Molecular Phylogenetics and Evolution 54 (2010) 783–809 Contents lists available at ScienceDirect Molecular Phylogenetics and Evolution journal homepage: www.elsevier.com/locate/ympev Molecular systematics of the marine gastropod families Trochidae and Calliostomatidae (Mollusca: Superfamily Trochoidea) S.T. Williams a,*, K.M. Donald b, H.G. Spencer b, T. Nakano c a Department of Zoology, The Natural History Museum, London SW7 5BD, UK Allan Wilson Centre for Molecular Ecology & Evolution, Department of Zoology, University of Otago, P.O. Box 56, Dunedin 9054, New Zealand c Department of Geology and Palaeontology, National Museum of Nature and Science, 3-23-1 Hyakunin-cho, Shinjuku-ku, Tokyo 169-0073, Japan b a r t i c l e i n f o Article history: Received 20 March 2009 Revised 14 August 2009 Accepted 10 November 2009 Available online 15 November 2009 Keywords: Systematics Diet Habitat Trochidae Calliostomatidae Evolution a b s t r a c t This study is the most extensive molecular study of the gastropod families Trochidae and Calliostomatidae published to date, in terms of both numbers of taxa and of gene sequences. As a result of Bayesian phylogenetic analyses of molecular sequence data from one nuclear gene and three mitochondrial genes, we propose dramatic changes to Trochidae family systematics, present the first molecular phylogeny for Calliostomatidae and include the first published sequence data for the enigmatic subfamily Thysanodontinae. Our phylogeny demonstrates that within the family Trochidae there is strong support for three subfamilies new to traditional classifications: Alcyninae subfam. nov., Fossarininae and Chrysostomatinae subfam. nov. As proposed, Alcyninae consists only of the nominotypical genus Alcyna, which is sister to all other trochids. The subfamily Fossarininae, as defined here, includes Fossarina, Broderipia, Synaptocochlea and ‘‘Roya” eximia and probably also Clydonochilus and Minopa. The subfamily Chrysostomatinae comprises the genera Chrysostoma and Chlorodiloma. Additional molecular support is also obtained for recently redefined Trochinae, Monodontinae, and Cantharidinae and for the traditionally recognised subfamilies Umboniinae and Stomatellinae. The subfamily Lirulariinae is not supported by the molecular data, but rather is incorporated into Umboniinae. We also demonstrate that the current concept of the subfamily Margaritinae (previously a trochid subfamily, but recently and provisionally assigned to Turbinidae) is not monophyletic. We provide preliminary evidence that whereas Margarella rosea (previously a member of Margaritinae) belongs in the trochid subfamily Cantharidinae, its presumptive congener M. antarctica is not a trochid, but instead clusters with the thysanodontine genus Carinastele. Based on the phylogenetic placement of C. kristelleae, we agree with previous proposals based on morphological data that Thysanodontinae are more closely related to Calliostomatidae than Trochidae. Both Calliostoma and Carinastele are carnivorous and if a sister relationship can be confirmed between Carinastele and Margarella antarctica it might mean that carnivory evolved twice in Trochoidea. The direction of dietary changes was not investigated in this study, but mapping diet onto the phylogeny suggests that true herbivory is predominantly a derived character. The new classification system also means that five trochid subfamilies are predominantly associated with hard substrata, one with soft substrata (Umboniinae) and two with algae and seagrass (Alcyninae and Cantharidinae). There has been a shift back to hard substrata in one umboniine clade. Two of three clades within Calliostomatidae were predominantly associated with hard substrata, but one Japanese clade is associated with sand. The finding of three new, unidentified species from very deep water means that Trochidae, like Calliostomatidae, now includes species found at bathyal depths. More deep-water species may be found as increased sampling leads to the discovery of new species. Crown Copyright Ó 2009 Published by Elsevier Inc. All rights reserved. 1. Introduction Trochoidea Rafinesque, 1815, is a highly diverse superfamily of marine gastropods, consisting of five families: Trochidae Rafinesque, 1815, Calliostomatidae Thiele, 1924, Turbinidae Rafinesque, * Corresponding author. Address: The Natural History Museum, Zoology Department, Cromwell Rd, London SW7 5BD, UK. Fax: +44 (0) 20 7942 5867. E-mail address: s.williams@nhm.ac.uk (S.T. Williams). 1815, Liotiidae Gray, 1850 and Solariellidae Powell, 1951. In this study we focus on two families – Trochidae and its likely sister taxon, Calliostomatidae (Williams et al., 2008). Of all the trochoidean families, Trochidae was thought to be the largest and most diverse in terms of diet and habitat (Hickman and McLean, 1990), but recent phylogenetic studies have suggested that some taxa traditionally thought to belong in Trochidae, are now excluded and these have been provisionally placed in Turbinidae or Seguenzioidea (Williams and Ozawa, 2006; Kano, 2008; Williams et al., 2008). 1055-7903/$ - see front matter Crown Copyright Ó 2009 Published by Elsevier Inc. All rights reserved. doi:10.1016/j.ympev.2009.11.008 784 S.T. Williams et al. / Molecular Phylogenetics and Evolution 54 (2010) 783–809 Nevertheless, as currently defined, the gastropod family Trochidae remains a large family including well in excess of 600 species and more than 60 genera (Appendix 1). Species are distributed throughout the Indian, Pacific and Atlantic Oceans, occurring in the tropical and temperate regions. They occur anywhere from the high intertidal to deep sea, although most species occur in shallow water. Species are commonly associated with hard substrata and are key taxa on many rocky shores and on coral reefs, where they are generally herbivores or detritivores. They are morphologically highly variable, with sizes ranging from minute (<0.5 cm) to very large (>20 cm). Trochid shells are renowned for their nacreous, pearly interior, and shells of some of the larger species are of commercial value, with the shell being used in the manufacture of buttons or as polished curios, or as inlay on laquerware. Calliostomatidae by contrast are a smaller family, with an estimated 250 species of medium size (Marshall, 1995a). Calliostomatids are unusual among vetigastropods in that they are carnivores, eating sessile invertebrates, predominantly cnidarians, especially hydroids, but also sponges and carrion. Other vetigastropod carnivores include key-hole limpets (Fissurellidae) that eat sponges and tunicates. Calliostomatids are found from shallow to deep-water (with many individual species occurring over this range of depths) and occur in all oceans, ranging from the tropics to polar latitudes. Although Trochoidea have been the recent focus of several morphological and molecular studies aiming to resolve family level relationships within the Vetigastropoda (e.g. Hickman and McLean, 1990; Hickman, 1996; Geiger and Thacker, 2005; Williams and Ozawa, 2006; Kano, 2008; Williams et al., 2008), the systematics of this group, particularly at the level of subfamilial relationships, are still in need of revision. An accurate phylogeny that correctly reflects systematic relationships among taxa is a crucial starting point for investigations into understanding how this large and very diverse group of marine molluscs has evolved. In this study we focus on the subfamilial systematics of the family Trochidae and present the first molecular phylogeny for Calliostomatidae. In order for molecular systematics to be considered a robust test of taxonomic classification, several points must be considered. It is important to include sufficient and appropriate outgroups, in order to adequately test for monophyly. In this respect, this study is able to focus on subfamily relationships as a result of earlier studies that redefined family level relationships within Trochoidea (e.g. Williams and Ozawa, 2006; Kano, 2008; Williams et al., 2008). Each taxonomic unit sampled (genus, subfamily, family, etc.) must be reciprocally monophyletic, and nest within the appropriate clade above. However, in order to test the concept of each taxonomic unit, the appropriate species must be included. Genera are defined by their type species, families by the type species of their nominotypical genus, and such taxa must thus be represented in genetic studies in order to be certain that the results obtained are typical of the genus or family. Genera (and other higher taxonomic units) must be monophyletic, but there are no accepted criteria (such as level of inter-specific genetic variation) that can be used to determine ‘cut-off’ points for inclusion or exclusion. We suggest that in order to define a genus, molecular analyses should recover a well-resolved clade including the type species, with a level of inter-specific genetic variation that falls within the range observed in other related genera and, ideally, morphological characters and/or biogeographic boundaries should distinguish it from other clades. For a family as large as Trochidae, the magnitude of the problem of obtaining a complete generic level revision is somewhat overwhelming. In this study we have sampled 110 trochid species from 42 trochid genera (Table 1) and aimed to include type species wherever possible, especially for nominotypical genera of families and subfamilies. We have also included many enigmatic and unusual taxa, not before included in genetic analyses. It was not our intention to test the validity of nominal species in this paper, although in some cases we have noted some discrepancies and paradoxes in species boundaries. 2. Materials and methods 2.1. Sampling and identification A total of 478 new sequences and 101 sequences from GenBank were analysed in this study. We obtained sequence from 114 nominal trochid species (sensu Williams et al., 2008) representing the subfamilies Stomatellinae Gray, 1840, Lirulariinae Hickman & McLean, 1990, Umboniinae H. & A. Adams, 1854, Trochinae Rafinesque, 1815, Cantharidinae Gray, 1857 and Monodontinae Gray, 1857. Previously published sequences were taken from Donald et al. (2005), Williams and Ozawa (2006) and Williams et al. (2008). The only potentially trochid subfamily not represented in this study was the subfamily Halistylinae Keen, 1958. Trochid sequences were analysed together with new and previously published sequences of species from the family Calliostomatidae (19 species in total), which was chosen as the outgroup on the basis of previous molecular studies showing it was most likely to be sister to Trochidae (Williams et al., 2008). We also included in this study new sequence from the type species of Carinastele Marshall, 1988, C. kristelleae, making this the first molecular study to include a representative of Thysanodontinae. Thysanodontinae have had a historically confused classification, and have in the past been included in Trochidae (Hickman and McLean, 1990), although more recently they were transferred to Calliostomatidae (Bouchet et al., 2005). Table 1 lists all species included in phylogenetic analyses, following the new subfamilial classification suggested by this study. New generic assignments suggested by this study have been listed in Table 1, but traditional names have been used in figures. A summary of the new classification and synonymies can be found in Appendix 1. All new Japanese samples were identified and sequenced by TN, all New Zealand species were identified by Bruce Marshall (Museum of New Zealand Te Papa Tongarewa) and HGS and sequenced by KMD and most others (except when used by KMD & HGS in previous studies) were identified and sequenced by STW (material from the Australian Museum was identified by Ian Loch; Japanese material from previous studies was identified by Tomowo Ozawa). Identifications were based primarily on Sasaki (2000), Poppe et al. (2006), Wilson (1993), Wells and Bryce (2000), Powell (1979) and Marshall (1995a, 1998a,c). Voucher specimens are kept at the Natural History Museum (London, UK), Museum of New Zealand Te Papa Tongarewa (Wellington, New Zealand), National Museum of Nature and Science (Tokyo, Japan), Muséum national d’Histoire Naturelle (Paris, France) and the Australian Museum (Sydney, Australia). Photos of most of the specimens used in this study are publicly available on MorphoBank at http://morphobank.geongrid.org/ permalink/?P223. The specimen Calliostoma consors (MorphoBank Number M24559) was incorrectly referred to as Calliostoma unicum in Williams et al. (2008). Previous GenBank entries have been updated to reflect this change. 2.2. DNA extraction, amplification and sequencing The DNA extraction and amplification protocols described by Williams and Ozawa (2006) were used to amplify portions of four genes: the nuclear 28S rRNA gene (28S: 1500 bp) and two mitochondrial genes: cytochrome oxidase subunit I (COI: 700 bp) and 16S rRNA (16S: 700 bp). A fragment of 12S rRNA (12S: 700 bp) was also amplified using the same PCR conditions as for the other Table 1 GenBank numbers for sequences from all species used in this study. Species are arranged alphabetically within families and subfamilies (as suggested by this study), but with traditional generic and specific names. New names or the need for taxonomic revision is indicated in a separate column. Registration numbers are given for voucher specimens are stored at the Natural History Museum, London (NHM), National Museum of Nature and Science, Tokyo, Japan (NSMT), Museum of New Zealand Te Papa Tongarewa, Wellington, New Zealand (NMNZ), Muséum national d’Histoire Naturelle, Paris (MNHN) and the Australian Museum, Sydney (AM). MorphoBank numbers refer to photos available online at http://morphobank.geongrid.org/permalink/?P223. Species New names/taxonomic comment Collecting locality Museum registration 28S sRNA Calliostoma antonii (Koch in Philippi, 1843) Veracruz, Pacific coast, Panama Calliostoma consors (Lischke, 1872) Calliostoma consors (Lischke, 1872) 1: Sugashima I., Mie Pref., Japan 2: Sugashima I., Mie Pref., Japan Calliostoma granti (Powell, 1931) Otago Harbour, New Zealand Calliostoma haliarchus (Melvill, 1889) Off Yagi, Iwate Pref., Japan Calliostoma iridium (Dall, 1896) Veracruz, Pacific coast, Panama Calliostoma javanicum (Lamarck, 1822) Bocas del Toro, Panama Calliostoma jujubinum (Gmelin, 1791) Bocas del Toro, Panama Calliostoma kiheiziebisu (Otsuka, 1939) Calliostoma ligatum (Gould, 1849) Off Kesennuma, Miyagi Pref., Japan Friday Harbour, Washington, USA Calliostoma multiliratum (Sowerby II, 1875) Horaizima, Otsuchi, Iwate Pref., Japan Calliostoma punctulatum (Martyn, 1784) 1: Cornwallis, Manukau Harbour, New Zealand 2: Cape Wanbrow, New Zealand NSMT Mo76813 NSMT Mo76812 NHM 20080936 – NHM 20080940 NMNZ M284057 NSMT Mo76814 NHM 20050731 NHM 20080937 NHM 20080938 – NHM 20080939 NSMT MoT76815 NMNZ M.287110 NMNZ M.287111 NSMT Mo76816 NMNZ M.287112 NSMT Mo76817 NMNZ M.287114 NHM 20080941 AB505225 AB505271 AB505316 AB505360 M24577 Calliostoma akoya (Kuroda in Ikebe, 1942) Off Zyogashima, Miura, Kanagawa Pref., Japan Off Hota, Chiba Pref., Japan Calliostomatidae Thiele, 1924 Subfamily Calliostomatinae, Thiele, 1924 Calliostoma aculeatum Sowerby III, 1912 Calliostoma shinagawaense Tokunaga, 1906 Off Zyogashima, Miura, Kanagawa Pref., Japan Whatipu, New Zealand Calliostoma tigris (Gmelin, 1791) Calliostoma unicum (Dunker, 1860) Calliostoma waikanae (Oliver, 1926) Tsuzizima, Amakusa–shi, Kumamoto Pref., Japan Snares Islands Calliostoma zizyphinum (Linnaeus, 1758) Ile Callot, Carantec, Brittany Subfamily Thysanodontinae (Marshall, 1988) Carinastele kristelleae (Marshall, 1988) Family status unresolved Margarella antarctica (Lamy, 1905) Trochidae Rafinesque, 1815 New subfamily Alcyninae Alcyna ocellata A. Adams, 1860 A 16S 12S MorphoBank ID AB505226 AB505272 AB505317 AB505361 M24576 GQ232378 GQ232352 GQ232281 GQ232313 M24557 EU530015 EU530118 – – – GQ232388 FN435323 GQ232289 GQ232324 M24559 GQ249711 – – GQ249802 M24761 AB505227 AB505273 AB505318 AB505362 M24578 GQ232380 GQ232354 GQ232283 GQ232316 M24521 EU530014 EU530117 – GQ232317 M24571 GQ232381 GQ232355 – GQ232318 M24552 AB505228 AB505274 AB505319 AB505363 M24582 GQ232383 GQ232357 GQ232285 GQ232320 M24511 AB505229 – AB505320 AB505364 M24579 GQ249708 – GQ249817 GQ249780 – – GQ249820 GQ249784 M24215 – AB505230 AB505275 AB505321 AB505365 M24580 GQ249709 – GQ249818 GQ249781 M24247 AB505231 AB505276 AB505322 AB505366 M24581 GQ249710 – GQ249821 GQ249785 M24248 EU530016 – – GQ232325 – S.T. Williams et al. / Molecular Phylogenetics and Evolution 54 (2010) 783–809 Calliostoma punctulatum (Martyn, 1784) COI Rank unresolved Snares Islands, New Zealand NMNZ M.287113 GQ249724 – – – Family status uncertain Generic status needs confirmation by inclusion of type species Galindez Island, Antarctica NHM 20080942 GQ232398 – – GQ232335 M24518 Cryptic species pair Nogamazima, Kumamoto Pref., Japan NSMT Mo76850 AB505232 AB505277 AB505323 AB505367 M24685 M24948 785 (continued on next page) 786 Table 1 (continued) Species New names/taxonomic comment Collecting locality Museum registration 28S sRNA Cryptic species pair Cryptic species pair Nogamazima, Kumamoto Pref., Japan Kewalo, Oahu, Hawaii – NHM 20080955 – AB505278 AB505324 AB505368 – GQ232376 GQ232350 GQ232279 – M24942 Subfamily Cantharidinae Gray, 1857 Cantharidella tesselata (A. Adams, 1851) Cantharidus tesselatus (A. Adams, 1851) Cantharidella tesselata (A. Adams, 1851) Cantharidus tesselatus (A. Adams, 1851) 1: Potato Point, Purakaunui Bay, New Zealand 2: Warrington, New Zealand Cantharidus bisbalteatus (Pilsbry, 1901) Generic status uncertain – not Cantharidus s.s. Cantharidus callichroa (Philippi, 1849) Generic status uncertain – not Cantharidus s.s. Cantharidus callichroa (Philippi, 1849) Cantharidus jessoensis (Scherenck, 1863) Generic status uncertain – not Cantharidus s.s. Generic status uncertain – not Cantharidus s.s. Alcyna ocellata A. Adams, 1860 A Alcyna ocellata A. Adams, 1860 B Cantharidus opalus (Martyn, 1784) Cantharidus purpureus (Gmelin, 1791) Cantharidus purpureus (Gmelin, 1791) Cantharidus striatus (Linnaeus, 1758) Generic status uncertain – not Cantharidus s.s. Diloma suavis (Phillippi, 1849) Pictodiloma suavis (Phillippi, 1849) Gibbula cineraria (Linnaeus, 1758) Gibbula fanulum (Gmelin, 1791) Gibbula magus (Linnaeus, 1758) Gibbula pennanti (Philippi, 1836) Generic status uncertain Gibbula rarilineata (Michaud, 1829) Generic status uncertain Gibbula umbilicalis (da Costa, 1778) Generic status uncertain Gibbula varia (Linnaeus, 1758) Jujubinus exasperatus (Pennant, 1777) Generic status uncertain Generic status uncertain Jujubinus suturalis (Adams, 1853) Kanekotrochus infuscatus (Gould, 1861) Generic status uncertain – not Jujubinus s.s Margarella rosea (Hutton, 1873) Cantharidus roseus (Hutton, 1873) Margarella rosea (Hutton, 1873) Micrelenchus capillaceus (Philippi, 1849) Micrelenchus capillaceus (Philippi, 1849) Micrelenchus dilatatus (Sowerby II, 1870) Micrelenchus dilatatus (Sowerby II, 1870) Micrelenchus huttonii (E.A. Smith, 1876) Cantharidus Cantharidus Cantharidus Cantharidus Cantharidus Cantharidus Micrelenchus huttonii (E.A. Smith, 1876) Cantharidus huttonii (A. Smith, 1876) Micrelenchus rufozona (A. Adams, 1853) Cantharidus rufozona (A. Adams, 1853) roseus (Hutton, 1873) capillaceus (Philippi, 1849) capillaceus (Philippi, 1849) dilatatus (Sowerby II, 1870) dilatatus (Sowerby II, 1870) huttonii (E.A. Smith, 1876) 16S 12S MorphoBank ID GQ434026 GQ434015 GQ434020 GQ434009 – GQ434028 GQ434017 GQ434022 GQ43401 M24265 AB505233 AB505279 AB505325 – M24196 AM048703 AM049338 AM048892 GQ232314 M24572 EU530017 EU530119 – – – AB505405 AB505280 AB505326 AB505369 M24197 GQ249739 GQ249696 GQ249830 GQ249799 M24249 GQ249725 GQ249679 GQ249806 GQ249757 – GQ249744 GQ249702 GQ249835 GQ249805 – – – GQ249836 – – GQ232387 GQ232361 GQ232288 GQ232323 M24543 AB505234 AB505281 AB505327 AB505370 M24219 AM048704 AM049339 AM048893 – M24522 GQ232391 GQ232363 GQ232293 GQ232328 M24525 GQ232392 GQ232364 GQ232294 GQ232329 – GQ232393 GQ232365 GQ232295 GQ232330 M24563 GQ232394 GQ232366 GQ232296 GQ232331 M24519 GQ232395 GQ232367 GQ232297 – – EF541180 EF541179 – – – GQ232396 GQ232368 GQ232298 GQ232333 M24561 EU530019 EU530121 GQ232299 – – AB505235 AB505282 AB505328 AB505371 M24198 GQ249726 – GQ249807 GQ249767 M24267 – GQ249742 GQ249743 GQ249727 – GQ434027 – GQ249834 – – GQ249809 GQ434021 GQ249682 GQ249699 – GQ249684 GQ249689 GQ434016 GQ434029 – – GQ249804 – – GQ249771 GQ434010 – – – – – M24268 GQ434023 GQ434012 – GQ249740 GQ249697 GQ249831 GQ249800 M24759 S.T. Williams et al. / Molecular Phylogenetics and Evolution 54 (2010) 783–809 Cantharidus opalus (Martyn, 1784) NMNZ M.288192 NMNZ M.288193 Nogamazima, Kumamoto Pref., Japan NSMT Mo76818 1: Omaezaki, Shizuoka Pref., Japan NHM 20050419 2: Manazuru, Kanagawa Pref., Japan – Otuchi, Iwate Pref., Japan NSMT Mo76819 1: Stewart I., New Zealand NMNZ M.287117 2: Stewart I., New Zealand – 1: Bay of Islands, New Zealand NMNZ M.287118 2: Bay of Islands, New Zealand – Ile Callot, Carantec, Brittany NHM 20080943 Susaki, Ogasawara Islands, Japan NSMT Mo76820 Wembury, Plymouth, UK NHM 20050420 Korcula I., Croatia NHM 20080378 Faro I., Algarve, Portugal NHM 20030329 Roscoff, France NHM 20080944 Lumbarda, Korcula I., Croatia NHM 20080375 Wembury, Plymouth, UK NHM 20080946 – – Lumbarda, Korcula I., Croatia NHM 20080387 Marble I., Duke Group, Queensland, Australia AM C419362 Katakuchihama, Kagoshima Pref., Japan NSMT Mo76821 1: Stewart Island, New Zealand NMNZ M.287119 2: Stewart Island, New Zealand – 1: Enderby Island, New Zealand – 2: Enderby Island, New Zealand – 1: Wellington, New Zealand – 2: Wellington, New Zealand – 1: Purakaunui Bay (rocks), New Zealand NMNZ M.288196 2: Purakaunui Bay (sandy rocks), New NMNZ Zealand M.288197 1: Tauranga, New Zealand NMNZ M284054 COI Micrelenchus rufozona (A. Adams, 1853) Cantharidus rufozona (A. Adams, 1853) 2: Urupukapuka Bay, New Zealand Micrelenchus sanguineus (Gray, 1843) Cantharidus sanguineus (Gray, 1843) 1: Warrington, New Zealand Micrelenchus sanguineus (Gray, 1843) Osilinus turbinatus (Born, 1778) Cantharidus sanguineus (Gray, 1843) 2: Warrington, New Zealand 1: Lumbarda, Korcula I., Croatia 2: Coral Bay Hotel, Cyprus.1 False Bay, South Africa Oxystele sinensis (Gmelin, 1791) East London, South Africa Oxystele tabularis (Krauss, 1848) East London, South Africa Oxystele tigrina (Anton, 1838) Cape Town, South Africa Oxystele variegata (Anton, 1838) Cape Town, South Africa Phasianotrochus irisodontes (Quoy & Gaimard, 1834) Dunsborough, Western Australia Prothalotia lehmanni (Menke, 1843) Dunsborough, Western Australia Scrobiculinus lepida (Philippi, 1846) Duke of Orleans Bay, Esperance, Western Australia Marukihama, Bounotsu, Kagoshima Pref., Japan Panglao, Philippines Thalotia attenuatus (Jonas, 1844) A Thalotia attenuatus (Jonas, 1844) B Tosatrochus attenuatus (Jonas, 1844) A; Cryptic species pair Tosatrochus attenuatus (Jonas, 1844) B; Cryptic species pair Thalotia conica (Gray, 1827) Unknown species ‘CP2203.SOL’ Unknown species ‘AT103_BC4864’ Dunsborough, Western Australia New genus and species (Vilvens, Warén and Williams, in preparation) New genus and species (Vilvens, Warén and Williams, in preparation) New subfamily Chrysostomatinae Chlorodiloma adelaidae (Philippi, 1849) Solomon Is. Big Bay, Vanuatu Adelaide, South Australia Chlorodiloma crinita (Philippi, 1849) Cape Leeuwin, Western Australia Chlorodiloma odontis (Wood, 1828) Chlorodiloma odontis (Wood, 1828) Swansea, Tasmania, Australia Edithburgh, South Australia Chlorodiloma odontis (Wood, 1828) Chrysostoma paradoxum (Born, 1778) Melbourrne, Victoria, Australia Santo, Vanuatu Subfamily Fossarininae (Bandel, 2009) Broderipia iridescens (Broderip, 1834) Marukihama, Bounotsu, Kagoshima Pref., Japan Benoki, Okinawa Pref., Japan 1: Green Point, Marrawah, Tasmania, Australia 2: Green Point, Marrawah, Tasmania, Australia 1: Auckland, New Zealand Broderipia iridescens (Broderip, 1834) Fossarina petterdi (Crosse, 1870) Fossarina petterdi (Crosse, 1870) Fossarina rimata (Hutton, 1884) Fossarina rimata (Hutton, 1884) ‘‘Roya” eximia (Nevill, 1869) Needs to be referred to new genus (Nakano & Marshall, in preparation) 2: Auckland, New Zealand 1: Shitiping, E. Taiwan MNHN 18153 NHM 20070155 MNHN 18160 MNHN 18222 NMNZ M.287132 NMNZ M.287133 – NMNZ M.287134 – MNHN 18152 GQ249741 GQ249698 GQ249832 GQ249801 – GQ249728 GQ249685 GQ249812 GQ249773 M24269 – GQ249686 GQ249813 – – GQ434031 GQ434019 GQ434025 GQ434014 – GQ434030 GQ434018 GQ434024 GQ434013 M33388 GQ249748 DQ061093 DQ061084 GQ249769 M24598 GQ249749 DQ061089 DQ061080 GQ249765 M24271 GQ249750 DQ061090 DQ061081 GQ249766 M24599 GQ249752 DQ061091 DQ061082 – M24600 GQ249751 DQ061092 DQ061083 GQ249770 M24601 EU530020 EU530122 – – M24530 EU530021 EU530123 GQ232302 GQ232338 M24536 GQ232402 GQ232371 GQ232305 GQ232341 M24943 AB505236 AB505283 AB505329 AB505372 M24199 GQ232377 GQ232351 GQ232280 – M24555 EU530022 FN435322 GQ232309 GQ232345 M24534 GQ232384 GQ232358 GQ232286 FN435321 M25184 FN435320 FN435319 FN435318 – M33710 GQ249712 AY858081 AY855317 GQ249759 M24509 GQ249713 AY858083 AY855319 GQ249760 M24584 GQ249714 – – – – – – M24586 GQ249788 – – AY858082 AY855318 – – GQ232386 GQ232360 GQ232287 GQ232322 M24509 S.T. Williams et al. / Molecular Phylogenetics and Evolution 54 (2010) 783–809 Osilinus turbinatus (Born, 1778) Oxystele impervia (Menke, 1843) NMNZ M284055 NMNZ M.287121 – NHM 20080408 – NMNZ M.287122 NMNZ M.287123 NMNZ M.287124 NMNZ M.287125 NMNZ M.287126 NHM 20070154/1, /2 NHM 20070156 NHM 20080950 – NSMT AB505237 AB505284 AB505330 AB505373 M24281 Mo76851 – EU530023 EU530139 – – – AM C458270 – GQ249690 GQ249824 GQ249793 – AM C458270 GQ249736 GQ249693 GQ249827 GQ249796 – NMNZ M.287151 – NSMT Mo76852 GQ249722 GQ249691 GQ249825 GQ249794 M24266 GQ249723 GQ249692 GQ249826 GQ249795 – AB505238 AB505285 – – M24282 787 (continued on next page) 788 Table 1 (continued) Species New names/taxonomic comment Collecting locality Museum registration 28S sRNA Needs to be referred to new genus (Nakano & Marshall, in preparation) 2: Shitiping, E. Taiwan – AB505239 AB505286 AB505331 AB505374 – Synaptocochlea asperulata (A. Adams, 1850) Santo, Vanuatu Synaptocochlea concinna (Gould, 1845) Hummocky I., Cape Capricorn, Queensland, Australia Benoki, Okinawa Pref., Japan MNHN GQ232399 – – 18159 AM C447162 EU530024 EU530140 – ‘‘Roya” eximia (Nevill, 1869) Synaptocochlea pulchella (A. Adams, 1850) Subfamily Monodontinae Gray, 1857 Austrocochlea brevis (Parsons & Ward, 1994) Pitt Water, Tasmania, Australia Porpoise Hole, Tasmania, Australia Austrocochlea diminuta (Hedley, 1912) Darwin, Northern Territory, Australia Austrocochlea porcata (Adams, 1853) Manly, New South Wales, Australia Austrocochlea rudis (Gray, 1826) Perth, Western Australia Diloma aethiops (Gmelin, 1791) Purakaunui Bay, New Zealand Diloma arida (Finlay, 1926) 1: Te Ngaru, New Zealand Diloma arida (Finlay, 1926) Diloma bicanaliculata (Dunker 1844, in Philippi, 1844) Diloma concamerata (Wood, 1828) 2: Purakaunui Bay, New Zealand Boulder Beach, Auckland, New Zealand Taranna, Tasmania, Australia Diloma constellatus (Souverbie in Souverbie & Austrocochlea constellatus (Souverbie in Souverbie Plum, SW New Caledonia Montrouzier, 1863) & Montrouzier, 1863) Diloma nigerrima (Gmelin, 1791) Concepcion, Chile Diloma piperinus (Philippi, 1849) Austrocochlea piperinus (Philippi, 1849) Benoki, Okinawa Pref., Japan Diloma radula (Philippi, 1849) Shitiping, Taitung, Taiwan Diloma subrostrata (Gray, 1835) Company Bay, Dunedin, New Zealand Diloma zelandica (Quoy & Gaimard, 1834) Diloma zelandica (Quoy & Gaimard, 1834) 1: Boulder Beach, Auckland, New Zealand 2: Warrington, New Zealand Monodonta australis (Lamarck, 1822) Isipingo, Kwa Zulu–Natal, South Africa Monodonta canalifera (Lamarck, 1816) A Monodonta canalifera (Lamarck, 1816) A Cryptic species pair Cryptic species pair 1: Katsuura, Kagoshima Pref., Japan 2: Okinawa, Japan Monodonta canalifera (Lamarck, 1816) B Cryptic species pair 1: Kagoshima, Japan Monodonta canalifera (Lamarck, 1816) B Monodonta labio (Linnaeus, 1758) Cryptic species pair 2: Kagoshima, Japan Darwin, Australia Monodonta perplexa perplexa (Pilsbry, 1889) 1: Toyama Monodonta perplexa perplexa (Pilsbry, 1889) 2: Hokkaido, Japan Subfamily Stomaetellinae Gray, 1840 Pseudostomatella decolorata (Gould, 1848) Pseudostomatella decolorata (Gould, 1848) 1: Benoki, Okinawa Pref., Japan 2: Benoki, Okinawa Pref., Japan 16S 12S MorphoBank ID – M24568 – – – AB505240 AB505287 AB505332 AB505375 M24283 NMNZ M.287127 NMNZ M.287128 NMNZ M.287129 NMNZ M.287130 NMNZ M.287131 NMNZ M.287135 NMNZ M.287137 – NMNZ M.287138 NMNZ M.287139 NHM 20070106 GQ249703 AY858088 AY855324 GQ249755 M24272 NSMT Mo76822 NSMT Mo76823 NMNZ M.287140 – NMNZ M.287141 NHM 20080947 – NMNZ M.287142 NMNZ M.287143 – NMNZ M.287144 NMNZ M.287145 NMNZ M.287147 – NSMT Mo76824 GQ249704 AY858086 AY855322 GQ249756 M24273 GQ249705 AY858084 AY855320 GQ249753 M24264 GQ249706 AY858087 AY855323 GQ249754 M24482 GQ249707 AY858085 AY855321 GQ249777 M24583 GQ249715 AY858100 AY855336 GQ249761 M24587 GQ249716 – – GQ249786 – – AY858098 AY855334 – M24588 GQ249717 AY858093 AY855329 GQ249762 M24590 GQ249718 AY858089 AY855325 GQ249758 M24591 EU530025 EU530126 GQ232290 – M24569 GQ249719 AY858095 AY855331 GQ249763 M24592 AB505241 AB505288 AB505333 AB505376 M24217 AB505242 AB505289 – AB505377 M24218 GQ249720 AY858099 AY855335 GQ249791 M24593 GQ249721 – – – – – AY858097 AY855333 GQ249790 M24594 EU530026 EU530127 – – – EU530027 EU530128 – – – DQ061095 DQ061087 – – M24595 GQ249732 – – – – – – – GQ249783 GQ249746 DQ061094 DQ061085 GQ249764 M24596 GQ249747 – – – GQ249787 – DQ061096 DQ061088 – M24597 AB505243 AB505290 AB505334 AB505378 M24275 AB505244 AB505291 AB505335 AB505379 M24276 S.T. Williams et al. / Molecular Phylogenetics and Evolution 54 (2010) 783–809 Austrocochlea constricta (Lamarck, 1822) COI Stomatella impertusa (Burrow, 1815) A Cryptic species pair Stomatella impertusa (Burrow, 1815) A Cryptic species pair Stomatella impertusa (Burrow, 1815) B Cryptic species pair Stomatella planulata (Lamarck, 1816) Stomatella planulata (Lamarck, 1816) Stomatia obscura Sowerby, 1874 Stomatia phymotis Helbling, 1779 Stomatolina angulata (A. Adams, 1850) Stomatolina rubra (Lamarck, 1822) Generic status uncertain MNHN 18154 2: Benoki, Okinawa Pref., Japan NSMT Mo76825 Esperance, Western Australia NHM 20080948 1: Teniya, Nago City, Okinawa Pref., Japan NHM 20080949 2: Marukihama, Bounotsu, Kagoshima Pref., NSMT Japan Mo76826 – Marukihama, Bounotus, Kagoshima Pref., Japan Santo, Vanuatu MNHN 18155 Marukihama, Bounotsu, Kagoshima Pref., – Japan Panglao, Philippines MNHN 18156 GQ232401 GQ232370 GQ232304 GQ232340 M24554 Hexham I., Queensland, Australia Tsuzizima, Kumamoto Pref., Japan EU530032 – – – – AB505249 AB505296 AB505340 AB505383 M24105 AM C447146 NSMT Mo76827 NHM 20080374 – NSMT Mo76828 NHM 20080372 – 1: Lumbarda, Korcula I., Croatia Clanculus cruciatus (Linnaeus, 1758) Clanculus denticulatus (Gray, 1826) 2: – Miyanohama, Ogasawara Islands, Japan Clanculus jussieui (Payraudeau, 1826) Lumbarda, Korcula I., Croatia Clanculus margaritarius (Philippi, 1849) Marukihama, Bounotsu, Kagoshima Pref., Japan Nogamazima, Kamiamakusa–shi, Kumamoto NSMT Pref., Japan Mo76829 1: Tsuzizima, Kumamoto Pref., Japan NSMT Mo76830 2: Koki Beach, Okinawa Pref., Japan – Lizard I., Queensland, Australia NHM 20080951 Dunsborough, Western Australia NHM 20070159 Port William, New Zealand NMNZ M.284056 1: Hawkes Bay, New Zealand NMNZ M.287146 2: Hawkes Bay, New Zealand – NMNZ 3: Auckland, New Zealand M.287148 4: Whatipu, New Zealand – Marukihama, Bounotsu, Kagoshima Pref., NSMT Japan Mo76831 S of Yate, SE New Caledonia NHM 20070143 1: Panglao, Philippines MNHN 18158 2: Marukihama, Bounotsu, Kagoshima Pref., NSMT Japan Mo76832 Tanjung Penawar, SE Johor, Malaysia NHM 20080658 Benoki, Okinawa Pref., Japan – 1: Eastern Beach, Auckland, New Zealand NMNZ M.287141 Clanculus microdon (A. Adams, 1853) Eurytrochus cognatus (Pilsbry, 1903) Generic status uncertain Eurytrochus cognatus (Pilsbry, 1903) Eurytrochus danieli (Crosse, 1862) Generic status uncertain Generic status uncertain Notogibbula preissiana (Philippi, 1848) Generic status uncertain Thoristella chathamensis (Hutton, 1873) Coelotrochus chathamensis (Hutton, 1873) Thoristella oppressa (Hutton, 1878) Coelotrochus oppressus (Hutton, 1878) Thoristella oppressa (Hutton, 1878) Thoristella oppressa (Hutton, 1878) Coelotrochus oppressus (Hutton, 1878) Coelotrochus oppressus (Hutton, 1878) Thoristella oppressa (Hutton, 1878) Trochus histrio (Reeve, 1848) Coelotrochus oppressus (Hutton, 1878) Trochus incrassatus (Lamarck, 1822) Trochus maculatus (Linnaeus, 1758) Trochus maculatus (Linneaus, 1758) Trochus radiatus (Gmelin,1791) Trochus stellatus (Gmelin, 1791) Trochus tiaratus (Quoy & Gaimard, 1834) Coelotrochus tiaratus (Quoy & Gaimard, 1834) M24277 GQ232400 GQ232369 GQ232303 GQ232339 M24944 EU530029 EU530130 GQ232307 GQ232343 M24514 AB505246 AB505293 AB505337 AB505380 M24278 AB505247 AB505294 AB505338 AB505381 M24279 GQ232403 GQ232372 GQ232306 GQ232342 M24558 AB505248 AB505295 AB505339 AB505382 M24280 GQ232404 GQ232373 GQ232308 GQ232344 M24562 GQ232379 GQ232353 GQ232282 GQ232315 M24524 – DQ093522 – AB505250 – – – – AB505384 M24106 GQ232382 GQ232356 GQ232284 GQ232319 M24516 AB505251 AB505297 AB505341 AB505385 M24078 AB505252 AB505298 AB505342 AB505386 M24107 AB505253 AB505299 AB505343 AB505387 M24109 EU530033 EU530133 – – – GQ232390 GQ232362 GQ232292 GQ232327 M24544 EU530028 EU530129 – GQ232337 M24533 GQ249738 GQ249695 GQ249829 GQ249798 M24760 – GQ249688 GQ249823 GQ249792 M24260 GQ249735 GQ249687 GQ249822 – – GQ249734 – – GQ249789 – AB505254 AB505300 – GQ249774 AB505388 M24110 GQ232405 GQ232374 GQ232310 GQ232346 M24520 GQ232406 GQ232375 GQ232311 GQ232347 M24527 AB505255 AB505301 AB505344 AB505389 M24111 GQ232407 – GQ232312 – EU530035 EU530135 – – – – M24940 – – GQ249778 M24603 (continued on next page) 789 Clanculus cruciatus (Linnaeus, 1758) AB505245 AB505292 AB505336 – S.T. Williams et al. / Molecular Phylogenetics and Evolution 54 (2010) 783–809 Subfamily Trochinae Rafinesque, 1815 Clanculus atropurpureus (Gould, 1849) Clanculus bronii (Dunker, 1860) 1: Santo, Vanuatu 790 Table 1 (continued) Species New names/taxonomic comment Collecting locality Museum registration 28S sRNA Trochus tiaratus (Quoy & Gaimard, 1834) Trochus viridis (Gmelin, 1791) Coelotrochus tiaratus (Quoy & Gaimard, 1834) Coelotrochus viridis (Gmelin, 1791) Coelotrochus viridis (Gmelin, 1791) New genus and species (Vilvens, Warén and Williams, in preparation) – NMNZ M.287150 – MNHN 18157 GQ249731 – GQ249733 – Trochus viridis (Gmelin, 1791) Unknown species ‘CP2466.BOA’ 2: Eastern Beach, Auckland, New Zealand 1: Ngawhiti Island, off Tata Beach, Tasman District, New Zealand 2: Snares Islands, New Zealand Vanuatu Generic status uncertain Manaduru, Kanagawa Pref., Japan AB505256 AB505302 AB505345 AB505390 M24486 Generic status uncertain Tateyama, Chiba Pref., Japan NSMT Mo76833 NSMT Mo76834 – NSMT Mo76835 NSMT Mo76836 NSMT Mo76837 NSMT Mo76838 NHM 20080952 NHM 20050421 NHM 20080953 NHM 20080953 NSMT Mo76840 NSMT Mo76841 AM C447174 Subfamily Umboniinae H. Adams & A. Adams, 1854 Conotalopia mustelina (Gould, 1861) Conotalopia ornata (Sowerby III, 1903) Tsubaki Onsen, Wakayama Pref., Japan 1: Kyoda, Okinawa Pref., Japan Ethaliella floccata (Sowerby III, 1903) 2: Kyoda, Okinawa Pref., Japan Ethminolia stearnsii (Pilsbry, 1895) 1: Tateyama, Chiba Pref., Japan Ethminolia stearnsii (Pilsbry, 1895) 2: Tateyama, Chiba Pref., Japan Ethminolia vitiliginea (Menke, 1843) Esperance, Western Australia Isanda coronata (H. & A. Adams, 1854) Karratha Back Beach, Western Australia Lirularia iridescens (Schrenck, 1863) Lirularia pygmaea (Yokoyama, 1922) 1: Monbetsu, Monbetsu City, Hokkaido, Japan 2: Monbetsu, Monbetsu City, Hokkaido, Japan 1: Tateyama, Chiba Pref., Japan Lirularia pygmaea (Yokoyama, 1922) 2: Tateyama, Chiba Pref., Japan Monilea lentiginosa Adams, 1851 Monilea smithi (Wood, 1828) Great Sandy Strait, SE of Urangan, Queensland, Australia 1: Kamae, Saiki, Oita Pref., Japan Monilea smithi (Wood, 1828) 2: Kamae, Saiki, Oita Pref., Japan Rossiteria nuclea (Philippi, 1849) 1: Katakuchihama, Kagoshima Pref., Japan Rossiteria nuclea (Philippi, 1849) 2: Katakuchihama, Kagoshima Pref., Japan Umbonium costatum (Valenciennes in Kiener, 1838) Umbonium giganteum (Lesson, 1833) Pusan, Korea 1: Hamamatsu, Shizuoka Pref., Japan Umbonium giganteum (Lesson, 1833) 2: Hamamatsu, Shizuoka Pref., Japan Umbonium moniliferum (Lamarck, 1822) 1: Youkakuwan, Kumamoto Pref., Japan Umbonium moniliferum (Lamarck, 1822) 2: Ako City, Hyogo Prefecture, Japan Zethalia zelandica (Hombron & Jacquinot, 1855) Zethalia zelandica (Hombron & Jacquinot, 1855) 1: Parengarenga Harbour, New Zealand Lirularia iridescens (Schrenck, 1863) 2: Tauranga, New Zealand NSMT Mo76842 NSMT Mo76843 NSMT Mo76844 NSMT Mo76845 NHM 20050422 NSMT Mo76846 NSMT Mo76847 NSMT Mo76849 NHM 20080954 NMNZ M.284059 NMNZ M.284058 16S 12S MorphoBank ID GQ249816 GQ249779 – GQ249819 GQ249782 M24604 – GQ249683 GQ249808 GQ249768 – GQ232385 GQ232359 – GQ232321 M24539 AB505257 AB505303 AB505346 AB505391 M24487 EU530036 EU530136 – – – AB505258 AB505304 AB505347 AB505392 M24488 AB505259 AB505305 AB505348 AB505393 – AB505260 AB505306 AB505349 AB505394 M24489 AB505261 AB505307 AB505350 AB505395 – GQ232389 – GQ232291 GQ232326 M24941 AM048705 AM049340 AM048894 GQ232332 M24523 GQ232397 EU530125 GQ232301 GQ232334 M24513 – EU530124 GQ232300 – M24513 AB505262 AB505308 AB505351 AB505396 M24490 AB505263 AB505309 AB505352 AB505397 – EU530038 EU530138 – GQ232336 – AB505264 AB505310 AB505353 AB505398 M24077 AB505265 AB505311 AB505354 AB505399 – AB505266 – AB505355 AB505400 – AB505267 AB505312 AB505356 AB505401 M24492 AM048706 AM049341 AM048895 GQ232348 M24546 AB505268 AB505313 AB505357 AB505402 M24493 AB505269 AB505314 AB505358 AB505403 – AB505270 AB505315 AB505359 AB505404 M24495 EU530039 – – GQ232349 M24510 GQ249737 GQ249694 GQ249828 GQ249797 – – GQ249700 GQ249833 GQ249803 M24758 S.T. Williams et al. / Molecular Phylogenetics and Evolution 54 (2010) 783–809 Ethalia guamensis (Quoy & Gaimard, 1834) Ethaliella floccata (Sowerby III, 1903) COI S.T. Williams et al. / Molecular Phylogenetics and Evolution 54 (2010) 783–809 three genes, but with an annealing temperature of 62 °C and magnesium chloride concentration of 2.75 mM. Sequence reactions were performed directly on purified PCR products using a BigDye Terminator v1.1 Cycle Sequencing Kit (Applied Biosystems) and run on an Applied Biosystems 3730 DNA Analyser automated capillary sequencer. Sequencing and PCR primers are listed in Williams and Ozawa (2006) except for 12S primers, which are from Oliverio and Mariottini (2001; 12SI) and Bandyopadhyay et al. (2008; 12S(-)). Alternative primers (Koufopanou et al., 1999) were sometimes used to amplify 16S for problematic taxa. All sequences have been deposited in GenBank (accession numbers in Table 1). Multiple 28S sequences differing in length and sequence were obtained from the Japanese Alcyna A. Adams, 1860 specimens. The shortest sequence, which was most similar to that obtained for the specimen of Alcyna from Hawaii, and those of other taxa, was used in this study. All Alcyna 28S sequences clustered together with the sequence used in this study in neighbour joining analyses, and thus our choice of sequence is unlikely to have affected phylogenetic relationships, but does suggest the presence of pseudogenes in this species. 2.3. Sequence analysis and phylogeny reconstruction Sequences were edited using Sequencher (v. 4.8, Gene Codes Corporation, Ann Arbor, MI). Alignment of COI was unambiguous, requiring no insertions and was done by eye in MacClade (v. 4.07 OSX; Maddison and Maddison, 2003), checking amino acid translations. Alignment of ribosomal genes was more complicated and these sequences were aligned using Clustal X (v. 2.0.9; Thompson et al., 1994, 1997) (with ‘delay divergent sequence’ set at 90– 95%, ‘gap-opening penalty’ = 20, ‘gap-extension penalty’ = 5; iterations after each step, or at the end of the alignment). Poorly aligned sites in rRNA alignments were identified using Gblocks (0.91b, Castresana, 2000) and removed from analyses. Parameters for Gblocks were set conservatively as: minimum number of sequences for a conserved position: 70% of total number of sequences; minimum number of sequences for a flanking position: 90% of the total number of sequences (or maximum number allowed, if lower); maximum number of contiguous non-conserved positions: 3; minimum length of a block: 5; and all gap positions allowed. After removal of ambiguous blocks, a total of 1412 bp of sequence from 28S remained to be used in phylogenetic analyses (95% of 1474 bp), 364 bp of sequence from 16S (57% of 636 bp) and 418 bp of sequence from 12S (52% of 789 bp). 2.4. Datasets Prior to undertaking any analyses, all 28S sequences from this study were analysed along with all 28S sequences from a previous study (Williams et al., 2008) including 12 vetigastropod families and two outgroup families in a neighbour-joining analysis in PAUP* (Swofford, 2002) to confirm that all taxa belonged in Trochidae or Calliostomatidae (tree not shown). Furthermore, an additional Bayesian analysis was performed using these same published 28S sequences from Williams et al. (2008), with the addition of the new 28S sequences for Margarella antarctica, Carinastele kristelleae, Alcyna ocellata and an unidentified specimen from Vanuatu (CP2466.BOA) from this study. This analysis allowed us to explore the relationship of the first two species to Calliostomatidae and to confirm that the latter two species belong in Trochidae (tree not shown). Phylogenies presented in this study were constructed using Bayesian methods implemented by MrBayes (v. 3.1.2, Huelsenbeck and Ronquist, 2001). Trees were obtained for each gene-sequence dataset (28S, Fig. 1; 12S, Fig. 2; 16S, Fig. 3 and COI, Fig. 4) and for combined-datasets, which included all species for which 28S 791 and 12S had been sequenced and all species for which all four genes were available (2-gene tree, 130 taxa, Fig. 5 and 4-gene tree, 110 taxa, Fig. 6). In some cases, gene sequences from different individuals (of the same species) were concatenated in order to obtain all four genes for a species. A second, larger 4-gene dataset (119 taxa) was also analysed using BEAST (v. 1.4.8, Drummond and Rambaut, 2007). Although this larger dataset included only a single representative of each species, it has more taxa because it included all taxa that had 28S sequence and at least two of the three mitochondrial-gene sequences (BEAST tree: Fig. 7). 2.5. Phylogeny reconstruction Models used in the Bayesian analyses were determined by MrModelTest (v. 2.1, J. Nylander, www.ebc.uu.se/systzoo/staff/ nylander.html). The COI dataset was further tested to see whether variation across codon positions would result in an improved likelihood. The best model for all datasets was determined to be GTR+G+I using the hierarchical likelihood-ratio tests. These models were used in all MrBayes analyses, with all parameters free to vary. In the combined analyses, variation was partitioned among genes and gene-specific model parameters were used (with all parameters free to vary independently within each partition). In addition, each gene was allowed to evolve at a different rate. The analysis for each dataset was run for 10,500,000 generations, with a sample frequency of 1000. The first 5501 trees were discarded, so that 5000 trees were accepted for each run after likelihood values had reached a plateau. The datasets were analysed in two independent runs, and the final tree was computed from the combination of accepted trees from each run (a total of 10,000 trees). Convergence between the two runs was tested by examining the potential scale reduction factors (PSRF) produced by the ‘sump’ command in MrBayes. All runs resulted in PSRF values of one for all parameters. Support for nodes was determined using posterior probabilities (PP, calculated by MrBayes). We also analysed the large dataset (119 taxa) using BEAST, to produce a phylogenetic hypothesis based on four genes (concatenated sequences from 28S, COI, 16S and 12S genes) using Bayesian inference with an uncorrelated relaxed, lognormal clock. This method allows for co-estimation of both phylogeny and divergence times and is thought to result in better trees than Bayesian analysis alone (Drummond et al., 2006). We did not, however, use any calibration dates, but rather fixed the mean rate of substitutions to one, as we were most interested in topology. The dataset included taxa that had 28S sequence and at least two of the three mitochondrial-gene sequences. One exemplar of each species was included, although in some cases it was necessary to concatenate sequences from different individuals of the same species. The Yule tree prior was used, which assumes a constant speciation rate among lineages, with lognormal prior for birth rate. Sequence variation was partitioned among genes and gene-specific model parameters were used with each gene allowed to evolve at a different rate. Models were chosen after several preliminary analyses (GTR+G+I for 16S and 12S, HKY+G+I for 28S and SRD06 for COI). On the basis of preliminary analyses, we changed Jeffries priors to lognormal priors for gtr-16S.cg, gtr-16S.ac and gtr-12S-cg. Based on MrBayes analyses we set the ingroup to monophyletic to ensure correct placement of the root. The analysis ran for 75,000,000 generations with sample frequency of 1000. All ESS values were greater than 1000 (some orders of magnitude greater). The final tree (Fig. 7) was a maximum clade credibility tree based on 67,500 trees (after burnin of 7500 generations) and support for nodes was determined using PP (calculated by BEAST). 792 S.T. Williams et al. / Molecular Phylogenetics and Evolution 54 (2010) 783–809 Fig. 1. Molecular phylogeny for Trochidae and Calliostomatidae based on single gene analysis of 28S rRNA produced by Bayesian analysis using MrBayes. Support values are posterior probabilities (PP); branches with PP < 50% were collapsed. Monophyletic trochid subfamilies are indicated by grey boxes, polyphyletic subfamilies are indicated by thick, vertical, grey lines. All trochid subfamilies are labelled on the right in grey text. Species names used in trees are ‘traditional’ names and do not reflect changes to taxonomy suggested by this study (see text and Table 1 for details). S.T. Williams et al. / Molecular Phylogenetics and Evolution 54 (2010) 783–809 793 Fig. 2. Molecular phylogeny for Trochidae and Calliostomatidae based on single gene analysis of 12S rRNA produced by Bayesian analysis using MrBayes. Support values are posterior probabilities (PP); branches with PP < 50% were collapsed. Monophyletic trochid subfamilies are indicated by grey boxes, polyphyletic subfamilies are indicated by thick, vertical, grey lines. All trochid subfamilies are labelled on the right in grey text. Species names used in trees are ‘traditional’ names and do not reflect changes to taxonomy suggested by this study (see text and Table 1 for details). 794 S.T. Williams et al. / Molecular Phylogenetics and Evolution 54 (2010) 783–809 Fig. 3. Molecular phylogeny for Trochidae and Calliostomatidae based on single gene analysis of 16S rRNA produced by Bayesian analysis using MrBayes. Support values are posterior probabilities (PP); branches with PP < 50% were collapsed. Monophyletic trochid subfamilies are indicated by grey boxes, polyphyletic subfamilies are indicated by thick, vertical, grey lines. All trochid subfamilies are labelled on the right in grey text. Species names used in trees are ‘traditional’ names and do not reflect changes to taxonomy suggested by this study (see text and Table 1 for details). S.T. Williams et al. / Molecular Phylogenetics and Evolution 54 (2010) 783–809 795 Fig. 4. Molecular phylogeny for Trochidae and Calliostomatidae based on single gene analysis of COI produced by Bayesian analysis using MrBayes. Support values are posterior probabilities (PP); branches with PP < 50% were collapsed. Monophyletic trochid subfamilies are indicated by grey boxes, polyphyletic subfamilies are indicated by thick, vertical, grey lines. All trochid subfamilies are labelled on the right in grey text. Species names used in trees are ‘traditional’ names and do not reflect changes to taxonomy suggested by this study (see text and Table 1 for details). 796 S.T. Williams et al. / Molecular Phylogenetics and Evolution 54 (2010) 783–809 Fig. 5. Molecular phylogeny for Trochidae and Calliostomatidae based on analysis of concatenated sequences from two genes (28S rRNA and 12S rRNA) produced by Bayesian analysis using MrBayes. Support values are posterior probabilities (PP); branches with PP < 50% were collapsed. Monophyletic trochid subfamilies are indicated by grey boxes, polyphyletic subfamilies are indicated by thick, vertical, grey lines. All trochid subfamilies are labelled on the right in grey text. The familial status of Margarella antarctica is uncertain. Species names used in trees are ‘traditional’ names and do not reflect changes to taxonomy suggested by this study (see text and Table 1 for details). S.T. Williams et al. / Molecular Phylogenetics and Evolution 54 (2010) 783–809 797 Fig. 6. Molecular phylogeny for Trochidae and Calliostomatidae based on analysis of concatenated sequences from four genes (28S rRNA, 12S rRNA, 16S rRNA and COI) produced by Bayesian analysis using MrBayes. Only taxa for which sequence from all four genes was available were included. Support values are posterior probabilities (PP); branches with PP < 50% were collapsed. All trochid subfamilies were recovered as monophyletic groups as indicated by grey boxes and are labelled on the right in grey text. Species names used in trees are ‘traditional’ names and do not reflect changes to taxonomy suggested by this study (see text and Table 1 for details). 798 S.T. Williams et al. / Molecular Phylogenetics and Evolution 54 (2010) 783–809 Fig. 7. Molecular phylogeny for Trochidae and Calliostomatidae based on analysis of concatenated sequences from four genes (28S rRNA, 12S rRNA, 16S rRNA and COI) produced by Bayesian analysis incorporating an uncorrelated relaxed, lognormal clock using BEAST. All taxa for which sequence was available for both 28S rRNA and at least two mitochondrial genes were included. Tree is a maximum clade credibility tree with median node height based on 67,500 trees. Support values are posterior probabilities (PP) (only shown where PP > 50%). All trochid subfamilies were recovered as monophyletic groups as indicated by grey boxes and are labelled on the right in grey text. No scale bar is provided because the root was arbitrarily set to one, and all other branch lengths are relative. Species names used in trees are ‘traditional’ names and do not reflect changes to taxonomy suggested by this study (see text and Table 1 for details). S.T. Williams et al. / Molecular Phylogenetics and Evolution 54 (2010) 783–809 3. Results and discussion This study is the most extensive molecular study published to date of the gastropod families Trochidae and Calliostomatidae, both in terms of number of taxa and of gene sequences. It includes many taxa at the level of species, genus, subfamily and even family for which there are no previous sequence data. This study in combination with a recent study by Williams et al. (2008) has resulted in dramatic changes to the systematics of the family Trochidae, especially at the subfamily level. As a result of this study, we propose two new trochid subfamilies (Alcyninae and Chrysostomatinae). We also find evidence to support recent redefinitions of three subfamilies (Trochinae, Cantharidinae, Monodontinae), two traditionally recognised subfamilies (Stomatellinae, Umboniinae) and one recently described family, with a change in rank (Fossarininae). On the basis of molecular data we no longer recognise Lirulariinae as a trochid subfamily as it is nested within Umboniinae and is here synonymized with the latter. We also show that the traditional concept of Margaritinae Thiele, 1924 (usually considered a trochid subfamily, but provisionally transferred to Turbinidae by Williams et al., 2008) is not monophyletic, with one of two species sampled (Margarella rosea) occurring in Trochidae and the other (M. antarctica) more closely related to Calliostomatidae. We include the first molecular sequence data for a member of the calliostomatid subfamily Thysanodontinae and show that it may be sister to M. antarctica. 3.1. Family Trochidae Rafinesque, 1815 The family Trochidae sensu Williams et al. (2008) is primarily a radiation of shallow-water species living intertidally or subtidally, although some species are found in deeper water (e.g. some South African Clanculus, Herbert, 1993) and rarely in bathyal depths (200 to 4000 m; e.g. unidentified species included in this study ‘CP2466.BOA’ was collected from 786 to 800 m and ‘CP2203.SOL’ was collected from 546 to 990 m; see below). Trochid species are predominantly tropical or sub-tropical, but do also occur in temperate oceans (e.g. Japan, New Zealand and southern Australia), often in association with hard substrata, but also infaunally and on macro-algae and seagrass. Diversity is highest in the Indo-West Pacific, but some species occur in the East Pacific, East Atlantic and Mediterranean. Only the subfamily Halistylinae occurs in the West Atlantic. All species are thought to be herbivores or detritivores (Herbert, 1993; Hickman and McLean, 1990). Seven clades were identified in the family Trochidae in our 4gene trees (Figs. 6 and 7). Five of these correspond to the subfamilies discussed in Williams et al. (2008): Trochinae, Monodontinae, Umboniinae, Stomatellinae and Cantharidinae. One corresponds to a new subfamily (Chrysostomatinae subfam. nov.) and the last corresponds to a recently described family, which we recognise at subfamilial rank (Fossarininae). Each of these seven subfamilies is discussed in detail below. One genus, Alcyna, did not fall into any of these subfamilies, but instead is sister to all other taxa. We have described a new subfamily for this genus, Alcyninae subfam. nov. The branching arrangement of subfamilies in Trochidae is pectinate, with monophyly of all subfamilies supported by PP = 100% in both 4-gene trees (Figs. 6 and 7; except Trochinae, Figs. 6 and 7: PP = 93%), and basal nodes supporting relationships between subfamilies were supported by PP P 96%, except for the sister relationship between Alcyninae and all other trochids, which has lower support (Figs. 6 and 7: PP = 94%, 77%). The relationship among subfamilies is: (((((((Cantharidinae, Stomatellinae) Umboniinae) Chrysostomatinae) Fossarininae) Monodontinae) Trochinae) Alcyninae). Each of these subfamilies is discussed in turn below, in the order that they appear in the phylogeny in Figs. 6 and 7. 799 3.1.1. New subfamily Alcyninae The genus Alcyna A. Adams, 1860 was long considered to be a subgenus of Thalotia Gray, 1847 but raised to full generic rank by Hickman and McLean (1990), and as such might be expected to belong in the subfamily Cantharidinae. Instead, our molecular data suggest that A. ocellata (the type species of Alcyna) is a basal member of the Trochidae that does not appear to belong in any previously defined subfamily, nor in the new subfamily described below (Figs. 6 and 7). Additional Bayesian analysis of 28S sequences from Alcyna with sequences from 12 vetigastropod families and two outgroup families (from Williams et al., 2008) confirm that this species is a trochid (tree not shown). We therefore propose that Alcyna should be referred to a new subfamily, Alcyninae (nominotypical genus Alcyna). Although Kuroda and Habe (1954; in Japanese) suggested that the radula of A. ocellata is ‘monodontine’, more recent studies suggest that the radula, external morphology and reproductive mode are all unique and unlike other trochoideans (C. Hickman, pers. comm., 2008). It is not yet known whether any other genera are likely to belong in this subfamily. The diagnosis for this subfamily is therefore a diagnosis of the genus Alcyna. Alcyna are minute (<5 mm), highspired, with convex whorls, and a closed umbilicus; the aperture is large, oval and smooth within, the outer lip simple, the ‘‘columella thickly calloused, bearing a nodular plait” (Wilson, 1993), and the shell lacks nacre. This genus is usually associated with algae or seagrass (Wilson, 1993; Sasaki, 2000). The genus occurs in the West Pacific and southern Australia. The two specimens used in this study were identified as A. ocellata, one from Japan (the type locality) and one from Hawaii. These two specimens are genetically distinct, have different shell colour patterns and likely represent different species (see MorphoBank M24685 and M24942). This genus is the subject of ongoing studies (C. Hickman, pers. comm., 2008). 3.1.2. Subfamily Trochinae Rafinesque, 1815 This subfamily was previously divided into three tribes by Hickman and McLean (1990): Trochini, Monodontini (equivalent to Gibbulini) and Cantharidini, but this study confirms previous molecular studies that showed the three are not collectively monophyletic and should be treated as distinct subfamilies, with some changes to their traditional compositions (Williams et al., 2008). This more restricted definition of the subfamily Trochinae is ‘‘a relatively well-defined entity with comparatively distinct shell characters”, being generally ‘‘larger [than species in Monodontinae and Cantharidinae], having a disjunct columella, four pairs of epipodial tentacles and a strongly papillate foot with hooded epipodial sense organs” (Herbert, 1993, 1998). The only extant genera traditionally referred to this subfamily on morphological grounds are Trochus Linnaeus, 1758 (along with its cold-water New Zealand subgenera Thorista Iredale, 1915 and Coelotrochus Fischer, 1879), Clanculus Montfort, 1810, Tectus Montfort, 1810 and Thoristella Iredale, 1915. Two other genera, Pulchrastele Iredale, 1929 and Rubritrochus Beck, 1995 are also provisionally placed in this subfamily on morphological grounds. The genus Tectus is not represented in this study, because although it shares conchologic, opercular, epipodial, and radular characters with Trochus, Williams et al. (2008) showed using molecular data that it does not belong in Trochidae; and may instead be a basal member of Turbinidae. Trochinae are monophyletic in both the 4-gene trees and the 28S tree, although with low support (Figs. 1, 6 and 7: PP = 62–93%). The Trochinae clade in our study includes the nominotypical genus Trochus (eight species, including type species, Tr. maculatus), Clanculus (seven species) and Thoristella (two species, including type species Th. chathamensis). It also includes Eurytrochus Fischer in Kiener, 1879 (two species, including type species E. danieli) and Notogibbula Iredale, 1924 (one species), which have traditionally been placed in Monodontinae, and a new, undescribed species (Fig. 7). Within 800 S.T. Williams et al. / Molecular Phylogenetics and Evolution 54 (2010) 783–809 Trochinae, tropical IWP Trochus are sister to a New Zealand radiation of Coelotrochus, Thoristella and Thorista. These together are sister to a more widely distributed Clanculus clade. The two species of the New Zealand genus Thoristella form a well-supported clade with the New Zealand Trochus (Thorista) viridis, and this clade is sister to the tropical Trochus species in the 4gene trees (Figs. 6 and 7). In the BEAST tree and the 2-gene tree, the tropical Trochus clade includes Tr. maculatus, Tr. histrio and Tr. incrassatus and the New Zealand clade includes Tr. (Thorista) viridis, Tr. (Coelotrochus) tiaratus, Thoristella chathamensis and Th. oppressa (Figs. 5 and 7). Marshall (1998a) commented on a ‘‘strong [morphological] similarity between adult Th. oppressa and juvenile Tr. viridis”. We propose that the New Zealand radiation, which can be distinguished morphologically from tropical Trochus by their smaller size and biogeographically by their exclusive distribution within New Zealand, should be treated as a single genus, distinct from Trochus. Three names are available: Thoristella Iredale, 1915, Coelotrochus Fischer, 1879 (type species Tr. tiaratus) and Thorista Iredale, 1915 (type species Tr. viridis); Coelotrochus is the oldest available and thus the valid name. Although currently used at generic rank, both Eurytrochus and Notogibbula have previously been regarded as subgenera of Gibbula Risso, 1826. Our trees show they are clearly unrelated to Gibbula (which belongs in the subfamily Cantharidinae; see below), although the type species of Notogibbula, N. bicarinata is needed to confirm that the sampled N. preissiana is representative. Eurytrochus danieli, E. cognatus and N. preissiana form a well-supported clade along with C. bronni, and these are sister to four species of Clanculus in the 4-gene tree (Fig. 7: Clanculus clade PP = 100%). The type species of Eurytrochus, E. danieli was originally described in Clanculus, so it might be best to refer Eurytrochus and N. preissiana to Clanculus. Alternatively, N. preissiana and C. bronni could be referred to the oldest alternative genus, Eurytrochus. Determination of the correct generic assignment for these taxa awaits the inclusion of more taxa, most notably N. bicarinata and the type species of Clanculus, C. pharonius. Four subgenera of Clanculus are represented in our trees: Clanculus Montfort, 1810 (represented by C. margaritarius), Mesoclanculus Iredale, 1924 (represented by C. microdon), Clanculopsis Monterosato, 1879 (represented by C. jussieui and type species C. cruciatus) and Eucheliclanculus Kuroda et al., 1971 (represented by type species C. bronni). The subfamily Trochinae also includes a new, undescribed species, collected from deep-water (786–800 m) off Vanuatu during the Muséum national d’Histoire Naturelle, Paris (MNHN) ‘BOA’ expedition (‘CP2466.BOA’ in trees; MorphoBank number M24539) (currently being described by Vilvens, Warén and Williams). Bayesian analysis of 28S sequences from the new species and 12 vetigastropod families and two outgroup families from Williams et al. (2008) confirm that this species is a trochid (tree not shown). We were unable to obtain 12S sequence for this species, so it is only represented in one figured combined-gene tree (the BEAST tree, Fig. 7). In this tree, the unknown species belongs in Trochinae and is sister to the Trochus/Coelotrochus clade (PP = 92%). 3.1.3. Subfamily Monodontinae Gray, 1857 Monodontinae are monophyletic in most analyses (Figs. 1 and 5–7: PP = 100%). The subfamily includes three genera, all of which are represented in our study: the nominotypical genus Monodonta Lamarck, 1799 (four species including type species, M. labio), Diloma Philippi, 1845 (11 species including type species, D. nigerrima), and Austrocochlea Fischer, 1885 (five species including type species, A. constricta). We move the traditionally monodontine genera Oxystele Philippi, 1847 and Osilinus Philippi, 1847 to Cantharidinae (see below). The genus Diloma, as traditionally described, is not monophyletic. Two species, D. constellatus and D. piperinus, are sister to the five Austrocochlea species in the large dataset 4-gene tree with high support (Fig. 7: PP = 100%), and should be included in this genus. The subgenus Pictodiloma Habe, 1946, type species D. suavis, belongs in the subfamily Cantharidinae and is here raised to generic rank. Both the genera Austrocochlea (with the inclusion of A. constellatus and A. piperinus) and Monodonta are monophyletic in the 4-gene trees. Cryptic species are evident within Monodonta canalifera based on our 28S tree (Fig. 1). Cryptic species have also been identified genetically within congeneric M. labio (Donald et al., 2005) and this group is the focus of ongoing studies by us all. Various taxa have now been excluded from Monodontinae. Cittarium pica, the only extant member of Cittarium Philippi, 1847, which was traditionally placed in this subfamily (Hickman and McLean, 1990), does not belong in Trochidae, but instead is likely a basal turbinid (Williams et al., 2008). Notogibbula was shown in a previous study to be likely to belong in Trochinae, based on the position of N. preissiana (Williams et al., 2008) and we confirm the position of N. preissiana in this study and also show that the genera Oxystele, Osilinus, Chrysostoma Swainson, 1840, Chlorodiloma Pilsbry, 1889 should be excluded from Monodontinae. The genera Margarella Thiele, 1893 Cantharidella Pilsbry, 1889 and Fossarina A. Adams & Angas, 1864 should also likely be excluded, although we have not confirmed this opinion with type species. As defined by this study, Monodontinae occur in both tropical and temperate Indo-Pacific habitats, in the intertidal and sublittoral on rocky substrates. Shells are predominantly imperforate; often have one or more columellar denticles and spiral ribs that are also apparent as apertural lirations on the inside of the shell. 3.1.4. Subfamily Fossarininae Bandel, 2009 The subfamily Fossarininae, forms a well-supported clade in most analyses (Figs. 1–3 and 5–7: PP = 82–100%). The Fossarininae clade includes four genera in our study: the nominotypical genus Fossarina (two species), Broderipia Gray, 1847 (one species), Synaptocochlea Pilsbry, 1890 (three species) and ‘‘Roya” eximia. Unfortunately we were unable to sample any type species from its component genera. Fossarina was formerly placed in Monodontinae, but Bandel (2009) recently erected a new family Fossarinidae based on F. mariei, which he placed within the superfamily Trochoidea. Although described as a new family, our data suggest that this clade is best recognized at the subfamily level. Other fossarinine genera Broderipia, Roya Iredale, 1912 and Synaptocochlea were previously classified in Stomatellinae (Hickman & McLean, 1990), although Hickman and McLean (1990) suggested that Synaptocochlea (along with Stomatolina Iredale, 1937 and Pseudostomatella Thiele, 1924) were problematic genera, including both stomatelline and ‘eucycline’ or ‘chilodontine’ species. As they included the type of Synaptocochlea, S. montrouzieri, in this group, and since S. concinna, which was included in this study, may prove to be a variant of S. montrouzieri, it is possible that some species in this group of ‘problematic’ taxa may actually belong in Fossarininae. The traditional position of Roya in Stomatellinae was on the basis of Iredale’s (1912) claim that its radula was similar to that of Gena Gray, 1842 (treated here as a subgenus of Stomatella Lamarck, 1816). More recently, Marshall (1981) transferred the genus from Trochidae to Siphonariidae Gray, 1827, as a synonym of Williamia Monterosato 1884, on the basis that shell characters (heterostrophic protoconch, shell morphology, colour and colour pattern) of the type species, R. kermadecensis, are indistinguishable from Williamia radiata nutata. He suggested it was likely that Iredale confused the radula sample with that of Stomatella oliveri, which Iredale examined at the same time. Sasaki (2000), however, placed a Japanese species, which he identified as ‘‘Roya” eximia and which we have sampled in this study (see MorphoBank M24282), within Trochidae on the grounds that it had epipodial tentacles and a S.T. Williams et al. / Molecular Phylogenetics and Evolution 54 (2010) 783–809 nacreous interior. Bruce Marshall (pers. comm., 2009) confirmed that this species, and another like it in New Caledonia, do not belong in Roya (=Williamia), but instead need to be referred to a new genus, a view supported by Hickman and McLean’s (1990) observation that the radula of ‘R.’ eximia is unlike that of B. iridescens and is unique among the Trochoidea. A new genus is being described by Nakano and Marshall. Herbert (1998) appears to be the first to suggest that Fossarina, Broderipia, and Synaptocochlea are closely related, but indicated that these taxa were too poorly known to be certain. He also included Clydonochilus Fischer, 1890 and Minopa Iredale, 1924, which are sometimes used as subgenera of Fossarina and are likely to be members of Fossarininae. Bandel (2009) included only Fossarina in his concept of Fossarininae, but also noted similarities in radular and protoconch characters between Fossarina and Broderipia. This subfamily was recovered in a molecular tree based only on Broderipia iridescens and Synaptocochlea concinna, but not named, in Williams et al. (2008). We note that the two species of Fossarina included in our work are only monophyletic in the BEAST tree (Fig. 7). If later studies with additional taxa show that the genus is not monophyletic, the subgeneric name Minos Hutton, 1884 could be raised to generic rank for its type species, F. rimata, if the type of Fossarina clusters with F. petterdi. All species in Fossarininae, as defined here, are small to minute (generally <1 cm, often <5 mm) and the nacreous layer is thin or apparently absent. The shells have few whorls (Synaptocochlea, Fossarina) or are limpet-like (‘‘Roya” eximia, Broderipia). Synaptocochlea and Fossarina have opercula, but ‘‘Roya” eximia and Broderipia do not (Sasaki, 1998). They do not have the ability of stomatellines to autotomise the metapodium (Hickman and McLean, 1990). Protoconch characters may prove to be useful in a diagnosis of Fossarininae (Bandel, 2009; D. Herbert, pers. comm. March 2009) and this subfamily is the focus of ongoing studies by the authors. Species from these genera are distributed throughout the tropical Indo-Pacific including Japan, southern Australia and New Zealand, as well as the Caribbean. 3.1.5. New subfamily Chrysostomatinae The new subfamily Chrysostomatinae, established here, is monophyletic in most analyses (Figs. 1, 2 and 4–7: PP: 90–100%). It includes only two genera in our study: the nominotypical genus Chrysostoma (represented by the type species, C. paradoxum) and Chlorodiloma (represented by three species, including type species, C. crinita). Donald et al. (2005) recently recognised Chlorodiloma as having generic rank, distinct from Austrocochlea. Both Chrysostoma and Chlorodiloma were long thought to belong in Monodontinae (Hickman and McLean, 1990; Donald et al., 2005), but Chrysostomatinae is one of the best-supported subfamilies in Trochidae in our study. Its composition remains uncertain, and more work is required to determine whether any other genera belong in this subfamily. Members of this subfamily typically have small (1–2 cm), smooth, low-spired, thick shells with green or orange columella. The aperture is large and smooth within; the outer lip is simple. The shell is nacreous and has a closed umbilicus. Chlorodiloma is endemic to Australia and Chrysostoma occurs in the central IWP and tropical Australia. 3.1.6. Subfamily Umboniinae H. & A. Adams, 1854 The subfamily Umboniinae is monophyletic in most analyses: (Figs. 1, 2 and 4–7: PP: 99–100%). It includes 10 genera in our study: the nominotypical genus Umbonium Link, 1807 (represented by three species), Ethalia H. Adams & A. Adams, 1854 (represented by type species, E. guamensis), Ethaliella Pilsbry, 1905 (represented by two species, including type species E. floccata), Zethalia Finlay, 1926 (represented by type species, Z. zelandica), Ethminolia Iredale, 1924 (two species), Isanda H. Adams & A. Adams, 1854 (repre- 801 sented by type species, I. coronata), Conotalopia Iredale, 1929 (two species), Monilea Swainson, 1840 (two species), Rossiteria Brazier, 1895 (represented by type species, R. nuclea) and Lirularia Dall, 1909 (two species). The inclusion of Lirularia, nested within Umboniinae shows that there is no molecular evidence to support the maintenance of a separate family for Lirulariinae. The type species, L. lirulata, was not included in this or previous studies, but molecular data for two species of the nominotypical genus Lirularia show that they belong in the subfamily Umboniinae. Morphological and behavioural evidence also links Lirularia with Umboniinae (filter feeders, monopectinate ctenidium with bursicles, snout tentacles and radular characters; McLean, 1986; Hickman and McLean, 1990; Herbert, 1992; Hickman, 1996). There is no molecular support for the separation of taxa into the tribes Umboniini H. Adams & A. Adams, 1854 and Talopiini Finlay, 1928 (=Monileini Hickman & McLean, 1990). Isandini Hickman, 2003 is represented only by a single taxon (Isanda coronata) and thus we are unable to comment on its validity. There are a significant number of putative genera that we were not able to sample (see Appendix 1) and, indeed, we were unable to include any representatives of the tribe Bankiviini Hickman & McLean, 1990. All genera with more than one species sampled are monophyletic in the combined-gene trees, with the exception of Conotalopia. The two sampled species of Conotalopia are not monophyletic in any tree, although they are paraphyletic in the 12S and COI trees (Figs. 2 and 4). Other genera not represented by multiple species in the combined-gene trees are monophyletic in single-gene trees, with the exception of Ethminolia, where three species are not monophyletic in the 28S tree. Ethminolia stearnsi is the type species of the subgenus Sericominolia Kuroda & Habe, 1954, which, if necessary, could be raised to generic rank. The majority of umboniine species live in relatively shallowwater, often in high-energy zones. They occur predominantly in tropical and sub-tropical Indo-West Pacific, but some species occur in temperate Australia and New Zealand (e.g. Zethalia) and Lirularia occurs in cold temperate to sub-tropical waters in northeast and northwest Pacific (Hickman and McLean, 1990; Herbert, 1992). Only one poorly known species, Monilea patricia (Philippi, 1851), occurs in the tropical East Pacific (Hickman and McLean, 1990). Lirularia, Umbonium and Bankivia Beck in Krauss, 1848 are filter feeders (Fretter, 1975; McLean, 1986; Hickman and McLean, 1990) and this mode of feeding is likely in other umboniine genera. Some genera, such as Ethalia and Ethminolia, are primarily depositfeeders and others may combine the two feeding modes (Hickman, 1985; Hickman and McLean, 1990; Herbert, 1992). When disturbed, some species exhibit a ‘swimming’ escape response, produced by thrashing movements of the foot (Hickman, 1985, 2003; Hickman and McLean, 1990; Herbert, 1992). 3.1.7. Subfamily Stomatellinae Gray, 1840 Members of Stomatellinae are distributed throughout the IndoPacific in temperate to tropical waters, ranging in depth from intertidal to shallow sublittoral on hard substrata. Shells are nacreous, low and flat (auriform) with a large aperture. Most auriform species lack an operculum, but some slightly higher-spired species have a small, vestigial operculum on the metapodium under the shell (Herbert, 1998). The foot has a highly extensible metapodium that cannot be withdrawn into the shell, but can be autotomised in response to disturbance and has the capability to regenerate (Hickman and McLean, 1990). The subfamily Stomatellinae is monophyletic in most analyses (Figs. 1, 2 and 5–7: PP = 52–100%). It includes four genera in our study: the nominotypical genus Stomatella Lamarck, 1816 (represented by two species, both belonging to the subgenus Gena Gray, 1850, including its type species S. planulata), Pseudostomatella (one 802 S.T. Williams et al. / Molecular Phylogenetics and Evolution 54 (2010) 783–809 species), Stomatia Helbling, 1779 (two species, including type species S. phymotis) and Stomatolina Iredale, 1937 (two species). All genera represented by more than one species are monophyletic in all trees. Cryptic species were recovered in Stomatella impertusa, with one specimen from Australia differing from two specimens from Japan and the Philippines, both genetically and in terms of colour pattern (see MorphoBank M24554, M24277 and M24944). Herbert (1998) suggested that Stomatellinae are closely allied to Cantharidinae and Gibbula (which also belongs in Cantharidinae sensu Williams et al., 2008) on the basis on a combination of characters including radular, protoconch, external anatomy and behaviour. This relationship between Stomatellinae and Cantharidinae is supported by our study. 3.1.8. Subfamily Cantharidinae Gray, 1857 The subfamily Cantharidinae is monophyletic in most analyses (Figs. 1, 2 and 5–7: PP = 75–100%). It is represented by species from 15 genera: the nominotypical genus Cantharidus Montfort, 1810 (represented by six species, including the type, C. opalus), Cantharidella Pilsbry, 1889 (one species), Gibbula Risso, 1826 (seven species, including type species G. magus), Jujubinus Monterosato, 1884 (two species, including type species J. exasperatus), Micrelenchus Finlay, 1926 (five species, including type species M. sanguineus), Oxystele (five species, including type species O. sinensis), Osilinus (represented by the type species O. turbinatus), Phasianotrochus Fischer, 1885 (one species), Prothalotia Thiele, 1930 (one species), Scrobiculinus Monterosato, 1889 (one species), Kanekotrochus Habe, 1958 (represented by type species K. infuscatus), Thalotia Gray, 1847 (two species, including type species T. conica), Margarella Thiele, 1893 (one species of two sequenced), as well as Diloma suavis and two unidentified specimens probably belonging to a single genus collected from deep-water off the Solomon Islands and Vanuatu (MNHN ‘Solomon 2’ and ‘Santo Marine Biodiversity Survey 2006’ expeditions). Traditionally Oxystele, Osilinus, Gibbula and Diloma suavis belong in Monodontinae. Donald et al. (2005) raised Oxystele, previously considered a subgenus of Diloma, to generic rank, a decision confirmed by this study. The position of Phorcus Risso, 1826 is unclear, as this genus was not included in our analysis, but morphological and allozyme studies suggest that it is related to Osilinus (Gofas and Jabaud, 1997) and as such it may also belong in Cantharidinae. This placement is supported by preliminary, unpublished molecular studies (J. Templado, pers. comm., 2009). Diloma suavis is the type species of a subgenus of Diloma, Pictodiloma Habe, 1946. We recommend that Pictodiloma be raised to generic rank and placed in Cantharidinae on the basis of this study. Williams et al. (2008) found molecular evidence to suggest Gibbula belonged in Cantharidinae, which was confirmed in this study. Thus Gibbulinae, often treated as a synonym of Monodontinae is, in fact, synonymous with Cantharidinae. The taxonomy of this group is in dire need of revision, as is evident from the fact that only one of the seven genera (Oxystele) represented by more than one species is monophyletic. Gibbula forms a moderately well-supported clade (Figs. 6 and 7: PP = 93%, 94%), but also includes Cantharidus striatus and the type species of Jujubinus, J. exasperatus and that of Osilinus, O. turbinatus. This topology suggests that Gibbula should probably be used to refer only to the clade containing the type species (G. magus), and the remaining Gibbula species could be given new generic status. We leave this decision for later studies involving more species from related genera, like Phorcus and more taxa from Gibbula; such a study is currently being undertaken in Spain (J. Templado, pers. comm., 2009). Species from the IWP attributed to Gibbula should probably be assigned to other genera (e.g. Agagus, Calliotrochus (Herbert, 1991, 1998) and Rubritrochus (Beck 1995)), some of which do not belong in Cantharidinae. Micrelenchus forms a clade along with Cantharidella tesselata, Cantharidus purpureus, Cantharidus opalus and Margarella rosea. As this clade includes both the type species of Cantharidus and Micrelenchus, and the genetic distances within this clade are similar to those of other trochid genera (e.g. Austrocochlea, Diloma), we propose that these taxa be referred to a single genus. Cantharidus predates Micrelenchus, therefore our definition of Cantharidus is a predominantly New Zealand radiation that excludes the genus Jujubinus, which has sometimes been used as a subgenus, and all Mediterranean (‘C.’ striatus) and Japanese taxa (‘C.’ bisbalteatus, ‘C.’ callichroa and ‘C.’ jessoensis) included in this study. Unexpectedly, Micrelenchus rufozona is sister to the rest of these New Zealand species and may be worthy of generic separation, but pending further data we retain it in Cantharidus. We were unable to sample the Australian type species of Cantharidella, C. picturata, or any other species from Australia, and so the position and synonymy of this genus remains unclear. Marshall (1998c) noted that all New Zealand ‘cantharidini’ (=Cantharidus, as defined here) have small to medium sized shells, with a small protoconch, with a narrowly pinched tip, all have three pairs of epipodial tentacles, most live on algae and under stones, but judging from gut contents, are detritivores. Marshall (1998c) also suggested that Australian and New Zealand cantharidines have been evolutionarily separated since at least the Early Eocene, when seafloor spreading in the Tasman Sea ceased, which might suggest that the Australian Cantharidella are not members of our concept of Cantharidus. Two unidentified specimens (two new species belonging to a single, new genus) collected from the MNHN expedition to the Solomon Islands and Vanuatu (‘CP2203.SOL’, ‘AT103_BC4864’; MorphoBank numbers: M25184, M33710) and Prothalotia lehmanni are sister to the New Zealand Cantharidus in the combined-gene trees (Figs. 6 and 7). The new species were collected from deepwater (546–990 m and 721–773 m, respectively) and are currently being described by Vilvens, Warén and Williams. Thalotia, as traditionally recognised, is not monophyletic in the 4-gene trees (Figs. 6 and 7). An alternative genus, Tosatrochus MacNeil, 1960 is available for T. attenuatus. Genetic sequence for COI as well as obvious morphological differences would suggest that we have two distinct species currently both identified as ‘Tosatrochus attenuatus’ (from the Philippines, MorphoBank number: M24555, identified from Poppe et al., 2006; and from Japan, MorphoBank number: M24199, identified from Sasaki, 2000; Fig. 4). ‘Cantharidus’ callichroa, ‘C’. jessoensis and ‘C’. bisbalteatus form a clade with Tosatrochus attenuatus, Pictodiloma suavis and Scrobiculinus lepida in the 4-gene trees (Figs. 6 and 7). Scrobiculinus lepida is closely allied with Phasianotrochus irisodontes in the 28S and COI trees (Figs. 1 and 4) and ‘Jujubinus’ suturalis is also a member of this clade in the 28S, 16S and BEAST trees (Figs. 1, 3 and 7). The generic assignments of ‘Cantharidus’ and ‘Jujubinus’ species in this clade are clearly in need of revision. The current concept of the genus Margarella is not monophyletic, with one species (M. rosea) in Cantharidinae and one (M. antarctica) sister to the thysanodontine species, Carinastele kristelleae. The genus Margarella was recently redescribed and on the basis of anatomy and behaviour has been placed within the ‘Gibbulini’ tribe of Trochinae (Zelaya, 2004). On the basis of shared behavioural characteristics, such as the ability to wrap the foot around algae, Margarella has been linked with ‘gibbuline’, cantharidine and trochine taxa (Hickman, 1996). New definitions of Cantharidinae and Monodontinae (sensu Williams et al., 2008 and this study) have resulted in the transfer of many taxa from Monodontinae (previously ‘Gibbulini’) to Cantharidinae. Therefore previous definitions based on morphological characters require updating. One possible radular character that might be useful for defining Cantharidinae is a reduced rachidian cusp and strong side denticles on the lateral tooth cusps, which was noted by Herbert (1998) in Margarella, Agagus, Jujubinus and to a lesser extent in Cantharidella and Thalotia. As currently defined, S.T. Williams et al. / Molecular Phylogenetics and Evolution 54 (2010) 783–809 this subfamily is predominantly a sub-tropical or temperate-water radiation. 3.1.9. Subfamily Halistylinae Keen, 1958 Halistylinae Keen, 1958 were provisionally retained in Trochidae by Williams et al. (2008) based on previous morphological studies (Hickman and McLean, 1990). We were unable to include any representatives of this group in our study to test this hypothesis. Some authors have been reluctant to assign this group to Trochidae (e.g. Ponder, 1985). 3.1.10. Subfamily Lirulariinae Hickman & McLean, 1990 See Umboniinae. 3.2. Turbinid subfamily Margaritinae Thiele, 1924 This subfamily is no longer considered to belong in Trochidae (Williams et al., 2008), nor is it monophyletic as currently defined. Previously it was divided into three tribes: Margaritini Thiele, 1924, Gazini Hickman & McLean, 1990 and Kaiparathinini Marshall, 1993. Recent molecular studies have shown that the nominotypical genus Margarites Gray, 1847 does not belong in Trochidae, but may instead belong in the family Turbinidae, although further work is needed to confirm its assignment (Williams et al., 2008). This study has shown that the current concept of the genus Margarella, another member of Margaritini, is not monophyletic. Neither of the two species included in this study is closely allied with Margarites; Margarella rosea is cantharidine and M. antarctica appears to be more closely allied with Calliostomatidae. The type species of Margarella, M. expansa, groups with M. antarctica on the basis of radula and soft tissue characters (Zelaya, 2004), suggesting that M. antarctica may be representative of the genus. If these relationships are correct, its Australasian relatives probably require a new generic name; Margarella rosea is here referred to Cantharidus, but the assignment of other species requires further work. Preliminary, unpublished molecular data for Gaza daedala Watson, 1879, the type species of Gaza Watson, 1879 suggest that this species is not a trochid, but that it is sister to Margarites as previously suggested by Hickman and McLean (1990) (STW, unpublished data). There are no published molecular data for any members of Kaiparathinini. 3.3. Turbinid subfamily Skeneinae Clark, 1851 The skeneimorph gastropods do not form a monophyletic group. It has been shown molecularly that some of these species belong in Seguenzioidea (those from the genera Adeuomphalus Seguenza, 1876, Ventsia Warén & Bouchet, 1993 and Xyloskenea Marshall, 1988; Kano et al., 2009), others in Turbinidae (Dillwynella Dall, 1889; Williams and Ozawa, 2006; Kano, 2008; Williams et al., 2008; and Cirsonella Angas, 1877, Bruceiella Warén & Bouchet, 1993, Lodderena Iredale, 1924; Kano, 2008). Some genera have been represented in molecular analyses, but their familial affinity is still not clear (Munditiella Kuroda & Habe, 1954; Williams and Ozawa, 2006; Kano, 2008; Williams et al., 2008). The nominotypical genus Skenea Fleming, 1825 has yet to be included in any study, but it is currently thought to belong in Turbinidae (Bouchet et al., 2005; Kano, 2008), making this a turbinid subfamily. It is possible that some skeneimorph taxa belong in Trochidae, but the subfamilial assignment of such species is uncertain. 3.4. Family Eucylidae Koken, 1896 The subfamily Eucylinae is no longer considered to belong in Trochidae or even Trochoidea. Instead molecular and morphological evidence shows that it belongs within the superfamily Segue- 803 nzioidea (Bouchet et al., 2005; Kano, 2008; Williams et al., 2008; Kano et al., 2009). This subfamily sensu Hickman and McLean (1990) is now divided into two groups at family-level rank (Eucylidae, which has no extant species and Chilodontidae; Bouchet et al., 2005; Kano et al., 2009). In total, six phylogenetic groups have been identified in Seguenzioidea: Seguenziidae, Chilodontidae, Calliotropidae, Cataegidae, Spinicalliotropis and ‘skeneimorph seguenzioids’ (Kano et al., 2009). Three of these families (Chilodontidae, Calliotropidae, Cataegidae) were previously thought to be subfamilies of, or closely allied with Trochidae until referred to Seguenzioidea by Bouchet et al. (2005). 3.5. Family Solariellidae Powell, 1951 The subfamily Solariellinae sensu Hickman and McLean (1990) is no longer considered to belong within Trochidae and was raised to family level status on the basis of molecular data by Williams et al. (2008). It is the focus of ongoing studies by STW. 3.6. Taxonomic issues at the species level Sequence variation based on COI distances (Kimura’s 1980 2parameter, K2P) and/or non-monophyly between terminal taxa suggests that some material identified as a single species, actually represents two species. These include Stomatella impertusa (COI K2P = 21%; Australia vs the Philippines and Japan), Alcyna ocellata (COI K2P = 15.8%; Hawaii vs Japan), Monodonta canalifera (no COI data for sp. B; 22 bp at 28S; both Japan), and Thalotia attenuatus (COI K2P = 9.4%; Japan vs Philippines). By contrast, four nominal species pairs showed very little genetic differentiation at the genes we sequenced: ‘Cantharidus’ bisbalteatus and ‘C.’ jessoensis (both Japan), Cantharidus huttonii and C. tesselata (previously Micrelenchus and Cantharidella; both New Zealand), Oxystele impervia and O. variegata (both South Africa), Austrocochlea rudis and A. constricta (both Australia). ‘Cantharidus’ bisbalteatus and ‘C.’ jessoensis are distinguishable primarily by shell colour and pattern and have almost allopatric distributions, overlapping only in the northern Sea of Japan. Sequences from COI are very similar (COI K2P < 0.5%) and from 28S are identical. Cantharidus huttonii and C. tesselata were once considered to be sufficiently distinct that they were placed into separate genera. These two species may occur sympatrically but can readily be distinguished by shell colour and pattern and presence/absence of an umbilicus. Sequences of COI are similar (COI K2P < 1.05%) and those of 28S are identical. Oxystele impervia and O. variegata were once considered to be a single species, but were separated on the basis of shell colour pattern, shape of the radular central tooth and allozyme differences (Heller and Dempster, 1991). Nevertheless, the mitochondrial genetic differences between Oxystele impervia and O. variegata are very small – COI sequence differs maximally by only 1.96% (in Donald et al., 2005) and are identical in this study. Sequences from 28S are also identical. Several allozyme loci show frequency shifts between these two nominal species, and one locus shows fixed differences (glycyl-leucine peptidase, Heller and Dempster, 1991). However, many allozyme markers are targets of selection (e.g. peptidase in Mytilus; Hilbish, 1985) and may reflect differences in diet, salinity or temperature stress perhaps as a result of the slightly different habitats they occupy. Notably, O. impervia occurs higher up the shore than O. variegata (Heller and Dempster, 1991). Austrocochlea rudis and A. constricta are distinguished by shell characters, primarily the presence of spiral cords and spire height and by colour pattern, including colour patterns that are unlikely to be ecophenotypic; A. rudis has a black margin to its outer lip and a large black smudge on the ventral side of the body whorl 804 S.T. Williams et al. / Molecular Phylogenetics and Evolution 54 (2010) 783–809 above the porcelain-white columellar thickening at the top of the aperture, neither of which occur in A. constricta (R. Willan, pers. comm., 2009). Sequences from COI are very similar (COI K2P < 0.5%) but the usually invariant (within species) 28S sequence differs by a single base. A third species, A. porcata, is also genetically (and morphologically) similar, but not as genetically close as A. rudis and A. constricta (COI K2P distance = 3.2–3.4%). Considered in the light of the marked genetic similarity at multiple genes (K2P COI differences in this study are all approximately 1% or less, whereas for 28S three of the pairs have identical sequences and the fourth pair differ by just 1 bp), some of the morphological differences between these taxon-groups do not seem to be compelling. Shells may grow with or without spines, depending on wave exposure (as demonstrated experimentally in Turbo cornutus by Ino, 1953); the presence or absence of ribs has been shown to be affected by environmental conditions (in Nucella emar ginata; Palmer, 1985); shell thickness depends on environment and exposure to predators (e.g. Littorina obtusata; Trussell, 1996) and both shell colour (e.g. Turbo cornutus; Ino, 1949) and radular characters (e.g. Littoraria spp.; Reid and Mak, 1999) can vary with diet. Nevertheless, we are unable to differentiate between very recent speciation, incipient speciation, hybridisation and ecological or population variants and our sampling is limited; therefore although it is likely that some of these pairs (in particular ‘Cantharidus’ bisbalteatus and ‘C.’ jessoensis and also Oxystele impervia and O. variegata) should be considered a single species, we suggest that further work is required before reaching a taxonomic decision. 3.7. Family Calliostomatidae Thiele, 1924 The family Calliostomatidae consists of about 250 extant species. It is distributed in all oceans, from the intertidal to bathyal depths (Marshall, 1995a; Hickman and McLean, 1990). Some species are found in association with algae, but stomach contents and field observations suggest that all (with one possible exception) are obligate carnivores as adults. Calliostomatids are typified by a protoconch with a reticulate sculpture and radular characters, including the presence of long, delicate, serrated rachidian (Hickman and McLean, 1990). Although Calliostomatidae were treated traditionally as a subfamily of Trochidae (e.g. Knight et al., 1960; Hickman and McLean, 1990) more recently they have been accorded familial rank (e.g. Marshall, 1995a,b; Bouchet et al., 2005) and this status was confirmed by recent molecular studies with three species (Williams et al., 2008). In this study, the split between Calliostomatidae and Trochidae is well supported in all analyses: (Figs. 1–7: PP = 99–100%, in some cases including Margarella antarctica and Carinastele kristelleae). Two subfamilies within Calliostomatidae are recognised in Bouchet et al. (2005): Calliostomatinae (with two tribes, Calliostomatini Thiele, 1924 and Fautricini Marshall, 1995) and Thysanodontinae. This study includes 18 calliostomatine species, all from the nominotypical genus Calliostoma, and one species from the subfamily Thysanodontinae Marshall, 1995 (discussed separately). Unfortunately we did not include the type species of Calliostoma, C. conulus, although we did include the species often thought to be the type species (C. zizyphinum) (type species designation discussed in Marshall, 1995b). Calliostoma zizyphinum, however, belongs to Calliostoma s.s. based on morphological grounds and is probably representative of the concept. The calliostomatine tribe Fautricini, Marshall, 1995 was not represented in this study. Three main clades are identified in Calliostoma in the BEAST tree (Fig. 7) and the 2-gene tree (Fig. 5) corresponding to a New Zealand/ European clade of four New Zealand species plus the C. zizyphinum from the western Europe, a clade of tropical American species from the Caribbean and tropical East Pacific and a clade with temperate Japanese and East Pacific species. Only two of these clades were recovered in the 4-gene tree based on taxa with complete sequences (Fig. 6), as COI could not be amplified for the New Zealand/European clade, therefore most of the discussion for this family refers to the BEAST tree and the 2-gene tree. Rankings within this family are uncertain, and, although we use subgeneric ranks within the genus Calliostoma, further work may support raising subgenera to generic rank and recognition of the clades as subfamilies. The temperate New Zealand/European clade appears to be a cold-water clade (Figs. 5 and 7: PP = 86%, 100%). The New Zealand clade includes the type species of Maurea (C. tigris), and thus corresponds to the New Zealand subgenus Maurea Oliver, 1926. Maurea is sister to C. zizyphinum (Fig. 5, PP = 86%), which represents the subgenus Calliostoma s.s. This subgenus is thought to include several other west European taxa, which were not sampled in this study (Ikebe, 1942; Marshall, 1995b). The tropical American clade includes four, tropical species collected from shallow-water (Figs. 5 and 7: PP = 100%). One of these species, Calliostoma jujubinum, is the type of the west Atlantic subgenus Elmerlinia, which Clench and Turner (1960) distinguished from Calliostoma s.s. on the basis of radular characteristics. Calliostoma javanicum also belongs in this subgenus. The remaining two species are East Pacific taxa and likely belong in different subgenera. The temperate Japanese/East Pacific clade includes species collected from deep-water (Figs. 5 and 7: PP = 100%). This clade includes the type species for Tristichotrochus Ikebe, 1942 (C. aculeatum), which was synonomised with Benthastelena Iredale, 1936 by Marshall (1995b). However, Marshall’s (1995b) concept of Benthastelena was of a group restricted to tropical and sub-tropical western Pacific. The fact that many of the species in this clade are distributed in temperate waters suggests that his concept may not apply. Moreover, this clade includes two species (C. haliarchus and C. unicum) that Marshall (1995b) suggested were worthy of (separate) subgeneric status, in which case Tristichotrochus may be an available name. Other available names for this clade, or subclades within it, are Akoya Habe, 1961 (type species C. akoya) and Otukaia Ikebe, 1942 (type species C. kiheiziebisu), both of which have been used at generic rank. Unfortunately, we were not able to sample the type of Benthastelena, C. katherina. 3.7.1. Subfamily Thysanodontinae Marshall, 1988 This subfamily was only recognised recently (Marshall, 1988) and includes three genera on the basis of morphological studies (Carinastele Marshall, 1988, Thysanodonta Marshall, 1988 and Herbertina Marshall, 1988). The subfamily Thysanodontinae was initially treated as a subfamily within Trochidae related to Calliostomatinae (Marshall, 1988), but more recently it has been treated as a subfamily within Calliostomatidae (Marshall, 1995b, 1998b; Bouchet et al., 2005). Our study would suggest that it is genetically more similar to Calliostomatidae than Trochidae, but has not been able to resolve its rank. We therefore continue to retain it within Calliostomatidae. Although we were not able to obtain any samples from the nominotypical genus, we were able to provide the first molecular data for any member of this family, Carinastele kristelleae, the type species of Carinastele. Limited sequences data (from 28S only) suggest that C. kristelleae is sister to Margarella antarctica (PP = 85%). As currently defined, Margarella is not monophyletic in our study, and it is possible that other species from this genus may be closely allied with M. antarctica, although the generic assignment of such species is not clear, as we have not included the type species M. expansa. In the 28S tree C. kristelleae and M. antarctica are sister to Calliostomatidae (not shown, node collapsed in Fig. 1: PP < 50%). We also included 28S sequences from M. antarctica and C. kristelleae in a Bayesian analysis with a previously published dataset includ- S.T. Williams et al. / Molecular Phylogenetics and Evolution 54 (2010) 783–809 ing 12 vetigastropod families and two outgroup families (from Williams et al., 2008). This tree (not shown) also showed that both species are sister taxa (with low support, PP = 63%) and are sister to Calliostomatidae (PP = 88%). They are not closely allied with Margarites, which does not belong in Trochidae, instead likely being a basal member of Turbinidae (Williams et al., 2008). The systematic relationship between C. kristelleae and M. antarctica is not clear from our study. The inclusion of M. antarctica in the Thysanodontinae seems highly unlikely as it is thought to be a herbivore, grazing on algae and does not have a radula consistent with the diagnosis for either Thysanodontinae or Calliostomatidae. Zelaya (2004) describes the first marginal tooth of M. antarctica as a ‘‘protolateromarginal plate”. Photos in Zelaya (2004) also show that all radular teeth are short and not typical of Thysanodontinae. It also lacks a reticulate pattern on its protoconch, but although typical of calliostomatids, some calliostomatids (e.g. Falsimargarita) do not have this character (A. Warén, pers. comm., 2009). Inclusion in the phylogeny of calliostomatid taxa, other than Calliostoma, may help to resolve these relationships. 3.8. Major shifts in habitat Members of the superfamily Trochoidea are often found in association with hard substrata, although species in some clades also occur infaunally and on macro-algae and seagrass. Previous traditional classifications meant that some subfamilies (e.g. the trochid subfamily Trochinae sensu Hickman & McLean, 1990) incorporated a wide range of habitat types, but based on the new classification of the family Trochidae, habitats are now generally fairly uniform within subfamilies, with five of the eight subfamilies (Trochinae, Monodontinae, Fossarininae, Chrysostomatinae and Stomatellinae) primarily associated with hard substrata. The subfamily Umboniinae is unique within the Trochidae, in that its members live infaunally in unconsolidated sediment with few exceptions (Herbert, 1992). All but three species in our study live on sand or mud intertidally, in some cases subtidally, to 20 or 30 m (McKnight, 1969; Higo et al., 1999). The three species that do not live infaunally (Lirularia iridescens, L. pygmaea and Conotalopia mustelina) form a well-supported clade in the 4-gene tree (Figs. 6 and 7: PP = 100%). Although Lirularia (which species is not indicated in the text) is capable of burrowing in unconsolidated sediment (Hickman and McLean, 1990), L. iridescens lives intertidally to 20 m on sea-plants (e.g. Zostera maxima, Phyllospadix iwatensis) (Higo et al., 1999), L. pygmaea lives intertidally to 20 m on rocks and algae (Higo et al., 1999; TN pers obs.) and Conotalopia mustelina lives intertidally on rocks and among algae (Higo et al., 1999; Sasaki, 2000). A second species of Conotalopia, C. ornata, lives on sand (Higo et al., 1999) and is not a member of this clade. McLean (1986) suggested that Lirularia might represent a step towards the specialisation in Umbonium, or a return to hard substratum of an infaunal umboniine. Our results suggest the latter. An association with seagrass or algae occurs in some Umboniinae (e.g. Lirularia and Conotalopia), Alcyninae and most Cantharidinae, suggesting that this specialisation has evolved at least three times in Trochidae. As currently defined, of these three subfamilies, Cantharidinae has the largest radiation of predominantly plantassociated species (Marshall, 1998c; Hickman, 2005). Molecular dating (Williams et al., 2008) and fossil evidence from extant genera suggest that Cantharidinae first diversified in the Upper Cretaceous (‘Jujubinus’ Monterosato, 1884, Turonian, 89–93.5 Myr and Gibbula Risso, 1826, Upper Cretaceous, 65.5–99.6 Myr; Knight et al., 1960). This timing coincides with the origins of seagrass and macrophyte dominated ecosystems, although red and brown algae have existed since before the Phanerozoic (Hickman, 2005). All species sampled in this study from the family Calliostomatidae occur in association with hard substrata, with the exception of 805 most of the species in the Japanese clade. All but two of the Japanese species sampled (C. akoya and C. multiratum, which occur on rocks), occur in deep-water on sand (Sasaki, 2000). These two species do not form a clade, suggesting that there have been two shifts back to hard substratum. Species of the New Zealand subgenus Maurea are predominantly associated with rocky or hard substrata, or with shell and bryozoan substrata and are distributed from the intertidal or shallow subtidal to deep-water (Marshall, 1995a). In our tree the only species sampled that is associated with shell and bryozoan substrata (C. waikanae) is in a derived position (Fig. 5). 3.9. Evolution of diet On the basis of fossil evidence, it has been shown that marine herbivory did not arise until the Mesozoic (Vermeij, 1977, 1987; Steneck, 1983). Vermeij and Lindberg (2000) argued on the basis of phylogenetic and paleontological evidence that marine herbivorous clades are highly derived and that microphagy and carnivory (of sessile invertebrates) are primitive modes of feeding among animals. They suggested that within the Vetigastropoda there are three clades with true herbivores: key-hole limpets (Fissurelloidea), haliotid abalones (Haliotidae) and the Trochoidea. Vermeij and Lindberg (2000) used a morphological phylogeny of trochoids (produced by Hickman, 1996) to map plant eating, and as a result suggested that herbivory evolved independently five times: in Turbinidae, Phasianellidae + Tricoliidae + Gabrielonidae and in the ‘trochid’ subfamilies Margaritinae, Tegulinae, and Trochinae. However recent molecular studies have dramatically changed the systematics of Trochoidea and led to a new phylogenetic hypothesis that in turn suggests new evolutionary patterns. Based on these studies the ‘pheasant shells’ (Phasianellidae + Tricoliidae + Gabrielonidae) are no longer considered to belong in Trochoidea, but rather belong in Phasianelloidea (Williams and Ozawa, 2006; Williams et al., 2008). Likewise, Trochinae has been split into three trochid subfamilies and both Margaritinae and Tegulinae have been removed from Trochidae and tentatively placed in Turbinidae (Tegulinae: Bouchet et al., 2005; Margaritinae and Tegulinae: Williams et al., 2008). Mapping herbivory onto this new phylogeny suggests that there are four main clades within Vetigastropoda with true herbivores and two within Trochoidea – of which only one is in Trochidae. Familial boundaries and subfamilial relationships within Turbinidae are still uncertain, but this study has produced a phylogeny that can be used to examine the evolution of herbivory within Trochidae. An important caveat to consider when examining diet, is that diet can change over time and seasons as well as opportunistically, and it is notoriously difficult to determine diet from gut contents as ingested items may not accurately reflect those items that are digested (C. Hickman, B. Marshall, J. Taylor, pers. comm., 2009). Further, we do not aim to produce here an exhaustive summary of all known data on trochoidean diet. Bearing these caveats in mind, however, trochids are thought either to be herbivores, feeding on algal films, diatoms and macro-epiphytes, or detritivores (Hickman and McLean, 1990). Many are primitive grazers that feed by rasping algae and small detritus off rocky surfaces. Others are true herbivores grazing directly on macro-algae. Filter-feeding occurs uniquely in the subfamily Umboniinae (Hickman and McLean, 1990; Herbert, 1992). Although Alcyna, and some umboniines (Lirularia and Conotalpoia) are also found in association with macro-algae or seagrass, the largest radiation of true herbivores occurs within Cantharidinae, although some subclades (e.g. Cantharidus, as defined here) are apparently detritivores (Marshall, 1998c). The cantharidine clade appears to be one of the most derived clades in the phylogeny, which is consistent with evolution from a more primitive feeding method. More basal clades like Trochinae and Monodontinae are found in association with hard substrata and species tend 806 S.T. Williams et al. / Molecular Phylogenetics and Evolution 54 (2010) 783–809 to be grazers or detritivores. Some primitive grazing on rocks occurs in species like Trochus maculatus, for which gut contents indicate a diet containing abundant fragments of algal species associated with rock and coral-rubble and filaments of blue-green algae (Taylor and Reid, 1984). Examination of other trochine genera and subgenera such as Clanculus and Infundibulops show gut contents include sediment and fine plant debris (Taylor and Reid, 1984), which may indicate a detritivorous diet. The position of Alcyna as sister to all other trochids is unexpected, if herbivory is indeed a derived character and if association with algae suggests herbivory, but the exact diet of this subfamily is not known so it is difficult to draw solid conclusions. In contrast, all species of Calliostomatidae investigated to date, either through examination of stomach contents or feeding observations, are carnivorous with the possible exception of Calliostoma ligatum. Most species eat sessile invertebrates including cnidarians (commonly hydroids, but also scleractinian corals and anemones), sponges and possibly bryozoans, but also carrion (fish and molluscs) and possibly polychaetes (Perron, 1975; Marshall, 1995a,b). Calliostoma ligatum is unusual in that its diet apparently consists of diatoms and detritus (Perron, 1975) and it has been observed grazing on Lamineria kelp (Chess, 1993). Cnidarians do occur on seagrass blades and it is difficult to identify cnidarians in gut contents (B. Marshall, pers. comm., 2009), so it is possible that this species is also a carnivore. If it is not, however, its position within Calliostoma is of interest: although relationships among the three clades are not well-resolved, if it has a basal position in the tree, carnivory in Calliostoma may have evolved from a diet of diatoms and detritus. Thysanodontines have not been observed feeding in the field, but Marshall (1988) argued convincingly that they are suctorial feeders, feeding on soft-bodied organisms on the basis of their extremely distinctive jaw and radula. Radular teeth are up to 1400 times as long as broad and are unique in the archaeogastropods in having ‘‘harpoon-like tips with backwards facing barbs” (Marshall, 1988). Marshall (1998b) found Thysanodonta wairua and Carinastele kristelleae associated with Stylasteridae (hydroids), which he presumed provided a food source. He suggested that the peculiar radular teeth might be used to break up and remove tissue from stylasterids (probably in conjunction with a protease). Herbert (1995) provided additional, circumstantial evidence in support of a carnivorous diet, stating that all his samples of Herbertina hayesi were found in crayfish traps, suggesting that they may have been attracted by bait. The apparent sister relationship shown in this study between C. kristelleae and M. antarctica (which needs to be confirmed with additional studies) is therefore extremely interesting. If this relationship is correct and these two are sister to Calliostomatidae then carnivory may have evolved twice within Trochoidea. polyphyletic as presently described. In this study we have highlighted several problem groups, without always being able to make any definite taxonomic recommendations. In some cases we were not able to meet all three requirements listed in the introduction and have left redefinition of these genera to later, more detailed studies. New generic assignments and taxonomic problems highlighted (but not resolved) by this study are indicated in Table 1. Acknowledgments We thank D. Herbert, B. Marshall, A. Warén, J. McLean, C. Hickman, J. Templado and one anonymous reviewer for making many useful comments that helped to improve the manuscript. We thank R. Willan, D. Herbert, A. Warén, C. Vilvens, T. Sasaki, J. Taylor and D. Reid for discussion about trochoidean taxonomy and systematics. We thank F. Fatih and P. Dyal for help in the lab. We especially thank P. Bouchet for providing many rare and interesting taxa and P. Maestrati for providing locality data. We are also very grateful to the following people for contributing or loaning specimens new to this paper or for their help in the field: C. Hickman, T. Ozawa, B. Marshall, C. Fraser, D. Reid, J. Taylor, M. Claremont, E. Glover, J. Jara, E. Gomez, D. Herbert, J. Hoffman, M. Malaquias, G. Williams, A. Moran, P. Moran, T. Haga, Y. Kano, P. Kuklinski, B. Chan, S. Higuchi, T. Hamada, H. Kinjo, N. Kaneko, S. Someya, M. Morley, A. Smith, R. Nikkula and S. Hills. This study was supported in part by a grant from the Natural Environment Research Council to SW (NE/C507453/1), a Grant-in-Aid for Scientific Research Project No. 207024 to TN from the Japan Society for Promotion of Science and funding from the Allan Wilson Centre for Molecular Ecology & Evolution to KD & HS. TN, KD and SW contributed equally to the sequencing and to data entry on MorphoBank. Photographs on MorphoBank were taken by KD and TN; we thank H. Taylor for photos of NHM specimens. SW did the analyses, prepared figures and tables and wrote the main text. HS and SW wrote the Appendix. All authors contributed to sample acquisition and editing and polishing the manuscript. Appendix 1. Summary of Generic Classification of Extant Trochidae The following lists the proposed classification of the Trochidae, down to the generic level. Several genera are omitted, as we cannot be confident of their subfamilial placement. For a similar list for the Calliostomatidae, see Marshall (1995b). Trochidae Rafinesque, 1815 4. Summary Alcyninae new subfamily The opportunity to greatly increase taxon sampling within Trochidae has resulted in dramatic changes to the systematics of this family. A list of included (and excluded) genera for each subfamily (as defined by molecular data, where available, otherwise following previous classifications) is given in Appendix 1. Although incomplete, this list acts as a starting point for the new classification scheme suggested here. New samples of Calliostoma and Carinastele, a thysanodontine, have enabled the first molecular investigation of relationships in the family Calliostomatidae. The new phylogeny provides a robust framework for more accurate taxonomic classification of Trochoidea, which is essential for understanding how this diverse group of marine gastropods evolved. This study has also highlighted problems with genus-level taxonomy, especially within Cantharidinae where many genera are Alcyna Adams, 1860 T (Type A. ocellata Adams, 1860) Cantharidinae Gray, 1857 (=Gibbulinae Stoliczka, 1868; new synonymy) Agagus Jousseaume, 1894 § (Type A. agagus Jousseaume, 1894) Calliotrochus Fischer, 1879 § (Type Turbo phasianellus Deshayes, 1863 = Margarita marmorea Pease, 1861) Calthalotia Iredale, 1929 § (Type Trochus arruensis Watson, 1880) Cantharidella Pilsbry, 1889  (Type Gibbula picturata Adams & Angas, 1864) Cantharidus Montfort, 1810 T (Type Trochus iris Gmelin, 1791 = Limax opalus Martyn, 1784) (=Micrelenchus Finlay, 1926 T [Type Trochus sanguineus Gray, 1843] = Plumbelenchus Finlay, 1926 T [Type Trochus capillaceus Philippi, 1848] = Mawhero S.T. Williams et al. / Molecular Phylogenetics and Evolution 54 (2010) 783–809 Marshall, 1998 T [Type Helix purpurea Gmelin, 1791]; new synonymy) Clelandella Winckworth, 1932 § (Type Trochus clelandi Wood, 1828) Gibbula Risso, 1826 T (Type Trochus magus Linnaeus, 1758) Jujubinus Monterosato, 1884 T (Type Trochus matoni Payraudeau, 1826 = Trochus exasperatus Pennant, 1777); possibly synonymous with Gibbula, further work is needed Kanekotrochus Habe, 1958 T (Type Trochus (Thalotia) yokohamensis Bock, 1878 = Zizyphinus infuscatus Gould, 1861) Komaitrochus Kuroda, 1958 § (Type K. pulcher Kuroda, 1958) Nanula Thiele, 1921 § (Type Gibbula tasmanica Petterd, 1879) Odontotrochus Fischer, 1880 § (Type Trochus chlorostomus Menke, 1843) Osilinus Philippi, 1847 § (Type Trochus turbinatus Born, 1778) Oxystele Philippi, 1847 T (Type Trochus sinensis Gmelin, 1791) Phasianotrochus Fischer, 1885  (Type Trochus badius Wood, 1828 = Bulimus eximius Perry, 1811) Phorcus Risso, 1826 § (Type P. margaritaceus Risso, 1826 = Monodonta richardi Payraudeau, 1826) Pictodiloma Habe, 1946 T (Type Trochus suavis Philippi, 1849); new rank Priotrochus Fischer, 1879 § (Type Trochus obscurus Wood, 1828) Prothalotia Thiele, 1930  (Type Trochus flindersi Fischer, 1878) Pseudotalopia Habe, 1961 ^ (Type P. sakuraii Habe, 1961) Scrobiculinus Monterosato, 1889  (Type Trochus strigosus Gmelin, 1791) (=Strigosella Sacco, 1896) Thalotia Gray, 1847 T (Type Trochus pictus Wood, 1828 = Monodonta conica Gray, 1827) Tosatrochus MacNeil, 1960 T (Type Thalotia aspera Kuroda & Habe, 1952 = Trochus attenuatus Jonas, 1844) Trochinella Iredale, 1937 § (Type T. perconfusa Iredale, 1937). Chrysostomatinae new subfamily Chlorodiloma Pilsbry, 1889 T (Type Monodonta crinita Philippi, 1849) Chrysostoma Swainson, 1840 T (Type Turbo nicobaricus Chemnitz, 1781 = Helix paradoxum Born, 1778). Fossarininae Bandel, 2009 new rank Broderipia Gray, 1847  (Type Scutella rosea Broderip, 1834) Clydonochilus Fischer, 1890 § (Type C. mariei Fischer, 1890) Fossarina A. Adams & Angas, 1864  (Type F. patula A. Adams & Angas, 1864) Minopa Iredale, 1924 § (Type Fossarina legrandi Petterd, 1879) Synaptocochlea Pilsbry, 1890  (Type Stomatella (Synaptocochlea) montrouzieri Pilsbry, 1890). Halistylinae Keen, 1958 Botelloides Strand, 1928 § (Type Onoba bassiana Hedley, 1911) Charisma Hedley, 1915 § (Type C. compacta Hedley, 1915) Halistylus Dall, 1890 § (Type Cantharidus (Halistylus) columna Dall, 1890). Monodontinae Gray, 1857 Austrocochlea Fischer, 1885 T (Type Monodonta constricta Lamarck, 1822) Diloma Philippi, 1845 T (Type Turbo nigerrimus Gmelin, 1791) Monodonta Lamarck, 1799 T (Type Trochus labio Linnaeus, 1758). Stomatellinae Gray, 1840 807 Microtis H. & A. Adams, 1850 § (Type M. tuberculata A. Adams, 1850) Pseudostomatella Thiele, 1924  (Type Stomatia papyracea Chemnitz, 1781 = Turbo papyraceus Gmelin, 1791) Stomatella Lamarck, 1816  (Type S. auricula Lamarck, 1816) Stomatia Helbling, 1779 T (Type S. phymotis Helbling, 1779) Stomatolina Iredale, 1937  (Type Stomatella rufescens Gray, 1847). Trochinae Rafinesque, 1815 Clanculus Monfort, 1810  (Type Trochus pharaonius Linnaeus, 1758) Coelotrochus Fischer, 1879 new rank T (Type Trochus tiaratus Quoy & Gaimard, 1834) (=Thorista Iredale, 1915 T [Type Polydonta tuberculata Gray, 1843 = Trochus viridis Gmelin, 1791] = Thoristella Iredale, 1915 T [Type Polydonta chathamensis Hutton, 1873]; new synonymy) Eurytrochus Fischer in Kiener, 1879 T (Type Clanculus danieli Crosse, 1862) Notogibbula Iredale, 1924  (Type Gibbula coxi Angas, 1867 = Gibbula bicarinata A. Adams, 1854) Pulchrastele Iredale, 1929 § (Type Calliostoma (Eutrochus) septenarium Melvill & Standen, 1899) Rubritrochus Beck, 1995 § (Type Gibbula pulcherrima A. Adams, 1855) Trochus Linnaeus, 1758 T (Type T. maculatus Linnaeus, 1758) Umboniinae H. & A. Adams, 1854 (=Lirulariinae Hickman & McLean, 1990) new synonymy Antisolarium Finlay, 1926 § (Type Solarium egenum Gould, 1849) Bankivia Beck in Krauss, 1848 § (Type Phasianella fasciata Menke, 1830) Camitia Adams & Adams, 1854 § (Type C. pulcherrima Adams & Adams, 1854) Conotalopia Iredale, 1929  (Type Monilea henniana Melville, 1891) Ethalia H. Adams & A. Adams, 1854 T (Type Rotella guamensis Quoy & Gaimard, 1834) Ethaliella Pilsbry, 1905 T (Type Ethalia floccata G.B. Sowerby III, 1903) Ethminolia Iredale, 1924  (Type E. probabilis Iredale, 1924) Isanda H. Adams & A. Adams, 1854 T (Type I. coronata A. Adams, 1854) Leiopyrga H. & A. Adams, 1863 § (Type Cantharidus lineolaris Gould, 1861) Lirularia Dall, 1909  (Type Margarites lirulata Carpenter, 1864) Monilea Swainson, 1840  (Type Trochus calliferus Lamarck, 1822) Parminolia Iredale, 1929 § (Type Monilea apicina Gould, 1861) Pseudominolia Herbert, 1992 § (Type Solariella splendens Sowerby, 1897) Rossiteria Brazier, 1895 T (Type Trochus nucleus Philippi, 1849) Umbonium Link, 1807  (Type Trochus vestiarius Linnaeus, 1758) Vanitrochus Iredale, 1929 § (Type Soleriella tragema Melvill & Standen, 1896) Zethalia Finlay, 1926 T (Type Rotella zelandica Hombron & Jacquinot, 1855). Genera of uncertain position Calumbonella Thiele, 1924 § (Type Gibbula gorgonarum P. Fischer, 1882) subfamily unresolved (Rolán et al., 2009) Hazuregyra Shikama, 1962 § (Type H. watanabei Shikama, 1962) 808 S.T. Williams et al. / Molecular Phylogenetics and Evolution 54 (2010) 783–809 Margarella Thiele, 1893  (Type Trochus expansus Sowerby, 1833) Family and subfamily status unresolved Tropidomarga Powell, 1951 § (Type T. biangulata Powell, 1951) Family and subfamily status unresolved. Excluded genera Chlorostoma Swainson, 1840, Cittarium Philippi, 1847, Margarites Gray, 1847, Norrisia Bayle, 1880, Tectus Montfort, 1810, Tegula Lesson, 1835 (not Trochidae, possibly basal members of Turbinidae; see Williams et al., 2008). Callistele Cotton & Godfrey, 1935 § (Type Astele calliston Verco, 1905) (possibly related to Chlorostoma; Marshall, 1995). Excluded subfamilies Eucyclinae Koken, 1897 sensu Hickman & McLean, 1990; including Agathodonta Cossmann, 1918, Bathybembix Crosse, 1893, Calliotropis Seguenza, 1903, Euchelus Philippi, 1847, Ginebis Otuka, 1942, Granata Cotton, 1957, Herpetopoma Pilsbry, 1889, Lischkeia Fischer in Kiener, 1879, Turcica A. Adams, 1854 (Seguenzioidea; see Kano, 2008; Williams et al., 2008; Kano et al., 2009) Margaritinae Thiele, 1924 = Margaritinae Stoliczka, 1868 (invalid; see Bouchet & Rocroi, 2005) (not Trochidae, possibly Turbinidae; see Williams et al., 2008) Solariellinae Powell, 1951 (Solariellidae; see Bouchet & Rocroi, 2005; Williams et al., 2008).  Type of genus not included in phylogeny and so placement and/or synonymy is provisional. T Type of genus included in phylogeny. ^ Based on preliminary unpublished genetic data (STW). § No species in genus included in phylogeny; Placement due to previous classifications and relationships. References Bandel, K., 2009. The slit bearing nacreous Archaeogastropoda of the Triassic tropical reefs in the St. Cassian Formation with evaluation of the taxonomic value of the selenizone. Berliner Paläontologische Abhandlungen 10, 5–47. Bandyopadhyay, P.K., Stevenson, B.J., Ownby, J.-P., Cady, M.T., Watkins, M., Olivera, B.M., 2008. The mitochondrial genome of Conus textile, coxI–coxII intergenic sequences and Conoidean evolution. Mol. Phylogenet. Evol. 46, 215–223. Beck, L.A., 1995. Rubritrochus, a new genus name for Gibbula pulcherrima A. Adams, 1855 and Gibbula declivis (Forskal, 1775) (Gastropoda: Prosobranchia: Trochidae). Arch. Molluskenkunde 124, 65–85. Bouchet, P., Rocroi, J.P., Fryda, J., Hausdorf, B., Ponder, W.F., Valdés, Á., Warén, A., 2005. Classification and nomenclator of gastropod families. Malacologia 47, 1–397. Castresana, J., 2000. Selection of conserved blocks from multiple alignments for their use in phylogenetic analysis. Mol. Biol. Evol. 17, 540–552. Chess, J.R., 1993. Effects of the stipe-boring amphipod Peramphithoe stypotrupetes (Corophioidea: Ampithoidae) and grazing gastropods on the kelp Laminaria setchellii. J. Crust. Biol. 13, 638–646. Clench, W.J., Turner, R.D., 1960. The genus Calliostoma in the Western Atlantic. Johnsonia 4, 1–80. Donald, K.M., Kennedy, M., Spencer, H.G., 2005. The phylogeny and taxonomy of austral monodontine topshells (Mollusca: Gastropoda: Trochidae), inferred from DNA sequences. Mol. Phylogenet. Evol. 37, 474–483. Drummond, A.J., Ho, S.Y.W., Phillips, M.J., Rambaut, A., 2006. Relaxed phylogenetics and dating with confidence. PLoS Biol. 4, e88. Drummond, A.J., Rambaut, A., 2007. BEAST: Bayesian evolutionary analysis by sampling trees. BMC Evol. Biol. 7, 214. Fretter, V., 1975. Umbonium vestiarium, a filter-feeding trochid. J. Zool. Lond. 177, 541–552. Geiger, D.L., Thacker, C.E., 2005. Molecular phylogeny of Vetigastropoda reveals non-monophyletic Scissurellidae, Trochoidea, and Fissurelloidea. Mollus. Res. 25, 47–55. Gofas, S., Jabaud, A., 1997. The relationships of the Mediterranean trochid gastropods ‘Monodonta’ mutabilis (Philippi, 1846) and ‘Gibbula’ richardi (Payraudeau, 1826). J. Mollus. Stud. 63, 57–64. Heller, J., Dempster, Y., 1991. Detection of 2 coexisting species of Oxystele (Gastropoda, Trochidae) by morphological and electrophoretic analysis. J. Zool. 223, 395–418. Herbert, D.G., 1991. A revision of the genus Agagus Jousseaume, 1894 (Mollusca: Gastropoda: Trochidae). J. Nat. Hist. 25, 883–900. Herbert, D.G., 1992. Revision of the Umboniinae (Mollusca: Prosobranchia: Trochidae) in southern Africa and Mozambique. Ann. Natal Mus. 33, 379– 459. Herbert, D.G., 1993. Revision of the Trochinae tribe Trochini (Gastropoda: Trochidae) of southern Africa. Ann. Natal Mus. 34, 239–308. Herbert, D.G., 1995. A new species of Thysanodontinae from South Africa (Mollusca: Trochoidea). Ann. Natal Mus. 36, 255–259. Herbert, D.G., 1998. Revision of the genus Calliotrochus Fischer, 1879 (Gastropoda: Trochoidea). Invert. Taxon. 12, 545–565. Hickman, C.S., 1985. Comparative morphology and ecology of free-living suspension feeding gastropods from Hong Kong. In: Morton, B., Dudgeon, D. (Eds.), The Malacofauna of Hong Kong and of Southern China. Hong Kong University Press, pp. 217–234. Hickman, C.S., 1996. Phylogeny and patterns of evolutionary radiation in trochoidean gastropods. In: Taylor, J.D. (Ed.), Origin and Evolutionary Radiation of the Mollusca. Oxford University Press, Oxford, pp. 177– 198. Hickman, C.S., 2003. Functional morphology and mode of life of Isanda coronata (Gastropoda: Trochidae) in an Australian macrotidal sandflat. In: Wells, F.E., Walker, D.I., Jones, D.S. (Eds.), The Marine Flora and Fauna of Dampier, Western Australia. Western Australian Museum, Perth. Hickman, C.S., 2005. Seagrass fauna of the temperate southern coast of Australia I: the cantharidine trochid gastropods. In: Wells, F.E., Walker, D.I., Kendrick, G.A. (Eds.), The Marine Flora and Fauna of Esperance, Western Australia. Western Australia Museum, Perth. Hickman, C.S., McLean, J.H., 1990. Systematic revision and suprageneric classification of trochacean gastropods. Nat. Hist. Mus. Los Angeles County Sci. Ser. 35, 1–169. Higo, S., Callomon, P., Goto, Y., 1999. Catalogue and Bibliography of the Marine Shell-bearing Mollusca of Japan. Elle Scientific Publications, Japan. Hilbish, T.J., 1985. Demographic and temporal structure of an allele frequency cline in the mussel Mytilus edulis. Mar. Biol. 86, 163–171. Huelsenbeck, J.P., Ronquist, F., 2001. MRBAYES: Bayesian inference of phylogenetic trees. Bioinformatics 17, 754–755. Ikebe, N., 1942. Trochid mollusca Calliostoma of Japan, fossil and recent. Jpn. J. Geol. Geogr. 18, 249–282. Ino, T., 1949. The effect of food on growth and coloration of the topshell (Turbo cornutus Solander). J. Mar. Res. 8, 1–5. Ino, T., 1953. Ecological studies of Turbo cornutus Solander – I. Changes of the spines on the shell due to their environments. Bull. Jpn. Soc. Sci. Fish. 19, 410– 414. Iredale, T., 1912. New generic names and new species of marine Mollusca. Proc. Malac. Soc. 10, 217–229. Kano, Y., 2008. Vetigastropod phylogeny and a new concept of Seguenzioidea: independent evolution of copulatory organs in the deep-sea habitats. Zool. Scr. 33, 593–594. Kano, Y., Chikyu, E., Warén, A., 2009. Morphological, ecological and molecular characterisation of the enigmatic planispiral snail Adeuomphalus (Vetigastropoda: Seguenzioidea). J. Mollus. Stud. 75, 397–418. Kimura, M., 1980. A simple method for estimating evolutionary rate of base substitutions through comparative studies of nucleotide sequences. J. Mol. Evol. 16, 111–120. Knight, J.B., Cox, L.R., Keen, A.M., Batten, R.L., Yochelson, E.L., Robertson, R., 1960. Archeogastropoda and some (mainly Paleozoic) Caenogastropoda and Opisthobranchia (Mollusca: Gastropoda). In: Moore, R.C. (Ed.), Treatise on Invertebrate Paleontology. Part I. Mollusca 1. Geological Society of America and Kansas University Press, Lawrence, Kansas, pp. 1–351. Koufopanou, V., Reid, D.G., Ridgway, S.A., Thomas, R.H., 1999. A molecular phylogeny of the patellid limpets (Gastropoda: Patellidae) and its implications for the origins of their antitropical distribution. Mol. Phylogenet. Evol. 11, 138–156. Kuroda, T., Habe, T., 1954. New genera of Japanese marine gastropods. Venus 18, 84–97. Maddison, D.R., Maddison, W.P., 2003. MacClade. Version 4.06 OSX. Sinauer Associates, Sunderland, MA. Marshall, B.A., 1981. The genus Williamia in the Western Pacific Mollusca Siphonariidae. N. Z. J. Zool. 8, 487–492. Marshall, B.A., 1988. Thysanodontinae: a new subfamily of the Trochidae (Gastropoda). J. Mollus. Stud. 54, 215–229. Marshall, B.A., 1995a. A revision of the recent Calliostoma species of New Zealand (Mollusca: Gastropoda: Trochoidea). Nautilus 108, 83–127. Marshall, B.A., 1995b. Calliostomatidae (Gastropoda: Trochoidea) from New Caledonia, the Loyalty Islands, and the northern Lord Howe Rise. In: Bouchet, P. (Ed.), Resultats des Campagnes MUSORSTOM, vol. 14. Mém. Mus. Natn. Hist. Nat. 167, 381–458. Marshall, B.A., 1998a. A review of the recent Trochini of New Zealand (Mollusca: Gastropoda: Trochidae). Mollus. Res. 19, 73–106. Marshall, B.A., 1998b. Food and feeding mode of Thysanodontinae (Mollusca: Gastropoda: Calliostomatidae). Mollus. Res. 19, 69–72. Marshall, B.A., 1998c. The New Zealand recent species of Cantharidus Montfort, 1810 and Micrelenchus Finlay, 1926 (Mollusca: Gastropoda: Trochidae). Mollus. Res. 19, 107–156. McKnight, D.G., 1969. Infaunal benthic communities of the New Zealand continental shelf. N. Z. J. Mar. Freshwat. Res. 3, 409–444. S.T. Williams et al. / Molecular Phylogenetics and Evolution 54 (2010) 783–809 McLean, J.H., 1986. The trochid genus Lirularia Dall, 1909: a filter-feeder? Am. Malac. Bull. 4, 109. Oliverio, M., Mariottini, P., 2001. A molecular framework for the phylogeny of Coralliophila and related muricoids. J. Mollus. Stud. 67, 215–224. Palmer, R., 1985. Quantum changes in gastropod shell morphology need not reflect speciation. Evolution 39, 699–705. Perron, F., 1975. Carnivorous Calliostoma (Prosobranchia: Trochidae) from the Northeastern Pacific. Veliger 18, 52–54. Ponder, W.F., 1985. A revision of the genus Botelloides (Mollusca: Gastropoda: Trochacea). South Australian Department of Mines and Energy Special Publication 5, pp. 301–327. Poppe, G.T., Tagaro, S.P., Dekker, H., 2006. The Seguenziidae, Chilodontidae, Trochidae, Calliostomatidae and Solariellidae of the Philippine Islands. Visaya. Conchology Inc. (Supplement 2, Occasional Publications). Powell, A.W.B., 1979. New Zealand Mollusca: Marine, Land and Freshwater Shells. Collins, Auckland. Reid, D.G., Mak, Y.-M., 1999. Indirect evidence for ecophenotypic plasticity in radular dentition of Littoraria species (Gastropoda: Littorinidae). J. Mollus. Stud. 65, 355–370. Rolán, E., Gonzalez-Porto, M., de Matos-Pita, S.S., 2009. The genus Callumbonella (Gastropoda, Trochacea) with the description of a new species from Namibia. J. Conchol. 39, 643–657. Sasaki, T., 1998. Comparative anatomy and phylogeny of the recent Archaeogastropoda (Mollusca: Gastropoda). The University Museum, The University of Tokyo, Bulletin No. 38, pp. 1–224. Sasaki, T., 2000. Trochidae. In: Okutani, T. (Ed.), Marine Mollusks in Japan. Tokai University Press, Tokyo, pp. 55–83. Steneck, R.S., 1983. Escalating herbivory and resulting adaptive trends in calcareous algal crusts. Paleobiology 9, 44–61. Swofford, D.L., 2002. PAUP*. Phylogenetic Analysis Using Parsimony (*and Other Methods). Sinauer Associates, Sunderland, MA. 809 Taylor, J.D., Reid, D.G., 1984. The abundance and trophic classification of molluscs upon coral reefs in the Sudanese Red Sea. J. Nat. Hist. 18, 175–209. Thompson, J.D., Gibson, T.J., Plewniak, F., Jeanmougin, F., Higgins, D.G., 1997. The CLUSTAL_X windows interface: flexible strategies for multiple sequence alignment aided by quality analysis tools. Nucleic Acids Res. 25, 4876– 4882. Thompson, J.D., Higgins, D.G., Gibson, T.J., 1994. CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties, and weight matrix choice. Nucleic Acids Res. 22, 4673–4680. Trussell, G.C., 1996. Phenotypic plasticity in an intertidal snail: the role of a common crab predator. Evolution 50, 448–454. Vermeij, G.J., 1977. The Mesozoic marine revolution: evidence from snails, predators and grazers. Paleobiology 3, 245–258. Vermeij, G.J., 1987. Evolution and Escalation: An Ecological History of Life. Princeton University Press, Princeton, NJ. Vermeij, G.J., Lindberg, D.R., 2000. Delayed herbivory and the assembly of marine benthic ecosystems. Paleobiology 26, 419–430. Wells, F.E., Bryce, C.W., 2000. Seashells of Western Australia. Western Australian Museum, Perth, Australia. Williams, S.T., Ozawa, T., 2006. Molecular phylogeny suggests polyphyly of both the turban shells (family Turbinidae) and the superfamily Trochoidea (Mollusca: Vetigastropoda). Mol. Phylogenet. Evol. 39, 33–51. Williams, S.T., Karube, S., Ozawa, T., 2008. Molecular systematics of Vetigastropoda: Trochidae, Turbinidae and Trochoidea redefined. Zool. Scr. 37, 483–506. Wilson, B., 1993. Australian Marine Shells. Prosobranch Gastropods. Part 1. Odyssey Publishing, Kallaroo. Zelaya, D.G., 2004. The genus Margarella Thiele, 1893 (Gastropoda: Trochidae) in the southwestern Atlantic Ocean. Nautilus 118, 112–120.