ARE a ae hee A ht < bem iia da м eo PO ee De E AA a ad cad Batts t po 9 vayan AAA Аль A la bern RTL E IA a her re Ам fs et at au, А LR de AAA tet ] залы ме A Maha a LE Тм вал UE DE ne tel иль kp Ree Er ET RE RENT бел e MEA Boa O Sorte ya oO NE L TE he he eae Re Br lan Wi Pa D АВ x L'or “he Ber R Shs Kote fe bak ARRAS a wi Ал Avia beer Июн Мм LEE wad ee DEAS EL MMe ns BR Ml: wee RE ore млн SATS RE MATE TENTE Аа SAS ues Зла Ah ae NET Vena wir LATE NID ии a, ee A De à REN EME" Пим, А TELE мл нач MAN A 5 2 Pw ee Be AS PEN va eck 4% NER DIENTE SEEN Kane NEN NN na u ERTL Ta RI E ETES OA mee ма. COCA Sa ot os da Sl Vi a TORT RR, A ЗНА a Е Aare ма АА RN RR N RE RS re Or RN NN Rs чо мам AN San в але. eh Ba ate ee um DÉTENTE NEAR MN вы Reh EH A A. Aa III AA Arora к туда Es а НА a nity Ae. мА AE А EE ACER ON ESP Si PORN ee TRES лю es erin RENE NEG Se kere ам ringe wee lee a a UM ATA AER Ne wein SUN Kam EEA, nen we in Seer tae Bae AN ae SV SEMA OREN I RU fee end Hiern! Bent an АРЕН et Ernte ep iat Rage nee 5 whey DITES КРУ 2 ote ры FAAP AE ER A kg ie tig RA ern AIN ee see; Ads ve uy Peter ese te Re $ na: went nn cating met in eer rere ee ae A od AA E Не рн IE ВАРНЫ te НЕНИЯ + price 4 и 4 2 Apacs ЕЕ" POE PE MNO LO rg AA AA e AA AA AA ARI cere ed £0 ere yee, BE ULL PEA art, UT TS rat ne Re Fre rag AREA £ НИ Ду ФБ ODOR НЯ ИЕР pee gg LAA RA Y rn... AO NAAA rue AAA E RS RS ver ae were regen PURO ен He hee ГА 3 Ce or AN Ри Oe ath AAA Pr oo AA ARA ER NR ET II AI A SR A EIA PER e ee brea ра AE И ad ОС A ANI IA e eee Pap mg II COR UE Wee ee РИ Fhe D DH gg à OT bd Sp a Ve Oa, pate be ot ba POA Pde ER уч ME ERRE oe ET ee PURE CEE EEE EUR ECC ae rer lehnen фаре LESE BI Sl ER AA SI ААА А wus pra y eae gent dh ade e aime Ahead RET an he ver ne 2. A Ata .. ep UP SEO a ts a der SO e + run per CORAN TES RARE ECOLES мы НАМ ПИТ Ferse tm dcden Fa sud ds $ 4 RADA ait gad hors ve pd CRE M et See EEE ОСИ OE ONS bea ие мм РЕ Lee ve IT Е $ ни пу EA : Fs a dears CROIENT Faye tied ge ge Аа ва а à Г АА, ВЫ ИРИ EI e ware АНН z Art НИ бер Pur atin tt wean ия РНР DE CORAN e. +.s A RN et CS O AAA A дв Pp SZ Mt ghia a dt a gh al PPP TE Ge PRT ee BA oe Pode A ORME RU Веди ве PR nr CE aes ee oe OCEAN И rose ta gras gi few р: vers ПАЛАТЫ ui РТ bb Ede i , SON ЛАНИТ ти ранние ро НОС ine + É El "er ke ee HE Pit HIDE MA Eu rt ЧЕТ O gare dope da à a en een ERDE DO ES is tye eu Qu # о нина ay ho are nf ar bs DE ee ghee an rt La аще pie № s ель a Da kop Lf REM AR Bee ре лу д" ane sy wed 3; QE VELO м AIT rie о Г Put HAS ARC ENTER as € sé 4 pneus “ioe Peer ren) RAA LT A ju QUE SUEDE 27 enh wi эр CAES ЕС r À ne y Y (ME m. CALE SUN ee перелетов ОХ Y : ы нев opie tabs Я watt ee EC eh Е an AR PESE NR] ARO A] LER, АНИ re bt oss + ON pte dus eg sie ay 48 Rip pit vr hf: Ford taht РНР, р. И ях | ae o | : 4 LTE RI SENT TEE RAT) ит, bo » oe, : iting honk > Ир нм ! . 15 Da PER AE IR i у 3 | ae м, te в к ck И path ae ee ‘ ag f ИЕН Sa at ; ' ЕТ, : 4 ee Pano he р , 14 . я я и Eh { LE y ne Ae E st Be fits sup р RR FELSEN Е Be ’ DOREEN BER SCH LOTS ; q te ret add a 1 НУ HWY bes etry ad A iS р ФИН ее ОС ran I вез АЕ и ak pak 0 Е a р аз, + AS eE ee DORE ER ETES м t y We ARRETE р НЯ, REKEN 4 viva ‘ de cn Pa наи A AE: ИЯ rine ! e dd A due ВАЛ EM Avie Gi eek НФ HARVARD UNIVERSITY e Library of the Museum of Comparative Zoology Br NR a VOL. 17 1978 MALACOLOGIA International Journal of Malacology Revista Internacional de Malacologia Journal International de Malacologie Международный Журнал Малакологии Internationale Malakologische Zeitschrift Publication dates Vol. 16, No. 2—17 September 1977 Vol. 17, No. 1—17 February 1978 MALACOLOGIA, VOL. 17 CONTENTS A. J. CAIN The deployment of operculate land snails in relation to shape ANCIRS ZERO RS NE NES ARS TT A Gl ОЕ Wilt ee ed ia ais 207 P. CALOW The evolution of life-cycle strategies in fresh-water gastropods ......... 351 М. В. CARRIKER and L. С. WILLIAMS The chemical mechanism of shell dissolution by predatory boring gastropods: a review and an hypothesis.................... 143 M. R. CARRIKER, L. G. WILLIAMS and D. VAN ZANDT Preliminary characterization of the secretion of the acces- sory boring organ of the shell-penetrating muricid gastropod COS DIRE CINERCA face Patrols a e ne А a ba 125 5. М. CHAMBERS Ап electrophoretically detected sibling species of ““Goniobasis floridensis’’(Mesogastropoda: Pleuroceridae)...................... 157 G. M. DAVIS Introduction. American Malacological Union - Systematics As- sociation Symposium Proceedings: Evolution and Adaptive Radiation of Mollusca: 12-13 July 19777 Naples, Florida ....... 20204. 00 163 H. HEATWOLE and A. HEATWOLE Ecology of the Puerto Rican camaenid tree-snails . ................. 241 K. E. HOAGLAND Protandry and the evolution of environmentally-mediated sex change. astway of the Mollusca: o a u ae ae сен 365 R. M. LINSLEY Locomotion rates and shell form in the Gastropoda ................. 193 M. MOUEZA et L. FRENKIEL Le systeme circulatoire et le jeu des siphons chez Donax zruUmeu/ us“ Mollusaue Eamellibranche: tl 2... Er a m ER ae: 117 J. MURRAY and B. CLARKE Changes of gene frequency in Cepaea nemoralis over fifty VER See О еее... 317. G. S. OXFORD The nature and distribution of food-induced esterases in ANSE TCH ESI ANUS men ed 331 Е. PERRON The habitat and feeding behavior of the wentletrap Ep/tonium BCE EU о KORE ee AA eu AS A 63 MALACOLOGIA CONTENTS (cont.) С. $. RICHARDS Genetic studies on Biomphalaria straminea: occurrence of a fourth allele of a gene determining pigmentation variations ........ N. W. RUNHAM Reproduction and its control in Deroceras reticulatum........... A. H. SCHELTEMA Position of the class Aplacophora in the phylum Mollusca......... A. A. SHILEYKO On the systematics of Trichia $. lat. (Pulmonata: Helicoidea: Hygromndael. 7:3... abscess ое eae eee eae V. A. VAIL Seasonal reproductive patterns in 3 viviparid gastropods ......... С. J. VERMEIJ and J. А. VEIL A latitudinal pattern in bivalve shell gaping .................. D. S. WOODRUFF Evolution and adaptive radiation of Cerion: a remarkably di- verse group of West Indian land snails. : . ..5. oc. ooo A Е. №. YOCHELSON An alternative approach to the interpretation of the phylogeny of 'antient MOIMMISKS* os 2. 5 oo ооо eee LE e. he к MUS. COMP. ZOOL. 7 Air LIBRARY FEB 27 1978 HARVARD UNIVERSITY 1978 ALACOLOGIA International Journal of Malacology г 3 _ Revista Internacional de Malacologia Journal International de Malacologie ” Международный Журнал Малакологии Internationale Malakologische Zeitschrift MALACOLOGIA Editors-in-Chief: GEORGE М. DAVIS ROBERT ROBERTSON Editorial and Subscription Offices: Department of Malacology The Academy of Natural Sciences of Philadelphia Nineteenth Street and the Parkway Philadelphia, Pennsylvania 19103, U.S.A. Associate Editors: JOHN B. BURCH University of Michigan, Ann Arbor ANNE GISMANN Maadi, A. R. Egypt Editorial Assistants: JUDITH DIAMONDSTONE LYNN HARTLEY SUSAN MILIUS MALACOLOGIA is published by the INSTITUTE OF MALACOLOGY (2415 South Circle Drive, Ann Arbor, Michigan 48103, U.S.A.), the Sponsor Members of which (also servingas ed- itors) are: J. FRANCES ALLEN, Emeritus Environmental Protection Agency Washington, D.C. CHRISTOPHER J. BAYNE, Vice-President ELMER G. BERRY, Emeritus Germantown, Maryland KENNETH J. BOSS Museum of Comparative Zoology Cambridge, Massachusetts JOHN B. BURCH, President MELBOURNE R. CARRIKER University of Delaware, Lewes GEORGE M. DAVIS, Executive Secretary-Treasurer ROBERT ROBERTSON CLYDE Е. Е. ROPER, President-Elect Smithsonian Institution Washington, D.C. W. D. RUSSELL-HUNTER Syracuse University, New York NORMAN F. SOHL United States Geological Survey Washington, D. C. RUTH О. TURNER, Alternate Museum of Comparative Zoology Cambridge, Massachusetts SHI-KUEI WU University of Colorado Museum, Boulder Institute meetings are held the first Friday in December each year at a convenient place. One subscriber may attend and vote by petitioning in advance. For information, address the President. Copyright, © Institute of Malacology, 1978 J. А. ALLEN Marine Biological Station, Millport, United Kingdom Е. Е. BINDER Museum d'Histoire Naturelle Genève, Switzerland me HeIGLARKE, Jr: National Museum of Natural History Washington, D.C., U.S.A. E. S. DEMIAN Ain Shams University Cairo, A. R. Egypt С. J. DUNCAN University of Liverpool United Kingdom Z. A. FILATOVA Institute of Oceanology Moscow, U.S.S.R. E. FISCHER-PIETTE Museum National d’Histoire Naturelle Paris, France A. FRANC L’Universite Paris, France м. EREITER University of Reading United Kingdom E. GITTENBERGER Rijksmuseum van Natuurlijke Historie Leiden, Netherlands A. N. GOLIKOV Zoological Institute Leningrad, U.S.S.R. A. V. GROSSU Universitatea Bucuresti Romania T. HABE National Science Museum Tokyo, Japan A. D. HARRISON University of Waterloo Ontario, Canada K. HATAI Tohoku University Sendai, Japan EDITORIAL BOARD B. HUBENDICK Naturhistoriska Museet Göteborg, Sweden A. M. KEEN Stanford University California, U.S.A. R. N. KILBURN Natal Museum Pietermaritzburg, South Africa M. A. KLAPPENBACH Museo Nacional de Historia Natural Montevideo, Uruguay J. KNUDSEN Zoologisk Institut & Museum Kdbenhavn, Denmark A. J. KOHN University of Washington Seattle, U.S.A. Y. KONDO Bernice P. Bishop Museum Honolulu, Hawaii, U.S.A. СМ. EAE McGill University Montreal, Canada JALEVER Amsterdam, Netherlands Ст. O National Taiwan University Taipei A. LUCAS Faculte des Sciences Brest, France N. MACAROVICI Universitatea “Al. I. Сига” lasi, Romania C. MEIER-BROOK Tropenmedizinisches Institut Tübingen, Germany (Federal Republic) H. K. MIENIS Hebrew University of Jerusalem Israel J. Е. MORTON The University Auckland, New Zealand В. NATARAJAN Marine Biological Station Porto Novo, India J. GKLAND University of Oslo Norway T. OKUTANI Tokai Regional Fisheries Research Labora- tory Tokyo, Japan W. L. PARAENSE Universidade de Brasilia Brazil J. J. PARODIZ Carnegie Museum Pittsburgh, U.S.A. С. М. PATTERSON University of Michigan Ann Arbor, U.S.A. W. F. PONDER Australian Museum Sydney A. W. B. POWELL Auckland Institute & Museum New ‘Zealand В. О. PURCHON Chelsea College of Science & Technology London, United Kingdom C. P. RAVEN Rijksuniversiteit Utrecht, Netherlands O. RAVERA Euratom Ispra, Italy N. W. RUNHAM University College of North Wales Bangor, United Kingdom 5. © SEGERSTRALE Institute of Marine Research Helsinski, Finland С. А. SOLEM Field Museum of Natural History Chicago, U.S.A. F. STARMUHLNER Zoologisches Institut der Universitat Wien, Austria W. STREIFF Université de Caen France J. STUARDO Universidad de Chile, Valparaiso Т. Е. THOMPSON University of Bristol United Kingdom Е. TOFFOLETTO Societa Malacologica Italiana Milano М. $. S. VAN BENTHEM JUTTING Domburg, Netherlands J. A. VAN EEDEN Potchefstroom University South Africa J.-J. VAN MOL Universite Libre de Bruxelles Belgium B. R. WILSON Western Australian Museum Perth C. M. YONGE Edinburgh, United Kingdom H. ZEISSLER Leipzig, Germany (Democratic Republic) A. ZILCH Natur-Museum und Forschungs-Institut Senckenberg Frankfurt-am-Main, Germany (Federal Republic) MALACOLOGIA, 1978, 17(1):1-56 ON THE SYSTEMATICS OF TRICHIA $. LAT. (PULMONATA: HELICOIDEA: HYGROMIIDAE) А. A. Shileyko! Institute of Biology and Soil Science, Far East Scientific Centre, USSR Academy of Sciences, Vladivostok, USSR ABSTRACT Systematic revision of Asiatic, Caucasian and some European species of the Trichia $. lat. group of the Trichiinae has shown that it is not a united genus but a complex of genera. All known representatives (8) of the Central Asian group, all species but 1 occurring in the Caucasus (14), and 9 representatives of the main European taxa are treated in the Systematic Part (31 species). The anatomical investigation comprises 29 species, including 25 of the 27 species known in the USSR. Of the 9 genera recognized, 3 are new: Hygrohelicopsis, Teberdinia, Plicuteria, and also 4 species: Leucozonella сапа, Hygrohelicopsis darevskii, Kokotschashvilia tanta, K. eberhardi. Apart from shell and traditional anatomical characters, great attention has been paid to peculiarities of the inner structure of the genitalia. It was thus found that Hygrohelicopsis darevskii does indeed possess 2 pairs of dart sacs, though the inner pair is not visible from the outside. In this species the right ocular retractor does not pass between penis and vagina, as in Trichia, but only near them, which is considered to be characteristic for He/icopsis in the “Helicellinae”” auct.; but from the totality of other characters Hygrohelicopsis is nevertheless regarded as belonging in the Trichiinae. Teberdinia was found to differ from other Caucasian forms by having a deep longitudinal groove on the surface of the penis papilla that separated off a lobe and from all other forms by having one of the intrapapillar cavities reaching into this lobe. In the genus Plicuteria the longitudinal vaginal plicae are subdivided by regular transverse prismatic folds forming a unique dense pattern on the internal wall of the vagina. Kokotschashvilia tanta and K. eberhardi can be distinguished, among other features, by the structure of the inner wall of the penis papilla, which in the latter is not smooth but plicate. The Asiatic group is distinguished from the European-Caucasian group by the mode of formation of the intrapapillar cavity or cavity system. In the former group it arises from the closing of a groove on the surface of the verge, the main phases of the process being evident in the material studied. In the latter group the cavities arise as paired structures in the thickness of the papillar wall and then assume more complex forms. The proposed new systematics of the Trichia $. lat. group are based on various characters of a different nature, taking into consideration as far as possible the organization of the animal as a whole, as only such an analysis can reflect on the evolution of the various groups and the true relations among them. INTRODUCTION The use of gross genital morphology in stylommatophoran pulmonates has helped Our understanding of systematics at the family level. Findings reflect relationships as they have evolved in nature. In practice, however, this approach has sometimes re- sulted in the creation of large genera com- prising the anatomical characters of the many species included. One result is that the conchological features of such genera have become so hazy as to be quite useless. The genus Trichia $. lat. (Trichiinae)2 is very interesting in this respect. A concho- logical diagnosis of the group is not practi- 1Present address: Benthos Laboratory, Institute of Oceanology, USSR Academy of Sciences, Krassikova ul. 23, Moscow 117218, USSR. 2The author, replacing a classification that recognizes 2 subfamilies, the Hygromiinae and Helicellinae, subdivides (Shileyko, 1972) the family Hygromiidae Tryon, 1866, into 7 subfamilies: Trichiinae Zilch & Jaeckel, 1962 Hygromiinae Tryon, 1866 Metafruticicolinae Shileyko, 1972 Monachinae Zilch, 1960 Cochlicellinae Shileyko, 1972 Ciliellinae (?) Shileyko, 1972 Geomitrinae Wenz, 1923 The “Helicellinae'” аист. are not considered to form a natural group but to consist of various hygromiid “vital forms” (Shileyko, 1972: 28-41). 2 SHILEYKO cable. The only diagnostic feature is the presence of 2 pairs of symmetrically situated dart sacs (stylophores). Other ana- tomical characters such as the shape and relative length of the flagellum, epiphallus, penis and spermathecal duct; the shape of the receptaculum seminis (spermatheca) and of the vagina; the number, location and nature of mucous glands and of their branches; and the presence of bends in the free oviduct are rather variable, and none of them is constant within the whole group. But one anatomical feature allows us to separate the present group from species of the genus Helicopsis, which be- longs in the “Helicellinae”” auct.: the right ocular retractor passes between the penis and vagina, not merely close to them, as in Helicopsis. This feature and its taxonomic value will be discussed in more detail be- low. The present paper shows that the genus Trichia is more complex than hitherto imag- ined. The probable phylogenetic relationships in the genus complex will be discussed after a systematic review of the species studied. This study includes the majority of species known from the USSR and also some eastern and central European forms allied to those from the European part of the Soviet Union. A total of 31 species that fall into 9 genera were considered as follows; 2 of these, marked by asterisks, were not examined. Genus Odontotrema Lindholm, 1927 O. diplodon Lindholm, 1927 Genus Leucozonella Lindholm, 1927 L. ferghanica Lindholm, 1927 L. caryodes (Westerlund, 1896) L. rubens (Martens, 1874) L. mesoleuca (Martens, 1882)* L. rufispira (Martens, 1874) L. retteri (Rosen, 1897) L. caria Shileyko, sp. nov. Genus Hygrohelicopsis Shileyko, gen. nov. Н. darevskii Shileyko, sp. nov. Genus Teberdinia Shileyko, gen. nov. T. zolotarevi (Lindholm, 1913) Genus Kokotschashvilia Hudec & Lezhawa, K. makvalae (Hudec & Lezhawa, 1969) K. tanta Shileyko, sp. nov. K. holotricha (Boettger, 1884) K. eberhardi Shileyko, sp. nov. K. phaeolaema (Boettger, 1886) Genus Caucasigena Lindholm, 1927 Subgenus Caucasigena $. str. С. (С.) armeniaca (L. Pfeiffer, 1846) C. (C.) tschetschenica (Retowski, 1914) С. (С.) rengarteni (Lindholm, 1913) С. (С.) eichwaldi (L. Pfeiffer, 1846) С. (C.) abchasica (Lindholm, 1927) * Subgenus Anoplitella Lindholm, 1929 С. (A.) schaposchnikovi (Rosen, 1911) Subgenus Dioscuria Lindholm, 1927 C. (D.) thalestris (Lindholm, 1927) Genus Plicuteria Shileyko, gen. nov. P. lubomirskii (Slossarski, 1881) Genus Trichia Hartmann, 1840 Subgenus Petasina Beck, 1847 Т. (Р.) unidentata (Draparnaud, 1805) Subgenus Trichia $. str. Т. (Т.) plebeia (Draparnaud, 1805) Т. (Т.) concinna (Jeffreys, 1862) T. (Т.) hispida (Linne, 1758) Т. (Т.) villosula (Rossmaessler, 1838) Т. (Т.) striolata (С. Pfeiffer, 1828) Т. (Т.) danubialis (Clessin, 1874) Genus Edentiella Polinski, 1929 Е. bakowskii (Polinski, 1924) This material was collected from the following geographical areas (Fig. 1): Odon- totrema and Leucozonella from Central Asia west of the great Tian-Shan! Mountains (Fig. 2); Hygrohelicopsis, Teberdinia, Ko- kotschasvilia and Caucasigena from the Caucasus (Fig. 3); Plicuteria, Trichia and Edentiella from various European countries (Fig. 4). SYSTEMATIC PART In the following text, shell descriptions are given only for the little-known, rare ог new species. The details of internal anatomy are mostly those relevant to taxo- nomic analysis. The characters of proximal reproductive structures do not distinguish genera and species; therefore the ovotestis, spermoviduct and albumen gland are not described. The terminology of the features of the reproductive tract is not uniform in the literature, and some features have not been previously used in taxonomic distinction; TRICHIA SYSTEMATICS x A \ MAN Y Вы > О Е = Agee | a 6 er e \ a £ ) ag TAN а 1 ER 4 2 | / = Le LT; с SEA PA ‹ 1 ( ane? if 5 с ppt) FINL / ae’ ~ I or ¿EP Lo == f SUR й Ss a È Oo se ` % “is / - Su | `? ES z oe ES \ Ах —— LT E ВР № / \ Y à > > si © \ % \ U 5 S R > = aufs © < FIG. 1. Map showing the 3 general areas in which the species discussed were collected. 4 SHILEYKO Diushambe a e = + wil Lake Balkash T RS eee, Чу, tA of ES TS oo FIG. 2. Inset A of Fig. 1, the source area of central Asian specimens. This map shows the locations of Central Asian Trichia s.l. collected in the Soviet Union (in the Kirghiz, Chatkal, Talas, and Hissar Mountains, i.e., in the western spurs of the Tien-Shan Mountains. Black circles = Odontotrema; black triangles = Leucozonella spp. Note: kul = lake; Uzbek SSR = Uzbekistan; Tadzhik SSR = Tadzhikistan, etc. therefore, certain terms used in this paper are defined and attention is drawn to par- ticular features. The free oviduct, i.e., that part of the female tract that continues the oviduct or uterine part of the spermoviduct, which runs from the point of departure of the vas deferens to that of the spermathecal duct, is here called “oviduct'””; “bend of the oviduct”” is used in specific differentiation. The following part of the tract, the vagina, is subdivided into an upper and lower portion: the upper lies in the mucous gland region above the dart sacs, and the lower included the dart sacs and continues to the genital atrium. The ‘‘dart sacs,’’ which are accessory Organs, are either rudimentary or contain lime shafts; ‘‘stylophore’’ is perhaps a shorter and more exact term for dart sac. In the distal part of the male genital tract several features have been accorded special attention. The penis consists of the “penis sheath’ and “рарШа” (‘’verge’’). In- side the walls of the inner verge surround- ing the seminal canal we find one or more “intrapapillar cavities,” i.e., a system of cavities, some of which may be separated by longitudinal tissue bands or by longi- tudinal septa. These communicate with the cavity of the penis sheath by a foramen, the “papillar lacuna’’ (or by several lacu- nae). The “papillar plicae,”” or folds, are ridges on the inner surface of the penis sheath. A “connective tissue penial mem- brane,” i.e., a thin membrane that may contain muscle fibers, may be stretched among the vas deferens, epiphallus and penis. Between the epiphallus and the distal parts of the penis sheath there may be groups of muscle fibers, here termed “penial muscle bands,’ either alone or associated with bands of connective tissue, the “connective tissue penial bands.” The term “perspective”” applied to the umbilicus or perforation of the shell de- notes that the whorls can be seen in widely or narrowly umbilicate or even perforate shells. TRICHIA SYSTEMATICS 5 Novy Aphon Sukhumi GEORGIA BL ACK (Grusinian SSR) 2 Stepanavan $ Lamy, TURKEY FIG. 3. Inset B of Fig. 1, the source area of specimens from the Caucasus. The locations of Caucasian Trichia s.|. collected in the Soviet Union. Open square = Teberdinia; black square = Caucasigena spp; open triangle = Kokotschashvilia spp; black triangle = Hygrohelicopsis. The crosses indicate that exact localities are not known but are located by district only. LIST OF ABBREVIATIONS AIG albumen gland р penis BO bend of oviduct ВЕ penial folds (plicae) CML* central mantle lobe PGr papillar groove (intrapapillar CPB connective tissue penial band cavity that still remains open) CPM connective tissue penial mem- РЕ papillar lacuna brane PMB penial muscle band D dart (stylet) PP penis papilla (verge) DS dart sac (stylophore) Pr prostate EDS outer (external) dart sac RML* right mantle lobe Ep epiphallus ROT right ocular retractor F flagellum RP retractor (muscle) of penis GA genital atrium SC seminal canal (duct) HD hermaphrodite duct SD spermathecal duct IA intrapenial appendix SIC septa in intrapapillar cavity IC intrapapillar cavity SOD spermoviduct IDS inner (upper) dart sac Sp spermatheca LML* left mantle lobe V vagina (lower portion) MFE main folds of epiphallus VD vas deferens MG mucous glands VF vaginal folds (plicae) Ov oviduct *The mantle collar and its lobes have been figured for some species, for what it is worth, but the feature is not further discussed as the material is not sufficient for any conclusions. SHILEYKO "sııqnday |елэрэ- uewisg = 449 “91/qnday эцпелэоцэ UeWIAaD = Yq ‘e!4eAo|soysazD = HSS9 ‘емазптона = э!бие!аз 398/q ‘dds е2/42//1 = sajfuei1] uado /е//азиару = aienbs Yoeıg ‘5иэ\ 294$ ueadoinz jo Bale ээ4п0$ au} ‘р ‘614 yo 9 1э$4| “py “Ol now AMVONNH CT = Viyisnv Á 4 52 À SIO” eneıstreid Yo y ie 2NOWOTO an ¿e > y e (o) ` Ve YIH33Z 9 SES 2N3PX4 E” < at ETATS Зт2атэт QNV10d if 31n MESIEM @ ЧитТрепо TRICHIA SYSTEMATICS 7 Genus Odontotrema Lindholm, 1927 The shell is lowly conical with a smooth angle on the periphery; it is finely hirsute and brown. The umbilicus is relatively wide and perspective. There is a fine lip in the aperture on which there are 2 teeth: a larger basopalatal tooth and a basocolumel- lar one. The dart sacs are club-shaped and weakly fused together. There are 4 un- branched mucous glands. The inner surface of the penis sheath is covered with clear- cut, fluted ridges; the surface of the verge bears a longitudinal groove. There are no cavities in the verge walls. The flagellum is a little shorter than the epiphallus. Genus monotypic. Odontotrema diplodon Lindholm, 1927 Figs. 5-9; Pl. I, 1 Two specimens were examined. These were collected from the scree at the base of the Chatkal Range, NW Tian-Shan Mountains, Kirghiz SSR, central Asia, in May 1972, and identified by me. The characteristic features of the inter- nal anatomy are as follows. The oviduct (uterus) part of the spermoviduct passes straight into the free oviduct without any curve. The 4 mucous glands are simple, nonbranching and radially arranged. The dart sacs are thin and long, the inner dart sacs fused with the outer at their base; the inner dart sacs are almost free of the vagina. The lower vagina is long, fusiform; slight longitudinal plicae run within the whole of the inner vagina, the “vaginal plicae.”” At the outlet of the dart sacs these plicae form in distinct lobes. The flagellum is shorter than the epiphallus; the latter curves twice and is held in this position by connective tissue bands containing muscle fibers. The penis is relatively very massive, fusiform; its inner surface bears branching, slightly crimped, long ridges. The verge is generally cylindrical, with a long, deep groove which disappears distally. The distal part of the verge is constricted by an incomplete circular groove. Except for the vas deferens (seminal canal), the verge does not contain any cavities. There are some long, weak connective tissue bands on the surface of the penis sheath. The spermathe- cal duct (truncus receptaculi) has no abrupt curves and passes smoothly to the elon- FIGS. 5-9. Odontotrema diplodon Lindholm, Chatkal range, NW Tian-Shan Mountains, Kirghiz SSR, May 1972. 5, reproductive tract; 6, penis, penis sheath partly removed; 7, cross-section of verge; 8, cross-section of epiphallus; 9, inner structure of vagina in dart sac region. 8 SHILEYKO gate-oval spermatheca (receptaculum semi- nis), which nearly reaches the lower edge of the albumen gland. Genus Leucozonella Lindholm, 1927 The shell is globose to lowly conical; in the latter case it may be angular at the periphery. Its color varies from light gray, yellowish, reddish to brown; оп the ре- riphery there is a light line, which is some- times very faint. The umbilicus may vary from dot-like to relatively wide. Sometimes it is half covered by the reflected columel- lar edge of the aperture. The dart sacs are globose or elongate. There are 3 or 4 mucous glands, having 2-3 branches. The inner surface of the penis sheath is smooth. The verge has a long groove, developed to various degrees, or it is closed, forming an intrapapillar cavity. Type-species: Helix 1874. Martens, rubens Leucozonella ferghanica Lindholm, 1927 Figs 10-75;,Pl. 1; 2 Two specimens were examined. They were collected in the Sary-Chileck Nature Reserve near Lake Kula-Kul, Chatkal Range, NW Tian-Shan Mountains, Kirghiz SSR, on 6 July 1966, by A. J. Jankowskaja and identified by I. M. Likharev. A descrip- tion of the shell is given by Likharev & Rammelmeyer (1952). The oviduct forms 2 sharp bends, and the walls of the tube are tightly pressed together (Fig. 10). The length of the straight part of the tube, the upper vagina, from the bend to the dart sacs, is 6-7 times its width. The dart sacs are globose; the sac region is separated from the lower vagina by a slight narrowing. The flagellum is thin, slightly longer than the epiphallus. The penis is very bulbous, fusiform. The verge bears a groove on its basal part (Figs. 12-14). The spermathecal duct is almost FIGS. 10-15. Leucozonella ferghanica Lindholm, Sary-Chileck Nature Reserve, Chatkal range, NW Tian-Shan Mountains, Kirghiz SSR, 6 July 1966. 10, reproductive tract; 11, inner structure of vagina in dart sac region; 12, penis, penis sheath partly removed; 13, 14, cross-sections of verge at different levels; 15, cross-section of epiphallus. TRICHIA SYSTEMATICS 9 straight and ends in a small spermatheca nearly spherical in form, which almost reaches the albumen gland. Leucozonella caryodes (Westerlund, 1896) Figs. 16-20; Pl. I, 3 Four specimens were examined. | col- lected them from the Talas Range, NW Tian-Shan Mountains, Kirghiz SSR, on 4 June 1972, and identified them. This species differs from L. rubens by its shell, which has considerably thicker walls and a narrower umbilicus (see PI. 1, 3b and 4b). The oviduct forms a loop-like bend (Fig. 6). Usually there are 3 mucous glands, each with 2 branches. The distance between the basal part of the mucous glands and the upper limit of the dart sacs is approxi- mately equal to the length of the lower vagina. The dart sacs are not globose but elongate, almost club-shaped. The lower vagina is thin; it is 2.5-4 times longer than it is wide. Vaginal plicae are very massive. The lobes at the outlet of the dart sac ducts are not well expressed. The flagellum 1 см is long, longer than the fine, curved, су- lindrical epiphallus. The different portions of the epiphallus are connected by short connective tissue bands (CPB, Fig. 16); there are also longitudinal muscle bands (PMB) on the surface of the fusiform penis. The inner surface of the penis sheath is smooth. The verge bears a sharp, long groove, the plane of which forms an acute angle with the sagittal papilla plane (Figs. 18, 20). The penis retractor loops around the epiphallus. The spermathecal duct is thin and slightly curved and merges indistinctly with the elongate-oval receptaculum seminis. The latter does not quite reach the albumen gland. Leucozonella rubens (Martens, 1874) Figs. 21-26; Pl. I, 4 Five specimens from the foothills of the Kirghiz Range (formerly Alexander Mountains), NW Tian-Shan Mountains, Kirghiz SSR, were examıned; collection (15 June 1972) and identification are mine. This shell is very much like that of “Euomphalia”” regeliana’ (Martens, 1882) FIGS. 16-20. Leucozonella caryodes (Westerlund), Talas range, NW Tian-Shan Mountains, Kirghiz SSR, 4 June 1972. 16, reproductive tract; 17, mantle collar with 3 lobes; 18, penis, penis sheath partly removed; 19, cross-section of epiphallus; 20, cross-section of verge. 10 SHILEYKO 22 FIGS. 21-26. Leucozonella rubens (Martens), foothills of the Kirghiz range, NW Tian-Shan Mountains, Kirghiz SSR, 15 June 1972. 21, reproductive tract; 22, penis, penis sheath partly removed; 23, mantle collar; 24, inner structure of vagina in dart sac region; 25, cross-section of verge; 26, cross-section of epiphallus. (cf. Pl. I, 4a, b, c and PI. Il, 5a, b, с); the shell of the latter species has a narrower umbilicus and in this respect it resembles L. caryodes, but it differs from it by being thinner. Nevertheless, despite these differ- ences among the 3 species, positive identifi- cation is possible only after dissection. The oviduct is rather narrow and long and bends suddenly (Fig. 21). There are 3-4 mucous glands, situated considerably higher than the dart sacs; each gland has 2-5 branches. The outer dart sacs are con- siderably larger than the inner ones; all are nearly spherical. The dart sac region tapers toward the vagina without any sudden nar- rowing. The vaginal plicae as well as the lamellae on them are indistinct. The flagel- lum is 2/5 to 1/4 the length of the epiphal- lus. The penis is fusiform to globose; the inside of the penis sheath is smooth, with- out any folds or lamellae. The verge bears a narrow, sharp, deeply incised groove on its surface. At the base of the papilla there is an appendix (ТА, Fig. 22), developed to various degrees, in the form of a protuber- ance, of a conical callus or (when maxi- mally developed) of a spongy lamina, partly superimposed upon the papilla. There is a connective tissue membrane between the penis and the epiphallus in this species. The spermathecal duct is thin and rather long; it ends in an oval sperma- theca that reaches the lower edge of the albumen gland. Leucozonella rufispira (Martens, 1874) Figs. 27-31; PI. Il, 6 Four specimens were examined. They were collected from the Anzob Pass, Hissar Range, W Tian-Shan Mountains, Tadzhik SSR, on 28 July 1968, by Z. Izzatulaev and identified by him. The oviduct forms a sudden bend, with the inner walls touching. As a rule there are 3-4 mucous glands sited around the upper vagina, all usually having 2 branches. The dart sacs are very massive and spheri- cal, and the inner ones are not as short as the outer ones. The length of the lower vagina exceeds its width 3-4 times. The vaginal plicae are rather clear and show TRICHIA SYSTEMATICS 11 FIGS. 27-31. Leucozonella rufispira (Martens), Anzob pass, Hissar Range, W Tian-Shan Mountains, Tadzhik SSR, 28 July 1968. 27, reproductive tract; 28, penis, penis sheath partly removed; 28, cross-section of verge; 30, cross-section of epiphallus; 31, mantle collar. distinct laminae. The flagellum is a little shorter than the sharply curving and slightly coiling epiphallus. The loop of the epiphallus is drawn to the penis sheath by connective tissue bands (CPB, Fig. 27). The penis is globose, and the inner surface of its sheath is smooth. The verge is fusiform and has a very deep, long, longitudinal groove. In addition, there is а transverse groove not wholly encircling the proximal part of the papilla. The papilla wall has no Cavities. The spermathecal duct is thin and almost straight, ending in an oval recep- taculum seminis and stopping considerably short of the lower edge of the albumen gland. Our observations differ somewhat from those of Likharev & Starobogatov (1967). According to these authors, there are 2 mucous glands on the very bend of the oviduct and the flagellum is 2/5 the length of the epiphallus. Leucozonella retteri (Rosen, 1897) Figs. 32-36; Pl. Il, 7 Four specimens were examined; they were collected in Kandara Valley, Hissar Range, W Tian-Shan Mountains, Tadzhik SSR, on 2 July 1967, by 2. Izzatulaev and identified by him. The oviduct forms only a slight bend. There are 4 mucous glands; they usually have 2 branches (1 specimen had 1 simple gland, Fig. 32). The dart sacs are massive, inflated, globose. The lower vagina is set off from the region of the dart sacs by a marked narrowing; it is long, narrow and cylindrical; its length is 5-6 times its width. The vaginal plicae are not distinct and form clear-cut lobes only at the entrance of the dart sacs. The flagellum is about as long as the epiphallus; the latter curves twice. There is a connective membrane between penis and epiphallus (CPM, Fig. 32); con- nective tissue bands also run along the penis sheath surface. The penis is massive, fusiform. The verge has a characteristic shape: its proximal part is cylindrical; it then suddenly widens out and ends in a conical distal part (Fig. 35). The seminal duct is fused to one side of the inner papilla wall, and it is embraced on all other sides by a vast intrapapillar cavity. A heavy crest-shaped plica runs on the inner surface of the penis. The spermathecal duct is 12 SHILEYKO FIGS. 32-36. Leucozonella retteri (Rosen), Kandara Valley, Hissar Range, W Tian-Shan Mountains, Tadzhik SSR, 2 July 1967. 32, reproductive tract: 33, mantle collar; 34, transverse section of papilla showing single intrapapillar cavity with crescent-shaped cross-section; 35, penis, penis sheath partly removed; 36, cross-section of epiphallus. slightly curved; an oval spermatheca reaches the lower edge of the albumen gland. Leucozonella caria Shileyko, sp. nov. Figs. 37-41; Pl. И, 8 Three specimens from the Hissar Range, W Tian-Shan Mountains, Tadzhik SSR, at the outskirts of the Khodzcha-Obi-Garm Rest Home, were examined. They were collected on 28 May 1968 by Z. Izzatulaev and identified by me. The holotype is at the Zoological Institute, USSR Academy of Sciences, Leningrad. The shell is small (8-9 тт) and very much like that of the European Trichia $. str., but it has a constant distinction: it has no more than 4.5-5 whorls, whereas in fully adult Trichia $. str. there usually are 6 whorls. The shell may vary in aspect from lowly conical to nearly conical; it is brownish-horny with a washed-out light line at the periphery. It is sculptured with fine and rather light radial lines. The body whorl is 1.5 times wider than the penulti- mate whorl. The shell is covered with long periostracal hairs, curved at the ends as in Trichia plebeia. When the hairs are lost, marks are left in their place that look like short radial wrinkles. Spiral sculpture 1$ absent. The nuclear whorls are glossy and not clearly limited from the adult whorls, and they have the same color. The whorls increase rather slowly in size, though more rapidly than European Trichia. The aper- ture is rounded inside; slightly away from the edge there is a low but wide, light- colored lip, which occupies the whole edge of the aperture in adults and is not limited to the basal part only. The umbilicus is narrow but perspective; though the colu- mellar edge of the aperture is slightly re- flected, it does not cover the umbilicus. The aperture is not deflected, moderately oblique. Measurements are as follows: Holotype Paratype (mm) (mm) Shell height 4.7 6.0 4.2 Shell width 7.8 Bo TT TRICHIA SYSTEMATICS ie р res FIGS. 37-41. Leucozonella caria Shileyko, sp. nov., holotype, Khodzcha-Obi-Garm Rest Home, Hissar Range, W Tian-Shan Mountains, Tadzhik SSR, 28 May 1968. 37, reproductive tract; 38, inner structure of vagina in dart sac region; 39, penis, penis sheath partly removed; 40, cross-section of verge; 41, cross-section of epiphallus. The shell of L. caria can be distinguished from that of L. retteri not only by its small size but also by its color and texture: L. retteri is light brown to reddish and has no hairs. In the structure of the verge, L. caria is characteristic for the Asiatic group and sharply’ differentiated from the European Trichia (cf. Fig. 160, IV, V). The oviduct is short and does not form a bend; there are 3 mucous glands, each with 2 or 3 branches. These are long, 1.5-2 times longer than the upper vagina. The dart sacs are relatively very massive, elon- gate; the outer sacs are closely pressed to the inner ones. The lower vagina is straight or curved. The flagellum is fine, not as short as the epiphallus, which forms a curve. The penis is elongate; between epiphallus and penis there is a ring-like swelling (bulla). The verge is small and oval and does not measure more than half the penis length. The intrapapillar cavity is as in L. retteri: it embraces the seminal duct, which adheres to the inner papilla wall on one side. The interior of the penis sheath is covered by numerous small longitudinal folds. There are only 2 pairs of vaginal plicae. At the base of the dart sac ducts they form well-developed lobes. The sper- mathecal duct is almost straight, ending in a small oval spermatheca that falls con- siderably short of the lower edge of the albumen gland. Genus Hygrohelicopsis Shileyko, gen. nov. The shell is very flattened, lilac-chest- nut-colored, with pale transverse spots. The aperture is rather large, with the body whorl rapidly increasing in height. The um- bilicus is narrow but deep, penetrating well into the shell. Unlike all other Trichiinae, in this genus, the right-hand ocular retrac- tor does not cross the distal part of the genitalia, passing between the penis and vagina, but only runs beside them. A further distinguishing mark is the seeming absence of the inner pair of dart sacs, which, however, are present internally. The flagellum is about as long as the penis and the epiphallus. Inside the verge the seminal duct is surrounded by a pair of intrapapil- lar cavities embracing it from 2 sides. 14 SHILEYKO Hygrohelicopsis darevskii Shileyko, sp. nov. Figs. 42-46; Pl. Ill, 9 Two specimens were examined. They were collected in Chegem Valley, N slope of the central Caucasus, USSR, at 200-2500 m above sea level, on 9-10 August 1965, by I. $. Darevski and identified by me. Holotype and paratype are in the Zoologi- cal Institute, USSR Academy of Sciences, Leningrad. The shell is very lowly conical, of a lilac-chestnut color, with pale, wide trans- verse spots and a silky gloss. It is sculptured with irregular, smooth, trans- verse wrinkles, particularly on the upper part of the body whorl; the basal part is smoother. In places there are weak spiral lines. The aperture is rather large, oblique and somewhat deflected. The aperture edges are simple, but the basal edge is slightly reflected. At the rim there is a heavy, snow-white Ир which is seen through the body whorl wall as a wide, white line with indistinct limits. During development not 1 lip alone but 2-3 lips are formed; earlier lips are also seen as wide, light lines. There are 5 whorls of moderately rapid growth. The body whorl increases rapidly in height so that the parietal wall of the aperture makes a rather acute angle with the periphery of the body whorl. This character easily distinguishes this species from Caucasigena schaposchni- kovi, in which the lower part of the palatal aperture wall is almost parallel to the pe- riphery of the body whorl (see PI. Ill, Эа and Pl. V, 19a and 20a). The umbilicus is narrow but deep. The nuclear whorls are smooth, lightly horn-colored; they are vaguely separated from the definitive whorls. The dimensions are Holotype | Рагатуре (mm) (mm) Shell height 6.0 6.0 Shell width 10.5 10.4 FIGS. 42-46. Hygrohelicopsis darevskii Shileyko, sp. nov., holotype, Chegem Valley, N slope of central Caucasus, USSR, 10 August 1965. 42, reproductive tract; 43, longitudinal section of dart sac region; 44, penis, penis sheath partly removed; 45, transverse section of verge showing paired intrapapillar cavities of crescent-shaped cross-section on either side of seminal duct, which is fused with the papillar wall at 2 opposite points; 46, cross-section of epiphallus. TRICHIA SYSTEMATICS 15 Anatomically this species stands apart from other representatives of the Trichia $. lat. group. The oviduct makes a sudden bend. There are 4 mucous glands; 1 or 2 are 2-branched. The upper and lower vagina in the dart sac region is bulbous; the dart sacs are very large. Externally only 1 pair of dart sacs is seen (Fig. 42) but upon dissec- tion one can see small rudimentary inner sacs fully covered by a common connective tissue sheath (Fig. 43). The flagellum 15 thin, about twice as long as the epiphallus. The epiphallus sharply curves twice and is held in this position by connective tissue bands. The penis is globose, inflated. The verge is cylindrical proximally and bulbous distally. The seminal duct is surrounded by a pair of intrapapillar cavities (Fig. 45). The penis sheath is smooth inside. The spermathecal duct is long, slightly curved, with a bulky rounded spermatheca almost reaching the albumen gland. It is emphasized that the right-hand ocular retractor (ROT, Fig. 42) is situated as is characteristic for the “’Нейсе!тае” auct.: i.e., it passes near the distal genitalia but not between them. We shall return to this point in the discussion. Genus Teberdinia Shileyko, gen. nov. The shell is brownish-yellow in color with a slightly inflated basal part; it is lowly conical with an umbilicus of medium size. The outer and inner dart sacs are closely apposed and fused along the greater part of their length, but the inner sacs are distinctly separate from the upper vagina. The seminal duct is surrounded by a pair of intrapapillar cavities. It is the only Caucasian form with a deep longitudinal groove on the surface of the penis papilla that separates off a lobe, and the only form so far observed in which one of the intrapapillar cavities reaches into the lobe. Genus monotypic. Teberdinia zolotarevi (Lindholm, 1913) Figs. 47-51; Pl. 111, 10 One specimen was examined anatomi- cally. It was collected from the Teberdia RIA GX FETE VF FIGS. 47-51. Teberdinia zolotarevi (Lindholm), Teberdia Nature Reserve, NW Caucasus, USSR, 24 July 1958. 47, reproductive tract; 48, inner structure of vagina in dart sac region; note lobes on vaginal folds; 49, penis, penis sheath partly removed; 50, cross-section of verge; 51, cross-section of epiphallus. 16 SHILEYKO Nature Reserve, NW Caucasus, USSR, by L. Arens on 24 July 1958; | identified it. Holotype and paratype (not fully adult) are at the Zoological Institute, USSR Academy of Sciences. The shell is lowly conical with a markedly bulbous basal part and flattened whorls. The color is brownish-yellow. It is sculptured with smooth radial wrinkles and spiral lines. The aperture is oblique, and in the basal (palatal) part it is twisted slightly to the right. The aperture edges are simple, slightly reflected. The columellar edge is more reflected. Near the edge there is a heavy white lip, visible through the shell as a wide, light line. The shell is perforate, and the aperture is slightly covered by the columellar edge. There are 5.75 whorls. The holotype shell height is 6.7 mm; width, 11.3 mm. The oviduct forms a smooth bend. There are 3 mucous glands, each with 2 branches. The dart sacs are small, closely pressed together in pairs and directed away from the upper vagina. The lower vagina is short, and cylindrical, with narrow but rather high plicae. At the base of the dart sac duct they form rather round lobes. The flagellum is longer than the straight cy- lindrical epiphallus. Between the flagellum and the penis there is a clear-cut ring (bulla). Internally the penis sheath bears thin long plicae. The verge is short, bag- like, blunt at the edge. A deep longitudinal groove occupies about half the length of the verge. Of the 2 intrapapillar cavities, 1 is adjacent to the seminal duct and the other characteristically extends into the longitudinal lobe separated by the above- mentioned groove (Fig. 50). The sperma- thecal duct is very thin and closely pressed to the spermoviduct; it ends in a bulky, bag-like receptaculum seminis that does not reach the albumen gland. Genus Kokotschashvilia Hudec & Lezhawa, 1969 The shell is lowly conical to top-shaped, white or horny, perforate to umbilicate and perspective. It is sculptured with radial lines, spiral lines or granules. The flagellum is about half as long as the epiphallus. There are 4 mucous glands, each with 2-4 branches. The dart sacs are massive and rounded. The receptaculum seminis is very bulky; when it is not full its walls are collapsed, so in this condition it looks atypical. The seminal duct is surrounded either by a pair of intrapapillar cavities or, when a longitudinal partition between the cavities is absent, by a single intrapapillar cavity with a crescent-shaped cross-section that embraces the seminal duct from 3 sides. The seminal duct may either adhere closely to the inner papilla wall on one side or on a thin, long band on that same side or also on the opposite side. Type-species: Helix holotricha Boettger, 1884 Kokotschashvilia makvalae (Hudec & Lezhawa, 1969) Figs. 52-56 The anatomy of 1 specimen was studied. It was collected at Balda village, Gegechkor region, NW Georgia, Grusinian SSR, on 3 May 1967 by G. Lezhawa and identified by V. Hudec. A description of the shell is given by Hudec & Lezhawa (1969a, b). The spermoviduct passes straight and without curving into the oviduct, which also shows no bend. Each of the 4 mucous glands has 2-3 branches. The inner dart sacs are slightly smaller than the outer ones. Inside the vagina there are, laterally, 2 pairs of long plicae. At the exit of the dart sac duct the vaginal plicae do not form any lobes. The flagellum is approximately half as long as the slightly curved epiphallus. A penial membrane, rather weakly developed, is present. The penis is fusiform, slightly bulbous. The penis sheath is smooth inside. The verge is fusiform. The seminal duct is held in place by 2 longitudinal bands. There is a fine channel in the papilla wall at the place of attachment of 1 of the bands. The spermathecal duct is moderately twisting, ending in a wide receptaculum seminis that almost reaches the albumen gland. Kokotschashvilia tanta Shileyko, sp. nov. Figs: 57-612P\. М Specimens were collected from alpine meadows near Lebarde village, Gegechkor region, NW Georgia, Grusinian SSR, on 5 July 1962 by M. G. Natsvlishvili and identi- fied by me. Three specimens were examined anatomically. The holotype and 6 paratypes (2 damaged) are at the Zoo- logical Institute, USSR Academy of Sciences, Leningrad. The shell is lowly conical to lowly top- shaped. Fresh shells are a light horn color. TRICHIA SYSTEMATICS 17 Sp MFE FIGS. 52-56. Kokotschashvilia makvalae (Hudec and Lezhawa), Balda village, Gegechkor region, NW Georgia, Grusinian SSR, 3 May 1967. 52, reproductive tract; 53, inner structure of vagina in dart sac region; 54, penis, penis sheath partly removed; 55, transverse section of verge; the 2 stalks joining the seminal duct to the papillar wall are the longitudinal bands, in cross-section, which attach the duct to the papilla; 56, cross-section of epiphallus. The color is darker above and under the periphery of the body whorl, forming vague darker lines. The shell is sculptured with irregular, rather fine radial lines; in places there are vague granulose and spiral lines, more clearly seen on the periphery and the basal part of the shell. A pale line runs above the suture. The aperture is roundly lunate, oblique and deflected. A narrow light spiral line starting a distance back of the aperture runs high on the shoulder of the whorl. The umbilicus is eccentric, and through it the penultimate whorl is fully visible. A pale line on the shell surface corresponds to the lip. The shell is similar to that of K. mak- valae, its nearest relative, but differs by its darker color (K. makvalae is uniformly white), by the presence of spiral sculpture, by a more oblique aperture, by a larger number of whorls (6.5-7.0 compared to 6) and by its definitely larger size. The largest of our specimens of K. makvalae has a diameter of 21.5mm compared to 22-27 mm т К. tanta. It differs from К. eberhardi in color and shape (К. eberhardi is yellowish and has a wider umbilicus and a higher spire) and by its much larger size. Shell measurements, including 5 of the paratypes, are as follows: Holotype Paratypes (mm) (mm) Shell height 16.8 16.7 18.0 14.6 16.3 17.8 Shell width 25:5 27.0 24.7 22.2 25.0 26.3 More distinct interspecific differences are in the reproductive anatomy, particu- larly in the inner structure of the verge. The oviduct makes a smooth bend; it is thin but becomes wider below. There are 4 mucous glands, each with 2-3 branches. The dart sacs are rather massive, globose. The vaginal plicae are heavy and form wide lobes. The flagellum is about half the length of the epiphallus. The latter is 18 SHILEYKO 40M FIGS. 57-61. Kokotschashvilia tanta Shileyko, sp. nov., paratype, Lebarde village, Gegechkor region, NW Georgia, Grusinian SSR, 5 July 1962. 57, reproductive tract; 58, inner structure of vagina in dart sac region; 59, penis, penis sheath partly removed; 60, cross-section of verge; 61, cross-section of epiphallus. attached to the penis by connective tissue bands. The penis consists of 2 parts: a globose fusiform proximal part and a cy- lindrical distal one; these are also con- nected by connective tissue bands (Fig. 57). The verge lies in the former. As in K. makvalae, the seminal duct is surrounded by a pair of intrapapillar cavities, but in this species one of the partitions separating the pair is practically absent; thus the seminal duct is closely united with the inner papilla wall on that side (cf. Figs. 55 and 60). As in K. makvalae, there is a narrow channel in the papillar wall along the seminal duct, and near it, moreover, a very fine capillary. The spermathecal duct is thin and slightly curving; it gradually merges with the bag-like receptaculum seminis (spermatheca), which was in a flac- cid state in all specimens studied. The spermatheca was apparently empty because of the time of collection (June). It is noteworthy that in К. holotricha and К. eberhardi, which were collected during June-July also, the spermathecas were empty, whereas in K. makvalae and K. phaeolaema, collected in early and mid- May, it appeared to be full (cf. Figs. 52, 57, 62, 67 & 71). Kokotschashvilia holotricha (Boettger, 1884) Figs. 62-66 Two specimens were examined. They were collected at Tsebelda village, near Sukhumi, in the Black Sea coastal region, Grusinian SSR by G. Lezhawa on 15 June 1968 and identified by V. Hudec. A good shell description is in Likharev €: Rammel- meyer (1952). The spermoviduct curves as it passes to the free oviduct, which itself is straight. There are 4 mucous glands, all with 2 branches. The dart sac region is set off from the lower part of the vagina, which is narrower. The vagina is short. Vaginal plicae occur in pairs and occupy a lateral position; they are abruptly cut off distally and connected by a transverse fold (Fig. 63); the lower part of the vagina and genital atrium is either smooth internally or has irregular fine wrinkles. The flagellum is TRICHIA SYSTEMATICS 19 62 DS MFE FIGS. 62-66. Kokotschashvilia holotricha (Boettger), Tsebelda village, near Sukhumi, Black Sea region of Georgia, Grusinian SSR, 15 June 1968. 62, reproductive tract; 63, inner structure of vagina in dart sac region; 64, penis, penis sheath partly removed; 65, cross-section of verge; 66, cross-section of epiphallus. short and conically tapering. It measures about half the length of the epiphallus, which is S-shaped and supplied with parti- cularly well-developed connective tissue bands containing muscle fascicles. The penis is fusiform and has longitudinal folds in the vicinity of the genital atrium. The verge is fusiform; the seminal duct is con- nected to the inner surface of the papilla wall by 1 single long band and surrounded on all sides by an intrapapillar cavity (Fig. 5); the papilla wall does not include any channels. Hudec and Lezhawa (1969a, b) did not comment on the S-shaped bend of the epiphallus in this species. Hesse (1931) pointed out that not all the mucous glands have 2 branches; some may be simple. Kokotschashvilia eberhardi, Shileyko, sp. nov. Figs. 67-70; PI. 111, 12 Specimens were collected in July 1969 between the villages of Sioni and Kazbegi, on the S slope of the central part of the main Caucasus range, Grusinian SSR, at an altitude of 2500-3000 m above sea level, by E. Clauss. They were identified by me. Four specimens were examined anatomi- cally. The holotype and 4 of the paratypes are at the Zoological Institute, USSR Academy of Sciences, Leningrad. The shell is compressedly conical to almost globose. Its coloration is character- istic: the background is a pale straw color that gradually becomes lighter near the sutures: above and under the periphery it may be darker, forming 2 dark bands. It has a sculpture of irregular radial wrinkles, more clearly expressed near the sutures. With a lens it can be seen that in fresh shells the wrinkles are lighter than the background, in fact almost white. There are granulose spiral lines, mostly on the basal part of the shell. The aperture is almost round; at its edge there is a well-developed white lip. This lip is cleariy visible from the outside as a whitish-yellow band; 2-3 such lips are formed during the life of the 20 SHILEYKO BO me St ‘ wer EE ER Bir Si sr FIGS. 67-70. Kokotschashvilia eberhardi Shileyko, sp. nov., paratype, between Sioni and Kazbegi villages, S slope of central Caucasus, Grusinian SSR, July 1969. 67, reproductive tract; 68, inner structure of vagina in dart sac region; 69, penis, penis sheath partly removed; 70, cross-section of verge. mollusk, and these earlier ones are also visible through the shell as light lines. The umbilicus, though narrow, is perspective. There are 6 whorls. The dimensions of our shells are: Paratypes (mm) TOW SO ee Ol Oto 10: 15.5 15.0 14.1 15.0 14.6 14. Holotype (mm) Shell height 10.3 Shell width 14.5 7 8 The shell of this species resembles that of К. makvalae and К. tanta in shape, though it differs in its smaller size, in color (K. makvalae is white and K. tanta horn- colored) and also in sculpture. There are clearer distinctions yet in the structure of the reproductive organs. The spermoviduct makes a sudden but smooth bend as it passes to the free ovi- duct, which itself is slightly bent. There are 4 mucous glands, each with 3-4 branches. The dart sacs are very massive and globose; the sac region is set off from the vagina by a slight narrowing of the lower vagina. In addition to the usual paired vaginal plicae in a lateral position, there are supple- mentary smaller vaginal plicae. Usually the basic plicae do not form lobes. The flagel- TRICHIA SYSTEMATICS 21 lum is thin, and its length is about half that of the epiphallus, which is also thin, cylindrical and weakly curving. The penis is massive, more or less curved. There are no muscle bands, only a fine connective tissue membrane stretched among penis, epiphal- lus and vas deferens. The inner surface of the penis sheath is smooth. The verge is bulky and bag-like and has a large papillar lacuna. The seminal duct is closely fused to one side of the inner papillar wall and surrounded on all other sides by a wide intrapapillar cavity as in K. tanta; however, the inner papillar wall facing this cavity is not smooth, as in K. tanta, but bears many lamellae and folds (Fig. 70). The sperma- thecal duct is almost straight and demar- cated from the receptaculum seminis, which has very thin walls. In the specimens studied the receptaculum walls were wrinkled and collapsed, the receptaculum apparently empty. The organ does not reach the albumen gland. The intense pig- mentation of the pallial nerve is note- worthy. Kokotschashvilia phaeolaema (Boettger, 1886) Figs. 71-76; Pl. IV, 13 Sixteen specimens were examined anatomically; 14 of these were collected on 14 May 1970 in Chegem Valley, which descends the N slope of the central Cauca- sus (USSR) as a tributary to the valley of the Terek; they were identified by me. Two specimens were collected from the Khunzah district of Daghestan on 26 August 1955 by T. Khasanov and identified by |. M. Likharev. As this species is little known, the shell is described here. It is lowly conical and almost globose, light horny, yellowish or chestnut in color, with a diffuse light line at the periphery, radial folds and distinct spiral ridges. The aperture is roundly lu- nate, slightly deflected and oblique. The umbilicus is narrow and may be half- covered by the reflection of the columellar lip. There is a thick white lip on the inside edge of the aperture. During life the animal forms 3-5 such lips, which can be seen translucently as radial bands of a lighter hue. On the outside, closely joined to them, there is a line darker than the back- ground of the shell. There are 6 whorls. The shell height (from 53 specimens) is 9-12 mm; the width, 10.5-16.0 mm. Lindholm (1913) described Helix (Fruticocampylaea) phaeolaema Boettger var. tenuitesta, which differs by its small, thin-walled, translucent shell (height, 9-10mm; width, 10.5-12mm), with a translucent lip and a yellowish or light brown horny color. As these characteristics are all seen in some specimens of our series, it is not necessary to separate this form. The oviduct makes a sudden sharp bend. There are 4 mucous glands, each usually with 2 branches, but sometimes we find a secondary branching in some particular gland. The dart sacs are well-developed and globose. The vaginal plicae are only moderately developed, but form clear-cut lobes near the opening of the dart sac duct. The length of the flagellum is 1/2 to 3/4 that of the epiphallus. The latter bends twice and is attached to the penis by a well-developed connective tissue membrane (CPM, Fig. 71) in which there are muscle bands. The penis is large, bag-like or glo- bose. Its sheath is smooth inside, except that there are a few longitudinal folds near the genital atrium. The papilla is very in- flated, globose or pear-shaped. The seminal duct is suspended from the papillar wall by 2 long bands; i.e., it is surrounded by a pair of intrapapillar cavities. In contrast to other species of this genus, the papillar walls are thick and contain additional intra- papillar cavities in the form of small scat- tered sinuses in the wall (Fig. 4), which therefore sometimes looks spongy. The receptaculum seminis is very massive and bag-like; its upper edge reaches the lower part of the albumen gland. The characteristic peculiarity of this species is the intense pigmentation of the spermoviduct (especially on the oviduct side), of the distal part of the intestine, of the mesentery and of the surface of the circumesophageal nerve ring. The gross morphology of the genitalia of this species was first described by Kalitina (1958: 159-160). Genus Caucasigena Lindholm, 1927 The shell is almost flat to lowly conical, smooth or ribbed, umbilicate or perforate (subgen. Dioscuria). Inside the aperture edge there is a heavy white lip. There are 3 or 4 mucous glands, each with 2 branches, and a tendency neither for secondary branchings nor for simple, 22 SHILEYKO FIGS. 71-76. Kokotschashvilia phaeolaema (Boettger), Chegem Valley, N slope of central Caucasus, USSR, 14 May 1970. 71, reproductive tract; 72, penis in copulating position; verge (papilla) is moved out, penis sheath is turned inside out; 73, penis, penis sheath partly removed; 74, cross-section of verge; 75, inner structure of vagina in dart sac region; 76, cross-section of epiphallus. nonbranching glands. Inside the papilla (verge) the seminal duct is very narrow. The longitudinal compartments of the intrapapillar cavity that lies on 1 side of the papilla are separated by longitudinal septa from one another and from the cavities (numbering from 3 to 9) that sur- round the seminal duct. The papillar lacunae are rather large. Type-species: Helix eichwaldi L. Pfeiffer, 1846. Subgenus Caucasigena s. str. The shell is pale, with brown spiral bands that sometimes are so markedly developed they form nearly the whole background of the shell; in this case there TRICHIA SYSTEMATICS 23 is a lighter line on the periphery. The radial sculpture varies from simple lines (ridges) to very rough ribs. The spiral sculpture appears either as fine striae or as very thin ribs. The flagellum is stout, conical, not more than 1/2 the length of the epiphallus, which is more or less straight. The inner dart sacs are the same size as the outer ones or slightly longer. The seminal duct is attached to the inner wall of the papilla by 1-3 long bands. Caucasigena (Caucasigena) armeniaca (L. Pfeiffer, 1846) Figs. 77-82; Pl. IV, 14 The specimens described are from Mt. Bzovdal (Mt. Todar) in the Stepanavan region, Armenian SSR. They were collected on 18 July 1951 and identified by N. N. Akramovsky. Two of the 6 specimens avail- able were dissected. Because of some confusion in the litera- ture, it is necessary to give a description of the shells. Records of this species in a number of districts of the main Caucasus | + By \ CPB LS QD range are erroneous: all those from the N slopes of the Great Caucasus are referable to С. rengarteni, which Likharev 8 Rammelmeyer (1952) mistakenly placed in synonymy with C. armeniaca and omitted from their index. The shell is lowly conical with a blunt angle at the periphery; it is relatively thick-walled, of a brownish horny color, with a light line at the periphery. The sculpture is very characteristic and consists of minute, smooth, radial ribs visible only under magnification and spiral periostracal ribs. At first sight they seem like the usual spiral lines, but when magnified 30-40 X and viewed with an oblique beam of light, they are clearly seen to be fine, sparsely but regularly spaced ribs; next to these are fine, rare, weakly curved hairs. The umbili- cus is at first wide, funnel-shaped, but it then sharply narrows though it remains perspective. The aperture is roughly lunate, very oblique; on the inner rim there is a somewhat thickened yellowish lip. The shell has 5 gradually increasing whorls. The dimensions of the 6 specimens are: 81 MFE FIGS. 77-82. Caucasigena (Caucasigena) armeniaca (L. Pfeiffer), Mt. Bzovdal (Mt. Todar), Stepanavan region, Armenian SSR, 18 July 1951. 77, reproductive tract; 78, inner structure of vagina in dart sac region; 79, cross-section of dart sacs; 80, penis, penis sheath partly removed; 81, cross-section of verge; 82, cross-section of epiphallus. 24 SHILEYKO Shell height (mm) Shell width (mm) 4.0 4.0 4.3 4.5 3.7 5.0 7:9:58:01:8:05 815, 6:74:87 The oviduct is short and straight. There are 4 mucous glands with 2 branches. The dart sacs are characteristic: the upper edges of the inner sacs pass beyond and curve over the elongate outer sacs, and their central portion curves medially (Fig. 77), sometimes almost bent over double. The inner pair of dart sacs is compressed from both sides (Fig. 79). The flagellum and epiphallus are both short; the former is thick and conical and about half as long as the latter, which is cylindrical and weakly curved. The penis is slightly bulbous, fusi- form. On the surface of the penis sheath there are slight connective tissue bands (CPB, Fig. 77). Between vas deferens, epi- phallus and penis is stretched a fine trans- parent connective tissue membrane (CPM). The interior of the penis sheath is smooth. The verge is club-shaped or fusiform, with a well-developed papillar lacuna. The semi- nal duct is attached to the internal papillar wall by 2 long bands; further longitudinal septa divide off another 2-3 cavities (Fig. 81), making a total of 4-5 cavities. The spermathecal duct is almost straight and fine; the spermatheca is small and rounded; it nearly reaches the albumen gland. Caucasigena (Caucasigena) tschetschenica (Retowski, 1914) Figs. 83-87; Pl. IV, 15 Six specimens were dissected. | collected them on a limestone cliff in Kurtatin Valley at about the middle course of Phiagdon River, central part of the N Caucasus, USSR, on 19 September 1970, and | identified them. The species was described by Retowski (1914) on the basis of 2 specimens col- lected by Kohnig on Mt. Bonoz-Mta in the Checheno-Ingush (‘’Chechnya’’). Since the FIGS. 83-87. Caucasigena (Caucasigena) tschetschenica (Retowski), Kurtatin Valley, middle course of Phiagdon River, central part of N Caucasus, USSR, 19 September 1970. 83, reproductive tract; 84, inner structure of vagina in dart sac region; 85,/penis, penis sheath partly removed; 86, cross-section of verge; 87, cross-section of epiphallus. TRICHIA SYSTEMATICS 25 species is little known, | here repeat its description, supplemented by my own observations. The shell is small with a rather wide umbilicus, about 1/5 the shell width. The shell shape is almost flat to compressed conical. The color, as a rule, is horny with a pale line on the periphery accompanied by dark horn-colored bands on both sides. There are 5 relatively prominent, slowly increasing whorls, irregularly and roughly ribbed, with fine spiral lines in the spaces among the ribs. The ribs are white. The last whorl is angled or keeled. The aperture is strongly oblique, rounded, with a heavy white lip inside it. The parietal edge of the aperture is short. The columellar edge is widened and slightly reflected. The range of shell measurements from 100 shells is as follows: shell height, 3.2-4.8 mm; width, 6.5-9.8 mm. Retowski gives the following sizes for his shells: Shell height (mm) 4-4.5 Shell width (mm) 7.7-8 The oviduct forms a rather smooth bend. There are 4 mucous glands, each with 2 branches. The dart sacs are elongate, and the rather large inner pair points dis- tinctly away from the upper vagina; the outer pair are almost the same size. The flagellum is conical, stout, about half as long as the epiphallus. The penis is fusi- form, globose. The male ducts are linked by -a fine translucent connective tissue membrane (CPM, Fig. 83). The verge is relatively very massive and bag-like. The seminal duct is attached to the internal papillar wall by only 1 longitudinal band; ¡.e., it is surrounded by 1 encircling cavity on 1 side of the papilla. In the remaining part of the verge, longitudinal septa stand- ing transversely divide the intrapapillar cavity into 3-5 parts (Fig. 86). The inside of the penis sheath is smooth, with vague Partitions in its distal part. The spermathe- cal duct shows no curves; it ends in an elongate-oval receptaculum seminis, which does not reach the albumen gland. Caucasigena (Caucasigena) rengarteni (Lindholm, 1913) Figs. 88-91; PI. IV, 16; PI. М, 17 Twenty specimens were dissected. These were collected in the central part of the N Caucasus (USSR) as follows: 10 specimens from the middle course of the Chegem River near Khushtosyrt village, on 14 May 1970; 6 from the upper Chegem near Chegem village, on 19 May 1970; 4 from the old settlement of Ara-Boran between the Chegem and Baksan valleys, on 17 May 1970. These were all identified by me. The shell characters are well given by Lindholm (1913), but he described the species from 3 specimens, whereas we had about 150 shells. It might now be added that some specimens have rare, wiry, short hairs on the surface of the shell. The dimensions may also be defined more exactly (Lindholm’s values were 7 and 14.5 mm, respectively). Shell height (mm) 5.8- 8.2 Shell width (mm) 7.3-15.5 In the same article Lindholm described Helix (Fruticocampylaea) gerassimovi and outlined the differences between that species and Helix rengarteni, but from his description it is not quite clear what characters other than size can distinguish these 2 species. | was able to familiarize myself with the type-series of both of Lindholm’s species (i.e., gerassimovi and rengarteni) and found that the differences between them depended on intrapopula- tional variability. Most of the specimens from Ara-Boran correspond to the diagnosis of H. gerassimovi, whereas in most of the series of C. rengarteni from the Chegem Valley one could see continuous gradations among the shells that invalidated the diagnosis of 2 separate species. Thus Н. gerassimovi is a synonym of Caucasigena rengarteni. On Plates IV and V, shell 16 is typical for “Helix” rengarteni and shell 17 for H. gerassimovi. The oviduct is almost straight. There are 4 mucous glands, each with 2 branches. The flagellum is thick, pointed and 1/2 to 3/4 the length of the epiphallus. The latter is straight or curved; it is connected with the penis by a very fine translucent connec- tive tissue membrane (CPM, Fig. 88). On the surface of the penis sheath there are connective tissue bands including muscle fibers. The penis consists of a fusiform, globose proximal part, in which the papilla is located, and a slightly curved cylindrical distal part. The proximal part is smooth inside, with only small wrinkles or folds; the distal part bears many long wrinkles 26 SHILEYKO “ДУ ) FIGS. 88-91. Caucasigena (Caucasigena) rengarteni (Lindholm), Khushtosyrt village, Chegem Valley, central part of N Caucasus, USSR, 14 May 1970. 88, reproductive tract; 89, distal part of genitalia, dissected; 90, cross-section of epiphallus; 91, cross-section of verge. FIGS. 92-95. Caucasigena (Caucasigena) eichwaldi (L. Pfeiffer), Chmi Village, Darial valley, near Ordjonikidze (North Osetia), central part of N Caucasus, USSR, 8 May 1970. 92, reproductive tract; 93, distal part of genitalia, dissected; 94, cross-section of verge; 95, cross-section of epiphallus. TRICHIA SYSTEMATICS 27 that merge with the folds in the genital atrium (Fig. 89). There are raised, rather heavy, vaginal plicae that form clear lobes at the mouth of the dart sac ducts. The verge is bulky and bag-like with a foramen in the shape of a wide slit and with large papillar lacunae. The seminal duct is held in place by 3 or 2 longitudinal bands; the intrapapillar cavity in the other half of the papilla is divided into 4-6 compartments by longitudinal septa. The spermathecal duct is straight or very weakly curved; the oval receptaculum seminis lies some distance away from the albumen gland (Fig. 88). For diagnostic characters, see below under C. eichwaldi. Caucasigena (Caucasigena) eichwaldi (L. Pfeiffer, 1846) Figs. 92-95; Pl. V, 18 Nineteen specimens were dissected. | collected them in the Darial Valley, central part of the М Caucasus (USSR), near Ordjonikidze (North Osetia) as follows: 10 specimens near Kazbegi village, оп 9 May 1970; 9 near Chmi village on 8 May 1970. The identification is mine. A good shell description is given by Likharev & Rammelmeyer (1952). | never- theless wish to point out that, although the shell of this species is rather variable as regards color, dimensions and particularly height of whorl, a keel is never found, nor any inclination toward forming an angle at the periphery, a character that distinguishes С. eichwaldi from С. rengarteni. As the spermoviduct passes to the ovi- duct there is a rather sharp curve. There are 4 mucous glands, all with 2 branches. The dart sacs are massive, slightly elongate or globose. The inner vaginal structure is the same as in С. rengarteni. The flagellum length is at most half that of the epi- phallus. The epiphallus does not make any sudden curve. The connective tissue mem- brane and penial band are as in C. rengar- teni. The penis sheath also consists of 2 parts. The verge is massive. In it, the seminal duct is fixed in place by 3 longi- tudinal bands; i.e., it is surrounded by 3 intrapapillar cavities. The cavity in the re- mainder of the papilla is divided into 6-9 parts by parallel, transversely arranged, longitudinal septa (Fig. 94). The spermathe- cal duct forms a marked curve at its base. The receptaculum seminis reaches the lower edge of the albumen gland. Differential diagnosis of C. rengarteni and C. eichwaldi The anatomical distinctions between these 2 similar species can now be formu- lated. In C. rengarteni the mucous glands lie approximately at the level of the upper edges of the dart sacs; the uterus is straight; the spermathecal duct is almost straight; the flagellum length is not less, but usually a little more, than half the epiphallus length. The intrapapillar cavity is divided into 4-6 portions. In the type- species, С. eichwaldi, the mucous glands lie considerably above the level of the upper edges of the sacs. The uterus is bent and so is the spermathecal duct. The flagellum length is usually less than half that of the epiphallus. The intrapapillar cavity is divided into 6-9 parts. Subgenus Anoplitella Lindholm, 1929 The shell is a light color with lilac spots, lines or marks; usually there are 2 or more less developed brown lines above and under the periphery of the body whorl. If these lines are well developed, the pale purple spots are absent. The shell is thick-walled, with a silky gloss. The umbilicus is more or less perspective. The flagellum is long and thin, a little longer than the curved epiphal- lus. The spermathecal duct is also long, thin and curved. The seminal duct is fused to the papilla on one side. Subgenus monotypic. Caucasigena (Anoplitella) schaposchnikovi (Rosen, 1911) Figs. 96-100; PI. М, 19, 20 Seventeen specimens were dissected. | col- lected these in the central part of the N Caucasus (USSR) as follows: 10 specimens from the old Ara-Boran settlement between the Chegem and Baksan valleys, on 17 May 1971, and 7 from Chegem Valley near Khushtosyrt village, on 15 May 1970. | identified them. The shell is flattened, sometimes almost completely flat, smoothly rounded along the periphery and glossy. The umbilicus is rather narrow but perspective: all the previous whorls are visible through it. A little back of the aperture edge there is a thick white lip. The adult has 2-4 earlier lips that are visible through the whorls concerned as light radial lines. The shell is 28 SHILEYKO FIGS. 96-100. Caucasigena (Anoplitella) schaposchnikovi (Rosen), Khushtosyrt village, Chegem Valley, central part of N Caucasus, USSR, 15 May 1970. 96, reproductive tract; 97, mantle collar; 98, penis, penis sheath partly removed; 99, cross-section of verge; 100, cross-section of epiphallus. sculptured with fine, irregular, radial striae and in places rare spiral lines. There are 6 whorls. The dimensions, shells, were as follows: measured from 65 Shell height (mm) 5.0- 8.5 Shell width (mm) 11.0-15.5 The original description was made from smaller specimens; the shell height was up to 6.5mm and the width up to 10.5 mm. In my collection the species is rep- resented by 2 ecological forms that differ in the color and pattern of the shell and that do not occur together. These forms cannot be distinguished by other concho- logical characters, nor are there any anatomical distinctions. In shady valleys, in which insolation is slight, one finds the form with 2 distinct spiral brown bands of about equal width, 1 of which runs above and the other below the periphery. The distance between them is approximately equal to the width of each band. The upper band is visible along all whorls (PI. V, 20). In places with intense insolation (e.g., the old settlement of Ara-Boran) the shell color is chestnut to almost lilac; on this background there are more or less numerous bluish-white irregular lines and spots. In some places these spots coalesce into a larger patch or they form a reticu- late pattern (Pl. V, 19). At the same time, in most specimens belonging to this second form, there are more or less developed spiral bands that are also characteristic for it. As is evident from study of his material at the Zoological Institute, USSR Academy of Sciences, Lindholm (1929) was dealing with such specimens when he described var. balkariensis. The oviduct is almost straight but may be slightly curved. There are 3 or 4 mucous glands, each with 2 branches. The oviduct and upper vagina gradually increase in bulk from the upper to the lower part, being rather inflated in the region of the dart sacs. The outer sacs are large; the inner sacs are considerably smaller and fused with the upper vagina and the outer sacs. The flagel- lum is thin, slightly longer than the TRICHIA SYSTEMATICS 29 epiphallus or equal to it, and coiling. The epiphallus is more or less coiled. There are well-developed connective tissue bands stretched between parts of the epiphallus and between the distal epiphallus and the distal penis. The penis is massive, elongate and smooth on the inside. The verge is sac-like with an opening forming a wide slot and with well-developed lacunae. The seminal duct adheres to the papillar wall on 1 side; the papillar cavity is subdivided into 4-6 parts (Fig. 99). The spermathecal duct is long, thin and coiled. The receptaculum seminis is small and oval; it reaches the albumen gland. Subgenus Dioscuria Lindholm, 1927 The shell is thin-walled, brittle, with a silky gloss and a large aperture; depressed, conical. The umbilicus is very narrow and almost completely covered by the reflec- tion of the columellar edge. The length of the flagellum is about half that of the epiphallus. The seminal duct is not separated from the papillar wall. Subgenus monotypic. Caucasigena (Dioscuria) thalestris (Lindholm, 1927) Figs. 101-105; PI. VI, 22 One not fully mature specimen was dis- sected. It was collected from Novy Aphon (New Athos), N of Sukhumi, Black Sea region of Georgia, Grusinian SSR, on 2 July 1913, by Nasonov, and identified by Likharev. A shell description is given by Likharev & Rammelmeyer (1952). The oviduct forms a slight bend. There are 3 mucous glands, each with 2-3 branches. These are situated slightly higher than the dart sacs. The inner dart sacs are a little larger than the outer ones. The vagi- nal plicae do not form lobes but are rela- tively well developed. The flagellum is about half the length of the epiphallus, which is wavy and curving. The penis forms 1 U-bend. A heavy muscle band between penis and epiphallus consists of several anastomosing strands. The verge is long and cylindrical. The seminal duct runs in the thickness of the papillar wall. The intra- papillar cavity lying in the wall on 1 side of the verge is divided into 3 parallel FIGS. 101-105. Caucasigena (Dioscuria) thalestris (Lindholm), Novy Aphon, N of Sukhumi, Black Sea region of Georgia, Grusinian SSR, 2 July 1913. Specimen not fully mature. 101, reproductive tract; 102, inner structure of vagina in dart sac region, 103, penis, penis sheath partly removed; 104, transverse section of verge; broken circle indicates boundary between external and internal layers of papillar tissue; 105, cross-section of epiphallus. 30 SHILEYKO transverse compartments by longitudinal septa (Fig. 104). The spermathecal duct is tortuous. The elongated receptaculum seminis does not quite reach the albumen gland. My data differ somewhat from those of Hesse (1931). According to his illustrations, the flagellum is slightly longer than the epiphallus; there is no indication of any penial muscle bands, and the spermathecal duct is short. Unfortunately not all Hesse’s drawings are accurate, as he generally ignored muscle and _ connective tissue attachments. On the other hand, it should be kept in mind that | have examined only 1 not fully adult specimen, and such a character as flagellum length may be т- fluenced by age variation. The same applies to the length and shape of the spermathe- cal duct. Genus Plicuteria Shileyko, gen. nov. The shell is depressedly conical with a sharp apex, covered with radial lines, yellowish-white, rarely hirsute; the suture is deep. The lip of the aperture is not thickened but weakly developed. Vaginal plicae appear each as a row of prismatic lamellae of equal size. The conical flagel- lum is half the length of the epiphallus or less. The epiphallus is almost straight. The spermathecal duct is very short. The re- ceptaculum seminis is irregularly pear- shaped. The verge bears on its surface 2 longitudinal diametrically positioned grooves. In its distal part the seminal duct hangs from the inner walls of the papilla on 2 long bands, and in the proximal part it is fused to 1 side of the papillar wall. Genus monotypic. It is not clear why Polinski (1924) did not single out this snail as a separate species or subspecies. According to his illustrations and description, he took into account features of the outer genitalia such as the sturdy flagellum, spermathecal duct and receptaculum seminis, which are distinct enough to separate it from other European Trichia. The shell of this snail is evidently distinctive also. At any rate, the new genus is unique in that the longitudinal vaginal plicae are sub- divided by regular transverse prismatic folds that form a dense pattern on the internal wall of the vagina, and the spermathecal duct is shorter than in any other form of this group. Plicuteria lubomirski (Slossarski, 1881) Figs. 106-111; Pl. VI, 23 One specimen was examined. It was collected near the town of Olomouc, Moravia, Czechoslovakia, on 16 April 1964 and identified by V. A. Hudec. As it emerges from the spermoviduct, the oviduct forms a definite bend. There are 4 mucous glands, each with 2-3 branches. The outer dart sacs are somewhat more massive than the inner ones, whose upper tips reach well beyond those of the outer sacs. The lower vagina is rather long and cylindrical. Its inner structure is most characteristic and unlike that of any other species; the vaginal plicae are clearly divided into regular rows of prismatic lamellae (Fig. 107). The male duct is elon- gate, neither curving nor twisted. The length of the conical flagellum is less than that of the thick, cylindrical epiphallus, which is connected to the vas deferens by a very fine membrane. The penis is fusiform; the penis sheath is smooth inside. The verge bears 2 shallow longitudinal furrows on its surface (Figs. 108, 109), as well as a few incomplete ring grooves (Fig. 108). In the proximal part of the verge the seminal duct is adjacent to the papillar wall on one side. On the opposite side it borders on a large intrapapillar cavity. Deep folds here reach into the verge wall from the papillar lacunae (Fig. 110). Distally (Fig. 109), the seminal duct lies encircled by 2 cavities of sickle-shaped cross-section. Their position, in conjunction with the thickness of the wall, supports the idea that cavities are thus formed during ontogenesis, most likely in the thickness of the papillar walls. The spermathecal duct is very thick, straight and short. The receptaculum seminis is irregularly pear-shaped and quite distant from the lower edge of the albumen gland. Genus Trichia Hartmann, 1840 The shell is low to depressed conical, rather brittle, of a brownish horn color, with a more or less narrow umbilicus. As a rule there are hairs on the shell surface, which are absent in the adult specimens of some species. A _ thickened lip is either absent or occupies the basal edge of the aperture. There are 4 mucous glands, usually with 2 branches each. The dart sacs are elongate and more or less club-shaped. Inside the verge the seminal duct is en- TRICHIA SYSTEMATICS 31 FIGS. 106-111. Plicuteria lubomirskii (Slôssarski), Olomouc, Moravia, Czechoslovakia, 16 April 1964. 106, reproductive tract; 107, inner structure of vagina in dart sac region; *stylophore ореп!пд; note that longitudinal folds have become rows of prismatic lamellae; 108, penis, penis sheath partly removed; 109, 110, cross-sections of verge at different levels; 111, cross-section of epiphallus. circled by a pair of intrapapillar cavities. There may be 1 further narrow slit-like cavity in the papillar wall. The type-species is Helix hispida Linne. It was formerly thought to be 7. filicina L. Pfeiffer, but Forcart (1958) proved that Hartmann, when he determined the genus Trichia with T. filicina as the type-species, really meant 7. hispida. Subgenus Petasina Beck, 1847 The shell is dome-like, with fine short hairs and a very narrow umbilicus. On its basal edge the lip sometimes bears a heavy swelling. The 4 mucous glands are simple, not branched, and some of them are only partly divided into 2. The inner pair of dart sacs is quite separate from the outer pair as well as from the upper vagina: the upper tips of the inner sacs reach well beyond the place of attachment of the mucous glands. The receptaculum seminis is very bulky and of a characteristic hammer shape. Type-species: Helix unidentata Drapar- naud, 1805. Trichia (Petasina) unidentata (Draparnaud, 1805) Figs. 112-118; Pl. VI, 24 Five specimens were examined anatomi- cally: 2 from central Czechia, Czechoslo- vakia, collected on 8 September 1969 by J. Buchar and identified by myself, and 3 from the High Tatra Mountains, Slovakia, Czechoslovakia, collected on 20 July 1964 by V. Hudec and identified by him. As the spermoviduct passes into the oviduct, it curves weakly; the free oviduct itself is also slightly bent. There are 4 mostly unbranched, somewhat kinky mucous glands. The inner dart sacs are long, well separated, club-shaped; the outer ones are considerably shorter. One of the specimens had a distinct knobby thickening on the inner sacs (Fig. 114). The lower vagina is inflated and fusiform. The vagi- nal plicae are weakly developed, but there 32 SHILEYKO 112 118 0,5см Sp F FIGS. 112-118. Trichia (Petasina) unidentata (Draparnaud), central Czechia, Czechoslovakia, 8 Septem- ber 1969. 112, reproductive tract; 113, penis, penis sheath partly removed; 114, inner structure of vagina in dart sac region; 115, cross-section of verge; 116, mantle collar; 117, cross-section of epiphallus; 118, dart sacs of another specimen from same location. are distinct lobes at the openings of the dart sac ducts. The flagellum is about 2/3 the length of the cylindrical, almost straight epiphallus. The membrane connect- ing the distal parts of the male duct is very fine and translucent. The penis is cylindri- cal, straight or curved. The seminal duct is attached to the papillar wall by 3 longitudi- nal bands and surrounded by a pair of intrapapillar cavities with a crescent-shaped cross-section. The spermathecal duct is fine and slightly wavy; it does not join the massive receptaculum seminis apically but from below (hammerhead shape). It is of interest that in specimens collected in July and September the spermatheca was full. Subgenus Trichia s. str. The shell is flattened; the contour of the whorls is not dome-like but conical. The umbilicus is wider than in representatives of the subgenus Petasina. The inner dart sacs are not so sharply separated. There are 4 mucous glands, each with 2 branches; in some glands the second branch is reduced. The mucous glands are attached con- siderably above the tips of the inner dart sacs. The receptaculum seminis is oval and rather small. Whether the conchological distinctions on which the 7richia hispida species group (7. plebeia, including “sericea,”” septentrio- nalis, concinna) is based (Forcart, 1965) are justified could not be investigated here for lack of sufficient material. Two species of this group (7. р/еБега, T. concinna) are discussed as separate species in this paper, although it has not been possible to deter- mine whether the anatomical distinctions found are species-specific or just due to seasonal variation. Trichia (Trichia) plebeia (Draparnaud, 1805) Figs. 119-127; Pl. МИ, 25, 26 Four specimens were dissected that had been collected from Grdlorez village, near Prague, Czechoslovakia, on 18 September TRICHIA SYSTEMATICS FIGS. 119-123. Trichia (Trichia) plebeia (Draparnaud), Grdlorez village, near Prague, Czechoslovakia, 18 September 1968. 119, reproductive tract; 120, inner structure of vagina in dart sac region; 121, penis, penis sheath partly removed; 122, cross-section of verge; 123, cross-section of epiphallus. 1968 by V. Hudec and identified by him (Figs. 115-119; Pl. V, 25). Another 4 speci- mens originated from Bodetal in the Harz region, German Democratic Republic. These were collected in June 1969 by E. Clauss and identified as Trichia “sericea” (Figs. 124-127; PI. VII, 26). The Czech specimens are as follows: there is a smooth curve where the sperm- oviduct continues as the free oviduct, which is also gently bent. There are 4 mucous glands, each with 2 or 3 branches. The dart sacs are elongate, opposite to one another and to the upper part of the vagina. The dart sac region is not set off but passes smoothly into a conically taper- ing lower vagina. The vaginal plicae are relatively well developed but form no lobes. The flagellum is about the same length as the epiphallus, which bends sharply 2-3 times; its various parts are connected by а membrane _ stretched between it, the penis and the vas deferens. The penis is fusiform and slightly curved. The verge is marked by a few incomplete ring grooves. The seminal duct is not attached to the inner walls of the papilla by thin longitudinal bands but is fused to the wall at 2 diametrically opposed spots, and the 2 nonadhering sides between them constitute the 2 intrapapillar cavities (Fig. 122). The spermathecal duct is somewhat convoluted; the spermatheca is small and oval and almost reaches the albumen gland. The German specimens agree in most respects. Note, however, that they rather resemble 7. concinna in others, such as the unwound position of the epiphallus, which is not held in the bent state by the connec- tive tissue membrane; the smooth outer aspect of the papilla, which lacks ringed grooves; and by the presence of short additional vaginal plicae. The spermatheca was larger in these specimens than in 7. plebeia from Czechoslovakia and did not reach the lower edge of the albumen gland. Trichia (Trichia) concinna (Jeffreys, 1862) Figs. 128-131; Pl. VII, 27 Four specimens were dissected. They were collected in the vicinity of Roznave, 34 SHILEYKO FIGS. 124-127. Trichia (Trichia) plebeia (Draparnaud), Bodetal, Harz region, German Democratic Republic, June 1969. 124, reproductive tract; 125, inner structure of vagina in dart sac region; 126, cross-section of verge; 127, penis, penis sheath partly removed. Slovakia, Czechoslovakia, by V. Hudec on 15 July 1962 and identified by him. The spermoviduct curves weakly at the emergence of the oviduct and at the bend of the oviduct. There are 4 mucous glands, each with 2-3 branches. The dart sacs are short and club-shaped, closely pressed to- gether and partly fused. The dart sac regions are vaguely offset from the cylindri- cal vagina. The flagellum is almost 1.5 times longer than the cylindrical, weakly curved epiphallus. As in both preceding species (7. plebeia, T. unidentata), the male ducts are bound by a fine, transparent, connective tissue membrane. The verge is cylindrical or slightly fusiform, with a large papillar lacuna. The inner structure of the papilla is as in 7. plebeia. The spermathecal duct is almost straight. The receptaculum seminis is rather bulky and _ irregularly shaped; it just fails to reach the albumen gland. Trichia (Trichia) hispida (Linné, 1758) Figs. 132-137; Pl. VII, 28 Forty specimens were examined. They were collected as follows: in the USSR, 11 specimens from the Lenin Mountains in Moscow on 12 July 1970; 13 specimens from the Ilyich Foundation settlement near the Serebryanka River, Moscow region, on 15 August 1970; 4 specimens from the garden of the Zoological Institute, USSR Academy of Sciences, Leningrad, on 28 September 1965; and 5 specimens from Stryisky Park, Lvov (Eastern Ukraine), on25 September 1969. All these were collected and identified by me. Further, 4 specimens were collected from a garden in the town of Quedlinburg, German Democratic Republic in April 1969 by E. Clauss and identified by him. Lastly, 3 specimens from the region of Cologne, German Federal Republic, were collected by H. Nordsieck and identified by V. Hudec. At the point where the oviduct emerges from the spermoviduct, there is a smooth curve; the oviduct is also slightly bent. There commonly are 4 mucous glands, rarely 3; usually each has 2 branches but secondary branching is not uncommon, so the total number of branches may amount to 12. On the other hand, some branches may be reduced, bringing down the total to 4 or 5. The inner pair of dart sacs may be TRICHIA SYSTEMATICS 35 FIGS. 128-131. Trichia (Trichia) concinna (Jeffreys), Roznave, Slovakia, Czechoslovakia, 15 July 1962. 128, reproductive tract; 129, inner structure of vagina in dart sac region; 130, cross-section of verge; 131, penis, penis sheath partly removed. a ESS INES ЕСН» FIGS. 132-137. Trichia (Trichia) hispida (Linne), Lenin Mountains, Moscow, USSR, 12 July 1970. 132, reproductive tract; 133, penis, penis sheath partly removed; 134, inner structure of vagina in dart sac region; 135, cross-section of verge; 136, mantle collar; 137, cross-section of epiphallus. 36 SHILEYKO developed to approximately the same degree as the outer, or they may be a little less developed. Their upper tips always considerably surpass those of the outer sacs, the mouth of the mucous glands is always considerably above the upper ends of the inner dart sacs. The lower vagina is rather long and cylindrical. The vaginal plicae are narrow but distinct. There are small lobes at the opening of the dart sacs. The length of the flagellum is equal to or greater than that of the epiphallus. Usually the latter is sharply curved and sometimes forms a loop. Between penis and epiphallus a connective tissue membrane attaches the epiphallus to the penis sheath. The penis is massive and fusiform. Its sheath is smooth inside. The verge is also fusiform. The seminal duct is more clearly separated from the papillar| wall than usual; i.e., it is suspended by 2 fine longitudinal bands. The spermathecal duct is gently curving and sometimes weakly twisting. The re- ceptaculum seminis is oval and almost reaches the albumen gland. Trichia (Trichia) villosula (Rossmaessler, 1838) Figs. 138-142; Pl. VIII, 29 Three specimens were dissected. These were collected from Babice, near Brno, Moravia, Czechoslovakia, on 16 August 1968 by V. Hudec and identified by him. The oviduct makes a rounded bend. There are 4 mucous glands, of which some have 2 branches. The dart sacs are very elongate, closely opposed to each other and to the upper vagina. They are massive; the tips of the inner sacs extend beyond those of the outer sacs. Anatomically this species stands apart from others of the subgenus Trichia because the tips of the inner dart sacs reach the base of the mucous glands. The vagina is long, slowly narrowing to the genital atrium. The vaginal plicae are rela- tively well developed; they do not form lobes. The flagellum is somewhat shorter than the epiphallus, which is smoothly curving and linked to the penis by connec- tive tissue bands in which there are muscle Ns ts 141 FIGS. 138-142. Trichia (Trichia) villosula (Rossmaessler), Babice, near Brno, Moravia, Czechoslovakia, 16 August 1968. 138, reproductive tract; 139, inner structure of vagina in dart sac region; 140, penis, penis sheath partly removed; 141, cross-section of verge; 142, cross-section of epiphallus. TRICHIA SYSTEMATICS 37 fibers. In addition, the vas deferens, epi- phallus and penis are bound together by a fine translucent membrane. The penis is fusiform and slightly curved. On approach- ing the genital atrium it rapidly narrows and forms a distinct bend. The verge is club-shaped, with large papillar lacuna and a wide slit for its outlet. Inside the penis sheath at the junction with the genital atrium there is a long fold. The seminal duct is attached to the papillar wall by 2 fine longitudinal bands. The spermathecal duct is thin, with an S-shaped bend. The receptaculum seminis is rounded and almost reaches the albumen gland. Judging from Polinski’s (1924) drawings, the specimen from the town of Zakopane, Galicia, in the High Tatra region of S Poland near the Czechoslovakian border, had a flagellum approximately equal in length to the epiphallus or even a little longer. A penial band was not present. Trichia (Trichia) striolata (C. Pfeiffer, 1828) Figs. 143-147; Pl. VIII, 30 Four specimens were examined. They were collected in the Cologne area, German Federal Republic, on 30 May 1963 and identified by V. Hudec. The free oviduct twice bends sharply, in the shape of an S. There are 4 mucous glands, of which some have 2 branches. Both the inner and outer dart sacs are of equal size. The inflated dart sac region is sharply offset from the thinner, lower part of the vagina. The vaginal plicae, which arise rather suddenly, do not form any lobes. The flagellum is about as long as the epiphallus or a little longer. The connective tissue membrane between the distal por- tions of the male duct is well developed. The penis is fusiform, massive and bulbous; the penis sheath is smooth inside; the distal MFE FIGS. 143-147. Trichia (Trichia) striolata (C. Pfeiffer), Cologne area, German Federal Republic, 30 May 1963. 143, reproductive tract; 144, inner structure of vagina in dart sac region; 145, penis, penis sheath partly removed; 146, cross-section of verge; 147, cross-section of epiphallus. 38 SHILEYKO part of the penis is cylindrical. The verge is club-shaped, with a very widely split fora- men and a large papillar lacuna. The semi- nal duct is attached to the papillar wall by 2 longitudinal bands. It differs from the other members of the subgenus in that 1 of the 2 intrapapillar cavities surrounding the seminal duct is larger than the other and has very friable, corroded walls with many hollows and lamellae (Fig. 146). The spermathecal duct repeats the bend of the oviduct in its basal part. It is thin and ends in an oval receptaculum seminis that almost reaches the lower edge of the albumen gland. Trichia (Trichia) danubialis (Clessin, 1874) Figs. 148-153 Two specimens were examined. They were collected near Petrzalka village, near Bratislava, Slovakia, Czechoslovakia, on 15 April 1965 by V. Hudec and identified by him. The oviduct is straight or faintly curv- ing. The mucous glands, originally 4, may be reduced to 2; some of them show multiple secondary branchings. The dart sacs are well developed and massive; the outer pair is larger than the inner. The lower vagina is very short and not set off externally. The 2 pairs of vaginal plicae occupy lateral positions and abruptly stop below the mouth of the dart sac ducts and before the genital atrium (Fig. 150); they do not show any lobes. The flagellum is 2/3 the length of the epiphallus and is thin and curving. In its distal part the epiphallus sharply curves twice. The penis sheath has an internal longitudinal fold. The verge is short and fusiform; it occupies only the proximal part of the cavity of the penis sheath. The seminal duct is attached to the papillar wall by 2 longitudinal bands. In addition to the 2 intrapapillar cavities thus surrounding it, there is another unpaired cavity in the papillar wall (Fig. 151). The papillar lacuna is very large. The spermathe- cal duct is short and thick, only weakly curving. The receptaculum seminis is pear- shaped and does not reach the lower part of the albumen gland. FIGS. 148-153. Trichia (Trichia) danubialis (Clessin), PetrZalka village, Bratislava area, Slovakia, Czecho- slovakia, 15 April 1965. 148, reproductive tract; 149, inner structure of vagina in dart sac region; 150, penis, penis sheath partly removed; 151, cross-section of verge; 152, cross-section of|epiphallus near penial retractor; 153, cross-section of epiphallus in proximal part. TRICHIA SYSTEMATICS 39 This species is usually treated as a sub- species of 7. striolata, but the conchologi- cal as well as anatomical distinctions are without doubt important enough to war- rant full species status. It should be noted that the photograph of 7. striolata in Lozek’s book (1956) depicts a specimen taken from the same place as the 7. danubialis we have studied. Unless one takes these forms to be sympatric, one may suppose that Lozek's 7. striolata is the snail here identified as 7. danubialis. The peculiarities of the outer mor- phology of the genitalia of 7. danubialis have been previously noted by Hudec (1964). Genus Edentiella Polinski, 1929 The shell is similar to that found in the representatives of the subgenus Petasina of Trichia but is distinguished by narrower umbilicus and by the absence of a tooth in the aperture. The mucous glands are long and rather well developed. The flagellum length is equal to the joint length of penis and epiphallus or a little shorter. The inner structure of the verge is as in the Cau- 0,5 см 154 casian Caucasigena; i.e., there is a system of intrapapillar cavities that are divided by longitudinal septa. The seminal duct is attached to the interior papillar wall by a single longitudinal band. Type-species: Helix edentula naud, 1805. Edentiella bakowskii (Polinski, 1924) Figs. 154-159; Pl. VIII, 31 Five specimens were examined. | col- lected them in a _ beech forest in the vicinity of Kvasi village, near the town of Rakhov, in the Transcarpathians (Ruthenia), Ukrainian SSR on 14 September 1969, and | identified them. As the spermoviduct passes to the ovi- duct, it suddenly narrows and makes a slight curve. Further down the oviduct gradually widens. The mucous glands are very well developed; they are about as long as the female tract between the attachment of the spermathecal duct and the genital atrium, or a little shorter. According to my observations, the glands, usually 6-8 in number, are not grouped in bundles but are located around the oviduct at the level of the spermathecal duct takeoff. Hudec Drapar- 158 FIGS. 154-159. Edentiella bakowskii (Pokinski), Kvasi village, Rakhov district, Transcarpathian region, Ukranian SSR, 14 September 1969. 154, reproductive tract; 155, penis, penis sheath partly removed; 156, cross-section of epiphallus; 157, cross-section of verge; 158, inner structure of vagina in dart sac region; 159, mantle collar. 40 SHILEYKO PLATE | (not actual size). 1a, b, с, Odontotrema diplodon Lindholm, Chatkal Range, NW Tian-Shan Mountains, Kirghiz SSR, May 1972. 2a, b, c, Leucozonella ferghanica Lindholm, Sary-Chileck Nature Reserve, Chatkal Range, NW Tian-Shan Mountains, Kirghiz SSR, 6 July 1966. 3a, b, c, Leucozonella caryodes (Westerlund), Talas Range, NW Tian-Shan Mountains, Kirghiz SSR, 4 June 1972. Да, b, с, Leucozonella rubens (Martens), foothills of the Kirghiz Range, NW Tian-Shan Mountains, Kirghiz SSR, 15 June 1972. TRICHIA SYSTEMATICS 41 PLATE II (not actual size). Ба, b, с, “Euomphalia”” regeliana (Martens), foothills of the Kirghiz Range, NW Tian-Shan Mountains, Kirghiz SSR. ба, b, с, Leucozonella rufispira (Martens), Anzob pass, Hissar Range, W Tian-Shan Mountains, Tadzhik SSR, 28 July 1968. 7a, b, c, Leucozonella retteri (Rosen), Kandara Valley, Hissar Range, W Tian-Shan Mountains, Tadzhik SSR, 2 July 1967. За, b, с, Leucozonella сапа Shileyko, sp. nov., holotype, Khodzcha-Obi-Garm Rest Home, Hissar Range, W Tian-Shan Mountains, Tadzhik SSR, 28 May 1968. 42 SHILEYKO PLATE III (not actual size). 9a, b, c, Hygrohelicopsis darevskii Shileyko, sp. nov., holotype, Chegem Valley, N slope of central Caucasus, USSR, 10 August 1965. 10a, b, с, Teberdinia zolotarevi (Lindholm), topotype, Teberdia Nature Reserve, NW Caucasus, USSR, 24 July 1958. 11a, b, с, Kokotschashvina tanta Shileyko, sp. nov., paratype, between Sioni and Kazbegi villages, slope of central Caucasus, Grusinian SSR, July 1969. TRICHIA SYSTEMATICS 43 PLATE IV (not actual size). 13a, b, с, Kokotschashvilia phaeolaema (Boettger), Chegem Valley, М slope of the central Caucasus, USSR, 14 May 1970. 14a, b, с, Caucasigena (Caucasigena) armeniaca (|. Pfeiffer), Mt. Bzovdal (Mt. Todar), Stepanavan region, Armenian SSR, 18 July 1951. 15a, b, с, Caucasigena (Caucasigena) tschetschenica (Retowski), Kurtatin Valley, middle course of Phiagdon River, central part of N Caucasus, USSR, 19 September 1970. 16a, b, c, Caucasigena (Caucasigena) rengarteni (Lindholm), Khushtosyrt village, Chegem Valley, central part of N Caucasus, USSR, 14 May 1970. 44 SHILEYKO PLATE V (not actual size). 17a, b, с, Caucasigena (Caucasigena) rengarteni (Lindholm), old settle- ment of Ara-Boran, between Chegem and Baksan valleys, central part of N. Caucasus, USSR, 17 May 1970. The form described by Lindholm under the name Helix gerassimovi. 18a, b, c, Caucasigena (Caucasigena) eichwaldi (L. Pfeiffer), Chmi village, Darial Valley, near Ordjonikidze (North Osetia), central part of N Caucasus, USSR, 8 May 1970. 19a, b, c, Caucasigena (Anoplitella) schaposchnikovi (Rosen), old settlement of Ara-Boran, between Chegem and Baksan valleys, central part of N Caucasus, USSR, 17 May 1970. 20a, b, c, Caucasigena (Anoplitella) schaposchnikovi (Rosen), Chegem Valley (damp, shady slope) near Khushtosyrt village, central part of N Caucasus, USSR, 15 May 1970. TRICHIA SYSTEMATICS 45 PLATE VI (not actual size). 21a, b, c, Xerocampylaea zelebori (L. Pfeiffer), Roumania. 22a, b, c, Caucasigena (Dioscuria) thalestris (Lindholm), Novy Aphon, N of Sukhumi, Black Sea region of Georgia, Grusinian SSR, 2 July 1913. Specimen not fully mature. 23a, b, с, Plicuteria lubomirskii (Slóssarski), Olomouc, Moravia, Czechoslovakia, 16 April 1964. 24a, b, c, Trichia (Petasina) unidentata (Draparnaud), central Czechia, Czechoslovakia, 8 September 1969. 46 SHILEYKO PLATE VII (not actual size). 25a, b, c, Trichia (Trichia) plebeia (Draparnaud), Grdlorez village, near Prague, Czechoslovakia, 18 September 1968. 26a, b, c, Trichia (Trichia) plebeia (Draparnaud), Bodetal, Harz region, German Democratic Republic, June 1969. 27a, b, c, Trichia (Trichia) concinna (Jeffreys), Roznave, Slovakia, Czechoslovakia. 15 July, 1962. 28a, b, c, Trichia (Trichia) hispida (Linné), Lenin Mountains, Moscow, USSR, 12 July 1970. (1965) notes the presence of 2 glands, each with 4 branches. Presumably these glands may either be sited independently or be grouped in 2 fascicles. The dart sacs are elongate, well separated from the upper vagina. The outer and inner pairs are developed to approximately the same degree. The lower vagina is cylindrical. The vaginal plicae are clearly formed. They may be weakly inflated near the outlet of the dart sac ducts, but they do not form lobes. The flagellum is 1.5-2 times longer than the straight epiphallus (in Hudec’s work the epiphallus is described as being curved); this organ passes smoothly into the penis without any demarcation. There is either no membrane between vas deferens, epiphallus and penis, or it is insignificant. The penis is TRICHIA SYSTEMATICS 47 PLATE VIII (not actual size). 29a, b, с, Trichia (Trichia) villosula (Rossmaessler), Babice, near Brno, Moravia, Czechoslovakia, 16 August 1968. 30a, b, c, Trichia (Trichia) striolata (С. Pfeiffer), Cologne area, German Federal Republic, 30 May 1963. 31a, b, c, Edentiella bakowskii (Polinski), Kvasi village, Rakhov district, Transcarpathian region, Ukranian SSR, 14 September 1969. slightly bulbous; its distal part is cylindri- cal. The verge is perfectly fusiform. The seminal duct is attached to 1 side of the inner papillar wall by a longitudinal band; i.e., it is encircled by 1 cavity. The oppo- site papillar wall encloses a system of intra- papillar cavities divided by septa (Fig. 157). These more or less parallel cavities anasto- mose and are also all connected with the lumen of the penis sheath by the papillar lacuna. The spermathecal duct is fine and passes very smoothly to the elongate-oval receptaculum seminis, which ends quite a distance short of the lower part of the albumen gland (Fig. 154). In 1 specimen studied, the receptaculum seminis had an isolated position. Taking into consideration the marked conchological similarity, one may suppose that 7. edentula (Draparnaud) and 7. bielzi (Schmidt) [see also Sods, 1917] are close to our species. Likharev & Rammelmeyer (1952), following Polinski (1924), classify it as a variety of 7. bielzi, noting that the dart sacs attach in the middle of the “uterus” (i.e., the part here differentiated into upper and lower vagina), which is characteristic for Polinski’s section Fil- icinella. Our material, however, shows that the dart sacs are so placed in many European Trichia. Thus the validity of Filicinella becomes questionable. The final answer to this question will depend on detailed investigation of the inner structure of 7. filicina, the type-species of this section. Note that conchologically 7. filicina is nearer to Trichia s. str. than to Edentiella. 48 le 2. 14. 15. 16. 17: SHILEYKO IDENTIFICATION KEY Species inhabiting central Asia Species inhabiting the Caucasus or Europe Shell aperture with 2 teeth Shell aperture without teeth . Shell globose Shell more or less depressed . Shell white, thick; body whorl lighter in color than other whorls; penis papilla without appendix near base Shell horny to reddish, uniformly colored, with white spiral band; walls moderately thick; penis papilla with appendix near base . Shell diameter 14 mm or more Shell diameter 11 mm or less . Shell surface with distinct periostracal hairs; diameter 9 mm or less Shell surface without periostracal hairs or diameter more than 9mm . Shell surface coarsely wrinkled Shell surface rather smooth . Shell diameter 15mm or more, surface with spiral lines; no groove on the penis papilla (i.e., it has become closed and is now an intrapapillar cavity) Shell diameter 14 тт or less, surface without spiral lines; there is a deep, open groove on the penis papilla . Species inhabiting the Caucasus Species inhabiting Europe . Shell distinctly hirsute Shell without hairs or with very short hairs . Shell with coarse, prominent radial ribs Shell not ribbed or sculptured with rib-striae . Shell diameter 9 mm or less Shell diameter 10 mm or more . Shell pale greenish, with large aperture, body whorl voluminous, umbilicus minute Shell does not have all features enumerated Shell periphery angular Shell periphery rounded Shell yellowish-brown, almost uniform in color; on the penis papilla there is a longitudinal groove Shell color is other; penis papilla without groove Shell depressed, with washed-out radial spots; inner pair of dart sacs not visible on external inspection of upper vagina Shell with more or less prominent spire; inner pair of dart sacs visible without dissection of upper vagina Shell nearly globose, diameter barely exceeding the height, uniformly brown or chestnut in color Shell more or less depressed, diameter markedly 2 dl Odontotrema diplodon 3, 4. 6. Leucozonella caryodes >} Leucozonella rubens Leucozonella mesoleuca Leucozonella caria 7. Leucozonella ferghanica 8. Leucozonella retteri Leucozonella rufispira 10. 23. Kokotschashvilia holotricha 11: Caucasigena tschetschenica 12. Caucasigena armeniaca 13: Caucasigena thalestris 14. Caucasigena rengarteni 15: Teberdinia zolotarevi 16. Hygrohelicopsis darevskii 17? Kokotschashvilia phaeolaema 18. 19. 20. TRICHIA SYSTEMATICS exceeding the height, gray or white, often with spiral bands Spiral bands are very faint or absent Spiral bands are distinct Shell diameter 19-22 mm; in the limits of penis papilla, seminal duct attached to inner papilla wall by 2 longitudinal membranes Shell diameter either less than 17mm or more than 22 тт; seminal duct closely adhering to 1 side of inner wall of papilla Shell diameter 17mm or less; inner wall of penis 49 18. 19: 27: Kokotschashvilia makvalae 20. papilla with numerous plicae Kokotschashbilia eberhardi Shell diameter 22mm or more; inner wall of penis papilla is smooth 21. Umbilicus is narrow Umbilicus is wide, perspective 22. Shell rather strongly striate Shell nearly smooth Kokotschashvilia tanta Caucasigena abchasica 22. Caucasigena eichwaldi Caucasigena schaposchnikovi 23: 24. 25. 26. 29. 28. 29 30. Shell yellowish-white, spire conical, aperture with fra- gile edges, almost without lip; the vaginal plicae are split into patterns of regular prismatic lamellae Shell does not have all features enumerated Shell of mature specimen with distinct periostracal hairs Shell of mature specimen without hairs or with rare faint hairs Hairs long Hairs rather short Shell aperture with basal tooth Shell aperture toothless Shell perforate; wall of penis papilla contains system of cavities, separated by septa Shell umbilicate; wall of penis papilla contains pair of symmetrically disposed cavities Shell diameter 6-9 mm Shell diameter 10-13 mm Shell widely umbilicate Shell moderately umbilicate Shell distinctly angulate; in addition to paired intra- papillar cavities, there is a narrow unpaired cavity Shell obtusely angulated; 1 intrapapillar cavity larger than the other and with corroded internal wall DISCUSSION some species Plicuteria lubomirskii 24. 25: 26. Trichia villosula Trichia plebeia Trichia unidentata 27: Edentiella bakowskii 28. 23 30. Trichia concinna Trichia hispida Trichia danubialis Trichia striolata reached the mountainous Analyzing the results obtained, | have concluded that the ancestral trichiae developed in 2 ways, 1 taken by the Asian forms and the other by the Caucasian- European forms. Asian forms: by eastward migration, regions of central Asia. These formed shells the common shape of which is now characteristic for most of the central Asian Helicoidea ( Leucozonella). Conchologically separate from these are Odontotrema diplo- don and Leucozonella сапа, the latter strongly resembling European Trichia $. str. 50 SHILEYKO We may assume that the verge of the initial primitive forms was a simple tube without the longitudinal groove on its surface and without any intrapapillar cavity (Fig. 160, I). In Recent Asiatic species, all the basic phases of intrapapillar cavity formation can be observed (Fig. 160, II-IV). At first the longitudinal groove on the surface of the verge is formed in the proximal part only (Fig. 160, Il, Leucozonella ferghanica); then it extends along the whole papillar length (Odontotrema diplodon), deepens (Leucozonella caryodes; Fig. 160, 111) and tends to become closed ([. rubens, L. rufispira); the final phase is the fully closed groove, which now forms an intrapapillar cavity (L. retteri, L. сама; Fig. 160, IV). In these forms the seminal duct adheres closely to the inner papillar wall on 1 side, and the intrapapillar cavity embraces it from all other sides. A remnant of the groove is the papillar lacuna that exists in all species discussed and by means of which the intrapapillar cavity connects with the cavity of the penis sheath. The circum- stance that most Asiatic species have an intrapapillar cavity that remains open testi- fies to their relatively primitive state. This conclusion is strenghtened by the fact that other Asiatic Hygromiidae are also the most primitive representatives of their groups (Shileyko, 1970). We must further conclude that, on the whole, the “Helicidae'” аист. had a wider distribution in the past than now and that necent Asiatic representatives of the group are relics from the Tertiary period. Caucasian-European forms: the second path of development was the formation of 2 groups: the Caucasian and European groups. In these the formation of the intra- papillar structures evolved by another principle than in the Asiatic group. Basically, intrapapillar cavities occur in pairs in the papilla walls, embracing the centrally placed seminal duct from 2 opposite sides; the papillar lacuna forms in a parallel manner. In other words, it is necessary to assume that the рарШаг lacunae of Asiatic and European-Caucasian species are not homologous. Similar types of papillar structure exist in such clearly independent groups as Hygrohelicopsis, Teberdinia, Plicuteria, Trichia s. str. and in some species of the genus Kokotschashvilia (Fig. 160, V). The very complex system of septate intrapapillar cavities such as are found in Edentiella and Caucasigena (Fig. 160, |X) might possibly have formed in the way shown by species such as Trichia striolata and 7. danubialis (Fig. 160, VII and VIII, respectively). It consists of the fragmenta- tion of 1 papillar wall, in which is thus formed a tertiary cavity, i.e., a derivative cf the secondary cavity, which then further disintegrates into a series of narrower cavi- ties. Nevertheless, the totality of characters of another order makes it necessary to recognize an independent origin for the European Zdentiella and the Caucasian Caucasigena. Within the genus Kokotschashvilia one can observe a number of variants in papilla structure that we are here attempting to derive from the same initial point of de- parture (Fig. 160, V). The papilla structure in K. makvalae (Figs. 55; 160, V) is nearest to this initial point: the seminal duct is surrounded by a pair of intrapapillar cavi- ties. A capillary runs in the papilla wall along 1 of the longitudinal tissue bands attaching the seminal duct to the inner papillar wall. In К. tanta this capillary is retained. The seminal duct is displaced toward 1 of the papillar walls and adheres to it closely; on the opposite side there still is a band of tissue separating the 2 cavities (Fig. 160, X). This division is absent in K. eberhardi, in which the cavity, now single, has a crescent-shaped cross-section, smooth- walled in not fully mature specimens and sinuously folded in mature specimens (Fig. 160, XI, ХИ). As a result, the type of papillar structure of К. eberhardi 1$ formally the same as in the Asiatic Leucozonella retteri and L. caria (Fig. 160, IV). To interpret this similarity as a purely formal one is justified from 2 considera- tions. In the first place, the formation of the intrapapillar cavity embracing the semi- nal duct т К. eberhardi has been traced in allied species of Kokotschashvilia (see above). In the second place, | have made a series of dissections of К. eberhardi at various stages of maturity which demon- strates that in specimens not fully adult the seminal duct is still attached by a longitudi- nal tissue band to the papilla wall opposite to the wall it adheres to, the band dividing the intrapapillar cavity into 2 chambers (Fig. 161) just as in adult K. tanta. In K. holotricha, the seminal duct does not adhere to the papillar wall on one side but is well removed, being held in a central position by a thin longitudinal band and 51 TRICHIA SYSTEMATICS “ешае/оэеца "У ‘ИХ “uawidads элпзеш 2 if “uaunoads элпзешиаи “pseysaga “Y ‘|X ‘EJUE] EIJIAYSEYIS10YOYH ‘X :5135а/ецз “9 ’moyıuyasodeyas D “pjemyola D ‘“иазлебиал “Y ‘eseiuauue tee а мае ‘eyjanuapz "ХЕ u 7 ‘WIA “e28/0/1S ем ‘ИЛ ЕКО ОЕ ТРЕ IA, “O18 “epidsiy *1 EuSIaUoS "L ‘в/ада/а eıyau] “11yS4ILUOQN| enaınaıld ‘иузлалер sisdoajayosbAY ‘A ‘eed “7 ‘alas “7 ‘Al “sapoAseo “7 ‘| “ed jewixoid ‘еэ/иецблау ejpeuozoona7 гиоро]а!р НЕ ‘11 ‘qed ¡e1sip ‘еэ/иеублау е//эиогоэпа7 ‘| :SW107 Buimojjo} ay) Aq рэциашаха se ’saıaads ueiseoned pue ueadoing и! Аилеэ palta e er) $эилеэ Jejjide die JO чоцешлоу JO sapow поел ‘||| Х-Л ‘$э1ээ4$ э1зе!$\у и! 210016 Jejjıded e шод} Аилеэ ле!дедели! au} JO чоцецло} епрелб ’Д|-|| "aqnı ajdwıs 8 se 3 1 ejjıded jeniu ‘| "papeys si 19mp jeulwas JO и013985-5$045) "sdıysuoireje4 4184} pue ‘3е| “Ss e/y9/4/ JO $Ч9пол6 зпомел ayi ul э1п19п3$ эблэл ul UONBIIeA ‘091 ‘914 92 SHILEYKO 2006 FIG. 161. Formation of intrapapiliar cavity in Kokotschashvilia eberhardi during ontogenesis (т cross-section). |, Initial paired cavities in papilla wall and broad contact between seminal duct (shaded) and inner papilla wall. Il, The cavities grow and there develops between them a longitudinal band that attaches the seminal duct to the papillar wall opposite the area of fusion with the wall. Ill, Connection of seminal duct and papillar wall by the band is lost by coalescing |cavities. IV, Except for the one point of adherence, the seminal duct is now completely isolated from the surrounding papillar wall. encircled by the intrapapillar cavity (Fig. 65). Finally, in K. phaeolaema additional sinuses have formed in the papilla wall (Fig. 160, XIII). Just as the papillar structures т Edentiella (Fig. 160, IX) are derived from Trichia striolata and T. danubialis (Fig. 160, VII, VIII), those of Caucasigena (Fig. 160, IX) are derived from Kokotschashvilia in a parallel and independent manner (Fig. 160, XI, ХИ, IX): in К. eberhardi the inner papillar wall opposite to the seminal duct shows clear traces of dissolution (Fig. 70; 160, XII). There still is 1 more variant in papillar structure: in Teberdinia there are both a pair of intrapapillar cavities and a deep groove on the penis surface; 1 of the cavities reaches into the lobe separated off by the groove (Fig. 160, VI). At the present time it is difficult to give any comparative morphological estimation of this variant as we do not yet know any related types of papilla structure. We might imagine the formation of a third cavity from a closing groove, in which case we would get a variant corresponding to 7. danubialis (Fig. 160, VIII). One might suppose that with the groove closing and with a simultaneous reduction of the band attaching the seminal duct to the papilla, there might arise the variant answering to Edentiella and Caucasigena (Fig. 160, IX). Assessment must be left to a later date, although the second of the modes discussed seems quite likely. If | have paid so much attention to the details of verge structure, it is because the richness and variability of intrapapillar structures will display and reflect the com- mon evolution of the group. No other feature offers us so rich a material for phylogenetic reconstruction. However, while considering variants of the papilla structure in detail, | am not proposing to attach excessive importance to the charac- ters of the papilla. The phylogenetic scheme (Fig. 162) here submitted rests as far as possible on the feature complex valuable for practical taxonomy. In the diagnosis of genera (see Systematic Part) various categories of characters are given, some of which need to be considered in discussing the proposed scheme of classifi- cation. In addition to the anatomical features treated in connection with the Asiatic group, we shall also consider the 4 concho- logical types occurring in that geographical area. The first type, exhibited by Odon- totrema diplodon (PI. |, 1), is very clearly distinguished by its aperture fold; the second type is the almost globose shell of Leucozonella caryodes, L. rubens and L. mesoleuca; third, we have the flattened shell type in L. ferghanica, L. rufispira and [. retteri; and the fourth is а purely European conchological type: [. сама (European Trichia are relatively small, thin-shelled, usually brown, usually hirsute). The primitive species obviously is Odon- totrema diplodon. Apart from the charac- ters of the penial papilla (open groove) already mentioned, we must consider the high degree of isolation of the dart sacs. Taking into consideration the oligomeriza- tion principle of homologous organs (Dogel, 1954), and the starting point of this principle, i.e., the multiple inception of newly formed organs, we must conclude that the ancestral forms had 4 equivalent dart sacs, each of which included a dart. The first phase of oligomerization was the (92) LO TRICHIA SYSTEMATICS "eIEJOLNS ‘1 ‘SAL /syergnuep ‘J ‘PAL ‘qed и! “(2142/41 ) e1Y2141 ‘AL Jesuipsagay ‘| /(еи1зезаа) 2142111 ‘ad “e/191091/4 “4 ‘EW91101U0PO ‘O “euldsyns 7 *y47 “suegna 7'947 У "7'417 /earueybiay 7 'y7] Гемез ‘7 ‘вэл :5аро/леэ ejjauozoanaT ‘27 гезиез ‘y ‘3; ‘ешае/оэеца ‘y dy !3erenyew "У ‘WU “ey9/130/04 *У ‘UM //рлецлада EIJIAYSEY2520 0 y ‘эм sisdo91jayolony ‘H “ejpanusp3 “3 :(емпэзо!а) eusbiseaney ‘а “(eusbiseone)) euaßisesney ‘D г(еаиаои\) euabiseoned “y “x9e\q ul рахлеш эле salsads |епзэе 0} Buipuodsa1109 sasejd ‘эблэл ay} JO syed Алецоцп|олэ шеш Z MOUS SMOJIY ‘эапз ajduwis e jo 110} eu} и! эблэл e pue soes злер ¡enba y }0 aouasaid ay} Aq ра7аазэелецо $120} |е435ээче Jo Xa[dW09 ay] 03 spuodsa1109 ease pajddi1s ayı *dnoJb “yey *$ e/y2/11 BY} jo auayos эзэцабо|Ача ‘291 ‘91+ 54 SHILEYKO loss of darts in the inner (upper) sacs. Further oligomerization of adventitious organs in the female tract causes their gradual reduction (maximal degree of reduction observed in Hygrohelicopsis). The presence of well-developed, clearly separated inner dart sacs approximately the same size as the outer sacs reaffirms the primitive condition of O. diplodon. The shell also shows the isolation of this species. From papillar and other features it is logical to consider Leucozonella ferghanica to be a primitive type in respect to other Asiatic species. The groove on the surface of the verge is only weakly traced, and the upper vagina is very long. [. caryodes is then a derivative; the groove on the verge surface is much longer and the upper vagina less so. The shell is more globose. L. rubens retains a globose shape but shows reduction of the inner dart sacs and a closer connection with the surface of the upper vagina. At the same time this species has acquired a number of quite important features such as a markedly shorter flagel- lum, more branched mucous glands, and the presence of an intrapapillar appendix, the purpose and origin of which is not clear. Leucozonella rufispira, another derivative of L. caryodes independent of L. rubens, is a distinct form by reason of other features. The flagellum length and degree of isola- tion of the dart sacs have not changed; the shape of the sacs has changed. The depth and general character of the papillar groove correspond to what we see in L. caryodes. The shell has acquired a more flattened form. Judging mainly by the shell, Leucozo- nella retteri is a derivative of L. rufispira. It has a qualitatively important new evolu- tionary feature, i.e., the final formation of the intrapapillar cavity, but generally ге- tains other characters of L. rufispira. In the series of the central Asiatic forms, Leucozonella caria is prominently distinct by its shell, which, in my view, like the shells of the European trichiae, retains the initial features characteristic for the ancestral forms. At the same time the species has the essential character that re- lates it to the true Asiatic group: its verge structure completely corresponds to L. retteri (Fig. 160, IV) and has nothing to do with European species. It is necessary to point out here that study of the external morphology of the genitalia does not engender understanding of the real essence of this species; using only the character of outer shape, we should regard L. caria as а member of European Trichia $. str., which would be very wide of the mark. To finish the discussion of the Asiatic group, we need to point out 1 further obviously primitive feature common to all these species: the regular internal longitudi- nal folding of the epiphallus which, in cross-section, gives to its lumen the shape of a multiradial regular star. Subsequently 2 neighboring longitudinal folds developed more strongly while the other folds were grouped on the opposite epiphallus wall. As a result there formed 2 main spermato- phore guide “'rails”” and a number of small ribs above them; in the furrows between these fit the rows of thorns on the sper- matophore surface (Fig. 163). Such a development of the 2 longitudinal folds is observed in the Caucasian and European species. As for the Asiatic species, the differentiation of the equal folds in the epiphallus lumen is only beginning, and distinctions between these 2 groups of folds are sometimes hard to make. The lumen of the epiphallus in cross-section retains in most cases the original type of multiradial, almost regular star (see Figs. 15122032) The European-Caucasian group of genera does not form one single developmental line, as 15 so well seen in the Central Asiatic forms, but is the product of some parallel and, to a certain degree, indepen- dent evolution, with the retention of a more or less definite complex of similar features. We do not find any anatomical characters permitting judgment on the evolutionary tendencies of the Caucasian and European groups. But the conchologi- cal distinctions between the European and Caucasian species are clearly seen. Generally characteristic for the European forms is a rather fragile, small- or medium- sized, usually hirsute shell of either a uni- form brown color or marked with a faint white line at the periphery, whereas the Caucasian forms are, as a rule, more solid, larger, usually not hirsute and show a variety of color besides brown (chestnut, white, lilac, greenish, yellowish, etc.). There are exceptions among the European forms, such as Plicuteria lubomirskii and Xero- campylaea zelebori (L. Pfeiffer) (Pl. VI, 21). The former also stands out anatomi- TRICHIA SYSTEMATICS 55 FIG. 163. Part of the spermatophore and trans- verse section of epiphallus (at right side) of Caucasigena (Caucasigena) eichwaldi. cally from the general run of species. The latter is discussed below. Almost all the European trichiae are forest animals. The Caucasian species are primarily con- nected with 2 types of biotope: (1) with rocks (Caucasigena s. str.) and (2) with either open and dry or damp slopes (the other forms, i.e., Hygrohelicopsis, Teber- dinia, Kokotschashvilia and the other 2 subgenera of Caucasigena). According to these habitats, we see 2 shell types: (1) either more or less flattened, ribbed shells, or (2) high to almost globose shells with relatively vague sculpture. The presence or absence of periostracal hairs on the shells of representatives of different species is not a constant feature: as a rule Kokotsch- ashvilia holotricha almost always is hirsute; Caucasigena rengarteni is generally not hir- sute, but there are some specimens with short hairs. | consider the presence of hairs to be a primitive character and believe that the disappearance of this feature in the Caucasian species is an independent and parallel development. Another conchological exception among European species is Xerocampylaea zele- bori, a species living among rocks as does the Caucasian group of Caucasigena s. str. It is now known that a rocky habitat usually produces a marked imprint on the aspect of the shell. The conchological similarity between these forms (flat, ribbed) caused earlier authors to regard Caucasigena as a separate section of the genus (or subgenus) Xerocampylaea Kobelt, 1871. As X. zelebori has not been carefully investigated anatomically, and its distribu- tion (Serbia, in part Roumania) is quite far from the Caucasus, | do not connect these groups, particularly as | am trying to establish the independence of the genus Caucasigena. Other evidence is that the shell of X. zelebori is otherwise quite clearly distinct from that of Caucasigena; the apex is acute and not obtuse, and the contour of the whorls is not dome-shaped. Lastly we must take into account similar features among ali the species examined. Except for the presence of 2 pairs of dart sacs, the only other common character is the presence of a papillar lacuna (in species with a completely formed intrapapillar cavity), by means of which the intrapapillar cavity communicates with the cavity of the penis sheath. We assume that when the opening into the cavity of the genital atrium is closed, the increased pressure of the intrapenial liquid helps to squeeze the papilla outside, in part owing to the presence of this lacuna. In copulation the papilla is not turned inside out but is moved forward (Fig. 72). These 2 features are the only ones common to all species discussed. Even the topography of the right-hand ocular retractor, which used to be taken as the basic anatomical criterion for the assignment of any species to either the “Helicellinae’’ auct. or the “Нудго- miinae,”” is not constant in the group studied. In the hygromiids living under arid conditions, the right ocular retractor passes alongside and not between the penis and vagina, and the shells of such groups have adaptive features such as very light color and strong calcification. Thus the retractor in question only passes near the distal part of the genitalia and not between them in Hygrohelicopsis darevskii (Fig. 42). Never- theless, the sum total of characters of another nature (shell, penial papillar struc- ture, pigmentation of integuments, ecology) indicates that H. darevskii obviously is a representative of the Trichiinae. From the above it is clear that, for the elucidation of a true and objective system, it is necessary to enlist as many and various characters as possible. This was attempted in the present work. 56 SHICENKO ACKNOWLEDGMENTS № is my pleasant duty to acknowledge gratefully the help of all those without whose kind assistance the present research would not have been completed. The scientists of the Zoological Institute, USSR Academy of Sciences, Leningrad, Drs. |. M. Likharev and Ya. I. Starobogatov, gave much helpful advice and instructions during the preparation of this manuscript. N. N. Akramovsky of the Zoological Institute, Armenian Academy of Sciences, furnished most valuable information with respect to Caucasigena armeniaca. | am greatly in- debted to Dr. V. Hudec, of the University of Prague, Czechoslovakia; Dr. E. Clauss, of the Institut für Pflanzenzüchtung at Quedlinburg, German Democratic Republic; and Mr. Z. Izzatulaev, of the Institute of Zoology, Academy of Sciences of Tadzhiki- stan at Dushanbe, for the important material they have provided. Some separate points of the discussion were constructively criticized by L. V. Shileyko and J. P. Ghubar, whose constant assistance and support | received while writing this paper. LITERATURE CITED DOGEL, V. A. 1954, Oligomerizatsia gomo- logichnÿkh organov. \zdatel’stvo Leningrad- skogo Universiteta, Leningrad, 366 р. FORCART, L., 1958, Trichia Hartmann, 1840—nomenklatorisch gültig. Archiv für Mol- luskenkunde, 87: 153-154. FORCART, L., 1965, New researches on Trichia hispida (Linnaeus) and related forms. Proceed- ings of the First European Malacological Con- gress: 79-93. HESSE, P., 1931, Zur Anatomie und Systematik palaearktischer Stylommatophoren. Zoologica, Stuttgart, 33(85): 1-59. HUDEC, V., 1964, Uber die Verbreitung der Schnecke Trichia striolata (C. Pfeiff.) in der Sudwestslowakei. Zoologicke listy, Brno, 13(3): 265-268. HUDEC, V., 1965, Systematische Stellung und Verbreitung von Trichia bakowskii in der Tschechoslowakei. Biologia, Casopis slovenskej Akademie Vied, Bratislava, 20(4): 245-259. HUDEC, V. & LEZHAWA, С. 1., 1969a, Drei neue Heliciden aus der Grusinischen SSR. Archiv flr Molluskenkunde, 99: 41-48. HUDEC, V. & LEZHAWA, С. 1., 1969b, Bemer- kungen zur Erforschung der Landmollusken der Grusinischen Sozialistischen Sowjetre- publik (11). Sbornik narodniho muzea и Praze, 25B(3): 93-155. KALITINA, 2. 1., 1958, К izucheniyu zonalnogo raspredeleniya nazemnykh mollyuskov severnykh sklonov Tsentralnogo Kavkaza i Vostochnogo Predkavkazya In Zdravo- okhranenie i meditsina v Severnoj Osetii, Ordzhonikidze, 7(1): 157-181. LIKHAREV, 1. М. € RAMMELMEYER, E. S., 1952, Nazemnye mollyuski fauny SSSR. Izdatel’stvo Akademii Nauk SSSR, Moscow- Leningrad, 511 pp. LIKHAREV, 1. М. & STAROBOGATOV, Ya. I., 1967, Materialÿ К faune mollyuskov Afgani- stana. In Mollyuski i ¡kh rol и biotsenozakh i formirovanii faun. \zdatel’stvo ‘‘Nauka,” Leningrad, 159-197. LINDHOLM, W. A., 1913, Neue Heliciden aus dem Kaukasus-Gebiete. Nachrichtsblatt der Deutschen Malakozoologischen Gesellschaft, 45: 137-144. LINDHOLM, W. A., 1929, Die gezähnten Helici- den des Kaukasus. Archiv für Molluskenkunde, 61: 205-211. LOZEK, V., 1956, К/с ceskoslovenskych mek- kysü. Bratislava. 437 pp. POLINSKI, W., 1924, Anatomisch-systematische und zoogeographische Studien über die Helici- den Polens. Bulletin de l’Académie des Sciences et des Lettres, series B, Krakowt 131-279, pls. 6-20. RETOWSKI, O., 1914. Materialen zur Kenntnis der Molluskenfauna des Kaukasus. /zviestiia Kavkazskii Muzei 6(4): 271-334. SHILEYKO, A. A., 1970, Ob’em sistema i filo- geniya gruppy Perforatella-Zenobiella- Chilanodon (Pulmonata, Helicidae). Zoologi- cheskii Zhurnal 49: 1306-1321. SHILEYKO, A. A., 1972, Some aspects of study of Recent non-marine gastropod mollusks. Results of Sciences. Zoology of Invertebrates, yol. 1, 188 p. (in Russian). $005, L., 1917, Zur Systematik und Anatomie der ungarischen Pulmonaten. Annales Musei Nati- onalis Hungarici, 15: 1-165. MALACOLOGIA, 1978, 17(1): 57-61 А LATITUDINAL PATTERN IN BIVALVE SHELL GAPING Geerat J. Vermeij and John A. Veil Department of Zoology, University of Maryland, College Park, Maryland 20742, U.S.A. ABSTRACT Bivalves with persistent posterior shell gapes comprise an increasingly smaller proportion of infaunal bivalve assemblages from Arctic to temperate to tropical coasts. This trend is not a taxonomic artifact, since it can be shown that deep-burrowing cold-water gapers are replaced in the tropics largely by tightly closing lucinids. We incerpret the increasing equatorward emphasis on complete shell closure among bivalves to reflect an increase in the intensity of predation. The latitudinal trend in predation and bivalve shell architecture is consistent with the microgeographical reduction in proportion of gaping from deep to shallow sediment layers. INTRODUCTION Although latitudinal gradients in diver- sity are well known and of extremely wide occurrence among both marine and non- marine organisms (see e.g. Fischer, 1960), much less is known about latitudinal varia- tions in animal architecture. Bakus (1969) has summarized gradients in algal morphol- ogy, and has related these to an equator- ward increase in grazing intensity. Bakus (1974) and Bakus & Green (1974) have described a gradient of increasing toxicity toward the tropics among holothurians and sponges. An increasing emphasis on calcifi- cation (thicker shells, stronger sculpture, etc.) among tropical as compared to cold- water species has been recognized both in gastropods (Graus, 1974; Vermeij, in press) and bivalves (Nicol, 1967). Besides the in- crease in the aragonite:calcite ratio in some bivalve shells toward the tropics (Lowen- stam, 1954; Waller, 1972), no other lati- tudinal variations in shell architecture of bivalves have been described to our knowl- edge. In this paper, we point out that bivalves with persistent posterior gapes are primarily polar and temperate in distribution, and suggest that the more complete shell pro- tection provided by tight valve closure in most tropical bivalves is related to an equa- torward increase in predation intensity. METHODS The latitudinal gradient in shell gaping may be documented in 2 ways. First, we (57) can estimate the number of species in each faunal province possessing a permanent posterior gape, and then compare this number with the total number of bivalve species in that province. Such estimates may be obtained by consulting major fau- nistic works; in our study, we have used Abbott (1974) for temperate North Amer- ica, Keen (1971) for tropical West America, and Abbott (1974) and Humfrey (1975) for the tropical western Atlantic. We have excluded from our ana.ysis all hard-bottom, nestling, and rock-boring bivalves, and have restricted our estimates to shallow-water soft-bottom clams living no deeper than 100 fathoms (278 m). Even within genera and families, we have rejected species which bore into rocks while retaining those which are found in soft sediments. Provincial boundaries used in this study conform to those proposed by Briggs (1974). The Arctic region is defined to extend south to the Gulf of St. Lawrence on the Atlantic coast, and (as the Aleutian Province) to Puget Sound on the Pacific coast. The Oregonian Province extends from Puget Sound to Point Conception, California; it is succeeded in a southward direction by the Californian Province, which extends to Magdalena Bay, on the outer coast of Baja California. The Panamic Province includes the outer coast of Baja California south of Magdalena Bay, and the mainland coast from the mouth of the Gulf of California to northernmost Peru. We have excluded from our analysis all species which are endemic to the offshore islands (Galapagos, Clipperton, Cocos, Revilla- gigedo), or to the Gulf of California. 58 VERMEIJ AND VEIL In the Atlantic, the Boreal Province extends from the Gulf of St. Lawrence and Newfoundland to Cape Cod, Massachusetts; the Virginian Province succeeds this prov- ince southward to Cape Hatteras, North Carolina. The Carolinian Province is here regarded as extending from Cape Hatteras to the area of Palm Beach, Florida. Al- though the northern Gulf of Mexico also belongs to the Carolinian Province (see e.g. Hedgpeth, 1953), we have not included bivalves from that area in our analysis. The West Indian Province comprises the east coast of Florida south of Palm Beach, as well as the Bahamas, West Indies, and the mainland coast of America from eastern Yucatan south. The Brazilian region faunis- tically belongs to the West Indian Province (see e.g. Vermeij & Porter, 1971), but is not included here because of insufficient faunistic data. A second way of describing the lati- tudinal pattern in posterior gape is to esti- mate the relative number of species with persistent posterior gapes at variously lati- tudinally separated localities. Species lists from various regional studies have been supplemented with collections made by the senior author from various parts of the world. RESULTS The genera in North American waters found to have persistent posterior gapes and included in our study are listed in Table 1. As Stanley (1970) and Runnegar (1974) have already pointed out, most gapers are deep-burrowing forms whose rate of burial varies from rapid (Ensis) to very slow (Mya, Cyrtopleura). Some species, however, are shallow burrowers; thus, Labiosa lineata (Say) is common on certain Thalassia flats in west Florida, where it is shallowly buried in muddy sand (Vermeij, personal observations). The cardiid Lophocardium is probably also a shallow burrower. Table 2 presents the proportion of species with posterior gapes in each of the generally recognized biogeographical prov- inces on the Pacific and Atlantic coasts of North America, as estimated from the tax- onomic works cited above. In general, trop- ical faunas have a much lower proportion of species with posteriorly gaping valves than do temperate faunas, but the steep- TABLE 1. North American genera of persistently gaping, infaunal clams. Solemyidae: Solemya Nuculanidae: Nuculana (only N. pernula Müller) Malletiidae: Tindaria Cardiidae: Papyridea, Lophocardium Veneridae: Saxidomus Semelidae: Cumingia Sanguinolariidae: Sanguinolaria (not $. nuttalli Conrad), Tage/us, Gari Solenidae: Si/iqua, Solen, Ensis, Solecurtus Mactridae: Mactra, Spisula (only $. dolabriformis Conrad), Labiosa, Tresus, Mactrellona (only М. clisia Dall) Hiatellidae: Cyrtodaria, Panomya, Panopea Myidae: Mya, Sphenia, Paramya, Cryptomya Pholadidae: Cyrtop/eura Lyonsiidae: Lyonsia (not L. arenosa Möller) | Ento- desma Thraciidae: Cyathodonta, Asthenothaerus Poromyidae: Poromya TABLE 2. Incidence of persistent posterior gaping in shallow-water infaunal bivalves from North America (data compiled from Keen, 1971; Abbott, 1974; Humfrey, 1975). Number of species Incidence Total of gaping Province Gapers infaunal % Eastern Pacific Aleutian 23 79 27 Oregonian 47 160 29 Californian 32 114 29 Panamic 58 448 13 Western Atlantic Arctic 11 36 31 Boreal 16 57 28 Virginian 17 70 24 Carolinian 20 137 15 West Indian 23 193 12 ness of the gradient is quite different on the 2 coasts. In the eastern Pacific, there is no significant difference between the Aleu- tian, Oregonian, and Californian Provinces in the incidence of gaping; but all 3 prov- inces show a sharply higher incidence of gaping than does the tropical Panamic Prov- ince (p < 0.005, chi-square tests). Differ- ences between adjacent provinces along the Atlantic coast are not significant (р > 0.05, chi-square tests), but there is an unmis- takable and rather uniform decline in the proportion of gaping species from north to south. To achieve the 0.05 level of signifi- cance between provinces, one must move at least 2 and sometimes 3| provinces depend- ing on the location within the gradient. Even within particular families, there is a trend for gaping species to be temperate BIVALVE GAPING AND LATITUDE 59 ог polar in distribution. Most mactrids with tightly closing valves (e.g. Rangia, most species of Spisula, Mulinia) are warm- temperate to tropical in distribution, while such typically gaping genera as Tresus and the European Lutraria are _ cold-water forms. The cold-water eastern Pacific Saxi- domus is one of the very few venerids with a persistent (albeit small) posterior gape. The gradient as described from faunistic studies is again evident when the clams from restricted localities are considered (see Table 3). The high percentage of gapers at Garrison Bay, an intertidal mud flat on San Juan Island in Washington State (48.5 N) (44%) contrasts sharply with the percent- ages in tropical Jamaica (9.1%) and the Seychelles (2.6%). DISCUSSION The observed equatorward decline in the incidence of gaping clams could be due to several causes, some trivial and others more interesting. Since gaping clams tend to be deep burrowers, it might be argued that tropical clams do not, on the whole, bur- row as deeply as temperate clams. This seems unlikely; not only do deposit-feeding tellinids often burrow deeply in the soft bottoms of both temperate and tropical regions, but the lucinids are a predomi- nantly tropical group characteristically found at great sediment depths (see Allen, 1958; Jackson, 1972, 1973). Both tellinids and lucinids are normally completely en- closed in their shells, and do not possess persistent gapes. Thus, the latitudinal gradi- ent in gaping is not an artifact of tax- опоту, nor does it reflect a restriction of tropical bivalves to shallow sediment depths. We must, therefore, conclude that the latitudinal gradient in gaping is a biological response to an external factor which is correlated with temperature. The close cor- relation with mean maximum temperature is evident from the difference in the steep- ness of the gradient on the two coasts of North America. In the eastern Pacific, water temperatures are cool and fluctuate little, and the north-south increase in tem- perature is comparatively small from Alaska to northern Baja California. On the east coast of North America, temperatures fluc- tuate wildly from winter to summer, and there is a steady rise in summer but not in winter temperatures from Canada to Flor- ida. How might temperature gradients influ- ence the incidence of gaping? Following the arguments of Bakus (1969, 1974), Bakus & Green (1974), and Vermeij (1977, in press) we believe that the increased intensity of various types of predation from high to low latitudes can account for the relative decline in species with gaping valves. It is becoming well! established that the intensity of crushing predation (by fishes, crabs, stomatopods, lobsters, etc.) increases toward warmer latitudes (for data see Menzel & Hopkins, 1955; Vermeij, 1977a). Data on other types of predation are scantier, but the diversity of predators which drill, swallow, or pry open their bivalve prey increases sharply toward the tropics. Among the important predators on bivalves, only the asteroids and birds are more prominent at high latitudes than in the tropics (Vermeij, in preparation). Several types of predators apparently take less time and require less specialized equipment to prey on gaping clams than on more tightly closing bivalves. Naticid gas- tropods do not normally drill gaping spe- TABLE 3. Incidence (1) of bivalve species with a persistent posterior gape (С) from total considered bivalve species (T) in selected localities. SPP gi Locality Lat. G T 1% Source Garrison Bay, Washington (intertidal) 48.5" М 4 9 44 GJV San Juan Islands, Washington (subtidal) 48.5° М 5 20 25 GJV Mugu Lagoon, California 34° N 3 15 20 Peterson, 1975 Alligator Harbor, Florida 30° N 3 18 17 GJV Sanibel Island, Florida 27° N 3 22 14 GJV Jamaica (Thalassia beds) 17° м 4 44 9 Jackson, 1973 Venado Beach, Panama Canal Zone TE 2 21 10 GJV Singapore US 4 27 15 Vohra, 1971 Mahé, Seychelles 37 2 76 3 Taylor, 1968 60 VERMEIJ AND VEIL cies of Таде/и$, but rather can attack them through the open anterior or posterior ends of the shell (Stump, 1975); they must resort to drilling in order to prey on vener- ids and other tightly closing clams. Gaping clams are digested more rapidly by the sand-dwelling asteroid Astropecten than are many tightly closing corbulids and venerids (Christensen, 1970; Massé, 1975). Extra- orally digesting sea stars may not need to pry open clams with a persistent gape, since the digestive juices from their stom- achs can be introduced into the victim by way of the siphon. Gastropods of the genus Busycon require a strong outer lip in order to pry open clams with tight valve closure, but could do the job with a much thinner lip when preying on Ensis and other gaping bivalves (Carriker, 1951; Paine, 1962, 1963). The North Atlantic snail Buccinum undatum Linn. must chip the valves of tightly closing clams, but need not employ this tactic in attacking gapers (Nielsen, 1975). Siphon-cropping fishes can readily snip off siphons from posteriorly gaping clams, but not so readily from tightly closing forms (Nesis, 1965; Edwards & Steele, 1968). Tropical acanthurid, scarid, and other fishes graze, scrape, and browse exposed rocky surfaces and grass flats ex- tensively (see e.g. Randall, 1967), and could have a disastrous impact on perma- nently gaping bivalves even if these were eaten or attacked only incidentally. Carriker & Van Zandt (1972) have pointed out that tight valve closure pre- vents metabolic products from escaping in the presence of predators, and could thus aid in preventing detection. This is obvi- ously not possible in bivalves with perma- nent gapes. The relative decline in gaping clams to- ward the tropics thus appears to go hand in hand with increasing protection from pred- ators, and is repeated on a microgeographi- cal scale from deep| to shallow sediment layers. Relatively few predators can attack deep-burrowing bivalves; some naticid gas- tropods, temperate asteroids, and warm- water rays can exploit deeply buried prey. Long-billed birds such as the curlew (Nu- menius) are generally incapable of taking deeply buried clams (see Burton, 1974). From the point of view of predation, then, moving down into the sediment is akin to heading away from the equator. LITERATURE CITED ABBOTT, R. T., 1974, American seashells, ed. 2, Van Nostrand Reinhold, New York, 663 p. ALLEN, J. A., 1958, On the basic form and adaptations to habitat in the Lucinacea (Lamellibranchia). Philosophical Transactions of the Royal Society of London, ser. B, 241: 421-484. BAKUS, G. J., 1969, Energetics and feeding in shallow marine waters. /nternational Review of General and Experimental Zoology, 4: 275-369. BAKUS, G. J., 1974. Toxicity in holothurians: a geographical pattern. Biotropica, 6: 229-236. BAKUS, С. J. & GREEN, С., 1974, Toxicity т sponges and holothurians. Science, 185: 951-953. BRIGGS, J. C., 1974, Marine zoogeography. McGraw Hill, New York, 475 р. BURTON, P. J. K., 1974, Feeding and feeding apparatus in waders: a study of anatomy and adaptations in the Charadrii. British Museum of Natural History, London, 150 p. CARRIKER, M. R., 1951, Observations on the penetration of tightly closing bivalves by Busycon and other predators. Ecology, 32: 73-83. CARRIKER, M. R. & VAN ZANDT, D., 1972, Predatory behavior of a shell-boring muricid gastropod. In H. E. WINN & B. L. OLLA, eds., Behavior of marine animals, I: inverte- brates. Plenum Press, New York, p. 157-244. CHRISTENSEN, A. M., 1970, Feeding biology of the seastar Astropectin irregularis Pennant. Ophelia, 8: 1-134. EDWARDS;cR> Ri Gira STEELE ЯН: 1968: The ecology of O-group plaice and common dabs at Loch Ewe. |. Population and food. Journal of Experimental Marine Biology and Ecology, 2: 215-238. FISCHER, A. G., 1960, Latitudinal variations in organic diversity. Evolution, 14: 64-81. GRAUS, R. R., 1974, Latitudinal trends in the shell characteristics of marine gastropods. Lethaia, 7: 303-314. HEDGPETH, J. W., 1953, An introduction to the zoogeography of the northwestern Gulf of Mexico with reference to the invertebrate fauna. Publications of the Institute of Marine Science, 3: 107-224. HUMFREY, M., 1975, Sea shells of the West Indies. Collins, London, 351 р. JACKSON, J. B. C., 1972, The ecology of mol- luscs of Thalassia communities, Jamaica, West Indies. Il. Molluscan population variability along an environmental stress gradient. Marine Biology, 14: 304-337. JACKSON, J. B. C., 1973, The ecology of mol- luscs of Thalassia communities, Jamaica, West Indies. |. Distribution, environmental physiol- ogy, and ecology of common shallow-water species. Bulletin of Marine Science, 23: 311-350. KEEN, А. M., 1971, Seashells of tropical West America, ed. 2. Stanford University Press, Palo Alto, Calif. 1064 p. LOWENSTAM, H. A., 1954, Factors affecting the aragonite:calcite ratios in carbonate-secreting marine organisms. Journal of Geology, 62: 285-322. BIVALVE GAPING AND LATITUDE 61 MASSE, H., 1975, Etude de l'alimentation de Astropecten aurantiacus Linné Cahiers de Biologie Marine, 16: 495-510. MENZEL, R. W. & HOPKINS, S. H., 1955, Crabs as predators of oysters in Louisiana. Proceed- ings of the National Shellfisheries Association, 46: 177-184. NESIS, K. N., 1965, Ecology of Cyrtodaria siliqua and history of the genus Cyrtodaria (Bivalvia: Hiatellidae). Malacologia, 3: 197-210. NICOL, D., 1967, Some characteristics of cold- water marine pelecypods. Journal of Paleon- tology, 41: 1330-1340. NIELSEN, A., 1975, Observations on Buccinum undatum L. attacking bivalves and on prey responses. Ophelia, 13: 87-108. PAINE, R. T., 1962, Ecological diversification in sympatric gastropods of the genus Busycon. Evolution, 16: 515-523. PAINE, R. T., 1963, Trophic relationships of eight. sympatric predatory gastropods. Есо/- ogy, 44: 63-73. PETERSON, C. H., 1975, Stability of species and of community of the benthos of two lagoons. Ecology, 56: 958-965. RANDALL, J. E., 1967, Food habits of reef fishes of the West Indies. Studies in Tropical Oceanography, Institute of Marine Science, University of Miami, 5: 665-847. RUNNEGAR, B., 1974, Evolutionary history of the bivalve subclass Anomalodesmata. Journal of Paleontology, 48: 904-940. STANLEY, S. M., 1970, Relation of shell form to life habits of the Bivalvia (Mollusca). Geo- logical Society of America Memoir 125: 1-296. STUMP, T. E., 1975, Pleistocene molluscan paleo- ecology and community structure of the Puerto Libertad region, Sonora, Mexico. Ра/ео- geography, Paleoclimatology, Paleoecology, ПУ: 177-226. TAYLOR, J. D., 1968, Coral reef and associated invertebrate Communities (mostly molluscan) around Mahé, Seychelles. Philosophical Trans- actions of the Royal Society of London, ser. B, 254: 130-206. VERMEIL, G. J., 1977, Patterns in crab claw size: the geography of crushing. Systematic Zoology, 26: 138-151. VERMEIJ, С. J., in press, The architecture of some gastropod shells. [In untitled book.] VERMEIL, С. J. & PORTER, J. W., 1971, Some characteristics of the dominant intertidal mol- luscs from rocky shores in Pernambuco, Brazil. Bulletin of Marine Science, 21: 440-454. VOHRA, F. C., 1971, Zonation on a tropical sandy shore. Journal of Animal Ecology, 40: 679-708. WALLER, Т. R., 1972, The functional signifi- cance of some shell microstructures in the Pectinacea (Mollusca: Bivalvia). /nternational Geological Congress, 24th Session, Montreal, Canada, Section 7, Paleontology, p. 48-56. nn taper) 4 Ty ae ho2 landet One (“a | 2 à ; Pup =. <, PR e be ВЯ ae > ae? 1 e 4 у > ~ 3 Invbives i Core (Wer, 1051; Prisa? o A ras né RSA MENT core У’ edt Herts linge nar tax Manan SE ES EPS are (me у N N lie) age) == ТУЧ ever Dis y N » aj b > aa Www u: we 4 2 , Cure + pi RUSTENNEN, AM, 1970, т cms, 1590 oy + ции wre ne aca м Y ST. © a tr Фо > Y > me O дне. > $ le ti ® 134, Ye er cy arte р! gu A 6% " | $" 1. Ce ÊTRE 7; ta f à con A ase er | | Lech Eb. МИЯ "us we à Der ' LA mn CNE Esser (Ht бы neniiv wet >, req mess e vows, 1 Na Au où er te ЕР АС (avt Ss м # 24 pal wre A Wl , # 4 tri ПИ > LL pl O) si niet 14 LATE 6 y ed ti ite it pa } EM . Le y AN = « are Г vip 1 „a Я ho у $ a u « = > | an á SO x a =. lee iis stu, а ~! ps Te Ate HA lia ' Е Wr us 7 $ tL 4 a h mie PL ATI. Ai y ci Ln de WES > ı@ : . < ( у ' 1 «А E «УГ E UN E po Dev ni & та paper я f om 19 9) 51 “ MITA, TN м : nai г + 8 ca ath >. terior LITER у р rein d : MEL FA q В à Fiz | в a 4 р WEHSTAM, И. A, Ума ang . ыы nanmnmiroselé Milka E = 4 ye ыы net ga wre er GN ca dre Ta MALACOLOGIA, 1978, 17(1): 63-72 THE HABITAT AND FEEDING BEHAVIOR OF THE WENTLETRAP EPITONIUM GREENLANDICUM Frank Реггоп1 Marine Biological Laboratory, Woods Hole, Massachusetts 02543, U.S. A. ABSTRACT Epitoniids (wentletraps) are mesogastropod prosobranchs which forage for and/or para- sitize a variety of benthic coelenterates. Epitonium greenlandicum (Perry) is circumboreal in distribution and occurs subtidally in the northwestern Atlantic. Laboratory and field observations indicate that this wentletrap feeds infrequently, and only on anemones. After feeding, Е. greenlandicum burrows into soft mud and can remain inactive for as long as 80 days. Wentletraps are able, at least over short distances, to locate anemones by chemotaxis. Although Е. greenlandicum can feed on at least 6 species of anemone, it shows а preference for Metridium senile and tends to parasitize this large anemone under laboratory conditions. In the shallow waters of the Bay of Fundy, М. senile occurs on exposed rocks and ledges and is not available to the primarily infaunal wentletraps. Consequently, Е. greenlandicum must forage for small anemones on a mixed mud-cobble bottom. It is suggested that epitoniids are able to occupy a purely ectoparasitic niche only in areas where there is both a stable supply of the host coelenterate and a nearby refuge from visual predators. INTRODUCTION The various groups of prosobranch gas- tropods which have evolved methods of feeding on coelenterate prey are listed by Robertson (1966, 1970) and by Perron & Turner (in press). Among these gastropods, wentletraps (family Epitoniidae) are cosmo- politan in distribution and range from the intertidal zone to abyssal depths. Detailed anatomical studies of epitoniids have been published by Taki (1956, 1957) and by Hochberg (1971), but surprisingly little in- formation is available on wentletrap natural history and ecology. The first significant contribution to our knowledge of feeding in epitoniid gastro- pods was Thorson’s (1957) suggestion that the entire family possibly is parasitic on sea anemones. Thorson based his statement on the behavior of the wentletrap Opalia creni- marginata (Dall), which lives in close asso- ciation with, and feeds on, the anemone Anthopleura xanthogrammica (Brand). Robertson (1963), in a review of the litera- ture on epitoniid feeding, noted that few wentletraps are found in association with coelenterates, and so modified Thorson’s statement by suggesting that wentletraps may grade from permanent ectoparasites to foraging predators on coelenterates. The available literature on wentletrap natural history consists of fragmentary ac- counts and brief reports on a variety of species. The present study, however, is a more detailed examination of the habitat and feeding behavior of a single species, E. greenlandicum (Perry). The data presented here contribute to a broader understanding of the ecology of wentletraps, an interest- ing group of gastropods. METHODS Field data on Е. greenlandicum were obtained using SCUBA techniques in 10-25 m of water at Eastport, Maine (44°54'N, 66 59'W). A total of approximately 12 hours of daytime underwater observations was distributed throughout the year (every other month from May, 1973—May, 1974 with a single observation in May, 1976). One nighttime dive was made in May, 1974. During these dives, notes were taken on the behavior of Е. greenlandicum and 1Present address: Department of Zoology, Edmondson Hall, University of Hawaii, Honolulu, Hawaii 96822, U.S.A. (63) 64 PERRON on the distributions and densities of several species of anemone. Anemone densities were determined from bottom transects and analysis of underwater photographs. Some specimens of Е. greendlanicum were fixed in 10% formaldehyde solution im- mediately after collection and saved for subsequent gut content analysis. Live wentletraps and anemones were collected and maintained 1st in a recirculating sea water system at the University of New Hampshire, and later in a flowing system at the Marine Biological Laboratories at Woods Hole, Massachusetts. Ten specimens of Е. greenlandicum were kept in a sea table arranged to approx- imate their natural habitat. Mud and rocks from the Eastport study site were placed in the sea table along with several specimens of the anemone Metridium senile Linn. Each of the 10 wentletrap shells was num- bered and daily records were kept of feed- ing frequencies, duration of feeding epi- sodes, and general feeding behavior. The feeding preferences of E. greenland- сит were experimentally determined by prey presentation trials similar to those described by Edmunds et al. (1974) for the nudibranch Aeolidia papillosa (Linn.). In these trials, wentletraps were placed 1 cm away from and facing single anemones. If the wentletrap everted its proboscis and attached its jaws to the anemone within 15 min, a positive feeding response was scored. А negative response was scored if no feed- ing took place within this time. The anem- ones presented to Е. greenlandicum were M. senile, Gonactinia prolifera (Sars), Stom- phia coccinea Mueller and Tealia -felina Linn. Ten anemones of each species were presented, 1 at a time, to each of 6 wentletraps. During these experiments, responses of anemones to predation were noted. The hypothesis that ЕЁ. greenlandicum may .locate prey through chemoreception was tested using 2 types of olfactometers. First, specimens of £. greenlandicum were tested using a 4-chambered olfactometer designed by Wood (1968). In this device, water from peripheral prey cells flows down ramps to a predator chamber, and exits through a central drain pipe. A $ес- ond set of tests was run using a simple plexiglas Y-maze 30 cm in length. Water flow rates were adjusted to 250 ml/min in both the 4-chambered olfactometer and the Y-maze. Wentletraps were run in groups of 8 in both sets of experiments. At the end of 8 hr, positive responses were scored for all wentletraps which had entered the stimulus arm of the Y-maze or the stimulus chamber of the 4-chambered olfactometer. Individual snails were removed from the olfactometer as they prepared to feed on the prey anemone. Stimulus and control chambers were randomly designated at the start of each trial. Only specimens of М. senile were used as prey. Preliminary Y-maze trials were run with- out prey to determine whether Е. green- landicum schools or trails. Such behavior patterns would bias the results of olfacto- meter trials in which groups of predators were run simultaneously. RESULTS Field observations The substrate at the Eastport, Maine study site consisted of exposed rocks and ledges surrounded by cobble encrusted with coralline algae in a fine mud matrix. Water temperatures ranged from 2.1°C in January to 12.5°C in August. Tidal currents were strong. Metridium senile was the most obvious anemone at Eastport. On rocky ledges, these suspension-feeding anemones often exceeded 20 cm in height, and formed aggregations of up to 160 individuals/m?. Metridium senile was rare (less than 0.1/m?) on the less exposed cobble bot- tom. In contrast to M. senile, the tiny (up to 6 mm in height) Gonactinia prolifera was common in the cobble-mud areas, but rare on exposed rock surfaces. G. prolifera pre- fers the undersides of rocks and dead mol- lusk shells where some space remains for water circulation. The distribution of this predatory anemone was extremely patchy and related to the availability of suitable habitats. Abundances ranged from O to 50 individuals/m?. Tealia felina and Stromphia coccinea were found with M. senile on the rock ledges, but also occurred on the cobble. Neither of these anemones forms aggrega- tions, and their densities were roughly 0.2/m?. Thirty specimens of Е. greenlandicum were collected during the study, and an WENTLETRAP FEEDING 65 average of 5 snails could be located per hr of intense searching. There were no appar- ent seasonal variations in the abundance of Е. greenlandicum. Twenty-eight specimens were found crawling on or among the cobbles and 2 were in the mud at the base of 1 large M. senile. Large anemones (М. senile and T. felina) were always carefully checked for associated wentletraps, but none were ever found with them on ex- posed rocks or ledges either during daylight or at night. Three specimens of EF. greenlandicum were observed feeding in the field. In each case, the anemone being attacked was a specimen of G. prolifera and the wentle- trap swallowed the small anemone whole. Wentletraps collected in the field gener- ally had empty guts. Of 12 animals col- lected throughout the year and examined in the laboratory, only 1 (collected May, 1973) had ingested material in its gut. This material contained nematocysts and was presumed to be anemone tissue. Laboratory observations and experiments Locomotion The locomotion of E. greenlandicum consists of 2 alternating movements. In the first, the shell remains stationary while the foot crawls forward and obtains a purchase in the substrate. In the second movement, the foot remains stationary while the co- lumellar muscle contracts and pulls the shell forward. Miller (1974) showed that this “discontinuous’’ form of locomotion, common to many burrowing gastropods such as the Turritellidae and the Terebri- dae, is related to slow locomotory rates and to a reduced ability of the foot to adhere to hard surfaces. Using Miller’s tech- niques, | measured the maximum adhesion (tenacity) of the foot of E. greenlandicum at less than 10 g/cm? foot area. In con- trast, Miller (1974) measured tenacities of up to 2792 g/cm? foot area for epifaunal gastropods. Prey preferences and anemone defensive reactions In the laboratory, Е. greenlandicum fed only on anemones [M. senile, С. prolifera, S. coccinea, Bunodactis stella (Verrill), Actinauge verrilli McMurrich and Diadu- mene leucolena (Verrill)], even though soft-bodied polychaetes, holothurians and tunicates were frequently available to them. The anemones В. stella, A. verrilli and D. leucolena were not found at the Eastport study site and therefore were not included in the prey preference experiments. When feeding on large anemones, wentletraps always attached their jaws to the column or base of the anemone (Fig. 1). Anemone tentacles were never bitten off and swallowed individually as has been reported for Е. tinctum (Carpenter) by Hochberg (1971). Small anemones (less than 1 cm in height) were simply grasped and swallowed whole. During feeding, anemone tissues could be seen slowly mov- ing through the extended proboscis. Although Е. greenlandicum is capable of feeding on several species of anemone, repeated laboratory observations suggested a strong preference for М. senile. This suspected preference was confirmed by the results of prey presentation trials involving 4 species of anemone (Table 1). Out of 240 such trials, only 77 positive feeding responses were scored. Metridium senile ac- counted for 75% of the positive responses, G. prolifera was 2nd in preference, and both 7. felina and S. coccinea were essen- tially ignored. Metridium senile responded | to wentle- trap attacks by (1) bending away from the predator, (2) extruding nematocyst-laden acontia, and (3) moving slowly away at approximately 1 cm/hr. Epitonium green- landicum was never deterred by these defensive reactions. Crawlinglat a mean rate of 45 cm/hr (N=5), E. greenlandicum readily overtook its prey, and once its jaws were attached to the anemone the wentle- trap was not easily dislodged. While the head and foot of Е. greenland- сит recoil violently from the sting of nematocysts, the outer surface of the proboscis appears less sensitive. When М. senile extruded acontia toward a feeding wentletrap, the operculum was withdrawn and the snail continued feeding with only the proboscis exposed to nematocysts. The defensive reactions of G. prolifera consisted of (1) using its tentacles to sting the wentletrap on the head or foot, (2) “walking” away, or (3) swimming. Stinging was an effective defense only when the anemone’s tentacles came into direct con- tact with the head or foot of the snail. Such stings caused wentletraps to recoil violently and lie inert for several min. During this period, G. prolifera would sometimes walk away from the wentletrap 66 PERRON FIG. 1. Two specimens of Epitonium greenlandicum One wentletrap has already attached its jaws to the proboscis partially everted. TABLE 1. The number of positive feeding re- sponses scored for each speimen of Epitonium greenlandicum (A-F) when presented with different species of anemone. Wentletrap Anemone Ab Gop) ae eo tal Metridium senile 10 10 8 10 9 10 57 Gonactinia prolifera 281020 54614 19 Stomphia coccinea ОО, 1000 1 Tealia felina OF 0.030700 0 attacking a large (height 8 cm) Metridium senile. anemone, while the other is approaching with its in the manner described by Robson (1966, 1971). In only 2 out of 60 trials did swimming occur in response to contact with the wentletrap’s proboscis. One anem- one swam 5 cm from the wentletrap and successfully avoided predation, while the other was eaten in spite of its attempts at escape. Tealia felina was never approached by wentletraps and showed no defensive reac- WENTLETRAP FEEDING 67 tions to the presence of Е. greenlandicum. The single S. coccinea which was attacked retracted its tentacles but did not swim. Feeding frequency and duration The 10 marked specimens of Е. green- landicum maintained in the laboratory sea table fed a combined total of 25 times during 6 months. Four specimens died after feeding only once. Table 2 gives the inter- vals between, and durations of, feeding episodes for the remaining specimens. Since the M. senile anemones in the sea table were all large, they could not be swallowed whole and usually survived several wentle- trap attacks. Each feeding episode lasted 4-10 hr (Table 2), after which the wentletrap bur- rowed in the mud at the base of the anemone on which it had just fed. Emerg- ing from the mud 12-80 days later, the wentletrap usually attacked the same anem- one. Location of prey through chemotaxis In the laboratory, wentletraps showed a TABLE 2. Durations of, and intervals between, laboratory feeding episodes of Epitonium green- landicum. Wentletrap Duration (hr) Interval (days) A 7,9 18 B 4,7 80 С 7,5,5,10,5,4 59,13,12,22,12 D 8,10,5 25,45 E 6,8,6,5 13,8,24 F 6,5,6,9 20,24 ,21 TABLE 3. Results of Y-maze prey and control trials, showing the numbers of Epitonium green- landicum snails moving into the stimulus arm, the non-stimulus arm, and the number not responding during the 8-hour trials. Stimulus Non-stimulus Not re- N* arm arm sponding Prey trials 88 68 7 13 Control trials 88 137 16 59 Wentletraps were run in groups of 8 in a total of 11 prey trıals and 11 control trials. **Since no prey was used in the control trials, both arms are non-stimulus arms. tendency to gather about injured anem- ones, and alternately everted and retracted their proboscises when approaching speci- mens of M. senile (Fig. 1). Eversion of the proboscis was never observed in the ab- sence of anemones. Such observations sug- gest that, at least in the laboratory, Е. greenlandicum may be able to locate prey through chemosensory means, and the re- sults of Y-maze olfactometer tests (Table 3) support this hypothesis. All responding animals showed a strong tendency to move toward the prey, and periodically everted their proboscises throughout the trials. In the control experiments, where no prey was placed in either arm of the Y-maze, nearly equal numbers of wentletraps moved to the right and left arms of the maze (Table 3). Wentletraps did not move in “schools,” nor did|they appear to follow each others’ mucous trails as has been reported for Littorina irrorata by Hall (1973). Seventy-three percent of the wentletraps entering the prey arm of the Y-maze at- tempted to feed during the 8-hr trials. These animals required 2.5-8 hr (x= 4.9 hr) to locate the prey in the 30 cm long Y-maze. Experiments using a 4-chambered olfac- tometer produced data which, when sub- jected to a x”-test, was random. In these trials, wentletraps tended to move about the central chamber everting their probos- cises, but less than 5% of the animals attempted to climb the ramps leading to either prey or control chambers. DISCUSSION Robertson (1963) stated that epitoniids grade from more or less permanent ecto- parasites, through temporary ectoparasites, to foraging predators, and he predicted that foraging species would feed infrequently and show high flexibility in their choice of prey. The laboratory and field observations in the present study generally support these predictions, and also elucidate some of the ecological factors which may determine feeding strategies within the family Epitoni- idae. Function of chemoreception In 1963, Robertson commented that the wentletrap £. rupicola (Kurtz) seemingly foraged at random and appeared unable to 68 PERRON locate anemones by chemosensory means. To my knowledge, the Y-maze data in the present study are the first reported evi- dence of chemotaxis in any epitoniid. How- ever, simple Y-mazes are rare in nature, and it should also be remembered that in these experiments wentletraps required a mean of 4.9 hr to find anemones only 30 cm away. Clearly, the results of these Y-maze tests do not prove that foraging wentletraps can locate distant anemones in their natural habitat. Wood (1968) designed the 4-chambered olfactometer to test rigorously the chemo- sensory abilities of the foraging neogastro- pod Urosalpinx cinerea (Say). The water currents in this olfactometer are swifter and more complex than would be encoun- tered in a Y-maze, and the inability of Е. greenlandicum to locate prey in this system may reflect the wentletrap’s actual perform- ance in the turbulent bottom currents at the Eastport study site. The most important site of chemorecep- tion in prosobranch gastropods is probably the osphradium (Kohn, 1961). In the car- nivorous Neogastropoda, a relatively long siphon directs water currents onto а Ы- pectinate osphradium, and many of these gastropods are known for their chemo- sensory prey-finding abilities (Kohn, 1961; Hathaway & Woodburn, 1961; Wood, 1968). Epitonium, a mesogastropod, has only a monopectinate osphradium, lacks an extended siphon, and therefore seems poorly adapted for locating distant prey through chemotaxis. Chemoreception may, however, serve im- portant functions for both ectoparasitic and foraging wentletraps. When a foraging E. greenlandicum senses a nearby anemone through chemoreception, the proboscis is everted and moved about, apparently at random. When the proboscis comes into contact with the anemone, random move- ment ceases and feeding begins. Therefore, although precise location of anemones may be through tactile stimulation of the everted proboscis, proboscis eversion itself is chemically stimulated. Everting the pro- boscis may also prevent the sensitive head and foot of the wentletrap from directly encountering the tentacles of a stinging anemone such as G. prolifera. Ectoparasitic wentletraps probably remain close enough to their prey so that only limited chemo- sensory abilities are necessary to prevent loss of contact with the host anemone. Dietary flexibility No epitoniid has ever been found to feed on other than coelenterate prey. Couthouy’s (1838) observation that Sca- laria subulata Couthouy [= Epitonium green- landicum] feeds on beef is almost certainly in error since he saw no proboscis everted during “feeding.” Fretter & Graham (1962) suggested that wentletraps may have an alternate method of feeding on an active animal such as an annelid or nemertine. Robertson (1963) considered this possibil- ity unlikely, and | must concur in light of the fact that specimens of Е. greenland- icum consistently rejected all non-coelenter- ate prey offered to them during the present study. Although several species of wentletraps are known to associate with, and feed on, anemones (Thorson, 1957; Robertson, 1963, 1970), little is known about the specificity of these associations. Epitonium greenlandicum can feed on at least 6 species of anemone, while the ectoparasites Opalia crenimarginata (Dall), E. tinctum, E. ши Pilsbry and Е. albidum (Orb.) have as yet been reported only on 1 or 2 hosts. Emlen (1968) stated that predators may be expected to specialize when food is abundant, and to feed more indiscrimi- nately as food becomes scarce (see also Menge, 1972). One extension of this pre- diction is that foragers should show lower prey specificity than do parasites. In terms of relative prey specificity within the Epi- toniidae, Е. greenlandicum seems to show the characteristics of a foraging generalist. However, since only fragmentary data exist on the diets of most epitoniids, other wentletraps may be able to feed on more species of coelenterates than is presently known. Wentletrap predation and anemone defenses The behavioral aspects of the relation- ship between Е. greenlandicum and its prey compare interestingly with published ac- counts of nudibranch-anemone associations. Although the nudibranch Aeolidia papillosa feeds on both 7. felina and M. senile in the northwestern Atlantic (Harris, personal communication), Waters (1973) and Edmunds et al. (1974) have demon- strated that A. papillosa shows а distinct preference for 7. felina over М. senile. In contrast to Е. greenlandicum, A. papillosa WENTLETRAP FEEDING 69 is often effectively repelled by the acontia of M. senile, and Harris (1973) reports that this nudibranch may be fatally stung when enveloped т M. senile acontia. In the laboratory, Е. greenlandicum strongly preferred М. senile over any of the other anemones tested. Epitonium green- landicum is protected from acontia by its shell and operculum, and Clench & Turner (1952) report that the esophagus of epi- toniids is lined with cuticle. However, the outer surface of the proboscis does not appear to be cuticularized, and the nature of its seeming immunity to nematocysts remains to be studied. Robson (1971) reported that the anem- one G. prolifera swims in response to chemicals secreted by A. papillosa, Cory- phella rufibranchialis (Johnston) and two other unidentified nudibranchs. Although Robson found no direct evidence that these nudibranchs feed on G. prolifera, the swim- ming response is probably an escape reac- tion to a predatory nudibranch. Conversely, E. greenlandicum rarely elicits a swimming response from С. prolifera even when physically disturbing the anemone. Stomphia coccinea swims to escape pre- dation by A. papillosa and certain starfish predators (Yentsch & Pierce, 1955; Robson, 1961). In the present study, how- ever, swimming was not observed in S. coccinea, and the results of prey presenta- tion experiments suggest that Е. greenland- icum rarely, if ever, feeds on this anemone in nature. Finally, 7. felina, although consistently ignored by Е. greenlandicum in the present study, ranked 4th out of 11 anemones in the prey preference hierarchy of A. papil- losa, and showed distinct defensive behav- ior in response to this nudibranch (Edmunds et al., 1974, 1976). These descriptions of anemone defensive reactions and the more extensive discussion of the subject in Edmunds et al. (1976) suggest that most of these behavior pat- terns are adaptations primarily directed against predatory nudibranchs rather than wentletraps. Nudibranchs are relatively fast [500 cm/hr for A. papillosa (Edmunds et al., 1976)] voracious predators, and Harris (1973) reports that predation by A. papillosa has a marked effect on the densities and age structures of М. senile populations. Wentle- trap attacks, however, rarely result in the deaths of any but the smallest anemones, and even when small anemones such as G. prolifera are involved, wentletraps move so slowly that they probably account for only a small fraction of anemone mortalities. Habitat and foraging behavior The laboratory data on Е. greenland- сит ргеу preferences suggest that this wentletrap is a facultative ectoparasite on the anemone M. senile. Although £. green- landicum may feed heavily on M. senile in some areas (perhaps in deep water), field observations at Eastport, Maine reveal that in shallow subtidal places, wentletraps probably spend most of their time foraging for small anemones (G. prolifera) in a habitat where M. senile is rare. The absence of wentletraps from the groups of М. senile anemones! on exposed rocky ledges is the result of a combination of factors. Epitonium greenlandicum 15 essentially an infaunal! animal and its type of locomotion is common to many burrow- ing gastropods. The low adhesion values obtained for the foot of Е. greenlandicumlin the present study show that this gastropod is poorly adapted for clinging to exposed rocks in strong currents. At a mean speed of 45 cm/hr, E. green- landicum requires several hr to cross a few meters. In order to prey on the ledge-dwel- ling anemones /at Eastport, wentletraps would have to cross wide expanses of bare rock on which they could be easily per- ceived by visual predators such as fish or crabs. Fish are known to feed on wentle- traps (Clapp, 1912; Wigley, 1956), and Homans €: Needler (1944) found “large quantities” of Е. greenlandicum in the guts of haddock caught off Nova Scotia. Also, since there is no mud on the rocky ledges, satiated wentletraps would not be able to conceal themselves by burrowing after each feeding episode. Thus restricted to the cobble-mud habi- tat, Е. greenlandicum must engage in a mixed strategy of foraging and temporary ectoparasitism. Wentletraps which encoun- ter large anemones on the proper substrate probably remain associated with them for some time. The 2 specimens of F. green- landicum found burrowed near 1 large М. senile are evidence that such associations exist in nature as well as in the laboratory. However, the sparse distribution of М. senile individuals on the cobble-mud bot- tom, the empty guts of nearly all the snails examined, and the fact that 28 of the 30 70 РЕВВОМ wentletraps observed т the field were not near large anemones, indicates that Е. greenlandicum is primarily a forager, at least at the Eastport study site. Epifaunally foraging wentletraps must remain exposed to visual predators during each foraging period, and although E. greenlandicum can subsist infaunally for up to 80 days after being satiated on a large anemone (Table 2), intervals between feed- ing episodes may be considerably shorter when only small anemones are available. At least in terms of predator avoidance, ecto- parasitsm would seem to be a more nearly optimal feeding strategy than is foraging. CONCLUSIONS It may now be possible to identify ecological factors which determine whether a given epitoniid will function as an ecto- parasite or as a forager. All known ecto- parasitic wentletraps occur where there is both a stable supply of the host coelenter- ate and a nearby refuge from visual preda- tors. On the west coast of the United States, the epitoniid Opalia crenimarginata feeds on the large anemone Anthopleura xanthogrammica and hides itself in the shell-gravel at the anemone’s base (Thorson, 1957). Epitonium tinctum is also reported to “live in close association with sea anem- ones in sand pockets and sand-filled crev- ices ...’” (Strong, 1941). In the Caribbean, Robertson (1963 and personal communica- tion) reports that Е. a/bidum burrows in sand pockets at the base of Stoichactis helianthus (Ellis). Off the Hawaiian Islands, E. ulu is a temporary parasite on the coral Fungia scutaria Lamarck. According to Bosch (1965), this wentletrap avoids direct light and feeds on the underside of the coral. Similar behavior is reported for Е. costulatum (Kiener) by Root (1958), as quoted in Robertson (1963). This Philip- pine wentletrap was always found on the underside of the coral Fungia or in the sand at the base of the coral. In the laboratory, specimens of 2. green- landicum behaved like temporary ectopara- sites, remaining associated with the same anemone for long periods. These wentle- traps had a relatively concentrated and stable supply of large anemones available to them and could easily burrow in the mud between feeding episodes. At the Eastport, Maine study site, however, groups of large anemones occurred only on hard substrates. Habitat restraints, especially the lack of hiding places near these anemones, pre- vented the wentletraps from being ecto- parasitic. The case of Е. greenlandicum suggests that epitoniids forage only when the neces- sary prerequisites for ectoparasitism [(1) a stable food supply, and (2) a convenient refuge from predators] do not exist. All wentletraps may, as Thorson (1957) origin- ally predicted, be facultative ectoparasites. ACKNOWLEDGEMENTS | thank Larry Harris and Ruth Turner for their support of this research. | also thank Brian Rivest and Alan Hulburt for diving assistance. Special thanks for techni- cal assistance are also due to Irene Thomp- son. LITERATURE CITED BOSCH, H. F., 1965, A gastropod parasite of solitary corals in Hawaii. Pacific Science, 19: 267-268. CLAPP, W. F., 1912, Collecting from haddock on the George’s Bank. Nauti/us, 25: 104-106. CLENCH, W. J. & TURNER, R. D., 1952, The genus Epitonium in the Western Atlantic; Part |. Johnsonia, 2: 249-288. COUTHOUY, J. P., 1838, Descriptions of new species of molluscs and shells, and remarks on several polypi found in Massachusetts Bay. Boston Journal of Natural History, 2: 53-111. EDMUNDS, M., POTTS, G. W., SWINFEN, R. C. & WATERS, V. L., 1974, The feeding prefer- ences of Aeolidia papillosa (L.) (Mollusca: Nudi- branchia). Journal of the Marine Biological Association of the United Kingdom, 54: 939-947. EDMUNDS, M., POTTS, С. W., SWINFEN, В. С. € WATERS, V. L., 1976, Defensive behaviour of sea anemones in response to predation by the opisthobranch mollusc Aeolidia papillosa (L.). Journal of the Marine Biological Associa- tion of the United Kingdom, 56: 65-83. EMLEN, J. M., 1968, Optimal choice in animals. American Naturalist, 102: 385-389. FRETTER, V. & GRAHAM, A., 1962, British Prosobranch Molluscs. Ray Society, London. 755 p. HALL, J. R., 1973, Intraspecific trail-following in the marsh periwinkle Littorina irrorata Say. Veliger, 16: 72-75. HARRIS, L. G., 1973, Nudibranch associations. Current Topics in Comparative Pathobiology, 2: 213-315. HATHAWAY, В. В. & WOODBURN, К. D., 1961, Studies on the crown conch Melongena corona Gmelin. Bulletin of Marine Sciences of the Gulf and Caribbean, 11: 45-65. HOCHBERG, F. G., 1971, Functional morphol- ogy and ultrastructure of the proboscis com- WENTLETRAP FEEDING 71 (Gastropoda: Society of plex of Epitonium tinctum Ptenoglossa). Echo (Western Malacologists), 4: 22-23. HOMANS, В. E: $. € NEEDLER, A. W. H., 1944, Food of the haddock. Proceedings of the Nova Scotia Institute of Science, 21(2): 15-49. KOHN, A. J., 1961, Chemoreception in gastropod molluscs. American Zoologist, 1: 291-308.- MENGE, B. A., 1972, Foraging strategy of a starfish in relation to actual prey availability and environmental predictability. Ecological Monographs, 42: 25-50. MILLER, S. L., 1974, [Adaptive design of loco- motion and foot form in prosobranch gastro- pods. Journal of Experimental Marine Biology and Ecology, 14: 99-156. ROBERTSON, R., 1963, Wentletraps (Epitoni- idae) feeding on sea anemones and corals. Proceedings of the Malacological Society of London, 35: 51-63. ROBERTSON, R., 1966, Coelenterate-associated prosobranch gastropods. American Malacologi- cal Union Annual Reports, 1965, 6-8. ROBERTSON, R., 1970, Review of the predators and parasites of stony corals, with special reference to symbiotic prosobranch gastro- pods. Pacific Science, 24: 43-54. ROBSON, E. A., 1961, Some observations on the swimming behaviour of the anemone Stom- phia coccinea. Journal of Experimental Biol- ogy, 38: 343-363. ROBSON, E. A., 1966, Swimming in Actiniaria. Symposia of the Zoological Society of Lon- don, 16: 333-359. RESUME ROBSON, E. A., 1971, The behaviour and neuro- muscular system of Gonactinia prolifera, a swimming sea anemone. Journal of Experi- mental Biology, 55: 611-640. ROOT, J., 1958, Rapa rapa in the Sulu Sea. Hawaiian Shell News, 7: 7-8. STRONG, A. M., 1941, Notes on Epitonium (Nitidoscala) tinctum (Carpenter). Nautilus, 55: 46-47. TAKI, Is., 1956, 1957, Anatomical study of Japanese Epitoniidae. Pts. 1 & 2. Bulletin of the National Science Museum [Tokyo], 3: 71-79; 176-132. THORSON, G., 1957, Parasitism in the marine gastropod family Scalidae. Videnskabelige Meddelelser fra Dansk naturhistorisk Forening; Kébenhavn 119: 55-58. WATERS, V. L., 1973, Food preferences of the nudibranch Aeolidia papillosa, and the effect of the defenses of the prey on predation. Veliger, 15: 174-192. WIGLEY, В. L., 1956, Food habits of George’s Bank haddock. Special Scientific Report— Fisheries No. 165, U.S. Dept. of the Inte- rior, 26 p. WOOD, L., 1968, Physiological and ecological aspects of prey selection by the marine gastro- pod Urosalpinx cinerea (Prosobranchia: Muricidae). Malacologia, 6: 267-320. MENISCHAC ES A PIERCE MDAC OSA “swimming” anemone from Puget Sound. Science, 122: 1231-1233. L'HABITAT ET LE COMPORTEMENT PREDATEUR D'EPITONIUM GREENLANDICUM Frank Perron Les epitoniidés sont des mollusques mésogastéropodes prosobranches qui fouillent pour trouver et/ou qui vivent en parasite chez une variété de coelentérés benthiques. Epitonium greenlandicum (Perry) est de distribution circumboréale et se présente au-dessous du niveau de la marée basse dans |’Atlantique du Nord. Des observations au laboratoire et sur place indiquent que ces gastéropodes ne se nourrissent que rarement, et seulement d’anémones. Une fois nourri, Е. greenlandicum se terre dans de la boue molle et peut y rester inactif pendant une période de jusqu’à 80 jours. Les epitoniidés ont la capacité, du moins dans de petites distances, de repérer les anémones par chémotaxis. Quoque Е. greenlandicum puisse se nourrir d'au moins 6 espèces d'anemone, il témoigne d'une préférence pour Metridium senile et tend à vivre en parasite ches cette grande anémone sous des conditions de laboratoire. Dans les eaux peu profondes de la Baie de Fundy, Metridium senile se trouve sur des rochers et des corniches déchaussés, et n’est pas accessible aux epitoniidés qui sont principalement infaunales. Par conséquent, Е. greenlandicum doit fouiller pour trouver de petites anémones sur un vasard caillouté. On suggère que les epitoniidés ne peuvent se nicher d'une manière purement ecto- parasitaire que dans des régions ou il y a à la fois une provision stable des coelentérés hôtes et à proximité un refuge des prédateurs. JADE: 72 PERRON ABCTPAKT МЕСТО ОБИТАНИЯ И ПОВЕДЕНИЕ ПРИ UMTAHMM 2PITONTUM GiusNLANDICUM Франк Е. Перрон ¿pitoniidae (Mesogastropoda:Prosobranchia) - это моллюски, питающиеся различными морскими анемонами. Ppitonium greenlandicum (Перри) по распределению циркумбореальный и встречается в северо- западной части Атлантического океана. Лабораторные и полевые наблюдения указывают Ha то, что этот моллюск питается редко и только анемонами. После питания в. greenlandicum погружается в мягкую тину, где он может оставаться неподвижным вплоть до 80-ти дней. №. greenlandicum обладает спо- собностью находить анемоны при помощи химического ощущения по крайней мере в пределах короткого расстояния. Несмотря на то что 1. greenlandicum может питаться по крайней мере шестью видами анемон, он предпочитает letridium senile и часто паразитирует эту крупную анемону в лабораторных условиях. В мелких водах залива Фунди М. senile, встречающийся на обнаженных скалах и выступах, недоступен 2* reenlandicum, главным образом обитающему в тине. Следовательно Zpitonium greenlandicum вынужден охотиться на мелких анемон на дне, покрытом тиной. Можно заключить, что “Pitoniidae обладают способностью занимать чисто паразитическую нишу только там, где существует постоянный запас анемон и где имеется ближележащее убежище от зрячих хищников. MALACOLOGIA, 1978, 17(1): 73-97 SEASONAL REPRODUCTIVE PATTERNS IN 3 VIVIPARID GASTROPODS Virginia A. Vail Tall Timbers Research Station, Rt. 1, Box 160, Tallahassee, Florida 32303, U.S.A. ABSTRACT Seasonal reproductive cycles in Campeloma geniculum (Conrad), Lioplax pilsbryi Walker and Viviparus georgianus (Lea), based on 12 consecutive monthly collections, are described in terms of activity levels of reproductive organs and brood characteristics. Gametogenic activity and occurrences of mature gametes showed seasonal patterns in each species as did secretory activity of accessory reproductive organs. In C. geniculum commencement of fertilization, as judged by the appearance of the smallest young in the pallial oviduct, followed maximum spermatogenic activity, maximum abundance of eupyrene sperm in the testis and eggs in the ovary, and the highest level of secretory activity in the albumen gland. Similar relationships between reproductive organ activity and brood produc- tion were seen in L. pilsbryi and V. georgianus (but not as clearly due to a greater degree of individual variation in these 2 species). Although reproduction by individuals was similar, the relative degree of synchronization of the individuals sometimes provided for different patterns in the populations. Animals of C. geniculum showed the least amount of individual variation and greatest synchronization, with fertilization and birth periods occurring during comparatively short periods of time in the population. In contrast, animals of L. pilsbryi and especially V. georgianus were unsynchronized, with the population exhibiting fertilization and birth at any time during the year. As a consequence, individual and population patterns were similar т С. geniculum but different in L. pilsbryi and И. georgianus. Greater synchronization in C. geniculum most clearly allowed observation of seasonal brood development. Initial fertilizations provided a brood of small-sized young, and con- tinued fertilizations added new members to the brood; as a result of the latter, the brood was not only larger but consisted of individuals in a size series. Subsequent birth of the larger members of the brood, generated from the earlier fertilizations, reduced the brood size; the remaining young in the brood increased in size prior to their birth. A few individuals contained both large and small young near the end of the birth period, indicating that more than 1 breeding cycle occurs in the life span. This pattern was also shown by individuals of L. pilsbryi and V. georgianus, but its appearance in their populations was obscured by lack of synchrony and smaller brood sizes. Specific details on fertilization, incubation and birth periods, individual reproductive habit and population pattern, number of young per brood and size of newborn are tabulated and compared to prior reports on congeners. Brood production is also shown in terms of comparative size relationships. Viviparus georgianus has the largest gravid adults (average diameter: 26.9 mm) and the largest newborn (> 7.5 mm diameter) from broods averaging 11.2 young per female. Lioplax pilsbryi has the smallest adults (average diameter: 15.7 mm) and moderate-sized newborn (> 4.5 mm diameter) from the smallest broods (average: 9.0 young per female). The intermediate-sized C. geniculum snails (average diameter: 18.2 mm) contained the smallest newborn (> 3.5 mm diameter) from the largest broods (27.7 young per female). The average number of young per female recorded here is the largest observed for these 3 species in this study; in each case the largest broods were comprised of young in a size series. INTRODUCTION Various aspects of reproduction among viviparid gastropods have received prior attention, but the findings have not yet provided a complete understanding of the process and events. Although many descrip- tive accounts of the structural reproductive system exist for these snails (cf. Vail, (73) 1977), specific functions of the reproduc- tive organs have not yet been defined in detail. Moreover, virtually nothing is known of seasonal activity cycles of the individual organs in relation to each other in each sex and to the ultimate production of young. Except for the works of Popoff (1907), Mattox (1937) and Bottke (1972, 1973), oögenic phenomena have been relatively 74 VAIL neglected. However, viviparid spermato- genesis has received considerable attention. Emphasis on the latter has been placed upon the mechanism of production of typical (eupyrene) and atypical (oligopy- rene) spermatozoa and morphology of mature sperm types (cf. Meves, 1903; Pol- lister, 1939; Pollister & Pollister, 1943; Hanson et al., 1952; Gall, 1961). Only Rossi (1968) and Barbato (1971) con- sidered seasonal occurrences of the 2 types of sperm. The functions and seasonal cycles of the accessory reproductive organs (viz., seminal vesicle and prostate gland in males; albu- men gland and seminal receptacle in females) have not been investigated for vivi- parids. However, these organs have been investigated in some oviparous pulmonates and their seasonal activities and hormonal regulation described (cf. Duncan, 1958; Smith, 1966, 1967; Runham & Laryea, 1968; Goudsmit, 1973). Numerous workers have published brief notes, observations which do not take into account seasonal changes, on size dimor- phism between adult male and female vivi- parid shells (e.g., Wood-Mason, 1881), sex ratios (Wood-Mason, 1881; Hubricht, 1943), birth (Frömming, 1928; Crabb, 1929) and size of newborn (Baker, 1928). Other workers conducted more detailed, quantitative, seasonal population studies that provided information on fertilization periods, duration of incubation and life spans, as well as on the aforementioned features (Van Cleave & Lederer, 1932; Van Cleave & Chambers, 1935; Van Cleave & Altringer, 1937; Medcof, 1940; Chamber- lain, 1958; Berry, 1974). Aspects of fecun- dity were considered by Annandale & Sewell (1921), Rohrbach (1937), Miroshnit- shenko (1958), Stanczykowska (1960), Stánczykowska et al. (1971), Samochwa- lenko & Stanczykowska (1972) and Berry (1974). Only Frömming (1931, 1940, 1956) reported general natural history observations. This report provides, for examples of each of 3 Nearctic viviparid genera, viz., Campeloma geniculum (Conrad), Lioplax pilsbryi Walker and Viviparus georgianus (Lea), a description of seasonal activity levels of gonads and accessory organs, analysis of seasonal brood chracteristics and an interpretation of the data showing how these viviparids successfully produce off- spring. The study was conducted on only 1 population of each species, and intra- specific and congeneric variations are unknown. However, the results offer a base line to which future studies can be com- pared. MATERIAL STUDIED The animals of Campeloma geniculum, Lioplax pilsbryi and Viviparus georgianus were from the same Florida populations studied by Vail (1977), who previously noted deposition of voucher specimens in specific museums. The former 2 species occurred at the same site and the latter 1 at a 2nd. METHODS Attempts were made to obtain con- secutive monthly collections of live animals over the span of an entire year: C. geni- culum and L. pilsbryi snails from Decem- ber, 1971 through November, 1972, and V. georgianus individuals from February, 1972 through January, 1973. However, high water levels occasionally prevented an adequate collection. April and July samples of C. geniculum and L. pilsbryi were un- obtainable until 1974. Specimens of L. pilsbryi were not available in February of 3 consecutive years, and for that same reason it was necessary to treat a collection of V, georgianus snails made on June 30, 1972 as the July sample. Individuals were sexed in the field on the basis of asymmetric, right tentacles (the right one being the copula- tory organ of males) and treated as detailed below. Each monthly collection of each species contained up to 30 females which were immediately isolated in individual jars of water to assure accurate analysis of indi- vidual broods regardless of possible abor- tion of young prior to laboratory study. Collecting was purposefully biased in that females of large size (i.e., of more probable reproductive age) were selected for brood analysis. Thus not all females encountered were included. Nevertheless, on occasion (particularly during high waters) a higher proportion of smaller females was obtained in the sample, and many of these animals were too young to be gravid. These isolated females were conchologically measured, de- shelled, fixed and preserved with any VIVIPARID REPRODUCTIVE PATTERNS 75 aborted young in the individual jars. The females were later examined for incubating young. Shell measurements were obtained from all young incubated in 10 randomly chosen gravid females. In addition, 5 females and 5 males per species per month were randomly chosen for histological examination of their gonads and accessory reproductive organs. Narcotization was accomplished with Diabutal (= sodium pentobarbital), fixation with Lavdowsky’s fluid (= AFA) and pre- servation with 70% ethyl! alcohol. Histologi- cal preparations involved paraffin embed- ding, sectioning at 10 um thickness, and staining with Harris’ hematoxylin and alcoholic eosin. Measurements of adult male and female shells were made with Vernier calipers, and of shells of incubating young with a cali- brated ocular micrometer. Other workers have employed length to denote shell size, but, because of frequent apical decollation among adult specimens examined in this .tudy, only diameter was used herein as an index of adult size and also, to be con- sistent, size of incubating young. For con- venience, all incubating young (usually pre- viously termed “embryos” regardless of developmental state) are hereafter desig- nated as “Fy” snails and all post-birth individuals as “P3” animals (or ‘‘adults’’) regardless of size or age. Thus, reference to F size is to the dimension of the Е1 shell diameter; however, reference to brood size is to the total number of Fy’s in the pallial oviduct. References are made later in this report to different seasons in northern Florida as follows: December, January and February constitute winter; spring includes March, April and May; summer contains June, July and August; and September, October and November comprise fall. SEASONAL ORGAN ACTIVITY An arbitrary percentage value was assigned to the state or level of activity of the different reproductive organs in each of the 10 individuals in the monthly samples and an “average state’’ was computed for each month to evaluate the composite activity in the population: low activity = 0%, moderate activity = 50%, and high activity = 100%. The findings are described in this section and illustrated in Figs. 1-3. Stages of both typical and atypical spermatogenesis were identified according to Meves’ (1903) descriptions. Levels of testicular activity were estimated on the basis of relative abundances of typical and atypical spermatogenesis and spermatozoa in histological sections. Possible sperm stor- age was judged by the presence or absence of sperm in the various male and female accessory organs, and, when present, by whether the gametes were free in the lumen or oriented with heads against cells lining the lumen. Levels of glandular activity in the dif- ferent accessory organs were evaluated in terms of the relative abundance of granular secretory products (cf. secretory phases of the albumen gland as described by Smith, 1966). The granular appearance of the secretory products may not be natural but an artifact resulting from reagents used for fixation and/or histological preparation (cf. Goudsmit & Ashwell, 1965). Campeloma geniculum Males. Testis (Fig. 1). Levels of sperma- togenic activity and abundance of sperma- tozoa varied seasonally in the population. Atypical spermatogenesis (ASG) occurred at highest levels during most of winter and early spring, and oligopyrene sperm (OS) were most abundant in mid- and late spring and late summer through mid-fall. Typical spermatogenesis (TSG) was most common from midsummer through fall, with a peak in mid-fall; eupyrene sperm (ES) were abundant during late fall and throughout winter, with a 2nd pulse in midsummer. They were at lowest level in mid-fall. Seminal vesicle (Figs. 1, SV; 4). The numerous villi (VI) are comprised of secretory cells surrounding a central stalk of muscle fibers (MS) extending from the outer circular muscle layer, and small, tri- angular, ciliated interstitial cells (IC) that alternate with the tips of the non-ciliated, columnar secretory cells (SC) adjacent to the lumen. Secretory products of unknown constitution and function appeared as faintly outlined bodies and were neither acidophilic nor basophilic. They were present throughout the year, although peak abundance was observed in mid-fall through winter and again in late spring and early summer (Fig. 1). Vacuoles, usually an indi- cation of relative glandular inactivity, were consistently present. No sperm, either free VAIL 76 "uoseas 18JUIM *|М “siseuabojewlads |еэ!9/з ‘9$. “ajoisen jeulas ‘AS !uoseas Jawuuns ‘NS ‘91263991 jeulwas *ys :uoseas Bulids ‘4$ “puej6 э1е35044 jo ио!бэл 1011915049 ‘Hdd ‘wads auasAdobijo ‘SO ‘Алело ‘O ‘риеб ajexsoid Jo uoibes ajppiw “Han ¿uoseas НЕ} "74 ‘wads auasAdna "63 !sısausßoreunads |еэ!а/зе ‘HS :руе|б ajeisoid jo uoiBai Joualue ‘Yqy ¿pue¡6 uswngje ‘OV "(1xa] 935) Bul] pijos e Aq payesıpuı SI (AS) ajoisan jeulWwas au} и! Ayını)9e A10191985 ¡euoseas AUe jo Y9ej зиэлеаае au] (Z) /Asgsjid ‘7 uj “wiads ausıAdna jo saduasind90 uO paseq эле зиоцемел (HS) э1езЧаээл ¡eulwas ‘(£) зпие/блоэаб sniedialA pue (Z) /Augsyıd xejdo!7 ‘(L) Wn¡n91usb ewojadwed ul зэгилцое чебло элзэпролАэл jeuoseag "5-1 SOI % 001 ES %0s С % 92 = 13 NS | dS IM ie ASS 11 | NS | dS | IM Er ee (oe | 1 rear Aca A e [ge SS ГЕ OO) 0 len) Е. SOS) Ste mola e EE Mar a a e => ГО) 5 O 0600 1000: CEE вау INON dW Еее es add FER Eid AS Se | 53 eee ere) 951 pts SO © > OO Г oo. = “wn OL :Saje9s И\/ “SNIIIA ‘LA ‘21921d9981 jeuiuues ‘HS ‘1189 A10181985 ‘Dg !ajosnw “SN -Чэшп| ‘7 ‘1189 |е13135493и! *9] :Wwuads auasAdna *s3 ‘189 |еи35лази! Jo Snajanu ‘Z [189 A1019198$ JO snajonu ‘| ‘(L) wads auasAdna разиэмо UJIM э|езааэал jeulwas pue (9) wads auaiAdna бипииезиоэ рие|б usunqje ‘(G) puej6 ajeysoid yo uoiBai |ецихола ‘(ÿ) ejoisan jeuiwes wnınaıuab ewojadwed jo suonses |ез3ибезеле "1-9 "SOI Pe a тв "TRAY [Ye wr ЖЕ. EN % a 4 hh my p Г и af y d > ; mM ol и у + À Ра ut we L PAS à ARR о à о 4 = hs i” 4 E a N e e Та ии их be Mn, 1 ) / ” Pr iy 0 y 4 Mo + Be D - E = | BEI Py Mi Sn. yy, pe ni gar 2 wi у ob 4,” 7 y de © 1 lo o M vr ee м, я A м Ar у 4 DOTE CET ES PR Pa D ge m i! sl e м.” er A un ” nf At Mi , su J y Di, W “A u int Ч y 4 [Ty q nm dh hy, ET Cl : m; Y. Y Y м ву a Wy lig УТ РУ у / ym ya 7 VIVIPARID REPRODUCTIVE PATTERNS И 0 a ie ls il a” ll, ly ty Y, ÿ "А ПИ Min, LI 7 7 27 E A, =! 78 VAIL in the lumen or attached to villi, were seen in any of the 60 glands studied. Prostate gland (Figs. 1, 5). Externally appearing to be a convoluted, undifferen- tiated duct, the prostate was found to be comprised of 2 histologically distinct regions: a basophilic, proximal part (Fig. 1, PPR) posterior to the tentacular base, and a strongly acidophilic part distal to that base (Fig. 1, APR). Common to both regions are large, unciliated, columnar secretory cells that extend from the walls to penetrate the cylindrical lumen, and small, triangular, ciliated interstitial cells alternating with the tips of these secretory cells adjacent to the lumen (Fig. 5). Activ- ity levels in the 2 regions overlapped seasonally but showed different patterns. Granules in the proximal region were un- common, although they were maximally abundant during early fall; the distal region displayed a longer period of maximum activity, viz., during the entire winter, although the granules were generally abun- dant at all times. No sperm were observed anywhere in the 60 prostates studied. Females. Ovary (Fig. 1, 0). Oocytes were most abundant in mid-spring and again in midsummer, least numerous т early winter and late spring, but common at all other times. The largest oocytes were 38.6 + 3.9 um (N = 10) in diameter. Albumen gland (Figs. 1, AG; 6). The secretory cells lining the central lumen of very numerous acini showed distinct sea- sonal variation in the abundance of strongly acidophilic granules. Little or no activity occurred from late fall through winter. Activity subsequently increased to a peak in mid-spring and then progressively declined in summer and fall. Eupyrene (never oligopyrene) sperm (Fig. 6, ES) were found in the albumen gland of some specimens throughout the year. Inasmuch as they were always rela- tively few in number, their presence could not be confidently evaluated on a seasonal basis. However, most of the PJ's from late fall through early spring collections con- tained sperm in the acini, and most PJ's from midsummer through mid-fall did not. When present, the sperm were usually oriented in the acini lumens, with heads at least against (penetrating?) the apical ends of the secretory cells, giving the appearance of being stored. Seminal receptacle (Figs. 1, SR; 7). The folds of the glandular lining of this organ include unciliated, columnar cells that con- tain small secretory granules that are only faintly acidophilic. These granules did not exhibit a seasonal occurrence, but there was a marked vacuolization among the cells from late summer through early winter. The presence of spermatozoa was highly seasonal. Most females contained large quantities of predominantly oligopyrene sperm free in the large lumen during the early winter and in the spring and early summer; also during the latter period, some (but relatively few) eupyrene sperm were Oriented against the cells of the glandular folds (Fig. 7, ES). Few females contained sperm in the seminal receptacle during the rest of the summer and throughout the fall. Pallial oviduct. Histological sections of this organ were not made from material in the monthly collections, but the glandular walls of this organ visibly appear thicker during the presence of younger incubating F1's (in general, mid-spring to mid-fall) and become thinner as the F1's advance т development (late fall through early spring). Lioplax pilsbryi Males. Testis (Fig. 2). The seasonal vari- ation in levels of spermatogenic activity and spermatozoan abundance in this popu- lation differed slightly from that of С. geni- culum. Atypical spermatogenesis (ASG) was most active during early and mid-fall; oligopyrene sperm (OS) displayed greatest abundance in spring and late fall. Typical spermatogenesis (TSG) was most active in late spring; eupyrene sperm (ES) were most abundant in early fall and least abundant in late fall. Seminal vesicle (Fig. 2, SV). Secretory products were neither acidophilic nor baso- philic, and large vacuoles comprised a major feature of the secretory cells. Vacuoles were abundant throughout the year, and no seasonality in secretory acti- vity could be ascertained. Sperm were not observed in any of the 50 organs examined. Prostate gland (Fig. 2). Externally appearing to be an undifferentiated duct, the prostate has а posterior, proximal region (PPR) that is less acidophilic than the distal region (APR). This differential staining may be the result of different amounts of a single secretion or of dif- ferent secretions. In both regions non- ciliated, tall-columnar secretory cells line VIVIPARID REPRODUCTIVE PATTERNS 79 the cylindrical central lumen, and small, triangular, ciliated interstitial cells alternate with the tips of the secretory cells. Sec- retory granules were always abundant in both regions, and no significant seasonal variation in secretory activity was perceived in the population sample. Neither eupyrene nor oligopyrene sperm were found in any of the 50 glands examined. Females. Ovary (Fig. 2, 0). Although common during the rest of the year, oöcytes were most abundant in the popula- tion in mid-spring through midsummer and again in early fall in contrast to peaks of shorter duration in С. geniculum. The average diameter of the largest oocytes was 75.1 + 3.5 um (N = 10). Albumen gland (Fig. 2, AG). Seasonal variation in the abundance of granules in the secretory cells contrasted to that observed т С. geniculum. In L. pilsbryi inactive acini were found in the population only in late fall, and maximum secretory activity occurred in early winter through spring. Eupyrene (never oligopyrene) sperm were found each month in some Р1’5 in albumen glands of all activity levels. Although accurate estimation of abundance was impossible, it was noted that propor- tionately more females contained sperm there during mid- and late spring, mid- summer and late fall. The sperm were located within the lumen of the acini, oriented with heads against (into?) the sec- retory cells. Sperm were also found in the albumen gland duct, and in both arms of the oviduct, in which they were unattached but with heads directed toward the albumen gland. Seminal receptacle (Fig. 2, SR). Although small, faintly acidophilic granules were observed, no perceptible seasonal secretory activity pattern was found in the population. Also, neither presence nor absence of sperm was seasonal. Indeed, only 4 of 55 Р1'’5 contained sperm (mostly eupyrene, but some oligopyrene), and of those only 2 had large numbers of these sperm. Pallial oviduct. The observed seasonal changes in this L. pilsbryi population were similar to those of С. geniculum. Viviparus georgianus Males. Testis (Fig. 3). The seasonal varia- tion of spermatogenic activity in this popu- lation contrasts with that observed in C. geniculum and L. pilsbryi as follows. In V. georgianus, atypical spermatogenesis (ASG) showed greatest activity in late fall through mid-winter, thereafter progressively decreas- ing to lowest levels in mid- and late spring, and then gradually increasing during the summer through mid-fall. An August decline is interpreted as reflecting inherent individual variation rather than an actual decrease in activity. Oligopyrene sperm (OS) were most abundant during early win- ter, only moderately abundant in mid- winter through early spring, and least abundant in late spring through early fall, after which time they began to increase in number. Typical spermatogenesis (TSG) was at highest levels during mid- and late spring and least active from late fall through mid-winter. Eupyrene sperm (ES) were most abundant from late spring through the summer, and least abundant in early winter. Seminal vesicle (Fig. 3, SV). Neither ciliated interstitial cells nor spermatozoa were seen in any of the 60 organs exam- ined. Secretory granules were most abundant in early winter, with only a slight decrease in relative numbers during mid- and late winter. Least secretory activity occurred from late spring through early fall. Vacuoles in the secretory cells became more abundant as secretory activity de- creased. Prostate gland (Fig. 3). Secretory activ- ity and also a peculiar regional differenti- ation showed seasonal variation. Internal regionalization of this gland was not notice- able in late winter specimens, and indi- viduals then showed variable, but generally low to moderate, secretory activity. Two distinct regions of the prostate were dis- cernible т most early and mid-spring specimens: the posterior 2/3 (PPR) con- tained moderately acidophilic granules, whereas the anterior 1/3 (APR) contained only strongly acidophilic granules. At this time both regions displayed moderate to high levels of secretory activity. During maximum prostate secretory activity, in late spring through early winter, 3 regions were discernible: a posterior region (PPR) with large, coarse, moderately acidophilic granules; a middle part (perhaps cor- responding to the red region in live animals; cf. Vail, 1977) (MPR) containing equally acidophilic but smaller, finer granules; and a strongly acidophilic anterior 80 VAIL section (APR) containing medium-sized granules. During the subsequent period of decreasing activity, only 2 regions were observed, viz. a moderately acidophilic posterior 2/3 and a strongly acidophilic anterior 1/3. Regional activity levels were almost always synchronous. Fig. 3 shows seasonal secretory activity patterns of the prostate regions for the population, but does not reveal individual variation within a month. Neither eupyrene nor oligopyrene sperm were found in any of the 60 organs studied. Females. Ovary (Fig. 3, 0). In contrast to those of C. geniculum and L. pilsbryi, oöcytes in V. georgianus were in greatest abundance in late winter, only slightly less abundant in late fall and, although still common throughout the rest of the year, showed a slight decrease in numbers in the spring. The diameter of the largest oocytes was 51.8 + 4.8 um (N = 10). Albumen gland (Fig. 3, AG). Maximum secretory activity in the population occurred in mid- and late winter, with a 2nd minor peak in early summer. Least activity occurred in late summer through mid-fall. Unlike the occurrences in С. geniculum and L. pilsbryi, none of the 60 albumen glands examined of V. georgianus snails contained either eupyrene or oligopyrene sperm. That observation may be erroneous, however, because the mature glands undergoing greater secretory activity sectioned poorly, with the tissue usually shredding. This problem was less severe in C. geniculum and L. pilsbryi. Seminal receptacle (Fig. 3, SR). Ciliated, columnar secretory cells, surrounding muscle fibers in the villi that project into the lumen of this organ, showed no seasonal variations in activity levels; sec- retory granules were common throughout the year. At least some Р1’5 from each month contained both eupyrene and oligopyrene sperm in the seminal гесер- tacle; most Р1’5 contained these sperm in mid- and late winter and mid- and late summer; fewest Р1’5 contained them in early summer and fall. Both oligopyrene and eupyrene sperm were found in the lumen, whereas only the latter occurred attached to the walls. Pallial oviduct. Monthly sections of the posterior portion of this organ were in- cluded in histological preparations of the albumen gland and seminal receptacle, but quantitative histological evaluation of the whole organ was not attempted. Histo- logical sections have provided some quali- tative information about the posterior, proximal end. Ciliated, tall-columnar sec- retory cells lining the lumen centrally con- tained many vacuoles; some of the vacuoles contained a subelliptical secretory body, whereas others contained what appeared to be degenerating sperm (both types?). Gross dissections provided the visual impression that the walls varied in thickness seasonally in a manner similar to that in C. geniculum and L. pilsbryi. BROOD PRODUCTION Information presented in this section was obtained from monthly size-frequency histograms of Е1'’$, seasonal frequencies of the different stages of brood development and both seasonal and non-seasonal analy- ses of the Px's. ; Different numbers of gravid Р1’5 and of F1's occurred from month to month, and therefore total numbers in each sample were converted to percentages; the latter serve as “frequency” in the histograms (Figs. 8, 15, 16) and in the figures (9-11) denoting the seasonal distribution of broods in different developmental stages. The constitution of individual broods varied both within a single month as well as seasonally. Some broods contained either only small F1's or only large F1's, others contained a size series of smaller to larger F4’s and still others contained Fy’s of 2 different broods, each from a different fer- tilization as evidenced by different size classes. As shown later, these different con- ditions are seasonal stages in brood devel- opment. Campeloma geniculum Fertilization, incubation and birth periods, and brood (F1) characteristics were highly seasonal. Monthly size-fre- quency histograms of the Fy population (Fig. 8) indicated that fertilizations, inter- preted according to the presence of smallest, youngest F1's in the pallial ovi- ducts, began about mid-spring and, although decreasing in frequency after mid- summer, continued to early fall. The histo- grams suggested that (1) incubation occurred over the winter, and birth, judged by the gradual disappearance of the largest, VIVIPARID REPRODUCTIVE PATTERNS 81 80 DEC APR Fe JAN MAY SEPT О Е. [0/6 Эс 10/129 10/221 40 20 OCT 807 ¡0/44 10/118 10/327 FREQUENCY (%) 80 MAR JULY NOV 604 9/53 10/133 10/204 бе pa Тв Gres bn pata Be OS SIZE (Е1 DIAM. IN MM) FIG. 8. Monthly size-frequency histogram of incubating Campeloma geniculum snails. Diam., diameter; F4, young developing within the pallial oviduct. Frequency: % of sample; fraction numerator: number of P1’s examined; denominator: number of F1's examined. "э1ещэ; | Ul SPOOJG Z JO 29U911N990 зпозиезпии$ ‘Hg Z :р0024 ajGuis e и! BunoA jews Ajuo ‘5-38 | :ро0.а ajBuis e и! BunoA abe Ajuo ’7-ıg | ‘18}UIM “| :JaWwuuns "ns ‘Butids ‘gS ¿seras azis e ul роола э|би!$ ‘5313$ “a¡euay риле1б-иоц ‘ON “1184 “14 "suoseas :essiosqy "TuaWwdojanap роол4 jo sabe]s :эзеи!рлО ‘(LL) snue/6108b snsediAIA pue (01) /Asgsııd xeydoı7 *(6) шпупэ!иаб ewojadwe) ui зиэшао|элэр poolq jo sabejs зиэлазуир yo (ajdwes jo %) sainuanbasy jeuoseag `| 1-6 "SOI %001 ES %05 O St 11 ns | dS | IM 11 | NS | dS | IM ON 48 à VAIL 1-49 1 $3143$ S-44 L Ц 82 VIVIPARID REPRODUCTIVE PATTERNS 83 oldest Fy y’s from the pallial oviducts, occurred in the spring; (2) only 1 brood, with an incubation period of perhaps as long as 12 months, was generated in a year; (3) birth of that brood was not completed before a 2nd fertilization period provided for a subsequent brood, i.e. there was no interim period of non-gravidity among the P1's. Seasonal analyses of the nature of the broods and also of brood sizes provided a more detailed, modified interpretation of some of these initial, provisional con- clusions. Most females with only small FJ's (arbitrarily, diameter < 1.0 mm) occurred in mid-spring through midsummer, those with Fy’s in a size series from mainly late summer through mid-fall (0.5-1.5 mm in August, 1.0-2.5 mm in September, and 2.0-3.5 mm in October) and those with just large F1's (> 3.0 mm) principally in late fall throughout the winter (Fig. 9). Also, only 9 of the 30 PJ's (i.e., 30.0%) in the March collection, all of reproductive size, were gravid, and just 2 (1 each in April and May) of the total of 279 gravid females simultaneously contained 2 broods, 1 with small F1's and 1 with large FJ's. Following early development of an initial number of Fy’s in the spring and early summer, continued fertilizations dur- ing the rest of the summer and early fall added new members to the brood, thus providing for a single brood in a size series during the latter period. The largest Fy’s in size-series broods, generated by fertilization in mid-spring to early summer, were born during the winter after an incubation period of 8-10 months. The smallest Fy’s in size-series broods, generated by fertili- zation in midsummer to early fall, matured to become the late fall and winter broods of only large F1's and were born in the spring after an incubation period of 8-10 months. Comparison of seasonal brood sizes supports these conclusions. Beginning with the appearance of small F1's in the pallial oviduct (cf. April, Fig. 8), the mean brood size (i.e. the mean number of young per brood) in Р1’$ with only small F1's gradually increased. The mean brood size increased markedly when F1's in a size series predominated (cf. Fig. 12). In late fall, when broods of large FJ's began to appear, the mean brood size decreased gradually as births commenced, reaching a low in the late winter and spring as births were concluded. Although the gravid Р1’5 examined throughout the study were of similar mean diameter, the mean brood size varied in a seasonal pattern, increasing from a low in February (large F1's) to highs in August and October (F1's in size series) and then decreasing thereafter (Fig. 12, Table 1). The unexpected lower mean brood size in September is attributed to the Fy’s (size series) occurring in smaller Р1’5 than did larger broods immediately before and after. In contrast to that seasonal independence of mean brood size in similar-sized Р1’$, analysis of the August Р1 collection (month of the maximum mean brood size) showed that mean brood size correlated with the size of the Pq’s (гр = 0.742, P< 0.01). In Table 2, composite data from all monthly collections show brood sizes in relation to PJ size classes. The smallest gravid Pq was 13.2 mm in diameter, and just 7 of 21 P7's < 14.0 mm were gravid. The F1's in those 7 P1's were probably generated by fertilization in the prior spring to summer. The P}’s came from fall (F1's in size series) and early winter (only large F1's) collections and had brood sizes of 2-7. The Р1’5 themselves may have been born as recently as the previous winter. Unfortunately, possibly gravid P1's were not saved in the other collections because enough larger females were present. Judged by sizes of largest F 7's in the pallial oviduct (cf. Fig. 8), newborn were about 2 3.5 mm in diameter. Lioplax pilsbryi Fertilization, incubation and birth periods, as well as brood (F7) characteris- tics, were again of seasonal nature. Monthly size-frequency histograms of the Е1 popula- tion (Fig. 15) indicated that (1) fertili- zation took place throughout the year, with highest frequency in the spring, de- creasing to lowest values in late fall and then increasing during winter; (2) births began in spring and lasted until mid- summer; (3) only 1 brood was produced in a year, after an incubation period of up to 12 months; (4) fertilization of a later brood began before birth of the previous brood was completed. Clarification of these interpretations is provided by analy- ses Of seasonal ontogenetic brood stages and of brood sizes. Broods comprised solely of small Е1’$ (diameter < 1.0 mm), in 6.1% of all gravid “492 UIM “| М :4эщип$ ‘NS :6и14 “ys ‘jenpiAipul пре “ta ¿mes 13 (pL) snuei6so0a6 ‘A pue (EL) tAsgsjid ‘7 ‘(у age ass osje ‘ZL) шп/пэ/иаб ewojadwe) ul sajewas Ly piaes6 yo azis ueaw pue azis роола ueaw usamiaq diysuonejal |еио5еэ$ “pPL-ZL ‘$9134 11 | NS | dS IM 11 | NS | dS IM Wee AS = dS IM VAIL (----) 371$ 00049 NVIM (—)WWNI'WVIO 14 NVIM yl el él 84 VIVIPARID REPRODUCTIVE PATTERNS TABLE 1. Monthly sizes of gravid Campe/oma geniculum Р, '5 and their broods. (Also see Fig. 12.) D Ч Е M A P, diam. (mm) x 16.8 te Se YES GIS S.D. 2.0 1.8 4.2 ee! 1.4 Range 13.2- 14.8- 14.7- 15.2- 17.6- 21.0 2019) 20:97 1915) 22:3 No. P,’s 27 21 19 9 10 Brood size (mean number of young per brood) x 10.7 6.1 4.1 5.9 5:9 S.D. 8.8 4.5 sl 4.9 3.9 Range 2- 1- 1- 1- lo 37 20 11 14 14 Total F, 's 200 128 78 53 59 85 Month M J J A 5 О М 17.6 16.7 11748 18.0 16.9 UT 17.0 22, 3.4 172 1.6 12 15 1.8 15.0- 15.0- 16.2- 14.1- 13.5- 14.9- 13.8- 25.4 2120 20.0 20.5 19.3 19.3 20.9 27 26 29 29 24 30 28 9.2 9.7 13.3 297, 20.8 727). 19.2 7.0 5.7 Ust 11.9 8.9 ae 9.1 1- 3- 1- 8- 6- 10- 3- 37 22 31 55 43 51 38 248 252 385 804 498 817 538 P1's, occurred only in spring through mid- summer, and then only in low frequencies (Fig. 10). The Fy y’s in a size series (1.5-4.4 mm in winter, < 0.5-4.2 mm in spring, < 0.5-4.8 mm in summer and 1.4-4.9 mm in fall), were observed in 43.3% of all gravid P1's, and occurred in all months except January, but were most frequent from early summer through mid-fall. Broods of only large Е1’$ (> 3.0 mm) were rare, in just 3.2% of all gravid P1's, and occurred only in mid-fall through early winter. The P1's simultaneously incubating members of 2 different broods were uncommon (9.7% of all gravid P1's) and of wide temporal occurrence, but showed greatest frequency in mid-winter through spring. Non-gravid P1's (13.3-63.2% of the monthly collec- tions, and 48.6% of the total number of females) were most frequent in mid-fall through mid-winter, but these were often too small (young) to be reproductively mature. Continued development of the 1st F1's formed plus subsequent fertilization occurring at intervals caused broods in a size series. Broods of only large Fy’s were composed of the formerly younger F ’s of the size-series condition remaining after the birth of the older Е1’5. Broods with 2 size groups of Fy’s occurred т P4’s that under- went fertilization to begin a new brood before the large Fy’s of the existing brood (i.e. the youngest F3's of the size-series condition) were born. On the basis of sea- sonal frequencies of Р1’5 with 1 brood of just large F1's, and of those with 2 broods, it is concluded that birth occurred in the winter after an incubation period of 10-12 months; this interpretation is supported by a seasonal comparison of brood sizes (see below). According to monthly size-fre- quency histograms (cf. Fig. 15), births con- tinued into spring after an incubation period of like duration. The mean brood size was always com- paratively small, but still varied seasonally in similar-sized Р1’5 (cf. Fig. 13). Smallest broods occurred during the winter and early spring, after which the brood size in- creased to a peak in late summer before decreasing during fall. Analysis of dif- ferent-sized Р1’5 in the August sample (which had the largest brood sizes) showed there was no significant increase in brood size with the increase in the size of P1's (гр = 0.244, P > 0.05). In Table 2 a composite of data from all monthly collections show mean brood size in relation to PJ size. The smallest gravid P was 13.8 mm in diameter, and only 4 of 69 P7's < 15.0 mm were gravid. Such relatively small P1's occurred in each monthly collection, whereas those 4 gravid P1's were found only in midsummer into early fall. The F1's in the 4 gravid P's, probably gene- rated by fertilization in the prior spring, constituted brood sizes of 3-10, all in a size series; the P1's themselves may have been born a year before (i.e., in the preceding winter or spring), and were nearly full-sized P1's. Judged by sizes of largest F1's (Fig. 15), newborn were about > 4.5 mm in diameter. Viviparus georgianus The majority of size-frequency histo- grams of the Е1 population (Fig. 16) dis- 86 VAIL TABLE 2. Composite data, combined from all monthly samples, showing average number of F,’s in Campeloma geniculum, Lioplax pilsbryi and Viviparus georgianus in relation to size of gravid P,’s. (& pre L. pilsbryi V. georgianus P, diameter No. P, 's No. No. P, 's No. F,'s «No: P;'s No. Е, 's (mm) examined | X+S.D. (Gael | examined x+S.D. (range) examined X+S.D. (range) 30.1-31.0 2 12.5 +0.7 (12-13) 29.1-30.0 9 11.33+4.35 (6-18) 28.1-29.0 16 12.43+7.18(1-23) 27.1-28.0 20 10.35+6.21 (1-23) 26.1-27.0 37 9.56+6.16(1-25) 25.1-26.0 1 37.0+ 0.0(-37-) 65 7.40+5.04 (1-20) 24.1-25.0 73 7.40+4.80(1-24) 23.1-24.0 O -------- 58 7.10+4.24 (1-17) 22.1-23.0 1 1.0+0.0(-1-) 28 5.25+3.54 (1-14) 21.1-22.0 O Een 10 3.81+1.83(1-6) 20.1-21.0 20 22.2+18.9(1-55) O CCE 19.1-20.0 31 21.3+12.1 (1-38) O --------- 18.1-19.0 56 16.0+12.0(1-50) 5 6.4 +3.17(4-13) 0 5: nee 17.1-18.0 60 16.0+12.0(1-45) 27 6.24+2.71(2-13) QA 16.1-17.0 57 13.51 8.7(1-36) 74 6.48+2.75(2-13) O0 =: CET 15.1-16.0 34 8.0+ 4.7(1-17) 44 6.57+2.82(2-12) 1 15.0+0.0(-15-) 14.1-15.0 12 7.2+ 4.0(1-14) 3 8.0 +2.60(5-10) 1 5.0+0.0(-5-) 13.1-14.0 7 3.9+ 2.1(2-7) 1 3.0 +0.0 (-3-) OÙ ее played a bimodal distribution that perhaps initially suggests the production of 2 broods in a year, with Pj’s typically becoming gravid with a new brood prior to the birth of an existing brood and simultaneously incubating the 2 broods for most of the year. However, 2 periods of fertilization and of birth are not evident in Fig. 16. Fertilization began in late fall, increased in frequency over the winter to a maximum in mid-spring, and then decreased into early fall. Births began in mid-winter and continued through mid- summer. A clearer picture of seasonal pattern(s) is provided by an evaluation of brood characteristics. Gravid P individuals contained Fj’s in 1 of 4 developmental conditions (Fig. 11). Broods of only small F1's (diameter < 1.0 mm) occurred in all months except October and November, but were most fre- quent in spring and early summer. The Е1'$ in a size series, found in some Pj’s each month (< 1.0-7.0 mm in winter, < 1.0-6.0 mm in spring, < 1.0-7.5 mm in summer and 3.0-7.5 mm in fall), were most fre- quent in midsummer through mid-fall and least common in spring. Broods of just large Е1’$ (> 5.0 mm) were most numerous in late fall into mid-winter, decreased in frequency into early summer, and were lacking in mid- and late summer. Some Р1 individuals simultaneously contained mem- bers of 2 different broods (large F1's and small F1's) in each monthly collection, but they were most abundant in late winter and early spring, and again in mid- and late fall. A few non-gravid females were found each month, but were only sporadically more common, viz., in November, Decem- ber and May (non-gravid P3's of reproduc- tive size; see below) and in September (all non-gravid Р1’$ of less than reproductive size). According to those findings, most P1's underwent initial fertilization in spring and early summer; continued fertilization into fall added new members to the brood, which now appeared as F7's in a size series. The subsequent appearance of broods con- taining only large F1's (in late fall to mid- winter) was a result of the birth of the largest F1 members in the previous size series. The presence of 2 broods in some P1's at that same time indicates that the large F1's were joined by small F1's from uncommon early fall fertilizations. Never- theless, some P y's exhibited lower fre- quencies of these different brood con- ditions during other seasons, indicating that fertilizations, sequential developmental stages and births evidently were able to occur at almost any time. Moreover, it is possible that, in some instances, some P7’s did not undergo continued fertilzations, and broods of only small Е1’5 may have directly become broods of just large Е1’5. Comparisons of seasonal brood sizes in the majority of Р1’5 that display the most fre- quent pattern in general support those con- clusions. УПРАВ О REPRODUCTIVE PATTERNS 87 60 DEC APR AUG 401 7/28 5/28 10/89 20 60 pa JAN MAY SEPT > ao! 5/18 10/74 10/64 > U 20 yA Lu I 60 O JUNE OCT Led 40 u 10/77 10/69 [= 4 LL. 20 MAR JULY NOV 6/21 10/68 10/42 40 20 a. Le © 2 4 6 Dr Avvo 6 2 4 SIZE (F1 DIAM. IN MM) FIG. 15. Monthly size-frequency histograms of incubating Liop/ax pilsbryi snails. Diam., diameter (mm); F4, young developing within the pallial oviduct. Frequency: % of sample; fraction numerator: number of P1's examined; denominator: number of F4’s examined. 88 VAIL 100 80 FEB 10 JUNE OCT 60 10/89 10/89 10/69 40 20 100 >= 80 MAR 30 JUNE NOV nt 60 10/81 10/106 10/68 > 40 У 20 Z Lu — 100 G 80 APR AUG DEC = = 10/76 10/119 10/55 LL. 20 100 80 MAY SEPT JAN 60 10 10/97 40 10/57 /99 20 246810 246810 246810 SIZE (F1 DIAM. IN MM) FIG. 16. Monthly size-frequency histograms of incubating Viviparus georgianus snails. Diam., diameter (mm); F4, young developing within the pallial oviduct. Frequency: % of sample; fraction numerator: number of P4’s examined; denominator: number of F4’s examined. УПРАВ О REPRODUCTIVE PATTERNS 89 Among Р1’5 incubating only small F7's, the greatest mean brood size (8.6 Е1’5/Р1) occurred in the spring. As brood develop- ment progressed and Py with only size- series broods were prevalent, the mean brood size increased over the summer (to 11.6-12.5 Е1’5/Р1) and then decreased to low levels in early and mid-fall. As births occurred, brood sizes decreased. Among P7’s with just large F1's, brood sizes were largest in late fall into mid-winter (3.8-4.4); Pı’s with 2 broods had smaller “сот- posite”” brood sizes in mid- and late fall (6.6 and 5.4, respectively), and showed larger values in late winter and early spring (9.2 and 11.6). Seasonal variation in those values reflects a combination of additions of small Fy’s by fertilizations and subtrac- tions due to births of large Fy’s. Gravid P1's were of similar mean size (diameter) throughout the study, although the mean brood size, without consideration of different kinds of developmental stages, varied in a seasonal pattern (Fig. 14). The mean brood size increased from its lowest value in early winter to an initial peak in late winter, then decreased to a low in late spring, after which it rose to a maximum value in late summer and then steadily decreased during the fall. This pattern, like that seen in the F7 size-frequency histo- grams (Fig. 16), is one of bimodal distri- bution, and suggests that births occurred within the entire population mostly in mid-spring into early summer and again in late fall and early winter. The earlier period, clearly identified in the histograms by low frequencies of the largest Fy’s, is characterized principally by the presence of small broods of only small F7's generated by recent fertilizations; the scarcity of large F1's here is indicative of the birth period. The latter period, characterized mainly by little broods of large Fy’s, is not a birth period; rather the small brood sizes are the result of the absence of small F1's from the recent fertilizations. Largest brood sizes were found in mid- and late summer (mainly F1's in a size series), with a secondary pulse in late winter and spring (high frequencies of P4’s containing 2 broods). Analysis of the August P sample (having the maximum brood size for the year) showed brood size and Р1 size were not correlated (гр = 0.057, P > 0.05). Table 2, listing composite data from all monthly collections, shows mean brood sizes in relation to PJ size classes. The smallest gravid Р1’5 were 15.6 and 15.0 mm in diameter; those were the only individuals among 19 P3's < 21.0 mm that were gravid, and just 12 of 31 P1's < 22.0 mm were gravid. The F7's in the 2 smallest gravid Р1’5, from the collection of June 30 constitute brood sizes of 15 (size series) and 5 (only small F7). Judged by sizes of the largest Р1’5 in the pallial oviduct (Fig. 16), newborn were about > 7.5 mm т diameter. DISCUSSION Relationship of Reproductive Organ Acti- vity and Brood Production When fertilizations and births occur in all individuals during a comparatively dis- crete period of time, the population re- flects a relative synchronization of indi- vidual members. Lesser individual variation in fertilization or birth periods presents a more synchronized pattern for the popula- tion than does greater individual variation. Moreover, the identification and description of any coordination (either simultaneous or sequential) between organ activities and brood production at both the individual and the population level becomes more dif- ficult with increasing individual variation. Some variation observed here within each monthly sample may have occurred because the animals were born at different times and underwent different rates of matura- tion. Among the 3 viviparids described herein, C. geniculum individuals exhibited the greatest synchronization, with fertilizations and births occurring in the population at discrete periods of time. Peak typical spermatogenesis, maximum abundance of eupyrene sperm in the testis, greatest abundance of oöcytes in the ovary, and highest levels of secretory activity by the albumen gland occurred at or prior to the commencement of fertilization, i.e. prior to the appearance of the smallest Fy’s in the pallial oviduct. Although not conclusively shown by the available data, copulation (circumstantially identified here by the presence of abundant eupyrene and oligopyrene sperm in the seminal receptacle and by eupyrene sperm in the albumen gland) is thought to have occurred during late fall through mid-winter in larger, older females, and in the other females in the 90 VAIL spring. Mating pairs were never seen in the field; higher water levels in winter and spring preclude their detection. Moreover, during periods of low temperatures the animals burrow into the substrate, where copulation in L. pilsbryi was observed to occur (see below). Following initial fertili- zations, organ and gland activities decreased somewhat (Fig. 1), even as fertilizations continued (at lower frequencies) (Fig. 8) to create broods with FJ's in a size series. The albumen gland remained relatively active, providing sufficient quantities of perivitel- line fluid for new embryos during that period. In contrast, greater variation between in- dividuals of L. pilsbryi and V. georgianus contributed to longer periods of fertili- zation and birth in those populations, and thus maturation rates and the time of the Ist fertilization varied. Peak abundance of oöcytes in the ovary of L. pilsbryi occurred during the time of highest fertilization frequencies, and the albumen gland was very active only during the earlier part of the long fertilization period (Figs. 2, 15). However, eupyrene sperm in the testis appeared to be un- common during this time. Mating pairs, partly embedded in the substrate, were found in the field in May and October. These observations support the conclusion that spring and fall are copulatory periods. This is suggested by the occurrence of eupyrene sperm at those times in the albumen gland (but not in the seminal receptacle), but do not explain the relative lack of sperm in that gland during the winter or their presence there in mid- summer. If sperm present in the albumen gland in July were received during copula- tion in the prior spring and stored there, it would appear possible that sperm received in the fall copulation could similarly be stored in the albumen gland in the winter. How- ever, few Р1’5 in the winter contained sperm in the albumen gland, but that lack may reflect failure of most of the investi- gated specimens to have mated in the fall. Some of the smallest F1's occurred in the pallial oviduct in all months, but times of lowest frequencies of those Е1’5 (fall and early winter) may be considered a post- fertilization period in most of the popula- tion; organ and gland activities were diminished then in the monthly samples. Least synchronization occurred in V. georgianus, where P 1 individuals showed comparatively long fertilization and birth periods. Mating pairs were never seen in Holmes Creek, although copulation in other populations was observed in the spring and fall (also during winter among snails т aquaria). Judged by frequencies of Py’s with sperm in the seminal receptacle (but not in the albumen gland), copulation in Holmes Creek must have occurred immedi- ately prior to a 1st fertilization period. At the same time, immediately preceding the highest frequencies of the smallest F7's in the pallial oviduct, eupyrene sperm were most abundant and greatest secretory activity of the albumen gland occurred (Fig. 3). Eggs were most numerous at 2 times in the population, once in the fertili- zation period and again in the post-ferti- lization period (identified by low fre- quencies of the smallest F3's). Eupyrene sperm were least abundant and albumen gland activity was lowest during the post- fertilization period (Fig. 3). Nevertheless, at least some individuals, both male and female, appeared to display a high level of gonadal and glandular activity in each month. Reproductive Organ Function It is generally conceded that eupyrene sperm from typical spermatogenesis are capable of fertilization. However, a variety of other functions have been hypothesized for the oligopyrene sperm from atypical spermatogenesis (Meves, 1903; Ankel, 1924, 1930; Gall, 1961; Nishiwaki, 1964; Rossi, 1968). Inasmuch as atypical sperma- togenesis and oligopyrene sperm exhibit high seasonal levels prior to those of typi- cal spermatogenesis and eupyrene sperm (Rossi, 1968; Barbato, 1971; partly con- firmed here), the hypothesis that oligo- pyrene sperm, by catabolic breakdown, provide a nutrient-energy source for eupyrene sperm (Hanson et al., 1952; Yasuzumi & Tanaka, 1958; Dembski, 1968) might be the most logical. Dembski (1968) also suggested that disintegrated atypical sperm in the seminal receptacle and pallial oviduct might provide nutrient matter for the female. However, the actual function of oligopyrene sperm has not yet been clearly demonstrated. Although Rossi (1968) and Barbato (1971) reported atypical spermatogenesis and oligopyrene sperm to be _ prevalent during colder, winter months, findings for VIVIPARID REPRODUCTIVE PATTERNS 91 all 3 species studied here showed atypical spermatogenesis and spermatozoa to be common during warmer periods also (Figs. 1-3). Moreover, oligopyrene and eupyrene sperm are not mutually exclusive in occur- rence either within the population or with- in a single individual, and both types have simultaneously been found in the pallial oviduct and seminal receptacle of females (Popoff, 1907; Hanson et al., 1952; Dembski, 1968; verified here). Seminal vesicle function in gastropods is commonly considered to be the storage of autosperm (i.e. sperm produced by the individual). Secretory activity to provide “nutrients” for the sperm would seem to be associated with the provision of such space. However, no individual of the 3 vivi- parids studied here contained sperm in that organ. The secretions might contribute to a kind of “culture medium” for sperm pass- ing through the tract. Liop/ax pilsbryi has a seminal vesicle of somewhat different struc- tural organization from that in С. geni- culum and У. georgianus (Vail, 1977), and that gland’s secretions in L. pilsbryi may or may not be different from those in the other 2 species. The same process may occur in the prostate gland in V. georgianus (Vail, 1977), and indeed in this species the seasonal regionalization shown Бу dif- ferential staining properties suggests that the same region of the prostate might secrete different products through time. These secretions presumably serve in the maintenance of autosperm, none of which were found stored in this gland. However, Kamaloney (1968) reported sperm stored in the prostate gland of Bellamya suma- trensis (Viviparidae: Bellamyinae). Following copulation, sperm deposited in the pallial oviduct are conveyed to the seminal receptacle along the sperm groove, although the mechanism for keeping at least most of these oligopyrene and eupy- rene gametes in that open groove is un- known. As previously noted, the walls of the pallial oviduct are seasonally thicker or thinner with glandular tissue, and the secre- tions may in part serve for sperm mainte- nance. The seminal receptacle in gastropods is considered the usual site for storage of allo- sperm (i.e., sperm received during copu- lation) in part because it produces а “nutritive” — secretion. However, few females of any of the 3 species studied here contained either oligopyrene or eupyrene sperm oriented with their head against (penetrating?) the cells of the inner surface of this structure. Very few L. pilsbryi females contained any sperm in the seminal receptacle, and the majority of C. geniculum and V. georgianus females with sperm in this organ contained them free in the lumen. This latter condition does not suggest prolonged sperm storage in this organ in these species. In С. geniculum and L. pilsbryi, but not in V. georgianus (perhaps due solely to shredded sections), eupyrene sperm were commonly found in the oviduct and in the albumen gland. Sperm were possibly being stored there as they were in the albumen gland and oviduct of V. contectus accord- ing to Dembski (1968). Copulation in some С. geniculum animals may occur several months prior to fertilization, necessitating a mechanism for comparatively long-term sperm storage in females. Such a condition would explain the availability of sperm that contribute to sequential fertilizations to provide broods with F1's in a size series. Sperm storage in the albumen gland would allow for simultaneous discharge of sperm and albumen (termed “perivitelline fluid” when surrounding a developing Fj within an acellular capsule) into the oviduct near the junction of the ovary and oviduct. Whereas Dembski (1968: 151) suggested for V. contectus that “it is possible that the disintegrated and ingested (eupyrene) sperm which were found in the albuminous gland are converted into yolk by the glandular cells of this gland.’’ Annandale € Sewell (1921: 235) previously reported (eupyrene?) sperm in the perivitelline fluid of /diopoma bengalensis (Bellamyinae). The presence of an egg in the proximal arm of the oviduct might initiate such a sperm- albumen discharge, or an initial discharge from the albumen gland might stimulate the release of oöcytes from the ovary to the oviduct. Similarly, fertilization in L. pilsbryi occurred throughout the year in the population, at irregular intervals in P4 individuals resulting in numerous occur- rences of broods with Е1 in a size series. This condition is likewise attributed to the storage of sperm for intermittent fertili- zation. It is nevertheless enigmatic that, while apparently containing more eggs in the ovary, L. pilsbryi Р1’$ release fewer eggs at a time and at apparently greater intervals than do C. geniculum and V. georgianus P's. 92 VAIL Few details of the nature of viviparid albumen and its functions(s) as perivitelline fluid are known. Annandale & Sewell (1921: 235) noted presumably calcareous spicules in the perivitelline fluid of /dio- рота bengalensis and suggested that the F1's use these spicules in their shell pro- duction. In his study on the “‘nutrition’’ of У viviparus Е1’$, Charin (1926: 66) recorded mucus with bound carbohydrates (mucopolysaccharides?) in the perivitelline fluid and the absence of ”... frei Kohlen- hydrate (auch Glykogen) und Lipoiden.” He also noted that the surrounding capsule, rich in lipids, was permeable to carbo- hydrates and lipids, and suggested that the gravid Pj continues to supply those sub- stances throughout development. Bottke (1972, 1973: 239) also indicated that the small, nearly yolkless (fertilized) eggs of У. contectus necessarily depend on ‘*, .. large amounts of secondary yolk pro- vided by the albumen gland and glandular celikse nofimtheai(pathialse oviduct! Alyakrinskaya (1969) and Samochwalenko & Stánczykowska (1972) noted that the capsules of V. fasciatus and V. viviparus contain proteins. Finally, Chatterjee & Ghose (1973) reported seasonal accumu- lations and declines of glycogen and lipids in “genital organs” of /. bengalensis (viz. gonad probably only testis, albumen gland and prostate gland) in relation to the breeding season (undescribed). Identification of specific viviparid acces- sory organ secretions and information on their functions such as Goudsmit & Ash- well (1965), Goudsmit & Neufeld (1966) and Goudsmit (1973) provided for Helix pomatia (Linn.), a terrestrial pulmonate, are presently lacking. The site of fertilization in viviparid females has been previously reported to be the seminal receptacle (von Siebold, 1836; Mattox, 1938) and the sperm groove on the floor of the pallial oviduct (Speyer, 1855). However, because sperm were observed stored in the albumen gland as well as present in the oviduct, and because encapsulated, very small F1's were found within the distal arm of the oviduct in all 3 species studied here (most commonly in V. georgianus), it is concluded that fertili- zation and capsule formation occur in the oviduct. Kamaloney (1968) also hypo- thesized the occurrence of fertilization in this site for Bellamya sumatrensis, but con- sidered fertilization in the seminal recep- tacle equally possible. This conclusion also concurs, in part, with that of Leydig (1850) for V. viviparus. Capsule form, with a chalaza or funi- culus (cf. Leydig, 1850: pl. 11, fig. 16), is similar in the 3 species studied here. The opening from the narrow oviduct into the comparatively spacious seminal receptacle is very small, and the chalaza perhaps results from a “pinching”” effect on the acellular capsular envelope as the Fy and peri- vitelline fluid contained therein are forced through that opening. Brood Production Seasonal reproductive patterns have been described for some viviparids, but the avail- able information does not document sufficient possible conspecific variations to permit adequate congeneric or conspecific comparisons at this time. This study, like most others, provides information on the reproductive cycle from observations of just 1 population of each species. Prelimi- nary observations (Vail, unpubl.) on many other populations of C. geniculum and V. georgianus indicate that maximum size of P1's, mean brood size and Е1 birth size, while relatively uniform within a popula- tion, are highly variable among populations. It is not yet known whether these other populations vary in the number of broods produced per year. Environmental factors influencing P1 size, brood size and F 1 birth size are, at best, poorly known. Until these factors are understood, their effects cannot be distinguished from possible generic or species specific traits (if any). Information on the 3 viviparids studied here is pre- sented in Tables 3 and 4 with similar data from the previous studies as a summary of current information rather than a com- parison of genera or species. Such com- parisons should not be attempted until intraspecific variation and environmental influences have been studied. Previous studies on viviparid reproduc- tive cycles emphasized analysis of the P4 population instead of the F7 population as was done here. Birth periods, age classes, life spans and incubation periods were extrapolated from seasonal size frequency histograms based on Р1 shell length. Brood sizes (total number of incubating young) and the size of Е1 individuals were fre- quently obtained from observations made without regard to season. In this study, it VIVIPARID REPRODUCTIVE PATTERNS TABLE 3. Life history features of North American Viviparidae. Data drawn from Chamberlain (1958), Medcof (1940), Van Cleave & Altringer (1937), Van Cleave & Chambers (1935), Van Cleave & Lederer (1932) and the present study. L, shell length; D, shell diameter. 93 Feature Campeloma decisum C. geniculum C. rufum C. tannum Smallest gravid P, 's 15.0 mm L 13.2 mm D 20 mm L 17.5 mm L Fertilization period at least March-May mid-spring- May-June “at all early fall seasons Incubation duration ? 8-10 months 11-12 months all seasons”’ Birth period March-June Feb.-April May (peak)-Sept. “summer” Ф age at 1st birth 2 years ? 2 years 3 years Non-gravid interval individual ? yes ? ? population ? mostly es ? Sex. size dimorphism no males mostly no males no males Life span (6:9) 0:3 years 7:2 0:2-3 years 0:5 years Viviparus Feature Lioplax pilsbryi L. sulculosa contectoides V. georgianus Smallest gravid P, 's 13.8 mm D 10-12 mm L 16-19 mm L 15.0 mm О Fertilization period continuous, peak in spring (?) early summer mid-spring- spring early fall Incubation duration 10-12 months 12 months 8-10 months 10-12 months Birth period Dec.-April May-June Feb.-June Jan.-July Q age at 1st birth ? 2 years 2 years 2 Non-gravid interval individual no ? yes possibly population no no yes no Sex. size dimorphism no yes yes no Life span (6:9) 2:2 1:2 years 1:3 years aie, TABLE 4. Sizes of large North American viviparid P, females, newborn and broods. Data drawn from references cited in Table 3. Species P, length (mm) C. decisum modes: 17-20 (up to 27.3) С. geniculum? modes: 24-26 (up to 34.2) C. rufum modes: 29-31 (up to 40) С. tannum modes: 16-20 (up to 47) L. pilsbryi® modes: 23-24 (up to 27.8) L. sulculosa modes: 16-17 (up to 23.3) V. contectoides modes: 27-29 (up то 40.7) У. georgianus® modes: 32-34 (up to 38.1) Newborn length (mm) Mean no. of young per p,b about 3.0 ? >4.0 4.1-27.7 (up to 4.6) (up to 55) mode: 4.0 3.042.1€ (2.5-6.0) (up to 90) ? 2.0-15.0€ (up to 26) >5.0 3.5-9.1 (up to 6.1) (up to 13) about 2.5 (up to 2.7) (up to 46) about 3.8 ? >1.4 5.2-10.4 (up to 8.2) (up to 25) 2512е5 of newborn judged from those of large Е, 's in pallial oviduct. bBrood sizes shown as range of monthly means. CMeans from different P, age classes. was seen that using shell length as an indi- cation of Py size is inadequate for these perennial animals because the apex can become eroded. It is possible that, due to erosion, 2 P1's could be in the same length-size class but be in different diameter-size classes. Thus, shell diameter, rarely modified by erosion, is a more reli- able index of Pj size. Information on brood production obtained only from 94 VAIL analyses of the Р1 population without con- sideration of the F7 population can be incomplete and misleading. Lacking would be data showing (1) whether P individuals undergo fertilization of a new brood prior to birth of the existing brood or experience a non-gravid interim between consecutive broods, (2) if members of a single brood are born within a brief period of time or sporadically over a longer period of time, or (3) how the brood production cycle of an individual compares to the cycle shown by the population, i.e., an individual might give birth to all her young at one time but it might be several months before all females in the population had given birth. Also, brood sizes and F7 dimensions ob- tained from samples taken without regard to season could be misleading as both change during the course of brood develop- ment (Figs. 8, 12-16). The best documented prior information on North American viviparid reproductive cycles (Tables 3, 4) concerns Campeloma rufum (Van Cleave & Altringer, 1937), C. tannum (Medcof, 1940), С. decisum (Chamberlain, 1958), Lioplax sulculosa (Van Cleave & Chambers, 1935) and Vivi- parus contectoides (Van Cleave & Lederer, 1932). The former 3 studies were based on seasonal (but not consecutive monthly) samples from the same population each, and the latter 2 each combined data from several populations. The findings of the present study resemble those from the aforementioned investigations only for a few features (Table 3, Сатре/ота: fertili- zation period; Lioplax: incubation dura- tion; Viviparus: birth period). All but lack- ing from those prior studies are data on age structure, life spans and seasonal brood production patterns. On the basis of current information the basic pattern of brood production appears to be similar in these 8 species representing 3 genera. In each there is 1 fertilization period and 1 birth period per year and incubation of young lasts nearly 1 year. Differences appear to occur only in the specific time and duration of each event (Table 3). However, information from the prior studies suggests that fertilization and birth periods overlap in individuals as well as in the population. While it is possible that 2 broods are simultaneously incubated, it is also possible that individual females experience a non-gravid interim between broods. Here a non-gravid interim was seen in individual С. geniculum but not in the population; simultaneous incubation of 2 broods was observed in L. pilsbryi and М. georgianus. Van Cleave & Lederer (1932) also reported that newborn male V. con- tectoides (< 4 mm in shell length) from several populations in Illinois and New York possessed a copulatory tentacle. Observations of V. georgianus Р1’5 from Holmes Creek indicated that tentacular asymmetry was not evident in individuals < 11 mm in diameter (newborn: > 7.5 mm). This delayed tentacular asymmetry was also observed in С. geniculum and L. pilsbryi. Sexual size (and age) differences have been reported in L. su/culosa and V. con- tectoides with females attaining a larger size and living 1 year longer than males (Van Cleave & Chambers, 1935; Van Cleave & Lederer, 1932). Sexual size dimorphism does not seem to occur in either L. pilsbryi or V. georgianus, but C. geniculum females are generally larger than the males (non- quantitative personal observations). The occurrence of sexual size dimorphism in C. geniculum is the 1st such record for Campeloma, but only because this is the 1st dioecious population so studied. Viviparid brood size has previously been considered in relation to season, PJ size and age, size of developing F1's and P 4 habitat. Brood size in C. geniculum, L. pilsbryi and V. georgianus increases sea- sonally following the onset of fertilization to a maximum and subsequently decreases as births occur (Figs. 12-14). This seasonal fluctuation is especially noticeable in С. geniculum, and is a consequence of pro- gressive brood development (Figs. 9-11). This is in agreement with those obser- vations by Van Cleave & Altringer (1937), Miroshnitshenko (1958), Stanczykowska (1960) and Samochwalenko & Stanczykowska (1972) on С. rufum, М, viviparus, V. fasciatus and V. fasciatus and V. viviparus, respectively. Brood size has also been related to Р1 size and age. Observations on different- sized С. geniculum and У. georgianus P1's show brood size increased with increasing Р1 size (and age). These observations con- cur with those of Medcof (1940), Miro- shnitshenko (1958), Stanczykowska et al. (1971) and Samochwalenko & Stan- czykowska (1972). In contrast, brood sizes in L. pilsbryi remained relatively constant among Py ’s of different sizes and ages. VIVIPARID REPRODUCTIVE PATTERNS 95 TABLE 5. Approximate diameter (mm) of Сатре/ота geniculum, Lioplax pilsbryi and| Viviparus georgi- anus at birth and maturation plus average and maximum diameter of gravid P, in each population. C. geniculum L. pilsbryi V. georgianus Birth size 23.5 24.5 27.5 Smallest gravid P, 13.2 13.8 15.0 x diameter of gravid P, 's 18.2=1.3 15.7+1.0 26.9+1.6 Largest P, collected 25.6 18.3 30.4 Also, a post-reproductive senescence, as Leon Co., Florida, yielded mean brood reported by Annandale & Sewell (1921), Van Cleave & Lederer (1932) and Van Cleave & Altringer (1937) for /diopoma bengalensis, V. contectoides and C. rufum, respectively, was not observed in the 3 species studied here. Table 2 shows average brood size in relation to Р1 size for each of these species. A composite sample was used in each instance because it was not possible to obtain a large sample containing Р1 in each size class during any 1 month. Campeloma geniculum and L. pilsbryi snails are approximately the same size at birth and both approximately triple their diameter before first becoming gravid. However, C. geniculum snails can grow to be much larger than L. pilsbryi individuals with a 7-fold increase of birth size (com- pared to only a 4-fold increase in the latter species). Viviparus georgianus animals approximately double their diameter between birth and first becoming gravid, and potentially can increase 4-fold in size. Table 5 summarizes these relationships. It is not known whether these relationships are a reflection of inherent differential growth rates, maturation rates, life spans or environmental factors or a combination of these factors. Reports relating brood size to Р1 habitat were not substantiated in this study. Annandale (1924), Prashad (1928), Rohr- bach (1937) and Samochwalenko €: Stán- czykowska (1972) observed that females in lotic populations carried more Fj’s per brood than did females in lentic situations. The 3 populations studied here showed wide variation in brood size even though all were from lotic habitats (C. geniculum and L. pilsbryi being from the same locality). In addition, preliminary analyses of quali- tative samples from lentic populations have provided a further comparison. A collection made in December, 1973 of C. geniculum snails (number of P4’s: 18; mean diameter: 18.0 mm), L. choctawhatchensis individuals (15; 11.0) and М. georgianus snails (13; 24.5) from Lake Talquin, a reservoir in sizes of 80.1, 48.5 and 23.4, respectively. These values are all greater than those reported here for the Chipola River and Holmes Creek congeners. ACKNOWLEDGEMENTS This study was funded, in part, by a Gerald W. Beadel Grant from Tall Timbers Research Station. The author expresses appreciation to William H. Heard of Florida State University for providing assistance and reviewing this manuscript, and to E. V. Komarek, Sr. of Tall Timbers Research Station for providing facilities and promot- ing basic research. LITERATURE CITED ALYAKRINSKAYA, 1. O., 1969, Morphological adaptations to viviparity in Viviparus viviparus (Gastropoda: Prosobranchia). Zoologicheskii Zhurnal, 48: 1608-1613. [In Russian with English summary.] ANKEL, W. E. 1924, Spermatozoen-Dimor- phismus und Befruchtung bei Bythinia tenta- culata L. und Viviparus vivipara L. Sencken- bergiana, 6: 1-12. |. ANKEL, W. E., 1930, Uber das Vorkommen und die Bedeutung Zwittriger Geschlechtszellen bei Prosobranchiern. Biologisches Zentralblatt, 50: 513-532. ANNANDALE, N., 1924, The evolution of the shell-sculpture in fresh-water snails of the family Viviparidae. Proceedings of the Royal Society of London, 96: 60-76. ANNANDALE, N. & SEWELL, R.B.S., 1921, The banded pond-snail of India (Vivipara benga- lensis). Records of the Indian Museum, 22: 215-292, 3 pl. BAKER, Е. C., 1928, The fresh water Mollusca of Wisconsin. Part 1. Gastropoda. Bulletin of the Wisconsin Geological and Natural History Survey, 70: xx + 507 p., 38 pl. ВАВВАТО, С., 1971, Studio sul comportamento stagionale de Viviparus ater (Crist. & Jan) dei Laghi d'lseo e di Gardo. Natura (Societa Italiana di Scienze naturali e Museo civico di Storia naturale, Milano), 62: 65-74. BERRY, A. J., 1974, Reproductive condition in two Malayan freshwater viviparid gastropods. Journal of Zoology, 174: 357-367. 96 VAIL BOTTKE, W., 1972, Zur Morphologie des Ovars von Viviparus contectus (Millet, 1813) (Gas- tropoda: Prosobranchia). 1. Die Follikelzellen. Zeitschrift für Zellforschung und mikro- skopische Anatomie, 133: 103-118. BOTTKE, W., 1973, Zur Ultrastruktur des Ovars von Viviparus contectus (Millet, 1813) (Gas- tropoda: Prosobranchia). Il. Die Odcyten. Zeitschrift für Zellforschung und mikro- skopische Anatomie, 138:239-259. CHAMBERLAIN, N. A., 1958, Life history studies of Campeloma decisum. Nautilus, 72: 22-29. м CHARIN, N., 1926, Uber die Nahrung des Em- bryo von Paludina vivipara. Bulletin de la Societe des Naturalistes de Voroneje, 1: 60-66. [In Russian with German summary]. CHATTERJEE, B. & GHOSE, K. C., 1973, Sea- sonal variation in stored glycogen and lipid in the digestive gland and genital organs of two freshwater prosobranchs. Proceedings of the Malacological Society of London, 40: 407-412. CRABB, E. D., 1929, Egg laying and birth of young in three species of Viviparidae. Nautilus, 42: 125-129. DEMBSKI, W. J., 1968, Histochemische Unter- suchungen über Funktion and Verbleib eu- und oligopyrener Spermien von Viviparus contectus (Millet, 1813), (Gastropoda: Proso- branchia). Zeitschrift für Zellforschung und mikroskopische Anatomie, 80: 150-179. DUNCAN, C. J., 1958, The anatomy and physiology of the reproductive system of the freshwater snail Physa fontinalis (L.). Proceed- ings of the Zoological Society of London, 131: 55-84. FROMMING, E., 1928, Der Vorgang der Geburt bei Viviparus viviparus. Archiv fur Mollusken- kunde, 60: 283-284. FROMMING, E., 1931, Ein Beitrag zur Ber- mehrung der Vivipara vivipara Muller. Blatter fiir Aquarien und Terrarienkunde, 42: 94-97. FROMMING, E., 1940, Beitrag zur Lebensweise unserer Deckelsump fschnecke V. viviparus L. International Revue der gesamten Hydro- biologie und Hydrographie, 40: 346-358. FROMMING, E., 1956, Biologie der mitteleuro- paischen Süsswasserschnecken. Duncan & Humblot, Berlin, 313 p. GALL, J. G., 1961, Centriole replication. А study of spermatogenesis in the snail Viviparus. Journal of Biophysical and Biochemical Cytology, 10: 163-193. GOUDSMIT, E. M., 1973, The role of galactogen in pulmonate snails. Malacological Review, 6: 58-59. GOUDSMIT, Е. М. & ASHWELL, G., 1965, Enzymatic synthesis of galactogen in the snail Helix pomatia. Biochemical and Biophysical Research Communications, 19: 417-422. GOUDSMIT, E. M. & NEUFELD, E. F., 1966, Isolation of GDP-L-galactose from the albumen gland of Helix pomatia. Biochimica et Biophysica Acta, 121: 192-195. HANSON, J., RANDALL, J. T. & BAYLEY, S. T., 1952, The Microstructure of the sperma- tozoa of the snail Viviparus. Experimental Cell Research, 3: 65-78. HUBRICHT, L., 1943, Notes on the sex ratios in Campeloma. Nautilus, 56: 138-139. KAMALONEY, A. C. B., 1968, The Reproductive system and aspects of development in the Malayan freshwater snail, Bellamya suma- trensis (Dunker) (Prosobranchia, Viviparidae). Unpbl. B. Sc. Honours Thesis, School of Bio- logical Sciences, University of Malaya, Kuala Lumpur; 41 text p., 1 pl., 22 fig., 10 tables on unnumbered p. LEYDIG, F., 1850, Ueber Paludina vivipara. Ein Beitrag zur naheren Kenntnis dieses Thieres in Embryologischer, anatomischer und histo- logischer Beziehung. Zeitschrift für wissen- schaftliche Zoologie, 2: 125-197, pl. 11-13. MATTOX, N. T., 1937, Oogenesis of Campeloma rufum, a parthenogenetic snail. Zeitschrift für Zellforschung und mikroskopische Anatomie, 27: 455-464. MATTOX, N. T., 1938, Morphology of Campe- loma rufum, a parthenogenetic snail. Journal of Morphology, 62: 243-261. MEDCOF, J. C., 1940, On the life cycle and other aspects of the snail Campe/oma, in the Speed River. Canadian Journal of Research, 18, Section D: 165-172. MEVES, F., 1903, Ueber oligopyrene und apyrene Spermien und uber ihre Entstehung, nach Beobachtungen an Paludina und Pygaera. Archiv für mikroskopische Anatomie und Entwicklungsgeschichte, 61: 1-84, 8 pl. MIROSHNITSHENKO, A. Z., 1958, Fecundity of the freshwater mollusc Viviparus viviparus L. Zoologicheskii Zhurnal, 37: 1635-1644. [In Russian with English summary]. NISHIWAKI, S., 1964, Phylogenetical study on the type of the dimorphic spermatozoa in Prosobranchia. Science Report of the Tokyo Kyoiku Daigaku, 11(B): 237-275. POLLISTER, A. W., 1939, Centrioles and chro- mosomes in the atypical spermatogenesis of Viviparus. Proceedings of the National Academy of Science, 25: 189-195. POLLISTER, A..W..& РОЦЫЗТЕВ, PR. E. 1943, The relation between centriole and centromere in atypical spermatogenesis of viviparid snails. Annals of the New York Academy of Sciences, 45: 1-48, 5 pl. POPOFF, M., 1907, Eibildung bei Pa/udina vivi- para und Chromiden bei Pa/udina und Helix. Archiv für mikroskopische Anatomie und Entwicklungsgeschichte, 170: 43-129, pl. 4-8. PRASHAD, B., 1928, Recent and fossil Vivi- paridae. A study in distribution, evolution and palaeogeography. Memoirs of the Indian Museum, 8: 153-251, pl. 19. ROHRBACH, F., 1937, Oekologische und mor- Phologische Untersuchungen an Viviparus (Bellamya) capillatus Frauenfeld und Viviparus (Bellamya) unicolor Olivier, unter Berück- sichtigung anderer tropischer Formen und im Hinblick auf phyletische Beziehungen. Archiv für Molluskenkunde, 69: 177-218. ROSSI, L., 1968, Variazioni stagionali della spermatogenesi tipica e atipica in Viviparus ater (Gastropoda: Prosobranchia). Archivio zoologico Italiano, 53: 315-330. RUNHAM, М. М. & LARYEA, A. A., 1968, Studies on the maturation of the reproductive system of Agriolimax reticulatus (Pulmonata: Limacidae). Malacologia, 7: 93-108. SAMOCHWALENKO, T. & STÄNCZYKOWSKA, A., 1972, Fertility differentiation of two VIVIPARID REPRODUCTIVE PATTERNS 97 species of Viviparidae (Viviparus fasciatus Mull. and V. viviparus L.) in some environ- ments. Ekologia Polska, 20: 479-492. [In English]. SIEBOLD, C. T. VON, 1836, Ferner Beobach- tungen über die Spermatozoen der wirbellosen Thiere. Archiv für Anatomie und Physiologie der wissenschaftlichen Medizin, 2: 232-255, pl. 10. SMITH, B. J., 1966, Maturation of the reproduc- tive tract of Arion ater (Pulmonata: Arioni- dae). Malacologia, 4: 325-349. SMITH, B. J., 1967, Correlation between neuro- secretory changes and maturation of the reproductive tract of Arion ater (Stylomma- tophora: Arionidae). Malacologia, 5: 285-298. SPEYER, O. W. C., 1855, Zootomie der Paludina vivipara. Inaugural-Dissertation, Marburg; Fis- _cher, Cassel, 46 p., 2 pl. STANCZYKOWSKA, A., 1960, Beobachtungen über die Gruppierungen von Viviparus fas- ciatus Müll. in dem Weichselarm Kon- federatka. Ekologia Polska, 8: 21-48. [In Polish with German summary]. STANCZYKOWSKA, A, MAGNIN, Е. 8 DUMOUCHEL, A., 1971, Etude de trois populations de Viviparus malleatus ‚ (Вееме) (Gastropoda, Prosobranchia) de la region de Montreal. |. Croissance, fecondite, biomasse, et production annuelle. Candian Journal of Zoology, 49: 1431-1441. VAIL, V. A., 1977, Comparative reproductive anatomy of 3 viviparid gastropods. Mala- cologia, 16: 519-540. VAN CLEAVE, Н. J. € ALTRINGER, D. A., 1937, Studies on the life cycle of Сатре/ота rufum, a fresh-water snail. American Natu- ralist, 71: 167-184. VAN CLEAVE, H. J. & CHAMBERS, R., 1935, Studies on the life history of a snail of the genus Lioplax. American Midland Naturalist, 16: 913-920. VAN CLEAVE, H. J. & LEDERER, L. G., 1932, Studies on the life cycle of the snail Viviparus contectoides. Journal of Morphology, 53: 499-522. WOOD-MASON, J., 1881, Notes on Indian land and fresh-water mollusks. No. 1. On the dis- crimination of the sexes in the genus Pa/udina. Annals and Magazine of Natural History, Ser. 5,8: 85-88. YASUZUMI, G. & TANAKA, H., 1958, Sperma- togenesis in animals as revealed by electron microscopy. VI. Researches on the sperma- tozoon-dimorphism in a pond snail, Cipango- paludina malleata. Journal of Biophysical and Biochemical Cytology, 4: 621-632, pl. 298-311. SA У PAGANAS et > u; oa re ЕР cure en EA FREE Fr Speers. "ИРА ew? AO ich omit" aun arg ву bis © ere Enter crient Acte er Te noie ii vin 7108 rank ИА Vem ith oo Ma et Kopie Soc у OF 30:23 era: ae ier „MH АХАИАТ 8 © IMUSUBAY vA Beier ¡ii piso © CE pio? St no rev era 85/35, VIMDAGOS. AUNAR? RC phen eth meirdenamib-1o0so) outer ICRA. Au VOribol--srestin en (eves) ое ы ie TE ID че ra cb renom gh Anika ob dstrtonemson ‘une guey rene = ene nm! ice de hr tos ren u JANO € roo SKATTSME CUR! A 2 { vu Ta ‘ Peiitercrurne rn nr АНА Pub ORE RC Ra AAI 101 ¿ooo Shuma, 3 АУРА г у 5 v me: 15 „Ah ndash залиты. iggy of eh ‚ eve wusm Of de VIDWIAAR, 5, 1554, РУО erates via Coya fe "DA © de Amore e и e ос vty or Lontk ihrwichis,, See Пр Of =" ‘vaike Ooguby IB) 22781. Zu ar Е В Ua ” FR ' DÍ STER aM. 1935 restricted ! Ур ehe for M now ; NN ern ин 3 4 Process of We 124) + E 4 à +1 р Не; - * mos 26 1853105. во e м Bern ААС FOLLIGTER RE y ua ou e a va à #4 ‘gie ang ir Eté ia N = = ofr MORAS CG “vip $ el Ph à ue á = т u % Е УзА fi } > “pr ch a > i = . -- sas ' cfd ic Pes tig à ь 4 ; > » | 7 9 и , E Hola € . №. № ОТ, As > > в} sa eventcid und 1: 4 ar; $ 6 . y W | к untef verwen 5) se traiter - Forte bin Jin 24 ef | РТ! gs . ri we y 20h de onungem, feet С ei eed) ) Md threat die, x 72-218, жом r p | м. Qu AE tiago coló a ' . > vs at DIE 4 903 in ue т р kg > ttrúpoda: Prossbreichiel в 4 j ao Iteflano, 53: 318330 ANSON, 3, АА | ви м № & LARYEA, AVAL mr Hi Вим e erouturation Gf ба оси 4 : al da vater of Agriolimas vase (Puls étés reh En 78 Ace). Васа а, 7 93 UBRICHT 1457 6 +. Y MMOCIWALENKO. TA Al 2¥ Campilura M у у (472 roid iierntiroon: 0! MALACOLOGIA, 1978, 17(1): 99-109 POSITION OF THE CLASS APLACOPHORA IN THE PHYLUM MOLLUSCA! Amelie H. Scheltema Woods Hole Oceanographic Institution, Woods Hole, Massachusetts 02543, U.S.A. ABSTRACT The Aplacophora are shell-less, vermiform mollusks found from the continental shelf regions of the world to depths of 9,000 m. They are grouped into a single class with 2 subclasses: Chaetodermomorpha (= Caudofoveata) and Neomeniomorpha (= Ventroplicida; Solenogastres sensu Salvini-Plawen), a classification which preserves early nomenclature based on Chaetoderma Lovén, validated by the International Commission on Zoological Nomencla- ture. The name “‘solenogaster’’ is reserved as a common noun like “clam” or “‘snail.”” The Aplacophora have several typical molluscan characters, namely a radula with associated buccal structures; a style sac and mucoid style; a coelom restricted to gonads, a dorsal pericardium, and kidneys; a heart consisting of a ventricle and 2 auricles; laterodorsal- ventral musculature; ventral musculature that bends the body and contained organs dorsally; a dorsal gut; mantle cavity and gills; a vestigial foot which secretes a slime trail; a nervous system of paired ganglionated cords, ladderlike, with pharyngeal ring and buccal ganglia; and finally a development which includes spiral cleavage and a protobranch-like larva. Most of these molluscan characters are not structurally like those of chitons with which they have often been classified; therefore, the Aplacophora are classified separately from the Polyplaco- phora. The 2 aplacophoran taxa resemble each other in their nervous system, coelom, haemocoele, musculature, and shape. Differences in integument between the 2 taxa may be due to reduction in the burrowing Chaetodermomorpha, and in the digestive system due to the obligate coelenterate feeding in the Neomeniomorpha. It is of phylogenetic importance that several characteristic molluscan structures have evolved in the Aplacophora independent of a shell. INTRODUCTION The Aplacophora are worm-shaped mol- lusks surrounded by a cuticle bearing cal- careous spicules; they inhabit the deep ocean basins and continental shelf and slope regions of the world. Observations from more than 380 collections made by the Woods Hole Oceanographic Institution, Oregon State University, Scripps Institution of Oceanography, and Centre National de Tri d’Oceanographie Biologique show that their greatest species radiation has been in the deep sea. They burrow into or creep on mud; some wrap themselves around alcy- Onarians upon which they feed. They are common in the deep sea, occurring in nearly all small-meshed epibenthic dredge hauls (Hessler & Sanders, 1967) and box cores taken from all depths to 9,000 m. In recent years the number of biological sur- veys of the deep sea has increased, and the discovery of numerous new aplacophoran species has rekindled interest in this group. Although most species are still to be des- cribed, some recent taxonomic works are those of Salvini-Plawen (1972, biblio- graphy, for his papers), Schwabl (1963) and Scheltema (1976). CLASSIFICATION The 2 classifications given in Table 1 are currently in use for living mollusks. Both classifications retain the 2 distinct taxa that have been recognized since the 19th cen- tury: one taxon (Neomeniomorpha = Ventroplicida = Solenogastres sensu Salvini- Plawen) is distinguished by a ventral groove containing a narrow foot, the other (Chae- todermomorpha = Caudofoveata) lacks a ventral groove and has an oral shield and a pair of ctenidia. It should be stressed that there has been no modification of member- ship in the 2 aplacophoran taxa in either of 1Contribution No. 3763 from the Woods Hole Oceanographic Institution. (99) 100 SCHELTEMA the classifications; only the hierarchical rank has been shifted, and in the second, names have been changed. The first classification is recommended for 3 reasons: 1) it conserves nomenclature in use since the early 1900's; 2) it separates the Aplacophora and Polyplacophora; and 3) it gives a hierarchical ranking to the 2 aplacophoran taxa that fits well with the arrangement of equally similar taxa т other molluscan classes. NOMENCLATURE Nomenclatural changes in the Aplaco- phora started in the 1930's with the belief that the genus Chaetoderma Lovén 1844 was a junior homonym of Chaetoderma Swainson 1839 (Pisces). The next available name was Crystallophrisson Mobius 1875, an orthographic horror variously spelled in the literature. Heppell (1963), who showed that Swainson used multiple spellings, re- quested and obtained from the Interna- tional Commission on Zoological Nomen- clature (ICZN) validation of the genera Chaetoderma Loven 1844 and Chaeto- dermis Swainson 1839 (Opinion 764, ICZN, 1966). Previous to the ICZN ruling, Boettger (1956), wishing to avoid an ordinal name based on "Crystallophrisson,'' used Caudo- foveata to replace Chaetodermatoidea; he also changed the name Neomeniida to Ventroplicida. As Chaetoderma Loven is a valid name, there is historical reason for preserving Pelseneer’s names Chaetoder- momorpha and Neomeniomorpha for the 2 major taxa of Aplacophora. The use of Solenogastres Gegenbaur 1878 for the Neomeniomorpha alone is particularly con- fusing, as Gegenbaur included the only 2 genera then known, Chaetoderma and Neomenia, under this name, which appears throughout the literature as synonymous with Aplacophora. “Зо!еподазег”” is best used with lower case and anglicized spel- ling, a term equivalent to “clam’’ or “snail.” APLACOPHORA AS MOLLUSKS The molluscan affinities of the Aplaco- phora, first noticed in the mid-1870’s, have long been discussed (see Hyman, 1967: 2-3, 68-70, for a review). The Aplacophora have been placed back and forth within and outside the phylum Mollusca for nearly a century; when considered true mollusks, it has usually been because of supposed simi- larities to chitons. Even Hyman, who ге- cognized the chitons and aplacophorans as separate classes, stated that such similarities “justify the inclusion of solenogasters in Mollusca’’ (1967: 69), and Fretter € Graham considered that the solenogasters are “undoubtedly related in some way to the chitons’’ (1962:8), although they placed them outside the Mollusca. The latter re- turned them to Mollusca in 1976 (p. 548) without discussion. Salvini-Plawen (Table 1) has grouped chitons and aplacophorans as a separate subphylum. The chitons and apla- cophorans together have been known as Amphineura, Aculifera, or Isopleura. At the time that Hoffman (1949) care- fully described the integument of the Apla- cophora, they had not been considered mollusks by Thiele (e.g., 1925: 12-14), one of the strongest voices of the preceding decades, because they seemed to him to lack the important molluscan characters of “shell, mantle, foot, and nephridia; further, in other mollusks the gonoducts do not issue from the pericardium’ (as summa- rized by Hyman, 1967: 68). Hoffman com- pared chiton and aplacophoran integuments in both mantle and foot, and by certain homologies between them, brought the 2 aplacophoran taxa back within the concept of mollusks. Two doubtful homologies that he proposed are (1) that the slime tracts of chitons equal the “shell glands” of apla- cophorans, a homology that calls for an impossible reorientation of gills in the apla- cophorans as noted by Hoffman himself (1949: 407), and (2) that the gland cells of the chiton foot equal the glands of the chaetodermomorph oral shield. However, he concluded that “In spite of my view that the Neomenioidea and Chaetoder- matida were thus early separated from each other and that their common stem was chiton-like in respect to the mantle and foot, | do not feel justified in regarding the Neomenioidea, Chaetodermatida, and chi- tons as 3 equivalent groups in the class Amphineura. The Neomenioidea and Chae- todermatida resemble each other exactly in more respects (e.g., in regard to the nerv- ous system and coelom) than either re- sembles the chitons” (1949: 424, here translated). Boettger (1956) elaborated on Hoffman APLACOPHORAN RELATIONSHIPS 101 TABLE 1. Classification of the living Mollusca. ! | Phylum Mollusca Class Monoplacophora Wenz 1940 Class Aplacophora von Ihering 1876 [= Solenogastres Gegenbaur 1878] Subclass Chaetodermomorpha Pelseneer 1906 Subclass Neomeniomorpha Pelseneer 1906 [= Ventro- plicida Boettger 1956] Class Polyplacophora [ex Polyplacophores] de Blainville 1816 (Classes Gastropoda, Pelecypoda, Scaphopoda, Cephalopoda) 11 Phylum Mollusca Subphylum Aculifera Hatschek 1891 [= Amphineura von Ihering 1876] Class Caudofoveata Boettger 1956 Class Solenogastres Gegenbaur 1878 [partim] Class Placophora von Ihering 1876 Subphylum Conchifera Class Tryblidiida Wenz 1939 [= Monoplacophora] (Classes Gastropoda, Bivalvia, Scaphopoda, Cephalopoda) ТА recent classification including the extinct mollusks is in Runnegar & Pojeta (1974), who place the Aplacophora and Polyplacophora by themselves in 2 separate subphyla. 2From Salvini-Plawen (1972). in a theoretical paper on early mollusks, and Salvini-Plawen (1967: 399) stated that from Hoffman’s work ‘‘der Placophoran- Verwandtschaft der aplacophoran Gruppen nicht mehr zu Zweifeln ist.” The Aplacophora bear many preemi- nently molluscan characters other than an integument that may or may not be homol- ogous to that of the chitons (see below). Despite their specialized vermiform shape, they have many archaic, or conservative, molluscan characters. A typical radula exists in several genera with a radular membrane issuing from a radula sac and bearing rows of teeth formed by odontoblasts (Figs. 1A, 2; Scheltema, 1972; Hyman, 1967, fig. 15). Although much of the published work indi- cates no more than serrate cuticularized pharyngeal epithelium, critical examination of some Chaetodermomorpha has shown the existence of a subradular as well as a radular membrane (Figs. 1A, 2; also Scheltema, 1972). However, Salvini-Plawen (1972: 240) considered that there is only a single basal cuticle. There are paired bol- sters, which are chondroid in some genera (Fig. 3B). There is no docoglossate, chiton- like dentition present in any Aplacophoran genus so far described. A style sac occurs in the Chaetoder- momorpha; a style in the form of a mucoid rod and gastric shield are present as well in the family Chaetodermatidae (Fig. 1) and in the genus Limifossor. (A description of this uniquely molluscan character is to be made more fully elsewhere.) There is no rod in the chiton style sac (Fretter, 1937). The Chaetodermomorpha also have a blind digestive gland, which empties into the posterior stomach (Fig. 1C; Wirén, 1892). The Neomeniomorpha, many of which are obligate predators on coelenterates (Salvini-Plawen, 1967), have a very differ- ent and probably specialized mid-gut sys- tem consisting of a wide tube thrown into folds and lack a separate digestive gland. There are no chiton-like esophageal glands (“sugar-glands”*) in the Aplacophora, а|- though salivary glands are found. The aplacophoran coelom, as in all mol- lusks is restricted to paired gonads, pericar- dium, kidneys, and the ducts therefrom; unlike other mollusks, the gonads empty directly into the pericardial cavity and gametes pass out of the pericardial cavity through coelomoducts (Fig. 4A). The heart and pericardial cavity are dorsal and mol- luscan in organization (Scheltema, 1973; and others, see Hyman, 1967). They may indicate that the early molluscan condition was a pair of ventricles and a paired peri- cardium, as found in WNeopilina, for the heart is bilobed during diastole and there are 2 large V-shaped lateral extensions of the pericardium (Chaetodermatidae: Fig. SCHELTEMA 102 APLACOPHORAN RELATIONSHIPS 103 > BOLA BOO a ( Y RE CON ana = om СХ 1.0 FIG. 2. Radula ribbon (heavy stippling) and sub- radular membrane (light stippling) of the limpet Acmaea testudinalis (A) and Prochaetoderma sp. (B; cf. Fig. 1). The position of the teeth in Acmaea is indicated; one transverse row is drawn. Scales in mm. AC; also Scheltema, 1973, figs. 1, 2). The chiton pericardial cavity, also large, does not receive gametes from the gonads. Histological evidence suggests functional kidneys in the Chaetodermatidae; the cells of the C-shaped coelomoducts, emptying the pericardium to the outside, are similar to the kidney cells in the protobranch Nucula (Fig. 3A, 3C). There are no experi- mental data on the function of the apla- cophoran coelomoducts. Laterodorsal-ventral musculature, ех- pressed serially in the Neomeniomorpha and in restricted body areas in the burrow- ing Chaetodermomorpha, may preserve the condition leading to reduction in pedal retractors of shelled mollusks including chi- tons (Salvini-Plawen, 1969). Ventral longi- tudinal muscles, found in both chitons and и aplacophorans, produce a dorsal bend (Fig. 5A-D). In the burrowing chaetoderms these muscles are weak (Fig. 4B). The digestive system is dorsally placed with a ventral mouth and anus (Figs. 1A, 3A, 4B). The dorsal bend appears during development in the late embryos of 2 species of neomenio- morphs (Pruvot, 1890; Thompson, 1960). The molluscan head-foot is not in great evidence in the vermiform Aplacophora. However, vestiges remain in both taxa. The Neomeniomorpha produce a sticky slime track along which they creep by the ciliary action of a greatly reduced foot; the head is free of the substrate and moves about by hydrostatic action (personal observation). The Chaetodermomorpha retain a ventral (= pedal)| sinus (Fig. 4B), and the genus Scutopus still has an external indication of a lost foot (Salvini-Plawen, 1972). In both taxa there 15 “а malleable ‘haemoskeleton’ that can be manipulated by the muscles of the body wall” as described by Morton (1967: 17) for the ancestral condition. A mantle is present in the sense that the epithelium which covers the outer body secretes a mucoid substance and calcium carbonate and forms a fold over a mantle cavity (Fig. 4A, 4C). Histochemical staining indicates that the cuticle of Proneomenia is composed of a glucoprotein complex ‘‘ten- tatively equated with an early mucoid stage in the evolution of the molluscan shell” but it is not the same as chiton cuticle, which is more specialized (Beedham & Trueman, 1968: 443). Both Hoffman (1949) and Salvini-Plawen (1972) con- sidered the spicular part of the chiton cuticle to be homologous to aplacophoran cuticle; however, Stasek (1972: 18) and Stasek & McWilliams (1973) pointed out that the homologous parts may be the spiculose integument of the dorsally turned mantle of the chitons and the nonspiculose mantle integument bordering the foot-fold FIG. 1. A: Sagittal section, anterior of Prochaetoderma sp. The most recently formed radula tooth lies adjacent to the odontoblasts at the blind end of the radula sac. The empty space in the head is that part of the pharynx lying between the paired ‘‘jaws’’ that protect and open the ventral mouth (not shown in section). B: Cross-section through mucoid style lying at anterior end of the style sac in Chaetoderma nitidulum. Two digestive cell types line the digestive gland lumen (“KGrnerzellen’’ and ‘’Keulenzellen’’; Wirén, 1892). С: Cross-section of anterior end of 100 ит mucoid style projecting from style sac (not shown) and gastric shield in Falcidens caudatus; the digestive gland opens into the stomach. Scales in mm. d digestive gland dl digestive gland lumen gs gastric shield j “jaws” m mucoid style od odontoblasts r radula sac sr subradular membrane st stomach 104 SCHELTEMA FIG. 3. A: Coelomoducts of Chaetoderma nitidulum show great histological similarity to the kidneys of the protobranch Nucula annulata (C). Both upper and lower limbs of the C-shaped coelomoducts of Chaeto- derma are evident, as well as the ventral bend of the intestine. B: Paired odontophore cartilages lie on either side of the radula sac with radula teeth in Prochaetoderma sp. C: Kidneys of the protobranch Nucula annulata. Scales in mm. cd coelomoducts i intestine k kidney oc odontophore cartilage and lying within the ventral groove of the Neomeniomorpha. The shape of the aplacophorans has greatly affected the extent of the mantle cavity, but typical molluscan characters are still evident in the posterior cavity into which the anus and coelomoducts empty p pericardial cavity ph pharynx r radula sac and which contains paired gills in the Chae- todermomorpha. The Neomeniomorpha also retain the mantle cavity in the form of a groove on either side of the vestigial foot. The nervous system is composed of paired cerebral ganglia and paired lateral and ventral cords with cross commissures, a APLACOPHORAN RELATIONSHIPS 105 g uu au СОК RL 0.1 FIG. 4. A: Sagittal section through posterior region of Falcidens caudatus. The gonad empties into the pericardial cavity, which is emptied by C-shaped coelomoducts (connection not shown). The intestine, above which lies a suprarectal commissure, bends ventrally to empty into the mantle cavity between a pair of gills. Muscles are not shown. B: Cross-section through the posterior half of a Chaetoderma nitidulum juvenile. The gonads, fused in adults, are still paired; the intestine is dorsal. There are 2 types of cell in the digestive gland (cf. Fig. 1B). The ventral sinus is probably homologous to the pedal sinus; a slight thickening of the muscle wall on either side is weakly expressed ventral longitudinal bands causing dorsal bending (Fig. 5). С: Dorsal frontal section through the posterior end of Falcidens caudatus. Both the ventricle in diastole and the pericardial cavity are bilobed. Scales in mm. a aorta g gonad 5 gill sinus au auricle gr gill retractor sc suprarectal commissure € mantle cavity i intestine v ventricle cd coelomoducts | longitudinal bands vs ventral sinus d digestive gland p pericardial cavity 106 5 0.0 SCHELTEMA O => = > S IN у) AL yyy a ) 7 Wi / Luly sn ur, S à \ БА ) ud Pay = <<- < Sn E Esla PS << = <= me Lia < PIS, HH < <<< < < DV. 1.0 Ae — FIG. 5. Flexed body in A: Neomenia herwigi Kaiser, 1976, a giant neomeniomorph from 130 m off the Falkland Islands (Kaiser, 1976); B: Chaetoderma sp. from 67 m S of Woods Hole, Massachusetts; C: Prochaetoderma sp. from 1600 m S of Woods Hole; D: Falcidens sp. from 650 m off Cape Hatteras, North Carolina. Scales in mm. primitive ladderlike plan as in chitons and Monoplacophora, but it is ganglionated rather than medullary (see review in Bul- lock & Horridge, 1965). The differences in the 2 aplacophoran taxa are given in Sal- vini-Plawen (1972). There is a typical mol- luscan buccal innervation and, like chitons, a suprarectal commissure. Thompson (1960), who observed the development of the late egg, early embryo and settlement stages of Neomenia, has published the most comprehensive account of the later development of an aplacopho- ran, which is remarkably like that of proto- branchs and scaphopods and not at all like that of chitons. Baba (1951) described and APLACOPHORAN RELATIONSHIPS 107 illustrated early spiral cleavage for a neo- meniomorph. DISCUSSION Aplacophora are true mollusks by the present definitions of that phylum, a mem- ber of which must express one or more of several diagnostic traits. No single key char- acter defines the phylum (Stasek, 1972). It is unfortunate that the Aplacophora have been considered mollusks chiefly by their superficial resemblance to chitons, for there are few shared characters between them except their archaic nervous systems (the “Amphineura’’) and primitive mucoidal mantles with spicules (‘‘Aculifera’’). Other proposed structural homologies seem highly improbable and have been influenced by the idea that aplacophorans really cannot be considered mollusks unless (1) an affin- ity to chitons can be shown (Hoffman, 1949) or (2) their position as a stem form of the Mollusca can be proven (Salvini- Plawen, 1969, 1972). From the structures briefly described in the previous section, the Aplacophora can certainly be included among the Mollusca on the basis of characters shared with other molluscan classes. To place the 2 taxa as subclasses in a single class makes a classifi- cation that gives a hierarchical ranking con- sistent with that of other equally similar molluscan subclasses. The differences be- tween the Chaetodermomorpha and Neo- meniomorpha in integument, form of mantle cavity, and digestive system set them apart from each other, but no more so than are prosobranch and opisthobranch gastropods. The differences in nervous systems stressed by Salvini-Plawen (1972: 320, fig. 50) are not as great as the differ- ences among the prosobranch gastropods with their separated or fused pleural and cerebral ganglia. and their ladderlike or unconnected pedal cords (Fretter € Graham, 1962). In fact, it is exactly the similarity in nervous systems in the 2 apla- cophoran taxa—ladderlike paired lateral and ventral cords which are ganglionated, and a suprarectal commissure (Fig. 4A, 4C)—that shows the close relationship between them; moreover, there is a dorsoterminal sensory organ in both. Equally important indications of similar- ity are the connection between gonads, pericardium, and coelomoducts in both taxa and their vermiform shape, conditions unique among the mollusks. The lack of a foot groove in the Chaetodermomorpha is most parsimoniously explained by loss owing to their burrowing habit. It is not clear that the simpler, thinner integument of the Chaetodermomorpha is the more archaic of the 2 taxa (Hoffman, 1949); it could just as well be that their cuticle is a secondarily derived state related to burrow- ing. The basis for Salvini-Plawen’s 2 apla- cophoran classes is the rigid application of Hennig’s cladistic methods. The method itself has several serious failings (see Szalay, 1977, for a critique); one is that it pre- cludes finding new phylogenetic relation- ships from the fossil record. Thus, the newly recognized fossil molluscan class Rostroconchia (Pojeta & Runnegar, 1976), probably ancestral to Pelecypoda and de- scendant from Monoplacophora, can have no place in a predetermined phylogeny such as devised by Salvini-Plawen (1968, 1969, 1972). Hennig’s method is inappropriate as applied to such a large, diverse group as the Mollusca with an unknown number of ex- tinct and unrecognized ancestral forms. It has led Salvini-Plawen (1972) to propose several evolutionary improbabilities. Con- sider 2 examples. First, according to Salvini-Plawen (1972: 284, fig. 35), the molluscan coelom was first (hypothetically) a pair of gonads and a small ductless pericardium. Then the next evolutionary steps were a (hypothetical) joining of gonad to pericardium and devel- opment of a pair of coelomoducts empty- ing the pericardium. Although this is the condition which presently exists in all Aplacophora except Phyllomenia (Salvini- Plawen, 1970), it is said to be secondarily derived after a (hypothetical) separation of gonads and pericardium, each with paired ducts, which gave rise to the condition found in all other living mollusks; replica- tion of the coelom then took place in the Monoplacophora. Only in the aplacophoran taxa are the gonads considered again to open into the pericardium in 2 separate evolutionary events. The original state of the molluscan coe- lom is conjectural, and its organogenesis is unknown in the Aplacophora and Mono- placophora. However, the large pericardium of the Monoplacophora, Aplacophora, and Polyplacophora, the bilaterality of all coe- 108 SCHELTEMA lomic spaces in the former 2, and the paired monoplacophoran ventricles, seem to argue against an original small, single coe- lomic widening simply to protect the heart. The matter of where the gonads emptied originally in the evolutionary history of the mollusks is left open; it may be that the unique gonad-pericardium connection of the Aplacophora is a derived state arising from their vermiform shape. In that case, Phyllomenia would be the 1 genus retaining a more ancestral condition. The second example of an improbable evolutionary event proposed by Salvini- Plawen (1972) is that the oral shield of the Chaetodermomorpha is a remnant of the molluscan foot. This singular homology forms the basis for making 2 classes of the Aplacophora and relates the occurrence of mucous gland cells that open beside the oral shield in Chaetoderma to those along the foot groove in Proneomenia (Hoffman, 1949). The chaetoderms are burrowing forms, moving headfirst by hydrostatic ac- tion. Gland cells are ubiquitous in the outer epithelium throughout the mollusks and appear wherever lubrication is needed functionally. The specialized innervation (Salvini-Plawen, 1972) and cuticularization of the chaetoderm oral shield seem to discount that it is part of an original gliding sole; one should not argue general- ized gland cells beside the oral shield and at the same time specialized innervation to It. The phylogenetic position of the Apla- cophora is not easy to ascertain. There is general agreement that the group is primi- tive and probably very old geologically. There must have been innumerable trials at calcification of a mucoid’ integument among precambrian forms for which we now have no direct evidence. The Apla- cophora were the only forms to survive that did not produce a point—or points—of calcification from which a shell could grow; each calcareous spicule is produced by a single cell (Hoffman, 1949). However, the aplacophoran integument with its spicules, although primitive, should not be con- sidered identical with the original mollus- can integument, for it, too, must be the result of selection.2 The most interesting phylogenetic infor- mation to be obtained from the Aplacoph- ora is that several characteristic molluscan structures have evolved independent of a shell. The notion that the adaptive func- tional possibilities of the mantle cavity have led to great diversity in the mollusks is upheld, for in becoming worms, apla- cophorans have not exploited this structure and retain quite a uniform morphological pattern. However, their unique specializa- tion of a worm shape has been well adapted to the deep sea benthos. ACKNOWLEDGEMENTS The conclusions in this paper are based on studies made of organisms dredged and sorted under National Science Foundation grants GB 6027X, GA 31105 and GA 36554; the studies were initiated under a grant from the Radcliffe Institute. | thank Rudolf Scheltema, Kristian Fauchauld, Judith Grassle, and Alison Stone Ament for carefully reading the manuscript. Peter Riser and Patricia Morse kindly provided the opportunity to observe a living neo- meniomorph on a recent visit to the North- eastern University laboratory at Nahant, Massachusetts. LITERATURE CITED BABA, K., 1951, General sketches of the devel- opment in a solenogastre, Epimenia verrucosa (Nierstrasz). Miscellaneous Reports of the Re- search Institute for Natural Resources (Japan), nos. 19-21: 38-46. BEEDHAM, G. E. & TRUEMAN, E. R., 1968, The cuticle of the Aplacophora and its evolu- tionary significance in the Mollusca. Journal of Zoology, 154: 443-451. BOETTGER, С. R., 1956, Beiträge zur System- анк der Urmollusken (Amphineura). Ver- handlungen der deutschen zoologischen Gesell- schaft, 1955, Zoologischer Anzeiger, Suppl. 19: 223-256. BULLOCK, T. H. & HORRIDGE, G. A., 1965, Structure and function in the nervous systems of invertebrates. Il. Mollusca: Amphineura and Monoplacophora. Freeman, San Francisco and London, p. 1274-1281. FRETTER, V., 1937, The structure and function of the alimentary canal of some species of Polyplacophora (Mollusca). Transactions of the Royal Society of Edinburgh, 59: 119-164. 2A recent comparative account on the integument by Rieger & Sterrer, not available to me at this writing, stated that “spicular skeletons . . . are, in some cases (Turbellaria and Mollusca), possibly relics indicating phylogenetic relationship.” Biological Abstracts, 62, No. 15000: RIEGER, В. М. & STERRER, W. 1975, New spicular skeletons in Turbellaria, and the occurrence of spicules in marine meiofauna. Il. Zeitschrift für die zoologische Systematik und Evolutionsforschung, 13: 249-278.) APLACOPHORAN RELATIONSHIPS 109 FRETTER, V. & GRAHAM, A., 1962, British Prosobranch Molluscs. Ray Society, London. 755 p. FRETTER, V. € GRAHAM, A., 1976, A Func- tional anatomy of invertebrates. Academic Press, London, New York & San Francisco. 589 p. HEPPELL, D., 1963, Chaetoderma Lovén, 1844 (Mollusca), and Chaetodermis Swainson, 1839 (Pisces): proposed addition to the Official List of Generic Names. Bulletin of Zoological Nomenclature, 20: 429-431. HESSPER ТВ: Re & SANDERS, He Ie, 1967; Faunal diversity in the deep-sea. Deep-Sea Research, 14: 65-78. HOFFMAN, S., 1949, Studien Uber das Integu- ment der Solenogastren nebst Bemerkungen Uber die Verwandtschaft zwischen den So- lenogastren und Placophoren. Zoologiska Bidrag fran Uppsala, 27: 293-427. HYMAN, 1. H., 1967, The Invertebrates. VI. Mollusca. McGraw Hill, New York. 1: 1-70. KAISER, P., 1976, Neomenia herwigi sp. n., ein bemerkenswerter Vertreter der Solenogastren (Mollusca, Aculifera) aus argentinischen Schelfgewassern. Mitteilungen aus dem Ham- burgischen zoologischen Museum und Institut, 73: 57-62. MORTON, J. E., 1967, Molluscs. Ed. 4, Hutchin- son, London, 244 p. POJETA, J., Jr. & RUNNEGAR, B., 1976, The paleontology of rostroconch mollusks and the early history of the phylum Mollusca. United States Geological Survey Professional Paper 968: 84 p. PRUVOT, G., 1890, Sur le développement d’une solénogastre. Comptes Rendus hebdomadaires des Séances de l’Académie des Sciences [Paris], 111: 689-692. RUNNEGAR, В. & POJETA, J., Jr., 1974, Mol- luscan phylogeny: the paleontological view- point. Science, 186: 311-317. = SALVINI-PLAWEN, 1. VON, 1967, Uber die Beziehungen zwischen den Merkmalen von Standort, Nahrung und Verdauungstrakt bei Solenogastres (Aculifera, Aplacophora). Zeit- schrift für Morphologie und Okologie der Tiere, 59: 318-340. SALVINI-PLAWEN, L. VON, 1968, Beiträge zur Systematik der niederen Mollusken. Proceed- ings of Symposium on Mollusca, Marine Bio- logical Association of India, Symp. Ser., 3(1): 248-256. SALVINI-PLAWEN, L. VON, 1969, Solenogastres und Caudofoveata (Mollusca, Aculifera): Organisation und phylogenetische Bedeutung. Malacologia, 9: 191-216. SALVINI-PLAWEN, L. VON, 1970, Phyllomenia austrina, ein phylogenetisch bedeutsamer Solenogaster (Mollusca, Aculifera). Zeitschrift für zoologisches Systematik und Evolutions- forschung, 8: 297-309. SALVINI-PLAWEN, L. VON, 1972, Zur Morphol- ogie und Phylogenie der Mollusken; die Beziehungen der Caudofoveata und der Solen- ogastres als Aculifera, als Mollusca, und als Spiralia. Zeitschrift für wissenschaftliche Zoo- logie, 184: 205-394. SCHELTEMA, A. H., 1972, The radula of the Chaetodermatidae (Mollusca, Aplacophora). Zeitschrift für Morphologie der Tiere, 72: 361-370. SCHELTEMA, A. H., 1973, Heart, pericardium, coelomoduct openings, and juvenile gonad in Chaetoderma nitidulum and Falcidens cau- datus (Mollusca, Aplacophora). Zeitschrift für Morphologie der Tiere, 76: 97-107. SCHELTEMA, A. H., 1976, Two new species of Chaetoderma from off West Africa. Journal of Molluscan Studies, 42: 223-234. SCHWABL, M., 1963, Solenogaster mollusks from southern California. Pacific Science, 17: 261-281. STASEK, C. R., 1972, The molluscan framework. т М. FLORKIN & В. T. SCHEER (eds.), Chemical Zoology, VII. Mollusca, р. 1-44. STASEK, С. В. & MCWILLIAMS, W. В., 1973, The comparative morphology and evolution of the molluscan mantle edge. Veliger, 16: 1-19. SZALAY, F. S., 1977, Ancestors, descendants, sister groups and testing of phylogenetic hy- potheses. Systematic Zoology, 26: 12-18. THIELE, J., 1925, Solenogastres. п KUKEN- THAL & KRUMBACH (eds.), Handbuch der Zoologie, 5(1): 1-14. THOMPSON, T. E., 1960, The development of Neomenia carinata Tullberg (Mollusca, Apla- cophora). Proceedings of the Royal Society of Landon, ser. B, 153: 263-278. WIREN, A., 1892, Studien Uber die Solen- ogastres. I. Monographie des Chaetoderma nitidulum Lovén. Konglige Svenska Veten- skaps-Akademiens Handlingar, 24(12): 66 p., 7 pl. 3 exe anne a COmcipet.aot :BET мо org Eater TAB НОАзвАМ ЗЫ Ва Я ohh en ‚ats (peer Фет: A TA” bests da ati И и ox) Shr AA gay. BNE» OPT ea gree а Ku EUR | avion! мо \ tee O ds pat STA ЗА“ АМТ fanthe A toria rod ain el Moca! anoet ть AD: 5 ES 5h ae a) o Rk. ui A NH, ee 2% я ae PORTE Коуч na гк as TU = сы АЯ MEA BA hte ASMA TE His HARO Мм oo iy agian’ iain ль. ‚Я. u. УМ { ом SHRM \# te ПА ‘in: MASA e MIC AAA ОТ. 341 Puede dis Berri ati . ЙЕ г <> wu MENU уе a | 4 weuyro и TUI? ONG 2 "7 or Гы aan > ur o «ОХ D > ot pret a u, | = ай { HARAN ‘ 1 и JA Г mis) oe = ЗА 13) 0 Vrai Mi > Ste Nay SHE ets O "E у ict 4 $ < CU 123) INTER ‘ it a $ * « A сих Fs — у von >. LE} | ri УЕ PR IAA 5 tó Six Vire mare wien 2 mb astas slertozibnewtaV chwolons 00 1. b ARE iiss (AAA he eatevésrevnl MV. eet AKT JAY deat nia. ese alneme No ite tar Sir EY qmtssiresyiey post Pc jos goal, Quer; PA 744 о visité, naine pres ATS a > EA ауте ne A h И ob ны are a FOBIA Era A ar Las eye ¿ET LACS NT u; HAN a en En) da LAC ee dy Oh" re Née obaerye so e ТР anoto с PST 1 Tiss ea ИАН, q ? a Sine 2. 3 $80 + “ttt A ATROS 6 а JADE alegar synegotydd A RARE HE AA che су J ИЗМАЗАЧИЮИ à = ме т | рые АУ: rra вип: karte Bl А green rot Mir tune такс АУМ ЛОХ GELBE we Dé AO Td EMAL ЧЕМ ре ety nll g Led un VE pr à itd Née mater D TT LU wlohe fa JE N > Г ; м a у $ wis na bit 4 . A 15 id EH ‘ e MALACOLOGIA, 1978, 17(1): 111-115 GENETIC STUDIES ОМ B/OMPHALARIA STRAMINEA: OCCURRENCE OF A FOURTH ALLELE OF A GENE DETERMINING PIGMENTATION VARIATIONS Charles S. Richards Laboratory of Parasitic Diseases, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, Maryland 20014, U.S.A. ABSTRACT Three alleles of a gene regulating pigmentation in Biomphalaria straminea have been described. Studies on their relationship to a 4th allele of the same gene, determining a phenotype in which black spots occur in the mantle but body and eyes lack black pigment, are reported. The spotted-mantle allele is recessive to wildtype and blackeye alleles, domi- nant over albino. Occurrence of 3 alleles of a gene deter- mining pigmentation variations in Bio- mphalaria glabrata (Say) was reported by Richards (1967). These variations, which show simple Mendelian inheritance, have provided excellent markers for genetic studies. Three similar pigmentation pheno- types (black wildtype pigmentation, black- eye, and albino) also occur in Biomphalaria straminea (Dunker) determined by 3 alleles (Richards, 1972, 1975). This has facilitated studies on variations in susceptibility to infection with Schistosoma mansoni in B. straminea. A 4th pigmentation phenotype, lacking black pigment in foot and eyes but with black mantle spots was observed in a colony of В. straminea (Richards, 1972). Studies on the genetics of this phenotype are here described. MATERIALS AND METHODS In 1971 a shipment of В. straminea from Sete Lagoas, Minas Gerais, Brasil, was received from Dr. Lobato Paraense. This sample of about 30 surviving snails in- cluded both albino and blackeye pigmenta- tion phenotypes (Richards, 1975). After isolating some of the snails of each pigment type the remainder were separated into an albino and a blackeye population in two jars (Fig. 1). When the albino colony was examined several months later, it was noted that some of the descendants of the origi- nal population had black pigmented mantle spots, although as in the albinos the eyes and body lacked black pigment. Thirteen (only 3 included in Fig. 1) of these spotted mantle snails were isolated as juveniles and observed through several generations of selection and selfing, and controlled mating experiments were рег- formed. Snails from which offspring by selfing had previously been obtained were mated for 1 week and then reisolated. The 3 pigment alleles similar to those in B. glabrata are: (C) wildtype with black- pigmented body, eyes, mantle collar, and mantle spots; (cB) blackeye with black- pigmented eyes and mantle spots but lack- ing black body or mantle collar pigment; and (c) albino with no black pigment. The C is dominant, cB recessive to C but domi- nant over c, and c recessive. The 4th allele in В. straminea is designated (cM), with black mantle spots but lacking black body, eyes, or mantle collar pigment (Fig. 2). RESULTS Albino and blackeye stocks established by isolation of snails from the original sample in 1971 and selected for homozy- gosity have bred true for these phenotypes for 5 years. A wildtype black-pigmented laboratory colony of B. straminea at The National Institutes of Health has bred true for this phenotype during 15 years of (111) 112 RICHARDS 30 B.STRAMINEA FROM BRASIL ALBINOS AND BLACKEYES REMAINDER IN 2 JARS 15 ISOLATED ALBINOS BLACKEYES MIX TURE a ЕЯ (> > BR à SPOTTED MANTLE С М 6 G cBcB 37.I1SO FAME u SELF SELFING Se ee ОВ o AN E eo velas Esta cc (FIG. 3) CcM CcM cMcM cMcM | | | ; cM cM 4C:1cM 17 С: 5 СМ 5C:9cM LIRE + Er Cr CcM ad cMcM y \ cMcM 3C:ıcM 2C:6cM FIG. 1. Origin of spotted mantle Biomphalaria straminea. Many of the snails isolated and observed are omitted from the diagram. Phenotypes are depicted as in Fig. 2. BIOMPHALARIA PIGMENTATION VARIATIONS 113 ALBINO PHENOTYPES Cs GENOTYPES сс cMcM \ cc Se a \ 4 cMcM к cMcM,CcN SPOTTED MANTLE BLACKEYE TYPE cMc cBc Ce cBcM CcM cBcB CcB Gern cMcM CcM (Se Ze cBcB (SS A> cMcM:CcM:CC (Me) ery il] cMcM,CcM GG FIG. 2. The 4 pigmentation phenotypes and their genotypes. Heavy arrows indicate true breeding lines. observation. When snails of the spotted mantle phenotype were observed in the interbreeding colony of albino В. stra- minea, 13 were isolated. Eleven of these produced only spotted mantle descendants (or spotted-mantle and albino descendants if heterozygous) for as many generations as observed. Albino lines derived from these snails bred true for the albino phenotype. Among the descendants of 2 of the spotted mantle snails originally isolated (including #11 in Fig. 1) sporadic wildtype snails occurred. In each case when a wildtype offspring of a spotted mantle snail was isolated, it proved to be heterozygous (CcM), producing offspring of both С and cM phenotypes in approximately 3:1 ratio. Homozygous wildtype snails derived from these heterozygotes bred true for the wild- type phenotype thereafter. The relationships among the 4 pigment alleles are illustrated in Fig. 3, which includes 2 crosses, each involving all 4 alleles but in different mating combina- tions. From the cross сВс X CcM 14 Fys were obtained, in phenotypic. ratio 6C:4cB:4cM. From the cross between 2 of me Fuss Ge A cBcM, 39 offspring were obtained in phenotypic. ratio 22C:8cB:9cM. Heterozygotes from these and other crosses were selfed with the following results: 16 CcM snails yielded 108C:43cM offspring (2.51:1) 12 cBcM snails yielded 85cB:30cM offspring (2.83: 1); and 19 cMc snails yielded 80cM:49c off- spring (1.63:1). Although blackeye pigment phenotype В. straminea typically produced mantle spots, these generally have less dense black pigmentation than wildtype B. straminea. Biomphalaria straminea snails of the spotted-mantle phenotype have fewer mantle spots (sometimes only 1) but these have dense black pigment like the wildtype B. straminea. This is particularly evident when a blackeye-spotted-mantle (cBcM) heterozygote is selfed; the blackeye off- spring have many pale mantle spots, while the spotted-mantle offspring have fewer but darker black spots. Like the wildtype and blackeye pigment types in B. glabrata (Richards, 1972), these phenotypes and the spotted-mantle in B. straminea have a network of cells with black pigment lining the hemocoel. These pigmented cells, like the mantle spots, are fewer in number in the spotted mantle snails than in the wildtype. DISCUSSION Data, including that from other selfed B. straminea snails and crosses in addition to those presented here, suggest that the spotted-mantle phenotype is determined by a 4th allele of the gene for pigmentation, and that this allele is recessive to wildtype RICHARDS 114 "SUOIJEUIQWUOD JUaIAJJIP и! Ing ’sajajje зиэшб!а y пе BUIAJOAUI yoea ’sasso19 OM] ‘$ *5) | I: IE 92:98 g91:9! 92:92 99:920! 98:22 mdL:DL md1:D2 g9r:J8 g92:99 e21:9r MM 91:92 I6:m92 92:96 md9:DL м a 923 923 ie 24D 2 AB Xp ap ABAD w2G:g22:)92! A PET? W2t :g91: 901 99 9:58 nl JE:WISI wI8 wIE:gIZ g222 9 92:96 92:90) gdl! 92%: вазе: а! 99 In? wow? WIW2 WOWD wa? a2a? 29 29 29 92 a? 823 829 £32949 00 4949 2 821195 822198 92$:) +! 92 2:94 LES D lo 2 49 ap ap As Ag Wt1g2t:29 wd x 289 ÿ — > ee re № 29 BIOMPHALARIA PIGMENTATION VARIATIONS 115 and blackeye alleles, dominant over albino. Spotted mantle is not as dependable a genetic marker as the other 3 pigment alleles. In some homozygous spotted-mantle lines, the mantle spots are distinguishable in juvenile snails and spotting occurs in essentially all snails. In other homozygous spotted-mantle lines, however, due either to delayed development or minimal expression of spotting, some snails cannot be dis- tinguished from albinos. In both aquarium populations and clonal stocks maintained by selfing, homozygous wildtype, blackeye, and albino stocks of B. straminea have bred true through many generations. Eleven of 13 juvenile spotted- mantle snails originally isolated bred true for the spotted mantle, or if cMc hetero- zygotes, produced only spotted-mantle and albino offspring. The other 2 gave rise to spotted-mantle stocks in which wildtype snails continued to appear sporadically. In some spotted mantle stocks it is possible that the pigment gene is unstable or that the pigmentation is modified by other genetic factors. If a similar change to black- eye or albino cccurs, this has not been recognized. Several different types of black pigmen- tation in B. glabrata have been reported (Richards, 1969, 1972). Occurrence of the spotted-mantle phenotype in B. straminea suggests an additional distinction between eye and mantle spot pigmentation. Based on the tissues involved and the distribution of pigment granules at least 5 forms of black pigmentation can be distinguished: mantle spots consisting of groups of epithelial cells with concentrated pigment, a loose reticulum of pigmented cells lining the hemocoel, pigmented eyes, diffuse pig- ment granules in the mantle epithelium (in B. glabrata), and pigment granules in the connective tissue of mantle collar and head-foot regions. It remains to be deter- mined if these pigment manifestations represent chemical differences or merely genetic regulation of the locations of pig- ment formation. LITERATURE CITED RICHARDS, C. S., 1967, Genetic studies on Biomphalaria glabrata: (Basommatophora: Planorbidae), a third pigmentation allele. Ма/а- cologia, 5: 335-340. RICHARDS, C. S., 1969, Genetic studies on Biomphalaria glabrata: mantle pigmentation. Malacologia, 9: 339-348. RICHARDS, C. S., 1972, Pigmentation variations in Biomphalaria glabrata and other Plan- orbidae. Ma/acological Review, 6: 49-51. RICHARDS, C. S., 1975, Genetics of pigmenta- tion in Biomphalaria straminea. American Journal of Tropical Medicine and Hygiene, 24: 154-156. TE iden #20} y wich sd a mem Y sn moitarestinem ег № № oo eles vire 10 ‘fsmubst} ars fuseng pig lo encllasol!s Be) a Вано mem : ana aires «= Mbits sent) NO 2 2 ¿Na { Demas)! amar Aran a „is nofres rt firma ‚Isenicvotlor PPT AOS no! воле maté et ZONAHMFS moles fergia singmr ‘arta кие BENCALE '9 aaa обет ву e Lg sth Say ¿2 © АЯ 9 alto! в ale averno, m 1 Bee ee aha umge td ini over 2 Э AORANDIA nee serie: ergarnamt A м 00 Mi mio td зая AN EAS HOOT Yo emul R?A!AS1I a Ez 5 he № р залюй ‚gnitiee yd 4 (206 rare onidle bins ? ав dguowb url be Daran ne €} net м Ban len -créitert “ol! s 1 Sek otinsmpstione vi 10 shane oun o Ln e % 197301 er О Aka fe 240 р | № + Ate, sE Y es ‘ 150 в y + se (à nos yd eitioonm,lai “Sand “a A 14900 эго si 4) 24501 | wl пай eilt зо okidienu si 9 09 a of «безо miii в wad ton 34n añ nu Nar nu F L sima cod area? snowttib tere 977039 fees) eves er > ty to mente ‚(ter Г Г 8 mi зоо m den mu. % Ce” a Cu? RE À 3 ch ny" am “ee —P 4 s = mt De Cle. ia? e500" We mm hé 4 MALACOLOGIA, 1978, 17(1): 117-124 LE SYSTEME CIRCULATOIRE ET LE JEU DES SIPHONS CHEZ DONAX TRUNCULUS, MOLLUSQUE LAMELLIBRANCHE Marcel Moueza! et Liliane Frenkiel2 RESUME Apres avoir étudié le systeme circulatoire de Donax trunculus Linn. à l’aide d’injections directes et recurrentes, les expériences de Chapman & Newell (1956) ont été reprises sur ce Mollusque, ainsi que des expériences complémentaires. Celles-ci démontrent que, contraire- ment à ce qui a lieu chez Scrobicularia plana, l'intervention d'un flux sanguin extérieur est nécessaire à l'extension des siphons chez Donax trunculus. INTRODUCTION Dans son étude sur les Tellinacea, Yonge (1949) postule que l'extension des siphons a lieu par un afflux de sang de la circulation generale-force dans les espaces hemocoeliens des parois siphonales—tandis que la retraction met en oeuvre les muscles rétracteurs tres developpes dans cette super-famille. Chapman & Newell (1956), ayant effectue une serie d’experiences sur Mya arenaria et Scrobicularia plana, esti- ment, dans le cas de cette derniere espece, et contrairement a Yonge, que le sang exterieur aux espaces hemocoeliens des parois siphonales n’intervient pas dans les mouvements des siphons. Pour eux, l’exten- sion est conditionnee par le jeu des muscles radiaires. Ces affirmations contradictoires nous ont amenes à reprendre l'analyse du jeu des siphons chez un Tellinacea, Donax trunculus Linn., après avoir effectue une investigation methodique du systeme circu- latoire. TECHNIQUES D'ÉTUDE La resection de la coquille a été pratiquee au disque à séparer monté sur tour rapide. Réalisée entre les adducteurs anterieur et postérieur et le rétracteur des siphons, elle permet de liberer le manteau. Effectuée autour de la charnière, elle permet d’acceder au coeur apres avoir découpe les expansions suprabranchiales et le pericarde. Des injections de bleu d’aniline, ou d’une solution preparee selon la technique de Ko Bun Hian (1973) ont été pratiquées dans le ventricule et la veine viscero- pedieuse. L’exploration du systeme lacun- aire a ete completee par des injections dans le velum, les siphons et la lacune palleale situee en arriere du muscle retracteur des siphons. Des ligatures des siphons ont ete realisees sur des animaux placees a des températures comprises entre OC et 5°C. Des destructions partielles ou totales du coeur ont ete realisees; la lacune poste- rieure au retracteur des siphons a ete sup- primee par destruction de sa paroi externe au thermocautere. Les resultats de ces diverses interventions ont ete observees apres un delai de 24 heures. Les techniques histologiques employees pour l'étude de la structure des siphons sont classiques: fixation au Bouin-Hollande, coloration par une variante du trichrome de Gomory citee par Gabe (1968) et par l'Azan. OBSERVATIONS Systeme Circulatoire Le coeur, situe en arriere et hors de la masse viscerale, est enveloppe par le peri- carde en relation avec des glandes peri- cardiques bien développées que White (1942) n'a pas trouvées chez Donax vit- tatus. Sensiblement fusiforme, le ventricule qui possède de fortes parois musculaires, donne Tinstitut National Agronomique, El Harrach, Alger, Algérie. 2U.S.T.A., Bab Ezzouar, Alger, Algérie. (117) 118 MOUEZA ЕТ FRENKIEL naissance а l’aorte anterieure (Fig. 1). Les oreillettes en position latéro-ventrale, com- muniquent entre elles au niveau de leur partie anterieure, sous le ventricule. Chaque oreil- lette communique sur sa face ventrale avec un gros tronc vasculaire, somme des vais- seaux branchiaux, et sur sa face interne, avec le ventricule. Le systeme arteriel débute par l'aorte anterieure qui passe sur la paroi dorsale de l’estomac, plonge dans la cavite viscerale, a gauche de l'oesophage, et donne naissance а toute une serie d’arteres dont le trajet a ete figure pour l'essentiel. Deshayes (1844) signale qu'il n’a pas réussi a faire penetrer l'injection dans l'aorte postérieure. De nombreuses injections ne nous ont pas permis de mettre en évidence une vasculari- sation partant de la région postérieure du ventricule, ni d’artere recurrente du systeme arteriel anterieur. Le systeme veineux est aussi developpe que le systeme arteriel, dense dans la masse viscerale, particulièrement au niveau de l'estomac et de la glande digestive. Son collecteur principal, la grosse veine viscero- pedieuse s’acheve dans le sinus veineux. Ce dernier communique avec le rein par un orifice situe entre les deux connectifs cere- bro-visceraux au niveau où ils sortent de la masse viscerale. Les injections dans les siphons, dans les lacunes palleales situees en arriere du re- tracteur des siphons, dans les lacunes du bord palleal en avant du muscle cruciforme, con- cordent a demontrer l'existence d'un vaste réseau de lacunes peripheriques, communi- quant entre elles, dont 2 sont importantes. La première, impaire, située entre l'adducteur antérieur, le retracteur anterieur et le pro- tracteur (Mouëza & Frenkiel, 1974, fig. 1), communique en bas, en avant avec les lacunes du manteau visceral, en arrière avec le vaisseau supérieur des branchies. Elle entre aussi en relation, en haut, avec les lacunes des bords du manteau. La seconde lacune, paire, situee en arrière du retracteur des siphons, communique largement avec les siphons, les lacunes du bord palleal, les branchies, le rein, et, par son intermédiaire, le sinus veineux. Les lacunes du bord palléal forment un système qui court tout autour du manteau. Elles communiquent avec les lacunes du siphon inhalant (Fig. 3), en arriere du muscle cruciforme, ainsi qu'avec la lacune impaire située en arriere de l'adducteur antérieur. Elles sont en outre en relation avec la lacune situee entre les dents de la charnière, ainsi qu'avec le reseau lacunaire des adducteurs antérieur et postérieur. Les Siphons Les siphons des Tellinacea sont entiere- ment séparés et tres mobiles, car ils pro- viennent, selon Yonge (1957), de la sou- dure des plis palléaux internes et, de ce fait, ne possedent pas de tentacules. Ils laissent donc pénetrer les dépots du fond, sans discri- mination, dans la cavite palleale. Cet état de fait se trouverait realise chez les Donacidae pour Donax vittatus (Yonge, 1949) et Egeria radiata (Purchon, 1963). Mais chez Donax gouldi (Pohlo, 1967) et D. denti- culatus (Wade, 1969), le siphon inhalant est pourvu d’une couronne de tentacules qui joue un rôle de filtre. Tel est aussi le cas de D. trunculus (Fig. 2). La paroi du siphon inhalant a une struc- ture qui se retrouve, a quelques variantes pres, dans tout le groupe des Tellinacea, structure decrite par Rawitz (1892) chez Psammobia vespertina, puis par Yonge (1949) chez Scrobicularia plana et Donax vittatus. Depuis, Chapman & Newell (1956) ont signalé que les couches de muscles circulaires decrites par Yonge sont en rea- lite formees de conjonctif riche en fibres collagenes et pauvres en cellules muscu- laires. Ces fibres, qui, en fait, ne sont pas circulaires forment un lacis et s’inserent obliquement sur les epitheliums interne et externe. Chez Donax trunculus (Fig. 4-8), la paroi du siphon inhalant est constituee par 2 couches musculaires épaisses, №, et L,, entourees par les assises conjonctives C,, C2 et C3 contenant de rares fibres musculaires. Une couche de fibres muscu- laires longitudinales №. s’individualise dans la couche C3 au contact de l'epithelium externe. Une couche similaire existe dans la couche C, au contact de l'epithelium interne, mais est moins bien individualisee. Des cloisons radiaires—muscles et fibres collagènes—courent d'un épithéelium a l'autre. L’assise C4, bien individualisée chez Scrobicularia plana, ainsi que chez Donax vittatus (Yonge, 1949), est irreguliere chez D. trunculus. La répartition des hemocoeles est sensi- blement differente dans un siphon retracte ou en extension. Dans le siphon retracte (Fig. 6) elles sont limitees a la couche Со. Les hemocoeles longitudinales dorsale et ventrale, decrites par Duval (1963) chez А AGP CPD ES VAV AIP EXTENSION DES SIPHONS DE DONAX API е antérieure; ADA, adducteur antérieur; ADP, adducteur e et supérieure; AN, anus; AP, artére des VST MC AN VVP SIV te antérieure; AA, arté — — © phon postérieur: AGP, artère é tracteur postérieur du pied; SE, siphon exhalant; SI, si ST, veine stomacale; VVP, veine viscéro-pédieuse. Lin dE >> a © Е = 5 .- O ex + E „eo a? E gos oesophage; P, péricarde; PA, palp t; V, ventricule; VAV, valvule auri ruciforme; O, oreillette; OE, inhalant; SIV, sinus veineux; SS, sac du stylet; V, FIG. 1. Systéme circulatoire de Donax trunculus. A, aor MC, muscle c 120 MOUEZA ЕТ FRENKIEL ел „№ | “ £ LA 14” ey #, te * + by ®. М. . 2 ak ENE FIG. 2. Extrémité du siphon inhalant montrant la lacune périphérique à la base des tentacules. FIG. 3. Coupe parasagittale des siphons montrant les relations des lacunes des siphons avec celles du velum. La coupe passe par les lacunes dorsale et ventrale du siphon inhalant. FIG. 4. Coupe longitudinale de la paroi du siphon inhalant. L'épithélium interne se plisse regulierement tandis que l'épithélium externe présente un plissement mineur produit par les muscles L, et un plissement majeur produit par les muscles L,. La répartition des faisceaux musculaires est a comparer avec celle observée sur coupe transversale (Fig. 6 et 8). Abréviations utilisées dans les figures 2 а 8: br, branchie; C,, С,, C,, couches conjonctivo-musculaires circulaires: cs, cellules sensorielles; ee, épithélium externe; ei, épithélium interne; |, lacune; L,, L,, L;, couches musculaires longitudinales; mc, muscle cruciforme; mr, muscle radiaire; n, nerf; se, siphon exhalant; si, siphon inhalant; ss, septum siphonal; t, tentacules siphonaux. EXTENSION DES SIPHONS DE DONAX FIG. 5. Siphons exhalant et inhalant rétractés. FIG. 6. Paroi du siphon inhalant rétracté. FIG. 7. Siphons exhalant et inhalant en extension. FIG. 8. Paroi du siphon inhalant en extension. Abréviations: voir sous Fig. 4. i car lili za 122 MOUEZA ЕТ FRENKIEL plusieurs especes, mais que cet auteur п’а trouvees chez D. trunculus, sont pourtant bien développées (Fig. 5 € 7). А l'extre- mite des siphons, elles s’ouvrent largement dans l’hémocoele circulaire, qui occupe toute l'épaisseur de la paroi siphonale dont la musculature est réduite et les differentes couches musculaires non individualisées a ce niveau (Fig. 2). L’hémocoele circulaire qui occupe une place importante a la base des siphons, se resoud a leur extremite dans les tentacules. Chez un animal anesthesie, dont les siphons sont en extension partielle, des expansions de I’hemocoele circulaire pénetrent le long des faisceaux de muscles radiaires (Fig. 8). Lorsque le siphon en extension a été ligaturé a sa base, on met en evidence des lacunes dans la couche C3 au contact de la couche L,. Ces lacunes communiquent avec les hemocoeles de la couche C, par l'intermédiaire des lacunes radiaires. Le siphon exhalant a une paroi plus mince que celle du siphon inhalant. Construite sur le méme plan, elle comporte les mémes couches musculaires et conjonc- tives, cependant moins bien individualisees. L'observation montre que l'extension des siphons est progressive, sans a-coups, comme l'ont note Chapman & Newell (1956), au contraire de la retraction qui peut étre rapide, voire brusque. Normale- ment, Donax trunculus ne sort jamais ses siphons au maximum de leur longueur. Toutefois, lorsque l'animal est place dans des conditions defavorables, en particulier dans une eau insuffisamment oxygenée ou la nourriture se fait rare, les siphons devien- nent tres longs et turgescents. Trevallion (1971) а montre experimentalement que des conditions defavorables entrainaient les mêmes effets chez Tellina tenuis. Dans les conditions normales d'existence, la turges- cence peut intervenir chez D. trunculus а la fois pour provoquer l'extension et l’elargissement des siphons, surtout du siphon inhalant. En augmentant le diametre de ses siphons, du simple au double, D. trunculus renforce les moyens qu’il a de s'opposer aux deplacements sous l’action des courants ou des vagues. Avec son pied et ses siphons etales, il offre le maximum de resistance au transport (Mouëza, 1972). Ceci est en contradiction avec l’idee de Yonge (1949), selon laquelle le volume sanguin n’est pas suffisant pour permettre, chez les Tellinacea, a la fois l'extension du pied et des siphons, et avec la théorie de Chapman € Newell (1956) qui lient l'extension des siphons à l'amincissement de leur paroi. De plus le fait qu'un animal anesthesié par le froid puisse garder ses siphons largement sortis, s'oppose а l’idée selon laquelle la contraction des muscles radiaires est le moteur principal de l’exten- sion. EXPÉRIMENTATION Apres ablation d'une valve de la coquille, le siphon exhalant est à peine sorti et le siphon inhalant s’allonge peu—5 mm environ—avec une forte tendance а se ге- dresser du côte de la valve absente. Le jeu des siphons apparait donc fausse, aussi a-t-on eu recours a une autre méthode pour montrer sans equivoque que la pression de l'eau contenue dans la cavité palleale n'intervient pas dans l'extension des siphons. Il suffit de resequer la plus grande partie de la valve en laissant en place les surfaces d'insertion des adducteurs, ainsi que celle des retracteurs des siphons, ce qui equivaut а l'ablation de la valve. Dans de telles conditions, les siphons sortent large- ment et leur jeu extension-rétraction est normal. La destruction du coeur a ete partielle— suppression du ventricule, suppression des oreillettes—ou totale. Dans le premier cas, le jeu des siphons demeure normal, tandis que dans le dernier, il n’y a jamais extension des siphons. La ligature des siphons a été pratiquee sur l’un ou l'autre ou sur les 2 siphons а la fois. Dans le cas de la ligature d’un seul siphon, quel qu'il soit, le jeu de l’autre de- meure normal. La ligature des 2 siphons a la fois entraine les mémes effets que ceux ob- serves dans le cas de ligature d'un seul siphon. La ligature est suivie d’une rétraction qui met en jeu seulement la partie proximale, la partie distale isolée par la ligature gardant méme longueur et méme diametre. L’allongement reste possible; il a lieu sans a-coup et de maniere progressive, par etirement de la partie proximale. Dans le cas d’une ligature du siphon inhalant en extension sub-totale, lorsque la ligature est proche de la coquille, l'animal ne peut que difficilement rentrer la partie distale, qui а gardé une longueur constante. Au prix d'une rétraction forcée, il peut n'en laisser dépasser qu'une tres faible portion, mais il doit pour cela, re- tracter la partie antérieure du siphon, en avant du chiasma du muscle cruciforme, ce qui n'est pas normal chez les Donax. EXTENSION DES SIPHONS ОЕ DONAX 123 Apres destruction de la lacune situee en arriere du rétracteur des siphons, les siphons ne s’allongent plus du tout, méme au cours de l'enfouissement qui reste possible. DISCUSSION Chapman & Newell (1956) estiment que le jeu des siphons chez Scrobicularia plana s'explique par l'action des fibres muscu- laires radiaires dont la contraction provoque l'élongation par redistribution du liquide antagoniste dans les espaces hemo- coeliens et l'amincissement des parois siphonales. L'extension, pour eux, apparait indépendante du flux sanguin et les parois gardent un volume constant à tous les stades du jeu des siphons, de telle sorte que, lors de la retraction, l'espace hemocoe- lien longitudinal qui participe au stockage du liquide antagoniste s’elargit. Chez Donax trunculus le jeu des siphons presente une certaine analogie avec celui de Scrobicularia plana, l'accent doit cependant étre mis sur les differences. L’experimenta- tion amene a tenir pour certaine l'interven- tion du flux sanguin exterieur aux siphons, et, pour cela, le pompage du coeur semble obligatoire et l’integrite des lacunes situees en arriere des retracteurs des siphons doit ötre conservee. Ceci suppose une large com- munication entre ces lacunes et les siphons, communication qui existe effectivement. L'injection du systeme arteriel permet exceptionnellement de mettre en evidence ces relations, ce qui, par contre, peut étre realise a partir d’injections dans le systeme lacunaire palleal ou dans la veine viscero- pedieuse. L’histologie apporte la preuve incontestable des liaisons entre les lacunes posterieures au retracteur des siphons et celles des siphons inhalant et exhalant, ainsi que des liaisons existant entre les lacunes du velum et du siphon inhalant. Enfin la meilleure preuve d’un apport exterieur de sang reside dans le fait qu’apres ligature, le trongon proximal s’allonge et repousse pro- gressivement le bout distal qui garde une longueur constante. Lors de la retraction, contrairement ä ce qui a ete observe chez Scrobicularia plana, les espaces hemocoeliens de la couche C, sont plats. Seules les hemocoeles longitudi- nales dorsale et ventrale gardent une certaine importance. Des que l'extension a lieu, les espaces hemocoeliens deviennent turgescents. Les lacunes de la couche C, augmentent de volume; celles de la couche C3 deviennent apparentes, l’epithélium externe se deplisse. Le volume sanguin admis peut entrainer ou non une augmenta- tion du diamètre, tres variable à longueur egale. L’épithelium interne du siphon inhalant offre peu de possibilités de deplis- sement. Il se pourrait que le diametre de la lumière soit quand même accru par le jeu des fibres radiaires comme le postulent Chapman €: Newell. ll n’a pas été possible d'apporter d'arguments sur ce point. L'epi- thelium interne du siphon exhalant possède la même structure que l'epithelium externe, argument en faveur d'une possiblite plus grande d'extension de sa lumiere. L'extension resulterait chez Donax, dans un premier temps, d'un relâchement des fibres musculaires longitudinales. || existe une position d'équilibre, nécessitant une dépense d'énergie minimale, au cours de laquelle les siphons et le pied sont legere- ment allonges, position dans laquelle on trouve frequemment les animaux sur le sediment. Le deuxième temps de l’allonge- ment est actif, et le coeur semble intervenir dans ces mouvements de fluide. Le fait que la destruction partielle du coeur n’entraine pas de disfonctionnement demeure inexpli- cable dans l'état actuel. L'extension sans a-coups, qui peut être lente ou rapide s'explique par la presence d'un réservoir. De fait, la destruction des lacunes situees en arrière des rétracteurs des siphons empêche toute extension. Le jeu des muscles radiaires et longitudinaux inter- vient, par contre, pour permettre les mouvements des siphons et leur rétraction. Celle-ci, rapide et brusque, se congoit aussi par la presence d’un reservoir, prét a rece- voir le sang brutalement chasse des siphons sous l'action de la contraction des muscles. TRAVAUX CITES CHAPMAN, G. & NEWELL, P., 1956, The role of body fluid in the movement of soft-bodied invertebrates. Il. The extension of the siphons of Mya arenaria L. and Scrobicularia plana (da Costa). Proceedings of the Royal Society of London, 145: 564-580. DESHAYES, G. P., 1844-1848, Exploration scientifique de l'Algérie. Zoologie, I. Histoire Naturelle des Mollusques, 609 p. DUVAL, D. M., 1963, The comparative anatomy of some lamellibranch siphons. Proceedings of the Malacological Society of London, 35: 289-295. GABE, M., 1968, Techniques Masson, Paris, 1 113 p. histologiques. 124 MOUEZA ЕТ FRENKIEL KO BUN HIAN, 1973, A new injection fluid for malacologists. Ma/acologia, 14: 440. MOUEZA, M., 1972, Contribution à l'étude de la biologie de Donax trunculus L. (Moll. Lam.) dans l'Algérois: ethologie en baie de Bou Ismäil. Tethys, 4: 745-756. MOUEZA, M. & FRENKIEL, L., 1974, Contri- bution à l'étude des structures palleales des Tellinacea. Morphologie et structure du manteau de Donax trunculus L. Proceedings of the Malacological Society of London, 41: 1-20. POHLO, R. H., 1967, Aspects of the biology of Donax gouldi and a note on evolution in Tellinacea (Bivalvia). Ve/iger, 9: 330-336. PURCHON, В. D., 1963, A note on the biology of Egeria radiata Lam. (Bivalvia, Donacidae). Proceedings of the Malacological Society of London, 36:251-271. RAWITZ, B., 1892, Der Mantelrand der Ace- phalen. Jenaische Zeitschrift für Naturwissen- schaft, 27: 1-232. TREVALLION, A., 1971, Studies on Tellina tenuis da Costa. Ill. Aspects of general biology and energy flow. Journal of Experi- mental Marine Biology and Ecology, 1: 95-122. WADE, B. A., 1969, Studies on the biology of the West Indian Beach Clam, Donax denticu- latus L. 3: Functional morphology. Bulletin of Marine Science, 19: 306-322. WHITE, K. M., 1942, The pericardial cavity and the pericardial gland in the Lamellibranchia. Proceedings of the Malacological Society of London, 25: 37-88. YONGE, C. M., 1949, On the structure and adaptations of the Tellinacea, deposit-feeding Eulamellibranchia. Philosophical Transactions of the Royal Society of London, Ser. B, 234: 29-76. YONGE, C. M., 1957, Mantle fusion in Lamel- libranchia. Pubblicazioni della Stazione Zoologica di Napoli, 29: 151-171. ABSTRACT The circulatory system of Donax trunculus Linn. is described, having been investigated by means of direct and recurrent injections, together with experiments on the movements of the siphons using Chapman & Newell’s (1956) methods. Our experiments show that a flow of blood into the siphons is necessary for their extrusion in D. trunculus. This is different in Scrobicularia plana. MALACOLOGIA, 1978, 17(1): 125-142 PRELIMINARY CHARACTERIZATION OF THE SECRETION OF THE ACCESSORY BORING ORGAN OF THE SHELL-PENETRATING MURICID GASTROPOD UROSALPINX CINEREA' Melbourne R. Carriker, Leslie G. Williams College of Marine Studies, University of Delaware Lewes, Delaware 19958, U.S.A. Dirk Van Zandt Research Tower, Analytical Department, Microscopy, Ethicon, Inc., Somerville, New Jersey 08876, U.S.A. ABSTRACT Results of a preliminary characterization of the secretion of the accessory boring organ (ABO) of the shell-boring, predatory, prosobranch gastropod Urosalpinx cinerea follyensis Baker are reported. The secretion was examined on ABOs of live snails normally extended through boreholes in valve models, and was studied ultrastructurally, histochemically, and physiologically. Hydrogen and chloride ions were monitored with microelectrodes. The ABO emerged from the foot of the snail bearing on its free distal surface most of the secretion to be used during the subsequent period of dissolution; only small amounts were released during ABO activity in the borehole. Blots of secretion from large snails touched to fragments of cover glass measured 1.5 to 2.0 mm in diameter, and contained an estimated 1 to 2 ug of dry material. Roughly a 3rd of the volume of the secretion evaporated during drying on blots, leaving a highly hygroscopic residue. Ultrastructurally the secretion con- tained minute particles resembling crystals and secretion granules. No acid mucopolysaccharides were identified in the secretory cells of ABOs stained with alcian blue and astra blue following fixation in several separate fixatives. The periodic acid-Schiff reaction was strongly PAS-positive. ABOs heated to 80°C did not etch. Those in a solution of papain produced no etchings, whereas others immersed in a trypsin solution etched conspicuously. The range of pH of the secretion was 3.8 to 4.0, whereas secretion free of seawater on the gland did not fall below 5.2. Maximum chloride ion concentration in the secretion ranged from 0.79 M to 1.71 M. Chloride ion concentration increased stepwise from the time of extension of the ABO generally to withdrawal from the borehole. Qualitative analysis of dry secretion by energy dispersive x-rays confirmed the presence of chloride and also disclosed sodium. Minute quantities of organic matter in the secretion were volatilized by heating. INTRODUCTION Although the pace of research on the biology of calcibiocavites (organisms that hollow out spaces in hard calcareous sub- strata) has quickened in the last decade (Carriker, Smith & Wilce, 1969; Carriker & Van Zandt, 1972b; Milliman, 1974; Golu- bic et al., 1975; Warme, 1975), studies on the physiological and chemical phases of penetration of hard calcareous substrates by organisms have lagged seriously behind the systematic, morphological, behavioral, and ecological aspects. In this paper we report the results of a preliminary characterization of the secre- tion of the accessory boring organ (ABO) of the shell-boring, predatory, prosobranch gastropod Urosalpinx cinerea follyensis Baker. Background for this investigation is found in papers by the following authors, who reported on preliminary physiological and chemical aspects of shell penetration in the following species. On Urosalpinx cinerea: Carriker & Chauncey, 1973; Carriker & Van Zandt, 1964; Carriker, Scott & Martin, 1963; Carriker, Van Zandt & Charlton, 1967; Carriker, Van Zandt & Charlton, 1972; Person et al., 1967; Smarsh TUniversity of Delaware College of Marine Studies, Contribution Number 111. (125) 126 CARRIKER, WILLIAMS AND VAN ZANDT et al., 1969; Zottoli & Carriker, 1974; on Nucella lapillus: Chetail & Fournié, 1969, 1970; Chetail, Binot & Bensalem, 1968; оп Polinices lewisi: Bernard & Bagshaw, 1969; and on Argobuccinum argus: Day, 1969. This literature is reviewed in detail by Carriker & Williams (1978) with reference to an hypothesis on the chemical mecha- nism of shell dissolution by predatory bor- ing gastropods. MATERIALS AND METHODS Animals Urosalpinx cinerea follyensis snails were collected in Wachapreague, Virginia, and were maintained in the laboratory with rapidly flowing seawater and oysters, Cras- sostrea virginica (Gmelin), or mussels, Mytilus edulis Linné. Large, actively feed- ing adult snails were employed in the study. During the winter, snails were main- tained in running seawater warmed (17-19°C) and allowed to reach gas phase equilibrium in a heat exchanger (Carriker & Van Zandt, 1973); maximum temperature in the summer was 24°C. Salinity of the seawater varied from 30.4 to 32.0 9/оо, and the pH ranged from 8.0 to 8.2. Stand- ard flourescent ceiling lights illuminated the seawater trays during the day and occasion- ally during the early evening, providing about 35 footcandles of light at the level of the trays. Only limited indirect daylight came to the front sides of trays through windows in adjoining cubicles. Collection of Secretion Collection of secretion from the nor- mally extended ABO of Urosalpinx cinerea was carried out on a valve model under a dissecting microscope (Fig. 1). Models were prepared as follows (Carriker & Van Zandt, 1972b: 178-182): on a Friday, snails deprived of food for about a week were placed in running seawater among well- shaped actively feeding oysters and mussels about 10 cm long. On the following Mon- day approximately 20 of the bivalves on which snails were boring were set aside for opening. The free valve and flesh were gently removed under water, and the remaining half-shell boring-snail preparation was then placed, inner surface of the valve facing up and the snail suspended under- neath, on a support cut from a plastic dish (fig. 6 in Carriker & Van Zandt, 1972b). Most of the snails completed penetration of the valve during the Monday-Friday period. Boring rate was about 0.3-0.5 mm per day, and it took snails about 8 hr to excavate a borehole completely once the emerging incomplete hole was visible through the translucent interior portion of the valve. Valve models were positioned one at a time on a manipulated stage in running seawater under a binocular microscope just prior to breakthrough, the inside surface of the valve elevated slightly above the meniscus of the seawater. Care had to be taken to position the inhalant siphon of the snail under water, or the snail would promptly abandon the borehole. The inner surface of the valve was then rinsed with distilled water and air-dried. Elevation of the model out of the seawater reduced the possibility of welling of seawater through the borehole during exchange of the ABO and the proboscis. When occasionally seawater did spill onto the inner surface of the shell, the surface was rinsed again with distilled water and blotted dry with absorbent paper. For the most part, however, the ABO (surround- ed by a sleeve of the stalk epithelium) generally extends from the foot into the borehole, and emerges from this sleeve as it comes close to the bottom of the borehole. This manner of extension of the ABO, in addition to voidance of seawater from the borehole by the propodium prior to entrance of the ABO, insured minimal, if any, dilution of the secretion by seawater (Carriker & Van Zandt, 1972b). This facili- tated collection of secretion free, or rela- tively free, of seawater. Samples contami- nated with seawater were discarded. When the inner surface of the valve was coarsely textured, we coated the surface with a thin layer (0.2-0.3mm) of hot paraffin just before breakthrough by the snail to reduce capillary creeping of sea- water over the shell. Snails readily rasped through this thin layer of paraffin. During very dry atmospheric conditions, when the secretion tended to dry on the gland, we increased atmospheric moisture in the vicinity of the gland by placing strips of cheesecloth across the valve a short distance from the borehole. Ends of the cloth immersed in running seawater around the model served as wicks to draw seawater over the shell and accelerated evaporation. Seawater did not wet the shell UROSALPINX ACCESSORY BORING ORGAN SECRETION 127 FIG. 1. Valve model for collecting secretion. Snail is attached to underside of valve, and glass bead on end of a glass stem is positioned in borehole to prolong use of hole by snail. Seawater is flowing slowly through the container. surface immediately around the borehole. Completion of boreholes was accelerated artificially by 1 of 2 methods: (a) by drill- ing into the shell almost to the emerging borehole with a 1mm bit and dental drill, and (b) by application of IN HCI with a small pipette to the shell over the emerging borehole. Action of the acid was immedi- ately stopped by addition of seawater. The surface of the shell was rinsed with distilled water just before artificial breakthrough. Accidental spillage of acid onto the ABO caused no response from the snail, whereas addition of distilled water to dilute the secretion for easier removal resulted in violent withdrawal of the gland. Glucose of approximately the same osmotic pressure as seawater was tolerated by the snail, but caused the secretion to form a film over the ABO. Occasionally for experimental reasons it was necessary to delay breakthrough for a few hours. This was done by coating the shell surface over the emerging borehole 128 CARRIKER, WILLIAMS AND VAN ZANDT with a relatively thick layer of molten par- affin. When ready to expose the borehole, we scraped the layer of paraffin off the valve. Abandonment of the borehole usually took place when the hole was large enough for the snail to extend its proboscis fully through it into the air; after unsuccessful attempts to locate flesh over the surface of the valve, the snail withdrew its proboscis and crawled away. In rare cases, snails con- tinued to enlarge the borehole without extending the proboscis through it, appar- ently because of the lack of stimulus required to initiate a feeding response and terminate shell boring. In order to prolong the use of the model for collecting secre- tion and provide as large a borehole as possible for full extension of the ABO, we frequently delayed abandonment of the hole by the snail. This was accomplished by gently positioning a tear-shaped glass bead or rod mounted on a stem over the borehole each time the ABO was with- drawn and before the proboscis was extended (Fig. 1); the false bottom over the borehole prevented discontinuation of shell boring. In the event that a snail deserted its borehole, the snail could often be attracted back by placing a small piece of fresh oyster or mussel flesh over the borehole. This procedure although of behavioral interest, interrupted the normal sequence of ABO-proboscis activity, and made collec- tion of secretion difficult. When the ABO bulged out of the bore- hole a distance at least its own diameter (about 1 mm in large snails), secretion was ready for collection. This was done by touching bits of cover glass, filter paper, glass hooks, or capillary tips to the secret- ing disc with the aid of a micromanipulator under a binocular microscope. Touching the surface of the ABO had to be done gently because the gland is heavily inner- vated (Nylen et al., 1969) and very sensi- tive to mechanical irritation and gradual changes in heavy pressure. A small frag- ment of mirror set at an angle beside the borehole and ABO, and visible through the binocular microscope, facilitated placement of the collecting device on the center of the secretory disc. Secretion was collected by 3 separate techniques: Capillaries. The tip of a capil- lary (Drummond microcap, 2 ul), directed by a micromanipulator and held on a hypo- dermic needle attached to a screw-driven suction apparatus, was touched to the exposed surface of the moist ABO. By capillarity, secretion rose in the capillary about a mm. The drop of secretion was then drawn to the middle of the capillary by aspiration and the ends of the capillary were sealed by flaming. In high humidity, the only condition under which the tech- nique works, it was possible to collect 2 or 3 successive samples of secretion in the same capillary during successive extensions of the ABO. Glass hooks. A 2-ul capillary was drawn to a fine, slightly hooked point in a small flame. After the ABO was extended in a relatively dry atmosphere, the partly jelled secretion was picked off the gland. The secretion, being viscid, formed a drop and soon dried, roughly 2/3 of it evaporating. In some cases it was possible to make 2nd and 3rd collections on the same glass hook on successive extensions of the ABO of 1 snail. Cover glass fragments. This proved the most successful procedure, and maximal quantities of secretion were obtained by it. After the ABO billowed fully out of the borehole, the flat surface of a fragment of cover glass (approximately 3 X 4mm) held on opposing edges by fine foreceps, was touched quickly to the disc of the ex- tended ABO. Snails did not seem to be irritated by sudden contact, and a thick blot of the full area of the exposed gland was generally collected on the glass surface. The surface of the glass picked up most of the secretion on the gland; a 2nd applica- tion of glass to the same gland before with- drawal resulted in very little additional secretion. This, and other observations, indicated that most of the secretion was released by the ABO when Ist extended. Secretion was exuded in a uniform layer over the entire crown of the gland, resulting in a thin cap-like sheet which tended to retain its form in seawater and in air. Because of their highly hygroscopic nature, collections of dry secretion were stored in a vacuum desiccator. Electron Microscopy For examination at low magnifications with the scanning electron microscope (SEM), ABO secretion was blotted on frag- ments of cover glass and dried in air. Each UROSALPINX ACCESSORY BORING ORGAN SECRETION 129 glass fragment was secured to a SEM stub with silver paint and coated with carbon and gold in a vacuum. For study of the ultrastructure of the secretion at higher magnifications, blots were fixed in osmium vapor immediately after collection, dehy- drated in an acetone-ethanol series, dried in a critical point dryer, and coated with car- bon and gold. Secretion in freshly collected incomplete boreholes was prepared as follows: snails were allowed to penetrate valve models of Myti/us edulis until radular and ABO activity were visible through the translucent inner aragonitic layer of the shell; snails were removed after the ABO had been in the hole about 30 min, and boreholes were dried immediately with a stream of air. Most of the shell surrounding the borehole was then removed with cut- ting pliers, and the shell preparation was mounted on a stub with silver paint, dried in an oven at 60°C overnight, coated with carbon and gold, and examined in the SEM. For ultrastructural examination by trans- mission electron microscopy (TEM), a very thin layer of ABO secretion was blotted on parlodion-coated copper EM grids. Grids were held one at a time with fine curved forceps and touched very lightly to the moist secretory disc of the ABO. Some grids were examined in the EM without coating under an electron beam of low intensity, while others were coated with carbon before viewing. Histochemistry ABOs were excised from the foot of active unanesthetized snails (Carriker & Van Zandt, 1972a describes the method). Glands were dropped immediately in one of the following histological fixatives: Bouin’s, Campy, formalin in seawater, Gilson, glutaraldehyde in seawater, Heiden- hain’s Susa, Petrunkewitsch, and Perenyi. Paraffin tissue sections, cut 4 um thick, were mounted on microscope slides and stained with (a) alcian blue and safranin-O, and (b) astra blue and Kernechtrot, to determine if acid mucopolysaccharides were present. Histochemical identification of neutral mucopolysaccharides and muco- proteins in sections of ABOs was carried out with the periodic acid-Schiff (PAS) reaction (Barka & Anderson, 1963). For general differentiation of tissues, sections were stained with Lillie’s modification of Masson’s trichrome stain and mounted in balsam saturated with salicylic acid. Physiology Heat. Since nearly all enzymes are irreversibly destroyed by heating to 80%: we applied heat to determine if this affected the capacity of ABOs to etch shell. Fourteen ABOs were excised from large snails. Seven ABOs were placed in a small screw-capped vial in 4 drops of clean seawater of pH 7.8, and the other 7 were left in 4 drops of seawater in a | сс dish. The vial, tightly capped, was then heated in a water bath to 85°C and left at this temperature for 2 min. Then heated ABOs were cooled. Both heated and control ABOs were transferred with a little of their own seawater to polished shell preparations in a moist chamber at about 24°C and left there for 18 hr. At the end of this time ABOs were flushed from the shell frag- ments, and shell surfaces were air-dried and examined with a light microscope for evi- dence of dissolution. Enzymes. The gross effect of the pro- teolytic enzymes, papain and trypsin, on the etching capacity of the secretion of excised ABOs was determined as follows. Nine glands were excised from adult snails. Three ABOs were placed in a 1% solution of papain (w/v in glass-filtered seawater), and 3 were put in a 1% (w/v in seawater) solution of trypsin. In each case, glands were allowed to remain in the enzyme solu- tions for about 2 sec and were then trans- ferred individually to polished shell with a small drop of the enzyme-seawater solu- tion. The 3 remaining ABOs, controls for the effects of enzyme solutions on ABO etching, were placed in filtered seawater and similarly transferred to polished shell with a drop of seawater. Drops of enzyme solutions were also placed on polished shell as controls for the effects these enzymes might have on shell dissolution independent of ABOs. Shell-gland-enzyme preparations were held in a moist chamber for about 15 hr. Shell surfaces were then rinsed, dried, coated with chromium in vacuum, and examined optically with incident illumi- nation. pH. The pH of secretion of intact, normally functioning ABOs of Urosalpinx cinerea was determined by Carriker, Van Zandt & Charlton (1967) in an oyster model. This device, however, presented 130 CARRIKER, WILLIAMS AND VAN ZANDT problems which were not possible to solve: seepage of fluids through the shell-glass juncture from the dying oyster within the model into the artificially bored hole, access to the disc of the ABO in the artificial borehole only from the side with the pH glass microelectrode, and extreme sensitivity of the system to electric fields which forced us to carry out deter- minations in still seawater. Development of the valve model (Carriker & Van Zandt, 1972b) presented an opportunity to check our earlier deter- minations of the pH of the ABO secretion in a normally functioning ABO under rela- tively normal conditions. One at a time adult snails which had completed а bore- hole т a Mytilus edulis valve were positioned on a submerged, mechanically manipulated stage in a container of slowly running seawater, the surface of the valve raised above the meniscus of the water and illuminated under a binocular microscope. A pH glass microelectrode (50 millivolts per pH unit, and relatively temperature independent) was held in a micromanipula- tor. The pH was recorded with a Beckman research pH meter and strip chart recorder, and the standard reference electrode was inserted in seawater flowing beside the valve model. The system was stable, and once calibrated, remained constant for several hours at a time. Eight series of complete recordings were made of the pH of secretions from the propodium, pro- podial transverse pedal furrow, buccal cavity, and ABO as they were moved into the borehole by snails. Chloride ion. In view of the low pH of the ABO secretion (Carriker, Van Zandt & Charlton, 1967) and dissolution of mol- luscan shell by it, the large quantities of NaCl in dried ABO secretion (Carriker, Van Zandt & Grant, 1972), and the lack of response of ABOs to the application of HCl, we decided to test for the presence of chloride ion activity. However, the small quantities of secretion released during shell penetration and its highly viscid, volatile nature prevented its examination by stand- ard analytical techniques. Accordingly, we selected an intracellular chloride ion microelectrode (Microelec- trodes, Inc.) with а tip 0.5 ит in diameter for the determination. The small size of the electrode permitted precise placement of the electrode tip on various parts of extended ABOs, The electrode consisted of a silver-silver chloride-coated platinum wire, and was filled with 0.5 M KCI saturated with AgCl. Possible interfering ions were bromide and sulfide. Potential was recorded with a Beckman research pH meter and strip chart recorder. A Faraday cage was used to shield the snail-valve preparation and electrodes from electrical interference. А 0.5 М КС! solution was arbitrarily desig- nated as the standard reference solution because a 0.5 M chloride ion concentration approximates the chloride ion concen- tration in seawater and in tissue fluids in marine invertebrates. Thus a difference in electrical potential, as measured with the chloride ion microelectrode, between the 0.5 М КС! standard and the ABO secretion would be proportional to the log of the chloride ion concentration of the ABO secretion. Chloride ion microelectrodes and strip chart recorder were calibrated by titration with standard КС! solutions (Whit- field, 1971). Fixed point calibration of the recorder was performed both before and after each recording period. During the course of recording from the ABO secre- tion, we used seawater flowing around the snail-valve preparation as a secondary stand- ard. This allowed a continual check on stability and drift of the recording appa- ratus. Individual Urosalpinx cinerea which had bored a hole in a valve of Crassostrea virginica were placed one at a time on a submerged mechanically manipulated stage in a container of slowly running seawater, and the surface of the valve, with the snail beneath it, was elevated above the meniscus of the water. A binocular microscope was positioned above the valve model and the boresite was illuminated by a microscope lamp. The double junction reference elec- trode was immersed in the seawater beside the snail. The surface of the valve model was rinsed and dried prior to use, and any seawater welling through the borehole as the proboscis withdrew was carefully blotted off with absorbent paper. As soon as the snail withdrew its proboscis and extended the ABO through the borehole, we applied the tip of the chloride ion microelectrode to the center of the gland. Maximum secretion was present at this point and there was least chance of contamination with chloride ions from seawater. Presence of chloride ions was supported by the use of qualitative energy dispersive UROSALPINX ACCESSORY BORING ORGAN SECRETION 131 x-ray analysis (EDAX) in combination with scanning electron microscopy. Employing the snail-valve model used for the chloride ion microelectrodes, we collected 6 suc- cessive blots of pure ABO secretion on the tip of a fragment of pure carbon. The frag- ment was supported on a micromanipula- tor, and the tip was applied to the crown of the ABO for a few seconds when it was 15+ extended. After each collection, and prior to the next extension of the ABO, the secretion was allowed to dry on the carbon. In preparation for viewing in the SEM and the elemental scanning analysis (EDAX), the carbon fragment was coated with carbon in a vacuum. RESULTS Release of Secretion by ABO Behavioral observations suggested that most of the secretion utilized in the incom- plete borehole during chemical activity collects over the secretory epithelium inside the stalk sleeve of the ABO while it is withdrawn within the foot during the rest- ing phase. As the gland slides into the incomplete borehole, it is initially en- veloped by the stalk sleeve, but billows quickly out of the sleeve and presses the drop of accumulated secretion against the bottom of the incomplete borehole. The suggestion that the ABO emerges from the foot holding most of the released secretion to be used during the subsequent period of dissolution was examined experimentally by separating different snails from their boreholes at short and long intervals after initiation of shell dissolution. Duration of chemical and rasping periods prior to final dissolution period, and duration of chemi- TABLE 1. Duration of chemical and rasping activi- ties of Urosalpinx cinerea in incomplete boreholes prior to removal of snails from the shell for deter- mination of amount of secretion in the boreholes. ABO in hole Rasping prior to removal Snail ABO т hole in hole of snail 1 35 min 10 min 75 sec 2 40 1.5 60 3 30 1.0 60 4 40 1.3 20 min 5 35 1.0 25 6 30 1.5 30 cal dissolution period before snail removal, are tabulated in Table 1. Boreholes were removed quickly, air- dried, coated with platinum-palladium, and examined with the SEM. The thickness and appearance of dried secretion visible under the SEM was similar for boreholes removed approximately 1 min and 20-30 min after initiation of dissolution. These observations support the hypothesis that most of the active non-volatile components of the secre- tion associated with shell dissolution are supplied from the start. The hypothesis was confirmed by the fact that we were able to collect only small additional amounts of secretion from the surface of the ABO in capillaries or by blotting on glass after initial protraction of the ABO. That small amounts of secretion con- tinue to be released during ABO activity in the borehole is suggested by the following experiment. Three ABOs were excised in rapid succession from resting snails, placed in 0.01 ml of seawater with a 1 ml hypo- dermic needle in a covered microdish and left there for 0.5 hr. At the end of this time, ABOs were washed vigorously in sea- water, and then placed in a drop of fresh seawater on polished shell in a humidity chamber for 15 min. A second set of 3 ABOs was treated similarly, except that they were soaked in seawater for 1 hr before washing and placing on polished shell. Pronounced etching occurred under all 6 ABOs. The continuous gentle pulsa- tory movements of the ABO in the bore- hole, ranging from 21 to 28 per min at room temperature, could be related to release of secretion. Physical Characteristics of the ABO Secretion Upon exposure to relatively dry air, the highly viscid secretion of normally ex- tended ABOs in a valve model jelled on the surface, the clear glaze increasing in firm- ness with increasing dryness. Secretion under the glaze in contact with microvilli remained fluid, facilitating removal of secretion films with glass hooks. Jelling also occurred under seawater, though not to the extent that it did in air, the film of secre- tion becoming a light translucent milky color in air. Dry secretion collected on glass or paper was highly hygroscopic and quickly accumulated water under conditions of 132 CARRIKER, WILLIAMS AND VAN ZANDT high atmospheric humidity. Dry secretion in incomplete boreholes rinsed with dis- tilled water as soon as removed from boring snails appeared to be unaffected ultrastructurally when examined with the SEM at mangifications up to 6,000 X. Full blots of secretion from large snails (40-45 mm in shell height) touched to frag- ments of cover glass from extended ABOs measured 1.5 to 2.0 mm in diameter, and contained approximately 1 to 2 ug of dry material, estimated from an approximation of the volume of the dried secretion. The pattern of dried secretion appeared to vary with speed of drying, humidity, tempera- ture, thinness of the secretory film, and drafts of air over the sample. Figure 2 illustrates the dendritic pattern commonly observed on thick blots allowed to stand at normal atmospheric conditions and room temperature for 1 or more days. Size of particles ranged from 0.7 um to 30 um. Smaller particles appeared crystalline т > £ 5 a A form (Fig. 3). About 65% of the secretion was volatile and evaporated on drying. Assuming that blots picked up about 80% of the free secretion, we estimated that the amount of secretion released by the ABO during 1 application in the borehole was roughly 2.5 to 5 ug. Appearance of dry secretion in an incomplete borehole in the shell of Myti/us edulis taken from a snail at the end of a normal period of chemical activity is illu- strated in Figure 4. Thickness of the secre- tion coat ranged from 0.03 to 1.7 ит, and varied with the amount of secretion applied to the borehole by the snail. As a result of drying during preparation for scanning electron microscopy, the secretion coat frequently peeled back and partially frag- mented, exposing incompletely dissolved shell prisms and lamellae beneath (Fig. 4, 5, 6). The crystalline-like particles present in the secretion blots were also commonly found on the secretion coat in boreholes FIG. 2. Blot of ABO secretion of Urosa/pinx cinerea on surface of glass illustrating pattern of drying. Blot remained in open room for 3 days prior to vacuum-drying and coating with metal. Scanning electron micrograph. FIG. 3. Magnification of smallest ‘‘crystals’’ in the secretion blot illustrated in Fig. 2. UROSALPINX ACCESSORY BORING ORGAN SECRETION 133 (Fig. 6, 7), where, however, because of previous rasping activity by the snail, it was difficult morphologically to separate particles from partly dissolved fragments of shell units dislodged by the radula during the previous rasping period. At high magnifications the dry secretion coat in the borehole resembled a relatively homo- geneous film (Fig. 7). A pattern of small particles ranging in diameter from about FIG. 4. Incomplete borehole of Urosalpinx cinerea in shell of Mytilus edulis, taken quickly from the snail after ABO was in borehole for 26 min preceded by a 2-min rasping period. The secretion coat curled away from walls of hole on drying, exposing portions of attached shell beneath. Scanning electron micrograph. FIG. 5. Higher magnification of borehole shown in Fig. 4 to show relative thickness of the secretion coat and the cleanly dissolved shell units beneath. Scanning electron micrograph. FIG. 6. Higher magnification of secretion coat and ‘“‘crystals’’ in borehole shown in Fig. 4. ‘’Crystals”’ are comparable in size to smaller ones in Fig. 3. Scanning electron micrograph. FIG. 7. Enlargement of 2 “‘crystals’’ shown in Fig. 6. Scanning electron micrograph. 134 CARRIKER, WILLIAMS AND VAN ZANDT 0.02 ит to 0.05 ит was observed in scan- ning electron micrographs of a thin blot of ABO secretion treated with osmium vapor. Transmission electron microscopy revealed additional morphological detail on the characteristics of secretion blots. After secretion was removed from an ABO by 1 or 2 successive blottings, microvilli from the ABO, unprotected by the lubricating secretion, adhered to the parlodion surface on the next application of a copper grid and were pulled off (Fig. 8). Flattened by adhesion and drying, microvilli (normally about 0.1 ит in diameter) measured about 0.2 ит in diameter. Free distal ends were smooth and slightly enlarged. Remaining microvillar surfaces were covered with prominent nodules; vesicles (about 0.04-0.06 ит in diameter), diffuse in appearance, were abundant close to micro- villar membranes, and some appeared in various stages of fusion with the nodules as if in the process of release. A thin secretion blot collected on a dry day on a parlodion-coated grid, immedi- ately dried further in a vacuum desiccator, and coated with carbon, consisted princi- pally of a background of finely granular to homogeneous material with occasional prominent fragmented bodies about 0.6 to 0.7 um long (Fig. 9). The latter, promi- nently membrane-bound, when not frag- mented and not flattened by drying would O2. correspond approximately to the size of the secretion granules in the secretory cells of the ABO (Nylen et al., 1969). Because of this similarity we will tentatively name these particles ‘’secretion granules” in this report. Thin secretion blots on parlodion-coated grids, likewise collected on a dry day, but held in a moist chamber for 1.5 hr before drying and coating with carbon, presented a rather different appearance (Fig. 10). Secretion granules were intact, and abun- dant, and scattered among what resembled crystallization of finely granular material into rosette patterns organized for the most part around the secretion granules. The latter ranged in diameter from 0.17 to 0.40 um, corresponding well in size with the granules described by Nylen et al. (1969). A final set of secretion blots on par- lodion-coated grids was collected and held in a moist chamber at room temperature for about 18 hr, then dried and examined ultrastructurally. The secretion dried and “crystallized” т dendritic patterns approxi- mating the low magnifications photo- graphed in Fig. 2. Individual “crystals” ranged in size from 0.04 to 1.3 um and their shapes were similar to those photo- graphed in Fig. 3. No secretion granules were evident, probably deteriorating during the long exposure to condensation water in the moist chamber. At high magnifications FIG. 8. Microvilli, diameter 0.17 to 0.22 um, of ABO pulled off on a parlodion-coated grid. Preparation placed immediately in vacuum desiccator, then coated with carbon. Arrows point to nodules on microvilli which are clearly in focus. Vesicles are located along surface of microvilli and free around them. Transmission electron micrograph. UROSALPINX ACCESSORY BORING ORGAN SECRETION 135 FIG. 9. Thin blot of secretion of the accessory boring organ on а parlodion-coated grid dried immediately after removal from the organ; coated with carbon. Large dark bodies, about 0.67 um long, appear like secretion granules. Remainder of dry secretion varies from finely granular to homogeneous in texture. Transmission electron micrograph. individual ‘‘crystals’’ exhibited a complex, fine, intracrystalline structure of variable design (Fig. 11). Histochemistry of ABO Secretory Epithelium The alcian blue test for acid mucopoly- saccharides was reliable after fixation of tissues in Bouin’s, buffered formalin, Gilson’s, buffered glutaraldehyde, and Heid- enhain’s Susa fluids. Mucous cells of the proboscis integument, esophagus, radular sac, salivary glands and ducts, pharynx of Leiblein, and pedal epithelium of Uro- salpinx cinerea gave strong positive results, whereas secretory cells of the ABO pro- duced a negative reaction. These results were confirmed by use of the stain astra blue. A pronounced negative reaction was obtained with this stain in the secretory cells of the ABO after fixation of tissues in Champy's, Gilson's, Perenyi’s, Petrunke- witsch’s, and buffered formalin fluids. Further evidence for the absence of acid mucopolysaccharides in the secretory cells of the ABO was obtained with the periodic acid-Schiff reaction applied to tissue sec- tions fixed in Gilson’s, Heidenhain’s Susa, and buffered formalin solutions. Most of the secretion in the epithelium of the ABO was strongly PAS-positive, the microvillar zone especially being distinctly rose colored after the alcoholic fixatives. Acid muco- polysaccharides are generally PAS-negative, or give only a weak reaction of question- able specificity. The PAS test identified pri- marily neutral mucopolysaccharides and mucoproteins. In ABOs fixed in Perenyi’s fluid and stained with Lillie's modification of Masson’s trichrome stain, we observed con- spicuous leucocyte-like cells packed with red spherules averaging 0.6 to 1.3 um in diameter. In several cases the leucocyte-like cells seemed to have burst and the red spherules were scattered in spaces adjacent to the cells. Spherules of similar size and color were distributed in short string-like rows in minute channels oriented radially from the proximal to the microvillar periphery of the secretory epithelium; 136 CARRIKER, WILLIAMS AND VAN ZANDT FIG. 10. Thin blot of secretion of the ABO on a parlodion-coated grid held in a moist chamber at room temperature for 1.5 hr, then dried and coated with carbon. Clear bodies, ‘secretion granules,’’ range in diameter from 0.17 to 0.40 um (average 0.33 ит). Finely granular patterns appear to represent “recrystallization'”” of the secretion during the period in the moist chamber. Transmission electron micrograph. although continuous strings could not be traced from sinus to periphery, they were visible in sections over various parts of the secretory epithelium, collectively giving the impression of continuous connection. Association, if any, between the spherules in the leucocyte-like cells and in the minute channels, is uncertain. The same pattern and distribution of red spherules were observed in other ABOs fixed in the other nitric acid fluids (Gilson’s and Petrunkewitsch’s) and like-wise stained with the trichrome solution. Trichrome stains are not histochemical in nature, so the color gave no clue to the identity of the spherules. That they are probably not secretion granules may be inferred from the considerably larger size of the spherules than the secretion granules. Physiology Heat. None of the 7 heated ABOs etch- ed polished shell, whereas all 7 control ABOs produced moderate etchings. Heated ABOs were firm and shrank from the effect of the heat, and control ABOs remained flaccid and soft. A preliminary experiment on inactivation of etching by heat was reported by Carriker, Scott & Martin (1963), and was repeated because of improved methods for testing the shell- penetrating capacity of excised ABOs. Present results confirm earlier findings. Enzymes. The effect of proteolytic enzymes on the etching capacity of the secretion of excised ABOs was tested with papain and trypsin. These results are sum- marized in Table 2. The 3 ABOs placed т UROSALPINX ACCESSORY BORING ORGAN SECRETION 137 FIG. 11. Thin blot of secretion of the ABO on a parlodion-coated grid, held in moist chamber at room temperature for about 18 hr, then dried and examined. Specimen was not coated with carbon, so electron beam intensity was kept low. “Crystals’’ are arranged in dendritic patterns, similar to those shown in Figs. 2 and 3. Transmission electron micrograph. | TABLE 2. Effect of enzymes on etching activity on polished shell of ABO excised from Urosalpinx cinerea: + minimal, ++++ maximal etching. Treatment Etching Activity 1 2 3 3 ABOs in seawater only +++ ++ HH 3 ABOs т papain solution 0 0 0 3 ABOs in trypsin solution +4 ++ ++ Papain solution only 0 0 0 Trypsin solution only 0 0 0 the papain solution did not etch polished shell, whereas ABOs immersed ш the trypsin solution produced strong etchings comparable to those of control ABOs in seawater (Carriker, Scott & Martin, 1963). Control drops of enzyme solutions did not etch. The ABOs were slightly mushy at the end of the experiment. pH. Hydrogen ion concentration of the ABO secretion, when the ABO extended normally through the borehole in a valve model and was covered with traces of sea- water, dropped consistently to levels rang- ing from 3.8 to 4.0. When the secretion was free of seawater, however, the pH did not fall below 5.2. Minimal pH of fluid in the buccal cavity when the snail took the electrode deep into its mouth was 5.7. The pH of the mucoid secretion within the transverse pedal furrow ranged between 7.0 and 7.8, the level seemingly more depend- ent on environmental conditions than on the secretion of the furrow itself. Chloride ion. Ten recordings were made of chloride ion concentration in secretion on ABOs extended normally through the borehole in a valve model. Duration of recordings ranged from 3 to 8 min, the time the ABO remained in the borehole. Temperature of the seawater during obser- vations was about 22°C. Recording 5 was aborted when the ABO was withdrawn, and the 10th recording was discarded because of electronic drift in the instruments. Maximum chloride ion concentration from the remaining 8 recordings ranged from 079M to 1.17M, levels con- spicuously above the standard 0.5 М con- centration in seawater. Increase in chloride ion concentration on the surface of the ABO with time followed a stepwise course until the ABO was withdrawn, generally after the maximum chloride ion concen- tration was reached (Fig. 12). In 1 case, the ABO remained in position in the borehole after maximum concentration was reached and thereafter for approximately 6 min the 138 CARRIKER, WILLIAMS AND VAN ZANDT 1.4 PA un | gos ; Faz = — 1 5 sr о Y = Г. \ г Pes O : O “op = g 9 ée < РО / 0 ao é = / FL / O LJ 0) OO O = > O O z © ie SOLS о O ya O O O O OO 0 o O DO O 0.6 0) 2 4 6 8 TIME (minutes) FIG. 12. Chloride ion concentration in secretion of the accessory boring organ measured over time. Chloride ion concentration was calculated from potential recordings at 15-sec intervals and is presented here as discrete data points. Dashed line and solid circles show stepwise increase in chloride concentration observed in Recording 7. This pattern was typical of the other runs with the exception of Recording 8 which is depicted here by solid line and open circles. In this case, a rapid increase in chloride ion con- centration was followed by a much longer period of decline in concentration. chloride ion concentration showed a step- wise decrease (Recording 8, Fig. 12). Five of the recordings were of sufficient dura- tion and free of artifacts to allow trans- position of the data at approximately 15-sec intervals from the recorder paper to graph form; all recordings demonstrated stepwise changes in chloride concentration. Of the total of 8 recordings, 6 demon- strated a continuous increase in concen- tration of chloride during the period of recording, and 2 showed a drop at the end. During most recordings, the tip of the electrode was immersed in a small pool of ABO secretion made by a slight depression in the surface of the ABO by the tip of the electrode. While extended in the borehole, the ABO consistently underwent gently UROSALPINX ACCESSORY BORING ORGAN SECRETION 139 rhythmic undulations. However, correlation between the frequency of undulations and stepwise changes in chloride ion concen- trations was not clear. Accidental jarring of the electrode against the ABO, or other mechanical irritations, caused premature withdrawal of the gland, and accounted for the brevity of the recordings. Ordinarily the ABO is extended for an average of 20 to 30 min or more (Carriker & Van Zandt, 1972b). Qualitative analysis by energy dispersive x-rays (EDAX) of dry ABO secretion on a carbon stub supported the presence of chloride, and in addition revealed the pres- ence of sodium. Peaks on the screen dis- closed a high concentration of chloride and a moderate concentration of sodium, and trace amounts of calcium and potassium. The carbon control showed none of these. Com- parison by Emmett Smith (personal communication) of the index of refraction of dried ABO secretion with standard NaCl also confirmed that NaCl was the major constituent (at least 90%) in the secretion. Spot analysis in the scanning electron microscope with EDAX of dried ABO secretion blotted on fragments of cover glass demonstrated that chloride and sodium found on the carbon stub were localized in the secretion particles described earlier (Figs. 2-3). Emmett Smith (personal communi- cation) in additional optical tests slowly heated dried secretion to 340°C and observed a slight browning suggestive of decomposable organic matter starting at 240°C. A portion of the secretion was then fumed over concentrated HCl, allowed to stand until no odor of HCl was evident, and treated with ninhydrin in ethanol followed by heating. There was no evidence of color change, suggesting that organic matter present in the secretion of the ABO was volatile and evaporated along with the solvent, presumably water. DISCUSSION By carrying a full load of secretion on the distal surface of the ABO into the borehole at the beginning of the chemical phase of penetration (Carriker & Van Zandt, 1972b), the snail is able to com- mence dissolution of shell immediately over the entire surface of the bottom of the incomplete borehole. Smaller quantities secreted thereafter could be necessary to supplement the initial amount of shell- dissolving chemicals, or could provide a vehicle for discharging ions (such as chlo- ride) from the gland as other ions (calcium, for example) pass into the gland (Carriker & Williams, 1978). Pulsatory movements of the ABO in the borehole should facilitate such ionic exchange. These movements possibly also serve as a pump to work the secretion into the dissolving shell surface, and mix the dissolving mineral and organic components of shell with the secretion. Because of its viscid nature, the ABO secretion is more readily applied and held at the surface of the shell in the borehole by the ABO than a watery fluid, which would tend to drain out of the borehole. Simultaneously the secretion probably pro- tects the delicate microvillar surface of the ABO (Nylen et al., 1969) from possible abrasion against the surface of the shell of the borehole during pulsatory movements of the ABO. Furthermore, the secretion undoubtedly serves as a radular lubricant during rasping of the shell surface, as well as a wet adhesive agent, which facilitates removal by radular cusps of loosened shell fragments from the surface of the borehole prior to swallowing. A further advantage is that the secretion coats shell fragments prior to swallowing in order to minimize laceration of epithelial surfaces of the alimentary canal as the fragments pass down the esophagus to the stomach and out the intestine (Carriker, 1977). Some of the secretion is probably lost in this way. The uniformly repeating pattern of small particles ranging in diameter from about 0.02 to 0.05 um in osmium-treated thin blots of ABO secretion is similar to the pattern of particles of approximately similar size (0.04-0.06 ит) seen close to and on microvilli (Fig. 8). Whether these are identical is problematical, though the similarity in size and close association of ABO secretion and microvilli would suggest that they could. Freshly released ABO secretion dried on filter paper likewise presents a granular appearance, though the size of the particles appears larger than that of particles associated with the microvilli. The larger bodies observed in dried secretion blots and ranging in diameter from about 0.17 to 0.40 um (Fig. 9, 10), because of their membrane-bound nature 140 CARRIKER, WILLIAMS AND VAN ZANDT and roughly comparable size (allowing for an increase in diameter due to flattening during drying), are suggestive of the secre- tion granules (average diameter 0.2 ит) described by Nylen et al. (1969) in trans- mission electron micrographs of secretory cells of the ABO. These authors suggested that these dense, membrane-bound secre- tion granules possibly constitute the patent cell constituent for extracellular use, and that vesicles could be other possible con- tributors of shell-dissolving chemicals. The appearance of these bodies outside of the secretory cells of the ABO in secretion blots supports the suggestion of Nylen et al. (1969). In view of the high concentration of chloride and sodium ions in the secretion, as well as the presence of some organic matter such as carbonic anhydrase, it is likely the crystal-like bodies, which ranged in size from about 0.04 to 1.3 ит in dry secretion, were NaCl crystals whose struc- ture had been modified by .organic and possibly other substances in the secretion (Fig. 11). Neutral mucopolysaccharides and muco- proteins were also found in the ABO of the naticid gastropod Polinices lewisi (Bernard & Bagshaw, 1969) and in the burrowing bivalve Lithophaga lithophaga (Jaccarini et al., 1968). The latter proposed that pene- tration of shell could be facilitated by com- plexing of calcium by neutral тисорго- teins, and this possibility should be kept in mind as an aspect of the shell-penetrating mechanism of Urosalpinx cinerea. When ABOs of live snails extended normally through the borehole in valve models, the pH of the secretion ranged between 3.8 and 4.0. This range was similar to that obtained earlier by Carriker, Van Zandt & Charlton (1967) in oyster models under less favorable instrumental con- ditions. The pure secretion on the surface of the ABO, free of seawater, however, gave a pH reading of 5.7. Whether the viscid nature of the pure secretion modified the reading, or the acidic element of the secretion was not released until the secre- tion came in contact with seawater, has to be determined. The observation that heated excised ABOs did not etch polished shell suggests that heat-inactivated enzymes which might hydrolyze organic matrix of shell, or vola- tilized, or otherwise altered, inorganic chemicals which might solubilize the mineral crystals. Experiments treating ABOs with enzymes confirm the proteinaceous nature of at least 1 necessary component of ABO secretion. Trypsin has a pH optimum of 7.8. The presumed catalytic mechanism of trypsin requires a substrate cationic group to react with an anionic group of the enzyme (White et al., 1968). Since ABO secretion maintains its pH at 5.0 when uncontaminated with seawater (pH about 8.0) and at 4.0 when in contact with sea- water, the effectiveness of trypsin in hydro- lyzing an active protein component of the ABO secretion would be greatly reduced if not completely neutralized. The relative activity of papain has a pH optimum in the range of 6.0 to 7.0 unless the substrate is electrically neutral or substrate charge is not involved in catalysis (Lehninger, 1970). In the latter case, relative activity of papain is independent of pH. The relative effi- ciency of papain, as compared to trypsin, in inactivating the etching ability of ABOs thus could be attributed to the pH char- acteristics of the ABO secretion and the charge characteristics of the active protein component in the ABO secretion. It is unlikely that this component is carbonic anhydrase because carbonic anhydrase requires a Zn*t* co-factor which is catalyti- cally active. The enzymatic or ionic ex- change functions of this protein substance remain to be explored. Discovery of substantial quantities of NaCl in the dry secretion (Smith, personal communication), and approximately 3 times the quantity of chloride ions in freshly released secretion than in seawater (confirmed independently by energy disper- sive x-ray analysis of dry secretion), was unexpected. The fact that all recordings of the secretion demonstrated stepwise increases in the concentration of chloride ions, supports our observations that the ABO continues to release small amounts of secretion after its placement in the bore- hole at least during the initial chemical phase of penetration. It has not yet been determined, as suggested in the case of Recording 8, whether decline of chloride ion concentration during the latter part of each chemical activity in the borehole is characteristic. Although correlation between the frequency of rhythmic undula- tions of the ABO in the borehole and step- wise changes in chloride ion concentration was not obvious, it is likely that the 2 functions are related. That ABO secretion might not be the UROSALPINX ACCESSORY BORING ORGAN SECRETION 141 only biological fluid involved in shell-pene- tration is suggested by the functional involvement of the proboscis and transverse pedal furrow in hole-boring (Carriker, 1977; Carriker & Van Zandt, 1972b). The accessory salivary glands empty into the buccal cavity and the anterior pedal mucous gland drains into the transverse furrow. The composition of these secre- tions and their role in shell-dissolution have yet to be determined. It is likely, however, that their function, if any, is minor as snails deprived of their ABO are unable to penetrate shell (Carriker & Van Zandt, 1972a). Rigorous proof of this assertion would require surgical or pharmacological inactivation of the salivary and pedal glands. Physiological and chemical results obtained in this investigation are discussed further by Carrriker & Williams (1978) with reference to an hypothesis on the chemical mechanism of shell-penetration by boring gastropods. ACKNOWLEDGMENTS Michael Castagna generousiy supplied live Urosalpinx cinerea follyensis from the eastern shore of Virginia. Special thanks go to Marie U. Nylen for collaboration in pre- liminary studies of the ABO secretion, and to Philip L. Levins for helpful suggestions in the early phases of research on the ABO secretion. Frank Medeiros photographed Fig. 1; scanning electron micrographs in Figs. 2-7 were taken in collaboration with Virginia Peters; transmission electron micro- graphs in Figs. 8-10, in collaboration with David Harling; and that in Fig. 11, with Peter Schaefer. Emmett M. Smith carried out optical tests on dry ABO secretion which we provided for him. Gary Charlton kindly prepared and contributed the pH glass microelectrode. Richard Srna provided helpful suggestions on the calibration and use of the chloride microelectrode. John Sherman and Takako Nagasi assisted with the EDAX analysis of dry secretion. Walter S. Kay prepared final prints of Fig. 1-11 for publication. Howard H. Chauncey, Julius Gordon, Myroslaw G. Harasewych, Philip Person, and Karl M. Wilbur kindly reviewed the manuscript and offered many _ valuable suggestions. Research was supported in part by Public Health Service Research Grant DE 01870 from the National Institute of Dental Research, and in part by a grant from the University of Delaware Research Foundation. Part of the Research was per- formed during the tenure of the senior author in the Systematics-Ecology Program, Marine Biological Laboratory, Woods Hole, Massachusetts. LITERATURE CITED BARKA, T. & ANDERSON, P. J., 1963, Histo- chemistry. Theory, Practice and Bibliography. Harper & Row, New York, 660 p. BERNARD, F. R. & BAGSHAW, J. W., 1969, Histology and fine structure of the accessory boring organ of Polinices lewisi (Gastropoda, Prosobranchia). Journal of the Fisheries Research Board of Canada, 26: 1451-1457. CARRIKER, M. R., 1977, Ultrastructural evidence that gastropods swallow shell rasped during hole boring. Biological Bulletin, 152: 325-336. CARRIKER, M. В. & CHAUNCEY, H. H., 1973, Effect of carbonic anhydrase inhibition on shell penetration by the muricid gastropod Urosalpinx стегеа. Malacologia, 12: 247-263. САВВ!КЕВ, М. R., SCOTT, D. В. € MARTIN, С. N., 1963, Demineralization mechanism of boring gastropods. In: SOGNNAES, R. F., Ed., Mechanisms of Hard Tissue Destruction. American Association for the Advancement of Science, Publication 75, p. 55-89. CARRIKER, M. R., SMITH, E. H. & WILCE, R. T., 1969, Penetration of calcium carbonate substrates by lower plants and invertebrates, an international multidisciplinary symposium. American Zoologist, 9: 629-1020. CARRIKER, M. R. & VAN ZANDT, D., 1964, Use of polished mollusk shell for testing demineralization activity of accessory boring organ of muricid boring gastropds. Biological Bulletin, 127: 365 (abstract). CARRIKER, M. R. & VAN ZANDT, D., 1972a, Regeneration of the accessory boring organ of muricid gastropods after excision. Transactions of the American Microscopical Society, 91: 455-466. CARRIKER, М. В. & VAN ZANDT, D., 1972b, Predatory behavior of a shell-boring muricid gastropod. In: WINN, H. E. € OLLA, В. L., Eds, Behavior of Marine Animals: Current Perspectives in Research, vol. 1, Invertebrates. Plenum, New York, p. 157-244. CARRIKER, M. R. & VAN ZANDT, D., 1973, Activity of the marine gastropod Urosalpinx cinerea in the absence of hibernation. Chesa- peake Science, 14: 285-288. CARRIKER, M. R., VAN ZANDT, D. & CHARL- TON, G., 1967, Gastropod Urosalpinx: pH of accessory boring organ while boring. Science, 158: 920-922. CARRIKER, M. R., VAN ZANDT, D. & GRANT, T. J., 1972, Mechanism of shell penetration by the boring muricid gastropod Urosalpinx cinerea. In: CARRIKER, М. R., Ed., A Decade of Whole Organism Biology. Systematics-Ecology Program, Marine Biologi- cal Laboratory, Woods Hole, Massachusetts, p. 16-18. 142 CARRIKER, WILLIAMS AND VAN ZANDT CARRIKER, М. В. & WILLIAMS, L. С., 1978, The chemical mechanism of shell dissolution by predatory boring gastropods: a review and an hypothesis. Malacologia, 17: 143-156. CHETAIL, M., BINOT, 0. & BENSALEM, M., 1968, Organe de perforation de Purpura lapillus (L.) (Muricidae): histochimie et histoenzymologie. Cahiers de Biologie Marine, 9: 13-22. CHETAIL, M. & FOURNIE, J., 1969, Shell-bor- ing mechanism of the gastropod, Purpura (Thais) lapillus: a physiological demonstration of the role of carbonic anhydrase in the dis- solution of CaC03. American Zoologist, 9: 983-990. CHETAIL, M. & FOURNIE, J., 1970, Mecanisme de perforation chez Thais lapillus L. (Gas- teropode Prosobranche, Muricide): mise еп evidence d'une entree d'ions calcium durant l'activite de l'organe de perforation. Comptes Rendus de l’Academie des Sciences, Paris, 271: 118-121. DAY, J. A., 1969, Feeding of the cymatiid gas- tropod, Argobuccinum argus, in relation to the structure of the proboscis and secretions of the proboscis gland. American Zoologist, 9: 909-916. GOLUBIC, S., PERKINS, R. D. & LUKAS, K. J., 1975, Boring microorganisms and micro- borings in carbonate substrates. In: FREY, R. W., Ed., The Study of Trace Fossils. Springer- Verlag, New York, p. 229-259. JACCARINI, V., BANNISTER, W. H. & MICALLEF, H., 1968, The pallial glands and rock boring in Lithophaga lithophaga (Lamel- libranchia, Mytilidae). Journal of Zoology, 154: 397-401. LEHNINGER, A. L., 1970, Biochemistry. Worth, New York, 833 p. MILLIMAN, J. D., 1974, Marine Carbonates, Part 1, Sedimentary Carbonates. Springer-Verlag, New York, 375 p. NYLEN, М. U., PROVENZA, D. V. & CAR- RIKER, M. R., 1969, Fine structure of the accessory boring organ of the gastropod, Uro- salpinx. American Zoologist, 9: 935-965. PERSON, P. A., SMARSH, A., LIPSON, S. J. & CARRIKER, M. R., 1967, Enzymes of the accessory boring organ of the muricid gas- tropod Urosalpinx cinerea follyensis, |. Aerobic and related oxidative systems. Biologi- cal Bulletin, 133: 401-410. SMARSH, A., CHAUNCEY, H. H., CARRIKER, М. В. & PERSON, P., 1969, Carbonic anhy- drase in the accessory boring organ of the gastropod, Urosalpinx. American Zoologist, 9: 967-982. WARME, J. E., 1975, Borings as trace fossils, and the processes of marine bioerosion. In: FREY, R. W., Ed., The Study of Trace Fossils. Springer-Verlag, New York, p. 181-227. WHITE, A., HANDEER, P. € SMITH, ЕЁ ER 1968, Principles of Biochemistry. McGraw Hill, New York, 1187 p. WHITFIELD, M., 1971, lon selective electrodes for the analysis of natural waters. Australian Marine Science Association Handbook 2, Hydrographic Office, R.A.N., Garden City, New South Wales, Australia. ZOTTOLI, R. A. & CARRIKER, M. R., 1974. External release of protease by stationary burrow-dwelling polychaetes. Journal of Marine Research, 32: 331-342. MALACOLOGIA, 1978, 17(1): 143-156 THE CHEMICAL MECHANISM OF SHELL DISSOLUTION BY PREDATORY BORING GASTROPODS: A REVIEW AND AN HYPOTHESIS" Melbourne R. Carriker and Leslie G. Williams College of Marine Studies, University of Delaware, Lewes, Delaware 19958, U.S.A. ABSTRACT The results of investigations of the authors and other researchers on predatory boring gastropods reported in earlier publications are analyzed to formulate a preliminary hypo- thesis on the physiological-chemical mechanism.of shell penetration by boring gastropods. The authors hypothesize that boring muricid and naticid gastropods employ a combi- nation of enzymes, an inorganic acid, and chelating agents in a hypertonic medium to facilitate dissolution of shell and intracellular transport of calcium during the chemical phase of valve penetration. Other chemicals, as yet unidentified, possibly also take part. Secretory granules and vesicles in the secretory epithelium of the accessory boring organ and in secretion released externally, organic matter in the secretion, and dissolution by the secretion of the insoluble fraction of the organic matrix of shell suggest the involvement of at least 1 enzyme in the hydrolysis of the organic portion of the shell. The enzyme carbonic anhydrase catalyzes the hydration of CO2 and is present in both the secretory epithelium and the free secretion. Hydrogen and chloride ions identified with microelectrodes, the chloride ions confirmed with energy dispersive x-ray analysis, suggest the role of HCl as one of the solubilizers of the mineralized component of shell. A chelating agent, localized in the secretory epithelium but not yet chemically identified, and a mucoprotein, also in the secretory epithelium, perhaps participate in sequestering calcium and other cations in the shell. Chelators possibly also function in removing soluble fractions of the organic matrix of shell. INTRODUCTION In other publications (Carriker, Van Zandt & Grant, 1972; Carriker, Williams & Van Zandt, 1978) we reported the results of a preliminary characterization of the secretion of the accessory boring organ (ABO) of the shell-boring prosobranch gas- tropod Urosalpinx cinerea follyensis Baker. In this paper we have drawn on these find- ings and those of earlier investigators to formulate an hypothesis on the physiologi- cal-chemical mechanism of shell penetration by boring gastropods (Carriker & Smith, 1969; Carriker € Van Zandt, 1972a; Carriker, Smith & Wilce, 1969; Carriker, Person, Libbin & Van Zandt, 1972). As background for this synthesis we include here a summary of existing knowledge of the secretory function of the ABO, chemi- cal aspects of penetration by other inver- tebrate calcibiocavites, and dissolution of bone by vertebrate osteoclasts. The only shell-boring gastropods in which the ABO has been examined in any detail are Urosalpinx cinerea (Say), Nucella lapillus (Linne), and Polinices lewisi (Gould). Probably because of the difficulty of the subject and the small size of the organisms, no one has yet investigated the physiological mechanism of penetration of calcareous substrates by Algae and Fungi (Golubic et al., 1975). LITERATURE REVIEW Cytology The secretory disc of the ABO of boring gastropods is composed of tall columnar epithelial cells arranged in groups. Each cell within a group undergoes primary branch- ing to form slender interlocking projections (Nylen et al., 1969). These projections undergo secondary branching to form a 1University of Delaware College of Marine Studies Contribution No. 112. (143) 144 CARRIKER AND WILLIAMS brush border of unusually long, densely packed microvilli (each about 0.1 ит in diameter) at the surface of the secretory disc (Fig. 1). Cell groups are further char- acterized by nerve fibers which penetrate the group basally and terminate in the region of primary branching. Each cell group is delimited by a membrane, the basal lamina, and is distinguished from neigh- boring cell groups by an interstitial space. The interstitial spaces contain muscles, capillaries, and nerve fiber bundles. Indi- vidual nerve fibers terminate on the mus- cles but not on the capillaries. The inter- stitial space in active ABOs widens to form a distal sinus at the level of primary branching (Carriker, 1943; Carriker & Van Zandt, 1972a; Fretter, 1946; Nylen et al., 1969; Derer, 1975). The secretory epithelium of the ABO possesses cytological features characteristic of highly active tissues. Although the fun- MV ZA SGA PB O FIG. 1. damental architecture of the secretory epithelium is the same in both active and resting ABOs, the epithelium of active ABOs is greatly amplified and more highly differentiated than their resting counter- parts. Active ABOs are characterized by increased length of the epithelial cells and their microvilli as well as by enlargement of the distal interstitial sinus (Nylen et al., 1969). Dense populations of mitochondria, especially prominent towards the distal end of secretory cells, are more abundant in active than in resting ABOs. In active ABOs most of the mitochondria in the distal por- tion of the cell are distributed in an area of cell contact with the hemocyanin-packed distal sinus while some smaller mito- chondria are located subjacent to the zone of primary branching. There is increased differentiation, especially of Golgi com- plexes, during periods of active boring. Conspicuous in the cytoplasm are dense Schematic diagram of distal portions of 2 cells in adjacent cell groups of the secretory epithelium of the accessory boring organ of Urosa/pinx cinerea. Individual cell groups are bound by a membrane, the basal lamina (BL). An interstitial space and distal sinus (DS) separate cell groups from one another. The labeled interdigitating projections (IP) in the zone of primary branching (PB) originate from parent cells out of the plane of focus in the diagram. Terminal bars (TB) are areas of cellular contact between interdigitating projections. The region of formation of terminal bars is proximal to the microvillar zone (MV) and is designated as the zona adherens (ZA). Mitochondira (Mi) are distributed next to the distal sinus and in the region of primary branching. Secretion granules (SG) are depicted in the region of primary branching and are presumably transported to this region from the proximal end of the cell. Nomenclature after Nylen et al. (1969). CHEMICAL SHELL-BORING BY GASTROPODS 145 membrane-bound secretory granules’ of rather uniform size, averaging about 0.2 um in diameter; other inclusions comprise mul- tivesicular bodies and single vesicles. Fusion of the secretory granules occurs in the zone of primary branching (Nylen et al., 1969). Histochemistry The following reports deal with intra- cellular substances in the gastropod ABO which seem to be concerned with secretion of shell-dissolving substances. In histo- chemical studies of the ABO of Urosalpinx cinerea, Person et al. (1967) discovered that cytochrome oxidase, succinate dehy- drogenase, and lactate dehydrogenase acti- vities are localized in the secretory cells. Oxidase activities of mitochondria-rich par- ticulate fractions isolated from homo- genates of whole active ABOs are exceed- ingly high and almost twice those of inactive ABOs. Chetail & Binot (1967) and Smarsh et al. (1969) determined histo- chemically (Häuslers method) that т Nucella lapillus and U. cinerea (гезрес- tively) carbonic anhydrase is present both in active and inactive ABOs as well as in other secretory tissues of the snails. The greatest carbonic anhydrase activity occurs in the proximal and microvillar zones of the secretory cells. Carbonic anhydrase activity is inhibited by sodium acetazol- amide. Initial histochemical research by Chetail & Fournie (1970) demonstrated the presence of adenosine triphosphatase in the distal rim of secretory cells and adjacent to limiting membranes of cell groups. Pre- liminary histochemical studies by Smarsh (personal communication) revealed strong aminopeptidase activity primarily in the epithelial cells of the ABO stalk. Chetail, Binot & Bensalem (1968) found, also histo- chemically, that lipase and alkaline phos- phatase are present in both active and inactive ABOs. Zottoli & Carriker (1974), using hide powder azure to search for a possible protease, did not find convincing amounts in whole excised ABOs of U. cinerea. Also histochemically, Smarsh et al. (1969) and Chétail & Fournie (1970) found calcium in red-staining granules, or microbodies, in active ABOs, but generally not in inactive ABOs, of Urosalpinx cinerea and Nucella lapillus, respectively. The cal- cium deposits found in N. /apillus are located at the base of the microvilli and in the zone of primary branching. A strong cobalt-binding material, possibly a chelator, was localized histochemically by Smarsh et al. (1969) in the microvillar region of the ABO of U. cinerea. The staining reaction is strongest in active ABOs, and there is little or no staining in inactive ones. Studies of the ABO of Nucella lapillus showed that secretory cells of the inactive gland are rich in glycogen, whereas those of the active gland are poor in this storage carbohydrate (Chetail, Binot & Bensalem, 1968). Nylen et al. (1969) in an ultra- structural investigation of the ABO of Urosalpinx cinerea confirmed these observa- tions. Bernard & Bagshaw (1969) estab- lished histochemically that the ABO of the naticid snail Polinices lewisi consists of 2 distinct epithelial regions, a peripheral one producing mucin and a central one elabo- rating a mucoprotein. This anatomical con- dition contrasts with that of the muricid ABO which contains only the central secretory disc (Carriker & Van Zandt, 1972a). ABO Secretion The following papers are concerned with substances in the released secretion of the gastropod ABO which are possibly involved with shell dissolution. Carriker, Van Zandt & Grant (1972) reported that freshly released secretion from normally extended ABOs of Uro- salpinx cinerea is highly viscid, microscopi- cally granular, varies in color from colorless to pale violet depending on the individual and that much of the secretion does not dissolve in seawater. The pH of ABO secretion was measured in Urosalpinx cinerea with a glass micro- electrode (Carriker, Van Zandt & Charlton, 1967). Continuous recording shows a pH range of 3.8 to 4.1. The cymatiid gas- {город Argobuccinum argus _ secretes Н2$04 (0.4 to 0.5 N) for boring into tubicolous polychaetes (Day, 1969). In this case, the acid is produced not by an ABO but by a large proboscis gland. Blots of secretion from normally ex- tended ABOs of Urosalpinx cinerea col- lected on filter paper, as well as secretion on the exterior of the ABO in histochemi- cal sections of the foot, test positively for carbonic anhydrase. The _ carbonic anhydrase activity of the secretion blots is 146 CARRIKER AND WILLIAMS inhibited by sodium acetazolamide (Car- riker & Chauncey, 1973). Secretion also contains cytochrome oxidase, a finding of considerable interest since this enzyme, as well as carbonic anhydrase, is almost always intracellular (Person et al., 1967). Secretion of whole active excised ABOs and halves of active excised ABOs etch polished mollusc shell. Dissolution com- mences as soon as the excised gland is applied to polished shell and continues at a greatly reduced rate for several hours. Con- trol tissues from other parts of the snail do not dissolve polished shell (Carriker, Scott & Martin, 1963; Carriker & Van Zandt, 1964). Excised ABOs heated to about 80°C (Carriker, Williams & Van Zandt, 1978), and purified bovine carbonic anhydrase likewise do not etch shell (Carriker & Chauncey, 1973). Excised ABOs treated with papain did not etch polished shell while those treated with trypsin retained their etching capabilities. Papain and trypsin solutions alone did not etch polished shell (Carriker, Williams & Van Zandt, 1978). Behavioral Experiments Carriker & Van Zandt (1972b) demon- strated during normal boring in prey shell that the ABO of live Urosalpinx cinerea dissolves shell in the incomplete borehole for periods ranging from a few minutes to about an hour. The relatively long periods of chemical activity alternate with brief periods of radular rasping. Rate of normal shell penetration by live snails, 0.3-0.5 mm per day, is reduced by sodium aceta- zolamide, an inhibitor of carbonic anhy- drase, in both U. cinerea (Carriker & Chauncey, 1973) and Nucella lapillus (Chétail & Fournie, 1969) when the in- hibitor is added to seawater in which the gastropods are maintained. Addition one at a time of CO», NaCl, and KCI to the sea- water accelerates the rate of hole-boring by N. lapillus (Chetail & Fournié, 1969). Comparative Calcibiocavitology Other Invertebrates. Chemical aspects of penetration of calcareous substrata by other invertebrate calcibiocavites have also received some attention. In the bivalve Lithophaga lithophaga histochemical evi- dence points to secretion by the pallial glands of a neutral mucoprotein, which is believed to complex calcium; apparently no acid mucopolysaccharides occur, and there is no indication of an acid (Jaccarini et al., 1968). Carbonic anhydrase is implicated in the mechanism of shell penetration by the burrowing sponge С/опа celata (Hatch, 1975) and by the burrowing barnacle Trypetesa nassarioides (Тигашег, 1968). Hatch (1975) postulated that the primary mechanism for dissolution of CaCO3 by С. celata involves a shift in the carbonate solubility product in the microenvironment of etching cells of the sponge. This shift is mediated through activity of carbonic anhydrase, producing hydrogen ions by hydration of CO2 within etching cells and possibly providing an optimal pH for break- down of the organic matrix of shell. Silen (1956) observed that quantitative analysis of dry shell invaded by the shell-burrowing bryozoan Penetrantia densa shows a higher content of phosphate ions than uninvaded shell. Finding no other differences, he con- cluded that phosphoric acid is used in shell penetration by this species. It is not clear whether account was taken of phosphorus of bryozoan soft tissues. Vertebrate Osteoclasts. Because osteo- clasts attack calcareous substrates, it is important to review briefly what is known about the process of bone dissolution by them as a possible aid in explaining shell penetration by gastropod ABOs. Although investigators have shown beyond a reason- able doubt that vertebrate osteoclasts resorb bone, the mode of action is still unproved (Hancox, 1972). The osteoclast is a large, generally multi- nucleated single cell possessing a prominent ruffled brush border and many membrane- bound vesicles in the cytoplasm which mor- phologically resemble lysosomes. Proof that these bodies contain lytic enzymes is still lacking, nor have enzymes been demon- strated in the altered bone near the ruffled border, but it seems likely that they are involved in degradation of the organic matrix. Carbonic anhydrase, cytochrome oxidase, and succinic dehydrogenase, as well as the hydrolases, acid phosphatase, leucine aminopeptidase, b-galactosidase, and b-glucosidase, have been identified in the cells. Acid hydrolases are of the lysosomal type such as might be involved in the breakdown of bone matrix (Hancox, 1972; Minkin & Jennings, 1972). Substrates. Secretion of the ABO dis- solves shell which consists of calcium car- CHEMICAL SHELL-BORING BY GASTROPODS 147 bonate crystals in an organic matrix of protein (some of it tanned), mucopolysac- charide, lipid, and glycoprotein (Grégoire, 1972: Wilbur, 1972); secretion of the osteoclast solubilizes bone which contains a calcium phosphate complex in an organic matrix of collagen fibers, mucopolysac- charides, and other substances (Pritchard, 1972). Ultrastructural studies suggest that ABO secretion initially dissolves the non- mineralized organic matrix surrounding shell units and then attacks the mineral crystals (Carriker, 1969), whereas osteo- clasts appear to remove bone mineral first and dissolve collagen fibers later (Hancox, 1972). Although carbonic anhydrase and aminopeptidase, as well as the respiratory enzymes cytochrome oxidase and succinic dehydrogenase, have been identified т both the ABO and the osteoclast, no mineral acids or chelators have been found in osteoclasts. DISCUSSION AND CONCLUSIONS Although still incomplete, information reported in the literature (Table 1) suggests a tentative hypothesis on the chemical mechanism of shell penetration by the ABO of muricid and naticid gastropods. We hypothesize that a combination of enzymes, an inorganic acid, and a chelating agent is employed in a hypertonic secretion to facilitate dissolution of shell and intra- cellular transport of calcium during the chemical phase of valve penetration. The hypothesis is based on cellular, histochemi- cal, and physiological observations and experimentation. The following discussion explores the shell-penetrating mechanism of boring gastropods in the context of this hypothesis. Our primary aim is to provide a conceptual framework that will stimulate further research toward elucidation of the chemical mechanism(s) of penetration of calcareous substrates by calcibiocavites (Carriker & Smith, 1969). Secretory Epithelium Microvilli facilitate exchange of soluble substances between secretory cells and their surroundings, and it is perhaps this purpose which is served by the unusually long microvilli of ABOs (Nylen et al., 1969). The ultrastructural organization of solute- transporting tissues is distinguished by a series of epithelial dead-end channels such as lateral spaces, basal infoldings, or intra- cellular canaliculi, and an active cation pump which parallels the distribution of mitochondria at the ion exchange site (Hochachka & Somero, 1973). The ABO secretory epithelium is clearly organized into 2 series of dead-end channels (Fig. 1). The hemocyanin-filled interstitial spaces and their terminal distal sinuses compose the somatic dead-end channels. The micro- villi compose the dead-end channels on the lumenal side of the secretory disc. This organization, together with the localization of mitochondria adjacent to both the distal sinus and immediately below the microvilli, is morphological evidence for ion exchange as well as secretory capabilities of the ABO epithelium. The presence of glycogen in the secretory epithelium of the resting ABO and its disappearance from active ABO tissue suggests rapid mobilization of gly- cogen for the synthesis and release of secretory products as well as the formation of adenosine triphosphate (ATP) to fuel ionic exchange pumps at the somatic and lumenal sites discussed above (Chetail & Fournié, 1969; Nylen et al., 1969). Histo- chemical identification of adenosine tri- phosphatase (ATPase) adjacent to the inter- stitial space in active ABOs is therefore a logical sequel to this generalization (Chetail & Fournié, 1970). The 2-fold increase in cytochrome oxidase activity in active, as opposed to resting, ABOs is likewise in agreement with the mobilization of gly- cogen and appearance of ATPase activity. Carbonic anhydrase present both within the secretory epithelium of the ABO and in the free secretion of the gland (Chetail & Fournié, 1969; Smarsh et al., 1969; Car- riker & Chauncey, 1973) catalyzes the hydration of С02 to HCO3 and H*. Intra- cellular carbonic anhydrase furnishes hydro- gen ions for secretion and perhaps is also involved in ion transfer (Koefoed-Johnsen & Ussing, 1960; Carter, 1972; Minkin & Jennings, 1972). The intracellular carbonic anhydrase activity possibly also serves as a mechanism for the production of high chlo- ride ion concentrations in the free secretion (Carriker, Williams & Van Zandt, 1978). Bicarbonate ions, their production cata- Iyzed by carbonic anhydrase, perhaps serve as counter ions for the intracellular trans- port of СГ from the blood plasma (Schof- feniels & Gilles, 1972). The extracellular transport of CI” could then follow the 148 CARRIKER AND WILLIAMS TABLE 1. Summary of information on the muricid-naticid accessory boring organ relative to shell dissolution. Conditions more pronounced in active than in inactive ABOs are indicated by **. Conditions more pronounced in inactive than in active ABOs are marked *. Conditions similar in active and inactive ABOs and cases where the distinction is not clear have no markings. 1. Cytology of secretory epithelium: Clusters of long secretory cells heavily sup- plied with capillaries and nerves Dense brush border of long slender micro- villi (0.1 um in diameter) **Dense populations of mitochondria **Golgi complexes (Derer, 1975; Nylen et al., 1969) 2. Physiology of ABO: a. General: Relatively long chemical period com- pared to short rasping period (Car- riker & Van Zandt, 1972b) Excised whole and half ABOs dissolve shell in normal and СО2-Нее atmosphere (Carriker & Van Zandt, 1972b) Bulk of secretion released into bore- hole initially, minor amount secre- ted thereafter (Carriker, Williams & Van Zandt, 1978) НС! and glucose do not stimulate with- drawal of ABO; distilled water does (Carriker, Williams & Van Zandt, 1978) b. Secretory epithelium: 1) Substances identified within it: **Cytochrome oxidase (Person et al., 1967) N Carbonic, anhydrase (Chetail & Fournie, 1969; Smarsh et al., 1969) Succinate dehydrogenase (Person et al., 1967) Lactate dehydrogenase (Person et al., 1967) Aminopeptidase (Smarsh, communication) Adenosine triphosphatase (Chetail & Fournie, 1970) Alkaline phosphatase (Chetail, Binot & Bensalem, 1968) Lipase (Chetail, Binot & Bensalem, 1968) **Chelator (Smarsh et al., 1969) PAS positive (mucoproteins, neutral mucopolysaccharides) (Bernard & Bagshaw, 1969; Carriker, Van Zandt & Grant, 1972) No acid mucopolysaccharides (Car- riker, Van Zandt & Grant, 1972) **Calcium (Chetail & Fournie, 1970: Smarsh et al., 1969) *Glycogen (Chetail, Binot & Ben- salem, 1968; Nylen et al., 1969) 2) Particles observed within it: Secretion granules (0.2 um average), multivesicular bodies, vesicles (Bernard & Bagshaw, 1969; Derer, 1975; Nylen, et al., 1969) c. Secretion released by secretory epithelium: personal 1) Characteristics: Viscid, granular, generally insoluble in seawater and in distilled water, highly hygroscopic, about 65% volatile, ‘'crystallizes’’ after exposure to atmospheric con- densate (Carriker, Van Zandt & Grant, 1972; Carriker, Williams & Van Zandt, 1978) 2) Substances identified in it: Organic matter and МаС! (Е. М. Smith, personal communication) Cytochrome oxidase (Person et al., 1967) Carbonic anhydrase (Carriker & Chauncey, 1973) Hydrogen ions (pH 3.8-4.1 with trace seawater; 5.2 pure) (Carriker, Van Zandt & Charlton, 1967; Car- riker, Williams & Van Zandt, 1978) Chloride ions (Carriker, Williams & Van Zandt, 1978) Sodium ions (Carriker, Williams & Van Zandt, 1978) Calcium ions, trace (Carriker, Wil- liams & Van Zandt, 1978) Potassium ions, trace (Carriker, Williams & Van Zandt, 1978) 3) Particles observed in it: Secretion granules (0.17-0.40 ит), SEM (Carriker, Williams & Van Zandt, 1978) Vesicles (0.04-0.06 um), SEM (Car- riker, Williams & Van Zandt, 1978) Red particles, trichrome stain (Car- riker, Williams & Van Zandt, 1978) 3. Factors affecting shell dissolution by ABO secretion: a. Accelerate: Increasing temperature (Carriker, Van Zandt & Grant, 1972) СО2, KCI, NaCl (live snails) (Chetail & Fournie, 1969) b. Decelerate: Sodium acetazolamide, in culture (Car- riker & Chauncey, 1973; Chetail & Fournie, 1969) c. Inhibit: Papain (excised ABO) (Carriker, Williams & Van Zandt, 1978) Heat (80°C) (excised ABO) (Carriker, Scott & Martin, 1963; Carriker, Williams & Van Zandt, 1978) Sodium acetazolamide, т shell-ABO preparation (Carriker & Chauncey, 1973) d. Uncertain effect: СО2 (excised ABO) (Carriker, Williams & Van Zandt, 1978) Trypsin (excised ABO) (Carriker, Wil- liams & Van Zandt, 1978) electrochemical gradient set up by the active extracellular transport of either Nat ог H* (Carriker, Williams & Van Zandt, 1978) or, alternatively, be exchanged for CHEMICAL SHELL-BORING BY GASTROPODS 149 HCO3 in the free secretion (Cotlove & Hogben, 1962; Rehm, 1972). Production of НС! by parietal cells of the gastric mucosa of vertebrates is visualized as being car- bonic anhydrase-dependent with separation of hydrogen and hydroxyl ions; the secreted hydrogen ions are linked to chloride ions (Maren, 1967). The extra- ordinary density of mitochondria and numerous microvilli in parietal cells (Ito & Winchester, 1963) is reminiscent of these features in the gastropod ABO (Nylen et al., 1969). Active transport of chloride and sodium ions also occurs in the aqueous humor of vertebrates where secretion of chloride ions (with or without sodium ions) is also blocked by carbonic anhydrase inhibitors (Garg & Oppelt, 1970). Secretion Membrane-bound secretion granules and vesicles in the secretory epithelium of the ABO (Nylen et al., 1969; Bernard & Bag- shaw, 1969; Derer, 1975) and in secretion blots (Carriker, Williams & Van Zandt, 1978) are possibly the extracellular carriers of carbonic anhydrase as well as other pre- cursors of the active dissolving agent in extracellular use. This view is consistent with Hatch’s (1975) observation that car- bonic anhydrase in Cliona celata is particle- or membrane-bound, as well as the fact that cross-membrane transport of such a large molecule as carbonic anhydrase (molecular weight, 30,000) or cytochrome oxidase (240,000) is unlikely. Although fusion of the secretion granules with the plasma membrane has been observed (Nylen et al., 1969), identification of their chemical constituents has not yet been accomplished. The ABO secretion is a complex chemi- cal mixture which is most probably elaborated by several pathways within the secretory epithelium. However, elaboration pathways and routes of extracellular dis- charge of secretory products are not yet clear. The brush border appears to serve as the surface of release (Nylen et al., 1969; Derer, 1975). The role of the secretory epithelium is complicated in that it is in- volved in the intracellular transport of cal- cium as well as in the production of secre- tion. The process of solubilization of mineral crystals and organic matrix during shell penetration can therefore be best understood by considering the activities of the ABO secretion and the secretory epithelium as being multiphasic interde- pendent processes. We suggest that a com- bination of enzymes and chelator(s), together with acidic and saline properties of the secretion, function in shell penetra- tion and calcium transport into the secre- tory epithelium. Enzymes The pure ABO secretion is a slightly acidic (pH = 5), hypertonic (NaCl), enzyme- bearing (carbonic anhydrase) fluid (Car- riker, Van Zandt & Grant, 1972). Enzymes or proteinaceous material other than car- bonic anhydrase is possibly present in the secretion as well. Cytoplasmic localization of aminopeptidase in the stalk epithelium (Smarsh, personal communication), organic matter in the secretion (Smith, personal communication), and inactivation of shell- dissolving activity of excised ABOs by heat and papain suggest this view (De Duve & Wattiaux, 1966; Scott, 1967; Carriker, Scott & Martin, 1963; Carriker, Williams & Van Zandt, 1978). That HCI alone is not responsible for shell dissolution is indicated by patterns of dissolution: HCl solubilizes the mineral crystals first when applied to a shell surface, whereas the ABO secretion tends to dissolve the organic matrix first (Carriker & Van Zandt, 1972b). Carbonic anhydrase by itself does not etch shell (Carriker & Chauncey, 1973). We postulate, therefore, that an enzyme or a chelator is involved in digestion of the fraction of organic matrix (nacrosclerotin and nacroin, Grégoire, 1972) insoluble in water and EDTA, and possibly in the dissolution of the mineral components of shell. Enzymes have not yet been identified, but perhaps are conchiolinase-like mole- cules which catalyze removal of the organic components of shell (conchiolin and peri- ostracum) (Carriker, 1969; Nylen et al., 1969; Travis & Gonsalves, 1969) prior to dissolution of the mineral component. Action of the ABO secretion is directed 1st toward dissolution of the organic matrix that surrounds each shell unit (prisms, lamellae; Taylor et al., 1969) and then to the intraunit organic matrix and its con- stituent CaCO3 crystals. Hydrolytic enzymes are assumed to function similarly in removal of organic matrix during resorp- tion of bone (Vaes, 1968). Hydrogen ions secreted by the ABO (Carriker, Van Zandt 150 CARRIKER AND WILLIAMS & Charlton, 1967; Carriker, Williams & Van Zandt, 1978) possibly contribute toward maintenance of optimal pH for enzymatic activity. Acid mucopolysaccharides are probably not involved in dissolution as they have not yet been identified in the ABO (Carriker, Williams & Van Zandt, 1978). Chelators А water-soluble, acetone-precipitable chelator present in the microvillar zone of the ABO (Carriker & Smith, 1969; Smarsh et al., 1969; see also Jenkins & Dawes, 1963) perhaps contribute to dissolution of the water-soluble nacrin in the organic matrix of shell as well as the mineral crystals. The insoluble nacrosclerotin and nacroin are presumably attacked by hydro- lytic enzymes (Grégoire, 1972). This suggestion is supported by Travis & Gon- salves (1969) who found that the organic matrix within shell units is soluble in water and in dilute concentrations of EDTA. They also demonstrated that organic sheaths of shell units solubilize with pro- longed treatment in mild concentrations of EDTA while a 3rd fraction is insoluble in prolonged treatment. Inorganic Acid The presence of hydrogen ions (Carriker, Van Zandt & Charlton, 1967; Carriker, Williams & Van Zandt, 1978) and chloride ions (Carriker, Williams € Van Zandt, 1978) in the free secretion, and of the chelator in the microvillar zone of the ABO, suggest that НС! perhaps functions in combination with the chelator to dissolve the CaCO3 in shell and bind calcium as a water-soluble compound for rapid removal. The pH 5 of the secretion corresponds to an H* concentration of 10-5М. The CI concentration reached levels as high as 1.71 М, suggesting that if it is secreted as HCl, the H* is used up and CI” concentration increases with time. This is in agreement with our results (Carriker, Williams & Van Zandt, 1978). Since methods used in measuring in vivo hydrogen and chloride ion concentrations allowed access to the secretion prior to, or independent of, its application to a CaCO3 substrate, one could argue that the disparity in H* and CI” concentrations was due to independent mechanisms for the release of the ions, or a reaction necessary for activating a com- ponent of the secretion, but not directly participatory in the shell-dissolution mecha- nism. We reiterate that НС! alone is not responsible for dissolution of shell. Shell Dissolution Jaccarini et al. (1968) proposed that the burrowing bivalve Lithophaga lithophaga, which apparently does not secrete acid, penetrates shell by complexing calcium through secretion of neutral mucoprotein with calcium-binding ability. In the same мау mucoprotein in the gastropod ABO (Bernard € Bagshaw, 1969; Carriker, Williams & Van Zandt, 1978) possibly also participates in complexing calcium during hole-boring. In a recent review of calcium- binding proteins Kretsinger (1976) noted that vitamin D-induced calcium-binding proteins appear in high concentrations in the vertebrate gut, kidney, and avian egg- shell gland. Also, active accumulation of calcium in mitochondria is apparently driven by an anion and/or Ht gradient as opposed to the Ма* gradient that drives the Cat2 pump of axons (Kretsinger, 1976; see also Wasserman & Taylor, 1969). The possi- bility of calcium-binding glycoproteins has been discounted since goblet (mucous) cell formation in the ABO, at least in muricid gastropods, is not evident. That calcium ions freed from shell in the borehole do enter the microvilli of active ABOs and pass into the foot of the snail has been demonstrated by Smarsh et al. (1969) and Chetail & Fournié (1970). Calcium localized histochemically by Chetial & Fournie (1970) appeared at the base of the microvilli in the zone of pri- mary branching of the secretory epi- thelium. The transmembrane flux of cal- cium is probably aided by carbonic anhy- drase (Istin € Kirschner, 1968) and adenosine triphosphatase, and as shown for intestinal mucosal cells, passage of calcium from the secretory cells seems to be dependent on sodium (Birge et al., 1972). A similar active transport of calcium takes place in the calciferous glands of earth- worms (Crang et al., 1968), and across the isolated avian shell gland (Ehrenspect et al., 1971). Active ATPase-dependent secretion of hydrogen ions by the ABO could estab- lish an electrochemical gradient, which in turn is counter-balanced by the passive influx of calcium ions (Chetail & Fournie, 1970). Outflux of sodium ions in the free secretion may likewise contribute to CHEMICAL SHELL-BORING BY GASTROPODS 151 balancing the influx of calcium ions, the transport possibly being coupled to the movement of chloride ions as suggested by Binder & Rawlins (1973) in the intestinal mucosa and by Garg & Oppelt (1970) in the aqueous humor. If chloride ions from the blood plasma are exchanged for HCO3 generated by car- bonic anhydrase in the secretory epithelium, then several explanations could account for the appearance of high chloride ion concentrations in the free secretion: 1) chloride ion extrusion follows an electro- chemical gradient set up by secretion of hydrogen ions, 2) chloride ion extrusion follows an electrochemical gradient set up by active sodium ion transport, and 3) chloride ions are exchanged for HCO3 produced catalytically by carbonic anhy- drase in the free secretion. If chloride ion transport is strictly dependent on the hydrogen ion gradient, then, considering the quantities of СГ in the free secretion, it would be difficult to account for the sim- ultaneous intracellular transport of calcium. If chloride ions are exchanged for HCO3 generated by carbonic anhydrase or if chloride ion extrusion follows an electro- chemical gradient set up by active sodium ion transport, then the remaining electro- chemical gradient would owe its ion trans- port potential to the Ht gradient, a con- dition that would favor intracellular trans- port of Са+2. The substantial quantities of NaCl (Carriker, Williams & Van Zandt, 1978) and carbonic anhydrase (Smarsh et al., 1969) in the free secretion possibly thus represent the involvement of one or both of these processes and suggest a pro- fitable line for further research. Comparable exchange pumps have been described in the gill of teleost fish where the net extrusion rate of sodium is identical to the potassium influx (Maetz, 1969) and in the avian salt gland which is carbonic anhydrase- dependent (Hochachka & Somero, 1973). In the case of the avian salt gland intra- cellular hydrogen and bicarbonate ions are exchanged for blood sodium and chloride ions. The cellular NaCl is then actively eliminated by a Na-K ATPase pump (Hochachka & Somero, 1973). Problems which must be faced in postulating a cal- cium chelator in the free secretion are its stability constant in a hypertonic medium, and competing ions at low pH (Moeller & Horwitz, 1960). An explanation for dissolution of mineral crystals of shell by the ABO secre- tion following digestion of the organic matrix is suggested by examination of the chemical constituents of the secretion in the context of their combined influence on the apparent (stoichiometric) solubility product of calcium carbonate (Pytkowicz, 1969). The apparent solubility product (ionization product) of CaCO3 is related to the thermodynamic solubility product by the expression Kips Ksp/fcafcoz where, Kin = apparent solubility product Kun = thermodynamic solubility product ee = calcium activity coefficient о = Carbonate activity coefficient 3 (Pytkowicz, 1969). In seawater the solution and precipitation of calcium carbonate are dependent on chlorinity (more precisely ionic strength), carbonate alkalinity, and pH. The apparent solubility product of calcium carbonate is dependent on ionic strength, ion pairing, and pH. Increases in ionic strength and ion pairing, or decreases in pH, cause decreases in the activity coefficients f¿¿ and fco3 with a resultant increase in K’sp, the apparent solubility product. The ABO secre- tion accomplishes all 3 of these functions. The concentration of Cl (up to 1.7 M!) and Na‘, as measured by ion specific elec- trode and electron dispersive x-ray analysis, respectively, indicate a very high salt con- tent in the secretion. If Na* concentration is as high as Cl concentration, then the average ionic strength of the secretion based on these ions alone would be 1.1 (1.7 maximally). For purposes of com- parison, the ionic strength of seawater is approximately 0.7. Calcium and carbonate activity coefficients would further decrease due to the рН of the secretion and ion pairing of CO3 2 and HCO3 with Nat (Gar- rels & Thompson, 1962; Pytkowicz, 1969). Finally, removal of free calcium ions by either chelation or ion transport across the secretory epithelium, would further serve to decrease the activity coefficient of cal- cium and render an increase т the apparent solubility product of calcium carbonate. Increases in concentration of NaCl and KCI in seawater, in which living Nucella lapillus were boring, accelerated the rate of shell penetration (Chétail & Fournie, 1969). Perhaps these ions when applied 152 CARRIKER AND WILLIAMS externally also participate in transmem- brane ionic fluxes which contribute to shell dissolution (see also Tormey, 1966). How- ever, if ion uptake was generated by either а Na-K ATPase or a Na-H ATPase acti- vated pump, one would not expect, assum- ing sodium uptake, that the КС! solution would have the observed effect. In view of the chloride ion concentration of the secre- tion, the observed influence of both KCI and NaCl would be consistent with either a СГ uptake coupled to active uptake of Nat or a CI"/HCO37 exchange. Deceleration of the rate of boring by live snails in seawater solutions of sodium acetazolamide (Chetail & Fournie, 1970; Carriker & Chauncey, 1973) can be explained by inhibition of carbonic anhydrase activity and the hydra- tion of СО2 to HCO3 and Ht (Carter, 1972). Presumably, this inhibition occurs intracellularly and disrupts hydrogen ion secretion (Chétail & Fournie, 1969) and possibly the chloride ion-secreting mecha- nism (Carriker, Williams & Van Zandt, 1978). The carbon dioxide experiments performed by Chetail & Fournie (1969) and by Carriker, Williams & Van Zandt (1978) seem to have conflicting results. Increased CO» in seawater significantly increased the rate of boring by Nucella lapillus in Chetail and Fournié's (1969) behavioral experiment. Carriker, Williams & Van Zandt (1978) conducted a shell-etch- ing bioassay in CO>-free seawater and found no difference between etchings pro- duced by ABOs in CO>-free seawater and ABOs in untreated seawater. Chetail & Fournie (1969) surmise from the results of their experiment that increased CO; levels in seawater caused a corresponding increase in intracellular CO j and a subsequent increase in production of hydrogen ions through carbonic anhydrase _ catalysis. Obviously, Carriker, Williams & Van Zandt (1978) were looking for a diminution in etching activity as a result of substrate limi- tation of carbonic anhydrase activity. Their results suggest 2 alternative explanations: 1) residual intracellular CO» was being utilized and limitation of carbonic anhy- drase substrate was not achieved by decreasing environmental CO5; 2) the activity of carbonic anhydrase in the free secretion does not require a source of CO) independent of the CaCO3 substrate. The latter alternative is consistent with the find- ings that carbonic anhydrase is a necessary component of the secretion which cannot by itself dissolve CaCOz, but which in some way possibly interacts with solubi- lized CaCO3 and function in its removal (Hatch, 1975; Carriker, Williams & Van Zandt, 1978). Failure of the ABO to with- draw from the borehole when 1 М HCl (a concentration 1000 times higher than that expected at the pH 4 of the secretion but nearly isotonic with respect to the Cl con- centration of the secretion) was inadver- tently spilled on it, whereas it withdrew quickly when distilled water was applied, indicates that the ABO is conditioned to the HCl in its own secretion and was thus not irritated by an additional amount (Carriker, Williams & Van Zandt, 1978). Total inhibition of etching by heat in the excised ABO-polished shell preparation seems to have resulted not only from inactivation of enzymes but also from volatilization of other shell-dissolving sub- stances. Possible Organic Substances in Secretion The possibility that organic matter secreted by the ABO includes succinic acid or lactic acid, or both, should not be dis- counted, even though these acids are generally found in tissues. According to Simpson & Awapara (1966) and Stokes & Awapara (1968), the main product of metabolic glucose degradation in the molluscs examined was succinic acid with small amounts of lactic acid. Crenshaw & Neff (1969) demonstrated that succinic acid dissolved shell at ‘the mantle interface in closed clams (Mercenaria mercenaria). These acids, if present in the ABO secre- tion, could also function in shell dissolu- tion. Person et al. (1967) reported active succinate and lactate dehydrogenases in the ABO, confirming the presence of succinic and lactic acids in the secretory epithelium, but not necessarily in the free secretion. Vertebrate Osteoclasts Information on the chemical mechanism of bone dissolution by osteoclasts, like that of shell dissolution by ABOs, is still in the alpha stage, and thus does not help much in explaining shell dissolution by ABOs. However, it is worth noting that, as in all shell-boring and burrowing invertebrates so far described, carbonic anhydrase also appears to be involved in bone dissolution by osteoclasts. Furthermore, observations CHEMICAL SHELL-BORING BY GASTROPODS 153 suggest that osteoclastic hydrolases partici- pate in degradation of bone matrix (Han- cox, 1972), lending support to сисит- stantial evidence for possible involvement of hydrolases in the breakdown of the insoluble fractions of the organic matrix of shell. ABO Activity in Borehole Initial release of the bulk of the secre- tion by the ABO as it 1st enters the bore- hole is important in starting dissolution promptly over the entire bottom of the borehole. Continued, though reduced, secretion would appear useful in extending the period of dissolution. What determines the time of withdrawal of the ABO from the borehole at the termination of the chemical period is not known, but retrac- tion could be triggered by exhaustion of the shell-dissolving agent in the secretion. The highly viscid, partially insoluble nature of the secretion probably aids application to, and holding the secretion in, the bottom of the incomplete borehole by the ABO. Myroslaw Harasewych (personal com- munication) calculated that for a hypo- thetical borehole 1.0 mm in diameter and 3.0 mm deep (borehole of average size for an adult Urosalpinx cinerea), the volume of shell removed would be 2.35 тт?. Assum- ing the shell to be pure calcite (true in the case of the oyster, except for the organic matrix and trace minerals), 6.4 mg (0.064 mmol) of CaCO3 would have to be removed. This would require 4.67 mg of НС! or 7.6 mg of succinic acid, or 0.064 mmol of a divalent chelating agent, and not take into account removal of portions of the chemically weakened shell by radular activity. An enzyme, on the other hand, is capable of catalyzing a reaction at rates of 104 to 106 (3.6 X 106 in the case of carbonic anhydrase, isozyme C) moles of substrate per mole of enzyme per min when the substrate is in solution or in suspension. At the slower rate, dissolution of the 2.36 mm3 of shell in 1 minute would require the production of 1078 mole of a hypothetical shell-dissolving enzyme. Rate of shell penetration by U. cinerea normally ranges from about 0.3 to 0.5 mm per day (Carriker & Van Zandt, 1972b), and dissolution occurs only at the surface of shell in the borehole. In nature, excava- tion of a borehole 3 mm in diameter takes about 6 days at the fastest rate of penetration. Harasewych's calculations demonstrate the extreme difference in the quantities of different kinds of secretion required to perform equivalent dissolution of shell. However, since radular activity alternates with chemical activity in hole- boring, and removes 10 to 20% of the surface area of the bottom of the incom- plete borehole to a depth of a few ит at each rasping period (Carriker, 1969), smaller quantities of shell solvents possibly are required than would be the case if shell penetration were entirely chemical. CONCLUSIONS We hypothesize that a combination of enzymes, an inorganic acid, and chelating agents is employed in a hypertonic secre- tion to facilitate dissolution of shell and intracellular transport of calcium during the chemical phase of valve penetration by boring gastropods. Secretion granules and vesicles in the secretory epithelium of the ABO and in the released secretion, organic matter in the secretion, and dissolution by the secretion of organic matrix of shell, especially the insoluble fractions, provide circumstantial evidence for the presence of enzymes. Hydrogen, chloride, and sodium ion concentrations demonstrate the hypertonic and acidic characteristics of the released secretion. A chelating agent, though not chemically identified, and a mucoprotein have been found in the sec- retory epithelium; the latter perhaps func- tions in chelation. The physiological-chemical mechanism of shell-penetration could prove a valuable tool to those physiologists and cellular biologists interested in developing models for elucidating mechanisms of ion transport and exchange (hydrogen, chloride, sodium, calcium), acid secretion (HCI), carbonic anhydrase enzymology, dissolution of cal- careous substrates (dental caries), and cal- cium-binding proteins and chelators, within the context of their own disciplines. It is a curious fact that, with the excep- tion of 1 group of freshwater polychaetes (Caobangia; Jones, 1969), all shell-excavating species so far known are inhabitants of estuarine or marine habitats. Calcibiocavitic algae are likewise predominantly marine (Golubic, 1969; Golubic et al., 1975). These facts suggest that seawater and 154 CARRIKER AND WILLIAMS estuarine water are characterized by factors that have favored the evolutionary develop- ment of the shell-penetrating mechanism. A better understanding of the chemical mechanism of shell penetration would be enlightening with respect to these factors and subsequently allow postulation of methods for control of predatory gastro- pods and other shell-penetrating organisms harmful to shellfish mariculture. ACKNOWLEDGMENTS The research was supported in part by a grant from the University of Delaware Research Foundation. LITERATURE CITED BERNARD, F. R. & BAGSHAW, J. W., 1969, Histology and fine structure of the accessory boring organ of Polinices lewisi (Gastropoda, Prosobranchia). Journal of the Fisheries Re- search Board of Canada, 26: 1451-1457. BINDER, J. H. & RAWLINS, C. L. 1973, Elec- trolyte transport across isolated large intestinal mucosa. American Journal of Physiology, 225: 1232-1239. BIRGE, S. J., GILBERT, H. R. & AVIOLI, L. V., 1972, Intestinal calcium transport: the role of sodium. Science, 176: 168-170. CARRIKER, M. R., 1943, On the structure and function of the proboscis in the common oyster drill, Urosalpinx cinerea Say. Journal of Morphology, 73: 441-506. CARRIKER, M. R., 1969, Excavation of bore- holes by the gastropod, Urosa/pinx: an analy- sis by light and scanning electron microscopy. American Zoologist, 9: 917-933. CARRIKER, М. В. € CHAUNCEY, Н. H., 1973, Effect of carbonic anhydrase inhibition on shell penetration by the muricid gastropod Urosalpinx cinerea. Malacologia, 12: 247-263. CARRIKER, M.R., PERSON, P., LIBBIN, R. & VAN ZANDT, D., 1972, Regeneration of the proboscis of muricid gastropods after amputa- tion, with emphasis on the radula and car- tilages. Biological Bulletin, 143: 317-331. CARRIKER, M. R., SCOTT, D. B. & MARTIN, G. N., 1963, Demineralization mechanism of boring gastropods. In: SOGNNAES, В. F., Ed., Mechanisms of Hard Tissue Destruction. American Association for the Advancement of Science Publication 75, p. 55-89. CARRIKER, M. R. & SMITH, E. H., 1969, Com- parative calcibiocavitology: summary and con- clusions. American Zoologist, 9: 1011-1020. CARRIKER, M. R., SMITH, E. H. & WILCE, R. T., 1969, Penetration of calcium carbonate sub- strates by lower plants and invertebrates, an international multidisciplinary symposium. American Zoologist, 9: 629-1020. CARRIKER, M. R. & VAN ZANDT, D., 1964, Use of polished mollusk shell for testing demineralization activity of accessory boring organ of muricid boring gastropods. B/ological Bulletin, 127: 365 (abstract). CARRIKER, M. R. & VAN ZANDT, D., 1972a, Regeneration of the accessory boring organ of muricid gastropods after excision. Transactions of the American Microscopical Society, 91: 455-466. CARRIKER, M. R. & VAN ZANDT, D., 1972b, Predatory behavior of a shell-boring muricid gastropod. In: WINN, H. E. 8 OLLA, В. L., Eds., Behavior of Marine Animals: Current Perspectives in Research, vol. 1, Invertebrates, Plenum, New York, p. 157-244. CARRIKER, М. R., VAN ZANDT, D. & CHARL- TON, G., 1967, Gastropod Urosalpinx: pH of accessory boring organ while boring. Science, 158: 920-922. CARRIKER, М. R. VAN ZANDT, D. & GRANT, T. J., 1972, Mechanism of shell penetration by the boring muricid gastropod Urosalpinx cinerea. т: CARRIKER, М. R., Ed., A Decade of Whole Organism Biology. Systematics-Ecology Program, Marine Biologi- cal Laboratory, Woods Hole, Massachusetts, p. 16-18. CARRIKER, M. R., WILLIAMS, L. G. & VAN ZANDT, D., 1978, Preliminary characteri- zation of the secretion of the accessory boring organ of the shell-penetrating muricid gastropod Urosalpinx cinerea. Malacologia, 17: 125-142. CARTER, М. J., 1972, Carbonic anhydrase: iso- enzymes, properties, distribution, and func- tional significance. Biological Reviews, 47: 465-513. 3 CHÉTAIL, M. €: BINOT, D., 1967, Mise en evi- dence et róle de l'anhydrase carbonique dans Гогдапе accessoire de perforation de Purpura lapillus L., Gasteropode Prosobranche. Comptes Rendus de l’Academie des Sciences, Paris, 264: 946-948. CHETAIL, M., BINOT, D. & BENSALEM, M. 1968, Organe de perforation de Purpura lapillus (L.) (Muricidae): histochimie et histo- enzymologie. Cahiers de Biologie Marine, 9: 13-22. CHETAIL, M. & FOURNIE, J., 1969, Shell-bor- ing mechanism of the gastropod, Purpura (Thais) lapillus: a physiological demonstration of the role of carbonic anhydrase in the dis- solution of CaCO3. American Zoologist, 9: 983-990. | CHETAIL, М. & FOURNIE, J., 1970, Mecanisme de perforation chez Thais lapillus L. (Gas- teropode Prosobranche, Muricide): mise en evidence d’une entree d’ions calcium durant l’activite de l'organe de perforation. Comptes Rendus de I’Academie des Sciences, Paris, 2 11831212 COTLOVE, E. & HOGBEN, C. A. M., 1962, Chloride. In: COMAR, C. L. & BRONNER, F., Eds., Mineral Metabolism, vol. 11, The Ele- ments, Part B. Academic Press, New York, p. 109-173. CRANG, В. E., HOLSEN, В. С. & HITT, J. B., 1968, Calcite production in mitochondria of earthworm calciferous glands. BioScience, 18: 289-301. CRENSHAW, М. A. & NEFF, J. M., 1969, Decal- cification at the mantle-shell interface т molluscs. American Zoologist, 9: 881-885. DAY, J. A., 1969, Feeding of the cymatiid gastropod, Argobuccinum argus, in relation to the structure of the proboscis and secretions of the proboscis gland. American Zoologist, 9: 909-916. CHEMICAL SHELL-BORING ВУ GASTROPODS 155 DE DUVE, С. & WATTIAUX, R., 1966, Func- tions of lysosomes. Annual Review of Physi- ology, 28: 435-492. DERER, M., 1975, L’organe de perforation de Thais lapillus L. (Gasteropodes, Prosobranche). Etude en microscopie optique et electronique. Archives d’Anatomie Microscopique et de Morphologie Experimentale, 64: 1-26. EHRENSPECK, G., SCHRAER, H. & SCHRAER, R., 1971, Calcium transfer across isolated avian shell gland. American Journal of Physi- ology, 220: 967-972. FRETTER, V., 1946, The pedal sucker and anal gland of some British Stenoglossa. Proceedings of the Malacological Society of London, 27: 126-130. GARG, L. C. & OPPELT, W. W., 1970, The effect of ouabain and acetazolamide on trans- port of sodium chloride from plasma to aqueous humor. Journal of Pharmacological and Experimental Therapeutics, 175: 237-247. GARRELS, R. M. & THOMPSON, M. E., 1962, A chemical model for seawater at 25°C and one atmosphere total pressure. American Journal of Science, 260: 57-66. GOLUBIC, S., 1969, Distribution, taxonomy, and boring patterns of marine endolithic algae. American Zoologist, 9: 747-751. GOLUBIC, S., PERKINS, R. D. & LUKAS, K. J., 1975, Boring microorganisms and microbor- ings in carbonate substrates, In: FREY, В. W., Ed., The Study of Trace Fossils. Springer- Verlag, New York, p. 229-259. GREGOIRE, C., 1972, Structure of the mol- luscan shell. In: FLORKIN, M. & SCHEER, B. T., Eds., Chemical Zoology, vol. 7, Mollusca. Academic Press, New York and London, p. 45-102. HANCOX, N. M., 1972, The osteoclast. In: BOURNE, G. H., Ed., The Biochemistry and Physiology of Bone, vol. 1, Structure. Academic Press, New York, p. 45-67. HATCH, W. 1., Jr., 1975, The implication of carbonic anhydrase т the _ physiological mechanism of penetration of carbonate sub- Strata by the marine burrowing sponge Cliona celata. Ph.D. thesis, Boston Univ. Graduate School, Biology. 159 p. HOCHACHKA, P. W. & SOMERO, G. N., 1973, Strategies of Biochemical Adaptation. Saunders, Philadelphia. 358 p. ISTIN, М. & KIRSCHNER, L. B., 1968, On the origin of the bioelectrical potential generated by the freshwater clam mantle. Journa/ of General Physiology, 51: 478-495. ITO, S. & WINCHESTER, R. J., 1963, The fine structure of the gastric mucosa in the bat. Journal of Cell Biology, 16: 541-577. JACCARINI, V., BANNISTER, W. Н. & MICALLEF, H., 1968, The pallial glands and rock boring in Lithophaga lithophaga (Lamel- libranchia, Mytilidae). Journal of Zoology, 154: 397-401. JENKINS, С. N. & DAWES, C., 1963, The possible role of chelation in decalcification of biological systems. In: SOGNNAES, R. F., Ed., Mechanisms of Hard Tissue Destruction. American Association for the Advancement of Science Publication 75, p. 637-662. JONES, M. L., 1969, Boring of shell by Caoban- gia in freshwater snails of Southeast Asia. American Zoologist, 9, 829-835. KOEFOED-JOHNSEN, V. & USSING, H. H. 1960, lon transport. In: COMAR, C.L. & BRON- NER, F., Eds., Mineral Metabolism, an Ad- vanced Treatise, vol. 1, Principles, Processes, and Systems, Part A. Academic Press, New York, p. 169-203. KRETSINGER, R. H., 1976, Calcium binding proteins. Annual Review of Biochemistry, 45: 239-283. MAETZ, J., 1969, Seawater teleosts: evidence for a sodium-potassium exchange in the branchial sodium-excreting pump. Science, 166: 613-615. MAREN, T. H., 1967, Carbonic anhydrase chem- istry, physiology, and inhibition. Physiological Review, 47: 595-781. MINKIN, C. & JENNINGS, J. M., 1972, Carbonic anhydrase and bone remodeling: sulfonamide inhibition of bone resorption in organ culture. Science, 176: 1031-1033. MOELLER, T. & HORWITZ, E. P., 1960, Chela- tion. In: COMAR, С. L. & BRONNER, F., Eds., Mineral Metabolism, vol. |, Principles, Processes, and Systems, part A. Academic Press, New York, p. 101-118. NYLEN, М. U., PROVENZA, D. V. & CAR- RIKER, M. R., 1969, Fine structure of the acessory boring organ of the gastropod, Uro- salpinx. American Zoologist, 9: 935-965. PERSON, P. A., SMARSH, A., LIPSON, S. J. & CARRIKER, M. R., 1967, Enzymes of the accessory boring organ of the muricid gastro- pod Urosalpinx cinerea follyensis, |. Aerobic and related oxidative systems. Blological Bulletin, 133: 401-410. PRITCHARD, J. J., 1972, General histology of bone. In: BOURNE, G. H., Ed., The Bio- chemistry and Physiology of Bone, vol. 1, Structure. Academic Press, New York, p. 1-20. PYTKOWICZ, В. M., 1969, Chemical solution of calcium carbonate in seawater. American Zoologist, 9: 673-679. REHM, W. S., 1972, Proton transport. In: HOKIN, L. E., Ed., Metabolic Pathways, vol. VI., Metabolic Transport. Academic Press, New York, p. 187-241. RUBIN, R. P., 1974, Calcium and the Secretory Process. Plenum, New York, 189 p. SCHOFFENIELS, Е. & GILLES, R., 1972, lono- regulation and osmoregulation in Mollusca. In: EFORRIN NE RERSCHEER 12 ЛЕО Chemical Zoology, vol. 7, Mollusca. Academic Press, New York and London, p. 393-420. SCOTT, B. L., 1967, The occurrence of specific cytoplasmic granules in the osteoclast. Journal of Ultrastructural Research, 19: 417-431. SILEN, L., 1956, On shell-burrowing Bryozoa and Phoronis from New Zealand. Transactions of the Royal Society of New Zealand, 84: 93-96. SIMPSON, J. W. & AWAPARA, J., 1966, The pathway of glucose degradation in some inver- tebrates. Comparative Biochemistry and Physi- ology, 18: 537-548. SMARSH, A., CHAUNCEY, H. H., CARRIKER, M. R. & PERSON, P., 1969, Carbonic anhy- drase in the accessory boring organ of the gastropod, Urosalpinx. American Zoologist, 9: 967-982. STOKES, T. M. & AWAPARA, J., 1968, Alanine and succinate as end-products of glucose degradation in the clam Rangia cuneata. Com- 156 CARRIKER AND WILLIAMS parative Biochemistry and Physiology, 25: 883-892. TAYLOR, J. D., KENNEDY, W. J. & HALL, A., 1969, The shell structure and mineralogy of the Bivalvia, Introduction, Nuculacea- Trigoniacea. Bulletin of the British Museum (Natural History), Zoology, Supplement 3, p. 1-125. TORMEY, J. M., 1966, Significance of the histo- chemical demonstration of ATPase in epithelia noted for active transport. Nature, 210: 820-822. TRAVIS, D. F. & GONSALVES, M., 1969, Com- parative ultrastructure and organization of the prismatic region of two bivalves and its possible relation to the chemical mechanism of boring. American Zoologist, 9: 635-661. TURQUIER, Y., 1968, Recherches sur la biologie des Cirripedes Acrothoraciques, 1. L'anhydrase carbonique et le mecanisme de perforation du substrat par Trypetesa nassarioides Turquier. Archives de Zoologie Experimentale & Generale, 109: 113-122. VAES, G., 1968, On the mechanisms of bone resorption. The action of parathyroid hormone on the excretion and synthesis of lysosomal enzymes and the extracellular release of acid by bone cells. Journal of Cell Biology, 39: 676-697. WASSERMAN, R. H. & TAYLOR, A. N., 1969, Some aspects of intestinal absorption of cal- cium with special reference to vitamin-D. In: COMAR, C. L. & BRONNER, F., Eds., Mineral Metabolism, vol. Ill, Calcium Physi- ology. Academic Press, New York, p. 321-403. WILBUR, K. M., 1972, Shell formation in mol- lusks. In: FLORKIN, М. & SCHEER, В. T., Eds., Chemical Zoology, vol. 7, Mollusca. Academic Press, New York and London, p. 103-145. ZOTTOLI, R. A. & CARRIKER, M. R., 1974, External release of protease by stationary burrow-dwelling polychaetes. Journa/ of Marine Research, 32: 331-342. MALACOLOGIA, 1978, 17(1): 157-162 AN ELECTROPHORETICALLY DETECTED SIBLING SPECIES OF “GONIOBASIS FLORIDENSIS” (MESOGASTROPODA: PLEUROCERIDAE) Steven M. Chambers Department of Zoology, University of Florida Gainesville, FL 32611, U.S.A. ABSTRACT Electrophoretic analysis of Goniobasis from the Ichetucknee River, Columbia Co., Florida, U.S.A., revealed that 2 morphologically very similar species are present at this site. The lack of common electromorphs at 8 of 18 loci argues strongly against gene flow between these sympatric species. One species is G. floridensis (Reeve, 1860), known through- out Florida, and the other has affinities to С. athearni Clench & Turner, 1956, a species found 280 km distant in the Florida Panhandle. Even though these species are widely divergent at the structural gene level, some individuals cannot be specifically identified without electrophoretic determination. INTRODUCTION The value of the techniques of gel electrophoresis of proteins for the detec- tion of sibling species was first demon- strated by Manwell 8 Baker (1970) т holothurians. Using gel electrophoresis, they discovered that two indistinct ““forms'” of the sea cucumber Thyonella gemmata are actually different species. Grassle & Grassle (1976) used similar methods and found that the marine polychaete worm Capitella capitata is actually at least 6 dis- tinct sibling species. In each case, discovery of sibling species was the result of the detection of protein bands, or electro- morphs, that had electrophoretic mobilities unique to a given species. This paper is a report on a sibling species differing from Goniobasis floridensis (Reeve, 1860), and detected by starch gel electrophoresis. Goniobasis floridensis is a common snail in the springs and spring-fed rivers and streams of northern Florida and adjacent parts of Georgia and Alabama. During a wider investigation of electrophoretic varia- tion in Goniobasis (Chambers, 1977), a site on the Ichetucknee River was selected as the site of a reference population, individ- uals of which could be run on all gels as a standard for the measurement of electro- morph mobilities of proteins of individuals from various populations. This site was chosen because it was close to Gainesville (where the laboratory work was per- formed), it had a very large population of Goniobasis which appeared to be typical G. floridensis, and preliminary work showed that these snails were nearly monomorphic with respect to isozyme patterns. ‚After using this population as a reference for comparison with several other G. floridensis populations, the wide divergence between the reference and these populations became obvious. Both the reference population and most G. floridensis populations have what shall be referred to as the “standard”” С. floridensis shell sculpture pattern. This pattern is comprised of axial costae and spiral cords that intersect to form nodules. The peripheral cord is usually the largest. The members of the Ichetucknee reference population were noted to differ in having the shell spire more eroded and less sharply sculptured, but this condition was attributed to abrasion against the bare lime- stone substrate of this rapidly flowing river. The shell morphology seemed well within the range of variation observed statewide in G. floridensis. Using the glutamate oxalacetate trans- aminase (Got) locus, which was fixed for an electromorph in the Ichetucknee refer- ence population not found in any of the G. floridensis samples, Goniobasis were sampled from other nearby sites in the Suwannee River system to indicate the extent of the divergent genotype of the Ichetucknee reference population (Cham- bers, 1977). All of these samples showed (157) 158 CHAMBERS the typical G. floridensis electromorph, Got??. Careful study of the morphology of new collections from the Ichetucknee River locality suggested that there might be two species at the site of the reference popula- tion, one of which, true Goniobasis flori- densis, had not been collected earlier. The electrophoretic analysis of samples of these morphologically very similar species is the subject of this investigation. MATERIALS AND METHODS All snails in this study were collected from the Ichetucknee River at the Florida Department of Transportation roadside park on U.S. 27, Ichetucknee Springs State Park, Columbia County, Florida. This exceptionally clear river is fed by Iche- tucknee Spring. Goniobasis snails were collected by hand and by use of a bottom sampling net. They were brought to Gaines- ville on the day of collection in unaerated jars or plastic boxes and placed in a 24- gallon aquarium that had been prepared in advance. The bottom of the aquarium had been spread with 2-3 cm of fine sand and the water was aerated and filtered by an outside filter. Snails to be electrophoretically exam- ined were removed from their holding aquaria the night before electrophoresis and kept in a clean bowl of water so that they could not feed. Only adult snails were used. The following day, each animal was extracted from its shell and placed in an individual centrifuge tube with an equal volume of the Tris-EDTA grinding buffer of Selander et al. (1971). They were then homogenized by sonication with a Branson Model W185 sonicater equipped with a micro tip. The samples were centrifuged at 4300 g for 5 minutes. The supernatant was drawn off and the pellet discarded. Fifty- two individuals of each species were examined. Horizontal starch gel electrophoresis was carried out with the samples applied to gels as Whatman No. 1 chromatography paper tabs soaked in the homogenate supernatant of a single snail and then lightly blotted. Following electrophoresis, gels were sliced and placed individually in stains for 14 different enzymes. These enzymes, with their abbreviations, are acid phosphatase (Acph), aldolase (Aldo), alkaline phos- phatase (Aph), esterase (Est), glutamate oxalacetate transaminase (Got), glyceralde- hyde-3-phosphate dehydrogenase (G3pd), hexanol dehydrogenase (Hexdh), leucine aminopeptidase (Lap), malic enzyme (Me), 6-phosphogluconate dehydrogenase (6Pgd), phosphoglucose isomerase (Pgi), phos- phoglucomutase (Pgm), sorbitol dehydro- genase (Sdh), and tetrazolium oxidase (To). Throughout the remainder of this paper, these enzymes will be referred to by their abbreviations. Staining solutions were pre- pared according to Bush & Huettel (1972) for Acph, Aph, Est, Hexdh, Lap, 6Pgd, and Pgi; according to Ayala et al. (1973) for G3pd, Me, and Pgm; according to Shaw and Prasad (1970) for Sdh and Aldo; and according to Selander et al. (1971) for Got. Five different gel buffer systems were used to resolve these enzymes. Gels using the Poulik (1957) system and 21 g of Con- naught starch, 21 g of Electrostarch, and 20 g of sucrose per 440 ml of gel buffer were used for Est and Aph. The same starch and sucrose mixture was used for the lithium hydroxide buffer system of Selander et al. (1971) for Pgi and Got. The remaining systems used 44 g of electro- starch per 440 ml of gel buffer. The Poulik (1957) system was used for Sdh, Lap, Me, and Pgm. Buffer B of Ayala et al. (1973) with the pH adjusted to 9.1 was used for G3pd, Aldo, Hexdh, and To. The histidine buffer system of Brewer (1970: 90) was used for Aph and 6Pgd with the following modifications: The concentration of the bridge buffer was 0.2 M and the pH of both the bridge and the gel buffers was set at 8.0. Voucher specimens were deposited in the Florida State Museum (FSM) as follows: G. floridensis (Reeve, 1860), Ichetucknee River, Florida. С. sp. (reference sp.), 24885, Ichetuck- nee River, Florida. G. albanyensis Lea, 1864, 24886, Chat- tahoochee River, Florida. С. athearni (Clench & Turner, 1956), 24887, Chipola River, Florida. The last two species listed were studied by Chambers (1977) and used for comparison with the first two listed species. 24884, RESULTS The results of the electrophoretic analy- sis are given in Table 1. Six loci (Acph-1, ELECTROPHORETICALLY DETECTED SIBLING SPECIES 159 TABLE 1. Allele (electromorph) frequencies, pro- portion of loci polymorphic, and average hetero- zygosities for 12 polymorphic loci in 2 species of Goniobasis from the Ichetucknee River. Locus Allele G. floridensis Goniobasis sp. Aph 102 1.00 — 100 = 1.00 Est-2 100 — 1.00 95 1.00 = Est-3 103 .038 .010 100 .962 .990 Got 100 = 1.00 96 .029 — 93 .971 = G3pd 100 .048 .942 95 .952 058 Lap-1 107 .038 = 102 .952 067 101 .010 = 100 _ .933 Lap-2 102 1.00 = 100 — 1.00 Ме 102 1.00 = 100 — 1.00 6Pgd 109 1.00 — 100 — 1.00 Ра! 109 .029 — 100 .971 1.00 Pgm 101 1.00 - 100 - .490 99 = .298 97 — 212 Sdh 102 = .087 100 — .913 98 1.00 = Loci polymorphic* .000 292 Loci polymorphic** .278 222 Average heterozygosity .021 .058 *A locus is considered polymorphic if no allele has a frequency greater than or equal to 0.99. **А locus is considered polymorphic if no allele has a frequency greater than or equal to 0.95. Acph-2, Aldo, Est-1, Hexdh, and To) were fixed for the same electromorph in both species samples and have been excluded from the table. Scoring of gels followed the criteria of Ayala et al. (1973). Gels were scored as autosomal loci with codominant alleles. The most common allele, or electro- morph, in the reference population for each locus was designated with the super- script! °°, with the other alleles at the locus designated by adding to 100 the distance in milimeters a band migrates anodal to the standard, or by subtracting from 100 the number of millimeters a band runs cathodal to the standard. Loci were designated for polymorphic loci when the genotype frequencies were found to approximate Hardy-Weinberg equilibrium. Genetic interpretations were further con- firmed for Got, G3pd, and Pgi by the presence of heterodimers in the hetero- zygotes. All bands migrated anodal from the origin. DISCUSSION The presence of 2 genetically distinct species, true G. floridensis and the refer- ence species, in the Ichetucknee River is confirmed by the data presented in Table 1. The lack of common alleles (Fig. 1) in these 2 samples at the Aph, Est-2, Got, Lap-2, Me, 6Pgd, Pgm, and Sdh loci is strong evidence that there is no gene flow between these sympatric species. No visible taxonomic shell character will consistently separate the 2 species (Fig. 2). The most useful characters of the reference species include the presence of freshwater sponges, the worn shell spire, and habitat preference for rock rather than vegetation. The overall aspect of the shell of the reference species is shorter and costae are not as strongly developed in most individuals, and the outer lip of the aperture is straighter than the more sigmoidly shaped outer lip of the С. floridensis aperture. Even so, some individuals, especially if they are abraded, are impossible to identify as one or the other species without electrophoretic deter- mination. For this reason, the two Iche- tucknee River species of Goniobasis can be considered sibling species. The degree of similarity between sets of isozyme data from different populations was estimated using the genetic identity (1) method of Nei (1972), which is the norma- lized probability of identity of alleles. The two Ichetucknee River samples of Gonio- basis have an | value of 0.468 (calculated for congruence of alleles over 18 loci), which is in the range of values found when comparing different non-sibling species of Drosophila (Ayala et al., 1974). The refer- ence species shows high identities with Goniobasis athearni\(0.911) and С. a/bany- ensis (0.862) (Chambers, 1977), 2 species found in the Florida Panhandle. These values are well within the range of those found between conspecific populations of Drosophila. The reference population 1$ most similar in shell morphology to G. athearni, which it resembles in overall form, differing only in more-developed costae and a less-developed peripheral cord (Fig. 3). The divergence between С. 160 СНАМВЕН$ 18 Маг 77 FIG. 1. Gel slice stained for Got. The origin (0), anodal (+), and cathodal (-) ends of the gel are indicated. From left to right patterns 1-11, 15, and 21 are members of the reference species (Goniobasis sp.) and indicate the Got’ electromorph. Patterns 12-14, 16-20, and 22-26 are of G. floridensis and indicate the Got?3 electromorph. The individual represented by pattern 23 is heterozygous (Got? 6/93), The scale is numbered in cm. | | || 4 || B FIG. 2. Goniobasis from the Ichetucknee River. floridensis. athearni (and the reference species) and G. floridensis on the genetic level is in con- trast to Clench & Turner’s (1956) belief that these 2 species are so similar that it was surprising to find them sympatric in the Chipola River. Morphological and genetic similarities suggest that the land 8| 1 НЕЙ Pea | | = 7 A-B, reference species (Goniobasis sp.). C-D, G. reference species would be best considered a population of G. athearni until studies of reproductive relationships can be made. The seemingly disjunct distribution of G. athearni in the Chipola River and the reference population 280 km distant in the Ichetucknee River has interesting parallels ELECTROPHORETICALLY DETECTED SIBLING SPECIES 161 | | | 7 8! dada D 6 dada 4 | | | FIG. 3. A-D, reference species (Goniobasis sp.). E-F, С. athearni, Chipola River. G-H, G. albanyensis, Apalachicola River. in other groups. Jackson (1975) has de- scribed the presence of a fossil aquatic turtle of the genus Graptemys in the Santa Fe River. Graptemys barbouri, to which Jackson has referred these fossils, is found today only in the Apalachicola River system, including the Chipola River. He offers the explanation that the Apalachi- cola and Suwannee (including the Iche- tucknee River) river systems are in close proximity where they drain adjacent river basins in southern Georgia. Similar patterns of distribution appear in other aquatic organisms (Jackson, 1975). The very low genetic identity between these sibling species of Goniobasis differs from the finding of Ayala et al. (1974) that sibling species of Drosophila are more closely related to each other than non- sibling species. The identity between the sibling species of Goniobasis is closer to the value for non-sibling species of Drosophila. This may indicate a fundamental difference between these groups in the process of evolutionary divergence. Mayr's (1963) suggestion that sibling species are no more closely related to one another than any other pair of good species may be true for one group, Goniobasis, but not for Drosophila. The sibling species of Gonio- basis may not be sibling species in the same sense as in the study of Ayala, et al. (1974), since all of the species in that study are members of the Drosophila willistoni group. The morphological simi- larity in these 2 species of Goniobasis per- haps is due to convergent evolution in shell sculptural characters. The relatively few characters on the shells of these snails would seem to make convergence more probable than in organisms with more com- plex external morphologies. Difficulties encountered when using shell sculptural characters for taxonomic deter- minations are probably not restricted to Goniobasis. Further use of electrophoretic techniques will very likely reveal more sibling species among this and other gas- tropod groups. ACKNOWLEDGMENTS Permission to collect Goniobasis in Iche- tucknee Springs State Park was granted by Maj. Jim Stevenson, Chief Naturalist, 162 СНАМВЕН$ Florida Division of Recreation and Parks. This work was supported by a grant from the Theodore Roosevelt Memorial Fund of the American Museum of Natural History, a Sigma Xi Grant-in-Aid of Research, and a National Science Foundation Grant for Improving Doctoral Dissertation Research in the Field Sciences. REFERENCES CITED AYALA, F. J., HEDGECOCK, D., ZUMWALT, G. S. & VALENTINE, J. W., 1973, Genetic variation in Tridacna maxima, ап ecological analog of some unsuccessful evolutionary line- ages. Evolution, 27: 117-191. AYALA, F. J., TRACEY, M. L., HEDGECOCK, D. & RICHMOND, R. C., 1974, Genetic differentiation during the speciation process in Drosophila. Evolution, 28: 576-592. BREWER, С. J., 1970, An Introduction to Iso- zyme Techniques. Academic Press, New York. 186 p. BUSH, G. L. & HUETTEL, R. N., 1972, Starch Gel Electrophoresis of Tephritid Proteins. International Biological Programme. 56 p. CHAMBERS, S. M., 1977, Genetic divergence during speciation in freshwater snails of the genus Goniobasis. Ph.D. Dissertation, Univer- sity of Florida. CLENCH, W. J. & TURNER, R. D., 1956, Fresh- water mollusks of Alabama, Georgia, and Florida from the Escambia to the Suwannee River. Bulletin of the Florida State Museum, 1: 97-239. GRASSLE, J. P. & GRASSLE, J. F., 1976, Sibling species in the marine pollution indi- cator Capitella (Polychaeta). Science, 192: 567-569. JACKSON, D. R., 1975, A Pleistocene Graptemys (Reptilia: Testudines) from the Santa Fe River of Florida. Herpetologica, 31: 213-219. MANWELL, С. & ВАКЕВ, С. М. A., 1970, Mole- cular Biology and the Origin of Species. University of Washington Press, Seattle. 394 p. MAYR, E., 1963, Animal Species and Evolution. Belknap Press, Cambridge, Massachusetts. 797 p. МЕ!, M., 1972, Genetic distance between popula- tions. American Naturalist, 106: 283-292. POULIK, M. D., 1957, Starch gel electrophoresis in a discontinuous system of buffers. Nature, 180: 1477-1479. SELANDER, R. K., SMITH, M. H., YANG, S. Y., JOHNSON, W. E. & GENTRY, J. B., 1971, Biochemical polymorphism and systematics in the genus Peromyscus. 1. Variation in the old- field mouse (Peromyscus polionotus). Studies in Genetics VI. University of Texas Publi- cation 7103: 49-90. SHAW, C. R. & PRASAD, R., 1970, Starch gel electrophoresis of enzymes——a compilation of recipes. Biochemical Genetics, 4: 297-320. LA К wit it LME rer atl. .ebtehest: ik арго Aare let Kom titi и № Ни a 200s mono oral tartine At ку OTR. DIA NS wis oh dress role vlats LP abiverd Als 50) AS NAAA Jon У Boney, ГУЛЯЯ dove sol muni Ват ‘vinggene @ olores À ER À 0 29067. altamar br yoolothoolbe a F IATA E Ad beams ty Mis? ét * bi re A wii Ain a 4 he УПА 4 itt miele ÿ | sate ЕН. Ne aS. E 2 bight waive 143 oro. boltiınd sd” bivorl fie wal a ver RRA О nom or в meagan sis р} Oye И Da лов: O51) af only A tl Um. aval Кок OUT AS TN (eat RE Maya "fant ee ld AQ), unbe an. AT vo own e bultimidua dd ode Hot bls Vor.) Wie ERA" solo in < | U j EE vo Sal vd it: uber Wolk HAT oni brides to А af оны ont RAD oc Пана o ‘Gest ии ME Port wit | ar bak 201127 papada). + rif oi Wel ny iG ator mark pier om fim FS nero seal. sel da rladır nha ih sx h ttc Noto 8 la Wi TA kan IT he wol DINO k Pfarr Ruquorg tai van ANG 1DOTE | SI Led | iy Mod) Bm. SINE Jud 606 aii пи в | hora, dl IB TEs iy LH BIT ULI HY 4 Mi Mn. Ног big een ao ai) tothe 29 IR зпонЕННЬ QU Te 1 ot Ole pane fan)” este 20 GS mitad San Dopage a. tu Cou , (lore ! y ge os doit nt aba, Lait 01 oldie Mie 2167 | Bom sip) A Il, uct bi $ | | “lo mul rip HA (rai ot ob Be 4 lue. Hina IS обеим UT ORION ANGL ES arkt wo, ybs bith. bia) ia wu } 15} Oso! 1.078 | io 008 bo) ИН Havre" | Bt idl ara Нас попала IA y oe oe «г: Die Sandler bru). IIA п incl! ЗОО anil io rro ar alero tw axhammue БОВ move. sn on i. Bhs ty Bir ive setts Wd maou” QUE LITE lin mh te Vera) || TU A dé: | > „€ ME natstvndas borra 1e dida ot Ben: An МУН en лез суб Ne | wily Darel) qoitace racy. oben у) ук уно: ni biquoia ad 14, ить » sit men nodos a AL and er tr, borer +} ml or au} Thue prit VAL ns Ho” balsas mn tv 6 ol я. ui Y NA у 4 dl р lest Led Pu à oe ВА „lan bee ater Boo” eta T | я | he AA с в и ваз mo evlbute toral: Миона | À JO vitae ay wm yore baby PRO UC una) тис har ИХ Kissen rap we ‘ting г. ро у Вано molidos Bi ites 17(1) INSTRUCTIONS FOR AUTHORS MALACOLOGIA publishes original studies on the Mollusca that are of inter- national interest and are of high scholarly standards. Both descriptive and experi- mental research results are acceptable pro- vided they are primarily or exclusively concerned with the phylum. Contributions include long monographs as well as moder- ately short research papers. Brief papers are not acceptable. MALACOLOGIA provides a forum for such different aspects of malacology as anatomy, comparative physiology, ecology, medical malacology, paleontology and systematics. Papers of only biochemical or physiological interest should be submitted elsewhere. Review articles are more appropriately submitted to Malacological Review (P.O. Box 801, Whitmore Lake, Michigan 48189, U.S.A.). All manuscripts submitted are reviewed by at least 2 malacologists. Articles are accepted with the firm understanding that they have not been submitted or published elsewhere in whole or in part. Manuscripts may be in English, French, German or Spanish, and should follow MALACOLOGIA style. They must contain a concise but informative Abstract sum- marizing not only the content but the results. Papers in languages other than English should include a translation of the Abstract into English. Authors desiring their abstracts translated into other languages must provide these. Care should be taken to include all necessary foreign accents. Manuscripts must be typed on one side of good quality white paper, double- spaced throughout, with ample margins, and are to be submitted in triplicate. Illustrations are likewise to be in triplicate (the 2 copies may be photocopies, etc.). Tables, figure captions and all footnotes are to be grouped (in this order) at the end of a manuscript, and all Ms pages (including the Abstract) are to be numbered se- quentially. Avoid internal page references (which have to be added in page proof). Make the hierarchy of headings within the text simple and consistent. Suggest an ab- breviated running title to be used at the top of each right hand page. Contributors in English are asked to use the Council of Biology Editors (CBE) Style Manual (Ed. 3, 1972), obtainable for $6.00 from the American Institute of Biological MALACOLOGIA 1978 Sciences, 1401 Wilson Boulevard, Arling- ton, Virginia 22209, U.S.A. MALA- COLOGIA follows most of the recommend- ations in this Manual. In particular, simplified practices such as the following are used: numbers above ten should not be written out except at the beginning of a sentence; percentages following a number are expressed as %, and abbreviations of measures (after a number): mm, ml, kg, etc. have no period (full stop), nor an “s” in the plural. Note that the international symbol for micron is now um, not u. Illustrations must be carefully prepared and so planned that they can be printed in 1 column or the full width of a page of the journal. The maximum size of a printed figure is 13.5X 20.0 cm (preferably not as high as this so that the caption does not have to be on the opposite page). Drawings and lettering must be in dark black on white, blue tracing, or blue-lined paper. Lines and dots should be thick enough to allow reduction by 1/2 or 1/3. This should be taken into consideration also in relation to the lettering. Letters and numbers must not be less than 2 mm in height, preferably larger, after reduction. Several drawings or photographs may be grouped together to fit a page, but drawings are not to be grouped with photographs. Photographs are to be glossy and high contrast. All illustrations are to Бе numbered sequentially as figures (not grouped as plates), and are to be arranged as closely as possible to the order in which they are first cited in the text. (Each figure must be cited in the text.) All original illustrations should be mounted, numbered, labeled or lettered and ready for the engraver. Scale lines are required for all figures and should be convenient lengths (e.g., “200 um”, not 163 um””). Magnifications in captions are not acceptable, and neither are photo- graphic reductions of line drawings. Captions should summarize what is shown in an illustration, and should not duplicate additional information given in the text. Each lettered abbreviation labeling an individual feature in a figure must either be explained in each caption (listed alpha- betically), or be grouped in one alphabetic sequence in a section near the beginning of the text (use the latter method if many abbreviations are repeated on different figures). Tables are to be used sparingly, and should be planned to fit 1 or 2 columns оп 1 page. Each table must be submitted double-spaced throughout on а separate manuscript page. Do not use vertical lines. All References cited in the text must be listed (bibliographies including uncited items are unacceptable). Each reference should be cited accurately (the Editors will spot check for accuracy) and should be in the style used in recent issues of MALA- COLOGIA—except that beginning with Vol. 16 journal titles will be cited complete and unabbreviated. For all manuscripts sub- mitted henceforth, disregard the abbrevi- ations in MALACOLOGIA, 1972, 11(2): 415-426. The journal uses the ampersand (&) for “and”; “et al.’’ may be used in the text, but not in the References. In addition to the volume number, complete page numbers of articles and books must be cited. If plates or maps, etc., are not included in the pagination they too must be cited. For books, the publisher and city are required. In systematic papers, synonymies should not give complete cita- tions but should relate by author, date and page to the References. Voucher specimens. In _ systematic papers, all new type-specimens must be deposited in museums where they may be consulted by other scientists. Beginning with Vol. 16 and when appropriate, MALACOLOGIA will also require that voucher specimens from other kinds of research be deposited in museums. Reprints. When they order 50 or more re- prints, authors will receive 25 additional re- prints gratis; additional copies may be ordered at the time proof is returned to the Editorial Office. Later orders cannot be considered. PAGE COSTS MALACOLOGIA requests authors with grant support to help pay publication costs. MALACOLOGIA requires subsidization for extra long papers. SUBSCRIPTION COSTS For Vol. 17, personal subscriptions are U.S. $12.00 and institutional subscriptions are U.S. $20.00. For information on Vol. 18, address inquiries to the Subscription Office. Vol. 17, No. 1 MALACOLOGIA CONTENTS A. A. SHILEYKO On the systematics of Trichia $. lat. (Pulmonata: Helicoidea: Hygromidae) a ea eee А nn = м va С. J. VERMEIJ and J. A. VEIL | | | A latitudinal pattern in bivalve shell gaping ............ SCANS N к o F. PERRON | ar The habitat and feeding behavior of the wentletrap | és Epitonium greenlandicum: ai ил et RES RE \. А. VAIL - Seasonal reproductive patterns in 3 viviparid gastropods . ....... Re psi A. H. SCHELTEMA Position of the class Aplacophora in the phylum Mollusca ...... DOUÉ С. $. RICHARDS и Genetic studies оп Biomphalaria straminea: occurrence of a ce À fourth allele of a gene determining pigmentation variations . .......... Y M. MOUEZA et L. FRENKIEL Le systeme circulatoire et le jeu des siphons chez Donax — ho ae trunculus, Mollusque Lamellibranche а a e a ie M. R. CARRIKER, L. G. WILLIAMS and D. VAN ZANDT Preliminary characterization of the secretion of the accessory boring organ of the shell-penetrating muricid gastropod Orosalpinx стегеа : à. 5.20% A pe SN M. R. CARRIKER and L. G. WILLIAMS The chemical mechanism of shell dissolution by predatory ‘ae boring gastropods: a review and an hypothesis ................... ME S. M. CHAMBERS An electrophoretically detected sibling species of “Gonio- ' basis floridensis”” (Mesogastropods: Pleuroceridae) .................1 me - M230, 2. VOL.17 NO. 2 1978 MALACOLOGIA International Journal of Malacology AMERICAN MALACOLOGICAL UNION - SYSTEMATICS ASSOCIATION SYMPOSIUM PROCEEDINGS Evolution and Adaptive Radiation of Mollusca 12-13 July 1977, Naples, Florida MALACOLOGIA Editors-in-Chief: GEORGE M. DAVIS ROBERT ROBERTSON Editorial and Subscription Offices: Department of Malacology The Academy of Natural Sciences of Philadelphia Nineteenth Street and the Parkway Philadelphia, Pennsylvania 19103, U.S.A. Associate Editors: Editorial Assistants: JOHN В. BURCH JUDITH DIAMONDSTONE University of Michigan, Ann Arbor LYNN HARTLEY ANNE GISMAN. SUSAN MILIUS Maadi, A. R. Egypt MALACOLOGIA is published by the INSTITUTE OF MALACOLOGY (2415 South Circle Drive, Ann Arbor, Michigan 48103, U.S.A.), the Sponsor Members of which (also serving as ed- itors) are: J. FRANCES ALLEN, Emeritus ROBERT ROBERTSON Environmental Protection Agency Washington, О.С, CLYDE Е. Е; ROPER, President-Elect CHRISTOPHER J. BAYNE, Vice-President Smithsonian Institution Oregon State University, Corvallis Washington, D.C. ELMER G. BERRY, Emeritus W. D. RUSSELL-HUNTER Germantown, Maryland Syracuse University, New York KENNETH J. BOSS NORMAN F. SOHL Museum of Comparative Zoölogy United States Geological Survey Cambridge, Massachusetts Washington, D, С. O A АННЫ RUTH D. TURNER, Alternate MELBOURNE R. CARRIKER Museum of Comparative Zoology University of Delaware, Lewes Cambridge, Massachusetts GEORGE М. DAVIS, Executive SHI-KUEI WU Secretary-Treasurer University of Colorado Museum, Boulder Institute meetings are heid the first Friday in December each year at a convenient place. One subscriber may attend and vote by petitioning in advance. For information, address the President. Copyright, © institute of Malacology, 1978 J. А. ALLEN Marine Biological Station, Millport, United Kingdom E. E. BINDER Muséum d'Histoire Naturelle Genève, Switzerland A. H. CLARKE, Jr. National Museum of Natural History Washington, D.C., U.S.A. Е. S. DEMIAN Ain Shams University Cairo, А. В. Egypt C. J. DUNCAN University of Liverpool United Kingdom Z. А. FILATOVA Institute of Oceanology Moscow, U.S.S.R. Е. FISCHER-PIETTE Museum National d Histoire Naturelle Paris, France A. FRANC L “Université Paris, France \. ЕВЕТТЕВ University of Reading United Kingdom E. GITTENBERGER Rijksmuseum van Natuurlijke Historie Leiden, Netherlands A. N. GOLIKOV Zoological Institute Leningrad, U.S.S.R. A. V. GROSSU Universitatea Bucuresti Romania T. HABE National Science Museum Tokyo, Japan A. D. HARRISON University of Waterloo Ontario, Canada K. HATAI Tohoku University Sendai, Japan MUS. COMP. ZOOL. LIBRARY JUL 27 1978 HARVARD UNIVERSITY EDITORIAL BOARD B. HUBENDICK Naturhistoriska Museet Göteborg, Sweden A. M. KEEN Stanford University California, U.S.A. R. N. KILBURN Natal Museum Pietermaritzburg, South Africa M. A. KLAPPENBACH Museo Nacional de Historia Natural Montevideo, Uruguay J. KNUDSEN Zoologisk Institut & Museum Kdbenhavn, Denmark A. J. KOHN University of Washington Seattle, U.S.A. Y. KONDO Bernice P. Bishop Museum Honolulu, Hawaii, U.S.A. Go Me EAE! McGill University Montreal, Canada JF EEVER Amsterdam, Netherlands CTEO National Taiwan University Taipei A. LUCAS Faculte des Sciences Brest, France N. MACAROVICI Universitatea “Al. I. Сига” lasi, Romania C. MEIER-BROOK Tropenmedizinisches Institut Tübingen, Germany (Federal Republic) H. K. MIENIS Hebrew University of Jerusalem Israel J. Е. MORTON The University Auckland, New Zealand В. NATARAJAN Marine Biological Station Porto Novo, India J. OKLAND University of Oslo Norway T. OKUTANI Tokai Regional Fisheries Research Labora- tory Tokyo, Japan W. L. PARAENSE Universidade de Brasilia Brazil J. J. PARODIZ Carnegie Museum Pittsburgh, U.S.A. С. М. PATTERSON University of Michigan Ann Arbor, U.S.A. W. F. PONDER Australian Museum Sydney A. W. B. POWELL Auckland Institute & Museum New Zealand R. D. PURCHON Chelsea College of Science & Technology London, United Kingdom С. P. RAVEN Rijksuniversiteit Utrecht, Netherlands О. RAVERA Euratom Ispra, Italy N. W. RUNHAM University College of North Wales Bangor, United Kingdom SG SEGERSTRALE Institute of Marine Research Helsinski, Finland G. A. SOLEM Field Museum of Natural History Chicago, U.S.A. F. STARMÜHLNER Zoologisches Institut der Universitat Wien, Austria W. STREIFF Universite de Caen France J. STUARDO Universidad de Chile, Valparaiso T. E. THOMPSON University of Bristol United Kingdom Е- ТОБЕОСЕЕТТО Societa Malacologica Italiana Milano W. $. $. VAN BENTHEM JUTTING Domburg, Netherlands J. А. VAN EEDEN Potchefstroom University South Africa J.-J. VAN MOL Universite Libre de Bruxelles Belgium B. R. WILSON Western Australian Museum Perth C. M. YONGE Edinburgh, United Kingdom EIZZEISSEER Leipzig, Germany (Democratic Republic) ARAUCA Natur-Museum und Forschungs- Institut Senckenberg Frankfurt-am-Main, Germany (Federal Republic) AMERICAN MALACOLOGICAL UNION - SYSTEMATICS ASSOCIATION SYMPOSIUM PROCEEDINGS Evolution and Adaptive Radiation of Mollusca 12-13 July 1977, Naples, Florida ТОКУ AN Take - Regina! Fittenes Remparct 1. rory Foly, Jaan W.L.FARAENSE Uintvertidace de Braslls Вой Jt. PAMNODIZ Creme Must rr Pretotus dl 5 С М, РА а ERSON узлу. Of ACT Ge Ann yee In 9 MOLTAIIO22A SIITAMIT2YE - VOIMU JADAOIO 29033909 тЫ ОВ sotculloN; to noitsibsA sviigsbArbns. molar, д BAT} Hint Ах Hin Зин AW. EE, POY | Mean? ПА Er (M ry. WE Я PUHRCHON ve ge ( Feng ol 1 rik toi tai r rt WE Че Ты | ‘ Y 5 TE far ii ‚sbhol? 281951 MOI vever ap | У are (Акиты se Cor France J. ErUARDO Universidad de Chie. + — | Waliparaao 1: E* THOMPSON University ol. Bret (ivitod ОТ я, FOFFOLETTO = Soclote Malszólogias. Hal Mano WE 5. VAN BENT Waurol y J у hol MO = УМЕ ¡bre de Bra “et м т р zur lJ | iW IN à a } > o ч ГО ИГТ Aira ЕО Peri - AJO ONGE Sodi ch . <>. 4 sn u irren ls tal Sagan = Aue т а =. 2 ve Demacsam | use Lhd Porto | и, er awit, Garren tedub! ie} MALACOLOGIA, 1978, 17(2): 163-164 INTRODUCTION George M. Davis President of the American Malacological Union, 1976-1977; Symposium moderator; Academy of Natural Sciences, Nineteenth and the Parkway, Philadelphia, Pa. 19103, U.S.A. Nearly two years before 12 July 1977, Professor A. J. Cain, then president of the Systematics Association (S.A.), and | dis- cussed the possibility of having a joint A.M.U.-S.A. Symposium concerned with molluscan evolution and adaptive radia- tions. Such a symposium would, for the first time, bring S.A. members together with A.M.U. members in a joint session. We organized the symposium to be held during the 43rd annual meeting of the A.M.U. and thus initiated the first formal international meeting of the A.M.U. In preparing this Symposium we were interested in current work that has a signif- icant impact on elucidating processes per- taining to evolution, adaptive radiations, and deployment. Accordingly we invited scholars studying aspects of molluscan biol- ogy as diverse as shell morphometrics, pop- ulation ecology, molecular genetics, paleon- tology, biochemistry, sexuality and repro- duction, and functional morphology. We wanted maximum interaction between speakers and audience keeping in mind that many A.M.U. members are amateur conch- ologists. We hoped that we could present for professional biologists and amateurs alike a broad view of some of the exciting work that is now being done in molluscan studies that transcends the levels of alpha taxonomy and faunistic compilations. To obtain a strong interaction between the speakers and the audience, the symposi- um was structured as follows: the 9 speak- ers served as a panel and were present for all papers. Each panelist was a discussant for one of the papers. Each speaker was given 40 to 45 minutes for the presenta- tion. A presentation was followed by com- ments and questions by the discussant. Then the paper was open to comments and questions from the panel and subsequently from the audience. The total period of discussion was 15 minutes and sometimes longer. We found this method worked ex- tremely well for stimulating and obtaining audience participation. As the total time allotted to each paper was 60 to 70 min- utes, a speaker had time to fully develop a topic. Each level of discussion and ques- tions opened up new avenues for discussion on the part of the audience. The order of the papers as presented in the symposium is given below along with the name of the discussant for each. In looking at this list and by reading the Proceedings, it is readily apparent that the majority of papers dealt with land snails (five of nine). | feel that this actually reflects the fact that most of the exciting and encompassing work in molluscan evolu- tion over the past 15 years has involved land snails. Three papers are concerned in one way or another with the adaptive significance of shell shape; two of these are with land snail faunas. Three papers are concerned with reproductive modes and the significance of these modes as regards evo- lution and deployment. The importance of molluscan genetics in assessing changes through time or degree of similarity within or between populations and species is stressed in three papers. The first two papers were presented by paleontologists as there can be no meaningful discussion of evolution without an integration of the fossil record within a framework of geo- logical change. These initial papers under- score the problems of working only with shells in attempting to assess relationships, and the fact that many paleontologists have, over the past decade, turned to studying living mollusks in order to inter- pret shell morphologies of living and fossil taxa. (163) 164 DAVIS The papers presented were: 1. Yochelson, E. L. Contrasting views in present-day literature of phylogeny of the early mollusks. Discussant: Robert M. Linsley (A.M.U.) 2. Linsley, R. M. Shell form and rates of locomotion in gastropods. Discussant: Ellis L. Yochelson (A.M.U.) 3. Cain, A. J. Why is a snail that high and that wide? Discussant: James J. Murray (A.M.U.) 4. Woodruff, David S. Evolution and adap- tive radiation of Cerion: a remarkably diverse group of West Indian land snails. Discussant: G. S. Oxford (S.A.) 5. Murray, J. and Clarke, B. (S.A.). Changes of gene frequency in Cepaea nemoralis: a fifty-year record. Discus- sant: A. J. Cain (S.A.) 6. Oxford, G. S. The nature and distribu- tion of food-induced esterases in helicid snails. Discussant: D. 5. Woodruff (A.M.U.) 7. Runham, N. W. Reproduction and its control in the slug Agriolimax (Dero- ceras) reticulatus. Discussant: P. Calow (S.A.) 8. Calow, P. The evolution of life-cycle strategies in fresh-water gastropods. Dis- cussant: K. E. Hoagland (A.M.U.) 9. Hoagland, K. E. The evolution of pro- tandry and labile sex determination in the Mollusca. Discussant: N. Runham (S.A.) We have added a tenth paper to the symposium proceedings. This addition by Heatwole & Heatwole on the camaenid land snails of Puerto Rico was made after the paper had been accepted for publica- tion in Malacologia. It very well comple- ments the papers by Woodruff and Cain in that it deals in depth with ecological and biological factors helping to explain the deployment of this Puerto Rican snail group. Such ecological work with land snails is very much needed, a point with which Professor Cain surely would agree, especially as such studies are essential to explain why there are non-random patterns in the height and breadth of shells of land snails not correlated with taxonomic af- finity. The lessons from these contributions are clear. Studies on the evolution and develop- ment of mollusks must be based on a firm foundation of systematics and nomen- clature. However, sound systematics can no longer be based on an analysis of shells alone, or on the simplistic analysis of a few conveniently studied morphological parts such as penis, radula, operculum. This was exemplified in Dr. Woodruff’s paper in which it was argued that 200 nominal species of the land snail genus Cerion, with remarkable shell morphological variation, reduced to possibly two species when pop- ulations were studied with reference to zoogeography, ecology, multivariate mor- phometrics, and molecular genetics. Ad- vances in and knowledge of systematics and evolution of molluscan groups will only come from broadly based studies con- ducted within a framework of population ecology, population genetics, and modern ecological theory. To these essential ele- ments one should add one or more areas of study such as functional morphology, re- fined quantitative analyses such as multi- variate analysis, molecular genetics, and zoogeography in time and space. MALACOLOGIA, 1978, 17(2): 165-191 AN ALTERNATIVE APPROACH TO THE INTERPRETATION OF THE PHYLOGENY OF ANCIENT MOLLUSKS Ellis L. Yochelson United States Geological Survey, Washington, D.C. 20560, U.S.A. ABSTRACT Some extinct Paleozoic mollusks have shell shapes which are judged to be distinct enough from those of living mollusks to warrant placement in extinct class level taxa. Recent publications placing several proposed extinct classes in the synonymy of living classes are judged to be based on poor definitions of class limits and excess speculation concerning unpreserved soft parts. More rigorous morphologic definition of the phylum and of the various classes based on the preserved hard parts might reduce some of the differences in interpretation among paleontologists. No clear arrangement of molluscan classes into subphyla is recognized. The origin of classes may constitute new major morphologic changes which are then exploited rapidly; adaptive radiation could be a pervasive process occurring at all taxonomic levels, including the class level. Much hypothesis and speculation is involved in any attempt to construct a higher classification of extinct forms, and a phylogenetic interpretation must be modified as new facts are obtained. There is no final and authoritarian paleontological viewpoint. INTRODUCTION Discussion of phylogeny can be enter- taining. Such writings have little impact on biostratigraphy, paleoecology, and any other facets of paleontology, but they can provide some generalizations which help to interpret the evolutionary history of a group of organisms. A discussion of phylog- eny includes both facts, subject to change as new fossils are found and old ones redescribed, and opinions, which are sub- ject to even greater modification. Any dis- cussion is useful in pointing out areas of uncertainty and in provoking comment which might lead to greater clarification. It is basic that no discussion of phylogeny be treated as the ultimate word. My sole earlier attempt to interpret evo- lution of the phylum Mollusca is a one page paper (Yochelson, 1963) in which | suggested a two-step evolution within the Mollusca. The first step in the early Paleo- zoic was a time of testing and experimenta- tion by higher categories and ultimately their extinction. The later development of the extant classes of mollusks constituted the second step. My current understanding is that this general pattern is plausible but the notion of steps is far too simplistic. | think there were a number of classes of mollusks which appeared at different times and that some of them are now extinct. Some neontologists such as Stasek (1972) and Salvini-Plawen (1968, 1972) attempt to interpret molluscan phylogeny with only limited reference to fossils; some paleontologists such as Runnegar & Pojeta (1974), Pojeta € Runnegar (1976) and Starobogatov (1970, 1974) draw extensive- ly on the fossil record and arrive at differ- ent conclusions. One can argue that an additional interpretation may be equally valid and so long as speculation is not treated as serious science, no great harm is done. It is human nature to want to clas- sify (Batten, 1960). My concept of the pattern to be seen in the phylogeny of moilusks is shown in Fig. 1. The question of when a group began is always subject to modification by new in- terpretation and new discovery. However, | think that the fossil record is fairly good for groups of high systematic rank and the base of range lines is unlikely to change much in the near future, at least. | presume that the first appearance of a fossil is about at the time a group appeared, though some persons like to speculate that ancestral forms appeared far earlier than their first records. | have not illustrated any of the genera (165) YOCHELSON Bent Flexed Nearly Straight Flexed Flexed & Torted Bivalve Univalve w 8 BIVALVIA SCAPHOPODA 425 O ROSTROCONCHIA NON-SHELLED ANCESTOR 575 STENOTHECOIDA ~ | Millions of Years © MATTHEVA = a 2 or 8 Part Shell Univalve Shell Closed at Apex pen 1 1 5 | | | | < т Е — © > N x ` \ SA [> RS OD ? x 1 La x = e > E 1 Ss x = ; S 1 = 1 | FIG. 1. Ranges of molluscan classes through the Paleozoic. А scale of geologic periods and radioactive dates is given to the left; interpretations of the gut appear at the top, just above a statement on geometry of hard parts. Class names are given in full capitals and their ranges end in the Paleozoic as indicated by a bar or continue to the Recent as indicated by an arrow; groups of probable class rank are indicated by superfamily name and those of possible class rank are indicated by generic name. discussed herein. It is better to refer to original photographs and original recon- structions than to interpret them through additional diagrammatic illustrations. My writing style is a bit “‘folksy’’ but whereas informality in a scientific review may mystify a reader, a generalized sketch of an organism may mislead. One aspect of my scheme is an ancestral generalized archimollusk which persisted from latest Precambrian through the early part of the Paleozoic; it is not a necessity as major change in larvae is an alternative. For the particular feature of this mythical form | choose to follow Fretter & Graham (1962: 1-6). My view is that as ecologic opportunity or genetic modification pre- sented opportunities, the archimollusk threw off classes as a St. Catherine’s wheel throws off sparks. | consider adaptive radia- tion to be as important in the origin of classes (Hotton, 1971) as it is at other taxonomic levels and | think this type of evolution happened rapidly. However, by mid-Paleozoic opportunities for major adap- tive radiation disappeared and competition from other more efficient animals was such that: “The Archi-mollusk sought a cleft his shame and grief to hide, crunched horribly his horny teeth, gave up the ghost, and died.” (Garstang, 1951: 41) THE TREND IN HIGH-LEVEL SYSTEMATICS Linnaeus, the father of all systematics, had three kingdoms: Animal, Vegetable, and Mineral. A decade ago, Whittaker had five, not counting the minerals, and PHYLOGENY OF ANCIENT MOLLUSKS 167 Russian paleontologists have proposed an- other. The term “taxonomic inflation‘ was coined by Sabrosky, and it aptly describes the change with increased study in level of elements of the Linnaean hierarchy. Genera become families, superfamilies become or- ders, classes become phyla, quite apart from increase in the number of species named. The number of taxa recognized at different levels increases at different rates, but they all increase; the two highest levels of taxa, class and phyla, increase at the slowest rate. The number of workers and the interest they generate in high-level clas- sification determines the rate at which changes occur; today few people search for new species of living birds, but many look for additional evidence of the ways families and orders might be related. The living world is not completely known, but the major biological surprises per number of workers is small compared to possible sur- prises remaining in 600 million years of relatively good fossil record. New major categories are probably going to be based on fossils (Yochelson, 197 1a). Consider the mollusks in this light. They were not even fully separated as a phylum by Linnaeus, most members being within the “Vermes mollusca,’’ though he did rec- ognize a group which is now the Bivalvia. Cuvier in 1797 was able to distinguish the classes Gastropoda and Cephalopoda; the Scaphopoda were recognized by Bronn in 1862. Though de Blainville had recognized the Polyplacophora in 1816, von Ihering in 1876 suggested that the Amphineura was a separate phylum. During the next 75 years, there were some attempts to name addi- tional classes and subphyla, particularly by German zoologists, but by and large, text- books of zoology discussed five classes of mollusks until 1957. At that time the concept of the class Monoplacophora Wenz (Knight, 1952) came clearly into focus, and the emphasis on phylogeny shifted to pale- ontology. Before leaving the Recent Mollusca, one may note that two decades after Neopilina was described, six shelled classes were gen- erally accepted. However, there may be seven or eight classes of extant mollusks, depending on whether one wishes to recog- nize the class Aplacophora or divide this group into the classes Solenogastres and Caudofoveata.1 Salvini-Plawen has written extensively on this issue and argues that there are eight classes of living mollusks. This is quite a taxonomic inflation over the five accepted in 1957 and inflation has a way of continuing. WHAT IS A MOLLUSK? The Mollusca, as conceived by neontolo- gists, is a large (Boss, 1971) and diverse phylum distinguished primarily on the basis of anatomical, as opposed to conchological (shell) features. It is impossible to produce an exclusive definition of the phylum with- out reference to the soft parts. Hyman (1967: 5) indicated that ‘‘Mollusks are one of the most definitely characterized groups of the animal kingdom, having at least two features, mantle and radula, not found else- where.’’ However, one of her characteristics is not exclusive, for Brachiopoda possess a mantle; some worms have a feeding struc- ture which shows similarities to a radula, whereas Pelecypoda do not. Neither feature is normally preserved after death of the animal. Likely features of ctenidia are peculiarly molluscan (John Morton, oral communication, 1977), but they, too, are not shown in fossils. Zoologists do not always study the same features as paleontologists and to some extent this is a division between those who look at soft parts and those who cannot. Too often the neontologist and the paleon- tologist talk at each other rather than to each other. A way to phrase the question could be: “how may a fossil be assigned, with a high degree of confidence, to the phylum Mollusca?” | suggested (Yochelson, 1961) a number of features which | think are common to many mollusks. These are: shell of calcium carbonate in several layers, prominent growth lines, logarithmic growth, inner structure prismatic, never fi- brous and not pierced by canals. Addition- ally, the shell commonly: lacks an apical foramen or attachment disk; is a univalve, and is bilaterally symmetrical. There may be septa within the shell and there may be an associated operculum. My aim was to invite discussion on the issue of what is a mollusk, but it was a total failure! After nearly two decades, | 1See A. H. Scheltema, 1978. Position of the class Aplacophora in the phylum Mollusca. Ma/acologia, 17: 99- 109. ED. 168 YOCHELSON can only say that | do not know of any features which are exclusive to fossil mol- lusk shells. A plexus of characters must be considered in order to assign a fossil to the Mollusca. The further one goes back into the fossil record, the greater the difference in gross morphology from living descend- ants. Most molluscan shells exfoliate from the steinkern when broken from the rock mat- rix, especially when the matrix is a fine- grained limestone. | suspect this results from diagenetic alteration of incremental shell layers combined with the nacreous laminar structure widespread in members of the various classes. Shells of early forms should have been aragonite which subse- quently reverted to calcite; the partially or fully calcitic shell was a later development within the Mollusca and perhaps associated with sedentary habit, though Pecten, for example, is a conspicuous exception. Some taxa in the Gastropoda and Pelecypoda have a shell in which the outer layer is calcite and the inner layer is aragonite. This combination of mineralogy seems unique to the phylum, but unfortunately it is not widespread enough in the Mollusca to help characterize them. Thus, those Cepha- lopoda which have a shell appear to utilize only aragonite. No shells generally regarded as mollusks contain appreciable amounts of calcium phosphate, and in my view, shells primarily phosphatic are immediately ex- cluded from the mollusks; other criteria have exceptions. The gross morphology of the preserved hard parts alone is not a sufficient criterion to place a fossil in the Mollusca. Tubes, lids, and sclerites, as represented by Mober- gella, styliolinids, tentaculitids, and tom- motiids are not mollusks, even though some mollusk hard parts are similar in shape; they are analogous pieces, rather than homologous. Morphology includes size, and early Paleozoic mollusks in the microfossil size range are uncommon. A philosophical issue is whether in the high level assignment of a fossil one takes a liberal or a restricted view. | tend towards restriction in that a fossil must fully satisfy me, subjectively, that it is a mollusk before | will accept it into the phylum. А nasty point is that of differentiating remains of fossil “‘worms’’ from those of mollusks. There are a limited number of shapes of “worm tubes’ today, but who is to say what the past may have witnessed? To take a tiny steinkern—an internal filling which preserves no evidence regarding the original shell—and say that it is a scaphopod be- cause it is curved or a gastropod because it is coiled seems to me to be poor science. Some fossils should not be assigned to phylum level on the basis of current infor- mation. | judge it far better to let some organisms remain incertae sedis than to assign them to the Mollusca. Keeping strange fossils separate focuses attention on them. Assigning them to the Mollusca pre- sumes more than we know and may lead to spurious data and correspondingly false phylogeny. Too much philosophy is stultifying. It is better to come to grips with organisms. In the following sections, my aim is to review the high-level taxa of mollusks. In the following treatment of the early history of the second largest animal phylum, generali- zations are necessary. It is well known that all generalizations are false, but | plead for mercy because of space limitations. There are some high-level categories for Paleozoic forms which have been proposed, but, have not been generally accepted and have dropped from usage. There seems little point in considering, for example, the Eopteropoda or the Coniconchia (Yochel- son, 1961). CEPHALOPODA The Cephalopoda have long been recog- nized as a major molluscan entity. D'Orbigny judged the Bellerophontacea, an extinct order of Paleozoic bilaterally symmetrical gastropods, to be cephalopods without septa. This was the only important piece of systematic confusion, and it was clarified 125 years ago. Most living cepha- lopods do not have calcareous hard parts, and their ancestors, even though with a phragmocone, show an incomplete fossil record (Donovan, 1977). The chambered Nautilus shell serves as a window to the past, but most living cephalopods demon- strate in a dramatic fashion that the pres- ent may be an exceedingly poor guide to the past. One would hope that even if Nautilus were extinct, hard parts of Sepia and Spirula might be sufficient to allow cham- bered shells of the past to be interpreted as cephalopods. That may be too generous an interpretation of our ability to relate fossils PHYLOGENY OF ANCIENT MOLLUSKS 169 to living organisms. Nautilus has been a mixed blessing because for years paleon- tologists attempted to interpret fossil ceph- alopods in the light of what was thought to be known of this organism. Thus, all cepha- lopods were assumed to be fierce carnivores and we argued fruitlessly whether extinct ones had a single pair or two pairs of gills. A fossil cephalopod is easy to recognize for the septa are pierced by a siphuncle, though fragments such as a node on the shoulder of a Pennsylvanian nautiloid, or a curved body chamber, are more difficult to assign. The septa/siphuncle combination is the “how” of a cephalopod. Much work over the last two decades has been devoted to gas production and its control within living cephalopods. The theory has been applied, apparently with a high degree of success to fossils, and | believe that we may speak of buoyancy as the “why” of cephalopods; Teichert (1967) has inter- preted evolution of the class in terms of buoyancy control. The record of the cephalopods shows a great deal of diversity in the past, more so than today. Possibly the biomass of living cephalopods is far greater than it was at any one time in the past, even when vari- ous extinct lineages are combined. This increase in biomass probably results from most modern cephalopods achieving the acme of molluscan buoyancy by loss of the shell. In much of the Ordovician and Silurian record of cephalopods, straight, chambered shells predominate. However, in the Late Cambrian a curved shell, Plectronoceras, occurs. Yochelson, Flower & Webers (1973) accepted this genus as the earliest cephalopod. They had found a Late Cam- brian high, curved, conical shell which con- tained multiple septa. It was their hypoth- esis that this form, named Knightoconus, was a monoplacophoran and was morpho- logically similar to the organism which gave rise to the Cephalopoda. Development of a siphuncle, a tube to pierce the septa, would have converted Knightoconus into ап or- ganism essentially like Plectronoceras. Runnegar & Pojeta (1974) have accepted this hypothesis, and attempted to extend the ancestral group downward by placing the Early Cambrian Tannuella in the cepha- lopod precursor group. Curved septate uni- valves are known in the Middle Cambrian, which might be related to Knightoconus, but it is premature to include any Early Cambrian genera in that lineage. A methodologic point should be empha- sized. Presumed relationship of Plectrono- ceras to Knightoconus was based on fea- tures of hard part morphology. Generalized soft parts were added to each fossil, but arguments based on soft parts were ancil- lary; paleontologists should deal primarily with facts of hard parts, and only second- arily with assumptions regarding features not preserved. The presumed relationship of Knighto- conus to Plectonoceras is plausible in part because the nature of the earliest cepha- lopod was clarified. Previously, Early Cam- brian straight-shelled Sa/terella and Vol- borthella were called early cephalopods. Although this notion was repeatedly denied during the 1950’s and 1960's, it still lin- gered in the literature. It is a classic case of organisms being assigned to the Mollusca for the flimsiest of morphologic reasons; their presence in the phylum and class confused interpretations. One can never be totally certain in science, and | can only hope that my decision to establish a new phylum based on Sa/terella (Yochelson, 1977) will stand. If nothing else, that ac- tion clearly removes the genus from the Mollusca. Starobogatov (1974) places most elon- gate tapering tubed shells in one class of mollusks, (‘‘Solinoconcha’’), distinguishing those with an opening at both ends (Sca- phopoda) from those which are open only at the aperture. The animal within these simple straight conical shells would as a next step secrete septa perforated by a siphuncle and the resulting animal would be an orthoconic cephalopod; it has been mentioned repeatedly that the earliest known cephalopods are not orthocones. Starobogatov is not clear as to the timing of these events or the selective pressures or other reasons for the dramatic changes. By extending his concepts, Starobogatov also develops a new classification within the Cephalopoda. | cannot accept Starobogatov’s deriva- tion of the Cephalopoda and | am satisfied with the idea presented by Yochelson, Flower, and Webers (1973). There are two general implications to this theory which should be repeated. First, in terms of mor- phology, Monoplacophora are the closest of all living groups to most published concepts of the ancestral mollusks; in contrast, many physiologists and anatomists judge living cephalopods to be the most advanced of all 170 YOCHELSON mollusks. If one can directly derive the most advanced from the most primitive form of mollusk, this suggests that these terms may actually have little meaning, at least in any evolutionary context. Second, it is generally assumed that most limpet- shaped monoplacophorans had the shell apex anterior, basing this assumption, in part, on morphology of living examples and, in part, on muscle scars of a few of the extinct forms. Acceptance of the rela- tionship of septate monoplacophorans as precursors carries with it the notion that some monoplacophorans had the shell apex directed toward the posterior. Living Cephalopoda are highly special- ized in a few traits but little specialized in other molluscan features (Dr. A. Bidder, oral communication, 1977). In a sense, this view supports a relationship of cephalopods to monoplacophorans. SCAPHOPODA My reason for mentioning elephant tusk shells next is that approximately as many persons have seen live Dentalium as have seen live Nautilus. Scaphopods live partially submerged in mud and hunt with their captacula, a complex of tentacle-like or- gans; Foraminifera no doubt regard them as fierce predators. They are carnivores, though part of their diet is organic detritus or bacteria in the sediment. The Scaphopoda and Bivalvia have in common the absence of a head. Whereas the Bivalvia lack a radula, this structure is present in the scaphopods. Accordingly, the pelecypods could not have been ancestors of the scaphopods, because it is unlikely that this complex structure could have been lost and then reacquired. The similar- ities there may be between the Scaphopoda and some Pelecypoda are probably the re- sult of convergence from a semi-infaunal life habit for both groups. Conchologically, the scaphopods are univalved in having a single hard part, a tapering curved shell open at both ends. The issue of what is a mollusk clearly applies to scaphopods. If scaphopods were extinct and soft parts were unknown, would they be called mollusks? | think not. Even if details of shell structure were avail- able, they still might not be considered molluscan, for the structure of calcareous tubes constructed by some worms is some- what similar to that of mollusk shells. The logarithmic curvature of the sca- phopod shell and its general regularity of shape, with little individual variation, is more suggestive of mollusks than of worms, but these are not unique differences. The tube of Pectinaria is open at both ends, curved, and fairly scaphopod-like except that it is constructed of agglutinated parti- cles. This worm resembles the scaphopods. The deep-sea quill worms, like Halcioneia, live in a curved tube, open at both ends, distinguished from the shell of a scaphopod only because it is a non-calcified organic sheath. In the past, did quill worms ever calcify this sheath? Perhaps | overempha- size the issue of distinguishing worm tubes, for the microstructure seen in recent tubes is quite different from that of mollusks. However, paleontologists seldom find mate- rial as well preserved as that just secreted and it is truly difficult to separate these two groups. Even in the Recent, Ditrupa has fooled zoologists who have not seen the soft parts of this worm. The oldest scaphopods that | have seen are Mississippian in age. Haas (1972) re- ported Devonian ones with convincing mol- luscan shell structure, though there is a slight chance that he sampled the body chamber of a curved cephalopod rather than a scaphopod. There are no reports of Silurian scaphopods. Ordovician scaphopods were described by Eichwald more than a century ago from the Ordovician of the Baltic platform, but have not been confirmed by any later investigators. My brief examination in 1962 of several of Eichwald's specimens shows them to be fragments of steinkerns 2-5 cm long. They could be fillings of the body chambers of nautiloids. Ordovician scaphopods were reported twice in recent years. Bretsky and Berming- ham (1970) transferred the Late Ordovi- cian “worm” Coleolus iowense James to the scaphopod genus Plagioglypta. Speci- mens have several thin layers which exfoli- ate singly, quite unlike mollusk shells. Fur- ther, Holocene Dentalium and Pennsyl- vanian Plagioglypta contain in their shell 0.01% to 0.02% Р.О; whereas С. iowense contains this compound in concentrations an order of magnitude larger, about 0.1% Р.О; (Jaresowich & Yochelson, unpub- lished). Runnegar & Pojeta (1974) and Pojeta & Runnegar, (1976: 43) also reported Middle Ordovician scaphopods. These are small PHYLOGENY ОЕ ANCIENT MOLLUSKS 171 tubes, curved, possibly open at both ends, but replaced by silica; because of replace- ment, nothing much can be learned of the original shell composition or structure. The cross-section of the tube is not circular and therefore unlike that of known living sca- phopods. The highly gregarious occurrence of these small tubes, like the gregarious occurrence of Coleolus ¡owense, would аг- gue more for the feeding of worms on decaying material than for scaphopods hunting prey beneath the surface. It is likely that the Paleozoic Sca- phopoda currently contain some forms which are worm tubes (for an example, see Yochelson, 1971b). It would be wiser to assign to the class only those forms for which typical molluscan shell structures can be demonstrated, rather than assume that all slightly curved tubes are scaphopods. One can pick the first known fossil scaphopod as Middle Ordovician, Devonian, or Mississippian. Regardless of which period is chosen, the time of origin thus attributed to scaphopods is much later than that of the cephalopods or any other extant class. Bouyancy may be the word for cepha- lopods but the “essence” of scaphopods escapes me. We need more scaphopod workers to develop new data on this most difficult class. Starobogatov (1974) has revised the Sca- phopoda. Placing them as a subclass of the order named Solinoconcha, he has included within them the Early Cambrian Voj/- borthella, a form that | do not consider to be a mollusk (Yochelson, Henningsmoen and Griffin, 1977) and the Middle Ordo- vician Polylopia, which might be a hyolith (Yochelson, 1968a). He also allies the class Xenoconchia Shimanskiy with the recon- stituted Solinoconcha so that the class is divided into two groups, those with an elongate tube open at both ends, and those with a tube closed at the apex. As noted, he derived the cephalopods from this latter group. | cannot accept Starobogatov’s views, but have not the space to refute them in detail. PELECYPODA, LAMELLIBRANCHIATA OR BIVALVIA If there is an essence of ‘‘pelecypodness” | believe it to be withdrawal from the epifaunal world by burrowing. Nucula and other protobranchs still process mud, mov- ing it inward by labial palps but most kinds of pelecypods filter water through their gills, assiduously pumping gallon after gal- lon, safe from many predators by hiding in the mud. Of course, there are exceptions to this generalization. A variety of bivalves live on the surface and Pecten is certainly active, but students of bivalves seem to be in agreement that pelecypods originally were dwellers within soft sediment and that surface dwelling—free, attached, cemented, or boring—is secondary. Thanks largely to the work of Pojeta (1971), the major groups of pelecypods can be documented back to the Ordovician. Pelecypods appear in the Early Ordovician and apparently underwent a dramatic radia- tion so that by the Middle Ordovician most lineages were established and by Late Ordo- vician even the pectinaceans appeared. For the animal to dig in, | think several features are needed. Obviously, two valves are required to open and close. This means ligament, teeth and sockets. Valves open only slightly and are rocked back and forth; teeth and sockets guide the valves. Finally, strong muscles are needed. This in turn means strong attachment to the inner surface of the shell on which it leaves a prominent scar inside. М muscles were not strongly attached, the valves could not be closed or rocked back and forth. There is a persistent myth of competi- tion during the Paleozoic between brachi- opods and pelecypods, with brachiopods slowly succumbing to the pressure of com- petition. To me, this seems most unlikely and Stanley (1968) has indicated that the great flowering of pelecypods occurred in the Jurassic, long after most kinds of bra- chiopods had become extinct. Paleozoic brachiopods and pelecypods often lived in different habitats; after all, brachiopods do not move, whereas most pelecypods do. In collections of fossils where both occur, nothing suggests competition; apparently there was sufficient food in the water for all filter feeders. Pelecypods are a fine example of a class which almost from its inception has т- cluded a great deal of diversity. In contrast, scaphopods show no diversity today or in the past. Since the Early Ordovician, pele- cypods have increased both in diversity and in number of individuals. In terms of biomass, they may be the most successful molluscan class. Pelecypods are also suc- cessful in that they live in marine, brackish 172 YOCHELSON and fresh water, whereas cephalopods and scaphopods are limited to marine conditions. During two centuries many schemes for arranging subdivisions of the class have been propounded. At least one scheme proposed recently by Russian workers (Nevesskaya et al., 1971) has not been critically evaluated by paleontologists in the western world. More new arrangements will result from the 1977 meeting in Lon- don on bivalves. There does seem to be consensus that protobranchs form a sub- class quite apart from other taxa in the class. The absence of a radula in the clams may be generalized as loss of the front part of the digestive tract in the Scaphopoda and further chopping off of the anterior portion in the Bivalvia (R. Palmer, oral communication, 1977). The necessity to feed by filtration may be an element in the clam success story. There is one major difficulty with the pelecypod story as recounted here. It con- cerns the genus Fordilla, which occurs in the late Early Cambrian and is widespread in the northern hemisphere. There is agree- ment that Fordilla is restricted stratigraph- ically and that it is a bivalved molluscan shell. A troublesome point is the early appear- ance of Fordilla and the lack of authentic records of the Bivalvia until the early Early Ordovician. There is a gap of nearly 75 million years. No acceptable theories aid us in evaluating gaps in the fossil record, but they are frequently associated with rare organisms. The record of the Bivalvia fol- lowing the dramatic late Early Ordovician radiation is one of steady increase in diver- sity and biomass. It is hard for me to imagine the Bivalvia originating with Fordilla, disappearing, and then reappearing with a great and dramatic diversification. Gaps do occur in the pelecypod record such as between Ordovician Babinka and younger lucinoids, but | do not think the apparent absence of Middle Cambrian to early Early Ordovician specimens can be dismissed. Fordilla has been variously classified. The current prevailing interpretation made repeatedly (Pojeta, Runnegar, and Kriz, 1973; Pojeta & Runnegar, 1974; Pojeta, 1975) is that Fordilla is a pelecypod. | disagree and hold that this genus could represent an extinct group of mollusks dif- ferent enough from pelecypods so that one might assign them a separate class rank. | base my argument on morphology of hard parts; the time gap is only secondary evi- dence, for the fossil record contains other gaps, though such gaps are usually associ- ated wtih rare groups. | have reexamined the type lot, as well as those figured by Pojeta, plus additional unillustrated specimens, and have collected a few. Based on this, a reconsideration of the genus is in preparation, but for the moment | can only present statements without much supporting documentation. Fordilla is a small shell showing closely spaced, fine, but prominent growth lines. Nothing is directly known of the original shell structure. The clean splitting away of shell from steinkern is inferential evidence of molluscan affinity. The shell is far smaller than many of the accepted pel- ecypod shells of the late Early Ordovician and the size range is near that of large living ostracodes. It does not show a promi- nent beak nor an obvious straight hinge line. А significant feature in my view is the lack of hinge teeth and sockets; a number of living pelecypod stocks, such as pholads Or pectinaceans, lack teeth or have them highly specialized, but so far as | can determine from the literature, no stock that lacks teeth is considered to be an ancestral group. Fordilla specimens from Siberia to be illustrated by Pojeta in a forthcoming paper are slightly asymmetrical near the center of the presumed dorsal edge; a slight overlap of valves at a single point is the most plausible interpretation. The notch in the margin of the steinkern is enigmatic. Internally, the steinkerns show a number of markings. Fordilla is interpreted as а pelecypod because it is reputed to have anterior and posterior adductors and a pallial line extending from the anterior adductor. | do not believe that published illustrations support such an interpretation. On none of the illustrations is there any clear evidence of a prominent adductor scar at the presumed posterior of the shell. The presumed pallial line is a greatly raised ridge whose proportions and shape are dif- ferent from those of living shells. | have not examined many tiny clam shells, but in larger ones the pallial line is not a promi- nent feature. The pallial line on Fordilla does not extend from anterior to posterior but only extends halfway across the shell and then disappears. There are living pel- PHYLOGENY OF ANCIENT MOLLUSKS 173 ecypods which have an interrupted pallial line, but like those which have lost hinge teeth, they are specialized forms. This “pallial'” line in at least one specimen is bifurcate (Pojeta, 1975, pl. 1, fig. 3). Near the middle of the valve, where the “pallial line” ends, there is a broad raised area, which in some individuals appears to be divided (Pojeta, 1975, pl. 1, fig. 1). Even allowing for shell curvature, this feature would appear to encroach further into the shell thickness than the pallial line. It also extends closer to the shell margin than the “pallial line. Pojeta (1975) considers Fordilla a part of the infauna, whereas Tevesz & McCall (1976) judge that it lived on the surface. Fordilla may be a pelecypod. If so in- terpreted, the inner features are so unique that, in my view, the genus should be placed in one subclass and all other known pelecypods be placed in a second subclass. That is not a good way of expressing the diversity found within the Pelecypoda. The discovery a few years ago of a living bivalved opisthobranch should make every- one who speculates about fossil mollusks exceedingly cautious (see Kay, 1968, for a review). Keen & Smith (1961, pl. 5) pic- ture this living snail carrying its bivalved shell; there is some similarity in size and shape of hard parts between Fordilla and Berthelinia! | suspect that Berthelinia is telling us that Mother Nature has a variety of ways of putting hard parts and soft parts to- gether and that we should not be too fixed in our interpretation of the interrelation- ship between the two. STENOTHECOIDA The genus Stenothecoides occurs from mid-Early Cambrian to early Middle Cam- brian; allied genera have about the same range. Like Fordilla it is widespread in the northern hemisphere. After the discovery of Neopilina there was a scramble to find muscle scars in cap-shaped univalves. Horny and Rasseti both illustrated some lateral markings on a slightly asymmetrical valve, plus fainter median markings. Not knowing what else to do with the creatures which were generally accepted as mollusks, Knight & Yochelson (1960) placed them in the Monoplacophora. In retrospect, it would have been better to treat them as Mollusca incertae sedis. Robison (1964) suggested a simple tooth and socket arrangement at the pointed end; anterior and posterior more often than not are misleading terms. Yochelson (1969) eventually published photographs of two specimens which are bivalved, asymmetrical and inequivalved. Akserina (1968) pub- lished the Class term Probivalvia, but Yochelson (1968b) also had discussed the concept and the name Stenothecoida in an abstract which had appeared earlier that year. There are no rules of nomenclature at this level, but there seems to be general acceptance of the principal of priority and Stenothecoida is the name used. Multiple independent discovery of a concept in sci- ence (Merton, 1961) is not proof that the concept is right, but it does suggest that the concept be examined seriously. Stenothecoida is an embarrassment to most phylogenetic schemes. The late Henning Lemche (oral communication, 1975) had worked out to his satisfaction a program of evolution for the entire inverte- brate world, except that he could not account for Stenothecoides. However, Runnegar & Pojeta (1974: 316) have dis- missed it out of hand. They wrote “We offer the alternative suggestion that it may have been a bivalve monoplacophoran, with the lower (smaller?) valve formed by the sole of the foot. A few living limpets form a second valve in this way, although in these cases the lower valve is cemented to the rocks.” The same view was expressed by Pojeta & Runnegar (1976: 44); both works refer to a paper by Yonge on Hipponix. Even if | were to grant that Steno- thecoides is a univalve-which | emphat- ically do not grant—there is no basis for assigning such a univalve to the Mono- placophora. Specimens are asymmetrical and the internal markings are quite unlike muscle scars seen in generally accepted fossil monoplacophorans. Although other points may be arguable, | believe that it is clear that Runnegar & Pojeta’s comparison of Stenothecoides with Cretaceous to Holocene Hipponix is spuri- ous and can be rejected. The bivalved Class Stenothecoida does exist. | have interpreted these strange little fossils as living on the surface, unattached, but dorsal-ventral like oysters. This may not be correct, but no other bivalved modes of life have been suggested. | point out that if Stenothecoides and its relatives can have two valves and 174 YOCHELSON yet not fit in the Bivalvia, perhaps this helps strengthen the case for а bivalved Fordilla that is not in the Bivalvia. About as many genera of Stenothecoida are known as are known of Scaphopoda, but these fossils are rarer than fossil sca- phopods. POLYPLACOPHORA The Polyplacophora are mollusks which have eight shell valves. The nature of my material forces me to be a paleoconcholo- gist. As a pragmatist, | prefer names which deal with hard parts, not soft tissue. Thus, | think it is high time we dropped the term Amphineura. Likewise, | should use Bi- valvia, though the informal term bivalve may cause confusion. The issue of Aplaco- phora or Caudofoveata and Solenogastres (Salvini-Plawen, 1972) is not mine to be involved with. The few examples of these soft-bodied creatures | have seen look im- pressively different from the polyplaco- phorans. Most living “‘coat-of-mail’’ animals cling to rocks. In part, this allows them to live in areas of strong wave action, but perhaps equally important this permits them to fight dessication. Lepidopleurus lies on mud in the deep sea and is regarded as more generalized by researchers of poly- placophorans. The Polyplacophora are more diverse than the Scaphopoda. They are certainly less diverse than the cephalopods, by sev- eral orders of magnitude. A point that | have attempted to make repeatedly is that distinctiveness is what separates one taxon- omic unit from another at a comparable level. Diversity is a measure of the variety within that level. All molluscan classes are distinctive, but some are diverse and some are not. The Polyplacophora are one that is not. In the mid-Paleozoic, isolated plates with insertion plates occur. These tiny pro- jections are what allow the plates to over- lap, yet retain flexibility. Some of these forms, well preserved in the Mazon Creek Shale, are virtually identical to living genera, including such details as girdle width, pattern of scales on girdle, and marginal spicules. The early Paleozoic forms seem to have been like Lepido- pleurus; likely they were cryptic, but could not cling so well to hard surfaces. The ES older genera may have had major overlap of the valves, but the valves might have been set in a long girdle, allowing flexi- bility of the animal. The oldest reported polyplacophorans are from the very latest Cambrian strata (Bergenhayn, 1960). The material consists entirely of steinkerns and | am not con- vinced that these truly are representatives of the group; they may be fragments of other fossils. Authentic earliest Early Ordo- vician material has been recovered recently (B. Stinchcomb, written communication, 1977). The oldest published fossils of this class that | accept without question occur in the late Early Ordovician (Smith & Toomey, 1964). They are associated with sponges and algae in some local build-ups of hard ground on a limy-bottom shallow sea. | cannot think of any fossil mollusk in the Cambrian which might serve as an ancestor to the polyplacophorans. Several points can be noted about living polyplacophorans. The mantle cavity is lateral, not posterior. It may well be that the paired lateral position is a result of the crawling and clinging habit of an elongate form. There has been speculation about segmentation of mollusks; the numerous ctenidia of some living forms show that pseudometameric repetition occurs within the mollusks and one need not get too excited about multiple repetition of organs. Although polyplacophoran plates usually look like the roof of a house and are easy to identify, some isolated fossil echinoderm plates have the same general shape. Murphy’s Law has been a dominant theme throughout the history of the mollusks. MATTHEVA Mollusks with more than one calcified shell tend to disarticulate after death. The living polyplacophorans provide one model of how several hard parts are used during life, but this need not be the only way more than one piece was assembled. The genus Matthevia occurs in the late Late Cambrian. Yochelson (1966) observed that two forms occurred together and inter- preted them as front and back pieces with two prominent muscle insertions in each piece. One of the mistakes | made at that time was to be overly generous with alter- native hypotheses, for | indicated that there might be small intermediate plates PHYLOGENY OF ANCIENT MOLLUSKS 175 (Yochelson, 1966, fig. ЗВ). | did not really believe it then, and am now fully satisfied that these smaller “plates” are only shell fragments. At the time | wrote, my guess, based on slight evidence, was that Matthevia lived on algal heads and used the weight of the anterior and posterior pieces to hold itself in place. Subsequently, | have seen speci- mens associated with stromatolites and the lithologic evidence of strong water motion in the areas of occurrence is compelling. This would certainly account for the wear of pieces after death and sorting of poste- rior from anterior pieces. Likely Matthevia had some flexibility and this permitted it to move among the algal heads. Mere weight to hold it in place was too simple an answer, but it had a basis in fact. Runnegar & Pojeta (1974) and Pojeta & Runnegar (1976: 44) have downgraded Mattheva from class status. In their view Matthevia is a primitive polyplacophoran. They have introduced a number of inter- mediate shapes of pieces and restored it to look something like a hedgehog or a stego- saur. | can think of many unacceptable points to their reconstruction, but an obvi- ous one is that this is hardly an adaptive shape for an animal clinging to rocks in turbulent water. Another basis for their taxonomic downgrading is to illustrate a form from the Early Ordovician presum- ably intermediate to more conventional polyplacophorans. Dr. Runnegar has kindly permitted me to examine specimens of the presumed intermediate (September, 1977), and to me this “missing link’’ is simply a thick valve of Chelodes with the typical single broad and shallow muscle insertion pocket; no polyplacophoran has more than one muscle insertion per plate. It has a clearly differentiated head or tail plate which by its absence of any muscle inser- tions is unlike a Matthevia. In my view the Class Mattheva is well-founded morpho- logically. HYOLITHA The hyoliths are an extinct group. Whereas Stenothecoides and Matthevia are uncommon fossils and occur only in a narrow time interval, Hyolithes and its allies are exceedingly abundant in the Cam- brian, provided of course one looks in the proper rock facies; they seem to have pre- ferred a moderately soft bottom. Hyoliths are rare in the Ordovician and exceedingly rare in younger beds, but they do occur in the Permian and thus had a duration of nearly 300 million years. They should not be ignored. Hyoliths are greatly elongated cones, closed at the apex. They are slightly curved and are bilaterally symmetrical. The tube- like shell has an operculum associated. There has been some confusion with “worm tubes” and with the opercula from such tubes, but the hyoliths stand as a moderately homogeneous group, though a few generic taxa still should be eliminated and moved to ‘’Vermes.”’ For decades, virtually no generic diver- sity was recognized, though many species were named. Currently, some dozens of genera are recognized. One might say the diversity approximates that of polyplaco- phorans, but that the fossil biomass may have been much larger. Two principal orders are recognized. The Orthothecida have a simple aperture, an operculum which is retracted into the tube, and a cross-section that is smoothly curved and non-angulate; it can be circular, oval, or bean-shaped. The Hyolithida are more complex. Most commonly their cross-sec- tion is triangular, with the base of the triangle extending forward to form a semi- circular shelf. The operculum covers this shelf, as well as the aperture. Protruding between the lateral edges of aperture and operculum are a pair of curved ‘‘whiskers,”’ which have received an inordinate amount of attention. The best guess is that they acted to stabilize the shell, like the out- rigger of a Polynesian canoe. It might be mentioned that peculiar as the whiskers are, there are a number of hyoliths—those in the Orthothecida—that did not have them. | have refrained from mentioning the phylum assignment of the hyoliths. For more than a century they were placed hither and yon, though most commonly at some vague position within the Mollusca. In 1963, Marek finally made a definite statement and placed them as a class of Mollusca. Not everyone agrees that they are mol- lusks. Runnegar et al. (1975) judged that they are different enough to form an ex- tinct phylum. Marek & Yochelson (1976) amplified and reinforced Marek’s original remarks that these fossils form a class of mollusks. For me, the most convincing argument in favor of Hyolitha being mol- 176 YOCHELSON lusks is the description by Runnegar et al. (1975, fig. 5) of typical molluscan cross- lamellar shell structure in a Permian speci- men. Runnegar (oral communication, 1977) holds that the ancestral mollusk possessed a dorsal epithelium which subsequently calci- fied; the reconstruction of soft parts for Hyolitha used by Runnegar et al (1975) indicates that the shell was not dorsal and the group is not molluscan. In proposing an extinct new phylum for consideration, | (Yochelson, 1977) suggested that while the shell of hyoliths does not look much like that of living mollusks, it is not so distinc- tive as to warrant phylum rank. There seems to be no room here for compromise. If the hyoliths are mollusks they are different enough to be a class, and if the hyoliths are not mollusks, they are different enough to be a phylum. The question is how different? In my view, Hyolitha belong within the Mollusca, but | cannot add new evidence to what has been presented; each systematist must decide for himself. To a large extent the notion of deposit feeding with palps or tentacles, which might have been present in ancestral mollusks, turns on the question of whether Hyolitha are included or excluded from the phylum. The Late Paleozoic class Xenoconchia (Shimanskiy, 1963, Shimanskiy and Barskov, 1970) has not been critically re- viewed; Runnegar and Pojeta (1974) did not mention it. On the basis of examina- tion of some specimens and discussion with Dr. Shimanskiy (September, 1975) my guess is that class rank may be unwar- rented. A few Mississippian specimens (cur- rently in the group) might be steinkerns of a nearly symmetrical Platyceras (Or- thonychia). However, Permian material is not so readily interpreted, for in some specimens one side is slightly flattened; others are circular in cross section. These shells may be related to the large Permian fossil Macrotheca. That genus fits within the Hyolitha but it is so large and expands so rapidly that status as a separate order probably is appropriate. Specimens of Xenoconchia are far rarer than those of Stenothecoida, and like them provoke major problems in interpretation and assignment. GASTROPODA In pelecypods there is a reasonably close relationship between the form of soft parts and the enclosing two valves, but in gastro- pods the relationship of shell to soft parts is far less obvious. Pelecypods can be de- fined moderately well by reference only to the hard parts, and externally shelled cephalopods can be defined quite well using only hard parts, but the gastropods are another matter. А gastropod 15 a mollusk whose soft parts have undergone torsion (twisting) of the nerve commissures and certain other soft parts. Over the years | have asked my neontological colleagues to phrase a class definition without reference to soft parts. The request is treated with amazement, amusement, or both, and | conclude that one cannot assign fossils to Gastropoda with as high a degree of assurance as with other classes of extant mollusks, except perhaps the scaphopods. One might turn this around and suggest that since torsion is unique, and as a consequence the gastro- pods show the greatest independence of shape between soft and hard parts, Gas- tropoda is the most noble and most ad- vanced of the molluscan classes. | do not believe it, but | put forth the claim to annoy those who extoll the wonders of living cephalopods. If one judges by the number of included lower-level taxa, gastropods are the most diverse of all mollusks. It is appropriate to emulate a gastropod and to stick my neck out to make progress. A class in system- atics may mean, approximately, a major mode of life. Originally, gastropods were grazing herbivores. The gastropods may have chosen to emphasize the radula, after torsion produced a neck and greater flexi- bility of movement. This mode of life was so successful that the gastropods were able to proliferate, undergo other anatomical changes, and invade the major niches oc- cupied by other classes. Thus, there are gastropods that float and swim, hunt through the bottom, filter water, and gas- tropods that cling to rocks. Incidentally, gastropods are the only mollusks to invade the land, and several stocks have done so. Biomass may not be as great as the pele- cypods, but gastropod diversity to me illus- trates the importance of brain, or at least head, over brawn. Gastropods were not the oldest herbi- vores; some “worms” have that distinction (Edhorn, 1977). There is an increase in snail diversity from Late Cambrian through Early Ordovician and speculation that they PHYLOGENY OF ANCIENT MOLLUSKS 177 inhabited algal stromatolites (Garrett, 1970). If this idea is substantiated, the gastropods may go down in biologic history as the ultimate grazing machine. Because septa are uncommon and com- plex suture patterns have not developed where septa are present, the shell of a fossil gastropod is not nearly so complex as that of а fossil cephalopod. However, shell structure can be elaborate and in some groups the shells are aragonite, in others calcite-aragonite, and in still others calcite. Diversification of shell composition, as well as anatomical change, may help explain the success of the class. Some shell shapes occur time after time among fossils and one has difficulty distinguishing conver- gence from true phyletic lines. However, the main lines are fairly clear, with asym- metrically coiled shells reasonably close to living pleurotomariaceans appearing in the Late Cambrian along with macluritaceans. More advanced, presumed single-gilled forms appeared later. Still later, the neritaceans, and opisthobranchs (Kollmann & Yochelson, 1976) appeared in the mid- Paleozoic. Pulmonates appeared by late Paleozoic time (a review by Solem & Yochelson is in preparation). Among the marine forms, the greatest change seems to have occurred in the Jurassic with a dra- matic increase in advanced prosobranchs. Naturally, the history of gastropods is not that simple, and | can think of at least four grave complications. First is the posi- tion of the Macluritacea; Linsley (in press) has considered their possible life-habit and discussion of their high-level systematic position can wait. Second is the position of the Bellerophontacea which will be treated later in the section. Third is the position of Aldanella, and fourth is the placement of Pelagiella and its allies. These last two may be combined under the general issue of “when did gastropods first appear’’? | have alluded earlier to the issue of worm tubes and | have noted the simplicity of the gastropod shell. A/danella was named from late Early Cambrian rocks in Massachusetts, but first received promi- nence as a member of the Tommotian fauna, an unusual Siberian assemblage of presumed early Early Cambrian fossils. Aldanella is closed at the apex, tubular, and is coiled in three dimensions and there- fore fits a superficial conchological concept of a gastropod. | do not agree with that assignment (Yochelson, 1975) for A/danella is based on tiny (1-2 mm) steinkerns, so that no information on the shell is avail- able. The Siberian specimens | have seen show a high degree of individual variation, more so than one would anticipate in gas- tropods. In addition, some __ illustrated steinkerns show a straight early portion of the tube not at all like a gastropod proto- conch. Finally, in a large collection of steinkerns of comparable size from Ordo- vician rocks on Spitsbergen (Bockelie & Yochelson, in press), the shape of A/danella is repeated in a population with so much individual variation that its non-molluscan nature is obvious. А/дапе//а is, in my view, a tiny worm tube and adds nothing but confusion to the phylogeny of mollusks. Pelagiella is another matter, and | do not doubt its molluscan nature. It is asym- metrically coiled, and in that feature is like a gastropod. However, the whorl expands at a rapid rate and the proportions of this small shell are quite unlike those of gener- ally accepted gastropods. Runnegar & Pojeta (1974) consider A/danella a pela- giellid, but | fail to see how the two groups can be considered related, for their shapes are distinct. In my opinion, the position of Pelagiella is comparable to that of Fordilla, in that both resemble an extant class, but appear early geologically. If it is necessary to bolster distinctiveness in hard part morpholo- gy with speculation regarding soft parts, | can, to my satisfaction, reconstruct Pe/agiella as an asymmetrically coiled non-torted mollusk. А reconstruction of hypothetical soft parts is not proof. For the moment, no clear-cut morphologic evidence can be de- rived from the hard parts to distinguish Pelagiella from the Gastropoda. There has not been time or inclination to study the Pelagiellacea carefully. In large measure, this is because specimens are uncommon and most are not well preserved. A few good specimens might provide a wealth of data and my guess is that when these are found, they will show that pelagiellids are not gastropods. It is now appropriate to turn to the Bellerophontacea, even though this essay is not much concerned with taxa below the class level. Bellerophontacea are coiled bi- laterally symmetrical shells that occur in the Paleozoic. The classic view is that these shells gave rise to the asymmetrical ones, but | still prefer to consider Bellero- phontacea as a specialized offshoot of the Pleurotomariacea (Yochelson, 1967). 178 YOCHELSON During the early 1960’s when the con- cept of Monoplacophora was new, Horny found specimens of the coiled Middle Or- dovician Cyrtolites which showed several sets of paired muscle scars. Subsequently, a few other coiled forms were found. | have found this to be interesting (Yochelson, 1967), but my conclusion a decade ago is the same as it is today; just because a few presumed monoplacophorans have coiled shells, it does not follow that Bellero- phontacea are a branch of the Mono- placophora. Starobogatov (1970), Runnegar & Pojeta (1974), Pojeta and Runnegar (1976: 32) and Runnegar & Jell (1976) moved all Bellerophontacea into the Monoplacophora mostly because of the superficial similarity of а bilaterally symmetrical shell т Cyrtolites and Bellerophontacea. In making this sweeping revision, several points have been ignored. Years ago, there was good evidence of lateral retractor muscles т bellerophonts (Knight, 1947); this has been further reinforced by additional finds. The close resemblance of the slit to that of pleurotomariacean gastropods has been ignored. In addition, the inductura, a sec- ondary deposit seen in many groups of gastropods, has not been emphasized. No undoubted monoplacophorans show such a deposit. | know of no theoretical recon- struction of coiled Monoplacophora which would place a mantle fold in the area where the inductura was deposited. | do not fully understand how the defi- nition of Monoplacophora came to be ex- panded to include coiled shells which show no sign whatsoever of dorso-lateral multiple paired muscle scars. Perhaps it is a mis- understanding of exogastric and endogastric coiling. Try as | may to follow the logic of those who state that the extinct Bellero- phontacea did not undergo torsion and therefore are Monoplacophora, | cannot. Let me say once again that it is impossible to determine whether an extinct form has undergone torsion. The hard part mor- phology indicates the Gastropoda as the class for Bellerophontacea. Secondary to this conclusion, reconstructions of pre- sumed bellerophontacean soft parts (Knight, 1952) show those of a gastropod fitting well into this shell form. Protowenella (Runnegar €: Jell, 1976), from the Middle Cambrian of Australia, 15 a tiny steinkern about 1mm across, judged by its authors to be the world's oldest bellerophontacean. | am reasonably certain that the paired lateral constrictions in the cross section are similar to those in the cross-section of undoubted worm tubes. It is circular reasoning to place fossils which cannot be assigned with any degree of confidence to any phylum in the Mollusca, draw inferences as to the soft parts, and use this as a basis for classification. ROSTROCONCHIA | am not the only one in recent years to have proposed classes of mollusks whose members are entirely extinct. In 1973, Pojeta and others proposed the class Rostroconchia and included specimens ranging from Early Cambrian through Permian age. Pojeta & Runnegar (1976) have elaborated upon their class concept. The rostroconchs are laterally com- pressed forms, some of which show a uni- valve protoconch, and are presumed to be non-torted mollusks. It is judged by the authors of the class that the shell split laterally as a consequence of growth, so that the animal was functionally nearly a bivalve. In one order, Ribeirioidea, a promi- nent plate occurs at the shortened, pre- sumed anterior edge of the shell, with a gape continuous from anterior to posterior. In my opinion, this hard part morphology is distinct at the class level from that of other mollusks. Two other orders, Ischy- rinioida and Conocardioida, emphasize a more symmetrical growth in terms of pre- sumed anterior and posterior, and a promi- nent posterior siphon, respectively. These are not as distinct from pelecypods as the ribeirioids, and are open to further specula- tion on their relationship to ribeirioids. Though | may accept the class, it does not follow that | accept all members of it. The group is certainly well developed in Late Cambrian - Early Ordovician; it is not reported by Pojeta & Runnegar (1976) to occur in the Middle Cambrian. The Early Cambrian record is based on two genera. One of them, Watsonella Grabau, comes from eastern Massachusetts in isolated boul- ders about the same age as those producing Fordilla. Others have seen a similarity be- tween the two, and | am quite satisfied that Watsonella is a subjective synonym of Fordilla. The other Heraultipegma, Early Cambrian genus is known from France and PHYLOGENY OF ANCIENT MOLLUSKS 179 Siberia. It does not weaken this class con- cept to note once again that evidence of molluscan affinities should precede assign- ment to class; phyllocarid crustacean аг- thropods are only one of the groups which have a shape convergent to that of the Rostroconchia. When Heraultipegma was first proposed, it was thought to be an arthropod. More recently, it has again been independently interpreted by two workers (Muller, 1975; Missarzhevskiy, 1976) as an arthropod. However, Runnegar & Pojeta (1974, fig. 4) accept it as a mollusk and as the ancestral rostroconch. They further in- dicate a series of younger morphological transition forms which lead to the principal subdivisions of the class. Some of these genera in the presumed phyletic lines of the class are coeval, rather than sequential. | have seen Muller’s specimens from France which are steinkerns; in my view they do not yield sufficient information to be assigned to any phylum; it remains to be proven that the class Rostroconchia appeared before the Late Cambrian. The dramatic radiation of the ribeiriids in the Late Cambrian might be the experimenta- tion of a group which has just developed and thereby been able to move into a new ecological niche. | can understand a new group appearing and diversifying rapidly. | cannot understand a new group being rare for 50-75 million years, and then evolving rapidly. It has been speculated (Yonge, 1953) that the Bivalvia might have been derived from a strongly compressed univalve. In this sense the Rostroconchia appear to be an ideal precursor to the pelecypods. If the pelecypods began in the Early Cambrian with Fordilla these two classes are nearly contemporaneous in time of origin, or at least in time of earliest known representa- tive. If the Bivalvia began in the early Early Ordovician, the two classes would be se- quential. Nevertheless, the step from uni- valve to bivalve is abrupt and it seems impossible to derive two centers of calcifi- cation from one centre of calcification ex- cept in a discontinuous manner. There is a large morphological gap between late Early Ordovician pelecypods and contemporane- ous rostroconchs. | cannot accept the con- cept as presented by Runnegar & Pojeta (1974; Pojeta & Runnegar, 1976) of a series of adult forms in a morphologic transition series linking the two classes. It is possible as suggested by Runnegar & Pojeta (1974) that the Rostroconchia were directly ancestral to the Scaphopoda. № my interpretation of the earliest sca- phopods is correct, they overlap only with the highly specialized conocarditids. Per- haps the two classes were derived from a common ancestor and are not more closely related. MONOPLACOPHORA It might have been more logical to begin this review of classes with the Monoplaco- phora, for it was the acceptance of a sixth class of living mollusks in the late 1950's which began all the current ferment con- cerning molluscan phylogeny. However, many problems surround this taxon, seem- ingly with several kinds of mollusks mixed together, so perhaps it is best kept till last. Malacologists should recall the paper by Knight (1952). He did not invent the con- cept of the Monoplacophora, as credit for that goes to Wenz (1940), but Knight did extend and clarify it. His conclusion was that the Monoplacophora were not particu- larly remarkable. Knight divided the class Gastropoda into the subclasses Anisopleura (gastropods in the traditional sense) and Isopleura, which in turn was subdivided into Monoplacophora and Polyplacophora. There the matter stayed until Lemche (1957) described Neopilina, a living mono- placophoran. All honor to Weopilina, for without it the neontologist would have payed slight attention to the fossil record and the paleoconchologist would probably never have had the courage to propose high- level molluscan taxa, all of whose members are extinct. One may ask, what is a mono- placophoran? The answer varies, but let me give my opinion for fossils. It is a bilater- ally symmetrical, wide molluscan univalve shell, most characteristically limpet-shaped, which shows prominent multiple muscle scars. Wenz was most impressed with paired scars of the Devonian Cyrtonella, a strongly arched form; Knight concentrated on Tryblidium, a flattened Silurian form. | suppose that it should not have come as any shock that a few monoplacophoran shells grew through several coils. Still, it was a surprise when Horny found Ordovi- cian Cyrtolites to have several whorls and multiple paired scars. Some of us do not yet have the concept of coiled Mono- 180 YOCHELSON placophora in perspective. As | mentioned under comments about Bellerophontacea, Cyrtolites does not have an inductura, and bellerophontaceans do; there are other mor- phologic differences between coiled mono- placophorans and bellerophontaceans. Let us now consider exceptions to my definition as regards shape. Not all mono- placophorans are absolutely _ bilaterally symmetrical. Yochelson (1958) described Early Ordovician Cyrtonellopsis from three specimens, the apexes of which are central, to the right, and to the left, respectively; there are no muscle scars known, so assign- ment of Cyrtonellopsis must carry with it some uncertainty. Even granting that, there should be a limit to the degree of asym- metry. | judge that assignment of Steno- thecoides as а monoplacophoran—regardless of whether it is a bivalve—to be incorrect, because the asymmetry shown by one valve greatly exceeds the asymmetry shown in any generally accepted monoplacophoran. It seems to me even more preposterous to consider Pelagiella as an asymmetrically coiled monoplacophoran. Muscle scars are not common in fossils. There are a greater proportion of scars in specimens of Monoplacophora than in specimens of other classes, but even so scars are rare. Patella and its allies show a scar quite unlike that of typical mono- placophorans but have a similar limpet-like shape. Horny (1961) has described Damilina, a Silurian patellid gastropod with a horseshoe-shaped muscle scar, so it is clear that the ranges of these two limpet shell forms overlap. | would be exceedingly cautious in assigning a fossil to the Mono- placophora. | am not satisfied that all taxa assigned to the group by Knight & Yochel- son (1960) are correctly placed. The number of scars in monoplaco- phorans is not constant. For a brief time, it was nice to think of eight plates in the Polyplacophora and eight scars in Tryblidium, but many genera have fewer than eight pairs. One tends to ignore Archaeopraga (Horny, 1963) which is the same size, shape, and age as 7ryblidium and has a single pair of elongate lateral scars. Bellerophontacean scars are not mono- placophoran scars, both because of their shape and because of their position within the shell; few bellerophontacean steinkerns yield data on musculature. One can argue that some nautiloid cephalopods were monoplacophorans, for the known muscle scars are similar in size, shape, and, to a certain extent, position (Mutvei, 1964). In- deed, if the relationship of Knightoconus as an ancestor to cephalopods is correct, that observation makes good sense. An impor- tant general feature is that muscle scars are prominent because the muscle is attached to the shell firmly. Weak attachment and non-attachment does not leave any shell scars. As | mentioned in connection with clams burrowing, firm attachment is neces- sary for some functions of the animals. For clamping against a hard substratum, strong attachment of muscles may be a prerequi- site (R. M. Linsley, oral communication, 1977). Let us consider Scenella. This genus occurs in the Early Cambrian and may straggle up to the early Middle Cambrian. One specimen is known at the end of this range which shows muscle scars, paired, but slightly asymmetrical. Further, the external shape is not similar to that of a typical Scenella, which has an elaborate radial Ornament and periodic development of a frill. | would say that the oldest mono- placophoran known for certain is Late Cambrian for it shows multiple sets of paired scars in a limpet-shaped shell. There is no good evidence of the class in older rocks, though the one early Middle Cam- brian specimen is suggestive. Even if this specimen is a slightly asymmetrical mono- placophoran, it may not be correctly as- signed to Scenella for it has a smooth external shell. The authentic Early Cam- brian Scenella may be а monoplacophoran or something else. The Early Cambrian Aktugaia (Missar- zhevskiy, 1976) has recently been described as a monoplacophoran because it has multi- ple paired muscles. | have not seen any specimens, but the assignment is suspect to me, for “worm” opercula such as Mober- gella or Discinella also show multiple paired scars. The wedge-shape of Aktugaia is not like that of these two genera, nor is it like the shape of a monoplacophoran. Aktugaia is tiny, as is Mobergella. In my opinion, the oldest undoubted monoplacophoran is still Late Cambrian. One may gather from the prior discus- sion that subdivision of the class is in a bit of a mess. Rosov (1975) added a fourth order to the three currently recognized. Runnegar & Jell (1976) placed two orders in synonymy and added the Bellero- PHYLOGENY OF ANCIENT MOLLUSKS 181 phontida. Anyone who thinks that high level classification is a slow methodical process has not been considering the Mono- placophora. Biomass for the class was always low, though Early Paleozoic specimens are more common than living Veopilina. The number of included genera, if one uses my defini- tion, is slightly greater than that of the Polyplacophora during the comparable time interval; if one uses the class limits of Runnegar & Pojeta, the number of Paleo- zoic genera is the same order of magnitude as in early Paleozoic gastropods. Presum- ably fossil monoplacophorans were herbi- vores; it is not fully clear what Neopilina eats. | do not think they were as mobile as gastropods or even some living limpet gas- tropods, and perhaps this explains in part their rarity relative to fossil gastropods. Next is the issue of endo- and exogastric coiling. Some fossil monoplacophorans had a larval stage, and if this stage was very long | warrant that the shell curved or even coiled for ease of movement, economy of material, and a host of other reasons. If the apex was strongly anterior, | can see how the larval coil in a wide, low-spired shell might be opposite from that of a living limpet gastropod, such as Pate//a. However, many presumed monoplacophorans are conical, not flattened. In these conical forms the apex is central or subcentral. The assumption has been made by some writers that the apex must curve or coil to the anterior. They gave no clear rationale for that assumption and | tend to reject it. In the absence of differentiated muscle scars, or indeed any muscle scars at all, one cannot determine anterior or posterior and thus judge ‘‘endogastric or exogastric”” coil- ing. For example, Pojeta & Runnegar (1976, pl. 15) illustrate a curved proto- conch of a species possibly belonging to Macroscenella. No muscle scars are known from that genus and this protoconch is squarely in the center of the shell above a smooth regular oval aperture. How does one know which is the right or the left side of the shell? There is concern about the position of the mantle cavity in coiled monoplaco- phorans. Neopilina has paired lateral mantle cavities, as do the polyplacophorans. If lateral cavities were ancestral and the soft parts were then crowded by increased cur- vature, the mantle cavity might have moved posteriorly or it might have moved anteri- orly. Knightoconus was restored with a posterior mantle cavity, but this was to align it with cephalopod soft parts, not for any intrinsic reason. It is well known that archaeomollusks had a posterior mantle cavity because that is the way Lankester thought of the ani- mal, and that torsion brought it forward. As | have never dissected an archaeomol- lusk, | do not know this. There is much concern about problems of fouling of the mantle cavity by discharge from the anus. In Neopilina the anus is posterior. Does it necessarily follow that anus and mantle cavity were always linked? Perhaps Cyrto- lites and similar coiled forms had an ante- rior mantle cavity and a posterior anus discharging over the foot. An anterior man- tle cavity need not be a result of torsion but might have been a consequence of the shell broadening anteriorly. Perhaps some difficulties lie in our presumptions about the ancestral form. | think that monoplacophoran shells coiled through several volutions were car- ried on the dorsal part of a foot, precisely as gastropods carry their shell. Because there was no torsion, the soft parts were probably less extensible and less flexible than those of gastropods. In particular, the scars are less than one-quarter of a volution inside the body whorl and the organism could not retract deeply into the shell. However, the animal functioned moderately well, for Cyrtolites is not rare. The model of the endogastric mono- placophoran, coiled with the shell balanced over the head, may have been mechanically plausible (Linsley, 1978) though | am still not convinced that the organism ever ex- isted as reconstructed. What evidence is available from the limpet-shaped Late Cam- brian monoplacophoran Kirengella indicates a subcentral position for an apex which curves toward the posterior. Until better evidence is presented | shall assume that all monoplacophorans which coil through more than one volution follow this same geometry. | fail to see why non-torsion and coiling over the head must be equated. If one can accept a similarity between Knightoconus and Plectronoceras, an increase in degree of coiling of a Knightoconus-like shell would result in a shell like Cyrto/ites. Muscle scars in Cyrtolites consist of several pairs, above on the dorsum, and a U-shaped ring below ending on the lateral slopes, situated slight- 182 YOCHELSON ly deeper in the whorl. If Cyrto/ites coiled as do gastropods, the attachment site for the foot retractors could be the U-shaped ring, and the discrete pair of scars on the dorsum near the aperture would be in a position to retract the head (D. Schindel, oral communication, April 1977). This way the head would be last in and first out. There is a point in homology of muscle scars which could be pursued a bit further. Pojeta & Runnegar (1976) show several members of the Rostroconchia with muscle scars. The general pattern is one or two pairs of discrete scars toward the presumed anterior, and a U-shaped scar posterior to these. If the muscles are in a similar posi- tion to those of Cyrto/ites, Runnegar & Pojeta have one group or the other back to front. It is more likely that the muscula- ture pattern is similar in both groups. | accept their interpretation of anterior in Rostroconchia, and thus have another line of evidence to support my orientation of the coiled monoplacophorans. An argument has been made that Cyrtolites and other coiled monoplaco- phorans had a posterior mantle cavity. This is based on the assumption of water flow- ing laterally into the shallow water cavity and out at the dorsum (Linsley, 1978). The morphologic evidence derived from shell shape suggests one orientation and the muscle scars suggest another. Whichever may eventually be judged to be correct, the coiled monoplacophorans occur sporad- ically beginning only in the Middle Ordo- vician. Linsley (in press) has argued that their morphology—assuming coiling was over the head—was appropriate to have given rise to the Bellerophontacea, but | feel that they occur geologically too late. As a general principle, | cannot put much faith in presumed morphologic series which do not occur in stratigraphic sequence. Let me now attack the “Helcionella heresy.’’ This genus and its allies are small, bilaterally symmetrical curved shells, very strongly compressed laterally; they char- acteristically occur in the late Early Cam- brian to early Middle Cambrian, but are found in slightly younger and older rocks. Commonly, there are periodic swellings and compressions of the shell, but some genera such as the Early Cambrian A/dabanella have smooth sides. The aperture is simple and shows no reentrant. In 1960, Knight, Batten and Yochelson did not know quite what to do with these mollusks, but finally followed Knight (1952) in relating them to the Bellerophontacea on the assumption that the helcionellaceans gave rise to forms which both developed a slit and elaborated the curvature to coiling of several whorls. Specimens are smaller than most bellero- phontaceans. Runnegar & Pojeta (1974), Pojeta & Runnegar (1976), and Runnegar & Jell (1976) regard the Helcionellacea as mono- placophorans. They are judged to be mono- placophorans because it is stated that the curvature is toward the anterior. | disagree both with the interpretation of curvature and with the class assignment. On the matter of curvature, as noted, it was suggested by Yochelson, Flower & Webers (1973) that one need not have anterior curvature in a high conical shell. Earlier, Knight (1952) schematically re- stored Helcionella with the shell gracefully curving toward the posterior so that the foot is partially covered by a train, as the aperture expands. Restoration with the curvature forward results in an organism very nearly as awkward as the hypothetical forward-coiled monoplacophoran. Some consideration has to be given to the point that these organisms crawled and that they could have crawled far more readily if the shell was carried on the foot, not balanced on the head. Yochelcionella has been described from the early Middle Cambrian (Runnegar & Pojeta, 1974). It is like He/cionella, but midway on the concave surface of the shell is an open tube. This is indicated as an inhalant siphon, partially because the ani- mal as interpreted by Runnegar & Pojeta would have crawled with this tube pointing forward. Because helcionellaceans are both compressed and small, lateral mantle cavi- ties are unlikely. Even assuming that Yochelcionella could have moved in an awkward position as suggested, the water flow is completely at odds to the general molluscan pattern. A general rule for water flow is “in below, across the gills, and out above,” though in some shell geometries there are modifications. For Yochelcionella the flow would be in above and out below. | do not see how this tube could function as an inhalant siphon, advancing through the water. Additional species of Yochelcio- nella with the open tube near the apex (Runnegar & Jell, 1976) make even less sense interpreted as functionally inhalant. The water stream would have to bifurcate PHYLOGENY OF ANCIENT MOLLUSKS 183 after entering to bathe the gills. However, if interpreted as an exhalant siphon, with the animal crawling away from the spent water, the tube makes a great deal of sense. For helcionellaceans without this tube, the train would have aided movement of water out of the posterior mantle cavity. To recapitulate, | hypothesize that coiled monoplacophorans had an anterior mantle cavity over the head and helcionellaceans a posterior mantle cavity. Water circulation hypotheses were divided from the shape of the hard parts. To return again to the hard parts of a class, it is important to note that a very large variety of shapes can be made from a univalve. | have mentioned slight asym- metry; let me now touch on lateral com- pression. Typical Monoplacophora have an elongate wide oval aperture in low shells. In conical shells, the aperture is still widely oval, and even in coiled forms the aperture is relatively wide. The Helcionella shell is strongly compressed laterally. Without speculating at all about soft parts, | would judge that the Helcionellacea exceed, by quite a degree, the morphologic variation of hard parts which should be allowed in the class Monoplacophora. | cannot help but add that nothing whatsoever is known of muscles in Helcio- nellacea. Nothing suggests that they had multiple paired muscles. By including Helcionellacea in the Monoplacophora | believe that Runnegar & Pojeta equate bilateral symmetry and a presumption of non-torsion of soft parts with the concept of Monoplacophora. If this is so, the concept of Rostrochonchia has no validity and should be placed in synonymy with Monoplacophora, for these forms are presumed to have been bilaterally symmetrical and non-torted! No matter what arguments might be raised about pos- sible loss of head or specialization of the foot, if non-torsion is equated with Mono- placophora, Rostroconchia is an unneces- sary term. Members of that class are closer in hard part morphology to a cap-shaped shell than are the bellerophontaceans which Runnegar & Pojeta transferred to Mono- placophora. Of course, rostroconchs do have addi- tional features of hard parts which suggest that they exceed the morphologic limits of Monoplacophora by about the same amount as do the Helcionellacea. If Rostro- conchia can, on features of hard part mor- phology, be differentiated as a class, | think any reasonable systematist should be willing to accept the elevation of Helcio- nellacea to class level. I do not know what is the proper systematic position of Scene//a, but would be willing to guess at this time that it might have been a wide helcionellacean, rather than an early monoplacophoran. However, there is a major difference in shape between this genus and the coeval Aldabanella. There is a great deal we still do not know. In closing this section, | trust that it will be clear that | do not auto- matically equate the ancestral hypothetical mollusk with Monoplacophora. SUPERPHYLA AND SUBPHYLA Valentine (1973a) is one of a long line of writers to suggest some interrelationships among the various animal phyla. There are a few objectors, but of late the invertebrate divisions of Proterostomia and Deutero- stomia seem to have found general accep- tance (Carter, 1965) and | judge there is general agreement among those biologists who deal with phylum level phylogeny that there may be some similarities or common ancestry among the mollusks, annelids and arthropods. | do not see what is gained by introducing the level of superphylum (Valentine, 1973b) to encompass one, two, or all three of these phyla, for this gives the impression that our level of knowledge and precision of classification is greater than it really is. Classification certainly does not stop at the phytum level and | concede that discussion of a possible se- quence of hypothetical events may provide insight, but to use the category super- phylum indicates a precise position in a nested hierarchy. As it has been employed, the superphylum also suggests that changes in morphology and their sequence are known as a far firmer series of events (Valentine & Campbell, 1975) than there is evidence to support. The question of relationships within the Mollusca is not so easily dismissed. Specula- tion on grouping of classes has been a topic of discussion for more than a century. | suggest that almost any relationship which might be proposed could be supported by some reference to earlier literature. Never- theless, the classic notion commonly con- sidered two divisions within the Mollusca, 184 YOCHELSON with the Aculifera (Amphineura or Aplaco- phora, and in some schemes even the Poly- placophora) giving rise to the Conchifera (univalved and bivalved molluscan groups). There were tacit assumptions within the Conchifera; that the Cephalopoda were in some way higher than other groups; that there was some relationship between the Scaphopoda and Pelecypoda; and that the Gastropoda, because of torsion, were the most aberrant of the Mollusca. Discovery of Neopilina and acceptance of the class Monoplacophora has led to some change in these concepts; modifica- tions have been added to the traditional interpretation of the ancestral protomollusk by Salvini-Plawen (1972) and others. Re- examination of the phylum has led to one rearrangement of the classes into three sub- phyla (Stasek, 1972). By adding data on the first occurrence of some classes in the fossil record and hypothesizing about the soft parts of extinct forms, Stasek’s scheme was then arranged by Runnegar & Pojeta (1974) into: Aculifera for those without a shell; Placophora for the multi-plated chitons; and the old Conchifera broken into the subphyla Diosoma and Cyrto- soma. Obviously, this is not the same scheme as the different subphyla and some- what different classes of von Ihering (1922) or Johansson (1952) or Harry (1969) or Salvini-Plawen (1972), to cite a few of the other alternatives. Pojeta & Runnegar (1976: 44-45) are the latest authors to con- sider this topic, but probably they are not the last. These authors recognize eight classes, one of which, the Rostroconchia, is extinct; the classes are rearranged as noted above into four subphyla. It puzzles me as to what tests we might apply to determine which of the schemes of subphyla arrange- ments is a more accurate summary of the evolutionary history of the phylum. | have arranged the extant classes and proposed and presumed extinct classes, which | have discussed earlier, into subdivi- sions based on geometry of the hard parts (Fig. 1 at top). | also included speculations on the geometry of the gut, which are reasonably conservative and closely follow those of most current writers (Fig. 1 at extreme top). These divisions of hard and soft parts do not coincide. If | were to use other features of actual and presumed soft part anatomy, such as presence or absence of a radula, a different grouping could be formed. Without emphasizing this approach ad nauseam | conclude that attention to any particular character would lead to other arrangements. Those who argue for mosaic evolution suggest this phenomenon is a common occurrence with different char- acters changing at different times. Cladists argue that changes are dichotomous. Neither position is probably entirely true. Theorists do not agree and this further reinforces my pragmatic view that since most of the classes do not appear to be linked in a sequence, perhaps there is no major intra-phylum grouping of Mollusca. At least, we should recognize that to group the classes will be a more subjective exer- cise than it was thought to be half a century ago in von lhering's day. PRECAMBRIAN EVENTS, MOLLUSK LARVAE AND SPECULATION There are no undoubted Precambrian mollusks. Glaessner (1969), followed by Runnegar & Pojeta (1974), interprets one Precam- brian fossil, Bunyerichnus, known only from its trace, as the trail of а mollusk. 1 do not know what this is, but it is not that of a crawling mollusk. Perhaps it is a frond of some sort. A new picture of the only known specimen (Häntzchel, 1975: W-49) shows how unlike a mollusk trail this was. Except for a few spicules and a radula, the Aplacophora lack hard parts, and to date this class is unknown in the fossil record, although Bunyerichnus has been alluded to as a possible aplacophoran. Salvini-Plawen (1968) judged Aplacophora to be primitive. Because | have not had any experience at all with these modern worm- like forms, my comments are uninformed speculation, but my guess is that these animals are secondarily simplified rather than primitive. First, there is a general trend in many groups of mollusks toward loss of the shell (Morton, 1963). Second, many aplacophorans occur at moderate depths, and shelled deeper water mollusks appear to be mostly derived from organ- isms moving downslope rather than relics of an early fauna (Clarke, 1962: 4). Thus, | would not include aplacophoran-like mol- lusks in my speculations concerning the Precambrian. On the other hand, my notion of the aplacophorans as secondarily naked may be PHYLOGENY OF ANCIENT MOLLUSKS 185 completely false. As Dr. John Morton put it (oral communication, 1977): “The con- clusion would seem irresistible from an ensemble of primitive soft parts (gonadial, pericardial and pallial morphology, as well as central nervous system), that however specialized in present habit and body form the living Aplacophora are, they diverged from the molluscan stem at a very lowly level.’” Perhaps the ancestral mollusk gave rise to the scaphopods, and lastly gave rise to the aplacophorans. The group does not fit anyone’s model of an early mollusk with all its apparently primitive features. The notion of organisms remaining un- changed since the dawn of the Cambrian is quite hard for me to accept, and if the Aplacophora are to be primitive, | prefer that the ancestral stock persisted longer. If this many-fold speculation is true, the rela- tionship between Polyplacophora and Aplacophora need not be close, even if both do possess spicules. But, as noted earlier regarding philosophy, too much speculation can be stultifying. | should at least now change the subject of specula- tion. Similarities in the preservation of various fossil mollusks argue for a common organic integument which developed prior to great radiation and diversification of soft part anatomy. Spicules are known in the perio- stracum of some pelecypods (Carter & Aller, 1975), as well as in the girdle of Polyplacophora (Beedham & Trueman, 1967) and in the Aplacophora (Beedham & Trueman, 1968). А spicular stage might have preceded formation of a true shell as Carter & Aller suggest, with the integument providing the organic matrix to regulate mineral deposition, so that only “то|- luscan’’ shell was grown. The sequence of events could have been: first, formation of an organic cover; second, major differentia- tion; and third, calcification of the integu- ment. | would follow Fretter & Graham (1962) who, among others, derive pre- mollusks from pre-turbellarians, in part by an increase in size. Suggestions as to devel- opment of the coelomic cavity in animals (Vagvolgyi, 1967; Salvini-Plawen, 1968) also lead to the notion of a turbellarian-like ancestor. Consequences of size increase would be the need for better digestion, solved by formation of a complete gut, and the need for more oxygen, solved by gills to provide a greater area for respiration. Development of a cuticle from mucus, and thus a slight protection of the gills, might then have followed. Runnegar & Jell (1976) suggest that there was a gradual size increase in mol- lusks, that is, members of the phylum were systematically smaller further back in time. It follows that the prospects of finding Precambrian individuals is exceeding slight. However, specimens of Dickinsonia more than 20mm in length are known, so | cannot subscribe to the doctrine that all Precambrian fossils are necessarily small. The issue of whether mollusks were tiny in the Cambrian and became larger through time depends on how one picks one’s data. If Heraultipegma is a rostroconch, it is larger than many others. If hyoliths are mollusks, the largest are in the Early Cam- brian, Scenella is larger than Cyrtonella, and so on. An important point that a paleontologist should bear in mind is the size of material traditionally studied; Tevesz & McCall (1976) have noted the importance of scales when attempting to reconstruct life habits of fossils. It is exceptional in the Paleozoic to break out from a rock a mollusk smaller than 2mm. By working with residues from acid solutions, one can obtain fossils much smaller, but as | have indicated, it is diffi- cult to establish that these remains are of mollusks. As a result of physiological needs, there is a maximum and minimum size for each kind of organism. Mature living mollusks do occur in the 1mm range. They are common, but they are atypical of the phylum. In the habitats in which they are found today, preservation as fossils is unlikely. Even if some tiny organisms should prove to be early growth stages of mol- lusks, they tell us little. There is a pro- found difference between the earlier and generally smaller planktonic state and the later and generally larger benthonic stage. Surface tension is significant for small ani- mals but not so much so for larger ones. | do not know what caused torsion, but to look for a mechanical interpretation of this process by suggesting how shells were car- ried by mature benthonic animals (Run- negar & Pojeta, 1974: 314) is about as logical as suggesting that during the years of burrowing, clams slowly reduced the size of the head to the point where it disap- peared. Levels of integration, loosely related to 186 YOCHELSON size of material, are important, for not all the same rules apply to all sizes of organ- isms. | think that most dramatic changes in fundamental morphology, the character- istics that distinguish classes, occurred dur- ing larval development. It is perhaps easier to evoke in my mind’s eye major ana- tomical changes at that stage than later in development. What appears to us as a major step in evolution may have been minor to the animals. A small change could have had a profound effect. This is why Garstang emphasized the advantages of tor- sion to the larva. The ontogenetic stage at which a change first occurs is not important. What is im- portant is that the change be perpetuated. We do not see missing links between classes in the fossil record. Dr. V. Jaanas- son (oral communication, 1977) suggests that for a few generations a small popula- tion (dithyrial population of literature) could consist of several morphological forms coexisting until the atypical form became genetically fixed. However, the mechanism for evolution is not so much my concern in this study. | speculate that we do not find missing links because none existed. Such speculation runs counter to the notions of most geneticists who see a gradual accumulation of small changes as the only way organisms evolve. Unfor- tunately, features such as fusion to form an open tube of the Scaphopoda are hard to envision as a series of small steps. ‘‘Hopeful monsters” are not in vogue, but | suspect that any intermediates among classes would be far more monstrous than those ancestral animals which first showed the prime fea- tures of the class. In another vein, new ideas are being proposed as to the choice an organism which developed hard parts had between calcium carbonate and calcium phosphate (Rhodes & Bloxam, 1971). New ideas are being proposed as to the mechanism which caused calcification originally (Rhodes & Morse, 1971) and as to the reason animals with hard parts diversified dramatically at the dawn of the Cambrian (Stanley, 1973). With all this intellectual fare to digest, consideration of the changes from Precam- brian to Cambrian cannot be taken without developing some cramps; we are closer to understanding these early events, but we have a long way to go before all is clear and logical. CRITIQUE Runnegar & Pojeta (1974; see also Pojeta & Runnegar, 1976) have written extensively on molluscan phylogeny. They do not consider the early mollusks in the same light as | do. It may be unfair of me to recast and simplify the views of these workers. Tabulation is often highly arti- ficial, but | employ it here as the easiest way to summarize differences in our inter- pretations. Runnegar & Pojeta Yochelson 1. Hyolitha an ex- 1. Hyolitha an extinct class tinct phylum. of mollusks. 2. The Rostroconchia 2. At least three extinct the only extinct classes; other groups class of mollusks. which may eventually deserve that rank. 3. Stenothecoides an 3.Stenothecoides and al- asymmetrical lied genera of extinct monoplacophoran. bivalves in class Steno- thecoida. 4. Mattheviaamem- 4. Matthevia the only mem- ber of class Poly- ber of class Mattheva. placophora. 5. Pelecypoda in 5. Pelecypoda in Early Early Cambrian. Ordovician; Fordilla pos- sibly a member of an ex- tinct bivalve class. 6. Gastropoda in Early 6. Gastropoda in Late Cam- Cambrian. brian; Pelagiellacea pos- sibly in an extinct class. Tiny coiled ‘‘worm”’ tubes resemble gastro- pod shells. 7. Scaphopoda in Mid- 7. Scaphopoda in Mississip- dle Ordovician. Pian or possibly Devon- ian; curved “worm” tubes resemble scapho- pod shells. 8. Monoplacophora 8. Monoplacophora restrict- a highly diverse ed to cap-shaped shells group; many genera plus a few coiled forms; without evidence Paired muscle scars re- of muscle scars. quired for definitive placement in the class. Helcionellacea probably in a distinct class. Runnegar & Pojeta recognize three of the six shelled extant classes in the Early Cambrian, whereas | judge that all extant classes appeared later. There are other clas- sificatory differences, but these appear to be the principal ones at high taxonomic levels. There are also several methods in ap- proach which might be noted. Thus, Runnegar & Pojeta make use of series of morphologic intermediates to link classes. | see practically no intermediates between classes. My emphasis is оп stratigraphic PHYLOGENY OF ANCIENT MOLLUSKS 187 occurrences, whereas many of the taxa in their presumed morphological series occur at the same time interval. They apparently view major changes as gradualistic, whereas | consider them to have been abrupt. It is my impression, though perhaps not a cor- rect one, that Runnegar & Pojeta discrimi- nate high level taxa because of assumptions regarding soft parts, whereas | would make a high-level separation on hard part mor- phology alone, and then try to interpret soft parts. DISCUSSION AND SUMMARY There is a reaction in science against spending time on definition, for in some ways it is sterile work. Nevertheless, unless one does consider this aspect, much of the current speculation on phylogeny may be futile. Thus, until there is agreement as to what criteria are needed to place a fossil within the mollusks or remove it from the phylum, controversy regarding position of the Hyolitha will continue. If Hyolitha are mollusks, our concept of the ancestral form will have to be re-evaluated. The problem of class definition is more complex, for there are more classes than there are phyla. Systematics may be rigor- ous and totally objective, but, according to one school of thought, systematics is a qualitative art; a species is defined as any group which a competent worker designates as a species. One should not lean on an authority to do one’s thinking, yet | can- not help but feel that this unsophisticated definition which calls on practical work as opposed to theory has more to commend it than most of us are willing to admit. Perhaps a class of mollusks should be de- fined as a group discussed in the earlier sections of this essay. A class must be distinct from all other classes. For the paleontologist, this means features of hard part morphology. Members of a class have a general morphology which differs from members of other classes. To me, this characteristic morphology implies invasion of a major new ecological niche, so that life habitat enters into the defini- tion of a class. Some adaptations which characterize and indirectly define a class are so broad that they allow the organisms to live in many environments and even to move into the habitats where members of + another class more characteristically live. Other adaptations result in a more re- stricted mode of life. For example, sca- phopods live in a very narrow habitat and pelecypods live in a variety of habitats, one of which happens to impinge on that of the scaphopods; | do not mean to imply either competition or a relationship be- tween members of these two classes but only use this phraseology to state that some pelecypods also live as part of the semi-infauna. “Major,” “niche,”” and “habitat” are all difficult concepts, and obviously | cannot define any of them precisely. | view the history of the mollusks as а series of experiments in morphology to test various habitats. It does not trouble me that some fossil forms are exceedingly rare, for some experiments were good and some were not good. The extant classes are the most suc- cessful experiments among the mollusks. They are successful, if one may use that term in the Animal Kingdom, not because they are alive today, but because they have persisted longer than any of the extinct classes. If any of the extinct classes have validity, then the Hyolitha and Rostro- conchia were the most successful ones, for they persisted through the Paleozoic. Bio- mass and diversity are measures of various kinds, but existence or extinction is the most basic measure in evolution. The term class, as | have employed it, allows a great deal of liberty for the paleontologist. However, liberty is not the same as license. In my opinion, repeated over and over, hard parts should be studied first and then hypothesis about soft parts added. | would also hold that it is far better to state that one has insufficient evidence and to indicate the higher sys- tematic position of an organism as /ncertae sedis than to force it into an accepted and clearly defined class. To state another truism, no one is fully objective. It has been written that “the eye beholds what the mind perceives.” In my mind one of the key features of evolution is adaptive radiation. | view this as essenti- ally instantaneous exploitation of a new feature, be it morphologic, physiologic, be- havioral, or otherwise novel. | further think that diversification based on exploitation and refinement of this new feature is a process which occurs at all systematic levels (Hotton, 1971) and which occurs exceed- 188 YOCHELSON ingly rapidly, once the feature has first appeared. There is no proof as to either the sta- bility of molluscan classes suggested by Runnegar € Pojeta or the more experi- mental aspect in the phylogeny of the Paleozoic mollusks that | postulated, but | suggest that ancillary data may be in- structive in supporting the latter view. In this regard, one of the most dramatic changes in class level systematics during the last two decades is the establishment and general acceptance of a large number of extinct classes of Echinodermata, most of which are found early in the fossil record (for a discussion of some of these see Sprinkle, 1976). In addition, there is emerging from current literature a sugges- tion that the conventional high level tax- onomy among the Arthropoda is under- going a similar revolution. These two phyla are not related, yet workers independently arrive at the notion of numerous extinct groups of major rank. Runnegar & Pojeta (1974) develop a simple, linear pattern for the evolution of mollusks. My pattern is more like an unpruned bush, but it does resemble the pattern of evolution now postulated for other phyla, whereas the linear pattern does not. To some people the notion of evolution carries with it the concept of perfect adap- tation to the environment. The increase in number of lower level categories through geologic time seems to be generally ас- cepted, and one might argue that a corol- lary is the relatively greater number of higher level taxa the farther one goes back- ward. Such philosophical remarks carry lit- tle real weight and elaborating them in an attempt to justify my systematic approach would be simply preaching to the con- verted; this might gain a small, though enthusiastic, response, but it would accom- plish little of substance. However, in spite of my scoffing there may be some major theoretical support coming my direction. “Quantum” evolution is based on the notion that rates of change are not constant. The doctrine of punctu- ated equilibrium carries with it the accept- ance of small, geologically early classes (Gould & Eldredge, 1977) and also the acceptance of non-evolution once a group is securely established in a restricted and stable habitat. To present the other side of the coin, there are at least three main areas of weakness in my arguments. First, | cannot come up with an unambiguous definition of every extinct class level taxon based on hard part morphology and must call on future work to supply this. | am even willing to suggest that with the best will in the world and the hardest work, we may be puzzled for generations as to the proper class-level placement of a few taxa. Second, with few exceptions | do not see a series of intermediates from one class to another, but insist on dramatic and rapid changes. Hopeful monsters are not in fashion and for those who view evolution as a con- tinuum, the absence of missing links is damning. At the risk of being accused of making a bad pun on systematics, | do think that speciation is trivial, and is not important in the long term history of a group. Third, calling on a non-calcified mollusk which might have persisted from late Precambrian to mid-Paleozoic as the source material for new classes is a piece of special pleading. My only defense is to argue that one might obtain greater varia- tion from the larval stages of an archaic type increasingly under the stress of com- petition than from larvae of more stabilized lines. | also must acknowledge that changes such as torsion could not have taken place if the mollusk upon which this change was about to be imposed did not already have a shell. In the face of these uncertainties, | hold that there are a number of classses of extinct mollusks. The variety to be seen within the early mollusks has been sam- pled, but it has not been fully explored or adequately documented. There may not be as many classes of extinct mollusks as there are of extinct echinoderms, but | would wager that posterity will recognize as many classes of extinct mollusks as there are extant classes. However, no one knows without the shadow of a doubt. Though | have used it before (Yochelson, 1977), | consider any paper of a speculative nature an appropriate place to quote the motto of my mentor, the late J. Brookes Knight: “Say not that this is so, but that this is how it seems to me to be as | now see the things | think | see.” LITERATURE CITED AKSERINA, N. A., 1968, Probivalvia—Novyy klass drevneyshikh mollyuskov [Probivalvia—A new class of ancient mollusks] (in Russian). Novye dannye ро geologii I poleznum PHYLOGENY OF ANCIENT MOLLUSKS 189 iskopaemum Zapadnov Sibiri. [New data about the ecology and useful leading fossils from West Siberia.] Zapadnosibirskoe Geo- logicheskoe Upravlenie, Novosibirskoe Geo- logicheskoe Upravlenie, 3: 77-86. BATTEN, R. L., 1960, The need to classify. Annual Report of the Smithsonian Institution, 1959: 509-522. BEEDHAM, G. E. & TRUEMAN, E. R., 1967, The relationship of the mantle and shell of the Polyplacophora in comparison with that of other Mollusca. Journal of Zoology, 141: 215-230. BEEDHAM, G. E. & TRUEMAN, E. R., 1968, The cuticle of the Aplacophora and its evolu- tionary significance in the Mollusca. Journal of Zoology, 154: 443-451. BERGENHAYN, J. R. M., 1960, Cambrian and Ordovician loricates from North America. Journal of Paleontology, 34: 168-178. BOGKERIE. STE ©: 180 МОСНЕ ОМ Es ein press, “Worm tubes” from the Valhallfonna Formation, Spitsbergen. Norsk Polarinstitut Arbok. BOSS, K. J., 1971, Critical estimate of the num- ber of Recent Mollusca. Occasiona/ Papers on Mollusks, Museum of Comparative Zoology, Harvard University, 3: 81-135. BRETSKY, Р. W. & BERMINGHAM, J. J., 1970, Ecology of the Paleozoic scaphopod genus Plagioglypta with special reference to the Or- dovician of eastern lowa. Journal of Paleon- tology, 44: 908-924. CARTER, G. S., 1965, Phylogenetic relations of the major groups of animals. In MOORE, J. A. (Ed.), Ideas in Modern Biology, Proceed- ings of the 16th International Congress of Zoology, Washington, D.C., August 23-27, 1963, 6: 427-448. CARTER, J. G. & ALLER, R. C., 1975, Calcifi- cation in the bivalve periostracum. Lethaia, 8: 315-320. CLARKE, A. H., 1962, Annotated list and bibli- ography of the abyssal marine molluscs of the world. Bulletin of the National Museum of Canada, 181: 114 p. DONOVAN, D. T., 1977, Evolution of the dibranchiate Cephalopoda. Zoological Society of London Symposium, 38: 15-48. EDHORN, A. S., 1977, Early Cambrian algal croppers. Canadian Journal of Earth Science, 14: 1014-1020. FRETTER, V. & GRAHAM, A., 1962, British Prosobranch Mollusca, their tunctional anatomy and ecology. Ray Society, London, 548 р. GARRETT, P., 1970, Phanerozoic stromatolites: noncompetitive ecologic restriction by grazing and burrowing animals. Science, 169: 171-173. GARSTANG, W., 1951, Larval forms and other Zoological verse. Saville Press, Blackwell, Oxford, 85 p. GLAESSNER, M. F., 1969, Trace fossils from the Precambrian and basal Cambrian. Lethaia, 2: 369-393. GOULD. $. J. & ELDREDGE, N., 1977, Punc- tuated equilibria: the tempo and mode of evolution reconsidered. Paleobiology, 3: 115-131. HAAS, W., 1972, Micro- and ultrastructure of Recent and fossil Scaphopoda. Proceedings of the 24th International Geological Congress, section 4: 15-19. HANTZCHEL, W., 1975, Trace fossils and prob- lematica. In: MOORE, R. C. (Ed.), Treatise on Invertebrate Paleontology, W, Miscellanea: supplement 1. University of Kansas Press, 269 p. HARRY, H. W., 1968, An alternate view on the phylogeny of the Mollusca. Marine Biological Association of India Proceedings Symposium, 3 (Mollusca) (1): 170-187. HORNY, R., 1961, New genera of Bohemian Monoplacophora and patellid Gastropoda. Czechoslovakia, Vestnik Ustredniho Ustavu Geologichého, 34: 299-302. HORNY, R. 1963, Archaeopraga, a new prob- lematic genus of monoplacophoran molluscs from the Silurian of Bohemia. Journal of Paleontology, 37: 1071-1073. HOTTON, N., Ill, 1971, Origins of vertebrate classes. North American Paleontological Con- vention, Chicago, 1969, Proceedings, H: 1146-1152. HYMAN, L. H., 1967, The Invertebrates, vol. 6, Mollusca 1. McGraw-Hill, New York, vii + 792 p. IHERING, H. VON, 1922, Phylogenie und Sys- tem der Mollusken. Archiv für Mollusken- kunde, 1: 1-115. JOHANSSON, J., 1952. On the phylogeny of the Mollusca. Zoologiska Bidrag Fran Uppsala, 29: 277-292. KAY, E. A., 1968, A review of the bivalved Gastropoda and discussion of evolution within the Sacoglossa. Zoological Society of London Symposium, 22: 109-134. KEEN, A. M. & SMITH, A. G., 1961, West American species of the bivalved gastropod genus Berthelinia. Proceedings of the Cali- fornia Academy of Science, 30: 47-66. KNIGHT, J. B., 1947, Bellerophont muscle scars. Journal of Paleontology, 21: 264-267. KNIGHT, J. B., 1952, Primitive fossil gastropods and their bearing on gastropod classification. Smithsonian Miscellaneous Collections, 114(13): 55 p. KNIGHT, J. B. & YOCHELSON, E. L., 1960, Monoplacophora. In MOORE, R. C. (Ed.), Treatise on Invertebrate Paleontology, |, Mol- lusca 1: 1-77-1-84. University of Kansas Press and Geological Society of America. KOLLMANN, H. А. & YOCHELSON, Е. L., 1976, Survey of Paleozoic gastropods possibly belonging to the subclass Opisthobranchia. Annalen der Naturhistorische Museum Wien, 80: 207-220. LEMCHE, H., 1957, A new living deep-sea mol- lusc of the Cambro-Devonian class Mono- placophora. Nature, 170: 413-416. LINSLEY, R.M., 1978, Locomotion rates and shell form in Gastropoda. Malacologıa, 17: 193-206. LINSLEY, R. M., in press. Shell form and the origin of gastropods. American Scientist. MAREK, L. & YOCHELSON, Е. L., 1976, As- pects of the biology of Hyolitha (Mollusca). Lethaia, 9: 65-82. MERTON, R. K., 1961, Singletons and multiples in scientific discovery: a chapter in the sociol- ogy of science. Proceedings of the American Philosophical Society, 105: 470-486. MISSARZHEVSKIY, V. V 1974, Novye dannye о drevneishihk okamenelostyakh rannego kembriya Sibirskoï platformy [New data on ancient fossils from the Early Cambrian of the Siberian Platform]. (in Russian). In Biostrati- 190 YOCHELSON grafiya i paleontologiya nizhnego kembriya Evropy i severnoi Azii. Akademiya nauk SSSR Geologicheskii Institut, Sibirskoe otdelenie, Institut geologii i geofiziki: 179-189. MISSARZHEVSKIY, V. V., 1976, New data on early Cambrian monoplacophors. Paleontologi- cal Journal (translation from Paleon- tologicheskir Zhurnal by American Geological Institute), 1976, 10(2): 234-236. MORTON, J., 1963, The molluscan pattern: evo- lutionary trends in a modern classification. Proceedings of the Linnean Society of Lon- „don, 174(1):53-72. MULLER, К. J., 1975, “Heraultia varensalensis” Cobbold (Crustacea) aus dem Unteren Kam- brium, der älteste Fall von Geschlechtsdi- morphismus. Paläontologische Zeitschrift, 49: 168-170. MUTVEI, H., 1964, Remarks on the anatomy of Recent and fossil Cephalopoda, with descrip- tion of the minute shell structure of belem- noids. Stockholm University Stockholm Con- tributions in Geology, 11(4): 79-102. NEVESSKAYA, |. A., SCARLATO, O. A. STAROBOGATOV, YA. |. € EBERZIA, A. C., 1971, New ideas on bivalve systematics. Paleontological Journal (translation from Paleontologicheskii Zhurnal by American Geo- logical Institute), 1971, 5(2): 141-155. POJETA, J., Jr., 1971, Review of Ordovician pelecypods. U.S. Geological Survey Profes- sional Paper, 695: 46 p. POJETA, J., Jr., 1975, Fordilla troyensis Bar- rande and early pelecypod phylogeny. Bul- letins of American Paleontology, 67: 363- 382. POJETA, J., Jr. & RUNNEGAR, B., 1974, Fordilla troyensis and the early history of the pelecypod mollusks. American Scientist, 62(6): 706-711. POJETA, J., Jr. & RUNNEGAR, B., 1976. The paleontology of rostroconch mollusks and the early history of the phylum Mollusca. U.S, Geological Survey Professional Paper, 968: 88 p. MS POJETA, J., Jr., RUNNEGAR, В. €: KRIZ, J., 1973, Fordilla troyensis Barrande: the oldest known pelecypod. Science, 180: 66-68. POJETA, J., Jr., RUNNEGAR, B.. MORRIS, N. J. & NEWELL, N. D., 1972, Rostroconchia: A new class of bivalve mollusks. Science, 144: 264-267. RHODES, D. C. & MORSE, J. W., 1971, Evolu- tionary and ecologic significance of oxygen- deficient marine basins. Lethaia, 4: 413428. RHODES IE MH BLOXAMamTaW. 21971, Phosphatic organisms in the Paleozoic and their evolutionary significance. North Ameri- can Paleontological Convention, Chicago, 1969, Proceedings К: 1485-1513. ROBISON, R. A., 1964, Late Middle Cambrian faunas from western Utah. Journa/ of Paleon- tology, 38: 510-566. ROSOV, S. N., 1975, A new order of Mono- placophora. Paleontological Journal (transla- tion from Paleontologicheskii Zhurnal by American Geological Institute), 1975 9(1): 39-43. RUNNEGAR, В. & JELL, P. A., 1976, Australian Middle Cambrian molluscs. Alcheringa, 1: 109-138. RUNNEGAR, B. & POJETA, J., Jr., 1974, Mol- luscan phylogeny: the paleontological view- point. Science, 186: 311-317. RUNNEGAR, В., POJETA, J. Jr., MORRIS, N. №, shAYEOR aa. BD aGAY COR MESES MCCLUNG, G., 1975, Biology of the Hyolitha. Lethaia, 8: 181-191. SALVINI-PLAWEN, L. v., 1968, Die ‘Funktions- Coelomtheorie’ in der Evolution der Mol- lusken. Systematic Zoology, 17: 192-208. SALVINI-PLAWEN, L. v., 1972, Zur Morpho- logie und Phylogenie der Mollusken: Die Beziehungen der Caudofoveata und der Solenogastres als Aculifera, als Mollusca und als Spiralia. Zeitschrift für wissenschaftliche Zoologie, Abt. A, 184: 205-394. SHIMANSKIY, V. N., 1963, The taxonomic posi- tion and content of Xenoconchia. Ра/еопто- logical Journal (translation from Paleonto- logicheskii Zhurnal, by American Geological Institute), 1963(4): 43-63. SHIMANSKIY, V. N. & BARSKOW, 1. S., 1970, New data on the order Toxeumorphida. Paleontological Journal (translation from Paleontologicheskir Zhurnal by American Geo- logical Institute), 1970, 4(3): 430-434. SMITH, A. G. & TOOMEY, D. F., 1964, Chitons from the Kindblade Formation. Oklahoma Geological Survey Circular, 66: 41 p. SPRINKLE, J., 1976, Classification and phy- logeny of “‘pelmatozoan’’ echinoderms. Sys- tematic Zoology, 25: 83-91. STANLEY, S. M., 1968, Post-Paleozoic adaptive radiation of infaunal bivalve mollusks—a con- sequence of mantle fusion and siphon forma- tion. Journal of Paleontology, 42: 214-229. STANLEY, S. M., 1973, An ecological theory for the sudden origin of multicellular life in the Late Precambrian. Proceedings of the National Academy of Sciences, 70: 1486-1489. STAROBOGATOV, YU. A., 1970, Systematics of early Paleozoic Monoplacophora. Paleonto- logical Journal (translation from Paleon- tologicheskii Zhurnal by American Geological Institute), 1970, 4(3): 293-302. STAROBOGATOV, YU. A., 1974, Xenoconchias and their bearing on the phylogeny and sys- tematics of some molluscan classes. Pa/eonto- logical Journal (translation from Paleon- tologicheskir Zhurnal by American Geological Institute), 1974, 8(1): 1-13. STASEK, C. R., 1972, The molluscan framework: In FEORKIN; М: & SHEERS BIST RES" Chemical Zoology, vol. 7, p. 1-44. Academic Press, New York and London. TEICHERT, C., 1967, Major features in cepha- lopod evolution. In TEICHERT, C. & YOCHELSON, E. L. (Eds.), Essays in Paleon- tology and stratigraphy. Department of Geology, University of Kansas, Special Publi- cation, 2: 162-210. MEVESZ> М: ($ Me МЕСАШЕ ТРО 6, Primitive life habits and adaptive significance of the pelecypod form. Paleobiology, 2: 183-190. VAGVOLGYI, J., 1967, On the origin of mol- luscs, the coelom, and the coelomic segmenta- tion. Systematic Zoology, 16: 143-168. VALENTINE, J. W., 1973a, Coelomate super- phyla. Systematic Zoology, 22: 97-102. VALENTINE, J. W., 1973b, Evolutionary paleo- ecology of the marine biosphere. Prentice Hall, Englewood Cliffs, New Jersey. PHYLOGENY ОЕ ANCIENT MOLLUSKS 191 VALENTINE, J. W. & CAMPBELL, С. A., 1975, Genetic regulation and the fossil record. American Scientist, 63: 673-680. WENZ, W., 1940, Ursprung und frühe Stammes- geschichte der Gastropoden. Archiv für Mol- luskenkunde, 72: 1-10. YOCHELSON, E. L., 1958, Some Lower Ordo- vician monoplacophoran mollusks from Mis- souri. Journal of the Washington Academy of Sciences, 48: 8-14. YOCHELSON, Е. L., 1961, Notes on the class Coniconchia. Journal of Paleontology, 35: 162-167. YOCHELSON, E. L., 1963, Problems of the early history of the Mollusca. Proceedings of the 16th International Congress of Zoology, Wash- ington, D.C. August 20-26, 1963, 2: 187. YOCHELSON, E. L., 1966, Mattheva, a proposed new class of mollusks. U.S. Geological Survey Professional Paper 532-B: B-1-B-11. YOCHELSON, Е. L., 1967, Quo vadis, Ве/- lerophon: in TEICHERT, C. & YOCHELSON, E. L. (Eds.), Essays in paleontology and strati- graphy. Department of Geology, University of Kansas, Special Publication, 2: 141-161. YOCHELSON, E. L., 1968a, On the nature of Polylopia. U.S. Geological Survey Professional Paper, 593-F: F-1-F-7. YOCHELSON, E. L., 1968b, Stenothecoida, a Proposed new class of Cambrian Mollusca. International Paleontological Union, Prague, Czechoslovakia, 1968: 34. YOCHELSON, E. L., 1969, Stenothecoida, a pro- posed new class of Cambrian Mollusca Lethaia, 2: 49-62. YOCHELSON, E. L., 1971a, Phylum and class nomenclature in systematics. Systematic Zoology, 20: 245-249. YOCHELSON, E. L., 1971b, The little known Pennsylvanian Cl/avulites reinterpreted as a “worm.” Journal of Paleontology, 45: 126-129. YOCHELSON, E. L., 1975, Discussion of Early Cambrian “mollusks.” Journal of the Geologi- cal Society of London, 131(6): 661-662. YOCHELSON, E. L., 1977, Agmata, a proposed extinct phylum of Cambrian age. Journal of Paleontology, 51: 437-454. YOCHELSON, Е. 1. FLOWER, В. H. € WEBERS, G. F., 1973, The bearing of the new Late Cambrian genus Knightoconus (Mol- lusca: Monoplacophora) upon the origin of the Cephalopoda. Letha/a, 6: 275-310. YOCHELSON, Е. L., HENNINGSMOEN, С. €: GRIFFIN, W. L., 1977, The Early Cambrian genus Vo/borthella in southern Norway. Norsk Geologisk Tidsskrift, 57: 133-151. YONGE, C. M., 1953, The monomyarian condi- tion in the Lamellibranchia. Transactions of the Royal Society of Edinburgh, 62(2): 443-478. August 20-26, Abstracts, ЕТ tet TA ик ‘ = o: 1 1, ‘ des ‘re un (alucypeds. U: Gealopyies Sun AP PE re мы à moral Paper, 605: Ak pe » STANLEY РОЗЕТА, dd. М: 1975. Fardilie rropeneis Gar rachwtion of infaune Аб РАД and cay Wendt phylcgeny. Hu soyuance -of manto fuir wad ‘eles +” Ameiva Paletrastagy 6 653. or. Jours el. Paranoia pi STANLEY, & М. 1973, Anton DEIN 4. Æ PUNNEGAR, 3 via tes audi ara E u Fol а к au i tartes VANO sf Mr Lala whine. р TOC Cl aps ‘of mM Py Moura A Si ei carey af Ésisruer, 107 MOG hr NTE TOA STARTEISATOV/ YU A 19705 у POETA. 1, à. & ОМ GIE. Ги pars “viecter Monitor Na qye - ee coriiciogy 44 «q ne © tua? a h mo rs ща. OS de Е ot. Kia О Mala | смо & Гагра! by вата bios erre Priest ‘om dis ие !, 1470, 83 266-302 Bic STAROBDGATOY, YUL A, 1974 РОТА. dí +. RUNNEGAI a. A ЦЕ oi Wir Bo: ir sw oe ory NI. и! vive ures BER PEA y of ave relación uh: ec res à Wun uO Ho . obj a AREA ww ron trom’ PAT 7 A 3 $e MEL re > y E € a geet A Prörel dy Ania NN) LIN [ #72 at! AG OMNETr A FOOT: 00D 774.804 13 A пены chute LITER о тт a AL AGE? Cs 4, 1871, Pre По a f FLOWKIN, © & SHE À Hinz Y > LA] M a 73. kv Ne ' COQ! vol. 7, Я 1445 ire fen; dee vif pal 1e ‘ ; London, o tt dt р sR ae 967 Ай» Града, DES bi h АВ . in] р ews ох Y TER à eo ung к N С и J 1 к nel # к ena) Essays: в» zer UL = , у shy al PIDE BONA, -ф ‘+ LAS 2097 MESA Ina ns 777208 sf K onsen NÉ 5 pa 3247 _t Je PM `.. TA © ме Сеть ol в € MCGALL, М f . lo Ma чо IN pa? A si ‘Law é- whee у р dat Vi à > a ‘. DUO fi ri ier ory Ga Pieri. Pi) iw +. 4 Е u Pe he Г CN USCIS is ia J ves saat СУПА 3 547, Ом tho- af has Past Lis Dei A 4 y JN E A fia cn, nd e conbermie VON ls ge I META. Pa 1 Arratia LON Hi: ET hee bl LEJTT ME J уч 107 at. Coslornáta UNE A у TU oe trie Grammar Eoology, 22: 277108 Mai Qe + da aran, \ ANT INE ad, AIG PUÉS: MAIS ed ero) и mario oh : ЕЛА (478, M "Erro CA ten dee MALACOLOGIA, 1978, 17(2): 193-206 ‘LOCOMOTION RATES AND SHELL FORM IN THE GASTROPODA Robert M. Linsley Department of Geology, Colgate University, Hamilton, N.Y. 13346, U.S.A. ABSTRACT Rate of locomotion in the Gastropoda correlates surprisingly well with the shape of the shell. In general, the fastest snails hold their shells in such a way that the frontal cross-section area is very small, the center of gravity and pressure point are very low, there is little or no torque, and essentially no surface ornamentation. If any of these factors is increased, the net result is a more slowly moving snail. The slowest snails maximize at least one of the parameters (i.e., extremes in ornamentation, very high center of gravity, etc.) and commonly achieve high values in more than one (i.e., snails with high torque also have a high center of gravity and a large frontal cross-sectional area). The very slowest gastropods are the “shell-draggers,'” those who drag their shell across the substrate rather than balancing it over the cephalopedal mass. Unlike most other organisms, the size of the snail does not seem to be a major determinant of speed. Possibly this is because “stride”” is of less significance in the locomotion of gastropods. It is presumed that the stromboids would exhibit considerable correlation between size and speed. Ciliary movers show the least correlation between size and speed. Snails typically move at such a sedate pace that there has been no reason to suspect that streamlining would be of any significance to them. Yet a qualitative anal- ysis suggests that the various parameters that reduce shell drag correlate well with locomotion rates in gastropods under labo- ratory conditions. Factors thought to be significant include total frontal cross- sectional area, height and relative positions of pressure point and center of gravity, surface smoothness and kind of symmetry. GEOMETRICAL CONSIDERATIONS Perhaps the problems of motion can best be appreciated if we imagine a prepos- terous snail carrying an elongate straight shell erect over its dorsal surface (Fig. 1). This shell obviously presents a large cross- sectional area, relative to the volume of the shell, for the fluid medium to act against if the animal tries to move. The center of gravity of the shell is located high above the substrate, and as the organism acceler- ates there will be a coupling action be- tween the shell and the locomotor organ (the sole of the foot). The pressure point will be located even higher than the center of gravity and equilibrium would be at- tained only when the pressure point is behind the center of gravity during motion. A great deal of energy would have to be expended to prevent this shell from tipping over during locomotion. Obviously any coiling will help. Both the center of gravity and pressure point will be lowered and the frontal cross- sectional area will be reduced. These results will enable the shell to be carried more readily through water. However, not all — — o > PRESSURE POINT CENTER OF GRAVITY LOCOMOTION Fig. 1. A preposterous gastropod with an un- coiled shell held erect dorsally. The center of gravity is high, the pressure point even higher. The large frontal cross-sectional area combines with these factors to make forward motion very difficult. (193) 194 LINSLEY modes of coiling are equal in this regard. The effectiveness of any particular coiling is primarily determined by the way a snail positions its shell while moving (Linsley, 1977). A shell held by a moving snail must be balanced. A limiting factor is that the aperture must be tangential, i.e. in a plane tangential to the ventral portion of the body whorl (Linsley, 1977) and held ap- proximately parallel to the substrate so the shell can be clamped against the substrate. A bilaterally symmetrical isostrophic shell (with zero translation, Raup, 1966) will of course be balanced, with the plane of symmetry of the shell continuing the plane of symmetry of the foot. It would be balanced whether the animal was untorted (monoplacophoran) or torted (gastropod). The cross-sectional area can be minimized either by making a very compressed shell or by large whorl overlap. However, a large degree of whorl overlap (a doubly anom- phalous shell) makes it geometrically diffi- cult to maintain a tangential aperture. In an asymmetrical (anisostrophic) shell, balance can be obtained by tilting the axis of coiling up (inclination) until the center of gravity of the shell and its contents is over the midline of the foot, or balance can be obtained by rotation of the axis of coiling (regulatory detorsion). Regulatory detorsion will swing the spire forward in hyperstrophic gastropods or backwards in orthostrophic forms. In fact every snail utilizes a combination of these two proc- esses to obtain a balanced shell, and the combination used is that which provides the lowest center of gravity while maintain- ing a tangential aperture (Fig. 2). However, as Vermeij (1971) has pointed out, the angle that the plane of the aperture makes with the axis of coiling (angle E) is depend- ent on the shape of the aperture. As the aperture becomes longer, angle E decreases. Since angle E is essentially a measure of the amount of inclination, a snail with an elongate aperture balances its shell pri- marily by regulatory detorsion (approach- ing 90°) and little inclination. The relative amounts of detorsion and inclination pro- duce different shell shapes. To simplify the discussion | begin with an analysis of shells whose generating Curves approximate a circle. The major difference among the two forms shown in Fig.3 is the amount of translation (T), although there are attendant changes т whorl expansion rates (W) and distance (D) isostrophic orthostrophic hyperstrophic Fig. 2. Shell balancing in Gastropoda. Isostrophic gastropods (a) are assumed to be the ancestral condition, torsion having brought the anus to a position over the head. In right-handed ortho- strophic shells (b) the spire projects to the ani- mal’s right side, while in hyperstrophic shells (с) the spire projects to the animal’s left side. Regu- latory detorsion swings the spire backwards in orthostrophic shells (d) but brings it forward in hyperstrophic forms (e). Inclination (f, g) swings the spire up so that the shell is now in a balanced position over the back of the snail. Figures h and i show the shell with tangentially constructed apertures, while all previous diagrams are drawn with radial apertures. from the axis of coiling. In general, halio- tiform, turbiniform and turritelliform shells are distinguished by a progressive increase in T accompanied by a decrease in W and D. Haliotiform shells approximate discoidal (planorbiform) shells in values of T, but discoidal shells have much lower values of LOCOMOTION IN GASTROPODA 195 A B Fig. 3. (A) Longitudinal section of a turbiniform shell. (В) Longitudinal section of a turritelliform shell, each with a circular generating curve. The whorl expansion rate and distance from the axis of coiling are highest in (A) while translation is greatest in (B). W. In a snail with a discoidal shell, such as Planorbis, the shell is held so that the axis of coiling is essentially parallel to the sub- strate and detorsion is O°. In contrast, a snail positions a haliotiform shell so that the axis of coiling is almost at right angles to the substrate; regulatory detorsion is still slight (10-20°). Vermeij (1975) suggested that the relative absence of discoidal shells since the Mesozoic is partially due to the appearance of crabs at that time. The dis- coidal shell is poorly adapted to resist the crushing pressure of crab chelae. While this may be a contributing cause | think that a discoidal shell’s high center of gravity and high pressure point suggest that it is poorly adapted to compete in modern oceans. The discoidal form is very common in fresh- water pulmonates because the pulmonary cavity acts as a buoyancy chamber and completely changes the analysis of forces acting during locomotion (Linsley, 1977). Thus | suggest that haliotiform shells are the equivalent of discoidal forms among marine prosobranchs. To reduce shell drag they have acquired the lowest center of gravity by positioning the shell primarily by inclination, achieving a proportionately small frontal cross-sectional area, and a pressure point positioned in line with and behind the low center of gravity. In order to construct a tangential aperture the W and D values have to be quite large, unlike pulmonate discoidal shells. Although | have not yet had the opportunity to study halio- tiform snails, | expect them to be very fast. In turbiniform shells with circular aper- tures the shell is balanced with an inclina- tion of 50-70° and 10-30° regulatory detor- sion (Vermeij, 1971). This results in a shell with a large frontal cross-sectional area, a high center of gravity and a high pressure point. Consequently snails with turbiniform shells tend to be fairly sedate in their rates of locomotion. However, because of the position of their retractor muscle (Linsley, in press a), they are well suited for rock- clinging and this seems to be a satisfactory trade-off for their speed limitations. It should be noted that while naticids hold their shells in a similar fashion, they “cheat” by inflating the foot with water to internalize their shell. As a result, an analy- sis of forces affecting their motion capabili- ties predicts them to be among the fastest snails, as indeed they are. As spire height increases through ап increase in T, with attendant decreases in W and D, it is possible for angle E to decrease as well. As a result higher-spired shells tend to achieve a lower center of gravity by employing more regulatory de- torsion and less inclination. In general this reduces frontal cross-sectional area, lower- ing the center of gravity and the pressure point, allowing more rapid locomotion. However, as the spire becomes still higher, the center of gravity moves farther from the aperture and the columellar muscle is flexed at an increasingly larger angle be- tween its insertion points in the foot and the shell. Thus as the spire elongates, the snail must work harder to hold the shell up and the muscle becomes positioned in a progressively less efficient attitude. As a result the turritelliform snail does not hold its shell above the foot, but drags the shell along the substrate. “Shell draggers” are among the slowest snails. Thus there is no way for a gastropod with a circular aperture to solve the geo- metrical problems in a way that allows it to be a “fast” snail. The only solution is an aperture elongated parallel to the direction of motion. And yet an aperture elongated in this manner is possible only in snails with a single gill. The primitive archaeo- gastropods with paired gills must maintain a mantle cavity (and aperture) sufficiently commodious for two separate currents through the cavity. Elongation of the aperture more or less parallel to the axis of coiling reduces incli- nation while retaining a tangential aperture and necessitates that shell balancing be 196 LINSLEY accomplished by regulatory detorsion. Any reduction of D simultaneously reduces frontal cross-sectional area, lowers the cen- ter of gravity, lowers the pressure point, and in general increases streamlining of the shell. These are all characteristics of the fastest snails. The preceding analysis seems to be valid for all gastropods supporting their shells above the substrate while moving. However, there are gastropods that are atypical in this regard (Linsley, Yochelson & Rohr, in press). In general they can be divided into three major categories with a number of subdivisions. These artificial groups can be characterized as the ‘‘motionless’’ forms, the “shell draggers’’ and the “‘leapers.”’ The motionless forms never move while adult. They are easily recognized by the presence of an attachment scar (e.g. Peta/o- conchus Lea) where the shell was cemented to the substrate, or by open or disjunct coiling (e.g. Si/iquaria Bruguiere, and Ver- micularia Lamarck) (Gould, 1969; Yochel- son, 1971; Peel, 1975; Rex & Boss, 1976). The shell draggers include a motley as- sortment of snails that rest their shells upon the substrate and move by arhythmic motion (Miller, 1974) wherein they extend the foot in front of the shell and then drag the shell up to it. These include high-spired snails (e.g. Cerithium Bruguiere) whose cen- ter of gravity is placed so far behind the aperture that to support the shell over the body would require undue energy. They also include snails whose aperture is off to the side of the shell (e.g. Conus Linnaeus). (The placement of the columellar muscle in Mirror at 45° Aquarium Stand this group is better situated for lifting the shell than in the high-spired forms, so although these animals typically rest their body whorl on the substrate they hoist it up more frequently than do the high-spired forms.) The final group of shell draggers are those shells with radial apertures (aper- tures whose plane includes the axis of coiling) (e.g. Architectonica Roding). The radial aperture (Linsley, 1977) makes it impossible for these animals to obtain pro- tection by clamping the shell against the substrate. Shells with radial apertures are rare in the modern gastropod fauna. The last group of atypical snails, the leapers, include genera like Strombus Lin- naeus, Aporrhais da Costa and Xenophora Philippi. These animals spend most of their time with the shell resting on the substrate; motion is accomplished by the foot thrust- ing against the substrate, which lifts the shell up and forward in the characteristic “leap” (Berg, 1972). Unlike the other two groups the leapers at times are active and can cover considerable distances in a short time. However, since their locomotion is so unlike that of the majority of gastropods, | have not included them in the analysis of locomotion. METHODOLOGY Time-motion study of gastropods was done with time-lapse movies using a Mi- nolta super 8 mm movie camera. An aquari- um was mounted on a stand with a mirror placed below it at a 45° angle. The camera was situated to the side of the aquarium Measured Grids Aquarium Fig. 4. Camera setup for taking motion picture studies of gastropods. LOCOMOTION IN GASTROPODA 197 and the picture was framed so that it included both the mirror and side of the aquarium (Fig. 4). This allowed motion of the snails to be recorded in all three dimen- sions; measured grids were placed along the top and back of the aquarium. Typically five individuals of one species were placed in the aquarium and their movements were filmed during a two-hour period. Fre- quently individuals would move actively for some time and then become quiescent. Moving them about the aquarium fre- quently initiated activity again. | made no attempt to elicit escape behaviors in the snails by introducing potential predators; | thus presume that | am measuring typical locomotion. Some snails vary widely in speed, but the majority are fairly uniform in their speeds. A few browsers have two speeds. One is fairly fast and the animal progresses in a fairly straight line. This | characterize as a “going somewhere” loco- motion. The other behavior consists of moving more sedately while the head swings back and forth and the snail pro- gresses more erratically, not in a straight line. This was a pattern found primarily in algae-grazers and | assume it is “feeding” locomotion. The most obvious objections to this approach are that the snails had light com- ing at them from all directions, even from below, and they were placed on an artifi- cial substrate. That they moved at all sug- gests that light below is no deterrent. Ob- servations on a few gastropods in their normal habitats revealed speeds consistent with those observed in the aquarium. It is my supposition that the nervous system of the Gastropoda is simple enough so that only a narrow range of behavioral responses are possible. In compiling data from the movies, | recorded three kinds of information: (1) fastest recorded speed, (2) average speed, and (3) total motion. The speed was timed from as many straight line “runs” of 2-30 cm as | could obtain on film. ‘’Fastest speed” was the fastest recorded time of all runs. “Average speed’’ was based on 2-20 runs. “Total motion” was based on total distance covered by each snail (whether it moved or not) and reduced to the average distance for all members of the species in cm/hr. All three of these measurements show positive correlation with shell shape; "average speed’’ provides the best correla- tion and “total motion’’ the poorest. The first movies were made in the sum- mer of 1975 in the Florida Keys and Sanibel Island. In January 1976 the author and eight students from Colgate University spent one month at the Smithsonian Insti- tution’s Marine Biology Station at Carrie Bow Key in Belize, thanks to arrangements made by Dr. Klaus Ruetzler. Analysis of resistance was made subjec- tively to determine whether a more sophis- ticated approach seemed warranted. Shells were evaluated for two characteristics: or- namentation and bilateral symmetry. The amount of ornamentation of the shell was divided into three grades: (1) smooth, (2) raised growth lines or incised sutures and (3) nodes or spines. Bilateral symmetry was used because it correlates with frontal cross-sectional area, height of center of gravity, height of pressure point and the amount of torque on the shell as it is moved through water. The geometrical con- straints cause snails with any appreciable spire that position their shells primarily by inclination to experience high levels of re- sistance. As a result, snails with round apertures and medium spires have the high- est level of asymmetry. As regulatory de- torsion increases, inclination and asym- metry decrease, as do all of the attendant factors listed above. Each snail studied was subjectively ranked from 1 to 3 on two categories: asymmetry, where “1” approaches bilateral symmetry and “3” represents the highest level of asymmetry; and ornamentation, where “’1””’ represents smooth shells and “3'' the presence of nodes or spines. The two numbers were then added and all snails were ranked from “2” to “6” with rank “2”’ being the most streamlined (and presumably the fastest) and rank “6” offering the most resistance (and hence presumably the slowest). In addition a rank 7" was added for the shell draggers, pre- sumably the slowest of all. Data were gath- ered for 57 species (Table 1). A student t-test was performed to compare the speeds recorded for each rank-group with those of every other rank-group. It was found that for groups 2 through 6 each group was significantly different to at least the 95% confidence level. From inspection of the data in Table 1 it is obvious that significant differences exist. If nothing else, | believe that this cursory and subjective treatment shows that more sophisticated analysis of the problem will be worthwhile. 198 LINSLEY TABLE 1. A pooled list of gastropod taxa investigated at Carrie Bow Key, Belize, and Key Largo and Sani- bel Island, Florida. In ranks 2 through 6 the first figure of the fraction refers to a subjective ranking of asymmetry with rank 1 approximating bilateral symmetry and 3 showing the greatest asymmetry. The second figure of the fraction is a similar ranking of ornamentation. Rank 1 represents smooth shells while rank 3 represents coarse ornamentation. The last number represents average speed. Rank 2 Rank 3 Rank 4 Busycon contrarium 1/1 4.0 Columbella mercatoria 1/2 0.9 Cassis flammea 2/2 1.8 Cyphoma gibbosum 1/1 1.3 Melongena corona 1/2 3.6 Cittarium pica 2/2 0.9 Сургаеа cervus 1/1 2.8 Melongena melongena 1/2 2.4 Crassispira cubana 1/3 0.4 Cypraea cinerea 1/1 4.3 Nassarius vibex 1/2 5.0 Littorina angulifera 2/2 2.4 Cypraea spurca 1/1 3.4 Trivia maltbiana 1/2 0.8 Littorina ziczac 2/2 0.6 acicularis Turbinella angulata 1/20.5 Murex pomum 1/3 1.6 Cypraecassis 1/1 5.6 И 22 Nerita fulgurans 2/2 2.4 testiculus 3 3 Nerita versicolor 2/2 2.6 Fasciolaria lilium 1/1 4.6 Pisania auritula 2122223 hunteria Planaxis nucleus 22187 Fasciolaria tulipa 1/1 6.5 Tegula fasciata 2/2 1.8 Hyalina avena MASAS Tegula lividomaculata 2/2 0.9 Marginella guttata 1/1 3.0 Thais rustica 2/2 2.4 Marginella lactea MARES Turbo canaliculatus 2/2 1.5 Marginella pruniosum 1/13.4 Vasum muricatum 1/3 0.3 Marginella sp. 1/1 5.3 Mitrella ocellata 1/1 1.9 u 1.6 Nitidella nitida 1/1 2.0 Oliva sayana 1/1 8.4 Polinices duplicatus 1/1 5.6 Polinices lacteus 1/1 3.4 Tonna maculosa 1/1 3.6 Average 829 Rank 5 Rank 6 Rank 7 (Shell-draggers) Cymatium sp. 2/3 0.6 Astraea phoebia 3/3 0.4 Batillaria minima 0.6 Leucozonia ocellata 2/3 0.5 Astraea tecta 3/3 0.3 Cerithium guinaicum 0.3 Nodilittorina 2/3 0.6 americana Cerithium litteratum 0.1 tuberculata Astraea tecta tecta 3/3 0.4 Conus jaspideus 0.3 Ocenebra minirosea 2/3 0.3 А 0.36 stearnsi Thais deltoidea 2/3 0.6 rege a Conus mus 0.8 Vexillum dermestinum 2/3 0.5 Conus regius 0.3 А 05 Muricopsis oxy tatus 0.1 9 Polystira albida 0.3 Average 0.35 One interesting sidelight is that nowhere in the analysis was size considered a rele- vant factor. In all other studies performed on the rates of locomotion of arthropods, mammals, fish or birds, the factor of “stride’’ (or its equivalent) had to be соп- sidered and thus there is a positive correla- tion between size and speed. In gastropods dependent upon cilia for their prime loco- motor force, the lack of a correlation be- tween size and speed would be understand- able, for the effectiveness of an individual cilium probably does not vary appreciably between young and adult. In fact adults may well have proportionately fewer cilia and thus move more slowly than the young. Miller’s study (1974) of locomotor types in the Gastropoda suggests that snails moving by muscular contractions show a constant number of waves on the foot within a species. Hence as an individual grows, its “stride'”” may increase and make an adult somewhat faster. However, be- tween species the kind or number of mus- cular contractions does not seem to corre- late with the speed of the organism. Although | have not yet assembled all the data, | believe that speed correlates with food preference. It would be nice if we could make simplistic correlations to the effect that “fast snails are carnivores and slow snails are herbivores.’’ Unfortu- nately the relationship is more complex than that. Although many gastropods are carnivorous, some are carnivores on tube worms, oysters, sea anemones, etc., prey that does not necessitate great speed from the predator. However, | do not think it coincidental that Natica, Oliva and Fascio- laria, three active carnivores, are also three of the fastest gastropods studied. Certainly it will be possible to make generalizations LOCOMOTION IN GASTROPODA 199 from the activity level to the rate of loco- motion and thence to shell shape. Although it may be premature, | propose life modes of some Paleozoic gastropods based on my estimates of their speed potential. INTERPRETATION OF PALEOZOIC BELLEROPHONTACEAN GASTROPODA Considerable controversy exists over the interpretation of isostrophic shells found in the Paleozoic. Some authors (Wenz, Thiele, Runnegar, Pojeta, Jell, etc.) consider all non-septate isostrophic shells as belonging to the Monoplacophora. Other authors (Batten, Knight, Linsley, Peel, Rollins, Yochelson, etc.) believe that while some of these shells are monoplacophorans, the great majority belong to the Gastropoda. The forms under discussion in this paper are presumed to be gastropods. Of all Paleozoic gastropods some mem- bers of the Bellerophontacea seem the best candidates for fast moving snails. The bel- lerophontaceans provide us with a wide array of forms which | interpret to be adaptations to a wide variety of life styles. At the risk of oversimplification, | focus on five genera from this superfamily: Ptomatis, Tropidodiscus, Euphemites, Knightites, and Bellerophon. The Middle Devonian Ptomatis Clarke (Fig. 5) can be taken as typical of a num- ber of taxa which have a large whorl expansion rate with little whorl overlap in the immature conch, and a widely expla- nate, bell-shaped tangential aperture in the adult stage. The center of gravity and pres- sure point are both low and, except for the presence of ornamentation in some species, this general form at first seems well adap- ted to fairly rapid motion. Alternatively, the explanate aperture implies that this was not a particularly active animal. Instead it suggests a form that had a broad foot Fig. 5. Reconstruction of Ptomatis. adapted to life on a firm clayey or silty substrate, possibly as a sluggish grazer or deposit feeder. | would place it in rank 5 for speed. One enigmatic aspect of Ptomatis is the thick, pustulose parietal deposit on mature specimens. In some specimens, presumably the most mature ones, the deposit even projects out into the aperture like a shelf (Knight, 1941, pl. 7, fig. 1f). | believe that this callus can be interpreted as the inser- tion area for the retractor muscles. In immature forms (before the advent of the flared aperture), the columellar retractor muscles would presumably be situated та characteristic bellerophontacean position, about one half volution inward from the aperture. This muscle position would allow protection by deep withdrawal. However, with the development of the bell-shaped aperture in adults, the center of gravity of the shell would be rapidly shifted anteri- orly. Since clamping presumably would now become more important than deep withdrawal, the columellar muscle inser- tions would migrate to the parietal lip. The extension of the shelf anteriorly would place the muscle insertions close to or at the center of gravity of the shell and thus maximize the effectiveness of clamping. Tropidodiscus Meek & Worthen (Fig. 6), from the Lower Ordovician to the Devo- nian, has a widely umbilicate, compressed shell with a deep slit. Because of the wide umbilicus it is geometrically possible for this form to have a tangential aperture. The shell has a relatively high center of gravity Fig. 6. Reconstruction of Tropidodiscus. 200 LINSLEY and high pressure point, but because of its narrow frontal profile | believe this animal to have been capable of considerable mo- bility. The curved form of the aperture would not allow it to clamp its shell against a hard substrate. Tropidodiscus may have lived on soft substrates. Indeed it is typically found in fine calcilutites associ- ated with reefy limestones such as the Anderdon limestone (Linsley, 1968) and Columbus limestone (Stauffer, 1957). | think that this snail was capable of con- siderable speed, but because of its high, compressed shell, it would have lived in low current environments like fine muds or lime sands in protected areas in the back reef zone. The shell form of Tropidodiscus suggests that it belongs in rank 4 for speed. The Upper Paleozoic genus Euphemites Warthin (Fig. 7) is a “fat” bellerophon with much overlap between successive whorls. One interesting feature of Euphemites is the variety of inductural deposits (second- ary shell deposits laid down on top of the primary shell) which cover the entire out- side of the shell (Moore, 1941). This indi- cates that various mantle or foot lobes covered the entire shell when it was active, and may explain why Euphemites has а tangential aperture rather than a radial aperture. The general form of Euphemites is well suited to rapid locomotion and if | were to look for an active predator among the Bellerophontacea this would be a likely OT? NS i a Fig. 8. Reconstruction of Bellerophon. candidate. Among modern gastropods, the Naticidae seem to be analogous in mode of life because the shell is also covered by the foot. While | would not expect Euphemites to have developed the shell-drilling capabili- ties of Natica, Euphemites may have been similar in that it burrowed into sediment with its shell essentially internalized. | be- lieve speed to have been in the range found in category 2. Another Upper Paleozoic form, Knight- ites Moore, is the bellerophontacean on which Knight based his 1952 reconstruc- tion. The shell has small umbilici, a slight flare to the tangential aperture in the adult, but is especially remarkable for periodically constructing paired “‘horns’’ or siphonal canal-like re-entrants in the aperture bor- dering the slit. Since the slit was the site for the exhalant current, it seems reason- able that the “‘horns’’ are indeed homol- ogous to the siphonal canal of modern gastropods in function as well as appear- ance. If this homology is true it makes no sense to reconstruct Knightites as a mono- placophoran (Runnegar & Pojeta, 1974; Pojeta & Runnegar, 1976), for in that reconstruction these inhalant currents would be positioned directly to the pos- terior, something only done by animals that put their heads in the sand, like clams. Knightites makes sense only if it is recon- structed as a gastropod, with torsion bring- ing the inhalant horns to the anterior where they could obtain clean oxygenated water and sample the chemistry of the area directly in front of them. Because of a high development of ornamentation | expect Knightites to have been sedate and prob- ably a grazer on algae fitting in category 5 for speed. The last bellerophontacean to be con- sidered is the long-lived genus Be//erophon Montfort (Silurian to Triassic), geometri- cally the most enigmatic of the group (Fig. 8). The shell has a low profile with many volutions which, as in Euphemites, is accomplished by considerable overlap be- tween the whorls of successive volutions. This combination of overlap and а паг- rowly phaneromphalous or anomphalous condition keeps the shell from having a tangential aperture. This general form is the only one | know of in gastropods that presumably was supported off the substrate in the normal dorsal position and that does not have a tangential aperture. Perhaps it was possible for the form to evolve this LOCOMOTION IN GASTROPODA 201 way because of its deep-withdrawal capa- bility. All known isostrophically coiled monoplacophorans that were unable to effect deep withdrawal have tangential apertures. Many specimens of Bellerophon show a “‘‘waterline’’ of inductural wash about half way up the shell, suggesting that a great deal of mantle and foot were exposed during locomotion. Possibly these animals were fast enough to avoid most predators and, when faced with the rest, depended on deep withdrawal for protec- tion. We know of no opercula from bellero- phontaceans. Perhaps in the absence of an operculum, narrow aperture and deep with- drawal capabilities have strong selective ad- vantage. Bellerophon typifies the problems faced by bellerophonts in general. In order to be fast, they had to keep the shell low and the only ways to do this were to have either considerable whorl overlap or a rapidly expanding generating curve. The rapidly expanding generating curve makes deep withdrawal impractical and as a con- sequence, probably limited effective de- fense to clamping. Clamping, to be effec- tive, requires a large foot and a firm sub- strate to hold onto; Ptomatis seems to have had both. Be//erophon used the other alter- native, large whorl overlap. In order to maintain a mantle cavity of sufficient breadth for effective separation of two inhalant streams, Bellerophon could not have wide umbilici. The geometric prob- lems of a shell with large whorl overlap and no umbilici made it impossible for this form to have a tangential aperture. Because species of Bellerophon have varying degrees of ornamentation | would place it either in category 3 or 4 for its speed. INTERPRETATION OF SELECTED OTHER PALEOZOIC GASTROPODA While the shell geometry of the Bellero- phontacea could have allowed some of them to be among the fastest of the Ar- chaeogastropoda, the geometry of the ma- jority of the members of the Macluritacea suggests that these were among the slowest of the Archaeogastropoda. Within that superfamily are two families, the Maclu- ritidae and the Onychochilidae, each repre- senting distinct adaptations (Linsley, 1977). The Onychochilidae range from the Upper Cambrian to the Lower Devonian and are medium-spired, presumably hyperstrophic gastropods with an elongate aperture tan- gential to the body whorl. This geometry suggests that the onychochilids were mobile animals carrying the shell dorsally with the spire swung forward over the head (Fig. 9). Shells of this sort would offer a high center of gravity and large frontal cross-sectional area and would therefore be quite slow, speed category 5 or 6. | suspect they were predominantly algae-grazers or possibly de- posit feeders, adapted to a hard substrate. In contrast, the shells of the Ordovician family Macluritidae are very low-spired, verging on discoidal (Fig. 10). All have ra- dial apertures with a sharp angulation on the right side of the shell, which presum- ably represented the position of the anus (Knight, Batten & Yochelson, 1960). Be- cause there is little translation during coil- ing in these discoidal shells, they are super- ficially convergent on the isostrophic form of the bellerophontaceans. But macluritids are distinct in having little or no overlap BN dpe Hy Ze Fig. 10. Reconstruction of Maclurites. 202 LINSLEY between successive whorls and, as well as an asymmetrical whorl profile, a slightly depressed spire and a radial aperture. Many are also associated with a heavy calcareous operculum. The radial aperture of the macluritids strongly suggests that these animals did not hold their shells over their backs; this has already been suggested for other reasons (Salter, 1859; Banks & Johnson, 1957). It is presumed that they were shell draggers (category 7) that normally rested the flat “base”” (left side) of the shell on the sub- strate with the anal angulation away from the substrate. The heavy shell of these snails is consistent with the idea that they must have only rarely, if ever, moved their shells. For defense they could clamp the thick operculum over their apertures. The almost total immobility | infer for this group suggests they may have been filter feeders. Knight (1952) suggested that ma- cluritids had only a single ctenidium (the left) and thus a unidirectional water cur- rent through the mantle cavity, inhalant at the portion of the aperture near the sub- strate and exhalant past the anal re-entrant. The Euomphalacea were considered by Knight, Batten & Yochelson (1960) to be closely allied to the Macluritacea and | fully concur, believing euomphalaceans to be derived in the Ordovician as the logical perfection of the macluritid stock. Again the dominant shell form is discoidal, but with variation between hyperstrophic and orthostrophic, even occasionally within the same species (Linsley, in press b). Within this taxon is found a large number of openly coiled shells (Yochelson, 1971; Linsley & Yochelson, 1973), an obvious adaptation that allows the shell to rest flat on the substrate. | interpret the life-mode of the euomphalids to be similar to that inferred for macluritids; essentially motion- less (category 7, shell-draggers), filter- feeding forms, adapted to a wide variety of substrates. The open-coiled forms were adapted to soft substrates, while the tightly coiled helical forms were better suited to firmer substrates. The Pleurotomariacea are the third ma- jor group of Paleozoic gastropods, more numerous and diverse than either the Ma- cluritacea and Euomphalacea or Bellero- phontacea. They diverged from the bellero- phontaceans by extending the spire out to the right where it is repositioned by swing- ing back and up through regulatory detor- Fig. 11. Reconstruction of Pleurotomaria. sion and inclination (Fig. 11). Most Pleuro- tomariacea have a tangential aperture and thus almost undoubtedly carried their shells balanced dorsally as they moved about. The geometry of pleurotomariaceans is limited by paired gills. They necessitate two streams of water flowing through the mantle cavity and separation of the two streams can be maintained only in a rela- tively capacious mantle cavity. As a result, the generating curve cannot depart very far from the circular in Pleurotomariacea. This shape in turn dictates that the shell be positioned with considerable inclination, giving rise to a high center of gravity, large cross-sectional area and high values of torque. Thus the majority of pleuroto- mariaceans will belong to categories 4, 5 or 6 and will move at a rather sedate pace at best, which leads me to believe they were grazers, deposit feeders or scavengers. If any pleurotomariaceans of the past adopted a carnivorous mode of life, they must have preyed on sessile animals or ones for which they could lie in wait. One group of Paleo- zoic Pleurotomariacea, the Raphisto- matinae, seems to have tried elongating the aperture by introducing a short siphonal canal for the left gill. In at least two genera (Scalites Emmons from the Ordovician and Tylozone Linsley from the Lower De- vonian, Fig. 12) the short siphonal canal combined with a medium spired shell to reduce inclination and to position the shell primarily by detorsion. This lowered the resistance and presumably allowed the snails to be among the most active of the pleurotomariaceans, perhaps in category 3. A second group that may have achieved more mobility than the majority of the superfamily is the Porcelliidae (Fig. 13). They achieved this by converging on the form of the bellerophontaceans so the adult shell approaches the isostrophic con- LOCOMOTION Fig. 13. Reconstruction of Porcellia. dition even though the embryonic shell is orthostrophic. Specifically they converged on the general form of Tropidodiscus, doubly phaneromphalous with a tangential aperture and a very deep slit extending over ninety degrees from the aperture. Pre- sumably everything inferred earlier about the mode of life of Tropidodiscus applies equally well to Porcellia. In fact the strati- graphic record suggests that Porcellia re- placed Tropidodiscus during the Devonian and continued through most of the Upper Paleozoic. As might be expected, pleurotomaria- ceans include few shell-draggers. A few possible Paleozoic candidates include Dir- hachopea Ulrich & Bridge (Upper Cambrian- Lower Ordovician) and Ca/aurops Whit- field (Lower Ordovician), both of which become uncoiled in the adult stage; they are certainly shell draggers, but may not belong in Pleurotomariacea. However, Еигу- zone Koken, Pagodispira Horny, Odon- tomaria С. Е. Roemer, Loxoplocus Fischer IN GASTROPODA 203 and Mastigospira LaRocque are all un- doubted pleurotomariaceans from Ordovi- cian to Devonian strata and all obtain open coiling. They may be considered either shell-draggers or completely motionless. These pleurotomariaceans achieve open coiling by increasing translation in contrast to the euomphalaceans that accomplish open-coiling by increasing distance or de- creasing whorl expansion rate. The Missis- sippian genus Rhaphischisma Knight may also have been a shell dragger because it appears to have a radial aperture. This is an interesting genus for the slit is positioned so close to the upper suture that it is impossible that the right gill was still pres- ent. Perhaps it should not be accepted as a pleurotomariacean without further investi- gation. Pleurotomariacea are readily recognized by their prominent apertural re-entrant. Other Paleozoic shells possessing a similar circular generating curve and a tangential aperture lack a re-entrant and are inter- preted as having a single gill. These are considered members of the Trochina. Like the modern Trochacea and the Pleuro- tomariacea they must have balanced their shells with much inclination. The Paleozoic forms, like their modern counterparts, were probably very slow and may have lived on a hard substrate, or even on rocks—their shell form is well suited for clamping. However, a few Paleozoic taxa deviate from this general pattern and deserve comment. Perhaps the best known deviants are the coprophagous Platyceratacea (Bowsher, 1955). This large and diverse group took up residence on the calyx of crinoids, apparently subsisting on the fecal matter. Since these snails were sedentary, some of them attained open coiling. The Tubinidae also have been placed in the Trochina and might have resembled the Platyceratacea in that both were relatively immobile. Open coiling is again common in this group, attesting to their lack of mobility. How- ever, the Lower Devonian Tubinidae gener- ally achieve open coiling by increasing D (the distance of the whorls from the axis of coiling) whereas the platyceratids gener- ally become openly coiled by an increase in T (translation rate). In both groups the shell geometry leaves the aperture free of interference from the spire. The last group that departs from the typical trochinid form that is mentioned in the Treatise (Knight, Batten & Yochelson, 1960) as 204 LINSLEY members of the suborder Trochina, is the Silurian to Devonian Oriostomatidae. These have radial apertures and a heavy, calcare- ous, multispiral operculum and probably should be transferred to the Euomphalacea (Linsley, in press b). The Murchisoniacea constitute an enig- matic group of high-spired Paleozoic snails. Knight, Batten & Yochelson (1960) have placed them in the Archaeogastropoda with some question, because although the Mur- chisoniacea have a slit positioned at or near mid-whorl, presumably indicating paired gills, they are thought to have been on their way towards eliminating the right gill and are considered ancestral to the Loxo- nematacea, placed by them in the Caeno- gastropoda. The two families within the Murchisoniacea are the high-spired Mur- chisoniidae and the moderate-spired Pletho- spiridae. Plethospirid shells are sufficiently low-spired that their shell could have been carried dorsally. Genera such as Ordovician to Silurian Plethospira Ulrich and Silurian to Devonian Diplozone Perner show the development of ап ill-defined siphonal canal which served to elongate the aper- ture, thus reducing inclination of the shell. | expect the Plethospiridae to have been more mobile (category 3 or 4) than most of the Pleurotomariacea. In contrast to the Plethospiridae, the majority of the Murchisoniidae (Fig. 14) are so high-spired according to a “corollary of Linsley’s Third Law,” they must have been shell-draggers (Linsley, 1977) and thus would have had lower activity levels than most pleurotomariaceans. At least four genera of the Murchisoniidae have made peculiar modifications of their shells that attest to this loss of mobility. Gasconadia Ulrich (Ordovician) and Lodonaria Dahmer (Devonian) have developed flaring bell- shaped apertures in the adult conch, the plane of which is roughly parallel to the axis of coiling or the ventral margin of the shell. A similar adaptation is made by the modern genera Cerithium Bruguiere and Distorsio Roding. Thus it would have been possible for them to clamp their aperture Fig. 14. Reconstruction of Murchisonia. against the substrate without elevating the spire. A form from the Lower Devonian, Biangularia Spitz, has a spire compressed parallel to the axis of coiling, somewhat reminiscent of Biplex. This shape would rest stably on the substrate without rolling. However, growth would have to occur in 180°-spurts as it does in the modern genus Biplex in order for the forms to be func- tional. Brilonella Kayser (Middle Devonian) is currently assigned to the Murchisoniidae, but may be a pleurotomariacean. It changes the direction of coiling so that the final whorl coils up and backward so the aper- ture opens near the spire. A conservative interpretation would suggest at least a be- havioral change between youth and ma- turity. Another group of high-spired Paleozoic gastropods is the Loxonematacea. | concur that they are possibly early Mesogastropoda because the sinus, which is at mid-whorl in Ordovician forms, frequently has migrated towards the upper suture in many subse- quent genera. Thus the right gill was pre- sumably lost. The great majority are so high-spired that they must have been shell- draggers, moving slowly and then only rarely. However, a few genera, for example Loxonema Phillips (Middle Ordovician to Mississippian) and Devonozyga Horny (Mid- die Devonian) are lower spired and con- sequently may have been able to hold their shells over their bodies while moving. If this were the case, the spire is high enough so that the shell would be positioned largely by regulatory detorsion with only modest amounts of inclination. Thus if the shell is held erect, these snails would fill the geometric qualifications found in mod- ern moderately fast snails. However, many specimens of the geo- metrically similar Palaeozygopleura Horny which | have collected from the Middle Devonian of New York have their shells covered with trepostomatous bryozoans, in what may have been a symbiotic relation- ship. These colonies frequently cover all the shell except the aperture. For a bryo- zoan to encrust in this manner, the encrus- tation must have occurred while the snail was not only alive but supporting the shell up off the substrate. On the other hand, the mass that the bryozoan colony would add to the shell certainly does not suggest that Palaeozygopleura was a very active gastropod. Even though some modern spe- cies of Thais are encrusted by red algae LOCOMOTION IN GASTROPODA 205 Fig. 15. Reconstruction of Scalaeotrochus. rather than Bryozoa there may be some similarity to these Paleozoic forms. If one can generalize from this example | con- clude that the Loxonematacea were rather slow gastropods with activity levels marked by category 7. Two other small Paleozoic taxa, the Pseudophoridae and Subulitacea, are worthy of comment. The Upper Paleozoic Pseudophoridae (Fig. 15) have а conical shell with a flat base. In most members a frill projects below the level of the base and serves to lift it up off the substrate, much as the frill of the modern xeno- phorids keeps the base of the shell from contacting the substrate (Linsley & Yochel- son, 1973). Linsley, Yochelson & Rohr (in press) have suggested that these extinct animals were immobile through much of their life and possibly had a mode of locomotion similar to the leap of the strombids and xenophorids. Most Subulitacea, which range from the Ordovician to the Permian, have an elon- gate aperture with a weakly developed si- phonal notch suggesting that they had only a single gill (Knight, Batten & Yochelson, 1960). These authors placed them in the Caenogastropoda. A few genera (Subulites Emmons, Cyrtospira Ulrich and Cerauno- cochlis Knight, Fig. 16) are so high-spired that they are best interpreted as having been shell-draggers. In both Cyrtospira and Ceraunocochlis the final whorl deviates so that its axis of coiling of the final whorl is at an angle to the axis of coiling of the earlier whorls. This change creates some- Fig. 17. Reconstruction of Soleniscus. thing approximating a sickle-shaped shell. This would widen the shell and increase the stability of these otherwise long, cylindrical shells as they rested against the substrate and make them less apt to roll in the presence of currents. However, most of the other subulitids are fusiform (Fig. 17) and geometrically would seem to have been capable of holding the shell erect over the body. Because of the elongated aperture, the shell would be positioned primarily by detorsion and the animal would seem to fulfill all prerequisites of a fast snail. Among the Paleozoic gastropods most Subulitacea along with some Beller- phontacea may be the best candidates for active carnivores. It has been the classical assumption that because nearly all modern archaeogastropods are grazing herbivores, all Paleozoic archaeo- gastropods must have inhabited precisely the same ecological niche. Generalizations based on surviving members of a formerly large and diverse taxon should be approached with a measure of caution and skepticism. | believe that there were probably many feed- 206 LINSLEY ing patterns available to the diverse Paleozoic archaeogastropods. Recent studies of the rhipidoglossate radula (Hickman, 1976) sug- gest that it is a very versatile scraping organ, readily adaptable to rasping a variety of substrates. Within modern snails it is possible to relate shell form to activity levels. Although gastropods have not been systematically treated yet in regard to relat- ing rates of locomotion to food preference, it seems probable that there is indeed a correlation. Paleozoic gastropods have shell forms that suggest a great diversity of activity levels and | thus infer that dietary preferences were almost as diverse in the Paleozoic gastropods as they are in snails of the modern oceans. The appearance of the Caenogastropoda in the Mesozoic allowed the development of behavioral modes far better adapted than their archaeogastropod counterparts. The relict archaeogastropods, the limpets, turbiniform and haliotiform snails, are taxa where a circular aperture was no great disadvantage. LITERATURE CITED BANKS, M. R. & JOHNSON, J. H., 1957, Maclu- rites and Girvanella in the Gordon River Lime- stone (Ordovician) of Tasmania. Journal of Paleontology, 31: 632-640. BERG, C. J., Jr., 1972, Ontogeny of the behavior of Strombus maculatus (Gastropoda: Strom- bidae). American Zoologist, 12: 427-433. BOWSHER, A. L., 1955, Origin and adaptation of platyceratid gastropods. University of Kan- sas Paleontological Contributions. Mollusca, Art. 5, p. 1-11, 2 pl. GOULD, S. J., 1969, Ecology and functional significance of uncoiling in Vermicularia spi- rata: an essay on gastropod form. Bulletin of Marine Science, 19: 432-445. HICKMAN, C. S., 1976, Form, function, and evolution in the archaeogastropod radula. Geo- logical Society of America Abstracts of 1976 Annual Meeting, 8: 417-418. KNIGHT, J. B., 1941, Paleozoic gastropod geno- types. Geological Society of America Special Paper 32: 1-510, 96 pl. KNIGHT, J. B., 1952, Primitive fossil gastropods and their bearing on gastropod evolution. Smithsonian Miscellaneous Collections, 117(13): 1-56, 2 pl. KNIGHT, J. B., BATTEN, R. L. & YOCHEL- SON, Е. L., 1960, In: Treatise on Invertebrate Paleontology (MOORE, В. C., Ed.), Part 1, Mollusca 1. Geological Society of America and University of Kansas Press, xxiii and 351 p. LINSLEY, R. M., 1968, Gastropods of the Mid- dle Devonian Anderdon Limestone. Bulletins of American Paleontology, 54: 333-465. LINSLEY, В. M., 1977, Some “‘laws’’ of gastro- pod shell form. Paleobiology, 3: 196-206. LINSLEY, R. L., in press a, Shell form and the Origin of the gastropods. LINSLEY, R. L., in press b, The Omphalocirri- dae: a new family of Paleozoic Gastropoda which exhibit sexual dimorphism. EINSEEY,; ¡RAMAS YOCHEELSON Е 93» Devonian carrier shells (Euomphalidae) from North America and Germany. [United States] Geological Survey Professional Paper 824: 1-26, 6 pl. LINSLEY, В. L., YOCHELSON, E. Li & ROHR, D., in press. MILLER, S. L., 1974, The classification, taxo- nomic distribution and evolution of locomotor types among prosobranch gastropods. Proceed- ings of the Malacological Society of London, 41: 233-272. MOORE, R. C., 1941, Upper Pennsylvanian gas- tropods from Kansas. State Geological Survey of Kansas, Lawrence, Bulletin 38(4): 121-163, ЗЕ PEEL, J. S., 1975, A new Silurian gastropod from Wisconsin, the ecology of uncoiling in Palaeozoic gastropods. Bulletin of the Geologi- cal Society of Denmark, 24: 211-221. POJETA, J., Jr. € RUNNEGAR, B., 1976, The paleontology of rostroconch mollusks and the early history of the phylum Mollusca. [United States] Geological Survey Professional Paper 968: 1-88, 54 pl. RAUP, D. M., 1966, Geometric analysis of shell coiling: general problems. Journal of Paleon- tology, 40: 1178-1190. REX, M. A. & BOSS, K. J., 1976, Open coiling in Recent gastropods. Malacologia, 15: 289-297. RUNNEGAR, В. € POJETA, J., Jr., 1974, Mol- luscan phylogeny: the paleontological view- point. Science, 186: 311-317. SALTER, J. W., 1859, Canadian organic remains, decade 1. Geological Survey of Canada, Mon- treal, p. 1-47. STAUFFER, C. R., 1957, The Columbus lime- stone. Journal of Geology, 65: 376-383. VERMEIJ, J., 1971, Gastropod evolution and morphological diversity in relation to shell geometry. Journal of Zoology, 163: 15-23. VERMEIL, С. J., 1975, Evolution and distribu- tion of left-handed planispiral coiling in snails. Nature, 254: 419-420. YOCHELSON, E. L., 1971, A new Upper De- vonian gastropod and its bearing on the prob- lems of open coiling and septation. Smith- sonian Contributions т Paleobiology, (3): 231-241, 2 pl. MALACOLOGIA, 1978, 17(2): 207-221 THE DEPLOYMENT OF OPERCULATE LAND SNAILS IN RELATION TO SHAPE AND SIZE OF SHELL A. J. Cain Department of Zoology, University of Liverpool, P.O. Box 147, Liverpool, England ABSTRACT Following the discovery of a regular bimodal distribution of shell height A versus shell width d in taxonomically unrelated terrestrial faunas of Stylommatophora, the land operculate snails of the world are examined. They show the same type of distribution of A versus d as do the stylommatophorans of the same faunas except in Africa and Madagascar. In the major groups of land operculates, the distribution is clearly different from that of their marine relatives. As with the stylommatophorans, different taxonomic groups of land operculates combine in different ways in different faunas to make up the same type of bimodal distribution. Surprisingly, land archaeogastropods and land mesogastropods seem to make up a single bimodal distribution, not two separate ones. Such reciprocal adjustment of the distributions of subgroups can only be by some type of interference, presumably competition. Since land operculate and land stylommatophoran distributions of A versus d overlap widely in most regions, these two major groups are presumably not in competition, although subgroups within them are, but are both subject to the same overall selective pressures determining the typical bimodal distribution. Some consequences of these findings for the interpretation of present-day zoogeographical distributions are pointed out, and the possi- bility is noted of much more general application of such diagrams (when appropriate measurements have been discovered). INTRODUCTION In a previous paper (Cain, 1977) and in current work (Cain, in preparation, a & b), it is shown that the height (h) and diame- ter (d) of shells of free-crawling fully re- tractile gastropods are not distributed at random with respect to each other but show distinct patterns that are character- istic, but by no means invariably so, of major taxonomic groups. In fully terrestrial Stylommatophora a bimodal distribution of h versus d is normally found, even in faunas taxonomically unrelated, except for certain subfamilies in wet tropical regions (Cain, in preparation, a). Scantily in tem- perate regions but throughout the tropics, and especially in the Caribbean and in southeast Asia, there exist side by side with the stylommatophorans many species of operculate land snails (prosobranchs), the result of five major and several minor in- dependent invasions of the land. In Europe, within a few feet or inches, species of prosobranchs and stylommatophorans can be found permanently coexisting in what appears to be the same habitat, and ap- parently in many other regions as well. These prosobranchs form a natural experi- ment in gastropod terrestrial life, in parallel with the stylommatophorans, and a com- parison will show how far they have under- gone parallel variation, and how far they retain the characteristic distributions of h and d of their marine relatives. To a large extent they appear to take on character- istics of the stylommatophorans. MATERIALS AND METHODS Measurements of height and diameter were made as described by Cain (1977). They were each made on a single adult shell chosen as representative (when, as usual, choice was possible) of each species currently recognized in the collections of the Academy of Natural Sciences, Phila- delphia. Although in groups not recently revised there may be more species than are really warranted, these extensive collections are unlikely to lack any genera or sub- genera showing markedly peculiar char- acters, having been built up over 75 years by H. A. Pilsbry (long the world author- ity), H. Burrington Baker, and many (207) 208 CAIN others, by collection and by exchange with other authorities all over the world. For the present purposes, the exact numbers of species in particular groups are not impor- tant, only the sampling of main shell varia- tions. Indeed, for many parts of the world no existing collection could give authorita- tive answers on species limits or numbers of existing species. Only a sample can be hoped for in groups so little worked; but as striking variations are readily noticed and collected and as museum material usually over-represents variation within species, it is likely that variation between species has been sampled adequately, if sketchily in some regions. The choice of a single specimen from a single locality may perhaps reduce the number of symbols on the diagrams, if any wide-ranging species is shown from a single region only. However, the regions taken (southeast Asia, Philippines, Central Amer- ica, Cuba, etc.) are large enough to have very few species in common, and no per- ceptible error is likely from this source. OVERALL VARIATION The principal land operculates are the archaeogastropod families Helicinidae and Hydrocenidae (superfamily Neritacea), and the mesogastropod superfamily Cyclo- phoracea and the families Pomatiasidae, and Aciculidae (= Acmidae) (both super- family Littorinacea). Although the Cyclo- phoridae have been elevated to a super- family (Tielecke, 1940) and this rank is accepted by Taylor & Sohl (1962), a num- ber of genera have not been allocated to a family within it. For the purposes of this paper (and with no pronouncement on rank intended), a single family, Cyclo- phoridae, has been retained within the superfamily. The Chondropomidae are re- tained as a subfamily of the Pomatiasidae. In various parts of the world particular species or genera have become virtually or wholly terrestrial: for example, the genus Blanfordia in Japan and Pomatiopsis lapidaria in North America (Rissoacea, Pomatiopsidae; G. M. Davis, personal com- munication), Geomelania in the mountains of the Greater Antilles (Rissoacea, Trunca- tellidae), and Cremnoconchus in the Indian mountain ranges (Littorinacea, Litto- rinidae). In New Caledonia a few species of neritids are found in very damp places on trees and bushes overhanging water. In the Assimineidae (Rissoacea) the subfamily Omphalotropidinae has terrestrial species from the Mascarene Islands, Japan, and the Marianas to New Zealand. Most of these other invasions of the land are small com- ponents of land operculate faunas and are not considered here. Figs. 1-3 show the overall variation in shell shape and size for each superfamily. The pattern most like any of those pre- viously reported is that of the Cyclo- phoridae (Fig. 2), which are distributed bimodally very much as is the stylommato- phoran fauna of either western Europe or North America (Cain, 1977; figs. 3, 4). The distribution is not at all that seen for mesogastropods in the sea (Cain, 1977: figs. 18, 19, North American species taken as representative) which is a wedge-shaped distribution covering the upper scatter in Fig. 2 but much wider at the higher values of h and approaching the bisector (indeed, spilling right across it at values of d below 0 10 20 30 mm FIG. 1. Variation in shell shape and size in terres- trial archaeogastropods, superfamily Neritacea. Main figure, Helicinidae; inset, Hydrocenidae to same scale. In all figures the vertical axis is the height of the shell 4, and horizontal, the width of the shell d. The line h = d is shown in each to facilitate comparison. OPERCULATE LAND SNAIL SHAPE AND SIZE 209 0 10 20 30 40 50 60 mm FIG. 2. Variation in shell shape and size in terrestrial mesogastropods, superfamily Cyclophoracea. 20 тт), filling in the gap left in Fig. 2 between the upper scatter and the bisector. Some mesogastropods, therefore, show one type of distribution of A and d in the sea and a very different type, like that of other terrestrial snails, on land; there is no pat- tern of distribution characteristic of meso- gastropods as such. The archaeogastropod Helicinidae (Fig. 1) correspond to a good part of the lower scatter of the Cyclo- phoridae; the marine archaeogastropods (Cain, 1977: Fig. 17), however, with very few exceptions, are scattered along the bisector; again, there is a marked difference between the scatters of the marine and the terrestrial representatives of the same major group. The Pomatiasidae (Fig. 3) show a truly remarkable distribution, reminiscent of the marine mesogastropods, and filling the gap between the upper and lower scatters of the Cyclophoridae; but unlike that of the marine mesogastropods the distribution does not expand upwards, lying instead more or less parallel to the bisector for d between 12.5 and 22.5mm, while at the same values generating a scatter that lies along the bisector and expands more or less equally on either side of it at higher values. This latter is like the marine archaeo- gastropod distribution (Cain, 1977: fig. 17). The Aciculidae (Fig. 3, inset) fall close to the upper scatter of the Cyclophoridae but, considered by themselves, could be a part of the marine distribution. The Hydro- cenidae (Fig. 1, inset), like the Aciculidae, could be considered a part of the marine distribution, or of the terrestrial. On overall distribution, therefore, in one major family a bimodal distribution highly reminiscent of that of stylommatophorans is found rather than the unimodal distribu- tion of its close relatives in the sea. In two others the distribution is certainly different from that of corresponding marine forms but to a lesser extent, and in two very small and unvarying families no conclusion can be reached. 210 CAIN 0 10 20 30 40 50 60 mm FIG. 3. Variation in shell shape and size in terrestrial mesogastropods, superfamily Littorinacea. Main figure, Pomatiasidae; inset, Aciculidae (= Acmidae) to same scale. Black dots, whole shells; circles, truncated shells, plotted with actual values of A after truncation. GEOGRAPHICAL VARIATION Deployment of the families The Helicinidae are found abundantly on the Caribbean islands, less so in the warmer parts of the American continents, and quite separately, often abundantly, in the Philippines, Malay archipelago, Austra- lasia, Polynesia, Micronesia and Hawaii. The Hydrocenidae occur thinly in Dalmatia, some Atlantic islands, South Africa, eastern and southern Asia, islands of the Pacific, and parts of Australasia, including New Zealand. Of the Cyclophoridae, the Cyclo- phorinae are widespread from Japan and China through India, southeast Asia, and the Philippines and throughout most of the larger (and many smaller) islands of the Pacific, including New Zealand, with a few species in Africa and islands to the east; one reaches as far westward as the shores of the Caspian Sea. The Hainesiinae com- prise a few species in Africa and Mada- gascar. The Pupininae are also from India, China, and Japan to northern Australia and New Zealand and out into Micronesia and Melanesia. The Alycaeinae reach from India and south China to the Philippines and Greater Sunda isles. The Diplommatinae reach somewhat further east to the Caro- lines, New Caledonia, and Queensland, but with one genus, Adelopoma, in South and Central America. The subfamily Poteriinae is wholly South and Central American and Antillean. Lastly the Craspedopomatinae are confined to Atlantic islands (Azores, Canaries, Madeira—the region sometimes re- ferred to as Macaronesia) and the Cochlo- stomatinae are southern European and North African, east to the Caucasus. The Pomatiasidae have one subfamily, Pomatiasinae, predominantly in Africa, Madagascar, Socotra, and the east African islands, with a few species reaching south- ern and western Europe and the Canaries and east to the Caucasus, and others in OPERCULATE LAND SNAIL SHAPE AND SIZE 211 India. The other subfamily, Chondro- pominae, is exclusively in the warmer parts of the New World, especially in the Antilles. The Aciculidae (= Acmidae, in Thiele, 1931) are wholly Palaearctic, main- ly around the Mediterranean and eastward to the Caucasus. These brief geographical indications show that these families are unequally dis- tributed in different parts of the world. The operculate land fauna varies greatly in taxonomic composition. We must therefore ask what parts of these overall distributions are found together in each major region, and what sorts of distributions are then made up. Oriental and Australasian On the Indian subcontinent pomatiasids and helicinids are few; the distribution is made up almost entirely of cyclophorids, and (Fig. 4) both scatters are represented. This distribution corresponds closely with those for western European or North American stylommatophorans (Cain, 1977) and shows the same bimodality. The distri- bution for North America is similar in all but the degree of extent to high values of the two scatters, the North American upper scatter reaching as high as h = 70 mm. However, the lower scatter seems itself to fall into two groups, a lower departing widely from the bisector and an upper (above about а = 20mm) running just below it. If such a diagram is made for southeast Asia (Fig. 5), an almost identical picture is obtained, with the same taxon- omic arrangement. This again suggests that the lower scatter is itself bipartite, being at medium values of d well below the bisector and at high ones (4 more than 20 mm) along and just below it. This bimodality is not apparent in the North American stylommatophoran scatter; there is a hint of it in the western European, which is d 30 40 50 mm FIG. 4. Land operculates of the Indian subcontinent, including Ceylon and Burma. Symbols on this and following figures: cross, Stylommatophora; stippled triangle, Helicinidae; black square, Hydrocenidae (Old World only); clear circle, Cyclophoracea; clear square, Pomatiasidae; black triangle, Aciculidae (Palaearctic only). In this and subsequent figures many species with small shells (A and d less than 5 mm) are omitted for clarity. 212 CAIN 0 30 40 50 60 mm FIG. 5. Land operculates of southeast Asia, including the Malaysian islands, Borneo, Celebes and the Sulu Archipelago, the Andamans and Nicobars, and Hainan. For explanation of symbols, see Fig. 4. confirmed when the Helicidae of all the Palaearctic are combined (Cain, in prepara- tion, b). The few Indian pomatiasids are in the lower scatter. In China the upper scatter hardly ex- tends above h = 15 and the lower beyond d = 30, but in essence they are the same as in the Indian fauna. In the Philippines (Fig. 6) the pattern is again recognizable, but now there are many more helicinids. They tend to occupy the lower part of the lower distribution, which again may be compo- site. In Australasia (in the wide sense) it is the helicinids almost exclusively which make up the lower distribution below d = 15 (Fig. 7). The New Zealand distribution (Fig. 8) has lost the lower scatter alto- gether, in which it resembles that of Europe (Fig. 14), and consists mainly of cyclophorids. New World In striking taxonomic contrast, the land operculates of Cuba (Fig. 9) are mainly pomatiasids and helicinids. Again there are two scatters, but the lower one is mainly helicinid and the upper pomatiasid. The few cyclophorids contribute to the width of the upper part of the upper scatter. A very few helicinids are tall-spired, but they do not move far into the upper scatter. A few very low-spired, almost planorboid pomatiasids are also found. Jamaica shows a similar picture without the cyclophorids. In both, the upper scatter is bent more toward the bisector than in Indian, Philip- pine, and Australasian faunas, because in the Greater Antilles many species of the upper distribution practice truncation of the shell. If complete specimens of these species could be found, the upper distribu- tion would be like that elsewhere or even steeper. In Central America (Fig. 10) the separation of families is clearer, with the Cyclophoridae at the upper ends of both scatters, Pomatiasidae at the lower end of the upper, and Helicinidae at the lower end of the lower. This separation poses some problems in ecology, which are discussed OPERCULATE LAND SNAIL SHAPE AND SIZE 213 0 10 20 d 30 40 mm FIG. 6. Land operculates of the Philippines, including Balabac. For explanation of symbols, see Fig. 4. 0 10 20 30 40 mm d FIG. 7. Land operculates of Australasia (wide sense, excluding the New Caledonian group and New Zealand), from the Moluccas and Bali to Fiji, Samoa, and the New Hebrides. For explanation of symbols, see Fig. 4. 214 CAIN 0 10 20 mm d FIG. 8. Land operculates of New Zealand. For explanation of symbols, see Fig. 4. later. North America has very few land operculates (Cain, 1977: fig. 10); they do not contradict the Antillean distributions. South America agrees well with Central America, with no high-spired pomatiasids. So far, it appears that land operculates in different parts of the world behave like stylommatophorans in respect to shell size and shape and that much the same two scatters are found anywhere, with the same family filling different parts of the scatters according to its occurrence in different parts of the world. The pomatiasids (Chondropominae) are responsible for most of the upper scatter in the New World, the helicinids for the lower part of the lower scatter; cyclophorids, when pres- ent, take the upper part of the lower scatter and occasionally that of the upper. In the Orient and Australasia, helicinids, when present, take the same role as in the New World and cyclophorids fill up the rest, with a few pomatiasids in the lower scatter in India. Ethiopian region In Africa (Fig. 11) and Madagascar (Fig. 12) we get a completely different picture. Almost the whole of the pomatiasid distri- bution that seemed remarkable before (Fig. 3) is concentrated in these regions. Unlike in the rest of the world, there is almost entirely a single scatter, running right up the bisector. If these species were removed from Fig. 3 and notional compensation were made for the truncation of most of the New World species, we would be left with a distribution of pomatiasids very similar to the upper scatter of Indian or Australasian land operculates or of Euro- pean or North American stylommato- phorans but not really like the wide spread 30 40 mm d FIG. 9. Land operculates of Cuba and the Isle of Pines. For explanation of symbols, see Fig. 4. OPERCULATE LAND SNAIL SHAPE AND SIZE 205 0 10 20 30 40 50 60 mm d FIG. 10. Land operculates of Central America, from Mexico to Panama, including Swan and Roatan Islands. For explanation of symbols, see Fig. 4. mm 40 30 20 10 0 10 20 E 30 40 mm FIG. 11. Land operculates of the Ethiopian region, including Socotra and nearby islands, and the Yemen. For explanation of symbols, see Fig. 4. 216 CAIN 0 10 20 30 40 50 60 mm FIG. 12. Land operculates of Madagascar, the Comoros, and the Mascarene Islands. For explanation of symbols, see Fig. 4. of marine mesogastropods (Cain, 1977: figs. 18, 19). The African and Malagasy distributions resemble, remarkably enough, that of the marine Archaeogastropoda (Cain, 1977: fig. 17). The apparent peculi- arities of the Pomatiasidae are therefore partly removed and partly shown to be a phenomenon concentrated in the Ethiopian and related regions. Madagascar appears to have a considerable radiation of pomatiasid types, but unlike in other groups (mam- mals, birds) it is similar, even generically (Thiele, 1931), to the African and not a spe- cial local development. Remarkably enough, a preliminary investigation of the African land stylommatophorans shows no departure from the usual bimodality (Fig. 13). Palaearctic In Europe east to the Caucasus and west to the Atlantic islands (Macaronesia), again a different picture is seen (Fig. 14). With one exception (a cyclophorid, Caspicyclo- tus, that reaches northwestward from the main distribution in southeast Asia as far as the Caucasus), all are in the upper scatter. The smallest are the endemic Aciculidae and a few Hydrocenidae. The Craspedo- pomatinae continue the line of the hydro- cenids and are confined to some of the Atlantic islands. On the mainland, the endemic Cochlostomatinae continue up- ward the line of the Aciculidae. Finally, there is a rather wide scatter of poma- tiasids, of genera endemic to the region but related to the African and Madagascan forms. Some of these European Poma- tiasinae are rather high-spired, but they are reminiscent of their African and Mada- gascan relatives. The Cochlostomatinae are remarkably homogeneous in shape and rather high-spired, but they do fit into cyclophorid distributions (upper scatter) in the Old World. The craspedopomatines are close to the bisector, but there are some cyclophorid forms similarly intermediate in position between the two scatters in the Indian and southeast Asian faunas (Figs. 4, 5). Nevertheless, the only picture which corresponds at all well with the European OPERCULATE LAND SNAIL SHAPE AND SIZE 217 + 30 40 mm FIG. 13. Land Stylommatophora of the former Belgian Congo (Pilsbry, 1919). Nine forms too large to show оп the diagram carry the upper distribution to h = 140, а = 72 mm. For explanation of symbols, see Fig. 4. is that for New Zealand, the other oceanic- temperate region investigated. The Еиго- pean distribution is certainly in the upper scatter of the Indian, Australasian or (com- pensated) Antillean faunas, except for the pomatiasines, which are not unlike the African or Madagascan forms but do tend to fill in the gap between the two scatters, although they are not truncated. Conclusions on distribution This survey shows that the land oper- culates, with the possible exception of the Hydrocenidae and Aciculidae and the definite exception of African and Malagasy pomatiasines, conform to patterns of distri- bution of h versus d that are characteristic of land gastropods generally or of particu- lar regions of the earth, not of their close relatives in the sea. All aciculids and hydro- cenids are above the bisector, but wide changes in shape occur in the other three families. Low-spired to planorboid poma- tiasids are found in the New World (Abbottella, Jamaica; etc.) and Old World (Cyclotopsis subdiscoidea, \ndia; Lithidion forbesianum, Socotra). High-spired 218 CAIN 0 10 20 mm FIG. 14. Land operculates of the Palaearctic re- gion (to the Caucasus) including Macaronesia (the Atlantic Islands). For explanation of symbols, see Fig. 4. helicinids occur in the New World (A/cadia (Idesa) charmosyne, Dominican Republic; Semitrochatella elongata, Cuba; Troschel- viana chrysochasma, Cuba). The tall cyclo- phorids are poteriines in the New World, cochlostomatines in Europe, various pupinines in southeast Asia and Austra- lasia, hainesiines in Africa, some cyclo- phorines and local pupinines in New Zea- land; depressed ones are cyclophorines and some alycaeines in the Orient and Africa. It is not the case, therefore, that wherever there is occasion for a high-spired (or low- spired) cyclophorid, the same forms taxon- omically are found. The filling-in of differ- ent areas in the scatters seems to be re- markably ad hoc. DISCUSSION Operculate and stylommatophoran scatters Since in most regions of the earth the operculate land snails and the land stylom- matophorans fall into much the same two scatters, filling them up irrespective of taxonomic relationships, it seems that these distributions must be dictated by some very pervasive ecological conditions on land. The largely unimodal distributions shown by the marine archaeogastropods, mesogastropods, and neogastropods are pre- sumably equally dictated by ecological con- ditions in the sea. It is true that the fully retractile marine euthyneurans (Cain, 1977: fig. 23) show a bimodal distribution very like that of their relatives, the Basommato- phora and Stylommatophora, in fresh water and on land. It might be argued that this is an ancestral feature, retained on coming out of the water. But in the first place, species of intermediate characteristics do occur occasionally in various terrestrial stylommatophoran and prosobranch groups, and what is to be explained is not their total absence but their rarity. Secondly, as shown already for the European and North American stylommatophorans (Cain, 1977: figs. 3, 4) and here for land operculates generally, within families, subfamilies, and occasionally even genera, variation between species can occur from the upper to the lower scatter or vice versa. If there is any control from the constitution of the an- cestral stock, it is so labile as almost to be a self-contradiction. Even more important is the way in which both stylommatophoran (Cain, 1977: figs. 6-8, Western European, and fig. 37, a-f, North American) and prosobranch families (this paper) combine to fill up each of their two scatters with little over- lapping. The cyclophorids (above, Figs. 4-7, 9, 10) are particularly instructive here. The impression given by the scatter diagrams is that there is a more or less fixed repertoire of both high-spired and medium- to low- spired shapes and sizes allowable in a re- gion and that the available stocks evolve to fill it. Such patterns as these can be brought about only by ecological con- straints that determine the extent of avail- able niches and by reciprocal interference, probably competition, between the avail- able stocks that causes them to share out the niches. It would follow, as already suggested (Cain, 1977), that the measure- ments h and d estimate properties of the shell which, however indirectly, are them- selves estimators of the type of adaptation of the species to filling a particular niche in competition with others. It happens that this can be done for land snails very largely (not completely since there are some over- laps) by a two-dimensional diagram. Raup (1966) has already shown a mutual ex- clusiveness between major groups of filter- feeding animals with shells, using three dimensions. Three matters arise from the present diagrams, one predictable on the hypothesis given above, and the other two not. The radulae of land operculates, unlike their shells, do not show much change toward those of stylommatophorans, but resemble OPERCULATE LAND SNAIL SHAPE AND SIZE 219 rather those of their marine relatives. If their modes of feeding are different from those of the stylommatophorans, they are probably not in direct competition and can overlap widely with them in h and d; in competition they would show a mutual exclusiveness. A considerable overlap in h/d between the European land prosobranchs, except the pomatiasids and the high-spired stylommatophorans, has already been pointed out. Examination of the Philippine land snails (Fig. 6, and Cain, in prepara- tion, a) shows a strong similarity of opercu- lates and stylommatophorans in the lower scatter, and a partial one at lower values of h and d in the upper. A rapid check on the Indian and South American land faunas gives similar results. The land operculates, on the whole, are not filling in the gaps between the stylommatophoran scatters but are coinciding with them, as would be expected. What is unexpected is that the land archaeogastropods should join with the land mesogastropods in filling up their scatters instead of overlapping them. Land archaeogastropod radulae are distinct from those of the mesogastropods. On the same reasoning as that given above, one would not expect effective exclusion in the distri- butions between these groups. Unfortunate- ly, so little appears to be known of the exact modes of feeding of herbivorous snails in the sea, still less of operculates on land, that only further research can solve the problem. Equally unexpected is the difference in shape of the overall operculate scatter for Africa and Madagascar. If, as suggested above, this shape defines the repertoire of available niches, then either there is some- thing very different about the Ethiopian region in respect of niches open to land operculates or their relationship with other land snails is different. It is unlikely that a difference in colonization is in question. The African land mass is old, and there has been time in Madagascar for other terrestri- al groups to undergo extensive adaptive radiation to occupy available niches. There is no reason to think that the land molluscs have not done so as well. The peculiar distribution of the land operculates on the h, d diagram must therefore mean a pecul- iar distribution of niches. The African land stylommatophorans seem to be orthodoxly bimodal. A tentative hypothesis It has been pointed out at various times that tree snails tend to Бе high-spired, although some that live on large flat leaves may be depressed. Although this idea needs detailed confirmation in regions where there are plenty of tree snails, it may point to a provisional hypothesis of the scatters found. Francis Bacon remarked that we can learn from error, but not from confusion; the disproof of a definite hypothesis may add considerable knowledge. If high-spired shells are usually of species that feed more or less vertically on rock faces and in trees, and low-spired ones are usually of species feeding on the ground or on horizontal surfaces, intermediate-shaped ones are pre- sumably dual-purpose species. Since land prosobranchs and stylommatophorans over- lap extensively in their distributions of h versus d within a fauna, presumably they are living in the same places and so are subject to the same selective pressures with respect to A and d—on the present hypoth- esis both groups are feeding on the same variety of surfaces. But they must be doing something different to be able (a) to co- exist each with the other major group and (b) to interfere with their relatives within each group in order to fill out the expected distribution. Although many factors must be acting, it seems likely that modes of feeding are the clue to both (a) and (b). The differences in radular teeth and jaws between prosobranchs and stylommato- phorans, while permitting a considerable overlap of diet during temporary super- abundances of particular foods, may allow the herbivorous stylommatophorans to con- centrate more on cutting and less on scrap- ing and the herbivorous prosobranchs, on the contrary, to scrape more than cut. (In particular circumstances, no doubt, there will be convergence, where the ecological situation allows, in this as well.) It might be thought that prosobranchs, considering their general habits on the shore and their direct invasion of the land many times over, might rely principally on fine turves of algae or other plants less relied on by shelled prosobranchs. Terrestrial algae are world-wide. They have even been isolated from the pavement outside the Marks & Spencer store in Liver- pool (Dr. John Eaton, personal communica- tion). But in the British Isles they are 220 CAIN sporadic and evanescent, quite unsuitable for any species to specialize on during the whole of its active season. In warmer and wetter countries they are more abundant and more permanent, as are other epi- phytes and surface plants (yeasts, bryo- phytes, etc.). If there is a feeding distinc- tion between land operculates and land stylommatophorans of the nature sug- gested, it is reasonable that land opercu- lates should be more abundant in warmer countries. Where it is wet and warm, crops of algae can be more or less permanent or can develop more quickly or persist for longer when the wet season arrives and snails come out of aestivation. The greater abundance of broad-leaved evergreen trees provides additional niches not seen in cold- er regions. Of the two scatters of h versus d, the upper one only reaches to high values of A in warmer countries. Almost all the large high-spired forms in the North American stylommatophoran fauna (Cain, 1977: fig. 4) are southern; high-spired forms in Europe are predominantly south- ern in both stylommatophorans and land operculates; and it is the high-spired forms that are cut back as one moves north into China from southeast Asia. In the far north, low-spired stylommatophorans are also cut back to small values in both North America and Eurasia (Cain, in preparation, b). Conversely, in tropical forest in both the Philippines and New Guinea, inter- mediate stylommatophorans become соп- spicuous in addition to those of the normal two scatters, as though intermediate habi- tats (or habits) are opened up (Cain, in preparation, a). It is possible that a large amount of the peculiarities of distribution of present-day snails really reflects the pres- ent distribution of meteorological factors in relation to their appropriate food crops. Interpretation of geographical distribution If this is so, a further point should be made. The restriction of helicinids virtually to the Caribbean and adjacent countries and the larger islands of the western Pacific is often quoted as a classic example of a relict distribution (e.g. Solem, 1959: 293). So, no doubt, it is; but if these are regions, both highly maritime, with a dampness of air and relative constancy of temperature that allows for better maintenance of standing crops of algae and epiphytes, there may be an excellent ecological reason for helicinids persisting in these disjunct re- gions, and even an excellent reason for retaining an archaeogastropod radula in those snails with small low-spired shells, as against a mesogastropod one. Similar con- siderations apply to the cyclophorids and pomatiasids. We need to distinguish two different meanings of the word primitive. Early members of a particular stock may be primitive both because they are early and because they are attempting to fulfil the same ecological role as their descendants but are not yet specialized for it; if it were possible to put them in competition with their descendants, they would be unsuc- cessful. But forms may appear early in the fossil record, rapidly adapt as well as pos- sible to their ecological roles, and then remain constant simply because they are still fulfilling the same role and the eco- logical opportunity for it is still available. They are not then imperfectly specialized for it, although they remain persistently primitive. There are various statements in the literature that, for example, the more primitive endodontid snails (Stylommato- phora) have a relict distribution in the mid-Pacific islands, being pushed out else- where by more advanced forms. But if they are specialized for the sort of environment now to be found only in such islands, they may be holding their own successfully against invasions by other forms and actual- ly excluding the so-called more advanced forms. In that case, what we have is a peripheral distribution at present of a par- ticular sort of environment and not at all a historical situation, with more primitive forms slowly being eliminated by competi- tion from others more specialized for the same ecological role. Without much further evidence it is dangerous to argue from an observed distribution to historical changes. Of course, one must consider only habitats in equilibrium—the severe damage often caused by man may well allow all sorts of introduced weed species to establish them- selves successfully. One should be very cautious, therefore, in trying to explain distributions like these as purely historical or due to the slowness of evolution (implying that primitive forms ‚ are still slowly adapting). The resemblance | of the A versus d scatter diagrams for such different faunas implies that the groups are in some sort of equilibrium within a fauna OPERCULATE LAND SNAIL SHAPE AND SIZE 221 and that their characteristics (of shell size and shape) are dictated by present ecologi- cal conditions. Use of diagrams The use of diagrams such as those shown in this paper can be extended to other groups of animals when one discovers the necessary measurements. The form of the scatter, if repeated in unrelated parts of the world, gives the repertoire of available niches; the mutual exclusion of taxon- omically defined scatters shows which taxonomic groups are apparently liable to interaction. Obviously, one could measure height and diameter of elephants and sun- birds, say, and get well-separated scatters on the same diagram; but wide separation does not indicate interaction, only con- tiguity combined with the sort of role changing within the same framework seen in Figs. 4, 7 and 9. Forms that overlap are not interfering, with respect to the char- acters measured (or rather, to their conse- quences and concomitants). Those that ex- clude each other show the results of inter- ference, probably of competition. In the example worked out in this paper, land operculates and land stylommatophorans appear to be free from interference from each other but members of each group compete with others in respect of char- acters associated with shell shape and size; at the same time, both groups, being equal- ly terrestrial, come under the same general ecological constraints and show the same overall pattern—except in Africa and Madagascar. When a pattern for a particular group is clearly recognized in several well- known faunas, it may be possible to use it to reconstruct the diversity in a badly known one. The overall scatter should be the same, and some idea of the extent of our ignorance will be given by the area within the expected scatter which is at present vacant. This method could be ap- plied, for example, to the contents of poorly fossiliferous strata or to undercol- lected or recently devastated faunas. Rensch (1959: 119) has already pointed to the relative constancy of percentage of small-sized shells in leaf litter land snail faunas of very different taxonomic com- position. He thinks, however, that the con- siderable variation in percentage of medi- um-sized and large land snails in the same forest faunas shows that “there is hardly any serious competition among the indi- viduals and the species,” which is not like- ly. An examination of detailed diagrams of the sort used in the present paper would be necessary to see whether there is evidence of reciprocal interference in the first place. ACKNOWLEDGMENTS | am indebted for criticism and advice to Professors J. D. Currey and M. H. Williamson; Drs. С. М. Davis, J. Eaton, Е. Hoagland, and R. Robertson; and Mr. R. W. Cowie. LITERATURE CITED CAIN, A. J., 1977, Variation in the spire index of some coiled gastropod shells, and its evolu- tionary significance. Philosophical Transactions of the Royal Society of London, Ser. B, 277: 377-428. CAIN, A. J., in preparation, a. Variation of terrestrial gastropods in the Philippines in rela- tion to shell shape and size. CAIN, A. J., in preparation, b. Variation in shell shape and size in land stylommatophoran faunas of the Palaearctic and Macaronesia. PILSBRY, H. A., 1919, A review of the land mollusks of the Belgian Congo chiefly based on the collections of the American Museum Congo Expedition, 1909-1915. Bulletin of the American Museum of Natural History, 40: 370 p. RAUP, D., 1966, Geometrical analysis of shell coiling: general problems. Journal of Paleon- tology, 40: 1178-1190. RENSCH, B., 1959, Evolution above the species level. London, Methuen, 419 p. SOLEM, A., 1959, Zoogeography of the land and freshwater Mollusca of the New Hebrides. Fieldiana: Zoology, 43: 240-359. TAVMEOR О. М. & SORE NIE 1962, Ап outline of gastropod classification. Mala- cologia, 1: 7-32. THIELE, J., 1931, Handbuch der systematischen Weichtierkunde, 1. Jena, Fischer, 778 p. Re- printed 1963, Amsterdam, Asher. TIELECKE, H., 1940, Anatomie, Phylogenie und Tiergeographie der Cyclophoriden. Archiv fiir Naturgeschichte, N.F., 9: 317-371. of аа. Can бо 40000. ar НА Part or tan. develop mors quickly or-peérticl fa: longer whi AMIS IWOMADA vas” and malla. сое Gut of asstivabon.. Vi greater али Gg» mid rat battens mares Drei ECO Ob! ee er Jensen Y Ne Ом O bat РМ О а А а | values of hi warmer counties, Newel A | the large Wegh-spirad forms м ve Norm see? rowed mer oft! ini Aurwriean styioma o de Sr Very Bn ied ¡0 977. fig ЧАТЫ RU ARS В din torms. In Europe are are veas: vi o) FT. mr eat ore ae в и И pex at tA rai LT Lin rt в pasate tae ров vemo № ‘aes a роет AG YORE Oth ГО wee ‚nokiregrieh touilddé sa ie LUS ant Asia. п ВАУ, ОБИ) вова rola уда | LA TN EU PAUSE We советом ат 4 "Re и их. 323 > 65 EN er Se ite Me (ie Oe! went rent Mates andi: rest ‘heel il riba Ev dee aruba LM TAN O ie СТ А Ora, ar iy 3490 «PUR HONS? Te? ARA CNY AO 7 rg tA et pole, В 1:44 2601 man; : ' 2 ARSTEE = "hoal sity 10 Veste + A Ota АТВ А ‘931 Wits АНТ . wi Ba en Are pe risa Ii ert 6 LE AT o? PO? СИ Y Peto ta eo ALONG оч: DO LY Fit teehee ofthe Sr MAL ON uit ir ARA" irony? pedi FON ine of Walz “iris: rie > \ 2 A a + \ nah vas à, | OPT O Se, И Site Sean TAR HU aS лия MA Fa Hay er. SOAR ar aa et a Pg PR O IA ff Heh ole И Spal jobs ' мым: + a? ta Bt LE 10% мал u een ово, Ai vos: mimar. dra 0 à N er Scan. | мрт ère, Aa ARA ian wo Ув e AR VS Ven pro de TRE ot A Y nest does ar SIACM3I102 Sabie NE VES TI Te Ve AT ado dat vo” wir to вез НОМ vwermwritor) ter A CRIME MY bre СНА bi; irn. I ee Bete hr aa att INOUE 30 i PAL 1 cad) o oe A Areal fü u alt #9965: - оба т . Ct Loro Eu 4 ‘byl Catia’ ty ря ВТ ‘ROW ana e ALTEN? LION AC a4 | NA A rie rite tar еек Yee. GT RST miró рен FA AV, i) retin core roy ee pue D Lo ae A ALLEN east Oe te rue УЗ буса 10 DN DA У 1900 Ly Ai mi sical ег. aan) > < ’n ran м: HANA AA a fs) Stra Penn ye Coy eT) | Г. i SI АА | Lun My er + Ny ные rien yy t MAN TTIENRNTE . es Ÿ ny в cu we tc the st i. С | HILO В № à er wo. I 0% 18H Le AL Mes Drm 110 y Gr) u | у y п ATOM WERDE The een pl DATE 4 SN ( ' ' m A. yarsas 7 ee ary (Ceara tov ) y" Л р ret EIA CEN MN fan ahé Mods are, ии ‚Ei pa ni me set of PAUTA within à faut MALACOLOGIA, 1978, 17(2): 223-239 EVOLUTION AND ADAPTIVE RADIATION OF CER/ON: A REMARKABLY DIVERSE GROUP OF WEST INDIAN LAND SNAILS? 2 David S. Woodruff Department of Biological Sciences, Purdue University, West Lafayette, Indiana 47907, U.S.A. ABSTRACT The pulmonates of the genus Cerion are remarkable for the extreme variability of their shells. The vast array of shell types present in Cuba and the Bahamas, coupled with past typological taxonomic practices, has resulted in the naming of over 600 “species.” The complex patterns of geographic variation are so confusing that several workers have despaired of the possibility of ever applying the biological species concept to this group. The allegedly haphazard distribution of fossil and living shell types has been attributed to the vagaries of hurricane dispersal. Stephen Jay Gould, of Harvard University, and | have been collaborating on a holistic study of the geographic variation, ecology, and evolution of these remarkable snails. We have begun to make sense of this extraordinary situation by detailed mapping of geographic variation in the field and by laboratory studies of multivariate morphometrics and biochemical genetics. So far we have found that the 200 “species” of the northern Bahamas reduce to a single pair of contrasting morphotypes: phenetic and genetic variation is continuous rather than capricious. Furthermore, while hurricanes may have played a role in establishing some distribution patterns, their importance appears to have been overestimated. In the northern Bahamas, we can discern an overall biogeographic pattern that appears to have evolved in situ during the late Cenozoic. As a result of these studies and others involving Cerion from elsewhere in the Bahamas, Cuba, Hispaniola, Puerto Rico, the Dutch Leeward Islands, and the Florida Keys, we anticipate a downward revision of close to two orders of magnitude in the number of valid species. Despite the enormous taxonomic literature on Cerion, we know remarkably little about its natural history or the adaptive significance of the variation in shell morphology. It was widely held that the snails are obligate halophiles, that they are restricted to a zone close to the shore, and that their abundance (up to 212 per square meter) stems from a lack of predation. Field observation over the last few years has shown that these beliefs are unfounded. Long-term ecological studies in the northern Bahamas have been initiated in an attempt to generate some basic information on population size, dispersion, dispersal, growth rates, predation, and mortality. The activities of over 1,500 individually marked snails have been followed periodically since November, 1973. Laboratory studies on the mechanical strength and thermal properties of the different shell types are also being conducted to try to establish the adaptive significance of the observed morphological variation. The results of these studies will be presented, and, in addition, attention will be drawn to a number of other outstanding problems (including Cerion's ability to estivate for protracted periods) Presented by these animals. In a broader context, Cerion will be shown to display one of nature's most impressive displays of morphological variety. While this diversity has been acquired without widespread speciation or extensive genetic differentiation, the group provides prime material for studies of the genesis and maintenance of morphological complexity. INTRODUCTION gradations, that naturalists do not like to rank them as dis- “Those forms which possess in tinct species, are in several re- some considerable degree the spects the most important to character of species, but which us.’’ (Darwin, 1859: 47). are so closely similar to some other forms, or are so closely “A genus in which many such linked to them by intermediate situations seem to occur is the 1This review is No. 10 т a series on The Natural History of Cerion. It is respectfully dedicated to William J. Clench and Ruth D. Turner who first introduced the author to Cerion. 2This research was supported, in part, by grants from the U.S. National Science Foundation to $. J. Gould and the author. (223) 224 WOODRUFF snail genus Cerion, considered “the most difficult genus of pulmonate mollusks to classify”’ (W. J. Clench, in litt.)” (Mayr & Rosen, 1956: 1). The West Indian pulmonate Cerion is celebrated in the evolutionary literature for its phenetic diversity. Intrapopulation varia- tion is rarely extraordinary in extent, but shells of each local population develop dis- tinctive features, and an astonishing range of forms is found in certain parts of Cuba and the southern Bahamas. Differences in shell size, shape, sculpture and coloration have been used to characterize almost 600 species and the evolutionary importance of these remarkable animals has been buried under an all but impenetrable taxonomic thicket. This is to be regretted as recent studies show Cerion to be prime material for the examination of a number of impor- tant biological problems including the rela- tionship between growth and form, the maintenance of genetic and phenetic varia- tion, the genodynamics of hybrid zones and the processes of geographic speciation, and the evolutionary biogeography of spe- cies distributions. In this paper | review the results of a decade’s work оп Cerion by Stephen Jay Gould and myself. Gould, a paleontologist and biometrist, was attracted by Cerion's basic diversity, by its good fossil record in the Pleistocene dunes, and by the pro- pensity of most individuals to display their entire ontogeny in an accessible shell. In contrast, | initially approached Cerion be- cause of its abundance and distribution pattern, characteristics that made it suitable for an experimental study of the relation- ship between the ecological genetics of natural populations and distribution. Our collaboration quickly transcended these personal interests and many of the studies reported below should be attributed to both workers rather than the author alone. While we have learned a great deal about Cerion only a small fraction of our findings have yet been reported. Accordingly, | will describe some of our on-going projects, both to report on our progress (or lack of it) and to indicate problems to which other investigators may hopefully be attracted. This paper, then, constitutes no more than a prolegomenon to the evolution and adap- tive radiation of these remarkable animals. NATURAL HISTORY Snails of the genus Cerion Roding are diverse and conspicuous members of the West Indian land snail fauna. They are found in the Florida Keys, Bahamas, Cuba, Cayman Islands, Hispaniola, Puerto Rico, Virgin Islands and the Dutch Leeward Islands. They are generally restricted to coastal areas where they аге typically found along a narrow strip of open vegeta- tion within 100 m of the shore. Cerions are abundant in their preferred habitat, often reaching densities of 10 adults/m? and at- taining colony sizes of 10% individuals. While the vast majority of cerions are typi- cally found hanging in the plants, to which they seal themselves with a thin epiphragm, a few species live in the leaf litter and sand. Very little is known about their ecology and behavior beyond the casual observation that they spend the greater portion of their adult lives in a state of estivation. Inactive during the day, they are thought to emerge and feed at night, when these areas are actively tenanted by mos- quitos and sandflies, and malacologists have retired from the scene. The most remarkable thing about Cerion is, of course, the great variation in shell morphology (Figs. 1-13). It has become apparent that nearly every local population has a diagnostic size, shape, sculpture and color. Most of the published work on Cerion is devoted to partitioning this varia- tion among nearly 600 species. (Mercifully, Clench (1957) has prepared an authorita- tive catalogue of this vast taxonomic litera- ture.) Plotting the locality records for these taxa produces a crazy-quilt distribution pat- tern for many species (see Gould & Wood- ruff, 1978: fig. 1). There is marked geo- graphic variation with adjacent populations being radically different from one another. Several species found on Long Island in the Bahamas are shown in Figs. 7-13; C. steven- soni and C. fernandina occur within 200 m of one another. Elsewhere in the Bahamas and Cuba different species replace one an- other in irregular fashion along the coasts. This unusual pattern of shell-type (morphotype) distribution is probably real; Cerion is an extremely well-collected group. The record does, however, embody two erroneous factors that must be noted at the outset. First, | can substantiate no case of sympatry between any of the species re- NATURAL HISTORY ОЕ CER/ON 225 FIGS. 1-13. The phenetic diversity of Cerion displayed by selecting average specimens of taxa discussed in the text. The bar represents 1 cm. 1, С. uva, Curagao; 2, С. bendalli, Little Abaco, Bahamas; 3, С. abacoense, Great Abaco, Bahamas; 4, C. glans, New Providence, Bahamas; 5, C. incanum, Big Pine Key, Florida; 6, C. pauli, Great Exuma, Bahamas. Bottom row, all from Long Island, Bahamas; 7, C. caerulescens 8, C. eximeum; 9, C. nudum; 10, C. sp. indet. (? dwarf caerulescens); 11, C. malonei; 12, C. (Umbonis) stevensoni; 13, C. fernandina. ported by earlier workers to coexist. Mayr’s (1963: 398) report for eastern Cuba is compromised by the presence of clear in- termediates in his sample. Other alleged cases arise from the custom of some earlier conchologists to artificially sort their sam- ples into dissimilar types without regard for the natural intrapopulational variation. Still another source of confusion comes when a sample contains both living and dead speci- mens. Crabs and various forces of nature carry dead shells vast distances and result in heterogeneous populations of shells. Er- rors of this type must be borne in mind when examining the literature and older museum collections. The second type of error in the record arises, in part, from the typological systematics practiced by the early Cerion specialists. Dissimilar forms were named as separate species without any regard for their interactions in the field. The truth of the matter is that wherever distinct populations come together in na- ture they appear to be able to interbreed. Thus, while the apparently haphazard dis- tribution of morphotypes in some areas is real, the pattern is the result of a mosaic of contiguously distributed morphotypes rather than the coexistence of different species in different combinations. Until the taxonomic status of these morphotypes is reestablished we can say almost nothing about the specific diversity of different islands or the genus as a whole. The so-called species of Cerion are char- acterized exclusively on the basis of shell characters. Four subgenera and 15 species groups have been proposed (Pilsbry, 1901-1902) on the basis of features which may or may not have phylogenetic signifi- cance. Two subgenera contain single spe- cies: Eostrophia, for C. anodonta from Oligocene deposits in Florida, and Cerion, for the type-species C. uva from Curagao. 226 WOODRUFF The name cerion, derived from the Greek word for a honey-comb, alludes to the resemblance of C. uva to an old-fashioned bee-hive (Fig. 1). The vast majority of living cerions are referred to the subgenera Strophiops and Diacerion depending on the form of the axial and parietal teeth in the aperture of the shell. However, as Pilsbry himself noted (1901-1902: 179), the dis- tinction between these groups breaks down in some forms. The remaining subgenus, Umbonis, is the only species group to have received any recent attention (Clench & Aguayo, 1952). The shells of this group are quite characteristic in their shape, in having spiral sculpture of numerous incised lines, and in agglutinating small sand grains to the outside of the shell. The group includes C. stevensoni from the northeast coast of Long Island (Fig. 12) and 13 other species from the Bahamas and Cuba. A dwarf form, C. turnerae, occurs on Great Inagua and reaches a length of 15mm (see Gould et al., 1974: fig. 1). The anatomy of several species of Cerion has been studied; Richter (1926) and Jaenicke (1933) did some careful work on C. uva from Curagao and C. glans from New Providence in the Bahamas. Unfor- tunately, it is difficult to assess the alleged differences in the digestive, nervous and reproductive systems until comparable studies have been made to establish the extent of intraspecific variation due to the age, behavior and physiological state of the snail. Dissections of some other species (including С. incanum from Florida) are figured by Bartsch (1920) and Pilsbry (1946), but there are numerous points of disagreement and the work will have to be repeated. Burch & Kim (1962) have described the karyotype of C. incanum: the diploid num- ber is 54. Twenty-seven pairs of chromo- somes have also been found recently in C. fernandina by Michael Goldman in my laboratory. Studies of genetic variation of Cerion have added a new dimension to our under- standing of the evolution and adaptive radi- ation of these snails (Woodruff, 1975a). Using electrophoretic techniques (Wood- ruff, 1975b) | have examined variation in more than 20 enzyme systems in several thousand snails from various parts of their range. | have identified more than 30 allozymes or structural gene products and begun to study their distribution in the various morphotypes (Gould et al., 1974; Woodruff, 1975b; Gould & Woodruff, 1978; Woodruff & Gould, in ргер.). Ап analysis of the observed genotype fre- quencies in each sample revealed that Cerion, a facultative hermaphrodite, is ap- parently outbreeding. Cerion populations (large and small, but with one notable exception reported below) studied to date have moderate amounts of genetic varia- bility: mean number of alleles per locus, 1.65-1.70; frequency of loci polymorphic per population, 0.15-0.30; and frequency of heterozygous loci per individual, 0.054-0.128. The systematic and evolu- tionary significance of these findings will be discussed in the second half of this review. As noted previously there were virtually no quantitative data available on the ecology of Cerion. Accordingly, in Novem- ber, 1973, | established two study sites on Great Abaco Island in the northern Bahamas. The sites were chosen to char- acterize high density populations of two morphotypes found throughout the Bahamas. One area (Rocky Point) is in- habited by С. bendalli which has a thin shell which is smooth or finely ribbed and mottled with brown and black (Fig. 2). The other area (Shipwreck) is inhabited by C. abacoense, a species with a heavy white shell with prominent ribs (Fig. 3). All snails at these study sites have been individually marked, in situ, a process which has a very slight effect on dispersal but no discernible effect on growth rates and survival. The populations have been censused repeatedly over a four year period and data are now in hand for over 1600 snails. As the data analyses are continuing much of what is reported here should be regarded as pre- liminary. At the start of these studies snail den- sity at Rocky Point averaged 13 adults per т? (range, 0-79). Similar densities occurred at Shipwreck. Typically over 90% of the adults at both sites are found off the ground attached to stems and leaves. They show a strong preference for certain plant species like the lily, Hymenocallis declinata. The juveniles (typically 10-20% of the pop- ulation) are found on the soil surface be- neath the shallow leaf litter or at the base of clumps of grass or bushes. Mark-release- multiple recapture studies indicate that the neighborhood area is less than 100 т? and that the effective population size (N of Wright, 1946) is approximately 1000 snails. NATURAL HISTORY OF CER/ON 227 Cerions appear to spend the greater part of their adult lives attached to the vegeta- tion in a state of estivation. They are typically encountered sealed (aperture up) by a thin epiphragm to the stems and branches of low bushes or the leaves of monocots. How long they remain in estiva- tion in nature is unknown and probably varies between individuals in response to spatial and temporal changes in the micro- environment. In the laboratory | find that the larger species will estivate for over 24 months if undisturbed. In the field, adults are found moving about at night and dur- ing the day following rain. They have been observed both moving among the leaves of plants and on the ground surface. Diet is still poorly defined. Contrary to Bartsch’s (1920: 6) statement that adults feed on fungal mycelia, other investigators have found vascular plant remains in the feces (Weston, in Mayr & Rosen, 1956; and June Chatfield, National Museum of Wales, in litt.). While cerions do not damage the plants with which they are associated they may graze on decaying plant material. Cerions at Rocky Point and Shipwreck are highly sedentary. Some snails have re- mained on the same plant for 2-3 years. With the assistance of Annette Adams in my laboratory | have devoted considerable attention to dispersal behavior as it is es- sential that we be able to estimate the magnitude of gene flow if we are ever to understand the microevolution of these ani- mals. As expected, adult dispersal patterns are leptokurtic. In the first six months (winter) of my study at Rocky Point the mean detected displacement was 107.7 cm (range, 0-1202 cm, N = 109). In the follow- ing six-month period (summer) the mean displacement was 215.7cm (range 0-2200 cm, N = 221). In each census pe- riod a significant number of adults (11-16%) were recaptured on the plant where they were first encountered. Using Batschelet’s (unpublished) ellipse of inertia technique to compare dispersal in succes- sive census periods | concluded that the adults move more during the wetter sum- mer months than during the winter. Dis- persal is not random with respect to direc- tion; it is clearly constrained by the size and shape of the habitat patch and by the distribution of plants within the patch. At Rocky Point, for example, the study site lies about 30m inland from the high water mark and the snails are distributed more or less continuously in the open vegetation that parallels the beach. The variance of dispersal along an axis parallel with the beach follows a pattern that we might expect for snails making random move- ments on a large habitat patch. (The vari- ance for the first 12 month period is greater than that for either component 6 month period, although it is not the sum of the two, as would be expected for completely random movements.) The move- ment of snails with respect to an axis at right angles to the shore shows a com- pletely different pattern: the 1 year vari- ance is actually less than that for either 6 month period. This indicates that the snails perceive the vegetation change along the beach front (and further inland) and move back into the habitat patch. The statistical interpretation of these data is difficult because several of the parameters are not normally distributed. However, if we make some simplifying (and reasonable) assumptions, it is possible to estimate some very important demographic parameters. If one assumes that generation time at Rocky Point is 5 years (see below) then | estimate the mean distance displaced per generation to be 200cm (range 100-400 cm). The evolutionarily important gene flow distance may then be estimated as the product of the mean distance trav- elled in a generation and the square root of the probability of leaving a deme or neigh- borhood (May et al., 1975). For a high density population of С. bendalli at Rocky Point my preliminary estimate for this parameter is 2.8m. The significance of this low value will be discussed later in the context of the maintenance of clines and narrow hybrid zones. Given its moderate size, Cerion’s abun- dance is remarkable; in its favored habitats it is far commoner than Helix, Cepaea and other land snails that have been subjected to ecological scrutiny. In our field work we have come to designate Cerion as being locally rare, uncommon, common and very common at average densities of 0.1, 1.0, 10 and 100 adults per m?, respectively. It is not unusual to find aggregations of living snails numbering between 100 and 150 individuals on a single bush. In this respect it is not unlike the Mediterranean land snail, Theba pisana. Density of dead shells can be even higher (up to 245 per m7), but this is misleading indicator of natural den- sity and dispersion. The shells accumulate 228 WOODRUFF on the ground surface and are resistant to decomposition. In some areas of the Bahamas these dead shells carpet the ground surface at densities ten times that of the living snails. In their favored habitat cerions are about ten times as abundant as the next most common land snail; usually a member of the genus Hemitrochus (Cepolis). Cerion's abundance has traditionally been attributed to a lack of predation; fires and hurricanes being generally recognized as the main factors determining a colony’s success. The common occurrence of areas with vast numbers of dead snails but very few living snails is cited as evidence for the significant role of natural disasters. Un- doubtedly, natural and man-made brush fires have a devastating effect on Cerion. The impact of a hurricane has yet to be formally established (fortunately my study sites have yet to be perturbed in this manner). Again, the traditional wisdom is that the colony is destroyed and a few fortunate, fertilized individuals survive to colonize new habitats into which they are introduced by the wind or water. Cerion's ability to survive immersion in sea water has been described by Bartsch (1912) and Mayr & Rosen (1956). One case might be cited as evidence that hurricanes may erad- icate some colonies: C. chrysaloides disap- peared from its former habitats on the southwestern coast of Grand Bahama Island following a “direct hit” in the 1930’s (Clench, 1938; Gould & Woodruff, 1978). On the other hand, the continued existence of Cerion on innumerable low lying cays in the Bahamas suggests that some individuals survive and repopulate these environments. High recapture rates of adults at Rocky Point and Shipwreck (80% after 6 months and 40% after 48 months) indicate that adult mortality is low. Observations there, and elsewhere, indicate that cerions do in fact experience significant losses to a num- ber of predators. Land crabs (Gecarcinus lateralis at Rocky Point) take a number of snails on the ground and may provide part of the selective force that causes the snails to estivate just above ground level. Typi- cally, these crabs will carry the snail back to the entrance of their burrow before breaking the shell into small pieces and removing the animal. Small piles of shell fragments, often with the lip intact, may be found around the edge of the sand pile surrounding an occupied crab hole. Rats also eat Cerion; one side of the shell is characteristically sheared off leaving both the lip and protoconch intact. Still another predator, possibly a bird, is implicated by the discovery in the pine forest of Grand Bahama Island of an “anvil'* surrounded by irregularly broken shell fragments. Finally, | note that cerions interact with still an- other organism which breaks a small round hole in the shell. It is not yet known whether the hole is made from within or without; it is subsequently repaired. As these anecdotes indicate, ме still know very little about the ecology and behavior of Cerion. The deficiencies are particularly apparent in the area of repro- ductive behavior where we badly need to establish whether there are specific behav- ioral traits that could serve as pre-mating isolating mechanisms between the various morphotypes. A knowledge of Cerion’s mating system and reproductive strategies is also required if we are to understand the adaptive radiation of these snails. Unfor- tunately, repeated attempts to induce courtship and mating in the laboratory have failed so we are still limited to some opportunistically gathered observations. | have seen mating once in the field, follow- ing an afternoon cloudburst in June on New Providence. Copulation is protracted, taking at least 2 hours, and is not recipro- cal. Pairs of snails lie on the ground surface with their apertures adpressed. Several mat- ing pairs were returned to the laboratory where additional matings between some of these pairs were observed. About 3 weeks after the original matings some snails began to excavate small egg-chambers at a depth of 2-4cm in moist sand. Batches of 4-7 eggs were layed in some of these chambers (details will be presented elsewhere: Wood- ruff, in preparation). The egg capsule is initially colorless but gradually turns opaque and white, suggesting a calcareous matrix. (Capsular fragments have been sent to Alex Tompa, University of Michigan, for analysis.) Juveniles began to hatch about 20 days later and emerged from the sand about a week after that. Studies of the growth and development of these snails (N = 212) are being conducted by Sarah Burgess in my laboratory. Not only will these observations provide us with informa- tion about growth but, as we have the parents of each group of juveniles, we hope to learn something about the heritability of shell traits too. NATURAL HISTORY OF CER/ON 229 Additional information about growth rates is coming from studies of marked juveniles at Rocky Point and Shipwreck. Unfortunately, the data on juveniles are not as extensive as that for adults as the former experience quite high mortality rates. Juveniles grow very erratically: in one 6 month period a snail might lay down a single rib while in the next 6 month period, perhaps two whorls. Although | suspect reproduction and hatching occur during the summer the subsequent growth pattern is not strongly seasonal. My pre- liminary conclusion is that individuals at these localities are typically 3 years old when they secrete the lip characteristic of the adult shell. This conclusion is sup- ported by laboratory growth experiments conducted by Ida Thompson (Princeton University). To calculate the generation time (estimated above to be 5 years) we need to know the age of sexual maturity. On this subject we have no information. However, as some snails live at least a decade after laying down the adult lip | have conservatively assumed that they are reproducing successfully in their second summer of adult life. | tentatively conclude that cerions are outcrossing, iteroparous hermaphrodites with generation times of 4-5 years and a longevity of perhaps twice that duration. RECENT STUDIES WHICH FOCUS ON THE SYSTEMATIC SIGNIFICANCE OF VARIATION IN CER/ON Bimini Islands Ernst Mayr's detailed study of the cerions of the Bimini Islands in the Ba- hamas constitutes the first notable contri- bution to the evolution of Cerion con- ducted in a modern framework (Mayr & Rosen, 1956). Their conclusions are worth noting here as they indicate the intracta- bility of the Cerion phenomenon and set the stage for our own work. Mayr con- ducted a very careful survey of the cerions of the six small islands in this group and found geographic variation to be pro- nounced but irregular. While each colony had its own diagnostic features, three major morphotypes were recognizable. These cor- responded to the species C. /erneri, C. biminiense and C. pillsburyi. Mayr & Rosen stress, however, that even though super- ficially each morphotype appears to be a separate species, each is allopatric, and adjacent colonies show signs of gene ex- change. Mayr & Rosen argued that these facts were best explained by assuming that the pattern was controlled by two antagonistic tendencies: a high degree of sedentariness and infrequent long-distance dispersal by hurricanes. The checkerboard pattern of morphotype distribution arises as a result of multiple colonizations of each island, extensive hybridization of types secondarily brought into contact with one another, and a steady extermination of colonies as a result of fires and hurricanes. They inter- preted the readiness with which the three morphotypes hybridized, in spite of pro- nounced morphological differentiation, as evidence that reproductive isolation is not easily acquired in this genus. In contrast, they агдие that shell characters are highly plastic and evolve rapidly during the early stages of colonization, adapting the snails to their new environment. They wrote “In view of the fact that these snails are her- maphrodites it is possible and probable that many if not most colonies are founded by a single fertilized adult. Two factors tend to promote rapid divergence in such col- onies. Comparatively high inbreeding among the early generations derived from the founder individual exposes homozygous genotypes more often to selection than in outbred populations. Also the impact of the new environment may lead to a drastic modification of the contents of the gene pool in the new population and conse- quently in the phenotype’’ (Mayr & Rosen, 1956: 43). They emphasize that this great morphological plasticity is an adaptation to the pioneer habitat in which Cerion lives: a habitat which changes continuously as a result of eustasy, plant succession, and natrual catastrophes. On the one hand Mayr & Rosen saw Cerion as an actively speciating group and on the other, as an extreme example of reticulate evolution on the _ intraspecific level. They concluded that the very char- acteristics (variation in shell morphology) which have adapted these snails so superbly to the habitat in which they live made it impossible to classify them in terms of the conventional categories of species and sub- species. 230 WOODRUFF Dutch Leeward Islands The type-species, C. uva, was described by Linnaeus in 1758. It is restricted to the islands of Curacao, Aruba and Bonaire which lie off the coast of Venezuela and some 800 km south of the rest of the range of the genus. In marked contrast to the situation in Cuba and the Bahamas only a single variable species was recognized by the early conchologists. Gould resolved to begin his study of the genus with this single, isolated taxon before confronting the more complex situations elsewhere. He was Originally attracted to Cerion because it provided all the features he could enu- merate for an ideal biometrical subject. First, adequate sample sizes are almost never a problem. Second, Сегоп shows a clear and sharp transition between the protoconch and later accretionary growth. This provides a criterion for the numbering of whorls as a reference for standardized measures at various stages of ontogeny. Finally, and unlike most snails, Cerion changes its direction of growth near the end of ontogeny and secretes a single, permanent adult lip. The bugbear of most biometrical studies of molluscs lies in the difficulty (if not impossibility) of sorting ontogenetic from static adult variation. This can be done unambiguously in Cerion and we can assess the true standardized variance of adult shells. C. uva offered one other attraction: it is among the world’s best known land snails from a biometrical point of view. It had been the subject of two major studies (Baker, 1924; Hummelinck, 1940; see also De Vries, 1974). Gould began by reanalyz- ing the published data for a series of 102 samples well distributed over the three islands. In contrast to earlier workers he applied a variety of multivariate statistical techniques to elucidate patterns of intra- and intersample variation (Gould, 1969). He found an excellent correspondence be- tween the data of Baker and Hummelinck on interregional diversity;samples sorted into four groups corresponding to Aruba, Bonaire, eastern and western Curagao. Diversity pat- terns did not change significantly during the 16 years between these studies. Next, Gould compared the early samples with a series of 69 modern samples he made himself (see Gould, 1971: unpublished data for mono- graph in preparation). Again, the patterns do not change over the 50 year period. Variation within each island or geo- graphic area appears to be influenced by environmental parameters. Surprisingly, the Cerion collected from volcanic areas on Curacao have larger shells than those found on limestone. This anomaly is probably not related to the substrate itself but to the microenvironmental features usually associ- ated with it. On Curacao the limestone areas are typically very dry, windswept and poorly vegetated; the volcanic areas are more sheltered and have a more mesic vegetation. Baker (1924) was the first to notice an inverse relationship between the number of whorls (and shell height) and the degree of exposure to the dry trade winds. He suggested that the evaporative effect of the wind reduced the snails’ ac- tivity period and hence its adult size. Gould found additional support for this hypothesis: plotting mean shell height for 15 samples he collected from similar micro- habitats on the first terrace along the northern coast of Curagao he found that shells were largest where the coast runs due east-west and the trade winds do not blow directly on shore. Elsewhere, where the coast runs north-south the shells are smaller (Gould, 1971: fig. 46). Coincidentally, the area where the coast runs east-west is also the area of volcanic soil. Still other evi- dence for environmental selection for shell size comes from fossil samples from Aruba and Curaçao. Gould (1971) found that shells from three middens on Curagao dat- ing to about 4000 y BP were larger than any living on this island today. Larger shells were also found in a midden on Aruba dated at 1500 y BP. Gould argued that this suggests that the island’s climate was moister during these periods than at present. If these interpretations are correct then the phenotypic variation within this single species is continuous rather than capricious and adaptive rather than adventi- tious. Little Bahama Bank Gould and | began our collaborative field work on the northernmost Bahamian bank: the Little Bahama Bank consisting of Grand Bahama, Little Abaco, Great Abaco and their fringing islands. This area was blessedly spared from visits by Maynard, one of the most exuberant conchological splitters to focus on this genus, and re- ported diversity was only 12 species. These NATURAL HISTORY OF CERION 231 names, plotted in the supposed areas of their occurrence, produce the “crazy-quilt'” distribution pattern traditionally associated with Cerion (Gould & Woodruff, 1978: fig. 1). After several trips to these islands it became clear, however, that the dozen or so names refer to two variable and imper- fectly isolated taxa (Gould & Woodruff, 1978). These semispecies, С. bendalli (smooth or finely ribbed and mottled shell) and C. abacoense (heavily ribbed white shell) are shown in Figs. 2, 3). This taxonomic simplification was not, however, immediately obvious to us as both species show considerable geographic variation. On our first survey trip, for example, we discovered a local situation where a population of Cerion bears a тог- phology sufficiently distinctive to warrant recognition as a new species by all previous criteria. Our resolution of the taxonomic status of this aberrant population, from Pongo Carpet on the northeast coast of Great Abaco, provided us with an oppor- tunity to test our combination of multi- variate morphometric and _ biochemical genetic techniques (Gould et al., 1974). This population is semi-isolated by man- groves on a narrow coastal strip and is in contact with other populations only in the north. Its morphology is highly distinctive (see Gould et al., 1974: fig. 3) but its patterns of covariation (as revealed by fac- tor analysis) cannot be distinguished from those of С. bendalli within whose range it occurs. These patterns of covariation among the 19 variables measured were, however, impressively different from those which characterize C. abacoense which is found 70 km away at the southern end of the island. Canonical analysis revealed a multivariate cline extending from Pongo Carpet north towards the area of potential contact with typical С. bendalli. Levels of morphological variability did not differ among samples; those of intermediate mor- phology show no increase (or decrease) in variability or other signs of hybridity. Elec- trophoresis showed the Pongo Carpet Cerion to be genetically indistinguishable from С. bendalli at the 18 structural gene loci surveyed. | could find no evidence for any genetic anomaly within the 12 popula- tions (7 from Pongo Carpet) sampled from northern Great Abaco. We concluded that the Pongo Carpet Сегоп are only a well- marked geographic variant within С. bendalli. \f this situation is any model for a revision of the genus, we predict that the number of species of Cerion may be re- duced almost 100-fold. We have subsequently extended our study to the Cerion of the entire bank and have been able to resolve a pattern involv- ing two variable semispecies (Gould & Woodruff, 1978). It turns out that C. abacoense is restricted to those present-day coasts which lie close to the edge of the Little Bahama Bank. We will refer to it as being a ‘‘bank edge” species. In contrast, С. bendalli is a “bank interior’ species, found in the interior of the islands and on coasts washed by the very shallow waters covering the bank today. In southern Great Abaco, the geography of the present-day coast relative to the edge of the bank is such that there are several places where the ranges of these 2 morphotypes come into contact and hybridization occurs. To re- solve the patterns of variation in each morphotype and to investigate the nature of their interaction we studied variation in 20 shell characters and 28 genetic loci in over 50 samples from across the bank. Morphometrically, the two morphotypes may be sorted unambiguously by factor analysis; 3 factor axes encompass over 96% of the information. All samples from the area where the two taxa meet plot in the intermediate phenetic field: the hybrids have intermediate phenotypes. Near Rocky Point the zone of allopatric hybridization (sensu Woodruff, 1973) is 1km wide and there is a gradual transition in mean mor- phology across the zone with no increase in intrasample variability. My study of allozyme variation for the same snails revealed very little concordance between the genetic and phenetic patterns. All samples of both C. bendalli and C. abacoense are markedly similar: Nei’s genetic identity (l) for 820 pairwise com- parisons averaged 0.98. No unique marker gene characterizes any region or mor- photype, though characteristic frequencies of certain variable alleles clearly separate Grand Bahamian from Abaconian Cerion in a statistical manner. Despite this overall genetic similarity the pattern of allozyme variation supports our decision to treat the Little Bahama Bank Сегоп as two semi- species. Although hybrids showed по т- crease in morphometric variability they showed significantly more genetic vari- ability (both within and between popula- tions) than snails collected elsewhere. They 232 WOODRUFF are also polymorphic for alleles not de- tected in either adjacent ‘‘parental’’ popula- tion. | have now resampled this hybrid zone and a more detailed discussion of its significance will be presented elsewhere (Woodruff & Gould, in prep.). It should be clear from the above discus- sion that there is no evidence that the various colonies of Cerion on the Little Bahama Bank were founded by single hur- ricane-transported waifs. The very high genetic similarity among the various рор- ulations sampled suggest that hurricanes have not played a significant role in the recent evolution of these snails. The situa- tion may be quite different on the tiny offshore cays surrounding the main islands; unfortunately | have not yet been able to study them genetically. | have, however, studied variation in populations of C. bendalli from Snake and Tuggy Cays which are adjacent to Great Abaco (Woodruff, 1975b). These populations show the same amount and type of genetic variation as the adjacent populations from the main island. Again, there is no evidence that these populations were derived from single founding individuals. For the Little Bahama Bank Cerion we reject the notion that the pattern is a ‘‘crazy-quilt’’ due to hurricane- dispersed morphotypes. Instead we con- clude that the pattern is quite coherent and evolved at a time when sea levels were lower and the populations on the various islands were in full genetic contact with each other. Great Bahama Bank Following our study of the Cerion т the northernmost Bahamian group of islands we turned our attention to the far more complex situation on the Great Bahama Bank. Here, there are a number of large islands (Andros, New Providence, Eleuthera, Cat, Exuma and Long), numer- ous small cays and several hundred de- scribed species. In the last five years Gould and | have undertaken exploratory surveys on all the main islands (collecting both living and fossil Cerion) and examined ma- terial in the collections of the major U.S. museums. Our studies on three of the large islands have reached the stage where | can announce some of our preliminary findings; detailed reports will be published else- where. New Providence (with the city of Nassau) has been scoured for Cerion for over a century and 82 species have been described from this relatively small island (207 km?). Yet we quickly realized that the situation here was very similar to the pattern we had elucidated on the Little Bahama Bank. There are two imperfectly separated entities: a ribby abacoense-like form (called С. glans on this island, see Fig. 4) and a mottled bendalli-like form (whose correct name we have yet to establish). Again, the ribby morphotype is a “bank edge’’ species found on the western edge of the island and on the offshore cays. The mottled morphotype inhabits the interior of the island and the eastern and southeast- ern coasts on the “bank interior’’ side of the island. The zone of interaction between these morphotypes is, however, different from that described on Abaco. While multi- variate clines mark the east-west transition across this island the hybrid zone is char- acterized by novel phenotypes and by un- usual intrapopulation variation in shell size. There is no smooth continuity of pheno- type, but rather a host of mixed pheno- types. This is borne out in a Q-mode factor analysis: the hybrid population is character- ized by shells which secrete the adult lip abruptly, without undergoing the character- istic changes in the axis of coiling typical of all other cerions. Genetically, the two morphotypes on New Providence are very similar, as they were on Abaco. On New Providence, however, the transitional popu- lations are not distinguished by any genetic anomalies. A paper in which we will sub- sume 80 taxa and describe the nature of the variation in these two semispecies and the significance of their interaction is now being prepared. We have made one survey trip to the Exumas and found, once again, that despite the plethora of available names there are basically two contrasting morphotypes pres- ent. There are numerous varieties of a bendalli-like mottled form (for which C. eximeum may be the oldest available name) on Great and Little Exuma and on the innumerable cays in the Ехита chain. There is a ribby abacoense-like form on some of the cays which lie close to the edge of the bank. Only two forms depart dramatically from one or other of these morphotypes. One, С. pauli (Fig. 6) from Great Exuma, matures at less than half the typical size of the mottled forms but has 10-12 whorls, or 2-3 more than normal. We NATURAL HISTORY ОЕ CERION 233 have no reason to regard it as anything more than a strongly dwarfed population of the mottled morphotype. It is also atypical in its habits, being found in piles of rotting vegetation near the mangroves on the “bank interior’’ coast. The other aberrant type is an isolated population of a member of the subgenus Umbonis on Great Guana Cay. Museum specimens indicate that this population, which was presumably derived from Cuba, is hybridizing with the local mottled forms. On Long Island, at the southeastern edge of the Great Bahama Bank, we found that our simplifying generalization about “bank edge”” and “bank interior’’ morpho- types was complicated by the incursion of several additional morphotypes. A few of the species described from Long Island are shown in Figs. 7-13. In the course of 3 field trips to this island we have mapped the distribution of these forms and uncov- ered by far the most interesting situation that we have yet encountered in Cerion. First, we discovered that two basic mor- photypes were present: a mottled form, C. eximeum, occurs along the western “bank interior’’ coast and a strongly ribbed form, C. caerulescens, occurs on the eastern coast adjacent to the edge of the bank. In addi- tion we found five other distinctive mor- photypes along parts of the hilly east coast. These taxa replace one another along the coast and we have now located 12 narrow hybrid zones between various combinations of these forms. Particularly spectacular zones involve the large, white, smooth- shelled С. fernandina, the squat, white, smooth-shelled С. malonei, and the dis- tinctive member of the subgenus Umbonis, C. stevensoni (Fig. 12). Defined on the basis of shell morphology these hybrid zones are quite narrow (some are only 100m wide), and may or may not be associated with marked changes in abun- dance or habitat. While the various species involved in these interactions are very simi- lar to each other genetically | found that at least two of the hybrid zones are char- acterized by significantly greater fluctua- tions in interpopulation allele frequencies and by the presence of novel genotypes absent from adjacent “parental popula- tions.” Michael Goldman, in my laboratory, has carefully checked intersample variation in adjacent populations away from one of these hybrid zones (fernandina-stevensoni) to ensure that our preliminary findings are not simply due to more intensive sampling of the intermediate populations. We see the elucidation of the genodynamics of these hybrid zones as important to the under- standing of the process of evolution of Cerion. п particular we will attempt to establish whether the zones are of primary ог secondary origin or whether they are a mixed group. It strikes us that marked geographic differentiation and parapatric speciation are quite possible in a group like Cerion. \nterestingly, parts of the complex distribution pattern on Long Island are relatively old; fossils indicate that some of these species have been in place for at least 110,000 years. Florida Keys A single endemic species, С. incanum, occurs in the Florida Keys and north as far as Key Biscayne in Miami. It is a medium- sized, typically smooth-shelled species and has little geographic variation (Fig. 5). | have recently discovered that, unlike а! other Cerion examined, С. incanum is large- ly invariant genetically. All the populations sampled in the Keys (sensu stricto) to date are monogenic; samples from Key Biscayne were polymorphic at an esterase locus. This contrasts markedly with the situation in Cerion from the Bahamas, Puerto Rico and Curacao which are typically polymorphic at 4-6 loci. In this respect C. incanum is more like North American populations of Rumina decollata than any other pulmonate gastropod described to date (Selander & Kaufman, 1973, 1975a). We will use this observation to standardize the nomenclature for the allozymes detected in Cerion (Woodruff & Burgess, in prepara- tion.) Henceforth, different allozymes will be compared directly to those in C. incanum and designated by their quantitative relative mobility. Gould and | made a second exciting discovery in the Florida Keys. In 1911 Paul Bartsch began a series of transplantation experiments aimed at showing that snails moved to a new environment would rapidly evolve to resemble the local (and pre- sumably optimal) Cerion morphotype. He transplanted several samples of Bahamian cerions to the Dry Tortugas and other Keys (Bartsch, 1920). Fires, hurricanes and other setbacks marred the experiments and the transplants bred true to their ancestral phenotype for as many generations as he 234 WOODRUFF could follow. He concluded that pheno- typic variation in Cerion is not under strict environmental control. As some of his transplants hybridized with the resident C. incanum, and as the hybrids were “епог- mously variable”” (a phenomenon Bartsch attributed to mutation), he further con- cluded that the “‘crazy-quilt” pattern of geographic variation in this genus was ad- “entitious rather than adaptive (Bartsch, 1949). While Bartsch’s experiments were inadequate to assess the process of local adaptation, and while his ideas of ‘‘mutat- ing hybrids” as the source of Cerion's diversity, are not supported by modern genetic theory, we are indebted to him for initiating these experiments nearly 60 years ago. We have discovered that the descend- ents of one of his transplantations still flourish today in a restricted area at the site of the original introduction. The in- troduced species hybridizes with С. incanum and the hybrids continue to show a range of variant phenotypes and a general level of variation far higher than that seen in homospecific populations. Bartsch docu- mented the early stages of his experiments in great detail and it has been possible to re-collect from the original Bahamian source population and confirm their identity genetically. We will describe this interesting experiment and discuss its evolu- tionary implication in the near future. Isolated Bahamian banks We have had limited field experience with the cerions on a number of the smaller island banks: San Salvador, Inagua, Rum and Conception. Rum Cay and Con- ception Island lie on small, isolated banks to the north and east of Long Island. Each is inhabited by a single, variable species of the ribby morphotype. Kathleen Ligare and | are establishing the genetic and phenetic relationship between these snails and the similar morphotypes on Long Island from which they might have been derived. Cuba The north and eastern coasts of Cuba are perhaps the center of Cerion’s remark- able diversity. We have begun a revision of this large island’s rich fauna; 147 “species and subspecies’’ according to Clench (1957) and Jaume (1975). We have examined all specimens in the great collection at the Museum of Comparative Zoology and pre- pared distribution maps of the purported taxa. The resolution of the resulting com- plex patterns will require additional field work. In the 1950’s Ernst Mayr collected Cerion in Cuba and his distribution map of forms on the Banes Peninsula on the north- eastern coast has done much to draw atten- tion to the genus and its problems (Mayr, 1963: 398-399, 1969: 17, 1970: 33). In this area Mayr found 7 highly distinct morphotypes replacing each other geo- graphically along a 50 km stretch of coast. Within four zones of contact presenting no ecological barriers to effective gene flow, he found populations which he interpreted as the hybrid products of secondary inter- gradation. Lynne Galler and Gould (in prep.) have been studying the narrow tran- sition zone between two of the most diver- gent morphotypes: C. moralesi and C. geophilus. Their biometrical studies show no increased variability in the geographical- ly intermediate samples. Several univariate and a multivariate cline cover the zone which is less than 1km wide. They find that the large morphological differences be- tween these taxa may arise from a small alteration in the rate of shell widening during the early phase of post-embryonic growth. The interplay of this simple event with the complex allometries of normal ontogeny produces the large adult dif- ferences. We must now confront the fascinating question: can any Cerion be transformed into any other Бу simple heterochronous changes in early ontogeny? Hispaniola, Puerto Rico and the Virgin Islands Gould and Paull (1977) completed a study of Cerion from the eastern end of its range: Hispaniola, Mona Is., Puerto Rico, and Necker and Anegada in the Virgin Islands. Here 11 names were available for a basic morphology that all students of Cerion have recognized as unique to this area (Pilsbry, 1901-1902). They performed a canonical analysis of morphological varia- tion in 23 samples from throughout this area and found that these samples are arranged along the most significant dis- criminator (the first axis: 59% of all infor- mation) in perfect geographic order. The morphological gradient runs from egg- shaped, finely and copiously ribbed shells NATURAL HISTORY OF CERION 235 with very obtuse apices (Virgin Islands) to more cylindrical, apically-pointed shells with fewer, stronger ribs (Hispaniola). In the light of this clinal pattern they reduce all available names to a single species, С. striatellum. DISCUSSION Prior to the commencement of our work three major attempts were made to inter- pret Cerion's complex diversity. Plate (1906, 1907) argued that the variation in shell morphology was adaptive. His inter- pretation was in a Lamarckian mode, how- ever, and was based on very limited per- sonal experience with the animals and much misinformation about areas he had not seen. Bartsch (1920, 1949) concluded that the pattern was not adaptive and resulted from chance events (colony extinc- tion and long-distance dispersal), evolution in isolation, and subsequent diversification among ‘‘mutating hybrids.’’ Mayr (1963; and Mayr & Rosen, 1956) interpreted the variation as adaptive but also argued that stochastic events play an important role in the development of the overall pattern. Mayr regarded each colony as an evolution- ary experiment whose ultimate fate is in- determinate. He interpreted the ‘‘crazy- quilt”” distribution pattern in terms of local extinction and recolonization with sub- sequent secondary contact between con- trasting morphotypes dispersed primarily by storms. He repeatedly cites Cerion as the classic example, in animals, of the acquisition of morphological differences without reproductive isolation (Mayr, 1963, 1969, 1970). In the past few years Gould and | have personally confronted about one third of the alleged diversity in Cerion. Our combi- nation of laboratory studies employing multivariate morphometric and biochemical genetic techniques is proving successful. Given time we have every reason to believe that Cerion's taxonomic overburden will be removed and the evolution and adaptive radiation of these snails can be exposed for direct investigation. While it should be clear that our bias is towards an interpretation involving both history and adaptation it is clear that we have not yet reached the point where sweeping generalizations are possible. It turns out that the earlier work- ers were partly right and partly wrong; the analogy to the situation in Cepaea—a prob- lem with too many solutions (Jones et al., 1977)—is striking. In the interim, how are we to regard these markedly different populations? Are they ecotypic races, morphospecies, or bio- logical species? At present we think Cerion may comprise a number of variable and polytypic semispecies. The actual number of these taxa is not yet known but we agree with Clench’s (1957) opinion that perhaps only 20% of the described species are valid. | emphasize that we still know nothing about reproductive isolating mech- anisms in Cerion; statements to the effect that reproductive isolating mechanisms are not easily acquired in this genus are based solely on the observation of shell types that are intermediate between various pairs of morphotypes. Until we identify the po- tential pre- and post-mating isolating mech- anisms and learn more about the reproduc- tive anatomy and behavior of various cerions, we simply cannot comment on the significance of the alleged interspecific dif- ferences. For example, we still do not know whether selfing plays any role in Cerion's mating system (as it does in early reproductive life in Partula taeniata (Mur- ray & Clarke, 1976)). Working in my labo- ratory, Daniel Chung has taken on some of these fascinating problems. The lack of marked genetic differences between the various cerions studied has little bearing on speciation per se. Twelve years of biochemical genetics of many or- ganisms provides little evidence for exten- sive reorganization of gene pools during speciation (Throckmorton, 1977). There is every indication that changes in a few loci are sufficient for speciation and that geo- graphic variation rather than genetic revolu- tion may be the critical prerequisite. As | can demonstrate no major differ- entiation in structural genes among the most widely divergent of Cerion’s morpho- types we are forced to look elsewhere for the underlying genetic determinants. Gould (1977) has argued that complex differences in form can often be traced to simple differences in developmental rates ex- pressed during ontogeny. Since these rates are probably controlled by regulatory genes (King & Wilson, 1975) it is not surprising that allozymic (structural gene products) variation is small. The possibility that minor developmental changes will translate into major differences in adult morphology 236 WOODRUFF is strongly enhanced in Cerion by the com- plex allometries that characterize growth: particularly the 3 divergent phases of juven- ile triangularity, mid-growth “barrelling,” and adult recurvature. Our vision is to reduce all this diversity to a simple system of ontogenetic growth gradients and their quantitative alteration. We have also tried to develop a “dynamic” approach to de- scribing shell variation and choose to work with patterns of covariation rather than static adult morphology. Our initial experi- ences have been most promising and it is pleasing to note that the raw data pre- sented in our first paper (Gould et al., 1974) has already been used as a basis for two subsequent studies (Sokal, 1976; Schueler € Rising, 1976). To give our enormous data sets even greater validity Kathleen Ligare has recently completed a study of the measurement errors associated with each morphometric trait. So far | have not discussed the possible adaptive significance of the variation seen in Cerion. | proceed into the difficult field of functional morphology with Darwin's remarks (quoted by Cain, 1954) on the danger of believing apparently trivial char- acters to be of no functional importance and in no way due to natural selection firmly in mind. It is not at all surprising that earlier workers, confronted with the puzzling and apparently random distribu- tion of morphotypes, gave up and decided that the pattern was due to drift and other random processes. It is only from our experience with Cerion in nature that we feel emboldened to ask questions about the adaptive significance of variation in the shells and other features. We have reached the stage where we can speculate about the possible adaptive sig- nificance of variation in shell size, shape, sculpturing and color. We proceed by look- ing for obvious correlations between mor- phometric traits and environmental para- meters. Overall shell size, for example, ap- pears to be correlated with humidity and the degree of shelter from the wind on Curacao. Unfortunately, this trend does not seem to explain patterns on the Great Bahama Bank where even greater inter- population variation occurs on single islands. Dwarfing occurs in some popula- tions of most of the main morphotype groups. | note that juvenile and adult shells are quite different in shape and wonder if this might be correlated with differences in microhabitat, shell function, and the mechanics of crawling (see Cain, 1977; and papers by Cain and Linsley in this sympo- sium). In Cerion the spire index (height/ max. diameter) of juvenile shells is 0.5-0.7; in adults the index is typically above 3.0 and can be as high as 7.0. Juveniles are typically found on the ground surface be- neath the leaf litter and are usually ori- ented apex up. In contrast, the adults are invariably found hanging on plants apex down. Lip size appears to be correlated with habit and habitat: small lipped shells are often associated with grass or leaf litter in humid microhabitats; cerions with large recurved lips are associated with drier areas which are exposed to the prevailing wind and often have rocky substrates. It is pos- sible that the greater surface area of an aperture with a large lip give the shells (which is fastened to the substrate or vege- tation by a thin epiphragm) greater stabil- ity in windy situations. | can not comment yet on the possible significance of variation in the apertural teeth which are present in most cerions in both juvenile and adult phases. Shell sculpture or ornamentation is another variable character of unknown adaptive significance. While in some areas ribbing appears to be correlated with hu- midity and with the presence of a calcare- ous substrate, in other areas this generaliza- tion does not hold. John Quensen, in my laboratory, is examining Vermeij’s (1975) suggestion that ribbing is an adaptive re- sponse to predator pressure in snails since prominent ribs confine the predator’s crushing force to the thickest part of the shell. His preliminary results indicate that overall shell size (weight and height) is more important than ribbing in determining a cerion’s ability to withstand crushing forces applied generally along the sides of the shell. He is now repeating these experi- ments using artificial crab chelae to apply forces at right angles to the shell’s long axis; a more realistic design in view of our findings at Rocky Point. Quensen has also been considering the possible adaptive sig- nificance of shell pigmentation. Mottled shells are initially hard to find as they hang on bush stems and blades of grass in the dappled sunlight and shadow, a clear case of disruptive coloration to our eyes (Gould & Woodruff, 1978: fig. 4). In contrast, white shells are conspicuous in the vegeta- tion (see Bartsch, 1968: fig. 24). Quensen found that when shells are compared be- NATURAL HISTORY OF CERION 237 neath a heat lamp in the laboratory the interior of a mottled shell averages at least 1°C warmer than the interior of an unpig- mented shell. In the Bahamas he found the effect of differential heating to be even greater: up to 4°C differences in Novem- ber, the actual difference being propor- tional to the air temperature. As the upper lethal temperature for Cerion is about 52.5 C it is quite possible that a pigmented shell exposed to the summer sun risks thermal death. The occurrence of unpig- mented shells in exposed coastal situations may be a thermoregulatory adaptation. We cannot even begin to discuss the possible adaptive significance of variation in characters other than those associated with the shell. Alleged interspecific variation in the radula and reproductive system are not adequately documented. Similarly, we can not yet account for the distribution of the snails themselves. Their absence from Jamaica and the Lesser Antilles is puzzling. Even within the areas where they occur we have no adequate theory to account for Cerion's microdistribution. Even the most basic assertions that cerions are halophiles (Mayr, 1963) or calciphiles (Clench, 1957) are contradicted by the occurrence of snails up to 15km from the coast on Grand Bahama and in volcanic (non-limestone) areas on Curagao. Cerions are remarkably like some vari- able plants, in which spectacular examples of highly localized, specially adapted eco- types which replace one another over dis- tances of a few meters are well known (Bradshaw, 1972). | see considerable prom- ise in attempting to reinterpret some of the patterns within the context of a model of parapatric differentiation rather than the allopatric model advocated by Mayr (1963). This is necessary because recent theoretical work on clines (Endler, 1977; Woodruff, 1978) indicates that the tradi- tional criteria used to distinguish between primary and secondary intergradation are inappropriate. We must carefully examine the possibility of explaining some aspects of the Cerion pattern in terms of con- temporary selection gradients and dispersal patterns. A continuously distributed popu- lation of 10°-10° snails cannot be con- sidered as though it were a single random mating population; Cerion is apparently broken up into small demes containing approximately 10° individuals. Each deme, even those in a more extensive population, occupies an area that is large relative to the gene flow distance and close evolutionary “tracking” of microenvironmental hetero- geneity is a real possibility. | am now beginning to look for evidence of this at the study sites on Abaco. The measurement of natural selection operating on various traits in nature is quite difficult; our initial approach involves looking for asymmetry in the various biometric traits. | am also ex- tending my investigation of gene flow to assess its magnitude in low (rather than high) density populations. Cerion provides prime material for the experimental investi- gation of natural clines and hybrid zones. Land snails are well suited to studies of evolution and adaptive radiation (Clarke et al., in press). The lessons of Cepaea (re- viewed by Jones et al., 1977), Theba pisana (Hickson, 1972; Nevo & Bar, 1976) and Helix pomatia (Pollard, 1975; Jarvinen et al.. 1976) on the role of natural selection and other agents in the microevolution of natural populations are particularly note- worthy. Studies of the variation of Partula in the Society Islands resulted in significant contributions to the theory of clines and parapatric differentiation (Clarke, 1968; Clarke & Murray, 1969; Murray, 1972). These studies, and others which focus on other organisms with low vagility, cast in- creasing doubt on the universality of the allopatric model of speciation (Bush, 1975; Endler, 1977; Woodruff, 1978). Finally, | note that recent work on genic variability and breeding systems in Helix aspersa and Rumina decollata (Selander & Hudson, 1976; Selander & Kaufman, 1973; 1975a; 1975b) have implications for evolutionary ecology far beyond the confines of mala- cology. It is our hope that Cerion will take its place among this small group of land snails that are of major significance to biology generally. REFERENCES CITED BAKER, H. B., 1924, Land and freshwater mol- luscs of the Dutch Leeward Islands. Occa- sional Papers of the Museum of Zoology, University of Michigan, 152: 1-158. BARTSCH, P., 1912, Planting Bahama cerions upon the Florida Keys. Yearbook of the Carnegie Institute of Washington, 11: 129-131. BARTSCH, P., 1920, Experiments in the breeding of cerions. Papers of the Department of Marine Biology, Carnegie Institute of Washing- ton, 282: 1-55. 238 WOODRUFF BARTSCH, P., 1949, Molluscan genetics: the role of hybridization, mutation, isolation, fixation and speciation in relation to taxonomy. Amer- ican Malacological Union News Bulletin and Annual Report, 1948: 2-4. BARTSCH, P., 1968, Mollusks. (Republication of Smithsonian Institution Series, 10(3), 1934.) Dover, New York, 111 p. BRADSHAW, A. D., 1972, Some of the evolu- tionary consequences of being a plant. Evo/u- tionary Biology, 5: 25-47. BURCH, J. B. & KIM, D. C., 1962, Chromosomes of Cerion incanum (Binney) (Mollusca: Gastropoda: Stylommatophora). Bulletin of the National Institutes of Health (Korea), 5: 181-186. BUSH, G. L., 1975, Models of animal speciation. Annual Review of Ecology and Systematics, 6: 339-364. CAIN, A. J., 1954, Animal species and their evolution. Hutchinson, London, 190 p. CAIN, A. J., 1977, Variation in the spire index of some coiled gastropod shells, and its evolu- tionary significance. Philosophical Transactions of the Royal Society of London, ser. B, 277: 377-428. CLARKE, B., 1968, Balanced polymorphism and regional differentiation in land snails. In: DRAKE, Е. T. (Ed.), Evolution and Environ- ment. Yale University Press, New Haven, p. 351-368. CLARKE, B., ARTHUR, W., HORSLEY, D. Т. & PARKIN, D. T., in press, Genetic variation and natural selection in pulmonate molluscs. (Manuscript for publication in: FRETTER, V. & PEAKE, J. (Eds.), Pulmonates. 2: System- atics, evolution and ecology. Academic Press, London). CLARKE, В. & MURRAY, J., 1969, Ecological genetics and speciation in land snails of the genus Partula. Biological Journal of the Linnaean Society, 1: 3142. CLENCH, W. J., 1938, Land and freshwater mollusks of Grand Bahama and the Abaco Islands, Bahama Islands. Memorias de la Sociedad Cubana de Historia Natural, Museo Poey, Universidad de la Habana, 12: 303-333. CLENCH, W. J., 1957, A catalogue of the Cerionidae (Mollusca-Pulmonata). Bulletin of the Museum of Comparative Zoology, 116: 121-169. CLENCH, W. J. & AGUAYO, C. G., 1952, The scalarinum species complex (Umbonis) in the genus Cerion. Occasional Papers on Mollusks, Harvard University, 1: 413-440. DARWIN, C., 1859, On the origin of species... John Murray, London. (Facsimile of the first edition published by Harvard University Press, Cambridge, 1964.) DE VRIES, W., 1974, Caribbean land molluscs: notes on Cerionidae. Studies on the Fauna of Curagao and other Caribbean Islands, 45: 81-112. ENDLER, J. A., 1977, Geographic variation, speciation and clines. Monographs in Popula- tion Biology No. 10, Princeton University Press, Princeton, 246 p. GALLER, L. & GOULD, S. J., in prep., Continu- ous phenetic transformation in a zone of interaction between two "species’’ of Cerion in Cuba: the reduction of complex form to simple generating factors in ontogeny. GOULD, S. J., 1969, Character variation in two land snails from the Dutch Leeward Islands: geography, environment, and evolution. Sys- tematic Zoology, 18: 185-200. GOULD, S. J., 1971, The paleontology and evo- lution of Cerion, Il: age and fauna of Indian shell middens on Curagao and Aruba. Breviora, 372: 1-26. GOULD, S. J., 1977, Ontogeny and phylogeny. Belknap Press of Harvard University Press, Cambridge, 506 p. GOULD, $. J... & PAULL, .G., 1977, «Natural history of Cerion. VII. Geographic variation of Cerion (Mollusca: Pulmonata) from the east- ern end of its range (Hispaniola to the Virgin Islands): coherent patterns and taxonomic simplification. Breviora, 445: 1-24. GOULD, 5: J. 87 WOODRUEE ADS 9738 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. GOULD, S. J., WOODRUFF, D. S. & MARTIN, J. P., 1974, Genetics and morphometrics of Cerion at Pongo Carpet: a new systematic approach to this enigmatic land snail. System- atic Zoology, 23: 518-535. HICKSON, T. G. L., 1972, A possible case of genetic drift in colonies of the land _ snail Theba pisana. Heredity, 29: 177-190. HUMMELINCK, P. W., 1940, Mollusks of the genera Cerion and Tudora. Studies of the Fauna of Curacao, Aruba, Bonaire, and the Venezuelan Islands, 2: 43-82. JAENICKE, J. H., 1933, Untersuchungen zur Anatomie und Verschiedenartigkeit der Cerion Arten der Bahamas als Beitrag zum Problem der Artentstehung. Jenaische Zeitschrift fiir .. Medizin und Naturwissenschaft, 68: 277-402. JARVINEN, O., SISULA, H., VARVIO-AHO, S-L. & SALMINEN, P., 1976, Genic variation in isolated marginal populations of the Roman snail, Helix pomatia L. Hereditas, 82: 101-110. JAUME, M. L., 1975, Catalogo de los moluscos terrestres Cubanos del genero Cerion (Mol- lusca-Pulmonata-Ceriidae). (Con una bibli- ografia general.) Catalogo de la Fauna Cubana, 37. Ciencias, ser. 4, 51: 1-47. JONES, J. S., LEITH, B. H. & RAWLINGS, P., 1977, Polymorphism in Cepaea: a problem with too many solutions? Annual Review of Ecology and Systematics, 8: 109-143. KING, М. С. & WILSON, A. С., 1975, Evolution at two levels in humans and chimpanzees. Science, 188: 107-116. MAY, R. M., ENDLER, J. A. & MCMURTIE, R. E., 1975, Gene frequency clines in the pres- ence of selection opposed by gene flow. American Naturalist, 109: 659-676. MAYR, E., 1963, Animal species and evolution. Belknap Press of Harvard University Press, Cambridge, 797 p. MAYR, E., 1969, Populations, species, and evolu- tion. Belknap Press of Harvard University Press, Cambridge, 453 p. MAYR, E., 1970, Principles of Systematic Zool- ogy. McGraw-Hill, New York, 428 p. MAYR, E. & ROSEN, C. B., 1956, Geographic variation and hybridization in populations of Bahama snails (Cerion). American Museum of Natural History Novitates, 1806: 148. NATURAL HISTORY OF CER/ON 239 MURRAY, J., 1972, Genetic diversity and natural selection. Oliver & Boyd, Edinburgh, 128 p. MURRAY, J. & CLARKE, B., 1976, Supergenes in polymorphic land snails. 1. Partula taeniata. Heredity, 37: 253-269. NEVO, E. & BAR, Z., 1976, Natural selection and genetic polymorphisms along climatic gradients. /n KARLIN, S. & NEVO, E. (Eds.), Population genetics and evolution. Academic Press, New York, p. 159-184. PILSBRY, H. A., 1901-1902, Family Cerionidae. Manual of Conchology, ser. 2, Pulmonata, Academy of Natural Sciences of Philadelphia, 14: 174-286. PILSBRY, Н. A., 1946, Land Mollusca of North America. Academy of Natural Sciences of Philadelphia, Monographs 3, 2(1): 158-169. PLATE, L., 1906, Die Artbildung bei den Cerion- Landschnecken der Bahamas. Verhand/ungen Deutschen Zoologischen Gesellschaft, Leipzig, 16: 127-138, 1 pl. PLATE, L., 1907, Die Variabilitat und die Artbildung nach dem Prinzip geographischer Formenketten bei den Cerion-Landschnecken der Bahama-Inseln. Archiv fur Rassen- und Gesellschafts-Biologie, 4: 433-470, 581-614. POLLARD, E., 1975, Differences in shell thick- ness in adult Helix pomatia L. from a number of localities in southern England. Oecologia, 21: 85-92. RICHTER, K., 1926, Zur Anatomie von Cerion glans Küster der Bahamas-Inseln. Jenaische Zeitschrift für Medizin und Naturwissenschaft, 62: 277-342. SCHUEFER- Е. _\М & RISING, 23. D:, 1976, Phenetic evidence of natural hybridization. Systematic Zoology, 25: 283-289. SELANDER, R. K. & HUDSON, R. O., 1976, Animal population structure under close т- breeding: the land snail Rumina in southern France. American Naturalist, 110: 695-718. SELANDER, R. K. & KAUFMAN, D. W., 1973, Self fertilization and genetic population struc- ture in a colonizing land snail. Proceedings of the National Academy of Science (United States), 70: 1186-1190. SELANDER, R. K. & KAUFMAN, D. W., 1975a, Genetic population structure and breeding systems. /n: MARKERT, С. L. (Ed.), /so- zymes, 4. Academic Press, New York, p. 27-48. SELANDER, В. К. & KAUFMAN, D. W., 1975b, Genetic structure of populations of the brown snail (Helix aspersa). |. Microgeographic varia- tion. Evolution, 29: 385-401. SOKAL, В. R., 1976, The Kluge-Kerfoot phenomenon reexamined. American Naturalist, 110: 1077-1091. THROCKMORTON, L. H., 1977, Drosophila systematics and biochemical evolution. Annual Review of Ecology and Systematics, 8: 235-254. VERMEIL, G. J., 1975, Marine faunal dominance and molluscan shell form. Evolution, 28: 656-664. WOODRUFF, D. S., 1973, Natural hybridization and hybrid zones. Systematic Zoology, 22: 213-218. WOODRUFF, D. S., 1975a, A new approach to the systematics and ecology of the genus Cerion. Malacological Review, 8: 128. WOODRUFF, О. S., 19756, Natural history of Cerion. V. Allozyme variation and genic hetero- zygosity in the Bahamian pulmonate Cerion bendalli. Malacological Review, 8: 47-55. WOODRUFF, D. S., 1978, Review of Geographic variation, speciation and clines by J. A. Endler. Science, 199: 1329-1330. WOODRUFF, D. S. & GOULD, S. J., in prep., Natural history of Cerion. XI. Genetics and morphometrics of the interaction between two semispecies of Cerion (Pulmonata) on Abaco, Bahamas. WRIGHT, S., 1946, Isolation by distance under diverse systems of mating. Genetics, 31: 39-59. A К ach UA ++ ned i vas Pa nt — pry) 64% a A В №, 1977 an * Sind Milan SE Ne Mare hand. a er We Aava ME of =. in,» 77 or sore mh eter ¡FIUAQDOW NA ab or N. о a 1 lo nen LE DIMTE кий, hes ARUN Ane ban noi CHAM). © NEA SHO MING hatin ort ni izogys Е eT A а i av AT Brant донный А. льва У ФА AROMA че ИА + send Chr > Pith morose à HADAS. (OH сиси A A TA TA om Aid nO. DD AN ES Cot paretd. norte в’ м) gar edn й en. ‚erh oo lefnde night nari legs mme w gene nd поро u wen a ee Aud. пола rat, од АА re sg opté PA AB Ta. УЖ ЕМС! Ww ] AA 31 hen ДАЛА Oo’ 10d! Da A ’ Pea hiba ,.. LOA ry Lis i Syd u dad Сала ce Пион Pug UT Юле. (¿vrai de Le tuba. A стен м р мм р фе Mica RA Ir td lA di 4 MV BE т 4 tes (anys CN ON, Decpuse.a! Pape: Merval Library ? 440 ки: N ТЕ. Car fie ty flyer Muy: re tiie TT swf ПА ot w A Стаи и, VO Dé VOLES, м. М. Tarioter ed мобы QUA i ag Vy mn т Gras Da 4158 A : Y 21-182 Caimi FF | А h "a : wig tas martian Y $ A ES TIR и Y в La 1 м 1 LIA "Lees 2 YN & 4 3 ve Lar» d Sa A e” Y L 7 de 3$ u m _ ig? Де pré po р МИ A? Ст fad cis & as u > er ON Y pone ir re clar tot ad Pta о aver LOREL У notée ‘бити | art: ОА" SOR ом ам aci MAR Ай не ha id niet Brn kait : M D Wit sor, A О vn a Risch CN 108%, de " Me nerel рых. Movie, ВО, Vo A ee Lo LENA Teen f ca” mere, end ann. | mice rf wha pumas Di Adel MONA onen А | Water vert\ felt Ager, к, ры NS Ln la II HCO MOL NA Nes CR) A D AMAR D 8 LO ME, М. 1. 1875, Chimie вл tartes Cubatros. del genera ware atra ie) (Onn, - D TE Cato 00 te PD in PI Al; 4 or YM ye E 4. & FUN “mu: In Cad ina Mady MUDONES Annual у yy aul Syrien E 108 que AE? | & eit} 4 ь Г) ran 7 = vient ong $ Scans VER: (072116, > rm. М. ERDLE 4. AD U er, Okie cha A в fwetio eae ВИ“ nes an Momracay, 30. СВОИ = ' LO Anima! pe gg, ‘Lat whan es tienne. Un Mir LA LT" via a TA A u fraiin® brasse 01 Maid on. Larter, 4G) в м. Es TITO, Ро a’ Sy in & yaw bhi Nuw Mas 279 р. awe ‘witli sad РЕ УИ >» eotetetiony: iota aan Они. Aner ian ere ~ a MALACOLOGIA, 1978, 17(2): 241-315 ECOLOGY OF THE PUERTO RICAN CAMAENID TREE-SNAILS Harold Heatwole! and Audry Heatwole2 ABSTRACT This paper compares the ecologies of the Puerto Rican camaenid snails. Interspecific differences in environmental tolerances and responses, and in life history patterns were assessed in relation to geographic distributional patterns. Finally, these aspects were viewed from the standpoint of adaptive strategy. There are five native species of Camaenidae in Puerto Rico, two of Caracolus (C. marginella, С. carocollus) and three of Polydontes (P. acutangula, P. luquillensis, P. lima). These occur in species pairs with С. marginella and P. lima predominating in the lowlands and С. carocollus and P. /uquillensis in the uplands. P. acutangula occupies a special upland habitat (leafy parts of vegetation). С. carocollus is the most eurytopic and occurs in both upland and lowland habitats, where it has different ecological characteristics. All five species of camaenid snails in Puerto Rico are primarily nocturnal. Juvenile snails tend to secrete themselves in crevices under objects, whereas adults more often are on tree trunks, or in the case of Polydontes acutangula, the leaves of trees or hedges. In wet montane forest a greater proportion of adult Caraco/us carocollus are on the trunks than is true in the hotter, drier lowland forest. More snails occur on large rather than small trees. There does not seem to be any selection of tree species other than by size; however, trees with scaly or flaking bark are not often used whereas trees with an abundance of shelter sites attract more snails. С. carocollus is not common where slopes are steep. Body temperature of animals in the shade is controlled by heat exchanges with the substrate and to a lesser extent with the air. In direct sunlight, body temperature rises as a result of absorption of radiant energy. As a result, and perhaps because of evaporative cooling and lags in temperature change in the body, air and substrate temperatures are not Precise predictors of snail body temperature. In Polydontes, the falling point (temperature at which a snail drops off a vertical substrate) was correlated with the habitat occupied by the species. Both species of Caraco/us had similar falling points. Neither lethal temperature nor thermal safety margin showed any consistent correlation with taxonomic or ecological groupings. Falling point is not dependent on body size, whereas lethal limit, and hence thermal safety margin are; juveniles have lower lethal limits and safety margins than adults. Survival of the adult С carocollus deprived of food is more than twice as long at 20°C than at 30°C. In Caracolus marginella survival at the two temperatures does not differ greatly and is similar to that of С. carocollus at 30°C. Juvenile С. carocollus has lower survival than adults at both temperatures tested. Most С. carocollus carry water in the mantle cavity. The amount carried is less when conditions are dry. However, in adults the water content of the body itself remains the same during dry periods and the animals continue to feed normally. С. carocollus from the lowlands has a high mortality in the very small juveniles with few animals surviving to maturity. Dry periods increase the relative mortality among small juveniles. In the moister uplands, juvenile mortality is less, most individuals surviving until adulthood. Other upland species, including one from a different family have survivorship characteristics similar to those of upland С. caroco//us. Lowland species tend to be between the extremes represented by upland and lowland populations of С. carocollus but have considerable juvenile mortality. Mating in upland species and in upland populations of С. carocollus seems to occur over a prolonged period centering on the dry season whereas in lowland species and lowland populations of С. carocollus mating tends to occur in the wetter parts of the year. Individuals stay in breeding condition for at least two months and during that time mate with more than one other snail. A given snail may breed in at least two consecutive seasons. In P. acutangula most egg-laying begins about November and continues till February or March; eggs take at least 1% months to develop. Sporadic oviposition may occur at other times of year. It deposits its eggs at a variety of epigean sites and on the surface of the ground under objects. Oviposition periods are not known for other species. After the mating peak, the albumen gland of С. carocollus begins to develop and reaches the peak of its cycle in June, after which the reproductive animals probably retreat to secluded refugia where they mature their eggs and oviposit. Embryonic development and hatching were not observed, but the timing of the rest of the cycle suggests that they 1Department of Zoology, University of New England, Armidale, N.S.W., 2351, Australia. 2Armidale Technical College, Armidale, N.S.W., 2350, Australia. (241) 242 HEATWOLE AND HEATWOLE probably occur before the onset of the subsequent late-winter and early-spring ‘‘dry’’ season. At Loiza Aldea, the peak of the albumen gland cycle is about a month later but the general seasonal pattern is similar. Growth of С. carocollus at El Yunque occurs primarily between March and August; the season is somewhat shorter at Е! Verde. Adulthood is reached in 3-6 years. P. acutangula and Р. luquillensis grow at all seasons and reach maturity after one year and two years respectively. С. carocollus lives at least up to ten years and probably longer. Longevity seems to be shorter in Р. acutangula and P. /uquillensis. A variety of foods are eaten by these camaenid species. The diet consists primarily of wood, bark, seeds and leaves of vascular plants, and of diatoms and other unicellular algae. Occasionally animal material and bryophytes are also eaten. There appear to be interspecific differences in diet as well as differences between localities within a given species. Snails tend to use a restricted area during their inactive periods (home site range [hsr] ). Juveniles set up hsr's well before maturity and the area of their hsr's does not differ significantly from those of adults from the same region. Size of mean Asr varies greatly between different study areas for a given species. A minimum of seven captures is required for accurately ascertaining the size of the Asr. Within the Asr, particular home sites were favored by individual snails. Hsr’s of different individuals overlapped greatly. С. carocollus maintained a rather stable size-structure throughout the study whereas there were temporal changes in that of Po/ydontes /uquillensis. At El Yunque the population of С. carocollus was increasing in density in late 1962 and into mid-1963 after which it declined. Densities were higher around the building than in the forest; both these areas had higher densities than the El Verde region. INTRODUCTION Selection can operate upon a population through climatic, edaphic, biotic factors or combinations thereof. There are different ways of responding to such forces, i.e. alternative adaptive modes are available. For example, a population can become adapted to extreme environmental heat by broadening its physiological temperature tolerances (withstanding heat) or behavior- ally by selecting only the cooler micro- habitats and changing its diel or seasonal pattern of activity (avoiding heat). Which of various alternative modes of adaptation that will occur depends on the genetic back- ground upon which selection is operating, as well as upon purely stochastic processes. The constancy of selective forces (dura- tion of a particular environmental complex) and the predictability of short term envi- ronmental changes can influence the repro- ductive pattern. For example, a species can have regular seasonal breeding periods con- trolled by internal rhythms, or reproduc- tion may be opportunistic and triggered by immediate weather conditions. A species may be long-lived and produce only a few young each year, or be short-lived and produce many young at a given reproduc- tive season; these patterns have environ- mental correlates (Tinkle et al., 1970). Matthews (1976) has reviewed general strategies which could be expected to be successful in different kinds of environ- ments. In ephemeral environments, the most successful strategy is to have good dispersal powers (to reach a new site when the old one changes) and to quickly exploit new areas by a rapid natural increase result- ing from high fecundity and short genera- tion time (r strategist). In stable and rela- tively equable environments an optimum strategy would be to have lower reproduc- tive rates and less mobility, but to develop more effective long term persistence in a given site through better defence against predators, superior competitive ability under prevailing conditions, and food specialization (K strategist). Finally, in pre- dictable but physically harsh environments, species that can survive physical rigors may become adapted through specialization (be- yond-K strategist). Ultimately, the mosaic of selective forces over an environmentally diverse area, interacting with the spectrum of possible responses by organisms, leads to adaptive radiations. Such radiations in turn can be expressed in biogeographic and ecological terms. One way of studying an adaptive radia- tion then, is to examine the present distri- bution of a cluster of related taxa, define their niches, habitat selection and biotic interactions, measure their tolerances of physical environmental factors, and com- pare their demographic and reproductive strategies—in short to conduct a compara- tive ecological study. One can then begin to assess the factors presently limiting their spatial distribution and evaluate the selec- tive forces operating upon them now and in the past. It is important to study eury- topic species in several different environ- ments in order to ascertain their ecological flexibility. PUERTO RICAN CAMAENID ECOLOGY 243 The need for such a broad comparative approach prompted the present study. Puerto Rico was especially suited as a study area because of the diverse array of habitats, topography and climate which oc- cur in a small area. The island is roughly 160 by 55km, yet has environments rang- ing from Montane Rain Forest to xeric scrub. The snail family Camaenidae was selected because (1) it contained a number of large, conspicuous, easily studied species occupying a broad spectrum of local habi- tats and geographic areas, (2) some species appeared to be highly specialized but others quite eurytopic, and (3) the geo- graphic distributions of the Puerto Rican species had been previously outlined (Van der Schalie, 1948). The organization of the present paper is unusual and requires some explanation. We have discussed our findings and their sig- nificance in six sections. The general reader can grasp the major results and concepts and gain an overall view of the study by reading only these chapters. Following the text are two appendices. Appendix 1 describes the study areas and Appendix 2 presents the detailed results, methods, tabular and graphic material and statistical treatments. When reference to tables or figures is made in the text, the specialist reader will want to turn to Ap- pendix 2 for documentation of the points being discussed, and may wish to read the description of the result accompanying the illustrative material in the appendix. THE GENERAL WORLD OF THE CAMAENIDAE The purpose of this section is to provide the reader with a basis for distinguishing camaenids from other land snails, and to give a general distribution of world-wide camaenids, with notes on general ecological information. Finally, we focus on the five native species of camaenids in Puerto Rico, indicating their probable phylogenetic rela- tionships, describing shell features, species- pairs based on shell features, and the gen- eral association of species of different shell morphologies. The Camaenidae (Pleurodontidae) The camaenids, like other stylommato- phoran pulmonates are hermaphroditic land snails. During copulation each member of the pair inserts its penis into the partner and reciprocal insemination is effected via a spermatophore. Morphologically the family is distinguished by: helices without a dart apparatus; penis continued in an epiphallus and a flagellum (the latter sometimes ves- tigial or wanting); spermathecal duct not branched; ovotestis a single mass of alveoli buried in, or closely appressed to the inside of the liver (Wurtz, 1955). There are about 40 genera (94 genera and subgenera), most of which contain only one or a few species. Only about 12 genera can be considered large. Most taxa are known only from the Recent although fossils are known from the Pleistocene of South Australia, the Pliocene and Pleisto- cene of Caribbean Islands, the Upper Mio- cene of Florida and a possible member of the family from the Paleocene or Eocene of south central United States (Zilch, 1959-1960). As one would expect from such a taxon- omically diverse family (only three of the 76 families of pulmonate snails have larger numbers of genera and subgenera; Te, 1976) there are a variety of habitats oc- cupied and niches filled. However, there is a prononounced tendency toward arboreali- ty (Burch, personal communication) and large size throughout the family. The dis- tribution of the extant camaenids includes southern and eastern Asia, New Guinea, Australia, Melanesia, Central America, South America and the Caribbean Islands. They do not now occur in Africa, the Palearctic or North America. Te (1976) lists 16 different patterns of distribution of pulmonate families and groups them into major categories of Old World-New World, Old World, New World, Australian Pacific, and Southern. He lists the Camaenidae in the first of these broad categories—with a South and East Asia-Australian Pacific- Middle and South America pattern (pattern IV). This is the only family included in this pattern and thus the distribution of camaenids is unique among the pulmonates. The Puerto Rican camaenids There are five native species of camaenid snails inhabiting Puerto Rico; Caracolus marginella (Gmelin), Caracolus carocollus (Linnaeus), Polydontes та (Еегиззас), Polydontes acutangula (Burrow), and Poly- dontes luquillensis (Shuttleworth) (Van der 244 HEATWOLE AND HEATWOLE FIG. 1. The Puerto Rican camaenid snails. A. Polydontes luquillensis; В, Caracolus carocollus; С, Polydontes acutangula; D. Caracolus marginella; E, Polydontes lima on tree trunk near Rio Piedras. Schalie, 1948; Zilch, 1959-1960; Aguayo, 1966). In addition, the introduced Cuban species Zachrysia auricoma havanensis Pilsbry occurs in the dry coastal limestones between Arecibo and Toa Alta (Van der Schalie, 1948); only the native species are considered in the present paper. Fig. 1 shows the appearance and relative sizes of Puerto Rican camaenids and Table 1 provides data on sizes, color and shell morphology; distribution patterns are por- trayed in Fig. 2. The vegetation zones of Fig. 2 are after Little & Wadsworth (1964). They can be ranked in a descend- ing order of wet to dry and cool to hot in the following order: Upper Luquillo Forest, Lower Luquillo Forest, Upper Cordillera Forest, Lower Cordillera Forest, Moist Coastal and Limestone Forests, Dry Coastal and Limestone Forests. In general terms the eastern highlands are the wettest and coolest and the southern coast the hottest and driest. The Puerto Rican camaenids did not adaptively radiate in situ from a single ancestral stem; rather early members of a number of related lineages probably reached the island independently. Conse- quently, the pattern of their evolutionary development was not a diverging into vari- ous niches and habitats from a single stock, but ecological adjustment through inter- action of various taxa to different, though similar genetic entities. Wurtz (1955) recog- nized four major taxonomic complexes within the American Camaenidae, Laby- rinthus, Caracolus, Pleurodonte, and Poly- dontes with Zachrysia. Bishop (in prepara- tion), on the basis of the presently avail- able anatomical evidence, suggests that Pleurodonte, Polydontes and Zachrysia form a single monophyletic lineage, while there is an unresolved trichotomy between this group, Labyrinthus and Caracolus. Labyrinthus is isolated in South America. Caracolus has a fossil representative in the Oligocene of Nebraska (Bishop, in prepara- tion), and Pleurodonte has a probable rela- ne FIG. 2. The distribution of Puerto Rican camaenid snails in relation to climax forest types. Distribu- tional data for snails from Van der Schalie (1948) and the author’s collections. Forest types after Little & Wadsworth (1964). Dashed lines = Upper Luquillo Forest; open area = Lower Luquillo Forest; cross-hatched areas = Upper Cordillera Forest; horizontal lines = Lower Cordillera Forest; Hatched from upper right to lower left = Moist Coastal forest; vertical lines = Moist Limestone Forest; hatched upper left to lower right = Dry Coastal Forest; vertically and horizontally hatched = Dry Limestone Forest. A. Location of the study areas. В. Distribution of С. carocollus (dots) and С. marginella (circles). Localities with both species indicated by half closed dots. С. Distribution of P. /uguillensis (circles) and P. lima (dots). D. Distribution of P. acutangula (dots). The boundary lines enclosing the Lower Cordillera and Lower Luquillo forests coincides very closely to the 155 т contour line and the 25°C isotherm. PUERTO RICAN CAMAENID ECOLOGY = ra LOIZA ALDEA A TCO > | = | 25 DR < Vig WARE À TTL CRE р => N > i > ©, 27 № — LEE > reese =f DO fd VO Ve BDO CSSS LS SS мя 245 246 HEATWOLE AND HEATWOLE TABLE 1. Shell morphology of adult Puerto Rican camaenid snails. Shell diameter No. of (mm) Species Form whorls xX Range Caracolus Low spire 5-6 56.9 44.0-65.01 carocollus Caracolus Low spire 4-5 34.8 31.0-37.4 marginella Polydontes High spire 4-5 34.9 28.0-38.02 luquillensis Polydontes High spire 4-5 28.0 18.5-30.13 lima Polydontes Flat above, 2-3 44.6 37.1-49.14 acutangula rounded below Shell weight Aperture di- (9) _ameter(mm) shell color x Range x Range and texture 20.8 19.9-26.8 25.7 21.5-27.6 Dark brown, smooth 13.4 11.4-15.0 14.6 13.0-16.8 Pale brown with red- dish brown stripe, smooth 22.7 21.2-24.2 15.2 11.2-17.1 Dark brown, smooth 16.7 14.1-24.4 12.3 11.5-13.6 Cream to pale brown, usually sculptured 17.5 15.9-19.1 23.9 22.1-25.7 White nearly through- out, smooth Tin the size range 44-50 mm, 7.5% were adults; in the size range > 50 mm, 84.1% were adult; the largest juvenile captured was 56 mm. Van der Schalie (1948) on the basis of 313 specimens from various parts of the island gives the range as 42-59 mm with most grouped between 44 and 55 mm. Zilch (1959-60) lists 27-65 mm. 2Zilch (1959-60) gives a very different range (35-70 mm) which is probably in error. 3Zilch (1959-60) lists 18-30 mm and Van der Schalie (1948) 19-32 mm. 4Zilch (1959-60) lists 30-50 mm. tive, Pleurodontites in the Miocene of Florida. Ancestral Caraco/us and Р/еиго- donte reached the West Indies and Poly- dontes and Zachrysia have developed there from antecedents to Pleurodonte. Pleuro- donte is now restricted to Jamaica and the lesser Antilles and has become extinct on the other Greater Antilles, where an ex- tensive radiation of its derivative, Ро/у- dontes, has occurred. A number of distinctive lineages within Polydontes can be recognized and have been given subgeneric rank (Wurtz, 1955). The Puerto Rican species are each assigned to a different subgenus, Р. acutangula to Parthenia, P. lima to Granodomus and P. luquillensis to Luquilla. Granodomus and Luquilla share an advanced feature of the penial complex, namely that there is a caecum on the flagellum. Parthenia is dis- tributed on Hispaniola and Puerto Rico, Luquilla is endemic to Puerto Rico, while Granodomus extends from Puerto Rico to the Virgin Islands. Parthenia is closer to the ancestral stock, Luqui//a is a montane rain- forest derivative, whereas Granodomus is better adapted to aridity and has a cor- respondingly wider distribution. The biogeographic and ecologic outcome of the interaction of these genetically simi- lar (confamilial) stocks in the varied envi- ronments of Puerto Rico is the subject of the following parts of this paper. Several subspecies have been described for C. carocollus, C. marginella and P. lima and are listed by Aguayo (1966). However, Van der Schalie (1948) has shown that in P. lima at a given locality one encounters the entire size range for the species, and both heavily sculptured and smooth indi- viduals occur together; the proportion of large, coarsely sculptured individuals de- creases with increasing altitude, but not in a way conducive to separating out sub- specifically named populations. Similarly, he found considerable variation in size in С. carocollus but no geographical cor- relates. We do not have data relevant to polymorphism in size or shell sculpturing at the infraspecific level and cannot make subspecific nomenclatural assessments. Con- sequently, although we follow the generic and specific nomenclature of the most re- cent author (Aguayo, 1966), we follow Van der Schalie (1948) in not employing trinomials for any of the Puerto Rican camaenids. All of the Puerto Rican species are PUERTO RICAN CAMAENID ECOLOGY 247 dextrally coiled and large. All have brown or gray bodies and can pull completely into the shell, except P. acutangu/a which has a bright lemon-yellow body with a brown margin to the foot and cannot completely retract. The species can be additionally distinguished on the basis of shell shape, color and dimensions (Table 1). Morphologically there are three groups, arranged as two species-pairs (each with a large and a smaller species) and a single species. These groupings are not necessarily indicative of closeness of genetic relation- ships (see above). One species-pair contains the two flat, round species, С. carocollus (large) and С. marginella (smaller); the sec- ond pair consists of the taller-spired, dome- shaped species, P. /uquillensis (large) and P. lima (smaller). Р. acutangula with its rather unusual shape is in a group by itself. These species are shown in Fig. 1. Distribution of the native Puerto Rican species The distribution from our data and that published by Van der Schalie (1948) is shown in relation to the major vegetation types in Fig. 2. In each of the morphologi- cal species-pairs One member tends to in- habit the cooler, more mesic areas and the other the more xeric, warm ones. The geographical separation is most distinct in the /ima-luquillensis species-pair. P. /uquillensis has a very restricted distribu- tion, being known only in the wettest, upland rainforests at the eastern end of the island (Fig. 2). By contrast P. /ima is relatively widespread in the various moist to xeric forests but is absent from the wet, cool forests and from the arid, hot regions. There are altitudinal correlates as well, Р. luquillensis occurring only in the higher mountains, Р. /ima from moderate eleva- tions down to coastal localities. As far as is known the two species are completely allo- patric and thus do not presently compete. P. acutangula is found in the wettest montane localities in Puerto Rico and at scattered localities in other mesic forests in the eastern third of the island. One locality is an exception (moist limestone forests on the northern coastal plain in the western third of the island; Fig. 2). C. carocollus and C. marginella although largely separated altitudinally and by forest type do show considerable geographic over- lap. С. carocollus is eurytopic, but primari- ly from moist areas, particularly in the uplands. It occurs in the wettest montane rain forest, in all of the mesic forest types including coastal ones and even in a few scattered localities from the drier vegeta- tion of the southern coast. С. marginella is almost entirely a coastal lowland form, being known only from one (roadside) lo- cality above 150m. It is most widespread on the northern coastal plain, being known only in a few localities on the more arid southern one. Thus, in the wet upland only C. carocollus is represented, but on the coastal plains both species occur and are recorded in the same locality by Van der Schalie (1948). However, the extent of overlap is not as great as would appear. In most lowland localities where С. carocollus is abundant, С. marginella is relatively un- common and vice versa. С. carocollus in- habits the moister pockets of habitat, and C. marginella the drier ones. The dry southern coast is nearly devoid of camaenids of all species. It is the hottest and most arid part of the island and per- haps exceeds the capacity of adaptation for this group. The few localities in which camaenids are known may represent pockets of unusually moist or cool habitat. It should also be mentioned that the boundaries of the vegetation types are not as precise as a line on the map might suggest; also the superimposition of Van der Schalie’s localities on the vegetation map involves a certain amount of plotting error. Consequently, undue significance should not be attached to dots that fall on or very near boundary lines. In summary, P. /uquillensis occurs т very wet, cool rain forests and is restricted to the higher altitudes of the Luquillo Mountains. P. acutangula is also restricted to the wet and mesic forests of the uplands but has a wider distribution. С. carocollus is eurytopic and widespread, but is most common in the cool, moist uplands and in the moister areas of the coastal habitats. P. lima is widespread in mesic to moderately dry areas at moderate elevations and down to the coastal plains. С. marginella is pri- marily on the north coastal plain in the drier habitats. The southern coastal plain is probably too hot or dry 10 support camaenids except in a few localities. In the moist uplands P. /uquillensis, P. acutangula and С. carocollus occur together. In the lowlands P. /ima, C. marginella and C. carocollus occur sympatrically, though the 248 HEATWOLE AND HEATWOLE latter may be partially separated from the others by habitat. Since for each species-pair the largest member is primarily from the cooler moist uplands and the smaller member from the warmer drier lowlands, it is tempting to speculate that moist, cool conditions select for larger size or that a warm dry environ- ment selects for small size. For the two most eurytopic species, the one that is most successful in the uplands (C. caro- collus) is large and the one that is most successful in the lowlands is small (Р. /ima). However, the variation within Р. /ima does not support this view, as large-sized indi- viduals are less common at higher than at lower elevations (Van der Schalie, 1948). The general characteristics of the Puerto Rican camaenids have been outlined above and their distributions described. The next section focuses on the finer details of their environmental relations, i.e. their microdis- tribution, their use of space and time and their maintenance activities. To accomplish these ends four study areas were selected that collectively included all five species. Details of these areas are given in Appendix Ake In Table 2 we present the 18 categories of data collected for the five species and the relative quantity of each type of data for each species. We have more data for some species than for others. For example, we have considerable data for 15 or 18 categories for С. carocollus and only 6 of 18 for C. marginella. DEFINING THE HABITAT AND NICHE There are various ways syntopic species can avoid competition. Niche separation can be achieved via temporal or spatial segregation and by use of different re- sources such as food, shelter, oviposition sites or combinations of the above. Conse- quently, if one wishes to ascertain the way species have evolutionarily adjusted to other species occupying the same general area, comparisons of activity cycle, habitat selection and use, food habits, and defense against other organisms are essential. Con- versely, the niche breadth of a given species can be appreciated by looking at these attributes in the various types of environ- ments occupied by a single species. The purpose of this section is to examine these aspects of the ecology of the Puerto Rican camaenids. The important points to be made in this section are: (1) The species are primarily nocturnal. (2) The daytime microhabitat occupied by juveniles differs markedly from that inhabited by adults in all species, with the juveniles occupying cooler, wetter and less exposed microhabitats than the adults. (3) These age differences are more pronounced in the Loiza Aldea habitat TABLE 2. Species and localities used for each of the studies reported т this paper. “X'” means that con- siderable data were obtained. ‘’+’’ indicates some data but less than was desired for conclusive statements; от means по data. Caracolus carocollus Е! Е! Loiza Type of data Yunque Verde Aldea Caracolus marginella Rio Piedras Polydontes Polydontes Polydontes luquillensis acutangula lima El Yunque El Yunque Rio Piedras Daily activity Habitat selection Home site ranges Food Defensive secretions Mantle water Size at mortality Copulation Eggs Albumen gland cycle Growth Population structure Population density Body temperature Temperature tolerance Stored reserves Evaporative losses Tolerance to water loss | | xxx + XXXXI|XXXI|X XX X | Spall |x++xx | XXKXX+X | X | IRON ETES RES а OCR ARS 5 el) ЕН [< elle 249 PUERTO RICAN CAMAENID ECOLOGY 0 = = 0 0 0 0 = 0 0 0 = = эл3эе % 8Z — — G G G с — v 9 L — — 9A119 JON 4эЧзеэм (pue|Mo}) 0 = = 0 0 0 0 = 0 0 0 = = ana НУ eu] зазиор//о4 0'9€ = — 0'00L — = 002 98% 0 0, 858 0 _ — эле % 9L = = 0 = v S L С a С — - 2A1198 LON Jay1eam (Руе|ч6!ч) 6 = = G = L С 0 0 L 0 = = annoy -Н\М S/Sua/inbn] sazuopAjog 6'&L = Eee 0'9z 0 0 Ур —5'55 0 De 212 vl 0 211198 % vl = E 6 С 51 G с LL 6 al 9 € 3A119€ JON Jayream (puejußıy) LL = L $ 0 0 v L 0 0 L | 0 BAILY -IW е/пбиезпэе sazuopAjog 0'5 = = 0 0 vl 0 0 0 о 65 0 — 211198 % LEL = = Gl LZ GZ 01 € L 81 ze 9 = эле JON Jayjeam (рие|мо]) v = = 0 0 С 0 0 0 0 с 0 = BAY “IV г//эи!блеш snjo9eJe) ited 052 5 0 0 ee Ze 02 ALL 6'5 0 — INTA 197 9 Zl 82 69 68 06 92 89 vl 6 — 9A119 JON el с с 0 0 с (5 0 L $ 0 = BAIOVW Buiuiex 6'5 LL = = 001 0 L'8 LE ES ore 188 19 _ эл!зэе % vobl L8z = = LE Zl ez 9 ve 68 z9 vl — эле JON Lg vz = = € 0 С L € 8 9 L = BAILY Apnojd 0'5 = = 0% 26% zs Cal EVA “Zio ee "Г 0 = эле % 959 3 — vl № Lev 98 19 LZL +oL zz = эле JON (pueluiu) 0% 0 0 0 € 8 L € $ L 0. — 9^135\ Auuns Sn//090129 Sn¡O9e1e) au ¡e101 0031 OOZL 009, 0091 00yLl 0051 0051 0011 000, 0060 0080 Анил Jayean $91594$ лэчзеэм ||\ Биии!бэа 1n0H "зелра!4 OIH 18 sauo „‚PuejMoj,, pue uolje1s ¡e9/60/01g эпБипд |3 au} зе apew элэм зиозеллезао , puejybiH,, ‘SIIBUS JO Аилцое UO Aep JO au] pue лэцзеам jo 193443 ‘€ 3719VL 250 HEATWOLE AND HEATWOLE where more extreme conditions prevail than in the cool, constantly moist El Yunque area. (4) P. acutangula differs in daytime habitat from the other upland species. (5) More adult С. carocollus are found on large trees than on small ones; they do not use trees with scaly or flaking bark, but otherwise do not seem to show preference for particular tree species, al- though individual trees may be favored because of an abundance of crevices or other shelter sites. (6) Few snails occur where slopes are very steep. Activity period All the species are primarily nocturnal and thus do not have temporal segregation. The 2,969 activity observations that were made are summarized in Tables 3 and 4 by species, time of day, and for С. carocollus (for which the most observations were available) by weather conditions. At night, the proportion of snails which were active was high whereas comparable daytime means were low. In general, the lowland species had fewer individuals active by day than did those from montane areas where conditions were not as harsh. During day- light hours, there were no consistent tem- poral trends in the proportion of snails which were active; rather, activity remained at a similar, low level throughout the day. The few high diurnal values occurred when absolute numbers of snails were very low and can thus be attributed to chance. Weather conditions affected diurnal ac- tivity in С. carocollus. On cloudy days there were twice as many snails active as on sunny days. On the other hand, rain inhibited activity, and the activity level was slightly lower than on sunny days. Im- mediate factors influencing diurnal activity are not evident from the field data, as temperature, humidity and light intensity all differ between cloudy and sunny days. However, humidity may be involved, as snails kept at high humidities in the labora- tory seemed to be more active than those TABLE 4. Nocturnal activity of snails in the mon- tane study areas. No. No. in- % Species active active Total active Caracolus carocollus 551 273 824 66.9 Polydontes acutangula 3 1 4 75.0 Polydontes luquillensis 319 88 407 78.4 in dry containers which were otherwise similar. By contrast, Cameron (1970b) found that temperature affected the ac- tivity of several species of helicid snails. The nature of the activity in which snails were engaged did not differ between day and night. Active snails usually foraged over vari- ous substrates. A number of copulations were observed during the day as well as at night. Habitat selection At night most species roam freely over the forest floor, tree trunks or almost any object within their home range, and with the exception of P. acutangula which is most often found on tree leaves, no inter- specific difference in foraging habitat was evident. During the day, however, there were interspecific as well as intraspecific differences in the sites selected for spend- ing the inactive period. On two occasions (31 Dec. 1964 and 30 Jan. 1965) an intensive search was made for С. marginella at the Rio Piedras study area, in the two locations where they had been most commonly found, i.e. on tree trunks and under objects (stones, logs or other debris). Both times juveniles were disproportionately represented under ob- jects, whereas the majority of individuals on tree trunks were adults, indicating dif- ferences in microhabitat between the two groups (Fig. 3). These data represent only a random sample of the sizes of animals found in the two places; the histograms of Fig. 3 cannot be used to compare relative snail densities in the different microhabi- tats, as no attempt was made to sample an equal proportion of the total number of tree trunks and objects on the ground. Juveniles of С. carocollus were not abundant in the Loiza Aldea study area. Only on 26 October 1965 were a sufficient number of juveniles found for analysis of habitat selection to be made. All juveniles were under rocks, many of them occupying small cracks or solution pits in the under- side of porous limestone boulders. Only three adults were found in the same loca- tions. By contrast, all individuals encoun- tered by carefully excavating plots of leaf litter were adults (Fig. 3). No individuals were seen on tree trunks. The small num- ber of young snails encountered on a num- ber of other occasions collectively supports the conclusion based on the large sample of PUERTO RICAN CAMAENID ECOLOGY 251 EN us С marginella RIO PEDRAS 31 DEC 1964 | 6: ON TREE TRUNKS 4, | 2 Ar. ze; ee, eee al UNDER OBJECTS 2 о RE АА, > o 20 30 40 50 60 =) a 5 [> > С morginelia o “4 RIO PEDRAS = | 30 Jan 1965 ON TREE TRUNKS 2 or —— — — —— _— A A u о [ 2 UNDER OBJECTS = © 20 30 40 50 60 Ww o = =} z 8; С. caracollus 6! LOIZA ALDEA 26 OCT 1965 4 IN LEAF LITTER UNDER ROCKS В 50 60 SHELL DIAMETER ( mm ) FIG. 3. Frequency of different size classes of Caraco/us marginella and Caracolus carocollus in different microhabitats. HEATWOLE AND HEATWOLE 252 “$9911 JO 2681104 ay} UO sjleus sauo хэе|а pue “1a111/ }еэ| UI 10 рипо.б ay] UO sjleus sauo рэззор ‘sajOY 291} и! JO SqUI[ 221} JO зихе ul знеиз sauo payoleu-ss019 ‘зип 821} UO зиеиз эзеэ!ри! sieq jo $и01392$ uado “(ge ‘614 ‘| xipuaddy) иепг ueg зе 152104 eapıy eZ107 104 зе43 ‘eulpy ел 18 painseaw эпБипд |3 104 езер jjejuley ‘эпБипд 13 wos е/пбиезпэе зазиор//од ‘D ‘(99-5961 S429A ayı yo a1ISOdWO9) еэле eapıy е2!0] ayl 18 $7//0204е5 snjo9eJe) “g ‘uonels jeoıbojoıg эпбипд |3 ayi Jesu $й//0204е2 $п/оэеле) “y ‘пезииел оз uonejes ul ззеиаечолониы зпомел Buowe sjleus пре JO uolnqinsIQ ‘ÿ “Old SMOILWAUSSEO JO 1432434 (%) TIVINIVU (wo) PUERTO RICAN CAMAENID ECOLOGY 253 26 October. Of the total of 30 such juven- iles found, 24 (80%) were under rocks and 6 (20%) on tree trunks; none were in litter. Adults found during the same observation periods were primarily in the leaf litter, with lesser numbers on tree trunks (see Fig. 4). C. carocollus shows similar behavior in the El Yunque study area except that juveniles occupy a greater variety of day- time microhabitats. Bromeliads do not oc- cur in the Loiza Aldea forest, but are abundant at El Yunque and are frequently occupied by snails. Bromeliads were care- fully examined on October 23-24, 1964, and all snails taken from them were meas- ured. One-fourth (N = 16) were juveniles. A slightly greater proportion (39%, N = 28) of those found under fallen palm petioles on the same date were juveniles, whereas only 15% (N = 13) of those on tree trunks were. Quantitative data were not obtained for the other species occurring in lower popula- tion densities. However, it was noticeable FIG. 5. Views of snails during their inactive periods at their home sites. A, Caracolus carocollus оп a banana petiole near the Biological Station. В, С. carocollus on the Biological Station building. С, many C. carocollus on the trunk of a broadleaf tree in the Downhill Plot. Note that many of the snails are marked. D, С. carocollus on the trunk of a Sierra Palm in the Downhill Plot. E, Po/ydontes acutangula on the stem of a hibiscus plant near the Biological Station. 254 HEATWOLE AND HEATWOLE in the Rio Piedras study area that juvenile P. lima were usually under rocks and adults were on tree trunks. At El Yunque most juvenile P. /uquillensis were inside curled-up Cecropia leaves (leaves either hanging т vegetation or on the ground), or inside fallen, rolled-up palm petioles. Although adults sometimes occurred in these places, they were usually encountered on tree trunks. P. acutangula juveniles were almost al- ways in rolled-up leaves (usually above the ground), in bromeliad axils, or the axils of banana trees; adults were more exposed, usually on the leaves of trees. Such ontogenetic differences in habitat 33 SNAILS OF NUMBER DBH may be common in land snails; Pollard (1975) found a similar situation in Helix pomatia. During searches of the study area for other purposes, notes were made on the exact location of each snail observed. Searches were made under debris and on the ground as well as on trees. Because of the small size of juveniles and their ability to use very small crevices, many were prob- ably overlooked, except when specific searches for them were made. However, the adults are conspicuous enough that many of them sheltering under debris or in leaf litter were found during a routine search. Thus, relative numbers of adults found in Gem) FIG. 6. Relationship of number of resident snails on tree trunks to tree diameter at breast height (dbh) in the Downhill Plot at the Е! Yunque Biological Station. All trees with dbh less than 5cm were grouped as saplings (Е represents mean and range for Euterpe globosa; О represents mean and range for all other species of saplings). M represents mean and range of banana plants (Musa cavendishii) at the level of their mean dbh. Dots represent Euterpe globosa and X's all other species. Broken lines indicate the limits of values for Е. globosa. PUERTO RICAN CAMAENID ECOLOGY 255 different microhabitats could be a useful index of species differences or temporal changes in behavior within a species. One of the most striking differences was between P. acutangula and the other species; 80% of the adult P. acutangula encountered above ground (during the part of the study when such data were re- corded) were on leaves, usually of Hibiscus, banana, or palm plants. The remainder were on tree trunks or stems of shrubs. Adults of all the other species in the Montane Rain Forest primarily occupied tree trunks (Fig. 4). In the El Yunque area, most adult C. carocollus spend the day on tree trunks with a relatively small proportion occupy- ing the ground or other more sheltered sites. The reverse was true at Loiza Aldea where most adult snails were in the leaf litter with only a few on tree trunks (Fig. 4). It appears that in less favorable habitats (or at least those with greater fluctuations in temperature and less rainfall) adult C. carocollus uses more sheltered (cooler and moister) sites than it does in more equable parts of the range. It was noticed that certain trees were used as a daytime resting site more often than others; some were observed to almost never have a snail present, whereas others had up to 19 individuals present at one time (Figs. 1, 5). At El Yunque more С. carocollus were generally found on large trees than on small ones (Fig. 6), but without differences in number of snails between tree species of the same size (Ap- pendix 2). Thus, size not the species of tree usually influences the number of snails inhabiting trees. Several notable exceptions occurred. Three of the larger trees never had any snails on their trunks. Two of these were tree-ferns (Cyathea arborea) which have densely scaly trunks. The third was a Swamp Cyrilla (Cyrilla racemiflora) in which the bark splits off in scales or thin plates and becomes spongy at the base of the tree. All the other species have smooth to moderately rough-textured or furrowed bark, but not scaly or flaking. Two trees of Cecropia peltata had num- bers of resident snails that fell within the range of Euterpe globosa; a third, however, was exceptional in having many more resident snails than expected for its size. Indeed, it had more than twice as many snails as any other tree in the plot. This tree had one unique feature, a rather extensive prop root system at its base under which snails were frequently observed sheltering. It was apparent in the field that snails were not uniformly distributed throughout the study areas (Fig. 7). С. carocollus avoids areas with very steep slopes. This may not be a direct response to topo- graphy, but to some factor in turn related to topography, e.g. amount of leaf litter or soil moisture. There is also another possible explanation. Snails were occasionally ob- served to fall from trees. Should this hap- pen over a very steep slope, the snails would probably roll to the bottom. CENTER | 5 CENTER Il FIG. 7. Distribution of Caracolus carocollus т the El Verde study areas. N, E, S, W refer to compass points. High and Low refer to the higher and lower parts of the topographic map to assist in orientation. Contour lines at 1 m intervals. Dots indicate centers of home ranges of snails captured more than once. Center | is the control area not irradiated by gamma radiation and Cen- ter Il the site that was radiated. Each center is 30 m in diameter. 256 HEATWOLE AND HEATWOLE Home site range Snails seem to have the sensory percep- tion and neural organization to recognize specific features of their external environ- ment and to use environmental cues in maintaining themselves in, or returning to specific locations. Some are known to re- strict their activities to rather small areas (Potts, 1975). The vineyard snail (Helix pomatia) tends to return to the same area each year to overwinter and when moved to a new site can accurately home (Edel- stam & Palmer, 1950; Pollard, 1975). Simi- larly, Blinn (1963) found that two terres- trial species (Mesodon thyroidus, Allogona profunda) “homed”” to overwintering sites in autumn, returning to the vicinity of their previous location every second year. The home site has been defined as the “location in which a snail passes its inactive period of any given day” and the home site range (hsr) as the “area in which the home sites occur over a prolonged period of time” (Heatwole et al., 1970). In the present study individual camaenid snails had a high fidelity to such sites and returned to them during their diurnal inac- tive period (Appendix 2, Figs. 8-10, Table 5). Home site range varied widely among individuals, values ranging from 0.08 т? to 43 т? for С. carocollus at El Yunque; representative ones are shown in Fig. 10. There was a regional difference in size of hsr. Mean areas of the hsr of С. caro- collus at El Verde were 3-8 times as large as for that species at El Yunque (Table 5). At El Verde, mean hsr was 21-59 m? in the two centers, including periods both before and after irradiating one center. For a given region, all species seemed to have approximately the same hsr areas or distances moved between home sites (Table 5). Too few data were available for P. /ima and С. marginella to permit accurate calcu- lation of Asr. However, both species did show an affinity for particular locations, only one individual of the latter species and none of the former having ever been found at more than one home site (Table 2) Juveniles do not disperse widely before setting up a home range, or if they do, they do so quickly and establish an hsr very early in life, i.e. well before reaching maturity (see Appendix 2). There did not seem to be exclusion of any individuals from an area by others. On a number of occasions snails hanging on walls or tree trunks were in direct physical FIG. 8. Home sites (dots) of Caracolus carocollus no. 17 recorded between 15 September 1962 and 31 July 1963, and enclosed in a polygon to indicate its home site range on the northwest corner of the EI Yunque Biological Station building. 257 PUERTO RICAN CAMAENID ECOLOGY ce Сб TLE 807 97% pex1eui SIIBUS JO Jaquinu ¡ezo | = OZ (082-92 ) OL = _ $49399 154 OM] иэамзэа ээиез$ир ueayy I (%L'L ) L (%8'5 ) pl (%7'0 ) | 0 JaqwnNn 454 билош А|зпоэиезиоа$ sjieus = = (0'05-0'5 )OZL — = 19}U99 JSY Mau 0] ||EM UJOAJ 9DUEISIP ueay = = (karl ae a = JequnnN pajuled элам S|¡gm $31 4981348 Buipying WOA} Aeme /sy би!лош sjleug TR ВЯ (48 LL) vv = $. (paze¡no¡e9 jou 454 jo ease) би!риипа e yo нем ayi uo Ájuo 1nq a9uo ue} 2100 Pain}ded sjieus = (ÿ'£t-£0'O ) 1S (1'699-9'0 ) O'LZ (2L'96r- 87 ) 8'se Base ueay\ Е = (%8’SL) ЧР! (%SG'EC) 96 (%S°VL) 55 SIIEUS jO 18qUNnN $9315 3J0W JO $ зе рэлпзае sjieus ||е 10} ueaW puelo) = (7'57-80`0 )ZZ (1'695-0'© ) 5'69 (LEvOL ) G02 (¿Lu) эбиел э315 awoY yo еэле ueayy =. = (%L'LL) 99 (%5'8 ) VE (%S°L ) v JequnnN (seinyded 61-1) $931$ элош JO $ зе рэлпзаеэ sjieus (acre ) Grete (ИЕН) E (еее (O0°962-S'0 ) ZOE (L'96ÿ-8 + ) ГУ (=) эбиел ais эшоч jo еэле ueayı (%b'6 ) $ (%9°L ) IL (%cC'OC) SL (%Z'SL) C9 (%L'LL) 6C J8quinnN (sainides 9-5) $931$ a10W JO $ зе рэлпзаеэ sjIeus = x 6S (9'05-9'0 ) 2s (SLv-80 ) cect (GSS-O1l ) SZL (Lu) $2415 элпз4еэ UaaMjaq ээиез$!р ueayy = (%TZ ) [4 (%6'9 ) 5 (%8'8 ) 95 (%S'9 ) ¡2 19QUINN $9115 OM} ÁJUO je зпа 391M} uey1 элош рэлпзаеэ sjieus Un ER S'S (Cal o-Cak т 89 (8"vL-2'0) LS (OLv-G'O )Z6LL (9'19-9'0 ) себ (ui) $9315 элпз4еэ иээмзэ4 aduejsip ueay\ (%9'91) S (%6'01) OL (%L'G ) LC (%5°6 ) 85 (%£'GL) Or JequnnN (5931$ OM) adIM} painides sjIeus (LE ) | (%0'51) CL (KEIN gcc (%6'9 ) 8 (%6'9 ) 81 915 auo AjUO 1e зпа a9uo чецз элош ралпзаеэ sjieus JO Jaquiny (%6° LL) ec (%7"S9) 09 (%1'95) L6 (%£'8€) 891 (%2°8S) val asuo ÁjuO painides sjleus yO лэашпмМ sısuayjınbn) y в/пбиезпэе “y Sn//0901€9 ‘7 uoleipey 1043409) uo1JeJS |еэ16о|018 anbun, |3 5п//оэолеэ ')—эраэл 13 "(0161 “¡e зэ ajom1eay Bas) ¿ou sem eae |0.14U09 au} “uo! -е!рел ещишеб оз pasodxa sem ease иоце!рел au] *(0461) ‘IE 19 э1омзеэНн WO] езер sapnjou! эрлал 13 “ParxJew sjieus JO Jaquunu |е303 ay] JO % AY} 03 12491 зэпел эбезиээ4э4 ay} 'зэбиел pue зэп|ел ueaw se passaidxa эле Seajy ‘(/sy) sabues ais эшоч pue ‘ззиэшалош JO $913 511эзэелецо ‘рэлпзаеэ sjleus Jo sıeqwnn ‘G 31941 258 HEATWOLE AND HEATWOLE NORTH SIDE de Sega | LA Ic a fan] EAST SIDE SOUTH SIDE WEST SIDE FIG.9. Home sites of Caracolus carocollus on the building of the Е! Yunque Biological Station. Each dot represents an observed location. Solid lines enclose all home sites for a given individual. Dotted lines indicate home site range around corner of building (upper sketch) or movements to a new area (lower sketches). Arrow indicates a home site off the building. Dates of capture for individual snails occur over periods of 42 days to 8 months and 15 days. contact with each other, sometimes in large clusters of individuals. Occasionally one snail was attached to the shell of another. Consequently, the hsr's of different indi- viduals broadly overlapped and if all were plotted on Figs. 8-10 one would be unable to distinguish individual outlines. Most С. carocollus favored a particular home site over others. Of those that were observed in only two sites, more than 80% of the observations were in one place and less than 20% in a second place (on some occasions, of course, given snails could not be found); even among snails that were observed occupying a number of different home sites, their use of such places was not evenly distributed, one to three sites being strongly favored and the remaining ones used only occasionally (Table 6). Data on home range (as opposed to home site range) were very few for any species. Active C. caroco//us were observed on the forest floor at El Yunque Biological Station as much as 20 т from their known hsr. By contrast, 157 marked P. /uquillensis on a retaining wall near La Mina on the average moved 7m or less in a horizontal dimension during their nocturnal activity (Table 7). Food Food habits of a given species differed in different areas. Eleven С. carocollus at El Yunque were observed feeding at night; none were seen feeding by day. The food consumed consisted of a wide variety of items. Three snails were observed eating dead, brown leaves (one leaf was Hibiscus, the other two unidentified), two were TABLE 6. Percentage of observations of Caracolus carocollus at different home sites at Е! Yunque. Only snails for which there were precise data on home site locations are included. Home sites numbered in order of decreasing use Number of Number of snails captures 1 2 3 4 5 6 7 8 9 10 № 12 10 72 81.9 18.1 13 80 60:07 22.6 174 28 159 41.4' 21.6 190 181 35 245 31.6 21.3 16.1 15.7 15.5 15 124 29164 И.З 113.97 92 6 68 29.4 20.2 11.4 9.8 9.8 9.8 9.8 6 80 22.1 57.2131 9.9 8.2 8.2 8.2 8.2 72 20 147 147 ОТ 1095 101-101 ое Ото 3 42 23.9" 14:3 9.4 9.4 7-2 122 Te. 7/2 VA Thee 2 31 16 eel Oat 9.4 9.4 9.4 6.7 6.7 6.7 6.7 6.7 6.7 1 ; 16 18.8 12.5 12.5 6.3 6.3 6.3 6.3 6.3 6.3 6318630673 121 Total 937 Total PUERTO RICAN CAMAENID ECOLOGY 259 GARBAGE PIT HEDGE HEOGE SHRUBS BUILDING FIG. 10. Some representative home site ranges from the El Yunque Biological Station area. Dots represent home sites, larger circles represent tree trunks. TABLE 7. Number of snails marked and recaptured at Rio Piedras (October 1963 to May 1965) and on two retaining walls at El Yunque (April 1965 to January 1966). All snails at Rio Piedras were captured at their original sites only. Those on retaining walls were at various vertical positions on the wall but in one horizontal dimension and no Asr's could be calculated. ‘ Total Number of times captured hed Ace: Species and locality 1 2 3 4 5 6 7 snails Retaining wall P. luquillensis Number of snails 73 30 25 9innt29 0.0 157 Distance between two furthest capture points (m) _ 6.7 5.1 61 36 — — Rio Piedras C. marginella Number always captured at same site 74 21 7 4 3 31 113 Number captured at more than one site (distance between furthest capture point in parentheses) 0 1 (3.2 m) 0 0 0 оо 1 P. lima’ Number always captured at same site 18 2 2 0 0 ORO 22 1No animals recaptured at other than the original site. eating unidentified live green leaves, two were eating large seeds (one was Ormosia krugi, the other unidentified), two fed on wet, discarded paper, and one each on arum roots and on flowers of the tree /nga vera. In the laboratory, snails readily fed on Hibiscus leaves, raw carrots and paper. Clearly the food preferences of this species are very wide. Only two individuals of other species were observed eating; both were P. /uquillensis eating dead, yellow Hibiscus leaves. 260 HEATWOLE AND HEATWOLE All macroscopically identifiable items in the feces were either parts of leaves, hard fibrous tissues (wood, bark, seeds, etc.) or animals. The last was relatively unimpor- tant and probably represents accidental in- gestion of dead animal material (Table 8). The proportion of leafy versus woody ma- terial in the feces varied more among locali- ties than among species. All three species analyzed from the El Yunque Biological Station had 73-78% of the items repre- sented by parts of leaves; the values for C. carocollus was 73%, whereas in the some- what drier El Verde area the value for this species was 54%, a greater difference than among species at Е! Yunque. In the dry lowlands, the only species studied (P. /ima) had a value similar to El Verde C. caro- collus. However, the type of leafy material eaten varied among species within one area, some relying more heavily on dicot species, others on monocot ones. This may reflect availability of different plants in the re- spective microhabitats of these species rather than on actual selection. There was a greater interspecific differ- ence in the content of fecal material on the microscopic (Table 9) than on the macro- scopic level. С. carocollus from both areas were similar in having predominantly diatoms represented in the feces, but dif- fered from Р. acutangula and Р. /uquillensis which had mostly spherocrystals (calcium oxalate), suggesting that they had fed to a relatively larger extent on araceous or other plants which contain these crystals, than TABLE 8. Macroscopic composition of feces of some camaenid snails of Puerto Rico. % of macroscopic items Amorphous Total Species and Monocot Dicot Grass leaf leaf Thin Insect locality leaf leaf blades tissue material fibers Wood Bark Seeds cuticle El Yunque Polydontes acutangula 34 21 6 17 78 8 14 0 0 0 Polydontes luquillensis 26 38 2 12 78 16 6 0 0 0 Caracolus carocollus 2 66 1 4 73 21 1 5 0 0 Е! Verde Caracolus carocollus 30 16 1 7 54 18 14 10 1 3 Rio Piedras Polydontes lima 13 37 5 2 57 18 9 14 2 0 TABLE 9. Microscopic composition of feces of some camaenid snails of Puerto Rico. М refers to total пит- ber of items. % of microscopic items Other Ca oxalate Species and locality Diatoms unicellular algae crystals Plant hairs Woodcells Leaf cells El Yunque Polydontes acutangula (N = 410) 5 0 73 0 11 11 Polydontes luquillensis (N = 371) 2 0 65 % 22 11 Caracolus carocollus (N = 348) 51 1 2 0 16 10 Е! Verde Caracolus carocollus (N = 170) 42 1 5 11 34 7 Rio Piedras Polydontes lima (N = 509) 0 Zi 8 2 14 6 PUERTO RICAN CAMAENID ECOLOGY 261 did C. carocollus. P. lima was very different from any other species in that it had mostly non-diatomaceous algae (mostly desmids) represented in its microscopic fecal elements. Qualitative inspection of the fecal sus- pension revealed the presence of items in addition to those listed in Tables 8 and 9. P. lima feces contained sand or grit (prob- ably accidentally ingested) and those of C. carocollus contained a whole mite, several pieces of bryophytes and an entire small invertebrate (either a parasite or a fly larva). Thus, although there is broad overlap in the food eaten by the various species, some degree of ecological segregation does occur, especially in terms of microscopic food items. Some of these differences are inde- pendent of the area investigated but others are area-dependent. The abundance of the types of food eaten would suggest that competition for food does not occur among syntopic species. However, snails may experience a shortage of high quality food of particular kinds even when accept- able food is in abundance (review Бу Butler, 1976), and more information about the nutrition of these camaenids would be needed to make a complete assessment of interspecific interactions. Defensive secretions P. acutangula is unable to completely withdraw into its shell. Furthermore, its shell is relatively thin compared to that of the other species. In the absence of good protection by the shell, other defensive mechanisms might be expected. This spe- cies exudes a conspicuous yellow slime over the surface of its body when molested. It is suggested that this secretion is defensive. The only other Puerto Rican camaenid with colored slime is P. /uquillensis. М sometimes exudes a very pale yellow secre- tion when molested. On one occasion, two P. luquillensis and two each of P. /ima, C. carocollus and C. marginella were kept in a desiccator jar at 20°C т a saturated atmos- phere with access to food and water. A second similar jar had the same species composition except that it lacked P. luquillensis. All the snails except P. luquillensis died in the first jar within 15 days, whereas all the snails in the second jar were alive and healthy in appearance at that time. This observation suggests that P. luquillensis has а deleterious effect оп other species of camaenid snails when in close contact with them. This hypothesis should be tested in a more rigorous way, with testing extended to include the effects of P. luquillensis and P. acutangula secre- tions on potential predators. The above two sections have indicated nearly complete overlap in activity cycle and broad overlap in food, but partial spatial separation among the camaenid snails of Puerto Rico. The next section investigates how they are affected by and adapted to their physi- cal environments. RESPONDING TO THE PHYSICAL ENVIRONMENT As discussed in the introduction, the physical environment has been shown to be an important selective agent in the life of snails generally. This section examines the physical parameters and relates interspecific differences in modes of adaptation to pat- terns of geographical distribution. Tempera- ture and water were the two parameters selected for study, as they were considered most likely to be of prime importance in regard to interspecific differences in adapta- tion. Body temperature Environmental temperatures are higher in the lowlands than in the uplands. Conse- quently, it might be expected that coastal species (or populations) would adapt to lowland conditions by (1) developing high- er temperature tolerances permitting main- tenance of elevated body temperatures and/ or (2) modifying heat exchange with the environment by behavioral or physiological means to the extent that body temperature does not rise above the level experienced by upland snails. (Because of potentially different thermal conditions in different parts of the habitats in a given region, species which are sympatric but ecological- ly segregated by microhabitat might show similar patterns even though perhaps on a reduced scale.) Finally, (3) differences in the habitat occupied by animals of differ- ent ages might be reflected in various as- pects of thermal adaptation. It was discov- ered that all of the above situations were present to some degree in the adaptation of the Puerto Rican camaenids. 262 HEATWOLE AND HEATWOLE All species in montane rainforest had similar body temperatures; none exceeded 24°C. All measurements on C. marginella (lowland) were above 25°C and in one case reached 35 C (Figs. 12-15). Although body and environmental tem- peratures are correlated, the latter are not precise predictors of the former (Figs. 12-14). Body and environmental tempera- tures were similar for animals in the shade or diffuse sunlight. Body temperature (Тв) correlates more closely with substrate tem- perature (Ts) than with air temperature (Ta), suggesting that heat conduction via the substrate plays a greater role in deter- mining body temperature than does ex- change with the air (Figs. 12, 13). Тв values lower than either TA or Ts are perhaps attributable to cooling by evaporation of water from the mantle cavity and/or general body surface. Several C. carocollus in the shade had body tem- peratures much higher than environmental ones (Figs. 12, 13). Some of these were obtained during a day when there were scattered clouds and their positions were 10 WN E66 a 6 on ar < 4 о < œ a та O I Zen NUMBER 5. 6) Y such that at times when the clouds were not in front of the sun, the snails would be in or near direct sunlight. The others were obtained just after a brief shower which resulted in sudden lowering of environ- mental temperatures; presumably the high snail temperature reflected a lag in cooling. The different morphologies and colors of the various species may be related to thermal adaptation. The lowland species (C. marginella, P. lima) are light in color whereas C. carocollus (which does extend to high altitudes) and P. /uquillensis which are restricted to upland rainforest are dark. The only montane species which is light in color is Р. acutangula. In contrast to other montane species which are tree trunk forms usually occurring in shaded places, it is often found on leaves, frequently in ex- posed situations. Animals in sunlight absorbed radiant energy and usually had Тв’$ above either TA or Ts (Figs. 12, 13). However, body temperature was usually lower than that of the black bulb thermometer (Твв) placed at the snail’s location (Fig. 14), suggesting 8. IM OA TEN OF CAPTURES FIG. 11. Relation of estimated hsr size to number of recaptures of Caracolus carocollus. PUERTO RICAN CAMAENID ECOLOGY 263 either that the snails were less effective absorbers of radiant energy reflected from their surroundings than was the black bulb, or that they were losing heat by evapora- tive cooling or conduction to the substrate. No marked interspecific differences were observed in the relationship of Tg to Tgg. The single snail in direct sun for which a Твв was obtained had a slightly higher temperature than the black bulb. Even so, the light-colored shell of C. marginella re- flects much of the radiant energy. After measuring the body temperature of two 35 30 O Lo] 25 ea) 2 >= O of 0/0 eo 4g O 20 e | Qo S ь® U e 15 15 20 animals, their shells were painted black. Within several minutes the Tg had raised Бу 4.7° in one and 8.9° т the other, although there had been no change т either air or substrate temperatures (Fig. 13). It is probable that the reflective light color of Р. acutangula and the lowland species is an adaptation preventing over- heating. С. carocollus is dark throughout its range, including the lowlands. However, in coastal localities it tends to occupy leaf litter and other cryptic habitats, rather 0: 8% : 25 30 35 Te (С) FIG. 12. Relationship of snail body temperature (Tg) to substrate temperature (Ts). Open symbols represent animals in direct sunlight, solid symbols those in the shade, and half solid ones animals in diffuse sunlight. Round symbols in upper part of graph (above 2 short parallel lines) represent Caracolus marginella, those in lower part (below short parallel lines) represent Caracolus carocollus. Triangular symbols represent Polydontes acutangula and X's Polydontes luquillensis (all in shade). Diagonal line is Y =Х. 2" indicates dots representing two identical values. 264 HEATWOLE AND HEATWOLE than tree trunks; in such places it would be less exposed to direct solar radiation. Evaporative cooling may also influence heat exchange. Р. acutangula cannot retract its body completely into the shell and thus would be more subject to evaporative cool- ing than other species of equivalent size. The occupancy by this species of a more 35 30 20 exposed habitat than other sympatric species may be facilitated by a combination of enhanced evaporative cooling and a light shell color. The relation of shell color and thermal ecology is not restricted to the camaenid snails, nor is it always in the same direc- tion. Jones (1973a, 1973b) suggests that in #_--------% o 25 30 Fa tete) FIG. 13. Relationship of snail body temperature to air temperature. Symbols as in Fig. 12. Dotted lines leading to asterisks indicate increase in body temperature of two Caracolus marginella after their shells had been painted black. PUERTO RICAN CAMAENID ECOLOGY 265 warm areas, a light colored shell would be an advantage to snails through reflection of radiant energy, and dark shells an advan- tage in cold areas because of the enhance- ment of heating; indeed morph frequencies in species with color polymorphism often show such trends. It has been shown in both Cepaea hortensis (Sedimair, 1956) and Cepaea nemoralis (Richardson, 1974) that there is a greater temperature tolerance in bandless (light colored) morphs than in darker ones. However, in the colder limits of its range in Iceland, bandless С. hortensis were more frequent in cold rather than in warm areas, i.e. the reverse to previously reported situations. Arnason and Grant (1976) suggest that this reversal near 3-5 3.0 O о e ZO 00 IR 20 ob do = ge 1-5 15 20 TBB the colder limits is because a dark shell not only absorbs heat more rapidly but also back-radiates more rapidly in the absence of solar input. Thus, a light shell is an advantage in either hot or cold extremes but a disadvantage under intermediate con- ditions. This conclusion is supported by Arnold's (1968) findings in Spain; the bandless morph of С. nemoralis was more frequent under extreme conditions (both cold-dark-humid and hot-dry-open), than under intermediate ones. Temperature tolerances Two endpoints of heat tolerances were used, the falling point (temperature at 25 30 35 EEN FIG. 14. Relationship of snail body temperature to black bulb temperature. Symbols as in Fig. 12. 266 HEATWOLE AND HEATWOLE 12 (°C) SAFETY MARGIN C.c C.m Plim Pa P luq FIG. 15. Comparison of body temperatures (dot- ted lines), falling points (lower, open figures), lethal points (solid figures) and thermal safety margins (upper, open figures) among five species of tree snails. Numbers above figures represent sample size. The asterisk indicates that one ani- mal was active on the bottom of the container and hence its falling point could not be deter- mined. Sample sizes are different for different measurements оп Caracolus carocollus (C.c) and Caracolus marginella (C.m) as some include juven- iles, others not (see text). Vertical lines repre- sent ranges, horizontal lines means, and rectangles two standard errors either side of the mean. Failure of rectangles to overlap indicates signifi- cant differences (approximately 5% rejection level). Safety margin is the temperature difference between the falling and lethal points (see text). which a snail dropped off a vertical sub- strate) and lethal point (temperature at which it failed to respond to tactile stimu- lation); see Appendix 2. P. lima (lowlands) had a significantly higher falling point than did P. acutangula and P. /uquillensis from the cooler moun- tains (Fig. 15). The latter two did not differ significantly from each other. The genus Caracolus, despite the fact that both its species occur in the lowlands, had generally lower falling points than Po/ydontes. С. carocollus and С. marginella did not differ significantly from each other. The lethal points of P. /ima (lowland) and P. /uquillensis (mountains) were not significantly different from each other, but were significantly higher than that of P. acutangula. The two species of Caracolus were not significantly different from each other or from P. acutangula, although the lethal point of С. marginella was somewhat higher than that of С. carocollus. The latter had a significantly lower lethal point than did P. /ima or P. luquillensis (Fig. 15). The temperature difference between the falling and lethal points is an indication of the safety margin that dropping from a vertical surface bestows upon an individual. Only adults were used in the analysis of safety margins because of the size- dependence of one of the factors involved in its calculation (lethal point; see Ap- pendix 2). The mean safety margins of С. caro- collus, С. marginella and P. /uquillensis were between 5.9°C and 6.8°C and were not significantly different from each other (Fig. 15). They were significantly higher than those of Р. /ima and P. acutangula (3.9°C and 4.3°C, respectively). The latter two species were not significantly different. Because of the inverse relation of size and lethal limit (but lack of a size effect on falling point) juveniles have lower safety margins than adults (Appendix 2). Temperature tolerances far exceeded body temperatures experienced in the field for all species except C. marginella in which several snails were encountered with a Tg within the lower range of falling points (Fig. 15). In other species the high- est Тв was more than 10°C below the lowest falling point measured. However, it should be remembered that body tempera- tures were not measured in the hottest part of the year, and that all Tg's for С. carocollus were from a montane popula- tion. Lowland С. carocollus would be ex- pected to have higher body temperatures. In comparison to snails from extreme habitats, the Puerto Rican camaenids have low temperature tolerances. For example, Schmidt-Nielsen et al. (1972) found that some desert species could tolerate tempera- tures of 55°C for over an hour. PUERTO RICAN CAMAENID ECOLOGY The mechanism of heat death was not ascertained in the present study. However, Grainger (1969, 1975) suggests that in Patella vulgata, Helix aspersa and Arianta arbustorum death results from an ionic imbalance at high temperatures which leads to failure of neuromuscular transmission. Effect of temperature on survival of food-deprived animals Thus far, high rather than low tempera- tures have been discussed. Although the subtropical climate of Puerto Rico would seem to provide few opportunities for ex- ceeding the cold tolerances of snails, low temperature may influence altitudinal dis- tribution in other ways. In С. carocollus, survival time during food deprivation was size-dependent at both 20 C and 30 С; small animals died more quickly than large ones (Fig. 18). Inasmuch as small snails tend to have higher oxygen consumption per gram of — O L 60 uy a 2 50 + < oc XX W 40 a = w 30 r = 20 < т a Ne uu ad ГО 20 30 267 body weight than do large ones (Gromadska & Przybylska, 1960), including C. carocollus (Stiven, 1970), size effect on survival may indirectly reflect a metabolic effect. An alternate hypothesis is that large animals have relatively larger stored energy reserves. Regardless of the mechanism in- volved, the ecological significance is that juvenile animals are less able to survive prolonged unfavorable conditions than are adults. Survival time of all ages was, less than half as long at 30°C as at 20°C. The elevated metabolic rate resulting from the higher temperature would cause faster utili- zation of energy reserves and an earlier death. For interspecific comparisons only data from adults were used. Several individuals of С. carocollus ate part of an index card that was inadvertently left under the lid of their container on day 102 of the experi- ment (arrow in Fig. 17). These animals survived much longer than would be ex- pected from extrapolation of the survivor- 40 50 60 70 SHELL DIAMETER (MM) FIG. 16. Relationship of lethal temperature and diameter of shell in Caracolus carocollus (dots) and Caracolus marginella (X's). Lines calculated by least squares method. 268 HEATWOLE AND HEATWOLE (%) E œ o INDIVIDUALS 8 40 20 PERCENT OF - A a he Ms _ _ AAA — о 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 TIME ( days ) FIG. 17. Survival of 18 Caracolus carocollus adults at 20°C (dots) and 17 at 30°C (X's) and of 24 Caracolus marginella adults at 20°C (squares) and 44 at 30°C (circles). Lines eye-fitted. Arrow indicates that the surviving С. carocollus at 20°C inadvertently received food (see text). Dotted line is ап extrapolation of survival curve to abscissa based on data prior to 102 days. ship curve to the abscissa using data prior to day 102 (dotted line in Fig. 17). They probably have cellulases in their guts as do many terrestrial snails (Mason, 1970), and were able to use the paper as food. The increased survival of these individuals illus- trates the importance of even brief oppor- tunities for feeding during generally un- favorable conditions. These animals were excluded from Fig. 18. Excluding them also from the survivorship curves in Fig. 17, maximum survival time of adult C. carocollus at 20°C is estimated to be about 108 days, in contrast to 47 days at 30°C, and 57 days for C. marginella at 20°C. Although maximum survival of C. caro- collus at 20°C was more than twice that at 30°C, in С. marginella maximum sur- vival at 20°C was less than 1% times that at 30°C (59 and 43 days respectively). Thus, temperature changes have greater ef- fect on C. carocollus than on C. marginella, perhaps reflecting different metabolic O; p's in the two species. However, the survival time of even C. marginella was sufficiently long (30-60 days) that it seems unlikely that thermally influenced inanition would often be responsible for differential mor- tality among species in nature. The mech- anism of interspecific differences in rate of depletion of food reserves was not directly ascertained. However, it could arise from (1) interspecific differences in relative amounts or types of energy stored or util- ized, e.g. the greater use of polysaccharides by Planorbis corneus during starvation as compared to some other species which use primarily lipids (Emerson, 1967), differ- ences among species in assimilation rate responses to (2) temperature as noted for various woodland snails by Mason (1970), or to (3) moisture as in the case of Biomphalaria [= Australorbis| glabrata which lowers metabolic rates under adverse moisture conditions and hence lengthens survival during starvation (von Brand et al., 1957). Whatever the mechanism of death, it is clear that at the temperatures prevailing at high altitudes C. caroco//us would use up its stored reserves more slowly than С. marginella and could be considered better adapted to such conditions. Ontogenetic differences in thermal ге- sponse and tolerance were noted (Appendix 2). Juvenile snails had a lower tolerance to high temperatures (Fig. 16) and at a given temperature survived a shorter period dur- ing food deprivation (Fig. 18) than did adults. Thus, it would be advantageous for young snails to avoid high temperatures. Their selection of the less exposed micro- habitats probably has adaptive significance in this regard. Water Moisture varies tremendously among habitats in Puerto Rico. For a given alti- tude there is a general decrease in moisture from east to west and from north to south, and at a given longitude from higher to lower elevations. Thus the Luquillo Moun- tains in the northeastern part of the island PUERTO RICAN CAMAENID ECOLOGY 269 IN DAYS SURVIVAL TIME 10 20 30 DIAMETER 40 50 60 (MM) FIG. 18. Relationship of survival of food-deprived Caracolus carocollus to shell diameter. Lower: at 20°C. Upper: at 30°C. Dots indicate juveniles, X's adults. Four С. carocollus which ate paper during the experiment are excluded. have the wettest climate and the southern coast (especially the southwestern portion) the driest. Adaptation to unfavorable mois- ture conditions would be expected to fol- low strategies similar to those outlined for temperature, e.g. low susceptibility to desiccation (as demonstrated in some snails by Van der Schalie & Getz, 1963) and/or behavioral means of avoiding dry condi- tions or of regulating water exchange with the environment. In this section we (1) describe the correlation between geographic distribution and tolerance to dry condi- tions, as expressed in terms of survival time, when access to water is denied, and in terms of evaporative loss, (2) consider the epiphragm and storage of mantle water as possible adaptations prolonging life dur- ing dry periods, and (3) evaluate size- effects on drought resistance. Mantle water and moisture content During dry periods or in dry localities C. carocollus has less water stored in the mantle cavity than under more moist con- ditions, suggesting that mantle water might be used by the animal for its biological "GL 'B14 ul se sjoquwidg ‘зиэшиладха ay) yo Buiuuibaq ayı зе рец Aau] 1918M эзиеш лэлазецм и!езэл оз раэмо|е элэм SJIBUS ay) Jeu} ul sauo 15214 au} WO} JajjIP SainBiy рэач53еч-5$049 ‘зазиришпц MOI Je 1day элэм зецз эзоцз (3) sainbiy 49e/q pue “sanipiuiny ч6!ч зе day sjewiue Wo. езер зиазалдал (9) Se1nbi} uedO “suol}!puod |езцашилааха зиэлазуир JapUN sjleus рииаешеэ jo эциз [PAIMINS "GL ‘OS oinôuon20 4 sisuelinbn) y owl| а ое! бош 9 591020205 9 .02 o0€£ .02 .0€ .02 .0€£ .0?2 .0O€ .02 .O€ : 3 4 9) 3 р 9 Be 3 9 3 9 À 3 9 3 9 Es El 3 9 3 9 3 3 9 3 э 0 ; = | Sue sE: = bob p 9 + Ole IVAIAUNS en +0€ Spa | | о 3WI1 Ol al | HEATWOLE AND HEATWOLE 270 ААА +09 -04 - 08 - 06 shop) ( PUERTO RICAN CAMAENID ECOLOGY needs during periods of moisture stress. However, experimental attempts to estab- lish whether this is true were inconclusive. The amount of water stored in the mantle cavity showed a lot of individual variation, values ranging from zero to over 7.5 ml. Within a given study area at a particular time, snails from different micro- habitats did not show consistent or sig- nificant differences in amounts of stored 190 140! м! / doy ) 120} œuf ( mg H,0 / gm total body Loss y о = mal WATER OF RATE MEAN 274 water (Fig. 22). It would appear that snails are mobile enough and move sufficiently often that mantle water is replenished if there is water available in their habitat. However, there exist seasonal fluctua- tions in amount of mantle water. From April through December of 1965 when there was considerable rainfall (see Fig. 4), adult С. carocollus at both El Yunque and Loiza Aldea had mean volumes of mantle 2 == CNE СЕ с Е ite 30° 20° P lima COME 20* Р acutangula BSD E 30* 20* Р luquillensis С. caracollus С. marginella FIG. 20. Rates of water loss from camaenid snails at different temperatures and at high humidities (open figures; C) and at low humidities (black figures; E). All snails lacked mantle water. Symbols as in Fig. 15. 272 water of about 2.5-3.5 ml. However, during the usually dry period of January through March (Fig. 4), the mean quantity of man- tle water was lower in both areas; further- more the mean March values for the drier Loiza Aldea site were significantly lower than those from El Yunque (Fig. 22). In January and March in Loiza Aldea, some snails were found which contained no man- tle water at all, a phenomenon never witnessed at Е! Yunque. In March at Loiza Aldea the snails contained on the average less than 0.5 ml of mantle water. Juvenile snails contained mantle water in proportion to their size, no differences HEATWOLE AND HEATWOLE occurring during the wet part of the year between the two study areas (Fig. 23). During the unusually dry conditions of February 1965 the Loiza Aldea snails suf- fered a slightly reduced body water content compared to that of snails from a wetter area at a time of more abundant rainfall (Appendix 2). The soft body parts had a water content of 86.2% and 89.3% in these two groups, respectively. However, the dry period had not prevented the Loiza Aldea adults from feeding, as they had gut con- tents of an even larger proportion of the body weight than did El Yunque snails (Table 10). TABLE 10. Mean weight of various body compartments and of mantle water and gut contents in C. carocollus. Shell Body Body water Mantle water Gut contents Total худ % total xwt %total Xwt %body xwt %body xwt %body Locality N wt g wt g g wt g wt g wt El Yunque 25 28.5 12.5 | 43:9 16.0 56.1 14.3 89.3 1:68410:0 1.9 11.9 Loiza Aldea 24 23.1 10.7 46.3 12.4 53:7 10.7 86.3 ATZE 2.6 21.0 70 60 50 15 = 20 Е = 10 30 TD = > 20 5 4 10 O ; ES C.caracollus C. marginella P lima P luquillensis P acutangula FIG. 21. Vital limit of water loss (percent of hydrated weight lost as water before death occurs) of camaenid snails. Black figures represent data obtained by desiccating snails in moist air at 20°C. Open figures for С. marginella, P. lima and P. /uquillensis represent data from all other treatments. Open figures for all other species represent data from all treatments including humid air at 20°C (see text and Appendix 2). Symbols as in Fig. 15. 273 PUERTO RICAN CAMAENID ECOLOGY "GL ‘Bi ul se sjoqwds "Aınf gz чо Bap|y 2z107 1e $9943 WO1} SJIEUS OL JO э|аше$ au] 10} 1da9xa элпб!} y9ea 104 GZ = М `$3204 Japun sjIeus sainBi} pajddi1s pue ‘sanea) uo sjieus sainbiy payojey AjassaAsues} ‘speijawoigq и! Sjleus sainbiy payoyey Ajjeuobeip ‘18}11| yea; ay) ul sjleus saunbi} P!jos ‘saas} uo sjleus эзеэтри! $элпб!{ иэао “seaie Apnıs OM] и! SN/J0904E9 SNjoDeseD 1¡Mpe JO Аилеэ эдчеш ay} и! 1эзем JO зипоше и! зэбиецо ¡euoseas “zz "II 5961 96! 1935$ 9nv Ane Sanne udV WW 933 930 AON 190 1935 ON AIN эм м UdV г X= - Ре A ++ 0 + — + + fl pulse re u ee er m 0 17 3NNTOA a a 30 > m YILVM + Le NI 3711NVMW o Le] © © ALIAY O ali М SOV) Ev, ZO JNONNA 13 rn ~ (1w) 274 HEATWOLE AND HEATWOLE TABLE 11. Comparison of survival times of adult and juvenile snails under a variety of treatments. Corre- lation analysis of the relation of survival time tosize (diameter of shell) involved a Spearman Rank Correla- tion test (one-tailed). Survival time in days Correlation of survival Mantle N Mean (range) time and size Species Temp Category water Juv., Ad. Juvenile Adult Ge Р С. carocollus 30 Control None 10,10 6.9( 6-7) Ц) 0.36 >0.05 30 Desicc. None 10,10 8.5( 7-10) 8.9( 7-10) 0.15 >0.05 20 Control None ‚5 65 (61-76) 74 (63-80) 0.75 <0.01 20 Desicc. None 5,5 14 (10-25) 66 (15-90) 0.67 <0.05 20 Desicc. With 5,5 43 (10-69) 52 (33-76) 0.46 >0.05 С. marginella 30 Control None 5,5 25 ( 5-67) 53 (16-71) 0.66 0.05>P>0.01 30 Desicc. None 5,5 50 (21-76) 38 (18-58) -0.14 >>0.05 20 Control None 5,5 36 (18-50) 52 (48-56) 0.34 >0.05 20 Desicc. None 5,5 69 (62-80) 81 (78-84) 0.88 <0.01 20 Desicc. With 5,5 65 (28-76) 82 (76-96) 0.81 <0.01 P. lima 20 Desicc. None 3,4 25 (19-30) 30 (30-30) 0.66 0.05>P>0.01 Р. acutangula 30 Control None 2,4 1 ( 1-1) 4.5( 3-6) 0.50 >0.05 30 Desicc. None 3,4 ae) 5 ( 3-6) 0.83 0.05>P>0.01 20 Control None 22 10.5( 9-12) 44 (37-51) = — 20 Desicc. None 2 12 (12-12) 44 (40-48) = = 40 . TZ 35 Е 30 > х = r . . > . . 5 2.5 ш . = | . . = A q | x = 20 < 2 2 Kos r ч х . 5 = u 10 x «Ke, x о xox . Ww ee . 2 z > 05 x о > > xx * a о ow = 9 10 15 20 25 30 35 40 45 50 55 60 SHELL DIAMETER ( mm ) FIG. 23. Relationship of size to amount of mantle water stored by juvenile Caraco/us carocollus at EI Yunque (dots) and Loiza Aldea (X's). Measurements were made during the wet part of 1965. PUERTO RICAN CAMAENID ECOLOGY 275 TABLE 12. Comparison of vital limits of adult and juvenile snails under a variety of treatments. Correla- tion analyses of the relation of vital limit to size (diameter of shell) involved Spearman Rank Correlation tests (two-tailed). P. acutangula not included because raw data for juveniles lost before these correlation analyses were run. Vital limit 2 , Correlation of vital Mantle NE > Mean (range) limit and size Species Temp. Category water Juv., Ad. Juvenile Adult E Р С carocollus 30 Control None 9,10 15.6( 8.2-29.1) 18.2( 8.1-25.2) 0.28 >>0.10 30 Desicc. None 5,10 21.5(17.8-25.6) 21.3(13.1-26.7) -0.003 >>0.10 20 Control None 5,5 11.5( 3.97-18.06) 20.1(15.8-28.6) 0.55 >0.10 20 Desicc. None 5,5 15.6( 9.4-24.3) 17.1(10.7-24.4) 0.14 >>0.10 20 Desicc. With 55 10.3( 5.2-12.5) 17.1( 7.4-24.3) -0.31 >0.10 C. marginella 30 Control None 2,5 14.2(13.5-14.8) 32.5( 7.1-16.9) 0.26 >>0.10 30 Desicc. None 5,5 21.6( 8.0-32.2) 23.5(12.3-41.7) -0.07 >>0.10 20 Control None 5,5 14.5( 8.5-21.3) 3.6( 1.3-17.9) -0.79 <0.02 20 Desicc. None 5,5 32.0(15.7-33.0) 18.4(13.1-19.1) -0.43 >0.10 20 Desicc. With 5,5 23.8(19.5-31.6) 20.0(13.9-28.0) -0.41 >0.10 P. lima 20 Desicc. None 3,4 44.7(35.7-51.0) 52.1(46.6-61.4) -0.39 >0.10 TABLE 13. Comparison of mean rates of loss of adult and juvenile snails under a variety of treatments. Correlation analyses of the relation of mean rate of loss to size (diameter of shell) involved Spearman Rank Correlation tests (one-tailed). P. acutangu/a not included because raw data for juveniles lost before these correlation analyses were run. Correlation of mean rate of loss and size Mean rate of loss Mantie N Mean (range) Species Temp. Category water Juv., Ad. Juvenile Adult TE в С carocollus 30 Control None 4,10 24.9 (14.0-41.6) 26.1(11.6-36.1) 0.28 >>0.05 30 Desicc. None 5,10 21.9 (13.8-24.6) 24.0(16.6-29.8) -0.49 0.05>P>0.01 20 Control None 5,5 1.65( 0.6-3.18) 2.7( 1.9-3.6) 0.50 >0.05 20 Desic. None 5,5 10.57( 8.6-15.3) 3.4( 1.5-7.1) -0.65 0.05>P>0.01 20 Desicc. With 5,5 9.83( 5.15-12.46) 4.1( 1.0-7.4) -0.64 0.05>P>0.01 С. marginella 30 Control None 2,5 6.3( 2.1-10.5) 3.10 2.2-5.1) -0.19 >>0.05 30 Desic. None 5,5 5.0 (1.44-8.24) 6.8( 3.65-7.78) 0.26 >>0.05 20 Control None 5,5 0.6 (0-3.2) 0.7( 0-3.6) 0.17 >>0.10 20 Desicc. None 5,5 3.7 (2.49-5.99) 2:3(21.6-2.2),5-0:72 <0.01 20 Desicc With 5,5 3.6 ( 2.52-5.26) 2.4( 1.8-3.0) -0.47 >0.10 P. lima 20 Desicc. None 3,4 18.3 (17.0-19.0) 17.4(15.5-20.5) -0.39 >>0.05 Desiccation loss and greater sensitivity to elevated Young snails desiccate more rapidly than conspecific adults and survive desiccation a shorter time (Appendix 2, Tables 11, 13). This is in marked contrast to the results of Riddle (1975) who found no effect of body size on rate of evaporative loss in live desert snails, Rabdotus schiedeanus. Adaptation in the Puerto Rican camaenid snails has followed the path of ontogenetic shifts in patterns of response to the en- vironment. Young snails, by selecting more sheltered habitats than adults reduce their water losses and lower the probability of overheating, thereby behaviorally com- pensating for their high rates of evaporative temperatures. That this compensation is not complete is evidenced by the fact that mortality among (and hence selection upon) juveniles is greater than among adults during periods of environmental stress. There was considerable variation in sur- vival time and rate of water loss among experimental treatments and species (only adults used in comparison). Р. /uquillensis was the only species that had consistently poor survival under all experimental treat- ments; it survived less than a week in humid conditions at 20°C if it did not have access to free water. Two species, C. caro- collus and P. acutangula, had low survival 276 HEATWOLE AND HEATWOLE at 30°C, both in low and high humidities (10 days or less), but high survival at cooler temperatures (up to 90 and over 50 days respectively) [Fig. 19]. С. marginella survived well under all conditions although its survival at 20° was longer than at 30°C. P. lima had survival times intermediate be- tween the sets of values previously dis- cussed; differences between temperatures were not great (Fig. 19). The two groups of snails (С. carocollus and С. marginella) that were allowed to retain their mantle water did not show significantly greater survival in comparison to similarly treated individuals without mantle water. Thus, either they had re- tained very little mantle water, or it was quickly lost or otherwise not available for replacing evaporative losses from the body (Fig. 19). The rates of water loss showed ап т- verse pattern to that of survival time, species and/or treatments with high rates of water loss showed low survival and vice versa. Thus, P. /uquillensis had consistently high rates of loss under all treatments (though evaporative rates were significantly lower at 20 C than at 30 C). P. acutangula and С. carocollus both had low rates of loss at 20°C and high ones at 30°C; С. marginella consistently had low rates of water loss and P. /ima intermediate values. These patterns are almost mirror images of those for survival time. Of the two physical factors affecting evaporation, temperature exerted a greater effect on the snails’ water loss than did humidity (water losses were higher at 30°C than at 20°C in most species). Indeed, the differences in dryness of the air between the control and experimental containers resulted in surprisingly small differences in rate of loss; only т P. /ima were rates significantly higher in the experimental containers (dry air) than in the control ones (moist air) [Appendix 2, Fig. 19]. Since some snails alter their metabolic rate with changes in relative humidity (von Brand et al., 1957, Riddle, 1975, 1977), the similarity of rates of loss at different humidities may have resulted from the snails lowering their metabolic rate in dry air and consequently reducing water loss below what it would otherwise have been under those conditions. Rate of evaporative loss in land snails is complex. It is very low in inactive snails, but stimuli such as pressing or pricking the mantle or shaking the snail results in marked increases, resembling those of active snails and perhaps related to mucus production (Machin, 1965). Although shell thickness and area, and aperture size have been suggested as variables influencing evaporative losses (Machin, 1967), it is known that considerable control over evaporation resides in the mantle itself and varies considerably over time and under different conditions (Machin, 1972). Such regulation by the mantle may have reduced water loss at the lower humidities. Although rates of evaporative loss tended to be higher (and survival time lower) in upland species than in lowland ones (Appendix 2, Figs. 19, 20), the vital limit (the % of total hydrated weight which can be lost as water before death occurs) showed a less clear-cut correlation; C. caro- collus, one of the upland species (P. luquillensis), and one lowland species (C. marginella) had similar vital limits (about 20%), whereas values nearly twice that high occurred in lowland P. /ima and even great- er ones in upland P. acutangula (Appendix 2, Fig. 21). The high vital limit of the last species may be an adaptation related to the possible necessity to use water frequently for evaporative cooling (see above). Machin (1967) demonstrated temperature lowering of the mantle of the snail Otala lactea during evaporative water loss. Several species of helicids (Cameron, 1970a) showed responses similar to those in camaenids in that differences in survival, rate of water loss and behavior under con- ditions of low humidity reflected dif- ferences in the moisture of the habitats occupied by different species. Although temperature, water loss, and at the longer survival times perhaps also deple- tion of energy reserves (see section on temperature), all probably had a synergistic effect on survival under the experimental conditions, the close inverse correlation be- tween rate of water loss and survival time suggests that water loss was the prime cause of mortality in most cases. Role of membranes The various species of snails had differ- ent behavioral responses to desiccation. Most species became inactive under low humidities and retracted into their shells. P. acutangula was an exception in that its body is too large to be enclosed completely PUERTO RICAN CAMAENID ECOLOGY 277 by the shell and it merely retracted as much as it could, reducing the exposed surface as much as possible; often the foot appeared curled and shriveled. The other species, when hanging on the glass walls, would sometimes secrete a clear mucus, which when dry would aid in attaching the snail to the container. It also sealed the opening around the attachment. However, whether this clear membrane retards water loss would depend on its permeability to water vapor. This was tested by comparing the rate of loss of С. marginella during periods when the membrane was present with those of the same individuals when they lacked it; no consistent differences were found (Appendix 2). Thus, the clear membrane does not retard water loss from the snails and its function would seem to be the aiding of attachment to a vertical substrate. Similarly von Brand et al. (1957) indicated that the mucus membrane of Biomphalaria glabrata did not seem to re- duce rate of evaporation. А different type of membrane, the epiphragm, is secreted across the opercular opening т С. marginella and P. lima; it does not occur in any of the upland species. It is hard and chalky in appear- ance, and occurred most commonly in ani- mals kept at 30°C, especially those over the desiccant. Survival was higher among snails which had an epiphragm most of the time than among those which had an epiphragm less often (Appendix 2); this finding suggests that the epiphragm might retard water loss. It can be concluded that the clear mem- brane does not influence evaporative rate from snails but that the epiphragm affords partial protection from desiccation and constitutes one of the modes of adaptation. It appears that similar adaptation occurs in other groups of snails as well. Grime & Blythe (1969) have suggested that differ- ences in the local distribution of Arianta arbustorum and Cepaea nemoralis resulted from the reluctance of the former to se- crete an epiphragm and the consequent sensitivity to desiccation which made it unable to colonize south-facing slopes. Machin (1967) found that species of snails from dry habitats tended to form thicker epiphragms than those from wetter ones. Rokitka & Herreid (1975a, 1975b) ob- served that Otala lactea had a greater tendency to form epiphragms at low humidities than at high ones, but surpris- ingly, that more snails formed epiphragms at low rather than at high temperatures; the number of epiphragms secreted by a given snail increased with duration of dormancy, and reached values up to 7. Response to relative humidity One of the C. carocollus subjected to the experimental relative humidity gradient was found to be between 50% and 63% r.h. at each of three consecutive daily observa- tions. By the fourth day it had moved to 20% r.h. where it remained inactive for nine days before being removed from the chamber. A second snail of this species moved between 48% and 60% r.h. during the first day (hourly observations) and then moved to the nearly saturated end of the gradient where it became inactive for three days, after which it was removed. It seems that this species does not have a consistent directional response to a gradient of rela- tive humidity. Effect of dry periods on survival In the lowland, mortality in C. caro- collus is highest among very small juveniles, especially during dry periods; in the up- lands where conditions are continually moist, most small snails survive to adult- hood and die at large size (Appendix 2). The characteristics of the upland popula- tion are paralleled by other upland species, even ones that are in a different family, are much smaller than the camaenids, and have a different ecology [Appendix 2, Fig. 23] (primarily a leaf litter form, Van der Schalie, 1948). In some studies, other snails from habi- tats less equable than rainforest, had a large proportion of juvenile mortality attributa- ble to drought (e.g. Wolda & Kreulen, 1973), whereas other factors, like preda- tion, were of greater significance in adults (Wolda, 1972). In other species, mortality by predation assumes greater significance at all stages and drought affects populations by preventing recruitment, rather than greatly affecting mortality (Potts, 1975). LIFE HISTORY AND POPULATION BIOLOGY Our intent in this section is to compare and contrast the species in terms of basic 278 HEATWOLE AND HEATWOLE TABLE 14. Seasonal occurrence of copulation and oviposition of some Puerto Rican camaenid snails (1962-1966). A = P. acutangula, М = С. marginella, Lq = P. luquillensis, and Lm = Р. lima. Number of copulations observed C. carocollus Number of egg clutches found (Е! Yunque) El El Loiza Other Month Yunque Verde Aldea species Р. acutangula С. carocollus Unidentified Total January 2 2 1 3 February 5 1 1 1 March 6 A 2 2 April 12 1 La 2 11 3 May 3 June 2 2 July 2 4 August 1 A,M 1 1 September October 1 November Lm 1 1 December 1 3 3 1Data from eggs deposited after captivity by snails from Loiza Aldea. life history parameters, growth, repro- ductive cycles, population structure, and factors contributing to population changes. Reproduction The mating season of С. carocollus is an extended one, copulation having been ob- served in every month except September and November (Table 14). The pattern seemed to vary geographically. In the up- lands the mating season is prolonged and centered on the dry season, but in the drier areas it is more seasonally restricted and tends to occur in the wetter parts of the year. Data are too few for the El Verde and Loiza Aldea sites, however, to permit a precise statement. Also in the Loiza Aldea area the fact that observations were made diurnally may have introduced a bias not present in the El Yunque region in that during the drier part of the year, almost no snails were active by day; they may have copulated at night when observations were not made. The fewer observations in the El Verde and Loiza Aldea area reflect at least in part the fact that fewer years were spent studying those areas than was true for El Yunque and do not necessarily indicate differences in incidence of copulation be- tween areas. Individual snails remain in breeding con- dition for at least two months during a given season and during that time copula- tion occurs more than once and with more than one other individual. Given snails may mate in at least two consecutive years and they remain reproductively active a number of years after achievement of adulthood (Appendix 2). The wide temporal separation (March and August) of the few copulation records for P. acutangula suggest an extended mating season. The only copulation record for P. /uquillensis occurred in April, the month of highest incidence т С. carocollus from the same area. These data are consis- tent with the view that these two upland species have a similar mating season. By contrast the lowland species were observed to copulate only during the wettest part of the year (С. marginella in August and Р. lima т November) and thus resemble low- land C. carocollus. At El Yunque copulation in C. caro- collus did not seem to be restricted to any particular type of place; copulating snails were observed on the ground, on stone walls, on tree trunks, beneath leaf litter and within dead, rolled-up Cecropia leaves above the ground, т curled-up palm petioles, and in the axils of bromeliads. Both copulations observed at Loiza Aldea took place on the ground (one pair was placed in a plastic bag; spermatophores were produced during the night). The ob- served copulation of P. /ima was on the ground, that of С. marginella on a tree trunk and those of P. acutangula on live tree leaves. At least in P. acutangula most egg-laying begins about November and continues until February or March. Eggs take at least 1% months to develop. Sporadic oviposition may occur at other times of year (Appen- dix 2). 279 PUERTO RICAN CAMAENID ECOLOGY CHEFS LAAL) "сазер pue $еэле зпо!лел LOI} зем peap Азизээл JO $11э4$ JO $9215 “pz ‘OI ARE LE) (ww) Vil3mvio $961 033 SI $961 wer - vos 530 U $74034 OW sos waz № ww ay» Y vw VZI07 INONNA 13 E sew oneAmmyg $961 МГ - 7961 230 717346 94102995 7 уземпн 30 SIVNOIAION! 280 HEATWOLE AND HEATWOLE The albumen gland in the El Yunque population usually averaged heavier than at Loiza Aldea, probably a reflection of the generally larger body size in the uplands. The seasonal cycle in the albumen gland 20 (9) GLAND Ll |) O z u = >) [se] = - 50 D =] 40 ч Е 5 S 20 O 10 25 = u E E 50 = 40 = 30 Q a 20 e 10 > JF M A M J 281 Loiza Aldea the peak and low points oc- curred in late July and May respectively (Fig. 26). Thus except for very brief periods, at least some individuals of С. carocollus т both localities are in ‘‘reproductive’’ condi- tion (capable of producing albumen) but that the number in such a state reaches a maximum in the summer months with lesser numbers at other times of year. The 1965 albumen cycle did not seem to be closely related to the particular weather conditions of that year. The in- crease in proportion of the population with functional glands began in late April 1965 but at a time when rainfall was still low— unusually so (compare Figs. 4 and 26); otherwise, however, the albumen gland cycle correlated rather well with rainfall. By contrast at Loiza Aldea, the two months of highest rainfall were at low points of the albumen gland cycle. The cycle does correlate with long term weather averages in that peaks in both areas occur during the warm summer months near the EE МОЕ ОАО, Jive) À 5980 М D: «Е FIG. 26. Temporal change in proportion of adults with functional albumen glands (wt > 0.85 9) in two populations of Caracolus carocollus. Sample sizes as in Fig. 25. tn HEATWOLE AND HEATWOLE 282 9961 р мое Ми TT = ee $961 ‘uoners 1291601018 anbun, |3 243 зе $7//0204е2 $п/оэеле JO YMOIH “LZ “OS 7961 96! 5961 Pave ие г Ч м 2095 v Pl lr WY Meana N Y W 3 nn sur 961 рано ] ‚or feral ¡007 087 ПЕН (ww) Н19мМ31 PUERTO RICAN CAMAENID ECOLOGY 283 beginning of the wet season (compare Figs. 4 and 26). The data on copulation and albumen gland cycles in С. carocollus all converge to indicate that reproduction occurs through- out most of the year but that there are seasonal peaks of activity; at El Yunque the peak of mating occurs in April, just before major development of the albumen glands. The albumen gland cycle peaks in June, about 2 months after mating. It is probable that development of ova follows immediately after the development of the albumen glands and that oviposition peaks in late summer well in time for egg devel- opment to occur before the onset of the dry season in late winter to early spring. The fact that in С. carocollus neither ani- mals with ripe ova or shelled eggs inside, nor eggs or even newly hatched snails were ever found suggests that just prior to devel- opment of ova (just after development of albumen gland) the animals secrete them- selves in places where they are not easily found and deposit their eggs there, the young not emerging until yolk stores are exhausted after hatching. The most likely place for such activities is underground as epigean habitats were extensively and т- tensively searched without success. Diaz- Piferrer (1962) was similarly unable to find eggs of the Cuban tree snail in the field. Other species were not studied as ex- tensively as C. caroco/lus. However, P. acutangula clearly has a somewhat different life cycle as it oviposits primarily in Janu- ary to April, at a time when С. carocollus is still primarily mating. Growth and longevity Most growth of С. carocollus took place in the middle of the year but little at other times (Figs. 27, 28). Growth rates of С. carocollus at EI Verde have previously been published (Heatwole et al., 1970). In general there was considerable growth from late May to August with little growth throughout the rest of the year. Some juveniles failed to grow at all for a particular growth season, or in some cases even for two consecutive ones. At the Е! Yunque site, growth seemed to begin in March and extend to at least August and perhaps several months later 560 | Г 3201 Е Е 280 2401 200 z 160 E 2 Ш Lo Caracolus carocollus 120 | EL YUNQUE BIOLOGICAL STATION 4 | a 80 23 wn ac OE AE и ee EURE EE REN RME 1962 1963 1964 1965 1966 FIG. 28. Growth of Caracolus carocollus captured infrequently over long periods at the El Yunque Biological Station. HEATWOLE AND HEATWOLE 284 ol АИ dv 8 9 b 2 OF 82 92 bz гг OZ BI gm а OF 8 эре HOYVAN te 62 ¿2 se £2 12 61 41 a cl SALAS Е AYUVNY833 fs € de Geez 12 61 2 Gi я 6 "эл! P8A185q0 эзер 158] YIM SI93UU09 aul) раззор !Peap рипо; sem |IBUS yoym зе азер salylußıs а “Pjai} ayı ul е/пбиезпэе 5азиор//од JO $эцэзп|о payorey А]маи JO YyYIMOID “GZ "II АВУПМУГ Of 82 92 v2 22 02 A 9 bl $961 si oz se 0€ 10% Se ss 113HS HLONS1 ( ww ) PUERTO RICAN CAMAENID ECOLOGY (Fig. 27). Thus, at El Yunque the growing season is somewhat extended beyond that at the drier El Verde site. Fig. 27 summarizes the growth of differ- ent animals than those appearing in Fig. 26 and which were caught infrequently over long time periods and indicates yearly growth but not seasonal patterns. This fig- ure, in conjunction with Fig. 27 and the growth curves from Heatwole et al. (1970), indicate individual variability of growth rates in С. carocollus. Slow growers took up to six years to reach maturity whereas fast growers could do so in three years. Such large differences in growth rates among individuals is not without ргесе- dence in terrestrial snails; Blinn (1963) found considerable variation in growth in Mesodon thyroidus and Allogona profunda and Wolda (1970) indicated a poly- morphism in growth rates in Cepaea nemoralis with fast growers reaching adult- hood in one year but slow growers taking three years. Similarly, Pollard (1975) indi- cates that growth rates among individuals from the same batch of eggs can vary widely, resulting in time of reaching matu- 60 ) ( mm 45 40 LENGTH 35 30 SHELL 25 20 26.282730 1723778577 и 55 JANUARY FEBRUARY 7 19 21 285 rity varying from two to five years. Richardson (1975) indicates great varia- bility in growth rates in Cepaea nemoralis and Giesel (1969) reports growth rate poly- morphism in a marine limpet, Acmaea digitalis. In the present study the mech- anism of growth rate differences among individuals was not ascertained. In Wolda’s study, at least part of such differences were genetically determined. Seasonal differences in growth rate were less pronounced in P. acutangula and in P. /uquillensis. The growth curves suggest that P. acutangula reaches maturity in about one year and P. /uquillensis in about two years (Appendix 2, Figs. 31, 32). The most data relevant to evaluation of longevity were obtained from C. carocollus. The С. carocollus with the greatest known age was marked on 23 Oct. 1962 as an adult and was recovered live on 8 Feb. 1970. The time between first and last capture was thus 7 years and 4 months. Since it was adult when first marked, it must have been at least 3 years older than that (the minimum time required to reach adulthood) for a minimum age of 10 years, де и |) So sl MARCH IS 17 1966 FIG. 30. Growth of 14 newly hatched Polydontes acutangula reared at 20°C in the laboratory. Dashed line indicates time after some hatchlings died and N values above figures indicate number of remaining live snails. HEATWOLE AND HEATWOLE 286 'UONEIS ¡e9160/0/g anbun, 13 ayı зе г/пбиезпэе sazuOpAjog 40 YIMOJO “LE "913 um aa NT 30 AON 150 143 MW д NN АМ вм YN 83 м 930 AON 190 1335 eee DAA ео ee à da мо с da cute 6 x oe à (ба | ог ¡ob E los ое PER ЗЕ а E 001 oz! — 55а 1 Ovi | oo: NOILVLS 2070 ПОМПА 3 ¡pa 1002 опбиоцпоо SAJUOPÁJOY ог ог HLON3I T13HS (ww) 287 PUERTO RICAN CAMAENID ECOLOGY Ov 08 oz! 002 | "Bulsayynjo эприп juanald оз А|элещ pajeiedas sainbiy лэмо| pue Jaddn "sısuajjınbn) sazuOpA/Od 40 YIMOIH "ZE “Ols Ant 596! ном Bbumiojas *anbunx 13 Ov [ora] 091 002 Ove ININIYNASVIN YVAN (ww ) 288 HEATWOLE AND HEATWOLE 4 months. A second, recently dead shell was found of the same age. In the El Yunque region there were six other snails with known ages of 8 or more years. Their ages were: 8 yrs, 8 yrs, 8 yrs 2 mo., 8 yrs 5 mo., 9 yrs 8 mo., and 9 yrs 9 mo. Snails up to a known age of 5 years were common. On 8 Feb. 1970 the region of the El Yunque Biological Station was searched for С. carocollus, 3 years and 10 months after the last snail had been marked in that area. None of the 18 juveniles had been previ- ously marked, but of the 83 adults found, 19 (23%) had been painted. Thus, almost one-quarter of the adult population were four or more years old. The individual histories of very few of them could be traced as some of the paint had flaked off of some animals and although one could tell they had been painted, their numbers could not be ascertained. Those with num- bers still legible were among the group of snails 8 or more years old listed above. The marked population in the Е! Verde area was not studied in 1970 and hence longevity data from that area are not avail- able. It is probable that Р. /uquillensis and P. acutangula are not so long-lived as C. caro- collus. The oldest known individuals of these two species were 3 years 5 months (2 individuals) and 2 years 4 months respec- tively. There are insufficient data to evalu- ate longevity for the other species. Population biology The size structure of the C. carocollus population was remarkably stable with the various size classes rather uniformly dis- tributed (Fig. 33) though there was usually a slight peak in the adult size range and the smallest size-groups of juveniles were роог- ly represented. With the exception of a large number of very small individuals in Sept. 25, 1965, there was no evidence of temporal peaks of reproductive activity and subsequent seasonal shifts in frequency dis- tribution of size classes. Such a stable size structure with a large number of adults might be expected in a long-lived species with low reproductive rates. С. carocollus is definitely long-lived after reaching the adult stage (see growth and longevity section); its reproductive rate is not known except that for a given year up to 68% of the adults at Е! Yunque have functional albumen glands (presumably heralding reproductive activ- ity) at the peak of the albumen gland cycle. Thus, at least that proportion must reproduce each year. P. luquillensis had a less stable size- structure than С. carocollus. In March 1965, there were 2 peaks in the size-frequency distribution, one consisting of adults and the other of medium-sized juveniles. During subsequent months the juvenile peak shift- ed progressively toward the adult one (Fig. 34) until they converged in early 1966. That this is not a regular yearly cycle is evidenced by the fact that the size-struc- ture in March 1965 and March 1966 were quite different. Rather, it seems that repro- duction had either been inhibited or there had been intense selective mortality among juveniles at a particular period which re- sulted in the bimodal pattern of March 1965. Subsequent months showed a gradual trend towards re-establishment of the origi- nal size-structure. Thus, although both Р. /uquillensis and C. carocollus were studied on the same wall, they showed different stabilities in population size-structure in ways suggesting that the former is more sensitive to im- mediate environmental influences than is the latter. The wide geographic distribution of the latter and the restricted one of the former would support this view. Also, P. /uquillensis reaches maturity in much short- er time and the average of many year classes which would tend to obscure year to year variation would be less likely than in the more slowly maturing С. carocollus. Environmental moisture seemed to af- fect population density. Even in such an equable environment as the montane rain forest at El Yunque, population density of С. carocollus showed long term changes. In 1962 and 1963 densities were relatively high but dropped during 1964 which was drier (see Eastern Interior summary of rain- fall, U.S. Department of Commerce, 1962-1966). Recovery had not occurred by 1965 or 1966 (both wet years) probably because of the long lag imposed by the long mean generation time of this species. The densities of C. caroco//us in the El Verde area was about 1/5 or less that of the highest density at the El Yunque forest and about the same as or less than the lower ones. It would appear that the wetter Е! Yunque forest generally supports higher population densities than does the forest at El Verde; the environs of the station build- ings appear to be especially favorable. PUERTO RICAN CAMAENID ECOLOGY 289 Although no quantitative data are avail- able, casual observations suggested that the density of С. carocollus at Loiza Aldea is much lower than at either upland site. There were insufficient recaptures for population density estimates of other species. Capture-recapture methods have been widely used for a long time to estimate population densities of animals. Methods have been continuously refined but even so, the most refined techniques demand either rather specialized attributes of the population sampled and/or very high sam- pling intensities, and are not strictly applic- able to many species. Parmenter (1976) has recently reviewed the literature on capture-recapture tech- niques and suggests that even the most refined ones are often grossly inaccurate when applied to the estimation of densities of most natural populations and warns against wholesale acceptance that the basic assumptions of the methods are met by the population under study; the inherent biases are reduced to reasonable levels only by sampling intensities higher than are gener- ally achieved with natural populations and density estimates are often of limited use unless only an order-of-magnitude accuracy is required. Capture-recapture methods may be especially unsuitable for studies of terres- trial snails. Parr et al. (1968) were unable to get reliable population density estimates of Helix aspersa using the Jolly (1965) method because random mixing of marked and unmarked animals did not occur. Snails aestivating on walls were captured fre- quently whereas those in holes were seldom found. A similar situation was found in the present study. In addition, snails maturing eggs were inaccessible for long periods of time and the assumption of equal catch- ability was clearly violated. As a result, population estimates close in time varied widely and variances were high. Conse- quently the estimates should not be con- sidered as precise, but rather as an indi- cation of order of magnitude only. Fig. 35 indicates the number of С. carocollus esti- mated to be present around the Е! Yunque Biological Station and on the Downhill Plot during the study. The best indicator of general trends would probably be those curves smoothed out to eliminate large fluctuations arising from sampling errors inherent in the method. It appears that the population was increasing in late 1962 and into mid-1963 after which it declined to a nearly constant level in the Downhill Plot but continued to gradually decrease around the Biological Station building. The latter decline may have resulted from disturbance to snails associated with the construction of a new wing onto the building, repeated painting of the walls, and other main- tenance. The estimated number of snails indi- cated in Fig. 35 represents population den- sities ranging from 1,285 to 7,070 per hectare for the Downhill Plot and 1,990 to 11,520 per hectare for the station grounds. However, the best estimates for providing a general idea of range in population densi- ties of the station ground would be the average of the estimates for the periods December 1962 to May 1963 (high den- sity) and July 1965 to February 1966 (low density). These values are 8,640 and 2,825 per hectare respectively and are probably representative of the range in population densities during the study. In the Downhill Plot, the average estimate for the period of high population density (March to July 1963) was 5,105 per hectare and that of the period of low density (July 1965 to January 1966) was 1,425 per hectare. The latter values are probably representative of general densities of the forest proper in that region. Thus, the population density of snails on the building grounds was 1.5-2 times that in the forest. Heatwole et al. (1970) have indicated a wide variation in population estimates from the El Verde region but most fall in the range of 300 to 400 snails in the radiation center (1,060 to 1,415 per hectare); there were about 2/3 that number in the control center (700 to 950 per hectare). There was little consistent temporal change in the estimates. Life history strategies When the various topics discussed above are combined, it is evident that the various species have adopted different life history strategies. Timing of reproductive events differs among species in the rain forest. In general, С. carocollus has a widely spread reproductive season but tends to avoid the wettest part of the year in the rain forest and the drier periods in the lowland areas. The proximate factors affecting the re- productive cycle and its spatial variation 290 HEATWOLE AND HEATWOLE were not ascertained. However, the fact that these snails, especially young ones, are moisture sensitive suggests that there may be survival value in adjusting appearance of hatchlings and/or young to coincide with favorably moist but not excessively wet periods. Diaz-Piferrer (1962) noted that reproduction in the Cuban tree snail (Polymita muscarum) appeared to be close- ly related to fluctuations of rainfall and apparently occurred at any season of the year if weather conditions were favorable. DECEMBER 27, 1964 INDIVIDUALS OF NUMBER MARCH 2, 1965 APRIL 17, 1965 20 24 26 32 36 40 44 48 52 56 60 64 SHELL DIAMETER (mm) JUNE 15, 1966 JULY 15, 1965 ofall O | A Pe alee AUGUST 27, 1965 SEPTEMBER 25, 1965 8 | m OCTOBER 23, 1965 DECEMBER 27, 1965 20 24 28 32 36 4 44 48 52 36 60 64 SHELL DIAMETER (mm) FIG. 33. Size-structure of the Caracolus carocollus population on the retaining wall near La Mina on different dates. PUERTO RICAN CAMAENID ECOLOGY 291 n = a =) a 2 Q z u бю e 6 4 2 о 10 ao AUGUST 27, 1965 W 2 6 5 4 TANT: о 10 o SEPTEMBER 25, 1965 2 о © 12 M 6 18 20 22 24 26 28 30 32 34 36 38 40 SHELL, SIZE (mm) OCTOBER 23, 1965 JANUARY 22, 1966 г pais Stl FEBRUARY 26, 1966 O0 LR 4 6 18 20 22 24 26 28 30 32 3 36 38 40 SHELL). SIZE ( mm) FIG. 34. Size-structure of the Polydontes luquillensis population on the retaining wall near La Mina on different dates. Owen (1964) found that the land snail Limicolaria martensiana bred ai all months of the year but with two seasonal peaks associated with the two annual wet and dry seasons in such a way that the newly hatched snails appeared during the wettest months. C. carocollus is a long-lived species with a seasonably stable size-structure whereas Р. luguillensis and P. acutangula from the same locality have much shorter life spans, and Р. /uquillensis, at least, shows seasonal changes in population size-structure. Long life may not be exceptional among large terrestrial snails; for example Dell (1953) found that Paryphanta busbyi took 7 years to attain a diameter of 54 mm. Within a species (С. carocollus) survivor- ship varied between localities, relatively greater mortality occurring among small snails in drier areas. The effect on eggs may have been even greater; no data on egg mortality were obtained. The various aspects of population and 292 HEATWOLE AND HEATWOLE DOWNHILL PLOT SNAILS © о u о 200 EL YUNQUE BUILDING AND GROUNDS a uy [ea] 2 = = [e] Ber _—— —— д S 00 N.D YW FM Aim J WAS ОНО 4 ЯМА WY JAS OW ID У ЕМА М ЛАЗО ШО ЗЕ МАМ 1962 1963 1964 1965 1966 FIG. 35. Change in estimated population density of Caracolus carocollus at the El Yunque study area. Estimates made by the method of Jolly (1965). reproductive biology have consequences for the ways these species can respond to selective influences. These will be discussed in the final section. SUMMARY AND CONCLUSIONS In the previous sections the adaptations of camaenid snails to their particular en- vironments were discussed in terms of their behavior, physiology and life history. It now remains to integrate these different aspects into a more complete picture of adaptive patterns. There are two major ways of viewing adaptation. One is to take a broadly com- parative approach and contrast the various modes of adaptation of related taxa oc- cupying different habitats or geographic regions. The second way is to study the opera- tion of natural selection at the single species level and examine the genetic basis of intraspecific adaptation. This latter ap- proach complements the former in that it provides a background of insight into the mechanics of adaptation against which the comparative aspects can be interpreted. Study of intraspecific variation (especial- ly color polymorphism) and adaptation at the species level in snails is well advanced and constitutes one of the important pillars of modern population genetics. By con- trast, the broadly comparative approach taken in the present study has previously scarcely been attempted for land snails. In this chapter we briefly review the major literature on the population genetics of snails as a background to a discussion of the adaptation of the Puerto Rican camaenids. The study of ecological genetics and selection in snails has centered around genera which are polymorphic for color and banding, especially Cepaea. The genetics of this polymorphism is partly known (Murray, 1963; Murray & Clarke, 1966, Cain et al., 1968) and thus ecological and population data can be interpreted against a certain background of genetic knowledge. Morph frequencies in nature show “area effects” in that adjacent and apparently similar areas have different morph fre- quencies but paradoxically, large, ecolog- ically diverse areas may have uniform fre- quencies (Cain & Currey, 1963a, 1968, Clarke & Murray, 1969). In other cases abundance of certain phenotypes согге- sponds to particular habitats or conditions which are consistent over a wide geographic area, or which show strong geographic cor- relates. Much of the research on Cepaea PUERTO RICAN CAMAENID ECOLOGY 293 has centered around assessment of the rela- tive role of various factors in influencing morph frequencies. Several categories of factors have been considered. (1) Drift: Genetic drift was one early suggestion (Lamotte, 1959). It was con- sidered that particular breeding units or demes showed random change in fre- quencies as a result of stochastic processes operating in small populations. Most later papers have tended to discount the impor- tance of drift relative to that of selection. However, random drift is not completely discounted, and it may operate under par- ticular conditions. Selander & Kauffman (1975) indicated that in Helix aspersa in a city, there were demes of only a few (up to 15) individuals which were isolated in part by low mobility. The pattern of vari- ance in allele frequencies suggested that drift was operating. Brussard (1974) found that at some loci in Cepaea, stochastic processes in small populations (fewer than 100 individuals) were strong enough to override any heterozygote advantage. (2) Climatic selection: Another view is that morph frequencies are expressions of the direct effect of selection. Clarke & Murray (1962a, 1962b) and Clarke et al. (1968) collected Cepaea in the same locali- ties from which collections had been made several decades earlier and found that there were consistent changes in morph fre- quencies amounting to a selective advantage or disadvantage of 5.2-6.2 per generation. In other cases, little change occurred and the constancy correlated with degree of habitat change; gene frequencies were more stable in habitats with little change. Some authors have attributed the selective effect to climate or microclimate (Arnold, 1968, 1969, 1971; Jones, 1973b). This appears to be likely in the case of close correlation of local and/or geographic patterns of fre- quencies with climatic factors if different morphs have different tolerances or re- sponses to environmental factors, as is true in Cepaea (see reviews by Lamotte, 1959; Wolda, 1967). Some workers have sug- gested that climatic effects operate where area effects are evident (Cain & Currey, 1963a, 1963b; Cain, 1968). Arnason & Grant (1976) have pointed to a close corre- lation of morph frequencies with tempera- ture, shading and humidity in C. hortensis near the northern limits of its range in Iceland. Morph frequencies showed con- stancy over large areas which had sharp discontinuities in habitat. It would appear that habitat and climatic continuities do not necessarily follow the same pattern and that morph frequencies are more closely related to the latter. Cain (1971) has re- viewed the work on morph frequencies in subfossil Сераеа and concluded that light colored, unbanded individuals were preva- lent in hotter times and the darker, banded ones in cooler periods. This also related to the present-day distributional pattern of these morphs (Arnold, 1968, 1969; Cain & Currey, 1963a-c; Carter, 1968; Cain, 1968). Similarly a white-lip seems to be associated with lowered temperature, or more prob- ably, dampness (see review by Cain, 1971). Morph frequencies in C. hortensis did not show as clear-cut a temporal pattern as did C. nemoralis. The temporal changes in pro- portions of these 2 species and their pres- ent distributions suggest that C. hortensis is relatively favored by a cooler climate and C. nemoralis by a warmer one (reviewed by Cain, 1971). Finally, Bantock & Price (1975) found that color pattern poly- morphism decreased in marginal open areas in ways expected if the effects were related to temperature. The most direct evidence of differential mortality among morphs was obtained by Richardson (1974) who found С. nemoralis dying in the field with symptoms of heat death; the morph fre- quencies of the dead snails departed from that of the population at large. There were fewer than expected of yellow, unbanded snails, a result expected on the basis of geographic and micro-geographic distribu- tions of morph frequencies and on labo- ratory tests of heat tolerances. (3) Predator selection: Cepaea colonies in woodlands tend to be heavily banded which gives a dark color matching that of the leaf litter, whereas colonies in short grasses have a lower percentage of banding and match the uniformly green background better. This background matching is at- tributed to “visual selection’’ such as the selective action of predatory birds against conspicuous color morphs (Cain & Sheppard, 1950, 1954; Currey et al., 1964; Carter, 1968; Cain & Currey, 1968). Two species of Cepaea sometimes occur together with rather different morph frequencies. Lamotte (1959) cited this as evidence against visual selection. Clarke (1960), how- ever, points out that different patterns in the two species can appear very similar; one species has a darker background color 294 HEATWOLE AND HEATWOLE whereas the other achieves the same effect by fusion of bands on yellow snails. Ban- tock & Bayley (1973) found that thrushes could not distinguish between visually simi- lar though structurally different snails of the two species. Richards & Murray (1975) have shown that a population of intro- duced Cepaea into the United States has responded in ways predicted by Clarke’s hypothesis. On the other hand, such intro- duced populations retain genetic character- istics traceable to their site of origin (Brussard, 1975). Color and pattern polymorphism may become balanced by the tendency of predators to develop search images and take disproportionately higher percentages of common varieties than of rarer ones. Clarke (1962) has called such selection “apostatic selection.” The color and pat- tern polymorphism in the African land snail Limicolaria martensiana has been postulated to be influenced by apostatic selection (Owen, 1963); that polymorphism is known to have existed for at least 8,000-10,000 years without loss of any of the types of morphs (Owen, 1966). Thus balanced polymorphism may be relatively stable for long periods. (4) Stability of co-adapted gene pools: Clarke & Murray (1969) proposed that co-adapted gene pools may account for area effects in species of Partu/a although they were hesitant to suggest that such a mechanism would cause divergence to the point of speciation. Basically the idea is that in species with low mobility, co- adapted gene complexes might arise, and frequencies of shell color genotypes might be determined as much by the genetic environment as by the external environ- ment; there may be evolutionary trends towards steepening of micro-geographic clines and the formation of sharp steps within them which would give rise to regions of comparatively uniform morph frequency, separated from other such regions by various transition zones; these regions need not correspond to discon- tinuities in the external environment (Clarke, 1966, 1968). However, one would suspect that all successful species would have co-adapted gene pools; a decision as to whether they play any special role in area effects still requires empirical documentation. Goodhart (1962) has suggested that when large populations are subjected to excessive mortality by disasters (floods in the case of his populations of Cepaea) random genetic assortment in the small, isolated populations would operate much as it does in the “founder effect’’ when a few individuals from a parent population, and carrying only a small part of the variability of the population at large, colonize a new habitat and give rise to a population con- taining markedly different properties from those of the parent population. Such an effect could account for local populations with different morph frequencies not ob- viously correlated with environmental features or predation pressures. “Роипаег”” populations might evolve different systems of balanced polymorphism in different areas when exposed to similar selective forces. Upon expansion and meeting of such populations they might maintain their co-adapted genetic integrity. It is likely that all of these factors, in various combinations in different regions, may be involved. For example, Jones (1973a) has suggested that although shell color may be influenced by the properties related to absorption of solar energy, ex- pression as “area effects” rather than clinal patterns 15 probably determined Бу со- adaptation. A final consideration is that observed morph frequencies may reflect differences in habitat selection by different morphs rather than the categories of effects men- tioned above (especially in heterogeneous environments). Different morphs of Cepaea nemoralis may select different environ- ments (Sedimair, 1956). Morph differences may not apply to all aspects of habitat selection, however, as in Bulimulus, Heatwole & Clarke (in press) were unable to detect differences among morphs in the selection of height or direction of exposure on tree trunks. Even in such well-studied groups as Cepaea there is not complete agreement as to the relative importance of the various types of selection or of random processes (e.g. Owen & Jones, 1974; Cain & Currey, 1963c versus Goodhart, 1963). Much work remains to be done before the evolutionary dynamics of any snail species can be under- stood. Perhaps for this reason most previous research has involved detailed study of the ecology of individual species or comparisons of a few aspects among several species. The broadly comparative, synoptic approach has been lacking. The present study is a step in that direction. PUERTO RICAN CAMAENID ECOLOGY 295 Clarke & Murray (1969) have suggested on the basis of work on Moorean Partula that tropical islands may provide a favor- able environment for sympatric speciation because of their combination of habitat diversity and equable environment and be- cause of a relative lack of resident fauna filling available niches. The camaenids of Puerto Rico, however, do not seem to have followed this path. In contrast to Aegean enids which seem unable to cross water barriers (Heller, 1976) almost every species of Puerto Rican camaenid represents inva- sion of a different stock from outside, and consequently evolution has centered around interactions of previously reproductively isolated taxa with each other and adjust- ment to specific climatic regimens rather than divergence in situ from a common stock. Randolph (1973) obtained experimental data indicating that a species of snail from a more variable physical environment ex- hibited a wider ‘’tolerance-niche’’ (tempera- ture and moisture primarily), higher r value, shorter life span, smaller litter size, shorter developmental time, more general behavior relative to its habitat, more general food requirements, and smaller biomass than did a species which occurred in a less fluctuating environment. We have not examined all these aspects for the camaenids; however, some of our results invite comparison. C. caroco/lus has a long life span, a rather long developmental time, a large (but unquantified) biomass, a con- siderable sensitivity to moisture at high temperatures, and although precise data are lacking it would appear to have a relatively low r value. In these characteristics it would seem to fit the expectation for a species inhabiting a relatively equable en- vironment as indeed it does in part of its range. However, in being a food generalist, having a small clutch size, and being rather temperature-tolerant it resembles more the expectations for a species from a fluctuat- ing environment. When all aspects are con- sidered collectively, it would appear that C. carocollus is primarily suited for existence in an equable environment. That it also lives in less equable areas can be attributed to its behavior; in such environments it selects only those micro-environments in which environmental oscillations are damped. This is especially true of the young, and more vulnerable, stages. Although not studied as well, Р. luquillensis seems to be more strongly an “equable-environment” type in that it is more sensitive to physical environmental extremes than С. carocollus, but less strongly so by virtue of its shorter life span. In reality it occurs only in the most equable environment in Puerto Rico. Both the lowland species, Р. /ima and С. marginella, have the expected broad toler- ances to the physical environment but in- sufficient is known of their other character- istics to evaluate whether they fulfil predic- tions in other regards. The relatively short life span of P. acutangula may reflect the fact that it lives in the more exposed parts (and consequently with greater fluctua- tions) of an otherwise equable environ- ment. However, perhaps it is not realistic to attach too much importance to relative values of life span. Among the species studied, the life span of P. acutangula and P. luquillensis were short only in compari- son to C. carocollus. In fact, all of these could be considered as having rather long life spans. Of perhaps more significance is the age-specific differences in survival. Furthermore, a short life span may not always be associated with extreme, fluctu- ating environments. Maiorana (1976) sug- gests that uncertain juvenile survival would favor a long adult life. She felt that a benign physical environment imposes no age-specific pattern of mortality on a species and life-history pattern would not be easily predicted. As a species moves into unfavorable environments, two opposite patterns would be expected, depending on whether the adult or juvenile stage were more susceptible to mortality. If the adult stage has a greater probability of survival she suggests that an iteroparous life with delayed maturity and reduced reproductive output might be characteristic whereas if survival of the adult stage were uncertain, early maturity and higher reproductive out- put may be favored. In the present study in the equable, “benign” rain-forest, sur- vival of all species tested was relatively high at all stages and there was an accumulation of individuals in the adult size classes. In the Loiza Aldea area, however, mortality was high among juveniles of С. carocollus, and it has the type of life-history pattern predicted by Maiorana to arise from such conditions. Taking all aspects into consideration as well as the data of Randolph mentioned above, it would appear that С. carocollus 296 HEATWOLE AND HEATWOLE arose under a rather equable environment but that it can persist elsewhere because of its behavioral avoidance of extremes and its pre-adaptation to a life-history strategy suitable under conditions of high juvenile mortality. The other forms are restricted to either upland or lowland areas and their characteristics correspondingly were almost certainly influenced by those environments. It is appropriate to ask what aspects of the environment were or are most important as selective agents. The physical factors will be considered first. Moisture has been clearly in- volved as (1) lowland species have a variety of adaptations permitting survival under dry conditions, e.g. secretion of an epiphragm, behavioral avoidance of exposed conditions, and tolerance to desiccation, and (2) mor- tality in young, lowland С. carocollus was observed to be especially high during drought. However, temperature has also played a role as exemplified by the lighter, more reflective color of lowland species and of the upland one that occurs in places exposed to solar radiation. P. acutangula is somewhat of an enigma in that it seems to be adapted to prevent excessive radiant heat loads, yet is exposed to higher desic- cation because of its inability to with- draw completely into its shell. However, its microhabitat, though often exposed to solar radiation is probably humid. It occurs where there is high rainfall, and humidity is probably often high even at the periphery of the canopy. Certainly it would have access to drinking water in the bases of bromeliads and other locations. The ex- posed body consequently may not be a detriment and may even enhance tempera- ture resistance (perhaps via evaporative cooling); such a situation would not be unique, as Mesodon roemeri and Helix aspersa can overcome the effects of high air temperature if sufficient moisture is present (Randolph, 1973; Potts, 1975). In summarizing the evolutionary re- sponses to the physical environment, it can be said that the Puerto Rican camaenids have the following modes of adaptation. C. carocollus: eurytopic, widely distributed, escapes environmental extremes rather than endures them, behaviorally adaptable. C. marginella: rather eurytopic, occupies coastal plain, endures environmental ех- tremes rather than escapes them, less adapt- able behaviorally. P. /ima: rather eurytopic, widely distributed except in wetter, cooler areas, endures rather than escapes extremes. P. acutangula: stenotopic, restricted in dis- tribution, occupies a rather well-defined habitat, endures rather than escapes ex- tremes. Р. /uquillensis: very stenotopic, very restricted distribution, neither endures nor escapes extremes but rather survives only in an equable habitat. Biotic interactions may also be involved as the morphologically most similar species tend to be either allopatric or to occupy different habitats. At this stage it is not possible to distinguish whether such habitat differences reflect the habitat preferences and adaptations of the original stocks, or whether competitive interaction among early stocks led to differences in habitat selection and adaptation to the physical environment. It is possible that once cer- tain stocks became established in Puerto Rico and filled particular niches, later im- migrant species could not become estab- lished because of competition with prior residents. Such a system could result in only mutually compatible or ecologically segregated species becoming established on the island. Whatever the inter-specific inter- action, competition for food does not seem to be important as the type of food eaten is abundant and there seem to be few significant dietary differences among syntopic species (but see Butler, 1976). Heavy predation might keep potentially competing populations below competing levels. However, little is known about pre- dation upon Puerto Rican camaenids; no indications of predation were observed dur- ing the study. Rats are a probable (recently introduced) predator upon young indi- viduals. The presumably protective slime secreted by Р. acutangula when disturbed suggests that perhaps in the exposed situa- tions in which it is found avian predators may be a potential problem. One of the major attributes of snails which has become evident in recent studies is the individual variability in a variety of characteristics. Color polymorphism has long been known in some groups (see above) and has been well studied. However, other more subtle characteristics have also been observed to be polymorphic in the present study. For example, there was a great difference in growth rates among indi- viduals of the same species. Randolph (1973) has shown that a moist substrate and access to drinking water are important for growth of snails. Consequently indi- vidual differences in access to moisture PUERTO RICAN CAMAENID ECOLOGY 297 could result in differences in growth rates among individuals. However, in the гат- forest such a situation is not likely to arise differentially for different individuals and yet great inter-individual differences in rate of growth were observed. It is more likely that such growth differences were genetical- ly based (see section on growth). The polymorphism in growth rate may have far-reaching consequences for popula- tion dynamics and evolution. Oosterhoff (1977) has shown experimentally that growth rate in Cepaea was influenced by various physical environmental factors and by population density. There was also a genetic component. He suggested that changes in growth rate and concomitant changes in size at adulthood would in turn affect juvenile mortality rates and rates of reproduction and thereby contribute to an adjustment of population density. The population data on the Puerto Rican camaenids are not appropriate for testing this model, and more empirical data are required before it can be ascertained whether it is applicable to this group. There are a number of other attributes of Puerto Rican camaenids which may in- fluence their mode of evolutionary re- sponse. For example, they occur in rather dense populations for animals their size, are relatively sedentary and show little migra- tion even as juveniles, are relatively long- lived and breed over a period of years with small clutches at each breeding and have multiple matings among hermaphroditic individuals. The dense population and the great longevity should result in genetic stability as effects of drift would be reduced and the genetic contribution of given indi- viduals would spread over a number of years. Also, as suggested for Cepaea by Williamson et al. (1977) such characteristics mean that adult density may decline for several years with little likelihood of popu- lation extinction. The multiple-mating system among hermaphrodites would also tend to minimize the effects of low densi- ties during population fluctuations (Murray, 1964). The combination of low mobility and high population density means that genetic innovations would spread slowly through a population and that stable genetic complexes would develop. One would expect, therefore, that there would be gradual improvement of fitness in re- spect to the environments occupied but not a marked ability to rapidly exploit new conditions (also low clutch numbers would militate against this). The impression gained is that of a suite of K-selected species. The different species differ in regard to various of the above attributes and may show varying tendencies in this direction. However, all appear to be generally cate- gorized as above. Their chief differences appear to be in the degree to which they are adapted to wet, cool as opposed to hot, dry habitats, and the amount of behavioral plasticity involved. ACKNOWLEDGMENTS This project was carried out under United States Atomic Energy Commission Contract AT-(40-1)-1833, with the Puerto Rico Nuclear Center of the University of Puerto Rico; we are grateful to Dr. Howard T. Odum for his aid and encouragement. The 1970 portion of the study was made possible by sabbatical travel funds from the University of New England. Dr. Hugh Ford read and criticized parts of the manuscript. The following persons aided various parts of the field and/or laboratory work or provided technical assistance: Sheila Blasini Austin, Isabel Colorado, Rita Amadeo, Abel Rossy, Julia Rossy, Zaida Miranda, Ana Vasquez, Sara Armstrong, Carlos Maestri, Frank Torres, Faustino McKenzie, Mary Lou Pressick, Clara Coulsen, Barbara Saylor Done, Joaquin Molinari, Elizabeth Cameron and Harry Wadleigh. Roy Wood- bury identified plants from the Е! Yunque study area and Dr. N. Prakash aided in identification of plant remains in snail feces. Drs. John B. Burch, Phillip Coleman and Carden Wallace provided literature. Viola Watt, Neva Walden, Russell Hobbs and Heather Powell aided in preparation of the manuscript. Dr. М. J. Bishop con- tributed heavily to the section dealing with the evolution and speciation of the Puerto Rican camaenids. LITERATURE CITED AGUAYO, C. G., 1966, Una lista de los moluscos terrestres у fluviales de Puerto Rico. Stahlia, 551-17: ARNASON, Е. & GRANT, P. R., 1976, Climatic selection in Cepaea hortensis at the northern limit of its range in Iceland. Evolution, 30: 499-508. 298 HEATWOLE AND HEATWOLE ARNOLD, В. W., 1968, Studies on Cepaea. VII. Climatic selection in Cepaea nemoralis (L.) in the Pyrenees. Philosophical Transactions of the Royal Society of London, ser. B, 253: 549-593. ARNOLD, R. W., 1969, The effects of selection by climate on the land snail Cepaea nemoralis (L.). Evolution, 23: 370-378. ARNOLD, R. W., 1971, Cepaea nemoralis on the East Sussex Downs, and the nature of area effects. Heredity, 26: 277-298. ВАМТОСК, С. В. & BAYLEY, J. A., 1973, Visual selection for shell size in Cepaea (Held.). Journal of Animal Ecology, 42: 247-261. BANTOCK! "CA RARA PRICEP И REITS, Marginal populations of Cepaea nemoralis (L.) on the Brendon Hills, England. 1. Ecology and ecogenetics. Evo/ution, 29: 267-277. BEARD, J. S., 1944, Climax vegetation т tropical America. Ecology, 25: 127-158. BLINN, W. C., 1963, Ecology of the land snails Mesodon thyroidus and Allogona profunda. Ecology, 44: 498-505. BRAND, T. VON, MCMAHON, P., & NOLAN, M. O., 1957, Physiological observations on starvation and desiccation of the snail Australorbis glabratus. Biological Bulletin, 113: 89-102. BRUSSARD, P. F., 1974, Population size and natural selection in the land snail Cepaea nemoralis. Nature, 251: 713-715. BRUSSARD, P. F., 1975, Geographic variation in North American colonies of Cepaea nemoralis. Evolution, 29: 402-410. BUTLER, А. J., 1976, A shortage of food for the terrestrial snail Helicella virgata in South Australia. Oecologia, 25: 349-371. CAIN, A. J., 1968, Studies on Cepaea. V. Sand- dune population of Cepaea nemoralis (L.). Philosophical Transactions of the Royal Society of London, ser. B, 253: 499-517. CAIN, A. J., 1971, Colour and banding morphs in subfossil samples of the snail Cepaea. Chap- ter 4 (р. 65-92) in: CREED, R., Ed., Ecologi- cal Genetics and Evolution. Blackwell Scien- tific Publications, Oxford. CAIN, A. J. & CURREY, J. D., 1963a, Area effects in Cepaea. Philosophical Transactions of the Royal Society of London, ser. B, 256: 1-81. GAIN, A. 3. & CURREY, 3. D., 1963b, Area effects in Cepaea on the Larkhill artillery ranges, Salisbury Plain. Journal of the Linnean Society (Zoology), 45: 1-15. CAIN AIR CURRE AUD M9SESC "Mine causes of area effects. Heredity, 18: 467-471. GAINS A MI SMCURRE JE D /M968MEcO- genetics of a population of Cepaea nemoralis (L.) subject to strong area effects. Philo- sophical Transactions of the Royal Society of London, ser. В, 253: 447-482. CAIN, А. J. & SHEPPARD, Р. M., 1950, Selec- tion in the polymorphic land snail Cepaea nemoralis. Heredity, 4: 275-294. CAIN, А. J. & SHEPPARD, P. M., 1954, Natural selection in Cepaea. Genetics, 39: 89-116. CAIN, А. J., SHEPPARD, P. M. & KING, J. M. B., 1968, Studies on Cepaea. |. The genetics of some morphs and varieties of Cepaea nemoralis (L.). Philosophical Transactions of the Royal Society of London, ser. B, 253:383-396. CAMERON, В. A. D., 1970a, The survival, weight-loss and behaviour of three species of land snail in conditions of low humidity. Journal of Zoology, 160: 143-157. CAMERON, R. A. D., 1970b, The effect of temperature on the activity of three species of helicid snail (Mollusca: Gastropoda). Journal of Zoology, 162: 303-315. CAMERON, В. А. D. & WILLIAMSON, P., 1977, Estimating migration and the effects of dis- turbance in mark-recapture studies on the snail Cepaea nemoralis L. Journal of Animal Ecology, 46: 173-179. CARTER, M. A., 1968, Studies on Cepaea. II. Area effects and visual selection in Cepaea nemoralis (L.) and Cepaea hortensis. Philo- sophical Transactions of the Royal Society of London, ser. B, 253: 397-446. CLARKE, B., 1960, Divergent effects of natural selection on two closely related polymorphic snails. Heredity, 14: 442-443. CLARKE, B., 1962, Balanced polymorphism and the diversity of sympatric species. Systematics Association Publication 4:47-70. CLARKE, B., 1966, The evolution of morph-ratio clines. American Naturalist, 100: 378-402. CLARKE, B., 1968, Balanced polymorphism and regional differentiation in land snails. Chapter 13 (©... 351-368) in: DRAKE, „Е. №, ЕЯ. Evolution and Environment. Yale University Press, New Haven. CLARKE, B., DIVER,C. & MURRAY, J., 1968, Studies on Cepaea. VI. The spatial and tem- poral distribution of phenotypes in a colony of Cepaea nemoralis (L.). Philosophical Trans- actions of the Royal Society of London, ser. B, 253: 514-548. CLARKE, B. & MURRAY, J., 1962a, Changes of gene-frequency in Cepaea nemoralis (L.). Heredity, 17: 445-465. CLARKE, B. & MURRAY, J., 1962b, Changes in gene-frequency in Cepaea nemoralis (L.). The estimation of selective values. Heredity, 17: 467-476. CLARKE, B. & MURRAY, J., 1969, Ecological genetics and speciation in land snails of the genus Partula. Biological Journal of the Linnean Society, 1: 31-42. CURREY, J. D., ARNOLD, R. W. & CARTER, M. A., 1964, Further examples of variation of populations of Cepaea nemoralis with habitat. Evolution, 18: 111-117. DANSEREAU, P., 1966, Studies on the vegeta- tion of Puerto Rico. 1. Description and inte- gration of the plant-communities. /nstitute of Caribbean Studies Special Publication 1: 1-45, 56-287. DELL, R. K., 1953, A contribution to the study of rates of growth in Paryphanta busbyi (Gray), (Mollusca, Pulmonata). Records of the Dominion Museum, 2: 145-146. DESMARAIS, A. P. & HELMUTH, B. T., 1970, Effects of 137¢, radiation on vegetation struc- ture and optical density at El Verde. Chapter D-2 (p. D-77 to D-102) in: A Tropical Rain Forest, A Study of Irradiation and Ecology at Е! Verde, Puerto Rico. United States Atomic Energy Commission Division of Technical In- ‚ formation Extension, Oak Ridge. DIAZ-PIFERRER, M., 1962, Reproduction of Polymita muscarum Lea, a Cuban tree snail. Caribbean Journal of Science, 2: 59-61. PUERTO RICAN CAMAENID ECOLOGY 299 EDELSTAM, C. & PALMER, C., 1950, Homing behaviour in gastropods. Oikos, 2: 259-270. EMERSON, D. N., 1967, Carbohydrate oriented metabolism of Planorbis corneus (Mollusca, Planorbidae) during starvation. Comparative Biochemistry and Physiology, 22: 571-579. GEISEL, J. T., 1969, Factors influencing the growth and relative growth of Acmaea digitalis, a limpet. Ecology, 50: 1084-1087. GOODHART, C. B., 1962, Variation in a colony of the snail Cepaea nemoralis (L.). Journal of Animal Ecology, 31: 207-237. GOODHART, С. B., 1963, “Area effects’’ and non-adaptive variation between populations of Cepaea (Mollusca). Heredity, 18: 459-471. GRAINGER, J. N. R., 1969, Heat death in Arianta arbustorum. Comparative Biochem- istry and Physiology, 29: 665-670. GRAINGER, J. N. R., 1975, Mechanism of death at high temperatures in Helix and Patella. Journal of Thermal Biology, 1: 11-13. GRIME, J. P. & BLYTHE, С. M., 1969, An investigation of the relationships between snails and vegetation at Winnats Pass. Journal of Ecology, 57: 45-66. GROMADSKA, M. & PRZYBYLSKA, M., 1960, Wplyw temperatur stalychi przemiennych na metabolizm oddechowy slimaka zaroslowego Arianta arbustorum L. Ekologia Polska, series A, 8: 315-324. HEATWOLE, H., 1962, Environmental factors influencing local distribution and activity of the salamander, Plethodon cinereus. Ecology, 43: 460-472. HEATWOLE, H., 1965, Report on snail project. In ODUM, H. T., Ed., The Rain Forest Project Annual Report FY-65 (p. 109-131). United States Atomic Energy Commission Report PRNC-61, Puerto Rico Nuclear Center, 220 p. HEATWOLE, H. & CLARKE, B., in press, Micro- habitat selection by color morphs of the snail, Bulimulus guadalupensis. Caribbean Journal of Science. HEATWOLE, H., MERCADO, N., & ORTIZ, E., 1965, Comparison of critical thermal maxima of two species of Puerto Rican frogs of the genus Eleutherodactylus. Physiological Zool- ogy, 38: 1-8. HEATWOLE, H., ROSSY, A., COLORADO, I., € AMADEO, R., 1970, Effects of radiation on a population of the Puerto Rican tree snail, Caracolus caracolla. Chapter E-1 (p. E-17 to E-24) in: A Tropical Rain Forest, A Study of Irradiation and Ecology at Е! Verde, Puerto Rico. United States Atomic Energy Commis- sion Division of Technical Information Exten- sion, Oak Ridge. HELLER, J., 1976, The biogeography of enid landsnails on the Aegean Islands. Journal of Biogeography, 3: 281-292. HOLDRIDGE, L. R., 1970, A system for repre- senting structure in tropical forest associa- tions. Chapter B-12 (p. B-127 to B-150) in: A Tropical Rain Forest, A Study of Irradiation and Ecology at Е! Verde, Puerto Rico. United States Atomic Energy Commission Division of Technical Information Extension, Oak Ridge. JOLLY, G. M., 1965, Explicit estimates from capture-recapture data with both death and immigration; stochastic model. Biometrika, 52: 225-247. JONES, J. S., 1973a, Ecological genetics of a population of the snail Cepaea nemoralis at the northern limit of its range. Heredity, 31: 201-211. JONES, J. S., 1973b, The genetic structure of the southern peripheral population of the snail Cepaea nemoralis. Proceedings of the Royal Society of London, ser. B, 183: 371-384. LAMOTTE, М. 1., 1959, Polymorphism of natural populations of Cepaea nemoralis. Cold Spring Harbor Symposia on Quantitative Biology, 24: 65-87. LITTLE, E. L. Jr. & WADSWORTH, F. H., 1964, Common Trees of Puerto Rico and the Virgin Islands. Agriculture Handbook 249, United States Department of Agriculture Forest Service, Washington D.C., 548 p. LOMNICKI, A., 1969, Individual differences among adult members of a snail population. Nature, 223: 1073-1074. MACHIN, J., 1965, Cutaneous regulation of evaporative water loss in the common garden snail Helix aspersa. Die Naturwissenschaften, 52: 18. MACHIN, J., 1967, Structural adaptation for reducing water-loss in three species of terres- trial snail. Journal of Zoology, 152: 55-65. MACHIN, J., 1972, Water exchange in the mantle of a terrestrial snail during periods of reduced evaporative loss. Journal of Experimental Biology, 57: 103-111. MAIORANA, V. С., 1976, Predation, submergent behavior, and tropical diversity. Evo/utionary Theory, 1: 157-177. MASON, C. F., 1970, Food, feeding rates and assimilation in woodland snails. Oecologia, 4: 358-373. MATTHEWS, E. G., 1976, /nsect Ecology. Uni- versity of Queensland Press, St. Lucia, 226 p. MURRAY, J., 1963, The inheritance of some characters in Cepaea hortensis and Cepaea nemoralis (Gastropoda). Genetics, 48: 605-615. MURRAY, J., 1964, Multiple mating and effec- tive population size in Cepaea nemoralis. Evo- lution, 18: 283-291. MURRAY, J. & CLARKE, B., 1966, The inheri- tance of polymorphic shell characters т Partula (Gastropoda). Genetics, 54: 1261-1277. ODUM, H. T., 1970, The Е! Verde study area and the rain forest systems of Puerto Rico. Chapter B-1 (p. B-3 to B-32) in: A Tropical Rain Forest, A Study of Irradiation and Ecol- ogy at El Verde, Puerto Rico. United States Atomic Energy Commission Division of Tech- nical Information Extension, Oak Ridge. ODUM, Н. Т. & PIGEON, В. F., 1970, A Tropi- cal Rain Forest, A Study of Irradiation and Ecology at Е! Verde, Puerto Rico. United States Atomic Energy Commission Division of Technical Information Extension, Oak Ridge. OOSTERHOFF, L. M., 1977, Variation in growth rate as an ecological factor in the landsnail Cepaea nemoralis (L.). Netherlands Journal of Zoology, 27: 1-132. OWEN, D. F., 1963, Polymorphism and popula- tion density in the African land snail, Limicolaria martensiana. Science, 140: 666-667. OWEN, D. F., 1964, Bimodal occurrence of breeding in an equatorial land snail. Ecology, 45: 862. 300 HEATWOLE AND HEATWOLE OWEN, D. F., 1966, Polymorphism in Pleistocene land snails. Science, 152: 71-72. OWEN, D. F. & JONES, J. S., 1974, Ecological genetics and natural selection in mollusks. Science, 185: 376-377. PARMENTER, C. J., 1976, The natural history of the Australian freshwater turtle Chelodina longicollis Shaw (Testudinata, Chelidae). Un- published Ph.D. thesis, University of New England, 210 p. PARR, M. J., GASKELL, T. J. & GEORGE, B. J., 1968, Capture-recapture methods of esti- mating animal numbers. Journal of Biological Education, 2: 95-117. PICO, R., 1954, Geografia de Puerto Rico. Parte |. Geografía Fisica, Editorial Universitaria, Rio Piedras, 243 p. POLLARD, E., 1975, Aspects of the ecology of Helix pomatia L. Journal of Animal Ecology, 44: 305-329. POTTS, D. C., 1975, Persistence and extinction of local populations of the garden snail, Helix aspersa in unfavorable environments. Oecologia, 21: 313-334. RANDOLPH, P. A., 1973, Influence of environ- mental variability on land snail population properties. Ecology, 54: 933-955. RICHARDS, A. V. & MURRAY, J., 1975, The relation of phenotype to habitat in an intro- duced colony of Cepaea nemoralis. Heredity, 34: 128-131. RICHARDSON, A. M. M., 1974, Differential climatic selection in natural population of land snail Cepaea nemoralis. Nature, 247: 572-573. RICHARDSON, A. M. M., 1975, Energy flux in a natural population of the land snail, Cepaea nemoralis L. Oecologia, 19: 141-164. RIDDLE, W. A., 1975, Water relations and humidity-related metabolism of the desert snail Rabdotus schiedeanus (Pfeiffer) (Helicidae). Comparative Biochemistry and Physiology, 51A: 579-583. RIDDLE, W. A., 1977, Comparative respiratory physiology of a desert snail Rabdotus schiedeanus, and a garden snail, Helix aspersa. Comparative Biochemistry and Physiology, 56A: 369-373. ROKITKA, M. A. & HERREID II, C. F., 1975a, Position of epiphragms in the land snail Ota/a lactea (Müller). Nautilus, 89: 23-26. ROKITKA, M. A. € HERREID II, С. F., 1975b, Formation of epiphragms by the land snail Otala lactea (Miller) under various environ- mental conditions. Nautilus, 89: 27-32. SCHMIDT-NIELSEN, K., TAYLOR, C. R., & SHKOLNIK, A., 1972, Desert snails: problems of survival. Symposia of the Zoological Society of London, 31: 1-13. SEDLMAIR, H., 1956, Verhaltens-, Resistenz- und Gehäuseunterschiede bei den poly- morphen Bänderschnecken Cepaea hortensis (Müll.) und Cepaea nemoralis (L.) Biologische Zentralblatt, 75: 281-313. SELANDER, R.K. & KAUFFMAN, D. W., 1975, Genetic structure of populations of the brown snail (Helix aspersa). Evolution, 29: 385-401. STIVEN, A. E., 1970, Respiration in the snail Caracolus caracolla and an estimate of the relative density and biomass of litter snails. Chapter 1-5 (р. 1-65 to 1-67) in: A Tropical Rain Forest, A Study of Irradiation and Ecology at Е! Verde, Puerto Rico. United States Atomic Energy Commission Division of Technical Information Extension, Oak Ridge. TE, G. A., 1976, A summary of pulmonate distribution information contained in Zilch’s 1959-1960 monograph: Gastropoda, Teil 2, Euthyneura. Malacological Review, 9: 39-53. TINKLE, D. W., WILBUR, H. M. & TILLEY, S. G., 1970, Evolutionary strategies in lizard reproduction. Evolution, 24: 55-74. UNITED STATES DEPARTMENT OF COM- MERCE, 1962-1966, Climatological Data, Puerto Rico and Virgin Islands Annual Sum- mary 12: 151-158. VAN DER SCHALIE, H., 1948, The land and fresh-water molluscs of Puerto Rico. Miscel- laneous Publications of the Museum of Zoology of the University of Michigan, 70: 1-134. VAN DER SCHALIE, H. & GETZ, L. L., 1963, Comparison of temperature and moisture re- sponses of the snail genera Pomatiopsis and Oncomelania. Ecology, 44: 73-83. WEAVER, J. D., 1961, Erosion surfaces in the Caribbean and their significance. Nature, 190: 1186-1187. WILLIAMSON, P., CAMERON, В. А. D. & CARTER, M. A., 1977, Population dynamics of the land snail Cepaea nemoralis L.: a six-year study. Journal of Animal Ecology, 46: 181-194. WOLDA, H., 1967, The effect of temperature on reproduction in some morphs of the land snail Cepaea nemoralis (L.) Evolution, 21: 117-129. WOLDA, H., 1970, Variation in growth rate in the land snail Cepaea nemoralis. Research in Population Ecology, 12: 185-204. WOLDA, H., 1972, Ecology of some experi- mental populations of the land snail Cepaea nemoralis (L.). I. Adult numbers and adult mortality. Netherlands Journal of Zoology, 22: 428-455. WOLDA, H. & KREULEN, D. A., 1973, Ecology of some experimental populations of the land snail Cepaea nemoralis (L.). Il. Production and survival of eggs and juveniles. Netherlands Journal of Zoology, 23: 168-188. WURTZ, C. B., 1955, The American Camaenidae (Mollusca: Pulmonata). Proceedings of the Academy of Natural Sciences of Philadelphia, 107: 49-143, 18 pl. ZILCH, A., 1959-1960, Gastropoda. Teil 2. Euthyneura. In: О. Н. SCHINDEWOLEF, Ed., Handbuch der Paläozoologie, vol. 6. Borntraeger, Berlin. 836 p. APPENDIX 1. Descriptions of the four study areas Four study areas were selected that col- lectively included all five species (Fig. 2). Two areas were in the Luquillo Mountains: (1) the vicinity of the Е! Yunque Biological Station of the University of Puerto Rico and (2) the EI Verde site used as an experimental radiation study area by the Puerto Rican Nuclear Center. C. carocollus, P. acutangula and P. /uquillensis all oc- PUERTO RICAN CAMAENID ECOLOGY 301 curred at both these sites. In addition (3) a lowland, dry forest near Loiza Aldea, in which С. carocollus was the principal resi- dent camaenid, was investigated. Small numbers of P. /ima and C. marginella were also present and one dead shell of an adult P. acutangula (perhaps transported there by some external agency) was found at this site. The final study area (4) was an open, lowland, park-like area surrounding the faculty residences of the University of Puerto Rico at Rio Piedras. Only P. /ima and С. marginella occurred there. General views of the study areas are shown т Figs. 36, 37 and the climate of each in Fig. 38. Each area will be briefly described. Е! Yunque Biological Station This study area is located in the Luquillo Mountains at the eastern end of the island (elevation 640 m). These moun- tains receive the highest rainfall of any place on the island. Mean annual precipita- tion at La Mina, near the Biological Sta- tion, is well over 500 cm. Distribution throughout the year is relatively uniform, the wettest month (December) averaging over 50cm and the driest one (March) about 25cm. Fog occurs nearly daily. Indi- vidual years may depart from this pattern somewhat. For example in 1965, January to April were relatively dry, May to August were wet and the rest of the year inter- mediate (see Fig. 4). Air temperature is remarkably uniform, mean monthly values ranging from about 19°C to 22°C (Fig. 38). Over short periods of time, temperatures are quite uniform. For example, air temperatures were con- tinuously recorded by a Friez Hygro- thermograph 15cm above the ground at the Biological Station for 6 weeks in 1963; the range in values for the weeks beginning 15 July, 26 July, 9 September, 27 September, 12 October and 10 November were respectively 18.5-19.8°C, 19.8-21.0°C, 18.8-21.0°C, 21.0-24.0°C, 22.0-24.2°C and 20.2-23.5°C. Relative humidities were con- tinuously recorded for the two weeks be- ginning 27 September and 10 November. The ranges were 85-98% and 78-97% ге- spectively. At El Yunque Peak, somewhat higher in elevation than the Biological Station, tem- peratures at the surface of the soil ranged from 11.5-36.5 C between September 1960 and January 1961 with the difference be- tween maxima and minima for individual months being 9-16.5 centigrade degrees. Corresponding values for 2.5cm above the soil were 11.5-32.0°C (individual monthly spans 11.5-13.5 centigrade degrees) and for 2.5 cm below ground were 11.5-29.0°C (in- dividual monthly spans 2-9.5 centigrade de- grees). These data are presented in more detail in tabular form by Heatwole et al. (1965). The vegetation consists of a wet forest type variously called Upper Luquillo Forest (Little & Wadsworth, 1964), Montane Rain Forest (Dansereau, 1966) and Subtropical Lower Montane Wet Forest (Holdridge, 1970). It is characterised by most species being evergreen, a closed canopy at 15m ог more, presence of tree ferns and an abundance of epiphytic bromeliads (Fig. 36). In the El Yunque study area, the Sierra Palm, Euterpe globosa, is the pre- dominant floral component and the forest there is a separate system or subsystem called Palm Brake by Beard (1944) and Odum (1970). In the study area a large number of broad-leafed trees are also pres- ent, including Croton poecilanthus, Cecropia peltata, Alchornea latifolia, Cyrilla racemiflora, Calycogonium squamulosum, Cordia borinquensis, Guarea ramiflora, Miconia sintenisii, Clusia krugiana, Ocotea leucoxylon and Psychotria berteroiana. The tree fern, Cyathea arborea, is also present. The immediate grounds of the Biological Station consisted of low vegetation bordered by a hedge of hibiscus and bananas, beyond which was the forest (Fig. 36). The hedges and the walls of the buildings were occupied by snails and were included in the study. This part of the study area is referred to as the “Station Grounds”: that part consisting of the forest itself is referred to as the “Downhill Plot.” El Verde This area and its vegetation and weather have been described in great detail by Odum & Pigeon (1970). The study sites were located at an elevation of 424-460 m on the western slopes of the Luquillo Mountains; they are in the Upper Luquillo zone of Little & Wadsworth (1964). The forest is Tabonuco Rain Forest (Odum 1970) which corresponds to the Lower Montane Rain Forest of Beard (1944) and Dansereau (1966) and to the Subtropical Wet Forest of Holdridge (1970). The closed 302 HEATWOLE AND HEATWOLE FIG. 36. Views of the El Yunque Biological Station general study area. Upper: Montane Rain Forest near La Mina. Lower: The grounds and building of the Е! Yunque Biological Station, near La Mina. PUERTO RICAN CAMAENID ECOLOGY 303 FIG. 37. Views of the El Verde (upper left), Loiza Aldea (upper right) and Rio Piedras (!ower) study areas. The photograph of the Loiza Aldea area shows the edge of the forest; underneath the canopy, there is almost no herb cover, only leaf litter and limestone rock. 304 HEATWOLE AND HEATWOLE LA MINA E AS ar en 60 40 го Е о 40870 RIO PIEDRAS AE O NET RAINFALL ОЕ МСА М. dis J AS O N OD MONTHS OF THE EE VERDE 30 3HN1VH3dW31 О) Ura Mie ARM Je ad YEAR ALAS AO AND FIG. 38. Mean monthly rainfall and temperature at the study sites. Data not available for Loiza Aldea and consequently data from San Juan are substituted (see Appendix 1). Data summarized from Picó (1954), Odum & Pigeon (1970), Desmarais and Helmuth (1970), and U.S. Dept. of Commerce (1966). canopy extends up to about 25 т; there is an open ground story in deep shade. The vegetation is characterized by buttressed roots, broad thin leaves, bromeliads, lianas, and trunk bark mottled with epiphytic growth (Odum 1970) (Fig. 37). Many of the same tree species occur here as at the Biological Station, including the Sierra Palm. However, at El Verde, the Tabonuco, Dacryodes excelsa, is the principal tree (Odum 1970). Rainfall is abundant throughout the year, exceeding 20 cm even in the driest month (April) and reaches 50cm in the wettest ones (November, De- cember). The seasonal pattern is roughly similar to that at La Mina but is slightly lower in absolute values (Fig. 38). Temper- atures are moderate with slightly higher values than at La Mina; mean monthly values range from about 21.5°C to 24°C. Two subareas were marked out for de- tailed study in relation to the radiation project carried out there. These are known as Center | and Center Il, each a circle of 30 т in radius. Both centers were mapped in detail by the Puerto Rican Nuclear Cen- ter (even to locations of individual trees); these maps were available for the present study. Loiza Aldea This study area was located near sea level in the vicinity of the coastal town of Loiza Aldea. The substrate was weathered limestone, with large chunks of limestone present on the forest floor. Little & Wadsworth (1964) have char- acterized the climax forest in this general area as Moist Coastal Forest (moist in PUERTO RICAN CAMAENID ECOLOGY 305 comparison to the much more xeric forests of the southern coast, not in comparison to the higher altitude forests). However, be- cause of its underlying porous limestone this forest has a more xeric appearance than its name implies. It more closely resembles the Seasonal-Evergreen Forest of the north-western and north central coasts described by Dansereau (1966) who also shows small areas of this forest further east, including the general vicinity of Loiza Aldea. The canopy is nearly closed. Around the edges there is a tangle of vines and other vegetation but inside the forest, the floor is relatively free of low vegetation. Rainfall in this area is sparser and more seasonal than in the mountains; in the absence of rainfall and temperature data from Loiza Aldea, those for San Juan which has approximately the same weather conditions (Pico, 1954) have been used instead. The driest month (February) re- ceives on the average only 6 cm of rain and the wettest one (November) about 17 cm. Temperatures are higher than in the mountains, the coolest months (January, February) have a mean value of about 24° C, and the hottest one (August) one of 27°C (Fig. 38). Rio Piedras The grounds of the faculty residences (elevation 30 m) consisted of lawns, hedges or shrubs, areas of herbs and low vines bordering a small stream, all interspersed with shade trees (Fig. 37). The mean an- nual rainfall is 189cm, with the lowest monthly mean being 8.3 ст for March and the highest one of 21cm for August (Desmarais & Helmuth, 1970) (see also Fig. 38). Temperature is relatively high with mean monthly values ranging from 23.5°С to 27.5°C (February and August, respect- ively). Being a residential area, this study site was exposed to the influence of humans such as trimming of lawns and shrubbery. These influences undoubtedly had some effect on the ecology of the snails. No evidence of overt interference of humans with snails, such as pulling them off trees or stepping on them was ever noticed. In summary, the study areas can be ranked in order of decreasing moisture and shade and increasing temperature as: La Mina, El Verde, Loiza Aldea, Rio Piedras. The study was carried out from Septem- ber 1961 to June 1966, with an additional study period in 1969-70. А preliminary report has appeared (Heatwole, 1965) and results relating to the effects of radiation at Е! Verde have been published (Heatwole et al., 1970). APPENDIX 2. Methods and results Not all aspects could be studied equally thoroughly for all species in all habitats. Table 2 is presented in order to provide the reader with a synoptic view of the cover- age. Activity cycle Snails encountered in the field were tallied as either active or inactive. The criterion for activity was that the tentacles were extended. Inactive snails (tentacles retracted) had their bodies completely withdrawn into the shell if on the ground or in a bromeliad axil, or had the foot extended and adhering to the supporting surface if on a tree trunk or wall. № time were limited, collection of activ- ity data was sometimes omitted for a given day. However, when such data were taken, the activity of all snails encountered on that day was recorded so as to prevent bias. Tables 3 and 4 indicate the activity of snails captured at different times of day and under different weather conditions. Habitat selection Individual trees in the Downhill Plot were categorized as to how many snails used them as their principal daytime resting site and the results tested to see whether trees were selected at random according to either size or species. There were 198 snails, omitting 26 whose principal resting sites were on dead stumps. Expected num- ber of С. carocollus (if distributed at ran- dom) associated with a given species of tree was calculated from the proportion of the total number of trees of that species repre- sented, grouping the 7 least common spe- cies together. The observed values were tested against the expected by a chi-square analysis. Snails were not distributed ran- domly with respect to tree species (Р << 0.005). In a similar analysis cate- 306 HEATWOLE AND HEATWOLE gorizing trees in 5cm intervals of dbh instead of by species, snails were not ran- domly distributed with respect to tree diameter (P<< 0.005); fewer than ex- pected were associated with small trees and more than expected were found on large ones. Because of the large number of cate- gories required, a contingency test could not be carried out to see whether selection of certain tree species occurred inde- pendently of the size effect. However, a graphic analysis permits a rough evaluation to be made (Fig. 6). The Sierra Palm, Euterpe globosa, was the most abundant species in the plot (53% of the live trees) and was represented by a wide size range of individuals. A plot of the number of snails associated with individual trees against tree dbh shows a direct, though loose relationship (Fig. 6). Values for most of the other species fall within or near the belt defined by Е. globosa. In order to ascertain whether distribu- tion was related to topography, the loca- tions of centers of home site ranges (see below) of C. carocollus from the EI Verde study area were plotted on a topographic map (Fig. 7). In Center | the northwest quadrant was virtually devoid of snails. This section had a very steep slope com- pared to other parts of the study area. In Center Il where slopes were more gentle, snails were found throughout most of the area and were much more evenly dis- tributed. Home site range Snails were individually marked (see sec- tion on growth). The location of each snail found by day was plotted on a map at each observation period (33 periods for the forest at El Yunque and 45 periods for the area around the building); numbers of re- captures of individual snails reached as high as 19. Some snails were found hanging within a few cm of the same spot on a wall or tree trunk on many different occasions over a period of more than 8 months (Figs. 8, 9). During some observation periods interspersed among the total number, they were found elsewhere, or not seen at all (probably under leaf litter, in bromeliads or other places where they could not be easily observed). The places a snail occupied dur- ing the inactive period can be designated its home site, and the polygon whose outline is delimited on the horizontal plane of a map by the most peripheral home sites of an individual snail, as its home site range (hsr) (Heatwole et al., 1970). The Asr is much smaller than the home range which would also include the area traversed by the active snai! at night. It should also be emphasized that the Asr should be thought of as a volume, as height on tree trunks adds a vertical component ignored in cal- culating Asr’s in the present study. Many of the snails were captured only one or two times (Table 5). These probably included individuals that died soon after their first capture, transients moving through the area, juveniles hatched near the end of the study, individuals that had only a small part of their Asr included in the study area, or those that had tendencies towards mobility rather than setting up an hsr (Lomnicki, 1969). The main distance between capture points for two-site cap- tures were practically identical for snails that were captured only twice and those that were captured more often (Table 5). The difference was not significant (Mann- Whitney U Test, one-tailed, P > 0.05). The greater the number of recaptures of an individual snail, the more accurately would be the outline of its Asr. Conse- quently, those hsr's based on few ге- captures might not be reliable indicators of the true hsr size. In order to ascertain the minimum number of captures needed to adequately estimate the size of the hsr, the mean areas were plotted against the num- ber of captures on which each was based (Fig. 11). Average estimates were relatively low up to six captures but were higher from seven or more captures. Thus seven captures seem to be about the minimum necessary to delimit an individual’s hsr. Variation among individuals is high and large numbers of individuals are required if the population is to be characterized. For the El Yunque population the mean size of hsr of snails captured 7 times or more was 7.2 т? (Table 5). It was thought probable that because of the lower density of trees around the build- ing than in the downhill plot, that snails in the former place might have larger hsr's than those in the latter. The respective mean areas of the hsr's in the two areas were 6.5 and 8.5 m? respectively, i.e. the differences were in the opposite direction to those predicted. However, the differ- ences were not significant (t = 1.18; 0.4 >Р> 0.2). Many of the snails on the PUERTO RICAN CAMAENID ECOLOGY 307 station grounds were not included in the above analysis as they were found only on the walls of the building and an Asr could not be calculated (see Figs. 8, 9). If juveniles disperse widely it would be expected that they would account for a greater proportion of single captures and captures at two sites than would be true for animals captured more often. Also their hsr's and distance between two capture points would be expected to be larger than those of adults. None of the above predic- tions proved true. The percentage of juveniles among single captures and those captured more than three times were quite close (58% and 55% respectively); only 36% of the animals cap- tured twice or more times were juveniles. The mean hsr size of juveniles and adults (captured seven or тоге times) was 7.67 т? and 7.18 m? respectively; the dif- ference was not significant (t = 0.29; Р> 0.50). Similarly, the distances between capture sites of snails that were captured twice did not differ significantly between adults and juveniles (Mann-Whitney U Test, one-tailed, P > 0.50). Food In addition to direct observations of feeding the diet of snails was assessed by collecting fresh feces, still clinging to the animals, in the field. The samples were kept either dry or preserved in FPA fixa- tive until analyzed. At the time of analysis, feces from all individuals for a given species and locality were lumped together and soaked in water until disintegrated into component pieces. The suspension was stirred until thoroughly mixed and a sub- sample taken and the larger fragments sieved out, floated in 70% alcohol in a petri dish and examined under a binocular microscope at X8 magnification. A grid in 1cm divisions was placed under the petri dish and the fragment closest to each cross- point of the grid identified and listed. There were 100 such points examined per species and locality. Casual inspection under a microscope revealed the presence of many smaller items not identifiable by the above method. Consequently, subsamples were taken from the stirred suspension by a dropper and placed on a glass slide with a drop each of safranine and glycerin, cov- ered with a coverslip and examined at a magnification of X200. There were 2-5 replicate mounts prepared per species and locality. Microscope fields on these mounts were examined at random and all identi- fiable items in the field noted. There were 20-40 fields examined per species. It is important to note that the various percentages indicated in Tables 8 and 9 do not refer to relative weights of different material eaten as the leaf and wood pieces were larger than the unicellular algae; dif- ferent-sized pieces within a given category were also given equal rank. Fungal hyphae were abundant and on occasion were noted to have penetrated cells of vascular plants. These fungi may represent molds growing on the dead leaves consumed. However, it is possible that the fungal growth occurred on the feces after it had left the animal (long strings often adhere to the animal for a considerable time after deposition). Because of the un- certainty as to whether mycelia were in- gested or not they were excluded from Table 9. They represented 17% (Е! Verde) to 29% (Е! Yunque) of the microscopic elements in the feces of С. carocollus. Body temperature Body temperatures (Tg) were measured in the field for all but the smallest species (P. lima) by insertion of a Schultheis quick- registering thermometer, accurate to + 0.1`С, into the opening of the mantle cavity. Air (TA) and substrate (Ts) tem- peratures at the site of the snail were then measured. In several instances black bulb temperatures (Tgp) were also obtained using a black bulb thermometer con- structed by inserting the sensitive end of a Schultheis thermometer into the center of a table-tennis ball, sealing the opening with glue and painting the ball with flat, black paint. Data were collected between 1300 and 1630 h under a variety of weather con- ditions during the period February-April, except for 3 readings on C. marginella in November. Thus, the total daily and seasonal ranges of temperatures are not represented, but rather interspecific com- parison is made of afternoon temperatures during the end of the dry season, the period when the highest temperatures were likely to be experienced. Temperature tolerances Animals for use in experiments on tem- perature tolerance were collected in two 308 HEATWOLE AND HEATWOLE localities, С. carocollus, P. acutangula and P. luquillensis from the Е! Yunque Bio- logical Station and C. marginella and P. lima from Rio Piedras. All animals were acclimated at 21° + 1°C and a 9L15D photo-period for 4-7 days before their tolerances were tested. During acclimation they were kept in containers with screen false bottoms below which a layer of water maintained high humidities. They were sup- plied with boiled Hibiscus leaves daily. The apparatus used to test tolerances consisted of a glass jar closed by a 2-hole rubber stopper through which passed a thermometer and a glass tube to supply air. A snail was placed in the jar which was then weighted and submerged in a 5,000 ml beaker of water. After the snail had re- sumed activity inside the jar, the water, initially at 21°C was heated at a rate of 1°C every 5 minutes. An air stone attached to an aquarium pump bubbled the water and promoted uniform distribution of heat within the bath. Two end-points were easily determined. The first was the drop- ping of the active snail from the wall of the jar to the bottom (the “falling point”). This may have been a behavioral response to high temperatures or perhaps the point at which viscosity of the mucus became too low to permit adhesion to a vertical surface. Its ecological significance is that in nature the snail would drop from a tree trunk to the ground where cooler condi- tions might be found beneath stones or among litter or vegetation. The second end- point was the cessation of response to tactile stimulation (lethal point). After the falling point had been reached, the jar was raised so that the top protruded above the surface of the water after each 5-minute heating interval (1°C rise in temperature). The stopper was removed and the snail prodded with a needle. The first tempera- ture at which it failed to respond was considered the lethal point. Large individuals of С. carocollus and С. marginella had higher temperature toler- ances than did small animals, the relation- ship between lethal point and shell di- ameter being linear (Fig. 16). „The slope of the regression line was 0.041° C/mm diam for С. carocollus and 0.104°C/mm diam for C. marginella; both were significantly differ- ent from zero (t-test; 0.025 > P > 0.010 and 0.050 > P > 0.025, respectively). The slope for С. marginella was not significantly different from that for С. carocollus (t-test; Р > 0.10) and thus size has the same relative effect on lethal limit in both species. The falling point was independent of size. Slopes of regression lines of falling point vs shell diameter did not differ sig- nificantly from zero (t-test; P > 0.50 for С. carocollus and 0.20 >P> 0.10 for С. marginella). Because of the dependency of lethal point on size, juveniles of С. carocollus and С. marginella were not used in making interspecific comparisons of this end-point; tests of other species involved only adults. Juveniles and adults were grouped, however, for interspecific comparisons of falling point since this showed no dependency on size. Effect of temperature on food- deprived animals The following experiment was designed to evaluate thermal effects on depletion of stored energy reserves in С. carocollus and С. marginella. С. carocollus was collected at El Yunque and С. marginella at Rio Piedras. They were acclimated for 8 days at 20° + 1°C and a 9L15D photo-period in a closed glass terrarium with about 1 ст of water at the bottom to provide a relative humidity near saturation. During this period it was as- sumed that all of the snails emptied their digestive tracts completely as no defecation occurred during the post-acclimation period. After acclimation, the С. carocollus snails were divided into two groups, each with approximately the same size-distribu- tion. One group was maintained under the acclimation conditions (20°C and near 100% r.h.) without food until all were dead. The second group was treated differ- ently only in that they were transferred to а cabinet at 30° + 1°C. These two tempera- tures were chosen because they represent typical ones in the field in the uplands and lowlands respectively during the end of the dry season when unfavorable conditions are most likely to occur (see section on body temperature). The lights were usually turned on at 0800 and off at 1700 h. Minor departures from this photo-period were the same for both groups. C. marginella was treated identically except that only adults were used. Mantle water and moisture content C. carocollus often carried a certain amount of water in the mantle cavity. It PUERTO RICAN CAMAENID ECOLOGY 309 was considered that such water might be a store usable during dry periods and that the quantity might accordingly vary season- ally or spatially depending on immediate environmental availability of water. To test this, at monthly intervals for part of the study, a sample of 25 snails from each of various microhabitats in the Loiza Aldea and Е! Yunque study areas were forced to yield their mantle water for measurement. All snails were taken well away from mapped areas in which population studies were being carried out. Each snail was repeatedly prodded until it had retracted into its shell as far as it could go. In so doing the mantle water was squeezed out and could be removed from the shell with a hypodermic syringe and the volume meas- ured. During the driest period of the study, a sample of 25 adult, inactive snails were collected from the El Yunque (23 Jan. 1965) and Loiza Aldea (15 Feb. 1965) study areas for ascertaining whether the drier condition of the lowland site would be reflected in reduced hydration levels of the snails there. The mantle water was first removed and measured and then the ani- mals killed and the shell, soft body parts and gut contents separately weighed. The soft body was then oven dried at 105°C to constant weight for determining moisture content. The samples of С. carocollus taken from El Yunque and from Loiza Aldea differed slightly in water content of the body and the differences were significant (t = 5.68, P< 0.001; two-tailed). The snails from El Yunque had a significantly greater average absolute body weight than those from Loiza Aldea (t = 6.23; P<0.001; two- tailed). The mantle water of species other than С. carocollus was not studied extensively. However, on 22 March 1965, 3 P. /ima were examined for mantle water; none contained any. On 28 August 1965, a sample of 25 adult Р. /uquillensis snails were obtained at the El Yunque Biological Station. Mean mantle water was 0.86 т! ($.Е. = 0.143; Range = 0-2.2 ml). Desiccation Standard laboratory desiccator jars with screen separating the upper and lower com- partments were used as experimental cham- bers for ascertaining evaporative loss, sur- vival time and vital limit of water loss under several conditions of temperature and humidity. One group of desiccators was maintained at 30” + 1°C; some of these contained water in the bottom com- partment (atmosphere nearly saturated) and the others contained CaCl, (dry atmos- phere). A second series of desiccators was maintained at 20° + 1°C, again subdivided into humid and dry groups. Snails used in the experiment were first kept in glass terraria with access to stand- ing water but not food, for 48 hrs. Their mantle water was then removed and 5 in- dividuals were placed in each desiccator jar. For some species and treatments juveniles and adults were both tested, in others only adults were used. In addition, a separate series of С. carocollus and С. marginella were treated identically except that the snails were allowed to retain their mantle water. The experiment was carried out at a photoperiod of 9L15D. The snails were not fed during the experiment nor did they have direct access to water. At intervals of one to three days the animals were removed, checked for pres- ence or absence of either a clear mucus membrane or an epiphragm, weighed, and returned to the chamber. Each terrarium was checked daily for dead animals. These were removed, weighed, and the vital limit of water loss calculated as the amount of weight lost in % of the total fresh weight at the beginning of the experiment (includ- ing shell). Since the experiment lasted a long time, a part of the weight loss was undoubtedly due to metabolic losses of carbon and constituted an unmeasured source of ex- perimental error. Neither the vital limit nor the rate of loss per unit body weight of the snails originally containing mantle water could be calculated as an undetermined part of their original weight resulted from their water load. Survival time during desiccation depend- ed on animal size; larger animals survived for longer periods than did smaller ones. In all cases except one (C. marginella at 30°C in dry air), mean survival of juveniles was lower than for conspecific adults under the same treatment; the correlation of size and survival time was significant in 7 of the 13 testable treatments (Table 11). The means of juveniles can not be used to compare treatments or species as absolute levels of survival times would depend on the relative 310 HEATWOLE AND HEATWOLE proportions of very young to nearly sub- adult individuals in the juvenile sample. Consequently, only data from adults are used for subsequent analyses. Within the size range of adults, size did not influence survival time (Spearman Rank Correlation Tests; P >0.05 in all treatments for every species, one-tailed). The vital limit is the percent of hy- drated body weight which can be lost as water before death occurs. In C. carocollus and P. acutangula the mean vital limits obtained during different treatments were not significantly different (Dice-Leraas graphic test) and consequently for these two species, values from all treat- ments were combined for interspecific com- parisons. In the remaining 3 species the mean vital limit obtained from desiccation at both humidities at 30°C and in dry air at 20°C were not significantly different and were grouped. However, the mean vital limit of snails of these three species sub- jected to moist air at 20°C was much (and significantly) lower than that for other treatments of the same species. This curi- ous result was probably not a chance phe- nomenon as it occurred in 3 of the 5 species (Fig. 21). No ready explanation for the function of such a response comes to mind. This result required that data from 20°C in moist air could not be validly combined with those from other treatments and consequently they are presented sepa- rately. Vital limit for a given species and treat- ment showed no significant correlation to size even in samples including juveniles and adults (Spearman Rank Correlation Test; P > 0.10 in all cases except one significant one; Table 12). It can be concluded that there is usually no ontogenetic change in vital limit of water loss in these species and that the difference between adults and juveniles in survival probably reflects differ- ences in rates of loss rather than in vital limits. To test this hypothesis, the rates of loss were analyzed. In 5 of the 11 compari- sons of adult and juvenile mean water loss, the observed values were higher in adults than in juveniles but in only one of these cases were the differences significant (Table 13). In the 6 remaining cases, juvenile mean rates of loss were higher than those of adults, the differences being significant in 3 of them (Table 13). However, within the adult size range, there was never a signifi- cant correlation between rate of loss and size (Spearman Rank Correlation Test; P >> 0.05 in all cases). Role of membranes Data on water loss from snails when clear membranes sealed their shell apertures and from periods when such a membrane was absent were available for 6 individuals. In 3 of them rates of loss averaged higher when there was no membrane, in 2 it averaged higher when there was, and in 1 the mean rates were identical for the 2 situations. A Wilcoxon Matched Pairs Signed Rank Test (matching the data by individual animals) indicated no significant difference between water loss values be- tween the two states (P >>0.05, one- tailed). There were sufficient data from snails with epiphragms to relate survival time of individual snails to the proportion of the time in which they had an epiphragm. A correlation analysis revealed that survival time was positively correlated with propor- tion of the time an epiphragm was in place (Spearman Rank Correlation Test; N = 10, г. = 0.76; 0.05 > P > 0.01, one-tailed). The line fitted to the data by least squares had the formula: Survival time = 21.7 + 0.54x where x is the proportion of the time the snail is protected by an epiphragm. A linear relationship between the two variables ap- peared reasonable from visual inspection. Other groups of С. marginella that formed epiphragms did so over too small a range of time to permit correlation analysis. The desiccation histories of all individual snails of the 30°C group that had days both in which there was an epiphragm and those in which there was no membrane of any kind, were examined on a day by day basis. A Wilcoxon Sums Matched Pairs Signed Ranks test in which the mean rate of loss when an animal had no membrane was matched against that of the same indi- vidual when it had an epiphragm, was carried out on both adults and juveniles (N = 16). The rate of loss was significantly lower when the epiphragm was present (0.05 > P > 0.025; one-tailed). The degree of reduction in water loss afforded by the epiphragm depended on the conditions of desiccation. The reduction was 9.7% in adults at 30°C but not over desiccant and 21.7% for those in the containers with CaCl, at 30°C. PUERTO RICAN CAMAENID ECOLOGY 311 Response to relative humidity For testing responses to relative humidi- ty, a gradient (0% r.h. to near saturation) employing various saturated salt solutions to maintain proper levels of relative hu- midity was set up in a chamber at 20° + 1°C as described by Heatwole (1962) ex- cept that small doors permitted introduc- tion of test animals without having to open the lid. Two adult C. carocollus snails were tested separately by introducing them into the chamber at the point where relative humidity was 50% and their subsequent location in the gradient periodically ob- served. Effect of dry periods on survival In order to ascertain size-structure of C. carocollus at death, the leaf litter at Loiza Aldea was carefully searched on 6 February 1965 for dead shells. Because small shells might disintegrate faster than larger ones and hence cause a bias in the size-structure obtained, only recently dead shells in which the periostracum was still intact were considered. Those which had weathered white or in which part of the periostracum had flaked off were discarded. Following this sample, there was an extended period without rain and the opportunity was taken to ascertain whether size-structure was altered by the drought; correspond- ingly a second sample was taken on 22 March 1965 for comparison. The samples of recently dead shells were divided into adults and into two size-classes of juveniles (less than 20mm in diameter and 20 or more mm in diameter). If the dry period had caused no dif- ferential mortality among size classes then the distribution of numbers in the 3 size classes should be the same in both samples. The pre-drought distribution was thus used as a basis for calculating the expected distribution for the post-drought sample and the observed values compared to it and tested by Chi-square. The two distributions were significantly different (X? = 9.5; 4.4. = 2; P< 0.01). The difference resulted from a relative increase in the proportion of animals in the smallest size category (from 29% to 48%) and a corresponding decrease in those in the larger juveniles (46% to 39%) and adults (25% to 19%). It was difficult to ascertain the size- structure of animals at mortality in the El Yunque area as few shells were found that retained the periostracum; it is probable that shells deteriorate more rapidly under the wetter conditions of the uplands. Despite a prolonged search through leaf litter only a small sample of any camaenid species was obtained. However, shells of the sagdid Platysuccinea portoricensis were obtained. The data for all species from both areas are compared in Fig. 24. Despite the small samples of С. carocollus from the upland it is clear that at El Yunque the majority of snails dying are large adults, rather than the very small juveniles, i.e. the opposite pattern from that prevailing in the Loiza Aldea region. P. luquillensis also had highest mortality among adults. This may well be a feature of the montane rain forest as a similar pattern also occurred in the sagdid snail from Е! Yunque. Only one P. acutangula dead shell (adult) was obtained in the El Yunque sample and only 5 C. marginella (3 juveniles, 2 adults) and 6 P. /ima (3 juven- iles, 3 adults) shells in the Loiza Aldea area, so these species unfortunately cannot be adequately compared. Reproduction At El Yunque copulation occurred from December to July inclusive but no records were obtained from August through No- vember for any year. Inasmuch as (1) field observations were primarily carried out by day when snails tended to be inactive, and (2) most copulations were observed during periods of cloudy weather or light drizzle, evidence of observed copulation might be influenced by weather. However, at El Yunque the largest numbers of copulations were recorded during the 3 driest months and none during the wettest season, the opposite expected on the basis of the above mentioned source of bias. Hence, it appears that seasonal differences in occur- rence of copulation are real and that mat- ing occurs chiefly in the driest months (February to April) but extends over an 8-month period. The results from the some- what drier El Verde site differ in that most copulation occurred later (June and July) when moisture was not at its lowest level. At Loiza Aldea, the driest study area in which С. carocollus occurred, 2 of the 3 copulation records obtained were in August and October, both wet months (Fig. 4). Some information was obtained on the 312 HEATWOLE AND HEATWOLE mating histories of individual C. carocollus. Snail no. 12 was known to mate at least twice within a 2-month period (on 22 Jan. 1963 with snail no. 49 and 11 March 1963 with snail no. 166). Snails no. 25 and no. 116 provided similar data, the former mated on 23 March 1963 with snail no. 38 and on 6 April 1963 with snail no. 75; the latter mated with snail no. 94 on 20 April 1963 and with snail no. 322 on 20 May 1963. Snail no. 395 was mating with no. 394 on 9 July 1964 and about 1 year later (25 July 1965) it was seen copulating with an unmarked snail. Of the 27 snails observed copulating that had been previously marked, 16 were first marked as adults, all but 4 of these had been first marked less than 1 year before they were observed copulating and little information on duration of the re- productively active span was obtained from them. The other 4 were observed copulat- ing the following periods of time after first being marked as adults; 2 years 11 months, 3 years 4 months, 1 year 1 month, 2 years 8 months. In addition two of the snails first marked as juveniles and later observed copulating were captured sufficiently often in between to know approximately when they achieved adulthood. They were known to copulate 1 year 6 months and 2 years 8 months after having become adult. In view of the facts (1) that many of the above snails were already adult when first marked and (2) that the longevity of the species is up to at least 10 years (see section on longevity), the reproductive period is prob- ably much longer than the above figures, which must be considered as minimum values. Eggs were only infrequently found and all those identified with certainty were of Р. acutangula except for one clutch of C. carocollus eggs; the former species definite- ly accounted for at least 8 of the 13 clutches found. The other clutches were in early stages of development and could not be identified but were probably P. acutangula. The eggs of P. acutangula are glossy white, with smooth hard calcareous shells that shatter when broken; they re- semble miniature chicken eggs in shape. The unidentified eggs were similar in size, appearance and number per clutch (Table 15). The single clutch of eggs of C. caro- collus (laid in a plastic bag by a captive snail) were longer and thinner than P. acutangula eggs. TABLE 15. Characteristics of eggs and clutches of camaenid snails at El Yunque. Data on clutch size were omitted for clutches found in leaf litter or on ground which were probably incomplete. The sample size of eggs measured were N = 7 for P. acutangula and N = 11 for the unidentified category. Unidentified P. acutangula eggs Mean Range Mean Range Egg length 9.6 8.3-10.4 8.7 7.9-9.9 Egg width 7.6 6.9-8.2 6.6 6.1-7.3 Clutch size 14.8 10-20 157. 1222 Eggs were encountered over a wide span of the year, at least from November through April for P. acutangula and per- haps for longer periods if the unidentified clutch in August were in fact of this spe- cies (Table 14). However, most eggs were found during the early (dry) part of the year. It is possible that during the wet season eggs of this species are in danger of being flooded. The sites of oviposition in- clude the wettest microhabitats in the forest and the eggs are thus probably pro- tected against desiccation during the dry season. Most P. acutangula clutches were found in the axils of bromeliads or of banana trees, both of which tend to collect water in the base. Three clutches were found on the ground, either in leaf litter or under stones. The unidentified clutches were all observed in these same 3 situa- tions. In the early part of the season most clutches were in early stages of develop- ment (and were not identified) whereas from January on, most were found at or near the hatching stage. There were several exceptions. One clutch of 11 P. acutangula eggs was found on 29 November 1964 in a bromeliad axil. The entire plant was brought into the laboratory and kept at about 20°C in a beaker of water. On 14 January 1965 (about 1% months later) most of the snails had just hatched and 1 (8.0 mm in greatest shell diameter) was just leaving the egg. In addition 1 egg contained a half grown embryo (greatest shell diame- ter 3.4mm) and one was merely fluid- filled. The other identifiable clutch that was found in an early stage of development was found on 22 January 1966. There were 9 eggs in the axil of a bromeliad accom- panied by an adult P. acutangula. Several days later there were 16 eggs present but PUERTO RICAN CAMAENID ECOLOGY 313 all subsequently disappeared before com- pleting development. Very little change in the state of the gonads was observed in C. carocollus, the only species dissected in large numbers. No individuals containing shelled eggs were ever dissected. Mean oviduct width of snails in the samples showed little seasonal varia- tion (3.5 тт-6.2 mm) with no consistent seasonal trends. Diameter of the ovotestis (monthly means 2.1 тт to 3.4 тт) simi- larly showed no consistent seasonal trends. However, if it is assumed that the weight of the albumen gland is an indication of reproductive state an idea of the reproduc- tive cycle can be obtained (Fig. 24). The appearance of albumen glands dif- fered considerably. The larger, presumably active ones were dark whereas smaller ones were paler. In July 1965 (Loiza Aldea) notes on gland color were made for each snail. The weight of dark glands ranged from 0.76 д to 1.52 д (x = 1.21; SE = 0.11; N = 8) whereas pale ones had equivalent values of 0.07 - 0.94g (x = 0.33; SE = 0.07; N = 17). If a weight of 0.85g (mid-point between smallest dark gland and largest pale gland) is used as the criterion for the albumen gland to be considered functional, the various samples can be divided into animals capable of producing albumen and those not capable of doing so. There was a seasonal pattern in relative proportion of those two types of animals (Fig. 26). There are few direct data relating albumen gland size to reproductive activity. One copulating pair of C. carocollus was captured at Loiza Aldea on 29 April 1965 (when albumen glands were generally low). One snail had an albumen gland of 0.463 д and the other one of 0.069g, both low values; the latter almost certainly indicates an inactive gland and the former was well below average values at peak season. An individual with a relatively small albumen gland (0.09 g) captured at El Yunque on 28 August 1965 had the duct of the spermatheca filled with sperm and 3 ani- mals with spermatophores on the same date had albumen glands of only 0.18, 0.20 and 0.42 g respectively. All these data suggest that mating may take place well in advance of the time the albumen gland is functional for the season. Sperm are probably stored for a considerable period of time prior to fertilization. The albumen gland cycle was not studied in detail in other species. However, one sample of 25 P. /uquillensis was ex- amined on August 28, 1965. Mean albumen gland weight was 0.91 g (В = 0.003-1.62 g; SE = 0.11). Oviduct diameter varied from 2.1-6.5mm (x = 4.2; SE = 0.14). Seven (28%) of the animals had the oviducts con- voluted and consequently had oviposited previously. This is a minimum estimate of animals that had previously bred as it is not known whether a convoluted oviduct will return to its former smooth condition between consecutive reproductive periods. Growth and longevity Snails were individually marked by painting numbers on their shells with non- toxic plastic paint. Their locations at each time of capture were plotted on a map and data on height above ground, activity, and type of substrate were recorded. Diameter of the shell, defined to be the distance between the peripheral edge of the lip and the opposite side of the shell along an imaginary axis passing through the apex of the whorl was measured with a Vernier calipers. However, it was difficult to make this measurement without dislodging the snails and field measurements of diameter displayed а rather large error. Conse- quently, it was felt that a more accurate picture of growth could be obtained from the increment in shell length (linear dimen- sion of the shell). In the laboratory a series of snails was used to determine the rela- tionship between diameter and length. Each shell was coated with glue and a string coiled on the upper surface midway Бе- tween the inner and outer edges of the whorls. When it was in place ink marks were made on the string at each complete revolution. The string was then uncoiled and the linear distances measured and plot- ted against the corresponding measurements of shell diameter (Fig. 39). When a juvenile snail was first captured its total shell length was determined by measuring the diameter and reading the length from Fig. 39. The edge of the lip was painted and upon recapture increments in length could be directly obtained by measuring the distance from the old lip mark to the new edge of the lip midway between the outer and inner edges of the whorl (Fig. 40). The new lip was then painted. Since snails do not grow after the umbilicus is closed and the lip becomes thickened, these two 314 HEATWOLE AND HEATWOLE 60 Е 50 Е Polydonte s F acutanqula 40 a 30 WwW r ш = 20 о го 60 100 140 LENGTH Caracolus caracollus Polydontes luquillensis 180 220 260 300 340 380 (тт ) FIG. 39. Relation of shell diameter and shell length in three species of camaenid snails (see Appendix 2 for further explanation). features were used as the criteria for adult- hood. Cameron & Williamson (1977) found that handling and marking snails increased the dispersal of animals from the study area. We have not evaluated this effect. However, in contrast to their methods (re- moval of snails from the study area, taking them to the laboratory for 2-4 days, mark- ing them and then scattering them around the study area again), we disturbed our snails very little, and the handling effect should have been minimal. The shapes of the three species on which growth data were obtained were different. Consequently, unit increase т shell diameter represented different in- creases in shell length among the various species. The curve relating these two measurements (Fig. 39) was used to obtain the shell length of snails at the time they were initially marked. Thereafter changes in shell length were measured directly. Growth studies on P. acutangula were based on fewer individuals than for C. carocollus. However, hatchlings were dis- covered on several occasions in the former species and the lower end of the growth curve is more complete. Clutches of eggs in the field were observed and young snails marked soon after hatching. Diameter at hatching ranged from 8.0mm to 10.3 mm (x = 9.02, SE = 0.268) or a mean shell length of about 22mm. Increase in shell length was practically linear, at least during the first month of life. For a clutch first observed on 14 January 1965, growth rate averaged about 0.56 mm per day (Fig. 29). Other, but less complete records gave growth rates ranging from 0.30 to 0.46 mm per day. These rates of growth of hatch- lings in the field are lower than the maxi- mum possible. Fourteen hatchlings raised in the laboratory at 20°C and supplied with abundant food (boiled Hibiscus leaves) and powdered calcium carbonate grew linearly at an average rate of 0.75mm per day for the first 22 days after hatching. After that time some died and the rate of growth of the remaining ones decreased (Fig. 30). For the larger juveniles, growth rates seemed to be relatively uniform throughout the year = FIG. 40. Photograph of Caracolus carocollus no. 354 showing method of marking. Notice the series of paint marks at the locations of the lip at different recapture periods. with no apparent seasonal limits as oc- curred in С. carocollus. However, гесар- tures are too few for most snails for precise details of growth pattern to be evident (Fig. 31). Р. luquillensis marked on a retaining wall near La Mina similarly showed rather linear increases of shell length throughout the year except for a few individuals that showed elevated growth rates in autumn (Fig. 32). It is estimated from these events PUERTO RICAN CAM AENID ECOLOGY 315 that it would take about 2 years to reach maturity. Growth data from other species were insufficient for interpretation. Population biology Because of the differences in diurnal microhabitat between juveniles and adults (see Habitat selection) it was difficult to quantitatively sample the population for size structure. Additional problems arose from the fact that juveniles are harder to see than adults. However, a special situa- tion permitted circumvention of these dif- ficulties for 2 species (С. carocollus and P. /uquillensis) and we obtained what we be- lieve to be a reliable estimate of size- structure. A stone retaining wall 43 m long located near La Mina was periodically censused by flashlight on humid or rainy nights when the wall was wet and when snails of all sizes were actively foraging over it. The relatively smooth surface permitted equal detection of all size classes when a slow and careful search was made. The wall was systematically searched, removing all snails Нот a given area before proceeding further, thereby eliminating a selective bias of the investigators. Sampling was termi- nated when about 100 C. carocollus had been found or unless weather changes like- ly to inhibit snail activity (e.g. wind beginning to blow) occurred. P. /uquillensis was not as abundant and sample sizes are consequently smaller. Occasional P. acutangula were encountered but not enough to be of use in estimating size structure. No comparable situation was found in the lowlands and no comparative data оп С. carocollus there, or on the strictly lowland forms was obtained. Estimates of population density were made by the Jolly (1965) method using the recapture data from marked snails. vievitetitneup SONS dom lanoltihbA лоу ete or Vorl ow afina] м 120 ee mud curia lai a muswoH zul: cart ou ib to noitnevmunig bettimmi ho N bs E 2330 С yor shi a ow 157 banicido “aw bas (anal Pi amiss pica joer 6 sos oval 4 = WLIO и elo m CE fiw brıinietsi anote о vd bozwariso (ipsiaoiro ao ВМ a) 1 паи’ atrbin ype: 19 bimur по 7Hpih on Us Yo ооо ии brie Tow cow New om ent 71 evo nigeria? Yievitan NOW 292 Вир finvieq sosie Doom vVvisvlals wols 6 haw, suse}. eae Via A zw ем эНТ .obera tew (fre lois bie slisne Ms omnivore: baue vils/is meer goibagoo sold ayu ‘nsvig | io Bs р a а be $ ы t . = en (644, tos Ae, 2 Qi rele VUS a ASTON DO MONTE infies pew pra motépirasvni ofr to on wullosenan зако) fo armaron | ai? lich quien lo bortrem wer inter aitor ddramnih ee! VE AS hal yt ber amos, 2 ar syade nor, Beran Wall RE ¡ASAS MEN" eine +6 ВО need bé 4,9) ‘gis Ven? TIME, a yl À | sare Au N he HOT MOOG Wold ah nn 79839 y RN CAT Civics ot ое. Unk Trmemuis ua hen am ey НМ ar EEE = I С нет Wirg Uneinos CIS fis WH? SAAS Oe a wun fon rye FEAT Joan "SAN оо LENA JON НИ 15015 br LP A 3 Е. a fai bey a ad OP TOULON вуз”. ni: und ЗА № ii qu OBVIO A tt 9} MO ВИ ами £ ' р irn Br avi AY iis (aoe) ah bins atin Wie st ort mu bnivie ont бы: rer Ч rfi “Bandit ix vs so WA NG anne ДАХ ao HER ny nf #19 a ae el a Cn а impor bralwal Vote ПОЛУ уров: 10° he ay У Утв, ry ОЕ бо poa ем ery E In wal $ VO) CHATA Я. Er) el" bombard (caer? Дот, art yd sham write Ft ang a van! alice beam mon Bist Mira A mt AAW Эт sho (Gz u ) L № м : ob | | ' nt raw > ‹ | | tower, thar aa р 3% ii, e г в. 4 sl Y м ori » a N “4 >lbisciós. WINE cel И ма Г uy га оси 4 Pay ; O min 00 y Ar y mer iwys alter hatchung АТОМ ya | sonne ced aa ivwetete 07 yrowtih E Growt ) D (GASTOS One CCE IF Я. 30% Far : = ' yá verlieg, port reves зе 1 rw hice eed day у | ny aniform . Pucha he NU MALACOLOGIA, 1978, 17(2): 317-330 CHANGES OF GENE FREQUENCY IN СЕРАЕА NEMORALIS OVER FIFTY YEARS James Murray! and Bryan Clarke2 ABSTRACT Populations of the land snail Cepaea nemoralis inhabiting the coastal sand dunes at Berrow, Somerset, England, have been under observation since 1926. The original survey was carried out by Professor A. E. Boycott and Captain Cyril Diver, who not only collected an extensive series of samples but also produced a careful map of their localities. Thus it is possible for us to make direct comparisons with samples taken in 1959-60, 1963, 1969, and 1975. Clinal patterns in the distribution of morph frequencies have remained essentially the same over the fifty-year period. However, there have been overall changes in the frequencies of two of the principal morphs. The allele for brown shell color (CP) has shown a decrease in frequency, while the allele for a single central band (UY) has shown an overall increase. The direction and consistency of these changes have been investigated by use of the С statistic and the analysis of covariance. These methods show clearly that there are differences in the frequency of both morphs among the localities and over the years. Despite one locality that shows an increase in brown, there is a consistent overall decrease in this morph, representing a mean selective disadvantage of about 3% per generation. With respect to the midbanded form, analysis of covariance indicates that there are real differences among localities in the direction or the amount of change. If we nevertheless calculate the mean selective advantage of this morph, we find it to be about 4% per generation. These differences can be interpreted in terms of changes in the habitats where the snails are found. INTRODUCTION Populations of the land snail Cepaea nemoralis (Linn.) inhabiting the sand dunes at Berrow in Somerset, England, provide an unusually good opportunity for measuring long-term changes in the frequencies of genotypes. Studies of this sort can be ac- complished only in rather special circum- stances. First, there must be a set of de- tailed, accurate, and localized data for some populations in the past. Second, the organism must have genetics sufficiently well understood for the reliable scoring of genotypes. Third, there should be evidence of genetic continuity over the intervening period so that continuous evolution may be distinguished from extinction and recoloni- zation. Finally, some idea of the selective significance of the characters under study is desirable. It is not surprising, therefore, that well-documented studies of gene frequen- cies over long periods are not at all common. It is even more unusual for all of the desirable conditions to be realized in a single example. Some of the best-docu- mented cases are the studies of color polymorphisms in the lady beetle Harmonia (Komai, 1956; Komai & Chino, 1969) and in the moths Biston (Kettlewell, 1965; Ford, 1971) and Panaxia (Fisher & Ford, 1947; Sheppard & Cook, 1962; Lees, 1970), and the work on chromosomal polymorphism in Drosophila pseudoobscura (Dobzhansky, 1958; Dobzhansky et al., 1964, 1966; Anderson et al., 1975). There have also been several studies of long-term changes in populations of the polymorphic land snail Cepaea nemoralis. Besides our own work based on the collections of Boycott and Diver (Clarke & Murray, 1962a, 1962b; Clarke, Diver & Murray, 1968) there are studies by Good- hart (1956, 1958) and van Heurn (1943, 1945). Some investigations have extended the time scale by comparing present-day and subfossil collections (Diver, 1929; Cur- rey & Cain, 1968; Cain, 1971). The most extensive and carefully docu- mented series of historical collections of Cepaea was taken by Professor A. E. Boycott and Captain Cyril Diver in 1926 on the sand dunes at Berrow. In earlier 1Department of Biology, University of Virginia, Charlottesville, VA 22903, U.S.A. 2Department of Genetics, University Park, Nottingham, NG7 2RD, England. (317) 318 MURRAY AND CLARKE papers (Clarke & Murray, 1962a, 1962b) we compared their data with collections taken in 1959 and 1960 from the same sites. We found that the populations were generally stable, but that there were signifi- cant changes in the frequencies of two of the principal morphs. We calculated the selection coefficient associated with a de- crease in the frequency of brown shell color to be about 6.2% and that associated with an increase of the single-banded (00300) phenotype to be about 5.2% per generation. Since the time of our first collections at Berrow, we have visited the area on a number of occasions, most recently in the summer of 1975. We are therefore able to report, in this paper, observations on gene frequencies extending over a total period of about half a century. METHODS The sampling area The sand dunes at Berrow have been described in a previous paper (Clarke & Murray, 1962a; see also Boley, 1944). They comprise an area of coastal dunes bordering the Bristol Channel for about 5km be- tween the villages of Berrow and Brean. The width varies between 300 and 600 т. The northern half of the area has been extensively disturbed, and our collections have been taken largely from the more stable southern half. Here the beach is bordered by a line of semi-mobile dunes reaching a height of about 5m. Farther inland there is a higher ridge of stabilized dunes that rises in places to 20 m. Between the two ridges and also on the landward side of the main ridge, the terrain consists of a chain of “slacks”” or rolling hollows that are frequently quite damp and occa- sionally contain standing water. А golf course threads its way along these slacks. Although the greens and fairways obviously represent a serious intrusion into the natural ecology, their net effect has been to resist the encroachment of more disrup- tive human developments. The flora of the dunes has been de- scribed by Boley (1944). Cepaea nemoralis is here particularly associated with clumps of /ris foetidissima but is more or less continuously distributed wherever’ the dunes are covered with herbaceous vegeta- tion. The road bordering the dunes on the eastern side marks a major change in the ecology. This transition is emphasized by the appearance of Cepaea hortensis, a close relative of C. nemoralis hardly ever found on sand dunes in England. Sampling Fig. 1 is a sketch map of the area showing the localities from which collec- о OO 200 300400500 YARDS FIG. 1. Sketch map of the sand dunes at Berrow, Somerset, England. The solid black patches represent collecting areas, drawn to scale. The prefix D indicates a locality sampled by us alone. Double dotted lines represent roads. The continu- ous line shows the approximate high water mark. CH. = churchyard; P.B. = pillbox; S.P. = sandpit; T.P. = trigonometric point. GENE FREQUENCY CHANGES IN СЕРАЕА 319 tions have been taken. Sites bearing the prefix D are those sampled by Boycott and Diver; sites prefixed B have been sampled only by us. Boycott and Diver described their sam- pling localities in detail and marked them on a six-inch Ordinance Survey map of the area. We have resurveyed the dunes with a prismatic compass, fixing the position of our localities with reference to permanent topographic features. Despite some inac- curacies in the Ordinance Survey map, we believe that we have been able to locate Boy- cott and Diver’s sites with a maximum error of about 15 m. During the summer of 1975 we attempted to recollect as many of the localities as possible. Of the 22 samples that we obtained, 14 represent localities originally collected by Boycott and Diver and the remaining 8 repeat our samples of 1959-1960. In addition we are reporting here on two intermediate sets of samples, 9 taken in 1963 and 16 in 1969. Scoring The snails were scored for age (mature or immature), condition, color, and band- ing according to the criteria given by Cain & Sheppard (1950, 1954) and Clarke (1960). The scoring is quite straightforward with the exception of one color class. There is at Berrow an unusual morph that we have designated “'pale brown.” It is a pale yellowish-brown, dusky, or tawny in color. In unbanded (00000) or middle- banded (00300) shells, this phenotype can usually be identified without difficulity. In fully banded (12345) shells, however, the color is easily confused with pink. Since the error involved in separating these two classes may be as large as 6%, we have not attempted to analyze pinks and рае browns separately. RESULTS Table 1 lists the phenotypes of the 5,615 live adult snails collected at Berrow in 1963, 1969, and 1975. The collecting localities are arranged in order from north to south, with successive samples from the same locality listed in chronological order. These data can be compared directly with Table 1 of Clarke & Murray (1962a). The column labeled 00345* includes not only 00345 but also 00045 and 00005. Dark browns are only scored as banded or unbanded because of the effect of this allele in partially suppressing the expression of banding. The numbers of shells with fused bands is given only with respect to the class of shells bearing all five bands. Unusual banding types are listed in a separate column. The principal findings of our former papers (Clarke & Murray, 1962a, 1962b) were (1) that there was a regular pattern in the distribution of alleles, recognizably similar in 1926 and 1960, and (2) that in frequencies of two phenotypes there had been major changes that were similar in magnitude and direction over the whole area. The consistency of these changes allowed us to calculate the relative selective values of the respective genotypes. The new data confirm the persistence of the general patterns of distribution over a fifty-year period. The centers of high frequency of all major phenotypes have remained in the same places, and over the years the morph-frequency clines have not shown any tendency to flatten out. These phenomena are apparent in Figs. 2-5, show- ing the distribution of the principal morphs in 1926, 1959-1960, and 1975. Fig. 2 shows the percentages of yellow shells. In all three sets of samples the frequency of yellow is highest in the central region, where individual samples range from 90 to 100%. Frequencies de- crease clinally in both directions. In the north the main alternative color morph is pink, while to the south there is also a significant proportion of dark brown shells. Fig. 3 shows the percentages of dark brown shells. All three sets of samples show a center of high frequency in the southern region with a clinal decrease northward. Over the northern half of the area, dark browns occur at such low frequency that chance seems to govern their presence in any particular sample. The decline in the frequency of brown, noted in the 1959-1960 samples (Clarke & Mur- ray, 1962a, 1962b) appears to have con- tinued during the period until 1975 (see below). Fig.4 shows the frequencies of the single-banded (00300) morph, expressed as a percentage of the banded yellow, pink, and pale brown shells. Unbanded and dark brown shells were excluded from this calculation because of their epistatic inter- actions with the gene producing 00300. In MURRAY AND CLARKE 320 SL6L 6961 5961 9461 - = =) ` sw === ` 6961 5961 9161 6961 5961 9461 9461 = - . - ee 33SS . ssuissuis = SL6L SL6L 6961 9461 . S 5335 - SL6L 6961 9161 6961 9461 9/61 SR ыы о особо DIODOS „sie = = 6961 s}UdWIWOD Oo AAA AHHH asada asoma SVOCLA-¢ GVETOA-L SVOZLA-L SVETOA-L GVOZLA-L SVOCLA-L GVOCLA-L SVXECLA-L SVOCLA-L SvOCld-L SVOCLA-L GVOCLA-L SPOZLA-L papueq ouıgje ¿ 000008.4-1 чес sadA Buıpu jensnu GVETL ul suoisn4 о = papuequn papueg зимо4а 44eq 97501 « 97500 | N suMoiq э|е4 00000 N= 97501 — оО т SAUId L 34 LLL S D aul as G 3 L6L $ S 62 = Œ E LoL v L 9L = 95а u er LVL С С GC = . E 881 9 L LL = zu, L6 G =< E F. ££a v =, 801 Е = “Ot = Га, = 87 3: = BS = = =a Lv L = GGL E vea zz coz (6 = (4 x cea L = LS С = HEC = 979 С = OC = L $ == 05а 8 = 86 — L L = L = 85 С = v = 6ca @ = LS С v С = Sca € = v9 L = PC 7; Res GC | = € = 9za ie = Lv = v LL = iG = Sl 5 ER С = ТУЗ 8 € Ge = E L | eva С = 66 L OL LE = € = SG 7: G el = 678 о = ae о © [== 88 8 $58 5 E (Syn (Ss) al oa gl о fo) = + < SMO||9A '97500 Se эм se $0000 pue 67000 10 siaquinu ay} apnjau! SUWN|09 „$00 SUL ‘6161 PUE ‘69GL ‘5961 Ul моллэя зе рэ}эа!10э 5//елошаи eseday yo sajdwes jo UOl}IsOdwos ay] ‘| JIGVL 321 GENE FREQUENCY CHANGES IN СЕРАЕА SL6L 6961 5961 9461 9161 9461 6961 5961 9161 6961 5961 9161 6961 5961 9461 6961 5961 6961 9461 6961 $961 9461 6961 9161 6961 = == == er Ss3==3=33 = = Шея == = == == ==> > . ss = - ui . ui — 32233 - - MODE 00 Se See - - = ` ian ee Dion are ` а Se - ОООО=О VOSOOSOUSOV OS о = ом mmmmmm-m о m aos om LE SHEZOA-L 00500 ouıgje A-L GVOCLA-L EA) oulqie A-L SvOCld-L GVOCLA-L GVOCLA-L 00500 oulqie A-L SVETOA-L Srezl oulqie д-1 SVOCLA-L GVOCLA-C GVOCLA-L SVECOA-L SvEZL oulge A-L LELL -MOKNOR ST | ELSE 851 01 SL сою = Ne 0 CO (O1 | 28 mm о N ¡FONO RA 8vL 9 SL LL 07 S ÿL 67 87 GS 1230 | 990 659 929 vsa са 05а 8va Эта Ova 138 054 322 MURRAY AND CLARKE 1959-60 1975 60 O, 640 © 730 9 97 oe 930 920 = pe 760 80 320° O 2 830 é 90 69@®@ +0 % YELLOW FIG. 2. Cepaea nemoralis at Berrow, Somerset, showing the percentages of yellow shells at various localities in successive samples from 1926 to 1975. The proportions are recorded to the nearest per cent, except when they are less than one per cent. all three sets, the frequency of this morph is higher in the south than in the north, with a local maximum in the center. Overall the frequency of 00300 has in- creased over time (see below). Fig. 5 shows the distribution of un- banded (00000) shells, expressed as a percentage of yellows, pinks, and pale browns. Dark browns have been excluded from the calculations to remove the effects of epistatic interactions of this allele with the locus controlling banding. Unbandeds, other than dark browns, are never com- mon. Nevertheless their occurrence seems to follow a pattern since there are two general areas where they appear in all three sets of samples. One of these is in the north, and the other is in the south. DISCUSSION These results appear to confirm and extend the conclusions of our previous papers. The populations of Cepaea nemo- ralis on the Berrow dunes have persisted without major disruptions for a period of at least half a century. The persisting patterns of morph frequency demonstrate that there has been no general extinction and recolonization. Therefore the changes that have been observed are the result of gradual evolution of the populations in situ. Moreover, the extent of the inhabited area, the sizes of the populations, and the consistency of the changes suggest that systematic rather than random forces are responsible. In order to investigate this GENE FREQUENCY CHANGES IN СЕРАЕА 323 1926 1959-60 1975 % BROWN FIG. 3. Cepaea nemoralis at Berrow, Somerset, showing the percentages of dark brown shells at various localities. consistency we have analyzed the frequen- cies of dark brown and 00300 among the localities and over the years. The analysis of dark brown is difficult because the decline of this phenotype with time has produced rather low frequencies in many of the recent samples. When they are compared with the 1959-60 collections, the eight samples taken in 1975 from the area of highest frequency (D40-D56) show five cases where the proportion of dark brown has decreased, two where it has remained approximately constant, and one where it has dramatically increased. For six of these localities there are scores for all five years of sampling, and these have been further studied by means of the G statistic and by the analysis of covariance. Considering first the G analysis of the years since 1959, the results set out in Table 2 show that although there is a strong association between sampling locality and color, the association of color with year of sampling is not significant. (The third association, locality and year, simply re- flects variations in sampling conditions, such as weather, that affect sample size differentially.) The final component, the interaction of the three variables, is also significant, reflecting the influence of sam- ple D56, where there has been an increase. When the analysis is extended to include the samples from 1926 (Table 3), this influence is submerged in the general trend toward a reduction in the frequency of the dark brown morph. The interaction is no longer significant, while the two compari- sons of interest (locality X color and year X color) are highly significant. The full impact of these statistics can be 324 MURRAY AND CLARKE 1926 1959-60 1975 se e? 268 17 08930 140 110 nd % 00300 FIG.4. Cepaea nemoralis at Berrow, Somerset, showing the percentages of single-banded (00300) shells at various localities. TABLE 2. G analysis of the distribution of frequencies of the dark brown morph by locality (D40, D48, D50, D52, D54, D56) and by year for the period 1959 through 1975 (1959-60, 1963, 1969, 1975). G Degrees of freedom Probability Gtotal = 779.17 38 << 0.001 Slocality X year = 659.81 15 << 0.001 locality X color = 83.81 5 << 0.001 color X year - 4.59 3 0:3>>P=0:2 Ginteraction = 30.96 15 < 0.01 TABLE 3. G analysis of the distribution of frequencies of the dark brown morph, as in Table 2 but with the addition of scores for 1926. G Degrees of freedom Probability Stotal = 1400.91 49 << 0.001 Glocality X year = 1155.31 20 << 0.001 Glocality X color = 155.33 5 << 0.001 Scolor X year = 70.70 4 << 0.001 interaction 19.56 20 05>P>0.3 GENE FREQUENCY CHANGES IN СЕРАЕА 1959-60 % 00000 FIG. 5. Cepaea nemoralis at Berrow, Somerset, showing the percentages of unbanded (00000) shells at various localities. TABLE 4. Analysis of covariance of the frequencies of the dark brown morph with time, for the period 1926 through 1975. SS dev. from reg. Deg. of freedom Mean square Е Р Within 208.97 18 11.61 Differences among regression coefficients 41.79 5 8.36 0.72 n.s. Common 250.76 23 10.90 Differences among adjusted means 790.84 5 158.17 14.51 < 0.001 Total 1041.60 28 seen from the analysis of covariance set out in Table 4. For these six localities over the whole period, the means of the frequencies of brown at the various localities are highly significantly different from one another, but the regressions of the frequencies of brown on elapsed time are not different from locality to locality. Hence we can with some justification consider the changes in the frequency of dark brown to be homogeneous and the populations to be subject to the same regime of selection. Using Bulmer’s method, outlined in Clarke & Murray (1962b), we can calculate the mean value of the selective coefficient associated with the gene for dark brown (C8) over the period 1926 to 1975 as 3.5% with a standard error of 1.1%. This figure 326 MURRAY AND CLARKE 25 20 % BROWN 1926 DATE OF SAMPLING 1960 1963 1969 1975 FIG. 6. Changes in the frequencies of dark brown shells in populations of Сераеа nemoralis at various localities at Berrow, Somerset. is not significantly different from that obtained in 1959-60. The changes are shown graphically in Fig. 6. In calculating coefficients of selection we have used a generation time of three years for Cepaea nemoralis (Clarke & Murray, 1962b). This estimate is based on the assumptions that an individual reaches maturity during its second autumn or third spring, when it is eighteen months or two years old, and that the adult population suffers a mortality rate of about 50% per year. Cain & Currey (1968) have suggested, from the size of laboratory bred snails and from the size distribution of young snails in nature, that juveniles do not reach maturity until they are at least two years old and that reproduction may not begin until the third year. Thus generation time would be increased to at least four years. However, they have assigned the size classes in wild-caught samples to year classes on the basis of laboratory bred individuals that grew to a maximum size of 7mm in a season. Comparable broods in our labora- tory are more successful. At 20 weeks a typical brood attained a mean size of 10mm with the largest animal reaching 15mm (N= 74; x= 10.02 mm; sx = 0.23). Thus it seems likely that the size classes reported by Cain & Currey are in fact one year younger than they have suggested, with animals maturing in their second year. Also, in our experience, individuals in good condition are capable of fertile mat- ing as soon as the lip is formed. The delays reported by Wolda (1963) probably reflect the artificially high densities on his experi- mental plots. Hence, we believe that three years is a reasonable estimate of the generation time. Indeed, it results in con- servative estimates of selective coefficients since the size of the coefficient of selection is directly proportional to the generation time (Clarke & Murray, 1962b). With respect to the middle-banded morph (00300), sample sizes are less of a problem, because most frequencies are in an intermediate range and increasing. All 22 samples taken in 1975 contain enough 00300 individuals for comparison with the 1959-60 series. There are twelve increases in frequency, three samples with essentially no changes and seven decreases. Since complete scores for all years are available for only nine localities, the 1963 and 1969 samples were excluded from the G analysis of the 00300 morph. Table 5 summarizes the comparison of 1975 with 1959-60. Both the main effects of interest are highly significant, but there is also a highly significant interaction component. It is clear, therefore, that although there are changes in the frequency of 00300 from place to place and from time to time, the changes are not consistent with one an- other. Extending the analysis to include the GENE FREQUENCY CHANGES IN СЕРАЕА 327 TABLE 5. G analysis of the distribution of frequencies of the 00300 morph by locality (all those sampled in 1975) and by year (1959-60 and 1975). G Gtotal = 1641.31 Glocality X year = 1335.21 Glocality X 00300 = 243.40 600300 X year = ogo Ginteraction = 54.09 Degrees of freedom Probability 64 << 0.001 21 << 0.001 21 << 0.001 1 < 0.01 21 < 0.001 TABLE 6. G analysis of the distribution of the frequencies of the 00300 morph, as in Table 5 but with the addition of scores for 1926. Fourteen samples are included for each of the years 1926, 1959-60, and 1975. G Gtotal = 2413.34 Glocality X year = 1992.03 Glocality X 00300 = 290.03 600300 X year — 90:02 Ginteraction = 41.26 Degrees of freedom Probability 67 << 0.001 26 << 0.001 13 << 0.001 2 << 0.001 26 < 0.05 1926 samples, there are fourteen localities with scores available from 1926, 1959-60, and 1975. The G analysis (Table 6) shows the same patterns as before, although in this instance the interaction component is only marginally significant. These results suggest that a formerly consistent trend toward the increase of 00300 по longer obtains, and that in recent years the frequencies of 00300 have been changing at different rates in different localities. The analysis of covariance can be per- formed for nine localities for which data are available in all five sets of samples. Unfortunately the result is not particularly enlightening. There is greater variability among the individual regression coefficients from 1959 to 1975 than from 1926 to 1975, but for statistical reasons it is only over the longer time span that one can detect significant differences. We can only say with confidence, therefore, that for some portion of the time the localities differ in the direction and magnitude of changes in the frequency of 00300. Under these circumstances it is doubtful whether one should speak of a mean value of the selective coefficient associated with the gene for middle-banded. The “mean” in this case is an average of disparate values rather than a statistical mean determined by a number of estimates of the same parameter. Nevertheless, we have calculated the average value of the selective coeffi- cient against the gene for unmodified banding (M~) as 4.0% with a standard error of 0.6%. The changes in the frequencies of 00300 are shown in Fig. 7. Finally, what can be said about the relation of these gene frequencies to the local ecology? In our previous paper (Clarke & Murray, 1962a) we noted two principal changes that had occurred at Berrow since the visit of Boycott and Diver. The first of these was the spread of sea buckthorn (Hippophae rhamnoides), a woody shrub that has become well estab- lished on the main ridge of the dunes. The primary effect of this plant is to reduce the area habitable by Cepaea, since the snails do not live in dense thickets. A probable secondary effect is to increase selective predation on Cepaea. The shrubs provide cover for the most important avian preda- tor, the song thrush (Turdus ericetorum). The “thrush anvils”” used for breaking snail shells are characteristically associated with shrubby cover. The second change apparent from our earlier work was а general stabilization of the dunes, especially the sea ridge, with marram grass (Ammophila arenaria) giving way to short, rabbit-grazed turf. According to the work of Cain & Sheppard (1950, 1954), both of these changes might be expected to favor yellow shells over browns and effectively unbanded shells (such as 00300) over more heavily banded morphs. From our 1975 field notes, three trends are apparent over the last fifteen years. First, the spread of sea buckthorn has continued. One of our sampling areas has 328 MURRAY AND CLARKE 30 25 20 % 00300 a 1926 1960 1963 1969 1975 DATE OF SAMPLING B 35 30 25 20 % 00300 1926 1960 1963 1969 1975 DATE OF SAMPLING FIG. 7a and b. Changes in the frequencies of single-banded (00300) shells in populations of Cepaea nemoralis at various localities at Berrow, Somerset. GENE FREQUENCY CHANGES IN CEPAEA 329 been completely lost to it, and several others show some degree of invasion. Second, there has been a partial reversal of the trend toward stabilization of the dunes. Foot traffic has increased dramatically with a resulting deterioration of the ground cover. Some of the dunes are again in motion, and two localities have been engulfed by sand. Third, the short turf is diminishing. Rabbits seem to be far less common, the grasses are longer and coarser, and there are more herbaceous weeds such as umbellifers, Ranuncu/us, and Rubus. Under these conditions it is still to be expected that yellow shells would be favored over pinks and browns, but the more disruptively patterned shells should gain in fitness over effectively unbanded ones. The shifting trends in the frequency of 00300 may reflect this changing regime of selection. If so, and if the current ecological changes continue, then we may expect to see in the future a continuing decline in the frequency of dark browns but a reversal of past increases in 00300, with rising frequencies of the more fully banded types. A rather different interpretation of the distribution of morph frequencies on sand dunes has been advanced by Cain (1968) based on an extensive survey of dune populations of Cepaea nemoralis around the British Isles. While generally recognizing the importance of visual selection by predators in these populations, he has also noted an association between the structural complexity of dune systems and the occur- rence of the dark brown morph, and has suggested that selection in favor of browns results from the ponding of cold air in the hollows of the dunes. In addition he has pointed out that the “effectively un- banded” 00300 morph is characteristic of warm, southwest trending dune slopes (Cain, 1968). Thus if climate were a direct selective factor, a change toward a warmer climate at Berrow could be responsible for both the decline of dark brown and the increase in 00300 (Prof. А. J. Cain, personal communication). Future observations at Berrow may indeed provide an opportunity to test the relative importance of climatic selection and visual selection in determining morph frequencies in Cepaea. Too often predic- tions based on the two hypotheses are confounded, as in the case where dark, uniform habitats tend to be associated with cold climate. At Berrow, however, the present trends in vegetation may allow us to separate these effects. A continuing negative association between the change of frequencies of dark brown and 00300 would be compatible with the hypothesis of climatic control. However if the decline in dark brown continues while the frequen- cies of 00300 stop increasing, begin de- creasing, or show local responses to changes in the character of the vegetation, then the hypothesis of control by visual selection would be supported. For this reason we intend to continue monitoring morph fre- quencies on the Berrow dunes. SUMMARY 1. Populations of the polymorphic land snail Cepaea nemoralis have been sampled at Berrow, Somerset, England over the past half century. 2. Patterns of distribution of the pheno- types have remained essentially the same over the entire period, with centers of high frequency of the principal morphs located in the same places. 3. There has been a consistent overall decrease in the frequency of the dark brown morph from 1926 to 1975 corre- sponding to a selective disadvantage of the allele for dark brown (СВ) of 3.5 + 1.1% per generation. 4. There has been an overall increase in the frequency of the middle-banded (00300) morph, but there is evidence for differences in the magnitude and direction of the changes at different localities. 5. These changes can be interpreted as responses to changing ecological conditions on the Berrow dunes. ACKNOWLEDGEMENTS Once again we should like to express our debt to Professor A. E. Boycott, F.R.S., and Captain Cyril Diver, C. B., C.B.E., for their meticulous pioneering field work and for the opportunity of using their unpublished observations. We should like to thank Mrs. Elizabeth Murray, Dr. Ann Clarke, Dr. and Mrs. Aldo Herrera de Araucho, and Mr. J. J. Murray, III for their help with the collections. The 1975 field work was carried out while one of us (J. M.) held a Sesquicentennial Fellowship 330 MURRAY AND CLARKE from the University of Virginia and was a guest of the Departments of Zoology and Genetics of Liverpool University. We should like to express our appreciation of the hospitality of Profesor A. J. Cain and the late Professor Р. М. Sheppard, F.R.S.; and we are grateful to the Science Research Council for financial support. We dedicate this paper to the memory of Professor Sheppard. LITERATURE CITED ANDERSON, W., DOBZHANSKY, Th., PAV- LOVSKY, O., POWELL, J. & YARDLEY, D., 1975, Genetics of natural populations. XLII. Three decades of genetic change in Drosophila pseudoobscura. Evolution, 29: 24-36. BOLEY, G. M., 1944, The vegetation at Berrow, North Somerset. 2. The sand dune succession. Proceedings of the Bristol Naturalists’ Society, 9: 510-520. CAIN, A. J., 1968, Studies on Cepaea. V. Sand-dune populations of Cepaea nemoralis (L.). Philosophical Transactions of the Royal Society of London, ser. B, 253: 499-517. CAIN, A. J., 1971, Colour and banding morphs in subfossil samples of the snail Cepaea. In CREED, R., Ed., Ecological Genetics and Evolution. Blackwell, Oxford, p. 65-92. CAIN, A. J. & CURREY, J. D., 1968, Studies on Cepaea. Ill. Ecogenetics of a population of Cepaea nemoralis (L.) subject to strong area effects. Philosophical Transactions of the Royal Society of London, ser. В, 253: 447-482. САМ FAT ac" ЗНЕРРАВО) “PMs 1950 Selection in the polymorphic land snail Cepaea nemoralis. Heredity, 4: 275-294. CAIN, A. J. & SHEPPARD, P. M., 1954, Natural selection in Cepaea. Genetics, 39: 89-116. CLARKE, B., 1960, Divergent effects of natural selection on two closely-related polymorphic snails. Heredity, 14: 423-443. CLARKE, B., DIVER, С. & MURRAY, J., 1968, Studies on Cepaea. VI. The spatial and temporal distribution of phenotypes in a colony of Cepaea nemoralis (L.). Philosophical Transactions of the Royal Society of London, ser. B, 253: 519-548. CLARKE, B. & MURRAY, J., 1962a, Changes of gene-frequency in Cepaea nemoralis (L.). He- redity, 17: 445-465. CLARKE, В. & MURRAY, J., 1962b, Changes of gene-frequency in Cepaea nemoralis (L.); the estimation of selective values. Heredity, 17: 467-476. CURREY, J. D. & CAIN, A. J., 1968, Studies on Cepaea. IV. Climate and selection of banding morphs in Cepaea from the climatic optimum to the present day. Philosophical Transactions of the Royal Society of London, ser. B, 253: 483-498. DIVER, C., 1929, Fossil records of Mendelian mutants. Nature, 124: 183. DOBZHANSKY, Th. 1958, Genetics of natural populations. XXVII. The genetic changes in populations of Drosophila pseudoobscura in the American Southwest. Evolution, 12: 385-401. DOBZHANSKY, Th., ANDERSON, W. W., PAV- LOVSKY, O., SPASSKY, В. & WILLS, С. J., 1964, Genetics of natural populations. XXXV. А progress report on genetic changes т populations of Drosophila pseudoobscura in the American Southwest. Evolution, 18: 164-176. DOBZHANSKY, Th., ANDERSON, W. W. & PAVLOVSKY, O., 1966, Genetics of natural populations. XXXVIII. Continuity and change in populations of Drosophila pseudoobscura in western United States. Evolution, 20: 418-427. FISHER, R. A. & FORD, E. B., 1947, The spread of a gene in natural conditions in a colony of the moth Panaxia dominula L. Heredity, 1: 143-174. FORD, E. B., 1971, Ecological Genetics. Ed. 3. Chapman & Hall, London. GOODHART, C. B., 1956, Genetic stability in populations of the polymorphic snail, Cepaea nemoralis (L.). Proceedings of the Linnean Society of London, 167: 50-67. GOODHART, C. B. 1958. Genetic stability in the snail Cepaea nemoralis (L.): a further ex- ample. Proceedings of the Linnean Society of London, 169: 163-167. HEURN, W. C. VAN, 1943, Stabiliteit van populaties van Cepaea nemoralis (L.). Basteria, 8: 59-63. HEURN, W. C. VAN, 1945, Stabiliteit van populaties van Cepaea nemoralis (L.), vervolg. Basteria, 9: 39-43. KETTLEWELL, Н. В. D., 1965, А twelve-year survey of the frequencies of Biston betularia (L.) (Lep.) and its melanic forms in Great Britain. Entomologist’s Record, 77: 195-218. KOMAI, T., 1956, Genetics of ladybeetles. Ad- vances in Genetics, 8: 155-188. KOMAI, T. & CHINO, M., 1969, Observations on geographic and temporal variations in the ladybeetle Harmonia. |. Elytral patterns. Pro- ceedings of the Japan Academy, 45: 284-288. LEES, О. R., 1970, The medionigra polymor- phism of Panaxia dominula in 1969. Heredity, 25, 470-475. SHEPPARD, P. M. & COOK, L. M., 1962, The manifold effects of the medionigra gene of the moth Panaxia dominula and the maintenance of a polymorphism. Heredity, 17: 415-426. WOLDA, H., 1963, Natural populations of the polymorphic landsnail Cepaea nemoralis (L.). Archives Néerlandaises de Zoologie, 15: 381-471. MALACOLOGIA, 1978, 17(2): 331-339 THE NATURE AND DISTRIBUTION OF FOOD-INDUCED ESTERASES IN HELICID SNAILS G. S. Oxford Department of Biology, University of York, Heslington, York YO1 500, United Kingdom ABSTRACT The variation, as revealed by gel electrophoresis, in primary and food-induced (secondary) esterases coded for by locus Es. 7 in the snail Cepaea nemoralis is described. Secondary modifications arise as a result of the ingestion of nettle (Urtica dioica) and other naturally occurring foods and can be eliminated by feeding snails on a diet of carrot and lettuce. The types of variation encountered demand at least three variable sites on the basic esterase molecule. Attempts at mimicking secondary esterases in vitro with chemical modifying agents have not been successful, although iodoacetamide appears to modify a site that distinguishes Primary from secondary zones. Of five other helicid snails investigated, only Cepaea hortensis showed food-induced modifications to primary esterase zones. Longer-term studies on the other four species are needed. Consideration of the time-course of secondary esterase induction and loss has led to the hypothesis that a bacterial component of the gut flora is selected by ‘inducing’ foods. This bacterial component, it is postulated, produces an enzyme which modifies the basic esterase molecule. INTRODUCTION Electrophoresis has been increasingly used over the last two decades to measure genetic variability within and between pop- ulations of a species and between different species. The great attraction of the method is the relative ease with which genetic information can be extracted from a gel phenotype. However, in most electro- phoretic surveys of natural populations little heed is given to the possibility that some enzymes may undergo post-transcrip- tional modifications that can lead to erro- neous genetic interpretations of gels. This may be especially misleading when the modifications do not always occur in all individuals. For example, if two electro- phoretically distinct isozymes are occasion- ally produced from a single primary pro- duct, naive interpretation of the zymogram may lead to (a) an overestimate of the number of loci revealed and/or (b) an overestimate of the number of alleles at a locus and/or (c) a false estimate of the heterozygosity at the locus. Confusion is compounded when the modified zones mimic zones that are primary products of other alleles at the same locus or even alleles at completely different loci. Fisher & Harris (1972), for instance, describe the production of secondary iso- zymes at a phosphoglucomutase locus (PGM3) in man. Some of the secondary isozymes can mimic the primary products of other РСМ. alleles; if these had not been recognized, three alleles would have been postulated (instead of two) and, of course, allele frequencies and heterozygosity would have been incorrectly estimated. The pro- duction of secondary isozymes of PGM3 in vivo appears to be a function of protein age. A similar situation has been reported by Oxford (1975) at an esterase locus in the snail Cepaea nemoralis (L.). In this case, not only is it often impossible to score allele frequencies in snails assayed straight from the wild, but the multiplicity of esterase zones (both primary and second- ary) which can exist in such animals led initially to a gross overestimate of the number of loci involved (Oxford, 1973a). From the point of view of surveys designed to assess the levels of poly- morphism in organisms, it is obviously important to determine the conditions under which secondary isozymes are pro- duced and to take steps to ensure that they do not occur under the electrophoretic procedures employed. Secondary isozymes are also of considerable interest in their (331) 332 OXFORD own right. For example, what is the biochemical nature of these secondary isozymes, especially in relation to the biochemical differences in the primary products of other alleles which they might mimic? Also, what is the distribution of secondary isozymes in related organisms and what selective advantage (if any) could be associated with the ability to generate such modified enzymes? In this paper, | will discuss these questions in relation to the secondary esterases found in Cepaea nemoralis. THE PATTERNS AND INDUCTION OF SECONDARY ISOZYMES IN ESTERASES OF CEPAEA NEMORALIS Digestive gland esterases of C. nemoralis that exhibit secondary modifications mi- grate to the cathodal end of polyacryla- mide disc gels during electrophoresis. The general methods of electrophoresis and the subsequent staining of esterases have been described elsewhere (Oxford, 1973a). The locus involved is Es. 7, at which six alleles have been identified (Oxford, 1973a, 1975, and unpublished results). These ‘alleles’ are electrophoretically defined and may con- tain more than one true allele, i.e., coding for different amino acid sequences, within a mobility class. The alleles code for primary products that fall into two distinct series on electrophoresis. The products of alleles Es. 77, Es. 12, and Es. 13 form one series, and alleles Es. 722, Es. 132, and Es. 742 form the other (Fig. 1). Within a series the allelic products are equidistant on a gel, and the distance between adjacent zones is the same in both series. The series of Est. 2a, Est. 3a, and Est. 4a (coded for by alleles Es. 122, Es. 13а and Es. 14a, respectively) is shifted slightly anodally compared to the zones in the other series (Fig. 1). The post-transcriptional (and almost cer- tainly post-translational) modifications to these primary zones involve anodal shifts in units equivalent to the distance between primary zones within a series. The major modification involves just one shift of position so that, for example, a modified Est. 1 has exactly the same electrophoretic mobility as the primary product Est. 2 (Fig. 1). As each primary zone can shift down one unit, it follows that Est. 1 and Est. 2a cannot be mimicked, whereas Est. 4 and Est. 5 can only be the products of modification. Esterase zones at the posi- tions of est. 2, Est. 3, Est. За, and Es 43 can be either primary products, secondary products, or a mixture of the two. The modification may go to completion so that there is no trace of the primary product, or it may be partial, yielding two zones in homozygotes. Further modifications are also possible. For example, electrophoresis of an extract from a snail homozygous for Est. 1 might reveal not only Est. 1 and Est. 2 but also very faint traces of Est. 3 and Est. 4. This suggests additional modifications, which again produce unit shifts in electrophoretic mobility. Obviously it is impossible to score allele frequencies at this locus while there is the possibility of confusing primary and secondary zones. It was the presence of four zones in some individuals, e. g., a snail of the genotype Fs. 77/Es. 13 with both primary and secondary esterases, which led to the idea that five closely linked loci were involved (Oxford, 1973a). Each locus was thought to control enzyme activity at one level in the gel and, at the time, five zones were recognized: Est. 1, Est. 2, Est. 3; Esti4: ama Estib: The conditions that induce secondary modifications have been identified (Oxford, 1975). Ingestion of nettle (Urtica dioica L.) produces secondary zones within a snail, whereas ingestion of carrot and lettuce suppresses them. Within an individual these processes are completely reversible. Collec- tions of snails from habitats that do not contain nettle indicate that other foods can induce secondary zones as well. In all samples taken so far, at least some indi- viduals have been shown to possess modi- fied esterases. If a survey of allele frequen- cies at locus Es. 7 is required, it is necessary to maintain snails on a carrot and lettuce diet for at least five to six months before extraction and electrophoresis (see below). PRIMARY AND SECONDARY VARIATION AND THE ESTERASE MOLECULE A consideration of the variation, both primary and secondary, in electrophoretic mobility of the esterase components de- scribed above demands at least three and FOOD-INDUCED ESTERASES IN HELICIDS 333 Secondary zones Primary zones + FIG. 1. The primary and major secondary esterase zones in C. nemoralis, showing the relationship between the two allelic series, the relative mobilities of allelic zones within a series, and the phenocopying of primary zones Бу some of the secondary zones. ид = upper gel. Migration is from cathode to anode. probably more variable sites in the basic esterase molecule. 1. The first and second variable sites on the esterase molecule are those which determine the mobility of an esterase zone within an allelic series. Presumably the difference in mobility between adjacent zones depends on the substitution of one amino acid for another such as to give the molecule an extra negative or positive charge, depending on whether one goes from Est. 1 to Est. 3 or in the reverse direction. The generation of three zones within each series demands a similar change at yet another site on the molecule. It is assumed that the differences in electro- phoretic mobility between zones within a series are due to unit charge changes (Oxford, 1975). 2. The shift observed between the two mobility series of allelic zones may be produced by the substitution at a third site of а negatively charged amino acid (under the electrophoretic conditions employed) for a neutral one (e.g., to convert Est. 1 to Est. 2a) or a positively charged amino acid for a neutral one (e.g., to convert Est. 2a to Est. 1). Henning & Yanofsky (1963) have shown that the charge expressed by an amino acid may vary according to the 334 OXFORD nature of adjacent amino acids in the polypeptide. It is therefore possible to observe different electrophoretic mobilities with changes involving the same amino acid at different positions in the protein. Also, the substitution of one amino acid for another may slightly change the conforma- tion of the molecule, thus exposing differ- ent charged groups. Mechanisms such as these may explain the non-unit mobility shift observed between the two series of allelic zones. Hypotheses based on a carbo- hydrate moiety (Karn et al., 1973), or more specifically sialic acid (Law, 1967), added or not added to a basic esterase molecule by another locus are doubtful be- cause an individual can be heterozygous for alleles in different series. The major secondary modification of all the primary esterases probably occurs at yet another site. How secondary isozymes are generated from primary ones is not clear. | have suggested before (Oxford, 1975) that they may be a result of deamination of a vulnerable glutamine or asparagine residue (charge change, O to -ive) or the blocking of an amino group by acetylation (charge change, tive to 0). Other mechanisms are also possible, e.g., the addition of small, charged side groups to the primary products. Sialic acid is not involved since incubation of primary and secondary esterases with neuraminidase does not alter electrophoretic patterns (Oxford, 1973a). However the increase in net negative charge is effected, it must exactly match the increase in negative charge displayed by allelic zones within a series. It was mentioned earlier that, besides the major modification that causes a unit shift toward the anode, some molecules undergo additional shifts of two and three units. These further modifications may occur at the same site as the major modification or at different sites on the molecule. Schwartz et al. (1965) have described a very similar esterase system in maize, consisting of seven alleles in two series. As in Cepaea, the charge difference between members of a series is constant but the two series, of three and four alleles, are shifted in mobility with respect to one another. Schwartz (1967) argued on the basis of experiments involving chemical modifica- tion of the esterases that electrophoretic differences between the isozymes within a series are unlikely to be due to differences in the net charge of amino acids making up the enzyme. He also suggested that differ- ent numbers of charged side groups con- jugated to the protein would not explain his findings. The hypothesis finally sugges- ted was that the allelic isozymes within a series had the same primary structure but differed in conformation and that the differences in electrophoretic mobility re- sulted from differential masking of charged groups. If this interpretation is correct, it is difficult to see how ‘allelic’ zones with different mobilities can be found within a single plant. In other words, how can different conformational forms Бе stabi- lized sufficiently to appear allelic in breed- ing experiments? In Cepaea there is, of course, the added complication of the food-induced modifications of primary es- terase zones. EXPERIMENTAL MODIFICATION OF ESTERASE ZONES Theoretically, it should be possible to identify the group or groups responsible for the differences between primary and secondary products by treating the ester- ases with chemicals known to modify particular molecular features. A chemical which converts primary zones into second- ary zones but which does not alter the mobility of existing secondary zones may be presumed to affect the molecular fea- ture responsible for the primary-to- secondary modification. Unfortunately, this approach has had only limited success so far. The only chemical agents found that result in a shift in the mobility of zones are formaldehyde, glyceraldehyde, and iodoacetamide. Treatment of extracts with 0.2M for- maldehyde for 3 hr at 37°C prior to electrophoresis results in a shift of both primary and secondary esterases of exactly two units toward the anode. There is a progressive increase, in unit steps, as the concentration is increased from 0.05M to 0.2M. After 0.2M no further change is produced. Identical results were achieved with glyceraldehyde, but acetaldehyde had no effect (Fig. 2). Since both primary and secondary zones are equally affected, this reaction gives no clue to the biochemical differences between them. FOOD-INDUCED ESTERASES IN HELICIDS 335 FIG. 2. A, Extract plus an equal volume of distilled water incubated for 3 hr at 37°C before electrophoresis. Primary zones are black; secondary zones are dashed. B, Extract plus an equal volume of 0.4M formaldehyde or glyceraldehyde incubated under identical conditions to A before electrophoresis. Note that all zones move two positions towards the anode. C, Control; conditions of incubation as for A. D, Extract plus an equal volume of 0.2M iodoacetamide incubated under identical conditions to A before electrophoresis. Note that only the primary zones show an increase in electrophoretic mobility. ug = upper gel. On the other hand, incubation of ex- tracts with 0.1M iodoacetamide for 3 hr at 37°C results in a slight anodal shift in the mobility of primary but not secondary zones (Fig. 2). If one assumes that the shift between one primary zone and the next in a series is equivalent to a unit charge difference, then iodoacetamide produces a shift of about 0.35 unit charge toward the anode. An increase in net negative charge is unexpected since most of the groups that iodoacetamide reacts with e.g., sulphydryl and amino, are probably negatively charged under conditions of electrophoresis. lodo- acetamide should neutralize these groups, thus producing a net increase in positive charge. Whatever the anomalies of this situation, it is clear that the iodoacetamide is reacting with a group which distinguishes primary and secondary esterases. The in- crease in electrophoretic mobility produced by iodoacetamide was not sufficient to mimic the shift that characterizes the two series of allelic products. Clearly, further work along these lines is necessary to clarify the situation. 336 OXFORD THE DISTRIBUTION OF FOOD-INDUCED ESTERASES IN OTHER SPECIES OF SNAIL Five other species of helicid snail, Cepaea hortensis (Muller), Helix aspersa Muller, Arianta arbustorum (L.), Monacha cantiana (Montagu), and Hygromia striolata (C. Pfeiffer), have been screened for their ability to produce food-induced esterases. A comparison of esterase patterns found on gels of digestive gland and kidney extracts and of crop juice has identified esterases possibly homologous to those produced by the Es. 7 locus in С. nemoralis. These esterases are present in the digestive gland but not in the kidney, and they are actively secreted into the crop juice where they function as extracellular digestive enzymes (Oxford, 1977, and Fig. 3). In all six species studied, a group of cathodal esterases conforms to this pattern; it was within these enzymes that food-induced modifications were sought. The experimental method used for the larger snails (С. hortensis, H. aspersa and А. arbustorum) was to feed the animals on carrot and lettuce for a period, hopefully to eliminate any food-induced zones gener- ated in the wild. A sample of digestive gland was taken and electrophoresed (Ox- ford, 1975), and the animals were put onto nettle for another period before resampling the digestive gland, usually by killing the snail. Of these three species, food-induced zones have only been detected in C. hortensis. Here only one allele (Es. 77) is present at what is certainly the homologous locus to that found in C. nemoralis. The secondary modifications are identical to those described for C. nemoralis; indeed the paucity of alleles makes C. hortensis much more amenable for studies on in- duced zones. With only one allele, Est. 1 must be a primary product and Est. 2 a secondary modification of it. The weaker secondary zones Est. 3 and Est. 4 are particularly clear in C. hortensis. Similar modifications have not been found in Arianta or Helix. Both species had been maintained on carrot and lettuce for only one month before the initial digestive gland sample was taken; in retrospect, this time may not have been long enough to eliminate any food-induced zones still re- tained from the wild (see below). The smaller species, Monacha cantiana and Hygromia striolata, were investigated for food-induced esterases by dividing a sample taken from the wild into two, half being fed on carrots and lettuce and half on nettle. Samples from both groups were extracted and electrophoresed after one month. No differences were observed be- tween the H. striolata subgroups, but in M. cantiana a new zone (Est. a, Fig. 3) was present in all the snails fed on nettle but in none of the snails fed on carrot and lettuce. This presumably represents the induction of another esterase rather than the modification of a previous zone. No other esterase zone was affected. A similar phenomenon has been observed in the esterases of the freshwater snail Potamo- pyrgus jenkinsi (Smith) by B. R. Johnson (personal communication). Compared with snails fed on dried lettuce or detritus, those fed on dead leaves develop an extra, powerful esterase zone after 14 days. With the exception of the apparent induction of a new zone in M. cantiana, the two Cepaea species seem to be the only ones of those tested which show food- induced modifications. However, longer term studies may show that modifications do occur in other species as well. It will be particularly interesting to examine the esterases of Cepaea sylvatica and С. vindo- bonensis from the point of view of modi- fiable zones. The test described above assumes that nettle can induce modifica- tions in all species (if modifiable esterases exist) and that carrot and lettuce produce no modifications. One or both of these assumptions may not be true for species other than Cepaea. AAA AAA FIG. 3. Tissue distribution of esterase zones in extracts of digestive gland (DG) and kidney (К) and in crop juice (CJ) of six species of helicid snails. Black and dashed zones preferentially hydrolyse a-naphthyl acetate (stain black/purple) and dotted zones preferentially hydrolyse f-naphthyl acetate (stain red) when mixed substrates are used. Relative intensities of esterases are indicated by the width of the bands, with faint esterases shown as dashed or dotted lines. The direction of migration is indicated by the arrows. 337 FOOD-INDUCED ESTERASES IN HELICIDS C. hortensis C. nemoralis о o A. arbustorum Н. азрегза - DE-=-=x ФЕ H. striolata M. cantiana = OU SERES NE o Ш Ш (a) vo 0 DL D £ м = ЕС © ©: 338 OXFORD THE LINK BETWEEN FOOD AND THE PRODUCTION OF SECONDARY ESTERASES IN CEPAEA Several hypotheses could be suggested to explain the link between nettle ingestion and the generation of modified esterase zones. Perhaps the most obvious explanation is that some component of ingested nettle acts directly on the esterase molecule, increasing its net negative charge. However, two reasons suggest that this is not the mechanism involved in this instance: (1) incubation of esterase extracts with aque- ous nettle extracts results in unaltered zones, and (2) the induction of secondary zones lags behind the ingestion of nettle by a week or so and continues for many weeks after nettle has been removed (see below). А second hypothesis is that nettle ingestion induces the production of a modifying substance, perhaps an enzyme, by the snail itself. The lag between nettle ingestion and the appearance of secondary esterases could thus be explained by the time needed to induce the intermediate substance. If this was the case, however, removal of nettle should stop the produc- tion of the intermediate substance and lead to the loss of secondary esterases. Assum- ing a rapid turnover rate for the primary esterase, the loss of secondary zones should take about the same time as their initial induction. A fast turnover rate for these esterases is very likely, considering the fact that they are actively secreted into the crop juice (Oxford, 1977). An examination of results from a num- ber of samples indicates that the induction of secondary esterases on a nettle diet is very much faster than their subsequent loss on a diet of carrot and lettuce (Table 1). In one experiment specifically designed to examine the rate of secondary esterase production, a batch of C. hortensis without secondary esterases was fed on nettle and samples were analyzed after various periods of time. Over half the snails had developed secondary esterases by day 13, and all possessed them by day 28. On the other hand, a sample of the same species collec- ted from a colony in which 90-100% of snails were known to exhibit secondary esterases still had 97% of individuals with secondary zones after 86 days on a carrot and lettuce diet. Other observations on TABLE 1. Percentage of snails with secondary esterases under different food regimes. No. Species Day % analyzed (a) Carrot/Lettuce — Nettle 1. С. hortensis 0 0 12 13 58 12 21 83 12 28 100 12 2. С. hortensis 0 0 15 40 71 7 3. С. nemoralis 0 0 24 86 61 18 (b) Wild > Carrot/Lettuce 4. C. hortensis 0 с90-100 — 86 97 32 5. C. nemoralis 0 24 25 19 31 13 97 0 14 6. С. nemoralis 0 100 40 59 100 2 203 0 34 both С. nemoralis and С. hortensis broadly agree with this pattern (Table 1). This marked asymmetry in the time for induction and loss of secondary esterases has led to a third hypothesis. The intestinal tract of snails, as in most animals, contains a resident population of microorganisms. The suitability of the gut environment for particular microorganisms must depend to some extent on the nature of the ingested food. Certain diets may select for some bacteria, for example, but against others. If ingested nettle imposed powerful selection for a bacterial component of the gut flora which produced and secreted an enzyme capable of modifying the Es. 7 esterases, then the temporal relationship between the ingestion of nettle and the appearance of secondary esterases would be explained. Carrot and lettuce as foods may impose mild selection against this bacterial com- ponent, reducing its importance over an extended period. This hypothesis would account for the asymmetry of response observed in the time for induction and loss of secondary esterases. As yet there is no evidence from Cepaea to support or refute It. A situation that parallels this hypothesis has been described for salivary amylase enzymes in man (Karn et al., 1973). The bacterial flora of the mouth produce an enzyme that can remove the carbohydrate moiety from the amylase molecule after FOOD-INDUCED ESTERASES IN HELICIDS 339 secretion, thus increasing its relative elec- trophoretic mobility. To be more speculative, if this /s the mechanism responsible for food-induced esterases in Cepaea, then it raises the question of whether it is of any selective advantage for secondary enzymes to be present if nettle (and other inducing foods) are present in the gut. Although the gross biochemical and physical properties of primary and secondary esterases appear the same (Oxford, 1971, 1973b, 1973c), we have no information about their relative hydrolytic efficiencies on natural sub- strates. If modified esterases are more efficient at the hydrolysis of natural esters present in nettle, then we have an intri- guing evolutionary situation in which the phenotype of an organism is adaptively changed, not as a result of the environ- mental stimulus itself but via an environ- mental effect on an intermediate organism. LITERATURE CITED FISHER, В. A. & HARRIS, H., 1972, ‘Second- агу’ isozymes derived from the three РСМ loci. Annals of Human Genetics, 36: 69-77. HENNING, U. & YANOFSKY, C., 1963, An electrophoretic study of mutationally altered A proteins of the tryptophan synthetase of Escherichia coli. Journal of Molecular Biology, 6: 16-21. КАВМ, В. C., SHULKIN, J. D., MERRITT, А. D. & NEWELL, В. C. 1973, Evidence for post-transcriptional modification of human salivary amylase (Amy, ) isozymes. Biochemi- cal Genetics, 10: 341-350. LAW, G. R. L., 1967, Alkaline phosphatase and leucine aminopeptidase association in plasma of the chicken. Science, 156: 1106-1107. OXFORD, G. S., 1971, The properties, genetics and ecogenetics of esterases in Cepaea. Ph.D. dissertation, University of Liverpool, Liver- pool, England. OXFORD, G. S., 1973a, The genetics of Cepaea esterases |. С. nemoralis. Heredity, 30: 127-139. OXFORD, G. S., 1973b, The biochemical proper- ties of esterases in Cepaea (Mollusca: Helici- dae). Comparative Biochemistry and Physi- ology, 45B: 529-538. OXFORD, G. S., 1973c, The molecular weight relationships of esterases in Cepaea nemoralis and С. hortensis (Mollusca: Helicidae) and their genetical implications. Biochemical Gene- tics, 8: 365-382. OXFORD, G. S., 1975, Food-induced esterase phenocopies in the snail Cepaea nemoralis. Heredity, 35: 361-370. OXFORD, G. S., 1977, Multiple sources of esterase enzymes in the crop juice of Cepaea (Mollusca: Helicidae). Journal of Comparative Physiology, 122: 375-383. SCHWARTZ, D., 1967, Ey esterase isozymes of maize: on the nature of the gene-controlled variation. Proceedings of the National Aca- demy of Sciences of the United States of America, 58: 568-575. SCHWARTZ, D., FUCHSMAN, L. & McGRATH, К. H., 1965, Allelic isozymes of the pH 7.5 esterase in maize. Genetics, 52: 1265-1268. Wet tbe acer! AE ESOS soon liste 0d sa А técond ty Pe jean Ne Ñ | a x ©. #5 e ra Lx Gece p PH tw the on ai Ma) Fs or adi poo OS CRM Keen fasten. 14 this we Plinio paar ¿sano ры те moa Shea, wom e Se eae har nf Pe : WW # fapl I MIMO YA von Y TAR Primar raté as tie 0 Coralia or мед (pie IDOUT 112 soiré une ay VAT: | ty MH On, Pel у r ascH Ba $ 0 к у тина iQ ! TOT ner \ И я MO q | an у 11 | IX ore rm? vor A | i d 4 am Là 4 1 } un ik A % Be 4 tl a у Y © mm | 3 y e у В ' i te GET ie № “À it ex 11 sn Ue р eu Y | Ms | 19 1 1 } Nag wen described for: gativary (sh vievisgabs al re ont né VFO norivns art to Mim e 28% Snihiveewad: siv dud tiseti ‚mainsgyo оное ns ne ee ой ньях ЕЕ > ber С has LT Чон" 1 teen phase фо у ws a 0 Moe BL I Ten Ur $ if Yo Shige, ey | у. NO RIM ne sun Way Му of he gut euro Mt TENUE MOUTON MATO pen ADF] Viper on The) ОЕ Об, oy, Cérivin diewermnäy select ion LA , Tor evarngie, QUE AGUAS ria ester’ néttic imposed Power y otoria conworen?t OPane sé y PAS which produced’ and’ setrered “an mabte of modiftving Me Es" Inu ter у Ito В 7 wd the ae 1 fates NOUS бя ERE wet’ « emuta a5 Toods МАЯ im | ti wirt this басен ant som) Tr Mut ous ood) perio Ti пузо un $ мо etyminairy om fus “cr dr me for Wmtion i ОО ГУ corevaced. As yar them Oe TOM Lenaes TO support OF абон thal Ване УЧЕТ ро en non (here et gl. 1079 e” ter Nore of dia mou product an плим thet сай’ janie Чье. сабо poiety: fri ihe ай 8$ mojeculo artery MALACOLOGIA, 1978, 17(2): 341-350 REPRODUCTION AND ITS CONTROL IN DEROCERAS RETICULATUM N. W. Runham Zoology Department, University College of North Wales, Bangor, Gwynedd, United Kingdom ABSTRACT The basic anatomy and histology of the gonad, hermaphrodite duct, carrefour, albumen gland, common duct, vas deferens, free oviduct, penial mass and bursa copulatrix of the reproductive system of Deroceras reticulatum are reviewed and related to their physiology. Evidence for nervous and endocrine integration of reproduction hermaphrodite species is discussed. INTRODUCTION Deroceras reticulatum (Müller) is ап important horticultural and agricultural pest in many countries. In Bangor it can normally be collected throughout the year; most easily at night, although the numbers are very dependent on weather conditions. Over the last ten years, together with a number of good research students, | have studied reproduction and the factors that control it in this animal. There are still many areas, however, where our under- standing is far from complete, and it is these areas that | emphasise in this paper. THE REPRODUCTIVE SYSTEM Like other pulmonates, this slug is hermaphrodite and the arrangement of the reproductive organs is typical of the sty- lommatophorans (Fig. 1). Although D. reticulatum is classified as а simultaneous hermaphrodite, with both sperm and ova in the gonad, in fact the sperm and those parts of the reproductive system that produce the spermatophore mature first. The maturation of oocytes and those parts of the tract that form the egg is completed later. This species is therefore said to be protandric. Although the degree of separation of the male and female phases is very clear in many pulmonates, in D. reticulatum there is considerable overlap. Reproduction can occur at any time of year but in the United Kingdom is most common in spring and autumn (Runham & Laryea, 1968). in this protandric Growth of the animal after hatching is at first slow but then increases considerably in parallel with the growth and maturation of the reproductive tract. At the end of this phase the growth rate decreases, and copulation and egg laying take place, quickly followed by death of the animal. D. reticulatum only lives for 6-12 months in the United Kingdom. Maintenance of the animals in the laboratory is easy as long as they are kept damp and cleaned regularly. They will eat a wide range of food materials. Long-term culture in this laboratory has proved to be much more difficult as there are often outbreaks of infection with Tetrahymena which, as they are transmitted in the egg, are impossible to eradicate. THE GONAD The gonad or hermaphrodite gland con- sists of acini connected via small collecting ducts to the main hermaphrodite duct. At hatching one to three acini are present, and these are very small swellings at the tips of outgrowths from the duct. Further out- growths from the duct increase the number of acini to about 100. At first they are rather globular, but as they increase in size they become more variably shaped: some are rather H-shaped, others irregularly lobed (Runham & Hogg, Ms). Examination of sections of the gonad reveals that four cell types are present: the male and female gametes and their associa- ted cells, the sertoli and follicle cells, respectively. The relative proportion of (341) 342 RUNHAM NS Me CGonad Hermaphrodite duct Carrefour Albumen gland Oviducal gland Prostate gland Trifid__- appendage Retractor | | Vas deferens nee DUE Bursa copulatrix E ule ir Free Penial oviduct sac Genital opening FIG. 1. Diagram of the reproductive system of Deroceras reticulatum (adapted from Bayne, 1966). these gametes and their state of maturity varies throughout the life cycle of the animal. Although these changes are con- tinuous, it is convenient to distinguish eight stages (Runham & Laryea, 1968): A, undifferentiated: at first only a few acini are present containing primary germinal cells; then acini increase to approximately 100. spermatocyte: recognisable tocytes and oocytes differentiate. B sperma- , C, spermatid: spermatogenesis has advanced to the spermatid stage, although many earlier stages are still present. D, early sperm: sperm are now present, but a thick layer of spermatogenesis stages lines the walls. E, late sperm: the acini have reached their maximum size, many sperm are present, and there 15 a clear space at the centre; the layer of cells in spermatogenesis is thinner, and the oocytes, which have been increasing in number since stage B, CONTROL OF DEROCERAS REPRODUCTION 343 have also increased in size, a few reaching the maximum diameter of 90 um; as this stage appears to be the most common, it probably lasts the longest time. F, early oocyte: mature oocytes predomi- nate, but all are present in follicles on the acinus wall. G, late oocyte: mature oocytes are present in the lumina of the acini, which have contracted in size. H, postreproductive: a columnar epithelium is present at the base of each acinus and can extend over most of the wall; this epithelium appears to separate the other cells from the wall of the acinus; it may be comparable with similar epithelia found in species that have a resting period between breeding seasons (Galan- gau, 1964). While apparently chaotic, the contents of the acini can be shown by reconstruc- tions to have in fact a very ordered arrangement (Runham & Hogg, Ms). On the wall of the acinus around the entry of the collecting duct there is a cubical epithelium—the germinal epithelium (Fig. 2). Around the edges of this are found the smallest oocytes; the oocytes increase in size the farther away they are from the germinal epithelium. The most mature oocytes are found on the wall opposite the collecting duct. Resorbing oocytes, when present, are mixed up with the mature oocytes. At first sight these sequences indicate that gametes and their associated cells are produced continually from the germinal ring and then move around the wall of the acinus, but it is possible that this arrangement is associated FIG. 2. Transverse section of gonad at late spermatozoon stage. D, hermaphrodite ductule; G, germinal epithelium; О, oöcytes. 344 RUNHAM with the enlargement of the acinus. Quanti- tative methods for estimating the rate of production of gametes must be developed to determine which of these views is correct. Perhaps the greatest problem associated with the gonad is the mechanism by which male and female gametes and their associa- ted cells differentiate from the one cell type present in the germinal epithelium. Recent observations (Runham & Hogg, Ms.) indicate that the germinal epithelium is continuous with a layer of sertoli cells. Cells above this layer differentiate into sperm and those beneath it differentiate into oocytes or follicle cells. There thus appear to be two separate compartments within the acinus, but it is uncertain how the cells come to adopt this distribution. Castration is a relatively simple opera- tion in this animal as the gonad is superficial and is contained in a blood sinus. Complete regeneration follows castra- tion in stage A and B animals within one month of the operation. Regeneration is rare in stage C animals, and | have only seen one example from an animal at stage D. The complete absence of such regenera- tion in the later stages has not been explained (Runham, 1976 and in prepara- tion). HERMAPHRODITE DUCT At hatching the hermaphrodite duct and the remainder of the reproductive tract are indistinguishable: it consists of a simple tube lined by a simple columnar or cubical epithelium. Within a few days, however, the cells start to fill with glycogen. The development of the carrefour determines the distal limit of the duct, and the proximal end is situated in the gonad. Until stage C the simple epithelium persists, but then the cells differentiate into ciliated and nonciliated. At either end of the duct only ciliated cells are present, but over most of its length these cells form a longitudinal band that occupies approximately one-third of the wall; the remainder is lined with nonciliated cells. During stage D sperm pass into the duct, which swells to accommo- date them. Sperm cause the duct to become white in appearance and apparently coiled. Further swelling of the duct is associated with a great extension of the nonciliated cells, which become very thin; remain columnar until (Hogg, unpublished the ciliated cells swelling is extreme data). The hermaphrodite duct therefore func- tions as a seminal vesicle. Within some of the nonciliated cells phagocytosed sperm at various stages of breakdown are found, but it is not certain if only abnormal sperm are so ingested. Other nonciliated cells are apparently involved in the passage of fluid material from the duct to the intercellular spaces and hence to the connective tissue. As the sperm are immotile, they must be carried by ciliary action through the collec- ting ducts into the main duct. It seems likely that the sperm transported in this way are carried in a fluid; but as sperm in the duct are very densely packed, this fluid must be removed, probably by the noncilia- ted cells. When sperm must pass out of the duct at copulation, it seems necessary to add fluid to the dense sperm mass. Over the outer surface of the her- maphrodite duct there is a thin layer of circular muscle fibres with an external layer of longitudinal fibres. These layers are thickened at the distal end of the duct where it passes into the carrefour. THE CARREFOUR One of the first structures to differ- entiate from the primary reproductive tract, the carrefour consists of a simple cylindrical process partially surrounded by а larger, flattened diverticulum (Fig. 3). Our observations agree with those on Helix (Lind, 1973), where the central process functions as a spermathecal sac for storage of foreign sperm and the larger outer structure is the fertilisation pocket (Nicho- las, Ms.) All the epithelia are ciliated, but how the direction of the animal’s own sperm and oocytes and foreign sperm are controlled is uncertain. Before it enters the carrefour the hermaphrodite duct narrows and forms a loose coil. ALBUMEN GLAND Developing from a pair of processes from the primary reproductive tract distal to the carrefour, the albumen gland en- larges and the walls of the sacs become more and more folded until a compound acinar structure is formed (Fig. 3). The CONTROL ОЕ DEROCERAS REPRODUCTION 345 FIG. 3. Transverse section of carrefour. A, hermaphrodite duct; S, spermathecal sac. epithelium of the primary duct continues into the albumen gland to line the acini. Cells from the connective tissue under the epithelium enlarge and send processes through the epithelium to the lumen. Secretion accumulates in these cells, which swell considerably; the lining epithelium becomes very stretched and thin. The lining cells develop cilia, which presumably help to transport the secretion (Bailey, 1970, 1973). Characteristically this secretion con- tains large amounts of galactogen; when fixed, this gland is exceedingly difficult to section because of its extreme hardness. Glycoprotein is also present in the secre- tion (Bayne, 1966). Fertilised eggs from the carrefour pass through the collecting duct area of the albumen gland, where they receive an aliquot of secretion, the perivitelline fluid that fills the central part of the egg. Since this secretion appears very concentrated, extra water may also be secreted here. Each zygote becomes surrounded by a characteristic amount of this perivitelline fluid so that some metering of the albumen gland secretion appears essential, but it has not been discovered how this is achieved. albumen gland acini; F, fertilisation pocket; H, COMMON DUCT Cell multiplication along one side of the primary reproductive tract, distal to the albumen gland, forms many tubular diverti- culae, which collectively form the prostate gland. At first only a simple epithelium is present, but cells that accumulate in the connective tissue send out processes that insinuate between the epithelial cells to reach the lumen. The cells differentiate into glandular cells; as these accumulate secretion, they swell and stretch the lining epithelium greatly. At least three types of secretory cell are present in this gland (Bailey, 1970, 1973). The oviducal gland develops on the duct wall opposite the prostate gland. Epithelial cells line the wall, and again glandular cells develop in the connective tissue beneath the epithelium and push up through it (Fig. 4). Two types of glandular cell appear to be present, and their enormous expan- sion considerably dilates the gland and attenuates the lining epithelium. The de- velopment of the prostate gland is normally complete before that of the oviducal gland starts. Eventually the epithelia lining both 346 RUNHAM FIG. 4. Developing oviducal gland. G, gland cells; L, lining epithelium; N, necks of gland cells pushing between lining epithelial cells. glands become ciliated, but the glands remain in open but narrow communication along their length (Fig. 5). Analysis of extracts of the common duct reveal the presence of a complex mixture of materials (Bayne, 1967), but it is not clear which cell type forms which constituent. At copulation sperm pass from the hermaphrodite duct into the common duct where they are encapsulated in a type of spermatophore known as a jelly mass. The prostate gland forms this jelly mass, but the details are unknown. One function of the rather lengthy courtship in these ani- mals is probably to trigger the processes that form the jelly mass. Egg shell material is deposited around the perivitelline layer of the egg as it passes down the common duct. The thickness of the shell increases as the egg passes down from the albumen gland, but it is not clear how the secretions produce two apparently distinct layers of egg shell. Calcium gran- ules are present in the egg shell (Tompa, 1976), but their source is uncertain. At the junction of the albumen gland and the common duct a small gland is present distinct from the rest of the common duct. At the moment its function is unknown, but in this position it could either add water to the albumen gland secretion or form the thin layer that surrounds the perivitelline fluid. Muscular tissue is present beneath the lining epithelium of the common duct as a thin meshwork. When the glandular cells develop they push through this layer, and as they expand the muscle layer becomes obscured. Transport of the sperm and jelly mass is probably by ciliary action of the lining epithelium, particularly in the area of the male groove. After copulation some of the partner’s sperm must be transported up the CONTROL ОЕ DEROCERAS REPRODUCTION 347 FIG. 5. Transverse section of common duct. L, lining epithelium; O, oviducal gland; P, prostate gland. common duct to the spermatheca. This requires a reversal of the normal ciliary current, but observations of this process have not been published in this species. In other species it appears to ascend the oviduct and oviducal gland (Lind, 1973). Movement of the completed egg, which must cause considerable distension of the tract, presumably requires muscular action, but again the details are obscure. VAS DEFERENS AND FREE OVIDUCT The vas deferens is a small tube with a well-developed muscular layer and a simple epithelial lining. What functions this struc- ture performs, in addition to conducting gametes, are unknown, as is the means by which this narrow tube transports the jelly mass. Large flask-shaped glandular cells open through the lining epithelium of the ovi- duct, and these must presumably add material to the outside of the egg. This secretion and its function have not been studied. THE PENIAL MASS Very little information has been pub- lished on this very complex structure. Anatomically it consists of a large penial sac on the tip of which are found the three processes of the trifid appendage, a distin- guishing character of this species. When everted during courtship, the penial mass forms a globular process on the outside of the animal. Surmounting this mass is a large conical process, the sarcobellum, which is highly mobile and is used to caress the partner. This sarcobellum is attached to the inner wall of the penial mass so that it is not everted, although undoubtedly it is a very erectile structure. The trifid appendage is everted during courtship, but it is often difficult to see because it is small. When the penis sac is everted, the openings of the oviduct and the bursa copulatrix open close together at the base of the everted mass, some distance from the opening of the vas deferens. Another function of courtship must be to align the slugs to bring the vas deferens openings 348 ВУМНАМ FIG. 6. Longitudinal section of penial mass. С, gland cells; $, sarcobellum; W, wall of репа! mass. opposite the bursa copulatrix openings. Accurate alignment does not appear to be essential, however, as animals fixed immedi- ately after copulation normally have the transferred jelly mass present within the cavity of the retracted penial sac. Eversion of the penial sac is accomplished by relaxation of the genital opening and pressure of blood blowing out the sac. The everted sac is very large and contains a high percentage of the animal’s blood together with any organs of the body that push into it. The sac is retracted by the репа! retractor muscle attached to the tip of the penis sac. Histologically the penis sac is shown to be complex (Fig. 6). The sarcobellum is very muscular, but at least three glandular areas are present. The functions of these glands in copulation or courtship are unknown. The trifid appendage processes contain a peculiar granular secretion that coats their surfaces when they are everted. BURSA COPULATRIX Often known as the spermatheca, this structure is mainly concerned with the breakdown of genital products rather than their storage; bursa copulatrix is the more accurate term (Bayne, 1973). This gameto- lytic gland varies considerably in size but is always found closely attached to the free oviduct. It is lined by a tall columnar epithelium and the contents of the lumen vary from sperm and jelly mass to partially digested eggs. In Helix the spermatophore is so shaped that when it passes up the duct of the bursa compulatrix by peristal- sis, sperm are expressed through a hole in the spermatophore wall; only sperm that escape make their way to the spermatheca, and the rest are digested in the bursa (Lind, 1973). The mechanism by which sperm are released from the jelly mass of D. reticulatum is unknown. NERVOUS CONTROL OF REPRODUCTION Sections of the reproductive tract ex- amined in the electron microscope reveal the presence of many nerve fibres in the connective tissue in all areas, but distinct nerves are only present around the penial sac area and passing along the hermaphro- dite duct to the gonad. It seems likely, however, that the fine nerve fibres are part CONTROL OF DEROCERAS REPRODUCTION 349 of a plexus that exists throughout most of the tract (Minker & Koltai, 1961). Severing the nerves to the reproductive system has so far only been achieved with the intesti- nal nerve, which sends a branch to the hermaphrodite duct. Eggs laid after this operation are perfectly normal; but if part of the visceral ganglion is also destroyed, the eggs are misshaped and polyembryonic (Button, unpublished data). Much work remains to be done in this area. ENDOCRINE CONTROL OF REPRODUCTION Transplants of undifferentiated common ducts into the haemocoel of male-stage and female-stage recipients differentiate differ- ently. The prostate gland became well developed in the male-stage animals and the oviducal gland in the female-stage animal. As the transplants were free in the haemo- coel, differentiation and maturation must have resulted from exposure to blood-borne hormones. The source of these hormones was unknown in these early experiments (Runham, Bailey & Laryea, 1973). Two sources of hormones exist in Deroceras—neurosecretory cells in the brain and the glandular dorsal bodies that lie on the surface of the cerebral ganglia and commissure (Fig. 7). Extirpation of the dorsal bodies slowed considerably the de- 100 ym. velopment of the oviducal gland and albu- men gland, and this effect is reversed by a transplant of dorsal bodies from another animal (Wijdenes 8 Runham, 1976 and unpublished). This gland is therefore the probable source of a hormone controlling the development of the oviducal and albumen glands. The source of the hor- mone controlling the prostate gland has not been determined. Recently, transplants of the penial mass indicate that this organ is also under hormonal control. A study of the maturation of the tract and the changes in the gonad revealed that they are closely related phenomena (Run- ham & Laryea, 1968). Castration stops or greatly slows tract maturation; this is revealed by cessation of mitotic division in differentiating tracts, by decreased incor- poration of radioactively labeled amino acids in more mature tracts, and by their weight. These changes are also associated with a massive increase in the amount of glycogen stored in the tissues of the body (Runham, 1976 and unpublished data). The interactions of the dorsal body and gonad have not yet been studied. Another function of the dorsal body appears to be the control of oocyte growth. Extirpation leads to reduced num- bers of large oocytes and an accumulation of smaller oocytes (Wijdenes & Runham, 1976). In Lymnaea, oocyte enlargement is associated with vitellogenesis (Geraerts & FIG. 7. Longitudinal section through the cerebral commissure and dorsal body. C, cerebral commissure; D, dorsal body. 350 ВУМНАМ Joosse, 1975), but this has not been studied in Deroceras. Many workers have attempted to repeat the work of Pelluet (1964) and Pelluet & Lane (1961), which established hormonal control of oogenesis by secretions from the tentacle and cerebral ganglia and control of spermatogenesis by cerebral ganglion secre- tions. In Bangor we have failed completely to repeat this work, but in Arion (Wattez, 1973) it seems likely that the tentacles do produce a secretion that inhibits oogenesis. Full clarification of these relationships really depends on developing simple accurate methods of estimating gamete production. CONCLUSIONS As is evident, we have some understand- ing of reproduction in D. reticulatum but there remain many areas where detailed information is lacking. In particular, inte- gration of the reproductive tract during copulation and egg laying and the mechan- isms that time maturation of various parts of the reproductive tract in this protandric hermaphrodite are virtually unknown. Other species of slug and most other pulmonates are even less well known. Environmental factors are known to influ- ence reproduction in temperate species, but again detailed information is lacking. Parti- cularly valuable would be a study of reproduction in species with a wide geo- graphical distribution to discover how cli- matic factors, expecially at the limits of their distribution, affect the complex proc- ess involved. LITERATURE CITED BAILEY, T. G., 1970, Studies on organ cultures of slug reproductive tracts. Ph.D. thesis, Uni- versity of Wales. BAILEY, T. G., 1973, The in vitro culture of reproductive organs of the slug Agriolimax reticulatus. Netherlands Journal of Zoology, 23: 72-85. BAYNE, C. J., 1966, Observations on the composition of the layers of the egg of Agriolimax reticulatus, the grey field slug (Pulmonata, Stylommatophora). Comparative Biochemistry and Physiology, 19: 317-338. BAYNE, C. J., 1967, Studies on the composition of extracts of the reproductive glands of Agriolimax reticulatus, the grey field slug (Pulmonata, Stylommatophora). Comparative Biochemistry and Physiology, 23: 761-773. BAYNE, C. J., 1973, Physiology of the pulmo- nate reproductive tract: location of sperma- tozoa in isolated, self-fertilizing succinid snails (with a discussion of pulmonate tract termi- nology). Veliger, 16: 169-175. GALANGAU, V., 1964, Le cycle sexuel annuel de Milax gagates (Drap.) et ses deux pontes. Bulletin de la Societe Zoologique de France, 89: 510-593. GERAERTS, W. P. M. & JOOSSE, J., 1975, Control of vitellogenesis and of growth of female accessory sex organs by the dorsal body hormone (DBH) in the hermaphroditic freshwater snail Lymnaea stagnalis. General and Comparative Endocrinology, 27: 450-464. LIND, H., 1973, The functional significance of the spermatophore and the fate of sperma- tozoa in the genital tract of Helix pomatia (Gastropoda: Stylommatophora). Journal of Zoology, 169: 39-64. MINKER, E. & KOLTAI, M. D., 1961, Unter- suchungen an isolierten Gastropodenorganen. Acta Biologica Hungarica, 12: 199-209. NICHOLAS, J., Ms., A light microscope study of the carrefour of Deroceras reticulatum (Agrio- limax reticulatus) Müller (Pulmonata: Lima- cidae). PELLUET, D., 1964, On the hormonal control of cell differentiation in the ovotestis of slugs (Gasteropoda: Pulmonata). Canadian Journal of Zoology, 42: 195-199. PELLUET, D. & LANE,N. J., 1961, The relation between neurosecretion and cell differentia- tion in the ovotestis of slugs (Gasteropoda: Pulmonata). Canadian Journal of Zoology, 39: 789-805. RUNHAM, N. W., 1976, The effects of castration on maturation of the reproductive tract of the pulmonate slug Agriolimax reticulatus. General and Comparative Endocrinology, 29: 293-294. RUNHAM, N. W., in preparation, Studies on the effect of castration in Deroceras reticulatum. RUNHAM, N. W., BAILEY, Т. С. € LARYEA, A. A., 1973, Studies of the endocrine control of the reproductive tract of the grey field slug Agriolimax reticulatus. Malacologia, 14: 135-142. RUNHAM, N. W. & HOGG, N., Ms., The gonad and its development in Deroceras reticulatum. Proceedings of the Sixth European Malacolog- ical Congress. Malacologia, 18. RUNHAM, N. W. & LARYEA, A. A., 1968, Studies on the maturation of the reproductive system of Agriolimax reticulatus (Pulmonata: Limacidae). Malacologia, 7: 93-108. ТОМРА, А. S., 1976, A comparative study of the ultrastructure and mineralogy of calcified land snail eggs. Journal of Morphology, 150: 861-888. WATTEZ, C., 1973, Effets de l'ablation des tentacules oculaires sur la gonade en crois- sance et en cours de régénération chez Arion subfuscus Draparnaud (Gastéropode Pulmoné). General and Comparative Endocrinology, 21: 1-8. WIJDENES, J. & RUNHAM, N. W., 1976, Studies on the function of the dorsal bodies of Agriolimax reticulatus (Mollusca: Pulmonata). General and Comparative Endocrinology, 29: 545-551. MALACOLOGIA, 1978, 17(2): 351-364 ТНЕ EVOLUTION OF LIFE-CYCLE STRATEGIES IN FRESH-WATER GASTROPODS P. Calow Department of Zoology, University of Glasgow, Glasgow, G12 800, Scotland, United Kingdom ABSTRACT From a review of the available data | examine the adaptive significance of differences in three aspects of the life cycles of fresh-water gastropods: (a) the “choice” between semelparity and iteroparity; (b) the “choice” between egg size and egg numbers; (с) the “choice” between gonochorism and hermaphroditism. Most temperate, fresh-water gastropods are annual and semelparous. On the other hand, marine and terrestrial species tend, in general, to be perennial and iteroparous. Since fresh-water species must have evolved from marine and possibly terrestrial ancestors, the question arises why iteroparous species should have evolved into a semelparous condition. In the gastropods the semelparous state seems to be associated with reproductive recklessness on the part of the parent, and this strategy has probably evolved in association with adaptations that ensure a greater chance of survival of offspring. As gastropods have invaded fresh water there has been a trend toward producing larger sized eggs, telescoping developmental stages into the egg, and thus producing larger, more fully developed hatchlings. These adaptations can be explained as a response to the challenge posed by inclement conditions. Some fresh-water gastropods produce larger eggs than others, and these differences can be correlated with differences in ecology. However, since the number of progeny that ultimately reach maturity depends both on the fitness of individual progeny and on their initial density, different strategies (in terms of the choice between egg size and numbers) can feasibly lead to equivalence in parental fitness under the same ecological conditions. This principle, of alternate equifit adaptations to the same environmental challenge, will be illustrated by reference to littoral species. Most pulmonates are hermaphrodite, but many prosobranchs have separate sexes and are viviparous. These differences are discussed in terms of current theory. INTRODUCTION In this review | consider the adaptive significance of various aspects of the life- cycle strategies of fresh-water gastropods. By life-cycle strategy | mean the complex set of adaptations which ensure that the products of reproduction reach a condition whereby they can themselves reproduce. It is assumed that selection will have operated in such a way that these strategies are optimally adapted to maximize fitness—in other words, so that they are matched (or nearly matched; see Rapport & Turner, 1977) to meet the challenge offered by a particular set of environmental conditions. Two types of study have been directed at life-cycle problems. Firstly, theoretical studies have sought to deduce the optimal strategies to be expected under specified ecological conditions from general evolu- tionary principles (Stearns, 1976). Sec- ondly, observational studies have sought to correlate particular strategies with particu- lar ecological circumstances and thereby to arrive inductively at the way selection optimizes life cycles under natural condi- tions (e.g. Tinkle, 1969). Clearly, the deductive and inductive methods ought ultimately to complement each other and to contribute jointly to a deeper under- standing of life cycles. The observational approach is most powerful when carried out on closely related species. Fresh-water gastropods pro- vide a good opportunity for this kind of study since there is already a considerable amount of reliable, descriptive information available in the literature on their life cycles. Here | intend to use these published data to consider what strategies are used by gastropods and when and why. Firstly, | consider what general patterns of life cycle are adopted by fresh-water gastropods; in this context | use Cole’s (1954) termi- nology of semelparity and iteroparity to (351) 352 CALOW distinguish between different patterns. The semelparous condition is when parents die after reproduction. The iteroparous condi- tion is when parents live on after reproduc- tion to reproduce again. Secondly, | con- sider to what extent fresh-water gastropods compromise between egg size and egg numbers. Since only a finite amount of energy is available for gamete production all organisms have to “choose” between producing a large number of small eggs or a small number of large eggs. Which strategy is adopted can usually be correlated with the prevailing ecological conditions (Calow, 1977; Wilbur, 1977), and several theoretical attacks have been made on this subject (Pianka, 1970). Thirdly and finally, | consider what methods are involved in reproduction in fresh-water gastropods in terms of gonochorism, hermaphroditism, parthenogenesis, and viviparity. Russell- Hunter (1964; in press) has already made some interesting comment on this subject from the point of view of the fresh-water pulmonates, and a recent theoretical paper by Heath (1977) provides a framework for my discussion here. LIFE-CYCLE PATTERNS OF FRESH-WATER GASTROPODS Little can be added, except in detail, to the reviews of Russell-Hunter (1961a, 1961b, 1964, in press) on the life cycles of fresh-water gastropods. He was able to distinguish between several patterns of life cycle in this group (Fig. 1). The annual condition, with a spring breeding season and complete replacement of one genera- tion by another, predominates (TYPE A); but a second, summer breeding season, either without (TYPE B) or with (TYPE C) a replacement, is common. A third breed- ing season in autumn, again either without (ТУРЕ. 09). ог withhe(TY¥PES E & Е) replacement, has been reported for several populations. Finally, а true perennial (sometimes just biennial) condition (TYPE G), though less common, can occur. Life- cycle types A, C, and F are semelparous; life-cycle types B, D, and E are quasi- iteroparous, in that they occur in special circumstances in species which are usually semelparous; life-cycle type G is the true iteroparous condition. Table 1 catalogues the data available on the life-cycle patterns of temperate gastro- SNAIL SIZE SPRING SUMMER FALL WINTER SPRING SUMMER FALL WINTER FIG. 1. Patterns of life cycle to be found т fresh-water gastropods. Circles, reproduction be- gins; triangles, egg capsules appear. ALL SPECIES number of species en © 27 VILLILILLS. == PROSOBRANCHIA A B С D E E G life-cycle type FIG. 2. Distribution of species between life-cycle types. pods using the above typology, and Fig. 2 gives a summary. Considering the distribu- tion of species between life-cycle types, a number of important points emerge: (1) LIFE CYCLES OF GASTROPODS 353 TABLE 1. Review of life-cycle type and fecundity characteristics in 37 species of fresh-water gastropods. Egg Life No./ cycle Size, No./ ind./ Species type mm capsule season Authority PULMONATA Ly mnaeidae Acella haldemani A 1х 0.6 3-12 ? Morrison, 1932 Lymnaea elodes B e) 20-47 ? Eisenberg, 1966 L. humilis (С ? 2 ? Van Cleave, 1935 A ? 6-17 ? McCraw, 1961 L. palustris ? 0.84 X 0.58 10-40 ? Bondesen, 1950 G ? ? ? McCraw, 1970 A ? ? ? Eckblad, 1973 G ? 20 250-370 Hunter, 1975 B ? — = Hunter, 1975 L. peregra ? 1.09 X 0.79 200 ? Bondesen, 1950 A ? 12-20 300-1100 Russell-Hunter, 1961a © ? — — Russell-Hunter, 1961a A ? ? ? Young, 1975 L. stagnalis ? 1.37 Х 1.00 100 2 Bondesen, 1950 G ? ? ? Fromming, 1956 G ? 100-150 1000-1500 Berrie, 1965, 1966 G ? ? ? De Coster & Persoone, 1970 L. truncatula F ? 7-9 % Walton & Jones, 1926 ? 0.78 X 0.53 15 ? Bondesen, 1950 € ? ? ? Heppleston, 1972 (© ? Y ? Bruce et al., 1973 Physidae Aplexa hypnorum ? 1.21 X 0.93 20 ? Bondesen, 1950 A % ? ? Hartog & de Wolf, 1962 A ? ? ? Vlasblom, 1971 Physa acuta ? 10 0:5 ? ? Bondesen, 1950 A Y ? ? Duncan, 1959 P. fontinalis ? 1Х 0.8 20 ? Bondesen, 1950 C ? 6-18 180 Wit, 1955 A ? 5-11 ? Duncan, 1959 A ? 15 174 Russell-Hunter, 1961a & b B 2? - _ Russell-Hunter, 1961a & b A ? ? to Girod, 1969 P. gyrina ? ? 100-200 ? Bondesen, 1950 — 0.79 X 0.72 13 272 DeWitt, 1954 A — - - DeWitt, 1955 B — — - DeWitt, 1955 e ? ? 700 Clampitt, 1970 P. integra B ? ? 1000 Clampitt, 1970 A ? ? ? Eckblad, 1973 P. virgata D ? 11-17 ? McMahon, 1975 Planorbidae Anisus rotundatus ? ? 8 ? Bondesen, 1950 G ? ? ? Marazanoff, 1970 Helisoma trivolvis A ? 20 1600 Eversole, 1974 G ? - — Eversole, 1974 Planorbis albus ? 0.73 X 0.50 5 ? Bondesen, 1950 A ? ? fé Russell-Hunter, 1961a & b Р. carinatus A ? ? ? Young, 1975 P. contortus ? 0.70 X 0.54 ? 2 Bondesen, 1950 A 1 X 0.50 3.5 34 Calow, 1972a С ? ? ? De Coster & Persoone, 1970 P. corneus ? 1.68 X 1.50 30-50 ? Bondesen, 1950 A ? Y ? Berrie, 1963 P. planorbis ? 0.78 X 0.68 30 ? Bondesen, 1950 A ? 2 — De Coster €: Persoone, 1970 P. vortex ? 0.76 X 0.60 20 ? Bondesen, 1950 A 2 ? ? De Coster & Persoone, 1970 Ancylidae Ancylus fluviatilis ? 1.42 X 1.20 2-3 ? Bondesen, 1950 A ies 4 46 Russell-Hunter, 1953 A 10023 4 50 Geldiay, 1956 A 13081220 3 31 Calow, 1972a 354 CALOW TABLE 1 (Continued). Life cycle Size, Species type mm A. lacustris ? 0.80 X 0.47 A ? Ferrissia rivularis A ? LS ? B 2 Hebetancylus excentricus G 1.82 Laevapex fuscus A 2.10 D = Е 2 PROSOBRANCHIA Hydrobiidae Amnicola limosa A 2% Bithynia tentaculata G 0.1 G ? G ? Potamopyrgus jenkinsi G P Valvatidae Valvata humeralis A ? V. piscinalis A 0.05 A ? Viviparidae Campeloma rufum G VP Campeloma spp.* G № Viviparus contectoides G № V. fasciatus G V V. malleatus G V У. viviparus G V Egg No./ No./ ind./ capsule season Authority 8-10 ? Bondesen, 1950 ? ? Russell-Hunter, 1953 1-7 9 Burky, 1971 _ = Burky, 1971 2 ? Nickerson, 1972 3-5 ? McMahon, 1976 3-5 42 McMahon, 1976 — ~ McMahon, 1976 -= = McMahon, 1976 ? ? Pinel-Alloul & Magnin, 1973 10 25-100 Lilly, 1953 ? ? Schaffer, 1953 ? ? Pinel-Alloul & Magnin, 1971 Р Р Robson, 1923 ? ? Gillespie, 1969 9-18 Pe Cleland, 1954 ? ? Gillespie, 1969 VP МР Van Cleave & Altringer, 1937 V V Medcof, 1940 V V Van Cleave & Lederer, 1932 № V Starczykowska, 1959 М V Stahiczy kowska et al., 1971 V V Miroshnitshenko, 1958 P = parthenogenetic; V = viviparous; * described as new spp. and different from C. rufum most species, particularly pulmonates, are semelparous and annual; (2) there may be intraspecific variations in life cycles—for example, Type À may transform to B, C, D, and Е in productive habitats (McMahon et al., 1974)—a point considered of particu- lar importance by Russell-Hunter (in press); (3) more of the type-G gastropods are exclu- sively iteroparous and are prosobranchs. However a few species with iteroparous populations also include populations show- ing other patterns. Also listed in Table 1 are records on the size and numbers of eggs produced per parent. Fresh-water gastropods produce in the order of 102-10? eggs per breeding season with diameters of about 1mm. This contrasts with marine gastropods, where eggs are usually 102-103 ит in diameter and where fecundity per breeding season may be as high as 10° eggs per parent (Fretter € Graham, 1964). In common with these marine prosobranchs, Melampus bidentatus, a primitive, saltmarsh pulmonate, produces approximately 33,000 very small eggs per individual per breeding season (Russell-Hunter & Apley, 1966; Apley et al., 1967; Apley, 1970; Russell- Hunter et al., 1972). Hence, the invasion of fresh waters by the gastropods, particularly the pulmonates, has been accompanied by a switch from emphasis on egg numbers to emphasis on egg size. This is а well- documented trend (Bondesen, 1950), which has been ascribed to the more stressful nature of the fresh-water habitat. The “dilute”” conditions present osmotic prob- lems, and there can be violent fluctuations in many of the physicochemical aspects of freshwater ecosystems. This has necessita- ted suppression of the sensitive planktonic stage and the telescoping of development into the egg (Fioroni & Schmekel, 1975). As a result, the egg must carry a more extensive yolk store and provide space for extended development. Hence, the empha- sis must be on egg size rather than numbers. LIFE CYCLES ОЕ GASTROPODS 355 EVOLUTION OF SEMELPARITY It seems very likely that the predomi- nant semelparous condition of fresh-water gastropods has evolved out of a primitive, iteroparous condition. Most marine proso- branchs are iteroparous (Fretter & Graham, 1964), and the primitive saltmarsh pulmo- nate Melampus bidentatus is iteroparous (Russell-Hunter & Apley, 1966; Apley et al., 1967; Apley, 1970; Russell-Hunter et al., 1972). The question therefore arises why the invasion of fresh waters from marine and perhaps terrestrial environments is associated with a shift from iteroparity to semelparity. Cole (1954), in his classic paper on life cycles, raised the opposite question: what is the advantage of a semelparous species becoming iteroparous? He was puzzled because a shift from semelparity to iteroparity makes little di- rect contribution to progeny production and hence fitness. However, Cole carried out his analysis on “ideal populations” in which there was no mortality between breeding seasons, and this is clearly a very special case. Whenever, on the other hand, there is a high probability of juvenile mortality, it can be shown theoretically that it is worthwhile preserving the parent as an insurance strategy even if this might lead to slightly reduced fecundity (Stearns, 1976). The advantage of iteroparity comes not from the direct contribution it might make to the production of progeny but because it gives the same parent several chances at replicating and spreading its own genes. Given no apparent disadvantage of iteroparity but the obvious advantage of an “insurance strategy,’’ the next question is why all species are not iteroparous or, more specifically, why iteroparous species should ever return to semelparity? This might be referred to as the ‘gastropod paradox” since it is the question raised most obviously by a comparative treatment of gastropod life cycles. There are several possible solutions to the “gastropod paradox.’’ There may be a direct advantage in truncating the life-span of the parent; for example, as Weisman (1882) believed, it may provide more food and space for the potentially more virile progeny. It is difficult to see, however, how such altruism could have evolved when the progeny of different parents apparently intermix completely and freely. Intermixing would mean that all progeny, not just those carrying the gene for the altruistic character, would be advantaged by the death of the parent. Hence, selection would be impotent in favoring the gene(s) for parental death. Another possibility is that semelparity may have evolved indirectly from the active selection of another, correlated character. Several theories on the evolution of life cycles are based crucially on the assump- tion that a high reproductive effort leads automatically to a reduced adult life-span (Cody, 1966; Williams, 1966a, 1966b; Gad- gil & Bossert, 1970; Charnov & Krebs, 1974; Leon, 1976). Whenever the gains (in terms of viable progeny) accruing from an increased reproductive effort are greater than the potential loss due to weakening of the parent, then the increased effort will be selectively favoured and will bring with it a shorter adult life-span. On this theory, there should be a negative correlation between reproductive effort and life-span. A favoured measure of reproductive ef- fort is the proportion of available re- sources (ingested or absorbed energy) used in gamete formation (Tinkle & Hadley, 1975), but data of this kind are rare. | have therefore tested the hypothesis in two ways. Where data were available, | have used the ratio energy output in gametes ter SN NO energy input from food as a direct index of effort (DEI). Otherwise | have used the ratio (E X EV)/SV = IEI (indirect index of effort) Е =no. eggs produced/breeding season EV =egg volume (calculated from 4/3nr° and assuming that eggs are spherical SV = parent volume (calculated from 1/3rr?h and assuming that snails approximate to a cone, where r = egg or shell radius and h = shell height) IEl assumes that energy loss per egg is proportional to egg volume and that the demand egg production makes on the parent is roughly proportional to the relative masses of the adults and the total eggs spawned. 356 IEl’s are calculable from the data given in Table 1, together with additional infor- mation (derived in the main from the source references) on the shell dimensions of adults. The results are given in Table 2. The mean IEI for the semelparous species was 2.06; for the iteroparous species it was 0.28. These values were significantly differ- ent (+ = 4.42, P< 0.05) and suggest that semelparous species expended about 10 times more effort in reproduction than the iteroparous species. Data on DEI’s are summarized in Table 3, using both the total absorbed energy (TA) and the nonrespired part of TA (NRA) as energy input terms over the reproductive period. Here it has been possible to compare effort indices in popu- lations of the same species showing differ- ent life-cycle patterns. In all cases semelpar- TABLE 2. Indirect indices of reproductive effort (1E1) for several species of gastropod. SEMELPAROUS SPECIES Lymnaea peregra 1.23-4.88 (median = 3.05) Physa fontinalis 217 P. gyrina 2.01 Planorbis contortus 1.18 Ancylus fluviatilis 1.42-2.36 (median = 1.89) AVERAGE* 2.06 SPECIES WITH ITEROPAROUS POPULATIONS Lymnaea stagnalis 0.34-0.51 (median = 0.43) L. palustris 0.03-0.35 (median = 0.19) Bithynia tentaculata 0.08-0.33 (median = 0.21) AVERAGE* 0.28 *based on median values. CALOW ous species-populations had values between 10 and 40% greater than the iteroparous species-populations. As judged by the “‘t test,”” average indices for semelparous spe- cies were significantly greater than those for iteroparous species (t = 3 to 4, Р < 0.05). Marine prosobranchs, which are pre- dominantly iteroparous, often produce more eggs than fresh-water gastropods but probably put less effort into reproduction. For example, egg output may be as much аз 10°/female/breeding season in this group, but egg diameter is often less than 0.1 mm. Hence with snails usually as large or larger than Lymnaea stagnalis, the IEI is 0.02. Grahame (1973) has calculated an energy budget for the iteroparous Littorina littorea. Here the DEI (based on TA) was 10.2, which is less than the figures quoted for fresh-water species in Table 3. There is good evidence, therefore, for the hypothesis that a negative correlation exists between reproductive effort and adult life-span, and this is almost certainly explicable by a causal link between these two variables. However, there is at least one important exception to the rule. Melampus bidentatus, the primitive salt- marsh pulmonate, produces about 33,000 eggs per female per breeding season (Russell-Hunter et al., 1972). It has ап IEI of 0.89 (calculated from the data of Apley et al., 1967, using egg and snail weight rather than volume) and a DEI (with NRA as denominator) of 80% (Apley et al., 1967). This key species, though iteropar- ous, puts considerably more effort into reproduction than many semelparous fresh- TABLE 3. Proportion of input energy (taken over the reproductive period), measured either as total ab- sorbed energy (TA) or as the nonrespired portion of TA (NRA), diverted into gamete production (REP). SEMELPAROUS SPECIES REP/TA REP/NRA Source Planorbis contortus 22 79 Calow, 1972a Ancylus fluviatilis 20 51 Calow, 1972a Ferrissia rivularis 15-20* 45 Burky, 1971 SPECIES WITH SEMELPAROUS ($) AND ITEROPAROUS (I) POPULATIONS Lymnaea palustris S 20 67 Hunter, 1975 1523 27 Bithynia tentaculata S 28* 55 Mattice, 1972 | 5 25 Calow, unpubl. Helisoma trivolvis $ 35* 69 Eversole, 1974 | 16* 33+ AVERAGES $ 23.8 61.0 | 7.8 28.3 *Calculated assuming that respiration is approximately equal to NRA. *From a caged population under conditions roughly equivalent to those that promote iteroparity. LIFE CYCLES OF GASTROPODS 357 water species. Hence, the relationship be- tween effort and life-span cannot be as simple and as straightforward as was ini- tially anticipated. An alternative way that the life-span might be indirectly shortened by the selection of another trait is if, under adverse conditions, semelparous species are those which risk effort in reproducing despite a possible adverse effect on the parent, whereas iteroparous species show restraint (Calow, 1973a). Selection here would be for reproductive recklessness or restraint on the part of the parent, and the length of the life-span would follow as an indirect consequence of the prime effect. This seems to be true in fresh-water triclads (Calow & Woollhead, 1977), and several lines of evidence support a similar hypothe- sis for the fresh-water gastropods. Firstly, because risk is likely to vary from place to place and time to time, it is probable that there will be spatial and temporal variation in the life-cycle strategies of reckless repro- ducers; this is indeed the case for fresh- water gastropods (Table 1). Secondly, many semelparous gastropods can be kept alive and in a reproductive condition for periods exceeding their ‘natural’ life-span by maintaining them in good, low-risk conditions in the laboratory (Comfort, 1957; Calow, 1973a). Thirdly, and finally, it is possible under laboratory conditions to obtain some т- dication of differential recklessness and restraint by measuring reproductive activi- ties under stress. Table 4, for example, illustrates a graded response to starvation in the reproductive output of several gastro- pods with different life-cycle patterns. In Bythinia tentaculata and Lymnaea stagnalis (with iteroparous populations) egg output stops quickly during starvation, whereas L. peregra (semelparous) shows a more slug- gish response, maintaining capsule produc- tion recklessly, for some time after starva- tion begins. In the iteroparous Littorina littorea gonadal activity ceases very promptly after the onset of starvation (Le Breton, 1971). There are associated differ- ences in the response of the gonad itself. This organ disappears rapidly during starva- tion of L. /ittorea. In В. tentaculata the gonad reduces rapidly in size during starva- tion, but gametogenesis still takes place for some time (Neuhaus, 1949). There was no histological data for Lymnaea peregra, but from evidence in Table 4 it seems that gonadal size and activity must be main- tained for a considerable time during starvation. It has also been shown that under a reduced food supply (rather than complete starvation) reproductive activity may be constrained in Helisoma trivolvis, a species with iteroparous populations (Ever- sole, 1974), but not in Planorbis contortus, a semelparous species (Calow, 1973a). What is urgently needed is more data on the relative partitioning of energy between growth, maintenance, and reproduction in semelparous and iteroparous species (Ever- sole, 1974). Data on Melampus bidentatus would also be interesting, since this is the species that did not seem to fit into the hypothesis that the life-spans of semelpar- ous species were shortened by high repro- ductive effort per se. It seems likely from the above that the evolution of semelparity in fresh waters has occurred indirectly from the combined selection for increased reproductive effort and reproductive recklessness. Clearly, these strategies can only be favored when prog- eny have an equal or better chance of TABLE 4. Average egg output per week in fully fed and completely starved snails in the laboratory. Weeks after initiation of starvation in experimental group 0 1 2 3 4 5 Source {Fed ES i a * = nc Calow, unpubl. Bithynia tentaculata | Starved $ 4 Fed 20.6 32.6 34.9 2273 38.3 28.2 Joosse et al., 1968 Lymnaea stagnalis Starved 22.4 13.4 0.1 0 0 0 Fedt 21 30 40 42 40 20 Calow, unpubl. L. peregra Starvedt 21 30 10 1 0.1 0 *data approximate. *signifies occurrence of capsule production—no quantitative data. 358 CALOW surviving than parents; that is, when high effort and recklessness bring real gains in transmission and multiplication of genes associated with these characters. | suggest that, paradoxically, this shift is a response to the harsh fresh-water conditions. The whole emphasis in the invasion of fresh water has been on the confinement of the sensitive, developmental stages within a protecting, egg-capsule envelope. Hence, larval mortality is likely to be much less here than in species that have a small, poorly developed, planktonic stage, and the adult phase will be “needed” less as an insurance policy. Under these circum- stances, it is probably advantageous, in terms of fitness, to invest more energy in progeny at the expense of putting the parent at risk. Given that the parent has a reduced chance of postreproductive sur- vival, it is also possible that “aging genes” (п the sense of Medawar, 1952, and Williams, 1957) may express themselves at this time and thereby fix, in genetic terms, what was originally a flexible, probabilistic response. Interestingly, species that have retained (or re-evolved) the iteroparous strategy are those which tend to inhabit small closed bodies of fresh water (particularly the prosobranchs). Under such conditions it is likely that there is more intense competi- tion for limited resources (space and food), more density-dependent control, and hence, a greater premium on the survival of a large “experienced” adult (Pianka, 1970; Calow, 1977). Moving from the poles to the tropics this “K selection” (Pianka, 1970) is supposed to become more intense and so iteroparity should become more frequent. However, many tropical fresh-water systems dry up seasonally and presumably demand as much, if not more, opportunism as in the more typical “‘r situations’’ of selection. Good studies on the life cycles of tropical gastropods are few, but they do point to opportunism and reproductive recklessness (Olivier & Barbosa, 1955a, 1955b; Pringle & Raybould, 1965; Jobin & Michelson, 1967; Sturrock, 1973). EGG SIZE AND NUMBERS In general, fresh-water gastropods invest more energy in reproduction than marine species but they partition it into fewer eggs. Even within the fresh-water species, though, there is some variation in egg size and egg numbers (Table 1). These differ- ences might ultimately be explained by differences in the ecology of each species, and more work on this problem would undoubtedly yield rewards. However, phylogenetic limitations may also be т- volved in this relationship. For example, those prosobranchs which are not vivi- parous produce small eggs about 103 ит т diameter and less. This could represent an inability of the prosobranch organization to produce larger spawn and may explain why several fresh-water species have adopted a viviparous mode of reproduction (see next section). It is also feasible that with complex adaptations, like life-cycle strate- gies, there may be more than one way of responding to a particular environmental challenge, and | illustrate that principle in this section. A most intense environmental challenge to a fresh-water fauna comes from water movements in fast-flowing streams and wave-swept lake shores. In the United Kingdom, two pulmonate gastropods are particularly successful in these habitats; the river limpet, Ancylus fluviatilis, and the more globose Lymnaea peregra. A smaller ramshorn snail, Planorbis contortus, is oc- casionally found at high densities in these habitats (Calow, 1973b, 1974). All three species are semelparous but show signifi- cant differences in the reproductive aspects of their life cycles. A. fluviatilis lays fewer eggs than L. регедга (Table 1), but both spend about 40 Joules in total egg produc- tion (Calow, 1972a). Hence, the number of Joules/egg is about 30-fold greater in A. fluviatilis than in L. peregra. P. contortus adopts a somewhat intermediate strategy, laying a smaller number of larger eggs than L. peregra but a larger number of smaller eggs than A. fluviatilis (Table 5). The reproductive products of the three species also differ in morphology. The capsules of Lymnaea peregra are globose and covered with a tertiary membrane (terminology after Bondesen, 1950) pro- duced in the reproductive tract. Alterna- tively, both Ancylus fluviatilis and Planor- bis contortus produce flattened capsules covered over with an extra secretion, a definite quaternary membrane of foot ori- gin in the former species and a loose mucoid secretion in the latter (Bondesen, 1950). These differences seem to have a marked effect on the survival of egg LIFE CYCLES ОЕ GASTROPODS 359 capsules. For example, my own observa- tions on marked capsules in both a small tarn and a large British lake suggest that of the capsules laid, approximately 90% sur- vive to hatch in A. fluviatilis, 70 to 80% in P. contortus, and 40 to 50% in L. peregra. The flattened capsules appeared to be much less vulnerable than the globose, and the quaternary membrane in A. fluviatilis may have endowed it with survival superiority. The larger eggs and richer reserves typical of Ancylus fluviatilis enable more complete, intracapsular development in this species. At 10°C, A. fluviatilis takes 26 days to hatch; on emergence it is roughly 6% the shell-free, dry weight of the adult (Calow, 1972a). Lymnaea peregra and Pla- norbis contortus may hatch out in under half that time (Calow, 1972a, and unpub- lished observation) but are only 1 to 2% of their final adult size by shell-free dry weight. The hatchlings of A. fluviatilis are likely to be better equipped than those of the other species to cope with the harsh environment; this is reflected in significant differences in survivorship curves between species (Fig. 3). The data for P. contortus and A. fluviatilis derive from a different habitat than those for L. peregra, so this complicates the comparison. However, as- suming the habitat effect is negligible (both are in fact very similar in general ecology), the possible influence of size of hatchlings on survivorship can clearly be seen. The survivorship curve for A. fluviatilis con- forms to a Deevey type II curve (Deevey, 1947) in which the rate of mortality and the mean life expectancy are constant throughout life. L. peregra and P. contortus curves conform more closely to a Deevey type Ill curve, in which mortality mainly affects the young, mean life expectancy increases with age, and median life expect- ancy is smaller than the mean. It is possible to combine all the above data by using the fitness-set analysis of Smith & Fretwell (1974). This is a graphic technique in which the abscissa plots both size and number of individual offspring (the relationship of these two variables is determined by assuming а fixed finite amount of energy available for reproduc- tion) and the ordinate plots the fitness of individual offspring. Straight lines through the origin of this graph are fitness func- tions, or lines of equal fitness for the parent (see also Levins, 1968). Greater slopes of the fitness function correspond to higher parental fitness. Using the data in Fig. 3 in combination with estimates of capsule survival it is possible to calculate the approximate fit- ness of the individual offspring as the combined probability of their survival to hatching and then to maturity. Joules/egg and total numbers of eggs produced/parent may be obtained from Table5. Points representing Ancylus fluviatilis, Lymnaea SURVIVORS RTE OEM TT MONTHS FIG. 3. Logarithmic survivorship curves. Those for Ancylus fluviatilis and Planorbis contortus are for populations from the littoral region of Malham Tarn (Calow, 1973b) and are based on sequential samples using the technique of Calow (1972b). The zurve for L. peregra is for a population from the littoral region of Loch Lomond, Scotland, and is based on a sampling technique in which stones of different, prede- termined size categories were collected from random sampling points; knowing the relative abundance of stones over the shore (determined from a random sample of 1,000), snail densities per stone of given size were converted, according to the method of weighted averages, to densities per “average stone.” In all cases densities were corrected to an initial level of 100. TABLE 5. Some quantitative data on reproductive output. No. Joules eggs/ in eggs/ lifetime/ lifetime/ snail snail J/egg Ancylus fluviatilis 30 ca. 40 133 Lymnaea peregra 1000 ca. 40 0.04 Planorbis contortus 50 ca. 40 0.80 360 CALOW > u fitness of individual progeny x 100 wo rs J/egg 20% no.eggs 0-1 0-5 10 400 80 zu FIG. 4. Fitness-set analysis. See text for further explanation. peregra and Planorbis contortus can be located roughly in the fitness-set space (Fig. 4). Since these points are calculated on the basis of information from specific populations at a specific time, they should not be taken as species-specific constants but rather as indices of the relative ability of the different species to cope with the littoral challenge. To facilitate comparison | have drawn in the fitness set that corre- sponds to perfect, annual replacement. The points for A. fluviatilis and L. peregra lie above this set, indicating their ability to cope with the littoral challenge. Further- more, both points lie close to but at opposite ends of the same set. Hence, the alternative strategies of “egg size” and “egg numbers” may balance to produce equifit strategies under similar environmental con- ditions. In comparison with the other results, the point for Planorbis contortus lies on a fitness set below that representing replace- ment. This shows that P. contortus is less well adapted for the littoral challenge than the other species. Hence, without con- tinuous replacement it is likely that littoral populations of this species might ultimately become extinct. There is evidence in one lake, however, that littoral populations may be sustained by a process of continuous immigration of P. contortus into exposed littoral sites from sheltered refuges (Calow, 1972a, 1974). METHODS OF REPRODUCTION Hermaphroditism is the dominant method of reproduction in the Mollusca, and the fresh-water gastropods, particularly the pulmonates, are no exception to this rule. For most phyla it is usually suggested that hermaphroditism is secondarily derived from a gonochoristic condition, but there is strong evidence that in the Mollusca it is the primary state (Morton, 1964). There are two advantages to hermaphroditism. First, the fact that all individuals are both male and female doubles the chance of any one individual meeting a mate. This is important either when population density is low or when individuals in the popula- tion are sluggish (Altenburg, 1934; Tomlin- son, 1966). Secondly, in principle a popula- tion of hermaphrodites can produce twice as many progeny as a population of gonochorists, because there are twice as many females (Maynard Smith, 1971). In semelparous snail populations where individuals are slow-moving and where densities can be very low during the breeding season, the advantages of her- maphroditism are obvious. It is interesting from this point of view that gonochorism in the Mollusca is only widespread through the more mobile Cephalopoda. There is a price to be paid, of course, for hermaphro- ditism: the cost of producing and maintain- ing two sets of reproductive apparatus in one animal is greater than the cost of building and maintaining single sets of apparatus in separate animals (Heath, 1977). This extra cost will become less tolerable as population density and the chances of encountering a mate increase, but it can be offset by any morphological adaptation that allows male and female organs to share common parts (ducts, etc.). In the fresh-water pulmonates, sharing of this sort can be extensive (Fig. 5); here even the gonad (ovotestis) is a shared organ. The latter is usually made up of several to many radiating acini, the lumina of which join at the origin of the her- maphrodite duct. In some species all acini can, at different times, contain stages of both oogenesis and spermatogenesis (Russell-Hunter & McMahon, 1976). The disadvantage of this “sharing strategy’ LIFE CYCLES OF GASTROPODS 361 FEMALE MALE FIG. 5. Hermaphrodite reproductive system of fresh-water pulmonates: OT = ovotestis; HD = hermaphrodite duct; AG = albumen gland; OOTH. G = oothecal gland; STH = spermatheca; PROST. G = prostate gland; PEN. C = penial complex. comes from the likelihood of self-fertiliza- tion. In many groups this leads to impaired vigor in the progeny (Heath, 1977). How- ever, even after twenty years enforced self-fertilization there was по apparent reduction in the viability of Lymnaea columella (Colton & Pennypacker, 1934). Anyway, most species of pulmonate are true simultaneous hermaphrodites and only the occasional species is protandric (Russell-Hunter & McMahon, 1976). In the prosobranchs the methods of reproduction are more diverse. Several species are gonochoristic, and at least two species are parthenogenetic. Of the latter, Potamopyrgus jenkinsi is interesting since in Europe over the course of this century it has rapidly invaded fresh-water from brack- ish habitats (Bondesen & Kaiser, 1949; Hubendick, 1950; Russell-Hunter & War- wick, 1957). This invasion has probably depended crucially on its parthenogenetic habit. As already noted, the small size of their spawn may have “forced” prosobranchs into viviparity. On the basis of arguments in an earlier section it is surprising that this strategy, which offers protection to young, should invariably be associated with itero- parity (Table 1). There are two possible explanations, which are not exclusive. Vivi- parity may mitigate recklessness since its success will depend upon the continuing survival of the parent through the brooding period. Also, physical limits to the brood pouch may limit fecundity, thereby pre- venting any adverse, proximate influence on the parent. It may also prevent the advan- tage ever shifting completely to progeny production at the expense of the parents. CONCLUSIONS AND FURTHER STUDIES In this paper | have advanced a broad theory of gastropod life-cycle strategies which is consistent with extensive but widely scattered malacological data and with the theoretical framework summarized in Stearns (1976). Further progress will require the collection of more quantitative data at the level of both population and individual and from both field and labora- tory studies. At the population level there is already an extensive body of data on life-cycle patterns (Table 1 and Fig. 1). What is urgently needed is more information on the quantitative aspects of the dynamics of molluscan populations. A fundamental need is information on population density at different times of the year, so that life tables and survivorship curves of the sort summarized in Fig. 3 can be constructed. Only when these kinds of data are available will it be possible to measure age-specific survivorship and the relative mortality schedules of iteroparous and semelparous species and to test the predictions that age-specific mortality should be more in- tense in the juveniles of iteroparous as opposed to semelparous populations. Fur- thermore, only when we know more about the population dynamics of fresh-water snails will it be possible to test out the consequences of “’r’’ and “К” selection on molluscan life cycles (Stearns, 1976) and to correlate particular life-cycle strategies with the demographic properties of populations occupying particular ecological conditions. Of specific interest, for example, would be comparisons between tropical and temper- ate populations, lotic and lentic popula- tions (Calow, unpublished data), and litto- ral and benthic populations. The major difficulty, of course, in these life-cycle studies has been, and will be, the develop- ment of methods for obtaining accurate density measures in complex habitats like weed beds and particularly stony shores. Techniques are slowly coming available (Calow, 1972b), and continuing efforts in this area are bound to be fruitful. At the level of the individual, more quantitative information is required on molluscan metabolism and energetics. In Particular, energy budgets should be con- structed at different ration levels, for only 362 CALOW in this way will it be possible to obtain a quantitative measure of degrees of reckless- ness and restraint. The 2 major predictions from this paper are that compared with iteroparous snails, individuals from semel- parous populations should (1) invest a greater proportion of their energy flow in reproduction and (2) recklessly maintain reproductive output despite disturbances in food supply. Histological data on gonadal morphology under various trophic condi- tions would also be useful. Finally, more attention should be paid to the correlation between the nature of the spawning products, both in size and numbers, and the nature of the ecological challenge experienced by a population, perhaps using the fitness-set analysis de- veloped here. This will depend on the availability of quantitative information on population dynamics. ACKNOWLEDGEMENTS | thank Professor W. D. Russell-Hunter for allowing me to see part of his review while it was still in press. LITERATURE CITED ALTENBURG, E., 1934, A theory of hermaphro- ditism. American Naturalist, 68: 88-91. APLEY, M. L., 1970, Field studies on life history, gonadal cycle and reproductive peri- odicity in Melampus bidentatus. Malacologia, 10: 381-397. APLEY, M. L., RUSSELL-HUNTER, W. D. & AVOLIZI, R. J., 1967, Annual reproductive turnover in the salt-marsh pulmonate snail, Melampus bidentatus. Biological Bulletin, 133: 455-466. BERRIE. A. D.. 1963. Life-cycle of Planorbarius corneus (L.). Nature, 198: 805-806. BERRIE, A. D., 1965, On the life cycle of Lymnaea stagnalis (L.) in the west of Scot- land. Proceedings of the Malacological Society of London, 6: 283-295. BERRIE, A. D., 1966, Growth and _ seasonal changes in the reproductive organs of Lym- naea stagnalis (L.) Proceedings ot the Malaco- logical Society of London, 37: 83-92. BONDESEN, P., 1950, A comparative morpho- logical-biological analysis of the egg capsules of freshwater pulmonate gastropods. Natura Jutlandica, 3: 1-208. BONDESEN, P. & KAISER, E. W., 1949, Hydrobia (Potamopyrgus) jenkinsi Smith in Denmark illustrated by its ecology. Oikos, 1: 252-281. BRUCE, В. G., ARMOUR, J. & CORBA, J., 1973, A further study of the epidemiology of ovine fascioliasis in Scotland and its control using molluscicides. Veterinary Record, 92: 518-526. BURKY, A. J., 1971, Biomass turnover, respira- tion, and interpopulation variation in the stream limpet Ferrissia rivularis (Say). Eco- logical Monographs, 41: 235-251. CALOW, Р., 1972a, The structural and functional dynamics of selected species populations of freshwater snails: towards a systems approach. Unpubl. Ph.D. thesis, University of Leeds. CALOW, P., 1972b, A method for determining the surface areas of stones to enable quantita- tive density estimates of littoral, stone-dwell- ing organisms to be made. Hydrobiologia, 40: 37-50. CALOW, P., 1973a, The relationship between fecundity, phenology and longevity: a systems approach. American Naturalist, 107: 559-574. CALOW, P., 1973b, Gastropod associations with- in Malham Tarn, Yorkshire. Freshwater Bi- ology, 3: 521-534. CALOW, P., 1974, Some observations on the dispersion patterns of two species-populations of littoral, stone-dwelling gastropods (Pul- monata). Freshwater Biology, 4: 557-576. CALOW, P., 1977, Ecology, evolution and ener- getics: a study in metabolic adaptation. In: MACFADYEN, A., Ed., Advances of Eco- logical Research, vol. 10. Academic Press, London and New York. CALOW, Р. & WOOLLHEAD, A. S., 1977, The relationship between ration, reproductive ef- fort and age-specific mortality in the evolution of life-history strategies—some observations of freshwater triclads. Journal of Animal Ecology, 46: 1-17. CHARNOV, Е. |. € KREBS, J. R., 1974, On clutch-size and fitness. /bis, 116: 217-219. CLAMPITT, P. T., 1970, Comparative ecology of the snails Physa gyrina and Physa integra (Basommatophora: Physidae). Ma/acologia, 10: 113-151. CLELAND, D. M., 1954, A study of the habits of Valvata piscinalis (Müller) and the structure and function of the alimentary canal and reproductive system. Proceedings of the Mala- cological Society of London, 30: 167-203. CODY, M., 1966, A general theory of clutch size. Evolution, 20: 174-184. COLE, L. C., 1954, The population consequences of life-history phenomena. Quarterly Review of Biology, 29: 103-107. COLTON, H. S. & PENNYPACKER, M., 1934, The results of twenty years of self-fertilization in the pond snail Lymnaea columella Say. American Naturalist, 68: 129-136. COMFORT, A., 1957, The duration of life in molluscs. Proceedings of the Malacological Society of London, 32: 219-241. DE COSTER, W. & PERSOONE, G., 1970, Ecological study of the Gastropoda in a swamp in the neighbourhood of Ghent (Bel- gium). Hydrobiologia, 36: 65-80. DEEVEY, D. S., 1947, Life tables for natural populations of animals. Quaterly Review of Biology, 22: 283-314. DEWITT, R. M., 1954, Reproduction, embryonic development and growth in the pond snail Physa gyrina (Say). Transactions of the Ameri- can Microscopical Society, 73: 124-137. DEWITT, R. M., 1955, The ecology of the pond snail Physa gyrina. Ecology, 36: 40-44. DUNCAN, C. J., 1959, The life cycle and ecology LIFE CYCLES OF GASTROPODS 363 of the freshwater snail Physa fontinalis (L.). Journal of Animal Ecology, 28: 97-117. ECKBLAD, J. W., 1973, Population studies of three aquatic gastropods in an intermittent backwater. Hydrobiologia, 41: 199-219. EISENBERG, R. M., 1966, The regulation of density in a natural population of the pond snail Lymnaea elodes. Ecology, 47: 889-906. EVERSOLE, A. G., 1974, Fecundity in the snail Helisoma trivolvis: experimental, bioenergetic and field studies. Ph.D. thesis, Syracuse Uni- versity, Dissertation Abstracts, 35: 5716-5717-B, Order no. 75-10, 538, 150 p. FIORONI, P. & SCHMEKEL, L., 1975, Develop- ment and habitat dependence in gastropods— an ontogenetic comparison. Forma et Functio, 8: 209-252. FRETTER, V. & GRAHAM, A., 1964, Reproduc- tion. In: WILBUR, K. M. & YONGE, C. M., Eds., Physiology of Mollusca, vol. 1. Academic Press, London and New York. FROMMING, E., 1956, Biologie der mitteleuro- paischen Süsswasserschnecken. Duncker & Humblot, Berlin. GADGIL, M. & BOSSERT, W. H., 1970, Life historical consequences of natural selection. American Naturalist, 104: 1-24. GELDIAY, R., 1956, Studies on local popula- tions of the freshwater limpet Ancylus fluvi- atilis Müller. Journal of Animal Ecology, 25: 389-402. GILLESPIE, D. M., 1969, Population studies of four species of molluscs in the Madison River, Yellowstone National Park. Limnology and Oceanography, 14: 101-114. GIROD, A., 1969, Ecologia dei fontanili lom- bardi. Malacofauna di alcuni fontanili a po- nente di Milano. Bolletino di Pesca Piscicol- tura e Idrobiologia, 24: 185-235. GRAHAME, J., 1973, Breeding energetics of Littorina littorea (L.) (Gastropoda: Proso- branchiata). Journal of Animal Ecology, 42: 391-403. НААТОС, С. DEN & DE WOLF, L., 1962, The life cycle of the water snail Aplexa hypnorum. Basteria, 26: 61-72. HEATH, D. J., 1977, Simultaneous hermaphrodi- tism: cost and benefit. Journal of Theoretical Biology, 64: 363-373. HEPPLESTON, P. B., 1972, Life history and population fluctuations of Lymnaea truncatula (Muller), the snail vector of fascioliasis. Jour- nal of Applied Ecology, 9: 235-248, 2 pl. HUBENDICK, B., 1950, The effectiveness of Passive dispersal, in Hydrobia jenkinsi. Zoo- logiska Bidrag fran Uppsala, 28: 493-504. HUNTER, R. D., 1975, Growth, fecundity and bioenergetics in three populations of Lymnaea palustris in upstate New York. Ecology, 56: 50-63. JOBIN, W. R. & MICHELSON, E. H., 1967, Mathematical simulation of an aquatic snail population. Bulletin of the World Health Or- ganisation, 37: 657-664. JOOSSE, J., BOER, M. H. & CORNELISSE, C. J., 1968, Gametogenesis and oviposition in Lymnaea stagnalis as influenced by gamma irradiation and hunger. Symposium of the Zoological Society of London, 22: 213-235. LEBRETON, J., 1971, Etude expérimentale de l'influence du jeune sur le cycle de la gonad male et du pénis chez Littorina littorea |. Haliotis, 1: 25-26. LEON, J. A., 1976, Life-histories as adaptive strategies. Journal of Theoretical Biology, 60: 301-336. LEVINS, R., 1968, Evolution in changing envi- ronments. Princeton University Press, Prince- ton, New Jersey. LILLY, M. M., 1953, The mode of life and the structure and functioning of the reproductive ducts of Bithynia tentaculata (L.). Proceedings of the Malacological Society of London, 30: 87-110. MARAZANOF, F., 1970, Contribution а l’etude écologique des mollusques des eaux douces et saumatres de Camargue. 2. Anisus rotundatus (Poiret, 1801). Annales de Limnologie, 6: 191-213. MATTICE, J. S., 1972, Production of a natural population of Bithynia tentaculata L. (Gastro- poda, Mollusca). Ekologia Polska, 20: 525-539. MAYNARD SMITH, J., 1971, What use is sex? Journal of Theoretical Biology, 30: 319-335. MCCRAW, B. M., 1961, Life history and growth of the snail, Lymnaea humilis. Transactions of the American Microscopical Society, 80: 16-27. MCCRAW, B. M., 1970, Aspects of growth of the snail Lymnaea palustris (Müller). Malacologia, 10: 399-413. MCMAHON, R. F., 1975, Effect of artificially elevated water temperatures on Physa virgata. Ecology, 56: 1167-1175. MCMAHON, R. F., 1976, Growth, reproduction and life-cycle in six Texan populations of two species of freshwater limpets. American Mid- land Naturalist, 95: 174-185. MCMAHON, R. F., HUNTER, R. D. & RUSSELL- HUNTER, W. О., 1974, Variation in aufwuchs at six freshwater habitats in terms of carbon biomass and of carbon : nitrogen ratio. Hydro- biologia, 45: 391-404. MEDAWAR, P. B., 1952, An unsolved problem in biology. Lewis, London. MEDCOF, J. C., 1940, On the life cycle and other aspects of the snail, Campeloma, in the Speed River. Canadian Journal of Research, 18D: 165-172. MIROSHNITSHENKO, A. Z., 1958, Fecundity of the fresh-water mollusc Viviparus viviparus L. Zoologicheskii Zhurnal, 37: 1635-1644. MORRISON, J. P. E., 1932, Studies on the life history of Acella haldemani ("‘Desh.” Binney). Transactions of the Wisconsin Academy of Science Arts and Letters, 27: 397-413, pl. 11-12. MORTON, J. E., 1964, Molluscs. Ed. 3. Hutchin- son University Library, London. NEUHAUS, W., 1949, Hungerversuche zur Frage der parasitären Kastration bei Bithynia ten- taculata. Zeitschrift fiir Parasitenkunde, 14B: 300-319. NICKERSON, R. P., 1972, A survey of enzyme and shell variation in 16 populations of the stream limpet Ferrissia rivularis (Say). Ph.D. dissertation, Syracuse University, Dissertation Abstracts, 33: 41588B, Order no. 73-7753. OLIVIER, L. & BARBOSA, F. S., 1955a, Seasonal studies on Australorbis glabratus Say from two localities in Eastern Pernambuco, Brazil. Publicoes avulsas Instituto Aggeu Magalhaes, 4: 79-103. OLIVIER, L. & BARBOSA, F. S., 1955b, Sea- sonal studies on Tropicorbis centimetralis in 364 CALOW Northeastern Brazil. Publicoes avulsas Instituto Aggeu Magalhaes, 4: 105-115. PIANKA, Е. R., 1970, On “r” and *“K” selection. American Naturalist, 104: 592-597. PINEL-ALLOUL, B. € MAGNIN, E., 1971, Cycle vital et croissance de Bithynia tentaculata L. (Mollusca, Gastropoda, Prosobranchia) du Lac Saint-Louis pres de Montréal. Canadian Jour- nal of Zoology, 49: 759-766. PINEL-ALLOUL, B. 8 MAGNIN, E., 1973, Observations sur le cycle vital et la croissance d'Amnicola limosa (Say) (Mollusca, Gastro- poda, Prosobranchia) du Lac Saint-Louis pres de Montréal. Canadian Journal of Zoology, 51: 311-313. PRINGLE, С. & RAYBOULD, J. N., 1965, The experimental study of water snails in a fish pond in Tanganyika. Il. Attempts to establish reproducible conditions. East African Medical Journal, 42: 289-295. RAPPORT, D. J. & TURNER, J. E., 1977, Economic models in ecology. Science, 195: 367-373. ROBSON, G. C., 1923, Parthenogenesis in the mollusc Paludestrina jenkinsi. British Journal of Experimental Biology, 1: 65-78. RUSSELL-HUNTER, W. D., 1953, On the growth of the fresh-water limpet, Ancy/us fluviatilis Müller. Proceedings of the Zoological Society of London, 123: 623-636. RUSSELL-HUNTER, W. D. 1961a, Annual varia- tions in growth and density in natural popula- tions of freshwater snails in the west of Scotland. Proceedings of the Zoological So- ciety of London, 136: 219-253. RUSSELL-HUNTER, W. D., 1961b, Life cycles of four freshwater snails in limited popula- tions in Loch Lomond, with a discussion of intraspecific variation. Proceedings of the Zoological Society of London, 137: 135-171. RUSSELL-HUNTER, W. D., 1964, Physiological aspects of ecology in nonmarine molluscs. In: WILBUR, К. М. € YONGE, С. M. Eads., Physiology of Mollusca, Vol. 1. Academic Press, New York and London, p. 83-126. RUSSELL-HUNTER, W. D., in press, Ecology of freshwater pulmonates. In: FRETTER, V., Ed., Pulmonates, vol. 2. Academic Press, New York and London. RUSSELL-HUNTER, W. D. & APLEY, M. L., 1966, Quantitative aspects of early life-history in the salt-marsh pulmonate snail, Melampus bidentatus, and their evolutionary significance. Biological Bulletin, 131: 392-393. RUSSELL-HUNTER, W. D., APLEY, М. L. & HUNTER, R. D., 1972, Early life-history of Melampus and the significance of semilunar synchrony. Biological Bulletin, 143: 623-656. RUSSELL-HUNTER, W. О. & MCMAHON, В. F., 1976, Evidence for functional protandry in a fresh-water basommatophoran limpet, Laeva- pex fuscus. Transactions of the American Microscopical Society, 95: 174-182. RUSSELL-HUNTER, W. D. & WARWICK, T., 1957, Records of Potamopyrgus jenkinsi (Smith) in Scottish fresh waters over fifty years (1905-56). Proceedings of the Royal Society of Edinburgh, ser. B, 66: 360-373. SCHAFFER, H., 1953, Untersuchungen zur Ökologie von Bithynia tentaculata. Archiv für Molluskenkunde, 82: 67-70. SMITH, C. C. & FRETWELL, S. D., 1974, The optimal balance between size and number of offspring. American Naturalist, 108: 499-506. STANCZYKOWSKA, A., 1959, Distribution and population dynamics of Viviparus fasciatus Müller on the River Konferatka. Ekologia Polska 5: 271-273. STANCZYKOWSKA, A., MAGNIN, E. & DU- MOUCHEL, A., 1971, Etude de trois popula- tions de Viviparus malleatus (Reeve) (Gastro- poda, Prosobranchia) de la region de Montreal. 1. Croissance, fécondité, biomasse and produc- tion annuelle. Canadian Journal of Zoology, 49: 1431-1441. STEARNS, S. C., 1976, Life-history tactics: a review of the ideas. Quarterly Review of Biology, 51: 3-47. STURROCK, R. F., 1973, Field studies on the population dynamics of Biomphalaria glabrata, intermediate host of Schistosoma mansoni on the West Indian island of St. Lucia. /nterna- tional Journal of Parasitology, 3: 165-174. TINKLE, D. W., 1969, The concept of reproduc- tive effort and its relation to the evolution of life histories in lizards. American Naturalist, 103: 501-506. TINKLE, D. W. & HADLEY, М. F., 1975, Lizard reproductive effort: caloric estimates and comments on its evolution. Ecology, 56: 427-434. TOMLINSON, J.. 1966, The advantages of her- maphrodism and parthenogenesis. Journal of Theoretical Biology, 11: 54-58. VAN CLEAVE, H. J., 1935, The seasonal life history of an amphibious snail Fossaria modi- cella, living on sandstone cliffs. Ecology, 16: 101-108. VAN CLEAVE, H. J. & ALTRINGER, D. A., 1937, Studies on the life cycle of Campeloma rufum, a freshwater snail. American Naturalist, 71: 167-184. VAN CLEAVE, H. J. & LEDERER, L. G., 1932, Studies on the life cycle of the snail Viviparus contectoides. Journal of Morphology, 53: 499-522. VLASBLOM, A. G., 1971, Further investigations into the life cycle and soil dependence of the water snail Aplexa hypnorum. Basteria, 35: 95-108. WALTON, С. L. & JONES, W. N., 1926, Further observations on the life-history of Limnaea truncatula. Parasitology, 18: 144-147. WEISMANN, A., 1882, The duration of life. In: Essays upon heredity and kindred biological problems. Oxford University Press, Oxford, England. WILBUR, H. M., 1977, Propagule size, number and dispersion pattern in Ambystoma and Asclepias. American Naturalist, 111: 43-68. WILLIAMS, G. C., 1957, Pleiotropy, natural selection and the evolution of senescence. Evolution, 11: 398-411. WILLIAMS, G. C., 1966a, Adaptation and natural selection. Princeton University Press, Prince- ton, New Jersey. WILLIAMS, G. C., 1966b, Natural selection, the costs of reproduction and a refinement of Lack’s principle. American Naturalist, 100: 687-690. WIT, W. F. DE, 1955, The life cycle and some other biological details of the fresh-water snail Physa fontinalis (L.) Basteria, 19: 35-73. YOUNG, M. R., 1975, The life cycles of six species of freshwater molluscs in the Worces- ter-Birmingham canal. Proceedings of the Malacological Society of London, 41: 533-548. MALACOLOGIA, 1978, 17(2): 365-391 PROTANDRY AND THE EVOLUTION OF ENVIRONMENTALLY-MEDIATED SEX CHANGE: A STUDY OF THE MOLLUSCA K. Elaine Hoagland Department of Biology, Lehigh University, Bethlehem, Pennsylvania 18015, U.S.A. and Department of Malacology, Academy of Natural Sciences, Philadelphia, Pennsylvania 19103, U.S.A. ABSTRACT Many molluscs in several taxonomic groups (Mesogastropoda, Bivalvia) and several ecological settings (parasites, filter-feeders, wood-borers) are protandrous hermaphrodites, with varying degrees of environmental determination of sex. | delineate the selective forces that have led to the establishment of protandry and labile sex determination in some, but not all, molluscs. | first report on a series of experiments on the sexual behavior and sex determination of four species in the mesogastropod genus Crepidula, and correlate differences | find with differences in the reproductive patterns of these species. | review the literature оп protandry and labile sex determination in molluscs and some other invertebrates in order to test the generality of the conclusions generated from the study of Crepidula. Species of Crepidula with planktonic larval development and that are rarely substrate- limited exhibit labile sex determination and have socially influenced sex ratios. Their gregarious behavior and female-induced delay of sex change appear to be mediated by pheromones, possibly through a common mechanism. Species of Crepidula lacking planktonic larvae also lack gregarious behavior, and sex change and sex ratio are independent of influence by other members of the species. Therefore, species patterns in control of sex change appear to be correlated with the mode of larval development and dispersal, and with substrate constraints. Protandry is advantageous—that is, it increases individual fitness—when one sex increases in fertility with age and size faster than the other. This is true in many molluscs where female fecundity is related to large body size, but male fecundity is related to mobility and therefore often to small size. Labile sex change that is influenced by the environment optimizes the size and age at sex change in protandrous species, thus maximizing individual fitness. This is important in species for which the optimal age at sex change varies from place to place and from generation to generation. Sedentary species with planktonic larvae are in this category. Gregariousness, in addition to protandry and labile sex change, ensures that each individual will reproduce. An isolated individual becomes female immediately upon metamorphosis and attracts spat, which become male, providing a mate. The advantage to colonizing populations of sedentary species is that every encounter between individuals is potentially productive of offspring. Molluscs that are largely sedentary as adults but possess dispersal stages and have substrates patchy in time and space have evolved protandry: the Calyptraeidae, oysters, wood-boring bivalves, and the parasitic mesogastropods are outstand- ing examples. INTRODUCTION Malacologists have long recognized the large number of molluscan species that possess some form of hermaphroditism. Several of these hermaphroditic molluscs are sedentary and reproduce via copulation. The pattern of sexuality in a copulating, sedentary organism is interesting to stu- dents of evolution because it affects the number of encounters resulting in off- spring, hence the chances that an individual will reproduce successfully. The list of molluscs that are proven protandrous her- maphrodites and those that show circum- stantial evidence of being protandrous is actively growing, as more and more species are studied biologically. Likewise, environ- mentally-mediated sex determination (often called labile sex determination) has been demonstrated in a number of molluscs, as well as other invertebrates and numerous (365) 366 HOAGLAND plants (Bacci, 1965; Charnov € Bull, 1977). Many of those with sexual lability and protandry are sedentary as adults. Environmentally-mediated sex determi- nation in the molluscan family Calyp- traeidae is not a new discovery. Gould (1917a, 1917b), Ishiki (1936), and Coe (1944) reported that members of the genus Crepidula within the Calyptraeidae are influenced physiologically and behaviorally by others of the same species. In C. fornicata and C. plana, the protandrous males are not only attracted to females, but attain the male phase earlier and remain as males longer when in contact with females (Coe, 1953). Protandry and labile sex determination occur in groups of molluscs that are widely separated phylogenetically. These character- istics are often but not always found in the same organism. Thus the interesting ques- tion is not so much how these modes of sexuality arose, but rather, why. What are the selective forces that lead to protandry and labile sex determination in certain molluscs (and other organisms)? | at- tempted to answer this question by using, as a model, species of Crepidula, which are all protandrous, low in mobility, and which copulate, but show varying degrees of sexual lability. | compared reproductive characteristics, sexual characteristics, group behavior, and general ecological data. | am able to generalize that substrate require- ments plus the type of larval development and dispersal, when superimposed on limited adult mobility, form predictable associations of reproductive and sexual traits in molluscs and perhaps other inverte- brates as well. The type of larval development and dispersal varies within Crepidula. Some species disperse as young adults after metamorphosis within a brooded egg sac, while others release planktonic larvae and are nearly sedentary as adults. The starting point for the research described in this report was my hypothesis that, comple- menting these different reproductive and dispersal mechanisms, there might have evolved differences in mechanisms of sex change and sexual behavior involved in getting the sexes together. Casual observations of four species of Crepidula in the laboratory did reveal striking species-level differences in sexual behavior, including gregariousness, degree of mobility of young males, permanence of mating pairs, and timing of sex change when isolated versus when maintained un- der crowded conditions. | quantified these differences through a series of experiments in the laboratory and in the field, and correlated them with other aspects of the life histories of the species. Finally, | reviewed the literature on protandrous Mollusca and other inverte- brates so that generalizations could be made on the evolutionary significance of protandry and sexual lability. | have con- cluded this report with a series of unan- swered questions which | hope will stimu- late further research and testing of the hypotheses developed in this paper. MATERIALS AND METHODS Choice of study organisms and localities Four species of Crepidula that vary in reproductive pattern and population struc- ture were examined. The first two species, C. fornicata (Linn.) and C. onyx Sow., have planktotrophic larval development pre- ceded by a period of brooded development. They naturally form large clusters or stacks of individuals in muddy bays where hard substrate would otherwise be difficult to find. A third species, C. plana Say, is sympatric with C. fornicata and has similar larval development, but lives inside empty shells or under stones and never forms stacks. However, several males may live attached to one female. The females of these three species are sedentary, and may remain for months in the same place, although they are capable of limited move- ment when placed in an unfavorable (e.g., oxygen-poor) environment. Young juveniles and males have limited mobility on hard substrates. Also sympatric with C. fornicata is the fourth species examined, C. convexa Say. It broods its young through metamorphosis and forms temporary male-female pairs, but does not form clusters. It is the smallest and most motile of the four species, but during brooding, female movement is virtu- ally nil. The study sites are described in Appen- dix A. Experimental methods Several laboratory and field experiments were conducted to answer specific ques- PROTANDRY AND LABILE SEX DETERMINATION 367 tions generated by the general hypothesis that sex determination and sexual behavior should differ in species with different modes of larval development and dispersal. These experiments are described here, to- gether with the rationale for performing them. Gregariousness and Sexual Behavior To quantify possible species differences in gregariousness and sexual behavior, | ran a series of laboratory experiments. Mem- bers of the three Atlantic species of Crepidula were obtained from Woods Hole, Massachusetts. C. onyx was collected in Balboa, California, and transported live to Woods Hole. Individuals of each species were sorted into four sexual categories: juvenile (J), no secondary sex characters present; male (М), presence of a fully developed penis; intermediate (1), presence of a degenerating penis, and female (F), absence of a penis or presence of a stump only, and presence of an oviduct. The reliability of each of these categories was confirmed by examination of the gonads and other internal reproductive organs of animals that were sacrificed. Snails in each sex category were placed in contact with those in every other category, in pairwise tests. Pairs were isolated in finger bowls with running sea water and observed for signs of gregariousness and sexual interac- tion, such as physical contact, over several intervals (24, 48 hours; 7 days; 1 month). | report the results after 48 hours. Longer intervals included sex changes; the shortest interval was insufficient for stabilized inter- action. The sample size was 20 individuals of each category, hence 40 per test. Field experiments were designed to determine if the patterns of gregarious behavior seen in the laboratory were dupli- cated under more natural conditions. Six clay pots of equal size were suspended from ropes into 1.5m of water off docks at the Woods Hole, Massachusetts, Yacht Club. These artificial substrates rested about 20cm above the water-mud inter- face. Six snails, consisting of a male-female pair of each of the three local Crepidula species, were allowed to attach to each of three of the pots while still in the laboratory. This was to determine if young are attracted to adults of the same species, or merely to any suitable substrate. Newly metamorphosed juveniles settling on the substrates were counted and removed weekly over a four-month period, June through October, 1973, and the ratio of young settling on pots with versus without adults of the same species was calculated. One further field experiment provided data on attractiveness of adults to juveniles. At several beaches in Woods Hole, С. fornicata and C. convexa co-occur on discontinuous substrates (small rocks, cob- bles, broken glass, and shells amidst sandy mud) in shallow water. Within two of these areas On two occasions in the summer of 1973, the numbers of newly metamor- phosed juveniles (shell length less than 3mm) of each species per substrate unit were counted. A total of 50 substrate units were examined for C. fornicata, and 255 for C. convexa. The surface area of each substrate unit was calculated, thereby deriv- ing the density of young snails on each substrate. Finally, the densities of males, females, and juveniles greater than 3mm long were calculated for the same sub- strates, in order to determine the relative attractive ability of these sex categories for metamorphosing young. Sex change and population density Laboratory experiments were conducted to test the null hypothesis that species with different modes of reproduction and dis- persal have the same degree of environ- mental determination of sex. The experi- ments also tested whether environmental determination of sex can indirectly affect the sex ratio of a population by altering the age at which an individual changes sex, relative to population density. | maintained 14 bowls each of С. fornicata and C. convexa of various initial densities and sex ratios, from June, 1972, to June, 1973, under natural temperatures in independent flowing sea water systems. The food supply to all bowls was the same. No recruitment to these artificial popula- tions was permitted. The final densities and sex ratios were recorded. In the field, | recorded the total adult density and the percentage of males for four populations of C. fornicata and five of C. convexa in May and/or August, 1972 and/or 1973. Discontinuous substrates such as stones were picked at random and all adults on them were sexed until N = 50 was reached (N = 200 for one very dense 368 HOAGLAND population from Rhode Island). The data were analysed for correlations between population density and sex ratio, and the results were compared with those obtained via the laboratory test of density and sex ratio. Two of the same field populations of С fornicata and С. convexa were sampled monthly during winter months and twice a month during summer (April through Octo- ber) to determine seasonal changes т population density, sex ratio, copulation, and egg-laying activity. Size and age at sex change Preliminary observations suggested that the size and possibly the age at sex change vary among populations of C. fornicata. To test the null hypothesis that size and age at sex change are constant within species, | amassed data on the size, age, and sex of individuals from five populations of C. convexa, four of C. fornicata, and one each of С. plana and С. onyx. If the hypothesis was refuted, | planned to look for relation- ships between total adult density or female density and the timing of sex change. | also wished to identify the effect of direct contact with females on the timing of sex change. Such relationships imply popula- tion interaction in the regulation of major events in the lives of individuals, and if present, require an interpretation in the light of current evolutionary theory. Over a period of three years, all indi- viduals of the four species of Crepidula in the study areas found to be intermediate between male and female were measured for shell length. The dry weights of the animals were estimated from a standard curve of length versus dry weight (body and shell) which was constructed previously (Hoagland, 1975). Time of year affected the number of individuals that were chang- ing sex, but did not affect either the age or size at sex change, so all intermediates from each population were pooled. This was necessary because the sample size of intermediates at any one sampling time was small. A note was made for each intermedi- ate as to its position—whether it was sitting on a female. Also, the age of each intermediate was estimated by counting winter growth lines (Sheldon, 1967), which allow age to be estimated within a range of about six months. RESULTS Gregariousness and sexual behavior | report results of the laboratory tests of gregarious behavior in the four species of Crepidula in Table 1. Some snails re- sponded to the test situation by crawling onto the shells of other snails; the former are called “‘respondents."’ Those that did nothing, but allowed other individuals to climb onto their shells, are called ‘‘re- |еазегз.” The bottom member of a pair rarely moved from its initial position and is assumed to have attracted the individual which crawled upon it. Interactions are species-specific. C. for- nicata and C. onyx respond alike and have the highest degree of gregariousness. C. convexa is least gregarious. Responses are always by smaller, less mature individuals to sexually advanced and/or larger indi- viduals. Females move toward tank aera- tors, but are otherwise sedentary. Crepidula fornicata, С. onyx, and to a lesser extent, C. plana males and juveniles cluster among themselves and are strongly attracted to larger individuals, especially females. Long-term observations revealed that pairings between male and female are often stable for as long as one year. However, C. convexa juveniles are not gregarious. Males pair only with females, and a pairing usually lasts less than three weeks. The field experiments using artificial substrates to determine if newly metamor- phosed young are attracted to adults of the same species gave the following results. The ratio of young settling on the three pots with adults versus the three without, was: С. fornicata, 722:232, and С. plana, 249:95. Therefore, for species with plank- tonic larvae, the presence of adults signifi- cantly increases the number of young that metamorphose on the same substrate. Cre- pidula convexa, which has по planktonic larval stage, did not appear on the pots. It does occasionally appear on such suspended experimental materials, probably because it is capable of floating on algae or in the water surface film, as | have observed at field sites in Barnegat Bay, New Jersey. Calculations of the density of newly metamorphosed C. fornicata snails with respect to the presence of older individuals on natural substrates are presented in Table 2. The density of metamorphosing young PROTANDRY AND LABILE SEX DETERMINATION 369 TABLE 1. Experiments on gregariousness. A summary of data after 48 hours. Response: % pairings* “Releaser”” Respondent fornicata опух plana сопуеха fornicata-onyx* E J 100 100 30 10 0 M 100 100 100 65++ 0 | 0 0 0 0 0 F 0 0 0 0 0 | J 60 50 25 0 0 М 75 70 80 201+ 0 | 0 0 0 0 0 Е 5 0 0 0 0 M J 70 60 25 10++ 0 M 100 ei TO Ба 0 | 0 0 0 0 0 Е 0 0 0 0 0 J J 65+++ 45 30.20: 0 0 М 0 0 0 0 0 | 0 0 0 0 0 F 0 0 0 0 0 *Percentage of the twenty individuals of a given sex which formed pairs by crawling on the back of one of the twenty members of the sex being tested as a releaser. #C. onyx as releaser and C. fornicata as respondent. ++Pairings are often transient; switching of mates was observed within 48 hours. +++ aggregations of contiguous indiciduals occur as well as pairings, and are included as positive responses. F = female M = male | = intermediate J = juvenile TABLE 2. Density of newly metamorphosed Crepidula fornicata young per cm? of substrate with respect to the presence or absence of older individuals. a дд No other Crepidula Females Males Mixed adults & juveniles Juveniles n° 10 10 10 10 10 x .18 15 ЗИ .66 37 Si .03 .20 .04 .25 .06 ANOVA TABLE, LOG-TRANSFORMED DATA Source of variation df ss ms F Among groups 4 .170 .042 5.68 Within groups 45 .336 .007 F(.05) = 2.61 for 4, 40 df 2.53 for 4, 60 df Significant individual comparisons (Ability to attract spat): 1. No Crepidula < mixed, male, and female 2. No Crepidula, juvenile and male < female *number of cobbles and rocks examined was not related to the density of adults, The attraction power may be _ ranked: but rather, to their presence or absence, probably because the response of larvae to adults is a threshold rather than a graded response. Because the variances were not homo- geneous (in fact, were roughly proportional to the means), | used a log transformation before conducting an analysis of variance. females > mixed juveniles and adults > males = juveniles > bare substrate, although the difference between females and mixed juveniles plus adults is not significant (p > .5). These data support the laboratory findings that C. fornicata juveniles are gregarious. Table 3 presents results for C. convexa 370 HOAGLAND TABLE 3. Density of newly hatched C. convexa per Littorina shell, with respect to the presence of older individuals on the same Littorina. Females Males Mixed adults & juveniles Juveniles n° 128 18 27 82 x 23 .78 1.04 1.22 са .21 .42 .50 os! Test for equality of variances: M = .19; 3 df; .975 < p < .99 ANOVA TABLE Source of variation df ss ms F Among groups 3 54.30 18.10 50.28 Within groups 250 89.57 .36 F(.05) = 2.60 for 3, © df Significant individual (Ability to attract young): 1. Female < male, mixed, and juvenile 2. Male < juvenile comparisons *Number of Littorina examined. All Littorina were approximately the same size. TABLE 4. Laboratory data on sex ratio after one year, varying the density and initial sex composition. C. fornicata* C. convexa* Initial composition Final composition Percent male M F 5 5 50 4 5 5 5 50 4 4 10 0 100 4 5 10 0 100 4 4 10 5 67 6 6 10 5 67 6 7 20 20 50 16 20 20 20 50 152220 50 0 100 28 18 50 0 100 24 20 50 25 67 327221 50 25 67 34 28 50 50 50 43 41 50 50 50 40 36 *substrate area: 458 cm? substrate area: 135 cm? M = male F = female on natural substrates. The attraction power for young less than 2 mm long is ranked: juveniles > mixed adults and juveniles > males > females. Females are significantly less attractive than each of the other three categories (in individual t tests, р < .05). Variation in hatching size, and the fact that пемЛу hatched young must begin life on the same substrate as the mother, pre- vented the scoring of density for newly hatched young on bare substrates. All the densities are low, especially that of juven- iles living with females, considering that the young are directly released by the females. Final composition Percent male M Percent male 44 1 6 14 50 0 6 0 44 0 6 0 50 0 6 0 50 111 6 46 220 20 44 34729 9 43 2732 6 61 2 38 5 55 О 32 0 60 4 48 8 55 6 51 11 51 218 3 53 4 71 5 The results are nearly the reverse of those for C. fornicata. Furthermore, out of the 255 Littorina shells observed with С. convexa snails attached, 146 carried adults only, 82 carried juveniles only, and a scant 27 carried both adults and juveniles. Attractive power of females for young appears to be nonexistent in C. convexa. Sex ratio and population density Table 4 presents data on the sex ratio of several artificial populations of C. fornicata PROTANDRY AND LABILE SEX DETERMINATION 371 and C. convexa maintained for one year in the laboratory. Regardless of initial condi- tions, most C. convexa snails eventually become females. There are differences among individuals in the tendency toward femaleness that possibly are genetic. Crepidula fornicata approaches and main- tains roughly a 50-50 sex ratio regardless of the initial ratios among the initial ratios test- ed. In dense populations there is a slightly higher percentage of males. A regression of sex ratio (percent male) after one year on density as the independent variable gives vw .28; the Spearman rank correlation г. = .61, {= 2.66, n= 14, and .02 df ’ Gunning Point Solitary 6.4 9.2 23 0.020 0.00002 23 Mated Solitary 12 12 0.014 0.012 25 10 64 80 (NS) *Dry weight data, log-transformed to normalize the data and to equalize variances. Age data are not analysed because the means are obviously uniform in the 3 populations. 375 РВОТАМОВУ AND LABILE SEX DETERMINATION 091 [7 v9 001 [901 SLI 61 62 $1910 12290 LGI6O LZILC 0 0 Але} 110$ pa}ew eoqieg xAuo 9D ÿLLT'O 9SLO'0 AAA 1074 gee Але} 110$ 005 07 OL EL vs 07 001 09 001 GC 001 0906 0 6v£9'0 £OC8'L t£0£ O 0786 L cl cl 81 9 81 ce 69 81 8 С = 1070`0 £vSL'O zOv0'0 “809€ TARA) vzSe 0 91280 8197'0 velec ve 69 81 8 с 00'9 06° 17`6 VETE GOV 00'8 cae 55'9 09V 09'51 рэазеш Але} 110$ рэзеш Але3 1105 poeyew (ZZ6L) 1 U (ZL61) IGN Jue yen e1e2/U10) “Y 007 007 9 ся GC 05 GL 00L Ofve'l LLL [4:33 20) COLL cl Le 9 LE [4 С v 59 86110 9/60`0 $695'0 BLILE CELC L EGLOC 8zcL'L 9960'p С [4 v 09 0081 00 CL 00 CL 66 6t 00 9 00'SL 00'9 08°02 Але} 110$ pajeu ÁJey1¡Oos parew yonben pseAaui/ $,ечзлей\| Ayısuap yjnpe AW (ajew %) оцел xes Aeyy N (1ybiam A1p *6) ajew 1596127 ajew }S2PIO N ээиечел ("6) 1ybiam Aup ueay\ N ээцелел (ou) abe ueayy "saıaads 914033 ue¡d :эбиецэ хэ$ зе 3215 pue aby '9319V_L 376 HOAGLAND 100 90 80 70 60 50 40 30 20 10 Percentage of Females Brooding Eggs Jan | Feb 'Mar 'Арг Months FIG. 5. The percentage of females brooding eggs, Crepidula convexa Woods Hole Yacht Club population, 1972 and 1973. individuals are invariably greater, showing that mated males delay sex change, and that sex change reflects interaction among individuals. Coe (1938a) and Gould (1917b and 1952) obtained similar results. A complete comparison with С. convexa is not possible because no mated intermedi- ates were found in that species. All males leave their mates prior to or at the beginning of sex change. The results concerning age, despite the low precision of the data, are interesting because they demonstrate that the differ- ences in size at sex change in C. fornicata are not due solely to differences in growth rates between populations while age at sex change remains the same. If growth rates alone were implicated, a trivial explanation such as differential food availability would be possible. The timing of sex change cannot be explained simply by the presence or absence of energy reserves in an organ- ism, because the explanation does not account for sex change being affected by interaction among individuals of a popula- tion. The relationship between population density and size at sex change is given in Figs. 6-9. For C. convexa, size at sex change is inversely related to population density (г? =.92) and to female density (г? = .94), but in both cases the regression lines have insignificant slope (Appendix B). Dense, crowded populations such as those living on eel grass (Zostera) mature at small size (are stunted), hence the inverse correla- tion. In С. fornicata, the general trend is for size at sex change to increase with popula- tion density. An exception is the colonizing Rhode Island population. My data show that such young, newly colonizing popula- tions have, on the average, earlier sex change than do populations in which older PROTANDRY AND LABILE SEX DETERMINATION 377 5.0 © 4.0 ® с = oO is O SO feb) n o pa oN Ne 2 v 2.0 = > ®) € a во =e || oP MBLB ‘72 O O 80 160 Adult Population Density (number рег т ° MV ow EB 320 E) 240 FIG. 6. Size at sex change as a function of adult population density, all species, mated individuals. The adult population density is the May value in each case. Abbreviations of localities are as in Appendix A. There were no mated intermediates found for Crepidula plana ог С. convexa. Key: С. fornicata O, С. onyx 0 age classes dominate, and in which there are more age classes (e.g., the Nahant population). Such populations as Nahant are more stable, and probably have lower adult mortality (especially density- independent mortality) than newly coloniz- ing populations. Warner (1975) showed a similar trend for protandrous Pandalus shrimp; later transformation occurred in populations with a wider range of ages in a population. Charnov (personal communica- tion) states that recruitment of an unusu- ally strong year class to a Pandalus jordani population results in a larger than usual percentage of young changing to female without functioning as male. Crepidula onyx and С. plana are in- cluded in Figs. 6-9. They fit well into the density versus size curves, reducing r? only slightly. The values of г? for С. fornicata 378 HOAGLAND Mean Dry Weight at Sex Change (G.) O 80 160 240 See Adult Population Density (number per m2) FIG. 7. Size at sex change as a function of adult population density, all species, solitary individuals. The adult population density is the May value in each case. Abbreviations of localities are as in Appendix A. Key: С. fornicata e, С. onyx а, С. plana O, С. convexa 0 alone improve in the following progression: total density versus size of mated inter- mediates (poorly correlated) < female den- sity versus size of solitary intermediates < female density versus size of mated inter- mediates < total density versus size of solitary intermediates. DISCUSSION Summary of differences among species of Crepidula The sexual behavior patterns of C. convexa and C. fornicata seem to be representative of patterns in non-planktonic and planktonic developers, respectively, of the family Calyptraeidae. In C. convexa, which lacks a planktonic dispersal stage, there is no attraction among newly hatched juveniles, nor between adults and juveniles. Coe (1953) studied the sexuality of several species that correspond to C. convexa in mode of larval development and dispersal, including С. norrisiarum Williamson, С. adunca Sow., and C. williamsi Coe (= C. striolata Menke, Hoagland, 1977b). Like C. convexa, matings were temporary, males left females at the time of their sex change thereby precluding stack formation, and protandrous sexual development was not environmentally regulated. С. williamsi possesses a two-week period during which newly metamorphosed young are not at- tracted by females. C. adunca young are PROTANDRY AND LABILE SEX DETERMINATION 379 Mean Dry Weight at Sex Change (G.) O 25 50 То) Female Population Density (number рег т?) 100 FIG. 8. Size at sex change as a function of female population density, all species, mated individuals. Key: С. fornicata O, С. onyx © not attracted to older individuals (Putnam, 1964). C. dilatata Lamarck and an as yet unnamed sibling species from South Ameri- ca possess different modes of larval de- velopment, and Gallardo (1977) indicated that the species with indirect development forms permanent stacks, but the species with direct development has temporary matings. At least one Ca/yptraea, C. chinen- sis, is protandrous and does not have labile sex determination. It has temporary mat- ings, does not form stacks, and is not gregarious. It also broods its young (Wyatt, 1960). Lack of attraction of juveniles to adults probably aids in juvenile dispersal, and results in increased outbreeding. Outbreed- ing is further increased by the temporary nature of male-female pairings. Females probably do attract males, but the response is limited to functioning males, for short durations. Timing of sex change is geneti- cally programmed rather than influenced by other members of the population. The result is a sex ratio that is determined by the age structure and genetic composition of a population, rather than by group interaction. A field study (Hoagland, 1975) revealed that, indeed, a C. convexa popula- tion possessed a stable age structure due to constant recruitment over the three-year study period, whereas C. fornicata had variable recruitment and an unstable age structure. Because C. convexa young radi- 380 HOAGLAND 2.0 © Ф LS с с o Е O x © n as AC = © ® = > Q BE wr = e MBL ‘73 = *MBL ‘72 oWHYC eW eMV е В O 25 Female Population Density (number per 50 TD 100 m2) FIG. 9. Size at sex change as a function of female population density, all species, solitary individuals. Key: С. fornicata e, С. onyx =, С. plana O, С. convexa © ate from the parents without the uncertain- ties of planktonic dispersal, their mode of sexual change and interaction is appro- priate. Under crowded conditions, C. convexa snails mature at a smaller-than-usual size, while timing of sex change is not affected. One can only speculate as to the mechan- ism, but the most obvious possibility is slower growth due to competition for food, or due to physical contact among the individuals. The behavior and sexuality of C. for- nicata are highly integrated among indi- viduals of a population. Individuals influ- ence the settling behavior and maturation rates of other individuals. Juveniles are attracted to adults, as is consistent with the planktonic mode of dispersal and the inherent danger of an individual settling and remaining in isolation. Spat which do settle alone rapidly become female and proceed to attract other C. fornicata snails. One could assume that outbreeding is accomplished via the planktonic stage. There are no data on degree of outbreeding in Crepidula. The age and size at sex change are factors that influence the demography of a protandrous population, for they deter- mine the female- and male-stage fecundities of the individuals. Size and age at sex change are variable among the C. fornicata populations studied. In non-colonizing pop- PROTANDRY AND LABILE SEX DETERMINATION 381 ulations, both size and age tend to increase with population density. Data for coloniz- ing populations provide another insight: there are few females already present in a very dense, new settlement of C. fornicata. Sex change occurs at a very small size and young age regardless of the high density. On the other hand, an established popula- tion with only moderate density but a high proportion of females has a greater size at sex change than could be predicted by total density alone. Hence the density of females is implicated in the timing of sex change in С. fornicata. In fact, in the case of mated males, female density explains more variability in size at sex change than does total population density (Appendix B). Crepidula fornicata responds oppositely from C. convexa, in which high density and the resultant crowding bring about a de- creased size at sex change. Even very dense populations of C. fornicata do not run out of substrate. C. fornicata snails may form stacks in an arc, as illustrated in Fig. 10. This capability insures the presence of substrate in this gregarious species. | have observed several males in one stack copulat- ing with the females of the same stack. The length of the penis increases with distance from the females, allowing a small male on the top of a stack to reach a female despite the presence of 3 or 4 males in between. This has the same effect as male mobility FIG. 10. A stack of Crepidula fornicata in life position. Sexes of the individuals are indicated. | = intermediate. in C. convexa. Small males of C. fornicata also have some mobility. In both species, polyandry is possible. There well could be sperm competition (Parker, 1970). Whether females accept sperm from more than one male is unknown, but they can store sperm for at least one year (Hoagland, 1975). The facts that the male phase is prolonged in C. fornicata within the range of a female and that pairs may stay together for over a year imply that a female mates more than once. Otherwise, no advantage (increase in fit- ness) would accrue to the male by staying with a female, and no advantage would accrue to the female by prolonging the male phase of young C. fornicata. One or the other sex (not necessarily both) must be advantaged by the prolonged pairing. The data available for C. onyx indicate that it is similar to C. fornicata in essential features of labile sexuality and gregarious- ness, as well as in reproductive and dis- persal modes, and the ability to form stacks. Stacked populations of C. onyx and C. fornicata are most common in muddy bays, where recruitment and population density are high. The population of C. fornicata at Nahant, Massachusetts, did not have stacks greater than three individuals, but animals taken from the area behaved as did other C. fornicata in the laboratory. Organisms expressing labile sex determi- nation are often phenotypically plastic in other ways. For example, C. fornicata and C. onyx vary their shape and intensity of pigmentation tremendously, depending on environmental conditions. The same factor that leads to sexual lability, the unpredict- ability of where an organism will settle and what selective regime it will occupy com- pared to that of its parents, also leads to physical and physiological plasticity. The only difference is that, in the case of sexual labiliy, organisms of the same spe- cies are the critical element of the environ- ment. Phenotypic plasticity itself is a character that is affected by natural selec- tion. Crepidula convexa and other species without planktonic young and labile sex determination also lack much phenotypic plasticity in color and shape. Rather, they possess discrete color morphs, genetically determined, in some populations (Hoag- land, 1977a). These species are thus charac- terized by low interaction between geno- type and environment to produce the phenotype. Another way of saying the 382 HOAGLAND same thing is that heritability of traits is high. A type of sexual behavior and popula- tion structure that is intermediate between the C. fornicata type and the C. convexa type is illustrated by С. plana. It produces dwarfed females like C. convexa, and is substrate-limited, for it lives underneath objects or inside them, and does not form stacks. With respect to gregariousness, it is intermediate between C. fornicata and C. convexa. It is quite possible that С. plana veligers, being strongly photonegative, settle together more because they are attracted to the same substrate than because they are truly gregarious. Coe (1938b, 1948) found that C. plana is also intermediate and highly variable in the time mated pairs stay together when compared with C. fornicata (long) and C. convexa (short). In reproduc- tion and dispersal, С. plana is like С. fornicata. \t possesses labile sex determina- tion (Gould, 1952). Coe (1953) demonstra- ted that the presence of a female delays sex change in males of С. plana and С. nivea C. B. Adams, ecological counterparts in the North American Atlantic and Pacific, respectively. About 90% of the males kept together with females in the laboratory did not change sex, whereas only about 30% of those without females remained male. In nature, 68% of the C. plana and 42% of the C. nivea remained male over the same length of time. Gould & Hsiao (1948) demonstrated individual differences т growth rates and sexual development in C. plana. They presumed these to be congeni- tal. My own experiments indicate that C. fornicata siblings also grow and mature at different rates. It is important that С. р/апа is interme- diate in both life history characteristics and in sexual behavior. It substantiates my hypothesis that the mode of reproduction, type of dispersal, and degree of substrate limitation interact to determine the optimal sexual behavior and population structure, in particular the sex ratio and recruitment patterns. Coe (1935, 1938a and b) claimed that some individuals of every Crepidula species are “true” males, incapable of sex change, but evidence is weak. In C. convexa, no evidence of “true” males or females, such as juveniles that fail to go through a male phase in the laboratory, has been found. However, C. fornicata individuals that ap- peared not to function as males have been seen in the course of this study. Of the 27% of the individuals in the colonizing Rhode Island population that were re- ported as females in 1972, most were equal in size to the males of the population. It appeared that these females were less than a year old, and may have passed through the male phase without functioning. From laboratory data and weekly examination of C. fornicata juveniles at the M.B.L. Beach locality, also representing colonization, it appeared that some individuals do bypass the functional male phase. The greater component of this capability may be environmental rather than genetic, because an individual selected at will and isolated often fails to develop a complete penis before it begins to resorb. Warner et al. (1975) found cases in fish where there is male dimorphism. Some members of a population are primary males, with one type of mating system, while others are protogynous males, with a different mating behavior. The presence of “true’’ males or females in the Calyptraei- dae, if they are shown to be genetically distinct from protandrous individuals, would provide an example of dimorphism within a sex in molluscs. Presently, there is no known behavioral difference of genetic origin within males of Crepidula. However, the youngest C. fornicata males often wander, while larger, older males are per- manently attached to a particular stack of females. Organisms with Comparable Sexuality Social interaction in sex determination is common among molluscs and other inverte- brates, but seems to be most prevalent in sedentary, territorial, and parasitic organ- isms with dispersive (e.g., planktonic) larval stages. It often accompanies the phe- nomenon of dwarf males that are parasitic on the females, or the phenomenon of protandry. The parasitic mesogastropod family Entoconchidae (Lützen, 1968) and the sedentary echiuroid genus Bonellia (Baltzer, 1931; Gould-Somero, 1975) illus- trate labile sex determination in which the sex of a juvenile is determined by whether it settles near or on a female. If it does, the juvenile becomes a dwarf male. How- ever, about 10% of Bonellia echiuroids differentiate sexually without regard to their associations with other Bonellia indi- viduals; thus there are “true’’ males and PROTANDRY AND LABILE SEX DETERMINATION 383 females (Leutert, 1975). Other genera of echiuroids lack sexual lability; interestingly, these genera are mobile and lack sexual dimorphism. Socially-mediated sex change has been reported in several fish, such as Anthias and Labroides (Smith, 1975) and Thalas- soma bifasciatum (Warner et al., 1975). Labile sex determination of a different sort occurs in the hermaphroditic fish A/vu/us marmoratus. |п this animal, both genetic and physical environmental factors such as temperature control sex determination (Harrington, 1971). Many plants also have environmental sex determination in which physical rather than social factors are important (Charnov & Bull, 1977). In molluscs, the physical environment can play a role in sex determination in pro- tandrous hermaphrodites; for example, starved animals tend to remain male (Coe, 1938a). This is because they do not have enough stored energy to accomplish the drastic anatomical remodeling required by the change from male to female. The polychaete Ophryotrocha puerilis has a sex reversal pattern whereby a female-female contact causes one of the pair to revert to male (Pfannenstiel, 1975). This pattern of labile sex determination does not occur in molluscs. However, sex alteration does occur in some protandrous oysters (Orton, 1927, 1936), in Тегеао navalis (Coe, 1941), and in Patella vulgata (Orton, 1919). It does not seem to be based on social encounters, but its control mechanism is not known. Bacci (1951, 1955) used the term “unbalanced hermaphrodite’ to describe protandrous animals in which the age at sex change varies, he presumed genetically, between populations. In this category he listed species of Crepidula and Patella, the shipworm genus Jeredo, the serpulid worm Pomatoceros triqueter (see also Ефуп, 1950), Ophryotrocha puerilis, and the starfish Asterina gibbosa.. The wide taxonomic range of these examples is noteworthy. It indicates that differences in the timing of sex change are easily evolved in organisms when the proper selective pressures are present. Bacci's claim that the differences in the age at sex change are genetically controlled needs better documentation, in the light of independent evidence that labile sex change exists in many species. The genetic mechanism for sex determina- tion itself has proven to be elusive, but is a critical key in our understanding of the relative importance of genes and environ- ment in sex change. | review the major references to pro- tandry in molluscs, but | do not attempt an exhaustive list. The important point is that there is a wide taxonomic range of protandrous species, indicating independent lines of evolution. With the exception of Lutzen and his co-workers, in the papers cited below, none of the authors com- mented on the possibility of social regula- tion of sex determination. The greatest concentration of protan- drous species is in the Mesogastropoda. Lutzen (1972) found protandry in Stilifer linckiae, a parasite. The male does not change sex in the presence of the female, thereby insuring the presence of both sexes in the starfish host which usually contains a low number of parasites. Protandrous Parasitic mesogastropods also include Robil- lardia cernica (Gooding & Lutzen, 1973), at least some members of Echineulima mittrei (Lutzen & Nielsen, 1975), and the ectoparasite Epitonium albidum (В. Robertson, personal communication). Pro- tandry occurs in the genus Trichotropis (Yonge, 1962) and in the Capulidae (Graham, 1954), Janthinidae and Scalidae (Ankel, 1926). Morton (1958) reported protandry in a gymnosomatous pteropod, the southern “dwarf form” of Clione limacina. Reports of protandry in the archaeogas- tropods are rare, but Fretter & Graham (1962) claim that protandry is occasionally found in Acmaea. Russell-Hunter & Mc- Mahon (1976) gave evidence of functional protandry in a fresh-water basommato- phoran limpet, Laevapex fuscus. In fact, many pulmonate and opisthobranch her- maphrodites approach functional pro- tandry; that is, the male organs mature before the female. The most surprising report of protandry in the gastropods is that by Smith (1967) on a population of Lora turricula, a member of the neogastropod family Tur- ridae. | know of no other protandrous neogastropod, though | expect that more will be discovered. The anatomy of L. turricula is closer to the hermaphroditic pulmonates and opisthobranchs than to the protandrous hermaphrodites of the Meso- gastropoda, in that there are two separate reproductive systems rather than a sinale ambisexual gonad. The independent origin 384 HOAGLAND of protandry in this organism is clear. Smith’s population of L. turricula is insu- lar, providing an advantage for hermaphro- ditism, but the evolutionary explanation for protandry must be sought in further details of the ecology of the species. Whether all populations of the species are protandrous is unknown. Turning to the bivalves, all species of the Teredinidae (wood-borers) so far studied are protandrous, alternating, or simultaneous hermaphrodites (Sigerfoos, 1908; Kofoid & Miller, 1927; Yonge, 1926; Coe, 1941; Turner, 1966; Herlin-Houtteville & Lubet, 1975). The adults are completely sedentary; larvae vary from completely planktonic to those released as crawling young. Some teredinids, including members of the genus Bankia, have been observed to copulate. Sperm are transferred to the female incurrent siphon via the male’s excurrent siphon. Protandrous Bankia gouldi shipworms, when living at high densities, mature and spawn when dwarf (R. Turner, personal communication). Protandrous Psiloteredo megotara males release sperm in diaphanous membranes which are actively sucked in by the female incurrent siphon. Turner & Cooley (un- publ.) found several individuals of P. megotara isolated in lobster pots. All were males, yet they were all larger than normal females of the species. The implication is that the sex change does not occur in isolated individuals, or at least is not triggered by the organism attaining a certain size, or by its being isolated. There are a few data that suggest that isolated males of Bankia gouldi remain males, while the presence of a second male triggers one to change sex (Culliney, 1969 and personal communication). This would be a reverse form of sexual lability, compared with Crepidula fornicata. М more data indicate that this is so, it may be related to the shipworms’ facultative copu- latory behavior. Isolated males are still able to broadcast sperm, so an isolated male is not reproductively dead. A juvenile attrac- ted to the same substrate at a later date could influence the first individual to change sex via a feminizing compound, if the gonad of Bankia is basically male, rather than female as it is in Crepidula (Le Gall & Streiff, 1975). Some unifying hypotheses on patterns of sexuality and their evolution Table 7 compares the essential popula- tion factors and sexual patterns in the Crepidula species and some other pro- tandrous and/or sexually labile molluscs. Bonellia is added because it is a classic example of labile sex determination, and information about its ecology is available. Can a single hypothesis explain the evolu- tion of sex in all of these organisms? The organisms of Table 7 have several features in common. The females have internal fertilization and are virtually seden- tary. There is a high degree of isolation of breeding units due to discontinuity of suitable substrates, but the species possess effective dispersal stages and may be called opportunistic species (MacArthur, 1962). They frequently colonize new areas, and their preferred substrates are transient. This is particularly true of shipworms, which consume their substrates, but the same effect is achieved with the shallow-water cobble and shell habitats of Crepidula, which are transient in geologic time and often shift during a period of several years. Ghiselin (1969) proposed that simultane- ous hermaphroditism is advantageous for individuals living in small, isolated popula- tions (but see Williams, 1975). Ghiselin felt that isolation (small deme size) could not favor protandry, because effective popula- tion size would not be increased by protandry. However, effective population size by itself has no relation to individual fitness. The critical factor is that under protandry coupled with sexual lability and/or chemical attraction of larvae (either by the species or by the substrate), each individual minimizes the age at which it first reproduces, increasing its reproductive potential. In С. fornicata, for example, isolated individuals rapidly become female and emit species-specific attractants; any second individual contacting the first 15 necessarily a juvenile at the point of metamorphosis or a male, because only such individuals are highly mobile. The juvenile in such a circumstance is guaran- teed to become a functional male, hence immediately to participate in reproduction. As a byproduct, every encounter between individuals can produce offspring. 385 PROTANDRY AND LABILE SEX DETERMINATION *лэрээ;} jeu snoonul-AJe1¡19 Ase}uapas, ‘ayqeiene езер 348!5!44п$4!++ т 5Й ШЪ жж —0—____д —— # * # Э315елеа # a1/seJed ¿poom # * ¿poom # 3 * al * ar * Sr * = * = * st * Ayısuap apow uonendod бБи!рээ- YIM pa]8/91109 01381 xaS эаеэнаае P HEMP jou = [o ö + + # = # + + = ‘dds eıyyzauog + == ? 6 + 150y 0} > te + de u ar 98/4941] ¿INS + _ (o) P Ajqissod 3504 O] — + + + + + 194331441 eu/¡naulyo3 (—) eJezobau # a lo) г — роом 0} + + + + элценпэе; + 0P34920/IS4 (=) # se г 9 # poom 0} + + + + элцезпэв} + 1p¡nob eıyueg = # 2 2 = = + = + = + + sısugulyd eaesdAjed + = 2 64109 tf + = = + + + + xAuo at = 2 6 4109 + + = = + + + + 2322!и10} + + o 6109 + yeam + = + + Je + euejd > in о Pp = = EN = A _ en Je ехэлиоэ = + ? P = = + = + = + + esunpe eınpıda4) ajewaj BUIPMOI2 эеша; jenpinıpul 2хэ$ ¿siimpe оз срэиши ¿peunsuos iAysyed эелле| wonez Алруезо14 S3193dS jo Japun e JEU pajejosí ajiqe] pademe eq Aew ajelmsqns ajelisqns 51403 -1111194 ээцэпзи! pajiemp Bulj119s BunoA sajıusanr 21211SQNS -yue|q ¡eulaju] Japun sajeuway3 [ENPIA e yo abueyo -1pul xas Би! xas BunoA e -uol}Juny jo juaul JO xaS 151 -94043$04 Е = “сэзелаазлали! Алезиара$ |елэла$ ul $илаззе4 элзопрол4эл pue xas jO uosisedwog ‘/ 319vVL 386 HOAGLAND Only in the non-planktonic Calyp- traeidae and possibly in the shipworms must | invoke another factor to explain the advantages of protandry: small differences in genetic proclivity to change sex. Such genetic differential guarantees the presence of both sexes in any single age cohort, even without the powerful aid of sexual lability. This is necessary to insure the reproductive potential of individuals. This factor, in- cidentally, is present in the other calyp- traeids as well. Protandry is not the only solution to problems caused by isolation of individuals or small demes. Simultaneous hermaphro- ditism coupled with facultative self-fertiliza- tion theoretically could accomplish the same things (minimization of age at first reproduction, maximization of reproductive events per lifetime, stabilization of the sex ratio) but with the additional cost of inbreeding. Although the sex ratio of protandrous animals is affected by the age at sex change of individuals making up a population, it is not the sex ratio that is under natural selection. Rather, it is the optimal age of individuals at sex change. Apart from providing flexibility in sex ratio, protandry is advantageous to indi- viduals if one sex increases in fertility with age (and size) much more rapidly than the other (Ghiselin, 1969; Warner et al., 1975). This is true of calyptraeids and shipworms, and probably molluscs in general. The female does not discriminate among males on the basis of age or size, hence male fecundity is not strongly correlated with either. Young males of Crepidula gain advantage by growing longer copulatory organs, but do not gain advantage from general size increase. In fact, smaller males are more mobile and have a better chance at fertilizing several females than do larger, more sedentary males. | have observed a burst of growth at the time of sex change in many individuals of C. fornicata, and Le Gall (1973, 1974) reported that growth is greater in individuals of C. fornicata the nearer they are to the base of a chain. Female fecundity is strongly related to age and size, for the number of eggs produced per season rises steeply with body size (Hoagland, 1975). In sum, plots of fecun- dity versus age or size are different for the two sexes. The differences are just as predicted by models of protandry based on maximization of individual fitness. Lack of male care of offspring and the relationship between size, sex, and fecundity explain why protandry and not protogyny is common in the Mollusca. Population density has been implicated in the adult sex ratios of several inverte- brates with labile sex determination, e.g., two parasitic nematodes (Caullery & Comas, 1928; Christie, 1929), monstrillid copepods (Malaquin, 1901; Anderson, 1961), some rotifers, cladocerans and Ophryotrocha (Coe, 1938b). Dense popula- tions tend to have proportionally more males, as described in this paper for С. fornicata and as Coe (1944) also demon- strated for Crepidula species. In the pro- tandrous species, total egg output per adult tends to decrease as population density increases, because each individual spends a greater proportion of its lifetime as a male. Reduction in egg output is probably counter-balanced by an increase in male fecundity. Unfortunately, there are no data on this critical parameter. However, Warner (1975) found variation in the time of sex change in protogynous fish and could relate it to total (male plus female) fecundity and hence to individual fitness. Ghiselin (1969) pointed out that brood- ing of young by the female often accom- panies protandry, and hermaphroditism in general. This pattern certainly holds for Crepidula and the family Calyptraeidae. Charnov et al. (1976) use a fitness set argument to demonstrate that brooding favors hermaphroditism because female ex- penditure of reproductive effort occurs later than male expenditure. The adaptive significance of differences among protandrous organisms The major distinction in Table 7 is between organisms with and without en- vironmentally-influenced sex change. Those without it are characterized by the possi- bility of substrate limitation. Most lack a planktonic larval stage. Juveniles are not attracted to adults, and females mature at a small size when crowded. In such organ- isms, risks of crowding are reduced by the absence of gregariousness; crowding is often associated with rapid maturation and early deaths of adults due to destruction of the substrate. Labile sex determination is linked to juvenile attraction to adults, and in most cases, lack of substrate limitation. The sex ratio of populations of species with en- РВОТАМОВУ AND LABILE SEX DETERMINATION 387 vironmentally mediated sex determination is correlated with population density. The sexually labile species that | have studied all have planktonic larvae, but all species with planktonic larvae are not sexually labile. Environmental influence in the tim- ing of sex change allows the size and age of an organism at sex change to be tailored to suit very localized conditions, even if the genetic constitution of local individuals reflects a much broader gene pool from a broader range of environments. Charnov & Bull (1977) elaborated on this idea for a variety of organisms, including plants. Crepidula fornicata, С. onyx, С. plana, Bonellia, and even the protogynous fish Thalassoma bifasciatum have planktonic development and dispersal coupled with low mobility as adults. In the protandrous calyptraeids without planktonic develop- ment, timing of sex change is under direct genetic control. On the basis of this known difference, one could predict that there is sexual lability in В. gouldi and other shipworms with planktonic stages, but not in those that hatch at a crawling stage. Mechanisms of control of sex behavior and sex change The basic mechanism to control be- havior in all the species of Crepidula and perhaps the other protandrous invertebrates is likely to be pheromonal regulation, although evidence is circumstantial. In the case of C. convexa, pheromonal involve- ment in sex change and gregarious behavior appears to be minimal, restricted to male- female interaction at the time of mating. In the case of С. fornicata, the responses of all individuals to one another are probably pheromone-mediated. Gregariousness and sexual lability probably have a common mechanism in calyptraeids, explaining why they are always coupled. The most direct evidence for phero- monal behavior control, as opposed to a tactile mechanism, was obtained by Gould (1919, 1952). He suspended males of C. plana above females and observed delays in sex change. My finding that the density of females, rather than total population den- sity, affects the mean age at sex change of new members of a population, indicates that a masculinizing pheromone resides in the female. However, discovery that mated males are more affected by female density than unmated males suggests that a tactile or localized component also could be a part of the behavioral mechanism, as suggested first by Coe (1938b). There is no evidence of a feminizing compound pro- duced by males; the female condition is physiologically dominant. That is, an isola- ted animal eventually becomes female (Le Gall & Streiff, 1975). QUESTIONS FOR THE FUTURE To explain the actual path of evolution of protandry in molluscs, it is necessary to know the genetic basis of sex determina- tion, the chemical basis for protandry, and the nature of the suspected pheromone(s) involved in sexual behavior. The chemical basis for protandry is not necessarily the same in all molluscs. Evolutionary relation- ships in the mesogastropods, particularly in the highly modified parasitic groups, could perhaps be seen through chemical analysis of sexual and behavioral mechanisms. So far, most of the thought in these areas is speculative. The mechanism controlling dwarfing, by which a male matures at a smaller-than-normal size even before it has run out of room to grow, is also unknown. One reproductive mode barely touched upon here, but which is an extension of protandry, is alternating sexuality within individuals. It should be examined in the light of this paper and the theoretical works of Warner (1975), Warner et al. (1975), Leigh et al. (1976), and Charnov et al. (1976). Species in which some popula- tions but not all have been reported to show protandry also are worthy of further study, particularly to see if sexual lability occurs with protandry. Ап examination of all invertebrates exhibiting protandry and/or sexual lability will include many which are not sedentary or substrate-restricted as adults, for ex- ample, the pandalid shrimp (Wenner, 1972; Charniaux-Cotton, 1965). How will these fit into the framework developed here? Also, it presently appears that highly mobile carnivores are not protandrous; exceptions should be sought. Finally, are there cases of invertebrates for which we predict protandry on the basis of other life history characteristics, but find it lacking? Answers to such questions will expand both theoretical and intuitive understanding of the evolution of protandry in molluscs and in other organisms as well. 388 HOAGLAND ACKNOWLEDGEMENTS | thank E. Charnov, G. Davis, and J. Murray for reading and commenting upon the manuscript. This work was supported in part by a Gibbs Fellowship from Harvard University and a contract #AT(49- 24)-0347 with the U.S. Nuclear Regulatory Commission. REFERENCES CITED ANDERSON, F. S., 1961, Effect of density on animal sex ratio. Oikos, 12: 1-16. ANKEL, W. E., 1926, Spermiozeugmenbildung durch atypische (apyrene) und typische Sper- mien bei Scala und Janthina. Verhandlungen der Deutschen Zoologischen Gesellschaft, Leipzig, 31: 193-202. BACCI, G., 1951, L’ermafroditismo di Ca/yptraea chinensis L. e di altri Calyptraeidae. Pubblica- zioni della Stazione Zoologica di Napoli, 23: 66-90. BACCI, G., 1955, La variabilita dei genotipi sessuali negli animali ermafroditi. Pubblica- zioni della Stazione Zoologica di Napoli, 26: 110-137. BACCI, G., 1965, Sex determination. /nterna- tional Series of Monographs in Pure and Applied Biology. Division: Zoology. KER- KUT, G. A., Ed., Pergamon Press, Oxford, 26: 306 p. 4 BALTZER, F., 1931, Echiuridae. In: KUKEN- THAL, Handbuch der Zoologie, de Gruyter, Berlin, 2: 62-168. CAULLERY, М. & COMAS, M., 1928, Le determinisme du sexe chez ип nematode parasite des larves de chironomes. Comptes Rendus... de L’Academie des Sciences, Paris, 186: 646-647. CHARNIAUX-COTTON, H., 1965, Hormonal control of sex differentiation in invertebrates. In: de HAAN, В. L. € URSPRUNG, H., Organogenesis, Holt, Rinehart & Winston, New York, p. 701-740. CHARNOV, Е. L. €: BULL, J., 1977, When is sex environmentally determined? Nature, 266: 828-830. CHARNOV, E. L., MAYNARD SMITH, J. & BULL, J., 1976, Why be a hermaphrodite? Nature, 263: 125-126. CHRISTIE, J. R., 1929, Some observations on sex in the Mermithidae. Journal of Experi- mental Zoology, 53: 59-76. COE, W. R., 1935, Sexual phases in Crepidula. Journal of Experimental Zoology, 72: 455-477. COE, W. R., 1938a, Conditions influencing change of sex in mollusks of the genus Crepidula. Journal of Experimental Zoology, 77: 401-424. COE, W. R., 1938b, Influence of association on the sexual phases of gastropods having pro- tandric consecutive sexuality. Biological Bulle- tin, 75: 274-285. COE, W. R., 1941, Sexual phases in wood-boring mollusks. Biological Bulletin, 81: 168-176. COE, W. R., 1944, Sexual differentiation т mollusks. Il. Gastropods, Amphineurans, Scaphopods, and Cephalopods. Quarterly Re- view of Biology, 19: 85-97. COE, W. R., 1948, Variations in the expression of sexuality in the normally protandric gastro- pod Crepidula plana Say. Journal of Experi- mental Zoology, 108: 155-169. COE, W. R., 1953, Influences of association, isolation, and nutrition on the sexuality of snails of the genus Crepidula. Journal of Experimental Zoology, 122: 5-19. CULLINEY, J., 1969, Larval biology and recruit- ment of the shipworms Teredo navalis and Bankia gouldi in the Newport Estuary, North Carolina. Ph.D. dissertation, Duke University, Durham, North Carolina, 174 p. FOYN, B., 1950, Sex and inheritance in the serpulid Pomatoceros triqueter. Nature, 165: 652. FRETTER, V. & GRAHAM, A., 1962, British Prosobranch Molluscs; their functional anat- omy and ecology. Ray Society, London, xvi + 7551p: GALLARDO, С. S., 1977, Two modes of development in the morphospecies Crepidula dilatata (Gastropoda: Calyptraeidae) from Southern Chile. Marine Biology, 39: 241-251. GHISELIN, M. T., 1969, The evolution of hermaphroditism among animals. Quarterly Review of Biology, 44: 189-208. GOODING, R. U. & LUTZEN, J., 1973, Studies оп parasitic gastropods from echinoderms. III. A description of Robillardia cernica Smith 1889, parasitic in the sea urchin Echinometra Meuschen, with notes on its biology. Konge- lige Danske Videnskabernes Selskab Biologiske Skrifter, 20(4): 1-22, 4 pl. GOULD, H. N., 1917a, Studies on sex in the hermaphrodite mollusk Crepidula plana. 1. History of the sexual cycle. Journal of Experimental Zoology, 23: 1-68. GOULD, H. N., 1917b, Studies on sex in the hermaphrodite mollusk Crepidula plana. 2. Influence of the environment on sex. Journal of Experimental Zoology, 23: 225-250. GOULD, H. N., 1919, Studies on sex in the hermaphrodite mollusk Crepidula plana. 3. Transference of the male producing stimulus through sea water. Journal of Experimental Zoology, 29: 113-120. GOULD. H. N., 1952. Studies on sex in the hermaphrodite mollusk Crepidula plana. 4. Internal and external factors influencing growth and sex development. Journal of Experimental Zoology, 119: 93-163. GOULD, H. N. & HSIAO, S. C., 1948, New experiments and observations on sexual in- stability in Crepidula plana. Biological Bulle- tin, 95: 255-256. GOULD-SOMERO, M., 1975, Echiura. In: GIESE As. «Ge &EBREARSE SiS! ECs, Reproduction of marine invertebrates. I. Annelids and Echiurans. Academic Press, New York, p. 277-311. GRAHAM, A., 1954, The anatomy of the prosobranch Trichotropis borealis Broderip and Sowerby, and the systematic position of the Capulidae. Journal of the Marine Biologi- cal Association of the United Kingdom, new series, 33(1): 129-144. HARRINGTON, R. W., 1971, How ecological and genetic factors interact to determine when self-fertilizing hermaphrodites of Rivulus mar- moratus change into functional secondarv males, with a reappraisal of the modes of intersexuality among fishes. Copeia, 1971: 389-432. PROTANDRY AND LABILE SEX DETERMINATION 389 HERLIN-HOUTTEVILLE, P. & LUBET, P. E., 1975, The sexuality of pelecypod molluscs. In: REINBOTH, R., Ed., /ntersexuality in the animal kingdom, Springer-Verlag, New York, p. 179-187. HOAGLAND, K. E., 1975, Reproductive strate- gies and evolution in the genus Crepidula (Gastropoda: Calyptraeidae). Ph.D. disserta- tion, Harvard University, Cambridge, Massa- chusetts, 360 p. HOAGLAND, K. E., 1977a, A gastropod color polymorphism: one adaptive strategy of phenotypic variation. Biological Bulletin, 152: HOAGLAND, К. E., 1977b, Systematic review of fossil and Recent Crepidula and discussion of evolution of the Calyptraeidae. Ма/асо/одга, 16: 353-420. ISHIKI, H., 1936, Sex changes in the slipper limpets, Crepidula aculeata and Crepidula walshi. Journal of Science, Hiroshima Univer- sity, Series B, Division 1, 4: 91-99. KOFOID EG? FATRSZAMIERERF FRE 761927, Biological section. In: HILL, C. L. & KO- FOID, C. A., Ed., Marine borers and their relation to marine construction on the Pacific coast. San Francisco Bay Marine Piling Com- mission, San Francisco, California, p. 188-343. LE GALL, P., 1973, Activation de la croissance par les mäles dans les chaines de Crepidula fornicata Phil. (mollusque, mésogastropode). Comptes Rendus... de L’Academie des Sci- ences, Paris, 276, series D (4): 615-617. LE GALL, P., 1974, Relations entre la croissance et la sexualité chez Crepidula fornicata Phil. Haliotis, 4: 101-105. LE GALL, P. & STREIFF, W., 1975, Protandric hermaphroditism in prosobranch gastropods. In: REINBOTH, R., Ed., /ntersexuality in the animal kingdom, Springer-Verlag, New York, p. 170-178. LEIGH, Е. С. Jr., CHARNOV, Е. L. & WAR- NER, R. R., 1976, Sex ratio, sex change, and natural selection. Proceedings of the National Academy of Sciences, U.S.A., 73: 3656-3660. LEUTERT, R., 1975, Sex-determination in Bonel- lia. In: REINBOTH, R., Ed., /ntersexuality in the animal kingdom, Springer-Verlag, New York, p. 84-90. LUTZEN, J., 1968, Unisexuality in the parasitic family Entoconchidae (Gastropoda: Proso- branchia). Malacologia, 7: 7-15. LUTZEN, J., 1972, Studies on parasitic gastro- pods from echinoderms. И. On Stilifer Broderip, with special reference to the struc- ture of the sexual apparatus and the reproduc- tion. Kongelige Danske Videnskabernes Selskab _ Biologiske Skrifter, 19(6): 1-18. LUTZEN, J. & NIELSEN, K., 1975, Contribu- tions to the anatomy and biology of Echineu- lima п. 9. (Prosobranchia: Eulimidae), parasitic on sea urchins. Videnskabelige Meddelelser fra Dansk Naturhistorisk Forening, 138: 171-199. MACARTHUR, R. H., 1962, Some generalized theorems of natural selection. Proceedings of the National Academy of Sciences, U.S.A., 48: 1893-1897. MALAQUIN, A., 1901, Le parasitisme evolutiv des Monstrillidae. Archives de Zoologie Ex- périmentale et Générale, 9: 81-232. MORTON, J. E., 1958, Observations on the gymnosomatous pteropod Cl/ione limacina (Phipps). Journal of the Marine Biological Association of the United Kingdom, 37: 287-297. ОВТОМ, J. H., 1919, Sex phenomena in the common limpet (Patella vulgata). Nature, 104: 373-374. ORTON, J. H., 1927, Observations and experi- ments on sex-change in the European oyster, Ostrea edulis. Journal of the Marine Biological Association of the United Kingdom, 14: 967-1045. ORTON, J. H., 1936, Observations and experi- ments on sex-change in the European oyster, Ostrea edulis L. 5. A simultaneous study of spawning in 1927 in two distinct geographical localities. Mémoires du Museum Royale d’Histoire Naturelle de Belgique, Series II, f. 3: 997-1056. PARKER, G. A., 1970, Sperm competition and its evolutionary consequences in the insects. Biological Reviews, 45: 525-567. PFANNENSTIEL, H.-D., 1975, Mutual influence on the sexual differentiation in the protandric polychaete Ophryotrocha puerilis. In: REIN- BOTH, R., Ed., /ntersexuality in the animal kingdom, Springer-Verlag, New York. р. 48-56. PUTNAM, D. A., 1964, The dispersal of young of the commensal gastropod Crepidula adunca from its host Tegula funebralis. Veliger, 6 (Suppl): 63-66. RUSSELL-HUNTER, W. D. & MCMAHON, R. F., 1976, Evidence for functional protandry in a fresh-water basommatophoran limpet, Laeva- pex fuscus. Transactions of the American Microscopical Society, 95(2): 174-182. SHELDON, R. W., 1967, Relationships between shell-weight and age in certain molluscs. Journal of the Fisheries Research Board of Canada, 24: 1165-1171. SIGERFOOS, C. P., 1908, Natural history, organ- ization, and late development of the Teredini- dae, or shipworms. Bulletin of the Bureau of Fisheries (U.S.A.), 37: 191-231. SMITH, C. L., 1975, The evolution of her- maphroditism in fishes. т: REINBOTH, R., Ed., /ntersexuality in the animal kingdom, Springer-Verlag, New York, p. 295-310. SMITH, E. H., 1967, The reproductive system of the British Turridae (Gastropoda: Toxo- glossa). Ме//дег, 10: 176-187. TURNER, R. D., 1966, A survey and illustrated catalogue of the Teredinidae. Museum of Comparative Zoology, Harvard University, Cambridge, Massachusetts, 265 p. WARNER, R. R., 1975, The adaptive significance of sequential hermaphroditism in animals. American Naturalist, 109: 61-82. WARNER, R. R., ROBERTSON, D. R. & LEIGH, E. G. Jr., 1975, Sex change and sexual selection. Science, 190: 633-638. WENNER, A. M., 1972, Sex ratio as a function of size in marine Crustacea. American Natu- ralist, 106: 321-350. WILLIAMS, С. C., 1975, Sex and evolution. Monographs in population biology. Princeton University Press, Princeton, New Jersey, 200 p. WYATT, H. V., 1960, Protandry and self-fertiliza- tion in the Calyptraeidae. Nature, 187: 520-521. YONGE, С. M., 1926, Protandry in Teredo norvegica. Quarterly Journal of Microscopical Science, new series, 70: 391-394. YONGE, C. M., 1962, On the biology of the mesogastropod Trichotropis cancellata Hinds, a benthic indicator species. Biological Bulletin, 122: 160-181. 390 HOAGLAND APPENDIX A. A description of the populations studied. Density * Description C. fornicata 10 Sparse, stable adult population size. Low recruitment. Nahant, Massachusetts On stones in tidepools. No substrate shortage. Stacks (N) rarely exceed 3. Many individuals 6-8 years old. Vineyard Haven, Martha’s 200 Dense, stable population, known since before 1900. Vineyard, Massachusetts Numerically constant over time. Large stacks on shells (MV) and stones in mud. Subtidal except for some washed onto beach. Waquoit Bay, Cape Cod, 200 Old, stable population, similar to Martha’s Vineyard, Massachusetts (W) but less extensive in area. Tiverton, Rhode Island (RI) 500 A colonizing population, less than 3 years old, at a new marina. Stacks on Mytilus edulis, undersides of docks. Marine Biological Lab. Beach, 15 Sparse, unstable population. High physical mortality Woods Hole, Massachusetts due to winter ice scour. Low food supply. Substrate of a (MBLB) few rocks and cobbles near stone jetties surrounded by clean sand. Essentially a colonizing population in 1972. Excess substrate. Stacks do not exceed 3. Subtidal. C. onyx Balboa Island, 160 Old, dense, numerically stable population. First collec- California (B) ted circa 1900. Stacks on restricted substrate (shells) in muddy sand. Subtidal. C. plana Woods Hole Yacht Club, 80 Sparse population on concave sides of glass and inside Woods Hole, Massachusetts shells. Sandy substrate, shallow subtidal, a few awash in (WHYC) surf attached to shells. C. convexa Gunning Point, Cape Cod, 80 Small individuals, dense population on eel grass and Massachusetts (GP) small Littorina littorea. Shallow to intertidal, sheltered bay with mud-sand substrate. Nobska Beach, Woods Hole, 10 Sparse population on rare L. /ittorea and occasionally Massachusetts (Nob) on stones. Wave-washed rocky intertidal point. Waquoit Bay, Cape Cod, 50 Small individuals, moderately dense population on eel Massachusetts (W) grass and shells. Shallow subtidal to intertidal, protected bay. Duxbury, Massachusetts 50 Moderately dense, numerically stable population on (D) shells and stones, sheltered sandy area. Shallow subtidal to intertidal. Woods Hole Yacht Club, 30 Sparse, numerically stable population on shells, espe- Woods Hole, Massachusetts cially L. /ittorea, stones, glass, and assorted debris. (WHYC) Shallow subtidal to intertidal. *Number рег m? of available substrate. PROTANDRY AND LABILE SEX DETERMINATION 391 APPENDIX B. Summary of the relationship between population density and the size and age at sex change—four Crepidula species. C. fornicata, C. onyx, and C. plana: 1. Population density vs. size (dry weight) at sex change solitary mated A. Total density ie eq: у = т еа: у = All planktonic species .325 .48 + .001 x .124 1.69 + .001 x В. 1. population removed .922 .25 + .004 x .756 1.21 + .009 x С. fornicata only, В. |. removed .991 .35 + .004 x .798 1724201052 B. Female density All planktonic species .500 .39 + .005 x .405 ELIO R. |. removed .776 .37 + .008 x .767 ая (0 x= С. fornicata only, В. |. removed .832 .46 + .008 x .890 1-20-03) pce 2. Population density vs. age at sex change, C. fornicata only A. Total density With В. 1. population .174 4.42 + .002 x .191 1247 Ех Without В. |. population .938 3.88 + .010 x .765 9.44 + .04 x* В. Female density With В. 1. population .367 4.15+.01 x .507 9.48+.04 x Without В. 1. population .832 АТ 02 x .877 9.34 +.12 x* С. сопуеха: 1. Population density vs. size (dry weight) at sex change A. Total density .92 .03 — .0002 x B. Female density .94 .04 — .0007 x 2. Population density vs. age at sex change: no correlation. *Slopes significantly different from zero. in, + “4 a = art « BOO. 4 rest « bre DER м Ganso $ в т TON ER, 5 sre oon a u ES ‘

Pe ARA cOn O, Faro, a cla keine Fr meee A Ai = SBE mas vacate 24 aa Me > 1 u me ee 2 pomo ква ra Cris vil 5 а 1 “A Nr cite Part Cae JU am 9% tal ae Ki Y ML PA A pa Me o luca НОМ imp (C000, fees ОХ EL 14) ro | vo ys | toc 00 muni | 14d к 000. - MO he у Инф - 009 , Us Modele 102 oP reel ТИ vtianab | iva An AAA À PP q Ne Nat na Proa io Mint а | зло sat 4 u ing A ма vi ' a TO d & | ” a t wi y rug № Y + e A м y 4 u ie u Ma у р | N 2 - > | 1 : = u INDEX TO SCIENTIFIC NAMES IN VOLUME 17, NOS. 1-2 An asterisk (*) denotes a new taxon abacoense, Cerion, 225, 226, 231, 232 arborea, Cyathea, 255, 301 Abbottella, 217 arbustorum, Arianta, 267, 277, 336, 337 abchasica, Caucasigena, 2, 49 Archaeogastropoda, 201, 204, 207-209, 216, 218- Acella 220, 383 haldemani, 353 Archaeopraga, 180 acicularis, Cypraea spurca, 198 Architectonica, 196 Aciculidae, 208-211, 216, 217 arenaria, Ammophila, 327 Acmaea, 383 arenaria, Mya, 117 digitalis, 285 arenosa, Lyonsia, 58 testudinalis, 103 Argobuccinum Acmidae, 208, 210, 211 argus, 126, 145 Actinauge argus, Argobuccinum, 126, 145 verrilli, 65 Arianta Aculifera, 100, 101, 184 arbustorum, 267, 277, 336, 337 “Aculifera,'’ 107 Arion, 350 acuta, Physa 353 armeniaca, Caucasigena, 2, 23, 43, 48, 51, 56 acutangula, Polydontes, 241-315 “Arum,” 259 Adelopoma, 210 aspersa, Helix, 237, 267, 289, 293, 296, 336, 337 adunca, Crepidula, 378, 385 Assimineidae, 208 Aeolidia Asterina papillosa, 64, 68, 69 gibbosa, 383 Aktugaia, 180 Asthenothaerus, 58 albanyensis, Goniobasis, 158, 159, 161 Astraea albida, Polystira, 198 americana, 198 albidum, Epitonium, 68, 70, 383 phoebia, 198 albus, Planorbis, 353 tecta, 198 Alcadia Astropecten, 60 charmosyne, 218 athearni, Goniobasis, 157-161 Alchornea auricoma, Zachrysia, 244 latifolia, 301 auritula, Pisania, 198 Aldabanella, 182, 183 Australorbis Aldanella, 177 glabrata, 268 Algae, 143, 219, 220, 261 avena, Hyalina, 198 Allogona Babinka, 172 profunda, 256, 285 bakowskii, Edentiella, 2, 39, 47, 49, 51 Alycaeinae, 210, 218 Bankia, 384 americana, Astraea tecta, 198 gouldi, 384, 385, 387 Ammophila barbouri, Graptemys, 161 arenaria, 327 Basommatophora, 218, 383 Amnicola Batillaria limosa, 354 minima, 198 Amphineura, 100, 101, 167, 174, 184 Bellamya “Amphineura,”* 107 sumatrensis, 91, 92 Ancylidae, 353 Bellamyinae, 91 Ancylus Bellerophon, 199-201 fluviatilis, 353, 356, 358-360 Bellerophontacea, 168, 177, 178, 180, 182, 199- lacustris, 354 202, 205 angulata, Turbinella, 198 Bellerophontida, 180, 181 angulifera, Littorina, 198 bendalli, Cerion, 225-227, 231, 232 Anisopleura, 179 bengalensis, Idiopoma, 91, 92, 95 Anisus berteroiana, Psychotria, 301 rotundatus, 353 Berthelinia, 173 annulata, Nucula, 104 Biangularia, 204 anodonta, Cerion, 225 bidentatus, Melampus, 354-357 Anoplitella, 2, 53 bielzi, Trichia, 47 schaposchnikovi, 2, 14, 27, 28, 44, 49, 51 bifasciatum, Thalassoma, 383, 387 Anthias, 383 biminiense, Cerion, 229 Anthopleura Biomphalaria xanthogrammica, 63, 70 glabrata, 111, 113, 115, 268, 277 Aplacophora, 99-109, 167, 174, 184, 185 straminea, 111-115 Aplexa Biplex, 204 hypnorum, 353 Biston, 317 Aporrhais, 196 Bithynia Araceae, 260 tentaculata, 354, 356, 357 (393) 394 MALACOLOGIA Bivalvia, 101, 166, 167, 170-174, 179, 365 Blanfordia, 208 Bonellia, 382, 384, 385, 387 borinquensis, Cordia, 301 Brachiopoda, 171 Brilonella, 204 Bromeliaceae, 253, 254, 273, 278, 296, 301, 304, 305, 312 Bryophyta, 220, 261 Bryozoa, 205 Buccinum undatum, 60 Bulimulus, 294 Bunodactis stella, 65 Bunyerichnus, 184 busbyi, Paryphanta, 291 Busycon, 60 contrarium, 198 Caenogastropoda, 204-206 caerulescens, Cerion, 225, 233 Calaurops, 203 Calycogonium squamulosum, 301 Calyptraea, 379 chinensis, 379, 385 Calyptraeidae, 365, 366, 378, 382, 386, 387 Camaenidae, 241-315 Campeloma, 354 decisum, 93, 94 geniculum, 73-97 rufum, 93-95, 354 tannum, 93, 94 canaliculatus, Turbo, 198 cantiana, Monacha, 336, 337 Caobangia, 153 capitata, Capitella, 157 Capitella capitata, 157 Capulidae, 383 Caracolus, 241-315 carocollus, 241-315 marginella, 241-315 Cardiidae, 58 *caria, Leucozonella, 1,2, *12, 13, 41, 48-54 carinatus, Planorbis, 353 carocollus, Caracolus, 241-315 caryodes, Leucozonella, 2,9, 10, 40, 48, 50-54 Caspicyclotus, 216 Cassis flammea, 198 Caucasigena, 2, 21, 39, 50, 52, 53, 55 abchasica, 2, 49 armeniaca, 2, 23, 43, 48, 51, 56 eichwaldi, 2, 26, 27, 44, 49, 51, 55 rengarteni, 2, 23, 25-27, 43, 44, 48, 51, 55 schaposchnikovi, 2, 14, 27, 28, 44, 49, 51 thalestris, 2, 29, 45, 48, 51 tschetschenica, 2, 24, 43, 48 caudatus, Falcidens, 103, 105 Caudofoveata, 99-101, 167, 174 cavendishii, Musa, 254 Cecropia, 254, 278 peltata, 255, 301 celata, Cliona, 146, 149 Cepaea, 227, 235, 237, 292-294, 297, 317-339 hortensis, 265, 293, 318, 331, 336-338 nemoralis, 265, 277, 285, 293, 294, 317-339 sylvatica, 336 vindobonensis, 336 Cephalopoda, 101, 166-170, 184 Cepolis, 228 Ceraunocochlis, 205 Cerion, 164, 223-239 abacoense, 225, 226, 231, 232 anodonta, 225 bendalli, 225-227, 231, 232 biminiense, 229 caerulescens, 225, 233 chrysaloides, 228 eximeum, 225, 232, 233 fernandina, 224-226, 233 geophilus, 234 glans, 225, 226, 232 incanum, 225, 226, 233, 234 lerneri, 229 malonei, 225, 233 moralesi, 234 nudum, 225 pauli, 225, 232 pillsburyi, 229 stevensoni, 224-226, 233 striatellum, 235 turnerae, 226 uva, 225, 226, 230 Cerithium, 196, 204 guinaicum, 198 litteratum, 198 cernica, Robillardia, 383 cervus, Cypraea, 198 Chaetoderma, 99, 100, 106, 108 nitidulum, 103-105 Chaetodermatida, 100 Chaetodermatidae, 101, 103 Chaetodermatoidea, 100 Chaetodermis, 100 Chaetodermomorpha, 99-101, 103, 104, 107, 108 charmosyne, Alcadia, 218 Chelodes, 175 chinensis, Calyptraea, 379, 385 choctawhatchensis, Lioplax, 95 Chondropomidae, 208 Chondropominae, 211, 214 chrysaloides, Cerion, 228 chrysochasma, Troschelviana, 218 Ciliellinae, 1 cinerea, Cypraea, 198 cinerea, Urosalpinx, 68, 125-146, 153 Cittarium pica, 198 Cladocera, 386 Cliona celata, 146, 149 Clione limacina, 383 clisia, Mactrellona, 58 Clusia krugiana, 301 coccinea, Stomphia, 64-67, 69 Cochlicellinae, 1 Cochlostomatinae, 210, 216, 218 Coleolus iowense, 170, 171 Columbella mercatoria, 198 columella, Lymnaea, 361 Conchifera, 101, 184 concinna, Trichia, 2, 32, 33, 35, 46, 49, 51 Coniconchia, 168 Conocardioida, 178 contectoides, Viviparus, 93-95, 354 contectus, Viviparus, 91, 92 contortus, Planorbis, 353, 356-360 contrarium, Busycon, 198 Conus, 196 jaspideus, 198 INDEX, VOL. 17 395 mus, 198 densa, Penetrantia, 146 regius, 198 Dentalium, 170 stearnsi, 198 denticulatus, Donax, 118 convexa, Crepidula, 365-391 dermestinum, Vexillum, 198 Copepoda, 386 Deroceras, 341-350 Cordia reticulatum, 341-350 borinquensis, 301 Deuterostomia, 183 corneus, Planorbis, 268, 353 Devonozyga, 204 corona, Melongena, 198 Diacerion, 226 Coryphella Diadumene rufibranchialis, 69 leucolena, 65 costulatum, Epitonium, 70 Diatomeae, 260, 261 Craspedopomatinae, 210, 216 Dickinsonia, 185 Crassispira Dicotyledoneae, 260 cubana, 198 digitalis, Acmaea, 285 Crassostrea dilatata, Crepidula, 379 virginica, 126, 130 dioica, Urtica, 331, 332 Cremnoconchus, 208 Dioscuria, 2, 21, 53 crenimarginata, Opalia, 63, 68, 70 thalestris, 2, 29, 45, 48, 51 Crepidula, 365-391 Diosoma, 184 adunca, 378, 385 divlodon, Odontotrema, 2, 7, 40, 48-54 convexa, 365-391 Diplommatinae, 210 dilatata, 379 Diplozone, 204 fornicata, 365-391 Dirhachopea, 203 nivea, 382 Discinella, 180 norrisiarum, 378 Distorsio, 204 onyx, 365-391 Ditrupa, 170 plana, 365-391 dolabriformis, Spisula, 58 striolata, 378 Donax williamsi, 378 denticulatus, 118 Croton gouldi, 118 poecilanthus, 301 trunculus, 117-124 Cryptomya, 58 vittatus, 117, 118 Crystallophrisson, 100 Drosophila, 159, 161 “Crystallophrisson,’’ 100 pseudoobscura, 317 cubana, Crassispira, 198 willistoni, 161 Cumingia, 58 duplicatus, Polinices, 198 Cyathea *eberhardi, Kokotschashvilia, 1, 2, 17, 18, *19, arborea, 255, 301 20, 49-53 Cyathodonta, 58 Echineulima Cyclophoracea, 208, 209, 211 mittrei, 383, 385 Cyclophoridae, 208-212, 214, 216, 218, 220 Echiuroidea, 382 Cyclophorinae, 210, 218 Edentiella, 2,6, 39, 49, 50, 52, 53 Cyclotopsis bakowskii, 2, 39, 47, 51 subdiscoidea, 217 edentula, Helix, 39 Cymatium, 198 edentula, Trichia, 47 Cyphoma edulis, Mytilus, 126, 129, 130, 132, 133, 399 gibbosum, 198 Egeria Cypraea radiata, 118 acicularis, 198 р eichwaldi, Caucasigena, 2, 26, 27, 44, 49, 51,55 cervus, 198 eichwaldi, Helix, 22 cinerea, 198 elodes, Lymnaea, 353 spurca, 198 elongata, Semitrochatelia, 218 Cypraecassis Endodontidae, 220 testiculus, 198 Enidae, 295 Cyrilla Ensis, 58, 60 racemiflora, 255, 301 Entoconchidae, 382 Cyrtodaria, 58 Entodesma, 58 Cyrtolites, 178-182 Eopteropoda, 168 Cyrtonella, 179, 185 Eostrophia, 225 Cyrtonellopsis, 180 Epitoniidae, 63, 72 Cyrtopleura, 58 Epitonium, 68 Cyrtosoma, 184 albidum, 68, 70, 383 Cyrtospira, 205 costulatum, 70 Dacryodes greenlandicum, 63-72 excelsa, 304 rupicola, 67 Damilina, 180 tinctum, 65, 68, 70 danubialis, Trichia, 2, 38, 39, 49-53 ulu, 68, 70 *darevskii, Hygrohelicopsis, 1,2, *14, 42, 48, 51,55 ericetorum, Turdus, 327 decisum, Campeloma, 93, 94 Euomphalacea, 202, 204 declinata, Hymenocallis, 226 “Euomphalia”’ decollata, Rumina, 233, 237 regeliana, 9, 41 deltoidea, Thais, 198 Euphemites, 199, 200 396 MALACOLOGIA Euryzone, 203 Euterpe globosa, 254, 255, 301, 306 Euthyneura, 218 excelsa, Dacryodes, 304 excentricus, Hebetancylus, 354 eximeum, Cerion, 225, 232, 233 Falcidens, 106 caudatus, 103, 105 fasciata, Tegula, 198 fasciatus, Viviparus, 92, 94, 354 Fasciolaria, 198 hunteria, 198 lilium, 198 tulipa, 198 felina, Tealia, 64-66, 68, 69 ferghanica, Lecozonella, 2, 8, 40, 48, 50-54 fernandina, Cerion, 224-226, 233 Ferrissia rivularis, 354, 356 filicina, Trichia, 31, 47 Filicinella, 47 flammea, Cassis, 198 floridensis, Goniobasis, 157-160 fluviatilis, Ancylus, 353, 356, 358-360 foetidissima, Iris, 318 follyensis, Urosalpinx cinerea, 125-142 fontinalis, Physa, 353, 356 Foraminifera, 170 forbesianum, Lithidion, 217 Fordilla, 166, 172-174, 177-179, 186 fornicata, Crepidula, 365-391 Fruticocampy laea, gerassimovi, 25 phaeolaema, 21 tenuitesta, 21 fulgurans, Nerita, 198 Fungi, 143, 307 Fungia scutaria, 70 fuscus, Laevapex, 354, 383 Gari, 58 Gasconadia, 204 Gastropoda, 101, 166, 167, 168, 176-178, 184, 186, 193-206, 207, 217, 351-364, 383 Gecarcinus lateralis, 228 gemmata, Thyonella, 157 geniculum, Campeloma, 73-97 Geomelania, 208 Geomitrinae, 1 geophilus, Cerion, 234 georgianus, Viviparus, 73-97 gerassimovi, Helix, 25, 44 gibbosa, Asterina, 383 gibbosum, Cyphoma, 198 glabrata, Biomphalaria [= Australorbis], 111, IIS 12681277 glans, Cerion, 225, 226, 232 globosa, Euterpe, 254, 255, 301, 306 Gonactinia prolifera, 64-66, 68, 69 Goniobasis, 158-161 albanyensis, 158, 159, 161 athearni, 157-161 floridensis, 157-160 gouldi, Bankia, 118, 384, 385, 387 Granodomus, 246 Graptemys barbouri, 161 greelandicum, Epitonium, 63-72 Guarea ramiflora, 301 guinaicum, Cerithium, 198 guttata, Marginella, 198 Gymnosomata, 383 gyrina, Physa, 353, 356 Hainesiinae, 210, 218 Halcioneia, 170 Harmonia, 317 havanensis, Zachrysia auricoma, 244 Hebetancylus excentricus, 354 Helcionella, 182, 183 Helcionellacea, 166, 182, 183, 186 helianthus, Stoichactis, 70 Helicellinae, 1 “Helicellinae,”’ 1,2, 15, 55 Helicidae, 212, 250, 276, 331-339 “Helicidae,” 50 Helicinidae, 208-212, 214, 218, 220 Helicoidea, 1, 49 Helicopsis, 1,2 Helisoma trivolvis, 353, 356, 357 Helix, 227, 344, 348 aspersa, 237, 267, 289, 293, 296, 336, 337 edentula, 39 eichwaldi, 22 gerassimovi, 25, 44 hispida, 31 holotricha, 16 phaeolaema, 21 pomatia, 92, 237, 254, 256 rengarteni, 25 rubens, 8 tenuitesta, 21 unidentata, 31 Hemitrochus, 228 Heraultipegma, 178, 179, 185 herwigi, Neomenia, 106 Hiatellidae, 58 Hibiscus, 253, 255, 258, 259, 301, 308, 314 Hipponix, 173 Hippophaë rhamnoides, 327 hispida, Helix, 31 hispida, Trichia, 2, 31, 32, 34, 35, 46, 49, 51 holotricha, Helix, 16 holotricha, Kokotschashvilia, 2, 18, 19, 48, 50, 53 hortensis, Cepaea, 265, 293, 318, 331, 336-338 humilis, Lymnaea, 353 hunteria, Fasciolaria lilium, 198 Hyalina avena, 198 Hydrobiidae, 354 Hydrocenidae, 208-211, 216, 217 *Hygrohelicopsis, 1,2, 5, *13, 50, 53-55 *darevskii, 1, 2, *14, 42, 48, 51, 55 Hygromia striolata, 336, 337 Hygromiidae, 1, 50 Hygromiinae, 1 “Hygromiinae,' 55 Hymenocallis declinata, 226 Hyolitha, 166, 175, 176, 186, 187 Hyolithes, 175 Hyolithida, 175 hypnorum, Aplexa, 353 Idesa charmosyne, 218 Idiopoma bengalensis, 91, 92,95 incanum, Cerion, 225, 226, 233, 234 INDEX, VOL. 17 397 Inga littorea, Littorina, 356, 357, 390 vera, 259 Littorina integra, Physa, 353 angulifera, 198 Invertebrata, 365, 366, 385, 387 irrorata, 67 iowense, Coleolus, 170, 171 littorea, 356, 357, 390 Iris ziczac, 198 foetidissima, 318 Littorinacea, 208, 210 irrorata, Littorina, 67 Littorinidae, 208 Ischyrinioida, 178 lividomaculata, Tegula, 198 Isopleura, 100, 179 Lodonaria, 204 Janthinidae, 383 Lophocardium, 58 jaspideus, Conus, 198 Lora jenkinsi, Potamopyrgus, 336, 354, 361 turricula, 383, 384 jordani, Pandalus, 377 Loxonema, 204 Kirengella, 181 Loxonematacea, 204, 205 Knightites, 199, 200 Loxoplocus, 203 Knightoconus, 169, 180, 181 lubomirskii, Plicuteria, 2, 30, 31, 45, 49, 51, 54 Kokotschashvilia, 2,5, 16, 50, 52, 55 Luquilla, 246 *eberhardi, 1,2, 17, 18, *19, 20, 49-53 luquillensis, Polydontes, 241-315 holotricha, 2, 18, 19, 48, 50, 53, 55 Lutraria, 59 makvalae, 2, 16-18, 20, 49, 50, 53 Lymnaea, 349 phaeolaema, 2, 18, 21, 22, 43, 48, 51-53 columella, 361 *tanta, 1,2, *16-18, 20, 21, 42, 49-51, 53 elodes, 353 krugi, Ormosia, 259 humilis, 353 krugiana, Clusia, 301 palustris, 353, 356 Labiosa peregra, 353, 356-360 lineata, 58 stagnalis, 353, 356, 357 Labroides, 383 truncatula, 353 Labyrinthus, 244 Lymnaeidae, 353 lactea, Marginella, 198 Lyonsia, 58 lactea, Otala, 276, 277 arenosa, 58 lacteus, Polinices, 198 Lyonsiidae, 58 lacustris, Ancylus, 354 Macluritacea, 177, 201, 202 Laevapex Maclurites, 201 fuscus, 354, 383 Macluritidae, 201 Lamellibranchiata, 171-173 Macroscenella, 181 lapidaria, Pomatiopsis, 208 Macrotheca, 176 lapillus, Nucella, 126, 143, 145, 146, 151, 152 Mactra, 58 lateralis, Gecarcinus, 228 Mactrellona, 58 latifolia, Alchornea, 301 clisia, 58 Lepidopleurus, 174 Mactridae, 58 lerneri, Cerion, 229 maculosa, Tonna, 198 leucolena, Diadumene, 65 makvalae, Kokotschashvilia, 2, 16-18, 20, 49, 50, Leucozonella, 2, 4,8, 49 53 *caria, 1,2, *12, 13, 41, 48-54 malleatus, Viviparus, 354 caryodes, 2,9, 10, 40, 48, 49, 51-54 Malletiidae, 58 ferghanica, 2,8, 40, 48, 50-54 malonei, Cerion, 225, 233 mesoleuca, 2, 48, 52 maltbiana, Trivia, 198 retteri, 2, 11-13, 41, 48, 50-54 _ mansoni, Schistosoma, 111 rubens, 2,9, 10, 40, 48, 50, 52-54 Marginella, 198 rufispira, 2, 10, 11, 41, 48, 50, 52-54 guttata, 198 Leucozonia lactea, 198 ocellata, 198 pruniosum, 198 lewisi, Polinices, 126, 140, 143, 145 marginella, Caracolus, 241-315 lilium, Fasciolaria, 198 marmoratus, Rivulus, 383 lima, Polydontes, 241-315 martensiana, Limicolaria, 291, 294 limacina, Clione, 383 Mastigospira, 203 Limicolaria Mattheva, 166, 174, 175, 186 martensiana, 291, 294 Matthevia, 174, 175, 186 Limifossor, 101 megotara, Psiloteredo, 384, 385 limosa, Amnicola, 354 Melampus linckiae, Stilifer, 383, 385 bidentatus, 354-357 lineata, Labiosa, 58 Melongena Lioplax corona, 198 choctawhatchensis, 95 melongena, 198 pilsbryi, 73-97 melongena, Melongena, 198 sulculosa, 93, 94 mercatoria, Columbella, 198 Lithidion Mercenaria forbesianum, 217 mercenaria, 152 Lithophaga mercenaria, Mercenaria, 152 lithophaga, 140, 146, 150 Mesodon lithophaga, Lithophaga, 140, 146, 150 roemeri, 296 litteratum, Cerithium, 198 thyroidus, 256, 285 398 MALACOLOGIA Mesogastropoda, 72, 157, 204, 207-210, 216, 218- 220, 365, 383, 387 mesoleuca, Leucozonella, 2, 48, 52 Metafruticicolinae, 1 Metridium senile, 63-69, 71, 72 Miconia sintenisii, 301 minima, Batillaria, 198 minirosea, Ocenebra, 198 Mitrella ocellata, 198 mittrei, Echineulima, 383, 385 Mobergella, 168, 180 Mollusca, 99-109, 165-191, 219, 360, 365-391 Monacha cantiana, 336, 337 Monachinae, 1 Monocotyledoneae, 260 Monoplacophora, 101, 106, 107, 166, 167, 169, 173, 178-184, 186 Monstrillidae, 386 moralesi, Cerion, 234 Mulinia, 59 Murchisonia, 204 Murchisoniacea, 204 Murchisoniidae, 204 Murex pomum, 198 muricatum, Vasum, 198 Muricopsis oxy tatus, 198 mus, Conus, 198 Musa cavendishii, 254 muscarum, Polymita, 290 Mya, 58 arenaria, 117 Myidae, 58 Mytilus edulis, 126, 129, 130, 132, 133, 390 nassarioides, Trypetesa, 146 Nassarius vibex, 198 Natica, 198, 200 Naticidae, 200 Nautilus, 168-170 navalis, Teredo, 383 Nematoda, 386 nemoralis, Cepaea, 265, 277, 285, 293, 294, 317- 339 Neogastropoda, 218, 383 Neomenia, 100 herwigi, 106 Neomeniida, 100 Neomenioidea, 100 Neomeniomorpha, 99-101, 103, 104, 107 Neopilina, 101, 167, 173, 179, 181, 184 Nerita fulgurans, 198 versicolor, 198 Neritacea, 208 Neritidae, 208 nitida, Nitidella, 198 Nitidella nitida, 198 nitidulum, Chaetoderma, 103-105 nivea, Crepidula, 382 Nodilittorina tuberculata, 198 norrisiarum, Crepidula, 378 Nucella lapillus, 126, 143, 145, 146, 151, 152 nucleus, Planaxis, 198 Nucula, 103, 171 annulata, 104 Nuculana, 58 pernula, 58 Nuculanidae, 58 nudum, Cerion, 225 Numenius, 60 nuttalli, Sanguinolaria, 58 ocellata, Leucozonia, 198 ocellata, Mitrella, 198 Ocenebra minirosea, 198 Ocotea leucoxylon, 301 Odontomaria, 203 Odontotrema, 2, 4, 7 diplodon, 2, 7, 40, 48-54 Oliva, 198 sayana, 198 Omphalotropidinae, 208 Onychochilus, 201 Onychochilidae, 201 onyx, Crepidula, 365-391 Opalia crenimarginata, 63, 68, 70 Ophryotrocha, 386 puerilis, 383 Opisthobranchia, 383 Oriostomatidae, 204 Ormosia krugi, 259 Orthonychia, 176 Orthothecida, 175 Otala lactea, 276, 277 oxytatus, Muricopsis, 198 Pagodispira, 203 Palaeozygopleura, 204 palustris, Lymnaea, 353, 356 Panaxia, 317 Pandalidae, 387 Pandalus, 377 jordani, 377 Panomya, 58 Panopea, 58 papillosa, Aeolidia, 64, 68, 69 Papyridea, 58 Paramya, 58 Parthenia, 246 Partula, 237, 294, 295 taeniata, 235 Paryphanta busbyi, 291 Patella, 180, 181, 383 vulgata, 267, 383 pauli, Cerion, 225, 232 Pecten, 168, 171 Pectinaria, 170 Pelagiella, 177, 180 Pelagiellacea, 166, 177, 186 Pelecypoda, 101, 167, 168, 170-173, 184, 186 peltata, Cecropia, 255, 301 Penetrantia densa, 146 peregra, Lymnaea, 353, 356-360 pernula, Nuculana, 58 Petaloconchus, 196 Petasina, 2, 39, 53 unidentata, 2, 31, 32, 34, 45, 49 phaeolaema, Helix, 21 phaeolaema, Kokotschashvilia, 2, 18, 21, 22, 43, 48, 51-53 INDEX, VOL. 17 399 phoebia, Astraea, 198 Porcellia, 203 Pholadidae, 58 Porcelliidae, 202 Phyllomenia, 107, 108 Poromya, 58 Physa Poromyidae, 58 acuta, 353 portoricensis, Platysuccinca, 279, 311 fontinalis, 353, 356 Potamopyrgus gyrina, 353, 356 jenkinsi, 336, 354, 361 integra, 353 Poteriinae, 210, 218 virgata, 353 Probivalvia, 173 Physidae, 353 Prochaetoderma, 103, 104, 106 pica, Cittarium, 198 profunda, Allogona, 256, 285 pillsburyi, Cerion, 229 prolifera, Gonactinia, 64-66, 68, 69 pilsbryi, Lioplax, 73-97 Proneomenia, 103, 108 pisana, Theba, 227-237 Prosobranchia, 72, 207, 218, 219, 351, 354, 355, Pisania 358, 361 auritula, 198 Proterostomia, 183 Placophora, 101, 184 Protowenella, 178 Plagioglypta, 170 pruniosum, Marginella, 198 plana, Crepidula, 365-391 Psammobia plana, Scrobicularia, 117, 118, 123, 124 vespertina, 118 Planaxis pseudoobscura, Drosophila, 317 nucleus, 198 Pseudophoridae, 205 Planorbidae, 353 Psiloteredo Planorbis, 195 megotara, 384, 385 albus, 353 Psychotria carinatus, 353 berteroiana, 301 contortus, 353, 356-360 Pteropoda, 383 corneus, 268, 353 Ptomatis, 199, 201 planorbis, 353 puerilis, Ophryotrocha, 383 vortex, 353 Pulmonata, 1, 223, 243, 341, 351-355, 360, 361, planorbis, Planorbis, 353 383 Platyceras, 176 Pupininae, 210, 218 Platyceratacea, 203 Rabdotus Platysuccinea schiedeanus, 275 portoricensis, 279, 311 racemiflora, Cyrilla, 255, 301 plebeia, Trichia, 2, 12, 32-34, 46, 49, 51 radiata, Egeria, 118 Plectronoceras, 169, 181 ramiflora, Guarea, 301 Plethospira, 204 Rangia, 59 Plethospiridae, 204 Ranunculus, 329 Pleuroceridae, 157 Raphistomatinae, 202 Pleurodonte, 244, 246 regeliana, ““Euomphalia,” 9, 41 Pleurodontidae, 243 regius, Conus, 198 Pleurodontites, 246 rengarteni, Caucasigena, 2, 23, 25-27, 43, 44, 48, Pleurotomaria, 202 51, 55 Pleurotomariacea, 177, 202-204 rengarteni, Helix, 25 *Plicuteria, 1,2, 6, *30, 50, 53 rengarteni, “Helix,” 25 lubomirskii, 2, 30, 31, 45, 49, 51, 54 reticulatum, Deroceras, 341-350 poecilanthus, Croton, 301 retteri, Leucozonella, 2, 11-13, 41, 48, 50-54 Polinices rhamnoides, Hippophaé, 327 duplicatus, 198 Rhaphischisma, 203 Ic. 245, 198 Ribeirioidea, 178 /ewisi, 126, 140, 143, 145 Rissoacea, 208 Polychaeta, 383 rivularis, Ferrissia, 354, 356 Polydontes, 241-315 Rivulus acutangula, 241-315 marmoratus, 383 lima, 241-315 Robillardia luquillensis, 241-315 cernica, 383 Polylopia, 171 roemeri, Mesodon, 296 Polymita Rostroconchia, 107, 166, 178, 179, 182-184, 186, muscarum, 290 187 Polyplacophora, 100, 101, 107, 166, 167, 174, Rotifera, 386 179-181, 184-186 rotundatus, Anisus, 353 Polystira rubens, Helix, 8 albida, 198 rubens, Leucozonella, 2,9, 10, 40, 48, 50, 52-54 pomatia, Helix, 92, 237, 254, 256 Rubus, 329 Pomatiasidae, 208-212, 214, 216, 217, 219, 220 rufibranchialis, Coryphella, 69 Pomatiasinae, 210, 216, 217 rufispira, Leucozonella, 2, 10, 11, 41, 48, 50, Pomatiopsidae, 208 52-54 Pomatiopsis rufum, Campeloma, 93-95, 354 lapidaria, 208 Rumina Pomatoceros decollata, 233, 237 triqueter, 383 rupicola, Epitonium, 67 pomum, Murex, 198 rustica, Thais, 198 400 MALACOLOGIA Sagdidae, 311 Salterella, 169 Sanguinolaria, 58 nuttalli, 58 Sanguinolariidae, 58 Saxidomus, 58, 59 sayana, Oliva, 198 Scalaeotrochus, 205 Scalaria subulata, 68 Scalidae, 383 Scalites, 202 Scaphopoda, 101, 166, 167, 169-172, 174, 179, 184, 186 Scenella, 180, 183, 185 Scenellacea, 166 schaposchnikovi, Caucasigena, 2, 14, 27, 28, 44, 49, 51 schiedeanus, Rabdotus, 275 Schistosoma mansoni, 111 Scrobicularia Diana Пу. 18, 123, 124 scutaria, Fungia, 70 Scutopus, 103 Semelidae, 58 Semitrochatella elongata, 218 senile, Metridium, 63-69, 71, 72 Sepia, 168 septentrionalis, Trichia, 32 “sericea,” Trichia, 32, 33 Siliqua, 58 Siliquaria, 196 sintenisii, Miconia, 301 Solecurtus, 58 Solemya, 58 Solemyidae, 58 Solen, 58 Solenidae, 58 Soleniscus, 205 Solenogastres, 99-101, 167, 174 Solinoconcha, 169, 171 Sphenia, 58 Spirula, 168 Spisula, 58, 59 dolabriformis, 58 spurca, Cypraea, 198 squamulosum, Calycogonium, 301 stagnalis, Lymnaea, 353, 356, 357 stearnsi, Conus jaspideus, 198 stella, Bunodactis, 65 Stenothecoida, 166, 173, 174, 176, 186 Stenothecoides, 173, 175, 180, 186 stevensoni, Cerion, 224-226, 233 Stilifer linckiae, 383, 385 Stoichactis helianthus, 70 Stomphia coccinea, 64-67, 69 straminea, Biomphalaria, 111-115 striatellum, Cerion, 235 striolata, Crepidula, 378 striolata, Hygromia, 336, 337 striolata, Trichia, 2,37, 39, 47, 49-53 Strombus, 196 Strophiops, 226 Styliolinidae, 168 Stylommatophora, 207-209, 211, 214, 216-221, 243, 341 subdiscoidea, Cyclotopsis, 217 subulata, Scalaria, 68 Subulitacea, 205 Subulites, 205 sulculosa, Lioplax, 93, 94 sumatrensis, Bellamya, 91, 92 sylvatica, Cepaea, 336 taeniata, Partula, 235 Tagelus, 58, 60 Tannuella, 169 tannum, Campeloma, 93, 94 *tanta, Kokotschashvilia, 1,2, *16, 17, 18, 20, 21, 42, 49-51, 53 Tealia felina, 64-66, 68, 69 *Teberdinia, 1, 2,5, *15, 50, 52,53, 55 zolotarevi, 2, 15, 42, 48, 51 tecta, Astraea, 198 Tegula fasciata, 198 lividomaculata, 198 Tellina tenuis, 122 Tellinacea, 117, 118, 122 tentaculata, Bithynia, 354, 356, 357 Tentaculitidae, 168 tenuis, Tellina, 122 tenuitesta, Helix phaeolaema var., 21 Terebridae, 65 Teredinidae, 384 Teredo, 383 navalis, 383 testiculus, Cypraecassis, 198 testudinalis, Acmaea, 103 Tetrahymena, 341 Thais, 204 deltoidea, 198 rustica, 198 Thalassia, 58, 59 Thalassoma bifasciatum, 383, 387 thalestris, Caucasigena, 2, 29, 45, 48, 51 Theba pisana, 227, 237 Thraciidae, 58 Thyonella gemmata, 157 thyroidus, Mesodon, 256, 285 tinctum, Epitonium, 65, 68, 70 Tindaria, 58 Tommotiidae, 168 Tonna maculosa, 198 Tresus, 58 Trichia, 1-56 bielzi, 47 concinna, 2, 32, 33, 35, 46, 49, 51 danubialis, 2, 38, 39, 49-53 edentula, 47 filicina, 31, 47 hispida, 2, 31, 32, 34, 35, 46, 49, 51 plebeia, 2, 12, 32-34, 46, 49, 51 septentrionalis, 32 “sericea,” 32, 33 striolata, 2,37, 39, 47, 49-53 unidentata, 2, 31, 32, 34, 45, 49 villosula, 2, 36, 47, 49 Trichiinae, 1, 13, 55 Trichotropis, 383 Tricladida, 357 triqueter, Pomatoceros, 383 Trivia maltbiana, 198 trivolvis, Helisoma, 353, 356, 357 Trochacea, 203 Trochina, 203, 204 Tropidodiscus, 199, 200, 203 Troschelviana chrysochasma, 218 Truncatellidae, 208 truncatula, Lymnaea, 353 trunculus, Donax, 117-124 Tryblidiida, 101 Tryblidium, 179, 180 Trypetesa nassarioides, 146 tschetschenica, Caucasigena, 2, 24, 43, 48 tuberculata, Nodilittorina, 198 Tubinidae, 203 tulipa, Fasciolaria, 198 Turbellaria, 108 Turbinella angulata, 198 Turbo canaliculatus, 198 Turdus ericetorum, 327 turnerae, Cerion, 226 turricula, Lora, 383, 384 Turridae, 383 Turritellidae, 65 Tylozone, 202, 203 ulu, Epitonium, 68, 70 Umbelliferae, 329 Umbonis, 226, 233 undatum, Buccinum, 60 unidentata, Helix, 31 unidentata, Trichia, 2, 31, 32, 34, 45, 49 Urosalpinx cinerea, 68, 125-146, 153 follyensis, 125-142 Urtica dioica, 331, 332 uva, Cerion, 225, 226, 230 Valvata humeralis, 354 piscinalis, 354 Valvatidae, 354 Vasum muricatum, 198 INDEX, VOL. 17 Veneridae, 58 Ventroplicida, 99-101 vera, Inga, 259 “Vermes,” 175 Vermicularia, 196 verrilli, Actinauge, 65 versicolor, Nerita, 198 vespertina, Psammobia, 118 Vexillum dermestinum, 198 vibex, Nassarius, 198 villosula, Trichia, 2, 36, 47, 49 vindobonensis, Cepaea, 336 virgata, Physa, 353 virginica, Crassostrea, 126, 130 vittatus, Donax, 117, 118 Viviparidae, 91, 354 Viviparus contectoides, 93-95, 354 contectus, 91, 92 fasciatus, 92,94, 354 georgianus, 73-97 malleatus, 354 viviparus, 92, 94, 354 viviparus, Viviparus, 92,94, 354 Volborthella, 169, 171 vortex, Planorbis, 353 vulgata, Patella, 267, 383 Watsonella, 178 williamsi, Crepidula, 378 willistoni, Drosophila, 161 xanthogrammica, Anthopleura, 63, 70 Xenoconchia, 171, 176 Xenophora, 196 Xerocampylaea zelebori, 45, 54, 55 Yochelcionella, 182 Zachrysia, 244, 246 auricoma, 244 havanensis, 244 zelebori, Xerocampylaea, 45, 54, 55 ziezac, Littorina, 198 zolotarevi, Teberdinia, 2, 15, 42, 48, 51 Zostera, 376 401 f A wa 2 My) Aare pe т 5 И. 1-Х: адом TA lira rro 52,10 wann ET il 2325 № $9 purine Gr at) ars сы Ba, TT) 11а, 122, 10a. BBE area ра ре OF лол ‚СВ ren У sven ICI „Bar эм M Dentro PAE done ON clones, 213 ESE TOS de ing sente Mierridiam (3-80.71 21 BIT weno Spa, 108 Ave subis Ô onu ia povidone, Trié, (Y воно no iii cums. Pura, 70 Bestes 107. af »e ar i ‘ta rica,” POSE moon мисс ns Foot est PTT tano KK Nidda 100 | ger smoAganey item, Micoife 301 nen, са ма ак лба CAT vera‘ SORTE, E Eoirrvyon, € И =. GAL Plo wanda an E BSS oriole foie, 2 PEZ tune Miles 710838 13 2» creel iteragoews vod лотос. 98-101; 107, BBP sam) Seal ainia Ta ОСИ Zit Ss ieruiclos Comes iia, бе BIE uno Авин, 164 te EN, 55 Arr буи. 98 E # pra, VUE avatars. Clyro м 30% apela, Lirrers, SS, 350 (eo TE e Lis 4 un armer u 1 56 $ 74.1 в pits ‘ ? ie 9 Me | Se. “iv 2083 пор WOR à at 300 Dekan? УПО à №. an OR ne ~ ws O TS | ту ae 4 buh? y EVA. CADA TO A7! yn E A Y 4%; ueno VENTA 57. м «та MIND if sera dit ‘viobliniGes,” 158; Ayvlonusrdisohosa, УП? See. 9% Fu 3 FU A ‘see te Cr Ar ‘à № Zum OS о Зе: SP re, | 7 ientarylote, Тонко и Мак. 168 ni. Tellite 122 tan *pagts, (edie Almen Terence, BS ' Teredaicina,. FEA or, € Trait 383 man, ES raté, Сурчасстай, cr ‚Als nn 15 cet >> url 125 h de Са US { TO 2777 LEE FRE Pam aus = wee acs o | fen Ones gers, 2,75, ER Г. За 77 MES A Moss, 36$ „As PLUS rond, E 68, 20 EEE УФ К >“ та rmatulto ae Trews, > Irena J Y “Us ws $? concirua, 2, 10,35. AB, UE, Y Paru: pid à 9% I a 27 4 a NT Fe, 34. 35, Mi, 40,81. (oy See "234, au, 43, %1 y PON OA ta . were, . spite 4.34. 0 47 496) Uri ¿DY AN 32 yA Ab. 49 vitale, 7, 36,41, В Me, A Trzhuwepy DES | > Trclactida, 207 trequater, Ярачн оке гов, 383 in + ronds 125 р ans iyi aria, 353, 366, 367; lroctnsss, 201 PR UN ek , ae" м py if | TA INN ГА re he к, № ere oo A и Ar … sl 4 nt y к MOR whens > Br big! Fu И TT CENT Fan à VUS CR NTU CE ARNO A eT Pere a ре val BR Ae \ т AU A4 f К Niet, | ee Gee ade wie Gl AP ole ot 4 | ON LE atc een ene de ie Sa ee Mer PLANE itd Or tm и 1 i } ys bi wt | Tr f L у taf j 4 N ] \ , 4 А i > м 4 А A r Si PV a “a 1 ï u у \ NA ae vie) TA M и р | he À Er ni i | | al Da / EN у y Mr mn Cale > у | 6 4 | dé er EL TETE Ma ; | Г AE у cet a een м Er ao 2 eg BL TE 20) В q O и e hat? № ah фм он | A О PS au LEE OT RL D DU AA AT Me e 205 PAUL ОА DOVE PAL EN LP" NU | Bigs, F4 "0 CRU ARRET nt : ara Big Les er = | 4 A 44 Ai Г atk | = р К "is NA MT if | Е Е ear. nd у К ча" ae LA Ne Kal NP by f wow mig Un AL re - il Ч o f a | y LU “y A» ir ih y » 1 CIM } i 13 tl Y Ar de pe) ri № 4 т 1 - D : À | y + i ? ñ A f a h e k x и и Y pe } y 4 ‘ 4 f i aight A { Wi $ в 17(2) INSTRUCTIONS FOR AUTHORS MALACOLOGIA publishes original studies on the Mollusca that are of inter- national interest and are of high scholarly standards. Both descriptive and experi- mental research results are acceptable pro- vided they are primarily or exclusively concerned with the phylum. Contributions include long monographs as well as moder- ately short research papers. Brief papers are not acceptable. MALACOLOGIA provides a forum for such different aspects of malacology as anatomy, comparative physiology, ecology, medical malacology, paleontology and systematics. Papers of only biochemical or physiological interest should be submitted elsewhere. Review articles are more appropriately submitted to Malacological Review (P.O. Box 801, Whitmore Lake, Michigan 48189, U.S.A.). All manuscripts submitted are reviewed by at least 2 malacofogists. Articles are accepted with the firm understanding that they have not been submitted or published elsewhere in whole or in part. Manuscripts may be in English, French, German or Spanish, and should follow MALACOLOGIA style. They must contain a concise but informative Abstract sum- marizing not only the content but the results. Papers in languages other than English should include a translation of the Abstract into English. Authors desiring their abstracts translated into other languages must provide these. Care should be taken to include all necessary foreign accents. Manuscripts must be typed on one side of good quality white paper, double- spaced throughout, with ample margins, and are to be submitted in triplicate. Illustrations are likewise to be in triplicate (the 2 copies may be photocopies, etc.). Tables, figure captions and all footnotes are to be grouped (in this order) at the end of a manuscript, and all Ms pages (including the Abstract) are to be numbered $е- quentially. Avoid internal page references (which have to be added in page proof). Make the hierarchy of headings within the text simple and consistent. Suggest an ab- breviated running title to be used at the top of each right hand page. Contributors in English are asked to use the Council of Biology Editors (CBE) Style Manual (Ed. 3, 1972), obtainable for $6.00 from the American Institute of Biological MALACOLOGIA 1978 Sciences, 1401 Wilson Boulevard, Arling- ton, Virginia 22209, U.S.A. MALA- COLOGIA follows most of the recommend- ations in this Manual. In particular, simplified practices such as the following are used: numbers above ten should not be written out except at the beginning of a sentence; percentages following a number are expressed as %, and abbreviations of measures (after a number): mm, ml, kg, etc. have no period (full stop), nor ап “s”” in the plural. Note that the international symbol for micron is now um, not u. Illustrations must be carefully prepared and so planned that they can be printed in 1 column or the full width of a page of the journal. The maximum size of a printed figure is 13.5 X 20.0 cm (preferably not as high as this so that the caption does not have to be on the opposite page). Drawings and lettering must be in dark black on white, blue tracing, or blue-lined paper. Lines and dots should be thick enough to allow reduction by 1/2 or 1/3. This should be taken into consideration also in relation to the lettering. Letters and numbers must not be less than 2 mm in height, preferably larger, after reduction. Several drawings or photographs may be grouped together to fit a page, but drawings are not to be grouped with photographs. Photographs are to be glossy and high contrast. All illustrations are to be numbered sequentially as figures (not grouped as plates), and are to be arranged as closely as possible to the order in which they are first cited in the text. (Each figure must be cited in the text.) All original illustrations should be mounted, numbered, labeled or lettered and ready for the engraver. Scale lines are required for all figures and should be convenient lengths (e.g., ‘200 um’’, not “163 um””). Magnifications in captions are not acceptable, and neither are photo- graphic reductions of line drawings. Captions should summarize what is shown in an illustration, and should not duplicate additional information given in the text. Each lettered abbreviation labeling an individual feature in a figure must either be explained in each caption (listed alpha- betically), or be grouped in one alphabetic sequence in a section near the beginning of the text (use the latter method if many abbreviations are repeated on different figures). Tables are to be used sparingly, and 4 у we | Vu р | } м n ; “ ; AS VITE 4 a a EIN ” у ne. oy р td 7 + AF | ant ta £a as | FD ee u eg page. De AE TR PAU D NU AIT Referenoss cite ie (a eu text, i py not nt fines pes on to the volume meh | + ‘wat ? ne ALES PA i Kay AN Vol. 17, №. 2 MALACOLOGIA 1978 CONTENTS AMERICAN MALACOLOGICAL UNION - SYSTEMATICS ASSOCIATION SYMPOSIUM PROCEEDINGS: EVOLUTION AND ADAPTIVE RADIATION OF MOLLUSCA; 12-13 JULY, 1977, NAPLES, FLORIDA G. M. DAVIS ОСНО EL: ae А O PEN gm O 163 . L. YOCHELSON An alternative approach to the interpretation of the phylogeny а Ми < AMI METIA RE AT 165 В. М. LINSLEY Locomotion rates and shell form in the Gastropoda ............<.... 193 A. J. CAIN The deployment of operculute land snails in relation to shape AC ВР > aoe ach eM ea Ri, И 207 D. S. WOODRUFF Evolution and adaptive radiation of Cerion: a remarkably di- m verse group of West Indian land snails...... 2... eee ee ee ee ee 223 H. HEATWOLE and A. HEATWOLE Ecology of the Puerto Rican camaenid tree-snails ...........,,,.,.. 241 J. MURRAY and B. CLARKE Changes of gene frequency in Cepaea nemoralis over fifty WSS O ONE о Ин Sb AN LE NE RCA 4 317 С. $. OXFORD The nature and distribution of food-induced esterases in NP PA (Mu O AN ал TA 331 М. М. RUNHAM Reproduction and its control in Deroceras reticulatum .....::..,..:,..: 341 Р. CALOW ; The evolution of life-cycle strategies in fresh-water gastropods .......... 351 К. Е. HOAGLAND Protandry and the evolution of environmentally-mediated sex changes a-stud Vink the MOLUSCH.. ann eae ET RAD HER ER 365 INDEX TO) VOLUME: 47) NOS. NE. ADO 2 AN ое 393 Acme Bookbinding Co., Inc. 100 Cambridge St. Charlestown, MA 02129 3 2044 072 160 427 bade He р da EL ЕО ПАНА Е ГА aid PMD verts à POULE ue 5 REP р pins d'à ry ates Иа COMPTE PAT ET ET TT VERS RU Ne ere pren pie # À 2 BI AENA Ud AO aa EITC o YA a gt E а a fh Mar ich ocd R O any a НИ АНН НЫ Gi rhe ny pea CENT dé hy HONTE AN HS веки, АТА MUR PU RD: SA OW TA DO CIA ES LAA РГ «+ varias “te En NEE ле Ue ELE IRIS IAE AN Aus ge PIE RETA НИ Pago i РИ, чи мин иен nio 4 Y) Bar ARA IA AA CET A TETE “ COUR LW Pig nib web LAC ETF ER + CPV eT Oe Oe tis Me a ere 11 CRT HA SE Ca nara: rd 6k abe pib ee iy OF ed Meee eee CNM учр, eit O ses bs be ons ER ARE aie) NE wa CAVE] , MEN Е Веру HE a ws : ира ее tas При бер wang Bay ‘ PE OER e DOTE ES ee TY ER WERE ETES АОН РОЙ. ветры ed oe ee 77 EI ET MENO AE ва SIR ENT НА lung dr» wii RE te Klee Oak COREA IRA ie BT DE ONO POTES Lips à Shed à PA TPE ET ONE рн RL De ke PTE ней aes eg Pa i ay PERO CO TL LE LEER I) NAAA Later Fame ITA PAU ES HS АТ APCE EEE UTP IR EE PTE TE de cul CAN PET TER AA IA ER Е АИ at Ре] ted RR AR A DRE COMITE LT ал ET AM ty Seth Ве лы SNOB EUAN PA Мм UNE LEE AMET ENS Ws ey. PRE sé HART RAM a BIS Un неа, APE a N Pm neun ENTE ILL Din ANNEE TO res TIT ARO PARENTS АА 6 к ВС SORT RS hy а а un Вам, uA CRETE HF AO 4g eng? НОА y ‘ a TRE io APU ap IBAN RUNTER se Ran, оо y ы O AL SATE cure a pd ды Aly bs a bon guy NETTER NS, DIENEN RER EA NAPO CA A a UA Launch Ow LASER RR NR Se cerns BAN AS AWE OA IO AY thy ste DE NC ан Mig Ma RTE Ok POT NOU Mee bg ans + NAT Sn ое жж: AZ NEE Ter eye sant SP OA ел TER SOU A TAT RUNS ами, ST NRZ qm A N an ASAS e AO ICAA NT har CA ль a, BRU STIER SERIE UA 0 yo eh PR RE ROUTE Met: E A CA EA Yo AA A CT EE RON ПВА CARRIERE hen area te ne warm ia La sur LR Ms 0 ля ` en Pe : т Пе, чье ан: ам и БА NE EE oh à AEE KIC Y NM AES aol, TARN а NETT RN nr N VRR CURA SOK cee Baie наи, о SEITE AS RR Ах ААА А А АТ WON ани а А tr ARE PLA ETES PTE Ee ча > VANS? АС А а CA M AEN ты Gwe С NM, NUD ре ки A АКН а ле a мо мета, EL A ARA ie Seda И 2 м Kran ne, VE Re 44 in ah AWA ER ма, а БЛАН CRE м AA QUE Фуа ES TN QE UNSEREN Ur Ач REE A ant а week ee CHG м дб: LA ame вы AAA DCR PET LL м чо им BQ ve do; У АЕ ann SAUNA ых ЗА LA MU tua rt wor EN Kinn АС wen ate ATEN AN ке Ааа bh ль АДА БА За LS BWA RTS" iba teeth bad ms RENTERES PWM Ae ieee beard LUN RARES ORNE SECO EU A ULA A RU ATE ae AA LR IRON o A A LR NA AO O ан EN Ma уу "уч ЕЮ ААА) ЗАЛАМ CAN ty NA € e N чин КУН, uw, Ки RSL ER RARE RUE ET A ES N. м НА RTE) à LORS чм, NR Lacio Umi ad a TES BLO MMR Ea e QT a e isa Moa rane A era Y AA A yy Et ROR +4 sde EUR we O НУ Ar RE EEE ag аа ENTRE ла at esa hy лая я me ая Аа FAS AL Roa he nr NT EIN HS EN N y ds ана Ильи ASS ERG, DL DE СЕН, ÍA SAA DE O A on esse AA AA N Мяу NT A 7 OR A ETES k ER EEE NORA TE US a AT OURS ALES l'a $. DEA АН АНА Ni АСОИ EN DON A А ны мона р лм ee WM Ar AAN nem MN EA REF AUS EP U RN Ar, Shih Au ua и web une О CDTI ME AS я nh АНА Ч; м RSR RON CONTENTER va НУ иен на Se NOUS Are ra РАЗЫ OA A a a Aa A AA ee ET EE EN АЗИИ Mea aA ug Re М a A AA a | HEC UPON. AS AAN WA а ик И RE DA D AR BE НН CP Mi wo, SA AMIR RR UR arn e Аа “si A RAT y tut ААА Ул, Vw ated boy ‘ VA iy Our CENTS MERE MEE Lt AS he RT AS oe rt N IR С А N ль gere mn 1:4 fy и ee bray Sehr Has tite TX CAC un WAS ee ‘ > y ма In tv wat, MAS | te bari Аа №, А > My A К an vn cant N Ar A wohin ad ed ENT LS x N 4 ON] Vapor da i и WAR an QUE Наль ума wo salary RE 04 Ty ANA % 4. A nm ~ hy Ww, A Upstate LAVE