<<

Invertebrate Biology 12 I(4): 27 1-324. 0 2002 American Microscopical Society, Inc.

On bivalve phylogeny: a high-level analysis of the () based on combined morphology and DNA sequence data

Gonzalo Giribet'%aand Ward Wheeler2

' Department of Organismic and Evolutionary Biology, Museum of Comparative Zoology, Harvard University; 16 Divinity Avenue, Cambridge, Massachusetts 021 38, USA Division of Invertebrate Zoology, American Museum of Natural History, Central Park West at 79th Street, New York, New York 10024, USA

Abstract. Bivalve classification has suffered in the past from the crossed-purpose discussions among paleontologists and neontologists, and many have based their proposals on single char- acter systems. More recently, molecular biologists have investigated bivalve relationships by using only gene sequence data, ignoring paleontological and neontological data. In the present study we have compiled morphological and anatomical data with mostly new molecular evi- dence to provide a more stable and robust phylogenetic estimate for bivalve molluscs. The data here compiled consist of a morphological data set of 183 characters, and a molecular data set from 3 loci: 2 nuclear ribosomal genes (1 8s rRNA and 28s rRNA), and 1 mitochondria1 coding gene (cytochrome c oxidase subunit I), totaling -3 Kb of sequence data for 76 rnollu bivalves and 14 outgroup taxa). The data have been analyzed separately and in combination by using the direct optimization method of Wheeler (1 996), and they have been evaluated under 1 2 analytical schemes. The combined analysis supports the monophyly of bivalves, paraphyly of protobranchiate bivalves, and monophyly of Autolamellibranchiata, , Hetero- conchia, , and s.I., which includes the monophyletic taxon An- omalodesmata. These analyses strongly support the conclusion that should not receive a class status, and that the heterodont orders Myoida and Veneroida are not mono- phyletic. Among the most stable results of the analysis are the monophyly of Palaeoheterodonta, grouping the extant trigoniids with the freshwater unionids, and the sister-group relationship of the heterodont families and , which together constitute the sister taxon to the remaining heterodont bivalves. Internal relationships of the main bivalve groups are dis- cussed on the basis of node support and clade stability.

Additional key words: Moll~sca,Bivalvia, Palaeoheterodonta, Heteroconchia, Heterodonta, 18s rRNA, 28s rRNA, cytochrome c oxidase I, morphology, direct optimization, scnsitivity analysis

Bivalve molluscs are characterized by a laterally ciliated labial palps connect the ctenidia to the mouth, compressed body with an external bivalved shell that and direct food particles into it. The extensible foot is is hinged dorsally. The valves are connected by a par- either elongated or laterally compressed. tially calcified elastic and are held together These modifications from the plesiomorphic mol- by 1 or 2 adductor muscles. There is no buccal or luscan condition have made it difficult to establish a radular apparatus, and the lobes are either phylogenetic scheme of the group. The problems arise joined or free ventrally. The spacious mantle cavity from the difficulty in homologizing certain structures extends upwards on each side of the visceral mass and useful for bivalve that are not present in the contains a pair of ctenidia suspended laterally. The cte- other molluscan classes. Paleontologists and neontol- nidia may be enlarged, lamellate and plicate. The ogists have disputed the monophyly and phylogenetic mouth and anus are located at opposite ends of the position of many groups, such as Anomalodesmata, body and the gut is typically convoluted. A pair of Protobrdnchia, and Palaeoheterodonta, while others us- ing molecular sequence data have openly questioned a Author for correspondence. the monophyly of Heterodonta, as well as the orders E-mail: ggiribet @ oeb.harvard.edu Veneroida and Myoida. 272 Giribet & Whceler

There have been several concerted attempts to re- laeotaxodonta (= Nuculoidea and Nuculanoidea), solve the contradictory systems of classification of bi- Cryptodonta (= Solemyoida), Pteriomorphia, Palaeo- valves proposed by paleontologists and neontologists. heterodonta, Heterodonta, and Anomalodesmata. Pur- C.M. Yonge and T.E. Thompson organized the sym- chon (1978) compiled a data matrix of 9 characters for posium Evolutionary Systematics of Bivalve Molluscs, 40 taxa (superfamilies) of bivalves that was analyzed which was published in 1978 (Phil. Trans. R. Soc. using a phenetic computer algorithm. Following an ex- Lond. B 284: 199-436). Two decades later, paleon- panded analysis (Purchon 1987b), bivalves werc di- to I ogi sts, neontol og ists, and molecular biologists pro- vided into 2 subclasses, and Lamelli- vided additional insights into bivalve phylogeny at the branchia, the latter containing 4 orders: Pteriomorphia, International Symposium on the Paleobiology and Mesosyntheta (= Trigonioida and Unionoida [includ- Evolution of the Bivalvia (Johnston & Haggart 1998) ing also , , and Leptonoi- and at a meeting on The Biology and Evolution of the deal), Anomalodesmata, and Gastropempta (= Bivalvia (Harper et al. 2000b).These efforts have not Heterodonta) (Fig. 1A). yet produced a single combination of morphological For the purposes of the present study, we lollow thc and molecular data, and conflicting hypotheses of bi- classification system of Beesley et al. (1998). Thus, a evolution remain. It is our aim to investigate classification system of 5 subclasses is followed prior previously proposed hypotheses by analyzing morpho- to the phylogenetic analyses. This classification is gen- logical and molecular characters of all the extant erally corroborated by the morphological analysis oc bivalve orders in a total-evidence framework. Waller (1998). Taxon names currently in use in the literature are noted where needed. This classification Previous classification systems of bivalves includes the following subclasses; representatives used in this study are listed in Table 2. Comparative anatomical studies of living bivalves Subclass Protobranchia. The classification of the have led to several classification schemes. Cox (1960) Protobranchia is unstable (Reid 1998). Nuculoidea and provided an excellent historical review of early at- Nuculanoidea are considered superfamilies of the or- tempts to classify the bivalves. Ridewood (1 903) rec- der Nuculoida by several authors, although certain ognized 3 orders of bivalves (Protobranchia, Eleuth- phylogenetic studies have suggested non-monophyly erorhabda, and Synaptorhabda) based on gill structure. of Nuculoida (Waller 1990, 1998; Morton 1996). Pelseneer (1906, 19 1 1) developed another system of Therefore, we have used the superfamilies Solemy- classification based on 5 grades of gill structure and oidea (2 of Solemya), Nuculoidea (4 species of assigned ordinal status to each grade: Protobranchia, ), and Nuculanoidea (2 species of Nuculan- Filibranchia, Pseudolamellibranchia, Eulamellibran- idae, 2 species of , and 1 of Neilonellidae). chia, and Septibranchia. Iredale (1939) added the Subclass Pteriomorphia (= Filibranchia). Five Isofilibranchia to distinguish the mytiloids, which he orders are recognized in this classification: Mytiloida, considered to differ sufficiently in gill structure from Arcoida, Pterioida, Limoida, and Ostreoida. Pojeta the other members of Filibranchia. Atkins (1 938) de- 978) separated mytiloids as a distinct subclass (Iso- scribed two types of latero-frontal ciliation on gill fil- (I aments and proposed division of the class into 2 filibranchia), but Waller (1 998) recognized the cate- groups, the Macrociliobranchia and Microciliobran- gory Pteriomorphia, regarding Mytiloida as the sister chia. Later workers proposed that other structures be group to the other pteriomorphs. Representatives of the used in classifying bivalves (stomach: Purchon 1960, 5 orders have been examined in this study. 1963, 1968; ctenidial-labial palp associations: Stasek Subclass Palaeoheterodonta. This group is com- 1963). Scarlato & Starobogatov (1975, 1978, 1979) posed of 2 orders, Trigonioida and Unionoida. How- and Starobogatov (1992) recognized 3 superorders of ever, some authors do not support the monophyly of bivalves, the Nuculiformii (= Protobranchia), Mytili- the group (e.g., Salvini-Plawen & Steiner 1996). Our forniii, and Conocardiiformii (= Anomalodesmata). In study includes representatives of both orders (2 this new classification, Mytiliformii contained unionids and 2 trigoniids). pteriomorphs, palaeoheterodonts, and heterodonts. Subclass Heterodonta. Classification of Hetero- Classifications based on single-character systems donta has not been resolved, even at the ordinal level have been criticized (Cox 1960; Newell 1965, 1969). (Prezant 1998). Heterodonta (as accepted by Vokes Newell (1965, 1969) summarized the available evi- 1968; Cox 1969; Newell 1969; Beesley et al. 1998) dence on shell structure and anatomy and presented a consists of 3 orders: the extinct Hippuritoida and the classification of Bivalvia that is now generally in use. extant Veneroida and Myoida. The large order Vene- In his scheme, 6 subclasses were recognized: the Pa- roida comprises - I8 superfamilies, of which IS are On bivalve phylogeny 27 3

Nuculoidca Nucincllidae Nuculanoidea Solcrnyidae relationship between Bivalvia and Scaphopoda, a view Solernyoida Nuculoidea shared by Stasek (I 972). The Diasoma concept (Run- Ptcriomorphia Nuculanoidea negar & Pojeta 1974; Pojeta & Runnegar 1976, 1985; Trigoninidea Trigoiiiidac Runnegar 1978; Pojeta et al. 1987) united Bivalvia, Unionoidca I’~ciinmorphia Cuassatclloidea Uniorioida Scaphopoda, and the extinct group Rostrocoachia (= Carclitoidca VciicmidaLucinoid;i Loboconcha of Salvini-Plawen 1980, 1985). Other au- Myoida Anomelodesmata thors also suggested monophyly of Scaphopoda and Anoinalodcsmata Myoida Bivalvia based on the foot structure (Hennig 1979; Vencroida Lauterbach 1984); Hennig (I 979) introduced the name A. I’urchon (I 987b) B. Salvini-l’lawen & Steiner (1996) Ancrypoda (“anchor foot”) for this clade. Some cla- distic analyses also supported this relationship (G6t- ting 1980; Lauterbach 1983; Salvini-Plawen & Steiner Solemyoidca (- Lipodoiita) Nuculoidca Solemyoidea 1996), but Waller (I 998) excluded Scaphopoda from Nuculoidca Niiculanoidea the Diasoma, considering scaphopods as sister group Nuculanoidca IMvtiloida TIIgoiiioida to Cephalopoda (Grobben 1886), while Haszprunar Uniorioida Piiinoidca (2000) placed Bivalvia as the sister group to Anomalodesrnata Ptcrioidea (Scaphopoda (Gastropoda + Cephalopoda)). Arcoida (= Ncotaxodonta) OstreoldCd Anoniioidea Steiner (1992) listed the putative synapomorphies of Ptcriomorphia the DiasomdLoboconcha: (1) Development of lateral IIctcIodonta Palaeohctcrodonta Anomalodesmata mantle folds that converge ventrally and enclose the c. Cope ( 1997) Heterodonta entire body, probably an adaptation to infaunal life; D Waller (1990, 1998) (2) a foot differentiated into a burrowing organ; and (3) an epiathroid nervous system with true pedal gan- Fig. 1. High-level phylogenetic relationships proposed for glia, all of which correspond in position and area of the Bivalvia. Leptonacea from Purchon (1 987b) comprises Galeominatoidca and Cyamoidca. innervation (Haszprunar 1988). Numerical and parsimony-based analyses of mol- luscan groups are routinely used to address phyloge- included in this study; the 4 superfamilies in the netic questions. However, only 2 phenetic analyses smaller order Myoida are included in this study. (Purchon 1978, 198713) and 2 parsimony analyses Subclass Anomalodesmata. Beesley et al. (1998) (Salvini-Plawen & Steiner 1996; Carter et al. 2000) followed the classification outlined by Morton (19824 assessing higher relationships among bivalves based in recognking a single order (Pholadomyoida) with 7 on morphological data have been published to date. superfamilies. This classification mainly based on is These studies agreed upon 2 main clades of bivalves, the paleontological work of Runnegar (1 974). How- Protobranchia (= Palaeotaxodonta) and Autolamelli- ever, Newel1 (1965) had divided the subclass into 2 branchiata’ (= Lamellibranchia), but d orders, Pholadomyoida and Poromyoida (= Septibran- conclusions (Fig. 1). Using a parsimony analysis of 42 chia or Septibranchida), a classification also followed characters for 14 terminal taxa, Salvini-Plawen & by Coan et al. (2000). We have included data on 2 Steiner (I 996) concluded that bivalves are monophy- species of the superfamily Pandoroidea (1 of Lyonsi- letic and divided into Protobranchia and Autolamelli- idae and I of ) and 2 species of Cuspida- branchiata. However, Palaeoheterodonta and Hetero- roidea. The superfamilies Thracioidea, Verticordioi- donta were not monophyletic (Fig. IS). dea, and are not included due to lack of Cope (1997) added numerous fossil taxa to analyses tissues for DNA samples. of extant bivalves. This produced a cladistic (but non- Phylogenetic relationships numerical) classification based primarily on shell mi- crostructural data. This unorthodox phylogenetic Although bivalves are well known morphologically, scheme had Palaeoheterodonta + Anomalodesmata as they have been a source of discord both in terms of the sister group to Pteriomorphia + Heterodonta (Fig. their relationships to other molluscan classes, and re- 1C). Morton (1 996) proposed a superfamilial phylo- lationships within the class. In the following paragraphs we discuss the most contradictory points. ‘ The term Autobranchia has been used in previous publi- Morphological studies cations to refer to Autolamellibranchiata GROBREN,1894. We use the correct Latin name, but we prefer to use the Based on anatomical and embryological data, term autobranch rather than autolamellibranchiate in the Lacaze-Duthiers (1856, 1857a, 1858) proposed a close colloquial sense. 274 Giribet & Wheeler genetic system of bivalves including numerous fossil and complete 18s rRNA sequences of the taxa Pro- taxa. In his phylogeny, Palaeoheterodonta was nested tobranchia (their Palaeotaxodonta), Pteriomorphia within Pteriomorphia. Heterodonta s.1. was monophy- (their Pteriomorphia and Tsofilibranchia), and Hetero- letic, with Anomalodesmata as a sister group to Myoi- donta (including Veneroida and Myoida). Palaeohet- da. In contrast, Veneroida was not monophyletic. Wal- erodonta and Anomalodesmata were not sampled, al- ler (I 990, 1998) proposed a cladistic classification though they placed Myoida within Anomalodesmata with monophyletic Protobranchia, Autolamellibran- in their table 1. The trees resulted in bivalve polyphyly chiata, Pteriomorphia, Palaeoheterodonta, Eulamelli- (Solemyu grouped with Gastropoda), pteriomorph par- branchia, Anomalodesmata, and Heterodonta (Fig. aphyly, and monophyly of Autolamellibranchiata and ID). Heterodonta. Hoeh et al. (1998) studied phylogenetic relation- Molecular studies ships among 14 species of bivalves based on sequenc- es of the mitochondria] cytochrome c oxidase subunit The first high-level phylogenetic analysis of bi- I (COI). These analyses, based on 613 bp, supported valves based on 18s rRNA sequence data included 13 monophyly of Autolamellibranchiata, Mytiloida, Ve- sequences (1 polyplacophoran, 2 gastropods, 8 pter- neroida, Unionoida, and Palaeoheterodonta, although iomorphs, and 2 veneroid heterodonts) and concluded monophyly of bivalves was not supported; the proto- that the 18s rRNA molecule did not recover bivalve branch fell within the non-bivalve molluscs. The a~i- monophyly in most of the analyses (Steiner & Muller thors concluded that the “molluscan bivalved body 1996). Giribet & Carranza (I 999) subsequently dem- plan may have evolved more frequently than tradition- onstrated that the polyphyly of Steiner & Muller al phylogenetic hypotheses suggest.” Another result (1996) was an artifact of taxon sampling. By adding suggested the monophyly of Mytiloida + Veneroida, 20 sequences to the data set used by Steiner & Miiller excluding Trigonioida. However, these results were (1 996), their analysis resulted in bivalve monophyly based on limited sequence data of only 5 of the 12 and recovered the subclasses Pteriomorphia and Het- recognized bivalve orders. Their study was the first to erodonta. However, both studies were based on poor include molecular data for the Trigonioida, and sup- representation of bivalve diversity, and stressed meth- ported the morphology-based hypothe odology more than the actual phylogeny of the group. rine group is the sister taxon of freshwater unionids, In another study of relationships among the molluscan comprising Palaeoheterodonta. classes, Winnepenninckx et al. (1996) used complete Subsequently, Canapa et al. (1999) studied relation- 18s rRNA sequences of 1 aplacophoran (Caudofov- ships among some autobranch bivalves based on 18s eata), 2 polyplacophorans, 7 gastropods, 1 scaphopod, rRNA sequence data (2 polyplacophorans, 2 gastro- and 13 bivalves (7 pteriomorphs and 6 veneroid het- pods, and 21 bivalves: 10 pteriomorphs, and 11 het- erodonts), with protostome worms and other groups as erodonts). When using polyplacophorans and gastro- outgroup taxa. None of the analyses supported bivalve pods as outgroups, bivalves were not monophyletic. monophyl y. But when only the 2 gastropods were used as out- A common problem in molecular phylogenetic anal- groups, Bivalvia, Pteriomorphia, and Heterodonta yses is limited taxon sampling, as in the cases of bi- were monophyletic. valves above. Adamkewicz et al. (1 997) alleviated pre- Other molecular studies have focused on lower level vious taxon sampling deficiencies by analyzing 500 bp relationships. Distel (2000) and Distel et a]. (2000) from all bivalve subclasses, including all orders except used 18s rRNA sequence data to elucidate relation- Limoida and Trigonioida (totaling 28 bivalves and 5 ships within , and its position within pterio- outgroup taxa). When all outgroup taxa were included, morph bivalves. Using 1 solemyid, 1 unionid, and 1 bivalves were polyphyletic. By removing gastropods myid as outgroups, pteriomorph monophyly was not and rooting the trees with polyplacophorans, bivalve obtained in any of the maximum-likelihood, parsimo- monophyly was recovered (Adamkewicz et al. 1997, ny, or minimum-evolution trees, due to the clustering figs. 2, 3), with the 2 anomalodesmatans and 2 pro- of with the . A recent study of sphaeriid tobranchs in a sister clade to the remaining bivalves. and corbiculid relationships analyzed -1 Kb of 28s The other protobranch (Nucula) was sister to a clade rRNA sequence data for 18 veneroid bivalves (Park & that contained 3 pteriomorphs. Pteriomorphs were 0 Foighil 2000). They obtained veneroid monophyly polyphyletic (distributed in 3 clades). Palaeohetero- with respect to 2 ostreoids, (although no myoid or an- donta was the sister clade to the monophyletic omalodesmatan species were sampled). The most im- Heterodonta. portant result of this study was to show that Corbi- Campbell et al. (1998) used a combination of partial culidae and Sphaeridae were not sister groups, since On bivalve phylogeny 275

Table 1. Outgroup taxa used in the analyses, with CenBank accession numbers.

18s rDNA 28s rDNA COI Class Polyplacophora Lepidopleurus cujetanus AF120.502 AF12056.5 AF 120626 Acanthochitona crinita AF120503 AF120566 AF 120627 Class Cephalopoda Nuutilus pompilius AF20764 1 AF41168X AFI 20628 Loligo pealei AF 120.505 AF120568 AF 1 20629 Sepia elegans AFI 20506-7 AF120569 Class Gastropoda Ha1ioti.s tuberculata AF12051l AF120570 Sinezona confusa AF1205 12 AF120571 AFl2063 1 Diodora grueca AFI 20.513 AF 120572 AF120632 Viviparus georgianus AFl20.5 16 AFl20574 AF 1 20634 Truncatellu guerinii AF120.517 AF 12057.5 AF 1 20635 Ralcis ehurneu AFl205 19 AFI 20576 AFI 20636 Peltodoris atromaculata AF120521 AFI 20577 AF 1 20637 Class Scaphopoda Antalis pilsbryi AF 120.5 2 2 AF 120579 AF I 20639 Rhahdus rectius AF120.523 AF120580 AF I20640 the formed a monophyletic group with ly divergent rates of molecular evolution may be re- veneroids and mactroids (with high nodal support). sponsible for the failure of molecular phylogenetic Thus, the monophyly of Corbiculoidea was not studies in recovering bivalve monophyly. No analysis supported. has combined data from more than one molecular Steiner & Hammer (2000) addressed pteriomorph source or has combined molecular data with morphol- and higher bivalvian relationships by using a wide rep- ogy in a character-based analytical framework. In the resentation of complete 18s rRNA sequences from bi- present study, we combine data from morphology and valves and other molluscs as outgroups. This elegant anatomy (1 83 characters), and molecules (complete study included nearly complete taxon sampling within 18s rRNA, D3 region of the 28s rRNA and -660 bp the pteriomorphs, along with most bivalve orders. Al- of the COT loci) of 62 bivalves (representing all extant though Bivalvia appeared diphyletic due to the hetero- orders and most superfamilies) with 14 outgroup taxa geneity of substitution rates among lineages, mono- (representing all extant classes of Testaria except Mon- phyly of Protobranchia, Heteroconchia, and oplacophora). We hope to resolve inconsistencies of Pteriomorphia was supported. However, Myoida and previous studies by providing the first total-evidence Veneroida were not monophyletic, and Anomalodes- investigation of all orders of the class Rivalvia. mata was nested within the heterodonts. Resolution within the Pteriomorphia showed conflict with mor- Methods phological hypotheses in the position of Taxonomic sampling and Arcoidea (Steiner & Hammer 2000, fig. 8). Another massive 18s rRNA sequence data analysis Molecular and morphological data of 5 classes of was published by Campbell (2000), again with most testarian molluscs were analyzed (Tables I, 2): Poly- bivalve orders and superfamilies represented. The re- placophora (2 spp.), Cephalopoda (3 spp.), Gastropoda sults were similar to those of Steiner & Hammer (7 spp.), Scaphopoda (2 spp.), Bivalvia (62 spp.). (2000), in supporting polyphyly of bivalves, mono- Within the bivalves, the 5 subclasses recogniLed by phyly of Pteriomorphia and Heteroconchia, and poly- Beesley et al. (I 998) were represented with the follow- phyly of Myoida and Veneroida. The Anomalodesmata ing number of terminal taxa: Protobranchia (1 1 ), Pter- nested within the Heterodonta. iomorphia (1 7), Palaeoheterodonta (4), Anomalodes- In summary, the taxonomic sampling of molecular mata (4), Heterodonta (27). All 12 orders of bivalves bivalve studies has improved, although some major were represented: 34 superfamilies (representing 74% lineages such as are not yet represented of bivalve superfamilial diversity according to Beesley in 18s rRNA data sets. It has been suggested that wild- et al. 1998) and 43 families (-45% of the familial 276 Giri bet & Wheeler

Table 2. Systematic list of the bivalve taxa used in the analyses (following Beesley et al. 1998). The list includes 62 species (but 63 exemplars, as Nucula sulcata is represented by 2 populations, one Mediterranean and one from the North Atlantic). The asterisk after the ordinal name indicates those orders that do not include all superfamilies. When no super- familial category is indicated, the order is represented by a single superfamily. Underlined taxon names are categories not supported by the analyses. When the symbols “18s” or “COI” are indicated, these sequences are from GenBank. All other sequences (GenBank accession codes indicated) have been obtained by the authors.

18s rDNA 28s rDNA COI Class Bivalvia Subclass Protobranchia Order Solemyoida Family Solemyidae Solernya velum AFl20524 AFl2O58 1 COI Solemyu reidi 18s Order Nuculoida Superfamily Nuculoidea Family Nuculidae Nuculu sulcuta MED AF 120525 AF120582 Nuculu .sulcuta ATL AF207642 AF207649 AF207654 Nuculu proxinzu AFl20526 AFl20583 AF 1 2064 1 Acilu custrensis AFl20527 AF 120584 Superfamily Nuculanoidea Family Yoldiidae Koldiu limatulu AF120528 AF120585 AF I20642 Koldia myalis AF207643 AF207650 AF207655 Family Nuculanidae Nuculana minutu AF120529 AF 120586 AFI 20643 Nuculunu pernulu AF207644 AF20765 I Family Neilonellidae Neilonellu subovata AF207645 AF207652 AF207656 Subclass Pteriomorphia Order Mytiloida Family Mytilidae Geukensia demissu 18s COI edulis 18s AFI 20587 COI Lithophuga lithophagu AF120530 AF 120588 AFI 20644 Order Arcoida Superfamily Arcoidea Family Arcidae Area none 18s Burhatiu harhatu AF207646 AFl20589 AF 120645 Family Striurcu lucteu AF 12053 1 AF 120590 AF 1 20646 Superfamily Limopsoidea Family Glycymeris insubrica AF207647 AF 1 2059 1 Order Pteroida Superfamily Family Ptcriidae Pteria hirundo AF 120532 AFl20592 AFI 20647 Superfamily Pinnoidea Family Atrina pectinata 18s AF120593 AF120648 On bivalve phylogeny 277

Table 2. Continued.

18s rDNA 28s rDNA COI Order Lirnoida Family Lima lirna AF120533 AFI 20.594 AFI 20649 Limuria hiuns AF120534 AFl20595 AF 120650 Order Ostreoida" Suborder Ostreina Superfamily Family Ostreidae Ctussostrea virginica 18s Ustrea edulis 18s AFI 20596 AFI 2065 1 Suborder Pectinina Superfamily Pectinoidea Family Pectinidae munimus 18s COI Chlamys varia 18s AFl20597 Family Spondylidae sinensis AF229629 Superfamily Family Anomiu ephippium AFl20535 AFl20598 Subclass Palaeoheterodonta Order Unionoida" Superfamily Unionoidea Family Psilunio littoralis AF 120536 AF120.599 AFl20652 Lampsilis cardium AF120537 AF 120600 AF120653 Order Trigonioida Family Neotrigoniu hednalli AF120538 Neotrigonia margaritacea AF4 1 1690 AF411689 COI Subclass Anornalodesrnata Order Pholadomyoida" Superfamily Pandoroidea Family Pandoridae Pandora a?*enosa AF 120539 AFI 2060 1 Family Lyonsia hyalina AF120540 AFl20602 AFl20654 Superfamily Cuspidarioidea Family cuspidatu AF I20541 -2 AF I20603 AF 120655 Myonera sp. AFl20544 AF 120605 Subclass Heterodonta Order Veneroida" Superfamily Carditoidea Family Carditidae calyculata AF120549 AF120610 AF 12O66O antiquata AF120550 AF120611 AF 12066 1 Superfamily Crassatelloidea Family Astartidae castanea AF 120% 1 AF 120612 AF I20662 278 Giribet & Wheeler

Table 2. Continued.

18s rDNA 28.5 rDNA COI Superfamily Chamoidea Family gryphoides AF 120545 AFI 20606 AF 120656 SuperSamily Lucinoidea Family Lucinidae Codukia cfr. orbiculata AF 1 20.546 AF 120607 AF I20657 Superfamily Galeommatoidea Family Galeomrnatidae Galeommu furtimi AF 120547 AF120608 AF I20658 Family Lasaeu sp. AF120548 AF 1 20609 AF 12065Y Superfamily Solenoidea Family ensis AF120555 AFI 2061 6 Superfamily Family Seinelidae Ahru di: prismaticu AFI 20554 Superfamily Family Cardiidae Parvicardium exiguum AF120553 AF 120614 AF I 20664 Frogum unedo 1as Superfamily Tridacnoidea Family Tridacnidae gigas 1as hippipus I as Superfamily Family po1.ymorphu AF12OS52 AF120613 AF I20663 Superfamily Mactroidea Faniily M actridae .suhtrcincata 1 8s AFI 206 I5 AF207657 nuttalli 18s Superfamily Arcticoidea Family Arc tic i dae islandicu 18s Family Calyptogena nzagnifica AFI 20556 AFl206 17 AF 120665 Superfamily Corbiculoidea Family Corbiculidae Corbiculu jhwninea AF 120557 AF120618 AF 120666 Family Sphaeriuin striatinum AFI 20558 AFI 206 19 AF 120667 Superfamily Veneroidea Family mercenaria AFI 20559 AF I20620 AF 120668 18s Order Myoida Superfamily Myoidea Family Mya nrenaria AFI 20560 AF 12062 I COI Family Varicorbula disparilis AF 12056 1 AF I 20622 AF I20669 On bivalve phylogeny 279

Table 2. Continued.

18s rDNA 28s rDNA COI Superfanily Gastrochaeonoidea Family Gustrochaena duhia AFl20562 AFl20623 AF 120670 Superfainily Hiatclloidea Family Hiulella arctica A F 1 20563 AFI 20624 Superfamily Pholadoidea Family Tercdinidae Rankia cavinatu AF I20564 AFl20625 AF120671 diversity). The analysis was limited to extant taxa. For references are provided in the character description details on collection data, see http://www.mcz. list. The inorphological character states were scored har vard. edu/ DepartrnentslJn vertZoo/giribetdata.h tm. for each terminal taxon whenever possible, and i F cod- ings were based on other terminal taxa, thi\ has been Data collection specified in the character description (Appendix I ). Morphological data. Morphological data were ob- Morphological data (Appendix 2) were entered in the tained from the literature and from direct observation program NDE v. 0.4.6 (Page 2000), and character op- of specimens as cited in character descriptions, result- timization over trees was conducted with MacClade v. ing in I83 characters, 182 of which were treated as 4.01 (Maddison & Maddison 2000). unordered (non-additive). Besides the many descrip- Molecular data. Complete 18s rRNA sequences of tions of characters referred to in the section describing 74 terminal taxa were analyzed (- 1,760-2,500 bases). the characters, 3 monographic family-level studies Of these, 60 terminal taxa were sequenced by the au- have been crucial (Boss 1982; Beesley et al. 1998; thors. The data set was complemented with 61 new Coan et al. 2000). sequences of the D3 region of the 28s rRNA loci The following morphological and anatomical stud- (-300-600 bases), and by 53 sequences of the mito- ies were consulted as primary literature sources: So- chondrial cytochrome c oxidase subunit 1 (COI) (660- 672 bases). lenzya velum (Drew 1900; Morse 1913; Gustafson & Lutz 1992); Solemya reidi (Gustafson & Reid 1988a); Details of DNA isolation, PCR amplification, and limatula (Drew 1899a; Kellog I9 15); Nuculana DNA sequencing are given in Edgecombe et al. (2002) minuta (Atkins 1936); Lithophaga lithophagu (B.R. and Giribet et al. (2002). Primers used for amplifica- Wilson 1979); Anomiu ephippium (Lacaze-Duthiers tion and sequencing can be found in Folmer et al. 1857b; Yonge 1977); Spondylus sinensis (S. american- (1994) and Giribet et al. (1996, 1999, 2002). Chromatograms obtained from the automated se- us in Yonge 1973); Neotrigonia margaritacea (Morton 1987~);Cardita culyculatu and Curdites antiquatu quencer were read and acsembled using the sequence (Yonge 1969); Astarte c'astanea (Saleuddin 1965, editing software Sequenchera 3 .O. Complete se- 1967); Chamu gryphoides (Yonge 1967); Galeomma quences were edited in Genetic Data Environment turtoni (M.L. Popham 1940; Bieler & Mikkelsen (GDE) software (Smith et al. 1994). The external 1992); sp. (M.L. Popham 1940); Dreissena po- primers IF and 9R (for the 18s rRNA loci), 28Sa lymorpha (Morton 1969; Pathy & Mackie 1993); Fra- and 28Sb (for the 28s fragment), and LCO1490 and HC02198 (for the COI fragment) werc excluded gum unedo (F. erugatum in Morton 2000a) ; Tridacna gigas and (Yonge 1980); Spisula from the analyses. All the new sequences have been subtruncata (Yonge 1948, 1982b); Calyptogena mag- deposited in GenBank (accession codes are given in Tables 1 and 2). n$cu (Boss & Turner 1980); Corhicula.fluminea (Brit- ton & Morton 1982); Myu arennria (Yonge 1982b); Phylogenetic analyses Varicorhula disparilis (Mikkelsen & Bieler 2001 ; also V. gibhu in Yonge 1946); Castrochuenu dubia (Carter Homology concept in sequence data. While most 1978); arctica (Yonge 197 1); carinata molecular analyses use strict base-to-base correspon- (Turner 1 966); Pandora arenosu (P. iizaequivalvis and dences (fixed alignments) for primary homology, this P. pinna in Allen 1954); Cuspidaria cuspiduta (see may introduce ambiguity and does not accommodate Yonge & Morton 1980; Morton 1987b). Additional sequences of substantially unequal length. Instead, our 280 Giribet & Wheeler first hypothesis of homology corresponds to secondary Sensitivity analysis. Character transformations were structural features (Giribet & Wheeler 2001) followed weighted differentially to observe how they affect phy- by a dynamic base-to-base correspondence (direct op- logenetic conclusions (sensitivity analysis sey1,~uWheel- timization method: Wheeler 1996). The ribosomal se- er 1995). Two analytical variables were examined: quences have been divided into unambiguously rec- insertion-deletion cost ratio, and transversion-transition ognizable homologous regions (see Giribet 2001). The cost ratio. When the transversion-transition ratio was set split was initiated by using internal primer regions, and at a value other than unity, the insertion-deletion cost then by identifying secondary structural features. Cor- was set according to the cost of transversions. In total, respondences among these regions are viewed as pri- 12 combinations of parameters were used in the anal- mary hypotheses of homology, analogous to the use ysis (insertion-deletion ratios of 1, 2, and 4; of morphological features to decide primary homolo- transversion-transition ratios of 1, 2, 4, and m). This gy. The protein-coding CO1 sequences were not divid- strategy allows discerning between stable relationships ed because we lacked internal primers or structural (those supported throughout the chosen range of param- predictions. eter values) and unstable relationships (those that Tn total, the 18s rRNA molecule was divided into appear only under particular parameter sets). 30 fragments, and the 28s rRNA region into 5 frag- Molecular data analysis. The 3 molecular data sets ments. Nomenclature of the secondary structural re- were analyzed independently and combined directly, gions of the 18s rRNA loci follows that of Hendriks with all characters weighted equally without regard to et al. (1 988). Particularly variable regions of the 18s source. These data sets are referred as 18s (1 8s rRNA rRNA loci are the following: E10-2 (fragment data set alone), 28s (28s rRNA datasetalone), COT biv10-2), E21-1 (biv21-1), 41 (biv41), and 47 (COI data set a=), and molecular (1 8S, 28S, and (biv47-2). These regions present high heterogeneity in COI). The COI data set, although from a protein- sequence length, with large insertions in the cephalo- coding gene, was analyzed at the DNA level without pods Nuutilus and Loligo and in the Anomalodesmata, specifying reading-frame constraints (because i ndels and therefore have been excluded from the analyses. were inferred). Moreover, preserving the reading frame The input files containing the unaligned sequences of may not yield the shortest (most parsimonious) clad- all terminal taxa, parameter files, and batch files are ograms, and because we are attempting to co-optimize available from the website http://www.mcz.harvard. 3 different sources of evidence (morphology, non-cod- edu/Departments/lnvertZoo/giribetdata.htm. ing genes, and protein-coding genes) using a sensitiv- Sequence data analysis: direct optimization. Se- ity analysis framework, this seems the most quence data were analyzed using the direct optimiza appropriate way to analyze the data. tion method (Wheeler 1996) and implemented in the Tree search commands executed in POY included computer program POY (Gladstein & Wheeler 1997; random addition sequence followed by a fast paral- Wheeler & Gladstein 2000). The method directly as- lelized tree-building step and by subtree pruning and sesses the number of DNA sequence transformations regrafting (SPR) and tree bisection and reconnection (evolutionary events) required by a phylogenetic to- (TBR) branch swapping. When classical swapping pology without the use of multiple sequence align- algorithms did not improve tree-length, the data ment. This is accomplished through a generalization were subjected to several rounds of tree-drifting and of existing character optimization procedures to in- tree-fusing (Goloboff 1999) to decrease tree length. clude insertion and deletion events (indels) in addition The entire search strategy was repeated up to 100 to base substitutions. The crux of the model is the times or until the results converged on the same re- treatment of indels as processes, as opposed to the pat- sult at least 3 times in independent replicates, as in terns implied by multiple sequence alignment. The re- previous analyses (Giribet et al. 200 1 ; Edgecombe sults of this procedure are directly compatible with et a!. 2002). parsimony-based tree lengths, and the method appears Morphological data analysis. A parsimony anal- to generate more efficient (thus simpler) explanations ysis of the morphological data set was performed with of sequence variation than multiple sequence align- the computer program NONA v. 1.9 (Goloboff 1998). ment (Wheeler 1996). Direct optimization, although The tree-search strategy adopted initially involved a computationally intense, is much less demanding than heuristic algorithm with random addition-sequence parsimony-based multiple sequence alignment algo- and TBR branch swapping. All characters were treated rithms. The method has also been demonstrated to as unordered (non-additive), except for character 1 15, yield more congruent results than multiple sequence and no specific weighting schemes were applied. Since alignments when using congruence among data sets as the traditional search combining random addition and a criterion (Wheeler & Hayashi 1998). TBR found the shortest tree length I out of 1000 On bivalve phylogeny 28 1 times, additional analyses were performed with the computers become available for exploring hypotheses beta version of TNT (Goloboff et al. 2000) using the in phylogenetic analysis. driven search (Goloboff & Farris 2001) option. Branch support (Bremer 1988, 1994) up to 4 extra steps was Results calculated using a heuristic procedure and holding a Morphological analyses maximum of 10,000 trees with NONA (command: -bs 4). The tree-search strategy adopted in NONA (h/lO;.__ Combined analysis. Morphological and molecular mult* 1000;max*) yielded trees of minimal length in I data (total)- were combined directly and analyzed using out of 1,000 replications, retaining 1,344 trees of 5 I4 the direct optimization method (Wheeler 1996) for the steps (CI = 0.44; RI = 0.83). Since these results were same 12 parameters that were applied to each of the unsatisfactory as to describe the total diversity of trees molecular data sets and following the same search (a single TBR island was found), we decided to apply strategy. The morphological transformations were more aggressive search algorithms implemented in the weighted as equal to the highest of the molecular costs program TNT incorporating sectorial searches, tree (= indel,), to diminish the putative overwhelming ef- fusing, and tree drifting (Goloboff 1999). A driven fect of molecular data vs. morphology. Bremer support search was conducted and repeated 5 times, finding a values were estimated using the heuristics procedure consensus of 32 nodes that stabilized already in the implemented in POY (-bremer -constrain “filename” first round, hitting minimum tree length 35 times in -topology “treetopology”). about 8 minutes (4 hitdmin) in an 866 MHz Pentium In total, we analyzed 6 data sets and 12 parameter 111 (256 Mb of RAM). sets per data set (72 analyses). POY analyses were run The strict consensus of those trees (Fig. 2) shows in a cluster of 292 Pentium 111 processors at 1,000 MHz monophyly of Conchifera, Castropoda + Cephalopo- connected in parallel using pvm software and the par- da, Scaphopoda + Bivalvia, and of the 5 molluscan allel version of POY (commands -parallel -jobspernode classes represented. Bivalvia and Autolamellibranchia- -2 in effect). The morphological analyses were run in an ta are both monophyletic with good Bremer support 866 MHz Pentium 111 processor. (bs > 4). However, the protobranchiate bivalves (su- Character congruence. Congruence between data perfamilies Nuculoidea, Solemyoidea, and Nuculanoi- sets (morphological and molecular) was measured by dea) are depicted as either monophyletic or paraphy- Incongruence Length Difference (ILD) metrics (Mick- letic, collapsing therefore in the strict consensus tree. evich & Farris 1981; Farris et al. 1995) (see Table 3). None of the fundamental trees supports the current or- This value is calculated by dividing the difference be- dinal category Nuculoida (= Nuculoidea + Nuculan- tween the overall tree length and the sum of its data oidea) (e.g., Beesley et al. 1998). Autolamellibran- components : chiata splits into 3 lineages: Unionidae, Trigoniidae, and a clade containing Pteriomorphia and Heterodonta, (LengthComh,nd - Sum Len~thI~~d~v~du~~Scd ILD = the latter including veneroids, myoids, and anomalo- LengthLamhmcd desmatan bivalves. Unionids are resolved either as sis- ter group to trigoniids (making Palaeoheterodonta [Eq. 11 monophyletic), or as sister to pteriomorphs + hetero- Character congruence was used ar a criterion to choose donts, thus making Palaeoheterodonta a paraphyletic our optimal tree-the tree that minimized character assemblage. Resolution within the third clade, pter- conflict among all the data. This is understood as an iomorphs + heterodonts, is poor, and only a few su- extension of parsimony (or any other minimizing cri- prafamilial relationships are obtained in all the shortest teria). In the same sense that parsimony tries to min- trees (Fig. 2). Heterodonta is paraphyletic, with Pter- imize the number of overall steps in a tree, the char- iomorphia as sister group to Galeomma turtoni. Struc- acter congruence analysis attempts to find the model ture within Pteriomorphia shows Arcoida as the sister that maximizes overall congruence for all the data group to the remaining pteriomorphs. Pteroida, Myti- sources. Obviously, trying to generalize an evolution- loida, Arcoida, and Limoida are monophyletic, but not ary model (viewed as an inferential model we use to the order Ostreoida, suborder Pectinina, or superfamily make sense of observations) for all taxa and all regions Pectinoidea. In addition, the superfamily Arcoidea is may be too simplistic, especially when evaluating di- paraphyletic since Glycymeris (Limopsoidea, Glycy- vergences ranging from the to the . meridae) is placed in between the families Noetiidae However, evaluating for two general parameters (gap/ and Arcidae. change ratio and transversion/transition ratio) is a start A few nodes that are supported in the morphological point in evaluating many other parameters when faster data set are Carditidae, Pandoroidea, Cuspidariidae, 282 Giribet & Wheeler

Lcpidopleurus cajctanus 1Acanthochitona crinita + , , Abru + Mactridae, Nautilus ponipilius Loligo pcalei Ensis + Arctica + Veneridae, and ( (Galeom- Sepia elcgans mu + Pteriomorphia)). Some of these nodes are, how- Haliotis tuberciilata Sinczona confusa ever, supported by a single extra step (Fig. 2), and Diodora gracca Viviparus georgianus monophyly of groups such as Arcticoidea, Cardioidea, Truncatclla gucrinii Cardiidae, Corbiculoidea, Myoida, Veneroida, and An- Balcia cburncd omalodesmata is not supported by the morphological >4 Aiitalis pilsbryi Khabdus rectius data set alone. >4 Solernya vclum Solcmva rcidi Congruence analysis The parameter set that minimizes incongruence among data sets is the one at gap/change ratio of 1 and transversion/transition ratio of 1, the topmost line of values in Table 3 (gaps are weighted equal to all Lanipsilk cardium 1 Neotrigonia bednalli base transformations). This parameter set-hereafter -Neotrigonia margaritace AstwLe castanea referred to as the optimal parameter set-reaches a Chama gryphoidcs maximum congruence of 0.04028. However, the 2 next Lasaca sp. Drcissciia polymorpha suboptimal parameter sets have very similar values, Parvicartlium cxiguum ILD = 0.04049 and 0.04089 (Table 3). Choice of ei- >4 Calyptogeiia inagnifica fluminea ther one of these 3 parameter sets may be conditioned striatum Mya arcndrid by the aggressiveness of the heuristic search strategy ". ,~iicotbula: dispiAs tiasti ochacna duhi:~ performed. Since the overall topology of the trees ob- Bankia carinata tained under these parameter sets is highly similar, we present the results based on the best parameter set for this particular search.

Partitioned molecular analyses 18s rRNA The analyses performed for the optimal parameter set yielded 2 trees of minimal tree length (3,979 steps), Callista chioiic and found minimum tree length 3 times. Neither tree Mcrccnaria iiicrccnaria Codakia orbiculata shows bivalve monophyly, and the strict coiiseiisus Galeomina turtoni Irlctea (Fig. 3) displays a polytomy of 6 clades: Nuculidae, Glycymeris insuhrica Solemyidae f Nuculanoidea, Arcoida, a clade con- Area node Barbatia barbata taining the remaining pteriomorphs, ((Carditidae + Ptcria hirundo Astartidae) Palaeoheterodonta), and a clade containing Geukcnsia dcinissa Mytilus edulis Gastropoda, Scaphopoda, Cephalopoda, and the re- Lithophaga lithophaga maining bivalves. Gastropoda, Protobranchia, Pterio- Anomia ephippiuni Pectcn maximus morphia, and Heterodonta are each polyphyletic. How- Chlamys varia Spondylus sincnsis ever, many resolved nodes correspond to conventional Lima lima taxa: Solemya, Nuculanoidea, Nuculoidea, Arcoida, Limaria hians Astarte + Carditidae, Palaeoheterodonta, Unionidae, Crassostrea virginica Neotrignnia, Mytilidae, Limidae, Pectinoidea, Pectin- Fig. 2. Strict consensus of 600 trees of 514 steps (CI = idae, Ostreidae, Galeommatoidea, Cardioidea, Tridac- 0.44; RI = 0.83) from a parsimony analysis of the morpho- ninae, Veneridae, and Mactridae. Almost all conven- logical data. Numbers on branches indicate Bremer support tional families represented by more than one species values as calculated in Nona (up to 4 extra steps, retaining are monophyletic, except for Arcidae (but Arcoida is 10,000 trees). Bivalves are represented by bold branches. monoph yletic). The strict consensus of all parameters explored sup- ports the following monophyletic groups: Polyplaco- phora, Conchifera, Coleoidea, Vetigastropoda, Denta- liidae, Nucula sulcata, Nuculanoidea, Nuculuna, On bivalve phylogeny 283

Table 3. Tree lengths, at 12 sets of parameter values, for the 4 individual and 2 combined data sets, and ILD’s for the combined analyses of all data. Parameters: gapkhange ratio (gap); transversiodtransition ratio (tv/ts). Individual data sets: 18s rDNA (1 8s);28s rDNA (28s); cytochrome c oxidase I (COI); morphology (mor). Combined data sets: molecular (mol = 18S, 28S, and COI); total (IXS, 28S, COI, and mor). Values for the parameter set that minimizes incongruence, i.c., has the lowest ILD (0.04028), appear in the topmost line; we refer to this as the “optimal parameter sct” (see text).

Individual Combined

gap tv/ts 18s 28.5 COI mor mol total ILD 1 1 3979 337 7060 514 11788 12389 0.04028 1 2 6094 472 1 I078 1028 18287 19468 0.04089 I 4 10157 71 1 18787 2056 30730 33049 0.04049 1 0 1971 I13 3765 514 6070 6665 0.0453 1 2 1 4784 378 7264 1028 1291 1 I4094 0.0454 1 2 2 7630 549 11417 2056 20353 22729 0.04738 2 4 13109 85 1 19403 41 12 34702 39497 0.05 1 I9 2 0 2655 142 3938 1028 7035 8236 0.05743 4 1 6120 439 7392 2208 14595 I703 I 0.05 I20 4 2 10183 65 I 11688 4416 23587 28499 0.05477 4 4 18101 1036 19948 8832 4 1006 SO9 17 0.05892 4 0 3747 I88 4079 2208 8469 10959 0.06725

Mytilidae, Barbatia + Glycymeris, Limidae, Pectinoi- remaining bivalves are resolved as a monophyletic dea, Pectinidae, Ostreidae, Palaeoheterodonta, Union- clade, with Palaeoheterodonta as sister group to the idae, Neotrigonia, Astute + Carditidae, Galeomma- remaining bivalves, but few suprafamilial relationships toidea, and Mactridae. Despite numerous unresolved are congruent with current classifications. Nuculanoi- nodes in the strict consensus, all the remaining dea and Astarte + Carditidae are among the results groupings have morphological support. that are congruent with most other data sets and pa- rameter sets. The strict consensus of all the analyses 28s rRNA performed with the COT data set alone yields a largely unresolved tree with a few supported nodes: Loligo The analyses performed for the optimal parameter + Sepia, Nuculanoidea, Yoldia, Arcoidea (Limopsoidea set yielded SO trees (until the buffer filled) of minimal is not represented in the COI data set), Limidae, tree length (337 steps), and found minimum tree length Unionidae, Dreissenu Myoidea, Guleommu Ban- 4 times. The strict consensus of the 50 trees obtained + + kia, Mytilus Geukensia, Carditidae, and Astarte for the 28s rDNA data is largely unresolved, perhaps + + Carditidae. because of the small size of the data set (-300 bp CO1 alignments within bivalves and among mol- used). This tree (not shown) supports polyphyly of luscs are not trivial, because the gene shows consid- Bivalvia, Pol yplacophora, and Gastropoda. The mono- erable length variation. (A sequence alignment is said phyletic groups found in all the fundamental trees are to be trivial when it is not parameter dependent, that Cephalopoda, Coleoidea, Vetigastropoda, Viviparus + is, generally does not have insertion/deletion events). Balcis + Truncutella, Nuculo sulcata, Atrina + Lyon- The typical length for all non-bivalve taxa studied is sia, Galeommatoidea, Carditidae, and a clade contain- 669 bp. This is also true for protobranchs, palaeohet- ing (Anomia (Mytilus (Limaria (Pandora + My- erodonts, anomalodesmatans, and few other pterio- onera)))).The strict consensus of all the trees found morphs and heterodonts. But the remaining bivalves under all parameter sets resolves only 2 clades, have sequences varying in length between 660 and 675 Coleoidea and Atrina + Lyonsia. bP. COT Combined molecular data (18S, 28S, COI) The CO1 tree for the optimal parameter set yielded I tree of 7,060 steps (Fig. 4). This tree does not sup- The analysis of all the molecular data combined, for port monophyly of Polyplacophora, Gastropoda, or the optimal parameter set, yielded 2 trees (L= 11,788). Cephalopoda. It also does not support bivalve mono- The strict consensus of the 2 trees (Fig. 5) shows bi- phyly because Nuculoidea and Solemyoidea nest with- valves as polyphyletic, because a clade of Nuculoidea in a clade containing gastropods and cephalopods. The + Solemyoidea is sister to a clade containing Scapho- 284 Giribet & Wheeler

Lepidopleurus cajetanus poda, Cephalopoda, and Peltodoris. The non-nuculoid, ____c Acanthochitona crinita Nucula proxima non-solemyoid bivalves form a clade, with the Palaeo- Acila castrensis heterodonta as sister to the remaining bivalves, includ- Nucula sulcata MED Nucula sulcata ATL ing Nuculanoidea, Pteriomorphia, Heterodonta, and Solemya velum Anomalodesmata. Additional clades obtained are Nu- Solemya reidi Nuculana minuta culanoidea, Mytilidae, (Pinnidae ( + Ostrei- Nuculana pernula Neilonella subovata dae)), Ostreidae, Arcoida, Nuculanoidea, Spondylidae Yoldia limatula + Limidae + Pectinidae, Heterodonta, Astartidae + Yoldia myalis Arca noae Carditidae, Lucinidae + Anornalodesmata, Anonialo- Striarca lactea desmata, Chamidae + Cardioidea, Cardioidea, Tridac- *Barbatia barbata Y Glycymeris insubrica ninae, Corbicula + Mactridae, Mactridae, Arcticoidea Mytilus edulis Veneridae, and Veneridae (see Fig. 5 for other Geukensia demissa + Lithophaga lithophaga clades) . ~ Anomia ephippium The strict consensus of all the parameter sets for all Lima lima c Limaria hians the molecular data analyzed in combination yielded L Spondylus sinensis Pecten maximus the following monophyletic groups: Polyplacophora, Chlamys varia Conchifera, Scaphopoda, Coleoidea, Vetigastropoda, Atrina pectinata Pteria hirundo Nucula sulcuta, Solemyu, Nuculanoidea, Mytilidae, Crassostrea virginica Limidae, Ostreidae, Pectinoidea, Pectinidae, Unioni- Ostrea edulis Astarte castanea dae, Neotrigonia, Palaeoheterodonta, Carditidae, AJ- Cardita calyculata tarte + Carditidae, Mactridae, Cardioidea (sensu Cardites antiquata Psilunio littoralis Schneider 1992), and Tridacninae. Lampsilis cardium Neotrigonia bednalli Neotrigonia margaritacea Combined analysis (morphological and Diodora graeca Haliotis tuberculata molecular) Sinezona confusa *Balcis eburnea Overall, the most congruent combined analysis of Viviparus georgianus all the data is derived from a gaplchange ratio = 1 and Truncatella guerinii Antalis pilsbryi a transversion/transition ratio = 1 (ILD = 0.04028; see Rhabdus rectius Table 3). This parameter set yielded a single tree of 1- 1- Peltodoris atromaculata Nautilus pompilius 12,389 steps (Fig. 6),and found it a single time. When Loligo pealei the combined tree is given as a constraint to the mor- /I Seoia eleuans L Pandora arenosa phological matrix, it requires 57 additional steps, a Lyonsia hyalina Cuspidaria cuspidata 10% increase in tree length. This indicates some con- Myonera sp. flict between the morphological and the molecular data Hiatella arctica - Codakia orbiculata sets. Neither the molecular nor the morphological trees Galeomma turtoni drive the final phylogenetic hypothesis; rather each in- - Lasaea sp. i dubia fluences different areas of the hypothesis. The clado- prismatica gram shows monophyly of Polyplacophora, Conchi- Parvicardium exiguum fera (bs = 43), Scaphopoda (bs = 59), Cephalopoda Fragum unedo (bs = log), and Bivalvia (bs = 20), bul not Gastro- - Tridacna gigas Hippopus hippopus poda, because Peltodoris is a sister taxon to Cepha- Sphaerium striatinum lopoda. The strict consensus tree obtained for the 12 - Mya arenaria parameter sets is also shown (Fig. 6). Figs. 7-10 show Varicorbula disparilis Dreissena polymorpha a schematic representation of the hypotheses obtained Bankia carinata for the 12 parameter sets explored, and Fig. 11 sum- marizes the hypothesis obtained for the combined analysis. Arctica islandica Mercenaria mercenaria relationships Outgroup Fig. 3. Strict consensus of 2 trees at 3,979 steps for the 18s Scaphopoda was found to be the sister group to Pel- rRNA data set yielded by the optinial parameter set. Mini- todoris + Cephalopoda (Fig. 6), not to Bivalvia, as mum tree length was found in 3 out of 100 replicates. proposed by the morphological data alone (Fig. 2). The Bivalves are represented by bold branches. clade Scaphopoda + Peltodoris + Cephalopoda ap- On bivalve phylogeny 285

Lepidopleurus CaJetanUs pears as the sister group to the remainder of Gastro- I 7Peltodoris atromaculata poda, and the 3 classes together constitute the sister Antalis pilsbryi group to Bivalvia. Lc Rhabdus rectius I IBalcis eburnea Nucula sulcata ATL Ingroup relationships

Several bivalve groups are stable to parameter change and are monophyletic in all the combined analyses (Fig. 6): Solemya, Nuculidae (N~tculaand Acila), Nuculanoidea (Yoldiu, Neilonella, and Nucu- lana), Nuculana, Yoldia, Pteriomorphia, Mytilidae (Lithophaga, Geukensia, and Mytilus), Arcoida Acanthochitona crinita (Area, Barbatid, Striarca, and GEycymeris), Ostreidae Diodora graeca (Ostrea and Crussostrea), Pectinoidea (Spondylus, Viviparus georg'rianus Pecten, and Chlamys), Pectinidae (Chlunzys and Pec- - Neotrigonia margaritacea ten), Limidae (Lima and Limariu), Palaeoheterodonta Psilunio littoralis (Unionidae and Neotrigonia), Unionidac (Psilunio Lampsilis cardium and Lumpsilis), Neotrigonia, Carditidae + Astartidae Lasaea sp. (Cardita, Cardites, and Astarte), Carditidae (Cardita - Cuspidaria cuspidata and Cardites), Cardioidea (Parviocardium, Fragum, Barbatia barbata Tridacnu, and Hippopus), Tridacninae (Tridacna and Striarca lactea Hippopus), Mactridae (Spisula and Tresus), Veneri- - Atrina pectinata dae (Callista and Mercenaria), and Dreissena + Ostrea edulis Myoidea (Myu and Varicorbula). Monophyly of the Varicorbula disparilis groups Bivalvia, Protobranchia, Autolamellibranchia- Dreissena polymorpha ta, Pteriomorphia, Heteroconchia, Anomalodesmata, Mya arenaria - and Heterodonta is not supported by all the parame- ters. However, Bivalvia, Autolamellibranchiata, and Anomalodesmata are monophyletic under most pa- rameter sets studied (Figs. 7, 10). The 2 parameters that disrupt their monophyly are the ones showing the highest gap costs, under which the highly autapo- morphic Cephalopoda is placed within the Anomal- odesmata. Heteroconchia is monophyletic under 7 parameter sets, whereas Heterodonta s.1. is monophy- letic under 6 parameter sets. In the tree yielded by the optimal parameter set (Fig. 6), Bivalvia is monophyletic, with the initial split di- viding the non-siphonate protobranchs (Solemyoidea + Nuculoidea) from the rest of bivalves (Nuculanoidea + Autolamellibranchiata). Protobranchia is paraphy- letic for all parameter sets explored so far. Autolamellibranchiata (Fig. 7: node 4) is monophy- letic under most parameter sets, including the optimal parameter set and all the nearest suboptimal ones. Au- tolamellibranchiata is divided into Pteriomorphia and Heteroconchia (= Palaeoheterodonta + Heterodonta s.1.). Pteriomorphia is monophyletic under all parameter sets (Fig. 8: node 1) with the mytiloids as the sister group to the remaining pteriomorphs (Fig. 8: node 2). Monophyly of Arcoida is supported under all parameter Fig. 4. Single tree at 7,060 steps for the COI data set yielded by the optimal parameter set. Minimum tree length was sets (Fig. 8: node 3), although monophyly of the su- found in 1 out of 100 replicates. Bivalves are represented perfamily Arcoidea is not supported, as the limopsoid by bold branches. 286 Ciiribet & Wheeler

Lepidopleurus cajetanus Glycymeris is generally placed between the families Ar- fiAcanthochitona crinita Diodora gracca cidae and Noetiidae. Monophyly of the remaining pter- Haliotis tubcrculilta Sinezona confusa iomorphs is stable (Fig. 8: node 4), although the internal Balcis cburnea relationships within the non-mytiloid non-arcoid pter- iomorphs are parameter-dependent (Fig. 8). For exam- ple, the order Pteroida, composed by the superfamilies

I Pterioidea (represented by Pteriu) and Pinnoidea (rep-

I resented by Atrina) is not monophyletic under the op- - timal parameter set, but is monophyletic under 3 other parameter sets. The position of the family Ostreidae is also unstable; it is the sister group to Pteriidae under Nucula sulcata MED the optimal parameter set, but to Pinnidae or to Pecti- Nucula sulcata ATL noidea under other parameter sets (Fig. 8). The order Psilunio littoralis Lampsilis cardium Ostreoida is not monophyletic under any parameter set. Ncotrigonia bednalli The suborder Pectinina (a member of the order Ostreo- Ncotrigonia margaritacca Anomia ephip ium ida) here represented by one member of' Anomioidea Lithophaga littophaaga (Anomiu) and 2 families of Pectinoidea (the pectinids Geukensia demissa Mytilus edulis Pecteri and Chlumys, and the spondy lid Spondylus) is Atrina pcctinata never monophyletic. Pteria hirundo Crassostrca virginica Heteroconchia, a group composed by the monophy- Ostrea edulis letic Palaeoheterodonta and Heterodonta s.1. (heter- Arcd noae Striarca lactca odonts including Anomalodesmata), is monophyletic Barbatia barbata Glycymeris msubrica under 7 parameter sets, including the optimal and 2 Neilonella subovata immediate suboptimal ones (Fig. 7: node 5). Palaeo- Yoldja limatula Yoldia myalis heterodonts are monophyletic for all parameter sets Nuculana minuta explored, with Unionidae as sister to Trigoniidae. Nuculana pernula Spondylus sinensis Some relationships among hcterodont groups for Lima lima the parameter sets explored are shown in Figs. 9 and Limaria hians Pecten maximus 10. Heterodont monophyly is obtained under 6 pa- Chlamys varia rameter sets (Fig. 9: node l), including the optimal Astarte castanea Cardita cdiyculatd and some immediate suboptimal ones. The first split Cardites antiquata Coddkia orbiculata within Heterodonta s.1. is between a clade Crassatel- - Lyonsia hyaiina loidea + Carditoidea-composed of the 2 carditids Pandora arcnosa Cuspidaria cuspidata (Curdita and Cardites) + the astartid (Asturte)-and Myoncra sp. the remaining heterodonts, which include the anom- - Hiatella arctica - Lasaea sp. alodesmatans. Monophyly of Carditidae + Astartidae Gastrochacna dubia is stable to parameter choice (Fig 9: node 2), as is the Dreissena olymorpha VaricorbuE dispdnlis monophyly of the non-carditid non-asturtid heter- i Mya arcnaria odonts (Fig 9: node 3). Relationships within the mod- 3haerium striatinum aleomma turtoni ern (non-crassatelloid, non-carditoid) heterodonts are Bankia carinata unstable to parameter choice, and only a few rela- Ensis ensis Abra prismatica tionships are stable. Among the stable groups are Car- Chama gryphoidcs Parvicardium exiguum dioidea, Tridacninae, Mactridae, Veneridae, and the Fra um unedo clade composed by the myoids (Mya and Vuricor- Trijacna gigas Hip opus hippopus bulu) + Dreissenu, oE which all are monophyletic for Corticula fluminea all parameters (Figs. 6, 10). A sister-group relatiun- Spisula subtruncata - Tresus nuttallii ship between Chamoidea and Cardioidea is also sug- Arctica ishdica gested by the data and found under most parameter Calyptogena inagnifica Mercenaria mercenaria sets (Fig. 9: node 4), as is the monophyly of Mactro- Callista chionc idea + Dreissenoidea + Myoidea + Arcticoidea + Fig. 5. Strict consensus of 2 trees at 11,788 steps for the Corbiculoidea + Veneroidea, and the monophyly of combined molecular sequence data (18S, 28S, and COI) its subgroups (Mactroidea (Dreissena + Myoidea)) yielded by the optimal parameter set. Minimum tree length and Arcticoidea + Corbuloidea + Veneroidea (Fig. was found in 2 out of 100 replicates. Bivalves are 9: node 5). However, monophyly of Arcticoidea (rep- represented by bold branches. On bivalve phylogeny 287

Lcpidoplcurus cajctanus Acanthochilona crinita - Sinczona confiisa Balcis churnca

Solcmya vchn 41 Solcmya reidi Nucula proximn i Ada . castrensis Nucula sulcata MED Nucula sulcata ATL I Nuculana niinuta Nuculana pcrnula Ncilonclla subovata Yoldia liinatula f Yoldia inyalis RIVALVIA I

c AlJTOI AMFl I I- RRANCHIAIA Lampsilis cardium Ncotrigonia hednal I i Ncotrigonia iriargaritacca Astartc castanca Cardita calyculata Clirditcs antiuuata Pandova arcnhsa Lyonaia hyalina Cuspidaria cuspidata Myoncra sp. ' f II Hiatclla arctica HE'I'EI3 ODONTA L.a,aca \I). Codakia oi-biculrtl~! Galcomma turtoni Fig. 6. Analyses 01' the combined Bankia carinata Gaatrochaena dubia morphological and molecular data. Abra prismatica Left: single tree at 12,389 steps 20 Cham gryphoidcs I'arvicardnrm cxiguim for- optimal parainctcr set. Min- 2'1 1h imum tree lcngth was found in I out 2 6 I of I00 replicates. Branches in bold Spisula suhtruncata Tresus nuttallii represent bivalves. Numbers on Mya arcnaria Dreisscna olymorpha nodes rcprcscnt Brcmer support Varicorbul disparilis Sphaeriuin striatinuni values. Lalyfopcna niagnifica Right: strict consensus of all trees Cov icu a iluniiiica ~ Arctica islandica obtained for the 12 piirameter sets Mcrccnxria mcrccnaria C>

node 1: Bivalvia node 2 cd Y,

node 3 node 4: Autolamellibranchiata 5

node 5: Heterocoiichia Palaeoheterodonta

I - ~ ,111 I 2 4 inf. I 2 4 in?. cal congruence plots for the 12 pa- rameter sets explored. = mono- phyletic; = non-monophyletic. TRANSVERSI0N:TRANSITIONRATIO falls within Heterodonta, its phylogenetic position tiple hypotheses of relationships, and by exploring within eulamellibranchs is not clear. Sister-group re- data using a sensitivity analysis approach, we can not lationships are suggested for Hiatellidae (2 parameter only generate hypotheses of relationships, but also sets, but not the optimal one). Under the optimal pa- evaluate their stability without the use of methods rameter set, Anomalodesmata is sister to all non- that perturb the data. These results should be consid- crassatelloid, non-carditoid eulamellibranchs. How- ered in the light of previous evidence supporting re- ever, when high parameter valueq are used (421 and lationships, some stable (which we consider well cor- 44 1 ), anomalodesmatans become paraphyletic with roborated), and some not. respect to cephalopods. This result may be explained by the high gap costs (of 8 and 16, respectively), Scaphopoda and the sister group of bivalves which might favor the clustering of species with large Molecular data (Winnepenninckx et al. 1996; Hoeh insertions. et al. 1998; Steiner & Hammer 2000; this study) do In summary, the results of the combined analysis not rupport Scaphopoda + Bivalvia. A sister-group for the optimal (most congruent) parameter set strong- relationship between these 2 classes has been pro- ly suggest paraphyly of protobranchiate bivalves, posed by several authors based on morphological ev- monophyly of Nuculanoidea Autolamellibranchiata, + idence (Gottmg 1980; Lauterbach 1983; Runnegar Autolamellibranchiata, Pteriomorphia, Palaeohetero- 1996; Salvini-Plawen & Steiner 1996). In contrast, donta, Heterodonta s.l., and Anomalodesmata, as well Waller (1998) and Haszprunar (2000) proposed a as paraphyly of Heterodonta S.S. and Veneroida, and closer relationship of Scaphopoda to Cephalopoda polyphyly of Myoida. Subclass-level status of and Gastropoda. While the morphological data com- Anomalodesmata is not supported. piled here do not support Waller's ( 1998) hypothesis, the molecular data place the Scaphopoda as sister to Discussion Cephalopoda + PeEtodoris (bs = 6), both in a clade The analyses presented here are the most extensive with the remaining gastropods. This adds support to study of bivalve phylogeny in terms of morphological the Cyrtosoma hypothesis semu Waller (1 998). A characters, molecular characters, molecular loci, and clade Scaphopoda + Cephalopoda is found in 3 of taxon sampling (families, superfamilies, and orders) the 12 parameter sets explored here, but not a single to date. By using character congruence among data parameter set supports Scaphopoda + Bivalvia. The sets as an optimality criterion to choose among mul- close relationship of Scaphopoda to Cephalopoda + On bivalve phylogeny 289

Gastropoda appears to be supported primarily by the odontophore, and associated buccal organs (character molecular data (bs = 10). We thus conclude that Sca- 85); presence of adductor muscles (character 106); phopoda are probably not sister to Bivalvia, although presence of a burrowing foot with anterior enlarge- more data are needed. This is clearly indicated by the ment (Character 109), also present in Scaphopoda; and unexpected behavior in sequence analyses of the gas- presence of an epiathroid nervous system with iden- tropod Peltodoris, which clusters between Scapho- tical innervation areas (character 123) (also present in poda and Cephalopoda under some parameter sets. Scaphopoda). The large amount of morphological ev- idence supporting bivalve monophyly, and the Bremer Bivalvia support value (bs = 20), as well as the stability of its monophyly to parameter choice when the morpholog- Bivalve monophyly has not generally been support- ical and molecular data are combined, strongly suggest ed by molecular studies (Steiner & Miiller 1996; Win- that bivalves are monophyletic. We obviously prefer nepenninckx et al. 1996; Adamkewicz et al. 1997; the conclusions based on our more extensive analyses Campbell et al. 1998; Hoeh et al. 1998; Steiner 1999), to those based on small subsets of taxa andlor even when taxon sampling is improved (Campbell individual fragmentary data sets. 2000; Steiner & Hammer 2000). Likewise, in the pre- sent study, the molecular data alone do not support Protobranchiate bivalves monophyly of Bivalvia, because a clade of protobran- chiate bivalves (Nuculoidea and Solemyoidea) appears Protobranchiate bivalves have been considered by nested within some outgroup taxa (Fig. 5). Similarly, many authors to be monophyletic and the most prim- the 18s rDNA analyses of Adamkewicz et al. (1997) itive group of bivalves because of the presence of a had Solemya, Yoldia, and 2 anomalodesmatans (Peri- plesiomorphic of ctenidia and their Cambrian or- ploma and Cuspiduria) clustering with outgroup taxa, igin. Classification of the Protobranchia has been a and Solemya was also related to the gastropods in the matter of contentious debate (Scarlato & Starobogatov 18s rDNA study of Campbell et al. (1998). 1979; Allen & Hannah 1986; Maxwell 1988). Beesley The lack of support for monophyly of bivalves with et al. (1998) adopted a classification system following molecular data has been interpreted to indicate poly- Maxwell (1988) and recognized 2 orders, Solemyoida phyletic origins of the bivalve body plan (Hoeh et al. and Nuculoida, the latter including the superfamilies 1998). As noted by these authors, their analyses had Nuculoidea and Nuculanoidea. In contrast, Waller limited taxon and character sampling (a maximum of (1998) considered Nuculoida paraphyletic ((Nuculo- 613 bp for 17 taxa: 1 polyplacophoran, 1 scaphopod, idea + Solemyoidea) Nuculanoidea). Our morpholog- 1 gastropod, and 14 bivalves). In our study, the tree ical tree does not support monophyly of Protobranchia obtained from the 54 COI sequences alone (using the but rather supports Solemyoidea, Nuculoidea, and same fragment as Hoeh et al. 1998) for the optimal Nuculanoidea as monophyletic groups. The combined parameter set supported polyphyly of Polyplacophora, data suggest a close relationship between Solem yoidea Scaphopoda, Cephalopoda, Gastropoda, and Bivalvia and Nuculoidea, as proposed by Waller ( 1998); this (Fig. 4), and did not support monophyly of Protobran- result is stable to parameter change (Fig. 7: node 2). chia, Autolamellibranchiata, Pteriomorphia, or Heter- Monophyly of Solemyoidea + Nuculoidea is support- odonta. These results suggest that COI sequence data ed by one unambiguous morphological synapomorphy : (at least from this CO1 fragment) lack sufficient phy- the presence of an adoral sense organ (character 141). logenelic signal to reconstruct higher molluscan rela- However, inclusion of more protobranch taxa could tionships, instead of indicating a polyphyletic origin of change the optimization of this character because ad- bivalves. oral sense organs have been observed in some nua- Our combined molecular and morphological data set lanoids not represented in the present analysis, such as supports the monophyly of bivalves under most pa- Nuculana fossa, N. pella, and Yoldia amygdulea rameter sets (Fig. 6 and Fig. 7: node l). Bivalves are (Schaefer 2000). supported by 10 unambiguous morphological syna- The clade Solemyoidea + Nuculoidea is not sister pomorphies (characters optimized using MacClade to Nuculanoidea, but rather to the clade Nuculanoidea 4.0): presence of pallial lines (character 24); body + Autolamellibranchiata, making Protobranchia a par- compressed laterally (character 45); absence of a dif- aphyletic group. This result is obtained under 10 of ferentiated head (character 46); presence of mantle the 12 parameter sets examined (Fig. 7: node 3), in- lobes (character 50) (also present in Scaphopoda); cluding the optimal parameter set. Monophyly of the presence of laterofrontal gill cilia (character 75); pres- protobranchiate bivalves is not found under any ence of labial palps (character 79); absence of radula, combination of parameters here explored. Giribet & Wheeler

node 1 : Pteriomorphia nodc 2 node 3: Arcoida

node 4 node 5 node 6 node 7

node 8 Pteroida Ostreoidea + Pectinoidca Pinnoidea 1 Ostrcoidca

I 2 4 inf, 1 2 4 inf, I 2 4 Illf TRANSVERSI0N:TRANSITIONRATIO Fig. 8. Summary tree for pteriomorph relationships and topological congruence plots for the 12 parameter sets explored.

4 = monophyletic; 1 = some of the Most Parsimonious Trees consistent with monophyly; [7 = non-monophyletic

Monophyly of Nuculanoidea + Autolamellibran- tigation, particularly of developmental characters re- chiata is supported by 3 unambiguous optimizations: lated to molluscan-cross formation, and of additional absence of prismatic structure in the shell (character data not sampled in this study. Addition of more ob- 7), which reverts in several lineages; absence of a hy- servations and more taxa could change how several pobranchial gland (character 62); and absence of a optimizations behave in the relationships presented molluscan cross during development (character 172). here. A molluscan cross is absent in all bivalves except so- lemyoids. A structure similar to a molluscan cross has Autolamellibranchiata been observed in Solemya reidi (Gustafson & Reid 1986) and S. velum (Gustafson & Lutz 1992), but no Members of Autolamellibranchiata are characterized data are available for Nuculoidea. Optimization of this as having modified gills that are not of the protobranch character might change when new data are added. The type. Autolamellibranchiata constitutes a clearly alternative hypothesis of protobranch monophyly is monophyletic group; this is stable to parameter choice supported by 2 unambiguous morphological synapo- (Fig. 7: node 4), and shows high Bremer support val- morphies: occurrence of extra- and intracellular diges- ues (bs = 27). The clade is supported by 14 unambig- tion in the midgut gland (character 87), and presence uous morphological optimizations: presence of proso- of a pericalymma larva during development (character gyrous umbones (character 26), shifting to orthogyrous 174). in a few lineages; having a mantle cavity occupied by Based on our analyses, we tentatively propose that gills lateral and posterior to the foot (character 63); bivalves with gills of the protobranch type are para- presence of reflected ctenidia (character 66); absence phyletic. This conclusion will require further inves- of palp appendages (character 81); absence of esoph- On bivalve phylogeny 29 1

nodc 1: Heterodonta s.1. node 2 nodc 3

Crassatelloidea Carditoidea Pandoroidea Cuspidarioidea Hiatelloidea Lasaeidae Lucinoidea Pholadoidea Gastrochaenoidea Tellinoidea Chamoidea node 4 6 Cardioidea node 5 nodc Solenoidea Mactroidea Dreissenoidea Mvoidea Sihaeriidae Vesicomyidae Corbiculidae w ,o Veneroidea

node 7 node 8 node 9 node 10

Arcticoidea + Vcncroidea Arcticoidca Corbiculoidea Mvoida

I 2 4 inl: I 2 4 Illf I 2 4 inf. 1 2 4 Ins TRANSVERSI0N:TRANSITTONRATIO Fig. 9. Summary tree for heterodont relationships and topological congruence plots for the 12 parameter set5 explored. = monophyletic; 17 = non-monophyletic. Only selected nodes are represented.

ageal ridges (character 86), secondarily originated in 1999). We propose that Autolamellibranchiata is the lucinids; presence of ciliated midgland ducts (character sister group to Nuculanoidea, and that it is composed 88); stomach with a crystalline style (character 89); of 2 main clades: Pteriomorphia and Heteroconchia. elongated major typhlosole (character 91); absence of a ventral surface of the foot (character 110); presence Pteriomorphia of a posterior pedal gland (character 114); presence of (character 115); presence of hemocyanin-like Pteriomorph relationships had received little atten- molecules (character 120); visceral ganglia larger than tion in modern phylogenetic studies, until the publi- pedal ganglia (126); and presence of nerve-type cation of the morphological analyses of Carter visceral connectives (character 127). (1990a), Waller (I 998), the molecular analysis of Autolamellibranch bivalves are also supported as Steiner & Hammer (2000), and studies interested in monophyletic in many of the partitioned analyses. the position of mytilids within bivalves (Distel 2000; Lack of support for monophyly of Autolamellibran- Distel et al. 2000). Our analyses clearly suggest that chiata in earlier studies may have resulted from limited pteriomorph bivalves are monophyletic, agreeing with taxon sampling, problems of alignment, or rooting the study of Steiner & Hammer (2000). This result is with uninformative outgroups (see Giribet & Carranza stable to parameter choice and has a Bremer support 292 Giribet & Wheeler

Mvoidca Drcisscna 4 Mvoidca Castrochaena I- Lasaca

Gastrochaena+ Lasaca+ (iiistrochaena + Cialcomma t Uankia Bankia + Galcomma Galeomma I- Bankia

Anoinalodcsmiila Hiatclla + Anomalodcsmata Hetcrodontd s.str.

Fig. 10. Topological congruence plots for the 12 parameter sets ex- I 2 4 inf plored showing relationships for se- 1 2 4 inf I 2 4 inf. w2lected myoid and anomalodesmatan heterodonts. = monophyletic; 0 TRANSVERSION: TRANSIT ION RATIO = non-monophyletic.

value of 22. Monophyly of Pteriomorphia is supported well corroborated by the data (Fig. 8: node 2; bs = only by a single unambiguous morphological optimi- 14), and is supported by 3 unambiguous changes: pres- zation: presence of egg cleavage with polar lobe for- ence of a byssal gape (character 20); presence of la- mation (character 171). However, the combined mo- terofrontal cilia of the microciliobranchiate type (char- lecular and morphological data strongly support this acter 76); and a type I11 ctenidial-palp association clade. (character 78). This result contrasts with the poGtion Mytiloidea constitutes a monophyletic group, sister of Mytiloidea in previous molecular analyses (Steiner to the remaining pteriomorphs. This result is stable to & Hammer 2000), but agrees with the most recent parameter changes, and is detected in 8 of the param- morphological analysis of pteriomorph relationships eter sets explored. The basal position of Mytiloidea (Waller 1998), and it is stable to parameter changes within Pteriomorphia has been proposed by other in- (Fig. 8: node 2). vestigators (i.e., Waller 1998; Steiner & Hammer 2000 Arcoida is sister to the remaining pteriomorphs [combined analysis I), although the molecular analyses (non-mytiloids, non-arcoids), which are monophy- published so far do not support the basal position of letic under I I parameter sets. This result contrasts Mytiloidea (Giribet & Carranza 1999; Steiner 1999; with the position suggested by the 18s rRNA trees Distel 2000; Steiner & Hammer 2000 [molecular anal- of Steiner & Hammer (2000), but agrees with their yses]). The basal position of Mytiloidea is supported combined tree of 18s rRNA and morphology, re- by our morphological data (Fig. 2) as well as by most solving a basal polytomy presented by Waller parameter sets for the combined morphological and (1998). This relationship is not supported by our molecular analyses. However, under certain parameter morphological tree (Fig. 2), which places the arcoids sets, mytiloids switch positions with arcoids, or appear as the sister group to the remaining pteriomorphs as the sister group of Heteroconchia, instead of to the (including mytiloids). The monophyly of the arcoi- remaining pteriomorphs. These results may be due to dan superfamily Arcoidea (Noetiidae and Arcidae) conflicting data in the molecular data sets, where my- is not supported by the data because the limopsoid tilids show large amounts of change (Distel2000). De- Glycymeris disrupts arcoidean monophyly under spite uncertainty in their sister-group relationships, our most parameter sets. Also, the molecular data fail to results corroborate the basal position of Mytiloida resolve the relationships among the 4 members of within Pteriomorphia. Arcoida. Further examination of the high-level ar- Monophyly of the non-mytiloid pteriomorphs is coidan relationships is needed. A similar lack of in- On bivalve phylogeny 293 ternal structure of the families was found for a more data analyses (Adamkewicz et al. 1997; Campbell extensive arcoidean analysis using 18s rRNA se- 2000; Steiner & Hammer 2000). However, data on Tri- quence data (Steiner & Hammer 2000). gonioidea were not available in the molecular studies. Monophyly of the remaining pteriomorphs is stable In contrast, Heteroconchia was not supported by the (Fig. 8: node 4; bs = 23) but the relationships among cladistic analysis of morphological data of Salvini- these groups (Pterioidea, Pinnoidea, Anomioidea, Li- Plawen & Steiner (1996), shell structure comparisons moidea, Ostreoidea, and Pectinoidea) are highly (Cope 1996, 1997, 2000), palaeontological compari- parameter-dependent and need further evaluation. For sons (Morton 1996), or molecular analysis of Cot data example, monophyly of the order Pterioida (Pterioidea (Hoeh et al. 1998). + Pinnoidea) is obtained under 3 parameter sets (not The data presented here support Heteroconchia as a including the optimal one), but monophyly of Ostreo- monophyletic group, which is diagnosed by a provin- ida (Ostreina + Pectinina) and Pectinina (Pectinoidea culum with differentiated dentition (character 38) and + Anomioidea) is not supported. Some other pteriom- dorsoventral muscles reduced to 2 pairs (or fewer) orph relationships are summarized in Fig. 8. The in- (character 105). Heteroconchia is divided into 2 stability of these results may be ascribed to poor taxon clades: Palaeoheterodonta and Heterodonta 5.1. sampling within the group, represented by only 10 of the 22 pteriomorph families. However, the support for Palaeoheterodonta monophyly of Pteriomorphia, Mytiloida, Arcoida, the remaining pteriomorphs, Limoidea, Ostreidae, and The name Palaeoheterodonta was given by Newell Pectinoidea constitutes a good basis for further studies (1965) to group the early Paleozoic actinodonts and of pteriomorph relationships. the extant unionoids and trigonioids. Palaeohetero- Previous molecular studies have failed to obtain donts were considered to be a monophyletic group pteriomorph monophyly (e.g., Steiner & Muller 1996; (Cope 1996) nested within the Pteriomorphia (Morton Adamkewicz et al. 1997; Campbell et al. 1998; Distel 1996), or sister to Anomalodesmata (Cope 1997) (Fig. 2000), possibly due to deficiencies in taxon or char- 1C). A recent suggestion excludes the trigonioids acter sampling, as demonstrated by more inclusive (Cope 2000). Previous morphological character anal- analyses (Campbell 2000; Steiner & Hammer 2000). yses did not support monophyly of Palaenheterodonvd In conclusion, our analyses indicate that pteriomorph (Purchon 1987b; Salvini-Plawen & Steiner 1996). Pur- bivalves constitute a monophyletic group, probably chon (1987b) proposed a relationship among Union- with the following internal structure: (Mytiloida (Ar- oidea and 3 superfamilies of heterodonts (Crassatel- coida ((Pinnoidea (Pterioidea + Ostreoidea)) (Pecti- loidea, Carditoidea, and Leptonoidea), with noidea (Anomioidea + Limoidea))))), a topology high- Trigonioidea as their sister group (Fig. 1A). Salvini- ly compatible with the combined tree of 18s rRNA Plawen & Steiner ( 1996) considered trigonioids to be and morphology presented by Steiner & Hammer the sister group to Pteriomorphia, while unionoids (2000). Ostreoida is polyphyletic, Pteroida and Pectin- were related to Heterodonta s.1. (Fig. 1B). The CO1 ina are weakly supported by the data, and the relation- sequence of Neotrigonia murgaritacea (Hoeh et al. ships among some of their superfamilies are unstable. 1998) was the first molecular data for a trigonioid and The position of the superfamilies Plicatuloidea and Di- suggested monophyly of Palaeoheterodonta, but did myioidea has not been addressed in the present not support Palaeoheterodonta as a sister group of het- analysis. erodonts. Monophyly of Palaeoheterodonta was cor- roborated by a broader taxonomic selection of COl Heteroconchia sequence data (Graf & 0 Foighil 2000) and is also supported by our molecular data (bs = 33) (with the Heteroconchia (sensu Cox 1960) includes all bi- exception of 28s rRNA) and for the total evidence valves with heterodont hinges. This well corroborated analyses for all the parameter set,. It is also seen in group (bs = 22) includes the Palaeoheterodonta and some of the shortest trees for the morphological data the Heterodonta s.1. (Heterodonta including Anomal- analysis. We conclude that Palaeoheterodonta is a odesmata) although, under some parameter sets, pa- monophyletic group (one of the best corroborated bi- laeoheterodonts and the oldest heterodonts (Carditoi- valvian clades according to our data), and is sister to dea + Crassatelloidea) do not form a clade with the Heterodonta s.1. Synapomorphies of Palaeoheterodonta remaining heteroconchs. Heteroconchia has been pre- include 3 characters with homoplasy outside the clade: viously proposed as a monophyletic group based on an aragonitic shell of simple prismatic structure (char- morphological data (Waller 1990, 1998; Starobogatov acter 7) with 3 shell layers (character 12); and 3 open- 1992; Carter et al. 2000) and on 18s rRNA sequence ings of ducts (character 102). Another synapomorphy 294 Giribet & Wheeler without homoplasy is presence of multiple acrosomal Monophyly is also yielded under all parameter sets for vesicles in the sperm (character 155; observed in sev- the 18s rRNA, CO1, and combined molecular analy- eral unionids and trigoniids; Healy 1989, 1996b; Lynn ses, as well as the morphological analysis (Fig. 2). 1994). Sperm characters are unambiguous synapomorphies for the group: 8 mitochondria in the midpiece (char- Heterodonta s.1. acter 167); a proximal centriole that splits open and unrolls to form a banded rootlet during the transitional Heterodonta s. I. comprises Anomalodesmata and the phase from spermatid to spermatozoa (character 169). heterodont orders Myoida and Veneroida. Classically, These synapomorphies and the strong molecular signal Anomalodesmata has been considered as the sister compel us to propose a sister-group relationship of group to a monophyletic Heterodonta (Myoida and Ve- these 2 superfamilies. The monophyly of this group is neroida), although other relationships have been pro- also corroborated by certain morphological features posed. Anomalodesmata has been suggested as a sister not coded in the matrix. For example, in Astartidae, group either to Myoida (Morton 1996; Salvini-Plawen the hinges, especially in Goodulliu spp., are extremely & Steiner 1996), to Palaeoheterodonta (Cope I997), or similar to those of Cuna spp. in the to Heteroconchia (Cope 2000). Morton (1 996) pro- (P. Middelfart, pers. comm.). Further analyses includ- posed monophyly of Heterodonta d., although this ing members of the families Condylocardiidae and was not supported by other morphological analyses will be required for a possible system- (Purchon 1987b; Salvini-Plawen & Steiner 1996). atic rearrangement of the 2 superfamilies. A sister- Cope ( 1997) considered Anomalodesmata and Heter- group relationship of Carditoidea + Crassatelloidea to odonta as having independent origins (thus Heterodon- the remaining heterodonts is found under 6 parameter ta s.1. would not be monophyletic), but later proposed sets, but under some suboptimal parameters this clade a clade Anomalodesmata + Heteroconchia (including groups with either pteriomorphs or palaeoheterodonts. unionoids but not trigonioids) (Cope 2000). These results are consistent with a previous hypothesis Our analyses suggest that heterodonts including an- that Crassatelloidea and Carditoidea are the most prim- omalodesmatans form a monophyletic clade Hetero- itive eulamellibranchs, and sister group to the remain- donta s,l., although neither Heterodonta s.s., Myoida, ing heterodonts (Yonge 1969). Other authors also pro- nor Veneroida is monophyletic. This result is consis- posed Astartidae as the most basal heterodont taxon tent with recent molecular analyses (Campbell 2000; represented in a 28s rRNA analysis, although no other Steiner & Hammer 2000). The internal structure of crassatelloids or carditoids were sampled (Park & 0 Heterodonta s.1. is not stable, probably because some Foighil 2000). Based on 18s rRNA sequence data, families were represented by a single taxon, or not Campbell (2000) suggested that Carditoidea was the represented at all. In Figs. 9 and 10, we have repre- most basal heterodont taxon, but no crassatelloids were sented the most stable nodes for the heterodont sub- sampled. Purchon (1 987b) removed Crassatelloidea, tree. Heterodonta s.1. can be divided into a clade Cras- Carditoidea, and Leptonoidea (= Galeommatoidea) satelloidea + Carditoidea (Fig. 9: node 2), and a clade from Veneroida, and placed them with Unionoidea in containing the rest of the eulamellibranchian taxa, in- the suborder Unionoida. The basal position of Cras- cluding Anoinalodesmata (Fig. 9: node 3). Heterodonts satelloidea and Carditoidea contrasts with the notion typically have a shell with crossed lamellar structure that members of Lucinoidea are the earliest diverging (character X), although this is found also in several heterodonts (Carter et al. 2000). Morton (1996) con- pteriomorphs, and therefore could be a symplesiomor- sidered Lucinoidea, Crassatelloidea, and Galeomma- phy for Autolamellibranchiata. Likewise, ventral man- toidea to form a clade of the most basal extant heter- tle fusion (character 51) is also found in Pterin and in odont bivalves, whereas members of Carditoidea were the mytilids. regarded as derived heterodonts. However, the data Crassatelloidea + Carditoidea. In the present presented here strongly suggest that the clade Cardi- study Crassatelloidea is represented by a single species toidea + Crassatelloidea is monophyletic and is the of the family Astartidae (Astuvte custuneu); the other sister group of the remaining heterodonts, including family, Crassatellidae, is not represented. Carditoidea Anomalodesmata. is represented by 2 species of Carditidae (Cuuditu cu- Unnamed node 3. This clade includes all heterodonts lyculatu and Curdites untiyuatu); the family Condy- except for members of the Crassatelloidea + Carditoidea. locardiidae is not represented. Monophyly of Carditi- This node is stable to parameter choice (Fig. 9: node 3), dae and of Carditidae + Astartidae is obtained and in all cases includes the former subclass Anomalo- throughout the spectrum of parameters for the com- desmata, which therefore does not warrant such a taxo- bined morphological and molecular data (Fig. 6). nomic rank. Few other heterodont relationships are stable On bivalve phylogeny 295

(Fig. 9), and many relationships suggested by the anal- many parameter sets suggest a convex relationship of yses are surprising when compared to classical taxono- both groups (i.e., Sphaerium and Corhicula nest next my. Some hypotheses are supported. Tridacninae appears to each other but do not form a clade). A character nested within Cardiidae (Schneider 1992, 1998; Schnei- supporting a putative sister-group relationship of Cor- der & 0 Foighil 1999), and therefore the taxon names hicula and Sphaerium is presence of a hypobranchial Tridacnoidea and Tridacnidae should be avoided. Car- gland (character 62). Hypobranchial glands of autola- dioidea (represented by 4 members of Cardiidae) and mellibranchs are restricted to a few pteriomorphs Chamoidea (represented by Chama) are sister taxa under (none of which were included in the present data set) most parameter sets (Fig. 9: node 4). Other stable nodes and some eulamellibranchs that incubate their larvae (Fig. 9) show monophyly of Mactroidea, Dreissenoidea, (Corbiculidae and Sphaeriidae, the only heterodonts and Myoidea (node 6) and monophyly of Dreissenoidea presenting such a structure; Morton 1977). and Myoidea (node 7). These relationships, especially Anomalodesmata. Anomalodesmata is undersam- the sister-group relationship of Dreissena and the 2 my- pled in terms of its diversity, with representative mem- oids (Mya and Varicorbula), were completely unexpect- bers of the families Pandoridae (Pandora urenmu), ed from a morphological viewpoint. Thus the orders Lyonsidae (Lyonsia hyalina), and Cuspidariidae (Cus- Myoida and Veneroida are both non-monophyletic. This pidaria cuspidata and Myoneru sp.). Rclationships result is further corroborated by many other unexpected within Anomalodesmata are therefore not extensively sister-group relationships of members of Myoida (see discussed in the context of the present siudy. Anomal- Fig. 10 for several relationships involving myoid taxa). odesmatans constitute a clearly monophyletic group Gastrochaena either appears as sister to Lasaea, or forms (Fig. 10) supported by molecular data alone. The com- a clade with Bankia and Galeomma, among other pos- bined morphological and molecular data also recover sibilities. Bunkia is sister to Galeomma in most analyses. monophyiy (Fig. 6); however, the morphological anal- The relationship of Galeomma and Bankia is supported ysis alone does not (Fig. 2). As mentioned above, an- by the' COI data, while a clade Galeomma + Lasaea, omalodesmatans are nested within heterodont bivalves, expected from morphology (but not supported by our and therefore their subclass rank is not justified. How- morphological analysis) is suggested by the 18s rRNA ever, the sister-group relationship of Anomalodesmata data (Fig. 3). These groups deserve additional study. Fi- is unstable to parameter-choice, requiring further study. nally, within the myoids, the behavior of Hiatella is un- The optimal parameter set places Anomalodesmata as expected, in being highly parameter-dependent. sister to the remaining heterodonts (excluding crassa- Other stable relationships within the derived heter- telloids and carditoids), as suggested by previous mo- odonts are the sister-group relationship of Arcticidae lecular analyses (Campbell 2000). Other parameters (Arctica) and Veneridae (Callista and Mercenaria), suggest a sister-group relationship to Hiatella (Fig. 1 Oj, making Arcticoidea (represented by Calyptogena, a but alternative possibilities exist. In agreement with re- member of Vesicomyidae, and by Arctica) non- cent analyses (Campbell 2000; Steiner & Hammer monophyletic; Corbiculoidea (represented by Corhic- 2000), Anomalodesmata is monophyletic and derived ula and Sphaerium) is not monophyletic under any pa- from heterodont bivalves, unlike previous molecular rameter sets, and generally is related to Arcticoidea analyses (Adamkewicz et al. 1997). In fact, the heter- and Veneroidea (Fig. 9: node 8). A relationship be- odont condition of anomalodesmatans had been report- tween Mactroidea, Dreissenoidea, Myoidea, Arcticoi- ed by earlier authors, who considered Anomalodesmata dea, Corbuloidea, and Veneroidea is suggested by the to be the sister group to Myoida (Morton 1996; Salvini- data (Fig. 9: node 5). This node of heterodonts is par- Plawen & Steiner 1996). Such a relationship makes ticularly interesting because it includes the 3 families Heterodonta a paraphyletic group (contra Cope 1997, of freshwater heterodont bivalves (Sphaeriidae, Cor- biculidae and Dreissenidae). The relative position of 2000; Waller 1990, 1998), unless it is rediagnosed to these 3 freshwater families was investigated with 28s include Anomalodesmata, as we propose. rRNA data (Park & 0 Foighil 2000). This analysis included several families of Heterodonta s.s. (therefore Phylogenetic conclusions no anomalodesmatans or myoids were included), and strongly suggested that Corbiculoidea (Corbiculidae The taxa analyzed here encompass a large spectrum and Sphaeriidde) is polyphyletic, with corbiculids of bivalve diversity, and offer a large and diverse set closely related to veneroids and mactroids, whereas of characters (morphological, anatomical, and molec- sphaeriids diverged earlier. No sister-group relation- ular). Many of the results obtained using character ship of Corbicula and Sphaerium is suggested by ei- congruence as an optimality criterion are stable to pa- ther the morphological or molecular data sets, although rameter choice, while others will require more data. 296 Giribet & Wheeler

Solemyoidea* vNuculoidea" Ih Nuculanoidea* Mytiloidea Arcoida Pinnoidea Ostreoidea Pteriomorphia Pterioidea Anomioidea Lirnoidea Pectinoidea Unionoidea - Palaeoheterodonta Autolarnellibranchiata I -4Trigonioidea 1- Crassatelloidea - Carditoidea Pandoroidea*** Heteroconchia Cuspidarioidea*** Hiatelloidea** Lasaeidae Lucinoidea Galeommatidae L Pholadoidea** Gastrochaenoidea** Tellinoidea Heterodonta s.I. Chamoidea Cardioidea Solenoidea Mactroidea L Dreissenoidea Myoidea"" Sphaeriidae Vesicom yidae Corbiculidae Arcticidae Veneroidea Fig. 11. Summary of bivalve relationships corresponding to the tree obtained for all the data analyzed in combination for the optimal parameter set. Asterisks: * protobranch bivalves; ** inyoid heterodont bivalves; '"** anomalodesmatans.

The phylogenetic tree proposed here (Fig. 11) large- valve relationships, and within Heterodonta s.1. The ly resembles the system presented by Waller (1990, relationships proposed can be summarized in a 1998), albeit with two major differences in basal bi- cladistic classification as follows: Bivalvia L~NNAEUS1758 Nuculoidea + Solemyoidea (Node PR-2 of Waller 1998) Nuculoidea GRAY1824 Solemyoidea GRAY1840 Nuculanoidea + Autolamellibranchiata Nuculanoidea ADAMS& ADAMS1858 Autolamellibranchiata GROBBEN1894 (= Autobranchia) Pteriomorphia BEURLEN1944 Mytiloida RAFINESQUE1815 (as Mytilacea) Non-mytiloid pteriomorphs Arcoida STOLICZKAI 87 1 (as Arcacea) Non-mytiloid, non-arcoidan pteriomorphs Heteroconchia COX 1960 Palaeoheterodonta NEWELL1965 Trigonioida DALL1889 (as Trigoniacea) Unionoida STOLICZKA187 1 Heterodonta NEUMAYR1884 (new definition) Crassatelloidea + Carditoidea Remaining hetesodonts (incl. Anomalodesmata) On bivalve phylogeny 297

If this new system withstands further testing, the sub- ans. Larval conchiferans typically have a discrete shell class Protobranchia should be regarded as pardphylet- gland with a pellicle () formed at the distal ic, as well as the subclass Heterodonta s.s., because it edge of the gland where the outer shell layer is precip- includes Anomalodesmata. At the ordinal level, major itated against the periostracum. In contrast, polypla- cophoran shell formation occurs under a thin layer of differences with previous classifications occur in the cuticle across a broad plate field (Eernisse & Reynolds Nuculoida (polyphyletic), Ostreoida (polyphyletic), 1994). Veneroida (paraphyletic with respect to Anomalodes- mata and Myoida), and Myoida (polyphyletic). Supra- Shell microstructure (characters 3-J3): data for numerous familial classifications are highly congruent with pre- bivalves have been reviewed by J.D. Taylor et al. (1 969, 1973), and Carter (1 990b). Generic groundplans have been vious classifications, with a few exceptions: Arcoidea applied when data for the represented terminal taxa were not includes Limopsoidea, Galeommatoidea iq paraphylet- available, unless more than onc species of the same ic, Cardioidea includes Tridacnoidea (as first suggested were represented. Codings for Scaphopoda from Reynolds by Schneider 1992), and Arcticoidea and Corbiculoi- & Okusu (I 999). dea form a clade that includes Veneroidea. The only 3. Mineralogic composition: (0)aragonite; (I) aragonite families of bivalves for which our hypotheses do not and calcite. Aragonite and calcite occur together con- agree with previous ones are Arcidae and Cardiidae. sistently in Mytiloidea, Pinnoidea, Pterioidea, Pecti- The sister group of bivalves is not yet clarified, but it noidea, Limoidea, and Ostreoidea, but not in Arcoidea seems that a sister-group relationship with scaphopods and Limopsoidea. Calcite also appears in shells of 2 is not supported. heterodont superfamilies, the extant superfamily Cha- moidea (reported only in Chama pellucida, not in Cha- Acknowledgments. Ken Boss, Gerhard Hasaprunar, Mich- ma gryphoides), and the extinct superfamily ele Nishiguchi, Diarmaid 0 Foighil, Patrick Reynolds, Jay Hippuritoidea (J.D. Taylor et al. 1969). Schneider, Gerhard Steiner, and an anonymous reviewer pro- 4. Nucreous structure (sheet nucre): dp. This is the vided invaluable comments and suggestions that helped to best-known and most widely studied shell structure. improve this manuscript. We are also indebted to Patrick Present in the Nuculoidea, Mytiloidea, Pinnoidea, Reynolds and especially to Clay Cook and Vicki Pearse for Pterioidea, Unionoidea, Trigonioidea, Pandoroidea, their comments and suggestions that helped to improve this and Pholadomyoidea (J.D. Taylor et al. 1969). article. We thank Gerhard Haszprunar, JosC Leal, Paula Mik- 5. Nacreous structure (lenticular ): dp. Present in kelsen, Diarmaid 0 Foighil, Kurt Schaefer, Jay Schneider, Nuculidae, Pteriidae, Mytilidae, Unionidae, Trigoni- Gerhard Steiner, and John Zardus for discussion on morpho- idae, Pandoridae, and Lyonsiidae. logical characters, and Matt Hahn for laboratory assistance. 6. Foliated structure: dp. Calcitic foliated structure that Jerry Harasewych, David Campbell, Gerhard Steiner, Penel- forms the calciostracum (or subnacreous layer) in the ope Barnes, Jay Cordeiro, 0 u L. Garcia, Lindsey T. Ostreoidea, Pectinoidea, Anomioidea, and Limoidea Groves, Peter Girguis, Olle Israelson, Eduardo Mateos, Con- (J.D. Taylor et al. 1969). stantine Mifsud, Paula Mikkelsen, Michele Nishiguchi, Dair- 7. Prismatic structure: (0) absent: (I) simple prismatic maid 0 Foighil, Cruz Palacin, Joong-Ki Park, and Elizabeth structure (aragonite); (2) simple prismatic structure Turner provided tissue samples. Pablo Goloboff and David (calcite); (3) composite prismatic structure. Maddison provided unpublished versions of their programs 8. Crossed lamellar structure: alp. for phylogenetic and character analysis, respectively. This 9. Complex crossed-lamellar structure: dp. research project was supported partially by a Lerner Gray 10. Homogeneous structure: u/p. Research Fellowship from the American Museum of Natural I I. Myostmcal pillars: a/p. History (G.G.), and by a NASA postdoctoral Fellowship 12. Shell layers: (0) three; (I) two. (G.G.). 13. Chalky lenses: dp. Chalky lenses are present in mem- bers of Ostreidae. Appendix 1 14. Flexible shell margin resulting from extension of per- iostracum beyond edge of calcified shell andor pre.s- Character descriptions. The designation a/p indicates the ence .J' secondary prismatic shell micro.c.tructure: dp. coding: (0) ubsent; (I)present. Matrix of morphological data A flexible shell margin resulting from Ihe extension of is in Appendix 2. the periostracum is found in all solemyoids (Beedham 1. Cuticle with spicules: a/p (Salvini-Plawen & Steiner & Owen 1965; Waller 1998). 1996). The classes Caudofoveata, Solenogastres, and 15. Protoconch shape: (0) cap-like (wider ihan long); (I) Polyplacophora present a dorsal surface (mantle) with tubular (longer than wide). According to Ponder & chitinous cuticle and aragonitic scales produced by sin- Lindberg (1 997), the primitive conchiferan protoconch gle or several cells, not present in any other molluscan condition is calcification of a small cap-shaped shell class. followed by incremental growth, as seen in bivc'1 1ves, 2. Discrete shell glund: a/p. Salvini-Plawen & Steiner scaphopods, cephalopods (Nuutilus;Arnold 1987), and (1996) coded for the presence of a shell in conchifer- monoplacophorans. However, in gastropods, the larval 298 Giribet & Wheeler

shell is not cap-shaped but tubular (Ponder & Lindberg Among gastropods, pallial muscles attached to thc shell 1997). Groundplan coding adopted. are not generally present (J.E. Morton & Yonge 1964) 16. Adult shell type: (0) univalve with one aperture; (I) except in limpet-like forms (Waller 1998). Among the univalve with two apertures; (2) bivalve. represented bivalves, a is absent in Atrina, 17. Shell coiling: a/p (Ponder & Lindberg 1997). Annmia and Sphuerium. 18. .shape: (0)equivalve; (I) inequivulve. The 25. : dp. Present in the represented nuculan- plesiomorphic state for bivalves seems to be equivalve oids, and in the heterodonts Spisula, Tresus, Abra, En- shells, which have become inequivalve in several sis, Callista, Mercenaria, Mya, Vuricorbula, Gustro- groups of pteriomorphs, myoids, and anomalodesma- chaena, Hiatella, Lyonsia, Cuspidaria, and Myonera. tans, as well as in a few groups of heterodonts (e.g., Coding restricted to bivalves with pallial line. Chama). Coding restricted to bivalves. 26. : (0) orthogyrous; (1)prosogyrous; (2) opistho- 19. Lateral expansions OJ' the shell (auricles): dp. Certain gyrous. In bivalves that obtain food by labial palps at pteriomorphs have lateral expansions of the shell (au- the anterior body end (Nuculidae, Solemyidae, and Nu- ricles) at each side of the umbo, such as in Pteria, cinellidae) and in those having a well-developed long Pecten, Chlamys, Spondylus, Lima, and Limaria. foot (Pisidiidae, Euglesidae, and ), the um- Coding restricted to bivalves. bones are shifted backward (Starobogatov 1992). The 20. Ryssul gape: a/p. An opening remaining for passage umbones of Ostreidae and Ensis are inconspicuous and of the byssus when the shell is closed: found in most thus have been coded as "?" Coding restricted to pteriomorphs, but not in Mytiloidea or Ostreoidea. bivalves. Coding restricted to bivalves. 27. Porphyrin-bused pigments: a/p. Porphyrin-based pig- 21. Anterior adductor muscle (or scar): (0) present; (I) ments are soluble in acid (Nuttall 1969), and they arc reduced; (2) absent (monomyarian condition). Pres- present in some pectinids, Pteria, Pinctuda, Malleus, ence of 2 muscle scars of similar size (isomyarian con- and Pinna. We have coded its presence for Pteria, dition) seems to be the plesiomorphic state for bi- Atrina (based on Pinna), Pecten, and Chlamys. valves. Certain bivalves are anisomyarian, with 28. Purple pigment in the internal shell layer: dp. This pig- reduction of anterior muscle (Mytilidae, Pteria, Atrina, ment cannot be extracted through the use of acids or Limidae, Ostreidae, Pectinidae, Spondylidae, Dreis- organic solvents (Morton et al. 1998); it is present in sena, and Tridacninae) (Yonge 1953a; Gilmour 1990). corbiculids and venerids. It has been coded as present The anisomyarian condition of certain pteriomorphs in Lusaea, Corbicula, Callistu, and Mercenaria. has led to the loss of the anterior adductor in a few 29. Shell tubules: (0)absent or restricted to early .shell or cases (Pteria, Pecten, Chlamys, Spondylus, Lima, Li- to inner shell layers; (1)present lhroughout the area maria, Anomiu, Ostrea, and Crassostreu). The anterior within the pallial line and penetrating the .shell to the adductor has also been lost in Tridacna and Hippopus. inner suqace of the periostracum. Tubules in most bi- Coding restricted to bivalves. valve groups are restricted either to carly ontogeny, as 22. Posterior adductor muscle (or scar): (0) present; (I) in Corbiculoidea (Tan-Tiu & Prexant 1989), or are reduced in size with respect to the anterior adductor. sparsely distributed and do not penetrate the cntire Certain bivalves are anisomyarian, with reduction of shell thickness. It is apparently only in Arcoida that the posterior muscle (Ensis, Gastrochaena, and tubules have a dense distribution across the entire area Bankia). inside the pallial line throughout ontogeny and pene- 23. Position of posterior pedal retractor scar relative to trate the entire shell to the inner surljce of the pcrios- posterior adductor .scar: (0) anterodorsally; (I) inset tracum (Waller 1990, 1998). Tubules with this distri- on the anterior, concave jace of a crescentic posterior bution occur in all extant arcoidean groups so far adductor scar. The pedal retractor insertions are studied (Morton 1978; Prezant 1990; Waller 1990, anterodorsal to the posterior adductor scar in most spe- 1998; Reindl & Haszprunar 1996), and thus we have cies of protobranchs and autobranchs, a position as- coded them as present in Arca, Barbatia, Striarca, and sumed to be plesiomorphic (Waller 1998). Within Pter- Glycymeris, adopting a groundplan coding for Arcoida. iomorphia, this position appears to be retained in the 30. External ligament: a/p (Owen 1959; Yongc 1973, Mytiloida, Arcoida, Limoida, and Pectinina (Waller 1978; Yonge & Morton 1980; Waller 1990). Thc lig- 1998), but not in Pinnoidea and Pterioidea. A posterior ament of boring forms ( and Teredinidae) is pedal retractor position that is inset on the anterior con- strongly reduced or absent in connection with the ne- cave face of a crescentic posterior adductor scar is ap- cessity of moving the valves in relation to each other parently restricted to Pinnidae and to the pteroidean (Starobogatov 1992). Noetiid and arcid ligamcnts have families Pteriidae, , and Isognomonidae been reviewed (Thomas et al. 2000). Coding restrictcd (Yonge 1953a, 1968; Cox 1969). Groundplan coding to bivalves. adopted. Coding restricted to bivalves with pedal re- 31. Ligament position: (0) amphidetic; (I) opisthodetic. tractor muscles. Yonge (1978, 1982b) suggested that primitive bivalves 24. Pallial line: a/p. Pallial lines are common throughout were more or less equivalve with an amphidetic exter- Bivalvia but absent in Scaphopoda and Cephalopoda. nal ligament. According to Yonge (1 982b), transfer of On bivalve phylogeny 299

the incurrent aperture to the posterior end produced dentition consists of 2 large diverging, blade-like teeth local enlargement of the posteriorly stretched external in the right valve that interlock with 2 deep and narrow opisthodetic ligament present in many modern bi- sockets in the left valve (Morton 1987c), as charactcr- valves. The ancestral ligament would have been 2- ized in Neotrigonia. The dysodont teeth of the Mytil- layered, but covered additionally by an external per- idae are not true cardinal tecth (Le Peniiec 1980), and iostracal layer. However, Waller ( 1990) concluded that thus we have decided to code mytilids as edentate, and the primitive ligament was opisthodetic. use the dysodont condition as a separate character. 32. Ligament type: (0) simple; (I) duplivincular; (2) ali- Coding restricted to bivalves. vincular; (3) transverse; (4)purivinculur. For defini- 40. Dysodont condition: a&. Coding restricted to bivalves. tions of these types of ligaments we follow Carter 41. Secondary teeth: a/p. Among the edentate monomy- (19904. arian bivalves, Spondylus and Plicatula are character- 33. Larval continues as an adult internal ligament ized by having secondary teeth (Yonge 1973). Coding (= resilium): dp (Waller 1998). Coding restricted to restricted to bivalves. bivalves. 42. Chomata: u/p (Waller 1998). Chomata are small tu- 34. Non-minerulized, non:fibrous medial core in the resi- bercles on short ridges on the hinge of the right valve lium, developed mainly ventrul to the : dp of Ostreidae, , and (Harry (Waller 1978, 1998). This type of resilium is present 1985; Waller 1998), but the members of thc gcnus in all species of Pectinoidea and unknown in other Crussostrea do not develop chomata (Slack-Smith pteriomorphs. Coding restricted to bivalves with 1998a). Coding restricted to bivalves. internal ligament. 43. Operculum: u/p. An operculum is present in all gastro- 3.5. Re,silifer: (0)absent; (I)present; (2)present us chon- pod larvae, although absent (secondarily) in the adults drophore. The ligament may sit in a hollowed out de- of most euthyneurans (Haszprunar 1988; Ponder & pression in the hinge plate known as the resilifer, lo- Lindberg 1997). cated internally just beneath the umbo. A 44. Principul growth axis: (0)anteroposterior, wilh mouth spoon-shaped, projecting resilifer (for example in the and anus at opposite ends of the shell; (1) dorsoven- mactrids) is termed a chondrophore. Coding restricted tral, with intestinal tract U-shaped, and mouth and to bivalves with internal ligament. anus near the same end of the shell (Waller 1998). 36. Lithoctesma: df~(Salvin-Plawen & Steiner 1994). A 45. Body compressed laterally: u/p (Salvini-Plawen & small calcareous plate or ossicle associated to the lig- Steiner 1996). Bivalves are unique among molluscs in ament is present in shells of Anomalodesmata. A lith- having a body compressed laterally. odesma effectively divides the ligament into 2 com- 46. Diferentiated head: (0) present; (1) absent (Salvini- pressive units assuring adequate abductive thrust of an Plawen & Steiner 1996). A head is absent in bivalves. otherwise wide ligament (Yonge & Morton 1980; Pre- 47. Snout: a/) (Ponder & Lindberg 1997). Coding restrict- zant & Carriker 1983). A lithodesma has also been ed to molluscs with a differentiated head (character 46, reported for some montacutids (Morton 198Oa; Allen state 0). 2000). Coding restricted to bivalves with internal 48. Torsion: a/p. Torsion of the shell and visccral hump in ligament. relation to the head-foot axis, followed by asymmctry 37. Pseudonymphue: dp. The pseudonymphae are modi- of the nervous system is present in gastropods. This fied ostracum secreted in advance of lamellar ligament phenomenon has been discussed at length in the liter- posteriorly and along the border between fibrous liga- ature (e.g., Spengel 1 88 I ; Naef I9 13; Wingstrand ment and ostracum (Waller 1990, 1998). Unlike true 1985; Haszprunar 1988: Bieler 1992; Pondcr 8.~ nymphae, pseudonymphae do not trap lamellar liga- Lindberg 1997: Waller 1998). ment in a groove along dorsal margin. These ligament- 49. Mantle covering visceral dorsal surface only: a/p support structures appear to be present in all mytiloids (Salvini-Plawen & Steiner 1996). A mantle covering except for those, such as Dacrydium, that have modi- the visceral dorsal surface only is present in fied ligament systems consisting only of a resilium, cephalopods and gastropods. possibly resulting from neoteny (Carter 1990a; see SO. Mantle lobes: dp (Salvini-Plawen & Steiner 1996). Waller 1990). Coding restricted to bivalves with Mantle lobes are present in scaphopods and bivalves. internal ligament. 51. Ventrul mantle ,fusion: u/p. Ventral mantle fusion is 38. Larval hinge appurutus (provinculum): (0)simple row present in scaphopods, mytilids, pteriids, and most het- of symmetrical teeth; (I) differentiated dentition; (2) erodonts (all those represented here except edentate (Cragg 1996). Provincula of numerous species Galeomma). have been illustrated (Le Pennec 1980; Lutz et al. 52. Part of the mantle epithelium secreting the periostruc- 1982a,b; Lutz 1985; Webb 1986, 1987; Goodsell et al. urn: (0) inner surjuce of the outer ,fold; (I) outer sur- 1992; Gustafson & Lutz 1992). Coding restricted to ,face of the middle fold; (2) between the middle and bivalves. outer folds. Saleuddin (1965) thoroughly discussed this 39. Adult hinge: (0) taxodont; (I) schizodont; (2) heter- subject. Field (1922) stated that the periostracum of odont; (3) desmodont; (4) edentute. The schizodont Mytilus comes from the outer surface of the middle 300 Giribet & Wheeler

fold as it is always adherent to this surface in sections. middle folds, siphons united with common outer ring Kessel (1 944) studied the histology of the mantle edge of sensory tentacles, incurrent opening fringed with fil- and the secretion of the periostracum in several bi- tering tentacles, excurrent aperture with valvular mem- valves (including Arctica islandica, brane; Type C: Also involving periostracal groove, of- echinata, and Ostrea) and concluded that in all cases ten united with very long siphons with a periostracal the inner surface of the outer fold is responsible for sheath (Yonge 1982a). Cuspidariids are coded as type secretion of periostracum. Brown (I 9.52) and Beedham B, as Yonge (1982a) found that the siphons are at- (I 958) studied the histology of the mantle edge of My- tached for half their length, and lack a pcriostracal cov- tilus edulis, Ostrea edulis, and Anodonta cygnea, ering. Coding restricted to bivalves with siphons. reaching the same conclusion. Saleuddin, however, 57. Type A siphons with the middle ,fi,ld great1.y reduced demonstrated that periostracum in Astarte is secreted and carrying the .sensory tentacles and ey~sin. the in- from the middle fold. Anodonta is used as a proxy for ner ,fold: a/p (type A +) (Yonge 1957, I982a). Certain unionids. Morton (200021) illustrated mantle margins bivalves have a greatly reduced middle fold (ventrally for Microjiragum erugatum, suggesting that in this spe- in the Tridacnidae), with sensory tentacles and eyes in cies, the periostracum originates between the middle the inner mantle folds. The inner folds alone constitute and the outer folds (proxy used for Fragum unedo). the siphons. 53. Cementation. to substrate by a calcareous secretion of 58. separation: (0) separated; (f ) ,joined. Coding the mantle margins: a/p. This type of cementation to restricted to bivalves with siphons. the substrate (as opposed to attachment by a calcified 59. Pallets: a/p. Calcareous siphonal pallets that close the byssus, typical of the Anomioidea) is found in Ostreoi- burrow when the siphons are retracted are typical of dea, certain of the Pectinidae (Hinnites), Spondylidae, Teredinidae. Coding restricted to bivalves. Plicatulidae, some species of , Chamoidea, 60. Fourth pallial aperture between the incurrent siphon Hippuritoidea, , and Cleidothaeridae and the pedal gape (or inner-mantle ,fiild.s ~fhrminga (Yonge 1979; Harper 1992; Harper et al. 2000a). For waste canal): dp.A fourth pallial aperture of unknown species in our study, this character has been coded as function is located between the excurrent siphon and present in Ostrea, Crassostrea, Spondylus, and Chuma. the pedal gape in some members of Solenoidea and 54. Adult incurrent: (0) anterior; (I) posterior only; (2) Mactroidea (Yonge 1948) and some representatives of absent (Yonge 193%; Salvini-Plawen & Steiner 1996). Anomalodesmata (Morton 198 I, 198.5; Prezant 1998; Yonge (1939a, 1957) proposed that lack of fusion of Harper et al. 2000a). From the tam represented in our the mantle lobes ventrally and the presence of an an- study, a fourth pallial aperture has been described for terior incurrent flow are primitive (plesiomorphic) Spisula suhtruncata (Yonge 1948), Pharidae (Atkins characters for bivalves. This condition is found in 1937b), and Lyonsiidae (Yonge 1952b; Narchi 1968; many protobranchs and pteriomorphs (e.g., Nucula, Harper et al. 2000a), but it is absent in Pandoridae Area tetragona, Glycymeris), whose pallial current en- (Allen 1954) and Cuspidariidae (Harper et al. 2000a). ters the cavity anteriorly and leaves posteriorly (Sa- Tresus nuttallii has the inner-mantle folds of a waste leuddin 1965). This condition also holds for some eu- canal but lacks the fourth aperture (Kellog 191.5). lamellibranchs. Most members of Lucinacea (Allen Coding restricted to bivalves. 1958), (M.L. Popham 1940), Galeomma 61. Mantle cavity separated by muscular septum (septi- turloni (M.L. Popham 1940), and Kellia suborhicularis branch condition): a/p. Present in the septibranch (Yonge 19.52a) (and Galeommatoidea in general) have anomalodesmatans. an anterior incurrent flow. In the majority of bivalves, 62. Hypohranchial gland: (0)present; (1) absent. 111 some however, the current enters and leaves posteriorly. molluscs, a hypobranchial gland lines the posterior in- 55. Siphons: a/p (Salvini-Plawen & Steiner 1996). Siphons ner wall of the mantle, typically above the ctenidia and are present in Nuculanoidea, Unionoidea, Anomalo- below the rectum (Morton 1977). A review of the hy- desmata, and in all heterodont bivalves except Cardi- pobranchial gland in Mollusca can be found in Yonge toidea. The genus Lithophuga exhibits extensive fusion (1947). Hypobranchial glands have been described in both above and below the excurrent opening that re- Nuculidae and Solemyidae but not Nuculanidae (Drew sults in an excurrent siphon (Pelseneer 191 1; B.R. Wil- 1901; Yonge 1939a,b; Morton 1977); and in some son 1979; Waller 1990). Other bivalves have functional members of Anomiidae (but absent in Anomia) (Atkins (but not morphologically defined) siphons delineated 1936; Yonge 1977); , Corbiculidae and by pallial fusions such as in Neolrigonia (Gould & Sphaeriidae (Morton 1977). It is assumed that the hy- Jones 1974; Morton 1987~).In Galeomma and La,saea, pobranchial glands of autobranchs are restricted to a as in other Galeommatoideans, one of the siphons few pteriomorphs (none of which are included in the might not be well developed (M.L. Popham 1940). present data set) and some eulamellibranchs that in- Coding restricted to bivalves. cubate their larvae. A familial groundplan coding has 56. Mantle margins and siphonal types: (0) type A; (1) been adopted. type B; (2) type C (Yonge 1957, 1982a). Type A: 63. Portion of mantle cavity occupied by gills: (0) both Union of inner folds only; Type B: Union includes lateral and posterior to the ,foot; (I) posterior to the On bivalve phylogeny 30 1

,foot. The gills are extended forward along the sides of reported A. sulcata as being type C( 1). Tevesz (1975) the body in Bivalvia (Waller 1998); however, the small considered the ctenidial ciliation of Neotrigonia to be posterior gills of Protobranchia are posterior to the of type D, as in the Unionoidea (Atkins 1937a), but foot. In Gastropoda and Cephalopoda, the gills are pos- Morton (1987~)showed thcm to be Type R(lb) terior to the foot. Coding restricted to molluscs with (ordinary filaments) as in many pteriomorphs. one pair of gills. 71. Type C gill with a groove at the ,free edge of the outer 64. Ctenidia: (0) present; (I) absent (Salvini-Plawen & demibranch (Type C(2): dp (Atkins 1937a). This con- Stejner 1996; Waller 1998). Gills with alternating leaf- dition has been observed in several Mactridae (but not lets or filaments occur in all molluscan classes except Spisula subtruncata [Atkins 1937al), Veneridac (in- Scaphopoda and Solenogastres (Waller 1998; Reynolds cluding Mercenaria mercenaria I Kellog 191 5 I), Myi- & Okusu 1999). Numerous modifications and loss of dae (including Mya arenaria [Kellog 1915; Yonge one or both ctenidia occurred within Gastropoda; het- 1923]), Pholadidae, , and Pharidae (Atkins erobranchs do not have ctenidia. Ctenidia are also 1936, 1937a). Coding restricted to bivalves with ciliary absent in septibranch anomalodesmatans. currents of type Cl . 6.5. Plicate ctenidia: (0) absent (nonplicate); (I) present 72. Chitinous rods of the gill ,filament with a major .struc- (Atkins 1937a). Coding restricted to autobranch tural enlargement in the base of the ,filament: dp.Mor- bivalves. ton (1987~)reported a similarity between the gill fila- 66. Demibranch: (0) not reflected; (I) reJected (Salvini- ments of Neotrigonia margaritaceu and certain Plawen & Steiner 1996). Coding restricted to molluscs pteriomorphs, in having the major structural enlargc- with gills. ment of the chitinous rods at the base (Mytilus edulis 67. Ctenidia: (0) eleutherorhabdic; (I) synaptorhabdic. and Harbutia virescens wcre shown as having a similar Coding restricted to Autolamellibranchiata. structure). On the contrary, the eulaniellibranchs (he 68. Ctenidial type: (0) protobranch (ctenidiobranch); (I) illustrated Anodontu woodiana) haw the major struc- Jilibranch; (2) eulamellibranch (Ridewood 1903; Pel- tural enlargement in a more distal position, and thus seneer 191 1; Atkins 1937a). In protobranch bivalves, the central stalk and apical components of each the gill consists of 2 rows of leaflets attached to a bran- filament are only flexibly supported. chial axis (Atkins 1937a). In filibranch ctenidia (most 73. Calcification oj' (f the chitinous rods o/' the gill ,$la- pteriomorphs and trigoniids), the individual filaments ments (gill spicules): dp. According to Atkins (1 938) are not united ventrally with their neighbors except by the Trigonioidea have calcareous gill spicules as seen opposing ciliary discs on the lateral bases of the fila- in the Unionoidea (Ridewood 1903). J.D. Taylor et al. ments, while the filaments are intimately united in the (1973) discussed that Trigonioidea and Unionoidea eulamellibranch ctenidia (Morton et al. 1998). Accord- may be the only bivalve groups to possess such spic- ing to Yonge (1977), Anomia ephippium (unlike other ules, which in fact should be calcified chitinous rods Anorniidae) exhibits tissue fusion, and therefore has of the gill filament, but Morton could not differentiate been coded as eulamellibranch type. Similarly, ostreids between the calcified chitinous rods of Neotrigonia and have some degree of inter-lamellar fusion and devel- other bivalves (Morton 1987~).We have adopted here opment of inter-filamental junctions, and therefore Morton's coding. have been coded as having the eulamellibranch 74. Ctenidial Jilament morphology: (0) homovhabdic; (I) ctenidial type. heterorhabdic (Atkins 1937a; Beninger & Dufour 69. Outer demibrunch: (0) present and consisting of as- 2000; Graf 2000). Coding restrictcd to Autobranchia. cending and descending lamellae; (I) upturned (type 75. Laterofrontal gill cilia: dp (Waller 1998). Laterofron- E gill of Atkins 1937a); (2) outer demibranch consist- tal gill cilia or cirri are unique to Rivalvia (Owen 1966, ing of the descending lamella only (type F gill of At- 1978; Waller 1998). Coding restricted to molluscs with kins 1937a); (3) absent (type G gill of Atkins 1937a). gills. Coding restricted to bivalves with reflected 76. Laterqfrontal cilia: (0) eulaterofrontal cilia, together demi branchs. with prolaterofrontal a (macrociliobranchiute); (1) 70. Eulamellibranch ciliary currents (gill type): (0) Type microlaterofrontal cilia (microciliobranchiate); (2) C; (I) Type D. This character reflects the ciliary types anomalous, together with paralaterojrontal cilia (At- of Atkins (1937b) found in eulamellibranch bivalves kins 1938; Salvini-Plawen & Steiner 1996; Waller with well-developed outer demibranch (types C and 1998). Coding restricted to bivalves with laterofrontal D). Since types E, E and G are considered in character gill cilia (gilled bivalves). 69, we do not account for them in this character. Also, 77. Abfrontal cilia: (0) present; (1) absent (Salvini-Plawen we have excluded from this character types A and B, & Steiner 1996). According to Salvini-Plawen & Stein- which are included in character 68 (coding for proto- er (1996), ctenidiobranch gills with abfontal cilia are branch and filibranch type of ctenidia). Yonge (1969) plesiomorphic for bivalves. reported the ciliary currents of the Carditoidea as being 78. Type of ctenidial-palp association: (0) type I; (1) type a modified type D (here considered as type D), similar II; (2) type III. Stasek (I 963) defined three main ana- to that of Astarte sulcatu, although Saleuddin (1965) tomical categories of associations between the ctenidia 302 Giribet & Wheeler

and the labial palps in bivalves. The categories were 90. Stonzuch couting/lining: (0)gastric .shield; (I) lurge1.y (I) in which the ventral tips of at least the first few or, cuticular (Salvin-Plawen & Steiner 1996). usually, of many anterior filaments of the inner demi- 91. Major typhlosole: (0) short; (I) elongated (Salvini- branch are inserted unfused into a distal oral groove (a Plawen & Steiner 1996). Coding restricted to bivalves. designation originated by Kellog 1915); (11) in which 92. Destination of the mujor lyphlosole and intestinal the ventral tips of the anteriormost filaments of the in- groove inside the st(imac1i in ,filter ,fi.eders: (0) in u.s- ner demibranch are inserted into and fused to a distal sociation with the left pouch; (I) on the left postc,rior oral groove; (111) in which the ventral tips of the an- stomach @or; (2) external to the left caecum; (3)en- terior filaments of the inner demibranch are not insert- ters the leji caecum (Purchon 1987a). Coding rcstrictcd ed into a distal oral groove, although the antero-ventral to bivalves with major typhlosole. margin of the inner demibranch may be fiised to the 93. Origin of major typhlosole and inte.rtiria1groovr wilhin inner palp lamella. Coding restricted to bivalves. the stomach that enters the left caecum and: (0)enter.7 79. Labial palp: a/p (Salvini-Plawen & Steiner 1996; within the left caecum; (I) ,fhrins u spirul coil within Waller 1998). Labial palps are present in all bivalves. the left caecum; (2) emerges and ends ,just outside the 80. Hypertrophied labial pulps: u/p. Hypertrophied labial Left caecum (Purchon 1987a). Coding restricted to taxa palps are found in members of Nuculoidea and showing state “3” in character 92. Nuculanoidea. Coding restricted to bivalves. 94. Stomach lype: (0) type I; (I) type 11; (2) t,ype Ill; (3) 81. Pulp appenduges: u/p (Salvini-Plawen & Steiner 1996; iype IV; (4) type V (Purchon 1987a). Data on stomach Waller 1998). Palp appendages are known only in Pro- structure were obtained by several authors (Purchon tobranchia (Stasek 196 I ). Coding restricted to bivalves. 1956, 1957, 1958, 1960, 1987a; Reid 1965; Dinainani 82. Pulp pouch: a/p (Waller 1998). The palp pouch is a 1967; Starobogatov 1992). Purchon (I9X7a) reviewed non-extensible prolongation of the posterior edge of 261 species of bivalves assigned lo 68 families. A each outer palp lamella that develops on the posterior stomach of type I has been described for tnembers of side of the palp appendage (Stasek 1965) of Nuculidae. Nuculidae, Malletidae, Nuculanidae; type 11 has been Some solemyids, such as Solemya reidi, have what ap- described for Cuspidariidae; type IIJ is present in mem- pears to be the homolog of a palp pouch (Reid 1980); bers of Pteroidea and Mytiloidea; type V is present in other species have only the palp appendage remaining certain members of Veneroida; a type IV stomach is with no trace of a palp pouch. Coding restricted to found in several groups of autobranchs. Coding bivalves with palp appendages. restricted to bivalves. 83. Adult with pedal retractor muscles: (0) present; (I) 95. Type I stomach. multiple looped: ..‘/I. All protobranchs absent. Post-settlement and members of Spon- except for of the Solemyoidea have a lengthened hind- dylidae lack pedal retractor muscles. Coding restricted gut. A single loop of hindgut on the right side of thc to bivalves. This character, thought to be convergent stomach is likely the primitive condition in Protobran- between Ostreidae and Spondylidae, is included be- chia (Allen 1978; Allen & Hannah 1989; Waller I WX), cause it is informative below the family level (i.e., it with multiple looping condition dcrived in several is not homoplastic between Ostrea and Crassostrea). groups: Nuculidae, Pristiglomidac, Neilonellidae, Spi- 84. Position of mouth relative to anterior adductor: (0) nulinae, and Ledellinac. Coding restricted to taxa with udjacent to posterior edge of udductor; (I) mouth lo- type 1 stomach. cated more posteriorly and not acljucent to posterior 96. Type 111 stomach with recqularly,fii,lded sorting urea: edge of adductor (Allen & Hannah 1986; Salvini- (0) present; (I) absent (Purchon 1987a). Coding Plawen & Steiner 1996; Waller 1998). Coding restrict- restricted to taxa with type I11 stomach. ed to bivalves. This character is regarded as a 97. Type III stomach with duct ori3ce.s: (0) .scattered; (I) synapoinorphy of N uculanidae. clustered (Purchon 1987a). Coding restricted to taxa 85. Rudula, odontophore, and associuted buccal organs: with type 111 stomach. (0)present; (I) absent (Waller 1998). Loss of the buc- 98. Type IV stomach with a conspicuous sorting urea on copharyngeal region with jaw, radular apparatus, sub- the anterior ,floor oj‘ the .stomach, empt.ying into the radular organ, buccopharyngeal glands, and buccal intestinal groove: (0) present; (I) absent (Purchon ganglia is a bivalve synapomorphy (Salvini-Plawen 1987a). Coding restricted to taxa with type IV stomach. 1988). 99. Type IV stomach with: (0)many duct oriJces .scattc>rcd 86. Esophageal ridges; (0) present; (I) absent (Salvini- or clustered; (I) duct ori$fices concentrated into u few Plawen & Steiner 1996). embayments (Purchon 1987a). Coding restricted to taxa 87. Digestion in midgut gland: (0)extracellular; (1)extra- with type IV stomach. and intracellular (J.E. Morton 1953; B.S. Morton 100. Type IV stomach with the major typhlosole that passes: 1983). (0)to the left pouch; (I) towards left caecum; (2).short, 88. Midgut gland ducts: (0) not ciliated; (I) ciliated posterior in position, not passing to either (Purchon (Salvini-Plawen & Steiner 1996). I987a). Coding restricted to taxa with type IV stomach. 89. Stomach with: (0) protostyle; (I) crystalline style 101, Opening of duct.r: (0) not in caeca; (I) in caeca (Salvini-Plawen & Steiner 1996). (Salvini-Plawen & Steiner 1996). On bivalve phylogeny 303

102. Num.ber of openings: (0) two; (I) three; (2) inany complex is near the posterior end of the sole of the (Salvini-Plawen & Steiner 1996). foot. It is consistently prescnt in juveniles but com- 103. Intestine: (0) normal; (I) reduced (Salvini-Plawen & monly absent in adult bivalves that do not rctain the Steiner 1996). A rcduced intestine consists of an ali- juvenile ability to secrete a byssus (Wallcr 1998). mentary canal reduced to a simple tube with 2 ducts 115. B~SSL~S:(0) uhsent; (I) present in larvae und adu11,v; to the sinall digestive diverticula and a narrow intestine (2) lost in udults. In that produce a larval but passing through the ventricle of the heart. In some rep- not an adult byssus, we assume that the adult has lost resentatives, the entire alimentary system is absent. the byssus secondarily, and therefore the character has This condition is found in Solemyidae. been treated as ORDERED. For Neotrigonia, wc fol- 104. Dorsal hood: a/p (Salvini-Plawen & Steiner 1996). low Could (1969); for Ga.stroc.Ciucna we follow Cartcr 105. Dorsovcntral muscles i-educed to two pairs or ,fkwer: ( 1978). (0)uhsent (more thnn two puirs); (I)present. Salvini- I Ih. Ontogenetic loss of ,foot inn~~diul~lyufier ,settJernznt: PIawen & Steiner (1996) used a similar character to a/j? (Waller 1998). Ontogenetic loss of the foot after characterize the Heterodonta s.I., which have reduced settlement is found in Ostreoidea (Harry 1985), and the number of dorsoventral muscles to 2 pairs or fewer thus we have coded it as present in 0,streu and in a few cases, vs. the plesiomorphic state of having Cra.r.sostreu. several dorsoventral muscle pairs. 1 17. Circumorul arms (= tentacles): ah,. Prescncc of‘ arms 106. Adductor muscles: u/p (Salvini-Plawen & Steiner is a synapomorphy for Ccphalopoda (Salvini-Plawcn 1996; Waller 1998). Adductor muscles are present in 1980; Boletxky 1988; Waller 1998). all the representatives of Bivalvia. 118. Kidneys: (0) tubular; (I) .suc-shaped;(2) I/-shuped 107. Siphuncular tub(): dp. (Brusca & Brusca 1990; Waller 1998). Thc basic mol- 108. Heart: (0)present; (1) uhsent (Waller 1998). A heart luscan kidney plan consists of a pair of tubular struc- consisting of a medial ventricle, at least one pair of tures, as in the Polyplacophora (Andrews 1988). This auricles, and an anterior aorta connecting with an open is also true in the early ontogeny of bivalves, the kid- haemocoelic blood circulation system is known only in neys becoming U-shaped in adult bivalvcs as kidney Mollusca and occurs in all molluscan classes except length increases (Raven 1966). Typically the 2 limbs Scaphopoda (Waller 1998; see Reynolds 1990). are structurally and functionally diffcrcnt, the proximal Groundplan coding adopted. limb being resorptive and the distal one excretory. 109. Burrowing ,foot with anterior enlargement: u/p U-shaped tubular kidneys occur throughout Bivalvia, (Salvini-Plawen & Steincr 1996). This type of foot is whereas Scaphopoda, Gastropoda, Ccphalopoda, and prcsent in scaphopods and bivalves, and is a modifi- Monoplacophora have sac-shaped kidneys (Pclseneer cation for burrowing. Similarities in the longitudinal 1906; Andrews 1988; Waller 1998). and transverse muscle fibers of the foot of protobran- 119. Intracellulur hemoglobin: u/p (Booth & Mangum chian bivalves and scaphopods, regarded by Steiner 1978). Hemoglobin is present in crythrocytes of mem- (I 996) as a synapomorphy for the 2 groups, are con- bers of the families Arcidae and Glycymeridae (Boyd sidered to be convergent by Waller (1998). 1998; Morton et al. 1998), Carditidae (citations in 110. Ventral surjbce of the ,fi,ot (sole): (0)present; (1) uh- Slack-Smith 1998b), and Culyptogena magn$ca (R.C. sent (Salvini-Plawen & Steiner 1996). Most autobranch Terwilliger et al. 1983). A familial groundplan coding bivalves have a foot without a ventral surface. The typ- has been adopted. ical foot has been modified in Cephalopoda, and there- 120. Hernocyunin-like molecules: (0) present; (I) absent. fore we have coded this character as inapplicable. This Hernocyanin-like molecules have bccn described for character has been discussed by Waller (1998: 21). Polyplacophora, Gastropoda, and Cephalopoda (Morse I1 1. Foot rnodijied to .fi>rman &icient creeping orgun: dp. et al. 1984; Mangum et al. 1987). Hemocyanin molc- Members of the Galeoinrnatoidea (Galeomma and La- cules have also been found in the protobranchs Sole- .saeu here) have developed a creeping habit with a rnyu velum, Nucula proximu, N. sulcatu, N. hanleyi, modified foot (M.L. Popham 1940). Acilu custrensis, Yoldiu lirnutula, and Y. thrucia

122. Cartilaginous cranium: dp. A cartilaginous cranium 130. Pallial eyes: (0) absent; (I) develop ,from the outer housing a brain formed by extensive fusion of ganglia mantle jolds; (2)develop ,from the middle mantle ,fi)lds; is found in Cephalopoda (Waller 1998). (3) develop ,from the inner mantle ,folds. Pallial eyes 123. Epiathroid nervous system with identical innervation are ectopic eyes with nervous links to the visceral gan- areas (but convergent elaboration of-‘ respective gun.- glia via the pallial nerves seen in species of Arcoidea, glia): a/p (Salvini-Plawen 1985; Hasxprunar 1988; Limopsoidea, Pterioidea, Limoidea, Pectinoidea, Car- Steiner 1992; Salvini-Plawen & Steiner 1996). This dioidea, and Laternulidae (Dakin 1928; Morton character is a putative synapomorphy for Loboconcha 2000b,c). Such eyes develop on either the outer, mid- (= Diasoma), although Waller (1998) regards it as con- dle, or inner mantle folds (Morton 2000b). Since their vergence because it is also present in Monoplacophora. positional homology is uncertain, we have chosen to 124. Lateral (pleural, visceral) nerve cords: (0) lateral or code this character as multistate. era1 nerve a major nerve encircling body, outside All the examined representatives of Arcoidea (Arca, shell muscles; (I) visceral nerve cord, inside shell mus- Barbatia; coded as “?” in Striarca) and Limopsoidea cles (Ponder & Lindberg 1997). Groundplan coding (Glycymeris) have ommatidium-like eyes developing adopted. on a sub-fold of the outer mantle fold-beneath thc 125. Viscerul and pedal ganglia: a/p (Waller 1998). Distinct periostracum (Waller 1980; Morton 1987a, 2000~). visceral and pedal ganglia occur in Bivalvia, Gastro- Eyes of this type are also apparently present in Lima pods, and Scaphopoda. In cephalopods, the specialized Lima but not in Ctenoides ,floridanus, wherc they occur nervous system is highly concentrated in the head re- on the middle folds as in Pectinoidea (Morton 2000b). gion, but homologs of the pedal ganglia have been rec- Complex pallial eyes developing on the middle man- ognized (Hasxprunar 1988). The presence of discrete tle folds (see reviews in (Morse & Zardus 1997; Mor- ganglia in the higher Conchifera but not in the Mon- ton et al. 1998) are present in virtually all shallow- oplacophora led Hennig (1979) to recognize this group genera of Spondylidae and Pectinidae (Morton et al. as the clade Ganglioneura (see also Lauterbach 1983). 1998; Morton 2000~).A thin cornea overlies a multi- 126. Visceral ganglia vs. cerebral ganglia: (0) smaller or cellular lens. Beneath this is a double retina composcd equal; (I) larger (Salvini-Plawen & Steiner 1996). of sensory and interstitial cells. Both retinas are of the Groundplan coding adopted following Salvini-Plawen inverse type, the optic fibers passing between lens and & Steiner (1996). retina. Below the retina is a light-reflecting layer or

127. Visceral connectives: (0) “(partly) chord-like; ” (1) tapetum derived from the underlying cellular pigment nerves (Salvini-Plawen & Steiner 1996). Groundplan layer (Morton et al. 1998; see Barber et al. 1967 for coding adopted following Salvini-Plawen & Steiner Pecten maximus; Morse & Zardus 1997 for Ch1amy.s (1 996). varia). 128. Cephalic tentacles: dp (Haszprunar 1988; Ponder & Simple pallial eyes (with an inverse type retina) de- Lindberg 1997). Presence of cephalic tentacles is re- veloping on the inner mantle folds are found in het- garded as a putative synapomorphy for Gastropoda. We erodonts and anomalodesmatans (Morse & Zardus follow Ponder & Lindberg (I 997) in not considering 1997; Morton et al. 1998; Morton 2000~).The 2 si- other cerebrally innervated structures of bivalves, phons of edule have about 100 eye- scaphopods, or cephalopods as true cephalic tentacles. bearing tentacles (Barber & Wright 1969), each eye 129. Cephalic (or cerebral) eyes: (0) absent; (1) open pit; consisting of a cup of reflecting cells that enclose some (2) closed eye; (3) coleoid eye. Paired cerebral eyes 12-20 receptor cells. Each eye has a thin cornea, a are found in gastropods and cephalopods, as well as in large oval multicellular lens with its long axis parallcl of bivalves. Similar paired cerebral eyes (also to the optic axis, and a single layer of columnar cells referred to as cephalic eyes or cerebral ocelli) were constituting the retina. The retina is of inverse type, noticed in adults of Mytilus edulis (Rosen et al. 1978). the nervous supply from the tentacular nervc to the They are small pigment-lined cups filled with a crys- sensory cells passing between lens and retina (Stasek talline material, located on the axial face of the first 1966; Barber & Wright 1969; Schneider 1992). This gill filament at the base, innervated from the cerebral type of simple eye has been observed on the ends of ganglia, and comprising pigment and ciliated sensory the tentacles of Parvicardium exiguum (J. Schneider, cells. They are restricted to some members of Pterio- pers. comm.) and P. pinnulatum (Meyer & Miibius morphia (Pelseneer 1899; Rosen et al. 1978; Morton 1872), and in Triducna sp. (Morton et al. 1998). It is et al. 1998). Coded as absent in all non-pteriomorphs. coded absent for Fragum unedo based on a member of Among the pteriomorphs, present in Mytilus, Arca, the same subfamily, Micrqfiagum erugatum (Morton Barbatia, Pteria, and Anomia; coded as “?” in Geu- 2000a). An eye of similar structure has been described kensia, Lithophaga, Striarca, and Glycymeris. We fol- for the anomalodesmatan truncata (Ada1 & low Ponder & Lindberg (1997) in coding the eyes of Morton 1973). Nautilus as putative homologs to the open pit of basal 131. Bivalve pallial tentacles in the mantle edge: ah? (Wal- gastropods. We have added a fourth state for the very 1er 1978). Pallial tentacles are present in numerous bi- special eye of coleoid cephalopods. valve groups, including Limidae, Ostreidae, Pectinidae, On bivalve phylogeny 30s

Spondylidae, Anomiidae, Trigoniidae, Chamidae, Gal- gill axes (Haszprunar 1983, 198Sd). The coding of this eommatidae, Tridacnidae, and Corbulidae. Coding character follows a groundplan as summarized by Ha- restricted to bivalves. szprunar (1983), with Mytiloidea (5 genera and l l 132. Type of pallid tentacles: (0) simple lobate extensions families investigated) having the AS0 outside of gill of mantle edge, poorly extensible; (I) complex autot- axes. Coding restricted to bivalves with ASO. omizing tentacles with bands oj‘ cells secreting preda- 139. Symmetry of paired ASO: (0) symmetrical; (I) asym- tor fepel1unt.s (Waller 1998). I imoid tentacles present metrical, with the kfi small or vestigial relative to the certain complex features not occurring in the tentacles right; (2) asymmetrical, with the lsft absent; (3) both of any other bivalves: (i) internal septa that subdivide reduced (Haszprunar 1983, I98Sd; Waller 1998). Ha- thc tentacle into a number of independent hydrostatic szprunar (1 983, 198Sd) summarized data on symmetry units for complex movements; (ii) rings of gland cells of the AS0 in Pteriomorphia, showing their symmct- that secrete sticky, predator-repelling mucus; and (iii) rical development in Mytilidae, Arcoidea, Limopsoidea autotomy, occurring either at the base of the tentacle (except Glycymerididae), Limidac, and Trigoniidac or at any of the septa (Gilmour 1963, 1967; Waller and greater development of the right abdominal sense 1998). Coding restricted to bivalves with pallial organ in Pinnidae, Pterioidca, Pectinoidea, Dimyidae, tentacles. Plicatulidae, Spondylidae, Ostrcidae, and Anomiidae. 133. Smsor.y mantle tentacle: a/p (Salvini-Plawen & Steiner Anornia ephippium has lost both ASOs (H 1996; Waller 1998). According to Waller (1998), a sin- 1983). Coding restricted to bivalves with ASO. gle retractile tentacle developed from the middle fold 140. Stenta’s (marginal) orgarz: a/p. A ciliated sense organ of the mantle in the region of the siphonal embayment in the middle fold of the mantle edge near thc anterior is a unique feature of Nuculanoidea. It is apparently end is universally present in Nuculanidae (Yongc absent only in Nuculanidae and in some members of 193%). It has been described for Nuculanu cornmututu but is consistently present in all other nu- (Stenta 1909); N. pernula, N. pella, Yoldia Eimutulu, culanoidean taxa (Brooks 1875; Yonge 1939a; Allen & Portlnndia isonota, and Malletia gigurzteu (Stoll I939), Sanders 1982, 1996; Boss 1982; Allen & Hannah N. minuta, N. pella, Yoldiella lucida, Mulletia obtusa- 1989). ta, and Yoldiu limatula (Yonge 193%); and several

134. “Pulp siphon: ” dp (Salvini-Plawen & Steiner 1996). species of , Neilonella, Ledella, Propeleda, Sil- We assume that the palp siphon of Salvini-Plawen & icula, Lametila, Nuculana, Malletia, Tindaria, and Yol- Steiner (1996) corresponds to the unique feeding ap- diella (Allen & Sanders 1973, 1982, 1996; Sandcrs & erturc of Nuculanoidea (Allen 1985), which marks the Allen 1977; Allen & Hannah 1989; Allen 1993; Allen place where the palp proboscides emerge from the et al. 1995). This has been termed “anterior sensc or- shell. This character was also used by Waller (1998: gan” (Allen 1985; Allen & Hannah 1986; Salvini- aperture for palp appendages). Plawen & Steiner 1996) or “anterior mantle sense or- 135. Dorsal pallial organ: dp. A pre-oral, unpaired pallial gan” (Waller 1998). Salvini-Plawen & Stciner coded gland is a unique organ of Pinnidae (Yonge 1953b). the presence of the anterior sense organ in Nuculoidea 136. Stempell’s organ: (0) absent; (I) present. This tube- and Nuculanoidea (their character 3X), but we follow shaped organ is situated immediately dorsal to the an- Yonge (1939a) and Allen (1 985) in considering that terior adductor muscle of some protobranchs (Nucula this character is a putative synapomorphy for nucleus, N. delphinodonta, and N.sulcata IDrew 1901 ; Nuculanoidea. Stempell 1898; Haszprunar 1985~1).Stempell’s organ I41. Adoral (or cephalic) sense organ: dp(Yonge I939a,b; has also been observed in Acila castrensis (Kurt Schae- Haszprunar 198%; Salvini-Plawen & Steiner 1996; fer, pers. comm.). It has been coded as “?” in N. prox- Waller 1998; Schaefer 2000). An adoral sense organ ima. It is absent in Malletia inequalis and Nuculnna has been described in several specics of Solemyidac pernula (Israelson, pcrs. obs. 1999), and thus we (Solemya togutu and S. reidi), Nuculidae (Aciln and assume that it is absent in all nuculanoids. Nuculu), Nuculanidae (Nuculanu), and Yoldiidae (Yol- 137. Abdominal sense organ (ASO): dp (Salvini-Plawen & dia) (see Schaefer 2000 for references). Absence of the Steiner 1996). An abdominal sense organ in the form adoral sense organ outside the protobranch bivalves of paired ectodermal thickenings on the posterior side has been coded as a groundplan assumption. For the of the posterior adductor near the anus is present protobranch species studied here, we follow the intcr- throughout Pteriomorphia (Thiele 1887, 1889; List pretations of Schaefer (2000). 1902; Clasing 1923; Studnitz 1931; Moir 1977; Yonge 142. Pedal reversal: u/p (Waller 1998). All known extant 1977; Haszprunar 1983, 1985c, 3985d; Morse & Zar- limoids have a unique foot that is rotated I80 degrees, dus 1997; Waller 1998), Trigonioidea (Pelseneer 189 I ; affecting the pedal nerves (Seydel 1Y09; Stuardo 1968; Haszprunar 1983, 1985d), and Unionoidea (Herbers Gilmour 1990). 1914 [cited in Waller 19981). Consequently, it has been 143. Osphradia: (0)present; (I) absent. Osphradia are ep- coded as present in all pteriomorphs and ithelial sense organs innervated by the visceral ganglia palaeoheterodonts. and functioning as chemoreceptors to test incoming

138. Position [f the ASO: (0) outside gill axes; (1) inside water (or outgoing water in some gastropods) (Krae- 306 Giribet & Wheeler

mer 1979; Haszprunar 1985a,b, 1987; Waller 1998). Geukensiu demissa, Arcticu islundica, and other mcm- They are present in all molluscan classes except Sca- bers of the families , Veneridae, Corbuli- phopoda and probably Monoplacophora (see Stork dae, Solenidae, Kelliidae, , and 1934; Charles 1966; Harry 1969; Haszprunar 1987). (see references in Reid 1990). Osphradia are present in Nautilus where they are called 147. Symbiotic zooxanthellae: a/p. Symbiotic zooxanthellae “interbranchial papillae,” but are absent in coleoid have been found in all members of Tridacninae (Yonge cephalopods (Naef 1923). Among bivalves, osphradia I980), other members of Cardiidae: i~ardissa have been described for Nuculu sulcata (Yonge 1939a, (Kawaguti 1950) and Fragum spp. (Ohno et al. 1995; 1947; Haszprunar 1987); Leda sulculata and Malletia Morton 2000a); proxy used for Fragum unedo), and in chilensis (Stempell 1898); Malletia gigantea (Stoll one species of the family Trapeziidae (Morton 1982b). 1939); Yoldiellu lucida (Haszprunar 1987); Mytilidae 148. Symbiotic zooxanthellae in the siphons and other re- (List 1902; Clasing 1923; Haszprunar 1987); Area gions of the mantle exposed to light: a,$. A major ad- noue (Spengel 188 I; Dakin I9 10; Haszprunar 1987); aptation in Tridacnidae is the enlargement of the si- Pecten (Dakin 1910); Anomiidae (Atkins 1936); Spon- phons for housing and exposure to light of thc dylidae (Dakin 1928); Unionidae (Freidenfelt 1897, symbiotic zooxanthellae. This involves hypertrophy of 1904; Kraemer 1981; Zaitseva & Sokolov 1981; So- the inner mantle fold on the upper surface, as well as kolov & Zaitseva 1982; Haszprunar 1987); Dreissena in the under surface of the middle mantle folds of polymorphu and verruc~sa(Haszprunar 1987); attached species (Yonge 1980). Venus casinn (Dakin 19 10); CorbiculaJYumineu(Dakin 149. Zooxanthella tube system: dp. A tube system (Man- I9 10); dactylus (Fnrster 1914; Haszprunar sour’s ducts) that contains zooxanthellae and connects 1987); Cerustoderma edule, Spisula .subtruncata, the digestive diverticula with the kidney (Microjirugum ., and . henslowanum erugatum; Morton 2000a), or with the kidney and the (Stork 1934). We have adopted a familial groundplan haemocoelic spaces of the siphonal tissues (Triducna coding. gigas; Norton & Jones 1992) is a putative synapomor- 144. Storuge vesicles in connective tissue: alp. Scattered phy of Fragiinae + Tsidacninae as proposed by Schnei- vesicles lined by large conical cells and surrounded by der (1992; see also Morton 200021 [but see Schneider fine strands of connective tissue may bind together the 19981). A genus-level coding has been adopted for this various organs of the visceral mass and spread into the character. mantle (Dakin 1910). These occur in several species 150. Gland of Deshuyes containing czllulolytic nitrogen- of Asturte, including A. castanea (Saleuddin 1967). fiing bacteria: a/p. Cellulolytic nitrogen-fixing bac- This character, although uninformative in the present teria have been isolated from the gland of Deshaycs in matrix, is a putative synapomorphy for the genus As- numerous teredinids (Popham & Dickson 1973; tarte, and it is included here for descriptive purposes. Waterbury et al. 1983). 14.5. Lum.inescent acinous glands: a/p. A special type of ac- Reproduction (characters 151-153): Sexual strategies vary inous glands has been shown to be luminescent in As- enormously among bivalves. Hermaphroditism occurs in turte sulcata (Saleuddin 1965). Glands of similar his- many bivalves species (Sastry 1979). It has been adopted as tological appearance and position are found in A. a reproductive strategy in virtually all representatives of An- castanea (Saleuddin 1967). Photogenic cells (of differ- omalodesmata, Galeommatoidea, Pectinoidea, Teredinidac, ent type) are also found in other few bivalves such as Ostreidae, and Sphaeriidae (Morton et al. 1998). Hermaph- in the pectinid genera Parvamussiunz and Propeamus- roditism may be (I) simultaneous (functioning as male and sium (Hicks & Marshall 1983, in , female at the same time), (2) consecutive (either protandric Gastrochaena grandis, and candida (Harvey or protogynic), or (3) alternating (functioning as male or 1952). female in regularly or irregularly alternating periods). In the 146. Intrucellulur ctenidial bacteria involved in sulphide- case of Mercenaria mercenaria, in which consecutive sex- dp. Symbiotic, chemoautotrophic uality is observed, a small proportion of individuals are gon- bacteria within bacteriocytes which play a role in nu- ochoristic. Coding restricted to documented cases; coding trition through the oxidation of sulphur are found in for Lasaea sp. follows 0 Foighil (pers. comm.), who pro- the gills of several bivalve families: Solemyidae, Nu- vided the specimens of this unidentified species and the cinellidae, Mytilidae, Lucinidae, Firnbriidae, Thyasiri- reproductive observations. dae, and Vesicomyidae (Reid & Brand 1986; Reid 1990; Distel 1998). Among the species represented 151. Number of gonoducts: (0)paired; (I)single, from pre- here, this symbiosis occurs in Solemya velum (Cavan- torsional left gonad; (2)single, .froin pretorsional right augh 1983), S. reidi (Felbeck et al. 1981; Felbeck gonad (Ponder & Lindberg 1997). 1983), Calyptogena magni$ca (Boss & Turner 1980; 152. Reproductive method: (0)free-spawning; (I) brooding Felbeck et al. 1981; Cavanaugh 1983), and Codakia to larvae; (2)brooding to ,juveniles; (3)ovoviviparous orbiculata (Berg et al. 1982). The remaining species (Mackie 1984; Kabat & 0 Foighil 1987; Kasyanov et have been coded as absent, although the absence has al. 1998; 0 Foighil & Taylor 2000). Coding of char- been documented only in Acila castrensis, Yoldia sp., acters 152 and 153 for Codukia orbiculatu based on On bivalve phylogeny 307

the congener C. orbicularis (Alatalo et al. 1984). Summers 1977); Ensis ensis (Casas & Subirana 1994); Fra- Codings restricted to documented cases. gum unedo (Healy 1995a; Keys & Healy 1999); Tridacna 153. Reproductive strategy: (0)planktotrophy; (I) strict le- gigas and Hippopus hippopus (Keys & Healy 2000); Spisula citotrophy; (2) direct development (Sastry 1979; Lutz suhtruncata based on S. trigonella (Healy 199%) and S. .so- et al. 1982a,b. Coding for Neotrigonia murgaritacea lidissima (Hylander & Summers 1977; Sousa et d. 1995); follows 0 Foighil & Graf (2000); coding for Varicor- Arctica islandica based on Trapezium suhlaevigatum (Healy hula disparilis follows Mikkelsen & Bieler (2001), 1995a); Corhicula$uminea (Kraemer 1983); Callistu chione who showed I to be separated from (Nicotra & Zappata 199 1 ); Dreissena polymorpha (FranzCn prodissoconch I1 by a distinct growth line, indicative 1983); Varicorbula disparilis based on Notocorhula vicaria of planktic development. Codings restricted to (Popham 1979); Bankia carinatu (Popham et al. 1974); documented cases. Lyonsia hyalina based on L. ventricosa (Kubo & lshikawa 1978); Cuspidaria cuspidutu based on Cuspidaria sp. (Healy Sperm (characters 154-169): A bivalve sperm consists typ- 19964. ically of an ellipsoid or conical nucleus, an acrosome of Codings for the outgroups are based on the following variable complexity, a middle piece consisting of usually 4- sources: Acanthochitona crinita based on A. garnoti and 5 mitochondria surrounding a pair of centrioles, and a fla- Lepidopleurus cajetanus based on the Lepidopleurinae Lep- gellum. This is seen in many other molluscs, but several tochiton usellus (Hodgson et al. 1988); Huliotis tuberculata modifications occur in diflerent groups. Detailed studies of based on H. midae (Hodgson & Foster 1992) and H. laevi- sperm ultrastructure are available for many bivalve species, gata (Healy et al. 1998); Diodora graeca based on D. aspera and a few broader comparative studies are available for bi- (Hodgson & Chia 1993); Sinezona confusa based on Sine- valves in general (Popham 1979), Veneroida (Healy 1995a), zona sp. (Healy 1990); Viviparus georgianus bascd on V. and Pteriomorphia (Healy et al. 2000). viviparus (Griffond 1980); Peltodoris atromaculata based on Codings used here are based on the following sources: several members of Dorididae (Healy & Willan 199 I); Nau- Nucula sulcata (FranzCn 1983); N. proxima based on N. tilus pompilius (Arnold & Williams-Arnold 1978); Sepia ele- hartvigiana (Popham & Marshall 1977); Mytilus edulis gans based on S. officinalis (Maxwell 1975); Loligo pealei (Hodgson & Bernard 1986; Sousa & Azevedo 1988; Sousa based on Loligo ,forbesi (Maxwell 1975); Antalis pilsbryi et al. 1995); Lithophaga lithophaga based on L. curta (Dan based on A. entails (Hou & Maxwell 1991). & Wada 1955); Barhatia barhata based on B. obliquata and B. foliata (Reunov & Hodgson 1994); Glycymeris insuhrica 154. Acrosomal vesicle: (0)present; (1) absent. An acro- based on G. holosericus (Healy 1996a; Healy et al. 2000); soma1 vesicle is absent in several Polyplacophora, but Striarca lactea (Healy et al. 2000); Pteria hirundo based on not in Lepidopleurinae (Hodgson et al. 1988). Pinctada sp. (Healy et al. 2000); Anornia ephippium based 155. Multiple acrosomal vesicles: (0) present; (I) absent. on A. trigonopsis (Popham 1979); Ostrea edulis (Popham The presence of multiple acrosomal vesicles in Velu- 1979; Sousa & Oliveira 1994; Sousa et al. 1995); Crassos- sunk ambiguus (Healy 1989), Anodonta grandis (Lynn trea virginica (Eckelbarger & Davis 1996); Pecten maximus 1994), and 3 species of Neotrigonia (Healy 1989, (Dorange & Le Pennec 1989); Chlamys varia based on C. 199613) has been interpreted as a putative synapomor- hastutu (C.A. Hodgson & Burke 1988); Spondylus sinensis phy for the Palaeoheterodonta (but see Peredo et al. based on S. nicoharicus (Healy et al. 2000); Limaria hians 1990; Rocha & Azevedo 1990). We have followed the based on L. fragilis (Healy et al. 2000); Atrina pectinata codings of Healy, assigning a groundplan coding for based on A. vexillum (Healy et al. 2000); Neotrigonia bed- Unionidae in order to be able to make use of this nalli (Healy 1989); Neotrigonia margaritacea (Healy potentially informative phylogenetic information. 1996b); Psilunio littoralis based on Velesunio ambiguus 156. Shape oj the acrosomal vesicle: acrosomal vesicle with (Healy 1989); Cardita calyculata based on Cardita muricata undifferentiated acrosomal contents (0); acrosomal (Healy 199Sb); Astarte castanea coded based on crassatel- vesicle with a thick posterolateral, highly electron- lids2 Eucrassatella cumingii,E. kingicola, and Talabrica au- dense basal ring (I); acrosrrmal vesicle with its con- rora (Healy 1995b); Galeomma turtoni based on Divariscin- tents diferentiated into a highly electron-dense ante- tilla yoyo, D. troglodytes, and Cintilla sp. (Eckelbarger et al. rior layer and a less dense posterior layer (2); broad 1990); Lasaea sp. based on L. subviridis (0 Foighil 1985a); acrosomal vesicle with a differentiated, apical wedge- Codakia orhicularis based on C. punctata (Healy 1995a); zone (3); acrosomal vesicle with a highly electron- Chama gryphoides based on C. macerophylla (Hylander & dense internal layer which recurves posteriorly, giving a double-layered effect through most of’ its length (4); 2This coding assumes monophyly of Crassatellidae + As- acrosomal vesicle with contents differentiated into a tartidae, and that the sperm anatomy described for 3 cras- very dense inner layer surrounded by less dense ma- satellids is the groundplan for the common ancestor of terial (5); acrosomal vesicle with contents diferenti- Astarte and the crassatellids. Astarte sulcata was investi- ated into a very dense outer layer .surrounding a less gated by means of light-microscopy by FranzCn (1955), dense material (6);acrosomal vesicle with anteriorly who found that the inidpiece contained only 4 mitochon- truncate projile ( “Pundoroidea type”) (7). dria and that the nucleus, although elongate was not Healy (1995a) differentiates among the acrosomal capped by an obvious acrosome. vesicle of his types A and B vs. C in that the members 308 Giribet & Wheeler

of type C have the electron-dense basal ring developed (Sousa & Azevedo 1988; Sousa et al. 1995) is clongate longitudinally. Here we code these three types within and refined apically into a peg-shaped structure pene- the same category (short conical acrosomal vesicle trating deep into the invagination of the acrosomal with an electron-dense basal ring). Differences within vesicle. this state could be considered as an additional charac- 164. Mitochondria1 midpiece: (0)present; (I) absent. A mi- ter. Several other autapomorphic types of acrosomal tochondrial midpiece forms part of the typical mollus- vesicles occur in molluscs. can sperm, but those of several cephalopods lack a 157. Acrosomal contents with a highly electron-opaque lay- midpiece (Healy 1996a). er associated with long radiating plates: (0)present; 165. Location OJ‘ mitochondria: (0) in midpiece; (I) en- (I)absent. In their review of‘pteriomorph spermatozoa, closed within a membrane sac; (2) two mitochondria Healy et al. (2000) proposed a putative synapomorphy running along lateral furrows in the nucleus (Maxwell for Pterioidea, Pinnoidea, and Pectinoidea, the pres- 1975; Arnold & Williams-Arnold 1978; Healy 1996a). ence of acrosomal contents with a highly electron- 166. Distribution of mitochondria in midpiece: (0)regularly opaque layer associated with long radiating plates, distributed around centrioles; (1) e,wentrically distrib- sometimes with additional differentiation zones. In Os- uted (C.A. Hodgson & Burke 1988; Healy 1996a). treidae, radiating plates are associated with the basal Coding restricted to molluscs with mitochondrial region of the acrosomal vesicle of Dendrostrea folium midpiece. but have not been detected in other ostreids examined. 167. Eight mitochondria in midpiece: u/p. Bivalve spcrm Lacking data on transverse sections of the acrosome in typically have a midpiece consisting of (usually) 4-5 other ostreids, we have coded them as “?.” This char- mitochondria surrounding a pair of centrioles, although acter has been investigated mostly for Pterioinorphia other configurations exist. This character recognizes (Healy et al. 2000) and cannot be coded for most other the presence of typically 8 mitochondria in members terminal taxa. of Carditidae and Crassatellidae, a configuration not 158. Position ofthe ucrosomal vesicle: (0)anterior; (I)pos- found in any other bivalve studied so far (Healy terior, at the side of the midpiece. A posteriorly posi- 199%). Coding restricted to molluscs with tioned acrosome (the “temporary acrosome” of Kubo mitochondrial midpiece. 1977) has been found in the families Lyonsiidae, La- J68. Membrane skirt: a/p (Maxwell 1975; Arnold & ternulidae, and Myochamidae (Kubo 1977, 1979; Kubo Williams-Arnold 1978; Healy 19963). & Ishikawa 1978; Pophdm 1979; Healy 1996a). Here 169. Proximal centriole splits open and unrolls 10 fiwm n we have coded Lyonsia hyalina based on the coding banded rootlet during the transitional phase from .yper- matid to spermatozoa: (0) absent (two centrioles pre- of L. ventricosa (Kubo & Ishikawa 1978). On the con- trary, Cuspidaria sp. (proxy used for C. cuspiduta) sent); (I) present. At the beginning of‘ spermiogenesis, 2 centrioles are present. In typical molluscan sperm, possesses an anteriorly positioned acrosomal vesicle both centrioles are conserved and are positioned at (Healy 1996a) resembling the one of Notocorbula vi- right angles to each other. The proximal centriole is curia (Popham 1979). The temporary acrosorne has not oriented perpendicular to the axoneme, and ihe distal been observed in any non-anomalodesmatan bivalve. one in line with the axoneme; the distal one forms the 159. Anterior act-osomal vesicle: (0) oriented ,filllowing the basal body of the flagellum (Dohmen 1983). In some longitudinal axis of the sperm; (I) arranged at a coiv cases, such in Latemula limicola, the proximal centri- siderable ungle to the longitudinal of the sperm. axis ole moves to the lateral side of the distal centriole, so An acrosomal vesicle arranged at a considerable angle that the 2 centrioles are parallel (Kubo & Ishikawa to the longitudinal axis of the sperm (sperm type of 1978). The reduction of one of the 2 centrioles during group B of Healy 1995a) is found in Lasuea subvividis, spermiogenesis maturation has been observed in Pseudophytina rugifera, Scintilla sp., Divariscintilla Carditidae and Crassatellidae (Healy 199%). yoyo, and D. troglodytes (0 Foighil 1985a,b; Eckelbarger et al. 1990). Developmental data (characters 170-1 81): Data on cleavage 160. Anterior nuclear fossa: a/p (Healy 1990). and germ layer formation are scarce and we have adopted 16J. Subacrosomal space contains granular material jbrm- different groundplans. We have used the developmental data ing an axial rod (perforatorium): dp. The subacroso- of Heath (1899) on Stenoplax heathiana (Ischnochitonina) ma1 material polymerizes at the time of the acrosome as a proxy for Lepidopleurus and Acanthochitonu (Lepido- reaction. However, in several species, the subacrosomal pleurina and Acmthochitonina, respectively) (see Pearse substance is organized into a more or less completely 1979). Data from Antalis sp. (Lacaze-Duthiers 1856, I 857a, preformed acrosomal filament or axial rod (Popham 1858) are used as a proxy for A. pilsbryi. Detailed studies 1979; Dohmen 1983). on cell lineage for bivalves are available for the following 162. Nucleus with eccentrically positioned ,flagellum: a@ species: Ostrea edulis (Fujita 1929), Lasmipma complanata (Healy 1996a). (Lillie 189.5) (proxy used for Psilunio and Lampsilis), Dreis- 163. Nuclear peg penetrating deep into the invagination cd sena polymorpha (Meisenheimer 1901), and Sphaerium the acrosomal vesicle: dp. The nucleus of Tridacna striatinurn (Woods 193 I) (proxy used for Sphaeriuni stria- maxima (Keys & Healy 1999) and Cerastoderma edule turn). Embryological data for Nucula spp. based on Nucula On bivalve phylogeny 309 delphinodonta (Drew 1901); Pecten muximus based on P. Reid 1986; Waller 1998). Coding restricted to taxa with tenuicostatus (Drew 1906); Chlumys varia based on C. has- pericalymma larva. tutu (C.A. Hodgson & Burke 1988); Codakia orbiculata 176. Trochophora larvae: (0)present; (I) absent. based on C. orbiculuris (Gros et al. 1997). 177. larvae: a/p. The molluscan veliger- larval stage- is a link between the trochophorc and the pediveliger 170. Mode of development: (0)unequal cleavage; (1)equal stage, and is represented in some members of Bivalvia cledvage; (2) yullzy meroblastic egg, with non-spiral and Gastropoda (Carriker 1990; see Cragg 1996 for cleavage, and direct development. The typical cleavage references). in molluscan eggs is a modification of holoblastic ra- 178. Pediveliger stage: a/p. The swimming-crawling, bi- dial cleavage (Verdonk & Van den Biggelaar 1983), valve-shelled pediveliger is a critical stage between and has been called spiral cleavage by E.B. Wilson planktonic and benthic existence in most bivalves. A (1892). All cephalopods seem to have development review of the literature by Carriker (1990) confirmed that involves a yolky meroblastic egg, non-spiral cleav- the presencc of a pediveliger in 31 familics and 66 age, and direct development (Bandel & Boletxky 1979; genera of bivalves. The pediveliger larva possesses a Waller 1998). 2-valved, hinged, mineralized shell; a strongly ciliatcd 171. Cleavage with polar lobe jbrmation: u/p. Polar lobe velum; and a densely ciliated, powerful foot (Carriker formation during cleavage has been found in numerous 1990). Detailed descriptions of pediveligers are those molluscs with spiral development (see Sastry 1979; of Ostrea edulis (Yonge 1926; Cole 1938), Crussostrea Verdonk & Van den Biggelaar 1983; Freeman & Lun- virginica (Galtsoff 1964), Mytilus edulis (Bayne 1971), delius 1992; Ponder & Lindberg 1997; and references Chlarnys hastatu (C.A. Hodgson & Burke 1988), and therein). Codakia orbiculuris (Gros et al. 1997). For a revision 172. Molluscan cross during development: (0)present; (I) on the subject and citations for the species here coded, absent. In molluscs, the ectoderm is divided into a pre- see Carriker (1 990). trochal and a post-trochal region by a band of ciliated 179. Statocysts in pediveliger .stage: (0) Jingle statolith in cells, the so-called prototroch cells. The pre-trochal re- each statocyst; (I)several statoconia in each statocyst. gion (the future head region) originates from the first Larvae of many species of gastropod and bivalve mol- quartet of micromeres (I a-d), formed at third cleavage. luscs develop statocysts in the swimming/crawling In all molluscs except bivalves, this first quartet forms stage prior to metamorphosis. In bivalve pediveligers, a typical structure, known as the molluscan cross (Ver- these statocysts may contain either a single statolith or donk & Van den Biggelaar 1983). Within the bivalves, several small statoconia (Cragg & Not1 1977; Carrikcr a structure similar to a molluscan cross has been ob- 1990; Cragg 1996). Coding restricted to bivalves with served in Solemya reidi (Gustafson & Reid 1986) and pediveliger. 180. Pallial eyes in pediveliger .stage: (0)present; ab- S. velum (Gustafson & Lutz 1992). (I) sent (Cragg 1996). Eyes of bivalve pediveligers lie 173 Apical tufi. on the larva: (0) present; (I) absent. An roughly at the center of each valve just beneath the apical tuft is common in the free-swimming larvae of larval shell, each consisting of a pigmented cpithelial the Bivalvia and in outgroups (Cragg 1996; Waller cup surrounding a central amorphous lens, the open 1998), although it is absent in some members of Os- end toward the exterior of the larva, and a nerve treidae, including Ostren edulis (Waller 1981) and leading inwards from each (Carriker 1990). virginica. Crasso.streu 181. : a/p. Glochidia have been described only 174 Pericalymma larvae: dp. Pericalymma larvae have for Unionoidea (see Hoggarth 1999), and thus all non- been described for only a few species of protobranch unionoids have been coded as absent, even though for bivalves: 4 nuculids, Nucula proxima, N. turgida, N. some of them the larval development is unknown. For delphinodonta, and Acila castreasis; 2 nuculanids, Nu- the species used in this study, the glochidiiim of Psi- culunu pernula and N. ,fi).r.vu; a yoldiid, Yoldia limatula; lunio littoralis has been illustrated by Altaba ( I992), and 2 solemyids, Solemyu reidi and 5'. velum (Drew and the glochidium of Lampsilis curdium was observed 1897, 1899a,b, 1901; Trevallion 1965; Gustafson & by the authors. Reid 1986, 1988a,b; Gustafson & Lutz 1992; Zardus 182. Swimming capacity through valval movrment: d12. & Morse 1998). Larvae of 2 other species, Nucula nu- This type of valval movement has been dcscribed for cleus and N. nitida, were raised by Lebour (1 938), but a few pectinids (Morton 1980b), and we have observcd aside from mentioning their barrel-shaped form, he it in Limaria hians and in Pecten maximus. provided no description (Zardus & Morse 1998). A 183. secreting a calcareous tube: a/'. Calcareous groundplan coding of absent has been adopted for the tubes are secreted by certain myoids and certain an- non-protobranch molluscs. omalodesmatans. From the taxa here reprcsented, cal- 175. Ciliation of test-cell larva: (0) in distinct hands; (I) careous tubes are built by Gustrochuenu (Carter 1978) entire test un$irmly ciliated (Drew 1901 ; Gustafson & and Bankia (Turner 1966). 3 10 Giribet & Wheeler

I I I I I I I i i I I I I I I I i i I I I I I I I i i I I I I I , I i i I I , I i i I I I I -1 /I c> 0 I I I I i I I I I I I I I

30C) 0 0 0 0 000 Ci 3000000 0000 C. C. C. C. C. I?. I?. C. 0 0 0 c. c, 0 30 0 0 0 0 0 000 -1 c. -1 I/ 30 0 0 0 0 0 000 0 c. c, 0 300 0 0 0 0 000 -1 ,i i ,/ I I I I I I I I I I0I0 I0I0 I 0 I0I0 10I0 I0 I0I0 I0 0 I0I0 10I0 I0 I0I0 I0 0 I0I0 I0I0

II 11 I1 I1 I1 I1 11 I I IC. P. C. 0 C. C. 3101 01 0101 0101 0101 I I -1 I c. I0I0 I C. I0I0 I0 I0 I0I0 0 0 0 0 C. 30 00 0 C. 04 0C. 0 C. 0 C. OC. I0 0 0 -1 0 C. I0I0 I C. I0IF. I C. I C. IC. 1-1 C. d 0 -1 c. 3101 0 C. 0101 0101 01 I I C. .i 4 -1 C. I0I0 I C. I0I0 I0 I0 I0 I0 C. d I -1 c. C. 0 C. 0 C. C. C. 0 P. I\. c. P. c. I?. 0. C. , 0 0 d c. /I c. C. C. 0 C. 0 C. C. C. 0 C. C. C. C. 0. c. 0. 0 I P. C. C3 Cl 0 C. I0 I0 C. I C. Id IC. I e. I c. I0 1-1 C. 0 I 0 -1

IdI /I IN IN I(U IC. IC. Id IC. I C. I ". I0 I0 I0I0 I0 I0I0 I0 I0 I0I0 1 C. I c. IC. I0 30 c70 0-I 04 Oil 00 00 00 00. 0 c. 0 c. Ie. I0 I0 I0I I I I I I I0 IF. 10I0 I c. I c. IC. I0 30 Ori 0 I0I0 I 000C.00 0e. 0 C. 0 P. IC. IC 10 I0IN 151 Id10 I010 re. I C. I C. I0IC 30 a0 0-1 0-1 0-1 00 00 00 00 0 c. 0 [j. do dc: IC? I I 1-10 -10 -10 -10 -10 ,I d -1 c. ,-I e. -1 I?. ii C. I rl 30 0-1 00 -1-1 r-Id 00 C.O 0 0 c. P. c. c. C. P. -1 0 C. 0 I0I I I0I0 I0 lrl 1-1 I0IP. I C. I C. IC. I0

10 I I 10 I0 0 1-1 -1 I0 IC. I c. I C. I0I0 IC3 I I 10 I0 0 I0 0 I0I0 I c. I C. I0I0 I0 I I I0I0 0 I0 0 I0I0 I c. I C. I0 I0 IC. I I I C. I C. c. I0 c. I e. I c. I C. I C. I C. I C. I0 I I I C. I c. C. I0 0 I0IC. I C. I C. I C. I I\. I0 I I I0 I0 0 10 0 I0I0 , I?. , I?. I0 I0 I0 I4 I0I0 0 I0 0 I0I0 I C. I C. I0I0 31 0-1 0ri 0ri 0 N 0 C. C. 0 e. 0 c. 0 C. 0 C. 0 e. 0 Ti ?. 0 c. 0 P. N c. 01 c. N 0. 0 c. C. 0 C. C. c. e. C. P. c. 0 C. o 30 00 P. 0 c. N c. N 0 -1 d 04 0-1 04 0-1 0-1 -IN

30 00 00 00 0 0 00 00 00 00 00 00 00 dC -1 0 -10 -1 0 30 00 do -10 -1 0 ,4 0 -10 -10 -10 -10 40 -10 0c 0 0 00 0 0 30 00 00 00 0 0 -10 -10 do -10 -10 -10 -10 00 0 0 00 0 0 30 00 00 00 0 0 -10 -10 ri0 do -10 do -10 00 0 0 I0I 0 30 00 00 00 0 0 00 00 00 00 00 00 00 00 0 0 -id -1 /I 30 00 00 00 0 0 00 00 00 00 00 00 00 00 0 0 -10 -1 0 30 00 d 0 -10 d 0 do -10 -10 -10 -1 0 <-+ 0 do -10 -1 0 00 0 0 30 00 00 0d 0 -1 ri 0 -10 -10 do -10 -10 do 0-1 0 4 0 c. 0 c. I0 I0 I0I0 I 0 I0 I0I0 I0 I0I0 I0 I0 0 00 0 0 I0 I0 I0 I0I 0 I0 I0I0 I0 I0I0 I0 I0 0 0 0. 0 ri

I I I I I 00 0 I I I 1 bil I I I I I I Nl 0 0 0 0 0 0 00 0 0 0 0 0 0 0 00 0 0 0 cl 0 0 0 0 do -1 0 I I I I 00 0 0 0 0 0 0 0 \i0 -1 0 I I 1 -1 I 1 -1 C3 c?

I0I0 I0I0 I0 I0 I0I0 I0 I0I0 I0 I0I 0 -1 c3 71 0 0 i\ c3 0 30 00 0-1 om om 0-1001 ON ON ON ON ON 00 00 00 00 00 00 00 00 00 00 00 0d 0-1 0d 0d 0-1 0-1 0-1 00 00 00 00 c-,c 00 30 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 I0I0 I010 10 I0 I0I0 I0 I0I0 I0 I0 I0NO NCJ NO mC> I0I0 1-1 1-1 1-1 1-1 1-1 Id 1-1 1-1 1-1 1-1 1-1 I -1 0d 0-1 051 0-1 30 00 0-1 0-1 Or1 c.-1 C.4 -1rl0-1 0d 0-1 I -1 00 00 -10 -1c -10 -10 I0 I0 I0 I0 I0 I0I0 I0I0 I0 I0I0 1-1 1-1 O.i 0-1 Or~I0-1 I0I0 1-1 1-1 15.1 I0I0 I0I0 I0 I0I0 I0 I000 00 00 <3c, I0I0 I0 I0 I010 I0 10 10 I010 10 Id I d 00 0O 00 (10

I0I0 I0 I0 I0 I0 10I0 I010 I0 I0I@. I c. 00 0 C. 00 I0I0 I0I0 I0 I0 I0 I0 I0 I0 I0 I0 I0 I000 0 0 00 I0I0 1-1 1-1 Iii 1-1 1-1 841-1 1-1 1-1 I -1 I r-I 1-1 ON 0 N om 00 00 -14 04 0-I -10 -10 -10 -10 ria -10 I000 00 00 0 0 00 I0I0 00 0.0 C.O 00 00 00 00 00 00 I0-10 -10 NO N 0 NO I0 I000 I0I0 -10 -10 -10 -10 -10 -10 do 00 00 00 0 0 00 I0 I000 00 00 00 00 00 00 00 00 I000 00 rl 0 -1 0 0 0 C. 0 C. 0 00 I01000000000 00 00 I000 00 00 0 0 0-1 C. 0 P. 0 00 I0I0 00 00 00 C.O C. 0 C. 0 I000 00 rl 0 -1 0 0 c C. I P. I 17.1 I I I 10.1 e.1 P.1 c.1 C. I C. I I I C.1 c. I 00 0 c, 00

C. c. 0 0 01 I I 00 00 00 C.O C. 0 C. 0 I0C.0 C. 0 1-1 0 -10 0 C. 0 C. 0 00 I1 0 00 00 00 c.0 C. 0 C. 0 I 0 c. rl r.. 4 0 ii 0-1 -1 0. 0 C. 0 00 I0 0 00 00 00 c.0 C. 0 C. 0 I 0 P. \-I c. -1 00 00 0 c. 0 C. 0 -1-1 I0 I4 -10 rl 0 -10 C.O c. 0 C. 0 I0-10 -10 -10 -10 0 C. 0 C. 0 00 I0 0 00 00 00 0.0 c. 0 C. 0 I000 00 0 5-1 0-1 -1 0. 0 C. 0 -10 I0 0 -10 rl 0 -10 C.O C. 0 C. 0 I00i4 04 00 00 -1 0 C. A. C. 0. -1 P. I 0. c. -I P. -1 c. d c. c. c. c. c. c. c. I e. 00 00 01 0 I-/ -I ,I 00 00 00 I0 0 00 00 00 00 ". 0 c. 0 I000 00 01 0 -1 0 0 C. 0 0. 0 d e. ri P. c. -1 e. d c. -1 c. -1 c. rl C. -1 C. rip. -1N -1N 7-1 0 -10 -1 /I d0dC 00 00 0 00 00 00 00 00 00 00 00 00 00 00 0 0 On bivalve phylogeny 31 1

I I I I 1 I 1 I 1 0 I I I I I I I I I I 0 I I I I I I I I I I 0 I I I I I c3 0 0000 0 I I I I I -10 00 8-1 1 0 0 00 I I I I I I 0 0 00 NN NN 01 m I I I I I 1 I I I I I I I , 0000 0 i' 0 0 00 d-1 , -1 4 -1-iri-iii

:.> 0 00 0000 00000 0 0 00 -1-1 -1 r/ 3-1 ,-I -1 -1 ,+ 0 0 00 d-1 -1 %-I -1-14-1-1 4 d r I ,-I 0000 00000 0 0 00 -14 -1 i-4 -Id-14-1 ?I /I -1-1 -1d-1 -1-1 5-1 d -1 -1 -1 -1 -1 d ,I -1 -1 -100 00000000c 00 00 00 0c 00 00 00 00 00 00 00 00 00 00 OC i3c-3 00 00 00 0c 00 I0I0 I0I0 I0 I0I0 I0 I0I0 Id CI 0 I/ 0 -10 -1c -10 00 00 00 00 00 00 00 00 00 00 00

rl I 8 I c. -1c. oc. 00 0c.00 0 c. 0 e. 61 I d c. -1 c. d P. d -I 7-1 c. 1-1 c. ri c. d P. 00 0 c. 0 c. 0 c. 0 .-1 0e. N-1 N c. c\1 e. do -1 c. -1 c. 0 1-1 0 -I 0d 0-1 0 e. 0 c. 04 0 e. 0".00 00 00 -10 d c. d c. 40 -1 I\. d C. <+ I %..I I 41-11 did1 1-1 I c. I e. 00 00 00 00 c.0 00 c. 0 0. c. e. c. c. c. 0 0 c. e. c. e. 0 c. 0. c. c. -1 c. e. c. c. c. c. -1 -1 c. c. c. c. -1 e. 0. c. In I e. I?. IC. 1-1 I c. I c. I e. I c.

I0I c. I C. I c. I 0. I e. I e. I e. I c. I c. I P. I c. 000e.00 00 0 P. 0 c. 0 C. 0 c. d c. -1 c. -10 do I c. I c. I c. I c. 00 0e. 00 00 c, e. n c. 0 c. 0 c. -10-1c. -10 d0 , c. I C. 1 0. I c. 00 OC. c.0 00 0 c. 0 e. 0 c. 0 c. 00 0c. 00 00 , I c. d c. d c. 4 e. 0 c. 0 c. 00 00 , 4 C. 4 c. d 0. 4 c. TI 0 ,-I c. 40 do , C. I e. I 0. I e. 0 c. 0 c. 00 00

0 c. 0 c. 0 e. 0 P. 0 ,--I 0 c. 0 c. 00 00 00 0-1 0-1 oc. i>d 0 c. 0 c. 0c. oc. 00 00 00. 00 00 0 0 0 0 0 0 C'(\. i'C.3 $-I c. 4 c. -1c. I e. I0 I0IC. I0 I0I0 I0 10 10. I<> 0. c. c. c. c. c. I c. I 0 c. P. I c. 1010 I0 1-1 1-1 1c. Id c. c. c. c. e. e. I c. I 10 c. c. I c. I0 I0 .. ." . . L I c. 54 c. -1c. oc. 00 -10 0c. 00 00 00 00 00 0c. 00 .ic. -1 c. dc. di. -10 do 0c. 00 00 00 -10 -10 -1c. -1c3 0-1 0 c. 0c. 00 00 00 00 0 c. 0 P. 0e. 00 00 00 00 0 c. c. c. c. c. c. c. 0 0 0. c- c. 0 c. 0 c. c. 0.O~ n.*. 0.n. n.n.. 0.n... 00 00 00 do rl0 do 00 00 00 00 do 00 00 00 -10 -10 -10 -10 ri0 do -10 -10 -10 -10 do -10 -10 do -10 do 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 I0I0 I0 I0 I0 I0 I0I0 I010 I0 I0I0 I0 I0I0 do -10 -10 rl0 ri0 d0 -10 -10 -10 -10 do -10 -10 do 71 0 -1 0 do -10 do -10 -10 do -10 -10 -10 -10 d0 do do 40 -10 -1 0 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 0e. 00 00 00 00 00 oe. 0 c. 0 e. 0 c. 0 c. 0 e. 00 00 00 00 00 00 00 00 00 00 00 00 00 00 0-1 0 ,-, 00 0 c. 0e. 00 0e. 00 00 00 00 00 00 00 00 00 00 00

04 0 c. Oil 04 0e. dC3 40 -10 00 00 00 00 00 00 00 00 0 I 01 0101 0 I O0-0 e0 00 00 00 0-1 -4-1 q-1 e0 00 01 01 0101 0 I000000 0-1 0d 0-1 0-1 0-1 0-1 0-1 0d 00 00 00 00 00 4-1 44 -id 0-1 0-1 0-1 0-10-104040d 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 d 0 -10 -10 id0 140 00 00 00 00 00 00 00 I0I -1 -10 -10 0-1 0-1 0d Oil 04 00 00 00 00 00 00 00 I0I00000 -1-1 -1-1 -10 ,I0 54-1 410 -10 -10 -10 40 10 -10 00 00 -10 -10 01 01 0101 0lNINl N1-11 -1Idl ri I N I m I ~-1~4 00 00 00 00 -10 do -10 -10 00 00 00 00 -10 40 0-1 0d

d0 -10 -10 ,I0 -104-1 dd dc. -Id d 0 40 -Id d c. 00 00 00 00 0c- -1-1 -14 4P. -1 e. 0 -1 00 00 00 00 00 00~ no-- 00 00 00 00 00 0 0 00 00 00 00 00 00 00 0d 04 0-1 04 0 <-I -1 -1 dd 0 ii 0 1-1 010 NO NON0 -1d -14 -14 0-1 0-1 0 ,-I 0 ,--I 0 #I 0 ,-I d 7-4 -14 -14 -14 0d 04 0d 0-1 0-1 0 d I Ti 0 >-I ad TI 0 -10 40 40 -10 -10 do do do d 0 00 40 -10 04 04 0d 0-1 0rl 0d 0d Od 0 ii -1 4 -1-1 04 0 -i 00 00 00 00 00 00 00 00 00 0 0 00 00 00 00 00 00 00 -10 00 00 00 00 N 0 TI 0 NO N 0

n i? n. 00 0 c. - n. - 0 c. 0 c. 0 e. 0 C. -1 P. -1 c. rl ,-I d c. --I c. d 0. -1 c. 00 00 00 00 00 00 c>0 00 0-1 0d 00 04 -1 0 00 -10 ON ON ON ON ON ON ON or% ON ON ON ON d N -IN ON 00 00 00 00 00 00 00 00 00 00 00 00 0 0 00 00 ru0 m0mo a0 NU NO NO NO NO NO NO NO 01 0 NO m 0 00 00 00 00 00 0-1 04 001 0d 0d 04 ON 0 -1 0-1 0-1 00 00 00 00 00 0-1 04 0ci 0-1 0-1 04 0-1 0 -1 0-1 0-1 00 00 00 00 00 00 00 00 00 00 00 00 0 0 00 00 -10 -10 -10 -10 00 40 -10 00 rl0 840 10 -10 0 0 do -10 00 00 00 00 c.0 P.O do -10 -10 d0 P.O -10 v-1 0 00 00

% i 0 -10 -10 -10 -10 4-1 0-1 0d 0-1 04 04 0-1 0-1 0-1 0-1 ad 04 04 0d0-1 c.4 0-1 0-1 0 d <4 ri -14 d-1 -1-1 0-1 04 -1-1 -14 00 00 00 00 c.0 00 00 00 rl0 crl0 do -10 00 00 00 nc> 00 00 00 00 e.O c.0 maNO 00 00 00 00 NO NO NO 00 0 ,-I 04 0d0-1 c.4 04 0d 0d 0-1 Or1 04 Or1 04 04 -1-1 -1-1 00 00 00 00 c.0 do -10 r-I 0 0 0 00 00 00 -10 00 00 00 0-1 0-1 Vd 0-1 c.4 d-1 drl 4d 0-1 04 04 0-1 -1-1 4-1 0-1 Vd 00 00 00 00 00 do -10 do 00 00 00 00 -10 40 -10 -10 d c. -1 c. -1 C. -1 c. -1 e. ci 01 rl N 4m -IN -1N dN -101 -1N 4N -1N -1m 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00

m -a v1 Y .-M E .3M E" ._m m 9 s 0 w Appendix 2. Continued. r r.3 r.3

Chlamys varia 0110010100 0000020:11 2001001001 0211100040 000011-001 0?010---00 010011010- -??1110210 0-00?10110 10-3---000 02010IOO11 O00I10020? 0010111002 1000001120 0000000000 00000??C0? ?0?00000?0 1100-011?0 000 Spondylus sinensis 0110010?00 0?00020110 20-1000001 0211100040 100011-001 0?5?0---00 OlCO15010- -??I110210 0-10110110 10-3---000 0201010011 0000?0020? 00?0111002 1000001120 0000000000 0??0041001 000000000? ???0-O???? 000 Anomia ephippium 0110010010 0100020101 20?0-00001 0210100040 000011-001 0?010---00 0100?1020? ?O-0110210 0-00110110 10-3---COO 0201010011 000110020? 0010111010 1000001130 0000000000 0??000?001 000000000? ???0-01110 000 Psilunio littoralis 0101101000 0000020000 0001010001 140--00?20 000011-001 00011??-00 0100011201 -110101010 0-00110110 1303---l1l 0101110011 000120020? 0010111000 0-00001??0 0000000000 02201--000 ?000000000 01?0-100-- 100 Lampsilis cardium OIO??????? ??00020000 0001010001 IL0--00?20 000011-001 000ll??-00 0100011201 -110101010 0-00110110 1303---111 0101110011 000120020? 0010111000 0-00001??0 0000000000 02201--000 ?0000000OO 01?0-loo-- 100 Neotrigonia bednalli 0101101000 1000020000 000100000~ 140--00110 000011-001 0?010---00 010001010- -100100010 0-00110110 12-3---111 0101110011 010120020? 0010111000 1000001100 00?0000000 0??01--000 ?00000000? ?????????? 000 Neotrigonia margaritacea 0101101000 1000020000 0001000001 140--00110 000011-001 00010---00 01OCO1010- -100100010 0-00110110 12-3---111 0101110011 010120020? 0010111000 1000001100 00?0000000 00101--000 ?00000000? ?????????? 000 Cardita calyculata 0100000110 1100020000 0001010001 1410100120 000011-001 1?010---00 0100011201 -??0101110 0-00110110 12-3---111 1201110011 0001100211 0010111000 0-00000--0 00?0000000 0220050001 100000101? ???0-10?-? 000 Cardites antiqnata 0100000110 1100020000 0001010001 1410100120 000011-001 1?010---00 0100011201 -??0101110 0-00110110 12-3---111 1201110011 0001100211 0010111000 0-00000--0 00?0000000 022??????? ?????????? ???&lo?-? 000 Astarte castanea 01000001?? ?100020000 0001010001 l40--00120 000011-001 1100100-00 010C011200 00-0101010 0-00110110 12-3---111 1201110011 000120020? 0010111000 0-00000--0 0011100000 0???050001 100000101? ???0-????? 000 Codakia orbiculata 0100003110 1000020000 0001010001 1410200120 000011-001 1?00100000 0100?1123- -??0101210 0-00100111 0303---??? 0201110011 000100020? 0010111000 0-00000--0 00?0010000 001001?000 0000000000 1100-111?? 000 Chama gryphoides 0100000110 1100020100 0001010001 1410200120 000011-001 1?11100000 0100111200 1???101110 0-00110110 1304------?201110011 000110020? 0010111000 1000000--0 00?0000000 0?0001?00? ?0?00????? Galeomma turtoni OIO??????? ??00020000 0001000000 --I0200140 000011-001 0?001??000 0100011200 00-0101210 0-00110110 1??3---??? ?201r10011 100110020? 0010111000 1000000--0 00?0000000 010001?01? ?100000??? ???0-01 Lasaea sp. OIO??????? ??00020000 0001010100 --10100120 000011-001 1?001??000 010C?1122- -??0101 0110110 1??3---111 ?201110011 100110020? 0010111000 0-00000--0 00?0000000 022001?01? 110C00000? ???0-00?-? 000 Dreissena polymorpha 0100000110 0100020001 1001010001 1?10000140 000011-001 1?011O0000 010001?200 00-0101110 0-001:0110 1324------1201110011 000110020? 0010111000 0-00000--0 0000000000 0?0001???? 1????????0 0100-0110? 000 Parvicardium exiguum 0iOOO00110 0100020000 000~010C911410200120 000011-CCl 1?0110100C: 0100 200 0??110111C C-001101:O :314------1201~10011000110020? 0010 003 O-CC000--0 0000C3CCCC OlO??????? ?????????? ???0-1:1?1 COO Fragum unedo 31GOCC0LLO 0100020000 05010:001 lil020C120 000011-C01 :20:10:000 0:0C11120G 00-0:0 0 C-CC113110 1314------a.9 12C11?0011 00CIlCC2~?3C13111000 0-0003C--O 00000~2101C 3?0001?000 000C3CCC%? ???0-?1??? 303 c Tridacna gigas C 1 0C C C C 11 C 0 10002 3C C 1 2 C C1C10 0 0 1 140 -- 0 0 -2 0 0 0 0 0 11 ?112C? ????~OIl~O0-00113111 1314------fi

12 0 111'2 C 11 0 3 0 110 3 2 0 ? C C 1311 13 0 3 100 0 0 0 0 - - 0 0 0 ? 3 0 0 CC222? ??C3-01?l 00'2 P Hippopus hippopus 0100 0 0 0 113 C 100 0 2 0 0 0 0 2 0 0 10I rjC 3 1 140 - - 0 0 12 0 0 0 0 01 : - 0 0 1 12C 11 C 103 0 C 1; C ? 112 C ? ? ? ? ? 1 C 11 10 0 - 0 0 1: 0 111 13 li------2 3 530110020? 00lOlIC33 12011100;: 13C0000--C 3C?0001:?0 000CSl?CC3 ?3CCC300C? ???C-C11?? 000 I"0 e 9 On bivalve phylogeny 313

I I 1 I Id I I I I Id Id

d' m ON mm -id

0000 -1 $1 d -1 -1 ,-I d -1 0000 >-I -1 d -1 $1 -i d -1 \-I >.-I -1 -1 -1 -1 -1 r-I , -1 ,-I I , I 0000 0000000 0000000 00 00 00 00 00 00 00 00 00 00 00 00 0-1 00 Oil 00 00 00 I0 I0I0 I0 10 10 10 10 10 10 10 I0 I0 I0I0 I0 I0I0 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00

0-1 or. OF. 00 0 >I 0 0 0 c. 0 c. 0-1 00 0 c. 0 e. -1 e. -I P. d I -1 P. I i c. -i 0. -11 -1 P. -10 -10 rl c. \ic. -1 d C. N $-I c. P. P. d -1 P. -1 m P. -I P. -1 -1 N 1-1 N c. d e. -1-1 -1-1 do lid -1-1 -1 ri -10 -1 -1 d /-I -Id -id dc. 00 00 Or1 0e. 00 0 0 0-1 0 0 00 0 c. 00 oc. -11 %+I r11 -11 -11 -1 I -11 -1 I 7-1 I -11 -1141 0 0 P. 0 0 0 8.1 0 -10 0 0 00 -1 0 rl 0 00 00 00 I c. e. c. I c. I c. c. I\. I 0 I c. I c. I0 I c. I c. I P. 0-1 P.P. or. 0d c. c. 0 c. 0-1 0 P. 0 c. 0 c. 0 c. 0 c. I -14 00 C.P. c. P. c. 0 c. 00 c. P. r I P. Ti e. 0 c. 0 c. 00 C. c. 1 P. 0 0 00 c. c. xi c. 0 0 (3 0 N c. N P. N 0 -10 , I c. d P. -1 0 -1 0 d P. -10 I I c. ,ic. -1 0 -10dP. 00 P. c. 0 P. /I 0 00 oc. 00 v c. 0 c. 0 0 0 0 0 P. 00 0 e. 0 P. 0 0 00 or. ri 0 -1 e. d C. /I 0 -10 he. 0-1 0 e. 0 c. 0 0 00 0P.

-1 0 0 P. 0 P. ri 0 00 0 P. 00 0 c. 00 0 P. 0 P. 00 0c. 00 0 P. 0 e. 00 00 0 P. 00 0 c. 00 0 c. 0 P. 00 0c. 0 0 0 C. <..< c. 00 00 0 P. -10 c. c. 00 0 P. -i C. s-I 0 i.1 e. I C. I C. 0 c. c. c. c. c. 0 P. P. c. e. e. I P. I e. I c. I c. 0 c. N-1 N C. 0 c. c. w c. -1 0 c. P. c. c. P. -1-1 -1 c. N c. N-1 0c. -10 -1 1.1 C. P. rl 0 -10 d P. -10 -1 c. -10 -1 P. -1 C. -1 0 r-l e. -i c. .-1 0 I P. I I 0 ,-, 0 -I 0 -+ c. \--I c. -10 -10 d c. -10 -1 e. -10 -1 c. 1-1 r. ,--I 0 s-I c. ri r. ,-I 0 ,-I c. ,-4 0 , I 0 00 00 00 0 c. 00 0 -1 01 ON 00 00 00 00 or. 00 00 00 00 00 P. a P. 0 c. 0 P. c. 00 0. c. P. rl C. m c. 0 c. 0 0. 0 P. P. c. c. P. c. c. 0 0. c. c. 0 c. c. -10 -10 -10 -10 -10 d 0 -10 ,+ 0 .-i0 ,-I 0 %-I 0 8-1 0 $-I 0 -10 -10 ria 110 >I0

$4 0 -1 0 d 0 -10 c-I0 ,.-I0 -10 -10 -10 -10 -1c -10 da -10 -1-1 00 00 0 0 00 00 00 00 00 00 00 00 00 00 00 00 00 0 0 0 0 00 00 00 00 00 00 00 00 00 00 0 c3 00 I0 I0 I 0 I0I0 I0 I0I0 I0 I0 I0I0 I010 10 %--I 0 -10 d 0 -10 -i 0 <-.I \-I $1 0 d 0 -1 0 -1 0 -1 0 -10 40 -10 -10 -10 -10 -1 0 -10 1-10 id0 -10 -10 -10 -10 -10 -10 do -10 -1a 00 00 0 0 00 00 00 00 00 00 00 00 00 00 00 00 00 00 0 P. 0 e. or. 0e. 00 00 00 00 or. 0 c. 0 P. 0 c. 0 P. 00 00 0 0 00 00 00 00 00 00 00 00 00 00 00 00 00 00 0 0 00 00 00 00 0c 00 00 00 00 00 00 00

00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 NI NI NINI NI hii NI mi NI mi mi m1-1 *I~INININI ,-I I dldI -11 li I 51 I -1 I -I I -1 I -1 I -1 I dldl-Il-1ldl~lNI 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 40 ii0 810 NO NO 010 I0 I0 I0I0 I0 I0 ION0 NO I0 I0NO 010 NO NO 00 00 00 I0 I0 I0 I0I0 I01000 00 I0I0 00 00 00 00 ri 0 -10 do 00 00 00 00 00 00 00 -10 -10 00 00 do,-I0 ,I0 %I0 *I TI1 41 I d' I * I I xt I C. I xt I e I XH I -0- I -1 IIXldIxtl -10 -10 -10 -10 \I 0 710 -10 -10 -10 -10 -10 -I\+ -10 -10 I003 -10 c.0

rio -10 do -10 \I0 ,do -10 -10 ri0 -10 do do -10 40 00 3-40 k.10

0 c. 0 c. 0 P. 0 0. 0 c. 0 P. 0 c. 0 c. 0 P. 0 P. 0 P. 0 c. 0 P. 0 c. 0 P. 0 00 0 0 00 00 00 0-1 00 00 00 00 00 00 00 00 00 0 ON 0 N ON ON ON ON ON ON ON ON -1N -1N Om -IN om d 00 0 0 00 00 00 00 00 00 00 00 00 00 00 00 00 0 010 01 0 NO NO NO NO NO NO NO NO NO (10 NO

0-1 0-1 ox+ 04 0 0-1 04 0-1 0-1 04 P.4 P.4 0-1 0-1 0-1 0-1 04 0-1 0-1 0 -1 0-1 0 0-1 0d 0-1 0-1 0d P.4 c.4 0-1 0-1 0-1 0-1 0d 0-1 4-1 0 -1 0-1 0 0-1 0d 0-1 0-1 0d P.4 c.4 0-1 0-1 0-1 0-1 0d 0-1 4-1 0 0 00 0 00 00 00 00 00 00 00 00 00 30 oo 00 00 0 c' 0 0 -1 rim ii N -1N -IN dN -1N rim -IN -1N -1N -1N dN -1N rim 401 1-1 0 0-i 0 0-1 0d 0-1 0-i Od 0-1 0-1 0-1 00 30 00 30 00 00 0 0

m 3 14 Giribet & Wheeler

References Atkins D 1936. On the ciliary mechanisms and interrela- tionships of lamellibranchs. Part I. New observations on Ada1 MN & Morton BS 1973. The fine structure of the pal- sorting mechanisms. Quart. J. Microsc. Sci. 79: 181-308. lial eyes of Laternula truncatu (Bivalvia: Anomalodes- 1937a. On the ciliary mechanisms and interrelation- mata: Pandoracea). J. Zool. Lond. 17 1 : 533-556. ships of lamellibranchs. Part 111: Types of lamellibranch Adamkewicz SL, Harasewych MG, Blake J, Saudek D, & gills and their food currents. Quart. J. Microsc. Sci. 79: Bult CJ 1997. A molecular phylogeny of the bivalve 375-421. mollusks. Mol. Bid. Evol. 14: 619-629. 1937b. On the ciliary mechanisms and interrelation- Alatalo P, Berg CJ, & D'Asaro CN 1984. Reproduction and ships of lamellibranchs. Part IV Cuticular fusion, with development in the lucinid special reference to the fourth apcrture in certain (Linne, 1758). Bull. Mar. Sci. 34: 424-434. lamellibranchs. Quart. J. Microsc. Sci. 79: 423-445. Allen JA 1954. On the structure and adaptations of Pandora 1938. On the ciliary mechanisms and interrelation- inaequivalvis and P. pinna. Quart. J. Microsc. Sci. 95: ships of lamellibranchs. Part VII: Latero-frontal cilia of 473-482. the gill filaments and their phylogenetic value. Quart. J. 1958. On the basic form and adaptations to habitat Microsc. Sci. 80: 346-436. in the Lucinacea (Eulamellibranchia). Phil. Trans. R. Soc. Bandel K & Boletzky S 1979. A comparative study of the Lond. B 24 I : 421-484. structure, development, and morphological relationships 1978. Evolution of the deep sea protobranch of chambered cephalopod shells. Veliger 2 1 : 3 13-354. bivalves. Phil. Trans. R. Soc. Lond. B 284: 387-401. Barber VC & Wright DE 1969. The fine structure of the eye 1985. The recent Bivalvia: their form and evolution. and optic tentacles of the mollusc Cardium edule. J. In: The Mollusca, Vol. 10. Evolution. Trueman ER, & Ultrastruct. Res. 26: 515-528. Clarke MR, eds., pp. 337-403. Academic Press, Orlando. Barber VC, Evans EM, & Land MF 1967. The fine structure 1993. A new deep-water hydrothermal species of of the eye of the mollusc, Pecten maximus. Z. Zellforsch. Nuculana (Bivalvia: Protobranchia) from the Guaymas Mikrosk. Anat. 76: 295-312. basin. Malacologia 35: 141-151. Bayne BL 1971. Some morphological changes that occur at 2000. An unusual suctorial montacutid bivalve from the metamorphosis of the larvae of Mytilus edulis. In: the deep Atlantic. J. Mar. Biol. Assoc. UK 80: 827-834. Fourth European Marine Biology Symposium. Crisp DJ, Allen JA & Hannah FJ 1986. A reclassification of the recent ed., pp. 259-280. Cambridge University Press, genera of the subclass Protobranchia (Mollusca: Bivalvia). J. Conchol. 32: 225-249. Cambridge. 1989. Studies on the deep sea Protobranchia: the Beedham GE 1958. Observations on the mantle of the subfamily Ledellinae (Nuculanidae). Bull. Br. Mus. Nat. lamellibranchia. Quart. J. Microsc. Sci. 99: 181-197. Hist. (Zool.) 55: 123-171. Beedham GE & Owen G 1965. The mantle and shell of Allen JA & Sanders HL 1973. Studies on deep-sea Proto- Solemya purkinsoni (Protobranchia: Bivalvia). Proc. Zool. branchia (Bivalvia); the families Siliculidae and Soc. Lond. 145: 405-430. Larnetilidae. Bull. Mus. Comp. Zool. 145: 263-310. Beesley PL, Ross GJB, & Wells A 1998. Moll~l 1982. Studies on the deep-sea Protobranchia (Bival- Southern Synthe CSIRO Publishing, Melbourne. via); the subfamily Spinulinae (family Nuculanidae). Bull. 1-1234 pp. Mus. Comp. Zool. 150: 1-30. Beninger PG & Dufour SC 2000. Evolutionary trajectories 1996. Studies on the deep-sea Protobranchia (Bival- of a redundant feature: lessons from bivalve gill abfrontal via): the family Neilonellidae and the family Nuculanidae. cilia and mucocyte distributions. In: Thc Evolutionary Bi- Bull. Nat. Hist. Mus. Lond. (Zool.) 62: 101-132. ology of the Bivalvia. Harper EM, Taylor JD, & Crame Allen JA, Sanders HL, & Hannah FJ 1995. Studies on the JA, eds., pp. 273-278. The Geological Society of London, deep-sea Protobranchia (Bivalvia); the subfamily Yoldiel- London. linae. Bull. Nat. Hist. Mus. Lond. (Zool.) 61: 11-90, Berg CJ, Alatalo P, Cavanaugh CM, Felbeck H, Jannasch Altaba CR 1992. Les naiades (Mollusca: Bivalvia: ) HW, & Somero GN 1982. Possible chemoautotrophic nu- dels Pai'sos Catalans. Butll. Inst. Cat. Hist. Nat. (Sec. trition in Bahamian bivalves. In: Proceedings of the 2nd Zool.) 60: 23-44. International Conference on Aquaculture Nutrition. Pruder Andrews EB 1988. Excretory systems of molluscs. In: The CD, Langdon C, & Conklin D, eds., Louisiana State Mollusca, Vol. 11. Form and Evolution. Trueman ER & University, Baton Rouge. Clarke MR, eds., pp. 381-448. Academic Press, New Bieler R 1992. Gastropod phylogeny and systematics. Annu. York. Rev. Ecol. Syst. 23: 311-338. Arnold JM 1987. Reproduction and Embryology of Nautilus. Bieler R & Mikkelsen PM 1992. Preliminary phylogenetic In: Nautilus, the Biology and Paleobiology of a Living analysis of the bivalve family Galeommatidae. Amer. Fossil. Saunders WB & Landman NH, eds., pp. 353-372. Malacol. Bull. 9: 157-164. Plenum Press, New York. Boletzky SV 1988. Cephalopod development and cvolution- Arnold JM & Williams-Arnold LD 1978. Spermiogenesis of ary concepts. In: The Mollusca, Vol. 12. Paleontology and Nautilus pompilius. I. General survey. J. Exp. Zool. 20.5: Neontology of Cephalopods. Clarke MR & Trueman ER, 13-26. eds., pp. 185-202. Academic Press, San Diego. On bivalve phylogeny 315

Booth CE & Manguni CP 1978. Oxygen uptake and trans- In: Skeletal Biomineralization: Patterns, Processes and port in the lamellibranch mollusc Modio1u.s demissus. Evolutionary Trends, Vol. 1. Carter JG, ed., pp. 297-41 1. Physiol. Zool. 5 1 : 17-32. Van Nostrand Reinhold, New York. Boss KJ 1982. Mollusca. In: Synopsis and Classification of Carler JG, Campbell DC, & Campbell MR 2000. Cladistic Living Organisms. Parker SP, ed., pp. 1092-1 166. perspectives on early bivalve evolution. In: The Evolu- McGraw-Hill Book Company, New York. tionary Biology of the Bivalvia. Harper EM, Taylor JD, Boss KJ & Turner RD 1980. The giant white clam from the & Crame JA, eds., pp. 47-79. The Geological Society of Galapagos Rift, Calyptogena rnagnifica species novum. London, London. Malacologia 20: 1 6 1 - 194. Casas MT & Subirana JA 1994. Chromatin condensation and Boyd SE 1998. Order Arcoida. In: Mollusca: The Southern acrosome development during spermiogenesis of Ensi,y en- Fauna of Australia, Vol. 5. Beesley PL, Ross sis (Mollusca, Bivalvia). Mol. Reprod. Dev. 37: 223-228. GJB, & Wells A, eds., pp. 253-261. CSIRO Publishing, Cavanaugh CM 1983. Symbiotic chemoautotrophic bacteria Melbourne. in marine invertebrates from sulphide-rich habitats. Nature Bremer K 1988. The limits of amino acid sequence data in 302: 58-61. angiosperm phylogenetic reconstruction. Evolution 42: Charles GH 1966. Sense organs (less Cephalopoda). In: 795-803. Physiology of Mollusca. Wilbur KM & Yonge CM, eds., 1994. Branch support and tree stability. Cladistics pp. 455-521. Academic Press, New York. 10: 295-304. Clasing T 1923. Beitrag zur Kenntnis des Nervensystems Britton JC & Morton BS 1982. A dissection guide, field and und der Sinnesorgane der Mytiliden. Jenaische Z. fiir laboratory manual for the introduced bivalve Corbicula Naturwissenschaften 59: 26 1-3 10. ,Pumineu. Malacol. Rev. Suppl. 3: 1-82. Coan EV, Valentich Scolt P, & Bernard f:R 2000. Bivalve Brooks WK 1875. An organ special sense in the laniel- of Seashells of Western North America. Marine bivalve mol- libranchiate genus Yoldiu. Proceedings Am. Assoc. Ad- lusks from Arctic Alaska to Baja California. Santa Barbara 23rd meeting, Hartford, vancement of Science, Museum of Natural History, Santa Barbara. 764 pp. Connecticut 1874: 79-82. Cole HA 1938. The fate of the larval organs in the meta- Brown CH 1952. Some structural proteins of Mytilus edulis. morphosis of Ostreu edulis. J. Mar. Biol. Assoc. UK 22: Q. J. Microsc. Sci. 93: 487-503. 469-484. Brusca RC & Brusca GJ 1990. Invertebrates. Sinauer Cope JCW 1996. The early evolution of the Bivalvia. In: Associates, Sunderland. 922 pp. Origin and Evolutionary Radiation of thc Molhsca. Taylor Campbell DC 2000. Molecular evidence on the evolution of JD, ed., pp. 361-370. Oxford University Prcss, Oxford. the Bivalvia. In: The Evolutionary Biology of the Bival- 1997. The early phylogeny of the class Bivalvia. via. Harper EM, Taylor, & Crame JA, eds., pp. 31-46. Palaeontology 40: 713-746. The Geological Society of London, London. 2000. A new at early bivalvc phylogeny. In: Campbell DC, Hoekstra KJ, & Carter JG 1998. 18s ribo- look The Evolutionary Biology of thc Bivalvia. Harper EM, somal DNA and evolutionary relationships within the Bi- Taylor JD, & Crame JA, eds., pp. 81-95. The Geological valvia. In: Bivalves: An Eon of Evolution-Palaeobiol- ogical Studies Honoring Norman D. Newell. Johnston PA Society of London, London. & Haggart JW, eds., pp. 75-85. University of Calgary Cox LR 1960. Thoughts on the classification of the Bivalvia. Press, Calgary. Proc. Malac. Soc. Lond. 34: 60-80. Canapa A, Marota I, Rollo F, & Olmo E 1999. The small- 1969. General features of the Bivalvia. In: Treatise subunit rRNA gene sequences of venerids and the on Invertebrate Paleontology, Part I, Mollusca 1. Moore phylogeny of Bivalvia. J. Mol. Evol. 48: 463-468. R, ed., pp. N2-Nl29. University of Kansas Press and Carriker MR 1990. Functional significance of the pediveliger Geological Society of America, Lawrence. in bivalve development. In: The Bivalvia-Proceedings of Cragg SM 1996. The phylogenetic significance of some an- a Memorial Symposium in Honour of Sir Charles Maurice atomical features of bivalve veliger larvae. In: Origin and Yonge, Edinburgh, 1986. Morton BS, ed., pp. 267-282. Evolutionary Radiation of the Mollusca. Taylor JD, ed., Hong Kong University Press, Hong Kong. pp. 37 1-380. Oxford University Press, Oxford. Carter JG 1978. Ecology and evolution of the Gastrochaen- Cragg SM & Nott JA 1977. The ultrastructure of the stato- acea (Mollusca, Bivalvia) with notes on the evolution of cysts in the pediveliger larvae of Pecten mmximus (L.) the endolithic habitat. Bull. Peabody Mus. Nat. Hist., Yale (Bivalvia). J. Exp. Mar. Biol. Ecol. 27: 23-36. Univ. 41: 1-92. Dakin WJ 1910. The visceral ganglion of Pecten, with some 1990a. Evolutionary significance of shell micro- notes on the physiology of the nervous system and an structure in the Palaeotaxodonta, Pteriomorphia, and Iso- inquiry into the innervation of the osphradium in the filibranchia (Bivalvia: Mollusca). In: Skeletal Biominer- Lamellibranchia. Mitt. Zool. Stat. Neapel 20: 1-40. alization: Patterns, Processes, and Evolutionary Trends, 1928. The eyes of Pecten, Spondylus, Amussium, Vol. 1. Carter JG, ed., pp. 135-296. Van Nostrand and allied lamellibranchs, with a short discussion on their Reinhold, New York. evolution. Proc. R. Soc. Lond. B 103: 355-365. 1990b. Shell microstructural data for the Bivalvia. Dan JC & Wada SK 1955. Studies oi' the acrosome. IV. The 316 Giribet & Wheeler

acrosome rcaction in some bivalve spermatozoa. Biol. Cycle and sulphide-oxidizing cnzymes in animals from Bull. 109: 40-55. sulphide-rich habitats. Nature 293: 29 1-293. Dinamani P 1967. Variation in the stomach structure of the Field IA 1922. Biology and economic valuc in sea-. Bivalvia. Malacologia 5: 225-268. Bull. US Bur. Fish. 38: 127-259. Disk1 DL 1998. Evolution of chemoautotrophic endosyn- Folmer 0, Black M, Hoeh W, Lutz R, & Vrijcnhock RC hioses in bivalves. Bivalve-bacteria chemosymbioses are 1994. DNA primers for amplification of mitochondrixl cy- phylogenetically diverse but morphologically similar. tochroine c oxidase subunit I from divcrsc metaman BioScience 48: 277-286. invertebrates. Mol. Mar. Bid. Biotechnol. 3: 294-299. 2000. Phylogenetic relationships among mytilidae Fiirster J 1914. Uber die Leuchtorgane und das Nervensys- (Bivalvia): 18s rRNA data suggest convergence in mytilid tem von Pholas dactylus. Z. Wisscnschaltliche Zool. 109: body plans. Mol. Phylogenet. Evol. 15: 25-33. 349-392. Distel DL, Baco AR, Chuang E, Morrill W, Cavanaugh C, FranzCn A 1955. Comparative morphological investigations & Smith CR 2000. Do mussels take wooden steps to deep- into the spermiogenesis among Mollusca. Zool. Ridr. sea vents? Nature 403: 725-726. Uppsala 30: 399-456. Dohmen MR 1983. Gametogene . In: The Mollusca, Vol. 1983. Ultrastructural studies of spermatoxoa in three 3. Development. Verdonk NH, Van den Biggelaar JAM, bivalve species with notes on evolution or elongated & Tonipa AS, eds., pp. 1-48. Academic Press, New York. sperm nucleus in primitive spermatozoa. Gamete Rcs. 7: Dorange G & Le Pennec M 1989. Ultrastructural character- 199-2 14. istics of spermiogene in Pecten wraxirnus (Mollusca, Freeman G & Lundelius JW 1992. Evolutionary implications Bivalvia). Invertebr. Reprod. Dev. 15: 109-1 17. of the mode of D quadrant specification in coclomates Drew GA 1897. Notes on the embryology, anatomy, and with spiral cleavage. J. Evol. Biol. 5: 205-247. habits of Yoldiu lirnatula, Say. Johns Hopkins Univ. Circ. Freidenfelt T 1897. Das zentrale Nervensystcm von 17: 11-14. Anodonta. Biol. Zbl. 17: 808-8 15. 1 X99a. The anatomy, habits, and embryology of Yol- 1904. Uber den feineren Bau dcs Visceral Ganglions dia lirnatula, Say. Mem. Biol. Lab. Johns Hopkins Univ. von Anodonta. Actd Universitatis Lundensis Sectio I1 4: 1-37. Medica Mathernatica Scientiae Rerum Naturalium 2 40: 1899b. Some observations on the habits, anatomy 1-28. and embryology of members of the Protobranchia. Anat. Fujita T 1929. On the early development of the common Anz. 15: 493-519. Japanese . Jap. J. Zool. 2: 353-358. 1900. Locomotion in Snlenomya and its relatives. Galtsoff PS 1964. The Anierican oyster. Fishery Bullctin or Anat. Anz. 17: 257-266. the Fish and Wildlife Service, United States Department I90 I. The life-history of Nucula delphinodonta of the Interior. 480 pp. (Mighels). Q. J. Microsc. Sci. 44: 3 13-392. Gilmour THJ 1963. A note on the tcntacles of hians 1906. The habits, anatomy, and embryology of the Lima (Gmelin) (Bivalvia). Proc. Malacol. Soc. Lond. 35: 80- giant (Pecten tenuicostatus,Mighels). Univ. Maine 85. Stud. 6: 1-71, Eckelbarger KJ & Davis CV 1996. Ultrastructure of the go- 1967. The defensive adaptations of Lima hians (Mollusca, Bivalvia). J. Mar. Bid. Assoc. UK 47: 209- nad and gametogcnesis in the eastern oyster, Crassostrea virginicu. 2. Testis and spermatogenesis. Mar. Bid. 127: 221. 89-96. 1990. The adaptive significance of foot reversal in Eckelbarger KJ, Bieler R, & Mikkelsen PM 1990. Ultra- the Limoida. In: The Bivalvia-Proceedings of a Memo- structure of sperm development and mature sperm mor- rial Symposium in Honour of Sir Charles Maurice Yonge, phology in three species of commensal bivalves Edinburgh, 1986. Morton BS, ed., pp. 247-263. Hong (Mollusca: Galeommatoidea). J. Morphol. 205: 63-75. Kong University Press, Hong Kong. Edgecombe GD, Giribet G, & Wheeler WC 2002. Phylogeny Giribet G 2001. Exploring the behavior of POY, a program of Henicopidae (Chilopoda: Lithobiomorpha): A com- for direct optimization of molecular data. Cladistics 17: bined analysis of morphology and five molecular loci. S60-S70. Syst. Entomol. 27: 31-64. Giribet G & Carranza S 1999. What can 18s rDNA do for Eernisse DJ & Reynolds PD 1994. Polyplacophora. In: Mi- bivalve phylogeny'? J. Mol. Evol. 48: 256-26 I. croscopic Anatomy of Invertebrates, Vol. 5: Mollusca I. Giribet G & Wheeler WC 200 1. Some unusual small-subunit Harrison FW & Kohn AJ, eds., pp. 55-110. Wiley-Liss, ribosomal RNA sequences of metamans. Am. Museum New York. Novit. 3337: 1-14. Farris JS, Kallersjo M, Kluge AG, & Bult C 1995. Testing Giribet G, Carranza S, BaguiiB J, Riutort M, & Ribera C significance of incongruence. Cladistics 10: 3 15-3 1 9. 1996. First molecular evidence for the existence of a Tar- Felbeck H 1983. Sulfide oxidation and carbon fixation by digrada + Arthropoda clade. Mol. Biol. Evol. 13: 76-84. the gutless clam Solernya reidi: an animal-bacteria Giribet G, Rambla M, Carranza S, BaguiiB J, Riutort M, & symbiosis. J. Comp. Physiol. 152: 3-11. Ribera C 1999. Phylogeny of the arachnid order Opiliones Felbeck H, Childress JJ, & Somero GN 1981. Calvin-Benson (Arthropoda) inferred from a combined approach of com- On bivalve phylogeny 317

plete 18s and partial 28s ribosomal DNA sequences and Solemyidae) as revealed by light and electron microscopy. morphology. Mol. Phylogenet. Evol. 1 1 : 296-307. Mar. Biol. 93: 41 1-427. Giribet G, Edgecombe GD, & Wheeler WC 2001. Arthropod 1988a. Larval and post-larval morphogcnesis in the phylogeny based on eight molecular loci and morphology. gutless protobranch bivalve Solenzycr reidi (Cryptodonta: Nature 413: 157-161. Solemyidae). Mar. Biol. 97: 373-387. Giribet G, Edgecombe GD, Wheeler WC, & Babbitt C 2002. 1988b. Association of bacteria with larvae of the Phylogeny and systematic position of Opiliones: A com- gutless protobranch bivalve SoIwrzyu reidi (Cryptodonta: bined analysis of chelicerate relationships using Solemyidae). Mar. Bid. 97: 389-40 I. morphological and molecular data. Cladistics 1 8: 5-70. Harper EM 1992. Post-larval cementation in the Ostreidae Gladstein DS & Wheeler WC 1997. POY: The Optimization and its implications for other cementing Bivalvia. 1. Moll. of Alignment Characters. American Museum of Natural Stud. 58: 37-48. History, New York. Harper EM, Hide EA, & Morton B 2000a. Relationships Goloboff PA 1998. Nona. Ver. 2.0. Program and documen- between the extant Anomalodesmata: a cladistic test. In: tation available at ftp.unt.edu.ar/pub/parsimony, New The Evolutionary Biology of the Bivalvia. Harper EM, York. Taylor JD, & Crame JA, eds., pp. 129-143. Thc 1999. Analyzing large data sets in reasonable times: Geological Society of London, London. solutions for composite optima. Cladistics 15: 41 5-428. Harper EM, Taylor JD, & Crame JA, eds. 2000b. The Evo- Goloboff PA & Farris JS 200 1. M,ethods for quick consensus lutionary Biology of the Bivalvia. The Geological Society estimation. Cladistics 17: S26-S34. of London, London. 494 pp. Goloboff PA, Farris JS, & Nixon K 2000. TNT: Tree analysis Harry HW 1969. An alternative view on the phylogeny of (1968: using New Technology. Ver. Beta test v. 0.2. Program and the Mollusca. In: Symposium on Mollusca Ernak- ulam, India), Proceedings, Part 1 pp. 170-187. Marine documentation available at ftp.unt.edu.ar/pub/parsimony. Biological Association India, Symposium Series 3. Goodsell JG, Fuller SC, Eversole AG, Castagna M, & Lutz of 1985. Synopsis of the supraspecific classification of RA 1992. Larval and early postlarval shell morphology of living oysters (Bivalvia: Gryphaeidae and Ostreidae). several venerid . J. Mar. Biol. Assoc. UK 72: 231- Veliger 28: 121-158. 255. Harvey EN 1952. Bioluminescence. Academic Press, New Gotting K-J 1980. Origins and relationships of the Mollusca. York. 649 pp. Z. Zool. Syst. Evo1ut.-forsch. 18: 24-27. Haszprunar G 1983. Comparative analysis of the abdominal Gould SJ 1969. The byssus of trigonian clams: phylogenetic sense organs of Pteriomorpha (Bivalvia). J. Moll. Stud. vestige or functional organ? J. Paleontol. 43: I 125-1 129. SUPPI. 12A: 47-SO. Gould SJ & Jones CC 1974. The pallial ridge of Neotuigon- 1985a. The fine morphology of the osphradial sense ia: functional siphons without mantle fusion. Veliger 17: organs of the Mollusca. Part 1 : Gastropoda Prosobranchia. 1-7. Phil. Trans. R. Soc. Lond. B 307: 457-496. Graf DL 2000. The Etherioidea revisited: a phylogenetic 1985b. The fine morphology of the osphradial sense analysis of hyriid relationships (Mollusca: Bivalvia: Pa- organs of the Mollusca. Part 2: Allogastropoda (Architec- leoheterodonta: Unionoida). Occ. Pap. Mus. Zool. Univ. tonicidae and Pyramidellidae). Phil. Trans. R. Soc. Lond. Mich. 729: 1-21. B 307: 497-505. Graf DL & 0 Foighil D 2000. The evolution of brooding 1985~.On the anatomy and fine-structure of a pe- characters among the freshwater pearly mussels (Bivalvia: culiar sense organ in Nucula (Bivalvia, Protobranchia). Unionoidea) of North America. J. Moll. Stud. 66: 157- Veliger 28: 52-62. 170. 1985d. The fine structure of the abdominal sense or- Griffond B 1980. Etude ultrastructurale de la spermatogb- gans of Pteriomorpha (Mollusca, Bivalvia). J. Moll. Stud. nese typique de Viviparus viviparus L., Mollusque 51: 315-319. GastCropode. Arch. Bid. 91: 445-462. 1987. The fine morphology of the osphradial sense Grobben C 1886. Zur Kenntnis und Morphologie und der organs of the Mollusca. 111. Placophora and Bivalvia. Phil. Verwandtschaftsverhaltnisse der Cephalopoda. J. Nat. Trans. R. Soc. Lond. B 315: 37-61. Hist. 33: 267-287. 1988. On the origin and evolution of major gastro- Gros 0, Frenkiel L, & Moueza M 1997. Embryonic, larval, pod groups, with special reference to the Streptoneura and post-larval development in the symbiotic clam Co- (Mollusca). J. Moll. Stud. 54: 367-441. dukia orbicularis (Bivalvia: Lucinidae). Invertebr. Biol. 2000. Is the Aplacophora monophyletic? A cladistic 116: 86-101. point of view. Amer. Malacol. Bull. 15: 115-130. Gustafson RG & Lutz RA 1992. Larval and early post-larval Healy JM 1989. Spermiogenesis and spermatoma in the rel- development of the protobranch bivalve Solemya velum ict bivalve genus Neotrigonin: relevance to trigonioid re- (Mollusca: Bivalvia). J. Mar. Biol. Assoc. UK 72: 383- lationships, particularly with Unionoidea. Mar. Biol. 103: 402. 75-85. Gustafson RG & Reid RGB 1986. Development of the per- 1990. Spermatozoa of the scissurellid gastropod Si- icalymma larva of Solemya midi (Bivalvia: Cryptodonta: nezona sp, (Prosobranchia, Pleurotomarioidea) with a dis- 318 Giribet &L Wheeler

cussion of pleurotomarioid and fissurelloid sperm sperm and spermiogenesis of three species of Mytilidae morphology. Zool. Scripta 19: 189-193. (Mollusca, Bivalvia). Gamete Res. 15: 123-1 35. 199%. Comparative spermatozoa1 ultrastructure and Hodgson AN & Chia F-S 1993. Spermatozoon structure of its taxonomic and phylogenetic significance in the bivalve some North American prosobranchs from the families Lol- order Veneroida. In: Advances in Spermatozoa1 Phylogeny tiidae (Patellogastropoda) and Fissurellidae (Archaeogas- and Taxonomy. Jamieson BG, Ausi6 J, &Justine J-L, eds., tropoda). Mar. Biol. 1 16: 97-1 0 I. pp. 155-1 66. Editions du MusCum Paris, Paris. Hodgson AN & Foster GG 1992. Structure of the sperm of 1995b. Sperm ultrastructure in the marine bivalve some South African archaeogastropods (Mollusca) from families Carditidae and Crassatellidae and its bearing on the superfamilies Haliotoidea, Fissurelloidea, and unification of the Crassatelloidea with the Carditoidea. Trochoidea. Mar. Biol. 113: 89-97. Zool. Scripta 24: 21-28. Hodgson AN, Baxter JM, Sturrock MG, & Bernard RTF 19962. Molluscan sperm ultrastructure: correlation 1988. Comparative spermatology of 11 species of Poly- with taxonomic units within Gastropoda, Cephalopoda and placophora (Mollusca) from the suborders Lepidopleurina, Bivalvia. In: Origin and Evolutionary Radiation of the Chitonina, and Acanthochitonina. Proc. R. Soc. Lond. B Mollusca. Taylor JD, ed., pp. 99-1 13. Oxford University 235: 161-177. Press, Oxford. Hodgson CA & Burke RD 1988. Development and larval I996b. Spermatozoan ultrastructure in the trigonioid morphology of the spiny scallop, Ch1umy.s hastutu. Biol. bivalve Nentrigoniu margaritacea Lamarck (Mollusca): Bull. 174: 303-3 18. comparison with other bivalves, especially Trigonioida Hoeh WR, Black MB, Gustafson RG, Bogan AE, Lutz KA, and Unionoida. Helgolander Meeresuntersuchungen 50: & Vrijenhoek RC 1998. Testing alternative hypotheses of 259-264. Neotrigoniu (Bivalvia: Trigonioida) phylogcnctic relation- ships using cytochrome c oxidase subunit I DNA Healy JM, Beaines KP, & Barclay DB 1998. Spermatozoa sequences. Malacologia 40: 267-278. of the Australian 'greenlip' abalone Haliotis laevigata Hoggarth MA 1999. Description of some of the glochidia of Donovan: ultrastructure and comparison with other gastro- the Unionidae (Mollusca: Bivalvia). Malacologia 41 : pods, especially other Haliotidae (Vetigastropoda, 1-1 18. Mollusca). Invertebr. Reprod. Dev. 34: 197-206. Hou ST & Maxwell WL 1991. Ultrastructural studies of Healy JM, Keys JL, & Daddow LYM 2000. Comparative spermatogenesis in Anla1i.s entalis (Scaphopoda, sperm ultrastructure in pteriomorphian bivalves with spe- Mollusca). Phil. Trans. R. Soc. Lond. B 333: 101-1 10. cial reference to phylogenetic and taxonomic implications. Hylander BL & Summers RG 1977. An ultrastructural anal- In: The Evolutionary Biology of the Bivalvia. Harper EM, ysis of the gametes and early fertilization in two bivalvc Taylor JD, & Crame JA, eds., pp. 169-190. The molluscs, Chama rnacerophylla and Spisula solidissimu Geological Society of London, London. with special reference to gamete binding. Cell Tissue Kes. Healy JM & Willan RC 199 I. Nudibranch spermatozoa: 182: 469-489. comparative ultrastructure and systematic importance. Iredale T 1939. Mollusca, Part I. Scientific Reports of thc Veliger 34: 134-165. Great Barrier Reef Expedition 1928-1 929 5: 209-425. Heath H 1899. The development of Zschnochilon. Zoologis- Johnston PA & Haggart JW 1998. Bivalves: An Eon of che Jahrbucher, Abteilung fur Anatomie und Ontogenie Evolution. University 01' Calgary Press, Calgary. 461 pp. der Tiere 12: 567-656. Kabat AR &L 0 Foighil D 1987. Phylum Mollusca, Class Hendriks L, van Broeckhoven C, Vandenberghe A, van de Bivalvia. In: Reproduction and Developnicnl of Marine Peer Y, & De Wachter R 1988. Primary and secondary [nvertebrates of the Northern Pacific Coast. Strathmann structure of the 18s ribosomal RNA of the bird spider MF, ed., pp. 300-353. University of Washington Press, Eurypelmu californicu and evolutionary relationships Seattle. among eukaryotic phyla. Eur. J. Biochem. 177: 15-20. Kasyanov VL, Kryuchkova GA, Kulikova VA, & Medved- Hennig W 1979. Taschenbuch der Speziellen Zoologie. Teil eva LA 1998. Larvae of Marine Bivalves and 1, Wirbellose I. Verlag Harri Deutsch, Zurich. 392 pp. Echinoderms. Science Publishers, Enficld. 288 pp. Herbers K 19 14. Entwicklungesgeschichte von Anodonta Kawaguti S 1950. Observations of the heart shell, Corculum cellensis SchrGter. Z. Wiss. Zool. 108: 1-174. cardissu (L.) and its associated zooxanthellae. Pac. Sci. 4: Herskovits TT, Hamilton MG, Cousins CJ, & Wall JS 1990. 43-49. Light-scattering and scanning transmission electron micro- Kellog JL 1915. Ciliary mechanisms of lamellibranchs with scopic investigation of the hemocyanin of the bivalve, Yol- descriptions of anatomy. J. Morphol. 26: 625-70 I. dia limatula (Say). Comp. Biochem. Physiol. B 96: 497- Kessel E 1944. uber Periostracum-bildung. Z. Morphol. 503. Okol. Tiere 40: 348-360. Hicks GRF & Marshall BA 1985. Sex selective predation of Keys JL & Healy JM 1999. Sperm ultrastructure of the giant deep-sea meiobenthic copepods by pectinacean bivalves clam Tridacna maxima (Tridacnidae: Bivalvia: Mollusca) and its influence in copepods sex ratios. NZ J. Mar. from the Great Barrier Reef. Mar. Bid. 135: 41-46. Freshw. Res. 19: 227-231. 2000. Relevance of sperm ultrastructure to the clas- Hodgson AN & Bernard RTF 1986. Ultrastructure of the cation of giant clams (Mollusca, Cardioidea, Cardiidae, On bivalve phylogeny 319

Tridacninae). In: The Evolutionary Biology of the Bival- Lutz RA, Mann R, Goodsell JG, & Castagna M I982a. Lar- via. Harper EM, Taylor JD, & Crame JA, eds., pp. 191- val and early post-larval development of Arctica is1andic.u. 205. The Geological Society of London, London. J. Mar. Bid. Assoc. UK 62: 745-769. Kraemcr LR 1979. Suprabranchial and branchial shelves of Lutz RA, Goodsell JG, Castagna M, Chapman S, Newcll C, bivalved molluscs: structural functional context of visceral Hidu H, Mann R, Jablonski D, Kennedy V, Siddall S, et ganglion, osphradium, and branchial nerves. Am. Zool. al. 1982b. Preliminary observations on the usefulness of 19: 959. hinge structures for identification of bivalve larvae. J. 198 1. The osphradial complex of two freshwater bi- Shellfish Res. 2: 65-70. valves: histological evaluation and functional context. Lynn JW 1994. The ultrastructure of the sperm and motile Malacologia 20: 205-2 16. sperinatozeugmata released from thc freshwater mussel 1983. Onlogenetic aspects of biflagellate sperm in Anodoizta grandis (Moll usca, B i val via, U nionidae). Can. Corbiculu ,fluminea (Muller) (Bivalvia: Sphaeriacea). J. Zool. 72: 1452-1461. Trans. Am. Microsc. Soc. 102: 88. Mackie GL 1984. Bivalves. In: The Mollusca, Vol. 7. Re- Kubo M 1977. The formation of ii temporary-acrosome in production. Tompa AS, Vcrdonk NH, & Van den Rigge- the spermatozoon of Laternulu limicolu (Bi valvia, laar JAM, eds., pp. 35 1-41 8. Acadcmic Press, Orlando. M~IILISCCL).J. Ultrastruct. Res. 61: 140-148. Maddison DR & Maddison WP 2000. MacCladc 4: Intcr- 1979. Ultrastructural studies on early events in fer- active Analysis of Phylogeny and Character Evolution. tilization of the bivalve Latrrrzulu lirnicolu. Protoplasma ver. 4.01. Sinaucr Associates, Sunderland. 100: 73-83. Mangum CP, Miller KI, Scott JL, Van Holdc KE, & Morsc Kubo M & Ishikawa M 1978. Organizing process of the MP 1987. Bivalve hemocyanin: structural, functional, and tempora~y-acrosomein spermatogenesis of the bivalve phylogenetic relationships. Bid. Bull. 173: 205-221. Lyonsia ventricosa. J. Submicrosc. Cytol. 10: 41 1-421. Maxwell PA 1988. Comments on “a reclassification of the Lacaze-Duthiers FJH 1856. Historie de I’organisation et du Protobranchia (Mollusca: Bivalvia)” by J. A. Allen and E dCveloppement du Dentale. Ann. Sci. Nat. (Zool.) 6: 225- J. Hannah (I 986). J. Conchol. 33: 85-96. 281. Maxwell WL 1975. Spermiogenesis of Eusepiu officinalis 1857a. Histoire de l’organisation et du dkveloppe- (L.), Loligo ,foubesi (Steenstrup), and Allotcuthis subulutu inent du Dentale. Annales des Sciences Naturelles 4, (L.) (Cephalopoda, Decapoda). Proc. R. Soc. Lond. B I9 I : Zoologie 7: 194-254. 527-535. 1857b. Voyage aux Iles Balkares OLI Recherches sur Meisenheimer J 190 1. Entwicklungsgeschichte von L‘anatomie et la Physiologie de Quelques Mollusques de Dveissensirx polynwrphu. Z. Wiss. Zool. 69: 1 - 1.17. la MCditerranCe. Victor Masson, Paris. 224 pp. Meyer HA & Mobius KA 1872. Die Prosobranchia und La- 1858. Histoire de L‘organisation, du Dkveloppemenf mellibranchia, Nebst Einem Supplement zu den Opistho- des Moeurs et des Rapports Zoologiques du Dentale. branchia. Verlag von Wilhelm Engelmann, Leipzig. Victor Masson, Paris. 287 pp. Mickevich MF & Farris JS 1981. Thc implications of Lambert 0, Taveau JC, Boisset N, & Lamy JN 1995. Three- congruence in Menidiu. Syst. Zool. 27: 143-158. dimensional reconstruction of the hemocyanin of the pro- Mikkelsen PM & Bieler R 2001. Vuricorbulu (Bivalvia: Cor- tobranch bivalve mollusc Nucula hanleyi from frozen- bulidae) of the western Atlantic: taxonotny, anatomy, lii’e hydrated specimens. Arch. Biochem. Biophys. 3 19: 23 I - habits, and distribution. Veliger 44: 27 1-293. 243. Moir AJG 1977. On the ultraslructurc of thc abdominal Lauterbach K-E 1983. Erbrterungen zur Stammesgeschichte sense organ of‘ the giant scallop, Plucopectun mugellurzic~us der Mdl~isca,insbesondere der Conchifera. Z. Zool. Syst. (Gmelin). Cell Tissue Res. 184: 359-366. Evo1ut.-forsch. 21 : 201-216. Morse ES 19 13. Observations on living Solenornj~a (velum 1984. Das phylogenetische System der Mollusca. and borealis). Biol. Bull. 24: 261-281. Mitteilungen der Deutschen Malakologischen Gesellschaft Morse MP & Zardus JD 1997. Bivalvia. In: Microscopic 37: 66-81. Anatomy of Invertebrates, Vol. 6A: Mollusca 11. Harrison Le Pennec M 1980. The larval and post-larval hinge of some FW & Kohn AJ, eds., pp. 7-1 18. Wiley-Liss, New York. families of bivalve molluscs. J. Mar. Biol. Assoc. UK 60: Morse MP, Meyhofer E, Otto JJ, & Kuzirian AM 1986. He- 601 -6 1 7. mocyanin respiratory pigment in bivalve mollusks. Lebour MV 1938. Notes on the breeding of some lanielli- Science 23 I : 1302-1 304. branchs from Plymouth and their larvae. J. Mar. Biol. Morton BS 1969. Studies on the biology of I)rc)i.s.c.enapo- Assoc. UK 23: 119-144. lymorpha Pall. I. General anatomy and morphology. Proc. Lillie FR 1895. The embryology of the Unionidae. A study Malac. Soc. Lond. 38: 301-321. in cell-lineage. J. Morphol. 10: 1-100. 1977. The hypobranchial gland in the Bivalvia. Can. List T 1902. Die Mytiliden des Golfes von Neapel. Fauna J. Zool. 55: 1225-1234. und Flora des Golfes von Neapel 27: 1-3 12. 1978. The biology and functional morphology of Lutz RA 1985. Identification of bivalve larvae and postlar- Philobrya minutu (Bivalvia: ) vae: A review of recent advances. Amer. Malacol. Bull. 185: 173-196. Special Edition No. 1: 59-78. 1980a. Some aspects of the biology and functional 320 Giribet & Wheeler

morphology (including the presence of a ligamental lith- 1923. Die Cephalopoden, Systematik. Fauna und odesma) of Montacutona cornpactu and M. olivacea (Bi- Flora des Golfes von Neapel: 863. valvia: Leptonacea) associated with coelenterates in Hong Narchi W 1968. The functional morphology of Lyonsiu cul- Kong. J. Zool. Lond. 192: 431-455. fornica Conrad, 1837 (Bivalvia). Veliger 10: 305-3 13. 19XOb. Swimming in Amusium pleuronectes Newell NU 1965. Classification of thc Bivalvia. Amer. (Bivalvia: Pectinidae). J. Zool. Lond. 190: 375-404. Museum Novit. 2206: 1-25, 198 I. The Anomalodesmata. Malacologia 2 1 : 35- 1969. Classification of Bivalvia. In: Treatise on In- 60. vertebrate Paleontology, Part N, Mollusca 6, Vol. I. Rival- 1982a. The functional morphology of Pavilirnya ,fva- via. Moore R, ed., pp. N205-N244. Geological Socicty of gilis (Bivalvia: Parilimyidae nov. ,f: Amer.) with a discus- America & University of Kansas, Boulder-Lawrcncc. sion on the origin and evolution of the carnivorous sep- Nicotra A & Zappata S 1991. Ultrastructure of the mature tibranchs and a reclassification of the Anomalodesrnata. sperm and speriniogenesis in Callista chiom (Mollusca, Trans. R. Soc. Lond. 36: 153-216. Bivalvia). Invertebr. Reprod. Dev. 20: 2 13-21 8. I 982b. The biology, functional morphology, and tax- Norton JH & Jones GW 1992. The : an Anatom- onomic status of Fluviolanatus suhtorta (Bivalvia: Tra- ical and Histological Atlas. Australian Centre for peziidae), a heteromyarian bivalve possessing Agricultural Research, Canberra. ‘zooxanthellae’. J. Malacol. Soc. Aust. 5: I 13-1 40. Nuttall CP 1969. Coloration. In: Treatise on Invertebrate Pa- 1983. Feeding and digestion in Bivalvia. In: The leontology, Part N. Mollusca 6, Vol. l. Bivalvia. Moore Mollusca, Vol. 5. Pt. 2. Saleuddin ASM & Wilbur KM, RC, ed., pp. N70-N72. Geological Society of Amcrica & eds., pp. 65-147. Academic Press, New York. University of Kansas, Boulder-Lawrcnce. 1985. Adaptive radiation in the Anomalodesmata. 0 Foighil D I Y85a. Fine structure of Lasneo .subvividis and In: The Mollusca, Vol. 10. Evolution. Trueman ER & Mysella tumida sperm (Bivalvia, Galeommatacea). Zoo- Clarke MR, eds., pp. 405-459. Academic Press, Orlando. morphology 105: 125-132. vives- I987a. The pallial photophores of Barhatia 1985b. Sperm transfer and storage in thc brooding cens (Bivalvia: Arcacea). J. Moll. Stud. 53: 241-243. bivalve Mysella tumidu. Bid. Bull. 169: 602-6 14. I987b. Siphon structure and prey capture as a guide 0 Foighil D & Graf DL 2000. Prodissoconch morphology to affinities in the abyssal septibranch Anomalodesmata. of the relict marine paleoheterodont Neotrigonia margar- Sarsia 72: 49-69. itacea (Mollusca: Bivalvia) indicates a non-planktotrophic 1987~.The functional morphology of Neotrigonia prejuvenile ontogeny. .I. Mar. Bid. Assoc. UK 80: 173- margaritacca (Bivalvia: Trigonacea), with a discussion of 174. phylogenetic affinities. Rec. Austr. Mus. 39: 339-354. 0 Foighil D & Taylor DJ 2000. Evolution of parcntal care 1996. The cvolutionary history of the Bivalvia. In: and ovulation behavior in oysters. Mol. Phylogenet. Evol. Origin and Evolutionary Radiation of the Mollusca. Taylor JD, ed., pp. 337-359. Oxford University Press, Oxford. 15: 301-313. 2000a. The biology and functional morphology of Ohno T, Katoh T, & Yamasu T 1995. The origin of algal- Fmgum evugatum (Bivalvia: Cardiidae) from Shark Bay, bivalve photo-symbiosis. Palaeontology 38: 1-2 1. Western Australia: the significance of its relationship with Owen G 1959. The ligament and digestive systcm in the entrained xooxanthellae. J. Zool. 25 1 : 39-52. taxodont bivalves. Proc. Malac. Soc. Lond. 33: 2 15-223. 2000b. The pallial eyes of Cfenoides floridanus 1966. Feeding. In: Physiology of Mollusca, Vol. TI. (Bivalvia: Limoidea). J. Moll. Stud. 66: 449-455. Wilbur KM & Yonge CM, eds., pp. 1-5 I. Academic Press, 2000c. The function of pallial eyes within the Pec- New York. tinidae, with a description of those present in Patinopeeten 1978. Classification and the bivalve gill. Phil. Trans. yessoensis. In: The Evolutionary Biology of the Bivalvia. R. Soc. Lond. B 284: 377-385. Harper EM, Taylor JD, & Cranie JA, eds., pp. 247-255. Page RDM 2000. NDE. ver. 0.4.8. Program available at The Geological Society of London, London. http://taxonomy .zoology.gla.ac.uk/rod/NDE/nde.html. Morton BS, Prezant RS, & Wilson B 1998. Class Bivalvia. Park JK & 0 Foighil D 2000. Sphaeriid and corbiculid clams In: Mollusca: The Southern Synthesis. Fauna of Australia, represent separate heterodont bivalve radiations into fresh- Vol. 5. Beesley PL, Ross GJB, & Wells A, eds., pp. 195- water environments. Mol. Phylogenet. Evol. 14: 75-88. 234. CSIRO Publishing, Melbourne. Pathy DA & Mackie GL 1993. Comparative shell morphol- Morton JE 1953. The functions of the gastropod stomach. ogy of Dreissena polymorpha, leucophaeata, Proc. Linnean Soc. Lond. 164: 240-246. and the “quagga” mussel (Bivalvia: Dreissenidae) in Morton JE & Yonge CM 1964. Classification and structure North America. Can. J. Zool. 71: 1012-1023. of the Mollusca. In: Physiology of Mollusca. Wilbur KM Pearse JS 1979. Polyplacophora. In: Reproduction of Marine & Yonge CM, eds., pp. 1-58. Academic Press, New York. Invertebrates, Vol. V. Molluscs: Pelecypods and Lesser Naef A 1913. Studien zur generellen Morphologie der Mol- Classes. Giese AC & Pearse JS, eds., pp. 27-85. lusken. l. Teil: Uber Torsion und Asyinmetrie der Gastro- Academic Press, New York. poden. Ergebnisse und Fortschritte der Zoologie 3: 73- Pelseneer P 189 1. Contribution a 1’Ctude des larnellibranch- 164. es. Arch. Biol. 9: 147-312. On bivalve phylogeny 32 1

1899. Les yeux branchiaux des lamellibranches. 1958. The stomach in the Eulamellibranchia; stom- Arch. Biol. 16: 97-103. ach types IV and V. Proc. Zool. Soc. Lond. 13 1 : 487-525. 1906. Mollusca. In: A Treatise on Zoology. 1960. Phylogeny in the Lamellibranchia. In: Pro- Lankester ER, ed., A. and C. Black, London. 355 pp. ceedings of the Centenary and Bicentenary Congress of 191 I. Les lamellibranchs de I’expCdition du Siboga. Biology, Singapore, Dec. 1958. Purchon RD, ed., pp. 69- Partie anatomique: 1. Siboga-Expeditie. Monograph 53: 82. University of Malaya Press, Singapore. 1-126. 1963. Phylogenetic classification of the Bivalvia, Peredo S, Garrido 0, & Parada E 1990. Spermiogenesis and with special reference to the Septi branchia. Proc. Malacol. sperm ultrastructure in the freshwater mussel Diplodon Soc. Lond. 3.5: 71-80. chilensis chilensis (Mollusca: Bivalvia). Invertebr. Reprod. 1968. The Biology of the Mollusca. Pcrgamon Press, Dev. 17: 171-179. Oxford. 560 pp. Pojeta J Jr. 1978. The origin and early taxonomic diversifi- 1978. An analytical approach to a classification of cation of pelecypods. Phil. Trans. R. Soc. Lond. B 284: the Bivalvia. Phil. Trans. R. Soc. Lond. B 284: 425-436. 225-243. 1987a. The stomach in the Bivalvia. Phil. Trans. R. Pojeta J Jr. & Runnegar B 1976. The paleontology of ros- Soc. Lond. B 3 16: 183-276. troconch mollusks and the early history of the phylum 1987b. Classification and evolution of the Bivalvia: Mollusca. Geological Survey of Professional Papers (US.) an analytical study. Phil. Trans. R. Soc. Lond. B 3 16: 277- 968: 1-88. 302. 1985. The early evolution of diasome molluscs. In: Raven CP 1966. Morphogenesis. The Analysis of Molluscan The Mollusca, Vol. 10, Evolution. Trueman ER & Clarke Development. Pergamon Press, Oxford pp. MR, eds., pp. 295-336. Academic Press, New York. Reid RGB 1965. The structure and function of the stomach Pojeta J Jr., Runnegar B, Peel JS, & Gordon Jr. M 1987. in bivalve molluscs. J. Zool. 147: 156-184. Phylum Mollusca. In: Fossil Invertebrates. Boardman RS, 1980. Aspects of the biology of a gutless species of Cheetham AH, & Rowel1 AJ, eds., pp. 270-435. Solemyu (Bivalvia: Protobranchia). Can. J. Zool. 58: 386- Blackwell Scientific Publications, Oxford. 393. Ponder WF & Lindberg DR 1997. Towards a phylogeny of 1990. Evolutionary implications of sulphide- gastropod molluscs: an analysis using morphological oxidizing symbioses in bivalves. In: The Bivalvia-Pro- characters. Zool. J. Linnean Soc. 119: 83-265. ceedings of a Memorial Symposium in Honour of Sir Popham JD 1979. Comparative spermatozoon morphology Charles Maurice Yonge, Edinburgh, 1986. Morton BS, ed., and bivalve phylogeny. Malacol. Rev. 12: 1-20. pp. 127-140. Hong Kong University Press, Hong Kong. Popham JD & Dickson MR 1973. Bacterial associations in 1998. Class: Bivalvia. In: Mollusca: The Southern the Bankiu uusti-alis (Lamellibranchia: Mollusca). Synthesis. Fauiia of Australia, Vol. 5. Beesley PL, Ross Mar. Biol. 19: 338-340. GJB, & Wells A, eds., pp. 19.5-234. CSIRO Publishing, Popham JD & Marshall BA 1977. The fine structure of the Melbourne. spermatozoon of the protobranch bivalve, Nucula Reid RGB & Brand DG 1986. Sulfide-oxidizing symbiosis hurtvigiana Pfeiffer. Veliger 19: 43 1-433. in Lucinaceans: implications for bivalve evolution. Veliger Popham JD, Dickson MR, & Goddard CK 1974. Ultrastruc- 29: 3-24, tural study of the mature gametes of two species of Bankiu Reindl S & Haszprunar G 1996. Shell porcs (caeca, aes- (Mollusca: Teredinidae). Aust. J. Zool. 22: 1-12. thetes) of Mollusca: a case of polyphyly. In: Origin and Popham ML 1940. The mantle cavity of some of the Ery- Evolutionary Radiation of the Mollusca. Taylor JD, ed., cinidae, Montacutidae and Galeommatidae with special pp. 115-1 18. Oxford University Press, Oxford. reference to the ciliary mechanisms. J. Mar. Biol. Assoc. Reunov AA & Hodgson AN 1994. Ultrastructure of the sper- UK 24: 549-587. matozoa of five species of South African bivalves (Mol- Prezant RS 1990. Microstructural shell and hinge ontogeny lusca), and an examination of early spermatogenesis. .J. of the Antarctic bivalve Lissarcu notorcadensis Melvill Morphol. 219: 275-283. and Standen. J. Mar. Biol. Assoc. UK 70: 841-855. Reynolds PD 1990. Functional morphology of thc perianal 1998. Subclass Anomalodesmata Introduction. In: sinus and pericardium of Dentaliurn. rectius (Mollu Mollusca: The Southern Synthesis. Fauna of Australia, Scaphopoda) with a reinterpretation of the scaphopod Vol. 5. Beesley PL, Ross GJB, & Wells A, eds., pp. 397- heart. Amer. Malacol. Bull. 7: 137-146. 40.5. CSIRO Publishing, Melbourne. Reynolds PD & Okusu A 1999. Phylogenetic relationships Prezant RS & Carriker MR 1983. Functional microstructure among families of the Scaphopoda (Mollusca). Zool. J. of the lithodesma of Mytilimeria nuttalli (Bivalvia: Linn. Soc. 126: 131-153. Anomalodesuiata). Veliger 2.5: 326-328. Ridewood WG 1903. On the structure ol‘ the gills of lamel- Purchon RD 1956. The stomach in the Protobranchia and libranchs. Phil. Trans. R. Soc. Lond. B 194: 147-284. Septibranchia (Lamellibranchia). Proc. Zool. Soc. Lond. Rocha E & Azevedo C 1990. Ultrastructural study of the 127: 5 1 1-525. spermatogenesis of Anodontu cygnea L. (Bivalvia, 1957. The stomach in the Filibranchia and Pseudo- Unionidae). Invertebr. Reprod. Dev. 18: 169- 176. lamellibranchia. Proc. Zool. Soc. Lond. 129: 27-60. Rosen MD, Stasek CR, & Hermans CO 1978. The ultrastruc- 322 Giribet & Wheeler

ture and evolutionary significance of the cerebral ocelli of in the subfamilies Clinocardiinae, Lymnocardiinae, Mytilus edulis, the Bay Mussel. Veliger 21: 10-18. Fraginae, and Tridacninae. Malacologia 40: 32 1-373. Runnegar B 1974. Evolutionary history of the bivalve Schneider JA & 0 Foighil D 1999. Phylogeny of giant clams subclass Anomalodesmata. J. Paleontol. 48: 904-939. (Cardiidae: Tridacninae) based on partial mitochondria1 1978. Origin and evolution of the Class Rostrocon- 16s rDNA gene sequences. Mol. Phylogenet. Evol. 13: chia. Phil. Trans. R. Soc. Lond. B 284: 319-330. 59-66. 1996. Early evolution of the Mollusca: the fossil re- Seydel E 1909. Untersuchen uber den byssusapparat der La- cord. In: Origin and Evolutionary Radiation of the Mol- mellibranchiaten. Zoologische Jahrbucher, Abteilung fur lusca. Taylor J, ed., pp. 77-87. Oxford University Press, Anatomie und Ontogenie der Tierc 27: 465-582. London. Shimek RL 1988. The functional morphology of scaphopod Runnegar B & Pojeta J Jr. 1974. Molluscan phylogeny: the captacula. Veliger 30: 213-221. paleontological viewpoint. Science 186: 3 1 1-3 17. Slack-Smith SM 199821. Order Ostreoida. In: Mollusca: The Saleuddin ASM 1965. The mode of life and functional anat- Southern Synthesis. Fauna of Australia, Vol. 5. Beesley omy of Astarte spp. (Eulamellibranchia). Proc. Malacol. PL, Ross GJB, & Wells A, eds., pp. 268-282. CSIRO Soc. Lond. 36: 229-257. Publishing, Melbourne. 1967. Notes on the functional anatomy of three I998b. Superfamily Carditoidea. In: Mollu. North American species of Astarte, A. undutu, A. castu- Southern Synthesis. Fauna of Australia. Vol. 5. Becslcy nu, and A. esquilmati Baird. Proc. Malac. Soc. Lond. 37: PL, Ross GJB, & Wells A, eds., pp. 322-325. CSIRO 38 1-385. Publishing, Melbourne. Salvini-Plawen LV 1980. A reconsideration of systematics Smith SW, Overbeek R, Woese CR, Gilbert W, & Gillevet in the Mollusca. Malacologia 19: 249-278. PM 1994. The Genetic Data Environment: an expandable 1985. Early evolution and the primitive groups. In: GUI for multiple sequence analysis. Comput. Appl. The Mollusca. Evolution. Trueman ER & Clarke MR, eds., Biosci. 10: 67 1-675. pp. 59-150. Academic Press, Orlando. Sokolov VA & Zaitseva OV 1982. Osphradial chcmorecep- 1988. The structure and function of molluscan di- tors in lamellibranch mollusks, Unio pictorum and gestive systems. In: The Mollusca, Vol. ll, Form and Anodonta cygneu. J. Evol. Biochem. Physiol. 18: 56-61. Function. Trueman ER & Clarke MR, eds., pp. 301-379. Sousa M & Azevedo C 1988. Comparative silver-staining Academic Press, New York. analysis on spermatozoa of various invertebrate species. Salvini-Plawen LV & Steiner G 1996. Synapomorphies and Int. J. Invertebr. Reprod. Dev. 13: 1-8. plesiomorphies in higher classification of Mollusca. In: Sousa M & Oliveira E 1994. An ultrastructural study of sper- Origin and Evolutionary Radiation of the Mollusca. Taylor matogenesis in Crussostrea ungulata (Mollusca, JD, ed., pp. 29-51. Oxford University Press, Oxford. Bivalvia). Mar. Biol. 120: 545-55 I. Sanders HL & Allen JA 1973. Studies on deep-sea Proto- Sousa M, Oliveira E, Carvalheiro J, & Oliveira V 1995. branchia (Bivalvia); prologue and the Pristiglomidae. Bull. Comparative silver staining of molluscan spermatozoa. In: Mus. Comp. Zool. 145: 237-262. Advances in spermatozoa1 phylogeny and taxonomy. Ja- Sastry AN 1979. Pelecypoda (excluding Ostreidae). In: Re- mieson BGM, Ausi6 J, & Justine J-L, eds., pp. 17-187. production of Marine Invertebrates, Vol. 5. Molluscs: Pe- Editions du Mustum Paris, Paris. lecypods and lesser classes. Giese AC & Pearse JS, eds., Spengel JW 1881. Die Geruchsorgane und das Nervensys- pp. 113-292. Academic Press, New York. tem der Mollusken. Z. Wissenschaftliche Zool. 35: 333- Scarlato OA & Starobogatov YI 1975. New data for con- 383. structing the system of Bivalvia. In: Molluscs, their Sys- Starobogatov YI 1992. Morphological basis for phylogeny tem, Evolution, and Significance in Nature. Likharev IM, and classification of Bivalvia. Ruthenica 2: 1-25. ed., pp. 3-244. Academy of Sciences USSR, Institute of Stasek CR 1961. The ciliation and function of thc labial Zoology Izdatel’stro ‘Nakua’ , Leningrad. palps of Acila castrensis (Protobranchia, Nuculidae), with 1978. Phylogenetic relations and the early evolution an evaluation of the role of the protobranch organs of of the class Bivalvia. Phil. Trans. R. Soc. Lond. B 284: feeding in the evolution of the Bivalvia. Proc. Zool. Soc. 217-224. Lond. 137: 5 1 1-538. 1979. General evolutionary patterns and the system 1963. Synopsis and discussion of the association of of the class Bivalvia. Trans. Zool. Inst. Acad. Sci. USSR ctenidia and labial palps in the bivalved Mollusca. Veliger 80: 5-38. 6: 91-97. Schaefer K 2000. The adoral sense organ in protobranch 1965. Feeding and particle-sorting in Yoldiu ensiferu bivalves (Mollusca): comparative fine structure with spe- (Bivalvia: Protobranchia), with notes on other nuculanids. cial reference to Nucula nucleus. Invertebr. Biol. 119: Malacologia 2: 349-366. I 8 8-2 14. 1966. The eye of the giant clam (TriduEucna maxima). Schneider JA 1992. Preliminary cladistic analysis of the bi- Occasional Papers Cal. Acad. Sci. 58: 1-9. valve family Cardiidae. Amer. Malacol. Bull. 9: 145-155. 1972. The molluscan framework. In: Chemical Zo- 1998. Phylogeny of the Cardiidae (Bivalvia): phy- ology, Vol. 7. Mollusca. Florkin M & Scheer BT, eds., pp. logenetic relationships and morphological evolution with- 1-44. Academic Press, New York. On bivalve phylogeny 323

Steiner G 1992. Phylogeny and classification of Scaphopoda. novelty among arcoid bivalves. In: The Evolutionary Bi- J. Moll. Stud. 58: 385-400. ology of the Bivalvia. Harper EM, Taylor JD, & Crame 1999. What can 18s rDNA do for bivalve JA, eds., pp. 279-289. The Geological Society of London, phylogeny‘? Response. J. Mol. Evol. 48: 258-261. London. Steiner G & Muller M 1996. What can 18s rDNA do for Trevallion C 1965. A Study of Detritus-feeding Bivalve bivalve phylogeny? J. Mol. Evol. 43: 58-70. Molluscs, and an Investigation on Detritus. Ph.D. Thesis. Steiner G & Hammer S 2000. Molecular phylogeny of the University of Southampton. Bivalvia inferred from 18s rDNA sequences with partic- Turner RD 1966. A Survey and Illustrated Catalogue of the ular reference to the Pteriomorphia. In: The Evolutionary Teredinidae. The Museum of Comparative Zoology, Biology of the Bivalvia. Harper EM, Taylor JD, & Crame Cambridge. 265 pp. JA, eds., pp. 11-29. The Geological Society of London, Verdonk NH & Van den Biggelaar JAM 1983. Early devel- London. opment and formation of the germ layers. In: The Mol- Stempell W 1898. Beitrage zur Kenntnis der Nuculiden. lusca, Vol. 3. Development. Verdonk NH, Van den Big- Zool. Jb. Suppl. 4: 339-430. gelaar JAM, & Tompa AS, eds., pp. 91-122. Academic Stenta M 1909. Uber ein neues Mantelrandorgan bei Leda Press, New York. commututu. Zool. Anz. 35: 154-157. Vokes HE 1968. Genera of the Bivalvia: A Systematic and Stoll E 1939. Malletia gigantea (Smith) ein beitrag zur Bibliographic Catalogue. Paleontological Research kenntnis der protobranchiaten Muscheln. Acta Zool. 20: Institution, New York. 307 pp. 339-438. Waller TR 1978. Morphology, morphoclines, and a new Stork HA 1934. Beitrage zur Histologie und Morphologie classification of the Pteriomorphia (Mollusca: Bivalvia). des Osphradiums. Arch. NCerl. Zool. I: 77-99. Phil. Trans. R. Soc. Lond. B 284: 345-365. Stuardo JR 1968. On the Phylogeny, Taxonomy, and Distri- 1980. Scanning electron microscopy of shell and bution of the Limidae (Mollusca: Bivalvia). Ph.D. Thesis. mantle in the Order Arcoida (Mollu Harvard University, Cambridge. 327 pp. Smithsonian Contr. Zool. 313: 1-58. Studnitz G 1931. Die Morphologie und Anatomie der Lima 198 1. Functional morphology and development of injuta, der Feilenmuschel, nebst biologischen Untersu- veliger larvae of the European oyster, Ostrea edulis LinnC. chungen an Lima hiarzs Gmelin. Zoologisches Jahrbuch, Smith. Cont. Zool. 328: 1-70. Abteilung Anatoniie 53: 199-3 16. 1990. The evolution of ligament systems in the Bi- Tan-Tiu A & Prezant RS 1989. Shell tubules in Corbicula valvia. In: The Bivalvia-Proceedings of a Memorial juminea (Bivalvia: Heterodonta): Functional morphology Symposium in Honour of Sir Charles Maurice Yonge, Ed- and microstructure. The Nautilus 103: 36-39. inburgh, 1986. Morton BS, ed., pp. 49-71. Hong Kong Taylor AC, Davenport J, & Allen JA 1995. Anoxic survival, University Press, Hong Kong. oxygen-consumption, and hemocyanin characteristics in 1998. Origin of the molluscan class Bivalvia and a the protobranch bivalve Nuculu suEcata Bronn. Comp. phylogeny of major groups. In: Bivalves: An Eon of Evo- Biochem. Physiol. A Comp. Physiol. 112: 333-338. lution-Palaeobiological studies honoring Norman D. Taylor JD, Kennedy WJ, & Hall A 1969. The shell structure Newell. Johnston PA & Haggart JW, eds., pp. 1-45. and mineralogy of the Bivalvia. Introduction. Nuculacea- Trigonacea. Bull. Br. Mus. Nat. Hist. (Zool.) Suppl University of Calgary Press, Calgary. 12s. Waterbury JB, Calloway CB, & Turner RD 1983. A cellu- 1973. The shell structure and mineralogy of the Bi- lolytic nitrogen-fixing bacterium cultured from the gland valvia 11. Lucinacea-Clavagellacea, Conclusions. Bull. of Deshayes in (Bivalvia: Teredinidae). Br. Mus. Nat. Hist. (Zool.) 22: 255-294. Science 221: 1401-1403. Terwilliger NB, Terwilliger RC, Meyhofer E, & Morse MP Webb CM 1986. Post-larval development of the tellinacean 1988. Bivalve hemocyanins-a comparison with other bivalves Abva alba, jabula, and Donux vittatus molluscan hemocyanins. Comp. Biochem. Physiol. B 89: (Mollusca: Bivalvia), with reference to the late larva. J. 189- 195. Mar. Biol. Assoc. UK 66: 749-762. Terwilliger RC, Terwilliger NB, & Arp AJ 1983. Thermal 1987. Post-larval development of the bivalvcs Nu- vent clam (Calyptogena magn$ca) hemoglobin. Science cula turgidu, Venus striatula, Spisula subtruncuta, and s. 219: 981-983. elliptica (Mollusca: Bivalvia), (with reference to the late Tevesz MJS 1975. Structure and habits of the “living fossil” larva). J. Mar. Biol. Assoc. UK 67: 441-459. pelecypod Nentrigonia. Lethaia 8: 32 1-327. Wheeler WC 1995. Sequence alignment, parameter sensitiv- Thiele J 1887. Ein neues Sinnesorgan bei Lamellibranchia- ity, and the phylogenetic analysis of molecular data. Syst. ten. Zool. Anz. 10: 413-424. Biol. 44: 321-331. 1 889. Die abdominalen Sinnesorgane der Lamelli- 1996. Optimization alignment: the end of multiple branchiaten. Z. Wissenschaftliche Zool. 48: 47-59. sequence alignment in phylogenetics? Cladistics 12: 1-9. Thomas RDK, Madzvamuse A, Maini PK, & Wathen AJ Wheeler WC & Gladstein D 2000. POY: the Optimization 2000. Growth patterns of noetiid ligaments: implicatioiis of Alignment Characters. Program and documentation of developmental models for the origin of an evolutionary available at ftp.amnh.org/pub/molecular, New York. 3 24 Giribet BL Wheeler

Wheeler WC & Hayashi CY 1998. The phylogeny of extant 19.57. Mantle fusion in the Lamellibranchia. Pubbl. chelicerate orders. Cladistics 14: 173-1 92. Stn. Zool. Napoli 29: 151-171. Wilson BR 1979. A revision of Queenshd lithophagine 1967. Form, habit, and evolution in the Chamidae mussels (Bivalvia, Mytilidae, Lithophaginae). Rec. Austr. (Bivalvia) with reference to the conditions in the MLIS.32: 435-489. (Hippuritacea). Phil. Trans. R. Soc. Lond. B 252: 49-105. Wilson EB 1892. The cell lineage of Nereis. J. Morphol. 6: 1968. Form and habit in species of hlalle~is(includ- 36 1-48 1. ing the “hammer oysters”) with comparative observations Wingstrand KG 1985. On the anatomy and relationships of of Isognomon isognomon. Bid. Bull. 135: 378-405. 1969. Functional morphology and evolution within recent Monoplacophora. Galathea Report 16: 1-94. the Carditacea (Bivalvia). Proc. Malac. Soc. Lond. 38: Winnepenninckx B, Backeljau T, & De Wachter R 1996. 493-527. Investigation of molluscan phylogeny on the basis of 18s 1971. On functional morphology and adaptive radi- rRNA sequences. Mol. Bid. Evol. 13: 1306-1317. ation in the bivalve superfamily Saxicavacca (Hiatella (= Woods FH 1931. History of the germ cells in Sphaerium Saxicava), Saxicavella, Panomya, Panope, Cyrtodaria). striatinum (Lam.). J. Morphol. 51: 545-595. Malacologia 1 1 : 1-44. Yonge CM 1923. The mechanism of feeding, digestion, and I 973. Functional morphology with particular refer- assimilation in the lamellibranch Mya. British J. Exp. Biol. ence to hinge and ligament in Spondylus and Plicatula and I: 1923-1924. a discussion of relations within the superfamily Pcctinacea 1926. Structure and physiology of the organs of (Mollusca: Bivalvia). Phil. Trans. R. Soc. Lond. B 267: feeding and digestion in Ostrea edulis. J. Mar. Bid. 173-208. Assoc. UK 14: 295-386. 1977. Form and function in the Anomiacea-Po- 19393. The protobranchiate Moll~isca;a functional dodesmus, Anomia, Patro, Enogmonia (A nomi i dae), Pla- interpretation of their structure and evolution. Phil. Trans. cunanomia, Placuna ( f. Amer. nov.). Phil. R. Soc. Lond. B 230: 79-147. Trans. R. Soc. Lond. B 276: 453-523. 1939b. On the mantle cavity and its contained or- 1978. Significance of the ligament in the classifica- tion of the Bivalvia. Proc. R. Soc. Lond. B 202: 23 1-248. gans in the Loricata (Placophora). Quart. J. Microsc. Sci. 1979. Cementation in bivalves. In: Pathways in Mal- 81: 367-390. acology. van der Spoel S, van Bruggen AC, & Lever J, 1946. On the habits and adaptations of Aloidis eds., pp. 83-106. Bohn, Scheltema and Holkema, Utrecht. () gibha. J. Mar. Biol. Assoc. UK 26: 358-376. 1980. Functional morphology and evolution in the 1947. The pallial organs in the aspidobranch gastro- Tridacnidae (Mollusca: Bivalvia: Cardiacea). Rec. Austr. pods and their evolution throughout the Mollusca. Phil. MU. 33: 735-777. Trans. R. Soc. Lond. B 232: 443-518. 1982a. Mantle margins with a revision of siphonal 1948. Cleansing mechanisms and the function of the types in the Bivalvia. J. Moll. Stud. 48: 102-103. fourth pallial aperture in SpisuZa suhtruncata (da Costa) 1982b. Ligamental structure in Mactracea and My- and Zutmria (L.). J. Mar. Biol. Assoc. UK 27: acea (Mollusca: Bivalvia). J. Mar. Bid. Assoc. UK 62: 585-596. 171-186. 1952a. Studies on Pacific coast mollusks. VI. A note Yonge CM & Morton BS 1980. Ligament and lithodesma in on Kelliu laperousii (Deshayes). Univ. Calif. Publ. Zool. the Pandoracea and Poromyacea with a discussion on evo- lutionary history in the Anomalodesmata (Mollu 55: 451-454. Bivalvia). J. Zool. Lond. 191: 263-292. 1952b. Structure and adaptation in Entodesma saxi- Zaitseva OV & Sokolov VA 1981. Cellular organization of cola (Baird) and Mytilimeria nuttallii Conrad. Univ. Calif. osphradia in the Lamellibranchia Unio peclorum and An- Publ. Zool. 55: 439-450. odonta cygnea. Arkh. Anat. Gistol. Embriol. 80: 9OkY7. 1953a. The monotnyarian condition in the Lamelli- Zardus JD & Morse MP 1998. Embryogenesis, morphology, branchia. Trans. R. Soc. Edinburgh 42: 443-478. and ultrastructure of the pericalymma larva of Acilu cus- 1953b. Form and habit in Pinnu carnea Gmelin. trensis (Bivalvia: Protobranchia: Nuculoida). Invcrtebr. Phil. Trans. R. Soc. Lond. B 237: 335-374. Biol. 117: 221-244.