Perec sheds heey Van Gotan ee NASON RAF a Oe ve Ro VNens Neiteita ee ASNT TN, Tol ee Se COT Os ae POs aC SONG Nae eA ag tl aan crn ac ER Eee TANIRcd eaatacheehinnnattcoaas F : AU CRN LESS aioe Aracnenena tentpouraansmetegarateae s A . iy oat * Cree ee Den ta A ay WV UPLAND, PONERIN MARNE GEN TO De Nate PLN IRe Meme RC WENA, it . i s Laas AM’ 7 NN ORHRSTC AT YRS Sry DAL Wtg ie AR sr Mbuly cine ge Netba MAR Tees Ie ? ; ath STMT SUR aR RCT eR PTH NEP RE reM TR eeemarT : t pa ITER TRE IPT Petra Rents tr Po anead gH Ca UR red By ne ; ash aed nr dat MOM REECLantCeTRT EAE MR CAER TR AUa Tee ULpnas eeu imesh aaaut NERUGL EBA ; SHEN TAPP PIEE ATi RET Hn dren Pen en taday PUT aEUIN Ee py rr asd Rut iNGad's Cy dee AAA IATA LaNt AA cede Raa s aaantn ea a va URSA aaa eee RYE aN ‘ aM REAPS WaT En AU Rien Mesaagncncudate peated a) . PRIORI APT IATA ERT ot RIP ST a hetaaGan death we N20 A MAAgS Sn iyriony wansaraaes Me Qenangaviegitiies . neatesy sath IU BOLSsamedncteaeyne UV vow tans EMA HN AGHNE HAM ie UD tan tit yok, putea anamaNon creuina fa wit ing teasing HET REM OES CE REST MET ie. Zena an NH BANA ARt Cane OMU UREA INK Irateha >Etoa Aaa et . . reer Pres Petes ee ete Te Sean ee UROL a PL end BRE oe ad NP AER ISM ER otis Baer AURA Nias AM sBulehoadatit mnsae meen swileceeAs RENAtON eee MUTANT Nia ae teeee meeaaan , . . t J ’ Peas eee ee ee Woe were Pen eee Cee On eee TAR Lary 0 Oey Sa ea oe : ‘el s Vendy ST AUPAEA Seve Vinchinn neat vaagN rags es pesb susUriby tela VOM A LA amen ilessyy \ehie beAMN ALN Oe ata 4 ? : canes cae Ried OA oan TA ateae LALA ILGAT ENN COL NNN Song NVI MIN Meibe ener sate aie Street Vi ie Bieerory eRe) ROM REN Aaa ies Ma es Vk bet toe be A BT ee fay a RIMLESA aA TACARHCDNUL acm Umae eae ceeanah even goat ine He anion err texe era 4 H ‘ Ait A at Veceihdtgeiadsgo. Sind ase Aanagdepaed af ye 4 ; at Vine vtsustasnotatneakinadoaee i Poh Re dare Aa ye wuaregdcavAdushcatatpenedancsoneae teeta sity : : eon ’ me \ Vel vey é ae Paya rs , : : i FERRY RE err MUinUea sean aye Aces og ' Nia bade} be eet Par ea SUN hg SPAY LOG DabEN nee Ledughdeviqoue sls ie SM ctey mad AAG aa PARES woe ae Hosea ahs enead aie Ns Wa ine TE A anny aed apie iter) in Hayes Deracge dacaguessaecasninemenstipsnegpe EU Rarer aan soln slatalmsscagasheneeamineiageasacenman races ins SY aaee Rennes : ous pate Haaren 5 on Nearness Asanti A cea eat 4 Rereresy : rasa Ore : me abla quis : Durer uynd ye sagen riers Rly act yes Ei oenibeiG Pate gcenieny sang hey Cry coast 1 boreal ye are MEV Tere beng ME Oth ede aeaiep aL Fanaptee oe dataips pe Da to satige 1d Le paecve eB Pivra sigagey eines gery Brett ‘i pinata phe Dy ‘ Vitasies vse hypsge qinaion at Be oeat sie nr ea : Se gueanres ai tips Hr deen ' ised ; ine hepa Boas yg pee ag Vie, tice SPANKS Les ‘ . poet) , tes crepe rar fate ptee es ‘ Tbe ohana Pipe edt aie MEET wae tags Pe ay Terai cs tik dy oo Peay) ol tein ste esid bead eg vii sto Peak it : z te aepeet taut rigs vel pe er Pharr Pree Re RO OSCR Le ener cca ere LY SP an ie Cle Lema TE Peary PeROLe TP EELOL SLOT TLE tenes fe . fod im . a ef COCR CEE ee Ry ee Saree : pista eteen Abies g Onsite paig eer Ota on fen Cerra hoacd mute oe wed ne PVNN Dialed EDN a tanh (iiss ‘ ne Gn poets Ag atnesng ie Fp aN es pay * jee Gio Ok WP y Gea aly Pee yt - Pepe sen ria Aa OEE Ti one : ds vr ae Reb APNE SH GEGEN Hie Uap ov 8 | C48 areas yrursiib ase WD ata ep Uagesteruliae ene # ro reer gr eeanys y guste cssionpe wpe eel? . Te biG Gon ct yrs fe ee i ‘ . Pope ate total aeere ‘ era nes : ‘ pow i Sphay fogsdedige pegrogad se hey ee ara Fpl ger anae din pellvae pretanneay nis 2 ae . ten ern , ata vee wane gue Weaievate Evi : , 2 ! san Wiads ene EO ee ge 8 re toeny EP ra tevasaiais oc ang ave ii on : rune) Fidcne Deve pordwes ane nee PY eee ee Pyar Terrase . 5 tae ' - i Pe PSE TDA ct state 0e a! ON dean geod ap Hey weed reper een Se é a deere Par davay in cacn vacua Wye ead hoe 689 Seg pt DNB AS an in . ve Va ky PE DW eg a pe are ead pee Pps eases sas vate Gn een Paneer Ser OSCR ee ene er ee eee ere Lee ee Garr ibaa yea aw Cara eer ein eps URUA Sg aus pseteune a H ‘ ‘ . ve irereniee Py ree EAT On UE TOTS PESTLE LeU EEE Petipa ego gusts teoeare gent 7 : AYER tte) ‘ PP AIA PENSE DLO OMS DEES Quer ta vp PE pate A Ge cat ee aa DOP Oa Pea UR PLOT TEL UR Le be NU toa LSELMCMLASHAPRD OD capeige nue pss) ; naeag SPA ARGH NYS Bre UeeDL TOW pocuted QeNERUETE EUs Re FUORI ENS ANOr EW bNEL oo Mian Upeciebe puge ‘ vie SOE Pen prin terreus CiMNGs alee oH ctw gee at bee cde Neg wot HE Terie : yes ee ae Ne TD DAMES SN aE Nay ' Peri ese Ai, Ca ae HP PION eg pin Mabey oe oes Oy pales vie Peroersorey Pree serra i ‘ Hae Toye se nue bead seub ite prec ee ce : 4 ’ woge Corer an Seangiuesy sid op eean sal eeyae th i : 2 Or aeete nT * UM pea tala bid se phase neers Se ONT an BENG as OTR att bn ne toe ‘ core PVE ASE anat pimiatiten stant yeh ae R an i 1 STAR Beppe Cate AECOR NO te are pCE ant goa citi , na no ve Petar RATS AV AZ GAHAN! SM ONI/ CME ENA OREM OEDIAD OA, tetiga ‘ , DERI cha tpg ale alee aces POE VE BOIL PEDSD bigtod peg die Nag Yu ta Hey nee Mya eyaney ind sete tay SUF RE Daw pith Bhaayiwaper U4 Bee : WANSTEAD EOD bo ged piety gin acre yipesbtune suntooe pei j ‘ ‘ . ' Ore Cee renner ert Siren ay ea ee mat ere ree . B ‘ pore hanes PUEDE a SU ROE ON TONE pst ‘ Wen das a . ‘ Vaciowiy OCT eee : rt : ee Ue Pes EEE OTTO Tee TT Te en ae OC ae re Vr) A avestaial ye Tee y een Me ee ee eM RE UIA MeO ee eee er Pe ; Ae, Oa, ION (einer ieay pEMAM stage ALS UOT UOT MEP EURO TAe TIES DLE Le Le SOnreT PO err TiC Peed tat i WEE pat UE EO SERN CBU REETON GEER GUD ROSE DNBILMUSDGI PAG Fea tiye ret eo atu ae ere ys PUN eee aU ee ST Lk ST SE ESE Ee EY POLE TI A ss al 1 HPP Cerne RUTULS MOU EE UOT CE Sneed COREE MESO TE Se ny Eee Per ee CET TeT tedy ere tet CP atest oe masevo re hiveat een Magi vied ues Pare Sr rrer erwrnne et aren i Bere rrereninon iy Ueerns oes ies * a Leauike A VPRO NG ABNEY, : , Lee Pinney Wyte stare wacneee CTU IE Re ES ORO TL aU Pee RT eros it Fai oe : HELE EN UI ORE SERRE ECP TU RN TENS ety ene eas Be rane TRV EE Seer ee TU TST STOLE LOS NPETEL Tet Rn Peter y Pris 5 epreses Fag arnt ad eg s Tyne ese eased Wag CH TUDE 2b Ep \te ga COIL A HRA 4 : eM Oe DEO MCT Dear ee SURO TORO Lore PECL ATT en Tea MONTE D pat ‘ By t 9 DY rene Ue ena uke Meee tna Or PON yaar Pete br Tet even Saye as PEE eNO FUN OLDIE hs WORE Ae DWF UNM PPUIV ere se PeNL Ce Oe eR RUSE eT LUM Oe mes Leet ania Pre at ae a > le) ass A = L158) TITUTION NOILALILSNI > = wn 2 5 Fe 6 Py $ = a = onc a, w ie} fo} a Es ° (a 4, @) ©) ninse nse NOILN 2 NY S\N NOILNS NOILNI iz ™ Canin LIBRA SAIYVUSITLIBRA NVINOSHLINS S3IYVYUEIT LIBRARIES SMITHSONIAN_INSTITUTION NOILNLILSNI_Nt ’ : SJIYvVYusIT LIBRA INSTITUTION NOILN Ssaluvygs LIBRA 1 ff y QXQY . 6Z W sa1uvygi INSTITUTION INSTITUTION INSTITUTION LIBRARIES SMITHSONIAN LIBRARIES SK z ee as pe ce ae a gl = | uvug IT = on = = = = = = fae = ras = 2 a 2 3 & 5 3 é = é = 2 E Z, E Zz B ountiene ; z : z 3 ws ITUTION NOILALILSNI_ NVINOSHLINS LIBRARIES SMITHSONIAN INSTITUTION NOILOLILSNI_ Nt on ah 4 z ” ; Z a z = o me a Y Yn,.° 2 v o — oe = [eed Z ate a = < c .< = < (lid = c 3 a m S ae GS GS ac S rea = co a oo 7 = 0 = a {o) pa fo) ma oO — (o) pas = a | za call a =! oA YUVYUEIT LIBRARIES SMITHSONIAN INSTITUTION NOILNLILSNI NVINOSHLINS SAZINYVYSIT LIBRARIES SI Le z S z = z = = = cS} = 9 = ° a & a 5 2 5 2 E 2 5 > - ¥ > = > | =) > > & Es) — % es) = y E xy a — - -— | od aes = x Ke * m : 2 m O 2 m ” m 2 w = wo < 7.) . £ o = MITUTION | NOILALILSNI_ NVINOSHLINS | S3 lyvudi Vl BRARI ES SMITHSONIAN INSTITUTION | NOILALILSNI_ NI ae = < = = — = . 2 = > = &, 2 (77) 4 ” sd 2 (7%) > z ” yvugitI_ LIBRAR! ES SMITHSONIAN _INSTITUTION NOILNLILSNI NVINOSHLIWS Sa byvusiT_LIB RARI ES Sh 2 uw = us 2 So 2 uw = ee oe e a ae = ee o < Y, = > -: > = >: ~ > rege 0 — es) = — Ss 7 — % = b 2 b = F = BS a Z i Z iy ae is z MITUTION NOILALILSNI — _S3IYVYEIT_LIBRARIES SMITHSONIAN INSTITUTION, NOILOLILSNI_ N\ | - (= BA py ow CAINS, SC RB. cc 2] CS EK wy x YyZ a Say cc OI = woe sey 2 ay = Y = ESS 5 CORRS ad “jl ro) None” a rms ZA = Zz = =! Zz pa] ve = WS” SJIMVYUGIT LIBRARIES SMITHSONIAN INSTITUTION NOILNLILSNI_NVINOSHLINS SSluvudi7 LIBRAR S a = eS fa * S Ee 5 a = } a i= es) YN \. 5 5 fe > = “ P re > We -_ > ‘= 2 = "2 = DB LAY = 2 ae — eae: rd ae Wy w =e = n & We 2 AK SAS ma a — wm a — ; << AN INSTITUTION NOILALILSNI YINGSHEINGS “Sailid Veneta oe BNR Wes SIT SON Ne eration pelt Zz a GUY, = a z = sae 2 GY ) 3 z eg 3 z 5 Nw a MNGi Ff - oO Y hf fp? ae oO 9) | kK ke Y ord Se) lo > z > > > 2G > = vz — Bs) eS 2) =e A ys ee) & = i= a = = - “iby ae = m S m Y m OARS 3 TG mn D —_ (27) . — wn — — WS S3IMVUGIT LIBRARIES SMITHSONIAN NOILNLILSNI NVINOSHLINS S3IYVYEIT _LIBRAR n z oe 7) z w ra , no Zz ah A 3 LD y Oo WN. a = = S = bh fp (@) WS J8 GH 2 8 B z g a GjJ 2X / 2 Gp = “XY Zz c zZ é 2 Uy = S\N ar a — ae oe = Re a tants wa '. AN INSTITUTION NOILOLILSNI NVINOSHLINS S31YVYaiT_ SMITHSONIAN _INSTITUTION NOILNLI 2 ws m™ 2 @ o 2 us = Mt = RQ ¢2) aE ty Se am pret oc “ wR S o a a“. vy Ab = a a < = AS < a = ‘> r= F NSW E > on a r, a = al ANS E = Z i Z ° Ae ee 2 LIBRARI ES SMITHSONIAN NSTITUTION NOILALI AN INSTITUTION NOILNLILSNI NVINOSHLIWS S3I¥VY¥ait : x x Wy, Me. A, Sepia na 7° “s WS S3IYVYGIT LIBRARIES SMITHSONIAN INSTITUTION NOILONLILSNI_ NVINOSHLIWS saluvugly LIBRAF See SMITHSONIAN At J tj NVINOSHLIWS AS SMITHSONIAN SMITHSONIAN NVINOSHLIWS NVINOSHLIWS SMITHSONIAN NVINOSHLIWS 7) > s = ” > $< = x “oy & = Wwe = YY (= ow Cc tc ser PASAY og c ee a i a 0 2 NWS a = = re) ea ro) = ° Sx ~ re) a 4 a) z ei = = = IAN INSTITUTION NOILOLILSNI NVINOSHLINS SSIYVYAIT LIBRARIES SMITHSONIAN INSTITUTION NOILNLI = z cine r = he z a z 40 a ow Ks o wo ° a@W 2) y x 5 D2 ANS = =e] = Gy —:) Ee : : aS : : CMY * : = = = - ~ a “Ys _ Fe (77) : (7p) é 2 2 8 ° a a 2 INS S3IYVYaIT_ LIBRARIES. SMITHSONIAN INSTITUTION NOILALILSNI NVINOSHLINS S3iuvad ITI BRAF = j/ AMERICAN —° MALACOLOGICAL Ms BULLETIN VOLUME 8 NUMBER 1 AUGUST 1990 CONTENTS Life cycle and sexual strategy of Saccostrea cucullata (Bivalvia: Ostreidae) from a Hong Kong mangrove. BRIAN MORTON................ 2.00... c cee ee 1 Ecological analysis of the living molluscs of the Texas panhandle. PRIS DN PURE M rec ts KA ee ee es OME ain aoe ove wee we Ma o Dees 9 Maturation of the reproductive tract of Neohelix major (Binney) (Gastropoda: Polygyridae). MARIA GABRIELA CUEZZO...........................0.... 19 SYMPOSIUM ON THE BIOLOGY OF PELAGIC GASTROPODS Morphology of the wings of euthecosomatous pteropods. DIETER FIEGE ....................... 27 Karyotype analysis in several pelagic gastropods. CATHERINE PnPreA HCH IMED NPR Leno econ yc ha eh ete Ss Gy eee wa ae Eas cease tee 37 Sampling requirements for epipelagic heteropod molluscs. ROGER R. SEAPY................... 45 In situ observations of feeding behavior of thecosome pteropod MO LUSGSiERONALDAW. GIEMER = eg ee oo ey Oa ee Ne ae ete 6 a le baegs 53 Mating behavior and spawning in two neustonic nudibranchs in the family Glaucidae. ROBIN M. ROSS and LANGDON B. QUETIN...........................0... 61 Bipolar variation in Clione, a gymnosomatous pteropod. RONALD W. Se ean, CAROL Ms CALLED ie po dou 2k fija aot ahs fey ha eats pba daawend es dba es 67 Annual cycle of Limacina retroversa in patagonian waters. GSE Re DADON 4. eo, 77 The taxonomy, distribution and biology of Atlanta gaudichaudi Souleyet, 1852 (Gastropoda, Heteropoda) from the Great Barrier Reef. Peo VIMN Aaa etn l et crane Mere Ne tienes es MALS SUAS Dy ea aw as sa cacsordniet Wi wks ore 85 POUMOGI Eee eat aoe ete nen eR UaNE MEM muni at Ns Glas sock aaa we wate wk aad ki teed one bw 95 Eds MGV EURATOTAES Tt YORE 5 et Ee Se Spee Se RECURS A Sco DA Oa 7 OE ne Oe cr i RR da 96 AMERICAN MALACOLOGICAL BULLETIN EDITOR-IN-CHIEF ROBERT S. PREZANT Department of Biology Indiana University of Pennsylvania Indiana, Pennsylvania 15705 MELBOURNE R. CARRIKER College of Marine Studies University of Delaware Lewes, Delaware 19958 GEORGE M. DAVIS Department of Malacology The Academy of Natural Sciences Philadelphia, Pennsylvania 19103 R. TUCKER ABBOTT American Malacologists, Inc. Melbourne, Florida, U.S.A. JOHN A. ALLEN Marine Biological Station Millport, United Kingdom JOHN M. ARNOLD University of Hawaii Honolulu, Hawaii, U.S.A. JOSEPH C. BRITTON Texas Christian University Fort Worth, Texas, U.S.A. JOHN B. BURCH University of Michigan Ann Arbor, Michigan, U.S.A. EDWIN W. CAKE, JR. Gulf Coast Research Laboratory Ocean Springs, Mississippi, U.S.A. PETER CALOW University of Sheffield Sheffield, United Kingdom BOARD OF EDITORS MANAGING EDITOR ASSOCIATE EDITORS RONALD B. TOLL Department of Biology The University of the South Sewanee, Tennessee 37375 W. D. RUSSELL-HUNTER Department of Biology Syracuse University Syracuse, New York 13210 CAROLE S. HICKMAN Ex Officio Museum of Paleontology University of California Berkeley, California 94720 BOARD OF REVIEWERS JOSEPH G. CARTER University of North Carolina Chapel Hill, North Carolina, U.S.A. ARTHUR H. CLARKE Ecosearch, Inc. Portland, Texas, U.S.A. CLEMENT L. COUNTS, III University of Maryland Princess Anne, Maryland, U.S.A. THOMAS DIETZ Louisiana State University Baton Rouge, Louisiana, U.S.A. WILLIAM K. EMERSON American Museum of Natural History New York, New York, U.S.A. DOROTHEA FRANZEN Illinois Wesleyan University Bloomington, Illinois, U.S.A. VERA FRETTER University of Reading Berkshire, United Kingdom ISSN 0740-2783 THOMAS R. WALLER Department of Paleobiology Smithsonian Institution Washington, D. C. 20560 JOSEPH HELLER Hebrew University of Jerusalem Jerusalem, Israel ROBERT E. HILLMAN Battelle, New England Duxbury, Massachusetts, U.S.A. K, ELAINE HOAGLAND Association of Systematics Collections Washington, D.C., U.S.A. RICHARD S. HOUBRICK U.S. National Museum Washington, D.C., U.S.A. VICTOR S. KENNEDY University of Maryland Cambridge, Maryland, U.S.A. ALAN J. KOHN University of Washington Seattle, Washington, U.S.A. LOUISE RUSSERT KRAEMER University of Arkansas Fayetteville, Arkansas, U.S.A. JOHN N. KRAEUTER Baltimore Gas and Electric Baltimore, Maryland, U.S.A. ALAN M. KUZIRIAN Marine Biological Laboratory Woods Hole, Massachusetts, U.S.A. RICHARD A. LUTZ Rutgers University Piscataway, New Jersey, U.S.A. GERALD L. MACKIE University of Guelph Guelph, Ontario, Canada EMILE A. MALEK Tulane University New Orleans, Louisiana, U.S.A. MICHAEL MAZURKIEWICZ University of Southern Maine Portland, Maine, U.S.A. JAMES H. McLEAN Los Angeles County Museum Los Angeles, California, U.S.A. ROBERT F. MCMAHON University of Texas Arlington, Texas, U.S.A. ROBERT W. MENZEL Florida State University Tallahassee, Florida, U.S.A. ANDREW C. MILLER Waterways Experiment Station Vicksburg, Mississippi, U.S.A. BRIAN MORTON University of Hong Kong Hong Kong JAMES J. MURRAY, JR. University of Virginia Charlottesville, Virginia, U.S.A. RICHARD NEVES Virginia Polytechnic Institute and State University Blacksburg, Virginia, U.S.A JAMES W. NYBAKKEN Moss Landing Marine Laboratories Moss Landing, California, U.S.A. A. RICHARD PALMER University of Alberta Edmonton, Canada WINSTON F. PONDER Australian Museum Sydney, Australia CLYDE F. E. ROPER U.S. National Museum Washington, D.C., U.S.A. NORMAN W. RUNHAM University College of North Wales Bangor, United Kingdom AMELIE SCHELTEMA Woods Hole Oceanographic Institution Woods Hole, Massachusetts, U.S.A. DAVID H. STANSBERY Ohio State University Columbus, Ohio, U.S.A. FRED G. THOMPSON University of Florida Gainesville, Florida, U.S.A. NORMITSU WATABE University of South Carolina Columbia, South Carolina, U.S.A. KARL M. WILBUR Duke University Durham, North Carolina, U.S.A. Cover. Late veliger of Atlanta gaudichaudi Souleyet, 1852. The biology of this species is discussed in a paper by Newman in this issue. This Paper is one in a series of papers that appear herein as part of the 1989 American Malacological Union Symposium on the Biology of Pelagic Gastropods. THE AMERICAN MALACOLOGICAL BULLETIN is the official journal publication of the American Malacological Union. AMER. MALAC. BULL. 8(1) August 1990 LIFE CYCLE AND SEXUAL STRATEGY OF SACCOSTREA CUCULLATA (BIVALVIA: OSTREIDAE) FROM A HONG KONG MANGROVE BRIAN MORTON DEPARTMENT OF ZOOLOGY UNIVERSITY OF HONG KONG HONG KONG ABSTRACT Stems, roots and pneumatophores of Hong Kong mangroves are colonised densely by Saccostrea cucullata (Ostreidae). A 13 month study of this oyster has shown that recruitment to the adult popula- tion takes place in a single extended phase over mid-summer, correlated with rising sea temperatures and falling salinities. New recruits generally mature into males, creating a heavily male-biased (78%) juvenile sex ratio. Such individuals, maturing in their first year, fertilize a predominantly female biased group of adults (two and three year olds) and then change sex in June to become females themselves. Some two year old oysters change sex, again in June, to become males, but an overall older female biased sex ratio is maintained. In this habitat, Saccostrea cucullata rarely lives beyond three years, such individuals constituting an insignificant component of the total population. Four year old oysters occur only rarely and are covered by successive recruitments. Such a life history tactic and sexual strategy of alternative hermaphroditism equips this species for opportunistic colonization of the high-zoned mangal and places it firmly, with other species shar- ing this rigorous habitat, at the r end of an r-K continuum. The rock-oyster Saccostrea cucullata (Born, 1778) is virtually ubiquitous on Hong Kong shores. It occurs as scat- tered, isolated, individuals on shores experiencing high ex- posure to wave action, but forms a dense band in the eulit- toral zone of shores experiencing less exposure and is par- ticularly obvious on estuarine shores where it encrusts aerial roots and stems of the seaward fringe mangrove plants Kandelia candel (L.) Druce and Aegiceras corniculatum (L.) Blanco and pneumatophores of Avicennia marina (Forsk.) Vierh. In such a situation it is sympatric with the bysally- attached mytilid Brachidontes variabilis (Krauss, 1848) (Mor- ton, 1988). Saccostrea cucullata is the type species of the genus (Stenzel, 1971) and has a wide Indo-Pacific distribution, from East (Day, 1974; Branch and Grindley, 1979) and South Africa (Lasiak, 1986) to the Pacific Islands, e.g. Guam (Stojkovich, 1977; Braley, 1982), and from New Zealand (Dinamani 1974; 1976) and Australia (Roughley, 1933) to Japan (Torigoe, 1981). Stenzel (1971) regards Saccostrea cucullata as a ‘superspecies’ and its component ‘species’, such as S. glomerata Gould, 1850, S. commercialis |redale and Roughley, 1933, S. echinata Quoy and Gaimard, 1835 and S. mordax Gould, 1850, as no more than geographic subspecies, ecomorphs or variants. Morris (1985) adds Ostrea circumsuta Gould, 1850 to this list. Dinamani (1976) could not find one valid morphological character to separate S. commercialis (from Australia) from S. glomerata (from New Zealand) and further found that many mangrove oysters from the South Pacific could be allied to S. glomerata and thus with S. cucullata. Similarly, the spiny S. echinata from Hong Kong and elsewhere, e.g. Inhaca Island (Macnae and Kalk, 1969), and the similarly spinose S. kegaki Torigoe and Inaba, 1981 from Japan (Torigoe and Inaba, 1981) are probably no more than sheltered habitat ‘forms’ of S. cucullata (Morris, 1985). Dinamani (1976) concluded that revision of this oyster assemblage could only be attempted after more detailed studies of all the ‘species’ of the S. cucullata group had been undertaken and laid the foundation for this by investigating the sexual cycle of S. glomerata from New Zealand (Dinamani, 1974). There is a considerable literature on the Caribbean mangrove oyster Crassostrea rhizophorae (Morton, 1983), but few detailed studies of its Indo-Pacific counterpart. Such in- formation as is available on reproduction has been reviewed by Lasiak (1986). Hong Kong’s rock and mangrove oysters have been identified as Saccostrea cucullata (Morris, 1985) and this study is thus of the type species of the genus. Until such time as the taxonomy of all Indo-Pacific ‘species’ has American Malacological Bulletin, Vol. 8(1) (1990):1-8 1 2 AMER. MALAC. BULL. 8(1) (1990) been stabilized, only limited comparison with them, e.g. S. glomerata (Dinamani, 1974), is possible although the analysis of reproduction in S. cucullata from seven locations by Lasiak (1986) has allowed tentative generalizations to be made. Morton (1985; 1988; 1990a) has studied the life history tactics and sexual strategies of Hong Kong’s other mangrove bivalves, i.e. Polymesoda (Geloina) erosa (Solander, 1786), Brachidontes variabilis and Gafrarium pectinatum (L., 1767), and this study permits comparison of Saccostrea cucullata with them to determine whether or not they are united by com- mon features that permit colonization and exploitation of the high-zoned mangrove environment. MATERIALS AND METHODS Every month (save April 1989) from October 1988 to October 1989, i.e. 13 (minus 1) months, a large sample (X= ~500) of Saccostrea cucullata was obtained from the mangrove at Ting Kok, New Territories of Hong Kong. Upon return to the laboratory they were separated into individuals and wet weighed to the nearest 1g. This allowed the construc- tion of monthly total weight frequency histograms. Oysters were subsequently fixed in 10% neutral formalin and, wherever possible, five individuals from each 2g weight class were sectioned. A piece of the body through the gonad was removed from each individual and, following routine histological procedures, sectioned at 6 um and stained in Ehrlich’s haematoxylin and eosin. Using previously defined criteria of developmental stages (Morton, 1982a, b; 1985; 1988; 1990a; Dudgeon and Morton, 1983), the gonads of Saccostrea cucullata were assigned stages of either: 1, primordial; 2, developing; 3, maturing; 4, mature; or 5, spent. Developmental stages for S. glomerata (= S. cucullata) are illustrated by Dinamani (1974). Data for each sex were also applied to the relative stages of development of hermaphrodites, and used to determine: (a), the timing and size of first maturity; (b), the timing and fre- quency of sex-reversal; (c), the subsequent pattern of gametogensis; and (d), the male/female sex ratio, overall, and with increases in size and from month to month. Using a test of proportions, the incidence of males in the total sample, ob- tained monthly over the 13 month period, was tested against the null hypothesis of equal proportions of both sexes, i.e. both 50%. A Chi-square test was applied to (a), the monthly samples and (b), to the different weight classes to determine if the sex ratio deviated from 1:1 in accordance with these parameters. This has allowed assessment of the reproduc- tive cycle and the sexual strategy adopted. On each visit, undertaken in mid-morning on a falling tide, three water samples were obtained, following temperature measurement in situ. Each sample was held in air-tight, dark glass bottles and upon return to the laboratory analysed as follows: the first to determine dissolved oxygen level; the second, BODs; the third salinity, pH, phosphate and nitrate-nitrite. The latter two were determined using a La Motte Water Analysis Kit (La Motte Chemicals Ltd.). RESULTS HYDROLOGY Morton (1990a) reported upon the hydrology of the Ting Kok mangrove in relation to the life cycle of Gafrarium pec- tinatum. The present study adds information to that one. Sea temperature fell to a low of 17°C in January 1989 and then progressively rose to 33.5°C in September 1989 (Fig. 1A). Thereafter it fell. Rainfall was not heavy in 1989, reflected in the low variation in salinity. Salinities were always > 25 9, reaching a maximum of > 30%%oo in May and July 1989 (Fig. 1A). Rainfall lowered salinity slightly in late summer and it re- mained low for October 1989 (28%). The exaggerated salinity curves typical of Hong Kong’s inshore waters (Morton and Wu, 1975) were thus not demonstrated in 1989. pH values were generally low in the winter months of December 1988 - January 19839, i.e. 7.6, and rose to 8.25 in August 1989 (Fig. 1A). Dissolved oxygen levels were consistently > 3 mg.I-1 with high values recorded in spring (March 1989) (5.6 mg-I-1) and July and August (mid-summer) (> 5 mg-I-') (Fig. 1B). BODs values were all < 4.0 mg.I-! with a low value of 1.2 mg-I"1 being recorded in January 1989 and a value of 2.4 mg-I-1 in July 1989. Highest values were recorded from February - June 1989, i.e. > 3 mg-l-1 (Fig. 1B). Nitrate-nitrite levels of 0.88 and 1.14 mg-I-' were record- ed for the months of October and November 1988, respec- tively, but thereafter values fell below 0.4 mg-I-1 and in July, August and September 1989 undetectable levels were pre- sent (Fig. 1C). Similarly with phosphate, high levels of 0.45 and 1.05 mg-I"' were recorded in October and November 1988, 35 +— pe a A eee | eo are ° ° ° ° << Ve eee © Temperature © Salinity 4 pH = Ae a a. _ Aa © Oxygen @ BODs © Nitrate—nitrite @ Phosphate I = 4 E 0.4 ¢: }. <= Se, Aw ao o—__@——_e-—_@ <—— $9 Yo s ° N D J F M A J J A s N 1988 1989 Fig. 1. Water quality parameters at Ting Kok mangrove, Tolo Har- bour, Hong Kong, over the period from October 1988 to October 1989. A, temperature, salinity and pH. B, dissolved oxygen and BODs. C, Nitrate-nitrite and phosphate. MORTON: SACCOSTREA LIFE HISTORY 3 respectively, but figures generally were < 0.05 mg/l thereafter (Fig. 1C). POPULATION DYNAMICS The highly variable shell form of Saccostrea cucullata precludes use of standard measurements in the construction of normal length frequency histograms that have been the basis for assessment of population dynamics in the Bivalvia. Instead, the total weights of intact animals have been used in the construction of equivalent histograms for this species (Fig. 2). The resultant histograms are thus less discriminating of age cohorts but nevertheless important generalisations can be made. Newly recruited juveniles of < 1.0g appeared in the population, attached to older individuals, during the period from March - September 1989. This seemed to occur as a single wave of recruitment, undivided into sub-cohorts, and coinciding with spring and summer in Hong Kong. With this information, the histograms were interpreted as follows. In October 1988, the major peak of ~ 4g represents settlement of that summer’s recruits. Other peaks of ~ 8g and ~ 13g represent settlements in 1987 and 1986, respectively. The few individuals of 19g and 22g are of unidentifiable age, but clearly older than three years. From November 1988 to February 1989, a similar population picture was obtained, as also in March 1989 ex- cept that the incidence of individuals weighing < 1g increased to comprise ~ 12% of the population. These new recruits were easily identifiable for the next six months, with the other age cohorts progressively increasing in weight. Peak recruit- ment appeared to occur in July and August with older (1988 and 1987) cohorts masked by dramatic settlement and growth of the younger animals. By September 1989 the rate of recruit- ment had slowed down and by October 1989 a picture similar to that seen in October 1988 was obtained with a large peak of 1989 recruits of ~ 4g followed by further peaks of ~ 8g and ~ 12g representing settlements in 1988 and 1987, respec- tively, and with a single individual of 17g of indeterminate age. It thus seems that Saccostrea cucullata lives for ~ 2 years, with only a few identifiable animals (1 - 2%) entering their third year and with even fewer (< 1%) entering their fourth and possibly fifth years. GROWTH Analysis of the histograms presented in figure 2 and identification of cohort peaks for each of the recruitment years has allowed an assessment of growth (Fig. 3). Somatic growth appears restricted to spring and summer months from March to September with no overall increasing trend being evident for all cohorts from October - February. With settlement oc- curring over the same period, it is clear that the newly recruited oysters reach an approximate weight of 5g very quickly. In their second year they attain a weight of ~ 12g and in their third a weight of ~ 15g. Somatic growth and gonad maturation, over the time frame monitored, thus coin- cide with rising sea (and air) temperatures, reduced salinities and high oxygen levels. SEXUAL CYCLE Analysis of sectioned gonads of Saccostrea cucullata shows that the species completes one sexual cycle each year (Table 1). In October 1988, the majority of gonads were either mature (Stage 4) or spent (Stage 5). This situation continued until November but by December 1988 the gonads were large- ly primordial (Stage 1) or immature (Stage 2). By February - March, gonads were maturing (Stage 3) with a majority reaching sexual maturity (Stage 4) in April - May 1989. Gonads were mature from this time until approximately August - September 1989, with the incidence of spent individuals in- creasing in October 1989, as in October 1988. For S. cucullata, therefore, spring in Hong Kong is the period of sexual matura- tion, summer when the gonads are mature and autumn and winter the time of gonadal decline and gametogenesis initia- tion respectively. SEXUAL STRATEGY Table 1 shows the stage of sexual maturity of sectioned individuals of, wherever possible, five individuals belonging to each 2g weight class. The overall sex ratio is biased towards males (50.5% males vs 36.6% females). This is significantly different (P = < 0.001). There is, moreover, a significant dif- ference (P = < 0.001) between the period from October 1988 to May 1989 and that from June 1989 to October 1989 with males predominating in the former and females in the latter. Significantly also (P = < 0.001), the sex ratio changes with increasing size and thus age. In the weight class from 0 - 1.99, 77.6% of all individuals were male. This figure pro- gressively decreased while the corresponding figures for females increased to 77.8% in the 16.0 - 17.9g weight class. In June 1989, however, 27.1% of all individuals were hermaphroditic. Such hermaphrodites could be divided into two sub-groups in the weight ranges 2.0 - 79g and 10.0 - 15.9g. The former hermaphrodites were spent males (Stage 5) and either maturing (Stage 3) or mature females (Stage 4). The latter were spent females (Stage 5) and either developing (Stages 1-3) or mature males (Stage 4). Such hermaphrodites constituted an average of 2.8% of the 2.0 - 7.9g individuals, and 5.8% of the 10.0 - 15.9g individuals, i.e. the latter were relatively twice as numerous as the former in their respec- tive weight classes. From the foregoing, a picture of the sexual strategy can be constructed. The majority (78%) of newly recruited in- dividuals mature into males during their first year of life leading to a juvenile male biased sex ratio from October to May. In June, many juveniles, now one year old, change sex into females leading to a female biased sex ratio from July to Oc- tober. Oysters in their second year of life, i.e. > 8.0 - 9.9g, are thus predominately females. In June of the following year, some spent females change sex again into males. Most do not, however, because older weight classes still are firmly female sex-biased. LIFE HISTORY TACTICS From the evidence before us of changes in the popula- tion structure and sexual strategy of Saccostrea cucullata at 20 OCT. 88 Frequency “lo Total AMER. MALAC. BULL. 8(1) (1990) 1987 1986 weight (g.) Fig. 2. The population structure of Saccostrea cucullata at Ting Kok mangrove, Tolo Harbour, Hong Kong over the period from October 1988 to October 1989. Dates above peaks indicate assessed years of recruitment. Ting Kok, and with associated hydrological data, a picture of the life history can be proposed. S. cucullata juveniles are recruited into the adult population in mid-summer when sea water temperatures are consistently > 30°C and salinities could be lowered due to summer rains. It is unknown if spawn- ing and recruitment occur as a distinct series of events related to the lunar cycle as is seen in other oysters (Walne, 1974; Andrews, 1979). They grow rapidly to a weight of about 5g and then growth ceases over the autumn period. Gameto- genesis begins in December, primordia developing some six months after settlement. Most juveniles mature into males (77.6%). MORTON: SACCOSTREA LIFE HISTORY 5 In late winter-early spring, when temperatures are ris- ing from their low of ~ 17°C in January and salinities are generally high, juveniles begin to grow again, their gonads maturing simultaneously, and mature and spawn in early to mid-summer at a weight of ~ 7.5g. Many of the spent males develop female gonads in June, so that a high percentage of hermaphrodites is recorded (27.1%) and a majority of the population, by a weight of 8.0 - 10.0g, is female (55.2%). It is perhaps significant that sea temperatures rose dramatically from 23°C to 33°C in May and salinities fell slightly over the same period from 30.5%/g9 to 28°/o9. The second year female oysters, now covered by a mass of settling juvenile males, overwinter, their gametes developing slowly. Somatic growth begins again in December-January and the females are mature by May. At this time, they spawn with the eggs fer- tilized externally by the now similarly mature males. In June of the following year, at a weight of ~ 12g, some of the spent females change sex into males so that in this month in- dividuals with a weight between 10.0 - 16.0g are maturing male and spent female hermaphrodites. Such individuals then enter their third year of life, overwintering with little growth and beginning to mature, either into males or females, in early spring. Shell growth begins again, later, in March to a modal weight of about 15g. Only occasionally were such individuals collected. For most individuals in the population, therefore, the life span is two years in which they are first predominate- ily male and then predominately female with the sex change occurring in June of each year. DISCUSSION The early research of W. R. Coe established the framework for our understanding of sexual strategies employed by the Bivalvia, notably the oysters (Coe, 1943). Coe’s review of oysters of the genus Crassostrea, i.e. C. virginica and C. gigas, demonstrated alternative sexuality with annual changes in sex. Amemiya (1929) showed for C. gigas that 25% of the females and 60% of the males in his experi- ment reversed their sex during one winter. Andrews (1979) has shown that about 96% of first year individuals (35 mm) of C. virginica are males with the reverse % in older oysters. Galtsoff (1964) estimates that about 10% of mature individuals of this species change sex annually. Roughley (1933) concluded that the Australian oyster, Ostrea commercialis (= Saccostrea cucullata) spawned first as males, though there appeared to be 2.7 times as many females as males in the population. Dinamani (1974) investi- gated the rock oyster Crassostrea glomerata (= S. cucullata) (Stenzel, 1971; Dinamani, 1976) and showed a similar situa- tion with reproduction occurring over a single period in sum- mer (January and February in southern hemisphere New Zealand) and with the majority of oysters of age one year and below male, and a greater percentage older than one year, female. Dinamani (1974) also gives data on gametogenesis of this oyster and provides photomicrographs of the gonads at various stages of development. Literature on reproduction in the oysters has been reviewed by Andrews (1979). The situa- tion seen in S. cucullata from Hong Kong conforms to the pat- tern established for C. virginica and other oviparous oysters of this genus, although the number of spawning periods evi- dent in any one year varies with species and location. C. madrasensis, for example, spawns twice in spring and autumn, in any one year in India (Stephen, 1980). Most studies of oysters have used cultivated individuals or those growing upon rocks; little information is available for Indo-Pacific species in the mangrove component of their habitat continuum. Available information has been reviewed by Morton (1983). The western Atlantic mangrove oyster NO GROWTH, 3rd year a REVERSAL p 3 ret Fig. 3. The pattern of growth of Saccostrea cucullata at Ting Kok mangrove, Tolo Harbour, Hong Kong, as determined from the monthly weight frequency histograms (Fig. 2). Superimposed upon these data is the predominant sex of each year class and the monthly stage of sexual 20 NO GROWTH GROWTH 15 aD = ter alu 3 3 eos } 1988 1689 maturity (1-5), as obtained from Table 1. 6 Table 1. Percentage incidence and stage of gonadal development of immature individuals (¢) of Saccostrea cucullata, males (0), females (1) ) and mean gonad condition (1-5) of individuals ranging in total weight from 1.0-1.9 g - 16.0-17.9 g obtained at monthly and hermaphrodites ( AMER. MALAC. BULL. 8(1) (1990) intervals from October 1988 to October 1989 from the Ting Kok mangrove, Tolo Harbour, Hong Kong. Figures at the bottom of each column represent the percentage incidence of immature individuals, males, females and hermaphrodites in that size category. Total weight (q)) redominant . Pi *% Incidence Sexual stage Month cs ey eo SE Ee 10-119 12-129 14-159 [16-179 [e | JO] Selo] ¢| oe 1988 | Gm @ sieig| d@@oOQR@o| © OW 37 1592)370| - | 5 |5.1| - me aus ra eer MOCO Ee co 26 658)316) - | 5 5.1 : Dec. StmOoMoM Oe bee GOO jua mes ne son 19991 PEO BPO $9 ES eehlaise o snss00fe4 ae Feb "S"| GS Seeladelsoolaat| ee [© | © forgesias| - | 3 [2] - BERS EAS|e| wo] | w etabnl- faba sa formenonmolacrle ne OE] OOOUS hes ioe awe ze dividuals Jun Peserse ee ae of ese - lspahosizrife-s|« AW4ace4 RPS O Be aa Be [mm || ba pwberpal «(| eae Aug ood foe Oe | © He | Z00R60K40) - [45 ae se. OS Oe eee | Ge lee | [3278 - f-5|¢8) Oct eee) ee OO ORISA 09 [oo 14 [436 4-5]4-5) - |e! 155 | 103 | 02 | 86 | 69 | wo | w2 | - | - /103/s0s369 “yo | 776 | 621 | ero | 483 | 379 | 360 | 333 | 2 | 222 x “9 69 | 24 | 254 | 414 | 552 | 460 | 513 | 625 Key: @ = Immature individuals Co) a Eee ele ealcr eer ot come Crassostrea rhizophorae was studied by Nascimento et al. (1980) in Brazil who showed that this species was also predominately male in its first year, with sex reversal occur- ring later. In its second year, C. rhizophorae achieves a shell length of 40mm, which is of marketable size. C. rhizophorae, however, breeds over the winter months from late summer (August - October) to March at a time of rising salinities (Bacon, 1970; Rojas, 1972; Wedler et al., 1978). The West African oyster, Crassostrea gasar (Adanson), in Sierra Leone, reproduces in winter (Hunter, 1969). In south west India, C. madrasensis (Preston) has two peaks of spawn- ing; a major one occurring from April - June with a second minor peak in September - October. This sub-tidal species <} =Hermaphrodites 1-5 =Stage of sexual development is separated in the Indian Sub-continent from the intertidal Saccostrea cucullata (Ansari and Ahmed, 1972). It is possible that in the Indo-Pacific there is a ‘super- species’ of intertidal oyster generally referred to as Saccostrea cucullata (Stenzel, 1971) but with a wide variety of geographic names attached to it. Lasiak (1986) has reviewed the available information on ‘‘S. cucullata’’ from seven locations throughout its broad range and shown that reproductive activity is con- tinuous in tropical populations in Singapore (Ling, 1970), India (Awati and Rai, 1931; Nagabhushanam and Birdarkar, 1977), Pakistan (Asif, 1980) and Guam (Braley, 1982). Sub-tropical or more temperate populations as for example in Australia (Roughley, 1933) and Southern Africa (Lasiak, 1986), however, MORTON: SACCOSTREA LIFE HISTORY i have a single annual cycle with spawning taking place in early summer. In Hong Kong, S. cucullata conforms to this latter pattern. Lasiak (1986) could detect only a few hermaphrodites (3 out of 333 individuals sectioned) in the population of S. cucullata from South Africa. In Hong Kong, the species is oviparous and exhibits alternative sexuality, a percentage of the population changing sex each year. Spawning and recruitment in the Bivalvia are sequen- tial but both are influenced by a variety of hydrological parameters. In the absence of confirmatory studies that ali these oyster ‘species’ are synonymous and can be categor- ized under the name Saccostrea cucullata, caution must be exercised in comparing the results obtained in Hong Kong with those obtained from elsewhere. However, a few generalizations can be made. The gonadal cycle of S. cucullata appears locally to be linked to temperature changes, as with this species in South Africa (Lasiak, 1986) but with this clearly not the case in tropical populations, e.g. Singapore (Ling, 1970). In tropical populations too, spawning can be in- duced by salinity reductions associated with heavy monsoon rains (Ling, 1970; Nagabhushanam and Birdarkar, 1977; Stephen and Shetty, 1981) but this is not the case either in South Africa (Lasiak, 1986) or in Hong Kong where local changes in salinity in the mangrove habitat (at least in the, admittedly dry, year of 1989) following rainfall are not so dramatic. Saccostrea cucullata occurs locally in a variety of subhabitats and clearly dominates the seaward mangrove habitat, encrusting stems, roots and pneumatophores. Of local ecological interest could be an answer to the question: how is this achieved? What aspects of the life history strategy of this species equip it for habitat dominance? In this context, Morton (1987, 1990b) has reviewed the life history tactics of southern Chinese fresh and brackish water bivalves and the sexual strategies of a wide spectrum of local bivalves. Saccostrea cucullata locally fits into the r end of an r-K continuum (MacArthur and Wilson, 1967; Pianka, 1970), be- ing highly opportunistic. The species grows fast and achieves sexual maturity in its first year of life. Growth of newly recruited juveniles halts over winter, but with the approach of spring at a time of high salinities and rising temperatures, somatic growth and gonad maturation proceed concurrently. This could be facilitated by the high levels of nutrients in the mangrove habitat (Fig. 1C). Juveniles predominantly mature into males as Morton (1988, 1990a, b) has shown for two other, albeit dioecious, seaward mangrove bivalves, i.e. Brachidontes variabilis and Gafrarium pectinatum. These species, too, have a male-biased juvenile sex ratio, which, it is thought, pro- gressively changes through enhanced juvenile mortality into a female dominance in subsequent year classes. Like S. cucullata, these species live for two to three years with the majority of all individuals in all species rarely living beyond two years. The population histograms for S. cucullata suggest that the species can live for greater than three years, but on- ly a few achieve this, successive recruitments possibly suf- focating older individuals as new recruits settle upon them. Thus in terms of the Bivalvia, local seaward mangroves are colonized by a guild of r-strategists, Saccostrea cucullata and Brachidontes variabilis being sympatric, the latter epibyssate on the former. Such a situation is different from the landward mangrove species, Polymesoda (Geloina) erosa, which occupies the K end of the r-K continuum, being long- lived, dioecious and with an overall female biased sex ratio (Morton, 1985, 1990b). In subtropical Hong Kong, all four mangrove associates breed in summer, Polymesoda and Saccostrea in a single phase in mid-summer, Brachidontes and Gafrarium in two phases in early summer and late summer/autumn. All species are tropical/subtropical, but it is thought possible that reproduction in Brachidontes and Gafrarium is divided into two phases by low summer salinities inhibiting either gametogenesis or spawning or both (Morton, 1987). In this context, Morton and Chan (1989) have shown that Polymesoda erosa is oligohaline and Gafrarium pec- tinatum, marine stenohaline, the latter, but not the former, therefore clearly responding dramatically to salinity fluctua- tions. Saccostrea cucullata and Brachidontes variabilis are, however, marine euryhaline and thus osmoconform to a wide range of salinities. A combination of morphological, behavioural, physio- logical and reproductive adaptatioris thus allow such bivalves to exploit the high-zoned mangrove habitat, throughout much of the Indo-Pacific. Such adaptations are reflected in life history tactics with the numerically dominant species, Saccostrea cucullata, opportunistically colonizing the lower components of the mangrove habitat. ACKNOWLEDGMENTS lam grateful to Mr. M. H. Lam and Ms. Lisa Ng for assistance with analysis of the samples and Mr. S. F. Leung for histological assistance. LITERATURE CITED Amemiya, |. 1929. On the sex change of the Japanese common oyster, Ostrea gigas Thunberg. Proceedings of the Imperial Academy of Tokyo 5:284-286. Andrews, J. D. 1979. Pelecypoda: Ostreidae. In: Reproduction of Marine Invertebrates, Vol. V. Molluscs: Pelecypods and Lesser Classes. A. C. Giese and J. S. Pearse, eds. pp. 293-341. Academic Press Inc., New York. Ansari, F. and M. Ahmed. 1972. Seasonal gonadal changes in the oyster Crassostrea glomerata Gould. Pakistan Journal of Zoology 4:35-43. Asif, M. 1980. The reproductive cycle in the population of Saccostrea cucullata (Born) from the coast of Karachi, Pakistan. Hydrobiologia 68:73-79. Awati, P R. and H. S. Rai. 1931. Ostrea cucullata (the Bombay oyster). Indian Zoological Memoirs 3:1-107. Bacon, P. R. 1970. Studies on the biology and cultivation of the mangrove oyster in Trinidad with notes on other shellfish resources. Tropical Science 12:265-278. Braley, R. D. 1982. Reproductive periodicity in the indigenous oyster Saccostrea cucullata in Sasa Bay, Apra Harbour, Guam. Marine Biology 69:165-173. Branch, G. M. and J. R. Grindley. 1979. Ecology of southern African estuaries. Part XI. Mngazana: a mangrove estuary in Transkei. Suid-Afrikaanse tydskrif vier Dierk 14:149-170. 8 AMER. MALAC. BULL. 8(1) (1990) Coe, W. R. 1943. Sexual differentiation in mollusks. 1. Pelecypods. Quarterly Review of Biology 18:154-164. Day, J. H. 1974. The ecology of Morrumbeene estuary, Mocambique. Transactions of the Royal Society of South Africa 41:43-97. Dinamani, P. 1974. Reproductive cycle and gonadial changes in the New Zealand rock oyster Crassostrea glomerata. New Zealand Journal of Marine and Freshwater Research. 8:39-65. Dinamani, P. 1976. The morphology of the larval shell of Saccostrea glomerata (Gould, 1850) and a comparative study of the lar- val shell in the genus Crassostrea Sacco, 1897 (Ostreidae). Journal of Molluscan Studies 42:95-107. Dudgeon, D. and B. Morton. 1983. The population structure and sexual strategy of Anodonta woodiana (Bivalvia: Unionacea) in Plover Cove Reservoir, Hong Kong. Journal of Zoology, London 201:161-183. Galtsoff, P S. 1964. The American oyster, Crassostrea virginica Gmelin. United States Fisheries and Wildlife Service. Fisheries Bulletin 64:1-480. Hunter, J. B. 1969. A survey of the oyster population of the Freetown estuary, Sierra Leone, with notes on the ecology, cultivation and possible utilization of mangrove oysters. Tropical Science 11:276-285. Lasiak, T. 1986. The reproductive cycles of the intertidal bivalves Crassostrea cucullata (Born, 1778) and Perna perna (Linnaeus, 1758) from the Transkei Coast, Southern Africa. Veliger 29:226-230. Ling, Y. Y. 1970. Studies on certain aspects of the biology of the local oyster Crassostrea cucullata (Born) in Singapore waters in rela- tion to its cultivation. Master’s Thesis, University of Singapore. 184 pp. MacArthur, R. H. and E. O. Wilson. 1967. The theory of island biogeography. Monographs in Population Biology 1:1-203. Macnae, W. and M. Kalk. 1969. A Natural History of Inhaca Island Mocambique. Witwaterstand University Press, Johannesburg. 163 pp. Morris, S. 1985. Preliminary guide to the oysters of Hong Kong. Asian Marine Biology 2:119-138. Morton, B. 1982a. Some aspects of the population structure and sex- ual strategy of Corbicula cf. fluminalis (Bivalvia: Corbiculacea) from the Pearl River, People’s Republic of China. Journal of Molluscan Studies 48:1-23. Morton, B. 1982b. The reproductive cycle in Limnoperna fortunei (Dunker, 1857) (Bivalvia: Mytilidae) fouling Hong Kong’s raw water supply system. Acta Oceanologia et Limnologia Sinica 13:312-324. Morton, B. 1983. Mangrove bivalves. In: The Mollusca. Vol. 6, Ecology. Wilbur, K. M. and W. D. Russell-Hunter, eds. pp. 77-138. Academic Press, New York. Morton, B. 1985. The reproductive strategy of the mangrove bivalve Polymesoda (Geloina) erosa (Bivalvia: Corbiculacea) in Hong Kong. Malacological Review 18:83-89. Morton, B. 1987. Comparative life history tactics and sexual strategies of the fresh and brackish water bivalve fauna of Hong Kong and southern China. American Malacological Bulletin 5:91-99. Morton, B. 1988. The population dynamics and reproductive cycle of Brachidontes variabilis (Bivalvia: Mytilidae) in a Hong Kong mangrove. Malacological Review 21:109-117. Morton, B. 1990a. The life cycle and sexual strategy of Gafrarium pec- tinatum (Bivalvia: Veneridae) in a Hong Kong mangrove. Malacological Review (in press). Morton, B. 1990b. Do the Bivalvia demonstrate environment-specific sexual strategies? A Hong Kong model. Journal of Zoology, London (in press). Morton, B. and K. Y. Chan. 1989. The salinity tolerances of four species of bivalves from a Hong Kong mangrove. In: Pro- ceedings of the Second International Marine Biological Workshop: the Marine Flora and Fauna of Hong Kong and southern China, Hong Kong. B. Morton, ed. pp. 1115-1122. Hong Kong University Press, Hong Kong. Morton, B. andR. S. S. Wu. 1975. The hydrology of the coastal waters of Hong Kong. Environmental Research 10:319-347. Nagabhushanam, R. and D. S. Bidarkar. 1977. Reproductive biology of the Indian rock oyster Crassostrea cucullata. Indian Jour- nal of Fisheries 24:135-142. Nascimento, |. A., E. M. da Silva, M. J. S. Ramos, and A. E. dos Santos. 1980. Development of the primary gonad in the mangrove oyster Crassostrea rhizophorae, age and length at first spawning. Ciencia e Cultura 32:736-742. Pianka, E. R. 1970. On r- and K-selection. American Naturalist 104:595-597. Rojas, A. V. 1972. Fijacion de la larva le la ostra da los bancos naturales de Bahia de Mochima y Laguna Grande. Boletin del Instituto Oceanografico, Universidad de Oriente, Cumana 11:97-106. Roughley, T. C. 1933. The life history of the Australian oyster, (Ostrea commercialis). Proceedings of the Linnean Society of New South Wales 58:279-332. Stenzel, H. B. 1971. Oysters. Part N . Bivalvia 3(3). /n: Treatise on In- vertebrate Paleontology, R. C. Moore, ed. pp. N953-N1224. The Geological Society of America Inc. and the University of Kan- sas Press. Stephen, D. 1980. The reproductive biology of the Indian oyster Crassostrea madrasensis (Preston). Aquaculture 21:139-146. Stephen, D. and H. P. C. Shetty. 1981. Induction of spawning in five species of bivalves of the Indian coastal waters. Aquaculture 25:153-159. Stojkovich, J. O. 1977. Survey and species inventory of representative pristine marine communities on Guam. Technical Report University of Guam Marine Laboratory 40:1-183. Torigoe, K. 1981. Oysters in Japan. Journal of Science of the Hiroshima University Series B, Division 1 (Zoology) 29:291-347, 36 plates. Torigoe, K. and A. Inaba. 1981. On the scientific name of Japanese spiny oyster ‘‘kegaki’’. Venus 40:126-134. 2 plates. Walne, P. R. 1974. Culture of Bivalve Molluscs: 50 years’ Experience at Conway. Fishing news (Books). Wedler, E., L. Perez, and J. Palacio. 1978. Ostricultura en la Ciénaga de Santa Marta Primera etapa. /nformacioon de un Proyecto especial de Colciencias. Santa Marta, Columbia. 1-64 p. Date of manuscript acceptance: 9 March 1990. ECOLOGICAL ANALYSIS OF THE LIVING MOLLUSCS OF THE TEXAS PANHANDLE RAYMOND W. NECK TEXAS PARKS AND WILDLIFE DEPARTMENT 4200 SMITH SCHOOL ROAD AUSTIN, TEXAS 78744, U.S.A. ABSTRACT The living molluscan fauna of the Texas Panhandle includes a total of 43 species: nine freshwater gastropods, 24 terrestrial gastropods (including one slug), and ten freshwater bivalves. Except for one terrestrial gastropod and one bivalve, all of these species are native to the central United States. Five freshwater and seven terrestrial gastropods, and six freshwater bivalves are here reported living in the Texas Panhandle for the first time. One terrestrial gastropod is a new record for Texas. Occurrence of freshwater gastropods and pelecypods is limited by quality and quantity of surface waters, whereas terrestrial gastropods are limited by distribution of surface soil moisture and cover objects. Highest diversity terrestrial faunas are associated with woodlands of plains cottonwood, Populus sargentii, but the mesic-adapted species can be found in moist floodplains even in the absence of trees. Until recently, the living molluscan fauna of the Texas Panhandle has been poorly known. General biological interest in this area has been limited in the past, at least partially be- cause of a presumed low species diversity. The Texas Pan- handle presents an extreme environment for both terrestrial and freshwater molluscs. The general aridity of the climate is made even more rigorous for molluscs by wide seasonal thermal extremes. For this study, the Texas Panhandle was defined as that part of Texas between the 100th and 103rd meridians and north of 34°18’40’’N, corresponding to a series of county boundaries between Cottle/Childress on the eastern end and Bailey/Parmer on the western end (Fig. 1). STUDY AREA Two major physiographic regions exist in the Texas Panhandle. The greater part of the area is within the High Plains where Quaternary eolian deposits overlie the Ogallala Formation of the Miocene and Pliocene. The High Plains is a nearly level plain rising gradually to the west. Potential natural vegetation for most of the Texas Panhandle is short and mixed grass prairies as observed by early European ex- plorers (Strout, 1971; Flores, 1984). Woody plants were few in species number and were generally restricted to canyons and steep slopes (Palmer, 1920). The eastern portion of the Panhandle is the second physiographic region and contains a mosaic of two units: 1) Permian redbeds or ‘‘badlands’’ with limited soil development and a mixed grassland/scrub community; 2) Quaternary alluvial plains with deep soils that naturally supported mixed grass prairie communities. The Permian badlands extend westward along the Canadian River. Gallery woodlands exist on Quaternary alluvial deposits along the larger water courses. Portions of the northeastern and southwestern Panhandle contain sand hills and ridges overlying the Ogallala and sup- port a mixed grass/herbaceous prairie community or a stunted woodland community. Analysis of the herpetofauna and woody flora reveals a mixed biota with regard to biogeographical affinities of the constituent species. Species present include those char- acteristic of central prairies, southern savannahs and brush- lands, and western montane areas along with a few represen- tatives of eastern forests (Palmer, 1920; Fouquette and Lind- say, 1955). The Texas Panhandle has a warm temperate, semiarid, continental climate. Mean annual temperature at Amarillo is 139°C, with record extremes ranging from —26.7°C to 41.1°C. The growing season averages 190 days from 17 April to 24 October. Summer days produce high temperatures (64 days with maximum above 32.2°C), but radiational cooling results in cool nights and early morning hours. Winter weather is characterized by strong cold fronts with occasional blizzards, but these fronts are usually followed by a warming trend after several days. Freezing temperatures occur on 110 days an- nually. Average annual precipitation is 485 mm with a record of 1110 mm in 1923. Most precipitation occurs during thunderstorms, which occur on an average of 49 days annual- ly. The above climatic data are taken from the Natural Fibers American Malacological Bulletin, Vol. 8(1) (1990):9-18 9 10 AMER. MALAC. BULL. 8(1) (1990) OKLAHOMA i fervor en 8 ' CARSON i DEAF SMITH ' RANDALL ‘ARMSTRONG DONLEY 7 NEW MEXICO WORTH i 1 lio 54 ee ee ye ee | ee | | I \ \ BRISCOE PARMER lensaee 1 | ' 91 | Tewiswer | | BAILEY (awe eae I rLovo MOTLEY ' I 1 | | | | ' 1 ' i TEXAS PANHANDLE ° 25 50 75 100 t N Scale in km Fig. 1. Panhandle of Texas giving collection localities referred to in text. Information Center (1987). The Texas Panhandle is drained by the Red/Canadian drainages of the Mississippi River system except for the ex- treme southwestern corner which is drained by Running Water Draw of the Brazos River drainage. Flowing waters in this region are quite restricted in width, depth, and permanence. Large streams have low flows with high concentrations of suspended red silt/clay and dissolved salts. Such water con- ditions are natural (E. James /n: Thwaites, 1905), but modern land use practices have further decreased the water quality of these streams. Water flows in these larger streams have been described as being ‘‘either dry or raging current’’ (Gould, 1906:42). Small feeder creeks are generally intermittent but a few creeks have perennial flows supplied by ‘‘sweetwater’’ springs supplied by the Ogallala Aquifer. Natural lakes are limited to the numerous playas on the High Plains surface (Guthery and Bryant, 1982). LITERATURE REVIEW Neck (1984) published the results of a survey of living terrestrial gastropods of various canyons of the Eastern Caprock Escarpment. Most published records (Henderson, 1909; Strecker, 1910; Walker, 1915; Clarke, 1938) were rejected as being samples of flood debris containing shells of undeter- minable, probably fossil, origin. Only a few published records of gastropods from this area were accepted as representing modern living populations (Pratt, 1965; Metcalf /n: Franzen, 1971; Metcalf /n: Bequaert and Miller, 1973). The opinions of Neck (1984) and the records accepted therein produced a list of 12 species of terrestrial gastropods known to be living along the Eastern Caprock Escarpment; two additional species were known from other areas of the Texas Panhandle. A subsequent review of terrestrial gastropods of the eastern United States (Hubricht, 1985) included several new records from the Texas Panhandle. The reports of Vallonia pulchella (Muller), V. cyclophorella (Sterki), aiid V. perspec- tiva Sterki from Potter and Randall counties (Hubricht, 1985:63, 65, 67) probably represent reworked fossil shells or introduced populations (Pilsbry, 1948; Bequaert and Miller, 1973; Pierce, 1975; Neck, 1989a) and are not accepted as native members of the living Panhandle fauna. Only limited records of urban molluscs of the Texas Panhandle are available (Neck, 1989a). Hubricht (1985) reported Succinea forsheyi |. Lea and S. indiana Pilsbry from Hemphill and Sher- man counties, respectively, although taxonomic uncertainty prevents complete udnerstanding of species relationships of the succineid snails. The Sherman County record was originally referred to S. vaginacontorta Lee by Franzen (1971), a taxon synonymized with S. indiana by Hubricht (1985). Records of freshwater gastropods from this area are also rare (Fullington, 1978), although Physella sp. and Planorbella sp. were observed in playas as early as 1876 (E. H. Ruffner /n: Baker, 1985). Clarke (1938) reported Physella virgata (Gould) living in ‘“‘Paladura Creek [northeast] of Canyon,’ which is actually the Palo Duro Canyon of the Prairie Dog Town Fork of the Red River. Singley (1893) reported P virgata and Planorbella trivolvis (Say) from a playa. Strecker (1910) reported P virgata and P. tenuis (Dunker) from creeks and tanks in Armstrong County. A survey of playa lakes of the Llano Estacado of Texas and New Mexico (Sublette and Sublette, 1967) revealed the presence of P trivolvis in five of 12 playas surveyed in the current study area. Rhodes and Gar- cia (1981) reported ‘‘snails (mostly Physa sp. and Lymnaea sp.)’’ from all eight playas that they surveyed from Castro and Swisher counties. Neck (1982) reported P virgata and P trivolvis from Lake Theo, Briscoe County. Several freshwater gastropods have been reported from playas in Castro County (Neck and Schramm, in press). Freshwater bivalves are less widely distributed in the Texas Panhandle than are freshwater gastropods. Clarke (1938) reported living Sphaerium striatinum (Lamarck) [as S. sulcatum (Lamarck)] and Uniomerus tetralasmus (Say) from the Prairie Dog Town Fork of the Red River, Randall County. The occurrence of U. tetralasmus in a stream of variable volume is related to the ability of this species to withstand complete desiccation of its habitat. Neck (1982) reported S. striatinum and an introduced population of Anodonta grandis Say in Lake Theo, an artificial impoundment on Holmes Creek, Caprock Canyons State Park, Briscoe County. Popula- tions of the widespread exotic Asian clam, Corbicula fluminea (Miller), occur in Greenbelt Reservoir, Donley County, and Lake Meredith, Hutchinson County (Neck, 1987b). A single record of S. striatinum from a playa is known (Neck and Schramm, in press). METHODS AND RESULTS Soil or benthic samples were either obtained by the author or received from other field workers as noted. Mollusc NECK: TEXAS PANHANDLE MOLLUSCS 11 Table 1. Occurrence of living nonmarine molluscs in the Texas Panhandle (includes both published reports and reports herein): O = previously published reports of living terrestrial accepted by Neck (1984); A = records in Singley (1893); B = Strecker (1910); C = Clarke (1938); S = Sublette and Sublette (1967); F = Fullington (1969, 1978); N2 = Neck (1982); N4 = Neck (1984); H = Hubricht (1985); N7 = Neck (1987b); NS = Neck and Schramm, in press; U = urban snails in Neck (1989a); single numbers refer to locality numbers in text; taxa are presented in phylogenetic sequence within each group. Species O A BC S F N2N4 H N7 NS U 123 4 5 6 7 8 9 10 11 12 13 FRESHWATER GASTROPODS Hebetancylus excentricus x Fossaria cockerelli x xX Fdalli X F. modicella Xx F. parva X Physella virgata X X X X Gyraulus parvus X X X Planorbella tenuis X X P trivolvis Xx Xx x Xx TERRESTRIAL GASTROPODS Vallonia gracilicosta xX V. parvula Xx Pupoides albilabris Gastrocopta cristata G. pellucida G. procera G. armifera xX G. abbreviata X 4 G. tappaniana Xx Vertigo ovata Xx Succinea forsheyi Xx X S. indiana Xx Xx Catinella cf. avara xX Xx Oxyloma retusa X Helicodiscus parallelus x x X H. singleyanus xX X X H. inermis X X xX X X Xx Deroceras laeve X x X Xx xX xX Hawaiia minuscula Xx Zonitoides arboreus x Euconulus trochulus X X X X Rumina decollata xX Rabdotus dealbatus Xx xX X X Stenotrema leai Xx X BIVALVES Anodonta grandis xX X X X A. imbecilis xX Uniomerus tetralasmus xX Xx Sphaerium striatinum X X X X Musculium transversum x Pisidium casertanum xX x P compressum X P. nitidum x P. punctiferum Xx Corbicula fluminea X xX x< x KK KK OK OK x KK XK x< ; ; as * A : “oo ee _ . . > Js = i. A , * Sa? ly 73 vv ‘ , q * é* - - & © ps ng ~ %.3* % . aS s* i *—, * *t | Sees 2 ¥ A all . vo be i . ae ‘ ty Fig. 4. Comparison of nucleolar organizer regions (NORs) in three Gymnosomata species (scale bar = 5 um). (A) silver-stained mitotic metaphase of Pneumodermopsis canephora (arrows point to NORs), (B) silver-stained mitotic metaphase of Pheumoderma atlanticum (arrows point to NORs), (C) Giemsa-stained mitotic metaphase of Pheumodermopsis paucidens (arrows on pair 2), (D) same metaphase, silver-stained (arrows on NORs). the pair 2 of the female could be called XX. The remaining pairs in the male karyotype are similar to the female except the absence of the smallest autosome pair. PTEROTRACHEA HIPPOCAMPUS One male could only be investigated in this species. Sixteen well-spread metaphases were analysed with a com- plement set of 2n=31. The karyotype (Fig. 7) shows heteromorphic chromosomes, comprising a large metacen- tric X, a telocentric Y; and a minute chromosome Yo, and 14 autosome pairs with two metacentric, two submetacentric, one subtelocentric-submetacentric, six subtelocentric and three telocentric. FIROLOIDA DESMARESTI In this species, several individuals of females and males were studied. But only two metaphasic cells in females and four in males could be karyotyped because of overlap- ping chromosomes. The overall large absolute chromosome size (Fig. 8A, B) of this species is probably the reason for this overlapping. The female diploid set consists of 2n=32 while the male shows 2n=31. In both sexes, the first pair is a large metacentric. Next, the second metacentric pair in the female corresponds to XX chromosomes while in the male sex heteromorphic chromosomes are observed comprising a large metacentric X, a large submetacentric-subtelocentric Y,; and a microchromosome Y>2. The remaining autosome pairs in the THIRIOT-QUIEVREUX: KARYOTYPE ANALYSIS 41 Table 3. Chromosome measurements and classification in six metaphasic cells of Paraclione longicaudata. Chromosome Relative length Arm ratio Centromeric index Classifica- pair no. Mean SD Mean SD Mean SD tion* 1 15.15 0.725 0.877 0.055 46.65 1.554 m 2 11.78 0.864 0.751 0.119 42.07 3.642 m 3 8.17 0.497 0.733 0.121 41.47 3.315 m 4 7.99 0.346 0.227 0.025 18.44 1.645 st 5 7.45 0.352 0.265 0.112 20.32 6.303 st 6 7.04 0.593 0.768 0.115 42.26 3.444 m th 6.59 0.439 0.366 0.070 26.45 3.667 sm 8 5.82 0.512 0.165 0.081 13.79 5,809 st 9 5.11 0.272 0.266 0.113 20.03 6.663 st 10 4.67 0.285 0.233 0.102 18.30 6.660 st 11 4.36 0.396 0.270 0.092 20.78 5.006 st 12 4.10 0.092 0.195 0.094 16.07 6.287 st 13 3.91 0.466 0.510 0.097 32.73 3.035 sm 14 3.85 0.542 0.363 0.202 24.84 10.174 st-sm 15 2.55 0.233 0.088 0.021 8.06 1.809 t 16 1.61 0.425 0.141 0.721 11.65 5,230 t * 16.00 Relative length Pneumodermopsis paucidens Paraclione longicaudata 10-11 t2 Chromosome pairs (] submetacentric [] sm-st, st-sm 13 14 15 16 HZ metacentric CL) subtelocentric J telocentric Fig. 5. ldeogram of Pneumodermopsis paucidens and Paraclione longicaudata, constructed from relative length and centromeric in- dex values. Chromosome morphology is indicated: dark stippled for metacentric, medium stippled for submetacentric, light stippled for submetacentric-subtelocentric or subtelocentric-submetacentric, open for subtelocentric, lozenge stippled for telocentric. m = metacentric; st = subtelocentric; sm = submetacentric; st-sm = subtelocentric-submetacentric; t = telocentric. female are four metacentric, four submetacentric, five subtelocentric and two telocentric. In the male, the smallest autosome pair is lacking. DISCUSSION THECOSOMATA The diploid number of 2n=28 observed in Hyalocylis Striata is different from the number found by Zarnick (1911). This difference in counts for the same species may result from the difficulty in evaluating chromosome number using older cytological techniques. Comparing data obtained with the same techniques, however, points out that H. striata shows the highest diploid number among Euthecosomata. Karyological data for all species studied in a previous paper (Thiriot-Quiévreux, 1988) and in the present work are recapitulated in table 4. The chromosome morphology of Limacina inflata (Orbigny), Peraclis reticulata (Orbigny) and Cymbulia peroni De Blainville is given following the nomenclature of Levan et al. (1964) after chromosome measurements. The chromosomal features among Thecosomata could be assigned the following trends: (i) from low to high diploid number (2n=20 to 2n=28 in Euthecosomata, and 2n=24 to 34 in Pseudothecosomata); (ii) towards an increasing propor- tion of subtelocentric-telocentric chromosomes (zero to five pairs in Euthecosomata, and one to six Pseudothecosomata). These features are unusual among Opisthobranchia. First, stable chromosome number within a family, or even an order, has been reported in many Opisthobranchia (Patter- son, 1969). However, chromosome number variability of 2n=26 to 2n=34 have been observed within the Cephalaspidea order (Curini-Galletti, 1985, 1988), and of one bivalent within the Aplysiidae family (Vitturi et a/., 1982b) and 42 AMER. MALAC. BULL. 8(1) (1990) G ip 2 A HK je ue KE AE ° fee | ed i@ «kh Ke €h QR Sn Pr cn | Fig. 6. Karyotypes of Pterotrachea coronata (scale bar = 5 um), (A), female, (B) male. Ly KX XR AA Ar rT OK AA AA — an Ah AA 2X reary: 12 3 14 Fig. 7. Male karyotype of Pterotrachea hippocampus (scale bar = 5 um). two Soleolifera families (see references in Patterson, 1969). Thus, the chromosome variability occurring among Thecosomata is especially large. Secondly, among the few karyological data recorded among Opisthobranchia, isobrachial chromosomes (i.e. metacentric or submetacentric) of a relatively homogeneous size are dominant in karyotypes (Curini-Galletti, 1988). Hence, Thecosomata differ greatly from the known Opisthobranchia. Moreover, if we assume that isobrachial chromosomes are plesiomorphic as suggested by Curini-Galletti (1988), the karyotypes of Limacina inflata show- ing only metacentric chromosomes should be considered as the most primitive. This point corresponds to the general agreement that the spirally coiled genus Limacina is the most primitive of the Euthecosomata (Lalli and Gilmer, 1989). The evolutionary cytotaxonomy within Thecosomata could be from Limacina inflata with low diploid number and metacentric chromosomes to Cymbulia peroni with high diploid number and subtelocentric-telocentric chromosomes. The Thecosomata constitutes an isolated order with no karyological relationship with the other opisthobranch orders and speculation about which ancestral molluscs might have given rise to the Thecosomata cannot be made. GYMNOSOMATA The two species studied in this paper share the same diploid number (2n=32) as the species previously investigated (Thiriot-Quiévreux, 1988). This supports a chromosome number stability within Gymnosomata. This number is close to the opisthobranch ancestral number of 2n=34 suggested by Schmeckel (1985). Pheumodermopsis paucidens reveals contradictory evolutionary features showing a plesiomorphic character (according to Curini-Galletti, 1988) with its highest proportion of isobrachial chromosomes, but also an apo- morphic character (Thiriot-Quiévreux, 1988 according to the hypothesis of Gold and Amemiya, 1986) with the presence of NORs on a metacentric chromosome. Paraclione longicaudata differs from the other species by its striking variability in relative length of the different chromosomes and by its high number of subtelocentric chromosomes. Therefore, karyological features among Gymnosomata are Characterized by modal chromosome number of 2n=32 and a variable proportion between metacentric- submetacentric and subtelocentric-telocentric chromosomes which, at present, preclude evolutionary implications. In conclusion, karyological analyses of the two orders, Thecosomata and Gymnosomata, lend support for a biphyletical origin, i.e. different opisthobranch ancestors, as generally accepted (see Van der Spoel, 1967, and Lalli and Gilmer, 1989). HETEROPODA Since Boveri (1890) reported 32 chromosomes in the females of Carinaria mediterranea Blainville and Pterotrachea mutica Lesueur (=P hippocampus), nothing has been pub- lished on the chromosomes of Heteropoda. The present data confirm a diploid complement of 2n=32 in the females of P coronata and Firoloida desmaresti. But the male diploid com- plement in the three species of Pterotracheidae here studied consists of 2n=31 with heteromorphic chromosomes sug- gesting a XY,Y2 sex mechanism. The remaining autosome pairs of the Heteropoda karyotypes show a variable propor- tion of the different morphological types of chromosomes within the three species studied. The pelagic group of Heteropoda is considered as a superfamily (Boss, 1982), a suborder (Lalli and Gilmer, 1989) or an order (Van der Spoel, 1976) of the Prosobranchia Mesogastropoda. Chromosome number of 2n=14 to 2n=34 has been reported within Mesogastropoda, excluding polyploid species of Cerithi- acea (see Patterson, 1969). In recent studies on Mesogastropoda, a chromosome number of 2n=32 in Rissoidae (Thiriot-Quiévreux and Ayraud, 1982), 2n=34 in Littorinidae (Janson, 1983; Vitturi et a/., 1986, 1988) and in Naticidae (Vitturi et a/., 1982a; Komatsu, 1985) has been found. In the hydrobioid Tricula aperta (Temcharoen), diploid chromosome number showed variations from 2n=32 to 34 in females and 2n =31 to 33 in males (Kitikoon, 1982). Thus, the diploid complement observed in Heteropoda ranges close to these numbers. THIRIOT-QUIEVREUX: KARYOTYPE ANALYSIS 43 Table 4. List of known karyological data for pelagic gastropods. Morphological types (no. of chromosome pairs of autosomes) Diploid sex sm-st no. chromo. m sm st-sm st t Opisthobranchia Thecosomata, Euthecosomata Limacina inflata 20 10 Creseis acicula (Rang) 20 5 5 Creseis virgula (Rang) 20 7 1 2 Clio pyramidata Linné 22 5 1 2 2 1 Cavolinia inflexa (Lesueur) 24 6 1 2 3 Hyalocylis striata 28 7 2 1 4 Thecosomata, Pseudothecosomata Peraclis reticulata 24 1 Cymbulia peroni 34 5 4 5 1 Gymnosomata Pneumodermospsis canephora 32 2 3 5 5 Pneumodermopsis paucidens 32 7 6 1 1 Pneumoderma atlanticum 32 3 5 4 4 Paraclione longicaudata 32 4 2 1 7 2 Prosobranchia Heteropoda Pterotrachea coronata female 32 XX 7 5 1 2 male 31 XY1Y2 7 4 1 2 Pterotrachea hippocampus male 31 XY 2 2 2 1 6 3 Firoloida desmaresti female 32 XX 4 4 5 2 male 31 XY 4 4 4 2 mH re wy an a XIN, ANI AVR AA A * Om me... aid Fig. 8. Karyotypes of Firoloida desmaresti (scale bar = 5 um), (A) female, (B) male. 44 AMER. MALAC. BULL. 8(1) (1990) Sex chromosomes have been observed in several species of Mesogastropoda with difference in male sex- determining mechanisms (YX vs. XO) (see references in Pat- terson, 1969; and Stern, 1975; Thiriot-Quiévreux, 1982; Kitikoon, 1982; Vitturi et a/., 1988). Therefore, the XY,Y2 male sex mechanism here suggested in Pterotracheidae is very unusual among molluscs and probably involves translocations between sex chromosomes and autosomes (White, 1973). This sex mechanism, even if Y2 is a microchromosome as in Firoloida desmaresti suggests a multiple sex-chromosome mechanism as reported by White (1973) in different taxons of the animal kingdom. However, further investigations on males and females of Heteropoda need to be carried out to confirm and elucidate this sex mechanism. ACKNOWLEDGMENTS Thanks are due to R. Voelksen for providing live animals from the plankton, to G. Quélart for her excellent technical assistance and to P. Chang for English corrections. LITERATURE CITED Boss, R. T. 1982. Mollusca. /n: Synopsis and classification of living organisms, S. Q. Parker, ed. pp. 1092-1096. McGraw-Hill, New York. 3 Boveri, T. 1890. Zellen Studien. Uber das Verhalten der chromatischen Kernsubstanz bei der Bildung der Richtungskorper und bei der Befruchtung. Jenaische Zeitschrift fur Naturwissenschaft 24:314-401. Curini-Galletti, M. C. 1985. Chromosome morphology of Philinoglossa praelongata (Gastropoda Cephalaspidea). Journal of Molluscan Studies 51:220-222. Curini-Galletti, M. C. 1988. Analyse du caryotype de Runcina coronata (Gastropoda Cephalaspidea). Cahiers de Biologie Marine 29:313-318. Gold, J. R. and C. T. Amemiya. 1986. Cytogenetic studies in North American minnows (Cyprinidae). XII. Pattern of chromosomal nucleolus organizer region variation among 14 species. Canadian Journal of Zoology 64:1869-1877. Howell, W. H. and D. A. Black. 1980. Controlled silver-staining of nucleolus organizer regions with a protective colloidal developer: a 1-step method. Experientia 36:1014-1015. Janson, K. 1983. Chromosome number in two phenotypically distinct populations of Littorina saxatilis Olivi and in specimens of the Littorina obtusata (L.) species complex. Journal of Molluscan Studies 49:224-227. Kitikoon, V. 1982. Studies on Tricula aperta and related taxa, the snail intermediate hosts of Schistosoma mekongi. |V. Chromosome studies. Malacological Review 15:21-42. Komatsu, S. 1985. Karyotypes of two species in two families of Pro- sobranchia. Venus (Japanese Journal of Malacology) 44:49-54. Lalli, C. M. and R. W. Gilmer. 1989. Pelagic snails. Stanford Univers- ity Press, Stanford, California. 259 pp. Levan, A., K. Fredga and A. A. Sanders. 1964. Nomenclature for centromeric position on chromosomes. Hereditas 52:201-220. Nakamura, H. K. 1985. A review of molluscan cytogenetic informa- tion based on the CISMOCH-computerized system for molluscan chromosomes. Bivalvia, Polyplacophora and Cephalopoda. Venus (Japanese Journal of Malacology) 44:193-225. Nakamura, H. K. 1986. Chromosomes of Archaeogastropoda (Mollusca Prosobranchia) with some remarks on their cytotax- onomy and phylogeny. Publications of the Seto Marine Biological Laboratory 31:191-267. Nakrassov, A. 1904. Untersuchung uber die Reifung und Befruchtung des Eies von Cymbulia peroni. Anatomisches Anzeiger 24:119-127. Patterson, C. M. 1969. Chromosomes of molluscs. /n: Proceedings of 2nd symposium on Mollusca. Ernakulam, Cochin, India. Vol. 2. pp. 635-689. Marine Biological Association of India, Ramanathapuram District, Madras State. Schmeckel, L. 1985. Aspects of evolution within the opisthobranchs. In: The Mollusca. E. R. Trueman and M. R. Clarke, eds. pp. 221-267. Academic Press, New York. Stern, E. M. 1975. The chromosomes of Viviparus viviparus (Say) (Streptoneura: Viviparidae). Malacological Review 8:107-108. Thiriot-Quiévreux, C. 1984. Chromosome analysis of Mytilus (Bivalvia Mytilidae). Marine Biology Letters 5:265-273. Thiriot-Quiévreux, C. 1988. Chromosome studies in pelagic opisthobranch molluscs. Canadian Journal of Zoology 66:1460-1473. Thiriot-Quiévreux, C. and N. Ayraud. 1982. Les caryotypes de quel- ques espéces de Bivalves et de Gastéropodes marins. Marine Biology 70:165-172. Thiriot-Quiévreux, C., J. Soyer, F. de Bovée, and P. Albert. 1988. Unusual chromosome complement in the brooding bivalve Lasaea consanguinea. Genetica 76:143-151. Van der Spoel, S. 1967. Euthecosomata, a group with remarkable developmental stages (Gastropoda Pteropoda). J. Norduun en Zoon, N. V., Gorinchem, Netherlands. 375 pp. Van der Spoel, S. 1976. Pseudothecosomata, Gymnosomata and Heteropoda (Gastropoda). Bohn, Scheltema & Holkema, Evren, B. V., Utrecht. 484 pp. Vitturi, R., M. B. Rasotto, and N. Farinella-Ferruzza. 1982a. The chromosomes of 16 molluscan species. Bollettino di Zoologia 49:61-71. Vitturi, R., M. Rasotto, N. Parrinello, and E. Catalano. 1982b. Sper- matocyte chromosomes in some species of the family Aplysiidae (Gastropoda, Opisthobranchia). Caryologia 35:327-333. Vitturi, R., E. Catalano, and M. Macaluso. 1986. Chromosome studies in three species of the gastropod family Littorinidae. Malacological Review 19:53-60. Vitturi, R., E. Catalano, M. Macaluso, and B. Zava. 1988. The karyology of Littorina neritoides Linnaeus, 1758) (Mollusca, Proso- branchia). Malacologia 29(2):319-324. White, H. S. D. 1973. Animal cytology and evolution. 3rd ed. Cambridge University Press. 961 pp. Zarnick, B. 1911. Uber den Chromosomencyclus bei Pteropoden. Verhandiungen der Deutschen Zoologischen Gesellschaft 20:205-215. Date of manuscript acceptance: 2 November 1989 SAMPLING REQUIREMENTS FOR EPIPELAGIC HETEROPOD MOLLUSCS ROGER R. SEAPY DEPARTMENT OF BIOLOGICAL SCIENCE CALIFORNIA STATE UNIVERSITY FULLERTON, CALIFORNIA 92634, U.S.A. ABSTRACT Previous populational studies of epipelagic heteropods have demonstrated high between-replicate variability. Design of sampling programs whose objectives are the determination of temporal and/or spatial populational differences must address the consequences of such variability. Net size, the volume of water filtered by the net and the number of replicates necessary to sample the target populations are essential considerations. Because of the increase in sample volume, use of a large net would be expected to result in a decrease in between-replicate variability and an increase in sample precision. Large nets also reduce the likelihood of net avoidance and, therefore, the potential problem of underestimating species abundances. The effect of sample volume on between-replicate variability in species densities was assessed by 30 min oblique tows to 50 m taken with 70 cm Bongo nets and a 226 cm ring net. The total volume of water filtered by the ring net averaged 36 times that filtered by the Bongo nets. Standard errors for the four most abundant species averaged 3.5 times greater for the Bongo nets than for the ring net. On average, the level of precision (ratio of standard error to mean) attained using the larger net (0.10) was almost twice that of the small net (0.18). Estimates of the number of replicate tows needed to sample the abundant species of Hawaiian heteropods were determined from a series of 22 tows taken with the 226 cm ring net towed obliquely to a target depth of 300 m. Based on levels of precision (D) ranging from 0.10 to 0.20 and the mean (m) and variance (s?) for each species from the 22 tows, the number of required replicates (n,) was computed for each species using the equation: n, = s2m-2D-2. Regression of the log variance and log mean for the eleven most abundant species had a high coefficient of determination (0.99). Based on this regression, variances can be predicted from sample means using Taylor’s power law: s2 = 0.883m1.-479. For the seven most abundant species of heteropods estimates of the number of required replicates at five levels of precision were compared with plots of cumulative (or running) mean density and cumulative variance using the density data from the 22 oblique tows. The cumulative mean became stable at the number of replicate tows corresponding, on average, to a 0.15 level of precision. On the basis of these results ten replicate tows would be needed to sample the four most abundant species. The heteropods are a holopelagic group of carnivorous gastropods that are highly mobile and locate their prey visually using well-developed eyes (reviewed by Lalli and Gilmer, 1989). Since the heteropods occupy a high trophic position as primary and secondary carnivores (Seapy, 1980; Richter, 1982), it is not surprising that they occur in low population densities. In the upper 100 m of the water column off southern California the carinarid heteropod Carinaria japonica Okutani, 1955, generally occurred at densities less than about 10 in- dividuals per 1,000 m3, although the maximal density was 142 per 1,000 m3 (Seapy, 1974). Abundances of eleven species of heteropods from the equatorial eastern Pacific were reported by McGowan and Fraundorf (1966). The three most abundant species were atlantids that averaged between 124 and 225 animals per 1,000 m3 from eight oblique tows to 140 m taken with ring nets having mouth diameters of 1.0 and 1.4 m. The maximal density (598 individuals per 1,000 m3) was recorded for Atlanta lesueuri Souleyet, 1852. Species of epipelagic heteropods from Hawaiian waters (Seapy, 1990) generally averaged less than five individuals per 1,000 m3, although densities between 10 and 20 per 1,000 m? were recorded for several species from shallower waters. In this study the maximal density was 58 A. /esueuri per 1,000 m3 between the surface and 50 m at night. A feature common to the studies of both McGowan and Fraundorf (1966) and Seapy (1990) was high variability in species abundances between replicate tows. Since abun- dances of epipelagic heteropods are generally low and American Malacological Bulletin, Vol. 8(1) (1990):45-52 45 46 AMER. MALAC. BULL. 8(1) (1990) between-replicate variability is high, an investigation of the sampling requirements for populations of these animals is needed and should be most beneficial to the planning of future sampling programs. The importance of such preliminary studies for the determination of optimal sample size and the number of required replicates has been widely acknowledged and was emphasized recently in a review of sampling by Andrew and Mapstone (1987). METHODS Epipelagic heteropods were collected from waters overlying the 2,000 m bottom depth contour at a distance of 11 to 20 km from the southwest coast of Oahu, Hawaii during two cruises of the R/V KILA, University of Hawaii in 1986. The effect of net size on between-replicate variability was in- vestigated by comparing the densities of heteropod species from a series of six, shallow-water oblique tows taken with paired, 70 cm Bongo nets (0.8 m2 combined mouth opening area) and a 226 cm ring net (4.0 m2). Three tows were taken with each type of net to a target depth of 50 m between 1031 and 1622 hrs on 23 November 1986. Tow duration averaged 30 min, and the towing speed was 2-3 knots. The ring net attained a maximal depth of 50 m during each tow and filtered an average of 5,112 m3 (range = 4,508-5,841 m3), while the Bongo nets achieved an average maximal depth of 40 m (range = 38-42 m) and filtered an average of 1,417 m3 (range = 1,086-1,700 m3). Thus, net size differed while tow duration was the same, resulting in a sample volume that was, on the average, 3.6 times greater for the ring net than for the com- bined Bongo nets. For each net type replicate variability was expressed in terms of the standard error (standard deviation divided by the square root of the number of samples). Sampling precision (the ratio of the standard error to the mean) was determined for the four most abundant species (mean densities greater than 4.0 individuals per 1,000 m) for the two types of nets. The numbers of replicate tows required to sample heteropod populations were investigated using a series of 22 oblique tows to a target depth of 300 m taken with a 226 cm ring net (mouth opening = 4.0 m2, mesh width = 0.5 mm). Tows were taken consecutively over a 24 hr period between 1524 hrs on 24 March and 1539 hrs on 25 March 1986, resulting in a total of 11 day and 11 night tows. The duration of the tows ranged from 32 to 43 min and averaged 38 min. The volume of water filtered by the net during each tow was determined using a T.S.K. Model Ol-210 Flow Meter, and averaged 6,973 m3 (standard deviation = 501 m3, range = 5,820 to 7,963 m3). The maximal depth achieved during each tow was obtained from a Benthos Time-Depth Recorder. Although the maximal target depth was 300 m, actual depths ranged from 240 to 400 m (average depth = 299 m; stan- dard deviation = 37 m). Based on vertical distribution and abundance data from an earlier study (Seapy, 1990), greater than 99% of the epipelagic heteropods off Hawaii occur in waters shallower than 200 m (which representes the upper one-half of the epipelagic zone off Hawaii; Young, et a/., 1980). Thus, oblique tows that extend deeper than 200 m can be considered to have adequately sampled the vertical range of the Hawaiian epipelagic heteropods. Since actual species abundances (ex- pressed as numbers per unit volume) differ both vertically and, for deeper-dwelling species, between day and night periods (Seapy, 1990), densities are most appropriately expressed in terms of numbers beneath a unit area of sea surface (area sampled = volume filtered/maximal depth of tow). In the pre- sent study the area sampled ranged from 15.4 to 31.9 m2 per tow, and the mean area was 23.7 m2. To standardize these differences in areas sampled, all density data were converted to numbers beneath 25.0 m2 of sea surface. The numbers of replicates needed to sample the seven most abundant species of heteropods collected by the 22 oblique tows taken in March were determined using a method described by Elliott (1971) for application with the freshwater benthos. This method has been used subsequently in several studies, e.g. Mcintyre et a/. (1984) and Vezina (1988) for the marine benthos and by Downing et al. (1987) for marine and freshwater zooplankton. The general formula is: n, = s2 m2D-2 (1) where: n, = number of required replicates; s? = sample variance; m = sample mean; D = precision (ratio of the stan- dard error to the mean). To apply this equation the investigator must select a required level of precision (D). This decision is of practical importance because the number of required replicates increases sharply as the level of precision in- creases. Elliott indicated that one should be able to tolerate a standard error equal to 20% of the mean (i.e. D = 0.20) for freshwater benthic populations. However, if one required a standard error equal to 10% of the mean (i.e. D = 0.10), the number of replicates would have to be four times as great. In addition to selecting the level of precision, equation (1) requires estimates of sample means and variances. A limited number of preliminary tows can give an approximate range of mean densities for the species under consideration. However, since high between-replicate variability has been reported (McGowan and Fraundorf, 1966; Seapy, 1990) for the epipelagic heteropods, high variances would be expected. Since a greater number of replicate samples appears to be necessary to make more reliable estimates of the variance (see Results), it would be beneficial if one could predict variance values based on means that had been obtained from a relatively small number of preliminary samples. For a variety of natural populations the mean has been shown (Downing, 1979; Downing and Cyr, 1985; Morin, 1985; Downing et al., 1987; Vezina, 1988) to be related to the variance by the power law of Taylor (1961): s2 = amb (2) where: s2 = sample variance; m = sample mean; a = a constant (the value of s? corresponding to a mean of 1.0, deter- mined from the regression equation); b = the regression coefficient (i.e. the slope of the regression line). In the pre- sent study log mean was plotted against log variance for the eleven most abundant species sampled by the 22 oblique tows in March 1986. Values for the constants a and b in SEAPY: NET SAMPLING OF HETEROPODS 47 equation (2) were obtained from a least squares regression of log mean on log variance. The density data from the 22 oblique tows were also used to calculate and plot cumulative (or running) means and cumulative (or running) variances for the seven most abun- dant species. The number of required replicates corre- sponding to five levels of precision (0.10, 0.125, 0.15, 0.175 and 0.20) were indicated on each cumulative mean plot for visual comparison with the cumulative mean and cumulative variance data. RESULTS Densities of the four most abundant species of heteropods from samples collected with 70 cm Bongo nets and a 226 cm ring net to a target depth of 50 m (Table 1) were not significantly different (Mann Whitney U test). Replicate variability resulting from the use of the two types of nets was assessed by computing standard errors (Fig. 1), which averaged 3.5 times greater for the Bongo nets than for the ring net and ranged from 1.8 (Firoloida desmaresti Lesueur, 1817) to 5.0 (Atlanta oligogyra Tesch, 1906) times greater for the Bongo nets. The level of sampling precision (Table 1) was, on average, almost twice as great for the ring net (mean D = 0.10) as for the Bongo nets (mean D = 0.18), which cor- responds to a nearly four-fold increase in replication requirements. A total of thirteen species of epipelagic heteropods were represented among the 22 oblique tows taken in March 1986 (Table 2). For the seven most abundant species (those with mean densities greater than four per 25 m2), no signifi- cant density differences (Mann Whitney U test) were obtained between the 11 day and 11 night tows (Table 3). Thus, the day and night data were pooled in the subsequent analyses. A log-log plot was constructed (Fig. 2) from the mean and variance values computed for the eleven most abundant species collected by the 22 replicate tows (Table 2). The resultant regression line was described by the equation: log s? = —0.125 + (1.479) (log m) (3) and had an extremely high coefficient of determination (r2 = 0.99), which was highly significant (t = 31.9; p <.01). Based on this regression analysis the relationship between the heteropod means and variances can be expressed by Taylor’s (1961) power law: s? = 0.883m!.479. (4) N 70-cm Bongo Nets (n=3) [| 226-cm (n=3 as Ring Net ) STANDARD ERROR (NO. 1,000 M73) ine) a—N z ss a ae Oo. S ° SS >, 2 xo 2 a, AL oO Oe oe “ % a “, a) “a — S. icy) ~ Kos < x) Qs on a =} 2 ~~ > Oo A [os fos @ Ss Fig. 1. Standard errors (numbers per 1,000 m3) for the four most abun- dant species of heteropods collected by oblique tows to a target depth of 50 m using 70 cm Bongo nets and a 226 cm ring net. The numbers of replicates required to sample the seven most abundant species were estimated (Table 4) at five different levels of precision, ranging from 10% to 20%, using the mean and variance values for the 22 replicate tows in Table 2. For comparison, the numbers of required tows were also computed for each species using variances determined from equation (4). Numbers of required replicates ranged from four to ten at the 20% level of precision, and from 14 to 40 at the 10% level (Table 4). These data illustrate the principle that an increase in the level of precision from 20% to 10% results in a four-fold increase in the number of required replicates. For this reason selection of an appropriate level of precision is of great importance in the design of a sampling program. The numbers of required replicate tows corresponding to each of the five levels of precision in Table 4 were includ- ed on cumulative mean plots for the five most abundant species (Figs. 3-7). The level of precision which appeared to be most appropriate for each species was then assessed visually from the plots. For Atlanta peroni (Fig. 3) the cumulative mean became relatively stable and approximated the final mean value of 32.8 per 25 m2 (indicated by the horizontal line in the plot) after the seventh tow. On the cumulative mean plot the eighth tow falls between the number of required tows at the 0.15 and 0.125 levels of precision. If one had only taken eight tows, however, the variance would have been overestimated by nearly 25% (the final variance Table 1. Mean densities (numbers 1,000 m-3), standard errors (SE) and precisions (D) for the four most abundant species captured by replicated oblique tows to a target depth of 50 m on 23 November 1986 using paired 70 cm Bongo nets and a 226 cm ring net. Bongo Net tows Ring Net tows (n = 3) (n = 3) Species Mean SE D Mean SE D Atlanta lesueuri 39.0 4.7 0.12 24.7 1.1 9.04 Firoloida desmaresti 11.5 2.3 0.20 10.3 1.3 0.13 A. turriculata d’Orbigny, 1836 16.5 4.8 0.29 8.2 1.5 0.18 A. oligogyra 9.6 1.2 0.13 441 0.2 0.05 48 AMER. MALAC. BULL. 8(1) (1990) Table 2. Mean densities, ranges and variances (numbers 25 m-2) of epipelagic heteropods captured by 22 replicate oblique tows to a target depth of 300 m on 24-25 March 1986. Species Mean Range Variance Atlanta peroni Lesueur, 1817 32.8 10.6-52.7 158.0 A. plana Richter, 1972 25.5 11.4-43.5 95.1 A. inflata Souleyet, 1852 18.4 8.9-37.3 54.4 Protatlanta souleyeti (Smith, 1868) 13.5 1.3-25.8 41.1 A. turriculata 8.7 2.7-21.3 18.8 A. lesueuri 5.1 0.8-11.4 10.3 A. meteori Richter, 1972 4.6 0-9.5 6.4 A. helicinoides Souleyet, 1852 2.3 0-6.1 3.0 A. oligogyra 1.3 0-6.5 2.3 A. fusca Souleyet, 1852 1.1 0- 4.2 1.0 A. tokiokai van der Spoel and Troost, 1972 (=A. inclinata Souleyet, 1852) 0.9 0-3.7 1.2 A. echinogyra Richter, 1972 0.1 0-1.4 0.1 Oxygyrus keraudreni (Lesueur, 1817) 0.1 O-1.1 0.1 500 100 o1) VARIANCE oO | 5 10 50 MEAN Fig. 2. Log-log plot of sample mean and variance for eleven species of heteropods obtained from 22 oblique tows to a target depth of 300 m. The regression equation for the plotted line is: log;9 variance = -0.125 + (1.479) (logio mean). value is indicated by the horizontal line in the plot). The cumulative variance declined to about the level of the final variance value at the eleventh tow, decreased and then in- creased gradually after that point. The plot of the cumulative mean for A. plana (Fig. 4) was similar to that for the preceding species. The cumulative mean approached the final mean value at the eighth tow and was very stable beyond that point. The eighth tow corresponded to the number of required replicates between the 0.15 and 0.125 levels of precision. The cumulative variance “‘overshot”’ the final variance by 33% at the eighth tow and did not decrease to the level of the final variance until the eighteenth tow. The cumulative mean for Atlanta inflata (Fig. 5) approx- imated the final mean value and remained nearly constant after the thirteenth tow. However, beginning with the seventh tow (corresponding to the 0.15 level of precision), the cumulative mean changed only gradually. Like A. plana, the cumulative variance ‘‘overshot’’ the final variance at the seventh tow, and then remained above the level of the final variance until the eighteenth tow. The cumulative mean for Protatlanta souleyeti (Fig. 6) decreased slowly between the fifth and twelfth tows, after which it gradually rose to nearly the level of the final mean value by the eighteenth tow. However, the cumulative mean was close initially to the final mean value between the seventh and tenth tows (corre- sponding to the 0.175 and 0.15 levels of precision), and changed only gradually after the seventh tow. The cumulative variance ‘‘overshot”’ (at four tows) and then ‘‘undershot’’ (at ten tows) the final variance, but became relatively stable at levels near the final variance after the eleventh tow. The plots of cumulative mean and variance for Atlanta turriculata (Fig. 7) were quite different from those of the preceding species. The cumulative mean remained below the Atlanta peroni oO (= z, Oa 5 200 = = 0: Cee rr Ee ase 5 z 40 Noo —0 z < ‘. Oo 9a WwW ~e. _®@ — = 30 th : re baler er eT Fo Tee nm = a ' fj 2 20 2079 tas ‘5 | Eee a eel aL eT ST a CR Coe aor, Coa (oe Conan ims am Foal aca fa 80 >* 60 . ee ° _ 4~e 2a 40 Sener ‘ 0 as \ e WW 7 ° ra ~e vi Oo e +e S ° e 22 20 be fees é d Se e e (0) a (a a a as (a a a en ed ay 5 10 15 20 TOW NUMBER Fig. 3. Densities (numbers per 25 m2) of Atlanta peroni from each of the 22 oblique tows taken to a target depth of 300 m (lower illustra- tion). Plots of the cumulative sample mean and cumulative sample variance in numbers per 25 m2 (upper illustration). The number of required replicates corresponding to five levels of precision (0.20, 0.175, 0.15, 0.125, and 0.10) are included on the plot of the cumulative mean. SEAPY: NET SAMPLING OF HETEROPODS 49 Table 3. Comparison of species densities (numbers 25 m-2) between day and night periods for the seven most abundant epipelagic heteropods captured 24-25 March 1986 off Oahu. Day-night differences in densities were tested by the Mann Whitney U statistic (s = significant at p < .05;ns = not significant). Day (n=11) Night (n = 11) Species Mean Variance Mean Variance Signif. Atlanta peroni 37.3 104.0 28.4 172.4 ns A. plana 24.5 75.0 26.5 113.3 ns A. inflata 18.4 26.9 18.5 81.8 ns Protatlanta souleyeti 16.5 36.3 10.4 27.6 ns A. turriculata 9.6 22.9 7.8 13.0 ns A. lesueuri 5.0 8.4 5.3 12.2 ns A. meteori 4.6 4.6 4.7 8.3 ns Atlanta plana eos 3 120 2 Table 4. The numbers of replicates required to sample the seven eo | OSG ~0—0—0 9 S most abundant species of epipelagic heteropods at five levels of preci- i el NE ce sion, using equation (1) from the text and the species means and ae o-=8 2 3 variances (n = 22 tows) from Table 1. The values in parentheses 9 2 are the numbers of required replicates based on means from Table Z 30 é om 1 and variances computed from equation (4). = ] ——s = ae te o—e_e—e—y—e = 20 1 See ' ; Level of Precision 4 + | S io: pops oe 10 Species 0.10 0.125 0.15 0.175 0.20 |S SD Posada) (a Cee, Sas Oe Ta | wrote T : re meee T gases | ' Atlanta peroni 15 9 7 5 4 e ] ° 14) (9) (6) (5) (4) ~& 40 a ( Ee | ee i hae a /j- A / A. plana 15 9 7 5 4 Spe \ ui sn Aa? (16) (10) (7) 8) A) az 6 | fiat A i A. inflata 16 10 7 5 4 (19) (12) (9) (6) (5) 5 10 15 20 Protatlanta souleyeti 23 14 10 7 6 TOW NUMBER (23) (15) (10) (7) (6) Fig. 4. Densities (numbers per 25 m2) of Atlanta plana from each pene Ulata a us _ : 3 of 22 oblique tows to a target depth of 300 m (lower illustration). Plots re , vy ee : = hy ie of the cumulative sample mean and cumulative sample variance in areeeuy 38 94 17 12 9 numbers per 25 m2 (upper illustration). The number of required sisi pe oe “ se replicates corresponding to five levels of precision (0.20, 0.175, 0.15, 0.125 and 0.10) are included on the plot of cumulative mean. final mean value between the second and ninth tows. At the tenth tow (corresponding to the number of required replicates between the 0.175 and 0.15 levels of precision), the cumulative mean approximated the final mean value and changed only gradually thereafter. The cumulative variance was far below the final variance (by a maximum of 94% at the sixth tow) until the twentieth tow. Cumulative mean and variance plots were also con- structed for Atlanta lesueuri and A. meteori, but are not in- cluded here. The cumulative mean plot for A. /esueuri oscillated between 6.0 (at the second tow), 3.8 (at the fifth tow) and 6.0 individuals per 25 m2 (at the tenth tow). After the eighth tow, however, the cumulative mean decreased gradually to a value close to the final mean at the fourteenth tow (corre- sponding to the number of required replicates between the 0.15 and 0.175 levels of precision). The cumulative variance followed a similar pattern to the cumulative mean, but was extremely stable and approximated the final variance after the tenth tow. For A. meteori the cumulative mean decreased from (40) (26) (18) (13) (10) initial high values of 6.5 per 25 m2 (the second through fourth tows) to a value close to the final mean at the eleventh tow (corresponding to the number of required replicates between the 0.15 and 0.175 levels of precision), and remained stable thereafter. Cumulative variance values were above (by a max- imum of 72%) the level of the final variance between the fifth and twenty-second tows. The above analyses of the cumulative mean plots sug- gest that the number of required replicate tows for the included heteropod species correspond, on average, to about the 0.15 level of precision. Inspection of the required number of replicates at a precision of 0.15 in Table 4 indicates that one would have to take ten replicate tows in order to sample the four most abundant species, while about eighteen tows would be required if one were to include all seven of the most abun- dant species. The above comparisons of the cumulative mean and cumulative variance plots (Figs. 3-7) indicated that for most species the cumulative mean stabilized at a much lower 50 AMER. MALAC. BULL. 8(1) (1990) number of tows than did the cumulative variance. Thus, although a preliminary sampling program for heteropod species populations would provide adequate estimates of species mean densities, the sample variances are likely to be inaccurate. Instead, the variances corresponding to the sample means should be determined from equation (4). DISCUSSION A preliminary sampling program is required when the research objectives include spatial and/or temporal com- parisons of population densities. The present study suggests that the characterization of zooplankton populations at a par- ticular place and time can be accomplished by taking replicated oblique tows that extend through the vertical ranges of the species under investigation. Preliminary samples are necessary for the determination of the appropriate sample size (in this case, the size of the net and the volume of water to be filtered by the net) and the number of required replicates. These considerations are extremely important in oceano- graphic studies because of the difficulty in obtaining work- ing time aboard research vessels and the high cost of ship time. Even if sampling time is not limiting, one must also take into account the costs and time required to process the col- lected samples. The relationship between sampling efficien- cy and costs have been discussed for marine populations in general by Andrew and Mapstone (1987) and for the zoo- plankton in particular by Cassie (1968), Wiebe (1971) and Downing et a/. (1987). Because heteropods occur in relatively low abundance, adequate numbers of specimens and representation by the species present in an area can best be achieved by filtering a relatively large volume of water during each tow. Large plankton nets, such as the 226 cm ring net used here, are preferred over smaller nets since they can filter a large volume of water in a relatively short time period. Large nets also Atlanta inflata 80 P~e—9~9_ 0 —0~ 0 S = o=0—e—e—p—e—0 Do 5 ee _o~ Zz Zz ae iS = 30 o—o Om WwW = 20 as dat eae S | “s—0 0 S 10 J rt ‘ 125 uy "40 a ys — SS e nw oo 20 e—e_ aN en ZN 4 e—e -®. oe © ae nar Se ae ee 2 OJ = [oh 5 ae oD (a ars Te Te es ae am TOW NUMBER Fig. 5. Densities (numbers per 25 m2) of Atlanta inflata from each of 22 oblique tows to a target depth of 300 m (lower illustration). Plots of the cumulative sample mean and cumulative sample variance in numbers per 25 m2 (upper illustration). The number of required replicates corresponding to five levels of precision (0.20, 0.175, 0.15, 0.125 and 0.10) are included on the plot of cumulative mean. reduce the sampling bias resulting from possible net avoidance and improve the chances that the less abundant species will be adequately represented (McGowan and Fraundorf, 1966; Cassie, 1968). Between-replicate variability is reduced and sampling precision is increased by using large nets (Wiebe, 1971), and the increased sample volume can result in a substantial decrease in sample variance and, thus, in the number of required replicates (e.g. Andrew and Mapstone, 1987; Downing et al., 1987). These relationships were supported by the present study, in which a reduction of between-replicate variability by two to five-fold and an in- Protatlanta souleyeti (e) 60 pe Pn z= --4 o~ 56, = 32 ©=9=0=9=0=5 9 =o 8 40 — So~ 9° QQ P a 20D % 20 5 WwW o 7° om = 15 eA o-*~, ~< en tt. woo = ° t Ne eo o !0 20 175 15 4 125 Lem URED Ce) TR Darl ie REL a Cee OY aera] be Orca Con Ge a Ca | 30 DENSITY (NO. 25 M2) me ine) [o) {eo} ee pent a SS es re / oe a a ae SS ae SS. J ae 10) e Lene Cae ya aes SN a a a ed rea) fn es eT Derg ea yt] 5 10 15 20 TOW NUMBER Fig. 6. Densities (numbers per 25 m2) of Protatlanta souleyeti from each of 22 oblique tows to a target depth of 300 m (lower illustra- tion). Plots of the cumulative sample mean and cumulative sample variance in numbers per 25 m2 (upper illustration). The number of required replicates corresponding to four levels of precision (0.20, 0.175, 0.15 and 0.125) are included on the plot of cumulative mean. Atlanta turriculata 4 os Da on He 20 ES 90-00-09 0-00- oO ©—o0—0—9__ 5 0-9-9 om s z2 10 - ae ae Dy re la 8-00 *-e pe os. 5 ' ' ) 20. 175 15 125 30 a pes 20 A HO P 7, ZENG *_ / ~e . AS ‘ fe e - as yee / e \ fe! N Me 2 fe) e “e eo Cp ea) NS SS a. a (ec oad] CS ees en Fh ae | 5 10 15 20 TOW NUMBER Fig. 7. Densities (numbers per 25 m2) of At/anta turriculata from each of 22 oblique tows to a target depth of 300 m (lower illustration). Plots of the cumulative sample mean and cumulative sample variance in numbers per 25 mz? (upper illustration). The number of required replicates corresponding to four levels of precision (0.20, 0.175, 0.15 and 0.125) are included on the plot of cumulative mean. SEAPY: NET SAMPLING OF HETEROPODS 51 crease in sampling precision that averaged nearly two-fold resulted from sampling with a 226 cm ring net compared with 70 cm Bongo nets. Reduction in the numbers of required replicate tows provides an added benefit of reducing the time period necessary for the collection of the samples. Even if one makes a conscious effort to ‘‘track’’ the body of water being sampled (such as McGowan and Fraundorf, 1966, did by using a reference drogue), the chances of sampling the same parcel of water decrease as the total sampling time increases. Anaiyses of the cumulative plots of species means and variances (Figs. 3-7) indicated that the numbers of replicate tows necessary to estimate the final variance values (after 22 tows) were higher (by several to many samples) than were needed to estimate the final means. Thus, the direct computa- tion of variances from preliminary sample means using Taylor’s power law is recommended. Taylor’s power law has been applied to mean and variance data for populations from a variety of different habitats. For marine benthic populations Vezina (1988) tabulated over 3,000 estimates of mean and variance and showed that although the data were obtained from populations in varied types of habitats, the least squares regression of log mean on log variance resulted in a high coef- ficient of determination (0.86). The variance could then be predicted from the mean using Taylor’s power law (s2 = amb = 1.641 m'.219). The value of b (1.219) was lower than the average of about 1.5 found for freshwater benthic populations (Downing, 1979; Downing and Anderson, 1985; Downing and Cyr, 1985). Cassie (1963) reported a value for b of 1.57 for zooplankton populations. The b value from the present study (1.479) is somewhat lower than these last two values, and is considerably lower than that (1.849) obtained by Downing et al. (1987) from an analysis of 1,189 sets of replicate samples of freshwater and marine zooplankton. The value for b ob- tained by Downing et a/. (1987) was based on a variety of sources whose density data ranged from 5 x 10-7 to 1.6 x 103 individuals per liter and sample values that ranged from 0.8 to 10® liters. To standardize these varied data, species densi- ties were transformed to numbers of individuals per liter. By applying a term for sample volume (Vc) in the equation for Taylor’s power law (s2 = ambVc), Downing et al. reduced the value of b to 1.622, which is much closer to that obtained in the present study. The results of Vezina’s (1988) analyses of the mean to variance relationship argue that abundance data should be given in units appropriate to the sample size used in the study. Expression of the abundance data from the present study in terms of numbers per 25 m2 was done in order to approximate the average area sampled by the 22 replicate tows (i.e., 23.7 m2). In turn, this resulted in density values that were close to the raw sample counts, e.g., the mean density of Atlanta peroni was 32.8 individuals per 25 m2 and the average number of A. peroni collected per tow was 30.8. Density data (in numbers per 1,000 m3) were tabulated for eight species of atlantids from the eastern Pacific Ocean (southeast tip of Baja California) by McGowan and Fraundorf (1966). Four series of oblique tows to 140 m were taken with ring nets of different sizes. However, possible avoidance of all nets except for the largest one (140 cm diameter) was suggested for several species. The average volume of water filtered by the four tows taken with the 140 cm net was 510 m3, and | converted their data from numbers per 1,000 m3 to numbers per 500 m3 for analysis here. Regression of log mean and log variance for the six most abundant species (range in mean density = 1.5 for A. fusca to 111.1 animals per 500 m3 for A. gaudichaudi Souleyet, 1852) resulted in the regression equation, log s2 = - 0.666 + (1.567) (log m), which had a very high coefficient of determination (r2 = 0.94). Expressed as Taylor’s power law, s? = 0.515m'.567, Superimposing this regression line on figure 2 results in remarkably close agreement (in terms of both the slope and elevation) with the regression for the eleven species of Hawaiian heteropods. The value of b (1.567) is only slightly greater than that found in this study (1.479) and matches the value of 1.57 reported by Cassie (1963). In conclusion, the present results suggest that the four most abundant species of epipelagic heteropods off Hawaii (i.e. those species whose densities exceeded 13 individuals per 25 m2) can be sampled by ten replicate oblique tows to a target depth of 300 m using a continuously-open, 226 cm ring net. To sample the three most abundant species (i.e. those whose mean densities exceed 18 per 25 m2) fewer tows (eight) would be needed, while considerably more tows (about 18) would be required to sample the sixth and seventh most abun- dant species (mean densities of about five per 25 m2). Ap- plication of Taylor’s power law to the mean and variance data for the eleven heteropod species sampled by the ring net in- dicated that variances can be predicted with a high degree of confidence from sample means (equation 4). Thus, the number of required replicates for a sampling program whose goal is the spatial and/or temporal comparison of heteropod species populations can be determined using equation (1) from estimates of population means, their corresponding variances (from equation 4) and the assumption of a 15% level of precision. ACKNOWLEDGMENTS The samples used in this study were collected during two cruises in 1986 of the R/V KILA, University of Hawaii. | am most grateful for the support of the officers, crew and members of the scien- tific party during these cruises, and to Richard Young for his collabora- tion in the sampling program. For the loan of the Bongo nets, | thank Jed Hirota. My appreciation is extended to Richard Young and Jed Hirota for their critical reviews of the manuscript. This study was sup- ported by National Science Foundation Grant OCE-8500593. LITERATURE CITED Andrew, N. L. and B. D. Mapstone. 1987. Sampling and the descrip- tion of spatial pattern in marine ecology. /n: Oceanography and Marine Biology Annual Review, Vol. 25, M. Barnes, ed. pp. 39-90. Aberdeen University Press, Aberdeen. Cassie, R. M. 1963. Microdistribution of plankton. in: Oceanography and Marine Biology Annual Review, Vol. 1, H. Barnes, ed. pp. 223-252. Allen and Unwin Ltd., London. 52 AMER. MALAC. BULL. 8(1) (1990) Cassie, R. M. 1968. Sampling design. In: Zooplankton Sampling, pp. 105-121. Monographs on Oceanographic Methodology 2. UNESCO, Paris. Downing, J. A. 1979. Aggregation, transformation and the design of benthos sampling programs. Journal of the Fisheries Research Board of Canada 36:1454-1463. Downing, J. A. and M. R. Anderson. 1985. Estimating the standing biomass of aquatic macrophytes. Canadian Journal of Fisheries and Aquatic Science 42:1860-1869. Downing, J. A. and H. Cyr. 1985. Quantitative estimation of epiphytic invertebrate populations. Canadian Journal of Fisheries and Aquatic Science 42:1570-1579. Downing, J. A., M. Perusse, and Y. Frenette. 1987. Effect of inter- replicate variance on zooplankton sampling design and data analysis. Limnology and Oceanography 32:673-680. Elliott, J. M. 1971. Some Methods for the Statistical Analysis of Samples of Benthic Invertebrates. Scientific Publication No. 25, Freshwater Biological Association, The Ferry House. 144 pp. Lalli, C. M. and R. W. Gilmer. 1989. Pelagic Snails. The Biology of Holopelagic Gastropod Mollusks. Stanford University Press, Stanford. 259 pp. McGowan, J. A. and V. J. Fraundorf. 1966. The relationship between size of net used and estimates of zooplankton diversity. Lim- nology and Oceanography 11:456-469. Mcintyre, A. D., J. M. Elliot, and D. V. Ellis. 1984. Design of sampling programmes. In: Methods for the Study of Marine Benthos, N. A. Holme and A. D. Mcintyre, eds. pp. 1-26. International Biological Programme Handbook No. 16, Second Edition. Blackwell Scientific Publications, Oxford. Morin, A. 1985. Variability of density estimates and the optimization of sampling programs for stream benthos. Canadian Journal of Fisheries and Aquatic Science 42:1530-1534. Richter, G. 1982. Mageninhaltsuntersuchungen an Oxygyrus keraudreni (Lesueur) (Atlantidae, Heteropoda). Beispiel einer Nahrungskette im tropischen Pelagial. Senckenbergiana maritima 14:47-77. Seapy, R. R. 1974. Distribution and abundance of the epipelagic mollusk Carinaria japonica in waters off southern California. Marine Biology 24:243-250. Seapy, R. R. 1980. Predation by the epipelagic heteropod mollusk Carinaria cristata forma japonica. Marine Biology 60:137-146. Seapy, R. R. 1990. Patterns of vertical distribution in epipelagic heteropod molluscs off Hawaii. Marine Ecology Progress Series 60:235-246. Taylor, L. R. 1961. Aggregation, variance and the mean. Nature, Lon- don 189:732-735. Vezina, A. F. 1988. Sampling variance and the design of quantitative surveys of the marine benthos. Marine Biology 97:151-155. Wiebe, P. H. 1971. A computer model study of zooplankton patchiness and its effect on sampling error. Limnology and Oceanography 16:29-38. Young, R. E., E. M. Kampa, S. D. Maynard, F. M. Mencher and C. F. E. Roper. 1980. Counterillumination and the upper depth limits of midwater animals. Deep-Sea Research 27A:671-691. Date of manuscript acceptance: 25 November 1989. IN SITU OBSERVATIONS OF FEEDING BEHAVIOR OF THECOSOME PTEROPOD MOLLUSCS RONALD W. GILMER MARINE SCIENCE DIVISION HARBOR BRANCH OCEANOGRAPHIC INSTITUTION 5600 OLD DIXIE HIGHWAY FORT PIERCE, FLORIDA 34946, U.S.A. ABSTRACT Recent in situ observations of thecosome pteropods were made during five cruises in tropical, temperate, and arctic waters of the North Atlantic Ocean, and during one astral summer season in Antarctic waters. The long quiescent periods employed by thecosomes to fish their mucous feeding webs and the apparent lack of a pumping mechanism to move water through the web suggests they rely on contact trapping of large, motile prey. Species of Cavolinia and Diacria were estimated to re- main in one location for at least 35 min to fish sequential webs. The external spherical webs used by euthecosomes and their ability to rapidly ingest them appear to be a unique feeding method among plankton that utilize mucous feeding structures. All euthecosomes observed during night dives entrapped numerous small crustaceans as the webs were withdrawn, and all specimens diver-collected at night contained small copepods in their guts. Crustaceans also accounted for up to 25% by volume of the gut contents of Limacina helicina (Phipps) preserved in situ in arctic waters during July and August, 1988. Individuals of a given species use webs of comparable dimensions in different water masses. Based on the observed feeding strategies and on the common ingestion of copepods by arctic Limacina and by temperate and tropical cavoliniid species, carnivory can not be precluded as a primary feeding habit for the Thecosomata. The Thecosomata comprise an order of common opisthobranch gastropods which exist in the holoplankton by means of parapodia for rapid swimming and by the use of large, external mucous webs to collect food (Lalli and Gilmer, 1989). Their feeding habits depend on buoyancy control and passive drifting. Consequently, they have greatly reduced the wall thickness of their external shell or replaced it with internal gelatinous conchae. They have no gills which can function for food gathering and resemble vermetid prosobranchs (Hughes and Lewis, 1974; Hughes, 1978) by using cilia on the surfaces of the mantle and footlobes to manipulate their feeding webs. Much of their behavior remains obscure due to their remote habitat, and the difficulty of observing them in an undisturbed state. In addition, their fragility makes col- lection of undamaged specimens difficult. Even carefully col- lected individuals display abnormal behavior in the laboratory where they only survive for brief periods. Using blue water scuba techniques (Hamner, 1975) to observe undisturbed thecosomes has been the most useful means to study their feeding habits. The feeding webs are usually so fragile and transparent that they are only visible in daylight with bright strobe lighting or by the delicate application of carmine particles to the web surfaces (Gilmer and Harbison, 1986). The euthecosomes (Limacina, Creseis, Styliola, Hyalocylis, Clio, Cavolinia, Diacria) use spherical webs attached directly to the ciliated footlobes on the wings. The pseudothecosome genera (Peraclis, Cymbulia, Corolla, Gleba) use large flat or funnel shaped webs that float above the wingplate and are attached to the animal by a proboscis com- posed of the footlobes. Thecosomes are extremely sensitive to turbulence and, at the slightest provocation, will abandon their feeding activity with rapid escape swimming. Although the abandoned webs are left floating intact, their transparent and fragile nature have thus far made it impossible to sam- ple them quantitatively. Information on thecosome diet is limited to a few qualitative descriptions of gut contents, fecal pellets, and web fragments (see review in Lalli and Gilmer, 1989). Based on these studies, thecosomes appear to be indiscriminate omnivores ingesting all size categories of prey from 1 um bacteria to copepods as large as 3 mm in length [seen in Limacina helicoides Jeffreys, Gilmer (pers. obs.)]. Thecosomes, however, are often categorized solely as herbi- vores (e.g. Morton, 1954; Silver and Bruland, 1981; Foster, American Malacological Bulletin, Vol. 8(1) (1990):53-59 53 54 AMER. MALAC. BULL. 8(1) (1990) 1987; Boysen-Ennen and Piatkowski, 1988) since many phytoplankton cells are captured and ingested with the web, and are prevalent in the fecal pellets. Herbivory is also con- sidered synonymous with mucous suspension feeding (Jorgensen, 1966). This label, however, ignores a much broader diet that often includes many protozoan and zoo- plankton prey items (e.g. Richter, 1977, 1983; Ishimaru et a/., 1988; Lalli and Gilmer, 1989) and seems generally in- appropriate to describe their feeding strategy. Although no detailed studies exist to document the relative frequency of the various prey fractions, the external web (Gilmer, 1972, 1974; Gilmer and Harbison, 1986) provides an obvious trapping mechanism for fast swimming organisms. From mid-July to mid-August, 1988, metazoan zooplankton comprised an average of 45% by volume of items in the guts of 28 subarctic Limacina helicina (Phipps) preserved in situ (Gilmer and Harbison, unpub. data). This paper describes observations, made with the use of scuba, of undisturbed thecosome individuals. These obser- vations, some lasting up to 15 min, expand on previous obser- vations (Gilmer, 1972; Gilmer and Harbison, 1986) and sug- gest ways that feeding webs could be produced, fished, and ingested. The term ‘‘mucous trapper’’ (Fallensteller), sug- gested by Richter (1977), is the most descriptive term relating to thecosome feeding behavior. MATERIALS AND METHODS Thecosome pteropods were observed and collected in hand-held glass jars by scuba divers during four cruises in the tropical and subtropical North Atlantic Ocean in May through August, 1986 (R/V ‘‘Oceanus’’ cruises 176, 177) in March, April, July and and August 1987 (R/V ‘“‘Oceanus’”’ cruises 184, 191), in the arctic and subarctic Atlantic Ocean in July and August, 1988 (R/V ‘“‘Endeavor”’ cruise, 182), and under sea ice in McMurdo Sound, Antarctica in November 1987. Individual thecosomes were observed and photo- graphed for up to 15 min in the upper 30 m of the water col- umn using standard blue water techniques (Hamner, 1975). To make feeding webs more visible, | used either carmine dye dispensed from a plastic squeeze bottle or strobe lighting. On night dives the absence of ambient light made webs easily visible with diving lights. Photographs were taken with a Nikonos V underwater camera fitted with 1:1 or 1:2 close-up lenses and backlit from 10 to 30 cm with one or two Nikonos SB-103 underwater strobes. Either Kodak Panatomic X, Technical Pan black and white film, or Kodachrome 64 color films were used. Some individuals were photographed in successive intervals of 20 to 30 sec to record feeding sequences. Web diameters were estimated to the nearest 10 mm from field photographs. Shell dimensions were measured to the nearest 0.2 mm using a dissecting microscope and ocular micrometer. Activity of arctic Limacina helicina was measured by a diver swimming horizon- tally through dense populations and randomly counting whether individuals were motionless or swimming. Counts were made until 10 swimming individuals were observed. RESULTS LIMACINA HELICINA (LIMACINIDAE: EUTHECOSOMATA) Limacina helicina retains its largest known size in subarctic Atlantic waters (Lalli and Gilmer, 1989). | measured feeding webs (Fig. 1a) up to 60 mm in diameter on specimens with 12.2 mm diameter shells (Table 1). L. helicina antarctica Woodward with shells measuring 4.6 mm in diameter (Table 1), fed from webs with estimated diameters of 20 mm. L. helicina is very sensitive to turbulence and usually abandons its feeding web in the presence of a diver. Careful placement of a camera framer around the animal can cause it to simply draw in the web. It is deflated like a balloon and appears to be drawn in dorsally between the wings near the large pallial opening (Fig. 1b). The web is withdrawn solely by ciliary ac- tions in 15 to 30 sec. Web production was not observed, but could be associated with a curious somersaulting behavior (Fig. 2) that is displayed by all of the euthecosome species | have ob- served. This behavior always takes place when swimming animals switch to a motionless, feeding posture (Fig. 1a). The animal slows its swimming speed, but at the same time in- creases wing motion and moves in a small arch roughly similar in dimension to its feeding web. At the apex, the animal moves in quick back and forth twists while continuing its path. Near the bottom of the arch when the shell is situated above the wings, the body quickly reserves position leaving the wings extended uppermost. Swimming motion stops immediately and the animal now hangs motionless in the water. The en- tire somersaulting sequence takes from eight to twelve sec to complete in all species that | have observed. Limacina helicina and L. retroversa (Fleming) sink slightly after somer- saulting, but attain neutral buoyancy within 5 sec. This con- dition then lasts for at least eight min (the longest observa- tion period), and is the only period when | have observed feeding webs in place. Neutral buoyancy, however, is also displayed by mating couples with no apparent aid from feeding webs or other mucus structures. On cruise ‘‘Endeavor’’ 182 Limacina helicina occurred in surface waters in a distinct layer between 5 and 28 m. More than 97% of the individuals | surveyed (n= 1200) were neutral- ly buoyant and motionless in their feeding posture. This percentage was similar on all dives made between 0800 hr and 1900 hr at various stations over 27 days. Feeding in- dividuals that were occasionally bumped by swimming Limacina showed no reaction to the contact and continued their quiescent posture. Several Limacina were even pulled a short distance in the water when the intruder became en- tangled in the web. Similar passive behavior by other feeding L. helicina was observed on four occasions when gammariid amphipods (>5 mm body length) blundered into the pteropod web. The amphipods immediately broke free of the web. After their encounter with the amphipod, two of the Limacina swam off and the other two remained quiescent but did not appear to set new webs during two minutes of further observation. GILMER: FEEDING BEHAVIOR OF THECOSOME PTEROPODS So Fig. 1a. Limacina helicina in subarctic waters. Lateral view of individual in motionless feeding posture (S, shell; MW, mucous feeding web) (scale bar = 4 mm); b, L. helicina, dorsal view, in final stages of withdrawing feeding web (MW, mucous feeding web) (scale bar = 4 mm); c, Cavolinia uncinata, with extended mucous feeding web (MW) and no pseudofeces retained off the posterior shell margin (scale bar = 10 mm); d, C. uncinata, with extended feeding web (MW) and aggregated mass (AP) of pseudofeces and fecal pellets retained on the posterior shell margin indicating an earlier web was set in the same location (scale bar = 10 mm); e, Corolla calceola from northwestern Atlantic slope water with large mucous feeding web in place (scale bar = 25 mm); f, C. ca/ceo/a with large mass of pseudofeces and feces after ingesting a feeding web (scale bar = 10 mm). FAMILY CAVOLINIIDAE (EUTHECOSOMATA) Within the upper 30 m, roughly 95% of the cavoliniids | observed maintained a quiescent feeding posture regardless of the time of day. The only exception to this is in turbulent mixing zones, such as langmuir cells or shear zones between warm and cold water masses (e.g. western edge of the Guif Stream and northwestern slope water interface), where our dive team has encountered vertical currents of approximate- ly 0.5 knot. Here cavoliniids are often abundant and rapidly swim to maintain their position, or are swept away in the cur- rent. These are exceptional circumstances, as most theco- somes retain their motionless, feeding posture within 3 m of 56 AMER. MALAC. BULL. 8(1) (1990) the surface on rough days (e.g. Beaufort wind force 7). Cavoliniids rapidly ingest their feeding webs in a man- ner similar to Limacina helicina. Webs are collected ventrally on the large expanse of the footlobes and funnelled into the mouth as condensed strings. At night, | have routinely ob- served Clio pyramidata Linné and Cuvierina columnella (Rang) withdraw their largest webs (Table 1) in 15 sec. Cavoliniids also display variable escape responses to the presence of a diver (Gilmer and Harbison, 1986). At one extreme of their behavior, individuals will flee from a small hand motion in- itiated up to 3 m away. Conversely, they will sometimes allow a diver to touch them several times before an escape response is induced. This sporadic escape behavior is characteristic of all thecosome species | have observed. Hand shading ap- plied to change the illumination on a feeding individual elicits no escape response unless associated with turbulence. | have observed Cavolina uncinata (Rang), C. triden- tata (Niebuhr), C. longirostris (Blainville), C. inflexa (Lesueur) and Diacria quadridentata (Blainville) withdraw feeding webs, and then enter a non-fishing period that lasts for observed periods of up to 12 min. During this period, no web is pre- sent but pseudofeces and fecal pellets are transported down the dorsal surface and are retained posteriorly (Fig. 1d), as described previously (Gilmer and Harbison, 1986). Photographs of C. uncinata, C. tridentata and D. quadriden- tata suggest they occupy one position long enough to set, fish, and ingest at least two webs. Figure 1c shows a specimen of C. uncinata with a web in place, but with no fecal matter or pseudofecal strings hanging from the posterior shell sur- face. Figure 1d shows another individual of this species with a web in place, but with fecal matter and pseudofecal strings present. This indicates a web was ingested and a new one set without swimming to a new location. Particle laden webs of five Cavolinia tridentata were observed during dives in northwestern Atlantic slope water. Four specimens left webs in place for five minutes, and one withdrew the web after three minutes, possibly because of diver turbulence. The latter specimen required almost one minute to withdraw its web. All individuals appeared to have produced webs in the same locations previously as fecal and pseudofecal material were present. These observations sug- gest Cavolinia and Diacria can remain in one location for at least 35 min to set and fish sequential webs for five min each, and digest each web during 12 min ‘‘non-fishing”’ periods. Occasionally a combined mass of fecal pellets and pseudofeces can be indentified in situ though | have never observed an intact abandoned web without having first disturbed an animal. All cavoliniid genera display the same somersaulting behavior as described for Limacina (Fig. 2). | have also ob- served an extended pattern of this behavior by several specimens of Creseis acicula (Rang) and Cavolinia longirostris. Initially, these individuals are motionless in feeding postures, but then sink away rapidly (approximately 5 cm/sec) for no apparent reason. After sinking 0.5 to 1.0 m, they somersault and again remain motionless. The somersault behavior in- Table 1. Comparison of maximum dimensions of shells and feeding webs (in mm) of euthecosome species by region in the North Atlantic (includes data from Gilmer and Harbison, 1986). Species Location Limacina helicina subarctic Atlantic McMurdo Sound, Antarctica Cuvierina columella north central Atlantic slope water N.W. Atlantic Clio pyramidata north central Atlantic slope water N.W. Atlantic Cavolinia longirostris north central Atlantic northern Sargasso Sea Gulf Stream axis slope water N.W. Atlantic C. uncinata Gulf Stream axis slope water N.W. Atlantic C. tridentata north central Atlantic northern Sargasso Sea Gulf Stream axis slope water N.W. Atlantic Diacria quadridentata Gulf Stream axis slope water N.W. Atlantic Canary Current northern Sargasso Sea ( ) No. of webs measured/No. with maximum dimension * excludes shell posterior to caudal septum visual estimation ae Shell length Max. Web Diameter (+ S.D.) 12.2 (diameter) 60 (24/3) 46 (diameter) 20 (3/1) 11.0* 120 (13/5) 11.0* 110 (5/3) 10.0 + 04 50 (10/8) 10.0 40 (5/2) 5.4 40 (7/3) 50 + 0.2 40 (5/1) 56 50 (2/1) 58 40 (3/2) 7.0 100 (7/2) 7.0 110 (9/5) 70 + 0.2 110 (3/3) 15.0 180 (2/1) 15.0 200 (1) 15.4 200 (2/2) 15.0 + 0.2 220 (6/4) 3.0 30 (2/2) 3.0 20 (2/1) 3.6 20 (3/1) GILMER: FEEDING BEHAVIOR OF THECOSOME PTEROPODS 57 volves the only swimming activity. | have observed several individuals each sink and somersault in this sequence up to four times in five min, with a net descent of approximately 3m. At the end of this sequence the animal either remains motionless with no apparent web for the duration of the obser- vation (up to five min), or swims off in a random direction out of the diving grid (>20 m). Small crustaceans often hover around feeding cavoliniids (Gilmer and Harbison, 1986), apparently attracted to the various surfaces as reported for larvacean mucous houses (Alldredge, 1972). During nighttime observations of Cavolinia uncinata (n=3), Clio pyramidata (n=15), and Cuvierina columnella (N=18), crustaceans (<1 mm length) were usually observed inside the web as it was ingested. Initially, the crustaceans were free swimming inside the spherical web area, but some appeared to be captured in the mucous walls as the web was withdrawn. Four specimens | observed and collected at night each had an estimated 20 to 30 crustaceans trapped in their webs as they withdrew them, and all had crustaceans in their guts (Table 2). Only smaller crustaceans are successfully captured. Hyperiid amphipods (3 to 4 mm) attracted by the dive lights were often caught in the webs, but easily broke free after several rapid swimming motions. Species occupying different water masses of the north Atlantic do not appear to alter their maximum web dimensions (Table 1). Cavolinia tridentata, C. longirostris, and C. uncinata were each observed with webs of similar dimension in cen- tral water masses of the temperate North Atlantic, the north- ern Sargasso Sea, the axis of the Gulf Stream, and in slope water along the northwestern Atlantic coast. FAMILY CYMBULIIDAE (PSEUDOTHECOSOMATA) The cymbuliids feed with enormous mucous webs that are often observed funnelled towards the footlobes as the animal lies below the web (Fig. 1e). Observations of Corolla and Gleba indicate that the web is slowly drawn in by the footlobes surrounding the mouth rather than the rapid, deflating balloon method observed with the spherical webs of euthecosomes. | have observed large numbers of Corolla calceola (Verrill) feeding in slope water regions of the northwest Atlantic. No apparent change in web size occurs during observations lasting up to 15 minutes. Food is ingested continuously and pseudofeces are released as long strands off the anterior side of the footlobes (see Lalli and Gilmer, 1989 for orientation in pseudothecosomes). | observed one specimen heavily laden with mucus and pseudofeces (Fig. 1f), suggesting that it had recently ingested a web. The mucus contained many copepods, larvaceans, and small diphyiid siphonophores in addition to the waste matter. It actively swam twelve meters horizontally before freeing itself from the mucus, and then swam downwards out of our diving range (> 30 m). | have no observations to indicate whether cymbuliids ever set sequential webs in one location. DISCUSSION Among oceanic suspension feeders that employ Fig. 2. Somersaulting behavior by Limacina helicina which occurs at the initiation of the motionless feeding posture. A similar behavior is displayed by all euthecosome genera. mucous structures, thecosomes appear to feed by a novel trapping strategy. Although some larvaceans (Alldredge, 1976) and all doliolids (Diebel, 1982) share a motionless feeding posture with thecosomes, these tunicates pump water through their mucous filters and feed on very small particles. Salps feed in a manner more analogous to thecosomes, but move constantly with pumping motions and expose their feeding webs to continuous new water (Madin, 1974; Harbison and Gilmer, 1976). Whether thecosomes remain in one location or sink slightly during feeding, they use no active transport of water through the web as in tunicates. Thecosomes ap- pear well suited to feeding by contract trapping of large motile organisms based on the occurrence of numerous crustaceans in and around the webs and in the gut contents of carefully collected specimens. Although nighttime observations are the most limited in number, they have provided the most information about the trapping ability of the thecosome webs. All individuals | have closely observed feeding at night (n=35) capture numerous small crustaceans as they withdraw the web. The few specimens also collected during these dives have all had small copepods in their guts (Table 2). These observations remain qualitative, since much of the web material and poten- tial food was undoubtedly lost during the capture. Intact crusta- ceans account for up to 25% by volume of intact items in guts of arctic Limacina helicina preserved in situ (Gilmer and Harbison, unpub. data). Copepods, however, are entirely ab- sent in the laboratory for as little as three hours after cap- ture. Small fragments of copepod exoskeleton are common in fecal pellets of L. helicina, especially segments of endopo- dites and thoraces. The estimated times for Cavolinia and Diacria to fish two webs is limited by my short observation periods. The actual fishing times and the number of webs set in any one 58 AMER. MALAC. BULL. 8(1) (1990) Table 2. Gut contents of cavoliniids collected on night dives with crustaceans observed inside their feeding webs. Species (Locality) Dominant taxa Max. dimension No. of (um) food items Cuvierina columnella (n=2) copepod naupli thorax <600 5 tintiniids lorica 120 14 (Northern Sargasso Sea) thecate dinoflagellates 60 18 Globigerina spp. 120 3 centric diatoms 30 24 Clip pyramidata (N=1) copepod juvenile thorax 1100 1 copepod nauplii thorax < 600 4 (Northern Sargasso Sea) tintiniids lorica 110 21 Radiolarians 150 4 centric diatoms 30 4 Cavolinia uncinata (n=1) copepod nauplii thorax <700 3 Limacina inflata juv. shell <300 2 (Florida Current) tintinnids lorica 140 6 thecate dinoflagellates 90 7 location are undoubtedly much greater, judging from the ex- tensive amount of pseudofeces that some individuals ac- cumulate. This quiescent behavior may explain how well developed hydrdoid colonies exist on the shells of some thecosomes. Kinetocodium danae Kramp often found on Diacria trispinosa (Blainville) (Lalli and Gilmer, 1989) has feeding polyps that could easily reach the web surface to prey upon attracted crustaceans. Hydroids | have observed on other thecosomes have feeding polyps either situated near the anterior portion of the shell, nearest the host feeding web, or have stalked feeding polyps that could reach the host web from other regions of attachment on the shell. The large, sheet-like webs used by pseudothecosomes and the slow methodical fashion of ingesting them is easily comparable to the feeding style of the vermetid prosobranchs (Hughes and Lewis, 1974; Hughes, 1978). In contrast, the spherical webs used by euthecosomes and their ability to rapidly ingest them appears to be unique among marine animals that feed with mucous structures. Much of the feeding behavior of thecosomes remains obscure. For instance, webs produced by Limacina appear to arrest sinking and provide neutral buoyancy during feeding, although mating couples display neutral buoyancy as well and have no feeding webs in place. Secondly, the somersaulting action displayed by all euthecosomes does not appear to coin- cide with the setting of a web, but always occurs prior to initi- ation of the feeding posture. Finally, it is unclear how free swimming copepods penetrate the walls of the euthecosome webs without adhering to them. Hopefully, future in situ obser- vations and collections will help to answer these questions and may ultimately help determine the precise, thecosome feeding strategy. ACKNOWLEDGMENTS | wish to thank Dr. G. R. Harbison for financial support and for allowing me time to make these observations during cruises. | also thank G. R. Dietzmann for valuable assistance during night dives. Research support was provided by N.S.F. grants OCE 85-16083, OCE 87-46136, and DPP 86-13388 to G. R. Harbison. This is Harbor Branch Oceanographic Institution Contribution No. 748. LITERATURE CITED Alldredge, A. L. 1972. Abandoned larvacean houses: a unique food source in the pelagic environment. Science 177:885-887. Alldredge, A. L. 1976. House morphology and mechanisms of feeding in the Oikopleuridae (Tunicata, Appendicularia). Journal of Zoology, Proceedings of the Zoological Society of London 181:178-188. Boysen-Ennen, E. and U. Piatkowski. 1988. Meso- and macro- zooplankton communities in the Weddell Sea, Antarctica. Polar Biology 9:17-35. Deibel, D. 1982. Laboratory-measured grazing and ingestion rates of the salp, Thalia democratica (Forskal) and the doliolid, Dolioletta gegenbauri (Uljanin) (Tunicata, Thaliacea). Journal of Plankton Research. 4:189-201. Foster, B, 1987. Composition and abundance of zooplankton under the spring sea-ice of McMurdo Sound, Antarctica. Polar Biology 8:41-48. Gilmer, R. W. 1972. Free floating mucous webs: a novel feeding adap- tation for the open ocean. Science 176:1239-1240. Gilmer, R. W. 1974. Some aspects of feeding in thecosomatous pteropod molluscs. Journal of Experimental Marine Biology and Ecology 15:127-144. Gilmer, R. W. and G. R. Harbison. 1986. Morphology and field behavior of pteropod molluscs: feeding methods in the families Cavoliniidae, Limacinidae and Peraclididae (Gastropoda: Thecosomata). Marine Biology 91:47-57. Hamner, W. M. 1975. Underwater observations of blue water plankton: Logistics techniques, and safety procedures for divers at sea. Limnology and Oceanography 20:1045-1051. Harbison, G. R. and R. W. Gilmer. 1976. The feeding rates of the pelagic tunicate, Pegea confoederata and two other salps. Lim- nology and Oceanography 21:517-528. Hughes, R. N. and A. H. Lewis. 1974. On the spatial distribution, feeding and reproduction of the vermetid gastropod Dendropoma maximum. Journal of Zoology, Proceedings of the Zoological Society of London 172:531-547. GILMER: FEEDING BEHAVIOR OF THECOSOME PTEROPODS Hughes, R. N. 1978. The biology of Dendropoma corallinaceum and Serpulorbis natalensis, two South African vermetid gastropods. Zoological Journal of the Linnean Society 64:111-127. Ishimaru, T., S. Nishida, and R. Marumo. 1988. Food size selectivity of zooplankton evaluated from the occurrence of Coc- colithophorids in the guts. Bulletin of Plankton Society of Japan 35:101-114. Jorgensen, C. B. 1966. Biology of Suspension Feeding. Oxford: Pergamon. 357 pp. Lalli, C. M. and R. W. Gilmer. Pelagic Snails: The Biology of Holoplanktonic Gastropod Mollusks. Stanford University Press, Stanford, California. 259 pp. Madin, L. P. 1974. Field observations on the feeding behavior of salps (Tunicata: Thaliacea). Marine Biology 25:143-148. Bie) Morton, J. E. 1954. The biology of Limacina retroversa. Journal of the Marine Biological Association of the United Kingdom 33:297-312. Richter, G. 1977. Jager, Fallensteller and Sammler (Zur Ernahrung planktischer Schnecken). Natur und Museum 107:221-234. Richter, G. 1983. Lebensformen und Nahrungsketten der Hochsee. Natur und Museum 113:131-138, 166-177. Silver, M. W. and K. W. Bruland. 1981. Differential feeding and fecal pellet composition of salps and pteropods, and the possible origin of the deep-water flora and olive-green ‘‘cells’’. Marine Biology 62:263-273. Date of manuscript acceptance: 25 November 1989 MATING BEHAVIOR AND SPAWNING IN TWO NEUSTONIC NUDIBRANCHS IN THE FAMILY GLAUCIDAE ROBIN M. ROSS LANGDON B. QUETIN MARINE SCIENCE INSTITUTE, UNIVERSITY OF CALIFORNIA AT SANTA BARBARA, SANTA BARBARA, CALIFORNIA 93106, U.S.A. ABSTRACT We observed the mating behavior and egg production rates of G/aucus atlanticus (Forster, 1777) and Glaucilla marginata (Bergh, 1868), neustonic nudibranchs of the family Glaucidae, collected from the western Pacific Ocean near Australia. Although the same sequence of mating behaviors occurred in both species, the timing of these behaviors and mating duration were different. The entire mating sequence from ‘‘kiss’’ to separation lasted about one hour in G. atlanticus and about one minute in G. marginata. Morphological differences that could be associated with this difference are discussed. Glaucids release short gelatinous strings of eggs at varying frequencies. Egg production rates in both species were directly related to food availability, though both species continued to produce eggs at lower rates for several days with little or no food. Both glaucids exhibited several characteristics of planktotrophic development: eggs were small (70 nm by 80 um), embryonic developmental times were short (2.5 to 3.0 days at 19°C), and veligers swam longer than a week. Only a small number of nudibranchs spend their en- tire life in the water column, and these species have few evolu- tionary adaptations to pelagic life. Several characteristics, such as no heavy shell and some swimming ability, that could be considered advantages in a pelagic habitat are also found in benthic nudibranchs. Some pelagic nudibranchs are flat- tened or otherwise modified and do not resemble benthic nudibranchs, but two species of Glaucidae, G/aucus atlanti- cus (Forster, 1777) and Glaucilla marginata (Bergh, 1868), are easily identified as eolid nudibranchs by their cerata. Like other nudibranchs, the Glaucidae are reciprocal hermaphro- dites, although little is known about their mating behavior and spawning characteristics (Lalli and Gilmer, 1989). Both glaucids are neustonic, occurring in the surface layer of tropical oceans. G/aucus atlanticus is circumtropical. Glaucilla marginata is restricted to the tropical Pacific Ocean. Both species float upside down on the air-seawater interface and neither is a good swimmer. Their distribution is primarily controlled by winds, as is also true of their cnidarian prey, Physalia, Velella, and Porpita (Lalli and Gilmer, 1989). Like other neustonic animals, they are countershaded, blue to pur- ple on their ventral surface which faces up, and white or silvery on their dorsal side facing down. Thompson and Bennett (1970) discovered that G. atlanticus stores nematocysts derived from their prey in cnidosacs at the tips of the cerata. The utilization of nematocysts as a defense against predators is well known in eolid nudibranchs, although the processes in- volved in the control of the discharge of these nematocysts are not well understood (Todd, 1981; Thompson and Bennett, 1970). There are several differences between Glaucus atlanticus and Glaucilla marginata that are important to a discussion of their mating behavior (Lalli and Gilmer, 1989). G. atlanticus is the larger of the two species, reaching a max- imum reported length of 43 mm (Miller, 1974). It has a long slender foot, and a long, strong, contractile penis armed with a chitinous spine. The long cerata are arranged in single rows in three to four clusters projecting from lobes on the sides of the body. The central ceras is much longer than those on the sides of the cluster. A single individual can have up to 85 cerata. G. marginata is smaller than G. atlanticus, with previous recorded sizes up to 12 mm, and up to 18 mm in this study. The penis is not armed and the cerata are arranged in multiple layers in four clusters. G. marginata can have twice as many cerata as G. atlanticus. In both species the reproduc- tive aperture is just posterior and level with the bottom of the first right ceratal arch. Despite their widespread distribution, observations and experiments on live Glaucus atlanticus and Glaucilla marginata are rare, perhaps because they are usually damaged when American Malacological Bulletin, Vol. 8(1) (1990):61-66 61 62 AMER. MALAC. BULL. 8(1) (1990) when they are collected with nets. In this paper we will com- pare the pattern and timing of mating behavior in these two pelagic nudibranchs. We will also describe and quantify spawning and egg production in both species, and the effect of starvation on egg production in G. marginata. Some obser- vations of the egg ribbons and embryonic developmental times will be described. However, veligers did not complete metamorphosis to adult in our laboratory. METHODOLOGY COLLECTION AND MAINTENANCE OF ANIMALS We found glaucid nudibranchs in the surf zone and blown onto the beaches of New South Wales, Australia, south of Sydney, in the austral summer of 1979-1980. G/aucus atlan- ticus were collected in early December from Morulga and Bawley Beaches, and Glaucus marginata were collected in early December from North Kioloa Beach, and early March from South Kioloa Beach. No glaucids were found on a search of the sand and surf zones of a group of beaches in this area from mid-December until late February. The beaches were regularly inspected by the manager of the Kioloa Field Sta- tion of the Australian National University (ANU). Surf zone temperatures were 20.5°C on 11-12 December, and 22.0°C on 2 March. We scooped glaucids from the surf into containers of seawater and returned them to the laboratory at ANU. At ANU, adult glaucids were kept in three 10.0 / rec- tangular seawater aquaria (two per aquarium) at room temperature (about 20°C) or at 19°C in shallow 1.0 / plastic boxes. Seawater in the containers was changed two to three times a week. Both glaucids were fed either fish food or pieces of Physalia sp. collected from the same beaches and frozen until use. MATING BEHAVIOR We observed the sequence and timing of mating behavior for three pairs of Glaucus atlanticus and six pairs of Glaucilla marginata. Mating behavior was observed and timed either in the aquaria or in large shallow containers. SPAWNING BEHAVIOR AND EGG PRODUCTION RATES Glaucids release strings of eggs, and do not lay down an egg mass like other eolid nudibranchs. Dimensions of egg capsules from seven strings released by G/aucus atlanticus were measured using a compound microscope with a calibrated micrometer. In addition, we measured the lengths of 20 strings and the number of eggs in each string for strings released four days after collection. Time intervals between release of strings were recorded for five individuals in order to calculate daily fecundities. Fresh spermatozoa were also measured using a compound microscope. For Glaucilla marginata, egg capsule size was measured in 11 strings released by adults collected in March. Ten Glaucilla marginata collected in March were used to evaluate the response of egg production rate to a decrease in food availability. Each adult was isolated immediately after collection, and removed from the jar and placed in fresh seawater at 12, 24 and 36 hr after collection. All the strings and the number of embryos per string in at least 15 strings were counted for each individual for each of the three 12 hr periods. The length of the string was also measured for the first 12 hr period. At the end of 36 hr, the total length and wet weight in grams of each individual were measured. We estimated the egg production rate for freshly collected animals and for animals kept without food for periods up to 36 hr. DEVELOPMENT Some egg strings were maintained in aerated glass jars at 19°C to determine embryonic developmental times, and to observe survival and behavior of the larvae after hatching. The age of the strings was known to within 12 hr. Embryos of both glaucids hatched. Veligers were inspected for con- tinued survival and swimming ability until they sank to the bottom of the jars and died. RESULTS MATING BEHAVIOR Mating in both Glaucus atlanticus and Glaucilla marginata involved a sequence of predictable and stereotypic behaviors. These behaviors were the same for both species and began when conspecific glaucids contacted each other. Although there was some variation in the exact timing of each behavior type in the sequence, total duration of mating was similar for all pairs of the same species. The first in the sequence of behaviors was the relatively brief ‘‘kiss’’. Partners oriented so that their mouths and ven- tral surfaces were touching, and the heads usually sub- merged. Shortly after the mouths joined, the penises emerged. The penises were greatly extended, and sometimes the end of one penis was wrapped around the other individual. The second major behavior was the intertwining of the two penises, and the two individuals began to couple. Initially the mouths separated, but the pair were still oriented ventral sur- face to ventral surface (Fig. 1a). Both glaucids arched the body and the cerata clusters toward their dorsal surfaces, keeping the cerata away from each other. During mating we observed individuals flinch and arch away when touched by the cerata of the mate. This observation indicated that a ceras could be stimulated by this contact to eject the cnidophages in the cnidosacs at the tip of the ceras, and the partner could be stung by the nematocysts from the cnidophage. Shortly after coupling began, the pair changed to a side by side orienta- tion with the ventral surfaces on the air-water interface and the penises lying between the two parallel bodies (Fig. 1b). During coupling the pair could be either head to head or head to tail, and often switched from one to the other once or twice during this time period. The pair lay quietly between changes in orientation. Penises were loosely intertwined until two-thirds through the coupling period when the penises became tight- ly intertwined (Fig. 1c). When the pair began to separate, the coil untwisted, the penises showed some thickening as they started to retract, and the nudibranchs flexed their bodies (Fig. 1d). Total retraction of the penises required a much longer ROSS AND QUETIN: MATING IN NEUSTONIC NUDIBRANCHS 63 Fig. 1. Glaucus atlanticus. Sequence of mating behaviors: (a) 7 min after initial contact, ventral surface to ventral surface; (b) 37 min after initial contact, coupled, penises in loose coil, ventral surfaces at the air/water interface; (c) 47 min after initial contact, coupled, penises in tight coil; (d) 58 min after initial contact, separated, penises beginning to retract, note thickening. Horizontal field width = 40 mm. period of time than separation of the two individuals. Although the pattern of behavior for the two species of glaucids was identical, the timing of each behavior and the duration of mating was very different. For Glaucus atlanticus, the ‘‘kiss’’ lasted about four min. Between four and nine min into the sequence the pair reoriented so they were parallel, usually head to tail, with the penises loosely coiled between them. The coil tightened 36 to 41 min into the sequence, re- mained tight for four to nine min, then loosened again. Separation took four to nine min and was complete 43 to 59 min after the initial joining. Retraction of the penis took up to an hour (19, 52 and 61 min). For Glaucilla marginata, the mating sequence was substantially shorter. The ‘‘kiss’’ lasted five sec. Pairs then twisted, and oriented parallel to each other with their ventral surfaces facing up and penises twisted around each other. Coupling lasted 60 sec (range 50 to 70 sec, SD = 84 sec). Separation was fast, and the penises retracted quickly, within 21 sec (range 10 to 35 sec, SD = 86 sec). The total time for mating was 65 sec, about 2% of the time taken by G/aucus atlanticus. The ability to mate again with the same or a different individual immediately after mating was investigated in both species. A pair did not remate after retraction of the penises, even if they were pushed together. But we observed that a third non-mated glaucid will extrude its penis in the presence of a mating pair. In Glaucilla marginata we saw these new in- dividuals mating with recently mated glaucids. SPAWNING AND EGG PRODUCTION Both species released their eggs in straight strings about 0.3 mm in diameter that sank slowly (about 50 m d-"). Strings were from 5.0 to 17.5 mm long for G/aucus atlanticus, and from 2.0 to 6.4 mm for Giaucilla marginata. In both species ova were individually encapsulated (primary membrane). Egg capsules were oval, and evenly spaced slightly less than one diameter apart within the egg string. A thin transparent tube (secondary membrane) surrounded the egg capsules within the mucous string (Fig. 2). Egg capsules of the two glaucids were similar in size: G. atlanticus, 60 to 75 um wide and 75 to 97 um long; and G. marginata, 58 to 67 um wide and 74 to 82 um long. Sperm of G. atlanticus were long and slender (129 pm by 0.9 um). Daily or hourly fecundity was a function of the number of embryos in a string, and the rate of string production. The number of eggs per string was a linear function of the length of the string, but smaller G/laucilla marginata released strings that were less than half the length and contained less than half the embryos of those released by G/aucus atlanticus (Fig. 3). The time interval between strings was also different for 64 AMER. MALAC. BULL. 8(1) (1990) the two species. Even after three to four days in the laboratory G. atlanticus produced 4-6 strings/hr (average for five in- dividuals was 3.8 strings/hr). G. marginata released 8.6 strings/hr during the first 12 hr after collection, twice the fre- quency of G. atlanticus. However, the frequency dropped to less than two strings an hour during the next 12 hr. Since the number of embryos per string ranged from 36 to 96, G. atlan- ticus released from 3300 to 8900 embryos/day, even after three to four days in the laboratory. The fecundity for freshly collected G. atlanticus was about the same, from 1850 to 9250 embryos/day. The relationship between size and egg production and the effect of starvation on egg production was quantified for Glaucilla marginata. First, although total length (TL) and wet weight of the ten individuals measured were significantly cor- related, egg production (EP in number/hr) was more closely related to total length (EP = 597.2 TL - 571.2, r2 = 0.56) (Fig. 4) than to wet weight (r2 = 0.27). The increase in egg pro- duction with increasing size was a function of both increased string production (strings/hr, rs = 0.705) and closer packing of egg capsules in a string (eggs per string, rs = 0.675), but not an increase in the length of the string with total length of the glaucid (rg = 0.421, ns) (Spearman’s rank correlation coefficient, p = 0.05; Siegel, 1956). The total length of G. marginata ranged from 12.6 to 179 mm, string production ranged from 3.3. to 12.9 strings/hr, and the average number of eggs per string ranged from 25.5 to 41.7 (n = 15 strings per individual). The average string length of the same 15 strings ranged from 3.8 to 5.1 mm, with 95% confidence in- tervals of 0.2 to 0.5 mm. Differenceg in average string length between individuals was not a significant factor in the increase in egg production with size. Egg production rates of starved Gi/aucilla marginata decreased significantly after 12 hours (Xr2 = 15.2, Friedman two-way ANOVA, p > 0.001; Siegel, 1956), but then remained the same during the remainder of the experiment (Fig. 5). Egg production rates in the first 12 hr after collection were 3.5 times rates in the second and third 12 hr intervals after collection, cs. Fig. 2. Glaucus atlanticus. Development at 19°C; embryos about 24 hours after release, multiple cell stage [e, egg capsule (primary mem- brane); t, internal tube (secondary membrane)]. Horizontal field width = 300 um. 100 80 60 40 20 Number of Eggs ) 5 10 15 20 Egg String Length (mm) Fig. 3. Relationship between egg string length and the number of eggs per string for Glaucus atlanticus (¢ = one pair) and Glaucilla marginata (o = 4 individuals). which were the same (U = 43, Mann Whitney U, p = 0.05, Siegel, 1956). G. marginata was still producing eggs after 36 hr with no food. Lower egg production rates in starved Gjlaucilla marginata were due to fewer strings produced/hr and fewer egg capsules per string. String production decreased to 36 to 39% of initial rates, and egg capsules per string to 59 to 71% of initial values. Although the decrease in egg capsules per string could be due to shorter strings or to greater spac- ing between the egg capsules or both, four of the ten G. marginata produced some very short strings with only one or two eggs per string during the second and third 12 hr periods of starvation. Thus string length could decrease dur- ing starvation. DEVELOPMENT Embryos of G/aucus altanticus began to divide after a few hours at 19°C. At about 24 hr, the embryos were multicellular. The egg capsules were still separated within the string, but the secondary membrane was thinner and con- stricted between the embryos (Fig. 2). Between 48 and 60 hr, embryos had beating cilia (trochophore), and the secon- dary membrane began to disintegrate. After about three days at 19°C, the half-shelled veligers moved slowly through the mucous string and swam away. The shell was initially ovoid but became coiled after a few days. Starved veligers of G. atlanticus swam continuously for seven to 11 days after hatching before sinking to the bottom of the containers, and dying before metamorphosis into juveniles. The shell was 89 pm by 104 pum, larger than the egg capsule. Embryonic developmental time for Glaucilla marginata was about the same, 2.5 to 3.0 days at 19°C. These embryonic developmental times were slightly slower than found for G. atlanticus veligers maintained at 25°C (2.0 days, Bebbington, 1986). The bilobed veligers of G. marginata swam continuously for 33 days after hatching before they died without metamorphosing. The shell was 96 um by 119 um. Thus both species showed characteristics of plankto- trophic development: small eggs, short embryonic ROSS AND QUETIN: MATING IN NEUSTONIC NUDIBRANCHS 65 600 400 200 Egg Production (number per hour) 1.2 1.3 1.4 1.5 1.6 1.7 1.8 Total Length (cm) Fig. 4. Glaucilla marginata. Relationship between animal size, total length, and egg production 12 hr after collection. 600 c = So 8 6 = 400 3S — oO o 3 a a 2 200 DE DoD 3 Ww £ 0 0 10 20 30 40 Time (h) Fig. 5. Glaucilla marginata. Decrease in egg production over time with no food available. Points are average egg production rates for three 12 hr intervals after collection for 10 individuals. developmental times, and veligers that spend several weeks swimming in the plankton. Veligers of Glaucus marginata sur- vived three times longer in the laboratory than did veligers of Glaucus atlanticus, but the veligers of G. marginata were given a mixture of marine phytoplankton which could have increased their survival times. We were unable to stimulate metamorphosis, but are unsure whether the inability of the veligers to metamorphose was due to inadequate diets or lack of the appropriate substrate as is necessary for most eolid nudibranchs (Harrigan and Alkon, 1978; Thompson and Brown, 1984). DISCUSSION Glaucus atlanticus found on the beaches of New South Wales in Australia were of the typical color pattern and within the size range found elsewhere in the world (Bennett, 1836; Bieri, 1966; Miller, 1974; Thompson and McFarlane, 1967). Glaucilla marginata have been recorded only once before from Australian waters (Thompson and Bennett, 1970). G. marginata collected in this study were 18 mm in length, much larger than previously reported (Thompson and Bennett, 1970), and the ventral surface was not brown but deep purple - similar to G. atlanticus. Differences in coloration could be due to dif- ferences in diet between the two groups of G. marginata, as found for other pelagic nudibranchs (Lalli and Gilmer, 1989). We never found the two species of glaucids together, but both species were always found with some of their cnidarian prey. Previous reports of mating behavior are sparse for Glaucus atlanticus and nonexistent for Glaucilla marginata (Lalli and Gilmer, 1989). Possibly because he did not observe the complete sterotypic mating sequence of G. atlanticus, Beb- bington (1986) stated that G. at/anticus paired laterally or ven- trally during copulation in the laboratory. We found that G. atlanticus pairs both laterally and ventrally, but at different times during the same mating sequence. Mating in Glaucus atlanticus was exceptionally long compared to other pelagic nudibranchs, whereas mating dura- tion in Glaucilla marginata was similar to other species of pelagic nudibranchs, one to fifteen minutes (Lalli and Gilmer, 1989). Maximizing the reproductive potential of each en- counter may be particularly important to a pelagic species that must depend on chance encounters to find a mate. In the pelagic realm there are several options for exchanging large amounts of sperm: an exchange of spermatophores as found in heteropods and thecosomes; prolonged mating as found in gymnosomes and G. atlanticus; mating with many partners sequentially in swarms or rafts as is true of one species of pelagic dendronotacean (Lalli and Gilmer, 1988). For the neustonic glaucids, wave and wind action at the surface make prolonged mating difficult. First, wave ac- ton will tend to separate partners. Second, both glaucids utilize nematocysts (Lalli and Gilmer, 1989). Contact with a mate could stimulate the contraction of the muscle complex around the cnidosac and the release of the cnidophage and nematocysts into the water, thus stinging the partner. In eolid nudibranchs, the nematocysts are in cnidophage cells inside the cnidosac. When the muscles surrounding the cnidosac contract, the cnidophage is ejected through the cnidopore or the epithelium at the tip of the ceras. If the cnidophage mem- brane ruptures on release, the nematocysts usually discharge (Greenwood and Mariscal, 1984). Stimulation of special neurosensory cilia which are concentrated at the ceras tip could cause contraction of the cnidosac wall (Todd, 1981). Mere pinching of the ceras with metal forceps (Thompson, 1976) or pressure on a cover slip (Greenwood and Mariscal, 1984) will stimulate ejection of cnidophage cells in many eolids. Both glaucid species actively avoid each other’s cerata, and individuals flinch when touched by the cerata of their part- ner during mating. For benthic eolid nudibranchs neither problem occurs. Wave action is minimal, and the cerata tips are oriented dorsally not to the side as is true of the glaucids. Morphological adaptations in Glaucus atlanticus ap- pear to have solved both of these difficulties. The chitinous spine on the penis of G. atlanticus may help prolong contact in the face of wave action, and thus may be a singularly im- portant morphological adaptation to long mating times (Miller, 1974). In contrast, pelagic nudibranchs with shorter mating times have unarmed penises. In addition, in G. atlanticus the cerata are fewer, longer and in a single layer, so are more easi- ly held away from the partner than in G/aucilla marginata. 66 AMER. MALAC. BULL. 8(1) (1990) Prolonged copulation may mean more sperm are exchanged, filling the seminal receptacle. If prey are available, fertile egg production can continue longer before finding another mate is necessary. This could be advantageous when finding a mate depends primarily on physical forces in the ocean and not active searching. With the exception of G. marginata, the other species of pelagic nudibranchs are more active swim- mers, and/or mate in swarms or cling to a surface flotsam (Lalli and Gilmer, 1989), so finding a mate is not as dependent upon physical forces in the ocean. Our observations of the frequency of egg string pro- duction and of the number of eggs per string for Glaucus atlan- ticus were different from those of Bebbington (1986) and Mac- nae (1954). Although Bebbington saw fewer ova per string, strings were produced much more frequently, leading to fecundity estimates about six times ours. Bebbington gives no information about how long his glaucids were kept in the laboratory or their feeding conditions. Macnae (1954) also found fewer ova per string and greater spacing between em- bryos than we did, but did not estimate frequency of string production. We observed that egg production in G. atlanticus increased within hours of ingesting a slurry of homogenized Physalia. There was no interval between strings, instead of the 10 to 15 min interval between strings found when this species was fed fish food. These combined observations on egg production in G. atlanticus in conjunction with the results of the experiment on the effect of starvation on egg produc- tion in Glaucilla marginata suggest that egg production in glaucids may be closely coupled to their recent feeding history. Thus the inter- and intraspecific differences in rates of egg production observed for the two glaucids may be due to differences in their immediate nutritional histories. There appear to be no major adaptative differences in reproduction and development between these neustonic eolids and their benthic relatives. Egg strings have the same basic form (hollow, cylindrical, capsule-filled cord) as benthic eolids, but the ribbon floats free as a short uncoiled string of eggs such as found in other pelagic nudibranchs instead of being attached on one side (Lalli and Gilmer, 1989). Many other species have only one ova per capsule (Hurst, 1967). The presence of a secondary membrane is not common, but its function is unknown (Eyster, 1986). Glaucids have relatively high fecundities compared to benthic eolid nudibranchs of the same size range (Harris, 1975; Rivest, 1978) and compared to the only other pelagic nudibranch, Phylliroe bucephala (Peron and Lesueur, 1810), for which we have fecundity estimates (Lalli and Gilmer, 1989). High fecundities and a direct linkage of egg production to food availability are valuable characteristics for these neustonic glaucids which have an unpredictable food source that occurs in large quantities. ACKNOWLEDGMENTS This research was conducted while L. B. Quetin was a Queen Elizabeth || Fellow in Marine Biology and R. M. Ross was a Visiting Fellow in the Research School of Biological Sciences at the Australian National University. We would like to thank Dr. E. E. Ball and the staff and scientists in RSBS and at the Kioloa Field Station for their sup- port and for the use of the facilities. Comments from two anonymous reviewers and a discussion with L. Harris helped clarify our thoughts. LITERATURE CITED Bebbington, A. 1986. Observations on a collection of Glaucus atlan- ticus (Gastropoda, Opisthobranchia). Haliotis 15:73-81. Bennett, G. 1836. Observations on a species of G/aucus, referred to the Glaucus hexapterygius, Cuvier. Proceedings of the Zoological Society of London 1836:113-119. Bieri, R. 1966. Feeding preferences and rates of the snail, /anthina prolongata, the barnacle, Lepas anserifera, the nudibranchs, Glaucus atlanticus and Fiona pinnata, and the food web in the marine neuston. Publications of the Seto Marine Biological Laboratory 14:161-170. Eyster, L. S. 1986. The embryonic capsules of nudibranch molluscs: Literature review and new studies on albumen and capsule wall ultrastructure. American Malacological Bulletin 4:205-216. Greenwood, P. G. and R. N. Mariscal. 1984. The utilization of cnidarian nematocysts by eolid nudibranchs: nematocyst maintenance and release in Spurilla. Tissue & Cell 16:719-730. Harrington, J. F. and D. L. Alkon. 1978. Larval rearing, metamorphosis, growth and reproduction of the eolid nudibranch Hermissenda crassicornis (Eschscholtz, 1831) (Gastropoda: Opistho- branchia). Biological Bulletin 154:430-439. Harris, L. G. 1975. Studies on the life history of two coral-eating nudibranchs of the genus Phestilla. Biological Bulletin 149:539-550. Hurst, A. 1967. The egg masses and veligers of thirty north-east Pacific opisthobranchs. Veliger 9:255-288. Lalli, C. M. and R. W. Gilmer. 1989. Pelagic Snails: The Biology of Holoplanktonic Gastropod Mollusks. Stanford University Press, Stanford, California. 259 pp. Macnae, W. 1954. On some eolidacean nudibranchiate molluscs from South Africa. Annals of the Natal Museum 13(1):1-52. Miller, M. C. 1974. Aeolid nudibranchs (Gastropoda: Opisthobranchia) of the family Glaucidae from New Zealand water. Zoological Journal of the Linnean Society 54(1):31-61. Rivest, B. R. 1978. Development of the eolid nudibranch Cuthona nana (Alder and Hancock, 1842), and its relationship with a hydroid and hermit crab. Biological Bulletin 154:157-175. Siegel, S. 1956. Nonparametric Statistics for the Behavioral Sciences. McGraw-Hill Book Co., New York. 312 pp. Thompson, T. E. 1967. Direct development in a nudibranch, Cadlina laevis, with a discussion of developmental processes in Opisthobranchia. Journal of the Marine Biological Association of the United Kingdom 47:1-22. Thompson. T. E. 1976. Biology of Opisthobranch Molluscs. Volume |. The Ray Society, London. 207 pp. Thompson, T. E. and |. Bennett. 1970. Observations on Australian Glaucidae (Mollusca: Opisthobranchia). Zoological Journal of the Linnean Society 49(3):187-197. Thompson, T. E. and G. H. Brown. 1984. Biology of Opisthobranch Molluscs. Volume II. The Ray Society, London. 229 pp. Thompson, T. E. and |. D. McFarlane. 1967. Observations on a col- lection of Glaucus from the Gulf of Aden with a critical review of published records of Glaucidae (Gastropoda, Opisthobran- chia). Proceedings of the Linnean Society of London 178(2):107-123. Todd, C. D. 1981. The ecology of nudibranch molluscs. Oceanography and Marine Biology, An Annual Review 19:141-234. Date of manuscript acceptance: 25 March 1990 BIPOLAR VARIATION IN CLIONE, A GYMNOSOMATOUS PTEROPOD RONALD W. GILMER MARINE SCIENCE DIVISION, HARBOR BRANCH OCEANOGRAPHIC INSTITUTION, FT. PIERCE, FLORIDA, 34946, U.S.A. CAROL M. LALLI DEPARTMENT OF ZOOLOGY, UNIVERSITY OF BRITISH COLUMBIA, VANCOUVER, BRITISH COLUMBIA, CANADA VET 2 AQ ABSTRACT The gymnosome Clione inhabits waters of the Arctic, North Pacific, North Atlantic, South Atlantic and Antarctic Oceans. This study reexamines bipolarity in this genus, utilizing new morphological tech- niques as well as observations of living animals in all locales. Differences in exernal anatomy are readily apparent in both larval and adult Clione from northern and southern areas, and scanning electron microscopy has revealed differences in the number and size of hooks in animals from different regions. The configuration of the radula differs between specimens from the northern and southern hemispheres, and Clione from the Antarctic lacks median radular teeth, in contrast to the well developed, sickle- shaped median teeth present in Clione from northern areas. Clione acts as a food specialist throughout its range, feeding either on the bipolar thecosome species Limacina helicina (Phipps) in cold waters or L. retroversa (Fleming) in temperate waters. Differences in maximum adult size of Clione can be related to the size of available prey. Although there do not appear to be differences in the feeding behavior of Clione from different areas, there are slight differences in swimming and escape behavior, particularly between specimens of Clione from northern and southern regions. The results of these anatomical and behavioral observations support the validity of recognizing two species of Clione, C. limacina (Phipps) in the northern hemisphere and C. antarctica (Smith) in the Antarctic and South Atlantic oceans. The gymnosome now called Clione limacina (Phipps) was originally illustrated in a publication by Martens in 1675, from specimens collected in waters off Spitzbergen and Greenland. However, Captain C. J. Phipps is credited with the first taxonomic description, which appeared in his 1774 book entitled A Voyage Towards the North Pole. Since these early descriptions, many publications have dealt with this species, the most abundant gymnosome in temperate and cold northern waters. It has been recognized as an important food for baleen whales and certain commercial species of fish (see Lalli and Gilmer, 1989). Since the late 1700s, a species of Clione has also been known to exist in cold waters of the southern hemisphere. Ear- ly illustrations and taxonomic history of the northern and southern Clione can be found in Rang and Souleyet (1852) under the respective names of Clio borealis (Rang and Souleyet) and C. australis (Rang and Souleyet). The southern Clione has been variously considered as a variety, form, or subspecies of C. limacina [see van der Spoel (1976) for synonymy], or as a separate species, C. antarctica, described by Smith in 1902. The southern Clione is less well known than the Arctic/Subarctic species, but preserved specimens were admirably described by Eliot (1907), who pointed out several anatomical differences between C. antarctica (Smith) and northern C. limacina. Additional anatomical studies of C. antarctica have been made by Meisenheimer (1906), Massy (1920, 1932) and Pruvot-Fol (1932), among others. Clione limacina is found throughout the central Arctic Ocean, and extends in Subarctic waters southward into both the North Atlantic and North Pacific to approximately 30° to 40° North latitude. The southern hemisphere Clione is found circumglobally in Subantarctic and Antarctic waters, extend- ing northward to about 40°. Because there are no connec- ting populations in tropical areas, these gymnosomes have American Malacological Bulletin, Vol. 8(1) (1990):67-75 67 68 AMER. MALAC. BULL. 8(1) (1990) been considered as one of several examples of bipolarity in plankton (Ekman, 1953). Our objectives here included direct comparisons of the morphological and behavioral differences between northern and southern populations of Clione in order to investigate their taxonomic distinction. We have compared our behavioral observations of several hundred living animals made during the Hudson 70 Expedition to Antarctic waters in 1970 by one of us (CML) and during two visits to Antarctica in 1987 and 1988 by the other of us (RWG), to those of living Clione observed in both the North Atlantic and North Pacific over a period of many years. We have also made anatomical and histological comparisons of specimens of Clione from several locations, using scanning electron microscopy to investigate detail not noted previously. METHODS Specimens of Clione were collected by plankton nets or by scuba divers at various locations throughout its range. Behavioral observations were noted in situ by divers or in laboratory aquaria. For anatomical studies, specimens representing a range of size categories were narcotized with MS222 (Ethyl m-aminobenzoate) and then preserved in either Bouin’s fluid or 4% formaldehyde. Radulae and hooks were dissected from specimens, and extraneous tissue was removed by soaking in a dilute bleach solution. After rinsing in an alcohol series, the buccal structures were mounted on stubs with double stick tape, coated with gold, and examined on either a Cambridge Stereoscan 250 or a Novascan 30 scanning electron microscope. In all, radulae were examined from over 20 specimens collected across the regions under study. Histological sections of 10 nm thickness were made of four Clione antarctica and stained using Cason’s modifica- tion of Mallory’s triple stain (Humason, 1962). RESULTS COMPARISON OF EXTERNAL ADULT MORPHOLOGY Adult specimens of Clione limacina from the North Pacific and North Atlantic are anatomically similar (Figs. 1a, b), despite differences in size. Clione attains a maximal size of about 30 mm in length in the North Pacific Ocean, whereas adult specimens in arctic and subarctic waters of the North Atlantic Ocean commonly reach a length of 70 to 85 mm (Table 1), the largest size of any gymnosome. The maximal size of C. limacina diminishes in more temperate waters of the North Atlantic Ocean; sexually mature specimens are usually less than 25 mm long in waters south of Nova Scotia, Canada, and are less than 12 mm long in the English Channel. Clione antarctica (Fig. 1c) attain a maximal length of about 42 mm, and it is strikingly different in external ap- pearance from the northern species of Clione. The head of C. antarctica is elongate compared with that of C. limacina, and the demarkation between the head and the remainder of the body is not as clear. The head comprises about one- quarter to one-fifth of the total body length, whereas the head of C. limacina is smaller relative to total body length, con- stituting less than one-seventh of the total body length of ex- tended, living animals. The anterior tentacles of C. antarctica are noticeably smaller than those of the northern hemisphere species, but both species have similarly sized posterior tentacles located in depressions on the dorsal surface of the head. Clione antarctica adults retain evidence of all three lar- val ciliary rings that are characteristic of gymnosome polytroch larvae, although the cilia can disappear. The first ring per- sists as distinct and separate protuberances encircling the mid-section of the head. The second and third rings persist as transparent bands that encircle the mid-trunk and the posterior tip of the body, respectively. In C. limacina, the anterior and middle larval rings disappear rapidly and com- pletely during metamorphosis of the polytrochous larva to the juvenile form; the third or posterior band can persist through the juvenile stage, but it is not evident in sexually mature adults, except in English Channel populations. The posterior tip of the body of adult C. antarctica is marked by the con- striction of the posterior larval ring, followed by an expanded triangular area. This is identical to the form of the posterior tip in young juvenile C. limacina but, in adults, the constric- tion is lost and the posterior end tapers gradually to a point. Adults of Clione antarctica also differ from the northern species in that the viscera extends three-quarters of the way to the posterior tip of the body. In C. limacina, the viscera can occupy the entire body of larvae and juveniles but, in adults, the visceral mass does not extend beyond the anterior one- half of the body; the posterior half of the adult body is a fluid- filled cavity. Body coloration is somewhat variable among in- dividuals but, in both species, the prehensile buccal cones are reddish-orange. The posterior tip is brightly pigmented in some specimens of Clione limacina, but it is usually transparent or only faintly colored in C. antarctica. The digestive gland of both species is usually yellowish-orange or orange, but is dark brown immediately after feeding. Col- oration in general is probably derived at least partly from diet, as Starved animals lose the striking red color from the cones. The wings and integument are transparent, but slightly less so in specimens of C. antarctica which can exhibit some degree of opacity. Eliot (1907) stated that the integument of preserved specimens of C. antarctica contained more “‘yellow spots’ than that of C. limacina. This statement apparently refers to oil droplets embedded in the integument, and Eliot’s statement is true when comparing adults. However, polytrochous larvae and young juveniles of C. limacina are very similar to C. antarctica adults in having numerous oil droplets scattered throughout the integument. Finally, the footlobes of Clione limacina tend to be short and wide. In comparison, all three lobes of C. antarctica are narrower and more elongate. COMPARISON OF THE BUCCAL MASS The feeding structures of Clione consist of three pairs of prehensile buccal cones (cephaloconi), paired hook sacs GILMER AND LALLI: BIPOLAR VARIATION IN CLIONE 69 Fig. 1. Comparisons of adult Clione from different localities; all in ventral view: a, Clione limacina from the Subarctic Atlantic Ocean; b, C. limacina from the North Pacific Ocean; ¢, C. antarctica from McMurdo Sound (scale bars = 5 mm). 70 AMER. MALAC. BULL. 8(1) (1990) Table 1. A comparison of predator-prey maximal sizes in different locations. Clione Limacina Location Maximal adult Maximal adult Reference* length (mm) diameter (mm) Subarctic Atlantic Ocean 70-85 (live) 11-12 (helicina) 2,4, 8 Nova Scotia 20-25 (live) <3.0 (helicina and 2 retroversa) English Channel 12 ca. 1.0 (retroversa) 5, 6 North Pacific Ocean <30 (live) 2.5 (helicina) 1,7, 8 Central Arctic Ocean 70-80 (live) ca. 11 (helicina) 3 Antarctic Ocean 42 (live) 5-6 (helicina) 8 “References 1 - Agersborg, 1923 2 - Conover and Lalli, 1972 3 - Ospovat, pers. obs. 5 - Lebour, 1931 6 - Lebour, 1932 7 - McGowan, 1963 4 - Lalli and Wells, 1978 containing curved chitinous hooks, and the radula. There is no obvious morphological difference between the buccal cones of Clione limacina and C. antarctica. Meisenheimer (1906) and Eliot (1907) claimed that, in C. antarctica, the dor- sal pair of cones was Spatially separated from the other two pairs and that the medial cones were the largest; in contrast, the cones were said to be equidistant in C. limacina, with the dorsal cones being largest. We were not able to confirm this to our satisfaction. Instead, it would appear that size and relative placement of the buccal cones is dependent on their state of eversion or contraction in living animals, or on their degree of contraction in preserved specimens. On the other hand, differences can be found in the structure of the chitinous hooks of the two species (Fig. 2). In all Clione, the hook sacs contain a variable number of hooks that can be everted from the sacs. These range in size from very small to fully developed, long, curved hooks. In the largest-sized adult Clione limacina from Subarctic waters of the North Atlantic Ocean, each hook sac contains approx- imately 60 hooks; these are arranged roughly in three rows, with the largest hooks being about 1.7 mm in length (Fig. 2a). In Clione from the North Pacific Ocean, the numbers and max- imal size of the hooks are reduced, which is related to the smaller size of the species in this area; in adults, each sac contains about 30 hooks, with the largest about 0.7 mm in length. The hooks of C. antarctica are more numerous and larger relative to body size than those of C. limacina. |n adult C. antarctica, each sac contains approximately 60 hooks, the largest is about 1.2 mm in length (Fig. 2b). The most striking differences between the two species occur in the morphology of the radula. The radula of Clione limacina (Figs. 3, 4) consists of a relatively wide ribbon that extends dorsally and ventrally over the tip of the odontophore. Each row of chitinous radular teeth consists of a single, cen- tral, sickle-shaped tooth and, in our specimens, up to 12 curved, pointed, lateral teeth on each side. The number of rows of teeth and the number of teeth in each row varies with the size of the individual examined, and these numbers are maximal in the large Subarctic Atlantic Ocean specimens (Fig. 3a). However, apart from relative numbers of teeth and relative 8 - Personal observations size differences in the teeth, there is little difference in the structure of the radula and radular teeth between North Atlantic Ocean and North Pacific Ocean Clione populations. In all specimens of C. limacina that we examined, the central tooth is well developed, even in young specimens. Denticula- tion of the central tooth varies from sharp points on newly formed teeth (Figs. 3c, 4b) to low, broad protuberances on older teeth (Fig. 3b), indicating wear from use. There is also no discernible difference in the general shape of the lateral teeth between populations of C. limacina (Figs. 3a, 4a). In contrast, the radula of Clione antarctica (Fig. 5) is extremely reduced, and this reduction is not proportionate to size differences between this species and the northern Clione. In Antarctic Clione, the minute radular ribbon is positioned differently on the odontophore. The ribbon splits just anteriorly to the radular sac and then extends forward as a broadening Y-shaped structure that spreads laterally over the tip of the odontophore (Fig. 5a). The lateral radular teeth (Fig. 5b) are smaller but similar in shape to those of C. limacina; the number of lateral teeth in each row is somewhat variable but is less than eight in all specimens examined. The most unusual feature is the absence of a central tooth. Although Eliot (1907) illustrated median teeth which he claimed were present in the first two rows of the radula, we have not been able to confirm this and concur with Pruvot-Fol (1932) who also did not find central teeth. Because of the lateral posi- tioning of the ribbon on the odontophore tip, the only place where such teeth could be present would be immediately anterior to the radular sac (Eliot’s ‘‘hindmost’’ rows). Examina- tion with scanning electron microscopy (Figs. 5c, d), revealed that the bases of adjoining lateral teeth were attached and that these eventually pull away and separate as the ribbon splits laterally. We believe that the small point of attachment of the lateral teeth in this area was seen at lower magnifica- tions by Eliot and confused with a denticle of a central tooth. BEHAVIORAL COMPARISONS All Clione feed on the thecosomatous pteropod Limacina by extracting the prey from its shell. In northern cold waters, Clione limacina feeds on L. helicina (Phipps), as does GILMER AND LALLI: BIPOLAR VARIATION IN CLIONE 1 Fig. 2. Scanning electron micrographs of the hooks from a single hook sac: a, Clione limacina (North Atlantic); b, C. antarctica (scale bars = 50 um in a; 200 um in b). C. antarctica in the Antarctic Ocean. In temperate waters of both the northern and southern hemispheres, Clione spp. co- exist with and feed on L. retroversa (Fleming). There are no significant differences in prey capture and feeding behavior between the two species of Clione. Maximal sizes attained by the predator and prey in different areas are linked (Table 1); both the prey and predator are largest in Subarctic waters and smallest in the English Channel. Antarctic Ocean species are intermediate in size. These size relationships in nature agree with the experimental laboratory results of Conover and Lalli (1972, 1974) suggesting that prey size over prey concen- tration and temperature is the major determinant of Clione size. Fig. 3. Scanning electron micrographs of the radula of Clione limacina from the Subarctic Atlantic: a, median and lateral radular teeth; b, older median teeth showing wear; c, newly formed teeth (scale bars = 25 umina, 5 uminb andc). 72 AMER. MALAC. BULL. 8(1) (1990) Swimming behavior of Clione has been observed in situ and in the laboratory. The majority of Clione limacina observed in the field are oriented head up and are simply maintaining their depth position with slow wing movements. During such periods of inactivity, subarctic specimens in water tempera- tures of less than 2°C move their wings dorsally and ventral- ly at arate of about 2 beats/sec. The posterior tip of an animal is often curved ventrally at this time. When disturbed, the animals swim rapidly and the orientation of the body can be in any direction away from the disturbance. Individuals with body lengths exceeding 70 mm can swim at speeds of at least 10 cm/sec in waters of —0.5°C. The body of an animal often flexes dorsally and ventrally during rapid swimming and turn- ing. Rapid wing motion can also be apparent during prey cap- ture, but this does not necessarily result in movement. Slow, rhythmic swimming is continuous during mating, but rapid swimming in a circular pattern appears to be necessary to break the pairing. Spawning individuals swim slowly, mov- ing their wings at about 2 beats/sec. Clione limacina also appears to be able to achieve neutral buoyancy. On numerous occasions, divers have observed C. limacina hanging motionless in a head down posi- tion, with the wings extended. No measurable sinking occurs at such times. In McMurdo Sound, Antarctica, Clione antarctica is generally found congregated near the undersurface of the ice shelf. It is generally sparsely distributed in water deeper than 20 m. Like C. limacina, this species also orients head up dur- ing slow rhythmic swimming, but it does not bend the posterior tip of the body ventrally. Wing motion is generally slow, at less than 2 beats/sec. When disturbed, however, large specimens (>25 mm long) accelerate the wing beat to 4 to 5 beats/sec and can swim at speeds of about 5 cm/sec. This species reacts differently to disturbance. Most often, disturbed C. antarctica will stop swimming motions, retract the head slightly, and remain motionless in the water. Animals seldom attempted to swim away from a disturbance, which is the most common reaction of C. limacina. Reproductive anatomy, copulation, spawning, and development are well known for Clione limacina, and these activities have been reviewed by Lalli and Gilmer (1989). Reproductive anatomy and behavior are not well known in the Antarctic species. Histological studies were made of four specimens of C. antarctica collected between November and January, when both spawning adults and veliger larvae were present. Specimens as small as 15 mm in length (preserved measurement) had oocytes in the gonad, sperm packed into a swollen genital duct, well developed mucous and albumen glands, and a large penis and prostrate gland. Two specimens of 30 to 40 mm length (live measurements) had spawned in the laboratory. Neither of these individuals showed any regres- sion of copulatory structures and both had masses of sperm in the genital duct, indicating that they were capable of func- tioning as both male and female. However, copulation has never been observed in this species, either in the field or in laboratory-maintained individuals. The size of the egg mass and the number of eggs spawned by three specimens of Clione antarctica varied with body size. The smallest individual (24 mm long, live measure- ment) produced an egg mass of 12 mm diameter containing 250 eggs. Two animals (33 mm and 38 mm long) spawned egg masses of 15 mm, with about 600 eggs produced by the smaller individual and about twice that many deposited by the larger animal. The largest egg mass was kept in the laboratory at 2°C (approximately 2.5°C above ambient temperature), and developmental times were recorded. On 12 November 1987, 24 hr after spawning, 20% of the eggs were at the second cleavage. At 400 hr, most embryos were spinning inside their egg capsules. About 10% had hatched as free-swimming veligers by approximately 20 days after spawning. Veligers of Clione antarctica (Fig. 6) are characterized by a bilobed velum and a small shell with two distinctive areas. The cap-like embryonic portion of the shell measures about 110 um long and 160 um at its widest diameter. The post- embryonic extension of the shell, which is formed after Fig. 4. Scanning electron micrographs of the radula of Clione limacina from the North Pacific: a, median and lateral radular teeth; b, median teeth (scale bars = 10 pm ina, 5 pm in b). GILMER AND LALLI: BIPOLAR VARIATION IN CLIONE 73 Fig. 5. Scanning electron micrographs of the radula of Clione antarctica from McMurdo Sound: a, configuration of the complete radula; b, lateral radular teeth; c, adjoined lateral teeth in the median furrow (arrow indicates fused teeth); d, enlargement of c (scale bars = 40 um in a, 8 pm in b, 12 um inc, 4 um in d). hatching, flares out like a collar from the embryonic shell and has distinctive, encircling growth rings. The shells differ from those of C. limacina veligers (see Lalli and Conover, 1976) in being more rounded posteriorly, with a shorter and broader embryonic portion. We have not been able to determine the size at which the larval shell is cast, nor have we observed loss of the velum and metamorphosis to the polytroch stage. The polytrochous larvae of Clione antarctica are similar to those of other gymnosome species in having three ciliary bands which encircle the body, but they already show the distinctive traits that separate them from the northern Clione. A comparison of the polytroch larvae (each 9 mm long) of C. antarctica and of C. limacina from the North Pacific (Fig. 7) shows the striking differences in head to body proportions and in footlobe shape that are also present in the adults. It is also clear that the protuberances underlying the cilia of the anterior larval band are much more prominent in C. antarc- tica larvae. At this stage, the Antarctic polytrochs already closely resemble the adults; the only major anatomical change that will accompany growth will be the gradual regression Fig. 6. Veliger larva of Clione antarctica (scale bar = 50 um). 74 AMER. MALAC of the cilia in the larval rings. The polytrochous larvae of C. limacina, however, will undergo a more dramatic meta- morphosis with rapid and complete loss of the anterior and middle ciliary bands, more gradual loss of the posterior ciliary ring, and a progressive lengthening of the trunk until the viscera are confined to the anterior half of the body. DISCUSSION Meisenheimer (1906) and Eliot (1907) were the first to present detailed accounts of Clione collected from the Antarctic. Both authors noted anatomical differences, but Meisenheimer preferred to regard these gymnosomes as a variety of C. limacina, whereas Eliot considered that the dif- ferences were such as to validate Smith’s establishment of C. antarctica. Eliot presented a list of nine major anatomical distinctions, and several minor ones, between C. antarctica and C. limacina. We concur with many of his points concern- ing C. antarctica: the head is larger relative to body length; the body of the Antarctic species is smaller; at least two lar- val bands are retained into the adult stage; the viscera ex- tend farther posteriorly; and the footlobes have a different shape and a narrower attachment. We are hesitant to accept that the buccal cones have a different arrangement, as this seems to be a preservation artifact. Nor do we agreee that there are more oil droplets in the integument than there are present in C. limacina; this appears dependent on the size Fig. 7. Polytrochous larvae of Clione antarctica from the Antarctic Ocean (left) and of C. limacina from the North Pacific (right) (ACB, anterior ciliary band; LFL, lateral footlobe; MCB, median ciliary band; MEL, median footlobe; PCB, posterior ciliary band). _ BULL. 8(1) (1990) (and age) of the animals which are compared. However, there is a tendency of Antarctic Ocean animals to develop a slight opacity of the integument compared to northern ones. On the other hand, we concur with Pruvot-Fol (1932) that Antarctic Ocean specimens do have a well-developed copulatory organ, consisting of a penis, prostrate gland and accessory organ, that is identical to that of C. limacina. We also share her view that median radular teeth are absent in C. antarctica, in con- trast to the well-developed and conspicuous median teeth found even in very small specimens of C. limacina. In addi- tion, our results show that the hooks of Antarctic Ocean animals are larger and more numerous relative to body size; this is probably correlated with the diminution of the radula and probably indicates a greater involvement of the hooks in extraction of prey from its shell. Behavioral differences between the two species do not appear to be so prounounced, but this could be due to fewer hours of observatoin of living Clione antarctica. Both C. limacina and C. antarctica feed on Limacina helicina or L. retroversa; we have not observed any significant difference in prey capture or in ingestion time from that described by Lalli (1970) and Conover and Lalli (1972). The Antarctic Ocean species, however, swims at slower rates than does its northern counterpart, and it usually responds to disturbance by cessa- tion of swimming rather than active movement away from a stimulus. Both clionids deposit free-floating, gelatinous egg masses. The size of egg masses and the number of eggs per mass are smaller in C. antarctica than in large, subarctic C. limacina, but are larger than in dwarf specimens of C. limacina from the English Channel (Lalli and Gilmer, 1989). In the northern species, hatching of the veligers coincides with periods of maximal phytoplankton abundance and with simultaneous hatching of Limacina veligers (Mileikovsky, 1970; Conover and Lalli, 1972). In the Antarctic Ocean, we have col- lected spawning adults from November through January, and Massy (1932) reported finding larvae of less than 3 mm length in October, November, December and February. There are differences in the size and shape of the veliger shell of C. antarctica, as well as between the polytroch larvae of the northern and southern species. It is interesting that C. antarc- tica, like several other gymnosome species (Lalli and Gilmer, 1989), displays neotenous characters, retaining external lar- val features and a relatively small size after reaching sexual maturity. In contrast, C. limacina usually undergoes a com- plete metamorphosis from the polytroch stage; a posterior larval band is present only in the small-sized individuals liv- ing in the English Channel (Lebour, 1931; Morton, 1958) or in the rare neotenous individual collected off Nova Scotia (Lalli and Conover, 1973). Although we do not agree on all the points of difference between northern and southern Clione as established by earlier workers, we do believe that the evidence presented here further strengthens the taxonomic distinctions separating C. limacina and C. antarctica. Although both clionids occupy identical ecological niches in cold water areas, feeding on two species of the shelled pteropod Limacina and exhibiting similar behavioral patterns, they are sufficiently different morphologically to justify their separation. In addition, the GILMER AND LALLI: BIPOLAR VARIATION IN CLIONE 15 species are spatially isolated. There is no evidence to sug- gest that there is any physical connection between the nor- thern and southern populations, so there is no possibility of interbreeding. However, it is evident from the similarity of the species in the two hemispheres that they have evolved from a common ancestor. ACKNOWLEDGMENTS We wish to acknowledge and thank Dr. T. R. Parsons for laboratory supplies and facilities and Dr. G. R. Harbison for shiptime on two Subarctic cruises in the North Atlantic. We also gratefully acknowledge the technical assistance of M. Weis, P. Linley, T. Smoyer, and P. Blades-Eckelbarger. Thanks are also due to G. Harbison, J. Lindsey, G. Dietzmann, D. Backus, and N. Wu for diving and photographic assistance, and to M. Ospovat for his diving observa- tions in the Central Arctic Ocean. Research support for RWG was provided by N.S.F. grants OCE-8209341, 8516083, 8746136 and DPP8613388 to G. R. Harbison; we also acknowledge research sup- port from NSERC grant 0006689 to T. R. Parsons. This is Harbor Branch Oceanographic Institution Contribution No. 749. LITERATURE CITED Agersborg, H. P. K. 1923. Gymnosomatous Pteropoda from Friday Harbor, Washington. Annals of Natural Science, Zoology, Series 10 6:391-402. Conover, R. J. and C. M. Lalli. 1972. Feeding and growth in Clione limacina (Phipps), a pteropod mollusc. Journal of Experimen- tal Marine Biology and Ecology 9:279-302. Conover, R. J. and C. M. Lalli. 1974. Feeding and growth in Clione limacina (Phipps), a pteropod mollusc. Il. Assimilation, metabolism, and growth efficiency. Journal of Experimental Marine Biology and Ecology 16:131-154. Ekman, S. 1953. Zoogeography of the Sea. Sidgwick & Jackson, Lon- don. 417 pp. Eliot, C. 1907. Mollusca. VI. Pteropoda. National Antarctic Expedition 1901-1904 (Natural History) 3 (Zoology and Botany): 1-15. British Museum, London. Humason, G. L. 1962. Animal Tissue Techniques. W. H. Freeman, San Francisco. 641 pp. Lalli, C. M. 1970. Structure and function of the buccal apparatus of Clione limacina (Phipps) with a review of feeding in gym- nosomatous pteropods. Journal of Experimental Marine Biology and Ecology 4:101-118. Lalli, C. M. and R. J. Conover. 1973. Reproduction and development of Paedoclione doliiformis, and a comparison with Clione limacina (Opisthobranchia: Gymnosomata). Marine Biology 19:13-22. Lalli, C. M. and R. J. Conover. 1976. Microstructure of the veliger shells of gymnosomatous pteropods (Gastropoda: Opisthobranchia). Veliger 18:237-240. Lalli,, C. M. and R. W. Gilmer. 1989. Pelagic Snails: The Biology of Holoplanktonic Gastropod Mollusks. Stanford University Press, Stanford, California. 259 pp. Lalli, C. M. and F. E. Wells. 1978. Reproduction in the genus Limacina (Opisthobranchia: Thecosomata). Journal of Zoology, Pro- ceedings of the Zoological Society of London 186:95-108. Lebour, M. V. 1931. Clione limacina in Plymouth waters. Journal of the Marine Biological Association of the United Kingdom 17:785-795. Lebour, M. V. 1932. Limacina retroversa in Plymouth waters. Journal of the Marine Biological Association of the United Kingdom 18:123-129. Martens, F. 1675. Spitzbergische oder grolandische Reise Beschreibung gethan im Jahr 1671. Schultzen, Hamburg. Massy, A. L. 1920. Mollusca. Ill. Eupteropoda (Pteropoda Thecosomata) and Pterota (Pteropoda Gymnosomata). British Antarctic (‘‘Terra Nova’) Expedition 1910, Natural History Report, Zoology 2:203-228. Massy, A. L. 1932. Mollusca: Gastropoda Thecosomata and Gym- nosomata. ‘‘Discovery’’ Reports 3:267-296. McGowan, J. A. 1963. Geographical variation in Limacina helicina in the North Pacific. /n: Speciation in the Sea. J. P. Harding and N. Tebble, eds. pp. 109-128. Systematics Association Publication No. 5. Meisenheimer, J. 1906. Die Pteropoden der deutschen Sudpolar- Expedition 1901-1903. Deutsche Sudpolar-Expedition |X. Zoologie. 1:94-153. Mileikovsky, S. A. 1970. Breeding and larval distribution of the pteropod Clione limacina in the North Atlantic, Subarctic and North Pacific Oceans. Marine Biology 6:317-334. Morton, J. E. 1958. Observations on the gymnosomatous pteropod Clione limacina (Phipps). Journal of the Marine Biological Association of the United Kingdom 37:287-297. Phipps, C. J. 1774. A Voyage Towards the North Pole. Bowyer & Nichols, London. 275 pp. Pruvot-Fol, A. 1932. Notes sur quelques gymnosomes de pro- venances diverses et diagnose d’un genre nouveau. Archives de Zoologie expérimentale et générale 74:507-529. Rang, P.C. A. L. and L. F. A. Souleyet. 1852. Histoire naturelle des Mollusques Ptéropodes. Monographie comprenant la descrip- tion de toutes les especes de ce groupe de Mollusques. J.-B. Bailliére, Paris. 86 pp. Smith, E. A. 1902. Mollusca. /n: Report on the Collections of Natural History of the ‘‘Southern Cross’: W. Clowes & Sons, Ltd., Lon- don. 344 pp. van der Spoel, S. 1976. Pseudothecosomata, Gymnosomata and Heteropoda (Gastropoda). Bohn, Scheltema, & Holkema, Utrecht. 484 pp. Date of manuscript acceptance: 3 October 1989. ANNUAL CYCLE OF LIMACINA RETROVERSA IN PATAGONIAN WATERS JOSE R. DADON DEPARTAMENTO DE CIENCIAS BIOLOGICAS FACULTAD DE CIENCIAS EXACTAS Y NATURALES UNIVERSIDAD DE BUENOS AIRES 1428 BUENOS AIRES, ARGENTINA ABSTRACT An annual cycle (April 1978 - April 1979) of Limacina retroversa (Fleming) in Patagonian shelf and surrounding waters (40 - 55°S) is analyzed. Both abundance and mesoscale distribution of this species in the area were strongly associated with the seasonal cycle. A stationary distribution pattern was evident in autumn and winter. The shelf population concentrated chiefly in a dense aggregation (maximum: 67,800 individuals per 1,000 m3) located to the northwest of the Malvinas Current. Individuals expatriated northward connected this aggregation with others carried by the core of the Malvinas Cur- rent. In October, the aggregation on the shelf began to disintegrate and proceeded slowly northward and off the shelf. The summer distribution pattern consisted of scattered individuals on the shelf and some aggregations moving along the slope. Hydrological dynamics together with diel vertical migrations appeared to be of crucial impor- tance in maintaining the observed distributions over shelf waters. As with many other groups of marine zooplankters, pelagic mollusks have been little studied in the Southwestern Atlantic. Boltovskoy (1973, 1975) and Magaldi (1974, 1977) com- pleted an inventory of the existing species and related the faunistic assemblages with megascale hydrological char- acteristics. However, their conclusions are based on a relative- ly small number of samples, which were not quantitatively, discontinuous in time, and obtained mainly from a reduced area between 35° and 40°S. In Patagonian waters south of 40°S, only subantarctic species can be found (Dadon, 1984, 1986). Though poor in diversity, pelagic mollusks in these cold waters are abundant, reaching densities of 55,000 per 1,000 m3 on the slope and outer shelf during winter. Limacina retroversa (Fleming) was found to be the predominant species; the other planktonic mollusks were L. helicina (Phipps) and Clio antarctica Dall. Seasonality in hydrographic conditions and biological production is very well known in Patagonian waters. Primary production peaks in September, October or November, de- pending on the latitude (Mandelli and Orlando, 1966; Carreto et al., 1981a). Zooplankton biomass increases in spring, follow- ing the period of high phytoplankton production (Ciechom- ski and Sanchez, 1983). This study is mainly concerned with seasonal patterns of Limacina retroversa in Patagonian waters. Monthly mesoscale distribution and changes in total abun- dance of the species during an annual cycle are analyzed and hypotheses are developed concerning the role of biological and environmental factors. MATERIALS AND METHODS Eighteen exploratory cruises were conducted in shelf and slope waters off Patagonia between April 1978 and April 1979 by the R/V ‘Walther Herwig”’ and ‘* Shinkai Maru”’ (see Ciechomski et a/., 1979; Cousseau et a/., 1979). These col- lections included 539 samples taken to the south of 40°S, all of which were analyzed for Limacina retroversa. Samples were collected during daytime and twilight with Bongo nets (mouth diameter = 60 cm; length = 330 cm) fitted with 330 or 505 1m mesh nets. On the continental shelf, oblique tows extended from the surface to approximately 20 m above the bottom; on the slope, from the surface to 100 m depth. The maximal tow depth was estimated by time- depth recorders. The nets were provided with digital flow meters in both mouth openings and were towed at 3.5 knots. In all cases, the volume of water filtered through the nets was 200 - 1000 m3. For counting, the samples were divided in aliquots with a Folsom plankton sample splitter until the final aliquot con- tained 180 to 300 individuals. Sparse samples were analyzed in their entirety. For large volumes, little or no significant differences were obtained either in the plankton volume or in the size com- position of the zooplankters, between the plankton volume filtered by the two meshes (Ciechomski and Sanchez, 1983). American Malacological Bulletin, Vol. 8(1) (1990):77-84 a 78 AMER. MALAC. BULL. 8(1) (1990) HYDROLOGY The Patagonian continental shelf is broad, with a gradual slope. Neritic waters are mostly of subantarctic origin with some admixture from continental run-off. Surface temperatures varied from 2.3 to 20.2°C; in general, temperature increased northward and westward. Surface salinity varied from 34.17 o/oo (slope waters) to 32.39 o/oo (in- ner shelf waters). For detailed information about the physical data of the studied cruises see Ciechomski et a/. (1979) and Cousseau et al. (1979). The mesoscale hydrology of the area is dominated by the Malvinas or Falkland Current, which is a branch of the West Wind Drift (Fig. 1). Satellite observations (Legeckis and Gordon, 1982) show the Malvinas Current to be a well de- fined, 100 km-wide belt of cold waters. The core of this cur- rent runs close to the western boundary of the slope. Surface waters overlying the shelf also flow in the same general direc- tion as the Malvinas Current, but at a considerably lower velocity. From the Gulf of San José (approximately 45° and 47°S) to the slope, the Patagonian Coastal Current (Brandhorst and Castello, 1971) flows Northeast through the shelf (Fig. 1). The contact between the Malvinas and Brazil currents constitutes the Subtropical-Subantarctic Front. This Front is located to the north of the Patagonian waters (Fig. 1) and con- stitutes a complex zone. It marks the limit of the distribution 40° Fig. 1. General hydrology of the study area (cold waters, full line; WWD, West Wind Drift; MC, Malvinas Current; PCC, Patagonian Coastal Current; Warm waters, hatched line; BC, Brazil Current). 50° 38°S 40° Oa «, ERE L/ eeee e 54° ahi Gal os 56° 70°W 68° 66° 64° 62° 60° 58° 56° 54° 52° Fig. 2. Distribution of Limacina retroversa in May 1978. (Circles denote stations. Abundances as follows: Black areas indicate >1,000 per 1,000 m3; cross-hatched areas, 1,000 - 100 per 1,000 m3; hachured areas, 100 - 10 per 1,000 m3; white areas (when sampled) <10 per 1,000 m3). of the subantarctic fauna in shelf waters and its position varies seasonally (Tseng, 1974; Olson et al., 1988). RESULTS SEASONAL DISTRIBUTION The mesoscale horizontal distribution of Limacina retroversa in Patagonian waters showed a well defined pat- tern during the 1978 - 1979 annual cycle (Figs. 2-6). Densities higher than 1,000 individuals per 1,000 m3 (maximum of 67,800 per 1,000 m3) were found in slope waters throughout the year and in intermediate and outer shelf waters from April to September, 1978. In slope waters, dense aggregates were recorded that were transported northward by the Malvinas Current throughout the year (Figs. 2, 3, 6). In general, ag- gregates traveling in slope waters seemed to replace each other easily and quickly. Other authors (Bigelow, 1926; Red- field, 1939) pointed out the tendency of L. retroversa to aggregate. The western limit for massive entrance of Limacina DADON: ANNUAL CYCLE OF LIMACINA RETROVERSA 79 Fig. 3. Distribution of Limacina retroversa in June 1978. Symbols as in figure 2. retroversa to the shelf coincided with the 100 m isobath to the south of 41°S (Figs. 3, 5). Between this dense strip and the coast, low densities were only occasionally registered, especially to the north of 47°S (Figs. 3, 4, 5). This low density or absence of the species constituted a tongue-like area which coincided with the path of the Patagonian Coastal Current. The holoplanktonic fauna depicted for this current is predominantly inner neritic [i.e. Sagitta friderici Ritter-Zahony, 1911 (Dadon and Mazzoni, 1989)]. Taking into account the geo- graphical range of L. retroversa in the area, this species is oceanic but can tolerate neritic conditions. In shelf waters, the highest densities were distributed in an oblong or pear-shaped configuration during the austral autumn and winter (Figs. 2, 3). The widest portion was located to the west of the Malvinas Islands and was continued as a band in a NNE direction onto the slope. This distribution was maintained without important changes practically from May until the beginning of spring. The only variations during this period were deformations and slight displacements of the ag- gregation towards the west (cf. Fig. 2 vs. Fig. 3). Same varia- tions were observed for Limacina helicina, chaetognaths and cladocerans patterns (Dadon, 1986). Ramirez (1981) and Car- reto et al. (1981b) pointed out that in winter, oceanic 70° W 68° 66° 64° 62° 60 58° 56 54 52 Fig. 4. Distribution of Limacina retroversa in October 1978. Symbols as in figure 2. euphausiids, amphipods and copepods occurred predomi- nantly on the Patagonian shelf, while typical neritic zooplankters were found in more offshore waters. At the beginning of spring, a series of important changes in the distribution of Limacina retroversa on the con- tinental shelf were observed (Fig. 4). The oblong aggrega- tion began to disintegrate. Each fragment was carried to the northeast, reaching the slope at latitudes below 40°S (Dadon, unpub. data). At least two of those isolated fragments were detected in October (Fig. 4), and two or three of them in November (Fig. 5). This process continued up to March 1979, when the last important concentration of organisms on the shelf was recorded between 43 and 45°S (Fig. 6). The sum- mer distribution pattern consisted of scattered individuals on the shelf and dense aggregations moving along the slope without penetrating the shelf, as shown between 39° and 42°S in figure 6. DENSITY IN RELATION TO ENVIRONMENTAL FACTORS Limacina retroversa was found in waters when the sur- face temperatures ranged from 4.1 to 188°C, and surface salinities were between 32.39 and 34.17 o/oo. The temperature 80 AMER. MALAC. BULL. 8(1) (1990) interval where the species was recorded in the Patagonian shelf waters (4 - 19°C) was almost the same as the one record- ed for the Southwestern Atlantic (see Spoel and Boltovskoy, 1981). Although densities higher than 1,000 per 1,000 m3 were found at 4.5 - 17.8°C and 33.07 - 34.15 o/oo, this range was mainly recorded at 5 - 99°C and >33.50 o/oo. These optima differ from those reported for northern hemisphere popula- tions of L. retroversa: 8 - 10°C and 34.5 - 35.0 o/oo in the North Atlantic (Chen and Bé, 1964); 7 - 12°C in the Gulf of Maine (Bigelow, 1926); 7.5 - 10°C in the Northeast Atlantic (Beckmann et al., 1987). To obtain a simple relation between selected en- vironmental factors and the species density in a given area, empirical relationships were sought. This analysis does not assume a direct cause-effect relationship, but, as Haedrich and Judkins (1979) pointed out, it constitutes a first necessary approximation to a very complex problem. Spatial variation in the density of Limacina retroversa density was compared to several environmental variables (depth, temperature, salini- ty and oxygen concentration) recorded during Cruises | and Il of R/V ‘‘Shinkai Maru’ (Cousseau et al., 1979). Associa- tion between density and each variable was determined by correlation analysis (Sokal and Rohlf, 1981). Data were nor- 38°S 40° 54° 56° 70° W 68° 66° 64° 62° 60° 58° 56° 54° 525 Fig. 5. Distribution of Limacina retroversa in November 1978. Sym- bols as in figure 2. mally distributed in all cases except for depth and density, for which the transformations x’ = x2 and x’ = In (x + 1) were applied, respectively. The abundance of Limacina retroversa and the depth are positively correlated (Table 1), indicating that these organisms are denser in deeper waters. The correlation be- tween density and temperature is negative, indicating that L. retroversa has a higher affinity for cold waters. However, the correlation between density and temperature decreases at greater depths. The strongest correlations between temperature and density were for the average temperature and for temperatures of the upper layers (Surface to 20 - 25 m). Confronted with the other environmental factors (Table 1), salinity shows lower correlation coefficients with density (r < 0.39; P < 0.05 or not significant). Dissolved oxygen con- centration and density were positively correlated in all cases and, in contrast to temperature and salinity, the average did not show a higher value than its components when considered individually. The highest correlation values were achieved for oxygen concentration at depths between 0 and 45 - 50 m. In order to establish the maximum proportion of the observed variation in the density of Limacina retroversa which can be explained in terms of the environmental variations, the 38°S 40° 50° 54° 56° 70°w 68° 66° 64° 62° 60° 58° 56° 54° 52° Fig. 6. Distribution of Limacina retroversa in March 1979. Symbols as in figure 2. DADON: ANNUAL CYCLE OF LIMACINA RETROVERSA 81 backward elimination and the forward selection procedures (see, for example, Drapper and Smith, 1966) were carried out considering the 25 environmental variables (Table 1) as predic- tors. The best regression equation was Y = -90.748 + 4308 X; + 2.036 X2 R2 = 0.52 where Y: density of L. retroversa; X,: oxygen concentration at 20 m; and X2: average salinity. Since the most commonly- measured factors are temperature and salinity at the surface, their efficiency in predicting the species density either in- dividually or in combination was analyzed for all months. Coef- ficients of determination ranked from 0.001 to 0.48 in all cases. DISCUSSION FACTORS GENERATING AND MAINTAINING MESOSCALE DISTRIBUTION Detailed analysis of distribution and abundance pat- terns implies the study of the relations between the organisms and the environment. From Pickford (1946) on, several authors have repeatedly contended that the distribution of plankton is primarily ruled, or at least potentially governed, by the distribution of the water masses that the plankters inhabit. In certain cases, clear evidences in favor of this hypothesis were provided (see the review of Haedrich and Judkins, 1979; in pelagic mollusks, Furnestin, 1978). In other cases, this rela- tion could be demonstrated only when intraspecific variations were considered (i.e. McGowan, 1963). In yet other cases, although the same species inhabited more than one water mass, there were remarkable differences in the density of in- dividuals present in each environment, pointing out the necessity for quantitative studies when faunistic areas are to be compared (Fasham and Angel, 1975; Dadon, 1984). In many cases, however, this hypothesis had to be rejected because the species were ubiquitous or were highly cosmopolitan. Limacina retroversa inhabits quite different water masses in the Southwestern Atlantic. It was collected on both sides of the Antarctic Convergence (Chen, 1968) and its geographical range extends northward to the Subtropical - Subantarctic Front. However, quantitative analyses (e.g. Bé and Gilmer, 1977: Fig. 7) have revealed that the area of highest densities is much less extensive and confined to the sub- antarctic region. Despite this clear association between L. retroversa and subantarctic waters, the present analysis of mesoscale distribution showed that the relationship is not sim- ple and that it depends on several factors. AS above men- tioned, high densities of L. retroversa in patagonian waters were correlated with highly oxygenated, cold pelagic (= deep) waters. This set of correlations clearly defines the core of the Malvinas Current as the most favorable habitat in the area. This current is probably the only way for L. retroversa to enter the region, although the aggregates recorded in the core (i.e. in slope waters) seem to be rapidly expatriated northward. High densities associated with shelf waters, even when in- fluenced by the Malvinas Current, established a stable distribution during autumn to winter. This stability implies the Table 1. Correlation between density of Limacina retroversa and en- vironmental variables (TD, tow depth; WA, weighted average for the sampled water column; r, correlation coefficient; **, PP<0.01; *, P<0.05; NS, P> 0.05). Environmental Variables r Depth (m) 0.5376 ie Temp (°C) 0 m -0.6252 ne * 10m -0.6276 gi ” 20-25 m -0.6188 ade 45-50 m -0.5316 se 65-70 m -0.3351 * 90-100 m -0.3122 ” TDm -0.6166 id ” WA -0.6546 sie Salinity (9/99) 0 m 0.3166 i v 10m 0.3206 . 4} 20-25 m -0.2140 NS 2 45-50 m 0.3408 i ” 65-70 m 0.2701 NS ue 90-100 m 0.3101 NS a TD m 0.2448 NS % WA 0.3898 ie Oz Conc (mg/l) Om 0.6416 es af 10 m 0.6320 he m2 20-25 m 0.6557 sis ” 45-50 m 0.6372 ee u 65-70 m 0.5359 ik y 90-100 m 0.5605 is ” TO m 0.5276 lg WA 0.5814 si existence of steady environmental conditions, responsible for generating and maintaining (or at least, allowing) such a pattern. Since plankton is transported passively, it is necessary to look for dynamic aspects of the oceanic environment. In an open area like the Patagonian shelf, a stable distribution necessarily implies a closed flow preventing massive, short- term expatriation of the organisms. Limocina retroversa in- habits the upper 150 m of the water column (Bé and Gilmer, 1977). Superficial hydrology of the area is predominantly unidirectional, i.e. through the area from south to north as an open flow. Conversely, deep layers are expected to flow at different velocities and directions. While on the slope the Malvinas Current moves fast and unidirectional at all depths, on the shelf there are dif- ferences in the movements of the waters at different levels, as indicated by the theoretical winter current field calculated by Lusquinos and Schrott (1982). Between the surface and 20 m, current vectors on the continental shelf have a direc- tion exclusively NNE, but at deeper levels (> 30 m), velocity decreases sharply and even reverses turning to the Southwest; this reversal is more important over the outer shelf south of 46°S (Fig. 7). Comparing the winter distribution pat- tern of L. retroversa with this current field, the accumulation zone can be seen to lay within the area where the direction of the flow reverses. Diel vertical migrations of L. retroversa have been observed by several investigators (Bigelow, 1926; Chen and Bé, 1964). Thus, individuals migrate through layers g2 AMER. MALAC. BULL. 8(1) (1990) 40° 50° Fig. 7. Theoretical winter field of currents (according to Lusquinos and Schrott, 1982) (thick arrows, 0 m; thin arrows, 50 m). which move at different velocities and even in opposite direc- tions. The bulk of the individuals in the area prevent massive expatriation, retrogressing during the day the way they pro- gressed during the night. On the other hand, irreversively ex- patriated individuals would gradually approach the continental slope, as a band connecting the aggregation area and the slope (Figs. 2, 3). This mechanism would explain the shape and the long-term stability of the aggregate on the continen- tal shelf. SEASONAL CHANGES Typically, a marked seasonality is shown in Patagonian waters, both for environmental and biological factors. During autumn and winter, the cooling of waters is uniform and smooth; on the contrary, warming of the water column dur- ing the warm period is sharp (Krepper and Bianchi, 1982). Temperature gradients between the coast and the slope became more distinct during spring and summer than the ones observed during fall and winter (Branshorst and Castello, 1971; Legeckis and Gordon, 1982). The primary production peaks during September in the northern area and this pulse displaces to the south reaching the southern extreme in November (Carreto et a/., 1981a). The highest chlorophyll con- centrations (100 - 200 mg/m?) were detected between 44° and 47°S (Carreto et al., 1981a). Zooplankton production follows a similar pattern, though slightly delayed (Ciechomski and Sanchez, 1983). During the 1978 - 1979 cycle, the highest zooplankton densities were registered by Ciechomski and Sanchez in November in the northern portion (42 - 44°S) and in summer in the south (51 - 53°S). Evidences of massive spawning in shelf waters during November 1978 were also found for oceanic zooplankters which, like L. retroversa, are transported to the outer shelf by the Malvinas Current (e.g. euphausiids; see Ramirez and Dato, 1983). According to this, a springtime pulse of L. retroversa should be expected coin- ciding with this increase in general zooplankton production. After November, the abundance of L. retroversa diminished as well as the area occupied by the species on the continental shelf. The last remnants of the dense core which had occupied the major part of the continental shelf were detected towards the end of summer, in March 1979, be- tween 43 and 45°S (Fig. 6). Over the rest of the shelf, and especially in the southern portion, L. retroversa was present only occasionally and always in very low densities (< 10 per 1,000 m3). Nonetheless, even when abundance over the shelf was lower in April 1979, densities up to 100 per 1,000 m3 were detected in neritic waters south of 47°S, indicating the begin- ning of a new cycle. According with these observations, not only the reproduction but also the massive inmigration seem to show a seasonal pattern. Seasonal patterns in abundance and/or horizontal dis- tribution have been previously described for Limacina retroversa in other regions (Redfield, 1939; Vane and Cole- brook, 1962; Paranjape and Conover, 1973). The annual cycle of this species on Patagonian shelf waters shows some similarities with that described by Redfield (1939) for the Gulf of Maine. In both areas, L. retroversa is alternately repatriated and expatriated in large numbers and cannot be maintained from one year to the next by means of in situ reproduction alone. The dynamics for this species and, probably, for most oceanic species of zooplankton, is predominantly regulated by migratory events. This is clearly evident in the core of the Malvinas Current (on the continental slope), where aggrega- tions are rapidly transported northward, but also in shelf waters. In the latter, even when seasonal changes in the local hydrology allow a temporary residence, and eventually reproduction, they cannot preclude a final expatriation. In this context, the Patagonian waters can be described as a subsidiary system, a one-way transit via connecting the West Wind Drift with the Transition Zone in which the long-term presence of the oceanic zooplankton depends on the produc- tion surplus coming from other systems. ACKNOWLEDGMENTS The author wishes to express his thanks to the Instituto Nacional de Investigacidn y Desarrollo Pesquero (Mar del Plata, Argentina) and especially to Juana D. de Ciechomski, for lending the samples; to Demetrio Boltovskoy, for his constant support and his critical revision of the manuscript; to Beatriz Gonzalez, for the revi- sion of the statistical methodology; and to Roger R. Seapy, for his valuable suggestions and styling of the English version. DADON: ANNUAL CYCLE OF LIMACINA RETROVERSA 83 LITERATURE CITED Bé, A. W. H. and R. W. Gilmer. 1977. A zoogeographic and taxonomic review of Euthecosomatous Pteropoda. /n: Oceanic Micro- paleontology, Vol. 1, A. T. S. Ramsay, ed. pp. 733-808. Academic Press, London. Beckmann, W., A. Auras and Ch. Hemleben. 1987. Cyclonic cold- core eddy in the eastern North Atlantic. Ill. Zooplankton. Marine Ecology Progress Series 39:165-173. Bigelow, H. B. 1926. Plankton of the offshore waters of the Gulf of Maine. Bulletin of the United States Bureau of Fisheries 40(2):1-509. Boltovskoy, D. 1973. Los Euthecosomados del Mar Argentino y adyacencias (Mollusca, Pteropoda). Physis (Buenos Aires) A, 32(85):395-411. Boltovskoy, D. 1975. Ecological aspects of zooplankton (Foraminifera, Pteropoda and Chaetognatha) of the Southwestern Atlantic Ocean. Veliger 18(2):203-216. Brandhorst, W. and J. P. Castello. 1971. Evaluacion de los recursos de anchoita (Engraulis anchita) frente a la Argentina y Uruguay. |. Las condiciones oceanograficas, sinopsis del conocimien- to actual sobre la anchojta y el plan para su evaluacion. Con- tribuciones del Instituto de Biologia Marina (Mar del Plata, Argentina) 166:-1-63. Carreto, J. |., R. Negri and H. Benavides. 1981a. Fitoplancton, pigmen- tos y nutrientes. Resultados de las campanas III y VI del B/| ‘Shinkai Maru”, 1978. In: Campanas de investigacion pes- quera realizadas en el Mar Argentino por los B/I ‘‘Shinkai Maru”’ y “Walther Herwig” y el B/P ‘Marburg’, anos 1978 y 1979. Resultados de la Parte Argentina, V. Angelescu, ed. pp. 181-201. Contribuciones del Instituto Nacional de Investigacion y Desar- rollo Pesquero 383, Mar del Plata, Argentina. Carreto, J. |., F C. Ramirez and C. Dato. 1981b. Zooplancton y produccidn secundaria. Parte Il. Distripucidn y variacidn esta- cional de la biomasa zooplanctonica. In: Campanas de investigacion pesquera realizadas en el Mar Argentino por los B/I “Shinkai Maru’”’ y ‘‘Walther Herwig”’ y el B/P ‘‘Marburg’’ anos 1978 y 1979. Resultados de la Parte Argentina, V. Angelescu, ed. pp. 213-232. Contribuciones del Instituto Nacional de Investigacidn y Desarrollo Pesquero 383, Mar del Plata, Argentina. Chen, Ch. 1968. Zoogeography of thecosomatous pteropods in the West Antarctic Ocean. Nautilus 81(3):94-101. Chen, Ch. and A. W. H. Bé. 1964. Seasonal distribution of Euthecosomatous pteropods in the surface waters of five sta- tions in the Western North Atlantic. Bulletin of Marine Science of the Gulf and Caribbean 14(2):185-220. Ciechomski, J. D. de, M. D. Ehrlich, C. A. Lasta and R. P. Sdnchez. 1979. Exploratory cruises carried out by the Fisheries Research Vessel ‘‘Walther Herwig’’, in the Argentine Sea, from May to November 1978. Organization and record of basic data ob- tained. Contribuciones del Instituto Nacional de Investigacion y Desarrollo Pesquero (Mar del Plata, Argentina) 374:1-312. (In Spanish and English). Ciechomski, J. D. de and R. P. Sanchez. 1983. Relationship between ichthyoplankton abundance and associated zooplankton biomass in the shelf waters off Argentina. Biological Oceanography 3(1):77-101. Cousseau, M. B., J. E. Hansen and D. L. Gru. 1979. The exploratory cruises carried out by the Japanese vessel ‘‘Shinkai Maru’’ in the Argentine Sea, from April 1978 to April 1979. Organiza- tion and record of basic data. Contribuciones del Instituto Na- cional de Investigacidn y Desarrollo Pesquero (Mar del Plata, Argentina) 373:1-625. (In Spanish and English). Dadon, J. R. 1984. Distribucidn y abundancia de Pteropoda Thecosomata en el Atlantico Sudoccidental. Physis (Buenos Aires), A, 42(102):25-38. Dadon, J. R. 1986. Distribucidn y abundancia de Pteropoda Thecosomata (Mollusca, Gastropoda) del Mar Epicontinental Argentino y adyacencias. Junio y julio de 1978. Physis (Buenos Aires), A, 44(106):43-50. Dadon, J. R. and H. E. Mazzoni. 1989. Aspectos biogeograficos de Thecosomata (Gastropoda) y Chaetognatha del Mar Argen- tino. In: Resumenes de la XIV Reunion Argentina de Ecologia, Universidad Nacional de Jujuy (Argentina). p. 161. Drapper, N. and H. Smith. 1966. Applied regression analysis. John Wiley & Sons, Inc. New York, London, Sydney. 407 pp. Fasham, M. J. R. and M. V. Angel. 1975. The relationship of the zoogeographic distributions of the planktonic ostracods in the North-east Atlantic to the water masses. Journal of Marine Biology Association of the United Kingdom 55:739-757. Furnestin, M. L. 1978. Planktonic molluscs as hydrological and ecological indicators. In: Pathways in Malacology, S. van der Spoel, ed. pp. 175-194. Bohn, Scheltema & Holkema, Utrecht. Haedrich, R. L. and D. C. Judkins. 1979. Macrozooplankton and its environment. In: Zoogeography and diversity of plankton, S. Van der Spoel and A. C. Pierrot-Bults, eds. pp. 4-28. Bunge Scientific Publishers, Utrecht. Krepper, C. M. and A. Bianchi. 1982. Balance caldrico del Mar Epicon- tinental Argentino. Acta Oceanographica Argentina 3(1):119-133. Legeckis, R. and A. L. Gordon. 1982. Satellite observations of the Brazil and Falkland Currents.- 1975 to 1976 and 1978. Deep- Sea Research 29(3A):375-401. Lusquinos, A. J. and A. G. Schrott. 1982. Corrientes en el Mar Epicontinental Argentino en invierno. Subsecretaria de Cien- cia y Tecnologia, Argentina. 81 pp. Magaldi, N. 1974. Moluscos holoplanctonicos del Atlantico Sudoc- cidental. |. Pteropodos euthecosomados colectados por el “‘Atlantis’” en Marzo 1971. Comunicaciones de la Sociedad Malacoldgica del Uruguay 4(27):1-34. Magaldi, N. 1977. Moluscos holoplanctonicos del Atlantico Sudoc- cidental. Ill. Heteropodos y pterdpodos de aguas superficiales brasilenas y uruguayas. Comunicaciones de la Sociedad Malacologica del Uruguay 4(33):295-320. Mandelli, E. F. and A. M. Orlando. 1966. La produccidn orgdnica primaria y las caracteristicas de la Corriente de Malvinas. Boletin del Servicio de Hidrografia Naval, Buenos Aires 3(3):185-195. McGowan, J. A. 1963. Geographical variation in Limacina helicina in the North Pacific. Speciation in the Sea. Systematics Associa- tion Publication No. 5:109-128. Olson, D. B., G. P. Podesta, R. H. Evans and O. B. Brown. 1988. Tem- poral variations in the separation of Brazil and Malvinas Cur- rents. Deep-Sea Research 35(12A):1971-1990. Paranjape, M. A. and R. J. Conover. 1973. Zooplankton of St. Margaret’s Bay. 1968 to 1971. Technical Reports of the Fishery Board of Canada 401. 83 pp. Pickford, G. E. 1946. Vampyroteuthis infernalis Chun. An archaic dibranchiate cephalopod. |. Natural history and distribution. Dana Reports 29:1-40. Ramirez, F. C. 1981. Zooplancton y produccidn secundaria. Parte I. Distribucion y variacidn estacional de los copépodos. /n: Campanas de investigacion pesquera realizadas en el Mar Argentino por los B/I ‘‘Shinkai Maru”’ y ‘‘Walther Herwig”’ y el B/P ‘Marburg’, anos 1978 y 1979. Resultados de la Parte Argentina, V. Angelescu, ed. pp. 202-212. Contribuciones del Instituto Nacional de Investigacion y Desarrollo Pesquero 383, Mar del Plata, Argentina. 84 AMER. MALAC. BULL. 8(1) (1990) Ramirez, F. C. and C. Dato. 1983. Seasonal changes in population structure and gonadal development of three Euphausiid species. Oceanologica Acta 6(4):427-433. Redfield, A. C. 1939. The history of a population of Limacina retroversa during its drift across the Gulf of Maine. Biological Bulletin 79:459-487. Sokal, R. R. and F. J. Rohlf. 1981. Biometry. W. H. Freeman and Co., New York. 859 pp. Spoel, S. Van der, and D. Boltovskoy. 1981. Pteropoda. /n: Atlas del zooplancton del Atlantico Sudoccidental y métodos de trabajo con el zooplancton marino, D. Boltovskoy, ed. pp. 493-524. In- stituto Nacional de Investigacidn y Desarrollo Pesquero, Mar del Plata, Argentina. Tseng, Y. C. 1974. Study of the surface boundary of the Brazil and Falkland Currents. In: Seminar on space applications of direct interest to developing countries. 2:160-173. Instituto de Pes- quisas Espaciais, Sao José dos Campos, Sao Paulo, Brasil. Vane, F. R. and J. M. Colebrook. 1962. Continuous plankton records: Contribution towards a plankton atlas of the North-Eastern Atlantic and the North Sea. Part VI: The seasonal and annual distribution of the Gastropoda. Bulletin of Marine Ecology 5(50):247-253. Date of manuscript acceptance: 2 May 1990. THE TAXONOMY, DISTRIBUTION AND BIOLOGY OF ATLANTA GAUDICHAUDI SOULEYET, 1852 (GASTROPODA, HETEROPODA) FROM THE GREAT BARRIER REEF, AUSTRALIA L. J. NEWMAN ZOOLOGY DEPARTMENT UNIVERSITY OF QUEENSLAND ST. LUCIA, BRISBANE, QUEENSLAND, AUSTRALIA 4067 ABSTRACT Investigations into the holoplanktonic gastropod fauna of the Great Barrier Reef, from 1985-1988, showed that Atlanta gaudichaudi Souleyet, 1852 was the most common atlantid species found within the surface waters around Lizard and Heron islands. The taxonomy of this species is reviewed par- ticularly with respect to two morphologically related species, A. peroni Lesueur, 1817 and A. plana Richter, 1972. Veligers of A. gaudichaudi are described for the first time. The mean abundance of A. gaudichaudi ranged from 0.2 to 17.3 animals/m3. Abundances were higher at Heron Island (Subtropical waters), especially in summer months. At Lizard Island (tropical waters) greatest numbers were found during the winter. Over 70% of the animals collected during both summer and winter months were veliger stages. Laboratory observations showed that adult A. gaudichaudi swim in a circular course with the aperture pointed towards the water surface; they often rest at the surface. A. gaudichaudi feeds on euthecosomatous and gymnosomatous pteropods. Cannibalism is also common. The family Atlantidae, which comprises approximate- ly 15 species, are the only heteropods that can completely retract into their shells. The shells are planispiral, with a thin keel surrounding the teleoconch. Due to the fragility of their shells, they are not collected easily intact. As a consequence, species descriptions are often incomplete and most identifica- tions dubious so that the taxonomy of atlantids remains con- fused, their distribution uncertain and their biology poorly understood. Tesch (1949) reviewed the taxonomy and distribution of atlantids and synonymized many of the nominal species. Richter (1974) examined the taxonomy of atlantids from the Indian Ocean and described several new species. Van der Spoel (1976) reviewed the taxonomy and distribution of heteropods from the world’s oceans. In the past, descriptions of the minute details needed for species differentiation, revealed under the light microscope, have been inadequate (Pilkington, 1970). Scanning electron microscopy (SEM) has proven to be an essential tool in the study of atlantid shell and radula morphology (Thiriot-Quiévreux, 1973; Richter, 1974). Shell morphology alone is inadequate to differentiate between closely related species. Species descriptions must deal with a range of characters such as morphology of the live animal, shell surface sculpture (well revealed by SEM), veliger shell morphology, shape of the operculum and the eyes, and radula tooth morphology. Richter (1974) used many of these characters in separating Indian Ocean species. Although Tesch (1949) emphatically stated that differences in radula tooth morphology only showed generic differences, this is not necessarily the case since Richter (1961, 1986, 1987) has shown that some atlantid species can be separated by differences in radula morphology. Tokioka (1961) also showed that some species can be distinguished by the shape of operculum. There are few records of atlantids from Australian waters. Smith (1888) reported Atlanta gaudichaudi Souleyet, 1852 from the Torres Strait and A. peroni Lesueur, 1817 ‘near’ Cape York during the Challenger Expedition of 1873-1876. Cot- ton (1932) recorded A. rosea Smith, 1888 (= A. peroni) from dredged material off south Australia. From material collected during the Siboga and Dana Expeditions, Tesch (1906, 1949) reported A. gaudichaudi from the Torres Strait and noted that this atlantid was predominantly found in the Indo-Pacific. Dakin and Colefax (1940) also commented that Atlanta spp. were common off the coast of New South Wales. The only detailed study on heteropods from Australian waters was con- ducted by Russell and Colman (1935) as a seasonal study American Malacological Bulletin, Vol. 8(1) (1990):85-94 85 86 AMER. MALAC. BULL. 8(1) (1990) of A. peroni in the waters of the northern Great Barrier Reef during the Great Barrier Reef Expedition of 1928 to 1929. Tesch (1949) found A. turriculata d’Orbigny, 1836 off the coast of New South Wales. Van der Spoel (1976) reported A. inclinata Souleyet, 1852 off northern Queensland and A. /esueuri Souleyet, 1852 and A. helicinoides Souleyet, 1852 from northwestern Australia. Ralph (1957) did not include atlantids in her review of the heteropods from New Zealand. The global distribution and the taxonomy of Atlanta gaudichaudi is uncertain as this species is easily confused with A. peroni and A. plana Richter, 1972 (Tesch, 1949; Thiriot- Quiévreux, 1973; Richter, 1974). Jamieson and Newman (1989), in their study of the sperm ultrastructure of A. gaudichauadi, used specimens from Heron Island, thus extending the known range of this species in Australian waters southward from the Torres Strait to southern Queensland. Swimming behaviour of atlantids has been briefly described by Tesch (1949), Wilbur and Yonge (1964), and Land (1982). Lalli and Gilmer (1989) reviewed the literature on atlantid biology and included some personal in situ observa- tions on these animals. Observations of atlantids feeding on pteropods and gastropod larvae were reviewed by Thiriot- Quiévreux (1973), Richter (1982), and Lalli and Gilmer (1989). Anatomical and histological studies of the digestive and reproductive systems were made by Martoja and Thiriot- Quiévreux (1975a, b) and Thiriot-Quiévreux and Martoja (1976). Organogenesis of the veligers of six atlantid species (not in- cluding A. gaudichaudi) have been documented in detail by Thiriot-Quiévreux (1969). Pilkington (1970) also documented the metamorphosis of an unidentified atlantid from New Zealand waters. This paper details the morphology of Atlanta gaudichaudi, collected within waters of the Great Barrier Reef (hereafter GBR) and provides for the first time descriptions of the live animal, veliger and radula for an Australian atlantid species. Data on seasonal distribution within the northern and southern GBR are presented, together with laboratory obser- vations on swimming behaviour, feeding and prey selection, and the metamorphosis of late veliger larvae. MATERIALS AND METHODS Plankton collections were made from surface waters around Lizard Island (northern GBR, 14°14’S, 145°27’E) and Heron Island (southern GBR, 23°27’S, 151955’E) (Fig. 1) from December 1985 to February 1988. Each sampling trip was of two to three weeks duration. Approximately 200 samples (20 samples per trip) were collected from the upper 1.0 to 1.5 m of the water column during daylight hours in summer and winter months. The bottom depth varied between 10-30 m at both locations. Quantitative plankton nets (with Oceanic Model No. 2030 flowmeters), 300 mm mouth diameter, with 0.2 mm mesh size were used. Tows were made once daily for 10 min duration, this being brief enough to minimize shell damage and ensure animals were alive after collection. All nets were equipped with soft cod ends (made of canvas) to further reduce shell damage. Immediately after collection, specimens were sexed and measured with an ocular micrometer on a 145° 150° YY j, UW X LIZARD ISLAND ~ \ 1s° — X HERON Ny QUEENSLAND ISLAND 100 200 = 300 km Fig. 1. Map of Queensland, Australia showing the locations of Lizard and Heron islands on the Great Barrier Reef. Wild M5 dissecting microscope (accuracy to 0.01 mm). Size was expressed as the maximal shell diameter, excluding the keel which is often broken after collection. Animals were subsequently preserved in 70% EtOH. The numbers of Atlanta gaudichaudi present in each sample (=abundance) were expressed as number of animals/m3; the volume filtered being calculated from flowmeter readings, net diameter and distance towed. Com- parisons of mean abundances between sampling periods were tested by One-Way ANOVA and Student’s t-tests (tested at the 95% confidence level) on log transformed data (Table 1). The relative percent of each life history stage was calculated for each sampling trip from the actual number of animals measured. Individuals were retained in the laboratory in aerated 250 ml containers for feeding observations, and 5 | containers for swimming behavioural studies. Seawater was changed twice daily. In feeding trials, Zooplankton prey was introduced individually and in groups to each container. Zooplankton was removed within four hours if it was not consumed by an atlantid. Live animals from plankton samples with food items in their oesophagus were induced to disgorge their gut con- tents by the addition of a drop of 5% gluteraldehyde. Alcohol preserved specimens for scanning electron microscopy were sonically cleaned for less than 10 sec, placed NEWMAN: ATLANTA GAUDICHAUDI FROM AUSTRALIA 87 on double-sided sticky tape on aluminum stubs and gold coated. Radulae were dissected from preserved animals with the buccal tissue intact, soaked in 10% KOH for 24-48 hours, then rinsed in distilled water. Due to the small size of the radulae (less than 0.3 mm in length) each radula was manipulated by fine insect pins and placed in a drop of distilled water on a stub covered with double-sided sticky tape. As the water evaporated, the radula was positioned on the stub and then coated. Photomicrographs were taken with 400 ASA Pan-X black and white film, at 20 KV with a Philips 505 scanning electron microscope. RESULTS AND OBSERVATIONS SPECIES DESCRIPTION Atlanta gaudichaudi Souleyet, 1852 (Figs. 2-4) (for synonymies see van der Spoel, 1976) The shell of this species is smooth and devoid of sur- face sculpture except for one fine spiral line on the outer edge of the apical whorls. The shell has four whorls, the outer whorl is relatively large and expands into an inflated aperture. The three apical whorls gradually increase in width. The spire is relatively low, not extending past the height of the body whorl. The keel does not penetrate between the outer whorl. The shell is transparent and colourless except for a faint red-brown colour at the base of the keel. Adults range from 1.00 to 2.25 mm in diameter. Sexually mature males are distinguished by a bilobed penis on the right side of the neck and, if present, a black and white oval spermatophore. Females are dis- tinguished by the presence of developing ova, seen clearly through the shell in the apical whorls. The eye tentacles are equal in length and do not ex- tend beyond the distal end of the proboscis. The body of the animal is colourless. The eyes are of Richter’s (1974) Type b, having a transverse slit present in the distal portion of the pigmented retina. The operculum is also Richter’s (1974) Type b, micro-oligogyre, or having only a small gyre on its periphery. The radula of Atlanta gaudichaudi is typically taenioglossan in morphology, with a formula of 2-1-1-1-2 (Figs. 3C, D). Veligers ranged in shell diameter from 0.04 to 0.40 mm. The shell is smooth and colourless, with a single spiral line on the outer edge of each whorl (Figs. 3A, B). In the smallest veligers, the velum is divided into four lobes, becoming six lobes prior to metamorphosis (Fig. 4B). The lobes, prior to metamorphosis, are only slightly wider than the shell diameter. The shell and body are colourless except for a red-brown or yellow-orange colouration of the digestive gland. Late veligers have well developed eyes with a clear lens and black retina, eye tentacles, and a small fin with a sucker. The aper- ture of the shell is indented and formation of a small keel can be seen. Juveniles ranged from 0.50 to 0.90 mm in shell diameter. Young males are distinguished by the presence of two small penial ‘‘buds’’ on the right side of the neck. Fig. 2. Scanning electron micrographs of Atlanta gaudichaudi: A, dorsal view; B, aperture view; C, close-up of apical whorls; D, umbilicus (A-C, scale bar = 1 mm; D, scale bar= 0.1 mm). 88 AMER. MALAC. BULL. 8(1) (1990) REMARKS Atlanta gaudichaudi can be easily confused with A. peroni and A. plana which are similar in shell morphology when examined by only light microscopy (Table 2). A. gaudichaudi mainly differs from A. peroni in maximum shell diameter and morphology of the apical whorls. A. gaudichaudi reaches a maximal diameter of 5 mm (twice the diameter recorded in the present study) whereas A. peroni can reach up to 11 mm (van der Spoel, 1976). Richter (1974), using SEM has shown that the apical whorls of A. peroni from the Indian Ocean have a distinct layer of callus running along the suture and a spiral line on the centre. In contrast, the apical whorls of A. gaudichaudi only have a faint spiral line on the outer margin. The shell aperture is larger in A. gaudichaudi than in A. peroni. The keel only penetrates between the outer whorl in large specimens of A. gaudichaudi whereas the keel strong- ly penetrates between the outer whorl of A. peroni (Tesch, 1949; van der Spoel, 1976). Differences are seen in the morphology of the lateral radular teeth of Atlanta gaudichaudi from different oceans as Fig. 3. Scanning electron micrographs of Atlanta gaudichaudi: A, late veliger, dorsal view; B, late veliger, aperture view; C, whole radula; D, median and lateral radular teeth (A, C, scale bar= 0.1 mm; D, scale bar= 10 um). specimens from Lizard and Heron islands do not possess a secondary spine on the inner margin as shown by Richter (1961) for A. gaudichaudi. Atlanta plana is also similar in shell shape to A. gaudichaudi. The most important difference is that the oper- culum of A. plana has fine hooks or spines on its central gyre (Richter, 1974), whereas the operculum of A. gaudichaudi does not have spines. In adults, the eyes are Type a (large pigmented retina and cuboidal shape) in A. plana compared with Type b in A. gaudichaudi (Richter, 1974). Adult A. plana can also be distinguished by the purple-violet colour of the apical whorls (van der Spoel, 1976). Veligers of A. plana were rarely collected in this study but they could be distinguished from those of A. gaudichaudi by their darker purple colour and the numerous fine spiral striae on the shell surface. SEASONAL DISTRIBUTION Atlanta gaudichaudi was the most common species en- countered, being found in all seasons at both Lizard and Heron islands. A. lesueuri, A. plana, A. helicinoides, and A. Table 1. T-Test of Log mean abundance of Atlanta gaudichaudi from Lizard and Heron islands (‘significant at 95% confidence level). Heron Summer vs. Lizard Summer Heron vs. Lizard Heron Winter vs. Lizard Summer | 0.197 | 6.398" | 6.679* | Heron Summer vs. Lizard Summer vs. Heron Winter Lizard Winter | 9.040* 3.885* | NEWMAN: ATLANTA GAUDICHAUDI FROM AUSTRALIA 89 Table 2. Comparison of the morphological features of Atlanta gaudichaudi, A. peroni and A. plana (from Richter, 1974; van der Spoel, 1976). Species Maximum Number of Apical Diameter (mm) | Whorls Whorls A. gaudichaudi 5 4 smooth, 1 spiral line A. peroni 11 5 smooth, Lesueur, 1817 1 spiral line, layer of callus A. plana 5 5 Richter, 1972 2 spiral lines Penetration Eyes Operculum Colour of Keel absent Type b Type b base of keel brown/red present Type b Type b base of keel brown absent Type a Type b apical whorls spines on gyre purple/pink fusca Souleyet, 1852 were rarely encountered in either sum- mer and winter months at Lizard or Heron Islands. A. tur- riculata and Oxygyrus keraudreni (Lesueur, 1817) were rarely collected and only at Lizard Island. Overall, the mean abundance of Atlanta gaudichaudi varied between 0.2 to 5.2 animals/m3 at Lizard Island and be- tween 2.2 to 17.3 animals/m3 at Heron Island. Abundances showed the greatest seasonal variation at Heron Island, where higher numbers were encountered in summer than winter months (Fig. 5). At Lizard Island greater numbers were found in winter, and the range of mean abundances was not as broad as at Heron Island. Comparisons of the mean abundances for combined summer versus winter samples (Fig. 6, Table 1) indicated high variability in the number of animals collected. Although, there was no significant difference in the overall abundance of Atlanta gaudichaudi between Lizard and Heron Islands (all sampling trips combined) there were highly significant seasonal differences both between and within sampling areas (95% confidence level). The percentages of each life history stage present in the waters off Lizard and Heron islands showed an overall similarity in population structure (Figs. 7, 8). The majority of individuals collected were veliger stages in both sampling seasons. Figure 7 shows that juvenile, male and female com- ponents each represented less than 10% of individuals from Lizard Island. At Heron Island, juveniles accounted for up to 20% (Winter 1986) and males and females less than 10% of individuals collected (Fig. 8). There are no clear seasonal dif- ferences in the ratios of the veliger, juvenile or adult stages for either Lizard or Heron islands. Usually, juveniles accounted for only a slightly larger ratio of animals compared to adults. A relatively higher ratio of juveniles were found in winter com- pared to summer for Heron Island. Generally there were more females than males at both sampling locations, although the actual numbers of adults were too low to permit statistical analyses. BIOLOGY Observations were made on the swimming behaviour of Atlanta gaudichaudi in the laboratory. The maximum period individuals could be kept alive was seven days. When swim- ming, individuals move in a circular course and the motion is jerky and sporadic (Fig. 9A). The shell aperture is directed towards the water surface and the foot is directed dorsally dur- ing swimming. The well developed foot muscles move the fin from side to side in a flapping motion which is counteracted f. . CO \ Bere , XG é B, \ 3} peprerrrrrr \ f \ a? ip) eF Fi 4 ed aie te -- FZ \ \ 4 (oS y {¢ ITM 1 Ss AX lp L , SS 2. oe. Zi a. Oe Ss” SK SX Sc Ry 4 YS pe: Fig. 4. Morphological features of Atlanta gaudichaudi: A, male with body extended (after Jamieson and Newman, 1989); B, late veliger, prior to metamorphosis (E, eye; ET, eye tentacle; F, fin; FT, foot; K, keel; O, operculum; P, penis; PR, proboscis; S, sucker; SH, shell; SP. spermatophore; V, velum). 90 AMER. MALAC. BULL. 8(1) (1990) by the swinging of the shell. Upward swimming always resulted in the animal hitting the water surface with its proboscis. Frequently, animals were seen to ‘attach’ to the water surface with their proboscis for a few minutes (Fig. 9B). When sinking, the head is tilted backwards, the fin and foot held vertically above. Sinking rates ranged from 0.5 cm/sec for juveniles, to 1.7 cm/sec for adults. Some were observed to slow their sinking rate and remain suspended in the water column. When currents were generated in the container, in- dividuals were seen suspended in a resting position with the aperture directed toward the water surface and the fin and foot held motionless above (Fig. 9C). When the currents sub- sided, all animals sank to the bottom. Once on the bottom, they spent a few minutes cleaning their shell, keel and foot with their proboscis. Most individuals, when disturbed with a fine probe, appeared to have mucus ‘threads’ attached to their aperture or foot. Suspended animals were also observed on a few occasions hanging to these threads in the water column. Atlanta gaudichaudi capture euthecosomatous pteropods by grabbing the prey’s shell with their fin and fin sucker. The fin sucker attaches to the shell, the fin wraps around it and the proboscis extends into the prey’s aperture (Fig. 10). Slowly, the prey’s wings are drawn out and into the atlantid’s oesophagus. The atlantid does not let go of its prey until the pteropod’s wings and viscera are consumed. In the case of creseid pteropods [Creseis acicula (Rang, 1828), C. virgula (Rang, 1838) and C. chierchiae (Boas, 1886)], the col- umellar muscles attaching the body to the shell were severed by the radula of the atlantid. Animals were also observed feeding on limacinid pteropods [Limacina trochiformis (d’Orbigny, 1836) and L. inflata (d’Orbigny, 1836)]. The atlantid’s sucker grasped the coiled shell while the proboscis removed the wings and viscera. Capture of food took from one minute to one hour, while digestion of ingested food, observed as it moved into the stomach, took up to 6 hours. On one occasion, A. gaudichaudi was seen to consume a Mean Abundance /m3 S85/86 W86 S86/87 W87 S$87/88 Sampling Seasons Fig. 5. Mean seasonal abundance of Atlanta gaudichaudi at Lizard and Heron islands, 1985-1988. Log Mean Abundance Lizard Is. Heron Is. Fig. 6. Summer and winter averages in the Log mean abundance of Atlanta gaudichaudi from Lizard and Heron islands, 1985-1988. gymnosomatous pteropod (Pneumodermopsis spoeli Newman and Greenwood, 1988). Other zooplankters offered to captive Atlanta gaudichaudi included copepods, prosobranch and bivalve veligers, crab larvae, mysids, ostracods, chaetognaths, salps, medusae, larval fish and radiolarians. None of these zooplankters were attacked or consumed by A. gaudichaudi. Cannibalism was frequently observed on smaller sized atlantids and veligers. Direct observations on food capture showed that A. gaudichaudi most frequently preyed on euthecosomatous pteropods. Examination of food material from the oesophagus of freshly caught Atlanta gaudichaudi revealed the presence of statocysts of euthecosomatous pteropods (as many as five pairs) and pairs of atlantid eyes. Identifiable regurgitated stomach contents also contained parts of limacinid and creseid pteropods and atlantids. Late stage veligers swam with the aid of their lobes as well as the small fin. On several occasions, the meta- morphic loss of the velar lobes of late veligers was observed under the dissecting microscope. At metamorphosis the lobes could be clearly seen being consumed by the radula. Con- sumption of the lobes took up to 4.0 hr, with removal of both sides of the velum at the same time. Late veligers were never seen with an asymmetrical velum as described by Pilkington (1970). DISCUSSION Atlanta gaudichaudi is the most common species of heteropod found within the waters of the Great Barrier Reef and is more abundant in the southern region (Subtropical waters), especially in summer months. In the Atlantic Ocean, A. gaudichaudi has a wide latitudinal range of distribution from 6O°N to 25°S (van der Spoel, 1976). Atlanta peroni was not found at either Heron or Lizard islands during this study and it is highly probable that reports of this species from off Low Isles (Russell and Colman, 1935) Relative % Relative % 80 60 40 20 NEWMAN: ATLANTA GAUDICHAUDI FROM AUSTRALIA Summer 1986 100 Winter 1986 £TS n=748 n=1176 & o a 3 7) ~ Ee es Veligers Juveniles Female Male Veligers Juveniles Female Mae Stage Stage Summer 1987 100 Winter 1987 n=39 n=952 80 60 ss o 2; 3 40 o ~ 20 Veligers Juveniles Female Male Veligers Juveniles Stage Stage 100 Summer 1988 n=517 Relative % Veligers Juveniles Female Male Stage Fig. 7. Relative percent of each life history stage of Atlanta gaudichaudi from Lizard Island, 1986-1988. 91 92 AMER. MALAC. BULL. 8(1) (1990) 100 Summer 1985/86 80 = Winter 1986 Qy n=334 WY n=77 80 SS NV : 60 : € Es : - eo c se : 2 : g : a : ‘3 40 2 Si! Ap E 3 . m4 e 4 e i : : By = , Soy , QQy Veligers Juveniles Female Male Veligers Juveniles Female Male Stage Stage 100 Summer 1986/87 100 Winter 1987 n=1153 n=213 re s o Ec E a Ss 7 Oo wv ~ 0 Veligers Juveniles Female Male Veligers Juveniles Female Male Stage Stage 100 Summer 1987/88 EOE, n=1736 80 604 40 Relative % 20 Veligers Juveniles Female Male Stage Fig. 8. Relative percent of each life history stage of Atlanta gaudichaudi from Heron Island, 1985-1988. NEWMAN: ATLANTA GAUDICHAUDI FROM AUSTRALIA 93 are erroneous. Support for this contention lies in the facts that: A. peroni is easily confused with A. gaudichaudi due to similarities in shell shape; ‘A. peroni’’ is the only species col- lected by these authors in the GBR region; their specimens were all relatively small in size (2.5 to 4.5 mm); they found no apparent seasonal change in abundance of ‘A. peroni”’ within the GBR Lagoon (also found for A. gaudichaudi at Lizard Island in the present study); smaller individuals were collected from July to September and larger animals were found from October to January (these authors did not account for veligers of “A. peroni’’). Similarly a higher ratio of juveniles to adults of A. gaudichaudi was found at Heron Island during winter months in the present study. The majority of individuals of At/anta gaudichaudi were in the veliger stage at both sampling locations. This could ac- count for this atlantid being overlooked or considered seem- ingly rare in past studies. The number of individuals varied greatly between sampling trips, but the relative percentage of veligers remained high. Richter (1968) also found high numbers of atlantid veligers within surface waters. In addi- tion to natural mortality, the high number of veligers could have been due to bias in sampling due to the small mesh size used, or to differences in diurnal vertical migration of the life history stages. Richter (1973) showed that heteropod larvae have a more extensive diurnal migration than the adults and concluded that a nocturnal maximum in surface waters is due to the vertical migration of larvae. Nocturnal samples were not collected during the present study. Frontier (1973) showed that veligers of A. gaudichaudi from the coast of Madagascar had a neritic distribution while the adults were oceanic and results from this study indicate a similar distributional pattern. Heteropods are known to swim with the aperture of the shell pointing towards the water surface and the foot directed dorsally. The eyes scan the water below in a slow smooth mo- tion for stationary prey that reflect the light from above (Land, 1982). Tesch (1949) described swimming as a continual jerky movement, keeping the animal afloat. Atlantids undulate their fin like a sculling blade (Wilbur and Yonge, 1964). My obser- vations indicate that atlantids use their fin for propulsion and their foot and shell as stabilizers. The flat surfaces of the shell, especially the thin keel, counteracts the motion of the fin. Atlantids also appear to rest at the water surface and remain suspended in water currents. It could also be possible that they attach themselves with their proboscis (radula) to other floating surfaces, such as abandoned euthecosomatous pteropod feeding webs or that they use mucus threads as a parachute. Lalli and Gilmer (1989, Fig. 11) observed an uniden- tified atlantid, in situ, suspended from mucus threads originating from the foot. Literature on the feeding biology of atlantids is scarce. Richter (1968) observed Atlanta peroni feeding on euthecosomatous pteropods. Thiriot-Quiévreux (1969, 1973) found; atlantids fed on creseid pteropods, gastropod veligers and atlantids; digestion was completed in 24 hr; and the prey is consumed one at a time. My observations showed that in- dividual A. gaudichaudi often had several pairs of creseid statocysts or atlantid eyes in the oesophagus at any one time, indicating that they feed on more than one prey at a time. Fig. 9. Diagram of observed behaviour of Atlanta gaudichaudi: A, swimming; B, resting at surface; C, sinking. Fig. 10. Diagram of Atlanta gaudichaudi consuming the pteropod, Creseis chierchiae. Richter (1982) examined the stomach contents of Oxygyrus keraudreni and showed that this species feeds mainly on gastropods (pteropods and heteropods) but also copepods and chaetognaths. My results indicate that A. gaudichaudi prefers to feed on creseid and limacinid pteropods under laboratory conditions. It appears from the limited literature that atlantids are selective predators, feeding mainly on other planktonic gastropods. These transparent carnivorous zooplankters probably remain suspended, and motionless at or near the water surface when they are not feeding. Swim- ming would only be employed for escaping predators and for active seizure of prey. ACKNOWLEDGMENTS Research was funded by the Australian Federation of University Women-Queensland Branch, the Hawaiian Malacological Society, the Australian Museum, the Coral Reef Society and the Zoology Depart- ment University of Queensland. Special thanks are given to the Direc- tors and staff of the Heron and Lizard Island Research Stations, Dr. 94 AMER. MALAC. BULL. 8(1) (1990) J. G. Greenwood for his advise and comments, Dr. S. van der Spoel for his critical review, Mr. P. Davie for his comments, the Electron Microscope Centre University of Queensland, Ms. M. Preker for assistance in the field and Mr. A. Flowers for technical support. LITERATURE CITED Cotton, B. C. 1932. Notes on Australian Mollusca, with descriptions of new genera and new species. Records of the South Australian Museum, Adelaide 4(4):537-547. Dakin, W. J. and A. N. Colefax. 1940. The Mollusca. The Plankton of the Coastal Waters off New South Wales. Part 1. Publications of the University of Sydney p. 205-209. Frontier, S. 1973. Zooplancton de la région de Nosy-Bé VI. Ptéropodes, hétéropodes- Premiére Partie: Especes holonéritiques et néritques-internes. Cahiers de Office de la Recherche Scien- tifique et Technique Outre-Mer, Série Océanographie 11(3):273-289. Jamieson, B. J. M. and L. J. Newman. 1989. The phylogenetic posi- tion of the heteropod Atlanta gaudichaudi Souleyet (Gastropoda, Mollusca), a spermatological investigation. Zoologica Scripta 18(2):269-278. Lalli, B. M. and R. W. Gilmer. 1989. Pelagic Snails: The Biology of Holoplanktonic Gastropod Mollusks. Stanford University Press. 259 pp. Land, M. F. 1982. Scanning eye movements in a heteropod mollusc. Journal of Experimental Biology 96:427-430. Martoja, M. and C. Thiriot-Quiévreux. 1975a. Convergence mor- phologique entre l’appareil copulateur des Heteropoda et des Littorinidae (Gastropoda, Prosobranchia). Netherlands Journal of Zoology 25(2):243-246. Martoja, M. and C. Thiriot-Quiévreux. 1975b. Données histologiques sur l’appareil digestif et al digestion des Atlantidae (Proso- branchia: Heteropoda). Malacologia 15(1):1-27. Pilkington, M. C. 1970. Young stages and metamorphosis in an atlantid heteropod occurring off southeastern New Zealand. Pro- ceedings of the Malacological Society of London 39:117-124. Ralph, P. M. 1957. A guide to New Zealand heteropod molluscs. Tuatara 6:116-120. Richter, G. 1961. Die Radula der Atlantiden (Heteropoda, Proso- branchia) und ihre Bedeutung fur die Systematik und Evolu- tion der Familie. Zeitschrift fur Morphologie und Okologie der Tiere 50:163-238. Richter, G. 1968. Heteropoden and heteropodenlarven im oberflachenplankton des Golfs von Neapel. Publicazioni della Stazione Zoologica di Napoli 36(3):347-400. Richter, G. 1973. Field and laboratory observations on the diurnal ver- tical migration of marine gastropod larvae. Netherlands Journal of Sea Research 7:126-134. Richter, G. 1974. Die Heteropoden der ‘Meteor-Expedition in den In- dischen Ozean 1964/65. ‘Meteor’ Forschung-Ergebnisse 17(D):55-78. Richter, G. 1982. Mageninhaltsuntersuchugen an Oxygyrus keraudreni (Lesueur) (Atlantidae, Heteropoda). Beispel einer Nahrungskette im tropischen Pelagial. Senckenbergiana Maritima 14(1/2):47-77. Richter, G. 1986. Zur Kenntnis der Gattung Atlanta II. Atlanta lesueuri Souleyet und Atlanta oligogyra Tesch. Archiv fur Mollusken- kunde 117(1/3):19-31. Richter, G. 1987. Zur Kenntnis der Gattung Atlanta Ill. Atlanta inflata, A. helicinoides, A. echinogyra und A. plana (Prosobranchia: Heteropoda). Archiv fur Molluskenkunde 117(4/6):177-201. Russell, F. S. and J. S. Colman. 1935. The zooplankton. IV. The oc- currence and seasonal distribution of Tunicata, Mollusca and Coelenterata (Siphonophora). Great Barrier Reef Expedition 1928-1929. Science Report on the Great Barrier Reef Expedi- tion 2(7):203-276. Smith, E. A. 1888. Report on the Heteropoda. Scientific Results of the Voyage of the H.M.S. ‘Challenger’, 1873-1876. 23 (Zoology):1-51. Spoel, van der S. 1976. Pseudothecosomata, Gymnosomata und Heteropoda (Gastropoda). Bohn, Scheltema and Holkema. Utrecht. 484 pp. Tesch, J. J. 1906. Heteropoden der Siboga Expedition. Siboga Report 51:1-112. Tesch, J. J. 1949. Heteropoda. Dana Report 34:1-55. Thiriot-Quiévreux, C. 1967. Descriptions de quelques véligeres planc- toniques de gastéropodes. Vie et Milieu 18(2A):303-315. Thiriot-Quiévreux, C. 1969. Organogenése du genre Atlanta (Mollusque, Hétéropode). Vie et Milieu 20(2A):347-396. Thiriot-Quiévreux, C. 1973. Heteropoda. Oceanography and Marine Biology Annual Review 11:273-261. Thiriot-Quiévreux, C. and M. Martoja. 1976. Appareil génital femelle des Atlantidae (Mollusca, Heteropoda). Vie et Milieu 26(2A):201-233. Tokioka, T. 1961. The structure of the operculum of the species of Atlantidae (Gastropoda: Heteropoda) as a taxonomic criterion, with records of some pelagic molluscs in the North Pacific. Publications of Seto Marine Biological Laboratory 9:267-332. Wilbur, K. M. and C. M. Yonge. 1964. Physiology of the Mollusca. Vol. 1, New York. Academic Press. 473 pp. Date of manuscript acceptance: 3 November 1989 SPECIAL PUBLICATIONS OF THE AMERICAN MALACOLOGICAL BULLETIN The Special Publication Series of the American Malacological Bulletin was begun to disseminate collected sets of papers with similar or related themes in a single volume. To date, three such issues have been published, each the result of a special convened symposium. The three Special Editions are PERSPECTIVES IN MALACOLOGY, PRO- CEEDINGS OF THE SECOND INTERNATIONAL CORBICULA SYMPOSIUM, and PRO- CEEDINGS OF THE SYMPOSIUM ON ENTRAINMENT OF LARVAL OYSTERS. Additional Special Editions are planned for the near future. PERSPECTIVES IN MALACOLOGY (Sp. Ed. #1, July, 1985) offers a wide range of papers dealing with molluscan biology of interest to professionals and amateurs alike. These papers were presented as part of a symposium held in honor of Professor M. R. Carriker at the time of his retirement and highlight a variety of recent advances in numerous facets of the study of molluscs. PERSPEC- TIVES IN MALACOLOGY offers insight into some of the frontiers of molluscan biology ranging from deep-sea hydrothermal vent malacofauna to chemical ecology of oyster drills. The PROCEEDINGS OF THE SECOND INTERNATIONAL CORBICULA SYMPOSIUM (Sp. Ed. #2, June 1986) contains numerous papers on this exotic bivalve that has become a significant ‘‘pest’’ organism of several power plants and other industries using cooling waters. The proliferation, spread, functional biology, attempts at industrial control, taxonomy, and many other topics of interest to the malacologist and industrial biologist are addressed in this important special publication. The third special edition of the American Malacological Bulletin, PROCEEDINGS OF THE SYMPOSIUM ON THE ENTRAIN- MENT OF LARVAL OYSTERS (Sp. Ed. #3, October 1986) contains important review papers on the larval biology of the American oyster Crassostrea virginica, as well as intriguing papers on factors that limit productivity of these bivalves and limitations that exist on their dispersal and survival. The impact of cutter-head dredges is addressed in this special edi- tion with special emphasis on the Chesapeake Bay system. To order your copies of PERSPECTIVES IN MALACOLOGY, PROCEEDINGS OF THE SECOND INTERNATIONAL CORBICULA SYMPOSIUM or PROCEEDINGS OF THE SYMPOSIUM ON ENTRAINMENT OF LARVAL OYSTERS, simply fill out the form below. Enclose check or money order made out to the AMERICAN MALACOLOGICAL BULLETIN. PERSPECTIVES IN 2ND INTERNATIONAL ENTRAINMENT OF MALACOLOGY CORBICULA SYMPOSIUM LARVAL OYSTERS Special Edition No. 1 Special Edition No. 2 Special Edition No. 3 AMERICAN MALACOLOGICAL BULLETIN AMU Members $10.00 _ $2000 Ss $14.00 Non-AMU Members $15.00 _ Ss $28.00 ss $20.00 Unitas Members $12.00 _ $22.00 $16.00 Institutions $25.00 ss $37.00 $28.00 Foreign Airmail $10.00 $10.00 $10.00 Foreign Seamail $500 $ 5.00 $ 5.00 SUBTOTALS $ $ $ TOTAL ENCLOSED $ (check or money order made out to: AMERICAN MALACOLOGICAL BULLETIN) Name: Send Orders To: Mailing Address: Dr. Clement L. Counts III AMU Secretary-Treasurer University of Maryland Eastern Shore Box 1106 Princess Anne, Maryland 21853 U.S.A. SB) IN MEMORIAM Max R. Matteson Alan Solem Thomas E. Thompson 96 CONTRIBUTOR INFORMATION The American Malacological Bulletin serves as an out- let for reporting notable contributions in malacological re- search. Manuscripts concerning any aspect of original, unpub- lished research, important short reports, and detailed reviews dealing with molluscs will be considered for publication. Each original manuscript and accompanying illustra- tions must be submitted with two additional copies for review purposes. Text must be typed on one side of 8% x 11 inch bond paper, double-spaced, and all pages numbered con- secutively with numbers appearing in the upper right hand corner of each page. Leave ample margins on all sides. ‘ Form of the manuscript should follow that outlined in the Council of Biology Editors Style Manual (fifth edition, 1983). This can be purchased from the CBE, 9650 Rockville Pike, Bethesda, Maryland 20814, U.S.A. Text, when appropriate, should be arranged in sections as follows: 1. Cover page with title, author(s) and address(es), and suggested running title of no more than 50 characters and spaces. Authors should also supply five key words, placed at the base of this page, for indexing purposes. 2. Abstract (less than 5% of manuscript length) 3. Text of manuscript starting with a brief introduction followed by methodology, results, and discussion. Separate sections of text with centered subtitles in capital letters. 4. Acknowledgments 5. Literature cited 6. Figure captions References should be cited within text as follows: Vail (1977) or (Vail, 1977). Dual authorship should be cited as follows: Yonge and Thompson (1976) or (Yonge and Thomp- son, 1976). Multiple authors of a single article should be cited as follows: Beattie et a/. (1980) or (Beattie et a/., 1980). All binomens should include the author attributed to that taxon the first time the name appears in the manuscript [eé.g. Crassostrea virginica (Gmelin)]. This includes non- molluscan taxa. The full generic name along with specific _ epithet should be written out the first time that taxon is referred to in each paragragh. The generic name can be abbreviated in the remainder of the paragraph as follows: C. virginica. In the literature cited section of the manuscript refer- ences must also be typed double spaced. All authors must be fully identified, listed alphabetically and journal titles must be unabbreviated. Citations should appear as follows: Beattie, J.H., K.K. Chew, and W.K. Hershberger. 1980. Dif- ferential survival of selected strains of Pacific oysters (Crassostrea gigas) during summer mortality. Pro- ceedings of the National Shellfisheries Association 70(2):184-189. Seed, R. 1980. Shell growth and form in the Bivalvia. In: Skeletal Growth of Aquatic Organisms, D.C. Rhoads and R.A. Lutz, eds. pp. 23-67. Plenum Press, New York. Vail, V.A. 1977. Comparative reproductive anatomy of 3 vivipa- rid gastropods. Malacologia 16(2):519-540. Yonge, C.M. and T.E. Thompson. 1976. Living Marine Molluscs. William Collins Son & Co., Ltd., London. 288 pp. Illustrations should be clearly detailed and readily reproducible. All line drawings should be in black, high quality ink. Photographs must be on glossy, high contrast paper. All diagrams must be numbered in the lower right hand corners and adequately labeled with sufficiently large labels to pre- vent obscurance with reduction by one half. Magnification bars must appear on the figure, or the caption must read Horizon- tal field width = xmm or xum. All measurements must be in metric units. All illustrations submitted for publication must be fully cropped, mounted on a firm white backing ready for reproduction, and have author’s name, paper title, and figure number on the back. All figures in plates must be nearly con- tiguous. Additional figures submitted for review purposes must be of high quality reproduction. Xerographic reproduction of photomicrographs or detailed photographs will not be accep- table for review. Abbreviations used in figures should occur in the figure caption. Indicate in text margins the appropriate location in which figures should appear. Color illustrations can be included at no extra cost to the author. Original illustra- tions will be returned to author if requested. Any manuscript not conforming to AMB format will be returned to the author for revision. Proofs. Page proofs will be sent to the author and must be checked for printer’s errors and returned to the printer within a three day period. Significant changes in text, other than printer errors, will produce publishing charges that will be billed to the author. Mailing. All overseas mailing must be done via airmail. The American Malacological Union will not be responsible for deferred publication of manuscripts delayed in surface mail. Charges. There are no mandatory page costs to authors lack- ing financial support. Authors with institutional, grant or other research support will be billed for page charges. The current rate is $35.00 per printed page. Acceptance and ultimate publication is in no way based on ability to pay page costs. Reprints. Order forms and reprint cost information will be sent with page proofs. The author receiving the order form is responsible for insuring that orders for any coauthors are also placed at that time. Submission. Submit all manuscripts to Dr. Robert S. Prezant, Editor-in-Chief, American Malacological Bulletin, Department of Biology, Indiana University of Pennsylvania, Indiana, Penn- sylvania, 15705-1090, U.S.A. Subscription Costs. Institutional subscriptions are available at a cost of $32.00 per volume. [Volumes 1 and 2 are available for $18.00 per volume]. Membership in the American Malacological Union, which includes personal subscriptions to the Bulletin, is available for $20.00 ($15.00 for students). All prices quoted are in U.S. funds. Outside the U.S. postal zones, add $5.00 seamail and $10.00 airmail per volume or membership. For subscription or membership contact AMU Secretary-Ireasurer, Dr. Clement L. Counts Ill, University of Maryland Eastern Shore, Box 1106, Princess Anne, Maryland, 21853, U.S.A. AMERICAN MALACOLOGICAL BULLETIN VOLUME 8 NUMBER 2 APRIL 1991 CONTENTS Genetic studies of Asiatic clams, Corbicula, in Thailand: allozymes of 21 nominal species are identical. VARAPORN KIJVIRIYA, E. SUCHART UPATHAM, VITHOON VIYANANT and DAVID S. WOODRUFF .......................000... Shell and hinge development of young Corbicula fluminea (Muller) (Bivalvia: Corbiculoidea). VICTOR S. KENNEDY, S. CYNTHIA FULLER and MMR RMPR TRUDE AS Fils 082 os Oe rai oh Die ae hE Minh aR IR UG cielo Uh Ga AM Soa pal 5 ak A distributional checklist of the freshwater unionids (Bivalvia: Unionoidea) of Kentucky. RONALD R. CICERELLO, MELVIN L. WARREN, JR. PM MEAUEN TERM A. SGHUS DEMS. 6 cre se ROE re eA on Sil Me Obes Ozarkian fresh-water mussels (Unionoidea) in the upper Eleven Point River, SOM SeGCNTY: TACINDELES SO, UPAE CURRIN 522 Eo ce oes cg x Li diy Pimp ate FO Assy sch ie anes as ge es Reintroduction of the spiny riversnail /o fluvialis (Say, 1825) (Gastropoda: Pleuroceridae) into the North Fork Holston River, Southwest Virginia and Northeast Tennessee. STEVEN A. AHLSTEDT .......................0.. 20.00.0625. Research Note: First record of net collected Ocythde tuberculata (Cephalopoda: Octopoda) from Peruvian waters. FRANZ CARDOSO ..............-.0 00. cee eee. SYMPOSIUM ON SYSTEMATICS, ANATOMY AND EVOLUTION OF WESTERN NORTH AMERICAN LAND MOLLUSCS Familial relationships and biogeography of the western American and Caribbean Helicoidea (Mollusca: Gastropoda: Pulmonata). WALTER B. MILLER and EDNA NARANJO-GARCIA .................0.0.0.0...000020.. A phylogenetic analysis and revised classification of the North American Haplotrematidae (Gastropoda: Pulmonata). BARRY ROTH .....................0....02.... Present status of the micromollusks of Northern Sonora, Mexico. EDNA NARANJO-GARCIA ie NU ee ee eee Sentenia: The next challenge: life styles and evolution. ALAN SOLEM........................... Review: Squid as Experimental Animals. John Messenger. ... 2... ...000 000 ccc le eee eee PAIN UN ORIRIENNIE iC ae OS n a ay oe pe EWS ta Cok Glau eee Sada NR re ek VN, EMU ER MNG Aaa DRO Cy hae Sane EER Rot ed CnC MNS ic tog tae RCE Sa EE AMERICAN MALACOLOGICAL BULLETIN EDITOR-IN-CHIEF ROBERT S. PREZANT Department of Biology Indiana University of Pennsylvania Indiana, Pennsylvania 15705 MELBOURNE R. CARRIKER College of Marine Studies University of Delaware Lewes, Delaware 19958 GEORGE M. DAVIS Department of Malacology The Academy of Natural Sciences Philadelphia, Pennsylvania 19103 R. TUCKER ABBOTT American Malacologists, Inc. Melbourne, Florida, U.S.A. JOHN A. ALLEN Marine Biological Station Millport, United Kingdom JOHN M. ARNOLD University of Hawaii Honolulu, Hawaii, U.S.A. JOSEPH C. BRITTON Texas Christian University Fort Worth, Texas, U.S.A. JOHN B. BURCH University of Michigan Ann Arbor, Michigan, U.S.A. EDWIN W. CAKE, JR. Gulf Coast Research Laboratory Ocean Springs, Mississippi, U.S.A. PETER CALOW University of Sheffield Sheffield, United Kingdom BOARD OF EDITORS ASSOCIATE EDITORS MANAGING EDITOR RONALD B. TOLL Department of Biology The University of the South Sewanee, Tennessee 37375 W. D. RUSSELL-HUNTER Department of Biology Syracuse University Syracuse, New York 13210 CAROLE S. HICKMAN Ex Officio Museum of Paleontology University of California Berkeley, California 94720 BOARD OF REVIEWERS JOSEPH G. CARTER University of North Carolina Chapel Hill, North Carolina, U.S.A. ARTHUR H. CLARKE Ecosearch, Inc. Portland, Texas, U.S.A. CLEMENT L. COUNTS, Ill University of Maryland Princess Anne, Maryland, U.S.A. THOMAS DIETZ Louisiana State University Baton Rouge, Louisiana, U.S.A. WILLIAM K. EMERSON American Museum of Natural History New York, New York, U.S.A. DOROTHEA FRANZEN Illinois Wesleyan University Bloomington, Illinois, U.S.A. VERA FRETTER University of Reading Berkshire, United Kingdom ISSN 0740-2783 THOMAS R. WALLER Department of Paleobiology Smithsonian Institution Washington, D. C. 20560 JOSEPH HELLER Hebrew University of Jerusalem Jerusalem, Israel ROBERT E. HILLMAN Battelle, New England Duxbury, Massachusetts, U.S.A. K. ELAINE HOAGLAND Association of Systematics Collections Washington, D.C., U.S.A. RICHARD S. HOUBRICK U.S. National Museum Washington, D.C., U.S.A. VICTOR S. KENNEDY University of Maryland Cambridge, Maryland, U.S.A. ALAN J. KOHN University of Washington Seattle, Washington, U.S.A. LOUISE RUSSERT KRAEMER University of Arkansas Fayetteville, Arkansas, U.S.A. CFS SRS Fok 0 Faint nthe a ant SB AERC REATS JOHN N. KRAEUTER Baltimore Gas and Electric Baltimore, Maryland, U.S.A. ALAN M. KUZIRIAN Marine Biological Laboratory Woods Hole, Massachusetts, U.S.A. RICHARD A. LUTZ Rutgers University Piscataway, New Jersey, U.S.A. GERALD L. MACKIE University of Guelph Guelph, Ontario, Canada EMILE A. MALEK Tulane University New Orleans, Louisiana, U.S.A. MICHAEL MAZURKIEWICZ University of Southern Maine Portland, Maine, U.S.A. JAMES H. McLEAN Los Angeles County Museum Los Angeles, California, U.S.A. ROBERT F. MCMAHON University of Texas Arlington, Texas, U.S.A. ROBERT W. MENZEL Florida State University Tallahassee, Florida, U.S.A. ANDREW C. MILLER Waterways Experiment Station Vicksburg, Mississippi, U.S.A. BRIAN MORTON University of Hong Kong Hong Kong JAMES J. MURRAY, JR. University of Virginia Charlottesville, Virginia, U.S.A. RICHARD NEVES Virginia Polytechnic Institute and State University Blacksburg, Virginia, U.S.A JAMES W. NYBAKKEN Moss Landing Marine Laboratories Moss Landing, California, U.S.A. A. RICHARD PALMER University of Alberta Edmonton, Canada WINSTON F. PONDER Australian Museum Sydney, Australia CLYDE F. E. ROPER U.S. National Museum Washington, D.C., U.S.A. NORMAN W. RUNHAM University College of North Wales Bangor, United Kingdom AMELIE SCHELTEMA Woods Hole Oceanographic Institution Woods Hole, Massachusetts, U.S.A. DAVID H. STANSBERY Ohio State University Columbus, Ohio, U.S.A. FRED G. THOMPSON University of Florida Gainesville, Florida, U.S.A. NORMITSU WATABE University of South Carolina Columbia, South Carolina, U.S.A. KARL M. WILBUR Duke University Durham, North Carolina, U.S.A. Cover. /o fluvialis (Say, 1825) is the logo of the American Malacological Union and the subject of a paper by Ahlstedt in this issue. THE AMERICAN MALACOLOGICAL BULLETIN is the official journal publication of the American Malacological Union. AMER. MALAC. BULL. 8(2) April 1991 GENETIC STUDIES OF ASIATIC CLAMS, CORBICULA, IN THAILAND: ALLOZYMES OF 21 NOMINAL SPECIES ARE IDENTICAL VARAPORN KIJVIRIYA DEPARTMENT OF BIOLOGY, FACULTY OF SCIENCE RAMKHAMHAENG UNIVERSITY RAMKHAMHAENG ROAD, BANGKOK 10240, THAILAND E. SUCHART UPATHAM VITHOON VIYANANT CENTER FOR APPLIED MALACOLOGY AND ENTOMOLOGY DEPARTMENT OF BIOLOGY, FACULTY OF SCIENCE MAHIDOL UNIVERSITY RAMA 6 ROAD, BANGKOK 10400, THAILAND DAVID S. WOODRUFF DEPARTMENT OF BIOLOGY AND CENTER FOR MOLECULAR GENETICS UNIVERSITY OF CALIFORNIA, SAN DIEGO LA JOLLA, CALIFORNIA 92093-0116, U.S.A. ABSTRACT Freshwater clams of the genus Corbicula, collected from 40 sites up to 1500 km apart in Thailand, and representing 21 nominal species, show no significant geographic variation at 24 electrophoretically detected allozyme loci and are most probably all referable to the widespread Asian species, C. fluminea (Muller, 1774). Thai Corbicula have very little genetic variability: mean number of alleles per locus (A) was very low: no variation (A = 1.0) was detected in 30% of the samples, in the remainder, A < 1.1; mean percentage of loci polymorphic in each sample, P = 4.59% (range: 0.0-12.5%); mean individual heterozygosity, H = 0.011 (range: 0.000-0.025) with one outlier where H = 0.058). The low level of population variability and very low individual heterozygosity suggest that most of the Corbicula in Thailand are facultative self-fertilizers. The small amount of genic diversity detected, and the observed genotype frequencies, are apparently maintained by limited outcrossing, at random with respect to shell phenotype, internal shell color and allozyme genotype. Eighty-eight percent of the samples, including one referable to Corbicula fluminea, cluster at Nei’s genetic distance values of D < 0.01. Only five samples from northeast Thailand stand slightly apart from the others. These very high genetic similarities, coupled with a lack of significant anatomical variation, provide no support for the recognition of more than one species in our samples. Twenty nominal Thai species are synonymized accordingly with C. fluminea; another seven nominal species are candi- dates for synonymy. Freshwater clams of the genus Corbicula (Bivalvia: Corbiculidae) occur throughout Asia, eastern Europe and Africa. One species, C. fluminea (Muller, 1774), is the com- monest freshwater bivalve in Asia and has recently been in- troduced to western Europe, North and South America (Mor- ton, 1986). In the United States it spread rapidly and, as a major economic pest, it has been the subject of considerable study and two international conferences (Britton, 1979, 1986). This research has shown that C. fluminea is quite variable in shell size, shape, sculpture and color: the same conchological characteristics upon which most Corbicula species have been defined. Morton (1979, 1986) reexamined such conchological variation in the numerous nominal species of Corbicula from China and Japan and found that such characters were American Malacological Bulletin, Vol. 8(2) (1991):97-106 97 98 AMER. MALAC. BULL. 8(2) (1991) taxonomically unreliable. He found that, wnen morphological variation was considered together with data on ecology, physiology, demography and reproductive behavior, only two biological species could be recognized in east Asia: freshwater C. fluminea, and estuarine C. fluminalis (Muller, 1774) (Mor- ton, 1986). He suggested that most of the other nominal Asian taxa are synonyms of these two species. This paper examines this hypothesis as it applies to the nominal species of Cor- bicula in Thailand. Most of the original species descriptions of Asian Cor- bicula are due to Prime (1864) and Prashad (1930, and references therein). Subsequently, Brandt (1974) cataloged the molluscs of Thailand and recognized 28 species of Corbicula, five of them new. These systematists worked almost entirely with conchological characters in defining their species and ignored geographic and ontogenetic variation and other biological attributes. The senior author has recently described variation in Corbicula representing 22 nominal species from 40 localities throughout Thailand (Kijviriya, 1990). In agree- ment with earlier reports, she found considerable conch- ological variation among clams referable to the different species recognized by Brandt (1974). She also found signifi- cant variation in shell size, shape, sculpture and color within and between allegedly conspecific populations. Gross anatomy and dissections of the siphon, digestive and reproductive systems, on the other hand, failed to reveal any taxonomically significant variation between any of the popula- tions sampled (Kijviriya, unpub. data). In this paper, we report on detectable genetic variation within and between these populations and conclude that a major taxonomic revision of Thai Corbicula is now required. Genetic variation in Corbicula was studied by elec- trophoretic surveys of multilocus allozyme patterns. Allozymes have proven extremely useful in recent systematic studies of molluscs. Intrapopulation allozymic variation has been used to establish mating systems (Selander and Kaufman, 1973; McCracken and Selander, 1980; Selander and Whittam, 1983), to reveal the existence of sibling species (Davis, 1983; Staub et al., 1990), to reveal cases of interspecific hybridization (Gould and Woodruff, 1986, 1987, 1990; Woodruff and Gould, 1987) and to study environmental effects on population struc- ture (Koehn and Hilbish, 1987; Nevo, 1988). Studies of inter- population allozymic variation have revealed the extent of geographic structuring and variation in widespread species (Gould and Woodruff, 1978; Grant and Utter, 1988) showing whether disjunct populations are or are not conspecific (Palmer et al., 1990; Woodruff et a/., 1988), revealed the historical pattern of a species dispersal (Mulvey et a/., 1988; Woodruff et a/., 1986, 1988), and helped define the limits of semispecies in syngameons (Woodruff and Gould, 1980; Woodruff, 1989). The comparison of allozymes among nominal species has supported numerous taxonomic revi- sions and phylogenetic discussions (Davis et a/., 1981; Ember- ton, 1988; Woodruff and Solem, 1990; Woodruff et a/., 1987). Given the demonstrable utility of such data on allozymic varia- tion, this approach was included in our study of the Thai Corbicula. Three research groups have previously reported studies on allozyme variation in Corbicula. Smith et al. (1979) found no detectable variation at 18 loci in five populations from across North America. In contrast, they found some variation at 12-17 loci in four Asian samples: the proportion of loci that were variable (P) in the Philippines was P = 0.17-0.23, in Hong Kong, P = 0.25, and in Japan, P = 0.76. The Asian clams were thus moderately to highly variable and the mono- morphism of the introduced North American populations was attributed to genetic drift associated with the founder effect during a single colonization. McLeod and Sailstad (1980) then reported consistent (year-round) low levels of genetic varia- tion at seven loci from one population in eastern U.S.A.: P = 0.14, and mean individual heterozygosity, H = 0.005. They concluded that their genetic observations supported the hypothesis that North American Corbicula are self-fertilizing. Subsequently, McLeod (1986) reported a study of variation in other populations from both eastern and western U.S.A. and again found little variation within samples. These earlier reports, especially that showing high genetic variability elsewhere in Asia, suggested that allozymic variation might be used in resolving the relationships among the Corbicula of Thailand. MATERIALS AND METHODS Clams were collected in 1985-1986 at 40 localities (Fig. 1, Appendix) from various provinces throughout Thailand. Most clams were collected from lotic habitats in the rivers or irrigation canals; exceptions were two lake localities (samples 1 and 23). Some clams were stored in liquid nitrogen im- mediately and hand-carried to the University of California, San Diego (UCSD) for allozymic screening and protocol develop- ment; others were maintained alive at Mahidol University and used for routine allozyme studies. Notes on clam density, shell morphology, and habitat type and water pH were recorded. Distributional and conchological criteria (Brandt, 1974) were used to identify 39 of the 40 samples to the species level. Voucher specimens from each sample were deposited in the Museum at the Center for Applied Malacology and En- tomology, Mahidol University, Bangkok. Voucher specimen lot numbers are given in the Appendix. ELECTROPHORESIS Proteins from 30 animals from each sample were studied by electrophoretic analysis using horizontal starch gel slabs following the techniques of Selander et a/. (1971). Prior to electrophoresis, individual clams were thawed arid quick- ly homogenized in 0.1-0.2 ml distilled water. The whole-body homogenate was centrifuged at 4,500 g for 10 min and the supernatant was absorbed onto Whatman No. 1 filter paper wicks (3x9 mm) and inserted into 12% (w/v) horizontal starch gels. Electrostarch Lot 146 (Otto Hiller, Madison, Wisconsin) was used throughout this study. Clams from different samples were run on each gel to facilitate comparisons and sample 17 from the Loei River (northeast Thailand) was used as a control group. Three buffer systems were employed to resolve the 14 enzyme systems and 24 allozymes studied (Table 1). KIJVIRIYA ET AL.: GENETIC VARIATION IN THAI CORBICULA 99 20°N 1O°N | : L 100°E 105°E Fig. 1. Location of sample sites in Thailand. Electrophoresis was carried out at constant voltage for 15 hr by which time a bromophenol blue marker dye had migrated 100-120 mm anodally. Following electrophoresis, each slab was cut into four or five slices and stained for a specific en- zyme following standard methods (Murphy et a/., 1990). Isozymes were numbered, and allozymes were assigned superscripts (a, b or c) in order of decreasing anodal mobili- ty. Commonly used enzyme abbreviations are typeset in italics to indicate the presumed genetic locus. STATISTICAL ANALYSES Data consisting of multilocus genotypes for individual clams scorable at all 24 loci were analyzed using the BIOSYS-1 computer program (Swofford and Selander, 1981). A locus was considered polymorphic if more than one allele was detected. Mean heterzygosity per individual (H) was estimated by direct count. Chi-square and Fisher exact tests were used to compare the frequency of individual genotypes with Hardy-Weinberg expectations for a panmictic population. Population structure was examined using fixation indices or F - statistics calculated for each locus and sample (Wright, 1978). Nei’s (1978) unbiased genetic distance coefficients (D) and Rogers’ (1972) genetic similarity coefficients (S) were calculated and clustered using the unweighted pair group averaging method (UPGMA) for the construction of phenograms and Wagner trees. RESULTS We obtained genetically interpretable results for 24 loci per sample. Descriptions of banding patterns and their inter- pretations are provided by Kijviriya (1990). Allozymic varia- tion in 40 samples from throughout Thailand is summarized in Table 2. 7 The mean number of alleles per locus (A ) was very low: no variation (A = 1.0) was detected in 12 of the 40 samples (30%) and in the remaining 28 samples, A < 1.1. The mean percentage of loci that were polymorphic in each Table 1. Electrophoretic buffers giving optimal resolution of 24 allozymic loci in Corbicula sp. Allozyme (E.C. No.) Abbreviation Buffer * Alkaline phosphatase (3.1.3.1) Alp TC 6.0 Aspartate aminotransferase (2.6.1.1) Aat-1 TC 6.0 Aat-2 TC 6.0 Esterase (alpha-naphthy! acetate) (3.1.1.-) Es-1 LiOH Es-4 LiOH Es-5 LiOH Glucose-6-phosphate dehydrogenase (1.1.1.49) G6pdh TC 80 Glucose-6-phosphate isomerase (5.3.1.9) Gpi TC 80 Isocitrate dehydrogenase (1.1.1.42) Idh-1 TC 6.0 Idh-2 TC 6.0 Leucine aminopeptidase (3.4.11.-) Lap-7 LiOH Lap-2 LiOH Malate dehydrogenase (NAD) (1.1.1.37) Mdh-1 LiOH Mdh-2 LiOH Malate dehydrogenase (NADP +) (1.1.1.40) | Mdhp-1 LiOH Mdhp-2 LiOH Peptidase (L-leucyl-L-alanine) (3.4.-) Pep-A-1 TC 6.0 Pep-A-2 TC 6.0 Peptidase (L-leucylglycylglycine) (3.4.-) Pep-B-2 LiOH Pep-B-3 LiOH Phosphoglucomutase (5.4.2.2) Pgm-1 TC 80 Pgm-2 TC 80 6-Phosphogluconate dehydrogenase (1.1.1.44) Pgdh TC 6.0 Xanthine oxidase (1.1.1.204) Xdh TC 6.0 *TC 60: 0.188 M Tris, 0.065 M citrate, adjusted to pH 60; diluted 1:9 for gels and 1:5 for electrodes (14 hr, 80 v). LiOH. solution A: 0.03 M LiOH, 0.19 M borate, pH 8.1; solution B: 0.008 M citrate, 0.05 M Tris, pH 8.4; 1A:9B for gels, A for electrode. TC 8.0. electrode: 0.678 M Tris, 0.157 M citrate, adjusted to pH 8.0; gel: 22.89 mM Tris, 5.22 mM citrate, pH 8.0 (14 hr, 100 v). 100 AMER. MALAC. BULL. 8(2) (1991) sample was P = 4.59% (range: 0.0-12.5%). Mean individual heterozygosity (H) was also very low: H = 0.011 (range: 0.000-0.025) with one outlier at Mekong River-5 (sample 15) where H = 0.058. As 24 loci were surveyed in every sample and as mean sample size per locus was 21-28, our results are adequate to conclude that the Thai Corbicula have very little genetic variability. Minor genetic variation was detected at 28 localities (70%) around Thailand; the remaining 12 were isogenic. Six- teen samples had 1 polymorphic locus, 8 had 2 polymorphic loci, and 4 had 3 polymorphic loci. This variation was detected in 5 loci, four of which were di-allelic and one was tri-allelic. In no case, however, were more than two alleles detected at a locus in a single population. There was no discernible geographic pattern to this variation although levels of P were slightly higher in the northeast (P = 0.065) than elsewhere (P = 0.03 - 0.046). Where a locus was polymorphic at a particularly locali- ty, the observed genotype frequencies were compared with the frequencies expected under panmixia. Using the chi- square test, the observations were in agreement with Hardy- Weinberg expectations in 40 out of 43 tests (P > 0.05). In the remaining three cases, two (both involving Lap-2) were also in agreement with Hardy-Weinberg expectation when the Fisher exact test was used. Only one case (involving Es-5 in sample 15) failed both tests: we detected 6 AA, 24 AB and no BB individuals. This accounts for the elevated heterozygosi- ty noted at this locality. In the Thai Corbicula, we found the simple morphological dichotomy between white and purple internal shell color to be generally inapplicable (Kijviriya, 1990); like Morton (1987) in Hong Kong, we recognized more intermediate color types. However, in eight parts of Thailand we found populations with approximately bimodal distributions of col- or morphs (samples: 9, 11, 15, 18-19 and 35-37). When these morph-sorted subsamples were compared they were found to be indistinguishable at the 24 allozymic loci. The genes we examined included three of the six loci found to show fixed differences between the contrasting morphotypes studied in Texas by Hillis and Patton (1982). Table 2. Allele frequencies for variable loci* in 40 samples of Thai Corbicula, with summary statistics of genetic variability at 24 loci**. Sample Es-5? Lap-2? Mdh-1? Pep-1? Pgdh° N A P H —- = 1 1.00 0.95 1.00 0.80 0.93 28 i 0.13 0.02 2 1.00 1.00 1.00 1.00 1.00 25 1.0 0.00 0.00 3 1.00 1.00 0.83 1.00 1.00 26 1.0 0.04 0.01 6 1.00 0.88 1.00 0.95 1.00 24 1.1 0.08 0.01 ii 1.00 0.93 1.00 1.00 1.00 26 1.0 0.04 0.003 9 0.90 0.93 0.92 1.00 1.00 25 t 0.13 0.02 10 1.00 1.00 0.88 1.00 1.00 25 1.0 0.04 0.01 11 1.00 1.00 0.97 1.00 0.78 26 1.1 0.08 0.02 12 1.00 1.00 1.00 0.87 1.00 26 1.0 0.04 0.01 13 1.00 1.00 1.00 0.85 1.00 27 1.0 0.04 0.01 14 1.00 1.00 1.00 0.87 1.00 26 1.0 0.04 0.01 15 0.60 0.85 1.00 0.85 1.00 25 131 0.13 0.06 16 1.00 0.87 0.93 0.95 1.00 27 14 0.13 0.02 18 0.23 1.00 1.00 0.93 1.00 26 1.1 0.08 0.03 19 0.10 0.95 1.00 1.00 1.00 26 1.1 0.08 0.01 20 0.00 0.85 1.00 0.80 1.00 26 te 0.08 0.02 21 0.00 1.00 1.00 0.85 1.00 26 1.0 0.04 0.01 22 0.20 1.00 1.00 0.90 1.00 25 1A 0.08 0.03 24 0.97 1.00 0.85 1.00 1.00 24 1.1 0.08 0.02 25 0.80 1.00 1.00 1.00 1.00 25 1.0 0.04 0.02 26 0.97 1.00 1.00 1.00 1.00 25 1.0 0.04 0.003 28 1.00 0.83 1.00 1.00 1.00 25 1.0 0.04 0.01 31 1.00 1.00 0.78 1.00 1.00 25 1.0 0.04 0.02 32 1.00 1.00 1.00 0.87 1.00 24 1.0 0.04 0.01 33 1.00 0.90 1.00 0.92 1.00 25 1 0.08 0.02 34 1.00 0.92 1.00 1.00 1.00 26 1.0 0.04 0.01 35 1.00 0.82 1.00 1.00 1.00 26 1.0 0.04 0.02 36 1.00 0.87 1.00 1.00 1.00 24 1.0 0.04 0.01 39 1.00 0.88 1.00 1.00 1.00 23 1.0 0.04 0.01 | *Each variable locus has two alleles except for Pdgh where c/a and c/b occur in samples 1 and 11, respectively. All other loci (see Table 1) were monomorphic. **Samples described in the Appendix. The following 11 samples are not shown but were identical to sample 2: 4, 5, 8, 17, 23, 27, 29, 30, 37, 38, 40 (i.e. N = 22-27, A = 1.0, P = 0,H = 0), where N = mean sample size per locus, A = mean no. alleles per locus, P = proportion of loci polymorphic, H = mean individual heterozygosity. KIJVIRIYA ET AL.: GENETIC VARIATION IN THAI CORBICULA 101 We tested the relationship between genic variability (P) and local abundance by simple regression analysis. As ex- pected, we found a significant negative correlation between abundance and sample area (r2 = 0.4). There was, however, no significant relationship between sample area or abundance and P (r2 = 0.13 and 0.09, respectively). Population variabili- ty was not simply a function of population size. We looked for ecological factors that could be associated with locally elevated levels of P The four most variable samples (samples 1, 9, 15 and 16) showed no con- sistent pattern with respect to water depth, water temperature, pH or habitat type. We are unable to distinguish such sam- ple sites ecologically from sites with isogenic clams. Genetic relationships between the 40 samples were studied by calculating various measures of genetic distance and clustering these values using various tree building algorithms. A phenogram (not shown) based on Nei’s genetic distance (D) did not discriminate between almost 90% of the samples; one based on Rogers’ genetic similarity resolved the trivial differences between some samples (Fig. 2). Both algorithms show that all the Corbicula studied from throughout Thailand are virtually identical to one another at the 24 allozyme loci examined. DISCUSSION The interpretations concerning allelic variation were made in a very conservative fashion. Genetically interpretable variation was seen in 5 loci: Es-5, Lap-2, Mdh-1, Pep-A-1 and Pgdh. |\n the majority of cases (39 out of 44), where a locus was polymorphic at a particular locality, only one of the two homozygote genotypes was observed. Both homozygote genotypes were, however, seen in at least one sample for 4 of the 5 polymorphic loci (Es-5 being the exception). Thus, our genetic interpretations are supported by the observation of all possible genotypes segregating in Thai Corbicula in the majority of the variable allozymes. Another laboratory using different techniques could find variation where we have found none as single-gel elec- trophoretic surveys are known to fail to detect about 20% of true sample variability (Ayala, 1982; Selander and Whittam, 1983). For this and other reasons, the comparison of electro- phoretic results between different laboratories is not recom- mended (Nei, 1987). However, some comments on the results of Smith et a/. (1979), the only other published survey of varia- tion in Asian Corbicula, are appropriate. Their study involved eight unique loci and ten that were shared with our survey. They too found variation in Mdh-1 and Pgdh in Asian Cor- bicula. In contrast, where they found Gpi-7 was variable in samples from Japan, Hong Kong and the Philippines, we detected no variation in Thailand. The other three polymorphic loci that we detected in Thailand were not studied by Smith et al. (1979). The number of individual clams studied per sample and the number of loci investigated per clam are adequate for the estimation of levels of genic variability. The approx- imate average values of A < 1.1,P = 005 andH = 0.01 are all very low relative to those estimated for many other [a i a lee elle eal 1 gustaviana 2 gustaviana 4 gustaviana 5 sp. Indet. 8 lydigiana 17 blandiana 23 lydigiana 27 fluminea 29 lydigiana 30 lydigiana 37 pingensis 38 noetlingi mm 40 lamarckiana 26 baudoni 12 leviuscula E 14 tenius E 32 noetlingi 13 leviuscula E 6 gustaviana 33 leviuscula r 4 Javanica 34 castanea 36 Iravadica 39 hearadl 28 virescens 35 messageri 16 baudoni E 3 gustaviana 10 lamarcklana 24 lydigiana 31 baudoni NNZNNANZZANZZZWZVZZZ2Z2Z0N0ZZ2Z2000 Z2Z200 0 NY 9 javanica 25 regia 11 tenius NE 15 arata NE 18 moreletiana NE 22 lamarckiana NE 19 bocourti NE 20 blandiana NE 21 solidula NE = ee ee | eee | 0.96 0.98 1.00 GENETIC SIMILARITY Fig. 2. Dendrogram showing relationships among 40 samples of Thai Corbicula generated by UPGMA clustering of Rogers’ (1972) genetic similarity (S) estimates based on 24 loci. The cophenetic correlation is 0.971. species of invertebrates. For example, Nevo et al. (1984) reported that for 361-371 species of invertebrates P = 0.38 and H = 0.10. Our estimates for the Thai Corbicula are also lower than those obtained by Smith et a/. (1979) who found P = 0.17-0.76 in the Philippines, Hong Kong and Japan. The average Thai values are thus closer to those estimated for North American populations where P = 0.04 andH = 0.002 (data summarized by Britton and Morton, 1986). Such mean values can, however, be misleading: the most variable sam- ple in Thailand (Sample 15) had P = 0.125 andH = 0.06 and is thus much more variable than most of the North American populations. Corbicula fluminea has, at various times, been regarded as dioecious, aS monoecious, aS a consecutive protandic hermaphrodite, and as a simultaneous hermaphrodite (Kraemer, 1979; Britton and Morton, 1979, 1986). The low level of population variability and very low individual heterozygosity suggest that most of the Corbicula in Thailand are facultative self-fertilizers. The reduced variability observed is similar to that found in other self-fertilizing molluscs (Selander and Kauf- man, 1973; Selander and Hudson, 1976; Hornbach et a/., 1980; 102 AMER. MALAC. BULL. 8(2) (1991) McCracken and Selander, 1980; Stoddard, 1983; Nevo et ai., 1984). Morton (1983) recognized that different populations of C. fluminea can have different sexual strategies; he noted dif- ferent degrees of self-fertilizatin in lotic and lentic habitats. The small amount of genic diversity detected in 70% of the samples must be maintained by either inbreeding among a few strains with limited outcrossing or by the self- fertilization of a large number of heterozygotes. We favor the former explanation as we failed to find the alternate homozygotes in 88% of the situations where we detected heterozygotes (each situation involved a specific polymorphic locus in a specific sample). Occasional random outcrossing would also account for observations that genotype frequen- cies conformed to Hardy-Weinberg expectations in 98% of the cases and the concomitant lack of a consistent pattern of heterozygote deficiencies expected for obligate self- fertilizing species. Linkage disequilibrium, which is common in self- fertilizing plants, was not detected in the variable samples of Corbicula. There was no significant association between alter- nate alleles at two or more polymorphic loci in any popula- tion. Thus, the variable populations in Thailand are not simply composed of two or more monogenic clones with limited inter- breeding. Instead the pattern we discovered suggests amphimixis is occurring today in some populations that have a recent history of automixis. The proportion of progeny pro- duced by self-fertilization (S) can be estimated from the pro- portion of heterozygous individuals in each subpopulation (Hamrick, 1983; Hillis, 1989). Using Pgdh genotype frequen- cies we estimated the frequency of self-fertilization in sam- ple 10 to be 13%; using Pep-A-7 we estimated the rates to be 29% in sample 1, 29% in sample 20, and 23% in sample 14; using Lap-7 our estimates were 68% in sample 7 and 57% in sample 16. These results indicate that multiple reproduc- tive modes may be used differentially in different localities. The very low levels of genic variation and small sam- ple sizes make it difficult to study population structure using Wright’s (1978) fixation indices or F- statistics. The over-all- loci values of F,, = -—0.13 indicates insignificant inbreeding within each sample. However, when the data are examined on a finer scale, eight significant individual locus F,, values were obtained in eight different samples; i.e. F,, > —0.20 for Es-5 in four samples, and for Lap-2 and Mdh-1, in two different samples each. In contrast, the larger F, = 0.46 indicates significant variation among samples. This value is similar to the average fixation index for self-fertilizing plants (Fy = 0.44; Hamrick, 1983). The reduction in individual heterozygosi- ty in the Thai Corbicula as a whole (due to both substructur- ing and inbreeding) is also quite substantial (F, = 0.39). The phenetic tree based on Rogers’ genetic similarity (Fig. 2) shows how all 40 samples of Thai Corbicula are almost identical at the loci studied. In terms of the more widely reported Nei’s genetic distances, 88% of the samples (35/40) cluster at averaged genetic distance values of D < 0.01; such values are smaller than their associated standard erros (Nei et al., 1986). Most of the Thai Corbicula are thus genetically identical or indistinguishable at a reasonably large sample of allozyme loci. Only five samples from northeast Thailand stand slightly apart from the others: samples 20-22 from the Mun River drainage and samples 18-19 from the Mekong River at Nakhon Phanom. Their genetic differentiation rests primari- ly on their having unusually high frequencies of Est-55 and to a lesser extent on their frequencies of Pep-A-15. Neither of these alleles are unique to these localities or even to north- east Thailand, however, so these samples are differentiated from the others at a barely significant level (D = 0.03-0.04). Not only were these regionally commoner alleles found elsewhere in Thailand, but we failed to detect them at two other Mekong River sites (Samples 14-15) between Nakhon Phanom and the point 300 km downstream where the Mun River enters the Mekong. The overall level of genetic differentiation seen among the 40 samples of Corbicula is well within that expected for conspecific populations of sexually reproducing species. Thorpe’s (1983) survey of 1111 published conspecific com- parisons showed that 98% of the genetic distance estimates were D < 0.10. Self-fertilizing species have received less at- tention, but generally show more geographic differentiation than outcrossing species, i.e. the observed values for Thai Corbicula are much lower than expected for a group of self- fertilizing congeneric species. The very high genetic similarities estimated here provide no support for the recogni- tion of more than one species of Corbicula in our samples. We hasten to add that even within a clade there is no simple relationship between genetic distance and taxonomic level. Some most closely related species are known that are very similar at their allozyme loci (Davis et a/., 1981; Gould and Woodruff, 1987). Nevertheless, the preponderance of the evidence on genetic differentiation associated with species of vertebrates (Avise and Aquadro, 1982), invertebrates generally (Nei, 1987), and molluscs in particular (Woodruff et al., 1988), supports the conclusion that our samples are conspecific. Furthermore, the degree of differentiation seen in the Thai Corbicula is less than that reported in other species of freshwater bivalve (Davis, 1983; Davis et a/., 1981; Kat, 1983; Kat and Davis, 1984). It is significantly less than that reported between clones of other automictic molluscs (D = 0.06-0.12) by McCracken and Selander (1980). The lack of genic differen- tiation in an area the size of Thailand was a little surprising. We failed to detect any evidence for even incipient specia- tion in clams from different river systems. Had we known, a priori, that the Thai Corbicula were primarily self-fertilizing we would have predicted even more differentiation between clones; as was found in North American slugs (McCracken and Selander, 1980) and Australian Thaira Stoddard, 1983, 1985). In the absence of comparable allozyme studies of Cor- bicula from elsewhere in Southeast Asia, we are unable to discuss the evolution or phylogenetic relationships of the Thai clams. Their close similarity probably reflects recency of divergence. Time for the accumulation of genetic variation has been limited and self-fertilization has worked constantly to eliminate heterozygotes and rare alleles. The degree of dif- ferentiation observed among the Thai Corbicula could have evolved in less than 200,000 years if Nei’s (1987) moderate calibration of the protein-electrophoretic clock (with a D value KIJVIRIYA ET AL.: GENETIC VARIATION IN THAI CORBICULA 103 of 1.0 equivalent to about 5 million years) is applicable. The allozymic similarity of the 40 samples of Thai Cor- bicula, together with their lack of taxonomically significant anatomical or conchological variation (Kijviriya, 1990), in- dicates that all are referable to a single species. Two arguments can be made that this species is C. fluminea. First, Morton (1986) and Britton and Morton (1986) have argued that most, if not all, of the Asian freshwater Corbicula are synonyms of C. fluminea. Second, Brandt (1974) identified C. fluminea as occurring at several localities in Thailand and we have studied variation in clams from one of these localities (sam- ple 27) and found that morphologically they are typical of C. fluminea. Furthermore, as we have discovered that these clams are genetically indistinguishable from all other Cor- bicula studied in Thailand, we propose the following taxa be synonymized with C. fluminea (Muller, 1774): Corbicula arata (Sowerby, 1877) (Brandt, 1974:313; pl. 27, fig. 73) C. baudoni Morelet, 1866 (Brandt, 1974:323; pl. 29, fig. 102) C. blandiana Prime, 1864 (Brandt, 1974:313; pl. 27, fig. 72) C. bocourti (Morelet, 1865) (Brandt, 1974:314; pl. 27, fig. 80) C. castanea (Morelet, 1865) (Brandt, 1974:317; pl. 27, fig. 79) C. gustaviana Martens, 1900 (Brandt, 1974:320; pl. 28, fig. 87) C. heardi Brandt, 1974:328; pl. 29, fig. 104 C. iravadica Hanley and Theobald, 1876 (Brandt, 1974:323; pl. 28, fig. 91) C. javanica (Mousson, 1849) (Brandt, 1974:315; pl. 27, fig. 82) C. lamarckiana Prime, 1864 (Brandt, 1974:316; pl. 27, fig. 76-77) C. leviuscula Prime, 1864 (Brandt, 1974:326; pl. 28, fig. 95) C. lydigiana Prime, 1861 (Brandt, 1974:316; pl. 27, fig. 74-75) C. messageri Bavay and Dautzenberg, 1901 (Brandt, 1974: 327; pl. 29, fig. 100) C. moreletiana Prime, 1867 (Brandt, 1974:321; pl. 28, fig. 89-90) . noetlingi Martens, 1899 (Brandt, 1974:319; pl. 28, fig. 88) . pingensis Brandt, 1974:324; pl. 28, fig. 93 . regia Clessin, 1879 (Brandt, 1974:320; pl. 28, fig. 86) . Solidula Prime, 1861 (Brandt, 1974:326; pl. 28, fig. 96 . tenuis Clessin, 1887 (Brandt, 1974:318; pl. 28, fig. 85) . virescens Brandt, 1974:324; pl. 29, fig. 101 As Brandt (1974) has provided fully referenced synonyms, notes on conchology, types, localities and distribution, and excellent photographs of representative shells, on the pages and plates cited above, such data are not repeated here. Kijviriya (1990) also provides color photographs of represen- tative shells (see Appendix for specific figure numbers). These taxonomic conclusions are based on the Thai material examined. We cannot assess the validity of any of these 21 taxa from elsewhere in Asia, e.g. Corbicula gusta- viana from Sumatra or C. javanica from Java. In addition, the following seven nominal Thai species were not collected (C. cyreniformis Prime, 1860; C. gubernatoria Prime, 1869; C. oc- cidentiformis Brandt, 1974; C. pisidiformis Prime, 1866; C. siamensis Prashad, 1929; C. vokesi Brandt, 1974) or were found in inadequate numbers for electrophoretic analysis (C. erosa Prime, 1861). We therefore have no opinion as to the validity of these taxa but suspect that they too are junior synonyms of C. fluminea. QNNNANNYO This, then, is our principal result: the freshwater clams of the genus Corbicula, collected from sites up to 1500 km apart in Thailand, show no significant geographic variation at 24 genetic loci and are referable to the widespread Asian species C. fluminea. This result would be confirmed if it were shown that the Thai Corbicula are closely related to better known C. fluminea from Hong Kong and North America. We predict that the Chinese clams will be very similar genetical- ly to those in Thailand (D < 0.10). Those of the Philippine islands could show greater differentiation associated with founder effects and perhaps reduced variability as in the case of the introduced North American clams. The reported higher levels of variation in Japan (P = 0.76, H = 0.23; Smith et al, 1979) are paradoxical and require confirmation. Perhaps the radically different level of variation in this peripheral popula- tion indicates that the Japanese clams have adopted a dif- ferent mating strategy to those used elsewhere. Alternative- ly, the island could have been colonized repeatedly by clams of diverse geographically origin. In parts of the United States, two morphotypes of Cor- bicula fluminea have been identified based on differences in color of the nacre: white or dark purple. The typical form has colored shells and the dark form has purple shells. Hillis and Patton (1982) found that in Texas the two morphs had fixed differences at six of 26 allozyme loci. McLeod (1986) reported a similar finding. Although Hillis and Patton (1982) and others have used these genetic differences as evidence for the presence of two species in North America, the analyses of Britton and Morton (1986) favor a one-species model. Britton and Morton argued that other conchological, ecological and behavioral traits are associated with this dimorphism in both North America and Hong Kong. These suites of associated characters are apparently local adaptive responses to en- vironmental variables and the contrasting morphotypes are thus better regarded as ecophenotypes. According to Britton and Morton (1986), allozymic variation has now been de- scribed in 22 North American populations of the white morphotype; P = 0.04 (range = 0-19%) and H = Oin 17 populations and H = 0.0013-0.0049 in 5 populations. The 3 samples of the purple morphotype that have been described were all monomorphic (P = 0, H = 0). No other genetic studies have been published, although work in progress on rRNA variation will reopen the question of whether more than one species is present in North America (D. Hillis, pers. comm.). A paradox to emerge from our study concerns Cor- bicula fluminea’s great ecological success despite its innate lack of genetic variability. We have found that one of the com- monest bivalves is as genetically depauperate in parts of its native range as it is in areas it has successfully colonized in the last 50 years. Although some workers have maintained that genic variability is positively related to ecological and evolutionary adaptability, it is clear that C. fluminea’s success is based on a narrow genetic base. McCracken and Selander (1980) found a similar pattern in some European slugs intro- duced into North America. Homozygous self-fertilizing species of slugs occupied a wider range of habitats and had superior colonizing abilities than some more heterozygous outcross- 104 AMER. MALAC. BULL. 8(2) (1991) ing species. Corbicula appears to have such a less variable general purpose genome. Clearly, the larger issues presented by Corbicula cannot be resolved from within Thailand; a broader geographic survey of genic variability and breeding systems in this widespread species is now required. Only when this is done will we begin to understand the relation- ship between genetics, the environment, and C. fluminea’s marked conchological, ecological and behavioral variation. ACKNOWLEDGMENTS We thank Captain Vinyoo Kijviriya, RTN, Dr. Usa Klinhom, Miss Duangduen Ratanaponglakha, Mr. Thongleum Saesonthi and Mr. Somyos Hatsanate for their assistance with collecting clams. Vimonsri Manasarn and M. Patricia Carpenter provided excellent technical assistance in the laboratory in Bangkok and San Diego, respective- ly. The senior author was supported by Ramkhamhaeng University and the National Research Council of Thailand and acknowledges the advice of Drs. Maleeya Kruatrachue and Rojana S. Keawjam. Work in San Diego was supported by the Academic Senate of the Univer- sity of California and, indirectly, by grants from the U.S. National Science Foundation, to DSW. We thank David M. Hillis and an anonymous reviewer for their comments on the manuscript. LITERATURE CITED Avise, J. C. and C. F. Aquadro. 1982. A comparative summary of genetic distances in the vertebrates. Evolutionary Biology 15:151-185. Ayala, F. J. 1982. The genetic structure of species. /n: Perspectives on Evolution, R. Milkman, ed. pp. 60-82. Sinauer, Sunderland, Massachusetts. Brandt, R.A.M. 1974. The non-marine aquatic Mollusca of Thailand. Archiv fur Molluskenkunde 105:1-423. Britton, J. C., ed. 1979. Proceedings of the First International Corbicula Symposium, Texas Christian University Research Foundation, For Worth, Texas. 313 pp. Britton, J. C., ed. 1986. Proceedings of the Second International Cor- bicula Symposium. American Malacological Bulletin Special Publication 2:1-239. Britton, J. C. and B. Morton. 1979. Corbicula in North America: The evidence reviewed and evaluated. /n: Proceedings of the First International Corbicula Symposium, J. C. Britton, ed. pp. 249-287. Texas Christian University Research Foundation, Fort Worth, Texas Britton, J. C. and B. Morton. 1986. Polymorphism in Corbicula fluminea (Bivalvia:Corbiculacea) from North America. Malacological Review 19:1-43. Davis, G. M. 1983. Relative roles of molecular genetics, anatomy, morphometrics and ecology in assessing relationships among North American Unionidae (Bivalvia). In: Protein Polymorphism: Adaptive and Taxonomic Significance, G. S. Oxford and D. Rollinson, eds. pp. 193-222. Academic Press, London. Davis, G. M. 1984. Genetic relationships among some North American Unionidae (Bivalvia): sibling species, convergence, and cladistic relationships. Malacologia. 25:629-648. Davis, G. M. and S. L. H. Fuller. 1981. Genetic relationships among recent Unionacea (Bivalvia) of North America. Malacologia 20:217-253. Davis, G. M., W. H. Heard, S. L. H. Fuller and C. Hesterman. 1981. Molecular genetics and speciation in Elliptio and its relation- ships to other taxa of North American Unionidae (Bivalvia). Biological Journal of the Linnean Society 5:131-150. Emberton, K. C. 1988. The genitalic, allozymic, and conchological evolution of the eastern Triodopsinae (Gastropoda: Pulmonata: Polygyridae). Malacologia 28:159-273. Gould, S. J. and D. S. Woodruff. 1978. Natural History of Cerion. VIII. Little Bahama Bank — a revision based on genetics, morpho- metrics and geographic distribution. Bulletin of the Museum of Comparative Zoology 148:371-415. Gould, S. J. and D. S. Woodruff. 1986. Evolution and systematics of Cerion on New Providence Island — A radical revision. Bulletin of the American Museum of Natural History. 182:389-490. Gould, S. J. and D. S. Woodruff. 1987. Systematics and levels of covariation in Cerion in the Turks and Caicos Islands. Bulletin of the Museum of Comparative Zoology 151(6):321-363. Gould, S. J. and D. S. Woodruff. 1990. History as a cause of Area Effects: An illustration from Cerion on Great Inagua, Bahamas. Biological Journal of the Linnean Society 40:67-98. Grant, W. S. and F. M. Utter. 1988. Genetic heterogeneity on different geographic scales in Nucella lamellosa (Prosobranchia, Thaididae). Malacologia 28:275-287. Hamrick, J. L. 1983. The distribution of genetic variation within and among natural plant populations: /n: Genetics and Conserva- tion, C. M. Schonewald-Cox, C. M. Chambers, B. MacBryde and L. Thomas, eds. pp. 333-348. Benjamin/Cummings, Menlo Park, California. Hillis, D. M. 1989. Genetic consequences of partial self-fertilization on populations of Liguus fasciatus (Mollusca:Bulimulidae). American Malacological Bulletin 7:7-12. Hillis, D. M. and J. C. Patton. 1982. Morphological and electrophoretic evidence for two species of Corbicula (Bivalvia:Corbiculidae) in North America. American Midland Naturalist 108:74-80. Hornback, D. J., M. J. McLeod, S. |. Guttman and S. K. Seilkop. 1980. Genetic and morphological variation in the freshwater clam, Sphaerium (Bivalvia: Sphaeriidae). Journal of Molluscan Studies 46:158-170. Kat, P. W. 1983. Morphologic divergence, genetics, and speciation among Lampsilis (Bivalvia:Unionidae). Journal of Molluscan Studies 49:133-145. Kat, PR W. and G. M. Davis. 1984. Molecular genetics of peripheral populations of Nova Scotian Unionidae (Mollusca:Bivalvia). Biological Journal of the Linnean Society 22:157-185. Kijviriya, V. 1990. Studies of the Asiatic clams (Corbicula Muhlfeld, 1811) in Thailand: Electrophoretic estimates of enzyme varia- tion and the use of anatomy as a species indicator. Doctoral Thesis, Mahidol University, Bangkok, Thailand. xii + 184 pp. Koehn, R. K. and J. B. Hilbish. 1987. The adaptive importance of genetic variation. American Scientist 75:134-141. Kraemer, L. R. 1979. Corbicula fluminea (Bivalvia:Sphaeriacea): The functional and morphology of its hermaphroditism. Bulletin of the American Malacological Union. 1979:40-49. McCracken, G. F. and R. K. Selander. 1980. Self-fertilization and mongenic strains in natural populations of terrestrial slugs. Pro- ceedings of the National Academy of Sciences of the U.S.A. 77:684-688. McLeod, M. J. 1986. Electrophoretic variation in U.S. Corbicula. Pro- ceedings of the Second International Corbicula Symposium. American Malacological Bulletin Special Publication 2:125-132. McLeod, M. J. and D. M. Sailstad. 1980. An electrophoretic study of Corbicula fluminea (Bivalvia:Corbiculacea) in the Catawba River. Bulletin of the American Malacological Union 1980:17-19. KIJVIRIYA ET AL.: GENETIC VARIATION IN THAI CORBICULA 105 Morton, B. 1979. Corbicula in Asia. Proceedings of the First Interna- tional Corbicula Symposium. J. C. Britton, ed. pp. 15-38, Texas Christian University Research Foundation, Fort Worth, Texas. Morton, B. 1983. The sexuality of Corbicula fluminea in lentic and lotic waters in Hong Kong. Journal of Molluscan Studies 49:81-83. Morton, B. 1986. Corbicula in Asia - up-dated synthesis, Proceedings of the Second International Corbicula Symposium. American Malacological Bulletin Special Publication 2:113-124. Morton, B. 1987. Comparative life history tactics and sexual strategies of the fresh and brackish water bivalve fauna of Hong Kong and southern China. American Malacological Bulletin 5:91-99. Mulvey, M., M. C. Newman and D. S. Woodruff. 1988. Genetic dif- ferentiation among West Indian populations of the schistosome- transmitting snail Biomphalaria glabrata. Malacologia 29:309-317. Murphy, R. W., J. W. Sites, D. G. Buth and C. H. Haufler. 1990. Pro- teins 1: Isozyme electrophoresis. /n: Molecular Systematics, D. M. Hillis and C. Moritz, eds. pp. 45-126. Sinauer, Sunderland, Massachusetts. Nei, M. 1978. Estimation of average heterozygosity and genetic distance from a small number of individuals. Genetics. 89:583-590. Nei, M. 1987. Molecular Evolutionary Genetics. Columbia University Press, New York. 512 pp. Nei, M., J. C. Stephens and N. Saitou. 1985. Standard errors of branching points in an evolutionary tree and their application to molecular data from human beings and apes. Molecular Biology and Evolution 2:66-85. Nevo, E. 1988. Genetic diversity in nature. Patterns and theory. Evolu- tionary Biology 23:217-246. Nevo, E., A. Beiles and R. Ben-Shlomo. 1984. The evolutionary significance of genetic diversity: ecological, demographic and life history correlates. In: Evolutionary Dynamics of Genetic Diversity, G. S. Mani, ed. pp. 13-213. Springer Verlag, Berlin. Palmer, A. R., S. D. Gayron and D. S. Woodruff. 1990. Reproductive, morphological and genetic evidence for two cryptic species of northeastern Pacific Nucella. The Veliger 33(4):331-344. Prashad, B. 1930. Revision of the Asiatic species of the genus Cor- bicula. |V. The species of the genus Corbicula from the Sunda islands, the Celebes and New Guinea. Memoirs of the Indian Museum 9:93-203. Prime, T. 1864. Notes on the species of the Family Corbiculidae with Figures. Reprinted in: Proceedings of the First International Cor- bicula Symposium, J. C. Britton, ed. pp. 15-38. Texas Chris- tian University Research Foundation, Fort Worth, Texas. Rogers, J. S. 1972. Measures of genetic similarity and genetic distance. University of Texas Studies in Genetics 7:145-153. Selander, R. K. and R. O.Hudson. 1976. Animal population structure under close inbreeding: The land snail Rumina in southern France. American Naturalist 110:695-718. Selander, R. K. and D. W. Kaufman. 1973. Self fertilization and genetic structure in a colonizing land snail. Proceedings of the National Academy of Sciences of the U.S.A. 70:1186-1190. Selander, R. K., M. H. Smith and S. Y. Yang. 1971. Biochemical polymorphisms and systematics in the genus Peromyscus. 1. Variation in the old-field mouse (Peromyscus polionotus). University of Texas Studies in Genetics 6:49-90. Selander, R. K. and T. S. Whittam. 1983. Protein polymorphism and the genetic structure of populations. In: Evolution of Genes and Proteins, M. Nei and R. K. Koehn, eds. pp. 89-114. Sinauer, Sunderland, Massachusetts. Smith, M. H., J. C. Britton and P. Burke. 1979. Genetic variability in Corbicula, an invading species. Proceedings of the First Inter- national Corbicula Symposium, J. C. Britton, ed. pp. 244-248. Texas Christian University Research Foundation, Fort Worth, Texas. Staub, K. C., D. S. Woodruff, E. S. Upatham and V. Viyanant. 1990. Genetic variation in Neotricula aperta, the intermediate snail host of Schistosoma mekongi: allozyme differences reveal a group of sibling species. American Malacological Bulletin 7:93-103. Stoddard, J. A. 1983. The accumulation of genetic variation in a parthenogenetic snail. Evolution 37:546-554. Stoddard, J. A. 1985. Analysis of species lineages of some Australian thiarids (Thiaridae, Prosobranchia, Gastropoda) using the evolutionary species concept. Journal of the Malacological Society of Australia 7:7-16. Swofford, D. L. and R. K. Selander. 1981. BIOSYS-1: A FORTRAN program for the comprehensive analysis of electrophoretic data in population genetics and systematics. Journal of Heredity 72:281-283. Thorpe, J. P. 1983. Enzyme variation, genetic distance, and evolu- tionary divergence in relation to levels of taxonomic separa- tion. In: Protein Polymorphism: Adaptive and Taxonomic Significance, G. S. Oxford and D. Rollinson, eds. pp. 131-152. Academic Press, London. Woodruff, D. S. 1989. Genetic anomalies associated with Cerion hybrid zones: the origin and maintenance of new electromorphic variants called hybrizymes. Biological Journal of the Linnean Society 36:281-294. Woodruff, D. S., M. P. Carpenter, W. B. Saunders and P. D. Ward. 1987. Genetic variation and phylogeny in Nautilus. In: Nautilus: The Biology and Paleobiology of a Living Fossil, W. B. Saunders and N. B. Landman, eds. pp. 65-83. Plenum Press, New York. Woodruff, D. S. and S. J. Gould. 1980. Geographic differentiation and speciation in Cerion — a preliminary discussion of patterns and processes. Biological Journal of the Linnean Society 14:389-416. Woodruff, D. S. and S. J. Gould. 1987. Fifty years of interspecific hybridization: Genetics and morphometrics of a controlled ex- periment involving the land snail Cerion in Florida. Evolution 41:1022-1045. Woodruff, D. S., L. L. McMeekin, M. Mulvey and M. P. Carpenter. 1986. Population genetics of Crepidula onyx: Variation in a California slipper snail recently established in China. Veliger 29(1):53-63. Woodruff, D. S. and A. Solem. 1990. Genetic variation in the Australian camaenid land snail, Cristiiabrum primum Solem. Veliger 33(2):129-139. Woodruff, D. S., K. C. Staub, E. S. Upatham, V. Viyanant and H. C. Yuan. 1988. Genetic variation in Oncomelania hupensis: Schistomsoma japonicum transmitting snails in China and the Philippines are distinct species. Malacologia 29:347-361. Wright, S. 1978. Evolution and the Genetics of Populations. Vol. 4, Variability Within and Among Natural Populations. University of Chicago, Chicago. 580 pp. Date of manuscript acceptance: 4 February 1991 106 AMER. MALAC. BULL. 8(2) (1991) APPENDIX The 40 samples are listed by number (1-40, in boldface type); locality; (relative abundance); and species name according to criteria of Brandt (1974). These sample numbers are placed by locality on the map of Thailand (Fig. 1). Clam abundance was scored of a scale from 1 (rare) to 5 (superabundant), where absolutes densities per square meter are approximately 10, 100, 500, 1000, and > 1000. In brackets following each species name, we give the voucher specimen’s lot number together with a reference to the color photograph (plate and figure number) of representative shells in Kijviriya (1990). Shells were given Mahidol University, Faculty of Science (MUFS) code numbers and deposited in the Museum of the Center for Applied Malacology and Entomology, Mahidol University. 1. Lake at Lampam, Phattalung province, (5), Corbicula gustaviana (MUFS-THO0-155, pl. 2, fig. 8B). 2. Ta Pi River-1, Chawang district, Nakhon Si Thammarat province, (2), C. gustaviana (MUFS-THOO-157, pl. 2, fig. 8D). 3. Min canal, Chawang district, Nakhon Si Thammarat province, (2), C. gustaviana (MUFS-THO0-156, pl. 2, fig. 8C). 4. Ta Pi River-2, Lampoo island, Muang district, Surat Thani province (3), C. gustavania (MUFS-THOO0-158, pl. 2, fig. 8E). 5. Tachin River, Muang district, Trang province, (4), C. sp. indet. (MUFS-THO0-191, pl. 4, fig. 23). 6. Ngae canal, Khlong Ngae Village, Sadao district, Songkhla pro- vince, (4), C. gustaviana. (MUFS-TH00-154, pl. 2, fig. 8A). 7. Saiburi River, Saiburi district, Pattani province, (5), C. javanica (MUFS-THO0-163, pl. 2, fig. 11A/B). 8. Maeklong River, Muang district, Ratchaburi province, (3), C. lydigiana (MUFS-THOO-168, pl. 2, fig. 12C). 9. Phetchaburi River, Muang district, Phetchaburi province, (2), C. javanica (MUFS-THO0-164, pl. 2, fig. 11C). 10. Phetchaburi irrigation canal, Muang district, Phetchaburi province, (2), C. lamarckiana (MUFSTHO00-172, pl. 3, fig. 13C). 11. Mekong River-1, Knhemmarat district, Ubon Ratchathani province, (2), C. tenius (MUFS-THOO0-189, pl. 4, fig. 21C). 12. Mekong River-2, Huai Klor, Nong Khai province, (3), C. leviuscula (MUFS-THOO-176, pl. 3, fig. 14C). 13. Mekong River-3, Phaeng village, Nakhon Phanom province, (3), C. leviuscula (MUFS-THO0-177, pl. 3, fig. 14D). 14. Mekong River-4, Muang district, Mukdahan province, (4), C. tenius (MUFS-THOO-188, pl. 4, fig. 21B). 15. Mekong River-5, Pakchom district, Nong Khai province, (3), C. arata (MUFS-THO00-142, pl. 1, fig. 1). 16. Suei River, near Nong Khai town, Nong Khai province, (5), C. baudoni (MUFS-TH00-144, pl. 1, fig. 2B). 17. Loei River, Muang district, Loei province, (1), C. blandiana (MUFS- THO0-148, pl. 1, fig. 3A). 18. Pho canal-1, Huai Pho, Nakhon Phanom province, (5), C. morele- tiana (MUFS-THO00-179, pl. 3, fig. 16A/B). 19. Pho canal-2, Huai Pho, Nakhon Phanom province, (5), C. bocourti (MUFS-THOO-150, pl. 1, fig. 4). 20. Mun River, Phibun Mangsahan district, Ubon Ratchathani pro- vince, (4), C. blandiana (MUFS-THO0-149, pl. 1, fig. 3B). 21. Pan River, Yang Talat district, Kalasin province, (3), C. solidula (MUFS-THOO-186, pl. 4, fig. 2). 22. Chi River, Muang district, Mahsarakham province, (3), C. lamarckiana (MUFS-TH00-173, pl. 3, fig. 13D). 23. Ubonrat reservoir, Knon Kaen province, (3), C. lydigiana (MUFS- THO0-167, pl. 2, fig. 12B). 24. Panthong canal, Panthong village, Chonburi province, (4), C. lydigiana (MUFS-THOO-170, pl. 2, fig. 12E). 25. Pra Sae Bon River, Glaeng district, Rayong province, (3), C. regia (MUFS-THO0-185, pl. 4, fig. 19A/B). 26. Chao Phraya River-1, Pamoke district, Angthong province, (3), C. baudoni (MUFS-TH00-146, pl. 1, fig. 2D). 27. Chao Phraya River-2, Krung Thon bridge, Sanghae area, Bangkok, (3), C. fluminea (MUFS-THOO-153, pl. 1, fig. 7A/B). 28. Chao Phraya River-3, Muang district, Nakhon Sawan province, (4), C. virescens (MUFS-THOO0-190, pl. 4, fig. 22). 29. Chao Phraya River-4, Kwae Noi River, Sena district, Ayutthaya province, (3), C. lydigiana (MUFS-THO0-169, pl. 3, fig. 12D). 30. Irrigation canal, Lopburi town, Lopburi province, (3), C. /ydigiana (MUFS-THO0-166, pl. 2, fig. 12A). 31. Ping River-1, Wuttikul bridge, Tak province, (3), C. baudoni (MUFS- THO0-147, pl. 1, fig. 2E). 32. Ping River-2, Naowaraj bridge, Chiang Mai province, (3), C. noet- lingi (MUFS-THOO-180, pl. 4, fig. 17D). 33. Ping River-3, Naowaraj bridge, Chiang Mai province, (3), C. leviuscula (MUFS-THOO-175, pl. 3, fig. 14B). 34. Yom River, Muang district, Prae province, (4), C. castanea (MUFS- THOO-151, pl. 1, fig. 5). 35. Kuang River, Muang district, Lamphun province, (3), C. messageri (MUFS-THOO-178, pl. 3, fig. 15). 36. Moei River-1, Ban Takham, Tak province, (2), C. iravadica (MUFS- THOO-162, pl. 2, fig. 10B). 37. Moei River-2, Mae Sot district, Tak province, (1), C. pingensis (MUFS-THO0-183, pl. 4, fig. 18A). 38. Fang River, Fang district, Chiang Mai province, (3), C. noetlingi (MUFS-THOO-181, pl. 4, fig. 17C). 39. Mae Lao River, Muang district, Chiang Rai province, (1), C. heardi (MUFS-THOO-160, pl. 2, fig. 9B). 40. Kaek River, Wang Thong district, Phitsanuloke province, (2), C. lamrackiana (MUFS-THOO0-171, pl. 3, fig. 13A/B). SHELL AND HINGE DEVELOPMENT OF YOUNG CORBICULA FLUMINEA (MULLER) (BIVALVIA: CORBICULOIDEA) VICTOR S. KENNEDY HORN POINT ENVIRONMENTAL LABORATORY, UNIVERSITY OF MARYLAND, CAMBRIDGE, MARYLAND 21613, U.S.A. S. CYNTHIA FULLER' RICHARD A. LUTZ INSTITUTE OF MARINE AND COASTAL SCIENCES, NEW JERSEY AGRICULTURAL EXPERIMENT STATION, COOK COLLEGE, RUTGERS UNIVERSITY, NEW BRUNSWICK, NEW JERSEY 08903, U.S.A. ABSTRACT Small specimens of Corbicula fluminea (Miller) were collected as juveniles newly released from adults, and larger specimens were extracted from sediment taken from natural beds of adults. Lengths of the prodissoconch | averaged 196.9 + 6.7 um(xX + SD; range = 187-213 um; n = 17); the average length for the prodissoconch II was 233.3 + 5.0 um (range 228-244 um, n = 10). The dissoconch had strong commarginal sculpturing, with less prominent radial striae also present. A ligament pit occurred in the posterior region of the shell hinge. Three cardinal and two lateral teeth developed, and shell tubules were noted in the early dissoconch. Over at least the last 65 years (McMahon, 1982; Counts, 1986), an introduced bivalve generally recognized as Corbicula fluminea (Muller) (Britton and Morton, 1986; but see Hillis and Patton, 1982; McLeod, 1986) has spread throughout the lower U.S. mainland, including sites in Maryland’s water- shed surrounding Chesapeake Bay (Counts, 1981; Kennedy and Van Heukelem, 1985). Adults brood eggs in demibranchs and release juveniles whose shell and hinge development we report on here. MATERIALS AND METHODS Adult Corbicula fluminea were collected in September in Nassawango Creek, Maryland, a short tributary of the Pocomoke River (approx. 38° 10’ N; 75° 27’ W), and were held in the laboratory in glass bowls of fresh water. Juvenile clams that they released were preserved in 95% ethanol. In addi- tion, sediment taken from the clam bed later in November was sieved (500 um) to retain sand and small clams that were col- lected with a pipette under a dissecting microscope. These 1Present Address: MEC Analytical Systems, Inc., 2433 Impala Drive, Carlsbad, California 92009, U.S.A. small clams, and larger juveniles collected from Whites Ferry in the Potomac River, Maryland (approx. 39° 09’ N; 77° 31’ W) in June, were also preserved in 95% ethanol. Preserved clams were prepared for scanning electron microscopy (Fuller et a/., 1989) by placement in 5.25% sodium hypochlorite for about 10 min to remove soft tissues. The resulting separated shells were rinsed in distilled water. Specimens less than or about 250 »m long were mounted on silver tape, with double coated tape used to prevent charg- ing in larger shells. Specimens were coated with about 600 A of gold-palladium and examined under an ETEC Autoscan scanning electron microscope. Orthogonal accuracy (i.e. equal magnification in horizontal and vertical directions) and ac- curacy in measurements were obtained by careful initial calibration of the horizontal and vertical magnifications, and by orienting shells such that anterior and posterior as well as dorsal and ventral margins were at equal working distances from the electron optical system (Fuller et a/., 1989). Shell dimensions were determined by comparison with standard grids photographed at the same magnification as each shell. Length measurements were made from the extreme anterior to posterior margins of each prodissoconch I, prodissoconch Il, or dissoconch. American Malacological Bulletin, Vol. 8(2) (1991):107-111 107 108 AMER. MALAC. BULL. 8(2) (1991) RESULTS AND DISCUSSION Shortly after recently collected adults were placed in glass bowls of fresh water, groups of discharged young could be seen on the bottom of the bowls near the parents. As reported earlier (Kennedy and Van Heukelem, 1985), newly released clams measured 210 - 250 um long (Kk = 236 um, n = 12). They had a relatively straight hinge (Fig. 1), and ex- hibited the D-shape common to many bivalve larvae. The pro- dissoconch | and much if not all of the commarginal pro- dissoconch II (Fig. 1) apparently had formed while the young were being brooded. The dissoconch, produced after release, exhibited a strong commarginal sculpture, with less promi- nent radial striae also present (Fig. 2). This external ap- pearance of the juvenile shell is of interest in that the shells of larger individuals do not exhibit such radial striae, whereas the commarginal sculpture remains prominent. The transitions in external appearance from a shell surface with no growth lines (prodissoconch |) to surfaces with increasingly distinct growth lines (prodissoconch Il, dissoconch) are shown in figure 3. Range of length for prodissoconch | was 187 - 213 um (xX + SD = 1969 + 6.7 um, n = 17) and for prodisso- conch II was 228 - 244 pm (XK = 2333 + 50 um,n = 10). Shells gradually changed from D-shaped to oval as the umbones developed (400 - 600 «um, Fig. 4), with the anterior margin becoming more prominent than the posterior margin over time (at least by about 600 um in length, Fig. 4). A change in shell shape occurs with growth: larger or adult clams possess a rounder anterior margin and more pointed posterior margin. The adductor muscle scars on the internal surface of most shells in figure 4 reflect the difference in shape be- tween the more rounded posterior and more elongate anterior adductor muscles (see also Britton and Morton, 1982). Internally, no denticles appeared in the provinculum (hinge) of newly released pediveligers (Fig. 5). Thereafter, a posterior ligament pit was first apparent at 257 um, becom- ing prominent at 411 um (Fig. 5). The appearance of the liga- ment pit after the dissoconch began to develop supports Lutz and Hidu’s (1979) hypothesis that development of this pit is one of the earliest changes to occur in shell morphology dur- ing metamorphosis in bivalves. Two lateral teeth and three cardinal teeth (anterior, central, posterior) were fully formed in each valve at lengths of 2 mm (Figs. 4, 5, 6). Small numbers of shell tubules (Tan Tiu and Prezant, 1989) can be seen in the internal shell surfaces of all but the smallest shells (229, 257 ym) in figure 4 (see also Figs. 5, 6). This complements Tan Tiu and Prezant’s (1989) report on tubules in shells of 7 to 38 mm long clams. Such tubules could develop only after metamorphosis in Corbicula fluminea, Fig. 2. Scanning electron micrograph of the exterior shell surface of the right valve of a 685 nm specimen of Corbicula fluminea, show- ing strong commarginal and weaker radial sculpture of the dissoconch. Fig. 1. Scanning electron micrograph of the exterior shell surface of the left valve of a 233 um specimen of Corbicula fluminea. PP marks the prodissoconch | - prodissoconch II boundary. Fig. 3. Higher magnification of the specimen in figure 2 in the areas of transition from prodissoconch | to prodissoconch II (PP), and from prodissoconch II to dissoconch (PD). KENNEDY ET AL.: YOUNG CORBICULA HINGE DEVELOPMENT 109 LEFT VALVE 229 686 RIGHT VALVE 257 Fig. 4. Scanning electron micrographs of interior surfaces of disarticulated shell valves of young specimens of Corbicula fluminea. Numbers are shell length (um). Specimens represent the Nassawango Creek, Maryland, population, except for the 2254 ym and 2548 um animals, which were collected at Whites Ferry, Maryland. 110 AMER. MALAC. BULL. 8(2) (1991) because they were found only in shells larger than those released by the adults. Tan Tiu and Prezant (1989) found tubules in their specimens to occur in a concentric band whose proximal border was about 700 um from the hinge. We found tubules to occur closer to the hinge in our specimens; small numbers of tubules can be seen in the umbo region of some shells in figure 4 (left valve: 613 um, 686 um, 1157 um; right valve: 671 um, 711 wm, 1066 um; see also the close-up views of some of these shells in figure 5). Tubules were fewer in number mm-2 in our specimens than reported by Tan Tiu and Prezant (1989). Our specimens are much smaller and presumably younger than those of Tan Tiu and Prezant (1989) so the few tubules that appear in the umbo region after release of young by adults may ultimately become covered as the shell enlarges and thickens with age, with the main proliferation of tubules oc- curring in older, larger individuals. LEFT VALVE 229 ACKNOWLEDGMENTS New Jersey Agricultural Experiment Station Publication No. D-32402-1-90, supported by New Jersey State funds, NSF Grant EAR-84-17011, and various NOAA Sea Grant awards to Rutgers University. Contribution No. 2197 of the Center for Environmental and Estuarine Studies. LITERATURE CITED Britton, J. C. and B. Morton. 1982. A dissection guide, field and laboratory manual for the introduced bivalve Corbicula fluminea. Malacological Review Supplement 3, 82 pp. Britton, J. C. and B. Morton. 1986. Polymorphism in Corbicula fluminea (Bivalvia:Corbiculoidea) from North America. Malacological Review 19:1-43. Counts, C. L. Ill. 1981. Corbicula fluminea (Muller) on the Delmarva Peninsula. Veliger 24:187-188. RIGHT VALVE 257 Fig. 5. Scanning electron micrographs of the hinge apparatus of specimens pictured in figure 4. Arrowheads point toward ligament pit. Numbers are shell length (um). KENNEDY ET AL.: YOUNG CORBICULA HINGE DEVELOPMENT 111 s ‘ a = Pose As, ESN Fig. 6. Enlarged view of the hinge apparatus of (top) the left valve of the 2254 um specimen of Corbicula fluminea and of (bottom) the right valve of the 2548 um specimen in figure 4. [CT = cardinal teeth (A = anterior cardinal tooth, P = posterior cardinal tooth); LT = lateral tooth]. 4 Counts, C. L. Ill. 1986. The zoogeography and history of the invasion of the United States by Corbicula fluminea (Bivalvia: Cor- biculidae). American Malacological Bulletin, Special Edition 2:7-39. Fuller, S.C., R. A. Lutz, and A. Pooley. 1989. Procedures for accurate documentation of shapes and dimensions of larval bivalve shells with scanning electron microscopy. Transactions of the American Microscopial Society 108:58-63. Hillis, D. M. and J. C. Patton. 1982. Morphological and electrophoretic evidence for two species of Corbicula (Bivalvia:Corbiculidae) in North America. American Midland Naturalist 108:74-80. Kennedy, V. S. and L. Van Heukelem. 1985. Gametogenesis and lar- val production in a population of the introduced Asiatic clam, Corbicula sp. (Bivalvia:Corbiculidae), in Maryland. Biological Bulletin 168:50-60. Lutz, R. A. and H. Hidu. 1979. Hinge morphogenesis in the shells of larval and early post-larval mussels (Mytilus edulis L. and Modiolus modiolus (L.)). Journal of the Marine Biological Association of the United Kingdom 59:111-121. McLeod, M. J. 1986. Electrophoretic variation in North American Cor- bicula. American Malacological Bulletin, Special Edition 2:125-132. McMahon, R. F. 1982. The occurrence and spread of the introduced Asiatic freshwater clam, Corbicula fluminea (Muller), in North America: 1924-1982. Nautilus 96:134-141. Tan Tiu, A. and R. S. Prezant. 1989. Shell tubules in Corbicula fluminea (Bivalvia: Heterodonta): functional morphology and micro- structure. Nautilus 103:36-39. Date of manuscript acceptance: 6 February 1991 iz A DISTRIBUTIONAL CHECKLIST OF THE FRESHWATER UNIONIDS (BIVALVIA:UNIONOIDEA) OF KENTUCKY RONALD R. CICERELLO KENTUCKY STATE NATURE PRESERVES COMMISSION FRANKFORT, KENTUCKY 40601, U.S.A. MELVIN L. WARREN, JR. SOUTHERN ILLINOIS UNIVERSITY AT CARBONDALE DEPARTMENT OF ZOOLOGY CARBONDALE, ILLINOIS 62901, U.S.A. GUENTER A. SCHUSTER EASTERN KENTUCKY UNIVERSITY DEPARTMENT OF BIOLOGICAL SCIENCES RICHMOND, KENTUCKY 40475, U.S.A. ABSTRACT A compilation of freshwater unionid records from Kentucky based on accepted literature reports, museum specimens, and personal collecting revealed that 103 taxa occur or occurred in the state. The fauna includes members of the Mississippian, Cumberlandian, and Atlantic Slope unionid regions. Eighteen taxa (17%) have been extirpated from the state or are extinct, and an additional 34 (33%) are rare in Kentucky or nationally. Significant new distributional information is presented, including the first Mississippi River record for each of 22 taxa and the first reported unionid bed from the lower Mississippi River. Of the 297 freshwater unionid (mussel) (Bivalvia: Unionoidea) taxa recognized from the United States and Canada (Turgeon et al., 1988), approximately 35% are known from Kentucky. Only Tennessee (Starnes and Bogan, 1988) and Alabama (J. D. Williams, pers. comm.) exceed Kentucky in unionid faunal richness. Our knowledge of the highly diverse Kentucky unionid fauna began with the taxonomic studies of C. S. Rafinesque in 1818-1819. He described many species from the Ohio River and its Kentucky tributaries (Rafinesque, 1820, 1831), but con- siderable confusion and controversy surrounds the identity and nomenclature of many of these taxa (Bogan and Parma- lee, 1983; Starnes and Bogan, 1988). Numerous studies followed that contributed significantly to our understanding of the distribution and zoogeographic relationships of the Ken- tucky fauna, including but not limited to Wilson and Clark (1914), Danglade (1922), Clench (1926), Ortmann (1926), van der Schalie (1939), Clench and van der Schalie (1944), Rosewater (1959), Neel and Allen (1964), Stansbery (1965, 1969), and Williams (1969). This historical information is com- plemented by a spate of recent (post-1969) distributional in- vestigations (Table 1). Even though a considerable body of literature exists regarding Kentucky unionids, only two compilations of distribu- tional information on the fauna have been published, each with certain limitations. Bickel’s (1967) checklist included 104 unionid taxa but was based solely on published records and contained no verified or new records. Schuster’s (1988) up- date of Bickel’s list included doubtful records, lacked much recent collection data, and did not identify the quality of col- lected specimens (e.g. live, subfossil, etc.). Moreover, neither effort assessed the distributional status of the fauna across the state. We attempt here to provide a comprehensive sum- mary of the historical and current distribution of the Kentucky unionid fauna as judged from personal collecting, museum holdings, and accepted literature records covering the period from 1818 through 1987. American Malacological Bulletin, Vol. 8(2) (1991):113-129 113 114 AMER. MALAC. BULL. 8(2) (1991) Table 1. Selected studies of Kentucky’s freshwater unionid fauna published after 1969. Drainage Source Ohio Warren and Call, 1983; Stansbery and Cooney, 1985; Miller et a/., 1986; Williams and Schuster, 1989. Tennessee Chandler, 1982; Sickel and Chandler, 1982; Warren and Call, 1983; Pharris et al., 1984; Sickel, 1985, 1987b. Blankenship and Crockett, 1972; Harker et ai., 1980; Starnes and Starnes, 1980; Clarke, 1981, 1983; Call and Parmalee, 1981; Sickel, 1982, 1987a, 1988; Starnes and Bogan, 1982; DiStefano, 1984; Miller et a/., 1984; Thompson, 1985; Ahlstedt, 1986; Schuster et a/., 1989. Curnberland Tradewater Harker et a/., 1981; Warren and Call, 1983. Green Clarke, 1981, 1983; Harker et a/., 1981; Taylor, 1983; Warren and Call, 1983. Salt Taylor, 1980a. Kentucky Williams, 1975; Houp, 1980; Taylor, 1981, 1984; Tolin and King, 1986. Licking Taylor and Spurlock, 1983. Kinniconick Warren et al., 1984. Tygarts Zeto, 1979, 1980; Taylor, 1980b. Big Sandy Tolin and Schettig, 1984. METHODS Spot distribution maps were prepared for each taxon known from the state and were used to generate the distribu- tional information reported here. The maps include locality information from accepted literature records, collections made by Kentucky State Nature Preserves Commission (KNP) per- sonnel, and examination of specimens in the following institu- tional or personal collections (acronym in parentheses): Academy of Natural Sciences of Philadelphia (ANSP), Carnegie Museum (CM), Eastern Kentucky University (EKU), Field Museum of Natural History (FMNH), H. D. Athearn (HDA), Illinois Natural History Survey (INHS), Museum of Comparative Zoology (MCZ), Murray State University Museum of Zoology (MSUMZ), Marshall University (MU), National Museum of Natural History (NMNH), Ohio State University Museum of Zoology (OSUMZ), University of Illinois Museum of Natural History (UIMNH), and University of Michigan Museum of Zoology (UMMZ). Specimens collected by KNP are or will be deposited at OSUMZ. Each locality for a taxon was mapped as either a recent, historial, or archaeological record. Recent records include post-1969 collections of live or fresh-dead specimens; historical records include specimens collected prior to 1970; and archaeological records include shells taken during archaeological excavations. Specimens collected post-1969 that were weathered-dry and those reported in recent literature were mapped as historial records, unless the author(s) indicated or it was apparent that the specimens were collected alive or fresh-dead. Annotations are presented alphabetically by genus and species within genus. Scientific and common names follow Turgeon et al. (1988). Cumberlandian endemics identified by Ortmann (1924), Starnes and Bogan (1988), or Gordon and Layzer (1989) are indicated. Each annotation includes: 1) let- ters denoting the drainage(s) (Fig. 1) from which the taxon is known based on recent (Fig. 2) and historical and archae- ological (Fig. 3) records; and 2) a brief statement regarding the known recent (post-1969) distribution. We use the terms “generally distributed’, ‘‘occasional’’, and ‘‘sporadic’’ follow- ing the definitions of Smith (1965) (Table 2). For taxa that are rare in Kentucky (Warren et a/., 1986; United States Fish and Wildlife Service, 1989, 1990) or those known from only one or two recent collections in a drainage, we present the stream or drainage name, county of occurrence (Fig. 4), and source of the record(s). For collections made from the Ohio River in bordering states, we indicate the adjacent Kentucky county (Fig. 4). As appropriate, we include a statement indicating whether the taxon is believed to be extirpated in Kentucky or extinct. Finally, the current state conservation status as defined by the Kentucky Academy of Science-Kentucky State Nature Preserves Commission (KAS-KNP) (Warren et al., 1986) is presented as E (= endangered), T (= threatened), or S (= special concern). Following the KAS-KNP status and a slash (/), the federal conservation status as defined by the United States Fish and Wildlife Service (USFWS) (1989, 1990) is abbreviated as E (= endangered), 2 (= candidate for listing), 3A (= persuasive evidence of extinction), 3B (= ques- tionable specific status), and 3C (= more abundant or wide- ly distributed than previously thought). RESULTS The following annotations document the occurrence and distribution of 103 unionid species in Kentucky (Table 3). Actinonaias ligamentina (Lamarck, 1819). Mucket. B, E, F, G, H, J, K, L, N, O, Q, S. Generally distributed in the upper Green, Licking, and Big Sandy rivers. Occasional in the Ohio, upper Cumberland (below the Falls), and Kentucky rivers. Sporadic in the lower Green River. Rare in the Tennessee (below Kentucky Dam), Livingston/Marshall counties (Sickel, 1985), and Salt, Nelson County (KNP), rivers. Actinonaias pectorosa (Conrad, 1834). Pheasantshell. Cumberlandian species. F, L, M. Generally distributed in the upper Cumberland River but sporadic above the Falls. Oc- casional in the Red River and Whippoorwill Creek of the lower Cumberland River. Alasmidonta atropurpurea (Rafinesque, 1831). Cumberland elktoe. Cumberlandian species. L, M. Occasional Table 2. Terminology and definitions regarding the current distribu- tion for a particular unionid taxon in Kentucky (from Smith, 1965). Generally distributed - implies that any suitable habitat within the prescribed area should be expected to yield specimens with a reasonably thorough search. Occasional - implies that suitable-appearing habitat may or may not yield specimens even after prolonged search. Sporadic - implies that encountering specimens of a given taxon can- not be predicted at all. CICERELLO ET AL.: KENTUCKY UNIONIDS 115 0 20 40 60 Kilometers ee Fig. 1. Major river drainages in Kentucky: A, Mississippi River mainstream; B, Ohio River mainstream and minor tributaries; C, Mississippi River tributaries; D, Terrapin Creek and other small tributaries of Obion River; E, lower Tennessee and Clarks rivers; F, lower Cumberland River; G, Tradewater River; H, lower Green River; J, upper Green and Barren rivers; K, Salt River; L, upper Cumberland River below Cumberland Falls; M, upper Cumberland River above Cumberland Falls; N, Kentucky River; O, Licking River; P, Kinniconick and Salt Lick creeks; Q, Tygarts Creek; R, Little Sandy River; S, Big Sandy River. 0 20 40 60 Kilometers Fig. 2. Locations of unionid collections made in Kentucky after 1969 (each dot represents one or more collections). and locally common in Marsh Creek (Call and Parmalee, 1981; KNP) and rare in Rock Creek (OSUMZ) (both McCreary County). The report from one locality in Horse Lick Creek, Jackson County (Ahlstedt, 1986), could represent a mis- identification. E/2. Alasmidonta marginata Say, 1818. Elktoe. B, F, J, K, L, M, N, O, R. Sporadic and rare in the upper Cumberland (below the Falls); Kentucky, where it is locally common; and Licking rivers. Known from a single record in the upper Green River, Hart County (KNP). TH. Alasmidonta viridis (Rafinesque, 1820). Slippershell mussel. B, F, H, J, K, L, M, N, O. Generally distributed in the Salt, Kentucky, and Licking rivers. Occasional and locally com- mon in the upper Cumberland River (below the Falls) and sporadic in the lower Cumberland River. Rare in Marsh Creek, McCreary County (Call and Parmalee, 1981; KNP). Amblema plicata plicata (Say, 1817). Threeridge. A, B, E, F, G, H, J, K, L, N, O, P, Q, R, S. Generally distributed in the Ohio, Tennessee, lower Cumberland, upper Green, up- per Cumberland (below the Falls), Kentucky, and Licking rivers. Occasional to sporadic in the lower Green, Salt, and Big Sandy rivers and Kinniconick and Tygarts creeks. Rare in the Mississippi River, Carlisle County (KNP), and Little Sandy River, Carter County (KNP). Anodonta grandis Say, 1829. Giant floater. A, B, C, D, E, F, G, H, J, K, L, N, O, P, Q, R, S. Generally distributed in the Tennessee, Salt, Kentucky and Licking rivers but avoids the headwaters of the last two drainages. Occasional in 116 AMER. MALAC. BULL. 8(2) (1991) Kilometers Fig. 3. Locations of unionid collections made in Kentucky before 1970. Dots represent historical collections (each dot represents one or more collections); triangles represent archaeological collections. Fig. 4. The counties of Kentucky. Mississippi River tributaries, and the Ohio, lower Cumberland, and Big Sandy rivers. Sporadic in the Mississippi, Tradewater, and Green rivers, and from Kinniconick Creek east to the Lit- tle Sandy River. Rare in the upper Cumberland River (below the Falls), McCreary/Wayne counties (Starnes and Bogan, 1982). Anodonta imbecillis Say, 1829. Paper pondshell. A, B, C, E, F, G, H, J, K, L, N, O. Sporadic in the Ohio, Tennessee, Tradewater, Salt, upper Cumberland (below the Falls), and Kentucky rivers. Known from one or two records each from the Mississippi River, Carlisle County (KNP); Mayfield Creek, Carlisle County (KNP); lower Cumberland River, Trigg County (MSUMZ); lower Green River, Daviess and Muhlenberg coun- ties (KNP); upper Green River, Barren (KNP) and Casey (W. R. Haag, pers. comm.) counties; and Licking River, Menifee Kilometers County (KNP). Anodonta suborbiculata Say, 1831. Flat floater. B, C, E, F, H, J. Sporadic in Mississippi River tributaries and in im- pounded reaches of the Tennessee and lower Cumberland rivers. Rare in the Ohio River and minor tributaries, Ballard (KNP) and Campbell/Pendleton (Stansbery and Cooney, 1985) counties; lower Green River, Henderson County (KNP); and upper Green River, Edmonson County (KNP). Anodontoides ferussacianus (Lea, 1834). Cylindrical papershell. B, F, G, H, J, K, L, M, N, O, Q. Sporadic and un- common in the Salt and Licking rivers. Known from one or two records each in the lower Cumberland River, Christian County (KNP); Tradewater River, Christian County (KNP); lower Green River, Christian and Grayson counties (KNP); up- per Cumberland River (above the Falls), McCreary County CICERELLO ET AL.: KENTUCKY UNIONIDS 117 (KNP); and Kentucky River, Owen County (MU). Arcidens confragosus (Say, 1829). Rock-pocketbook. A, B,C, E, F, G, H, J. Occasional to sporadic in the lower Ohio, Tennessee, and lower Cumberland rivers. Rare in the Mississippi River, Carlisle and Hickman counties (KNP); Obion Creek, Hickman County (KNP); Tradewater River, Crittenden County (KNP); lower Green River, Butler and Ohio counties (KNP); and upper Green River, Warren County (KNP; OSUMZ). Cumberlandia monodonta (Say, 1829). Spectaclecase. B, J, L. Formerly occurred in the Ohio (Call, 1900; MCZ; OSUMZ), upper Green (Stansbery, 1965), and upper Cumberland (below the Falls) (Neel and Allen, 1964) rivers. Known to persist in the upper Green River in Mammoth Cave National Park, Edmonson County (KNP), and in the upper Cumberland River based on one live specimen found below Wolf Creek Dam (Miller et a/., 1984). E/2. Cyclonaias tuberculata (Rafinesque, 1820). Purple wartyback. B, E, F, H, J, K, L, N, O, S. Generally distributed to occasional in the Ohio, upper Green, upper Cumberland (below the Falls), and Licking rivers. Occasional to sporadic in the Tennessee (below Kentucky Dam) and lower Cumberland rivers. Cyprogenia stegaria (Rafinesque, 1820). Fanshell. B, E, F, H, J, K, L, N, O, Q, S. Occasional in the upper Green and lower Licking rivers. Rare in the Tennessee River (below Kentucky Dam), Livingston/Marshall counties (D. C. Wade, pers. comm.) and Tygarts Creek, Carter County (OSUMZ). T/E. Dromus dromas (Lea, 1834). Dromedary pearlymussel. Cumberlandian species. E, F, L. Formerly occurred in the Ten- nessee and lower Cumberland rivers as judged from archaeo- logical records (Casey, 1986) and in the upper Cumberland River (below the Falls) (Wilson and Clark, 1914; Neel and Allen, 1964). Considered extirpated. E/E. Ellipsaria lineolata (Rafinesque, 1820). Butterfly. B, E, F, H, J, L, N, O. Generally distributed in the Ohio River and the Tennessee and lower Cumberland rivers (below the dams). Sporadic in the upper Green, where it is locally common, and Kentucky rivers. Known from one or two records each in the lower Green River, Butler County (KNP), and Licking River, Pendleton County (KNP). Elliptio crassidens (Lamarck, 1819). Elephant-ear. A, B, E, F, G, H, J, K, L, N, O, P, Q, R, S. Generally distributed in the Ohio River. Occasional in the Tennessee and lower Cumberland rivers but avoids Kentucky and Barkley reser- voirs. Sporadic in the upper Green, upper Cumberland (below the Falls), and Big Sandy rivers. Rare in the lower Green River, Butler and Ohio counties (KNP); Kinniconick Creek, Lewis County (Warren et a/., 1984); Tygarts Creek, Greenup County (C. R. Burchett, Jr., pers. comm.); and Little Sandy River, Carter County (KNP). Elliptio dilatata (Rafinesque, 1820). Spike. B, E, F, H, J, K, L, M, N, O, P, Q, R. Generally distributed in the upper Green, upper Cumberland (below the Falls), Kentucky, and Licking rivers, and Kinniconick Creek. Occasional in the Ohio, Tennessee, and lower Cumberland rivers but avoids Kentucky and Barkley reservoirs. Sporadic in Salt River and Tygarts Creek. Rare in the lower Green River, Ohio County (KNP); upper Cumberland River (above the Falls), McCreary/Whitley counties (KNP); and Little Sandy River, Carter County (KNP). Epioblasma arcaeformis (Lea, 1831). Sugarspoon. Cumberlandian species. E, F, L. Archaeological records are available for the Tennessee and lower Cumberland rivers in Livingston and Lyon counties (Casey, 1986) and Todd Coun- ty (OSUMZ). Also known historically from the Big South Fork Cumberland River, Pulaski County (Wilson and Clark, 1914). Presumed extinct (Stansbery, 1970, 1971). E/3A. Epioblasma biemarginata (Lea, 1857). Angled riffleshell. Cumberlandian species. L. Known historically from the up- per Cumberland River (below the Falls), Pulaski County (Johnson, 1978). Presumed extinct (Stansbery, 1971). E/3A. Epioblasma brevidens (Lea, 1831). Cumberlandian combshell. Cumberlandian species. L. Known historically from the upper Cumberland River (below the Falls). Persists in Buck Creek, Pulaski County (EKU), and the Big South Fork Cumberland River, McCreary County (EKU; KNP), where it is occasional to sporadic. E/2. Epioblasma capsaeformis (Lea, 1834). Oyster mussel. Cumberlandian species. L. Known historically from the up- per Cumberland River (below the Falls). Presently rare in Buck Creek, Pulaski County (EKU), and Big South Fork Cumber- land River, McCreary County (EKU; KNP). E/2. Epioblasma flexuosa (Rafinesque, 1820). Leafshell. B, E, F, H, J, K, N. Known historically from the lower Ohio River (Goodrich and van der Schalie, 1944; KNP), and from the Ohio River adjacent Kenton (NMNH; OSUMZ; UMMZ) and Camp- bell (OSUMZ) counties. Known from prehistoric sites or sub- fossil specimens from the Ohio River, Henderson (Parmalee, 1960), Ballard (KNP), and Livingston (MSUMZ) counties; Ten- nessee River, Livingston/Marshall counties (Casey, 1986); and Cumberland River, Livingston/Lyon (Casey, 1986) and Todd (OSUMZ) counties. Described by Rafinesque (1820) from the Green, Salt, and Kentucky rivers; Johnson (1980) regarded these records as spurious. Presumed extinct (Stansbery, 1970, 1971). Epioblasma florentina florentina (Lea, 1857). Yellow blossom. Cumberlandian species. L. Formerly occurred in the upper Cumberland River (below the Falls), Cumberland, Pulaski, and Russell counties (Wilson and Clark, 1914; Neel and Allen, 1964; Johnson, 1978). An archaeological site specimen (Todd County; OSUMZ) could represent this taxon or E. florentina walkeri. Considered extirpated. E/E. Epioblasma florentina walkeri (Wilson and H. W. Clark, 1914). Tan riffleshell. Cumberlandian species. L. Historically known from the upper Cumberland River (below the Falls), Pulaski and Russell counties (Neel and Allen, 1964; Stansbery, 1976a). See E. florentina florentina annotation. Pro- bably extirpated. E/E. Epioblasma haysiana (Lea, 1833). Acornshell. Cumber- landian species. L. Formerly known from the upper Cumber- land River (below the Falls) (Wilson and Clark, 1914; Neel and Allen, 1964; UMMZ). Considered extirpated. E/3A. Epioblasma lewisii (Walker, 1910). Forkshell. Cumber- landian species. B, L. Formerly occurred in the Ohio River, Kenton County (OSUMZ), and the upper Cumberland River (below the Falls), Pulaski and Russell counties (Walker, 1910; 118 AMER. MALAC. BULL. 8(2) (1991) Neel and Allen, 1964). Presumed extinct (Stansbery, 1970, 1971). E/3A. Epioblasma obliquata obliquata (Rafinesque, 1820). Catspaw. B, H, J, L, N. O. Known historically from the Ohio River adjacent Indiana (Call, 1900; Goodrich and van der Schalie, 1944) and Kenton County (Johnson, 1978), and at archaeological sites adjacent Henderson (Parmalee, 1960) and Lewis (OSUMZ) counties. Also formerly occurred in the lower Green River, Butler County (Clench and van der Schalie, 1944; Johnson, 1978); upper Green River, Hart (EKU) and War- ren (OSUMZ) counties; upper Cumberland River (below the Falls), Cumberland (Neel and Allen, 1964), Pulaski (Johnson, 1978), and Russell (EKU) counties; Kentucky River (Rafinesque, 1820); and Licking River (Johnson, 1978, 1980; MCZ). Persists only in the Green River (Stansbery, 1971). E/E. Epioblasma obliquata perobliqua (Conrad, 1836). White catspaw. B. This federally endangered unionid (as E. sulcata delicata) could have occurred in the lower Ohio River (Hog- garth, 1990) but is presumed extirpated. -/E. Epioblasma personata (Say, 1829). Round combshell. B. Known only from the Ohio River at Cincinnati, Kenton County (Stansbery, 1971; Johnson, 1978). Presumed extinct (Stansbery, 1970, 1971). -/3A. Epioblasma propinqua (Lea, 1857). Tennessee riffle- shell. B. Known only from the Ohio River at Cincinnati, Ken- ton County (Johnson, 1978). Presumed extinct (Stansbery, 1970, 1971). -/3A. Epioblasma sampsonii (Lea, 1861). Wabash riffleshell. B, H, N. Known historically from the Ohio River at Cincinnati, Kenton County (Johnson, 1978). Also collected from archaeo- logical sites in the Ohio River, Henderson County (Parmalee, 1960); lower Green River, McLean County (D. H. Stansbery, pers. comm.); and the Kentucky River, Woodford County (Call and Robinson, 1983; D. H. Stansbery, pers. comm.). Pre- sumed extinct (Stansbery, 1970, 1971; United States Fish and Wildlife Service, 1984). Epioblasma stewardsoni (Lea, 1852). Cumberland leaf- shell. Cumberlandian species. L. Known only from the up- per Cumberland River (below the Falls), Pulaski County (Johnson, 1978). Presumed extinct (Stansbery, 1970, 1971). E/3A. Epioblasma torulosa rangiana (Lea, 1839). Northern riffleshell. B, H, J, K, N, O. Known historically from the Ohio River, Kenton County (MCZ); upper Green River, Edmonson (Ortmann, 1926; KNP), Green (MCZ; UMMZ2), Hart (Stansbery, 1965; KNP; OSUMZ), Taylor (MCZ), and Warren (MCZ; OSUMZ; UMMZ) counties; Salt River, Nelson (KNP; OSUMZ) and Spencer (OSUMZ) counties; Kentucky River, Franklin County (KNP; OSUMZ); and Licking River, Bath (KNP; OSUMZ) and Pendleton (OSUMZ) counties. Known from archaeological sites in the lower Green, Butler County (Patch, 1976), and Kentucky, Woodford County (Call and Robinson, 1983; D. H. Stansbery, pers. comm.) rivers. As judged from recently collected fresh-dead specimens, persists only in the upper Green River (KNP; OSUMZ). E/2. Epioblasma torulosa torulosa (Rafinesque, 1820). Tubercled blossom. B, E, F, J, K, O. Present historically in the Ohio River, Kenton (MCZ; OSUMZ) and McCracken (Parma- lee, 1967) counties; lower Tennessee River (D. H. Stansberry, pers. comm.); lower Cumberland River, Livingston County (MSUMZ); upper Green River, Warren County (Johnson, 1978; KNP; MCZ); Salt River, Nelson County (KNP; OSUMZ); and Licking River, Pendleton County (OSUMZ). Prehistoric records are available for the Ohio River adjacent Henderson (Parmalee, 1960) and Lewis/Greenup (OSUMZ) counties. Also may have occurred in the Kentucky River (Rafinesque, 1820; Danglade, 1922). Possibly extirpated. E/E. Epioblasma triquetra (Rafinesque, 1820). Snuffbox. B, F, H, J, K, L, N, O, P,Q. Occasional in the middle Licking River. Sporadic in the upper Green, upper Cumberland (below the Falls), and upper Kentucky rivers and Kinniconick and Tygarts creeks. S/-. Fusconaia ebena (Lea, 1831). Ebonyshell. A, B, E, F, J, L, O. Generally distributed and common in the Ohio, lower Tennessee, and lower Cumberland rivers; rare in the Mississippi River, Hickman County (KNP). Fusconaia flava (Rafinesque, 1820). Wabash pigtoe. B, E, F, G, H, J, K, L, N, O, P, Q, R, S. Generally distributed in the Ohio and Licking rivers. Occasional to sporadic in the Tennessee, lower Cumberland, lower and upper Green, Salt, Kentucky, Little Sandy, and Big Sandy rivers and Kinniconick and Tygarts creeks. Fusconaia subrotunda (Lea, 1831). Long-solid. B, E, F, H, J, K, L, N, O, S. Sporadic and rare in the Tennessee, up- per Green, Licking, and Big Sandy rivers. T/-. Glebula rotundata (Lamarck, 1819). Round pearlshell. B. Known from a single collection from the Ohio River at Louisville, Jefferson County (OSUMZ). Hemistena lata (Rafinesque, 1820). Cracking pearlymussel. B, J, L, N. Known historically from the Ohio River, Jefferson (NMNH), Carroll (OSUMZ), and Kenton (OSUMZ) counties; the upper Green River, Edmonson (MCZ) and Hart (Clench, 1926; Ortmann, 1926; Stansbery, 1965) counties; the upper Cumberland River (below the Falls) (Wilson and Clark, 1914; Neel and Allen, 1964); and the Ken- tucky River (Rafinesque, 1820). Could persist in the upper Green River. E/E. Lampsilis abrupta (Say, 1831.) Pink mucket. B, E, F, J, K, L. Sporadic and rare in the lower Ohio River (United States Fish and Wildlife Service, 1985); Tennessee River (below Ken- tucky Dam), Livingston/Marshall counties (Sickel, 1985, 1987b; United States Fish and Wildlife Service, 1985); and upper Green River, Edmonson, Hart, and Warren counties (KNP; OSUMZ). E/E. Lampsilis cardium (Rafinesque, 1820). Plain pocket- book. B, E, F, G, H, J, K, L, M, N, O, P, Q, R. S. Generally distributed in the upper Cumberland (below the Falls), Ken- tucky, and Licking rivers and Kinniconick Creek. Occasional in the upper Green River, Tygarts Creek, and Big Sandy River. Sporadic in the Ohio, lower Green, Salt, and upper Cumber- land (above the Falls) rivers. Known from one or two collec- tions each in the Tennessee River, McCracken County (KNP); lower Cumberland River, Livingston and Logan counties (Sickel, 1987a; KNP); and Little Sandy River, Carter County (KNP). Lampsilis fasciola Rafinesque, 1820. Wavy-rayed CICERELLO ET AL.: KENTUCKY UNIONIDS 119 lampmussel. B, F, J, K, L, M, N, O, P,Q. Generally distributed in the upper Cumberland River (below the Falls). Occasional in the Kentucky River; sporadic in the lower Cumberland, up- per Green, upper Cumberland (above the Falls), and Licking rivers. Rare in Kinniconick Creek, Lewis County (Warren et al., 1984). Lampsilis ovata (Say, 1817). Pocketbook. B, E, F, H, J, K, L, M. Occasional and locally common in the upper Green River. Rare in the lower Ohio River, Ballard County (Miller et al., 1986), and the Tennessee River, Livingston/Marshall coun- ties (Sickel, 1985, 1987b). Wilson and Clark (1914) were told that L. ovata was transplanted into the upper Cumberland River above the Falls. E/-. Lampsilis siliquoidea (Barnes, 1823). Fatmucket. B, C, E, F, G, H, J, K, N, O, P, Q, R, S. Generally distributed from the upper Green River east to the Little Sandy River. Occa- sional in the lower Green and Big Sandy rivers. Rare in the Ohio River, Campbell County (W. R. Haag, pers. comm.); Obion Creek, Hickman County (KNP); and lower Cumberland River, Christian and Trigg counties (KNP). Lampsilis teres (Rafinesque, 1820). Yellow sandshell. A, B, C, E, F, G, H, J, K, N, O. Sporadic in the Ohio, Tennessee, upper Green, and Kentucky rivers. Rare in the Mississippi River and tributaries, Carlisle (KNP) and Hickman counties (KNP; NMNH); lower Green River, Butler/‘Muhlenberg and Ohio counties (KNP); Salt River, Spencer County (KNP); and Licking River, Bath (KNP) and Pendleton (OSUMZ) counties. Lasmigona complanata complanata (Barnes, 1823). White heelsplitter. A, B, C, E, F, G, H, J, K, N, O, R, S. Occa- sional in the Salt and Licking rivers; sporadic in the Ten- nessee, lower Cumberland, lower Green, and Kentucky rivers. Rare in the Mississippi River, Hickman County (KNP); Obion Creek, Hickman County (KNP); Ohio River, McCracken Coun- ty (EKU); Tradewater River, Crittenden County (KNP); upper Green River, Warren County (MSUMZ; OSUMZ); and Big Sandy River, Lawrence County (Tolin and Schettig, 1984). Lasmigona compressa (Lea, 1829). Creek heelsplitter. B, Q, R. Rare in Tygarts Creek, Carter and Greenup counties (Zeto, 1979; Taylor, 1980b), and the Little Sandy River, Carter County (OSUMZ). T/-. Lasmigona costata (Rafinesque, 1820). Fluted-shell. B, F, G, H, J, K, L, N, O, P, Q, R, S. Generally distributed in the upper Cumberland (below the Falls) and Licking rivers. Oc- casional in the upper Green, Salt, Kentucky, and Big Sandy rivers; sporadic in Tygarts Creek. Known from one or two localities each in the Ohio River, Campbell/Pendleton coun- ties (Stansbery and Cooney, 1985); lower Cumberland River, Logan County (KNP); Kinniconick Creek, Lewis County (War- ren et al., 1984); and Little Sandy River, Carter County (KNP; MU). Lasmigona subvirids (Conrad, 1835). Green floater. Q. Known only from a single collection from Tygarts Creek, Greenup County (Zeto, 1979, 1980). T/-. Leptodea fragilis (Rafinesque, 1820). Fragile papershell. A, B, C, E, F, G, H, J, K, L, N, O, PR, Q, R, S. Generally distributed to occasional in the Ohio, Tennessee, upper Cumberland (below the Falls), Licking, and Big Sandy rivers. Sporadic in the Mississippi River and tributaries, lower Cumberland, upper Green, Salt, Kentucky, and Little Sandy rivers and Tygarts Creek. Rare in the Tradewater River, Crit- tenden/Webster counties (KNP); lower Green River, Butler and Ohio counties (KNP); and Kinniconick Creek, Lewis County (Warren et al., 1984). Leptodea leptodon (Rafinesque, 1820). Scaleshell. B, J, L, N. Formerly known from the Ohio River adjacent Indiana (Goodrich and van der Schalie, 1944), near Cincinnati, Ken- ton County (Stansbery, 1970), and at Constance, Boone Coun- ty (CM). Also occurred in the lower Ohio (Rafinesque, 1820) and Kentucky (Vanatta, 1915) rivers. Last collected from the upper Green River, Hart County (Stansbery, 1965), and the upper Cumberland River (below the Falls), Cumberland and Russell counties (Wilson and Clark, 1914; Neel and Allen, 1964). Possibly extirpated. E/2. Lexingtonia dolabelloides (Lea, 1840). Slabside pearlymussel. Cumberlandian species. F. Discovered recently in the Red River, Logan County (KNP; MSUMZ). Possibly ex- tirpated. -/2. Ligumia recta (Lamarck, 1819). Black sandshell. A, B, E, F, H, J, K, L, N, O, Q, S. Occasional in the upper Green and upper Cumberland (below the Falls) rivers; sporadic in the Ohio and Tennessee rivers. Rare in the Mississippi River, Hickman County (KNP), lower Cumberland River, Livingston/ Lyon counties (Sickel, 1982; MSUMZ); lower Green River, Butler County (KNP); Kentucky River, Henry/Owen counties (Tolin and King, 1986); Licking River, Pendleton County (KNP); Tygarts Creek, Greenup County (C. R. Burchett, Jr., pers. comm.); and Big Sandy River, Lawrence County (Tolin and Schettig, 1984). Ligumia subrostrata (Say, 1831). Pondmussel. B, C, E, G, H. Sporadic in the Tennessee River; rare in an Ohio River floodplain lake, McCracken County (KNP); and the Tradewater River, Hopkins County (Warren and Call, 1983). Medionidus conradicus (Lea, 1834). Cumberland moc- casinshell. Cumberlandian species. F, L. Generally distributed to occasional in the upper Cumberland River (below the Falls). Known only from Red River, Logan County, in the lower Cumberland River (KNP; OSUMZ). Megalonaias nervosa (Rafinesque, 1820). Washboard. A, B, E, F, G, H, J, K, L, N, O, Q, S. Generally distributed in the Tennessee and lower Cumberland (below Barkley Dam) rivers. Occasional in the Ohio, upper Green, Kentucky, and Licking rivers. Sporadic in the lower Green and Salt rivers. Rare in the Mississippi River, Fulton and Hickman counties (KNP); Tradewater River, Crittenden County (KNP); Tygarts Creek, Greenup County (W. R. Haag, pers. comm.); and Big Sandy River, Lawrence County (Tolin and Schettig, 1984). Obliquaria reflexa Rafinesque, 1820. Threehorn warty- back. A, B, E, F, G, H, J, K, L, N, O. Generally distributed in the Ohio and Tennessee rivers. Occasional in the lower Cumberland, upper Green, and Licking rivers. Sporadic in the lower Green and Kentucky rivers. Known from single localities in the Mississippi River, Hickman County (KNP); and Salt River, Spencer County (OSUMZ). Obovaria olivaria (Rafinesque, 1820). Hickorynut. A, B, E, F,H, J, L, N. Generally distributed and common in the Ohio River. Rare in the Mississippi River, Hickman County (KNP); 120 AMER. MALAC. BULL. 8(2) (1991) Tennessee River, Livingston/Marshall counties (MSUMZ); and Kentucky River (Williams, 1975). Obovaria retusa (Lamarck, 1819). Ring pink. B, E, F, H, J, L, N. Persists only in the Tennessee River (below Ken- tucky Dam), Livingston/Marshall counties (Sickel, 1985), and upper Green River, Edmonson and Hart counties (Stansbery, 1965; KNP). E/E. Obovaria subrotunda (Rafinesque, 1820). Round hickorynut. B, E, F, H, J, K, L, N, O, Q, R, S. Occasional in the upper Cumberland (below the Falls) and Big Sandy rivers. Sporadic in the upper Green, Kentucky, and Licking rivers and Tygarts Creek. Rare in the Ohio River, Campbell/ Pendleton counties (EKU); lower Cumberland River, Logan County (KNP); and Little Sandy River, Boyd County (MU). Pegias fabula (Lea, 1838). Little-wing pearlymussel. Cumberlandian species. F, L. Rare in Horse Lick Creek, Jack- son County (DiStefano, 1984); Kennedy Creek, Wayne County (KNP); Little South Fork Cumberland River, McCreary/Wayne counties (Starnes and Starnes, 1980; Starnes and Bogan, 1982); Big South Fork Cumberland River, McCreary County (Ahistedt, 1986); and Whippoorwill Creek, Logan County (KNP). Known from one live specimen from Buck Creek, Pulaski County (Stansbery, 1976b; OSUMZ). E/E. Plectomerus dombeyanus (Valenciennes, 1827). Bankclimber. B, E. Known only from the Tennessee River (Ken- tucky Lake), Trigg County (Chandler, 1982; Pharris et al., 1984). Plethobasus cicatricosus (Say, 1829). White wartyback. B, F. Formerly known from the Ohio River, Kenton (NMNH; UMMZ) and Carroll (OSUMZ) counties. Also Known from archaeological sites along the Ohio River in Henderson (Parmalee, 1960), Livingston (Casey, 1986), and Oldham (KNP) counties; and in the lower Cumberland River in Livingston/Lyon (Casey, 1986) and Todd (OSUMZ) counties. Possibly extirpated. E/E. Plethobasus cooperianus (Lea, 1834). Orange-foot pimpleback. B, E, F, H, J, K, L. Rare and only known to per- sist in the Ohio River, Ballard County (Miller et a/., 1986; Williams and Schuster, 1989), and lower Tennessee River, Liv- ingston/Marshall counties (Sickel, 1985; MSUMZ). E/E. Plethobasus cyphyus (Rafinesque, 1820). Sheepnose. B, E, F, H, J, L, O. Generally distributed and common in the Ohio River. Occasional in the upper Green River; sporadic in the Tennessee (below Kentucky Dam) and Licking rivers. S/-. Pleurobema clava (Lamarck, 1819). Clubshell. B, E, F, J, K, N, O. As judged from recently collected fresh-dead specimens, probably persists in the upper Green River (KNP; OSUMZ). E/2. Pleurobema coccineum (Conrad, 1834). Round pigtoe. B, E, F, H, J, K, L, N, O, Q. Occasional in the upper Green and upper Cumberland (below the Falls) rivers. Sporadic in the Ohio, Kentucky, and Licking rivers. Rare in the lower Ten- nessee River, Livingston/Marshall counties (MSUMZ). Pleurobema cordatum (Rafinesque, 1820). Ohio pigtoe. B, E, F, H, J, K, L, N, O, Q. Generally distributed in the Ohio River. Occasional in the Tennessee and lower Cumberland (below the dams), upper Green, and Kentucky rivers. Known from one record each in the lower Green River, Butler County (KNP), and Tygarts Creek, Greenup County (W. R. Haag, pers. comm.). Pleurobema oviforme (Conrad, 1834). Tennessee club- shell. Cumberlandian species. F, L. Sporadic and rare in the lower and upper Cumberland River (below the Falls). Persists in Big South Fork Cumberland River, McCreary County (KNP); Buck Creek, Pulaski County (Schuster et a/., 1989); Horse Lick Creek, Jackson/Rockcastle counties (KNP); Kennedy Creek, Wayne County (Ahlstedt, 1986); Little South Fork Cumberland River, McCreary/Wayne counties (Starnes and Bogan, 1982; Ahistedt, 1986); and Whippoorwill Creek, Logan County (KNP). E/2. Pleurobema plenum (Lea, 1840). Rough pigtoe. B, H, J, L, N, O. Sporadic and rare in the upper Green River, Butler, Edmonson, Hart, and Warren counties (KNP, OSUMZ). E/E. Pleurobema pyramidatum (Lea, 1840). Pyramid pigtoe. B, E, F, H, J, K, L, N, O. Sporadic in the upper Green River, Butler, Edmonson, Hart, and Warren counties (KNP); rare in the Tennessee River, Livingston/Marshall counties (MSUMZ). E/3B. Potamilus alatus (Say, 1817). Pink heelsplitter. A, B, C, E, F, G, H, J, K, L, N, O, P, Q, R, S. Generally distributed to occasional throughout the state but sporadic in the Mississippi River. Known from single collections in Obion Creek, Hickman County (KNP), and Tradwater River, Crittenden County (KNP). Potamilus capax (Green, 1832). Fat pocketbook. B, F, H. Sporadic and rare in the Ohio River from the Wabash River, Union County, downstream to Ballard County (KNP), and the extreme lower Cumberland River, Livingston County (Sickel, 1987a). E/E. Potamilus ohiensis (Rafinesque, 1820). Pink papershell. A, B, C, E, F, H, J, K, L, N, O, S. Generally distributed to oc- casional in the Mississippi and lower Ohio rivers upstream to the Wabash River. Sporadic in the Tennessee and lower Cumberland rivers. Known from one or two collections each in Obion Creek, Hickman County (KNP); lower Green River, Butler County (KNP); upper Cumberland River (below the Falls), Laurel (KNP) and Pulaski (EKU) counties; Kentucky River, Franklin and Henry/Owen counties (Tolin and King, 1986); and Big Sandy River, Lawrence County (Tolin and Schettig, 1984). Potamilus purpuratus (Lamarck, 1819). Bleufer. A, B. Known only from two specimens from the Mississippi River, Fulton County (KNP), and one specimen from an Ohio River oxbow lake, Ballard County (OSUMZ). Ptychobranchus fasciolaris (Rafinesque, 1820). Kidneyshell. B, E, F, H, J, K, L, N, O, P, Q, R, S. Generally distributed to occasional from the upper Green River east to Tygarts Creek. Occasional in the lower Cumberland River. Rare in the Ohio River, Lewis County (EKU); Tennessee River, Livingston/Marshall counties (MSUMZ); Little Sandy River, Carter (KNP) and Greenup (OSUMZ) counties; and Big Sandy River, Floyd (OSUMZ) and Lawrence (Tolin and Schettig, 1984) counties. Ptychobranchus subtentum (Say, 1825). Fluted kidneyshell. Cumberlandian species. F, L. Occasional in the upper Cumberland River (below the Falls). T/-. CICERELLO ET AL.: KENTUCKY UNIONIDS 121 Quadrula apiculata (Say, 1829). Southern mapleleaf. A, B, E, H, J. Sporadic in the lower Ohio and Tennessee rivers. Known from one record each in the Mississippi River, Hickman County (KNP), and lower Green River, Butler County (KNP). Quadrula cylindrica cylindrica (Say, 1817). Rabbitsfoot. B, E, F, H, J, K, L, N, O, S. Sporadic in the lower Ohio River, Henderson (Williams and Schuster, 1989) and McCracken (INHS) counties; Tennessee River (below Kentucky Dam), Liv- ingston/Marshall counties (Sickel, 1985); lower Cumberland River, Logan County (KNP); and upper Green River, Green, Hart, and Taylor counties (KNP; OSUMZ). E/-. Quaarula fragosa (Conrad, 1835). Winged mapleleaf. B, F. Known historically from the Ohio River, Oldham (Call, 1900) and Kenton (LaRocque, 1967) counties; and the lower Cumberland River, Lyon and Trigg counties (Wilson and Clark, 1914). A recently collected fresh-dead specimen from the lower Ohio River, Ballard County (KNP; OSUMZ), could represent this or an undescribed species (D. H. Stansbery, pers. comm.). E/3C. Quacarula metanevra (Rafinesque, 1820). Monkeyface. B, E, F, H, J, L, N, O. Generally distributed to occasional in the Ohio, Tennessee and lower Cumberland (below the dams), and Licking rivers. Sporadic in the upper Green River. Known from a single record in the Kentucky River, Henry/Owen coun- ties (Tolin and King, 1986). Quaarula nodulata (Rafinesque, 1820). Wartyback. A, B, E, F, G, H, K, N, O. Generally distributed in the Ohio and Tennessee rivers. Occasional in the lower Cumberland, lower Kentucky, and Licking rivers. Rare in the Mississippi River, Carlisle and Hickman counties (KNP); Tradewater River, Crit- tenden County (KNP); lower Green River, Ohio County (KNP); and Salt River, Spencer County (KNP). Quadrula pustulosa pustulosa (Lea, 1831). Pimpleback. A, B, C, E, F, G, H, J, K, L, N, O, P, Q, R, S. Generally distributed in the Ohio, Tennessee, upper Green, upper Cumberland (below the Falls), and Licking rivers. Occasional to sporadic in the lower Cumberland (below Barkley Dam), lower Green, Salt, Kentucky, and Big Sandy rivers and Tygarts Creek. Known only from one or two collections each in the Mississippi River, Hickman County (KNP); Obion Creek, Hickman County (KNP); Kinniconick Creek, Lewis County (Warren et al., 1984); and Little Sandy River, Carter County (KNP). Quadrula quadrula (Rafinesque, 1820). Mapleleaf. A, B, C, E, F, G, H, J, K, L, N, O, Q, R, S. Generally distributed in the Ohio, Tennessee, and lower Cumberland rivers. Occa- sional to sporadic in the Mississippi, lower Green, upper Green, Salt, Kentucky, Licking, Little Sandy, and Big Sandy rivers. Rare in Obion Creek, Hickman County (KNP), and Tradewater River, Crittenden County (KNP). Quaadrula sparsa (Lea, 1841). Appalachian monkeyface. Cumberlandian species. L. Ortmann (1912) examined two specimens from the upper Cumberland River (below the Falls), Cumberland and Pulaski counties; these records are included provisionally, but the specimens apparently are not extant. Neither Starnes and Bogan (1988) nor Gordon and Layzer (1989) recognized either sparsa or intermedia as occurring in the Cumberland River, and the specimens reported by Ortmann (1912) could have been based on tuberosa. E/E. Quaarula tuberosa (Lea, 1840). Rough rockshell. Cumberlandian species. L. Formerly occurred in the upper Cumberland River (below the Falls) (Wilson and Clark, 1914; UIMNH; UMMZ). Presumed extinct (Ahlstedt, 1983). Simpsonaias ambigua (Say, 1825). Salamander mussel. B, J, K, N, O, P, Q, R. Occasional to sporadic in the Kentucky, Licking, and Little Sandy rivers and Tygarts Creek. Known only from two records in Kinniconick Creek, Lewis County (Warren et a/., 1984). T/2. Strophitus undulatus (Say, 1817). Squawfoot. B, F, H, J, K, L, M, N, O, P, Q, R, S. Occasional to sporadic from the upper Green River east to the Big Sandy River; unknown from recent collections in the upper Cumberland River (above the Falls). Toxolasma lividus (Rafinesque, 1831). Purple lilliput. B, F, H, J, K, L. Sporadic and rare in the upper Cumberland River (below the Falls). Known from a single recent collection in the lower Cumberland River, Todd County (OSUMZ). E/2. Toxolasma parvus (Barnes, 1823). Lilliput. B, C, E, H, J, K, L, M, N, O, R. Sporadic in the Ohio, lower Green, and Salt rivers. Known from one or two records each in Mississippi River tributaries, Fulton County (KNP); Tennessee River (Ken- tucky Reservoir), Calloway/Trigg and Livingston/Marshall counties (Chandler, 1982); and Little Sandy River, Greenup County (MSUMZ). Toxolasma texasensis (Lea, 1857). Texas lilliput. E. G. Known only from Cypress Creek, Marshall County (KNP); Smith Ditch, Union County (Warren and Call, 1983); and Tradewater River, Hopkins County (KNP). Tritogonia verrucosa (Rafinesque, 1820). Pistolgrip. A, B, C, E, F, G, H, J, K, L, N, O, P,Q, R, S. Generally distributed to occasional in the Ohio, Tennessee, lower Cumberland, lower and upper Green, Salt, upper Cumberland (below the Falls), Kentucky, Licking, and Big Sandy rivers and Kin- niconick Creek. Sporadic in Tygarts Creek and Little Sandy River. Known from one location each in the Mississippi River and Obion Creek, both Hickman County (KNP). Truncilla donaciformis (Lea, 1828). Fawnsfoot. B, E, F, G, H, J, K, L, N, O. Occasional in the Tennessee River; sporadic in the Ohio, upper Green, and Licking rivers. Known from one record each in the lower Cumberland River, Liv- ingston/Marshall County (MSUMZ); Salt River, Soencer Coun- ty (KNP); and Kentucky River, Henry/Owen counties (Tolin and King, 1986). Truncilla truncata Rafinesque, 1820. Deertoe. A, B, E, F, G, H, J, K, L, N, O, P, Q, S. Occasional in the Ohio, Ten- nessee (below Kentucky Dam), Salt, and Licking rivers. Sporadic in the Green and Kentucky rivers and Tygarts Creek. Rare in the Mississippi River, Hickman County (KNP); upper Cumberland River (below the Falls), Rockcastle County (KNP); and Kinniconick Creek, Lewis County (C. R. Burchett, Jr., pers. comm.). Uniomerus tetralasmus (Say, 1831). Pondhorn. B, C, E, F, G, H, K. Rare and known only from one or two collections each in the Tennessee River, McCracken County (KNP); lower 122 AMER. MALAC. BULL. 8(2) (1991) Cumberland River (Barkley Reservoir), Trigg County (MSUMZ); Tradewater River, Webster County (Warren and Call, 1983); and lower Green River, Daviess (Warren and Call, 1983) and Ohio (MSUMZ) counties. Villosa fabalis (Lea, 1831). Rayed bean. B, J, K, N, O. Known historically from the Ohio River, Kenton County (Lea, 1870); upper Green River, Grayson (KNP), Green (HDA), Hart (Stansbery, 1965; KNP), and Warren (ANSP; UMMZ) coun- ties; Salt River, Spencer County (KNP); Kentucky River, Gallatin and Scott counties (KNP); and Licking River, Pendleton County (KNP). The most recent collection of live or fresh-dead specimens is from the upper Green River (Stansbery, 1965). E/2. Villosa iris (Lea, 1829). Rainbow. B, F, H, J, K, L, N, O, P, Q. Generally distributed in the upper Cumberland River (below the Falls); occasional in Kinniconick Creek; and sporadic in the lower Cumberland, upper Green, Kentucky, and Licking rivers and Tygarts Creek. Villosa lienosa (Conrad, 1834). Little spectaclecase. B, E, F, H, J, K, L, N, O, P, Q, R, S. Occasional in Kinniconick Creek; sporadic in the upper Green, Salt, and Kentucky rivers. Known from single collections in the Tennessee River, Graves County (INHS); Licking River, Bath County (KNP); Little Sandy River, Carter County (KNP); and Big Sandy River, Lawrence County (Tolin and Schettig, 1984). S/-. Villosa ortmanni (Walker, 1925). Kentucky creekshell. H, J. Endemic to and occasional in the upper Green River, where it persists in the Barren, Green, and Nolin rivers. E/2. Villosa taeniata (Conrad, 1834). Painted creekshell. Cumberlandian species. F, L. Occasional in the upper Cumberland River (below the Falls); sporadic in the lower Cumberland River. Villosa trabalis (Conrad, 1834). Cumberland bean. Cumberlandian species. L. Occasional and rare in the up- per Cumberland River (below the Falls). Persists in Buck Creek, Pulaski County (Schuster et a/., 1989); Rockcastle River, Jackson, Laurel, Pulaski, and Rockcastle counties (Thompson, 1985; KNP; OSUMZ); and the Big and Little South Forks Cumberland River, McCreary and Wayne coun- ties (Starnes and Bogan, 1982; KNP). E/E. Villosa vanuxemensis vanuxemensis (Lea, 1838). Moun- tain creekshell. Cumberlandian species. F. Sporadic and rare in the lower Cumberland River. Specimens reported from the upper Cumberland River (below the Falls) by Neel and Allen (1964) are considered mis-identified T lividus (Stansbery, 1976c; Starnes and Bogan, 1982). T/-. DISCUSSION The 103 unionid taxa reported herein for Kentucky (ex- cluding the provisional record for Q. sparsa) compares favorably with the 104 taxa listed by Bickel (1967) and 103 species presented by Schuster (1988). Bickel’s (1967) list can be reduced to 89 taxa, including two not recognized by us (Actinonaias [=Venustaconcha] ellipsiformis (Conrad, 1836) and Pleurobema edgariana [=Fusconaia cor (Conrad, 1834))), by using Turgeon et a/. (1988) as a nomenclatural standard. Fifteen additional species reported here (Alasmidonta atropur- purea, Anodonta suborbiculata, Epioblasma biemarginata, E. propinqua, E. sampsonii, E. stewardsoni, Glebula rotundata, Lasmigona subviridis, Lexingtonia dolabelloides, Obovaria retusa, Plectomerus dombeyanus, Potamilus purpuratus, Quaodrula apiculata, Q. tuberosa, and Toxolasma texasensis) were discovered in Kentucky subsequent to Bickel (1967), mentioned in literature he did not review, or simply overlooked. The only qualitative differences between Schuster (1988), who also used the nomenclature of Turgeon et a/. (1988), and our list are that he included V. ellipsiformis as a dubious member of the fauna, and we added Q. tuberosa. Glebula rotundata is a Gulf Coastal species generally found within about 200 km of the Gulf (Gordon, 1983) and is probably not a perma- nent member of the fauna. Faunal richness varies considerably between and within drainage systems (Tables 3 and 4). For example, only a single species (Anodonta grandis) is known from Terrapin Creek (region D) whereas at least 80 are known from the Ohio River. Eighty-seven taxa are known from the Cumberland River of Kentucky but only 11 of these have been documented above Cumberland Falls, a major barrier to dispersal. Con- trary to previous reports that the mainstream Mississippi River downstream from St. Louis is either devoid of unionids or sup- ports few species because of the Missouri River silt or mud load (Bartsch, 1916; van der Schalie and van der Schalie, 1950; Oesch, 1984), 22 taxa have been found in the river adja- cent Kentucky, including 19 from one bed in Hickman County. Kentucky’s rich fauna is the result of the presence of three unionid assemblages (Ortmann, 1924, 1926; Clench and van der Schalie, 1944). The unique Cumberlandian fauna is restricted to the Cumberland and Tennessee river basins from their headwaters downstream to Clarksville, Tennessee and Muscle Shoals, Alabama, respectively (Ortmann, 1924, 1925). Twenty-one members of this fauna as defined by Ortmann (1924), excluding the problematic Quadrula sparsa, and refined by Johnson (1980) and Gordon and Layzer (1989) are known from the Cumberland River in Kentucky (regions F, L, and M). Ortmann (1924) also suggested that several other Cumberlandian unionids had invaded the Ohio River drainage but did not name them. Johnson (1980) identified this group as including Epioblasma flexuosa, E. personata, E. propinqua, E. sampsonii, Lampsilis abrupta, Plethobasus cicatricosus, and Toxolasma lividus. Two Cumberlandian species, Dromus dromas and E. arcaeformis, were recently discovered at archaeological sites in the the lower Cumberland and Ten- nessee river drainages (Casey, 1986) outside the Cumber- landian region and are probably also members of this group. The Interior Basin or Mississippian fauna is broadly distributed throughout the Mississippi and Ohio drainages (van der Schalie and van der Schalie, 1950) and includes 73 taxa in Kentucky. Johnson (1980) subdivided this area to create the Ohioan Region which in addition to Mississippian species includes several (16) of Ohioan origin (e.g., Epioblasma obli- quata, Hemistena lata, Obovaria retusa, and Villosa fabalis). He also mentined the Gulf Coastal Region, which includes the Mississippi River drainage downstream from the Ohio River exclusive of the Ozarkian Region, but did not identify CICERELLO ET AL.: KENTUCKY UNIONIDS 123 Table 3. Unionid distribution by major river drainage (A, Mississippi River mainstream; B, Ohio River mainstream and minor tributaries; C, Mississippi River tributaries; D, Terrapin Creek and other small tributaries of Obion River; E, lower Tennessee and Clarks rivers; F, lower Cumberland River; G, Tradewater River; H, lower Green River; J, upper Green and Barren rivers; K, Salt River; L, upper Cumberland River below Cumberland Falls; M, upper Cumberland River above Cumberland Falls; N, Kentucky River; O, Licking River; P, Kinniconick and Salt Lick creeks; Q, Tygarts Creek; R, Little Sandy River; S, Big Sandy River) (*= included provisionally). Taxa A B Cc D E F G H J K L M N O P Q R S) Actinonaias ligamentina X Xx X X X X X X X x X X pectorosa X x X Alasmidonta atropurpurea marginata X Xx X X viridis X X X X X 93 cm, and current velocity increases from moderate or swift to very swift. (Channel dimensions were measured at raceways.) There could also be a downstream change in substratum composition, but this has not been measured. All mussels and mussel shells were collected by hand from gravel bars, riffles, and shallow areas along deeper raceways and pools (9-10 June 1990). The collection is a “grab” sample that was not obtained using quantitative sampl- WARREN: OZARKIAN MUSSELS 133 Se <2 Fig. 2. Two views of the Eleven Point River, Oregon County, Missouri, showing (a) the relatively small channel at the upper end of Reach 1, looking downstream from below the Highway 99 bridge near Thomasville, and (b) the larger spring-fed channel at the lower end of Reach 3, looking downstream from below the Highway 19 bridge near Greer (February, 1991). ing techniques (cf. Miller and Payne, 1988). The sample is presumably representative of the relative abundance of various species, but it is not appropriate for estimating the size, density, or demographic characteristics of local mussel populations. Most of the specimens collected were dead. Some could have been residues of muskrat predation, but many ap- pear to have died when they were dislodged by flood waters and displaced outside of the river channel, where they perished after the river returned to its banks. Water levels in the river were about 23-25 cm higher than normal at the time of our survey because of unusually heavy spring rainfall, but the water levels were receding. Live individuals were collected as one sample and stored on ice. Dead individuals were collected from three separate reaches of the river (Fig. 1), cleaned, and stored dry. Reach 1, the farthest upstream, is 5.6 km long and runs from the Thomasville acess to a point below Posy Spring (0-5.6 km below Thomasville). Reach 2 is 9.4 km long and runs from the Posy Spring locality to the Cane Bluff access (5.6-15.0 km below Thomasville). Reach 3 runs 11.4 km from Cane Bluff to the Highway 19 bridge (15.0-26.4 km below Thomasville). Dead shells were classified by valve condition. ‘Fresh shells’ are specimens with the nacre and periostracum intact and well-preserved. This category includes all paired valves that were still joined at the dorsal ligament. ‘Eroded shells’ have weathered, worn, or poorly preserved nacre and periostracum. Some broken shells were included in this category, but all retain the dorsal hinge line and are at least half complete. Specimens were identified using a comparative col- lection at the Illinois State Museum, Springfield. Nomenclature follows Turgeon et a/. (1988). All specimens are being curated at the Illinois State Museum. RESULTS Our collections in the upper Eleven Point River pro- duced a total of 14 species of Unionacea (Table 1). All are represented by live or fresh-dead individuals. Nine species are upstream records for taxa previously documented by Oesch (1984) at or below the Highway 19 bridge in Missouri, including Elliptio dilatata (Rafinesque), Fusconaia ozarkensis, Pleurobema coccineum (Conrad), Lampsilis cardium 50 ¢ \ \ 40 Ne cai Bi Nake sae i°2) : \ Sagconrt . wo \ = \ N g 30 }- @ Fusconaia ae 5 ozarkensis \ Cc v Lampsilis Vv ca reeviana brevicula re) @@ Venustaconcha Poy pleasi e 20: |- A Ptychobranchus oO occidentalis ro) @ Other Species Bobet) nw eh, NS. glee 2 oesnec ea ad 10 + et _ a eae ee Cereal Reach 2 Sample Areas Reach 1 Reach 3 Fig. 3. Variation in the species composition of fresh-water mussels in three reaches of the upper Eleven Point River, Missouri. The four Ozarkian species in the collection (Fusconaia ozarkensis, Lampsilis reeviana brevicula, Venustaconcha pleasi, and Ptychobranchus oc- cidentalis) are plotted individually. All 10 of the other taxa in the col- lection are Mississippian species. 134 AMER. MALAC. BULL. 8(2) (1991) Table 1. Fresh-water mussels from the Upper Eleven Point River, Oregon County, Missouri. Dead mussels Reach 1 Reach 2 Reach 3 Live Percent Taxon! mussels Fresh Eroded Total Fresh Eroded Total Fresh Eroded Total Total Total Anodonta imbecillis 0 0 0 0 1 0 1 0 0 0 1 6 Say, 1829 Lasmigona costata 1 0 0 0 2 0 2 4 1 5 8 5.2 (Rafinesque, 1820) Strophitus undulatus 1 2 0 2 0 0 0 0 0 0 3 1.9 (Say, 1817) Elliptio dilatata 0 0 0 0 0 0 0 4 1 5 5 3.2 (Rafinesque, 1820) Fusconaia flava 0 0 0 0 2 0 2 0 0 0 2 1.3 (Rafinesque, 1820) F. ozarkensis 5 8 1 9 2 1 3 5 1 6 23 14.8 (Call, 1887) Pleurobema coccineum 1 0 0 0 0 1 1 3 2 5 7 4.5 (Conrad, 1834) Lampsilis cardium 0 2 0 2 5 2 7 1 1 2 11 7.1 (Rafinesque, 1820) L. reeviana brevicula 4 9 3 12 19 3 22 9 2 11 49 31.6 (Call, 1887) L. siliquoidea 0 0 0 0 1 1 2 1 1 2 4 2.6 (Barnes, 1823) Venustaconcha pleasi 5 15 1 16 1 0 1 0 0 0 22 14.2 (Marsh, 1891) Villosa iris 0 2 1 3 0 0 0 1 0 1 4 2.6 (Lea, 1829) V. lienosa 0 2 0 2 0 0 0 0 0 0 2 1.3 (Conrad, 1834) Ptychobranchus occidentalis 2 0 1 1 5 2 7 4 0 4 14 9.0 (Conrad, 1836) Number of individuals 19 40 7 47 38 10 48 32 9 41 155 99.9 Number of taxa 7 7 5 8 9 6 10 9 7 9 14 — 1Nomenclature follows Turgeon et a/. (1988). Taxonomic ordering is phylogenetic above the genus level (after Davis and Fuller 1981:246). (Rafinesque), L. reeviana brevicula (Call), Venustaconcha pleasi (Marsh), Villosa iris (Lea), V. lienosa (Conrad), and Ptychobranchus occidentalis (Conrad). Johnson (1980:133) lists an additional location for L. r brevicula (as V. reeviana) at Riverton. Of the live individuals in our collection, two F. ozarkensis and three V. pleasi were gravid females. Only two species documented by Oesch (1984) in the Eleven Point River are missing from our collections: Cyclonaias tuberculata (Rafinesque) and Ligumia recta (Lamarck). Oesch found C. tuberculata near the Highway 19 bridge, and recorded both species downstream near the Missouri-Arkansas border at Billmore, Missouri (Oesch 1984:117, 197). Their absence in our collections could be due to sampling error, limited zoogeographic ranges, or both. The five remaining taxa are new records for the Eleven Point River in Missouri. These species include Anodonta im- becillis Say, Lasmigona costata (Rafinesque), Strophitus un- dulatus (Say), Fusconaia flava (Rafinesque), and Lampsilis sili- quoidea (Barnes). All of these species have been documented elsewhere in the Black River Basin of southern Missouri and northern Arkansas (Gordon et al., 1984; Oesch, 1984; Utter- back, 1917), and it was not surprising to find them in the up- per Eleven Point River. Corbicula fluminea (Muller), the in- trusive Asian clam, was noted in all three reaches but was not collected. There is a significant correlation between the species counts of live and dead individuals in the collection (Pearson’s r = .73; df = 12; p < .01; see Rock, 1988:168). This test in- dicates that the two samples are proportionately representa- tive of the same population. Nevertheless, large quantitative collections from stratified sample areas would undoubtedly provide more reliable data on the species composition of ex- tant communities. Judging from the relative frequencies of individuals in the collection, the mussel community in the upper Eleven Point River is dominated by four species. All are endemic Ozarkian taxa (Gordon, 1980; Johnson, 1980), and together they comprise 69.6% of the collection. In decreasing order of abundance, these are Lampsilis r brevicula (31.6%), Fusconaia ozarkensis (14.8%), Venustaconcha pleasi (14.2%), WARREN: OZARKIAN MUSSELS 135 Table 2. Geographical variation in the species composition and diversity of fresh-water mussel com- munities along the Upper Eleven Point River in Oregon County, Missouri. Variable Reach 1 Leading dominant Venustaconcha pleasi Species (S) 8 Individuals (N) 47 Percent Ozarkian 50.0 species (%S)! Percent Ozarkian 80.9 individuals (%N)! Simpson index of species .21 dominance (D) Shannon index of species 1.69 richness (H’) Reach 2 Reach 3 Lampsilis Lampsilis reeviana reeviana brevicula brevicula 10 9 48 4) 40.0 33.3 68.7 51.2 125 13 1.73 2.02 1 Ozarkian species represented in the collection include Fusconaia ozarkensis, Lampsilis reeviana brevicula, Venustaconcha pleasi, and Ptychobranchus occidentalis (cf. Gordon 1980; Johnson 1980). and Ptychobranchus occidentalis (9.0%). The remaining species are members of the Mississippian faunal province and have extensive ranges in the Mississippi River basin (Johnson, 1980; Oesch, 1984). There is significant geographical variation in the species composition of dead shells collected from the three sample reaches (Tables 1-2; Fig. 3). The three leading dominants in Reach 1, the upstream section, are Venusta- concha pleasi (34.0%), Lampsilis r. brevicula (25.5%), and Fusconaia ozarkensis (19.1%). In Reach 2 there is a substan- tial drop in the abundance of V. pleasi (2.1%). The leading dominant in Reach 2 is L. r brevicula (45.8%), followed by Ptychobranchus occidentalis (14.6%) and L. cardium (14.6%). L. r brevicula is also the leading dominant in Reach 3, the downstream section, although its relative abundance drops to only 26.8%. The second leading dominant in Reach 3 is F. ozarkensis (14.6%), followed by L. costata, Elliptio dilatata, and P. coccineum at 12.2% each. Geographical variation in species composition is ac- companied by substantial downstream decreases in the relative abundance of Ozarkian elements and corresponding increases in Mississippian elements (Table 2; Fig. 3). Ozarkian taxa comprise 50% of the species documented in Reach 1, 40% in Reach 2, and 33% of the species in Reach 3. There is a parallel decline in the relative abundance of Ozarkian in- dividuals, although the curve is even steeper. Ozarkian species account for more than 80% of individuals in Reach 1, but only about 69% in Reach 2 and 51% in Reach 3. These trends could reflect either a downstream decrease in the density of Ozarkian species, an upstream decrease in the density of Mississippian taxa, or both. There also appears to be geographical variation in species diversity (Table 2). Although the numbers of species (S) and individuals (N) are fairly constant among reaches, there are discernible patterns in indexes of species dominance and species richness. The Simpson index (D; Simpson, 1949)—a measure of dominance concentration that correlates inversely with the evenness of species proportions—peaks in Reach 2 and is lowest in Reach 1. This pattern reflects the unusually high relative abundance of Lampsilis r. brevicula in Reach 2 and the relatively uniform proportions of various taxa in Reach 3. The Shannon index (H’; Shannon and Weaver, 1949)—a measure of species richness (see Magurran, 1988)— increases monotonically downstream from Reach 1 to Reach 3. This trend indicates that species diversity is relatively low in upstream areas dominated by Ozarkian taxa, but increases downstream where Mississippian taxa are relatively abundant. It should be cautioned that these patterns are based on rather limited data, and additional fieldwork is needed to confirm their validity. DISCUSSION AND CONCLUSIONS In a discussion of the prospects of rare and en- dangered fresh-water mussels in Arkansas, Harris and Gor- don (1987) suggest that certain sections of the Eleven Point River and several other streams may offer suitable habitat for sustaining endangered species. However, in an assessment of the remaining habitat available to Venustaconcha pleasi, Oesch (1984:153) remarks that most of the Eleven Point River offers only ‘‘marginal habitat’’ because of its many tributary springs. Oesch (1984:229) suggests elsewhere that water from large springs is too cold and nutrient-poor for the survival of most mussel species. In a similar vein, Buchanan (1984:85) states that cold springs tend to reduce the abundance and species diversity of mussels in the Current River basin, the next drainage east of the Eleven Point Basin. Despite these suggestions, our collections indicate there is a significant community of mussels in the upper reaches of the Eleven Point River, where spring water ac- counts for much of the river’s discharge. It is true that the total number of known species (14) is low in comparison with the 136 AMER. MALAC. BULL. 8(2) (1991) Spring River, Black River, and other large streams in the White River Basin (see Gordon, 1982; Gordon et a/., 1984; Utter- back, 1917). However, the Eleven Point River apparently pro- vides suitable habitat for four endemic Ozarkian species (Fusconaia ozarkensis, Lampsilis r. brevicula, Venustaconcha pleasi, Ptychobranchus occidentalis), which together comprise almost 70% of the individuals collected. Moreover, three of these species have appeared on state or national lists of rare and endangered fresh-water mussels (F. ozarkensis [Nord- strom et al., 1977]; V. pleasi [Nordstrom et al., 1977]; P. oc- cidentalis [Stansbery, 1971]). None is currently on federal or state lists of protected species (Fish and Wildlife Service, 1989; Missouri Department of Conservation, 1986), but F ozarken- sis and V. pleasi are rare/uncommon in Missouri (Dennis Figg, pers. comm., 1990). There is significant downstream variation in species composition in the study area, which could be related to the inflow of numerous springs and a considerable increase in stream discharge. The abundance of Venustaconcha pleasi in the upstream sample area (Reach 1) suggests it is primarily a headwater species that could be most common in the upper reaches of Ozark streams (Fig. 3). Spatial pat- terning in the relative abundance of V. pleas/ has not been documented elsewhere, but our observations are consistent with Gordon’s (1982) qualitative study of the White River in Arkansas, where V. pleasi is restricted to the upper reaches of the stream above Fayetteville. Lampsilis r brevicula is the most abundant species in the middle and lower sample areas (Reaches 2-3), where most of the tributary springs enter the river. This correlation is consistent with Oesch’s (1984:229) suggestion that the various subspecies of L. reeviana are unusually tolerant of streams below the outlets of major springs. Fusconaia ozarkensis has a bimodal pattern of abun- dance, with peaks in the upper and lower sample areas (Reaches 1 & 3), while Ptychobranchus occidentalis parallels the distribution of L. r brevicula with a peak in the middle sam- ple area (Reach 2). Considering the rather small samples of F. ozarkensis and P occidentalis, the significance of their com- positional patterns is questionable. Other taxa, all of which are Mississippian species, tend to increase monotonically downstream. There is a downstream increase in species diversity in the study area that correlates inversely with downstream decreases in the qualitative and apparent quantitative abun- dance of Ozarkian taxa and individuals. The Ozarkian decline could reflect increased competition with Mississippian faunal elements in the lower reaches of the river, although quan- titative collections are needed to test this possibility. The upper Eleven Point River in Oregon County, Missouri, is a potential refuge for the conservation of Ozarkian mussel communities. Further research is needed to measure the size, density, and demographic structure of mussel populations in this area, as well as to assess their habitat tolerances and reproductive systems in greater detail. The section of river discussed here is currently being protected in accordance with the National Wild and Scenic Rivers Act, and it appears to provide appropriate habitat for fostering the survival of rare Ozarkian species. ACKNOWLEDGMENTS | thank C. E. Colten, J. A. Ferguson, S. McGuire, R. B. McMillan, S. K. Santure, and K. B. Tankersley for their help with col- lections in the field. | am grateful to R. D. Oesch for confirming several species identifications. D. Figg, of the Missouri Department of Con- servation, kindly provided information on rare and endangered species in Missouri. | thank S. Sternes of the U. S. Geological Survey in Rolla, Missouri, for supplying discharge and gage-height data. | would also like to thank A. E. Bogan, J. A. Ferguson, R. B. McMillan, J. R. Purdue, R. D. Oesch, B. W. Styles, M. D. Wiant, and two anonymous reviewers for comments on the manuscript. LITERATURE CITED Bogan, A. E. and P. W. Parmalee. 1983. Tennessee’s Rare Wildlife, Volume II: The Mollusks. Tennessee Wildlife Resources Agency, Nashville. 123 pp. Bretz, J. H. 1965. Geomorphic History of the Ozarks of Missouri. Missouri Geological Survey and Water Resources, Rolla. Buchanan, A. C. 1984. Naiades of the Current River Basin, Missouri. American Malacological Bulletin 2:85. Call, R. E. 1895. A study of the Unionidae of Arkansas, with inciden- tal reference to their distribution in the Mississippi Valley. Trans- actions of the Academy of Science of St. Louis 7:1-65. Davis, G. M. and S. L. H. Fuller. 1981. Genetic relationships among recent Unionacea (Bivalvia) of North America. Malacologia 20:217-253. Fish and Wildlife Service. 1989. Endangered and threatened wildlife and plants; animal notice of review. Federal Register 54(4):553-579. Gordon, M. E. 1980. Recent Mollusca of Arkansas with annotations to systematics and zoogeography. Proceedings of the Arkan- sas Academy of Science 34:58-62. Gordon, M. E. 1982. Mollusca of the White River, Arkansas and Missouri. Southwestern Naturalist 27:347-352. Gordon, M. E., P. A. Durkee, H. M. Runke, and H. J. Zimmerman. 1984. Mussel Fauna of the Black and Spring Rivers in North- eastern Arkansas. Report to U. S. Army Corps of Engineers, Little Rock. 27 pp. Gordon, M. E., L. R. Kraemer, and A. V. Brown. 1980. Unionacea of Arkansas: historical review, checklist, and observations on distributional patterns. Bulletin of the American Malacological Union 1979:31-37. Harris, J. L. and M. E. Gordon. 1987. Distribution and status of rare and endangered mussels (Mollusca: Margaritiferidae, Unionidae) in Arkansas. Proceedings of the Arkansas Academy of Science 41:49-56. Isom, B. G. 1969. The mussel resource of the Tennessee River. Malacologia 7:397-425. Johnson, B. M. and J. K. Beadles. 1977. Fishes of the Eleven Point River within Arkansas. Proceedings of the Arkansas Academy of Science 31:58-61. Johnson, R. |. 1980. Zoogeography of North American Unionacea (Mollusca: Bivalvia) north of the maximum Pleistocene glacia- tion. Bulletin of the Museum of Comparative Zoology 149:77-189. Magurran, A. E. 1988. Ecological Diversity and Its Measurement. Princeton University Press, Princeton. 179 pp. Miller, A. C. and B. S. Payne. 1988. The need for quantitative sampl- ing to characterize size demography and density of freshwater mussel communities. American Malacological Bulletin 6:49-54. Missouri Department of Conservation. 1986. Checklist of Rare and WARREN: OZARKIAN MUSSELS 137 Endangered Species of Missouri. Missouri Department of Con- servation, Jefferson City. 20 pp. Nordstrom, G. R., W. L. Pflieger, K. C. Sadler, and Walter H. Lewis. 1977. Rare and Endangered Species of Missouri. Missouri Department of Conservation and USDA Soil Conservation Service, Jefferson City, Missouri. 129 pp. Oesch, R. D. 1984. Missouri Naiades: A Guide to the Mussels of Missouri. Missouri Department of Conservation, Jefferson City. 270 pp. Ortmann, A. E. 1917. A new type of the nayad-genus Fusconaia. Group of F. Barnesiana Lea. Nautilus 31:58-64. Pflieger, W. L. 1975. The Fishes of Missouri. Missouri Department of Conservation, Jefferson City. 343 pp. Rock, N. M. S. 1988. Numerical Geology. Springer-Verlag, Berlin. Sauer, C. O. 1920. The Geography of the Ozark Highland of Missouri. University of Chicago Press, Chicago. 245 pp. Shannon, C. E. and W. Weaver. 1949. The Mathematical Theory of Communication. University of Illinois Press, Urbana. 125 pp. Simpson, E. H. 1949. Measurement of diversity. Nature 163:688. Stansbery, D. H. 1970. Eastern freshwater mollusks (I): the Mississippi and St. Lawrence River systems. Malacologia 10:9-22. Stansbery, D. H. 1971. Rare and endangered freshwater mollusks in eastern United States. In: Proceedings of a Symposium on Rare and Endangered Mollusks (Naiads) of the U. S., S. E. Jorgensen and R. W. Sharp, eds. pp. 5-18f. Ohio State University, Columbus. Turgeon, D. D., A. E. Bogan, E. V. Coan, W. K. Emerson, W. G. Lyons, W. L. Pratt, C. F. E. Roper, A. Scheltema, F. G. Thompson, and J. D. Williams. 1988. Common and Scientific Names of Aquatic Invertebrates from the United States and Canada: Mollusks. Special Publication 16. American Fisheries Socie- ty, Bethesda, Maryland. 277 pp. Utterback, W. |. 1915-1916. The naiades of Missouri. American Midland Naturalist 4:41-53, 97-152, 181-204, 244-273, 311-327, 339-354, 387-400, 432-464. Utterback, W. I. 1917. Naiadgeography of Missouri. American Midland Naturalist 5:26-30. van der Schalie, H. and A. van der Schalie. 1950. The mussels of the Mississippi River. American Midland Naturalist 44:448-466. Vineyard, J. D. and G. L. Feder. 1974. Springs of Missouri. Missouri Geological Survey and Water Resources, Rolla, Missouri. 266 pp. Warren, R. E. 1991. Fresh-water mussels as paleoenvironmental in- dicators: a quantitative approach to assemblage analysis. /n: Beamers, Bobwhites, and Blue-Points: Tributes to the Career of Paul W. Parmalee. J. R. Purdue, W. E. Klippel, and B. W. Styles, eds. Illinois State Museum (in press). Date of manuscript acceptance: 21 January 1991 REINTRODUCTION OF THE SPINY RIVERSNAIL /O FLUVIALIS (SAY, 1825) (GASTROPODA: PLEUROCERIDAE) INTO THE NORTH FORK HOLSTON RIVER, SOUTHWEST VIRGINIA AND NORTHEAST TENNESSEE STEVEN A. AHLSTEDT WATER RESOURCES AQUATIC BIOLOGY DEPARTMENT TENNESSEE VALLEY AUTHORITY NORRIS, TENNESSEE 37828, U.S.A. ABSTRACT lo fluvialis was reintroduced in 1978 to two sites in the lower North Fork Holston River, along the Virginia and Tennessee border, and subsequently at one upstream site above Saltville, Virginia, in 1979. Reproduction by transplanted /. fluvialis was first observed in 1986 at Cloud Ford, 2.4 km downstream from the two lower transplant sites. Quantitative sampling in 1987 and 1988 resulted in population estimates of 1.75/m2, which are comparable to healthy self-sustaining populations in the Clinch and Powell rivers. Qualitative sampling in 1990, at the upper transplant site also confirmed reproduction. This successful re-introduction demonstrates that transplants can be considered a viable, long-term technique for re-establishing aquatic snails in river systems that were previously polluted but are now in recovery. The spiny riversnail, /o fluvialis (Say, 1825), is a large (50 mm length) freshwater prosobranch snail occurring in the Clinch, Powell, and Nolichucky rivers. Shells are generally thick, tapered at both ends, high-spired (Fig. 1), and vary in color from tan to dark brown or olive green. Shells have six or seven whorls (coils) and there is lateral wrinkling of the epidermal surface. Two phenotypes (smooth and spiny forms) are found in the tributaries of the upper Tennessee River system (Adams, 1915). Shell sculpture on the smooth form consists of folds or parallel ridges that are slightly nodulous, without pronounced spines, and the spiny form has from one to 12 (generally seven) well-developed spines. Historically, /o fluvialis is endemic to the Tennessee River and was reported as far downstream as Muscle Shoals, Alabama (Lewis, 1876; Hinkley, 1906), but it was considered more common in the upper Tennessee River and larger tributaries upstream from Chattanooga (Adams, 1915; Clench, 1928; Parmalee and Bogan, 1987). With impoundment of the Tennessee River system in the 1930s and 1940s, recent mollusk studies and environmental assessments in the Ten- nessee River and its tributaries indicate that /. fluvialis was extirpated from these streams. Presently, only the headwaters of the upper Clinch and Powell rivers, and the lower Nolichucky River support native populations of /. fluvialis. Industrial pollution has been a major factor in the ex- tirpation of /o fluvialis from the North Fork Holston River. From Fig. 1. Phenotypes of the spiny riversnail /o fluvialis (smooth-left 48.3 mm, spiny-right 47.1 mm). 1894 to 1972, the Olin Chemical Corporation maintained a soda ash plant on the banks of the North Fork Holston near Saltville, Virginia. Chloride wastes and other chemicals from this facility were discharged into the river and later into settl- American Malacological Bulletin, Vol. 8(2) (1991):139-142 139 140 AMER. MALAC. BULL. 8(2) (1991) =a KM 10.| RT. 23 BRIDGE MENDOTA Fig. 2. North Fork Holston River Transplant Sites, 1978 and 1979. ing lagoons (Ahlstedt, 1979). As a result of chemical discharge, aquatic organisms in the lower 128 km of the North Fork were effectively eliminated, especially freshwater mollusks (Adams, 1915; Ortmann, 1918). Adams (1915) reported collecting |. fluvialis in 1900 above the alkali plant upstream from Saltville, the type locality. This was the last reported sighting of |. fluvialis in the North Fork. The chemical plant was closed in 1972 because it could not afford to comply with water quality standards. Freshwater mussels have yet to recolonize this reach of the river (Stansbery, 1972; Stansbery and Clench, 1974; Stansbery and Stein, 1976; Ahlstedt, 1979). From 1971 until 1976, the Tennessee Valley Authority (TVA) conducted a biological sampling program to monitor the anticipated recovery of both fish and macroinvertebrate populations in the North Fork Holston River. Routine monitor- ing indicated that fish and benthic macroinvertebrates were recovering rapidly. The objective of this subsequent investiga- tion was to reintroduce and re-establish /o fluvialis into the North Fork. MATERIALS AND METHODS lo fluvialis was first reintroduced initially in July and August, 1978, at two sites in the lower North Fork Holston, and at one additional site in the upper North Fork, in October 1979. Collections of /. fluvialis were obtained from the upper Clinch River in July and August 1978, at the following sites (2) UPSTREAM STATE KM |5 ¢ ok BRISTOL UPPER SALTVILLE VIRGINIA 15 KILOMETERS and kilometer (km) locations: Kyles Ford (km 305) and Wallens Bend (km 309), Hancock County, Tennessee. A total of 716 live specimens were collected by snorkeling and handpick- ing individual snails from bedrock ledges and boulder substrata. In October 1979, an additional 557 specimens were collected from the same two sites. Both the smooth and spiny forms were collected and included various size classes (13-53 mm). Specimens were transported in insulated coolers con- taining river water to the following locations in the North Fork Holston River: 270 specimens to Click Island (km 10.1), and 446 specimens to upstream of the State Route 23 bridge (km 15), Scott County, Virginia; and 557 specimens upstream from Saltville (km 137), Smyth County, Virginia (Fig. 2). Transport time was approximately 1 hr to the lower two transplant sites on the North Fork, and 2 hr to the upper site upstream of Saltville, Virginia. At each site, specimens were removed from the coolers and placed in nylon mesh bags anchored in riffle areas to reduce stress. Shell measurements (mm) were made with vernier calipers and included: length or height, measured from the tip of the apex to the tip of the anterior canal; and width, measured from the periphery of the last whorl behind the outer lip of the aperture to the widest spine on the op- posite side of the shell. Specimens were numbered with white water-proof paint. Numbers were placed on the widest part of the shell behind the outer lip of the aperture. Small in- dividuals were marked with an ‘X’. Snails were placed by hand in creases along bedrock ledges and among boulders to AHLSTEDT: REINTRODUCTION OF /O 141 reduce passive transport in current, or to keep them from tumbling into silted areas. Because of insufficient funding and the long-term nature of monitoring the three transplant sites, every effort was made to qualitatively sample at least one site yearly, after the final transplant in 1979. Samples were col- lected by snorkeling, wading, and using a water scope dur- ing low-flow condition to determine whether the species was surviving and possibly reproducing. lo fluvialis densities were determined from 1987 data using a stratified random sampling design. A total of 43/m2 samples were taken with a 1/m? quadrat sampler along a transect in five areas. In each stratum, 4% of the area was sampled. Substratum within each of the 43 quadrats was visually searched and sampled by snorkeling. /. fluvialis col- lected were measured, counted, and returned to the substrata. Population estimates were calculated by the area-density method (Everhart et a/., 1975). RESULTS AND DISCUSSION During October 1986, while monitoring a transplant of the endangered birdwing pearly mussel, Lemiox rimosus Rafinesque, 1831 [=Conradilla caelata (Conrad, 1834)], two live specimens (one spiny and one smooth) of /o fluvialis were found in the lower North Fork Holston at Cloud Ford (km 7.7), Sullivan County, Tennessee. Neither specimen had any numerical markings or was measured (estimated to be 20-25 mm in length), and both were returned to the river. Both specimens were assumed to be the result of reproduction and downstream colonization from the original transplant of |. fluvialis at Click Island (km 10.1), since Cloud Ford (km 7.7) is located 2.4 km downstream of the snail transplant site. Time constraints and high water conditions prevented additional sampling for /. fluvialis in the North Fork in 1986. In October 1987, during quantitative sampling for L. rimosus at Cloud Ford, 43 quadrat samples yielded 68 specimens of /. fluvialis for amean density of 1.60/m2. Positive evidence of reproduc- tion was documented since the smallest specimen recovered in October 1987 measured 12.5 mm, and the smallest specimen transplanted into the North Fork measured 13.7 mm. Resampling at Cloud Ford in October 1988, resulted in 34 |. fluvialis collected from 31 quadrat samples for a mean den- sity of 1.00/m2. Sampling conditions in 1988 were poor at Cloud Ford because of abundant aquatic vegetation, and ex- tremely low water levels from a prolonged drought. Annual transplant evaluations for L. rimosus were not conducted by TVA at Cloud Ford during fall 1989, but large numbers of |. fluvialis were observed visually by the author during Oc- tober 1989, while wading at Cloud Ford. Currently, popula- tions are now estimated at 1,665 individuals or 1.75/m2 at Cloud Ford. These population estimates are comparable to spiny riversnail populations sampled by the author from a number of collecting sites on the Clinch and Powell rivers. Visits to /o fluvialis transplant sites at Click Island (km 10.1) and State Highway 23 bridge (km 15) in June 1989 pro- vide only qualitative evidence of reproduction and downstream colonization in the river, based on the extent of downstream colonization and the different sizes and numbers of individuals observed. The transplant site in the upper North Fork Holston (km 137) upstream from Saltville, Virginia had not been sampled since September 1985, when four live snails were found. All four were large adults, of which two retained par- tial markings of white epoxy paint. No evidence of reproduc- tion was observed at this site. In February 1990, reproduc- tion of /. fluvialis was documented at the upper transplant site on the North Fork. A total of 26 live specimens were collected qualitatively during 2.5 man-hours of sampling time. Specimens ranged in length from 10.3 - 50.7 mm. The smallest specimen, transplanted to this site in 1979, measured 18.4 mm. As judged by the large size of some of the specimens found, they are likely to be survivors from the original transplant. The white epoxy numbers marked on the shells at all three transplant sites have since worn away; however, lateral growth increments (growth rests) on some of the larger snails, especially at the upper transplant site are light-green (indica- tion of new growth) and more pronounced than the dark green color of the shell when it was first transplanted from the Clinch River. Based on these findings, /. fluvialis could be relatively long-lived. It is unknown whether /o fluvialis at the upper site in the North Fork is colonizing the river downstream. The first shoal, located approximately 275 m downstream, was sampled for 1 man-hour, but no /. fluvialis were found. Sampling was inadequate, because of high-flow conditions and deeper water habitat present at this shoal. At present, it is reasonable to assume that /. fluvialis will continue to reproduce and eventually colonize the upper North Fork. CONCLUSIONS Transplantation or reintroduction of species is one technique used successfully for a number of aquatic and ter- restrial species throughout the world. Biological data indicated that the North Fork Holston River was in recovery following pollution abatement. The success of the /o fluvialis transplants into the North Fork confirmed the suitability of this technique for re-establishing aquatic snails in previously impacted river systems. However, establishment of a viable reproducing population through transplants is a long-term process. In the North Fork, documented evidence of successful reproduction (juveniles) did not appear until 9 years post-introduction. Population estimates in the lower North Fork Holston (Cloud Ford) are now comparable to remnant populations in the Clinch and Powell rivers. Since the two reproducing popula- tions of /. fluvialis are separated by nearly 128 km of river, suitable sites between these two are recommended for addi- tional transplants, as well as other sites in recovering streams within its historic range. ACKNOWLEDGMENTS | wish to thank K. Henn, R. Smith, and D. Hill for assisting with the collection and transplant of /o fluvialis; and C. Saylor, G. Hickman, and R. Biggins for monitoring the transplant sites. 142 AMER. MALAC LITERATURE CITED Adams, C. C. 1915. The variations and ecological distribution of the snails of the genus /o. Memoirs National Academy Sciences 12(2):1-184. Ahlstedt, S. A. 1979. Recent mollusk transplants into the North Fork Holston River in southwestern Virginia. Bulletin of the American Malacological Union for 1979:21-23. Clench, W. J. 1928. /o fluvialis turrita Anthony. Nautilus 42(1):36. Everhart, W. H., A. W. Eipper, and W. D. Youngs. 1975. Principles of Fishery Science. Cornell University Press, Ithaca, New York. 288 pp. Hinkley, A. A. 1906. Some shells of Mississippi and Alabama. Nautilus 20(3):34-36; 20(4):40-44; 20(5):52-55. Lewis, J. L. 1876. lo and its habits. American Naturalist 10(6):321-326. Ortmann, A. E. 1918. The nayades (freshwater mussels) of the up- per Tennessee drainage. With notes on synonymy and distribu- tion. Proceedings of the American Philosophical Society 57(6):521-626. Parmalee, P. W. and A. E. Bogan. 1987. New prehistoric distribution . BULL. 8(2) (1991) records of /o fluvialis (Say, 1825) (Gastropoda:Pleuroceridae) in Tennessee with comments on form variation. Malacology Data Net, 2(1/2):42-54. Say, T. 1825. Descriptions of some new species of fresh water and land shells of the United States. Journal of the Academy of Natural Sciences of Philadelphia 5:129. Stansbery, D. H. 1972. The mollusk fauna of the North Fork Holston River at Saltville, Virginia. Bulletin of the American Malacological Union for 1971:45-46. Stansbery, D. H. and W. J. Clench. 1974. The Pleuroceridae and Unionidae of the North Fork Holston River above Saltville, Virginia. Bulletin of the American Malacological Union for 1974:33-36. Stansbery, D. H. and C. B. Stein. 1976. Changes in the distribution of lo fluvialis (Say, 1825) in the upper Tennessee River system (Mollusca:Gastropoda:Pleuroceridae). Bulletin of the American Malacological Union for 1976:28-33. Date of manuscript acceptance: 17 December 1990 RESEARCH NOTE FIRST RECORD OF NET COLLECTED OCYTHOE TUBERCULATA (CEPHALOPODA: OCTOPODA) FROM PERUVIAN WATERS FRANZ CARDOSO ; DEPARTAMENTO DE MALACOLOGIA Y CARCINOLOGIA MUSEO DE HISTORIA NATURAL UNIVERSIDAD NACIONAL MAYOR DE SAN MARCOS APARTADO 14-0434, LIMA-14, PERU ABSTRACT Ocythoe tuberculata Rafinesque, 1814, is recorded from net captured animals for the first time from Peruvian waters, thus confirming its presence in the southeastern Pacific Ocean. Three specimens were captured in oceanic waters at depths from 10 to 180 m. These records are consistent with the known epipelagic distribution of O. tuberculata. The literature on cephalopods contains numerous records of the pelagic octopod Ocythde tuberculata Rafinesque, 1814 (see review in Roper and Sweeney, 1976). Collectively, these reports suggest that O. tuberculata is distributed in warm waters, however, it has not been record- ed from the tropics or from the southern hemisphere except for captures in the eastern South Atlantic off South Africa (Voss, 1967a) and in the far eastern Indian Ocean off Australia (Roper and Sweeney, 1976; Lu and Phillips, 1985). As part of my current studies on the pelagic cephalo- pods from Peruvian waters, three net collected Ocythoe tuber- culata were located in the collections of the Instituto del Mar del Peru (IMARPE) and the Museo de Historia Natural, Universidad Nacional Mayor de San Marcos (MUSM). These specimens represent the first net captures of O. tuberculata from Peruvian waters. The measurements and indices used herein are defined by Pickford and McConnaughey (1949) and Roper and Voss (1983). Ocythoe tuberculata Rafinesque, 1814 Material examined. - 1 female, ML = 66.8 mm, R/V HUMBOLDT, Cruise 8202, 17° 02’S, 77° 42.5’W, 180 m, 988/400 mesh pelagic trawl, 24 Feb 1982, T. Dioses leg., IMARPE M46.011. - 1 female, ML = 105.0 mm, EUREKA XLVII, 14° 24'S, 77° 13’W, 30-40 m, purse seine, 27 Feb 1982, W. Elliott and M. Niquen leg., IMARPE M46.012. - 1 male, ML = 15.6 mm, R/V SNP-1, Cruise 8805-06, 16° 49.6’S, 72° 49.4’W, 10-40 m, 434/400 mesh pelagic trawl, 22 May 1988, A. Chipollini leg., MUSM 460101. The specimens, typical of the species, have arm pairs Il and II much shorter than arm pairs | and IV, ventral water pores only, web greatly reduced, and a complicated mantle- funnel locking apparatus that is unique among the octopods. The ventral surface of the mantle of the females is covered by tubercles in a reticulate pattern. The much smaller male lacks the dermal tubercles and has the right arm Ill hec- tocotylized and enrolled in a membranous sac. All three specimens are sexually inmature (Table 1). Distribution. - The three specimens were taken singly in Peruvian waters, between latitudes 14° and 18°S (Fig. 1). The two females were collected in association with subtropical surface-water masses. The male was collected in association with coastal cold-water masses (21.3°C, S = 34.987 °%/po9). All specimens were captured with non-closing nets in oceanic waters. DISCUSSION Roper and Sweeney (1976) summarized the records of Ocythde tuberculata from the Mediterranean Sea, the North and eastern South Atlantic Oceans, and the North Pacific Ocean and described the first occurrence of the species from the Indian Ocean off Australia. Lu and Phillips (1985) gave the known range of O. tuberculata in Australian waters as from New South Wales to the Great Australian Bight. Based on analyses of stomach contents, Shchetinnikov (1986) listed O. tuberculata as a food item of the giant ommastrephid squid Dosidicus gigas from off Peru. The Peruvian specimens described here represent the first net captures of O. tuber- culata from the eastern South Pacific Ocean. The known depth distribution of Ocythde tuberculata American Malacological Bulletin, Vol. 8(2) (1991):143-144 143 144 AMER. MALAC. BULL. 8(2) (1991) Table 1. Measurements (in mm) and indices of Ocythde tuberculata (a = approximate value; b = arm enrolled in sac). IMARPE M46.011 Specimen Female Mantle length 66.8 Mantle width 34.18 Head width 19.6 Arm lengths L R | 118.0 113.7 Il 71.4 71.0 Wl 69.3 61.9 IV 108.7 98.4 Hectocotylized arm length — Total length 182.2 Mantle width Index 51.08 Head width Index 29.3 Arm length Index 64.8 Mantle arm Index 56.6 Hectocotylized Index — No. gill lamellae 17 PTA. DONA MARIA 78° 76° 7a° 72° Fig. 1. Localities of specimens of Ocythde tuberculata from Peruvian waters. indicates that it is limited to near-surface waters (Voss, 1967b; Roper and Young, 1975). Capture records by Roper and Sweeney (1976) from the Sargasso and the Mediterranean seas (0 to 200 m) and my data (10 to 180 m) indicate that O. tuberculata inhabits the epipelagic zone. Ocythoe tuberculata is known from six areas: Mediter- ranean Sea; North and eastern South Atlantic Oceans; North and eastern South Pacific Oceans; southeastern Indian Ocean. It is an epipelagic and possibly cosmopolitan species in warm oceanic waters. Its occurrence in tropical waters re- main unconfirmed. IMARPE MUSM M46.012 460101 Female Male 105.0 15.6 63.8 11.5 41.4 9.8 L R L R 231.4 230.0 40.6 42.7 171.4 181.6 21.0 21.5 150.8 150.0 17.2 9.3° 199.2 215.0 43.4 43.8 _ 9.3 330.0 60.6 60.8 73.7 39.4 62.8 70.1 72.3 45.4 35.6 — 21.2 18 10 ACKNOWLEDGMENTS The author wishes to thank Dr. C. F. E. Roper, Smithsonian Institution, Dr. G. Lamas, Museo de Historia Natural, Lima, and two anonymous reviewers for offering valuable suggestions on an earlier version of the manuscript. | am also grateful to A. Chipollini, M. Niquen and T. Dioses for providing information about capture data. LITERATURE CITED Lu, C. C. and J. U. Phillips. 1985. An annotated checklist of the cephalopoda from Australia waters. Occasional Papers from the Museum of Victoria 2:21-36. Pickford, G. E. and B. H. McConnaughey. 1949. The Octopus bimaculatus problem: A study in sibling species. Bulletin of the Bingham Oceanographic Collection 12(4):1-66. Roper, C. F. E. and M. J. Sweeney. 1976. The pelagic octopod Ocythée tuberculata Rafinesque, 1814. Bulletin of the American Malacological Union for 1975:21-28. Roper, C. F. E. and G. L. Voss. 1983. Guidelines for taxonomic descrip- tions of cephalopod species. Memoirs of the National Museum of Victoria 44:49-63. Roper, C. F. E. and R. E. Young. 1975. Vertical distribution of pelagic cephalopods. Smithsonian Contributions to Zoology No. 209:1-48. Shchetinnikov, A. S. 1986. Geographical variability of food spectrum of Dosidicus gigas (Ommastrephidae). /n: Resources and perspectives of the utilization of the squids of the world Ocean. Sbornik Nauchnykh Trudov VNIRO MPX SSSR. pp. 143-153. Moskva, Izd. VNIRO. (English summary). Voss, G. L. 1967a. Some bathypelagic cephalopods from South African waters. Annals of the South African Museum 50(5):61-88. Voss, G. L. 1967b. The biology and bathymetric distribution of deep- sea cephalopods. Studies in Tropical Oceanography 5:511-535. Date of manuscript acceptance: 9 May 1990 SYSTEMATICS, ANATOMY AND EVOLUTION OF WESTERN NORTH AMERICAN LAND MOLLUSCS A SYMPOSIUM IN HONOR OF WALTER B. MILLER ORGANIZED BY F. G. HOCHBERG AND BARRY ROTH SANTA BARBARA MUSEUM OF NATURAL HISTORY AMERICAN MALACOLOGICAL UNION LOS ANGELES, CALIFORNIA 28 - 29 JUNE 1989 FAMILIAL RELATIONSHIPS AND BIOGEOGRAPHY OF THE WESTERN AMERICAN AND CARIBBEAN HELICOIDEA (MOLLUSCA: GASTROPODA: PULMONATA) WALTER B. MILLER EDNA NARANJO-GARCIA DEPARTMENT OF ECOLOGY AND EVOLUTIONARY BIOLOGY UNIVERSITY OF ARIZONA TUCSON, ARIZONA 85721, U. S. A. ABSTRACT Nordsieck’s (1987) Revision des Systems der Helicoidea, the latest published classification of the superfamily, is further revised. By showing that anatomical characters of certain xanthonychid sub- families are as distinctive as the ones now used to designate familial rank for other helicoid families, we reaffirm Schileyko’s (1978) elevating the Humboldtianinae to familial rank and we propose ;aising the Helminthoglyptinae, sensu Nordsieck, to familial rank. We support the placing of Monadenia by Nordsieck into a separate subfamily and we suggest that it could belong in the Bradybaenidae. The discrete anatomical characters of the American helicoid families and their discontinuous geographical distribution bring into question the theory of a putative continuous radiation and evolu- tion from an Asian origin. Recent geophysical data have shown that large parts of eastern Asia and western America were formed from Gondwanian terranes that migrated tectonically from the south Pacific Ocean. We suggest that most helicoid families, Asian and American, arrived at their separate destinations, passively, via these Pacifican terranes. The classification of the land snail superfamily Helicoidea, formerly called Helicacea, has been the subject of many studies and revision within the past 50 years. Pilsbry (1939) provided the first concise key which designated the distinguishing familial characters of the Helicidae, Helicellidae, Bradybaenidae, Helminthoglyptidae, Polygyridae, Camaeni- dae, and Sagdidae, all of which he included in the Helicacea. Zilch (1959-1960) removed the Polygyridae and the Sagdidae from the Helicoidea and placed them in the superfamily Polygyracea (now Polygyroidea, in accordance with Article 29a of the /nternational Code of Zoological Nomenclature). Solem (1978) further revised the Helicoidea by placing the Camaenidae in its own superfamily. The most recent revision of the Helicoidea was made by Nordsieck (1987) in which he recognizes the following families: Sphincterochilidae Zilch, 1960; Xanthonychidae Strebel and Pfeffer, 1880; Brady- baenidae Pilsbry, 1934; Hygromiidae Tryon, 1866; Helicidae Rafinesque, 1815. In considering the familial relationships within a super- family, one is naturally led to speculate on the evolution of these families from a single ancestral group and their dispersal from some point of origin to their present distribu- tion. The most generally accepted theory on the biogeography of the Helicoicdea has been that the ancestral helicoid probably arose in the Palearctic during the Mesozoic era (Pilsbry, 1894, 1939) from whence the ancestors of the Euro- pean helicoids could easily spread throughout Europe while the ancestral xanthonychids invaded the Americas via the Bering land bridge no later than early Eocene and probably earlier (Pilsbry, 1894). We have been studying the biogeography of the western North American helicoids for many years, especial- ly in northwest Mexico, and we have been puzzled by their total absence from a large part of Sonora where a continuous radiation would have been expected from a trans-Beringean southward radiation. We have also been puzzled by the total absence of helicoids from a large part of northwestern South America which would have had to be traversed in order to reach Peru and Argentina. Recent determinations in geophysics that most of eastern Asia and a large part of Northwest America were not part of the Asian continent or the American continent, American Malacologica! Bulletin, Vol. 8(2) (1991):147-153 147 148 AMER. MALAC. BULL. 8(2) (1991) respectively, during the Mesozoic, but rather were part of Gondwanian land masses located in early Mesozoic in the south-central Pacific Ocean led us to investigate the possibility that Asian and western American and Caribbean helicoids could have arrived at their present destinations via the Gond- wanian terranes from a Pacifican origin. Our concurrent studies on the familial characteristics of the American helicoids have also shown that several of Nordsieck’s xan- thonychid subfamilies are sufficiently different, anatomically, to warrant familial status. Together, the pronounced anatomical differences and the discontinuous geographical distributions of these subfamilies have provided a compelling stimulus to investigate available evidence which could support a theory of a Pacifican origin for the superfamily. Detailed analyses and resulting determinations follow. FAMILIAL RELATIONSHIPS Pilsbry (1939) placed all Western American and Carib- bean Helicoidea into the single family Helminthoglyptidae. He recognized no less than eight strongly differentiated groups which he ranked as subfamilies. Only seven, however, were named: Helminthoglyptinae Pilsbry, 1939; Sonorellinae Pilsbry, 1939; Humboldtianinae Pilsbry, 1939; Cepoliinae Pilsbry, 1939 (emendation for Cepolinae Hoffman, 1928, a homonym); Xanthonychinae Strebel and Pfeffer, 1879 (as Xan- thonychidae); Lysinoinae Hoffman, 1928; Epiphragmophori- nae Hoffman, 1928. Of these seven subfamilies, Zilch (1959-1960) lumped Lysinoinae into Humboldtianinae and Epiphragmophorinae into Helminthoglyptinae. Subsequent- ly, Schileyko (1978) raised Humboldtianinae to familial rank. The use of the name Helminthoglyptidae was brought into question by Baker (1959) when he showed that the name Xanthonychidae Strebel and Pfeffer, 1879 had considerable priority over Helminthoglyptidae Pilsbry, 1939. Furthermore, Baker showed in a precise chronological review of family names from 1867 to 1958 that the Old World family Brady- baenidae had been considered repeatedly to include many New World subfamilies of the Xanthonychidae. Indeed there were no precise, consistent characters that could be used to separate Bradybaenidae from Xanthonychidae. Nevertheless, Baker condescended to allow the use of the name Brady- baenidae with this statement (1959): ‘‘Since the sizes of families are matters of convenience and/or custom, we Americans, North and South, can leave to the wisdom of our Old World colleagues the advisability of a separate family for the genera of their home lands’. To date, European mala- cologists have continued to use the name Bradybaenidae for what Baker considered to be ‘‘Old World Xanthonychidae’’, while many American malacologists have continued to use the name Helminthoglyptidae for New World Xanthonychidae. All helicoid families, Old World as well as New World, are characterized by having a reproductive system equipped with a dart and mucus gland apparatus associated with or in close proximity to the vagina; certain genera in this super- family without a dart apparatus are believed to have become secondarily simplified in their evolution from dart-bearing ancestors. The principal characters used in separating families within the Helicoidea are the type and shape of the mucus glands and the position of their insertion into the vagina or the dart sac. Pilsbry (1939) published a simple key of distinguishing characters to differentiate each family which can be summarized as follows: |. HELICIDAE: medium or large snails, usually with banded shells, having one dart sac with two tubular, simple or branching mucus glands inserted close to its base, the spermatheca on a long duct which usually bears a branch. Il. HELICELLIDAE (now in Hygromiidae): of medium or small size, with the dart sac often twinned, sometimes want- ing, the tubular mucus glands when present inserted well above it on the vagina; spermathecal duct medium or short, never branching. Ill. HELMINTHOGLYPTIDAE (now Xanthonychidae): dart sac or sacs and mucus glands present, the latter club- shaped, globular or irregular (not tubular or finger-shaped), inserted close to the base of the dart sac. Pilsbry then briefly referred to the Bradybaenidae ‘‘of Eastern Asia’ as having irregular type mucus glands open- ing through an accessory sac on the dart sac or sometimes directly at the base of the latter. It was clear from the above definitions that Pilsbry’s Helminthoglyptidae and Bradybaenidae were a catch-all group for those species whose mucus glands were not tubular or finger-shaped. Schileyko (1978) reviewed the superfamily Helicoidea, with emphasis on anatomical characters. He recognized that the Humboldtianinae formed a consistently distinct group characterized by four compact dart sacs arranged circumferentially high on the vagina, and four globular, compact mucus glands also arranged circum- ferentially above the dart sacs and inserting directly into the vagina at the level of the dart sacs. He therefore raised this subfamily to familial rank as a distinct and separable taxon whose character differences were of equal magnitude as the character differences used in separating the Helicidae and Hygromiidae. Miller (1987) agreed with Schileyko and showed that the genus Bunnya Baker, 1942, was characterized by a similar set of dart sacs and mucus glands, except that their number was three instead of four, and thus this genus should be included in the Humboldtianidae. Most recently, Nordsieck (1987) published a revision of the taxonomy of the Helicoidea based on a detailed study of the dart apparatus and mucus glands of the various groups. He illustrated diagrammatically the various configurations of the system and defined each of the numerous families, sub- families, and tribes, many of them newly erected. He then prepared an elaborate cladogram based not only on the dart apparatus and mucus glands but also on the presence or absence of accessory seminal vesicles, chromosome number, presence or absence of accessory dart sacs, presence or absence of spermathecal diverticulum, and the position of the spermatheca (either along the spermoviduct or bent away from it.). The main thrust of Nordsieck’s cladogram, however, was to concentrate on the evolution of the Helicidae, Hygromiidae, and Bradybaenidae. A similar detailed analysis would have been highly desirable for the Xanthonychidae, MILLER AND NARANJO-GARCIA: AMERICAN HELICOIDEA 149 although it is recognized that there are large gaps in our knowledge of anatomical characters and chromosome numbers of the numerous xanthonychid subfamilies. Nord- sieck affirmed that the correct familial name for the Western American and Caribbean Helicoidea should be Xanthonychi- dae as Baker (1959) had shown, and he recognized the follow- ing nine subfamilies in the Americas: Monadeniinae Nord- sieck, 1987; Helminthoglyptinae Pilsbry, 1939; Cepoliinae Pilsbry, 1939; Epiphragmophorinae Hoffman, 1928; Tricho- discininae Nordsieck, 1987; Lysinoinae Hoffman, 1928; Xan- thonychinae Strebel and Pfeffer, 1880; Metostracinae Nord- sieck, 1987; Humboldtianinae Pilsbry, 1939. By returning the Humboldtianinae to the Xanthonychidae as a subfamily, he disagreed with Schileyko’s familial ranking for this group. He concurred that Bunnya belonged with the Humboldtianinae and erected the new tribe Bunnyini for this genus. Nordsieck’s revision of the Helicoidea provides the most authoritative classification of the superfamily at this time. Disagreements with his findings must be substantiated with evidence. In addition, the dictates of the International Code of Zoological Nomenclature must be followed if we are to avoid chaos in nomenclature. During the past many years, we have been examining the reproductive anatomies of most of the numerous genera of the western North American helicoids and, in many cases, most or all of their individual species. The genera most throughly examined are the following: Helminthoglypta Ancey, 1887; Micrarionta Ancey, 1880; Xerarionta Pilsbry, 1913; Eremarionta Pilsbry, 1913; Plesarionta Pilsbry, 1939; Monadenia Pilsbry, 1895; Sonorella Pilsbry, 1900; Sonorelix Berry, 1943; Mohavelix Berry, 1943; Tryonigens Pilsbry, 1927; Greggelix Miller, 1972; and Eremariontoides Miller, 1981. Ad- ditional genera also critically examined but from only a few representative species are the following: Epiphragmophora Doring, 1875; Averellia Ancey, 1887; Humbolatiana l|hering, 1892; Lysinoe H. and A. Adams, 1855; Cepolis Montfort, 1810; Bunnya Baker, 1942. In all of our examinations, we were impressed repeated- ly by certain distinguishing characters that stood out markedly from all others and were consistently occurring in the species of what are now classified by Nordsieck as three subfamilies, namely the Humboldtianinae, the Helminthoglyptinae (sensu Nordsieck, i.e. minus Monadenia which had been included by Pilsbry), and the Monadeniinae. In the genus, Humbolat- iana, all of our species had four compact, vesicular mucus glands, circumferentially arranged around the vagina and in- serting directly into it; additionally, they had four dart sacs also circumferentially arranged around the vagina and located immediately below the mucus glands. This arrangement was strikingly unlike any other system found in any other helicoid except Bunnya. In the Helminthoglyptinae, all dart bearing species had one or both mucus glands consisting of wide membranes that wrapped around various parts of the anterior end of the reproductive tract such as the penis, and/or the vagina, and/or the artrial sac. Again, this arrangement was strikingly different from any other helicoid system. Finally, in the Monadeniinae, all species were equipped with only a single, tubular mucus gland which inserted on a large swelling of the anterior end of the reproductive tract, a structure somewhat resembling the accessory sac (Nordsieck’s Neben- sack) found in some of the Bradybaenidae. Although Pilsbry (1939) considered the anterior muscular swelling to be simply an atrium, the similarity with the bradybaenid accessory sac could not be ignored; furthermore, the tubular mucus gland would, by Pilsbry’s own definition, exclude this subfamily from the Xanthonychidae. We came to the conclusion, therefore, that the anatomical differences found in the Humboldtianinae and in the Helminthoglyptinae were as distinctive and of equal or greater magnitude as those now used to designate familial rank for other helicoid families, namely Helicidae, Hygromiidae, Bradybaenidae, and Sphincterochilidae. Accord- ingly, we concur with Schileyko that the Humboldtianinae should be raised to familial rank. We further propose that the Helminthoglyptinae, sensu Nordsieck, also be raised to familial rank. Because Miller (1970, 1973, 1981) already pro- vided evidence that certain genera in Pilsbry’s Sonorellinae (Nordsieck’s Sonorellini) probably arose from different helmin- thoglyptine ancestors, namely Mohavelix Berry from Hel- minthoglypta micrometalleoides Miller, 1970, and Eremarion- toides Miller from Eremarionta greggi Miller, 1981, the Sonorellini must be considered polyphyletic and unacceptable as a taxon. To date, there are no convincing data available to indicate the immediate dart-bearing ancestors of Sonorella Pilsbry, Sonorelix Berry, Greggelix Miller, and Tryonigens Pilsbry. Classification of the helminthoglyptid genera at the subfamilial level, therefore, will have to await more sophisti- cated methods of analysis probably involving chromosome banding and DNA hybridization. Finally, we also concluded that the Monadeniinae were more closely similar, anatomically, to the Bradybaenidae than to the Xanthonychidae, Humboldt- ianidae, or Helminthoglyptidae so that they should be classified either as a bradybaenid subfamily or raised to familial rank. At this time, we recommend the conservative approach of leaving them in subfamilial status as a fourth bradybaenid subfamily along with Bradybaeninae Pilsbry, 1924, Aegistinae Kuroda and Habe, 1955, and Helicostylinae Ihering, 1909. A simple key can be erected as follows to separate the Humboldtianidae and the Helminthoglyptidae from the Xanthonychidae: 1. One or more mucus glands membranous ........... Such pths dh Patapon ec yn tates Helminthoglyptidae Mucus glands vesicular ........................ 2 2. Mucus glands and dart sacs compact, multiple, seated high on the vagina ................... Humboldtianidae Mucus glands and dart sacs not thus............... Additionally, figure 1 shows the most probable phylogeny of the families, based on the concept that membranous mucus glands are probably ancestreal to vesicular glands. In summary, the classification of the Western American and Caribbean Helicoidea, lumped by Nordsieck (1987) into the single family Xanthonychidae, is now proposed as follows: 150 AMER. MALAC. BULL. 8(2) (1991) Helicoidea Rafinesque Xanthonychidae Strebel and Pfeffer Cepoliinae Pilsbry Epiphragmophorinae Hoffman Trichodiscininae Nordsieck Lysinoinae Hoffman Xanthonychinae Strebel and Pfeffer Metostracinae Nordsieck Humboldtianidae Pilsbry Humboldtianinae Pilsbry Bunnyinae Nordsieck Helminthoglyptidae Pilsbry Bradybaenidae Pilsbry Monadeniinae Nordsieck BIOGEOGRAPHY After determining that the anatomical characteristics of the Western American and Caribbean Helicoidea were suf- ficiently distinct to separate at least four different families, namely Xanthonychidae, Bradybaenidae, Humboldtianidae, and Helminthoglyptidae, our attention turned to their possi- ble evolution and ultimate dispersal from a common origin. Figure 2 shows the general distribution of the Western American and Caribbean Helicoidea and their East Asian relatives. To account for this nearly circum-Pacific distribu- tion, many terrestrial malacologists theorized that these helicoids had a Eurasian origin and migrated into the Americas via a Bering land bridge during the Tertiary period (Pilsbry, 1894; Pilsbry, 1948; Gregg, 1959). This theory seemed to us to leave too many important questions un- answered. For example, in order to attain the current distribu- tion in the Americas (Fig. 2), the theory presumed a long and narrow dispersal along the west coast of North America into Mexico and Central America followed by a radiation eastward to the main Caribbean islands as well as many of the lesser Antilles, ultimately as far as the Bahamas and southern Florida. In the meantime, however, these helicoids failed to reach the central and eastern United States although there is ample evidence that during that same time period the Polygyridae were able to populate extensive areas from southern Florida to New England to Washington, Oregon, and northern California and nearly all of central and western Mex- ico. The Helicoidea also would have crossed the isthmus of Panama to reach vast areas of Peru and northwestern Argentina but they failed to leave any trace in Panama, Col- ombia, and Ecuador; they also failed to disperse into Venezuela and Brazil. More recently, as a result of a five year study of helminthoglyptid distribution in Sonora, Mexico, by one of us (ENG), we found a complete absence of helmintho- glyptids from the latitude of Hermosillo to the Sinaloan border (Naranjo-Garcia, 1988); yet we found that region to be well populated by other families of large snails such as bulimulids and polygyrids. Such gaps in distribution, unexplainable by geological or ecological events due to the presence of other families of large snails in these gaps, presented a serious flaw in the theory of a Bering land bridge migration. During the past four decades, several biogeographers Xanthonychidae Humboldtianidae Helminthoglyptidae mucus glands and dart sacs compact, multiple, seated high on vagina mucus glands and dart sacs not seated high on vagina one or more mucus mucus glands glands membranous vesicular; none membranous Fig. 1. Probably phylogeny. began to question the widely invoked theories of palearctic origins for much of the New World biota. Croizat (1952) and Melville (1966, 1981) suggested that the exchange of biota through a Bering land bridge was relatively insignificant. Melville (1966), in order to explain the distribution of angiosperms along the Pacific Rim, proposed the existence of a mesozoic land mass in the south-central Pacific, which he named Pacifica, that broke up and migrated tectonically to accrete to the continental margins of Asia and America. Nur and Ben-Avraham (1977) suggested that the circum- Pacific mountain belts could have been the result of past con- tinental collisions similar to those associated with the Alpine belt. They proposed that these collisions were made by parts of a continental mass situated in the South Pacific Ocean, Pacifica (referring to Melville’s 1966 article), which disag- gregated during the Mesozoic and spread out on the Kula, Farallon, Phoenix, and Pacific plates eventually to reach con- tinental margins. Subsequently, additional geophysical and geological evidence appeared in the literature in support of the former existence of Pacifica (Kamp, 1980; Davis et al., 1978; Coney et a/., 1980; McGeary and Ben-Avraham, 1981; Nur and Ben-Avraham, 1982). Then Jones et al. (1982) showed convincing evidence, supported by the work of Tarduno et al. (1986), that western North America consists of accreted ter- ranes which originated thousands of kilometers to the south and west of their present position. Jones et a/. also provided evidence to show that Permian terranes, originally formed in the Tethys Sea, had also accreted to form a large part of eastern Asia. Kulm et al. (1986), while studying the subduc- tion zone of Oregon and Washington which possesses ac- creted terranes, found communities of clams and tube worms (Calyptogena sp.) similar to those found in the accretionary complexes of Japan, the Philippine Plate, and other locations around the Pacific Ocean. Figure 3 is a schematic composite model of Pacifica land mass migrations during the Mesozoic era, according to these cited authors. In the light of this mounting volume of evidence, we studied the possibility that the circum-Pacific Ocean distribu- tion of the Helicoidea could be much better explained by theorizing a center of origin on a Mesozoic Pacifican land mass, Pacifica, which broke up into several parts that were ultimately carried, as Nur and Ben-Avraham suggested, to form large parts of western North and South America and east Asia. These terranes provided a passive means of dispersal for ancestral populations of helicoids to dock at various MILLER AND NARANJO-GARCIA 151 1 Ny \ oot \ \ / ge Figure 2. Approximate areas of distribution of Western American and Caribbean Helicoidea and Asian Bradybaenidae: B, Bradybaenidae (other than Monadeniinae); C, Cepoliinae (Xanthonychidae); E. Epiphragmophorinae (Xanthonychidae); He, Helminthoglyptidae; Hu, Hum- boldtianidae; M, Monadeniinae (Bradybaenidae); X, Xanthonychidae (other than Cepoliinae and Epiphragmophorinae). discrete parts of the Americas and Asia. With this theory, the absence of any Helicoidea from Panama, Colombia, Ecuador, and a large area of Sonora did not need to be attributed to mass extinctions. Moreover, the time period involved in the break-up and migration of the Pacifican terranes during the Mesozoic would have permitted ample isolation for the evolu- tion of separate families and subfamilies on each different terrane. We came to the conclusion that the Pacifican theory did indeed better explain the circum-Pacific distribution of the helicoids than the trans-Beringean theory. To explain the cur- rent distribution of the families and subfamilies of these helicoids, we hypothesize the following vicariance patterns: 1) the arrival of the ancestral Bradybaenidae in eastern Asia, with subsequent dispersal along the shore of the Tethys Sea to south-central Asia and into southern Europe; 2) the arrival of the ancestral Helminthoglyptidae in western North America, docking along what is now California and Baja California, with eventual dispersal eastward as far as west Texas and Chihuahua; 3) the arrival of the ancestral Xanthonychidae, other than Epiphragmophorinae and Cepoliinae, along the shores of southwestern Mexico and Central America; 4) the arrival of the ancestral Epiphragmophorinae in the vicinity of Peru, with eventual dispersal southeasterly as far as northwest Argentina and southern Brazil; 5) the arrival of the ancestral Cepoliinae into the Caribbean region on the Greater Antilles terranes which Burke et al. (1984) stated had origins in the Pacific Ocean during the Mesozoic and migrated to collide with the Bahamas long before the formation of the Isthmus of Panama. In the case of the Humboldtianidae, conchological and anatomical characters indicate a closer evolutionary relation- ship to the European helicoids than to the other American helicoids (Schileyko, 1978). Additionally, their widespread distribution not only over the entire Mexican Plateau but also, as fossils, as far north as east-central Wyoming led us to nypothesize that they must have been indigenous to the North American craton after it separated from Laurasia. They are, therefore, the one American helicoid family that apparently did not evolve from Pacifican ancestors. Although the theory of a Pacifican origin for the East Asian, Caribbean, and Western American helicoids (except- 152 AMER. MALAC. BULL. 8(2) (1991) Figure 3. Schematic composite model of estimated Pacifica land mass migrations during Mesozoic era. Heavy lines mark estimated posi- tions of continental areas at end of Mesozoic, according to Dietz and Holden (1975). Horizontal dashed lines mark presumed Pacifican ter- ranes that accreted to the continents according to Jones et a/. (1982) and Burke et a/. (1984). Diagonal dashed lines mark estimated posi- tion of Pacifica at beginning of Mesozoic according to Nur and Ben- Avraham (1977). Major Pacific plates during Mesozoic according to Zonenshayn et a/. (1984): ANT, Antarctica; AUS, Australia; EURA, Eurasia; FAR, Farallon plate; KUL, Kula plate; NAM, North America; PAC, Pacific plate; PHO, Phoenix plate; SAM, South America. ing the Humboldtianidae) satisfactorily explains the distribu- tion of the current populations, it does not rule out the possibili- ty of some limited Tertiary trans-Beringean migration. As stated earlier, we consider the Monadeniinae to be more closely related to the Bradybaenidae than to any other American helicoid group. The possibility exists, therefore, that this bradybaenid subfamily could have migrated across the Ber- ing land bridge from northeast Asia. Conchological characters would support a very close relationship between Japanese Euhadra species and Monadenia species. Anatomically, however, the Monadeniinae are very different from Japanese Euhaadra or from northeast Siberia Bradybaena. Accordingly, it appears to us that the ancestral Monadeniinae evolved on a separate Pacifican terrane that ultimately docked on the North American continent somewhere on the Canadian or southern Alaskan coast. ACKNOWLEDGMENTS We are especially indebted to Barry Roth for making us aware of the Pacifica theory and suggesting that our helicoids could have originated there. Moreover, he assisted greatly in bringing to our at- tention the fact that Monadenia appeared to have greater affinity with the Bradybaenidae as well as in providing the key and the cladogram used in this article to differentiate between the Xanthonychidae, Helminthoglyptidae, and Humboldtianidae. We also wish to thank both Eric Hochberg and Barry Roth for their encouragement and their helpful suggestions in the preliminary review of this manuscript. The Consejo Nacional de Ciencia y Tecnologia provided grant support for the studies in Sonora by one of us (ENG). LITERATURE CITED Baker, H. B. 1959. Xanthonychidae (Pulmonata). Nautilus 73(1):25-28. Burke, K., C. Cooper, J. F. Dewey, P. Mann, and J. L. Pindell. 1984. Caribbean tectonics and relative plate motions. Geological Society of America Memoirs 162:31-63. Coney, P. J., D. L. Jones, and J. W. H. Monger. 1980. Cordilleran suspect terranes. Nature 288:329-333. Croizat, L. 1952. Manual of Phytogeography. Uitgeverij Dr. W. Junk, The Hague. 587 pp. Davis, G. A., J. W. H. Monger, and B. C. Burchfiel. 1978. Mesozoic construction of the cordillera ‘‘collage’’, central British Colum- bia to central California. In: Mesozoic paleogeography of the western United States, Pacific Coast Paleogeography Sym- posium 2, D. G. Howell and K. A. McDougall, eds. pp. 1-32. Pacific Section, Society of Economic Paleontologists and Mineralogists, Los Angeles. Dietz, R. S. and J. C. Holden. 1975. Reconstruction of Pangaea; breakup and dispersion of continents, Permian to present. Journal of Geophysical Research 75(26):4939-4950. Gregg. W. O. 1959. Derivation of the Helminthoglyptinae with par- ticular reference to the desert forms. American Malacological Union, Pacific Division Twelfth Annual Meeting:45-46. Jones, D. L., A. Cox, P. Coney, and M. Beck. 1982. The growth of western North America. Scientific American 247(5):70-84. Kamp, P. J. J. 1980. Pacifica and New Zealand - proposed eastern elements in Gondwanaland history. Nature 288(5792):659-664. Kulm, L. D., E. Suess, J. C. Moore, B. Carson, B. T. Lewis, S. D. Ritger, D. C. Kadko, T. M. Thornburg, R. W. Embley, W. D. Rugh, G. J. Massoth, M. G. Langseth, G. R. Cochrane, and R. L. Scam- man. 1986. Oregon subduction zone: venting, fauna, and car- bonates. Science 231(4738):561-566. McGeary, S. E. and Z. Ben-Avraham. 1981. Allochthonous terranes in Alaska: implications for the structure and evolution of the Bering Sea shelf. Geology 9(12):608-614. Melville, R. 1966. Continental drift, Mesozoic continents and the migra- tion of the angiosperms. Nature 211:116-120. Melville, R. 1981. Vicarious plant distributions and paleogeography of the Pacific region. In: Vicariance Biogeography: A Critique, G. Nelson and D. E. Rosen, eds. pp. 238-274. Columbia Uni- versity Press, New York. Miller, W. B. 1970. A new species of Helminthoglypta from the Mojave Desert. Veliger 12(3):275-278. Miller, W. B. 1973. Saltational speciation in American Helmin- thoglyptidae (Gastropoda: Pulmonata). Bulletin of the American Malacological Union for 1973:44. Miller, W. B. 1981. A new genus and a new species of Helmintho- glyptid land snails from the Mojave Desert of California. Pro- ceedings of the Biological Society of Washington 94(2):437-444. Miller, W. B. 1987. A new species of Bunnya (Gastropoda:Pulmonata: Humboldtianidae) from western Mexico, with notes on its life cycle and familial relationships. The Veliger 29(3):308-312. Naranjo-Garcia. E. 1988. Systematics and biogeography of the Helminthoglyptidae of Sonora. Doctoral Dissertation E9791 1988 272, University of Arizona. University Microfilms International, Ann Arbor, Michigan. Nordsieck, H. 1987. Revision des Systems der Helicoidea MILLER AND NARANJO-GARCIA 153 (Gastropoda:Stylommatophora). Archiv fur Molluskenkunde 118(1/3):9-50. Nur, A. and Z. Ben-Avraham. 1977. Lost Pacifica continent. Nature 279(5632): 41-43. Nur, A. and Z. Ben-Avraham. 1982. Oceanic plateaus, the fragmen- tation of continents, and mountain building. Journal of Geo- physical Research 87(B5):3644-3661. Pilsbry, H. A. 1894. Guide to the study of Helices. Manual of Conchology 9. Academy of Natural Sciences of Philadelphia, Philadelphia. 366 pp. Pilsbry, H. A. 1939. Land Mollusca of North America (north of Mex- ico). Academy of Natural Sciences of Philadelphia, Mono- graphs Number 3, 1(1), Philadelphia. 573 pp. Pilsbry, H. A. 1948. Land Mollusca of North America (north of Mex- ico). Academy of Natural Sciences of Philadelphia, Monographs Number 3, 2(2), Philadelphia. 592 pp. Solem, A. 1978. Classification of the land mollusca. /n: Pulmonates, 2A, V. Fretter and J. Peake, eds. pp. 49-97. Academic Press, London. Schileyko, A. A. 1978. Land mollusks in the superfamily Helicoidea. Fauna of the USSR, Mollusks 3(6), USSR Academy of Science, Leningrad. 384 pp. Tarduno, J. A., M. McWilliams, W. V. Sliter, H. E. Cook, M. C. Blake, Jr., and |. Premoli-Silva. 1986. Southern hemisphere origin of the Cretaceous Laytonville limestone of California. Science 231(4744):1425-1428. Zilch, A. 1959-1960. Euthyneura. /n: Gastropoda, Handbuch der Palaozoologie 6(2), W. Wenz, ed. 834 pp. Gebruder Born- traeger, Berlin. Zonenshayn, L. P., L. A. Savostin, and A. P. Sedov. 1984. Global paleogeodynamic reconstructions for the last 160 million years. Geotectonics 18(3):181-192. Date of manuscript acceptance: 26 July 1990 = ‘ oe a : & : f we ‘eel — ta A PHYLOGENETIC ANALYSIS AND REVISED CLASSIFICATION OF THE NORTH AMERICAN HAPLOTREMATIDAE (GASTROPODA: PULMONATA) BARRY ROTH DEPARTMENT OF INVERTEBRATE ZOOLOGY SANTA BARBARA MUSEUM OF NATURAL HISTORY SANTA BARBARA, CALIFORNIA 93105, U. S. A. ABSTRACT A revised classification of the North American Haplotrematidae (=Haplotrematinae) is presented, based on character compatibility analysis of 25 characters of the musculature, reproductive system, radula, kidney, pericardium, and shell. Two genera are recognized -- Ancotrema and Haplotrema, the latter with three subgenera, Ancomena, Geomene, and Haplotrema, sensu stricto. The hypothesized age of the Haplotrematinae is 42.5 million yr (late middle Eocene), when the temperate Haplotrematinae and its tropical out-group, the Austroselenitinae, are assumed to have diverged. The phylogenetic hypothesis indicates an early dichotomy between Ancotrema and Haplotrema, with the species of Ancotrema retaining more characteristics of the hypothetical ancestor. Other evolutionary trends in- clude: (a) a tendency toward reduction of vaginal musculature; (b) a parallel tendency toward greater penial complexity; (c) migration of the origin of the penial retractor muscles from the columellar mus- cle bundle to the floor of the lung (presumably increasing mechanical efficiency); (d) parallel instances of gigantism in the Pacific Northwest; (e) size reduction southward along the Pacific Coast, correlating with the presence/severity of a summer dry season; (f) a homoplastic continuation of the loss or reduction of cusps on the teeth of the radula. A lectotype is designated for Selenites vancouverensis forma hybrida Ancey, 1888. The Haplotrematidae are a specialized family of rhytidoidean land snails restricted to the Western Hemisphere. Baker (1941) divided the family into the subfamilies Haplotrematinae (North American) and Austroselenitinae (Antillean and South American). The Haplotrematinae con- stitute a holophyletic group (sensu Ashlock, 1971), based on characters of the kidney, ureter, sigmoid loops of the hindgut, salivary gland conformation, jaw, and radula. In the current classification (Pilsbry, 1946), which draws heavily on the detailed anatomical studies of Baker (1931, 1941), all species of Haplotrematinae are assigned to the genus Haplotrema Ancey, 1881. [The classification of Zilch (1960) merely strews all previously proposed genera and subgenera of Haplotrematidae as subgenera under the genus Haplotrema.] Table 1 compares the classifications of Baker (1931) and Pilsbry (1946), both of which used the non-standard category ‘section,’ with the classification proposed in this paper. Since the time of those authors’ writings, methods have been developed that enable systematists to analyze patterns of character variation in accordance with stated criteria in order to develop hypotheses about the evolutionary history of groups of organisms. The purpose of this study is to apply some of these methods to derive a classification of the species of Haplotrematinae that is supported by as many indepen- dent characters as possible and therefore represents the best available estimate of the phylogenetic relationships of these taxa. The method used is character compatibility analysis (Estabrook, 1972; Le Quesne, 1982), which has its roots in the observations of Wilson (1965) and Le Quesne (1969, 1972). Character compatibility analysis is the method of choice where, as here, one seeks the tree exhibiting the most con- gruence in the synapomorphy pattern of the taxa under study (Strauch, 1984). The analysis incidentally provides a test of how well the characters studied and utilized by Baker (1931, 1941) support his own classification. This inquiry grew out of my preparation of: (a) a treat- ment of the Haplotrematidae for the Council of Systematic Malacologists ‘‘Checklist of the Non-marine Mollusca of the United States and Canada’’ (Pratt, in prep.) and (b) an identi- fication manual of the land snails and slugs of California for the California Departments of Food and Agriculture and Fish and Game. This analysis and discussion are beyond the scope of either of those works, and so are presented in this paper. American Malacological Bulletin, Vol. 8(2) (1991):155-163 155 156 AMER. MALAC. BULL. 8(2) (1991) Table 1. Comparison of classifications of the North American Haplotrematidae (=Haplotrematinae). Baker (1931) Pilsbry (1946) This Paper Genus Haplotrema Ancey, 1881 subg. (Haplotrema, s.s.) sect. Haplotrema, s.s. sect. Geomene Pilsbry, 1927 subg. (Ancotrema) Baker, 1931 sect. Ancotrema, S.s. sect. Ancomena Baker, 1931 Genus Haplotrema subg. (Haplotrema, s.s.) sect. Haplotrema, s.s. sect. Geomene subg. (Ancomena) sect. Ancomena, s.s. sect. Greggiella Baker, 1941 Genus Haplotrema subg. (Haplotrema, s.s.) subg. (Geomene) subg. (Ancomena) Genus Ancotrema sect. Ancotrema MATERIALS AND METHODS Character compatibility analysis (Estabrook, 1972, 1978) was used to identify patterns of agreement and disagreement among characters in the data set and to identify which characters were most useful in constructing an estimate of the phylogenetic relationships among the 16 valid specific and infraspecific taxa assigned to the Haplotrematinae. This analysis was performed manually according to the method described by Meacham (1981), and subsequently using the Fortran program CLINCH by Kent L. Fiala, with identical results. Examples of the use of character compatibility analysis include Duncan (1980) and Meacham (1980), where further references to its theoretical and mathematical basis can be found. The basic structure of the phylogenetic hypothesis was established from the primary clique(s) of characters derived through compatibility analysis. Then rejected characters and character state trees were re-evaluated for the additional in- formation they could provide. Anatomical data are lacking for some of the taxa of the Haplotrematinae. Four taxa regarded by previous authors as species either have not been dissected or the available anatomical information is incomplete. They are therefore ex- cluded from the analysis: tentative suggestions for their place- ment are given in the discussion. The subspecific relation- ship between Haplotrema duranti duranti (Newcomb, 1864) and H. d. contintentis (Baker, 1931) is maintained, although only shell and radular data are available from H. a. duranti. One other taxon is regarded as a species inquirenda. Sources of data are listed below. The abbreviations BR and TP designate locality numbers from the collections of the author (San Francisco, California) and of Timothy A. Pearce (Ann Arbor, Michigan), respectively. In addition to the anatomical material cited, approximately 900 lots of shells were examined, mainly in the Santa Barbara Museum of Natural History, California Academy of Sciences, and the author’s collection. TAXA INCLUDED AND DATA SOURCES alameda Pilsbry, 1930. Baker (1931); author’s dissec- tion of nominal subsp. fie/di Pilsbry, 1930, from Kern River canyon, 1.6 km southwest of Democrat Hot Springs, Kern County, California (BR 957). caelatum (Mazyck, 1886). Baker (1931, 1941). concavum (Say, 1821). Baker (1931), Pilsbry (1946); additional information on genitalia from Webb (1943). Subsp. minus (Ancey, 1882) is a synonym (Pilsbry, 1946). duranti (Newcomb, 1864). Radula from topotypic subsp. duranti (Baker, 1931); other anatomy based on H. d. continentis Baker, 1931 (Baker, 1931). hybridum (Ancey, 1888). Baker (1931); author’s dissec- tion from Cape Blanco, Curry County, Oregon (BR 1173). Figure of genitalia by Webb (1961) shows some characters as in Ancotrema sportella, others equivocal [e.g. bulge on left side of vagina labeled ‘‘pilaster’’ resembles muscular collar found in this position in A. sportella (Gould, 1846) and other species]. Although Baker (1931) reported intergradation be- tween ‘‘the semidecussatum form of hybridum’”’ and A. sportella in western Washington, in northwest California and southwest Oregon hybridum occurs at the same localities as typical A. sportella and Haplotrema vancouverense (Lea, 1839) without intergradation. Lectotype designation in a following section fixes type locality as Astoria, Oregon. ‘Form’ semidecussatum Gratacap is not an available name. The original publication (as Macrocyclis vancouveren- sis [Lea] var. semi-decussata) was not intended as the pro- posal of a new taxon but merely the reference to a museum label (apparently from Thomas Bland) accompanying ‘‘a very interesting specimen...from Astoria, Oregon’ (Gratacap, 1901:340). The text goes on to quote the opinion of H. A. Pilsbry that ‘‘it should not be accorded position and is probably var. hybrida Ancey’’ (Gratacap, /oc. cit.). Baker (1931) provided diagnostic information but treated it as in- frasubspecific ‘‘form’’ of ‘‘Haplotrema sportella hybridum.”’ The anatomy is reported to be the same as in A. hybridum (Baker, 1931). ‘Var?’ depressa Ancey, 1888, is an unavailable infra- subspecific name. keepi (Hemphill, 1890). Baker (1931); Webb’s (1961) figure of genitalia of specimen from Dog Creek, Shasta Coun- ty, California, and author’s dissection from Signal Butte, Shasta County (TP 861026-1215), differ in characters of the penis and spermathecal duct and could represent a different, undescribed species. minimum (Ancey, 1888). Baker (1931); author’s dissec- tions from Duncan Point, Sonoma County (BR 1466); Point Reyes Peninsula, Marin County (BR 1490, BR 1640, BR 1642); San Pablo Dam Road above San Pablo Reservoir, Contra Costa County (BR 316); Sharps Park, San Mateo County (BR 104); Malpaso Creek, Monterey County (BR 1503), California. ‘‘Forms’’ occidentale (Hemphill In: W. G. Binney, 1892), ROTH: NORTH AMERICAN HAPLOTREMATIDAE as VA tenue (Hemphill In: W. G. Binney, 1892), and ke/seyi (Hemphill, 1911) are synonyms (Pilsbry, 1946). sportella (Gould, 1846). Baker (1931); author’s dissec- tions from along Woods Creek, west of Philomath, Benton County, Oregon (BR 1681); Moonstone Beach, Humboldt County, California (TP 831228-1). Figures of genitalia by Webb (1961) corroborate some characters. Porter’s (1965, 1968) figures of the genitalia of a specimen from near Corvallis, Oregon, show an extremely short, saccular spermathecal duct with a pointed end, not otherwise reported in the species. A specimen that | dissected from Woods Creek, about 15 km west of Corvallis (BR 1681), has the elongate spermathecal duct and globose spermatheca normal in the species. Perhaps Porter’s (1965, 1968) material was abnormal, or the structures were misinterpreted. transfuga (Hemphill /n: W. G. Binney, 1892). Baker (1941). vancouverense (Lea, 1839). Baker (1931); author’s dissections from Brookings, Curry County, Oregon (BR 357); Luffenholtz Beach, south of Trinidad, Humboldt County, Cali- fornia (BR 354). Figures of genitalia by Webb (1961) and Porter (1965, 1968) add no new information. ‘‘Forms’’ vellicatum (Forbes, 1850) and chocolatum (Dall, 1905) are synonyms (Pilsbry, 1946), although a lectotype designation for chocolatum may be needed to confirm this (cf. remarks by Henderson 1936:257). voyanum (Newcomb, 1865). Baker (1931, 1941); author’s dissection from Bidden Creek, Trinity County, California (BR 1229). TAXA NOT INCLUDED IN THE ANALYSIS catalinense (Hemphill In: W. G. Binney, 1890). Not dissected. costatum Smith, 1957. Not dissected. guadalupense Pilsbry, 1927. Not dissected. humboldtense Pilsbry, 1946. Based on material that Baker (1931) included in H. voyanum, Haplotrema voyanum humbolatense Pilsbry, 1946, was described in such vague terms and with such a generalized type locality (“‘Humboldt and ‘Klamath’ counties,’ Calif.) that even a meaningful search for the type population would be difficult. Whether Klamath County, Oregon, was meant (Baker, 1931:421) or the former California county of the same name, as in Pilsbry’s interpreta- tion, is not known. Until such time as the type material can be matched up with a population somewhere, H. v. humbolat- ense must remain a species inquirenda. kendeighi Webb, 1951. Originally described as a subspecies of Haplotrema concavum; regarded as a species by Hubricht (1956, 1985). Dissection by Hubricht (1956) in- completely described. CHARACTERS Appendix A lists the characters, character states,and hypothesized character state trees used in the analysis. Characters of the musculature, reproductive system, radula, kidney, pericardium, and shell are included. No weighting was assigned a priori to the characters of any one system. The polarity of character state transformations was in- ferred mainly on the basis of out-group comparison at various levels of generality, and in some cases by correlation of equivocal characters with the direction of other, better- grounded transformations. For example, the reduction of radular tooth secondary cusps (apomorphous states of characters 3-5) parallels a well-established, general trend in carnivorous land snails and slugs toward a ‘‘slicing’’ or “‘stab- bing’’ tooth morphology (Watson, 1915; Solem, 1974). Some use was made of the assumption that the most common state of a character is primitive. The Haplotrematidae have no significant fossil record, so no inferences could be based on stratigraphic precedence. Direction of ‘‘chorological progression’ — suggested by Hennig (1966) as one method of inferring polarity — could be based, for example, on the observation that the Klamath Mountains are a refugium preserving vegetation (Whittaker, 1961; Axelrod, 1976) and perhaps a land snail fauna (Roth, 1981) more like that of the later Cenozoic than any other region in the west. This would lead to the hypothesis that character states found in Haplotrema voyanum and H. keepi, residents of the Klamath region, are primitive within their clades. However, one of the uses of a phylogenetic hypothesis is to illuminate the biogeographic history of a group, and excessive reliance on this form of reasoning would introduce circularity into a biogeographic analysis. Chorological progression therefore was considered only in the spirit of ‘reciprocal il- lumination’ (Hennig, 1966; Page, 1987). Table 2 shows the distribution of character states among the 12 taxa included in the analysis. A hypothetical ancestor, with all characters in the inferred primitive condi- tion, is added. Its character states are not necessarily those of the Austroselenitinae, because the species of that tropical group have specializations of their own [for example, the slender, elongate atrium of Austroselenites (Zophos) concolor (Férussac, 1820) (Baker, 1941:pl. 9, fig. 6)]. Comments on specific characters follow. The right ocular retractor is free from the genitalia (character 1, state B) in all Austroselenitinae dissected (Baker, 1931, 1941). Posterior position of the external genital orifice (character 2, state B) is an apomorphy of the Haplotrematinae (and perhaps Haplotrematidae) in general, but plesio- morphous within the group. Haplotrema generally enter the shell of a prey snail through the aperture or through a hole broken just behind the peristome (pers. observ.); relocation of the genital apparatus backward probably allows the head and foreparts to reach farther inside the shell. A genital orifice close to the base of the right ommatophore (characters 2, state A) in Haplotrema duranti and H. caelatum probably represents a space-related reversal in these small-sized species. On San- ta Barbara Island, California, H. duranti preys on the pupillid snail Nearctula rowelli (Newcomb, 1860), entering the shell through the side of the whorls of the spire (F. G. Hochberg, pers. comm., 1979); it probably does not require the extreme elongation of the foreparts needed by species that go up the body whorl of their prey. Character 4 (multiple/reduced number of cusps on cen- tral tooth of radula) is kept separate from character 5 158 AMER. MALAC. BULL. 8(2) (1991) Table 2. Distribution of character states among species of North American Haplotrematidae and hypothetical ancestor. Character No. Taxon Symbol 1-5 6-10 11-15 16-20 21-25 caelatum CAE AABBB CAABA AABAA ACBAB BAABB concavum CON ABABA CAABA BABBB ACBAB BBDBA duranti DUR AAABA CAABB BABBB ACBAA ABCBB hybridum HYB BBAAA AAABB AAABB BAABC ABAAA keepi KEE BBABA CAAAB CABBB BCBCA ABDBA minimum MIN BBABA CABBB AABBB BABAA ABDBA alameda ALA BBABA CAABB BABAB ABBAA ABBBB sportella SPO BBAAA AAABB AAABB BAABC ABAAB transfuga TRA BBABB CAABB BABAB ABBAA ABBBA vancouverense VAN BBABA CBBBB ABBBB BACAA ABDBA voyanum VOY BBAAA BAABB AAABB BABAC ABAAB ancestor ANC BBAAA AAABB AAABB BAAAC ABAAA (presence/absence of cusps on central tooth) because multi- ple cusps could be lost without passing through a stage of reduction in number. Migration of the origin of the penial retractor muscle from the columellar bundle to the floor of the lung (character 6) probably represents a transition toward greater mechanical efficiency. A tendency for the main penial chamber to be differentiated into apical and basal parts (character 13) is also seen in the out-group Austroselenitinae. RESULTS PRIMARY CLIQUES Two primary cliques of 17 characters each were de- rived. The cladograms based on these cliques are shown in figure 1. Symbolic abbreviations of species names used in the figures and in the following discussion are given in Table 2. Characters 3, 7, 8, 9, 12, 15, and 22 are compatible with all others in the analysis (all but character 8 are characters unique to one analyzed species) and are therefore common SPO, TRA, ANC HYB VOY MIN VAN KEE ALA DUR CON CAE 22(A) 10(A),21(B) 9(A),19(C) 1(A) 4(B),6(C),13(B),24(B) 6(B),18(B) 3(B),15(A), to both primary cliques. In addition to these ‘‘consensus”’ characters, both cliques include characters 1, 4, 6, 10, 13, 18, 19, 21, and 24. The two cliques differ only in the presence of character 16 (clique A) or character 17 (clique B). The com- ponents of cladograms A and B grouping (ANC,((SPO,HYB), VOY)) and (DUR,(CON,CAE)) are isomorphous in both clado- grams. The cladogram based on clique A, with character 16 (atrium with/without stimulator), contains an unresolved trichotomy. The cladogram based on clique B, with the multi- state character 17 (vagina with shining muscular collar/with sphincteric thickening/without collar or thickening), is more fully resolved and is accepted as the best estimate of actual patterns of ancestry and descent in the Haplotrematinae. OTHER CHARACTERS 2. An external genital orifice close to the right om- matophore in DUR and CAE is a probable homoplasy associ- ated with small size. 5. Absence of cusps on the central tooth of the radula SPO, TRA, ANC HYB VOY MIN VAN ALA KEE DUR CON CAE x 17(C) 17(B) B Fig. 1. Cladograms based on the two primary cliques: (A) based on clique with character 16; (B) based on clique with character 17. Abbrevia- tions of taxa as in Table 2. Numbers on branch segments designate apomorphic states of characters as defined in Appendix (shown in B only where different than in A). ROTH: NORTH AMERICAN HAPLOTREMATIDAE 159 TRA, TRA, ALA KEE DUR CON CAE ALA DUR CON CAE KEE 110A) Reversal Fig. 2. The right-hand branch of cladogram B (from Fig. 1) with in- ferred apomorphies and reversals of character 11 (A) and character 16 (B). Component (KEE,(DUR,(CON,CAE))) is rotated around node x in right diagram for clarity; the order of branching remains unchanged. is a homoplasy, occurring independently in TRA and CAE. 11. Sharp papillae are present on the penial wall of (TRA,ALA), DUR, and CON; blunt papillae are present in KEE. As shown in figure 2, derivation of blunt papillae from sharp, and one reversal (to papillae absent in CAE), are sufficient to incorporate this character into the basic cladogram. 14. A penial sheath is present in CAE and (TRA,ALA); homoplastic origins are indicated. 16. A stimulator is present in the atrium in (TRA,ALA) and (DUR,(CON,CAE)) but absent in KEE. As shown in figure 2, asingle reversal is sufficient to incorporate this character into the basic cladogram. 20. The relative length of the kidney is greatest in ((MIN VAN),((TRA,ALA),KEE,DUR)) and least in ((SPO,HYB), VOY); a partial reversal involving (CON,CAE) is indicated. When recoded C — A — B, this character becomes compati- ble with the primary clique. 23. A basic trend from broad to narrow radial ribs is established but the absence of ribs is homoplastic. Unribbed shells apparently have been derived from shells with broad, narrow, or threadlike ribs. The presence of broad ribs in CAE represents at least a two-step reversal. 25. Minute, wavy spiral striation may be generally distributed through the Haplotrematinae but its expression masked on ribbed shells. Its distribution in the group studied approximately parallels that of smooth shells, which is homoplastic. DISCUSSION Figure 3 presents the hypothesis of the phylogeny of the Haplotrematinae produced by this study. Five characters from five systems (radula, musculature, kidney, reproductive system, and shell) support the dichotomy between the group (ANC,((SPO,HYB),VOY)) and the rest of the subfamily. The former group consists of the species that Baker (1931) referred to the nominate ‘‘section’” of Ancotrema. Neither primary clique contains an apomorphy defining Ancotrema or distinguishing it from the hypothetical ancestor; the group is paraphyletic. A plausible alternative coding of character 24, with state A (coarse spiral striae cutting tops of radial ribbing) as apomorphous, would distinguish Ancotrema but is not com- patible with characters 6 and 18. Similarly, round-topped radial ribs wider than their interspaces could be apomorphous in Ancotrema. An alternative, less parsimonious interpretation, which would allow VOY to be grouped with (SPO,HYB) on the basis of these shell characters, is that migration of the penial retrac- tor muscle to the floor of the lung (character 6, states B, C) and elongation of the vagina (character 18, states B, C) are homoplastic. Obsolescence of radial ribs on the penult and body whorls distinguishes HYB from SPO. In conjunction with increased size, it is a noteworthy convergence upon VAN and occurs in the same region — the moist forests of the Pacific Northwest — where other cases of gigantism in land mollusks are known. Chromosome number, although known for only two species of the Haplotrematinae and, therefore, not entered in the compatibility analysis, supports the distinction between Ancotrema and Haplotrema. The haploid chromosome number of VAN is 30, that of SPO, 29 (Burch, 1965). The higher number is considered to be apomorphous. Four characters, all having to do with the reproductive system, distinguish (MIN VAN) from ((TRA,ALA),(KEE,(DUR, (CON,CAE)))). (MINVAN) is defined by the autapomorphy of a swollen vas deferens (character 8, state B). This group con- tains the type species of Ancomena (VAN, Haplotrema van- couverense) and is here recognized as a restricted subgenus Ancomena of Haplotrema. The other species included in Ancomena by Baker (1931), TRA, ALA, and KEE, are removed from that subgenus. TRA is distinguished from ALA by the absence of a cusp on the central tooth and by minor quan- titative differences in shell sculpture and genitalia. Ancotrema Haplotrema | | cet Haplotrema, es eee ANC SPO HYB VOY MIN VAN TRA ALA KEE DUR CON CAE 3(B),11(A), 15(A),22(A) 9(A),11(C), 10(A),20(B), 16(B),19(C) t 21(B) 1(A) 11(B),16(A),17(B) 4(B),6(C),13(B),20(A),24(B) 6(B),18(B) Fig. 3. Phylogenetic hypothesis of the Haplotrematinae, and genus/subgenus boundaries. Underlining denotes reversal; other con- ventions as in Fig. 1. 160 AMER. MALAC. BULL. 8(2) (1991) The remaining species are a rather heterogeneous lot, although with a clearly hierarchic set of relations. It is easy to see why Baker (1931) and Pilsbry (1946) resorted to three- layered classifications using the infrasubgeneric category “‘section,’ but not so easy to decide on a single set of rules for dividing this array into subgenera. (CAE,CON) is holophyletic, with three defining apomorphies. The subgenus name Geomene, based on CON, has priority over Greggiella, based on CAE. However, four characters distinguish CAE from CON. KEE is holophyletic and distinguished by five characters, including four autapomorphies. My decision has been to recognize no monotypic subgenera and to accept two heterogeneous subgenera (in addition to Ancomena) in Haplotrema — Haplotrema, sensu stricto, and Geomene. Table 3 presents a revised classification of the Haplotrematinae based on this analysis. PLACEMENT OF TAXA NOT ANALYZED catalinense. Differs from Haplotrema duranti principally in larger size (diameter 5.7-6.3 mm versus 5.2 mm). Fine and partly obsolete ribbing occurs also in several mainland popula- tions of H. d. continentis. Dissection of both H. catalinense and H. d. duranti is needed. Haplotrema s.s. costatum. Character of ribbing suggests relationship to Haplotrema caelatum, but shell sculpture is not diagnostic in subgenus Geomene. Subgenus uncertain. guadalupense. Differs from Haplotrema duranti main- ly in absence of ribbing and smooth surface. Haplotrema s.s. kendeighi. Probable sibling species of Haplotrema con- Table 3. Revised classification of the North American Haplotremat- idae. Superfamily Rhytidoidea Pilsbry, 1895 Family Haplotrematidae Baker, 1925 Subfamily Haplotrematinae Baker, 1925 Genus Haplotrema Ancey, 1881 Subgenus Haplotrema, sensu stricto H. (H.) duranti (Newcomb, 1864) H. (H.) a. duranti H. (H.) d. continentis Baker, 1931 H. (H.) catalinense (Hemphill In: W. G. Binney, 1890) H. (H.) guadalupense Pilsbry, 1927 H. (H.) keepi (Hemphill, 1890) H. (H.) alameda Pilsbry, 1930 H. (H.) transfuga (Hemphill in: W. G. Binney, 1890) Subgenus Geomene Pilsbry, 1927 H. (G.) concavum (Say, 1821) H. (G.) kendeighi Webb, 1951 H. (G.) caelatum (Mazyck, 1886) Subgenus Ancomena Baker, 1931 H. (A.) vancouverense (Lea, 1839) H. (A.) minimum (Ancey, 1888) Subgenus uncertain H. costatum Smith, 1957 Genus Ancotrema Baker, 1931 A. sportella (Gould, 1846) A. hybridum (Ancey, 1888) A. voyanum (Newcomb, 1865) cavum. Hubricht (1956) distinguished it from H. concavum by the color of the foot and by the large atrium, ‘‘at least twice the size of that of [H. concavum],” even in smaller animals. None of the reported characters associates H. kendeighi with any West Coast species of Haplotrema. Geomene. HISTORY OF THE HAPLOTREMATINAE Chambers (1987) calculated the average chromosomal rate of evolution for land pulmonates at 0.021 karyotypic changes per lineage per million yr. This figure is derived from the fossil record of land snail genera (Chambers, 1987), which (a) represents minimum maximum ages and (b) is far more fragmentary than that of marine genera. The ‘‘short- weighting” that results from these features is somewhat counterbalanced by the tendency for paleontological systematics — based, necessarily, on shell characters alone — to construe genera broadly and by the frequency of con- vergence in pulmonate shell form. After excluding Haplotrema (=Ancotrema + Haplotrema of this study) from Chambers’s data set to avoid circularity, and recalculating the rate of karyotypic changes based on a recent Cenozoic time scale (Berggren et a/., 1985), using mean rather than maximum ages for the fossils of an epoch or sub-epoch, the data suggest an approximate age of 42.5 million yr (late middle Eocene) for the Haplotrematinae. It must be stressed that this estimate is based solely on an average derived from a wide variety of land pulmonates. Fossil evidence indicates that during the middle Eocene a diverse land mollusk fauna including genera of present-day tropical distribution existed at middle latitudes in North America (Roth, 1984, 1988; Roth and Megaw, 1989). By the late middle Eocene many of these genera had begun to retreat southward and a latitudinal stratification of land mollusk faunas had become evident. The dichotomy between the temperate Haplotrematinae and its tropical out-group, the Austroselenitinae, could have arisen at this time. The phylogenetic hypothesis indicates an early dichotomy between Ancotrema and Haplotrema, with the species of Ancotrema retaining more characteristics of the hypothetical ancestor. Other trends include: (a) a tendency toward reduction of vaginal musculature as one moves toward the right side of the cladogram in figure 3; (b) a parallel tendency toward greater penial complexity; (c) migration of the origin of the penial retractor muscles from the columellar muscle bundle to floor of lung (presumably increasing mechanical efficiency); (d) parallel instances of gigantism in the Pacific Northwest; (e) size reduction southward along the Pacific Coast (partly, but not wholly, on islands) correlating with the presence/severity of a summer dry season; (f) a homoplastic continuation of the loss or reduction of cusps on the teeth of the radula. The dichotomy between the widespread eastern North American species, Haplotrema concavum, and a West Coast species (the highly derived H. caelatum) occurs higher on the tree (i.e. later in time) than the origins of all other major West Coast lineages. The vicariance was probably not related to the Laramide orogeny (which would require condensing most ROTH: NORTH AMERICAN HAPLOTREMATIDAE 161 of haplotrematine evolutionary history into the later Eocene) but to a later climatic event, such as the late Tertiary emergence of an arid environment in the American Southwest (Axelrod, 1979) or Pleistocene glacio-pluvial changes, which were relatively mild in coastal southern California (Johnson, 1977) but profound in the continental interior. Several other characteristically ‘‘eastern’’ taxa such as the genera Stenotrema Rafinesque, 1819 (Polygyridae), and Hendersonia Wagner, 1905 (Helicinidae), were present in the North American cordillera as recently as early (and possibly mid- dle) Miocene time (Berry, 1953; Roth and Emberton, unpub. data), but the timing of their extinction in the cordillera is uncertain. LECTOTYPE DESIGNATION FOR SELENITES VANCOUVERENIS FORMA HYBRIDA ANCEY, 1888 Three localities were cited in the original publication of Selenites vancouverensis forma hybrida Ancey, 1888: “Oregon, dans la region du Fleuve Columbia; Portland, Oregon (Dore); territoire de Washington.’ Type material was not specified. Baker (1931) designated Portland as the type locality. C. F. Ancey’s collection was dispersed by the shell dealer Geret (Dance, 1966). One syntype, purchased from Geret by S. Stillman Berry in the early years of this century, is now in the Santa Barbara Museum of Natural History. On a printed label of ‘‘Geret, Conchyliologiste/Naturaliste’’ an unknown hand gives the locality as Astoria, Oregon. A second hand has written in pencil, “TYPE”. The specimen is a mature shell, 24.2 mm in maximum diameter, 11.0 mm in height, with 6.2 whorls, agreeing with the original description but slightly lower-spired than Ancey’s figure. The specimen, SBMNH 35134 (Fig. 4), is here designated lectotype of Selenites van- couverensis forma hybrida. The type locality of the taxon is therefore Astoria, Clatsop County, Oregon, rather than Portland as designated by Baker (1931). Ancey (1888) stated that he was maintaining the name “hybrida’”’ under which J. H. Thompson had designated the taxon in his private collection. The name seems to have had even wider currency among amateur malacologists of the time. Hemphill (1890a) listed ‘‘Selenites concava var. hybrida’’ in asales catalog and, apparently unaware of Ancey’s (1888) proposal, described the same taxon as Selenites van- Fig. 4. Lectotype of Selenites vancouverensis forma hybrida Ancey, 1888, SBMNH 35134; top and basal views. Diameter 24.2 mm. couverensis var. hybrida (Hemphill, 1890b), likewise based on specimens from Astoria (Coan and Roth, 1987). ACKNOWLEDGMENTS | am grateful to Paul Scott and Eric Hochberg for help with type and other specimens from the S. S. Berry collection at the Santa Barbara Museum of Natural History. Tim Pearce loaned specimens from his collection. David R. Lindberg (University of California Museum of Paleontology) provided the CLINCH analysis. Messrs. Pearce and Lindberg read drafts of the manuscript. Walt Miller discussed the results with me. Ellen and George Moore collected Ancotrema sportella in the Corvallis area in response to my request. | thank the anonymous reviewer who provided the HENNIG86 maximum-parsimony analysis. LITERATURE CITED Ancey, C. F. 1888. Descriptions de mollusques terrestres. Le Naturaliste, sér. 2, 2:188-190. Ashlock, P. D. 1971. Monophyly and associated terms. Systematic Zoology 20:63-69. Axelrod, D. |. 1976. History of the coniferous forests, California and Nevada. University of California Publications in Botany 70:1-62. Axelrod, D. |. 1979. Age and origin of Sonoran Desert vegetation. California Academy of Sciences Occasional Papers No. 132:1-74. Baker, H. B. 1931. The land snail genus Hap/otrema. Proceedings of the Academy of Natural Sciences of Philadelphia 82:405-425, pls. 33-35. Baker, H. B. 1941. Some Haplotrematidae. Nautilus 54:130-136, pl. 9. Berggren, W.A., D. V. Kent, J. J. Flynn and J. A. Van Couvering. 1985. Cenozoic geochronology. Geological Society of America Bulletin 96:1407-1418. Berry, S. S. 1953. A terrestrial molluscan faunule from the Miocene of Montana. Annual Report of the American Malacological Union for 1953:23 (abstract). Burch, J. B. 1965. Chromosome numbers and systematics in euthyneuran snails. Proceedings of the First European Malacological Congress: 215-241. Chambers, S. M. 1987. Rates of evolutionary change in chromosome numbers in snails and vertebrates. Evolution 41:166-175. Coan, E. and B. Roth. 1987. The malacological taxa of Henry Hemphill. Veliger 29:322-339. Dance, S. P. 1966. Shell Collecting, An Illustrated History. Berkeley, University of California Press. 344 pp. Duncan, T. 1980. A cladistic analysis of the Ranunculus hispidus com- plex. Taxon 29:441-454. Estabrook, G. F. 1972. Cladistic methodolgy: a discussion of the theoretical basis for the induction of evolutionary history. An- nual Review of Ecology and Systematics 3:427-456. Estabrook, G. F. 1978. Some concepts for the estimation of evolu- tionary relationships in systematic botany. Systematic Botany 3:146-158. Farris, J. S. 1983. The logical basis of phylogenetic analysis. /n: Ad- vances in Cladistics, Vol. 2. N. |. Platnick and V. A. Funk, eds. pp. 7-36, Columbia University Press, New York. Farris, J. S. 1988. HENNIG86 reference, version 1.5. Port Jefferson Station, New York (published by author). Gratacap, L. P. 1901. Catalogue of the Binney and Bland Collection of the terrestrial air-breathing mollusks of the United States and Territories in the American Museum of Natural History, with enumeration of types and figured specimens, and sup- plementary notes. Bulletin of the American Museum of Natural 162 AMER. MALAC. BULL. 8(2) (1991) History 14:335-403. Hemphill, H. 1890a. Catalogue of North American shells collected and for sale by Henry Hemphill. San Diego, privately printed. 21 pp. Hemphill, H. 1890b. New varieties of western land shells. Nautilus 4:41-43. Henderson, J. 1936. The non-marine Mollusca of Oregon and Washington — supplement. University of Colorado Studies 23:251-280. Hennig, W. 1966. Phylogenetic Systematics. Urbana, University of Illinois Press. 263 pp. Hubricht, L. 1956. Haplotrema kendeighi Webb. Nautilus 69:126-128. Hubricht, L. 1985. The distributions of the native land mollusks of the eastern United States. Fieldiana Zoology, new series, no. 24. viii + 191 pp. Johnson, D. L. 1977. The late Quaternary climate of coastal California: evidence for an ice age refugium. Quaternary Research 8:154-179. Kluge, A. G. 1989. A concern for evidence and a phylogenetic hypothesis of relationships among Epicrates (Boidae, Serpentes). Systematic Zoology 38:7-25. Le Quesne, W. J. 1969. A method of selection of characters in numerical taxonomy. Systematic Zoology 18:201-205. Le Quesne, W. J. 1972. Further studies based on the uniquely derived character concept. Systematic Zoology 21:281-288. Le Quesne, W. J. 1982. Compatibility analysis and its applications. Zoological Journal of the Linnean Society 74:267-275. Meacham, C. A. 1980. Phylogeny of the Berberidaceae with an evalua- tion of classifications. Systematic Botany 5:139-172. Meacham, C. A. 1981. A manual method for character compatibility analysis. Taxon 30:591-600. Page, R. D. M. 1987. Graphs and generalized tracks: quantifying Croizat’s panbiogeography. Systematic Zoology 36:1-17. Pilsbry, H. A. 1946. Land Mollusca of North America (north of Mex- ico). Monograph No. 3, Academy of Natural Sciences of Philadelphia 2(1):i-viii, 1-520. Porter, C. A. 1965. Comparison of genitalia of two sympatric species of Haplotrema. Nautilus 79:19-23. Porter, C. A. 1968. Comparative morphology of the reproductive tracts of some of the Pacific Northwest pulmonates and its use in their taxonomy. Northwest Science 42:71-79. Pratt, W. L., ed. In preparation. Checklist of the non-marine Mollusca of the United States and Canada. Council of Systematic Malacologists. Roth, B. 1981. Distribution, reproductive anatomy, and variation of Monadenia troglodytes Hanna and Smith (Gastropoda: Pulmonata) with the proposal of a new subgenus. Proceedings of the California Academy of Sciences 42:379-407. Roth, B. 1984. Lysinoe (Gastropoda: Pulmonata) and other land snails from Eocene-Oligocene of Trans-Pecos Texas, and their paleoclimatic significance. Veliger 27:200-218. Roth, B. 1988. Camaenid land snails (Gastropoda: Pulmonata) from the Eocene of southern California and their bearing on the history of the American Camaenidae. Transactions of the San Diego Society of Natural History 21:203-220. Roth, B. and P. K. M. Megaw. 1989. Early Tertiary land mollusks (Gastropoda: Pulmonata) from Sierra Santa Eulalia, Chihuahua, Mexico, and the origins of the North American arid- land mollusk fauna. Malacological Review 22:1-16. Smith, A. G. 1957. Snails from California Caves. Proceedings of the California Academy of Sciences, ser. 4, 29:21-46. Solem, A. 1974. Patterns of radular tooth structure in carnivorous land snails. Veliger 17:81-88. Strauch, J. G., Jr. 1984. Use of homoplastic characters in compatibility analysis. Systematic Zoology 33:167-177. Watson, H. 1915. Carnivorous slugs of South Africa. Annals of the Natal Museum 3:107-256. Webb, G. R. 1943. The mating of the landsnail Haplotrema concavum (Say). American Midland Naturalist 30:341-345. Webb, G. R. 1951. A new landsnail, Haplotrema concavum kendeighi (Mollusca: Pulmonata). 7ransactions of the Kansas Academy of Science 54:78-82. Webb, G. R. 1961. The phylogeny of American land snails with emphasis on the Polygyridae, Arionidae, and Ammonitellidae. Gastropodia 1:31-45, 47-49, 51. Whittaker, R. H. 1961. Vegetation history of the Pacific Coast states and the ‘‘central’”’ significance of the Klamath Region. Madrono 16:5-22. Wilson, E. O. 1965. A consistency test for phylogenies based on con- temporaneous species. Systematic Zoology 14:214-220. Zilch, A. 1959-1960. Gastropoda, Teil 2, Euthyneura. Handbuch der Palaozoologie 6 (2):1-400 (1959); 401-834 (1960). Note added after revision. One anonymous reviewer of this paper performed a maximum-parsimony analysis (Farris, 1983; Kluge, 1989) of this data set using the program HENNIG86 (Farris, 1988). Four equally (and maximally) parsimonious trees were derived. One tree was isomorphous with the tree shown in figure 3; a second dif- fered in having the branch leading to KEE and that leading to (TRA,ALA) reversed. Two other trees contained the grouping (((MIN,VAN),KEE),((TRA,ALA),(DUR,(CON,CAE)))). A classification based on this additional analysis might include KEE in the subgenus Ancomena and treat (TRA,ALA) as a group coordinate in rank with (DUR,(CON,CAE)). Date of manuscript acceptance: 20 June 1990 ROTH: NORTH AMERICAN HAPLOTREMATIDAE 163 APPENDIX A. LIST OF CHARACTERS, CHARACTER STATES, AND CHARACTER STATE TREES Character No. 1. Right ocular retractor: in penioviducal angle (A), free from genitalia (B); B — A. Character No. 2. External genital orifice: close to base of right om- matophore (A), distant from right ommatophore (B); B — A. Character No. 3. Radula: with several lateral teeth bicuspid (A), with one lateral tooth bicuspid (B); A — B. Character No. 4. Central tooth of radula: with multiple cusps (A), with reduced number of cusps (0-1) (B); A — B. Character No. 5. Central tooth: with cusps (A), lacking cusps (B); A — B. Character No. 6. Penial retractor muscle origin: on left side of col- umellar muscle (A), on floor of lung with strands from columellar bun- dle (B), on floor of lung (C); A — B — C. Character No. 7. Penial retractor muscle insertion: at summit of penis with vas deferens entering laterally (A), on vas deferens, which enters penis apically (B); A — B. Character No. 8. Vas deferens: slender throughout (A), with swollen, epiphallus-like portion (B); A — B. Character No. 9. Penis: with flagelloid caecum (A), without (B); B — A. Character No. 10. Penis: with diverticulum (A), without (B); B — A. Character No. 11. Penial chamber wall: without papillae (A), with sharp, thornlike papillae (B), with blunt papillae (C); A — B — C. Character No. 12. Apical penial chamber: well differentiated (A), not well differentiated (B); A — B. Character No. 13. Principal penial chamber: differentiated into apical and basal parts (A), not differentiated thus (B); A — B. Character No. 14. Penial sheath: present (A), absent (B); B — A. Character No. 15. Penial chamber: with apical papilla (A), without (B); B — A. Character No. 16. Atrium: with ‘‘stimulator’’ (A), without (B); B — A. Character No. 17. Vagina: with shining muscular collar (A), with sphincteric thickening (although not a muscular collar) (B), without collar or thickening (C); A — B — C. Character No. 18. Vagina: extremely short (i.e., spermathecal duct origin directly above atrium) (A), somewhat less than to approximately equal to length of penis (B), 1.5 times length of penis (C); A — B — C. Character No. 19. Base of spermathecal duct: weakly dilated (A), strongly dilated (B), with a dual dilation (C); B — A — C. Character No. 20. Length of kidney: 1.5 times its base (A), approx- imately 2 times its base (B), over 2 times its base (C); C — B — A. Character No. 21. Length of pericardium: approximately 24 length of kidney (A), approximately 2 length of kidney (B); A — B. Character No. 22. Embryonic shell: with radial riblets on last /2 whorl (A), smooth throughout (B); B — A. Character No. 23. Shell: with distinct, round-topped radial ribs as broad or broader than interspaces (A), with round-topped radial ribs narrower than interspaces (B); with fine, threadlike to laminar radial ribs (C), without ribs (except for low undulations of shell surface) (D); D-—-A-—-B-C. Character No. 24. Shell: with coarse spiral striae cutting tops of radial ribbing (A), without (B); A — B. Character No. 25. Shell: with minute, close, wavy spiral striae (A), without (B); A — B. PRESENT STATUS OF THE MICROMOLLUSKS OF NORTHERN SONORA, MEXICO EDNA NARANJO-GARCIA ESTACION DE BIOLOGIA CHAMELA-DEPARTAMENTO, DE ZOOLOG/A INSTITUTO DE BIOLOGIA, UNIVERSIDAD NACIONAL AUTONOMA DE MEXICO, APARTADO POSTAL 70-153, MEXICO, D.F. 04510, MEXICO ABSTRACT Information about the terrestrial micromollusk fauna of Sonora State is compiled. Fifteen genera and 22 species are reported to occur. Preliminary field work has added new locality records, as well as two species new to the state: Euconulus fulvus (Muller, 1774), and Lamellaxis sp. There are seven widely distributed species in the state. Thysanophora hornii (Gabb, 1866) is very well represented. The distribution of Gastrocopta ashmuni (Sterki, 1898) enlarges the Southwestern Molluscan Province southward to central Sonora. Karolus consobrinus primus De Folin, 1870, a Mid-western Mexican Pro- vince species, has affinities to the south and penetrates south and central Sonora. RESUMEN Se conjunta la informacion sobre la micro malacofauna terrestre del Estado de Sonora, se han registrado 15 géneros y 22 especies en el estado. Trabajo de campo de prospeccidn ha permitido anadir nuevos registros de localidades, lo mismo que dos nuevas especies para el estado: Euconulus fulvus (Muller, 1774), y Lamellaxis sp. Hay siete especies ampliamente distribuidas. Thysanophora hornii (Gabb, 1866) es la mejor representada. Gastrocopta ashmuni (Sterki, 1898) extiende hacia el sur a la Provincia Malacoldgica del Suroeste, en el centro de Sonora, Karolus consobrinus primus De Folin, 1870, es un elemento de la Provincia del Oeste Medio de México que penetra el sur y centro de Sonora. The micromollusks of Mexico have received little at- tention. Because of their small size they are difficult to locate and to collect. Recent explorations have increased our knowledge of the malacofauna of Sonora. Early works, which include some locality records from Sonora deal primarily with faunas of other regions. Other information is widely scattered throughout the literature. Dall (1896), in his report for the In- ternational Boundary Commission, included microgastropods from various places in Mexico except Sonora. Drake (1953, 1956), who conducted some surveys in the state, added five new records. Furthermore, he mentioned an almost completed monograph on the non-marine mollusks of the region and wrote of a collection placed in the Instituto de Biologia, México, but | have not been able to locate either. Drake did not give a faunal listing for Sonora. In a paper on terrestrial snails from northern Mexico, Pilsbry (1953) cited eight records from Sonora. Branson et a/. (1964) wrote a note on new localities of snails, which included Vallonia perspectiva Sterki, 1893 as a new record for the state. Bequaert and Miller (1973) dis- cussed mainly the Arizona molluscan fauna from a biogeographical viewpoint. Their work included the Sonoran fauna shared with Arizona. An important contribution of that work was the updating of the nomenclature of the included taxa. Habitat information is scant and scattered, although some highlights are given in the literature, e.g. Pilsbry (1948) discussed the general features of the environment where pupillids are found. Bequaert and Miller (1973) gave additional details for several species. The other works cited above did not discuss habitat. The purpose of this study is to present information on the distribution of microgastropods of Sonora, based on material collected from 1983 to 1988. | have tried to review most literature references, provide a point of departure for other investigations, and contribute toward the condensation of the widely dispersed information concerning the Mexican terrestrial mollusk fauna, in general. The state of Sonora is located in the northwest of Mex- ico, adjacent to the Gulf of California; approximately 26° to 32°N and 108° to 115°W. It is bordered on the west by the Gulf of California, to the north by Arizona, to the east by Chihuahua state and to the south by Sinaloa state. Plains and mountains American Malacological Bulletin, Vol. 8(2) (1991):165-171 165 166 AMER. MALAC. BULL. 8(2) (1991) show several biotic communities that Brown (1982) summa- rized for the region. The largest biome in Sonora is the Sonoran Desert which occupies most of the western half of the state. The southern and southeastern regions bear the Sinaloan Thornscrub. The Sinaloan Deciduous Forest extends from the south to the north in and a narrow strip, bordered to the east by Madrean Evergreen Woodland (Gentry, 1982), and to the north by the Chihuahuan Desert and the Madrean Evergreen Woodland. METHODOLOGY While searching for Sonorella (macromollusks) (Naranjo-Garcia, 1988) several microgastropods were found coincidentally; therefore, the majority of the samples are from central and north Sonora sites inhabited by this genus (from the Madrean Evergreen Woodland; Lower Colorado River Subdivision of the Sonoran Desert, Semidesert Grassland, to the Chihuahuan Desertscrub). In addition, an exploration to the south yielded several samples from that region. Samples were taken from the same Sonorella locality or a place close by. Leaf litter, humus, soil accumulations, and drift Gastrocopta cochisensis G. contracta G, pilsbryana Vertigo ovata Cecilioides sp. Lamellaxis sp. Helicodiscus eigenmanni Euconulus fulvus Thysanophora proxima Fig. 1. Species with a single record for Sonora, numbers in the map are my localities (see Appendix). Localities of species shown in this map are from the following authors: Gastrocopta contracta Arroyo San Rafael, San Bernardo (Pilsbry, 1953:161); G. cochisensis Arroyo 8 km S from Guaymas (Pilsbry, 1953:162); G. pilsbryana Brushy area in mountains 29 mi. S of Cumpas Hwy 10 (Branson et a/., 1964:104); Vertigo ovata Rio Sonoyta at Sonoyta (Bequaert and Miller, 1973:183); Helicodiscus eigenmanni San Bernardino (Pilsbry, 1948:630); Thysanophora proxima Rio Bavispe 21 mi. S Agua Prieta (Branson et al., 1964:103). wash material were examined for micromollusks, then sampled. In the laboratory, soil samples were sifted by a series of sieves (3.36 mm, 1.68 mm, and 0.84 mm) to separate the different sizes of shells. As a result of these investigations, new localities are added for the microsnail fauna of Sonora. Figures 1 to 5 pre- sent the distribution of the species known for Sonora. RESULTS AND DISCUSSION A total of 15 genera and 22 species have been documented for Sonora (Table 1). The primary source of data for this list of species is from the papers of Pilsbry (1948 and 1953) and Bequaert and Miller (1973); some localities shown in the maps are from those authors and from Branson et al. (1964). The systematics follows the classification suggested by Solem (1978), Burch and Jung (1988), and Abbott and Boss (1989). Thirty-six new localities have been added to the records (see Appendix). The species taken in this survey have all been record- ed previously from Sonora, except for two taken in my field samples that are new species records for the state: Lamellaxis sp. and Euconulus fulvus (Pilsbry and Ferriss, 1906). FAMILY PUPILLIDAE Gastrocopta (Immersidens) ashmuni (Sterki, 1898) (Fig. 4) is extensively distributed in Arizona, New Mexico, trans-Pecos Texas, and Chihuahua (Bequaert and Miller, 1973). It was considered by those authors as representative ( i ah 2\ 2 2 | 3 3 x a © Pupoides albilabris @ Glyphyalinia indentata @ Hawaiia minuscula Fig. 2. Distribution of Pupoides albilabris, Glyphyalinia indentata, and Hawaiia minuscula in Sonora. NARANJO: MICROMOLLUSKS OF SONORA 167 \ { \ 26) Ber fe) \ } ey aay \A ) 25 \ octezuma _>—~ J gf feo y, é HF mosilio oO Gastrocopta pellucida e@ Chaenaxis tuba O Pupisoma minus a Karolus consobrinus Fig. 3. Distribution of Gastrocopta pellucida, Chaenaxis tuba, Pupisoma minus, and Karolus consobrinus primus in Sonora. * wey fio V0 ee { \ 7) O26 ig. Mectezuma eS a Hermosillo © Gastrocopta ashmuni ® G. dalliana O G. d. bilamellata 4 G. cristata Fig. 4. Distribution of Gastrocopta ashmuni, G. dalliana, G. d. bilamellata, and G. cristata in Sonora. a octezuma —-$—. @ tf ae a ee a Sa @ \dllonia perspectiva ® Helicodiscus singleyanus @ Thysanophora hornii Fig. 5. Distribution of Vallonia perspectiva, Helicodiscus singleyanus, and Thysanophora hornii in Sonora. of Arizona, New Mexico, trans-Pecos Texas, Chihuahua and Sonora; it is known from northern Sonora. Gastrocopta |. dalliana dalliana (Sterki, 1898) (Fig. 4) is broadly distributed in east-central Arizona, also NW Chihuahua and Baja California Sur (Bequaert and Miller, 1973). It is apparently widely distributed throughout Sonora. Gastrocopta |. dalliana bilamellata (Sterki and Clapp, 1909) (Fig. 4) occupies south-west and south-central Arizona. It is considered as representative of the very arid part of SW Arizona and NW Sonora (Bequaert and Miller, 1973). Its distribution seems to be over the west of Sonora with two records to the east. G. d. dalliana and G. d. bilamellata have not been found to co-occur in the state. Further collections are needed in order to obtain a better understanding of these subspecies ranges. Bequaert and Miller (1973) showed that the two subspecies overlap in central Arizona and suggested interbreeding among populations. Gastrocopta (Immersidens) cochisensis (Pilsbry and Ferriss, 1910) (Fig. 1) is known from a single locality in Sonora (Pilsbry, 1953). With a broad distribution in southeastern Arizona (Bequaert and Miller, 1973), it has been recorded also from Animas Peak, Hidalgo County, New Mexico (Metcalf, 1976). Gastrocopta (Gastrocopta) pellucida (Pfeiffer, 1841) (Fig. 3) is widespread in Texas and Florida, and scattered throughout the southeastern United States (Hubricht, 1985). It occupies SW United States from Texas to SE California, as well as Mexico and Central America south to Panama, and the Antilles (Bequaert and Miller, 1973). Those authors sug- 168 AMER. MALAC. BULL. 8(2) (1991) gested that Gastrocopta pellucida is a neotropical native that advanced from the south, and Mexico could be its center of dispersal. Gastrocopta (G.) cristata (Pilsbry and Vanatta, 1900) (Fig. 4) is found mainly in central and western Texas, but ranges from Oklahoma to Arizona (Hubricht, 1985). It is com- mon in southern Arizona, but is also known from Nebraska, Kansas and New Mexico (Bequaert and Miller, 1973). Its widely separated localities suggest an extensive range in Sonora. There is one record of Gastrocopta (Albinula) contrac- ta (Say, 1822) (Fig. 1), for Sonora (Pilsbry, 1953). It is widely distributed in the eastern United States (Hubricht, 1985) and also occurs in western, central, eastern and north-eastern Mexico (Bequaert and Miller, 1973). Gastrocopta (Vertigopsis) pilsbryana (Sterki, 1890) (Fig. 1) is known from a single record for the state (Branson et al., 1964). Pilsbry (1948) considered it one of the most com- mon species of the highlands of New Mexico and Arizona. It also occurs in southern Utah, trans-Pecos Texas, and northern Mexico (Bequaert and Miller, 1973). Chaenaxis tuba (Pilsbry and Ferriss, 1906) (Fig. 3) is broadly distributed in Sonora. The range of Chaenaxis tuba as shown by Bequaert and Miller (1973) is extended by several kilometers; to the west (present in the south and north ends of Sierra El Viejo range), to the south close to Sierra Alamos, and to the east near the border of Chihuahua state (found at Cerro Prieto). Bequaert and Miller (1973) regard this species as one of the most distinctive snails of the southwestern arid fauna. Pupisoma minus Pilsbry, 1920 (Fig. 3) is a synonym of P macneilli (Clapp, 1918) according to Hubricht (1985). P macneilli occurs mainly in Florida and the southeastern United States. The species has been identified by shell material on- ly (Pilsbry, 1953, this account), and for that reason | have decided to continue using P minus for Sonora until live material becomes available for further studies. Pupoides (Pupoides) albilabris (Adams, 1841) (Fig. 2) is widely distributed in the eastern United States (Hubricht, 1985), in South and North Dakota, Utah and Arizona (Bequaert and Miller, 1973). In Sonora a scattered distribution suggests that it is widespread throughout the state. Vertigo (Vertigo) ovata (Say, 1822) (Fig. 1) is widely distributed in the eastern United States, rare in Arizona, and perhaps also in Sonora. There is one record for the state (Bequaert and Miller, 1973). FAMILY VALLONIDAE Vallonia perspectiva Sterki, 1893 (Fig. 5) has a dis- persed distribution in eastern United States (Hubricht, 1985). It has an extended distribution in the highlands of Arizona, New Mexico and trans-Pecos Texas (Bequaert and Miller, 1973). The scattered localities of the species in Sonora sug- gest a large latitudinal range. FAMILY FERUSSACIIDAE Karolus consobrinus primus De Folin, 1870 (Fig. 3) is a snail mainly from eastern Mexico; for example Veracruz, from alluvial deposits of Rio Antigua (von Martens 1890-1901), Tampico, Tamaulipas from river debris (Pilsbry, 1907), Veracruz Table 1. Species found in Sonora and localities. Numbers represent my localities (see Appendix). Asterisks indicate taxa shared with Arizona. SPECIES LOCALITIES FAMILY PUPILLIDAE Gastrocopta (Immersidens) ashmuni 5, 10, 11, 16, 28,35 * G. (I.) dalliana 10, 19, 20, 21, 34 : G. (I.) d. bilamellata 28, 35 7 G. (I.) cochisensis 3 G. (Gastrocopta) pellucida 8, 11, 15, 22, 24, : 25, 26, 27, 35 G. (G.) cristata * G. (Albinula) contracta G. (Vertigopsis) pilsbryana i Chaenaxis tuba 16, 17, 19, 21, 22, 7 23, 27, 33 Pupisoma minus 28 Pupoides albilabris 19, 27, 34 ; Vertigo (Vertigo) ovata ii FAMILY VALLONIIDAE Vallonia perspectiva : FAMILY FERUSSACIIDAE Karolus consobrinus primus 32 Cecilioides sp. 9 FAMILY SUBULINIDAE Lamellaxis (Allopeas) gracilis Z L. sp. 26 FAMILY SPIRAXIDAE Pseudosubulina sp. FAMILY HELICODISCIDAE Helicodiscus singleyanus 28 H. eigenmanni a FAMILY VITRINIDAE Glyphyalinia indentata 3, 4, 5, 7, 8, 10, ‘i 11, 15, 16, 18, 23, 26, 28, 35, 36 Hawaiia minuscula 11 i FAMILY EUCONULIDAE Euconulus fulvus 26 . FAMILY THYSANOPHORIDAE Thysanophora hornii 1, 2, 5, 6, 8, 9, 11, rs 12, 13, 14, 15, 16, 17, 19, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 33, 35 T. proxima and Necaxa, Puebla (Baker, 1930). There are two records of the species for Sonora: Pilsbry (1953) and this account. Cecilioides sp. (Fig. 1) was cited by Bequaert and Miller (1973) to be present in Sonora; however they did not specify a locality. | have a sample from Sierra Alamos, south Sonora. It is an element of a tropical family. FAMILY SUBULINIDAE Lamellaxis (Allopeas) gracilis (Hutton, 1834) was record- ed for the state by Drake (1953:156) who mentioned that it was found ‘‘as far north as Hermosillo’ without giving precise NARANJO: MICROMOLLUSKS OF SONORA 169 localities. It is found in Arizona, and was introduced in the southern United States (Bequaert and Miller, 1973). Lamellaxis sp. (Fig. 1) from southern Sonora differs from Lamellaxis gracilis by having widely spaced transverse ribs and a curved outer lip. The lip of L. gracilis is straight. Both show a high spire. This species is known from a single locality, Sierra Alamos. FAMILY SPIRAXIDAE Drake (1953) added Pseudosubulina to the species list for Sonora with a question mark, and no precise locality. The genus has been recorded in eastern, southeastern and southern Mexico (von Martens, 1890-1901). FAMILY HELICODISCIDAE Helicodiscus (Helicodiscus) eigenmanni Pilsbry, 1890 (Fig. 1) has once been recorded in the state by Pilsbry (1948). Bequaert and Miller (1973) considered it as a Southwestern snail, but it is also found in the south-central United States, and other scattered records exist for Colorado and South Dakota, Chihuahua and Puebla (central Mexico). Helicodiscus (Hebetodiscus) singleyanus (Pilsbry, 1889) (Fig. 5) is found in several localities in the eastern United States (Hubricht, 1985), west to southeastern South Dakota, Colorado, Kansas, Oklahoma, and Texas. In Mexico there are records from Tamaulipas and Sonora (Bequaert and Miller, 1973). It has two records in Sonora which could suggest a widespread distribution. FAMILY VITRINIDAE Glyphyalinia indentata (Morelet, 1851) (Fig. 2) is widespread latitudinally in eastern Sonora. It is also broadly distributed in the central-eastern United States (Hubricht, 1985), west to Oklahoma, Texas, New Mexico, Utah, and Arizona, and from Mexico to southern Guatemala (Bequaert and Miller, 1973). Hawaiia minuscula (Binney, 1841) (Fig. 2) is extensive- ly distributed in the eastern United States (Hubricht, 1985). It could be the commonest snail of Arizona. It is found from North America to Central America and the Antilles. Nowadays it is almost cosmopolitan due to introductions (Bequaert and Miller, 1973). FAMILY EUCONULIDAE Euconulus fulvus (Miller, 1774) (Fig. 1) was not known previously from Sonora. This account includes a single record from Sierra Alamos. It is a Holoarctic snail, and it is found over most of the United States (Bequaert and Miller, 1973). FAMILY THYSANOPHORIDAE Thysanophora hornii (Gabb, 1866) (Fig. 5) is the com- monest microgastropod of Sonora. It is also found in Arizona, New Mexico, and southern Texas (Hubricht, 1985). It is main- ly known from the northern states of Mexico, to Sinaloa and Jalisco (Bequaert and Miller, 1973). These authors regarded T. hornii as an atypical Neotropical invader. Thysanophora proxima Pilsbry, 1899 (Fig. 1) has one record for the state by Branson, et a/. (1964). This seems to be the northest locality since it was described from Uruapan, Michoacan (Pilsbry, 1899 and 1903) in west-central Mexico. Pilsbry gave other localities in Michoacan state. Thysanophora hornii is the most widely distributed species in the state, followed by Glyphyalinia indentata, Gastrocopta pellucida and Chaenaxis tuba. It should be noted that several species, such as Gastrocopta contracta, G. cochisensis, G. pilsbryana, Vertigo ovata, Thysanophora proxima, and Helicodiscus eigenmanni, known from previous publications, as well as three species, Cecilioides sp., Euconulus fulvus and Lamellaxis sp., from my field samples, are known only from single records (Fig. 1). Other specimens display an apparently spotty distribu- tion, e.g. Gastrocopta cristata (Fig. 2), Vallonia perspectiva (Fig. 5), Helicodiscus (Hebetodiscus) singleyanus (Fig. 5), and Pupisoma minus (Fig. 3). Nevertheless, the records are so scattered as to suggest a possibly wider distribution. The present distribution of Gastrocopta (Immersidens) ashmuni (Fig. 4) indicates an extension of the Southwestern Molluscan Province to the south, while Karolus consobrinus primus (Fig. 3), indicates an extension of a Mid-western Mex- ican Province to the north. The available information does not yet allow the identification of a biotic community with particular significance for the distribution of microsnails in Sonora. The new localities for the Sonoran fauna represent an additional contribution to the regional molluscan biogeography as discussed by Bequaert and Miller (1973), by supporting the Southwestern Molluscan Province suggested by Henderson (1931), and providing evidence that it extends farther south than Henderson indicated. The records of snails from the southern part of Sonora clearly represent popula- tions with affinities to a biogeographical zone containing tropical components, such as species of the families Ferussaciidae, Subulinidae and perhaps Spiraxidae. The micromolluscan fauna of Sonora shows a strong affinity with the Southwestern Molluscan Province of Hender- son (1931), with an additional weak introduction from the south, from what could be called a Mid-western Mexican Pro- vince to be better understood when more information becomes available. More field work needs to be done in the state in order to reach a more comprehensive conclusion. ACKNOWLEDGMENTS | wish to thank Dr. Walter B. Miller, James Hoffman, Ronnie Sidner and Diana Warr for their companionship in the field, and Teresa Olivera and Oscar J. Polaco from the Laboratorio de Prehistoria, In- stituto Nacional de Antropologjia e Historia for allowing me to utilize their installations in the identification of some material. Ricardo Ayala, Tim A. Pearce, and two anonymous reviewers gave various ideas. LITERATURE CITED Abbott, R. T. and K. J. Boss, eds. 1989. A Classification of the Living Mollusca. American Malacologists, Inc. Melbourne, Florida. 189 pp. Baker, H. B. 1930. Mexican mollusks collected for Dr. Bryant Walker in 1926. Part Il. Auriculidae, Orthurethra, Heterurethra and Aulacopoda. Occasional Papers Museum of Zoology, Univer- sity of Michigan 220:1-46. 170 AMER. MALAC. BULL. 8(2) (1991) Bequaert, J. C. and W. B. Miller. 1973. The mollusks of the Arid Southwest, with an Arizona Check list. The University of Arizona Press, Tucson. 271 pp. Branson, B. A., C. J. McCoy and M. E. Sisk. 1964. Notes on Sonoran gastropods. Southwestern Naturalist 9(2):103-104. Brown, D. E., ed. 1982. Biotic Communities of the American Southwest-United States and Mexico. Desert Plants Special Issue 4(1-4), 342 pp. Burch, J. B. and Younghun Jung. 1988. Land snails of The Univers- ity of Michigan Biological Station Area. Walkerana 3(9), 177 pp. Dall, W. H. 1896. Report on the mollusks collected by the Interna- tional Boundary Commission of the United States and Mex- ico, 1892-1894. Proceedings of the United States National Museum 19:333-379. Drake, R. J. 1953. Study of the species (and distributions) of the non- marine mollusk fauna of Sonora. American Philosophical Socie- ty Year Book for 1952:154-157. Drake, R. J. 1956. Further study of the species of the nonmarine mollusks fauna of Sonora and the preparation of a monograph on the species of landshells in particular of Sonora. American Philosophical Society Year Book for 1955:131-132. Gentry, H. S. 1982. Sinaloan Deciduous Forest /n: Biotic Communities of the American Southwest-United States and Mexico. Brown, D. E., ed. pp. 73-77. Desert Plants Special Issue 4(1-4). Henderson, J. 1931. Molluscan Provinces in the western United States. University of Colorado Studies 18(4):177-186. Hubricht, L. 1985. The distribution of the native land mollusks of the eastern United States. Fieldiana Zoology new series No. 24, 191 pp. Martens, E. von. 1890-1901. Biologia Centrali-Americana. Terrestrial and Fluviatile Mollusca. London. i-xxviii + 706 pp. Metcalf, A. L. 1976. First records of two species of land snails in New Mexico. Southwestern Naturalist 21(3):399-414. Naranjo-Garcia, E. 1988. Systematics and biogeography of the Helminthoglyptidae of Sonora. Doctoral Dissertation. Dept. of Ecology and Evolutionary Biology. University of Arizona. Tuc- son, Arizona. 105 pp. Pilsbry, H. A. 1899. Descriptions of new species of Mexican land and fresh-water mollusks. Proceedings of the Academy of Natural Sciences of Philadelphia 1899:391-402. Pilsbry, H. A. 1903. Mexican land and freshwater mollusks. Pro- ceedings of the Academy of Natural Sciences of Philadelphia 1903:761-789. Pilsbry, H. A. 1907. Descriptions of new Mexican land shells. Nautilus 21(3):26-29. Pilsbry, H. A. 1948. Land Mollusca of North America (North of Mex- ico). Academy of Natural Sciences of Philadelphia, Monograph No. 3. Part 2:i-xlviit+ 521-1113. Pilsbry, H. A. 1953. Inland Mollusca of Northern Mexico. Il Urocop- tidae, Pupillidae, Strobilopsidae, Valloniidae and Cionellidae. Proceedings of the Academy of Natural Sciences of Philadelphia 105:133-167. Solem, A. 1978. Classification of the land Mollusca. /n: Fretter, V. and Peake, eds. 49-97 pp. Pulmonates Vol. 2A. Date of manuscript acceptance: 17 April 1990 NARANJO: MICROMOLLUSKS OF SONORA Appendix KEY OF LOCALITIES: 1. on 10. 11. 12. 13. Cerro Gallardo, ca. 1.5 km due S of Rancho Gallardo buildings, in north-facing rhyolite rockpiles, 31° 17.8’N, 109° 23.4’ W, elev. 1500 m. . At Rancho Seco, along Carbo - Rayon road. 29° 41’N, 110° 40.1’ W. . Along road to Microondas installations, Sierra Mariquita ca. 40 km W Cananea at jct road - Hwy 2, 319 03.5’ N, 110° 22.5’ W, elev. ca. 2000 m. . El Quince, ca. 7.8 miles by road W Cananea at El Quince, Sierra Elenita, 29° 40.1’ N, 110° 47.5’ W. . Sierra Las Minitas, ca. 2 km due SE of Rancho El Jucaral buildings; in rhyolite rockslide, 31° 11.1’ N, 109° 04.7’ W, elev. ca. 1400 m. . 5 miles South of Cucurpe, along Rio San Miguel Horcacitas, on N facing slope, 30° 15’ N, 110° 43.5’W, elev. ca. 830 m. . Sierra San Ignacio, ca. 2 km E from El Naranjo, 27° 14’ 51.8” N, 108° 45’ 38” W. . Sierra San Ignacio ca. 1 km E Rancho Agua Salada, 27° 15’ 03” N, 108° 46’ 15” W, elev. ca. 600 m. . Sierra Alamos, Arroyo Las Piedras, S Alamos. 26° 59’ 22” N, 108° 57’ 22” W. At S edge of Sierra El Pinito ca. 10.1 km W Aribabi on road Imuris- Cananea (Hwy 2), N facing limestone slide. 30° 52.7’ N, 110° 42.7’ W. High, N facing rockslide on slope above arroyo, along Rayon- Cerro de Oro road, at ca. 4.2 miles from Cerro de Oro. 29° 38.9’ N, 110° 36.6’ W. Ravine along road from Rayon to Cerro de Oro, in small rock- pile at base of left bank of arroyo ca. 2.6 mi from Cerro de Oro. 29° 38.3’ N, 110° 37.7’ W. Mina San Jose (Zona de Minas) along dirt road from Rayon to Cerro de Oro. 29° 37’ 17’ N, 110° 37’ 48.8” W. . Rancho La Tarasca (Rancho Las Rastritas) on banks of Rio Los Lobos. 29° 37.5’ N, 110° 38.5’ W. . Behind hill of arroyo along Rayon-Cerro de Oro road, ca. 4.2 mi from Cerro de Oro. 29° 39.6’ N, 110° 35.3’ W. . Sierra Batamote, along left side of La Estrella - Sahuaripa road, on hill facing Rancho El Torreoncito. Mina El Milagro. 28° 57’ 09.7” N, 109° 32’ W. . Cerro La Mona, on left side of road from Hermosillo to Mazatan, ca. 21 km E Hermosillo. 29° 02’ 54.4’’ N, 110° 39’ 22.5” W. . At left side of road, on road from Sahuaripa to Hermosillo, ca. 29° 01.5’ N, 109° 18.7’ W. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 171 Sierra El Viejo, north end of range, on N-facing limestone piles, 30° 24.1’ N, 112° 22.5’ W, elev. ca. 550 m. Sierra El Viejo, SW part of range, along W running wash, on granite rockpiles, 30° 18.7’ N, 112° 20’ W, elev. ca. 540 m. Rancho Tres Marias, W of Alamos, ca. 1 km on road to Conicarit from junction with Hwy 1 (Alamos-Navojoa). 27° 06’ 56” N, 109° 09’ 18” W. Cerro Prieto, ca. 1.5 km west Ejido 18 de Agosto on basalt rocks, ca. 319 14.5’ N, 109° 16.2’ W. At S edge of Sierra El Pinito, ca. 12.8 km W Aribabi, along Im- uris - Cananea road (Hwy 2) on W facing limestone rockslide, 30° 52’ 20’ N, 110° 43’ 10” W. Low and small range along right side of road from Altar to Sasabe. 30° 58.7’ N, 111° 45.4’ W. On asmall hill in front of La Estrella (near El Novillo) on left side of road from Hermosillo to Sahuaripa, ca. 28° 56.8’ N, 109° 38.5’ W. At ESE part of Sierra Alamos, (Mina abandonada) Arroyo Las Piedras, 26° 59’ 37.5’’ N, 108° 56’ 36.4” W. Arroyo Cerro de Oro, ca. % mile from Cerro de Oro on road to Rayon. ca. 29° 36.8’ N, 110° 37.8’ W, elev. ca. 600 m. Cerro de Oro (WBM) sample. Arroyo Cerro de Oro, ca. % mile from Cerro de Oro on road to Rayon. ca. 29° 36.8’ N, 110° 37.8’ W, elev. ca. 600 m. NW end of Sierra Pico, in igneous rocks at base of cliffs, along road from El Plomito to Puerto Libertad, at 23.8 km from El Plomito; ca. 365 m. ca. 30° 04’ 16.2’”’ N, 112° 25’ 18.2” W. Sierra Pozo Verde, Cerro El Sasabe, NE slope, ca. 4 km S of Sasabe, along road to Altar, 319 26.3’ N, 111° 33.8’ W, elev. ca. 1050 m. Arroyo Los Alamos, E side Rio Yaqui, ca. 15 km S El Novillo, 28° 58.1’ N, 109° 37.5’ W, elev. ca. 260 m. SW part of Sierra Batamote. On ravine along La Estrella- Sahuaripa road few miles from La Estrella, ca. 27° 57’ N, 109° 32’ W. Sierro El Viejo SW part of range, along W running wash, on a limestone rockslide, 30° 18.7’ N, 112° 20.2’ W, elev. ca. 540 m. Sierra El Viejo N end of range on granite rockpiles, 30° 23.8’ N, 112° 23.6’ W, elev. ca. 500 m. SW end of Sierra Los Embudos, circa V2 km E of Rancho El Jucaral. 31° 11.8’ N, 109° 05.2’ W. Sierra Mariquita, along road to Microondas installations, ca. 40 km W Cananea at jct road - Hwy 2, 31° 02’ 22” N, 110° 22’ 25” W. SENTENTIA THE NEXT CHALLENGE: LIFE STYLES AND EVOLUTION ALAN SOLEM' DEPARTMENT OF ZOOLOGY FIELD MUSEUM OF NATURAL HISTORY ROOSEVELT ROAD & LAKE SHORE DRIVE CHICAGO, IL 60605-2496, U.S.A. ABSTRACT Knowledge of Western North American land snails, especially those inhabiting desert or semi- arid regions, still rests in the exploratory-descriptive stage. Technological advances in transportation have shifted travel from the ‘hiking and horseback” two month camping trips made early in this cen- tury by Pilsbry and Ferriss, to the current 4-wheel drive vehicles used in weekend dashes. Inevitably technology will shift to other modes of transport — but the dust and sweat of rock moving will remain. We have found more species, and accumulated data on what lives in additional mountain ranges, but we still know almost nothing of these snail’s life histories. Generalizations concerning their pat- terns of variation, ecology, and structure must be abstracted from the early descriptive-faunistic papers of Pilsbry and Ferriss (1905, 1910, 1915) or the seminal report by Gregg (1960) on genital reductions in true desert taxa. These snails do not have a ‘‘natural history’’ elucidated. Without these data, at- tempts to interpret their evolution and biogeography will be simple academic exercises. A brief commentary on Australian arid zone taxa Suggests many avenues for North American investigations, and that the current ‘‘hyperinflation’”’ of suprageneric taxonomic units probably results from our ignorance of desert snail biology and evolution. There are a happy few of us who have experienced the joys, frustration, rewards, bruises, thrill of discovery, and comradeship of collecting land snails in the deserts of the world (e.g. Ferriss, 1904). In Western North America, for nearly 30 years, this has been the basic provenance of Walter Miller and his succession of graduate students, here christened “‘Miller’s Mob’’. At the same time, there are those who “‘put up with”’ the chosen few. A significant gap in this symposium is the failure to permit Walt’s gracious wife a few words on being the ‘‘wife of a snailer’’. This subject has been reviewed delightfully by Frances Norman Young (1954) in her ‘‘A Word From The Wife Of A Tree Snail Hunter’, referring to her hus- band’s ‘‘obsession’’ with Florida Tree Snails of the genus Liguus. In her husband’s words, ‘‘...snail hunting combines all the pleasures of an Easter Egg Hunt with the dangers and thrills of exploration in tropical jungles’ (Young, 1954: 151). All of us desert rats who have backed into a cactus while 1Published posthumously with minor editorial revision. stooping to pick up a shell, trod on a rattlesnake, grabbed a centipede or large scorpion while turning a rock, can testify that desert joys are not second to those of the Everglades. And it is the thrill of exploration that at least partly guides current work. Pick an isolated and unvisited moun- tain range or two, drive to the foothills, hike in, and demolish a few rock piles to extract some “‘live snails’. A part of this process is to verbally share your efforts by shouting messages to people on the next hillside, using the arcane language of “‘snailer’s progress’’ as the rock pile is changed to an ‘‘open pit’. First, there are the near surface ‘“‘bones’’ — weather- beaten long dead shells; then ‘‘fresh shells’’ — snails dead only a year or two; excitement and assurance of success mounts at the first sighting of the characteristic coils of dried excrement that prove living snails are nearby; ‘‘juveniles’’ — first live examples discovered; and finally the triumphant call of ‘‘live adult”. Although not part of ‘‘Miller’s Mob’’, | did share some weekend camps with Walt Miller, ‘‘Doc’’ Gregg, and Munroe Walton. This was in the early 1960’s, when | cut my teeth on American Malacologica! Bulletin, Vol. 8(2) (1991):173-175 173 174 AMER. MALAC. BULL. 8(2) (1991) desert snails. But my interests turned to the Pacific Basin, and then to the semi-arid areas of Australia. Back in the laboratory, the snails were fixed and preserved, genitalia dissected out, cleared and prepared for study. ‘‘Miller’s Mob’’ followed the technique developed by “‘Doc”’ Gregg and perfected by Walt Miller, of making perma- nent slide mounts of stained genitalia. Differences are noted from previously known species, illustrations prepared, and publications extracted from longer theses. Gradually our knowledge of the land snail distribution patterns grew and the number of described species increased. But surprisingly little new data were added concern- ing life history, evolution, or biogeography. A dedicated amateur, Munroe L. Walton, whose ambition was to collect every species of land snail in the Western United States from its type locality, kept a few live snails in his basement, even- tually permitting a series of notes on longevity of different genera (Walton, 1963, 1970). And ‘‘Doc’’ Gregg (1960) pub- lished an abstract of a major work tracing the evolution of desert helicoid genera by reduction in genital structures. It remains necessary to use the reports of Pilsbry and Ferriss (1905: 226-227; 1910: 47-53; 1915: 363-364) for generalizations as to ecology and variation patterns among the species. They pointed out that ‘‘Our work is a recon- naissance rather than a complete survey”’ (Pilsbry and Ferriss, 1915: 363). The many added finds made by Walt Miller and his students confirm this view, and even in the “‘best explored ranges’’, such simple facts as the total range of any species remain a mystery. THE MODERN VIEW As is evidenced by the contributions of this symposium, malacologists have been exposed to modern ideas concern- ing cladistics, continental drift and ‘‘Pacifica’’, phyletics, and vicariant biogeography. Undoubtedly land snail data will have much to contribute to discussions of these subjects. But our understanding of the significance and evolutionary origin of observed structural differences is very poor. Without such understanding, interpretation of differences will be on the most basic phenetic level, and hence of little import. We see differences in genital structures among species of the many genera, and assume that these indicate that speciation has occurred. We see similarities in structure among taxa from different ranges, but are these convergent adaptations to similar environmental parameters? Or are they evidence for speciation by vicariation? We cannot answer such simple questions as: How long do the species live? How long before they become sexually mature? Are there seasonal variations in genital structure? When do they reproduce? Are mating and egg laying in the same “‘activation’’, or separated by a period in diapause? What percentage of the year are they active in different vegetation zones and ranges? How many times during a year are they activated? What do they feed on? Especially under the stress of desert conditions, when the number and length of activations will be minimal, are there competitive interactions among sympatric species or not? Has the reproductive cycle of species shifted from that found in areas with a ‘‘dependable”’ annual period of rain to a different pattern in places where rain comes only occasion- ally and unpredictably? Are there structural changes that can be correlated with increased aridity? Did the snails col- onize the arid areas? Or did increased aridity come to the snails? For many of the arid zone land snails in Western North America, the currently available data on genital structures, ranges, and climatic parameters could enable initial estimates of environmental correlation and patterns of changes. But much basic life history observation will need to be accom- plished before the full evolution of these fascinating taxa can be understood. TRANSCONTINENTAL COMPARISONS For the past 15 years | have been monographing the camaenid land snails from the western two-thirds of Australia, and puzzling over their patterns of structural and ecological specializations. The basic monographs are now completed, identifying 356 species (234 new) in 51 genera (24 new), with many more species to be discovered. From this massive data set, a number of patterns have emerged, usually correlating closely with moisture regimes. Where there is a dependable and predictable annual period of rain, maturation takes four years, there are considerable seasonal shifts in the size of individual reproductive organs (reduced in size during late wet season and through most of the dry season in order to maximize food storage), little or no evidence of feeding specializations, and generally very large genitalia. Where the annual rainfall is greatly reduced, but still dependable and predictable, patterns of maturation and seasonality remain, but feeding specialization become common, and both the size and complexity of the genitalia tends to be reduced. In areas where the rainfall is sporadic, unpredictable, and usually slight (relieved by rare torrential drenchings), the snails show major feeding shifts, must be ready to reproduce ANY time it rains (hence seasonal shrinkage of reproductive organs is no longer possible), the need for increased food storage has resulted in functional shifts among organs, and there has been drastic decrease in general body size and terminal genital organs. These altera- tions decrease energy use and thus improve the chance of survival in marginal desert areas. These changes are repetitive and recur in unrelated taxonomic lineages. | anticipate that equivalent patterns can be discovered in the New World taxa, and believe that the hunt for and probable recognition of such patterns is a necessary prelude to sound phylogenetic models. After a lengthy period of neglect, malacologists are suddenly paying attention to family level classifications and problems, [see for instance Schileyko (1989, abstract)]. Only a decade ago, Solem (1978: 93) listed but one superfamily and three families in the Helicacea. Schileyko (this sym- posium) now splits the old Helicacea into three superfamilies, 11 families, 33 subfamilies, and 18 tribes - after subtracting a few groups. Such ‘‘malacological hyperinflation” hopefully SOLEM: LIFE STYLES 175 will stop short of the example set by ornithologists, where about 12,500 generic names are available for only 9,000 species of birds! The amount of analytic work that has gone into the revisions of Schileyko and Nordsieck is significant, but | think they are flawed by a failure to appreciate the plastic- ity and adaptability of the molluscan body in adjusting to liv- ing conditions in semi-arid and desert area. Plus the fact that these changes have happened many times. It is clear that much remains to be done before we can hope to understand the phylogeny of land snails and establish an adequate phylogenetic classification. My own view is that the many semi-arid and desert faunas will provide many problems and special opportunities for study, and that loss or size reduction of structures has been a fundamental and repeated evolutionary sequence. Until this is recognized, pat- terns analyzed, and the consequences of such convergences allowed for in phylogenetic reconstructions, the proposed classifications will be less than robust. LITERATURE CITED Ferris, J. H. 1904. Southwestern shells. Nautilus 18(5):50-54. Gregg, WO. 1960. Derivation of the Helminthoglyptidae with particular reference to the desert forms. American Malacological Union, Annual Reports for 1959, pp. 45-46. Pilsbry, H. A. and J. H. Ferriss. 1905. Mollusca of the southwestern states, |: Urocoptidae; Helicidae of Arizona and New Mexico. Proceedings of the Academy of Natural Sciences of Philadelphia 57:211-290. Pilsbry, H. A. and J. H. Ferriss. 1910. Mollusca of the southwestern states: IV. The Chiricahua Mountains, Arizona. Proceedings of the Academy of Natural Sciences of Philadelphia 62:44-144. Pilsbry, H. A. and J. H. Ferriss. 1915. Mollusca of the southwestern states, VII: The Dragoon, Mule, Santa Rita, Baboquivari, and Tucson Ranges, Arizona. Proceedings of the Academy of Natural Sciences of Philadelphia 67:363-416. Schileyko, A. A. 1989. Taxonomic status and phylogenetic connec- tions in the Helicoidea. American Malacological Union, pro- gram (abstract), p. 46. Solem, A. 1978. Classification of the land mollusca. /n: Pulmonates Volume 2A, Systematics, Evolution and Ecology. V. Fretter and J. Peake, editors. pp. 49-97. Academic Press, New York. Walton, M. L. 1963. Length of life in West American land snails. Nautilus 76(4):127-131. Walton, M. L. 1970. Longevity in Ashmunella, Monadenia and Sonorella. Nautilus 83(3):109-112. Young, F.N. 1954. A word from the wife of a tree snail hunter. Everglades Naturalist 2(3):148-151. Date of manuscript acceptance: 21 July 1990 BOOK REVIEW Squid as Experimental Animals Gilbert, D. L., W. J. Adelman, Jr., and J. M. Arnold, eds. Plenum, New York, 1990. 516 pp. $75.00. It is nice to see a new book about cephalopods because until recently there were precious few of them. Squid as Experimental Animals weighs in at just over 17 Ibs., with over 500 pages, 22 chapters, 34 contributors and no fewer than 3 editors. It is a successor to a modest handbook of 74 pages with 6 authors, published in 1974, and entitled A Guide to the Laboratory Use of the Squid Loligo pealei, and its size reflects both the advantages and disadvantages of having more knowledge. Of course, it is a great tribute to squids that many biologists devote at least their summers to studying them and the idea of bringing together the work of North American squid experts in a single volume like this obvious- ly has much to commend it. The editors are to be con- gratulated on their enterprise and the book will undoubtedly be very useful. However, expansion of a field should not automatically mean expansion of all treatments of it, and in my view, this guide book would have been immeasurably im- proved if it had been kept shorter. There are six sections. The first, Evolution, History and Maintenance, comprises 58 pages and is, as its name sug- gests, a hotch potch. A lengthy chapter on rearing and cultur- ing squids (Hanlon) is the meat here and there is also a little natural history, though alas without any pictures. It also in- cludes two bizarre indulgences, in one of which the evolu- tion and intelligence of cephalopods are despatched in a mere half a dozen pages. The second section, Mating Behaviour and Embryology (25 pages long), comprises two chapters by the same author (Arnold) so that there is uniformity here and some useful data are clearly presented. | was sorry, however, to see that the only pictures in the entire book of squids behav- ing showed such strange, grey-looking creatures. For | do not believe that Loligo pealei maintains such a uniform ap- pearance during courtship and copulation and indeed the pic- tures belie some of the author’s own description. The third and fourth sections (99 pp. and 172 pp., respectively) constitute the bulk of the book. They are broad- ly devoted to what has come to be known as ‘“‘axonology’’, though the editors divide the ten chapters somewhat arbitrarily into Neural Membranes (Section III) and Cell Biology (Section IV). The treatment here is very uneven indeed. First in length: in the shortest chapter, the cut-open axon technique is described in only seven pages yet the authors of the cytoskeleton chapter are allowed to run to 68 printed pages! The median length is 20 pages but the fact that the chapter by two of the editors ( Adelman and Gilbert) is 40 pages long seems evidence that they do not rate brevity particularly high and so were hardly likely to demand it from their contributors. More seriously, the treatment is also uneven, more so, in fact, than is usually associated with a multi-author work: were all the authors really given the same brief by the editors? If they 177 were and some authors ignored it then the editors should have called the transgressors to order. But Adelman and Gilbert themselves (Ch. 7) present such an uneasy mix of history, tabulated data and banal figures that we can hardly expect much better of the other authors. Some of the chapters in this part of the book will un- doubtedly be very useful for beginners in the field and we may mention here the chapters by Bezanilla and Vandenberg (Ch. 9), Stanley (Ch. 11), Rice, Mueller and Adelman (Ch. 12), R. S. Cohen, Pant and Gainer (Ch. 13), and Weiss, Meyer and Langford (Ch. 15). On the other hand some will disappoint: the chapter on the potentially fascinating axoplasm by Brown and Lasek (Ch. 14) is so long that this reviewer at least had great difficulty in seeing the wood for the trees. How | wish the editors had been tougher and bullied the authors into tell- ing us just what it was they had found out. In fact this brings me to my chief concern with this part of the book: what ex- actly is being found out these days about the axon? For ex- ample, internal dialysis (Ch. 8) and optical measurements (Ch. 10) are all very well but | could not discover what exactly they have added to our understanding of the axon. Indeed Gould and Alberghina (Ch. 16) frankly admit that the function of most lipids in the axon is ‘‘a mystery’, although they only come clean after 46 pages! We all knew that J. Z. Young, Cole, Curtis, Hodgkin, Huxley and Baker were “‘giants’’ but, my goodness, they really do seem to have creamed off the very best bits, didn’t they? Just where is ‘‘axonology’’ going? The fifth section, Sensory Systems, which comprises 67 pages, offers the reader the best value for money in the volume. Saibil and Meinertzhagen on the eye (Chs. 17 and 18) and Budelmann on the statocysts (Ch. 19) produce no verbiage, and, as well as providing admirable summaries these chapters offer the kind of practical technical informa- tion a potential reader of this book might need, often moreover in an easily accessible form. And they summarise important and elegant work. The last section (Part VI), Integrated Systems, is another hotch potch, a short account of an organophoshorus detoxifying enzyme nestling cheek by jowl with an interesting account of gas transport (Mangum) and a nice chapter on physiological performance (O’Dor, et al.), which clearly is about what | take to be an “integrated system’, i.e. an animal. There is no overview or summary; but there is a sub- ject index. References are wisely left attached to their ap- propriate chapters. So overall the book is like the curate’s egg, excellent in parts. Whether it could be said to meet the Editors’ aim of being ‘‘an introduction to squid biology”’ is a moot point, however, and one that readers of the American Malacological Bulletin will need to ponder. Obviously it will be very useful for graduate students and for workers experienced with other animals coming new to Loligo. | have already drawn atten- tion to the sort of chapter | would expect to be really useful to such people: Saibil’s chapter on the eye would be my model. But what a pity the editors did not include an introduc- tory chapter with a few diagrams to show the beginner the main external features of Loligo and the major organ systems - a re-working of L. W. Williams (1909) in fact. However | do not want to go on about the editors. Lots of the authors are to blame for prolixity and even more im- portantly for a feature that as a Zoologist | found especially disppointing. The biology of these wonderful animals is 178 somehow missing in this book: it is as if all the boring bits got put in and the exciting parts left out! Those of us who are fortunate enough to have worked with living squid or the ravishingly beautiful tissues that can be dissected out of freshly killed squid will not be able to use this text to try to convey to our less fortunate students and colleagues the sheer beauty and excitement of many of the systems that the con- tributors to this volume discuss. —John Messenger, Dept. cf Zoology University of Sheffield Sheffield 10, England OLOGICAL UNION 199] 57TH ANNUAL MEETING Berkeley, CA THE AMERICAN MALACOLOGICAL UNION of MALACOLOGISTS BERKELEY, CALIFORNIA JUNE 30 - JULY 5, 1991 The 57th annual meeting of the American Malacological Union will be a combined meeting with the Western Society of Malacologists, held June 30 - July 5, 1991 at the Clark Kerr Campus of the University of California, Berkeley. The conference center provides a comfortable and convenient complex of meeting facilities, adja- cent dining facilities, and guest accommodations in a Spanish style complex separated by lawns, enclosed gardens and courtyards. The Campus is nestled into 43 acres overlooking San Francisco Bay and stretching upward into the Berkeley Hills. The conference center will offer a special package of meals and accommodations, in a combination of residence halls and suites. A variety of hotel and motel housing will be available to those who wish to stay off campus. The Bay Area is filled with diverse cultural and natural history attactions as well as recreational and sight- seeing opportunities. It is famous for its restaurants and fine cuisines. Berkeley summers are free of rain and are regularly cooled (sometimes chilled) by the renowned San Francisco Fog. Three symposia are planned: MARINE BIVALVE RESEARCH IN THE NEXT CENTURY, A REVIEW OF THE CURRENT STATE OF OUR KNOWLEDGE AND DIRECTIONS FOR THE FUTURE (Organized by Drs. Paul H. Scott, Brian Morton, and Eugene V. Coan) MOLLUSCAN TAPHONOMY AND PALEOECOLOGY (Organized by Drs. Carole S. Hickman and Michael P. Russell) MOLLUSCAN BIOGEOGRAPHY OF THE PACIFIC BASIN (Organized by Drs. David R. Lindberg and Geerat J. Vermeij) In addition to the symposia, contributed papers, and poster presentations, scheduled events will include field trips, an outdoor barbecue, a joint AMU/WSM auction, a dessert reception at the Museum of Paleontology, and a 4th of July Banquet in the Great Hall of the Faculty Club. Fossil and Recent mollusk collections in the University of California Museum of Paleontology and the California Academy of Sciences in San Fran- cisco will be available to meeting participants before, during, and after the meeting. For further information please contact: Carole S. Hickman President, AMU Museum of Paleontology University of California Berkeley, California 94720 Telephone (415) 642-3429 FAX (415) 642-1822 179 SPECIAL PUBLICATIONS OF THE AMERICAN MALACOLOGICAL BULLETIN The Special Publication Series of the American Malacological Bulletin was begun to disseminate collected sets of papers with similar or related themes in a single volume. To date, three such issues have been published, each the result of a special convened symposium. The three Special Editions are PERSPECTIVES IN MALACOLOGY, PRO- CEEDINGS OF THE SECOND INTERNATIONAL CORBICULA SYMPOSIUM, and PRO- CEEDINGS OF THE SYMPOSIUM ON ENTRAINMENT OF LARVAL OYSTERS. Additional Special Editions are planned for the near future. PERSPECTIVES IN MALACOLOGY (Sp. Ed. #1, July, 1985) offers a wide range of papers dealing with molluscan biology of interest to professionals and amateurs alike. These papers were presented as part of asymposium held in honor of Professor M. R. Carriker at the time of his retirement and highlight a variety of recent advances in numerous facets of the study of molluscs. PERSPEC- TIVES IN MALACOLOGY offers insight into some of the frontiers of molluscan biology ranging from deep-sea hydrothermal vent malacofauna to chemical ecology of oyster drills. The PROCEEDINGS OF THE SECOND INTERNATIONAL CORBICULA SYMPOSIUM (Sp. Ed. #2, June 1986) contains numerous papers on this exotic bivalve that has become a significant ‘‘pest’’ organism of several power plants and other industries using cooling waters. The proliferation, spread, functional biology, attempts at industrial control, taxonomy, and many other topics of interest to the malacologist and industrial biologist are addressed in this important special publication. The third special edition of the American Malacological Bulletin, PROCEEDINGS OF THE SYMPOSIUM ON THE ENTRAIN- MENT OF LARVAL OYSTERS (Sp. Ed. #3, October 1986) contains important review papers on the larval biology of the American oyster Crassostrea virginica, as well as intriguing papers on factors that limit productivity of these bivalves and limitations that exist on their dispersal and survival. The impact of cutter-head dredges is addressed in this special edi- tion with special emphasis on the Chesapeake Bay system. To order your copies of PERSPECTIVES IN MALACOLOGY, PROCEEDINGS OF THE SECOND INTERNATIONAL CORBICULA SYMPOSIUM or PROCEEDINGS OF THE SYMPOSIUM ON ENTRAINMENT OF LARVAL OYSTERS, simply fill out the form below. Enclose check or money order made out to the AMERICAN MALACOLOGICAL BULLETIN. PERSPECTIVES IN 2ND INTERNATIONAL ENTRAINMENT OF MALACOLOGY CORBICULA SYMPOSIUM LARVAL OYSTERS Special Edition No. 1 Special Edition No. 2 Special Edition No. 3 AMERICAN MALACOLOGICAL BULLETIN AMU Members $10.00 $20.00 $14.00 Non-AMU Members $15.00 $28.00 $20.00 Unitas Members $12.00 $22.00 = $16.00 Institutions $25.00 $3700 _ $28.00 Foreign Airmail $10.00 $10.00 $10.00 Foreign Seamail $ 5.00 Ss 00g eee oe $ 5.00 SUBTOTALS $ nae : a eae $ TOTAL ENCLOSED $ (check or money order made out to: AMERICAN MALACOLOGICAL BULLETIN) Name: Mailing Address: Send Orders To: Richard E. Petit Acting AMU Secretary-Treasurer P. O. Box 30 North Myrtle Beach, South Carolina 29582 U.S.A. 180 IN MEMORIAM Irene E. Beeler Yoshio Kondo Jane Zager 181 Ahlstedt, S. A. 139 Cardoso, F. 143 Cicerello, R. R. 113 Cuezzo, M. G. 19 Dadon, J. R. 77 Fiege, D. 27 Fuller, S. C. 107 Gilmer, R. W. 53, 67 Kennedy, V. S. 107 Kijviriya, V. 97 Actinonaias 114 Alasmidonta 114 Amblema 115 Ancomena 159 Ancotrema 159 Anodonta 11, 15, 134 Anodontoides 116 Arcinins 117 Atlanta 47, 85 Bradybaenidae 148 Bunnyinae 150 Carinaria 47 Catinella 11 Cavolinia 27, 56 Cecilioides 168 Cepoliinae 148 Chaenaxis 168 Clio 27, 56 Clione 67 Corbicula 11, 97, 107 Corbiculidae 97 Creseis 27 Cumberlandia 117 Cuvierina 56 Cyclonaias 117 Cyprogenia 117 Dereceros 11 Diacria 56 Dromas 117 Ellipsaria 117 Elliptio 117, 133 Epioblasma 117 Epiphragmophorinae 148 Euconulidae 169 Euconulus 11, 169 Euniomerus 11 Euthecosomata 54 Ferussaciidae 168 Firoloida 40 Fusconaia 118, 133 Gastrocopta 11, 166 Geomene 159 Glaucidae 61 Glaucilla 61 Dates of Publication Volume 8(1), August, 1990 Volume 8(2), April, 1991 INDEX TO VOLUME 8 (1 and 2) AUTHOR INDEX Lalli, C. M. Lutz, R. A. Messenger, J. Miller, W. B. Morton, B. Naranjo-Garcia, E. Neck, R. W. Newman, L. J. Quetin, L. B. Ross, R. M. Glaucus Glebula Glyphyalinid Gryaulus Haplotrema Haplotrematidae Haplotrematinae Hawaiia Hebetancylus Helicellidae Helicidae Helicodiscidae Helicodiscus Helicoidea Helminthoglyptidae Hemistena Humboldtianidae Hyalocylis lo Karolus Lamellaxis Lampsilis Lasmigona Leptodea Lexingtonia Ligumia Limacina Limacinidae Lysinoinae Medionidus Megalonais Metostracinae Monadeniinae Musculium Neohelix Obliquaria Obovaria Ocyth6e Ostreidae Oxyloma Paraclione Pegias Physaria Physella 182 67 107 177 147 1 147, 165 9 85 61 61 PRIMARY MOLLUSCAN TAXA INDEX 61 118 169 11 159 155 155 11, 169 11 148 148 169 11, 169 147 148 118 149 38 139 168 168 118, 133 119, 134 119 119 119 27, 54, 77 54 148 119 119 150 149 11 19 117 119 143 1 11 39 120 11 11 Roth, B. 155 Seapy, R. R. 45 Schuster, G. A. 113 Solem, A. 173 Thiriot-Quiévreux, C. 37 Upatham, E. S. 97 Viyanant, V. 97 Warren, M. L., Jr. 113 Warren, R. E. 131 Woodruff, D. S. 97 Pisidium 11 Planorbella 11 Plectomerus 120 Plethobasis 120 Pleurobema 120, 133 Pneumodermopsis 38 Potamilis 120 Protatlanta 48 Pseudosubulina 169 Pterotrachea 39, 40 Ptychobranchas 120, 133 Pupillidae 166 Pupsioma 168 Pupoides 11, 168 Quaadrula 121 Rhabdotus 11 Rumina 11 Saccostrea 1 Selenites 161 Simpsonaias 121 Sonorellinae 148 Sphaerium 11 Spiraxidae 169 Stenotrema 11 Strophitus 121, 134 Subulinidae 168 Succinea 11 Thysanophora 169 Thysanophoridae 169 Toxolasma 121 Trichodiseininae 150 Tritogonia 121 Truncilla 121 Uniomerus 121 Unionidae 113, 131 Vallonia 11, 168 Vallonidae 168 Venustaconcha 134 Vertigo 11, 168 Villosa 122, 134 Vitrinidae 169 Xanthonychidae 148 Xanthonychinae 148 Zonitoides 11 Illustrations should be clearly detailed and readily ‘reproducible. All line drawings should be in black, high quality ‘ink. Photographs must be on glossy, high contrast paper. All diagrams must be numbered in the lower right hand corners and adequately labeled with sufficiently large labels to pre- SR anUacapl and atconipanying ifluctra: _ vent obscurance with reduction by one half. Magnification bars ibmitted with two additional copies for review | _—- must appear on the figure, or the caption must read Horizon- st be typed on one side of 812 x 11 inch _ tal field width = xmm or xum. All measurements must be in “Spaced, and all pages numbered con- ~ metric units. All illustrations submitted for publication must ith numbers appearing in the upper right hand (oe fully cropped, mounted on a firm white backing ready for F ‘page. Leave ample margins on all sides. reproduction, and have author’s name, paper title, and figure th ‘manuscript should follow that outlined i in _ number on the back. All figures in plates must be nearly con- logy Editors Style Manual (fifth edition, 1983). _tiguous. Additional figures submitted for review purposes must sed from the CBE, 9650 Rockville Pike, _ be of high quality reproduction. Xerographic reproduction of __ photomicrographs or detailed photographs will not be accep- en propriate should be arranged i in sections table for review. Abbreviations used in figures should occur sae | the figure caption. Indicate in text margins the appropriate ith title, author(s) and address(es), and ne location in which figures should appear. Color illustrations can inning title of no more than 50 = Os included at extra cost to the author. Original illustrations ers and spaces. Authors should also supply —_will be returned to author if requested. ords, placed at the base of this page, for — Any manuscript not conforming to AMB format will be returned to the author for revision. Proofs. = eee will be sent to the author and must be aie ieee iene Onion will not be ienenabe ft deferred Publication of manuscripts delayed i in surface mail. c Charges. There are no mandatory page costs to authors lack- 2 IQ: financial support. Authors with institutional, grant or other research ‘support will, be billed for page charges. The current ~ rate is $35, 00 per printed page. Acceptance and ultimate Publication 1 is in ‘no ‘way based on ability to pay page costs. ae Reprints. Order forms and reprint cost information will be sent en out ie first tiie that taxon is sipisnied with page proofs. ‘The author receiving the order form is gh. The generic name can be abbreviated responsible for i insuring that ees for any coauthors are also the paragraph as follows: C. virginica. laced at that time. re cited section of the manuscript refer- ee double le spaced. All authors must Submission. Submit all inuactipts to Dr. Robert S. Prezant, oda Editor-in-Chief, American Malacological Bulletin, Department ot Biology, Indiana University of eeSyvanla, Indiana, Penn- _ sylvania, 15705-1090, DS Act Poa Subscription Costs. inetitutional subscriptions are available on ak a cost of $32. 00 per volume. [Volumes 1 and 2 are available for. $18.00 per volume}. ‘Membership in the American ree Malacological Union, which includes personal subscriptions Bulletin, is available for $20.00 ($15.00 for students). prices quoted are in US. funds. Outside the U.S. postal ones, add $5.00 seamail and $10.00 airmail per volume or membership. For subscription or membership contact Acting — AMU Secretary-Treasurer, Richard E. Petit, P.O. Box 30, North oe Beach, eae Carolina, 29582, U.S.A... rae. es AMERICAN MALACOLOGICAL BULLETIN VOLUME 9 NUMBER 1 Biannual Journal of the American Malacological ‘Union CONTENTS a Editorial Comments vs.204540s.0G+ shad veda sanyh sd avons sdiehs sine s doh dPdadela vy beclencead 4 iii Observations on the anatomy of the scaphopod mantle and the description ' of a new family, the Fustiariidae. GERHARD STEINER..................... eu Ba oe cans 1 The freshwater mussels (Bivalvia: Unionoidea) of the upper Delaware River drainage. DAVID L. STRAYER and JONATHAN RALLEY...................... 21 The Naiades (Bivalvia: Unionoidea) of the Delmarva Peninsula. CLEMENT L. COUNTS, III, THOMAS S. HANDWERKER and ; ROMAN ‘V. JESUEN sccc400 coisa cee es inten hs bon be bowen daw badee et beua de. 27 The influence of oxygen availability on oxygen consumption in the freshwater clam Musculium partumeium (Say) (Bivalvia: Sphaeriidae). 13 DANIEL J. HORNBACH ooo ecdb cece cdaaa wes seebuvebeyedenebelewtilhcn nt 3D INTEGRATIVE NEUROBIOLOGY AND BEHAVIOR OF MOLLUSCS Sexual conflict and the mating systems of simultaneously hermaphroditic gastropods. JANET L. LEONARD ........................000... eee eee 45 Substratum associations of natural populations of Iceland Scallops, Chlamys islandica Miller 1776, on the northeastern Grand Bank of Newfoundland. KENT D. GILKINSON and JEAN- MARC GAGNON ...................0000000 2005 - 59 Anatomical and behavioural studies on vision in Nautilus and Octopus. W. R. A. MUNTZ ..... Sp AE aa H RaS Cohn a SoA S CR ein oe Dove ag eae 69 Complex learning in Octopus bimaculoides. JEAN BOAL ...... 00.0.0. 75 Mating behavior of the freshwater pulmonate snail, Physa gyrina. THOMAS J. DOWUCD v4 2.00025 vi tate dcd cee cs ou ou eee Ri ea ee ee 81 Reproductive patterns and seasonal occurrence of the Sea Hare Aplysia brasiliana Rang (Gastropoda, Opisthobranchia) at South Padre Island, Texas. NED E. STRENTH and JAMES E. BLANKENSHIP ..........................-0000- 85 Sententia: Variation in sense organ design and associated sensory capabilities among closely related molluscs. P. V. HAMILTON .............0.. 000 ccc cece ee eeeeee 89 Research Note: Sperm storage and evidence for multiple insemination in a natural population of the freshwater snail, hae AMY R. WETHINGTON and ROBERT T. DILLON, JR. 2... 2.0.06. .00ccccceceeee teen ne eeees 99 AMNOUNCEMERE «2c 2t0052 cds doce tan Heese lanes Mea Pore Oe ee ee eee 103 AMERICAN MALACOLOGICAL BULLETIN ROBERT S. PREZANT, Editor-in-Chief BOARD OF EDITORS Department of Biology Indiana University of Pennsylvania Indiana, Pennsylvania 15705 ASSOCIATE EDITORS MELBOURNE R. CARRIKER College of Marine Studies University of Delaware Lewes, Delaware 19958 GEORGE M. DAVIS Department of Malacology The Academy of Natural Sciences Philadelphia, Pennsylvania 19103 R. TUCKER ABBOTT Melbourne, Florida, U.S.A. JOHN A. ALLEN Millport, United Kingdom JOHN M. ARNOLD Honolulu, Hawaii, U.S.A. JOSEPH C. BRITTON Fort Worth, Texas, U.S.A. JOHN B. BURCH Ann Arbor, Michigan, U.S.A. EDWIN W. CAKE, JR. Ocean Springs, Mississippi, U.S.A. PETER CALOW Sheffield, United Kingdom JOSEPH G. CARTER Chapel Hill, North Carolina, U.S.A. ARTHUR H. CLARKE Portland, Texas, U.S.A. CLEMENT L. COUNTS, III Princess Anne, Maryland, U.S.A. THOMAS DIETZ Baton Rouge, Louisiana, U.S.A. WILLIAM K. EMERSON New York, New York, U.S.A. DOROTHEA FRANZEN Bloomington, Illinois, U.S.A. VERA FRETTER Berkshire, United Kingdom ROGER HANLON Galveston, Texas ROBERT C. BULLOCK, Ex Officio Department of Zoology University of Rhode Island Kingston, Rhode Island 02881 BOARD OF REVIEWERS JOSEPH HELLER Jerusalem, Israel ROBERT E. HILLMAN Duxbury, Massachusetts, U.S.A. K. ELAINE HOAGLAND Washington, D.C., U.S.A. RICHARD S. HOUBRICK Washington, D.C., U.S.A. VICTOR S. KENNEDY Cambridge, Maryland, U.S.A. ALAN J. KOHN Seattle, Washington, U.S.A. LOUISE RUSSERT KRAEMER Fayetteville, Arkansas, U.S.A. JOHN N. KRAEUTER Baltimore, Maryland, U.S.A. ALAN M. KUZIRIAN Woods Hole, Massachusetts, U.S.A. RICHARD A, LUTZ Piscataway, New Jersey, U.S.A. GERALD L. MACKIE Guelph, Ontario, Canada EMILE A. MALEK New Orleans, Louisiana, U.S.A. MICHAEL MAZURKIEWICZ Portland, Maine, U.S.A. JAMES H. McLEAN Los Angeles, California, U.S.A. ROBERT F;} MCMAHON Arlington, Texas, U.S.A. RONALD B. TOLL, Managing Editor Department of Biology University of the South Sewanee, Tennessee 37375 W. D. RUSSELL-HUNTER Department of Biology Syracuse University Syracuse, New York 13210 THOMAS R. WALLER Department of Paleobiology Smithsonian Institution Washington, D. C. 20560 ANDREW C. MILLER Vicksburg, Mississippi, U.S.A. BRIAN MORTON Hong Kong JAMES J. MURRAY, JR. Charlottesville, Virginia, U.S.A. RICHARD NEVES Blacksburg, Virginia, U.S.A. JAMES W. NYBAKKEN Moss Landing, California, U.S.A. A. RICHARD PALMER Edmonton, Canada WINSTON F- PONDER Sydney, Australia CLYDE F; E. ROPER Washington, D.C., U.S.A. NORMAN W. RUNHAM Bangor, United Kingdom AMELIE SCHELTEMA Woods Hole, Massachusetts, U.S.A. DAVID H. STANSBERY Columbus, Ohio, U.S.A. FRED G. THOMPSON Gainesville, Florida, U.S.A. NORMITSU WATABE Columbia, South Carolina, U.S.A. KARL M. WILBUR Durham, North Carolina, U.S.A. Cover. Io fluvialis (Say, 1825) is the logo of the American Malacological Union. THE AMERICAN MALACOLOGICAL BULLETIN is the official journal publication of the American Malacological Union. AMER. MALAC. BULL. 9(1) ISSN 0740-2783 CONTENTS Editonale@omunents? sy... ion ide ees een on te a Gs A Se eee hee ne bad dae owe iil Observations on the anatomy of the scaphopod mantle and the description of a new family, the Fustiariidae. GERHARD STEINER ...........................0005. 1 The freshwater mussels (Bivalvia: Unionoidea) of the upper Delaware River drainage. DAVID L. STRAYER and JONATHAN RALLEY...................... 21 The Naiades (Bivalvia: Unionoidea) of the Delmarva Peninsula. CLEMENT L. COUNTS, HI, THOMAS S. HANDWERKER and ROMAN. YV: JESTEN 3 cos sect o os eek os pee a ed wh A ee ieee ow edb ga RE 27 The influence of oxygen availability on oxygen consumption in the freshwater clam Musculium partumeium (Say) (Bivalvia: Sphaeriidae). DANIEL. JeHORNBACH .. y 4-s6enn Gucn tor aat aaci Veda @atne ss kobe se ne Deina tens waalen 39 INTEGRATIVE NEUROBIOLOGY AND BEHAVIOR OF MOLLUSCS Sexual conflict and the mating systems of simultaneously hermaphroditic gastropods. JANET i. LEONARD 2... 2900 fe 20 sae ns FEE WR Oa Pee eo aes awss 45 Substratum associations of natural populations of Iceland Scallops, Chlamys islandica Miiller 1776, on the northeastern Grand Bank of Newfoundland. KENT D. GILKINSON and JEAN- MARC GAGNON ..............0.....00000 02 eeee 59 Anatomical and behavioural studies on vision in Nautilus and Octopus. VOR AI DUUIN Zoe 52 Are 208, 55 Beda d cepa Sper Me ce eat ae apc risa URIS fo kat tibia 69 Complex learning in Octopus bimaculoides. JEAN BOAL ...........0.0000000 00 ccc eee 75 Mating behavior of the freshwater pulmonate snail, Physa gyrina. FEHONIAS A SDOW VG is 26102 ae ld, ee Gad dohad okt AR Sesi as oawieee ciS3 wad owas 81 Reproductive patterns and seasonal occurrence of the Sea Hare Aplysia brasiliana Rang (Gastropoda, Opisthobranchia) at South Padre Island, Texas. NED E. STRENTH and JAMES E. BLANKENSHIP......................0..0.00.5. 85 Sententia: Variation in sense organ design and associated sensory capabilities among closely related molluscs. P. V. HAMILTON ...........0...0. 0.00.00 cee eee eee 89 Research Note: Sperm storage and evidence for multiple insemination in a natural population of the freshwater snail, Phyra. AMY R. WETHINGTON and ROBERE TT. DUGLON,. Rin « 2 ee onto bene onan t peduuae tena th waaee dads 99 ATIMOUNGELTICTI tee rea eee eee tees ante Ne ee AR Oe ee TE eS ati she es 103 EDITORIAL COMMENTS The American Malacological Bulletin has strived over the years to offer the malacological community a diverse and high quality series of articles on molluscan biology. Since its inception in 1983, the AMB has undergone an evolution leading to the package you now hold in your hands. To celebrate the changes and evolution of the Bulletin, as well as the continued success of the American Malacological Union, we have altered our journal appearance. As you can note, the articles are in a different format, one that we hope satisfies both the professional and aesthetic tastes of our readers. The publication date of a particular issue will no longer be given in that issue, because it can not always be guaranteed that this date represents the exact date of distribution (see ICZN Code, Art. 21). The publication date of a current issue will therefore be published in a later issue. Additionally, to denote these changes, there are alterations in our contributor information that clarify and highlight expectations of our journal and the desire to increase the level of submissions dealing with new molluscan taxa. The new cover stock also reflects these changes. We continue to hope our readership and members of AMU find the American Malacological Bulletin an important contribution to malacology. We appreciate the sup- port of our Reviewers, Editorial Board, AMU Members and Council and general readership. RSP, Editor-in-Chief lll Observations on the anatomy of the scaphopod mantle and the description of a new family, the Fustiariidae Gerhard Steiner Institute of Zoology, University of Vienna, A-1090, Vienna, Austria Abstract. The mantle of five genera of the order Dentaliida and ten genera of the order Gadilida are investigated using histological and ultrastructural methods. The anterior mantle edge is divided into three functional units: 1) a frontal epithelium; 2) a central fold; 3) an inner gland region. In the Dentaliida the frontal epithelium is differentiated as an outer gland region and the central fold carries an annular ciliary organ. Rhabdus rectius (Carpenter) (Dentaliida) has a pair of ciliated slits instead of the annular organ. Neither feature is found in the Gadilida, where the frontal epithelium forms sensory papillae. Different types of ciliary receptors in the anterior mantle edge of both orders are present. The inner gland region in the Dentaliida consists of epithelial gland cells only; in the Gadilida both epithelial and subepithelial gland cells are present. The epidermis of the pallial cavity is low cuboidal but features ciliary rings in the anal region. The posterior pallial edge or pavillon of the gadilids has a powerful ciliary organ to produce water currents and ciliated ridges on the dorsal mantle process; both structures are lacking in the dentaliids. Fustiaria differs from other genera of the Dentaliida by lacking a ventral bolster of connective tissue at the posterior mantle margin. Abundance and distribution of subepithelial gland cells vary between genera. The validity of the family Rhabdidae is supported and a new family, the Fustiariidae, is proposed. The anatomy of the Scaphopoda has received the least Boissevain (1904) described the posterior mantle edge, the attention of all of the classes of the molluscan subphylum so-called pavillon, and reported a small ciliated area and scat- Conchifera. Early accounts, e.g. Deshayes, 1825; Lacaze- tered gland cells. Reynolds (1988) investigated sensory cells Duthiers, 1856-7; Fol, 1889; Plate, 1892; and Simroth, 1894, fo the pavillon of Rhabdus rectius (Carpenter) (Dentaliida). however, contained excellent studies of scaphopod This paper presents a comparative anatomy of the man- morphology and histology that remain valid in many aspects. tle in selected members of the orders Dentaliida and Gadilida Although almost all investigations were made on a single (= Siphonodentaliida) to demonstrate the anatomical divers- genus, Dentalium (=Antalis) (order Dentaliida), the results ity in Scaphopoda and to provide new information for a bet- have been generalized subsequently for all Scaphopoda. ter founded classification. Lacaze-Duthiers (1856-7), Plate (1892), and Boissevain (1904) described the histological structure of the mantle in MATERIALS AND METHODS Antalis entalis (L.). They noted an outer glandular region at the anterior mantle edge, including two to three types of The species studied are listed in Table 1. Mediterranean gland cells, a ring of connective tissue supporting the man- species were collected by means of SCUBA or by a triangle tle opening, and an inner glandular region consisting of dredge from 8-20 m in Rovinj and Piran (Peninsula of Istria, mucus-producing cells. Plate (1892) found that the genus Yugoslavia). In Norway, animals were gathered by a ‘*Mini- Cadulus lacked the outer gland region, and later Odhner Sanders’’ epibenthic sledge and box-core sampler from (1931) confirmed. Boissevain (1904) reported a ring of ciliated 120-590 m in the fjords near Blomsterdalen, Bergen. The cells between the gland regions that she considered to be a animals from Puget Sound, Washington, U.S.A., were col- sensory organ. Gabe and Prenant (1950) included the anterior lected with a triangle dredge from 75-100 m near Waldrun mantle edge of A. entalis in their histological studies of the Island. scaphopod connective tissue. In the midregion of the animal, ciliated ridges were HISTOLOGICAL PREPARATION found just anterior to the anal opening in Antalis entalis and Living animals were narcotized with isotonic MgCl, A. dentalis (L.) by Lacaze-Duthiers (1856-7) and Fol (1889). solution (1:2 to 1:1 with sea-water) or MS 222 added to the Distaso (1906) described them in more detail and also claimed sea-water (1:4000 - 1:2000). Tissues were fixed in 2-4% for- the presence of an osphradial sense organ. Leon (1895) and malin or 4% glutaraldehyde buffered in sea-water. Helly’s American Malacological Bulletin, Vol. 9(1) (1991):1-20 | i) Table 1. Systematic list of the species examined (USNM, United States Na- tional Museum of Natural History, Smithsonian Institution; NMNZ, National Museum of New Zealand; 'Museum National d’Histoire Naturelle, Paris; *Gareth Davies, University of Edinburgh; *Zoologisk Museum, Kopenhagen; *prepared for TEM methods) Order DENTALIIDA Family Source Dentaltidae Dentalium laqueatum Verrill, 1885 USNM 765278 D. neohexagonum Sharp and Pilsbry, 1897 Graptacme calamus (Dall, 1889) Antalis dentalis (L., 1767) A. entalis (L., 1758) A. inaequicostatum (Dautzenberg, 1891) A. occidentalis (Stimpson, 1851) A. vulgaris (Dacosta, 1778) A. sp. BS 660 A. sp. P 927 II A. sp. 930 A. sp. Q 719 Fissidentalium candidum (Jeffreys, 1877) F. majorinum (Mabille and Rochebrune, 1889) F. megathyris (Dall, 1899) F. zelandicum (Sowerby, 1860) Fustiaria rubescens (Deshayes, 1825) F. sp. Rhabdidae Rhabdus rectius (Carpenter, 1864) Laevidentaliidae Laevidentalium callipaplum (Dall, 1889) Entalinidae Entalina quinquangularis (Forbes, 1843) Bathoxiphus ensiculus (Jeffreys, 1877) B. sp. S 153 Heteroschismoides subterfissum (Jeffreys, 1877) Pulsellidae Pulsellum lofotense (M. Sars, 1865) P. salishorum Marshall, 1980 P. sp. BS 940 P. sp. P 937 1 Annulipulsellum euzkadii Scarabino, 1986 Striopulsellum sandersi Scarabino, in lit. Siphonodentaliidae Siphonodentalium grandis (Verill, 1884) S. lobatum (Sowerby, 1860) S. spectabilis Verill, 1885 S. vitream M. Sars, 1851 Wemersoniellidae Wemersoniella turnerae Scarabino, 1986 Gadilidae Gadila fraseri Nickles, 1979 G. metivieri Scarabino, in lit. G. sp. Cadulus aberrans Whiteaves, 1887 . arctatus Jeffreys in Locard, 1898 . cylindratus Jeffreys, 1877 . delicatulus Suter, 1913 . jeffreysi (Monterosato, 1875) . propinquus Sars, 1878 subfusiformis (M. Sars, 1865) sp. 68 sp. P927 I ANANANAAN Monterey Bay USNM 801250 Rovinj* MBI? Piran* Bergen, Norway* Rovinj NMNZ BS660 NMNZ P927 II NMNZ P930 NMNZ Q719 AT 186? USNM 709081 Galathea St716? NMNZ BS496 Piran NMNZ sand lagoon San Juan Islands USNM 678797 USNM 765451 Thalassa Y 378! Bergen, Norway* Incal WS OI! NMNZ S153 Nordatlante 85! ES 218? Bergen, Norway* San Juan Islands NMNZ BS940 NMNZ P937 I BIOGAS IV, DS63! Nordatlante P19! Incal WS 02! ES164? Incal DS 09! Greenl. Exp. 423 Incal WS 07! Galathea St. 101° BIOGAS V, DS66! NMNZ P939 Monterey Bay Thalassa Z 407! ES 272 NMNZ BS544 BIOGAS V, DS66! BIOGAS VI, DS82! Bergen, Norway* USNM 803468 NMNZ P927 I AMER. MALAC. BULL. 9(1) (1991) fixative produced good fixation of cytoplasmatic components in small specimens and pieces of tissue. Bouin’s fluid (aq.) penetrates tissues rapidly, decalcifies, but often damaged cilia- tion. This fluid was used also for decalcification after one of the other fixatives. Fixation times ranged from 2-24h. The specimens were washed in 70% ethanol + NH, (3 drops/l00ml) to remove the picric acid. After dehydration in ethanol, tissues were transferred to methyl benzoate as an antemedium, benzol, 40% paraffin and embedded in 58-60°C Paraplast. Sections of 5-7 wm were taken and stained with Azocarmine and Anilineblue-Orange G after Heidenhain (AZAN), Haematoxilin and Eosin after Mayer (HE) or Kernechtrot + Picroindigocarmine (KP). Recipes for fixation and staining were modified after Adam and Czihac (1964). Museum samples, usually kept in alcohol, were treated in the same way starting with decalcification or dehydration. SEM PREPARATION The same fixatives were used as for histological pur- poses. Tissues were dehydrated in grades of acetone, critical point dried with liquid CO, and sputter coated with a 200 A gold-layer. The samples were examined in a Jeol JMS 09 scanning electron microscope. TEM PREPARATION Two methods were applied as follows: 1) 4 hr in 3% glutaraldehyde in phosphate buffer 0.1M pH 7.2-7.4 (after Sorensen) or cacodylate buffer 0.1M pH 7.2; postfix for 2 hr in 2% OsO, 0.1M in the corresponding buffer. To 100ml 0.1M buffer, 10g sucrose were added to adjust the osmolar- ity for sea-water. 2) 2 hr fixation in glutaraldehyde- paraformaldehyde (1% formalin + 2.5% glutaraldehyde in 0.IM cacodylate buffer pH 7.2) and postfix for 2 hr in 2% OsO, in the same buffer. All fixatives were cooled to 4°C. Ethylen diamine tetra acetic acid (EDTA) was used for decalcification. After dehydration in ethanol the samples were embedded either in Epon-Araldite (Mollenhauer, 1964) or in Spurr’s medium (Spurr, 1969). Semi-thin sections (0.5-1 yam) were Stained with 0.1% toluidine blue solution. Sections of 70-80 nm were made on a Reichert OMU-3 microtome with glass or diamond knives, and treated for 45 min with a saturated uranyl-acetate solution (Watson, 1958) and for 5 to 8 min with a 0.4% solution of lead-citrate (Venable and Cog- geshall, 1965). Sections were studied under a Zeiss EM9/S2 transmission electron microscope. RESULTS ANTERIOR MANTLE EDGE GENERAL CHARACTERISTICS The anterior pallial opening is supported by a ring of cartilage-like connective tissue as well as by muscles. It can STEINER: SCAPHOPOD MANTLE ANATOMY 3 Fig. 1. A. Scaphopod in a schematic longitudinal section; B. Anterior mantle edge in the order Dentaliida (e.g. Antalis, semischematic longitudinal section; C. Anterior mantle edge in the order Gadilida (e.g. Pulsellum), semischematic longitudinal section; D. Posterior mantle edge in the order Dentaliida (e.g. Antalis), semischematic longitudinal section, insert (not to scale) showing cross-section of the pavillon proper; E. Posterior mantle edge in the order Gadilida (e.g. Entalina), semischematic longitudinal section, inserts (not to scale) showing cross-sections of the pavillon proper and the valve ciliary organ (ciliation drawn on left side only); a, anus; ae, attachment epithelium; as, annular sinus; bc, buccal cavity; ca, ciliated area; cf, central fold; clc, ciliated ledge cells; co, ciliary organ; ctr, connective tissue ring with periostracal groove; drm, dorsal retractor muscle; e, esophagus; f, foot; fe, frontal epithelium; go, gonad; i, intestine; ig, inner gland region; 1, lacunae; Ic, ledge cells; Im, longitudinal muscles; mc, mantle cavity; mo, mouth opening; n, nerve; nm, neuropil; og, outer gland region; pcr, preanal ciliary ridges; pp, pavillon proper; rm, radial muscles; s, socket of connective tissue; sm, sphincter muscle; smg, supramarginal groove; st, stomach; vb, ventral bolster; vco, valve ciliary organ; vm, valve muscle; x, identifies axis of cross-section in lower insert of E; scale bars = 100 pm. 4 AMER. MALAC. BULL. 9(1) (1991) exhibit different gland regions. The whole complex (see Figs. 1A-C, 2) can be divided into: 1) a frontal epithelium; 2) a central fold; 3) an inner gland region. In all scaphopods the central fold consists of a ring of connective tissue and the muscle apparatus. The former is a wedge-shaped mass of extracellular matrix in which col- lagen and scaffolding fibers (fibers grillagées) are embedded (Gabe and Prenant, 1950) (Fig. 3). The orientation of the fibers is predominantly radial. The term ‘‘cartilage-like’’ in- dicates the histological and functional resemblance of this tissue to vertebrate cartilage; it does not indicate homology. Towards the anterior aperture, collagen and muscle fibers become more abundant, replacing the extracellular matrix. Thus, the cartilage-like appearance is lost (Figs. 2, 4). This ring provides an attachment site for the pallial muscles on both proximal and distal sides. The periostracal groove of the outer mantle epithelium has the shape of a con- centric ring. Thus, constant shape and diameter are main- tained by connective tissue for the shell-forming epithelia. In fixed animals the anterior mantle margin is always retracted, so the groove opens to the anterior (Figs. 1, 4). Towards the anterior aperture the central fold includes a 4 tp Oy 0 ay ome Or ee as Di gy ate late nek Sr tins 8 Fh me Pt SF ee ree Pett . < wee te hg SI. Seaweed x EE Re ens e < Fig. 31. Entalina quinquangularis, cross-section of the posterior mantle edge (formalin, H.E., 7 wm); 1, lacunae; s, ventral socket of connective tissue; fibres of the valve muscle (arrows); crests of nerve- and glia cells (arrow-heads); scale bar = 50 um. Fig. 32. Entalina quinquangularis, TEM micrograph of the valve ciliary organ of the posterior mantle edge, horizontal section (glutar-paraformaldehyde - OsO,, Spurr); note the long ciliary rootlets descending between the densely packed mitochondria in the ciliated cell; n, nucleus of ciliated cell; vm, valve muscles; basal lamina (arrow); scale bar = 5 um. Fig. 33. Entalina quinquangularis, cross-section of the pavillon proper (Bouin, AZAN, 7 pm); ciliated ledge cells (arrows); scale bar = 50 pm. Fig. 34. Cadulus subfusiformis, longitudinal section of the posterior mantle edge (Bouin, AZAN, 5 ym); note the additional ring fold (arrow-heads) and the ciliated ledge cells (arrows); ae, attachment epithelium; go, gonad (ovary); pp, pavillon proper; sgc, subepithelial gland cells; vco, valve ciliary organ; scale bar = 25 pm. They are inflated cell processes with the perikaryon below the epidermis. The cytoplasm looks empty. Elongated mito- chondria are present close to the cell apex, and neurotubules extend vertically through the process. The cell surface shows only small microvilli, but it is provided with two to eight cilia with the common 9x2+2 pattern and a rootless centriole. The outer pavillon epithelium is not as folded as the inner one; thus, its cells are of more constant cuboid shape and size (4-4.5 um). The microvilli covering the surface are about 1.7 um long. Numerous dark vesicles, mitochondria and RER are contained in the cytoplasm. The nucleus is close to the basal lamina. TEM reveals groups of two or three gland 16 AMER. MALAC. BULL. 9(1) (1991) cells in regular intervals. The dark cells closely resemble the secretory cells of the inner epithelium. All examined Gadilida possess an additional ring-fold distal and parallel to the supramarginal groove (Fig. 34). Some of the gadilid species studied show subepithelial gland cells opening through the inner epithelium of the pavillon. The species are Pulsellum lofotense (M. Sars), Wemersoniella turnerae Scarabino, and all Cadulus species (Fig. 34) except Cadulus sp. (NMMH 803468) (Table 3). DISCUSSION FUNCTIONAL ASPECTS The tubular nature of the peripedal pallial cavity can be considered a consequence of a rounding-off process of the mantle-shell-complex when adopting a burrowing habit (Pojeta and Runnegar, 1985; Runnegar and Pojeta, 1974; Run- negar, 1978; Salvini-Plawen, 1981, 1984). This process is recapitulated in scaphopod ontogeny (Lacaze-Duthiers, 1856-7; Kowalevsky, 1883). As in bivalves, the mantle cavity is not only the site of respiration, defecation and excretion, but also of food intake (Dinamani, 1964a, b). The organs of the mantle edges protect the animal against, and put it in touch with, its environment. GLANDS Experimental and detailed histochemical data on the functions of the various gland regions of the mantle epithelium are lacking. Therefore, their functional interpretation must be based on structure and the observations on living animals alone. The main function of the gland cells of the anterior mantle edge is probably mucus production. They closely resemble the mucus cells described in other marine inverte- brates by Storch and Welsch (1972). This resemblance, the staining properties and the observations on living animals, indicate a mucoid nature of the secretions which have the same optical properties as water and can be discharged in copious amounts. Proteinaceous components may also be assumed, as the circum-nuclear cytoplasm is rich in rough endoplasmic reticulum. The basal portions of the cells are frequently associated with muscle fibers. Members of the order Dentaliida, where a prominent outer gland region is present, secrete enormous amounts of acid mucoid sub- stances. The roles of the different types of gland cells are not clear. In Antalis, Fustiaria and Rhabdus the secretion is a sticky mass with the same optical properties as water. Gainey (1972) and Poon (1987) reported a feeding cavity in the sediment made by the foot after the animal is dug in. Gainey (1972) supposed that this cavity might be coated with mucus. At least for Dentaliida it is likely that this coating originates from the outer gland region of the anterior mantle edge. The weakly developed gland cells in the frontal epithelium of Gadilida are hardly capable of providing a mucus sheet to support a feeding cavity. In living Entalina, Pulsellum or Cadulus such massive mucus secretion has never Table 3. Characters of the posterior mantle margin: (++) abundant; (+) present; (+) scarce, few; (—) absent; (?) no information. Aperture Valve Mechanism Pavillon slit slit ring dorsal ventral subepith. ciliated sphincter dorso- ciliary ciliated subepith. addi- hori- verti- sinus flap bolster = gland area muscle ventral organ ledges gland tional zontal cal cells con- cells _ring-fold strictors Antalis + — + + + + + + _ - = + = Dentalium + — + + + + + + 4 ae Fissidentalium + — + + + + + + = = = + hs Graptacme + - + + + + + + — = = + = Fustiaria + — - + = a ne ai pani = Laevidentalium ? a ? ? ? ? ? ? ? ? ? ? ? Rhabdus + - + + + + + + — - = + a Entalina a + + — - + + + Bathoxiphus as + + = aS aie + ae - Heteroschismoides = te = = = ? = aa ot + te 2 ? Pulsellum = + - aa = + a = te an ate + + Annulipulsellum = + = = = + - 7 ap + + 7 + Striopulsellum = + 4 - = + + rt = ip Siphonodentalium a + + - = + + aft = + Wemersoniella = + + - ~ + + + ++ + Gadila - + — - _ + — = + ae 4. = 4. Cadulus a + — - = + _ = a ane a. ae at STEINER: SCAPHOPOD MANTLE ANATOMY 17 been observed. Another function of the mucoid secretions of the anterior mantle edge appears to be the cleansing of the pallial opening from particles and the protection against disturbance by other organisms. The latter view is supported by the obser- vation that moribund animals are invaded by protozoans, e.g. ciliates. The inner gland region, common to all scaphopods, acts as a lubricator for the foot. This is not only important for locomotion and sealing of the pallial aperture but also for feeding. Gainey (1972) reported that the dentaliids Antalis pseudohexagonum Ihering (Henderson) and Graptacme eboreum (Conrad) collect food particles in the mantle cavity by ciliary action of the dorsomedian foot-furrow before they are ingested. The fact that gadilid scaphopods have epithelial as well as subepithelial gland cells in this region could point towards a compensation for the less elaborated outer gland region, compared to dentaliids. The glandular region anterior to the preanal ciliary ridges has been regarded homologous with the hypobranchial gland of other molluscs by Distaso (1906). The absence of gills as a location reference and the lacking detailed struc- tural analysis lend little weight to this hypothesis. However, it is likely that the mucoid secretion produced there lubricates fecal material and/or binds other particles in the mantle cavity before they are removed through one of its openings. It seems justified to assume that the subepithelial gland cells, which are most similar to those of the anterior mantle edge, secrete mucus for cleansing and repellant purposes. Numerous authors (Simroth, 1894; Yonge, 1937; Stasek and McWiliams, 1973; Shimek, pers. comm.) ascribe the ability of shell reabsorption and truncation, as well as the formation of secondary apical shell features (pipes, plugs, lobes or slits) to the gland cells of the pavillon. Although this study provides no evidence for either of these functions, there are clear indications for the temporary nature of most apical shell features; they should therefore not be used as reliable systematic characters. SENSORY RECEPTORS The assumption that the receptor cells in the epithelia of the mantle margins have sensory qualities is based on their similarity with sensory cells of other invertebrates. The com- mon charactes are: an electron lucent cytoplasm of almost empty appearance, the presence of one or more cilia with or without sunken-in bases, neurotubules in parallel orienta- tion to the cell axis, a long, subepithelial, dendritic process, the absence of nuclei in the epithelial region, and, if present, the collar of eight to nine stereomicrovilli. Chemo- and mechanosensory functions are ascribed to these types of ciliary receptor cells. Sensory cells similar to type 1 and la of the anterior mantle edge of Antalis and Fustiaria have been described for the siphons of the bivalves Donax serra Roding, D. sordidus Hanley and Solen capensis Fischer by Hodgson and Fielden (1984, 1986), as well as for Schizochilus caecus L’Hardy (Neorhabdocoela, Kalyptorhyncha) by Ehlers and Ehlers (1977). The collar-receptor of type 2 is of a very common type found in numerous metazoan taxa (Haszprunar, 1985; Salvini-Plawen, 1988). Structures bearing certain similarity with the papillae of the frontal epithelium of Entalina and Pulsellum are reported for the tentacles of the bivalve Placopecten magellanicus (Gmelin) by Moir (1977), although in the lat- ter the papillae exhibit collar-receptors with nuclei. The ciliary organ of the dentaliid anterior mantle margin probably has no sensory qualities. The vividly beating cilia in living animals suggest the function of water current generation. The length of the cilia, compared to the pallial orifice diameter, is not sufficient to produce an effective flow of water in the mantle cavity unless the opening is maximal- ly constricted. Alternatively, the ciliary organ could be responsible for water exchange over the frontal epithelium and thus facilitate chemoreception. The dorsolateral slits in the anterior mantle edge of Rhabdus rectius probably represent true sensory organs. This remains to be confirmed by TEM investigations. Contrary to the opinion of Distaso (1906), no osphradial sense organs are developed in the epithelium beneath the visceral ganglia. Distaso may have misinterpreted the pre-anal ciliary ridges and the medullary nature of the visceral connectives in this area. No detailed descriptions of the sensory structures of the posterior mantle edge of Scaphopoda have been published. Reynolds (1988) gave a short report on the occurrence of multiciliated receptors in Rhabdus rectius (Carpenter), but no comparison with the corresponding structures in Entalina quinquangularis (Forbes) can be made at present. APICAL VALVE-ORGAN ANTALIS - TYPE The arrangement of muscles and connective tissue in Antalis spp. indicates that the width of the opening is regulated by the dorsal section of the annular fold, while the ventral part acts as an abutment for the musculature. The dorsal sec- tion closes the aperture when it is pressed against the ventral part and against the semi-annual projection on the dorsal side. The action of the sphincter alone is probably not responsible for this; blood pressure in the sinus certainly helps to close the valve. The muscle fibers transversing the dorsal section of the sinus open the valve. FUSTIARIA - TYPE Due to the lack of a ventral bolster and a sphincter muscle, rising blood pressure is likely to be the only force 18 AMER. MALAC. BULL. 9(1) (1991) lowering the large dorsal flap. Like in the Antalis-type, decreasing pressure and contraction of the dorsoventral mus- cle fibers lift the flap to open the pallial orifice. ENTALINA - TYPE The closure of the mantle opening to a vertical slit is probably effected by the dorsoventral muscle bundles on either side of the orifice. Dorsally and ventrally attached to con- nective tissue sockets, they constrict the orifice by approx- imating its lateral walls. Increasing haemolymph pressure in the surrounding lacunae is probably synergetic. The antagon- istic force is provided by the radially arranged muscle fibers between inner and outer mantle epithelia. WATER CIRCULATION Respiration, defecation and excretion depend on the currents of water through the mantle cavity. Its surface has to meet the respiratory demands of the organism, as ctenidia are not present. Yonge (1937) gave a detailed account of the currents in the pallial cavity of Antalis entalis. He described the apical orifice as the inhalent and exhalent opening, observing a steady inflow of water due to ciliary action, followed by a violent expulsion, which is caused by a sud- den retraction of the foot. In general, these results can be confirmed, although some complementary remarks can be made. The inhalent water currents are produced by the ciliary organ of the pavillon and the preanal ridges in Gadilida, while in Dentaliida the ciliary beat of the more numerous preanal ridges alone generate the flow. The metachronal waves of the cilia can be easily observed through the transparent shells of Rhabdus, Pulsellum and Cadulus. Contrary to Yonge’s reports, the apically entering water leaves the mantle cavity through the anterior aperture. This was observed in Entalina quinquangularis, Pulsellum lofotense, Cadulus subfusiformis, and Antalis occidentalis (Stimpson). Suspended carmine par- ticles and sperm could be followed through the mantle cavity. During the flow of water through the pallial cavity, the foot can only play a minor role in producing the inhalent cur- rent. The piston-like actions of the foot causing the rapid apical exhalent current occur about twice a minute in relaxed specimens of Antalis and Fustiaria (Dentaliida). The gadilid species showed intervals of one minute or more between ex- pulsions. During the expulsions and during burrowing ac- tion the ciliary beat ceases; it is resumed a few seconds later. SYSTEMATICS IMPLICATIONS Systematics below the order level are solely based on shell and radula characters. Chistikov (1975a, b) attempted to include some criteria of the soft body. The results of the present study provide detailed in- formation on the differences of the mantle edges between the scaphopod orders, but also allow some statements at the fami- ly level. DENTALIIDA The anterior pallial edge of the order Dentaliida is characterized by a conspicuous outer gland region, a strong ring of cartilage-like material, as well as massive fibrous con- nective tissue and muscles in the central fold, a circular ciliary organ, and an inner gland region with epithelial gland cells only. The average number of preanal ciliary ridges ranges from 12 to 15. In most representatives of this order the valve organ of the posterior mantle edge shows a dorsal and a ventral elevation of the pallial wall in combination with an annular blood sinus. Closing of the horizontal slit is achieved by lowering the dorsal flap by means of muscular action and haemolymph pressure. No ciliary tracts are developed on the ledges of the pavillon proper. These features are shared by most of the examined den- taliid genera, such as Antalis, Dentalium, Fissidentalium and Laevidentalium. Thus, analysis of this organ system would suggest the grouping of these genera in the family Dentaliidae. With the exception of Laevidentalium, this grouping agrees with that of Scarabino (1979). The genus Fustiaria differs from the general scheme in three aspects: 1) The radially arranged fibers of connec- tive tissue in the central fold of the anterior mantle edge are not developed; 2) The relative abundances of the two types of sensory cells are completely different, and an additional kind of ciliary receptor is present (type la); 3) The most com- plex difference to the Antalis-type organization concerns the posterior mantle edge. These differences, together with the particularities of the shell, the radula (Scarabino, 1979), and the foot (Steiner, in press), justify the separation of Fustiaria from the Dentaliidae. Thus, a new monogeneric family, the Fustiari- idae, is proposed. FUSTIARIIDAE (FAM. NOV.) DIAGNOSIS — Dentaliid Scaphopoda with smooth, polished, thin-walled and transparent shells of slender shape, moderature curvature and circular cross-section. The ven- tral side of the apex can have a straight, deep slit or a small notch. The rachis tooth of the radula differs from that of Dentaliidae by showing a flat superior edge. The ciliary organ of the anterior pallial margin consists of five to six rows of cells. The posterior mantle orifice can be closed by a dorsal flap. No ventral bolster, sphincter muscle or glands are elaborated there. Instead, subepithelial gland cells are abun- dant in the pavillon. The pedal sinus is divided by a hori- zontal septum. TYPE GENUS — Fustiaria Stoliczka, 1868. The genus Rhabdus, represented by R. rectius in this study, has a STEINER: SCAPHOPOD MANTLE ANATOMY 19 remarkable specialization at the anterior mantle margin. A pair of dorso-lateral invaginations at the periphery of the outer gland region are developed, and the ciliary organ is wanting. The slit-like invaginations can either be considered denovo differentiations of the outer gland replacing the lost ciliary organ, or a differentiation of the ciliary organ itself. The family Rhabdidae was erected by Chistikov (1975a), who emphasized the peculiar morphology of radula and shell. The special elaboration of the anterior pallial edge supports the separation from the other dentaliid Scaphopoda. GADILIDA The anterior mantle edge of Gadilida differs from the corresponding organ of Dentaliida by the lack of an outer gland region and the ciliary organ. The frontal epithelium is elaborated into papillae which carry ciliary sensory cells. The amount of cartilage-like and fibrous connective tissue in the central fold varies species-specifically. The inner gland region exhibits subepithelial as well as epithelial gland cells. The average number of ciliary rings in the preanal area (4 to 8) is lower than in the Dentaliida. The posterior pallial edge is characterized by a power- ful ciliary organ of large ciliated cells arranged in a single row encircling the mantle opening. Dorsoventral muscle fibers close the aperture by approximating the lateral walls of the pallial opening, thus producing a vertical slit. No distinct an- nular blood sinus is developed. The ledges of the pavillon proper are lined by ciliary tracts. At present the only anatomical character suitable for a classification within the Gadilida is the different abundances of papillae in the frontal epithelium of the anterior mantle edge. These structures were found in great numbers in the examined species of Entalina, Bathoxiphus, Heteroschis- moides, Pulsellum, Annulipulsellum, Striopulsellum and Wemersoniella. Comparatively few papillae are developed in Siphonodentalium, Cadulus and Gadila. For the common ancestor of Dentaliida and Gadilida the following differentiations of the mantle can be assumed: 1) a rigid ring formed by connective tissue and muscles to support the anterior pallial opening, 2) an area of epithelial gland cells inside the mantle cavity (= inner gland region), 3) several ciliated, annular ridges in the preanal region, perhaps preceded by a ventral glandular field (=hypobranchial gland of Distaso), 4) the posterior mantle opening with an elaborated dorsal extension, the pavillon proper. According to this list, the outer gland region and the presence of a ciliated organ at the anterior pallial aperture are differentiations within the Dentaliida, while the subepithelial gland cells of the inner gland region and the ciliation of the pavillon proper have developed in Gadilida only. ACKNOWLEDGMENTS I thank V. Scarabino and B. Metivier (Mus. Hist. Nat., Paris), B. Marshall (Nat. Mus., New Zealand). J. Knudsen (Zoologisk Museum, Kopenhagen), G. Davies (University of Edinburgh) and R. Houbrick and D. Bohmhauer (Nat. Mus. Nat. Hist., Washington, D.C.) for generously loaning scaphopod specimens. Collection of living Scaphopoda was finan- cially supported by travel-grants of the Bundesministerium fiir Wissenschaft und Forschung and by a grant-in-aid of Sigma Xi, The Scientific Research Society. I am indebted to my supervisor, Professor L. Salvini-Plawen, for his help and for the critical reading of the manuscript. LITERATURE CITED Adam, H. and G. Czihac. 1964. Arbeitsmethoden der makroskopischen und Mikroskopischen Anatomie. Fischer Verlag, Stuttgart. 583 pp. Boissevain, M. 1904. Beitrage zur Anatomie und Histologie von Dentalium. Jenaische Zeitschrift fur Naturwissenschaften 38:553-572. Chistikov, S. D. 1975a. Some problems of scaphopod taxonomy. In: Molluscs, Their Systematics, Evolution and Significance in Nature, pp. 21-23. Academii Nauka CCCP (Moskow). Chistikov, S. D. 1975b. Some problems in the classification of the order Dentaliida (Mollusca; Scaphopoda). Fifth Meeting on Molluscs, Len- ingrad, 1975. Malacological Review 1978, \1:71-73 (abstract). Deshayes, G. P. 1825. Anatomie et monographie du genre Dentale. Memoires de la Societe d’histoire naturelle de Paris 3:1-54. Dinamani, P. 1964a. Burrowing behavior in Dentalium. Biological Bulletin 126:28-32. Dinamani, P. 1964b. Feeding in Dentalium conspicuum. Proceedings of the Malacological Society, London 36(1):1-5. Distaso, A. 1906. Sull anatomie degli Scaphopoda. Zoologischer Anzeiger 29:271-278. Ehlers, U. and B. Ehlers. 1977. Monociliary receptors in interstitial Pro- seriata. and Neorhabdocoela (Turbellaria, Neoophora). Zoo- morphology 86:197-222. Emerson, W. K. 1951. Nomenclatural notes on the Scaphopoda: the subgenus Dentale DaCosta. Nautilus 65(1):17-20. Emerson, W. K. 1952. Nomenclatural notes on the scaphopod Mollusca: the type species of Fustiaria and Pseudantalis. Proceedings of the Biological Society of Washington 65:201-208. Fol, H. 1889. Sur ’'anatomie microscopique du Dentale. Archives de Zoologie Experimentale ét Generale (2)7:91-148. Gabe, M. and M. Prenant. 1950. Données histologiques sur le tissue con- jonctiv du Dentale. Bulletin de la Societe Zoologique de France 75:10-18. Gainey, L. F., Jr. 1972. The use of foot and captacula in the feeding of Dentalium. Veliger 15:29-34. Haszprunar, G. 1985. The fine morphology of the osphradial sense organs of the Mollusca. I. Gastropoda: Prosobranchia. Philosophical Trans- actions of the Royal Society, London B307:457-496. Hodgson, A. N. and L. J. Fielden. 1984. The structure and distribution of peripheral ciliated receptors in the bivalve Donax serra and Donax sordidus. Journal of Molluscan Studies 50:104-112. Hodgson, A. N. and L. J. Fielden. 1986. The ultrastructure of ciliated cells from the siphon of Solen capensis (Mollusca, Bivalvia). Journal of Molluscan Studies 52:161-168. Kowalevsky, M. A. 1883. Etude d’embryologie du Dentale. Annales du Musee d Histoire Naturelle de Marseille Z.\(7):\-54. Lacaze-Duthiers, H. 1856-7. Histoire de organisation et du developpement du Dentale. Annales des Sciences Naturelles, Paris 4, Zool. 6:225-281; 7:5-S1, 171-225; 8:18-44. Leon, H. 1895. Zur Histologie des Dentaliummantels. Jenaische Zeitschrift 29:411. Moir, A. J. G. 1977. Ultrastructural studies on the ciliated receptors of the long tentacles of the giant scallop, Placopecten magellanicus (Gmelin). Cell Tissue Research 184:367-380. Mollenhauer, H. H. 1964. Plastic embedding mixtures for use in electron microscopy. Stain Technology 39:111-114. Odhner, N. H. 1931. Die Scaphopoden. Further Zoological Results of the Swedish Antarctic Expeditions 1901-1903 2(5):1-8. Pilsbry, H. A. and B. Sharp. 1897. Scaphopoda. Manual of Conchyology 17:1-280. Plate, L. H. 1892. Uber den Bau und die Verwandtschaftsbeziehungen der Solenoconchen. Zoologische Jahrbiicher der Anatomie 5:301-386. Pojeta, J. and B. Runnegar. 1985. Origin and Diversification of the Mollusca. In: The Mollusca, Vol. 10. Trueman and Clarke, eds. pp. 1-57. Academic Press, London. Poon, P. A. 1987. The diet and feeding behavior of Cadulus tolmiei Dall, 1897. Nautilus 101(2):88-92. Reynolds, P. D. 1988. The structure and distribution of ciliated sensory recep- tors in the Scaphopoda (abstract). American Zoologist 28(4):140A. Runnegar, B. 1978. Origin and evolution of the class Rostroconchia. Philosophical Transactions of the Royal Society London B.284:319-333. Runnegar, B. and J. Pojeta. 1974. Molluscan phylogeny: The paleontological viewpoint. Science 186(4161):311-317. Salvini-Plawen, L. 1981. On the origin and evolution of the Mollusca. In: Atti dei Convegni Lincei; Convegno Internazionale ‘‘Origine dei grandi Phyla dei Metazoi’: pp. 235-293. Accademia Nazionale dei Lincei, Roma. Salvini-Plawen. L. 1984. Die Cladogenese der Mollusca. Mitteilungen der 20 AMER. MALAC. BULL. 9(1) (1991) deutschen malakologischen Gesellschaft 37:89-118. Salvini-Plawen, L. 1988. Annelida and Mollusca - a prospectus. Microfauna Marina 4:383-396. Scarabino, V. 1979. Les scaphopodes bathyaux et abyssaux de 1’Atlantic oc- cidental: Nouvelle classification pour l’ensemble de la Class. Doc- toral Dissertation, Universite d’Aix-Marseille. 154 pp. Shimek, R. L. 1989. Shell morphometrics and systematics: a revision of the slender, shallow-water Cadulus of the North-Eastern Pacific. Veliger 32(3):233-246. Simroth, H. 1894. Scaphopoda. Jn: Klassen und Ordnungen des Thierreichs, Bronn, G. ed. Vol. 31:356-467. Spurr, A. R. 1969. A low viscosity epoxy resin embedding medium for elec- tron microscopy. Journal of Ultrastructural Research 26:31-41. Stasek, C. R. and R. W. McWilliams. 1973. The comparative morphology and evolution of the molluscan mantle edge. Veliger 16(1):1-19. Steiner, G. In press. The organizations of the scaphopod pedal musculature. Journal of Molluscan Studies. Storch, V. and U. Welsch. 1972. The ultrastructure of epidermal mucus cells in marine invertebrates (Nemertini, Polychaeta, Prosobranchia, Opisthobranchia). Marine Biology 13:167-175. Venable, J. H. and R. Coggeshall. 1965. A simplified lead citrate stain for use in electron microscopy. Journal of Cell Biology 25:407-408. Watson, M. L. 1958. Staining of tissue sections for electron microscopy with heavy metals. Journal of Biophysical and Biochemical Cytology 4:475. Yonge, C. M. 1937. Circulation of Water in the Mantle Cavity of Dentalium entalis. Proceedings of the Malacological Society, London 22(6):333-336. Date of manuscript acceptance: 28 April 1991 The freshwater mussels (Bivalvia: Unionoidea) of the upper Delaware River drainage David L. Strayer and Jonathan Ralley Institute of Ecosystem Studies, The New York Botanical Garden, Box AB, Millbrook, New York 12545, U.S.A. Abstract. We surveyed the unionoidean fauna of 44 sites in the upper Delaware River drainage of New York during 1990. Seven species of unionoideans were found living in the basin, including the endangered Alasmidonta heterodon (Lea, 1829). Four other species are known historically from the upper Delaware basin, but now are either absent from the area or restricted to a few local sites. Neither calcium concentration nor stream size is a good predictor of unionoi- dean species richness in the study area. We hypothesize that sediment stability could regulate the occurrence of unionoideans in the streams of the upper Delaware drainage. The Delaware River is one of the major rivers of the northern Atlantic Slope. Ortmann (1919) showed that the lower Delaware basin contained a rich Atlantic Slope fauna, includ- ing the endangered species Alasmidonta heterodon (Lea, 1829) and the southernmost known population of Margaritifera margaritifera (Linneaus, 1758). Because the up- per Delaware is known for its high water quality, we felt that similarly rich communities of unionoideans could live up- stream of the area surveyed by Ortmann. There is little pub- lished information on the unionoideans of the upper Delaware basin. Marshall (1895) reported six species from unspecified sites in the Delaware River system in New York. Harman (1975) published a brief article focusing on the effects of anthropogenic disturbances on the molluscan community of the Delaware’s headwaters. We surveyed the waters of the up- per Delaware drainage in New York in 1990 to determine whether A. heterodon lived in this area and to assess the cur- rent status of the freshwater mussel community in general. THE STUDY AREA Our survey covered the streams in the Delaware River basin in New York (Fig. 1). Streams range in size from head- water brooks to the Delaware River itself, which has a mean annual discharge of 160 m?/sec at Port Jervis (our station 1) (Zembrzuski et al., 1983). Most streams in the study area have fairly high gradients, and sediments consist chiefly of cobbles, gravel, and coarse sand. The water in most streams is very clear and somewhat soft (Table 1). Most of the watershed is forested, although there is some agriculture, and villages and small cities are scattered along the Delaware River and its major tributaries. The largest municipalities in the basin are Port Jervis (pop. 8699), Mon- ticello (6306), and Hancock (1526), so urban pollution is not pronounced. The larger streams in the upper Delaware basin are used heavily for recreation (boating, fishing). The major current anthropogenic impacts on the streams in the drainage probably arise from the three large reservoirs of the New York City water supply system. These reservoirs alter the hydro- logical and thermal characteristics (all three reservoirs are hypolimnetic release) of downstream waters (the lower East and West Branches of the Delaware River, the upper mainstem of the Delaware River, and the middle Neversink River). About 30 m?/sec of water is diverted out of the basin from these reservoirs to supply drinking water for New York City (Zembrzuski et al., 1983). METHODS We visited 44 sites on the upper Delaware River drainage during periods of low, clear water between July and September, 1990, collecting mussels by handpicking while wading or snorkeling. Most specimens were identified and returned immediately to the stream. Voucher specimens (chiefly dead shells) have been deposited in the New York State Museum (NYSM) and Academy of Natural Sciences at Philadelphia (ANSP). In addition to our field collections, we searched the collections of the National Museum of Natural History (USNM), American Museum of Natural History (AMNH), and NYSM for specimens of unionoideans from the upper Delaware basin. Mussel nomenclature follows that of Turgeon et al. (1988). Water samples were collected in clean polyethylene bottles and analyzed for calcium by plasma emission using a Perkin-Elmer ICP/6000. American Malacological Bulletin, Vol. 9(1) (1991):21-25 pA N ie) Fig. 1. Location of the Delaware River basin and sampling sites on the up- per Delaware River basin. Inset shows the Delaware (D) and nearby drainages (H = Hudson, S = Susquehanna) in New York, New Jersey, and Penn- sylvania; the study area is stippled. Site numbers on the main figure corre- spond to those given in Table 1. Open circles show sites where unionoideans were not found, small black circles show sites where only Elliptio complanata was found, and large black circles show sites where at least two species of unionaceans were found. Dotted lines are county boundaries. RESULTS AND DISCUSSION The waters of the upper Delaware River drainage now support seven species of unionoideans (Table 1). Four other species are known only through historical records. Marshall (1895) reported Ligumia nasuta (Say, 1817), Lampsilis cariosa (Say, 1817), and L. radiata (Gmelin, 1791) from the upper Delaware system, but we saw no trace of these species in 1990 [Marshall’s report of Anodontoides ferussacianus (Lea, 1834) probably is based on a misidentification of an Anodonta sp.]. Margaritifera margaritifera is represented by a single shell (AMNH 164659) taken from ‘‘lake at Camp Welmet near Narrowsburg, Sullivan Co., NY”’ by H. S. Feinberg in 1949. Unfortunately, we were unable to get access to the lake at Camp Welmet (Silver Lake, not Lake Welmet, which, con- fusingly enough, is not on the property of Camp Welmet) in 1990 to assess the status of this population. Of the seven species still living in the upper Delaware basin, Elliptio complanata is by far the most abundant and widespread. In fact, we found E. complanata at every site where unionoideans were present. Although many authors have commented on the broad ecological tolerances of this species (e.g. Ortmann, 1919; Clarke and Berg, 1959; Strayer, 1987), we know surprisingly little about what behavioral, physiological, or ecological adaptations allow this species to succeed over such a broad range of habitats. There are old, indefinite reports of Alasmidonta AMER. MALAC. BULL. 9(1) (1991) heterdon from New York (Marshall, 1895; Letson, 1905), but ours are the first reliable records of this species from New York (there are no museum lots of this species from New York in the NYSM, AMNH, USNM, ANSP, University of Michigan Museum of Zoology, or Museum of Comparative Zoology collections.) There is apparently a healthy popula- tion of this species in the lower 12-18 km of the Neversink River. Our findings raise the obvious possibility that A. heterdon could still persist in other tributaries of the upper Delaware (or in the river itself) in New Jersey or Pennsylvania. The other two species of Alasmidonta (A. undulata and A. varicosa) also were abundant in the Neversink River drainage. In addition, we found A. undulata in the West Branch of the Delaware River above Cannonsville Reservoir. It is possible that small numbers of A. varicosa may live in the upper West Branch as well, although we did not find it there in 1990. Harman (1975) reported A. marginata Say, 1818, a species that resembles A. varicosa, but which probably does not occur in the Delaware basin, in the upper West Branch. Anodonta implicata is found in small numbers in the lower Neversink River near Port Jervis. A. implicata is parasitic on anadromous shad and herring (Alosa spp.), and is found typically in low-gradient coastal rivers and ponds (Johnson, 1946; Davenport and Warmuth, 1965; Smith, 1985; Strayer, 1987). Our records from the upper Delaware River system are interesting for two reasons. First, the reach of the Neversink River occupied by A. implicata is a relatively high- gradient, stony, upland river, unlike the coastal sites typical- ly frequented by this species. Second, although large numbers of American shad [Alosa sapidissima (Wilson, 1811)] run upstream to well above the junction of the East and West Branches of the Delaware, we found no trace of A. implicata in most of the mainstem, even in such apparently suitable habitat as the huge, quiet pool at Narrowsburg. This obser- vation suggests that some ecological factor other than the distribution of the host fish determines the current distribu- tion of A. implicata in the Delaware system. One of the most striking impressions from our work was just how poor the unionoidean communities were over large parts of the upper Delaware basin. Unionoideans appar- ently were absent at many (30%) of the sites that we surveyed, even though the streams were large enough to support union- oideans and were not obviously polluted. In some of these cases (e.g. stations 30, 33), the sediments consisted mainly of well rounded cobbles, and probably are too coarse and too unstable for unionoideans. Other sites (e.g. stations 18-20) apparently have suitable substrata, high water quality, and diverse fish communities, but no trace of unionoideans. We do not know what is keeping unionoideans out of these sites. One obvious possibility that we believe we can rule out is inadequate dissolved calcium. Although many waters in the basin are soft, there is no relationship between calcium con- STRAYER AND RALLEY: DELAWARE BASIN MUSSELS 23 Table 1. Distribution of unionid bivalves in the upper Delaware River basin in 1990 (numbers show numbers of living animals collected; d, old, dead shells found; D, recently dead shells found). & $ & = ¥ ¢ § Ny o o VS) v o) > y $ 2 Wi + & § C . S 3 Sy & o. e ‘ S § § cS S$ SE EF S& Fo SE é £ €§& EF $F #§ FF SF FS 1. Delaware River, Port Jervis 8.7 1.5 2 d 2. Delaware River, Mongaup 1.2 28 3. Delaware River, Pond Eddy 1 26 4. Delaware River at Roebling Bridge near Minisink dd 0.5 D 5. Delaware River, Narrowsburg Te. 2:2 d 6. Delaware River, Skinner’s Falls 1 2 7. Delaware River, Stalker (PA) 1 D 8. Delaware River, Lordsville Ae, 0.5 d 9. Neversink River, Port Jervis 2:2 4 10. Neversink River, Huguenot 7.8 3 128 2 1 2 2 11. Neversink River, Graham Road 6 64 1 20 3 5 12. Neversink River, Roses Point 6 52 3 76 1 26 13. Neversink River, Oakland Valley 5.8 1.5 1 14. Neversink River, Bridgeville 5.6 1.5 D 1 15. Neversink River at mouth of Sheldrake Stream 235 57 51 2 16. Neversink River, Ranch Hill Road 2 2 11 1 17. Neversink River, Woodbourne 3.9 1.5 D 18. Basher Kill, Galley Hill Road 0.7 19. Basher Kill, Westbrookville 1 20. Basher Kill, Wurtsboro 7.4 0.8 21. Delaware and Hudson Canal, Bova Road 0.7 5 22. Sheldrake Stream at mouth 7.8 1.3 41 18 23. Sheldrake Stream, Thompsonville 1 2 24. Sheldrake Stream, Ranch Hill Road 1 9 D 25. Mongaup River, Route 97 0.3 26. Mongaup River south of Swinging Bridge Reservoir 5.9 0.8 27. West Branch Mongaup River, Gale Road 7.9 0.5 28. Ten Mile River below Route 97 0.2 29. East Branch Ten Mile River, County Rte. 23 3.6 0.5 2 30. Callicoon Creek at mouth 9.2 0.5 31. Callicoon Creek, Hortonville 0.5 32. East Branch Callicoon Creek below Route 52 8.2 0.5 33. North Branch Callicoon Creek 2 miles above Hortonville 9.6 0.2 34. East Branch Delaware River, Peas Eddy we) 1 60 35. East Branch Delaware River, Fish’s Eddy 1.5 11 36. East Branch Delaware River, Downsville 5.8 0.5 37. East Branch Delaware River, Margaretville 8.7 0.5 38. West Branch Delaware River, Hancock 1 D 39. West Branch Delaware River, Hale Eddy 6.8 1 1 Table 1. (continued) AMER. MALAC. BULL. 9(1) (1991) é * ra s $ § 8 s Z ie ng S Vy 40. West Branch Delaware River, Deposit 0.5 d 41. West Branch Delaware River, Walton 11.5 1 D 42. West Branch Delaware River, Hamden 2.5 2 43. West Branch Delaware River, Delhi 14.9 2 55 44. West Branch Delaware River, Bloomville 14.9 1 d centration and either unionoidean density or species richness (Fig. 2). Harman (1975) believed that the operation of the New York City water supply reservoirs eliminated most unionoi- deans from tailwater reaches, but our richest sites were downstream of such a reservoir on the Neversink River. A second piece of evidence that shows the poverty of the upper Delaware River unionoidean community is that 58% of the sites that contain unionoideans contain only one species, Elliptio complanata. The dominance of E. com- planata is especially striking in the main Delaware River, where we found only a single, old shell of Anodonta implicata along with more than 500 living or recently dead specimens of E. complanata. The main Delaware is a large river with a rich fish fauna, and would be expected to sup- port several (6-10) species of unionoideans, as was the case on the Susquehanna and lower Delaware rivers (Ortmann, 1919; Clarke and Berg, 1959; Harman, 1970). Clarke (1986) recently found that parts of the upper Connecticut River that formerly supported several species of unionoideans now con- 8 species richness calcium (mg/l) Fig. 2. Species richness of unionoideans in streams of the upper Delaware basin as a function of calcium concentrations (r = 0.22, NS). Y cy & Qy s < > < ¥ ~S Q AS < e y & & < RY Fo” AN eos c ao S Ra NES ve ¥S Nia SES ai os £ § Ys &9 OS Vv ¥ a £ ey ee & v .o :_ y oar R a vyY VS Vw Ve Ve Se tain only E. complanata. He suggested that the operation of hydroelectric dams could have eliminated most of the unionoidean species without, however, suggesting why E. complanata would be resistant to these recent environmental changes. Marshall’s (1895) old records of species such as Lampsilis cariosa from the basin suggest that some union- oidean species could have likewise been extirpated from the mainstem Delaware River. It is possible that past episodes of pollution, perhaps from wood processing industries in the Delaware basin (Myers, 1986), could have destroyed the unionid fauna in some streams, but it is unclear why only E. complanata would be able to recolonize these reaches once the pollution stopped. Finally, unionoideans are highly localized even in sites where several species are present (e.g. stations 10-12). At these sites, there often are sharp boundaries between dense (> 1 individual/m*), multispecific beds of mussels and areas en- tirely devoid of mussels. These sharp boundaries do not generally correspond to obvious changes in environmental conditions (e.g. sediment grain size, current velocity, water depth), although a detailed, quantitative study like that of Salmon and Green (1983) probably would uncover statistically significant differences in environmental conditions between mussel beds and nearby areas devoid of mussels. An alterna- tive hypothesis is that the mussel beds represent areas of relatively stable sediments. It is well known (e.g. Leopold et al., 1964; Richards, 1982) that most stream sediments are set in motion by floods every year or two. The instability of sediments poses obvious problems for the long-lived Unionoidea, which could be displaced, crushed, or buried when the sediments in which they live are moved. Vannote and Minshall (1982) showed that sediment stability was a ma- jor factor regulating the local distribution of mussels in the Salmon River canyon, Idaho. We suggest that sediment stabili- ty is generally important to mussels in streams, and that the STRAYER AND RALLEY: DELAWARE BASIN MUSSELS 2D highly local mussel beds that we observed in the Neversink and elsewhere represent not particularly favorable conditions of sediment grain size, current velocity, and so on, but rather areas in which the sediments have not been moved for some time (a decade or so?), or are stable during critical periods such as during recruitment of juveniles. We have devoted some space to this speculative discus- sion of potential controlling factors because we feel that, despite a large volume of research on unionoidean ecology, there is little real understanding of what controls unionoidean distribution and abundance in streams. Why does the lower Neversink River contain a rich community of unionoideans, including an endangered species, while other apparently suitable sites nearby support only one species or no unionoi- deans at all? Until we can answer questions like these, it will be difficult to formulate intelligent management schemes to protect our remaining unionoidean communities. ACKNOWLEDGMENTS We are grateful to the USNM, AMNH, NYSM, Paul Greenhall, Walter Sage, and Norton Miller for allowing us to examine their collections of unionoideans, to Ed Blakemore for help in depositing voucher snecimens, and to Kathy Schneider and Paul Novak for help in the field. Kathy Schneider also provided encouragement and logistical support. This work was funded by the United States Fish and Wildlife Service through The New York Natural Heritage Program of the Department of Environmental Conservation, with additional support from a grant to the Institute of Ecosystem Studies from the General Reinsurance Corporation. This is a contribution to the program of the Institute of Ecosystem Studies of The New York Botanical Garden. LITERATURE CITED Clarke, A. H. 1986. Unionidae of the upper Connecticut River, a vanishing resource. Nautilus 100:49-53. Clarke, A. H. and C. O. Berg. 1959. The freshwater mussels of central New York. Memoirs of the Agricultural Experimental Station, Cornell University 367:1-79. Davenport, D. and M. Warmuth. 1965. Notes on the relationship between the freshwater mussel Anodonta implicata Say and the alewife Pomolobus pseudoharengus (Wilson). Limnology and Oceanography 10:R74-R78. Harman, W. N. 1970. New distribution records and ecological notes on central New York Unionacea. American Midland Naturalist 84:46-58. Harman, W. N. 1975. The effects of reservoir construction and canalization on the mollusks of the upper Delaware watershed. Bulletin of the American Malacological Union for 1974:12-14. Johnson, R. I. 1946. Anodonta implicata Say. Occasional Papers on Mollusks 1:109-116. Leopold, L. B., M. G. Wolman and J. P. Miller. 1964. Fluvial Processes in Geomorphology. W. H. Freeman, San Francisco. 522 pp. Letson, E. J. 1905. Check list of the Mollusca of New York. Bulletin of the New York State Museum 88:1-112. Marshall, W. B. 1895. Geographical distribution of New York Unionidae. 48th Annual Report of the New York State Museum: 47-99. Myers, F. D. 1986. The Wood Chemical Industry in the Delaware Valley. Prior King Press, Middletown, New York. Unpaginated. Ortmann, A. E. 1919. A monograph of the naiades of Pennsylvania. Part III. Systematic account of the genera and species. Memoirs of The Carnegie Museum 8:1-384. Richards, K. 1982. Rivers: form and process in alluvial channels. Methuen, London. 361 pp. Salmon, A. and R. H. Green. 1983. Environmental determinants of unionid clam distribution in the Middle Thames River, Ontario. Canadian Journal of Zoology 61:832-838. Smith, D. G. 1985. Recent range extension of the freshwater mussel Anodonta implicata and its relationship to clupeid fish restoration in the Con- necticut River system. Freshwater Invertebrate Biology 4:105-108. Strayer, D. 1987. Ecology and zoogeography of the freshwater mollusks of the Hudson River basin. Malacological Review 20:1-68. Turgeon, D. D., A. E. Bogan, E. V. Coan, W. K. Emerson, W. G. Lyons, W. L. Pratt, C. F E. Roper, A. Scheltema, F. G. Thompson and J. D. Williams. 1988. Common and scientific names of aquatic in- vertebrates from the United States and Canada: mollusks. American Fisheries Society Special Publication 1\6:\-277. Vannote, R. L. and G. W. Minshall. 1982. Fluvial processes and local lithology controlling abundance, structure, and composition of mussel beds. Proceedings of the National Academy of Sciences 79:4103-4107. Zembrzuski, T. J., P. M. Burke, R. J. Archer and J. A. Robideau. 1983. Water resources data. New York Water year 1982. Volume 1. Eastern New York excluding Long Island. United States Geological Survey Water Data Report NY-82-1:1-267. Date of manuscript acceptance: 27 March 1991 The Naiades (Bivalvia: Unionoidea) of the Delmarva Peninsula Clement L. Counts, III’; Thomas S. Handwerker? and Roman V. Jesien* 'Coastal Ecology Research Laboratory, Department of Natural Sciences, University of Maryland Eastern Shore, Princess Anne, Maryland 21853, U. S. A. 2Small Farms Institute, Department of Agriculture, University of Maryland Eastern Shore, Princess Anne, Maryland 21853, U. S. A. 3Center for Estuarine and Environmental Studies, University of Maryland System, Cambridge, Maryland 21613, U. S. A. Abstract. A survey comprising 307 stations of the freshwaters of the Delmarva Peninsula of Delaware, Maryland, and Virginia was conducted to deter- mine the species diversity and zoogeographic distribution of unionid bivalves. The unionid fauna included Elliptio fisheriana (Lea, 1838), E. complanata (Lightfoot, 1786), Lampsilis radiata (Gmelin, 1791), Leptodea ochracea (Say, 1817), Anodonta cataracta Say, 1817, A. implicata Say, 1829, Ligumia nasuta (Say, 1817), Strophitus undulatus (Say, 1817), Alasmidonta undulata (Say, 1817), and the rare and endangered A. heterodon (Lea, 1830). Although a review of the literature and a survey of museum collections revealed records for FE. dilatata (Rafinesque, 1820), L. cariosa (Say, 1817) and A. varicosa (Lamarck, 1819), these species were not found during the field survey. No populations of freshwater unionids were found south of the Maryland/Virginia state line. Comparisons of collections in this study with those reported in the literature and in museums indicate a general decline in diversity. The federally listed rare and endangered A. heterodon is reported from the Choptank drainage in Maryland. No populations of the zebra mussel, Dreissena polymorpha (Pallas, 1771) were found. Rhoads (1904), in a brief examination of the molluscan fauna of Delaware, remarked that the literature on the molluscan fauna of Delaware was practically nonexistent. This was true in 1904 and is still true nearly 80 years later, not only for Delaware but for the entire Delmarva Peninsula. Rhoads (1904), after collecting at three localities, reported the presence of ten unionid species: Lampilis nasutus (Say, 1817) [= Ligumia nasuta (Say, 1817)], L. radiatus (Gmelin, 1791) [= L. radiata (Gmelin, 1791)], L. cariosus (Say, 1817) [= L. cariosa (Say, 1817)], L. ochraceus (Say, 1817) [= L. ochracea (Say, 1817)], Unio complanatus (‘Solander’ Dillwyn, 1817) [= Elliptio complanata (Lightfoot, 1786)], U. fisherianus Lea, 1838 [= E. fisheriana (Lea, 1838)], Anodonta cataracta Say, 1817, Strophitus edentulus (Say, 1829) [= S. undulata (Say, 1817)], S. undulatus, and Alasmidonta marginata varicosa (Lamarck, 1819) [= A. varicosa (Lamarck, 1819)]. This list remains the most complete faunal record for the naiades of the entire Delmarva peninsula. Since the report of Rhoads, several other works have mentioned the unionid fauna of the peninsula only in pass- ing (Ortmann, 1919; Johnson, 1970; Davis and Fuller, 1981; Davis et al., 1981; Clarke, 1981; Davis, 1984). Ortmann (1919) mentioned Elliptio fisheriana from the Chester River system of Maryland. Johnson (1970), in his treatment of the freshwater bivalves of the Atlantic Slope, documented only the unionid species inhabiting the waters draining the western shore of Chesapeake Bay. He (Johnson, 1970) made reference to only E. fisheriana (Lea, 1838), which he considered a junior synonym of FE. lanceolata, as having its type locality at the head of the Chester River in Kent Co., Maryland. Davis and Fuller (1981) reported Lampilis radiata (Gmelin, 1791) from Sussex Co., Delaware, without more precise locality data. Davis (1984) reported FE. fisheriana and EF. dilatata (Rafinesque, 1819) from Concord Pond, Sussex Co., Delaware, E. complanata from Deep Creek, Sussex Co., Delaware, and the Sassafras River, Kent Co., and Chester River, Queen Anne’s Co., Maryland. Clarke (1981) noted the presence of Alasmidonta undulata at ‘*Choptank Mills,’ Kent Co., and A. varicosa at the head of Red Clay Creek, New Castle Co., Delaware. These few reports constitute the published informa- tion available on the freshwater naiades of the Delmarva Peninsula. Given this paucity of information, the potential threat of exotic species introductions, e.g. Dreissena poly- morpha (Pallas, 1771) (Counts et al., 1991; Handwerker and Counts, 1991), and the geographic features unique to the peninsula, the present study was undertaken to provide a re- cent baseline study of species diversity and zoogeographic distribution of the freshwater unionids indigenous to the Delmarva Peninsula. The distribution of the Asian clam, Cor- bicula fluminea (Miller, 1774) on the Delmarva Peninsula is discussed in Handwerker et al. (1991). American Malacological Bulletin, Vol. 9(1) (1991):27-37 27 28 AMER. MALAC. BULL. 9(1) (1991) UESCRIPTION OF THE STUDY AREA The Delmarva Peninsula, as defined here, extends from Cecil Co., Maryland, east of the Susquehanna River, and New Castle Co., Delaware, and includes all that land lying between the Susquehanna River and Chesapeake Bay on the west and the Delaware River and Bay and Atlantic Ocean on the east, south to Cape Charles, Virginia. The peninsula is bisected on an east-west axis by the brackish waters of the Chesa- peake-Delaware Canal. Hence, the lower portion of the penin- sula is an island. The study area included the entire state of Delaware, the nine Eastern Shore counties of Maryland (from Cecil Co. in the north to Somerset - Worcester counties in the south) and Virginia (Accomack and Northampton coun- ties). Geologically, the Delmarva Peninsula is composed of deposits ranging from the Late Cretaceous Potomac Group in the north to the Pleistocene sands of the southern half of the peninsula (Stephenson ef al., 1933). The freshwaters of the Delmarva Peninsula drain into the Chesapeake Bay on the west or into the Delaware Bay or Atlantic Ocean on the east. River systems draining into Chesapeake Bay are generally longer and wide than those of the eastern portions of the peninsula that drain into the Atlantic Ocean, either directly or via the Delaware River and Bay. Regardless of the drainage, the streams of the peninsula are tidal and saline for major portions of their length. The transition from brackish to freshwater is usually abrupt and occurs at mill dams. In some cases (e.g. the Sassafras River) the tidal, brackish portion of the stream is substantially longer than the freshwater. The freshwater drainages of the Maryland portion of the peninsula have been described by Carpenter (1983). Generally, streams flow at a slow rate (0.25 - 3.74 m?/sec) and have a maximum discharge rate ranging between 30.0 - 212.25 m?/sec and minimum discharge rates ranging between 0 - 0.37 m?/sec (Carpenter, 1983). All ponds of the Delmarva Peninsula were originally impounded to store water for livestock and to power grain mills or, in the case of those on the upper peninsula, manufac- turing facilities. While none of the mills are now operational, the ponds have been preserved and many are maintained as recreational areas. The Delmarva Peninsula is also the northern-most point at which cypress swamps occur on the Atlantic coast. Many of the streams of southern Delaware and Somerset, Wicomico, and Worcester counties, Maryland (e.g. Nanticoke River, Pocomoke River, Dividing Creek), drain these swamps and these streams are typical ‘‘blackwater’’ systems. Many of the freshwater streams of the peninsula are channelized and characterized by steep banks that lack vegeta- tion other than grasses. These streams, and their headwaters, serve to drain cultivated fields. Because of the intense agricultural development of the peninsula, freshwaters show a high degree of eutrophication south of the Chesapeake- Delaware Canal. Most of the stream systems in this agricultural region are little more than drainage ditches between cultivated fields, particularly those of the Virginia counties. METHODS A review of the literature (Rhoads, 1904; Ortmann, 1919; Clarke, 1981; Davis and Fuller, 1981; Davis, 1984) was conducted in conjunction with a survey of the unionid col- lections of the Academy of Natural Sciences of Philadelphia (ANSP) and Delaware Museum of Natural History (DMNH) for collections made on the Delmarva Peninsula. This museum survey was conducted to verify published records and to record and consolidate any unpublished records reflected by these collections. These records were collated and localities listed were then surveyed for the presence of species historically reported or collected. A survey of major drainage systems of the Delmarva Peninsula (Appendix, Fig. 1) was conducted from August 1989 through August 1990. The survey included 307 stations. A description of all stations surveyed is on file at the Academy of Natural Sciences of Philadelphia. Stations were defined as the section of the stream 100 m above and below the point Fig. 1. Station locations at which naiades were collected in the northern half of the Delmarva Peninsula (Delaware and Maryland). COUNTS ET AL.: DELMARVA NAIADES 29 at which the stream was entered. Stream surveys included tidal and brackish water portions of streams where these con- ditions occurred, along the Chesapeake Bay and the Delaware River estuary. Bivalves were collected by hand and small dredge and by screening substrata. Representative specimens were placed in the collections held at the Coastal Ecology Research Laboratory at the University of Maryland Eastern Shore (UMES). Collections made at sites reported in the literature or which were represented in the ANSP and DMNH were com- pared to determine changes in species diversity. All localities reported from those sources were surveyed when locality data were sufficient to relocate the original site. RESULTS Ten species of unionids were found in the waters of the Delmarva Peninsula. Historic records were found for three additional species. Station numbers from the present study are given in parentheses for surveys conducted at localities listed in published works and the museum records. Station localities are provided in the appendix. Elliptio fisheriana (Lea, 1838) Published Records. Chester River System: Head of the Chester River (Kent Co.) MD [Type Locality (Lea, 1838)]; Stations 5 (Ortmann, 1919), 8 (Ortmann, 1919), 15 (Rhoads, 1904), 23 (Davis, 1984), 28 (Rhoads, 1904). Museum Materials Examined. Stations 5, 13, 17, 18, 23, 24, 28. Also examined were specimens from an unnamed body of water, Chestertown, Kent Co., MD (DMNH 174253) and from Sussex Co., DE (ANSP 345052). Records from Present Study. Stations 11, 17, 29, 30, 31, 34, 35) = 30. There has been confusion as to the systematic posi- tion of Elliptio fisheriana. Ortmann (1919) noted that a close morphological similarity exists between E. fisheriana and E. cupreus and believed that FE. fisheriana could be a lowland race of E. cupreus. However, he also noted that he could not detect intergrades between the two species. He further noted that specimens collected in White Clay Creek, Chester Co., Pennsylvania (Hartman and Michener, 1874), could in fact have been collected in Delaware. Johnson (1970) reported E. fisheriana to be a synonym of E. lanceolata (Lea, 1828) which he considered to be a highly variable species (25 synonyms listed for E. lanceolata); however, Davis (1984) found E. fisheriana to be genetically distinct from E. lanceolata. No published reports exist regarding the glochidia, soft tissue anatomy, or breeding season of EF. fisheriana. Elliptio complanata (Lightfoot, 1786) Published Records. Stations 4 (Davis, 1984), 24 (Davis, 1984). An additional record was published for the Chester River, Queen Annes Co., MD (Davis, 1984). Museum Materials Examined. Stations 3, 4, 12 - 14, 17, 18, 21, 22 - 24, 28, 43, 46. Additional specimens: Elk River, Sandy Cove, Cecil Co., MD (DMNH 75226); an unnamed ditch, Petersburg, Kent Co., DE, No Date, (DMNH 131124); Christmas Creek, Newark, New Castle Co., DE (ANSP 366090); Small Creek, Kent Co., DE (ANSP 358279). Records from Present Study. Stations 1, 6, 9, 10, 12, 13, 17 - 19, 21, 22, 25 - 27, 30 - 32, 34, 37, 48, 49, 53, 54, 59, 61, 63. Elliptio complanata can demonstrate local morpho- logical variability in both shell shape and coloration. Further, the species has the widest zoogeographic distribution of those unionid bivalves indigenous to the Atlantic Slope, ranging from the Apalachicola River drainage of Florida to the St. Lawrence drainage and Interior Basin of Canada (Burch, 1973; Clarke, 1973). Perhaps it is because of this wide variability and zoogeographic distribution that numerous synoyms ex- ist for the species [124 in Johnson (1970)]. Soft tissue anatomy has been described by Ortmann (1911) and Reardon (1929). Elliptio complanata is reported to breed from April through July or August. The host species is the anadromous yellow perch, Perca flavescens (Mitchill) (Lefevre and Cur- tis, 1912; Matteson, 1948). E. complanata is the most com- mon species of freshwater bivalve encountered in the Atlan- tic drainage (Clarke and Berg, 1959) and is typically found in lakes, ponds, rivers, and small streams on all types of substrata except very soft mud. Clarke and Berg (1959) reported that often this is the only species found in a par- ticular locality and, if other species of unionids are present, it is the most abundant. Our survey confirms this finding. Elliptio dilatata (Rafinesque, 1820) Published Records. Chesapeake Bay drainage. Nanticoke River System: Concord Pond, Sussex Co., DE (Davis, 1984) Station 23). Elliptio dilatata is an Interior Basin species that closely resembles FE. complanata and is common throughout its range (Clarke and Berg, 1959) in large and small rivers in either rapid or slow-flowing reaches, as well as in lakes on rocky, gravel, sand, or mud substrata (Clarke and Berg, 1959). Soft tissue anatomy was described by Ortmann (1911). Reproduc- tion occurs in the spring although Ortmann (1919) reported gravid females were found in Pennsylvania from May through August. Glochidia are retained in the marsupium until August (Ortmann, 1919; Clarke and Berg, 1959). The host fish in unknown. No specimens of Elliptio dilatata were found during the present study but a single published record is reported. Since the voucher specimen for this record was not found in the collections of the Academy of Natural Sciences of Philadelphia, it is believed that the specimens were late iden- 30 AMER. MALAC. BULL. 9(1) (1991) tified as &. complanata. An examination of the collection locality failed to reveal the presence of this species. Therefore, no verified specimen of EF. dilatata has been taken from the Delmarva Peninsula. Alasmidonta undulata (Say, 1817) Published Record. Chesapeake Bay Drainage. (Station 15) (Clarke, 1981). Museum Materials Examined. Stations 15 and 17. Record from Present Study: Station 17. Clarke (1981) published a single locality for Alasmi- donta undulata on the Delmarva Peninsula, and two lots from the Choptank River system are at ANSP. We located a single specimen during our survey. Ortmann (1919) noted that Alasmidonta undulata is gravid from July to the following June and Clarke (1981) reported collecting gravid females between August and Oc- tober. The host fish is unknown. A. undulata is reported to occur in moderately flowing streams, from rivers to creeks, and is most abundant on gravel and sand substrata, being ab- sent from mud (Ortmann, 1919; Clarke and Berg, 1959; Clarke, 1981). The species is also found in lakes on sand and gravel substrata (but growth can be stunted in these habitats) and reaches its maximum size in stream outlets located down- stream of lakes (Clarke and Berg, 1959). The species is reported to be commonly associated with Elliptio complanata and secondarily with Strophitus undulatus (Clarke and Berg, 1959; Clarke, 1981). Ortmann (1911) described the soft tissue anatomy. Alasmidonta varicosa (Lamarck, 1819) Published Record. Station 47 (Clarke, 1981). Museum Materials Examined. Station 47. Clarke (1981) published a single locality for Alasmi- donta varicosa on the Delmarva Peninsula. However, we were unable to locate the species and a survey of the Clarke’s local- ity failed to reveal the presence of the species. Alasmidonta heterodon (Lea, 1830) Museum Materials Examined. Station 17. One lot (ANSP 174899) collected by G. A. Coventry, August 1939, ‘*Delaware.”’ Although Alasmidonta heterodon has been reported from streams to areas adjacent to the Delmarva Peninsula (Clarke and Berg, 1959; Johnson, 1970; Clarke, 1981), it has never been described from the waters of the peninsula-proper. The species was first collected on the peninsula in August 1939 by G. A. Coventry without specific collection data other than ‘*Delaware.’’ One population was found during this study in (Station 17) Norwich Creek, a tributary of Tuckahoe Creek, Choptank River system, near Hillsboro, Maryland. The population is located just within the Talbot Co. line and ap- pears to be locally abundant. A second population, which we were unable to locate, was reported from Long Marsh Ditch (pers. comm., Maryland Nature Conservancy, 1991). Because of its rare and endangered status, no collections were made. Alasmidonta heterodon has been described as in- conspicuous with a disjunctive distribution along the Atlan- tic coast (Clarke and Berg, 1959; Clarke, 1981). The species is bradytictic and Clarke and Berg (1959) noted that its breeding season is not well known with gravid females be- ing reported in February and April. Clarke (1981) noted gravid females have been found in June and, in the Tar River, North Carolina, in late August. The fish host species is not known. Details of soft tissue anatomy are presented in Clarke (1981). The reported habitats include medium-sized rivers or rather slow-moving rivers of varying size on substrata of gravel, sand, or muddy sand, and sometimes among submerged aquatic vegetation (Clarke and Berg, 1959; Johnson, 1970; Clarke, 1981). The population at Norwich Creek (Station 17) is living in slow-moving water over a sandy substratum. Clarke and Berg (1959) also noted that A. heterodon was associated commonly with Elliptio complanata and Strophitus undulatus in central New York. Although we found neither of these species in direct association with A. heterodon at Norwich Creek, contemporaneous historical collections from this locality indicated the presence of E. complanata, S. un- dulatus, as well as Anodonta cataracta, E. fisheriana, and A. undulata. Anodonta cataracta Say, 1817 Published Record. Seaford, Sussex Co., DE (Rhoads, 1904). Museum Materials Examined. Stations 8, 11, 15, 17, 24, 55, 61. Other materials examined: Leonards Brick Pond, Chop- tank River System, Cambridge, Dorchester Co., MD. (col- lections made before 1930, a note with the lot states that the locality now has dwellings) (DMNH 87400): Tributary of Brandywine Creek, Christina River System, Greenville, New Castle Co., DE (ANSP 182963). One lot from Cambridge, Dorchester Co., MD. (ANSP 132477) is from a fish pond, with no outlets to streams. Another lot (ANSP 355544) gives a locality of ‘‘Wye Mills, Norwich Creek, Talbot Co., MD.’ Wye Mills is located in Queen Anne’s Co. and Norwich Creek is in Talbot Co. ap- proximately 8 km east of Wye Mills. Records from Present Study. Stations 1 - 3, 6, 11, 20, 24, 29, 30, 32, 34, 35, 38, 40 - 42, 51, 52, 55 - 63, 65. This was the second most commonly found freshwater bivalve on the Delmarva Peninsula. Clarke and Berg (1959) report the species to be common in lakes and ponds and streams varying in size from large rivers to small creeks, it is most abundant on substrata of sand or mud. They further noted that it was the only species found in soft and substrata of ponds and backwater areas. On Delmarva, the species has COUNTS ET AL.: DELMARVA NAIADES 3] been collected historically from small streams but our col- lections were entirely from the small mill ponds, usually from sand-silt substrata. Anodonta cataracta is reported to breed from the mid- dle of July to the following April or May (Clarke and Berg, 1959). The host fish species is unknown. Details of soft tissue anatomy are presented in Reardon (1929). Anodonta implicata Say, 1829 Museum Materials Examined. Stations 4, 5, Il, 13. One lot (ANSP 355543) gives a locality of ‘“Wye Mills, Norwich Creek, Talbot Co., MD.’ Wye Mills is located in Queen Anne’s Co. and Norwich Creek is in Talbot Co. approximately 8 km east of Wye Mills. Records from Present Study. Stations 12, 13, 16, 31, 49, 50, 60. Anodonta implicata is found most commonly in sand or gravel substrata and, very rarely, in mud (Clarke and Berg, 1959; Johnson, 1970). Johnson (1970) notes that the species seems to prefer stream habitat although it can be found in coastal ponds with an unobstructed outlet to the ocean. A. implicata was found during our study in ponds without such direct access to the ocean in the Delaware River and Bay drainage (Stations 49, 50, 60). Ortmann (1919) reported Anodonta implicata to be bradytictic (winter breeders) with larvae present in the mar- supium between July and September. Johnson (1970) reported gravid females in Massachusetts in early May and June. Larvae are released the followiong spring and the host species is the anadromous alewife, Alosa pseudoharengus (Wilson, 1811) (Clarke and Berg, 1959). No report of soft tissue anatomy is known. The glochidia were described by Johnson (1946). Unionids most commonly associated with A. implicata are A. cataracta, Lampsilis radiata radiata, L. ochracea, and Elliptio complanata (Clarke and Berg, 1959). There are no published records for A. implicata on the Delmarva Peninsula. Lampsilis radiata radiata (Gmelin, 1791) Published Record. Sussex Co., DE (Davis and Fuller, 1981). Museum Materials Examined. Stations 3, 5, ll - 13, 24. Ad- ditional materials were examined from Grays Branch, Chop- tank River System, near Denton, Caroline Co., MD (ANSP 106007). Collections from Present Study. Stations 1, 2, 12, 13, 33, 44. Lampsilis r. radiata is distributed widely over the Atlantic Slope (Johnson, 1970). Clarke and Berg (1959) noted that the breeding season appears to begin in August and end the following August. It is not known if this implies con- tinuous breeding or if a hiatus occurs between breeding years. The host fish is unknown but these authors suggested that many of the species serving as hosts for L. siliquoidea (Barnes, 1823) also serve as hosts for glochidia of L. r. radiata [bluegill, Pomoxis annularis Rafinesque; black crappie, P. nigromaculatus (LeSueur); largemouth bass, Micropterus salmoides (Lacepede); smallmouth bass, M. dolomieui dolomieui Lacepede); white bass, Roccus chrysops (Rafinesque); yellow perch, Perca flavescens (Mitchill); eastern sauger, Stizostedion canadense (Smith); and yellow pikeperch, S. vitreum (Mitchill)]. Lampsilis r. radiata is found typically on gravel or sand substrata and occasionally on mud. It occurs in lakes and rivers of all sizes but can be absent from smaller ponds and creeks (Clarke and Berg, 1959). Lampsilis cariosa (Say, 1817) Published Record. Seaford, Sussex Co., DE (Rhoades, 1904). Museum Materials Examined. Stations 24, 28. The anatomy of Lampsilis cariosa was described by Lea (1838). Ortmann (1911) found it to be similar to that of L. ventricosa (Barnes, 1823). The length of the breeding season is unknown but Clarke and Berg (1959) believed the species to be bradytictic. The host species is unknown. L. cariosa is found in riffles and shoals of large to medium-sized streams in fine to coarse gravel, usually in sand bars (Clarke and Berg, 1959; Johnson, 1970). There are only three records for Lampsilis cariosa from the Delmarva Peninsula, all from the Nanticoke River system. The species was not encountered during our study even though collections were made at the same stations as the historic records. This absence of L. cariosa could not be explained on the basis of misidentification. Leptodea ochracea (Say, 1817) Published Record. Seaford, Sussex Co., DE (Rhoads, 1904). Museum Materials Examined. Stations 5, 24, 28, 55. Collections from Present Study. Stations 1, 2, 60, 61. Ortmann (1919) reported Leptodea ochracea to be restricted to the tidal portions of the Delaware River. Clarke and Berg (1959) noted the species occurs in ponds, canals, and slow-flowing portions of rivers. The soft tissue anatomy was described by Reardon (1929). Very little is known con- cerning the glochidia, breeding season, or host species (Ort- mann, 1919; Johnson, 1947; Clarke and Berg, 1959) although Johnson (1970) reported finding gravid females in early May at Plymouth, Massachusetts, and thought the species was bradytictic. Johnson (1970) also believed that the glochidia probably parasitize migratory fish because L. ochracea is restricted generally to the lower reaches of streams having direct connections with the Atlantic Slope. — ieumia nasuta (Say, 1817) Published Records. Station 15 and Seaford, Nanticoke River System, Sussex Co., DE (Rhoads, 1904). Museum Materials Examined. Stations 15, 24, 28. Additional materials: Ditch, Choptank River System, near Petersburg, Kent Co., DE, 1939 (ANSP 175862); headwaters of the Chop- tank River, Medford, Mills, Sussex Co., DE, 1939 (ANSP 174904). Record from Present Study. Station 45. While there are historical collections that have placed Ligumia nasuta in various streams and ponds of Delmarva, it was encountered at only a single station (45) during our study. Clarke and Berg (1959) report Ligumia nasuta to breed from August to the following June. The host fish species is unknown. The preferred habitat appears to be ponds, lakes, and slack water portions of streams and canals on sand and mud substrata (Clarke and Berg, 1959). The species is usually associated with Elliptio complanata and Lampsilis radiata. Strophitus undulatus (Say, 1817) Published Record. Station 15 (Rhoads, 1904). Museum Materials Examined. Stations 13, 17, 47. Record from Present Study: Station 17. There is some debate as to the life-cycle of Strophitus undulatus. Lefevre and Curtis (1912) and Clarke and Berg (1959) report that the species can complete its development in parental marsupia but note that Baker (1928) reported glochidia completed development after attachment to largemouth bass, Micropterus salmoides (Lacepede, 1802), and the northern creek chub, Semotilus atromaculatus (Mitchill, 1818). Both of these species are found in waters draining into Chesapeake Bay (Lee, 1980; Lee and Platania, 1980; Hocutt et al., 1986). The habitats reported for Strophitus undulatus include small rivers and creeks on substrata of mud, sand, or fine gravel (Clarke and Berg, 1959; Johnson, 1970). S. undulatus is reported in association with Elliptio complanata and Alasmidonta undulata (Clarke and Berg, 1959). Con- temporaneous collections at ANSP confirm this association in Andover Creek, Maryland, as do our own collections in Norwich Creek, Talbot Co., Maryland. These collections further indicate associations with Lampsilis radiata, Anodonta cataracta, A. implicata, E. fisheriana, and Alasmidonta heterodon. DIVERSITY Unionid bivalves were found at 56 (18.2%) of the 307 stations examined (Table 1). Of the 13 species of unionids historically reported from waters of the Delmarva Peninsula, 10 are now present (Table 1). The highest number of species 32 AMER. MALAC. BULL. 9(1) (1991) present at a single locality was 5 (Station 17) followed by one station (Station 2) with 4 species, 8 stations with 3 species, 10 stations with 2 species, and 36 stations with a single species present (Table 1). When these data are compared with those for the 21 historically identifiable stations represented in museum collections or published in the literature only 2 sta- tions show no change in species composition. Additionally, 14 stations show a decline in the number of species present, 2 stations show an increase in species diversity, and 3 sta- tions have lost species but gained new ones. Therefore, the trend is towards a decrease in unionid species diversity. A review of historical collections revealed that Ellip- tio complanata was associated commonly with Anodonta cataracta and E. fisheriana and E. fisheriana with A. cataracta (Table 1). Collections from our survey indicate that E. complanata is commonly associated with E. fisheriana and E. fisheriana with A. cataracta but that E. complanata now is associated closely with Lampsilis radiata (Table 1). The three most commonly encountered unionids on the Delmarva Peninsula, both historically and at present, are Anodonta cataracta, Elliptio complanata and E. fisheriana (Table 2). Neither FE. dilatata, Lampsilis cariosa nor Alasmidonta varicosa were found during our survey. Table 2 summarizes species composition changes for all historical and present collections of unionids on the Delmarva Peninsula. DISCUSSION The unionid fauna of Delmarva is composed entirely of species associated with the greater Atlantic Slope fauna (Ortmann, 1919; Clarke and Berg, 1959; Johnson, 1970; Clarke, 1981). The origination of this fauna probably occurred in much the same manner as that of the freshwater fishes of the peninsula. In view of the need for a fish host for the bivalves to complete their development, the population of the peninsula by these two groups should have been simultaneous. Hocutt et al. (1986) noted that Chesapeake Bay is the drowned channel of the Susquehanna River and that the lower sea levels associated with interglacial periods facilitated the dispersal of several upland fish species to the peninsula. This process of rising and falling sea levels in the bay occurred many times (Flint, 1957) with the lowest level occurring during the Wisconsonian glaciation (Lougee, 1953). Thus, with the free movement of fishes and unionids during these periods, it is not surprising that Delmarva’s unionid fauna is like that of the rest of the Atlantic Slope. Once established, some isolation of unionid popula- tions in the major drainages of the peninsula could have occurred (Sepkowski and Rex, 1974). Further, extinction of local populations can occur and, given the disjunct popula- tions of such species as Alasmidonta heterodon, this has un- doubtedly occurred. Given the life cycle and physiological COUNTS ET AL.: DELMARVA NAIADES Table 1. Species present at all stations (A, Species historically present but now absent; B, Species historically present and now present; C, New record for the species). Station Number Taxon 1 2 3 4 5 6 7 8 9 10 ll 12 3 Elliptio fisheriana — E. complanata Cc E. dilatata = = — —_ = eas — = = == = = a Alasmidonta undulata — = = = _ = = = = A. varicosa — — = _ A. heterodon — Anodonta cataracta Cc Cc A. implicata — = = Lampsilis r. radiata C C A — Cc C > | > | | fo) QO | | fo) fo) | w ee] | | | | | | >>ow | | wo | L. cariosa Leptodea ochracea Ligumia nasuta = = = = _ = = BES Strophitus undulatus = = pes = = = A >| >> | | | | | | Station Number 14 15 16 7 18 19 20 21 22 23 24 25 26 Elliptio fisheriana — A — B A = = — — E. complanata A — _— B B @) wet E. dilatata = = _ a ae Alasmidonta undulata — A — B — — -- — A. varicosa a A. heterodon = = = B Anodonta cataracta —_ A — A — — Cc — a — A. implicata — _ (é = = as = = = = Lampsilis r. radiata — = = = = L. cariosa _ = = a Leptodea ochracea ae = Ligumia nasuta — A — = = = = _ = _ Strophitus undulatus — A — B _ = = = = = ie] wo >>> >> | | | | >>>> | w | | | Station Number vA 28 29 30 31 32 33 34 35 36 a WwW oo Ww \o _— — Cc Cc C Cc ee = Elliptio fisheriana a E. complanata E. dilatata = = = _ a = = Alasmidonta undulata = = = = = A. varicosa _— = = = = = A. heterodon = == Anodonta cataracta — — Cc Cc — Cc — Cc Cc — — Cc — A. implicata — — — Lampsilis r. radiata — os = L. cariosa _— — — Leptodea ochracea = Ligumia nasuta = Strophitus undulatus _ = = = = = fo) >> | aa aa fe’ | ene | io) | | | | | | | | > | | | | | | | | | | Station Number 40 41 42 43 44 45 46 47 48 49 50 51 52 Elliptio fisheriana = E. complanata — — — Cc — — A — Cc Cc = — a E. dilatata = = = = = Alasmidonta undulata _ aes = = = _ = _ A. varicosa — = A = = = _ = A. heterodon — — Anodonta cataracta Cc Cc Cc — _ — Cc — — = _ (ei Cc A. implicata Cc Cc = es Lampsilis r. radiata — ~ — =_ C = = = — L. cariosa Leptodea ochracea = = = = a = = = = = = = = Ligumia nasuta A = = _ = = Strophitus undulatus = — a a = (continued) Table 1. (Continued) Taxon AMER. MALAC. BULL. 9(1) (1991) Station Number Elliptio fisheriana E. complanata Cc C = = E. dilatata Alasmidonta undulata A. varicosa — — —_— — A. heterodon Anodonta cataracta — — B Cc A. implicata Lampsilis r. radiata = -_ = — L. cariosa 53 54 55 56 S57 S58 59 60 61 62 63 64 65 =) =e Ca 6 ee Co UC ae CLR. COC £ AL =. a Cc = = = = a Leptodea ochracea — = A = Ligumia nasuta Strophitus undulatus requirements of unionids, opportunities for expansion of zoogeographic ranges are limited in these habitats and thus, the observed pattern of distribution of unionids in coastal rivers is not explained easily. Sepkowski and Rex (1974) offered three hypotheses to explain the distribution of unionid fauna along coastal rivers. The first proposes that gravid mussels could attach to the feet of aquatic birds and are thence transported into neighboring systems. They noted that van der Schalie (1945) rejected this particular statement of the hypothesis but noted that predatory birds can carry the tissues of glochidial-parasitized fish from one stream to another. This form of the hypothesis seems more reasonable given the size and weight of unionid bivalves although the size, depth, and degree of canopy cover of most Delmarva Peninsula streams do not lend themselves to large- scale use by predatory birds. The second hypothesis hinges on parasitism of secon- dary and peripheral fishes. Myers (1938, 1951) classified fishes on the basis of their salinity tolerance. Primary fishes have little or no tolerance to sea water; secondary fishes are restricted usually to freshwater but have a salinity tolerance sufficient to cross narrow bands of salt water. Thus, secondary fish hosts could move freely into the bay from one drainage and enter another. Salinity tolerance studies by Musick (1972) and Lee (1976) suggested that primary fishes can tolerate greater salinity concentrations than suggested by Myers. Because of this others (e.g. Hocutt et al., 1986) have sug- gested that the salinity tolerance classification scheme is in- valid. The presumed primary fish host of Lampsilis r. radiata and Strophitus undulatus, Micropterus salmoides, tolerated a salinity of 12.9 ppt and another primary fish host of L. r. radiata and Elliptio complanata, Perca flavescens, tolerated a salinity of 13 ppt. Thus, these fishes could move more freely among the drainage systems of the peninsula than was suspected previously. Unionid glochidia, however, are in- tolerant of saline conditions (Cvancara, 1970) and their move- ment on a fish host into saline conditions seems unlikely unless physiological isolation occurs within the fish host’s tissues. It seems more likely that the usual mode of move- ment of unionid species on fish hosts occurred during times of high freshwater input into the system (e.g. rains, etc.). These events would dilute the saltwater barriers between drainages and this seems to agree with the third hypothesis of Sepkowski and Rex (1974) that stream capture and flooding could play a role in dispersal. This dispersal mechanism could also explain the absence of unionids on the Virginia portion of the Delmarva Peninsula. These streams are short and shallow and most are tidal and brackish. Furthermore, the waters of the Chesapeake Bay on the west side of the peninsula are more saline [22 - 28 ppt (Bashore and Kelly, 1987)] than those observed farther north in Maryland. The extensive use of mill dams that act to sharply demarcate fresh and salt waters are essentially ab- sent from the Virginia portion of the peninsula. Thus, the movement of unionids from one drainage to another in the southern end of the peninsula, even while attached to a host fish species, requires traversing waters of even higher salinities than those farther north. Even major weather events that result in significant amounts of rainfall could not dilute appreciably waters of these salinities to permit the passage of species from one drainage to another. Another factor contributing to the absence of unionid species in this region is the frequency with which the streams dry up during periods of drought. Should a unionid species become established in a southern penin- sula stream, it could find itself without water during the sum- mer months. The absence of species usually associated with ponds (Elliptio complanata, Anodonta cataracta, Lampsilis r. radiata, Leptodea ochracea, Ligumia nasuta) is puzzling because many of the existing ponds have been stocked with fish. The collections made during our survey indicate a general trend toward decline in species diversity. While there has not been a historical trend towards development of heavy industry on the Delmarva Peninsula during the past 100 years, there has been an increase in population and an intensifica- tion of agricultural production. This has led to the channel- COUNTS ET AL.: DELMARVA NAIADES Sh ization of many streams for both flood control and increased drainage of farm fields. This process could have contributed to habitat loss for some species because several historical localities are now channelized. Furthermore, the increased application of agricultural chemicals could have played a role in diversity decline by acting either directly upon the mussels or upon their fish hosts species. Because many of the fish hosts for the species of unionids on Delmarva are as yet unknown, the extent of this factor is unknown. The zebra mussel, Dreissena polymorpha, has yet to be found along the Delmarva Peninsula. The point at which direct introduction from abroad is likely to occur is the Chesapeake Bay and Delaware Canal. The Maryland Depart- ment of Natural Resources is currently monitoring these waters for evidence of this exotic species. ACKNOWLEDGMENTS The authors wish to thank Gene Handwerker and Nicholas Counts for their help during sampling. Thanks are also due David O’ Neill, University of Maryland Eastern Shore, for his help in collecting specimens. We would like to also thank Dr. Rudiger Bieler, Delaware Museum of Natural History and Dr. Arthur E. Bogan, Academy of Natural Sciences of Philadelphia, and their respective staffs for assisting us in locating historical materials described in this paper. The paper benefited from the comments of two anonymous reveiwers whose assistance is gratefully acknowledged. LITERATURE CITED Baker, F. C. 1928. The fresh water Mollusca of Wisconsin. Part 2. Pelecypoda. Wisconsin Geological and Natural History Survey, Bulletin 70(2). 492 pp. Bashore, T. L. and K. L. Kelly. 1987. Chesapeake Bay Marine Environmental Assessment March 1986 - August 1986. Chesapeake Research Con- sortium (Gloucester Point, Virginia) Publication 123. 51 pp. Burch, J. B. 1973. Freshwater Unionacean Clams (Mollusca: Pelecypoda). Biota of Freshwater Ecosystems Identification Manual No. 11, United States Environmental Protection Agency (Washington, D.C.). 176 pp. Carpenter, D. H. 1983. Characteristics of streamflow in Maryland. Depart- ment of Natural Resources, Maryland Geological Survey, Report of Investigations No. 35. 237 pp. Clarke, A. H. 1973. The freshwater molluscs of the Canadian Interior Basin. Malacologia 13(1/2):1-509. Clarke, A. H. 1981. The tribe Alasmidontini (Unionidae: Anodontinae), Part I: Pegias, Alasmidonta, and Arcidens. Smithsonian Contributions to Zoology Number 326. 101 pp. Clarke, A. H. and C. O. Berg. 1959. The freshwater mussels of Central New York, with an illustrated key to the species of northeastern North America. Cornell University Agricultural Experiment Station Memoir 367: 1-79. Counts, C. L., II, T. S. Handwerker and E. B. May. 1991. The potential for infestation of Chesapeake Bay by the zebra mussel, Dreissena polymorpha (Bivalvia: Dreissenacea). Proceedings of the 1990 Chesapeake Research Conference. In press. Cvancara, A. M. 1970. Mussels (Unionidae) of the Red River Valley in North Dakota and Minnesota, U.S.A. Malacologia 10:57-92. Davis, G. M. 1984. Genetic relationships among North American Unionidae (Bivalvia) sibling species, convergence and cladistic relationships. Malacologia 25(2):629-648. Davis, G. M. and S. L. H. Fuller. 1981. Genetic relationships among Re- cent Unionacea (Bivalvia) of North America. Malacologia 20(2):217-253. Davis, G. M., W. H. Heard, S. L. H. Fuller and C. Hesterman. 1981. Molluscan genetics and speciation in Elliptio and its relationship to other taxa of North American Unionidae (Bivalvia). Biological Journal of the Linnean Society 15:131-150. Flint, R. F. 1957. Glacial and Pleistocene Geology. John Wiley and Sons, New York. 553 pp. Handwerker, T. S., C. L. Counts, III and R. V. Jesien. 1991. Distribution of Corbicula fluminea (Bivalvia: Corbiculidae) in the Chesapeake Bay drainage of the Delmarva Peninsula and its use in aquaculture. Proceedings of the 1990 Chesapeake Research Conference. In press. Hartman, W. D. and E. Michener. 1874. Conchologia Cestrica. Philadelphia. 114 pp. Hocutt, C. H., R. E. Jenkins and J. R. Stauffer, Jr. 1986. Zoogeography of the fishes of the Central Appalachians and Central Atlantic Coastal Plain. In: The Zoogeography of North American Freshwater Fishes, C. H. Hocutt and E. O. Wiley, eds. pp. 161-211. John Wiley and Sons, New York. Johnson, R. I. 1946. Anodonta implicata Say. Occasional Papers on Mollusks, Museum of Comparative Zoology 1:109-116. Johnson, R. I. 1947. Lampsilis cariosa Say and Lampsilis ochracea Say. Occasional Papers on Mollusks, Museum of Comparative Zoology 1:145-156. Johnson, R. I. 1970. The systematics and zoogeography of the Unionidae (Mollusca: Bivalvia) of the southern Atlantic Slope region. Bulletin of the Museum of Comparative Zoology 140(6):263-450. Lea, I. 1838. Descriptions of new freshwater and land shells. Transactions of the American Philosophical Society, New Series, 6:1-154. Lee, D. S. 1976. Aquatic zoogeography of Maryland. Atlantic Naturalist 31(4):147-158. Lee, D. S. 1980. Micropterus salmoides (Lacepede). In: Atlas of North American Freshwater Fishes, D. S. Lee, C. R. Gilbert, C. H. Hocutt, R. E. Jenkins, D. E. McAllister and J. R. Stauffer, Jr., eds. p. 608. North Carolina State Museum of Natural History, Raleigh. Lee, D. S. and S. P. Platania. 1980. Semotilus atromaculatus (Mitchill). In: Atlas of North American Freshwater Fishes, D. S. Lee, C. R. Gilbert, C. H. Hocutt, R. E. Jenkins, D. E. McAllister and J. R. Stauffer, Jr., eds. p. 361. North Carolina State Museum of Natural History, Raleigh. Lefevre, G. and W. C. Curtis. 1912. Studies on the reproduction and arti- ficial propagation of freshwater mussels. United States Bureau of Fisheries, Bulletin 30:105-201. Lougee, R. J. 1953. A chronology of postglacial time in eastern North America. Science Monthly 76(5):259-276. Matteson, M. P. 1948. Life history of Elliptio complanatus (Dillwyn, 1817). American Midland Naturalist 40:690-723. Musick, J. A. 1972. Fishes of Chesapeake Bay and the adjacent Coastal Plain. Virginia Institute of Marine Science, Special Scientific Report 65:175-212. Myers, C. S. 1938. Fresh-water fishes and West Indian zoogeography. Smith- sonian Report 1937:339-364. Myers, C. S. 1951. Freshwater fishes and East Indian zoogeography. Stan- ford Ichthyological Bulletin 4:11-24. Ortmann, A. E. 1911. A monograph of the najades of Pennsylvania. Memoirs of the Carnegie Museum 4:279-347. Ortmann, A. E. 1919. A monograph of the naiades of Pennsylvania. Part III. Systematic account of the genera and species. Memoirs of the Carnegie Museum 8(1):1-384. Reardon, L. 1929. A contribution to the knowledge of the anatomy of the 36 AMER. MALAC resh-water mussels of the District of Columbia. Proceedings of the United States National Museum 75(11):1-12. Rhoads, S. N. 1904. A glimpse at the shell fauna of Delaware. Nautilus 43:63-67. Sepkowski, J. J., Jr. and M. A. Rex. 1974. Distribution of freshwater mussels: coastal rivers as biogeographic islands. Systematic Zoology 23(2):165-188. Stephenson, L. W., C. W. Cooke and W. C. Mansfield. 1933. Chesapeake . BULL. 9(1) (1991) Bay Region. International Geological Congress, XVI Session, United States, 1933, Guidebook 5: Excursion A-5. 49 pp. van der Schalie, H. 1945. The value of mussel distribution in tracing stream confluence. Papers of the Michigan Academy of Arts and Letters 20:355-373. Date of manuscript acceptance: 17 May 1991 APPENDIX. Collection station numbers and localities examined for the presence of unionid bivalves on the Delmarva Peninsula. The geographic locations of stations are identified in figure 1 by station number. Stations are listed from north to south first in the Chesapeake Bay drainage and secondly in the Delaware Bay Atlantic Ocean drainage. Previously reported stations are referenced in parentheses. (DMNH, Delaware Museum of Natural History; ANSP, Academy of Natural Sciences of Philadelphia; UMES, University of Maryland Eastern Shore). Only those stations where unionid bivalves were found are listed. A complete list of all station localities, including those where no naiades were found, is on file at ANSP. Chesapeake Bay Drainage SUSQUEHANNA RIVER 1. Susquehanna River, mud flat at Veterans Hospital, Perry Point, Perry- ville, Cecil Co., MD (UMES 2258, 2259, 2260, 2261) BOHEMIA RIVER SYSTEM . Little Bohemia Creek, Bohemia Mills, Bunker Hill Road, Cecil Co., MD (UMES 2397, 2398, 2399) N SASSAFRAS RIVER SYSTEM 3. Sassafras River, MDSR 299 bridge, Cecil-Kent Co. line, MD (ANSP 358302, 358303, 358304, 358305; UMES 2408) 4. Sassafras River, under US 301 bridge, Cecil-Kent Co. line, MD (Davis, 1984; ANSP 346882, 346889, 347868) CHESTER RIVER SYSTEM 5. Ratcliff Creek, MDRT 661, Chestertown, Kent Co., MD (Ortmann, 1919; ANSP 65093, 65095, 65096; DMNH 78556) 6. Island Creek, MDSR 213 bridge, north of Centreville Queen Anne’s Co., MD (UMES 2402, 2403) 7. Granny Finley Branch, MDSR 213 bridge, Queen Anne’s Co., MD (UMES 2409) 8. Southeast Creek, MDSR 213 crossing, Queen Anne’s Co., MD (Ort- mann, 1919; ANSP 66311) 9. Red Lion Branch, Dudley Corse Road, Queen Anne’s Co., MD (UMES 2395) 10. Red Lion Branch, Red Lion Branch Road, Queen Anne’s Co., MD ll. Unicorn Branch, Unicorn Community Lake, Queen Anne’s Co., MD (ANSP 358299, 358300, 358301; UMES 2378, 2379) 12. Chester River, MDSR 313 bridge, Millington, Kent-Queen Anne’s Cos.., MD (ANSP A9484, 358297, 358298, 358308; UMES 2307, 2308, 2309) 13. Andover Branch, Peacock Corner Road crossing, east of Millington, Kent Co., MD (ANSP A9482, 355555, 355801, 358291, 358292, 358293; UMES 2480, 2481, 2482) 14. Chester River, US 301 bridge, Kent-Queen Anne’s Co. line (ANSP 346890) CHOPTANK RIVER SYSTEM 15. Choptank River, ‘‘Choptank Mills,” RT 207 at Mud Mill Pond, Kent Co., DE (Rhoads, 1904; Clarke, 1981; ANSP 85224, 85226, 85264) 16. Tuckahoe Lake, Tuckahoe Creek, Tuckahoe State Park, Crouse Mill Road bridge, Caroline-Queen Anne Cos., MD (UMES 2375) 17. Norwich Creek, RT 404 bridge, Queen Anne’s Co., MD (ANSP A9487, A9488, Al0261, Al10262, 355556, 355802, 358285, 358286, 358287, 358288, 358289, 358290; DMNH 41138; UMES 2471, 2472, 2473, 2474) 18. Mason Branch, Tuckahoe Creek, MDSR 304 bridge, Caroline-Queen Anne’s Cos., MD (ANSP 358294, 358295, 358296; UMES 2265) 19. Watts Creek, Legion Road bridge, Caroline Co., MD (UMES 2380) 20. Williston Lake at spillway, Denton Road, Caroline Co., MD (UMES 2400) 21. Hog Creek, Hog Creek Road bridge, Caroline Co., MD (ANSP 358306; UMES 2483) 22. Beaverdam Branch, 1.5 km west of Matthews at MDSR 328 bridge, Talbot Co., MD (ANSP 358309; UMES 2484) NANTICOKE RIVER SYSTEM 23. Concord Pond, just north of DESR 20, CR 516, 1.6 km east of Seaford, Sussex Co., DE (Davis, 1984; ANSP 345054, 352551, 358277) 24. ‘‘Deep Creek,’’ just below Concord Pond spillway, RT 516, east of Seaford, Sussex Co., DE (Davis, 1984; ANSP 346863, 346864, 346865, 346866, 346867, 347863, 347864, 347865, 347866, 347867, 349333, 349337, 349338, 358272, 358273, 358274, 358276, 358278; UMES 2476) 25. Gravelley Branch, Collins Pond, just east of Coverdale Crossroads, Sussex Co., DE (UMES 2243) 26. Marshyhope Creek, DESR 16 bridge, just east of Hickman, Kent Co., DE (UMES 2233) 27. Nanticoke Branch, DESR 18 bridge, just east of junction with CR 533, Sussex Co., DE (UMES 2235) 28. Nanticoke River, US 13 bridge, Seaford, Sussex Co., DE (Rhoads, 1904; ANSP 84837, 88219, 88370, 301003, 301004, 345058; DMNH 75214) 29. Trap Pond, Trap Pond State Park, Sussex Co., DE (UMES 2266, 2267) 30. Chipman Pond, CR 465 bridge at dam, Sussex Co., DE (UMES 2262, 2263, 2264) 31. Meadow Branch, just downstream of Horsey’s Pond, DESR 24 bridge, Laurel, Sussex Co., DE (UMES 2276, 2277, 2279) 32. Quantico Creek, Quantico Creek Road, east of Quantico, Wicomico Co., MD (UMES 2469, 2470) 33. Rewastico Pond, Rewastico Creek at dam, Athol Road, Wicomico Co., MD (UMES 2252) 34. Barren Pond, Barren Creek at dam, Wicomico Co., MD (UMES 2249, 2250, 2451) 35. Tyndall Branch, Fleetwood Pond, CR 484 at dam, Sussex Co., DE (UMES 2298, 2297) 36. Record Pond, at dam, Laurel, Sussex Co., DE (UMES 2293) 37. Gales Creek, Galestown Pond, Galestown-Reliance Road at spillway, Galestown, Dorchester Co., MD (UMES 2287, 2289) COUNTS ET AL.: DELMARVA NAIADES 37 38. Butler Mill Branch, Craigs Pond, public boat access at spillway, CR 542A, Sussex Co., DE (UMES 2290, 2291) 39. Nanticoke River, Williams Pond, Tharp Road bridge, Seaford, Sussex Co., DE (UMES 2300) 40. Bucks Branch, Hearn Pond, CR 544A at spillway, Sussex Co., DE (UMES 2299) 41. Smithville Lake, Opossum Hill Road at dam, Caroline Co., MD (UMES 2377) WICOMICO RIVER SYSTEM 42. Wicomico River, Riverside Boat Ramp Park, Salisbury, Wicomico Co. , MD (UMES 2241) 43. Leonard Pond Run, Leonard Pond, Leonard Pond Park, US 13 bridge, Salisbury, Wicomico Co., MD (DMNH 48638) POCOMOKE RIVER SYSTEM 44. Nassawango Creek, Red House Road bridge, Worcester Co., MD (UMES 2374) 45. Unnamed tributary (A) of Nassawango Creek, Nassawango Road just southeast of Pennewell Road, Worcester Co., MD (UMES 2219) Delaware River and Bay, Atlantic Ocean Drainage CHRISTINA RIVER SYSTEM 46. Brandywine Creek, Brandywine Park at Dam near Brandywine Zoo, Wilmington, New Castle Co., DE (DMNH 44090) 47. Red Clay Creek at Yorklyn, New Castle Co., DE (Clarke, 1981; ANSP 85227, 85228) 48. Becks Pond, SR 48 at dam and swimming beaches, New Castle Co., DE (UMES 2269, 2270) APPOQUINIMINIK RIVER 49. Shallcross Lake at Greylag, CR 428 bridge at dam, New Castle Co., DE (UMES 2282, 2283) 50. Silver Lake, CR 442 bridge at dam, east of Middletown, New Castle Co., DE (UMES 2281) 51. Wiggins Mill Pond, Road 446, below spillway, New Castle Co., DE (UMES 2406) 52. 33: 60. 6l. 65. BLACKBIRD CREEK Pond, Road 463A, below spillway, New Castle Co., DE SMYRNA RIVER SYSTEM Paw Paw Branch, Road 40 bridge, east of Thomas Corner, New Castle Co., DE (UMES 2396) LEIPSIC RIVER SYSTEM . Massey’s Millpond, Road 42 bridge, Kent Co., DE (UMES 2407) ST. JONES RIVER SYSTEM . Silver Lake, Silver Lake Recreation Area, Dover, Kent Co., DE (ANSP 45991, 45992; UMES 2376) Tidbury Creek, Derby Pond, US 13A, south of Wyoming, Kent Co., DE MURDERKILL RIVER SYSTEM McGinnis Pond, Road 378, southwest of Lexington Mill, Kent Co., DE (UMES 2393) . McColley Pond, DESR 15 bridge at dam and spillway, Mordington, Kent Co., DE (UMES 2392) MISPILLION RIVER SYSTEM Blairs Pond, Road 443, west of Milford, Kent/Sussex Cos., DE (UMES 2382, 2385) Griffith Lake, Road 443 at spillway, west of Milford, Kent/Sussex Cos., DE (UMES 2384, 2386, 2388) Haven Lake, US 113 bridge at spillway, Milford, Kent/Sussex Cos., DE (ANSP 103011; UMES 2389, 2390, 2391) . Silver Lake, DESR 36 bridge ai spillway, Milford, Kent/Essex Cos., DE (UMES 2401) CEDAR CREEK SYSTEM . Cedar Creek, Cubbage Pond, CR 214 bridge, Sussex Co., DE (UMES 2284, 2285) INDIAN RIVER SYSTEM . Millsboro Pond, DESR 24 at dam, Millsboro, Sussex Co., DE (ANSP 85835) Ingram Pond, Public Fishing Area at dam, CR 328, Sussex Co., DE (UMES 2298) The influence of oxygen availability on oxygen consumption in the freshwater clam Musculium partumeium (Say) (Bivalvia: Sphaeriidae) Daniel J. Hornbach Department of Biology, Macalester College, St. Paul, Minnesota 55105, U.S.A. Abstract. The O, consumption of the freshwater clam Musculium partumeium Say was determined as a function of clam size and O, availability. The rate of O, consumption by large individuals was influenced significantly by O, availability whereas the rates of smaller individuals were not altered significant- ly. One reason for the difference in response for large and small animals was the greater degree of availability in O, consumption among smaller individuals. A number of studies have examined the influence of low O, availability on metabolism in marine bivalves (e.g. Newell, 1979 and references therein). Much of this work is predicated upon the supposition that the tidal exposure of many bivalves in marine habitats regularly subjects them to periods of low O, availability and, thus, these forms could have responded evolutionarily to this selection pressure. Few studies, however, have examined the influence of reduced O, availability in freshwater bivalves (Burky, 1983; Burky et al., 1985; Hornbach, 1985). While there is not a cyclical tidal change in O; availability in freshwater systems, there are a number of factors which could affect O, ac- cessibility. Diurnal rhythms of O, concentration in littoral areas are known to occur as a result of shifts in the dominance of photosynthetic or respiratory activities in these highly pro- ductive areas. Seasonal shifts in O, availability in the pro- fundal zones of lakes associated with temperature stratifica- tion and subsequent turnover events are also well documented. Probably one of the most severe of freshwater habitats, in terms of changes in O, availability, are ephemeral ponds. In these habitats, O, levels can be high in the spring as the pond fills, but as water levels decline in the mid to late summer, O, availability often becomes restricted. Also, many ephemeral ponds have heavy loads of organic debris which can contribute to anoxic conditions in very shallow water dur- ing any season. One of the most common inhabitants of ephemeral ponds are bivalves of the family Sphaeriidae, the fingernail and pea clams. Within this family, the genus Musculium has the largest population of species found in temporary habitats (Burch, 1975). A moderately large body of information is available concerning life history adaptations in this genus for survival in ephemeral ponds (see e.g. Hornbach and Childers, 1986), yet little information is available concern- ing physiological adaptations for this harsh habitat (Burky, 1983). This study examines the influence of O, availability and body size on O, consumption in the clam, Musculium partumeium (Say). MATERIALS AND METHODS Clams utilized in this study were collected in May - July 1986 from a temporary pond located at the Kathryn Ord- way Natural History Area in Inver Grove Heights, Minnesota. In all cases, clams were removed from the pond when sedi- ment temperatures were near 17°C (range 16.5 - 18.0°C). Specimens were returned to the laboratory and maintained at 17°C in filtered pond water (Whatman ‘‘42’’, effective pore size 2.5 wm) for 6-12 hr before metabolic rates were deter- mined. The shell length (anterior-posterior dimension) was measured to the nearest 0.1 mm with a dissecting microscope fitted with a stage mounted micrometer. To examine the effect of size and O, availability on the metabolic rate of Musculium partumeium, the following ex- perimental design was utilized. The metabolic rates for three replicates of each of four size groups of clams were deter- mined at six different levels of O, availability. This yielded 72 values of O, consumption. Water with various levels of O, availability were pre- pared by mixing air-saturated, filtered pond water with nitrogen-saturated, filtered pond water in proportions of 1:0, 3:1, L:1, 1:3, 1:7 and 0:1, thus giving six levels of O, availa- bility. Because of the variability in the mixing process, changes in barometric pressure, etc., mixtures prepared on different days had varying Pos. The averages and ranges (in parentheses) of Po, values [measured by the azide-modified American Malacological Bulletin, Vol. 9(1) (1991):39-42 he) 40 AMER. MALAC. BULL. 9(1) (1991) Table 1. Summary of numbers, sizes and range of metabolic rates of Musculium partumeium utilized in this study. The coefficient of variation was calculated as the standard deviation as a percentage of the mean. Experiment Number of Mean Mean Range of Mean Number clams per shell ash-free metabolic rates coefficient of bottle length dry weight (ul O, AFDW-! variation (mg) hr-') (%) 1 10 1.2 0.05 0.68-4.78 35.7 2 5 2.6 0.31 0.44-2.26 33.3 3 5 4.0 0.79 0.81-1.46 8.3 4 3 6.2 2.59 0.62-1.41 18.5 Winkler method (American Public Health Association, 1980)] were 140 mm Hg (117-166), 116 mm Hg (98-129), 93 mm Hg (75-109), 71 mm Hg (57-85), 58 mm Hg (42-72) and 41 mm Hg (27-50) for the mixtures (air-saturated water: nitrogen- saturated water) 1:0, 3:1, 1:1, 1:3, 1:7 and 0:1, respec- tively. O, consumption of groups of 3-10 clams (depending on size) (Table 1) was measured after the specimens had been placed in 40 ml glass-stoppered bottles containing one of the pond water mixtures at 17°C. Control chambers (bottles with no clams) were analysed similarly as controls to correct for changes in O, concentration not directly attributable to clam metabolism. After 15-20 hr, the O, remaining in each chamber was measured with the azide-modified Winkler method. Duplicate 10.0 ml aliquots of water from each chamber were titrated with approximately 0.0025 N sodium thiosulfate (standardized with potassium bi-iodate) and a digital buret calibrated to 0.01 ml. Results were corrected for the volume of Winkler reagents used (0.5 ml each) in the fix- ation process. Additional details of this method can be found in Hornbach ef al. (1983). Values as mg O, were converted to pl O, or Po, (mmHg) at STP with equations presented in Pierce et al. (1973). Ash free dry weights (AFDW) were deter- mined for all groups of clams as the difference in weight after drying to constant weight at 100°C and ashing at 500°C. ee = RESULTS A Figure | shows the effect of O, availability on the rate - of O, consumption for each of the four size categories of E Musculium partumeium utilized in this study. Multiple regres- co) sion analysis for each size class, with log. (Qo) (ul O, mg = AFDW-' hr-') as the dependent variable and Po, (O, tension a mm Hg) and log. (AFDW) as the independent variables, oO showed that Po, significantly influenced Qo, for the two largest size classes (4.0 mm — F=17.0, 1,16 df, p=0.0009; 6.2 mm — F=4.77, 1,17 df, p=0.045). For the two smaller size classes, there was no significant influence of Po, on Qo, (1.2 mm — F=1.44, 1,17 df, p=0.248; 2.6 mm — F=1.35, 1,17, df, p=0.263). This lack of a statistically significant in- fluence was due partially to the large degree of variability in the O, consumption rates for these size groups of clams (Table 1). An overall regression analysis resulted in the follow- ing equation: log.(Qo,)=—0.21—0.169 [log.(AFDW)]+0.003 (Po,); r?=0.44 (F=26.37, 2,70 df, p=0.0001). The standar- dized partial regression coefficients for log.(AFDW) and Po, are —0.66 and 0.28, indicating that AFDW had a higher degree of influence on metabolic rate than did O, availability. Bayne (1971) derived an index for assessing the degree of influence of O, availability on O, consumption. If the con- sumption availability curve is hyperbolic, then Qo,=Po,/k, + k, (Po,). When Po,/Qo, is plotted against Po2, a straight line is produced with k, the intercept and k, the slope. A high value of the ratio k,/k, indicates a greater O,-dependence of O, consumption. Regressions of these values were performed for each size class of clam to obtain estimates of k, and k,. The results are shown in Table 2. DISCUSSION Several studies have shown that sphaeriid clams are capable of surviving various periods of anoxia (Juday, 1908; Shell Length Po, (mm Hg) Fig. 1. Relationship between oxygen availability and oxygen consumption for four sizes of Musculium partumeium. HORNBACH: OXYGEN CONSUMPTION OF MUSCULIUM 4l Table 2. Indices of oxygen dependence of metabolism for various sizes of Musculium partumeium. Indices are from the equation Po,/Qo, = k, +k, (Po,) and r? is the coefficient of determination. Mean Shell k, k, r? k,/k, Length (mm) 1:2 14.95 0.38 0.36 39.45 2.6 17.77 0.68 0.21 26.21 4.0 19.54 0.64 0.80 30.63 6.2 26.83 0.78 0.66 34.53 Eggleton, 1931; Thomas 1963, 1965; Gale, 1976; Holopainen and Jonasson, 1983; Way ef al., 1983; Holopainen, 1987). Nonetheless, few studies have examined the influence of O, availability on metabolism (Hornbach, 1985). Waite and Neufeld (1977) indicated that for the sphaeriid, Sphaerium simile (Say), there was little influence of O, availability on O, consumption over the range of 32-80 mm Hg Po. Out- side this range the influence of O, availability on O, con- sumption varied with temperature. Berg et al. (1962) found that O, consumption declined and reduced O, availability in Pisidium casertanum (Poli). Buchwalder (1983) found that Musculium partumeium and M. lacustre (Miiller) dis- played a dependence of O, consumption on O, availability and that this dependence varied seasonally. The greatest independence was found during periods of low O), availability and could help to explain how some sphaeriids can grow and reproduce under hypoxic conditions (Thomas, 1963, 1965; Way et al., 1980). In this study I found that there is an effect of O, availability on O, consumption and that the effect ap- pears to be size dependent (Fig. 1). There was considerable variability in the rates of consumption that were ascertained (Table 1). This high degree of variation could be due to behavioral differences in individuals (e.g. differences in locomotory or pumping activities) or differences in the reproductive condition of mature individuals. No attempt was made to observe individual variation in these parameters, however it was noted that in the vast majority of cases, all individuals were moving actively in the chambers, though it is unknown if this occurred continuously throughout the 15-20 hr experimental period. The fact that the variation in O, con- sumption is not significantly higher for brooding clams (6.2 mm size class - Table 1) suggests that reproductive condition does not alter significantly the variation noted by this technique. Newell (1979), based on studies of marine bivalves by Bayne (1971, 1973) and Taylor and Brand (1975), claimed that small animals, with high weight-specific metabolic rates tend to have higher k,/k, ratios than larger conspecific animals, indicating a decreased ability in smaller organisms to regulate O, consumption. Based on the k,/k, values in Table 2, Musculium partumeium = 2.6 mm appears to follow an op- posite trend, that is, larger animals tend to have larger values of k,/k, indicating a greater independence of O, consump- tion on O, availability. M. partumeium in the 4.0 and 6.2 mm size classes are reproductively active and often contain extra- marsupial larvae (Hornbach et al. , 1980; unpubl data). It is possible that the larvae displayed a regulatory pattern from the adults and thus mask the response of the adult. For ex- ample, Burky and Burky (1976) have shown that over 50% of the O, consumption of adults of the sphaeriid, Pisidium walkeri Sterki, could be attributed to the respiration of in- cubating embryos. Although the 1.2 mm clams had the highest k,/k, ratio (Table 2), the r? of the relationship used to attain this value was low (Table 2). In fact, for clams of mean shell length 1.2 or 2.6 mm, there was no statistically significant influence of O, availability on O, consumption. This could indicate that these size clams are regulators. However, Table | indicates that one reason for the lack of a statistical relationship be- tween O, availability and O, consumption in these size groups is due to the high individual variation in metabolic rates. The reasons for this variability are not known. However, in temporary ponds, many juvenile individuals do not grow dur- ing the summer. Apparently this lack of growth is required because only juveniles can overwinter in the dry pond (Way et al., 1980; Hornbach and Childers, 1986). These over- wintering juveniles grow to adults in the spring, reproduce and die during the summer. There is some evidence, however, that if temporary ponds remain wet through the fall, some portion of the juveniles produced in the spring and summer grow and reproduce in the fall (Hornbach ef al. , 1980). There are varying genotypes in the population (McLeod ef al., 1981). It is possible that each is associated with a different metabolic rate as juvenile and this contributes to the observed variability (Table 1). The fact that juvenile Musculium partumeium must survive extremely long periods of anoxia, compared to many marine bivalves, could explain why their patterns of O, consumption dependency on O, availability do not follow those described by Newell (1979) for marine species. The results of this study indicate that O, availability can influence O, consumption in Musculium partumeium but that the influence is size-specific. Furthermore it appears that the degree of variability in O, consumption is also influenced by size. ACKNOWLEDGMENTS I wish to thank Karen Moulton, Tony Deneka and Robert Dado for their assistance in collecting specimens and technical assistance in the laboratory. LITERATURE CITED American Public Health Association. 1980. Standard Methods for the Ex- amination of Water and Waste Water, \Sth ed. American Public Health Association, Washington, D.C. 1134 pp. 42 AMER. MALAC. Bayne, 8. L. 1971. Oxygen consumption by three species of lamellibranch mollusc in declining ambient oxygen tension. Comparative Biochemistry and Physiology 40A: 955-970. Bayne, B. L. 1973. The response of three species of bivalve mollusc to declin- ing oxygen tension at reduced salinity. Comparative Biochemistry and Physiology 45A: 793-906. Berg, K., P. M. Jonasson and J. Ockelmann. 1962. The respiration of some animals from the profundal zone of a lake. Hydrobiologia 19: 1-39. Buchwalder, S. H. 1983. Seasonal respiratory adaptations to temperature and oxygen availability in the freshwater clams Musculium partumeium (Say) and Musculium lacustre (Muller). Master’s Thesis. Universi- ty of Dayton, Ohio. 62 pp. Burch, J. B. 1975. Freshwater Sphaeriaceans Clams (Mollusca: Pelecypoda) of North America. Malacological Publications, Hamburg, Michigan. 96 pp. Burky, A. J. 1983. Physiological ecology of freshwater bivalves. In: The Mollusca, Vol. 6, Ecology. W.D. Russell-Hunter, ed. pp. 281-327. Academic Press, New York. Burky, A. J. and K. A. Burky. 1976. Seasonal respiratory variation and ac- climation in the pea clam, Pisidium walkeri Sterki. Comparative Biochemistry and Physiology 55A: 109-114. Burky, A. J., D. J. Hornbach and C. M. Way. 1985. A bioenergetics ap- proach to life-history tactics: Comparisons of permanent and tem- porary pond populations of the freshwater clam, Musculium par- tumeium (Say). Hydrobiologia 126:35-48. Eggleton, F. E. 1931. A limnological study of the profundal bottom fauna of certain freshwater lakes. Ecological Monographs 1:231-331. Gale, W. F. 1976. Vertical distribution and burrowing behavior of the fingernail clam, Sphaerium transversum. Malacologia 15:401-409. Holopainen, I. J. 1987. Seasonal variation of survival time in anoxic water and the glycogen content of Sphaerium corneum and Pisidium am- nicum (Bivalvia, Pisidiidae). American Malacological Bulletin 5:41-48. Holopainen, I. J. and P. M. Jonasson. 1983. Long-term population dynamics and production of Pisidium (Bivalvia) in the profundal of Lake Esrom, Denmark. Oikos 41:99-117. BULL. 9(1) (1991) Hornbach, D. J. 1985. A review of metabolism in the Pisidiidae with new data on its relationship with life history traits in Pisidium casertanum. American Malacological Bulletin 3:187-200. Hornbach, D. J. and D. L. Childers. 1986. Life-history variation in a stream population of Musculium partumeium (Bivalvia: Pisidiidae). Journal of the North American Benthological Society 5:263-271. Hornbach, D. J., C. M. Way and A. J. Burky. 1980. Reproductive strategies in the freshwater sphaeriid clam, Musculium partumeium (Say), from a permanent and a temporary pond. Oecologia (Berlin) 44:164-170. Hornbach, D. J., T. E. Wissing and A. J. Burky. 1983. Seasonal variations in the metabolic rates and Q)9-values of the fingernail clam, Sphaerium striatinum Lamarck. Comparative Biochemistry and Physiology 76A :783-790. Juday, C. 1908. Some aquatic invertebrates that live under anaerobic condi- tions. Transactions of the Wisconsin Academy of Science, Arts and Letters 16:10-16. McLeod, M. J., D. J. Hornbach, S. I. Guttman, C. M. Way and A. J. Burky. 1981. Environmental heterogeneity, genetic polymorphism and reproductive strategies. American Naturalist 118:129-134. Newell, R. C. 1979. Biology of Intertidal Animals. Marine Ecological Surveys, Ltd., Kent, Faversham. 757 pp. Pierce, R. J., R. L. Wiley and T. E. Wissing. 1973. Interconversion of units in studies of the respiration of aquatic organisms. Progressive Fish Culturist 35:207-208. Thomas, G. J. 1963. Study of a population of sphaeriid clams in a temporary pond. Nautilus 77:37-43. Thomas, G. J. 1965. Growth in one species of sphaeriid clam. Nautilus 79:47-54. Waite, J. and G. Neufeld. 1977. Oxygen consumption by Sphaerium simile. Comparative Biochemistry and Physiology 57A:373-375. Way, C. M., D. J. Hornbach and A. J. Burky. 1980. Comparative life history tactics of the sphaeriid clam, Musculium partumeium (Say), from a permanent and a temporary pond. American Midland Naturalist 104:319-327. Date of manuscript acceptance: 9 July 1991 INTEGRATIVE NEUROBIOLOGY AND BEHAVIOR OF MOLLUSCS Organized by ROGER T. HANLON MARINE BIOMEDICAL INSTITUTE UNIVERSITY OF TEXAS MEDICAL BRANCH and ALAN M. KUZIRIAN MARINE BIOLOGICAL LABORATORY AMERICAN MALACOLOGICAL UNION WOODS HOLE, MASSACHUSETTS 4 - 5 JUNE 1990 43 ed Sexual conflict and the mating systems of simultaneously hermaphroditic gastropods Janet L. Leonard* Department of Zoology, University of Oklahoma, Norman, Oklahoma 73019, U. S. A. Abstract. Here I review the predictions, for hermaphroditic gastropods, of recent developments in mating systems and sexual conflict theory. Sexual conflict theory predicts that hermaphrodites should have a species-specific preferred sexual role. The Hermaphrodite’s Dilemma model explores the conse- quences of this and predicts that all hermaphrodite mating systems should be based on reciprocity with cheating in a preferred role. Traditional models based on Bateman’s principle predict that the male role will be preferred. Experimental and observational evidence from Navanax inermis (Cooper) indicate that the female role is preferred, contrary to predictions from Bateman’s principle, and that the mating system is based on sperm-trading which serves to enforce reciprocation, preventing individuals from specializing in the female sexual role. Comparison of Navanax to other hermaphrodites suggests that the preferred sexual role is that which offers control of fertilization. This, the gamete-trading model, predicts that all hermaphroditic gastropods with sperm storage and a gametolytic gland should demonstrate a preference for the female role and a mating system based on sperm trading. This model and the Hermaphrodite’s Dilemma model of strategies in a situation of sexual conflict make specific predictions about the behavior of hermaphroditic gastropods. The available literature on opisthobranchs and pulmonates suggests several interesting tests of these models but the available data are insufficient to support or refute the predictions. The mating systems of euthyneuran gastropods require investigation from the standpoint of modern mating systems theory. Recent progress in understanding the evolution of reasons: 1) it could provide new and useful insights into the species-typical reproductive behaviors stems largely from the biology and evolutionary ecology of taxa of interest; 2) by analysis of mating systems in terms of a conflict of interests examining the mating systems of a diverse array of organisms between the sexes (Orians, 1969; Trivers, 1972; Emlen and in terms of sexual conflict and/or sexual selection, we should Oring, 1977; Parker, 1979; Hammerstein and Parker, 1987). be able to identify useful systems for testing some of the Because such sexual conflict is assumed to be a product of assumptions and predictions of modern mating systems the differential selective pressures associated with reproduc- theory. tion through sperm versus reproduction through eggs (Parker, The gastropods are particularly interesting in this 1979), it should exist in all anisogamous organisms (Bateman, respect because they offer: 1) a diverse array of sexual systems 1948). Although work in this field (despite Ghiselin, 1974; and reproductive strategies; 2) a variety of very complex Williams, 1975) has concentrated on a small number of taxa reproductive behaviors; 3) complex genitalia, the anatomy (largely vertebrates and insects), almost all of which have of which is important taxonomically in many groups. Where separate sexes, there has been increasing interest in the role genital anatomy is varied sufficiently to be a useful taxonomic of sexual conflict and/or sexual selection in the mating systems character at the levels of genus, subgenus and species, as in of a wider array of organisms, such as plants, including some groups of gastropods (Mead, 1943; Rudman, 1974; hermaphroditic forms (Willson and Burley, 1983; Bronstein, Gilbertson, 1989; Patterson, 1989), sexual selection is likely 1988; Galen and Rotenberry, 1988; Nakamura and Stanton, to have been important (Eberhard, 1985). Similarly, where 1989) and some animals with sequential hermaphroditism there are elaborate forms of courtship and copulatory (primarily fish, see Charnov, 1982, 1986; Shapiro and behavior, and particularly where there is significant diversi- Boulon, 1982; Warner, 1982; Warner and Lejeune, 1985) and ty within a taxonomic group, one predicts that sexual con- more recently simultaneous hermaphroditism (e.g. serranid flict and/or sexual selection has been important as a selec- fishes; Fischer, 1980, 1984; Fischer and Petersen, 1987; and tive force. the polychaete, Ophyotrocha; Berglund, 1985; Sella, 1985, In this paper, I 1) present predictions as to the types 1988). of sexual behavior and mating systems expected for her- Extension of the analysis of sexual conflict to a diverse maphroditic gastropods if sexual conflict is important and array of taxa and modes of sexuality is important for two 2) discuss a few well-known sexual phenomena in the two *Address correspondence to: Mark O. Hatfield Marine Science Center, predominantly hermaphroditic subclasses of gastropods Oregon State University, Newport, Oregon 97365, U. S. A. (Opisthobranchia and Pulmonata) in light of predictions of American Malacological Bulletin, Vol. 9(1) (1991):45-58 45 46 AMER. MALAC. BULL. 9(1) (1991) it models based on the assumption of sexual conflict. These two subclasses, collectively (loosely) termed the euthyneuran gastropods, are predominantly outcrossing simultaneous hermaphrodites. Hermaphroditic species offer exciting opportunities to test: 1) the hypothesis that sexual conflict exists (Leonard, 1990), because as Parker (1979) has pointed out, sexual conflict has been widely assumed but there is little evidence that it exists; 2) alternate hypotheses as to the source and nature of sexual conflict (Leonard and Lukowiak, 1991). SEXUAL CONFLICT SEXUAL CONFLICT AND HERMAPHRODITE MATING SYSTEMS Sexual conflict is a conflict of interests between the two parties to a mating encounter such that one individual has more to gain (less to lose) by mating than the other does. The idea of sexual conflict is based on the common observa- tion that among many species of animals, males are “‘eager’’ to mate with virtually any available female, to the extent of risking death and/or serious injury in fighting other males for access to females, whereas females are ‘‘coy’’. This phenomenon, which Darwin (1874) considered paradoxical, is usually explained by Bateman’s (1948) principle, i.e. that males are more eager to fertilize eggs than females are to get their eggs fertilized because the fitness of females is limited by the resources available for egg production, whereas the fitness of males is only limited by the availability of females. For a simultaneous hermaphrodite, sexual conflict arises if and when there is more to be gained from mating in one sexual role than in the other. In the population as a whole, reproductive success through eggs must exactly equal reproductive success through sperm (R. A. Fisher, 1958). However, the distribution of reproductive success across the population could well differ for sperm and eggs (i.e. the variances differ, see Charnov, 1979). When this is the case, there is a potential asymmetry in the pay-offs of the two sex- ual roles to an individual, particularly in a single encounter, and an individual’s overall fitness (its reproductive success relative to the rest of the population) will depend in part on how it divides its reproductive effort (or reproductive oppor- tunities) between the two sexual roles. That is, an individual which specializes (differs from the population average of 50:50 reproduction through sperm versus eggs) in one of the roles (that which is less costly in terms of energy expenditure, mating time, risk, etc., see below) could be able to achieve greater than average fitness. This means of course that other individuals in the population will find themselves specializ- ing inadvertently in the more costly role, which should result in below average fitness. Charnov (1979:2482) discussed the implications of Bateman’s principle for pair-mating hermaphroditic animals, including gastropods, and recognized that, in simultaneous hermaphrodites; ‘‘There must often be a conflict of interest between mating partners — as a recipient each should be in- clined to accept sperm (not necessarily for fertilization of its own eggs) in order to give its sperm away. As a donor, one should be selected to induce one’s partner to use the new sperm in fertilization’’. He went on to suggest that both the complicated reproductive anatomy and the elaborate precopulatory behaviors of animals such as gastropods ‘‘are explicable when one realizes that the interests of the part- ners are often in conflict’’. In hermaphrodites this sexual con- flict is direct, in that each individual is in direct competition with all other individuals, including its mate, for fitness. On the other hand in dioecious species sexual conflict (with the exception of conflict over parental care) is an epiphenomenon of intrasexual competition (Hammerstein and Parker, 1987). Thus, in simultaneous hermaphrodites sexual conflict should lead to the evolution of a preference for mating in a particular sexual role, a preference that will be shared by all individuals of the species. If all individuals prefer the same role, the interests of two individuals meeting for a mating encounter will be in direct conflict. In hermaphrodites, each mating encounter could be expected to involve competition between members of the pair for the preferred role. Mating systems in hermaphrodites should reflect or represent a resolution of, this competiton for the preferred role (Leonard, 1990, unpub. data; see also below). THE ORIGIN OF SEXUAL CONFLICT If sexual conflict exists, the preferred sexual role will be consistent within a species, since the term implies an in- herent advantage to one sexual role, but the favored role could vary between species, depending on the source of sexual con- flict. Charnov based his arguments (Charnov, 1979, 1982) on the explicit assumption of Bateman’s principle, i.e. that male fecundity is limited by access to eggs while female fecundity is not limited by sperm availability. While Bateman’s princi- ple is the most widely accepted explanation of sexual con- flict, there are two general types of alternative explanation: mating time or rate (Baylis, 1981; Sutherland, 1986); and con- trol of fertilization or risk (Alexander and Borgia, 1979; Leonard and Lukowiak, 1984, 1985, 1991). The first of these, the rate hypothesis, argues that a discrepancy between males and females in the minimum interval between matings, could be a source of sexual conflict. That is, if after a mating en- counter, one sex (say the female) must spend a prolonged period in yolking up eggs, or some form of parental care (gestation, brooding, etc.) before being able to increase her fitness by a second mating encounter, whereas the male can inseminate many females in that same time period then one could expect the female to be ‘‘coy’’ and the male, ‘‘eager’’, even in the absence of differential energy expenditure. For a discussion of how mating rate can favor the evolution of LEONARD: GASTROPOD MATING SYSTEMS 47 male parental care see Baylis (1981). The second type of hypothesis is based on the idea that the important dichotomy between reproducing through eggs and reproducing through sperm may not be differential costs in a fixed currency (such as energy or time) but rather a dif- ferential probability that the investment that is made will pay off in the form of zygotes. If gamete production involves use of an exhaustible resource, then in economic terminology, optimal depletion of that resource (i.e. depletion which max- imizes profit) must be based not only on accounting costs (the energy, time, etc. required for a particular type of reproductive act) but also the opportunity cost (that is the cost of a reproductive act now must include the value of that act if it were made at some future date, or the pay-off that could have been achieved from the other role, etc.) (Nicholson, 1978; A. C. Fisher, 1981). Alexander and Borgia (1979) suggested that an important difference between the sexes could be the extent to which control is exerted over the fate of the gametes that they produce. Alexander and Borgia (1979) argued that females, in general, produce fewer gametes not because they are limited by energy availability but rather because they invest more in ‘‘following-up’’ on that gamete, insuring that it will be fertilized. That is, one can think of females as adopting a “‘risk-averse’’ reproductive strategy whereas males have a ‘‘risk-prone’’ reproductive strategy. This type of model has some interesting implications for hermaphrodites. The unique feature of hermaphrodite sexual strategies is obviously the opportunity to choose between reproduction through eggs or sperm. If, as is required for sexual conflict, the variances of these two types of strategy differ, which should be preferred, the high variance strategy or the low variance strategy? Extension of Bateman’s principle to hermaphrodites (Bateman, 1948; Charnov, 1979) suggests that the male role (the high variance strategy) will be preferred (see above). However, Gillespie (1977) has demonstrated that where two genotypes yield equal average offspring number, but have unequal variances, the one with the lower variance will offer a fitness advantage. This suggests that given an op- portunity to choose between a high and a low variance reproductive strategy, selection would favor the individual that opted for the low variance strategy (Leonard and Lukowiak, 1991). In Gillespie’s (1977: 1012) words, ‘‘the addition of a stochastic element to the offspring number of a genotype will effectively lower the fitness of that genotype as measured by its mean frequency in the next generation’. A practical problem in using Gillespie’s principle to predict the favored sexual role for a hermaphroditic species is that we have little empirical evidence as to the relative variances of offspring production through male versus female function. In general, however, male reproductive success has been assumed to have the greater variance. Under this assumption, predictions from Gillespie’s principle would be very different than those from Bateman’s principle. The gamete-trading model (Leonard and Lukowiak, 1984, 1985), developed from a comparison of the mating systems of the aglajid opisthobranch, Navanax inermis (Cooper) (Leonard and Lukowiak, 1984, 1985) (for discus- sion of the validity of the genus see Rudman, 1974; Gosliner, 1980), and a serranid fish (Fischer, 1980), is based on the premise that the preferred role for a simultaneous hermaphro- dite will be the one that controls fertilization, i.e. is the last to make an irrevocable commitment of gametes. The sexual role that controls fertilization affords the greatest certainty that the investment made will result in zygotes. That is, the preferred role will have the greatest certainty of parenthood. Considered from the standpoint of selection if two reproduc- tive options will have the same mean pay-off but one is less risky than the other, an individual with limited resources to invest will do better to play it safe (Gillespie, 1977; Philippi and Seger, 1989). Such competing hypotheses as to the source of sexual conflict can be tested in hermaphrodites by finding species for which two of the hypotheses make opposite predictions as to the sexual role that should be favored and then deter- mining which role is in fact preferred. For example, ex- perimental studies designed to test the conflicting hypotheses that in Navanax inermis, the male role is preferred as would be expected from Bateman’s principle and the egg-trading model (Fischer, 1980; see discussion in Leonard and Lukowiak, 1991), versus the conflicting hypothesis that the female role is preferred (based on the gamete-trading model), indicate a preference for the female role (Leonard and Lukowiak, 1991). In this way, the debate over the factors responsible for sexual conflict can be moved from a theoretical level to an experimental one. DOES SEXUAL CONFLICT SHAPE MATING SYSTEMS? The question of the importance of sexual conflict in shaping mating systems could be best answered by compara- tive studies. That is, we need to decide whether the mating systems observed in simultaneous hermaphrodites conform to what would be expected if sexual conflict were important. A model in the form of a game of strategy, termed Hermaph- rodite’s Dilemma, has been developed to analyze the situa- tion that arises given a conflict of interest between two simultaneous hermaphrodites in a mating encounter (Leonard, 1990). This model can provide qualitative predictions as to the type of behavior and mating system to be expected under a variety of conditions. HERMAPHRODITE’S DILEMMA Briefly, Hermaphrodite’s Dilemma is a two-person, 48 AMER. MALAC. BULL. 9(1) (1991) non-zero-sum, conditional game of strategy; the available decisions are: 1) to offer to assume both roles in a mating encounter (the cooperate decision), or 2) to mate only in the preferred sexual role and then desert without allowing the partner to mate in the preferred sexual role (the defect deci- sion). Assuming that both players are, being simultaneous hermaphrodites, prepared to mate in both roles at all times, and that there is an advantage to mating in one role, an in- dividual ought to be willing to assume that role in any and all encounters. Therefore, its decision in a mating encounter is not whether or not to mate in the preferred role, but whether or not it should also mate in the non-preferred role. The model predicts that the best strategy over a wide variety of condi- tions will be one that combines cooperating (reciprocation) most of the time, with a certain low level of defection (= cheating), which could or could not be contingent on the part- ner’s prior behavior, depending on circumstances [i.e. w, the probability of encountering a partner again, (Axelrod and Hamilton, 1981), and whether the pay-off matrix corresponds to Prisoner’s Dilemma or Chicken]. That is, given sexual con- flict, the mating system of simultaneous hermaphrodites should be based on reciprocation, with cheating in a species- typical preferred sexual role occurring at a relatively low frequency, and the existence of mechanisms to reduce vulnerability to cheaters (Leonard, 1990). Furthermore, the model suggests that mechanisms should exist to prevent cheating and/or ‘‘punish’’ cheaters. Although the arguments developed here should apply to all hermaphroditic gastropods, the discussion will focus on the pulmonate and opisthobranch (lumped here as ‘‘euthy- neuran’’) gastropods because they offer a wide array of forms of (often bizarre) reproductive behavior [communal parental care (Rose and Hoegh-Guldberg, 1982), hypodermic copula- tion (Rivest, 1984), chain copulation (Aplysia and other species), and elaborate and bizarre forms of courtship (Helix, Limax and some other stylommatophorans)] that have been little studied. Even the familiar cases (i.e. Helix, Limax, Aplysia) have not been studied in light of modern mating systems theory. The analysis presented here suggests new in- terpretations of familiar phenomena in gastropod biology and proposes tests of important assumptions and predictions of mating system theory using comparative and experimental studies of these gastropods. The Hermaphrodite’s Dilemma model predicts that where unilateral copulation is the rule, 1) mechanisms for enforcing reciprocation exist [such mechanisms could include explicit alternation of sexual roles, such as that described for Navanax (Leonard and Lukowiak, 1984, 1985)], 2) effective reciprocity is achieved by random mating within a small mating group [there is some evidence to suggest that Aplysia californica Cooper form small, relatively stable mating clusters (Kupfermann and Carew, 1974)], or 3) willingness to copulate in the less preferred role is maintained by the con- ditions of the Chicken matrix (Riechert and Hammerstein, 1983). For example, Lymnaea stagnalis (L.) are always ready to copulate as females, the preferred role according to the gamete-trading model, but become willing to copulate as males after periods of isolation (v. Duivenboden and ter Maat, 1985). Alternatively, if the mating system were not based on reciprocation, one would expect to see aggressive attempts to force copulation by individuals acting in the favored sex- ual role, with the victim attempting to avoid copulation and/or to retaliate by assuming the favored role, as could be the case in some leeches (Leonard, unpub. data). The problem faced by a pair of simultaneous her- maphrodites in a mating encounter differs from that described in the classical Prisoner’s Dilemma (Luce and Raiffa, 1957; Davis, 1983) or in the formal Hermaphrodite’s Dilemma model (Leonard, 1990) in that a player has information about the decision made by its partner, and in general, one player will have to make the first move. One possible tactic in this situation would be the ‘‘Quick-Draw’’ approach; making the first move and assuming the preferred role, leaving the part- ner to either assume the non-preferred role or pass up the chance to mate. If the partner agreed to assume the non- preferred role, it would be possible for the initiator to either ‘‘cheat’’ (by leaving without reciprocating) or to reciprocate, by offering to mate in the non-preferred role also. This type of tactic could result in the evolution of either 1) ‘‘Hit-and- Run’’ mating encounters, such as those of some leeches (see above), or 2) insistence by the party of the second part on strong assurance that the initiator will reciprocate, before the party of the second part assumes the non-preferred sexual role. The simultaneous reciprocal copulation found in many nudibranch and stylommatophoran gastropods and clitellate annelids (earthworms and some leeches) could have evolved in this way. In species in which mating is not (or cannot be) simultaneously reciprocal, an individual pursuing a strategy such as Tit-for-Tat should advertise its willingness to reciprocate, in order to attract potential partners or to avoid rejection by a partner that it has located. Assuming the non- preferred sexual role in the first mating would be a way of accomplishing this. For such a tactic to be successful, there would have to be some protection against ‘‘cheating’’ by the partner. Such tactics appear to be employed as part of a Tit- for-Tat strategy in the mating systems of certain serranid fishes (Fischer, 1980, 1984) and at least one opisthobranch gastro- pod, Navanax inermis (Leonard and Lukowiak, 1984, 1984, 1991). EUTHYNEURAN GASTROPODS GENERAL REPRODUCTIVE BIOLOGY (for review see Tompa et al., 1984) All but a very few species of the euthyneuran (Sub- LEONARD: GASTROPOD MATING SYSTEMS 49 classes Opisthobranchia and Pulmonata) gastropods are simultaneously hermaphroditic. Among euthyneuran groups, gonochorism is common only in the opisthobranch order Acochlidioidea (Hadfield and Switzer-Dunlap, 1984; but see Wawra, 1988). While many authors refer to protandry in these taxa, this usually means only that histological investigations show mature sperm present before eggs have matured. For example, in Limacina, an opisthobranch considered to be pro- tandrous, copulation appears to be simultaneously reciprocal between mature males (Lalli and Wells, 1978; other examples in Hadfield and Switzer-Dunlap, 1984). This usage has been common in the molluscan literature since at least the end of the last century (e.g. Pelseneer, 1895). Storage of allosperm (and often autosperm) is common and individuals could receive sperm which they will later use to fertilize eggs, long before eggs have been formed in the ovotestis. True sequen- tial hermaphroditism, in which an individual is purely a sperm donor at one stage of its life and only a sperm recip- ient at another, is rare, if it in fact exists, in euthyneuran snails (Geraerts and Joose, 1984; Hadfield and Switzer-Dunlap, 1984; Tompa, 1984), although it is not uncommon in the prosobranch gastropods. One must be cautious, therefore, in interpreting references to functional protandry (see discus- sion in Ghiselin, 1965). Reproduction through true parthenogenesis (probably automitic) has been reported for one stylommatophoran slug (Hoffmann, 1983; see also discussion in Tompa, 1984). Although the capacity for self-fertilization is not uncommon in pulmonates, and apparently exists in a few opisthobranchs (Hadfield and Switzer-Dunlap, 1984), cross-fertilization is the rule and the vast majority of euthyneuran gastropods act as simultaneous hermaphrodites throughout their reproduc- tive lives. The anatomy of the reproductive system of euthyneuran gastropods can be summarized as very complex, highly diverse, and taxonomically valuable at the species level. As Eberhard (1985) pointed out, these characteristics are in- dicative of rapid evolution, probably as a result of sexual selection. This, in turn, suggests that these gastropods should be characterized by diverse, complex and fascinating sexual behavior and mating systems. Unfortunately, our under- standing of the reproductive biology of these gastropods is very scanty. For most species only the gross anatomy of the genitalia has been described; less is known at the histological level, and there have been relatively few studies that have looked directly at the physiology of various parts of the reproductive tracts. In many cases it is difficult to imagine how the genitalia would look when everted and how the various parts would fit together during copulation (see Reeder, 1986). Also, most of the available information on sexual behavior consists of casual or anecdotal observations. In terms of sexual behavior, euthyneuran gastropods can be divided into three groups; those in which copulation is normally simultaneously reciprocal, those in which it is unilateral, and those in which copulation is unilateral and chains of copulating individuals occur commonly. As a broad genralization, simultaneous reciprocal copulation occurs in taxonomic groups in which the penis and common genital aperture (= vaginal pore) are close together on the body, while unilateral copulation is typical of taxonomic groups in which these structures are widely separated. Speaking again very generally, one can say that simultaneous reciprocal copulation is characteristic of the stylommatophoran pulmonates (exceptions include Stenotrema, Webb, 1948), whereas the basommatophorans have unilateral, and sometimes chain, copulation (Geraerts and Joose, 1984; Tompa, 1984). Among the opisthobranchs, simultaneous reciprocal copulation (and/or sperm transfer) is found in most of the nudibranchs, notaspideans, saccoglossans and pteropods, whereas unilateral and/or chain copulation (or sperm transfer) is more common in cephalaspideans, anaspi- deans, pyramidelloideans and acochlidioideans (Hadfield and Switzer-Dunlap, 1984). True copulation is the general mode of sperm transfer in euthyneuran gastropods, but hypodermic insemination and aphallic spermatophore transfer have evolved several times among opisthobranchs (Hadfield and Switzer-Dunlap, 1984). Some land slugs deposit sperm on the tip of the partner’s penis (Gerhardt, 1933; Tompa, 1984). Aphallic sperm transfer apparently occurs in some basom- matophorans (Geraerts and Joose, 1984). In the nudibranch Aeolidia sperm are deposited on the outer surface of the female gonopore (Longley and Longley, 1984). OPPORTUNITY FOR SPERM COMPETITION SPERM STORAGE Although the functions of the myriad organs of gastropod reproductive systems are poorly understood, one can say that both opisthobranchs and pulmonates typically have organs for storage of allosperm and there is some evidence that allosperm could require a period of residence in the sperm storage organ before becoming competent to fer- tilize eggs (Hadfield and Switzer-Dunlap, 1984). The period of residence in the sperm storage organ is uncertain for most species. Hadfield and Switzer-Dunlap (1984) reported that while some opisthobranchs show a close association between copulation and egg laying, with egg laying following copula- tions at a fairly predictable interval, this is not always the case. Individuals could copulate many times without spawn- ing. One Navanax inermis, under laboratory conditions, copulated 24 times in the female role before laying an egg mass (Leonard and Lukowiak, 1985). The record for the number of egg masses laid between copulations is apparent- ly still held by an Aplysia californica taken to Woods Hole, Massachusetts, and held in isolation by MacGinitie (1934). That individual laid eggs for more than four months and the 50 AMER. MALAC. BULL. 9(1) (1991) first (5S egg masses were fertile, demonstrating that Aplysia can store sperm for at least two and one-half months and that the sperm storage organ can hold enough sperm to fertilize at least 15 egg masses. In Navanax, one isolated individual laid nine egg masses over the course of a month. The first five egg masses developed normally; the last four were inviable. After being allowed to copulate again (once as a female and once as a male) it laid two egg masses. The first, a small inviable one, was laid within 24 hours of copulation and a large viable egg mass was laid 24 hours later. No further eggs were laid although the individual was maintained in the laboratory for another month (Leonard, unpub. data). In one case, a Navanax laid a fertile egg mass after 32 days of isolation (Leonard and Lukowiak, 1985). In Navanax both the frequency and size of egg masses vary, and this could reflect size of the animal, food availability and/or sperm stores (Leonard, unpub. data). In isolated Hermissenda crassicornis (Escholtz), an aeolid nudibranch, Rutowski (1983) reported that some individuals produced as many as eight egg masses that were at least partially fertile. He reported that Hermissenda isolated upon collection produced an average of 3.7 + 2.5 fertile egg masses within 24 days. He reported 20 days as the longest interval between fertile egg masses but did not indicate how long sperm can be stored after a copula- tion. In that study, sperm-deleted animals laid egg masses approximately three days after receiving an intromission. Another aeolid, Phestilla melanobranchia Bergh, can lay an average of 1.5 fertile egg masses per day for two weeks before depleting the sperm received in a single mating (Harris, 1975). The possibility of self-fertilization complicates analysis of the relationship between egg laying and copulation in both basommatophoran and stylommatophoran pulmonates. In at least two species of the basommatophoran Lymnaea, egg lay- ing begins sooner in mated than in isolated individuals (Horstmann, 1955; Boray, 1964; van Duivenboden, 1983, 1984). Horstmann (1955) established that this effect was mediated by the presence of allosperm in the gametolytic gland. However, van Duivenboden (1984) reported Lymnaea raised in groups laid fewer eggs than did isolated individuals. NUMBER OF MATES There are at least two reports of apparent monogamy in opisthobranchs. Lalli and Wells (1978: 103) concluded, from anatomical evidence, that in the pelagic pteropod Limacina inflata (d’ Orbigny) a spermatophore is formed from the prostate gland which must be reciprocally transferred to a partner, because “‘all L. inflata of the proper size have either a well-developed prostate gland or a spermatophore received from another individual, but never both structures’’. Since this species is anatomically protandrous, an individual presumably mates only once in its lifetime. A different type of monogamy, involving iteroparity, has been reported from aeolid nudibranchs of the genus Phestilla. Rudman (1981: 408) reported that for two species from Tanzania, ‘‘At a very early stage individuals would pair with another individual of the same species and they would remain nestled together for their whole lifespan, except when egg-laying or feeding.’ In- dividual mature Phestilla deposit one or two egg masses per day for several weeks (Harris, 1975). However, as Hadfield and Switzer-Dunlap (1984) suggested, in most opisthobranch species individuals probably mate with a number of different individuals over their lifespan. Tompa (1984) considered that multiple mating is pro- bably the rule for stylommatophorans and cited evidence from Cepaea nemoralis (L.) (Murray, 1964) that an average brood is sired by two individuals and that 10-20 spermatophores have been found in an individual. Lind (1988) concluded that sperm competition will often occur in Helix pomatia L. under field conditions. Multiple paternity has also been reported for basommatophorans (Mulvey and Vrijenhoek, 1981; Monteiro et al., 1984; Rudolph and Bailey, 1985). In summary, the available information, while scanty, suggests that while various types of monogamy can occur, most euthyneuran gastropods mate with several sexual partners over their lives and that there is considerable potential for sperm competi- tion and multiple paternity of egg masses. Furthermore, the ability to store sperm for long periods of time can create special problems relative to certainty of paternity. Studies of gastropod reproductive anatomy and physiology from the standpoint of sexual selection and/or sexual conflict acting through sperm competition may shed light on both the func- tion of gastropod reproductive tracts and the evolutionary biology and speciation of gastropods. PREDICTIONS FROM THEORY As stated above, the Hermaphrodite’s Dilemma model predicts that the mating systems of simultaneous _her- maphrodites, including gastropods should a) be based on reciprocity and b) involve a detectable level of “‘cheating’’ in a favored role. In some euthyneuran gastropods, reciprocity in the mating system is obvious and here the novel prediction is that of ‘‘cheating’’ in a favored role. In other groups reciprocity is not obvious or not known and thus is, in and of itself, a strong prediction of the model. A further problem is to predict which sexual role will be preferred. Bateman’s principle and the egg-trading model of hermaphrodite mating systems (Charnov, 1979; Fischer, 1980; see also discussion in Leonard and Lukowiak, 1991) predict that the male role will be preferred except where the male contributes something other than sperm to the mating (i.e. parental care, a nutri- tional investment, etc.) which makes male parental invest- ment larger than female parental investment. Because one cannot determine which investment is larger a priori, I LEONARD: GASTROPOD MATING SYSTEMS 5] assume here, for the sake of simplicity, that the egg-trading model consistently predicts a preference for the male role. The gamete-trading model (Leonard and Lukowiak, 1984, 1985) predicts that where the female controls fertilization the mating system will be based on sperm trading. That is, the female role will be preferred in general in euthyneuran gastropods, and that exceptions should be found in those species that lack a gametolytic gland and/or sperm storage. Thus, for most species of euthyneuran gastropod the egg- trading model predicts a preference for the male role and the gamete-trading model predicts a preference for the female roles. Specific predictions and tests of these models are discussed below. CASE STUDIES OF MATING SYSTEMS NAVANAX INERMIS AND SPERM-TRADING There has been some progress in the analysis of sex- ual conflict in one of the unilaterally copulating opistho- branchs, Navanax inermis (Leonard and Lukowiak, 1984, 1985, 1987a, 1991), which has served as the stimulus for development of the gamete-trading and Hermaphrodite’s Dilemma models. The mating system of Navanax 1s based on reciprocation in that pairs of individuals actively alter- nate sexual roles over a series of copulations. This mating system is analogous in many respects to that of the egg-trading serranid fishes (Leonard and Lukowiak, 1984, 1985; Leonard, unpub. data), in that two individuals remain together for a series of copulations but there is no long-term pair bond. The major difference is that in serranids the male role is preferred as predicted by Bateman’s principle (Fischer, 1980; Leonard, unpub. data). In Navanax, both qualitative observations (Leonard and Lukowiak, 1984, 1985) and experimental tests (Leonard and Lukowiak, 1991) indicate that it is female, rather than the male, sexual role that is preferred. The preference for the female sexual role in Navanax has been hypothesized to be a consequence of female control of fertilization (Leonard and Lukowiak, 1984, 1985, 1991). In Navanax, as in most euthyneuran gastropods (Pruvot-Fol, 1961; Tompa, 1984; Geraerts and Joosse, 1984; Hadfield and Switzer-Dunlap, 1984), there is both a sperm storage organ and a gametolytic gland (Rudman, 1974), and ovulation is not tied to copulation. Consequently, sperm transferred to a partner could be ‘‘wasted’’ in that they could be digested rather than stored, and if stored they could or could not be used for fertilization while they (the sperm) are still viable. In the serranids, on the other hand, female court- ship displays are closely tied to ovulation and the ‘‘male’’ has reliable information as to the onset and duration of spawning (Fischer, 1980). Therefore, the ‘‘male’’ has greater control over the fate of its gametes than does the ‘‘female,”’ and this could be generally true of externally-fertilizing fishes (Alexander and Borgia, 1979). The gamete-trading model predicts that simultaneous hermaphrodites will prefer the sex- ual role that offers control of fertilization (Leonard and Lukowiak, 1984, 1985, 1991). Therefore, the female role should be preferred in most euthyneuran gastropods (Leonard and Lukowiak, 1985). The mating system of Navanax, termed sperm-trading, represents a mirror image of the egg-trading system. In Navanax a sexual encounter is initiated by an individual track- ing down, courting and copulating in the male role (indicating willingness to reciprocate by starting out in the less-preferred sexual role). The available evidence suggests that males main- tain intromission until the partner reciprocates (Leonard and Lukowiak, 1984, 1985, 1987a, 1991). This could serve to en- force reciprocation by preventing the partner from mating again as a female before serving as a male to its current part- ner. The (indirect) evidence that ‘‘cheating’’ occurs in Navanax is twofold: 1) reciprocation does not always occur, and 2) the complexity and variability of behavior observed during alternation of sexual roles suggests that cheating at- tempts may be occurring at this time (Leonard and Lukowiak, 1985, 1987a). In Navanax, sperm transfer is not directly ob- servable and at present there is no evidence as to whether or not “‘subtle cheating’’ (Trivers, 1971) in the form of failures to transfer sperm during a copulation, or the transfer of sub- standard quantities of sperm, is occurring. BIOMPHALARIA AND SPERM-SHARING Evidence for what could represent a type of ‘‘subtle,”’ probably even ‘‘victimless’’ cheating on a sperm-trading mating system, does exist in planorbid basommatophoran pulmonates of the genus Biomphalaria. Using genetic markers, Monteiro et al. (1984) demonstrated that Biom- phalaria copulating as males sometimes transfer, to a female partner, sperm that they have received from a previous part- ner. That is, a snail copulating as a male may inseminate its partner with allosperm instead of, or along with, autosperm. This phenomenon appeared paradoxical at first glance because Bateman’s principle can not explain how an in- dividual would benefit by distributing someone else’s sperm. The most probable functional explanation of this phenomenon is that it is a form of ‘‘cheating’’ in a sperm-trading mating system, serving to allow a snail to receive more sperm from a partner than it gives up (in autosperm) (Leonard and Lukowiak, 1987b; Monteiro et al. , 1987). This prediction was made on the basis of the gamete-trading model and analogy with Navanax (Leonard and Lukowiak, 1985, 1987b). This is a strong prediction because although copulation in Biom- phalaria glabrata (Say), at least, is typically unilateral and reciprocation can occur (Brenner, 1990), the details of the mating system are as yet unknown. That is, it remains to be shown that the mating system is based on reciprocation with courtship performed by the male, and a preference for the female role, as is required for sperm-trading. Sperm- oa) i) parcelling, in which only a small quantity of sperm (too lit- tle to fill the sperm storage organ) is transferred in a single copulation, is expected to form part of sperm-trading mating systems (Leonard and Lukowiak, 1984, 1985). M.E.B. Valadares-Ribeiro has obtained evidence from studies with genetic markers of sperm-parcelling in B. tenajophila (pers. comm. from W. Monteiro). Sperm-parcelling has yet to be demonstrated directly for Navanax, although there is some evidence that a single copulation as a female does not com- pletely replenish depleted sperm stores (Leonard, unpub. data, see also above). OTHER MATING SYSTEMS WITH UNILATERAL COPULATION To date, Navanax is the only species for which repeated alternation of sexual roles in a copulatory bout has been described. Further investigation could show that this is not uncommon in unilaterally copulating gastropods since very little is known about the copulatory behavior of most of them. However, it is clear that repeated alternation is not characteristic of some taxa, e.g. Lymnaea, Aplysia, etc. There are three types of mating interaction that have been described for these species: single non-reciprocal copulations, a single alternation of sexual roles, and chain copulation. LYMNAEA AND OTHER BASOMMATOPHORANS A single alternation of sexual roles commonly occurs in the lymnaeid basommatophorans, Lymnaea stagnalis (L.) (Noland and Carriker, 1946; Barraud, 1957; van Duiven- boden, 1984) and Stagnicola elodes (Say) (Rudolph, 1979a) and perhaps also in the stylommatophoran genus Partula (Lipton and Murray, 1979). In contrast, reciprocation does not occur commonly in the planorbid basommatophoran Bulinus globosus (Morelet) (Rudolph, 1979b). The mechanisms by which reciprocation is enforced (or cheating prevented) in cases of a single alternation of sexual roles have not been studied in detail. In S. elodes and L. stagnalis, male sexual behavior has been shown to be induced by copula- tion as a female (Rudolph, 1979a; v. Duivenboden and ter Maat, 1985) and this could also be the case in Navanax (Leonard and Lukowiak, 1991). In Lymnaea, courtship is a male behavior and individuals appear to be always willing to copulate as females (van Duivenboden and ter Maat, 1985), which suggests that the female sexual role is preferred. Also, copulatory plugs have been reported in S. elodes and B. globosus (Rudolph, 1979a, b). That of S. elodes probably prevents a second copulation as a female for two to three hours, while in B. globosus the copulatory plug is presumed to be ineffective in preventing a second copulation. It would be interesting to know more about the relationship of these plugs, apparently common in basommatophorans (Geraerts and Joosse, 1984), to the mating system. These instances of apparently unilateral and/or single AMER. MALAC. BULL. 9(1) (1991) reciprocal copulations offer exciting opportunities to test the Hermaphrodite’s Dilemma model, because they appear at first glance to contradict the predictions of the model. The first prediction of the model is that of reciprocity. Thus in order to be consistent with the model, the cases of unilateral copulation must actually represent part of a reciprocal in- teraction [which is not impossible, the single alternation of Lymnaea stagnalis could involve a period of hours between the two copulations during which the first male rides on the shell of the first female before she begins to reciprocate (Leonard, v. Duivenboden and ter Maat, unpub. data)] or the unilateral copulations must represent a form of ‘‘cheating’’ obviously derived from a reciprocal mating system. The cases of a single reciprocal copulation are also puzzling. Single reciprocal copulations would be consistent with the predictions of the Hermaphrodite’s Dilemma only under Game of Chicken conditions. That is, if there is a preferred role, an individual that has taken that role in the first copulation would have no reason to remain and reciprocate by assuming the less preferred role, unless there were a shortage of other mates available. Because many of the basommatophorans, such as Lymnaea and Physa, for ex- ample, typically occur in dense populations this seems unlike- ly. Such population densities suggest that Prisoner’s Dilem- ma conditions should be in operation but it is axiomatic (Axelrod and Hamilton, 1981) that under Prisoner’s Dilem- ma, reciprocation (such as Tit-for-Tat) can only evolve where the last move of the interaction is known. Otherwise, each player would benefit by refusing to reciprocate on the last move (see Leonard, 1990). Therefore, the Hermaphrodite’s Dilemma model predicts that single reciprocal copulations must involve as yet identified mechanisms for enforcing reciprocation. Specifically, where a gametolytic gland ex- ists, it should be the case that the male is able to prevent the female from leaving before reciprocating by assuming the male role with its former partner. The egg-trading model, on the other hand, would predict that individuals should com- pete for opportunities to copulate as males. CHAIN COPULATION Chain copulation consists of a mating interaction between three or more individuals in which the individual in front acts only as a female, while each middle individual acts both as a male (to the individual in front of it) and as a female (to the individual behind it) while the last individual acts only as a male. This phenomenon has been observed in a variety of euthyneuran gastropods, particularly basom- matophoran pulmonates and tectibranch opisthobranchs (see various planorbids, Precht, 1936; Duncan, 1975; Kuma, 1975; Geraerts and Joosse, 1984; Hadfield and Switzer- Dunlap, 1984; Franc, 1986). Chain (and even ring) forma- tion appears possible in virtually all species that normally copulate unilaterally, and some that usually copulate LEONARD: GASTROPOD MATING SYSTEMS 53 reciprocally (e.g. Phyllaplysia taylori Dall, Beeman, 1970a, b), and is particularly common under crowded laboratory con- ditions. In some taxa, such as Aplysia spp., however, chain copulation appears to occur commonly in the field, and must be regarded as a normal feature of the mating system (P. Fischer, 1869; MacGinitie and MacGinitie, 1968; Ricketts et al., 1968; Kupfermann and Carew, 1974; Leonard, unpub. data). Chain copulation has also been described from laboratory observations for Acera bullata Miller (Legendre, 1905). Geldiay (1956) concluded that chain copulation was the rule rather than the exception for Lake District populations of the freshwater limpet, Ancylus fluviatilis Miller, where chains of as many as seven individuals have been observed. Wesenberg-Lund (1939) reported for Lymnaea that chains of three individuals were not uncommon in the field (see also Crabb, 1927; Noland and Carriker, 1946; Barraud, 1957; v. Duivenboden, 1984) and that the female will next act as male to a nearby individual. There are also reports of simultaneous reciprocal copulation in Lymnaea (Klotz, 1889; Crabb, 1927). In other species, chain copulation is probably largely an artifact of laboratory conditions [e.g. Physa fontinalis (L.) (Duncan, 1959)] and rare, if it occurs at all, in the field. For example, in Navanax chains and/or rings of three or four copulating individuals occur commonly in the laboratory, but are very rare in the field (Leonard and Lukowiak, 1985). Rivest (1984) described group hypodermic copulation in two species of the nudibranch Palio, but this appears to be the exception, the rule being simultaneously reciprocal hypo- dermic copulation. The Hermaphrodite’s Dilemma model predicts that chain copulation should represent an obvious derivative from a reciprocal mating system. The data available in the literature are not adequate to confirm or refute this prediction and further observations are required before we can understand chain copulation as a mating system. In Lymnaea, mating in- teractions typically involve a single alternation of sexual roles between members of a pair (Noland and Carriker, 1946; Bar- raud, 1957; v. Duivenboden, 1984; v. Duivenboden and ter Maat, 1985; Leonard, v. Duivenboden and ter Maat, unpub. obs.) and it could be the case that chain copulations occur under conditions of high density and could represent ‘‘cheating’’ on a successively reciprocal system. If so, the gamete-trading model predicts that the ‘‘cheating’’ will con- sist of females avoiding male behavior, whereas the egg- trading model would predict that individuals should compete for opportunities to copulate as males. Another possibility is that mechanisms exist for reciprocation within the chain interaction. That is, individuals in chains could remain in the chain until they have copulated equally often in both roles. Some observations in both Lymnaea (Wesenberg-Lund, 1939) and Aplysia californica (Leonard and Lukowiak, 1983; Leonard, unpub. data) suggest that, as in Navanax (Leonard and Lukowiak, 1987a, 1991), individuals begin to act as males after acting as females. Also, laboratory observations indicate that chains of copulating A. californica can break and reform and individuals can copulate several times before mating ac- tivity ceases, with some indication that females (individuals at the front of the chain) tend to act as males to either the animal at the end of the chain or a nearby individual in the subsequent copulation (Leonard and Lukowiak, unpub. data). The data, however, are too scanty to allow us to tell whether individuals alternate sexual roles within chains. In Aplysia, courtship is initiated by the individual that will act as a male (Kupfermann and Carew, 1974; Leonard and Lukowiak, 1983) as is the case in Lymnaea and Navanax but there is as yet no clear evidence that the female is preferred as predicted by the gamete-trading model. THE STRANGE CASE OF ARIOLIMAX: SELF- MULTILATION? HERMAPHRODITES AS ‘“*CASTRATING FEMALES’’? Another intriguing observation is the report for a stylommatophoran slug, Ariolimax, that ‘‘they frequently gnaw off the penis at the close of copulation,’ (Mead 1943: 675). A certain percentage of large individuals in Ariolimax appear to lack completely a penis (Heath, 1916; Mead, 1943), whereas in others it is underdeveloped (Heath, 1916). Heath (1916), having hypothesized that the penis must be lost and then regenerated in this species, collected 200 individuals in an enclosure and after several weeks was able to observe two instances of copulation. He described the courtship process and stated that copulation was unilateral and that in both cases the penis was chewed off as soon as the animals began to draw apart. He indicated that in at least one case the am- putation was initiated by the female who was then joined in amputation by the “‘possessor of the intromittent organ con- cerned’’. Upon dissection Heath found that in two of the in- dividuals (the females in the copulations), the amputated penis extended from the genital pore internally to the distal end of the seminal receptacle. Heath found this phenomenon understandably perplexing and offered two possible explana- tions: 1) that the amputated penis serves as a sperm plug; 2) that the behavior is an artifact of disturbance by the observer and has evolved as a means of rapid separation when escape is necessary. Because the amputation process took over 10 min (Heath, 1916), it seems unlikely that it is an effective defense against predators but it could serve to prevent dessica- tion. Heath mentioned that copulation is nocturnal and that intromission had lasted several hours before the animals began to separate so it may be the case that copulations starting late in the night might create a risk of dessication in the morning sun unless there was a way of rapidly terminating them. Sex- ual conflict theory can add 1) the possibility that the func- tion of the sperm plug is more to prevent the mate from acting as a female again (keep other sperm out) than to prevent loss of sperm, and/or 2) the suggestion that in these her- 54 AMER. MALAC. BULL. 9(1) (1991) mmaphrodites an individual that amputated the penis of its mate could increase its own reproductive success as a male by decreasing the number of effective male rivals. The Hermaphrodite’s Dilemma model would predict that this behavior, if it is not merely a defense mechanism against danger of dessication or the like, must be a means of enforcing reciprocation or ‘‘cheating’’ on a reciprocal mating system. For example, if as predicted by the gamete- trading model, the female role is preferred it could be the case that once an individual has accumulated enough allosperm to fill its own sperm storage organ, it could not have any ‘‘reason’’ to mate as a male and could amputate its own penis, leaving it as a sperm plug to prevent its mate from receiving more sperm before egg-laying, thereby insur- ing paternity. It is barely conceivable that an individual could be able to regenerate its penis in time to get to use it in reciprocal mating interactions in order to obtain a new load of sperm after using the previous batch. In any case, the gamete-trading model predicts that Ariolimax which lack a penis should remain willing to copulate as females whereas the egg-trading model would predict that, because her- maphrodites should copulate as females in order to get an opportunity to copulate as males, that an individual lacking a penis ought to be unwilling to copulate as a female. Similar- ly, if the goal of copulating as a male is getting an oppor- tunity to be female, as is predicted for euthyneuran gastropods by the gamete-trading hypothesis, then Ariolimax should be reluctant to act as a male to an individual that lacks a penis. The egg-trading model on the other hand predicts that an in- dividual copulating as a male should not be fussy and should accept a mate with or without a penis of its own. Therefore, both gamete-trading and egg-trading predict that individuals lacking a penis will be unlikely to be involved in copulations but the egg-trading model predicts that that will be due to ‘‘coyness’’ of the amputee, whereas the gamete-trading model predicts that the individual lacking a penis will be unattrac- tive or rejected as a (female) mate. The copulation of these common banana slugs of the northwest coast of the United States seems to demand further attention. We need to know: 1) whether this amputation is a defensive response or whether it occurs as a normal part of the sexual behavior; 2) how commonly this amputation occurs; 3) who amputates the penis of whom; and 4) whether this amputation occurs after an individual’s first copulation or only in older individuals who may have mated with several partners. We also need to know how often these animals copulate over their life-span, and/or between egg-layings in order to understand the significance of this. Perhaps these slugs are effectively monogamous, at least as males, each in- dividual mating once upon attaining adulthood and losing its penis in the process, with occasional individuals surviving long enough to regenerate the penis and copulate as males a second time? Its amazing how little we know about such common and conspicuous animals. SIMULTANEOUSLY RECIPROCAL COPULATION STYLOMMATOPHORANS Both Hyman (1967) and Franc (1968) made the inter- esting generalization that basommatophorans have unilateral copulation associated with a short, simple courtship per- formed by the individual that will act as the male, whereas stylommatophorans have simultaneously reciprocal copula- tion preceded by lengthy, elaborate, and often bizarre court- ship behavior. The usual explanation of this phenomenon has been mechanistic; i.e. that the behavior serves to facilitate coordination between the partners to allow simultaneous reciprocal intromission, and most of the experimental work has focused on that aspect of the behavior (i.e. Helix, Jep- pesen, 1976; Lind, 1976; Chung, 1986; Adamo and Chase, 1988; Giusti and Andreini, 1988). However, because many opisthobranchs, particularly nudibranchs, have simultaneous reciprocal copulation without lengthy or notably peculiar courtship behavior (Hadfield and Switzer-Dunlap, 1984; Leonard, unpub. data; see also aeolids below), it is difficult to argue that simultaneous reciprocal copulation must be ac- companied in evolution by such bizarre mechanisms as the love-dart of Helix, the ingestion of caudal mucous globules, the aerial performance of Limax maximus (L.), etc. (see Hyman, 1967; Franc, 1968; Tompa, 1980). An obvious func- tional or adaptive explanation is that these elaborate court- ship behaviors have evolved through sexual conflict. Specifically, the Hermaphrodite’s Dilemma model pre- dicts that they all serve to prevent ‘‘cheating’’ on the recipro- cal mating system, which should take (according to the gamete-trading model) the form of individuals attempting to act as females, receiving sperm, without offering any of their own. Both Meisenheimer (1907) and Lind (1976) reported that in Helix, that an individual (A) that inserts its penis into the vagina of its partner (B) will immediately withdraw its penis unless B simultaneously inserts its (B’s) penis into A’s vagina (but see Chung, 1987). This is consistent with the idea that courtship serves to prevent an individual from acting only as a male. Similar reluctance to act as a male before the part- ner does should be seen in other species with elaborate court- ship and simultaneous reciprocal copulation. The courtship therefore, should enforce reciprocity, specifically by prevent- ing individuals from acting only as females. An egg-trading model, based on Bateman’s principle would predict the op- posite; that is the courtship serves to enforce reciprocity by preventing individuals from ‘‘cheating’’ by acting only in the male role. One would predict therefore that in these elaborate courtships there should be evidence that individuals are 1) coy as females, refusing to allow intromission until they have an opportunity to act as a male; and 2) eager as males, LEONARD: GASTROPOD MATING SYSTEMS oe, competing with each other for the first intromission. Charnov (1979) suggested that such ‘‘complicated precopulatory displays’ should serve (under Bateman’s principle) to induce the partner to use the sperm received to fertilize eggs. These hypotheses should be testable by experimental and com- parative studies of common species of stylommatophorans. AEOLIDS While Lind (1976) reported that unilateral copulations were rare in Helix, Rutowski (1983) found that 49% of all copulations in Hermissenda were non-reciprocal. Rutowski (1983) discussed this phenomenon in terms of sexual con- flict and Charnov’s (1979) prediction, considering the possibility that the failure of one individual to intromit after everting its penis was the result of an effort by its partner either to deflect the penis of its partner or to give sperm quick- ly without receiving any. That is, that one individual was at- tempting to ‘‘cheat’’ by mating only in the male role. He con- cluded that this was unlikely because many of the “‘missed individuals’’ were sperm-depleted and would have benefited from receiving sperm. If, as the gamete-trading model (Leonard and Lukowiak, 1984, 1985) predicts, the female sexual role is preferred, the “‘cheater’’ in a unilateral copulation would be the individual receiving but not giving sperm. These unilateral intromissions in Hermissenda could represent ‘‘cheating’’ by ‘‘deliberately’’ missing the target. However, since Rutowski (1983) reported that sperm was ejaculated into the water as a result of these ‘‘missed’’ intromissions this seems im- probable. It seems very unlikely that emission of sperm into the water would be more adaptive than transferring it to a partner, especially because Rutowski’s (1985) observation that Hermissenda ingest any sperm left on the gonopore suggests that the caloric content of sperm is not trivial. I agree, therefore, with Rutowski’s conclusion that the high frequen- cy (49%) of copulations in which only one individual achieves intromission is probably a consequence of whatever factors have selected for extremely rapid copulation in this species. However, on the assumption that the female sexual role is preferred in Hermissenda, I suggest that cheating was represented in Rutowski’s observations by those copulations (5% of the total) which were unilateral because only one in- dividual everted its penis, because in these cases individuals received sperm without giving any in return (or wasting any). A more detailed study of the mating behavior of Hermissenda as a function of the sperm stores of the interacting individuals might serve to test this possibility. In particular, Rutowski’s observation that most of the ‘‘missed individuals’’ in only semi-successful reciprocal copulation attempts were sperm- depleted is intriguing. One would like to know if sperm- depleted animals behave differently during mating encounters, and/or if their depleted status is detectable by partners who then treat them differently. One would expect that, if there is any difference, sperm-depleted individuals ought to be both more willing to receive sperm, and more attractive as female partners, than individuals with full sperm stores. Observations from another aeolid, Aeolidia papillosa Bergh, are also suggestive of cheating in a mating system based on reciprocation. The sexual behavior of Aeolidia is very similar to that of Hermissenda (Longley and Longley, 1982, 1984); encounters are very brief and usually simul- taneously reciprocal. However, in Aeolidia there is no copula- tion, sperm packets are deposited on the partner’s gonopore (Longley and Longley, 1984). The Longleys observed one individual which copulated repeatedly (over a period of days) without producing sperm packets. These authors also reported that the quantity of sperm transferred in a copulation was determined by both 1) the duration of the copulation, which is correlated with the size of the smaller partner, and 2) the rate of sperm transfer, which is related to the number of autosperm remaining in the ampulla. This raises the possibility that Aeolidia could engage in what Trivers (1971) termed ‘‘subtle cheating.’’ That is, an Aeolidia could cheat by engaging in a reciprocal mating when it has relatively few autosperms available, and thereby receive more sperm than it gives to its partner. Beeman (1970a) also observed instances in which only one member of a pair of reciprocally copulating Phyllaplysia taylori Dall trans- ferred sperm to its partner, since the other’s ampulla was empty. Observations on many species with simultaneous reciprocal copulation mention that unilateral copulations sometimes occur (e.g. Helix: Herzberg and Herzberg, 1962; Lind, 1976; opisthobranchs: Hadfield and Switzer-Dunlap, 1984; including Aeolidia papillosa: Longley and Longley, 1984; Hermissenda crassicornis: Longley and Longley, 1982; Rutowski, 1983; Melilbe: Agersborg, 1922) which could be considered ‘‘cheating.’’ In summary, there is some evidence that “‘cheating’’ can occur occasionally in species with simultaneous reciprocal copulation, in the form of unilateral copulations and/or ‘‘subtle cheating.’ However, from the available evidence one cannot say with confidence that ‘‘cheating’’ does or does not occur in these species. The evidence does suggest that studies directed to the analysis of sexual conflict in this group would be very rewarding. DISCUSSION The review and analysis presented here suggest that 1) gastropods offer a broad array of reproductive phenomena that require explanation in terms of mating systems theory; 2) Hermaphrodite’s Dilemma model makes nontrivial predic- tions about the mating systems of hermaphroditic gastropods that may serve to test the model; 3) for many euthyneuran gastropods the egg-trading and gamete-trading models make opposing predictions, making this group a useful means of 56 AMER. MALAC. BULL. 9(1) (1991) distinguishing between the two models. While experimental studies can be used to determine the preferred sexual role for a given species (Leonard and Lukowiak, 1991), com- parative studies can also be useful here. Review of the range of mating systems found in simultaneously hermaphroditic serranid fishes (Leonard, unpub. data) provides strong sup- port for the existence of sexual conflict since it provides con- firmation of the hypothesis that the male sexual role is preferred in these simultaneous hermaphrodites, as assumed by Fischer (1980, 1984) and predicted by Charnov (1979). The evidence for this is twofold. First, in all species studied, ‘‘cheating,’” on a reciprocal mating system, whether as streak- ing or as extra-pair spawning in the monogamous Serranus tigrinus (Bloch) (Pressley, 1981), is a male behavior; cheaters ‘‘cheat’’ in order to fertilize someone else’s eggs, not to get their own eggs fertilized. Second, where mates become a defensible resource, large, dominant individuals become male (e.g. S. fasciatus (Jenyns) and S. baldwini (Evermann and Marsh) (Petersen, 1990). In serranids, then, the harem- based mating systems are exceptions that prove the egg- trading rule. These simultaneously hermaphroditic fish pro- vide strong evidence that sexual conflict both exists and is important in shaping mating systems. However, both the gamete-trading and the egg-trading models predict (indeed the gamete-trading model assumes) that serranids prefer the male role so that the serranids do not allow us to distinguish between the egg-trading and gamete-trading models (contrary to Fischer, 1987). The gastropods therefore, offer an exciting opportunity not only to test the assumption that sexual con- flict exists but also to distinguish between models based on different assumptions about the source of sexual conflict. The variety of reproductive behavior and physiology found within the gastropods should allow us to identify species that can be used to test hypotheses about the relative importance of energy investment, mating time and control of fertilization in sexual conflict. In this paper I have attempted to demonstrate that analysis in terms of sexual conflict makes specific predic- tions about gastropod mating systems that may allow us to elucidate the adaptive significance of many bizarre phenomena in gastropod reproductive biology. The available literature on gastropod sexual behavior suggests a number of interesting test cases for the Hermaphrodite’s Dilemma model, but does not, in itself, provide sufficient data to test the model. The chief difficulty interpreting the available in- formation on gastropod reproduction in terms of sexual con- flict or sexual selection is that one can seldom determine from the available descriptions which types of behavior represent the rule and which the exceptions. Where the initial studies were not informed by mating systems theory or selection thinking, crucial information is apt to be lacking, even when there have been numerous detailed studies of the behavior, as in Helix (Leonard, unpub. data). Mating systems theory has a lot to offer to the study of euthyneuran gastropods and euthyneuran gastropods have a lot to offer to the study of mating systems theory. ACKNOWLEDGMENTS The author was supported by grants from the NIMH and NSF during part of the period of manuscript preparation. LITERATURE CITED Adamo, S. A. and R. Chase. 1988. Courtship and copulation in the terrestrial snail Helix aspersa. Canadian Journal of Zoology 66:1446-1453. Agersborg, H. P. v. W. Kjerschow. 1921. Contribution to the knowledge of the nudibranchiate mollusk, Melibe leonina (Gould). American Naturalist 55:222-253. Alexander, R. D. and G. Borgia. 1979. On the origin and basis of the male- female phenomenon. In: Sexual selection and reproductive competi- tion in insects. M. S. Blum and N. A. Blum, eds. pp. 417-440. Academic Press, Inc., New York. Axelrod, R. and W. D. Hamilton. 1981. The evolution of cooperation. Science 211:1390-1396. Barraud, E. M. 1957. The copulatory behavior of the freshwater snail (Lymnaea stagnalis L.). British Journal of Animal Behaviour 5:55-59. Bateman, A. J. 1948. Intra-sexual selection in Drosophila. Heredity 2:349-368. Baylis, J. R. 1981. The evolution of parental care in fishes, with reference to Darwin’s rule of male sexual selection. Environmental Biology of Fishes 6:223-251. Beeman, R. D. 1970a. The anatomy and functional morphology of the reproductive system in the opisthobranch mollusk, Phyllaplysia taylori Dall, 1900. Veliger 13:1-31. Beeman, R. D. 1970b. An ecological study of Phyllaplysia taylori Dall, 1900 Gastropoda: Opisthobranchia with an emphasis on its reproduction. Vie et Milieu Sec. A 21:189-211. Berglund, A. 1986. Sex change by a polychaete: Effects of social and reproductive costs. Ecology 67:837-845. Boray, J. C. 1964. Studies on the ecology of Lymnaea tomentosa, the in- termediate host of Fasciola hepatica. Il. The sexual behaviour of Lymnaea tomentosa. Australian Journal of Zoology 12:231-237. Brenner, V. A. 1990. The effect of isolation on the reproductive behavior and fecundity of Biomphalaria glabrata (Mollusca: Pulmonata). Undergraduate Honors Thesis, Department of Zoology, University of Oklahoma, Norman. 11 pp. Bronstein, J. L. 1988. Limits to fruit production in a monoecious fig: con- sequences of an obligate mutualism. Ecology 69:207-214. Charnov, E. L. 1979. Simultaneous hermaphroditism and sexual selection. Proceedings of the National Academy of Sciences, U.S.A. 76:2480-2484. Charnov, E. L. 1982. The Theory of Sex Allocation. Princeton University, Press, Princeton, New Jersey. 355 pp. Charnov, E. L. 1986. Size advantage may not always favor sex change. Journal of Theoretical Biology 119:283-285. Chung, D. J. D. 1986. Stimulation of genital eversion in the land snail Helix aspersa by extracts of the glands of the dart apparatus. Journal of Experimental Zoology 238:129-139. Chung, D. J. D. 1987. Courtship and dart shooting behavior of the land snail Helix aspersa. Veliger 30:24-29. Crabb, E. D. 1927. The fertilization process in the snail Lymnaea stagnalis appressa Say. Biological Bulletin 53:67-109. Darwin, C. 1874. The Descent of Man and Selection in Relation to Sex (2nd ed.). Rand McNally and Co., Chicago. 672 pp. LEONARD: GASTROPOD MATING SYSTEMS =| Davis, M. D. 1983. Game Theory. Revised edition. Basic Book, Inc., New York. 250 pp. Duncan, C. J. 1959. The life cycle and ecology of the freshwater snail Physa fontinalis (L.). Journal of Animal Ecology 28:97-117. Duncan, C. J. 1975. Reproduction. Jn: Pulmonates, Vol. 1. Functional Anatomy and Physiology. V. Fretter and J. Peake, eds. pp. 309-365. Academic Press, London. Duivenboden, Y. A. van. 1983. Transfer of semen accelerates the onset of egg-laying in female copulants of the hermaphrodite freshwater snail Lymnaea stagnalis. International Journal of Invertebrate Reproduc- tion 6:249-257. Duivenboden, Y. A. van. 1984. Sexual behavior of the hermaphrodite freshwater snail Lymnaea stagnalis. Doctoral Dissertation. Vrije Universiteit, Amsterdam. 109 pp. Duivenboden, Y. A. van and A. ter Maat. 1985. Masculinity and receptiv- ity in the hermaphrodite snail, Lymnaea stagnalis. Animal Behaviour 33:885-891. Eberhard, W. G. 1985. Sexual Selection and Animal Genitalia. Harvard University Press, Cambridge, Massachusetts. 244 pp. Emlen, S. T. and L. W. Oring. 1977. Ecology, sexual selection, and the evolu- tion of mating systems. Science 197:215-223. Fischer, E. A. 1980. The relationship between mating system and simultaneous hermaphroditism in the coral reef fish, Hypoplectrus nigricans (Serranidae). Animal Behaviour 28:620-633. Fischer, E. A. 1984. Egg trading in the chalk bass, Serranus tortugarum, a simultaneous hermaphrodite. Zeitschrift fir Tierpsychologie 66:143-151. Fischer, E. A. 1987. Mating behavior in the black hamlet - gamete trading or egg trading? Environmental Biology of Fishes 18:143-148. Fischer, E. A. and C. W. Petersen. 1987. The evolution of sexual patterns in the seabasses. BioScience 37:482-489. Fischer, P. 1869. Sur l’accouplement et la ponte des Aplysiens. Comptes Rendus de | ‘Academie des Sciences, Paris 69:1095-1097. Fisher, A. C. 1981. Resource and Environmental Economics. Cambridge University Press, Cambridge, England. 284 pp. Fisher, R. A. 1958. The Genetical Theory of Natural Selection. Second re- vised edition. Dover Publications, Inc., New York. 291 pp. Franc, A. 1968. Class des Gasteropodes (Gastropoda Cuvier, 1798). In: Traite de Zoologie Tome V, Fasc. 3. P.-P. Grasse, ed. pp. 1-925. Masson et Cie., Paris. Galen, C. and Rotenberry, J. T. 1988. Variance in pollen carryover in animal- pollinated plants: Implications for mate choice. Journal of Theoretical Biology. 135:419-429. Geldiay, R. 1956. Studies on local populations of the freshwater limpet Ancylus fluviatilis Miller. Journal of Animal Ecology 25:389-402. Geraerts, W. P. M. and J. Joosse. 1984. Freshwater Snails (Basommatophora). In: The Mollusca, Vol. 7, Reproduction. A. S. Tompa, N. H. Ver- donk and J. A. M. van der Biggelaar, eds. pp. 141-207. Academic Press, Orlando, Florida. Gerhardt, U. 1933. Zur Kopulation der Limaciden. I. Mitteilung. Zeitschrift fiir Morphologie und Okologie der Tiere 27:401-450. Ghiselin, M. T. 1965. Reproductive function and the phylogeny of opisthobranch gastropods. Malacologia 3:329-378. Ghiselin, M. T. 1974. The Economy of Nature and the Evolution of Sex. University of California Press, Berkeley. 346 pp. Gilbertson, L. H. 1989. A new species of Holospira (Gastropoda: Pulmonata) from Arizona, with the reproductive anatomies of H. arizonensis and H. chiricahuana. Veliger 32:308-312. Gillespie, J. H. 1977. Natural selection for variances in offspring numbers: a new evolutionary principle. American Naturalist 111:1010-1014. Giusti, F. and S. Andreini. 1988. Morphological and ethological aspects of mating in two species of the family Helicidae (Gastropoda Pulmonata): Theba pisana (Mueller) and Helix aperta Born. Monitore Zoologica Italiano (N.S.) 22:331-363. Gosliner, T. M. 1980. Systematics and phylogeny of the Aglajidae (Opisthobranchia: Mollusca). Zoological Journal of the Linnean Society 68:325-360. Hadfield, M. G. and M. Switzer-Dunlap. 1984. Opisthobranchs. In: The Mollusca, Vol. 7, Reproduction. A. S. Tompa, N. H. Verdonk and J. A. M. van der Biggelaar, eds. pp. 209-350. Academic Press, Orlando, Florida. Hammerstein, P. and G. A. Parker. 1987. Sexual selection: games between the sexes. In: Sexual Selection: Testing the Alternatives. J. W. Brad- bury and M. B.Andersson, eds. pp. 119-142. John Wiley and Sons. Harris, L. G. 1975. Studies on the life history of two coral-eating nudibranchs of the genus Phestilla. Biological Bulletin 149:539-550. Heath, H. 1916. The conjugation of Ariolimax californicus. Nautilus 30:22-24. Herzberg, F. and A. Herzberg. 1962. Observations on reproduction in Helix aspersa. American Midland Naturalist 68:297-306. Hoffmann, R. J. 1983. The mating system of the terrestrial slug Deroceras laeve. Evolution 37:423-425. Horstmann, J. -J. 1955. Untersuchungen zur Physiologie der Begattung und Befruchtung der Schlammschnecke Lymnaea stagnalis L. Zeitschrift fur die Morphologie und Okologie der Tiere. 44:222-268. Hyman, L. H. 1967. The Invertebrates. Vol. 6. Mollusca I. McGraw-Hill Book Co., New York. 792 pp. Jeppesen, L. L. 1976. The control of mating behaviour in Helix pomatia L. (Gastropoda: Pulmonata). Animal Behaviour 24:275-290. Klotz, J. 1889 . Beitraege zur Entwicklungsgeschichte und Anatomie des Geschlechtsapparate von Lymnaeus. Jenaische Zeitschrift fur Natur- wissenschaften 23:1-40. Kuma, E. 1975. Studies on the behaviour of Bulinus (Physopsis) globosus (Morelet). Zoologischer Anzeiger 194:6-12. Kupfermann, I. and T. J. Carew. 1974. Behavior patterns of Aplysia cali- fornica in its natural environment. Behavioral Biology 12:317-337. Lalli, C. M. and F. E. Wells. 1978. Reproduction in the genus Limacina (Opisthobranchia: Thecosomata). Journal of Zoology 186:95-108. Legendre, R. 1905. Notes biologiques sur Acera bullata Mull. Archives de zoologie experimentelle et generale, Ser. 4, Vol. 4:vi-xiv. Leonard, J. L. 1990. The Hermaphrodite’s Dilemma. Journal of Theoretical Biology 147:361-372. Leonard, J. L. and K. Lukowiak. 1983. The behavior of Aplysia californica: Ethogram and ethological analysis of mating. Society for Neuroscience Abstracts 9:534. Leonard, J. L. and K. Lukowiak. 1984. Male-female conflict in a simultaneous hermaphrodite resolved by sperm-trading. American Naturalist 124:282-286. Leonard, J. L. and K. Lukowiak. 1985. Courtship, copulation and sperm trading in the sea slug, Navanax inermis (Opisthobranchia: Cephalaspidea). Canadian Journal of Zoology 63:2719-2729. Leonard, J. L. and K. Lukowiak. 1987a. Ethological mechanisms of sexual role alternation in Navanax inermis (Opisthobranchia). In: Neurobiology, Molluscan Models. H. H. Boer, W. P. M. Geraerts and J. Joosse, eds. pp. 242-247. Monographie Koninklijke Nederlandse Akademic. van Wetenschappen, North Holland Publ. Co., Amsterdam. Leonard, J. L. and K. Lukowiak. 1987b. Sperm-sharing in Biomphalaria snails. Nature 325:737-738. Leonard, J. L. and K. Lukowiak. 1991. Sex and the simultaneous her- maphrodite: testing models of male-female conflict in a sea slug. Animal Behaviour 41:255-266. Lind, H. 1976. Casual and functional organization of the mating behaviour sequence in Helix pomatia (Pulmonata, Gastropoda). Behaviour 59:162-202. Lind, H. 1988. The behaviour of Helix pomatia L. (Gastropoda, Pulmonata) in a natural habitat. Videnskabelige Meddelelser fra Dansk Naturhistorisk Forening. 147:67-92. Lipton, C. S. and J. Murray. 1979. Courtship of land snails of the genus Partula. Malacologia 19:129-146. 58 AMER. MALAC. BULL. 9(1) (1991) Longley, R. D. and A. J. Longley. 1982. Hermissenda: Agonistic behavior or mating behavior? Veliger 24:230-231. Longley, A. J. and R. D. Longley. 1984. Mating in gastropod mollusk Aeolidia papillosa: behavior and anatomy. Canadian Journal of Zoology 62:8-14. Luce, R. D. and H. Raiffa. 1957. Games and Decisions. John Wiley and Sons, New York. 509 pp. MacGinitie, G. E. 1934. The egg-laying activities of the sea hare, Tethys californicus (Cooper). Biological Bulletin 67:300-303. MacGinitie, G. E. and N. MacGinitie. 1968. Natural History of Marine Animals (2nd ed.). McGraw-Hill Book Company, New York. 523 pp. Mead, A. R. 1943. Revision of the giant West Coast land slugs of the genus Ariolimax Moerch (Pulmonata: Arionidae). American Midland Naturalist 30:675-717. Meisenheimer, J. 1907. Biologie, Morphologie und Physiologie des Begat- tungsvorgangs und der Eiablage von Helix pomatia. Zoologische Jahrbicher, Abteilung fur Systematik 25:461-502. Monteiro, W., J. M. G. Almeida, Jr. and B. S. Dias. 1984. Sperm-sharing in Biomphalaria snails: a new behavioral strategy in simultaneous hermaphroditism. Nature 308:727-279. Monteiro, W., J. M. G. Almeida, Jr. and B. S. Dias. 1987. Monteiro et al. reply. Nature 325:737-738. Mulvey, M. and R. C. Vrijenhoek. 1981. Multiple paternity in the her- maphroditic snail, Biomphalaria obstructa. Journal of Heredity 72:308-312. Murray, J. 1964. Multiple mating and effective population size in Cepaea nemoralis. Evolution 18:283-291. Nakamura, R. R., M. L. Stanton and S. J. Mazer. 1989. Effects of mate size and mate number on male reproductive success in plants. Ecology 70:71-76. Nicholson, W. 1978. Microeconomic Theory. Dryden Press, Hinsdale, Illinois 768 pp. Noland, L. E. and M. R. Carriker. 1946. Observations on the biology of the snail Lymnaea stagnalis appressa during twenty generations in laboratory culture. American Midland Naturalist 36:467-493. Orians, G. H. 1969. On the evolution of mating systems in birds and animals. American Naturalist 103:589-603. Parker, G. A. 1979. Sexual selection and sexual conflict. /n: Sexual Selec- tion and Reproductive Competition in Insects. M. S. Blum and N. A. Blum, eds. pp. 123-166. Academic Press, New York. Patterson, C. M. 1989. Morphological studies of a Tahitian succineid, Suc- cinea (Kondosuccinea) wallisi. Malacological Review 22:17-23. Pelseneer, P. 1895. Hermaphroditism in Mollusca. Quarterly Journal of Microscopical Science 37:19-46. Petersen, C. W. 1990. The relationships among population density, individual size, mating tactics, and reproductive success in a hermaphroditic fish, Serranus fasciatus. Behaviour 113:57-80. Philippi, T. and J. Seger. 1989. Hedging one’s evolutionary bets, revisited. Trends in Ecology and Evolution 4:41-44. Precht, H. 1936. Zur Kopulation und Eiablage einiger Planorbiden. Zoologischer Anzeiger 115:80-89. Pressley, P. H. 1981. Pair formation and joint territoriality in a simultaneous hermaphrodite: the coral reef fish Serranus tigrinus. Zeitschrift fur Tierpsychologie 56:33-46. Pruvot-Fol, A. 1961. Les organes genitaux des opisthobranches. Archives de Zoologie experimentelle et general. 99:185-224. Reeder, R. L. 1986. A new species of Helminthoglypta (Gastropoda: Pulmonata: Helminthoglyptidae) from the Cuyamaca mountains of Southern California. Veliger 28:444-447. Reinboth, R. 1962. Morphologische and funktionelle Zweigeschlechtlichkeit bei marinen Telostieren (Serranidae, Sparidae, Centracanthidae, Labridae). Zoologische Jahrbicher, Physiologie 69:405-480. Ricketts, E. F. and J. Calvin, revised by J. W. Hedgpeth. 1968. Between Pacific Tides. Stanford University Press, Stanford, California. 614 pp. Riechert, S. E. and P. Hammerstein. 1983. Game theory in the ecological context. Annual Review of Ecology and Systematics 14:377-409. Rivest, B. R. 1984. Copulation by hypodermic injection in the nudibranchs Palio zosterae and P. dubia (Gastropoda, Opisthobranchia). Biological Bulletin 167:543-554. Rose, R. A. and I. Hoegh-Guldberg. 1982. A brood-protecting nudibranch with pelagic lecithotrophic development. Journal of Molluscan Studies 48:231-232. Rudman, W. B. 1974. A comparison of Chelidonura, Navanax and Aglaja with other genera of the Aglajidae (Opisthobranchia: Gastropoda). Zoological Journal of the Linnean Society 54:185-212. Rudman, W. B. 1981. Further studies on the anatomy and ecology of opisthobranch molluscs feeding on the scleractinian coral Porites. Zoological Journal of the Linnean Society 71:373-412. Rudolph, P. H. 1979a. The strategy of copulation of Stagnicola elodes (Say) (Basommatophora: Lymnaeidae). Malacologia 18:381-389. Rudolph, P. H. 1979b. An analysis of copulation in Bulinus (Physopsis) globosus (Gastropoda: Planorbidae). Malacologia 19:147-155. Rudolph, P. H. and J. B. Bailey. 1985. Copulation as females and use of allosperm in the freshwater snail genus Bulinus (Gastropoda: Planor- bidae). Journal of Molluscan Studies 51:267-295. Rutowski, R. L. 1983. Mating and egg mass production in the aeolid nudibranch Hermissenda crassicornis (Gastropoda: Opisthobranchia). Biological Bulletin 165:296-285. Sella, G. 1985. Reciprocal egg trading and brood care in a hermaphroditic polychaete worm. Animal Behaviour 33:938-944. Sella, G. 1988. Reciprocation, reproductive success and safeguards against cheating in a hermaphroditic polychaete worm, Ophryotrocha diadema Akesson, 1976. Biological Bulletin 175:212-217. Shapiro, D. Y. and R. H. Boulon, Jr. 1982. The influence of females on the initiation of female-to-male sex changes in a coral reef fish. Hor- mones and Behavior 16:66-75. Sutherland, W. J. 1986. Measures of sexual selection. Oxford Surveys in Evolu- tionary Biology (1985) 2:90-91. Tompa, A. S. 1980. The ultrastructure and mineralogy of the dart from Philomycus carolinianus (Pulmonata: Gastropoda) with a brief survey of the occurrence of darts in land snails. Veliger 23:35-42. Tompa, A. S. 1984. Land Snails (Stylommatophora). Jn: The Mollusca, Vol. 7, Reproduction. A. S. Tompa, N. H. Verdonk and J. A. M. van der Biggelaar, eds. pp. 47-140. Academic Press, Orlando, Florida. Tompa, A. S., N. H. Verdonk and J. A. M. van den Biggelaar (eds.). 1984. In: The Mollusca, Vol. 7, Reproduction. Academic Press, Orlando, Florida. 486 pp. Trivers, R. L. 1971. The evolution of reciprocal altruism. Quarterly Review of Biology 46:35-37. Trivers, R. L. 1972. Parental investment and sexual selection. Jn: Sexual Selection and the Descent of Man. B. Campbell, ed. pp. 136-176. Aldine Publishing Company, Chicago. Warner, R. R. 1982. Mating systems, sex change and sexual demography in the rainbow wrasse, Thalassoma lucasanum. Copeia 1982:653-661. Warner, R. R. and P. Lejeune. 1985. Sex change limited by paternal care: a test using four Mediterranean labrid fishes, genus Symphodus. Marine Biology 87:89-99. Wawra, E. 1988. Strubellia paradoxa (Strubell, 1892) (Gastropoda: Opisthobranchia) von den Salomon-Inseln. Zoologischer Anzeiger 220:163-172. Webb, G. R. 1948. The mating of Stenotrema fraternum (Say). Nautilus 62:8-12. Wesenberg-Lund, C. 1939. Biologie der Siisswassertiere, Wirbellose Tiere. Verlag Julius Springer, Vienna. 817 pp. Williams, G. C. 1975. Sex and Evolution. Princeton University Press, Princeton, New Jersey. 200 pp. Willson, M. F. and N. Burley. 1983. Mate Choice in Plants. Princeton Uni- versity Press, Princeton, New Jersey. 244 pp. Date of manuscript acceptance: 16 November 1990 Substratum associations of natural populations of Iceland Scallops, Chlamys islandica Muller 1776, on the northeastern Grand Bank of Newfoundland Kent D. Gilkinson' and Jean- Marc Gagnon? 'LeDrew, Fudge and Associates Limited, P. O. Box 9370 Sta. B, St. John’s, Newfoundland, Canada AlA 2Y3 2Department of Biology, Memorial University of Newfoundland, St. John’s, Newfoundland, Canada AIB 3X9 Abstract. Live, Iceland scallops, Chlamys islandica, were enumerated and their occurrences assigned to substratum coarseness grades along five photographic transects (8-10 km in length) covering areas of the northeastern Grand Bank of Newfoundland. Scallops were disproportionately (53-94%) associated with the coarsest grade of substratum comprising dense gravel-cobble (80-100% by area). Overall, scallops were uncommon to rare on predominantly sand substrata. Average densities of scallops per photograph (5.4 m?) ranged from 0.5 to 13.8 in cobble fields and from 0.02 to 1.7 on open sand. It is hypothesized that Iceland scallops on the northeastern Grand Bank are aggregated on coarse substrata because of a strong propensity towards byssal attachment at all post-larval life history stages. A survey of substratum associations of extant species of Chlamys reveals that, with few exceptions, association with coarse substrata is common within the genus. The Iceland scallop Chlamys islandica Muller, 1776, is a subarctic-boreal species, extending from Hudson Strait, N.W.T., south to the Massachusetts region (Lubinsky, 1980). Its bathymetric range extends to about 180 m and over the expansive Newfoundland continental shelf, it can occur at commercial densities (Naidu and Cahill, 1989). Apart from mostly anecdotal accounts, very little is known about sub- stratum associations of natural populations of this species and the family Pectinidae in general. To date, research on sub- strata associations has focused on the settlement and growth of pectinid spat on artificial collectors within an overall aquaculture context. While it is generally known that pectinid spat will settle on substrata such as algae, hydrozoans, bryo- zoans and various artificial surfaces (Fraser, 1983), the natural settlement substrata for C. islandica are unknown (Wallace, 1982). Subsequent to spat settlement, it is believed that juvenile scallops display an overall movement from primary settlement substrata to substrata where they will reside dur- ing juvenile and adult stages. In the case of C. islandica, this could take place about one year after settlement when they have attained a shell height of >5 mm (Wallace, 1982). Based on the contents of scallop dredges, Naidu (1988) found that Newfoundland populations of Iceland scallops are found normally at depths greater than 55 m, usually on hard bottom of variable substratum composition including mix- tures of sand, gravel, shell fragments, rocks and boulders. Dense, commercial concentrations of Iceland scallops are known to occur on St. Pierre Bank (Newfoundland Grand Banks) (Naidu and Cahill, 1989) in areas with sediments characterized as Sable Is. gravel, which is a mixture of gravel and < 10% sand (Fader et al., 1982). Most of the specimens of North American continental shelf Chlamys islandica in the collection of the National Marine Fisheries Service (Woods Hole) were collected from coarse substrata comprising gravel, sand-gravel, till and sand (Theroux and Wigley, 1983). Wiborg (1962, cited in Vahl and Clausen 1980) stated that C. islandica lives on coarse sediments or on hard bottom. In a shallow water study in the Gulf of St. Lawrence, Jalbert et al. (1989) determined that C. islandica occurred most fre- quently on coarse substrata, primarily cobble and gravel. The objective of this investigation was to identify natural substratum associations in offshore populations of Iceland scallops from the Newfoundland Grand Bank. Based on a series of photographic transects across areas of the north- eastern Grand Bank, spatial distribution patterns of the mega- fauna and major substratum types (coarseness grades) were described (Schneider et al., 1987). The scope of this paper is threefold. First, Iceland scallop-substrata associations are documented for the northeastern Grand Bank. This is followed by an assessment of physical and behavioural mechanisms which could potentially influence these associations. Final- ly, observed patterns of substrata association in Chlamys islandica are compared with existing information on substrata and habitats occupied by extant Chlamys spp. STUDY AREA The Grand Banks of Newfoundland represents the most extensive shallow (< 200 m) feature on the continental American Malacological Bulletin, Vol. 9(1) (1991):59-67 Bb) 60 AMER. MALAC. BULL. 9(1) (1991) sheif of eastern Canada and is comprised of a series of ma- jor banks: St. Pierre, Green, Whale and Grand. The study area is situated on the northeastern edge of the Grand Bank of Newfoundland in the vicinity of the Hibernia oilfield (Fig. 1). Water depths range from 70 to 100 m. The most complete description of this area in terms of the surficial sediments and seabed processes is found in Barrie et al. (1984). Sedimentary cover typically consists of reworked sand and gravel deposits, generally < 2 m thick. Bedforms include sand ripples (< 1 m wide), sand megaripples (5-10 m wide), sand ribbons (100 m to 1 km wide) and sand ridges (general- ly > 3 km in width). Coarse sediment is incorporated into the sand, in places. Coarse sediment consists of gravel and cobble-sized clasts (Grand Bank Gravel) which are believed to represent Pleistocene glacial deposits which have been reworked subsequently into coastal environments by an early Holocene marine transgression as a result of a eustatic rise in sea level. Overall, sedimentary bedforms on the Grand Banks are believed to be dynamic and are being reworked by unidirectional storm-driven currents, ocean currents and extreme waves. Meigs C] Sand ridge field e i ees Grand Bank gravel . - Boundary sand Pad Lag gravel and sand ribbons Prapeettred CT Unsurveyed area «] contours in meters a Tg scale (km) yo Fig. 1. Location of photographic transects on the northeastern Grand Bank. METHODS Five photographic transects, ranging in length from 8 to 10 km, were conducted across the northeastern edge of the Grand Bank (Fig. 1). Colour 35 mm slide photographs were taken at 10s intervals with the BRUTIV system (Vilks, 1984) which consists of a sled-mounted camera (aligned ver- tically) towed 3 m above the seabed. Technical details are provided in Schneider et al. (1987). The area defined by each photograph was 5.4 m*. Each frame (slide) was examined under low magnifica- tion (X16) and classified into one of six substratum categories based on a gross classification scheme of sediment texture and estimated areal coverage: l- pure sand; 2- sand with scattered cobble; 3- scattered gravel on sand; 4- sand with cobble and shell: 5- 50 to 80% cobble and gravel; 6- > 80% cobble and gravel. Sand and gravel were distinguished on the basis of textural differences. While gravel is defined as particles with diameter > 5mm (ASTM, 1988), much of the sediment in categories 5 and 6 were dominated by large-sized rocks (> 20 mm diameter) and are referred to as cobble. All identifiable Iceland scallops were enumerated by superimposing a 5 X 5 grid over each slide and recording the number of organisms in each cell. Analyses were car- ried out on the sum of counts in each slide. Megafauna less than about 2 cm in shell height were visible, but generally could not be identified to the species level. Scallops were classified as live if they were in normal life position (epi- faunal) with normal colouration (i.e. not bleached). While there could be a tendency to overestimate the abundance of live scallops by inclusion of articulated, collapsed cluckers (sensu Naidu, 1988), there were probably instances where live, ‘“bleached-looking’’ scallops were classified as dead. All identifiable scallops displayed a sculpture of coarse ribs and had unequal hinge ‘‘ears,’’ features which distinguish the Iceland scallop from the sea scallop Placopecten magellanicus (Gmelin, 1791). The only scallop reported from the northeast- ern Grand Bank has been the Iceland scallop (Mobil, 1985). For each transect, the frequency distributions of live scallops, by substratum type, were tested statistically for significant departures from random distributions based on the proportion of available substratum types (G-test for goodness of fit, Sokal and Rohlf, 1981). It was assumed that scallops would be distributed randomly at similar densities for all substratum types. RESULTS The bedforms which the photographic transects crossed are shown in figure 1. Transect Northwest (NW), which over most of its length traversed previously unsurveyed territory, crossed a sand ridge field with areas of scattered cobble and gravel. Transect Northeast (NE) intersected two sand ridge fields with minor amounts of cobble, gravel and shell. Transect East (E) crossed sand ribbons developed on top of a lag gravel. Transect Southwest (SW) crossed a gravel and cobble field (Grand Bank gravel) with a regular alterna- tion of sand and cobble and Transect Southeast (SE) crossed a gravel and cobble field (Grand Bank gravel) which in- tersected a sand field (boundary sand) at its eastern end. Substratum composition varied between transects (Fig. GILKINSON AND GAGNON: SCALLOP SUBSTRATUM ASSOCIATIONS 61 LIVE SCALLOPS SUBSTRATUM 100 NW 100 NW TRANSECT TRANSECT (9) (744) 50 50 123 4 5 6 123 4 5 6 100 NE 100 NE TRANSECT TRANSECT (516) (753) 50 50 123456 123456 » 100 E 100 E Z TRANSECT TRANSECT 189 732 S 50 ee 50 oS WwW oc WL io 0 1234656 123456 100 Sw 100 Sw TRANSECT TRANSECT (4041) (723) 50 50 ° 123 4 5 6 0 123 45 6 100 SE 100 SE TRANSECT TRANSECT (2462) (539) 50 50 12345 6 123 4 5 6 FINE === COARSE FINE == COARSE SUBSTRATUM CATEGORY Fig. 2. Frequencies of occurrence of (i) live Iceland scallops by substratum category and (ii) substratum categories along photographic transects. Numbers in parentheses refer to total number of enumerated live scallops (left col- umn) and photographs examined (right column) along each transect. 2). Transects E and SE had a high percentage (50-60% ) oc- currence of predominantly coarse substrata, comprised of gravel and cobble (some shell), whereas Transects NE and NW had high occurrences of primarily sand substrata (> 60% in both cases). Along Transect SW there was more or less equal representation of fine and coarse substrata. Along those transects which included the two coarsest substratum categories, scallops were aggregated on coarse substrata (i.e. gravel and cobble) (Fig. 2). Along Transects NE, E and SE, greater than 80% of the total number of live scallops occurred on the coarsest substratrum (category #6). This was in spite of the fact that predominantly fine substrata occurred at frequencies ranging from 15 to 80% along these transects. In particular, along Transect NE where the pure sand substratrum category (#1) had a frequency of occurrence of 65%, approximately 85% of the scallops occurred on cob- ble substrata which had a frequency of occurence of only 10%. Along Transect SW, approximately 10% of the scallops were counted from substrata comprised predominantly of sand and scattered cobble (Fig. 2). It is noted that this occurred along the transect with the highest densities of live scallops (Fig. 3). Frequencies of scallops along these transects showed a highly significant deviation from a random distribution across all substratum types (G-test statistic range: 176-2869, p< <001). Transect NW crossed a sand ridge field with areas of scattered gravel. Along this entire 10 km transect, represented by 744 photographs, only 9 live scallops were counted. There was considerable variation in the substratum- specific densities of live scallops between transects (Fig. 3). Average densities on the coarsest substratum (> 80% cob- ble and gravel) ranged from 0.5 to 13.8 scallops per photograph (0.09 to 2.5 scallops/m?) while average densities on the next coarsest substratum (50% to 80% cobble and gravel) ranged from 0.1 to 5.7 scallops per photograph (0.02 to 1 scallop/m?). The highest densities on fine substrata occurred on Transect SW with an average density of 1.7 scallops per photograph (0.3 scallops/m?) on a substratum consisting primarily of open sand with scattered cobble. It was noted previously that this transect displayed the highest densities of scallops on most substratum categories. On the remaining transects, densities ranged from 0.02 to 0.4 scallops per photograph (0.003 to 0.07 scallops/m?) on the two finest substratum categories. DISCUSSION It is obvious that there is a strong association between Iceland scallops and coarse substrata (gravel, cobble) on the northeastern Grand Bank. At the outset, there are several ex- planations which could account for the observed aggregated distribution of scallops on coarse substrata. These include (1) substratum-specific predation pressure and (2) behavioural and physical mechanisms maintaining scallop-substratum associations. PREDATORS The underlying premise of substratum-specific preda- tion pressure is that scallops suffer heavy mortality from predators after movement of juveniles or adults onto fine 62 AMER. MALAC. BULL. 9(1) (1991) LIVE SCALLOPS (avg. number/frame) SUBSTRATUM CATEGORY []NE KNE Fig. 3. Average densities of live Iceland scallops by substratum type along photographic transects. substrata. As mentioned previously, evidence to date on pec- tinid spat substratum preference rules out primary settlement on fine substrata (1.e. sand). While disproportionate preda- tion pressure due to predator-substrata associations cannot be ruled out as a contributing factor, we believe that predators on the northeastern Grand Bank are not responsible for the skewed distribution of scallops on coarse vs. fine substrata. Potential predators of Iceland scallops on the Grand Banks are listed in Table 1. Of these, only the pleuronectids, American plaice (Hippoglossoides platessoides Fabricius, 1780) and yellowtail flounder (Limanda ferruginea Storer, 1839), are known predators of Iceland scallops on the Grand Banks (Pitt, 1976; Naidu and Meron, 1986) and other con- tinental shelf regions (Langton and Bowman, 1981). Naidu and Meron (1986) determined that Iceland scallops occurred in plaice stomachs with a frequency of 22% on St. Pierre Bank. They found that Iceland scallops were susceptible to predation until the age of five years, at which point they achieved a size refuge which was a function of predator mouth gape. While probably not accounting for the rarity of scallops on fine substrata on the northeastern Grand Bank, American plaice would enhance the contrast in distribution of scallops between coarse and fine substrata through predation on these relatively rare occurring individuals on fine substrata. Other large predatory fish such as Atlantic cod (Gadus morhua Linnaeus, 1758) are not known to be associated with any particular substratum type and are typical opportunistic feeders. Examination of cod stomach contents from the Grand Banks reveals a very low incidence of Iceland scallops (G. Lilly, pers. comm.). Little is known about invertebrate predators of Iceland scallops. The Buccinidae and the Naticidae are probably the Table 1. Potential predators of the Iceland scallop, Chlamys islandica, on the northeastern Grand Bank of Newfoundland. Substratum Predator! Fine Coarse MOLLUSCA Buccinidae P P Naticidae pt CRUSTACEA Majidae- includes Hyas spp., P P and Chionocetes opilio ECHINODERMATA Asteroidea Asterias vulgaris PE Leptasterias sp. p* CHORDATA Rajidae Pt Gadidae P P Zoarcidae P P Cottidae P P Pleuronectidae K* Pseudopleuronectes americanus Limanda ferruginea ' -Identified along photographic transects (Schneider et al., 1987) * -Typical substratum association P-potential predator K-known predators (Langton and Bowman, 1981; Naidu and Meron, 1986) two major predatory gastropod groups on the Grand Banks. Species lists are incomplete for the study area and ecological relationships are poorly documented. However, from studies conducted in coastal areas, it is known that adult Buccinum undatum (Linnaeus, 1758) can be attracted over considerable distances (> 50 m/day) in search of bivalves which are their primary prey, on substrata including sand, mud and rock GILKINSON AND GAGNON: SCALLOP SUBSTRATUM ASSOCIATIONS 63 (Himmelman, 1988; Jalbert et al., 1989). Buccinid snails were common over all photographic transects and substratum types, although never abundant. The maximum average den- sity for a transect was 0.4 snails/photograph (0.07 snails/m?) (Schneider et al. , 1987). Although uncommon in photographs, naticid gastropods were observed along the study transects on sand substrata. Most species of Naticidae are relatively stenotypic and prefer sand or muddy substrata (Golikov and Sirenko, 1988). If these gastropods were exerting a heavy mortality on scallops, one would expect to find a surface ac- cumulation of empty scallop shells; this was not observed. Alternatively, naticids could be preying upon small scallops in the size range below the limits of resolution in photographs (i.e. < 2 cm). Crabs are known predators of various scallop species (Elner and Jamieson, 1979; Lake er al. , 1987). While poten- tially important predators of scallops such as majid crabs (Hyas spp. and Chionocetes opilio O. Fabricius, 1780) were common over most transects, they were not restricted to a particular substratum (Schneider ef al. , 1987). The extent of predation by various echinoderms (e.g. Asteroidea) on scallops is unknown, however, it is noted that asteroids were primarily associated with the coarsest substrata on the transects (Schneider et al/., 1987) and, therefore, would not be expected to cause excessive mortalities on sand substrata. In conclusion, it is considered unlikely that Iceland scallops are concentrated on coarse substrata due to differential sur- vivorship from intense predation pressure on fine substrata. BEHAVIOURAL AND PHYSICAL MECHANISMS MAINTAINING SCALLOP- SUBSTRATUM ASSOCIATIONS The most plausible explanation for the association of Iceland scallops with coarse substrata is the propensity, at all life history stages, towards byssal attachment to a stable substratum. It is known that a high percentage of individuals in a population of Iceland scallops are attached to the substratum by the byssus at any given time. Frequencies of 76% (laboratory) and 97% (field) byssally attached adult scallops have been reported by Naidu and Meron (1986) and Vahl and Clausen (1980), respectively. From diving obser- vations in West Greenland, Pedersen (1989) reported that Iceland scallops were attached to the substratum by the byssus, large scallops were attached directly to the substratum while small scallops were attached to larger scallops or empty shells. Vahl and Clausen (1980) postulate that because Chlamys islandica cannot recess on coarse sediments, it remains in danger of being swept away by currents and because C. islandica tends to occur in habitats with strong currents (Wiborg, 1962 fide Vahl and Clausen, 1980) byssal attach- ment remains necessary at all sizes. An important aspect in assessing the importance of water movements in shaping the distribution of scallops is the current speed required to dislodge byssally attached scallops. Gruffydd (1976) determined ‘*wash-away’’ velocities for Iceland scallops which ranged from 21 cm/s for 10-20 mm individuals to 26 cm/s for 65-70 mm individuals. For the northeastern Grand Bank of Newfoundland, Barrie et al. (1984) hypothesized that periodic, high unidirectional flow velocities (> 50 cm/s) occur, possibly every year to every few years at depths less than 110 m. Under these extreme flow conditions, the coarsest sediments become mobile and move as bedforms. At other times, maximum tidal current velocities (15 cm/s, Mobil, 1985) in the study area are less than those required to ‘‘wash-away’’ scallops. Therefore, within the study area, scallops are probably washed away infrequently although major storms would have the capability of dislodg- ing large numbers of scallops. At present, the extent of the impact of such extreme events is unknown. Swimming activity would make Chlamys islandica susceptible to being ‘“‘washed-away’’ from preferred substrata. Gruffydd (1976) determined that all sizes of Iceland scallops, and particularly medium-sized (30-40 mm shell height) in- dividuals, displayed a tendency to swim. However, Vahl and Clausen (1980) considered swimming activity to be a relative- ly rare phenomenon, with individual scallops making, on average, a swimming excursion every 31 days. This is in spite of the fact that byssus production is a minor item in the energy budget of C. islandica (Vahl and Clausen, 1980). There are limited data regarding conditions which initiate the swimming response. While Gruffydd’s (1976) experiments showed that the swimming response was strongest at the fastest current speed (15 cm/s), Vahl and Clausen (1980) determined the flight reaction to be less evident during periods when current speeds were strong (about 50 cm/s) and speculated that this was pro- bably due to the high risk to scallops associated with being carried to unsuitable habitats in strong currents. Byssal at- tachment rate of adult C. tehuelcha (d’Orbigny, 1835) in- creased dramatically over relatively small changes in current velocity, from 65% attachment at 6.6 cm/s to 90% at 8.3 cm/s (Ciocco et al., 1983 fide Orensanz et al., in press). Patterns of swimming behaviour in C. islandica may be determined by a combination of factors including habitat type and critical current velocity (in terms of initiating the swimming response) specific to these habitat types. SUBSTRATA AND HABITATS OCCUPIED BY EXTANT CHLAMYS SPP. In the evolution of the Bivalvia, neotenous retention of the byssus was an adaptive break-through in terms of physical stabilization, giving rise to an invasion of new habitats by epifaunal species (Stanley, 1972). The oldest (Triassic) pectinids are of the adult-byssate Chlamys type (Triassic) while the emergence of post-Triassic free-living pec- tinids evolved from byssate forms (Stanley, 1972). 64 AMER. MALAC. BULL. 9(1) (1991) Table 2. Approximate maximum sizes (shell height) and substrata and depths occupied by extant Chlamys spp!! Shell Height Habitat (mm) Substratum NORTHWEST ATLANTIC Chlamys islandica (Muller, 1776) 100 gravelly sand, shell rock SOUTHWEST ATLANTIC C. benedicti (Verrill and Bush, 1897) 13 ? C. mildredae (Bayer, 1943) 38 undersides of rocks C. sentis (Reeve, 1853) 38 undersides of rocks C. ornata (Lamarck, 1819) 40 undersides of rocks C. imbricata (Gmelin, 1791) 44 undersides of rocks C. multisquamata (Dunker, 1864) 59 rock crevices C. patagonica (King and Broderip, 1832) 79 consolidated sand, shell C. tehuelcha (d’Orbigny, 1835) 100 consolidated sand; shell- gravel, rocky bottoms EASTERN ATLANTIC C. furtiva (Loven) 19 muddy, gravelly sand C. striata (Muller, 1776) 19 muddy sand, gravel, shell C. tigerina (Muller, 1776) 25 sandy mud, gravel, rock C. multistriata (Poli, 1795) 30 2 C. tincta (Reeve, 1853) 30 ? C. sulcata (Muller, 1776) 40 2 C. distorta (da Costa) 50 ? C. flabellum (Gmelin, 1791) 50 7 C. septemradiata (Muller, 1776) 51 mud C. nivea (MacGillivray, 1825) 60 ? C. varia (Linnaeus, 1758) 64 rocks, muddy gravel, shell ICELAND C. islandica (Muller, 1776) 110 clay, sand, shell BERING SEA C. behringiana (Middendorff, 1849) ? ? C. pseudislandica (MacNeil, 1967)? 75 ? EASTERN PACIFIC C. jordani (Arnold, 1903) ? ? C. lowei (Hertlein, 1935) ? ? C. amandi (Hertlein, 1935) 40 uf C. incantata (Hertlein, 1972) 60 offshore (200 m) C. rubida (Hinds, 1845) 60 rocks, gravel, shell C. hastata hastata (Sowerby, 1842) 64 rocks, gravel C. hastata hericius (Gould, 1850) 83 rocks, sand, mud SOUTHWEST PACIFIC C. dichroa (Suter, 1909) 42 2 C. zelandiae (Gray, 1843) 30 undersides of rocks C. gemmulata (Reeve, 1853) 30 ? C. kiwaensis (Powell, 1933) 33 de C. zeelandona (Hertlein, 1931) 35? 4 C. atkinos (Petterd, 1886) 38 v4 C. luculenta (Reeve, 1853) 40 ? C. lentiginosa (Reeve, 1853) 40 coral reefs C. taiaroa (Powell, 1952) 43 ? C. funebris (Reeve, 1853) 50 ? C. australis (Sowerby, 1847) 60 He Depth? offshore (to 220 m) 9 inshore (upper subtidal) inshore (< 15 m) inshore (to 4 m) inshore (< 6 m) inshore/offshore (6-56 m) offshore (to 300 m) inshore/offshore (< 60 m) inshore/offshore (< 200 m) inshore/offshore inshore/offshore (to 550 m) inshore/offshore (to 2000 m) ‘5 offshore (to 850 m) inshore/offshore (to > 90 m) inshore inshore/offshore (11-183 m) offshore inshore/offshore (to 1000 m) inshore/offshore (to 300 m) offshore (40-150 m) inshore/offshore inshore/offshore (2-60 m) inshore/offshore (2-175 m) offshore Abbott and Dance, 1986 inshore/offshore (to 183 m) inshore/offshore (5-150 m) inshore/offshore (to 152 m) offshore (to 100 m) inshore (to 30 m) inshore (to 30 m) inshore?/offshore inshore inshore/offshore? offshore inshore inshore?/offshore inshore? offshore Source this study; Theroux and Wigley, 1983; Naidu and Cahill, 1989 Abbott, 1974 Abbott, 1974 Abbott, 1974; Rehder, 1981 Abbott and Dance, 1986 Abbott, 1974 Abbott, 1974 Orensanz et al., in press; O. Iribarne, pers. comm. Orensanz et al., in press; O. Iribarne, pers. comm. Tebble, 1966 Tebble, 1966 Madson, 1949; Tebble, 1966 Abbott and Dance, 1986 Abbott and Dance, 1986 Abbott and Dance, 1986 Tebble, 1966 Abbott and Dance, 1986 Allen, 1953; Tebble, 1966 Abbott and Dance, 1986 Allen, 1953; Tebble, 1966 Madson, 1949 Bernard, 1983 MacGinitie, 1959; Bernard, 1979 Bernard, 1983 Bernard, 1983; Keen, 1971 Abbott and Dance, 1986 Rehder, 1981; Kozloff, 1983 Rehder, 1981; Bourne, 1987 Rehder, 1981 Abbott and Dance, 1986; Powell, 1979 Abbott and Dance, 1986 Abbott and Dance, 1986 Powell, 1979 Powell, 1979 MacPherson and Gabriel, 1962 Abbott and Dance, 1986 Abbott and Dance, 1986 Powell, 1979 Abbott and Dance, 1986 Abbott and Dance, 1986 GILKINSON AND GAGNON: SCALLOP SUBSTRATUM ASSOCIATIONS 65 Table 2. (continued) Abbott and Dance, 1986 Abbott and Dance, 1986 Abbott and Dance, 1986; Powell, 1979 Bull, in press Young and Martin, 1989; R. McLoughlin, pers. comm. Young and Martin, 1989; R. McLoughlin, pers. comm. Powell, 1979 Bernard, 1983 Kuroda et al., 1971 Abbott and Dance, 1986 Abbott and Dance, 1986 Kuroda et al., 1971 Abbott and Dance, 1986; Kuroda et al., 1971 Abbott and Dance, 1986; Kuroda et al., 1971 Abbott and Dance, 1986 Kuroda et al., 1971 Abbott and Dance, 1986; Kuroda et al., 1971 Abbott and Dance, 1986; Kuroda et al., 1971 Abbott and Dance, 1986) Silina and Pozdnyakova, 1990 Kuroda et al., 1971 Abbott and Dance, 1986 Abbott and Dance, 1986 Shell Height Habitat (mm) — Substratum Depth? Source C. scabricostata (Sowerby, 1915) 60 offshore C. squamosa (Gmelin, 1791) 60 inshore C. dieffenbachi (Reeve, 1853) 64 inshore/offshore (to 35 m) C. delicutula (Hutton, 1873) 70 gravel and shell offshore (to 200 m) C. asperrimus (Lamarck, 1819) 100 muddy sand to sand* inshore/offshore (to 100 m) C. bifrons (Lamarck, 1819) 150 sandy and coarse bottoms inshore/offshore (to 100 m) C. consociata (E. A. Smith, 1915) ? ? inshore/offshore (to 182 m) NORTHWEST PACIFIC C. albida (Arnold, 1906) 2 offshore (to 200 m) C. princessae (Kuroda and Habe) 23? sand, shell offhsore (to 200 m) C. asperulata (Adams and Reeve, 1850) 25 2 inshore (to 20 m) C. albolineata (Sowerby, 1887) 25 2 inshore C. empressae (Kuroda and Habe) 30? sand, shell offshore (to 200 m) C. irregularis (Sowerby, 1842) 40 rocks, gravel inshore/offshore (to 600 m) C. jousseaumei (Bavay, 1904) 40 fine sand inshore/offshore C. larvata (Reeve, 1853) 40 2 offshore C. farreri nipponensis (Kuroda) iu rock, gravel inshore/offshore (to 60 m) C. lemniscata (Reeve, 1853) 50 sand, shell inshore/offshore (to 300 m) C. squamata (Gmelin, 1791) 75 rock, gravel inshore/offshore (to 50 m) C. gloriosa (Reeve, 1852) iP ? offshore C. rosealbus (Scarlato) 90 silty-sand with pebbles, inshore/offshore (13-2030 m) rocks (rarely sand, shells) C. nobilis (Reeve, 1852) 118? rocks inshore (to 20 m) INDIAN OCEAN C. ruschenbergerii (Tyron, 1869) 75 ? offshore C. senatoria (Gmelin, 1791) 75 2 offshore C. townsendi (Sowerby, 1895) 150 ? inshore (to 20 m) Abbott and Dance, 1986 'This is not a complete taxonomic listing. Bernard (1983) considers Hinnites to be a subgenus, however, because Hinnites spp. attach to the substrate by cementation rather than by a byssus this group has been excluded from analyses. In certain instances a species may occupy two geographic regions, however, in order to simplify the table the species is recorded for only one region. An exception was made in the case of C. islandica. ?Those species with maximum depth distributions of 30 m are considered inshore species. Note that this division is arbritrary and that not all deepwater occurrences are necessarily offshore. 3There is some debate over whether or not living C. islandica occurs in the eastern Pacific although it is reported to occur in the Arctic (Bernard, 1979; Lubinsky, 1980). Bernard (1979) considers C. islandica recorded from Point Barrow, Alaska (MacGinitie, 1959) to be C. pseudislandica while the more southerly occurring specimens from this collection he considers to be C. rubida. ‘Although widespread on a variety of soft substrata, both juveniles and adults usually found attached by byssus to available solid objects, i.e. rocks, other bivalves, pier pilings (Young and Martin, 1989). While there is limited ecological information on many During the Paleozoic, bivalves were primarily re- stricted to nearshore habitats and it was largely after the Paleozoic that the bivalvia spread offshore to attain their pre- sent distributions, replacing the previously dominant articulate brachiopods (Stanley, 1972). From a survey of the habitats occupied by extant species within the genus Chlamys, it is seen that there is a high proportion (73%) of species with an offshore distribution, or at least ranging from shallow water to offshore depths (Table 2). of these species, particularly with respect to substrata associa- tions, it would appear that members of the genus have radiated into a variety of habitat types. Substrata associations range from the undersides of shallow water boulders and coral reefs (e.g. subtropical and tropical species) to deepwater muds (e.g. Chlamys septemradiata Miiller, 1776). Kauffman (1969) classifies byssate species of Chlamys as byssate fissure- dwellers. The typical habitats of this group are the under- 66 AMER. MALAC. BULL. 9(1) (1991) sides of rocks, crevices and fissures, reef tunnels, spaces in- side root bundles of aquatic plants and similar niches with good water circulation, weak light and good protection from strong wave or current action. The occurrence of C. septemradiata on deep-water, flocculent muds (Allen, 1953) represents a unique substratum association within the genus. In fact, in this habitat, C. septemradiata serves as a stable settling surface for other sessile invertebrates which other- wise would be subjected to siltation. Examining species-specific maximum sizes within the genus, it is seen that while most shallow water crevice species are small (< 60 mm), there are several very small (< 30 mm) species which are found distributed offshore to great depths (Table 2). The Iceland scallop is one of the largest species within the genus Chlamys, attaining a maximum size of about 100 mm in the Newfoundland regions although on the northeastern Grand Bank most scallops are between 60 and 80 mm in shell height (Naidu and Cahill, 1989). The observed life habit orientation of Chlamys islandica on the northeastern Grand Banks is a fully exposed position. Because of the nature of the substratum (dense gravel, cobble), in most instances adult Iceland scallops must assume an epifaunal position on the exposed, upper surfaces of rocks. Only in the case of irregular occurrences of boulders would Iceland scallops be afforded the opportunity to assume a cryptic habit by attaching to the undersides. However, C. islandica could be compensated for this apparent lack of refuge through heavy biofouling by barnacles and soft corals in particular, which is often observed on the external sur- faces of the upper valve (KDG, pers. obs.). While this may not decrease the risk of predation from chemosensory orient- ing predators (see Lake et al., 1987), presumably it would be an advantage in the case of visually cueing predators, in particular, fish. Epizoic associations have been studied for several species. Epizoic sponge cover of the valves of C. varia (Linnaeus, 1758) and C. asperrima (Lamarck, 1819) is known to provide protection from predatory starfish (Forester, 1979; Chernoff, 1987; Pitcher and Butler, 1987). Chlamys dieffen- bachi (Reeve, 1853) is almost invariably enveloped in living sponge (Powell, 1979) while C. hastata and C. rubida (Hinds, 1845) are regularly colonized by sponges that form thick coatings (Kozloff, 1983). In summary, from the results of this study, and the observations of others, it would appear that Chlamys islandica is restricted to habitats with coarse substrata although this includes a range of sediment types from gravelly sand to gravel, cobble and shell mixtures. This would appear to be due to a requirement for byssus attachment at all life history stages. Within the genus, other species known to be bysally attached to substrates as adults include: C. asperrima (Young and Martin, 1989), C. varia (Rodhouse and Burnell, 1979), C. irregularis Sowerby, 1842, C. squamata Gmelin, 1791, C. farreri Jones and Preston, 1904 and C. nobilis Reeve, 1853 (Kuroda et al., 1971). Frequency of byssal attachment to the substratum decreases with age in C. tehuelcha although the capacity to form a byssus is not lost in the largest individuals (Orensanz et al. , in press). While occupation of habitats with coarse substrata appears to be typical of members of the genus, there is at least one example (C. septemradiata) of radiation into a deepwater habitat characterized by flocculent muds, presumably with a consequential loss of byssus attachment. ACKNOWLEDGMENTS We thank the Centre for Cold Ocean Research and Engineering (Memorial University of Newfoundland), the Atlantic Geosciences Centre (Bedford Institute of Oceanography, Dartmouth, Nova Scotia), the Offshore Geotechnics Program of the Federal Panel on Energy Research and Develop- ment, T. Folkes, M. Lewis and the Captain and crew of the C.S.S. HUD- SON for logistic support. We thank J. A. Hutchings and S. Naidu for review- ing the manuscript, G. Carmichael for graphics, and D. Pitcher for assistance with data analyses. We are grateful to S. Shumway, M. Bricelj, O. Iribarne, P. Young and R. McLoughlin for providing unpublished data or informa- tion sources for Table 2. LITERATURE CITED Abbott, R. T. 1974. American Seashells. The Marine Mollusca of the Atlantic and Pacific Coasts of North America. Second Edition. Van Nostrand Reinhold Company, New York. 663 pp. Abbott, R. T. and P. Dance. 1986. Compendium of Seashells. American Malacologists Inc. Melbourne, Florida. 411 pp. Allen, J. A. 1953. Observations on the epifauna of the deep-water muds of the Clyde Sea area, with special reference to Chlamys septemradiata (Miller). Journal of Animal Ecology 22(2):240-260. ASTM. 1988. Standard test methods for classification of soils for engineer- ing purposes. Jn: Annual book of ASTM (American Society for Testing and Materials) standards. Section 4. pp. 283-292. Philadelphia, Pennsylvania. Barrie, J. V., C. F. M. Lewis, G. B. Fader and L. H. King. 1984. Seabed processes on the northeastern Grand Banks of Newfoundland; modern reworking of relict sediments. Marine Geology 57:209-227. Bernard, F. R. 1979. Bivalve mollusks of the Western Beaufort Sea. Con- tributions in Science, No. 313. Natural History Museum of Los Angeles County. 80 pp. Bernard, F. R. 1983. Catalogue of the living bivalvia of the eastern Pacific Ocean: Bering Strait to Cape Horn. Canadian Special Publications of Fisheries and Aquatic Sciences No. 61, 102 pp. Bourne, N. 1987. Status of invertebrate fisheries off the Pacific coast of Canada. Canadian Technical Report of Fisheries and Aquatic Sciences No. 1576. 158 pp. Bull, M. In press. New Zealand. In: Scallops: Biology, Ecology and Aquaculture. S. E. Shumway, ed. Elsevier Science Publishers, Amsterdam. Cernohorsky, W. O. 1972. Marine Shells of the Pacific. Vol. I. Pacific Publica- tions, Sydney. 411 pp. Chernoff, H. 1987. Factors affecting mortality of the scallop Chlamys asper- rimus (Lamarck) and its epizoic sponges in South Australian waters. Journal of Experimental Marine Biology and Ecology 109:155-171. Elner, R. W. and G. S. Jamieson. 1979. Predation of sea scallops, Placopecten magellanicus, by the rock crab, Cancer irroratus, and the American lobster, Homarus americanus. Journal of the Fisheries Research Board of Canada 36:537-543. GILKINSON AND GAGNON: SCALLOP SUBSTRATUM ASSOCIATIONS 67 Fader, G. B., L. H. King and H. W. Josenhans. 1982. Surficial geology of the Laurentian Channel and the western Grand Banks of New- foundland. Marine Sciences Paper 21. Geological Survey of Canada Paper 81-22. Forester, A. J. 1979. The association between the sponge Halichondria panacea (Pallis) and scallop Chlamys varia (L.): a commensal- protective mutualism. Journal of Experimental Marine Biology and Ecology 36:1-10. Fraser, D. I. 1983. Observations on the settlement of pectinid spat off the west coast of Scotland in 1982. International Council for the Ex- ploration of the Sea. CM 1983/K:40. Golikov, A. N. and B. I. Sirenko. 1988. The naticid gastropods in the boreal waters of the western Pacific and Arctic Oceans. Malacological Review 21:1-41. Gruffydd, L. D. 1976. Swimming in Chlamys islandica in relation to cur- rent speed and an investigation of hydrodynamic lift in this and other scallops. Norwegian Journal of Zoology 24:365-378. Himmelman, J. H. 1988. Movement of whelks (Buccinum undatum) towards a baited trap. Marine Biology 97:521-531. Jalbert, P., J. H. Himmelman, P. Beland and B. Thomas. 1989. Whelks (Buc- cinum undatum) and other subtidal invertebrate predators in the northern Gulf of St. Lawrence. Naturaliste Canadian (Revue Ecologie et Systematics) 116:1-15. Kauffman, E. G. 1969. Form, Function, and Evolution. In: Treatise on In- vertebrate Paleontology. R. C. Moore, ed. pp. N130-N205. The Geological Society of America and the University of Kansas. Keen, A. M. 1971. Seashells of Tropical West America. Marine Molluscs from Baja California to Peru. Second Edition, Stanford University Press, Stanford, California. 1064 pp. Kozloff, E. N. 1983. Seashore Life of the Northern Pacific Coast. An Il- lustrated Guide to Northern California, Oregon, Washington and British Columbia. Douglas and McIntyre, Vancouver/Toronto. 370 pp. Kuroda, T., T. Habe and K. Oyama. 1971. The seashells of Sagami Bay. Collected by His Majesty The Emperor of Japan. Maruzen Co., Ltd., Japan. 489 pp. Lake, N. C. H., M. B. Jones and J. D. Paul. 1987. Crab predation on scallop (Pecten maximus) and its implication for scallop cultivation. Journal of the Marine Biological Association of the United Kingdom 67:55-64. Langton, R. W. and R. E. Bowman. 1981. Food of eight Northwest Atlantic pleuronectiform fishes. National Oceanic and Atmospheric Admin- istration (NOAA) Technical Report NMFS SSRF-749. 16 pp. Lubinsky, I. 1980. Marine bivalve molluscs of the Canadian central and eastern arctic: faunal composition and zoogeography. Canadian Bulletin of Fisheries and Aquatic Sciences No. 207. 111 pp. MacGinitie, N. 1959. Marine mollusca of Point Barrow, Alaska. Proceedings of the United States National Museum 109(3412):59-208, pls. 1-27. MacPherson, J. H. and C. J. Gabriel. 1962. Marine Molluscs of Victoria. Melbourne University Press, Melbourne, Australia. 475 pp. Madsen, F. J. 1949. The Zoology of Iceland: Marine Bivalvia. Volume IV, Part 63. Ejner Munksgaard, Copenhagen and Reykjavik. 85 pp. Mobil. 1985. Hibernia development project environmental impact statement. Vol. Ila. Biophysical Environment. Mobil Oil Canada Limited, St. John’s, Newfoundland. Naidu, K. S. 1988. Estimating mortality in the Iceland scallop, Chlamys islandica (O. F. Miiller). Journal of Shellfish Research 7(1):61-71. Naidu, K. S. and F. M. Cahill. 1989. The quest for Iceland scallops, Chlamys islandica (O. F. Miiller) on the Grand Bank, NAFO Division 3LNO. Canadian Atlantic Fisheries Scientific Advisory Committee Research Document 89/12. 31 pp. Naidu, K. S. and S. Meron. 1986. Predation of scallops by American plaice and yellowtail flounder. Canadian Atlantic Fisheries Scientific Ad- visory Committee Research Document 86/62. 25 pp. Orensanz, J. M., M. Pascual and M. Fernandez. In press. Scallop resources from the southwestern Atlantic. In: Scallops, Biology, Ecology and Aquaculture. S. E. Shumway, ed., Elsevier Science Publishers, Amsterdam. Pedersen, S. A. 1989. Inshore scallop resources, Chlamys islandica, in the Nuuk Area west Greenland. Northwest Atlantic Scientific Council Research Document 89/20. 16 pp. Pitcher, C. R. and A. J. Butler. 1987. Predation by asteroids, escape response, and morphometrics of scallops with epizoic sponges. Journal of Ex- perimental Marine Biology and Ecology 112:233-249. Pitt, T. K. 1976. Food of yellowtail flounder on the Grand Bank and a com- parison with American Plaice. International Commission for the the Northwest Atlantic Fisheries Research Bulletin 12:23-27. Powell, A. W. B. 1979. New Zealand Mollusca. Marine, Land and Freshwater Shells. William Collins Publishers Ltd., Hong Kong. 500 pp. Rehder, H. A. 1981. The Audubon Society Field Guide to North American Seashells. New York. 894 pp. Rodhouse, P. G. and G. M. Burnell. 1979. Jn situ studies on the scallop Chlamys varia. Pp. 87-98. In: Progress in Underwater Science Vol. 4. J. C. Gamble and J. D. George, eds. pp. 87-98. Pentech Press, Plymouth, United Kingdom. Schneider, D. C., J- M. Gagnon and K. D. Gilkinson. 1987. Patchiness of epibenthic megafauna on the outer Grand Banks of Newfoundland. Marine Ecology Progress Series 39:1-13. Silina, A. V. and L. A. Pozdnyakova. 1990. Growth of the scallop Chlamys rosealbus in the Sea of Japan. The Soviet Journal of Marine Biology (Translated) 16(1):32-36. Sokal, R. R. and F. J. Rholf. 1981. Biometry. The Principles and Practise of Statistics in Biological Research. Second Edition. W. H. Freeman and Company, San Francisco. 859 pp. Stanley, S. M. 1972. Functional morphology and evolution of byssally at- tached bivalve mollusks. Journal of Paleontology 46(2):165-212. Tebble, N. 1966. British Bivalve Seashells. A Handbook for Identification. Alden Press, Oxford. 212 pp. Theroux, R. B. and R. L. Wigley. 1983. Distribution and abundance of east coast bivalve molluscs based on specimens in the National Marine Fisheries Service Woods Hole Collection. NOAA Technical Report NMFS SSRF-768. 172 pp. Vahl, O. and B. Clausen. 1980. Frequency of swimming and energy coast of byssus production in Chlamys islandica (O. F. Miller). Journal du Conservation International Exploration de la Mer 39(1):101-103. Vilks, G. 1984. Report of Cruise 83033, Bedford Institute of Oceanography, Atlantic Geosciences Centre. Geological Survey of Canada internal report, Dartmouth, Nova Scotia. 51 pp. Wallace, J. C. 1982. The culture of the Iceland scallop, Chlamys islandica (O. F. Miller). I. Spat collection and growth during the first year. Aquaculture 26:311-320. Young, P. C. and R. B. Martin. 1989. The scallop fisheries of Australia and their management. CRC Critical Reviews in Aquatic Sciences 1(4):615-638. Date of manuscript acceptance: 1 October 1990 Anatomical and behavioural studies on vision in Nautilus and Octopus W. R. A. Muntz Department of Ecology and Evolutionary Biology, Monash University, Clayton, Victoria 3168, Australia Abstract. Nautilus is a cephalopod that is primitive in many respects, and is often considered to be a ‘‘living fossil’. The eye of Nautilus is apparently a primitive feature, acting as a pin-hole camera and lacking any lens or other dioptric apparatus. In contrast, in Octopus and most other coleoid cephalopods, there is a well formed spherical lens. The basic structure of the retina is similar in the two animals, but there are also a number of important differences: the microvilli of the receptors of Nautilus do not form a regular rectilinear array as they do in Octopus; the microvilli from neighbouring receptors overlap, which does not occur in Octopus; the supporting cells have a different structure; the nuclei of the supporting cells and receptor cells are distributed either side of the basement membrane in Octopus, but not in Nautilus, cilia are present in the retina of Nautilus but not Octopus; and the myeloid bodies are much more developed in Nautilus. Both behavioural experiments and calculation show that, as expected on anatomical grounds, visual acuity and sensitivity are much better in Octopus than Nautilus. Reasons for the limitations in the visual capabilities of the two animals are discussed. Nautilus is the last surviving genus of a group that arose in the Triassic, and has apparently changed little since Cretaceous times, at least as far as we can judge from the shell. The animal shows many apparently primitive features, such as the lack of an ink sac or chromatophores, the ex- ternal shell, the funnel formed of two overlapping lobes, and the simple pin-hole camera eye lacking any lens or other diop- tric apparatus (e.g. Morton, 1967). The ancestry of the genus has been discussed by, among others, Teichert and Matsumoto (1987), who concluded that it can truly be called a “‘living fossil’’. The first octopods, on the other hand, are found in the Upper Cretaceous (Donovan, 1977), and the group must be counted among the most developed invertebrates in ex- istence. The octopus eye, which resembles that of most other coleoid cephalopods, has in contrast to Nautilus a very well- developed lens and superficially looks remarkably similar to the eyes of vertebrates. The present paper compares the struc- ture and function of the eyes of these two very different cephalopods. STRUCTURE OF THE EYE IN NAUTILUS AND OCTOPUS The most obvious difference between the eyes of Nautilus and Octopus is the complete lack of any dioptric apparatus in the former genus. The pupil in Nautilus opens directly to the sea, and the eye must act as a pin-hole camera. The eyes of octopuses in contrast have well developed spherical lenses, which, as in most fishes, have focal lengths about 2.5 times their radius (Matthiessen’s ratio) and are well corrected for spherical aberration (Sivak, 1982; Sroczynski and Muntz, 1985). The overall size of the eyes of the two animals is, however, similar, and both animals have contrac- tile pupils that are elongated in the horizontal direction (Muntz, 1977; Hurley er al., 1978). Figure 1 shows the general appearance of the eyes of Nautilus pompilius Linnaeus and Octopus vulgaris Lamarck. Descriptions of the retinal anatomy of Nautilus can be found in Barber and Wright (1969), Muntz and Raj (1984), and Muntz and Wentworth (1987); and of Octopus in Young (1962a, 1971) and Yamamoto et al. (1965). These papers also give references to earlier work. Following convention, in this paper the segments of the receptors facing the light, which contain the photopigment, will be referred to as the distal or outer segments, and the nuclear region as the proximal or inner segment. The basic elements of the retina in both Octopus and Nautilus are the receptor cells, with distal segments consisting of a central core from which the microvilli (which contain the visual pigment) radiate outwards, and the supporting cells with their processes lying between the receptor cell outer segments (Figs. 2, 3). The packing of the receptor cells is roughly similar in the two species. Thus, there are about 20,000 receptor cells mm-? in N. pompilius, varying little over the retina (Muntz and Raj, 1984), and between 18,000 and 55,000 mm? in O. vulgaris, depending on retinal position (Young, 1971). Although basically similar, there are however also a number of important differences between the two species, which can be summarised as follows. American Malacological Bulletin, Vol. 9(1) (1991):69-74 69 70 AMER. MALAC. BULL. 9(1) (1991) Fig. 1. Left, vertical section through the eye of Nautilus (from Willey, 1902). Right, vertical section through the eye and optic lobe of Octopus (from Young, 1962b). (1) Transverse sections through the outer segments of the retinal receptors of Nautilus show that there are usually five or six (occasionally four or seven) bundles of microvilli running out from each receptor body to the bodies of neighbouring receptors. The receptors thus form a roughly hexagonal array (Fig. 2a). The microvilli from neighbouring receptors within a given bundle often interdigitate, although the extent of this interdigitation is not clear. The processes of the supporting cells run out in groups between these bundles of microvilli. In contrast in Octopus the receptor outer segments form a rectilinear array, with the microvilli oriented vertically or horizontally with respect to gravity (Fig. 3a). The microvilli of each receptor remain strictly segregated from those of the neighbouring receptors, with no interdigitation. (ii) The structure of the supporting cells is quite dif- ferent in the two animals. In Nautilus each cell has a number of fine microvillous processes which project out between the receptor outer segments in groups, whereas in Octopus each supporting cell has a single process, which is much larger and contains screening pigment. In the former animal the cell nuclei of both the receptors and the supporting cells lie distal to the basement membrane, whereas in Octopus the support- ing cell nuclei lie distal and the receptor cell nuclei proximal to the basement membrane. (iii) In Nautilus the supporting cells have cilia, as well as the microvillous processes that extend between the recep- tors. It is not certain whether the receptor cells have cilia as well. Ciliary structures have not been reported in the retina of any other adult cephalopod, although the photosensitive organs of many animals have receptors of ciliary origin, or cilia, presumed not to be photosensitive, intermingled with the receptors (Vanfleteren, 1982). (iv) The inner segments of Nautilus photoreceptors have complex myeloid bodies, which often have the ap- pearance of a tubular structure, or a series of wavy plates (Fig. 2). It has been argued that this apparently complex struc- ture consists of a series of dimpled plates, stacked in register above each other (Muntz and Wentworth, 1987). In Octopus and the other coleoid cephalopods, the myeloid bodies are reduced to a few membranous strands. It is interesting that some of the characteristics by which the Nautilus retina differs from that of adult octopuses also have been found during the development of the embryos of coleoid cephalopods. Thus in the cuttlefish Sepiella Japonica Sasaki, embryos have cilia on both receptor and sup- porting cells, the nuclei of both the receptor cells and the supporting cells lie distal to the basement membrane, and the supporting cells send long microvillous processes out among the whole length of the receptor outer segments (Yamamoto, 1985). Work in progress shows a similar situa- tion in the embryos of the Australian octopuses Octopus pallidus Hoyle and O. australis Hoyle (Wentworth and Muntz, unpub. data). BEHAVIOURAL STUDIES To date, no studies of vision have been carried out with Nautilus using any form of learnt behaviour, and it is not known how far the animals are capable of learning. However, Nautilus shows two well developed forms of innate visual behaviour, the positive phototactic response and the optomotor response, which have been used to determine the animals’ visual acuity, and also their absolute and spectral sensitivities (Muntz and Raj, 1984; Muntz, 1986, 1987). As we should expect for an animal with an eye having the simple optics of a pin-hole camera, visual performance in Nautilus is very poor compared to that of animals with lens bearing camera eyes. The minimum separable visual acuity, for example, measured using the optomotor response, lies between 5.5° and 11.25°, which agrees well with values calculated on the basis of the gross dimensions of the eye and pupil, and with expectations based on photographing a visual test chart using a scale model of the eye (Muntz and Raj, 1984). This can be compared with values of about 5’ MUNTZ: VISION IN NAUTILUS AND OCTOPUS fA sup. n.ret. Fig. 2. Diagram of the structure of the retina of Nautilus, as seen in tangential (above) and radial section (below). The diagram is not to scale: in particular the horizontal dimensions have been exaggerated compared to the vertical dimensions for clarity. The mean length of the distal segments is in fact about 360 um, and of the proximal segments 100 »m, and the mean centre to cen- tre distance between adjacent receptors 3.5 um: there is little variation over the retina (bas.m., basement membrane; dist., receptor distal segment with microvilli radiating from a central core; pr.s., receptor proximal segment; sup., processes of supporting cells; m., myeloid body; on., optic nerves; n.sup., nucleus of supporting cells; n.ret., nucleus of retinal cell). obtained with octopuses, various fishes and two aquatic mam- mals (Table 1): even with the most favourable estimate of 5.5° the performance level of Nautilus is over 60 times worse than for these other aquatic animals. By using the positive phototactic behaviour of Nautilus it has also been possible to determine its absolute sensitivity to tungsten light (Muntz, 1987). In itself this result is not par- ticularly useful, because the spectral output of the tungsten source used was very different from that of the light to which the animal will be exposed in its natural environment. However, given a knowledge of the spectral transmission of the water in which the animals live, the spectral quality of the daylight reaching the surface of the sea, and the animals’ own spectral sensitivity, it is possible from these results to 0000 00008 9000 0F0% 0 cr ePoneeqooste® THT TTS iu tiit ~~ dist. Ay otn gre ete ttid Lidedis tit Lit wwe UU iid TTT TTT TTT TTT Tf Apia LiL e> \s e—-s~ ere prerniy Litt ~e r= — dit sees = 7. ITT Ui titiiit THrnryT pitti teri tidittistiiiiiditiay 00 090600 O09 0 0900000900 00%%0 Meee [uvupeecueuuce TITTY TT Ty LLU THTITETT ITT 98 © 000 © O50 00809 0900 00% 90000 codehooocce pouty jis TTT ue TTT tt TTTTTTIATIT ITT TTT 000 ee ©f © 00 06900009 00 0960_00 0000 Littl euuvecl o> SNe S TTT Soe cece Cetete bas.m. eff. Fig. 3. Diagram of the structure of the retina Octopus, as seen in tangential (above) and radial section (below), from Young (1962a). Not to scale: in fact the distal segments vary between about 60 um and 180 pm in length depending on retinal position, while the proximal segments are about 90 pm long and the individual rhabdomeres about 5 ym in width (Young, 1962a) (rh., rhabdome; pig., pigment granule at centre of rhabdome; ret., retinal cell; ].m., limiting membrane; dist., distal segment of receptor; bas.m., basal membrane; pr.s., proximal segments of retinal cells; col.f., fine dendritic collateral of retina cell; p.l., retinal nerve plexus; eff., ending of efferent fibre in retina; ep., epithelial cell; n.sup., nuclei of supporting cells; sup. , processes of supporting cells). calculate the maximum depth at which surface light would be visible at all to Nautilus. Reasonable estimates for the first two factors are available, and the animals’ spectral sensitiv- ity was taken to be the same as the absorption spectrum of its visual pigment (see Muntz, 1987 for details). It appears that some daylight should be visible to Nautilus down to 800 m, which is slightly deeper than the maximum depth at which the animal is found. This is, however, considerably less than the maximum depth at which daylight should be visible to deep sea fishes, which has been calculated by Clarke and Denton (1962) as over 1000 m. Calculations based on the dimensions of the Nautilus eye also indicate that Nautilus will be less sensitive, by about 2 log units, than a fish or cephalopod that has a camera eye with a lens obeying A AMER. MALAC. BULL. 9(1) (1991) Matthiessen’s ratio (Muntz and Raj, 1984). Finally, the positive phototactic behaviour has also been used to determine spectral sensitivity directly (Muntz, 1986). The sensitivity curve obtained agreed well with the absorption spectrum of the extractable visual pigment, which was itself well fitted by Dartnall’s (1953) visual pigment nomogram for an A,-based pigment with its maximum at 467 mm. In contrast to Nautilus, octopuses learn visual discriminations very readily, and a great deal of information is now available on their visual capabilities (Wells, 1978; Messenger, 1981 for reviews). Most of this work has con- cerned higher visual functions, such as the ability to discriminate shapes, the mechanisms by which such discriminations are learnt, and the function of the various parts of the central nervous system. Comparatively little work has been done on the animals’ more basic visual capabilities, such as sensitivity or visual acuity, which are probably more directly related to the optics of the eye and the structure and function of the retina, and which can be compared to the data that are available for Nautilus. In the case of sensitivity, for example, there appear to have been no studies at all carried out on coleoid cephalopods using learning, and only one study involving innate behaviour, in which the spectral sensitivity of Loligo pealei Lesueur larvae was measured using the positive phototactic response in a manner rather similar to that used with Nautilus. The results were also similar show- ing a smooth bell-shaped spectral sensitivity curve maximal at around 480 mm and compatible with a single visual pig- ment (White, 1924). A few behavioural studies have been carried out on the visual acuity of octopuses. Thus Sutherland (1963), us- ing a training situation, obtained an estimate of 17’ for Octopus vulgaris, and Packard (1969), using the same species and the optomotor response, found that in very small specimens (<3-22g) acuity improved with size. The most recent studies on acuity in octopuses (Muntz and Gwyther, 1988a, 1989) used fully grown animals and a two choice learning situation, and the stimuli were gratings of equally spaced black and white stripes oriented vertically, horizon- tally, or obliquely at 45° The animals were trained to discriminate these gratings from each other or from a uniform gray stimulus, and visual acuity was taken as the separation between the bars of the gratings where performance reached chance levels. The results showed that the minimum separable visual acuity of O. australis and O. pallidus is about 5’ (Fig. 4). With gratings close to the animals’ threshold, performance with the vertical gratings was best, and with the horizontal gratings worst, but the effect was not large. The ability of Octopus pallidus and O. australis to discriminate distances has also been determined behavioural- ly, using the animals’ tendency to attack the nearer of two stimuli presented simultaneously (Muntz and Gwyther, 1988b). Assuming that the animals are using accommoda- Table 1. Minimum separable visual acuities, in minutes of arc, of various aquatic animals measured behavioural- ly using gratings. Learnt discriminations were used in all cases except Nautilus where the optomotor response was used. Species Acuity Reference MAMMALS Harbour seal 8.3 Schusterman and Balliet, 1970 Phoca vitulina Linnaeus Stellar Sea Lion 71 Schusterman and Balliet, 1970 Eumetopias jubata (Schreber) TELEOST FISHES Convict fish 4.9 Yamanouchi, 1956 Microcanthus strigatus (Cuvier and Valenciennes) Minnow 10.8 Brunner, 1934 Phoxinus laevis Linnaeus Skipjack tuna a) Nakamura, 1968 Katsowomis pelamis Linnaeus Little tuna 7.4 Nakamura, 1968 Euthynnus affinis (Cantor) Cichlid fish 5.8 Baerends et al., 1960 Aequidens portalegrensis (Hensel) CEPHALOPODS Nautilus 330-670 Muntz and Raj, 1984 Nautilus pompilius Octopus 5.0 Muntz and Gwyther, 1988a Octopus pallidus O. australis MUNTZ: VISION IN NAUTILUS AND OCTOPUS 7 tion to estimate distance, which various tests indicated is the most likely mechanism, the animals can detect blurring of points on the retinal image comparable in size to a single retinal receptor, and lens displacements of around 10 um. Finally, training experiments have shown that Octopus vulgaris can discriminate the plane of polarised light (Moody and Parriss, 1961), and also it has been shown that two species of decapod larvae orient themselves to the plane of polarised light (Jander et al., 1968). It is not known whether Nautilus has this ability. DISCUSSION The evolution of the camera eye has attracted interest ever since Darwin [1859 (reprinted 1958)] listed it as one of the ‘‘organs of extreme perfection and complication’’, and wrote that to believe that such organs could have been formed by natural selection seems ‘‘absurd in the highest possible degree’’. Darwin’s solution to the problem was to suggest that the eye must have arisen through numerous inheritable grada- tions, each of which was useful to its possessor, and that while strictly we should look for such gradations among the animal’s lineal ancestors, we are usually forced to look at living species of the same group to see what gradations are possible. The eye of Nautilus could be taken as such a grada- tion on the route to the complex eyes of the more recent response Probability of correct 0 10 20 30 40 Visual angle (minutes) Fig. 4. Visual acuity of Octopus, measured behaviourally for various stimulus combinations. The animals weighed between 59 and 1134g. A, vertical gratings against grey; +, horizontal gratings against grey; X, oblique gratings against grey; Ml, vertical gratings against horizontal gratings. Data from Muntz and Gwyther (1988a, 1989). cephalopods. Even though Nautilus vision is poor, never- theless its visual behavior is precise, in that in the optomotor response they follow the stripes accurately without visible lag, and in the phototactic situation if the difference between the stimuli is well above threshold the brighter light is chosen on every occasion. The eyes are also stabilised with respect to gravity by means of the statocysts (Hartline et al. , 1979). These facts suggest that vision is important to the animal. It is not, however, clear what use they make of such poor vision in their normal life. The habitat of Nautilus often has strong currents, and the optomotor behaviour could be related to holding station under these conditions. It could also be that the positive phototactic behaviour is related to bioluminescence, which is a major source of light at depth in the sea. Nautilus is often trapped in association with deep water bioluminescent shrimps, which also feed on decaying animal material, and moving towards bioluminescence could help take the animals towards their food. Finally, Nautilus shows diurnal vertical migrations (Carlson er al. , 1984; Ward et al., 1984), and vision could be a factor in this behaviour. Without further information on the normal behaviour of the animals however, these remain speculations. In the case of Octopus we have no behavioural evidence on its visual sensitivity, although presumably it is considerably better than that of Nautilus. The minimum separable visual acuity of 5’ for Octopus is comparable to that of fishes and aquatic mammals (Table 1). Nevertheless, it is not clear why the acuity is not even better than this. The retinal mosaic is rather finer than necessary for the acuity that is in fact achieved (Muntz and Gwyther, 1988a), and the pupil size is large enough that diffraction will not be limiting. In Eledone cirrhosa (Lamarck), another octopod, spherical aberration is far less than would be needed to limit acuity to this level (Sroczynski and Muntz, 1985). Furthermore, terrestrial animals can achieve much better acuities; in the case of humans, for example, the minimum separable lies between 0.5’ and I’ (e.g. Senders, 1948), and in the American kestrel, Falco sparvernius Linne, the acuity, measured behavioural- ly using square wave gratings, is 0.19’ (Fox ef al., 1976). Since, however, the minimum separable acuity, measured be- haviourally, has been found to be about 5’ for all aquatic animals where it has been measured, it could be the environ- ment itself that is limiting. While not very many data are available on the subject, it is clear that high spatial frequen- cies are particularly heavily attenuated by the water body itself, and it could be that the ability to resolve very fine detail is consequently irrelevant (see Muntz, 1990 for further discussion). The ability of octopuses and other coleoid cephalopods to discriminate the plane of polarisation of light is usually attributed to the regular rectilinear array of the microvilli of their receptors (Moody and Parriss, 1961). Nautilus lacks such a rectilinear array. Nevertheless, the microvilli within any 74 AMER. MALAC. BULL. 9(1) (1991) given bundle remain parallel to each other, and so plane polarised light should still be able to affect the receptors dif- ferentially, even though there would be no precise relation- ship between the plane of polarisation and the receptors stimulated. It would be interesting to know whether Nautilus can show any differential response to polarised light. LITERATURE CITED Baerends, G. P., B. E. Bennema and A. A. Vogelzang. 1960. Uber die Anderung der Sehscharfe mit dem Wachstum bei Aequidens portalegrensis (Hensel) (Pisces, Cichlidae). Zoologische Jahrbucher 88:67-78. Barber, V. C. and D. E. Wright. 1967. The fine structure of the sense organs of the cephalopod mollusc Nautilus. Zeitschrift fur Zellforschung und micoscoskopichse Anatomie 102:293-312. Brunner, G. 1934. Uber die Sehscharfe der Elritze (Phoxinus laevis) bei verschieden Helligkeiten. Zeitschrift fur vergleichende Physiologie 21:296-316. Carlson, B. A., J. N. McKibben and M. V. DeGruy. 1984. Telemetric in- vestigation of vertical migration of Nautilus belauensis in Palau (Western Caroline Islands, Pacific Ocean). Pacific Science 38:183-188. Clarke, G. L. and E. J. Denton. 1962. Light and animal life. Jn: The Sea, Vol. 1, M. N. Hill, ed. pp. 456-468. John Wiley, New York. Dartnall, H. J. A. 1953. The interpretation of spectral sensitivity curves. British Medical Bulletin 9:24-30. Darwin, C. 1958. The Origin of Species. Reprint of Ist edition (1859), J. W. Burrow, ed. Penguin Books, Harmonsworth. 477 pp. Donovan, D. T. 1977. In: Symposium on the biology of cephalopods, M. Nixon and J. B. Messenger, eds. Symposia of the Zoological Society of London 38:277-285. Fox, R., S. W. Lehmkuhle and D. H. Westendorf. 1976. Falcon visual acui- ty. Science 192:263-265. Hartline, P. H., A. C. Hurley and G. D. Lange. 1979. Eye stabilization by statocyst mediated oculomotor reflex in Nautilus. Journal of Com- parative Physiology 132:117-1226. Hurley, A. C., G. D. Lange and P. H. Hartline. 1978. The adjustable ‘* pinhole camera’ eye of Nautilus. Journal of Experimental Zoology 205:37-44. Jander, R., K. Daumer and T. H. Waterman. 1963. Polarised light orienta- tion by two Hawaiian decapod cephalopods. Zeitschrift fur vergleichende Physiologie 46:383-394. Messenger, J. B. 1981. Comparative physiology of vision in molluscs. Jn: Handbook of Sensory Physiology, vol. VU/6C, Comparative physiology and evolution of vision in invertebrates, H. Autrum, ed. pp. 93-200. Springer-Verlag: Berlin, Heidelburg, New York. Moody, M. F. and J. R. Parriss. 1961. The discrimination of polarised light by Octopus: A behavioural and morphological study. Zeitschrift fur vergleichende Physiologie 44:269-291. Morton, J. E. 1967. Molluscs (4th edition). London. Hutchinson. 244 pp. Muntz, W. R. A. 1977. Pupillary responses of cephalopods. Jn: Symposium on the biology of cephalopods, M. Nixon and J. B. Messenger, eds. Symposia of the Zoological Society of London 38:277-285. Muntz, W. R. A. 1986. The spectral sensitivity of Nautilus pompilius. Journal of Experimental Biology 126:513-517. Muntz, W. R. A. 1987. Visual behaviour and visual sensitivity of Nautilus pompilius. In: The biology and paleobiology of a living fossil. W. B. Saunders and N. H. Landman, eds. pp. 231-244. Plenum Press, New York. Muntz, W. R. A. (1990). Stimulus, environment and vision in fishes. Jn: The visual system of fish. R. H. Douglas and M. B. A. Djamgoz, eds. pp. 491-511. Chapman and Hall, London. Muniz, W. R. A. and J. Gwyther. 1988a. Visual acuity in Octopus pallidus and Octopus australis. Journal of Experimental Biology 134:119-129. Muntz, W. R. A. and J. Gwyther. 1988b. Visual discrimination of distance by octopuses. Journal of Experimental Biology 140:345-355. Muntz, W. R. A. and J. Gwyther. 1989. The visual acuity of octopuses for gratings of different orientations. Journal of Experimental Biology 142:461-464. Muntz, W. R. A. and U. Raj. 1984. On the visual system of Nautilus pompilius. Journal of Experimental Biology 109:253-263. Muntz, W. R. A. and S. L. Wentworth. 1987. An anatomical study of the retina of Nautilus pompilius. Biological Bulletin 173:387-397. Nakamura, E. L. (1968). Visual acuity of two tunas, Katsuwonus pelamis and Euthyunus affinis. Copeia 1:41-49. Packard, A. (1969). Visual acuity and eye growth in Octopus vulgaris (Lamarck). Monitore Zoologica Italiana (N.S.) 3:19-32. Schusterman, R. J. and R. F. Balliet. 1970. Visual acuity of the harbour seal and the stellar sea lion underwater. Nature, London 226:563-564. Senders, V. L. (1948). The physiological basis of visual acuity. Psychological Bulletin 45:465-490. Sivak, J. G. 1982. Optical properties of a cephalopod eye. Journal of Com- parative Physiology 147A:323-327. Sroczyrski, S. and W. R. A. Muntz. (1985). Image structure in Eledone cirrhosa, an octopus. Zoologische Jahrbuker 89:157-168. Sutherland, N. S. (1963). Visual acuity and disrimination of stripe widths in Octopus vulgaris Lamarck. Pubblicationes Stazione Zoologica Napoli 33:92-109. Teichert, C. and T. Matsumoto. 1987. The ancestry of the genus Nautilus. In: Nautilus: The Biology and Paleobiology of a Living Fossil. W. B. Saunders and N. H. Landman, eds. pp. 25-32. Plenum Press, New York. Vanfleteren, J. R. 1982. Monophyletic line in evolution? Ciliary induced receptor membranes. In: Visual Cells in Evolution. J. A. Westfall, ed. pp. 107-136. Raven Press, New York. Ward, P. D., B. Carlson, M. Weekley and B. Brumbaugh. 1984. Remote telemetry of daily vertical and horizontal movement by Nautilus in Palau. Nature, London 309:248-250. Wells, M. J. 1978. Octopus: Physiology and Behaviour of an Advanced In- vertebrate. London: Chapman and Hall. 417 pp. White, G. M. 1924. Reactions of the larve of the shrimp, Palaemonetes vulgaris, and the squid, Loligo pealii, to monochromatic light. Biological Bulletin 47:265-273. Willey, A. 1902. Contribution to the natural history of the pearly Nautilus: Zoological results based on material from New Britain, New Guinea, Loyalty Islands and elsewhere, collected during the years 1895, 1896 and 1897: Part 6:691-683. University Press Cambridge, England. Yamamoto, T. 1985. Ontogeny of the visual system in the cuttlefish, Sepiella japonica. 1. Morphological differentiation of the visual cell. Journal of comparative Neurology 232:347-361. Yamamoto, T., K. Tasaki, Y. Sugarara and A. Tonosaki. 1965. Fine struc- ture of the octopus retina. Journal of Cell Biology 25:345-365. Yamanouchi, T. 1956. The visual acuity of the coral fish Microanthus strigatus (Cuvier and Valenciennes). Publications of the Seto Marine Biology Laboratories 5:133-156. Young, J. Z. 1962a. The retina of cephalopods and its degeneration after optic nerve section. Philosophical Transactions of the Royal Society of London B245:1-18. Young, J. Z. 1962b. The optic lobes of Octopus vulgaris. Philosophical Trans- actions of the Royal Society of London B245:19-58. Young, J. Z. 1971. The Anatomy of the Nervous System of Octopus vulgaris. Oxford University Press. 690 pp. Date of manuscript acceptance: 9 November 1990 Complex learning in Octopus bimaculoides Jean Boal Curriculum in Ecology, University of North Carolina, Chapel Hill, North Carolina, 27599-3275, U.S.A. Abstract. In order to investigate complex learning in Octopus bimaculoides (Pickford and McConnaughey), I presented subjects with a series of com- binations of mollusc shells. Combinations consisted of two shells of one type and an odd shell of another type. The shells were suspended in the octopuses’ home tanks, and the animals were rewarded with food for correctly grabbing the odd shell. Associative learning was demonstrated by the subjects’ eventual mastery (70 - 100% success rates) of each combination in a series (A+ B— B-—), (C+ D— D-) ... By mastery of new combinations of the same stimull, (A+ D— D-), (C+ B- B-) geees subjects demonstrated transfer of learning. Learning improved across successive combinations, evidence for learning set formation. However, because octopuses did not learn to choose the odd stimulus when trained only with non-repeating combinations, no evidence in- dicated that the octopuses formed the relative class concept of oddity. The demonstration of complex learning in cephalopods could provide important insight into the evolution of cogni- tion. Most research on complex learning has focused on higher vertebrates. However, an ecological approach suggests that complex learning could evolve whenever it was adap- tively advantageous (Shettleworth, 1984). Several investiga- tions have suggested a connection between complex learning abilities and sociality (see Humphrey, 1976; Essock-Vitale and Seyfarth, 1986). Wells (1978) has argued, however, that because the predominantly solitary octopuses have no obvious means of self defense and live in a highly competitive en- vironment, they also could show complex learning. The ex- periments reported here test this proposal. I presented Octopus bimaculoides (Pickford and McConnaughey) with three objects, two alike and one dif- ferent, and rewarded them with small pieces of squid if they grabbed the odd one. Using this methodology, I could pose a series of problems differing in the complexity of learning required for successful mastery. The first question was, could the octopuses learn to choose a particular shell, in repeated presentations of the same combination? Simple associative learning of this sort has previously been demonstrated with Octopus vulgaris (Cuvier) (Wells, 1978). The second question was, could the octopuses still pick the correct shell if known shells were arranged into new com- binations? Positive results would show transfer of learned response tendencies. Thirdly, as the animals gained experience in learning shell combinations, would they improve at learning new but similar tasks? An ability for learning to learn, or developing a learning set, has not been shown previously in an in- vertebrate, although the related task of learning reversals has (Mackintosh, 1965; Morrow and Smithson, 1969). Lastly, could the octopuses eventually generalize and immediately choose the odd shell when presented with new combinations? Generalization has been shown in tactile discriminations for Octopus vulgaris (Wells and Young, 1970). This task is particulary significant because oddity is an abstract concept, defined only in relationship to other ob- jects and not by any attribute of the object itself. So far, only higher invertebrates have shown evidence of forming such relative class concepts (Thomas, 1980; Lombardi et al., 1984; Thomas and Noble, 1988). METHODS EXPERIMENT I Subjects were three wild caught adults of undetermined sex which had been living in the laboratory a full six months before experiments began. In the laboratory, they were noc- turnal and not easy to clock-shift. Therefore, experiments were performed at night under red light. The octopuses were trained initially to take small pieces of frozen squid from a rod and, later, to grab a single plastic triangle or square on the end of a rod to get a food reward. On days when they were unsuccessful with dis- crimination tasks, I fed them after trials using the rod alone, up to their minimum daily intake. Trials consisted of presenting three stimuli, two alike and one different. Six different combinations were used (Appendix 1). The combinations presented were composed of mollusc shells varying in color, texure, and shape, except for combinations four and five, which consisted of plastic shapes which varied in both texture and shape. I suspended the stimuli on nylon monofilament in the American Malacological Bulletin, Vol. 9(1) (1991):75-80 75 16 AMER. MALAC. BULL. 9(1) (1991) Experiment | Experiment Il Fig. 1. Apparatus for presenting combinations of shells. Location of the odd, positive stimulus (+) was determined randomly. In Experiment I, the shells were suspended on monofilament; in Experiment II, they were attached to acrylic rods. octopuses’ tanks (Fig. 1). The location of the odd object was determined randomly, with the constraint that in half of the presentations it was in the front half of the tank, and in half it was in the back. A subject was then given two minutes in which to grab one of the stimuli. Responses were usually im- mediate. A correct response was promptly rewarded with a small piece of squid. I gave each subject eight trials per learning session, ten to fifteen minutes apart, with two learn- ing sessions per day, ll sessions per week. In Experiment I, each of the six combinations was pre- sented for 11 sessions or until all subjects reached a success rate of greater than 50% for three successive sessions. The second combination was cut short because one of the shells was shattered by a particularly vigorous grab. Octopuses were then retested for two or three sessions with each familiar com- bination. They were then each given an equal number of presentations with three of eight arbitrarily chosen new com- binations of the same, familiar stimuli. Positive, rewarded shells remained positive and negative, unrewarded shells re- mained negative; however, the particular combinations of positive and negative stimuli were new. Experiment I ran for a total of 119 sessions across 12 weeks. Response rates averaged 43% for the three octopuses. For each octopus, sessions with fewer than two grabs were eliminated from the study. EXPERIMENT II Subjects were three freshly caught Octopus bimacul- oides, just reaching sexual maturity, one female and two males. Initial training was carried out as described above. Trials consisted of presentation of three stimuli, two alike and one different, as before. All stimuli used were mollusc shells (Appendix 2). Shells were presented in a line (Fig. 1) to cor- rect for the location bias found in Experiment I. I gave the octopuses ten trials each learning session, spaced five to ten minutes apart, with one session per day, six days per week. In this experiment, every combination presented to a subject (each trial) was novel. Sixty pairs of shells were used, with every type of shell presented both as a positively reward- ed odd shell and as a negative pair, in order to control for the possibility that subjects had attended to some attribute other than oddity. Each type of shell was therefore seen by the octopuses only twice within each week: once as a positive single shell, and once as a negative pair of shells. The only way to solve this problem successfully would be to employ the relative class concept of oddity. Combinations of shells were determined randomly with the constraint that positive odd shells differed from the negative pair by two of the three features of color, texture, or type (bivalve or gastropod) (Appendix 2). Orders of presentation and locations of the odd shells were randomized. Simple reinforcement trials were given at the begin- ning and end of each training session and randomly inter- spersed among the oddity discrimination problems. These trials consisted of presenting the octopus with the single plastic triangle or square used in initial training. Octopuses were rewarded for grabbing the shape (no choice or dis- crimination was involved). Response rates in discrimination tasks were 89%. To reduce any inadvertent cuing, beginning with the second week of trials a gauze curtain was draped between the experimenter and the octopuses, with red lights only on the octopuses’ side. RESULTS EXPERIMENT I Performances of the three octopuses were statistically indistinguishable (contingency table for three subjects ver- sus correct or incorrect response: x? = 0.694, d.f. = 2, P > 0.70). Results were therefore pooled. Octopuses showed a strong bias towards objects in the front of the tank in the first experiment (y? = 107.15, d.f. = 1, P < 0.01); subjects grabbed front stimuli more frequent- ly and less accurately. Response rates were low and not related to success rates (Figs. 2, 3). Could the octopuses learn to choose correctly one ob- ject out of a combination with repeated presentations? Suc- cess rates for the first combination, in terms of correct choices as a percentage of total grabs (Fig. 2), showed clearly that the combination was learned and retained. Despite the small sample size and wide variability in performances, learning curves for the first 11 sessions of all six combinations (Fig. 3) demonstrated that the octopuses were able to learn to choose the correct object consistently. Could they still choose the correct shell if known shells were recombined in new ways? This task required the ability BOAL: LEARNING IN OCTOPUS 100 Percent Success Number of Grabs 0 5 10 15 20 Session Number Fig. 2. Mean success rates (squares) (number of correct grabs as percen- tage of total grabs) with standard errors and total number of grabs (diamonds) across sessions, for the three octopuses on the first discrimination combination. to remember five (combination two had been broken) simultaneous discriminations, or positive or negative at- tributes of at least five of the ten different stimuli. Their high success rate with the familiar arrangements shows good reten- tion over this time period, and also shows that five simultaneous discriminations can be mastered by these oc- topuses. That they performed equally well on the original and new combinations (t = 1.01, d.f. = 26, P > 0.20) is evidence of transfer of learning (Fig. 4); they could use the learned information about member shells of a combination in a new context. Could the octopuses improve at the task of learning new combinations? Variability was too high and sample sizes too small for comparisons to be made among slopes and in- tercepts of the six learning curves in the series. However, per- formances on days one and five (Fig. 5) (or for combination two, the last two completed learning sessions) across the six sequential combinations showed a trend toward improvement in first-day performances. For the sixth combination, a second observer was present on the first day and not on the fifth, a difference that could have affected performances. A paired comparisons test on all six combinations for the two days showed significant effects for both day and combination (days, F = 18.98, d.f. = 1,5, P < 0.01; combinations, F = 10.81, d.f. = 5,5, P < 0.025). The octopuses thus both learned the individual combinations and improved across the series. Sam- ple sizes were too small to control for any effects of shells (combinations 1-3, and 6) versus plastic objects (combina- tions 4-5). The last question was, could the octopuses generalize and choose the odd object after learning a number of dif- ferent combinations? Their very high success rates on day one for the later combinations suggested that they might have formed the relative class concept of oddity. Experiment II was undertaken to explore this possibility further. Percent Success TW an 2a 2 oO r-) i~ Qo 2 —E 3 z 0 0 0 24 6 8 1012 Session Number 3 4 100 0 6 8 1012 0 8 1012 0 6 8 1012 0 2 4 6 Fig. 3. Mean success rates and standard errors for each of the six combina- tions (n=3). Dotted lines indicate the total number of responses. Note the above-random performance on all combinations, despite wide variation. 100 80 60 40 Percent Success 20 Original Rearranged Combinations Combinations Fig. 4. Mean success rates and standard errors for the three octopuses with five original combinations (n=13) and with eight new combinations of familiar stimuli (n=15). 78 AMER. MALAC. BULL. 9(1) (1991) EXPERIMENT II Performances of the three octopuses were once again statistically indistinguishable (contingency table for three sub- jects versus correct or incorrect response: x? = 0.132, d.f. = 2, P > 0.90). Results were therefore pooled. The performance of the octopuses in choosing the odd object showed no clear improvement over time (Fig. 6). When mean success rates are separated by week (or by natural breaks), performances appeared slightly, but not significantly, better than completely random (33%). The means by week were 38, 36, 36, 47, and 40% (n=18). The periodicity found in this learning curve (Fig. 6) was unexpected. It did not correspond to the six session work weeks and was not evident in all three individuals’ per- formances (Fig. 7). 100 80 aa [] Day1 HM Day5 Percent Success 20 1 2 3 4 5 6 Combination Fig. 5. Mean success rates and standard errors for days one and five for each of the six sequential combinations. 100 Percent Success Number of Grabs 0 6 12 18 24 30 Session Number Fig. 6. Mean success rates (squares) with standard errors (n=3) and total numbers of responses (diamonds) across sessions in Experiment IT when every combination presented was novel. DISCUSSION Octopus bimaculoides, a small octopus from the southern California coast, is a solitary, nocturnal predator, feeding primarily on gastropods but also on_ bivalves, polychaete worms, fishes, and crabs (Forsythe et al., 1984). It lives in dens or burrows and interacts rarely with other octopuses (Lang, 1990). Generations do not overlap, which precludes the level of social learning available to even the most solitary of mammals. Laboratory-reared octopuses have the reputation of being slower to learn than those caught wild, perhaps an indication that learning takes place normally in their environment. The particular tasks in these experiments do not relate directly to any activities known to be performed in the wild. However, they are comparable to experimental tests used to assess cognitive abilities in a wide range of species (Thomas, 1980). In Experiment I, the octopuses’ mastery of the simple task of learning to choose a particular shell after repeated presentations of the same combination is clear evidence of associative learning. This result is consistent with findings for Octopus vulgaris (Wells, 1978) as well as other in- vertebrates (Corning et al., 1976). Transfer of learning was also evident from _per- formances on new combinations of familiar stimuli, an in- dication that their learning was not tightly dependent on con- text. Whether they based their choices on learned positive identities, learned negative identities, or both, cannot be deter- mined from this experiment. However, response patterns sug- gest that they did in fact learn both. The octopuses normally sat in one of the top front corners of the tanks during trials. During the first experiment, they quickly stopped swimming to the far side of the tank (approximately an extra 20 cm) in order to grab the far stimulus. Therefore, they had to decide not only which stimulus to grab, but which stimuli not to grab, since the two nearby shells might both be negatives. They were highly successful at learning these discriminations, an indication that they had, in fact, learned both positive and negative shell identities. That learning was retained over time was shown in the transfer of learning trials. The octopuses’ success rate with the first combination was 80% , when retested after five weeks of experience with other combinations. Informal observations suggest that these animals might show retention even for months. Learning set formation, or learning to learn, implies that something beyond recognition of objects has been learned. It could be only to attend carefully to relevant stimuli. Or, it could involve remembering previous choices and out- comes so as to arrive more quickly at correct solutions. Although learning set formation has not been shown previous- ly in invertebrates, the related task of learning reversals has been mastered by isopods (Morrow and Smithson, 1969) and BOAL: LEARNING IN OCTOPUS Fe) OCTOPUS 1 100 100 Success Rate $ 100 Session Fig. 7. Mean success rates across sessions for each subject in Experiment II when every combination presented was novel. by octopuses (Mackintosh, 1965). Now that learning set for- mation has been indicated in octopuses, experiments with a longer learning set (many more combinations) could make comparisons with vertebrate species possible. The results from the Experiment II showed no evidence of octopuses learning the concept of oddity. The results were surprising, after the octopuses’ promising first day per- formances towards the end of the learning set. It is possible that more time was needed; the first experiment took place across three months as opposed to only five weeks for the second experiment. Or, it could be that repeated learning of exemplars, as in the first experiment, could succeed where the repeated novel presentations did not. It also could be that the curtain used during the second experiment prevented cu- ing that biased the first experiment. The apparent periodicity in the learning curve for this experiment has no clear explanation. The learning curves for the first experiment also showed temporal variability in per- formance, especially prior to mastery of the task, but no ap- parent periodicities. Sanders (1977) has documented octopus learning curves to be multiphasic, corresponding to possible transitions between short- and long-term memory processes. However, his experiments were examining retention of a learned task across hours as opposed to performances of new tasks across days. I expect that the pattern observed in this experiment was simply an artifact of the small sample size and short duration of the experiment. There were a number of problems in Experiment I. First, response rates were low, perhaps because, with so many trials each day, the smallest practical food rewards still add- ed up to more than their normal daily intake levels. Another related explanation is that older, laboratory-habituated animals seem to have smaller appetites and to be less responsive in general. Second, choices during the first trial of each train- ing session were significantly less accurate, as compared to subsequent trials (x? = 22.884, d.f. = 7, P < 0.01). Results from this experiment were therefore conservative. Third, animals were significantly more likely to grab stimuli in the front half of the tank. The experimental design was balanced, however, for front and back placement of the odd object. These three problems appeared to be addressed successfully in Experiment II by switching to a linear presentation, pro- viding simple reinforcement trials at the beginning of each session, limiting trials to ten per day, and using younger, freshly caught animals. Clearly, octopuses are capable of some forms of com- plex learning. While it remains to be seen if these octopuses can master a relative class concept such as oddity, these ex- ploratory experiments suggest that the abilities underlying the formation of a learning set have evolved in an invertebrate, as well as in vertebrates (Pearce, 1987). They provide fur- ther evidence for convergences in function despite divergences in physiology between invertebrates and vertebrates (Packard, 1972; Corning et al., 1976). ACKNOWLEDGMENTS I would like to thank W. M. Kier for sharing his expertise and for use of his laboratory, R. H. Wiley for his encouragement and thoughtful critiques, and A. Smith for his camaraderie in times of aquatic crisis. I would also like to thank my anonymous reviewers for their helpful comments. This research was supported by the University of North Carolina Curriculum in Ecology and Sigma Xi Grants-in-Aid of Research. 80 AMER. MALAC. BULL. 9(1) (1991) LITERATURE CITED Corning, W. C., J. A. Dyal and R. Lahue. 1976. Intelligence: an invertebrate perspective. In: Evolution, Brain and Behavior: Persistent Problems. R. B. Masterson, W. Hodos, H. Jerison, eds. pp. 215-263. John Wiley & Sons, New York. Essock-Vitale, S. and R. M. Seyfarth. 1986. Intelligence and social cogni- tion. In: Primate Societies. B. B. Smuts, D. L. Cheney, R. M. Seyfarth, R. W. Wrangham, T. T. Struhsaker, eds. pp. 452-461. University of Chicago Press, Chicago. Forsythe, J. W., R. H. DeRusha and R. T. Hanlon. 1984. Notes on the laboratory culture of Octopus bimaculoides, the California mudflat octopus. American Malacological Bulletin 2:92. Humphrey, N. K. 1976. The social function of intellect. Jn: Growing Points in Ethology, P. P. B. Bateson and R. A. Hinde, eds. pp. 303-317. Cambridge University Press, New York. Lang, M. A. 1990. Population dynamics and life history of Octopus bimaculoides. Abstract. American Malacological Union, 56th An- nual Meeting Program, p. 47. Lombardi, C. M., C. C. Fachinelli and J. D. Delius. 1984. Oddity of visual patterns conceptualized by pigeons. Animal Learning and Behavior 12(1):2-6. Mackintosh, N. J. 1965. Discrimination learning in the octopus. Animal Behavior Supplement 1, 129-134. Morrow, J. E. and E. L. Smithson. 1969. Learning sets in an invertebrate. Science 164:850-851. Packard, A. 1972. Cephalopods and fish: the limits of convergence. Biological Reviews 47:241-307. Pearce, J. M. 1987. An Introduction to Animal Cognition. Lawrence Erlbaum Associates, London. 328 pp. Sanders, G. D. 1977. Multiphasic retention performance curves: fear or memory? Symposia of the Zoological Society of London 38:435-445. Shettleworth, S. J. 1984. Learning and behavioral ecology. Jn: Behavioral Ecology: An Evolutionary Approach. J. R. Krebs, N. B. Davies, eds. pp. 170-194. Sinauer, Sunderland, Massachusetts. Thomas, R. K. 1980. Evolution of intelligence: an approach to its assess- ment. Brain, Behavior and Evolution 17:454-472. Thomas, R. K. and L. M. Noble. 1988. Visual and olfactory oddity learn- ing in rats: what evidence is needed to show conceptual behavior? Animal Learning and Behavior 49(3):395-409. Wells, M. J. 1978. Octopus: Physiology and Behavior of an Advanced In- vertebrate. Chapman and Hall, London. 417 pp. Wells, M. J. and J. Z. Young. 1970. Stimulus generalization in the tactile system of Octopus. Journal of Neurobiology 2(1):31-46. Date of manuscript acceptance: 5 November 1990 APPENDIX 1 Combinations in Experiment I Combination |: Noetia ponderosa (Say) (+), Anomia simplex Orbigny (—). Combination 2: Anomia simplex Orbigny (+), Chione cancellata (Linne’) (—). Combination 3: Aequipecten gibbus (Linne’) (+), Mercenaria mercenaria (Linne’) (—). Combination 4: White square plastic grid (3 x 3 x 1 cm) (+), white square flat plastic chip (3 x 4 x .4 cm) (—). Combination 5: white PVC adaptor fitting, threaded by slip, six-sided middle section (2.9 x 4 cm) (+), white PVC pipe section (4.1 cm diam. x 1.6 cm width) (—). Combination 6: Crassostrea virginica (Gmelin) (+), Busycon contrarium (Conrad) (—). APPENDIX 2 Combinations in Experiment II Trials in Experiment II consisted of presenting combinations of three shells, two of one species and one of another. Species were chosen random- ly with the constraint that the two types of shells of a combination must differ on two of three features (texture: 1=smooth, 2=ridged, 3=pointed; type: l=gastropod, 2=bivalve or slipper; color: 1=light, 2=dark, 3 =patterned). Species included were: Aequipectin gibbus (Linne) (2,2,3); Amphi- dromus entobaptus (Dohrn) (1,1,1); Anadara brasiliana (Lamarck) (2,2,1); A. ovalis (Bruguiere) (2,2,1); A. ovalis (Brugiere) (2,2,1); Anomia simplex Orbigny (1,2,1); A. simplex (1,2,1); A. simplex (1,2,2); Argopecten sp. (2,2,1); A. sp. (2,2,2); A. sp. (3,2,3); A. sp. (2,2,3); Arca zebra Swainson (2,2,3); Architectonia nobilis Roding (2,1,3); Babylonia areolata (Link) (1,1,3); Bursa sp. (3,1,3); Chicoreus cichoreum (Gmelin) (3,1,3); Chione cancellata (Linne’) (2,2,3); C. cancellata (2,2,1); C. cancellata (2,2,1); C. cancellata (2,2,1); C. paphia Linne’ (2,2,3); Conus pulcher Lightfoot (1,1,1); Crassostrea virginica (Gmelin) (1,2,2); C. virginica (2,2,1); C. virginica (1,2,2); C. virginica (2,2,1); Crepidula fornicata (Linne) (1,2,1); C. fornicata (1,2,1); C. fornicata (1,2,2); C. fornicata (1,2,1); Dodinia discus (Reeve) (1,2,1); Ficus subintermedia (Orbigny) (1,1,1); Geukensia demissa (Dillwyn) (2,2,2); Helicostyla sp. (1,1,3); Haliotis asinia Linne(1,2,2); Liguus virineus (Linne’) (2,1,3); Lucina pectinata (Gmelin) (2,2,1); Marisa cronuarietis (Linne) (1,1,3); Melongena corona Gmelin (3,1,3); Mercenaria mercenaria (Linne) (1,2,1); Murex fulvescens Sowerby (3,1,1); Natica sp. (1,1,1); N. stellata Chenu (1,1,1); Noetia ponderosa (Say) (2,2,1); N. ponderosa (2,2,1); N. ponderosa (2,2,2); Oliva sayana Ravenel (1,1,1); O. servicea Roding (1,1,3); Phalium granulatum (Born) (1,1,3); Pitar morrhuana (Linsley) (1,2,2); Pleuroplaca sp. cf. glabra (Dunker) (2,1,1); Sinum perspectiuum (Say) (1,2,1); Sisula solidissima (Dillwyn) (1,2,1); Tagelus plebius (Lightfoot) (1,2,1); Tectarius cf. Coronatus valenciennes Gmelin (3,1,3); Telescopium telescopium (Linne’) (1,1,2); Tellina alternata Say (2,2,1); Trachycardium egmontium (Shuttleworth) (2,2,1); TZ. egmontium (2,2,2); Turbo sp. (1,11); T. sp. cf. petholatus Linne’ (1,1,2); Turritella sp. (2,1,3); Vexillum rugosum (Gmelin) (2,1,3). Mating behavior of the freshwater pulmonate snail, Physa gyrina Thomas J. DeWitt Department of Biology, State University of New York, Binghamton, New York 13902-6000, U.S.A. Abstract. The typical mating sequence of freshwater pulmonate snails includes mounting, positioning, preputium eversion, intromission and dismount- ing. Previous studies report that sperm recipient behavior is passive or absent, and fecundity increases with size. I examined the mating sequence, mate rejection and fecundity of Physa gyrina (Say) removed from the Fenway in Boston, Massachusetts. Mate rejection occurred in any of four stereotypical ways: 1) shell swinging; 2) shell jerking; 3) increasing distance to the gonopore; 4) making head/preputium contact. The jerking behavior corresponds to what others described as sham copulation. Shell swinging was similar to parasite escape behavior. Fecundity increased with the size of sperm recipients but decreased with the size of sperm donors. Differences in methodology could explain discrepancies in the interpretation of mating behavior among studies. Many studies have been conducted on the reproduc- tive biology of freshwater pulmonate snails (Basom- matophora; Geraerts and Joosse, 1984). Several behavioral studies, however, have reported conflicting observations within and among closely related taxa. This variation could be due to genetic variation and phenotypic plasticity, as well as differences in research methodology. The most complete information on basommatophoran reproductive behavior is for the sister families Lymnaeidae and Physidae. Physid snails have been studied recently in terms of predator induced life history shifts (Crowl, 1990; Crowl and Covich, 1990), macrophyte herbivory (Sheldon, 1987), sperm storage (Wethington and Dillon, 1991), parasite defense strategies (Townsend and McCarthy, 1980) and mate choice patterns (T. J. DeWitt, unpub. data). Physa (=Physella) gyrina (Say) is widely distributed (Clarke, 1981) in lentic environments, especially where pollu- tion or high temperatures exclude other snails (Clampitt, 1970; Harman, 1974). Physid snails, like lymnaeids, are herma- phrodites that usually cross fertilize but are capable of self fertilization (Colton, 1918; R. M. DeWitt and Sloan, 1959). Although the male genital tract can mature briefly before that of the female, it is unclear as to whether physids are ever functionally protandric (Duncan, 1959). Furthermore, snails can receive sperm regardless of the sexual maturity of female tracts, but can not oviposit (Duncan, 1959). Peak breeding in the field for Physa gyrina is reported to occur from April through June in Iowa (Clampitt, 1970) and Michigan (R. M. DeWitt, 1955), although mature snails collected any time of the year will breed in the laboratory as long as water temperature exceeds 10°C (R. M. DeWitt, 1955). This corresponds to the April to May peak reported for P. fontinalis Linné by Duncan (1959) at a similar latitude. In this paper, I report observances taken on the reproductive behavior of Physa gyrina removed from the field. I examined the typical mating sequence, mate rejection, and fecundity. Conflicting reports on mating behavior within and between pulmonate snail species could be due to research methodology. METHODS On six independent occasions between February and May 1990, 300-500 snails were collected from the same loca- tion in the Fenway in Boston, Massachusetts (NW of Beth Israel Hospital). Snails were collected between 1200 and 1600 hours. Dip nets were used to collect litter from the bottom of the Fenway (1-1.5 m depth). Snails were removed from the litter, by hand, to a bucket containing Fenway water. On all occasions, Physa gyrina made up over 98% of snail fauna [other species, in order of abundance, included Planorbella campanulata (=Helisoma campanulatum) (Say) and Stagnicola elodes (=Lymnaea palustris) (Say)]. Each collec- tion, including all specimens, was brought back to the laboratory and placed in an aerated, filtered 38 / aquarium with approximately 20 / of water (half from the collection site, the other half was conditioned tap water). Population densities were greater than, but similar to those observed in the field during collection. Snails were kept at 20°C near a natural light source. They were fed canned or boiled vegetables (carrot, pea, chick pea). This diet was supple- mented by unidentified algae in the aquaria. Preliminary observations of snails were made to define criteria for successful versus failed copulation. Criteria I used to define successful mating were that 1) the sperm donor was in place at the shell margin of the recipient; 2) the preputium was placed in the immediate area of the gonopore; 3) the seminal vesicle within the preputium contained a milky American Malacological Bulletin, Vol. 9(1) (1991):81-84 81 82 AMER. MALAC. BULL. 9(1) (1991) substance; 4) no agonistic behavior occurred within the first 30 seconds of mating. TREATMENT 1 After approximately 18 hours from the time of collec- tion, interacting Physa gyrina pairs were observed. As soon as the nature of each interaction was determined, pairs were removed and measured from the shell apex to the most distal portion of the shell margin. Sperm recipients from successful matings were removed to mason jars for a week and the number of eggs they laid was recorded. This continued until 12 successful matings were observed on each of the six oc- casions (72 observations total). TREATMENT 2 An additional 90 Physa gyrina pairs were observed as in treatment one. In treatment two, however, pairs were not interrupted, but allowed to finish their interactions. RESULTS The sizes of sperm donors and recipients in treatment one were 7.32 + 0.96 and 7.90 + 1.12 (mean + sd), respec- tively. The generalized mating sequence for Physa gyrina was similar to that described for Lymnaea stagnalis (Linne’ ) (van Duivenboden and Maat, 1988), a closely related taxon, ex- cept that it occurred more rapidly. The sequence of successful mating included mounting (mate selection?), positioning, eversion of the preputium, intromission and dismounting. This process was sometimes interrupted by mate rejection but lasted approximately 10-20 minutes when successful (Fig. 1). Rejection of mates was observed to occur in four (ye ee | | Mounting | al) \ / / \ / \e/: | Positioning < ce Mate rejection 3 F Shell swinging | Eversion of preputium Shell jerking AL Increasing distance v to gonopore Head/preputium contact wy, V [ ae | Intromission / Vv | Dismounting | Fig. 1. Mating sequence of Physa gyrina. Thickness of arrows approximates the frequency that path is followed. Fig. 2. Stereotypical mate rejection behavior by sperm recipients. a) shell swinging, b) shell jerking, c) increasing distance to gonopore, d) head/ preputium contact. stereotypical ways (Fig. 2). In order of decreasing frequen- cy, intended sperm recipients performed 1) shell swinging, 2) shell jerking, 3) shell positioning such that sperm donors could not reach the gonopore, 4) head to preputium contact (“‘biting’’). Shell swinging was characterized by a 180-270° left and right twisting of the shell about the stationary foot. Shell jerk- ing was rapid contraction of the shell toward the foot, with gradual release. This was generally repeated several times until the preputium was withdrawn. Shell positioning by in- tended sperm recipients occurred such that the distance from their shell margin to their gonopore was increased, effectively placing the gonopore out of reach of the sperm donor’s preputium. I could not observe whether head to preputium contact included radular action but think it likely given that physid snails rarely cease radular movement and that the reac- tion from sperm donors was always immediate, violent withdrawal of the preputium, usually followed by immediate dismount. The first three forms of agonistic sperm recipient behavior generally resulted in two or three re-attempts at mating by sperm donors before they dismounted. I commonly observed chain copulations of up to five snails, acting alternately as sperm donors and recipients. Furthermore, triads were observed on three occasions while collecting in the field. Polyandrous copulation, which has only been clearly described (among basommatophorans) for Physa fontinalis (Duncan, 1959), was not observed. The range of snails observed to oviposit in this study was 6.3-10.9 mm. Larger sperm recipients proved significantly more fecund (R? for log transformed data = 0.58, p,j. = 0.0001). However, the size of sperm donors was negatively correlated with sperm recipient (log transformed) fecundity (pearson correlated coefficient = -0.44, p,,. < 0.01), even DeWITT: MATING BEHAVIOR OF PHYSA GYRINA 83 when the size of sperm recipients was controlled for (partial correlation coefficient = -0.34, p,,. = 0.025). DISCUSSION Describing the mating behavior of freshwater gastropods can be accomplished by field or laboratory studies. The former may be made difficult by turbid water, detritus, invasiveness of techniques used to locate snails and low pro- bability of finding mating pairs. For these reasons, laboratory investigation is usually employed. This is acceptable if it is reasonable to believe organisms are behaving as they would in the field. Observations from field studies generally con- cur with laboratory findings (e.g. R. M. DeWitt, 1954b). However, it is likely that the laboratory techniques employed affect the behavior of animals being studied, and account for some of the variation in reported behavior. R. M. DeWitt (1954a) reported 7 mm as the minimum size at which Physa gyrina would oviposit. However, this research of P. gyrina, and studies on other physids, suggest that the size at first oviposition varies among populations and species (Duncan, 1959; Clampitt, 1970; McMahon, 1975; Crowl, 1990). This variation could be due to physical variables such as temperature (McMahon, 1975) and plastic life history strategies. P. virgata (Gould) (Crowl, 1989; Crow] and Covich, 1989) is induced, by the presence of crayfish (predators), to delay reproduction until larger size is attained. I have observed similarly varied life history patterns in P. heterostropha Say in streams with and without goldfish in New York (unpub. data). R. M. DeWitt (1954a) reported a lack of chain copula- tions in Physa gyrina. However, I observed chain copulation in the laboratory and field. Chain copulations have been reported for laboratory populations of several other freshwater pulmonates (e.g. Barraud, 1957; Duncan, 1959). My observation of increasing fecundity as a function of snail size concurs with work on several freshwater pulmonates (e.g. R. M. DeWitt, 1954b; de Wit, 1955; Hunter, 1975; McMahon, 1975). However, this is the first report that the size of sperm donors is negatively correlated with the fecundity of sperm recipients. It will be interesting to see if studies employing different methods obtain similar results. The data from the present study do not allow me to address why larger sperm donors could have suppressed sperm recipi- ent fecundity. Many studies report that females are passive during mating (R. M. DeWitt, 1954a; Duncan, 1959). Van Duiven- boden and ter Maat (1988) state ‘‘characteristic female mating behavior is absent’’. I observed several stereotypical female behavior patterns which I interpret to be mate rejection (Fig. 2). Two of these correspond to behavior reported by other investigators. Shell swinging is similar to weak parasite avoidance reported for Physa fontinalis in response to chemicals from predatory leeches (Townsend and McCarthy, 1980). Shell jerking is similar to the description of ‘‘false coupling’’ given by Barraud (1957) and ‘‘sham copulation’’ given by van Duivenboden and ter Maat (1988) for Lymnaea stagnalis. Barraud (1957) describes false coupling as when the female gonopore remains unoccupied while the preputium is bent around her shell margin, despite contorted retractions of the female. Van Duivenboden and ter Maat describe sham copulation as ‘‘characterized by strong withdrawal of the forepart of the female, after which she relaxes again. ..while the preputium remains in place (under the female shell margin — not in the gonopore).’’ Barraud (1957) report that males may move away without copulation after this event and that copulation was seldom successful. Van Duivenboden and ter Maat (1988) report that nearly all copulations were eventual- ly successful but ‘‘sham copulations occur frequently (= 50% of the pairs)”’. I believe the phrase ‘‘sham copulation’’ could be a misnomer because the interpretation as mate rejection is strongly supported in this study and that of Barraud (1957). Natural selection should provide strong selection pressure for successful copulatory mechanisms between conspecifics (Ridley, 1983) unless there is a benefit to unsuccessful mating behavior. Let us examine potential costs and benefits of sham (or unsuccessful) copulation. The only benefit I could associate with sham copulation would be the ethological argu- ment that the practice is valuable for future ‘‘true’’ copula- tion. Noland and Carriker (1946) show that prior experience increases copulatory success in Lymnaea stagnalis. However, later studies by Barraud (1957) and van Duivenboden and ter Maat (1988) refute this. Costs of ‘‘sham copulation’ could go beyond time and energy expenditure. Predation risk may be greater for animals in a pair versus solitary animals [e.g. Gammarus pulex (Linne); Ward, 1986]. Thus, the value of sham copulation is dubious. I suspect that the long period of isolation used in many studies made sperm donors less willing to give up copulation attempts, thereby reducing the occurrence of successful mate rejection and delaying or preventing straightforward copulation. Researchers often use the isolate-unite technique (e.g. Noland and Carriker, 1946; Barraud, 1957; Duncan, 1959; van Duivenboden and ter Maat, 1988) to study the reproduc- tive behavior of gastropods. This protocol is effective for stimulating copulation, but could lead to overlooking phenomena of more realistic situations for field populations (e.g. mate choice). The protocol could introduce artifact problems as well [e.g. sperm deprivation during isolation evenutally leads to initiation of self fertilization (van Duiven- boden, 1983), thereby reducing receptivity to mating]. It is probable that a more generalized view of mating behavior in gastropods can be obtained by supplementing the isolate-unite technique with studies using multiple wild-caught snails in as close conditions to their natural environment as 84 AMER. MALAC. BULL. 9(1) (1991) possible (i.e. population density, water, temperature, light, eic.). The use of focal pair sampling (perhaps videotaping) would also be more likely to yield valuable details on mating over multiple pair observation (e.g. van Duivenboden and ter Maat, 1988). The suggested techniques allow elucidation of structured mating, mate rejection and other phenomena oc- curring in wild populations. ACKNOWLEDGMENTS This research was made possible by academic support from Dr. Fred A. Wasserman, to whom I am greatly obliged. LITERATURE CITED Barraud, E. M. 1957. The copulatory behaviour of the freshwater snail (Linnaea stagnalis L.). British Journal of Animal Behaviour 5:55-59. Colton, H. S. 1918. Self fertilization in the air-breathing pond snails. Biological Bulletin of Woods Hole 35:48-49. Clampitt, P. T. 1970. Comparative ecology of the snails, Physa gyrina and P. integra (Basommatophora: Physidae). Malacologia 10:113-151. Clarke, A. H. 1981. The Freshwater Mollusks of Canada. National Museums of Canada, Ottawa, Canada. 446 pp. Crowl, T. A. 1990. Life-history strategies of a freshwater snail in response to stream performance and predation: balancing conflicting demands. Oecologia 84:238-243. Crowl, T. A. and A. P. Covich. 1990. Predator-induced life-history shifts in a freshwater snail: A chemically mediated and phenotypically plastic response. Science 247:949-95]1. de Wit, W. F. 1955. The life cycle and some other biological details of the freshwater snail Physa fontinalis (L.). Basteria 19:36-73. DeWitt, R. M. 1954a. Reproduction, embryonic development and growth in the pond snail, Physa gyrina Say. Transactions of the American Microscopical Society 73:124-137. DeWitt, R. M. 1954b. Reproductive capacity in a pulmonate snail (Physa gyrina Say). American Naturalist 88:159-164. DeWitt, R. M. 1955. The ecology and life history of the pond snail, Physa gyrina. Ecology 36:40-44. DeWitt, R. M. and W. C. Sloan. 1959. Reproduction in Physa pomilia and Helisoma duryi. Animal Behaviour 7-81-84. Duncan, C. J. 1959. The life cycle and ecology of the freshwater snail Physa fontinalis (L.). Journal of Animal Ecology 28:97-117. Geraerts, W. P. M. and J. Joosse. 1984. Freshwater snails (Basommatophora). In: The Mollusca, Volume 7: Reproduction, K. M. Wilbur, ed. pp. 141-207. Academic Press, New York. Harman, W. N. 1974. Snails (Mollusca: Gastropoda). In: Pollution Ecology of Freshwater Invertebrates, C. W. Hart, Jr. and S. L. H. Fuller, eds. pp. 274-312. Academic Press, New York. Hunter, R. D. 1975. Growth, fecundity, and bioenergetics in three popula- tions of Lymnaea palustris in upstate New York. Ecology 56:50-63. McMahon, R. F. 1975. Effects of artificially elevated water temperatures on the growth, reproduction and life cycle of a natural population of Physa virgata Gould. Ecology 56:1167-1175. Noland, L. E. and M. R. Carriker. 1946. Observations on the biology of the snail Lymnaea stagnalis appressa Say during twenty generations in the laboratory culture. American Midland Naturalist 36:467-493. Ridley, M. 1983. The Explanation of Organic Diversity: the Comparative Method and Adaptations for Mating. Clarendon Press, Oxford, United Kingdom. 272 pp. Sheldon, S. P. 1987. The effects of herbivorous snails on submerged macrophyte communities in Minnesota lakes. Ecology 68:1920-1931. Townsend, C. R. and T. K. McCarthy. 1980. On the defence strategy of Physa fontinalis (L.), a freshwater pulmonate snail. Oecologia 46:75-79. van Duivenboden, Y. A. 1983. Transfer of semen accelerates the onset of egg-laying in female copulants of the hermaphrodite freshwater snail, Lymnaea stagnalis. International Journal of Invertebrate Reproduc- tion 6:249-257. van Duivenboden, Y. A. and A. ter Maat. 1988. Mating behaviour of Lymnaea stagnalis. Malacologia 28:53-64. Ward, P. I. 1986. A comparative field study of the breeding behaviour of a stream and a pond population of Gammarus pulex (Amphipoda). Oikos 46:29-36. Wethington, A. R. and R. T. Dillon, Jr. 1991. Sperm storage and multiple insemination in a natural population of the freshwater snail, Physa heterostropha. American Malacological Bulletin 9(1):99-102. Date of manuscript acceptance: 22 January 1991 Reproductive patterns and seasonal occurrence of the Sea Hare Aplysia brasiliana Rang (Gastropoda, Opisthobranchia) at South Padre Island, Texas Ned E. Strenth' and James E. Blankenship? ‘Department of Biology, Angelo State University, San Angelo, Texas 76909, U.S.A. 2The Marine Biomedical Institute, The University of Texas Medical Branch, Galveston, Texas 77550, U. S. A. Abstract. Monthly collections of Aplysia brasiliana Rang were made at South Padre Island, Texas over an 18 month period. Seasonal changes in distribu- tion of weight classes support the existence of a maximum life cycle of approximately 1 year. Large numbers of juveniles in spring collections are produced by the reproductive activities of overwintering adults or possibly recruitment of larvae from other areas. The spring recruitment is followed by the disap- pearance from the population by large overwintering specimens. Spring juveniles increase in size and weight by late summer. Early fall collections yielded no specimens. A minor period of secondary recruitment can occur during late fall or early winter. Winter collections, when successful, yielded reduced numbers of large adults. A number of investigations conducted on naturally oc- curring aplysiid populations have centered upon duration of life cycles, seasonal changes in size or weight, reproductive activities, and recruitment into the population by recently metamorphosed juveniles (Miller, 1960; Carefoot, 1967; Usuki, 1970; Audesirk, 1979; Sarver, 1979; Gev et al., 1984). Following a review of these studies, Carefoot (1987) concluded that, while considerable variation does exist in some species, most sea hares do, in fact, exhibit an annual life cycle. This conclusion is consistent with current information for the sea hare Aplysia brasiliana, which is common to the Gulf of Mexico (Strenth and Blankenship, 1978a). For Florida populations of A. brasiliana, both Krakauer (1969) and Hamilton et al. (1982) concluded that this species exhibits a life cycle of approximately one year. Krakauer’s (1969) con- clusions relative to A. brasiliana (as A. willcoxi Heilprin) were, however, based upon only a one year series of collec- tions of relatively small sample sizes. The study by Hamilton et al. (1982) was based upon large sample sizes but unfor- tunately was limited to the months of March through June. Krakauer (1969) also characterized the life cycle of A. brasiliana in Florida as having *‘two waves of settling’’ with a major “‘spawning period’’ in late March and early April. Following several years of preliminary field work, this study was undertaken in an effort to clarify the life cycle of A. brasiliana. METHODS Collections were made on a continuous monthly basis in the south Laguna Madre at South Padre Island, Texas, at depths of < 1 m, from July, 1977 through December, 1978. All known habitats of Aplysia brasiliana were established dur- ing preliminary field work in 1975 and 1976. While each of these habitats was surveyed during each trip, collection tech- niques and success varied with the season. Winter and spring collections were most productive during early morning low tides. These low tides trapped nocturnally active specimens in shallow grass flats which are common along the western coastal margin of the island. Recently metamorphosed juveniles were collected in the spring from artificial habitats of broken concrete, brick and stone, which had been deposited near the east abutment of the old causeway. During summer months, swimming specimens were collected with dip nets at night near the lights of fishing piers located along the western side of South Padre Island. Specimens were also found on rocks of the boat slip at the Coast Guard Station and the channel side of the north jetty. Following collection, specimens were weighed to the nearest gram. RESULTS AND DISCUSSION DURATION OF LIFE CYCLE The results (Fig. 1) of this study support the existence of a maximum life cycle of approximately | year for Aplysia brasiliana at South Padre Island, Texas. Summer collections, such as those of 1977, were generally characterized by the presence of specimens in the 50 to 300 gm weight range. These animals increased in size and weight during the late fall and winter. By March of the following spring (Fig. 1), American Malacological Bulletin, Vol. 9(1) (1991):85-88 85 86 AMER. MALAC. BULL. 9(1) (1991) 60 so Jul, 1977 as Jan, 1978 | 40 Jul, 1978 = N = 69 ee N = 58 ie N =222 fo) fo) FA Z [1 1m ra fo) Z = 60 60 60 5 Aug, 1977 | 45 Feb, 1978 | 40 Aug, 1978 & a j N= 131 as N = 30 _ N = 91 S) cS ° q | mr ° PA me 2 U A 60 60 =e Sep. 1977 ] 4, Mar, 1978 | 40 Sep, 1978 5 20 Bi 20 N = 58 20 Nene = Ps ° mm ra Z [1 rm ° ° 60 60 ce Oct, 1977 Apr, 1978 Oct, 1978 N= 0 N = 97 N= 0 Ss 20 . £ 60 60 2 4s Sue We a3 May, 1978 | 4 Nov, 1978 E 20 AV Aly 2+ DALE Me 126 20 Meee 7 olamll A ii 5 a 6 PATA ws 60 60 60 ag Dec, 1977 | |. Jun, 1978 | 4 Dec, 1978 ay N = 70 = N = 106 am N = 15 gnanens g fe) mit) a fe) rc re) vate were ° 100 200 300 400 500 600 700 ° 100 200 300 400 500 600 700 © 100 200 300 400 500 600 7 8 Individual Specimen Weights in 50 gm Intervals Fig. 1. Numbers of specimens of Aplysia brasiliana in each 50 gm weight class of monthly collections at South Padre Island, Texas from July, 1977 through December, 1978. the population was characterized by specimens in the 250 - 600 gm range. The marked appearance of large numbers of recently recruited juveniles in the April, 1978 collection, as well as moderate numbers of large specimens in the 450 - 700 gm size, clearly revealed the overlap of generations. These very large specimens were absent from the popula- tion by the following month. Similar life cycle lengths have been reported by Carefoot (1967) for A. punctata Cuvier, by Audesirk (1979) for A. californica Cooper, and by Usuki (1970) for A. kurodai (Baba) and A. juliana Quoy and Gaimard. RECRUITMENT The results of this study (Fig. 1) confirm the presence of a major period of recruitment of recently metamorphosed juveniles during the spring of the year. The April, 1978 col- lection clearly revealed the marked appearance of large numbers of specimens in the | - 5O gm weight class. The presence of specimens weighing less than 50 gm in the col- lections of June, July and August of 1978 initially suggested a continuous 5 month period of recruitment. This does not, however, appear to be the case. Based upon field observa- tions made during 1975, 1976 and 1977, the spring recruit- ment of 1978 appeared unusually large. Hundreds of swim- ming specimens were observed in May. Small to medium sized specimens were very abundant on the rocks of the jetty and Coast Guard Station. A marked decline in numbers of specimens was evi- dent by early summer and, despite an extensive collecting effort, the July collection resulted in only 22 specimens. It appears that the unusually large spring recruitment of 1978 was followed by a mid-summer crash in the population. While the June, July and August collections were characterized by specimens under 50 gm, there were no specimens in the 1 - 10 gm range which had characterized the April collection. The small specimens present in the June, July and August collections of 1978 were therefore considered to be spring recruits which were unable to increase in size due to com- petition, rather than recent recruits to the population. Col- lections made during the summer of 1977 (Fig. 1) as well as those of 1976, appear to support this conclusion. These col- lections were characterized by specimens in the 50 - 300 gm range, as well as the absence of smaller specimens in the 1 - 50 gm range. The summer collections of 1977 are therefore considered typical for this species. The presence of one 12 gm specimen as well as several others weighing less than 50 gm in the November, 1977 col- lection (Fig. 1) supports the existence of a minor period of secondary recruitment during the late fall or early winer. The life cycle of Aplysia brasiliana at South Padre Island appears to be characterized by the presence of a major period of spring recruitment followed by a very minor period of secondary recruitment during late fall or early winter. Usuki (1970) reported similar findings for A. kurodai in the Sea of Japan. Aplysia brasiliana appears to exhibit a relatively high reproductive potential during much of its life cycle. Upon cap- ture, specimens were observed to engage readily in copulatory and egg laying activities throughout most of the year. Despite this fact, recruitment is clearly not a continual process. Re- cent life history studies (e.g. Sarver, 1979; Gev et al. , 1984) have established the relationship of the seasonal abundance STRENTH AND BLANKENSHIP: APLYSIA REPRODUCTION PATTERNS 87 of select algal species with the timing of metamorphosis and subsequent recruitment of juveniles into naturally occurring populations of various aplysiid species (see Carefoot, 1987, for review). Mature veliger larvae of Aplysia brasiliana readily metamorphose in the presence of the red alga Callithamnion and, to a lesser degree, Polysiphonia (Strenth and Blanken- ship, 1978b). While Callithamnion can be found throughout most of the year (Sorensen, 1979), it reaches its maximum abundance at South Padre Island during March (Penn, 1974). This seasonal abundance of Callithamnion just precedes the major period of recruitment of juvenile A. brasiliana into the population in April. While many additional environmental factors should be considered, it appears possible that this seasonal peak of Callithamnion could be a major contributing factor in the timing of the spring recruitment of juvenile A. brasiliana at South Padre Island, Texas. SEASONAL OCCURRENCE The disappearance of specimens from otherwise normally occupied habitats during the months of September and October (Fig. 1) appears to be a normal aspect of the life cycle of Aplysia brasiliana at South Padre Island. Krakauer (1969) also failed to collect specimens of this species in Florida during this same time period. Studies on other aplysiids have reported similar decreases in numbers of specimens during the fall of the year. Audesirk (1979:413) reported a ‘‘nearly total disappearance of animals’’ for A. californica during October, November and December. Gev et al. (1984:69) reported that ‘‘The Aplysia season ends in September”’ for A. depilans Gmelin and A. fasciata Poiret along the Mediterranean coast of Israel. Both Audesirk (1979) and Gev et al. (1984) related this noticeable drop in numbers of specimens during the fall of the year to the decline of one year class, which in turn is replaced by the succeeding generation. Results obtained during the course of this study do not support the above premise as it relates to Aplysia brasiliana at South Padre Island. While some minor recruitment could occur during November or December following the absence of collectable fall specimens, there is no major shift in modal weight class during the fall such as that observed for the month of April, 1978 (Fig. 1). In addition, average weights of specimens increased from August to November during both years. The average weight during 1977 increased from 139 gm (N = 131) in August to 220 gm (N = 142) in November. During 1978, the average weight increased from 67 gm (N = 91) in August to 206 gm (N = 26) in November. The demonstrated presence of continual weight increase in specimens from August to November, as well as the absence of large numbers of recently recruited juveniles in the early winter collections, serve to support the existence of a standing population of A. brasiliana which is character- ized by slight to moderate increases in size of individual specimens rather than a population undergoing decline and replacement. A suitable explanation to account for the location of specimens during September and October does not appear forthcoming from observations made during the course of this study. While discounting reproductive migration in general, Carefoot (1987:204) states that the theory of feeding migra- tions is often ‘‘attractive in that it accounts for seasonal gaps in abundance.”’ Neither the theory of reproductive migration nor that of feeding migration appear to provide a feasible ex- planation (see below) to account for the disappearance of Aplysia brasiliana at South Padre Island during the early fall of the year. Aplysia brasiliana is a known burrower (Aspey and Blankenship, 1976). Individual specimens could be under- going continual or prolonged intermittent periods of burrow- ing in the soft substratum of the south Laguna Madre during September and October. It should be emphasized that this hypothesis is conjectural and not supported by field work. It should be noted, however, that decreased foraging activities by burrowed A. brasiliana during September and October could possibly facilitate a rebound in the standing crop of Callithamnion, which in turn could account for the secondary period of recruitment during late fall or early winter. MIGRATION In his review of migration theory as it relates to life cycles of aplysiids, Carefoot (1987:203) states that ‘‘This theory of migration has fallen into disfavour from lack of sup- porting evidence’’. While individual specimens were observed to exhibit localized movements in association with foraging behavior, the current study provides no support for the migra- tion theory as it relates to Aplysia brasiliana at South Padre Island, Texas. This could be due in part to the somewhat con- fined nature of the south Laguna Madre as well as the presence in the lagoon of both adult food (Laurencia and Gracilaria) as well as metamorphosing substrata (Callitham- nion and Polysiphonia) for developing larvae. Juveniles and adults were collected from the exact same habitats during dif- ferent times of the year. While specimens of Aplysia brasiliana are found oc- casionally beached on the Gulf of Mexico side of the island, the nature of the offshore substrate does not appear to favor the attachment of suitable benthic marine algae. Consequently, the immediate offshore environment of South Padre Island appears to offer little if any favorable habitat for completion of all or part of the life cycle of A. brasiliana. While not surveyed in this study, deeper off-shore reefs are known habitats (Tunnell and Chaney, 1970) for A. brasiliana and could provide for variations in life cycles not observed dur- ing the course of this study. 88 AMER. MALAC. BULL. 9(1) (1991) CONCLUSIONS The results of this study are consistent with and sup- port the conclusions of both Krakauer (1969) and Hamilton et al. (1982) that the maximum length of the life cycle of Aplysia brasiliana is approximately one year. This species exhibits a major period of recruitment during the spring followed by a marked loss of large overwintering specimens from the population. A minor period of secondary recruit- ment can occur during late fall or early winter. The reason for the absence of collectable specimens during the early fall currently remains obscure. ACKNOWLEDGMENTS The presentation of the results of this study during the American Malacological Union meeting at Woods Hole was made possible by a Faculty Development Grant to one of us (N.E.S.) from Angelo State University. Field studies were supported by a National Science Foundation grant (BBS 8711368) to one of us (J.E.B.). Appreciation is extended to the faculty and staff of the Pan American University Marine Laboratory at South Padre Island, Texas. This study would not have been possible without their support. A very special thanks goes to Barbara Strenth for preparation of the figure. LITERATURE CITED Aspey, W. P. and J. E. Blankenship. 1976. Aplysia behavioral biology: II. Induced burrowing in swimming A. brasiliana by burrowed con- specifics. Behavioral Biology 17:301-312. Audesirk, T. E. 1979. A field study of growth and reproduction in Aplysia californica. Biological Bulletin 157(3):407-421. Carefoot, T. H. 1967. Studies on a sublittoral population of Aplysia punctata. Journal of the Marine Biological Association of the United Kingdom 47:335-350. Carefoot, T. H. 1987. Aplysia: Its biology and ecology. Annual Review of Oceanography and Marine Biology 25:167-284. Gev, S., Y. Achituv and A. J. Susswein. 1984. Seasonal determinants of the life cycle in two species of Aplysia found in shallow waters along the Mediterranean Coast of Israel. Journal of Experimental Marine Biology and Ecology 74:67-83. Hamilton, P. V., B. J. Russell and H. W. Ambrose. 1982. Some characteristics of a spring incursion of Aplysia brasiliana into shallow water. Malacological Review 15:15-19. Krakauer, J. M. 1969. The ecology of Aplysia willcoxi Heilprin at Cedar Key, Florida. Master’s Thesis, University of Florida, Gainesville, Florida, U.S.A. 87 pp. Miller, M. C. 1960. A note on the life history of Aplysia punctata Cuvier in Manx waters. Proceedings of the Malacological Society of Lon- don 34:165-167. Penn, G. J. 1974. Seasonal periodicity of dominant, benthic, marine macroalgae of the north jetty of Brazos Santiago Pass, Texas, as related to environmental factors. Master’s Thesis, Pan American Univers- ity, Edinburg, Texas, U.S.A. 45 pp. Sarver, D. J. 1979. Recruitment and juvenile survival in the sea hare Aplysia juliana (Gastropoda: Opisthobranchia). Marine Biology 54:353-361. Sorensen, L. 1979. A Guide to The Seaweeds of South Padre Island, Texas. Gorsuch Scarisbrick Publishers. Dubuque, Iowa, U.S.A. 123 pp. Strenth, N. E. and J. E. Blankenship. 1978a. On the valid name of the com- mon Texas and Florida species of Aplysia (Gastropoda, Opistho- branchia). Bulletin of Marine Science 28(2):249-254. Strenth, N. E. and J. E. Blankenship. 1978b. Laboratory culture, meta- morphosis and development of Aplysia brasiliana Rang 1828 (Gastropoda: Opisthobranchia). Veliger 21(1):99-103. Tunnell, J. W. and A. H. Chaney. 1970. A checklist of the mollusks of Seven and One-half Fathom Reef, Northwestern Gulf of Mexico. Contribu- tions in Marine Science 15:193-203. Usuki, I. 1970. Studies on the life history of Aplysiidae and their allies in the Sado district of the Japan Sea. Science Reports of Niigata University, Series D (Biology) 7:91-105. Date of manuscript acceptance: 21 December 1990 Sententia Variation in sense organ design and associated sensory capabilities among closely related molluscs P. V. Hamilton Department of Biology, University of West Florida, Pensacola, Florida 32514, U. S. A. Abstract. Knowledge of ontogenetic and interspecific variation in structural and functional properties of an organ is prerequisite to establishing valid generalizations about the organ’s role within a group. However, for many molluscan sense organs, generalizations have become established with minimal knowledge of such variation. Review of data on the gross structure, optical properties and visual responses of 44 gastropods provides a clear example of this problem. Lens structures range from crystalline to gelatinous, with concomitant differences in refractive index, and degree of structural and optical homogeneity. Refractive index measurements from differential interference microscopy indicate that gastropod lenses could be partly or completely corrected for spherical aberration. Photoreceptor separation distances vary from about 3 to 25 ym, and photoreceptor abundances from about five to 100,000 per eye. Estimates of anatomical resolution vary from 0.25 to 14 degrees, and overlap considerably with values for arthropods and vertebrates. Visual responses range from simple taxis to the ability to detect an object’s orientation. Reviews of data on the structure and function of opisthobranch rhinophores, and on scallop eye structure, also reveal greater variation than is typically appreciated. More attention must be given to variation in molluscan sense organ structure and function in order for this field to develop more fully. The field of molluscan biology has been expanding its horizons over the past quarter century from its earlier em- phasis on studies of taxonomy, shell morphology and anatomy (Solem, 1974), to greater exploration of life histories, ecological relationships, and behavior. Although few scien- tists with formal training in animal behavior specialized in studying molluscs 20 years ago, this is no longer true. Most animal behaviorists are still attracted to the more rapidly mov- ing arthropods and vertebrates, but cephalopods have always presented a unique challenge, and the value of certain gastropods as neuroethological models has attracted numerous neurobiologists and some behaviorists, especially those in- terested in a reductionist approach. A computer search of literature citations dealing with the behavior of molluscs clear- ly demonstrates the rapid expansion of this subdiscipline (Fig. 1). More than 7,600 citations in Biological Abstracts have in- cluded information about molluscan behavior or sensory biology over the past 20 years, and the current rate of in- crease is about 500 citations per year. The usual pattern of development in many biological subdisciplines is for initial studies to be conducted on one or two easily obtained species thought to be typical of a group, and for initial generalizations to be developed for the group based on the findings obtained for these ‘type’ species. Later, as is required for any field to mature, significant efforts must be directed toward consideration of a broad range of species. Information from comparative studies provides opportunities 15 13 bk Zz uw O 11 a Ww a 9 7 be er ees Qe a Faces ey a pl ed 1970 1975 1980 1985 1990 YEAR Fig. 1. Number of publications dealing with the behavior or sensory biology of molluscs, as a percentage of all publications dealing with molluscs, over the past two decades. These data are based on a computer search of Biological Abstracts, and should be considered only an approximation of the general trend. to determine the phylogenetic limits of plasticity in the group, and often leads to revision of generalizations for the group and reconsideration of whether the initially studied species are indeed typical of the group. Despite the substantial body of information already ac- quired on the behavior and sensory biology of molluscs, this American Malacological Bulletin, Vol. 9(1) (1991):89-98 89 90 AMER. MALAC. BULL. 9(1) (1991) knowledge base is limited in several critical ways. Much of ‘he information contained in the 7,600 citations noted above concerns just a handful of species, and the levels of in- trageneric and intrafamilial variation have been hardly ex- amined. In other cases intragroup variation has been examined and found to be considerable, but this does not seem widely appreciated. And finally, some of the ‘type’ species recog- nized by early investigators of molluscan behavior and sensory biology do not seem particularly representative in the con- text of the broader base of knowledge available today. Col- lectively, these factors would limit the validity of generaliza- tions made about the behavior and sensory biology of any animal group. However, this is especially true for molluscs, as they have undergone an extensive adaptive radiation into a wide range of habitats, and they exhibit substantial plasticity in sense organ structure and behavior (Seed, 1983; Audesirk and Audesirk, 1985). It seems, therefore, that substantially more attention needs to be given to the variation that exists within molluscan groups in the areas of behavior and sensory biology in order to resolve these problems and allow this field to achieve a more mature level of development. The goal here is to con- vince the reader of that viewpoint by examining three selected examples involving a range of taxonomic levels. First, eye structure and the behavioral function of vision in the entire Class Gastropoda will be reviewed in some detail. Then, briefer comparisons will be made of the structure and ap- parent behavioral function of a pair of head tentacles termed ‘rhinophores’ in the gastropod Subclass Opisthobranchia, and of the structure of the eyes in the bivalve Family Pectinidae (scallops). GASTROPOD EYES AND VISUALLY MEDIATED BEHAVIOR The primary example involves vision in gastropod molluscs. It has been traditional for zoologists (e.g. Messenger, 1981) to recognize three structural grades of eyes among the Gastropoda: a) the open cup or pit eye, in which the intraocular space is unfilled and not isolated from the surrounding medium by a cornea (e.g. Patella); b) the filled cup eye, in which a gelatinous (and hence low refractive index) material fills the intraocular space, but in which a cornea is lacking (e.g. Haliotis); c) the closed cup eye, in which a soft or hard lens is present in the intraocular space, and in which a cornea is present (e.g. Nerita, Strombus, Littorina). Recognition of three grades of structure in gastropod eyes can describe, at best, but three points within what is a wide range of ocular designs. Furthermore, because two of the three traditionally recognized grades are exhibited only by quite primitive gastropods (e.g. Patella, Haliotis), this three grade scheme cannot reflect adequately the range of varia- tion present in gastropod eyes. Land (1981, 1984) and Cronin (1988) have reviewed cer- tain aspects of eye structure and optical properties in in- vertebrates, or particularly in molluscs, and limited efforts have been made to tabulate data on gastropod eyes (Zunke, 1978; Messenger, 1981; Land, 1984). Although many excellent studies have been made of retinal ultrastructure and neurophysiological responses, these studies have tended to provide incomplete or no information about whole eye struc- ture or, importantly, the eye as a complete optical system. Thus, there are fewer published data than some might suspect on which to base any general conclusions about gastropod vision. A more detailed comparative view of eye structure in this group is provided here (Table 1) by pooling most of the published data, which covers about 32 species, with previously unpublished data on an additional 12 species. Few authors who have studied gastropod eyes have reported retinal surface areas, or photoreceptor separation distances, abundances, and densities. Thus, most of the data in Table 1 are derived from my computations based on il- lustrations and data included in the references cited. Average intraocular diameter is the mean of the major axis (distance from pupil to retina along optic axis) and the minor axis (distance between opposite retinal surfaces along a line per- pendicular to the major axis and mid-way along it). Retinal areas were computed based on an ellipsoidal model, using the major and minor axes described above, with subtraction of that portion of the ellipsoid’s total surface area corre- sponding to the aperture, where no photoreceptors are located. In most cases, photoreceptor separation distances were obtained directly from an author’s measurements or il- lustrations. However, for Pterocera (=Lambis) lambis (Linneé) and Aplysia californica Cooper, the separation distances used here are | ym greater than the indicated photoreceptor diameters; this adjustment corresponds to a minimal thickness of supportive ceils separating adjacent receptors. Receptor knowledge base is limited in several critical ways. Much of the information contained in the 7,600 citations noted above concerns just a handful of species, and the levels of in- trageneric and intrafamilial variation have hardly been ex- amined. In other cases intragroup variation has been examined sides whose length equals the receptor separation distance). A hexagonal model matches the actual spatial positions of photoreceptors seen in ideal sections taken tangential to the inner retinal surface, for a variety of gastropods (e.g. Hamilton et al., 1983). Receptor densities were computed from retinal area and receptor abundance values. For lens sizes, a single value is given for spherical lenses, and the order of the two values listed for oblong lenses represents the lengths parallel and perpendicular to the optic axis, respectively. Shell length was used as a measure of body length, except in species where the shell is clearly reduced or absent. The length HAMILTON: VARIATION IN MOLLUSCAN SENSE ORGANS oI Table 1. Variation in eye structure among selected gastropod molluscs. Taxon Adult Mean Retinal Receptor Length _Intraoc. Area Separ. (mm) Diam.(um) (mm?) _ Dist.(um) PROSOBRANCHIA Haliotis discus Reeve 150 840 1.703 10.0 Turbo castanea Gmelin 16 514 0.555 anf Neritina reclivata (Say) 15 248 0.161 5.7 Littoraria irrorata (Say) 19 233 0.109 4.4 Littorina littorea (Linné) 22 143 0.052 12.0 Tectarius muricatus (Linne) 19 178 0.074 4.5 Strombus luhuanus Linne 40 1184 3.025 6.0 Pterocera lambis (Linne) 90 1023 2.893 12.0 Lioplax pilsbryi Walker 28 220 0.121 5.5 Pomacea paludosa (Say) 50 561 0.775 4.0 Elimia curvicostata (Reeve) 15 103 0.028 4.5 Marginella sp. 9 120 0.031 6.5 Nassarius vibex (Say) 13 197 0.085 6.9 Melongena corona (Gmelin) 76 273 0.166 7.9 OPISTHOBRANCHIA Bulla gouldiana Pilsbry 50 327 0.285 12.5 Aplysia brasiliana Rang 178 507 0.613 16.0 A. californica Cooper 216 471 0.572 16.0 Hermissenda crassicornis — — — — (Eschscholtz) Tritonia diomedea (Bergh) _ — _— — Mean for 16 nudibranchs _ = — = PULMONATA Melampus bidentatus Say 13 101 0.018 7.4 Lymnaea stagnalis (Linne) 28 140 0.050 10.0 Biomphalaria glabrata (Say) 17 212 0.121 14.8 Strophocheilus sp. 120 350 0.338 23.8 Euglandina rosea (Ferussac) 63 177 0.083 5.0 Helix aspersa Miller 36 231 0.134 6.3 Succinea putris (Linne) 17 97 0.026 14.7 Limax flavus Linne 88 209 0.132 15.0 Agriolimax reticulatus Muller 43 94 0.013 20.0 Receptor Receptor Lens Reference/Source Abund. Density Diam. (mm?) (um) 19665 11550 n/a Tonosaki (1967) 89840 161980 385+ a 5735 35540 235+ a 6565 60470 170 Hamilton et al. (1983) 415 8030 110 Newell (1965) 4210 57030 141 a 97020 32080 732 Gillary (1974), Gillary and Gillary (1979) 23200 8020 1023 Prince (1955) 4610 38210 165x150 a 55940 72170 440x322 a 1600 57030 112x104 a 835 27330 102 a 2050 24250 150x165 a 3075 18510 215x175 a 2110 7390 300x225 Jacklet and Colquhoun (1983) 2765 4510 388x297 a 2585 4520 450x400 Jacklet and Geronimo (1971), Herman and Strumwasser (1984) 5 — 35x50 Stensaas et al. (1969) s] — 150x130 Chase (1974) 8 — 57 Hughes (1970) 375 21210 63xl00 a 585 11610 110x100 — Stoll (1973) 640 5290 163x131 Schall and Baptista (1990) 690 2040 291x240 Oswaldo-Cruz and Bernardes (1982) 3825 46210 175x150 a 3890 29120 225x200 Eakin and Brandenburger (1975), Brandenburger (1975) 145 5490 60x70 Zunke (1978) 680 5160 145x175 Kataoka (1975, 1977) 40 3150 87x67 Newell and Newell (1968) + = Additional lens protrusion through aperture. a = Based on author’s previously unpublished observations. values used were those stated in the references indicated in Table 1 or, when authors failed to indicate the sizes of the animals studied, those given as average adult sizes in ap- propriate basic references. Starting at the largest scale, one can consider the relative sizes of the eyes in adults of various species. Figure 2 shows the relationship between the mean diameter of the intraocular space and the body length of adults for 26 species distributed among three subclasses. Two points are clear from this analysis. First, the size of the intraocular space varies fairly widely among gastropods, with Strombus luhuanus Linne having an average intraocular diameter more than 12 times greater than Agriolimax reticulatus Miller. Second, average intraocular diameter is significantly correlated with body length when all three subclasses are pooled together (r=0.43, P=0.030), and for the subclasses Prosobranchia (n=14, r=0.62, P=0.019) and Pulmonata (n=9, r=0.75, P=0.019) when analyzed separately. Opisthobranchs were not analyzed separately because reasonable measures of in- traocular diameter are available for only three species. Gastropod eyes can be positioned within a substantial peri- optic sinus, as in littorinids (Newell, 1965; Hamilton ef al. , 1983), or they can be closely surrounded by connective tissue, as in Aplysia californica (Herman and Strumwasser, 1984). As would be expected, lens size (as measured by area in mid-saggital section) is highly correlated with average in- traocular diameter (r=0.87, P< 0.0001) and the amount of o2 AMER. MALAC. BULL. 9(1) (1991) Aplysia californica® r=0.43, P=0.03 200 Aplysia brasiliana @ € £ 150 Haliotis discus @ ac kK oO @ Strophocheilus Zz W 1004 Limax flavus - ® ar e@fielongena Pterocera lambis 5 corona < @Pomacea paludosa @ 50 ® ——w Strombus luhuanus @ gi @ Turbo castanea cern nk — 0 200 400 600 800 1000 1200 MEAN INTRAOCULAR DIAMETER (um) Fig. 2. Relationship between the mean diameter of the intraocular space and the adult body length of 26 gastropods from three subclasses. The species associated with the outlying points are identified. Significant positive cor- relations also exist for just the prosobranchs (n=14) and just the pulmonates (n=9). retinal surface area (r=0.97, P<0,0001). The abalone Haliotis discus Reeve lacks a lens, as reportedly do Trochus and Patella (Hilger, 1885), but the eyes of all other gastropods surveyed here have a lens or lens-like material located in the intraocular space. Optical properties of lenses are influenced partly by their shape. Lens shapes vary from spherical or nearly so, as in the littorinids, strombids and Lymnaea, to ellipsoidal, as in various prosobranchs, the aplysiids and most pulmonates (Table 1). Some amount of vitreous material is present between the lens and retina in most species, although Dorsett (1986) suggests this is not the case. Past confusion concern- ing the presence or absence of vitreous material in the eyes of various species could have been due to the susceptibility of vitreous material to dissolution during histological pro- cessing (Charles, 1966; Hamilton et al., 1983). The lens of Aplysia californica was originally il- lustrated as being spherical, and filling the entire intraocular space (Jacklet et al., 1972). This report led logically to the conclusion that its eye could not form a sharp image (Messenger, 1981; Dorsett, 1986) because of insufficient distance for light rays to be brought to focus on the retina (Land, 1981). However, the illustration in Jacklet et al. (1972) is based apparently on an earlier section of the eye in which no aperture was visible, indicating that the section was oblique rather than longitudinal (Jacklet, 1969). A more recent study by Herman and Strumwasser (1984) has shown clearly that the A. californica lens and intraocular space are quite ellip- soidal, as is the case in A. brasiliana Rang. This finding will hopefully stimulate reconsideration of the visual capabilities of this species. In addition to these general shape patterns, the lenses of some gastropods exhibit unique shape variants. The opisthobranch Navanax (=Aglaja) inermis (Cooper) possesses a distinctly bilobed lens, the function of which is unknown (Eskin and Harcombe, 1977). In the eye of Turbo castanea Gmelin, a portion of the lens protrudes through the aperture, and this protrusion of the lens has a shorter radius of curvature than the main body of the lens located within the intraocular space. Because the focal length of a curved refractive surface is directly proportional to its radius of curvature, the protruding portion of lens causes the entire T. castanea lens to have a shorter focal length than it would otherwise have, making focus of light on the retina more like- ly. A wider field of view should also result from this ‘fish- eye’ lens, but this is probably less significant for 7? castanea. Similar lens protrusions were described for Turbo creniferus Kiem. and Nerita polita Linné (Hilger, 1885), and Neritina reclivata (Say) also possesses a distinct lens protrusion. Gastropod lenses vary considerably in hardness. The literature contains numerous pictures of shattered lenses, and it appears that hard lenses are more common than soft lenses in gastropods. However, exact description of lens hardness is difficult, and besides apparent lens hardness (as indicated by degree of shattering when sectioned) seems to be in- fluenced somewhat by the fixative and embedding medium used when processing eyes for histological examination (Hamilton et al. , 1983). Hardness is generally correlated with refractive index, a physical property directly relevant to vi- sion. Authors have frequently noted a concentric pattern of stain uptake by gastropod lenses (e.g. Newell, 1965; Jacklet and Colquhoun, 1983; Gibson, 1984), and have inferred from this that such lenses vary concentrically in composition, and presumably refractive index. Gibson (1984) reported that polyhedral subunits, apparently composed of protein, were packed more densely towards the lens center in //yanassa obsoleta (Say). Refractive index patterns or gradients within lenses or other structures can be measured exactly from frozen sec- tions using differential interference microscopy. Land has suc- cessfully employed this technique with various invertebrates (e.g. Land and Burton, 1979), although data for the lenses of gastropods have not previously been published. In this technique, the distance that an interference fringe is displaced at any given point depends upon the refractive index at that point, as well as light wavelength and specimen thickness, both of which can be controlled. Preliminary data for the marsh periwinkle, Littoraria irrorata (Say), reveal a distinct refractive index gradient within its spherical lens, which is only 170 wm in diameter (Fig. 3). [Reid (1986) moved this species from the genus Littorina.] This gradient closely matches the theoretical gradient required for complete correction of spherical abberation (the curve in Fig. 3; based on Fletcher et al., 1954), a major source of potential image degradation. Preliminary data obtained for the ellipsoidal Aplysia brasiliana lens also indicate a refrac- tive index gradient, ranging from about 1.40 at the periphery to about 1.51 at the core. The Turbo castanea lens varies in HAMILTON: VARIATION IN MOLLUSCAN SENSE ORGANS 93 REFRACTIVE INDEX ee e@ 1.40 \ e — 0.0 0.2 0.4 0.6 0.8 1.0 tas POSITION ON UNIT RADIUS _f CORE PERIPHERY Fig. 3. Refractive index values at different points in the 170 wm diameter spherical lens (f/r=2.71) of the marsh periwinkle, Littoraria irrorata, as measured using differential interference microscopy. The curve indicates the refractive index gradient required for perfect correction of spherical aber- ration in a spherical lens (f/r=2.70) surrounded by a medium whose refrac- tive index is 1.365, which is the average value of cornea and vitreous material. refractive index from 1.40 at the cornea to 1.45 near the lens center, so even soft lenses can have refractive index gradients. The optical significance of a refractive index gradient is unclear for species with relatively soft lenses, and lenses with non-spherical surfaces. Substantial spherical aberration could exist in such eyes, or spherical aberration could be eliminated if corneal surfaces are parabolic, instead of spherical. In ad- dition to reducing spherical aberration, refractive index gra- dients also produce a shorter-than-expected focal length, a result which could be as or more important in some species. Gastropod eyes also vary substantially in retinal and photoreceptor properties (Table 1). Nudibranchs must be con- sidered separately because they clearly exhibit extreme reduc- tion of the eye. For example, Hughes (1970) found that 16 species of nudibranchs had an average of only eight (8) photoreceptors per eye. A survey of the 26 non-nudibranch species on which data are available reveals that retinal areas vary from 0.013 mm? (Agriolimax reticulatus) to 3.025 mm? (Strombus luhuanus), a factor of about 230 times. Most gastropods’ photoreceptors contain few or no melanin pig- ment granules, and so contrast strongly with supportive cells. The photoreceptor counts or estimates reported in the literature have been obtained by various methods. My data on photoreceptor size and spacing have been obtained from sections tangential to the inner retinal surface. As shown in Table 1, adjacent photoreceptors are separated by distances varying from 2.7 wm (Turbo castanea) to 23.8 pm (Strophocheilus sp.) Photoreceptor abundances per eye vary from 40 (Agriolimax reticulatus) to 97,020 (Strombus luhuanus), a factor of about 2,500 times. These photoreceptor abundance values should be viewed only as estimates because at least some gastropods possess two or more receptor types, and these may not be easily distinguished at the light microscope level. Also, photoreceptor densities are known to differ in different regions of the retina in Littoraria irrorata (Hamilton et al. , 1983) and Aplysia californica (Herman and Strumwasser, 1984). Variation in lens or retinal properties can even be substantial within a family or genus. The greatest variation encountered thus far appears to be among the littorinids. When compared with Littorina littorea (Linné), Littoraria irrorata has twice the retinal area and a three times shorter receptor separation distance, which results in almost 16 times more receptors per eye (Table 1). L. irrorata is active in air, while L. littorea appears principally active when submerged in water. That difference in behavioral ecology is probably associated with the substantial difference in eye structure, because the degree to which light is refracted at the cornea depends greatly upon the refractive index of the surrounding medium. Jectarius muricatus (Linné), another littorinid, is active in air. Although it has a retinal area intermediate be- tween the other two littorinids, it has a receptor separation distance and total receptor abundance that are much more similar to the littorinid active in air, L. irrorata. These similarities in retinal design are related presumably to the higher light levels present in air. Unfortunately, the data available for the eye of Littorina scutulata Gould (Mayes and Hermans, 1973) do not allow evaluating the eye as an optical system. Variation in eye structure in littorinids can be explained by differences in behavioral ecology, but the situation is less clear in other cases of variation between closely related gastropods. Within the Strombidae, Strombus luhuanus and Pterocera (=Lambis) lambis have similarly sized eyes, but S. luhuanus has twice as closely spaced receptors and hence four times more receptors than P. lambis. The receptor which Prince (1955) illustrated for P. lambis is 6 wm in diameter; however, he indicated that the average receptor diameter is Il ~m, and his estimate of total receptor density (‘‘something approaching’’ 10,000/mm~?) generally agrees with the estimate in Table 1, which is based on a receptor separation distance of 12 um. Both strombid species are active in shallow water. S. luhuanus is about half the size and travels almost nine times more slowly than Pterocera (=Lambis) lambis (Berg, 1974), yet the S. /uhuanus eye seems capable of resolving greater detail. Berg (1974), who studied ten strombids, noted specifically that S. /uhuanus seemed ‘‘to be able to sense the position’’ of a predatory cone snail, but he did not speculate on the sensory modality involved. No obvious dif- ferences are apparent in eye structure between the two Aplysia species listed, beyond the disagreement mentioned earlier about shape of intraocular space and lens. This is somewhat surprising because the species seem to exhibit basic behavioral differences. A. brasiliana is an excellent swimmer and 1s ac- tive principally at night. In contrast, A. californica apparently does not swim at all, and is diurnally active in the lab and field; whether it may also be nocturnally active under 94 AMER. MALAC. BULL. 9(1) (1991) natural conditions remains unknown (Hamilton, 1986; Leonard and Lukowiak, 1986). Vision in various species can most accurately be com- pared, not by any of the numbers in Table 1, but rather by knowing the resolving power of the eye. Several factors can influence resolution, but one useful estimate of resolution is the angular separation of adjacent receptors relative to the ‘optical center’ of the eye (the posterior or proximal nodal point). Resolution measures or estimates have been published for Littorina littorea (Newell, 1965), Littoraria irrorata (Hamilton et al., 1983), Strombus luhuanus (Land, 1984), Haliotis discus (Land, 1981), Biomphalaria glabrata (Say) (Schall and Baptista, 1989), and Strophocheilus sp. (Oswaldo- Cruz and Bernardes, 1982), and reasonable estimates can be computed for a few other gastropods. These resolution values are given in figure 4, along with comparative data for selected arthropods and vertebrates from Kirschfeld (1976). Clearly the resolving powers possessed by gastropod eyes exhibit a wide range, and they overlap considerably with the resolv- ing powers of arthropod and vertebrate eyes. As Audesirk and Audesirk (1985) suggests, the assump- tion that vision plays only a minor role in gastropod behavior has had an inhibitory influence on careful studies of visually- mediated behavior in this group. It is commonly believed that the structurally-simple eyes of gastropods only mediate sim- ple phototaxis or skototaxis, orientation toward or away from light or dark areas, respectively. However, critical experiments have rarely been done that could allow discrimination between true phototaxis or skototaxis, and form vision, however crude or simple. Hopefully, the existing collection of anecdotes and VERTEBRATES and GASTROPODS ARTHROPODS 0 Frog, Cat 0 @ Strombus luhuanus Lizard, paies a 1 @ Turbo castanea Dragonfly 8 1 Jumping Spider ®@ Littoraria irrorata 2 2 @ Tectarius muricatus Bec Housetly ¢ Bat, Fruitfly @ @ Helix aspersa RESOLUTION (degrees) A A RESOLUTION (degrees) @ Littorina littorea S t Beet 7 g Strophocheilus eae ais 7 2 Biomphalaria glabrata 14he@ Haliotis discus 414 Fig. 4. Measures and estimates of anatomical resolution for nine species of gastropod, with comparative data for selected vertebrates and arthropods. ae 25 Tectarius A o muricatus F o : : Su 20 Littoraria: Helix Ss irrorata: aspersa a : : z a E sd é . 15 Turbo : ci my o castanea-: : r) Ww : : : a : : : = 10 : ae z : : 2) : A = - : M= VERTICAL O 5 : AO (r=0.98, P=0.0169) uw z @ @ = HORIZONTAL, BASAL tl : Yu (r=0.95, P=0.0457) a) E Fru e 4 = HORIZONTAL, CENTRAL 0 (r=0.99, P=0.0044) paca a CD IT (aI CRT (IIE Daa eae ay a Te | 0 1 2 3 4 5 6 ANATOMICAL RESOLUTION (degrees) Fig. 5. Measures and estimates of anatomical resolution, and behavioral ‘detection’ thresholds for oriented responses to three types of targets (ver- tical stripe, centrally-positioned horizontal stripe, basally-positioned horizontal stripe), for four species of gastropod. weakly founded assertions about gastropod vision will even- tually be replaced with carefully obtained data delineating their visual capabilities. Standardized behavioral measures of visual detection have been obtained for four of the gastropod species discussed here (Hamilton and Winter, 1982, 1984), and these compare well with the previously mentioned anatomical resolution measures and estimates (Fig. 5). Also, it is clear that at least Littoraria irrorata can distinguish details of an object’s orien- tation; it preferentially moves toward a vertical black stripe on a white background when presented with horizontal or diagonal stripes having equivalent width and contrast (Hamilton and Winter, 1982). Because the L. irrorata eye is not qualitatively different from that of many other gastropods, it could well be that other species can distinguish such visual detail too. Both Zectarius muricatus and Turbo castanea show some ability to discriminate target orientation (Hamilton and Winter, 1984). An unblocked view of the sky is required for Aplysia brasiliana to maintain its swimming direction, which suggests sensitivity to complex visual cues (Hamilton and Russell, 1982a). In summary, the assortment of optical tricks en- countered in the eyes of various vertebrates, including fish eye lenses and lenses with refractive index gradients, are also found among gastropods. This high degree of variation in gastropod eyes should not be viewed as counter-intuitive. The fishes, for example, are less diverse in habitat and general morphology than the gastropods, yet their eyes exhibit an ex- tensive range of adaptations correlated with habitat and behavioral strategy (Lythgoe, 1980; Fernald, 1988). As Land (1981) has noted, there is no clear break within the range of resolving powers exhibited by the eyes of animals. There real- ly is no such thing as an image forming eye as distinct from HAMILTON: VARIATION IN MOLLUSCAN SENSE ORGANS 95 a non-image forming eye. There are only degrees of need for visual detail among animals, and degrees of image quality provided by eyes. For gastropods and other invertebrates, assumptions about vision, based on vertebrocentric biases, need to be replaced by more hard data and a genuine com- parative perspective. OPISTHOBRANCH RHINOPHORES This is one of the most obvious cases of unappreciated variation in molluscan sensory biology, and a classic exam- ple of why biologists should avoid assigning names to struc- tures that are based on assumed functions. The rhinophores are a pair of tentacles located near the eyes on the dorsal sur- face of the head, and hence in the same location as the pair of tentacles termed the ‘cephalic tentacles’ in prosobranch gastropods. The name ‘rhinophore’ literally means bearer of the nose or nasal sense. This name appears to have been coined by Bergh (1879), who worked principally with the predatory nudibranchs, a group in which the rhinophores do seem involved in distance chemoreception in many species. MacFarland’s (1966) treatise on opisthobranchs beautifully illustrates the structural diversity of nudibranch rhinophores. In many species, rhinophoral sheaths are present, as well as numerous lamellae projecting laterally from a central axis. The greatly increased surface area provided by the lamellae is itself suggestive that nudibranch rhinophores have a chemo- sensory function in this carnivorous group. However, if one takes a more comparative approach, and looks at other opisthobranch taxa, it is clear that other feeding strategies exist, and that the so-called rhinophores can have a variety of different structural features and sen- sory functions. In the herbivorous Aplysia, for example, each rhinophore has a simple gross structure, with no basal sheath and no lamellae. A pigmented groove on the distal half of the rhinophore receives most of the innervation (Fig. 6; from Ronan, 1990). Despite this much simpler gross structure, aplysiid rhinophores have been implicated in chemorecep- tion, mechanoreception (touch, waves, currents) and even photoreception (Frings and Frings, 1965; Jahan-Parwar, 1972; Audesirk, 1975; Chase, 1979; Jacklet, 1980; Hamilton and Russell, 1982b). It could be that variation exists in rhinophore sensory function, even among Aplysia species, since substan- tial variation exists in morphology, activity rhythms and swim- ming behavior among aplysiids (Eales, 1960; Hamilton, 1986). Inadequate appreciation for the variation in rhinophore morphology, feeding strategy and key behavioral traits be- tween nudibranchs and other opisthobranchs has led to the incorrect assumption that the sensory functions associated with the rhinophores of a predatory group (the nudibranchs) automatically apply to other opisthobranch groups. A general under-appreciation for diversity within opisthobranchs could Fig. 6. External view showing the pigmented groove (PG) of the Aplysia brasiliana rhinophore (left), and internal view showing the rhinophoral nerve (RN) and its innervation pattern (right), based on analysis of serial sections (from Ronan, 1990, with permission). A fully extended rhinophore is 15-20 mm long in adults. be partly responsible for this problem, but the literal mean- ing of ‘rhinophore’ is probably the primary factor. Begin- ning ethologists are taught the importance of selecting names for the discrete behaviors included in an ethogram based on spatial and temporal features of the movement pattern in- volved, rather than on the presumed adaptive function of the behavior. For example, the cyclic lateral movement of the siphon which appears when various neogastropods become alerted to a prey’s proximity would be named something like ‘siphon waving’ rather than ‘odor searching’. Anatomists are presumably taught some similar rule, and certainly the ‘issue’ of opisthobranch rhinophores would be less confused today if such a standard had been followed in the 1800’s. The con- fusion that the term rhinophore has caused over the years sug- gests that we could be better off just referring to these opisthobranch structures as cephalic tentacles, as we do for the seemingly homologous paired tentacles of prosobranch gastropods. Admittedly, the term ‘tentacle’ suggests a tactile sensory function, but mechanoreception seems a general pro- perty of virtually all such structures, regardless of what ad- ditional sensory capability they can exhibit. SCALLOP EYES Scallops (family Pectinidae) are swimming bivalve molluscs which possess about 60-100 eyes distributed along the mantle edges. The detailed studies of Land (1965) on Pecten maximus (Linne) revealed the presence of a unique double-layered retina in each eye, and an equally unique 96 AMER. MALAC. BULL. 9(1) (1991) optical system dependent upon a reflector at the rear of the eye, whose surface is described as spherical. Importantly, Land’s analyses show little or no space between the rear of the lens and the retina, or between the retina and the reflec- tive argentea. This is in contrast to Dakin’s (1928) earlier study of Pecten maximus, which reported a space between the lens and retina about 20% of the eye’s length. Butcher (1930) found no space between the lens and retina of Pecten (=Argopecten) gibbus (Linné) but he found a space occupy- ing about 40% of the eye’s length between the retina and argentea of this species. In a preliminary study of the eye of Argopecten irradians (Lamarck), Wooters (1989) found a space having a similar size and location to that reported by Butcher, and an argentea which appeared more parabolic than spherical. Finally, considerable variation exists in the sizes of the eyes possessed by individuals in at least some scallop species. This variation could be due to differences in developmental or regenerative growth, but this appears unstudied with the exception of Butcher’s (1930) work on Pecten (=Argopecten) gibbus. Perhaps the apparent inter- specific differences in scallop eye structure are due to different investigators describing eyes at different growth stages in dif- ferent species. As with gastropod molluscs, any genuine interspecific differences in scallop eye structure could also be associated with differences in behavior or ecology. Argopecten irradians are found in water less than 4 m deep, particularly in beds of the seagrass Thallasia, whose blades reach as high as 0.4 m above the bottom. In contrast, A. gibbus and Pecten maximus are both found in water deeper than 10 m, typically on substrata consisting of sand and shell fragments. Hence, these three scallops differ in the amount of biologically rele- vant visual detail in their environments, as well as in water depth and associated light characteristics. How these habitat differences or behavioral traits could be associated with dif- ferences in eye structure is unknown. Wooters (1969) found that A. irradians orient visually toward grassbeds when released in sand patches adjacent to grassbeds, but little is known of the behavioral role of vision in other scallops. In summary, detailed comparative studies need to be completed to determine exactly what ontogenetic and in- terspecific variation exists in scallop eye structure and op- tical properties. Although the eyes of the various scallop species seem generally similar, the available information on eye structure suggests that some interspecific differences could exist in optical properties, especially as regards focal point position and the degree of spherical aberration. Insufficient attention to interspecific variation can easily lead to confu- sion, and possibly even errors in interpretation. For exam- ple, McReynolds and Gorman (1970) studied the elec- trophysiological properties of the eye of Pecten (=Argopecten) irradians, but included a ‘‘Pecten eye’’ diagram based on Dakin’s (1928) illustration of the eye of Pecten maximus. It seems prudent to exercise caution in extrapolating structural, physiological and behavioral results across scallop species until detailed comparative studies are conducted. CONCLUSION This review of gastropod vision, opisthobranch rhinophores and scallop eyes demonstrates that there exists considerable variation in sensory system design and capabilities among molluscs, even within families or genera. In the case of gastropod eyes, considerable progress has been made in documenting substantial variation, and the problem seems that the observed variation is underappreciated. In the case of scallop eyes, interspecific variation has barely been documented, yet preliminary results encourage detailed com- parative study in the future. The case of opisthobranch rhinophores seems somewhat intermediate; appreciation that the function implied by the name rhinophore might not apply to many opisthobranchs has existed at least since Arey (1918), yet little progress has been made toward obtaining a satisfac- tory comparative view of the behavioral and ecological cor- relates of the variations in structure found in this group. Additional evidence of substantial interspecific varia- tion will likely be obtained eventually for other sense organs, such as the statocyst and osphradium. Statocyst structure has been examined for a modest number of molluscs (Budelmann, 1988), but the possible behavioral-ecological correlates of observed structural diversity have hardly been explored. The literature contains some interesting and apparently paradox- ical cases which beg for analysis. For example, the benthic gastropod Pomacea paludosa (Say) possesses about 3,000 sen- sory cells in its statocyst, while a mere 13 cells are present in the statocyst of Aplysia limacina de Blainville (=A. fasciata Poiret), which is one of the many aplysiids which has been reported to swim as well as crawl on the bottom (Dijkgraaf and Hessels, 1969; Stahlschmidt and Wolff, 1972). Analysis of osphradial variation and its basis across molluscs is com- plicated by the fact that the name osphradium has sometimes been assigned to structures that are clearly not homologous. The presence of this substantial interspecific variation emphasizes the necessity of communicating about particular organisms and sense organs more carefully and, where possi- ble, more exactly. The limits and bases of variation would be more easily understood if more authors would include specific information about the sizes of the actual animals they studied, as well as the relationship between sense organ dimensions and body size in the species. Zunke’s (1978) study of the eye of Succinea includes detailed coverage of onto- genetic variation and, unfortunately, it is unusual in doing so. Failure to consider intraspecific or ontogenetic variation can lead to generalizations just as premature and flawed as those based on too few or atypical species. Autrum (1979) pointed out that it is clearly a bad habit to speak just of ‘‘the HAMILTON: VARIATION IN MOLLUSCAN SENSE ORGANS 97 fly’’ in scientific writings. Likewise, it can be a bad habit to speak just of ‘‘the littorinid’’, ‘‘the rhinophore’’ or “‘the scallop’’, depending on what species, system or capability is being discussed. Finally, as more comparative data become available, it should become possible to discard some of the less infor- mative terms in general use for sensory structures. For ex- ample, there have existed in the invertebrate literature several terms for structurally simple eyes. ‘Pigment cup eyes’ and ‘pinhole eyes’, which both lack lenses, have widely accepted definitions based on structural features (see Land, 1981). However, ‘eyespot’ and ‘ocellus’ both mean ‘‘a very small simple eye formed in invertebrates.’’ As has so often been the case in invertebrate zoology, the emphasis seems to have been on establishing the existence of a difference as com- pared to the vertebrate eye, rather than on recognizing the various invertebrate eyes as being distinct structures worthy of independent study and understanding. As a relatively new field, many gaps exist in our knowledge of molluscan behavior and sensory biology. A clearer understanding would be provided about sensory capabilities within groups if more species were examined, and this would also allow more thorough investigation of the basis of existing variation across species. However, more ef- fort should also be devoted to completing the data sets for species about which some information is already available. In many cases, only data on structure of sense organs or sen- sory tissues are available from which to draw inferences about function; obviously the availability of data from electro- physiological and behavioral studies would be desirable in such instances. For example, scallop eye structure and elec- trophysiological responses have received some study, but there is little understanding of how well they can see or what adap- tive value vision has in nature. A reasonably complete range of data are available for only a few cases, mostly involving species that serve as model systems for neurobiologists. Thus, future researchers in molluscan sensory biology should be encouraged to flesh out existing stories, as well as to expand our understanding of variability within the group by examin- ing as-yet-unstudied species. ACKNOWLEDGMENTS I thank W. Jones for assistance in computing partial surface areas of ellipsoids, S. Miller for drawing figure 6, and C. N. D’Asaro and J. S. Wooters for assistance and critical comments. Portions of this research were sup- ported by NSF grants BNS79-16358 and BNS83-08186, and by an award from UWFP’s research committee. This is contribution 91-4 of the Institute for Coastal and Estuarine Research. LITERATURE CITED Arey, L. B. 1918. The multiple sensory activities of the so-called rhinophores of nudibranchs. American Journal of Physiology 46:526-532. Audesirk, T. E. 1975. Chemoreception in Aplysia californica. 1. Behavioral localization of distance chemoreceptors used in food finding. Behavioral Biology 15:45-55. Audesirk, T. and G. Audesirk. 1985. Behavior of gastropod molluscs. Jn: The Mollusca, Volume 8, A. O. D. Willows, ed., pp. 1-94. Academic Press, New York. Autrum, H. 1979. Preface. In: Handbook of Sensory Biology, Volume VII/6A, H. Autrum, ed. pp. v-vii. Springer-Verlag, Berlin. Berg, J. C. 1974. A comparative ethological study of strombid gastropods. Behaviour 51:274-322. Bergh, R. 1879. On the nudibranchiate gasteropod Mollusca of the north Pacific Ocean, with special reference to those of Alaska. Proceedings of the Academy of Natural Sciences 1879:71-132, 1880:40-127. Brandenburger, J. L. 1975. Two new kinds of retinal cells in the eye of a snail, Helix aspersa. Journal of Ultrastructure Research 50:216-230. Budelmann, B-U. 1988. Morphological diversity of equilibrium receptor systems in aquatic invertebrates. In: Sensory Biology of Aquatic Animals, J. Atema, R. R. Fay, A. N. Popper and W. N. Tavolga, eds. pp. 757-782. Springer-Verlag, New York. Butcher, E. O. 1930. The formation, regeneration, and transplantation of eyes in Pecten (gibbus borealis). Biological Bulletin 59:154-164. Charles, G. H. 1966. Sense organs (less cephalopods). /n: Physiology of Mollusca, Volume II, K. M. Wilbur and C. M. Yonge, eds. pp. 455-521. Academic Press, New York. Chase, R. 1974. The electrophysiology of photoreceptors in the nudibranch mollusc, Tritonia diomedia. Journal of Experimental Biology 60:707-719. Chase, R. 1979. Photic sensitivity of the rhinophore in Aplysia. Canadian Journal of Zoology 57:698-701. Cronin, T. W. 1988. Vision in marine invertebrates. Jn: Sensory Biology of Aquatic Animals, J. Atema, R. R. Fay, A. N. Popper and W. N. Tavolga, eds. pp. 403-418. Springer-Verlag, New York. Dakin, W. J. 1928. The eyes of Pecten, Spondylus, Amussium and allied lamellibranchs, with a short discussion on their evolution. Proceedings of the Royal Society of London, B 103:355-365. Dijkgraaf, S. and H. G. A. Hessels. 1969. Uber bau und funktion der statocyste bei der schnecke Aplysia limacina. Zeitschrift fiir vergleichende Physiologie 62:38-60. Dorsett, D. A. 1986. Brains to cells: the neuroanatomy of selected gastropod species. In: The Mollusca, Volume 9, A. O. D. Willows, ed. pp. 101-187. Academic Press, New York. Eakin, R. M. and J. L. Brandenburger. 1975. Understanding a snail’s eye at a snail’s pace. American Zoologist 15:851-863. Eales, N. B. 1960. Revision of the world species of Ap/ysia (Gastropoda: Opisthobranchia). Bulletin of the British Museum of Natural History, Zoology 5:268-404. Eskin, A. and E. Harcombe. 1977. Eye of Navanax: optic activity, circadian rhythm and morphology. Comparative Biochemistry and Physiology 57A:443-449. Fernald, R. D. 1988. Aquatic adaptations in fish eyes. In: Sensory Biology of Aquatic Animals, J. Atema, R. R. Fay, A. N. Popper and W. N. Tavolga, eds. pp. 435-466. Springer-Verlag, New York. Fletcher, A., T. Murphy and A. Young. 1954. Solutions to two optical prob- lems. Proceedings of the Royal Society of London, A 223:216-225. Frings, H. and C. Frings. 1965. Chemosensory bases of food-finding and feeding in Aplysia juliana (Mollusca: Opisthobranchia). Biological Bulletin 128:211-217. Gibson, B. L. 1984. Cellular and ultrastructural features of the adult and the embryonic eye in the marine gastropod, /lyanassa obsoleta. Journal of Morphology 181:205-220. Gillary, H. L. 1974. Light-evoked electrical potentials from the eye and op- tic nerve of Strombus: Response waveform and spectral sensitivity. Journal of Experimental Biology 60:383-396. 98 AMER. MALAC. BULL. 9(1) (1991) Gillary, H. L. and E. W. Gillary. 1979. Ultrastructural features of the retina and optic nerve of Strombus luhuanus, a marine gastropod. Journal of Morphology 159:89-116. Hamilton, P. V. and M. A. Winter. 1982. Behavioural responses to visual stimuli by the snail, Littorina irrorata. Animal Behavior 30:752-760. Hamilton, P. V. and B. J. Russell. 1982a. Celestial orientation by surface- swimming Aplysia brasiliana Rang (Mollusca: Gastropoda). Journal of Experimental Marine Biology and Ecology 56:145-152. Hamilton, P. V. and B. J. Russell. 1982b. Field experiments on the sense organs and directional cues involved in offshore-oriented swimming by Aplysia brasiliana Rang (Mollusca: Gastropoda). Journal of Ex- perimental Marine Biology and Ecology 56:123-143. Hamilton, P. V., S. C. Ardizzoni and J. S. Penn. 1983. Eye structure and optics in the intertidal snail, Littorina irrorata. Journal of Comparative Physiology 152:435-44S. Hamilton, P. V. and M. A. Winter. 1984. Behavioural responses to visual stimuli by the snails Jectarius muricatus, Turbo castanea, and Helix aspersa. Animal Behavior 32:51-57. Hamilton, P. V. 1986. Swimming tracks of Aplysia brasiliana, with discus- sion of the roles of swimming in sea hares. Veliger 28:310-313. Herman, K. G. and F. Strumwasser. 1984. Regional specializations in the eye of Aplysia, a neuronal circadian oscillator. Journal of Comparative Neurology 230:593-613. Hilger, C. 1885. Beitrage zur Kenntnis des Gastropodenauges. Morpho- logisches Jahrbuch 10:351-371. Hughes, H. P. I. 1970. A light and electron microscope study of some opisthobranch eyes. Zeitschrift fiir Zellforschung und mikroskopische Anatomie 106:79-98. Jacklet, J. W. 1969. Electrophysiological organization of the eye of Aplysia. Journal of General Physiolgy 53:21-42. Jacklet, J. W. and J. Geronimo. 1971. Circadian rhythm: population of in- teracting neurons. Science 174:299-302. Jacklet, J. W., R. Alvarez and B. Bernstein. 1972. Ultrastructure of the eye of Aplysia. Journal of Ultrastructure Research 38:246-261. Jacklet, J. W. 1980. Light sensitivity of the rhinophores and eyes of Aplysia. Journal of Comparative Physiology 136:257-262. Jacklet, J. W. and W. Colquhoun. 1983. Ultrastructure of photoreceptors and circadian pacemaker neurons in the eye of a gastropod, Bulla. Journal of Neurocytology 12:673-696. Jahan-Parwar, B. 1972. Behavioral and electrophysiological studies on chemoreception in Aplysia. American Zoologist 12:525-537. Kataoka, S. 1975. Fine structure of the retina of a slug, Limax flavus L. Vision Research 15:681-686. Kataoka, S. 1977. Ultrastructure of the cornea and accessory retina in a slug, Limax flavus L. Journal of Ultrastructure Research 60:296-305. Kirschfeld, K. 1976. The resolution of lens and compound eyes. Jn: Neural Principles in Vision, F. Zettler and R. Weiler, eds. pp. 354-372. Springer-Verlag, Berlin. Land, M. F. 1965. Image formation by a concave reflector in the eye of the scallop, Pecten maximus. Journal of Physiology 179:138-153. Land, M. F. and F. A. Burton. 1979. The refractive index gradient in the crystalline cones of the eyes of a euphausiid crustacean. Journal of Experimental Biology 82:395-398. Land, M. F. 1981. Optics and vision in invertebrates. In: Handbook of Sen- sory Physiology, Volume VII/6B, H. Autrum, ed. pp. 471-594. Springer-Verlag, Berlin. Land, M. F. 1984. Molluscs. In: Photoreception and Vision in Invertebrates, M. A. Ali, ed. pp. 699-725, Plenum Press, New York. Leonard, J. L. and K. Lukowiak. 1986. The behavior of Aplysia californica Cooper (Gastropoda: Opisthobranchia): I. Ethogram. Behaviour 98:320-360. Lythgoe, J. N. 1980. The Ecology of Vision. Clarendon Press, Oxford. 244 pp. MacFarland, F. M. 1966. Studies of Opisthobranchiate Mollusks of the Pacific Coast of North America. Memoirs of the California Academy of Science, Vol. VI, 546 pp.. Mayes, M. and C. O. Hermans. 1973. Fine structure of the eye of the proso- branch mollusk Littorina scutulata. Veliger 16:166-168. McReynolds, J. S. and A. L. F. Gorman. 1970. Photoreceptor potentials of the opposite polarity in the eye of the scallop, Pecten irradians. Journal of General Physiology 56:376-391. Messenger, J. B. 1981. Comparative physiology of vision in molluscs. In: Handbook of Sensory Physiology, Volume VII/6C, H. Autrum, ed. pp. 93-200. Springer-Verlag, Berlin. Newell, G. E. 1965. The eye of Littorina lilttorea. Proceedings of the Zoological Society of London 144:75-86. Newell, P. F. and G. E. Newell. 1968. The eye of the slug Agriolimax reticulatus (Mull.). Symposium of the Zoological Society of London 23:97-111. Oswaldo-Cruz, E. and R. F. Bernardes. 1982. Morphological and functional observations on the eye of Strophocheilus (Mollusca, Gastropoda, Stylommatophora). Brazilian Journal of Medical and Biological Research 15:161-174. Prince, J. H. 1955. The molluscan eyestalk: using as an example, Prerocera lambis. Texas Reports on Biology and Medicine 13:323-339. Reid, D. G. 1986. The littorinid molluscs of mangrove forests in the Indo- Pacific region: the genus Littoraria. British Museum (Natural History), London. 227 pp. Ronan, M. P. 1990. A review of opisthobranch rhinophores with a study of Aplysia rhinophore structure. Master’s Thesis, University of West Florida, Pensacola, Florida. 113 pp. Schall, V. T. and D. F. Baptista. 1989. Structural and functional analysis of the eye of Biomphalaria glabrata (Mollusca, Gastropoda, Basom- matophora). Brazilian Journal of Medical and Biological Research 22:497-508. Seed, R. 1983. Structural organization, adaptive radiation, and classifica- tion of molluscs. In: The Mollusca, Volume 1, P. W. Hochachka, ed. pp. 1-54. Academic Press, New York. Solem, G. A. 1974. The Shell Makers. John Wiley, New York. 289 pp. Stahlschmidt, V. and H. G. Wolff. 1972. The fine structure of the statocyst of the prosobranch mollusc Pomacea paludosa. Zeitschrift fiir Zellforschung und mikroskopische Anatomie 133:529-537. Stensaas, L. J., S. S. Stensaas and O. Trujillo-Cendz. 1969. Some morphological aspects of the visual system of Hermissenda crassi- cornis (Mollusca: Nudibranchia). Journal of Ultrastructure Research 27:510-532. Stoll, C. J. 1973. Observations on the ultrastructure of the eye of the basommatophoran snail, Lymnaea stagnalis (L.). Proceedings of Koninklijke Nederlandse Akademie van Wetenschappen, Ser. C, 76(4):1-11. Tonosaki, A. 1967. Fine structure of the retina in Haliotis discus. Zeitschrift fiir Zellforschung und mikroskopische Anatomie 79:469-480. Wooters, J. S. 1989. The ecology of vision in the bay scallop, Argopecten irradians (Pectinidae: Bivalvia). Master’s Thesis, University of West Florida, Pensacola, Florida. 57 pp. Zunke, U. 1978. Bau und Entwicklung des auges von Succinea putris (Linne, 1758) (Mollusca, Stylommatophora). Zoologischer Anzeiger 3:220-244. Date of manuscript acceptance: 16 April 1991 Research Note Sperm storage and evidence for multiple insemination in a natural population of the freshwater snail, Physa Amy R. Wethington and Robert T. Dillon, Jr.! Department of Biology, College of Charleston, Charleston, South Carolina 29424, U. S. A. Abstract. Although a number of estimates are available regarding the capacity of freshwater pulmonates to store sperm after laboratory mating, few such data are available for wild-collected snails, where the recency of mating is necessarily unknown. We collected 35 Physa heterostropha pomilia (Conrad) from a local population and held each in isolation for 60 days, rearing all egg masses. We then used protein electrophoresis to determine the LAP genotype of each parent and a sample of its offspring. Five of the parental snails (total 60 day fecundities from 300 - 600 progeny) were found to be homozygous at the LAP locus yet producing approximately 50% heterozygous offspring. In 4 of these 5 cases, no significant difference was detected in offspring genotype frequencies over 60 days, suggesting both that mating has generally been recent and that reservoirs of stored sperm are, as a rule, large. In the fifth case, the frequency of heterozygous offspring increased significantly, suggesting multiple insemination. Multiple insemination and sperm storage have obvious adaptive significance to colonizing species such as freshwater pulmonates. The capability to store sperm, especially sperm con- tributed by multiple partners, has the potential to lessen the severity of genetic drift by increasing the effective popula- tion size represented by a small number of survivors or col- onizers. And the capacity of an individual pulmonate snail to store sperm can be prodigious. In Bulinus, laboratory mating has been reported to provide enough sperm to fer- tilize from 1000 - 2000 eggs up to 4400 eggs, depending par- ticularly upon the reproductive condition of the snail acting as female (Rudolph, 1983; Rudolph and Bailey, 1985). In Biomphalaria, longevity of stored sperm has been variously estimated as 25 - 68 days (Paraense, 1955), 42 days (Richards, 1973), and more than 100 days (Monteiro et al., 1984). Vianey-Liaud et al. (1987) reported a mean of about 50 days with a range from 3 days to 127. Cain (1956) reported some exogenously fertilized egg production by Lymnaea stagnalis Say up to 116 days after isolation. All these studies have in- volved lab crosses with pigment variants (usually albinism) as a genetic marker. Rollinson and Wright (1984) used isozyme markers to demonstrate sperm storage up to 70 days after laboratory mating of Mauritian Bulinus. The effect of a delay in oviposition by the mother (as by a severe winter or desiccation on a birds foot, for exam- ple) has been investigated by Rudolph and Bailey (1985). Ap- parently Bulinus can store viable exogenous sperm through a minimum of seven weeks of starvation, eight weeks of low ‘Address correspondence to RTD temperature (10 - 15°C) or four weeks of desiccation. Rudolph and Bailey (1985) also reported some fairly strong evidence that Bulinus can store sperm from more than a single male simultaneously. Although short-lived copulatory plugs have been described in several pulmonate species, behavioral observations nevertheless suggest that multiple in- semination could be common in laboratory situations (Rudolph, 1979a, b; van Duivenboden and ter Maat, 1988). Duncan (1959) reported that he had not observed reciprocal copulation in either Physa fontinalis (L.) or P. acuta (Drap.). But we have observed both reciprocal copulation and multi- ple mating behavior in our laboratory populations of Physa, without genetic markers for confirmation. Using isozyme markers, Mulvey and Vrijenhoek (1981) found strong evidence of multiple paternity in clutches of eggs laid by isolated wild-caught Biomphalaria. It could not be determined, however, if the sperm used to fertilize these eggs came from two different exogenous sources, or from a com- bination of one outcross and selfing. It does in fact seem that fertilization can proceed with endogenous and exogenous sperm simultaneously, at least in some situations (Paraense, 1955; Monteiro et al., 1984; Rollinson, 1986). But multiple insemination has been conclusively documented by Rollin- son et al. (1989), using stocks of Bulinus cernicus (Morelet) homozygous for three different alleles at the Gpi locus. As important as they are, data from lab crosses such as these do not directly address the likelihood of genetic drift after a population crash or founder event. A second set of American Malacological Bulletin, Vol. 9(1) (1991):99-102 O° 100 AMER. MALAC. BULL. 9(1) (1991) variables is involved: how recently will an arbitrarily chosen pulmonate snail have mated? On the average, how healthy and fecund will it be? Rollinson (1986) has performed ex- periments bearing on this question using several species of African Bulinus. He isolated individual wild-caught snails and collected and reared their eggs up to 60 days. He then examined isozyme phenotype of both parent and offspring at four polymorphic loci. The large majority of these parents did lay demonstrably outcrossed eggs, apparently using stored sperm. Stored sperm was used for at least 41-44 days. Intriguingly, Rollinson found one Bulinus scalaris (Dunker) clearly producing offspring from two different fathers. The great majority of all observations to date have in- volved lab matings of the tropical planorbids Bulinus and Biomphalaria. So to test the generality of the sperm storage phenomenon in natural populations of temperate pulmonates, we performed an analysis similar to Rollinson’s on a local population of Physa. METHODS Physa heterostropha pomilia (Conrad) is a widespread and variable species, found throughout eastern North America (Wurtz, 1949; Burch and Tottenham, 1980). Recently Te (1978, 1980) has suggested that this taxon, along with the majority of North American species, be separated into the resurrected genus Physella. But we agree with Taylor (1988) that the distinction between Physella and Physa (s.s.) is more pro- perly at the rank of subgenus. We collected Physa heterostropha pomilia from a pond at Charles Towne Landing, a state park within the city limits of Charleston, South Carolina. The snails did not appear to be common on this nor on any subsequent trip, and we felt it possible that individual encounters and matings could be infrequent in this population. Voucher specimens have been deposited in the Academy of Natural Sciences of Philadelphia. For an initial survey of polymorphism, we homogen- ized 62 snails in a 7% sucrose solution, buffered at pH 7.4 with 0.05 M tris (hydroxymethyl) aminomethane and H;PO,, to which xylene cyanole had been added as marker. Samples were centrifuged and horizontal starch gel electrophoreses was performed on the supernatant using standard techniques (Dillon and Davis, 1980; Dillon, 1982, 1985). The 14% starch gels were made of 3 parts Sigma starch: | part Electrostarch and AP6 buffer, diluted 19:1. AP6 buffer is 0.04 M citric acid (monohydrate) adjusted to pH 6 with N—(-3-aminopropyl) morpholine (Clayton and Tretiak, 1972). Gels were run for approximately 4.5 hr at 40 volts under refrigeration. They were then sliced and stained for leucine aminopeptidase (LAP) using the recipe of Shaw and Prasad (1969). Two isozymes were found to be segregating in a fashion consistent with Hardy Weinberg expectation in the Charles Towne Landing population, at gene frequencies of 0.58 and 0.42. We have designated the more common isozyme ‘‘LAP 100’’, and the isozyme migrating 3 mm faster in our gel con- ditions ‘‘LAP 103’’. We returned to Charles Towne Landing in May, 1989, and collected 35 adult Physa, placing each in a separate 10 oz plastic cup of pond water with a plastic petri dish cover. We fed them commercial Tetra-Min ‘‘Conditioning’’ food for plant-eating fish (Jennings et al., 1970), and changed the water with fresh, aerated pond water periodically. Each parent was checked daily and transferred to a new cup when an egg mass was produced, in a fashion similar to that of Rollinson (1986). Water was changed once a week for the adults (along with newly laid egg masses) and once every other week for the juvenile snails when they had grown to a size at which this could be done safely. The experiment was terminated in July (after 60 days), by which time 29 of the parents had reproduced, some pro- lifically and others much less so. We then determined the genotypes of each parent and a sample of early laid offspring at the LAP locus. Of the 15 LAP homozygotes identified, four were producing all homozygous progeny and the re- mainder were producing high frequencies of heterozygotes, obviously using exogenous sperm. We did not find any case where a homozygous parent was producing entirely hetero- zygous offspring, as could be expected from a single outcross to the opposite homozygote. But we selected for further study the five largest sibships from the Il including heterozygotes. We then compared the frequency of heterozygotes in 20 to 30 offspring from the earliest egg masses in these sibships to a similar sample from the last laid sibships. RESULTS Data on the fecundity of these five snails (A through E) during the 60 day study period are presented in figure 1. The figure shows the cumulative number of juveniles sur- viving to countable size. It can be seen that all parents con- tinued to lay eggs throughout the entire period, although the rate slowed, especially after day 45. Total viable egg masses ranged from 15 to 19, and the total countable offspring ranged from about 300 to 600. Table 1 shows the frequency of heterozygotes in 20 to 30 offspring of the first laid sibships and the frequency of heterozygotes in similar sized samples from the last laid sib- ships. Genotype frequencies were not significantly different from 1:1 (chi square, Yates corrected) in any of the five groups of first laid offspring. This is consistent with Mendelian ex- pectation if each mother had mated with a single heterozygous father. Table 1 also shows that in four of the five cases, there was no significant difference between the genotype frequen- cies in the first laid sibships and the last laid sibships. WETHINGTON AND DILLON: SPERM STORAGE IN PHYSA 101 Table 1. LAP genotypes among the offspring of five parent Physa, first laid sibships compared to last Number of Number of chi-square Homozygotes 100/103 Heterozygotes 20 15 1.44 16 5 16 17 6.63** 2 18 17 13 0.36 11 7 19 15 1.04 9 14 13 17 0.13 12 11 laid sibships. Parent LAP Day Number Genotype of Oviposition A 103/103 4,9 58,60 B 100/100 7,11 51,52,53,56 Cc 100/100 2,8 56,57,58 D 103/103 2 44,51,56 E 103/103 7 51,52,56 **P < 0.01 The one significant value of chi square (contingency test, Yates corrected) is shown in the offspring of parent B, where there was an unexpected excess of heterozygotes among the off- spring laid in the final days of the experiment. DISCUSSION We do not know whether the total fecundities reflected in figure 1 overestimate or underestimate production by a single Physa founding a new population in the wild. On the one hand, we attempted to provide food in excess and pro- tected the juveniles from predation. But competition among juveniles could have been an important factor in our plastic cups, especially on occasions when as many as three egg masses were laid overnight. Although it can be seen that all parents continued to lay eggs for the entire 60 day period, there was a reduction in the number of viable embryos per egg mass through time. For example, the last three egg masses laid by snail E, on days 57, 58, and 60, contained no viable embryos at all. Qualitatively these results are similar to those obtained by E Offspring 10 20 30 40 50 60 Days Fig. 1. Cumulative 60-day fecundity (offspring surviving to countable size) of the five Physa examined for sperm storage. Duncan (1959) with English Physa fontinalis, although P. heterostropha pomilia fecundities seem to be higher. As previous work suggests a gradual shift from exogenous to endogenous sperm (Paraense, 1955; Cain, 1956), we initial- ly interpreted this observation as evidence of depleted sperm stores. But Table | shows that after 60 days and as many as 600 offspring, none of these individuals seems to have ex- huasted its reservoir of stored sperm. In retrospect, our deci- sion to terminate the experiment was premature. The excess of heterozygotes observed among the pro- geny of parent B is not easily explained as the result of a single pair mating. If the offspring were fathered by a single hetero- zygous individual, one would expect a 1:1 ratio of homo- zygous to heterozygous progeny. This is indeed the ratio observed in all early sibships, laid on days 2 through 11 (Table 1). But the sample of snail B progeny from days 51, 52, 53, and 56 contained 18 heterozygotes and only 2 homozygotes, very significantly different from 1:1. Even combining all 53 progeny examined from parent B, one still obtains a signifi- cant excess of heterozygotes (goodness-of-fit chi square = 4.79). Nor can these results be explained by a single cross to a homozygous father, as no homozygous progeny would have been expected at all. To argue that self fertilization played any role, one would need to postulate that parent B was at least partly self fertilizing to start but increasingly shifted to exogenous sperm as the days in isolation passed, quite counter to all previous observations on other pulmonates. By far the most likely explanation for the results from parent B is multiple insemination by both a heterozygote and a homozygote for the opposite allele. The first eggs seem to have been fertilized by sperm from the former, and the last eggs by the latter, in a fashion similar to that inferred for Biomphalaria (Mulvey and Vrigjenhoek, 1981) and well documented for Bulinus (Rudolph and Bailey, 1985; Rollin- son, 1986; Rollinson et al., 1989). Clearly more work is called for, possibly using multiple loci as markers for dif- ferent parents. But if multiple insemination is in fact 102 AMER. MALAC. BULL. 9(1) (1991) widespread, individual pulmonate snails surviving coloniza- tion events, hard winters or severe storms could potentially represent a great deal of genetic variation indeed. Given the apparently large capacity Physa hetero- stropha pomilia displays for sperm storage, it would be in- teresting to see how frequently an average snail mates, and in what capacity. The great majority of freshwater pulmonate species show either simultaneous development of both reproductive tracts, or are slightly protandric (Russell-Hunter and McMahon, 1976; Rudolph, 1983). The male organs develop before the female organs in P. fontinalis, although environmental conditions in the wild do not in general favor mating until both organ systems are mature (Duncan, 1959). It is not clear whether fully mature pulmonates of any sort prefer to mate as a certain sex, whether a snail’s sexual role can change with size or sperm stores, or how often snails switch roles in single encounters. Multiple insemination in- troduces questions of sperm competition and sperm ‘‘shar- ing’’ (Monteiro et al., 1984; Vianey-Liaud et al., 1987). ACKNOWLEDGMENTS We thank Charles Towne Landing State Park and Mike Dorn for ac- cess to the Physa population and Dr. Margaret Mulvey for her advice on pulmonate culture. LITERATURE CITED Burch, J. B. and J. L. Tottenham. 1980. North American freshwater snails: species list, ranges, and illustrations. Walkerana 3:1-215. Cain, G. L. 1956. Studies on cross-fertilization and self-fertilizasion in Lymnaea stagnalis appressa Say. Biological Bulletin 111:45-52. Clayton, J. W. and D. N. Tretiak. 1972. Amine-citrate buffers for pH con- trol in starch gel electrophoresis. Journal of the Fisheries Research Board of Canada 29:1169-1172. Dillon, R. T., Jr. 1982. The correlates of divergence in isolated populations of the freshwater snail, Goniobasis proxima. Doctoral Dissertation, University of Pennsylvania, Philadelphia. 183 pp. Dillon, R. T., Jr. 1985. Correspondence between the buffer systems suitable for electrophoretic resolution of bivalve and gastropod isozymes. Com- parative Biochemistry and Physiology 82B:643-645. Dillon, R. T., Jr. and G. M. Davis. 1980. The Goniobasis of Southern Virginia and Northwestern North Carolina: genetic and shell morphometric relationships. Malacologia 20:83-98. Duncan, C. J. 1959. Life cycle and ecology of the freshwater snail, Physa fontinalis. Journal of Animal Ecology 28:97-117. Jennings, A. L., K. N. DeKock and J. A. von Eeden. 1970. An evaluation of five different kinds of experimental freshwater snail food. Wetensk Bydraes Potchefstroomse Universitat B (29):1-14. Monteiro, W., J. M. G. Almeida, Jr. and B. S. Dias. 1984. Sperm sharing in Biomphalaria snails: a new behavioral strategy in simultaneous hermaphroditism. Nature 308:727-729. Mulvey, M. and R. C. Vrijenhoek. 1981. Multiple paternity in the her- maphroditic snail, Biomphalaria obstructa. Journal of Heredity 72:308-312. Paraense, W. L. 1955. Self and cross-fertilization in Australorbis glabratus. Memorias do Institato Oswaldo Cruz 53:285-291. Richards, C. S. 1973. Genetics of Biomphalaria glabrata (Gastropoda: Pulmonata). Malacological Review 6:199-202. Rollinson, D. 1986. Reproductive strategies of some species of Bulinus. In: Proceedings of the Eighth International Malacological Congress, L. Painter, ed. pp. 221-226. Hungarian Natural History Museum, Budapest, Hungary. Rollinson, D., R. A. Kane and J. R. L. Lines. 1989. An analysis of fer- tilization in Bulinus cernicus (Gastropoda: Planorbidae). Journal of Zoology (London) 217:295-310. Rollinson, D. and C. A. Wright. 1984. Population studies on Bulinus cernicus from Mauritius. Malacologia 25:447-464. Rudolph, P. H. 1979a. An analysis of copulation in Bulinus (Physopsis) globosus (Gastropoda: Planorbidae). Malacologia 19:147-15S. Rudolph, P. H. 1979b. The strategy of copulation of Stagnicola elodes (Say) (Basommatophora: Lymnaeidae). Malacologia 18:381-389. Rudolph, P. H. 1983. Copulatory activity and sperm production in Bulinus (Physopsis) globosus (Gastropoda: Planorbidae). Journal of Molluscan Studies 49:125-132. Rudolph, P. H. and J. B. Bailey. 1985. Copulation as females and use of allosperm in the freshwater snail genus Bulinus. Journal of Molluscan Studies 51:267-275. Russell-Hunter, W. D. and R. F McMahon. 1976. Evidence for functional protandry in a freshwater basommatophoran limpet, Laevapex fuscus. Transactions of the American Microscopical Society 95:174-182. Shaw, C. R. and R. Prasad. 1970. Starch gel electrophoresis of enzymes - a compilation of recipes. Biochemical Genetics 4:297-320. Taylor, D. W. 1988. New species of Physa (Gastropoda: Hygrophila) from the western United States. Malacological Review 21:43-79. Te, G. A. 1978. The systematics of the family Physidae (Basommatophora: Pulmonata). Doctoral Dissertation, University of Michigan, Ann Arbor. 325 pp. Te, G. A. 1980. New classification system for the family Physidae (Pulmonata: Basommatophora). Archiv fur Molluskenkunde 110:179-184. Van Duivenboden, Y. A. and A. ter Maat. 1988. Mating behavior of Lymnaea stagnalis. Malacologia 28:53-64. Vianey-Liaud, M., J. Dupouy, F. Lancastre and H. Nassi. 1987. Genetical exchanges between one Biomphalaria glabrata (Gastropoda: Planorbidae) and a varying number of partners. Memorias do In- stituto Oswaldo Cruz 82(4):457-460. Wurtz, C. B. 1949. Physa heterostropha (Say). Nautilus 63:2-7. Date of manuscript acceptance: 10 September 1990 SPECIAL PUBLICATIONS OF THE AMERICAN MALACOLOGICAL BULLETIN The Special Publication Series of the American Malacological Bulletin was begun to disseminate collected sets of papers with similar or related themes in a single volume. To date, three such issues have been published, each the result of a special convened symposium. The three Special Editions are PERSPECTIVES IN MALACOLOGY, PRO- CEEDINGS OF THE SECOND INTERNATIONAL CORBICULA SYMPOSIUM, and PROCEEDINGS OF THE SYMPOSIUM ON ENTRAINMENT OF LARVAL OYSTERS. Additional Special Editions are planned for the future. PERSPECTIVES IN MALACOLOGY (Sp. Ed. #1, July 1985) offers a wide range of papers dealing with molluscan biology of interest to professionals and amateurs alike. These papers were presented as part of a symposium held in honor of Professor M. R. Carriker at the time of his retirement and highlight a variety of recent advances in numerous facets of the study of molluscs. PERSPECTIVES IN MALACOLOGY offers insight into some of the frontiers of molluscan biology ranging from deep-sea hydrothermal vent malacofauna to chemical ecology of oyster drills. The PROCEEDINGS OF THE SECOND INTERNATIONAL CORBICULA SYMPOSIUM (Sp. Ed. #2, June 1986) con- tains numerous papers on this exotic bivalve that has become a significant ‘‘pest’’ organism of power plants and other industries using cooling waters. The proliferation, spread, functional biology, attempts at industrial control, taxonomy, and many other topics of interest to the malacologist and industrial biologist are addressed in this important special publication. The third special edition of the American Malacological Bulletin, PROCEEDINGS OF THE SYMPOSIUM ON ENTRAIN- MENT OF LARVAL OYSTERS (Sp. Ed. #3, October 1986) contains important review papers on the larval biology of the American oyster Crassostrea virginica, as well as intriguing papers on factors that limit productivity of these bivalves and limitations that exist on their dispersal and survival. The impact of cutter-head dredges is addressed in this special edition with special emphasis on the Chesapeake Bay system. To order your copies of PERSPECTIVES IN MALACOLOGY, PROCEEDINGS OF THE SECOND INTERNATIONAL CORBICULA SYMPOSIUM and PROCEEDINGS OF THE SYMPOSIUM ON ENTRAINMENT OF LARVAL OYSTERS, simply fill out the form below. Enclose check or money order made out to the AMERICAN MALACOLOGICAL BULLETIN or include VISA or MasterCard number and expiration date. PERSPECTIVES IN 2ND INTERNATIONAL ENTRAINMENT OF MALACOLOGY CORBICULA SYMPOSIUM LARVAL OYSTERS Special Edition No. 1 Special Edition No. 2 Special Edition No. 3 AMERICAN MALACOLOGICAL BULLETIN AMU Members $10.00 $20.00 $14.00 Non-AMU Members $15.00° ye $28.00 $20.00 Unitas Members $12.00 SP Q008 je Bes $16.00 Institutions $25.00 $37.00 a $28.00 Foreign Airmail $10.00 $10.00 $10.00 Foreign Seamail $ 5.00 $5.00 $ 5.00 SUBTOTALS \ ee eee a $ $ TOTAL ENCLOSED $ (check or money order made out to: AMERICAN MALACOLOGICAL BULLETIN) Name: MasterCard/VISA No. Mailing Address: Expiration Date: Send Orders To: Richard E. Petit Acting AMU Secretary-Treasurer Card Holder P. O. Box 30 Signature: North Myrtle Beach, South Carolina 29582, U.S.A. 103 CONTRIBUTOR INFORMATION The American Malacological Bulletin serves as an out- let for reporting notable contributions in malacological re- search. Manuscripts concerning any aspect of original, unpublished research, important short reports, and detailed reviews dealing with molluscs will be considered for publication. Each original manuscript and accompanying illustra- tions must be submitted with two additional copies for review purposes. Text must be typed on one side of 8% x Il inch bond paper, double-spaced, and all pages numbered con- secutively with numbers appearing in the upper right hand corner of each page. Leave ample margins on all sides. Form of the manuscript should follow that outlined in the Council of Biology Editors Style Manual (fifth edition, 1983). This can be purchased from the CBE, 9650 Rockville Pike, Bethesda, Maryland 20814, U.S.A. Text, when appropriate, should be arranged in sections as follows: 1. Cover page with title, author(s) and address(es), and suggested running title of no more than 50 characters and spaces. Authors should also supply five key words, placed at the base of this page, for indexing purposes. 2. Abstract (less than 5% of manuscript length) 3. Text of manuscript starting with a brief introduc- tion followed by methodology, results, and discus- sion. Separate sections of text with centered sub- titles in capital letters. 4. Acknowledgments 5. Literature cited 6. Figure captions All binomens must include the author attributed to that taxon the first time the name appears in the manuscript [e.g. Crassostrea virginica (Gmelin)]. This includes non-molluscan taxa. The full generic name along with specific epithet should be written out the first time that taxon is referred to in each paragragh. The generic name can be abbreviated in the re- mainder of the paragraph as follows: C. virginica. References should be cited within text as follows: Hillis (1989) or (Hillis, 1989). Dual authorship should be cited as follows: Yonge and Thompson (1976) or (Yonge and Thomp- son, 1976). Multiple authors of a single article should be cited as follows: Beattie et al. (1980) or (Beattie et al., 1980). In the literature cited section of the manuscript refer- ences must also be typed double spaced. All authors must be fully identified, listed alphabetically and journal titles must be unabbreviated. Citations should appear as follows: Beattie, J.H., K.K. Chew, and W.K. Hershberger. 1980. Dif- ferential survival of selected strains of Pacific oysters (Crassostrea gigas) during summer mortality. Pro- ceedings of the National Shellfisheries Association 70(2):184-189. Hillis, D.M. 1989. Genetic consequences of partial self- fertilization on population of Liguus fasciatus (Mollusca: Pulmonata: Bulimulidae). American Malacological Bulletin 7(1):7-12. Seed, R. 1980. Shell growth and form in the Bivalvia. In: Skeletal Growth of Aquatic Organisms, D.C. Rhoads and R.A. Lutz, eds. pp. 23-67. Plenum Press, New York. Yonge, C.M. and T.E. Thompson. 1976. Living Marine Molluscs. William Collins Son & Co., Ltd., London. 288 pp. Illustrations should be clearly detailed and readily reproducible. All line drawings should be in black, high quali- ty ink. Photographs must be on glossy, high contrast paper. All diagrams must be numbered in the lower right hand cor- ners and adequately labeled with sufficiently large labels to prevent obscurance with reduction by one half. Magnifica- tion bars must appear on the figure, or the caption must read Horizontal field width = xmm or xpm. All measurements must be in metric units. All illustrations submitted for publica- tion must be fully cropped, mounted on a firm white back- ing ready for reproduction, and have author’s name, paper title, and figure number on the back. All figures in plates must be nearly contiguous. Additional figures submitted for review purposes must be of high quality reproduction. Xerographic reproduction of photomicrographs or detailed photographs will not be acceptable for review. Abbreviations used in figures should occur in the figure caption. Indicate in text margins the appropriate location in which figures should appear. Color illustrations can be included at extra cost to the author. Original illustrations will be returned to author if requested. Any manuscript not conforming to AMB format will be returned to the author for revision. New Taxa. The Bulletin welcomes complete descriptions of new molluscan taxa. Establishment of new taxa must con- form with the International Code of Zoological Nomenclature (1985). Descriptions of new species-level taxa must include the following information in the order as given: higher taxon designation as needed for clarity; family name with author and date; generic name with author and date; Genus species author sp. nov. followed by numeration of all figures and tables; complete synonymy (if any); listing of type material with holotype and any other type material clearly designated along with complete museum catalogue or accession infor- mation; listing of all additional non-type material also with full museum deposition information; type locality; diagnosis and full description of material done in telegraphic style in- cluding measurements and zoogeographic distribution as necessary; discussion; etymology. Descriptions of new supra- specific taxa should include type species (for new genus) or type genus (for new family), diagnosis and full description done in telegraphic style, and list of included taxa. Proofs. Page proofs will be sent to the author and must be checked for printer’s errors and returned to the printer within a three day period. Significant changes in text, other than printer errors, will produce publishing charges that will be billed to the author. Mailing. All overseas mailing must be done via airmail. The American Malacological Union will not be responsible for deferred publication of manuscripts delayed in surface mail. Charges. There are no mandatory page costs to authors lack- ing financial support. Authors with institutional, grant or other research support will be billed for page charges. The current rate is $35.00 per printed page. Acceptance and ulti- mate publication is in no way based on ability to pay page costs. Reprints. Order forms and reprint cost information will be sent with page proofs. The author receiving the order form is responsible for insuring that orders for any coauthors are also placed at that time. Submission. Submit all manuscripts to Dr. Robert S. Prezant, Editor-in-Chief, American Malacological Bulletin, Depart- ment of Biology, Indiana University of Pennsylvania, Indiana, Pennsylvania, 15705-1090, U.S.A. Subscription Costs. Institutional subscriptions are available at a cost of $32.00 per volume. [Volumes | and 2 are available for $18.00 per volume]. Membership in the American Malacological Union, which includes personal subscriptions to the Bulletin, is available for $25.00 ($15.00 for students). All prices quoted are in U.S. funds. Outside the U.S. postal zones, add $5.00 seamail and $10.00 airmail per volume or membership. For subscription or membership information contact AMU Acting Secretary-Treasurer, Richard E. Petit, P. O. Box 30, North Myrtle Beach, South Carolina, 29582, U.S.A. } AMERICAN MALACOLOGICAL BULLETIN VOLUME 9 NUMBER 2 Biannual Journal of the American Malacological Union CONTENTS 1991 AMU/WSM SYMPOSIUM ON MARINE BIVALVES The Bivalvia: Future directions for research, BRIAN MORTON.......................00 000 107 The earliest bivalves and their descendants. BRUCE RUNNEGAR and JOHN POJETA, JR. ..... 0.000000 cence tee nee eeeas 117 Systematics, evolution, and distribution of mussels belonging to the genus Mytilus: an overview. RAYMOND SEED ..........0..00.0.0000 000 ccc cence eee ee 123 The Australasian Protocardiinae revisited (Bivalvia: Cardiidae). ; JEAN-MAURICE POUTIERS ...............0 000000002 ce eee Le... 139 Preliminary cladistic analysis of the bivalve family Cardiidae. JAY A. SCHNEIDER: . 2. koe necees ees eyes see 400K bebe na deus doe acsehe eee 145 Preliminary phylogenetic analysis of the bivalve family Galeommatidae. RUDIGER BIELER and PAULA M. MIKKELSEN .......................0005 00005 157 Seasonal variations in brood size of Lasaea cf. nipponica (Bivalvia: Galeommatoidea) in Hong Kong. BRIAN MORTON ....................0..0 0.000000: 165 Reproductive ecology of the Antarctic bivalve Lissarca notorcadensis (Philobryidae). ROBERT S. PREZANT, MERRILL SHOWERS, RAY L. WINSTEAD and CAROL CLEVELAND ........................05: he msa ke 173 The evolution of the hindgut of the deep-sea protobranch bivalves. JORIN Av ALLEN soc hide cet ed Hh Oped $hbb0ci cds araniddskaeusa ot pha eee ee 187 Prismatic shell formation in continuously isolated (Mytilus edulis) and periodically exposed (Crassostrea virginica) extrapallial spaces: explicable by the same concept? MELBOURNE R. CARRIKER ...................... 193 A new approach to the study of bivalve evolution, MARY ELLEN HARTE................... 199 Research on marine bivalves in the People’s Republic of China. LHUANG QIOUTAN os eee eniceetee ei e ke aS eu ad an dn ti 84H oe 4 He ed nee ee: 207 PImanctal RepoMt sist cqcane8n aed oe eae bd oa dna deen se 4S5GR den bd pe chen chee ee ee 217 1 MOMOTIONE Soh Sica arin tek ae a DG teas oR FEN ond Ow G54 eNO eas A 218 AMERICAN MALACOLOGICAL BULLETIN ROBERT S. PREZANT, Editor-in-Chief BOARD OF EDITORS Department of Biology Indiana University of Pennsylvania Indiana, Pennsylvania 15705 ASSOCIATE EDITORS MELBOURNE R. CARRIKER College of Marine Studies University of Delaware Lewes, Delaware 19958 GEORGE M. DAVIS Department of Malacology The Academy of Natural Sciences Philadelphia, Pennsylvania 19103 R. TUCKER ABBOTT Melbourne, Florida, U.S.A. JOHN A. ALLEN Millport, United Kingdom JOHN M. ARNOLD Honolulu, Hawaii, U.S.A. JOSEPH C. BRITTON Fort Worth, Texas, U.S.A. JOHN B. BURCH Ann Arbor, Michigan, U.S.A. EDWIN W. CAKE, JR. Ocean Springs, Mississippi, U.S.A. PETER CALOW Sheffield, United Kingdom JOSEPH G. CARTER Chapel Hill, North Carolina, U.S.A. ARTHUR H. CLARKE Portland, Texas, U.S.A. CLEMENT L. COUNTS, III Princess Anne, Maryland, U.S.A. THOMAS DIETZ Baton Rouge, Louisiana, U.S.A. WILLIAM K. EMERSON New York, New York, U.S.A. DOROTHEA FRANZEN Bloomington, Illinois, U.S.A. VERA FRETTER Berkshire, United Kingdom ROGER HANLON Galveston, Texas ROBERT C. BULLOCK, Ex Officio Department of Zoology University of Rhode Island Kingston, Rhode Island 02881 BOARD OF REVIEWERS JOSEPH HELLER Jerusalem, Israel ROBERT E. HILLMAN Duxbury, Massachusetts, U.S.A. K. ELAINE HOAGLAND Washington, D.C., U.S.A. RICHARD S. HOUBRICK Washington, D.C., U.S.A. VICTOR S. KENNEDY Cambridge, Maryland, U.S.A. ALAN J. KOHN Seattle, Washington, U.S.A. LOUISE RUSSERT KRAEMER Fayetteville, Arkansas, U.S.A. JOHN N. KRAEUTER Baltimore, Maryland, U.S.A. ALAN M. KUZIRIAN Woods Hole, Massachusetts, U.S.A. RICHARD A. LUTZ Piscataway, New Jersey, U.S.A. GERALD L. MACKIE Guelph, Ontario, Canada EMILE A. MALEK New Orleans, Louisiana, U.S.A. MICHAEL MAZURKIEWICZ Portland, Maine, U.S.A. JAMES H. McLEAN Los Angeles, California, U.S.A. ROBERT F; MCMAHON Arlington, Texas, U.S.A. RONALD B. TOLL, Managing Editor Department of Biology University of the South Sewanee, Tennessee 37375 W. D. RUSSELL-HUNTER Department of Biology Syracuse University Syracuse, New York 13210 THOMAS R. WALLER Department of Paleobiology Smithsonian Institution Washington, D. C. 20560 ANDREW C. MILLER Vicksburg, Mississippi, U.S.A. BRIAN MORTON Hong Kong JAMES J. MURRAY, JR. Charlottesville, Virginia, U.S.A. RICHARD NEVES Blacksburg, Virginia, U.S.A. JAMES W. NYBAKKEN Moss Landing, California, U.S.A. A. RICHARD PALMER Edmonton, Canada WINSTON F: PONDER Sydney, Australia CLYDE F. E. ROPER Washington, D.C., U.S.A. NORMAN W. RUNHAM Bangor, United Kingdom AMELIE SCHELTEMA Woods Hole, Massachusetts, U.S.A. DAVID H. STANSBERY Columbus, Ohio, U.S.A. FRED G. THOMPSON Gainesville, Florida, U.S.A. NORMITSU WATABE Columbia, South Carolina, U.S.A. KARL M. WILBUR Durham, North Carolina, U.S.A. Cover. Io fluvialis (Say, 1825) is the logo of the American Malacological Union. THE AMERICAN MALACOLOGICAL BULLETIN is the official journal publication of the American Malacological Union. AMER. MALAC. BULL. 9(2) ISSN 0740-2783 AMERICAN MALACOLOGICAL BULLETIN VOLUME 9 NUMBER 2 Biannual Journal of the American Malacological Union CONTENTS 1991 AMU/WSM SYMPOSIUM ON MARINE BIVALVES The Bivalvia: Future directions for research. BRIAN MORTON.........................0..-. 107 The earliest bivalves and their descendants. BRUCE RUNNEGAR ANE OEE OME PAee Ve cer rete te ok setae kgm Aula ws Waa e cave gestae Das ardalas Cele ou 17 Systematics, evolution, and distribution of mussels belonging to the genus Mytilus: an overview. RAYMOND SEED .............00.0000 0 ccc ccc eens 123 The Australasian Protocardiinae revisited (Bivalvia: Cardiidae). JEAN-MAURICE POUTIERS ...........0.0. 00000. eee eee eens 139 Preliminary cladistic analysis of the bivalve family Cardiidae. TAYCAN SCHNENDE Ro ee fe cei he ae od HESS Sor ha cre tol Eke bo dew oeee 145 Preliminary phylogenetic analysis of the bivalve family Galeommatidae. RUDIGER BIELER and PAULA M. MIKKELSEN .........................0..00.. 157 Seasonal variations in brood size of Lasaea cf. nipponica (Bivalvia: Galeommatoidea) in Hong Kong. BRIAN MORTON. .........................55- sete se LOD Reproductive ecology of the Antarctic bivalve Lissarca notorcadensis (Philobryidae). ROBERT S. PREZANT, MERRILL SHOWERS, RAY L. WINSTEAD and CAROL CLEVELAND ................ 00.0000 cee eee 173 The evolution of the hindgut of the deep-sea protobranch bivalves. JOHINGA CALEB: 26 joe oo eee nyo sehen as es eee wade pi Osawa Renee 187 Prismatic shell formation in continuously isolated (Mytilus edulis) and periodically exposed (Crassostrea virginica) extrapallial spaces: explicable by the same concept? MELBOURNE R. CARRIKER ...................... 193 A new approach to the study of bivalve evolution. MARY ELLEN HARTE................... 199 Research on marine bivalves in the People’s Republic of China. CHUANG .OIGIAN 6 23 2 sachet eta eet ene g wie dea 25 Be oie eas 207 EinlaniGral RE DOrtss ar. wren eee hiciae halen ential A aie A A ahs Ha aye send Wat Rh Beale WA a as 217 Le MeO ri phe eter Sass yale ee ssre RE ade ag A eee ns ae 0 ARRAN pragactl Ngee Pade SACO ee noe ta se ee 218 —— SYMPOSIUM ON MARINE BIVALVES Organized by PAUL H. SCOTT SANTA BARBARA MUSEUM OF NATURAL HISTORY EUGENE V. COAN SANTA BARBARA MUSEUM OF NATURAL HISTORY BRIAN MORTON DEPT. OF ZOOLOGY AND SWIRE MARINE LABORATORY UNIVERSITY OF HONG KONG AMERICAN MALACOLOGICAL UNION WESTERN SOCIETY OF MALACOLOGISTS BERKELEY, CALIFORNIA 1 - 3 JULY 1991 105 The Bivalvia: Future directions for research Brian Morton Department of Zoology and The Swire Marine Laboratory, The University of Hong Kong, Hong Kong The theme of this paper and attendant symposium is ‘*Future Research Directions for the Bivalvia’. This paper presents a personal view about a class of animals for which we share a common interest and enthusiasm. My involvement with the Bivalvia goes back many years when I first began researches upon Dreissena polymor- pha (Pallas), then and now a continuing problem in the ex- ploitation of fresh waters in Europe. I left Dreissena in 1969, when I completed my Ph.D. and, at the same time, left Great Britain to take up a teaching post in Hong Kong. I believed I had left Dreissena for ever. Just this year, however, I have been asked to write the introductory chapter for a book on Dreissena, the stimulus for publishing such a volume being the introduction of this species into North America via the Great Lakes and the economic consequences of the introduc- tion that are already being felt. I will return to the general topic of introductions later, but it is worth noting, in pass- ing, that the Bivalvia, despite their sedentary habits, do have a tendency to travel to the most unlikely places. I was a student of Professor R. D. Purchon who was, in turn, a student of the late Sir Maurice Yonge, arguably the father of modern research upon the Recent Bivalvia. I first met Maurice in 1967 when he visited the University of London to talk to the student Biology Society of which I was Secretary. He and I, of course, discussed Dreissena for he was working on it too and so a friendship was founded that lasted for 20 years until his death in 1986. I was responsible for organizing a symposium on the Bivalvia in his honour but which was, in the event, in his memory, at the IX Malacological Congress in Edinburgh in 1986. He died just a few months before the meeting was convened. The Pro- ceedings of that Symposium were subsequently published in 1990 (Morton, 1990a). Maurice and his students of the Bivalvia, for exam- ple, G. Owen, J. A. Allen, T. H. J. Gilmour, R. G. B. Reid, A. D. Ansell and R. D. Purchon, have had a powerful im- pact upon not only our understanding of the Recent repre- sentatives of this class, but also upon the Mollusca and Zoology in general for much of this century, notably in Great Britain. Some of these students have emigrated to Canada and the U.S.A., for example, where their influence, and thus Maurice’s, persists. North America too has fostered endemically students of the Bivalvia through such eminent scientists as K. M. Wilbur, R. D. Turner, S. M. Stanley, M. R. Carriker, K. J. Boss, J. Pojeta and N. D. Newell, not forgetting such legendary characters as W. H. Dall, W. R. Coe, T. C. Nelson and V. L. Loosanoff. This is not, however, going to be a discussion about famous bivalve malacologists and you have to forgive me if I have not mentioned your name or that of a mentor you think significant and comparable with those just identified. I am similarly not ignoring equally famous bivalve malacologists from, for example, Europe (V. Scarlato), Australia (B. Runnegar) and elsewhere: the only point I am trying to make is that there are rich geneologies of eminent zoologists who have made the Bivalvia their own and stamped their mark upon a group that today is close to serving as a classic model of adaptive radiation. Some of our most eminent zoologists are students of the Bivalvia. The class is of importance and interest. With this in mind, I looked back over the last twenty- five years of The Journal of Molluscan Studies, so renamed after 1976 from its official status as the Proceedings of the Malacological Society of London, one of malacology’s most prestigous societies and which, in 1993, will celebrate its centenary. I counted the number of papers published in each volume, their page length and noted which of them were con- cerned with the Bivalvia. The results of this simple analysis are shown in figure 1 wherein it can be seen that (A), the number of papers published annually has increased (the quantum jump in 1982 results from the initiation of the publication of Research Notes) although (B), page length, has decreased somewhat (the decline in 1982 again results from the initiation of the publication of Research Notes). Look, however, at figure IC. The number of papers published on the Bivalvia since a heyday in 1967, when 67% of the papers were on this class, has declined progressively. So, I suspect, has the variance. In the 1960’s and 1970’s, volumes would contain between 25% and 45% information on the Bivalvia. In the 1980’s such figures were between 10% and 25%. In 1990, the figure is approximately 10% . Extrapolating American Malacological Bulletin, Vol. 9(2) (1992):107-116 107 108 AMER. MALAC. BULL. 9(2) (1992) 6041 A Number of Papers.volume@ ! . + ee e ] y=12.69+1.532x ZaeT 40 | r=0.832 2 4 e o ee 2 E = = 20% 0 J 16 B. Average page length of each paper e e e 12 y=12.794-0.178x ° r=0.508 ‘ a all oO e : i. e e 44 Oo I 80) C.Papers on the Bivalvia e 60 4 r SS e y=41.36-1.197x Ee ® r0.675 e e - Pee ee e at a en < o ’ e e 0+ Tr T T T T 7 + T T + r 7 1966 1970 1974 1978 1982 1986 1990 Fig. 1. A, Numbers of papers; B, average page length of each paper; C, % numbers of papers on the Bivalvia published in Journal of Molluscan Studies (1967-1990). such a graph, I have concluded that The Journal of Molluscan Studies will cease to publish papers on the Bivalvia by about the time the parent society celebrates its centenary. I have been invited to give the plenary lecture on the Bivalvia at the Conference to celebrate the Society’s centennial: it rather looks as though I will be talking to myself! The editors of the Journal over these years have been N. B. Eales, A. Graham and J. D. Taylor, all gastropod workers, but I am not suggesting uncharitably that they are responsible for this decline. Rather, it appears that there is a declining interest in this class of the Mollusca, at least in Great Britain. I have also taken an American malacological journal, Veliger, appropriate because of its affiliation to the Cali- fornia Malacazoological Society, and performed the same analysis of the published papers. Here, a similar trend is ap- parent with regard to the numbers of papers being published (Fig. 2A) although their average length (Fig. 2B) is increas- ing, unlike papers in the Journal of Molluscan Studies. So, too, however (Fig. 2C), is the incidence of papers on the Bivalvia, again in contrast to the Journal of Molluscan Studies. We thus see a fundamental difference in trends with respect to the Bivalvia in the British and American journals. I do not, however, wish to make too much of this, after all there are dozens of malacological journals and an analysis of them all would be needed to obtain a clearer picture. I will make one point, however. The loss to North America of some of C. M. Yonge’s students, for example, R. G. B. Reid, T. H. J. Gilmour, W. D. Russell-Hunter and P. V. Fankboner, and their potential to, in turn, engender students of the Bivalvia can help us to explain the decline in British studies of the Bivalvia and the increase in North American. Scientific emigration from Britain has been going on for decades and is a damning indictment of British Government policy with regard to its science and its scientists. But is there a wider trend with regard to research upon the Bivalvia? With the help of my colleague, Professor J. Britton, I have conducted a library search of BIOSIS for the numbers of papers published on the Mollusca, Bivalvia, Gastropoda and Cephalopoda since 1969, i.e. the last 22 years. Numbers of papers published on the Mollusca have, as might be antici- pated, increased from 379 in 1969 to 801 in 1990, i.e. an overall increase of 422 or 113.5% (Table 1). Numbers of papers on the Cephalopoda, Gastropoda and Bivalvia have increased by 13 (11.9%), 232 (128.9%) and 177 (135.1%), respectively. The largest overall percentage increase is, thus, upon the Bivalvia. Looking at publications upon the three classes in terms of their percentage contribution to the molluscan body of literature (Fig. 3), trends become apparent. The relative number of papers published on the Cephalopoda has declined, since 1969, by as much as ~ 50%. Research on the Gastropoda seemed to peak in about 1972, cor- responding with a trough in publications on the Bivalvia. 805 A.Number of Papers volume ! e e e Y 604 eue cane = ee e uo Po) e e € ee = oar oe? ig S Th e . y=0.313x+45 136 od e 20) ere r=0.237 ° e g°@ . 20 12'> y=0.146x+4.534 r=0.686 Cc Papers on the Bivalvia @ y=0.217x+12.342 e r=0.355 1 5 9 13 17 21 25 29 33 1989/90 Volume Fig. 2. A, Number of papers; B, average page length of each paper; C, % numbers of papers on the Bivalvia published in Veliger (1967-1990). MORTON: BIVALVIA - FUTURE DIRECTIONS 109 Table 1. Numbers of papers published on the Mollusca and the three largest classes (plus percentages) between 1969 and 1990 (BIOSIS data). Mollusca Gastropoda Cephalopoda Bivalvia No. % No. % No. % 1969 379 180 47.5 68 17.9 131 34.6 1970 359 185 51.5 68 18.9 106 29.5 1971 459 250 54.5 68 14.8 141 30.7 1972 524 296 56.6 68 13.0 160 30.5 1973 511 272 53.2 68 13.3 171 33.5 1974 561 270 48.1 68 12.1 223 39.8 1975 669 315 47.1 86 12.8 268 40.1 1976 716 353 49.3 88 12.3 275 38.4 1977 728 355 48.8 88 12.1 285 39.1 1978 717 323 45.0 88 12.3 306 42.7 1979 708 355 50.1 88 12.4 265 37.4 1980 747 357 47.8 92 12.3 298 39.9 1981 829 396 47.8 92 11.1 341 41.1 1982 811 375 46.2 92 11.3 344 42.4 1983 686 325 47.4 92 13.4 269 39.2 1984 833 376 45.1 92 11.0 365 43.8 1985 856 428 50.0 93 10.9 335 39.1 1986 843 422 50.0 81 9.6 340 40.3 1987 880 425 48.3 81 9.2 374 42.5 1988 930 453 48.7 82 8.8 395 42.5 1989 944 453 48.0 81 8.6 410 43.4 1990 801 412 51.4 81 10.1 308 38.5 Since 1974 the percentage contribution of papers on the Gastropoda to the body of molluscan literature has remained relatively stable between 45% and 50%. Since the early 1970’s, however, publications on the Bivalvia have increased from around 30% to achieve virtual parity with the Gastropoda at between 40-45%. Thus, the global relative significance of the Bivalvia to working scientists has in- creased. Relatively more papers are being published on the Bivalvia today, than upon any other molluscan class. Again, I do not wish to make too much of these figures, after all BIOSIS is not comprehensive, but it may be making you think about the remarks made earlier concerning the British and American malacological journals. Do such figures match up 60 5 ° y=50.917-0.15x be e e r=0.34 ie ae = ° ere Sen res ——e_@ 9 7.) e e a a i * Pine ee fon Ball Le say be ~~ a * x o Pe ae S Q pa a y=33.004+0.486 a a4 r=0.740 ¢ Gastropoda ° ‘ : 4 Bivalvia ® Cephalopoda 207 J. y=16.061-0.337x BY — = _ r=0.853 a2, © ee = . a a s #2 a a a a 19) T 19 T T T T — Sat is = T i ies T 7 i | 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 Fig. 3. Relative numbers of papers published annually on the Gastropoda, Bivalvia and Cephalopoda (BIOSIS data: 1969-1990). with those obtained from the Journal of Molluscan Studies and The Veliger? The short answer is that they do not and we thus see that our ‘academic’ malacological journals are not reflecting, in their contents, the full extent of the research currently being undertaken on the Bivalvia. I now wish to change the direction and introduce this Symposium but, also, to highlight some areas where I believe more research upon the Bivalvia is needed. I will, however, return to the above figures and arrive at a general conclusion. The Bivalvia are an ancient group having their origins in the Palaeozoic. Fordilla and Pojetia are considered to be the oldest bivalves but yet, as Pojeta and Runnegar (1976) point out, we still have not reconstructed an adequate working pic- ture of how such an animal was organized in its shell (Fig. 4). Since the Palaeozoic, the class has radiated, leaving a rich series of fossils and thus a good fossil history that enabled Newell (1965) to lay the basis for a sound, workable, system of classification that, with subsequent minor amendments, serves as a solid base for our understanding of the group’s adaptive radiation. A symposium convened in 1977, organized by C. M. Yonge and T. E. Thompson (1978), under the auspices of The Royal Society, attempted, successfully, to marry the works of bivalve palaeontologists and Recent anatomists and we see arising from this venture a much greater appreciation of the need for the two groups of scien- tists to work together. In this symposium we explore such subjects further in the Evolution and Systematics Session but 10 AMER. MALAC. BULL. 9(2) (1992) there is, nevertheless, still much scope for closer co-operation between the two groups of scientists. The researches of Stanley (1986a, b) and Vermeij (1989a, b) among others, have examined the rich repertoire of bivalve fossils to present ideas on the causes of marine extinctions and thereby explaining the past and present pat- Fig. 4. Four possible explanations of the shell muscle insertions of the Ear- ly Cambrian pelecypod Fordilla. Adductor muscles are cross hatched; radial pallial muscles are stippled; muscles extending from the shell to the inner surface of the mantle are diagonally shaded; pedal muscle insertions are black. Arrows indicate possible water flow in and out of the mantle cavity. Note that if B were correct, the whole of the posterior end of the shell would be effectively sealed (after Pojeta and Runnegar, 1976). terns of geographical restriction. We discuss bivalve biogeography in this Symposium in the Biogeography/Evolu- tion Session. Anatomical studies on the Bivalvia, particularly Recent representatives of what are considered to be ancient lineages, are, however, still needed to help us understand more fully how the Bivalvia has managed to achieve the overall success it so richly, and clearly, enjoys. More importantly, however, such studies would help us to understand the anatomy of the rich array of fossil lineages without Recent representatives. Recently, Purchon (1987) suggested how grades of organization evolved within the Bivalvia to link lineage with lineage, so allowing us to better visualize its evolution. We know much about the shell and ligament struc- ture, mantle fusions and gill, palp and stomach structure of many bivalve families. With continuing expansion and refine- ment, such information should allow us to make much more intelligent guesses at fossil body structure and cladistics could be a useful tool in this respect. The adaptive radiation of the Bivalvia, boosted by a new era of success in the Mesozoic, continues to the present day (Stanley, 1977) and Vermeij (1977) and Taylor (1981) have suggested that such a diversification can be correlated with a ‘Mesozoic Revolution’ involving the adaptive radiation of a new Suite of ecologically important predators that exploited the Bivalvia as a major source of food (Fig. 5). Such preda- tion pressure was, thus, deeply felt by the bivalves, effectively driving them underground (Stanley, 1977) with concomitant adaptations for deep burrowing. It could also have fostered the exploitation of rocky shores by more modern hetero- myarian heterodonts and the evolution of a coral host/ bivalve borer symbiosis and modifications to the borer’s shell in tropical species (Fig. 6) (Morton, 1990b). The indepen- dent appearance of cementation in many clades of bivalves 500F © Diversity of bivalve genera 10 = 2 © Rate of evolution of bivalve genera = 8 e 3s a a 2 8 c=) / 3 3S LO ° = = Pe io i] Se ao 8 0 cos 410 i 0 w PALAEOZOIC MESOZOIC CAENOZOIC A 2016p ¢ Rate of evolution of 20 g siphonate bivalves bd = o Diversity of predatory Doe § gastropod families o & € : ry © ro} a s 9 3 2 e = & Ys a" i z e on oO B 2 0 4 — t — MESOZOIC CAENOZOIC Fig. 5. A, rates of bivalve evolution at the family and generic levels (after Stanley, 1973); B, rates of bivalve and predatory neogastropod evolution at the family level during the Caenozoic (after Taylor, 1981). MORTON: BIVALVIA - FUTURE DIRECTIONS il in the Palaezoic and Mesozoic can also be linked to preda- tion pressure (Harper, 1991). Patterns of evolution in the Bivalvia are beginning to emerge. Additional evidence, however, points to the success of the Bivalvia and to their importance in aquatic ecosystems. Sanchez-Salazar et al. (1987a, b) have shown how on a small sandy beach in North Wales, Great Britain, the cockle Cerastoderma edule Linné, is a far more important prey item for foraging crabs [Carcinus maenas (Linné)] and Oyster- catchers (Haematopus ostralegus Linne) than was conceived of previously. On this one small hectare bay, some 18 million juvenile cockles are consumed annually by these predators. Griffiths (1990) has suggested that the same pressures could be acting upon the gregarious bivalves of rocky shores and we can identify a wide range of mesogastropod, neogastropod and opisthobranch snails (Natica; Melongena, Buccinum, Philine), crabs (Carcinus; Thalamita), fish (Pleuronectes) and birds (Haematopus; Crocethia) that exploit bivalves as food. I believe bivalves to be much more important as primary con- sumers and prey in aquatic food chains than presently ap- preciated. This could already be recognized, but literature on the subject is published in non-malacological journals (just as with the above papers) such that they become mere names and numbers in papers dealing not so much with the animals themselves but with their presence or absence in com- munities. In studies of benthic assemblages, they have thus become relegated to a statistic. I have pointed out (Morton, 1991b) that the Mesozoic Revolution also affected the Bivalvia in other ways, driving some of them into the deep seas where many still retain primitive characters and could thus constitute living fossils. I suggested, for example, that Bathyarca could be a living cryptodont (Morton, 1982) while Allen and Sanders (1969) have suggested that Nucinella is a living actinodont. Waller (1971) has suggested that the Propeamussidae are similarly primitive. The deep seas have, however, also fostered the evolution of remarkable structures that have enabled some lineages to pursue predatory careers. I refer of course to the septibranch Anomalodesmata studied by Yonge (1928), Allen (1983), Knudsen (1979), and Reid and Reid (1974) and leading me to suggest (Morton, 1991b) that, in this environment, they have become the agents for natural selection rather than a consequence of it. But, how much, in reality do we know of the lifestyle of such animals? If I were to suggest that.a group of modern ungulates possessed representatives that are sedentary, ambushing, carnivores residing upon the Himalayas, think of the thousands of researchers who would desperately seek to find them. The mythical Yeti is a case in point. Why not for the septibranchs too then? For that is what they are, in essence, and urgently call out for greater study as very few of them have been studied alive. There is, I thus believe, still much opportunity for significant research upon the evolution and adaptive radia- tion of the Bivalvia. Bivalves are also important economically. The gregar- ious shallow water representatives of the Pterioida and Heterodonta are two major lineages of bivalves that not only predators exploit as a major class of prey, but which also com- mand our attention as food. Such animals as clams, cockles, mussels, oysters and scallops have been exploited as a human food resource since the Neolithic. Their importance is recognized in their contribution to the composition of the kitchen middens of such early people. Their importance con- tinues today and many are now the subjects of thriving mariculture industries as reflected in the Food and Agriculture Organization (1989) figures for 1987 as presented in Table 2. Such data are, however, in reality, only the tip of a vast underwater iceberg of what the Bivalvia really constitute as a human food resource and their exploitation has led to a massive literature, often unreferred to by malacologists. In today’s ‘economic’ world I believe that academic malacologists must surely aim for an enhanced co-operation with mariculturists. We are aware of the researches and mariculture efforts of workers upon the oysters that have, through Crassostrea gigas (Ventilla, 1984), revitalized dy- ing industries based upon less hardy species and of the over- whelming success of the Japanese scallop (Ventilla, 1982) and pearl oyster industries, but is there any significant research being undertaken on the wider range of other potentially cultivatable bivalves? One example of such a success has been the effort to re-establish overexploited and thus endangered Table 2. Fishery statistics for the Mollusca - 1987. (After: Food and Agriculture Organization of the United Nations, 1989). Commodity Production (metric tonnes) Abalone meat (frozen) ......... 0.000000 1,634 Smalls sGirOZEM)arsqscea ceeds setsee fencicucne ane eustensciac. gee vgnsPecacneterc ee 3,964 Oyster meat (MOZEN) 0. ois se ose ne ens a ea ete S 2,974 Mussel meat (frozen) ............ 00000 11,242 Scallop meat (frozen) ........ 000. eee 38,866 Clams, cockles etc., meat (frozen) ...............0.00 2000. 66,634 Cuttlefish (frozen) ......0...00 00.0000 cee eee eee eee eee. 37,831 Squidss(frozen): anz.c. score gees eis Snares eis onsale eat 563,720 Octopus:(frozZen) sce howe ecy neds e ee Reet has eee eben ke 53,183 General cephalopods (frozen) ..............0..00 000s e eee 158,092 General molluscs (frozen). ........0.0 00.050 cece eee eee 127,194 OySterss(Aried) sche sic ances ee aiigs evs ce ansee tas there nvavaneds Gen aaa eeaecat abe 3,934 Guttlefishs (dried) cccie ce enon tes ech seers on ac Sees enone ss vamiec ava enars 242 Squids (Grid) aise. saproreaitass ceatlsiggn pose a ayn eerie wcedeseuasdevaerae 48,431 OCtOpUs: (CEE) a fare cots. sbi davare shies ar sever eilacet nurs outs sranduens slqeene dune Go eeacions 80 Squids (smoked) .......... 000.000. 5,684 General cephalopods (dried, salted, etc.) ...............00.. 15,431 General molluscs (dried, salted, etc.) ..............0.20.2040. 8,640 Total 1,147,776 Bivalvia (Total) 123,650 112 AMER. MALAC. BULL. 9(2) (1992) Lithophaga teres L. malaccana L. laevigata L. aristata Petricola lapicida Gregariella coralliophaga Coralliophaga coralliophaga Jouannetia cumingjii Fig. 6. Lateral views of the shells of coral boring and nestling bivalves. Also illustrated are end-on views of the posterior valve margins of L. laevigata and L. aristata (after Morton, 1991a). Pacific island stocks of giant clams. Induced breeding has led not only to the probability of their conservations but also to the commercial development of them as a fishery resource, all the more significant because they do not have to be fed! Bivalves are economically important in other ways: as borers of wood, stone and plastic in the sea, so eloquently evaluated by R. D. Turner and her Harvard school. But, apart from this group, how many others are researching this economical- ly important and fascinating group of bivalves? We also are aware of the importance of mussels (and oysters) in the monitoring of an almost globally declining in- shore aquatic environment. Their propensity to acquire large amounts of sewage bacteria, red tide toxins and a suite of trace metals and organochlorines, may have reduced the impor- tance of the Bivalvia in the public’s eye as potential food (in the absence of widescale depuration technology) but they are, nevertheless, virtually ideal indicators of environmental stress (Akberali and Trueman, 1985). In polluted Hong Kong, we have identified Perna viridis (Linné) (Mytilidae) (Lee, 1985), and, most recently, Tapes philippinarum (A. Adams and Reeve) (Veneridae) as highly resistant final indicators of sewage pollution. As such, they dominate polluted hard and soft shores, respectively, adding new significance to the ecological importance of the Bivalvia in such habitats but also posing the question: why is it that we have not exploited such a protein-producing potential for our own benefit? Such animals are also as important, in today’s polluted world, as good monitoring sentinels of ambient pollution loadings - but how many such ‘mussel watch’ networks have been de- veloped, despite their repeated advocacy (National Academy MORTON: BIVALVIA - FUTURE DIRECTIONS 113 of Sciences, 1980). I have also to ask why it is that such research, when undertaken, is published not in malacological journals but in their more applied competitors, for example, Aquaculture and Marine Pollution Bulletin. Academic malacologists seem to have a mental block about the significance of research on the applied aspects of their sub- ject and bivalve workers seem particularly aloof to such work. Some ‘opportunistic’ bivalves too are important foul- ing agents, particularly in fresh waters. I earlier mentioned the subject of my first researches, Dreissena polymorpha, spread from its home base in the Caspian Sea throughout Europe during the 19th century Industrial Revolution and on into Great Britain. In 1989 it was introduced into the Great Lakes and, if there are any lessons to be learnt from history, it will invade progressively most of the Americas. The economic consequences of its spread will be enormous and yet I have to ask the question: how many reputable bivalve malacologists have stepped in to research it? My impression is few and, if true, then the story of Dreissena, as with Cor- bicula flumina (Miller) and its introduction into North America and subsequent spread (McMahon, 1983), will be left in the hands of engineers and aquatic ecologists whose relationship with the academic malacological fraternity who should be at the forefront of such research, is, at best, tenuous. Regretably, Dreissena and Corbicula were not discussed at this Symposium, nor even this Conference. Dare I suggest it, but modern bivalve researchers could do no better than to study Dreissena and Corbicula for they are what we need: living proof of the importance of the Bivalvia and which no gastropod, save vectors of schistosomiasis, can even remote- ly approach in terms of economic significance. There are many other examples of exotic introductions: Mytilopsis sallei Récluz introduced into the Pacific from the Atlantic (Morton, 1987a) and Musculista senhausia (Benson) introduced into the southern from the northern Pacific (Willan, 1987). With Pacific trade now outstripping that across the Atlantic, there will be many more such voyages by the Bivalvia (Carlton, 1987). It seems to me that bivalve workers must dirty their hands with such animals that appear, superficially, to be so uninspiring, because, in reality, it is these that are so clearly the visible proof of the success of the Bivalvia in our modern world. My close friend and colleague, J. C. Britton, and I have, over the years, researched Corbicula fluminea here in America and elsewhere. Though others continue to argue with us (and I have no problem with that: arguments lost, won and then lost again are the stuff of science), we believe that only one species has been introduced but that it is poly- morphic in terms of shell form, texture and colour and, most fascinating of all, sexual expression (Britton and Morton, 1986). We believe we are essentially seeing, through genotypic and phenotypic plasticity, evolution in action and yet how many other bivalve workers are actively engaged in sorting out this riddle? Open up any volume of the Biological Journal of the Linnean Society of London to see the extent of the literature on polymorphism in the Gastropoda. My own paper on polymorphism in Hong Kong Corbicula (Morton, 1987b) was dismissed by a reviewer as wrong - there are simply two species involved - revealing the extent of one of our peers knowledge on this subject in the Bivalvia. I should add that a subsequent paper by Tsoi ef al. (1990) has demonstrated that there is no significant allozyme difference between the two morphs while Kijviriya et al. (1991) have shown, using the same electrophoretic approach, that 21 nominal species of Corbicula in Thailand are all assignable to C. fluminea. Cain (1988) has recently published a paper on polymorphism in deep-burrowing Macoma_balthica (Linné) and suggests that it results from apostatic selection. If this is true, then research on this topic has just received another intellectual injection that should be married into our current concept of how the Bivalvia has evolved and radiated as a result, at least in part, by natural selection through predation. I am aware of the inherent difficulties in undertaking discrete genetic studies upon randomly externally cross- fertilizing bivalves when the sex of the gamete donors can- not often be determined until autopsy. Such research should, however, be a challenge for us, not a hinderance, and that such animals, gregarious, numerous, prolific, and economically important as well as sessile bioindicators of changing environmental regimes should be receiving far more of our attention than they do currently. With the modern technology of DNA fingerprinting before us, research upon the Bivalvia should, already, have seen a quantum leap in output, but it clearly has not. Even DNA technology is un- necessary, however, for some bivalves, which possess self- fertilizing representatives, e.g. the Galeommatoidea, and O Foighil (1989) has been able to come to important conclu- sions regarding the significance of planktotrophy versus direct development in the dispersal of representatives of one cosmopolitan genus of this superfamily, Lasaea. But there are many other lineages of hermaphroditic bivalves about which virtually nothing is known, for example, the Anomalodesmata (Morton, 1985). I recently published a paper (Morton, 1991a) in which I argued that life history tactics and reproductive strategies among a suite of bivalves occupying a freshwater to marine continuum were related to the microhabitats occupied (Table 3). This model will, hopefully, be tested by others. I hesitate to ask this, however, but why is it that the Bivalvia so ob- viously sessile, so obviously in intimate reality with a sweep of aquatic habitats from mountain pools to the abyss have not become the model for such studies and further, since we now know that many of their life history traits can be environmen- tally regulated, that they also become the model to help us understand how such natural and unnatural perturbations 114 AMER. MALAC. BULL. 9(2) (1992) Table 3. The sexual strategies adopted by Hong Kong freshwater, estuarine and intertidal bivalves (After Morton, 199 1a). Sexual strategy adopted Hermaphroditic: Hermaphroditic/ Dioecious: (brooding) dioecious: Strongly Pronounced environmentally female bias regulated sex (brooding ratio (brooding) Freshwater Musculium lacustre Corbicula (small lotic) Pisidium clarkeanum _ fluminea Pisidium annandalei Freshwater (large lentic) in Anodonta) Anodonta woodiana Limnoperna fortunei Dioecious: Dioecious: Dioecious: Dioecious Overall Slight overall (Alternative Slight male female bias; female bias; sexuality in bias (not female bias male bias Saccostrea): significant) in juveniles in juvenile Slight overall (non-brooding) (non-brooding) | Brachidontes male bias; (non-brooding) male bias in juveniles (non-brooding) Freshwater Corbicula cf. (large lotic) fluminalis Mangrove Polymesoda erosa Mangrove Brachidontes variabilis Mangrove Gafrarium pectinatum Harbour Mytilopsis sallei Saccostrea cucullata Intertidal Perna marine viridis Donax semigranosus work at the individual, population and species levels. C. M. Yonge was fascinated by the Bivalvia and laid the foundation for our modern understanding of them. I have read recently the paper by Mikkelsen and Bieler (1989) about the yo-yoing galeommatoidean Divariscintilla yoyo Mikkelsen and Bieler. I am also reminded of the work I did upon Chlamydoconcha orcutti Dall (Fig. 7A) (Morton, 1981) which is one of the most remarkable of bivalves. It has a minute shell, a weirdly anteriorly monomyarian musculature and very strange defensive appendages. The female also possesses a, possibly parasitic, dwarf male (Fig. 7B). This animal is as exotic as any gastropod. The Galeommatoidea must be one of the most fascinating superfamilies of the Bivalvia. Every study of their numerous representatives speaks of their strange adaptations. Yet, all would agree that the taxonomy and systematics of the group, at every level, are a mess. We desperately need someone to sort them out because I believe that their story is one of the strangest yet to be told. We should be fascinated by such animals and encourage work upon them for they not only expand our understanding of the full extent of bivalve radiation but also open up new research horizons into the origins, through neoteny, of commensalism and, possibly, parasitism. Bivalves, alone among the Mollusca, are overwhelm- ingly economically and ecologically significant and yet they seem, returning to my first figures, to be of declining interest to academic malacologists. Clearly, however, there is not an overall decline in interest (Fig. 3). Is it possible that we as academic malacologists are not in step with globally chang- ing perceptions of the Bivalvia? Is it, further, possible that we are missing a golden harvest of research money and careers that could ultimately be more rewarding? It is clear that few applied aspects of our chosen class were part of this symposium, i.e. their pollution monitoring potential, mariculture significance and their importance as significant fresh water and marine foulers. It could be useful to convene MORTON: BIVALVIA - FUTURE DIRECTIONS HS Dwarf Defensive A male papilla Pheromone organ Shell Defensive papilla Exhalant Inhalcn ~ Siphon _——s : cht Te Smm Foot f / / Pedal Byssus Byssus gape — _ oland Mucous Shell Embryonic shell ye Byssus Fig. 7. Chlamydoconcha orcutti. A, female; B, dwarf male (after Morton, 1981). joint meetings of academic and applied bivalve workers on specific topics. I am aware, for example, of the regularly organized pectinid workshops (the 8th was convened in 1991, in Cherbourg, France) that have brought together a rich mix of pure and applied malacologists to great effect. If, therefore, you agree, in any way with my analysis of the current situa- tion with regard to trends and omissions in bivalve research, then perhaps we can persuade the American Malacological Union to assist us in bringing more academic and applied bivalve malacologists together so that the two groups of scien- tists can begin to benefit, more fully, from the closer co- operation that such meetings would foster. ACKNOWLEDGMENTS I am very pleased that Paul Scott and his colleagues seized the in- itiative in organizing a Symposium and subsequent Workshop on the Bivalvia. I am also grateful to Paul for convening this Symposium and to the Councils of the American Malacological Union and the Western Society of Malacologists for agreeing to host the meeting. I am further grateful that I had the opportunity to address the Opening Session. LITERATURE CITED Akberali, H. B. and E. R. Trueman. 1985. Effects of environmental stress on marine bivalve molluscs. Advances in Marine Biology 22:101-198. Allen, J. A. 1983. The ecology of deep sea molluscs. In: The Mollusca. Vol. 6. Ecology. W. D. Russell-Hunter, ed. pp. 29-75. Academic Press, Inc., Orlando, Florida. Allen, J. A. and H. L. Sanders. 1969. Nucinella serrei Lamy (Bivalvia: Pro- tobranchia), a monomyarian solemyid and possible living actinodont. Malacologia 7:381-396. Britton, J. C. and B. Morton. 1986. Polymorphism in Corbicula fluminea (Bivalvia: Corbiculacea) from North America. Malacological Review 19:1-43. Cain, A. J. 1988. The colours of marine bivalve shells with special reference to Macoma balthica. Malacologia 28:289-312. Carlton, J. T. 1987. Patterns of transoceanic marine biological invasions in the Pacific Ocean. Bulletin of Marine Science 41:452-465. Food and Agriculture Organization of the United Nations. 1989. Year book. Fishery Statistics for 1987. Vol. 65. FAO Fisheries Series No. 33. pp. 369. Griffiths, C. L. 1990. Spatial gradients in predation pressure and their in- fluence on the dynamics of two littoral bivalve populations. In: The Bivalvia. Proceedings of a Memorial Symposium in Honour of Sir Charles Maurice Yonge, Edinburgh, 1986. B. Morton, ed. pp. 321-332. Hong Kong University Press, Hong Kong. Harper, E. M. 1991. The role of predation in the evolution of cementation of bivalves. Palaeontology 34:455-460. Kijviriya, V., E. S. Upatham, V. Viyanant and D. S. Woodruff. 1991. Genetic studies of Asiatic clams, Corbicula, in Thailand: allozymes of 21 nominal species are identified. American Malacological Bulletin 8:97-106. Knudsen, J. 1979. Deep sea bivalves. In: Pathways in Malacology. S. van der Spoel, A. C. van Bruggen and J. Lever, eds. pp. 195-224. Bohn, Scheltema and Holkema, Utrecht. Lee, S. Y. 1985. The population dynamics of the green mussel, Perna viridis (L.) in Victoria Harbour, Hong Kong - dominance in an polluted en- vironment. Asian Marine Biology 2:107-118. McMahon, R. F. 1983. Ecology of an invasive pest bivalve, Corbicula. In: The Mollusca. Vol. 6. Ecology. W. D. Russell-Hunter, ed. pp. 505-561. Academic Press, Inc., Orlando, Florida. Mikkelsen, P. M. and R. Bieler. 1989. Biology and comparative anatomy of Divariscintilla yoyo and D. troglodytes, two new species of Galeom- matidae (Bivalvia) from stomatopod burrows in Eastern Florida. Malacologia 31:175-195. Morton, B. 1981. The biology and functional morphology of Chlamrydoconcha orcutti Dall with a discussion on the taxonomic status of the Chlamydoconchacea (Mollusa: Bivalvia). Journal of Zoology, Lon- don 195:81-122. Morton, B. 1982. Functional morphology of Bathyarca pectunculoides (Bivalvia: Arcacea) from a deep Norwejian fjord with a discussion of the mantle margin in the Arcoidea. Sarsia 67:269-282. Morton, B. 1985. Adaptive radiation in the Anomalodesmata. In: The Mollusca. Vol 10. Evolution. E. R. Trueman and M. Clarke, eds. pp. 405-459. Academic Press, New York. Morton, B. 1987a. Recent marine introductions into Hong Kong. Bulletin of Marine Science 41:503-513. Morton, B. 1987b. Polymorphism in Corbicula fluminea (Bivalvia: Cor- biculacea) from Hong-Kong. Malacological Review 20:105-127. Morton, E. (Editor) 1990a. The Bivalvia. Proceedings of a Memorial Sym- posium in Honour of Sir Charles Maurice Yonge, Edinburgh, 1986. Hong Kong University Press, Hong Kong. 355 pp. Morton, B. 1990b. Corals and their bivalve borers - the evolution of a sym- biosis. In: The Bivalvia. Proceedings of a Memorial Symposium in 116 AMER. MALAC. BULL. 9(2) (1992) Honour of Sir Charles Maurice Yonge, Edinburgh, 1986. B. Morton, ed. pp. l-46. Hong Kong University Press, Hong Kong. Morton, B. 199la. Do the Bivalvia exhibit environment-specific sexual strategies? A Hong Kong model. Journal of Zoology, London 223:131-142. Morton, B. 1991b. ‘Cockles and mussels - alive, alive O’. Inaugural Lec- ture. University of Hong Kong, Supplement to the Gazette. Vol. XXXVIII. No. 1. pp. 1-20. National Academy of Sciences. 1980. The International Mussel Watch. Na- tional Academy of Sciences, Washington, D.C. 248 pp. Newell, N. D. 1965. Classification of the Bivalvia. American Museum ; Novitates No. 2206:1-25. O Foighil, D. 1989. Planktotrophic larval development is associated with a restricted geographic range in Lasaea, a genus of brooding, her- maphroditic bivalves. Marine Biology 103:349-358. Pojeta, J. and B. Runnegar. 1976. Fordilla troyensis and the earliest history of pelecypod mollusks. American Science 62:706-711. Purchon, R. D. 1987. Classification and evolution of the Bivalvia: an analytical study. Philosophical Transactions of the Royal Society, Series B 316:277-302. Reid, R. G. B. and A. M. Reid. 1974. The carnivorous habits of members of the septibranch genus Cuspidaria (Mollusca: Bivalvia). Sarsia 56:47-S6. Sanchez-Salazar, M. E., Griffiths, C. L. and R. Seed. 1987a. The interac- tive roles of predation and tidal elevation in structuring populations of the edible cockle, Cerastoderma edule. Estuarine, Coastal and Shelf Science 25:245-260. Sanchez-Salazar, M. E., Griffiths, C. L. and R. Seed. 1987b. The effect of size and temperature on the predation of cockles Cerastoderma edule (L.) by the shore crab Carinus maenas (L.) Journal of Experimental Marine Biology and Ecology 111:181-193. Stanley, S. M. 1973. Effects of competition on rates of evolution, with special reference to bivalve mollusks and mammals. Systematic Zoology 22:486-506. Stanley, S. M. 1977. Trends, rates and patterns of evolution in the Bivalvia. In: Patterns of Evolution. A. Hallam, ed. pp. 209-250. Elsevier, Amsterdam. Stanley, S. M. 1986a. Anatomy of a regional mass extinction: Plio-Pleistocene decimation of the western Atlantic bivalve fauna. Palaios 1:17-36. Stanley, S. M. 1986b. Population size, extinction, and speciation: the fis- sion effect in Neogene Bivalvia. Paleobiology 12:89-110. Taylor, J. D. 1981. The evolution of predators in the late Cretaceous and their ecological significance. In: Chance, Change and Challenge: The Evolv- ing Biosphere. P. L. Forey, ed. pp. 229-240. The British Museum (Natural History), London and Cambridge University Press, Cambridge. Tsoi, S. C. M., S. C. Lee, W. L. Wu and B. Morton. 1990. Genetic varia- tion in Corbicula fluminea (Bivalvia: Corbiculoidea) from Hong Kong. Malacological Review 23:81-90. Ventilla, R. F. 1982. The scallop industry in Japan. Advances in Marine Biology 20:309-382. Ventilla, R. F. 1984. Recent developments in the Japanese oyster culture industry. Advances in Marine Biology 21:1-57. Vermeij, G. 1977. The Mesozoic marine evolution: evidence from snails, predators and grazers. Paleobiology 3:245-258. Vermeij, G. 1989a. Geographical restriction as a guide to the causes of ex- tinction: the case of the cold northern oceans during the Neogene. Paleobiology 15:335-356. Vermeij, G. 1989b. Invasion and extinction: the last three million years of North Sea pelecypod history. Conservation Biology 3:274-281. Waller, T. R. 1971. The glass scallop Propeamussium, a living relict of the past. Report of the American Malacological Union Pacific Division 1970:5-7. Willan, R. C. 1987. The mussel Musculista senhausia in Australasia, another aggressive alien highlights the need for quarantine at ports. Bulletin of Marine Science 41:475-489. Yonge, C. M. 1928. Structure and function of the organs of feeding and diges- tion in the septibranchs, Cuspidaria and Poromya. Philosophical Trans- actions of the Royal Society of London, Series B 216:221-263. Yonge, C. M. and T. E. Thompson (Editors). 1978. Evolutionary systematics of bivalve molluscs. Philosophical Transactions of the Royal Society of London, Series B 284:199-436. Date of manuscript acceptance: 25 November 1991 The earliest bivalves and their Ordovician descendants Bruce Runnegar! and John Pojeta, Jr? ‘Department of Earth and Space Sciences, Institute of Geophysics and Planetary Physics, and Molecular Biology Institute, University of California, Los Angeles, California 90024, U. S. A. 2United States Geological Survey, National Center, Mail Stop 970, Reston, Virginia 22092, U. S. A. Abstract. Replicated shell microstructure in a specimen of Fordilla troyensis Barrande from the Early Cambrian of Greenland confirms a close relation- ship between Fordilla Barrande from the North Atlantic region and Pojetaia Jell from the Early Cambrian of Australia and China. These genera could either be stem-group bivalves which predated the last common ancestor of living members of the class or the earliest known representatiaves of the extant subclasses Isofilibranchia and Palaeotaxodonta. Two other bivalved molluscs from the Middle Cambrian of Australasia and Scandanavia (Tuarangia Mackinnon and Pseudomyona Runnegar) had D-shaped valves, a single central adductor muscle, and shells formed of foliated calcite. They have been interpreted as early pteriomorphian bivalves and also as bivalved monoplacophorans; each displays some features of larval bivalves but at shell sizes that are larger than is typical for bivalve prodissoconchs. If Tuarangia and Pseudomyonia were early pteriomorphian bivalves the Class Bivalvia might well be diphyletic. However, it is more likely that the Palaeotaxodonta had a fordillid ancestor and that all other extant subclasses of Bivalvia were derived from this paraphyletic group. Other proposed genera of Cambrian Bivalvia (Buluniella Jermak, Cycloconchoides Zhang, Hubeinella Zhang, Lamellodonta Vogel, Oryzoconcha He and Pei, Praelamellodonta Zhang, Xianfengoconcha Zhang, Yangtzedonta Yii) are either not molluscs or are junior synonmys of Fordilla and Pojetaia. The search for Cambrian ancestors of the bivalves that in age. In South Australia, Pojetaia runnegari Jell occurs in are so obvious in Ordovician epicontinental marine strata has three of four successive trilobite zones of the middle part of yielded a variety of fossils that have been offered as early the Early Cambrian (zones of Abadiella huoi (Zhang), representatives of the Class Bivalvia (Table 1). Of these, only- Pararaia tatei (Woodward) and P. janeae Jell; Bengston et Fordilla Barrande (Pojeta, 1975) and Pojetaia Jell (Runnegar al., 1990). This stratigraphic range is approximately and Bentley, 1983) are regarded widely as true bivalves but equivalent to the Chiungchussu and Tsanglangpu stages of even they could represent stem group taxa in that they could the Chinese Early Cambrian succession and to the Atdabanian predate the latest common ancestor of all living members of and Botomian stages of the Siberian Platform. Chinese oc- the class (Morris, 1990; Waller, 1990). Two Middle Cam- currences of Pojetaia are from the early Tsanglangpu stage brian taxa, Pseudomyona Runnegar and Tuarangia Mackin- (Chen and Wang, 1985; He and Pei, 1985) and are probably non (which are obviously related to one another), have been equivalent in age to the zone of Nevadella Raw of North regarded as early pteriomorph bivalves by some workers America (P. A. Jell, pers. comm.). (Mackinnon, 1982; Berg-Madsen, 1987) but Runnegar (1983) Fordilla is found in the middle Early Cambrian rocks considered them to be bivalved monoplacophorans, analogous in eastern North America, Greenland and Denmark (Pojeta, to living bivalved opisthobranch gastropods. The purpose of 1975) and in Tommotian and Atdabanian strata of the Siberian this brief review is to reassess the status of these four genera Platform (Jermak, 1986, 1988). There is no longer any and their role in the early history of the class. All other pro- evidence that the first appearance of Pojetaia predates posed Cambrian Bivalvia are either junior synonyms of signficantly the origin of Fordilla (e.g. Jell, 1980); the two Fordilla and Pojetaia or are other kinds of fossils that have taxa probably originated at approximately the same time and been mis-identified as bivalves (Table 1). The most notable then coexisted in different biogeographic regions (Redlichian of the latter kind is Lamellodonta simplex Vogel, which and Olenellian realms; Kobayashi, 1972) for about 10 million Havlicek and Kfiz (1978) showed to be a deformed obolellid years. brachiopod. Tuarangia paparua Mackinnon and T: gravgaerdensis Berg-Madsen occur in approximately coeval late Middle Cam- BIOSTRATIGRAPHY OF CAMBRIAN brian (Boomerangian) strata in New Zealand and Denmark BIVALVES (Berg-Madsen, 1987). Pseudomyona is known only from its type locality in northwestern Queensland which is early Both Fordilla and Pojetaia are middle Early Cambrian Middle Cambrian (Floran) in age (Southgate, 1986). American Malacological Bulletin, Vol. 9(2) (1992):117-122 7 118 AMER. MALAC. BULL. 9(2) (1992) Table 1. List of generic and specific names that have been proposed for fossils considered to be Cambrian bivalves (type species are identified by an asterisk). Only two Early Cambrian species (Fordilla troyensis and Pojetaia runnegari) are certainly bivalves. Our assessment of the remaining taxa, based in most Cases upon an examination of original material, is given in the right hand column. Buluniella Jermak, 1986 B. borealis* Jermak, 1986 Cycloconchioides Zhang, 1980 C. elongatus Zhang, 1980 C. venustus Zhang, 1980 Fordilla Barrande 1881 F. troyensis* Barrande, 1881 F. sibirica Krasilova, 1977 Hebeinella Zhang, 1980 H. formosa* Zhang, 1980 Lamellodonta Vogel, 1962 L. simplex* Vogel, 1962 Oryzoconcha He and Pei, 1985 O. prisca* He and Pei, 1985 Pojetaia Jell, 1980 P. ovata Chen and Wang, 1985 P. runnegari* Jell, 1980 Praelamellodonta Zhang, 1980 P. elegansa* Zhang, 1980 Pseudomyona Runnegar, 1983 Myona queenslandica* Runnegar and Jell, 1976 Tuarangia Mackinnon, 1982 T. paparua* Mackinnon, 1982 Middle Cambrian bivalve? T. gravgaerdensis Berg-Madsen, 1987 Middle Cambrian bivalve? Xianfengoconcha Zhang, 1980 X. elliptica Zhang, 1980 X. rotunda Zhang, 1980 X. minuta Zhang, 1980 Yangtzedonta Yii, 1985 Y. primitva* Yi, 1985 Fordilla troyensis? stenothecoid stenothecoid Early Cambrian bivalve F. troyensis stenothecoid inarticulate brachiopod Pojetaia runnegari P. runnegari Early Cambrian bivalve stenothecoid Middle Cambrian bivalve? stenothecoid stenothecoid stenothecoid unique, unidentified microfossil Berg-Madsen (1987) also illustrated a single specimen of Tuarangia from a glacial erratic in north Poland; the presence of the conodont Westergaardodina tricuspidata Miiller in the same boulder suggested an early Late Cambrian age for the source rock. HIGHER TAXA OF CAMBRIAN BIVALVES When Jell (1980) described Pojetaia runnegari he noted its similarities to Fordilla troyensis (size, shape, cardinal hinge teeth, opisthodetic ligament, etc.) and placed it in the same family (Fordillidae) and order (Fordilloida) as Fordilla. In contrast, Runnegar and Bentley (1983) emphasised similarities between Fordilla and Ordovician isofilibranch bivalves such as Neofordilla Krasilova and Modiolodon Ulrich and they therefore referred Fordilla to the Mytilacea. Pojetaia, on the other hand, was regarded as a primitive nuculoid palaeo- taxodont. This interpretation placed the origin of these two subclasses (Isofilibranchia and Palaeotaxodonta) within the Fordillidae during the Early Cambrian. One of the most distinctive features of Pojetaia is seen on phosphatic internal molds of the shell (Figs. 1D-E). The surface of most internal molds is covered with imbricated polygonal cells that are inclined in a consistent way and become smaller toward the growing margin of the valve. These cells were interpreted as the impressions of the ends of near-vertical prisms formed of aragonite fibers by Run- negar and Bentley (1983) and Runnegar (1985), but this in- terpretation has been challenged by Carter and Clark (1985) and Carter (1990), who regard the structure as the imprints of exceptionally large (30 ym) tablets of nacre. Although we do not agree with this suggestion because nacre tablets are never imbricated, we must admit that the microstructure of the Pojetaia shell is not yet well understood. Nevertheless, a clearly homologous microstructure was present in the Fordilla shell (Figs. 1A-B). A steinkern of F troyensis discovered in Greenland by John S. Peel has the same kind of cellular network as internal molds of Pojetaia, except that in Fordilla the cells are more elongated than they are in Pojetaia. As in Pojetaia, each cell has closely-spaced transverse marks which could be casts of individual mineral (aragonite?) fibers. The importance of this unusual shell microstructure is that it reunites Fordilla and Pojetaia into a monophyletic group (Fig. 2). Thus Wailer (1990) treated this microstruc- ture as an autapomorphy of the Fordilloida and, on that basis, regarded the fordillids as an extinct stem group which diverged from the line leading to modern bivalves prior to the origin of the class. In Waller’s phylogenetic tree all modern bivalve higher taxa are derived directly or sequen- tially from Early Ordovician palaeotaxodonts. It is unlikely that the distinctive shell microstructure of Fordilla and Pojetaia would be lost independently in lines leading to both the Palaeotaxodonta and the Isofilibranchia so the independent derivation of these subclasses from the Fordillidae is not supported by the new microstructural evidence. However, inclined (but near-vertical) aragonitic prisms of the type postulated by Runnegar and Bentley (1983) in Pojetaia have been observed in the outer shell layer of a Devonian nuculoid [Palaeoneilo filosa (Conrad); Carter, 1990: 159-162] and so it is still possible that the Early Ordovician palaeotaxodonts are direct descendants of a fordillid such as Pojetaia. A quite different shell microstructure unites Tuarangia and Pseudomyona (Fig. 3G). Once again, this microstruc- ture is known only from phosphatic replicas of the inner sur- faces of recrystallized shells. It has been interpreted as replicated foliated calcite by comparison with modern ex- amples and crystallographic analysis (Mackinnon, 1982; Run- negar, 1984); it must be a primary feature of the shell rather than a secondary diagenetic artefact for the foliated calcite RUNNEGAR AND POJETA: EARLIEST BIVALVES 119 telah H \ “3 % Fs a H i cb i Notes, x Fig. 1. Replicated shell microstructure of Fordilla troyensis Barrande (A-B) and Pojetaia runnegari Jell (C-E), Early Cambrian bivalves. A, dolomitic internal mold of right valve, Early Cambrian, Greenland, showing replicas of structures interpreted by Runnegar and Bentley (1983) as casts of the ends of composite prisms formed of fibrous aragonite (shell length = 3.25 mm). B, enlargement of antero-ventral part of A. C, exterior of right valve; shell is 1.0 mm long. D-E, scanning electron micrographs of the antero-ventral part of internal mold of right left valve; the images have been electronically inverted to give the impression of the actual structure rather than its negative cast. 120 AMER. MALAC. BULL. 9(2) (1992) Tuarangia Pseudomyona Middle Cambrian Fordilla Pojetaia Early Cambrian foliated calcite shell central adductor muscle lateral hinge teeth amphidetic ligament prismatic aragonite shell posterior adductor muscle cardinal hinge teeth opisthodetic ligament bivalved ancestor? Fig. 2. Tree illustrating the phylogenetic hypotheses discussed in the text. It is concluded that Fordilla Barrande and Pojetaia Jell did not share a com- mon bivalved ancestor with Tuarangia Mackinnon and Pseudomyona Runnegar. stops abruptly at the edges of muscle scars (Fig. 3F) and the trend of the folia is related to their position along the valve margin (Runnegar, 1983). A similar microstructure has been observed in the Middle Cambrian snorkel-bearing univalve Eotebenna Runnegar and Jell (Runnegar and Jell, 1976; Run- negar, 1983) but is otherwise known from fossil molluscs until the (presumably independent) origins of foliated calcite in the patelloid limpets and pectinoid bivalves some time dur- ing the early post-Cambrian Paleozoic. In addition to a foliated calcite shell, 7iarangia and Pseudomyona each had an amphidetic hinge, D-shaped valves and lateral but not cardinal teeth (Fig. 3). They both prob- ably also possessed the central adductor muscle seen in Pseudomyona (Runnegar, 1983). However, it is not clear whether the “‘protoconch’’ of Pseudomyona (Figs. 3A, 3C-D) is homologous with the central ‘‘ligament pit’’ of Tuarangia (Fig. 3E). Mackinnon (1982) and Berg-Madsen (1987) have re- garded Tuarangia and Pseudomyona as early pteriomorphian bivalves belonging to the extinct Order Tuarangiida. They cited the amphidetic ligament, straight hinge, lateral teeth and foliated calcite shell as pteriomorphian characters. The pro- blem with this interpretation is that this collage of characters is not characteristic of the early pteriomorphians so far discovered in Ordovician strata (Pojeta and Runnegar, 1985). DISCUSSION By the end of the Ordovician, at least five of the ex- tant subclasses of bivalves had appeared: Palaeotaxodonta; Isofilibranchia [Boss (1982) and Waller (1990) consider the Isofilibranchia to be a superorder of the Pteriomorphia]; Anomalodesmata; Heteroconchia; Pteriomorphia (Pojeta and Runnegar, 1985). On both stratigraphic and morphological grounds the palaeotaxodonts are regarded frequently as primitive and the direct or indirect ancestors of both the Pteriomorphia and the Heteroconchia (Palaeoheterodonta + Heterodonta) (Pojeta and Runnegar, 1985; Waller, 1990). It is possible (but less likely) that the nuculoid palaeotaxodonts were derived from an actinodont heteroconch (Babin and Gutierrez-Marco, 1991). Waller (1990) used the distinctive microstructure of Pojetaia and Fordilla as a synapomorphy for the Fordilloida, which he regarded as the sister group of the rest of the Bivalvia. He supported this taxonomic decision with the assumption that the fordillids had lost a preexisting nacreous inner shell layer and had not yet acquired a fibrous layer in their ligament. This allowed him to reinstate the palaeotax- odonts as the earliest members of the crown group. Morris (1990) came to a similar conclusion but for less explicit reasons. Although some laterally-compressed Cambrian mol- luscs apparently had prismato-nacreous aragonitic shells (Runnegar, 1985), there is no evidence that nacre was present in (or absent from) the ribeiroid rostroconch ancestors of the bivalves (Runnegar, 1983). Similarly, the ligaments of Pojetaia and Fordilla are not preserved and so could have been fibrous, granular or unmineralized; in any case, the ligament of nuculids is not fibrous and Waller (1990) has therefore sug- gested that the granular ligament of Nucula Lamarck is a derived condition. [Based on evidence obtained from well- preserved Devonian nuculoids, Carter (1990) also proposed that the granular ligament of Nucula is derived from an ancestral weakly-mineralized to non-mineralized condition. ] Given these and other uncertainties we tentatively maintain the fordillids within the crown group for the time being. As mentioned above, it is possible that Pojetaia was an early palaeotaxodont. Pojeta and Runnegar (1985) recognized four major kinds of Ordovician pteriomorphs: pterineid pteriaceans; cyr- todontids; ambonychiids; a probable ancestral limid, Pro- lobella? Ulrich. Under some existing classifications the pterineids and the limid would be placed together in the order Pterioida but this grouping makes little sense in an Ordovician context because the pectiniform shell of Prolobella? is unlike- ly to be homologous with the shells of younger Pectinacea. We therefore agree with Waller (1978) and Johnston (1991) who assigned the ambonychiids and limoids to the superorder Prionodonta, sensu Boss (1982). The duplivincular ligament of the pterineids, cyrto- dontids, and ambonychiids strongly suggests that they con- stitute a monophyletic group (Prionodonta + Eupterio- morphia). Many authors have considered the cyrtodontids to RUNNEGAR AND POJETA: EARLIEST BIVALVES 121 Fig. 3. Middle Cambrian bivalved molluscs, Pseudomyona queenslandica (Runnegar and Jell) (A, C-D, F-G) and Tuarangia gravgaerdensis Berg-Madsen (B, E). A, C-D, phosphatic internal mold viewed from right? side and posterior? and anterior? ends; note lateral teeth (arrows) and univalved protoconch. B, internal mold of right? valve showing well-developed lateral teeth. E, end-on view of internal mold showing crenulations caused by interlocking teeth and central ‘‘ligament pit’’. F, surface of internal mold showing casts of overlapping calcite folia (upper left), edge of adductor muscle scar (ink line), and smooth surface of adductor scar; polygons near edge of adductor scar are interpreted as casts of myostracal prisms. G, enlargement of replicated foliated calcite on surface of internal mold. be the most primitive pteriomorphs and they derive them either from the cycloconch actinodonts (Pojeta and Runnegar, 1985; Johnston, 1991) or ‘‘actinodont’’ palaeotaxodonts (Waller, 1990). Johnston (1991) has described a Silurian eupteriomorph (Umburra cinefacta Johnston) which he regards as ‘‘more primitive dentally’’ than any known cyr- todontid. If this were true, then the common ancestor of the prionodonts and the eupteriomorphs would have had an equivalved shell, a duplivincular ligament and ‘‘actinodont’’ hinge teeth. Although this hypothetical ancestor approximates some of the morphological features of Tuarangia and Pseudo- myona, it almost certainly would have had anterior and posterior adductor muscles (Johnston, 1991) rather than the single central adductor muscle of Pseudomyona and a 122 AMER. MALAC. BULL. 9(2) (1992) nacreous or crossed-lamellar aragonitic shell (Carter, 1990) instead of the foliated calcite shell of Pseudomyona and Tuarangia. Thus the postulated genetic connection between Tuarangia/Pseudomyona and the Pteriomorphia (Mackinnon, 1982; Berg-Madsen, 1987) remains tenuous and is not sup- ported by the evidence currently available. As there is no character apart from the bivalved condition in common be- tween the fordillids and either Pseudomyona or Tuarangia it is difficult to sustain the hypothesis that they once shared a common bivalved ancestor (Fig. 2). On the contrary, the presence of foliated calcite in the pseudobivalved univalve Eotebenna pontifex Runnegar and Jell (Runnegar and Jell, 1976) is an indication that Pseudomyona and Tuarangia are not true bivalves and that the Class Bivalvia is therefore monophyletic. ACKNOWLEDGMENTS We thank Dr. J. S. Peel, Geological Survey of Greenland, Copenhagen for the specimen of Fordilla troyensis illustrated in figure 1 and Dr. Vivianne Berg-Madsen, Geological Institute, Stockholm University for the photographs of Tuarangia gravgaerdensis. Mr. Paul Belasky kindly translated some of the Russian papers. Dr. Joseph G. Carter and an anonymous reviewer pro- vided helpful suggestions for improvements to the manuscript. LITERATURE CITED Babin, C. and J-C. Gutierrez-Marcos. 1991. Middle Ordovician bivalves from Spain and their phyletic and paleogeographic significance. Palaeon- tology 34, 109-147. Berg-Madsen, V. 1987. Tuarangia from Bornholm (Denmark) and similarities in Baltoscandian and Australasian late Middle Cambrian faunas. Alcheringa 11:245-259. Bengston, S., S. Conway Morris, B. J. Cooper, P. A. Jell and B. Runnegar. 1990. Early Cambrian fossils from South Australia. Memoirs of the Association of Australasian Palaeontologists 9, 1-364. Boss, K. J. 1982. Mollusca. Jn: Synopsis and Classification of Living Organisms, S. P. Parker, ed., pp. 945-1166. McGraw-Hill, New York. Carter, J. G. 1990. Evolutionary significance of shell microstructure in the Palaeotaxodonta, Pteriomorpha and Isofilibranchia (Bivalvia: Mollusca). /n: Skeletal Biomineralization: Process and Evolutionary Trends, Volume I, J. G. Carter, ed., pp. 135-296. Van Nostrand Reinhold, New York. Carter, J. G. and G. R. Clark. 1985. Classification and phylogenetic significance of molluscan shell microstructure. Jn Mollusks. Notes for a Short Course, T. W. Broadhead, ed., pp. 50-71. University of Tennessee Studies in Geology 13, Knoxville. Chen, Y. and Z. Wang. 1985. A bivalve of the Lower Cambrian Xinji For- mation in western Henan Province. Earth Science - Journal of the Wuhan College of Geology 10(2):27-29. Havlicek, V. and J. Kriz. 1978. Middle Cambrian Lamellodonta simplex Vogel: “‘bivalve’’ turned brachiopod Trematobolus simplex (Vogel). Journal of Paleontology 52, 972-975. He, T. and F. Pei. 1985. The discovery of bivalves from the Lower Cam- brian Xinji Formation in Fangcheng County, Henan Province. Journal of the Chengdu College of Geology 1985 (1):61-66 (in Chinese with English summary). Jell, P. A. 1980. Earliest known pelecypod on Earth - a new Early Cam- brian genus from South Australia. Alcheringa 4:233-239. Jermak, V. V. 1986. Rannekembriyskie fordillidy (Bivalvia) severa sibirskoy platformy [Early Cambrian Fordillidae (Bivalvia) from the northern Siberian platform]. Trudy Sibirskoe Otdelenie Institut Geologogii i Geophysiki Akademiya Nauk SSSR 669:183-188. Jermak, V. V. 1988. Stroenie zamochnogo apparata, mikrostruktura rakoviny i obraz zhizni rannekembriyskikh fordillid (Bivalvia) [Structure of hinge apparatus, shell microstructure and mode of life of Early Cambrian fordillids (Bivalvia)]. Trudy Sibirskoe Otdelenie Institut Geologogii i Geophysiki Akademiya Nauk SSSR 720:179-184. Johnston, P. A. 1991. Systematics and ontogeny of a new bivalve, Umburra cinefacta, from the Silurian of Australia: implications for pterio- morphian evolution. Alcheringa 15, 293-319. Kobayashi, T. 1972. Three faunal provinces in the early Cambrian Period. Proceedings of the Japan Academy 48, 242-247. Mackinnon, D. I. 1982. Tuarangia paparua n. gen. and n. sp., a late Mid- dle Cambrian pelecypod from New Zealand. Journal of Paleontology 56:589-598. Morris, N. J. 1990. Early radiation of the Mollusca. Jn: Major Evolutionary Radiations, P. D. Taylor and G. P. Larwood, eds. pp. 73-90. Claren- don Press, Oxford. Pojeta, J. 1975. Fordilla troyensis Barrande and early pelecypod phylogeny. Bulletins of American Paleontology 67:363-384. Pojeta, J. and B. Runnegar. 1985. The early evolution of diasome molluscs. In: The Mollusca, Volume 10, E. R. Trueman and M. R. Clarke, eds., pp. 295-336. Academic Press, Orlando. Runnegar, B. 1983. Molluscan phylogeny revisited. Memoirs of the Associa- tion of Australasian Palaeontologists 1:121-144. Runnegar, B. 1984. Crystallography of the foliated calcite shell layers of bivalve molluscs. Alcheringa 8:273-290. Runnegar, B. 1985. Shell microstructure of Cambrian molluscs replicated by phosphate. Alcheringa 9:245-257. Runnegar, B. and C. Bentley. 1983. Anatomy, ecology and affinities of the Australian Early Cambrian bivalve Pojetaia runnegari Jell. Journal of Paleontology 57:73-92. Southgate, P. N. 1986. The Gowers Formation and Bronco Stromatolith Bed, two new stratigraphic units in the Undilla portion of the Georgina Basin. Queensland Government Mining Journal 85:407-4l1. Vogel, K. 1962. Muscheln mit Schloiszahnen aus dem spanischen Kambrium und ihre Bedeutung fiir die Evolution der Lamellilbranchiaten. Akademie der Wissenschaften und der Literatur in Mainz, Abhandlungen der Mathematisch-Naturwissenschaftlichen Klasse 1962 (4):194-244. Waller, T. R. 1978. Morphology, morphoclines and a new classification of the Pteriomorpha. Philosophical Transactions of the Royal Society of London B284:345-365. Waller, T. R. 1990. The evolution of ligament systems in the Bivalvia. In: The Bivalvia — Proceedings of a Memorial Symposium in Honour of Sir Charles Maurice Yonge, Edinburgh, 1986, B. Morton, ed. pp. 49-71. Hong Kong University Press, Hong Kong. Yu, W. 1985. Yangtzedonta — a problematica Bivalvia from the Meishucunian Stage, China. Acta Micropalaeontologica Sinica 2:401-408 (in Chinese with English summary). Zhang, R. 1980. On the earliest bivalve fauna — bivalves from Lower Cam- brian Tianheban Formation, Xianfeng, Hubei. /n: Scientific Papers on Geology for International Exchange prepared for the 26th Inter- national Geological Congress. 4, Stratigraphy and Palaeontology, Publishing House of Geology, Beijing, pp. 121-129 (in Chinese with English summary). Date of manuscript acceptance: 16 January 1992 Systematics evolution and distribution of mussels belonging to the genus Mytilus: an overview Raymond Seed School of Ocean Sciences, University of Wales-Bangor, Menai Bridge, Gwynedd, LL59 SEY, United Kingdom Abstract. Despite their scientific and commercial interest and their widespread distribution throughout the cooler waters of both northern and southern hemispheres, the taxonomy of mussels belonging to the genus Mytilus remains controversial. This paper reviews the systematics of this group, albeit with particular emphasis on the smooth-shelled mussels of the M. edulis complex, and stresses throughout the need for a multidisciplinary approach. Multivariate analysis of allozyme and morphometric data obtained for mussels worldwide now provides compelling evidence for the existence of three distinct evolutionary lineages: M. edulis; M. galloprovincialis; M. trossulus. No single taxonomic character discriminates unequivocally among these taxa though certain characters, either individually or in combination, are virtually diagnostic. All three lineages occur in northern waters but only M. edulis and M. galloprovincialis have so far been recorded from the southern hemisphere. Whether these taxa are accorded full specific status will require an agreed operational definition of biological species. Future research should focus on the biological mechanisms that maintain the distinctive characteristics of these mussels across vast distances despite the occurrence of hybridisation and the massive potential for larval dispersal. The origin, evolution and distribution of mussels within the genus are discussed. The genus Mytilus is one of the most cosmopolitan of all marine genera, occurring at higher latitudes in all oceans and major seas of both northern and southern hemispheres. It is found intertidally and subtidally, in estuarine and fully saline habitats, attached by means of byssal threads to a wide variety of hard or semiconsolidated substrata. In view of its widespread distribution, as well as its scientific and commer- cial importance, it is perhaps surprising that the taxonomy and systematics of this extensively studied genus still remains a somewhat controversial issue (e.g. Gosling, 1984; McDonald et al., 1991). Much of the early taxonomy of Mytilus was based sole- ly on morphological features, particularly those pertaining to the shell. However, ontogenetic and environmentally in- duced variation in shell characteristics (e.g. Seed, 1968, 1973, 1978; Lewis and Seed, 1969), combined with the complex interactions that are now known to exist among several taxa within this genus, has produced an extremely confused and largely erroneous taxonomy (Koehn, 1991). In a comprehen- sive review of the genus, Lamy (1936) recognised the follow- ing smooth-shelled mussels as distinct species: M. edulis Linnaeus from north temperate waters; M. galloprovincialis Lamarck from the Mediterranean Sea; M. trossulus Gould from the Pacific coast of North America, M. chilensis Hupé and M. platensis Orbigny from the east and west coasts of South America respectively; M. planulatus Lamarck from Australia and New Zealand. He also described M. desola- tionis Lamy (= M. kerguelensis Fletcher) from the Kerguelen islands in the southern Indian Ocean. These taxa, however, were reported by Soot Ryen (1955) as geographical subspecies or races of the M. edulis species complex. Other taxa previously considered to be subspecies of M. edulis include the Californian bay mussel, M. diegensis Coe (Soot Ryen, 1955) and M. aoteanus Powell from New Zealand (Fleming, 1959) together with M. kussakini and M. zhirmunskii from the Pacific coast of Asia (Scarlato and Starobogatov, 1979). Mytilus californianus Conrad, a distinctively different species of large body size and divergent ecology to M. edulis, is identified readily by the presence of radiating ribs on the shell (Soot Ryen, 1955). M. coruscus Gould (= M. crassitesta Lischke) is a thick-shelled, ribbed mussel with minute crenulations along the ventral margin close to the apex (Kira, 1962). Unfortunately, however, we have little or no detailed information regarding this mussel and its systematic status thus remains uncertain. Recently, Vermeij (1989) has specu- lated that M. californianus and M. coruscus could in fact com- prise a single species with geographical variations in the prominence of the radiating ribs. However, because both of these mussels are distinguished easily from the smooth- shelled mussels of the M. edulis group, they will not be con- sidered in any detail in this paper. The use of enzyme electrophoresis to characterise in- dividual and population differences in genetic composition, together with multivariate techniques applied to both enzyme and morphometric phenotypes, have assisted greatly in elucidating the systematics and taxonomic status of species of smooth-shelled mussels (e.g. McDonald and Koehn, 1988; Varvio et al., 1988; McDonald et al., 1991). Although three American Malacological Bulletin, Vol. 9(2) (1992):123-137 123 124 AMER. MALAC. BULL. 9(2) (1992) taxa have been identified (Mytilus edulis, M. galloprovincialis and M. trossulus), hybridisation has been reported at most locations where the ranges of these mussels coincide and con- sequently this has led to considerable speculation regarding their taxonomic status (e.g. Skibinski et al. , 1983; Gosling, 1984, 1992; McDonald and Koehn, 1988; Johannesson et al. , 1990; Vain6dla and Hvilsom, 1991). In this paper I shall docu- ment briefly the evidence for the existence and distribution of these three relatively distinct mussels, albeit with particular emphasis on the taxonomic validity of the Mediterranean mussel M. galloprovincialis, which was originally thought to be restricted to European coasts but which now appears to be far more widely distributed (e.g. Wilkins et al. , 1983; Lee and Morton, 1985; Grant and Cherry, 1985; McDonald and Koehn, 1988; McDonald ef al., 1991). Much less in- formation is available currently concerning the systematics, distribution and ecological characteristics of M. trossulus. SYSTEMATIC CHARACTERISATION OF MYTILUS A) ENZYME ELECTROPHORESIS: Allozyme characters have assisted greatly in clarifying the complex biosystematics of the genus Mytilus. Despite the large number of enzymes that are potentialy available for study, in practice only a few have sufficiently high levels of variation to be of significant taxonomic value (e.g. Ahmad ef al., 1977). Earlier studies on the M. galloprovincialis-M. edulis complex (reviewed by Gosling, 1984) used various combinations of six loci; esterase D (Est-D), leucine aminopeptidase (Lap-1), glucose phosphate isomerase (Gpi), aminopeptidase (Ap), peptidase 2 (Lap-2) and phosphoglucomutase (Pgm). More recently, octopine dehydrogenase (Odh) and mannose phosphate isomerase (Mpi) have also been incorporated into the suite of enzymes used to differentiate between these smooth-shelled mussels (e.g. Skibinski, 1983; Grant and Cherry, 1985; Varvio et al. , 1988; McDonald and Koehn, 1988). None of these loci, however, discriminate unequivocally between M. edulis and M. galloprovincialis, but according to Varvio et al. (1988) the Mpi locus is ‘virtually diagnostic’. McDonald and Koehn (1988) similarly found that Mpi was diagnostic in almost all the allopatric populations of Mytilus that they studied although a combination of other loci with large differences in allele frequency could also effectively discriminate between dif- ferent taxa. When a combination of four allozyme loci were used, Sanjuan et al. (1990) found that the probability of misclassification was exceedingly low (1.5 x 10-7); indeed 99% of all individual mussels in their samples could be assigned correctly on the Est-D genotype alone. Although a less well studied enzyme, leucyl glycyl glycine peptidase is also reported to provide an almost perfect discrimination between M. edulis and M. galloprovincialis (Grant and Cherry, 1985). The principal loci used in studies of Mytilus genetics, especially those which have proved to be most valuable in taxonomic studies, are comprehensively reviewed by Gosling (1992). Beaumont et al. (1989) examined allele frequencies at three loci (Est-D, Mpi, Odh) in mixed populations of Mytilus edulis and M. galloprovincialis from two physically contrasted sites, Rock and Polzeath, in the Camel estuary in south-west England, and a pure population of M. galloprovincialis from Langebaan lagoon on the west Cape coast of South Africa. Their results, summarised in Table 1, reveal markedly dif- ferent allele frequencies between these two mussels, par- ticularly with respect to the Est-D and Mpi loci. At Rock, the Mpi locus proved to be less effective at differentiating M. edulis and M. galloprovincialis than the Est-D locus, whilst both of these loci were rather poor discriminators in the South African and Polzeath populations. Odh genotypes did not appear to be particularly good discriminating characters in any of the populations studied though the data of Varvio et al. (1988) did allow clear discrimination of M. galloprovincialis populations on the basis of Odh allelic com- position. A further feature of the Rock mussel population was the disparity in the percentage of M. edulis compared to M. galloprovincialis that were misidentified by the Mpi locus. This is owing to the fact that the Mpi® allele (the characteristic M. galloprovincialis allele) was present in the M. edulis population at a frequency of 0.197 but the reverse was not true as the M. edulis allele, Mpi'®°, was present in the M. galloprovincialis population at a frequency of only 0.053 (Table 1). Two salient features would thus appear to emerge from the use of single locus genotypes as characters for discriminating between Mytilus edulis and M. galloprovin- cialis. Firstly, a locus can give good discrimination in one Table 1. Allele frequencies at three loci in sympatric populations of Mytilus edulis and M. galloprovincialis from Rock and M. galloprovincialis from Polzeath and South Africa (after Beaumont ef al., 1989). Allele frequencies Locus Alleles Rock S. Africa Polzeath (relative M. edulis M. gall. M. gall. M. gall. mobility) Esterase-D 60 0.014 _ — _— (Est-D) 82 0.021 0.941 0.802 0.360 100 0.936 0.059 0.198 0.640 118 0.029 _ —_— — Mannose 63 0.197 0.947 0.882 0.722 phosphate 100 0.796 0.053 0.118 0.278 isomerase 133 0.007 — — — (Mpi) Octopine 60 0.007 0.006 — 0.015 dehydro- 70 0.111 0.530 0.540 0.634 genase 77 0.014 = 0.030 — (Odh) 100 0.799 0.226 0.120 0.227 106 0.014 0.006 — 0.015 112 0.055 0.232 0.310 0.109 SEED: SYSTEMATICS EVOLUTION AND DISTRIBUTION OF MYTILUS 125 population but poorer discrimination in another, and, second- ly, a locus may be diagnostic for one species, but not the other, within any single mussel population (Beaumont et al., 1989). Variations in genotype, whether on a local or geographical scale, could of course reflect differential patterns of en- vironmental selection rather than distinct evolutionary back- grounds (e.g. Murdock ef al., 1975; Koehn ef al., 1980; Gartner-Kepkay ef al., 1983; Johannesson er al., 1990; Tedengren et al. , 1990). The marked differences in allele fre- quencies reported for the mussel populations at Rock, however, cannot easily be attributed to such causes since these mussels occur within mixed clumps and are thus presumably subjected to identical environmental conditions. On the basis of five allozyme loci, Koehn e7 al. (1984) were able to separate samples of putative Mytilus edulis from several sites throughout eastern North America into three distinguishable groups, though one of these involved separa- tion at a single locus (Lap) and was not, therefore, thought to represent a distinct taxonomic group. The other two groups, however, were very different at several loci and this led these authors to suggest that one of these groups represented a hitherto unrecognised species. Subsequently, this was given additional support by Varvio ef al. (1988) who showed that this mussel was most similar to Mytilus from the Baltic Sea (see also Bulheim and Gosling, 1988) and which is now recognised as M. trossulus, a species reported previously only from parts of the Pacific coast of North America (e.g. McDonald and Koehn, 1988). More recently McDonald eft al. (1991) used a multivariate technique to analyse the electrophoretic data at eight loci in over a thousand mussels collected from a total of 45 sites in the northern and southern hemispheres. Allozyme data (71 characters) were reduced and displayed using principal component analysis which locates the ortho- gonal axes accounting for the greatest amount of variation in the multidimensional space. This analysis defined three distinct clusters of individuals in the northern hemisphere samples but only two clusters in southern hemisphere mussels, albeit with some intermediate individuals mainly from those sites where these mussels come into contact and hybridisation occurs (Figs. 1A,B). Each cluster in the north- ern hemisphere could be assigned to an extant species, Mytilus edulis, M. galloprovincialis or M. trossulus, based on the examination of principal component scores of individuals from locations where the identity of mussels had been designated previously (e.g. McDonald and Koehn, 1988; McDonald et al., 1990). The Mytilus edulis cluster in the southern hemisphere, comprising mussels from South America, the Falklands and Kerguelen islands (= ‘‘South American mussels’’), were most similar to northern hemisphere M. edulis, although they did contain alleles that were characteristic of all three north- ern mussels. The reason for this is that many loci of these ‘*South American’’ mussels contained alleles that in northern hemisphere mussels were common only in M. gallopro- vincialis or M. trossulus. Blot et al. (1988) has similarly shown that mussels from the Kerguelen islands were more similar genetically to northern M. edulis than to M. galloprovincialis. Mussels from Australia, Tasmania and New Zealand (= ‘‘Australian mussels’’) formed the second south- ern hemisphere cluster with principal component scores similar to M. galloprovincialis from the northern hemisphere, though once again there were some differences in allele fre- quencies, particularly at the Mpi and Est-D loci. Mussels from South Africa have similar allelic frequencies to M. gallopro- vincialis from the Mediterranean Sea and south-west England (Grant and Cherry, 1985; Beaumont et al., 1989; Table 1). B) MORPHOMETRIC CRITERIA: Overall shell morphol- ogy in Mytilus is subject to considerable phenotypic varia- tion. Such environmental control of shape (and growth rate) is readily demonstrated by transplanting mussels from one habitat to another and recording the resulting changes in morphology [see Seed and Richardson (1990) and references therein]. Moreover, such environmentally induced variations are further confounded by ontogenetic changes in shape brought about by allometric growth (Seed, 1968, 1973, 1978, 1980). Similar trends are exhibited by both Mytilus edulis and M. galloprovincialis resulting in a considerable degree of con- vergence so that, in some populations, shell characters merge until identification on gross morphology alone becomes dif- ficult or impossible. Intermixing of morphological characters in sympatric populations could also be due to hybridisation between these two mussels (e.g. Seed, 1972, 1974). In some populations, however, differences in gross shell morphology between M. edulis and M. galloprovincialis can be extreme- ly pronounced. At Rock, for instance, where these two mussels occur in mixed populations, M. galloprovincialis has a significantly taller shell with a steeper ligamentary angle (the angle subtended by the ventral and ligamentary margins, see Fig. 2A) than M. edulis. Maximum shell width lies closer to the ventral margin and consequently the ventral aspect of M. galloprovincialis is much flatter when viewed in cross section (Figs. 2B, C). Typically M. galloprovincialis has a more pointed, beaked or ventrally incurved shell with a rather triangular shaped outline whereas M. edulis is more round- ed anteriorly, has a more elongate, cylindrical shell with a straight or even slightly convex ventral margin. Thus, at Rock, several shell features combine within single individuals to produce mussels which are quite distinctive in their overall external appearance (Figs. 3H, I). Furthermore, these dif- ferent morphologies are maintained amongst all size ranges of mussels strongly suggesting that they are, in fact, genetical- ly rather than environmentally controlled (Beaumont ef al. , 1989; Seed, 1990). Many of these features also recur in mussels from different parts of their geographical range (Figs. 126 AMER. MALAC. BULL. 9(2) (1992) -4-3-2-10 123 45 6 Principal component 1 N —_— Cc ® Cc ° Q = fo) (°) a M. trossulus Qa oO Cc = oO =4=3-2=-10 12 3 4 5.6 M. galloprovincialis N ® ~ © c © > © = Cc fe) Cc © O O 2 4 6 8 10 6) 2 4 6 8 10 12 Canonical variate 1 Fig. 1. First and second principal components of allozyme data for A, northern hemisphere, and B, southern hemisphere mussels. To aid visual comparison outlines have been drawn subjectively around the northern hemisphere clusters. C, D, First and second canonical variates of morphometric data for northern and southern hemisphere mussels respectively. Lines separating the northern hemisphere clusters have been drawn to aid comparison (after McDonald et al., 1991). 3, 4). In the mussel populations studied by Beaumont et al. (1989) the anterior adductor muscle to shell length ratios ({aams/sl]x10) in Mytilus edulis were consistently and significantly larger and the elongated scar more conspicuous than in its congener. The dark blue hinge plate in M. edulis is typically a more gently curved structure whereas in M. galloprovincialis it is usually paler in colour and describes a much tighter arc with the posterior end more closely delimited from the adjacent shell margin (Figs. 2B, C). Both the hinge plate to shell length ratio ({hp/sl]x10) and the length to width of the posterior byssal retractor scar (lbrs/wbrs) are significantly larger in M. edulis; in the latter ratio this is due almost entirely to variations in scar width rather than scar length. On virtually all of the morphometric criteria used by Beaumont et al. (1989), M. galloprovincialis at Rock were statistically indistinguishable from conspecifics from Polzeath and South Africa. Frequency distributions of several of these morphometric characters are illustrated in figure 5 and show that whilst the mean values between these two mussels are markedly different, there is, nevertheless, a considerable degree of overlap in the ranges of these individual shell characters. The value of the anterior adductor muscle scar and hinge plate as taxonomic characters for separating Mytilus edulis and M. galloprovincialis has been reported by several workers. In most of these studies (e.g. Lewis and Seed, 1969; Seed, 1978; Wilkins et al., 1983; Grant and Cherry, 1985; SEED: SYSTEMATICS EVOLUTION AND DISTRIBUTION OF MYTILUS 27 1 it = sw(x0.5) = Fig. 2. A, Terminology of shell characters: a, position of maximum shell width along the dorso-ventral axis; aams, anterior adductor muscle scar; dm, dorsal margin; hp, hinge plate; Ibrs, length of byssal retractor muscle scar; Im, ligamentary margin; pm, posterior margin; sh, shell height; sl, shell length; sw, shell width; vm, ventral margin; wbrs, width of byssal retrac- tor muscle scar. Anterior end and transverse profiles of B, Mytilus gallo- provincialis and C, M. edulis (after Beaumont et al., 1989). Lee and Morton, 1985) these characters have been considered separately but Verduin (1979) and Sanjuan ef al. (1990) achieved a more effective separation when these were com- bined into a single taxonomic index. Other taxonomic characters previously used to separate these two mussels in- clude the colour of the mantle edge, which is typically yellowish-brown in M. edulis and deep purple-violet in M. galloprovincialis, and the presence (M. edulis) or absence (M. galloprovincialis) of longitudinal rays of deeper colour in the shell (e.g. Hepper, 1957; Lewis and Seed, 1969). By using four shell characters together with mantle edge colour and the genotypes of three enzyme loci (see p. 124) to identify Mytilus edulis and M. galloprovincialis Beaumont et al. (1989) were then able to test the reliability of each individual character against a final identification based on all eight characters. Table 2 shows the percentage of mussels that would have been misidentified using single tax- onomic characters. The main point to emerge from this analysis was that no single character existed which allowed the certain identification of all mussels within these three populations. Overall, however, certain characters were clearly more reliable than others, though the diagnostic value of each character varied, sometimes quite markedly, both within and between sites. This applied equally to both morphometric and genetic characters. On average, single locus genotypes proved to be somewhat poorer diagnostic characters than the polygenic morphometric characters though significant dif- ferences between populations were more easily detected by the electrophoretic than by the morphometric data (Beaumont et al., 1989). It is clear from the above that individual morphological characters in Mytilus can vary, often on an exceedingly localised scale, and are therefore of limited taxonomic value though certain characters, or combinations of characters, do permit the separation of these two mussels with a high degree of confidence at least in certain populations. Multivariate techniques, on the other hand, have proved to be more suc- cessful in discriminating between mussels within the M. edulis species complex. Using a canonical variates analysis of 19 different morphometric characters in those samples from northern hemisphere locations where allozyme analysis had indicated previously the presence of a single species, McDonald et al. (1991) were able to resolve three distinct clusters corresponding to M. edulis, M. galloprovincialis and M. trossulus (Fig. 1C). Somewhat surprisingly, the best discrimination was between M. edulis and M. galloprovin- cialis, a long standing taxonomic problem in this genus. Canonical variates analysis finds the linear functions of the morphological variables with coefficients that maximize the distance between groups that have been previously identified using some other criteria, in this case allozyme characters. When the functions from the canonical variates analysis of northern mussels were applied to southern hemisphere samples, southern M. edulis was found to be morphologically intermediate between northern M. edulis and M. trossulus; southern and northern M. galloprovincialis, by contrast, were remarkably similar to each other (Fig. 1D). Characters wich have been considered previously useful for distinguishing M. edulis and M. galloprovincialis, such as the adductor mus- cle scar and hinge plate, also contributed most to the canonical variates analysis. Thus, whilst some overlap occurred in the canonical variates, most individual mussels in these pure samples could be identified from shell characters alone when multivariate functions of all 19 morphometric variables were used. In- dividual characters, on the other hand, even those which are known to show the greatest variation between taxa, exhibited considerable overlap when these were considered singly. McDonald et al. (1991) also calculated canonical functions for each pair of northern mussels because all of the known areas of overlap between these mussels involve only two taxa. Results indicate that a linear combination of all characters in the canonical variate gives total separation in the case of Mytilus edulis and M. galloprovincialis (Fig. 6A) and an almost total separation of M. edulis and M. trossulus (Fig. 6B). For M. galloprovincialis and M. trossulus, which share several morphological traits, particularly with regard to their overall shell shape (Figs. 3, 4) and small size of the anterior adductor scars and hinge plates, there was a somewhat greater 128 AMER. MALAC. BULL. 9(2) (1992) Fig. 3. Mytilus trossulus from: A, Tillamook, Oregon; B, Newport, Oregon. M. edulis from: C, Stony Brook, New York; D, Portland, Maine; E, Aarhus, Denmark, F, Falkland Islands; G, Mar del Plata, Argentina; H, Rock, S. W. England and M. galloprovincialis from: I, Rock, S. W. England (scale bar in cm). degree of overlap (Fig. 6C). The posterior byssal retractor scar of M. trossulus, however, is characteristically much nar- rower than that of M. galloprovincialis of comparable shell length. Further research is now required to determine whether the morphometric differences described by McDonald et al. (1991) for pure mussel samples persist in areas of overlap and hybridisation. Moreover, by incorporating additional mor- phological characters into the multivariate analysis it would seem likely that an even better discrimination of these mussels could be achieved. C) OTHER CRITERIA: Quite apart from the genetic and morphometric differences described above, Mytilus edulis and M. galloprovincialis are also known to vary in several other important respects. Figure 7 shows that, at Rock, spawning in M. edulis occurs mainly during May and June whereas M. galloprovincialis does not spawn until late July or August when seawater temperatures for this geographical locality are maximal. The cyclical pattern of reproduction is also less pro- nounced in M. galloprovincialis with significant proportions of fully ripe individuals persisting throughout much of the year. These differences are documented in detail elsewhere (Seed, 1971) but, in summary, extensive hybridisation at this particular site in south-west England seems most unlikely, a conclusion which is broadly supported by electrophoretic data (e.g. Skibinski er al. , 1983; Beaumont et al. , 1989). Tem- poral differences in spawning activity between these mussels have been reported similarly at another site in south-west England (Croyde) where M. galloprovincialis also had a greater estimated annual fecundity than M. edulis (Gardner and Skibinski, 1990); at a second site (Whitesand), however, there was a higher degree of genetic mixing resulting from reduced levels of variability in the timing of spawning and fecundity (Fig. 8). Hybridisation can be induced artificially in the laboratory and when Mytilus edulis and M. galloprovincialis are crossed they produce fertile hybrids which can then backcross to the parent form to produce viable offspring (Lubet et al. , 1984). There would appear to be little evidence, therefore, of any absolute reproductive barrier or genetic in- SEED: SYSTEMATICS EVOLUTION AND DISTRIBUTION OF MYTILUS 129 Fig. 4. Mytilus galloprovincialis from: A, Los Angeles, California; B, Victoria Harbour, Hong Kong; C, Albany, Western Australia; D, Huon River Estuary, Tasmania; E, Venice, Italy; F, West Cape coast, S. Africa; G, Newquay, S. W. England; H, Polzeath, S. W. England; I, Parede, Portugal; J, Vigo, Spain (scale bar in cm). compatibility between these two mussels. However, whilst Lubet ef al. (1984) apparently were unable to detect any adverse effects on viability, growth or mortality amongst the F' hybrids, recent research has shown that the mortality rates of hybrid larvae can be substantially higher than those of pure M. edulis or pure M. galloprovincialis larvae (Table 3). Mytilus edulis and M. galloprovincialis can exhibit markedly different levels of infection by certain parasitic organisms (e.g. Seed, 1969, 1978; Coustau ef al., 1990; Hillman, 1990). Such differences appear to have a genetic rather than an ecological basis, and because these parasites can influence fitness through their effects on fecundity and condition, they could provide a potentially important selec- tive force in sympatric mussel populations. Table 4 shows that on average approximately 30% of the M. edulis popula- tion at Rock is infested by the peacrab Pinnotheres pisum Penn. whereas in M. galloprovincialis the level of infesta- tion is less than 2%. Several immunological (e.g. Bisignano et al., 1980; Brock, 1985), histopathological (e.g. Hillman, 1990) and chromosomal (e.g. Thiriot-Qui¢vreux, 1984; Dixon and Flavel, 1986; Pasantes ef al/., 1990) investigations are available for Mytilus, albeit with somewhat equivocal results. Significant differences in sperm size and morphology also have been described (e.g. Drozdov and Reunov, 1986; Hodgson and Bernard, 1986; Crespo ef al., 1990). Whilst the cytological differences reported within this genus are clearly insufficient to prevent hybridisation, they could, nonetheless, be partially responsible for maintaining species separation and could also presumably serve as useful tax- onomic characters. Only recently has the analysis of mitochondrial DNA variation been used in taxonomic studies of marine mussels (e.g. Skibinski, 1985; Blot ef al., 1990). Several pure and mixed populations of Mytilus edulis and M. galloprovincialis have been studied (e.g. Edwards and Skibinski, 1987; Fisher and Skibinski, 1990) and, whilst significantly different mtDNA genotypes were reported, none was perfectly diag- nostic. There is little evidence, therefore, to suggest that 130 AMER. MALAC. BULL. 9(2) (1992) Table 2. Percentage of mussels which would have been misidentified using single taxonomic characters (from Beaumont et al. , 1989). Overall Mantle n shape aams! hp? colour Raying Mpi Est-D Odh i) Rock: Mytilus edulis 64 0 0 4.7 0 17.2 23.8 3.2 25.0 M. galloprovincialis 76 34.2 17.1 10.5 6.6 1.3 7.9 9.4 36.5 ii) S. Africa: M. galloprovincialis 38 234 5.3 10.5 0 0 21.1 28.9 21.1 ili) Polzeath: M. edulis 3 0 0 0 0 33.3 66.6 0 33.3 M. galloprovincialis 81 11.3 19.4 9.7 0 322 48.4 88.7 32.3 Mean 16.8 11.8 88 1.9 6.1 26.7 35.1 298 '52 Anterior adductor muscle scar and hinge plate, respectively. mtDNA variation provides any greater overall diagnostic power than allozyme variation in distinguishing between the different forms of Mytilus though mtDNA studies should ultimately lead to an improved understanding of both the population biology and taxonomy of this genus (Edwards and Skibinski, 1987). ORIGINS AND DISTRIBUTION With a geological record extending back for less than two million years, the genus Mytilus is of relatively recent origin (Seed, 1976). Amongst the smooth-shelled taxa, M. edulis generally is considered to be the ancestral species ap- 50 30 30 10 55 95 135 (aams/sl)x10 ance Percent oO Q parently having evolved from some more primitive infaunal or semi-infaunal modiolid stock (Stanley, 1972; Seed, 1990). M. edulis is widely distributed throughout the temperate latitudes of both hemispheres. In the northern hemisphere it occurs along the eastern seaboard of North America as far south as Cape Hatteras in North Carolina, but evidence now suggests that this species is absent from the Pacific coast of the north American continent (McDonald and Koehn, 1988). In Europe it extends from the Arctic waters of the White Sea and northern Norway southwards to north Africa (Seed, 1976; Suchanek, 1985) although recent work by Sanjuan et al. (1990) suggests that mussels along the whole of the Iberian peninsula could in fact be M. galloprovincialis, and that the E Rock M.g. | | S.Africa M.g. | gy Polzeath M.g. 80 1201.5 3.96.3 12 28 44 37 53 69 ale mm aus Ibrs/wbrs (a/sh)x100 Lig.angle Fig. 5. Frequency distributions of several morphometric characters in Mytilus edulis and M. galloprovincialis. Mean values denoted by arrowheads. Note the similarity between the three M. galloprovincialis samples and how these differ from M. edulis (for abbreviations see figure 2) (after Beaumont et al. , 1989). SEED: SYSTEMATICS EVOLUTION AND DISTRIBUTION OF MYTILUS 131 M.edulis 0 ne 3 -16 -14 -12 -10 -8 -6 -4 -2 0) e.g canonical variate M.trossulus Numbers -1 012 3 4 5 6 7 8 9 10 M.trossulus M.galloprovincialis 16 18 20 22 24 26 g.t canonical variate Fig. 6. Distribution of canonical variates for pairs of species from the northern hemisphere (after McDonald et al., 1991). southern limit of M. edulis is probably further north than was suspected previously. It is present in Iceland (Varvio et al., 1988) and in Hudson Bay (Koehn, 1991), but its occurrence in Greenland, Novaya Zemlya and along the Arctic coast of Canada is still in question. In the southern hemisphere M. edulis occurs in the Falkland islands and along the east and west coasts of South America (as M. platensis and M. chilensis respectively). Mussels from the Kerguelen islands (= M. desolationis) are tentatively regarded as M. edulis (McDonald et al., 1991). Mytilus galloprovincialis also occurs in temperate waters of both hemispheres but its range extends into much warmer latitudes than M. edulis. This mussel is thought to have evolved from the M. edulis stocks which were present originally both on the Atlantic and Mediterranean coasts (Barsotti and Meluzzi, 1968). The warmer conditions which developed in the Mediterranean and the reduced contact be- tween the Mediterranean and Atlantic during one of the Pleistocene ice ages favoured the differentiation of these stocks - a process which is probably still in progress (Seed, 1978). Recent studies on mtDNA suggest a divergence time between M. edulis and M. galloprovincialis which is consistent with palaeontological evidence (Fisher and Skibinski, 1990). Northerly migration of M. galloprovincialis probably oc- curred as the ice cap retreated, and in Europe this mussel is now present along much of the Atlantic coasts of Britain, France and Ireland where it coexists and hybridises to vary- ing degrees with M. edulis (e.g. Seed, 1978; Gosling and Wilkins, 1981; Skibinski et al., 1983). M. galloprovincialis has also been introduced to areas far removed from its region of origin and in each case the intro- duced population is strikingly similar, both genetically and morphologically, to Mediterranean populations of this mussel. In the northern hemisphere its presence has been confirmed in California (McDonald and Koehn, 1988), Japan (Wilkins et al., 1983), Hong Kong (Lee and Morton, 1985) and along the east China coast northwards as far as the border between Korea and the Soviet Union (McDonald et al., 1991). These introductions were probably relatively recent events though M. galloprovincialis could have been present in California (as M. diegensis) since the turn of the century (McDonald and Koehn, 1988). In the southern hemisphere it occurs in South Africa (Grant and Cherry, 1985) and is widely distributed (as M. planulatus) throughout Australasia (McDonald et al., 1991); its absence from South America is intriguing given the long history of trading between this Continent and countries bordering the Mediterranean. Different allele frequencies between southern and northern populations of Mytilus have led to speculation that many southern mussel populations (of both M. edulis and M. galloprovincialis) could be native rather than introduced. Sup- port for this view is provided by the occurrence of Mytilus- like fossils or subfossils in Australasia (Fleming, 1959; Don- Table 3. Summary of laboratory fertilisation and larval survival experiments (Beaumont, Matin and Seed, unpub.). Treatment! Survival (%) after Abnormal 3 days? 9 days? larvae (%) pure lines: e/e;, g/g 69 70 44 ns = ns hybrids: e/e; g/e 62 37 38 ‘Each treatment consisted of 12 replicates. All replicates started with 50 or 100 x 10° eggs. 3Cultures maintained at constant larval densities (numbers.ml-') by adjusting volume. **p < 0.01; ns, not significant 132 AMER. MALAC. BULL. 9(2) (1992) A M.edulis Percent %cover J FMAM J J ASONDJIFMAMJS JAS | 108 8 9111315 17161312121110 9 9 1013161817 © Fig. 7. Reproductive cycles of A) Mytilus edulis and B) M. galloprovincialis at Rock, S. W. England. Open columns denote ripe individuals, stippled columns spent individuals C) Area occupied by reproductive follicles in histological sections of mantle tissue; asterisks indicate onset of the main spawning periods (after Seed, 1971). 0.25 oO M.galloprovincialis aan e M-edulis | 0.20 @ hybrids o ” ” =| EB 0.15 oO = 0.10 ig @ | Cc =| oO ®o 2 0.05 22.5 32.5 42.5 52.5 22.5 32.5 42.5 52.5 Shell length (mm) Fig. 8. Total annual fecundity as a function of genotype and shell length in mussels from A) Croyde and B) Whitsand, S. W. England (after Gardner and Skibinski, 1990). SEED: SYSTEMATICS EVOLUTION AND DISTRIBUTION OF MYTILUS 133 Table 4. Incidence of Pinnotheres pisum and the proportions of Mytilus edulis and M. galloprovincialis in low shore mussels at Rock!. M. edulis Date % infected (n) 1. 19522 _ 2. Nov 1966; Jan 1968 30.5 (128) 3. Jun 1968 45.3 (316) 4. May 1968-Aug 1969 30.1 (718) 5. Oct 19853 22.6 (230) 6. Oct 1989 Mean (total) 32.3 (1392) M. galloprovincialis _ Proportion (%) % infected (n) M.e. M.g. — 15 85 4.5 (112) — - 2.8 (212) 16 84 1.4 (768) — _ O (148) 14 86 — 17 83 1.8 (1240) 15.5 84.5 ‘All mussels exceeded the min. length (3.35cm) at which infection occurs. From Hepper (1957). 3Larger more heavily infected mussels less abundant than in earlier collections. ner and Jungner, 1981; Kerrison and Binns, 1984) and South America (Johnson, 1976). The possibility still remains, however, that native species of Mytilus could have interbred subsequently with, or been largely displaced by, introduced mussels of northern origin. The absence of Mytilus from aboriginal shell middens and raised-beach deposits in South Africa and from early museum collections in Japan and South Africa (Wilkins et al., 1983; Grant and Cherry, 1985) is con- sistent with the view that the present populations of M. galloprovincialis were introduced. Because M. galloprovin- cialis is widespread in the South Pacific, introductions into the northern Pacific need not, however, have originated in Europe (Koehn, 1991) though the genetic similarity between what are believed to be introduced populations and Mediter- ranean M. galloprovincialis would tend to argue against this view. Mytilus trossulus has a rather disjunct distribution oc- curring in the colder waters along both sides of the Atlantic and Pacific oceans. It is present on the west coast of North America from central California to Alaska (McDonald and Koehn, 1988), along the Pacific coast of the Soviet Union (McDonald et al. , 1991), in the Maritime Provinces of north- eastern Canada (Koehn ef al., 1984) and in the Baltic Sea (Varvio et al., 1988; Bulnheim and Gosling, 1988). Varvio et al. (1988) have suggested a relatively ancient (1-2myr) northern origin for this lineage which probably evolved from some cold tolerant genotype during the Pleistocene glacial period; this could explain why its present distribution is broadly confined to regions just south of areas that were previously ice covered. Koehn (1991) argues that M. trossulus could in fact be a zoogeographical remnant of what was once a far more widely distributed mussel. To date, M. trossulus has not been recorded in the southern hemisphere. Mytilus californianus is restricted to the Pacific coast of North America where it ranges from the Aleutian islands in Alaska to northern Mexico (Seed, 1976). M. coruscus oc- curs in Japan, and on the Pacific coast of Asia in China, Korea and Siberia (Scarlato, 1981). The geographical ranges of these two mussels, therefore, overlap with those of M. trossulus and M. galloprovincialis though M. californianus and M. cor- uscus are fairly easily differentiated from these smooth- shelled mussels on shell characteristics alone. The global distribution of Mytilus, based largely on the extensive survey by McDonald et al. (1991), is illustrated in figure 9. This survey, however, was not intended to include the small scale sampling which will clearly be required in order to establish the precise geographical and ecological ranges of the various taxa, as well as the extent of hybridisa- tion. The small scale variations in distribution are well il- lustrated by reference to one of the sites studied by McDonald et al. (1991), Posjet Bay in the Soviet Union, where mussels from an intertidal beach contained only M. trossulus whilst mussels from a floating dock just a few meters away were all M. galloprovincialis. At sites in Britain and Ireland, where M. edulis and M. galloprovincialis coexist, M. galloprovin- cialis often predominates on wave exposed shores, particularly at higher tidal elevations, whereas protected bays and estuaries are more typically favoured by M. edulis (e.g. Gosling and Wilkins, 1977, 1981; Skibinski et al., 1983; Skibinski and Roderick, 1991; Gosling and McGrath, 1990). M. galloprovicialis is known to have stronger byssal attachment than M. edulis (Gardner and Skibinski, 1990, 1991) and also possesses shell features that enhance physical stability on hard surfaces (Seed, 1978, 1990). Such attributes could explain the apparent success of this mussel in high energy environments. It is interesting to note, therefore, that M. californianus, which shares several features with M. galloprovincialis (e.g. shell shape, strong byssal attachment) also predominates on wave exposed shores (Harger, 1972; Seed and Suchanek, 1992). TAXONOMIC RELATIONSHIPS Mytilus taxonomy has relied traditionally on mor- phological shell characters but these are greatly influenced by environment, and their diagnostic value is therefore often questionable. Enzyme electrophoresis, restriction analysis of 134 AMER. MALAC. BULL. 9(2) (1992) (_)M.galloprovincialis @ M.edulis ¥ M.trossulus €) M.californianus Fig. 9. Global distribution of the three smooth-shelled mussels, Mytilus edulis, M. galloprovincialis and M. trossulus (mainly from McDonald et al., 1991). The distribution of M. californianus is also shown. mtDNA and amino-acid sequencing are relatively free of en- vironmentally induced changes and these techniques, together with immunological and cytological studies are now playing an increasingly important role in the systematic characterisa- tion of Mytilus worldwide. At present, no single character exists which separates the three smooth-shelled taxa, M. edulis, M. galloprovincialis and M. trossulus unequivocally, though certain characters and combinations of characters are clearly more diagnostic than others. Recently, McDonald er al. (1991) effectively discriminated among these taxa using a multivariate approach and in all cases analyses of allozyme and morphometric data gave concordant results. Furthermore, mtDNA sequence variation data are so far broadly consis- tent with the taxonomic judgements based on both allozyme and morphometric data (Koehn, 1991). Such studies serve to emphasise the value of a multidisciplinary approach in resolv- ing complex taxonomic problems. Whether Mytilus edulis, M. galloprovincialis and M. trossulus should be considered as separate species has been the focus of considerable discussion (e.g. Gosling, 1984; Skibinski ef al., 1983; Blot et al., 1988; Bulnheim and Gosling, 1988; Johannesson ef al. 1990; Vainola, 1990). One reason for the reluctance to consider these taxa as distinct species is that in areas of geographical overlap allozyme characters indicate varying degrees of hybridisation and in- trogression (e.g. Skibinski and Beardmore, 1979; Gosling and Wilkins, 1981; Skibinski et al. , 1983; McDonald and Koehn, 1988; Koehn, 1991; Vainola and Hvilsom, 1991) with the con- commitant mixing of morphological characters (e.g. Seed, 1972, 1974). Unfortunately, there is no generally accepted maximum amount of hybridisation which two taxa can ex- hibit and still be considered separate species. Hybrid zones of these mussels vary in size and are spatially complex with pure, mixed and hybrid populations occurring in a patchwork pattern. The most geographically widespread hybridisation occurs between M. edulis and M. galloprovincialis existing from the Biscay coast of France or even northern Spain to parts of northern Britain. Hybridisation between North Sea M. edulis and Baltic M. trossulus, by contrast, occurs over a relatively narrow zone in the Danish Belt Sea. Contact between M. edulis and M. trossulus in North America is poor- ly documented, but from the available evidence hybridisa- tion occurs at several sites in the upper reaches of the Gulf of St. Lawrence. In central California M. galloprovincialis, M. trossulus and their hybrids are present. No hybridisation has so far been reported in the region near the border between Korea and the Soviet Union where there is contact between M. trossulus and M. galloprovincialis, but this could simply reflect the lack of detailed information for this particular geograpical area. Hybrid zones between these mussels also appear to be relatively stable indicating that although gene flow does occur, the parent forms can still maintain their genetic (and morphological) integrity. In south-west England the propor- tions of Mytilus edulis and M. galloprovincialis have remained virtually unchanged over a period of almost 40 years (Table 4) despite the occurrence of hybridisation. Moreover, these proportions are virtually identical amongst all size categories of mussels (but see Gardner and Skibinski, 1990). We have no clear evidence therefore that M. galloprovincialis in this SEED: SYSTEMATICS EVOLUTION AND DISTRIBUTION OF MYTILUS 135 particular geographical locality is gradually replacing M. edulis; this is perhaps surprising in view of the higher fecundity and competitive edge that M. galloprovincialis ap- parently enjoys over its congener (e.g. Gardner and Skibin- ski, 1988, 1990, 1991; Skibinski and Roderick, 1991). If direc- tional selection in favour of M. galloprovincialis is occurring then it is obviously being offset by the immigration of M. edulis from other localities. Current genetic and morpho- metric data suggest that gene flow between M. edulis and M. galloprovincialis in the Rock population is limited. This of course partly reflects the different reproductive cycles of these two mussels at this site (Fig. 7) although in laboratory ex- periments we now know that hybrid larvae can experience heavy mortality rates (Table 3) thus presumably contributing to the temporal genetic stability in sympatric populations of these two mussels (see also Gardner and Skibinski, 1988; Gosling and McGrath, 1990). A considerable amount of selection against hybrid individuals could therefore con- ceivably occur before juvenile mussels are actually recruited to the established population. Despite the lack of any absolute reproductive barrier and the massive potential for dispersal via a planktonic lar- val stage that can last for several weeks, populations of Mytilus edulis, M. galloprovincialis and M. trossulus comprise relatively homogenous groups each maintaining a unique genetic and morphological phenotype across vast distances. This distinctiveness warrants recognition and it is perhaps appropriate therefore to consider these taxa as three distinct species despite the occurrence of localised hybridisation (McDonald et al., 1991). Differences in mtDNA fragments, sperm size and structure, as well as chromosomal variations (p. 129) further support the taxonomic interpretation that the genetic differences between these mussels are quite substan- tial and that they ought therefore to be accorded separate and equal systematic status. In an earlier paper (Seed, 1978) I have argued that M. galloprovincialis could be an emerging species, reaching specific status in certain parts of its geographical range whilst freely interbreeding elsewhere. Far from straining the biological species concept this merely em- phasises the problems inherent in extending the concept geographically. Tentative synonomies of Mytilus are sum- marised in Table 5. It is clear from the Mytilus galloprovincialis contro- versy that a multidisciplinary approach is required if the com- plex systematics of the genus Mytilus are to be satisfactorily resolved. In addition to further research using allozyme and morphometric characters, particularly applied to the hither- to poorly studied populations in the southern hemisphere, promising areas for future work include: 1) studies of reproductive cycles and fecundity in sympatric populations; 2) measurements of survival, growth and physiological parameters in natural and laboratory hybrids; 3) studies of abnormal development, growth and survival in pure and Table 5. Simplified and tentative synonymies of Mytilus spp. i) M. edulis Linnaeus (=M. platensis Orbigny M. chilensis Hupé M. desolationis Lamy = M. kerguelensis Fletcher) ii) M. galloprovincialis Lamarck (=M. diegensis Coe M. planulatus Lamarck M. aoteanus Powell M. zhirmunskii Scarlato and Starabogatov) ili) M. trossulus Gould (=M. kussakini Scarlato and Starabogatov) iv) M. californianus Conrad v) M. coruscus Gould [=M. crassitesta Lischke (=M. californianus ?)] hybrid larvae; 4) reciprocal transplants of mussels between the ranges of the different taxa; 5) comparisons of sperm mor- phology and an extension of the mtDNA, karyological and immunological studies of mussels from pure and hybrid populations. The aim of this research should not be to deter- mine once and for all whether these Mytilus taxa are ‘good species’. This is unresolvable without an agreed operational definition of a biological species and in any case is perhaps a somewhat semantic question largely devoid of biological interest. Instead, future research should concentrate on the biological processes which keep these taxa distinct despite the widespread occurrence of hybridisation. LITERATURE CITED Ahmad, M., D. O. F. Skibinski and J. A. Beardmore. 1977. An estimate of the amount of genetic variation in the common mussel Mytilus edulis. Biochemical Genetics 15:833-846. Barsotti, G. and C. Meluzzi. 1968. Osservazioni su Mytilus edulis L. e Mytilus galloprovincialis Lmk. Conchiglie 4:50-58. Beaumont, A. R., R. Seed and P. Garcia-Martinez. 1989. Electrophoretic and morphometric criteria for the identification of the mussels Myrilus edulis and M. galloprovincialis. In: Reproduction, Genetics and Distributions of Marine Organisms, J. S. Ryland and P. A. Tyler, eds. pp. 251-258. Olsen and Olsen, Fredensborg, Denmark. Bisignano, G., A. di Pietro, O. C. Grillo and I. Picerno. 1980. Serological studies of the genus Mytilus. Comparison of proteinic antigens ex- tracted from M. edulis and M. galloprovincialis. Revue International d’ Oceanographie Medicale 59:31-39. Blot, M., C. Thiriot-Quievreux and J. Soyer. 1988. Genetic relationships among populations of Mytilus desolationis from Kerguelen, M. edulis from the North Atlantic and M. galloprovincialis from the Mediter- ranean. Marine Ecology Progress Series 44:239-247. Blot, M., B. Legendre and P. Albert. 1990. Restriction fragment polymor- phism of mitochondrial DNA in subantarctic mussels. Journal of Ex- perimental Marine Biology and Ecology 141:79-87. Brock, V. 1985. Immuno-electrophoretic studies of genetic relations between populations of Mytilus edulis and M. galloprovincialis from the Medi- terranean, Baltic, East and West Atlantic and East Pacific. In: Proceed- ings of the Nineteenth European Marine Biology Symposium, P. E. Gibbs, ed. pp. 515-552. Cambridge University Press, Cambridge. 136 AMER. MALAC. BULL. 9(2) (1992) Bulnheim, H. P. and E. Gosling. 1988. Population genetic structure of mussels from the Baltic Sea. Helgolander Meeresunters 42:113-129. Coustau, C., C. Combes, C. Maillard, F. Renaud and B. Delay. 1990. Pro- sorhynchus squamatus (Trematoda) parasitosis in the Mytilus edulis Mytilus galloprovincialis complex: specificity and host-parasite rela- tionships. /n: Pathology in Marine Aquaculture, T. C. Cheng and F. O. Perkins, eds. 8 pp. Academic Press, New York and London. Crespo, C. A., T. Garcia-Caballero, A. Beiras and J. Espinosa. 1990. Evidence from sperm ultrastructure that the mussel of Galician estuaries is Mytilus galloprovincialis Lamarck. Journal of Molluscan Studies 56:127-128. Dixon, D. R. and N. Flavell. 1986. Study of the chromosomes of Mytilus edulis and Mytilus galloprovincialis. Journal of the Marine Biological Association of the United Kingdom 66:219-228. Donner, J. and H. Jungner. 1981. Radiocarbon dating of marine shells from southeastern Australia as a means of dating relative sea-level changes. Annales Academiae Scientiarum Fennicae (Geologica-Geographica) 131:1-44. Drozdov, A. L. and A. A. Reunov. 1986. Morfologiya gamet s’edobnoi midii iz Belogo, Yaponskogo morei i Avachinskoi gybi (The morphology of gametes of the blue mussel Mytilus edulis from the White and Japan Seas and Avacha inlet). Biologiya Morya 1986:52-55. Edwards, C. A. and D. O. F. Skibinski. 1987. Genetic variation of mito- chondrial DNA in mussel (Mytilus edulis and M. galloprovincialis) populations from South West England and South Wales. Marine Biology 94:547-556. Fisher, C. and D. O. F. Skibinski. 1990. Sex-biased mitochondrial DNA heteroplasmy in the marine mussel Mytilus. Proceedings of the Royal Society of London B 242:149-156. Fleming, C. A. 1959. Notes on New Zealand Recent and Tertiary mussels (Mytilidae). Transactions of the Royal Society of New Zealand 87:165-178. Gardner, J. P. A. and D. O. F. Skibinski. 1988. Historical and size-dependent genetic variation in hybrid mussel populations. Heredity 61:93-105. Gardner, J. P. A. and D. O. F. Skibinski. 1990. Genotype-dependent fecun- dity and temporal variation of spawning in hybrid mussel (Myrilus) populations. Marine Biology 105:153-163. Gardner, J. P. A. and D. O. F. Skibinski. 1991. Biological and physical fac- tors influencing genotype-dependent mortality in hybrid mussel popula- tions. Marine Ecology Progress Series 71:235-243. Gartner-Kepkay, K. E., E. Zouros, L. M. Dickie and K. R. Freeman. 1983. Genetic differentiation in the face of gene flow: a study of mussel populations from a single Nova Scotian embayment. Canadian Journal of Fisheries and Aquatic Sciences 49:443-451. Gosling, E. M. 1984. The systematic status of Mytilus galloprovincialis in Western Europe: a review. Malacologia 25:551-568. Gosling, E. M. 1992. Chapter 6. Genetics and evolution. /n: The Mussel Mytilus, E. M. Gosling, ed. (In press). Gosling, E. M. and D. McGrath. 1990. Genetic variability in exposed-shore mussels, Mytilus spp., along an environmental gradient. Journal of Experimental Marine Biology and Ecology 104:413-419. Gosling, E. M. and N. P. Wilkins. 1977. Phosphoglucoisomerase allele fre- quency data in Mytilus edulis from Irish coastal sites: its ecological significance. In: Biology of Benthic Organisms, B. F. Keegan, P. O’Ceidigh and P. J. S. Boaden, eds. pp. 297-309. Pergamon Press, London. Gosling, E. M. and N. P. Wilkins. 1981. Ecological genetics of the mussels Mytilus edulis and M. galloprovincialis on Irish coasts. Marine Ecology Progress Series 4:221-227. Grant, W. S. and M. I. Cherry. 1985. Mytilus galloprovincialis Lmk. in southern Africa. Journal of Experimental Marine Biology and Ecology 90:179-191. Harger, J. R. E. 1972. Competitive co-existence: maintenance of interacting associations of the sea mussels Mytilus edulis and Mytilus cali- fornianus. Veliger 14:387-410. Hepper, B. T. 1957. Notes on Mytilus galloprovincialis Lamarck in Great Britain. Journal of the Marine Biological Association of the United Kingdom 36:33-40. Hillman, R. E. 1990. The implications to the taxonomy of Mytilus of histopathological observations of Mytilus edulis collected from the mussel watch program. Journal of Shellfish Reseach 8: p.452 (Abstract). Hodgson, A. N. and R. T. F. Bernard. 1986. Observations of the ultra- structure of the spermatozoon of two mytilids from the south-west of England. Journal of the Marine Biological Association of the United Kingdom 66:385-390. Johannesson, K., N. Kautsky and M. Tedengren. 1990. Genotypic and phenotypic differences between Baltic and North Sea populations of Mytilus edulis evaluated through reciprocal transplantations. II. Genetic variation. Marine Ecology Progress Series 59:211-219. Johnson, L. 1976. Informe sobre una prospeccion arquelologica en Magallanes. Anales. Instituto de la Patagonia 7:87-94. Kerrison, A. R. and M. A. Binns. 1984. A midden excavation - Royal Tasmanian Botanical Gardens, Hobart. Royal Society of Tasmania. Papers and Proceedings \18:53-63. Kira, T. 1962. Shells of the western Pacific in Colour. Hoikusha Publishing Company, Osaka. Koehn, R. K., R. I. E. Newell and F Immerman. 1980. Maintenance of an aminopeptidase allele frequency cline by natural selection. Pro- ceedings of the National Academy of Sciences 77 :5385-5389. Koehn, R. K. 1991. The genetics and taxonomy of species in the genus Mytilus. Aquaculture 94:125-147. Koehn, R. K., J. G. Hall, D. J. Innes and A. J. Zera. 1984. Genetic dif- ferentiation of Mytilus edulis in eastern North America. Marine Biology 79:117-126. Lamy, E. 1936. Revision de Mytilidae vivantes du Museum National d’Histoire Naturelle de Paris. Journal de Conchyliologie 80:66-102. Lee, S. Y. and B. S. Morton. 1985. The introduction of the Mediterranean mussel Mytilus galloprovincialis into Hong Kong. Malacological Review 18:107-109. Lewis, J. R. and R. Seed. 1969. Morphological variations in Mytilus from south-west England in relation to the occurrence of M. galloprovincialis Lamarck. Cahiers de Biologie Marine 10:231-253. Lubet, P., G. Prunus, M. Masson and D. Bucaille. 1984. Recherches experimentales sur I’hybridisation de Mytilus edulis L. et M. gallo- provincialis Lmk. (Mollusques:Lamellibranches). Bulletin de la Societe Zoologique de France 109:87-98. McDonald, J. H. and R. K. Koehn. 1988. The mussels Mytilus galloprovin- cialis and M. trossulus on the Pacific coast of North America. Marine Biology 99:111-118. McDonald, J. H., R. K. Koehn, E. S. Balakirev, G. P. Manchenko, A. I. Pudovkin, S. O. Sergiyevsky, K. V. Krutovsky. 1990. On species identity of the ‘‘common mussel’’ inhabiting Asiatic coasts of the Pacific Ocean. Biologiya Morya 1990:13-22. McDonald, J. H., R. Seed and R. K. Koehn. 1991. Allozymes and mor- phometric characters of three species of Mytilus in the Northern and Southern Hemispheres. Marine Biology 111:323-333. Murdock, E. A., A. Ferguson and R. Seed. 1975. Geographical variation in leucine aminopeptidase in Mytilus edulis L. from the Irish coasts. Journal of Experimental Marine Biology and Ecology 19:33-41. Pasantes, J., M. J. Martinez-Exposito, A. Martinez-Lage and J. Mendez. 1990. Chromosomes of Galician mussels. Journal of Molluscan Studies 56:123-126. Sanjuan, A., H. Quesada, C. Zapata and G. Alvarez. 1990. On the occur- rence of Mytilus galloprovincialis Lmk. on the N.W. coast of the Iberian Peninsula. Journal of Experimental Marine Biology and Ecology 143:1-15. Scarlato, O. A. 1981. Bivalve mollusks of temperate latitudes of the western SEED: SYSTEMATICS EVOLUTION AND DISTRIBUTION OF MYTILUS 137 portion of the Pacific Ocean. Opred Faune SSSR 126:1-461. Scarlato, O. A. and Y. I. Starobogatov. 1979. The systematic position and distribution of mussels. Jn: Commercial Bivalve Molluscan Mussels and their Role in the Ecosystem, O. A. Scarlato, ed. pp. 106-111, Zoological Institute of the Soviet Academy of Sciences, Leningrad. Seed, R. 1968. Factors influencing shell shape in the mussel Mytilus edulis. Journal of the Marine Biological Association of the United Kingdom 48:561-584. Seed, R. 1969. The incidence of the peacrab Pinnotheres pisum Penn. in the two types of Mytilus (Mollusca: Bivalvia) from Padstow, south- west England. Journal of Zoology, London 158:413-420. Seed, R. 1971. A physiological and biochemical approach to the taxonomy of Mytilus edulis L. and Mytilus galloprovincialis Lmk. from south- west England. Cahiers de Biologie Marine 12:291-322. Seed, R. 1972. Morphological variations in Mytilus from the French coasts in relation to the occurrence and distribution of Mytilus galloprovin- cialis Lmk. Cahiers de Biologie Marine 13:357-384. Seed, R. 1973. Absolute and allometric growth in the mussel Mytilus edulis L. (Mollusca: Bivalvia). Proceedings of the Malacological Society of London 40:343-357. Seed, R. 1974. Morphological variations in Mytilus from the Irish coasts in relation to the occurrence and distribution of Mytilus galloprovin- cialis Lmk. Cahiers de Biologie Marine 14:1-25. Seed, R. 1976. Ecology. In: Marine Mussels: Their Ecology and Physiology, International Biological Programme 10, B. L. Bayne, ed. pp. 13-65. Cambridge University Press, Cambridge. Seed, R. 1978. The systematics and evolution of Mytilus galloprovincialis Lamarck. In: The Genetics, Ecology and Evolution of Marine Organisms, J. A. Beardmore and B. Battaglia, eds. pp. 447-468. Plenum Press, New York. Seed, R. 1980. Shell growth and form in the Bivalvia. In: Skeletal Growth of Aquatic Organisms, D. C. Rhoads and R. A. Lutz, eds. pp. 23-67. Plenum Press, New York. Seed, R. 1990. Taxonomic and evolutionary relationships within the genus Mytilus. In: The Bivalvia - Proceedings of a Memorial Symposium in Honour of Sir Charles Maurice Yonge 1986, B. Morton, ed. pp. 97-110. Hong Kong University Press, Hong Kong. Seed, R. and C. A. Richardson. 1990. Mytilus growth and its environmental responsiveness. In: The Neurobiology of Mytilus edulis. G. B. Stefano, ed. pp. 1-37. Manchester University Press, Manchester. Seed, R. and T. H. Suchanek. 1992. Chapter 4. Population and community ecology of Mytilus. In: The Mussel Mytilus, E. M. Gosling, ed. (In press). Skibinski, D. O. F. 1983. Natural selection in hybrid mussel populations. In: Protein Polymorphism: Adaptive and Taxonomic Significance. Systematics Association Special Volume 24, G. S. Oxford and D. Rollin- son, eds. pp. 283-298. Academic Press, New York and London. Skibinski, D. O. F. 1985. Mitochondrial DNA variation in Mytilus edulis L. and the Padstow mussel. Journal of Experimental Marine Biology and Ecology 92:251-258. Skibinski, D. O. F. and J. A. Beardmore. 1979. Genetic study of intergrada- tion between Mytilus edulis and Mytilus galloprovincialis. Experientia 35:1442-1445. Skibinski, D. O. F., J. A. Beardmore and T. F. Cross. 1983. Aspects of the population genetics of Mytilus (Mytilidae; Mollusca) in the British Isles. Biological Journal of the Linnean Society 19:137-183. Skibinski, D. O. F. and E. E. Roderick. 1991. Evidence of selective mor- tality in favour of the Mytilus galloprovincialis Lmk. phenotype in British mussel populations. Biological Journal of the Linnean Socie- ty 42:351-366. Soot-Ryen, T. 1955. A report on the family Mytilidae (Pelecypoda). Allan Hancock Pacific Expeditions 20:1-174. Stanley, S. M. 1972. Functional morphology and evolution of bysally attached bivalve mollusks. Journal of Paleontology 46:165-212. Suchanek, T. H. 1985. Mussels and their role in structuring rocky shore communities. In: The Ecology of Rocky Coasts, P. G. Moore and R. Seed, eds. pp. 70-96. Hodder and Stoughton, Sevenoaks. Tedengren, M., C. Andre, K. Johannesson and N. Kautsky. 1990. Genotypic and phenotypic differences between Baltic and North Sea populations of Mytilus edulis evaluated through reciprocal transplantations. III. Physiology. Marine Ecology Progress Series 59:221-227. Thiriot-Quievreux, C. 1984. Chromosome analysis of three species of Mytilus (Bivalvia: Mytilidae). Marine Biology Letters 5:265-273. Vainola, R. 1990. Allozyme differentiation between Baltic and North Sea Mytilus populations: a reassessment of evidence from transplantations. Marine Ecology Progress Series 67:305-308. Vainola, R. and M. M. Hvilsom. 1991. Genetic divergence and a hybrid zone between Baltic and North Sea Mytilus populations (Mytilidae: Mollusca). Biological Journal of the Linnean Society 43:127-148. Varvio, S. L., R. K. Koehn, R. Vainola. 1988. Evolutionary genetics of the Mytilus edulis complex in the North Atlantic region. Marine Biology 98:51-60. Verduin, A. 1979. Conchological evidence for the separate specific identity of Mytilus edulis L. and M. galloprovincialis Lmk. Basteria 43:61-80. Vermeij, G. J. 1989. Geographical restriction as a guide to the cause of ex- tinction: the case of the cold northern oceans during the Neogene. Paleobiology 15:335-356. Wilkins, N. P., K. Fujino and E. M. Gosling. 1983. The Mediterranean mussel Mytilus galloprovincialis Lmk. in Japan. Biological Journal of the Linnean Society 20:365-374. Date of manuscript acceptance: 11 December 1991 The Australasian Protocardiinae revisited (Bivalvia: Cardiidae) Jean-Maurice Poutiers Laboratoire de Biologie des Invertebres marins et Malacologie, Museum national d’ Histoire naturelle, 55, rue de Buffon, 75005 Paris, France Abstract. The Protocardiinae, the oldest unquestionable group of the family Cardiidae, has a long geological existence in Australasia. Six Recent Australasian species have been recognized and are traditionally referred to Nemocardium s.1. Two of them belong to Pratulum, a genus-group now restricted to relatively deep biotopes in warm temperate waters of Australasia. Considering its characteristic sculptural pattern and restricted Australian-New Zealand distribution since the Cretaceous, Pratulum is considered here to be a distinctive genus. The genera Microcardium and Trifaricardium are reported for the first time from Australasia. The other Australian species of Protocardiinae are classified in Lyrocardium and Frigidocardium, with a single species remaining in Nemo- cardium s.s. Since two new species of Microcardium were obtained in the upper bathyal zone by the French oceanographic ex- pedition MUSORSTOM.-1 in the Philippine Islands (Poutiers, 1981), further investigations allowed the accumulation of new data on Recent and fossil Cardiidae. This has rendered necessary the publication of a series of papers devoted to this family, and specially to members of the subfamily Proto- cardiinae. This paper is intended only as a preliminary note on the subject, in which the present author gives some results of his studies on Australasian Protocardiines, and discusses the taxonomy, biogeography and paleontology of the group. This note is based on an extensive suite of fossil and Recent material from the Indo-Pacific area, including material collected since 1977 during several oceanographic expeditions organised by the Paris Museum national d’ Histoire naturelle (Richer de Forges, 1990) as well as historic and recently col- lected material obtained from various institutions. Detailed analysis of the material used in this study, and descriptions of new taxa illustrated here will appear in subsequent publications. Since the subfamily Protocardiinae was established (Keen, 1951), there have been only a few studies on this group as a whole (Keen, 1969, 1980; Habe, 1977; Kafanov and Popov, 1977; Popov, 1977). To date, most authors have adopted the taxonomic scheme advanced by Keen. FOSSIL RECORD The protocardiines are the oldest unquestionable representatives of the Cardiidae. Their fossil record extends back to the Upper Triassic and they diversified largely dur- ing the Mesozoic. This is especially true for Protocardia Beyrich, 1845, the type-genus of the subfamily, which is known from Upper Triassic to Upper Cretaceous. Only one other genus, Septocardia Hall and Whitefield, 1877, alleged- ly belonging to the subfamily Cardiinae, occurs in the Up- per Triassic, but its placement in the Cardiidae has not re- ceived a general acceptance among taxonomists (Kafanov and Popov, 1977). Two taxa, Nemocardiunt Meek, 1876, and Pratulum Iredale, 1924, bridged the gap between Mesozoic and Cenozoic (Keen, 1950, 1980; Skwarko, 1983), and they still occur in the Recent fauna. They are the only members of the family to show this stratigraphic distribution. After a second period of radiative evolution during the Tertiary, Protocardiines nowadays survive mainly on the outer shelf and upper continental slope. This is a rather deep-water habitat for the family that occurs typically in the shallow waters of coastal to subtidal environments. Thus, in a way, Recent Protocardiines can be considered as a relict deep-water group within the family Cardiidae. BIOGEOGRAPHY The subfamily Protocardinae is present in nearly every tropical region, and the western Pacific Ocean is the area of maximum species diversity with 15 of the 26 known species worldwide (Fig. 1). The genus Microcardium Thiele, 1934, is a good example. No less than three species are known from Philippine waters, two of which have been described recent- ly (Poutiers, 1981). The most common of these, Microcardium aequiliratum Poutiers, 1981 (Fig. 2a), lives in fine soft bot- toms at depths of 180m to 350m. In addition to the Western Pacific Ocean, Protocardiines occur also in the Indian Ocean (six species), Eastern Pacific (five species), as well as in the Western and Eastern Atlantic Ocean (three and one species respectively). In Australasia, five Recent species of Protocardiines are known from Australia (two of which occur also in New American Malacological Bulletin, Vol. 9(2) (1992):139-144 139 140 AMER. MALAC. BULL. 9(2) (1992) Philippine Sea e Scueh. a 2 gegen Usea BORNEO ae ee cergses et YD MJAVA L SFrivon > Arafura Sea Timor Sea | Tropic of Capricorn 110°E A 120° 4 Va oP a7 NEW GUINEA = SSP PACIFIC OCEAN sO ° y 8s 4 oO 10 N [@) Equator oO Coral Sea Fig. 1. Simplified map of Australasia and neighbouring areas. Caledonia), and one from New Zealand. They are traditional- ly allocated to the broadly defined genus Nemocardium, as are all Recent species of the Protocardiinae (Keen, 1969, 1980). This placement seems to be supported by the rather conservative aspect of conchological characters in the sub- family, and by a frequent occurrence of secondary con- vergence that makes the differentiation of lineages sometimes problematical. According to the climatic zonation of the world oceans, two main biogeographical areas are recognized currently in Australasia. In these areas, a number of faunal regions and provinces have been defined for the intertidal and shelf zones, on the basis of distribution and the degree of endemism of the benthic and demersal organisms (Wilson, 1971; Briggs, 1974: Pielou, 1979). Thus, northern Australia, New Caledonia and its dependencies (Fig. 1) are included in the tropical Indo- West Pacific Region, whereas southern Australia and northern New Zealand correspond to a warm temperate area forming the Southern Australian and the Northern New Zealand Regions. This biogeographical pattern is reflected in the distribu- tion of Protocardiines in Australian waters. Nemocardium 1s present in the tropical region, represented by N. bechei (Reeve, 1840), the only living species of Nemocardium s.s. (Fig. 2b). Pratulum occurs in the temperate region, repre- sented by P. thetidis (Hedley, 1902), type-species of the genus (Fig. 2c). P. thetidis was thought by Hedley (1902) to be only a local variety of Cardium striatulum Sowerby, 1834, an older but incorrect name for the New Zealand species P. pulchellum (Gray, 1843) (Fig. 2d). However, Iredale (1924) correctly dis- tinguished these species when he erected Pratulum. Pratulum appears to be confined nowadays to the temperate waters of Australasia. The only record of Pratulum from tropical Australasia is that of Fischer-Piette (1977). This seems to be an error based on the incorrect labelling of an old specimen, and I have been unable to find any member of this genus in the large samples collected recently in that region by the Paris Museum national d’Histoire naturelle. However, paleontologists often consider Pratulum a Tethyan Indo-Pacific element, like Nemocardium s.s. For example, Darragh (1985), using a similar approach to Fleming (1962, 1967), analysed the composition of the molluscan fauna to define a tentative biozonation of the Tertiary in southeastern Australia and included Pratulum under the Tethyan Indo- POUTIERS: AUSTRALASIAN PROTOCARDIINAE 141 Sethe hen ee Fig. 2. External views of Recent and Tertiary species of Protocardiinae, showing diversity of sculpturing among genera (AMS, Australia Museum, Sydney; MNHN, Museum national d’Histoire naturelle, Paris; MV, Museum of Victoria, Melbourne). a) Microcardium aequiliratum, length 26.5 mm, Philippines, MUSORSTOM-2 Stn 17 (MNHN); b) Nemocardium bechei, length 40.0 mm, New Caledonia Stn 542 (MNHN); ¢) Pratulum thetidis, length 16.0 mm, southern Australia (AMS C145776); d) P. pulchellum, length 22.0 mm, New Zealand (MNHN); e) N. antisemigranulatum, length 39.2 mm, middle Miocene of Australia (MV P128253-57); f) N. edwardsi, length 50.1 mm, Paleocene of Paris Basin (MNHN); g) Frigidocardium sp., length 10.9 mm, middle Miocene of Australia (MV P123418); h) P. hemimeris, length 8.6 mm, middle Miocene of Australia (MV P123419); i) P. ornithopetronicum, length 11.4 mm, upper Oligocene of Australia (MV P123422-36); j) P. proterothetidis, length 12.4 mm, Pliocene of Australia (MV P30788); k) Microcardium sp. A, length 11.1 mm, north- western Australia (AMS C145746); 1) Microcardium sp. B., length 11.3 mm, New Caledonia, MUSORSTOM-6 Stn DW428 (MNHN); m) Trifaricardium sp., length 11.6 mm, northeastern Australia (AMS C145814). 142 AMER. MALAC. BULL. 9(2) (1992) Pacific heading, but did not record Nemocardium from that region. Therefore, a re-evaluation of the generic status of Australian species seemed necessary to decide if there is any correlation between the past and the present biogeography of Protocardiines. TAXONOMY Nemocardium s.s. is well known from Cenozoic strata. It is a typical Tethyan Indo-Pacific element, with maximum species diversity in the Paleogene of Europe, where about 30 nominal species are recorded (Keen, 1950; Tremlett, 1950; Popov, 1977). Nemocardium exhibits a characteristic, strongly discrepant, outer sculpture with well marked radial ribs on the posterior area; the remainder of the surface is quite smooth and reveals only numerous, fine and low subsurface radial riblets. Internally, the marginal crenulations appear much stronger on the posterior area, in accordance with the outer radial sculpture. On the posterior margin, crenulations correspond with radial ribs but, on the ventral and anterior margins, they correspond with interstices. In the Cenozoic of New Zealand, Nemocardium is represented by Varicardium Marwick, 1944, known from the Upper Eocene to the Middle Miocene (Beu and Marwick, 1990). Species of Varicardium have been first referred to the Mesozoic genus Protocardia (Suter, 1914; Finlay, 1924) in view of the concentric folds of their anterior slope. However, as already pointed out by Marwick (1944) himself, it appears to be a case of secondary convergence and Varicardium more probably represents an offshoot of the Nemocardium line in New Zealand (Boreham, 1965; Keen, 1980). In Australia, Nemocardium is known from the Miocene onwards, represented by the species N. antisemigranulatum (McCoy, 1877) (Fig 2e). McCoy (op. cit.) described it as a species of “‘Cardium (Protocardium)’’, an unjustifiable emen- dation for Protocardia. However, this latter taxon is confined to the Mesozoic, and Stewart (1930) has given good reasons for separating it from Nemocardium at the generic level. Since then, this species has often been classed in Pratulum (Darragh, 1970). However, it shows strong affinities with the Recent species N. bechei, and with typical Nemocardium species from the Paleogene of Europe (Fig. 2f), and fits well in that genus. This kind of confusion in the generic place- ment of species is rather common in the Protocardiinae. The first confusion between Nemocardium and Pratulum was due to Iredale (1927) himself, who differentiated the Australian specimens of N. bechei under the name ‘‘Pratulum pro- batum’’, i.e. only three years after he created the genus Pratulum, the true affinity of P- probatum has been shown by Wilson and Stevenson (1977). In addition, Australian Cenozoic Protocardiines are almost invariably referred to Pratulum, without any consideration to their true generic status. For example, it was found that collections of the Museum of Victoria contain a Miocene species of the genus Frigidocardium Habe, 1951 (Fig. 2g) which had been con- fused with P. hemimeris (Tate, 1887). This seems to be the first fossil record of Frigidocardium in Australia. Frigido- cardium is characterized by an homogeneous outer sculpture, whereas both Pratulum and Nemocardium exhibit two clear- ly differentiated areas. Recent species of Pratulum have a distinctive sculptural pattern. As noted above, the outer surface has two clearly differentiated areas as in Nemocardium s.s. but, in Pratulum, radial ribbing is not confined to the posterior area. Moreover, the anterior and median areas of Pratulum spp. have a fine secondary sculpture of irregularly concentric, anastomosing threads crossing the radial ribs. The marginal crenulations are not clearly stronger posteriorly as they are in Nemo- cardium but, as in that genus, the crenulations do correspond with the radial sculpture. The study of Australian fossils reveals that, despite the above mentioned errors in generic placement, Pratulum actually is present in Australia during the whole Cenozoic, and probably also in Upper Cretaceous (Skwarko, 1983). The known Australian Cenozoic species referrable to Pratulum are: Cardium hemimeris Tate, 1887 (Fig. 2h), Protocardia ornithopetronica Chapman and Crespin, 1928 (Fig. 21), N. (Pratulum) proterothetidis Lud- brook, 1955 (Fig. 2j). In New Zealand, Pratulum also has a long history (Finlay and Marwick, 1937; Marwick, 1944) and there is evidence that it has been present from the Early Paleocene onwards (Hornibrook and Harrington, 1957; Beu and Marwick, 1990). However, none of the Cenozoic Euro- pean species referred formerly to Pratulum actually belong to this genus (Gilbert and Van de Poél, 1970). Thus, Pratulum appears to be a distinctive genus with a mainly Australian-New Zealand distribution, and the claim that it represents a Tethyan Indo-Pacific element seems un- founded. Recent Pratulum comprises three species, one in New Zealand and two in Australia (one undescribed). In spite of morphological similarities between Partulum and Microcardium Thiele, 1934, the latter seems to be a distinct lineage of tropical affinities, with a completely dif- ferent history. Its primary diversification center could be Central America. Morphologically, Microcardium can be distinguished from Pratulum by the occurrence of concen- tric scales in the interstices of radial ribs of the posterior area. Microcardium is here reported for the first time in the extant Australasian fauna, with one new species in Australia (Fig. 2k), and one in New Caledonia (Fig. 21). There do not seem to be any fossil records of this genus in Australia. Trifari- cardium Kuroda and Habe, 1951, is also recorded for the first time in that region, with a new Recent species occurring in Australia and the Coral Sea (Fig. 2m). This genus closely resembles Frigidocardium, but can be distinguished by its strong, beaded concentric threads on the anterior slope. The other Australian species of the subfamily Proto- POUTIERS: AUSTRALASIAN PROTOCARDIINAE 143 cardiinae can be referred to Frigidocardium (two species) and Lyrocardium (one species), with a single species (Nemo- cardium bechei) remaining in Nemocardium s.s. In their com- prehensive account of Western Australian Cardiidae, Wilson and Stevenson (1977) dealt with these taxa, but described the species of Frigidocardium under the heading of ‘‘Nemo- cardium (Microcardium)”’: DISCUSSION If the subfamily Protocardiinae represents a mainly deep-water, relict group in the Recent fauna, it appears to be well diversified in Australasia. The bathymetrical distribu- tion of its member taxa reveals once more the conservative character of the upper bathyal zone (Lozouet, 1990). Moreover, the recent exploration of the bathyal area of South West Pacific around New Caledonia reveals that a number of taxa descend directly from the Mesozoic fauna (Vacelet, 1977; Ameziane-Cominardi et al., 1987; Bouchet, 1987; Bourseau et al., 1987) in this relatively stable, refuge area, which represents a survival of the eastern rim of Gond- wanaland. In itself, this new information justifies the efforts of faunistic research made in this region. ACKNOWLEDGMENTS I wish to thank Philippe Bouchet (Museum national d’Histoire naturelle, Paris), Thomas Darragh (Museum of Victoria, Melbourne), George Kendrick (Western Australian Museum, Perth), Neville Pledge (South Australian Museum, Adelaide), Winston Ponder (Australian Museum, Sydney) and Shirley Slack-Smith (Western Australian Museum, Perth) for communication of material on which this study is based. I am also grateful to Thomas Darragh for valuable assistance in Australian biostratigraphy and relevant literature. LITERATURE CITED Ameziane-Cominardi, N., J. P. Bourseau and M. Roux. 1987. Les crinoides pédoncules de Nouvelle-Calédonie (sw-Pacifique): Une faune bathyale ancestrale issue de la Mesogee mesozoique. Compte-rendus heb- domadaires des Seances de l'Académie des Sciences, Paris (II) 304:15-18. Beu, A. G. and P. A. Maxwell. 1990. Cenozoic Mollusca of New Zealand. New Zealand Geological Survey Paleontological Bulletin (58):1-518. Boreham, A. U. E. 1965. A Revision of F. W. Hutton’s Pelecypod Species Described in the Catalogue of Tertiary Mollusca and Echinodermata (1873). New Zealand Geological Survey Paleontological Bulletin (37):1-84. Bouchet, P. 1987. L’exploration de la faune profonde de Nouvelle-Caledonie ou ala découverte des mondes perdus. Lettre d’Information Greco Ecoprophyce (4):84-87. Bourseau, J. P., N. Ameziane-Cominardi and M. Roux. 1987. Un crinoide pedonculé nouveau (Echinodermes) représentant actuel de la famille Jurassique des Hemicrinidae: Gymnocrinus richeri nov. sp. des fonds bathyaux de Nouvelle-Caledonie (sw. Pacifique). Compte-rendus heb- domadaires des Seances de l’Académie des Sciences, Paris (II) 305:595-599. Briggs, J. C. 1974. Marine Zoogeography. McGraw-Hill Book Company, New York. 475 pp. Darragh, T. A. 1970. Catalogue of the Australian Tertiary Mollusca (except Chitons). Memoirs of the National Museum of Victoria (31):125-212. Darragh, T. A. 1985. Molluscan biogeography and biostratigraphy of the Tertiary of southeastern Australia. Alcheringa 9:83-116. Finlay, H. J. 1924. The molluscan fauna of Target Gully: Part 1. Transactions of the New Zealand Institute 55:495-516. Finlay, H. J. and J. Marwick. 1937. The Wangaloan and Associated Molluscan Faunas of Kaitangata-Green Island Subdivision. New Zealand Geological Survey Paleontological Bulletin (15):1-140. Fischer-Piette, E. 1977. Revision des Cardiidae (Mollusques Lamelli- branches). Memoires du Muséum national d'Histoire naturelle Nouvelle Série (A) 101:1-212. Fleming, C. A. 1962. New Zealand Biogeography. A paleontologist’s ap- proach. Tuatara 10 (2):53-108. Fleming, C. A. 1967. Cenozoic History of Indo-Pacific and other Warm- Water Elements in the Marine Mollusca of New Zealand. Venus 25(3-4):105-117. Glibert, M. and L. Van de Poél. 1970. Les Bivalvia fossiles du Cenozoique etranger des collections de I’Institut Royal des Sciences Naturelles de Belgique. VI (fin): Oligodontina (2) Astartedontina et Septibranchida. Memoires de l'Institut royal des Sciences naturelles de Belgique (2) 84:1-185. Habe, T. 1977. Systematics of Mollusca in Japan. Bivalvia and Scaphopoda. Zukan-no-Hokuryuka, Tokyo. 372 pp. (In Japanese). Hedley, C. 1902. Mollusca. Part I. Brachiopoda and Pelecypoda. In: Scien- tific results of the trawling expedition of H.M.C.S. ‘‘Thetis’’, off the coast of New South Wales, in February and March, 1898. Part 5. Memoirs of the Australian Museum (4):285-324. Hornibrook, N. de B. and J. H. Harrington. 1957. The status of the Wangaloan Stage. New Zealand Journal of Science and Technology (B) 38:655-670. Iredale, T. 1924. Results from Roy Bell’s molluscan collections. Proceedings of the Linnean Society of New South Wales 49:179-278. Iredale, T. 1927. New molluscs from Vanikoro. Records of the Australian Museum 16:73-78. Kafanov, A. I. and S. V. Popov. 1977. K sisteme kaynozoyskikh kardioidey (Bivalvia). [On the system of cenozoic Cardioidea (Bivalvia)]. Paleontologicheskij Zhurnal 1977 (3):55-64. Keen, A. M. 1950. Notes on'the history of Nemocardium (family Cardiidae). Journal de Conchyliologie 90 (1):23-29. Keen, A. M. 1951. Outline of a proposed classification of the pelecypod family Cardiidae. Conchological Club of Southern California Minutes 111:6-8. Keen, A. M. 1969. Family Cardiidae Lamarck, 1809. In: Treatise on In- vertebrate Paleontology, Part N, Volume 2, Mollusca 6, Bivalvia, R. C. Moore, ed. pp. N583-N590. The University of Kansas, Lawrence, and the Geological Society of America, Boulder. Keen, A. M. 1980. The Pelecypod Family Cardiidae: a taxonomic summary. Tulane Studies in Geology and Paleontology 16 (1):1-40. Lozouet, P. 1990. Mollusca Gastropoda: Eumitra recentes de la région néo- calédonienne et Charitodron fossiles de I’Oligocene superieur d’Aqui- taine (Mitridae). In: Résultats des Campagnes MUSORSTOM, Volume 7, A. Crosnier and P. Bouchet, eds. Mémoires du Museum national d Histoire naturelle (A) 150:205-222. Marwick, J. 1944. New Zealand Fossil and Recent Cardiidae (Mollusca). Transactions of the Royal Society of New Zealand 74(3):255-272. McCoy, F. 1877. Prodromus of the Palaeontology of Victoria. Decade 5S. Geological Survey of Victoria, Melbourne. 40 pp. Pielou, E. C. 1979. Biogeography. John Wiley & Sons, New York. 351 pp. Popov, S. V. 1977. Mikrostruktura rakoviny i sistematika kardiid. [The shell structure and system of the Cardiids]. Trudy paleontologicheskogo Instituta Akademija Nauk SSSR 153:1-124. Poutiers, J. M. 1981. Mollusques: Bivalves. In: Résultats des Campagnes MUSORSTOM.1 - Philippines (18-28 Mars 1976). Mémoires ORSTOM 144 AMER. MALAC. BULL. 9(2) (1992) (91):325-356. Richer de Forges, B. 1990. Les campagnes d’exploration de la faune bathyale dans la zone économique de la Nouvelle-Calédonie. In: Résultats des campagnes MUSORSTOM, Volume 6, A. Crosnier, ed. Mémoires du Museum national d'Histoire naturelle (A) 145:9-54. Skwarko, S. K. 1983. Cenomanian (Late Cretaceous) Mollusca from Mount- norris Bay, Arnhem Land, northern Australia. Bureau of Mineral Resources, Australia, Bulletin 217:73-83. Stewart, R. B. 1930. Gabb’s California Cretaceous and Tertiary Type Lamellibranchs. Academy of Natural Sciences of Philadelphia Special Publication (3):1-314. Suter, H. .1914. Revision of the Tertiary Mollusca of New Zealand, based on type material. Part I. New Zealand Geological Survey Paleon- tological Bulletin (2):1-64. Tremlett, W. E. 1950. English Eocene and Oligocene Cardiidae. Proceedings of the Malacological Society of London 28 (4-5):115-133. Vacelet, J. 1977. Une nouvelle relique du Secondaire: un representant ac- tuel des Eponges fossiles Sphinctozoaires. Compte-rendus heb- domadaires des Séances de l'Academie des Sciences, Paris (II) 285:509-S11. Wilson, B. R. and K. Gillett. 1971. Australian Shells. Reed, Sydney, Melbourne. 168 pp. Wilson, B. R. and S. E. Stevenson. 1977. Cardiidae (Mollusca, Bivalvia) of Western Australia. Western Australian Museum Special Publica- tion (9):1-114. Date of manuscript acceptance: 31 October 1991 Preliminary cladistic analysis of the bivalve family Cardiidae Jay A. Schneider Department of Geophysical Sciences, University of Chicago, Chicago, Illinois 60637, U. S. A. Abstract. Phylogenetic relationships within the bivalve family Cardiidae have been examined by cladistic analysis. Thirty-six of the approximately 180 cardioid supraspecific taxa are analyzed, including members of each of the generally recognized cardiid subfamilies, plus the cardioid families Lahilliidae, Lymnocardiidae, and Tridacnidae. Data for each taxon have been taken from a single species. For the outgroup, a hypothetical ancestor has been constructed from data for the carditids Cyclocardia ventricosa (Gould) and Cardita variegata Bruguiére. The data consist of 54 characters and 170 character states. Results indicate that the families Lahilliidae, Lymnocardiidae, and Tridacnidae should be given subfamilial status within the Cardiidae. Septocardia is removed from the Cardiinae and placed in its own subfamily, and the subfamily Protocardiinae is found to be paraphyletic. The Laevicardiinae, as proposed by Keen (1936, 1951, 1969, 1980), is shown to be polyphyletic: Cerastoderma is a lymnocardiine; Dinocardium is a cardiine; Clinocardium is the type genus of the Clino- cardiinae. In addition to Laevicardium, only Habecardium and Fulvia remain in Laevicardiinae. The Trachycardiinae is found to be a monophyletic taxon within the Cardiinae. The subfamilies Clinocardiinae, Tridacninae, Lymnocardiinae, and Fraginae form a monophyletic clade. Sawkinsia, long considered a tridacnid, belongs with the Cardiinae. Bivalves of the family Cardiidae, or cockles, display acuticostatum in Cardium (Bucardium), which Keen (1980) a wide spectrum of shell shapes, ribbing and ornamentation indicates is known from only the Miocene to Recent. patterns, hinge morphologies, and numerous other conch- Boss (1971), Kafanov and Popov (1977), Keen (1980) ological features. Their complex morphology, accompanied and Ponder et al. (1981) placed the enigmatic Hemidonax in by their good fossil record, allows the cardiids to be evaluated the Cardioidea. However, Scarlato and Starobogatov (1979) evolutionarily, ecologically, functionally, and phylogenetically argued that Hemidonax is aligned with the Donacidae. After in considerable detail. examination and comparison of the shell and of the anatomy The higher-level taxonomy of the Cardiidae has been (both external and internal) of Hemidonax to both cardiids more thoroughly studied than that of most other groups of and donacids, I cannot justify placing Hemidonax as a bivalves (Dall, 1901; Stewart, 1930; Keen, 1936, 1937, 1951, member of the Cardioidea. However, neither can I place 1969a, 1980; Fischer-Piette, 1977; Kafanov and Popov, 1977; Hemidonax within the Donacidae. Instead, I favor placing Popov, 1977; Wilson and Stevenson, 1977; Voskuil and Onver- Hemidonax as incertae cedis within the order Veneroida, until wagt, 1989). Kafanov and Popov (1977) made the only detailed a phylogenetic analysis of the Veneroida is undertaken. attempt to reconstruct the phylogenetic history of the group. Virtually all cardiid taxonomy is based on hard parts, Traditionally, the superfamily Cardioidea comprises: with the exception of Starobogatov’s (In: Kafanov and Popov, (1) the Cardiidae; (2) the extinct, southern hemisphere 1977) study of stomach structure. Most of what is known Lahilliidae; (3) the brackish-water Lymnocardiidae, confined about the anatomy of cardiids comes from the study of the to eastern Europe and southwestern Asia; (4) the Tridacnidae, common cockle, Cerastoderma edule (Linnaeus) (see or giant clams [Keen (1969b), Kafanov and Popov (1977), and Menégaux, 1890; Johnstone, 1899; Zugmayer, 1904; Kiipfer, Scarlato and Starobogatov (1979) have placed the giant clams 1915; Roche, 1925; Atkins, 1937; Graham, 1949), which is in a separate superfamily]. Kafanov and Popov (1977) con- taken as a model for the entire family. Furthermore, Russian tended that the Lahilliidae belonged to the Arcticoidea. malacologists (Kafanov and Popov, 1977; Popov, 1977; However, as noted by Finlay and Marwick (1937) and Mar- Taktakishvili, 1987) place Cerastoderma within the subfam- wick (1944), the hinge of Lahillia is of the cardiid, not ily Lymnocardiinae - which, less Cerastoderma, Keen (1969a, arcticoid type; the lack of external ornament is apparently 1980) considered to be a separate family within the super- a case of convergence. This classification has not been taken family Cardioidea. The only treatment of comparative for granted, and representatives of the Tridacnidae, Lahillii- anatomy of the cardiids is that of Pelseneer (1911). dae, and Lymnocardiidae are included in this analysis. The goals of this study are to: (1) determine the status Additionally, Cardium acuticostatum d’Orbigny, 1842, and content of subfamilies erected by previous workers; (2) is included. Wilckens (1904) placed the Cretaceous C. propose a preliminary phylogenetic hypothesis for the family. American Malacological Bulletin, Vol. 9(2) (1992):145-155 145 146 AMER. MALAC. BULL. 9(2) (1992) Characters and character states are briefly described herein and will be treated fully in future publications. MATERIALS AND METHODS There are about 180 generally accepted cardioid genera and subgenera (see Keen, 1969a, 1980; Kafanov and Popov, 1977; Vokes, 1980; Taktakishvili, 1987). It is not feasible cur- rently to run a computer-driven cladistic program for such a large number of taxa. The 36 taxa chosen in this study in- clude at least one representative of each of the cardiid sub- families accepted by Keen (1969a, 1980), Kafanov and Popov (1977), and Voskuil and Onverwagt (1989). As stated above, the tridacnids (Tridacna), \ahilliids (Lahillia), and lymno- cardiids (Hypanis) are also represented in this analysis (the suffix -ids is used in a vernacular sense until their taxonomic placement is discussed thoroughly). Kafanov and Popov’s (1977) and Keen’s (1980) classification scheme for the taxa analyzed in the present study is given in appendices | and 2. A cladistic analysis of the 36 taxa with 54 characters comprising 170 character states (appendices 3 and 4) was made using PAUP 3.0d (Swofford, 1989). The accelerated transformation option (ACCTRAN) was used, and steps were not added to terminal taxa with polymorphisms. Synapomor- phies for each node are presented in appendix 5. Character states were encoded from a single species of each genus or subgenus. Most of the taxa are represented by their type species. Exceptions are those taxa for which outgroup Palaeocardita Septocardia Protocardia ISEPTOCARDIINAE IPROTOCARDIINAE Integricardium A Lahillia JLAHILLIINAE Nemocardium Ss Habecardium 7 Fulvia ©N Laevicardium Granocardium 2 Trachycardium Vasticardium &~» Acrosterigma Hedecardium Agnocardia Orthocardium os Cardacuti S S» Austrocardium JCARDIINAE nN = , Vepricardium RS Cardium 2 3 Xe Dinocardium Bucardium Chrysocardium Sawkinsia Acanthocardia Rudicardium Plagiocardium o Parvicardium a , Apiocardia S, <7 Trigoniocardia vw Fragum . \ Tridacna —_ ITRIDACNINAE ec, Cerastoderma Hyoanie | [LYMNOCARDIINAE Clinocardium [CLINOCARDIINAE LAEVICARDIINAE FRAGINAE Fig. 1. Majority-rule consensus tree of 50 most parsimonious trees. Nodes 21 and 22 supported by 60% of trees; nodes 28 and 30 supported by 70% of trees; node 31 supported by 90% of trees; all other nodes supported by 100% of trees. Synapomorphies supporting each node given in appendix 5. Cardium Bucardium Vepricardium y Hedecardium Cardiinae 4 °cc-errreooo eee Trachycardiini IEEE se | aevicardiini n-- -_—_eoereree— Dinocardium —_—_—_—_:.:. nnn ee Clinocardiinae = Nemocardium Protocardiinae Habecardium — ae Fragum 7 4 y Trigoniocardia C= A § heanthocaria Fraginae \ Orthocardium Plagiocardium \ ——— Parvicardium Fig. 2. Evolutionary scenario for the Cardiidae from Kafanov and Popov (1977). Only those taxa included in both the present study and Kafanov and Popov (1977) are shown. (1) material of the type species was unavailable, or (2) anatomical material was not available for the type (e.g. Nemocardium, for which the type species is extinct), but was available for another species generally assigned to that tax- on. Taxa represented by species other than the type are listed in appendix 6. Therefore, this analysis should be taken as a phylogeny for these species only. Character states presented may not be constant throughout all species of a given genus. Because there is considerable disagreement over what con- stitutes a genus or a subgenus in the Cardiidae, all terminal taxa will be considered to have equal rank as genera; no distinction will be made between genera and subgenera, ex- cept as noted in the text. Citations in the character list (appendix 3) refer to previous discussions of that character. Except for the infor- mation on Cyclocardia (see below), character 12, and character 8 for Parvicardium (from Pelseneer, 1911), all character states were encoded from examination of specimens. The Cardiidae are generally accepted as having been derived from a member of the Carditoidea [Cox, 1949; Keen, 1969a, 1980; Newton, 1986; but see Morris (1978) and Morris et al. (1991)]. These authors have postulated an evolutionary scenario of Palaeocardita originating from some primitive carditid or permophorid, with Septocardia then originating from Palaeocardita. Palaeocardita is usually placed with the Carditidae (Chavan, 1969). However, examination of the one species of Palaeocardita available to me, Palaeocardita silberlingi Newton et al., has led me to place this species within the ingroup Cardiidae, on the basis of its cardinal teeth, which are arranged as in Septocardia and Protocardia [see Newton et al. (1987) for a discussion of this species]. Therefore, to represent the outgroup, a hypothetical ancestor was constructed with information from the Recent carditids Cardita variegata Bruguiére and Cyclocardia ventricosa SCHNEIDER: CLADISTIC ANALYSIS OF THE CARDITDAE 147 (Gould). Character states for C. ventricosa were taken from information in Yonge (1969). For characters 6 and 30, the two carditids provided conflicting information, and hence the states are scored as missing (‘‘?’’). Most characters are unordered. It was possible to con- struct character state trees based on ontogeny for characters 23 (shell shape), 24 (ribbing pattern), 29 (mosaicostracum), and 40 (rib flares). RESULTS AND DISCUSSION Fifty most parsimonious trees of 208 steps (consistency index = 0.566) were found. The 50% majority-rule consensus tree is presented in figure 1, which can be compared with two previously presented evolutionary scenarios. Kafanov and Popov (1977) produced a phylogram based on two key characters stomach structure (analyzed by Ya. I. Starobogatov) and Popov’s (1977) work on shell micro- structure. Kafanov and Popov (1977) considered 38 taxa [only those taxa represented in both my analysis and that of Kafanov and Popov (1977) are shown in figure 2; they considered neither the origin of the Cardiidae nor the group’s Mesozoic history]. Only nine of the 28 extant taxa were examined for stomach structure. Starobogatov’s study of stomach structure rests heavily on the presence/absence and position of sorting areas, as described by Purchon (1960a). Purchon’s (1960a) description of the cardiid stomach came from the study of Cerastoderma edule by Graham (1949). Starobogatov (In: Kafanov and Popov, 1977) stated that the SA-3, or posterior sorting area, is absent in Cerastoderma based on examina- tion of C. glaucum (Bruguiére) (Ya. I. Starobogatov, pers. comm.) and Hypanis. However, the posterior sorting area (labeled SAP) is the most prominent structure in Graham’s ( 1949) figure of the stomach of C. edule. While promoting the utility of using stomach structure to elucidate the higher- level phylogeny of the Bivalvia (Purchon, 1959, 1960a, b), Purchon (1960a:481) warns that ‘*...it is not easy to make an objective analysis of the occurrence and identities of the various sorting areas. The presence or absence, and the degree of development of the various sorting areas has a profound effect on the appearance of the interior of the stomach, and could obscure more fundamental issues such as the course taken by the major typhlosole and the intestinal groove. ..the occurrence of sorting areas can only be used with the greatest caution for phylogenetic purposes.’’ Although seven of the 20 anatomical characters in the present analysis concern the stomach, none relate to the sorting areas. Popov’s (1977, 1986) classification of bivalve shell microstructure conflicts with those of Carter (1980, 1989), Carter and Clark (1985), Carter and Lutz (1989) and Watabe (1984). The only systemically useful microstructural char- acters that I have found so far concern the relationship of the ornament to the rest of the shell (characters 28, 29 and 40). The only cardiid phylogeny suggested by Keen is found in her description of cardiid evolution (Keen, 1980). I have constructed a phylogram (Fig. 3) based on that description. In the present analysis, Palaeocardita silberlingi is located at the base of the cladogram, followed by Septocardia, Protocardia, and then the rest of the Cardiidae. This topology is in agreement with the ideas of early cardiid evolution sug- gested by Cox (1949), Keen (1969a, 1980) and Newton (1986). However, the monophyly of Palaeocardita is questionable (C. R. Newton, pers. comm.), and the more common species, including the type P austriaca (Hauer) and P. crenata (Goldfuss) could be carditoids, whilst P. silberlingi is a cardiid. Due to the uncertainty of the taxonomy of Palaeo- cardita, I refrain from placing P. silberlingi in a higher tax- on within the Cardiidae. Septocardia was placed in its own family, the Septo- cardiidae, in the superfamily Tridacnoidea by Kafanov and Starobogatov (In: Kafanov and Popov, 1977). Septocardia is clearly a primitive cardiid and it does not share any of the derived features of Tridacna. Likewise, Septocardia does not belong in the derived subfamily Cardiinae as indicated by Keen (1969a, 1980). Septocardia is here placed in the cardiid subfamily Septocardiinae. The subfamily Protocardiinae has been understood to include the genera Protocardia, Integricardium, Jurassi- cardium, and Nemocardium (Kees, 1969a, 1980). My results indicate that this is a paraphyletic group. /ntegricardium is more closely related to Lahillia. The paraphyly of the Pro- tocardiinae has been acknowledged implicitly for some time. McLearn (1933) erected Onestia as a subgenus of /ntegri- cardium, the former was considered a genus by McLearn Carditoidea Palaeocardita Septocardia Protocardiinae fe ae ae Laevicardiinae Incacardium Cardium Bucardium Orthocardium Trachycardiinae Fraginae Acanthocardia Granocardium Fig. 3. Evolutionary scenario reconstructed from Keen (1980:24-30). 148 AMER. MALAC. BULL. 9(2) (1992) (1945) and Day (1978), but not by Keen (1969a, 1980). Day (1978) postulated that /ntegricardium is ancestral to Onestia, which is in turn ancestral to Lahillia. He also placed Onestia in the Lahilliidae. Present results indicate that the family Lahilliidae should be relegated to a subfamily (Lahilliinae) within the family Cardiidae (as originally proposed by Finlay and Marwick, 1937), and should include /ntegricardium. Nemocardium is the sister taxon to the Laevicardiinae. This group is in turn the sister taxon to the rest of the cardiids. Because of its change in ribbing pattern from that similar to Nemocardium as a juvenile, to that of Fulvia as an adult, Habecardium has been recognized as transitional from the Protocardiinae to the Laevicardiinae (Glibert and van de Poel, 1970; Keen, 1980). Glibert and van de Poel (1970) erected Habecardium as a subgenus of Laevicardium, into which some of the species of Habecardium had been placed previously. Keen (1980) placed Habecardium as a subgenus of Nemocardium. Popov (1977) and Kafanov and Popov (1977) also placed Habecardium as a subgenus of Nemocardium, but did not recognize it as transitional to Laevicardium and Fulvia, placing the latter two taxa in the Cardiinae. Besides ribbing pattern (24:2), the Laevicardiinae are united by the number of ctenidial plicae (5:2), tentacles that extend only to the bottom of the posterior adductors (10:0), presence of complex eyes (12:1), a centrally located right caecum (18:1), and shape of the cardinal teeth (43:3 and 45:3). Keen (1969a, 1980) placed all the Cenozoic proto- cardiines in the genus Nemocardium. Other authors (Fischer- Piette, 1977; Popov, 1977; Wilson and Stevenson, 1977; Noda, 1988; Voskuil and Onverwagt, 1989) have raised some of the subgenera to the generic level. It is suspected strongly that the subtraction of Habecardium from Nemocardium would leave the latter as a monophyletic group. I decline here to place Nemocardium within a subfamily. A systematic analysis which includes all of the subgenera of Nemocardium as in Keen (1969a, 1980) represented, plus the Laevicardiinae, is in progress. Jurassicardium is a monotypic taxon known from only a few specimens. Only the type material is sufficiently well preserved to be of systematic use, and I have not examined it. The remainder of the Cardiidae comprise those forms typically accepted to constitute the taxa Cardiinae, Trachy- cardiinae, Fraginae, Clinocardiinae, Lymnocardiidae, and Tridacnidae. Two monophyletic clades can be distinguished within this unnamed taxon. One clade, here considered the subfamily Cardiinae, contains the taxa placed in the Trachy- cardiinae and most of the taxa placed in the Cardiinae by Keen (1969a, 1980), and all of taxa placed in the Cardiinae (except for Laevicardium and Fulvia) by Kafanov and Popov (1977). In agreement with Kafanov and Popov (1977), the present results indicate that Trachycardium and the related taxa Acrosterigma and Vasaticardium do not constitute a separate subfamily but are members of the Cardiinae. The least derived monophyletic group within the Cardiinae contains the taxa Acanthocardia, Rudicardium, Sawkinsia, and Chrysocardium. Synapomorphies of this clade are cardiiform shell shape (23:5), tuberculate spines (28:5) and irregular cross-striae (30:1). Rudicardium is considered either a subgenus of Acanthocardia (Keen, 1969, 1980; Popov, 1977) or a synonym of it (Voskuil and Onverwagt, 1989). These two taxa are united by a suite of hinge characters: in- complete anterior cardinal socket (42:1); shape of the cardi- nal teeth (43:8 and 45:10); hinge plate overlapping the right posterior lateral socket (48:1). Cox (1941) erected Sawkinsia as a genus of cardiid. Vokes (1953) placed Sawkinsia in the Tridacnidae, and was followed by Rosewater (1965), Keen (1969b), and Jung (1976). Stasek (1962) considered the resemblance between Sawkinsia and the tridacnid Hippopus to be a case of convergence. Sawkinsia does not share any of the derived characters of Tridacna, nor any of Tridacna’s notable autapomorphies: (1) there is no loss of the anterior lateral teeth; (2) the spines are tubercles, not wide and gently curved; (3) nor is there any evidence of the rotation of the shell about the animal. According to the present phylogenetic hypothesis, Sawkinsia is a member of the subfamily Cardiinae. Woodring (1982) erected the genus Chrysocardium in the subfamily Fraginae based on a single left valve. Chryso- cardium shares not only a lunule flap touching the beak (25:3) with Sawkinsia, but three characters found nowhere else in the Cardiidae: hinge inversion (36:1) (described for Sawkin- sia by Cox, 1941); weak myophorous buttress (38:1); double keel (41:1). Chrysocardium should be considered a synonym of Sawkinsia, however C. aurum Woodring, appears to be valid. Except for missing data, the characters for two taxa are scored identically (see appendix 4). The next monophyletic group includes Bucardium, Cardium, Vepricardium, and Dinocardium, and is united by seven synapomorphies. The close relationship of the first three taxa to each other has been recognized by numerous authors (see Keen, 1969a; Kafanov and Popov, 1977). The position of Dinocardium, however, remains uncertain. Keen (1951, 1969a, 1980) placed Dinocardium in the Laevicardiinae. Kafanov and Popov (1977), in dismantling the Laevicardiinae, tentatively placed Dinocardium in the Cardiinae; Kafanov (1980:298) called the taxonomic position of Dinocardium ‘*most mysterious.’’ The next node within the Cardiinae contains Austro- cardium and Cardium acuticostatum, and is united by lack of lunule flap (25:0) and shape of the anterior cardinal (45:5). Freneix and Grant-Mackie (1978) erected the Cretaceous Austrocardium as a monotypic taxon. Wilckens (1904) placed the Cretaceous form C. acuticostatum in Cardium (Ringi- carcium) [=Cardium (Bucardium)], which is otherwise a Miocene to Recent taxon (Keen, 1969a, 1980). The results of my analysis indicate that C. acuticostatum belongs in SCHNEIDER: CLADISTIC ANALYSIS OF THE CARDITDAE 149 Austrocardium. There are three other Cretaceous species that differ little from C. acuticostatum and Austrocardium. These are: (1) C. denticulatum Baily, which was placed by Dartevelle and Freneix (1957) in Acanthocardia (Acanthocardia); (2) C. (Bucardium) lillei Freneix and Grant-Mackie (specimens of which had originally been described as C. acuticostatum); (3) Schedocardia ? waiparana Freneix and Grant-Mackie. These species should be placed provisionally in Austrocar- dium, as they share the apomorphies of Austrocardium but not those of either Bucardium or Acanthocardia. Except for cases of missing data, the posterior gape of C. acuticostatum (33:1) is the only character not scored identically to that of Austrocardium (appendix 4). The posterior gape is convergent with that of Cardium. Hedecardium, Orthocardium, and Agnocardia form a monophyletic clade. Hedecardium has been considered variously as a subgenus of Vepricardium (Keen, 1969a, 1980), a genus closely related to Vepricardium (Popov, 1977), and a genus in the Protocardiinae (Marwick, 1960; Maxwell, 1978). The latter authors derived Hedecardium from Nemocardium on the basis of its discrepancy in thickness in the ribs across the shell. However, the rib discordance in Hedecardium is not comparable to that in Nemocardium. In Hedecardium, four to six ribs on the posterior slope are split with a furrow running down the middle, and the remaining posterior ribs are reduced in strength (as happens numerous times in the Cardiidae; it was the basis of Keen’s [1936] sub- family Laevicardiinae, shown to be polyphyletic). This change from all ribs of equal width to the rib discrepancy seen in Hedecardium can be seen in the growth stages of a single shell. Likewise, the early growth stages of Hedecardium are circular, and circular shells are unknown in any form of Nemocardium. As a juvenile, Hedecardium would strongly resemble Orthocardium. It is recommended that Hedecar- dium and Orthocardium be considered as distinct genera. Orthocardium has been considered a subgenus of Vepricardium (Keen, 1969a) or of Cardium (Keen, 1980) or a genus of fragine (Popov, 1977; Kafanov and Popov, 1977). Here, Orthocardium is united with Agnocardia and Hede- cardium by concave ribs (32:1), a condition otherwise unknown in the Cardiidae. The last clade within the Cardiinae comprises Grano- cardium, Trachycardium, Acrosterigma, and Vasticardium, and is united by one character, ovate shell shape (23:3). The latter three taxa, united by three synapomorphies, are usual- ly placed in the Trachycardiinae (Keen, 1969a, 1980) or the tribe Trachycardiini (Kafanov and Popov, 1977) within the Cardiinae. Current results support the latter. The other major clade of cardiids is united by five synapomorphies and generally contains forms that have been assigned to the Clinocardiinae, Lymnocardiidae, Tridacnidae, and Fraginae. Of these the least derived is Clinocardium. This taxon was placed in the Laevicardiinae by Keen (1951, 1969a, 1980), but has come to be accepted as the type of the sub- family Clinocardiinae (Kafanov and Popov, 1977; Kafanov, 1980; Voskuil and Onverwagt, 1989). The next node en- countered is Cerastoderma and Hypanis, which is the sister taxon to the Tridacna and Fraginae. Synapomorphies of lymnocardiids + Tridacna + Fraginae are medium labial palps (1:1), functional byssus in adult (6:2), posterior cardinal socket angle (44:1) and shape of the anterior cardinal (45:9). Yonge’s (1936) and Stasek’s (1962) suggestion that the ancestry of Tridacna’s was close to that of Cerastoderma is upheld by the results. Giant clams should be considered as the sub- family Tridacninae within the Cardiidae. Cerastoderma and Hypanis form a monphyletic group. Therefore, as has been argued by eastern European mala- cologists for some time (Kafanov and Popov, 1977; Popov, 1977; Taktakishvili, 1987; Basch, 1990), the brackish-water forms should be subfamily Lymnocardiinae, and contain Cerastoderma. Furthermore, the results support Kafanov and Popov’s (1977) contention that the Lymnocardiinae are related closely to the Fraginae. Five of the seven characters that unite Tridacna and the Fraginae are anatomical: tentacle pattern (9:2); large valvule (11:2); centrally located style sac (15:1); presence of a raised bar on the stomach floor (17:1); posteriorly located left caecum (19:0). The Fraginae is united by the presence of ventral appendages on the foot (7:1), absence of a peri- phonal suture (8:1), and presence of a mosaicostracum (29:1). The least derived taxa, Plagiocardium and Parvicardium, were placed in the Fraginae by Kafanov and Popov (1977) but in the Cardiinae by Keen (1969a, 1980). Apiocardia, Trigonio- cardia, and Fragum are united by eight synapomorphies, all based on hinge characters. Finally, Trigoniocardia and Fragum are united by six synapomorphies, five of which are anatomical characters: short labial palps (1:0); fewer than ten ridges on the palps (2:1); the inner palp connected to the bot- tom of the inner demibranch (3:1); fewer than 20 ctenidial plicae (5:0); type 4 gut (13:3); quadrate shell shape (23:2). From the above discussion it can be concluded that the family Cardiidae includes nine subfamilies: Septo- cardiinae; Protocardiinae; Lahilliinae; Laevicardiinae; Cardiinae; Clinocardiinae; Tridacninae; Lymnocardiinae; Fraginae. Taxa usually assigned to the Trachycardiinae form a monophyletic group within the Cardiinae. Dinocardium is a cardiine closely related to Cardium and Vepricardium. Sawkinsia (=Chrysocardium) is transferred from the Tridac- ninae to the Cardiinae, as it is related closely to Acantho- cardia and Rudicardium. The Protocardiinae (as presented in the Treatise on In- vertebrate Paleontology) is paraphyletic. Integricardium is a member of the Lahilliinae. Nemocardium is the sister taxon to the Laevicardiinae. The subfamilies Clinocardiinae, Tridac- ninae, Lymnocardiinae, and Fraginae form a monophyletic group. 150 AMER. MALAC. BULL. 9(2) (1992) ACKNOWLEDGMENTS Numerous people and institutions have allowed me to examine their collections and have lent specimens. Foremost, I would like to thank the Academy of Natural Sciences of Philadelphia, my supervisor at the Academy, G. Rosenberg, and G. Davis and the rest of the staff of the depatments of Malacology and Paleontology of that institution. I would like to thank the following people and institutions for allowing me to examine their collec- tions: T. Waller, W. Blow, R. Hershler, R. Houbrick, J. Harasewych (United States National Museum); M. Hinckley, N. Eldredge (American Museum of Natural History); C. Coney (Los Angeles County Museum of Natural History); M. Baker, M. Nitecki, J. Voight, R. Bieler (Field Museum of Natural History). I would like to thank the following people for lending specimens: W. Ponder, Australian National Museum; A. Beu, New Zealand Geological Survey; J. A. Grant-Mackie, University of Auckland; S. Freneix, J. -C. Fisher, and Y. Gayrard, Museum National d’ Histoire Naturelle, Paris. I examined specimens of Palaeocardita and Septocardia at C. R. Newton’s laboratory at Syracuse University. I learned techniques for the study of shell microstructure at J. G. Carter’s laboratory at the University of North Carolina- Chapel Hill. The following agencies have generously given me financial assistance for my research: The Academy of Natural Sciences of Philadelphia (Jessup Fund), the National Capitol Shell Club, the Western Society of Malacologists, the Paleontological Society, Conchologists of America, the Santa Barbara Shell Club, the Hinds Fund of the University of Chicago, Sigma Xi, the Gurley Fund of the Department of Geophysical Sciences of the University of Chicago, and the Amoco Corporation. During part of my graduate study I was supported by a grant from the National Science Foundation (EAR-90-05744) to D. Jablonski. I would also like to thank the following people at the University of Chicago who have assisted me in my research: D. Jablonski, M. LaBarbera, O. Draughn, K. Roy, P. Wagner, and D. Miller. LITERATURE CITED Atkins, D. 1937. On the ciliary mechanisms and interrelationships of Lamellibranchs. Part III: Types of lamellibranch gills and their food currents. Quarterly Journal of Microscopical Science 79:375-421. Basch, O. 1990. Cardiidae (Mollusca, Lamellibranchiata) Pontskog kata u Hrvatskoj. Paleontologia Jugoslavica 39:1-158. Boss, K. 1971. Familial affinities of Hemidonax (Bivalvia). Nautilus 85:9-12. Braun, R. 1954. Zum Lichtsinn augenloser Muscheln. Zoologische Jahrbucher 65:194-208. Carter, J. G. 1980. Guide to bivalve shell microstructures. Jn: Skeletal Growth of Aquatic Organisms, D. C. Rhoads and R. A. Lutz, eds. pp. 645-673. Plenum Press, New York. Carter, J. G., ed. 1989. Skeletal Biomineralization: Patterns, Processes and Evolutionary Trends. Volume I. Van Nostrand Reinhold, New York, 832 pp. Carter, J. G. and G. R. Clark, III. 1985. Classification and phylogenetic significance of molluscan shell microstructure. University of Tennessee Department of Geological Sciences Studies in Geology 13:50-/1. Carter, J. G. and R. A. Lutz. 1989. Bivalvia (Mollusca). In: Skeletal Biomineralization: Patterns, Processes and Evolutionary Trends. Volume II, Atlas and Index, J. G. Carter, ed. pp. 5-28. Van Nostrand Reinhold, New York. Chavan, A. 1969. Superfamily Carditacea Fleming, 1820. In: Treatise on Invertebrate Paleontology, Part N, Vol. 2, Mollusca 6, Bivalvia, R. C. Moore, ed. pp. 543-561. University of Kansas and the Geological Society of America, Lawrence, Kansas. Cox, L. R. 1949. Moluscos del Triasico Superior del Peru. Instituto Geologico del Peru, Boletin 12:5-48. Dall, W. H. 1901. Synopsis of the family Cardiidae and of the North American species. Proceedings of the United States National Museum 23:381-392. Dartevelle, E. and S. Freneix. 1957. Mollusques fossiles du Cretace de la Cote occidentale d’Afrique du Cameroun a 1’Angola, II. Lamelli- branches. Annales du Musée Royal du Congo Belge, Tervuren (Belgique), serie 8, 20:271 pp. D’Attilio, A. and G. E. Radwin. 1971. The intritacalx, an undescribed shell layer in mollusks. Veliger 13:344-347. Day, W. R. 1978. Onestia McLearn, an unusual cardiacean pelecypod from the Lower Cretaceous of Australia and Canada. Bureau of Mineral Resources, Geology and Geophysics, Bulletin 192:37-44. Finlay, H. J. and J. Marwick. 1937. The Wangaloan and associated molluscan faunas of Kaitangata-Green Island subdivision. New Zealand, Depart- ment of Scientific and Industrial Research, Geological Survey Branch, Palaeontological Bulletin 15:140 pp. Fischer-Piette, E. 1977. Revision des Cardiidae (Mollusques Lamelli- branches). Mémoires du Museum National d’Histoire Naturelle 101A:212 pp. Freneix, S. and J. A. Grant-Mackie. 1978. New Upper Cretaceous Cardiidae (Bivalvia) from New Zealand and New Caledonia. New Zealand Journal of Geology and Geophysics 21:505-51S. Graham, A. 1949. The molluscan stomach. Transactions of the Royal Society of Edinburgh 61:737-778. Hamilton, G. H. 1969. The taxonomic significance and theoretical origin of surface patterns on a newly discovered bivalve shell layer, the mosaicostracum. Veliger 11:185-194. Johnstone, J. 1899. Cardium. Liverpool Marine Biological Committee Memoirs 11:84 pp. Jung, P. 1976. The Eocene genus Sawkinsia (Bivalvia: Tridacnidae). Eclogae geologicae Helvetiae 69:743-751. Kafanov, A. I. 1980. Systematics of the subfamily Clinocardiinae Kafanov, 1975 (Bivalvia: Cardiidae). Malacologia 19:297-328. Kafanov, A. I. and S. V. Popov. 1977. On the system of the Cenozoic Cardioidea (Bivalvia). Paleontological Journal 1977:307-314. Keen, A. M. 1936. Revision of cardiid pelecypods. Proceedings of the Geological Society of America (1935) 48:367. Keen, A. M. 1937. Nomenclatural units of the pelecypod family Cardiidae. Bulletin du Musee royal d’Histoire naturelle de Belgique 13:1-22. Keen, A. M. 1951. Outline of a proposed classification of the pelecypod family Cardiidae. Minutes of the Conchological Club of Southern California 111:6-8. Keen, A. M. 1969a. Superfamily Cardiacea Lamarck 1809. In: Treatise on Invertebrate Paleontology, part N, volume 2, Mollusca 6, Bivalvia, R. C. Moore, ed. pp. 583-594. University of Kansas and the Geological Society of America, Lawrence, Kansas. Keen, A. M. 1969b. Superfamily Tridacnacea Lamarck 1819. In: Treatise on Invertebrate Paleontology, part N, volume 2, Mollusca 6, Bivalvia, R. C. Moore, ed., pp. 594-595. University of Kansas and the Geological Society of America, Lawrence, Kansas. Keen, A. M. 1980. The pelecypod family Cardiidae: a taxonomic summary. Tulane Studies in Geology and Paleontology 16:1-40. Kishinouye, K. 1894. Note on the eyes of Cardium muticum Reeve. The Journal of the College of Science, Imperial University, Tokyo 6:279-285. Kifer, M. 1915. Entwickslungsgeschichte und neuro-histologische Unter- suchungen an Sehorganen am Mantelrande der Pecten-Art mit anschliessenden vergleichend-anatomischen Betrachtungen. Viertel- jahrsschrift der Naturforschenden Gesellschaft in Zurich 60:568-591. Marwick, J. 1944. New Zealand fossil and Recent Cardiidae (Mollusca). Transactions of the Royal Society of New Zealand 714:255-272. Marwick, J. 1960. Early Tertiary Mollusca from Otaio Gorge, South Canter- bury. New Zealand Geological Survey Bulletin 33:32 pp. Maxwell, P. A. 1978. Taxonomic and nomenclatural notes on some New Zealand Cenozoic Mollusca, with descriptions of new taxa. New SCHNEIDER: CLADISTIC ANALYSIS OF THE CARDITDAE 151 Zealand Journal of Zoology 5:15-46. McLearn, F. H. 1933. Pelecypods of the Lower Cretaceous Clearwater For- mation, Northern Alberta. Transactions of the Royal Society of Canada, section 4, series 3, 27:139-156. McLearn, F. H. 1945. Revision of the Lower Cretaceous of the Western In- terior of Canada (2nd Ed.). Geological Survey of Canada, Paper 44-17:1-14. Menegaux, A. 1890. Recherches sur la Circulation des Lamellibranches Marins. J. Dodivers, Besancon. 296 pp. Morris, N. J. 1978. The infaunal descendants of the Cycloconchidae: an outline of the evolutionary history and taxonomy of the Heteroconchia, super- families Cycloconchacea to Chamacea. Philosophical Transactions of the Royal Society of London 284B:259-27S. Morris, N. J., J. M. Dickins and K. Astafieva-Urbaitis. 1991. Upper Paleozoic Anomalodesmatan Bivalvia. Bulletin of the British Museum of Natural History (Geology) 47:51-100. Nagel, W. 1897. Uber rathselhafte Organe en den Siphopapillen von Cardium oblongum. Zooligischer Anzeiger 20:406-409. Nakazima, M. 1964. On the differentiation of the crenated folds in the midgut gland of the eulamellibranchia (VI): crenated folds in the Cardiacea. Venus 23:143-148. Newton, C. R. 1986. Late Triassic bivalves of the Martin Bridge Forma- tion, Hells Canyon, Oregon: taphonomy, paleoecology, paleo- zoogeography. United States Geological Survey Professional Paper 1435:7-22. Newton, C. R., M. T. Whalen, J. B. Thompson, N. Prins and D. Delalla. 1987. Systematics and paleoecology of Norian (Late Triassic) bivalves from a tropical island arc: Wallowa Terrane, Oregon. The Paleon- tological Society, Memoir 22:1-83. Noda, H. 1988. Molluscan fossils from the Ryukyu Islands, southwest Japan. Part 2. Gastropoda and Pelecypoda from the Shinzato Formation in the middle part of Okinawa-jima. Science Reports of the Institute of Geosciences, University of Tsukuba, section B, 9:29-85. Pelseneer, P. 1911. Les lamellibranches de 1’ expedition du Siboga, partie anatomique. Siboga Expeditie 53a:125 pp. Ponder, W. F., P. H. Colman, C. M. Yonge and M. H. Colman. 1981. The taxonomic postion of Hemidonax Morch, 1871 with a review of the genus (Bivalvia: Cardiacea). Journal of the Malacological Society of Australia 5:41-64. Popov, S. V. 1977. Mikrostruktura rakovinii i sistematika kardiid. Akademiia Nauk SSSR 153:1-124. Popov, S. V. 1986. Composite prismatic structure in bivalve shell. Acta Palaeontologica Polonica 31:3-26. Purchon, R. D. 1959. Phylogenetic classification of the Lamellibranchia with special reference to the Protobranchia. Proceedings of the Malacological Society of London 33:224-230. Purchon, R. D. 1960a. The stomach in the Eulamellibranchiata; stomach types IV and V. Proceedings of the Zoological Society of London 135:431-489. Purchon, R. D. 1960b. Phylogeny in the Lamellibranchia. In: Proceedings of the Centenary and Bicentenary Congress of Biology, Singapore, 1958, R. D. Purchon, ed. pp. 69-82. University of Malaysia Press, Singapore. Roche, W. L. 1925. Note on the microscopic anatomy of the tentacular sense organ of Cardium edule. Journal of the Royal Microscopical Society 1925:154-158. Rosewater, J. 1965. The family Tridacnidae in the Indo-Pacific. Indo-Pacific Mollusca 1:347-396. Scarlato, O. A. and Ya. I. Starabogatov. 1979. Osnovye cherty evolyutsii i sistema Klassa Bivalvia, morfologiya, sistematika i filogeniya mollyuskov. Trudy Zoologicheskogo Instituta, Akademiia Nauk SSSR 80:5-38. Stasek, C. 1962. The form, growth, and evolution of the Tridacnidae (giant clams). Archives de Zoologie Experimentale et Generale 101:1-40. Stewart, R. B. 1930. Gabb’s California Cretaceous and Tertiary type lamellibranchs. Academy of Natural Sciences, Special Publication 3:314 pp. Stoliczka, F. 1871. Cretaceous fauna of southern India, volume III. The Pelecypoda, with a review of all known genera of this class, fossil and recent. Palaeontologia Indica, series 6, 3:537 pp. Swofford, D. 1989. Phylogenetic Analysis Using Parsimony (PAUP). Illinois Natural History Survey. Taktakishvili, I. G. 1987. Sistematika i filogeniya Pliotsenovikh kardiid paratetica. Akademiia Nauk Gruzinskoi SSR, Tbilisi. 247 pp. Vokes, H. E. 1953. Preliminary classification of the genera of the Pelecypoda (unpublished, cited in Rosewater 1965). Vokes, H. E. 1977. Cardiidae (Mollusca: Bivalvia) from the Chipola For- mation, Calhoun County, Florida. Tulane Studies in Geology and Paleontology 13:143-189. Vokes, H. E. 1980. Genera of the Bivalvia: a Systematic and Bibliographic Catalogue. Paleontological Research Institution, Ithaca, New York. 307 pp. Vokes, H. E. 1984. Notes on the genus Agnocardia (Mollusca: Cardiidae) with the description of a new species from the Pliocene of Florida. Tulane Studies in Geology and Paleontology 18:37-45. Vokes, H. E. 1989. Neogene paleontology in the northern Dominican Republic, 9. The family Cardiidae (Mollusca: Bivalvia). Bulletins of American Paleontology 97:95-161. Voskuil, R. P. A. and W. P. H. Onverwagt. 1989. Inventarisation of the Re- cent European and west African Cardiidae (Mollusca, Bivalvia). Gloria Maris 28:49-96. Watabe, N. 1984. Shell. Jn: Biology of the Integument, 1, Invertebrates, J. Bereiter-Hahn, A. G. Matoltsy, K. Sylvia-Richards, eds. pp. 448-485. Springer-Verlag, New York. Weber, F. L. 1908. Uber sinnesorgane des Genus Cardium. Arbeiten aus dem Zoologische Institut Wien and Triest 17:187-220. Wilckens, O. 1904. Revisions der Fauna der Quiriquina-Schichten. Neues Jahrbuch fur Mineralogie, Geologie, und Palaontologie, Beilage-Band 18:181-284. Wilson, B. R. and S. E. Stevenson. 1977. Cardiidae of Western Australia. Western Australia Museum, Special Publication 9:114 pp. Woodring, W. 1982. Geology and Paleontology of Canal Zone and adjoin- ing parts of Panama. Description of Tertiary mollusks (Pelecypods: Propeamussidae to Cuspidariidea; additions to families covered in P 306-E; additions to gastropods, cephalopods). United States Geological Survey Professional Paper 306-F:541-759. Yonge, C. M. 1936. Mode of life, feeding, digestion and symbiosis with zoo- xanthellae in the Tridacnidae. British Museum (Natural History) Great Barrier Reef Expedition, 1928-1929, Scientific Reports 1:283-321. Yonge, C. M. 1969. Functional morphology and evolution within the Carditacea (Bivalvia). Proceedings of the Malacological Society of London 38:495-527. Zugmayer, E. 1904. Uber Sinnesorgane an den Tentakeln des Genus Cardium. Zeitschrift fur Wissenschaftliche Zoologie 1904:478-508. Date of manuscript acceptance: 25 September 1991 152 AMER. MALAC. BULL. 9(2) (1992) APPENDIX 1. APPENDIX 2. Classification of taxa used in this study following Kafanov and Popov (1977). Classification of taxa used in this study according to Keen (1969a, b; Taxa not listed are not discussed therein. 1980). Taxa not listed are not considered in any of Keen’s (1969a, 1969b, 1980) papers. Superfamily Cardioidea Family Cardiidae Superfamily Cardiacea Subfamily Protocardiinae Family Cardiidae Protocardium Subfamily Cardiinae Integricardium Septocardia Nemocardium Granocardium Habecardium Vepricardium (Vepricardium) Subfamily Cardiinae Tribe Cardiini Cardium Tribe Vepricardiini Vepricardium Vepricardium (Orthocardium) Vepricardium (Hedecardium) Cardium (Cardium) Cardium (Bucardium) Acanthocardia (Acanthocardia) Agnocardia Acanthocardia (Rudicardium) Bucardium Acanthocardia (Agnocardia) Hedecardium Plagiocardium Tribe Trachycardiini Parvicardium Trachycardium Subfamily Trachycardiinae Acrosterigma Trachycardium Vasticardium Tribe Laevicardiini Fulvia Tribe Dinocardiini Dinocardium Subfamily Clinocardinae Tribe Clinocardiini Acrosterigma (Acrosterigma) Acrosterigma (Vasticardium) Subfamily Protocardiinae Protocardia Integricardium Nemocardium (Nemocardium) Nemocardium (Habecardium) Clinocardium Subfamily Fraginae Subfamily Fraginae Fragum Tribe Fragini Trigoniocardia (Trigoniocardia) Trigoniocardia Trigoniocardia (Apiocardia) Apiocardia Subfamily Laevicardiinae Fragum Laevicardium (Fulvia) Tribe Acanthocardiini Laevicardium (Dinocardium) Rudicardium Cerastoderma Acanthocardia Clinocardium Tribe Parvicardiini Family Lahillidae Plagiocardium Lahillia Parvicardium Family Lymnocardiidae Orthocardium Hypanis Subfamily Lymnocardiinae Superfamily Tridacnacea Tribe Cerastodermatini Family Tridacnidae Cerastoderma Tridacna Tribe Hypanini Sawkinsia Hypanis Superfamily Tridacnoidea Family Septocardiinae Septocardia Family Tridacnidae Tridacna Superfamily Arcticoidea Family Lahilliidae Lahillia SCHNEIDER: CLADISTIC ANALYSIS OF THE CARDITDAE 153 APPENDIX 3. List of characters and character states. I. Anatomy A. Labial palps 1. Length: 0) short, 1) medium, 2) long 2. Number of ridges on palps: 0) ridges absent, 1) <10, 2) 10 - 19, 3) 20 - 29, 4) >29 3. Connection of inner palp with ctenidia: 0) connects behind inner demibranch, 1) connects with bottom of inner demibranch B. Ctenidia 4. Inner demibranch/outer demibranch relation: 0) Outer demibranch does not overlay inner demibranch, 1) Outer demibranch partially underlain by inner demibranch 5. Number of plicae: 0) <20, 1) 20 - 39, 2) 40 - 59, 3) 60 - 79, 4) 80 - 99, 5) >99 C. Foot Byssal apparatus: 0) absent in adult, 1) present in adult, but non- functional, 2) functional in adult (see Pelseneer, 1911) 7. Ventral appendages: 0) absent, 1) present D. Siphons and tentacles 8. Periphonal suture: 0) present, 1) absent (see Pelseneer, 1911) 9. Tentacle pattern: 0) absent, 1) numerous, in both mantle fold and on siphonal area, 2) numerous, in mantle fold only, 3) few, in mantle fold and siphonal area, 4) few on siphonal area only 10. Dorsalmost extension of tentacles: 0) bottom of adductors, 1) to middle of adductors, 2) to top of adductors, 3) beyond top of adductors ll. Valvule (see Pelseneer, 1911) 0) absent, 1) small, 2) large 12. Eyes (see Kishinouye, 1894; Nagel, 1897; Zugmayer, 1904; Weber, 1908; Pelseneer, 1911; Braun, 1954) 0) simple, 1) complex E. Gut 13. 0) Type 1, 1) Type 2, 2) Type 3, 3) Type 4, 4) Type 5 F. Stomach (see Graham, 1949 and Purchon, 1960a, for general descrip- tions of bivalve stomachs) 14. T3 (tertiary typhlosole): 0) absent, 1) present 15. Position of style sac: 0) posterior, 1) central 16. Tl (major typhlosole) curved (see Nakazima, 1964): 0) yes, 1) no 17. raised bar: 0) absent, 1) present 18. Position of right caecum: 0) right side of stomach, 1) central 19. Position of left caecum: 0) posterior to right caecum 1) caeca parallel, 2) anterior to right caecum 20. Accessory left caeca: 0) absent, 1) present Il. Shell A. General 21. Posterior margin: 0) digitate, 1) crenulate, 2) smooth 22. Rib number: 0) absent, 1) less than 70, 2) greater than 70 23. shell shape: 0) carditaform, 1) quadrate - long, 2) quadrate - short, 3) ovate, 4) circular, 5) cardiiform, 6) oval, 7) cerastiform, 8) trigonal, 9) elliptical, 10) oblique Character state tree: ((((5,(7,8)6)4,10,3)2,9)1)0 24. Anterior/central rib pattern: 0) concentric, 1) radial, equal in width to posterior ribs, 2) concentric, changing to radial, 3) rib dis- cordance, 4) radial, thinner than posterior ribs, 5) none Character state tree: (((2)4,5)0,3)1 25. Lunule flap (see Kafanov, 1980, pp. 298-299): 0) absent, 1) raised, does not block beak, 2) blocks beak but does not touch it, 3) touches beak, 4) strongly folded over beak 26. Ridges on lunule flap: 0) absent, 1) present 27. Growth line strength: 0) strong, 1) weak 28. Spines: 0) lumpy nodes, 1) none, 2) round knobs, 3) A-shaped, separate shell layer from ribs, 4) hollow posterodorsally, 5) tubercles, 6) gently curved, 7) hollow keel Spines are defined as emanating from the top of the ribs. The ‘‘spines’’ of trachycardiines emanate from the side of the ribs, and are considered separately (character 40). 29. Mosaicostracum (‘‘spines’’ on fragines): 0) none, 1) beads, 2) scales Mosaicostracum was first described by Hamilton (1969). Keen (1980) and Vokes (1977, 1989) referred to this layer as the intritacalx (D’Attilio and Rad- win, 1971). Carter (1989) considers intritacalx synonymous with mosaicostracum. This character is linearly ordered. 30. Cross-striae: 0) simple, 1) irregular, 2) absent 31. Internal rib expression: 0) weak, 1) strong 32. Ribs concave: 0) no, 1) yes 33. Posterior gape: 0) absent, 1) present 34. Dorsal nymph extension, 0) absent, 1) present 35. Posterior umbonal buttress: 0) absent, 1) present 36. Hinge inversion: 0) absent, 1) present 37. Nymph groove: 0) absent, 1) present 38. Myophorous buttress: 0) strong, 1) weak, 2) absent 39. Adductor scar relief: 0) strong, 1) weak 40. Rib flares (‘‘spines’’ on trachycardiines): 0) absent, 1) strong, 2) weak This character is linearly ordered. 41. Double keel: 0) absent, 1) present B. Hinge teeth 1) Right cardinal teeth 42. Anterior socket: 0) complete, 1) incomplete 43. Posterior cardinal shape: 1) 1, 2) 2, 3) 3, 4) 4, 5) 5, 6) 6, 7) 7, 8) 8, 9) 9 44. Angle of posterior cardinal socket to horizontal: 0) greater than 40 degrees, 1) equal or less than 40 degrees 45. Shape of anterior cardinal: 0) 0, 1) 1, 2) 2, 3) 3, 4) 4, 5) 5, 6) 6, 7) 7, 8) 8, 9) 9 Rest of hinge 46. Right anterior lateral teeth - ventral tooth continues up into umbo: 0) no, 1) yes 47. Right anterior lateral teeth - ventral tooth inserts into socket: 0) no, 1) yes 48. Right posterior laterals: hinge plate overlaps socket: 0) no, 1) yes 49. Anterior lateral teeth: 0) absent, 1) present 50. Left posterior lateral teeth: 0) weak or absent, 1) strong 51. Nymph overlies posterior cardinals: 0) no, 1) yes 52. Left anterior lateral horizontal: 0) no, 1) yes 53. Left anterior lateral socket: 0) absent or weak, 1) strong 54. Left posterior lateral socket: 0) absent or weak, 1) strong 154 Data matrix for cladistic analysis. 66g? AMER. MALAC. BULL. 9(2) (1992) APPENDIX 4. signifies missing data. X indicates character state 10. A-D indicate polymorphisms. A: states 1 and 2; B: states 0 and 1; C: states 2 and 3; D: states 1, 2, and 3. Taxa outgroup Palaeocardita Septocardia Protocardia Integricardium Lahillia Nemocardium Habecardium Fulvia Laevicardium Granocardium Cardium acusticostatum Austrocardium Chrysocardium Sawkinsia Hedecardium Orthocardium Vepricardium Cardium Agnocardia Bucardium Acanthocardia Rudicardium Plagiocardium Clinocardium Dinocardium Trachycardium Vasticardium Acrosterigma Parvicardium Apiocardia Trigoniocardia Fragum Cerastoderma Hypanis Tridacna 1 2 3 4 5 123456789012345678901234567890123456789012345678901234 22014100101021010 227272229792 799027727799727015141170210110012100281X100100000 ~) ~ ) ~ ~ ~) ~) ) ) ~ ~) i=) a) ) a) ~~ oo) oo) ~~ i) j=) - a i=) nN i=) o i) i=) o Ne j=) j=) No Ne ) i=) i=) j=) j=) j=) j=) Oo i=) ~ ~ ~ ~ a) ~ ~) ~) ~) ~ ~) ~] ~) ~ x) ~) ~) ~ wn N So - oO =) So So o o j=) fo) nN =) fo) oo j=) x =) So oS So io) o 231051103210200100101161101102000000020001508000100000 230151001310200101101151411102100100121000202101100000 240041001310200100200131D01100000000020100202001110000 220041001210200100200131101100000000020200406000110000 011122012320301110000121301120000000020001609000120011 12012A0031104001001011710?1102000000020001519001100000 140112004710400100101171101102001000020001919??70000700 Synapomorphies for interior nodes. Nodes numbered as in figure 1. Terminal SCHNEIDER: CLADISTIC ANALYSIS OF THE CARDITDAE 155 APPENDIX 5. taxa not diagnosed. Node Synapomorphies (Character: State) 2:2, 5:1, 9:1, 11:1, 12:1, 14:0, 16:1, 20:0, 43:1 23:1, 25:1 24:0, 28:1, 30:2, 38:2, 49:1 27:1 21:2, 22:0, 23:9 23:2, 43:2, 45:2 24:4, 26:3, 48:1 5:2, 10:0, 12:1, 18:1, 24:2, 43:3, 45:3 39:1, 42:1 1:2, 6:1, 13:2, 19:1, 23:4, 24:1 5:4, 10:3, 25:2 23:5, 28:5, 30:1 25:3, 36:1, 38:1, 41:1 42:1, 43:8, 45:10, 48:1 21:0 18:1, 25:4, 26:1, 31:1, 34:1, 39:1, 42:2, 48:1 37:1, 46:1 2535 5393. 2329 4:0, 19:2, 50:1 25:0, 45:5 28:2, 30:0 32:1 42:1 23:3 25:1, 28:1, 40:1 10:2, 40:2, 43:4, 45:6 9:3, 23:6, 42:1, 43:5, 45:4 1:1, 6:2, 44:1, 45:9 13:4, 23:7 9:2, 11:2, 15:1, 17:1, 19:0, 23:8, 30:0 7:1, 8:1, 29:1 42:0, 43:6, 47:1, 50:2, 51:1, 52:1, 53:1, 54:1 1:0, 2:1, 3:1, 5:0, 13:3, 23:2 APPENDIX 6. Taxa represented by species other than type. Palaeocardita, type species Palaeocardita austriaca (Hauer). Species examined: Palaeocardita silberlingi Newton et al. Integricardium, type species Integricardium dupinianum (d’Orbigny). Species examined: Integricardium globulum (Whitfield). Lahillia, type species: Lahillia angulata (Philippi). Species examined: Lahillia larseni (Sharman and Newton). Nemocardium, type species Nemocardium semiasperum (Deshayes). Species examined: Nemocardium bechei (Reeve). Laevicardium, type species Laevicardium oblongum (Gmelin). Species examined: Laevicardium laevigatum (Linne). Granocardium, type species Granocardium carolinum (d’Orbigny). Species examined: Granocardium dumosum (Conrad). Agnocardia, type species Agnocardia claibornense (Aldrich). Species examined: Agnocardia dissidepictum (Woodring). Trigoniocardia, type species Trigoniocardia granifera (Broderip and Sowerby). Species examined: Trigoniocardia antillarum (d’Orbigny). Tridacna, type species Tridacna gigas (Linneaus). Species examined: Tridacna maxima (Roding). Hypanis, type species Hypanis plicatum (Eichwald). Species examined: Hypanis colorata (Eichwald). ~ Preliminary phylogenetic analysis of the bivalve family Galeommatidae Ridiger Bieler' and Paula M. Mikkelsen? ‘Department of Zoology, Field Museum of Natural History, Roosevelt Road at Lake Shore Drive, Chicago, Illinois 60605, U.S.A. 2Harbor Branch Oceanographic Museum, Harbor Branch Oceanographic Institution, 5600 Old Dixie Highway, Ft. Pierce, Florida 34946, U.S.A. and Department of Biological Sciences, Florida Institute of Technology, Melbourne, Florida 32901, U.S.A. Abstract. A preliminary phylogenetic analysis of species assigned to the ill-defined family Galeommatidae, plus selected others, was attempted in an effort to clarify the relative value of various systematic characters. Aspects of dealing with large numbers of equally parsimonious trees in cladistic analyses and intrinsic problems of analyses based on unordered multistate characters are addressed briefly. The analysis (Hennig86, implicit enumeration) of 18 characters with 46 character states for 20 species yielded 164 equally parsimonious trees (length 52, ci 53, ri 72), displaying five distinctly different branching patterns. Separate consensus trees were produced for the five groups. Of the five topologies, one is considered most likely in an evolutionary context, and is discussed in detail. Three consistent species groups were recognized: 1) Divariscintilla group (eight species); 2) Galeomma-Ephippodonta group (six species); and 3) Scintillona-Ceratobornia group (three species). Two specialized anatomical characters were analyzed for relative systematic value: 1) ‘“‘hanging’’ foot morphology, which could have evolved more than once within the Galeommatoidea; 2) flower-like organs, a possible synapomorphy of the Divariscintilla species-group. Results also indicated that the monospecific Phlyctaenachlamys Popham, 1939, is a junior subjective synonym of Divariscintilla Powell, 1932, and suggested that the generic limits of Galeomma Sowerby In: Turton, 1825, and Ephippodonta Tate, 1889, should be reexamined. Data matrix construction further iden- tified potentially valuable, but currently unusable, characters in need of further investigation: occurrence of flower-like organs; foot morphology including byssus gland(s); homologies of hinge teeth; ligament/resilium apparatus; shell microstructure; reduction/loss of ctenidial interlamellar junctions and outer demibranch; presence/extent of midgut typhlosole, innervation of pallial tentacles, and sperm structure. ‘‘The interchanges of characters and the multiplicity of forms in mantis shrimp burrows, in New Zealand. separated by apparently trifling details of structure make this Anatomical description of the five Floridian species group one of the most perplexing I have ever tried to review.’ has resulted in redescription of the genus (Mikkelsen and — William H. Dall, 1899:875. Bieler, 1989), incorporating an interesting suite of complex Galeommatoidean bivalves are recognized by a suite characters reflective of the clams’ specialized nature and of character specializations, character reductions and possi- habitat. Two of these characters (“‘hanging’’ foot morphology, ble cases of convergence (Boss, 1965; Morton and Scott, flower-like organs, discussed below), although diagnostic of 1989). Common traits include a muscular foot modified for the genus, are also known from species assigned to other snail-like locomotion, a byssus gland present in the adult, genera, sometimes placed in other nominal families, of species-specific arrangements of sensory papillae and Galeommatoidea. tentacles, anterior-to-posterior water flow through the man- ‘*Hanging’’ foot morphology involves a bipartite foot tle cavity, and eulamellibranch ctenidia. Trends within the with a muscular anterior crawling portion, and an elastic group are toward shell reduction, with corresponding reduc- posterior extension. A ciliated ventral groove extends from tions in sculpture, hinge structure and adductor muscles, in- the primary byssus gland in the antero-ventral part to the ternalization of the shell by mantle lobes, reduction of the terminal, internally-lamellar byssus adhesive gland (see Mik- outer demibranch, commensalism, and reproductive special- kelsen and Bieler, 1989). This morphology is present in ization, including hermaphroditism, brooding, dwarf males members of the five Floridian Divariscintilla species (Mik- (B. Morton, 1976, 1981; O Foighil, 1985a), spermatophores kelsen and Bieler, 1989: figs. 18, 19, 21, 22; in press), D. (O Foighil, 1985a, b), and mating behavior (Mikkelsen and maoria (see Judd, 1971: 351-352, figs. 1-4; pers. obs.), and Bieler, in press). Phlyctaenachlamys lysiosquillina Popham, 1939 (: 64, figs. Five species in the galeommatid genus Divariscintilla 1, 7). It is also known from three members of Lasaeidae (= have been described recently from eastern Florida (Mikkelsen Erycinidae): Parabornia squillina Boss, 1965 (: 4, fig. 3; pers. and Bieler, 1989, in press). All five are co-occurring com- obs.); Ceratobornia longipes (Stimpson, 1855) (Dall, 1899: mensals with a single species of burrowing mantis shrimp 889, pl. 88, figs. 10, Il, 13); C. cema Narchi, 1966 (: 515, [Lysiosquilla scabricauda (Lamarck)] that inhabits shallow- Nes; 2. 3). water sand flats. The only other known member of the genus Flower-like organs are located on the anterior surface is the type species, D. maoria Powell, 1932, also commensal of the visceral mass, just ventral to the labial palps. They American Malacological Bulletin, Vol. 9(2) (1992):157-164 157 158 AMER. MALAC. BULL. 9(2) (1992) are mushroom-shaped, without major nervous supply, with a ‘‘head’’ composed of onion-shaped secretory units empty- ing into the anterior pallial cavity (Mikkelsen and Bieler, 1989: fig. 23). Their number is species-specific and cons- tant, except in Divariscintilla yoyo which has from three to seven organs in a close cluster. Their function is undeter- mined, but is probably pheromonal, related to intraspecific communication with potential reproductive partners or with free-swimming veliger larvae (Mikkelsen and Bieler, 1989, in press). Flower-like organs are present in members of four of the five Floridian Divariscintilla species (Mikkelsen and Bieler, 1989: figs. 26, 27; in press), D. maoria (see Judd, 1971: 352, figs. 2, 4, PP; pers. obs.), Phlyctaenachlamys lysiosquillina (two in number, pers. obs.), Vasconiella jef- freysiana (P. Fischer, 1873) (see Cornet, 1982: fig. 5), and the lasaeid Parabornia squillina (single, pers. obs.) The family Galeommatidae is ill-defined at present. It is differentiated traditionally from other Galeommatoidea mainly by hinge teeth, generally described as irregular, edentulous, or with weak tubercular cardinals (e.g. Chavan In: Moore, 1969a; Kay, 1979). This reflects general reduc- tion in the hinge, rather than any defined synapomorphy for the group. In this context, the systematic value of the two specialized characters relative to traditionally employed characters (e.g. hinge teeth, ctenidia) is of interest. In response, a phylogenetic analysis of species assigned to Galeommatidae, plus selected others, was attempted. Characters involving the shell, mantle, mantle cavity, reproduction, and ecology were included, and type species of genera were used whenever possible. The analysis is con- sidered preliminary, in the sense that it is partially based on limited literature data, and as such does not propose to resolve phylogenetic relationships for the family. It has, however, shown the distribution of foot- and flower-like-organ- characters within the group, and perhaps more importantly, identified characters which could not be used at this time due to insufficient data. MATERIALS AND METHODS TAXA Ingroup: Species were selected from as many galeom- matid genera as possible, but dependent on those with ade- quate, available anatomical data. Type species were includ- ed whenever possible. Twenty species [18 in Galeommatidae, two in Lasaeidae (= Erycinidae)] were included as members of the ingroup (Appendix 1). Data for most non-Floridian species were based on published literature, but in several cases were supplemented or verified by original observations of specimens (Appendix 1). Outgroup: Out initial attempts used either or both of the galeommatoidean (but presumably non-galeommatid) species Montacuta substriata (Montagu, 1808) and Lasaea rubra (Montagu, 1803) as outgroups. This was abandoned when it became clear that these taxa, with their own special- izations (and unresolved taxonomic questions) introduced additional homoplasy complicating the efforts to estimate in- group relationships. Instead, a hypothetical bivalve (all character states = 0) was used as the outgroup. This bivalve, as defined by our character set, approximates closely a generalized member of Lucinoidea, e.g. Lucina (see Chavan In: Moore, 1969b), except in morphology of the foot which in lucinid species is specialized for burrowing (Yonge and Thompson, 1976). Abbreviations of repositories are as follows: AMS, Australian Museum, Sydney; BMNH, Natural History Museum, London; CAS, California Academy of Sciences, San Francisco; FSBC, Florida Marine Research Institute, Department of Natural Resources, St. Petersburg; NMP, Natal Museum, Pietermaritzburg, South Africa. CHARACTERS Characters used in the analysis were dependent upon those features that could be adequately coded from statements in the literature. Particular attention was paid to those characters which have been employed in genus- and family- level descriptions, e.g. hinge structure. Additional characters were evaluated but could not be used (see Discussion). Eighteen characters were used, involving the shell (ten), anatomy (six), reproduction (one), and life habit (one) (Appendix 2). Commensalism was superimposed upon the completed trees to visualize the taxonomic distribution of commensal versus non-commensal species. Hinge characters comprised five of the ten shell characters used in this analysis, yet they proved exceedingly difficult to code with regard to the nature and numbers of teeth present (see Discussion). In the absence of data con- cerning tooth homologies, we coded functional presence/ absence states for both cardinal and lateral teeth (characters 5 and 6), i.e. present and interlocking, or present and noninterlocking, or absent. The location of both lateral teeth (character 7), either anterior, posterior or both, was also coded. Thickened ridges along the hinge line (character 8) in several species could not be interpreted as modified lateral teeth with confidence, so they were coded separately as present/absent. The extent of mantle coverage over the shell (character 10) could not be determined reliably from preserved material nor from published reports based on preserved material alone. Therefore, one question-mark (for Vasconiella jeffreysiana) exists in the coding of this character in the final data set. Inter- preting the degree of internalization was also a problem. Some species descriptions indicated complete internalization (e.g. Coleoconcha opalina, see Barnard, 1964). However in two species with such descriptions (Chlamydoconcha orcutti, see B. Morton, 1981; Phlyctaenachlamys lysiosquillina, see BIELER AND MIKKELSEN: PHYLOGENETIC ANALYSIS OF GALEOMMATIDAE 159 Popham, 1939), an umbonal foramen actually exists connec- ting the external environment and the cavity containing the shell (B. Morton, 1981, and pers. obs., respectively). The mantle tissue with this type of opening is not retractable, even upon preservation. These two cases show a higher degree of mantle fusion than that seen in, e.g. Divariscintilla troglodytes, where retraction exposes more than half of the shell. However, in view of the difficulties experienced with C. orcutti and P. lysiosquillina, all cases of mantle fusion preventing complete retraction, regardless of the degree, were coded identically. The character ‘‘dymantic tentacles’ refers to two single dorsal tentacles, one anterior and one posterior, which are used in dymantic, or defensive, display (see B. Morton, 1975). They were coded as present when present morphologically, even when dymantic behavior had not been documented (e.g. for Galeomma turtoni, see Popham, 1940). They were cod- ed separately from other, non-dymantic tentacles, which usually exist in lateral pairs. ANALYSIS The final data matrix appears in Appendix 3. The Hennig86 program package (version 1.5; Farris, 1988) was used for this analysis on a 486-class IBM-compatible per- sonal computer. Tree generation utilized ‘‘implicit enumera- tion’ (ie), an algorithm that guarantees finding all shortest equally parsimonious trees. The terms consistency index (ci) and retention index (ri) are employed as defined by Kluge and Farris (1969) and Farris (1989), respectively. No a priori assumptions were made _ regarding * o o eo o character importance (weighting) or evolutionary direction (ordering of multistate characters). The use of unordered character states avoids bias in tree development. However, the algorithm can find it more parsimonious to interpret the (initially presumed plesiomorphic) state of the outgroup as autapomorphic and the initially presumed synapomorphic state of the ingroup as a symplesiomorphy, shared with the hypothetical ancestor of both ingroup and outgroup. It is therefore necessary to scrutinize every resulting tree for this occurrence [e.g. as discussed further below, all resulting trees assumed some extent of mantle coverage (character 10) for the hypothetical ancestor]. RESULTS Based on the rigorous ie-algorithm, 164 equally parsi- monious trees (length 52, ci 53, ri 72) resulted from the analysis. Each tree was analyzed and most (96%) could be assigned to one of five distinct tree topologies. The remain- ing trees (4%) were combinations of the five scenarios. Four of the five topologies (56% of the trees generated) are here considered less likely in an evolutionary context because they are based on assumptions such as complete shell coverage by the mantle in the hypothetical ancestor (character 10 state 3), reversal from lost to interlocking lateral hinge teeth (character 6), a flattened limpet-like ancestor (character 17), or a large number of character state reversals (as opposed to parallel acquisition). The major species groups (discussed below) in these trees were recognizable but often as grades rather than clades. * * * KK K KK * * * Cc Co Cho Em Eo Gtu Gta Ps Do Dm Dc ODI Ot Dy Pl Fig. 1. Nelson (strict) consensus tree of the one of five tree topologies that is here considered most likely in an evolutionary context (based on 44% of all trees generated) [Length = 52, consistency index (ci) = 53, retention index (ri) = 72, * = species known to live commensally with another invertebrate species, OG = hypothetical bivalve outgroup (for other acronyms see Appendix 1)]. 160 AMER. MALAC. BULL. 9(2) (1992) The Nelson (strict) consensus tree for the fifth topology, here considered most likely in an evolutionary con- text (within the limits of the current dataset), is presented in figure 1. Three monophyletic groups were distinct: 1) Divariscintilla group (eight species), including Phlyctaena- chlamys lysiosquillina (Pl), with Parabornia squillina (Ps) forming the sister group to Divariscintilla; 2) Galeomma- Ephippodonta group (six species), including Coleoconcha opalina (Co) and Chlamydoconcha orcutti (Cho); 3) Scintillona-Ceratobornia group (three species), comprising Scintillona zelandica (Sz), S. bellerophon (Sb), and Ceratobornia cema (Cc). Scintilla stevensoni (Ss), S. violescens (Sv), and Vasconiella jeffreysiana (Vj) were not affiliated clearly with any of the three groups, but the two Scintilla species (Ss, Sv) appear to be not monophyletic. The relationships between the three major branches remain unresolved. Within the Divariscintilla-group, several stable subunits can be recognized. Based on the common character state of anterior shell prolongation (character 2), Dy, Dt and Pl group together. The taxa Do and Dm are linked because of their low degree of shell coverage by the mantle (character 10). Five species (Dc, DI, Dt, Dy and PI) fall together because of shell reduction (character 0). Parabornia squillina (Ps) joins this clade because of the elongated foot (character 4), absence of lateral hinge teeth (character 7) and presence of a flower-like organ (character 13). The species of the Galeomma-Ephippodonta-group always group together based on the synapomorphies of lateral hinge ridges (character 8), beaded shell sculpture (character 9) and dimantic tentacles (character 12), but with equivocal distinction between the two nominal genera. Two additional taxa (Co and Cho) join the clade based on overall limpet- shape (character 17; Co only), presence of dwarf males (character 16), and a number of character state losses (e.g. cardinal and lateral hinge teeth, characters 5 and 6). The clade of the Scintillona-Ceratobornia-group is determined by the indented hinge plate (character 4), non- interlocking cardinal hinge teeth (character 5), the low degree of shell coverage by the mantle (character 10), and by the specialized foot morphology (character 14, Scintillona spp. only). Two characters in the analysis were interpreted con- sistently as autapomorphies and not synapomorphies: inequivalve shells and ventral shell notch (characters | and 3). The state changes of two other characters (7 and 15; lateral teeth position and adductor muscles) could not positively be placed and could have happened in any of several branches of the trees. ‘‘Hanging’” foot morphology (character 14) was not confined to a single group, and may therefore be convergent. It does appear in most members of the Divariscintilla group, but also in Ceratobornia cema (Cc). Flower-like organs were primarily confined to members of the Divariscintilla group, but are also present in Vasconiella (Vj), a taxon of uncertain affiliation at this point. Commensal species (Fig. 1, *) are distributed widely on all trees, confirming a trend toward commensalism in the superfamily, but not defining any tax- onomic group. DISCUSSION This analysis, that could not employ a number of recognizably valuable characters (see below), and that resulted in such a high number of equally parsimonious trees, is ob- viously a preliminary one. However, without over-interpreting the results, five generalizations can be made. 1) The described Divariscintilla species and monotypic Phlyctaenachlamys ap- pear to form a monophyletic group. If treated as one genus, Divariscintilla Powell, 1932, has priority over Phlyctaena- chlamys Popham, 1939. 2) The generic allocation of species in Galeomma versus Ephippodonta needs further study. 3) The postulated close relationship (same genus) of Scintilla stevensoni and S. violescens is questionable. 4) The relation- ships between Ceratobornia cema and the two studied Scintillona species also warrant further investigation. 5) The relationship between Parabornia squillina and Divariscintilla species, usually placed in different nominal families, needs additional study. Improvement on the data set using more specimen- (versus literature-) based data is of course indicated before strong taxonomic decisions can be made as a result of phylogenetic analysis of this group. Character states are too often ambiguous in written descriptions, even more so if in- terpreted from line drawings or photographs. Furthermore, an appreciation of variability of characters within a species (see below) is seldom available in the literature. We assume that the complex ‘“‘hanging’’ foot mor- phology (with associated glandular structures) and flower- like organs represent synapomorphies within Galeom- matoidea, but the extent of the groups defined by them presently remains unclear. Perhaps the most valuable result of this study was the identification of numerous characters that, although poten- tially valuable, could not be used in the analysis due to lack of data on species not studied by us. These are discussed below as suggestions for needed comparative investigations and/or as useful items to include in future species de- scriptions. 1. The occurrence of both specialized characters em- phasized in this study (“‘hanging’’ foot morphology, flower- like organs) requires additional documentation. Examination of additional Phlyctaenachlamys lysiosquillina specimens dur- ing this study suggested that flower-like organs may vary within a species. These organs were not mentioned or in- dicated in drawings of anatomy or histological sections in BIELER AND MIKKELSEN: PHYLOGENETIC ANALYSIS OF GALEOMMATIDAE 161 the excellent original description by Popham (1939). Nor were they present in one of Popham’s paratypes (BMNH 1939.5.10.2) examined by us. Therefore, we trust that flower- like organs were not merely overlooked by Popham. Never- theless, each of three specimens from the Australian Museum (C165143) had two well-defined flower-like organs. Whether this reflects populational, reproductive or seasonal variation is unknown. In addition to flower-like organs, two other features of Phlyctaenachlamys lysiosquillina were clarified during ex- amination of the Australian Museum specimens, and are worthy of mention here: 1) the existence of an umbonal foramen (see above); 2) the presence of ctenidial interlamellar junctions (in inner demibranch, approximately mid-gill) which had been stated as absent by Popham (1939: 72). 2. Hanging foot morphology is complicated. Oldfield (1955, 1961) showed multiple byssus glands in various galeom- matoideans which compliment our findings in Divariscintilla. In Lasaea rubra, a ‘‘subsidiary byssus gland’’ empties into the canal of the main byssus gland with its associated byssogenous lamellae in the posterior “‘heel’’ of the foot (Oldfield, 1955: 233-234, fig. 4, BGl, BG2, BL). In Monta- cuta substriata, several subsidiary glands ‘‘open by long, slender ducts, into the extreme anterior end of the byssus [ventral] groove’’ (Oldfield, 1961: 270, fig.7, BG1-3, BL), mir- roring the condition seen in Floridian Divariscintilla species. In light of these data, as mentioned previously (Mikkelsen and Bieler, 1989), the anteroventral “‘mucous gland’’ in the two-part foot of Ceratobornia cema should be reevaluated. Behavioral observations and histochemical techniques would be valuable in this area. 3. As was implied above, galeommatoidean hinge teeth are difficult to interpret. Problems such as small subumbonal tubercles which may or may not interlock, or lateral ridges which may or may not be true lateral teeth, are not uncom- mon. Unresolved questions of this kind involving tooth homology prevented rigorous coding of the hinge teeth, and will require ontogenetic studies to resolve with certainty. Also concerning the hinge, the external ligament and resilium have been described in a variety of ways, e.g. with or without nymph/resilifer/socket, triangular or oblique, subumbonal between teeth or posterior (for examples, see Chavan In: Moore, 1969a). Whether these are real differences or mere variation in wording must await reanalysis of hinge structures. 4. Coney (1990) and ourselves (Mikkelsen and Bieler, 1989) have illustrated and described shell microstructure in several galeommatoidean species. Additional species should be investigated using more consistent and rigorous methods. 5. Second only to the hinge in traditional taxonomic use in Galeommatoidea are the ctenidia. Two potentially useful characters could perhaps be quantified for use in an analysis such as this. 1) Relative size of the outer demibranch has been subjectively recorded but with implied quantifiable differences, e.g. ‘‘much shorter’ (Scintillona zelandica; J. E. Morton, 1957: 185), ‘slightly [reduced]’’ (Ephippodonta macdougalli; Woodward, 1893), ‘‘longer, dorso-ventrally than the inner’’ (Chlamydoconcha orcutti; B. Morton, 1981). Com- plete loss of the outer demibranch is characteristic of Montacutidae. 2) Reduction in number or complete loss of interlamellar junctions have been correlated with extensive expansion/contraction of the mantle (Popham, 1939; Narchi, 1966), or with the incubation of larvae in the suprabranchial chamber (B. Morton, 1981). Four of the five Floridian Divariscintilla species are known to brood and to have inter- lamellar junctions. The presence of interlamellar junctions in the type species, D. maoria, was also confirmed during this study (from Australian Museum specimens, collected by W. Judd, C165142). 6. The presence and extent of a midgut typhlosole ap- pear to differ among species. For example, it has been reported as absent in Ceratobornia cema (see Narchi, 1966), Phlyctaenachlamys lysiosquillina (see Popham, 1939), and Montacuta spp. (Oldfield, 1961), and as present in Chlamydoconcha orcutti (see B. Morton, 1981), Divariscin- tilla spp. (Mikkelsen and Bieler, 1989, in press), and Galeomma takii (see B. Morton, 1973). 7. Innervation of the various pallial tentacles could reveal patterns and possible homologies. Many species (e.g. Ceratobornia cema, Scintilla violescens, Galeomma turtoni) show unpaired tentacles along the dorsal midline; in Divariscintilla yoyo, these are known to be innervated by branches from both pallial nerves (Mikkelsen and Bieler, 1989: fig. 31). The innervation of ‘‘dymantic’’ tentacles (B. Morton, 1975, 1976) in Galeomma and Ephippodonta species is especially important to this line of inquiry. 8. Finally, in this group where reproductive complex- ity is the rule, sperm structure could be a conservative, valuable indicator of phylogenetic relationships. Reported morphologies include elongated curved heads with collared acrosomes [Divariscintilla spp., Eckelbarger et al., 1990, Lasaea australis (Lamarck, 1818), O Foighil, 1988], elongated straight heads with cone-shaped acrosomes [Mysella tumida (Carpenter, 1864), O Foighil, 1985b], and oval heads (Chlamydoconcha orcutti, B. Morton, 1981). ACKNOWLEDGMENTS We thank Terrence Gosliner (CAS), Richard N. Kilburn (NMP), William G. Lyons (FSBC), Winston F. Ponder (AMS), and Kathie Way (BMNH) for the loan of specimens. Walter Sage (American Museum of Natural History, New York) and Paul Scott (Santa Barbara Museum of Natural History, California) provided needed literature. Two anonymous reviewers are thanked for their comments. The research was supported in part by the Smithsonian Marine Station at Link Port; the cooperation of Dr. Mary Rice and her staff is gratefully acknowledged. This is Harbor Branch Oceanographic Institution Contribution No. 878 and Smithsonian Marine Station Contribution No. 290. 162 AMER. MALAC. LITERATURE CITED Angloy, M. 1988. Galeomma turtoni Sowerby in Turton, 1825. La Conchiglia 20(236-237):12-13. Arakawa, K. Y. 1961. A note on the animal of Scintilla violescens collected in Genkai Sea. Venus, Japanese Journal of Malacology 21(2):143-146. Barnard, K. H. 1964. Two new genera of Erycinacea (Bivalvia) from South Africa. Proceedings of the Malacological Society of London 36(1):33-37. Bernard, F. 1897. Anatomie de Chlamydoconcha orcutti Dall, lamellibranche a coquille interne. Annales des Sciences Naturelles, Zoologie et Paléontologie 4:221-252. Boss, K. J. 1965. A new mollusk (Bivalvia, Erycinidae) commensal on the stomatopod crustacean Lysiosquilla. American Museum Noviates No. 2215: 1-11. Chavan, A. 1969a. Superfamily Leptonacea Gray, 1847. In: Treatise on In- vertebrate Paleontology, Part N, Vol. 2, Mollusca 6, Bivalvia, R. C. Moore, ed., pp. N518-N537. Geological Society of America and University of Kansas, Lawrence, Kansas, pp. N491-N952. Chavan, A. 1969b. Superfamily Lucinacea Fleming, 1828. In: Treatise on Invertebrate Paleontology, Part N, Vol. 2, Mollusca 6, Bivalvia, R. C. Moore, ed., pp. N491-N518. Geological Society of America and University of Kansas, Lawrence, Kansas, pp. N491-N952. Coney, C. C. 1990. Bellascintilla parmaleeana new genus and species from tropical eastern Pacific, with a review of the other, ventrally notched galeommatid genera (Bivalvia: Galeommatacea). Nautilus 104(4):130-144. Cornet, M. 1982. Anatomical description of Vasconiella jeffreysiana (P. Fischer, 1873) (Mollusca, Bivalvia, Leptonacea). Journal of Molluscan Studies 48(1):36-43. Dall, W. H. 1899. Synopsis of the Recent and Tertiary Leptonacea of North America and the West Indies. Proceedings of the United States Na- tional Museum 21(1177):873-897, pls. 87-88. Eckelbarger, K. J., R. Bieler, and P. M. Mikkelsen. 1990. Ultrastructure of sperm development and mature sperm morphology in three species of commensal bivalves (Mollusca: Galeommatoidea). Journal of Mor- phology 205:63-75. Farris, J. S. 1988. Hennig86 Reference. Version 1.5 (18 pp. manual distributed with program diskette). Farris, J. S. 1989. The retention index and the rescaled consistency index. Cladistics 5:417-419. Judd, W. 1971. The structure and habits of Divariscintilla maoria Powell (Bivalvia: Galeommatidae). Proceedings of the Malacological Society of London 39:343-354. : Kay, E. A. 1979. Hawaiian Marine Shells. Reef and Shore Fauna of Hawaii, Section 4, Mollusca. Bernice P. Bishop Museum Spec. Publ. 64(4). Bishop Museum Press, Honolulu. 653 pp. Kisch, B. S. 1958. Vasconiella jeffreysiana (P. Fischer). Proceedings of the Malacological Society of London 33(1):21-24, pl. 3. Kluge, A. G. and J. S. Farris. 1969. Quantitative phyletics and the evolution of anurans. Systematic Zoology 18(1):1-32. Kuroda, T. 1945. New Japanese shells (Part 6). Venus, Japanese Journal of Malacology 14:29-42. Kuroda, T. and I. Taki. 1961. On a new species of Scintilla (Galeommatidae) from Japan. Venus, Japanese Journal of Malacology 21(2):141-142. Mikkelsen, P. M. and R. Bieler. 1989. Biology and comparative anatomy of Divariscintilla yoyo and D. troglodytes, two new species of Galeom- matidae (Bivalvia) from stomatopod burrows in eastern Florida. Malacologia 3\(1):1-21. Mikkelsen, P. M. and R. Bieler. In press. Biology and comparative anatomy of three new species of commensal Galeommatidae, with a possible case of mating behavior in bivalves. Malacologia. BULL. 9(2) (1992) Mittre, M. H. 1847. Notice sur l’organisation des Galeomma. Annales des Sciences Naturelles, Zoologie (serié 3) 7:169-181, pl. 5. Morton, B. 1973. The biology and functional morphology of Galeomma (Paralepida) takii (Bivalvia: Leptonacea). Journal of Zoology 169(2):133-150. Morton, B. 1975. Dymantic display in Galeomma polita Deshayes (Bivalvia: Leptonacea). Journal of Conchology 28:365-369. Morton, B. 1976. Secondary brooding of temporary dwarf males in Ephip- podonta (Ephippodontina) oedipus sp. nov. (Bivalvia: Leptonacea). Journal of Conchology 29:31-39. Morton, B. 1981. The biology and functional morphology of Chlamydoconcha orcutti with a discussion on the taxonomic status of the Chlamydo- conchacea (Mollusca: Bivalvia). Journal of Zoology 195:81-121. Morton, B. and P. H. Scott. 1989. The Hong Kong Galeommatacea (Mollusca: Bivalvia) and their hosts, with descriptions of new species. Asian Marine Biology 6:129-160. Morton, J. E. 1957. The habits of Scintillona zelandica (Odhner) 1924 (Lamellibranchia: Galeommatidae). Proceedings of the Malacological Society of London 32:185-188. Narchi, W. 1966. The functional morphology of Ceratobornia cema, new species of the Erycinacea (Mollusca, Eulamellibranchiata). Anais da Academia Brasileira de Ciéncias 38(3-4):513-524. O Foighil, D. 1985a. Form, function, and origin of temporary dwarf males in Pseudopythina rugifera (Carpenter, 1864) (Bivalvia: Galeom- matacea). Veliger 27(3):245-252. O Foighil, D. 1985b. Sperm transfer and storage in the brooding bivalve Mysella tumida. Biological Bulletin 169:602-614. fe) Foighil, D. 1988. Random mating and planktotrophic larval development in the brooding hermaphroditic clam Lasaea australis (Lamarck, 1818). Veliger 31(3/4):214-221. O Foighil, D. and A. Gibson. 1984. The morphology, reproduction and ecology of the commensal bivalve Scintillona bellerophon spec. nov. (Galeommatacea). Veliger 27(1):72-80. Odhner, N. H. 1924. Papers from Dr. Th. Mortensen’s Pacific Expedition 1914-16. XIX. New Zealand Mollusca. Videnskabelige Meddelelser fra Dansk naturhistorisk Forening i Kobenhavn 77:\-90, pls. 1-2. Oldfield, E. 1955. Observations on the anatomy and mode of life of Lasaea rubra (Montagu) and Turtonia minuta (Fabricius). Proceedings of the Malacological Society of London 31:226-249. Oldfield, E. 1961. The functional morphology of Kellia suborbicularis (Mon- tagu), Montacuta ferruginosa (Montagu) and M. substriata (Montagu), (Mollusca, Lamellibranchiata). Proceedings of the Malacological Society of London 34:255-295. Ponder, W. F. 1967. Observations on the living animal and mode of life of some New Zealand erycinacean bivalves. Transactions of the Royal Society of New Zealand, Zoology 10(3):21-32. Popham, M. L. 1939. On Phlyctaenachlamys lysiosquillina gen. and sp. nov., a lamellibranch commensal in the burrows of Lysiosquilla maculata. British Museum (Natural History), Great Barrier Reef Expedition 1928-29, Scientific Reports 6(2):61-84. Popham, M. L. 1940. The mantle cavity of some of the Erycinidae, Mon- tacutidae and Galeommatidae with special reference to the ciliary mechanisms. Journal of the Marine Biological Association of the United Kingdom 24(2):549-587. Powell, A. W. B. 1932. On some New Zealand pelecypods. Proceedings of the Malacological Society of London 20(1):65-72, pl. 6. Powell, A. W. B. 1979. New Zealand Mollusca. Collins, Auckland, xiv + 500 pp., 82 pls. Tate, R. 1889. Descriptions of some new species of marine Mollusca from south Australia and Victoria. Transactions and Proceedings and Report of the Royal Society of South Australia 11:60-66, pl. 11. Tebble, N. 1966. British Bivalve Seashells: a Handbook for Identification. British Museum (Natural History), London, 212 pp., 12 pls. BIELER AND MIKKELSEN: PHYLOGENETIC ANALYSIS OF GALEOMMATIDAE 163 Woodward, M. F. 1893. On the anatomy of Ephippodonta macdougalli, Tate. Proceedings of the Malacological Society of London \(1):20-26, pl. 2. APPENDIX 1. Taxa included in the analysis. Reference(s) are sources of literature or original (pers. obs.) data (T = type species of the genus; M = type species of monotypic genus). Reference(s) GALEOMMATIDAE (18) M Chlamydoconcha orcutti Dall, Bernard, 1897; Chavan In: Moore, 1884 (Cho) 1969a; B. Morton, 1981. M Coleoconcha opalina Barnard, Barnard, 1964; pers. obs. 1964 (Co) (NMP A1747). T Divariscintilla maoria Powell, 1932, 1979; Judd, 1971; Powell, 1932 (Dm) Coney, 1990; pers. obs. (AMS C165142). D. troglodytes Mikkelsen and Bieler, 1989. Mikkelsen and Bieler, 1989 (Dt) D. yoyo Mikkelsen and Bieler, 1989. Mikkelsen and Bieler, 1989 (Dy) D. n. sp. ‘‘heart-shaped’’ Mikkelsen and Bieler, in press. (De) D. n. sp. “‘yellow’’ (Dl) D. n. sp. ‘‘white’’ (Do) Ephippodonta macdougalli Tate, 1889 (Em) Mikkelsen and Bieler, in press. Mikkelsen and Bieler, in press. Tate, 1889; Woodward, 1893; Chavan In: Moore, 1969a; pers. obs. (CAS 077807). E. oedipus B. Morton, 1976. B. Morton, 1976 (Eo) T Galeomma turtoni Turton, 1825 (Gtu) Mittre, 1847; Popham, 1940; Tebble, 1966; Chavan In: Moore, 1969a; B. Morton, 1973; Angloy, 1988; pers. obs. (CAS 41198). G. takii (Kuroda, 1945) (Gta) Kuroda, 1945; B. Morton, 1973; B. Morton and Scott, 1989. Popham, 1939; Coney, 1990; pers. obs. (AMS C165143). M Phlyctaenachlamys lysiosquillina Popham, 1939 (Pl) Scintilla stevensoni Powell, 1932, 1979; Ponder, Powell, 1932 (Ss) 1967. S. violescens Kuroda and Arakawa, 1961; Kuroda and Taki, Taki, 1961 (Sv) 1961; B. Morton, 1976. T Scintillona zelandica Odhner, 1924; J. E. Morton, (Odhner, 1924) (Sz) 1957; Chavan In: Moore, 1969a. S. bellerophon fe) Foighil fe) Foighil and Gibson, 1984. and Gibson, 1984 (Sb) M UMsconiella jeffreysiana (P. Fischer, 1873) (Vj) LASAEIDAE (2) Ceratobornia cema Narchi, Kisch, 1958; Cornet, 1982; Coney, 1990. Narchi, 1966. 1966 (Cc) M Parabornia squillina Boss, Boss, 1965; pers. obs. 1965 (Ps) (FSBC 16943). Yonge, C. M. and T. E. Thompson. 1976. Living Marine Molluscs. Collins, London, 288 pp. Date of manuscript acceptance: 21 October 1991 APPENDIX 2. Characters and character states used in the phylogenetic analysis. SHELL 0. Size relative to mantle. 0 = subequal, | = significantly smaller than mantle. 1. Size of valves relative to each other. 0 = equivalve, 1 = inequivalve. . Prolongation. 0 = equilateral, | = anteriorly, 2 = posteriorly. 3. Ventral notch. 0 = absent, | = notched in one valve, 2 = notched in both valves. 4. Hinge plate. OQ = not indented, 1 = indented. 5. Hinge - cardinal teeth. nN 0 = interlocking, 1 = present, not interlocking, 2 = absent. 6. Hinge - lateral teeth. 0 = interlocking, 1 = present, not interlocking, 2 = absent. 7. Hinge - lateral teeth (position). 0 = anterior + posterior, | = anterior only, 2 = posterior only. 8. Hinge - lateral thickened ridges. 0 = absent, | = present. 9. Sculpture. 0 = not reticulate/beaded, | = reticulate/beaded. ANATOMY 10. Extent of mantle coverage. 0 = none, | = margins only, 2 = complete, but retractable, 3 = complete, with fusion preventing complete retraction. 11. Elongated tentacles (excluding dymantic tentacles). 0 = absent, | = present. 12. Dymantic tentacles. 0 = absent, | = present. 13. Flower-like organs. 0 = absent, | = present. 14. Foot structure. 0 = cylindrical anterior + blunt heel, | = cylindrical anterior + elongated heel, 2 = blade-like anterior, without distinct heel. 15. Adductor muscles. O = subequal, | = posterior reduced, 2 = anterior reduced, = both absent. OTHER 16. Reproductive - dwarf male. 0 = absent, | = present. 17. Life habit. 0 = not flattened limpet-like, | = flattened limpet-like. AMER. MALAC. BULL. 9(2) (1992) 164 APPENDIX 3. Data matrix (? = character state unknown; - character state not applicable; OG = hypothetical bivalve outgroup; for other acronyms see Appendix 1). CHARACTERS Taxa 0000000000000 00 0 0 0 OG Co ee — oe ee = Se ee Dm De DI Do Dt Dy Em Eo Gtu Gta Pl 003 0001 1 1 000 0 00000 0 0 0000 0 0 0 00 0 0 Ss 2? 0 ? Sv 0 0 0 l Sz Sb 0 00 0 0 Vj Ps Cho Cc Seasonal variations in brood size of Lasaea cf. nipponica (Bivalvia: Galeommatoidea) in Hong Kong Brian Morton Department of Zoology and The Swire Marine Laboratory, The University of Hong Kong, Hong Kong Abstract. Monthly samples of Septifer virgatus Weigmann, 1837 (Mytilidae) from an exposed rocky shore in Hong Kong were examined for Lasaea cf. nipponica. Evidence for assigning this name to the Hong Kong species of Lasaea is given. Lasaea cf. nipponica is hermaphroditic and produces two broods of crawl-away juveniles each year, in spring (May) and autumn (October to December). Although brood size is a function of parental size, spring parents brood large numbers (x = 81.5/parent) of larger (~ 320-330 «xm shell length) juveniles; autumn parents brood fewer (x = 24/parent) numbers of smaller (~ 300-320 ym shell length) juveniles. Greater variation in the size of juveniles comprising the spring broods, however, suggests that a seasonal comparison of this factor could be non-significant and requires further study. This study suggests that variations in brood size (and possibly juvenile size) results from seasonal differences in environmental factors acting upon the parents. Brood size (and possibly juvenile size) is therefore likely to vary not just seasonally, but across the spectrum of populations encompassed by the species range. Morton and Scott (1989) reported upon a species of Lasaea from Hong Kong shores and to which they ascribed the name Lasaea rubra (Montagu, 1803). Such a designa- tion for a species with a presumed North Atlantic Ocean distribution (Keen, 1983) could be considered intrepid, ex- cept that Ponder (1971) believes the species to have a much wider distribution, encompassing the Pacific Ocean, than hitherto presumed, possibly because of artificial introduc- tions. Regional workers, e.g. Habe (1977), and researchers of the Chinese marine Mollusca, e.g. Bernard er al. (in press), refer to the Asian species of Lasaea as L. undulata (Gould, 1861). Morton and Scott (1989) pointed out, however, that the Hong Kong species does not possess the prominent, widely spaced, concentric rings that are believed typical of L. un- dulata. O Foighil (1988) has, however, suggested that the presence or absence of heavy concentric rings is a poor character for separating species of Lasaea. The taxonomic status of the Hong Kong species of Lasaea will be discussed. The Hong Kong species of Lasaea, i.e. Lasaea cf. nip- ponica, occurs amongst an ~ | mm wide band of mussels (Septifer virgatus Wiegmann, 1837) colonizing the area of mean high water neap tide (MHWNT) on exposed rocky shores. From September 1989 to August 1991, monthly samples of S. virgatus were inspected for Lasaea. These were, in turn, examined for the presence of brooded juveniles, with a view to: (a) obtaining characters which would aid in the identification of the local species of Lasaea; (b) reveal the pattern of reproduction and, thus, brooding undertaken by the species locally; (c) correlate features of brood and juvenile sizes with adult size to assess the extent of in- traspecific variation in this species. Coincidentally, hydro- logical parameters were investigated so as to provide a measure of the environmental changes the local species of Lasaea experiences and possible correlations with such im- portant activities as gametogenesis and the release of crawl- away juveniles. MATERIALS AND METHODS Each month, between September 1989 and August 1990, samples of Lasaea cf. nipponica were obtained from standard (20 x 20 cm) random samples of Septifer virgatus along the exposed shores at Cape d’Aguilar, Hong Kong Island. The samples were fixed immediately in 5% neutral formalin. Each individual in each sample was measured along its greatest length to the nearest 0.1 mm using a dissecting microscope and eye-piece graticule. Wherever possible, ten individuals from each 0.1 mm size class so obtained were dissected and the ctenidia inspected for brooded juveniles. When found, they were isolated from each ctenidium and counted. Each one was then measured along its greatest length to the nearest | jam using a compound microscope and eye- piece graticule. Numbers of juveniles in the left and right ctenidia were separated, initially, but a subsequent test for differences showed there to be none and the two sets of in- formation were pooled subsequently to give a total figure for the numbers of juveniles/parent. The results obtained will be discussed in relation to the prevailing hydrological climate American Malacological Bulletin, Vol. 9(2) (1992):165-171 165 166 AMER. MALAC. BULL. 9(2) (1992) of Cape d’Aguilar and which was assessed by similarly monthly-obtained water samples that were analyzed in terms of temperature, salinity, pH and dissolved oxygen using standard procedures. TAXONOMY Bernard et al. (in press) record Lasaea undulata Gould, 1861 from the coast of China. This is considered to be the senior synonym for both L. nipponica Keen, 1938 and Kellia minutissima Habe, 1960 (Habe, 1977). Morton and Scott (1989), however, reported upon L. rubra (Montagu, 1803) from Hong Kong, noting the lack of strong concentric rings in local specimens but which are, apparently, characteristic of L. undulata (Keen, 1938). As this study will later show, however, the Hong Kong Lasaea is quite unlike L. rubra in terms of breeding season and brood size. O Foighil (1989) has suggested that the Hong Kong species of Lasaea produces crawl-away juveniles; this study will confirm that observation. O Foighil (1989) further points out that the species of Lasaea found in Kagoshima, Japan, referred to as L. undulata, produces planktotrophic larvae and appears to be identical, in terms of valve morphology, to the similarly planktotrophic Australian species L. australis (Lamarck, 1818). O Foighil (1989) further points out that in Kagoshima, there appears to be two species of Lasaea, one with planktotrophic larvae (L. undulata?) and one with crawl-away juveniles. In Japan, Taiwan and Hong Kong, species of Lasaea also have crawl-away juveniles. Keen (1938) compared her species, L. nipponica, from Japan with two other species, also from Japan, i.e. L. striata Tokanuga, 1906 and L. undulata, and dismissed both as possible conspecifics, the former being far too big (9.5 mm shell length), the latter too heavily ringed concentrically. L. undulata with a maximum shell length of ~ 5 mm is also larger than L. nipponica (2.9 mm shell length). The largest individual of the Hong Kong species of Lasaea ever found is 4.0 mm long, reproductively mature parents falling between 2.1 - 3.6 mm in shell length. Thus, in terms of shell form and size and the degree of con- centric ringing, the Hong Kong species is most like L. nip- ponica. If it is true, moreover, that L. undulata produces planktotrophic larvae, whereas the second widely distributed species in Japan (plus Taiwan) produces crawl-away juveniles, then it seems possible that this species, and the Hong Hong species, is L. nipponica and is distinct from L. undulata despite its synonomy with the former species by recent workers, e.g. Habe (1977). In the absence of good comparative data it has, however, been confused with L. undulata, the lat- ter name eventually coming to gain widest coinage. For the above reasons, I refer here to the Hong Kong species as Lasaea cf. nipponica. Scanning electron micrographs of two paratypes of L. nipponica are illustrated in figures 1A and B (California Academy of Sciences Reg. No. 7227 and 7228, respectively) (Invertebrate Zoology Cat. No. 065920). These individuals were collected from Watanoha, Rikuzen, N.E. Matsusima, Honshu, Japan by S. Nomura. A scanning elec- tron micrograph of an individual of Lasaea cf. nipponica from Cape d’Aguilar, Hong Kong is illustrated in figure IC. They appear to be conspecific. Subsequent information on the reproductive biology of L. cf. nipponica in Hong Kong will, hopefully, provide for a better interpretation of this species and allow separation of it from others. RESULTS Figure 2 shows the numbers of individuals of Lasaea cf. nipponica obtained from the standard Septifer virgatus samples over the course of one year from September 1989 - August 1990. Two peaks are obvious, the first in January 1990, the second in May 1991, when densities were ~ 1750 m?* Fig. 1. A, B. Two paratypes of Lasaea nipponica (Keen, 1938) from Honshu, Japan (CAS Reg. No’s. 7227 and 7228, respectively). C, Lasaea cf. nip- ponica from Cape d’ Aguilar, Hong Kong. MORTON: LASAEA CF. NIPPONICA IN HONG KONG 167 10005 Numbers of individuals /sample Y%Brooding parents 1989 1990 Months Fig. 2. Lasaea cf. nipponica. Numbers of individuals obtained in each monthly sample from Septifer virgatus beds at Cape d’ Aguilar, Hong Kong (stippled areas indicate the % number of brooding parents in each sample). 160 e—e May o----o October - December 120 80 Numbers of juveniles 3/ parent {O) 6 40 = 12.93x- 11.8 =0.545 oe 2.0 2.5 3.0 3.5 Parental shell length (mm) Fig. 3. Lasaea cf. nipponica. The numbers of juveniles/parent, expressed as a function of parental shell length, for both spring (May) and autumn (October - December) broods in Hong Kong. and 2375 m-, respectively. At other times of the year, densi- ties were low, all < 600 m-? and for most of the time < 125 m-*. Overlain on figure | is the incidence of brooding individuals. Peaks in overall numbers in the population thus match peaks in the incidence of brooding. Juveniles were ob- tained from L. cf. nipponica parents in October, November and December 1989 and in May 1990, with the latter month providing the greatest % of brooding individuals, i.e. > 36%. Figure 3 compares the data obtained for parental shell length and numbers of juveniles/parent. As might be expected, brood size in both seasons is a function of parental size; big- ger parents have bigger broods. From October to December, brooding individuals of up to 3.4 mm shell length were ob- tained with a mean of 24 shelled juveniles/parent and a max- imum of 44 juveniles/parent, of 4.4 mm shell length. In May 1990, the picture was different, brooding individuals of up to 3.6 mm shell length were obtained with a mean of 81.5 juveniles/parent, although a maximum number of 151/parent, of 3.3 mm shell length, was recorded. Obtained values for brood sizes between the two periods have been tested for significance using an Analysis of Variance and are significant- ly different (F = 19.835; P < 0.001). Figure 4 compares the data obtained for parental shell length (mm) and juvenile length (um). In May 1990, average juvenile shell length, for each 0.1 mm increase in parental shell length, ranged between 308-335 ~m. From October to December, values ranged between 298 and 323 ym. Juvenile shell length, at both times of the year, did not increase with 350 7 e—e May O---O October - December 340 Juvenile length (um) w 3 y = 30.72 + 0.619x poe | eg 320 5 y = 31.11 + 0.0625x 310 1 50.0251 300 : | I cD pe Lt oe ee eee oe! 3.0 2.0 2.5 3.5 Parental shell length (mm) Fig. 4. Lasaea cf. nipponica. Juvenile shell length, expressed as a function of parental shell length, for both spring (May) and autumn (October - December) broods in Hong Kong (mean + 1 S.D.). 168 AMER. MALAC. BULL. 9(2) (1992) parental shell length, 1.e. bigger parents did not have bigger juveniles. Such data are significantly different (P = 0.05). Figure 5 compares juvenile length (um) and numbers of juveniles/parent. For May 1990, a mean juvenile length of 325 ym was obtained which did not vary significantly with respect to the numbers of juveniles present. From October to December, mean juvenile lengths varied between ~ 300-310 um, with an apparent decline in length with increas- ing numbers. This decline is not, however, statistically signifi- cant although it is true that the two sets of seasonal data are significantly different (P = 0.05). The results obtained from this study suggest that spring parents (May) brood larger numbers (x = 81.5/parent) of big- ger juveniles (~ 320-330 wm); autumn parents (October - December) brood fewer (x = 24/parent), smaller (~ 300-320 pm), juveniles. The large standard deviations in juvenile shell lengths (Figs. 3 and 4), however, cast doubt on the signifi- cant difference obtained for the data between the two seasons. Note, for example in figure 4, the wide standard deviations in juvenile shell lengths obtained for spring parents of 2.6, 2.9, 3.0 and 3.4 mm shell length. This would suggest that in May, a variety of developmental stages were present in the broods whereas from October - December, broods were more uniformly at a similar stage of development. DISCUSSION Species of the cryptic galeommatoidean genus Lasaea, occupy the high intertidal virtually world-wide and have aroused interest because of their smallness (< ~-5 mm), their adaptations to such a high-zoned life (Ballantine and Morton, 1956; Morton, 1956, 1960; Morton et al. , 1957) and their mode of reproduction, i.e. simultaneous herma- hese May o----O October - December ° ° = 0.028 \ ear Juvenile length (um) y = 30.83 - 0.0126 xi r= -0.0947 280 0 40 80 120 160 Numbers of juveniles. parent Fig. 5. Lasaea cf. nipponica. Juvenile shell length, expressed as a function of the numbers of juveniles/parent, for both spring (May) and autumn (October - December) broods in Hong Kong (mean + 1 S.D.). phroditism and self-fertilization, with ctenidial brooding (Oldfield, 1955; O Foighil, 1987; O Foighil and Eernisse, 1988). Seed and O’Connor (1980) followed seasonal changes in population structure of Lasaea rubra on the north-east coast of Ireland and suggested that the population comprised three to four age groups, but with few individuals living beyond their third year. McGrath and O Foighil (1986) also studied the population dynamics of L. rubra in south-east Ireland and showed essentially the same pattern, i.e. the polymodal population picture suggested a life span of 2-3 years, with only a few individuals entering their fourth year. The work of Oldfield (1964), Seed and O’Connor (1980) and McGrath and O Foighil (1986) all suggest that Lasaea rubra in the British Isles reproduces in a single phase over summer, typically from May to October, with maximum numbers of brooded, crawl-away, juveniles occurring from June-July and with peak recruitment occurring in August. Oldfield (1964) studied the life cycle of L. rubra in south- west England and estimated an incubation period of about two months. McGrath and O Foighil (1986) have summar- ized the literature on reproduction in L. rubra and of other species, notably L. subviridis from the north-east Pacific. The latter species broods juveniles year round but with peaks oc- curring at different times of the year, according to location. L. rubra hinemoa (from New Zealand) and L. australis (from Australia) similarly brood juveniles year round (Booth, 1979; Roberts, 1984). Table | summarizes the data on the above species of Lasaea and compares it with information obtained about the species reported upon in this study. The Hong Kong species of Lasaea shows some similarities with other species, e.g. the maximum size of brooding adults (3.6 mm) approximates that recorded for L. subviridis (Glynn, 1965; O Foighil, 1985; Beauchamp, 1986). In other respects, however, notably with regard to the juveniles, the species is distinct, i.e. brood size ranges from means of 24 - 81.5/parent, in autumn and spring, respectively, whereas for other species it is far fewer. Similar- ly, crawl-away juvenile length is approximately half that of the other species hitherto investigated. Moreover, the local species broods juveniles in two phases, i.e. spring and autumn, whereas L. subviridis broods juveniles year round (O Foighil, 1985) and L. rubra, in the north-eastern Atlantic, is a summer brooder (Oldfield, 1955, 1964; Seed and O’Con- nor, 1980; McGrath and O Foighil, 1986). L. rubra does, however, brood over the same time frame as L. cf. nipponica in Hong Kong, i.e. from May - ~ November. Russell and Huelsenbeck (1989) report upon variations in brood size and brood structure in the small venerid bivalve Transenella confusa Gray, 1982, from California. In this species, although broods are present in some individuals throughout the year, brood size was significantly lower in winter although it is also a function of adult length (as shown here for Lasaea cf. nipponica). Moreover, the relative pro- MORTON: LASAEA CF. NIPPONICA IN HONG KONG 169 Table 1. Aspects of the reproductive biology of species of Lasaea (after McGrath and O Foighil, 1986) compared with those identified for L. cf. nipponica from Hong Kong. Species Lasaea rubra L. subviridis L. rubra hinemoa L. cf. nipponica Expression of sexuality hermaphroditic hermaphroditic hermaphroditic hermaphroditic Brooding period summer year round year round two phases (May, October - December) Brood size 6-22 2-39 2-33 x = 24 (autumn) (juveniles/parent) 1-21 4-30 4-71 x = 81.5 (spring) Maximum size of 2.4 3.3-3.5 4.1 3.6 brooding adult (mm) Size of released 500-600 530-650 500-600 280-340 juveniles (um) 600-700 510-640 Oldfield, 1955; Glynn, 1965; Booth, 1979; this study Seed and O’Connor, McGrath and Beauchamp, 1986; 1980 O Foighil, 1986 O Foighil, 1985 portions of developmental stages within the broods of 7? con- fusa, varied seasonally, as is possibly the case with Lasaea cf. nipponica in Hong Kong, suggesting variation due to en- vironmental factors. In Hong Kong, many subtropical species of bivalves show two phases of recruitment, in spring and autumn, e.g. the freshwater species Limnoperna fortunei, Corbicula fluminea and Musculium lacustre (Morton, 1987) as well as the estuarine species, Brachidontes variabilis, Saccostrea cucullata and Gafrarium pectinatum (Morton, 1988, 1990a,b). A similar pattern is emerging for marine intertidal species, e.g. Septifer virgatus (Morton, unpub. data). It is thought possible that an otherwise single peak of reproduction is divided into two, early and late, phases by either high mid- summer temperatures or low salinities as a result of the im- pact upon Hong Kong of the tropical south-east Monsoon (Morton, 1991). In the case of the freshwater species, sum- mer rain-induced flooding can flush away gametes or newly- released juveniles. Such a suggestion needs verification and is being investigated. If it is true, however, then the local species of Lasaea, i.e. Lasaea cf. nipponica, could be com- pared with L. rubra in terms of a summer peak in reproduc- tion being divided into early and late phases by environmental extremes. Sea water temperatures in Hong Kong in May and from October - December are, however, far higher than these reported to be optimal for reproduction in both L. rubra (~ 7°C - 14.5°C) (Seed and O’Connor, 1980; McGrath and fe) Foighil, 1986), to which it has been compared (Morton and Scott, 1989), and L. subviridis (8 - 15°C) (O Foighil, 1985) and range between 26°C (May) and 28°C declining to 19°C between October and December. Salinity, pH and dissolved oxygen levels are less variable during the brooding months than at other times of the year (Table 2). Similarly, with regard to the numbers of brooded juveniles (x = 24 - 81.5/parent) and their maximum size (280-340 um), the local species is different from all other species of Lasaea for which information is available (Table 1). In comparison with all other species, Lasaea cf. nipponica in Hong Kong broods, approximately, twice the number of half the size juveniles. On the basis of earlier described taxonomic evidence and the data set out above, it is, therefore, concluded that the local species of Lasaea is best designated the name, Lasaea cf. nipponica. O Foighil (1986) has shown that Lasaea subviridis can brood juveniles to either the Prodissoconch I or Pro- dissoconch II stage (Carriker and Palmer, 1979; Waller, 1981), the former lacking and the latter possessing commarginal striae. O Foighil suggests that prodissoconch morphology is environmentally regulated such that the former morphology is produced by submerged adults, the latter by intertidal parents. This study similarly suggests that for Lasaea cf. nip- ponica, juvenile size is different for spring and autumn Table 2. The ranges of various sea water parameters at Cape d’Aguilar, Hong Kong, and those prevailing during the periods (May and October - December) when Lasaea cf. nipponica is brooding juveniles. Range May October - December Temperature (°C) 17-30 26 28-> 19 Salinity (°/o9) 26-33 32 31-32 pH 7.7-8.4 8.1 8.3-8.1 Dissolved oxygen (mg/I) 5.6-7.8 6.8 6.0-6.9 Most probable wave heights 2.0 29 (m) (50% exceedance) (After Apps and Chen, 1973) 170 AMER. MALAC. BULL. 9(2) (1992) clutches. This observation, however, due to the great varia- tion in the sizes of juveniles comprising, particularly, the spring broods, requires further verification and such data need to be thus treated with caution. In Hong Kong, however, wave heights are higher in autumn than in spring (Apps and Chen, 1973) (Table 2) although the pattern of tidal fluctuations is similar at these times of the year (Morton and Morton, 1983). Autumn parents are thus more frequently immersed (by higher waves) than spring parents. It is thus possible that brood size (and possibly juvenile size) is related to immer- sion period which, if true, is similar to the situation reported upon by O Foighil (1986) for L. subviridis. Russell and Huelsenbeck (1989) similarly show for Transenella confusa in California that brood size is smaller in winter than at other times of the year. Brooded juveniles of Lasaea do not possess feeding structures (Oldfield, 1964; Beauchamp, 1986; O Foighil, 1986). It is, therefore, likely, since the parental size range is the same in spring and winter, that differential environ- mental factors act upon the Lasaea cf. nipponica parents in spring and summer to limit brood size in winter. This could be because of seasonal differences in percentage time immersed, acting via seasonal variations in wave height (Table 1). It is possible that such factors act upon the physiological environment of the parental mantle cavity, as suggested for Ostrea edulis by Waller (1981) and L. subviridis by O Foighil (1986). The results of this study upon Lasaea cf. nipponica and that of Russell and Huelsenbeck (1989) upon the bur- rowing Transenella confusa suggest that brood size is en- vironmentally regulated. So, apparently, is prodissoconch morphology in L. subviridis. Apart from the obvious implica- tions of the former observations with regard to subsequent recruitment and thus population dynamics, such studies col- lectively suggest that the wide variation in shell form typical of species of Lasaea is phenotypic, accounting for the many taxonomic problems surrounding the representatives of this genus. ACKNOWLEDGMENTS I am grateful to Mr. T. K. Cheung of the Electron Microscope Unit of The University of Hong Kong for SEM assistance. LITERATURE CITED Apps, R. F. and T. Y. Chen. 1973. Sea waves at Waglan Island, Hong Kong. Royal Observatory, Hong Kong. Technical Notes. No. 36. pp. 1-18, figs. 1-35. Ballantine, D. and J. E. Morton. 1956. Filtering, feeding and digestion in the lamellibranch Lasaea rubra. Journal of the Marine Biological Association of the United Kingdom 35:241-74. Beauchamp, K. A. 1986. Reproductive ecology of the brooding, herma- phroditic clam Lasaea subviridis. Marine Biology 93:225-235. Bernard, F. R., Y. Y. Cai and B. Morton. In press. Catalogue of the Marine Bivalve Molluscs of China. Hong Kong University Press, Hong Kong. Booth, J. D. 1979. Common bivalve juveniles from New Zealand Leptonacea. New Zealand Journal of Marine and Freshwater Research 13:241-254. Carriker, M. R. and R. E. Palmer. 1979. Ultrastructural morphogenesis of prodissoconch and early dissoconch valves of the oyster Crassostrea virginica. Proceedings of the National Shellfisheries Association 69:103-128. Glynn, P. W. 1965. Community composition, structure and interrelation- ships in the marine intertidal Endocladia muricata - Balanus glandula association in Monterey Bay, California. Beaufortia 148:1-198. Habe, T. 1977. Systematics of Mollusca in Japan: Bivalvia and Scaphopoda. Hokuryu-kan Publishing Co. Ltd., Tokyo. 372 pp. Keen, A. M. 1938. New pelecypod species of the genera Lasaea and Crassinella. Proceedings of the Malacological Society of London 23:18-32. McGrath, D. and D. O Foighil. 1986. Population dynamics and reproduc- tion of hermaphroditic Lasaea rubra. Ophelia 25:209-219. Morton, B. 1987. Comparative life history tactics and sexual strategies of the fresh and brackish water bivalves of Hong Kong and southern China. American Malacological Bulletin 5:91-101. Morton, B. 1988. The population dynamics and reproductive cycle of Brachidontes_ variabilis (Bivalvia: Mytilidae) in a Hong Kong mangrove. Malacological Review 21:1-9. Morton, B. 1990a. The life cycle and sexual strategy of Gafrarium pectinatum (Bivalvia: Veneridae) in a Hong Kong mangrove. Malacological Review 23:53-62. Morton, B. 1990b. Life cycle and sexual strategy of Saccostrea cucullata from a Hong Kong mangrove. American Malacological Bulletin 8:1-8. Morton, B. 1991. Do the Bivalvia exhibit environment-specific sexual strategies? A Hong Kong model. Journal of Zoology, London 223:131-142. Morton, B. and J. Morton. 1983. The Sea Shore Ecology of Hong Kong. Hong Kong University Press, Hong Kong. 350 pp. Morton, B. and P. H. Scott. 1989. The Hong Kong Galeommatacea (Mollusca: Bivalvia) and their hosts, with descriptions of new species. Asian Marine Biology 6:129-160. Morton, J. E. 1956. The tidal rhythm and action of the digestive system of the lamellibranch Lasaea rubra. Journal of the Marine Biological Association of the United Kingdom 35:563-86. Morton, J. E. 1960. The responses and orientation of the bivalve Lasaea rubra (Montagu). Journal of the Marine Biological Association of the United Kingdom 39:5-26. Morton, J. E., A. D. Boney and E. D. S. Corner. 1957. The adaptations of Lasaea rubra (Mont.), a small intertidal lamellibranch. Journal of the Marine Biological Association of the United Kingdom 36:383-405. fe) Foighil, D. 1985. Fine structure of Lasaea subviridis and Mysella tumida sperm (Bivalvia: Galeommatacea). Zoomorphology 105:125-135. O Foighil, D. 1986. Prodissoconch morphology is environmentally modified in the brooding bivalve Lasaea subviridis. Marine Biology 92:517-524. fe) Foighil, D. 1987. Cytological evidence for self-fertilization in the brooding bivalve, Lasaea subviridis. International Journal of Invertebrate Reproduction and Development (Amsterdam) 12:83-90. ) Foighil, D. 1988. Random mating and planktotrophic juvenile develop- ment in the brooding hermaphroditic clam Lasaea australis (Lamarck, 1818). Veliger 31:214-221. O Foighil, D. 1989. Planktotrophic juvenile development is associated with a restricted geographic range in Lasaea, a genus of brooding, her- maphroditic bivalves. Marine Biology 103:349-358. O Foighil, D. and D. J. Eernisse. 1988. Geographically widespread, non- hybridizing, sympatric strains of the hermaphroditic, brooding clam Lasaea in the northeastern Pacific Ocean. Biological Bulletin MORTON: LASAEA CF. NIPPONICA IN HONG KONG 171 175:218-229. Oldfield, E. 1955. Observations on the anatomy and mode of life of Lasaea rubra (Montagu) and Turtonia minuta (Fabricius). Proceedings of the Malacological Society of London 31:226-47. Oldfield, E. 1964. The reproduction and development of some members of the Erycinidae and the Montacutidae (Mollusca, Eulamellibranchia). Proceedings of the Malacological Society of London 36:79-119. Ponder, W. F. 1971. Some New Zealand and subantarctic bivalves of the Cyamiacea and Leptonacea with descriptions of new taxa. Records of the Dominion Museum, Wellington 7:119-141. Roberts, D. 1984. A comparative study of Lasaea australis, Vulsella spongiarum, Pinna bicolor and Donacilla cuneata (Mollusca: Bivalvia) from Princess Royal Harbour, Western Australia. Journal of Molluscan Studies 50:129-136. Russell, M. P. and J. P. Huelsenbeck. 1989. Seasonal variation in brood structure of Transennella confusa (Bivalvia: Veneridae). Veliger 32:288-295. Seed, R. and R. J. O’Connor. 1980. Shell shape and seasonal changes in population structure in Lasaea rubra (Bivalvia: Erycinidae). Journal of Molluscan Studies 46:66-73. Waller, T. R. 1981. Functional morphology and development of veliger juveniles of the European oyster Ostrea edulis Linne. Smithsonian Contributions to Zoology 328:1-70. Date of manuscript acceptance: 3 December 1991 Reproductive ecology of the Antarctic bivalve Lissarca notorcadensis (Philobryidae) Robert S. Prezant,' Merrill Showers, Ray L. Winstead! and Carol Cleveland? ‘Department of Biology, Indiana University of Pennsylvania, Indiana, Pennsylvania 15705-1090, U. S. A. Invertebrate Zoology Section, Gulf Coast Research Laboratory, Ocean Springs, Mississippi 39564, U. S. A. Abstract. The reproductive ecology of museum deposited specimens of the small philobryid Antarctic bivalve Lissarca notorcadensis was examined. Common to southern polar waters, the philobryids are typically epifaunal in habitat. L. notorcadensis, collected from sites adjacent to the Antarctic Peninsula, are found frequently attached byssally to cidaroid echinoid spines. While 95% of all mature clams display either ova or sperm, there is a quantitative shift in populations examined from male to female dominance with some indications of protrandy. L. notorcadensis are synchronous pallial brooders producing 9-15 shelled offspring regardless of adult female size. Juveniles that have already produced dissoconch shell, apparently crawl from the parental infrabranchial chamber directly onto the same spine as their parent. The newly released juveniles lack fully developed digestive diverticula but are sustained by residual yolk reserves. Possible adaptations for brooding in the adult include an absence of the anterior adductor muscle thereby increasing pallial volume, and filibranch ctenidia that allow compact brood development. We suspect that these iteroparous bivalves release their broods in late austral winter to early summer. A single distinct collection from a site close to South Georgia reflects differences between zoogeographic areas. As environmental threats to the relatively pristine ecosystem of Antarctica grow, it is imperative that we examine its present malacofauna and ecological ‘‘sanctity’’. The philobryid bivalves are a common and widely distributed group of southern polar molluscs (Powell, 1960; Tevesz, 1977) with intriguing reproductive biologies. Primarily thought to be dioecious, the philobryids brood their young in their in- frabranchial chamber well past prodissoconch stages (Prezant, 1990). The circumantarctic philobryid Lissarca notorcadensis Melvill and Standen has been reported to have a small pro- dissoconch I and a large prodissoconch II (Prezant, 1990), a pattern more characteristic of bivalves producing plankto- trophic larvae (Jablonski and Lutz, 1980; 1983). The pro- posed discrepancy in larval shell form could be a retention of a primitive character from a nonbrooding ancestry. Philobryids, while Tertiary in origin (Morton, 1978), have in fact an interesting mix of apomorphic and plesiomorphic characters (outlined in Tevesz, 1977; Morton, 1978). The Philobryidae are known poorly. Tevesz (1977) defines the family as filibranch bivalves with a posteriorly directed, ventral ciliary tract and a ligament pit situated be- tween two rows of denticles (the provinculum). According to Tevesz (1977) there are eight valid, extant genera, dis- tinguishable by variations in their ligament pit, denticles, adult hinge teeth, external shell ornamentation, and adductor muscles. The family is also characterized by nonfused man- tle lobes, lack of siphons, reduction or absence of anterior adductor muscle, and the capacity to brood young to a juvenile stage. There has, to date, been only a single detailed report on the reproductive ecology of any philobryid. Richardson (1979) reported on the reproductive ecology of a shallow water population of Lissarca miliaris (Philippi) from Borge Bay, Signy Island, Antarctica. Other than this, Dell (1964), reported about 400 “‘young shells’? associated with the ctenidia of the Antarctic philobryid Philobrya capillata Dell. An incidental report on phyilobryid reproduction is given in Bernard (1897). L. notorcadensis, while sometimes found as an epibiont on kelp holdfasts, gorgonians, hydroids, sponges, and large solitary ascidians, is frequently found byssally at- tached to cidaroid echinoid species [especially species of the genus Ctenocidaris (J. H. Dearborn, pers. comm.)] (Fig. 1). The present paper details some aspects of the reproductive biology of L. notorcadensis from deeper Antarctic waters. Viviparity in antarctic fauna appears to be a common phenomenom but understanding of this stems from studies of relatively few species (see Pearse et al., 1991). In the very early 1900s, Giard (1905) suggested that antarctic organisms would tend to have a disproportionately large number of brooders that bypassed larval stages. Thorson (1950) proposed that the large number of polar brooders could be attributed to the short austral summers and restricted food supplies. Un- doubtedly, the philobryids are well adapted to brooding but this may not be an adaptation to polar conditions but instead could be a conservative trait reflective of the group’s phylogeny. Possible adaptations of adult female phylobryids for brooding young to late stage juveniles include the absence of an anterior adductor muscle, filibranch ctenidia with a reduced outer demibranch, and a one way flow of pallial water American Malacological Bulletin, Vol. 9(2) (1992):173-186 173 174 AMER. MALAC. BULL. 9(2) (1992) currents originating along the anterior portion of the mantle cavity. The relationship between small adult size and internal brooding is discussed by Strathmann and Strathmann (1982). This paper has a more narrow focus and attempts to explore the overall reproductive ecology of a single, small antarctic Fig. 1. Lissarca notorcadensis attached to sea urchin spines. Various distribu- tions and concentrations are displayed by these three ‘‘typical’’ examples of spines recovered. species with respect to brood size, maturation characteristics, and distribution within habitat. METHODS Specimens of Lissarca notorcadensis were obtained from the Smithsonian Oceanographic Sorting Center of the U. S. National Museum. Voucher specimens have been ac- cessed into the Smithsonian Institution, U. S. National Museum of Natural History (USNM Cat. #860303). Speci- mens were collected during expeditions of the early to mid- 1970s and included the following: Islas Orcadas Cruise 575 from 12 May 1975 using a 10’ Blake Trawl at depths of 132 - 143 m from 53°38.0’S/038°01.8’W; 2 June 1975 using a 5’ Blake Trawl at depths of 161 - 210 m from 56°23.8’S/027° 24.6’W; and 31 May 1975 using a 5’ Blake Trawl at depths of 121 - 228 m from 56°42.8’°S/026°69.7’W; and the HERO Cruise of 3 January 1972 using a Blake Trawl at a depth of 110 m from 64°47.3’S/64°07.4 to 64°06.3’W. Locations are denoted on a station map (Fig. 2). More specific informa- tion on the Islas Orcadas cruise can be found in DeWitt (1976). Most of these cruises were under the supervision of Dr. John H. Dearborn, University of Maine. While only spine material was available to us and positive identification of the specific host proved difficult, it is most likely that the host echinoid was one of three common cidaroids present at these collec- tion sites: Ctenocidaris geliberti, C. perrieri or C. speciosa (J. Dearborn, pers. comm.). Specimens were preserved originally in 10% formalin and transferred to isopropyl (3) South Georgia Island 2) INDIAN OCEAN ATLANTIC OCEAN @ * \ “a Antarctic ‘} Peninsula ¥ ! 90°W PACIFIC OCEAN Fig. 2. Antarctic station map. Exact locations are noted in the text. Site 1, 3 January 1972 collection; Site 2, 31 May and 2 June 1975 collections; Site 3, 12 May 1975 collection. PREZANT ET AL.: LISSARCA REPRODUCTIVE ECOLOGY 175 alcohol within two weeks of collection (B. Burch, pers. comm.). They were transferred to 70% ethanol in February 1986. To obtain specific data on sex ratios, stages of develop- ment, and reproductive and brood capacity, clams from the 3 January 1972, 31 May and 2 June 1975 samples were divid- ed into six size classes based upon maximum height (<1.0mm, 1.0-1.99mm, 2.00-2.99mm, 3.00-3.99mm, 4.00-4.99mm, and >5.00mm). A further breakdown of size classes at the 0.5mm level was also instituted in order to resolve more finely the possibility of an ontogenetic shift in sex ratio. Fifty specimens from each size class had their length, height and breadth measured to the nearest 0.1mm. Each specimen was then opened and examined under a dissec- ting scope to determine sex, number of mature ova, number of nonshelled embryos, and number of shelled juveniles in brood. Maturity of ova was determined by size, mature ova being much larger than immature. Histological subsamples were taken to confirm maturity of ova. There was a clear distinction between shelled and nonshelled embryos with, in our samples, very few intermediates. To determine if there was any segregation of stages or sex along individual echinoid spines, 15 spines from 12 May 1975, and 34 spines from 31 May 1975 were photographed carefully from two angles to expose all attached clams. These two collections were used because of their sampling time proximity. Only spines containing five or more bysally attached clams were examined. Each individual clam was numbered and the numbers keyed into similarly numbered photographs. Each numbered clam was then removed from the spine, measured to the nearest 0.01 mm and examined for sex, mature ova, nonshelled and shelled brood. Each clam was placed into one of three spine sectors determined in two ways. First, the entire spine from base to tip was divided in- to thirds. The distance of each specimen was measured from the base of the spine to the point of byssal attachment on the spine. These clams were then placed in categories of top, mid- dle or bottom third of a total spine. To account for segrega- tion within a population on a spine, clams were also measured in terms of their position within a particular spine popula- tion. Thus, the lowest attached clam represents the position of the lowest measurement; the highest clam on a spine represents the top measurement. With these two demarca- tions, the individual spine populations were again divided into thirds and clams were denoted as being in one of the three intrapopulation sectors. Statistical evidence for our conclusions are given along with the collation of data in the results section. All statistical decisions were made at the ~ = 0.05 level, however the actual probabilities in each case are also given. One way analyses of variance (ANOVA) were conducted to test for possible dif- ferences among the four specified groupings (i.e. male, female with ova, female with nonshelled embryo, female with shelled juveniles) with respect to length, height and breadth of the clams. Tukey’s Test was also conducted as a follow-up specific comparison test to determine pairwise differences among the four groups. Strength of Association measures were also calculated. One-way ANOVAs were also conducted to test for differences among sample dates with respect to the number of nonshelled embryos and number of shelled juveniles in brooding females. Since the data consisted of counts, square- root transformations were used. Combining the data from all spines examined, chi- square tests were done to compare the groupings within populations along the bottom, middle, and top thirds of the entire spine and also the populations demarcated by the up- permost and lowest clam (i.e. within spines). Pairwise specific comparisons were conducted when the initial analysis showed significance. Further chi-square tests were performed to ex- amine the relative abundance of each grouping within a seg- ment of the spine. Chi-square tests and pair-wise specific comparisons were also used to determine if different patterns existed be- tween any specific grouping with respect to the distribution of clams along the bottom, middle and top third of the entire spine (and also within a population on a spine). RESULTS Sexes typically appear separate in Lissarca notor- cadensis although a small percent (less than 5%) of the population examined, especially in the 3.5 - 4.0 mm size range, possess both testes and ovaries. Another indication of protandric hermaphroditism includes proportionally more males than females of smaller size. A compilation of total number of males versus females in the combined 31 May 1975 and 2 June 1975 samples reveals that approximately 73% of the mature population for the size class between 3.00 and 3.99 mm were males (n = 159); between 4.00 and 4.50 mm 55% were males (n = 166); between 4.51 and 4.99 mm 35% were males (n = 101); and over 5.00 mm 39% were males (n = 31) (Fig. 3). These numbers do not change substantially when specimens from 3 January 1972 are incorporated into the calculations (respectively males then compose 66% , 54%, 33%, no specimens in the largest size range were collected on this date). Thus there is an inversion from approximately 70% of the total population being male at sizes under 4.00 mm to about the same percent being female at sizes above 4.50 mm. About equal numbers are diagnosed as male or female in the 4.00 - 4.50 mm range. Recall it is also in this size range where we find most specimens possessing both gamete types. An examination of mean lengths, heights and breadths for each grouping reveals a clear trend (Table 1). For all three size variables measured, the males are the smallest, the 176 AMER. MALAC. BULL. 9(2) (1992) Protandry PERCENT 80,—— 60} ge | a+ aera Neer 4.00-4.50 4.51-4.99 5.00 « TOTAL HEIGHT te) 3.00-3.99 ~~ MALE —+~ FEMALE Fig. 3. Total percent male versus female in incremental size classes. The graph represents cumulative totals in each size class from 31 May and 2 June 1975. The cross-over point also represents the approximate size class that has shown the greatest number of simultaneous hermaphrodites. females with ova are next to the smallest, brooding females with nonshelled embryos are next to the largest and brooding females with shelled juveniles are the largest. Statistically, for all three variables, males were significantly smaller than both brooding female groups, while females with ova were significantly smaller than females with shelled juveniles. Although males were not significantly smaller than females with ova and females with ova were not significantly smaller than females with embryos, the trend remains intact. As would be expected, the correlation between the three measurement variables is strong. For example, the correlation coefficient in a sample of males (n = 110) is 0.95 for length and height. Strength of Association measures for length, height, and breadth respectively are 0.105, 0.125 and 0.113. This means, for example, that 10.5% of the variation observed in the length data can be accounted for by differences in the specific groupings identified, whereas 89.5% of the varia- tion is due to the combination of the infuence of other fac- tors (e.g. environment) on length not accounted for in this analysis. A mature female Lissarca notorcadensis generates relatively few (approximately 25-35) large eggs packed with Table 1. Average (+ S.D.) lengths, heights, and breadths (in mm) of the four groupings in cumulative samples of Lissarca notorcadensis (n = 195) from May, June 1975 and January 1972. Grouping Length Height Breadth Males 3.98 + .57 4.75 + .89 2.64 + .56 Females with ova 4.14 + 50 5.00 + .87 2.81 + .53 Females with embryos 4.43 + 36 5.56 + .60 3.09 + .35 Females with shelled juveniles 4.59 + 40 5.81 + .64 3.27 + .39 yolk. Mature eggs average 140 ym in diameter. There is lit- tle variation in number of ova produced with adult growth as most clams examined containing mature ova, contained the same approximate number. Eggs are most likely fertil- ized in the pallial cavity immediately after release through the paired oviducts. Lissarca notorcadensis broods its young in the in- frabranchial chamber, showing several adaptations that could enhance brooding capabilities. The filibranch ctenidium, with firm but few ciliary junctions, allows spacious volume for development of a brood of few but large embryos and juveniles. Additionally the absence of an anterior adductor produces a spacious pallial cavity for the developing brood. We suspect that the pallial current, flowing from anterior to posterior, also support brood maintenance. There are no in- dications of any material tissue connections to the develop- ing brood. A preshelled embryo and shelled juvenile are the most common brood stages found in our samples (although very rarely we found partially shelled embryos with merely incipient shell formation). The two stages commonly found in our samples are found independently and do not overlap in a single brood. Both stages contain large quantities of yolk droplets. In the shelled juvenile these yolk droplets surround the developing digestive diverticula. TEMPORAL ASPECTS. Males form about 60% of the total population of specimens collected on 31 May and 2 June 1975 but only 40% on 3 January 1972 in geographically closely situated samplings. The sample from near South Georgia had 45% males on 12 May 1975 (Fig. 4). Taking into account the disparate years of sampling, there appears to be a decline in females brooding nonshelled young from January to June. Numbers are quite low for total females with shelled brood in January 1972 but reach a higher plateau in the May-June 1975 samples (Fig. 4). Females with mature ova in January 1972 and 12 May 1975 compose 40% of the total female population rising to over 60% in late May and early June 1975 (Fig. 4). Over 50% of the total female population in January 1972 were collected with early broods. Only 22% of the females on 2 June 1975 contained early broods. In all these data there is the problem of temporal gaps and various sampling sites that can confound the results. Nevertheless, the trends appear in our limited sample set. DISTRIBUTION ON SPINES Size Distribution. Size class distributions of specimens from 12 and 31 May 1975 and 2 January 1972, still attached to spines are displayed on the percentage polygon in figure 5. Clams between 2.1 - 4.0 mm in height dominate collec- tions from 12 May 1975 (N = 208). The collection from 31 May 1975 (N = 414) shows a slight bimodal distribution with peaks in size classes between 1.6 - 3.0 and 3.6 - 5.0 mm. The 177 PREZANT ET AL.: LISSARCA REPRODUCTIVE ECOLOGY PERCENT FEMALE FEM/OVA FEM/B1 FEM/B2 MALE WN JUNE 2, 75 MAY 31, 75 ZZ MBY 12, 75 MM JAN. 3, 72 Fig. 4. Proportion of male versus female clams from the four collection dates. The two groups of bars to the left of the vertical line compare total male and female. To the right of the vertical line, the bars represent the break-down of stages of females only. Thus, for 3 January 1972, about 42% of all females had mature ova. FEM/OVA = females with mature ova; FEM/BI = females with nonshelled brood; FEM/B2 = females with shelled brood. 3 January 1972, N = 85; 12 May 1975, N = 348; 31 May 1975, N = 575; 2 June 1975, N = 187. PERCENT POPULATION 40 3.6-4.0 4.6-5.0 5.6-6.0 6.5 « TOTAL HEIGHT 2.6-3.0 1.6-2.0 —S—~ 3 JAN. 1972 Fig. 5. Length percentage polygon for populations occurring on spines only at three collection dates. The x-axis represents total height in mm; y-axis percent of total population from specific collection date. —+~ 31 MAY 1975 12 MAY 1975 178 AMER. MALAC. BULL. 9(2) (1992) size distribution of the January 1972 collection (N = II15) is similar to that of 12 May 1975 in that 77.5% of the 12 May 1975 population examined falls between 2.6 and 4.0 mm length and 67.9% of the January 1972 collection falls into the same size range. Only 37.6% of the 31 May 1975 population falls into this size range and it is only this population, as well, that has adults forming a substantial component above the 5.1 mm length. Sex and Stage Distribution. For the 49 spines analyzed (Fig. 6), a trend for the entire spine distribution of decreasing numbers of clams from bottom to top third of the spines was detected for all groupings. Indeed, except for im- mature clams, more clams are found on the bottom third of the spines than the upper two thirds combined. In general, few clams were found on the upper third of the spines. Figure 7 diagramatically shows the distribution and significance of all groupings analyzed along the entire spine. A similar representation is seen in figure 8 for within spine population distribution. Even though all groupings exhibited the same general trend of decreasing numbers from bottom to top third, the data were examined further for possible differences between groupings in the relative ‘‘steepness’’ of the overall decreas- TOTAL NUMBER ing pattern. No significant difference was found in the distribution pattern of clams among the bottom, middle and top third of the entire spine between males and total females (p > .18), males and females with ova (p > .87), females with nonshelled embryos and females with shelled juveniles (p > .93), total females and immature clams (p > .05), females with embryos and immature clams (p > .50), females with shelled juveniles and immature clams (p > .66), females with ova and all brooding females (p > .23) and all brooding females and immature clams (p > .39). A significant dif- ference, however, in the pattern was found between males and all brooding females (p < .04), males and immature clams (p < .001) and females with ova and immature clams (p < 03). A significantly greater proportion of males than brooding females was found on the bottom third compared to the top third of the spines (p < .01) and also on the mid- dle third when compared to the top third (p < .05). No dif- ference in distribution pattern was significant for these two groupings when comparing the middle with the bottom third of the spines (p > .49). A significantly greater proportion of males than immatures was found on the bottom third com- pared to the middle (p < .004), on the bottom compared to the top (p < .001) and on the middle compared to the top (p < 03). A significantly greater proportion of females with 200 150 + 100; 50 + - Jt MALE Ml BOTTOM FEMALE FEM/OVA @ZZ MIDDLE FEM/B1 FEM/B2 IMMATURE [|Top Fig. 6. The distribution of clams along total spine sectors. Bottom = bottom third of total spine; Middle = middle third of total spine; Top = top third of total spine. y-axis is total number of clams of specific stage or sex. FEM/OVA = females with mature ova; FEM/B1 = females with nonshelled brood; FEM/B2 = females with shelled brood; Immature = clams, usually less than 3.0 mm in height, with nondifferentiated gametes. 49 spines analyzed from 12 and 31 May 1975. PREZANT ET AL.: LISSARCA REPRODUCTIVE ECOLOGY 179 ova than immatures were found on the bottom third compared to the top (p < .02), however the distribution patterns of these two groupings were not significantly different when compar- ing the bottom with the middle (p > .08) and comparing the middle with the top (p > .13). Examining the bottom third of the spines, the region with the greatest concentration of clams, reveals that there is a progressive decline in total number from male to total female (p < .05) and, within females, from mature females with ova to females brooding nonshelled embryos (p < .05) and shelled juveniles (p < .05) (Fig. 6). Because spines only infrequently have clams attached along their entire length, separate measurements of abundance were taken by dividing the length of spine that did contain clams into thirds. This accounts for what we have termed within population distributions. Figure 9 denotes the cumulative distribution of clams on 49 spines for 12 and 31 Males Females Females with ova Females with nonshelled embryos Brooding females combined Females with shelled juveniles Immatures Fig. 7. A diagramatic representation of clam grouping distributions along entire spine length. Bot = bottom third of spine, Mid = middle third, Top = top third; > = greater than, < = less than, circled symbols are statistical- ly significant. Females Females with shelled juveniles Females with ova Females with nonshelled embryos Brooding females combined Immatures Fig. 8. A diagramatic representation of clam groupings distributed within a given spine population. Symbols and abbreviations as in figure 7. 180 AMER. MALAC. BULL. 9(2) (1992) May 1975. The trend seen on the entire spine of decreasing numbers of clams from bottom to top is not exactly seen in the within spine populations (compare Figs. 7 and 8). Some of the groupings were more evenly distributed within the spine. Figure 10 shows distribution of each grouping on the spines compared against themselves, with the exception of juveniles, again demonstrating the reduction in numbers with ascent on the spine. Possible differences between groupings in distribution patterns of clams among the bottom, middle and top thirds within the spine population were also checked. No signifi- cant differences were found in the distribution patterns be- tween males and females (p > .34), females with ova and all brooding females (p > .64), males and all brooding females (p > .84), females with nonshelled embryos and females with shelled juveniles (p > .24), males and females with ova (p > .23), females with ova and immature clams (p > .15) and females with embryos and immatures (p > .16). A significant difference in pattern was found between males and immatures (p < .05), brooding females and im- matures (p < .05) and females with shelled juveniles and immatures (p < .05). Except for the comparison between males and immatures, all of these significant pairings are dif- ferent than those found to be significant for the entire spine. Similar to the earlier test for the entire spine, a significantly greater proportion of males than immatures was found in the TOTAL NUMBER bottom third compared to the top third (p < .001) however no statistical difference was exhibited between the middle and top (p > .09). A greater proportion of females than im- matures was found in the bottom third compared to the top (p < 01) and more in the middle than the top (p < .05). No significant difference was observed between the bottom and middle (p > .45). A greater proportion of all brooding females than immatures was observed in the bottom third of the population than the top (p < 01) but no difference in pattern was detected for these two groups between the bot- tom and middle (p > .27) and between the middle and the top (p > .14). This same result is exhibited when compar- ing females with juveniles and immatures. Since females with shelled juveniles are a part of the grouping of all brooding females, the previous result could be due to the overriding influence of the females with juveniles in this grouping. In particular, a greater proportion of females with juveniles than immatures was observed in the bottom third than the top (p < .05), however no differences were detected between the bottom and middle (p > .05) and middle and top (p > .77). Brood Distribution. Table 2 shows the average num- ber of nonshelled embryos and shelled juveniles found in broods from each collection. The average number of non- shelled embryos within a brooding female was significantly higher in the 31 May 1975 sample (avg. = 14.4) than in either 140 120-- 100 + ZA MALE MM BOTTOM FEMALE FEM/OVA ZZ, MIDDLE nw \))) i LZ: FEM/B1 FEM/B2 IMMATURE [|Top Fig. 9. Total number of clams in each sector from each sex and stage within population on echinoid spines. 49 spines analyzed from 12 and 31 May 1975. PREZANT ET AL.: LISSARCA REPRODUCTIVE ECOLOGY 181 PERCENT 60 50 + 40 + 30 + 20 + 10 5 MALE FEMALE OVA MH BOTTOM @ZZ MIDDLE EA \ MI Z Li IMMATURE LA B1 B2 [|Top Fig. 10. Percent of total sex and stage in each sector of echinoid spine. Bars for females with ova, nonshelled and shelled brood, represent percent of total female population, not total population. Population as per figure 6. the 2 June 1975 (avg. = 10.9, p < .05) or 12 May 1975 (avg. = 9.5, p < 05) samples. Furthermore, the 3 January 1972 sample (avg. = 12.9) is significantly higher than the 12 May 1975 sample. There are no statistically significant differences between the 31 May 1975 and 3 January 1972 samples, between the 2 June 1975 and 3 January 1973 samples, or between the 2 June 1975 and 12 May 1975 samples. In contrast, when comparing the number of shelled juveniles within a brooding female among the sample dates, no significant differences were found (p > .05). The 3 January 1972 sample was not included in the analysis since it contained only a single adult brooding shelled juvenile. PROPOSED LIFE CYCLES Based on samples from 3 January 1972, 31 May and 3 June 1975, the following life cycle is proposed for Lissarca notorcadensis (see Fig. 11). Fertilization occurs in the man- tle cavity after release of mature eggs in the early austral sum- mer or late spring. During the summer, when females dominate in total numbers over males, many females are brooding early stage, nonshelled embryos. Shell development begins during the late summer and early autumn with simultaneous redevelopment of mature ova. Many females brooding late stage young also possess mature ova. In late autumn and early winter, males dominate females in abun- dance and ova are fully developed. Among females, those with mature ova dominate at a time when males dominate the total population. Shelled juveniles are released in late winter to early summer. Only a single adult was found brooding shelled juveniles in the January 1972 sample. This basic cyclic trend (Fig. 11) is superimposed on populations where each stage can be found at some level in the population year round and is characteristic for stations sampled in January 1972, and 31 May and 2 June 1975. That for the 12 May 1975 sample is slightly offset, possibly re- flecting the proximity to South Georgia. It is possible that the conversion of some portion of the population from males to females also occurs at a specific season, perhaps just after austral summer, accounting for some of the differences in sex and stage noted. Some of the basic features of this life cycle are noted in Table 3 along with a comparison of the same features for L. miliaris. DISCUSSION The Antarctic Ocean benthos, below the 33 meter limit of anchor ice formation, offers a rich and diverse assemblage including close to 900 species of molluscs (Moss, 1988). Most of our knowledge of the benthos reflects research that has occurred over the past three decades. Many collected materials have been deposited in various museums and much remains to be discerned from these stored collections that essentially remain untouched. The Smithsonian Institution Oceanographic Sorting Center represents a rich repository of historical collections of Antarctic molluscs. Many of these collections were ob- tained during expeditions of the early to mid-1970s. The present research examined specimens available from those 182 AMER. MALAC. BULL. 9(2) (1992) Table 2. Average number (+ S.D.) of nonshelled embryos and shelled juveniles retained in brood of females from various sampling times and sites. The collection for 12 May 1975 is adjacent to South Georgia Island. The other collections are closer to the Antarctic Peninsula. Note that the 3 January 1972 sample had only a single female brooding shelled juvenile. Average number Average number of nonshelled of shelled Date N embryos N juveniles 3 Jan 1972 24 12.9 + 4.1 1 10.0 12 May 1975 32 9.5. + 2.7 15 8.9 + 2.5 31 May 1975 35 15.1 + 4.2 27 10.4 + 4.9 2 June 1975 17 10.9 + 3.2 10 9.0 + 3.1 collections and as such was limited to samples outlined in the methodology. Before discussing the results per se, we must ask, can we lend confidence to interpretations based on a rather disparate set of Antarctic collections? The tentative answer is yes, at least until we have a more complete set of data with more closely set samples over a much longer period of time and from more defined benthic sites. Additionally, the clear clues offered in the analyzed data set in conjunc- tion with data from the work of Richardson (1979) on Lissarca miliaris, lends additional credence to our speculation. L. notorcadensis offers a classic example of a polar brooder lean- ing heavily towards the ‘‘k-selection’’ side of an r-K con- tinuum [as designed by Pianka (1970)]. K-selected species vie for maximum competitive ability in a ‘‘saturated’’ environ- ment (as outlined by Southwood, 1977) with slow growth, delayed maturation, iteroparity, low fecundity and large yolky eggs. It is recognized, however, that an examination of liv- ing populations through time is called for to refine our con- cept of reproductive ecology of this philobryid. Berkman et al. (1991) examined recently the plankto- trophic development of the Antarctic scallop Adamussium col- becki (Smith) and suggest that various anatomical features prohibit brooding. Included in these restraints are a filibranch gill without “‘strong tissue connections’, an open mantle cavity without mantle edge connections, and a lack of siphons. Considering these same characters in Lissarca notorcadensis, it is doubtful that these inhibit internal pallial brooding. L. notorcadensis brood their young in their relatively spacious infrabranchial chamber. Increased volume for brooding results from a lack of anterior adductor muscle and a reduced outer ctenidial demibranch. We suggest that flexibility of the fili- branch ctenidial filaments allows the growing brood to essen- tially impinge upon the gills without disruption of function. The ctenidial filaments are relatively short but ‘‘stout’’. Burne (1920; p. 238), in fact, suggests that the gills ‘‘present a somewhat embryonic appearance.’ The anterior flow of pallial water across the gills, as discerned for other philobryids by Tevesz (1977) and Morton (1978), can serve to efficiently aerate the developing young, which presumably are retained in brood for an extended period of time. Large yolky eggs, as found in Lissarca notorcadensis, are typical of direct developing bivalves (Matveva, 1978). Similarly large eggs are found in other brooding philobryids (eg. L. miliaris; Richardson, 1979) as well as several other brooding bivalves including species of Lasaea Beauchamp, 1986; McGrath and O Foighil, 1986), Transennella tantilla (Gould) (Asson-Batres, 1988) and Kellia suborbicularis (Montague). Thorson (1950) suggested that eggs of many polar marine invertebrates are large and have dense concentrations of yolk. This suggests that reproduction in polar habitats is expensive energetically and would require large energy stores (Clarke, 1979). Vance (1973a, b) suggested that it is more ef- ficient to produce fewer but larger eggs in a uniform, stable environment as opposed to a variable or unstable environ- ment. The large and yolky eggs of the philobryids, once fer- tilized, give rise to slowly developing young that are retained in the pallial cavity past prodissoconch stages and into an early dissoconch stage. When released, the young of L. notor- cadensis have a full complement of organs including a near- ly complete digestive system. They retain, however, a quantity of yolk that surrounds the developing digestive diverticula. It is not certain whether the newly released brood actively feed but it is unlikely. Similar stores of apparent nutrients (yolk) in released young are also found in direct developing Lasaea (O Foighil, pers. comm.) and in another Antarctic bivalve, Kidderia subquadratum Pelseneer (Shabica, 1974). In the latter, Shabica (1974:59) also found that ‘‘digestive diverticula of the embryos remain nondifferentiated until ap- proximately two months post-liberation suggesting that no active uptake of particulate nutrients occurs while the em- bryos are developing within the maternal organism or even immediately after liberation.’ Additionally Shabica found that ‘‘large amounts of yolk’’ were widely distributed in the early Table 3. Comparative data for Lissarca notorcadensis and L. miliaris. Data for L. miliaris is adapted from Richardson (1979). Sex size ratio signifies comparative sizes of males and females at various stages. Sex ratio indicates total number of male versus female. 'See text for clarification. L. miliaris L. notorcadensis Depth 5-9 m (to 121 m) 100-400 m (to 1000 m) Maximum height 6.0 mm 7.0 mm Height at maturity 2.5 mm 2.9 mm Brooding height 3.1 mm 3.6 mm Average brood 60 9 Brood range 54-70 6-16 Pre-release size 0.57 mm 1.1 mm Sex size ratio Ll female > male! Preshelled brood Feb. - July Summer (Jan.; some year round) Partial shelled Aug. - July Shelled brood Nov. - Dec. By May Recruitment Dec. - Feb. June - Dec! PREZANT ET AL.: LISSARCA REPRODUCTIVE ECOLOGY 183 male < female JANUARY mostly nonshelled brood shelled shelled brood juveniles initiates released \ AN ova developing male > female J JUNE MAY Fig. 11. Proposed life cycle for Lissarca notorcadensis. ova developed embryo of K. subquadratum but later surrounded develop- ing organs. In fact, as noted by Shabica (1974: iii), yolk was still present in newly released embryos, which can continue development ‘‘...at least for short periods, outside the maternal organism.’’ We have predicted that L. notorcadensis release shelled brood in late winter to early summer. In cases where the brood is released in later winter, these presumed nutrients may well serve to satisfy nutrient requirements prior to the end of the austral winter and up to a time when primary production increases in the Southern Sea. It is also possible that the final stages of development occur post-release as they do in K. subquadratum. On the other hand, Ralph and Everson (1972) speculate that the brooded young of another Antarctic bivalve, K. bicolor (Martens), could remain in the parental mantle cavity for ‘‘their first winter’’ prior to release. This is certainly possible for L. notorcadensis as well and in fact the dissoconch shell deposited prior to release could indicate this extended brood. There is, however, no way of knowing with certainty until monthly samples are available. The brood of Lissarca notorcadensis develops syn- chronously, as in the intertidal Galeommatoidean Lasaea rubra (McGrath and O Foighil, 1986). Prior to brood release, the developing young have all deposited significant dissoconch shell material and young are released as juveniles. Numerous females examined that contained developing broods were also well into a renewed oogenesis cycle (unpub. data). There ap- pears to be a decrease in number of mature ova to early non- shelled brood and finally shelled juveniles. Unlike L. rubra (McGrath and O Foighil, 1986) and L. miliaris (Richardson, 1979), ultimate brood size in L. notorcadensis is very small and is independent of adult size. The smallest as well as the largest specimens averaged 9.4 shelled young. Similarly, there was an average of 12.1 nonshelled young in broods of com- parable sized animals and these in turn were generated from a potential field of at least 25 eggs. The decline is specifical- ly unaccounted for but we assume a loss once eggs are re- leased into the infrabranchial chamber which is essentially open to the environment since the adult mantle edges remain unfused. Decline in the potential number of recruits was noted by Shabica (1974) who discussed ‘‘sloughing-off of eggs’’ in the brooding Antarctic bivalve Kidderia subquadratum. Young of Lissarca notorcadensis are released presumably from the pallial brood chamber directly into their parent’s microhabitat. At time of release the byssal gland is fully functional and young can attach immediately to either the home spine or the shell of another member of the popula- tion. In either case, the young are released into the immediate vicinity of the adult and hence obtain whatever benefits are derived from this microhabitat. An association between bivalve and sea urchin is not unique to this philobryid. Axinodon symmetros (Jeggreys) is, for instance, associated with deep water echinoids (Ockelmann, 1965). The symbiosis for L. notorcadensis appears to be commensalistic with the bivalve being held above the sediment and in the water col- umn where filtration efficiency could be increased due to lack of sedimentary interference. Additionally, the bivalves ob- tain a ready dispersal agent while nestled among protective spines. Dell (1965) suggests that the epifaunal habits of some small Antarctic organisms on motile hosts is related to the ‘lack of motile larval stages’’ (Dell, 1965). The lack of other hard substrata must also be a factor in deeper waters. According to Fretter and Graham (1964), dioecious species of molluscs often have more females than males. They suggest this is related to males dying at a younger age or size. This could be the case in Lissarca notorcadensis but, like several other small brooding bivalves, such as Transenella tantilla (Kabat, 1985; Asson-Batres, 1988), L. notorcadensis could be a protandric hermaphrodite. Evidence for this in- cludes the comparatively larger females found in the popula- tions examined and the cross-over at around 4.0 mm height (probably more than coincidentally also the size range where we located several simultaneous hermaphrodites) from male to female dominated populations. T: tantilla show male dominance up to a size of 3.5 mm with females ranging from 3.5 - 7.0 mm in length (Kabat, 1985). It is possible that all specimens found on a single spine are members of the same family line. This, of course, could limit outbreeding. Pro- tandry would, on the other hand, increase reproductive suc- cess by insuring availability of members of both sexes. Ockelmann and Muus (1978) found that several species of the Montacutidae show reverse geotactic responses just prior to release of their larval brood. Thus, for example, Mysella bidentata (Montagu), a normally infaunal bivalve, becomes negatively geotaxic and will emerge from the sedi- ment and crawl upward upon available vertically oriented ob- jects. Upon reaching the highest point available, the veliger 184 AMER. MALAC. BULL. 9(2) (1992) brood is released. Ockelmann and Muus (1978) suggest this behavior is stimulated by a ‘“‘specific substance released by the young when they become ready to begin life on their own.’ Most specimens of Lissarca notorcadensis, including reproductively active adults, occur on the lower portions of the host spines. There are two likely explanations for this rela- tionship. First, there could be an inherent or external stimulus that attracts males and females to the lower third of the spine. Extrinsically there could be a stimulus released by the sur- face of the urchin that attracts males and mature females or perhaps these specific bivalves are positively geotaxic. On the other hand, there could be an intrinsic stimulus, such as endocrine or horomonal shifts, that are activated when either males or females reach maturity and these chemical shifts stimulate a downward migration. Similarly, one sex or stage could preferentially react to stimuli indicated above and in turn ‘‘attract’’ the other sex through an allellochemic interaction. A second possible explanation is that males prefer- entially remain near the spine’s bottom third and, being pro- tandric hermaphrodites, a conversion of sex would result in a preferential appearance of more mature females along the spine base as well. The positioning of mature males and females along the basal third of the spine would, of course, promote successful fertilization. Brooding females are not preferentially clumped along any region of the spine. This could be indicative that brood are released along the length of the spine as a result of an upward migration of many of the brooding females after fertilization near the spine’s base. There is also an apparent seasonal influence in terms of sex dominance. As noted in the results, there is a transi- tion of dominance from close to 60% male in May and June 1975 to about 40% in January 1975 and 45% on 12 May 1975 at a very different site. By eliminating the 12 May 1975 sam- ple, which is situated close to the north side of South Georgia Islands (see Fig. 2) thus strongly influenced by this land mass, certain trends relegated to an austral summer and spring become clearer (Fig. 12). According to Dell (1972) the South Georgia area represents a distinct biogeographic subdivision of the Antarctic Region. This site is north of the Antarctic convergence and represents a milder climate (see Simpson, 1976). In fact, South Georgia is the northernmost distribu- tion limit for many Antarctic species (Dell, 1965) and the northern side of the island is especially productive (DeWitt et al., 1976). Thus this single site of the four sampled is placed in the South Georgian District (Dell, 1972) and results of a reproductive ecology study might be expected to differ from the other three. Another interesting difference between the two May samples is in size class distribution. The 12 May sample, again from South Georgia, contained no specimens larger than 4.5 mm in length. All other samples had larger bivalves and that from 31 May 1975, in fact, had about 20% of its population longer than 4.5 mm. It is again possible that the less stable and more seasonal South Georgia site could yield bivalves that grow slower. What is more likely is that this population is on the distributional fringe for this species and thus is subject to an earlier demise. The other three sam- ple sites are placed in the Continental Antarctic Division PERCENT 70 60 + 50 + 40 + = 30+ \ ZZ 7 20 + A 7 DA i ~ 10 - MALE FEMALE MR JAN. 3, 72 FEM/OVA AZ MAY 31, 75 FEM/B1 FEM/B2 [| JUNE 2, 75 Fig. 12. Removing the 12 May 1975 (South Georgia Island) sample, more easily shows specific trends outlined in text. Again, female stages to right of vertical line represent percent of total female population only. Population numbers as per figure 4. PREZANT ET AL.: LISSARCA REPRODUCTIVE ECOLOGY 185 (Dell, 1972). The 31 May and 2 June 1975 samples are very similar to each other and clearly different from January 1972. This holds true for all stages sampled (Fig. 4). An additional difference of some note is that of total size class distribution between 12 and 31 May 1975 and 3 January 1972. Specimens sampled on 12 May 1975 show a unimodal distribution with most specimens being between 2.6 and 4.0 mm in height. The January 1972 sample shows a similar distribution. The sample from 31 May 1975, however, has a more uniform distribution with a strong indication of bimodality (see Fig. 5). We have little information on growth rates for any philobryid. Extrapolating from commarginal lines and incremental growth measurements, Richardson (1979) suggested that Lissarca miliaris deposited growth rings slowly, with only one ring deposited in 3 - 15 month old clams. Indeed, Nicol (1966) suggested that, along with small size of Antarctic bivalves, these organisms are slow growing and this is indicated by numerous, narrowly spaced growth rings. Although L. notorcadensis is almost certainly slow growing, with growth rings already deposited in its dissoconch shell prior to release from the brood, it is difficult to extend the measurements suggested by Richardson to this philobryid. A recent review by Bourget ef al. (1991), in fact, casts doubt on many reports that use shell growth marks to determine or substantiate bivalve age. The differences and similarities in population size classes between Lissarca notorcadensis and L. miliaris are perhaps reflective of temporal and environmental differences as discussed elsewhere in this paper. At this point it is uncer- tain if the bimodal distribution represents a failed reproduc- tive event or a natural depression in seasonal reproduction in an iteroparous bivalve reproducing more than once per year. The philobryids are a relatively speciose and common component of the Antarctic malacofauna and yet there is very little information available on their reproductive ecology aside from Richardson’s (1979). Data have been extracted from that paper for comparative purposes and are outlined on Table 2. Most importantly, any comparison drawn between these two species must be done in light of the fact that Richardson used a population of philobryids attached to a brown alga from waters less than 10 m in depth (although this species can reach depths to 121 m). Specimens of L. notorcadensis examined here are from waters no shallower than 110 m (this philobryid has been collected from depths as deep as 1000 m). Ice scouring is an important factor controlling shallow water fauna of Antarctica down to depths of 15 m (Moss, 1988). L. miliaris is restricted to protected embayments of the near sublittoral zone. Additionally, Richardson’s samples of L. miliaris were obtained from within the bounds of the Weddell Drift, a highly productive zone of the Southern Ocean (Moss, 1988). The two philobryids attain approximately the same maximum size although Lissarca miliaris reaches sexual maturity at a slightly smaller size than L. notorcadensis. Additionally, L. notorcadensis is larger when it reaches brooding size. In the more heterogeneous shallow waters in- habited by L. miliaris, it might be predicted that fecundity would be higher. L. miliaris broods 54 to 70 juveniles, directly increasing in number with adult size. L. notorcadensis, on the other hand, broods a consistently smaller number re- gardless of adult size. Size of juveniles at release reflects the differences in brood number. Pre-release size for juveniles of L. miliaris is half that of L. notorcadensis. This of course reflects specifically the volume available for brood in the two species and the number of young able to be packed into the infrabranchial chamber. There appears to be no size difference between males and females of Lissarca miliaris while this varies in L. notor- cadensis. In the latter, females reach a larger size than males, perhaps reflecting protandry as discussed above. It is clear that brooding females in L. notorcadensis are larger than males as well as sexually mature but nonbrooding females. The general life cycles of the two species have some basic similarities in terms of development but there are im- portant temporal differences noted in the populations ex- amined. Preshelled young are evident from February to July in L. miliaris but likely dominate in the austral summer for L. notorcadensis. In the latter species, however, some preshelled brood are evident in all samples examined. Fully shelled juveniles dominate from November to December in L. miliaris and can be found at some level in all our samples of L. notorcadensis but are presumed dominant in May and almost absent in January. Recruitment for L. miliaris and L. notorcadensis 1s from December to February and likely June through December respectively. The latter significant time shift again seems to reflect habitat differences in populations studied. Pearse et al. (1991) review the contrast in the ‘‘physi- cally unstable shallow benthos’’ and the ‘‘physically stable deeper benthos’’. In oligotrophic waters deeper than 20 - 30 m many benthic invertebrates produce pelagic lecithotrophic larvae that yield larger juveniles more resistant to predation pressures. L. notorcadensis extends this to release of brood- ed juveniles that are not only large but protected by the host urchin. As L. notorcadensis is one of the most common molluscs of the antarctic region, the success of this mode of development seems apparent. Grahame and Branch (1985) review different modes of reproduction in marine in- vertebrates and recognize that Antarctica represents an in- triguing case where primary production and potential paucity of food could regulate reproductive strategies. They suggest that more information on these south polar organisms is need- ed to fully appreciate their modes of reproduction. While we are able to discern much information through the use of museum material, to fully appreciate the unique evolutionary game plans developed by these small brooding molluscs, we must promote longer term studies of the living organisms. 186 AMER. MALAC. BULL. 9(2) (1992) ACKNOWLEDGMENTS Specimens for this research were kindly loaned by the Smithsonian Oceanographic Sorting Center. The assistance of G. Hendler, J. Norenberg, M. Fissler and J. Houbrick are particularly noted in this loan. Additional assistance in the quantitative aspects of this research were supplied by K. Snyder, S. Magaro and S. Logan, and we thank them for their efforts. Many thanks to J. H. Dearborn for helpful information on Antarctic cruises and echinoderms. Thanks also to Mr. Doug Shumar for drafting figures 2, 7 and 8. We appreciate the efforts of two anonymous reviewers who helped to clarify the text and refine some of our ideas. This research was funded by a Faculty Research Award to RSP of Indiana University of Pennsylvania. LITERATURE CITED Asson-Batres, M. A. 1988. Reproduction and growth of the brooding bivalve Transennella tantilla. Veliger 30:257-266. Beauchamp, K. A. 1986. Reproductive ecology of the brooding, her- maphroditic clam Lasaea subviridis. Marine Biology 93:225-235. Berkman, P. A., T. R. Waller and S. P. Alexander. 1991. Unprotected larval development in the Antarctic scallop Adamussium colbecki (Mollusca: Bivalvia: Pectinidae). Antarctic Science 3:151-157. Bernard, M. F. 1897. Etudes comparatives sur la coquilles des lamelli- branches. (Part 2) II. Les generes Philobrya et Hochstetteria. Journal de Conchology 45:5-47. Bourget, E., H. Bernard and V. Brock. 1991. Testing hypotheses concern- ing the formation of shell growth marks in marine pelecypods. Canadian Journal of Zoology 69:535-546. Burne, R. H. 1920. Mollusca. Part IV - Anatomy of Pelecypoda. British Antarctic (‘‘Terra Nova’’) Expedition, 1910. Natural History Report; Zoology 11:233-256, IV pls. Clarke, A. 1979. On living in cold water: K- strategies in Antarctic benthos. Marine Biology 55:\1-119. Dell, R. K. 1964. Antarctic and Subantarctic Mollusca: Amphineura, Scaphopoda and Bivalvia. Discovery Report 33:93-250. Dell, R. K. 1965. Marine Biology. /n: Antarctica, T. Hatherton, ed. pp. 129-152. F. A. Praeger, Inc., New York. Dell, R. K. 1972. Notes on nomeclature of some Mollusca from Antarctica and southern South America. Records of the Dominion Museum 8:21-42. DeWitt, H. H. 1976. ARA Islas Orcadas cruise 5. Antarctic Journal of the U.S. 11:48-49. DeWitt, H. H., J. D. McCleave and J. H. Dearborn. 1976. Ecological studies of fishes and echinoderms during ARA Islas Orcada cruise 5. Antarctic Journal of the U.S. 11:49-53. Fretter, V. and A. Graham. 1964. Reproduction. Jn: Physiology of Mollusca, Vol. I, K. M. Wilbur and C. M. Yonge, eds. pp. 127-164. Academic Press, New York. Grahame, J. and G. M. Branch. 1985. Reproductive patterns of marine in- vertebrates. Oceanography and Marine Biology Annual Review 23:373-398. Giard, A. 1905. La Poecilogonie. 6th International Congres Zoologie, Berne. pp. 617-646. Jablonski, D. and R. A. Lutz. 1980. Molluscan larval shell morphology - ecological and paleontological applications. In: Skeletal Growth of Aquatic Organisms. D. C. Rhoads and R. A. Lutz, eds. pp. 323-377. Plenum Press, New York. Kabat, A. R. 1985. The allometry of brooding in Transennella tantilla (Gould) (Mollusca: Bivalvia). Journal of Experimental Marine Biology and Ecology 91:271-279. Matveeva, T. A. 1978. Adaptations to egg bearing in some marine bivalves. Malacological Review 11:116-117. McGrath, D. and D. O Foighil. 1986. Population dynamics and reproduc- tion of hermaphroditic Lasaea rubra (Montagu) (Bivalvia, Galeom- matacea). Ophelia 25:209-219. Morton, B. 1978. The biology and functional morphology of Philobrya munita (Bivalvia: Philobryidae). Journal of Zoology 185:173-196. } Moss, S. 1988. Natural History of the Antarctic Peninsula. Columbia Uni- versity Press, New York. 208 pp. Nicol, D. 1966. Descriptions, ecology, and geographic distribution of some Antarctic pelecypods. Bulletin of American Paleontology 51:1-102. Ockelmann, K. W. 1965. Developmental types in marine bivalves and their distribution along the Atlantic coast of Europe. Jn: Proceedings of the First European Malacological Congress, London, 1962, L. R. Cox and J. F. Peake, eds. pp. 25-35. Conchological Society of Great Britain and Ireland and the Malacological Society of London. Ockelmann, K. W. and K. Muus. 1978. The biology, ecology and behaviour of the bivalve Mysella bidentata (Montagu). Ophelia 17:1-93. OFoighil, D. 1986. Prodissoconch morphology is environmentally modified in the brooding bivalve Lasaea subviridis. Marine Biology 92: 517-524. fe) Foighil, D. 1988. Random mating and planktotrophic larval development in the brooding hermaphroditic clam Lasaea australis (Lamarck, 1818). Veliger 31:214-221. Pearse, J. S., J. B. McClintock and I. Bosch. 1991. Reproduction of antarctic benthic marine invertebrates: tempos, modes, and timing. American Zoologist 31:65-80. Pianka, E. R. 1970. On r- and K- selection. American Naturalist 104:592-597. Powell, A. W. B. 1960. Antarctic and subantarctic Mollusca. Records of the Auckland Institute and Museum 5:117-193. Prezant, R. S. 1990. Microstructural shell and hinge ontogeny of the Antarctic bivalve Lissarca notorcadensis Melvill and Standen. Journal of the Marine Biological Association of the United Kingdom 70:841-855. Ralph, R. and I. Everson. 1972. Some observations on the growth of Kid- deria bicolor (Martens) (Mollusca: Lamellibranchiata) at South Georgia. British Antarctic Survey Bulletin 31:51-54. Richardson, M. G. 1979. The ecology and reproduction of the brooding antarctic bivalve Lissarca miliaris. British Antarctic Survey Bulletin 49:91-115. Shabica, S. V. 1974. Reproductive biology of the brooding antarctic lamellibranch Kidderia subquadratum Pelseneer. Master’s Thesis, Oregon State University. 143 pp. Simpson, R. D. 1976. Physical and biotic factors limiting the distribution and abundance of littoral molluscs on Macquarie Island (sub-antarctic). Journal of Experimental Marine Biology and Ecology 21:1]-49. Strathmann, R. R. and M. F. Strathmann. 1982. The relationship between adult size and brooding in marine invertebrates. American Naturalist 119:91-101. Tevesz, M. J. S. 1977. Taxonomy and ecology of the Philobryidae and Limop- sidae (Mollusca: Pelecypoda). Postilla 171:1-64. Thorson, G. 1950. Reproductive and larval ecology of marine bottom in- vertebrates. Biological Review 25:1-45. Vance, R. R. 1973a. On reproductive strategies in marine benthic invertebrates. American Naturalist 107:339-352. Vance, R. R. 1973b. More on reproductive strategies in marine benthic in- vertebrates. American Naturalist 107:353-361. Date of manuscript acceptance: 28 February 1992 The evolution of the hindgut of the deep-sea protobranch bivalves J. A. Allen University Marine Biological Station, Millport, Isle of Cumbrae, Scotland, KA28 OEG Abstract. The evolution of the hindgut of the deep-sea Protobranchia is considered in relation to their feeding strategy. One of the cornerstones of their success in colonizing soft abyssal sediments lies in their ability to digest organic materials extracellularly in relatively low concentration. The time needed to digest such materials contained in a continuous column of sediment in the hindgut is maximized by a much elongated gut. Elongation has been accompanied by a variety of configurations in its course, these maximize the length of the tube that can be housed withih the space of the visceral mass. It would appear that this has been a major evolutionary concern, if not the major evolutionary concern, of the deep-sea protobranchs and has proved to be of considerable taxonomic important. The Palaeotaxodonta, which comprise the vast majority of the deep-sea (> 500m) protobranchs, are all deposit feeders and commonly as many as fourteen species can be collected at a single station both on the lower slope and the abyssal plain (unpub. data). Analysis of the gut contents of specimens representing various species from a single station indicates that all are consuming a similar fraction of the soft sediments. Most, if not all, feed via their palp proboscides in the sur- face layers of the abyssal sediments and, probably for the most part, on the surface layer itself. In such a situation it is natural to speculate what, if anything, separates the ecological niches of the cohabiting species. In terms of Hutchinsonian Ecology, Allen (1985) has suggested that habit and space could be the separating factors. A more extreme view would be that there is no niche separation, that competition is low and that specia- tion is neither driven nor restricted by stress. For instance, high pressure could be considered stressful, however having become adapted, in reality all physical factors of deep-sea basins are stable and predictable over geologic time and despite being extreme are unlikely to be stressful. Animal density is low and there is circumstantial evidence that preda- tion is remarkably low (Oliver and Allen, 1980; Turekian et al., 1975), but this may or may not be an indication of the level of competition. Diversity is high however measured (Sanders and Hessler, 1969) with a range of frequency of oc- currence of species from common to rare. Community struc- ture in the deep sea is not atypical compared with other marine communities. Apart from indications from morphology of small variations in habit and function (i.e. some species may be less mobile and deeper burrowing than others) it must be concluded that we are as yet largely blind to the delineation of niche for the infaunal protobranchs of the deep sea. THE PROTOBRANCH GUT There is one morphological variation that is particular- ly obvious in the Palaeotaxodonta, namely the configuration of the hindgut. In naive terms it could be concluded that in their evolution the main thrust has been how best to accom- modate an elongate hindgut within a small body space rather than changes in external shell morphology to gain ecological advantage in a monotonous environment. External morpho- logical variations are subtle, often to the extent that one taxon can be barely recognized from another (Allen and Hannah, 1986). Numerically the protobranchs dominate the particulate feeding abyssal bivalve fauna (Clarke, 1962; Sanders ef al., 1965). (Note: Carnivorous septibranchs are well represented in the deep sea.) One thing, possibly the only one, that ex- plains this dominance is a difference in the digestive physiology of the protobranchs as compared with the lamellibranchs. Little is known of the digestive processes in protobranchs but what is would indicate a major difference in the function of the digestive diverticula (Owen, 1956, 1980; pers. obs.) as compared with that of lamellibranchs. In the organically rich sediments and water column of shelf seas this difference could be of little significance, but in the relatively impoverished sediments and water column of the abyss (Gage and Taylor, 1991), digestive efficiency is vitally important. The efficiency of the selective process to extract scarce organics within a slurry of fine sediment, together with the time taken to digest skeletal proteins also becomes important. To maximize extracellular digestion of refractive food material in relatively low concentrations in fine sediments, and the material attached to silt particles, requires an American Malacological Bulletin, Vol. 9(2) (1992):187-191 187 188 AMER. MALAC. BULL. 9(2) (1992 increased residence time within the gut. An elongated hind- gut allows for greater passage time of sediment and a larger surface of contact between digestive epithelium and gut con- tent, allowing for a more complete digestive and adsorptive process. The elongate gut has to accommodate in a small and confined space and this has been accomplished in a variety of ways. Also, an efficient method of transporting a very long column of sediment through a narrow pipe had to be achieved. To address the last point first, movement through the gut of protobranchs could be dependent, at least in part, on the use of the stomach and the style sac as a squeezable bulb, the pressure so created driving the fluidized gut contents. This is in contrast to the lamellibranchs in which ciliary transport appears to be the driving force. Anatomical observations in- dicate that pressure can be exerted in two ways. One is via the surrounding musculature basal to the stomach epithelium and the second is via modified anterior pedal retractors, which cradle the ventral side of the stomach (unpub. data). When contracted the latter squeeze the dorsal side of the stomach against the inner shell below the umbo. In the protobranchs the dorsal surface of the stomach lies immediately adjacent to the epithelium of the body wall that lines the shell. THE HINDGUT OF THE NUCULANOIDEA The basic design of the hindgut of the nuculanoideans, the most common of the deep-sea protobranchs, is a loop on the right side of the visceral mass (Fig. la). The loop ex- tends from the ventral limit of the style sac dorsally posterior to the style sac and stomach, and thence on the right side in a sweeping antero-ventral curve to the mouth and anterior adductor muscle and from there dorsally and posteriorly along the dorsal side of the viscera, passing through the heart and dorsal to the kidney and posterior adductor muscle to the anus. In most species a single typhlosole is present along the whole length of the hindgut. From this simple configuration a limited number of different types of more complex configurations can be derived. The range of types of configuration is present in a number of families or subfamilies and these types could have evolved more than once in the history of the group. Although there are exceptions, in general protobranch species with more elongate hindguts are found deeper in the ocean than those with short hindguts. For example, Tindaria callistiformis Verrill and Bush, 1897, from depths between 3305 and 5042m, has a hindgut volume per unit animal that is half as large again as T. hessleri Sanders and Allen, 1977, from depths between 1739 and 2339m (Sanders and Allen, 1977). More recent measurements of length show that this applies to other protobranch genera, thus Yoldiella subcir- cularis Odhner, 1960, from 3250-5987m, has a hindgut that is more than four times the length of Y. inconspicua Verrill and Bush, 1898, from 1102-4829m, in animals of the same size (pers. obs.). The protobranchs are not alone in showing this phenomenon, thus similar increases in gut length with depth were shown in the case of a series of species of the tellinoidean genus Abra living at increasing depths (Allen and Sanders, 1966). In many species it would appear that the extended single loop simply takes a wider course around the right side of the body. The diameter of the lumen of the hindgut may also be increased. As seen in the tindariids (Sanders and Allen, 1977) the lengthening of the single loop at its most extreme involves the penetration of the visceral tissue con- taining the ventral portion of the loop into the mantle of the right side to lie close to the inner muscular fold of the man- tle edge. (Fig. 1B). In less extreme cases the ventral limit of the loop lies at the margin of a visceral fold that in large part overlies the right palp. In most cases the lengthened hind- gut also penetrates deep into the foot ventral to, and in some cases, anterior to the pedal ganglia. The configuration of the path taken and the diameter of the gut is quite specific and in very many cases the identification of the species can be accurately based on this one feature. In the progression from a single loop, one sequence involves the coiling of the loop on the right side. This pro- cess can be easily illustrated by manipulating a length of rope and taking hold of the anterior limit of a loop and coiling once, twice and eventually more times in a clockwise direc- tion. Individual species have their own characteristic form and number of coils (Figs. 2A-F). Multiple coils have suffi- cient volume to cause displacement of the stomach to the left, and to limit the digestive diverticulum of the right side to the centre of the coil (Allen and Hannah, 1989). The second sequence from a single loop is but a varia- tion of that described above. Here the anterior section of the loop extends to the left side of the body passing between the oesophagus and the anterior adductor muscle. On the left side of the body a loop is formed and this, depending on the species, may or may not be coiled in the manner described above (Figs. 2G, H). The morphological consequence of this penetration to the left side of the viscera is to displace the mouth posteriorly. This latter form of hindgut configuration is confined to relatively few species. Much more common is a third sequence in which the hindgut on reaching the dorsal margin of the visceral mass to the stomach passes first to the left side of the body, makes a loop around the periphery of the visceral mass, then passes to the right side of the body and describes a single loop of similar extent to that on the left (Fig. 3A). Various extensions and modifications to this left and right looped configuration occur in different species. The most common is where right and left loops are doubled on each side. In rarer cases addi- tional loops can be formed on either side. In species with a vertically elongate foot, lengths of gut can be doubled or quadrupled parallel to the style sac and reach far ventral within the foot (Figs. 3B, C). ALLEN: PROTOBRANCH HINDGUT 189 SS Ar, eee Fig. 1. A, A diagrammatic representation of a generalized nuculanid protobranch from the right side to show the major morphology of the body organs; B, a similar diagrammatic representation of a tindariid protobranch as seen from the right side (for identification of the organs see (A); C, a diagrammatic representation of a vertical cross-section of a tindariid protobranch through the lines a-b in (B) to show the position of the hindgut. aa, anterior adductor muscle; bg, byssal gland; cg, cerebral ganglion; ft, foot; gi, gill; hg, hindgut; ht, heart; ki, kidney; me, mantle edge; pa, posterior adductor muscle; pg, pedal ganglion; pl, palp; pp, palp proboscides; pr, pedal retractor muscle; so, anterior sense organ; ss, siphons; st, stomach; vg, visceral ganglion. 190 AMER. MALAC. BULL. 9(2) (1992) Fig. 2. Line drawings of the course of the gut in eight species of Yoldiella to illustrate the different courses taken by the hindgut with increasing length of the hindgut. The drawings are from the right side, the end of the line to the left represents the anus; the end of the curved line to the right, the mouth; the stomach is shown as a ‘pear*shaped outline. The course of the hindgut is shown as a solid line on the right side of the body and as a dotted line on the left. C Fig. 3. Line drawings of the different courses taken by the hindgut of three species of Yoldiella in which the hindgut loops to both the left and right of the body. The course on the left side is differentiated by a dotted line. Note in all these species the point of passing from left to right and from right to left is posterior to the postero-dorsal limit of the stomach. In all of the configurations of the second and third sequences described above, the passage of the hindgut from one side of the body to the other occurs either anteriorly between oesophagus and anterior adductor muscle or mid- dorsally immediately posterior to the stomach. It can be pic- tured that, as the hindgut develops, it is these two points that provide the gateway to the extension to the left side of the body. It is of note that in no species as yet seen does penetra- tion to the left occur at both points in an individual species. In relatively few species penetration of the hindgut to the left side of the body occurs both postero-dorsal to the stomach and anterior to the stomach (but posterior to the oesophagus) within the mass of the viscera. As in the second and third sequences this has occurred in well-separated tax- onomic groups (Lametilidae and Yoldiellinae) (Allen and Sanders, 1973, unpub. data). THE HINDGUT OF THE NUCULOIDEA In the case of the Nuculoideans, no species has yet been described in which there is a hindgut that describes a single loop to the right. Nevertheless, there is much con- formity within the Nuculoidean taxa. Thus, the Nuculidae all have a coiled hindgut predominantly dorsal and inclined to the right of the stomach, while the Pristiglomidae have multiple loops to the right and the left passing from side to side postero-dorsal to the stomach but with loops carried dorsal to the stomach and viscera (Sanders and Allen, 1978; Rhind and Allen, 1992). DISCUSSION The disposition of the hindgut provides additional clear evidence of the differentiation of the Nuculoida into the Nuculoidea and Nuculanoidea - already well-established on other morphological and functional grounds (e.g. the presence or absence of an anterior inhalent current, anterior mantle sense organ and posterior inhalent and feeding apertures) (Fig. 4). Hindgut differences clearly differentiate the two nuculoi- dean families, the Nuculidae and the Pristiglomidae. Similar universal differences cannot be seen in the case of the nuculanoideans. In a few families there is a predomi- nant hindgut configuration (Fig. 4). These more clear-cut ex- amples include the Nuculaninae, which possess a single loop to the right that is not particularly lengthened, and the Malletiidae and the Tindariidae in which the single loop is extended (Sanders and Allen, 1977, 1985). In the latter fami- ly the hindgut is usually carried into the right mantle lobe. In the Spinulinae, multiple coils are present on the right side of the body in most species (Allen and Sanders, 1982). In contrast, species of the Ledellinae (Allen and Hannah, 1989), the Yoldiellinae and the Neilonellidae (unpub. data) exhibit a wide range of configurations. Whether or not these groups ALLEN: PROTOBRANCH HINDGUT 19] Protobranchia Nuculoida Cae aes Solemyoida No coils or loops ie Nucinellidae Nuculoidea Solemyidae sae le Nuculidae Tindariidae Many coils to right Enlarged loop into right mantle Pristiglomidae Many loops to right & left Neilonellidae Loop to right; loop on r.passes to | and coils; coils to right Nuculanidae Lametilidae Complex looping |.& r. ant & post. to stomach Nuculaninae Yoldiidae Malletiidae Loop to right Loop to right Spinulinae Yoldiinae Yoldiellinae Loop to right; coils to right Loop to right Loop to right; loops to F&I; coils to right Ledellinae Loop to right; loops to r &l.; coils to right oe 5 Siliculidae Loop to right; loop to r. &| Fig. 4. Possible evolutionary relationships within the Protobranchia. are discrete entities or whether they are polyphyletic is not yet absolutely certain. In the case of the Ledellinae, the sub- family comprises a series of closely related species in which other characters, particularly those of the shell, show not only close relationship but are sufficiently distinct as to delineate a suprageneric taxon. This is less clear in the Yoldiellinae (unpub. data), which comprises a very large number of species, but without the strongly marked shell characters displayed by the Spinulinae and Ledellinae. Overall the dif- ferences in shell characters are so slight as to be impossible to arrange the species into distinct units. The Neilonellidae are presently under review, but the indications are that this family exhibiting a variety of gut configurations is monophyletic (pers. obs.). LITERATURE CITED Allen, J. A. 1978. Evolution of the deep-sea protobranch bivalves. Philosophical Transactions of the Royal Society of London 284B:387-401. Allen, J. A. 1983. The ecology of deep-sea molluscs. Jn: The Mollusca, 6, Ecology, W. D. Russell-Hunter, ed. pp. 29-75. Academic Press Inc. (London) Ltd. Allen, J. A.1985. The Recent Bivalvia: their form and evolution. In: The Mollusca, 10, Evolution. E. R. Trueman and M. R. Clarke, eds. pp. 377-403, Academic Press Inc. (London) Ltd. Allen, J. A. and F. J. Hannah. 1986. A reclassification of the Recent genera of the subclass Protobranchia (Mollusca: Bivalvia). Journal of Concholgy, 32:225-249. Allen, J. A. and H. L. Sanders. 1966. Adaptations to abyssal life as shown by the bivalve Abra profundorum (Smith). Deep Sea Research, 13:1175-1184. Allen, J. A. and H. L. Sanders. 1973. Studies on deep-sea Protobranchia. The families Siliculidae and Lametilidae. Bulletin of the Museum of Comparative Zoology Harvard, 150:1-36. Clarke, A. H. 1962. Annotated list and bibliography of the abyssal marine molluscs of the world. Bulletin of the National Museum of Canada, 181:1-114. Gage, J. D. and P. A. Tyler. 1991. Deep-sea Biology. a Natural History of the Organisms at the Deep-sea Floor. Cambridge University Press. 504 pp. Newell, N. D. 1969. Classification of Bivalvia. In: Treatise on Invertebrate Paleontology, Part IV, 1, Mollusca Bivalvia. R. C. Moore, ed. pp. N205-N224. Geological Society of America and the University of Kansas. Oliver, G. and J. A. Allen. 1980. On the adaptations of the Limopsidae (Mollusca: Bivalvia) of the abyssal Atlantic. Philosophical Transac- tions of the Royal Society of London, B, 291:77-125. Owen, G. 1956. Observations on the stomach and digestive diverticula of the Lamellibranchia. I]. The Nuculidae. Quarterly Journal of Microscopical Science 97 :541-567. Owen, G. 1973. The fine structure and histochemistry of the digestive diver- ticula of the protobranchiate bivalve Nucula sulcata. Proceedings of the Royal Society, B 183:249-264. Rhind, P. M. and J. A. Allen. 1992. Studies on the deep-sea Protobranchia (Bivalvia); the family Nuculidae. Bulletin of the British Museum of Natural History (Zoology) 58(1):1-33 Sanders, H. L. and J. A. Allen. 1973. Studies on deep-sea Protobranchia: Prologue and the Pristiglomidae, Bulletin of the Museum of Com- parative Zoology Harvard 145:263-314. Sanders, H. L. and J. A. Allen. 1977. Studies on deep-sea Protobranchia: the family Tindaridae and the genus Pseudotindaria. Bulletin of the Museum of Comparatiave Zoology Harvard 148:29-59. Sanders, H. L. and J. A. Allen. 1985. Studies on deep-sea Protobranchia (Bivalvia); the family Malletiidae. Bulletin of the British Museum of Natural History (Zoology) 49:195-238. Sanders, H. L. and R. R. Hessler. 1969. Ecology of deep-sea benthos. Science 163:1419-1424. Sanders, H. L., R. R. Hessler and G. R. Hampson. 1965. An introduction to the study of deep-sea benthic faunal assemblages along the Gay Head - Bermuda transect. Deep-sea Research 12:845-867. Turekian, K. K., J. K. Cochran, D. P. Kharkar, R. M. Cerreto, J. R. Vaisnys, H. L. Sanders, J. F. Grassle and J. A. Allen. 1975. Slow growth rate of a deep-sea clam determined by 2?8Ra chronology. Proceedings of the National Academy of Sciences, U.S.A. 72:2829-2832. Date of manuscript acceptance: 20 April 1992 Prismatic shell formation in continuously isolated (Mytilus edulis) and periodically exposed (Crassostrea virginica) extrapallial spaces: explicable by the same concept? Melbourne R. Carriker College of Marine Studies, University of Delaware, Lewes, Delaware 19958, U.S.A. Abstract. Many aspects of molluscan biomineralization and shell formation yet remain unexplained. A comparison of formation of prismatic shell in two species of bivalves is presented that suggests, as indicated by a current hypothesis, that soluble organic matrix precedes and regulates the type and form of biocrystals. In Mytilus edulis Linné prisms develop at the shell margin in an extrapallial space continuously closed to seawater by the emerging periostracal sheet, the mantle margin remaining in place. In Crassostrea virginica (Gmelin), however, margins of mantle lobes are frequently withdrawn into the mantle cavity, exposing prismatic shell surfaces to seawater. Prism-secreting mantle cells, even though they could be precisely repositioned over growing prisms, probably do not control the growth of these prisms; rather, once nucleated in the soluable matrix, each biocrystal with its accompanying matrix probably mediates mineralization and shape and size of that prism and subsequent prisms in the prismatic column. Thus, after the secreting epithelium is extended to the valve edge, bathing growing prisms in extrapallial fluid, development is resumed without deformation. Studies of the morphology and development of EXTRAPALLIAL SPACE molluscan shell have proliferated during the past thirty years. Yet many aspects of shell formation, such as biomineraliza- Among bivalve molluscs, extracellular shell formation tion, for example, remain imperfectly understood. Of primary generally occurs between the mantle shell-secreting interest are the cellular physicochemical mechanisms that epithelium and the inner surface of each of the two shell valves bring about nucleation, orientation, size, micromorphology, in the very thin extrapallial space. Within this space, the and polymorphic type of the biocrystals, the basic microstruc- solubility product of the minerals being incorporated in shell tural units of shell (Simkiss and Wilbur, 1989; Carriker, 1991; microstructures is probably surpassed by a change in the con- Rosenberg and Hughes, 1991). The present overview explores centration of precursor ions. The distance between epithelial principally the micromorphological interplay at bivalve shell and shell surfaces is so small that transfer of mineral ions margins of a) the shell-secreting epithelium, b) developing and organic molecules from the secretory epithelium to the prismatic microstructures, and c) freshly secreted valve surface could occur virtually by direct contact (Simkiss periostracum in a) an extrapallial space continously closed and Wilbur, 1989). to seawater, and b) one periodically exposed to flooding by In many bivalves the extrapallial space is closely sealed seawater. from seawater. This closure is effected by the newly emerg- Is it conceivable that prismatic microstructures, formed ing periostracal sheet, which maintains contact between the in two such apparently disparate anatomical microenviron- outer mantle margin and the free edge of the valve (Saleud- ments, are the products of the same biophysicochemical din and Petit, 1983). As the periostracal sheet is pressed from mechanism? It would appear so. The existence of such a the periostracal groove, it becomes temporarily attached to similar mechanism is indicated by the present comparison the outer surface of the middle mantle fold by adhesive of formation of prismatic shell in two species of molluscan epithelia. Blood pressure in the hemocoel of the outer man- bivalves, both in the Subclass Pteriomorpha: Mytilus edulis tle fold holds the secreting epithelium tightly against the Linne (Family Mytilidae, Order Mytiloida) and Crassostrea periostracum (Dunachie, 1962-63). In this way, the ex- virginica (Gmelin) (Family Ostreidae, Order Pterioida). The trapallial space is completely enclosed, and shell growth at mechanism is interpretable in terms of the current hypothesis the valve margin takes place without dilution by environmental that organic matrix of shell is an organized medium that serves seawater. Species with continuously isolated extrapallial as a mediator of biomineralization (Simkiss and Wilbur, 1989) spaces include the marine bivalve Mytilus edulis (Figs. 1-3) in the extrapallial space. (Dunachie, 1962-63; Mutvei, 1972; Bayne, 1976; Rosenberg American Malacological Bulletin, Vol. 9(2) (1992):193-197 193 194 AMER. MALAC. BULL. 9(2) (1992) Figs. 1-3. Scanning electron micrographs (SEM) of valves (6 cm long) of the shell of Mytilus edulis, washed, air dried, fractured with a hammer, pieces mounted and coated with gold. Fig. 1. A fractured surface (f) perpendicular to prismatic shell margin (e) in mid-ventral region of the valve, showing the edge enveloped by periostracal sheet (p). Inner edge of sheet (in life arising in the periostracal groove of the mantle edge) now stuck to inner surface of the valve(s). Growing prisms (m) lie inside periostracal sheet (p) (horizontal field width = 0.9 um). Fig. 2. Enlargement of figure 1 at outer edge (right) of the V-shaped break in periostracal sheet (p), prism tips (m) to the left beneath it. Fracture surface (f) [horizontal field width (hfw) = 45 pm]. Fig. 3. Enlargement of prism ends (m) in figure 1 (hfw = 18 um). Fig. 4. SEM of clear, new, single, dried, periostracal sheet (p) slightly folded under (to the left), older part of the sheet (to the right) containing a small developing prisms (m), at the mid ventral edge of the left valve of a spat of Crassostrea virginica (1.2 cm high) that had set on a glass surface (g); oyster flesh was pulled away leaving periostracal sheet in place against the glass surface. Specimen was dehydrated in alcohol, mounted, oven dried, and coated with carbon and gold (hfw = 20 um). Figs. 5-6. SEMs of inner surface of prismatic shell of C. virginica raised in a hatchery until | cm high, then grown in local estuary for 2 months, and finally laboratory cultured where food was added and seawater was changed daily for 10 days to allow oysters to deposit shell in undisiurbed conditions. Valve surfaces were cleaned with a soft brush under tap water, pieces were sawed out under running tap water to avoid contamination wiih shell dust, dehydrated in alcohol, mounted, and coated with carbon and gold. Fig. 5. Interior surface of the mid-ventral prismatic margin of right valve, a short distance inward from valve edge. Prisms (m) are considerably larger than those in figure 4, and there is still substantial organic matrix (c) (conchiolin) between prisms (hfw = 40 um). Fig. 6. The same general region of interior surface of prismatic margin as in figure 5 but farther inside the shell edge. Prisms are larger, and have ‘‘crowded out’’ more of the organic matrix (c) (hfw = 40 um). Fig. 7 SEM of three dimensional view of young full formed prisms (m) in a section of right valve of Crassostrea virginica (same lot as in figures 5-6) fractured by breaking and showing growth layers (1). The preparation was treated with full strength bleach for 1 min before dehydration to dissolve superficial organic matrix (hfw = 40 pm). CARRIKER: PRISMATIC SHELL FORMATION 195 and Hughes (1991), the freshwater clam Amblema plicata perplicata (Conrad) (Saleuddin and Petit, 1983), and the freshwater and estuarine invader Dreissena polymorpha (Pallas) (Morton, 1960). In a second major group of bivalve molluscs, pallial attachment to the interior surface of the valves occurs well inside valve margins, allowing deep retraction of the ventral half of the mantle lobes and flooding of extrapallial spaces with seawater. Mantle withdrawal is characteristic, for ex- ample, of species of the Anomiodea, Limoidea, Ostreoidea, Pectinoidea, and Pinnacoidea. In these taxa, structural changes associated with monomyarianism (Yonge, 1953) have resulted in secondary pallial attachment, tenuous mantle- periostracal contact, and a thin, inconspicuous external shell periostracum (Taylor et al., 1969). A notable example of a species that periodically exposes its extrapallial spaces is the Eastern Oyster Crassostrea virginica (Galtsoff, 1964; Car- riker et al., 1980; Carriker, 1991). Another is the San Diego Scallop Pecten diegensis Dall (Clark, 1974). In species with periodically exposed extrapallial spaces, the margin of the valves tapers to a thin edge, and the forming periostracum arising in the periostracal groove is extremely thin and appears to possess little stability. The region of the mantle lobes between the single adductor mus- cle and ventral margins can be physically highly active. In Crassostrea virginica, for example, in which the only attach- ment of the mantle lobes to the valves is at the circumference of the adductor muscle, the lobes can extend some distance beyond the edge of the valves, withdraw deeply inside the shell, form ridges, and roll into a temporary channel to facilitate rejection of undigestible particles in mucus. These movements can involve a small part of the ventral half of the lobes, or all of it, depending on the intensity of stimulation received by tentacles of the middle and inner mantle folds. In an oyster whose valves are closed, mantle margins are nor- mally retracted to midway between the margin of the gills and edges of the valves (Galtsoff, 1964; Carriker, 1991). Man- tle contraction exposes newly forming surfaces of shell to a wash with seawater, and breaks any connection that might have existed between periostracal sheets and valve edges. ORGANIC SHELL MATRIX Organic matrix of molluscan shell is considered an organized, genetically programmed medium that in some way functions to nucleate minerals, determine the mineral phase (polymorph), and regulate crystallographic orientation, microarchitecture, growth, and size of mineralized shell microstructures (Simkiss and Wilbur, 1989; Crenshaw, 1990). A current view is that insoluble matrix forms a structural framework, and that soluble matrix present in and around it in the extrapallial fluid serves as a nucleating surface; in solution, soluble matrix possibly can act as an inhibitor, con- trolling the thickness of layers (lamina) of microstructures (Simkiss and Wilbur, 1989; Crenshaw, 1990). Structurally speaking, each biocrystal originates as an ion cluster that grows into a critical nucleus attracted to, or formed at, the charged surface of the insoluble matrix. Each nucleus, upon addition of ions from the extrapallial fluid, develops into a small crystal, and this grows into a definitive biocrystal within the insoluble matrix. The biocrystal and ac- companying internal and external soluble and insoluble matrix constitute the microstructural unit. It is possible there is feed- back from the growing biocrystal surface to secretory man- tle cells over the biocrystal that could facilitate coordination of microstructural formation. Shell microstructures ultimately reach a general size and shape consistent with the micro- morphological-mineralogical type characteristic of its shell region (Wilbur, 1974; Carriker et al., 1980; Simkiss and Wilbur, 1989: Crenshaw, 1990; Carriker, 1991). Composition of the insoluble organic matrix varies in different microstructural regions of a single valve (Simkiss and Wilbur, 1989; Crenshaw, 1990), probably accounting, in part, for such dissimilar microstructures as prismatic, foliated, chalky, and myostracal in, for example, Crassostrea virginica. PRISM FORMATION IN AN OPEN EXTRAPALLIAL SPACE A closer examination of how prisms appear to form at the growing margin of the valves of Crassostrea virginica explains how the developing concepts of microstructural for- mation reviewed in the previous paragraphs could apply, not only in isolated, but also in periodically exposed extrapallial spaces. Mantle-edge activity and formation of shell prisms at the margin of valves have been studied a) in young live oysters in seawater under a binocular microscope (Tomaszewski, 1982), b) by inserting pieces of thin glass between the edge of the mantle and the valve of live oysters, and removing the glass at regular intervals for analysis under the light microscope (Galtsoff, 1964), and c) viewing with a scanning electron microscope the newly deposited periostracum and developing biocrystals at the edge of the left valve of young oysters that had set on small pieces of thin glass (Carriker et al., 1980; Carriker, 1991). Clark (1974) investigated the growing margin of Pecten diegensis with time lapse photography and scanning electron microscopy. During prism formation, the oyster, valves gaping slightly and pumping seawater, extends both left and right mantle-lobe margins a short distance beyond their respective, thin, mineralized, valve edges. In the temporarily enclosed extrapallial space of each valve, the oyster releases onto and beyond the mineralized valve margins from the periostracal grooves, a thin, clear, viscous, sometimes stringy, sheet of 196 AMER. MALAC. BULL. 9(2) (1992) periostracum. When the margin of the left valve is affixed to a hard substratum (like shell or glass), the new periostracal sheet is laid over this (Fig. 4); the sheet on the right valve if off of the substratum and that on the left valve, if the valve is no longer attached to the substratum, are extended into and remain suspended in sewawater. Contact with seawater probably firms the previously liquid periostracal sheets enough that they maintain their form and position even though suspended in seawater. During release of the fluid perio- stracum, the mantle expands and retracts actively, spreading the secretion over previously deposited periostracal layers and beyond these to form new ones. New left and right marginal periostracal sheets probably never adhere to each other dur- ing shell formation because oysters, with valves slightly apart, continue pumping seawater during shell deposition. It is likely periostracum soon looses its surface adhesiveness when ex- posed to seawater. The first crystallites, becoming visible under the magnification of the scanning electron microscope as very small, roughly rounded, randomly distributed bodies (0.01 pan or less), and presumably surrounded by extrapallial fluid supersaturated with respect to the minerals being deposited, appear embedded within the new periostracal matrix, over and beyond the previously mineralized shell edge (Fig. 4). As the thin, wafer-like biocrystals grow in diameter and thickness, their margins approach each other, apparently ‘*squeezing’’ organic matrix between them (Figs. 5, 6). Once lateral boundaries of prisms come close, prism growth becomes primarily lengthwise, resulting at maturity in long slender needle-like structures generally polygonal in cross section (Fig. 7). In micrographs, biocrystals are shown as growing embedded within the sheets of organic matrix (Fig. 4) rather than on the surface (Carriker et al., 1980; Wilbur and Saleuddin, 1983; Simkiss and Wilbur, 1989). It would seem that in the oyster initial biocrystal for- mation can occur virtually independent of a stable shell margin substratum, at least in those parts of the freshly formed periostracal sheet suspended in seawater beyond the edge of the mineralized shell margin. Whether, as Clark (1974) suggested for the scallop, stability on a firm foundation is a requirement for orderly shell growth in the oyster, is unclear. This seems unlikely in view of the sequence of prism develop- ment represented in the micrographs (Figs. 4-6). Being exposed to environmental seawater during periods of withdrawal of the mantle, the freshly growing, ap- parently slightly viscid marginal valve surface could be vulnerable to contamination by clays, silts, and other suspend- ed particles. This could explain the presence of foreign par- ticle and chemical contaminants in shell (Simkiss, 1965; Car- riker ef al., 1982, 1991). The degree of ‘‘stickiness’’, if any, of the hardening periostracum has not been determined. Adventitious impurities are probably characteristically pre- sent in all molluscan shells that possess transiently open ex- trapallial spaces, and especially of those species that inhabit coastal and estuarine waters that tend to be highly turbid (Car- riker, 1967; Carriker et al., 1980; Carriker, 1986). COMMON MECHANISMS AND FURTHER INQUIRIES Enclosed, continuously isolated extrapallial spaces, like those of Mytilus edulis, would seem to offer an ideal microenvironment for the growth of prismatic biocrystals (Figs. 1-3). By contrast, periodic withdrawal of mantle lobes and the ensuing flooding of extrapallial spaces with seawater, as in such species as Crassostrea virginica (Figs. 4-6), would appear to pose a disadvantage in the process of marginal prism formation. Yet, this strikes me as not being the case, because bivalves with open biomineralization systems form shell ap- parently as successfully as those with closed systems; both groups contain species that are probably equally successful biologically. One can conjecture that in the Bivalvia, bio- mineralization evolved first as an extracellular process on open mantle edges (Stasek, 1972). Hence evolution of the genetically programmed organic matrix of shell appeared first in open systems, and was retained as a basic component of the biochemical mechanism of biomineralization as closed systems evolved in other species. To what extent intermediate anatomical configurations between open and closed systems, if they exist, might shed information on the evolution of closed systems, would be worth investigating (Simkiss, 1974). What utility, if any, the closed system could have over an open one relative to the mechanism of shell formation and survival of the bivalve, is not clear. There is the obvious possibility that shell produced in isolated margins is freer of adventitous particles and chemicals than that in open margins; whether this is, or is not, an advantage to a bivalve is still a question, and a point worth pursuing in view of the increas- ingly contaminated waters of coastal regions of the world (Simkiss, 1965; Carriker, 1976). Because of the hypothesized importance of the genet- ically programmed organic matrix, whether in open or closed mantle margins, there is no longer any need to presume, as one of the possibilities given by Carriker et al. (1980), that a precise spatial microassociation could exist between man- tle epithelial shell-secreting cells and developing microstruc- tural units growing apposed to them, nor that specific clusters of secretory cells must be repositioned over the ends of biocrystals first formed by them before each withdrawal of mantle lobes. Although prismatic shell formation, or any shell growth for that matter, is understood imperfectly (Simkiss and Wilbur, 1989), even scantier knowledge is available on the control of the transition in microstructure and in mineralogy in the same developing shell valve. There are such CARRIKER: PRISMATIC SHELL FORMATION 197 common changes, for example, as marginal prismatic shell to foliated (in oysters) or nacreous (in mussels) inside the valves as the shell increases in size, or the introduction of aragonitic myostracum over calcitic folia (in oysters) as the adductor muscle scar migrates marginward during shell enlargement (Wilbur and Saleuddin, 1983). Although the basic mechanisms of biocrystal forma- tion could be common to all biological systems (Crenshaw, 1990), as shown in Mytilus edulis and Crassostrea virginica the physicochemical microenvironment in which biomineral- ization occurs can vary. Comparative studies of the similarities and differences in the anatomical structures and processes of shell formation in different taxa could provide a greatly needed perspective (Wilbur, 1974). And because biomineral- ization is largely about ions and structures, initially at molecular and cellular levels, and subsequently at the organismal level (Simkiss and Wilbur, 1989), comparative investigations could be exceedingly fruitful when approached multidisciplinarily. Crenshaw (1990) noted that most of the progress toward an understanding of biomineralization at all phylogenetic levels has been made as a result of research on invertebrate animal systems, especially those of molluscs. The present overview would suggest that Mytilus edulis and Crassostrea virginica, because of the strikingly different arrangement of the microspaces in which shell formation takes place in these species, would serve as excellent models in the further unravelling of the fascinating complexities of biomineraliza- tion and mineralized skeletal formation in both plants and animals. ACKNOWLEDGMENTS Special thanks go to Dirk Van Zant, Virginia Peters, John Ewart, Robert Palmer and Robert Prezant for collaboration in the original preparation of specimens and scanning electron microscopy. Robert Bowden II prepared prints of the micrographs. I am also grateful to Miles Crenshaw, Karolyn Mueller Hansen, Robert Prezant, Kenneth Simkiss, and Herbert Waite for valuable comments on the manuscript. LITERATURE CITED Bayne, B. L., ed. 1976. Marine Mussels: Their Ecology and Physiology. In- ternational Biological Programme 10, Cambridge University Press, London. 506 pp. Carriker, M. R. 1967. Ecology of estuarine benthic invertebrates: a perspec- tive. In: Estuaries, G. H. Lauff, ed. pp. 442-487. American Associa- tion for the Advancement of Science, Washington, D. C. Carriker, M. R. 1976. The crucial role of systematics in assessing pollution effects on the biological utilization of estuaries. Jn: Estuarine Pollu- tion Control and Assessment. pp. 487-506. Office of Water Planning and Standards. U. S. Government Printing Office, Washington, D. C. Carriker, M. R. 1986. Influence of suspended sediment on biology of oyster larvae in estuaries. American Malacological Bulletin, Special Edi- tion No. 3:41-49. Carriker, M. R. 1992. The shell and ligament. In: The Biology, Culture, and Management of the Eastern Oyster, A. F. Eble, V. S. Kennedy, and R. I. E. Newell, eds. University of Maryland Office of Sea Grant (in press). Carriker, M. R., R. E. Palmer, and R. S. Prezant. 1980. Functional ultramor- phology of the dissoconch valves of the oyster Crassostrea virginica. Proceedings of the National Shellfisheries Association 70:139-183. Carriker, M. R., C. P. Swann, and J. W. Ewart. 1982. An exploratory study with the proton microprobe of the ontogenetic distribution of 16 elements in the shell of living oysters (Crassostrea virginica). Marine Biology 69:235-246. Carriker, M. R., C. P. Swann, R. S. Prezant, and C. L. Counts, II]. 1991. Chemical elements in the aragonitic and calcitic microstructural groups of shell of the oyster Crassostrea virginica: a proton probe study. Marine Biology 109:287-297. Clark, G. R., I. 1974. Calcification on an unstable substrate: marginal growth in the mollusk Pecten diegensis. Science 183:968-970. Crenshaw, M. A. 1990. Biomineralization mechanisms. Jn: Skeletal Biomineralizations: Patterns, Processes and Evolutionary Trends. J. G. Carter, ed. pp. 1-9. Van Nostrand Reinhold, New York. Dunachie, J. F. 1962-63. The periostracum of Mytilus edulis. Transactions of the Royal Society of Edinburgh 65:383-4l11. Galtsoff, P. S. 1964. The American Oyster Crassostrea virginica (Gmelin). Fishery Bulletin, Fish and Wildlife Service, U. S. Department of the Interior, U. S. Government Printing Office, Washington. 480 pp. Morton, B. 1969. Studies of biology of Dreissena polymorpha Pall. 1. General anatomy and morphology. Proceedings of the Malacological Society of London 38:301-321. Mutvei, H. 1972. Formation of nacreous and prismatic layers in Mytilus edulis L. (Lamellibranchiata). Biomineralization Research Reports 6:96-100. Rosenberg, G. D. and W. W. Hughes. 1991. A metabolic model for the deter- mination of shell composition in the bivalve mollusc, Mytilus edulis. Lethaia 24:84-96. Saleuddin, A. S. M. and H. P. Petit. 1983. The mode of formation and the structure of periostracum. In: The Mollusca, Volume 4, Physiology, Part 1, A. S. M. Saleuddin and K. M. Wilbur, eds. pp. 199-234. Academic Press, New York. Simkiss, K. 1965. The organic matrix of the oyster shell. Comparative Biochemistry and Physiology 16:427-435. Simkiss, K. 1974. Cellular aspects of calcification. In: The Mechanisms of Mineralization in the Invertebrates and Plants, N. Watabe and K. M. Wilbur, eds. pp. 1-31. University of South Carolina Press, Columbia, South Carolina. Simkiss, K. and K. M. Wilbur. 1989. Biomineralization. Cell Biology and Mineral Deposition. Academic Press, New York. 337 pp. Stasek, C. R. 1972. The molluscan framework. Jn: Chemical Zoology, M. Florkin and B. T. Scheer, eds. pp. 1-44. Academic Press, New York. Taylor, J. D., W. J. Kennedy, and A. Hall. 1969. The shell structure and mineralogy of the Bivalvia. Introduction. Nuculacea-Trigonacea. Bulletin of the British Museum (Natural History), Supplement 3, 125 pp. Tomaszewski, C. 1981. Cementation in the early dissoconch stage of Crassostrea virginica (Gmelin). Master's Thesis, University of Delaware 94 pp. Wilbur, K. M. 1974. Recent studies of invertebrate mineralization. In: The Mechanisms of Mineralization in the Invertebrates and Plants, N. Watabe and K. M. Wilbur, eds. pp. 79-108. University of South Carolina Press, Columbia. Wilbur, K. M. and A. S. M. Saleuddin. 1983. Shell formation. In: The Mollusca, Volume 4, Physiology, Part 1, A. S. M. Saleuddin and K. M. Wilbur, eds. pp. 235-287. Academic Press, New York. Yonge, C. M. 1953. The monomyarian condition in the Lamellibranchia. Transactions of the Royal Society of Edinburgh 62:443-478. Date of manuscript acceptance: 28 November 1991 A new approach to the study of bivalve evolution Mary Ellen Harte 1180 Cragmont Avenue, Berkeley, California 94708, U. S. A. Abstract. Systematic and evolutionary studies of the Bivalvia have been based mostly on obvious conchological characters, but such characters could often reflect parallel adaptations and not phylogenetic relationships. Of the various biomolecular techniques capable of measuring genetic relationships among taxa, radio-immuno-assay (RIA) is particularly suited for bivalve studies. RIA measures genetic distance between taxa by measuring how much antibody made against proteins of one species binds with proteins from another. The results correlate highly with DNA hybridization, DNA sequencing and microcomplement results. RIA is unique in that it can extract information from proteins preserved in fossil and recent calcareous matrices. Because most molluscan collections consist predominantly of dry shells, RIA could prove to be very important for their evolutionary studies. I am using RIA to study parallel evolution in Veneridae and to develop a phylogenetic outline of the family. Venerid taxonomy is based currently on obvious conchological similarities. Preliminary results indicate that: 1) obvious conchological similarities can be parallel adaptations; 2) closely related species could exhibit wide conchological divergences in response to different life strategies; 3) Veneridae is deeply divided, but its origin is monophyletic. More accurate classifications could need to depend more strongly on anatomical, biomolecular and biogeographical data. Bivalve systematics and studies of bivalve evolution parallelism. Both ‘‘parallel’’ and ‘‘convergent’’ have been have been based historically upon conchological characters used to describe the acquiring of superficially similar because mollusc collections, both recent and fossil, consist characters through evolutionary time as similar adaptations primarily of dry shells. Shells are durable, and relatively easy to similar selective pressures by genetically distant taxa. When to transport, preserve and examine. Thus, bivalve evolution interpreting data on clusters of extant taxa, such a has often been discussed in terms of conchological diversity phenomenon can only be inferred, because data on ancestral adapting to environmental diversity (e.g. Stanley, 1970, 1977a, species are absent. Depending on the definition of the points 1981). of reference, the inference can be wrong. The superficially But how often do conchological similarities reflect similar but genetically distant taxa of A and B, for example, genetic versus adaptive similarities? How large a role does could appear to be convergent. Yet the immediate ancestor adaptive parallelism play in evolution? Such questions have of A could be much more similar to that of B, so that A and led to the development of evolutionary systematics, a school B are really diverging currently, even if their overall history that argues that systematics should be based on genetic, not is one of convergence. Thus, to advance beyond inferrence, morphological similarities. Thus, systematics is intended to both extinct and extant points of reference must be defined, reflect the evolutionary processes of species and higher taxa, as well as the scale and scope of characters that comprise and not merely the groupings of morphologically similar ob- the presumed parallelism. For most studies, such data are jects (De Queiroz, 1988; Lindberg, 1989). rarely available. To succeed, fossils of ancestral taxa are need- Biomolecular techniques have provided some insight ed from which both morphological and genetic data can be into adaptive parallelism, and resulting phylogenies of various extracted. extant groups of animals (e.g. Sibley and Ahlquist, 1991; Opportunities to do this within the Bivalvia exist Lowenstein, 1985; Jope, 1980) suggest that such parallelism through the biomolecular technique radio-immuno-assay is more prevalent than previously suspected. Parallelism was (RIA). RIA measures immunologically genetic distance be- an important component in the early evolution of Bivalvia tween taxa by measuring how much antibodies made against (Stanley, 1974); apparent cases occur frequently throughout proteins of one species bind with proteins from another. The Cardiidae (Savazzi, 1985) and Veneridae. More knowledge method is comparable roughly to DNA hybridization in that about the nature of the process might be gained by specifically the products derived from the DNA, the proteins, are essen- examining how and where parallel evolution actually occurs tially hybridized via the mediation of the immunological reac- throughout the evolution of a large successful family of tion. The immunological distances resulting from RIA cor- organisms. relate highly with DNA hybridization (Sibley and Ahlquist, There are inherent limits in examining adaptive 1991), microcomplement fixation (Lowenstein et al., 1981) American Malacological Bulletin, Vol. 9(2) (1992):199-206 1 200 AMER. MALAC. BULL. 9(2) (1992) and DNA sequencing results (Lowenstein and Scheuenstuhl, 1991). RIA can detect extremely small amounts of protein (Luft and Yalow, 1974) and can distinguish closely related species (Lowenstein et al., 1981; Lowenstein and Ryder, 1985) or compare widely divergent groups (Lowenstein, 1981). A unique feature of RIA is that it can provide information on proteins preserved in fossils as old as 60 million years (Westbroek et al. , 1979) and in calcareous matrices (Lowen- stein, 1981; Lowenstein et al., 1982, 1991; Molleson, 1982; Rainey et al., 1984; Collins et al., 1988). RIA has been used successfully on such disparate groups as algae (Olsen- Stojkovich et al., 1986), gymnosperms (Price et al., 1987), brachiopods (J. M. Lowenstein, pers. comm.), and reptiles, birds, and mammals (Lowenstein, 1981). Within the Bivalvia, the Veneridae (Heterodonta: Veneracea) is an example of a large, global, diverse family with a rich fossil record; 500 or more extant species are classified into approximately 12 subfamilies, with 50 extant and 55 extinct genera, and 150 extant and 99 extinct subgenera. The earliest venerid fossils are approximately 130 million years old. Present in a wide variety of marine ecosystems, members of the family are characterized by hav- ing three cardinal teeth in each valve, and sometimes up to three anterior teeth (one in the left valve and two smaller in- terlocking ones in the right). A lunule, escutcheon, and pallial sinus are usually present; valves have concentric sculpture ranging from smooth to pronounced, and sometimes radial and divaricate sculpture, as well. Venerid taxonomy is controversial, with several dis- crepancies among recent systematic works (Fischer-Piette and Delmas, 1967; Keen, 1969; Fischer-Piette, 1975; Fischer- Piette and Vukadinovic, 1975, 1977). No consistent, com- prehensive conchological descriptions exist for the sub- families, genera and subgenera. Genera, especially of minute clams, continue to be moved among subfamilies or changed by workers (Bernard, 1982; Lindberg, 1989), who must weigh which characters indicate true phylogenetic alliance versus parallel adaptations. Little genetic information exists on the Veneridae. RNA sequencing data exist on three taxa (Bowman, 1989) but the amount is inadequate to draw significant conclusions about the family. Phylogenetic estimations are mostly based on conchological characters (e.g. Parker, 1949; Casey, 1952; Fischer-Piette and Vukadinovic, 1977) or shell microstruc- ture (Shimamoto, 1986). The taxonomic confusion, lack of genetic information, and the size, age and diversity of Veneridae all highlight the potential gains to be had through RIA in understanding the evolution and systematics of the family. Geographic and anatomical data indicate parallel evolu- tion within Veneridae. Jones (1979) anatomically examined four species of the sub-family Chioninae: the west Pacific Chione (Austrovenus) stutchburii (Wood), the west Atlantic Chione cancellata (Linneaus) and Mercenaria mercenaria (Linneaus), and the east Pacific Chione californiensis (Broderip). The genus Chione is characterized by large, well- spaced concentric cords or lamellae, and strong radial ribs on the valves. Mercenaria has fine, closely spaced concen- tric threads that merge medially into smooth areas but can develop into weak, low lamellae laterally on the valves; a prominent, rugosely sculpted nymph is also present (Figs. If-g). Jones (1979) observed that anatomically, the three North American species (the western Atlantic Ocean and eastern Pacific Ocean) were allied much more closely to each other than to C. stutchburii, despite the conchological differences, indicating either a parallelism in anatomy among the American chionines, or in conchology between C. stutchburii and the American Chione spp. Similarly, Keen (1969) places two geographically disparate but conchologically similar species, Anomalocardia brasiliana (Linneaus) and Cryptonomella producta (Anton) within the chionine genus Anomalocardia Schumacher, which she characterized as thick shells, with undulating concentric folds crossed by radial riblets, and large, impressed lunules. Harte (1992) observed similarities in sculpture, profile and nymphs between Anomalocardia s. s. and species of Chione and Mercenaria, however, and A. brasiliana overlaps geographically with Chione cancellata, which might indicate a common genetic origin. Past workers have occasionally classified Anomalocardia as a subgenus of Chione (Olsson, 1932; Parker, 1949). As a preliminary test of RIA on venerid shells, I used RIA to determine whether the above two cases inferred conchological parallelism. The results are presented here. MATERIALS AND METHODS RIA data were obtained for two groups of taxa (Fig. 1) to test for parallel evolution. Nomenclature follows that of Keen (1969). The first group included: Chione cancellata (CC), C. (Austrovenus) stutchburii (CS) and Mercenaria mercenaria (Linneaus) (MM). Another west Pacific chionine species, Timoclea (Glycydonta) marica Linneaus) (TM) was also included. The second test group included Anomalocardia brasiliana (AB) and A. (Cryptonemella) producta (PR), which were compared with the first group. Macrocallista nimbosa (Lightfoot) (NI) (Veneridae: Pitarinae), was used as an outgroup to the two test groups of chionines. RIA analyses on the above species were carried out under the direction of Dr. Jerold Lowenstein at the Medical Center of the University of California at San Francisco, where he has refined the method with over a decade of research on various groups of organisms. For each species, approximately 10 g of shell was ground to a powder and placed in 100 cc of 0.2M EDTA, a calcium chelating agent, for two days to dissolve the CaCo,. The resulting protein solution was then HARTE: BIVALVE EVOLUTION — A NEW APPROACH 201 Fig. 1. a. Anomalocardia brasiliana, (shell length (s.1.) = 3.0 cm]; b. A. (Cryptonemella) producta (s.1. = 3.2 cm); ¢. Chione cancellata (s.1. = 2.8 cm); d. Timoclea marica (s.1. = 2.2. cm); e. C. (Austrovenus) stutchburii (s.1. = 3.8 cm); f-h. Mercenaria mercenaria. f. interior with rugose nymph indicated by white bar; g. rugose nymph (section between white bars = top; 5.5 cm, bottom). tested for adequate reactivity against a pre-existing antibody made against Arctica islandica Linne. If the test was suc- cessful, a rabbit was injected at two points along its neck with 2.0 cc total of equal parts shell protein solution (antigens), and Freund’s adjuvant, an agent used to stimulate antibody response. Every two weeks thereafter, the rabbit was further injected with 1.0 cc of shell protein solution. At the end of two months, the rabbit was drained of its serum, containing the antibodies made against the shell proteins. A 0.1 dilution of the antigens for each species was made and tested against the antibodies of all the species. Antigens of species A were placed in a cup of a plastic microtiter plate and allowed to bind with the plastic for one hour. The excess was removed, and the cup was coated with 0.5% soy serum for five minutes to bind with the remaining plastic not bound by antigens A’. Excess serum was removed, and the rabbit antibodies of species B were added to the cup to bind with antigens A’ for two days. Excess antibodies were removed, and the cup was rinsed with soy serum again. Radioactively tagged Iodine-125 goat antibodies produced against rabbit gamma globulin (GARGG) were then added to the cup for one day to bind with the antibody B-antigen A’ complexes 2.2 cm); h. exterior (s.1. = 6.8 cm); i. Macrocallista nimbosa (s.1. = 8.6 cm, present. The cup was then rinsed with water, dried, placed inside a radiation counter tube, and its radioactivity measured. Additionally, for each set of antibodies, control cups without antigens were treated as above to determine the levels on nonspecific binding of the antibodies to the soy rinse serum. The radioactive count of an empty radiation counter tube represented the radioactive standard, the level of background radioactivity. The resulting matrix of raw data was adjusted by first subtracting the nonspecific binding levels of the controls from tests involving their respective antibodies. The matrix was then divided by the radioactive standard, and the resulting quotients were expressed as percentages. The immuno- distance (ID) for each reaction was calculated as follows: IDa’p = 100 logy (A’A/A’B); IDap’ = 100 logy) (B’B/B’A); IDag = 1/2 UDap + IDap)); where A’A, for example, is the reaction of the antigens (A) with the antibodies (A’) of species A. A Fitch-Margoliash unrooted tree of the taxa was calculated from the resulting lower diagonal ID matrix using the program PHYLIP 3.1. 202 AMER. MALAC. BULL. 9(2) (1992) RESULTS Results of the reactivity tests indicated that protein con- tent varied noticeably among individual shells from the same species and even from the same lot, with some shells often not having adequate amounts of protein. Unknown and dif- fering concentrations of antigens and antibodies do not in- terfere with interpretation of the resulting immunological distances, however, because of the nature of the immuno- logical reaction, the experimental design, and the im- munological distance algorithm. In RIA, the immunological reaction is a two step process: a plastic substrate is coated with antigen, which is subsequently exposed to reaction by antibodies. Because the binding substrate for antigens has a high affinity and limited capacity for antigens, the first step is a saturation reaction that requires little dissolved protein (i.e. antigen) to saturate all sites. Thus, different protein con- centrations still result in saturation of binding sites. (This ex- plains the fact that in preliminary trials, some antigen solu- tions ultimately yielded functioning antibodies, despite previous reactivity tests indicating that these solutions had practically no proteins present). The binding of the antibodies to antigens is an equilibrium reaction, however, so the amount of binding is proportional to antibody concentration. For dif- ferent antibody concentrations, this is compensated by the counter reactions, which are incorporated as proportions in- to the immunological distance algorithm. Consider the ex- ample of antigens of species A reacting against two concen- trations of antibodies from species B: weak (B’B = 20, B’A = 4), and strong (B’B = 40, B’A = 8). The resulting IDa’z remains unchanged because, as explained above, the concen- trations of antigens A and B do not affect their reactions with the antibody, A’. The resulting IDap: also remains unchanged, whether the strong or weak antibodies are used, because the B’B/B’A term is equivalent in both cases (20/4 = 40/8 = 5). Thus, ID ag remains unchanged, regardless of the concentration of antibody used. While shells recently separated from their animals (i.e., within days) exhibited adequate amounts of protein, there were no other clearcut predictors of adequate protein levels. Beachworn specimens often did not have adequate amounts of protein. Adequate protein solutions of all the above species were obtained, however, and antibodies were made against them. Results are presented in Table | and figure 2. The matrix indicates that a deep division exists between Chioninae and Pitarinae, represented by NI, although NI is allied with the chionine taxa through TM, with the distance between NI and TM, 42, comparable to other distances among the chionine taxa (e.g. distances between TM and MM, and CS and MM). The unrooted Fitch-Margoliash tree groups the six chionine taxa in three pairs: AB and TM, CC and MM, PR and CS. All three pairs are conchologically dissimilar; the second pair are west Atlantic species and the third pair cc MM PR NI Fig. 2. An uprooted Fitch-Margoliash tree derived from Immunological Distance data on venerid taxa (abbreviations in text) (S. D. = 17.3, S. S. = 1.2). are west Pacific species. Using CC as a point of reference, the tree indicates parallel evolution of either the cancellate sculpture of Chione twice in CS and TM or parallel evolu- tion of the rostrate posterior and primarily concentric sculpture of Anomalocardia once in PR. Because no data are available on ancestral species, neither possibility assumes priority or an arbitrary direction. The matrix indicates that AB and TM are most closely allied (25), followed by PR and CS (39), and CM and MM (43). The matrix data agree well with the anatomical analyses of Jones (1979). DISCUSSION Much has been observed about bivalve adaptations (e.g. Carter, 1968; Stanley, 1970; Seilacher, 1974; Savazzi, 1985). The burrowing paradigm of Seilacher (1974) required that valve sculpture be perpendicular to the direction of bur- rowing, asymmetrical in crossection, and reduced medially (perimeter smoothening). Later experimentation and obser- vations have supported this paradigm (Stanley, 1977b; Stanley, 1981; Savazzi, 1985). A clam burrows anteriorly, and assumes a life position with the posterior closest to the sediment sur- face. From this it is logical to assume that the anterior will facilitate burrowing and anchorage. The posterior, especial- ly of shallow burrowers, coming into contact with the substratum only towards the end of burrowing, can contribute little towards it, and in cardiids posterior sculpture often does not conform to the burrowing paradigm (Savazzi, 1985). Be- ing the point closest to the surface and predators, however, the posterior probably functions more towards predatory defense and reducing surface scour of sediment around the shell, thereby preventing disinterment. These are useful perspectives for analyzing venerid adaptations. Most chionine clams burrow sluggishly and shallow- HARTE: BIVALVE EVOLUTION — A NEW APPROACH 203 Table 1. Immunological distances among various venerid taxa. Chione Timoclea Mercenaria Chione Anomalocardia Anamalocardia Macrocallista stutchburii marica mercenaria cancellata brasiliana producta nimbosa Group 1 Chione stutchburii Timoclea marica 55 Mercenaria mercenaria 46 43 Chione cancellata 59 106 34 Group 2 Anomalocardia brasiliana 28 25 27 39 A. producta 37 77 77 103 56 Outgroups =-Macrocallista nimbosa 128 42 134 163 87 112 ly, with the posterior tip positioned within | cm of the sedi- ment surface (Stanley, 1970). The shells are moderately thick, prosogyrous, and subovate with a slightly angular posterior; most have strong valve ornamentation. In each species, the unique set of variations among these characters reflects a unique balance and compromise of adaptive strategies. This is illustrated in the above results, which not only indicate con- chological parallelisms among west Pacific and North American chionine clams, but an extensive conchological diversification between closely related species. The latter is especially well illustrated between Mercenaria mercenaria and Chione cancellata. M. mercenaria is a large, thick shelled, moderately rapid bur- rower (Stanley, 1970) of subdued, predominantly concentric sculpture. Such sculpture aids burrowing, while size and thickness help keep it anchored in the sediment (Kauffman, 1969). The clam adjusts burrowing depth (1-2 cm between posterior and sediment surface) and life position to sediment type, and inhabits an unusually broad range of environmental conditions (Stanley, 1970); this ability to adapt to sediment changes probably accounts at least partly for its wide exploita- tion of habitats. In contrast, C. cancellata is a small, thick shelled, slow burrower (Stanley, 1970) with well spaced, sharp, concentric lamellae, slightly corrugated from underly- ing, well spaced radial ribs. Its life position is with the posterior near or at the surface, and its habitat is comparative- ly restricted (Stanley, 1970). Stanley (1981) showed that the strong sculpture inhibited burrowing. Furthermore, a rough comparison of the ornamentation and burrowing rate indices of these and other Carribean Chione species (Stanley, 1970) indicates that burrowing rate is inversely correlated to the development of the concentric sculpture, a relationship sup- ported on a broader scale throughout bivalvia (Kauffman, 1969). The strong sculpture of C. cancellata compensates posteriorly by reducing scour, however (Stanley, 1981); once interred, the sharp lamellae probably aid anchorage and discourage nipping of the short siphons or firm grippage by predators. Radial ribs not only confer rigidity to the shell (Kauffman, 1969) but in this case evidently strengthen the lamellae basally and the resulting corrugations of the lamellae (Stanley, 1981). Lack of life history data precludes comprehensive com- parisons of conchological features and life strategies for the pair of west Pacific chionines, Chione stutchburii and Anomalocardia producta. While they have different profiles and concentric sculpture, they have similar radial patterns. Radial ribs are virtually absent anteriorly, appearing faintly medially and predominating the posterior sculpture. Posterior radial ribbing offers an acceptable approximation to being perpendicularly oriented to water currents, thereby reducing surface scour (Savazzi, 1985), which might explain its presence in both species, although not as a parallel adapta- tion but as a commonly derived one. The concentric sculpture in both species is more subdued than that of C. cancellata, indicating that it functions in them more as a burrowing aid than anchor. For the more streamlined species, C. stutchburii, it could enable more rapid burrowing, enabling exploitation of less stable habitats; indeed, the species inhabits a wide variety of sediments, ranging from less stable sediments of sand and gravel to a more stable muddy sand within estuaries and enclosed bays (Beu and Maxwell, 1990). For A. producta, concentric sculpture might compensate at least partly for its lack of streamlining. The third linked pair of chionines, Timoclea marica and Anomalocardia brasiliana, are the most closely linked of the pairs immunologically, yet exhibit wide disparities in profile, sculpture and geography. Their different adaptive pathways exhibit interesting parallels. Both live near the sur- face, are similar in size and probably equally slow infaunal burrowers. 7. marica has sculpture closely similar to Chione cancellata: well developed lamellae (though not as widely spaced) that flare posteriorly, corrugated by well separated radial ribs. A plausible inferrence is that it functions similarly, inhibiting burrowing but serving as an anchor, an anti-scour mechanism posteriorly, and to discourage predators. Similar functions in Anomalocardia brasiliana probab- ly are accomplished through its unusual profile rather than its relatively subdued sculpture, which aids burrowing. A. brasiliana and Chione cancellata are equally slower bur- rowers (Stanley, 1970); the blunt, relatively obese anterior pro- file probably inhibits burrowing. The compensations, however, are several. Angular to rostrate posteriors serve to 204 AMER. MALAC. BULL. 9(2) (1992) elevate the siphonal flow with a minimum of shell excretion while permitting the center of gravity to remain relatively deep, although streamlining is sacrificed (Stanley, 1970). The rostrate posterior maximizes this effect. The life position in the sediment is perpendicular to the sediment surface (Stanley, 1970), and this minimizes the posterior surface vulnerable to scouring and predators. Such advantages could result in parallel selection for this trait, and, indeed, this trait appears in several subfamilies of Veneridae [e.g. Lepidocardia floridella (Gray), Pitarinae; Eumarcia paupercula (Holten), Tapetinae; Timoclea malonei (Vanatta) and T. peresi Fischer- Piette and Vukadinovic, Chioninae]. Another difference between Timoclea marica and Anomalocardia producta is that the posterior dorsal margins of T. marica are finely crenulated, while those of A. producta slightly interlock by means of a long, shallow fold in the right valve. Hypotheses on defensive functions of crenulated margins include increasing resistance of the shell to compres- sion from shell-crushing predators (Waller, 1969), restricting predatory access of starfish, and creating a tight seal (Carter, 1968), thereby preventing release of diagnostic chemicals into the environment, and increasing survival times in a predator’s digestive tract (Vermeij, 1987). Restricting predatory access and creating a tight seal might be effected equally by marginal folding, and function similarly. Jones (1979) observed, for ex- ample, that marginal folding effectively keeps the posterior dorsal margin closed while siphons are extended, and sug- gests that the resulting marginal overlaps might deter polydorid polychaete pests. Additionally, both marginal in- terdigitation and folding can thicken the marginal juncture, discouraging boring predators. Which adaptation is ultimately chosen might depend ultimately on slight differences in ontogeny and environment. The data indicate, then, that Timoclea marica and Anomalocardia brasiliana are closely related species ex- hibiting different sculpture, profiles and posterior margins that function similarly in aiding anchorage and discouraging predators. In both species, efficiency in burrowing is sacrificed for advantages in anchorage and _ predatory discouragement. The Fitch-Margoliash tree indicated parallel evolution of either rostrate posteriors and their accompanying characteristics, or cancellate sculpture. Arguments for the former are advanced, above. Parallel adaptation of the cancellate sculpture characteristic of Chione and Timoclea could occur in various species because of the cumulative ad- vantages offered by ribbing, and the ontogenetic ease with which strong, well spaced concentric sculpture can be modified into structures that aid anchorage and defense (lamellae) or burrowing (cords or ridges), facilitating evolu- tion into different life strategies. Indeed, this transition can be seen within several species of venerids, such as Mercenaria mercenaria, where juveniles, more vulnerable to disinter- ment, have widely spaced low, anchoring, concentric lamellae (Pratt and Campbell, 1956) that gradually become closely spaced, more subdued threads, medially worn smooth in adults. The RIA data have several ramifications for venerid systematics, which, for the above chionines, is beset with definitional problems. As developed by Keen (1969), the genus Anomalocardia is paradoxical: except for Anomalocardia s. §., the concentric sculpture of the included subgenera are concentric cords, not the undulating folds given in the generic definition (Fig. la-b). The result is a de facto assemblage of chionine species with rostrate posteriors and primarily con- centric sculpture. Immunological distances indicate the features are a parallel adaptation, although the tree allows an interpretation of them as a derived adaptation. Besides having different concentric sculpture, Anomalocardia producta differs from A. brasiliana in lack- ing a rugose nymph and crenulated margins, and in having pronounced posterior radial ribs. While not as obvious a trait as a rostrate posterior, the rugose nymph of A. brasiliana is a distinct marker for linking the taxon conchologically to im- munological allies, such as Mercenaria. Matrix data support the morphological transition pro- posed by Harte (1992) linking Anomalocardia s. s. to Mercenaria and two subgenera of Chione, Lirophora and Iliochione, and Mercenaria. Besides a rugose nymph, present at various degrees in all four taxa, specific traits include con- centric undulations as the predominant sculpture, with a posterior that ranges from angular (Mercenaria) to rostrate (lliochione, some Lirophora and Anomalocardia). Several systematic changes are indicated by the above data. Matrix data indicate that Cryptonomella is more close- ly related to Chione (Austrovenus) stutchburii than Anomalocardia; it should stand as a separate genus until it can be demonstrated to be closely linked to a senior genus. C. (A.) stutchburii is more closely allied to Anomalocardia and Mercenaria than to Chione and should stand as a separate genus until its intergeneric relationships are further clarified. The relationships of the other subgenera of Chione (Chionista Keen, 1958; Chionopsis Olsson, 1932; Panchione Olsson, 1964 and Securella Parker, 1949) should be clarified before deciding the systematic relationships of Chione to other chionine taxa. The immunological distances between Macrocallista nimbosa and most of the chionine taxa indicate a deep divi- sion within the family, an observation of systematic workers, as well. With some exceptions, venerid taxa fall roughly into two groups: 1) those with well developed anterior lateral teeth, and simple valve sculpture (Pitarinae, et al.); 2) those with little or no anterior lateral teeth, and often strong valve sculpture (Chioninae, ef al.). Although venerid classification has been in a state of constant flux, some past workers have tended to sometimes merge existing subfamilies within these HARTE: BIVALVE EVOLUTION — A NEW APPROACH 205 two groups (Fischer, 1887; Dall, 1902; Jukes-Browne, 1914). In contrast, Frizzell (1936) elevated the ten subfamilies that existed by then into separate families. Keen (1969) returned them to subfamilies, and the relatively close alliance of M. nimbosa to Timoclea indicates that the family is monophyletic. CONCLUSIONS Radio-immuno-assay has been used successfully in conjunction with conchological and anatomical analyses to help illustrate various evolutionary processes within Veneridae. RIA and conchological analyses of Pacific and Atlantic Ocean chionines indicate that between distantly related species there exists one or more strong conchological parallelisms, which have as their bases adaptations towards predatory defense, positional stability, and burrowing efficiency. RIA analyses also indicate that closely related species have undergone extensive conchological diversification. This diversification probably results from adaptation to different lifestyles, or utilizing different adaptations for similar func- tions. The conchological differences between Mercenaria mercenaria and Chione cancellata reflect different life strategies for predatory defense and positional stability. M. mercenaria appears to rely on burrowing abilities, size and thickness, while C. cancellata relies on sculpture, thickness, and a more specialized habitat. Conchological differences between Timoclea marica and Anomalocardia brasiliana, in contrast, reflect different conchological strategies that effect parallel adaptive compromises, and cope with parallel lifestyles. Ramifications for systematics include separating Cryp- tonomella from Anomalocardia, separating Austrovenus from Chione, and reassessing the systematic relationships of Anomalocardia, Timoclea, Mercenaria and Chione. A deep division exists between the subfamilies Pitarinae and Chioninae, although a single, relatively close link between them indicates that the family is monophyletic. Geographic and anatomical data, and obscure conchological traits can sometimes indicate genetic alliances, and should be utilized in assessing systematic relationships. RIA has the potential for proving examples of parallel evolution between extant bivalve taxa and their ancestors. Further research utilizing fossil taxa is planned. ACKNOWLEDGMENTS I gratefully acknowledge the technical assistance of Dr. Jerold Lowen- stein and Gary Scheuenstuhl. Materials for this study were donated by the Museum of Paleontology, University of California at Berkeley, and the Museum of Comparative Zoology, Cambridge. I thank Dr. John Harte, Dr. David Lindberg, and Dr. Jerold Lowenstein and two anonymous reviewers for their constructive advice and criticism. This research was supported under NSF grant BSR-9009189, through the Rocky Mountain Biological Laboratory. LITERATURE CITED Bernard, F. R. 1982. Nutricola n. gen. for Transennella tantilla (Gould) from the northeastern Pacific (Bivalvia: Veneridae). Venus (Japanese Journal of Malacology) 41:146-149. Beu, A. G. and P. A. Maxwell. 1990. Cenozoic Mollusca of New Zealand. New Zealand Geological Survey, Bulletin No. 58, 518 pp. Bowman, B. H. 1989. Non-clocklike evolution in the Ribosomal RNAs of bivalve molluscs. Doctoral Dissertation. Department of Biochemistry, University of California at Berkeley. 245 pp. Carter, R. M. 1968. On the biology and paleontology of some predators of bivalved Mollusca. Palaeogeography, Palaeoclimatography, Palaeoecology 4:29-65. Casey, R. 1952. Some genera and subgenera, mainly new, of Mesozoic hetero- dont lamellibranchs. Proceedings of the Malacological Society of Lon- don 29:121-176. Collins, M. J., G. B. Curry, G. Muyzer, P. Westbroek, T. Zomerdijk and R. Quinn. 1988. Serotaxonomy of skeletal macromolecules in living terebratulid brachiopods. Jn: Immunological Approaches in Geological Research, G. Muyzer, ed. pp. 41-59. Krips repro, Meppel. Dall, W. H. 1902. Synopsis of the family Veneridae and of the North American Recent species. Proceedings of the U. S. National Museum 26:335-411. De Queiroz, K. 1988. Systematics and the darwinian revolution. Philosophy of Science 55:259-283. Fischer, P. 1887. Manuel de Conchyliologie et de Paleontologie Conchyli- ologique. Paris. 1,008 pp. Fischer-Piette, E. 1975. Revision de Venerinae (Lollusques Lamellibranches). Memoires du Museum National d'Histoire naturelle, Nouvelle Serie, Serie A, Zoologie 93:1-64. Fischer-Piette, E. and D. Delmas. 1967. Revision des mollusques Lamelli- branches du genre Dosinia Scopoli. Memoires du Museum National d Histoire naturelle, Nouvelle Serie, Serie A, Zoologie 47A:1-91. Fischer-Piette, E. and D. Vukadinovic. 1975. Revision de Circinae. Journal de Conchyliologie 112(1-2):3-74. Fischer-Piette, E. and D. Vukadinovic. 1977. Suite de revisions des Veneridae (Mollusques Lamellibranches) Chioninae, Samarangiinae et comple- ment aux Venus. Memoires du Museum National d'Histoire naturelle, Nouvelle Serie, Serie A, Zoologie 106:1-186. Frizzell, D. L. 1936. Preliminary reclassification of veneracean pelecypods. Bulletin of the Belgium Royal Museum of Natural History 5:1-84. Harte, M. E. 1992. An eastern Pacific Mercenaria and notes on other chionine genera (Bivalvia: Veneridae). Veliger 35:137-140. Jones, C. C. 1979. Anatomy of Chione cancellata and some other chionines (Bivalvia: Veneridae). Malacologia 19:157-199. Jope, M. 1980. Phylogenetic information derivable from fossil brachiopods. In: Biogeochemistry of Amino Acids, P. E. Hare, T. C. Hoering, and K. King, Jr., eds. pp. 83-94. John Wiley & Sons, New York. Jukes-Browne, A. J. 1914. A synopsis of the family Veneridae. Parts I and II. Proceedings of the Malacological Society, London \1:58-94. Kauffman, E. G. 1969. Form, function, and evolution. Jn: Treatise on In- vertebrate Paleontology. R. C. Moore, ed. pp. N129-205. Geological Society of America and University of Kansas Press: Lawrence, Kansas. Keen, A. M. 1969. Veneridae. Jn: Treatise of Invertebrate Paleontology, R. C. Moore, ed. pp. N671- N688. Geological Society of America and University of Kansas. Lawrence. Lindberg, D. 1989. Transennella Dall versus Nutricola Bernard (Bivalvia: Veneridae): an argument for evolutionary systematics. Journal of Molluscan Studies 56:129-132. Lowenstein, J. M. 1981. Immunological reactions from fossil material. Philosophical Transactions of the Royal Society of London B292:143. Lowenstein, J. M. 1985. Radioimmunoassay of extinct and extant species. In: Hominid Evolution: Past, Present, and Future. P. V. Tobias, ed. pp. 401-410. Alan R. Liss, Inc., New York. 206 AMER. MALAC. BULL. 9(2) (1992) Lowenstein, J. M., T. Molleson and S. L. Washburn. 1982. Piltdown jaw confirmed as orangutan. Nature 299:294. Lowenstein, J. M., W. M. Rainey and J. L. Betancourt. 1991. Immunospecific albumin in fossil pack rat, porcupine and hyrax urine. Natur- wissenschaften 78:26-27. Lowenstein, J. M. and O. A. Ryder. 1985. Immunological systematics of the extinct quagga (Equidae). Experientia 41:1192-1193. Lowenstein, J. M., V. M. Sarich and B. J. Richardson. 1981. Albumin systematics of the extinct mammoth and Tasmanian wolf. Nature 291:409-411. Lowenstein, J. M. and G. Scheuenstuhl. 1991. Immunological methods in molecular paleontology. Philosophical Transactions of the Royal Society, London 333:375-380. Luft, R. and R. S. Yalow. eds. 1974. Radioimmunoassay: Methodology and Applications in Physiology and in Clinical Studies. Georg Thieme, Stuttgart. 195 pp. Molleson, T. 1982. Shrunken heads — but whose? New Scientist 96:318. Olsen-Stojkovich, J., J. A. West and J. M. Lowenstein. 1986. Phylogenetics and biogeography in the Cladophorales complex (Chlorophyta): some insights from immunological distance data. Botanic Marina 29:239-249. Olsson, A. A. 1932. Contribution to the Tertiary Paleonotology of northern Peru. V. Miocene Mollusca. Veneracea. Bulletin of American Paleon- tology 19:120-121. Parker, P. 1949. Fossil and Recent species of the pelecypod genera Chione and Securella from the Pacific Coast. Journal of Paleontology 23:577-593. Pratt, D. M. and D. A. Campbell. 1956. Environmental factors affecting growth in Venus mercenaria. Limnology and Oceanography 1:2-17. Price, R. A., J. Olsen-Stojkovich and J. M. Lowenstein. 1987. Relationships among the genera of Pinaceae: an immunological comparison. Systematic Botany 12:91-97. Rainey, W. E., J. M. Lowenstein, V. M. Sarich and D. M. Magor. 1984. Sirenian molecular systematics — including the extinct Steller’s sea cow (Hydrodamalis gigas). Naturwissenschaften 71:586-588. Savazzi, E. 1985. Adaptive themes in cardiid bivalves. Neues Jahrbuch fur Geologie und Palaontologie Abhandlungen 170:291-321. Seilacher, A. 1974. Fabricational noise in adaptive morphology. Systematic Zoology 22:451-465. Shimamoto, M. 1986. Shell microstructure of the Veneridae (Bivalvia) and its phylogenetic implications. Tohoku University, Science Reports, Second Series (Geology) 56:1-39. Sibley, C. G. and J. E. Ahlquist. 1991. Phylogeny and Classification of Birds. Yale University Press, New Haven. 1,056 pp. Stanley, S. M. 1970. Relation of shell form to life habits of the Bivalvia (Mollusca). Geological Society of America, Memoirs 125:1-296. Stanely, S. M. 1974. Effects of competition on rates of evolution, with special reference to bivalve mollusks and mammals. Systematic Zoology 22:486-506. Stanley, S. M. 1977a. Trends, rates and patterns of evolution in the Bivalvia. Developments in Palaeontology and Stratigraphy 5:210-250. Stanley, S. M. 1977b. Coadaptation in the Trigoniidae, a remarkable family of burrowing bivalves. Palaeontology 20:869-899. Stanley, S. M. 1981. Infaunal survival: alternative function of shell ornamen- tation in the Bivalvia (Mollusca). Paleobiology 7(3):384-393. Vermeij, G. J. 1987. Evolution and Escalation: An Ecological History of Life. Princeton University Press, Princeton. 527 pp. Waller, T. R. 1969. The evolution of the Argopecten gibbus stock (Mollusca: Bivalvia) with emphasis on the tertiary and Quaternary species of eastern North America. Journal of Paleontology, Supplement to No. 5, 43:1-125. Westbroek, P., P. H. van de Meide, J. S. van der Wey-Kloppers, R. J. van der Sluis, J. W. de Leeuw and E. W. de Jong. 1979. Fossil macromolecules from cephalopd shells; characterization, im- munological response and diagenesis. Paleobiology 5:151-167. Date of manuscript acceptance: 25 November 1991 Research on marine bivalves in the People’s Republic of China Zhuang Qiqian Fujian Institute of Oceanology, 34 Hai Shan Road, Xiamen, Peoples Republic of China Abstract. This paper summarizes the published literature on marine bivalves in the People’s Republic of China from 1950 to 1991. The 221 citations listed in the bibliography contain only significant works as viewed by the author. The summary includes systematic research on 23 families of bivalves distributed in the Chinese seas and encompasses the Arcidae, Glycymerididae, Limopsidae, Mytilidae, Pinnidae, Pteriidae, Isognomonidae, Pectinidae, Plicatulidae, Ostreidae, Chamidae, Cardiidae, Tridacnidae, Mactridae, Mesodesmatidae, Tellinidae, Solenidae, Veneridae, Myidae, Corbulidae, Pholadidae, Teredinidae and Laternulidae. The breeding biology and experimental ecology of 12 economically important bivalves is addressed for the following species; Crassostrea rivularis (Gould), Crassostrea gigas (Thunberg), Mytilus edulis (Linnaeus), Perna viridis (Linnaeus), Sinonvacula constricta (Lamarck), Tegillarca granosa (Linnaeus), Chlamys farreri (Jones and Preston), Chlamys nobilis (Reeve), Argopecten irradians (Lamarck), Ruditapes philippinarum (Adams and Reeve), Pinctada martensii (Dunker), and Pinctada maxima (Jameson). China is a large coastal country, with 18,000 kilometers of coastline, a vast expanse of seas, and over 14,000 kilometers of islandic coastline. There are four seas, the Bohai Sea, the Yellow Sea, the East China Sea and the South China Sea, with an area of 4.75 million square kilometers, including more than 6,000 islands. Marine bivalve resources are very rich throughout the region. Three marine molluscan faunal provinces for China seas have been identified: 1) a warm temperature region that includes the Yellow Sea and Bohai Sea; 2) a subtropical region that includes the East China Sea, the north-western coast of Taiwan and the northern coast of Hainan; 3) a tropical region including the south-eastern coast of Taiwan, the coast of the southern tip of Hainen Island and the area south of them. The suffix to this paper offers more than 200 citations concerning Chinese marine bivalves. This obviously does not include the entire Chinese bivalve literature but rather a cir- cumspect summary of the significant works. The majority of the papers were written in Chinese, with English abstracts or with English headings only, though a small number were written in English. There have also been several articles that were published in general magazines, but are not included in this summary. SUMMARY OF MARINE BIVALVE RESEARCH IN CHINA 1. Systematics and Biogeography In China, the study of marine bivalve taxonomy started much later than in western Europe. Before October 1949 (The Liberation), foreign and domestic scientists had published few comprehensive systematic works. However, between 1950-1990, 40 years were dedicated to intensive sampling and investigation along China’s coastal seas, that is, starting from the most northern sea, the Yalu River Estuary, to the southern most area, the Xisha Islands. Between 1958-1960, a nation- wide general exploration was completed in the coastal seas; from 1959 to 1962 an investigation on the Beibu Gulf of the South China Sea was completed. Owing to the extensive number of sampling stations and frequency of seasonal in- vestigation, a large number of specimens were obtained. These specimens have been identified and are housed cur- rently in the Institute of Oceanology, Academia Sinica. Thanks to the extended collecting plan, we have been able to report on the distribution of Chinese coastal and near sea bivalves and summarize the zoogeography of the region. From 1980-1984, under the leadership of Chinese National Bureau of Oceanography, these investigations were completed. The above work was done by relevant universities of the coastal provinces, each with a mission to collect and preserve large numbers of samples. The investigation of the coastal regions was also combined with sublittoral zone research (0 to 5 m). Further, since 1990, the Chinese National Bureau of Oceanography has been organizing the coastal provinces to launch large-scale islandic investigations. Further study is anticipated on the coastal resources, followed by a program of marine bivalve resource development and utilization. Marine bivalves hold an important position among marine molluscs, due to their use as a human food and as feed for domestic fowl and prawns, and thus they are viewed as a key resource in the national economy. Our research is based primarily on bivalve species that can be developed for American Malacological Bulletin, Vol. 9(2) (1992):207-215 207 208 AMER. MALAC. BULL. 9(2) (1992) a market economy or have some economic impact, such as members of the Ostreidae and Teredinidae. Many articles have been published in China about the breeding, growth and ecology of Crassostrea rivularis (Gould, 1861), and Teredo navalis Linnaeus, 1758, resulting in enhanced oyster breeding and a shipworm prevention and treatment program. Additional research on marine fisheries, physical oceanography, and pollution has yielded large numbers of specimens that can be used for detailed studies of the systematics, ecology and zoogeography of bivalves. On the basis of these scientific materials, taxonomists have published many monographs and treatises on a variety of families and genera, and have discovered and described many new genera and new species. As basic survey work continues, taxonomists continue to engage in systematic research on the families that have economic value, such as the Arcidae, Mytilidae, Pin- nidae, Pectinidae, Mactridae, Veneridae, Tellinidae, Pholadidae, Corbulidae, Laternulidae and Myidae. 2. Biology and Ecology The biology and ecology of intertidal and infaunal benthic bivalves has been superficially studied. Naturally, there has been some analysis of the variation in mollusc species composition and density. The ecology and biology of local species of Donax, Moerella and Cultellus has been studied, and the population structure of Perna viridis (Linnaeus, 1758) has been examined in several regions. General research on benthic community structure in three inner bays in the Bohai Sea, has yielded data on the relatively high densities of Scapharca subcrenata (Lischke, 1869); Chlamys farreri (Jones and Preston, 1904), Atrina pectinata (Linnaeus, 1758), and Paphia undulata (Born, 1778), suggesting possible sites to develop and utilize in the future. 3. Experimental ecology and mariculture of economically important species Oysters. The literature on Ostreidae breeding far exceeds that of other families. The primary species utilized in China are Crassostrea rivularis and C. gigas (Thunberg, 1793). C. rivularis is common in the Pearl River Estuary, and is the primary species harvested in Guangdong Province. It is also found in lower densities in the Fujian Province. Results of research on the feeding habits, breeding seasons, predators, and artificial rearing of larvae are included in the literature reviewed below. Beside the traditional cultivation method of “throwing stones’’, a new raft culture technique has produced high yields. C. gigas is a well known species in the Fujian Province, and reports on its culture were recorded in Xiapu County in the early 16th Century. The annual harvest of this species in 1986 was 44,200 tons. Mussels. There are three species of economically important mussels in China: the Purple mussel, Mytilus edulis Linnaeus, 1758; the Hardshell mussel, M. crassitesta Lischke, 1868; and the Jade mussel, Perna viridis. M. edulis is distributed in north China, M. crassitesta is distributed in Liaoning, Shangdong and Zhejiang Provinces, but P. viridis is found only in Fujian and Guangdong Provinces. Cultivation of marine mussels began in China in 1958, at which time studies were undertaken to study mytilid life history, feeding habits, sexual maturation, growth, meat con- dition, method of culture, development of larvae, and artificial rearing of spat. Mussel rearing conditions, feeding re- quirements and treatment of larval culture water has been reported in numerous publications. Artificial spat rearing of these three mussel species has been successful. The primary mussel breeding technique in China utilizes hanging rafts. Due to the large acreage devoted to culture, juveniles are usually collected from sea-weed rafts. New types of ecologically sound breeding programs have been tested, such as combined mussel-seaweed culture, where the nitrogenous waste products of mussels are utilized as fertilizer for the sea-weed, and in turn the growth of the seaweed im- proves the mussel culture. Chinese Razor Clam. Sinonovacula constricta (Lamarck, 1818) lives in coastal waters with lower salinity than usually found in the northern and southern mid-littoral zones, however, artificial breeding of razor clams has been successful in the Zhejiang and Fujian Provinces. Traditional culture techniques have been practiced in those areas where the juvenile clams naturally settle, allowing easy collection, redistribution and harvest. By the end of the 1950’s, malacologists had examined the reproductive biology of razor clams. Analysis of these data has provided an effective forecast system for rake flat productivity, and has allowed the establishment of harvest quotas. In this way, the number of clams harvested has been increasing. From 1970-1980, malacologists experimented with natural juvenile collection. These studies included monitoring temperature, salinity, the relationship between rate of survival of juvenile clams and growth, as well as the influence of the environment on the rate of razor clam incubation and the rate of gonad maturation. Ark Clams. Tegillarca granosa (Linnaeus, 1758) is widely distributed along the Chinese coast, with artificial breeding mainly occurring in the Zhejiang, Fujian and Guangdong Provinces. Traditionally, the Shangdong Province supplies juvenile bloodclams that are then dispersed in the mid-lower intertidal zone and fed until they reach a preset size, after which they are spread, taking 1-2 years to reach marketable body weight (500 gms/80 shells). A production of 750 - 4,500 kg/mu is an optimistic yield. Extensive study on this species has been neglected and further investigations are desirable. Scallops. Three species within the Pectinidae are commercially ZHUANG: CHINESE BIVALVE LITERATURE 209 exploited in China, Chlamys farreri (Jones and Preston, 1904), C. nobilis (Reeve, 1852) and Argopecten irradians (Lamarck, 1819). C. farreri is distributed throughout the Yellow Sea and Bohai Sea, as well as to the south of the East China Sea, along the eastern Zhejiang coastal sea. C. nobilis is mainly found in the coastal area of the Fujian and Guangdong Provinces, and A. irradians has been introduced from the United States. The adductor muscle of Chlamys farreri is the only tissue used in the production of dried scallop muscle, ganbei. In the mid-1950’s, the famous Chinese malacologist, Tchang Si, developed methods for studying the feeding and growth of scallops, and provided a good foundation for the culture of this species. In 1974, Chinese scientists began to collect natural scallop spat and successfully experimented with arti- ficial feeding and cultural production of C. farreri. In 1979, artificial culture experiments produced 0.92 million spat/m?. Many research laboratories in North China have continued investigations of reproductive cycles, natural spat collection, artificial breeding and the study of an immense standard seeding, which has pushed the scallop fishery to the verge of mass production. During this period, breeding methods were changed from rafting and hanging techniques to field spreading. Chlamys nobilis is an economically important species that has been cultivated in South China. Owing to the suc- cess of artificial breeding, cultivation has been modified to a method similar to the one used in mussel culture, that is, growing scallops on seaweed rafts. Argopecten irradians is an important species that originated from the Atlantic Ocean. This species grows rapid- ly and produces a marketable scallop (mean length of 5 cm) in one year. In 1982, Professor Zhang Fusui (Institute of Oceanology, Academia Sinica) introduced breeding stock of A. irradians and has had excellent results in North China, with a higher fecundity than usual. The species has gradual- ly spread down to northeast Fujian, creating a system of autumn artificial spat setting suitable to the thermal limita- tions in South China. Currently scientists are conducting a series of large scale experiments on the development of high density spat-rearing for this species. In 1984, Professor Zhang and his colleagues introduced the Japanese scallop, Patinopecten yessoensis (Jay, 1857) into China in hopes of utilizing this species for the production of dried scallop ad- ductor muscles (ganbei). Littleneck Clam or Flower Clam. Ruditapes philippinarum (Adams and Reeve, 1850) is distributed broadly along the Chinese coast. In the north, the intertidal and sub-tidal zones have large beds of this clam, however, this is primarily due to the presence of strict resource protection and planned gathering. Artificial culture has been seldom practiced. Nevertheless, the Fujian Province has an important project on the artificial breeding of the Littleneck clams, including research on reproduction mechanisms, spermatogenesis, in- fluences of chemicals and sexual products on hastening par- turition, life history, growth of spats, and feeding habits. Laboratory breeding success, combined with mass produc- tion needs, have facilitated experiments in outside earth ponds including egg production, larval cultivation, and mass field production. Pearl Oysters. The major species in China is Pinctada martensii (Dunker, 1850), which is used in pearl culture and is an important enterprise in Guangdong, Guangxi and Hainan Provinces. Research on this species has encompassed artificial breeding techniques including nucleus insertion, induced polyploidization, and observation of gonads during triploid- ization. Scientists have successfully completed inter-species hybridization of P. martensii, P. chemnitzii (Philippi, 1847) and P. maxima (Jameson, 1901), and have studied hybrid chromosomes and their zymograms. Many studies have been undertaken on P- maxima. Owing to its scarce distribution in nature, large shell, and good quality and high priced pearls, artificial breeding experiments have been completed. However, it must be noted that many hurdles still must be overcome to utilize this species. Along with the seven important bivalves mentioned above, research has begun on the breeding and ecology of other bivalves such as Coelomactra antiquata (Spengler, 1802), Meretrix meretrix (Linnaeus, 1758), Mactra veneri- formis Reeve, 1854, Musculus senhousei (Benson, 1842), Atrina pectinata (Linnaeus, 1758), Laternula marilina (Reeve, 1860), Saxidomus purpuratus (Sowerby, 1852), Scapharca broughtoni (Schrenck, 1867) and Potamocorbula laevis (Hinds, 1843). CONCLUSIONS The study of systematics has naturally focused on families and genera that are economically important, leav- ing ignored several important families without commercial potential. Limited by restricted resources, we have only described species along the Chinese coast. Most studies have been limited to the identification of species, morpholog- ical descriptions, geographical distributions and a discussion of nomenclatural problems. The systematic treatment of superfamily and subspecific levels and the evolutionary rela- tionships of bivalves have been inadequately studied. Moreover, most systematic research has been based chiefly on external morphology. The study of shell microstructure and functional morphology has been meager. Due to the over emphasis of research on commercial species and the limited funding in general, the study of bivalve ecology has been neglected, especially research on autecology and synecology. Much attention has been paid to bivalve 210 AMER. MALAC. BULL. 9(2) (1992) experimental ecology and breeding, but spat ecology needs further study. Facing the outside world and absorbing Euro- American advanced technology is required to enhance Chinese cultivation of marine bivalves. ACKNOWLEDGMENTS I thank Ms. Lin Huigiong who helped collect the literature and critical- ly read the manuscript. Paul Scott of the Santa Barbara Museum of Natural History and Robert Prezant of Indiana University of Pennsylvania provided assistance with editing and translating this article. Eugene Coan of the Los Angeles County Museum consulted on the cited literature. PRINCIPLE LITERATURE ON CHINESE MARINE BIVALVES 1950-1959 Tchang Si and Lou Tzekong. 1956. A study on Chinese oysters. Acta Zoologica Sinica 8(1):65-94. Tchang Si, Tsi Chung-yen et Li Kie-min. 1956. Recherches sur la reproduc- tion et la croissance d’un petoncle comestible Chlamys farreri (Jones et Preston). Acta Zoologica Sinica 8(2):235-253. Tchang Si, Tsi Chung-yen et Li Kie-min. 1958. Recherches sur les tarets des cotes du sud de la Chine. I. Acta Zoologica Sinica 10(3):242-257. Tchang Si and Lou Tzekong. 1959. Oyster. Scientific Publishing House. Beijing. 156 pp. Tchang Si, Tsi Chung-yen et Xie Yu-kan. 1959. Moeurs de s’alimenter chez V’Ostrea_ rivularis Gould. Oceanologia et Limnologica Sinica 2(3):163-179. 1960-1969 Cai Naner. 1963. Studies on the life history of Mytilus edulis Linnaeus. Studia Marina Sinica 4:81-94. Cai Yingya. 1966. A preliminary survey on Bivalvia from Fujian coast. Jour- nal of Zoology 2:76-80. Li Kie-min. 1965. Description d’ especes nouvelles et nouveaux records des taretes des cotes de la Chine. Studia Marina Sinica 8:1-7. Scarlato, Q, A. 1965. Studies on Tellinacea (Bivalvia, Mollusca) of China coasts. Studia Marina Sinica 8:27-114. Tchang Si, Tsi Chung-yen et Li Kie-min. 1960. Etude sur les Pholades de la Chine et description d’especes nouvelles. Acta Zoologica Sinica 12(1):63-87. Tchang Si and Hwang Hsuiming. 1964. On the Chinese species of Solenidae. Acta Zoologica Sinica 16(2):193-206. Wang Zhenrui. 1964. Preliminary studies on Chinese Pinnidae. Studia Marina Sinica 5:30-41. Xu Fengshan. 1964. Studies on the Cardiid molluscs of the China Seas. Studia Marina Sinica 6:82-98. Zhirmunsky, A. V. and Chu Li-chun. 1963. The cell thermostability of sym- patric species of Donax in relation to the temperature condition of their habitat. Acta Zoologica Sinica 15(1):21-27. Zhuang Qiqian. 1964. Studies on Chinese species of Veneridae (Class Lamellibranchiata). Studia Marina Sinica 5:43-106. 1970-1979 Marine Drug Group. Institute of Oceanology, Academia Sinica. 1976. In- habitory effect of the extract of the common clam Venerupis (Amygdala) philippinarum (Adams et Reeve) on the experimental tumor in rodents. Studia Marina Sinica \1:389-396. Nie Zhongging, Wang Zhongyuan, Niu Xiduan, Ji Meifang, Shen Juefen, Chen Wenhua and Liu Lihua. 1979. Artificial rearing of mussel (Mytilus edulis) larvae on a large scale. Acta Oceanologica Sinica 1(1):138-156. Station of Experimental Molluscan Ecology. Institute of Oceanology, Academia Sinica and Marine Aquaculture Experimental Station of Yan- tai District. 1977. A study on the artificial rearing of mussel spats. Scientia Sinica 20(2):244-255. Tang Chungti, Hsu Chentsu, Huang Munchie and Lu Sulian. 1975. Studies on the parasitic disease of Chinese razor clam (Sinonovacula con- stricta Lamarck) in the northern estuary of Chiulung River, Fujian. Journal of Xiamen University (Natural Science) 2:161-177. Tang Chungti and Xu Zhenzu. 1979. The ‘black root’’ disease of the razor clam in estuary of Jiulong River, Fujian. Acta Zoologica Sinica 25(4):336-346. Wang Zhenrui. 1978. A study of the Chinese species of the family Pteriidae (Mollusca). Studia Marina Sinica 14:101-11S. Zhuang Qigqian. 1978. The Tridacnids of the Xisha Island, Guangdong Province, China. Studia Marina Sinica 12:133-139. Zhuang Qigqian. 1978. Studies on the Mesodesmatidae (Lamellibranchia) off the Chinese coast. Studia Marina Sinica 14:69-73. 1980-1989 Cai Lizhe and Li Fuxue. 1988. On several fundamental allometries of the green mussel Perna viridis. In: Proceeding on Marine Biology of the South China Sea. pp. 93-102. China Ocean Press. Cai Yingya and Zhuang Qiqian. 1985. A new species of Psammobiidae (Mollusca, Bivalvia). Tropic Oceanology 4(3):64-66. Cai Yingya, Zhang Zhiqiang. 1986. Reproduction of Sanguinolaria violacea Lamarck. Marine Science 5:32-34. Chen Boyun. 1985. A preliminary study on the embryonic development and food for larvae of Solen roseomaculata. Marine Science Bulletin 4(6):40-44. Chen Jindi, Shen Guoying and Zhuang Chunquan. 1984. Preliminary studies on the use of cyanide in the cultivation of the razor clam. Journal of Xiamen University (Natural Science) 23(2):217-223. Chen Jindi and Xu Zhenzu. 1989. Effects of Zn? + on embryonic and larval developments of Sinonovacula constricta and detoxification of EDTA. Journal of Oceanography in Taiwan Strait 8(2):150-155. Chen Jindi. 1989. Effects of Cu2+ on larval development of Sinonovacula constricta and detoxification of EDTA. Journal of Oceanography of Taiwan Strait 8(3):242-246. Chen Junhau and Chen Guiging. 1987. Crystal structure and micro-structures of pearl and shells of pearl oyster in the South China Sea. Nanhai Studia Marina Sinica 8:135-142. Chen Pingjian and Wang Yingeng. 1985. The preliminary study on the biology of Cultellus scalprum (Gould) I. Morphological observation. Journal of Xiamen Fisheries College 2:34-41. Chen Pingjian, Ye Qiwang and Li Biquan. 1986. The preliminary study on the biology of Cultellus scalprum (Gould) II. Distribution and Living- habitats. Journal of Xiamen Fisheries College 1:46-57. Chio Leecheung and Lee Yunfan. 1983. A report on the raft culture of oyster (Ostrea rivularis Gould). Transactions of the Chinese Society of Malacology 1:149-156. Cui Guangfa. 1981. Influence of temperature and specific density of sea water on the development and morphology of clam larvae. Marine Sciences 1:37-38. Cui Guangfa, Yu Yeshao. 1985. Preliminary study on artificial breeding of Mactra quadrangularis. Marine Sciences 3:36-40. Fang Qi, Wei Xinmin. 1988. Study on suitable species and density of food for free-swimming larvae of bay scallop (Argopecten irradians). Journal of Oceanography of Taiwan Strait 7(4):331-338. Fang Yunchiang. 1982. A study on the reproductive mechanism of Ruditapes Philippinarum II. Ultrastructural study on promoting of egg maturity by NH,OH sea water. Journal of Oceanography of Taiwan Strait 1(2):99-105. Guan Futian, Han Yiping and Qu Weigong. 1989. On the habitat and growth of Musculus senhousei (Benson). Journal of Fisheries of China ZHUANG: CHINESE BIVALVE LITERATURE 211 13(3):181-188. Guo Shimao, Chen Chengchong and He Lixuan. 1986. A preliminary report on the biology of Pinna pectinata Linnaeus. Transactions of Chinese Society of Malacology 2:102-111. Guo Shimao and Chen Chengchong. 1987. A preliminary study on the arti- ficial rearing of larval pen shell Pinna pectinata Linnaeus. Marine Sciences 1:34-39. He Jinjin, Qi Qiuzheng and Wei Xinmin. 1981. A study on the food and feeding habit of the clam spat. Journal of Fisheries of China 8(2):99-106. He Jinjin, Wei Xinmin and Xu Zhangcheng. 1984. A study on the food of the larvae of Sinonovacula constricta. Journal of Oceanography in Taiwan Strait 3(2):208-216. He Jinjin, Wei Xinmin and Xu Zhangcheng. 1986. Study on food substrates of Sinonovacula constricta (Lamarck) spat. Journal of Fisheries of China 10(1):29-40. He Jinjin, Wei Xinmin, Fang Qi and Xu Zhangcheng. 1989. Artificial breeding of larval and juvenile Sinonovacula constricta (Lamarck) by combina- tion of indoor and outdoor conditions. Journal of Oceanography in Taiwan Strait 8(3):233-241. He Yichao and Zhang Fusui. 1983. The effect of temperature on the em- bryonic development of the mussel Mytilus edulis Linnaeus. Trans- actions of Chinese Society of Malacology 1:132-142. He Yichao and Zhang Fusui. 1986. Variation in range of effective temperature for embryonic development of mussel (Mytilus edulis Linnaeus). Transactions of Chinese Society of Malacology 2:88-93. Huang Jiachi. 1984. Study on the embryonic development of Laternula (Exolaternula) marilina and comparison with that of Sinonovacula constricta. Journal of Xiamen University (Natural Science) 23(2):224-229. Huang Jiachi and Xu Zhenzu. 1989. An experimental study of the effect of some environmental factors on the hatching rate of eggs of Sinonovacula constricta (Lamarck). Acta Ecologica Sinica 9(4):310-314. Huang, Z. G, Lee S. Y. and Mak P. M. S. 1983. The distribution and popula- tion structure of Perna viridis (Bivalvia, Mytilacea) in Hong Kong waters. In: Proceeding of Second International Workshop of Malacology, Hong Kong and South China. pp. 465-471. Hong Kong University Press. Jiango Weigo and Wei Yiyao. 1982. A preliminary observation of the chromosomes in several species of pearl oysters and their hybrids. Nanhai Studia Marina Sinica 3:\M1-118. Jiang Weigo, Wei Yiyao and Li Gang. 1983. Studies on the cultivated in- terspecific hybridization between pairs of Pinctada fucata, P. chem- nitzi and P. maxima (Mollusca, Bivalvia). If. Observation on the chromosomes of fertilization and hybrids. Tropic Oceanology 2(4):316-320. Jiang Weigo and Wei Yiyao. 1986. Study on the caryotypes of six species of the Pinctada (Pteridae, Bivalvia). Transactions of Chinese Society of Malacology 2:58-64. Jiang Weigo, Li Gang, Lin Yueguang and Qing Ning. 1987. Induced polyploidization in pearl oyster Pinctada martensi (Dunker). Tropic Oceanology 6(4):37-45. Jiao Jucheng, Liu Hongyao and Wang Rucai. 1985. The preliminary study on the semi-artificial collection of Venus clam’s spats. Marine Sciences 6:41-44. Jin Qizeng, Wei Yiyao and Jiang Weigo. 1981. On the artificial rearing of the larvae and juvenile pearl oyster Pinctada martensii (Dunker). Nanhai Studia Marina Sinica 2:107-116. Jin Qizeng, Wei Yiyao and Jiang Weigo. 1982. On the artificial rearing of larvae and juvenile of pearl oyster Pinctada martensii I. The rearing of the larvae and juvenile. Nanhai Studia Marina Sinica 3:99-10. Jin Qizeng and Zhou Baicheng. 1983. Studies on prevention and elimina- tion of fouling organisms on the pearl oyster Pinctada fucata Gould. I. Effects of fouling organisms on the survival and growth of the pearl oyster. Tropic Oceanologoy 2(2):141-147. Li Fenglan. 1983. Studies on the Arcidae of the Xisha Island, Guangdong Province, China. Studia Marina Sinica. 20:223-230. Li Fenglan. 1983. Studies on Chinese species of the family Arcidae II. Anadarinae. Transactions of Chinese Society of Malacology 1:31-44. Li Fenglan. 1984. A study of the Arcidae from China coasts I. Arcinae. Studia Marina Sinica 23:145-161. Li Fenglan. 1985. Studies on the Chinese species of the family Arcidae III. Striarcinae. Studia Marina Sinica 25:153-159. Li Fenglan. 1986. Studies on species of Glycemeridae off the China coasts. Transactions of Chinese Society of Malacology 2:22-29. Li Gang, Jiang Weigo and Wei Yiyao. 1983. Studies on the cultivated in- terspecific hybridizations between pairs of Pinctada fucata, P. chem- nitzi and P. maxima (Mollusca, Bivalvia) III. Comparative study on zymograms. Tropic Oceanology 2(4):321-328. Li Shixiao and Wang Renmei. 1982. Study on bioaccumulation and excre- tion of zinc-65 in mussel Myrilus edulis and their mechanisms. Marine Sciences 3:26-28. Liang Guangyao, Chen Shengtai and Xu Guoling. 1983. Report on artificial incubation of Ostrea rivularis Gould. Marine Sciences 5:41-43. Liang Xianyuan, Lin Huiqiong, Wu Pingru and Liu Yuan. 1986. Comparison on morphology of Pinnidae (Mollusca, Lamellibranchia) from the Fujian coast. Tropic Oceanology 5(1):13-19. Liang Xianyuan. 1988. The periostracum of Pinnidae as revealed by scan- ning electron microscopy. Jn: Proceeding of Marine Biology of South China Sea. pp. 87-89. China Ocean Press. Liao Chengyi, Xu Yingfa and Wang Yuanlong. 1983. Reproductive cycle of the scallop Chlamys farreri (Jones et Preston) at Qingdao. Journal of Fisheries of China 7(1):1-14. Lin Bishui, Huang Bingzhang and Wu Tienming. 1982. The spawning- facilitation effects of some chemical substances and sexual substance on clam (Ruditapes philippinarum). Marine Sciences 5:31-34. Lin Bishui, Wu Tianming and Huang Bingzhang. 1983. The effects of temperature and salinity on the growth and development of spats of the clam (Ruditapes philippinarum). Journal of Fisheries of China 7(1):15-24. Lin Bishui and Wu Tianming. 1986. The relation of temperature and salini- ty to the survival and growth of the spat of Sinonovacula constricta. Journal of Fisheries of China 10(1):41-50. Lin Bishui, He Jinjin, Wei Xinmin and Xu Zhangcheng. 1987. Preliminary study on artificial spating of Atrina (Servatrina) pectinata (Linnaeus). Journal of Oceanography in Taiwan Strait 6(3):260-268. Lin Jiahan and Ouan Teyio. 1983. Studies on ultrastructure of oocyte in the formation of yolk granules in oyster. Journal of Xiamen University (Natural Science) 22(3):355-360. Lin Ruicai, Chen Min and Lin Bishui. 1989. Effects of temperature and salini- ty on attachment and metamorphosis by bay scallop larvae Argopecten irradians (Lamarck). Journal of Oceanography in Taiwan Strait 8(1):60-66. Lin Yuefu, Chen Rongzhong, Zhou Xiulan, Yin Weiping and Wang Chusheng. 1985. Effects of mercury on the GPT activity of Perna viridis. Journal of Oceanography in Taiwan Strait 4(1):38-43. Liu Mingxing, Bao Wanyou and Zhang Shoulin. 1983. The seasonal varia- tion of some trace metals in the Ruditapes philippinarum in Jiaozhou Bay. Oceanologia et Limnologia Sinica 14(1):22-29. Liu Quanshun. 1984. The effect of some environmental factors on the arti- ficial breeding of Sinonovacula constricta. Marine Science Bulletin 3(5):41-45. Liu Zhongchuan. 1981. Natural spats collection of the scallop Chlamys farreri (Jones and Preston). Transactions of Oceanology and Limnology 3:60-62. Liu Zhongchuan. 1985. Large scale semi-artificial gathering of larval scallops for bumper harvest. Marine Sciences 6:45-47. 212 AMER. MALAC. BULL. 9(2) (1992) Liu Zhongchuan and Wang Yunjie. 1986. Shortening of Chlamys farreri cultivation period. Marine Sciences 4:40-41. Luo Wei and Gao Qiongzhen. 1988. Development of giant clams Tridacna crocea Lamarck. Tropic Marine Research 3:108-117. Lou Zikang, Liu Xiangsheng, He Yichao, Li Shuying and Zhang Fushui. 1982. A report on the experiment of mussel eggs collection. Journal of Fisheries of China 6(2):139-145. Lou Zikang, Liu Xiangsheng, Chen Zhaohua, Zhang Xiufeng and Zhang Naishi. 1982. Experiment on rearing mussel spats in late autumn by raising water temperature. Journal of Fisheries of China 6(1):43-49. Lou Yousheng, Lin Xianghui and Zhao Heping. 1983. An investigation on exploitation of scallop spatfall waters in Zhimao Wan. Journal of Fisheries of China 7(3):267-278. Meng Zhaomei and Li Xuezhang. 1983. Studies on the transplantation of- the mantle cell-piece and formation of pearl-sac in Pinctada maxima. Transactions of Chinese Society of Malacology 1:95-100. Ni Chunzi and Tseng Wenyoung. 1985. Preliminary assay of coliform in sea water and several edible bivalves of Tolo Harbour, Hong Kong. Tropic Oceanology 4(3):80-84. Nie Zhongqing and Ji Meifang. 1980. Studies on food for Mytilus edulis larvae. Marine Fisheries Research 1:85-96. Qi Qiuzheng, Lin Bishui, Wu Tianming, Xu Ruian and Yang Mingyue. 1981. Experiments on in-door induced breeding of Ruditapes philippinarum. Journal of Fisheries of China 5(3):235-243. Qi Qiuzheng and Yang Mingyue. 1984. The growth and development of the planktonic larvae, spats and young shells of Sinonovacula constricta. Journal of Oceanography in Taiwan Strait 3(1):90-96. Qi Qiuzheng. 1987. The life history of the clam (Ruditapes philippinarum). Journal of Fisheries of China 1\(2):1M-119. Qi Qiuzheng and Yang Mingyue. 1988. The growth and development of the clam Ruditapes philippinarum. Journal of Fisheries of China 12(2):1-1. Qi Xiang and Fang Yongqiang. 1989. Spermatogenesis of Ruditapes philip- pinarum. Journal of Oceanography in Taiwan Strait 8(3):247-250. Qiu Liqiang and Li Yongfang. 1986. Propagative period of oysters Crassostrea rivularis (Gould) in the Sheng-zheng Bay. Transactions of Chinese Society of Malacology 2:112-117. Qiu Wenren, Fu Subao, Zhou Dongtian, Zhu Ming and Shi Bingzhang. 1983. Studies on the artificial rearing of the planktonic larvae of Philippine clam Ruditapes philippinarum (Adams et Reeve) in the ponds, Jinjiang, Fujian. Journal of Xiamen University (Natural Science) 22(4):514-523. Shangguan Bumin, Xu Zhenzu and Liu Zhenzong. 1989. Studies on the gonadial maturity of razor clam Sinonovacula constricta (Lamarck). Journal of Xiamen University (Natural Science) 28(1):79-82. Shen Chiheng, Chen Baiyun and He Guansheng. 1985. The effect of infra- sound upon the growth and survival of larvae of Sinonovacula con- stricta (Lamarck). Marine Science Bulletin 4(1):58-62. Shi Bingzhang, Fu Subao and Qiu Wenren. 1984. Studies on the spawning of Philippine clam Ruditapes philippinarum (Adams and Reeve) in artificial rearing earth ponds. Journal of Xiamen University (Natural Science) 23(2):211-216. Song Zhile and Zhao Yushan. 1985. A preliminary experiment on the culture of great standard seeds of Chlamys farreri (Jones and Preston). Trans- actions of Oceanology and Limnology 1:46-49. Song Zhile, Zhao Yushan and Sun Jinshen. 1986. The experiment on rais- ing survival rate of Chlamys farreri (Jones and Preston) in temporary feeding stage. Marine Sciences 4:31-36. Su Baxian, Lao Zigian, Gao Xiang, Lai Yexing and Wu Yunhui. 1986. A preliminary report on the hemagglutinin of thirteen species of marine shellfish from Guangxi coast. Tropic Oceanology 5(1):81-84. Wang Chusheng, Yin Weiping, Lin Yuefu, Cheng Rongzhong, Zhou Xiulan and Yang Feng. 1985. Assay of 6-aminoleavulinic acid dehydratase in blood of Arca granosa Linnaeus. Journal of Oceanography in Taiwan Strait 4(2):141-146. Wang Qingcheng, Kou Baozeng, Liu Yongfeng and Li Wenji. 1986. A primary approach to the subject of rearing spats of Japanese scallop (Pecten yessoensis) at optimum temperature. Marine Science 2:49-50. Wang Renmei, Li Shixiao and Liu Fayi. 1984. Bio-accumulation, elimina- tion and combination of antimony by Mytilus edulis. Marine Sciences 1:31-32. Wang Ruccai, Gao Jie, Zhang Lianging and Wu Yuanqi. 1987. Studies on the techniques of collecting spats of scallop (Chlamys farreri) in natural sea area. Journal of Shangdong College of Oceanography 17(3):93-100. Wang Ruccai. 1989. An analysis of composition and content of amino acid in the adductor muscle of Farrer’s scallop. Journal of Ocean University of Qingdao 19(2):12-19. Wang Ruccai and Yu Ruihai. 1989. The utilization of algal excretion in the processes of bay scallop parent stocking. Marine Science 6:55-56. Wang Ruccai. 1989. Analysis of foodstuffs for Farrer’s scallop Chlamys (Azumapecten) farreri (Jones and Preston). Journal of Ocean University of Qingdao 19(2):36-48. Wang Weide and Wang Huichong. 1983. Artificial stimulation and hasten- ing of egg-laying of hard clams (Meretrix meretrix). Marine Sciences 2:33-35. Wang Wenxiong and Xu Zhenzu. 1988. The rearing behavior and the ef- fects of salinity on the burrowing activity of the spat of Sinonovacula constricta. Journal of Xiamen University (Natural Science) 27(3):333-337. Wang Wenxiong and Xu Zhenzu. 1989. The taxic behavior in the larvae of Crassostrea gigas. Journal of Xiamen University (Natural Science) 28(3):301-305. Wang Xinyuan. 1989. Tolerance of the blood cockle Anadara granosa L. and Philippine clam (Ruditapes philippinarum Adams and Reeve) to ammonia in sediments. Marine Sciences 5:51-54. Wang Yixing. 1984. Shortening of the maturation process of Pinctada martensii (Dunker). Marine Sciences 1:36-39. Wang Zhenrui. 1980. Study on Chinese species of Isognomonidae (Mollusca). Studia Marina Sinica 16:131-141. Wang Zhenrui. 1980. Study on Chinese species of the family Pectinidae (Mollusca, Bivalvia) I. A new species of the subfamily Pro- peamussinae. Oceanologia et Limnologia Sinica 11(3):259-262. Wang Zhenrui. 1982. Study on species of Plicatulidae off China coasts. Marine Sciences 1:23-27. Wang Zhenrui. 1983. Studies on Chinese species of the family Pectinidae Il. Chlamydidae (1. Chlamys). Transactions of Chinese Society of Malacology 1:45-54. Wang Zhenrui. 1983. Studies on the Mytilidae of the Xisha Islands, Goangdong Province, China. Studia Marina Sinica 20:213-221. Wang Zhenrui. 1983. Studies on the Chinese species of the family Pectinidae III. Chlamydinae (Genus Bractechlamys). Oceanologia et Limnologia Sinica 14(6):531-535. Wang Zhenrui. 1983. Studies on Chinese species of the family Pectinidae III. Chlamydinae (A new species and three new records of the genus Cryptopecten). Oceanologia et Limnologia Sinica 14(1):402-405. Wang Zhenrui. 1984. Studies on species of Pectinidae of the China coasts. II. Subfamily Amussinae. Studia Marina Sinica 22:245-253. Wang Zhenrui. 1984. Studies on Chinese species of the family Pectinidae VI. Subfamily Propeamussinae. Oceanologia et Limnologia Sinica 15(6):598-604. Wang Zhenrui and Qi Zhongyan (Tsi Chungyen). 1984. Study on Chinese species of the family Mytilidae (Mollusca, Bivalvia). Studia Marina Sinica 22:199-242. Wang Zhenrui. 1985. Studies on Chinese species of the family Pectinidae VII. Chlamydinae (Genus Semipallium). Oceanologia et Limnologia Sinica 16(6):502-506. Wang Zhenrui. 1989. Studies on the Chinese species of the family Pectinidae VIII. Subfamily Pectininae. Studia Marina Sinica 30:171-183. Wang Zichen, Liu Jiming and Zhu An. 1984. A preliminary survey on the biology of Mactra chinensis in Yalu Jiang River Outfall. Journal of ZHUANG: CHINESE BIVALVE LITERATURE pale) Fisheries of China 8(1):33-44. Wei Piping, Sun Jinshan and Ma Shaozheng. 1986. The reproduction, growth and distribution of Tegillarca granosa in Rushan Bay. Journal of Fisheries of China 10(1):87-94. Wei Liping and Guan Futian. 1986. Growing and breeding habits of Potamocorbula laevis. Transactions of Chinese Society of Malacology 2:94-101. Wei Yiyao, Jiang Weigo and Li Gang. 1983. Studies on the cultivated in- terspecific hybridization between pairs of Pinctada fucata, P. chem- nitzi and P. maxima (Moll. Biva.) I. Cultivated hybridizations and observations on the hybrids. Tropic Oceanology 2(4):309-315. Wei Yiyao, Jiang Weigo and Jin Qizeng. 1985. Study on the artificial rear- ing of larvae and juvenile pearl oysters Pinctada chemnitzi. Journal of Fisheries of China 9(1):13-18. Weng Dequan. 1982. A preliminary observation on the ecology of the spat of Chlamys nobilis (Reeve). Marine Sciences 5:38-39. Wu Rong. 1985. The genetics and breeding of oysters. Journal of Fisheries of China 9(2):207-215. Wu Tianming and Lin Bishui. 1987. Effects of several environmental factors on survival rate of young shells of Sinonovacula constricta (Lamarck). Journal of Oceanography in Taiwan Strait 6(2):120-126. Wu Yulin, Cui Keduo, Liu Yumei, Hou Lanying and Lou Qingxiang. 1983. Laboratory experiment on the accumulation and depuration of mer- cury by Arca (Anadara) subcrenata Lischke. Oceanologia et Lim- nologia Sinica 14(1):30-34. Xiao Shunghong. 1984. Oyster breeding by hanging into sea water in flood area. Marine Sciences 3:49-52. Xie Yukan and Xu Zhijian. 1980. The early feeding habits of the larvae of Pinctada maxima. Acta Zoologica Sinica 26(1):80-85. Xie Yukan and Lin Biping. 1983. On handling the pearl oysters before and after nucleus-insertion. Journal of Fisheries of China 7(3):229-234. Xie Yukan, Lin Biping and Feng Yongqin. 1988. A record on the tropical culture of Pinctada fucata (Gould). Tropic Marine Research 3:118-123. Xie Yukan and Lin Biping. 1988. The pearl culture of Pinctada fucata (Gould) dividing two periods in the climate zones. Tropic Marine Research 3:1-9. Xiong Daren and Cai Yingya. 1981. A report on Lamellibranchia of Zhanjiang coast. Journal Zhanjiang College of Fisheries 1:1-10. Xu Fengshan. 1980. Two new species of Bivalvia (Mollusca) from the East China Sea. Oceanologia et Limnologia Sinica 11(4):337-340. Xu Fengshan. 1984. Preliminary study on the Protobranchia (Mollusca) from the shallow waters of China II. Nuculidae. Studia Marina Sinica 22:179-188. Xu Fengshan. 1984. Preliminary study on the Protobranchia (Mollusca) from the shallow waters of China I. Nuculanidae. Studia Marina Sinica 22:167-177. Xu Fengshan. 1986. The new records and list of Bivalves from the Jiaozhou Bay. Transactions of Chinese Society of Malacology 2:30-41. Xu Fengshan. 1987. New species and new records of Myidae from the China coast. Oceanologia et Limnologia Sinica 18(5):437-441. Xu Zhangcheng, Hong Liqiang, He Jinjin, Wei Xinmin and Lin Bishui. 1989. Toxic effects of oil and oil dispersant on straight hinge of Sinonovacula constricta. Journal of Oceanography in Taiwan Strait 8(1):68-73. Xu Zhijian and Li Kongkai. 1980. A preliminary study on the rearing of artificially fertilized larvae of Pinctada maxima. Journal of Fisheries of China 4(3):295-302. Xu Zhijian and Xu Yijiang. 1983. Preliminary observation of the structural change of the aggregate size and weight of pearl oyster. Marine Sciences 3:36-38. Yang Jiadong. 1984. Extraction of vitamin B12 from Arca subcrenata Lischke and effect of 60-Co on B12 content. Marine Sciences 2:29-31. Yang Qingming and Zhang Qixin. 1985. Preliminary study of ear-hanging breeding technique of scallop. Marine Sciences 4:39-40. Yang Xuefang and Sun Changxiang. 1981. A study on the artificial rearing of Chlamys farreri spats. Transactions of Oceanology and Limnology 1:64-69. Yang Xuefang. 1982. On the effect of high temperature upon the mussels (Mytilus edulis L.) spats. Marine Sciences 5:35-37. You Zhongjie and Wang Yinong. 1989. Additions to the marine Bivalvia of the Nanji Island of China. Journal of Zhejiang College of Fisheries 8(1):17-27. Yu Weiping and Yan Sixu. 1986. Chemical modificiation of functional groups of alkaline phosphatase from Sinonovacula constricta. Journal of Xiamen University (Natural Science) 25(5):562-567. Yu Yeshao, Yu Zhihua. 1985. A preliminary analysis of the cause of the mor- tality of Meretrix meretrix (Linnaeus). Marine Sciences 4:46-48. Zhang Fusui, He Yichao, Liu Xiangsheng, Ma Jianhu and Lou Zikang. 1980. The breeding seasons of mussel (Mytilus edulis L.) in Jiaozhou Bay, Shangdong Province, China. Oceanologia et Limnologia Sinica 11(4):341-350. Zhang Fusui, He Yichao, Li Shuying, Liu Xiangsheng, Ma Jianhu, Chen Zhaohua and Zhang Xiufeng. 1981. Observations on the growth of mussels in Jiaozhou Wan. Journal of Fisheries of China 5(2):133-146. Zhang Fusui, Li Shuying, He Yichao, Liu Xiangsheng, Ma Jianghu and Yu Shuoen. 1983. Studies on the yield of eggs collected from parent mussels cultured in Jiaozhou Bay, Shangdong Province, China. Trans- actions of Chinese Society of Malacology 1:121-131. Zhang Fusui, Lou Zikang, Ma Jianghu, Yu Shuoen, Liu Xiangsheng and Li Shuying. 1984. The development of seed mussel resources in Jiaozhou Bay, Shangdong Province, China. Studia Marina Sinica 22:115-125. Zhang Fusui, He Yichao, Ma Jianhu and Liu Xiangsheng. 1984. Renovation research on artificial rearing of mussel spats I. The improvement of collector and inhibition of bacteria. Oceanologia et Limnologia Sinica 15(6):590-597. Zhang Fusui, He Yichao, Ma Jianghu, Liu Xiangsheng, Li Shuying, Qi Lingxin, Yu Shuoen and Yan Shuisheng. 1984. The introduction of the Japanese scallop Patinopecten yessoensis (Jay) into China, its spat- rearing and experimental cultivation. Marine Sciences 5:38-45. Zhang Fusui, Ma Jianghu, He Yichao, Li Shuying, Liu Xiangsheng and Yu Shuoen. 1985. Comparison of feeding effects of various feeds in rearing larvae of Mytilus edulis L. Studia Marina Sinica 25:91-101. Zhang Fusui, Li Shuying, Liu Xiangsheng, He Yichao and Ma Jianghu. 1986. A study on the meat condition of mussel (Mytilus edulis) in Jiaozhou Bay, Shangdong Province, China. Transactions of Chinese Society of Malacology 2:80-87. Zhang Fusui, He Yichao, Ma Jianghu and Liu Xiangsheng. 1986. Comparison to two crops of scallop (Chlamys farreri) larvae in growth and develop- ment. Transactions of Chinese Society of Malacology 2:118-120. Zhang Fusui, He Yichao, Liu Xiangsheng, Ma Jianghu, Li Shuying and Qi Lingxin. 1986. A report on the introduction, spat-rearing and ex- perimental culture of bay scallop Argopecten irradians Lamarck. Oceanologia et Limnologia Sinica 17(5):367-374. Zhang Fusui and He Yichao. 1987. A report on the experimental cultivation of bay scallop and kelp by turns. Marine Sciences 6:1-6. Zhang Fusui, He Yichao, Liu Xiangsheng and Ma Jianghu. 1989. Mass pro- duction of commercial seed bay scallop in China. Jn: Current Topic in Marine Biotechnology. pp. 306-310. Zhang Jianzhong and Li Fuxue. 1988. Determination of age of the Asian hard clam Meretrix meretrix (Linnaeus). Journal of Fisheries of China 12(3):251-258. Zhang Junye. 1983. A study on the artificial rearing of spats of Saxidomus purpuratus and their transplanting into the sea. Transactions of Chinese Society of Malacology 1:157-164. Zhang Qixin and Yang Qingming. 1983. A preliminary inquiry of the most suitable collecting period for scallop. Marine Sciences 4:34-37. Zhang Qixin and Yang Qingming. 1984. Breeding depths of scallop (Chlamys farreri) on raft. Marine Sciences 4:33-34. 214 AMER. MALAC. BULL. 9(2) (1992) Zhang Qixin and Jiang Wenfa. 1987. Semi-artificial picking of Chlamys farreri in Autumn. Marine Sciences 5:46-47. Zhang Shumei and Zeng Shunqin. 1982. Determination of mercury in venerupis philippinarum (Adams and Reeve) from Jiaozhou Bay. Marine Sciences 1:19-22. Zhang Tianfo, Gu Tangxiu and Xu Xianyi. 1982. Rapid and simple method for the determination of BHC and DDT pesticide residues in Meretrix sp. Oceanologia et Limnologia Sinica 13(6):510-513. Zeng Tianling. 1989. Study on the modalities of bacterial contamination of mussels. Journal of Xiamen University (Natural Science) 28(1):101-105. Zhou Dongtian, Zhu Ming, Shi Bingzhang, Qiu Wenren and Fu Subao. 1984. Studies on the artificial culture of Philippine clam Ruditapes philip- pinarum (Adams and Reeve) in the ponds, Jinjiang, Fujian. Journal of Xiamen University (Natural Science) 23(4):515-522. Zhou Zongcheng, Ni Chunzhi, Li Zhitang, Zeng Huoshui and Zhang Nanfeng. 1983. The relationship between the number of fecal coliform and oyster diseases in a oyster hatchery. Journal of Oceanography in Taiwan Strait 2(2):133-136. Zhuang Qiqian. 1983. Two new species of Mactridae (Mollusca, Bivalvia) off the Chinese coast. Oceanologia et Limnologia Sinica 14(1):88-91. Zhuang Qiqian. 1984. Studies on the Chamidae (Bivalvia) off the China coast. Studia Marina Sinica 22:191-198. Zhuang Qiqian, Lin Huigiong and Liang Xianyan. 1981. On the species of the genus Ruditapes (Mollusca, Lamellibranchia, Veneridae) off China coast. Studia Marina Sinica 18:207-215. Zhuang Qiqian and Cai Yingya. 1982. Studies on the Laternulidae off the Chinese coast. Oceanologia et Limnologia Sinica 13(6):553-561. Zhuang Qiqian and Cai Yingya. 1983. Studies on the Corbulidae (Bivalvia) off Chinese coasts. Transactions of Chinese Society of Malacology 1:57-68. 1990-1991 Gao Shihe and Xu Zhenzu. 1990. Study on commodity food for planktonic larvae of Sinonovacula constricta. Marine Science Bulletin 9(3):63-68. Guan Futian and Wei Liping. 1990. The breeding habit and the embryonic development of Musculus senhousei (Benson, 1842). Transactions of Chinese Society of Malacology 3:\17-123. Guan Yinling and Li Yongfan. 1990. Comparative studies on some physiological and biochemical indices of the cultured oysters in the Zhujiang River Estuary. Transactions of Chinese Society of Malacology 3: 110-116. He Yichao and Zhang Fusui. 1990. The influence of environmental salinity on various development stage of the bay scallop Argopecten irradians Lamarck. Oceanologia et Limnologia Sinica 21(3):197-204. Jiang Weigo, Li Gang, Lin Yueguang, Xu Gaoqiang and Qing Ning. 1990. Observation on the gonad of triploidy in Pinctada martensii (D.). Tropic Oceanology 9(1):24-30. Jiang Wenfa, Liu Qishun and Xiao Peihua. 1990. Large scale experiments on high density seedling rearing of bay scallops. Transactions of Oceanology and Limnology 4:65-69. Jiang Yu, Wu Ruimin and Zhang Ligui. 1991. A study on artificial breeding of Ostrea gigas in large areas. Fujian Fisheries 3:44-48. Li Fenglan. 1990. Studies on species of the family Limopsidae (Bivalvia) off the China coasts. Transactions of Chinese Society of Malacology 3:19-23. Li Pilian, Gao Xialing, Hong Yichun, Kang Zhuang and Huang Kunshou. 1990. Autumn artificial spat rearing and experimental culture of bay scallop Argopecten irradians. Tropic Oceanology 9(2):16-22. Li Xiaoxu. 1990, On the development of gills in the oyster Crassostrea gigas. Marine Sciences 2:13-17. Lin Bishui and Wu Tianming. 1990. Temperature and salinity related to the survival, growth and development of the larvae and spat of Stnonovacula constricta. Journal of Fisheries of China 14(3):171-178. Lin Ruccai, Lin Bishui and Chen Min. 1991. Study on effects of temperature and salinity on migrating behavior, growth and survival of juvenile bay scallop Argopecten irradians (Larmack). Journal of Oceanography in Taiwan Strait 10(2):133-138. Liu Zhengzong, Shanguan Bumin and Xu Zhenzu. 1990. On the ultrastruc- tural features of spermatogenesis of razor clam, Sinonovacula con- stricta (Lamarck). Journal of Xiamen University (Natural Science) 29(1):81-84. Luo Wei. 1991. Larval growth of giant clam Tridacna crocea Lamarck. Tropic Oceanology 10(1):71-77. Meng Zhaomei and Xing Kongwu. 1991. The effect of various factors on the nucleus-insertion of the black lipped oysters Pinctada margaritifera (Linnaeus). Oceanologia et Limnologie Sinica 22(1):8-13. Wang Meilin. 1990. The karyotype of Chlamys farreri. Journal of Ocean University of Qingdao 20(1):81-85. Wang Qingcheng and Li Wenji. 1990. Primary study on artificial rearing spats of Arca (Anadara) inflata Reeve. Transactions of Chinese Society of Malacology 3:124-127. Wang Yuanlong, Yang Xiaoyan, Sun Guang and Zhang Dehua. 1991. Research on the technique for preserving seedings of bay scallop (Argopecten irradians) in the sea. Transactions of Oceanologia and Limnologia 4:79-87. Wang Zhenrui. 1990. A study on species of the family Pectinidae I. Genera Volachlamys and Annachlamys off the China coasts. Transactions of Chinese Society of Malacology 3:13-18. Wang Zhenrui. 1990. Study on offshore species of the family Limidae from waters off China. Studia Marina Sinica 31:163-174. Xiang Jianhai and Chen Qiu. 1991. Study on the possibility of hybridization and breeding between Argopecten irradians, Chlamys farreri and Patinopecten yessoensis II. A study on the cytogenetic basis of the hybridization between the three species of scallops. Annual Research Report of the Experimental Marine Biology Laboratory, Institute of Oceanology, Academia Sinica pp. 137-139. Xu Fengshan. 1990. Preliminary study on the Protobranchia (Mollusca) from the China Seas III. Malletidae and Tindariidae. Oceanologia et Lim- nologia Sinica 21(6):559-562. Xu Fengshan. 1990. Study on the Septibranchia from Chinese waters. Studia Marina Sinica 31:177-184. Xu Fengshan. 1990. The Bivalvia in the deep-water area of the East China Sea. Studia Marina Sinica 31:185-193. Xu Guogiang, Lin Yueguang, Li Gang and Jiang Weigo. 1990. A preliminary study on the induction of gynogenetic diploid and ‘‘Hetwig Effect’’ in pearl oyster Pinctada martensii (D.). Tropical Oceanology 9(2):1-7. Xu Zhangcheng, Wei Xinmin and He Jinjin. 1990. Effects of some ecological factors on growth and survival of juvenile Sinonovacula constricta. Journal of Oceanography in Taiwan Strait 9(1):62-68. You Zhongjie, Wang Yinong, Zhu Xinding and Li Renwei. 1990. Preliminary study on population comparison and growth of the clam (Moerella iridescens). Marine Science Bulletin 9(6):35-40. You Zhongjie, Wang Yinong, Yan Zhengrong and Guo Shen. 1991. Growth and seasonal change in size structure of the clam, Gomphina veneriformis. Journal of Oceanography in Taiwan Strait 10(1):52-58. You Zhongjie, Wang Yinong and Zhang Jialiang. 1991. Preliminary study on life history of Moerella iridescens (Benson). Marine Science Bulletin 10(3):50-55S. You Zhongjie and Wang Yinong. 1991. A preliminary study on the artificial seed rearing of Moerella iridescens. Transactions of Oceanology and Limnology 3:55-61. Zhang Fusui, Ma Jianghu, He Yichao, Liu Xiangsheng, Li Shuying and Qi Lingxin. 1991. A study on the meat condition of the bay scallop in Jiaozhou Bay. Oceanologia et Limnologia Sinica 22(2):97-103. Zhang Fusui, He Yichao, Liu Xiangsheng, Ma Jianghu, Qi Lingxin and Li Shuying. 1991. Growth and mortality of bay scallop Argopecten irradians cultured at various water-layers in Jiaozhou Bay. Journal of Fisheries of China 15(1):42-47. ZHUANG: CHINESE BIVALVE LITERATURE 215 Zhang Guofan, Gao Yuemian. 1990. The assimilation efficiencies of Arca inflata fed on different single alga. Transactions of Chinese Society of Malacology 3:103-109. Zhang Xinjian and Xing Yinping. 1990. Studies on anti-cancer activity of Meretrix meretrix nucleic acid. Oceanologia et Limnologia Sinica 21(1):88-91. Zhao Zhijiang, Li Fuxue and Ke Caihuan. 1991. On the gonad development and reproductive cycle of the clam Paphia undulata. Journal of Fisheries of China 15(1):1-9. Zhen Zhinan and Li Fuxue. 1990. Seasonal changes in soft part weight and biochemical composition of the bivalve Cyclina sinensis. Tropic Oceanology 9(2):8-15. Zhen Zhinan and Li Fuxue. 1991. The study on reproductive cycle of Cyclina sinensis. Tropic Oceanology 10(1):86-92. Zhou Lihong and Yu Zhihua. 1991. The effect of darkening on the spawning of Sinonovacula constricta (Lamarck). Marine Science 2:71-72. Zhou Wei. 1991. The biological zero of gonad development of bay scallop [Argopecten irradians (Lamarck)]. Journal of Fisheries of China 15(1):82-84. Date of manuscript acceptance: 24 April 1992 AMERICAN MALACOLOGICAL UNION FINANCIAL REPORT 1991 Income and Expenses Income: SOO Dues: (GiCcatePONies) da.coaiis nau aah nateine a adie ge Riana ee been a wus 6 Aaa eens $ 3,626.00 1OOS Dues Call Care CONES) sinks ttc de iain bee COE ey Boake ce ate Redon owe e eed 11,931.00 1OOZ sues; (allCateg Ones). 1.25 er aaqungte cy ee Mee See NO RMaG hie oan He eae aaa ee 9,167.00 eke MEMbDEnsnipaicnt 3. ethan cate Sees Gara e Ga fina Aare os mae alee wee nae ae 400.00 BULICLIN SUUSCTIPUONS = VOINDIE + 5 Sircce hun ad ee et gl eats. « dt ans oer wae was at 1,817.00 Bullen Subscnpuions = Volume 9 piso i eket tecuusagete ke oe ieehebuseeeditegeesnas douse’ 2,375.00 Bullen SUOSCTIPtOns = VOICE NO tence caes : eee ae totaal oe A weraine a sala ae® 1,364.00 Bulleun, Sobscripuons— volume ll. .< sarees pet dsas cose Sea gear ies ds4e uae ee eekaeas 32.00 Bulletin back 1ssUCS Sc SUPPICIMIENtS 20 wo 22 aS eSea ance ewe Oe dnd 4 oe ete eee ee See kee eee 839.00 Page charges and reprints). saa «gas aa baa eh in cane chee e a nekatn ener Se Ee wha 1,930.85 Other Publications ~s.¢4 1% sec aw a eee eee Sakae ee he aaa s aoe eee Cee 193.00 HESIMEL SALES Were. tantra wean ele aa ia Sei aha Pam edn she tie eee 1,557.00 NEG) ACSI Sees are x ca 5 seen ae ph Aas a Mac ie RAM Sa Se Pela PD Sy bck tk GeO Seen Alcanshts ested gee 188.00 Student. Award Donavon... ysii040e were en ahicweus wach anes bo Radke eee eae meee 250.00 72 VC) 0101 § eae ae ee ar Pee eT OS ns Seeded ean eee eg eee ee eae eer eee ee 806.41 Interest Call aCcOunts)icck. 26 chest ee nag wy elsoaelSieister ao aia alates whe wane Pde ad ake wales aie wees 3,492.57 INMISCelanGOUS INCOING 2. ....s 550.8 Bosc wer nek ae ee etend aoe nuseiwet inane dureewa ee 466.88 Sy MpOsiM MUNG DONANONS a5 vanata mies ec Gag da hueen «ie GR RON ed Weg as Stee owNY 1,150.00 $41,585.71 Expenses: WETICExCXPCUSCS 1d batts rren hice OeR ORR Ane 8 Pode a eal doh mck a pa Paledgaesae wae TOS 1,937.63 POSCAS Cis ee eee wie sare oes cones we a Sa Ue Re pene Magia tae aA ee y 1,437.75 Dues invOter OrBamiZ ations baa.-tr witha eee eee tee ee aoa a i ee ute coe ela N asad onal 345.00 Branciisevanid COrporadOn recsa haan aewticcd eaGh ag auaeur ns shen haa Rane ay gaa a rene 100.00 Misukance and BONG! Gece aie una dat awe wn Saeed dd aemoeean tal ead dowd 383.00 PSCCOUDEIN Sc 42. tee sere eaey Meee Fick GR ES Copa hited arpa chpraen sda a wel ee NG dist ai ati eas ancdcate ae brathe Ges 280.00 UPaAVeIPEXOENSES ico eee ie eter eS all on eri lea a cise § ceaed wena UN Sa. sc Boe ae aS 2,065.00 IE Re Sy eee eee ae emer me te eT 8 clack Ua teat erese, Geaat ene eliaaianig hath Acerca tall Aaah a 140.00 Banke Charles Gases ak eM an a aN ead add cond ta ark ale dared tnd sana bailed 159.28 SW POSIM I CRC NSCS ine recreate ane dutta) «cate whe thee dates ng eaves tor nate aaNnahe Siaiminaileid a wie 2,968.00 BUS UTES es etter etre RS tr cle i dete es Sn ie ech ene ip Gl st nda ath Ni desert tate Baa tay 932.90 Student; papel A Wats nines cece ited ct Mig tive atnc we ke wba Rating be eee ane Weer ne 750.00 L991 Meeting expenses neat eco ees oe ee eee watotanng Mius hee hiadegneeenss aes: 331-20 MOP OMICS INS ca carte eae ee eeu eect is a rd eres aunty Ene en mw aac Chaay ie aac une eae 118.45 SILL CUTI ESL) PENCE Te on ORA Re RE, eal. flay Bec het ae Alea rt esanie Melvaisd ls ane if. lactase ek keer t 6,273.87 BUC: = TEDTINts FOE Rls WOOO ae eye sees tas ko Ge ee ig eae tneeh oy Singhs eae aA 140.45 INEWSIELLETS A(EWO) aeons renee ece eee eh eee eccnd ted co re wansk teeth RNa cle Me cece eh eat concn cre Manat. apes 2,130.03 $20,492.56 INTC CESS TMC CAS lea SSCS areas ester cas ger capre cheat ee ret eta e hey meee Scr see heh ara AK hk AiNsbalatallon cit al chchada Oe 21,093.15 Gashieassetszas 0 fealty Oiler rte eee en BPI eae et ele Sade a Meee er thle «Snes ae 76,948.81 Baee ND Pe ITN CL CASE tree ee ones yr eke ete ade fee te aah mean OR ne ieee ee ct ale 2 Obie a 21,093.15 $98,041.96 217 IN MEMORIAM Anita Brown Croft Ralph W. Dexter MacKenzie Gordon, Jr. 218 INDEX TO VOLUME 9 (1 and 2) AUTHOR INDEX Allen, J. A. 187 Handwerker, T. S. 27 Ralley, J. 21 Bieler, R. 157 Harte, M. E. 199 Runnegar, B. 117 Blankenship, J. E. 85 Hornbach, D. J. 39 Schneider, J. A. 145 Boal, J. 75 Jesien, R. V. 27 Seed, R. 123 Carriker, M. R. 193 Leonard, J. L. 45 Showers, M. 173 Cleveland, C. 173 Mikkelsen, P. M. 157 Steiner, G. 1 Counts III, C. L. 27 Morton, B. 107, 165 Strayer, D. L. 21 DeWitt, T. J. 81 Muntz, W. R. A. 69 Strenth, N. E. 85 Dillon, R. T., Jr. 99 Pojeta, J. 7 Wethington, A. R. 99 Gagnon, J.-M. 59 Poutiers, J-M. 139 Winstead, R. L. 173 Gilkinson, K. D. 59 Prezant, R. S. 173 Zhuang, Q. 207 Hamilton, P. V. 89 PRIMARY MOLLUSCAN TAXA INDEX Alasmidonta 22, 30 Fulvia 148 Navanax 51 Anodonta 22, 30 Fustiaria 13 Nemocardia 139 Anomalocardia 200 Fustiariidae Fam. Nov. 18 Octopus 69, 75 Antalis 12 Gadilida 7 Parabornia 160 Aplysia 85, 92 Galeomma 160 Pecten 96 Arctica 201 Habecardium 148 Perna 208 Argopecten 209 Helix 54 Phlyctaenachlamys 160 Ariolimax 53 Hemidonax 145 Physa 81, 99 Biomphalaria 51 Hermissenda 55 Pinctada 209 Cardiidae 145 Hypanis 146 Pojetaia 7 Cardiinae 145 Laevicardiinae 148 Pratulum 139 Cardium 146 Laevicardium 148 Protocardia 139 Cerastoderma 147 Lahillia 147 Protocardiinae 139, 146 Ceratobornia 160 Lahillidae 148 Pseudomyona 117 Chione 200 Lampsilis 31 Pterocera 93 Chlamys 59, 209 Lasaea 165 Ruditapes 209 Clinocardiinae 146 Leptodea 31 Sawkinsia 148 Clinocardium 146 Ligumia 32 Septifer 165 Crassostrea 193, 208 Lissarca 173 Septocardia 139, 147 Cryptonomella 200 Littorina 93 Sinonvacula 208 Dentaliida 5 Lymnaea 52 Strophitus 23, 32 Dinocardium 148 Lymnocardiinae 146 Tegillarca 208 Divariscintilla 157 Lyrocardium 143 Timoclea 200 Elliptio 22,29 Macrocallista 200 Tridacnidae 149 Entalina 14 Mercenaria 200 Tridacninae 149 Ephippodonta 160 Microcardium 139 Trifaricardium 142 Fordilla 7 Musculium 39 Tuarangia 117 Fraginae 146 Mytilus 123, 193, 208 Unionoidea 21, 27 Fridigocardium 143 Nautilus 69 Dates of Publication Volume 9(1), December, 1991 Volume 9(2), August, 1992 219 CONTRIBUTOR INFORMATION The American Malacological Bulletin serves as an out- let for reporting notable contributions in malacological re- search. Manuscripts concerning any aspect of original, unpublished research, important short reports, and detailed reviews dealing with molluscs will be considered for publication. Each original manuscript and accompanying illustra- tions must be submitted with two additional copies for review purposes. Text must be typed on one side of 8% x Il inch bond paper, double-spaced, and all pages numbered con- secutively with numbers appearing in the upper right hand corner of each page. Leave ample margins on all sides. Form of the manuscript should follow that outlined in the Council of Biology Editors Style Manual (fifth edition, 1983). This can be purchased from the CBE, 9650 Rockville Pike, Bethesda, Maryland 20814, U.S.A. Text, when appropriate, should be arranged in sections as follows: 1. Cover page with title, author(s) and address(es), and suggested running title of no more than 50 characters and spaces. Authors should also supply five key words, placed at the base of this page, for indexing purposes. . Abstract (less than 5% of manuscript length) . Text of manuscript starting with a brief introduc- tion followed by methodology, results, and discus- sion. Separate sections of text with centered sub- titles in capital letters. 4. Acknowledgments 5. Literature cited 6. Figure captions All binomens must include the author attributed to that taxon the first time the name appears in the manuscript [e.g. Crassostrea virginica (Gmelin)]. This includes non-molluscan taxa. The full generic name along with specific epithet should be written out the first time that taxon is referred to in each paragragh. The generic name can be abbreviated in the re- mainder of the paragraph as follows: C. virginica. References should be cited within text as follows: Hillis (1989) or (Hillis, 1989). Dual authorship should be cited as follows: Yonge and Thompson (1976) or (Yonge and Thomp- son, 1976). Multiple authors of a single article should be cited as follows: Beattie et al. (1980) or (Beattie et al., 1980). In the literature cited section of the manuscript refer- ences must also be typed double spaced. All authors must be fully identified, listed alphabetically and journal titles must be unabbreviated. Citations should appear as follows: Beattie, J.H., K.K. Chew, and W.K. Hershberger. 1980. Dif- ferential survival of selected strains of Pacific oysters (Crassostrea gigas) during summer mortality. Pro- ceedings of the National Shellfisheries Association 70(2):184-189. Hillis, D.M. 1989. Genetic consequences of partial self- fertilization on population of Liguus fasciatus (Mollusca: Pulmonata: Bulimulidae). American Malacological Bulletin 7(1):7-12. Seed, R. 1980. Shell growth and form in the Bivalvia. In: Skeletal Growth of Aquatic Organisms, D.C. Rhoads and Why R.A. Lutz, eds. pp. 23-67. Plenum Press, New York. Yonge, C.M. and T.E. Thompson. 1976. Living Marine Molluscs. William Collins Son & Co., Ltd., London. 288 pp. Illustrations should be clearly detailed and readily reproducible. All line drawings should be in black, high quali- ty ink. Photographs must be on glossy, high contrast paper. All diagrams must be numbered in the lower right hand cor- ners and adequately labeled with sufficiently large labels to prevent obscurance with reduction by one half. Magnifica- tion bars must appear on the figure, or the caption must read Horizontal field width = xmm or xpm. All measurements must be in metric units. All illustrations submitted for publica- tion must be fully cropped, mounted on a firm white back- ing ready for reproduction, and have author’s name, paper title, and figure number on the back. All figures in plates must be nearly contiguous. Additional figures submitted for review purposes must be of high quality reproduction. Xerographic reproduction of photomicrographs or detailed photographs will not be acceptable for review. Abbreviations used in figures should occur in the figure caption. Indicate in text margins the appropriate location in which figures should appear. Color illustrations can be included at extra cost to the author. Original illustrations will be returned to author if requested. Any manuscript not conforming to AMB format will be returned to the author for revision. New Taxa. The Bulletin welcomes complete descriptions of new molluscan taxa. Establishment of new taxa must con- form with the International Code of Zoological Nomenclature (1985). Descriptions of new species-level taxa must include the following information in the order as given: higher taxon designation as needed for clarity; family name with author and date; generic name with author and date; Genus species author sp. nov. followed by numeration of all figures and tables; complete synonymy (if any); listing of type material with holotype and any other type material clearly designated along with complete museum catalogue or accession infor- mation; listing of all additional non-type material also with full museum deposition information; type locality; diagnosis and full description of material done in telegraphic style in- cluding measurements and zoogeographic distribution as necessary; discussion; etymology. Descriptions of new supra- specific taxa should include type species (for new genus) or type genus (for new family), diagnosis and full description done in telegraphic style, and list of included taxa. Proofs. Page proofs will be sent to the author and must be checked for printer’s errors and returned to the printer within a three day period. Significant changes in text, other than printer errors, will produce publishing charges that will be billed to the author. Mailing. All overseas mailing must be done via airmail. The American Malacological Union will not be responsible for deferred publication of manuscripts delayed in surface mail. Charges. There are no mandatory page costs to authors lack- ing financial support. Authors with institutional, grant or other research support will be billed for page charges. The current rate is $35.00 per printed page. Acceptance and ulti- mate publication is in no way based on ability to pay page costs. Reprints. Order forms and reprint cost information will be sent with page proofs. The author receiving the order form is responsible for insuring that orders for any coauthors are also placed at that time. Submission. Submit all manuscripts to Dr. Robert S. Prezant, Editor-in-Chief, American Malacological Bulletin, Depart- ment of Biology, Indiana University of Pennsylvania, Indiana, Pennsylvania, 15705-1090, U.S.A. Subscription Costs. Institutional subscriptions are available at a cost of $32.00 per volume. [Volumes | and 2 are available for $18.00 per volume]. Membership in the American Malacological Union, which includes personal subscriptions to the Bulletin, is available for $25.00 ($15.00 for students). All prices quoted are in U.S. funds. Outside the U.S. postal zones, add $5.00 seamail and $10.00 airmaii per volume or membership. For subscription or membership information contact AMU Acting Secretary-Treasurer, Richard E. Petit, P. O. Box 30, North Myrtle Beach, South Carolina, 29582, U.S.A. \ i t { er > ‘ i 1 { ] ” ie | ate oe oth | J te FS CR FU EN FS Ge ww eee euegwy pt re ite Bs 5 Neen oe at wives ee our eyes SS ae! eee ba i aa a Seb sea or =a SPE 6 FP UNTRUE Soe EVES SE S 79) = wn = w = & Ww Ff Ww ae. . WW A wy 4 =e | re = Pe 4 ~~ = 4 = i = a =. a = BS “¢ = EG WHYS a < = a = = 4A SNXQ's 4 = & fs) a ro) 5 OT ewe 3 co” STITUTION NOILMLILSNI NVINOSHLINS S3INYVYGIT LIBRARIES SMITHSONIAN INSTITUTION NOILNLILSNI NYIA — - z= Ee = z re S) os S) = ) = FS © E 2a Ee = = wD .é ~ eae =) = -. 2 5 io] \ = >" = > 3 > i= >: ‘ i = - 2 a 2 = 0 z fs 2 D 2 r z rn J1YVYGIT_LIBR is NOI \ a SES = gp NOHLALILSNI pSdtYVUGIT LIBRARIES SMM “= Ss FE < = < = ZZ & £E EQN : z : 3 = 2S 5 EN 32 g ue 4 g 2M 2 E MOQ 2 E zZ FE 2 E \. 8 = = = Bini tem 5 = £ ay ME STITUTION NOILNLILSNI LIBRARIES SMITHSONIAN INSTITUTION NOILNLILSNI NVIt NOILNLILSNI n = wn > w = non = Z a “ & a Ne aa uw a | ac a a = a. 4Y = cc a < = < a < cA = < a 4% nd iS a = a Gy c oc e 4 Pre) = oo = ao FY et oO at as ro) = oO — fe) — ro) as Fa NL 2 ad ra =i a UVUGIT LIBRARIES SMITHSONIAN INSTITUTION. NOILNLILSNI NVINOSHLINS SAIuVaatT LIBRARIES SMI ae 5 Gly, ® 5 2 AY . 5 x = > © GL e > AWE > = P| co hig C4. a _ 2 A INS — 9 E 4 = ee oe o = WS" 2 = ; a 7) = an = 7 = bs z STITUTION NOILNIILSNI_ NVINOSHLINS, S3IYVUGIT_ LIBRARIES, SMITHSONIAN INSTITUTION NOILMLILSNI_ Nv | 5 Wy 2 = fy, 5 S z ee NS PB : S Z : Ay, : i3 ~ NX LY yt : MD ~WNX @ An EN oO VOM ae oO ac oO . BA . . AN $Y fi’ = Z. = Zz = ENS 3 | = NS 5 = 5 = 5 = NY 5 = a fee J1YVYGIT LIBRARIES SMITHSONIAN INSTITUTION NVINOSHIINS S31uvu gy LIBRARIES SMI Ww Ww oc < a 0 Bel NOILNLILSNI NOILALILSNI NOILALILSNI STITUTION NOILALILSNI NVINOSHLINS S3IYVYSIT LIBRARIES SMITHSONIAN INSTITUTION NOILNLILSNI NVI INSTITUTION Ssaluvugi INSTITUTION INSTITUTION INSTITUTION JIYVUYGIT_LIBRARIES SMITHSONIAN NOILALILSNI LIBRARIES SMI NVINOSHLINS S3INVUIT_ LIBRARIES a >. NVINOSHLINS S3SIYVYEIT LIBRARIES : z & R = g z g Z NS ) eg off bs : z AY SS Zz = A Lf ; Zz 4 Cf: nf Z 2 z NS Bi EOIR HiGi_ i& LLY 4 N ‘ yy; " Y X “jo = *N E 2% : sw = SN ” > 7) Fra Yn 2 ” . a, 4 ISTITUTION NOILNLILSNI NVINOSHLINS S3IY¥VYSIT LIBRARIES SMITHSONIAN INSTITUTION NOILONLILSNI NVI z «. f 2 7 z _ wT XS n ud « _ ms oc NA et a. oc . : BW : * : . oc LQ oe a a \ 5 2 a N ro 3 3 JIYVYdIT LIBRARIES SMITHSONIAN _INSTITUTION NOILALILSNI _NVINOSHLIWS S3lyvugiI1T LIBRARIES SMI : = -) > =) STITUTION NOILNLILSNI 1NVYGIT LIBRARI STITUTION NOILNLILSNI STITUTION NOILLNLILSNI 1yvugit * STITUTION buvaugit Nn wy Lia 2 eee 7 mee ar ORT Ee FY he ww . A eh ae ele Se id Re o- eum w a wo ae wn z uw & ul Fe wy a ul ‘Ss e 5 \. & = a 7 ce . c = c ty ING: - c = i Cc = a co So YW 3 ra S = S = Zz er) z a z a IS S3JIYVUSIT LIBRARIES SMITHSONIAN INSTITUTION NOJLALILSNI NVINOSHLINS S3IYVYEIT LIBRARIES = - = es re a s Fe fs) o 2 wo = o re 2 0 5 2 5. rand VS a = Bs = z a b = 0 “Yom 2 m B m = 7) <2 w = 7) . = 7) N NOILOLILSNI NVINOSHLINS S3IMVYEIT LIBRARIES SMITHSONIAN z 7) = ” z ne ” = b < = < eS = Ws = < = 4 » Pr Lens NY ASS z 4 a 4 = S a PENS F D z } 2 S Gt ¢ 6 SWC GC a O : 2% = z EN Z E : =o) © = > = Fa . 2 a 2 a » 2 a & iS Saiuvy¥aIt LIBRARIES SMITHSONIAN _ INSTITUTION NOILNLILSNI NVINOSHLINS S3luvUugiy — —_ (ip) — — us = oS M = us z oe a yg" “ye i ~ =i aS o = = EPipp x < = = = = ie’ age Se ee = > = z 77) = 2 “oO ie ra 7) Fa 7) N INSTITUTION NOILNLILSNI_NVINOSHLINS S3IYVYSIT LIBRARIES INSTITUTION ra 7 -—e ” > n = ” 7% u G = a = 2 = = ow = oc = w os a dpe < = _< a < an < S ac = = = a = a 2 = e 5 Ee 5 a ez = = el z a z a S S3J!I¥YVYUdEIT LIBRARIES SMITHSONIAN INSTITUTION NOILNLILSNI NVINOSHLINS S3IYVYUSIT LIBRARIES 6 ~ S ~ o aS 5 ~. = w == 00 om = x o o E 2 5 = = 2 NA 5 = zs > ¥ > re > ING = > = 2 - 2 = 2 \ = 2 e m z m 2 m 2 e z = a 7) be ee 7) N INSTITUTION NOILNLILSNI SAIYVYEIT LIBRARIES SMITHSONIAN INSTITUTION NOILNLILSN =z we n a n za we wn z f wn < SK 7 < = 1,¢8 \S = < . = = . = Zz 4K ye Z WS 4 = = SWS =. 5 = Gh ENS 2 5 oa a Go? ac O ysl xr NS \ } 2 o = 2 = = “iy = YS = = = b 1 : a 2 a a = c= js SaldvagdiT_ LIBRARIES SMITHSONIAN INSTITUTION NOILMLILSNI_NVINOSHLIWS S31YVUdIT_ LIBRARIES us a uw a us 2 Su 2 wz : ° = wz a ae 2 a Sy a Cc a c = oO re is 4 = GF WB « = = $ er = a o q - = . = az) 2) ; IN INSTITUTION > NOILALILSNI _NVINOSHLINS ~°4 luvud rh LIBRARI ES_ SMITHSONIAN” INSTITUTION = NOILNLILSNI o 2 BY = = S wy, = SI Ake = 2 ww 5 as E te = 32 iz 3 SY = ad = = = = = 2 WAY & » = 2 a ponstieas monte Arms meted Hi afleAe enh Maher ‘ we 4 i : F ‘ A : : A 7 4 TAP RaeMSM LA) Hed ADH RN REDS th: fills soszhs ayalfem 28 hot ‘ G 5 it gh : ; rar! ‘ 4 fetes B aK ‘ wiles Hed ehe i ie £ ‘ : WaMEhAMEP EDN AeA IN: AVE ee panagieay Ponca Hehe otc ida Nit hu Zit, MARS ERASE he RAR eT SED Fos Sadan Da ees Fs: edie 34x 7 . ‘i Boe Mates aR fous . at 1 i f iS pap oee eens hari heweiian barn Ait 862i) int te ares “ 5 H . a 3 ‘ Fn eM leeGE ie > P F 4 tit : - : ; Ser eee a : i A : WEL ¥a2 bt Fs heteetsOohehivamsbeeeg find Ti yyatatiaheding teh 3 Gov Bobb AML Gall ene dato nds Cah RiGteonewhite Ry satsloncactn we¥et ve cht al ebiehs fas abc AN Ces nocaitntsd fa mathe rerhs Pui nert Dahle ohae Haeie aay ; CHU het eebanedaiensis 7 5 Peat REE EE Madde eh a hee Oot etinsi2zirton sft i Bradlee meni nebote®s hefifeet Dif fad ther Pep bat . =Bneritesttoh Sorc ne Writ Rey PC rey fe yon f° MEER f shea e MED monatanecahlPahanatadansh tices} LAOUAOZE taney thhvebdults niManped he ne ORO : Dip aion nel eh " Dale Po un tne AE Er é K ERMA ter sia Finhi mt ; amatee Me hencud arBolte Oar e : , Gs 5 : : ‘ ; ahh aa Rea eH oH ad Bah sbynnndy Doce hayek abe fn gustste Meal ag aug AEE ANTES ot iat a peteelFaA Svanbe ses nes Wey ern BR2e ee om RENs Crbenetat abe neha i Coren ee Ailes hiodte at zh 4 i itpaide ot A OCr anne oAbheh fis eaten ry Ce Sree PR ian gee : i is ‘ Mancgir gna hy Fie tub staf ene? : é ‘ae Rina mA Mati sasian by Ioaiat reat ne Derg ede hone t PnP Our eh Gar hes E PARLE LE BL Boake O errr : b Daliew ani ny fink hee h ; 4 F deta tonne Re pene * eth ‘ cree, Canling. nantedrdietitoe i : abet ‘ : be Conse ediownty OsF Sardatedctnpe dys Bene tiiehanoh Ba Carradine i ved ; eg eee} Rip tnbesiee a straw e i He: eh atom Pata vie Mite Roby pore a i B : Cfo Dit bes ani Ae santo” : ; 4 Tenant beet 4 aie be 1 (Pade ohte eh Fe! ‘ cee i ere Shp s ares yee f : ‘ yeep ; pererentn ; : 4 ‘ : 1A Eee ; F “ suraey ake Thc ates : hav aey Pay : og" , a a] ; eytusye wes bE ; ; ; prerereneces) ! Torkeeraeeeicaye ea omega ‘ ° Pee terwinant : ThE : Sa rcewrenyea te Seri et) : Sena gayaentey ony oture ing, wu PUR QulPeRsetnts henneert nity es Ore Rene rmcn Nera Seen weurwe nature st 2) ste goaeyee R meiaimpewey SLENIE TN AY ity Panuperecyyy Tsar ayy luca Peay sustbire la acs 1g oe F 4 eign PPR NYRI mu yee Ra BY 4 o Sirah y alge ne Nes ' BURRESS a ete Gee Pacis epee tomas ceaey ‘ ert : , VAuyeweus TP ue a aie a . Hisyeune evs Gaels UATETVAVAL POP OM as | Ua gfe hawt Brkae eaya) ore a8 Ba ‘ . eae TRUSS LOU enteg Us He TYE Ne Tanger WC UPUee Teh Gtaeeb ape thal ey é . Tan is > ; af ‘ UU VER NTE aD PEE UND S ¥ elatate ‘ Se hey SPSL : Wiese ea TUS ae mya RA EEN VERE yoin Veale sy Vaeye gen sant 4 wt ter ‘ ir tte . 3: SEU Ae pH yeeeare pee ‘ fi Boh “ ee sete , x 3 Be rg v : : ruse yee , : ‘ mm ‘ . ; ‘ Hah rae pa é Paapeawy MUTE YNUT Cry EUS HSE EDP ALTE HERES ay weds ta cs eadtees ‘ 5 Wiese Paes: 1 Dea : P paalceeet hep teen NV NP MEN SOT OCT TREATS Ch Se RE LOY UaU SHG hy ensure ey yeN pts veh agas VO ays we fo : : te, 5 SGaewbarueyrus amgenxnnerauenih yey pheed VENY LATE OL WAR Woy Gy Tig dei sy Heras ap upe borat tw ’ in ve sath ‘ 2, th nie . : Beetercarr intent tiers CUE ELEN UF WoW WN BWP u ae wet eae) Hue gee ERIN 2 Malta vd ets Pere ve i oy Fi ah eee a ; i a Vorgrcatans ,weunaa Ss yha sanyensieata diets b eit aeh (eee) eek AN Tek ws eet ashae ty aah a : : ‘ Hos) Mere rector errrare! Ae Yen gree ea ra ms wyn 29 CONTE saeERtS ey ePiteealy PR ee Ti AR r f Ney : ™ 7 pape tra Ot en rtt fucwnen va ; ‘ Se ea yar ls ( , Ure Gn sare ms 7 j 4 , eyes ryay UENO re rey THD, “pee i Tb rQt aes tye Glget POM CHELON Livin ; tat ' tA : a rs j iceman VENMIIE PEST APUL gwar 2 Uefa Plate nah : , aL Nee PUTS PURSE SRES TUE REY AUF ALA EMU UE WOVE AYEE EEE BTW ENE ae Pr Ope WAWig a aye a i ia) : : 4 rine ‘ Sy ora ON Pare VEU CS Fy =H Lerten CoN Mene rer Us bett Peat OUDE QE APL WUE NLM wee thane haan bit a8 BAM Ag KI - : peer 8, MEU TLEUH MU TORTS PUENTE RY SHORE MED URES TST HORS OER ENE re ry Un Ee gr Ue , , \ Ceti} aL a) P . he alae PVR ESET HPP UNE W nant u ens ee OV 3 whe age ae oD i ne : ie r wees hy « 3 Se i ‘ Re eecnuianes A , ; Heyou sey Ter gre UNE , Vase 6 ‘ SEEM een oo : ar : x ‘ we Marys ecu sUOVOUTE TUAUEAC GT Osteen OO YEE CE worcaany Taypere ses : target . ; My a ry nacre ravers PSE pauress ara ss . ernest oe Nice Cay F eile ee ‘ tenes Faris yee” UrLeny ae evade Me resets Una iti a : : t ‘ i aie ies aioe Vrs tonnn on Cyr oats hott ‘ : one ; _ : On epuraiepIQONeyye Us CLT LTY: Vive : exons \ ‘ : ‘ act aren undeea ae TT NEEM Wet w eytaaee Ke yp igenies nH os 5 : seo enw Serr ws Wma) BER FUE TLS Pa UTYPSENEUE Ly Wentgnaee eas was bud Poco ay ay : fi 7 marr oNaRw we 2U Re ‘i Lim te gals sua dee a Cine . oo ’ a ‘ trea Sain PN FMA EoiF geoy eT TWY TAPE RS DENOTE ES PRD CE RX COAT APRS . ay al 1 ae yshes ane ye p tee tot at toe fad er bebeteerient CEM Rereere mnt ‘ vA ti . ’ 2 3 He : meen rays era rete FEE ween . OP ae boat UR) : . i Pa" Veen ser Ur ae? Yl Fy era rwee abe VIL HeE Wel Cee vines neath aVag see ' f : ' F eres una sub \=Wh 8 COUT aN EAN rRLG Mar anne ; ‘ : . ‘ 3 * Rete ie te sate ee hee FOP NAR PUNE UE VE yes ‘ Hehe 3 va : , F SEU TOPETUE NU PERE OaNt LURE IA VER UTR Z THEM AU DAE see DeES YS YR Tay Pricer tin Exe Petr mtreKnOn git Ae ey is , ; eG pncrpeerareramamibatucamsiyentene 1232824 are ‘ : é 7 wr eer year eas Bap aye RAHA LE d : : ‘ ‘ Praaveutenneyearsatacey CRE ErE U ‘ Rs : 6 é ' He = NYRR ON EAL yey RYLES NEUEN NTN ERA ‘ i 4 DPV EAENTY WY RBUCI PPV eT AT n : ‘ : wre Ssleraus eos evils: 1) f 4 4 f Whe, saraptieemene Te re aE 4 ar E ‘ wey Ferre ha ven evte RTE LAUT O ERTS saikety? “ 5 5 Singlet ieeet ed te atead Mia weg raentyet4s® ; ei Tethtetetiecte is ca ryriieue ubprvewey ‘ ath UE! t hay Seennzeas ery unws - i ; MER EDIE t , seeks : a F Pn ee rer ra wreuN erie yb YEU RYT 2 ¢ é : wateruprevenrnnertssteevsirstoee ii ; ep aay sr = : $ ’ ‘ 4 Agha ese TY rah Fae |e i i, i Bey resis eRe ys UNS: } PU IMIS PY FH 99D SDD 2 Sep DV PET’ ; 4 aruravererererase7aruivs 9: viyeNst2 ow 73) nett ‘ ans 7 inane wa 7378 =F YN = i ' 4 eee Dara esse tata E i Wyant . rage TD MERAY Sper AML OT UNI EN : ‘ 2 H H 3 : i ciate oenyapurnoyns ries) i rt ‘ Bin h eee Peal a eked ak ees ; ‘ F . seveadmnpnenmarvenenesean a ty : on A 4 z : periten Pere r ese 8 uP ; ieee Ven dy é Hf : i : 4 wit PE inal e A : , sireaneeaveriaeen enh ard-#1 2422) Me 5 V4 ! ey tx . 5 apasy a O7is PUIATLE 7% a ; ise fe naan ares WEN YS bres eta ‘ , : pen eye 7s a5 svi EAN Spe eat 2 3 \ Mi ‘ ‘ " Patt Desay i et tal 59 hers f : A % ite ng rey h are ath Tenn i : : : awe wesia : Cer fis a . i i ayrds Wel) fa t diay ay er 3; 4 3 rae POOH