Ocinebrina edwardsii (Payraudeau, 1826)

Barco, Andrea, Houart, Roland, Bonomolo, Giuseppe, Crocetta, Fabio & Oliverio, Marco, 2013, Molecular data reveal cryptic lineages within the northeastern Atlantic and Mediterranean small mussel drills of the Ocinebrina edwardsii complex (Mollusca: Gastropoda: Muricidae), Zoological Journal of the Linnean Society 169 (2), pp. 389-407 : 397-402

THE OCINEBRINA EDWARDSII COMPLEX

In the four-MOTUs scenario, supported by the ABGD method only (yet within the confidence limit of the GMYC analysis), the four major lineages (I–IV) comprised the following samples:

MOTU-I: specimens morphologically identified as O. piantonii and O. carmelae from the Gulf of Gabès.

MOTU-II: specimens morphologically identified as O. edwardsii from the Aegean Sea, Ionian Sea, southwestern Sicily and Lampedusa Island.

MOTU-III: specimens morphologically identified as O. edwardsii from the Tyrrhenian Sea (including

Nominal taxon Type locality [listed localities] Type material Currently accepted as Notes

Murex aciculatus Lamarck, 1822: 176 Vannes area ( France) Not known O. aciculata Type species of Ocinebrina Jousseaume, 1880 and Corallinia Bucquoy & Dautzenberg, 1882 (in Bucquoy, Dautzenberg & Dollfus, 1882: 24)

Murex acrisius Nardo, 1847: 59–60 Gulf of Venice ( Italy) 14 syntypes MSNVE 21980 (Crocetta O. aciculata Originally referring to: Chiereghin (unpublished*):

et al., 2012) figs. 713–714. Subjective junior synonym ( Crocetta et al., 2012)

Ocinebrina edwardsi [sic!] var. apiculata Pallary , Tanger ( Morocco) Not known O. edwardsii

1902a: 314

Murex baeticus Reeve, 1845 : pl. 32 (fig. 162) Not known 3 syntypes NHMUK 1972024 – one in O. edwardsii Probable subjective senior synonym of Murex hybridus Houart (2001) as BMNH 1621845

Ocinebrina buzzurroi Cecalupo & Mariani, 2008 : Borj el Hissar ( Tunisia) Holotype MNHM 33489 (Cecalupo et al., O. corallinoides Subjective junior synonym ( Crocetta et al., 2012)

in Cecalupo et al., 2008: 90, pl. 43 (figs. 1–7) 2008)

Ocinebrina carmelae Cecalupo, Buzzurro & Borj el Hissar ( Tunisia) Holotype MNHM 33491 (Cecalupo et al., O. carmelae New subjective junior synonym of O. piantonii

Mariani, 2008: 98, pl. 48 (figs. 1–8) 2008) (? = Murex hybridus )

Amyclina compacta Nordsieck, 1968: 140 , pl. 23 Portovenere ( Italy) Holotype nr./1, Senckenberg Mus. O. edwardsii

(fig. 80.45) ( Houart, 2001)

Ocinebrina corallinoides Pallary, 1912: 221 , plate Sfax, Gulf of Gabès ( Tunisia) Not known O. corallinoides

(fig. 48)

Murex corallinus Scacchi, 1836: 12 (fig. 15) Baia ( Italy) Neotype MZN Z7010 (Crocetta et al., O. aciculata Subjective junior synonym ( Houart, 2001; Crocetta 2012) et al., 2012)

Murex costulatus Nardo, 1847: 55–56 (sp. 8) Cres Island area ( Croatia) Not known Originally referring to: Chiereghin (unpublished*): figs. 693–694. Hereby first considered subjective junior synonym of Fusus helleri . Primary junior homonym of Murex costulatus Schröter 1805 and Murex costulatus Risso, 1826

Ocinebrina edwardsi [sic!] var. crassata Pallary , Tanger ( Morocco) Not known Hereby considered as belonging to the Ocinebrina

1902b: 12, pl. 1 (figs. 10–11) edwardsii complex

Ocinebrina cyclopus Monterosato, 1884: 112 Palermo ( Italy) Syntypes in MCZR ( Settepassi, 1977) O. edwardsii

Purpura edwardsii Payraudeau, 1826: 155 , pl. 7 Corsica ( France) Not known O. edwardsii Type species of Dentocenebra Monterosato, 1917

(figs. 19–20) (= Ocinebrina Jousseaume, 1880 )

Ocinebrina erronea Cecalupo, Buzzurro & Sfax and Kerkennah Islands Not designated O. hispidula Subjective junior synonym (Cossignani & Ardovini,

Mariani, 2008: 92, pls. 42 (figs. 5–6), 44 (figs. ( Tunisia) 2011), hereby assigned to Cecalupo et al. (2008)

7–16) and 45 (figs. 1–10) since Muricopsis erroneus Settepassi, 1977 is unavailable ( ICZN, 1999: Art. 11.4)

Ocinebrina aciculata exilis Houart, 2001: 19 (figs. NW of Bou Grara Sea ( Tunisia) Holotype MNHN 0362 (Crocetta et al., O. corallinoides Subjective junior synonym (Bonomolo & Buzzurro,

8–9), 51 (fig. 62), 62, 143 (figs. 167–168), 176 2012) 2006; Cecalupo et al., 2008; Crocetta et al., 2012)

(figs. 324–325)

Fusus helleri Brusina, 1865: 8 [Zadar, Šibenik, Hvar, Not known O. helleri = Murex hellerianus Brusina, 1866: 63 (unnecessary Dubrovnik, Budva] ( Croatia spelling emendation: Houart, 2001)

and Montenegro)

Ocinebrina edwardsi [ sic!] var. hispidula Pallary, Gulf of Gabès ( Tunisia) 3 syntypes MNHN 1001 ( Houart, 2001; O. hispidula

1904: 231, pl. 7 (fig. 18) Giannuzzi-Savelli et al., 2003; present

paper)

Murex hybridus Aradas & Benoît, 1876: 272 , pl. 5 Palermo ( Italy) Not known O. hybrida Probable subjective junior synonym of Murex baeticus

(fig. 9)

Murex inconspicuus Sowerby G.B. II, 1841: 5 (67), Jersey Island (Bailiwick of Not known O. aciculata Subjective junior synonym ( Houart, 2001; Crocetta

plates (figs. 81) Jersey) et al., 2012)

Murex inglorius Crosse, 1865: 213 , pl.6 (fig.4) Not known Holotype MNHN 0993 ( Fair, 1976; O. ingloria

Houart, 2001; Giannuzzi-Savelli et al.,

2003; present paper)

Ocenebra inordinata Houart & Abreu, 1994: 123 , Madeira Island Holotype MMF 25429 (Houart & Abreu, O. inordinata

129 (figs. 11–13) 1994)

Ocinebrina labiosa Monterosato, 1884: 112 [Chioggia and Venice] ( Italy) Presumably not existing Originally referring to: Chiereghin (unpublished*): figs. 695–696.

New subjective junior synonym of Murex inglorius

Merex [sic!] labiosus Nardo, 1847: 55–56 (sp. 9) Kvarner Gulf ( Croatia) Presumably not existing Ocenebra erinaceus Originally referring to: Chiereghin (unpublished*): figs. 695–696. Currently accepted as Ocenebra erinaceus , and here considered as a member of the O. edwardsii complex. Primary junior homonym of Murex labiosus Gray, 1828 and Murex labiosus Wood, 1828

Ocinebrina leukos Houart, 2000: 464 (figs. 7–10) La Isleta, Lanzarote ( Spain) Holotype MNHN 0966 ( Houart, 2000; O. leukos

present paper)

Ocinebrina miscowichi Pallary, 1920: 40 , plate Essaouira ( Morocco) 32 syntypes MNHN 0176 and 0177 O. miscowichae Spelling emended by Houart (1997)

(figs. 5–6) ( Houart, 2001; present paper)

Ocinebrina nicolai Monterosato, 1884: 112 [ Corsica, Sardegna, Lipari] ( Italy Syntypes in MCZR ( Settepassi, 1977) O. nicolai

and France)

Murex orcomenus Nardo, 1847: 57–58 (sp. 11) Northern Adriatic Sea Not known Originally referring to: Chiereghin (unpublished*): figs. 699–700.It is probably a senior synonym of Murex inglorius , but its status as nomen oblitum ( ICZN, 1999: Art. 23.9.1.2) should be verified

Ocinebrina paddeui Bonomolo & Buzzurro, 2006 : Off Capo Caccia , Alghero ( Italy) Holotype MNHN 29909 (Bonomolo & O. paddeui

1 (figs. 1–6) Buzzurro, 2006)

Murex pereger Brugnone, 1873: 10 , plate (fig. 17) African shores Not known O. hybrida Subjective junior synonym ( Houart, 2001). Primary junior homonym of Murex pereger Beyrich, 1854

Murex edwardsi [sic!] f. perigmus De Gregorio, Mediterranean Not known O. edwardsii Originally referring to: Bucquoy et al. (1882): pl. 2

1885: 253 (fig. 3). Subjective junior synonym ( Houart, 2001)

Murex edwardsi [sic!] f. perilus De Gregorio, Mondello ( Italy) Not known O. edwardsii

1885: 253

Ocinebrina piantonii Cecalupo, Buzzurro & Borj el Hissar ( Tunisia) Holotype MNHM 33490 (Cecalupo et al., O. piantonii Probable subjective junior synonym of Murex baeticus

Mariani, 2008: 96, pl. 49 (figs. 1–10), pl. 50 2008)

(figs. 1–6)

Murex pistacia Reeve, 1845 : pl. 34, fig. 174 Not known 3 syntypes NHMUK 1972021 (Crocetta O. aciculata Subjective junior synonym ( Houart, 2001; Crocetta

et al., 2012) et al., 2012)

Murex pumilus Küster, 1869 (in Küster & Kobelt, Not known Not known O. edwardsii Primary junior homonym of Murex pumilus Broderip,

1839–1878: 118, pl. 35: figs. 8–10) 1833 and Murex pumilus A. Adams, 1853

Ocinebrina purpuroidea Pallary, 1920: 39 , plate Rabat ( Morocco) 8 syntypes MNHN 0931 ( Houart, 2001) O. purpuroidea

(figs. 16–17, 24)

Ocinebrina reinai Bonomolo & Crocetta, 2012 : (in Procida Island ( Italy) Holotype MNHN 24566 (Crocetta et al., O. reinai

Crocetta et al., 2012: 180, figs. 1 K-1 L, 2A-2F, 2012)

2I, 2 L, 3)

Murex semiclausus Küster, 1869: 111 , pl. 34 (figs. Not known Not known O. edwardsii

6–7)

Murex subaciculatus Locard, 1886: 164 Toulon ( France) Not known O. aciculata Originally referring to: Murex aciculatus in Hidalgo (1870): pl. 13 (figs. 7–8); Murex aciculatus var. curta Bucquoy et al. (1882) : page 25. Subjective junior synonym ( Houart, 2001; Crocetta et al., 2012)

Fusus titii Stossich, 1865: 31 Northern Adriatic Sea Not known O. aciculata Subjective junior synonym ( Houart, 2001; Crocetta

et al., 2012)

Ocinebra [sic!] wardiana Baker, 1891: 134 , pl. 11 ‘Australia’ (see discussion in Holotype CAS 20698 ( Vokes, 1994) O. aciculata Subjective junior synonym ( Houart, 2001; Crocetta

(fig. 5) Vokes, 1994) et al., 2012)

Murex weinkauffianus Crosse, 1866: 274 , pl. 8 Zadar ( Croatia) Not known O. helleri Subjective junior synonym ( Houart, 2001)

(fig. 4)

Corsica, type locality of O. edwardsii ); specimens morphologically identified as O. hispidula (from the Gulf of Gabès); specimens from the Mediterranean and Atlantic coasts of Spain and Portugal, including the Canary Islands; part of the specimens from the Venice Lagoon.

MOTU-IV: specimens from the Adriatic Sea and a few localities in the Ionian Sea.

In the eight-MOTUs scenario, supported by all three methods (ABGD, GMYC, TCS), the eight major lineages maintained a geographical coherence, and only MOTU-A corresponded exactly to MOTU-I.

We cannot decide with the present data between the two scenarios (four-MOTUs or eight-MOTUs). Pending the availability of more reliable and congruent data from the nuclear genes to support such a decision, we discuss the geographical patterns in the O. edwardsii complex focusing on the more detailed (albeit admittedly less conservative) eight-MOTUs scenario.

In the eight-MOTUs scenario, all the specimens attributed preliminarily to O. edwardsii were distributed among seven MOTUs: clade B, including specimens from localities in the Aegean Sea; clade C from the Ionian Sea, southwestern Sicily and Lampedusa Island; clade D2 from the Atlantic coast of Spain and Portugal, the Canary Islands and Málaga Province; clade E from several localities across the Tyrrhenian Sea; clade F from Pellestrina, a locality within the Venice Lagoon; clade G from Campania (Pozzuoli and Palinuro); and clade H from the Adriatic Sea and a few localities in the Ionian Sea.

Of these groups, we provisionally identify MOTU-E as representing the ‘true’ O. edwardsii , because it included the specimens sampled in Corsica, its type locality ( Payraudeau, 1826). Among our samples, this group is apparently restricted to the Tyrrhenian Sea (eastern Corsica, eastern Sardinia , northern Sicily and western Italian coast: Fig. 4 ) .

The sister lineage to clade E is composed of the D1 and D2 lineages (clade D). The D2 lineage included specimens morphologically ascribed to O. edwardsii sampled along the Atlantic coasts of Spain and Portugal, from the Canary Islands and from Málaga Province; the specimens ascribed to O. leukos from the Canary Islands; and specimens from populations referred in the recent literature to O. [cf.] nicolai from Galicia ( Rolán, 1983) (samples 963, 965, 1033, 1058, 1059, 1061–1063), from Barbate (sample 1282) ( Gofas, 2011) and from off Ponta da Piedade–Algarve (sample 1294) ( Afonso et al., 2011) ( Fig. 4 ). The specimens of O. edwardsii sampled along the Atlantic coast of Spain and Portugal displayed a rather distinctive shell morphology with respect to the most common forms found in the Mediterranean. Ocinebrina leukos was described from Lanzarote (Canary Islands) as living sympatrically with, yet morphologically easily distinguished from, local O. edwardsii ( Houart, 2000) . The latter, however, has been recorded as highly variable morphologically, even in the same locality ( Houart, 2000), and, according to our molecular data, O. leukos may fall within the variability of this lineage (D2). Ocinebrina nicolai was originally described from Corsica, Sardinia and Lipari Islands ( Monterosato, 1884). Specimens from the Monterosato collection (see Settepassi, 1977: 38–44), however, are morphologically different from those figured recently under this name (e.g. Rolán, 1983; Houart, 2001; Afonso et al., 2011; Gofas, 2011). Therefore, the association of the name O. nicolai with one of our MOTUs is postponed pending the genetic analysis of topotypical specimens.

According to the ABGD and GMYC analyses, clade D also included the specimens from Gabès ascribed to O. hispidula , although a distinct well-supported clade (D1) for this lineage has been recovered in all the topologies. The large genetic distance between specimens identified as O. hispidula from Gabès and specimens classified preliminarily as O. edwardsii from Sicily (clade C) has been reported recently by Barco et al. (2013). Ocinebrina hispidula has also been reported from other Atlantic and Mediterranean localities, such as Mogador ( Ardovini & Cossignani, 2004), Sicily ( Settepassi, 1977; Giannuzzi-Savelli et al., 2003), Cyprus ( Houart, 2001; Öztürk, Buzzurro & Benli, 2004) and the Baleares ( Pons-Moyà & Pons, 2002). Although these records may have been based on extremely spiny morphotypes of other taxa of the O. edwardsii complex, or on mislabelled specimens, according to our analyses (ABGD and BEAST- GMYC), O. hispidula is included in a more widely distributed MOTU (clade D). Even in the less conservative eight-MOTUs scenario, MOTU-D probably represents a single species ranging from the northeastern Atlantic to the western and central Mediterranean Sea, which has recently colonized the Gulf of Gabès with a population displaying a strikingly deviating morphology ( Fig. 4 ), a phenomenon well known in the Gabès molluscan fauna ( Cecalupo et al., 2008), and with a distribution compatible with that of other intertidal molluscs ( Calvo et al., 2009). Furthermore, the young age of the Gulf of Gabès ( Burollet, Clairefond & Winnock, 1979; Stocchi, Colleoni & Spada, 2009) would not be compatible with the recent evolution of an endemic species. Ocinebrina erronea Cecalupo, Buzzurro & Mariani, 2008 (see Table 2 for nomenclatural details) has been synonymized recently with O. hispidula by Cossignani & Ardovini (2011). A single specimen of the O. erronea morphotype (sample 1053) was included in our dataset and, in all the analyses, it fell within the O. hispidula clade (group D1), indicating that it is not a distinct species.

Ocinebrina piantonii and O. carmelae were described as new species from the Kerkennah Islands by Cecalupo et al. (2008), living sympatrically (and syntopically) amidst the rhizomes of Posidonia oceanica (L.) Delile, 1813. The morphological differences between the two nominal taxa are indeed very subtle, possibly ending up as just two colour forms: a white phenotype ( O. piantonii ) and a dark phenotype ( O. carmelae ). Furthermore, the latter is indistinguishable from O. hybrida ( Aradas & Benoît, 1876) (which is possibly a synonym of O. baetica ( Reeve, 1845) : see Table 2), a rare species for which we had no specimens for DNA extraction. We assayed two specimens from the Gulf of Gabès, representing the two morphotypes, which proved clearly to belong to a single species (clade A), not related phylogenetically to O. hispidula . They fall into an eastern Mediterranean group including clades B (Aegean Sea) and C (Ionian Sea).

Specimens collected at four localities across the Aegean Sea (clade B) were indistinguishable morphologically from those classified preliminarily as O. edwardsii from other Mediterranean localities. The COI and 16S sequences of these specimens were particularly divergent from all the others, allowing the identification of a well-distinct mitochondrial lineage. Furthermore, the mitochondrial data also suggest a well-defined population structure within the group, with three deeply separated clusters corresponding to distinct sampling localities (northern, central and southern Aegean Sea) ( Fig. 4 ).

Another clade (C) is represented by the specimens collected in the Ionian Sea, southwestern Sicily and Lampedusa Island. Similar to group B, clade C is well defined geographically ( Fig. 4 ), but the shells are rather indistinguishable from specimens ascribed morphologically to O. edwardsii from other Mediterranean localities. The sister group relationship with the Aegean clade suggests an allopatric origin of both lineages from an ancestral eastern Mediterranean clade (see below).

Ocinebrina ingloria was described originally by Crosse (1865) without specifying any locality. Other names may apply to this entity (e.g. Murex orcomenus Nardo, 1847 , see Table 2), but the assessment of the status was beyond the scope of this study. Neglected for over a century, it was recorded as O. ingloria recently from the Adriatic Sea, from Grado and the Venice Lagoon (which was interpreted as the most likely locality of the type material) by Houart (2001) and from Palermo by Giannuzzi-Savelli et al. (2003). The population at Grado seems to be extinct, together with other intertidal molluscan species, as a result of the intense human activity in the area (see Crocetta, 2011), whereas the record from Palermo seems likely to be incorrect because such a morphotype has never been found there (G. Bonomolo, unpublished). We have assayed specimens from the Venice Lagoon (morphologically and geographically representing the nominal taxon O. ingloria ) which, despite evident morphological differences, were grouped in clade H with other specimens classified morphologically as O. edwardsii from the Adriatic Sea and a few localities in the Ionian Sea ( Fig. 4 ). Apparently all specimens of the O. edwardsii complex from the Adriatic Sea belong in the same lineage as O. ingloria , again suggesting that shell variation, as currently approached within this group, is rather misleading, and that a single species is involved in this lineage.

The sole geographical exception to clade H is represented by some specimens sampled at Pellestrina (clade F), within the area of the Venice Lagoon. All specimens from this site were indistinguishable from each other morphologically (see, for example, vouchers 1046_1 and 1046_3: Fig 3B , Supporting Information Fig. S4 ). After the first analyses split the first lot of specimens from this locality (sample 1046) into groups F and H, we collected other specimens from the same locality (sample 1293) in order to rule out the possibility of contamination, and obtained the same results. Thus, according to this evidence, two morphologically indistinguishable MOTUs live sympatrically at this site in the Venice Lagoon. The sister clade to MOTU-F from Pellestrina is clade G from Pozzuoli and Palinuro (central Tyrrhenian Sea). Clades F and G were clearly distinct MOTUs in all of our species delimitation analyses, and both lineages possibly represent relict populations isolated on the two sides of the Italian peninsula.

Figures

Figure 3. Consensus tree of a final sample of 104 trees obtained by BEAST on the cytochrome oxidase subunit I (COI) alignment. The black bars on the right delimit the Molecular Operational Taxonolllmic Units (MOTUs) as defined by the three species delimitation methods employed [Automatic Barcode Gap Discovery (ABGD), Generalized Mixed Yule- Coalescent (GMYC) and TCS]. Posterior supports> 0.95 are reported, and only for the nodes subtending the MOTUs. Voucher shells for each MOTU are figured (shells not to scale). 3A, MOTUs A–C; 3B, MOTUs D1–H.

Figure 4. Same tree as in Figure 3 obtained by BEAST on the cytochrome oxidase subunit I (COI) alignment, and distribution maps of the vouchers, grouped in the clades resulting from the COI analysis.