some specimens from Ficuzza (Palermo, western Sicily) and those of the savii e brachycercus group, suggesting a possible distinct specific status for the former population. On the other hand, Nappi et al. (2006) on the basis of morphological analysis, considered that the Sicilian Microtus (Terricola) falls in the southern group of savii, having, however, some distinctive features in the anterior part of the first molar. Conversely, Piras et al. (2010) did not detect any significant difference between the Sicilian populations and the Savi vole from the Italian peninsula. Species of the subgenus M. (Terricola) are the fossil voles present, locally in great abundance, in many Sicilian deposits, since the late MiddleeLate Pleistocene. However, until the 1990s, information on its occurrence was limited to a specimen from La Costiera (PA; Burgio and Kotsakis, 1987). Previously, limited attention was given by paleontologists to non-endemic small mammals of Sicily and very little (if any) sediment from the 170 Quaternary deposits known on the island had been sieved for collecting microfauna (Petruso et al., 2008). Even fewer works, specifically addressed to the detailed analysis of the Savi vole morphology, have been produced to clarify the relationships between the Sicilian populations and the peninsular ones. The first finding, reported by Burgio and Kotsakis (1987) from the site of La Costiera, consisted of a unique remain (a mandible fragment) referred to as “Pitimys¼Microtus (Terricola) cf. savii” and doubtfully ascribed to the Late Glacial Castello Faunal Complex (FC). Subsequently, Tagliacozzo (1993) documented the diffusion of Microtus (Terricola) in the early Holocene deposit of Uzzo cave (San Vito lo Capo, TP) and confirmed its occurrence in pre-Holocene (Late Glacial) levels. Bonfiglio et al. (1997) described samples from two new fossil sites and referred them to M. (T.) ex gr. savii: the Late Pleistocene sample from Contrada Pianetti (RG), later referred by Bonfiglio et al. (2002) to the Grotta di San Teodoro e Pianetti FC (MIS4-2), and a sample from the Late Pleistocene layer of site K22 (San Vito lo Capo, TP). The authors stressed the differences between the two fossil populations and suggested they could have derived from two distinct dispersal events from the continent, considering the one that gave rise to the Contrada Pianetti vole as older. The first comparative study of the extant Savi vole from Sicily, an extant population from Aspromonte and the Late Pleistocene sample from site K22, was carried on shortly after (De Luca, 1998). In the comparison, all the analyzed populations were attributed to the savii group, with the exception of the sample from Calabria, which showed morphological features closer to the multiplex-subterraneus group. A deeper and more extended comparative analysis of the fossil material of this species from more numerous Late PleistoceneeEarly Holocene Sicilian sites (San Teodoro Cave, Acquedolci, ME; Uzzo Cave, Castello shelter, PA; K22) was carried on by Petruso (2003), who attributed all the Sicilian samples to the savii group and detected a more complex and assorted coexistence of different characters in the Maltese sample from the Last Glacial deposit of Ghar Dalam (Malta; Storch, 1974). The comparison noted the wide size variation among the populations, ascribable to different micro-environmental features. Locatelli (2007) analyzed the late Middle Pleistocene-Late Pleistocene vole remains from Isolidda 3 cave (San Vito lo Capo, TP), and the results show a marked difference between these old Sicilian populations and the aforementioned younger ones. A detailed study of the San Teodoro vole remains is reported in Bonfiglio et al. (2008), pointing out the morphological differences between samples coming from different sectors and different layers of the cave. Despite the efforts of the last 10 years, the details of the dispersal history of this vole in Sicily and the question if it reached the island only one time, and then evolved in Sicily, with sporadic contacts with the mainland populations, or if it reached the island more than once, is still particularly challenging. Morphological divergence in respect to the peninsular populations due to endemic processes is expected as a result of geographical isolation, while greater similarities with the peninsular samples of different ages would be indicative of possible connections with the mainland. The aim of this study is to detect, by the morphometrical analysis of the first lower molar of Microtus (Terricola) ex gr. savii of extant and fossil Sicilian populations and their comparison with peninsular and Western European samples, the possible similarities between different populations that could suggest a strict phylogenetic relation. Thus the dispersal events of this vole during late MiddleeLate Pleistocene could be outlined, hypothesizing the presence of temporary land bridges with the mainland, also forming a contribution to the reconstruction of the Sicilian paleobiogeography. 2. Materials and methods The systematic study was carried on by means of morphometric analysis on voles’ first lower molars from 22 localities, including 13 fossil deposits and 9 extant populations. The location of the sites is shown in Fig. 1. The total number of samples examined was 34, as in some fossil sites with a clear stratigraphic succession (Isolidda 3 cave, Cavallo cave, Grotta Grande di Scario cave, Uzzo cave and San Teodoro cave), specimens from different levels have been analyzed separately. The samples used in the analysis have been chosen for both chronological and geographical reasons. With regard to Sicilian samples, all the known fossil sites with Microtus (Terricola) have Please cite this article in press as: Petruso, D., et al., Phylogeny and biogeography of fossil and extant Microtus (Terricola) (Mammalia, Rodentia) of Sicily and the southern Italian peninsula based on current dentalmorphological data, Quaternary International (2011), doi:10.1016/ j.quaint.2011.03.013 D. Petruso et al. / Quaternary International xxx (2011) 1e12 3 Fig. 1. Location map. been considered (with the exception of La Costiera unique remain, which has been lost), while the extant populations come from localities with different geographical features (some are located near coastal areas, and others close to mountain chains). For the Italian peninsula samples, sites coeval with the Sicilian ones and located in different regions of Italy have been considered, in order to separate geographical data from chronological ones. Eventually, two extant populations belonging to Microtus (Terricola) subterraneus (France) and Microtus (Terricola) lusitanicus (Spain) have also been considered as a sort of “out-group” for comparisons. The list of the localities, fossil and extant ones, together with the age of their deposits (single or distinguished into different levels), the distinctive code used in the diagrams, the number of specimens of each sample and the references for the chronology of fossil sites, are reported in Table 1. Measurements have been taken according to the plan proposed by Van der Meulen (1973) and modified by Masini (in Bonfiglio et al., 1997; Masini et al., 2005) with the integration of some measures according with the Brunet-Lecomte and Chaline (1992) scheme (Fig. 2) with a Leica MZ8 orthoplan microscope. At Ferrara university it was provided with a Parker 4508DM positioning stage (straight-line accuracy 2 mm). The analysis has been carried out by bivariate diagrams using the measurements and morphological indexes computed from the single measures such as: AL (A/L*100), A2A (A2/A*100), BW (B/ W*100), CW (C/W*100), DW (D/W*100), DE (D/E*100), L45 (L4/ L5*100). To trace the evolutionary pattern and to find out possible relations between different populations, the mean values of each measurement or morphological indexes are plotted on the abscissa, while on the vertical axis discrete intervals of time (time slices), related to important paleoclimatic events are reported. Principal Component Analysis (PCA) with correlation metric has been carried out on the means of measurements and indexes of the considered samples using PAST software (Hammer et al., 2001). 3. Results The analysis demonstrated that extant Sicilian populations are fairly separate from continental ones. Fossil populations show a more diversified pattern. 3.1. Size variation Fig. 3 illustrates the variation in size related both to time and to geography inferred by the length of M1. Most of the fossil Sicilian populations are on the average larger than the continental ones. The greatest size was reached during MIS 3 (San Teodoro trench 03e04 and square A B C, TEO0304ABC), the Late Glacial (Castello Shelter e RC) and especially in the ancient Holocene (Uzzo Mesolithic II and Neolithic levels, UZ mII e UZn). However, very small sized or “normal” sized samples (namely Contrada Pianetti, Oriente Cave and San Teodoro square D) occur among the fossil Sicilian voles of these time slices. In older samples, the size difference between insular populations (Isolidda 3) and continental ones (Cavallo cave, Grotta Grande di Scario) seems to be less relevant, even though Sicilian samples are already slightly larger. 3.2. Morphological variation Beside the size variation, the analysis demonstrated important morphological variation in tooth shape. The relative proportions of the anteroconid (ACC, the anterior portion of the first lower molar, Fig. 2) respect to the total length of M1 have been investigated by measures A, A2, L and the related indexes, AL and A2A. It is possible to notice a difference between insular and continental populations (Figs. 4 and 5); the latter have longer anteroconid complex (A) and anterior cap (A2) relative to the Sicilian Savi voles. The extant Savi sample from Melendugno, in the Salento peninsula, represents an exception, as it is closer to Sicilian populations than to peninsular ones. As the ACC represents the most variable part of the first lower molar, this part of the tooth is particularly important for the taxonomy of the group and its morphology has been investigated more in the details. In order to analyze the degree of constriction of the triangles, three parameters (D, E, W) and their indexes have been used (Figs. 6 and 7). The variable D measures the degree of constriction of the anterior cap; E describes the width of the triangles T6 and T7 that in the maximum development shapes a rhombus (Fig. 2); W conveys the width of the next pair of triangles (T5eT4), the so-called pitymyan rhombus (Fig. 2), peculiar for the morphology of Microtus (Terricola). Peninsular populations are Please cite this article in press as: Petruso, D., et al., Phylogeny and biogeography of fossil and extant Microtus (Terricola) (Mammalia, Rodentia) of Sicily and the southern Italian peninsula based on current dentalmorphological data, Quaternary International (2011), doi:10.1016/ j.quaint.2011.03.013 4 D. Petruso et al. / Quaternary International xxx (2011) 1e12 Table 1 Localities. List of the localities, fossil and extant ones, together with the age of their deposits (single or distinguished into different levels), the distinctive code used in the diagrams, the amount of each sample and the literature used as reference for the chronology of fossil sites. Country Area Locality Code Taxon Age N France Spain Italy Italy Italy Italy Italy Italy Italy Italy Apulia Tuscany Tuscany Calabria Sicily Sicily Sicily Sicily Sarcey-Rhone (FR) Lamilla-Burgos (E) Melendugno (LE) San Miniato (PI) Cascina (PI) Aspromonte (RC) Borgo Lupo (TP) Piana di Gela (CL) Mazzarino (CL) Uzzo Cave (TP) sub lus Mel Smin Cas Asp BLu Gel Maz UZmI M. (T.) subterraneus M. (T.) lusitanicus M. (T.) savii M. (T.) savii M. (T.) savii M. (T.) s. brachycercus M. (T.) s. nebrodensis M. (T.) s. nebrodensis M. (T.) s. nebrodensis M. (T.) ex gr. savii 5 27 32 38 45 30 22 26 12 39 Tagliacozzo (1993) Italy Sicily Uzzo Cave (TP) UZmII M. (T.) ex gr. savii 32 Tagliacozzo (1993) Italy Sicily Uzzo Cave (TP) UZn M. (T.) ex gr. savii 15 Tagliacozzo (1993) Italy Italy Sicily Sicily Cala Mancina (TP) Castello shelter (PA) MAN RC M. (T.) ex gr. savii M. (T.) ex gr. savii 60 6 Martini et al. (in press-b) Petruso (2003) Italy Sicily OR M. (T.) ex gr. savii 41 Martini et al. (in press-a) Italy Italy Campania Apulia Oriente Cave -Favignana island (TP) Serratura Cave (SA) Cavallo Cave (LE) layer B SER CavB M. (T.) ex gr. savii M. (T.) ex gr. savii Extant Extant Extant Extant Extant Extant Extant Extant Extant Early Holocene Mesolithic 1 Early Holocene Mesolithic 2 Early Holocene Neolithic Early Holocene Late Glacial - Holocene (MIS2e1) Late Glacial (MIS2) Late Glacial (MIS2) Late Glacial (MIS2) 96 29 Italy Sicily K22 (TP) K22 M. (T.) ex gr. savii Würm (MIS3e2?) 35 Italy Sicily TEO98 M. (T.) ex gr. savii Würm (MIS3) 22 Italy Sicily TEOABC M. (T.) ex gr. savii Würm (MIS3) 12 Bonfiglio et al. (2008) Italy Sicily TEOD M. (T.) ex gr. savii Würm (MIS3) 16 Bonfiglio et al. (2008) Italy Apulia S.Teodoro Cave (ME) trench 1998 S.Teodoro Cave (ME) trench 2003e04 ABC sectors S.Teodoro Cave (ME) trench 2003e04 D sector Cavallo Cave (LE) layer F Martini (1993), Bertolini et al. (1996) Palma di Cesnola and Borzatti von Löwestern (1964) Petruso (1996), De Luca (1998), Di Maggio et al. (1999) Bonfiglio et al. (2008) CavF M. (T.) ex gr. savii Würm (MIS3) 45 Italy Malta Italy Sicily Contrada Pianetti (RG) Ghar Dalam Cave (Malta) Grotta Grande di Scario (SA) trench A CP GD GGSA M. (T.) ex gr. savii M. (T.) melitensis M. (T.) ex gr. savii Würm (MIS3) Würm? Post-Tyrrhenian (MIS5d-a) 31 28 18 Italy Campania Grotta Grande di Scario (SA) trench F GGSF M. (T.) ex gr. savii Post-Tyrrhenian (MIS5d-a) 24 Italy Apulia Cavallo Cave (LE) layer M CavM M. (T.) ex gr. savii 45 Italy Apulia San Sidero 3 (LE) SS3 M. (T.) sp. 24 De Giuli (1983) Italy Italy Apulia Campania Cavallo Cave (LE) layer N Grotta Grande di Scario (SA) trench C lower CavN GGSCi M. (T.) ex gr. savii M. (T.) ex gr. savii Post-Tyrrhenian (MIS5d-a) Post-Tyrrhenian (MIS5 s.l.) Eemian (MIS5e) Pre-Tyrrhenian (MIS6) Palma di Cesnola and Borzatti von Löwestern (1964) Bonfiglio et al. (1997) Storch (1974) Abbazzi and Masini (1998), Boscato and Ronchitelli (2004), Ronchitelli et al. (1998, 2008) Abbazzi and Masini (1998), Boscato and Ronchitelli (2004), Ronchitelli et al. (1998, 2008) Sarti et al. (2002) 33 48 Italy Campania Grotta Grande di Scario (SA) trench C upper GGSCs M. (T.) ex gr. savii Pre-Tyrrhenian (MIS6) 25 Italy Italy Italy Sicily Sicily Sicily Torre Isolidda 3 (TP) US 15, lower Torre Isolidda 3 (TP) US 13, middle Torre Isolidda 3 (TP) US 12, upper TI3i TI3m TI3s M. (T.) sp. M. (T.) sp. M. (T.) sp. Pre-Tyrrhenian? Pre-Tyrrhenian? Pre-Tyrrhenian? 42 45 45 Sarti et al. (2002) Abbazzi and Masini (1998), Boscato and Ronchitelli (2004), Ronchitelli et al. (1998, 2008) Abbazzi and Masini (1998), Boscato and Ronchitelli (2004), Ronchitelli et al. (1998, 2008) Martini et al. (in press-c) Martini et al. (in press-c) Martini et al. (in press-c) Campania well separated from insular ones, with the exception of the oldest samples, as Sicilian (Isolidda 3) and peninsular (Scario) samples are closer to each other. From MIS 3 to Holocene, all Sicilian populations are very homogeneous in the EW diagram (Fig. 6), showing slight variation as do the living ones, but having smaller values (T6eT7 rhombus respect to the pitymyan rhombus relatively smaller than in the continental populations). On the other hand, the anterior complex of Sicilian populations since MIS 3 underwent an opening (DE index), with the higher value found in the living populations, thus increasing the distance from the northern samples that maintain rather small values. Considering all samples together, although EW does not show a great variation (standard deviation of the mean values of EW of each sample is 2.64), DE is Reference for the chronology of fossil sites highly variable (8.85) and extant northern populations are clearly separated from the Sicilian ones. The Apulian living population from Melendugno plots with Sicilian samples, while the Aspromonte Microtus (Terricola) brachycercus is always within the multiplex-subterraneus group and the two Central Italy Savi populations. A different pattern is shown by pre-Tyrrhenian samples. The Isolidda 3 populations are close to the contemporary continental samples from Grotta Grande di Scario, both in the EW and in DE diagrams. Fossil Apulian populations from levels M and N of Cavallo cave are well isolated from all the other ones in showing plesiomorphic features (namely, high values of DE and small values of EW). Please cite this article in press as: Petruso, D., et al., Phylogeny and biogeography of fossil and extant Microtus (Terricola) (Mammalia, Rodentia) of Sicily and the southern Italian peninsula based on current dentalmorphological data, Quaternary International (2011), doi:10.1016/ j.quaint.2011.03.013 D. Petruso et al. / Quaternary International xxx (2011) 1e12 5 Fig. 2. Nomenclature and measures. (a) Nomenclature of dental parts of the first lower molar in occlusal view. (b) Plan of measurements, according to Van der Meulen (1973) and modified by Masini (in Bonfiglio et al., 1997). (c) Plan of measurements, according to Brunet-Lecomte and Chaline (1992) scheme in Masini et al. (2005). 3.3. PCA The PCA (Fig. 8) considers the first two components that account for 75% of total variance (47.14% and 28.17% respectively, Table 2). Interpretation of the morphological meaning of components 1 and 2 is not particularly straightforward because most variables are correlated to both axes (see factor loadings in Table 3). Variables expressing the size of teeth are loaded positively on the first component, and negatively on the second. Variation in size therefore ranges along the diagonal with larger teeth scattering mostly in the II square of the diagram. The variables and indexes that account for the elongation of the ACC (including AC2) are negatively correlated on both axes, so that samples with more elongated ACC tend to scatter in the III square. The same is true for E and EW, while D, DE and DW are positively correlated to both components Fig. 3. Length. Dispersion diagram illustrating the variation in size of the first lower molar in time. The abscissa represents the mean value of L, the ordinate discrete intervals of time. Sicilian samples are grouped inside the boxes. so that teeth with an open AC2 tend to scatter in the first square. Measure B and index BW also are both positively correlated to components 1 and 2, but they are more weighted on the first one. Therefore, teeth that tend to be smaller, with elongated ACC and AC2, narrow AC2, wide T6eT7 rhombus and narrow constriction between the pitymyan rhombus and AC2, are distributed on the lower left side of the diagram, where M. (T.) subterraneus and M. (T.) lusitanicus samples are placed. Thus, the PCA confirms the previous results. Populations of the same area and the same age tend to plot together in the PC1e2 diagram, with the exception of the Apulian samples. With regard to extant populations, samples from Calabria and Central Italy are Fig. 4. AL. Dispersion diagram illustrating the variation in time of the anteroconid length compared to total length. The abscissa represents the mean value of the index AL, the ordinate discrete intervals of time (the legend of time slices is reported in Table 2). Sicilian samples are grouped inside the boxes. Please cite this article in press as: Petruso, D., et al., Phylogeny and biogeography of fossil and extant Microtus (Terricola) (Mammalia, Rodentia) of Sicily and the southern Italian peninsula based on current dentalmorphological data, Quaternary International (2011), doi:10.1016/ j.quaint.2011.03.013 6 D. Petruso et al. / Quaternary International xxx (2011) 1e12 Fig. 5. A2A. Dispersion diagram illustrating the variation in time of the relative proportion of the anteroconid. The abscissa represents the mean value of the index A2A, the ordinate discrete intervals of time. Sicilian samples are grouped inside the boxes. clearly separated from those from Sicilian localities and from Apulia. The first are close to the multiplex-subterraneus voles, while the others tend to disperse on the opposite side. Fossil samples from Isolidda 3 are rather homogeneous and disperse close to Calabrian, Central Italian and subterraneus voles. Fossil samples from Grotta Grande di Scario are more scattered, and plot rather close to Isolidda 3 and the peninsular samples. The Sicilian populations of younger age (MIS2e3) are rather isolated (high degree of endemism), plotting alone on the positive part of Axis 1 (47.14% Fig. 7. DE. Dispersion diagram illustrating the variation in time of the closing of the anterior cap. The abscissa represents the mean value of the index DE, the ordinate discrete intervals of time. Sicilian samples are grouped inside the boxes. variance) and the negative part of Axis 2. The sample of Microtus (Terricola) from Malta stands isolated in the same square and is characterized by extremely negative values of Component 2. The fossil sample for the Hyblean Plateau (Contrada Pianetti) is also separate from the other Sicilian samples, falling very close to the continental fossil populations. Apulian samples from Cavallo cave are the most scattered, notwithstanding that they come from a single geographic location. In particular, there is a great difference between the ancient populations (CavM and CavN), which show very marked “savimorph” characters, and the more recent ones (CavF and CavB), which seem to be closer to the peninsular samples. San Sidero 3 plots in an intermediate position among extant Apulian Terricola, fossil Savi vole from Grotta Grande di Scario and the less “endemic” Würmian Sicilian samples. 4. Discussion The morphometric study of Southern Italian fossil Microtus (Terricola) carried on by comparative analysis produced a series of results and raised some questions. 4.1. Size variation Fig. 6. EW. Dispersion diagram illustrating the variation in the width of the anterior cap related to the width of the pytimian complex. The abscissa represents the mean value of the index EW, the ordinate discrete intervals of time. Sicilian samples are grouped inside the boxes. The comparison of the length of the first lower molar gives an immediate idea of the body size variation occurring in Sicilian samples and in the southern peninsula ones. Mammals in a “normal” continental regime may reduce or increase their size, always to a slight extent, mainly in relation to variation in climatic and environmental conditions (Bergmann rule), while on islands, body size usually varies in one direction only and its variation is ruled mainly by several ecological factors typical of the island environment (e.g. area of the island, availability of resources, reduced predatory pressure etc.; Lomolino, 2005). By and large, according to the socalled “island rule” (Foster, 1964; Van Valen, 1973; Sondaar, 1977), geographic isolation on islands is expected to produce “giant” small mammals and “dwarf” large mammals, i.e. small mammals should acquire a larger size respect to their mainland counterparts, while the reverse is expected in large mammals. The increase in size of Please cite this article in press as: Petruso, D., et al., Phylogeny and biogeography of fossil and extant Microtus (Terricola) (Mammalia, Rodentia) of Sicily and the southern Italian peninsula based on current dentalmorphological data, Quaternary International (2011), doi:10.1016/ j.quaint.2011.03.013 7 D. Petruso et al. / Quaternary International xxx (2011) 1e12 Fig. 8. PCA. Principal Component Analysis on the symmetric correlation matrix, Joliffe cut-off 0.7. Samples from different periods and different areas are grouped inside the boxes. The explanation for the abbreviations is reported in Table 1. small mammals may be explained by competitive interactions, that is by the benefit that a larger size gives in the individual exploitation of food supply in consideration of the small area of the island; the bigger they are, the better they prevail in the competition within the populations of the same species and/or with individuals of the other species. Sicily is a sort of limiting case for this rule, as it is a large island (the largest in the Mediterranean Sea) and it is separated from the mainland only by a sea corridor, the Strait of Messina. This particular geographical and geological configuration made the isolation unstable, at least from the Late Middle Pleistocene onwards, thus reducing the above-mentioned effects on the faunal association and the body size variation. Considering the mean value of the length (L) of each sample, Sicilian Savi voles are on average some 5% (computed on the mean L of each sample) larger than the peninsular populations of Terricola, but a distinction among samples of different age has to be made. The extant Sicilian populations are rather variable in size, but they are quite small, about the same size as the extant mainland ones. This homogeneity between insular and mainland populations could be due to frequent human and faunal exchanges between the island and the mainland, and to the influence of humans on the insular mammal ecology (with the introduction of foreign taxa, terrestrial predators, habitat fragmentation for agriculture purposes and domestication) that has reduced, perhaps annulled, the insular condition and consequently the endemic evolutionary pressure (Sarà, 1998; Petruso et al., in press). A rather different situation can be observed in populations of the Early Holocene, when the Sicilian Microtus (Terricola) reaches its maximum size, and in samples from Table 2 PCA eingenvalues and variance. The eigenvalues and the percentage of variance of the first 10 components of the PCA analysis are reported. PC Eigenvalue % Variance 1 2 3 4 5 6 7 8 9 10 8.9559 5.35137 1.33966 1.10228 0.838771 0.565951 0.302736 0.235157 0.104046 0.0813695 47.136 28.165 7.0508 5.8015 4.4146 2.9787 1.5933 1.2377 0.54761 0.42826 Late Glacial and MIS3, where the difference with the coeval peninsular samples is 7.88% (computed among the means of L of each sample). Such a large size is in accordance with the general trend of size increase predicted for insular small mammals by the “island rule”. Even if this is not a true gigantism episode, the increment in size is conspicuous and is likely to be the result of isolation, perhaps due to the reduced interspecific competitive pressure, as the small mammal community is poor and other species of arvicolids are absent on the island. It should be noted, however, that the MIS3 time slice includes also two smaller-sized populations, from San Teodoro D (dated older than 33 ka) and, even more so, from Contrada Pianetti, which are in the range of extant populations of the Savi vole from Sicily and from the peninsula. A climate effect on the size variation can be excluded for several reasons. Firstly, extant samples from Tuscany are of the same size as those from Sicily, even though climatic conditions are different. Secondly, samples from MIS3 (younger levels of San Teodoro cave) and samples from the early Holocene at Uzzo Cave (UZ mII, UZ nI), have more or less the same size, even though climatic conditions are expected to be much different. Furthermore, a direct correlation between climate and body size has been excluded also by Piras et al. (2010), based on the study of extant populations. On the other Table 3 PCA factor loadings. Component 1 L A A2 W C B E D WT L4 L5 AL A2A BW CW DW EW DE L4L5 0.2905 2.40E-01 0.1044 0.2588 0.2757 2.48E 01 0.1614 0.2651 0.274 0.3117 0.3143 0.1395 0.167 0.2182 0.1651 0.2013 0.1336 0.2003 0.234 Component 2 0.172 0.2765 0.3831 0.2183 0.04708 0.1046 0.3271 0.2116 0.1295 0.09115 0.04492 0.3323 0.2864 0.1642 0.08125 0.3131 0.2487 0.3244 0.1283 Please cite this article in press as: Petruso, D., et al., Phylogeny and biogeography of fossil and extant Microtus (Terricola) (Mammalia, Rodentia) of Sicily and the southern Italian peninsula based on current dentalmorphological data, Quaternary International (2011), doi:10.1016/ j.quaint.2011.03.013 8 D. Petruso et al. / Quaternary International xxx (2011) 1e12 hand, even in peninsular populations a very high variation in size occurs between samples from different geographical locations (e.g. Serratura Cave, Campania on the Tyrrhenian side, and Cavallo Cave layer B, Southern Salento, at the Ionian Sea), confirming that different ecological conditions may influence body size. When considering the older samples (time slices 6, 7 and 8, that is MIS 5 aed, MIS 5 e and MIS 6) the Sicilian samples from Isolidda 3 (TI3), here referred to MIS6 (?), are smaller than those of MIS 3 e Early Holocene, and have roughly the same size of extant Sicilian and peninsular Savi voles. When compared to samples of the same age, they are close to or even smaller than the peninsular ones (Grotta Grande di Scario). At Isolidda 3 the vole occurs after the disappearance of the endemic dormouse Leithia sp., but together with very endemic small mammals such as Maltamys and Crocidura, forming a poorly diversified assemblage that is suggestive of isolated conditions. In such a context, the small “normal” size of the vole might suggest the absence of endemism. Isolidda populations were rather close in time to the dispersal event of this species on the island, and therefore they had not yet achieved endemic features. 4.2. Morphological variation The morphology of the anteroconid allows recognition of two distinct groups among the extant samples. According to Petruso (2003), they can be named as “savimorph” (from the Savi vole) and “subterraneomorph” (from the common pine vole). Even if almost all the Microtus (Terricola) considered here, except Microtus (Terricola) lusitanicus, and Microtus (Terricola) subterraneus, are referred to as M. (T.) ex gr. savii (i.e. they have a simplex M3, Chaline and Graf, 1988), the anterior portion of the first lower molar shows a remarkable degree of variation that indicates the morphological distinction. In “subterraneomorph” populations, the anterior cap is more constricted (lower B and D, as can be seen in the DE diagram of Fig. 7) and the anteroconid is more elongated and articulated (higher A, A2 and E in Figs. 4e6). Conversely, in the “savimorph” populations the anterior cap is less constricted, and the anteroconid is shorter and looks more squat and asymmetric. This means that samples belonging to the first group plot closer to the M. (T.) subterraneus and M. (T.) lusitanicus populations, while the other ones plot on the opposite side. The populations from Tuscany (San Miniato e Smin e and Cascina e Casc e, Figs. 4e6), and the one from Aspromonte (Calabria), are part of the subterraneomorph group, while all the Sicilian populations and the one from Meldendugno (Mel e Apulia, Figs. 4e6), belong to the savimorph group. These observations, although based on a small number of samples (n ¼ 7), support the conclusions of Nappi et al. (2006) and Castiglia et al. (2008), who found a differentiation among Northern-Central and Southern Italy populations, more than the suggestion of Piras et al. (2010), who denied such differences. On the other hand, as already reported by Petruso (2003), the sample from Aspromonte is fairly distinct from the savimorph M1s. It is close to the Central Italy samples as well as to M.(T.) lusitanica and M.(T.) subterraneus and must be referred to the species M.(T.) brachycercus. The occurrence of morphological distinction of M.(T.) brachycercus respect to the other southern Microtus (Terricola) populations apparently does not support the conclusions of Nappi et al. (2005), while it has been recently confirmed by Piras et al. (2010). Considering the Early Holocene, Late Glacial and MIS3 time slices, the savimorph group is still visible, particularly in plots which report the ACC elongations (AL and A2A, Figs. 4 and 5). The savimorphs are also identifiable considering the width of T6-T7 complex (index EW, Fig. 6), with the exception of CavB, which displays, for this feature only, subterraneomorph values. On the other hand, considering the width of the anterior cap (index DE, Fig. 7), even though the two morpho-groups are still distinguishable, in the savimorph populations there is a trend towards enlargement of the anterior cap in younger samples. The Sicilian fossil populations from Oriente cave, Uzzo cave, Cala Mancina, K22, Castello Shelter, San Teodoro cave as well as the extant Sicilian samples can be clearly assigned to the savimorph group, notwithstanding the aforementioned differences in size. The Salento (Apulia) sample of Cavallo F (MIS 2) and the Campanian sample of Serratura (Late Glacial) belong to this group as well, while sample CavB (Late Glacial) is distinguished by a very wide T6eT7, a feature that could denote a drift due to isolation or, perhaps, the phenotypic effects of migrations from central to north Italian peninsula. When considering also the older time slices, interpretation of the results is less straightforward. The samples from Grotta Grande di Scario are rather homogenous and show intermediate characters between the two morpho-groups, fairly closer to the subterraneomorph group. The tooth looks quite elongated, but not as much as the extant Central Italy ones. The younger sample from the Serratura cave (SER) is close to Grotta Grande di Scario (Fig. 8), but it has the smallest size (Fig. 3). The three samples from the oldest Sicilian deposit of Isolidda 3 are clearly distinct from the younger Sicilian samples. They are close to the extant subterraneomorph populations from Central Italy and to Grotta Grande di Scario samples (pre- and post-Tyrrhenian in age). They definitely cannot be included in the savimorph group and apparently they are not even very suitable forerunners for the later Sicilian Savi vole. The similarity with the continental Scario voles coupled with the aforementioned “normal” size may be further confirmation that the Isolidda 3 vole is not much modified by endemic evolution. The vole from Contrada Pianetti (MIS3) is somewhat distinct from the other Sicilian MIS3-Holocene populations, being smaller and morphologically closer to the Grotta Grande di Scario southern peninsular samples, thus suggesting a low degree of endemization. The vole samples from Cavallo cave represent a remarkable case, as some of them show peculiar characters that cannot be found in any other sample. Firstly, there is a great difference among samples of different age. While younger samples (CavF e MIS3 and CavB e MIS2) are more similar to those coeval from the peninsula (see in AL, A2A and DE diagrams of Figs. 4, 5 and 7 respectively), the oldest ones (CavN e MIS5e and CavM e MIS5d-a) are noticeably different from any other population in having a rather shorter ACCeAC2 and the widest neck of AC2 (D, DE). Curcio et al. (2005) already noticed a peculiarity in the Apulian populations, interpreted as linked to the role of glacial refuge assumed by this area during cold climatic oscillations. However, the most distinct morphology is shown by the sample from the “warm” Eemian layer N from Cavallo cave deposit. Cavallo cave is located at the extreme south of the Salento peninsula, and therefore it is peripheral with respect to the species geographical range. The observed morphological divergence could be interpreted as the result of geographic isolation during the Eemian interglacial, while the minor divergence observed in the younger sample CavM is possibly the result of contacts with population from the peninsula that migrated southwards in response to the cooler climatic fluctuations of MIS 5. The Late Glacial vole from Cavallo cave (CavB) is distinct from the other savimorphs for having a wide T6eT7 rhombus, a feature which can denote contacts with more northern populations. On the other hand, the sample from the MIS5aed Apulia site San Sidero 3 (SS3) has typical savimorph characters and is indistinguishable from smaller-sized Sicilian samples and is also very close to the extant populations of Savi vole from Apulia (Mel). Microtus (Terricola) melitensis from the Maltese site of Ghar Dalam (GD) is different from Microtus (Terricola) savii in having some strong subterraneomorph features (like the high EW and the low DE values, Figs. 6 and 7) coupled to a shorter ACC (AL and A2A in Figs. 4 and 5) and a wide constriction between the T6 and T7 and Please cite this article in press as: Petruso, D., et al., Phylogeny and biogeography of fossil and extant Microtus (Terricola) (Mammalia, Rodentia) of Sicily and the southern Italian peninsula based on current dentalmorphological data, Quaternary International (2011), doi:10.1016/ j.quaint.2011.03.013 D. Petruso et al. / Quaternary International xxx (2011) 1e12 9 Fig. 9. Dispersals. This figure shows the distribution in time of the sites considered in the analysis and the location of the dispersal events. The MIS curve changes scale at 0.12 Ma. Boxes and arrows indicate the time and direction of dispersals. the pitymyan rhombuses (high B and BW). It has a fairly large size, comparable with the bigger M1 from the Sicilian sites, indicative of its endemic status, and is characterized by poorly differentiated enamel walls of the dental triangles, a feature that is primitive within genus Microtus (Martin, 1987; Petruso, 2003). Its M3 is of the simplex type, as in the savii group (Petruso, 2003). In summary, the Ghar Dhalam sample shares a mixture of characters, some primitive and other more evolved. The most parsimonious hypothesis is that the Maltese species derives from some primitive Sicilian populations belonging to the savii group e the Isolidda vole could be a suitable ancestor e that underwent a genetic drift resulting from long isolation in a small island such as Malta, and having the very conspicuous size recognized in the analysis (Fig. 3). Unfortunately, the age of this dispersal phase is unknown as no numerical dating exists, but a relative chronological attribution to the Würmian, provided by Storch (1974) is available for this taxon, thus placing the Maltese taxon coeval with the Sicilian Savi vole, and opening several phylogenetic and paleogeographic problems. 4.3. The role of climate To attempt to find out which are the factors acting on the variation of the tooth size and morphology, the role that climatic and environmental fluctuations could have played in generating such variation must be considered. As usual in the continental record of mammals, the sample does not allow fine correlation of the frequent and short climatic variations of the MiddleeLate Pleistocene with the modifications in dental structure, as the paleontological documentation is fragmentary in time (for example there are no populations from MIS4) and space (samples are not present for every area for every period) and the dating of the deposits is not always very precise. Nevertheless, it is possible to concentrate only on the major climatic events for checking evidences that support or contradict the existence of a significant climatic influence on the characters analyzed. If the morphological variation of the anteroconid of Microtus (Terricola) was correlated to environmental factors (change in temperature and humidity for instance), there should be a high variation of measurements and indexes in conformity with the major climatic fluctuations of the last glacial cycle. However, considering the AL, A2A and EW indexes (see Figs. 4e6) for the Sicilian samples during MIS3, Late Glacial and the ancient Holocene (a time span characterized by a dramatic climatic change), there is no significant variation. Furthermore, even if the DE index displays a strong difference between MIS3 and ancient Holocene (Fig. 7), this difference is even stronger in living Sicilian populations, and the variation is linear (a continuous increase) and so difficult to associate with climatic variations (ancient Holocene had dramatically different climatic conditions than LGM, Late Glacial and nowadays, with a strong increase in the mean temperature and humidity). This supports the hypothesis of a morphological stasis in extinct populations of Microtus (Terricola), thus being in agreement with Maul et al. (1998) and Piras et al. (2009). According to Masini et al. (2008), climate affects insular biodiversity mainly as a promoter of dispersals from the mainland, and seems to have a minor impact on the evolution of the resident endemic species, that, in strongly isolated systems, can last for a long time with minor, or no morphological modifications. In the samples analyzed and in the characters considered, the effects of climate cannot be detected. This does not mean at all that morphology is not susceptible to environmental changes, but its influence cannot be read in the dataset, perhaps because other factors more strongly affect the variables analyzed. In these samples, the variation seems to be more correlated to phylogenetic factors, that is to the history of each population and their ancestors, at least on a finer scale. It is possible to detect traces of the climate molding in similarities among populations which are well separated in time and space. For example, the similarities between some extant and fossil Apulian samples (Mel, SS3, CavM) with the Sicilian ones is rather hard to explain by a very strict phyletic link. It could be better explained as a homoplasy due to parallel evolution, likely as a response to quite similar environmental conditions between Sicily and the Salento peninsula. Furthermore, it is possible to detect a northesouth trend, with northern samples characterized by elongated teeth with Please cite this article in press as: Petruso, D., et al., Phylogeny and biogeography of fossil and extant Microtus (Terricola) (Mammalia, Rodentia) of Sicily and the southern Italian peninsula based on current dentalmorphological data, Quaternary International (2011), doi:10.1016/ j.quaint.2011.03.013 10 D. Petruso et al. / Quaternary International xxx (2011) 1e12 constricted anteroconid, and southern samples (with the exception of the Calabrian population, that has been also assigned to a different species), whose antericonid is shorter, compressed and made up of highly confluent triangles. These characteristics are the same pointed out by Nappi et al. (2006) and seem to be related to the geography of the area. Thus, if at a local scale it is not possible to correlate the variation of these features to environmental factors, traces of this relation are recognizable at a regional scale. 4.4. Dispersals to Sicily According to the data gathered and discussed above, M. (Terricola) voles underwent at least two different dispersal events from Italian peninsula to Sicily (Fig. 9) during Middle and Late Pleistocene. This hypothesis is supported by the marked difference occurring among the older Sicilian samples from Isolidda 3 and the younger Würmian (MIS3) populations, which resemble recent and fossil populations of Savi vole from the southern Italian peninsula. Although the occurrence of a gap in the documentation between Isolidda 3 and the younger samples requires caution, the hypothesis of a direct phyletic link between Isolidda 3 and the younger voles would imply the development of an amount of homoplasies, due to parallel evolution on the island and on the mainland, which is rather improbable. Furthermore, some smaller-sized Würmian populations such as those from Contrada Pianetti or San Teodoro square D are suitable to be considered as new arrivals and ancestors of the endemic Savi vole populations of the island. Eventually, the dispersals of several taxa to the island during the Würmian render the hypothesis of the evolution in isolated conditions since the lastbut-one glacial (MIS6) even more unlikely. The following interpretation is presented. The ancestor of the vole from Isolidda 3 first reached the island during an ancient low stand phase of the sea level, probably during MIS6 (no numerical dating at present). The absence in Isolidda 3 samples of southern savimorph feature, their close similarity with the peninsular populations from the site of Grotta Grande di Scario (GGSC attributed to MIS6, Table 1) as well as their small size, support the hypothesis of such an old dispersal for this vole, and that the age of Isolidda should be rather close to that of the dispersal event. On the other hand, the second dispersal from the peninsula involved Microtus (Terricola) savii populations already characterized by well-developed “southern” savimorph features. The dispersal might have occurred during the low stand phase related to the glacial peak of MIS4. Possibly during the following MIS3, characterized by alternating cool and relatively warmer fluctuations, might have been a period of partial isolation for the Sicilian populations, allowing the voles to develop a large size. It is difficult to assess if the climate deterioration of MIS2 and the following Late Glacial, determining the probable connection with the mainland that introduced the Epi-Paleolithic man onto the island, did affect the evolution of the Sicilian voles by the introduction of continental populations or groups of individuals from the mainland. Actually, if it was so, such contacts did not left traces on the vole’s size, which are still large in the Late Glacial sites and particularly so in the early Holocene. 5. Conclusions The morphometric analysis of the first lower molars of Microtus (Terricola) from Sicily, Malta, peninsular Italy, France and Spain proved to be helpful in the attempt to reconstruct the phylogenetic relationships between Sicilian Savi vole and the peninsular ones as well as to detect the dispersal events from the mainland to the island. The following points can be drawn. 1. The analysis allows identification, within the Savi vole latu sensu, of two main morphotypes or morpho-groups: “subterraneomorh” e for the affinity with the common pine vole morphology e and “savimorph” e from the southern Savi vole populations. 2. Concerning the extant populations, there is a clear separation between Sicilian Microtus (Terricola) savii and the central peninsular populations (which show subterraneomoroph features). These results are in agreement with Nappi et al. (2006) and Castiglia et al. (2008), but dissent from Piras et al. (2010). The Apulia population is close to Sicilian ones, in agreement with the occurrence of a “southern” group of Savi voles as affirmed by Nappi et al. (2006). The small size of the extant Savi vole of the island is likely due to the anthropogenic impact that reduces or cancels the effects of geographic isolation. As a by-product, although based on few samples, the morphological distinction of the Calabrian vole referred to M. (T.) brachycercus, already noticed by Petruso (2003), is supported by the analysis, in agreement with Piras et al. (2010). 3. Since the MIS3 till the early Holocene, the morphology of the fossil Sicilian Savi vole is close to the Southern Italy populations. Several Sicilian populations, however, show markedly larger size than continental ones, while a few others are of the same size as the continental voles. Larger size is interpreted as an effect of geographic isolation. This would suggest that voles of the savii group during the Würmian glaciation dispersed from the mainland, and then became isolated, developing endemic features during the late part of the glaciation (Fig. 9). The possibility that some exchanges with the mainland repeatedly occurred during the cooler phases cannot be ruled out. This reconstruction is apparently in contrast with Castiglia et al. (2008), who suggested a species status for the Sicilian vole based on the divergence indicated by molecular data. Isolation and drift may have played an important role in determine such a high divergence. 4. Pre-Tyrrhenian Sicilian populations from the site Isolidda 3, referable to the last-but-one glaciation (MIS6), are distinctly different from post-MIS4 populations, and are closer to coeval population from the Tyrrhenian side of the peninsula, thus suggesting that they originate from an older dispersal event that took place during the sea - level low stand of MIS6 (Fig. 9). 5. The M. (T.) melitensis sample from Ghar Dalam (Malta) is a mosaic of subterraneomorph and savimorph features, and occupies an isolated position. The result of the comparisons, as well as the primitive enamel pattern of the M1, apparently rules out its derivation from a Late Pleistocene Savi vole dispersed from Sicily, and support the hypothesis of an older dispersal. 6. Climatic changes do not seem to have a direct effect on size and morphology of the insular voles, while dispersals from the mainland, likely stimulated by climatic driven environmental changes and facilitated by the eustatic effects of climatic changes, are responsible for species substitution in the island. In summary, the work provides a large amount of original information regarding the Sicilian voles, and a new piece of information to the reconstruction of the evolutionary history of Microtus (Terricola) in the Southern Italian regions. Further studies, regarding both extant and fossil populations could help in unraveling the problems that are still open. Acknowledgements The present work results from the following individual contributions: Daria Petruso, most of the Sicilian samples and Ghar Please cite this article in press as: Petruso, D., et al., Phylogeny and biogeography of fossil and extant Microtus (Terricola) (Mammalia, Rodentia) of Sicily and the southern Italian peninsula based on current dentalmorphological data, Quaternary International (2011), doi:10.1016/ j.quaint.2011.03.013 D. Petruso et al. / Quaternary International xxx (2011) 1e12 Dalam; Elisa Locatelli, Isolidda 3, Cala Mancina cave, Oriente cave; Giovanni Surdi, Grotta Grande di Scario cave, Serratura cave, extant samples from Cascina and S. Miniato; Chiara Dalla Valle, Cavallo cave; Federico Masini and Benedetto Sala coordinated the work. The paper has been elaborated conjunctly by the research teams of Palermo and Ferrara Universities. The work is supported by Italian MIUR (PRIN 2008, 2008RTCZJH), by the Ferrara University (FAR funding 2007) and by the Palermo University (Athenaeum Grants e FAR-year 2008). Elisa Locatelli was supported by the PhD program “Scienze and Tecnologie per l’Archeologia e i Beni culturali” at Ferrara University; Giovanni Surdi is supported by PhD fellowship program “I micromammiferi tardo quaternari della Sicilia e del Sud della penisola italiana”, Palermo University. The authors thank Prof. Laura Bonfiglio of the University of Messina, for the material of Grotta San Teodoro; Prof. Annamaria Ronchitelli and Prof. Paolo Boscato of the Department of Environmental Sciences “G. Sarfatti”, Siena University, for the material from Grotta Grande di Scario; Prof. Fabio Martini of the Department of Antiquity Sciences, Florence University, for the material from Isolidda 3, Cala Mancina cave, Oriente cave and Cavallo cave; Prof. Maurizio Sarà for the extant Sicilian samples of Borgo Lupo, Mazzarino, Gela, and for the sample from Aspromonte. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.quaint.2011.03.013. References Abbazzi, L., Masini, F., 1998. L’association faunistique du sondage ‘F’a la Grotta Grande de Scario. In: Ronchitelli, A. (Ed.), Structures d’habitat dans la Grotta Grande de Scario (S.Giovanni a Piro, Salerno-Italie). Proceedings of The XIII International Congress of Prehistoric and Protohistoric Sciences, 2nd Section, Forlì, Italia, pp. 153e163. Bertolini, M., Fedozzi, S., Martini, F., Sala, B., 1996. Late Glacial and Holocene climatic oscillations inferred from the variations in the micromammal associations at Grotta della Serratura (Marina di Camerota, Salerno, South Italy). Il Quaternario 9 (2), 561e566. Bonfiglio, L., Insacco, G., Marra, A.C., Masini, F., 1997. Large mammals, amphibians, reptiles from a new fissure filling deposit of the Hyblean Plateau (South Eastern Sicily). Bollettino della Società Paleontologica Italiana 36 (1e2), 97e122. Bonfiglio, L., Marra, A.C., Masini, F., Pavia, M., Petruso, D., 2002. Pleistocene faunas of Sicily: a review. In: Waldren, W.H., Ensenyat, J.A. (Eds.), World Islands in PrehistorydInternational Insular Investigations, V Deia International Conference. British Archaeological Research International Series 1095, vol. 38, pp. 428e436. Bonfiglio, L., Esu, D., Mangano, G., Masini, F., Petruso, D., Soligo, M., Tuccimei, P., 2008. Late Pleistocene vertebrate bearing deposits at San Teodoro Cave (NorthEastern Sicily): preliminary data on faunal diversification and chronology. Quaternary International 190, 26e37. Boscato, P., Ronchitelli, A., 2004. Paléosurfaces du Paléolithique moyen: l’example de Scario (Salerno e Italie du sud). In: Actes du XIVème Congrès UISPP, 2001, Liège (Belgique). British Archaeological Research International Series 1239, pp. 283e291. Brunet-Lecomte, P., Chaline, J., 1992. Morphological convergences versus biochemical divergences in the Holoarctic ground voles: Terricola and Pitymys (Arvicolidae, Rodentia). Neue Jahrbuch Geologie Palaeontologie Mittelhandlungen 12, 721e734. Burgio, E., Kotsakis, T., 1987. Presenza di Pitymys (Mammalia, Arvicolidae) nel Pleistocene della Sicilia. Il Naturalista Siciliano 4 (10), 27e34. Castiglia, R., Annesi, F., Aloise, G., Amori, G., 2008. Systematics of the Microtus savii complex (Rodentia, Cricetidae) via mitochondrial DNA analyses: paraphyly and pattern of sex chromosome evolution. Molecular Phylogenetics and Evolution 46, 1157e1164. Chaline, J., 1972. Les rongeurs du Plèistocene moyen et superieur de France. Cahiers Paléontologique, Paris. Chaline, J., Graf, J., 1988. Phylogeny of the arvicolidae (Rodentia): biochemical and paleontological evidence. Journal of Mammology 69 (1), 22e33. Contoli, L., 2003. On subspecific taxonomy of Microtus savii (Rodentia, Arvicolidae). Hystrix, Italian Journal of Mammalogy 14, 107e111. Contoli, L., Nappi, A., 2009. Microtus brachycercus. In: Amori, G., Contoli, L., Nappi, A. (Eds.), Fauna d’italia, mammalia ii: erinaceomorpha, soricomorpha, lagomorpha, rodentia. Calderini, Milano, pp. 535e538. 11 Curcio, M.T., Contoli, L., Di Canzio, E., Kotsakis, T., 2005. Preliminary analysis of the first lower molar variability in Late Pleistocene and living populations of Terricola savii (Arvicolidae, Rodentia). Geologia Alpina 2, 91e98. De Giuli, C., 1983. Le faune pleistoceniche del Salento. 1.- La fauna di San Sidero 3. Quaderni Museo Pal. di Maglie 1, 45e84. De Luca, C., 1998. Nuovi ritrovamenti di Microtus (Terricola) in un deposito pleistocenico in località Piana di Sopra, Penisola di San Vito lo Capo. Unpublished Thesis, Scienze Geologiche, A.A. 1997/1998. Università degli Studi di Palermo. Di Maggio, C., Incandela, A., Masini, F., Petruso, D., Renda, P., Simonelli, C., Boschian, G., 1999. Oscillazioni eustatiche, biocronologia dei depositi continentali quaternari e neotettonica nella Sicilia nord-occidentale (Penisola di San Vito Lo Capo e Trapani): un approccio interdisciplinare. Il Quaternario 12 (1), 25e49. Foster, J.B., 1964. Evolution of mammals on islands. Nature 202, 234e235. Galleni, L., Stanyon, R., Tellini, A., Giordano, G., Santini, L., 1992. Karyology of the Savi pine vole, Microtus savii (de Sélys Longschamps, 1838) (Rodentia, Arvicolidae): G- C-, DA/DAPI- and AluI-bands. Cytogenetics and Cell Genetics 59, 290e292. Galleni, L., Stanyon, R., Tellini, A., Giordano, G., Santini, L., 1994. Taxonomy of Microtus savii (Rodentia, Arvicolidae) in Italy: cytogenetic and hybridization data. Journal of Mammalogy 75, 1040e1044. Galleni, L., 1995. Speciation in the Savi pine vole, Microtus savii (de Sélys-Longschamps, 1838) (Rodentia, Arvicolidae): a theoretical biology approach. Bollettino di Zoologia 62, 45e51. Galleni, L., Stanyon, R., Contadini, L., Tellini, A., 1998. Biogeographical and kariological data of the Microtus savii group (Rodentia, Arvicolidae) in Italy. Bonner Zoologische Beiträge 47, 277e282. Hammer, Ø., Harper, D.A.T., Ryan, P.D., 2001. PAST: paleontological statistics software package for education and data analysis. Palaeontologia Electronica 4 (1). Jaarola, M., Martinkova, N., Gunduz, I., Brunhoff, C., Zima, J., Nadachowski, A., Amori, G., Bulatova, N.S., Chondropoulos, B., Fraguedakis-Tsolis, S., GonzalezEsteban, J., Jose Lopez-Fuster, M., Kandaurov, A.S., Kefelioglu, H., Da Luz Mathias, M., Villate, I., Searle, J.B., 2004. Molecular phylogeny of the speciose vole genus Microtus (Arvicolinae, Rodentia) inferred from mitochondrial DNA sequences. Molecular Phylogenetics and Evolution 33, 647e663. Locatelli, E., 2007. I micromammiferi fossili della Sicilia. I giacimenti di Isolidda 3, Grotta d’Oriente e Cala Mancina. Unpublished thesis, LS Scienze preistoriche A.A. 2006/2007, Università di Ferrara. Lomolino, M.V., 2005. Body size evolution in insular vertebrates: generality of the island rule. Journal of Biogeography 32, 1683e1699. Martin, R.A., 1987. Notes on the classification and evolution of some North American fossil Microtus (Mammalia, Rodentia). Journal of Vertebrate Paleontology 7, 70e73. Martini, F., 1993. Grotta della Serratura a Marina di Camerota. In: Culture e ambienti dei complessi olocenici. Garlatti e Rizzai, Firenze. Martini, F., Lo Vetro, D., Baglioni, L., Colonese, A.C., De Curtis, O., Di Giuseppe, Z., Locatelli, E., Sala, B., Mazza, P., Tusa, S., Primi risultati sulle nuove ricerche stratigrafiche a Grotta d’Oriente (Favignana, TP). Scavi 2005. In: Dai Ciclopi agli Ecisti, Società e territorio nella Sicilia Preistorica e Protostorica, Atti della XLI Riunione Scientifica IIPP, Firenze, in press-a. Martini, F., Lo Vetro, D., Colonese, A.C., Di Giuseppe, Z., Locatelli, E., Sala, B., Tusa, S., Dati preliminari sul Mesolitico di Grotta di Cala Mancina (San Vito Lo Capo, TP): paletnologia e ambiente. In: Dai Ciclopi agli Ecisti. Società e territorio nella Sicilia Preistorica e Protostorica, Atti della XLI Riunione Scientifica IIPP, Firenze, in press-b. Martini, F., Lo Vetro, D., Colonese, A.C., Di Giuseppe, Z., Locatelli, E., Sala, B., Tusa, S., Nuove ricerche a Grotta Racchio - Gruppo dell’Isolidda (San Vito Lo Capo, TP): primi risultati. In: Dai Ciclopi agli Ecisti. Società e territorio nella Sicilia Preistorica e Protostorica, Atti della XLI Riunione Scientifica A IIPP, Firenze, in press-c. Masini, F., Giannini, T., Abbazzi, L., Fanfani, F., Delfino, M., Maul, L.C., Torre, D., 2005. A late Biharian small vertebrate fauna from the lacustrine succession of San Lorenzo (Santarcangelo Basin, Basilicata, Italy). Quaternary Intenational 131, 79e93. Masini, F., Petruso, D., Bonfiglio, L., Mangano, G., 2008. Origination and extinction patterns of mammals in three Central Western Mediterranean Islands in the Late Miocene to Quaternary. Quaternary International 182, 63e79. Maul, L., Masini, F., Abbazzi, L., Turner, A., 1998. The use of different morphometric data for absolute age calibration of some South e and Middle European arvicolid populations. Palaeontographia Italica 85, 111e151. Nappi, A., Brunet-Lecomte, P., Montuire, S., 2005. Dental morphology of Microtus (Terricola) voles from Calabria (Southern Italy) and relationships with M. (T.) savii (Rodentia, Arvicolidae). Hystrix. Italian Journal of Mammalogy 16 (1), 75e85. Nappi, A., Brunet-Lecomte, P., Montuire, S., 2006. Intraspecific morphological tooth variability and geographical distribution: application to the Savi’s vole, Microtus (Terricola) savii (Rodentia, Arvicolinae). Journal of Natural History 40 (5, 6), 345e358. Palma di Cesnola, A., Borzatti von Löwestern, E., 1964. Gli scavi dell’Istituto Italiano di Preistoria e Protostoria nel Salento durante l’ultimo trienno. In: Atti della VIII e IX Riunione Scientifica, Firenze, pp. 24e43. Petruso, D., 1996. Le faune a mammiferi del Pleistocene medio e superiore della Penisola di San Vito. Correlazione con i terrazzi marini. Unpublished thesis, Scienze Geologiche, A.A. 1994/1995, Università degli Studi di Palermo, pp. 1e96. Please cite this article in press as: Petruso, D., et al., Phylogeny and biogeography of fossil and extant Microtus (Terricola) (Mammalia, Rodentia) of Sicily and the southern Italian peninsula based on current dentalmorphological data, Quaternary International (2011), doi:10.1016/ j.quaint.2011.03.013 12 D. Petruso et al. / Quaternary International xxx (2011) 1e12 Petruso, D., 2003. Il contributo dei micromammiferi alla Stratigrafia e Paleogeografia del Quaternario continentale siciliano. PhD thesis, Geologia del Sedimentario, Università di Napoli e Palermo. Petruso, D., Cusumano, M., Domina, D., Pavia, M., Masini, F., 2008. Una valutazione tassonomica quantitativa preliminare della documentazione di vertebrati del Quaternario della Sicilia. Atti del Museo Civico di Storia Naturale di Trieste 53 (Suppl.), 189e232. Petruso, D., Sarà, M., Surdi, G., Masini, F., Le faune a mammiferi della Sicilia tra il Tardoglaciale e L’Olocene. Biogeographia, in press. Piras, P., Marcolini, F., Raia, P., Curcio, M.T., Kotsakis, T., 2009. Testing evolutionary stasis in first lower molar shape of fossil Italian populations of Terricola savii (Arvicolidae, Rodentia) by means of geometric morphometrics. Journal of Evolutionary Biology 22, 179e191. Piras, P., Marcolini, F., Raia, P., Curcio, M.T., Kotsakis, T., 2010. Ecophenotypic variation and phylogenetic inheritance in first lower molar shape of extant Italian populations of Microtus (Terricola) savii (Rodentia). Biological Journal of The Linnean Society 99, 632e647. Ronchitelli, A., Abbazzi, L., Accorsi, C.A., Bandini Mazzanti, M., Bernardi, M., Masini, F., Mercuri, A., Mezzabotta, C., Rook, L., 1998. Palethnological, palynological and paleontological data on the Grotta Grande of Scario e Salerno (Campania Southern Italy, 40 020 2100 N/15 280 3100 E). In: De Leo, G. (Ed.), Proceedings of First International Congress on Science and technology for The Safeguard of Cultural Heritage in The Mediterranean Basin, Palermo, pp. 1529e1535. Ronchitelli, A., Boscato, P., Masini, F., Petruso, D., Surdi, G., Accorsi, C.A., Torri, P., 2008. The Grotta Grande of Scario (Salerno, Italy): a spot on the archeology and the environment during the last interglacial (OIS5) of the Mediterranean region. In: Atti del Convegno Internazionale INQUA-SEQS (Subcommission on European Quaternary Stratigraphy) “The Quaternary of Europe from Atlantic seaboard to continental Interior”, Rennes, pp. 46e47. Sarà, M., 1998. I mammiferi delle isole del Mediterraneo. L’Epos, Palermo. Sarti, L., Boscato, P., Lo Monaco, M., 2002. Il Musteriano finale di Grotta del Cavallo nel Salento: studio preliminare. In: Origini e preistoria e protostoria delle civiltà antiche, vol. 22. Bonsignori editore, Roma, pp. 45e109. Shackleton, N.J., 1995. New data on the evolution of Pliocene climatic variability. In: Vrba, E. (Ed.), Paleoclimate and evolution. Yale University Press, pp. 242e248. Sondaar, P.Y., 1977. Insularity and its effect on mammal evolution. In: Hecht, M.K., Goody, P.C., Hecht, B.M. (Eds.), Major patterns in vertebrate evolution. Plenum, New York, pp. 671e707. Storch, G., 1974. Quartäre Fledermaus e Faunen von der Insel Malta. Senckenbergiana 55 (1/5), 407e434. Tagliacozzo, A., 1993. Archeozoologia della Grotta dell’Uzzo, Sicilia. Da un’economia di caccia ad un’economia di pesca e di allevamento. Supplement. Bollettino della Società Paleontologica Italiana, N.S. 84, 1e278. Tougard, C., Brunet-Lecomte, P., Fabre, M., Montuire, S., 2008. Evolutionary history of two allopatric Terricola species (Arvicolinae, Rodentia) from molecular, morphological, and palaeontological data. Biological Journal of The Linnean Society 93 (2), 309e323. Van der Meulen, A.J., 1973. Middle Pleistocene smaller mammals from the Monte Peglia, (Orvieto, Italy) with special reference to the phylogeny of Microtus (Arvicolidae, Rodentia). Quaternaria 17, 1e145. Van Valen, L., 1973. A new evolutionary law. Evolutionary Theory 1, 1e33. Wilson, D.E., Reeder, D.M., 2005. In: Mammal Species of The World. A Taxonomic and Geographic Reference, third ed. Johns Hopkins University Press, Baltimore, p. 2. Please cite this article in press as: Petruso, D., et al., Phylogeny and biogeography of fossil and extant Microtus (Terricola) (Mammalia, Rodentia) of Sicily and the southern Italian peninsula based on current dentalmorphological data, Quaternary International (2011), doi:10.1016/ j.quaint.2011.03.013