Biologia 69/6: 771—779, 2014Section ZoologyDOI: 10.2478/s11756-014-0372-x

The Cochlostoma (Holcopoma) westerlundi group in Italy(Caenogastropoda: Cochlostomatidae)

Mariastella Colomba1, Fabio Liberto2, Armando Gregorini1, Walter Renda3,Agatino Reitano4 & Ignazio Sparacio5

1Universita di Urbino “Carlo Bo”, Dip. Scienze Biomolecolari, via Maggetti 22 (loc. Sasso), 61029 Urbino (PU), Italy;e-mail: mariastella.colomba@uniurb.it2Strada Provinciale Cefalu-Gibilmanna n◦ 93, 90015 Cefalu (PA), Italy3via Bologna 18/a, 87032 Amantea (CS), Italy4 Via Gravina 77, 95030 Tremestieri Etneo (CT), Italy5Via E. Notarbartolo 54 int. 13, 90145 Palermo, Italy

Abstract: Currently, the Cochlostoma (Holcopoma) westerlundi (Paulucci, 1879) group includes three subspecies inhabitingsouthern Italy up to southeastern Sicily. C. w. westerlundi (Paulucci, 1879) is limited to southern Calabria, C. w. yapigium(Westerlund, 1885) is widespread across the Salento (the southeastern extremity of the Apulia region) and C. w. dionysii(Paulucci, 1879) is endemic to the environs of Siracusa (SE Sicily). There is also a fourth taxon, C. paganum (Westerlund,1885) described for Otranto (LE, Apulia), considered a synonym of C. w. yapigium. Up to now, the molecular geneticsof C. westerlundi s.l. have been unknown and the morphological data of several populations are still lacking. Hence,the systematic classification of the group is tentative. Aiming at filling this gap, mtDNA (16S rDNA and COI ) partialsequences were investigated and, in addition, the reproductive apparatus of C. w. westerlundi was described for the firsttime. Molecular sequences and anatomical data were used to test the taxonomic and phylogenetic status of the examinedpopulations. Maximum Likelihood and Bayesian analysis revealed three clusters, strongly supported, corresponding to thethree taxa. For the first time, synonymy between paganum and yapigium was confirmed by molecular evidence. Geneticdistances between groups (DxyJC) ranged from 2.6% to 5% (16S rDNA) and from 6.3% to 8.3% (COI ). Molecular andmorphological data led us to suggest elevating the three subspecies to the species rank.

Key words: Cochlostoma westerlundi group; 16S rDNA; COI ; molecular phylogeny; taxonomic revision

Introduction

In zoological systematics, morphologically similar in-dividuals are often grouped under the same nomi-nal taxon. This method of individual classification(not always practicable in terrestrial gastropods) gavebirth, over the past nearly sixty years, to severalslightly different classifications of the Cochlostomawesterlundi s.l. morpho-types. Currently, the C. wester-lundi group comprises four nominal taxa, three ofwhich are considered subspecies: C. w. westerlundi(Paulucci, 1879), inhabiting very restricted areas ofsouthern Calabria; C. w. yapigium (Westerlund, 1885),which is widespread in southern Apulia; and C. w.dionysii (Paulucci, 1879), endemic to southeasternSicily (Fig. 1). The fourth taxon, C. paganum (West-erlund, 1885) described for Otranto (LE, Apulia), isconsidered a synonym of C. w. yapigium (Forcart 1965;Pintér & Szigethy 1976; Bodon et al. 1995; Ferreri et al.2005; Bank 2012). As far as concerns a brief overview ofthe Cochlostoma Jan, 1830 system (at least that of theItalian species), Sacchi (1954) accounted yapigium asdistinct from westerlundi while Alzona (1971) reported

westerlundi, yapigium and dionysii as different species,and paganum as a subspecies.Comparative anatomical data on the shell and gen-

italia are extremely limited. In males of C. w. yapigium,the penis is simple, gradually tapering at the end; infemales, the sperm receptacle is absent (personal data;see also Ferreri et al. 2005). In C. w. dionysii, the penisis more or less cylindrical and narrows abruptly at theend (Reitano et al. 2009), whereas nothing is knownabout the female reproductive system. Finally, the gen-ital anatomy of C. w. westerlundi has never been stud-ied. Cochlostoma westerlundi s.l. prefers limestone envi-ronments and possesses very low active dispersal ability(due to limited vagility), which is likely to have dras-tically reduced the gene flow, thus suggesting strongisolation and (presumably) divergence among popula-tions.Recently, the combination of genetic data and

morphological characters (i.e., shell sculpture, genitalanatomy) has proven useful in unraveling taxonomicissues in other groups of terrestrial mollusks (Uit DeWeerd & Gittenberger 2004; Liew et al. 2009; Colombaet al. 2011). We herein applied a similar approach to

c©2014 Institute of Zoology, Slovak Academy of Sciences

772 M.S. Colomba et al.

Fig. 1. Geographic distribution of taxa ascribed to the Cochlostoma westerlundi group. • – yapigium; � – westerlundi; – dionysii;◦ – C. paganum; � – new records for dionysii: Priolo, Palazzolo Acreide, Eloro and contrada Palombara (Siracusa province, SR, Sicily).

solve the systematics of C. westerlundi s.l. and providethe first comprehensive molecular phylogenetic studyof the taxon group by means of partial sequences oftwo mitochondrial genes [16S rRNA and cytochromeoxidase c subunit I (COI )] analyzed by Bayesian In-ference and Maximum Likelihood algorithms. Finally,we suggest a paleogeographic scenario which might ac-count for the current spatial distribution and evolution-ary history of the organisms under study.

Material and methods

SpecimensTwo hundred and fifty-eight (258) specimens were analyzedfor both morphological and biometrical key features of shellsand animals. Individuals were drowned and fixed in 70–90% ethanol. The reproductive apparatus was extracted bymeans of a scalpel, scissors and forceps. Photographs weretaken with a digital camera. The maximum height and max-imum diameter of the shell along with some parts of gen-italia were measured (in mm) using digital calipers. Illus-trations of genitalia were sketched using a camera lucida.

Voucher specimens were deposited at the I. Sparacio collec-tion (CS), F. Liberto collection (CL) and W. Renda collec-tion (CR). Twenty-one additional specimens including sixC. w. yapigium, three C. paganum, six C. w. dionysii andsix C. w. westerlundi were employed for the molecular char-acterization based on multigenic sequence analysis.

Examined material

C. w. yapigiumLecce, Nardo, masserie, 40◦08′14′′ N, 18◦01′16′′ E, 60 m,19.III.2009, leg. RendaW., 7 specimens, (CL); Lecce, Nardo,Porto Selvaggio, 09.XI.2009, leg. Ferreri D. & Renda W., 7specimens, (CL); Lecce, Lizzanello, 08.XI.2009, leg. FerreriD., 5 specimens, (CL); Lecce, Bosco La Lizza, 08.XI.2009,leg. Ferreri D., 7 specimens, (CL); Otranto, Valle dell’Idro,08.XI.2009, leg. Ferreri D., 6 specimens, (CL); Gallipoli,Pineta, IV.1980, leg. Pirozzi N., 5 specimens (CS); Lecce,Acaya, 4.V.1993, 11 specimens (CS); Lecce, Bosco La Lizza,08.XI.2009, leg. Ferreri D., 2 specimens employed for molec-ular analyses, labeled as CYBL; Lecce, Nardo, Portoselvag-gio, 09.XI.2009, leg. Ferreri D. & Renda W., 2 specimensemployed for molecular analyses, labeled as CYNA; Lecce,

Cochlostoma (Holcopoma) westerlundi in Italy 773

Table 1. Origin of dissected specimens and data on mtDNA partial sequences generated in the present study.

Taxon Collection site Region/Country mtDNA sequence Length (bp) Isolate GenBank ID

dionysii Siracusa (SR) Sicily/Italy 16S 436 CDSR JN030448westerlundi Mt. Consolino (RC) Calabria/Italy 16S 447 CW JN030449yapigium Lecce (LE) Apulia/Italy 16S 438 CYLE JN030450yapigium Nardo (LE) Apulia/Italy 16S 438 CYNA KC609730yapigium Bosco La Lizza (LE) Apulia/Italy 16S 438 CYBL KC609732paganum = yapigium Otranto (LE) Apulia/Italy 16S 438 CYOTR JN162680dionysii Siracusa (SR) Sicily/Italy COI 645 CDSR JN030453westerlundi Mt. Consolino (RC) Calabria/Italy COI 654 CW JN030452yapigium Lecce (LE) Apulia/Italy COI 647 CYLE JN030451yapigium Nardo (LE) Apulia/Italy COI 647 CYNA KC609731paganum = yapigium Otranto (LE) Apulia/Italy COI 647 CYOTR JN162681yapigium Bosco La Lizza (LE) Apulia/Italy COI 647 CYBL KC609733roseoli Moraca valley Montenegro COI 638 HNHM 95594 KF812521

loc. CAT, 08.XI.2009, leg. Ferreri D., 2 specimens employedfor molecular analyses, labeled as CYLE;

C. paganumValle dell’Idro, Otranto, 08.XI.2009, leg. Ferreri D., 3 speci-mens employed for molecular analyses, labeled as CYOTR.

C. w. dionysiiNoto, Noto Antica, 20.VIII.1993, legit Sparacio I., 7 spec-imens (CS); idem, 24.IV.2005, legit Brancato A., 3 speci-mens (CL); idem, 26.VI.2005, leg. Brancato A., 13 speci-mens (CL); Melilli, R.N.I. “Grotta Palombara” 37◦06′22′′

N, 15◦11′45′′ E, 130 m., 16.V.2009, leg. Liberto F., 8 speci-mens, (CL); Priolo Gargallo, Cava Sorciaro, 37◦09′49′′ N,15◦09′24′′ E, 95 m, 09.V.2010, leg. Liberto F., 7 speci-mens (CL); idem, leg. Sparacio I., 12 specimens, (CS);idem, 28.XI.2010, leg. Sparacio I., 6 specimens (CS);Melilli, Sorgente Belluzza, 37◦13′16′′ N, 15◦06′21′′ E,104 m, 21.XI.2010, leg. Liberto F., 10 specimens (CL);Melilli, Santa Caterina, 37◦12′06′′ N, 15◦06′12′′ E, 200 m,21.XI.2010, leg. Liberto F., 7 specimens, (CL); Palaz-zolo Acreide, Pianetti, 37◦02′14′′ N, 14◦58′37′′ E, 500 m,05.III.2011, leg. Sparacio I., 28 specimens (CS); idem, leg.Sparacio I., 8 specimens (CL); Siracusa, Santa Lucia, 50 m,VIII.2011, leg. Brancato A., 5 specimens (CL); Siracusa,Belvedere, Passo Piano, 37◦06′00′′ N, 15◦12′04′′ E; 130 m,VIII.2011, leg. Brancato A., 3 specimens (CL); Siracusa, neipressi dell’Orecchio di Dionisio, 37◦04′36′′ N, 15◦16′33′′ E,30 m, VIII.2011, leg. Brancato A., 6 specimens (CL); Sira-cusa, pressi del Teatro antico, leg. Sparacio I., 10.IX.1982, 4specimens (CS); idem, leg. Sparacio I., 28.XI.2010, 8 spec-imens (CS); Sortino, Necropoli di Pantalica, 37◦08′28′′ N,15◦01′51′′ E, 250 m, XI.2009, leg. Reitano A., 10 specimens(CL); Noto, Eloro, 18.VIII.1993, leg. Sparacio I., 2 speci-mens (CS); Siracusa, Spinagallo, 37◦00′11′′ N, 15◦10′50′′ E,119 m, 01.IV.2012, leg. Reitano A., 15 specimens (CL); Sir-acusa, Belvedere Nord, 17.III.2009, leg. Renda W., 6 speci-mens employed for molecular analyses, labeled as CDSR.

C. w. westerlundiPazzano, Monte Stella, 23.X.2009, leg. Renda W., 5 spec-imens (CR); idem, 09.V.2010, leg. Renda W., 2 specimens(CR); Stilo, Monte Consolino, 38◦28′37′′ N, 16◦27′53′′ E,550 m, 03.VII.2008, leg. Renda W., 12 specimens (CR);idem, 14.III.2009, leg. Renda W., 3 specimens (CR); idem,14.III.2009, leg. Renda W., 5 specimens, (CL); idem,14.III.2009, leg. Renda W., 6 specimens employed for molec-ular analyses, labeled as CW.

DNA extraction, amplification and sequencingDissected specimens originated from the following popula-tions/collection sites (Table 1); C. w. yapigium: Lecce (LE),Nardo, Portoselvaggio (LE), and Bosco La Lizza (LE), Apu-lia; C. paganum: Otranto (LE), Apulia; C. w. dionysii Sira-cusa (SR), Sicily; C. w. westerlundi: Monte Consolino, Stilo(RC), Calabria. Samples were stored at –20◦C in test tubes.The entire specimen was used for total DNA extraction (byWizard Genomic DNA Purification Kit, Promega). For eachpopulation/collection site, para-voucher specimens – sensuGroenenberg et al. (2011), i.e., different specimens than theones used for DNA analysis, but from the same sample orpopulation/collection site – were stored at the University ofUrbino.

COI amplicons (645–654 bp) were obtained by theprimers LCO 1490 (5’-GGTCAACAAATCATAAAGATATTGG-3’) and HCO 2198 (5’-TAAACTTCAGGGTGACCAAAAAATCA-3’) according to Folmer et al. (1994); 16SrDNA fragments (436–447 bp) were amplified by the univer-sal forward primer (5’-CCCGCCTGTTTACCAAAAACAT-3’) reported in Mitani et al. (2009) and the reverseprimer (5’-TATCTCAATCCAACATCGAGG-3’) (designedherein). PCR reactions were conducted as follows: for 16S,95◦C for 5 min; 95◦C for 1 min, 55◦C for 1 min, 72◦C for1 min (35 cycles); 72◦C for 5 min; for COI, 95◦C for 5 min;95◦C for 1 min, 42◦C for 1 min, 72◦C for 1 min (37 cycles);72◦C for 10 min. To remove primers and unincorporatednucleotides, amplification products were purified using theWizard SV gel and PCR Clean-up kit (Promega). Sequenc-ing of the purified PCR products was carried out using au-tomated DNA sequencers at Eurofins MWG Operon (Ger-many). Chromatograms of each amplified fragment were vi-sually inspected for reading mistakes by the sequencer. Foreach collection site, the generated sequences were depositedin GenBank (Table 1).

Phylogenetic analysesAll sequences were visualized with BioEdit Sequence Align-ment Editor 7 (Hall 1999), aligned with the ClustalW op-tion included in this software and refined by eye. For bothdatasets, the strict molecular clock hypothesis could not berejected (P < 0.99, tested by MEGA 5.0.3). Phylogeneticanalyses were conducted in BEAST 1.6.1 (Drummond &Rambaut 2007) using the *BEAST implementation (Heled& Drummond 2010). A series of initial runs was performedto optimize priors and runtime parameter choices to obtaineffective sampling sizes (ESS) above 500 for all estimatedparameters. Parameter estimates were gained from com-bined log files. The best-fit evolution model of nucleotide

774 M.S. Colomba et al.

Fig. 2. Maximum Likelihood consensus tree obtained by the combined analysis of 16S rDNA and COI datasets (substitution method:HKY+G, G = 0.1627). All positions containing gaps and missing data were eliminated; the tree is drawn to scale, with branch lengthsmeasured in the number of substitutions per site. Numbers above branches represent bootstrap values (after 1000 replicates) andBayesian posterior probabilities (in parentheses). A timeline in million years before present for the C. westerlundi complex evolutionperformed by MEGA is reported below the tree.

substitution resulted in HKY with empirical base compo-sition; the Yule Process tree prior for mitochondrial datawith piecewise linear population size model was applied witha UPGMA-generated tree as the starting point. Five sin-gle runs were combined with the LogCombiner 1.6.1 im-plemented in the software package BEAST. Trees from allruns were combined to produce an ultrametric consensustree using TreeAnnotator 1.6.1. The first 2× 106 trees werediscarded as burnin. Support for nodes is expressed as pos-terior probabilities.

Moreover, the Maximum Likelihood (ML) method wasperformed by MEGA 5.0.3 (Tamura et al. 2011); accord-ing to the “Find Best DNA Models (ML)” option avail-

able in this software, HKY + G resulted the best-fitsubstitution model. Bootstrap values were assessed after1000 replicates. C. patulum (Draparnaud, 1801), C. ele-gans (Clessin, 1879), C. septemspirale (Razoumowsky, 1789)and C. roseoli (A.J. Wagner, 1901) (see also Webster etal. 2012) were used as out-groups to root phylogenetictrees. In this regard, please note that the C. roseoli COIsequence (uploaded by us, GenBank ID: KF812521, Ta-ble 1) is from the same population as the C. roseoli 16Ssequence in Webster et al. (2012). Both Bayesian andMaximum Likelihood analyses were performed either withsingle (16S or COI ) or combined (16S-COI ) moleculardatasets.

Cochlostoma (Holcopoma) westerlundi in Italy 775

Fig. 3. Shells of Cochlostoma w. yapigium from Nardo, Portoselvaggio (LE, Apulia) (A); C. w. dionysii from Siracusa (SR, Sicily)(B); C. w. westerlundi from Monte Consolino (Stilo, RC, Calabria) (C).

Results

Molecular dataThree COI haplotypes, restricted to single geographicareas, Sicily, Calabria and Apulia, were recorded. Thesame was done for the 16S rDNA sequences, includ-ing three haplotypes each of which was unique forone of the three Italian regions. Both Bayesian andML analyses resulted in phylogenetic trees, the topolo-gies of which were largely consistent, showing threeclades with a strong branch support. Figure 2 illus-trates the ML consensus tree obtained by the com-bined analysis of 16S rDNA and COI datasets. Three

distinct clusters are clearly evident. The first one in-cludes C. w. yapigium and C. paganum specimens(CYLE1-2, CYNA1-2, CYBL1-2 e CYOTR1-3); thesecond includes C. w. dionysii specimens (CDSR1-6)and, finally, the third includes C. w. westerlundi spec-imens (CW1-6). It is interesting to note that, in thefirst cluster, C. w. yapigium and C. paganum spec-imens are intermixed. In addition, by comparison ofthe mtDNA nucleotide sequences, it was found that16S amplicons of C. w. yapigium were identical tothose of C. paganum and the same was found for COI,which, for the first time, confirmed by molecular ev-idence the synonymy between the two nominal taxa

776 M.S. Colomba et al.

(hereafter not further distinguished and indicated as“yapigium”).Divergence among groups (DxyJC) ranged from

2.6% to 5% (16S rDNA) and from 6.3% to 8.3% (COI ).By entering a standard COI rate estimate of ca. 1% permillion years per site (see Donald et al. 2005) into the“calibrating MolClock” option included in MEGA soft-ware, it was determined that the C. westerlundi groupshared an exclusive common ancestor up to 5.01 mil-lion years ago (Mya), when C. westerlundi s. str. di-verged from the yapigium-dionysii ancestor. Under thesame model, it was further estimated that yapigium anddionysii were separated by about 2.57 Mya. Finally, themost recent common ancestor (TMRCA) of C. wester-lundi group and C. roseoli, both belonging to the sub-genus Holcopoma Kobelt et Moellendorf, 1899, mightbe dated back to about 7.14 Mya.

Morphological dataIn this study we describe, for the first time, the re-productive system of C. (H.) w. westerlundi specimensfrom Calabria, although, because of its restricted dis-tribution area and relatively small populations, it waspossible to examine only five individuals (one male andfour females).

Cochlostoma w. yapigiumShell (Fig. 3A). Protoconch smooth, with no ribs.Teleoconch conical, ribbed, with rounded, regularly in-creasing whorls and deep sutures. Base of the last whorlwell-rounded. Shell yellowish with small brown-reddishspots arranged in interrupted bands. Ribs oblique, con-densed and regularly spaced, slightly prominent onlyat the end of the last whorl. Last whorl slightly ris-ing towards the aperture. Aperture rounded more orless oblique with a well-developed lip. Columellar lobewell-developed and abruptly curved, covering umbili-cus. Shell wider and more elongated in females. Males:shell height 7.24–8.10 mm, diameter 3.90–4.00 mm; fe-males: height 7.52–9.02 mm, diameter 4.19–4.43 mm.Male genitalia (Fig. 4A). Simple penis, cylinder-

conic, gradually tapering to the end with a long, thinapex and internal penial ductus.Female genitalia (Fig. 4B). Sperm receptacle ab-

sent. Bursa copulatrix medium-sized with a short andnarrow duct originating near the proximal portion.

Cochlostoma w. dionysiiShell (Fig. 3B). Protoconch smooth, with no ribs. Teleo-conch conical, prominently ribbed with seven roundedwhorls and deep sutures. Base of the last whorl wideand rounded. Ribs oblique, prominent, close, regularin space, becoming thicker and slightly prominent onlyat the end of the last whorl. Last whorl slightly ris-ing towards the aperture. Aperture rounded with awell-developed lip. Columellar lobe well-developed andabruptly curved, concealing umbilicus. Shell wider andmore elongated in females. Males: shell height 8.40–8.94 mm, diameter 4.40–4.70 mm; females: height 8.62–9.65 mm, diameter 4.30–4.90 mm.

Male genitalia (Fig. 4C). Simple penis, more orless cylindrical and narrowing abruptly at the end; thinapex, shorter than in yapigium.Female genitalia (Fig. 4D). Sperm receptacle ab-

sent. Bursa copulatrix medium-sized with a short(shorter than in yapigium) and narrow duct originat-ing near the proximal portion.

Cochlostoma w. westerlundiShell (Fig. 3C). Protoconch without any ribbing. Teleo-conch conical, elongated, densely ribbed, with roundedwhorls and deep sutures. Base slightly angled and rel-atively narrower than in yapigium and dionysii. Shellyellowish-brown with small reddish marks arranged inthree interrupted bands. Ribs on first whorls slightlyoblique, prominent, widely spaced, thin, becoming in-distinct towards the aperture. Last whorls with five toeight thin, parallel, spiral ribs. Last whorl slightly ris-ing towards the aperture. Aperture rounded, with du-plicated broadly-reflected peristome only at the basaland columellar side, inner lip well-developed and sepa-rated from the outer lip by a furrow. Columellar lobewell-developed and abruptly curved, covering the um-bilicus. Distance between base of the last whorl andthe upper edge of the columellar lobe wider than inyapigium and dionysii. Shell wider and more elongatedin females. Males: shell height 7.50–9.60 mm, diameter2.50–3.85 mm; females: height 10.00–11.00 mm, diam-eter 4.00–5.00 mm.Male genitalia (Fig. 4E). Penis simple, robust,

broad, rough on the inner edge; apex short and bentbackwards.Female genitalia (Fig. 4F). Sperm receptacle ab-

sent. Bursa copulatrix wide with a long duct (twice aslong as in dionysii and yapigium).

Comparative notesThe main differences between yapigium and dionysiirelate to: (i) rib count (ribs are fewer and more widelyspaced in dionysii); (ii) size and morphology of the lastwhorl of the shell (wider and more rounded in dionysii);(iii) penial apex (less elongated in dionysii); (iv) lengthof the bursa duct (slightly shorter in dionysii).Compared to yapigium-dionysii, westerlundi is

characterized by (i) an angled shell base (rounded inyapigium and dionysii); (ii) the first whorls showingspaced ribs becoming thinner and more widely spacedin the last whorls (regular in yapigium and dionysii);(iii) the final whorl showing blurred lines at the open-ing (contrary to dionysii and yapigium where the finalwhorl is ribbed) and some parallel, thin, spiral lines(absent in yapigium and dionysii); (iv) final portion ofthe last whorl slightly ascending (markedly ascendingin yapigium and dionysii); (v) the edge of the peristomepoorly developed, with the exception of the columellarpart (entirely well developed in dionysii and yapigium);(vi) a big distance between the columella and peris-tome, much greater than in dionysii and yapigium; (vii)a more robust penis, ending with a stocky apex (thinapex in yapigium and dionysii); and finally, (viii) fe-

Cochlostoma (Holcopoma) westerlundi in Italy 777

Fig. 4. Genitalia of Cochlostoma w. yapigium male (A) and female (B) from Nardo, Portoselvaggio (LE, Apulia); C. w. dionysii male(C) and female (D) from Siracusa (SR, Sicily); and C. w. westerlundi male (E) and female (F) from Monte Consolino (Stilo, RC,Calabria).

males showing a long channel of the bursa copulatrix(much longer than in yapigium and dionysii).

Discussion

The present study provides original data accumulatedby this research team during the last few years onC. westerlundi s.l. from Italy, a very interesting andmorphologically conservative group ascribed to the sub-genus Holcopoma showing very low active dispersalability and a patchy scattered geographic distribution.Main results include: (i) characters from morpholog-ical structures (i.e., shell and genital anatomy) showsome differences among the three taxa (C. w. yapigium,

C. w. dionysii and C. w. westerlundi); and (ii) molecu-lar phylogenetic analysis revealed three clusters exactlycorresponding to the three subspecies. Taking into ac-count the phenotypic features, nucleotide divergencesand topology of the phylogenetic trees, we strongly sug-gest elevating these three subspecies to the higher rank.Hence, in the remainder of this paper – in line with Al-zona (1971) – these three taxa will be referred to asspecies.In many animal groups, including Gastropoda,

species-level identification based on molecular mark-ers is still contentious due to two main factors: (i) ge-netic distances are often expressed in different units,which makes them difficult to compare and (ii) there

778 M.S. Colomba et al.

is no constant or universal molecular clock. There isno simple, “routine” way of interpretation of geneticdistances while assessing species distinctness. In fact,distance values are applicable only within a group ofclosely related species, and we are not aware of suchdata for the Cochlostomatidae. Nevertheless, withinthe C. westerlundi group, nucleotide divergences arein line with genetic distances reported as discriminat-ing among a few Caenogastropoda species. For exam-ple, 16S rDNA distances ≥0.03% were reported for sev-eral Ovulidae (Schiaparelli et al. 2005); 1.3% betweenKelletia kelletii and K. lischkei and 1.5% between Pe-nion chathamensis and P. sulcatus (Hayashi 2005); lessthan 4–5% for Conus species (Duda et al. 2009); andranging from 0.30% to 12.40% between specimens ofcongeneric neogastropod species (Zou et al. 2011). ForCOI distances, values of at least 3% were reported forHydrobia (Wilke et al. 2000), ranging from 0.3% to26.6% in Trochidae (Donald et al. 2005); ≥1.5% in By-thinella (Bichain et al. 2007); from 2.1% to 19.8% in40 Neogastropoda species (Zou et al. 2011). Therefore,genetic differences among the three taxa under study(DxyJC, 2.6–5% for 16S rDNA; 6.3–8.3% for COI ) maybe considered as the result of a certain degree of isola-tion and divergence, i.e., the very low active dispersalability of these animals is likely to have drastically re-duced the gene flow and given rise to three geneticallyindependent species. All these findings led us to sug-gest that the three clusters, supported by strong boot-strap values and very high posterior probability, rep-resent three phylogenetic species which we refer to asC. westerlundi, C. yapigium and C. dionysii. Regardingthe morphological characters, the shell phenotype andgenital architecture in C. yapigium and C. dionysii arerather conservative. In fact, these two taxa inhabitinggeographically well separated areas (Apulia and Sicily)have maintained quite homogeneous features. On thecontrary, C. westerlundi s. str. appears to be quite dif-ferent. Such an evident differentiation might be justifiedby a longer evolutionary history (it was the first one tosplit out from the group) and geographical and ecolog-ical isolation of the taxon (small populations occurringin restricted mountainous areas of southern Calabria).A possible paleogeographic reconstruction of the

history of the C. westerlundi group dates back to thevery end of the Miocene, in the Messinian (7.2–5.3million years ago), when a series of tectonic move-ments caused the closure of the connections between theMediterranean and the Atlantic Ocean. The Mediter-ranean became a huge lake and gradually began to dryup; many lands emerged from the waters and new con-nections between areas previously isolated by the seawere settled. Divergence times reported herein wouldbe in line with La Greca (1967), who hypothesizedthe ancestor of the Cochlostoma (Holcopoma) as oneof the numerous groups of terrestrial mollusks which,at the beginning of the Messinian period, expandedfrom the Balkan peninsula to southern Italy up to Sicilythanks to the partial desiccation of the Mediterranean(see also Sacchi 1954; Thake 1985, Giusti et al.1995;

Colomba et al. 2010). According to our results, theseparation between C. roseoli and C. westerlundi s.l.might be roughly dated back to about 7.14 millionyears ago. Then, in our opinion, the banishing of a rel-atively small group of specimens in isolated mountain-ous areas – combined with the low active dispersioncapacity of these calciphilous organisms – reduced thegene flow within the original population, thus originat-ing C. westerlundi s. str. which separated from the restof the group about 5.01 Mya. During Pliocene, the re-opening of the Strait of Gibraltar caused the Mediter-ranean Sea level to rise and the formerly associated ter-ritories became separated, creating overall conditionsof greater insularity. In the late Pliocene (about 2.57Mya), these populations, already fragmented, divergedinto two separate taxa: C. yapigium and C. dionysii. Al-ternatively, Girod & Sacchi (1967) suggested a passiveintroduction of C. dionysii by humans in Sicily since itwas known only for the archaeological area of Siracusa.However, C. dionysii is fairly widespread throughoutthe Iblei mountains, even in areas not interested by hu-man settlements (see Fig. 1). Hence, in our opinion, itis to be considered a native species.As far as concerns the ecology of these organisms,

Cochlostoma westerlundi s.l. are rupicolous and cal-ciphilous; C. yapigium show a marked preference forsunny habitats and are widespread across the Salento.C. dionysii prefer shady areas and often take refuge inrocky cracks. Occasionally, they occur in other typesof environment; for example, we found several speci-mens in the bark of poplars (Populus sp.) and willows(Salix sp.) along river banks. They are also widespreadon the eastern side of the Iblei Mts. (SE Sicily) fromthe first outcrops of limestone a few meters above sealevel (Eloro Cape, SR) up to medium altitudes (670 ma.s.l., Palazzolo Acreide, SR). C. westerlundi are usu-ally found both in shady areas and sunny rocks. Theyare extremely localized, with populations showing verylow numerical density, which makes this species vulner-able to extinction and, therefore, worthy of attentionand more protection.In conclusion, we hope that the present findings

may (i) shed some light on several aspects, neverfocused on before, of the origin and evolution ofCochlostoma (Holcopoma) westerlundi s.l. and (ii) pro-vide new data to support a taxonomic revision of thegroup by elevating the three subspecies to the speciesrank.

Acknowledgements

The authors wish to thank D. Ferreri (Lecce, Italy), A.Brancato (Siracusa, Italy) and A. Corso (Siracusa, Italy)for contributing to specimen collection in the field and S.Giglio (Cefalu, Italy) for taking pictures.

References

Alzona C. 1971. Malacofauna Italica. Catalogo e bibliografia deimolluschi viventi, terrestri e d’acqua dolce. Atti della Societa

Cochlostoma (Holcopoma) westerlundi in Italy 779

Italiana di Scienze Naturali e del Museo Civico di Storia Nat-urale di Milano 111: 1–433.

Bank R.A. 2012. Fauna Europaea: Gastropoda, Cochlostomidae.Fauna Europaea version 1.1. http://www.faunaeur.org (ac-cessed 06.07.2012)

Bichain J.-M., Gaubert P., Samadi S. & Boisselier-DubayleM.C. 2007. A gleam in the dark: Phylogenetic species delimi-tation in the confusing spring-snail genus BythinellaMoquin-Tandon, 1856 (Gastropoda: Rissooidea: Amnicolidae). Mol.Phylogen. Evol. 45 (3): 927–941. DOI: 10.1016/j.ympev.2007.07.018

Bodon M., Favilli L., Giannuzzi Savelli R., Giovine F., GiustiF., Manganelli G., Melone G., Oliverio M., Sabelli B. &Spada G. 1995. Gastropoda Prosobranchia, HeterobranchiaHeterostropha. In: Minelli A., Ruffo S. & La Posta S.(eds), Checklist delle specie della fauna italiana, Fascicola14,Edizioni Calderini, Bologna, 60 pp. http://www.comitato.faunaitalia.it/docs/F14.DOC (accessed 10.07.2013)

Colomba M.S., Gregorini A., Liberto F., Reitano A., Giglio S. &Sparacio I. 2010. Molecular analysis of Muticaria syracusanaand M. neuteboomi from Southeastern Sicily (Gastropoda,Pulmonata, Clausiliidae). Biodivers. J. 1 (1-4): 7–14.

Colomba M.S., Gregorini A., Liberto F., Reitano A., Giglio S. &Sparacio I. 2011. Monographic revision of the endemic Helixmazzullii (De Cristofori & Jan, 1832) complex from Sicily andre-introduction of the genus Erctella Monterosato, 1894 (Pul-monata, Stylommatophora, Helicidae). Zootaxa 3134: 1–42.

Donald K.M., Kennedy M. & Spencer G.S. 2005. Cladogenesisas the result of long-distance rafting events in South PacificTopshells (Gastropoda, Trochidae). Evolution 59 (8): 1701–1711. DOI: 10.1554/04-553.1

Drummond A.J. & Rambaut A. 2007. BEAST: Bayesian evolu-tionary analysis by sampling trees. BMC Evol. Biol. 7: 214–257. DOI: 10.1186/1471-2148-7-214

Duda T.F.Jr, Kohn A.J. & Matheny A.M. 2009. Cryptic speciesdifferentiated in Conus ebraeus, a widespread tropical marinegastropod. Biol. Bull. 217 (3): 292–305. PMID: 20040753

Ferreri D., Bodon M. & Manganelli G. 2005. Molluschi terrestridella provincia di Lecce. Thalassia Salentina 28: 31–130. DOI:10.1285/i15910725v28p31

Folmer O., Black M., Hoeh W., Lutz R. & Vrijenhoek R. 1994.DNA primers for amplification of mitochondrial cytochromec oxidase subunit I from diverse metazoan invertebrates. Mol.Mar. Biol. Biotechnol. 3 (5): 294–299. PMID: 7881515

Forcart L. 1965. Rezente Land- und Susswassermollusken dersuditalienischen Landschaften Apulien, Basilicata und Cal-abrien. Verh. Naturforsch. Gesell. Basel 76: 59–184.

Girod A. & Sacchi C.F. 1967. Considerazioni biogeografiche sullamalacofauna pugliese. Atti della Societa Italiana di ScienzeNaturali e del Museo Civico di Storia Naturale di Milano 106(4): 258–274.

Giusti F., Manganelli G. & Schembri P.J. 1995. The non-marinemolluscs of the Maltese Islands. Monografie Museo RegionaleScienze Naturali Torino 15: 1–607. ISBN: 9788886041249

Groenenberg D.S.J., Neubert E. & Gittemberg E. 2011. Reap-praisal of the “Molecular phylogeny of Western PalaearticHelicidae s.l. (Gastropoda: Stylommatophora)”: When poorscience meets GenBank. Mol. Phylogen. Evol. 61(3): 914–923. DOI: 10.1016/j.ympev.2011.08.024.

Hall T.A. 1999. BioEdit: a user-friendly biological sequence align-ment editor and analysis program for Windows 95/98/NT.Nucleic Acids Symposium Series 41: 95–98.

Hayashi S. 2005. The molecular phylogeny of the Buccinidae(Caenogastropoda: Neogastropoda) as inferred from the com-plete mitochondrial 16S rRNA gene sequences of selected rep-resentatives. Mollusc. Res. 25 (2): 85–98.

Heled J. & Drummond A.J. 2010. Bayesian inference of speciestrees from multilocus data. Mol. Biol. Evol. 27 (3): 570–580.DOI: 10.1093/molbev/msp274

La Greca M. 1967. Considerazioni sulla origine del popolamentofaunistico pugliese. Archivio Botanico e Biogeografico Ital-iano 43: 297–320.

Liew T.-S., Schilthuizen M. & Vermeulen J.J. 2009. System-atic revision of the genus Everettia Godwin-Austen, 1891(Mollusca: Gastropoda: Dyakiidae) in Sabah, northern Bor-neo. Zool. J. Linn. Soc. London 157 (3): 515–550. DOI:10.1111/j.1096-3642.2009.00526.x

Mitani T., Akane A., Tokiyasu T., Yoshimura S., Okii Y. &Yoshida M. 2009. Identification of animal species using thepartial sequences in the mitochondrial 16S rDNA gene. LegalMedicine 1 (Suppl 1): S449–450. DOI: 10.1016/j.legalmed.2009.02.002

Pintér L. & Szigethy A.S. 1976. Schnecken aus Sizilien [Szicíliaicsigák]. Soosiana 4: 27–38.

Reitano A., Liberto F., Sparacio I. & Giglio S. 2009. I mol-luschi terrestri della R.N.I. “Grotta Palombara” (Melilli, Si-cilia sud-orientale) (Gastropoda Architaenioglossa, Neotae-nioglossa, Stylommatophora). Il Naturalista siciliano S. IV,33 (1-2): 177–205.

Sacchi C.F. 1954. Note di malacologia terrestre pugliese. Boll.Zool. 21(1): 51–76. DOI: 10.1080/11250005409439190

Schiaparelli S., Barucca M., Olmo E., Boyer M. & CanapaA. 2005. Phylogenetic relationships within Ovulidae (Gas-tropoda: Cypraeoidea) based on molecular data from the16S rRNA gene. Mar. Biol. 147 (2): 411–420. DOI:10.1007/s00227-005-1566-0

Tamura K., Peterson D., Peterson N., Stecher G., Nei M. & Ku-mar S. 2011. MEGA5: molecular evolutionary genetics anal-ysis using Maximum Likelihood, Evolutionary Distance, andMaximum Parsimony methods. Mol. Biol. Evol. 28 (10):2731–2739. DOI: 10.1093/molbev/msr121

Thake M.A. 1985. The biogeography of the Maltese islands, il-lustrated by the Clausiliidae. J. Biogeogr. 12 (3): 269–287.

Uit De Weerd D.R. & Gittenberger E. 2004. Re-evaluatingCarinigera: molecular data overturn the current classifica-tion within the Clausiliid subfamily Alopiinae (Gastropoda,Pulmonata). J. Mollusc. Stud. 70 (4): 305–318. DOI:10.1093/mollus/70.4.305

Webster N.B., Van Dooren T.J. & Schilthuizen M. 2012. Phy-logenetic reconstruction and shell evolution of the Diplom-matinidae (Gastropoda: Caenogastropoda). Mol. Phylogen.Evol. 63 (3): 625–638. DOI: 10.1016/j.ympev.2012.02.004

Wilke T., Rolán E. & Davis G.M. 2000. The mudsnail genusHydrobia s.s. in the northern Atlantic and western Mediter-ranean: a phylogenetic hypothesis. Mar. Biol. 137 (5-6):827–833. DOI: 10.1007/s002270000407

Zou S., Li Q., Kong L., Yu H. & Zheng X. 2011. Comparingthe usefulness of distance, monophyly and character-basedDNA barcoding methods in species identification: A casestudy of Neogastropoda. PLoS One 6 (10): e26619. DOI:10.1371/journal.pone.0026619.

Received August 18, 2013Accepted February 10, 2014