Organisms Diversity & Evolution

, Volume 12, Issue 4, pp 349–375

Systematics and phylogenetic species delimitation within Polinices s.l. (Caenogastropoda: Naticidae) based on molecular data and shell morphology

Authors

    • School of Biological SciencesThe University of Queensland, St. Lucia
    • Zoologisches Forschungsmuseum Alexander Koenig
  • Daniel Tapken
    • Department of Biochemistry I - Receptor Biochemistry, Faculty of Chemistry and BiochemistryRuhr University Bochum
  • Tim Dahlmann
    • Department of Biochemistry I - Receptor Biochemistry, Faculty of Chemistry and BiochemistryRuhr University Bochum
  • Heike Wägele
    • Zoologisches Forschungsmuseum Alexander Koenig
  • Cynthia Riginos
    • School of Biological SciencesThe University of Queensland, St. Lucia
  • Michael Hollmann
    • Department of Biochemistry I - Receptor Biochemistry, Faculty of Chemistry and BiochemistryRuhr University Bochum
Original Article

DOI: 10.1007/s13127-012-0111-5

Cite this article as:
Huelsken, T., Tapken, D., Dahlmann, T. et al. Org Divers Evol (2012) 12: 349. doi:10.1007/s13127-012-0111-5

Abstract

Here, we present the first phylogenetic analysis of a group of species taxonomically assigned to Polinices sensu latu (Naticidae, Gastropoda) based on molecular data sets. Polinices s.l. represents a speciose group of the infaunal gastropod family Naticidae, including species that have often been assigned to subgenera of Polinices [e.g. P. (Neverita), P. (Euspira), P. (Conuber) and P. (Mammilla)] based on conchological data. The results of our molecular phylogenetic analysis confirm the validity of five genera, Conuber, Polinices, Mammilla, Euspira and Neverita, including four that have been used previously mainly as subgenera of Polinices s.l. Our results furthermore indicate a close relationship of members of the Polinicinae to Sinum—a genus traditionally placed in the naticid subfamily Sininae. We furthermore present conchological analyses to determine the validity of shell characters used traditionally in species designation in the genus Polinices. Our data reveal several characters (e.g. protoconch, operculum colour, parietal callus) to be informative, while many characters show a high degree of homoplasy (e.g. umbilicus, shell form). Among the species arranged in the genus Polinices s.s., four conchologically very similar taxa often subsumed under the common Indo-Pacific species P. mammilla are separated distinctly in phylogenetic analyses. Despite their striking conchological similarities, none of these four taxa are related directly to each other. Additional conchological analyses of available name-bearing type specimens and type figures reveal the four “mammilla”-like white Polinices species to include true P. mammilla and three additional species, which could be assigned to P. constanti (replacement name for P. dubius), P. jukesii and possibly P. tawhitirahia, based on protoconch and operculum characteristics.

Keywords

PolinicesMolecular systematicsConchologyPolinicinaeSininaeNaticinaeBarcoding

Introduction

Polinices sensu latu represents one of the most speciose genera within the infaunal caenogastropod family Naticidae, including species assigned taxonomically to Polinices sensu strictu (s.s.) (Polinices (Polinices)) or to species that were described as being members of subgeneric taxa of Polinices, such as P. (Neverita), P. (Euspira), P. (Conuber) and P. (Mammilla) (see Cernohorsky 1971; Marincovich 1977; Majima 1989). Members of Polinices s.l. are distributed widely, occurring predominantly in tropical waters of the Indo-Pacific region with only a few species living in the Atlantic Ocean, Mediterranean Sea, Caribbean Sea and the Eastern Pacific. Taxonomic assignment of Polinices s.l. species to the traditional subfamilial group Polinicinae is based on the presence of a corneous operculum (Marincovich 1977; Majima 1989; Kabat 1991).

The genus Polinices Montfort, 1810 is based on the description of the purely white-shelled type species Polinices albus Montfort, 1810 (type locality: Ambon Island, Indonesia, fide Kabat 1990; see Supplement). Objective synonymy of this species with Nerita mammilla Linnaeus, 1758 was assured by the action of Kabat (1990), who designated the lectotype of Nerita mammilla as the neotype of Polinices albus (see Linnaeus 1758). Thus, Kabat prevented the well-established and broadly used generic level taxon Polinices to be discarded, should its type species, Polinices albus, for which no type material could be located (Kabat 1990), be considered a nomen dubium. The genus Polinices erected by Montfort was used later as the type genus for the subfamilial group Polinicinae Gray, 1847 (Montfort 1810). As molecular, anatomical, biogeographical, or ecological data are difficult to obtain for this species group, the characters most commonly used for species differentiation within the Polinicinae are the size and colour of the operculum, the arrangement of the funicle within the umbilicus, the shell form and colouration as well as the size and colouration of the protoconch (e.g. Risso 1826; Agassiz 1837; Chenu 1842; Récluz 1844; Philippi 1849; Tryon 1886; Garrard 1961; Cernohorsky 1971; Marincovich 1977; Majima 1989; Bandel 1999; Kabat 2000).

Conchologically, species assigned to Polinices s.l. are very similar to each other and are generally characterized by plain white or monochrome, glossy, ovate to pyriform-shaped shells; a brownish, corneous operculum; a medium-to-thick parietal callus; and a partly or completely filled umbilicus (Fig. 1). Based on intra-specific variation of these features and striking inter-specific similarities of Recent Polinices s.l. species, a large number of species with questionable taxonomic status have been described to date (see Supplement).
https://static-content.springer.com/image/art%3A10.1007%2Fs13127-012-0111-5/MediaObjects/13127_2012_111_Fig1_HTML.gif
Fig. 1

Species analysed in this study. aPolinices sp.2 [#70-2, MNHN#IM-2009-5170]. bPolinices sp.3 [#70-6, QM#MO80747]. cPolinices flemingianus (Récluz, 1844) [#141-1, MNHN#42645]. dPolinices sp.4 [#D6, QM#MO80750]. ePolinices mellosus (Hedley, 1924) [#59-4, MNHN#IM-2009-5167]. fPolinices cumingianus (Récluz, 1844) [AMS#C434459]. gPolinices peselephanti (Link, 1807) [AMS#C451672]. hPolinices albumen (Linnaeus, 1758) [SBD#026719]. iPolinices mediopacificus Kosuge, 1979 [MNHN#42646]. jPolinices uber (Valenciennes in Humboldt, 1832) [#30-1, MNHN IM-2009-5172]. kPolinices sp.1 [#51-1, MNHN IM-2009-5174]. lConuber sordidus (Swainson, 1821) [AMS#EBU30442]. mConuber conicus (Lamarck, 1822) [#80-2, $$]. nConuber incei (Philippi, 1851) [#100-1, AMS#C399745]. oMammilla priamus (Récluz, 1844) [#07-1, MNHN IM-2009-5179]. pMammilla simiae (Deshayes in Deshayes & Edwards, 1838) [#77-1, MNHN IM-2009-5177]. qMammilla melanostomoides (Quoy & Gaimard, 1832) [#87-1, MNHN 42649]. rMammilla melanostoma (Gmelin, 1791) [#25-1, MNHN IM-2009-5176]. sMammilla caprae (Philippi, 1850) [#123-1, MNHN IM-2009-5178]. tSinum haliotoideum (Linnaeus, 1758) [#97-1, AMS#C451594]. uSinum sanctijohannis (Pilsbry & Lowe, 1932) [#35-1, MNHN#IM-2009-5162]. vEuspira lewisii (Gould, 1847) [#104-1, QM#MO80751]. Pictures of specimens analysed in earlier studies (Euspira, Neverita) can be found in Huelsken et al. (2006, 2008). Enlarged images of the protoconchs are shown in the small inserts. Bars 0.5 cm

The problem of highly similar conchological features used for species identification, however, is not restricted to Polinices s.l. species, but is found in different (sub)generic lineages within the entire family Naticidae (Bandel 1999; Huelsken et al. 2011b). Differences in the shape and extent of the parietal callus, shell shape, thickness of the funicle and size and form of the umbilical cavity, observed in (sub)generic taxa within the Polinicinae are often limited to the degree of character expression only. Consequently, Cernohorsky stated, that “…umbilical and opercular characters are not always in agreement nor do they follow a pre-diagnosed generic pattern” (1971: 169). This statement concurs with analyses of Troschel (1856–1863) and Bandel (1984) who regarded members of the subfamilial groups Naticinae and Polinicinae to be congeneric based on similarities in the morphology of their radulae. Popenoe et al. (1987), amongst others, stated in their compilation of the late Cretaceous subfamilial naticid taxon Gyrodinae that the convergent development and inconstant characteristics of umbilical features complicate the classification within the entire family Naticidae.

Not surprisingly, the generic classification within the Polinicinae has changed frequently during the last two centuries. Due to the lack of distinct and characteristic conchological features, members of the subfamilial taxon Polinicinae (Polinices, Conuber Finlay and Marwick 1937, Euspira Agassiz in Sowerby, 1837, Mammilla Schumacher, 1817, Neverita Risso, 1826) have been treated as subgenera of Polinices (e.g. Cernohorsky 1971; Marincovich 1977; Majima 1989) or have been considered to be closely related to the subfamilial taxon Sininae (e.g. the genus Mammilla; Cernohorsky 1971; Kabat 1996). These examples support scepticism in the application of conchological characters in cladistic analyses, because of their highly homoplasious nature (Kool 1993) caused by analogous adaptations to environmental constraints. This is particularly true for the infaunal Naticidae, all of which are burrowing species with a seemingly identical ecology and a predatory feeding behaviour that relies on drilling of the shells of their prey (Cernohorsky 1971; Huelsken 2011).

Yet, empty shells are often the only available information source with which to identify gastropod species. As scientific names are assigned formally to type specimens, which in gastropods most often are available only as empty shells, analysis of conchological characters is the boon and bane of taxonomic assignments: in most cases, species determination based on shells of type specimens allows reliable taxonomic assignment of recent and fossil species. In other cases, type lots are missing, shells are broken, the few available characters are homoplasious or available species descriptions are not informative enough for reliable species identification.

In the present study, we employ a multilocus molecular phylogenetic analysis to investigate the relationships within the genus Polinices s.s. and its association with the (sub)generic taxa Conuber, Euspira, Mammilla, Neverita and Sinum that have been regarded traditionally as closely related. Analyses of conchological characters of molecularly characterized specimens serve to estimate the validity of traditionally used characters and their usage in type specimen assignment to Polinices s.s. species. Additionally, we provide species names and taxonomic descriptions for several plain white Polinices s.s. species, which can be separated from P. mammilla by phylogenetic analyses.

Materials and methods

Throughout the manuscript, the term ‘Polinices’ is to be taken sensu stricto (s.s) unless mentioned otherwise. Polinices sensu latu (s.l.) refers to species assigned to Polinices s.s. and to species that have previously been assigned to subgenera of Polinices [e.g. P. (Mammilla), P. (Euspira)].

Material examined

Specimens (Fig. 1) analysed were collected by diving, snorkelling, and dredging from several spots around Lizard Island and Dingo Beach in Queensland, Australia or were on loan from the Australian Museum, Sydney, Australia (AMS), the Queensland Museum, Brisbane, Australia (QM) or the Muséum National d'Histoire Naturelle, Paris, France (MNHN) (Table 1). Additional specimens were obtained from 9 of 457 trawl samples taken during the Great Barrier Reef Seabed Biodiversity Project (see Pitcher et al. 2007). Collected specimens have been vouchered in the malacological collection at the Queensland Museum (QM) or have been taken from previous work (Hülsken et al. 2006; Huelsken et al. 2008). The criteria for an a priori definition of species and genera were based on previously published taxonomic descriptions (e.g. Röding 1798; Schumacher 1817; Swainson 1840; Récluz 1850; Philippi 1850; Garrard 1961; Cernohorsky 1971; Marincovich 1977; Majima 1989; Kabat 1991; Hülsken et al. 2006; Huelsken et al. 2008).
Table 1

Specimens analyzed in this study, with specimen numbers, collection sites and museum voucher numbers

Species, author

Collection site

Voucher number/ reference

Outgroup taxa

 Bostrycapulus pritzkeri (Collin, 2005)

Edwards Beach, Balmoral, New South Wales, Australia

Colgan et al. 2007; AMS#C335468

 Strombus luhuanus (Linnaeus, 1758)

Heron Island, Queensland, Australia

Colgan et al. 2007; AMS#C203214

 

DB-TH225 Dingo Beach, Whitsunday Islands, Queensland, Australia

QM#MO80745

 Cypraea annulus (Linnaeus, 1758)

Heron Island, Queensland, Australia

Colgan et al. 2007; AMS#C203215

 Pyrazus ebeninus (Bruguiere, 1792)

SI-TH116 on muddy sand at low tide, Dunwich, North Stradbroke Island, Australia

QM#MO80742

 Oliva amethystina (Röding, 1798)

DB-TH158 on muddy sand at low tide, Dingo Beach, Whitsunday Islands, Queensland, Australia

QM#MO80743

 

DB-TH159 on muddy sand at low tide, Dingo Beach, Whitsunday Islands, Queensland, Australia

QM#MO80744

 Strombus dilatatus (Swainson, 1821)

Seabed material, GPS −24.386282 152.593168, 25.0 m depth, Australia, SBD#035346

QM#MO80144

 

Seabed material, GPS −24.127022 152.212747, 34.7 m depth, Australia, SBD#026638

QM#MO80145

 

Seabed material, GPS −24.467679 152.969458, 27.1 m depth, Australia, SBD#026756

QM#MO80146

 

Seabed material, GPS −24.127022 152.212747, 34.9 m depth, Australia, SBD#026638

QM#MO80147

 

Seabed material, GPS −20.935000 150.455000, 39.9 m depth, Australia, SBD#023154

QM#MO80148

 

Seabed material, GPS −24.386282 152.593168, 25.0 m depth, Australia, SBD#035346

QM#MO80149

 

Seabed material, GPS −24.066716 152.134289, 31.9 m depth, Australia, SBD#035215

QM#MO80150

 

Seabed material, GPS −21.089767 150.936033, 31.9 m depth, Australia, SBD#026848

QM#MO80151

Subfamily Naticinae, Guilding 1834

  

GenusTectonaticaSacco, 1890

  

 Tectonatica sagraiana (Orbigny, 1842)

#47-1, #47-2, #47-6, #47-8, #47-9 Campese Bay, Isola del Giglio, 7–10 m depth

Huelsken et al. 2008

 Tectonatica cf. rizzae (Philippi, 1844)

C82, C126, C127, C131, egg masses, Pt. delle Secche, Isola del Giglio, 18 m depth

Huelsken et al. 2008

Subfamily Sininae Wenz, 1941

GenusSinumRöding, 1798

  

 Sinum haliotoideum (Linnaeus, 1758)

Swan Island, Shoalwater Bay, Great Barrier Reef

AMS#C451594

 Sinum sanctijohannis (Pilsbry & Lowe, 1932)

#35-1 Baja California, Mexico

MNHN#IM-2009-5162

Subfamily Polinicinae Gray, 1847

  

GenusEuspiraAgassiz in J. Sowerby, 1837

  

 Euspira nitida (Donovan, 1804)

#114-5 Cala dell´Allume, Isola del Giglio, 9 m depth

MNHN#IM-2009-5163

 Euspira catena (Da Costa, 1778)

#127-2 Island of Terschelling, Netherlands

MNHN#IM-2009-5164

 Euspira fusca (de Blainville, 1825)

#126-1 off Olhão, Portugal, dredged in 380–400 m depth

MNHN#IM-2009-5165

 Euspira intricata (Donovan, 1804)

#120-1 Cala dell´Allume, Isola del Giglio, 6 m depth

MNHN#IM-2009-5166

GenusPolinicesMontfort, 1810

  

 Polinices albumen (Linnaeus, 1758)

North of Fraser Island, -24.485999 153.100564, 29.3 m depth, SBD#026719

QM#MO80152

 Polinices cumingianus (Récluz, 1844)

#96-1 Fraser Island, Great Barrier Reef, Australia

AMS C434459

 

#96-2 South Passage, Shark Bay, Western Australia, Australia

MNHN#IM-2009-5169

 Polinices mellosus (Hedley, 1924)

#59-3 Casuarina Beach, Lizard Island, Australia, 2–3 m depth

MNHN#IM-2009-5167

 

#59-4 Casuarina Beach, Lizard Island, Australia, 2–3 m depth

MNHN#IM-2009-5168

 Polinices flemingianus (Récluz, 1844)

#141-1 Vanuatu: Santo Island, Aore Island

MNHN#42645

 Polinices mediopacificus (Kosuge, 1979)

Bohol Sea, off Palimacan Island, Philippines

MNHN#42646

 Polinices peselephanti (Link, 1807)

#99-1, #99-2, #99-3 Marble Island, Duke Group, Great Barrier Reef, Australia

AMS#C451672

 Polinices sp. 1

#51-4 (IM-2009-5174), #51-5 (IM-2009-5175), #51-7, #51-10, #51-11, #51-13, #51-14, #51-15, #51-16, #51-17, #51-18, #51-19, #51-20, #51-21, #51-22, #51-64, #70-18 Casuarina Beach, Lizard Island, Australia, water depth 2–3 m

MNHN#IM-2009-5174; MNHN#IM-2009-5175

 

#DB-TH143, #DB-TH144, #DB-TH145, Dingo Beach, Whitsunday Islands, Australia, water depth 0–1 m

QM#MO80740

 

#VM2.1-#VM2.5, Santo Island, Vanuatu

MNHN#42647

 

#M5.1 , 9°35.5´N 123°43.3´E/123°44.3´E Panglao Island, Doljo Point, Philippines, 0–2 m soft mixed intertidal platform, fringe mangrove, seagrass

MNHN#M8.1

 

#51-24 Nabq National Park, Sinai, Egypt, water depth 0–3 m, coral sand

QM#MO80741

 Polinices sp. 2

#70-1 (IM-2009-5170), #70-2 (IM-2009-5171), #70-3, #70-9, #70-14, #70-16, #70-17 Casuarina Beach, Lizard Island, Australia, water depth 2–3 m

MNHN#IM-2009-5170; MNHN#IM-2009-5171

 

#70-8 Latalata Island, West of Halmahera, Indonesia

QM#MO80746

 Polinices sp. 3

#70-6, #70-10 , #70-12 on coral sand, water depth 2–3 m, Casuarina Beach, Lizard Island, Australia

QM#MO80747–QM#MO80749

 Polinices sp. 4

#D1, #D3, #D4, #D5, #D6 (QM#MO80750) Between Town Beach and Entrance Point, Broome, WA, Australia, on muddy sand close to shoreline

QM#MO80750

 Polinices uber (Valenciennes in Humboldt, 1832)

#30-1 Cholla Bay, Puerto Penasco, Sonora, Mexico

MNHN#IM-2009-5172

 

#30-3 Cholla Bay, Puerto Penasco, Sonora, Mexico

MNHN#IM-2009-5173

GenusMammillaSchumacher, 1817

  

 Mammilla caprae (Philippi, 1850)

#123-1 Cedros Island, Costa Rica, dredged at 400 m depth

MNHN#IM-2009-5178

 Mammilla melanostoma (Gmelin, 1791)

#25-4 Casuarina Beach, Lizard Island, Australia, water depth 1 m

MNHN#IM-2009-5176

 Mammilla melanostomoides (Quoy & Gaimard, 1832)

#87-2 Santo Island, Segond Channel, Vanuatu

MNHN#42649

 Mammilla priamus (Récluz, 1844)

#07-1 Philippines

MNHN#IM-2009-5179

 Mammilla simiae (Deshayes in Deshayes & Edwards, 1838)

#77-1 Bernier Iland, Western Australia, Australia

MNHN#IM-2009-5177

GenusNeveritaRisso, 1826

  

 Neverita aulacoglossa (Pilsbry & Vanatta, 1908)

#95-1, #95-2, #95-3 QLD, Australia, Great Barrier Reef, Fraser Island, offshore of Sandy Point, S of Mooan 25° 19.945´S, 153° 0.524´ E. On & in sand, LT, sandbank

AMS#C412600

 Neverita delessertiana (Récluz in Chénu, 1843)

#19-1, #19-3, #19-5, #19-9 Clearwater, Florida

Hülsken et al. 2006

 

#19-11 Cedar Key, Florida

Hülsken et al. 2006

 Neverita didyma (Röding, 1798)

Taiwan

Collin 2003

 Neverita duplicata (Say, 1822)

#21-2, #21-3, #21-4 Intertidal ocean, Little Talbot Island, Jacksonville, Florida

Hülsken et al. 2006

 Neverita lewisii (Gould, 1847)

#104-1 Bainbridge Island, Tecoma, Pudget Sound, Washington, USA, under rocks and pebbles

QM#MO80751

 Neverita josephinia (Risso, 1826)

#46-4, #46-6, #46-16 Giglio Campese, Campese Bay, depth: ca. 10 m, on sand ground, Isola del Giglio, Toscana, Italy, water depth 10–12 m

Huelsken et al. 2008

 Neverita reclusiana (Deshayes, 1839)

#33-1 Cholla Bay, Puerto Penasco, Mexico

Huelsken et al. 2008

GenusConuberFinlay & Marwick, 1837

  

 Conuber conicus (Lamarck, 1822)

Great Sandy Strait, SE of Urangan 25°19.980´S, 152°55.933`E; in & on muddy sand, on sandflat, QLD, Australia

AMS#C412917

 Conuber incei (Philippi, 1853)

Crowdy Head, S beach 31°52.290´S 152°42.250´E, NSW, Australia, intertidal sand, surfing beach

AMS#C399745

 

Lennox Head Beach, access via Pacific Pde/Ross St. junction 28°47.200´S 153°35.600´E, 0–0.5 m sand in surf zone

AMS#C414237

 Conuber sordidus (Swainson, 1821)

Dingo Beach, Queensland, Australia, crawling on muddy sand at low tide

QM#MO80752

 

Careel Bay, Pittwater, NSW

AMS# EBU30442; Colgan et al. 2006

Collected animals were anaesthetised with 0.25 M MgCl2, fixed in 75–85 % EtOH and subsequently stored in 94 % EtOH. Altogether, our data set is based on 87 specimens representing 32 naticid species in eight traditional (sub)genera from three traditional subfamilies, the Polinicinae (Polinices, Conuber, Neverita, Mammilla, Euspira, Payraudeautia), Sininae (Sinum) and Naticinae (Tectonatica) (Table 1). The genus Tectonatica was chosen as an internal outgroup to root the ingroup and to test the relationship of Polinicinae and Sininae.

We additionally selected several members of the caenogastropod families Strombidae [Strombus dilatatus (Swainson, 1821), Strombus luhuanus (Linnaeus, 1758)], Batillariidae [Pyrazus ebeninus (Bruguiere, 1792)], Calyptraeidae [Bostrycapulus pritzkeri (Collin, 2005)], Olividae [Oliva amethystina (Roeding, 1798)] and Cypraeidae [Cypraea annulus (Linnaeus, 1758)] for outgroup comparison (see Table 1).

Nucleic acid isolation and sequence analysis

Total DNA was extracted from ethanol/RNAlater (Qiagen, Hilden, Germany)-preserved tissue using a modified protocol of the DNeasy Extraction Kit (Qiagen) (Huelsken et al. 2011a) and stored in 0.1 mM Tris-EDTA pH 7.4. A 447-bp fragment of the COI gene, 264 bp of the H3 gene, 476 bp of the 16S gene, 401 bp of the 18S gene and 352 bp of the 28S gene were amplified and sequenced from each specimen. Amplification reactions were performed with iProof polymerase (Bio-Rad Laboratories, Munich, Germany) on MJ Research thermocyclers (MJ Research, Watertown, MA). Amplification primers were P388 (5´-gcttttgttataattttytt-3´) and P390 (5´-cgatcagttaaaartatwgtaat-3´) for COI, P263 (5´-cctcatcgttacaggcccgg-3´) and P266 (5´-actggatgtccttgggcatg-3´) for H3, P213 (5´-cgcctgttaccaaaaacat-3´) and P214 (5´- ccggtctgaactcagatcacgt-3´) for 16S, P398 (5´-cgtgttgatyctgccagt-3´) and P399 (5´- tctcaggctccytctccgg-3´) for the partial 18S gene and P1017 (5´-acccsctgaayttaagcat-3´) and P1018 (5´- aactctctcmttcaragttc-3´) for the partial 28S gene fragment [primer sequences taken from Colgan et al. 2007 (3´end of 28S rRNA), Huelsken et al. 2008, 2011a].

PCR products were purified using the JETSORB Gel Extraction Kit (Genomed, Löhne, Germany) and both strands were sequenced on an ABI 3130xl automated sequencer using the PCR primers and a BigDye Terminator v3.1 sequencing kit (both Applied Biosystems, Foster City, CA). Sequences of Neverita, Euspira, Tectonatica and Payraudeautia species had been obtained by us in the context of other studies (Hülsken et al. 2006; Huelsken et al. 2008). Pictures of these species can be found in the respective publications or under the Morphobank project ID#189 (O'Leary and Kaufman 2007).

Phylogenetic analyses

The phylogenetic trees (Figs. 2 and 3, Table 2) were calculated with MrBayes v3.1.2 (Ronquist and Huelsenbeck 2003), while the NeighborNet analysis (Fig. 4) was performed using SplitsTree v4.0 under the LogDet model (Huson 1998; Huson and Bryant 2006). Sequence distances have been demonstrated to represent evolutionary distances between species (Makowsky et al. 2010). Thus, genetic distances between taxa and clades were calculated using PAUP*4.0b10 (Swofford 2003) (Tables 3 and 4). Sequences of all specimens analysed were uploaded to GenBank (accession numbers FJ263801–FJ263889, GQ328724–GQ328743 and FJ623464–FJ623465) and the concatenated alignment was deposited in TreeBASE (Sanderson et al. 1994).
https://static-content.springer.com/image/art%3A10.1007%2Fs13127-012-0111-5/MediaObjects/13127_2012_111_Fig2_HTML.gif
Fig. 2

Phylogram obtained through Bayesian inference based on the concatenated data set (COI, 16S, 18S, 28S, H3) for a reduced number of taxa. Posterior probabilities are indicated at the nodes. Branches supported by values > 0.95 are indicated in bold. Polytomies are due to the cut-off value specified for the consensus tree (50 % used as the default value in MrBayes)

https://static-content.springer.com/image/art%3A10.1007%2Fs13127-012-0111-5/MediaObjects/13127_2012_111_Fig3_HTML.gif
Fig. 3

Phylogram obtained through Bayesian inference based on the COI gene fragment. Posterior probabilities are indicated at the nodes. Branches supported by values > 0.95 are indicated in bold. Polytomies are due to the cut-off value specified for the consensus tree (50 % used as the default value in MrBayes)

Table 2

Arrangement of naticid taxa in the phylogenetic trees derived from different gene fragments. Bold taxa are supported by posterior probability > 0.95, capitalized taxa are arranged para- or polyphyletically

Gene fragment

Taxa arrangement

mtCOI

(Sinum(Neverita(Tectonatica(Euspira(Conuber(Mammilla, Polinices))))))

mt16S rRNA

(Mammilla(Sinum, Tectonatica, Euspira, Conuber, Neverita)(Polinices))

ncH3

(Outgroup, Neverita(Tectonatica, Conuber)(Euspira, Conuber)Mammilla, Polinices)

nc28S rRNA

(Mammilla, Polinices, Conuber)(Sinum, Tectonatica, Euspira, Neverita)

nc18S rRNA

(Sinum(Tectonatica(Conuber(Neverita(Mammilla, Polinices, Euspira)))))

ALL

(Tectonatica(Conuber, Neverita(Sinum(Euspira(Mammilla(Polinices))))))

https://static-content.springer.com/image/art%3A10.1007%2Fs13127-012-0111-5/MediaObjects/13127_2012_111_Fig4_HTML.gif
Fig. 4

NeighborNet network based on the concatenated data set (COI, 16S, 18S, 28S, H3). Bootstrap values are indicated

Table 3

Inter- (normal) and intra-specific (bold) genetic distances calculated for the analysed Polinices species based on the COI gene fragment (average±standard deviation)

 

P. alb (n = 1)

P. cum (n = 3)

P. flem (n = 1)

P. sp. 1 (n = 30)

P. med (n = 1)

P. mel (n = 5)

P. sp. 2 (n = 9)

P. sp. 3 (n = 3)

P. sp. 4 (n =5)

P. pes (n = 3)

P. ube (n = 2)

P. albumen (n = 1)

--

          

P. cumingianus (n = 3)

0.083±0.000

0.003±0.003

         

P. flemingianus (n = 1)

0.071±0.000

0.085±0.000

--

        

P. sp. 1 (n = 30)

0.111±0.009

0.101±0.008

0.106±0.009

0.089±0.002

       

P. mediopacificus (n = 1)

0.089±--

0.081±0.000

0.087±--

0.081±0.004

--

      

P. mellosus (n = 5)

0.065±0.002

0.074±0.002

0.064±0.001

0.095±0.010

0.085±0.004

0.003±0.002

     

P. sp. 2 (n = 9)

0.082±0.002

0.062±0.004

0.075±0.003

0.089±0.012

0.075±0.002

0.064±0.002

0.005±0.006

    

P. sp. 3 (n = 3)

0.064±0.006

0.071±0.005

0.059±0.004

0.092±0.010

0.075±0.004

0.058±0.005

0.058±0.005

0.013±0.003

   

P. sp. 4 (n = 5)

0.075±0.001

0.082±0.001

0.073±0.001

0.111±0.009

0.098±0.001

0.071±0.002

0.077±0.003

0.062±0.005

0.001±0.001

  

P. peselephanti (n = 3)

0.071±0.001

0.0071±0.001

0.076±0.005

0.093±0.011

0.062±0.003

0.066±0.002

0.067±0.003

0.055±0.005

0.079±0.002

0.001±0.002

 

P. uber (n = 2)

0.087±0.000

0.090±0.001

0.092±0.000

0.081±0.009

0.062±0.002

0.078±0.001

0.074±0.001

0.071±0.004

0.099±0.001

0.070±0.003

0.009±−−

Table 4

Inter- (normal) and intra-clade (bold) genetic distances calculated for clades 1–3 of P. sp. 1 based on the COI gene fragment (average genetic distance±standard deviation)

 

Clade 1

Clade 2

Clade 3

P. sp. 1 [clade 1]

--

  

P. sp. 1 [clade 2]

0.089±0.01

0.077±0.01

 

P. sp. 1 [clade 3]

0.086±0.01

0.091±0.01

0.010±0.02

For the Bayesian analysis, 15 x 106 generations were calculated saving every 1,000th tree. The first 3,000 trees were discarded as burn-in. In the phylogenetic analyses (single gene fragment; concatenated data set) protein-coding gene fragments (COI, H3) were analysed using the model NY98 implemented in MrBayes (triplet code: metmt) to consider differences in omega variation across sites (neutral/purifying/positive selection of positions). Ribosomal gene fragments were calculated using a model predicted by MrModeltest (Nylander 2004). The GTR+G+I model was used for the 16S gene fragment, while the HKY85 model was used for the 18S and 28S gene fragments. Ambiguously aligned parts of the rRNA sequences were excluded from the analysis. All gene fragments were analysed as being unlinked. Phylogenetic analyses were performed separately for each gene fragment as well as for a combined data set (Figs. 2, 3 and 4, see Supplementary Figs. S1, S2). Bootstrap analysis of the molecular data set in the network analysis using SplitsTree v4.0 was performed with 1,000 replicates using the LogDet model.

Conchological analyses and studies of type material

In order to identify the validity of shell characters used in species identification and generic classifications of Polinices species, we analysed 27 conchological and one developmental character (Table 5, Supplementary Table S1). The characters were chosen based on their usage in species descriptions and the fact that they had been proposed to vary between the analysed species (see Tryon 1886; Murray 1966; Cernohorsky 1971; Marincovich 1977; Majima 1989; Kabat 1996; Aronowsky 2003). The character states were coded binarily and plotted on the phylogenetic tree (concatenated data set, Fig. 2) to calculate the consistency index (CI) and the retention index (RI) for each character using MacClade v4.06 (Maddison and Maddison 2006). All characters were equally weighted (see Supplementary Table S1).
Table 5

Results of the conchological analyses performed with MacClade v4.06 for the entire data set (Polinices s.l.) and for the reduced data set (Polinices s.s.) listed for each character. Type Type of coding (o ordered; u, unordered), States number of morphological states; Steps total number of steps in the phylogenetic tree, CI consistency index, RI retention index, EW embryonal whorls, FEW first embryonal whorl

   

Polinices s.l.

Polinices s.s.

 

Character

Type

States

Steps

CI

RI

States

Steps

CI

RI

1

Protoconch color

u

2

6

0.17

0.81

2

1

1.00

1.00

2

No. of EW

o

8

24

0.28

0.81

7

9

0.78

0.97

3

Size of FEW

o

8

35

0.21

0.72

8

20

0.39

0.83

4

Shell color

u

3

9

0.22

0.56

3

4

0.50

0.50

5

Color pattern

u

2

6

0.17

0.58

2

2

0.50

0.00

6

Shell shape (ratio height to width)

o

4

12

0.25

0.53

3

6

0.33

0.00

7

Aperture height ratio

o

5

11

0.36

0.84

2

2

0.50

0.00

8

Suture

u

2

4

0.25

0.67

2

0

0.00

0.00

9

Subsutural wrinkels

u

2

1

1.00

1.00

1

0

0.00

0.00

10

Umbilicus

u

3

16

0.12

0.48

3

8

0.22

0.42

11

Sulcus

u

2

3

0.33

0.00

2

2

0.50

0.00

12

Funicle

u

2

4

0.25

0.57

1

0

0.00

0.00

13

Umbilical callus

u

2

1

1.00

1.00

1

0

0.00

0.00

14

Inner lip

u

2

5

0.20

0.69

1

0

0.00

0.00

15

Operculum surface

u

2

1

1.00

1.00

1

0

0.00

0.00

16

Operculum color

u

4

4

0.50

0.71

2

1

1.00

1.00

17

Operculum size

u

3

2

1.00

1.00

1

0

0.00

0.00

18

Shell solidity

u

2

4

0.25

0.25

1

0

0.00

0.00

19

Shell texture

u

2

3

0.33

0.33

1

0

0.00

0.00

20

Aperture size

u

2

4

0.25

0.75

1

0

0.00

0.00

21

Aperture shape

u

3

7

0.29

0.58

1

0

0.00

0.00

22

Parietal callus

u

2

1

1.00

1.00

1

0

0.00

0.00

23

Parietal callus thickness

u

2

4

0.25

0.63

1

0

0.00

0.00

24

Posterior apertural angle

u

2

5

0.20

0.73

1

0

0.00

0.00

25

Parietal callus merge

u

3

5

0.40

0.81

1

0

0.00

0.00

26

Parietal callus size

u

2

5

0.33

0.78

1

0

0.00

0.00

27

Whorl expansion

u

2

5

0.20

0.56

2

2

0.50

0.00

28

Egg mass structure

u

2

1

1.00

1.00

1

0

0.00

0.00

In order to address quantitative variations of similar trait values within species and between closely related species, continuous characters (e.g. protoconch size, ratio of height to width, number of embryonic whorls) were coded as ‘ordered’ while all remaining characters were coded ‘unordered’ (see Wiens 2001). The analysis was performed for species taxonomically assigned to Polinices s.s. as well as for the entire data set (Polinices s.l.).

In the course of our molecular phylogenetic analyses, we came across a number of white Polinices species, which differed from P. mammilla but for which an unequivocal taxonomic assignment was difficult. We analysed a subset of 16 conchological key characters (characters A–P) for each of these species and compared those with characters determined by investigating existing type specimens in museum collections (Tables 6, 7). Unfortunately, many type specimens of white Polinices species are missing. In these cases, type figures or type descriptions were used to extract available information on shell features.
Table 6

Morphological shell characters and their ranges, of Polinices molecularly analysed in this study. Numbers in brackets refer to character numbers used in the prior conchological analysis (see Table 5). A Protoconch colour: W, white; B, black; Br, brown. B Number of protoconch whorls. C Size of first whorl of protoconch; in μm. D Shell colour: W, white; Y, cream-yellowish; W(+C), white background, occasionally with faint brownish pattern; B, distinct brownish colour pattern. E Shell shape: ratio of shell height to shell width; F Shell solidity: M, thick and massive. G Operculum: C, corneous. H Operculum size: A, same size as aperture. I Operculum colour: B, black; H, honey-coloured; HB, honey-coloured with black streak; B, black. J Aperture size: 1, < 60 % of shell height; 2, 60-70 % of shell height; 3, 70-80 % of shell height. K Funicle: 1, funcicle prominent; 0, funicle not discernible. L Umbilical structure: 0, umbilicus open; 1, umbilicus fully closed, 2, umbilicus partly closed, leaving a distinct opening anteriorly. M Columellar shape: S, straight columella. N Transition of parietal callus to umbilical callus: 1, same width; 2, narrowing. O Parietal callus: T, thick. P Distribution: IP Indo-Pacific, EP Eastern Pacific

Species

Characters

A (1)

B (2)

C (3)

D (4)

E (6)

F (18)

G (15)

H (17)

I (16)

J (20)

K (12)

L (10)

M (14)

N (25)

O (23)

P

P. albumen

W

1.50–1.75

570±30

O

0.71

M

C

A

H

0.70

1

1

S

1

T

IP

P. cumingianus

W

1.75

750±70

W+C

1.04±0.09

M

C

A

H

0.59±0.02

1

0

S

2

T

IP

P. mellosus

W

1.25–1.45

690±20

Y

1.20±0.06

M

C

A

H

0.59±0.08

0

1

S

1

T

IP

P. flemingianus

W

1.20

640

W

1.12

M

C

A

Hb

0.55

0

1/2

S

1

T

IP

P. mediopacificus

Br

2.50

560

W

1.02

M

C

A

H

0.65

0

0

S

1

T

IP

P. sp. 1

B

2.00–2.25

370±67

W

1.29±0.07

M

C

A

H

0.67±0.07

0

0.7

S

1

T

IP

P. sp. 2

W

1.25–1.50

770±60

W

1.27±0.1

M

C

A

H

0.71±0.06

0

1/2

S

1

T

IP

P. sp. 3

W

1.25–1.50

660±60

W

1.28±0.02

M

C

A

H

0.67±0.04

0

1

S

1

T

IP

P. sp. 4

W

0.90–1.15

870±70

W

1.09±0.03

M

C

A

B

0.82±0.01

0

1/2

S

1

T

IP

P. uber

Br

2.35–2.75

720±10.0

W

1.26±0.03

M

C

A

H

0.68±0.04

0

2

S

1

T

EP

P. peselephanti

W

1.75

1,250

W+C

1.09±0.02

M

C

A

H

0.60

1

0

S

2

T

IP

Table 7

Morphological shell characters and their range distribution, of P. sp. 1 through P. sp. 4 as well as possible name-bearing type specimens. Numbers in brackets refer to character numbers used in the preceding conchological analysis (see Table 5). A Protoconch colour: W, white; B, black; Br, brown. B Number of protoconch whorls. C Size of first whorl of protoconch; in μm. D Shell colour: W, white; Y, cream-yellowish; W(+C), white background, occasionally with faint brownish pattern; B, distinct brownish colour pattern. E Shell shape: ratio of shell height to shell width; F Shell solidity: M, thick and massive. G Operculum: C, corneous. H Operculum size: A, same size as aperture. I Operculum colour: B, black; H, honey-coloured; HB, honey-coloured with black streak; B, black. J Aperture size: 1, < 60 % of shell height; 2, 60-70 % of shell height; 3, 70-80 % of shell height. K Funicle: 1, funcicle prominent; 0, funicle not discernible. L Umbilical structure: 0, umbilicus open; 1, umbilicus fully closed, 2, umbilicus partly closed, leaving a distinct opening anteriorly. M Columellar shape: S, straight columella. N Transition of parietal callus to umbilical callus: 1, same width; 2, narrowing. O Parietal callus: T, thick. P Distribution: IP Indo-Pacific; EP Eastern Pacific

Species

Characters

A (1)

B (2)

C (3)

D (4)

E (6)

F (18)

G (15)

H (17)

I (16)

J (20)

K (12)

L (10)

M (14)

N (25)

O (23)

P

Polinices sp. 1

B

2.00–2.75

370±67.0a

W

1.29±0.07a

M

C

A

H

0.67±0.07a

0

1/2

S

1

T

IP

Polinices sp. 2

W

1.25–1.50

770±60.0a

W

1.27±0.1a

M

C

A

H

0.72±0.06a

0

1

S

1

T

IP

Polinices sp .3

W

1.25–1.50

660±60.0a

W

1.28±0.02a

M

C

A

H

0.65±0.03a

0

1/2

S

1

T

IP

Polinices sp. 4

W

0.90–1.15

870±70.0a

W

1.09±0.03a

M

C

A

B

0.82±0.01a

0

1

S

1

T

IP

Nerita mammilla Linnaeus, 1758

ind.

ind.e

W

1.26b

M

?

A

?

0.65b

0

1

S

1

T

?

Mammillaria tumida Swainson, 1840 [Mamma albula Chemnitz, 1781 [non-binomial]

ind.

ind.

W

1.28b

M

?

A

?

0.51b

0

1

S

1

T

?

Natica candidissima Le Guillou, 1842

?d

?

?

W

?

M

C

A

H

?

0

0

S

1

T

IP

Natica pyriformis Recluz, 1844

B

2.25–2.50a

410±10.0a

W

1.14±0.06a

M

?

A

?

0.61±0.06a

0

1

S

1

T

IP

Natica dubia Récluz, 1844

W

1.15–1.25a

656±97.0a

W

1.00±0.06a

M

?

?

?

0.59±0.02a

0

2

S

1

T

?

Natica cygnea Philippi, 1850

?

?

?

W

1.24c

M

?

?

?

0.61c

0

1

S

1

T

?

Natica virginea Philippi, 1850

?

?

?

W

1.21c

M

?

?

?

0.65c

0

2

S

1

T

?

Natica galactites Philippi, 1851

?

?

?

W

1.10c

M

?

?

?

0.68c

0

2

S

1

T

IP

Natica deiodosa Reeve, 1855

W

1.25–1.50a

740±40.0a

Y

1.03±0.14a

M

C

A

H

0.61±0.07a

0

1/2

S

1

T

IP

Natica jukesii Reeve, 1855

W

1.50–1.75a

790±50.0a

W

1.01±0.04a

M

C

A

H

0.58±0.05a

0

1/2

S

1

T

IP

Natica phytelephas Reeve 1855

W

2.00b

820A

W

1.01±0.03a

M

?

?

?

0.66±0.04a

0

0

S

1

T

IP

Natica vavaosi Reeve, 1855

?

?

?

W

1.16B

M

?

?

?

0.67B

0

2

S

1

T

IP

Polinices controversus Pritchard & Gatliff, 1913

W

1.75b

1.500b

W

0.90b

M

?

?

?

0.62b

1

2

S

2

T

IP

Polinices mellosus (Hedley, 1924)

W

1.25–1.50a

770±10.0a

Y

1.05±0.06a

M

C

A

B

0.58±0.09a

0

1/2

S

1

T

IP

Polinices putealis Garrard, 1961

Br

1.75b

320A

W

1.12±0.06a

M

?

?

?

0.59±0.03a

0

0

S

1

T

AU

Polinices tawhitirahia Powell, 1965

W

1.75b

622b

W

1.02b

M

C

A

B

0.66b

0

1/2

S

1

T

NZ

aAverage values for n±2 (sampled specimens of P. sp. 1–P. sp. 4; for types: holotype plus paratypes, or syntypes)

bValues from holotype only

cValues measured for figured type

dData unknown

eIndeterminate (i.e., broken shell)

Shell height (h), shell width (w) and aperture height were measured from vertically positioned shells or from drawn and pictured shells (apex up and basal lip down; see Fig. 1). Further data was compiled by analyses of the shell form, shell colour, protoconch morphology, umbilicus morphology and operculum colouration (Tables 6, 7; Supplementary Table S1). Protoconch morphology was measured according to Solsona and Martinell (1999). The size of the first embryonic whorl (FEW) and the number of embryonic whorls (EW) were measured using a digital binocular. The data matrix was uploaded to Morphobank project ID#189 (O'Leary and Kaufman 2007).

Results

Phylogenetic analyses

Partial sequences of two mitochondrial genes (COI, 16S) and three nuclear genes (28S, 18S, H3) were determined resulting in a concatenated alignment of 1,852 bp. In the phylograms, species were arranged into seven monophyletic groups, representing Conuber, Euspira, Mammilla, Neverita, Polinices, Sinum and Tectonatica (see Figs. 2, 3 and 4, Supplementary Figs. S1, S2). The assignment of the identified species to these monophyletic genera was similar in four single gene topologies (COI, 16S, 28S, 18S) and the combined analysis (Table 2). However, the relationship of generic clades, especially Mammilla and Euspira, differed in the various tree analyses (Figs. S1, S2). The gene H3 showed low resolution, not all genera were recognized and analysis resulted in a comb-like topology (Fig. S1).

In the analysis of the concatenated data set (Fig. 2), Tectonatica presents the most basal naticid taxon followed by a polytomy of Conuber, Neverita and a clade comprising Sinum, Euspira, Mammilla and Polinices. The sister taxa Mammilla and Polinices represent the most derived genera. The placement of Sinum as sister taxon to the clade Euspira/Mammilla/Polinices and even monophyly, is challenged in all the single gene analyses. Highest congruence with results obtained from the concatenated data set can be seen in the analysis of the COI gene (Table 2, Fig. 3).

The NeighborNet analysis of the concatenated gene fragments (Fig. 4) was congruent with the respective phylogenetic reconstructions. Polinices and Mammilla were well separated by distinct branches from all other monophyletic genera. There was a strong phylogenetic signal for the monophyly of Mammilla, but the taxon was nested within the Polinices species, rendering Polinices paraphyletic. Phylogenetic signal (recognizable in the long edges) for the monophyly of the genera Conuber, Euspira, Neverita, Sinum and Tectonatica was also high. However, a conflict was obvious in Sinum, represented here with only two species.

With few exceptions the grouping of species was identical in all phylogenetic reconstructions, placing species into supraspecific taxa according to their a priori taxonomic assignment (Figs. 2, 3 and 4, Supplementary Figs. S1, S2). However, our phylogenetic analyses revealed some unexpected species placements. First, individuals assigned to the Eastern Pacific species Euspira lewisii (Gould, 1847) grouped within the genus Neverita in the COI tree (Fig. 3). Second, our phylogenetic analyses revealed four conchologically similar, well-separated and highly supported plain-white species within Polinices (P. sp. 1 to P. sp. 4 in the following). P. sp. 1, P. sp. 2 and P. sp. 3 are very similar in shell structure and, at first glance, appear all to be referable to the common moon snail P. mammilla (Figs. 1, 5). They share a glossy, all-white shell, a closed-to-partly-open umbilicus and a honey-coloured, corneous operculum. The fourth all-white, glossy-shelled taxon (P. sp. 4) is distinguished only by an entirely black-coloured operculum.
https://static-content.springer.com/image/art%3A10.1007%2Fs13127-012-0111-5/MediaObjects/13127_2012_111_Fig5_HTML.gif
Fig. 5

Pictures of type specimens and protoconchs of aNerita mammilla Linnaeus, 1758 [ZMUU#386] b Mamma albula Chemnitz, 1758 [non-binomial, ZMUC] and cNatica pyriformis Recluz, 1844 [BMNH#1991089.1]. For further information see Table 1. Bars 0.5 cm

Haplotype analyses of the mitochondrial cytochrome oxidase (COI) and 16S gene fragments of 30 specimens of P. sp. 1 resulted in a strict separation of the specimens into three clades reflecting different localities (Fig. 3, Supplementary Fig. S1). The specimen in branch 1 was collected in Egypt (Nabq National Park, Sinai; #51–24), the specimens in clade 2 were collected in Indonesia (#M5.1), Lizard Island (Queensland, Australia; #70–18) and Vanuatu (VM2.1–2.5) and those in clade 3 at the Great Barrier Reef, Australia (Lizard Island, Whitsunday Islands).

Sequence distances of the mitochondrial COI gene fragments

The lowest genetic distance (uncorrected p-distance) was observed between species of Polinices and Mammilla (9.0 %±1.0). Other comparisons of species in distinct genera have genetic distances ranging from 13 %±2.0 to 16 %±2.0. Notably, these values did not reflect any subfamilial assignment of the species: intra- and inter-subfamilial distances are similar between genera of different subfamilies.

Within Polinices, species showed p-distances of 5.5 %±0.5 [P. peselephanti (Link, 1807) - P. sp. 3) to 11.1 %±0.9 (P. sp. 1 - P. albumen (Linnaeus, 1758)]. P. sp. 1 had the largest genetic distance to the remaining Polinices species. The species is closely related to P. mediopacificus (Kosuge, 1979) (8.1 %±0.4) and P. uber (Valenciennes in Humboldt, 1832) (8.1 %±0.9) and had p-distances ranging from 8.9 % to 11.1 % to remaining Polinices species (Table 3). Intra-specifically, specimens of P. sp. 1 differed in 8.6 %±1.0 to 9.1 %±1.0 genetic distance (Table 4). Clade 2 within P. sp. 1 showed a high intra-specific average p-distance of 7.7 %±0.5 and member specimens comprised collecting sites with wide geographic distribution (Vanuatu–Philippines–UK). However, specimens from Vanuatu in clade 2 showed no genetic divergence at all (VM2.1–5). Similarly, the 21 specimens of P. sp. 1 in clade 3 from the Great Barrier Reef showed only 1.0 %±1.5 genetic distances (Whitsunday Islands– Lizard Island). By contrast, P. cumingianus (Récluz, 1844), P. mellosus (Hedley, 1924), P. uber, P. sp. 2, P. sp. 3 and P. sp. 4 had low intra-specific divergence ranging from 0.1 %±0.2 to 1.3 % ± 0.3 (Table 3), even between specimens of the same species collected from widely separated localities (e.g. P. sp. 2 collected from the Great Barrier Reef and Indonesia). Thus, the genetic divergence between the P. sp. 1 clades was similar or even higher than the divergence between other taxonomically distinct Polinices species (see Tables 3 and 4).

Conchological analysis

Our conchological analysis for the entire set of taxa (Polinices s.l.) revealed low CI and medium to high RI values for many shell characters (Table 5). Only five characters (9, 13, 15, 17, 22, 28) were identified with autapomorphic features, separating Tectonatica (9, 15), Sinum (13, 17, 22), Conuber (28) or M. caprae (17) from the remaining species. As they were the only members of the Naticinae in this study, only the Tectonatica species show subsutural wrinkles (9) and, obviously, a calcareous, white operculum (15). The shells of Sinum species differ from the remaining genera by the absence of an umbilical callus and a parietal callus (13, 22) and a strongly reduced operculum (17). Conuber is the only naticid genus whose members are known to produce gelatinous, sand-free egg masses (28) instead of a sand collar.

Other characters (colour of protoconch, number of embryonic whorls, aperture height/total height ratio, morphology of the suture, colouration of the operculum, aperture size and shape, thickness, shape, size and transition of the parietal callus) showed low to medium CI values ranging from 0.12 to 0.40 and medium to high RI values ranging from 0.25 to 0.84 (see Table 5).

Low CI and low/medium RI values were calculated for shell colour and colour pattern, funicle morphology, shell solidity, and shell texture (Table 5). Most of these characters showed overlapping states in particular in Mammilla, Sinum and Neverita species (e.g. depressed shell, spiral grooves) despite the fact that these groups were not directly related to each other in the phylogenetic reconstruction (see Figs. 2, 3 and 4).

When character analysis was applied to Polinices s.s. only, the results clearly demonstrated that only a few discrete shell characters differ between the species while many characters are identical or missing (Table 5). Only two characters (1, 16) were identified with autapomorphic states, separating P. sp. 1, P. uber, P. mediopacificus (1), P. flemingianus and P. sp. 4 (16) from the remaining species. Character 1 united P. sp. 1, P. uber and P. mediopacificus, which have brownish-to-black protoconchs. P. flemingianus features a black streak on its honey-coloured operculum while P. sp. 4 has an entirely black operculum (16). Of the remaining features, the number of embryonic whorls and the protoconch size showed low/medium CI and high RI values. Other characters, such as shell colour, colour pattern, umbilical morphology, ratio of total height to aperture height and whorl expansion showed low/medium CI and low RI values in Polinices s.s.

Protoconch colour, the size of the first embryonic whorl and number of protoconch whorls were observed to show little intra-specific variability (see Table 6). However, variations in protoconch whorl size ranged from 5 % to 25 % in most species (e.g. P. sp. 1, P. sp. 2, P. sp. 3, P. sp. 4, P. albumen) for which more than three specimens had been analysed (see Tables 6 and 7).

By contrast, shell colour, shell shape and umbilicus morphology were observed to vary strongly between adult and juvenile specimens. Adult specimens of P. sp. 1, P. sp. 2, P. sp. 3 and P. mellosus predominantly possessed a closed umbilicus and a pyriform shell shape (R[w/h] > 1.1). Ratio of height to width (R[h/w]) and umbilical morphology, however, were observed to differ considerably within adults and between adult and juvenile specimens as juvenile specimens of P. mellosus and P. sp. 2 from Lizard Island have a globose shell (R[h/w] = 0.9–1.1) and a partially open umbilicus with an anterior cleft-like opening (Supplementary Fig. S3).

P. mellosus and P. sp. 2 feature identical shells (Table 6) with the yellowish-cream shell colouration in P. mellosus as the only differentiating character. However, shell colouration in P. mellosus is less intense, covers parts of shells only or is missing entirely in juvenile specimens, thus impeding clear species identification (Supplementary Fig. S3). Similarly, it is known that P. cumingianus can feature a considerable range of colouration from faint brownish horizontal bands to an entirely brown shell and shows a large size range of the umbilical callus (Cernohorsky 1971). Despite the fact that those characters were described to differentiate P. cumingianus from P. peselephanti, both species can identified reliably only by differences in protoconch size (Table 6).

Identification of white Polinices s.s. species

Using the data from our conchological analyses, we were able to identify most of the phylogenetically determined Polinices species based on a subset of conchological characters or by at least one discrete conchological character typical to a certain species (characters A–P; Table 6). Thus, taxonomic assignment to valid species was possible for most Polinices species in the phylogenetic analyses, such as P. albumen, P. cumingianus, P. flemingianus, P. mediopacificus, P. mellosus, P. peselephanti, P. uber (Table 6). Our taxonomic assignments are in agreement with most previously published species descriptions (e.g. Marincovich 1977; Majima 1989; Kabat 2000).

The three “mammilla”-like taxa P. sp. 1, P. sp. 2 and P. sp. 3 as well as P. sp. 4 differed in colouration and size of the first whorl of the protoconch ("first embryonic whorl", FEW), the total number of embryonic whorls of the protoconch (EW) and the colouration of the operculum. All P. sp. 1 specimens had a black protoconch with 2.25–2.75 EW and a FEW of 370±67.0 μm, P. sp. 2 specimens showed a white protoconch with 1.25 EW and a FEW of 660±60.0 μm and P. sp. 3 specimens had a slightly larger white protoconch with 1.25–1.50 EW and a FEW of 775±60.0 μm. The protoconch morphologies of the latter two species are therefore virtually identical to those of P. flemingianus (1.25 EW, FEW = 640 μm), P. mellosus (1.25 EW, FEW = 690±17.0 μm), P. cumingianus (1.75 EW, FEW = 700 μm) and P. mediopacificus (1.25 EW, FEW = 680 μm) and were even similar to P. sp. 4 (0.9–1.15 EW, FEW = 870±70.0 μm) (Table 6).

However, P. sp. 2 and P. sp. 3 specimens clearly differed from P. flemingianus by the colouration of the operculum (brown with a black streak in P. flemingianus, honey-coloured in P. sp. 2 and P. sp. 3) and from P. mellosus by the shell colour (P. mellosus: cream-coloured to yellowish; P. sp. 2 and P. sp. 3: purely white). Due to identical shell and overlapping protoconch morphology, P. sp. 2 and P. sp. 3 could not be differentiated unambiguously from each other (Table 6) while P. sp. 4 could be differentiated from P. sp. 1, P. sp. 2, P. sp. 3 and P. flemingianus by its slightly larger white FEW (870±70.0 μm), a slightly smaller EW (0.9–1.15) and its entirely black operculum.

Taxonomic considerations of P. sp. 1 through P. sp. 4

Our conchological analyses clearly demonstrated that species assigned to Polinices s.s. are very similar with regard to shell characters as they usually are characterized by plain white or monochrome glossy, ovate-shaped shells, overlapping protoconch features, a honey-coloured to black corneous operculum, a medium to thick parietal callus and a partly or completely filled umbilicus (see Fig. 1). Probably as a consequence of the intra-specific variation of these features and striking inter-specific similarities of recent Polinices species (e.g. partly and completely filled umbilici), a large number of species with questionable taxonomic status have been described to date: at least 55 plain white Polinices species have been proposed or described, including 21 from the Indo-Pacific region, 10 with unknown type locality and 24 from regions other than the Indo-Pacific (Supplementary Table S2). Many of these taxa are now regarded as junior synonyms of other Polinices species (Tryon 1886; Cernohorsky 1971; Marincovich 1977; Kabat 2000).

Our conchological analyses of type species revealed P. sp. 1 to be conspecific with P. mammilla based on the colouration and size of the protoconch, while P. sp. 2 and P. sp. 3 are considered to be referable to P. jukesii (Reeve, 1855) and P. dubius (Récluz, 1844), respectively. As N. dubia Recluz, 1844, however, is junior homonym of the fossil species Natica dubia Römer, 1836 we herewith introduce a replacement name for Natica dubia Recluz, 1844, Polinices constanti Huelsken and Hollmann, to maintain taxonomic stability.

However, we emphasise that the assignment of P. sp. 2 and P. sp. 3 to P. jukesii and P. constanti is preliminary as the existing types in the dry collection of the NHM (London) unfortunately cannot be analysed with molecular methods due to the lack of preserved tissue. P. sp. 4, by contrast, is similar to P. tawhitirahia Powell, 1965 based on the colouration of the operculum and the size of the protoconch. Detailed species descriptions and discussions of these four species follow below.

Family Naticidae Guilding, 1834

Subfamily Polinicinae Gray, 1847

Genus Polinices Montfort, 1810

Polinices mammilla (Linnaeus, 1758) [= Polinices sp. 1 in the preceding discussion]—Figs 1k, 5a

Nerita mammilla Linnaeus, 1758; Syst. Nat. ed. 10, pl. 52 [fide Kabat 1990]

+„Mamma albula Chemnitz, 1781“; Syst. Conch. Cab., 5: 280, pl. 189, Figs. 1928–31, (non binomial)

+Albula mammilla Röding, 1798; Mus. Bolten., p. 20, (ref. Chemnitz, op. cit., Figs. 1928–31)

+Mammillaria tumida Swainson, 1840; Treat. Malac. p. 345 (ref. Chemnitz, op. cit. Figs. 1928–31)

+Natica pyriformis Récluz, 1844; Proc. Zool. Soc. Lond. pt. 11: 211

+Natica albula Récluz, 1851; J. Conchyl. 2(2): 194 (ref. Rumphius), pl. 22 Fig. E)

+Natica ponderosa Philippi, 1849; Syst. Conch. Cab. 2nd ed. 2(1): 32 pl. 4, Figs. 9–10

+Natica cygnea Philippi, 1850; Syst. Conch. Cab. 2nd ed. 2(1): 80, pl. 12, fig. 6

Natica mammilla (Linnaeus, 1758); Reeve (1855), Conch. Icon., pl. 7, fig. 27

Natica mammilla (Linnaeus, 1758); Sowerby (1883), Thes. Conchyl., 5: 85, pl. 3, Figs. 28–30

Polinices (Polinices) mammilla (Linnaeus, 1758); Lass, Bern. P. Bish. (1943), Mas. Bull., 119: 210, pl. 36, Figs. 4–5

+Polinices pyriformis (Récluz, 1844); Habe & Kosuge (1967), Stand. book Jap. shells col., 3: 45, pl. 18, fig. 7

+Polinices (Polinices) tumidus, (Swainson, 1840); Cernohorsky (1971), Rec. Auckland. Inst. Mus., 8 (December) p.190, Figs. 49–50

P. mammilla (Linnaeus, 1758); Torigoe and Inaba (2011), sp. 93, pp. 37–38.

Description

Shell

Shell up to 60 mm in height, ovate-pyriform to pyriform, glossy white, occasionally with light, ill-defined brownish striae or brownish spots on the shoulder of the bodywhorl, giving it a rusty appearance. Ratio of shell height to shell width 1.29±0.07 in specimens analysed in this study (n = 30) and 1.26 in the lectotype (n = 1). Aperture wide and semi-ovate, ratio from aperture height to total height 0.67±0.07 in analysed specimens (n = 30) and 0.65 in the lectotype specimen (n = 1). Umbilicus completely covered by a heavy callus in adult specimens; a small anterior umbilical groove may be present in juveniles but also in adult specimens. Parietal callus extends into umbilical callus without sulcus.

Protoconch

Brownish to black, 2.00–2.25 whorls, size of first embryonic whorl 370±67 μm in specimens analysed in this study (n = 30). Protoconch in lectotype broken (see discussion below).

Operculum

Corneous, light brown in colour.

Foot

Propodium white and long (>2 times shell). Mesopodium white, overlapping the protoconch, leaving only a quarter of the shell surface visible.

Distribution

Indo-west Pacific to Easter Island, Red Sea.

Differential diagnosis

P. mammilla can distinctly be differentiated from any other Indo-west Pacific white Polinices species by its smaller and black protoconch with at least 2.00 embryonic whorls (EW = 2.00–2.25, FEW = 370±67 μm).

Material examined

Type specimens of Nerita mammilla (ZMUU#769), Mammillaria albula [= Mamma albula, non-binomial] [ZMUC (Cernohorsky 1974)] and Natica pyriformis (BMNH#1991089.1-3, BMNH#1845.6.24.56-58, MHNG#2017, MHNG#2018). For molecularly and morphologically analysed specimens see Table 1.

Discussion

Based on the black protoconch and the pyriform white shell, P. sp. 1 is identical to the description of a ZMUU specimen (#769) designated as lectotype for Nerita mammilla Linnaues, 1758 by Kabat (Kabat 1990). Conchological analyses of type material of Nerita mammilla and synonymized species, however, have revealed new information that will be discussed in the following, in order to retain taxonomic stability in this important Polinices taxon.

The name P. mammilla has been used traditionally for white-shelled Indo-Pacific Polinices species with closed umbilicus (e.g. Majima 1989; Kabat 1990), which have a “…protoconch reddish to black” (Kabat 1990: p. 17). Although Kabat examined the type specimen (1990), he did not mention that the protoconch of the lectotype of Nerita mammilla Linnaeus, 1758 (ZMUU#386) is broken and filled with a blackish sand grain (Fig. 5a). It is therefore impossible to determine the exact proportions and the colouration of the protoconch of the lectotype of Nerita mammilla Linnaeus, 1758.

Motivated by the perceived uncertainty of the taxonomical validity of Nerita mammilla Linnaeus, 1758, the original collecting site of which is “Bahamas” in the Caribbean Sea, Cernohorsky (1971) suggested to use the next available name Mammillaria tumida Swainson, 1840 for the Indo-Pacific white Polinices species with a black protoconch (Fig. 5b, Table 7). Unfortunately, the protoconch of the syntype of M. tumida (= Albula mammilla Röding, 1798) at ZMUC (Cernohorsky 1971), which is based on Mamma albula Chemnitz, 1758 [non-binomial; in Martini and Chemnitz 1769–1829] and for which the type locality is unknown, is broken, too (see Fig. 5b). Therefore, even with this type specimen of a junior synonym it is impossible to determine the exact proportions and the colouration of the protoconch of P. mammilla.

The next available name for this taxon is Natica pyriformis Récluz, 1844 (our Fig. 5c). The three specimens in the collections of the NHM (London) (BMNH#1991089), "syntypes" according to Kabat et al. (1997), have been collected in the Philippines as is noted on the NHM collection label. In the original description, Australia is mentioned as an additional locality. Two of the syntypes have black protoconchs (BMNH#1991089.1; BMNH#1991089.3) of 410 μm±10 and 2.25–2.50 EW while the protoconch of the third syntype (BMNH#1991089.2) is broken (Table 7). The protoconchs of three additional "possible syntypes" (Kabat et al. 1997) at the NHM (London) (BMNH#1845.6.24.56–58) are blackish, while those of five further "possible syntypes" (Kabat et al. 1997) found at the MHNG (MHNG#2017, 2 specimens, MHNG#2018, 3 specimens) are white. Although the operculum is unknown in N. pyriformis, shell shape, protoconch morphology, type locality and morphology of the umbilicus is virtually identical between Recent P. mammilla and the specimen BMNH#1991089.1 of N. pyriformis and very similar to the other five specimens of N. pyriformis (syntypes BMNH#1991089.2 and BMNH#1991089.3; possible syntypes BMNH#1845.6.24.56-58) found at the NHM. In order to maintain taxonomic stability in this taxon, we designate the specimen No. BMNH#1991089.1 (black protoconch, FEW: 400 μm, EW 2.25; our Fig. 5c) as lectotype and the other five specimens at the NHM as paralectotypes (BMNH#1991089.2–3, BMNH#1845.6.24.56-58) of N. pyriformis.

While it can no longer be determined whether the neotype of Nerita mammilla or the type of Natica tumida originally had a black protoconch or not, the lectotype of Natica pyriformis clearly does so. As all three taxa are believed to be conspecific (P. tumidus and P. pyriformis: Cernohorsky 1971; Majima 1989; Kabat 1990; P. mammilla and P. pyriformis: Majima 1989; Kabat 1990; P. mammilla and P. tumidus: Majima 1989; Kabat 1990) and in order to retain taxonomic stability, we follow Kabat's concept of P. mammilla as the earliest name for the Indo-Pacific white-shelled, glossy Polinices specimens with a black protoconch, with P. tumidus and P. pyriformis being junior synonyms.

Polinices jukesii (Reeve, 1855) [= Polinices sp. 2 in the preceding discussion]—Figs. 1a, 6g
https://static-content.springer.com/image/art%3A10.1007%2Fs13127-012-0111-5/MediaObjects/13127_2012_111_Fig6_HTML.gif
Fig. 6

Analysed type specimens or figured type specimens of taxa that could potentially represent Polinices sp. 2, Polinices sp. 3 or Polinices sp. 4. aNatica controversa Pritchard & Gatliff, 1913 [MV#F7695]. bNatica dubia Récluz, 1844 [BMNH#1991085] (= P. constanti Huelsken and Hollmann, herein; replacement name). cNatica deiodosa Reeve, 1855 [BMNH#1991069]. dUber mellosum Hedley, 1924 [AMS#C20058]. eNatica phytelephas Reeve 1855 [BMNH#1991096]. fPolinices putealis Garrard, 1961 [AMS#C63344]. gNatica jukesii Reeve, 1855 [BMNH#1991067]. (h) Polinices tawhitirahia Powell, 1965 [Auckland Museum #71242]. iNatica vavaosi Reeve, 1855 [figured type]. jNatica galactites Philippi, 1851 [figured type]. kNatica cygnea Philippi, 1850 [figured type]. lNatica virginea Philippi, 1850 [figured type]. For further information see Table 1. Bars 0.5 cm

Natica jukesii Reeve, 1855; Gen. Natica, Conch. Icon. 9: sp. 84, pl. 19, Figs. 84a,b

Polinices jukesii (Reeve 1855); Torigoe and Inaba (2011), sp. 84, p. 33, Pl. 2, Fig. 11.

Description

Shell

Shell morphologically similar to P. mammilla but smaller in maximum size with up to 34 mm in height (type specimens: 31–34 mm; specimens analysed in this study: 13–25 mm), pyriform to ovate in shape. Ratio of shell height to shell width 1.27±0.01 in specimens analysed with molecular methods (n = 12) and 1.01±0.05 in type specimens (n = 3; BMNH 1991067.1-3). Aperture wide and semi-ovate, ratio from aperture height to total height 0.72±0.06 in specimens analysed with molecular methods (n = 12) and 0.58±0.05 in type specimens (n = 3). Umbilicus completely covered by a heavy callus in adult specimens; a small anterior umbilical groove may be present in juveniles. Parietal callus extending into umbilical callus without a sulcus.

Protoconch

White, 1.25–1.50 whorls, size of first embryonic whorl 770±60 μm (n = 12) in specimens analysed in this study, 1.50–1.75 whorls, size of first embryonic whorl 790±50 μm in type specimens (n = 3).

Operculum

Corneous, light brown in colour.

Foot

Propodium white and long (>2 times shell). Mesopodium white and short, overlapping protoconch, leaving only a quarter of the shell surface visible.

Distribution

Central Indo-Pacific, East-Australia

Differential diagnosis

This species differs only in protoconch morphology slightly from P. constanti (P. jukesii: EW = 1.25 – 1.50, FEW = 770±60 μm; P. constanti: EW = 1.25–1.50, FEW = 660±60 μm). As the shells of both species do not appear to have further differentiating characters, the identification of the species will be impossible in cases where protoconch features are identical. Similarly, the shells of P. jukesii are virtually identical to those of P. flemingianus. However, P. flemingianus can be unequivocally differentiated from P. jukesii by the black streak on its otherwise light brown operculum (Fig. 1, Table 6) and also has a slightly smaller protoconch (P. jukesii: FEW = 770±60 μm, EW = 1.25–1.50; P. flemingianus: FEW = 640 μm, EW = 1.25). When operculi are missing, species of P. flemingianus and P. jukesii can be separated from each other based only on molecular data.

Material examined

Syntypes of Natica jukesii, BMNH 1991067.1-3, for additional specimens analysed in this study see Table 1.

Discussion

see discussion of P. constanti [replacement name for Polinices dubius (Récluz, 1844)].

Polinices constanti Huelsken and Hollmann, herein (replacement name for Polinices dubius (Récluz, 1844) [= Polinices sp. 3 in the preceding discussion]—Figs 1b, 6b

Natica dubia Récluz, 1844 [non Römer, 1836]; Proc. Zool. Soc. London (for 1843) 11(130): 209–210

+Polinices dubius (Récluz, 1844); Torigoe and Inaba (2011), sp. 116, p. 44.

Description

Shell

Shell morphologically identical to P. mammilla, P. jukesii, P. flemingianus and P. cf. tawhitirahia but the specimens analysed here were smaller in maximal and average size, with up to 32 mm in height (type specimens: 30–32 mm; specimens analysed in this study: 13–19 mm). The shell is pyriformly-ovate with a ratio of shell height to shell width of 1.28±0.02 in specimens analysed with molecular methods (n = 3) and 1.00±0.06 in type specimens (n = 2). Aperture wide and semi-ovate with a ratio of aperture height to total height of 0.65±0.03 (n = 3). Umbilicus completely covered by a heavy callus in adult specimens sometimes showing a small anterior umbilical groove. Smaller specimens investigated here predominantly exhibit an anterior umbilical groove. Parietal callus extending into umbilical callus without a sulcus.

Protoconch

White, 1.25-1.50 embryonic whorls, size of first embryonic whorl 660±60 μm in specimens molecularly analyzed in this study (n = 3), EW = 1.20, size of first embryonic whorl 656±97 in type specimens (n = 2).

Operculum

Corneous, light brown in colour.

Foot

Propodium white and long (> 2 times shell length). Mesopodium white and short, overlapping the protoconch, leaving only a quarter of the shell surface visible.

Distribution

Central Indo-Pacific, East Australia.

Differential diagnosis

The species is virtually identical to P. jukesii and can be differentiated only by the slight differences in protoconch features (see above). P. flemingianus can be differentiated from P. constanti only by its black streak on the operculum, as both species show identical protoconch features (P. flemingianus: EW = 1.25, FEW = 640 μm; P. constanti: EW = 1.25–1.50, FEW = 660±60 μm). P. mellosus can only differentiated from P. constanti by its yellowish-cream colouration (P. mellosus: EW = 1.25–1.45, FEW = 690±20 μm; P. constanti: EW = 1.25–1.50, FEW = 660±60 μm) (Table 6 and 7). When operculi are missing, species of P. flemingianus and P. constanti can be separated from each other based only on molecular data.

Material examined

Syntypes of Natica dubia, BMNH 1991085.1-2, for additional specimens analysed molecularly and morphologically, see Table 1.

Discussion

While being identical to P. mammilla in most conchological characters, P. sp. 2 and P. sp. 3 both have significantly larger white protoconchs that are virtually identical among the two taxa. Based solely on shell features, assignment to the following species appears possible (original combinations given): N. candidissima Le Guillou, 1842, N. dubia Récluz, 1850, N. cygnea Philippi, 1850, N. virginea Philippi, 1850, N. galactites Philippi, 1851, N. jukesii Reeve, 1855, N. phytelephas Reeve, 1855, N. vavaosi Reeve, 1855, N. deiodosa Reeve, 1855, N. controversa Pritchard & Gatliff, 1913, Uber mellosum Hedley, 1924, P. putealis Garrard, 1961 and P. tawhitirahia Powell, 1965 (see Table 7).

Type specimens are available only for 8 out of those 13 taxa: N. dubia, N. jukesii, N. phythelephas, N. deiodosa, N. controversa, U. mellosum, P. putealis and P. tawhitirahia. Reliable species identification is possible only if operculum colouration as well as protoconch dimensions are known [see results about P. flemingianus and P. cf. tawhitirahia (= P. sp. 4) as an example of the importance of operculum colouration]. In our analysis we therefore focussed on those seven species out of the eight with available type material for which the operculum colour and protoconch dimensions are known: N. dubia, N. jukesii, N. controversa, N. deiodosa, P. putealis, U. mellosum and P. tawhitirahia, thus excluding N. phytelephas (Fig. 6).

Conchologically, P. controversus is identical to P. peselephanti based on the depressed shell shape, an open umbilicus, the presence of an umbilical callus and the large FEW, which reaches nearly 1,500 μm with 1.75 EW (P. peselephanti: 1.25 EW, FEW = 1,200 μm). P. putealis is a deep sea species, found at >100 m depth off South-East Australia (type locality: Botany Bay, Sydney, NSW). It has a brownish protoconch, 1.75 EW and a FEW of 320 μm. P. tawhitirahia is the only naticid species reported to have an almost black operculum (Powell 1965) and might be conspecific with P. sp. 4. P. mellosus and P. deidosus can also be excluded from the list as both are of creme-yellowish shell colour. P. mellosus furthermore can be differentiated from P. sp. 2 and P. sp. 3 based on molecular results (see Figs. 2, 3 and 4, Table 3). These differences in shell morphology allow excluding these three species as candidate taxa for P. sp. 2 or P. sp. 3.

Thus, only N. dubia and N. jukesii remain as possible name-bearing types for the two unknown species, as they show similar shell morphology, operculum colour and protoconch characters as P. sp. 2 and P. sp. 3. Based on the observation that P. jukesii has a slightly larger protoconch than P. dubius, we conclude that P. jukesii is conspecific with our P. sp. 2 while our P. sp. 3 is referrable to P. dubius (see Table 7). N. dubia Recluz, 1844, however, is a junior homonym of Natica dubia Römer, 1836 who used this name for a fossil naticid specimen. We therefore introduce Polinices constanti Huelsken and Hollmann, herein as a replacement name (nomen novum) for Natica dubia Recluz, 1844. Etymology: a patronym honoring Constant A. Récluz who described this species first in 1844 (as Natica dubia) and who made extensive contributions to naticid taxonomy.

However, we emphasize that the reference of the molecularly defined P. sp. 2 and P. sp. 3 to P. jukesii and P. constanti, respectively, cannot be verified by molecular analysis as the type material of P. jukesii and P. constanti kept at the NHM (London) does not include preserved tissue.

It is noteworthy that the ratio of height to width and the umbilical morphology differ between type specimens and specimens analysed in this study in P. jukesii and P. constanti. The specimens in both type lots are significantly larger than the specimens used in the molecular analyses, with differences ranging between 5 and 19 mm. As shown for P. jukesii and P. mellosus, the ratio of height to width may vary strongly between juvenile and adult Polinices specimens, ranging from 0.9 to 1.20. It is worth noting that the studied specimens of P. mammilla (n = 30) also vary considerably in height/width ratio, from 1.09 to 1.50. The variability of this character is also reflected in the low statistical values (CI: 0.33) in our conchological analysis of Polinices species. This goes along with results from empty shell material of P. mammilla from Lizard Island showing a height to width ratio range of 0.8–1.5 (own observations of TH). Similarly, the umbilical morphology varied in P. jukesii, P. mammilla, P. flemingianus and P. mellosus, resulting in low statistical values (CI: 0.22, RI: 0.42) in our conchological analysis of Polinices species. At present, we therefore predict the shell shape (i.e. ratio of height to width) and the umbilical morphology of (Table 5, character 10) to be too variable to provide characters for reliable species identification in P. jukesii and P. constanti.

To our knowledge, no type material of species described earlier by Le Guillou and Philippi is available at this stage (N. candidissima Le Guillou, 1842, N. cygnea Philippi, 1850, N. virginea Philippi, 1850 and N. galactites Philippi, 1851). Should type material of Le Guillou and Philippi be found in the future, conchological analysis of the protoconchs and the operculi of such type specimens may change the taxonomic assignments for the species here referred to P. jukesii and P. constanti. As both species are virtually identical, synonymies of putative conspecific taxa are difficult or impossible to discuss. We therefore refrain from providing any synonymy for either of the two species. Such synonymies can be attempted only when protoconch or operculum features for the considered synonymous species become known, which could happen only if type specimens can be located in the future (see discussion).

Polinices cf. tawhitirahia Powell, 1965 [= Polinices sp. 4 in the preceding discussion]—Fig. 1e, 6h

P. tawhitirahia Powell, 1965, Rec. Auck. Inst. Mus., 6(2), Figs. 22(1–3), p. 163

+P. mellosum (Hedley 1924), in Majima (1989), Figs. 18–19, p. 48 [not mellosus Hedley, 1924]

+P. mellosum (Hedley 1924), in Kabat (2000), Fig. 31, p. 72 [not mellosus Hedley, 1924]

+P. pyriformis (Récluz, 1844), in Okutani (2000), Fig. 29, pl. 126 [not pyriformis Récluz, 1844]

+P. mellosus (Hedley, 1924) Torigoe and Inaba (2011), sp. 94, p. 38, Pl. 2, Fig. 14. [not mellosus Hedley, 1924]

Description

Shell

Shell morphologically similar to P. mammilla, P. flemingianus, P. constanti and P. jukesii: The shell is plain white, 13–20 mm in height, globose to slightly pyriform with a ratio of shell height to shell width of 1.09±0.03 (n = 5). Aperture is wide and semi-ovate; the ratio of aperture height to total height is 0.82±0.01 (n = 5). Umbilicus is completely covered by a heavy callus. Parietal callus thick, filling posterior apertural angle. Parietal callus extending into umbilical callus without a sulcus.

Protoconch

White, 0.9–1.15 whorls, size of first embryonic whorl 870±70 μm in specimens analyzed in this study (n = 5); white, 1.75 whorls and 622 μm in the holotype.

Operculum

Corneous and entirely black operculum.

Distribution

New Zealand, West-Australia, Indonesia (Ambon).

Differential diagnosis

This species can be differentiated from other plain white Polinices species by its entirely black operculum and its white protoconch with only 0.9–1.1 whorls and a large first embryonic whorl of 870±60 μm.

Material examined

For specimens analysed molecularly and morphologically see Table 1.

Discussion

Shell characters are very similar to most of the other plain white Polinices species analysed in this study. However, in contrast to the honey-coloured corneous operculi of other white Polinices species, P. sp. 4 has an entirely black corneous operculum. Other distinguishing features are the larger size of the first protoconch whorl (FEW: 870 μm±70.0) and the low number of embryonic whorls (EW: 0.9–1.15). A glossy white Polinices species with a black operculum and a white protoconch has been described from New Zealand by Powell (1965) as P. tawhitirahia (his Fig. 7G). Interestingly, a white Polinices specimen with a black operculum (Kabat 2000, his Fig. 31, p. 72) was also found during the 1990 Rumphius Biohistorical Expedition to Ambon (Indonesia). Furthermore, specimens with shell characters identical to our material have also been pictured by Majima (1989; text: Figs. 18 and 19, on p. 48), by Okutani (2000; pl. 126, fig. 20) and by Torigoe and Inaba (2011; pl. II, fig. 14). Kabat, Majima as well as Torigoe and Inaba erroneously named the species “P. mellosum (Hedley 1924)”, while Okutani erroneously named it “P. pyriformis (Récluz, 1844)”. Although the operculum of the type of P. pyriformis is unknown, that species has to be synonymized with P. mammilla based on protoconch morphology (300–400 μm) and protoconch colouration, as discussed above and thus must possess a honey-coloured operculum. Therefore, the specimen figured by Okutani as P. pyriformis with a black operculum cannot be P. pyriformis but instead is conspecific with our P. sp. 4.

Based on overlapping similarities in operculum and shell morphology we therefore conclude that P. sp. 4 can be assigned to P. tawhitirahia as differences in protoconch morphology of the holotype (NZ71242) of P. tawhitirahia and P. sp. 4 (FEW = 622 μm; EW = 1.75 in P. tawhitirahia vs FEW = 870 μm±70.0; EW = 0.9 1.15 in P. sp. 4) lie within the margin of morphological variability (5–25 %, see Results section) generally observed in white Polinices species. Supporting this conclusion, Kabat (2000) also synonymises his Ambon Polinices specimen with black operculum with P. tawhitirahia from New Zealand. The diverse collecting sites (North Western Australia/Ambon/Japan vs New Zealand) suggests that this species either has a very broad distribution range or, alternatively, the specimens investigated in this study represent an additional species with an entirely black operculum, in which case this opercular feature would occur in at least two different taxa of white Polinices. However, given the wide specimen distribution and the fact that most operculi are missing in museum specimens, we cannot exclude unequivocally that P. sp. 4 does not represent P. tawhitirahia (Table 7, Fig. 6).

Discussion

Our study represents the first approach to clarify the questionable species identifications and phylogenetic relationships within the conchologically rather uniform taxon Polinices and its predicted closely related taxa Conuber, Euspira, Mammilla, Neverita and Sinum (Cernohorsky 1971; Marincovich 1977; Kabat 1991, 1996). The monophyletic grouping and the high genetic divergence demonstrate clearly that Conuber, Euspira, Mammilla and Neverita indeed represent independent genera and not merely subgenera of Polinices (Figs. 2, 3 and 4, Supplementary Figs. S1, S2).

Beyond the clarification of supraspecific taxa relationships within Polinices s.l., our phylogenetic analyses allow species differentiation within the group of conchologically uniform white Polinices species. This includes taxonomical evaluations and re-descriptions of the common P. mammilla (= P. sp. 1), the formerly synonymized species P. jukesii (= P. sp. 2) and P. constanti (= P. sp. 3) as well as P. cf. tawhitirahia (= P. sp. 4). Our analyses furthermore prove for the first time that many conchological characters traditionally used in the description and identification of species in Polinices s.l. are homoplasious and of low information value due to identical expression in distinct clades or intra-specific variability.

Discussion of conchological analyses

The low CI values indicate that most traditionally used shell features have a wide range of intra-specific variability or occur simultaneously in various Polinices s.l. genera, thus revealing strong homoplasy. However, some shell characters appeared to be informative with respect to the phylogenetic pattern of the taxa, which was revealed by the differences between high RI values and low CI values for the same characters (Table 5). Several characters listed here were identical (e.g. operculum surface, aperture size, colour of columellar callus) or showed intra-specific variability in all analysed Polinices s.l. species (e.g., umbilical structure). This pronounced uniformity of shell characters in Polinices clearly is the reason for problematic taxonomic assignments in this species group.

The apparent autapomorphic nature of some characters (e.g. subsutural wrinkels) can be an effect of insufficient taxon sampling. Other characters, however, seem to be helpful in discriminating between genera (e.g. operculum size and umbilical callus in the Sininae; egg mass structure in Conuber) or species (e.g. operculum size in M. caprae; protoconch colour in P. sp. 1, P. uber, P. mediopacificus; operculum colour in P. flemingianus and P. sp. 4). Characters with high CI/RI values include protoconch features (EW, FEW, colouration), the presence or absence of an umbilical callus, parietal callus features, operculum features and egg mass morphology. These characters are highly informative when used for species identification in a set of closely related Polinices species (see Polinices) and are even informative (but with lower CI/RI values) in discriminating Polinices s.l. species despite their intra-generic variability. However, characters such as shell shape, shell colour and general umbilical characters (anterior cleft) are less informative in discriminating either Polinices s.l. or Polinices s.s. species, owing to their intra-specific variability and convergent occurrences.

Our results therefore confirm predictions that developmental characters are highly informative in taxonomic assignments of gastropods (Bouchet 1989). However, protoconch features coincide between closely related Polinices s.l. species (Table 6), indicating a need for additional discrete morphological characters for reliable species identification. Unfortunately, little is known about shell variability (shape and umbilicus) with regard to ontogeny. Given the fact that adult and juvenile specimens in P. mellosus, P. sp. 1 and P. sp. 2 show highly variable shell morphology and shell colouration, investigation on a larger scale than is presented here is needed.

Given our results, we cannot confirm Kool’s sweeping criticism (1993) regarding the usefulness of shell character in phylogenetic and systematic analyses. Of course, shell characters in Polinices species (and probably in all naticids) are subject to evolutionary convergence due to analogous adaptations to environmental constraints based on identical predatory behaviour and their burrowing way of life (Bandel 1999). However, shell characters need to be analysed carefully before they can be rejected as uninformative (Vermeij and Carlson 2000). This is particularly important for the identification of species via DNA barcoding and phylogenetic approaches and for the subsequent assignment of type specimens of gastropods for which, in most cases, only empty shells are available. In conchologically homogeneous groups such as the moon snails, shell characters may not be informative enough to resolve taxa in phylogenetic and cladistics analyses, but can certainly be used in species identification when described and categorized accurately (see also Aronowsky 2003).

Phylogenetic and systematic considerations

In our study definitions of species and genera were based on the monophyletic arrangement of each of the taxa in the phylogenetic trees and the high genetic divergence (13 %–16 %) between the groups. Any species assignment thus follows a phylogenetic species concept.

Genus Polinices Montfort, 1810

Based on molecular data, the genus Polinices can be divided in two groups. One group comprises P. sp. 1 (= P. mammilla), P. uber and P. mediopacificus, while the second group contains P. albumen, P. cumingianus, P. mellosus, P. flemingianus, P. sp. 2 (= P. jukesii), P. sp. 3 (= P. constanti), P. sp. 4 (= P. cf. tawhitirahia) and P. peselephanti (Figs. 2, 3 and 4, Supplementary Figs. S1S2). Species in the second Polinices group are related more closely to each other and showed lower intra-specific geographic resolution. This may be largely a consequence of limitations in taxon sampling, as most specimens were sampled in one region only or only one specimen per species was available for sequencing (e.g. P. flemingianus).

The patterns of genetic variation in the monophyletic P. mammilla (= P. sp. 1) between Egypt (clade 1), Indonesia, New Caledonia (clade 2) and the Great Barrier Reef (clades 2/3) correspond to the phylogeographic category II of Avise (2000) in which it is assumed that the different mitochondrial haplotypes originated either from hitherto unidentified sympatric species or from previously isolated lineages with restricted genetic connectivity (see Thomaz et al. 1996; Avise 2000). Thus, the mtDNAs appear to have diverged in allopatry, with a secondary admixture of populations in the northern Great Barrier Reef. This is supported by the fact that two genetically distinct lineages have been found on Lizard Island (clades 2/3). However, the specimens from Vanuatu and Lizard Island in clade 2 are also clearly separated from clade 3 by the slower evolving protein-coding histone H3 gene fragment (see Supplementary Fig. S2). Thus, the apparent genetic divergences between the three P. mammilla lineages support a separation at species level. However, as all specimens are conchologically identical and grouped together in a single monophyletic taxon, more data will be needed to test whether these clades may indeed represent different species.

The phylogenetic separation of the morphologically virtually identical P. constanti and P. jukesii (Figs. 2, 3 and 4) could either indicate the existence of non-monophyletic species with two mitochondrial lineages caused by incomplete lineage sorting or hybridisation, or indicate the existence of two independent species (Davison 2000; Funk and Omland 2003; Meyer and Paulay 2005; Huelsken et al. 2011b). The genetic divergence between P. jukesii and P. constanti is identical to, or even higher than, values calculated for conchologically well-separated Polinices species in this group, such as P. mellosus, P. cf. tawhitirahia and P. cumingianus (Table 4). Additionally, virtually identical mitochondrial and nuclear sequences for P. jukesii were obtained from two independent and not directly connected localities in the Philippines and the Great Barrier Reef indicating strong genetic connectivity. The two taxa are furthermore strictly separated by the slower evolving protein-coding histone H3 gene fragment, with P. jukesii sharing the same genetic information with P. cumingianus, P. mediopacificus and P. peselephanti in this gene fragment (see Supplementary Fig. S2). The molecular data in combination with slight differences in protoconch morphology therefore rather reject the hypotheses of hybridisation or incomplete lineage sorting in P. jukesii and P. constanti but support the idea that the two taxa are separated at the species level.

Genus Mammilla Schuhmacher, 1817

Considering the paraphyletic arrangement of Polinices, with Mammilla grouping within the former clade in some of the phylogenetic analyses (Figs. 2, 3 and 4), the question may be posed whether Mammilla should be classified as a subgenus of Polinices. Despite a close genetic relationship between Mammilla and Polinices, both taxa are well separated by their distinct conchological characters (see Fig. 1). The phylogenetic trees show increasing resolution of Mammilla and Polinices from slow to fast evolving genes (Figs. 2, 3; Supplementary Figs. S1, S2), suggesting that Mammilla and Polinices represent two independent genera which were separated from each other more recently.

As mentioned above, the genus Mammilla in earlier classifications has been described occasionally as closely related to the Sininae. In his compilation of the Naticidae from Fiji, Cernohorsky (1971) stated that Eunaticina (subfamily Sininae) “… may represent an intermediate group between Mammilla and Sinum…”. Mammilla and Sinum thus have been considered closely related taxa by Cernohorsky (1971), an idea that, amongst others, was later also taken up by Kabat (1996). Our morphological analyses indicate a high conchological concordance between members of these two taxa, such as the depressed shell shape, the thin shell, the shell texture, the widened aperture and the reduced operculum in some Mammilla species. As the molecular data presented here support a very close relationship between Mammilla and Polinices species, we conclude that the apparently similar shell characters in Sinum and Mammilla must have evolved independently at least twice within the Naticidae.

Conuber Finlay & Marwick, 1937

The genus Conuber is best suited to illustrate the extensive variability and high similarity of conchological features in species of the Polinicinae. Shell features characterizing valid Conuber species are also found in Neverita (compare C. incei) or Polinices (compare C. conicus, C. sordidus) (Fig. 1). Not surprisingly, C. incei has been assigned to the genus Neverita based on its depressed shell form, widened aperture and thin parietal callus (e.g. Hacking 1998). Based on the pyriform shell form, ratio of aperture height to total height and the morphology of parietal callus and columellar callus, C. conicus and C. sordidus were often assigned to the genus Polinices (e.g. Marincovich 1977; Booth 1995; Morton 2008). However, our phylogenetic analysis clearly demonstrated Conuber representing a distinct monophyletic taxon, thus contradicting the view of Conuber as a subgenus of Polinices as proposed by Finlay and Marwick in 1937 (p. 53).

The distinctive character defining the genus Conuber is the large, sausage-shaped gelatinous egg mass without sand grain incorporation, which differs from the typical sand collar found in all other naticids (Murray 1962, 1966; Booth 1995). This feature represents an autapomorphic character for Conuber (see also Riedel 2000) as it has not yet been found in any other naticid genus.

Neverita Risso, 1826

The genus Neverita was characterized by homogeneous conchological characters such as a depressed shell shape, an ovate aperture, a large parietal callus, a greatly enlarged body whorl and a typical umbilical area containing a large, distinctive funicle (see Cernohorsky 1971; Marincovich 1977; Majima 1989). Its placement as a subgeneric taxon within the genus Polinices, however, was based on the occurrence of several conchological characters that could be assigned to both taxa. For instance, P. albumen, P. peselephanti and P. cumingianus have “Neverita”-like depressed to slightly globose shells and thus have often been considered to belong to Neverita, either at the generic level or at the subgeneric level as in Polinices (Neverita) (e.g. Cernohorsky 1971; Majima 1989). The concept of Neverita as a subgenus of Polinices can now be rejected because the taxa Neverita and Polinices each form statistically well-supported monophyletic clades in our analyses. In consequence, similar or even identical shell characters have evolved separately in these two genera.

To our surprise, the widely known and well-investigated (e.g. Bernard 1967; Grey et al. 2007; Cook and Bendell-Young 2010) Northern Pacific species E. lewisii groups within Neverita as sister species to the Australian N. aulacoglossa in the COI tree. This confirms morphology-based cladistics with E. lewisii and E. heros grouping within Neverita (Aronowsky 2003). Although our phylogenetic placement is based only on sequence data obtained from the mitochondrial COI gene fragment, we conclude from the data sets of Aronowsky (2003) and this study that E. lewisii (and probably its sister species E. heros) should definitely be assigned to Neverita (see Table 1).

The existence of a separate (sub)genus Glossaulax Pilsbry, 1929 within Neverita (e.g. Marincovich 1977; Majima 1989) appears doubtful. The (sub)genus Glossaulax is defined by an umbilical callus that covers the umbilicus entirely and is divided into anterior and posterior lobes by a narrow transverse groove (Majima 1989). In the present study, the type species of the (sub)genus Neverita (Glossaulax), N. (G.) reclusiana, is grouped together with its nominate sister species N. (G.) didyma but not with the Australian N. (G.) aulacoglossa (Figs. 2, 3 and 4). The species are in fact separated from each other by species which are distinctly assigned to Neverita s.s. (e.g. N. delessertiana, N. duplicata) (Figs. 2 and 3).

Our data, albeit somewhat preliminary, supports the proposed validity of N. didyma (Indo-Pacific) and N. aulacoglossa (Eastern Australia) as distinct species and reject their former synonymisation under the name N. didyma (Kabat 2000). However, more sequences for N. didyma (presented here by one sequence from Taiwan, AF550509; Strong 2003) are needed for clarification.

Genus Euspira Agassiz in J. Sowerby, 1837

According to the analyses presented here, Euspira is not a subgenus of Polinices (Marincovich 1977) but represents a valid genus related closely to Conuber and Neverita. Similar to earlier results, Payraudeautia intricata (Donovan, 1804) exhibits a high genetic similarity with and thus groups within, Euspira in all genetic analyses. Synonymisation of Payraudeautia with Euspira is therefore appropriate (see Table 1; Huelsken et al. 2008).

The genus Euspira Agassiz in Sowerby, 1837, is based on the fossil European species Natica glaucinoides Sowerby, 1812 from the Middle Eocene, by subsequent designation (Bucquoy et al. 1883). The genus is characterized by a globose to elongate-globose shell with a partly-to-fully open umbilicus, abutting to an impressed suture, a slender umbilical callus, convex whorls and a turreted spire (Bandel 1999). Species assigned to Euspira therefore show many shell characters (e.g. umbilical morphology, shell shape) that are identical to those in other naticid genera (e.g. Natica, Tectonatica). Understandably, Bandel (1999) criticized the application of these shell characters in the establishment of a separate genus, Euspira, in particular since neither operculum nor protoconch of the type species of Euspira s.s. is known. Thus, a conchological analysis of N. glaucinoides and other taxa assigned to Euspira is needed to re-evaluate the taxonomic validity of Euspira.

Genera Sinum Röding, 1798 and Tectonatica Sacco, 1890

The two Sinum species analysed in this study are sister taxa in the tree based on the concatenated data set and in the 18S gene analysis (Fig. 2, Fig. S1). The separated placement of Sinum together with genera of the Polinicinae matches previous contentions that the Sininae more likely represent a genus within the Polinicinae (Finlay and Marwick 1937; Oyama 1969). However, the basal arrangement of Sinum species in both, the nuclear 18S tree and the mitochondrial COI tree, favour the hypothesis for the Sininae being a true subfamilial taxon independent of Polinicinae. This view appears to be contradicted by the basal arrangement of Tectonatica in the tree based on the concatenated data set, which, however, cannot be observed in either of the single analyses. Phylogenetic analyses of the entire family including more species and genera not investigated yet are needed to clarify these taxonomic puzzles.

Acknowledgements

We are grateful to Dr. Philippe Bouchet of the Muséum National d'Histoire Naturelle (MNHN, Paris, France) and Ian Loch of the Australian Museum (AM, Sydney, Australia) for providing ethanol-preserved specimens and for comments and help with taxonomic questions. We also thank Kathie Way and Amelia MacLellan for access granted to the mollusc collection of the Natural History Museum (NHM, London, UK) and for the permission to publish the photos taken of the syntypes of N. dubia Récluz, 1844, N. flemingiana Récluz, 1844 and N. jukesii, Reeve 1855. We also thank Dr. John Healy of the Queensland Museum (QM, Brisbane, Australia) and Dr. Brian Wilson of the Victoria Museum, (VM, Melbourne, Australia) for access to the malacological collection and seabed material, and for permission to publish the photos taken of the holotype of N. controversa Pritchard & Gatliff, 1913. We are grateful to Dr. Yves Finet (MHNG, Geneva, Switzerland), Dr. Wilma Blom (Auckland Museum, Auckland, New Zealand), Dr. Erica Sjöllin (Museum of Evolution, Uppsala University, Uppsala) and Prof. Ole S. Tendal (Natural History Museum of Denmark, Copenhagen) for permission to publish the photos of the syntypes of N. flemingianus Récluz, 1844, P. tawhitirahia Powell 1965, Nerita mammilla Linnaeus, 1758 and Mamma albula Chemnitz, 1758 [non-binomial], respectively. We thank Dr. Ralph Tollrian (Ruhr University Bochum, Germany) and Dr. Dustin Marshall (University of Queensland) for making their digital binocular microscopes available for taking pictures of protoconchs. We would like to thank Hugh Morrison of Kingsley, Western Australia, for providing ethanol-preserved specimens for sequence analyses and Dr. Lyle Vail and Dr. Anne Hogget of the Lizard Island Research Station, Queensland, Australia, for their kind help during collecting. We additionally thank Dr. Nico Michiels and Dr. Nils Anthes (University of Tuebingen) for the possibility to work on the Lizard Island Research Station. We would like to thank Annette Tolle and Björn Peters for expert sequencing. This paper is an output from the Great Barrier Reef Seabed Biodiversity Project, a collaboration between the Australian Institute of Marine Science (AIMS), the Commonwealth Scientific and Industrial Research Organisation (CSIRO), Queensland Department of Primary Industries & Fisheries (QDPIF) and the Queensland Museum (QM); funded by the CRC Reef Research Centre, the Fisheries Research and Development Corporation and the National Oceans Office led by R. Pitcher (Principal Investigator, CSIRO), P. Doherty (AIMS), J. Hooper (QM) and N. Gribble (QDPIF). We also wish to thank the field team led by D. Gledhill (CSIRO), the crew of the FRV Gwendoline May (QDPI & F) and RV Lady Basten (AIMS). Field and laboratory work were conducted in accordance with the Great Barrier Reef Marine Park Authority (GBRMPA) permit G05/16526.1 to T.H. and M.H. T.H. is funded by a grant from the Deutsche Forschungsgemeinschaft (Hu 1806/1-1, Hu 1806/2-1) and the Malacological Society of Australasia.

Supplementary material

13127_2012_111_Fig7_ESM.jpg (136 kb)
Fig. S1

Phylograms obtained through Bayesian inference based on the 16S and 28S gene fragments. Posterior probabilities are indicated at the nodes. Branches supported by values > 0.95 are not shown. Polytomies are due to the cut-off value specified for the consensus tree (50% used as the default value in MrBayes) (JPEG 136 kb)

13127_2012_111_MOESM1_ESM.tif (489 kb)
High resolution image (TIFF 488 kb)
13127_2012_111_Fig8_ESM.jpg (121 kb)
Fig. S1

Phylograms obtained through Bayesian inference based on the 16S and 28S gene fragments. Posterior probabilities are indicated at the nodes. Branches supported by values > 0.95 are not shown. Polytomies are due to the cut-off value specified for the consensus tree (50% used as the default value in MrBayes) (JPEG 136 kb)

13127_2012_111_MOESM2_ESM.tif (432 kb)
High resolution image (TIFF 432 kb)
13127_2012_111_Fig9_ESM.jpg (34 kb)
Fig. S1

Phylograms obtained through Bayesian inference based on the 16S and 28S gene fragments. Posterior probabilities are indicated at the nodes. Branches supported by values > 0.95 are not shown. Polytomies are due to the cut-off value specified for the consensus tree (50% used as the default value in MrBayes) (JPEG 136 kb)

13127_2012_111_MOESM3_ESM.tif (3.5 mb)
High resolution image (TIFF 3567 kb)
13127_2012_111_MOESM4_ESM.docx (95 kb)
Table S1Characters and character states analysed in the conchological analysis (DOCX 95 kb)
13127_2012_111_MOESM5_ESM.docx (130 kb)
Table S2Taxa of white Polinices species worldwide, with type information and literature reference to the original descriptions. Altogether, 55 taxa have been found in the literature and are listed alphabetically in three groups. Group 1: White Polinices taxa (21) from the Indo-Pacific region. Group 2: White Polinices taxa (10, plus 1 non-binomial name) with unknown type locality. Questionable type localities based on assumed species synonymy are marked with a ?. Group 3: White Polinices taxa (24) from regions outside the Indo-Pacific area. H holotype, S Syntype, pS possible syntype, P paratype, pP possible paratype, NT neotype, L lectotype, PL paralectotype, pPL possible paralectotype, Uk type unknown (DOCX 129 kb)

Copyright information

© Gesellschaft für Biologische Systematik 2012