¥A/
Molecular Marine Biology and Biotechnology (1997] 6(1), 1-20
Phylogeny and relationships of pleurotomariid gastropods (Mollusca: Gastropoda): an assessment based on partial 18S rDNA and cytochrome c oxidase I sequences
M.G. Harasewych* V2 variable region of the 18S RNA. Relationships of Department of Invertebrate Zoology, National Museum the genera and species within Pleurotomariidae are of Natural History, Smithsonian Institution, fully resolved using "total molecular evidence" Waslvngton, DC 20560. U.S.A. consisting of partial sequences of 18S rDNA and COI and including data on length variation within S. Laura Adamkewicz the inserts. Department of Biology, George Mason University, Fairfax, Virginia 22030, U.S.A. Introduction Judith A. Blake, Deborah Saudek, Tracy Spriggs, and Carol J. Bult The superfamily Pleurotomarioidea is recorded The Institute for Genomic Research, Rockville, from Upper Cambrian deposits (Knight et al., 1960; Maryland 20850, U.S.A. Tracey et al., 1993) and comprises one of the oldest undisputed gastropod lineages. Its diversity at all taxonomic levels was greatest during the Paleozoic, Abstract when these animals were the most numerous, The phylogenetic position of the ancient family varied, and abundant of all gastropods. Pleurotoma- Pleurotomariidae within the Molluscan class Gas- rioideans were conspicuous components of shal- tropoda, as well as the relationships of its Recent low-water marine communities throughout the genera and species, were assessed using an iterative, Paleozoic and Mesozoic, but Cenozoic fossils are two-gene (18S rDNA and cytochrome c oxidase I) rare, occurring in deep-water facies (for review, see approach to phylogeny reconstruction. In order to Hickman, 1976). Of the approximately 1500 species orient the Pleurotomariidae within Gastropoda, par- that have been described, only 24, all members of tial 18S rDNA sequences were determined for 7 the family Pleurotomariidae, are living today. The pleurotomariid and 22 other gastropods that span Recent species are limited to tropical and temperate the major groups within the class as well as for latitudes, occurring along the margins of continen- one cephalopod and two polyplacophorans, which tal tectonic plates that bound the western shores of serve as outgroups. Cladistic analyses of a sequence the Atlantic, Pacific, and Indian Oceans as well as of approximately 450 base pairs (bp) near the 5' end along southeastern Asia, northwestern Australia, of the 18S rDNA support the monophyly of the and intervening islands (for review, see Anseeuw following higher gastropod taxa: Patellogastropoda, and Goto, 1996). All living species are members Vetigastropoda, Neritopsina, Apogastropoda, and of upper continental slope communities (approx. its subclades Caenogastropoda and Heterobranchia. 100-1000 m), most are restricted to hard substrates The 18S rDNA sequences and 579 bp of cytochrome and steep to vertical slopes, and tend to be the nu- c oxidase I (COI) analyzed separately and together, merically dominant, large gastropods in these habi- indicate that Pleurotomariidae are included within tats (Harasewych et al, 1988). Vetigastropoda but comprise a clade that is the sister Pleurotomarioideans are most readily identified group to the other families referred to this order. on the basis of having dextral, conispiral shells with Monophyly of the Pleurotomariidae is also sup- narrow, generally deep emarginations or "slits" ported by the unique presence of seven separate along their outer lips that give rise to a spiral band inserts (ranging in length from 1 to 68 bp) within the or selenizone along or near the periphery of the shell (Figure 1). The discovery of the first living pleurotomariid in 1856 prompted an era of intense '"'Correspondence should be sent to this author. anatomic scrutiny of these rare "living fossils," with *^ 1997 Blackwell Science, Inc. every new collection of specimens generating a se-
// 2 M.G. Harasenych et al.
Figure 1. Representatives of fossil and living Pleurotomariidae, (A) Pleurotomaria angUcus (Sowerby, 1818), type species of Pleurotomaria, lurassic. France. (B) Perotrochas quoyanus (Fischer and Bernai'di, 1856). Recent, Lesser Antilles. (C) En- teinnotrochus adansonianus (Crosse and Fischer, 1861), Recent, Lesser Antilles. Reproduced from Knight et al., 1960, pi. 131 (A) and from Ball, 1889, pi. 37 (B and C).
ries of publications (for review, see Bayer, 1965). surellidae -I- Haliotidae -I- Fissurellidae), Cocculi- Pleurotomariids liave since been regarded almost noidea, Trochoidea, and the extinct Macluritoidea universally as the most primitive living gastropods and Murchinsinoidea, as well as two other "early (e.g., Thiele, 1929; Taylor and Sohl, 1962; Boss, side-branches," the patellogastropod limpets and 1982; Vaught, 1989) and appeared as such in para- the Neritopsina. Following a rapid succession of digms of gastropod evolution (e.g., Bouvier, 1887; articles providing data on new suites of anatomic Yonge, 1947; Morton and Yonge, 1964; Graham, and ultrastructural characters as well as on newly 1985; Salvini-Plawen, 1985; Brusca and Brusca, discovered taxa, Haszprunar (1988) published a hy- 1990; for a detailed review, see Bieler, 1992). The pothesis of gastropod phylogeny in which he pro- combination in pleurotomarioideans of an asym- posed the successive emergence of 13 higher taxa metrically coiled shell and symmetrically paired (Figure 2A). More recently, Ponder and Lindberg palliai organs (including gills, auricles, kidneys, (1996, 1997) published parsimony-based phyloge- and hypobranchial glands) has been regarded as netic analyses of Gastropoda based on anatomic transitional between extinct, planispirally coiled, characters (Figure 2B) that differed from Haszpru- bilaterally symmetrical ancestors, generally be- nar's hypothesis in the composition and placement lieved to be bellerophonts, and asymmetrical mod- of several key families and orders, but concurred ern gastropods. in placing the patellogastropod limpets as the sister Interest in gastropod phylogeny has had a renais- taxon to the remaining gastropods and the pleuroto- sance over the past two decades, prompted in large mariids as closely related to the trochids. part by discoveries of primitive, previously un- Few DNA sequence-based phylogenies con- known groups, primarily from deep-sea hydrother- taining more than five gastropods have been pub- mal vents, and fueled by developments of molecular lished to date (Emberton et al., 1990; Tillier et al,, techniques and software for cladistic analysis, as 1992, 1994, 1996; Rosenberg et al., 1994, 1997), all well as by the more widespread application of the based on small portions (approx. 150-700 bp) from principles of phylogenetic systematics. The first of various domains within the large ribosomal mole- the major works to reassess gastropod evolution was cule (28S rRNA). Most focused on evolutionary rela- that of Golikov and Starobogatov (1975). These au- tionships of higher taxa within Gastropoda, but thors recognized that the patellogastropod limpets, several (Tillier et al., 1992, 1994; Rosenberg et al., together with some Paleozoic groups, manifested a 1994, 1997) spanned the class as a whole. These r series of archaic characters that were neither shared studies, for the most part, supported themonophyly with nor could be derived from other gastropods. of Vetigastropoda and Apogastropoda [sensu Pon- They placed these taxa in the subclass Cyclobran- der and Lindberg, 1997 - Caenogastropoda + Het- chia to segregate them as an earl}', monophyletic erobranchia), but differed in the placement of such offshoot fi-om the remaining gastropods, and consid- basal groups as the Neritopsina (with Bivalvia in ered the pleurotomarioideans to be the most primi- Rosenberg et al, 1994, 1997) and the patellogastro- tive of the Recent forms among the remaining pod limpets. Relationships of the patellogastropod gastropod taxa. limpets have proved especially difficult to resolve, Salvini-Plawen (1980) divided the basal gastro- as they emerge as a sister group to all other Gastro- pods into the Vetigastropoda, in which he included poda (Tillier et al., 1992), as a sister group to either the Pleurotomarioidea (Pleurotomariidae + Scis- Vetigastropoda, Neomphalina -I- Apogastropoda, Molecular phylogeny of pleurotomarüd gastropods 3
Figure 2. Recent morphologj'-basecl P*TELl.DaASTPOPail* - POLVPi-rtCOPHLWA
hypotheses of gastropod phylogeny. Ci>]CULWIlOAE COCíUL íUfíf^MIft - ríAUílLOIDÉA TEIELLOIÜEA - CATEUOÜASTUDPOQA NEPiTOPSIHA (A) Phylogenetic reconstruction of - LfPÇTÏLUOlDEA 1 Haszprunar (1988, Figure 5), with no- - RSSüBELLOlDEA - PtEJPOTOWAfllCrtlEA uetlGASTBOPCOA menclature and taxon rank modified/ • PISSUHELLOIDEA - TTICCMIDAE . SCis5ur4ttb
case ai'e represented in the present - CEniTHllD4E
CAZWOÜBTÍÍOPDUA CAEHDGAS1 Rf^POOA study (see Table 1). The "Hot Vent - MEDTÍE^IOQLOSSA C" limpets reported by Haszprunar - 10NNC}I[>LA APOGASTBOPOOA (1988) were subsequently described - KJfiJCOIDEA - COMQ<íl£A as family Neolepetopsidae by HFTEROElñArJGHIA iVicLean (1990) and included in Patel- HFl£ftOBA4NCHm l'rRAlJiPELtOlCftA - 0PISTHüBJlAí4L^t^|.-, logastropoda by Ponder and Lindberg EUtHYMEURA - Pm.fd<ííATA (1996, 1997),
or Heterobranchia (Tillier et al., 1994), or within letic. Most recently. Ponder and Lindberg (1996, Caenogastropoda (Rosenberg et al., 1997), de- 1997) added the superfamily Lepetelloidea and the pending on such variables as outgroup, included family Senguenziidae to the Vetigastropoda, re- taxa, and the algorithm used for tree construction. moved the Neomphalidae, grouping it with other In the only study to contain a pleurotomarüd (Ro- "Vent ta:xa," and considered Pleurotomarioidea to senberg et al., 1994), the single species emerged be the sister taxon of Trochidae + Seguenziidae within Caenogastropoda at the base of the Neogas- (see Figure 2B). tropoda. Sequences of the small subunit ribosomal The family Pleurotomariidae is of Triassic (Ladi- RNA (18S rRNA) molecule have been determined nian) origin (Tracey et al., 1993). The majority of for relatively few (<5) gastropods, generally in the the dozen genera included in this family are extinct, context of broader studies such as the evolution of and all have been defined typologically on the basis the Metazoa (e.g., Field et al., 1988; Ghiselin, 1988; of the presence or absence of a few easily recogniz- Winnepenninckx et al., 1994). able shell characters used sequentially. Most mod- The composition of the superfamily Pleurotoma- ern authors concede that generic distinctions are rioidea has also undergone considerable revision not always clear and that many taxa are not readily during the course of renewed scrutiny. Knight et al. assigned into existiiig genera (Cox, 1960; Hickman, (1960) included within the Pleurotmarioidea 1976; Szabó, 1980). The genus Pleurotomaria De- 20 extinct families as well as the living families france, 1826, is generally regarded as having ranged Pleurotomariidae, Scissurellidae, and Haliotidae, a from the lower Jurassic to the Lower Cretaceous scheme followed in most contemporary classifica- (Knight et al., 1960), while the genera Pemtrochus tions (e.g., Hyraan, 1967; Hickman, 1984; Vaught, Fischer, 1885, Entemnotrochus Fischer, 1885, and 1989). Salvini-Plawen (1980) added Fissurellidae Mikadotrochus Lindholm, 1927, have been pro- to Pleurotomarioidea, which he included in Vetigas- posed for living pleurotomariids. Some authors re- tropoda. Boss (1982) added Neomphalidae, but re- gard these as, at best, subgenera of Pleurotomaria moved Fissurelloidea as a separate superfamily. In a (e.g., Dall, 1889; Hickman, 1976, 1984; Abbott and reassessment of the Vetigastropoda, Salvini-Plawen Dance, 1982), while most others (e.g.. Knight et al., and Haszprunar (1987) identified other anatomic 1960; Bayer, 1965; Benfrika, 1984; Vaught, 1989; characters to support the monophyly of Vetigast- Endo, 1995; Anseeuw and Goto, 1996) accord some ropoda, which they considered to include the or all of them generic rank. A synopsis of available superfamilies Fissulrelloidea, Trochoidea, and information on living Pleurotomariidae has recently Pleurotomarioidea (containing the families Pleuro- been published (Anseeuw and Goto, 1996). To date, tomariidae, Scissurellidae, and Haliotidae), with there have been no published studies of phyloge- Neomphalidae appended as "incertae sedis." These netic relationships within the famil)'. authors noted that no synapomorphies could be The principal objectives of this study are to re- found to unite the pleurotomarioidean families, and construct the major outlines of gastropod evolution that Pleurotomarioidea was most likely polyphy- on the basis of partial sequences of the 18S rDNA 4 M.G. Harasewych et al. gene and to assess the ph3'logenetic position and the smaller E-10-1 insertions found in Nautilus and composition of the superfamily Pleurotomarioidea in the patellogastropod and cocculiniform limpets. on the basis of these data. Corroboration and better Therefore, the alignment encompassing all taxa resolution of the relationsliips of the genera and spans 653 positions (Figure 3), although the longest species within the family Pleurotomariidae are pro- constituent sequence is 537 bp. Seven autapomor- vided by analyses of sequence data from the mito- phic, single-nucleotide insertions were excluded chondrial cytochrome c oxidase subunit I gene. from phylogenetic analyses, as were 10 regions, ranging in length from 1 to 171 bp, that could not be reliably aligned (Figure 4B). Of the remaining Results 392 alignable sites, 199 were constant and 73 were Partial sequences comprising approximately 450 parsimony uninformative. Cladistic analysis of the base pairs (bp) near the 5' end of the 18S rDNA 120 informative characters using the branch and gene were determined for 27 gastropods, one cepha- bound algorithm of PAUP 4.0.0d42 produced 4656 lopod, and one polyplacophoran, and aligned with most parsimonious trees with a length (L) of 433, homologous regions of the complete gene sequences consistency index (CI) of 0.621, retention index (RI) of ope polyplacophoran and two gastropods ob- of 0.744, and rescaled consistency index (RC) of tained from GenBank (see Table 1). The portion of 0.481. The strict consensus of these trees is illus- the gene used in this study corresponds to positions trated in Figure 5A. Results of bootstrap and jack- 60 to 515 of the ] 8S rRNA sequence of Onchidella knife analyses (1000 replicates) are superimposed céltica as reported by Winnepenninckx et al. (1994, on the consensus tree. Characters were reweighted p. 101). Among the ta3Perotrochus midas) owing to the pres- 60 most parsimonious trees of length 229.95 (CI = ence of insertions and deletions, especially in the 0.779; RI = 0.845; RC = 0.659). The strict consensus regions of the helices 10, E-lO-l, and 11 (see VVin- of these trees, with bootstrap and jackknife support nepenninckx et al. 1994, Figure 1). The large inser- for nodes, is shown in Figure 5B. tion in this region that was found in all Quantitative comparisons of the 18S rDNA se- Pleurotomariidae is of questionable homology to quence-based phylogeny with morphology-based
Table 1, Locality data, tissues extracted, voucher specimen information, and sequence accession numbers for taxa used in this study.
SíjqiiL^ni:e ^rij'ii-ince
y'iuir.Wrr accR&sion nrri ^s:na
Taxon CoMeciicjn localiiv n-.iireijul riiimbrr 18S i;,iiidrr GOl
Class Polyplacophür:! Acaí\thQ)]lt!í.na jirjioniro (Lischkr. 1ü"¿l Gmßauk ex \ViiLLirp<'¡;iimckx .\ ak. 1993 EMBL X70210 Cnviioc.'u/.ï" ~-írf,Vri (Miiidfiuioril. ia-17) Bsiiiütjld, B.C., CanaJii Biuxid musLli- USNM 888657 CSDB L78876
Ctaís Cßplialüpttdii ¡VaiFf/íus .ícioiinitoílí« L-jililluijl. 1786 PajJLia, New GtLLr]GEi imral rmisde USNjVI 8S5r,78 GSDB L78377
Clj^s CüSliopuJ.i PatcllojiïaBlFopQds Acinoeu niüra Rntrlike. líi^rí BomfiFid, B.C., C.iiada Hiici-.il niLisrliJ USNM 888640 CSDB L78878 O Huno ni^yr-ürwula IReeve. 1854) MiTjub+í. Jíipaii Buuiinl miisili! USK.vl 888623 GSDB 1,8879
Cocriilinirorniia CnucuJJnidiH! Coi:(.:utinu ilics=ii!-J McU:.;n í^ il.i-.iscwych, 1995 Huc«l iiiusde USNM 888655 GSDB L78880 Lepeltjlluidea iVoíDLTrorpr /ioijbr>i;^í Mv.Lt;.iíi ^ H^rdíví\\y(li, 1995 Whuk iiiiimiil USNM üUSbjb GSDB L;888:
NprdupHÍno Nf^riio ver.Tj'rijMfGrnelin. 1791 üjj Pine Kay. Id , USA BUCCHI niu.icie L'SNM 888658 CSDB L78882 Neritina recH^'íila {Say. IHZ^) iig Pille Key, FL, USA nm;i;ol raustlo USNM 888659 GSDB L78883 Molecular phylogeny of pieurotomariid gastropods 5
Table 1. Continued
Sequence Sequenct;
Voucher accession 3CrA:ssinn Tax on Colleclion locniitv material number 18S number COI
Vetigaslropoda Fissiirelloidea Diodora cayenensis (Lamarck, 1822) Sebastian [n)(!t. FL. USA Buccai mustie USNM 888660 GSDB L78884 CSDB L78908 Halioloidea Holiolis ítj/csrríis- Swainson. 1822 Bamfield, B,C.. Canada Bucea) muscie USNM 888642 GSDB L78885 Trochidae /\si(aeo coelato (Gmelin, 1791) Berrv (s., Bahamas Buccal mnsile USlNM 888603 CSDB L78886 GSDB L78909 Citlarium pica [Linné. 1758) jamaica Bucea) musc)e USNM 888661 GSDB L78887 P)eurrjloinarioidea Eníemnoírochiis adonsonianus (Crosse & Fisc)ier. 1861) Guadeloupe Buccal muscle USNM 888647 GSDB L78888 GSDB L78910 Eníp;n;ioííodl(i5 rumphii (SciTepman, 1879) Aniami-O-Shima. )apan Bucea) nuiscle USNM 888698 CSDB L78889 GSDB L78911 Perolrochus quovanius (Fisc)ier K: F^ernardi, 1856) Guadeloupe Buccal uuiscle USNM 888646 GSDB L7He90 CSDB L7Ö915 Pecotrochus iucaya P.M. Bayer. 1964 Bahamas Buccal muscle USNM 888619 GSDB L78891 CSDB L78916 Perohoclnis nwuieri Harase\vych & Askew. 1991 Charleslon. se. USA Buccal muscle USNM 388662 GSDB L78892 GSDB L78914 Perolrochus midas P.M. Bayer. 1965 Goulding's Cay. Bahamas Buccal muscle USNM 888645 GSDB L78893 GSDB L78913 Peroirochus Icremochii Kuroda. 1955 lapan Buccal muscle USNM 888617 GSDB L78894 GSDB L78912
Caeaogastrûpoda Cerithiidae Cerilhium alivlum (Born. 1778) Sebaslian Inlel, FL, USA Buccal muscle USNM 888663 GSDB L78895 GSDB L78907 Xenoplioroidea Xenophora esiilum Reeve, 184,3 Minabe. Japan Buccal muscle USNM 888631 GSDB L78896 Tormoidea Pusitrilon oregonense (Redfie)d. 1848) Dutch Haibor. AK, USA Buccal muscle USNM 888534 CSDB L78897 Miiricoidea Oliva sayona Ravene). 1834 Fl. Pierce. FL, USA Buccal muscle USNM 888605 GSDB L78898 Conoidea Hosítdo cinérea [Born. 1778) Pi. Pierce, PL, USA Buccal muscle l S\M 888611 GSDB L78899
Heterobranchia Pyiamide)]oidea Fargoo busihona Bailsc):. 1909 Fl. Pierce. FL. USA Whole animal USNM S88638 GSDB L78900
Eulhvneiira
Opislhobianchia Huniinoea onliHarinn (d'Oibignv, 1841) Ft. Pierce. FL. USA Buccal musr:le USNM 888664 GSDB L78901 Aplysia doclyloinela (Rang. 1828) Minabe. Japan Buccal muscle USNM 888624 CSDB L78902
Pulnionata Onchidella cellica (Cuviíjr. 1817) Genbank ex VViunepenninck,\ (-1 al.. 1994 EMBL X70211 Physa Iwleroslmpha Say. 1822 llhaca. NY, USA Bu< ( .:! muscle USNM 888613 GSDB L78905 Limicoíaria konibeal (Bru^uière. 1792) Genbank ex VVinnepenninckx et al.. 1992 EMBL X66374 Liniax mo.viinus Liuní-'. 1758 Silver Sjirlnp. MD, USA Buccal muscle USNM 888604 CSDB L78906
EN4BL indicates European Molecular Biology Laboratory Data Library; GSDB. Genome Sequence Data Bank; USNM, niollusk collection, National Museum of Natural History, Smithsonian Institution.
phylogenetic hypotheses are based on reanalyses of To better resolve the relationships among the sequence data for subsets of shared taxa (shown Vetigastropoda, a subset of taxa that included the in upper-case letters in Figure 2) and subsequent seven pieurotomariid and four nonpleurotomariid manipulation of tree structure using MacClade vetigastropods, as well as the two Neritopsina and (Maddison and Maddison, 1992). Resulting in- Cerithiuin and Hastula (as representatives of Caeno- creases in tree length and indices are reported in gastropoda), was reanalyzed. Although the align- Table 2, ment of nucleotides was not altered, elimination 6 M.G. Horasevvych et al.
ID ion
AcaíiClmpleuia T.AAiïr.A'CTiGACTTT- " • - -Ly,-C"A[A01I.;"wv\Caï:AMTaX'rCAWAMIX3iïrrATCATTTCTTMaiCCTA" "O\C"T"0C''TAel 71 Cryprt:cichitor, . . .T" . .C.•.T .G. .C .G ACG.A. CG^. . . -c.CA""r. HO G . . .•! •• ..G...GT rCcmxXMGG. .G.G. 'OCC. " . . .T .c. .c A. .A.C..C. . .G "03W.AG.GQG.". 93 Cellanö r. .T" cr,- ronncrjxc- .G.COJ.TC. " .C -G. .C A-O^.G. .C. . .G T. .C. .T-.GO;.". .C. ..".• 9Ö .r ....A....JC.G --.GA".- . . r.. . 'U ".".CGGTiarr "" .TT'";ccc. ".. "G. .C ...."G..c.• .G .CG.A. . . .Ain"; rk.O\.'i -.cr-l'.G. . 1 . . . Slj •• . . .T" .A • •• •• G. . G..C. . . . IB . • ç. .A..C..-"- ,-.(-• •' .A...C.-"- "" .-G. • IfiO IBO 300 AcanciBple'jva ""irrÄ 'A;-'.CTGTGr:TAATTCTAGiOrrAATA2ATGACGTTCA''GCTCCGa^ " " "TTT- *"' TT GC'- 'AGG^A'^ -.AGi CC":'l l'.A'i'r ii5 Crypcochimn t^uLLlua C ... CAACGAT' ."TC. . , ' Ul.G .A'",".... • .J Acraea AA. ... CAACG. ."C.GTC.T. .CCC"' 'CIT. .CCTVAICT..Ç-TK!r.'i.y:r, " A. CG .A. . Ce: lara AA. . .W M cmcG. ."".G."A. ..C. ••••c""GCii::.A'- • G; Ml Ca^cnj ; u'A . . .G 164 . .A. .A C-M>\. .T. . - . . .T. .CXXiTi'-XTT. . .LXiGC"ai;.CÏCGT- .T"Ce. .GC .G. . .G. . . . .G I7T ""C... c G. .CAAß^.. " . . . .T. . . . .C C l'îl Nericijia 151 Diodora C-AG.TTO C . .Cec- -".=A- G .A. . . . . C ; wi HalioCis C-AC.A.A C"- . . .C"T .A' G .G. . . . .C ist AsLraea C"AG.ATA C"' . .CQC" "" .A 'G G .. .c 157 Citcarium C".';.CAIA C . .Ö3Z~ .A"-G .G. . .. .c m E. ac3ansor"ü.¿y;ü C OZ-^-ZC. . " " " "" C .CGC •".."" 15H E. rufipt'ii •,. Q TG^'^'"' " - . .. - C.crcC ' " \5B P. quoyanus C OGAG':. . " H•H.,_^.,^ (-^Q-. i 57 P. Icicaya .... C . ..•_'GACC.." .C.C•-"" ".."" 157 P. ireureri c ("SAIX.." "".CCTC""" .."" 15K P. niiclss i- rG.AC^'.." .C.CGG"*" •'".,"" I5P, P. ceCBnsc.M i -i- U-CG " -"•* C CCC' • J5â Cei-Lthiudi C.,V-.C . ." C.CAC G. .A. . . .R. IS4 Xenophura "-:::' " . . C. A C . . " C" C. ce " " • G . . A. . . . .1' . . .Y lie Rjsicriton C.TAC. ." G"-- C.• G. .A. . ist 01 i-^ c:r.AC.." C C.CG-" " G. .A. . 15-î rastula C. .AC. . " c c.•" "G. .A. . 151 Fàrçfca •,. A.C C.TAC. .A ••- C.CG " C. .A. . T51 H^tminoca C C. :G.G.A ""C.• T. .A. . 151 f^lysiii •.. C CTKE. .A -C.• •- :'. .A. . 15; Onchidella G ... .C'A. . .A ""- "c.cr G 151 LuTiicoUria G .... i.-TT.AC. . A " " C. tr-." " • c. . A. . 15J Physa Lirrew •• C .'i'V.-.C. .A C.CG T. .A. . iu Figure 3. Aligneci partial sequences of the molluscan 18S rDNA gene spanning helices 6-17, and corresponding to positions 60-515 of the 18S rRNA sequence of OncbJdeUa céltica as reported by VVinnepennincLx et a!. [1994, p. 101). Among the included taxa the length of the gene between these two positions ranges from 430 to 537 bp owing to the presence of insertions and deletions, especially in the regions of the helices 10, E-10-1, and 11. All sequences were confirmed by at least two sequencing reactions in one direction and a third sequencing reaction in the opposite reaction. Ambiguous base assignments are noted using lUPAC symbols. Dots (•] represent nucleotides identical to those oí Acan- thopleum, dashes (-) represent missing data, and quotation marks (") represent gaps inserted during alignment. Molecular phviogeny of pleiurotomariid gastropods 7 230 ?£;"ATCAñGA TCAíM- -fïïJX" "C" "TC" •GG" 1B 310 COT ?:07XAT"TCTCAATA 204 Cfypcochif."-:-! A. .C"" 2(Jb fsä'Jl;i lus TTGCAA. •CGTT . GA " " - •MTCQCAAATA- .G". 275 Aomea .GÄ"-" CAA" .AA. 269 Celläna • ••'-•«'• .AA. 2Í6 Coccüllna 2Da NococraLer 332 Nerita 202 Nericijia 203 Diodora --•c. 211 Haliocis T. 2Û7 Asctaeñ T. ¡se C: f.l-xurl'jni 221 E. .sclanscniarjui "YT.G.;. ^ •pGG••GC^CCCACCA.Ca30CCCÏJrT^TTCGCrr:H^G(KIrTC^G^•,^•ÏV^CG.A. 2f!l E. runîMlii ;•:•, :-;.C :'lG"•cCCCAOCAC•CL\'TrijrTirxr"CGCOUCCrTTCCCGAACC. A. 281 P. quoysHLis "':s.GA YTa7r3a3.Acxix32"CcccccACCAccrrr• Figure 3. Continued of taxa that contained putatively nonhomologous appear to be unique to the family. These insertions, insertions enabled the alignment to be condensed which vary in length from 1 to 68 nucleotides, are to 556 positions. All of the regions excluded from concentrated in the helices 10 and E-10-1 (Table 3, the original analysis could be unambiguously Figure 4C). Two of the seven insertions vary in aligned, although not all were present in all taxa. length among pleurotomariid taxa, each appearing All Pleurotomariidae were found to contain seven to contain one or more insertions or deletions. Five insertions in the V2 region (De Rijk et al., 1992) that autapomorphic, single-nucleotide insertions from 8 M.G. Harasewych et al. 500 Acanthopieura ACr-TTGIGCraATCGTAirJ'; Lr"Ac-"Ga:;' CCOGOSA ' aji'ATCTrrcHACTKTKTCCcci'ATrA" Acmrr ^•iX<-Tfv;.13K3ATf;iT.CCTACC . . .T -.A...... G. .".GT-."". . •. .ex: G..AC . .C.. . .C • .G. .CG. .T.axj.. .c... 365 ..."G.."".C...... GG.. "T.i-rr•.G.'xc.. .A. .. '.T.TG.CA.G T .."....".T..G. . CG. c. ce J62 Cfell^na. ..."... .""C. . . •" . .c."" .AJ-.m ..A.' •ICC... A.G r ..'!...." .T.'3A. .CG.C.CC-C Cocculina ..-"CGC . .TA. . "".TTT--"T.C '" . .A. . . . •.AA. . .CAA,-,. . . .A G-.. A' J'A. .ÍXJTC. .ce. .A.C. . . 294 NoCLOcrûLei- . .C •" C...... GG.AGTTCTTAACA.GCTC.TA. . . . '. .CGC.CGAA..AC C...G.. .CG.aX.'.G c... fjerita • A • A 351 . ^ •.. . MeriLina 392 LJiocbira . . • -.G". . .G. . . •..CG. .CA. .. .A G...G.. . .A. . .c 300 iteliotis ..'-.CIA. .G. . . C. . • . .C. . . .A <-. . .G.. . .A.. .c •;.... 2S6 257 Astraea . .-".CTA. .G. . . . .T: .A. ' . .C... .A G...G.. . .A. . .'• T. . . . Ciutârium . . -".CTA. .G...... ce ..rn-'... 'G •..CG. . .A A . .T. . .C T. . . . 31.'; E. ad^nspruanus ' G AT G C ^ r-- "" Ahi ' . . . . • . .CG A .C i-x:... .c... 37? 372 E. rurtphii ..." .C.A'l'.G...... C..G..C."".A. ' . .03 A .C. CJC. .. .c . . P. qijqygnus ..." .CAT.G...... C .G. .TT."".A • '" S. A. .rr, ".A T. .C. ;•;•:•... .c .. 177 379 P. iucaya C. .G. .TT. "• .A. • ". .CG A .C. ACCC....C... P. rmureri. " C AT G C G .••';'. " ' .A. ". .CG A .C. ACCC....c... 360 p. midas . . .".C.AT.G. . . 3.. T: . ' ' . A • . .CG A .C CCC...C... 389 p. tetanöChil ...".C.AT.G...... C . 3. .T"."" .A. -' ' . .CG A .C. cx".r... .C... 3 S« Cerithiuri ' . .TT .A. ". .C A -.A.GA. .CG c 290 ". .C ".A -.A.GA. 281 ..." C. . . .."T."".A. •" • ' . .r A ".A.GA. 2ëii C • . . -T. -" .A. ". .C A ".A.GA. c 289 Hflstiijla C • . . "T. "" .A. " . .:• A ".A.GA. c 2B9 Fargoa " . .C A -.A.G.. 3ÜB Í^OTiinoea - . .C A ".A.G.. -¡(¡6 .^Ç'iysia ,. •. ::vvi- .T. " . .C A " .A.G.. 309 crchittelU • ". .C A T.".A.G". 107 ijimi col aria " . .C A ".A.G.. c 306 Phyp-"; " C A - .A.G . 306 • • • ,. ' . "i".ACT. ' " • . .C A -.A.G.. c 3(17 540 5^.0 SBO ATCCri • ITGTAACGGGTMC ' C =C"1AG=VA'ICA •• U.X?r! iIGATT-: -V.GGÍi.GAGGGAC^ATGí^GAAACEGr "AC 7\CATCCAAGG " AilTOTA :CAaXGr.GCAAATT 3» CrypCochi ter; 392 ^•• G QA. . . . 4Si Acnaea C. .C C. G. . . . C 459 • Cal lare G "... C.-C C. .. G. . . . .c.. 420 Cûcculirfi G. A. . . .C. . 392 pJotociraLer C. T T .C. . 424 iiiericfi G " G " C 388 ^jtîriLirid G " G " r 3B9 iTíicdíiríi 197 Hfâiiotis G " 393 • Astraea GA G " 394 • .. Cirtariijm G '• " 411 E. adansorü-sri G •' . . . .G. . C .. r 46& K- runphii G G " ....G. C .. .C.. 469 P. quoyamJS G ce G G " C .C. . 474 P. lucaya G ce G G C .C . 476 P. iTBureri. .C. . 477 p. inidcJä G. . . .ce . . . G. . " G ... . C .c.. 43S P. fierarrecñi i C .C. . 463 367 Ceritfiiurri r r- Xenopbora 378 FusitJficcn . . .T.A.C. . . . .G '• -•' .C. . itiâ Olivs G " r 386 . , .T.G.C ...... G. •• .G " 3B6 tíasculñ c Farpoa 405 M Hamlnoaa G r 4U3 Apiysia . . .T G GV T T ... .T.... .c.. 40(î Cnchidel Ui .c.. 4D.Í .. 1- Licnicolariâ G " A. -" 401 403 i-¡ .'• LLTTCIX ...... •• -G •' Fisure 3. Continued other portions of the sequence were excluded from RC = 0.778). Figure 6B shows the strict consensus all analyses. An initial maximum parsimony analy- of the six most parsimonious trees (L = 196; CI = sis (exhaustive search option, 94 informative char- 0.872; RI = 0.918; RC = 0.801) produced when acters) was performed on data from which the seven the analysis was repeated with tlie seven insertions insertions (totaling 94 nucleotides) unique to Pleu- included in the data set (105 informative charac- rotomariidae were excluded. Figure 6A illustrates ters). the strict consensus of the resulting 36 most paj'si- Figure 7 contrasts the phylogenetic information raonious trees (L = 168; CI = 0.851; RI = 0.914; contained in the length polymorphisms of these in- Molecular phylogeny of pleurotowariid gastropods 9 Figure 3. Continued t; • 620 630 640 650 Acêinchopieura Aca:?.CTCCT03:.H•«:GA•TA£r;GAa;ajAfATAACA'c-;j.; ri.a-'/.r- 441 Crvptochi-tín v .. 443 Nautil-js A.. .c - T.cr,.T,. ...••••... 'íi.o Acnse* A. . .C.A 'Ca':. . . .GOL'. . vil Cellam A. . .C.A.T T.-GCO-: Y.C. ,"2 Coccuiina A. . .C.A c. . . . C. . . '144 NotOcraCEr- A. .TC.A.T..A " ".-,.-.G C• 474 tterica A •• C. . . 440 Naritína T • C.. . 441 EHotora A. .•?:•.AT.. .A " c.. . 449 Haliocis A. .'.r.AT. . .A " •• C. . . 444 AstlBsa A. .TC.A. . . .A " C. . . 44f CittariLm A. .TC.A. . . .A " C. . . 463 E. udflnsoTLÍsriu^: " C, . . W.i K. nirrpiLLi " C. . . b2l ?. quoyanus " c. . . "^26 r". luc^ys " C. . . ^28 P. jTHLireri " " C. . . 528 P. [nidas " C. . . 537 P. tei-ämäch:. " '• C... 534 Oerithi'jT. ". . .AA.C. . . 439 Xencptotö '• • • • A.c... .-:3c FVjsicriton " A.C. . . 4j8 OUva "....A.C... 4J6 HastUla " A.c. . . •;3B Fargoa C " C... 457 Hamirioea c " C... 455 fplysía C " A C. . . 460 Onchidella C " C... 456 Liinicoiiuria " .c A.G " c.. . 45? Physa C • " C... 454 LiiIBX C " C. . . 456 Figure 4. Distribution of variable 5 7 8 9 10 11 12 13 14 15 16 17 sites along tlie 18S rDNA gene among B the 32 taxa in this study (Table 1). (A) C Positions of helices (from Winnepen- ninckx et a!., 1994, Figure 1]. (B) Posi- tions of bases excluded from phylogenetic analyses due to uncer- tain homology or alignment (bars) or because they were autapomorphic in- sertions (*). (C) Location of the seven insertions unique to Pleurotomarii- dae. Asterisks (*) indicate single base insertions. 50 100 150 200 250 300 350 400 450 500 550 600 650 SITE sériions with that contained in their sequence. Of ing to positions 32 to 611 of the COI sequence of the 94 aligned positions in the inserts, only 11 are DrosophUa yakuba (Folmer et al., 1994, Table 2), parsimony-informative, producing 9 trees (L - 27; Of the 579 nucleotides, 333 were constant and 59 CI = RI = RC = 1.0], A consensus tree is shown were parsimony-uninformative. Analysis of the 187 in Figure 7A. Figure 7B illustrates the strict consen- informative cites using the exhaustive search algo- sus of three trees (L = 22; CI 0,864; RI = 0.885; RC rithm produced a single most parsimonious tree = 0.764) based on an analysis of the distribution (L = 540; CI = 0.669; RI = 0.581; RC = 0.388), of the 19 insert segments (Figure 8) among pleuroto- shown in Figure lOA with bootstrap and jackknife mariid taxa. support for nodes superimposed. Alignment of the partial sequences for the mito- The "total molecular evidence" data matrix con- chondrial cytochrome c oxidase I gene (Figure 9) sisted of 1154 characters (556 18S rDNA -I- 579 COI produced an unambiguous alignment correspond- + 19 insert segments), of which 755 were constant 10 M.G. Harasewvch et al. Figure 5. Phylogenetic relationships U^POLYPiACOPHORA • NAUTILOlDEA among major groups of Gastropoda, PAT EL LOUAS T ROPODA Nautilus, and two Polyplacophora COCCUilNIFORMIA based on approx. 450 bp of 18S rDNA sequence from which autapomorphic NERlTOPSIWA insertions and regions of uncertain alignment have been removed. (A) «^LELinOTOMARIIDAE Consensus of the 4656 shortest trees (L = 443; CI = 0.621; Rl = 0.744; RC VETtGAËTnOPODA = 0.481) resulting fxoni a branch and bound search. [B] Consensus of 60 shortest trees (L = 229.95; Cl = 0.779; RI = 0.845; RC = 0.659) pro- CAEN0GASTF1ÖP0DA duced by a branch and bound search of a data set in which characters were APOGASTHOPODA reweighted by the maximum value of the rescaled consistency index (base HETEHOB RANCHEA weight = 1). Bootstrap proportions given as percentages above the nodes J and jackknife proportions given as percentages below the nodes. An as- terisk (*) indicates that the node was not supported in a 50% majority rule bootstrap or jackknife consensus tree. Table 2. Comparisons of 18S rDNA sequence-based phylogeny of the class Gastropoda with contemporary, morphology- based hypotheses.* Phylogenetic hypothesis length CI RI RC Taxa [N) Haszpruirar (Figure 2A) 360 0.65 0.53 0.34 14 18S rDNA tree+ 338 0.69 0.61 0.42 14 Ponder & Lindberg (Figure 2B) 371 0.63 0.50 0.32 15 18S rDNA tree* 338 0.69 0.62 0.43 15 *TQ facilitate compaiisons. trees were pruned to contain ttie same sets of taxa. When t-vigt\er taxa were represented by oMiltiple species in the 18S data set. a single representative was selnired with the objective of minimizing the number of deletions and ambiguous base as.signments. Patellogastropoda is represented by Cellana: Neritopsina by Nerita: Pteurotomariidae by Perotrochus lucaya: Trocliida« by Astraea: Euthyneura and Opisthobranchia by Haminoea: Pulmonata by Oncliidnllo. Polyplacophora were excluded from the data set, and Naulilus was selected as the outgroup. Comparisons are based on taxa identified by Upper Case in Figure 2, 'A single most parsimonious tree resulted from an anahsis using the reduced subset of taxa. 'Three equally parsuironious trees differing only in the topology of taxa within Caenogastropoda resulted from an analysis \ising the reduced subset of taxa. One of tlies(i frees, which agieed with the generally accepted topology of Caenogastropoda included in Ponder and Lindberg (Figure 2B), wa.s used as the basis for comparison. and 275 were parsimony-informative. An exhaus- sequence data resuhed in a phylogenetic hypothesis tive search yielded a single most parsimonious tree for the class Gastropoda (Figure 5) that is, for the (L - 712; CI = 0.729; RI = 0.658; RC = 0.479), most part, similar to those based on morphologic shown in Figure lOB. Bootstrap and jackknife val- data (Figure 2], Quantitative comparisons (Table 2] ues are given for each node. indicate that the phylogenetic hypotheses advo- cated by Haszprunar (1988) are, as a whole, more congruent with the molecular data than those pub- Discussion lished by Ponder and Lindberg (1996, 1997). On the Maximum parsimony analyses of the conservative, basis of molecular data, living Gastropoda repre- unambiguously aligned portions of the 18S rDNA sented in our study are divided into four groups. Molecular phylogeny of pleurotornariid gastropods 11 Table 3. Insertions within the 18S rDNA gene tliat are unique to Pleurotomariidae.* Insertion Length (bp) Position in aligned sequence (Figure 3) Segment 1 3 211-213 1 2 1 220 2 3a 3 247^249 3 3b 1 250 4 3c 4 251-254 5 4 7 276-277, 284-285, 290-291, 293 6 5 6 296-299, 310 7 6a 1 314 8 6b 1 315 9 6c 8 316-323 10 6d 10 324-333 11 6e 1 334 12 6f 2 335-336 13 6g 1 337 14 6h 1 338 15 6i 20 339-358 16 6j 1 359 17 6k 23 360-382 18 7 1 422 19 •Positions are referenced to llie aligned sequences stiown in Figure 3. Insertions 3 and 6 vary in lengLli among pleurotornariid taxa, and tiave been subdivided into segments that could be scored as either present or absent in individual taxa (Figure 8) and sen.'e as the basis for the cladogram in Figure 6B. Figure 6. Phylogenetic relation- ships among Vetigastropoda, Neritop- sina, and Caenogastropoda based on dicdofo 18S rDNA sequences. Five autapo- ^- morphic, single-nucleotide inser- ^ HjfröiiS 100 tions are excluded from analyses. (A) AS{T:!C.I -^ iCO A consensus tree of 36 most pai-simo- TROCHIDAE CUtarium Cillariiim • nious trees (L = 168; CI = 0.851; Rl E.adansonianus -n = 0.914; RC = 0.778) resulting bom 60 98 96 Enteninolrochus a branch and bound search of the data eo F:uij:rhu set ft-om which the seven insertions P.C¡'j the limpets, the Neritopsina, the Vetigastropoda, and Signor, 1991) with early Paleozoic origins (Tra- and the Apogastropoda, each representing radia- cey etal., 1993). Relationships between these clades tions within an extinction-resistant clade (Erwin are not robustly resolved, suggesting that diver- 12 M.G. Harase\v)'ch et al. Figure 7. A comparison between the phylogenetic information contained in the length polymorphisrns of the insertions unique to Pleurotomarii- Eníemnoírochus dae with that contained in their se- quence. (A) Consensus of 9 trees (L = 27; CI = RI - RC = 1.0) based only on sequences of the inserts. (B) Perotrochus Group "A" Consensus 3 trees (L = 22; CI = 0.864; Rl = 0.885; RC = 0.764) based on an analysis of the distribution of insert segments among pleurotoma- Perotrochus tiid taxa. Bootstrap proportions given Group "B' as percentages above the nodes and A B jaclclsniFe proportions given as per- centages below the nodes. An asterisl< (*) indicates that the node was not supported in a 50% majority rule bootstrap or jackknife consensus tree. 1 111111111 for their monophyly. Our data do not strongly sup- Insert Segment 1234567890 123456789 port the monophyly of Cocculiniformia, but do indi- Asiratia cate that Cocculinidae and Lepetelloidea are more E. adansonianus + + + + + + + + + - -- + + - + + + + closely related to each other and to the patellogas- E. rumphii + + + + + + ++-Í-- -- + + - + + + + P. quoyanus + + + - + + + + -- + + + + - + - + + tropod limpets than to Neritopsina or Vetigastro- P. lucaya + + + - + + + + -- + + + + - + - + + poda, respectively, as suggested by Ponder and P. maureri + + + - + + + + -- + + + + - + - + + P midas + + + - + + + + + + + + + ++ + - + + Lindberg (1996, 1997). P. leramachii 4-J. + - + + + + + + +-- + + + - + + Manipulation of the tree to produce a monophy- letic Cocculiniformia emerging basal to Neritopsina Figure 8. The distribution of 19 insert segments among Pleurotomariidae. These insert segments (characterized (ITaszprunar hypothesis) increases the tree length in Table 3) are absent in Astraea and all nonpleurotoma- by 9 steps. Placing Cocculina as a sister taxon to riid taxa within our study. Inserts are scored as present Neritopsina lengthens the tree by 11 steps, while (-I-) or absent (~) without regard to their sequence. making the lepetelloidean limpet Notocrater the basal vetigastropod increases the tree length by 17 steps. The most parsimonious placement of Noto- gences were ancient and rapid. Philippe et al. (1994) crater within Vetigastropoda is as sister tajxon to the calculated that rapid (S40 Million-year) radiations nonpleurotomariid vetigastropods, which increases are beyond the limits of resolution of the entire 18S tree length by 12 steps. Although data are lacking rDNA gene. for Cocculina, Notocrater contains a portion of the Perhaps the most anomalous result of our study large patellogastropod insert (positions 16-20), add- is the grouping of the cephalopod Nautilus in a ing support for its relationship to the patellogastro- clade together with the patellogastropod and coc- pod limpets. culiniform limpets. Although the node connecting It should be noted that all the taxa grouped in Nautilus to the limpet taxa is not strongly corrobo- this clade are anomalous in containing insertions rated by either bootstrap or jackknife proportions, that are highly variable in sequence and length, moving Nautilus to its accepted position as a sister difficult to align, of uncertain homology, and are group to the Gastropoda increases the length of the absent in other gastropods as well as in the chitons. tree by 9 steps. The monophyly of patellogastropod Although these insertions were excluded from the limpets is strongly supported (bootstrap and jack- data set and did not contribute to the structure of knife proportions of 100). The two species in our the tree, we conjecture that these ta^a may have study contain uniquely shared insertions (e.g., posi- undergone inversions, translocations, or changes tions 16-27, 32, 86, 643) and a deletion (position other than point mutations in this region of the gene 62) that provide additional, independent support that are not homologous among all taxa. Branch Molecular phylogeny of pleurotomariid gastropods 13 i'rAf.r-íxr.íAciíU.'i'it;^;; nj-x; 1 ï vrrrriÄ;!:;: wr.raes c.T.'Tr.. . .G. .AC.A. G.A. . . .A. . />^-. i .A. .A. . / •.. .cr.AG.G. .G. .c:. . . .1- . .G .A t'-J £:. acfensoniñiíus c.c.(7rG. . . .G.CTC.A. .-rc. . . .A. . . .A. ,A. .A. .A. .C. ..Ar. :. .A. C.7. ... .c .."G.G.. C •íí . . . .GTT, . . .G.CC. .A. .c.. . .A. . ..AC. . . .A. ..T.O.^., .A. c. :. .G.. . . . i.. A, . " A...O!'... . .A.r-. .A. .c.. . . A . . . .A. .C. .A, . . .c. .A. .T. . .':-.ü. .G. .0. .C. G.T. ...G. ...'• "• C 99 P. iiiufes' A,..cTr.., . .A.r?. .A. .r . . .A. . . .A. r^. .A, . . .c. .A. .T. . .7.r;..G. .G. .c. C.T. . . .G. . . .• • C 99 p. TOijster; A. . .GT. . . •. .A.CT. .A. . r.. . .A. . . .A. .c. .A. .(• .A. .A. .A. .0. ..• ... .G. .c. .a:, i . . . .G.C. .' •. .C, . . 100 p. quoyoniis C...AT... '. .ce:-. .A. . .A. . . .A. .c. .A; .r. .A. .A. .A. .A. .C. .AC. .. .G. .C. .,A.'.T. . . .G. . . .• ..C... 10Ö P. lucays C...GT... -. .;;.C7. .A. . .A. . . .A. .c. .;c.G. .A. .A. .A. .A. .C..AC... .G. .C. .C'.T. ...G...... C. . . lOC CeriLhiíJir, . .G.A:T. . •crc.c. .TC. . .A. . . . " . .c. .A. . . . .A. .T. . C. . . C. T. ... .O':... . .A . .G. .A J9 110 120 130 140 170 180 190 200 DicóDra CA'rGC^H'iTnT..va'AATiTi':^!';'-U7irr.^G\;GCxGrATGATAA'rixKaKATiTQ3GAACiGA':"AGiTCG'^ .Ai'•rwr Ascraea . .C 'I' C. .A. .A C. .T T T. .T. .C.íG. .A. . .C.'í'. .A ' C. .C. .ACT. E, fdar.sQ'n i ,ín;i G C A. .A. .A. .A. .A A. .T 'i' ( TG GG. . . .AC. . . .A. .C , .ACG. E. fim^^i A A. .A. .A T. .T r. .T. . .GlC. .G. .(G 'IG.C.. .A A. .1 . .AG.G. 200 ÎV ccrsrrachjl C..T C G. .G A. .A A A. .T. .(JCl. . .A. .CC.T. . .C. ; . .G. .C. .G. .G. .AG. . 195 P. niidas C. .T C G. .G A. .A A A. .T. .GCl. . .A. .GC.T. . .G. ; ..G. .G. .G. .G. .W3.T. 199 P. [íBueret 1 C C. . . .A. .A. -A. .A. .A ...A.. . .C. .A. T. .aT. . .A. .te.T. ..AG.G. .A. .G. .G. .G. .AC.T. 200 P. cfucyaniis . .A . .C .Ai'.G. .G. .C. .C. 300 P. ] ^jcays C. . . .A. .fC7. .Ai-.C. .G. .G. .G. ino Cerithium C.. • • c.. T.A. . .A. .A. .A...... T. . .C. T. . . .'.-G .(T.\ . .... G. .A. ..-•G. T. 199 1D[) Diodorra TTCCTCmc "TAATAA " hTmXV -T03aGrrGCClX;GTrCV>i •• TAA " TprcTTATAIlC 'TCKiGViQiaTriGiiAaGGQaïGCTCaA' i Ci ; ITTi .V Astraea .T. .G. .T.IT. .AC.ce.T. .G. . .G.AC. .TIG"" .C.r'TT.CT ITT. .T. .A. .T.T" .G. E. adansoniâiUi;: TC. ..AT.AC.A. . . . .cr. .G.G.A. . . .T.GTJG. .A. .T. .A. .A A.T~ G. .A. 299 E. r^jnçhii .G. .A.T;'. .A:\AC .AT. .OCC.AC. . .T.GCA. .G. .G A A. . . . 1. . .T 3 DO P. CfîraJivichii . . . .A.Ti . .Al'. . . .A. .A. .AT. .G.. .AcrnroGA T. .A. .G •..'ic. .A. .G. .G. P. miitis .Y. .C. . . . .A.T. . .AV. . . .A. .A. .AT. .G. . .AC. IGIGÜCA T. .A. .•."; A.'". .A. .C. .C. 299 p. rrtîL^erErL .G..G. .G. .A.'iG .... .AG. : . .G. ;CT.GC. .AC.G.- .Gcr. .T. .A. .A. .G. .T- JOB p. quQyanUi .G. .A.TT. .AG.l. .G. .GC. .AC.C.G.GCG T. .A. .A. .G. 30(1 P. lucays .T. .G. cerlthium .T.TG. .AC.T. .A. 310 320 J30 340 JÜO 3 Si 390 ': • : ; ATTG-XC; •• '?CJPG1TJ1AAA•GGGTCATGCTCK " CGTlGTCmGAWPrAACTAl-.' • •'P.'TCTTIGQ'. : " TGQ'^. • GiGTG"! CTTGT.ATTT: AGCr: 397 A A. .T ....-•.. .G.:G"A C..TCGC. . ". .A-.TC.ÜG. . .TG . .A-GG^.'". . . • A. .G G .19.1 E. 3íÍLnson.L¿ijiu¿ A .G CT.AG. .A.C. .ce C. .C. .T. .A. .A. .G. .T. . .G. . .Tl .G. . .G. . . .G. .A A. .T. .G. .G. . .C. . . .G 399 E. rijnpliii A. .A. .G. .A. .AT.A.G. .A.C. . .C C. .G. .C. .A. .A. .A. .¡C. .G.G.'ri C.T A. .A. .A. . .G. 4D(J 1^. tcraiTïich ;• i C. .G A.'..%,GAA. . . .C. .A A. .C A. .A. .T. , .G.A.TC A. .A. .C. A. .A. .AA. .G. 399 P. nudas C. ...AT.AG.AA....G..A A..G A. .A. .T...G.A. G A..A..C..A..A..AA.T G A 399 P. rauerer;. A. . . .AT.AG.AA C. .A. .•:• G. .A. .T. . .G.A. .:'; OC.A. .G. .A. .A. ,AA,T G. . .G. . . .G 400 P. quciyonuE A. . . .AT,A,G.GA A. .G G. .A. .T. . .G.A.-G.' A. . A . . ". . A. .G. . AA. 1' G. . .G. . . .G ÎOO P. lucai'a A. . . .Ar.A,G.GA A. .G G. .A. .T. . .G.A. ]'; . .G. .A. .A. .G. .A. .G. ..AA.T G. . .G. . . .G 430 4R(; 490 D' octari 497 Ai 1 raea ' \ T.'^. .7. .A. "A. .AG.ff: 'G ... •... :. .G. .a:.. . . ¡GGA T. .G. . • A. ..." 49-1 E. odariËOiiiani . .^ . .A. . . .A. ./; "" ^^r .• ;.. . .AG. .7 G. .•.G. .AL.A. . . .T. .A. ,s. .G G. .G. 499 E. . .C. .T. .C. .G. .C. .G. .AA.'^. .G. .AG.,;...- . .A. . .G. .GC.A. .AV.A. . . .T. ,G. .A. .T. . A-- .G. . .C son V. teraiTöchii .T. .G. .G. ..AA.G. :r.. .AC.G .A. . . .AG. .1 . .A. . .'i-.A. .AC.A. . . .T. .A. .A . GG •;59 P. .T. .G. .G. .A.A.C. .T. . .AG... .A. . . .AG. .T. .G. . .'i .A. .A.'.A. . . .T. .A. .A. .( :• •199 p. fTHtícrer] .T. .AA.T. .T. . . .A. .AG.A, .G. . . .AG. .T . .'l'.A. .G;.A.. . .T. .G. .G. .A. .Gi. 'l'Jlî P. quoyarLu£ .A. .AC..AG.G. . . .AG. .7. .A. . .7.A. .G: .A T. .G. .A A G. .A. .00 500 P. luqaya .A. .AG..AG.G .-..G. .-1 . .A. . .'V.A. .GT.A T. .G. .A A G. .A. .CC 500 Ceri c^um .A. .AGG..-.G.A AG. .G. .G. .GC. . . .TC.T G. .A AA.G. .T. .CG. .C 499 540 'JÉO •J.cdois. TATK7n7i'L'IG;CGTlGGCGGríGTTC<.C'?G0a:CT.ATT.ACr.A7Gn'TÜrJÍ«»3A71GT.A¡iTrr.AAT.A7:'T>; Ascraea . . . .AL'.ce.A A. .A. .G AG. E. fldansorianu.- . .C. ". .G. .A. .GC.T A. .A. .C. .A. .A A. . .G.-: . .A. . . . .'.'). .G. .A. . runphii . . . .A. . .G.G. .AC.i . .A A. .C. .A. .A A. .ACT. .A. ....G..G..A.. cerflTTïîcli 1 i .T. .AC.T. .A. . .lA ¡'. .A. .AC.A. .G. .A. .A A. .,AC. .C.T. .C C .T. .AC.T. .A. . .G. r. .A. .AG.A. .f;. .A. .A A. .AC. .C.T. .G G .G. ..V:.T. .A. . .G. /. .T A A. .G. .G. .A. .A. .AG.A. qijnyariJJ, .T. . .C.T. .A. . .C.T A G. .C. .A. .A. ..AG.A. .G G. ;-.'9 lucaya .T. . .C.T. .A. . .C. i ....".. .A G. .G. .A. .A. .AG.A. .C C. Gerithr^ïïTi .T. .AG.TC.T. .ce. . . .T. .AC.G. .A. .T G. iG.T. T. .G. .G. .G. . . . .AG. . . .G. 578 Figure 9. Aligned partial sequences of the moUuscan mitochondrial cytochronie c oxidase subunit I gene corresponding to positions 32 to 611 of the COI sequence of Drosophila yakuba (Folmer et ah, 1994, Table 2). All sequences were confirmed by at least t-wo sequencing reactions in each direction. Ambiguous base assignments are noted using lUPAC symbols. Dots (•) represent nucleotides identical to those of Diodora, and quotation marks (") represent gaps inserted during alignment. 14 M.G. Harasev\ych et al. (1988) and Hickman (1988), and supported by lim- ited 28S RNA sequences (Rosenberg et al., 1994). However, when Polyplacophora were excluded from the data matrix and Nautilus was specified to be the outgroup, Neritopsina emerged between Vetigastropoda and Apogastropoda as advocated by • F JUiflSûntBfiiK Ponder and Lindberg (1996, 1997). Thus, 18S se- quence data appear insufficient to resolve conclu- P. 'ff.',l'mcnn sively the relationships of Neritopsina. The monophyly of the Pleurotomariidae is strongly supported by 18S sequence data as well as by the presence of seven insertions unique to the family. Contrary to the predictions of the morph- ology-based phylogenies, the Pleurotomariidae B emerge as the sister group of the remaining vetigas- tropods, which also comprise a highly robust clade, Figure 10. Phylogenetic relationships among Recent rather than as a sister group to the family Trochidae. pleurotomariici species and genera. [A) The single most Remarkably, the monophyly of the Vetigastropoda, paisimonious tree resulting from an analysis (exhaustive which is well supported by morphologic evidence search option) of 579 bp fi-om the mitochondrial cyto- (Salvini-Plawen and Haszprunar, 1987; Ponder and chrome c oxidase 1 gene (L = 540; CI 0.669; RÍ = 0.581; Lindberg, 1996, Figure 11.2), is but weakly sup- RC = 0.388). Bootstrap proportions given as percentage above the node and jackknife proportions given as per- ported by sequence data (bootstrap and jackknife centage below the node; (*) indicates that the node was not proportions < 50% with both raw and reweighted supported in a 50% majority rule bootstrap or jackknife characters), suggesting an ancient divergence be- consensus tree. (B) The single most parsimonious tree tween the two clades. Although Pleurotomarioidea resulting from an analysis (exhaustive search option) of date to the Cambrian, the remaining taxa appear the "total molecular evidence" of 1154 characters (556 later•Trochoidea first appear in the Ordovician, 18S rDNA + 579 COI + 19 insert) (L = 712; CI = 0.729; and earliest records of Haliotoidea and Fis- RI = 0.658; RC = 0.479). Bootstrap proportions given as surelloidea are from the Triassic (Tracey et al., 1993). percentages above the nodes and jackknife proportions Reanalyses of the 18S rDNA sequence data from given as percentages below the nodes. a subset of taxa for which the sequences could be reliably aligned in their entirety (Figure 6) resulted lengths for these taxa are two to three times as long in an identical division of Vetigastropoda into pleu- as for other species in the tree, raising the possibiHty rotomariid and nonpleurotomariid clades, with that they may be grouped on the basis of long branch each clade strongly supported by bootstrap and attraction. Several authors (e.g., Termier and Ter- jackknife proportions of 100. However, monophyly mier, 1968; Shileyko, 1977; Ponder and Lindberg, of the Vetigastropoda remains weakly supported. 1997; TiUier et al., 1992) have speculated that the The relationships of the pleurotomariid taxa to each patellogastropod limpets may not be gastropods, other as well as to other vetigastropods, predicted and suggested that the monophyly of the Gastro- by the 18S sequences, were confirmed by sequence poda requires further corroboration. However, coc- from the mitochondrial gene for cytochrome c oxi- culiniform Umpets, which together with the dase I (Figure lOA). The phylogeny based on this patellogastropod limpets comprise a monophyletic gene sequence confirms the monophyly of the Pleu- group in our study, have been regarded as unambig- rotomariidae and supports the monophyly of the uous gastropods in all previous classifications. nonpleurotomariid vetigastropods, specifically re- The monophyly of the Neritopsina is strongly futing a close relationship between Pleurotomarii- supported, perhaps in part because both representa- dae and Trochidae. tives in our study are members of the same subfam- There have been differing opinions among mor- ily (Neritinae). The position of the Neritopsina phologists about the relationships of vetigastropod within Gastropoda is less certain and strongly influ- taxa. McLean (1984] commented that the Tro- enced by the outgroup. When Polyplacophora choidea "share so many features with the Pleuroto- served as the outgroup, Neritopsina emerged basal mariidae that their derivation from the group is to Vetigastropoda, as hypothesized by Haszprunar readily perceived." However, Hickman (1984) Molecular phylogeny of pleurotomaríid gastropods 15 noted that the radula of Pleurotomariidae "had little tirely concordant with the morphology-based in common with any living gastropod, including hypotheses of Bayer (1965). The inclusion of both the [then included] two other pleurotomariodean Japanese [P. teramachii) and Caribbean (P. midas) families [Scissurellidae, Haliotidae] of more recent large-shelled species in a clade separate from the geological appearance." While McLean (1984) hy- small-shelled species indicates that the divergence pothesized that the Haliotidea are the "limpet-like between Perotrochus "Group A" and "Group B" oc- derivatives" of the Pleurotomarioidea, Haszprunar curred before the closing of the Tethys during the [1985) noted that Haliotidae and Trochoidea are Oligocène. further linked by shared aberrant chemosensory The clade comprising Apogastropoda, with each structures. of the constituent groups (Caenogastropoda and In the light of the molecular evidence, the mor- Heterobranchia) broadly represented, is well sup- phologic data and analyses of vetigastropod phylog- ported by the 18S rDNA sequence. Monophyly of eny should be reappraised with special attention the Caenogastropoda is more robustly supported to the precise composition of higher taxa and the than that of the Heterobranchia. Although unsup- resulting distribution of characters. The usage of ported in the strict consensus tree, Heterobranchia Pleurotomarioidea has been inconsistent in molecu- emerges as a weakly supported clade in bootstrap lar as well as morphologic studies, Tillier et al. and jackknife analyses (61% bootstrap and 61% (1994), for example, concluded that "Fisurelloidea jackknife proportions). The seven heterobranch taxa is the sister group of Pleurotomarioidea and Tro- in this study contain 15 to 20 more nucleotides in choidea" based on 28S rDNA sequences, because the E-10-1 helix than do any of the caenogastropods, they used Haliotis as the exemplar of Pleurotomari- providing further evidence for their monophyly in oidea. Thus, what appears to be a contradictory find- the form of unique insertions. Resolution of phylo- ing is, in fact, concordant with our resrdts. genetic relationships within Apogastropoda is the Within Pleurotomariidae, the monophyly of the subject of ongoing research, and will be reported in genus Entemnotrochus was well supported by 18S a separate paper. sequences, even without data contained in the in- The gene coding for the small subunit ribosomal sertions. The insertions were required to corrobo- RNA (18S RNA) was the first to be used for se- rate the monophyly of Perotrochus, as well as its quence-based phylogenetic inference (Field et al., small-shelled (Perotrochus "Group A," Bayer, 1965) 1988), and has subsequently been applied primarily and large-shelled [Perotrochus "Group B," Bayer, to resolve relationships of phyla and classes. WiÜiin 1965) clades. Information contained in the se- Mollusca, this gene has been applied to studies of quences of the seven inserts unique to Pleurotomari- bivalve evolution (Rice et al., 1993; Kenchington et idae was sufficient to distinguish Perotrochus and al. 1994), in which it has successñüly resolved fam- to resolve it into small-shelled and large-shelled ily and in some cases genus-level taxa. The present groups, while length polymorphisms of these in- study, which represents the most dense and wide- serts were able to differentiate Entemnotrochus spread sampling of taxa within a single molluscan from the large-shelled Perotrochus (Figure 7). By class published to date, reveals the 18S gene to be contrast, the nonpleurotomariid vetigastropod fam- capable of broad but uneven levels of taxonomic ilies could not be resolved by 18S sequence data. resolution, in some instances being able to resolve Cittarium, which emerged anomalously at the base genera, in others failing to resolve superfamilies and of the clade, could be returned to its predicted posi- orders. An empirical observation is that higher taxa tion as sister taxon to Astraea by increasing tree that contain insertions are more finely resolved than length by only one step. those taxa that do not, even when the insertions are Cytochrome c oxidase I sequence data strongly not included in the analyses. It further appears that support the monophyly of the genus Perotrochus larger insertions produce finer resolution than and of its constituent subclades Peroirocii us "Group smaller insertions, but this hypothesis requires con- A" and Perotrochus "Group B," and are capable of firmation. A corollary of this conjecture is that the resolving closely related, parapatric species. Ivlanip- 18S gene should prove more useful for resolution ulation of tree topology to make the genus Entemno- of phylogenetic relationships among cephalopods, trochus monophyletic increases tree length by one patellogastropod and cocculiniform limpets, and step. The single most parsimonious tree resulting heterobranchs, all of which contain insertions, and from the exhaustive search analysis of the "total less informative about relationships among Caeno- molecular evidence" is extremely robust and en- gastropoda and nonpleurotomarioidean vetigastro- 16 M.G. Harasewvch et al. pods, which lack insertions. It is possible that species within Pleurotomariidae. A similar ap- Caenogastropoda and nonpleurotomarioidean veti- proach may be fruitful for resolving relationships gastropods may have insertions in other variable within groups such as Caenogastropoda that have regions of the 18S gene that would increase the piovedrefractory to resolution by IBS data. Alterna- resolving abilitj' of those regions. tively, the addition of data for one or more slowly The distribution of variable and conserved re- evolving genes may enhance the resolution of gions in the 18S RNA molecule was described by deeper branches within Gastropoda, Mollusca, and De Rijk et al. (1992) and later refined by Winnepen- Metazoa. ninckx et al. (1992). Kenchington et al. (1994) iden- tified variable regions on a finer scale that were useful for resolving relationships within the class Experimental Procedures Bivalvia and noted a region of potential utility for Sample collection and preparation studies of Gastropoda. They emphasized the neces- All of the species sequenced were collected ex- sity of determining the entire gene sequence for pressly for this study, frozen while living, trans- phylogenetic studies in order to capture data from ported to the laboratory on dry ice, and maintained all variable regions throughout the gene. Recog- at - 80°C until DNA was extracted. Specimens were nizing the inevitable conundrum of molecular sys- collected by a variety of techniques including hand tematists, Lecointre et al. (1993) assessed the collecting, trapping, dredging, trawling, or use of comparative benefits of adding sequence length ver- the Harbor Branch Océanographie Institution's sus those of adding taxa and concluded that the JOHNSON-SEA-LINK and CLELIA manned sub- impact of species sampling on tree topology and mersibles. A list of taxa, collection localities, tissue robustness is greater than the impact of sequence extracted, and voucher specimen information are length. These authors recommended that sequenc- given in Table 1. ing efforts be modified to sequence fewer nucleo- tides per species while increasing the number of DNA extraction representatives of presumed monophyletic taxa and to include several outgroups, an approach that we Specimens were thawed and maintained on ice have followed in the initial phase of this study. A while the buccal muscles were dissected out. Small dense sampling of taxa also allowed us to recognize shreds of muscle (totaling 1-2 mm=') were placed in patterns in the distribution and lengths of insertions a 1.5-ml Eppendorf tube containing 300 ¡xl of CTAB and deletions among related taxa and to interpret buffer at 60°C and ground using a disposable pestle these variations in a phylogenetic context. (Kontes catalog no. 749520) in an electric drill. The resolution achieved by combining data sets When the animal was minute, the shell was from multiple genes demonstrates the utility of an crushed, shell fragments were removed with for- iterative, multistep, multigene approach to phylog- ceps, and the entire remaining tissue mass was used eny reconstruction. In the initial step, the phy- for extraction. Individual animals were used for all logenetic relationships of the taxon of interest extractions. The extraction protocol follows Adam- (Pleurotomariidae) were assessed in a broad con- kewicz and Harasewych (1996), with the following text. The 18S rDNA gene proved to be sufficiently modifications. The initial incubation at 60°C was variable to resolve many of the relationships among extended to between one and two hours, and the higher taxa within the molluscan class Gastropoda, tissue was ground again after 30 to 45 minutes. A especially in those taxa that contain one or more second chloroform-isoamyl alcohol extraction was large insertions. Both sequences and length poly- added. All centrifugation steps were extended to 10 morphisms within the insertions fmther improved minutes at 10,000 rpm. Absorbances at 260 and 280 resolution. After monophyly of the Pleurotomarii- nm were measured for all DNA preparations. When dae had been established and appropriate sister taxa the ratio of absorbances at 260/280 nm was less than identified, finer-scale relationships on the order of 1.5, the preparation was treated with proteinase K genera and species were resolved with supplemen- and reextracted. tal sequence data from the more rapidly evolving mitochondrial gene. "Total molecular evidence," Primer selection consisting of partial sequences of IBS rDNA and Sequence data for taxa in Table 1 were obtained COI and including data on length variation within from a region of approximately 450 bp at the 5' end the inserts in the IBS rDNA gene, was sufficient to of the small subunit of ribosomal DNA (IBS rDNA) resolve fully the relationships of the genera and that contains the variable regions Vl and V2 (De Molecular phylogeny of pleurotowaaid gastropods 17 Rijk et al., 1992], reported b^' Kenchington et al. automated DNA sequencers using fluorescence Dye (1994) to be useful in inferring molluscan phy- Terminator Cycle Sequencing kits. Sequence data logeny. The primers (forward, 5'-GCCAAGTAGCA- were edited and uploaded to a relational database TATGCTTGTCTC-3', and reverse, 5'-AGACTTGCC- called PHYLO. 18S rDNA target regions were se- TCCAATGGATCC-3') were from Holland et al. quenced and confirmed by at least two DNA se- (1991). Two additional, internal primers were de- quencing reactions in one direction and a third signed by the authors for sequencing (forward, 5'- reaction in the opposite direction. The COI target GGCGGCGACG(T/C)ATCTTTC-3', and reverse, 5'- regions were sequenced and confirmed by at least GAAAGAT(A/G)CGTCGGCGGC-3']. Sequence data two DNA sequencing reactions for each primer. Re- for pieurotomariid taxa and selected outgroups gions of ambiguity are noted using the lUPAC sym- were also obtained from a 579-bp fragment of mito- bols. Sequence fragments for each gene were chondrial cytochrome c oxidase subunit I (COI) downloaded from PHYLO and assembled into a using the "universal" primers for COI of Folmer et consensus sequence using modified TIGR Assem- al. (1994) (forward, 5'-GGTCAACAAATGATAAA- bler software (Fleischmann et al., 1995). The con- GATATTGG-3', and reverse, 5'-TAAACTTCAGGG- sensus sequence was edited using a modified ABI TGACCAAAAAATCA-3'). Sequence Editor software. Sequences have been posted to the Genome Sequence Data Bank (URL PCR amplification http://www.ncgr.org), and accession numbers are PCR amplifications were performed using a Perkin- included in Table 1. Elmer 480 thermal cycler. PCR reaction mixtures Multiple sequence alignment had a total volume of 50 |JL1, containing 200 to 500 ng of genomic DNA, 1.25 U of Amplitaq Taq poly- The confirmed consensus sequences from each or- merase (Perkin-Elmer), 200 |xm of each dNTP, 0.25 ganism were dov\mloaded from PHYLO in multiple |JLM of each primer, 1.5 raM MgClz, and 5 ¡xl of FASTA file format (Pearson and Lipman, 1988) and a lOX Perkin-Elmer PCR reaction buffer. All PCR viewed with Genetic Data Environment (GDE) ver- reactions were performed as "hot start" reactions sion 2.0 (Smith et al., 1994). The sequences were with the following thermal cycler parameters: five aligned using the multiple sequence alignment algo- minutes at 95°C, addition of Taq polymerase, 30 rithm (MSA) of Sutton (Bult et al., 1997). For the cycles of 45 seconds at 94°C for denaturing, two protein coding gene, COI, the "protein assisted" minutes at 50°G for annealing, and one minute at feature of MSA was employed, which uses an opti- 72°C for extension. Following the 30 cycles, a five- mal alignment of amino acid translations of the minute final extension at 72°C was performed and genes to guide the nucleotide alignment. Aligned the PCR reaction mixtures were then held at 4°C sequences are available in nexus format via FTP until use. For each PCR product 5 ¡xl was evaluated at ftp.tigr.org/pub/data/gastropod_18s.aln (18S on a 1% (w/v) agarose gel. If a PCR reaction was rDNA sequences) and ftp.tigr.org/pub/data/gastro- imsuccessful, 1 ^x.\ of the initial reaction mixture pod_col.aln (cytochrome c oxidase I), or from the was used as template for a second round of amplifi- corresponding author. cation. The thermal cycler profile for reamplifica- tions was the same as for the initial PCR with two Phylogenetic analyses exceptions: the number of cycles was decreased Phylogenetic analyses using parsimony were con- from 30 to 20 and the annealing temperature was ducted using PAUP 4.0.0d45 (Swofford, 1996; beta raised from 50°C to 52°C. If a single band was ob- test version) on a Power Macintosh 6100/66. All served, the reaction mixture was transferred to a characters were treated as unordered. For 18S rDNA Microcon-100 microconcentrator (Amicon, Inc.] for sequences, autapomorphic insertions (present in a the removal of primers, unused dNTPs, and salts. single taxon) and regions of uncertain homology, An initial wash with 200 |xl of 40% (v/v) isopropa- including insertions not present in all taxa (Figure nol was performed prior to a wash with 450 |j,l of 4B], were excluded from the data set prior to maxi- sterile, distilled water. Samples were concentrated mum parsimony analysis using the branch and to a final volume of 10 to 20 jxl, and had a DNA bound search algorithm. In subsequent analyses, concentration of 50 to 800 ng/[xl. subsets of taxa with fewer regions of uncertain ho- mology were analyzed using maximum parsimony DNA sequencing and data management and the exhaustive search option. Bootstrap and Purified PCR products (100-200 ng DNA per reac- jackknife analyses (lOOO replicates) were performed tion) were sequenced on Applied Biosystems 373A using the "fast" stepwise addition option. 18 M.G. Harase\'\ych et al. To compare 18S rDNA sequence-based gastropod Jon Norenburg, Department of Invertebrate Zoology, phylogen)' with prevailing morphology-based phy- National Museum of Natural History, Smithsonian logenetic hypotheses, subsets of the data matrix Institution; Dr. Peter Fankboner, Simon Fraser Uni- were created by pruning to contain taxa overlapping versity; Dr. Bruce Saunders, Bryn Mawr College; Dr. with the morphology-based trees under consider- John B. Wise, Houston Museum of Natural History; ation (taxa appearing in upper case letters in Figure Dr. Paula Mikkelsen, Delaware Museum of Natural 2). When higher taxa were represented by multiple History. Special thanks are due Mr. Yoshihiro Goto species in the molecular data set, a single represent- and Mr. Paul Callomon for providing frozen speci- ative containing the minimum number of deletions mens of Japanese pleurotomariids as well as for and ambiguous base determinations was selected. making possible the senior author's field work in Following analysis (branch and bound option] of Japan. The assistance of Ms. Molly Ryan with the the reduced subset, the most parsimonious tree that illustrations is gratefrdly acknowledged. We are es- most closely resembled the topology of the mor- pecially gratefrd to Dr. David L. Swofford of the phology-based tree to which it was being compared Laboratory of Molecular Systematics, Smithsonian was imported into MacClade 3.01 (Maddison and Institution, for providing access to various beta test Maddison, 1992), where branches were rearranged versions of PAUP 4.0. This is Smithsonian Marine to correspond with their position in the morphol- Station at Link Port Contribution Number 421. ogy-based tree. Resulting changes in tree length and character indices were noted. References Phylogenetic signal contained in the seven inser- tions unique to Pleurotomariidae was assessed by Abbott, R.T., and Dance, S.P. (1982). Compendium of Sea- analyzing only the information in these insertions. shells. New York: E. P. Duttoa. Adamkewicz, S.L.,and Harasewycli.M.G. (1996). Sj'stematics A phylogeny of the seven Recent Pleurotomariidae ajid biogeography of the genus Donax (Bivalvia: Donaci- was calculated from these isolated insertion se- dae) in eastern Nort)i America. Am Malacolog Bull quences. Individual insertions and deletions were 13(l/2):g7-103. then defined in terms of 19 segments (ranging from Anseeuvv, P., and Goto, Y. (1996). The Living Pleurotomarii- 1 to 22 nucleotides) that varied between the pleuro- dae. Osal commanding. 29. Report on the MoDusca, Part 2. Gastro- Hyman, L.H. (1967). The Invertebrates, vol 6, MoUusca 1. poda and Scaphopoda. Bull Museum Comp Zöol Harvard New York: McGraw-Hill Book Co. 18:1-492, pis. 10-40. Kenchington, E.L., Roddick, D.L., Singh, R.K., and Bird, Cf. De Rijk, P., Neefs, J-M., Van der Peer, Y., and De Wächter, (1994). Analysis of small-subunit rRNA gene sequences R. (1992). Compilation of small ribosomal subunit RNA from six families of molluscs. / Mar Biotechnol 1:215- sequences. Nucleic Acids Res 20 (suppl):2075-2089. 217. Emberton, K.C, Kuncio, G.S., Davis, G.M., Phillips, S.M., Knight, J.B., Cox, L.R., Keen, A.M., Batten, R.L., Yochelson, Monderevvicz, K.M., and Guo, Y.H. (1990). Comparison E.L., and Robertson, R. (1960). Molhisca-Gastropoda, sys- of recent classifications of stylommatophoran land-snail tematic descriptions. In: Treatise on Invertebrate Paleon- families, and evaluation of large-ribosomal-RNA sequenc- tology, part I, MoUusca 1. Moore, R.C. (ed.). Lav\Tence, ing for their phylogenies. Malacologia 31:327-352. KS: University of Kansas Press and the Geological Society Endo, H. (1995). Slit Shells: Pleurolomariidae. ivianazuru, of America, 169-331. Japan: Endo Museum. Lecointre, G., Philippe, H., Le, H.L.V., and Le Guyard, H. Erwin, D.H., and Signor, P.VV. (1991). Extinction in aji extinc- (1993). Species sampling has a major impact on phyloge- tion-resjstajitclade: the evolutionary history of the Gastro- netic inference. Mol Phylogenet Evol 2(3):205-224. poda. In: The Unity of Evolutionary Biology. Proceedings Maddison, W.R, and Maddison, D.R. (1992). MacClade. Anal- of the Forth International Congress of Evolutionary and ysis of Phylogeny and Character Evolution, version 3. Sun- Syslematic Biology, l Dudley, E.G. (ed.). Portland, Ore: derland. Mass: Sinauer Associates. Dioscorides Press, 152-160. jVlcLean, J.H. (1984). A case for the derivation of the Fissurelli- Field. K.G., Olsen, G.|., Lane, D.J., Giovanonni, S.]., Ghiselin, dae from the Bellerophontacea. Malacologia 25(l):3-20. M.T., Raff. E.G., Pace, N.R., and Raff, R.A. (1988). Molecu- McLean, J.H. (1990). Neolepetopsidae, a new docoglossate lar phylogeny of the animal kingdom. Science 239: limpet family from the hydrot)iermal vents and its rele- 748-753. vance to patellogastropod evolution. / Zool Land Fleischmann, R.D., Adams, M.D., White, O., Clayton, R., Kirk- 222:485- 528. ness, E.F, Kerlavage, A.R., Bult, G.J. et al. 1995. Whole- Morton, J.L., and Yonge, CM. (1964). Classification and struc- genome random sequencing and assembly of Haemophi- ture of the moJIusca. In: Physiology of the MoUusca, vol 1. lus influenzae Rd. Science 269:496-512. Wilbur, K.M., and Yonge, CM. (eds.). New York: Academic Folmer, O., Black, M., Hoeh, W., Lutz, R., and Vrijenhoek, R. Press, 1-58. (1994). DNA primers for amplification of mitochondrial Pearson, W.R., and Lipman, D.J, (1988). Improved tools for cytochrome c oxidase subunit 1 from diverse metazoan biological sequence comparison. Proc Nat Acad Sei USA invertebrates. Mol Mar Bid Biotechnol 3(5):294-299. 85:2444-2448. Ghiselin, M.T. (1988). The origin of molluscs in the light of Philippe, H., Chenuil, A., and Adoutte, A. (1994). Can the molecular evidence. In: Harvey, P.H., and Partridge, L. Cambrian explosion be inferred through molecidar phy- (eds.) Oxford Surveys in Evolutionary Biology, vol 5. New logeny? Development 1994 suppi: 15-25. York: Oxford University Press, 66-95. Ponder, W.F, and Lindberg, D.R, (1996). Gastropod phylog- Golikov, A., and Starobogatov, Y.I. (1975). Systematics of pro- eny•challenges for the 90s. In: Origin and Evolutionary sobranch gastropods. Malacologia 15(l):185-232. Radiation of the MoUusca. Taylor, f.D. (ed.). Oxford, UK: Graham, A. (1985). Evolution within Gastropoda: Prosobran- Oxford, University Press, 135-154. chia. In: The MoUusca, vol 10, Evolution. Trueman, E.R., Ponder, W.F,, and Lindberg, D.R. (1997). Towards a phylogeny and Clarke, M.R. (eds.) Orlando, Fla: Academic Press, of gastropod molluscs•an an analysis using morphologi- 157-186. cal characters. Zool ] Linnean Soc (in press). Harasewych, M.G.. Pomponi, S.A., and Askew, T.M. (1988). Rice, E.L., Roddick, D., and Singh, R. (1993). A comparison Spongivory in pleurotomariid gastropods. Nautilus of molluscan (Bivalvia) phylogenies based on palaeonto- 102(3):92-98. logical and molecular data. Mot Mar Biol Biotechnol Haszprunar, G. (1985). The fine morphology of the osphradial 2:137-146. sense organs of the MoUusca, I: Gastropoda•Proso- Rosenberg, C, Kuncio. G.S., Davis, CM., and Harasewych, branchia. Philos Trans R Soc Land B 307:457-496, 1-V. M.G. (1994). Preliminary ribosomal RNA phylogeny of Haszprunar, G. (1988). On the origin and evolution of major gastropod and unionoidean bivalve mollusks. In: Molecu- gastropod groups, with special reference to the Strepto- lar Techniques and Molluscan Phylogeny. Harasewych, neuva. / Mollusc Stud 54:367-441. M.G., and Tillier, S. (eds.) NautUus SuppJ 2:111-121. Hickman, C.S. (1976). Pleurotomaria (Archaeogastropoda) in Rosenberg, C, Tillier, S. Tillier, A., Kuncio, G.S., Hanlon, the Eocene of the northeastern Pacific: a review of cenozoic R.T., Masselot, M., and Williams, C.J. (1997). Ribosomal biogeography and the ecology of the genus. / Paleontol RNA phylogeny of selected major clades in the MoUusca. 50(6):1090-1102. / Mollusc Stud (in press). Hickntan, C.S. (1984). Pleurotomaria: pedigreed persever- Salvini-Piawen, L. von (1980). A reconsideration of systemat- ajice? In: Living Fossils. Eldridge, N., and Stanley. S.M. ics in the MoUusca (phylogeny and higher classification). (eds.). New York: Springer Verlag, 225-231. Malacologia 19:249-278. Hickman, C.S. (1988). Archaeogastropod evolution, phylog- Salvini-Plawen, L. von (1985). Early evolution and lire primi- eny and systematics: a réévaluation. Malacolog HevSuppl tive groups. In: The MoUusca, vol 10, Evolution. Trueman, 4:17-34. E.R. and Clarke, M.R. (eds.). Orlando, Fla: Academic Press, Holland, P.W., Hacker, A.M., and Williams, N.A. (1991). A 59-151. molecular analysis of the phylogenetic affinities of Sacco- Salvini-Plawen, L. von, and Haszprunar, G. (1987). The Veti- glossus cambrensisBramhell &Cole (Hemicordata). Philos gasti'opoda and the S)'stematics of streptoneurous Gastro- Trans R Soc Load Biol 332:185-189. poda (MoUusca). ¡Zool Land 211:747-770. 20 A'/.G. Harasenych et al. Shileyko, A.A. (1977). The symmetry of the Docoglossa and niques and Molluscan Phylogeny. Harasewych, M.G., and the problem of the origin of the order. Bull Moscow Nat Tillier, S. (eds.). Nautilus SuppI 2:122-140. Socß/o/Se;-81 (3):60-65 (English summary). See Boss and Tillier, S., Masselot, M., and Tillier, A. (1996). Phylogenetic lacobson, 1977:494, for EngUsh summary. relationships of tlie pulmonate gastropods from rRNA se- Smith, S., Overbeelv, VV.R., Woese, C.R., Gilbert, W., and Gib quences, and tempo and age of the stylommatophoran level, P.ÎVI. (1994). The genetic data environment: an ex- radiation. In: Origin and Evolutionary Radiation of the pandable GUI for multiple sequence analysis. Comp AppI Mollusca. Taylor, J.D. (ed.). Oxford, UK: Oxford University Biosci 10:671-675. Press, 267-284. Swofford, D.L, (1996). PAUP, Phylogenetic Analysis Using Tracey, S.,Todd,J.A,,and Erwin, D.H. (1993). Mollusca: Gas- Parsimony, version 4.0.0d42 (beta test version). Washing- tropoda. In: The Fossil Record 2. Benton, M.J. (ed.). Lon- ton, DC. Laboratory of Molecular Systeixiatics, Smithson- don: Chapman and Hall, 131-167. ian Institution. Vaught, K.C. (1989). A Classification of tlie Living Mollusca. Szabó, ). (1980). Lower and Middle Jurassic Gastropods from Melbourne, Fla. American Malacologists, Inc. the Bakony Mountains (Hungary), Part II: Pleurotomaria- Winnepenninckx, B., Backeljau, T., Van de Peer, Y., and De cea and Fissurellacea (Archeogastropoda). Annal Histor- Wächter, R. (1992). Structure of the small ribosomal sub- ico-Naturales Musei Nat Hungarici 72:49-71. unit RNA of the pulmonate snail Limnicolaria kambeul, Taylor, D.W., and Sohl, N.F. (1962). An outline of Gastropod and phylogenetic analysis of the Metazoa. FEES Lett classification. Malacologia 1(1):7~32. 309:123-126. Termier, H., and Termier. G. (1968). Evolution et Biocinèse. Winnepenninckx, B., Backeljau, T, and De Wächter, R. (1993). Les ¡nvértebvés dans l'histoire du Monde Vivant. Paris: Complete small ribosomal subunit RNA sequence of the Masson & Cie. chiton Acanthopleurajapónica (Lischke, 1873) (Mollusca, Thiele, ¡. (1929). Handbuch der systematischen Weichtier- Polyplacophora). Nucleic Acids Res 21:1670. kundo. Jena: Gustav Fischer, l(l):l-376. Winnepenninckx, B.,T. Backeljau, and R. De Wächter. (1994). Tillier, S., Masselot, M., Hevré, P, andTillier, A. (1992). Phylo- Small ribosomal subunit RNA and the phylogeny of Mol- génie moléculaire des Gastropoda (MoUusca) fondée sur lusca. In: Harasewych, M.G., and Tillier, S. (eds.) Molecu- le séquençage partiel de l'ARN ribosomique 28 S. Comptes lar Techniques and Molluscan Phylogeny. Nautilus suppl Rendus Acad Sel (Paris) Series 3, 134:79-85. 2:98-110. Tillier, S., Masselot, M., Guerdoux, J., and Tillier, A. (1994). Yonge, CM. (1947). The Palliai Organs in the Aspidobranch Monophyly of major gastropod taxa tested ft'om partial Gastropoda and their evolution throughout the Mollusca. 28S rRNA sequences, with emphasis on Euthyneura and Philos Trans R Soc Lond B 232(591):443-518, pi. 18. Hot-Vent Limpets Peltospiroidea. In: Molecular Tech-