DUREE ВАД Eu st dao adela fee AE Se Ya paa ta in ed a da or HARVARD UNIVERSITY e Library of the Museum of Comparative Zoology и с, CA DRE № EA MW A vay р у a si a | у 5 4 ONE PAU р fit р iy hs ON ya DS AA ON ey LIBR ARY VOL. 36, no. 5 pe 1995 nl Al 7 A N 0 ; 9 E | ) 3 N h 1800 } Al ий WAALS Fle КМ $ > : N Ar YU J NY I \_ i | x р HA YA E © VB LR > D] у | u i UNIVERSIT ae - 2 р ы o | À | и ue г vi о | Е a | E ? м 1 | | \ A Е / / | EN \ pa \ | A] ly ? | = | aff | A A ~ <= \ \ y | к С р TS à a . x a À Г / 1) у “| dd EL ] + Yul À y | Е т À \ ve if | Pas | 5 a \ x i 4 HQE Sr ANS / N / | E eae, ja j ur / f ух À Aa [ i y> ) { f \ 1 \ N a { >= | ` Г г qe ;: а к 7 2 : = 1 N й а E j — ER j à % г \ au в Ш 5 f hy ni Sp y yr - * 4 y С 4 UT Al of | | р) | Ng À у x \ 4 IR ~ Br BR va = ES en A) ; ; mn x \ po > г — | Pi < _ ¢ SER 7 > à = 1 ‘aa y 1) a j | dat = N ` ) A (Ry Ar d ae fr | 5 J Г \ 2 Na + aly a u al Journal of Malacology — хм A A | y) ur Eee à | 1 , ? ] 4 ( : ия и Г oa 7 ы A и Е EN i IT \ 215 Ya À a th D j 4 = me à | N > Kir п q Г \ Intern nacio onal de Malacologia AAA р h \ | + 3 RE Ps e £ Er 29 = > 7 7 р | = A 4 в er ) к \ р. Jou urnal International ‚der Malacologie A и a LA N Es Ау = } \ Loin y м Pare Ma PA | Y _ Журнал. Manaxo norm — ыы ИК" i hy {| $ Yo A u Eu LL O Ka Pay > à 3 > У | ан в. NUE ine e a. i) t Sy Ph Pan у pn) я > 7 dort RESP A NE MO “In nternationale Malakologische Zeitschr ft fd Л | >, | ce К ne ts | i : x F Y 2 a ie | x н | RUN ER и Rn Ye ERDE AU 3 Qe Wey ey / INT UFER ERE A WY a > Sy | ий м O: Le ut ¡MAR r у в“ Ye I; Ey ; JR Mes E SU Y | L, у , an AN ae 1 м А К | N ) A | po a у he trek thee L » Te = qe i Cs > A ue 2 EAS 4) ( \ 4 M? у ид a Bo 00 | AN MALACOLOGIA ^ | и y © h Editor-in-Chief: | | N: inh \ ме В ca) GEORGE M. DAVIS 2 | y y Editorial and Subscription Offices: М И ah Ate, is 2 | : Department of Malacology ae E 117 NÉ _ The Academy of Natural Sciences of Philadelphia Bau. , 1900 Benjamin Franklin Parkway. ART Le _ Philadelphia, Pennsylvania 19103-1195, U.S.A. ‹ CAL à rt ( 4 i ] ex у №, ya / ‘Co-Editors: | E И г | EUGENE COAN м € CAROL JONES — и California Academy of Sciences. Denver, CO * MA STE San Francisco, CA x N A Va ane Г a - Assistant Managing Editor: > mf ; Go ; и "| CARYL НЕВТЕВМАМ ^ | Pe a Associate Editors: 5% à : > ~ 1 es a = os Ê er PRE + > ANNE GISMANN — ey University of Michigan Maadi i Ann. Arbor . A | Egypt >.) a ae E ay A MALACOLOGIA a ey the INSTITUTE OF. MALACOLOGY, fe Sponsor M | ni which EEE serving a as ai are: Ar | KENNETH u. BOSS | eh “JAMES NYBAKKEN _ a №. Museum of Comparative Zoology Moss Landing Marine Laboratory aay Cambridge, Massachusetts California | Ум I JOHN BURCH, President ; CLYDE = E. ROPER, as 7 > \ > . À Le i Rat PN. MELBOURNE R.GARRIKER | 12%", Washington, a ^_ Ne University of Delaware, Lewes | | 9 41 W. D. RUSSELL- HUNTER LEN “GEORGE M. DAVIS | | Y и Sec retary and Treasurer | Syracuse University, New York — $ | MEET SHI-KUEI WU ; = sy MN д | CAROLE 3. НСКМАМ pa ulder ¿University of MEN р, e University ae Museum, Bo т | Ales nt | President- Elect г << rie ALA 2 : Oe №. ( \ с | / т AUS IO ДИ Fa) у и _ Participating Members | | | E MUND GITTENBERGER JACKIE L. VAN o y Secretary, UNITAS MALACOLOGICA | _ Treasurer, UNITAS MALACOLOG IC À by Rijksmuseum van Natuurlike Op LS -Koninklijk Belgisch Instituut _ x mak CPE sie _ Historie bins д“ 1 voor D af KK. LEO Re Leiden, Netherlands | ме Brussel, Belgium — AR Y PAN gt, Y VA | 4 AR Lee > IE he N, IR sf e EN A Emeritus Members | LA Mae Harty J. FRANCIS ALLEN, Emerita Pars à | ROBERT ROBERTSON _ | eN Environmental Protection Agency ane The Academy of À Natural Scienct 4] NE D.C. у й el EP Hope Pennsylvania | 1% b cs Al Mus 5 Y de | Le Wir 7 pk и 1 % y | я 4 | 5 — - ы - - . — - я Bin a u a 3 a ¡O ее IO ae a A АЕ qi, vd < 7 Y Copyright O 1995 by the Institute of Malacology. y SIN) à \ Rz) i Bu nn Ма E > 2 IR e A En, O rt У ес A 4 A Ved hi ; Pr taire À W р AL ут АД N Behe | ри KART DE PN A4 ony } Kr ae} 0 d Hl 4 и / + } ) 4 ‘à A FUGA A MIDA 5 eN 1995 EDITORIAL BOARD J. A. ALLEN Marine Biological Station Millport, United Kingdom R. BIELER Field Museum Chicago, U.S.A. E. E. BINDER Museum d’Histoire Naturelle Geneve, Switzerland A. J. CAIN University of Liverpool United Kingdom P. CALOW University of Sheffield United Kingdom J. G. CARTER University of North Carolina Chapel Hill, U.S.A. R. COWIE Bishop Museum Honolulu, HI., U.S.A. A. H. CLARKE, Jr. Portland, Texas, U.S.A. В. С. CLARKE University of Nottingham United Kingdom R. DILLON College of Charleston SC, U.S.A. С. J. DUNCAN University of Liverpool United Kingdom О. J. EERNISSE University of Michigan Ann Arbor, U.S.A. E. GITTENBERGER Rijksmuseum van Natuurlijke Historie Leiden, Netherlands F. GIUSTI Universita di Siena, Italy A. N. GOLIKOV Zoological Institute St. Petersburg, Russia $. J. GOULD Harvard University Cambridge, Mass., U.S.A. A. V. GROSSU Universitatea Bucuresti Romania T. HABE Tokai University Shimizu, Japan R. HANLON Marine Biomedical Institute Galveston, Texas, U.S.A. J. A. HENDRICKSON, Jr. Academy of Natural Sciences Philadelphia, PA, U.S.A. D. M. HILLIS University of Texas Austin, U.S.A. K. E. HOAGLAND Association of Systematics Collections Washington, DC, U.S.A. B. HUBENDICK Naturhistoriska Museet Góteborg, Sweden S. HUNT Lancashire United Kingdom R. JANSSEN Forschungsinstitut Senckenberg, Frankfurt am Main, Germany В. М. KILBURN Natal Museum Pietermaritzburg, South Africa М. A. KLAPPENBACH Museo Nacional de Historia Natural Montevideo, Uruguay J. KNUDSEN Zoologisk Institut & Museum Kobenhavn, Denmark А. J. KOHN University of Washington Seattle, U.S.A. A. LUCAS Faculté des Sciences Brest, France C. MEIER-BROOK Tropenmedizinisches Institut Túbingen, Germany H. K. MIENIS Hebrew University of Jerusalem Israel J. E. MORTON The University Auckland, New Zealand J. J. MURRAY, Jr. University of Virginia Charlottesville, U.S.A. R. NATARAJAN Marine Biological Station Porto Novo, India J. OKLAND University of Oslo Norway T. OKUTANI University of Fisheries Tokyo, Japan W. L. PARAENSE Instituto Oswaldo Cruz, Rio de Janeiro Brazil J. J. PARODIZ Carnegie Museum Pittsburgh, U.S.A. J. P. POINTER Ecole Pratique des Hautes Etudes Perpignan Cedex, France W. Е. PONDER Australian Museum Sydney SI ENG Academia Sinica Qingdao, People’s Republic of China D. G. REID The Natural History Museum London, United Kingdom N. W. RUNHAM University College of North Wales Bangor, United Kingdom S. G. SEGERSTRLE Institute of Marine Research Helsinki, Finland A. STANCZYKOWSKA Siedlce, Poland F. STARMÜHLNER Zoologisches Institut der Universität Wien, Austria У. |. STAROBOGATOV Zoological Institute St. Petersburg, Russia W. STREIFF Université de Caen France J. STUARDO Universidad de Chile Valparaiso S. TILLIER Museum National d’Histoire Naturelle Paris, France В. D. TURNER Harvard University Cambridge, Mass., U.S.A. J.A.M. VAN DEN BIGGELAAR University of Utrecht The Netherlands J. А. van EEDEN Potchefstroom University South Africa N. H. VERDONK Rijksuniversiteit Utrecht, Netherlands B. R. WILSON Dept. Conservation and Land Management Kallaroo, Western Australia H. ZEISSLER Leipzig, Germany A. ZILCH Forschungsinstitut Senckenberg Frankfurt am Main, Germany MALACOLOGIA, 1995, 36(1-2): 1-14 MORFOLOGÍA DEL ESTOMAGO Y PARTES BLANDAS EN MYTELLA STRIGATA (HANLEY, 1843) (BIVALVIA: MYTILIDAE) María Villarroel* y José Stuardo* ABSTRACT The anatomy of Mytella strigata, a species found in lagoons on the Pacific coast of central Mexico, is compared with that of the М. charruana of the Atlantic and of other marine Mytilidae, giving particular emphasis to the morphology of the stomach. The siphons belong to type A (Yonge, 1948), the ctenidia to type B(1) (Atkins, 1937), and the stomach to type Ш (Purchon, 1957) with sorting mechanisms of type B (Reid, 1965). Possible relationships between mor- phological adaptations and life habits are discussed. Key words: stomach morfology, Bivalvia, Mytilidae, Mytella strigata, anatomy. INTRODUCCIÓN Muchos de los 23 géneros de la familia Mytilidae (Soot-Ryen, 1955) han sido estudia- dos desde el punto de vista anatómico fun- cional, entre ellos, Mytilus (White, 1937; Owen, 1974), Botula y Lithophaga (Yonge, 1955), Musculus (Merril 4 Turner, 1963), Xe- nostrobus (Wilson, 1967), Ааша (Вошла) (Fankboner, 1971), Limnoperna (Morton, 1973), Modiolus (Pierce, 1973; Morton, 1977), Musculista (Morton, 1974), y Brachi- dontes (Paiva Avelar & Narchi, 1984a, b). El género Mytella está representado en la Provincia Panámica por cuatro o cinco espe- cies, de las cuales cuatro se registran corrien- temente en territorio mexicano: Mytella guy- anensis (Lamarck, 1819), М. strigata (Hanley, 1843), M. speciosa (Reeve, 1857), y M. tum- bezensis (Pilsbry & Olsson, 1935) (Keen 1971; García-Cubas 4 Reguero, 1987); aunque Bernard (1983) considera a las dos ultimas como sinónimos. La distribución de М. strigata en el Pacífico se extiende desde Guaymas, Sonora, Méxi- co, hasta el sur de El Salvador e Islas Ga- lápagos; pero, ocurre también en el Atlántico desde Venezuela hasta Argentina (Keen, 1971). En México, М. strigata se encuentra en abundancia en las lagunas costeras de Agiabampo, Topolobampo, Yavaros y Huiza- che-Caimanero de la costa del Golfo de Cali- fornia (García-Cubas 8 Reguero, 1987) y en la costa del Pacífico, en las lagunas de Nuxco y Chautengo del estado de Guerrero (Stuardo & Villarroel, 1976; Villarroel, 1978); Laguna de Cuyutlán y Bahía de Manzanillo (Colima) (Cobo et al., 1978); y en la costa de Oaxaca (Holguín 4 González, 1989). Algunas especies del género Mytella han sido estudiadas en consideración a su im- portancia ecofisiológica y a su potencialidad como fuente de alimento. De М. guyanensis se tiene información acerca de su tamaño y concentración de metales (De Lacerda & Lima, 1983; De Lacerda et al., 1983), ma- durez sexual (Sibaja, 1986); sobrevivencia y capacidad de aislamiento en diferentes sali- nidades (Leonel 4 Silva, 1988). Mytella strigata ha sido estudiada desde el punto de vista de su biología, ecología, prác- ticas experimentales de cultivo y morfometría (Stuardo & Rivera, 1976; Estévez, 1975; Stu- ardo 8 Estévez, 1977; Sibaja, 1985), a su contenido de glucógeno y grasa (Reprieto & Stuardo, 1975) y de metales pesados (Paez- Osuna et al., 1988), pero se conocen sólo observaciones generales sobre sus partes blandas. Sin embargo, un detallado estudio anatómico-funcional sobre М. charruana, rea- lizado por Narchi & Galvao-Bueno (1983), nos permite comparar sus resultados con los obtenidos en M. strigata y extrapolarlos a nivel genérico. MATERIALES Y MÉTODOS Se obtuvieron 200 ejemplares de dife- rentes edades de las lagunas de Nuxco (100°47’N, 100°49’М/ y Chautengo (99°02’N, 99°09’W) (Guerrero; Nov. 1974-Mayo 1975) y TEsc. Biología, Lab. Invertebrados, Univ. Michoacana de San Nicolás de Hidalgo, Apdo. Postal 59-3, Morelia 58021, Michoacán, México. “Fac. Cienc., Biol. Rec. Nat., Depto. de Oceanografía Univ. Concepción, Casilla 2407, Concepción, Chile. 2 VILLARROEL & STUARDO de la Laguna de Cuyutlan, (19°02’N, 104°19’W) (Colima; Ago. 1977-Jul. 1978). La colecta se realizö de forma manual despren- diendo del lodo los grupos de ejemplares que yacen sobre el fondo. Las disecciones anatómicas se realizaron en ejemplares adultos de 50 a 70 mm de lon- gitud, previamente fijados en formol al 5% y preservados en alcohol etílico al 70%, utili- zando un microscopio estereoscópico Zeiss. Para precisar los detalles de poco contraste se utilizó el colorante rojo neutro. Las figuras de las estructuras internas se hicieron directamente al microscopio utili- zando ejer. plares fijados, y ejemplares vivos para el borde del manto. En la descripción de la musculatura, se usó la terminología de Graham (1934a, b) y en la del estómago las de Graham (1949) con las modificaciones sugeridas por Owen (1953) y Purchon (1957, 1958, 1960) у en especial por Dinamani (1967). La Figura 3 que muestra el interior del estómago, se hizo después de realizar un corte en la pared dorsal desde el esófago hasta el saco del estilo. Los términos dere- cho e izquierdo, aplicado a las estructuras estomacales, se refieren a esta línea media. RESULTADOS Y DISCUSIÓN Concha. La concha de М. strigata es mitiliforme, generalmente algo cóncava en su parte ven- tral, y de forma aguzada ó más ó menos en- sanchada anteriormente. Los umbos son subterminales a casi terminales. El margen dorsal es regularmente curvado. En la cara externa, se observan a menudo estrías radia- les finamente marcadas en el tercio anterior y escasas, pero más marcadas, en la mitad posterior, sobre todo internamente. El peri- óstraco es brillante, y su coloración variable entre amarillo verdoso claro a casi negro, uniforme o sombreado de verde o pardo amarillento en los márgenes anterior, dorsal posterior y especialmente el ventral. Prefe- rentemente en los ejemplares juveniles o me- dianos, se observan bandas oscuras radiales o entrecruzadas que resaltan sobre una su- perficie más clara o como manchas zigza- geantes o jaspeadas de color pardo. En la cara interna se observan de 2 a 4 pliegues radiales, a manera de dientes en el margen anterior. El ligamento es muy alar- gado llegando hasta la mitad de la concha. La impresión del aductor anterior es relativa- mente grande y la del aductor posterior es grande y redondeada; sobre esta última y ha- cia adelante de ella, se encuentra la huella del retractor posterior (Fig. 1a). La coloración interna es violácea oscura, brillante. La longitud máxima de la concha consta- tada por nosotros en esta especie es de 80 mm, aunque tales tamaños corresponden aparentemente sólo a unos pocos individuos en cada población. Sibaja (1985), en un es- tudio del crecimiento de la concha en una población de M. strigata de la Playa de Le- panto, Puntarenas, Costa Rica, encontró que es alométrico y que el largo es un parámetro adecuado para evaluar crecimiento. Sin em- bargo, la concha crece más en altura y es comparativamente más esférica que en M. guyanensis. Las medidas mínimas y máxi- mas obtenidas por este autor fueron de 10 y 42.6 mm y los promedios calculados iguales a 24.9, 10.8 y 7.8 mm para el largo, ancho y alto, respectivamente. Estas medidas de lon- gitud máxima son inferiores a las constata- das por Estévez & Stuardo (1977) en pobla- ciones de la costa Pacífica mexicana, en donde observaron máximos de hasta 70 mm, con promedios de hasta 49.6 mm en diver- sas lagunas. Manto y Aberturas Sifonales. Los lóbulos del manto en М. strigata se encuentran unidos en la región dorsal, en toda su longitud, desde el extremo anterior hasta la parte dorsal del área anterior del músculo retractor pedal posterior (Fig. 1a), a diferencia de otros mitilidos (e.g. Mytilus edu- lis) en donde la unión llega hasta la mitad de la región del aductor posterior (Bullough, 1958). En cambio, en Mytella charruana, Brachidontes darwinianus y B. solisianus esta unión llega hasta la región posterior al aduc- tor posterior (Narchi & Galvao-Bueno, 1983; Paiva Avelar 4 Narchi, 1984a, b). El borde del manto está compuesto de un delgado pliegue externo, un pliegue medio, moderadamente delgado y de un pliegue in- terno relativamente grueso, con su borde en- rollado y con tentáculos cortos o papilas digi- tiformes. En M. charruana el borde interno es liso, sin fusiones y totalmente libre; además, los bordes medios y externo son contínuos, musculares y unidos en la región posterior para formar un sifón (Narchi 8 Galvao-Bueno, 1983). Según White (1937), en Mytilus edulis este pliegue es ciliado y muy sensible al tacto MORFOLOGÍA DE MYTELLA STRIGATA 3 / = 7225 Zee — FIG. 1. Mytella strigata. (a) Manto izquierdo. (b) Vista ventral del tracto digestivo. (c) Cavidad paleal, sin branquia izquierda ni gländula digestiva. A, ano; AA, aductor anterior; Aa, aorta anterior; AE, abertura exhalante; Ai, abertura inhalante; AM, arteria del manto; AP, aductor posterior; Ap, aorta posterior; C, ctenidio; CD, capuchön dorsal; CV, conectivo visceral; DD, ductos de los diverticulos digestivos; E, esófago; Es, estómago; G, gónada; Gp, glándula pericärdica; |, intestino; M, mesosoma; NL, nervio paleal; NP, nervio del palpo; P, pie; Pe, pericardio; PL, palpo labial; PU, papila urogenital; R, recto; RA, retractor pedal anterior; Ri, riñón; RP, retractor pedal posterior; SE, saco del estilo; V, ventriculo. 4 VILLARROEL 8 STUARDO y a repentinos cambios de intensidad lumi- nosa. Aberturas Exhalante e Inhalante. El pliegue interno de los lóbulos del manto se une posteriormente, en su tercio superior, para formar la abertura exhalante con los márgenes lisos y protruidos debilmente. Así mismo, esta abertura queda limitada ventral- mente por la parte terminal de la membrana branquial, que forma un septo triangular. La unión continúa por un trecho corto, pero el resto del borde queda libre, dejando una gran abertura ventral que corresponde a la abertura inhalante, con papilas ó tentáculos digitiformes sólo en su región posterior, de manera similar a lo observado en M. charru- ana (Narchi 8 Galváo-Bueno, 1983). Yonge (1948) clasifica a este tipo de aberturas como sifones del tipo A, aunque no las considera sifones verdaderos. No es posible establecer diferencias entre las especies de Муе/а utilizando las carac- teristicas de las papilas de los bordes del manto, como lo sugiriera Soot-Ryen (1955). Debido a su gran variación y, en consecuen- cia, al menos en este caso, parecen no re- presentar un carácter morfológico de valor taxonómico. Las digitaciones o tentáculos del manto se observaron de color blanco y poco arborescentes en ejemplares peque- ños (juveniles); en cambio, en ejemplares grandes (maduros), tanto los márgenes del manto, las digitaciones y el septo presen- taron manchas pardas oscuras y las digita- ciones más arborescentes. Al igual que en otros mitílidos, en época de maduración sexual, cada lóbulo del manto se observa engrosado por las gónadas que se extienden en ellos. Músculos Aductores y del Pie. El aductor posterior en Mytella strigata tiene una posición semejante a los de M. charruana y de Mytilus edulis. Sin embargo, entre estos dos géneros se puede apreciar una diferenciación en lo que respecta a los retractores del pie. En Mytella, la rama pos- terior del retractor posterior alcanza hasta la mitad del aductor posterior (Fig. Та, с, RP, AP); en cambio, en Mytilus el retractor pos- terior queda sólo adosado al aductor poste- rior. Además, en Mytella se distinguen dos paquetes de retractores pedales; ésto es, los anteriores están separados de los posterio- res (Fig. la, с, RA, RP), como en Choro- mytilus, en cambio en Mytilus los haces están contíguos. Las diferencias funcionales no se han determinado pero, aparentemente están ligadas a una mayor actividad del animal. Pie. El pie de М. strigata (Fig. 1c, р), como en otros mitílidos, es pequeño, de color pardo obscuro y con un profundo surco posterior. No hay una variación apreciable en la forma del pie respecto de las otras especies cono- cidas. Ctenidios. En M. strigata, al igual que en M. charru- ana, la demibranquia interna es un poco menor que la externa. De acuerdo a Atkins (1937) pertenece al tipo de ctenidio B(1) ca- racterístico de Mytilidae y Pinnidae. No hay extensión supraxial (Fig. 1b, c, que muestra parte de las lamelas del lado derecho). Son branquias planas y homorrábdicas con una distribución de tractos ciliares descritos en detalle en М. charruana por Narchi 4 Galvao- Bueno (1983). Palpos Labiales. Los palpos de M. strigata, como los de M. charruana (Narchi 8 Galvao-Bueno, 1983) y Modiolus metcalfei (Morton, 1977), son muy grandes y alargados, en comparación con los de las especies de Mytilus en que son casi triangulares. Su longitud es tal, que al- canza hasta el pie y la parte anterior del me- sosoma (Fig. 1с, PL, М); hay numerosos sur- cos que se aprecian por transparencia (Fig. 1c). Entre las láminas externa e interna de los palpos derechos e izquierdos, hay un surco ciliado profundo, que conduce a la boca. Tubo Digestivo. El curso y las características del aparato digestivo en M. strigata (Fig. 1b, c) son, en general, similares a los de M. charruana, ex- cepto el largo y posición de la vuelta anterior del intestino, que en M. strigata es lateral izquierda, larga, y en M. charruana es más dorsal y más corta. En ambas especies el saco del estilo y el intestino salen en forma paralela separadamente hacia atrás, termi- nando sobre el tercio anterior del aductor posterior en M. charruana, y sobre el ano, MORFOLOGÍA DE MYTELLA STRIGATA 5 más atrás del aductor posterior en M. stri- gata. Con respecto a Mytilus edulis y las espe- cies de Xenostrobus hay una notable dife- rencia en el curso de la vuelta anterior del intestino, ya que en ellas, el intestino al volver desde atrás, después de separarse del saco del estilo, pasa de dorsal derecho a ventral al esófago y se vuelve a dirigir hacia atrás por el lado izquierdo (Bullough, 1958: fig. 141; Wil- son, 1967: fig. 3). En los otros géneros que han sido estudiados por diversos autores, que se mencionan en el siguiente párrafo, no se ha descrito el curso del tubo digestivo. Estómago. Estómagos de varios géneros de Mytilidae han sido descritos anteriormente en detalle. Así ocurre para tres especies de Modiolus: M. modiolus (Nelson, 1918; Reid, 1965), M. undulatus y М. striatulus (Dinamani, 1967). Dos especies de Lithophaga: L. nasuta (Pur- chon, 1957) y L. gracilis (Dinamani, 1967); Limnoperna fortunei (Morton, 1973); Muscu- lista senhausia y Modiolus melcalfei (Morton, 1977); Adula (Botula) falcata (Frankboner, 1971); Mytilus edulis (Graham, 1949; Reid, 1965); Perna viridis, Arcuatula sp. y Botula cinnamomea (Dinamani, 1967); Mytella char- гиапа (Narchi 4 Galvao-Bueno, 1983); y Brachidontes solisianus y B. darwinianus dar- winianus (Paiva Avelar & Narchi, 1984a, b) La nomenclatura adoptada en la siguiente descripción, es la propuesta por Owen (1953). El estómago de M. strigata (Fig. 2d, e), al igual que en M. charruana, es largo y aplas- tado dorsoventralmente, diferente del de Mytilus edulis que es corto. En su estructura interna se asemeja más al estómago de Perna viridis, descrito por Dinamani (1967). Como se ilustra en las Figuras 2b y 3, el esófago (E) desemboca anteriormente (E’) y presenta repliegues longitudinales internos (RF) que terminan en una elevación que cir- cunda su entrada al estómago; un surco lon- gitudinal corre a cada lado de esta elevación. Por ambos lados del estómago (Figs. 1b, 2a, c-e) entran numerosísimos ductos (más o menos 34) provenientes de los divertículos digestivos (DD) y posteriormente, salen jun- tos el saco del estilo (SE) y el intestino (I). Externamente y en vista dorsal (Fig. 2c) son también conspicuos: el extremo anterior del ciego seleccionador de alimento (CSD); hacia el lado izquierdo el capuchón dorsal (CD); debajo y hacia atrás, la bolsa izquierda (Bl); en la linea media un bolsillo en el que se transparentan surcos (Ca) y hacia la izquierda del mismo, el área de selección posterior (APS). Ventralmente (Fig. 2a), la estructura más notable es el ciego seleccionador de ali- mento (CS), aplanado y casi tan largo como el estómago mismo, en el que se transpa- renta el curso del tiflosol mayor y el surco intestinal. A la derecha de este último, y bajo el capuchón dorsal, nace el intestino (I) junto al saco del estilo, pero no dentro de él; am- bos se encuentran estrechamente unidos. Internamente, hay similitud con los rasgos generales descritos para otros géneros, pero se constatan diferencias en el ciego selec- cionador de alimento, la distribución de los ductos de los divertículos digestivos y la con- figuración del repliegue axial, como se des- cribe a continuación. En M. strigata el tiflosol mayor (Fig. 3, TY) emerge del intestino (I) y corre hacia adelante por el lado derecho sobre el piso del es- tómago, se curva hacia la izquierda y entra al ciego seleccionador de alimento (CS). Dentro de este ciego (Fig. 2b) da una vuelta com- pleta en espiral en dos planos; luego sigue hacia atrás hasta llegar casi al extremo del ciego, se curva hacia arriba manteniendo el mismo plano y, nuevamente, en sentido in- verso, sigue hasta el fondo del saco y se vuelve a curvar hacia arriba. Manteniendo el mismo plano anterior llega a la mitad del ciego, y siguiendo una amplia curvatura toma el sentido espiral de la primera vuelta y aban- dona el ciego por encima de donde entró. En total da cuatro vueltas, dos en cada sentido. En Modiolus modiolus, al igual que en Mytilus edulis, el tiflosol mayor presenta una sola vuelta; sin embargo, el ciego es más corto en Modiolus (Nelson, 1918; Reid, 1965: fig. 6). En Modiolus undulatus el tiflosol mayor entra en un plano y regresa en otro sobre sí mismo, ésto es, da una sola vuelta en dos planos; en cambio en Arcuatula sp. también experimenta una vuelta, pero en el mismo plano. En Mytilus striatulus el ciego está poco desarrollado, apenas como una depresión. El tiflosol muestra un pequeño embahiamiento y termina más arriba en la pared izquierda. Situación semejante, aun- que más sinuosa, se presenta en Lithophaga gracilis y en Botula cinnamomea (Dinamani, 1967). El término del tiflosol mayor (TY”), en M. strigata y Perna viridis se presenta una vez 6 VILLARROEL 8 STUARDO (b) (e) FIG. 2. Mytella strigata. Estómago. (a) Vista ventral. (b) Detalle del ciego seleccionador de alimento. (с), (а) у (e), Vistas dorsal, lateral derecha e izquierda respectivamente. (a), (c), (d) y (e) están en la misma escala. APS, Area de sección posterior; Bl, bolsa izquierda; Ca, ciego del APS; CD, capuchón dorsal; CS, ciego seleccionador; CSD, prolongación dorsal del CS; DD, ductos de los divertículos digestivos; E, esófago; Е’, abertura esofágica; |, intestino; Ra, repliegue axial; RL, repliegue lateral; SE, saco del estilo; Sl, surco intestinal; TY, tiflosol mayor; TY”, término TY. MORFOLOGÍA DE MYTELLA STRIGATA Y < FIG. 3. Interior del estómago de Mytella strigata abierto por un corte longitudinal dorsal. AS, área de selección; APS, AS posterior; Bl, bolsa izquierda; Ca, ciego del APS; CD, capuchón dorsal; CS, ciego seleccionador; D, dientes del EG; DD, ductos divertículos digestivos; E, esófago; EG, escudo gás- trico; |, intestino; Ra, repliegue axial; Rd, repliegues dorsales; RF, repliegues esofágicos; RL, repliegue lateral; SE, saco del estilo; SI, surco intestinal; SI’, comienzo del SI; TN, tiflosol menor; TY, tiflosol mayor; TY’, término del TY. que el tiflosol emerge del ciego selecciona- Modiolus modiolus (Nelson, 1918) y en dor de alimento, sobre la pared izquierda la- Mytilus edulis (Graham, 1949) termina dentro teral a la abertura esofágica. En cambio, en del ciego seleccionador de alimento. 8 VILLARROEL 8 STUARDO Todos los ductos de los divertículos diges- tivos del lado derecho (18), están distribuidos en una línea a intervalos uniformes, y desem- bocan en una “depresión” a la derecha del tiflosol mayor, que se comunica anterior- mente con el ciego seleccionador de ali- mento y, posteriormente, conduce hacia el área posterior de selección. Dicha distribu- ción es semejante a la de Botula cin- namomea y más o menos similar en Mytilus edulis, pero distinta a la de Lithophaga gra- cilis, en la que los ductos están ordenados еп tres grupos (Dinamani, 1967: fig. 8). A la derecha y arriba de esta “depresión,” hay dos repliegues dorsales (Fig. 3, Rd), con un surco profundo entre ellos, que comienza como un pequeño reborde; los repliegues salen desde la plataforma anterior al ciego del área posterior de selección y terminan en la entrada del capuchón dorsal. Las dimen- siones en su origen son mayores que el tiflo- sol mayor y van disminuyendo gradualmente hasta casi desaparecer en su término. En la parte dorsal del estómago, a la derecha de los repliegues antes descritos, se encuentra un área de selección (AS) con nu- merosos pliegues y surcos finos y uniformes que corresponden al tracto anterodorsal de Reid (1965). Este tracto relaciona, en parte, al capuchón dorsal con el ciego del área pos- terior de selección. Su unión directa está im- pedida por un área lisa junto al ciego del área posterior de selección y por una prolonga- ción del escudo gástrico en la entrada del capuchón dorsal. El capuchón dorsal (CD), a diferencia de las otras especies en que se ha descrito, pre- senta un área de pliegues y surcos que sin duda corresponde a un área de selección. Su presencia en esta estructura (considerada sólo de almacenamiento), junto con los otros caracteres ya mencionados, implica una mayor adaptación a la selección de particu- las. Los ductos de los divertículos digestivos del lado izquierdo se abren tanto en el área deprimida que se encuentra entre el escudo gástrico y el repliegue axial, a la entrada de la bolsa izquierda (Bl) como dentro de la misma. El conjunto del ciego seleccionador de ali- mento con los ductos de los divertículos di- gestivos, la continuación del tiflosol mayor y el surco intestinal, representan el mecanismo principal de selección del estómago y corres- ponde al tipo B de Reid (1965). El surco intestinal (SI) se origina cerca de la bolsa izquierda (SI’), sigue a la izquierda del tiflosol mayor en todo su curso por el piso del estómago, entra luego al ciego selecciona- dor de alimento y regresa por la derecha del tiflosol hacia el intestino. Un repliegue aplanado (RL), denominado repliegue lateral por Dinamani (1967) en Perna viridis y descrito anteriormente por Graham (1949) como “fold” en Mytilus edu- lis, comienza donde termina el tiflosol mayor. Este repliegue pasa por las paredes del ciego seleccionador; al salir, gira hacia la derecha y sigue bajo el esófago formando una plataforma; se vuelve luego hacia atras y corre en dirección paralela al tiflosol mayor sobre el lado derecho del estómago. En el hecho, constituye una pared delante del área de selección posterior, y deja entre él y el surco intestinal una pequeña área plegada que, quizás, represente un área menor de se- lección. En M. strigata, el área posterior de selec- ción, ocupa una especie de bolsillo sobre la pared dorsal izquierda (Fig. 3, APS), con un gran desarrollo de repliegues y surcos, en comparación a las otras especies. Efectiva- mente, en las especies de Modiolus descritas por Dinamani (1967) no hay área posterior de seleción, sino que aparenta haber pequeños abultamientos sobre la pared derecha. En M. undulatus tiene la forma de una depresión baja; en M. striatulus es más pronunciada y en Arcuatula existe un área a manera de ca- nal que se extiende desde el intestino, donde va también el tiflosol menor. El ciego del área posterior de selección (Ca), que de acuerdo a Reid (1965) no es funcional en Mytilus edulis, está bien de- sarrollado еп М. strigata, presentando pliegues y surcos bien notorios; éstos desempeñan probablemente una función se- leccionadora. Además, el ciego está casi separado del área principal, situación que tampoco presenta Perna viridis. El área posterior de selección está limitada ventral y anteriormente por un surco pro- fundo, que corresponde al surco de rechazo (SR) descrito para algunos eulamelibran- quios (Owen, 1953; Dinamani, 1967); su fun- ción es la de drenar el área de selección pos- terior y llevar las partículas al surco intestinal. Otra zona del estómago que presenta un área de pliegues muy finos, comparable solo a la descrita por Dinamani (1967: fig. 4a) en Perna viridis, se encuentra sobre el pliegue axial (Ra). Este repliegue axial ocupa casi todo el piso del estómago desde el ciego se- MORFOLOGÍA DE MYTELLA STRIGATA 9 leccionador de alimento hasta la bolsa izquierda y es tan prominente como en las especies de Modiolus. El escudo gástrico (EG) no es muy grande. Presenta dos prolongaciones anteriores ha- cia el capuchón dorsal; una muestra su ex- tremo romo y la otra un par de fuertes dien- tes. Detrás de estos últimos, existe un área deprimida con varias líneas longitudinales de dientes quitinosos muy pequeños. Este carác- ter, que indudablemente ayuda en la mejor desintegración de las partículas alimenticias, no ha sido descrito en otros mitílidos. Recto. El recto, después de dejar la cavidad peri- cárdica y atravesar al ventriculo, desciende por la parte posterior del aductor posterior y desemboca en el ano (A). Es difícil adjudicar alguna ventaja funcional al plan del recto atravesando la cavidad peri- cárdica de estos mejillones y sus modifica- ciones o derivar este ordenamiento de al- guna explicación embriológica o fisiológica. A este respecto, en M. strigata se observa otra diferencia con Mytilus edulis, ya que el recto luego de abandonar el ventrículo y ca- vidad pericárdica pasa dorsalmente sobre el complejo de músculos comprendidos por el aductor posterior, el retractor posterior del biso y el retractor pedal posterior. Por otra parte, Pierce (1973) encontró diferencias en la morfología interna del recto de los mitílidos estudiados por él. Compa- rando sus resultados con lo encontrado por nosotros en M. strigata el tiflosol del recto es algo similar al de Modiolus demissus grano- sissimus, aunque un poco más aplastado y completamente diferente al de Mytilus edulis, Modiolus squamosus e Ischadium recurvum, en los que el tiflosol no está bien definido. Sistema Circulatorio. Quitando las valvas de ejemplares donde la gónada se encuentra en estado de reposo о indiferenciado (Fig. 1a), se aprecian en primer plano las ramificaciones de la aorta anterior (Aa), la aorta posterior (Ap) y las ar- terias del manto anterior y posterior (AM). La aorta anterior se bifurca delante del re- tractor anterior del pie, originando la arteria anterior del manto (AM) y otra rama que se dirige hacia las vísceras. Las ramificaciones de la arteria anterior del manto cubren la mi- tad anterior del mismo. La rama más anterior de esta arteria forma un circuito cerrado, como se aprecia en la Figura Та. La aorta posterior sólo irriga la parte superior del manto. La arteria posterior del manto aparece por debajo de la parte anterior del retractor posterior del pie, y se ramifica repetidas veces cubriendo la mitad posterior del manto. El corazón se encuentra dentro de la ca- vidad pericárdica (Fig. 1с), ubicada casi pos- terior a los músculos retractores pedales an- teriores. Las aurículas, cubiertas por la glándula pericárdica, están tan alargadas hacia atrás que llegan a alcanzar las terminaciones de los retractores posteriores del pie, sin cur- varse luego hacia el lado contrario. La glán- dula pericárdica presenta un aspecto granu- lar aceitoso. La sangre entra a las aurículas lateral- mente desde numerosos senos pequeños que están ligeramente tapizados y oscureci- dos por los órganos de Keber, y abandona el ventrículo anteriormente por la aorta anterior, que desemboca en un bulbo aórtico que sale del pericardio. El ventrículo (V) es alargado y está atrave- sado por el recto (R) en toda su longitud. Como en la mayoría de los mitilidos (Fig. 1c), el ventrículo está suspendido desde cuatro puntos: anteriormente de la aorta y el recto; posteriormente del recto y lateralmente de las aurículas. El recto pasa longitudinalmente a través de todo el lúmen del ventrículo y de ahí hacia atrás a través de todo el largo de la cavidad pericárdica. Tal secuencia de recto y ventrículo ha sido descrita en varios mitílidos (Field, 1922; White, 1942; Jegla & Greenberg, 1968) y en particular para Modiolus squamo- sus (Pierce, 1973). En cambio, la suspensión de los ventrículos de “Modiolus” demissus e Ischadium recurvum es completamente diferente del plan típico de los mitilidos, y resulta de un modelo modificado del paso del recto a través de la cavidad pericárdica. Efectivamente, el recto pasa solamente a través de la porción anterior del ventrículo y luego, emergiendo desde la superficie dorsal del ventrículo, se arquea dorsalmente en su propia envoltura a lo largo del techo de la cavidad pericárdica, para después subir en el extremo posterior de la cavidad. La mitad posterior del ventrículo, no soportada por el recto, cuelga libremente en la cavidad peri- cárdica, y su extremo anterior está suspen- dido y anclado por las aurículas y el recto. Según Pierce (1973), una consecuencia fisio- 10 VILLARROEL 8 STUARDO lógica obvia de su ordenamiento es que la dirección del batir ventricular en estas dos especies es postero-anterior, más que lateral a medio como en la mayoría de los mitílidos. Organos Excretores. El riñón (Figs. 1a, с, Ri) en М. strigata se encuentra a ambos lados de la base de la branquia. Desde la parte anterior del retrac- tor del pie, donde se ensancha un poco, pasa lateralmente por toda la masa visceral, con- servando más o menos el mismo diámetro hasta el aductor posterior, donde se ensan- cha notablemente y desciende finalmente por debajo del mismo. Esta situación es diferente en Mytilus edulis, donde la parte an- terior llega hasta la región de los palpos y la parte posterior hasta el límite posterior del aductor (Bullough, 1958: fig. 141) al igual que en Brachidontes darwinianus y B. solisianus (Avelar & Narchi, 1984 a, b). El riñón, a ambos lados de la masa vis- ceral, aparece como una bolsa pardo oscura con digitaciones notables hacia el extremo anterior y el posterior. Está unido por una angosta banda de tejido renal al organo de Keber o “glándula pericárdica” (Gp). Hay una abertura interna muy poco visible en el peri- cardio y una abertura externa de la papila urogenital (PU). A esta papila se abren uno al lado del otro el ducto renal y el gonoducto. Sistema Nervioso El sistema nervioso, en su aspecto general, no presenta variaciones notables en el análi- sis comparativo. Los dos ganglios cerebrales se encuentran situados posteroventralmente a los bordes de la boca. Un conectivo cerebro-visceral los une al ganglio visceral y una bifurcación de éste (el conectivo cerebro-pedal), al ganglio pedal. Cada ganglio visceral se encuentra en posición antero-ventral al músculo aductor posterior en la línea de fijación de los cteni- dios. Los dos ganglios pedales están es- trechamente unidos en la parte más pro- funda del extremo proximal del pie. El nervio más prominente del ganglio ce- rebral es el nervio anterior del manto. En el caso del ganglio pedal, es el nervio pedal el que pasa hacia abajo en el pie; en el del gan- glio visceral es el nervio posterior del manto que corre por toda la periferia de éste. Sistema Reproductor. Las gónadas (Fig. 1a, С) se encuentran ex- tendidas en los lóbulos derechos e izquierdo del manto y entre los órganos de la masa visceral. Las regiones ventrales de las góna- das derecha e izquierda llenan completa- mente el mesosoma (M). Los gonoductos se abren a lo largo del ducto renal en la papila urogenital (PU). CONCLUSIONES Mytella strigata, al igual que М. charruana y Modiolus demissus, es una forma infaunal o semiinfaunal que se entierra en el sustrato blando para ganar estabilidad y protección, aunque el presentar biso, la capacita para fijarse a cualquier tipo de sustrato vecino duro o semiduro. En las lagunas estudiadas se encontró viviendo en fondos areno-limosos con alto contenido de materia orgánica y restos de conchas, formando bancos de extensión y abundancia considerables (Stuardo & Villar- roel, 1976; Stuardo 8 Estévez, 1977; Villar- roel, 1978). En estos bancos, como con- secuencia de la falta de sustrato duro, los ejemplares se fijan unos a otros, de modo que llegan a formar masas como “racimos” que se mantienen sobre el sustrato y parcial- mente enterradas, debido a la tranquilidad de las aguas y a la falta de corrientes notorias. Sin embargo, cualquier tipo de sustrato duro (rocas, raíces de mangle, etc.) y por su- puesto, sustratos artificiales, determinan la fijación muy numerosa de ejemplares. En la Laguna de Chautengo los bancos se encon- traron concentrados en la mitad oriental y so- bre todo en el sector noreste; en la laguna de Nuxco, cubrían gran parte de los fondos, salvo en las zonas más profundas, pero la mayor concentración correspondió a los sec- tores marginales y en particular, a la región oriental cercana a la barra y al canal, como lo ilustran los mapas de Stuardo 8 Estévez (1977). En la Laguna de Cuyutlán existían bancos en zonas con características típicamente la- gunares, pero al abrir artificialmente el Canal Ventanas, desaparecieron por el cambio ha- cia condiciones marinas. Mytella como Modiolus, no anida en el lodo con el biso como ocurre en Musculus, Mus- culista y Amygdalum, a los que Morton (1977) considera más especializados. Por lo tanto, MORFOLOGÍA DE MYTELLA STRIGATA 11 Mytella puede considerarse una forma inter- media entre los que anidan y los altamente especializados como Mytilus, Septifer y Lim- noperna de la epifauna, como lo sugieren Narchi & Galvao-Bueno (1983). Considerando distintas opiniones, los ca- racteres que refuerzan su estado infaunal o semiinfaunal son: condición anisomiaria ex- trema; concha de contorno triangular con umbos recurvados o con forma de gancho, que la capacita para anclarse mejor o adher- irse a las superficies redondeadas de ejem- plares de mejillones vecinos; palpos y bran- quias bien desarrollados, que junto a los grandes retractores posteriores favorecen la ventilación y eliminación de partículas de sedimento; y un estómago con complejos tractos de selección para la selección de par- tículas de alimento. Estas características definirian también una condición primitiva entre los Mytilidae (Yonge & Campbell, 1968). Efectivamente, la reducción de la región anterior de la concha en los Mytilidae está acompañada por la disminución del tamaño del aductor anterior. Ya Yonge (1953) lo su- girió como una adaptación en estos organis- mos gregarios, para elevar la región posterior de la concha, de modo que no sean obstru- ídas las corrientes inhalantes de individuos muy cercanos. Una alternativa propuesta por Stasek (1966), sugiere que la reducción an- terior puede haber evolucionado en los mití- lidos de regiones tropicales, donde la pro- ductividad es baja, como una forma de aumentar la captación de alimento. Según Morton (1977), la condición heteromiaria es primitiva e indica que el hábito de enterrarse precede al epifaunal. Por otra parte, Stanley (1970) relaciona el área de la corriente inhalante y el grado de bombeo con el tamaño y estructura de los órganos de bombeo (las branquias), los que podrían ser más importantes en este as- pecto. Es probable que еп М. strigata y М. charruana la disminución de la abertura de los bordes del manto ayude a impedir la en- trada de fango a la abertura exhalante y a favorecer la tasa de bombeo. También es posible que la presencia de gran cantidad de papilas en el borde del manto, juegue un mayor papel en la detección de la calidad del agua circundante, y la filtración y rechazo de las partículas de limo-arcilla en aguas cal- mas, о de arena fina en hábitat rocosos con alta energia, como se ha observado en Mytilus edulis y en Perna perna (Narchi 8 Gal- vao-Bueno, 1983). Los palpos de M. strigata y M. charruana presentan mayor tamaño y, por consiguiente, mayor número de surcos y pliegues que los de Mytilus edulis, lo que indica un mayor grado de selectividad de particulas alimenti- cias (materia orgánica particulada, fito y na- noplancton). Esta situación se corrobora también por la presencia de un gran número de filamentos branquiales. Otros caracteres importantes que apoyan la adaptación de Mytella a los fondos blan- dos y explican su forma de alimentación son: el estómago de tipo Ш de Purchon (1957) con varias áreas de selección, y dentro del mismo, el rol del ciego seleccionador de ali- mento. Este actuaría más como un reservorio temporal de alimento que como un ciego de selección, capacitando a esta especie para períodos de ayuno largos, como lo trata de demostrar Dinamani (1967) con base en sus observaciones en Perna viridis. Pero, tam- bién parece plausible considerar a este ciego como una estructura seleccionadora, basán- donos en que М. strigata es una especie que vive tanto en aguas de alta turbidez como en aguas limpias. Hay obviamente otras estruc- turas en el estómago que contribuyen a que esta función sea llevada a cabo con mucha eficiencia; este es el caso de un tiflosol mayor muy largo y de una posible área de selección en el capuchón dorsal; la presencia muy desarrollada del área posterior de selec- ción con su respectivo ciego, y otros carac- teres no menos importantes como son: el área del escudo gástrico con pequeños di- entes quitinosos, no descrita en otros mitíli- dos y que representa una estructura po- derosa en la desintegración de particulas; el intestino muy largo con un tiflosol bien de- sarrollado; y, por último, la salida separada del saco del estilo y del intestino, de manera similar a lo observado en M. charruana y Musculista senhausia descritos por Narchi & Galväo-Bueno (1983) y Morton (1974), res- pectivamente. RESUMEN Utilizando principalmente muestras fijadas de ejemplares colectados en las lagunas costeras mexicanas de Nuxco, Chautengo y Cuyutlán se describe la morfología de la con- cha y de las partes blandas, especialmente del estómago, en Mytella strigata. Se le com- para con las descripciones publicadas de M. charruana y especies de mitílidos de los gé- 12 VILLARROEL 8 STUARDO neros Modiolus, Mytilus, Lithophaga, Perna, Arcuatula y Botula. Al igual que М. charruana posee sifones del tipo A (Yonge, 1948), ctenidios del tipo B(1) (Atkins, 1937) y el es- tómago es del tipo Ш (Purchon, 1957) con mecanismos de selección de tipo B (Reid, 1965). Se discute la posible relación entre adaptaciones morfológicas y sus hábitos de vida. AGRADECIMIENTOS Agradecemos a la Comisión del Río Balsas y a la Comisión Federal de Electricidad de México, que en su momento otorgaron el apoyo financiero para desarrollar este estu- dio; a los Institutos de Ingeniería y Ciencias del Mar y Limnología de la Universidad Na- cional Autónoma de México, por las facili- dades otorgadas durante la realización del mismo; a los colegas Edmundo López, a Sandra Rubio de la Universidad Michoacana de San Nicolás de Hidalgo y a Zoila Castillo de la Universidad Nacional Autónoma de México por sus valiosas críticas y a Lourdes Espinoza por el entintado de las figuras. LITERATURA CITADA ATKINS, D., 1937, On the ciliary mechanisms and interrelationships of Lamellibranchs. Part Ill: types of Lamellibranch gills and their food cur- rents. Quarterly Journal of microscopical Sci- ences, N.S., 79: 375-421. BERNARD, F. R., 1983, Catalogue of the living Bi- valvia of the Eastern Pacific Ocean: Bering Strait to Cape Horn. Canadian Special Publication of Fischeries and Aquatic Sciences, 61: 102 pp. BULLOUGH, W. S., 1958, Practical invertebrate anatomy. Macmillan & Co., New York, 483 pp. COBO, D., M. VILLARROEL, L. TREVIÑO, M. T. ORTA, R. ITURBE, R. MORALES 4 C. RANGEL, 1978, Determinación de los efectos de la termo- eléctrica de Manzanillo en la flora y fauna de la Laguna de Cuyutlán, Colima. Instituto de Inge- niería, Universidad Nacional Autónoma de Méxi- co-Comisión Federal de Electricidad, Proyecto 7038. 67 pp. DE LACERDA, (. D. & М. В. W. LIMA, 1983, Size and metal concentration in the mangrove mussel Mytella guyanensis (Lamarck). Brazilian Journal of Medical Biological Research. 16: 421. DE LACERDA, L. D., N. R. W. LIMA, W. C. PFEIF- FER 8 M. FISZMAN, 1983, Size and metal con- centration in the mangrove mussel Mytella guya- nensis (Mollusca: Bivalvia) from Baia de Sepetiva Brasil. Revista de Biologia Tropical, 31: 333-335. DINAMANI, P., 1967, Variation in the stomach structure of the Bivalvia. Malacologia, 5: 225- 268. ESTÉVEZ, J., 1975, Aspectos generales de la bio- logía y ecología del mejillón Mytella strigata (Han- ley, 1843) en las costas del estado de Guerrero. Tesis Profesional Facultad de Ciencias, Univer- sidad Nacional Autónoma de México. 63 pp. FANKBONER, P. V., 1971, The ciliary currents as- sociated with feeding, digestion and sediment removal in Adula (Botula) falcata Gould, 1851. Biological Bulletin, 140: 28-45. FIELD, |. A., 1922, Biology and economic value of the sea mussel Mytilus edulis. United States Bu- reau of Fisheries, Bulletin, 38: 127-259. GARCIA-CUBAS, A. & М. REGUERO, 1987, Conocimiento actual de los mitílidos tropicales de México. Memorias del Encuentro Regional sobre producción de mejillón. Ensenada, Baja California. 17-19 Nov. FONDEPESCA, Instituto de Investigaciones Oceanográficas de la Univer- sidad de Autónoma de Baja California y SEPESCA Baja California: 139-165. GRAHAM, A., 1934a; b, The cruciform muscle of lamellibranchia; the structure and relationships of lamellibranch possessing a cruciform muscle. Proceedings of the Royal Society of Edinburgh, 54: 17-30; 158-187. GRAHAM, A., 1949, The molluscan stomach. Transactions of the Royal Society of Edinburgh, 61: 737/18: R HOLGUIN Q., O. E. & A. C. GONZALEZ, 1989, Atlas de los moluscos marinos del estado de Oaxaca. Secretaría de Educación Pública. Insti- tuto de Ciencias Marinas, Atlas CICIMAR 7: 221 pp. JEGLA, Т. С. & М. J. GREENBERG, 1968, Struc- ture of the bivalve rectum, I. Morphology. Veliger, 10: 253-263. KEEN, A. M., 1971, Sea shells of tropical west America. Marine Mollusks from Baja California to Peru, 2da. ed. Stanford University Press, Stan- ford, California. 1064 pp. LEONEL, В. М. V. & I. М. SILVA, 1988, Estudo de sobrevivencia e de capacidade de isolamento de Mytella guyanensis (Moll. em diferentes salini- dades. Revista Nordestina Biologia, 6: 34-41. MERRILL, А. $. & В. О. TURNER, 1963, Nest build- ing in the bivalve genera Musculus and Lima. Veliger, 6: 55-59. MORTON, B., 1973, Some aspects of the biology and functional morphology of the organs of feeding and digestion of Limnoperna fortunei (Dunkner) (Bivalvia: Mytilacea). Malacologia, 12: 265-281. MORTON, B., 1974, Some aspects of the biology, population dinamics, and functional morphology of Musculista senhausia Benson (Bivalvia, Myti- lidae). Pacific Science, 28: 19-33. MORTON, B., 1977, The biology and functional morphology of Modiolus metcalfei (Bivalvia: Mytilacea) from the Singapore mangrove. Mala- cologia, 16: 501-517. MORFOLOGÍA DE MYTELLA STRIGATA 13 NARCHI, W. & М. $. GALVAO-BUENO, 1983, Anatomía funcional de Mytella charruana (d’Or- bigny, 1846) (Bivalvia: Mytilidae). Boletim de Zoologia da Universidade de Sáo Paulo, 6: 113- 145. NELSON, T. C., 1918, On the origin, nature, func- tion of the cristalline style in lamellibranchs. Journal of Morphology, 31: 53-111. OWEN, G., 1953, On the biology of Glossus hu- manus (L.) (Isocardia cor Lam.). Journal of Ma- rine Biology Association of United Kingdom. 32: 85-106. OWEN, G., 1974, Studies on the gill of Mytilus edu- lis: the eulaterofrontal cirri. Proceedings of the Royal Society of London, B, 187: 83-91. PAEZ-OSUNA, F.; G. IZAGUIRRE-FIERRO, В. 1. GODOY-MEZA, Е. GONZALEZ-FARIAS & J. 1. OZUNA-LOPEZ, 1988, Metales pesados еп cua- tro especies de organismos filtradores de la re- gión costera de Mazatlán: técnicas de extrac- ción y niveles de concentración. Contaminación Ambiental 4: 33-41. РАМА AVELAR, W. Е. P. & W. NARCHI, 1984a, Functional anatomy of Brachydontes solisianus (Orbigny, 1846) (Bivalvia: Mytilidae). Boletim de Zoologia da Universidade de Sáo Paulo 8: 215- 237. РАМА AVELAR, W. E. Р. & W. NARCHI, 1984b, Anatomia funcional de Brachidontes darwin- ianus darwinianus (Orbigny, 1846) (Mollusca: Bi- valvia). Papéis Avulsos de Zoologia, Säo Paulo, 35: 331-359. PIERCE, 5. K., 1973, The rectum of “Modiolus” demissus (Dillwyn) (Bivalvia: Mytilidae); a clue to solving a troubled taxonomy. Malacologia, 12: 283-293. PURCHON, R. D., 1957, The stomach in the Fili- branchia and Pseudo-lamellibranchia. Proceed- ings of the Zoological Society of London, 129: 27-59. PURCHON, R. D., 1958, The stomach in the Eu- lamellibranchia; stomach type IV. Proceeding of the Zoological Society of London, 131: 487-525. PURCHON, R. D., 1960, The stomach in the Eu- lamellibranchia, stomach types IV and V. Pro- ceedings of the Zoological Society of London, 35: 251-271. REID, R. G. B., 1965, The structure and function of the stomach in bivalve molluscs. Journal of Zo- ology, 147: 156-184. REPRIETO, J. Е. 8 J. STUARDO, 1975, Determina- ción del contenido de glucógeno y grasa en el “mejillón de laguna” Mytella strigata en dos la- gunas costeras del Estado de Guerrero. México. Informe Técnico Interno, Centro de Ciencias del Mar y Limnología, Universidad Nacional Au- tónoma de México. 22 pp. SIBAJA, W. G., 1985, Dimensiones de la concha del mejillón Mytella strigata Hanley (Bivalvia: Mytilidae) de la playa de Lepanto, Puntarenas, Costa Rica. Revista de Biología Tropical, 33: 59- 60. SIBAJA, W. G., 1986, Madurez sexual en el mejillón chora Mytella quayanensis Lamarck, 1819 (Bi- valvia: Mytilidae) del manglar en jicaral, Punta Arenas, Costa Rica. Revista de Biologia Tropi- cal, 34: 151-155 SOOT-RYEN, T., 1955, A report on the family Myti- lidae. Allan Hancock Pacific Expedition, 20: 1-175. STANLEY, S. M., 1970, Relation of the shell form to life habits in the Bivalvia (Mollusca). Geological Society of America, Memories, 125: 296 pp. STASEK, С. R., 1966, Views on the comparative anatomy of the bivalved Mollusca. Malacologia, 5: 67-68. STUARDO, J. & J. ESTÉVEZ, 1977, Aspectos gen- erales de la biología y ecología del “mejillón de laguna” Mytella strigata (Hanley, 1843) en dos lagunas costeras de Guerrero, México. Anales del Instituto de Biología de la Universidad Na- cional Autónoma de México, Serie Zoologia, 1: 187-229. STUARDO, J. 8 В. RIVERA, 1976. Sobre la biología de los bancos del mejillón Mytella strigata y prácticas experimentales de cultivo en la laguna de Chautengo, Gro. In: Informe final de la Sera. etapa del Programa uso de la zona costera de los Estados de Michoacán y Guerrero. Convenio Comisión del Río Balsas-UNAM. Subprograma de Biología. Centro de Ciencias del Mar y Lim- nología de la Universidad Nacional Autónoma de México, 2: 188-223. STUARDO, J. & М. VILLARROEL, 1976, Aspectos ecológicos y distribución de los moluscos en las lagunas costeras de Guerrero. México. Anales del Centro de Ciencias del Mar y Limnología de la Universidad Nacional Autónoma de México, 3: 65-92. VILLARROEL M., M., 1978, Adaptaciones funcio- nales y relaciones ecológicas de los macroinver- tebrados bentónicos, especialmente moluscos, en tres lagunas costeras de Guerrero, México. Tesis de doctorado, Centro de Ciencias del Mar y Limnología, Universidad Nacional Autónoma de México. 189 pp. WHITE, K. M., 1937, Mytilus. Liverpool Memories Biological Committee, 7: 117 pp. WHITE, K. M., 1942, The pericardial cavity and the pericardial gland of the lamellibranchia. Pro- ceedings of Malacological Society of London, 25: 37-88. WILSON, B. R., 1967, A new generic name for three recent and one fossil species of Mytilidae (Mollusca-Bivalvia) in southern Australia with re- descriptions of the species. Proceedings of Ma- lacological Society of London, 37: 279-295. YONGE, C. M., 1948, Formation of siphons in Lamellibranchia. Nature, 1961: 198-199. YONGE, C. M., 1953, The monomyarian condition in the Lamellibranchia. Transactions of the Royal Society of Edinburg, 62: 443-478. YONGE, C. M., 1955, Adaptation to rock boring in Botula and Lithophaga (Lamellibranchia, Myti- lidae) with a discussion on the evolution of this 14 VILLARROEL 8 STUARDO habit. Quarterly Journal of Microscopical Sci- ence, 96: 383-410. YONGE, С. М. & J. I. CAMPBELL, 1968, On the heteromyarian condition in the Bivalvia with spe- cial reference to Dreisena polymorpha and cer- tain Mytilacea. Transactions of the Royal Society of Edinburg, 68: 21-43. Revised Ms. accepted 25 Aug. 1993 MALACOLOGIA, 1995, 36(1-2): 15-27 LABORATORY EXPERIMENTS ON THE INFLUENCE OF FOOD AVAILABILITY, TEMPERATURE AND PHOTOPERIOD ON GONAD DEVELOPMENT IN THE FRESHWATER MUSSEL DREISSENA POLYMORPHA Jost Borcherding Zoological Institute, University of Cologne, Physiological Ecology, Weyertal 119, D-50923 Kóln, Germany ABSTRACT (1) Two groups of the mussel Dreissena polymorpha were collected in March (E1) and Sep- tember (E2). Each group consisted of individuals at different stages of the annual gonadal cycle (e.g. females: in E1 had high numbers of oocytes still increasing in size, females in E2 had empty but recrudescent gonads). (2) Each group was kept for three months under nine different combinations of temperature (5°, 12°, 19°C), food availability (high and low), and photoperiod (LD 16:8 and 8:16). At each temperature (except at 5°C with E1), a high availability of food resulted in significantly larger gonads. (3) When food availability was high, the maximum gonad volume always occurred at 12°С. A low availability of food caused a progressive decrease in gonad volume at increasing temperatures. An influence of the photoperiod (tested at low food availability at all three temperatures) was not established. (4) The increase in oocyte size to maturity in the E1-series correlated positively with increasing temperatures. Simultaneously, the number of oocytes decreased significantly, and it appears that oocytes were reabsorbed in order to support maturation of the remaining oocytes. (5) An ample supply of food influenced oocyte development positively. No influence of photoperiod was observed. (6) The results are discussed with respect to the spread of D. polymorpha. Key words: Dreissena polymorpha, reproduction, laboratory experiments, gonads, oocyte size frequency distribution, temperature, food availability, photoperiod. INTRODUCTION The recent introduction of the zebra mus- sel Dreissena polymorpha (Pallas) into North American rivers and lakes (e.g. Hebert et al., 1989; Roberts, 1990) has evoked the in- creased interest of ecologists and water util- ities as well as the general public in this “pest” species. As in Europe some 30 years ago, zebra mussel research in North America has until now focused mainly on the ability to disrupt in water supplies and influence the aquatic food web (cf. Nalepa & Schloesser, 1993). In the 1970s, the focus of research in Europe shifted to the physiological ecology of this interesting freshwater species, partic- ularly in the context of their potential use in biomonitoring studies (Neumann & Jenner, 1992). Research into a species’ reproduction is essential for an understanding of its ecology and thus, of its ability to spread (Sastry, 1979). Among the freshwater bivalves, D. polymorpha is the only species to reproduce via a pelagic larva. This type of reproduction, which 1$ common in marine species, 15 char- 15 acterized by high fecundity, and is probably one reason as to why the zebra mussel has been able to spread rapidly in lakes and rivers presenting favourable conditions (Sprung, 1989). In Europe, the zebra mussel has colonized waters with various trophic and temperature conditions (e.g. Stánc- zykowska, 1977). As reviewed by Sastry (1979) for marine bivalves, both temperature and food availability can influence the course of the annual gametogenetic cycle, the rate of oocyte development, the number of differ- entiating oocytes (and hence the gonad size), and the onset of spawning. This is exempli- fied by the bay scallop Argopecten [= Ae- quipecten] irradians, where a minimum tem- perature of 20°C and an abundance of food are essential for the successful maturation of oocytes (Sastry, 1968). The influence of environmental conditions on the reproduction of D. polymorpha is poorly understood. Onset of the spawning season in spring was correlated with a tem- perature threshold of about 12°C at three dif- ferent locations in two central European lakes (Borcherding, 1991). In the upper hypolim- 16 BORCHERDING TABLE 1. Mean daily level of energy for each mussel (Joule/day, see text) under the various conditions employed during the experiments. El = experiment with mussels before the spawning season, E2 = experiment with mussels after the spawning season, НЕЁ = high food level, LFL = low food level. food photo- level period HFL LD 16:8 El EEE LD 16:8 LFL LD 8:16 HFL LD 8:16 E2 ELE LD 8:16 EEE LD 16:8 nion of lakes, where there is a phase shift in annual temperature cycles, the onset of spawning was delayed until late summer. However, food availability may also affect go- nad development, since reductions in oocyte numbers and delays in maturation of the first oocyte cohort were observed where only a relatively low level of food was available (Borcherding, 1991). The aim of this study was to investigate the influence of food avail- ability, temperature and photoperiod on quantitative aspects of gonad growth and oocyte development in D. polymorpha. This information forms a suitable basis for further- ing our understanding of how the zebra mus- sel might become more widespread not only in Europe, but also in North America. MATERIAL AND METHODS Sampling and Storage In order to conduct experiments with ani- mals at different stages of gonad develop- ment, mussels were collected from the Fúh- linger See at a depth of 2-3 m (FS-2m) before (30 March 1987-experiment E1) and after (21 Sept. 1987-experiment E2) the spawning season (Borcherding, 1991). In the labora- tory, the mussels were cleaned, divided into nine groups of about 70-90 individuals, and stored in well-aereated aquaria (30 |) filled with deionized water containing 1% sea wa- ter and calcium chloride (final concentration = 160 mg Call). The water was changed at least every ten days during the course of each experiment (approx. three months). At the same time, microbial growth was re- moved from the mussels and from the aquar- ium walls. temperature (°C) 5 12 19 3.4 13.3 19.3 0.2 0.5 0.9 0.2 0.5 0.9 6.2 13.6 20.0 0.2 0.6 0.9 0.2 0.6 0.9 Experiments The range of conditions used in the various experimental treatments are outlined in Table 1. Chlamydomonas rheinhardtii served as food and was added once per day. The amount of food available per day for each mussel was calculated as follows: the num- ber of cells added each day and the number of mussels per day in a given aquarium was summed for the total experimental phase. As investigated in two control experiments with- out mussels, 25% of the algae were assumed to be unavailable to the mussels due to their sedimentation. The sum of the cells was di- vided by the sum of the mussels giving the mean number of cells available to each mus- sel per day. This value was multiplied by the energy content of С. rheinhardtii (weight 4.6 + 0.12x10 '' g/cell, with about 50% organic С and an energy content of 45 J/mgC [Finlay & Uhlig, 1981], this equals 1.04x10 © J/cell,), yielding the mean availability of energy to each mussel per day (Table 1). The quantity of algae at the high food level (HFL) should have covered the metabolic demand to- gether with an additional amount of energy for growth. The metabolic demand was cal- culated after respiration data (for methods, see Sprung & Borcherding, 1991) obtained from mussels of the Heider Bergsee that ranged from about 50 ul O,/day in winter to about 500 ul O,/day in early summer (for a mussel of 20 mm shell length; Sprung, un- publ. data). These values equals about 1 to 10 J/day, following the conversion of Wieser (1986) with 20.3 J per ml oxygen. The HFL was equivalent to the mean availability of food in the Heider Bergsee, a lake with sub- optimal feeding conditions (Sprung, 1989; Borcherding, 1991). The low food level (LFL) was about 5% of the HFL and should thus GONAD DEVELOPMENT OF DREISSENA 1% have represented starvation conditions. In- creases in the availability of energy with in- creasing temperatures were based on the respiration data named above, from which the increases in the metabolic requirements of Dreissena were calculated for the experi- mental temperatures. Experiments were per- formed under both long and short day con- ditions (LD 16:8 and LD 8:16; Table 1). Histological Procedures and Measurements Histological techniques had to be used to estimate gonad size on account of the close association between the gonads and other tissues of the visceral sack (Borcherding, 1991). For each experimental group, the soft portions of 16 mussels were divided into the visceral sack (gonads, digestive gland, stom- ach, parts of the digestive tract and adductor muscle, byssus gland) and the remaining body tissues (gills, mantle etc.). Each visceral sack was fixed in Bouin-Allen’s fluid, dehy- drated and embedded in paraffin. The vis- ceral sack was sectioned transversally with a Leitz microtome. Up to 20 sections (10 um) were taken continuously along the visceral sack. These were stained with Mayer’s hae- malaun and eosin and mounted in Canada balsam (Adam & Czihak, 1964). The areas of the gonads and the entire vis- ceral sack from each section were measured with an image analysis system (SIS GmbH, Múnster). The mean tissue areas for each mussel were multiplied by the length of the visceral sack (distance from the first to last section) in order to calculate the volume of the gonads (GV), the volume of the visceral sack (VV) and the gonad index (Gl = propor- tional volume of the gonads in the visceral sack). The image analysis system was also used to evaluate sections of gonad tissue from each female in experiment E1. The diameters of 150 to 200 oocytes with clearly visible nu- clei (to ensure that each section passed through the centre of the oocyte) were mea- sured. The proportional volume of gamete in gonad tissue was estimated from the mean of 10 single gray-scale analyses of each fe- male. Knowing the gonad volume (see above), these data were then used to calcu- late the number of oocytes per female. For the mature oocytes (diameter range 40-65 um) identified within the gonads, a mean oocyte diameter was computed for all sizes above 40 um. Otherwise, the mean diameter of all oocytes was calculated. Details regard- ing these measurements and calculations are given in Borcherding (1990). Data Evaluation and Statistical Procedures Gonad size and the size of other morpho- metric and physiological parameters of iteroparous molluscs are usually closely re- lated to body size (reviewed by Bayne & Newell, 1983). In order to obtain estimates for a mussel with a standard size of 20 mm shell length (SL), regressions of the original data for GV, VV, Gl and the corresponding SL were fitted to the allometric equation y = aSLP; y is the predicted value for GV, VV or Gl, as appropriate, which was calculated for a standard animal using the parameters a and b. The regressions, with 95% confidence intervals, were calculated according to Sachs (1984). Since a regression of oocyte numbers against SL was not possible, the “change in oocyte numbers” (CON) during each experi- ment was estimated as follows. A theoretical number of oocytes was calculated using the SL of each female analysed at the end of the experiments and the regression equation for oocyte numbers in the initial population (the- oretical number of oocytes = 30.72*SL* 14: Borcherding, 1990). The differences between the number of oocytes in the females at the end of the experiment and the theoretical number of oocytes yields the CON. A lack of any overlap between 95% сопй- dence intervals usually allowed a simple sta- tistical investigation of the differences be- tween two means (Sachs, 1984). Otherwise, the means were compared pairwise with the non-parametric Mann-Whitney U-test. The influence of various biotic and abiotic factors on the parameters measured was evaluated by analysis of variance (four-way-ANOVA or three-way-ANOVA). As in all cases the inves- tigated variables depended on a covariate, the equality of variance was tested with the Fmax test of Hartley for the remaining vari- ance of the certain regressions (Underwood, 1981; Sachs, 1984). The results of this test never showed any significant differences be- tween the variances of the certain groups (p > 0.05). Every ANOVA was followed by a sec- ond one, in which the non-significant factors were excluded if the probability of the F-ratio increased to more than 0.25 (Underwood, 1981: 587-588). The final ANOVA of each ex- periment was followed by multiple classifica- 18 BORCHERDING tion analysis (MCA), which evaluated the di- rection and strength of these influences (Schuchard-Ficher et al., 1982). The statisti- cal procedures were calculated on a person- al computer using SPSS/PC+ (Uehlinger, 1988). RESULTS The mean rate at which shell length in- creased in each experimental group was 1.8 um/d (range 0-9.4 um/d) over three months. This rate was low compared to that of pop- ulations in natural environments (e.g. up to 75 um/d in the ljsselmeer, Вл de Vaate; 1991; up to 60 um/d in the River Rhine, Jantz 8 Neu- mann, 1992). Further, there was no tendency towards lower rates of shell growth in larger mussels (e.g. Bji de Vaate, 1991; Jantz 8 Neumann, 1992), and there was no relation- ship between shell growth and temperature or food availability. Mortality was generally low (0-2%) in the course of both experi- ments. An exception was during the first four weeks of one experiment (E1:19°C, НЕЕ, LD 16:8, mortality = 38%), the reason for which was not apparent. Experiments with Mussels Before the Spawning Season (Е1) At the time of their collection (end of March), mussels had well-developed gonads in the final stage of gametogenesis. Water temperature at the sampling site was about 5°C when the experiments at 5°, 12° and 19°С were run. After 90 days, gonad volume (GV) and the gonad index (Gl) of mussels kept at 19°С were significantly lower than those of mussels at lower temperatures (U- test, p < 0.05). Under low food conditions (LFL), both GV and the visceral sack volume (VV) decreased as temperature was in- creased (Fig. 1). Under conditions of high food availability (HFL), GV and Gl were sig- nificantly higher at 12°С than at the higher and lower temperatures (Fig. 1). If the results obtained under different experimental condi- tions at each temperature are compared, the positive influence of HFL on the gonads was significant only for groups at 12°С (U-test, р < 0.05). The results from short-day condi- tions (LD 8:16, tested only at LFL, data not shown) never differed significantly from those under long-day conditions. In the experiments with mussels before 60 $ E © B 40 HFL a © LEL E20 start after 90 days E 60 o = 40 > É HFL a 20 5 LFL 8 start after 90 days a, 30 E E 20 . © > LO HFL 5 QD start after 90 days LFL ics 5 12 19 temperature [°C] FIG. 1. Dreissena polymorpha (20 mm shell length): Gonad volume, visceral sack volume and gonad index of a standardized specimen at the start (30 March 1987) and end (30 June 1987) of the exper- iment in relation to temperature and food level un- der the long photoperiod (LD 16:8). All values were calculated using regression analysis, resulting in the asymmetrical 95% confidence intervals indi- cated by the vertical lines. (E1) as well as after the spawning season (E2) the four-way-ANOVA confirmed that there was no significant influence of the photope- riod (probabilitity of the F-ratios 0.352 to 0.614). After Underwood (1981), those fac- tors can be excluded from the ANOVA, in which the significance of the F-values is p > 0.25 (in E1 this was also the case for the sex with p = 0.714 [GV] and p = 0.646 [Gl], re- spectively). Thus, the results of E1 were fi- nally investigated with a two-way-ANOVA (Table 2). For GV and Gl the total variance of all measurements (n = 96) was significantly influenced by all factors considered in the GONAD DEVELOPMENT OF DREISSENA 19 TABLE 2. Dreissena polymorpha: summary statistics for the two-way-ANOVA and MCA, for all data (n = 96 for each group of values) in E1, describing the influence of the various factors on the dependent parameters GV and Gl. gonad volume (GV) gonad index (Gl) percentage percentage of total of total significance variation significance variation ofF accounted for of F accounted for covariate (SL) p < 0.001 36.4 p < 0.001 1122 temperature p < 0.001 19.4 р < 0.001 27.0 food level р = 0.397 — р = 0.038 2.9 2-way interactions р.= 0.172 — р = 0.052 — multiple r? 0.561 0.412 two-way-ANOVA (p < 0.001). Because there were no significant two-way interactions for both variables (p > 0.05) and because they operate independently on the variables of in- terest (Underwood, 1981), the influence of the main factors can be discussed without any restrictions. Apart from the expected sig- nificant influence of shell length as a covari- ate, temperature was the main influence on variability in both sets of data (Tab. 2). As shown above, the MCA confirmed that a tem- perature of 12°С had the greatest influence on the GV, and on the Gl in particular (Fig. 2). HFL had only a low, but nevertheless signif- icantly positive, effect on the Gl (ANOVA; Ta- ble 2). Factors considered in the ANOVA ac- counted for 56.1% of the total variance for all the GV values, whereby 36.4% of the vari- ance was contributed by SL and 19.4% by temperature. About 44% of the total variance was due to other factors not considered in this analysis. With Gl, this proportion of un- accounted variation was almost 59%, tem- perature was the main influence (279%), fol- lowed by SL (11.2%). Only a small proportion of the variance was accounted for by the availability of food (2.9%; Table 2). Analysis of the oocytes was used to pro- vide information on factors inducing the dif- ferent stages of maturity, and whether these factors might correspond to those controlling gonad size. Even though the oocytes of each mussel can vary to a certain extent under any given set of conditions, a description of the developmental tendencies may assist the recognition of environmental influences. The basis of the following classification into three types was a comparison of the oocytes at the beginning and end of each experiment. Type 1 (minor changes): After three months, the oocyte size frequency distribu- tions showed no clear deviation from the un- imodal distribution in the initial population. The mean diameter of all the oocytes altered only slightly. Type 1 was found only in spec- imens from the experimental groups at 5°C and 12°C (Fig. 3), with no significant differ- ences between these temperatures. HFL had a slightly positive influence on the mean oocyte diameter (OD) and change in oocyte numbers (CON) during the course of the ex- periments. Type 2 (occurrence of mature oocytes): Mature oocytes were clearly visible in the his- tological slides. There was a distinct peak in numbers of large oocytes prior to spawning (Fig. 3). Mature oocytes were not identified in mussels at 5°C. Oocyte maturity remained almost constant at this temperature. At 12°C about 36% of females had a ripe oocyte co- hort, and at 19°C all the females, other than a few individuals of type 3 (see below) be- longed to type 2. In the groups containing ripe oocytes at the end of the experiments at 12°C and 19°C, LFL resulted in a somewhat smaller OD and a further reduction in oocyte numbers (CON) in comparison to HFL. A fur- ther reduction in the number of oocytes, up to their total elimination, was evident from the latter parameter for mussels at 19°C, com- pared with those at 12°C (Fig. 3). Type 3 (reduction in oocyte size): There were no oocytes in the larger size classes, and the mean oocyte diameter decreased significantly. Up to three mussels in each group (except 12°C, HFL, LD 16:8) had sig- nificantly reduced numbers of oocyte. There was no relationship to the level of food avail- ability or photoperiod (not shown in Fig. 3). Since a significant negative correlation was found between OD and CON (Spearman № © gonad index [%] gonad volume [mm] change in oocyte numbers [%] oocyte diameter [ит] S 12 19 temperature [°С] FIG. 2. Dreissena polymorpha: Mean values for go- nad index, gonad volume, change in oocyte num- bers, and oocyte diameter at three temperatures along with the overall means in E1 (dotted lines). All values were adjusted for the independent factors and covariates given in Tables 2 and 3. BORCHERDING Rank Correlation, p = 0.0002), CON was used as the covariate for OD and vice versa. As for the GV and Gl, the photoperiod had no significant influence, and thus, this factor was excluded from the finally investigated ANOVA. There were no significant two-way interactions of the main effects (Table 3). Overall, the two-way-ANOVA demonstrated the significant effects of all the factors as well as the covariates on both these variables (p < 0.001). In addition to the distinct influence of the covariates (25.2% and 21.8%, respec- tively), both temperature and availability of food affected oocyte development signifi- cantly. Altogether, these factors accounted for about 47% of the variance in all the mea- surements (n = 60, for details see Table 3). The MCA indicated that an increase in tem- perature resulted in an increase in oocyte size, accompanied by a decrease in oocyte numbers (Fig. 2). Furthermore, the positive influence of HFL was clearly valid for OD, and to a lesser extent to CON. Experiments with Mussels After the Spawning Season (E2) Mussels used in this experiment were col- lected about one month after the spawning season (during the mussels’ resting stage, cf. Borcherding, 1991) from FS-2m when the water temperature was 18.6°C. During the 90 days of this experiment, the mussels were either maintained at this temperature or at reduced temperatures of 12°С and 5°С (Ta- ble 1). For GV and Gl the total variance of all mea- surements (п = 96) was significantly influ- enced by all the factors considered in the three-way-ANOVA (p < 0.001, photoperiod excluded after the four-way-ANOVA because p = 0.461 [GV] and p = 0.366 [Gl], respec- tively). As there were no significant two-way interactions or even three-way interactions for both variables (p > 0.05), the main factors should operate independently on the vari- ables (Underwood, 1981). At all tempera- tures, gonads in the mussels at HFL were significantly larger than those at LFL (Fig. 4, below). This result was confirmed for GV by the three-way-ANOVA (p < 0.001, Table 4). Apart from the significant influence of shell length as a covariate on both variables (p < 0.001), temperature showed nearly the same effect on gonad growth as in the experiment with mussels prior to spawning (E1). The maximum value at 12°С occurred for Gl un- GONAD DEVELOPMENT OF DREISSENA 21 ARAB EBS care COTES um] 2 RAIN Ps PA hay oocyte size classes [diameter 20 40 0 20 40 0 20 40 start HFL LFL HEL LFL HFL LFL E De 19°C FIG. 3. Dreissena polymorpha (20 mm shell length): Relative frequency of oocytes in the oocyte size classes (y-axes show 4-um classes) of standardized specimens at the start of the experiment (30.3.87) and the different conditions in E1 at the end of the experiment (30.6.87). Types 1, 2, 3: are explained in the text. n = number of females studied, а = mean oocyte diameter for all oocytes, d* = mean diameter of oocytes in the mature fraction only (stripped bars), g = total number of oocytes per mussel (x10%) in the initial population; c = CON, the change in the total number of oocytes relative to the initial population (%). If an exact analysis of the oocytes was not possible, the histograms were estimated and are presented with hatched lines. der all conditions (three-way-ANOVA: p < tions at increasing temperatures also corre- 0.001) and GV under HFL conditions (Fig. 4; sponds with the results of experiment El. three-way-ANOVA: p < 0.001, Table 4). The Overall, there were distinct similarities be- decrease in GV in mussels under LFL condi- tween the trends found in experiments El 22 BORCHERDING TABLE 3. Dreissena polymorpha: summary statistics for the two-way-ANOVA and MCA, for all data (n = 60 for each group of values) in E1, describing the influence of the various factors on the dependent parameters OD and CON. oocyte diameter (OD) change in oocyte numbers (CON) percentage percentage of total of total significance variation significance variation of F accounted for of F accounted for covariate p < 0.001 25:2 р < 0.001 21.8 temperature р = 0.016 11.6 р = 0.001 20.3 food level p = 0.003 9.6 р = 0.040 4.8 2-way interactions р = 0.374 — р = 0.420 — multiple r? 0.464 0.469 and E2. Only food availability had a stronger effect in E2, especially at 5°С. DISCUSSION The extent to which a species can spread, as well as its success in a given environment, is related mainly to those factors which can limit reproduction (Sastry, 1979). Often the causal relationship between these limiting factors and such physiological processes as reproduction can only be evaluated in con- trolled laboratory experiments since the mul- titude of environmental factors in the field may conceal the true relationships. Bayne (1976) named three main aspects of repro- duction limited by environmental factors: ga- metogenesis, larval development, and meta- morphosis into a young adult. The first stage of reproduction, gametogenesis, creates the source of material for the subsequent steps, and should be described not only quantita- tively (e.g. gonad size) but also qualitatively (e.g. stage of maturity). Gonads As expected for experimental groups with limited food availability, the gonad size of D. polymorpha under these experimental condi- tions was always significantly smaller than in the field population at the start of the exper- iment. In order to compensate for the in- creased metabolic requirements with т- creasing temperatures, the supply of energy at 19°C was about 4.5 times that at 5°C (cor- responding to a three-fold increase in the en- ergy requirement for a temperature increase of 10°C). Despite this, gonad volume still de- creased with increasing temperatures at the low food availability (Figs. 1, 4). Thus it is possible that the increased food supply was not sufficient to compensate for the т- creased metabolic demands at higher tem- peratures. The decrease in gonad volume in mussels at 19°C, together with the high level of food availability indicated a similar trend. Despite these reservations in interpreting the data, the following trends were observed. The largest gonad volumes were measured in both experiments after three months at 12°С (although only three temperatures were tested) and at high food availability (Figs. 1, 4). This shows that the results were basically independent of the initial stage of gonad de- velopment. The increase in the gonad index during the course of the experiment in nearly all the groups of mussels kept at 5°С and 12°С indicated intensified gonadal growth compared with other tissues, at lower tem- peratures. In addition, and despite the reser- vations outlined above about the interpreta- tion of the data at 19°С, the distinct negative influence of higher temperature on gonad volume was revealed. The statistical analysis showed that gonad volume and the change in oocyte numbers were influenced in a similar manner (e.g. for temperature, see Fig. 2). This means that go- nad volume was mainly a function of the number of oocytes. Thus it should be possi- ble to compare the results of the present study with measurements of eggs spawned by Mytilus edulis after storage under various conditions, as reported by Bayne et al. (1978). These authors demonstrated a higher fecundity in mussels at 11°С than at 18°С, and a reduction in the number of eggs spawned when no food was available. To- gether with estimations of the “scope for growth,” Bayne et al. (1978) concluded that GONAD DEVELOPMENT OF DREISSENA 23 © SI 3 LFL Я 20 TD = 5 oD 10 start after 90 days fsa! | Е 45- E e S 30 - % HEL Я = 15 - — 3 LFL iS Start after 90 days i T T % = Я 10 — . = > HFL 5. 5b start after 90 days LFL T T | 5 12 19 temperature [°C] FIG. 4. Dreissena polymorpha (20 mm shell length): gonad volume, visceral sack volume, and gonad index of a standardized specimen at the start (21 Sept. 87) and end (23 Dec. 87) ofthe experiment, in relation to temperature and food level under the short photoperiod (LD 8:16). All values were calcu- lated using regression analysis, resulting in the asymmetrical 95% confidence intervals indicated by the vertical lines. fecundity in M. edulis depends mainly on the energy available for gamete production. This conclusion seems to be valid for D. polymor- pha as well because the results indicate that higher fecundity (i.e. gonad size or number of oocytes) was related to a sufficient supply of energy. Sastry (1968), working on A. irradians, re- ported slightly higher gonad indices (the con- tribution of gonad to the body weight) at 15°C and 20°C in fed mussels, compared to starved mussels. However, these differences were not significant. In the present study, go- nad indices based on volume or weight were not sufficient to reveal all the trends. For ex- ample, if E2 values for the gonad index only were taken into account, then the significant differences between gonad volume in the mussels at HFL and LFL at 5°C and 19°C would not have been recognized (Fig. 4). The conclusive evidence for the significant influ- ence of food availability on gonad develop- ment could only be drawn from the absolute values (Table 4). On the other hand, the sim- ilarity in gonad indices, along with different gonad volumes (e.g. E2: 19°С conditions, Fig. 4), implies that the gonads are supported at the cost of other tissues under conditions of environmental stress (i.e. low food avail- ability). In contrast to temperature and food avail- ability, photoperiod never had a significant influence on the gonad development in D. polymorpha (Tables 2-4). The results of E2 (Fig. 4) were similar to those of Gimazane (1971, cited by Sastry, 1979), who found that photoperiod had no significant effect on ga- metogenesis in Cerastoderma [= Cardium] edule with gonads at the resting stage. Pho- toperiod also had no influence on the gonads and oocytes in D. polymorpha at the end of gametogenesis (E1: Tables 2, 3). However, Bohlken & Joosse (1982) reported that an LD 16:8 induced a relatively early maturation of the female reproductive system and a high rate of egg production in the gastropod Lym- naea stagnalis. The possibility of photoperiod inducing similar effects in the zebra mussel, for instance under conditions of high food availability or longer experimental periods, can only be clarified with appropriate exper- iments. Oocytes Information on the gonad size only 1$ not sufficient for assessing the different stages of maturity. A better evaluation is provided by the mean oocyte diameter (of either all the oocytes or just the пре oocytes), which can be used to approximate the stage of maturity (Sastry, 1979). Food availability influenced oocyte size in the same manner as gonad volume, but the influence of temperature was totally different on both the above-mentioned variables in D. polymorpha from E1. Oocyte diameter, which 15 related to the stage of ma- turity, increased with temperatures (Fig. 3). This was the opposite effect to that wit- nessed for gonad size and the change in 24 BORCHERDING TABLE 4. Dreissena polymorpha: summary statistics for the three-way-ANOVA and MCA, for all data (n = 96 for each group of values) in E2, describing the influence of the various factors on the dependent parameters GV and Gl. gonad volume (GV) significance of F covariate (SL) p < 0.001 temperature p < 0.001 food level p < 0.001 sex 21005] 2-way interactions (t-f) р = 0.051 2-way interactions (t-s) p = 0.893 2-way interactions (f-s) р = 0.518 3-way interactions SAS multiple r? oocyte numbers (Fig. 2). This indicates that the maturation of oocytes, even if it was only a low portion of the total number, was re- stricted to higher temperatures (a minimum of 12°C for the temperatures tested). Similar conclusions have been drawn for Crassostrea virginica (Loosanoff 8 Davis, 1952) and A. irradians (Sastry, 1966). Sastry (1968) reported that bay scallops from North Carolina developed oogonia when exposed to a sub threshold temperature of 15°C, but oocyte growth did not occur even though they were supplied with ample food. After transferring these scallops to higher temper- atures (20°C and 25°C), oocyte growth be- gan immediately when sufficient food was available (Sastry, 1968). However in bay scal- lops from Massachusetts, the cytoplasmatic growth phase of oocytes was initiated at 15°C, at 5°C only oogonia developed (Sastry & Blake, 1971). Sastry (1970) suggested that such variations between populations may be an adaptive response to geographic differ- ences in temperature and food production. In contrast to A. irradians, the zebra mussel was able to develop only a fraction of its oocytes to maturity at temperatures of 12°C (Fig. 3), even when the availability of food was so low that gonad size was reduced to less than 30% of its initial value (Fig. 1). This might occur at the expense of body reserves (e.g. in M. edulis, Gabbott & Bayne, 1973; Gabbott, 1975). On the other hand, the de- crease in oocyte numbers and the maturation of parallel oocyte cohorts (Fig. 3) suggested that some of the oocytes were reabsorbed in order to support a remaining, smaller portion of the oocytes. The possibility of oocyte re- gonad index (Gl) percentage percentage of total of total variation significance variation accounted for of F accounted for 48.3 p < 0.001 23.9 9.6 p = 0.001 13.0 9.0 р = 0.049 2.9 —— pl 051185 — — р = 0.428 — = p = 0.831 — = p = 0.669 — — р = 0.867 — 0.672 0.399 sorption in D. polymorpha was discounted by Walz (1978), although gonad size and oocyte numbers were not evaluated, casting a doubt on the conclusions of Walz' study. The re- sorption of oocytes in response to environ- mental stress is a common adaptation in many bivalves (e.g. С. virginica, Loosanoff & Davis, 1951; А. irradians, Sastry, 1966; М. edulis, Bayne et al., 1978, 1982). It was also described recently in D. polymorpha during periods of starvation (Sprung & Borcherding, 1991), with electron microscopy providing di- rect evidence for resorption processes dur- ing the same experiments (Bielefeld, 1991). Oocytes in the mussels studied in E2 were poorly developed at the end of the experi- ment, which made an extensive analysis of the oocytes difficult. Using microscopy, it was possible to gain an impression of the stage of oocyte development. This confirmed that nearly all the trends outlined above for the gonad volume (maximum values at 12°С, minimum values at 19°C, a negative influ- ence of LFL) were also valid for the stage of oocyte maturity. This is in clear contrast to the situation in E1 (Fig. 2). In autumn, at the onset of the gametogenetic cycle in D. poly- morpha, the increase in gonad volume could be attributed mainly to the proliferation of new germ cells, and only to a small extent to the growth of oocytes (Borcherding, 1991). An identical situation occurred with the mus- sels in E2. Gonad size increased but, in con- trast to E1, the oocytes did not mature at 12°C and 19°C. Two factors may have been responsible: (1) an endogenous component and/or (2) the necessity for low temperatures during a certain phase of the annual repro- GONAD DEVELOPMENT OF DREISSENA 25 ductive cycle, perhaps to initiate or synchro- nize certain gametogenetic processes in D. polymorpha. To summarize, it appears that fecundity (i.e. number of oocytes, gonad volume) was influenced mainly by food availability under the different experimental conditions, both in spring and in autumn. However, maturation of the oocytes (i.e. their size) was affected positively by increased temperatures only in spring prior to spawning, while in autumn there was no effect on the stage of maturity for mussels at the onset of gametogenesis. Conclusions (1) Temperature: Borcherding (1991) pointed out that environmental temperature across the year may limit the further spread of D. polymorpha in three ways. First, if tempera- tures remain above 12”C throughout the year, oocyte maturation and spawning may be- come desynchronized within a population (possibly the stimulus for initiating maturation is lacking), thus an important prerequisite for fertilization with this type of reproduction would be lost. Second, temperatures fail to rise above the apparent threshold of 12°С in cold monomictic lakes, then spawning cannot occur and fertilization of the eggs would not be possible (Sprung, 1987). Third, a low am- plitude in annual temperatures might be in- sufficient for temporal control (e.g. deep-sea species, Mackie, 1984), whereas high ampli- tudes might be unfavourable if rates of res- piration and ingestion are not balanced during periods of rapidly changing temperatures (Bayne & Newell, 1983). The results of the experiment with mussels at the onset of ga- metogenesis (E2, 19°С) appear to correspond with the first suggestion. Although high tem- peratures can support oocyte maturation (if gonads are in another phase of the gameto- genetic cycle, cf. E1), the growth of oocytes was reduced at high temperatures. Possibly a spell of low temperatures is necessary after spawning in order to initiate a new gameto- genetic cycle. On the other hand, the results of E1 revealed that oocytes did not mature at low temperatures (5°C), thus higher temper- atures (here a 12°C threshold) are required prior to spawning, which confirms to the sec- ond of these suggestions. (2) Food Availability: Fecundity in D. poly- morpha was undoubtedly reduced under conditions of low food availability (cf. Bor- cherding, 1992). This was confirmed by the increased reduction in oocyte numbers under reduced food availability (indicated in E1). However, low food availability alone, with no unfavourable temperature conditions (e.g. rapidly changing temperatures), was not suf- ficient to prevent the first stage of reproduc- tion. The results of E1 and other starvation experiments (Sprung & Borcherding, 1991) showed that a small portion of the oocytes in D. polymorpha could mature under adverse conditions of food availability. Thus, low food concentrations alone may restrict, but do not prevent, the spread of D. polymorpha in such an environment. A restriction in the spread of this species would appear to be influenced only by the ambient temperature conditions. ACKNOWLEDGEMENTS This study formed part of a Ph. D. thesis submitted to the University of Cologne. My cordial thanks are due to Prof. Dr. D. Neu- mann for suggesting the theme, for his sup- port and for discussions in the course of the work. | am grateful to Dr. M. Sprung for con- tributing numerous ideas to this study and Dr. F. Bairlein for help with statistics. Thanks are also due to Dr. D. Fiebig for improving the English text and to two anonymous review- ers. This study was supported by a grant from the Deutsche Forschungsgemeinschaft to Prof. Dr. D. Neumann (contract No.: Ne 72/26-1) and a Graduierten Stipendium from Nordrhein Westfalen to the author. LITERATURE CITED ADAM, H. 4 G. CZIHAK, 1964, Arbeitsmethoden. Gustav Fischer Verlag, Stuttgart. BAYNE, B. L., 1976, Aspects of reproduction in bivalve molluscs. Pp. 432-448, in: м. WILEY, ed., Estuarine processes, Vol. 1. Academic Press, New York. BAYNE, В. L., A. BUBEL, Р. A. GABBOTT, D. В. LIVINGSTONE, D. М. LOWE & М. М. MOORE, 1982, Glycogen utilisation and gametogenesis in Mytilus edulis L. Marine Biology Letters, 3: 89- 105. BAYNE, B. L., D. L. HOLLAND, M. N. MOORE, D. М. LOWE & J. WIDDOWS, 1978, Further studies on the effects of stress in the adult on the eggs of Mytilus edulis. Journal of the Marine Biological Association of the United Kingdom, 58: 825-841. BAYNE, В. L. & В. С. NEWELL, 1983, Physiological energetics of marine mussels. Pp. 407-515, in: 26 BORCHERDING K. M. WILBUR, ed., The Mollusca. Vol. 4, Physiol- ogy, Part 1. Academic Press, New York. BIELEFELD, U., 1991, Histological observations of gonads and digestive gland in starving Dreissena polymorpha. Malacologia, 33: 31-42. BlJ DE VAATE, A., 1991, Distribution and aspects of population dynamics of the zebra mussel, Dreissena polymorpha (Pallas, 1771), in the lake lisselmeer area (The Netherlands). Oecologia, 86: 40-50. BOHLKEN, S. & J. JOOSSE, 1982, The effect of photoperiod on female reproductive activity of the freshwater pulmonate snail Lymnea stagnalis kept under laboratory breeding conditions. /nter- national Journal of Invertebrate Reproduction, 4: 213-222. BORCHERDING, J., 1990, Die Reproduktionsleis- tungen der Wandermuschel Dreissena polymor- pha. Ph.D. thesis, University of Cologne. BORCHERDING, J., 1991, The annual reproduc- tive cycle of the freshwater mussel Dreissena polymorpha Pallas in lakes. Oecologia, 87: 208- 218. BORCHERDING, J., 1992, Morphometric changes in relation to the annual reproductive cycle in Dreissena polymorpha—a prerequisite for bio- monitoring studies with zebra mussels. Pp. 87- 99, in: D. NEUMANN & Н. A. JENNER, eds., The zebra mussel Dreissena polymorpha: ecology, biologi- cal monitoring and first applications in the water quality management. Limnologie Aktuell, Vol. 4. Gustav Fischer Verlag, Stuttgart, Jena, New York. FINLAY, B. J. 8 G. UHLIG, 1981, Caloric and car- bon values of marine and freshwater protozoa. Helgolánder Meeresuntersuchungen, 34: 401- 412 САВВОТТ, P. A., 1975, Storage cycles in marine bivalve molluscs: an hypothesis concerning the relationship between glycogen and gametogen- esis. Pp. 191-211, in: H. BARNES, ed., Proceed- ings, 9th European Symposium on Marine Biol- ogy University Press, Aberdeen, Scotland. САВВОТТ, Р. А. & В. L. BAYNE, 1973, Bio- chemical effects of temperature and nutritive stress on Mytilus edulis L. Journal of the Marine Biological Association of the United Kingdom, 53: 269-286. GIMAZANE, J. P., 1971, Introduction à l'étude ex- périmentale du cycle sexuel d'un mollusque bi- valve Cardium edule L. Analyse des populations, évolution de la gonade et action de quelques facteurs: nutrition, temperature, photoperiode. Ph.D. thesis, University of Caen. HEBERT, P. D. N., B. W. MUNCASTER 4 G. L. MACKIE, 1989, Ecological and genetic studies on Dreissena polymorpha (Pallas): a new mollusk in the Great Lakes. Canadian Journal of Fisheries and Aquatic Sciences, 46: 1587-1591 JANTZ, В. & D. NEUMANN, 1992, Shell growth and aspects of the population dynamics of Dreissena polymorpha in the River Rhine. Pp. 49-66, in: D. NEUMANN & Н. A. JENNER, eds., The zebra mus- sel Dreissena polymorpha: ecology, biological monitoring and first applications in the water quality management. Limnologie Aktuell, Vol. 4. Gustav Fischer Verlag, Stuttgart, Jena, New York. LOOSANOFF, V. |. 8 Н. С. DAVIS, 1952, Temper- ature requirements for maturation of gonads of northern oysters. Biological Bulletin (Woods Hole), 103: 80-96. MACKIE, G. L., 1984, Bivalves. Pp. 351-418, in: K. M. WILBUR, ed., The Mollusca. Vol. 7, Repro- duction. Academic Press, New York. NALEPA, T. F. & D. W. SCHLOESSER, 1993, Zebra mussels—Biology, impacts, and control. Lewis Publishers, Boca Raton, Ann Arbor, London, To- kyo. NEUMANN, D. 8 H. A. JENNER, 1992, The zebra mussel Dreissena polymorpha: ecology, biologi- cal monitoring and first applications in the water quality management. Limnologie Aktuell, Vol. 4, Gustav Fischer Verlag, Stuttgart, Jena, New York ROBERTS, L., 1990, Zebra mussel invasion threat- ens U.S. waters. Science, 294: 1370-1372 SACHS, L., 1984, Angewandte Statistik. -6. Aufl. Springer Verlag, Berlin Heidelberg New York. SASTRY, A. N., 1966, Temperature effects in re- production of the bay scallop, Aequipecten irra- dians Lamarck. Biological Bulletin (Woods Hole), 130: 118-134. SASTRY, A. N., 1968, Relationships among food, temperature and gonad development of the bay scallop, Aequipecten irradians Lamarck. Physio- logical Zoology, 41: 44-53 SASTRY, A. N., 1970, Environmental regulation of oocyte growth in the bay scallop, Aequipecten irradians Lamarck. Experentia, 26: 1371-1372 SASTRY, A. N., 1979, Pelecypoda (excluding Os- treidae). Pp. 113-292, in: А. С. GIESE & J. $. PEARSE, eds., Reproduction of marine invertebrates, Vol. V. Academic Press, New York. SASTRY, А. М. & М. J. BLAKE, 1971, Regulation of gonad development т the bay scallop, Ае- quipecten irradians Lamarck. Biological Bulletin (Woods Hole), 140: 274-283. SCHUCHARD-FICHER, C., K. BACKHAUS, U. HUMME, W. LOHBERG & W. PLINKE, 1982, Multivariate Analysenmethoden. Springer Verlag, Berlin. SPRUNG, M., 1987, Ecological requirements of developing Dreissena polymorpha eggs. Archiv für Hydrobiologie/Supplement, 79: 69-86. SPRUNG, M., 1989, Field and laboratory observa- tions of Dreissena polymorpha larvae: abun- dance, growth, mortality and food demands. Ar- chiv für Hydrobiologie, 115: 537-561. SPRUNG, M. & J. BORCHERDING, 1991, Physio- logical and morphometric changes in Dreissena polymorpha (Mollusca; Bivalvia) during a starva- tion period. Malacologia, 33: 179-191. STÄNCZYKOWSKA, A., 1977, Ecology of Dreis- sena polymorpha (Pall) (Bivalvia) in lakes. Polish Archive of Hydrobiology, 24: 461-530. GONAD DEVELOPMENT OF DREISSENA 27 UEHLINGER, H.-M., 1988, SPSS/PC+ Benutzer- laboratory experiments and in lake Constance. handbuch, Vol. 1. Gustav Fischer Verlag, Stutt- Il. Reproduction. Archiv für Hydrobiologie/Sup- gart—New York. plement, 55: 106-119. UNDERWOOD, A. J., 1981, Techniques of analysis WIESER, W., 1986, Bioenergetik—Energietransfor- of variance in experimental marine biology and mationen bei Organismen. Thieme, Stuttgart, ecology. Oceanography and Marine Biology, an New York. Annual Revue, 19: 513-605 WALZ, N., 1978, The energy balance of the fresh- water mussel Dreissena polymorpha Pallas in Revised Ms accepted 25 January 1994 PA em че . = u } ie u > > Fe o © ALL Y Er ta > > : ww. MALACOLOGIA, 1995, 36(1-2): 29-41 RELAXING TECHNIQUES FOR FRESHWATER MOLLUSCS: TRIALS FOR EVALUATION OF DIFFERENT METHODS В. Araujo, J. М. Remon, D. Moreno & M. A. Ramos Museo Nacional de Ciencias Naturales (CSIC), José Gutiérrez Abascal, 2, 28006 Maarid, SPAIN ABSTRACT Twelve different methods of relaxing freshwater molluscs were tested to find the most suit- able for future research and conservation in scientific collections. In addition to drowning, different concentrations of the following agents were tested: phenoxyethanol, MS 222, clove oil, pentobarbital, sodium pentobarbital, diethylether, chloroform, and urethan. Also menthol, lime tree, and valerian were tried. Tests were made with species of main groups of freshwater molluscs: Pisidium amnicum, Corbicula fluminea, and Unio sp. among bivalves; Melanopsis sp., Bithynia tentaculata, Valvata piscinalis, Potamopyrgus jenkinsi, Pseudamnicola cf. luisi, and Horatia sturmi among prosobranch gastropods; and Lymnaea peregra and Ancylus fluviatilis among pulmonate gastropods. Relaxation condition of specimens after narcosis, response to fixative fluids, action time and availability of narcotic agents were considered for evaluation. There was considerable variation between species in their susceptibility to narcotic agents, suggesting that many factors may be involved in the response of freshwater molluscs to narcotization. Among tested agents, sodium pentobarbital and especially, pentobarbital, were the most suitable for relaxing freshwater molluscs. No overdosing troubles were registered in trials with pentobarbital. Results with menthol were unpredictable, although it may be used over a wide range of species. Key words: freshwater molluscs, pulmonates, prosobranchs, bivalves, narcotization, relaxing agents, techniques, fixation. INTRODUCTION Anatomical studies in such areas as taxon- omy, physiology, ecology and systematics of molluscs, and the increased interest in mor- phometrics (Meier-Brook, 1976b; Emberton, 1989), have generated a need for improved relaxing techniques for the study of molluscs. The preservation of specimens in a life-like position is also important for natural history collections. A search of the literature yielded only a few papers dealing specifically with methods for relaxing freshwater molluscs. Runham et al. (1965) reviewed the results of previous au- thors while trying both narcotics and anaes- thetics on different species of freshwater and terrestrial gastropods, showing that methods vary considerably in effectiveness among species. Meier-Brook (1976a) compared the effectiveness of two different agents (pento- barbital and sodium pentobarbital) on spe- cies of Basommatophora, Prosobranchia and Bivalvia. Meier-Brook (1976b) tested the effect of different levels of extension after narcosis on the anatomical measurements of a Planorbis species. More recently, Girdle- 29 stone et al. (1989) used different concentra- tions of three volatile products to reversibly anaesthetize specimens of Lymnaea stagna- lis (Linnaeus, 1758). The experiences of previous authors to- gether with our own, led us to carry out ten- tative tests in order to obtain more information before designing the present experiment. Since the aim is to obtain properly extended animals for immediate dissection and/or fix- ation, selection of methods does not take into consideration whether animals might be al- lowed to recover, as other authors have re- quired (Michelson, 1958; Girdlestone et al., 1989). Therefore, we employ the term narco- tization in the sense of Runham et al. (1965). According to them, the best narcotic will relax the animals “in as life-like a position as pos- sible and to such an extent that they do not contract when fixed.” Therefore, we tested several narcotics in different concentrations to various species ofthe main groups of fresh- water molluscs. Some methods previously cited for narco- tization of marine molluscs (Smaldon & Lee, 1979) or other animals (Lincoln & Sheals, 1979) are used in this paper for the first time 30 ARAUJO, REMÓN, MORENO 8 RAMOS for freshwater molluscs. “Drowning” was also tried to test the effect of the progressive lack of oxygen on some species. From the eleven species tested, only Lymnaea peregra (Müller, 1774), Bithynia tentaculata (Linnaeus, 1758), and Potamopyrgus jenkinsi (Smith, 1889) were previously used in similar studies by other researchers. The purpose of this investigation was to discover the most suitable relaxing method for each species or species group of fresh- water molluscs, with special attention to the repeatibility of the procedures in future re- search. Relaxation (degree of extension) of speci- mens after narcosis, response to fixative flu- ids, action time and availability of the narcotic agents, were all considered in evaluation of their effectiveness. MATERIALS AND METHODS Before planning a definitive protocol, we tried several products and concentrations, as well as different fixation routines and labora- tory conditions. These early experiences will be called “pretests” in this paper, and the subsequent ones will be simply called “tests.” Pretests These were carried out with menthol, add- ing crystals to cover the water surface of the jar, and with sodium pentobarbital (1% and 2%), urethane (2% and 4%), MS 222 (0.05%, 0.1% and 0.2%), clove oil (from 3 to 30 drops depending on the water volume), phenoxy- ethanol (1%), magnesium chloride (7.5%), magnesium sulphate (7.5%) and chloral hy- drate (5%) among narcotic agents. The vol- ume of the glass jars was twice that of the tested solution (7 ml, except 60 ml for Unio sp.) and solutions were prepared with de- chlorinated tap water (for exceptions see re- sults). Pretests were made in covered jars in a refrigerator (10-13°С). Menthol was also tested at room temperature (20-28°С). The effect of “drowning” over the specimens was proved in 15 ml of water (100 ml for Unio sp.). Pretests were regularly checked in order to monitor changes in the general aspect of an- imals. After checking the lack of response to mechanical stimulus (by use of a needle), specimens were killed by different methods: 70% ethanol at room temperature and at 12°C below zero (+ 2°C), liquid nitrogen after ten seconds and hot water (60°С) after five seconds. Fixation was always carried out in 70% ethanol. Five specimens of Potamopyr- gus jenkinsi, three of Valvata piscinalis (Müller, 1774), two of Bithynia tentaculata and Pisidium amnicum (Múller, 1774) and one of Corbicula fluminea (Múller, 1774) and Unio sp., depending on specimen availability, were used in each test. Tests The results of the pretests were used as the basis for choosing the agents and estab- lishing the conditions for the following tests: Several species of the main groups of freshwater molluscs were chosen in order to test if responses to narcotic agents were sim- ilar within each group. Trials were made with Pisidium amnicum, Corbicula fluminea, and Unio sp. among bivalves; Melanopsis sp., Bithynia tentaculata, Valvata piscinalis, Pota- mopyrgus jenkinsi, Pseudoamnicola cf. luisi Boeters, 1984, and Horatia sturmi (Rosen- hauer, 1856) among prosobranch gastro- pods; and Lymnaea peregra and Ancylus fluviatilis (Muller, 1774) among pulmonate ba- sommatophoran gastropods. No species of the family Planorbidae were tested as they had been included in studies by Michelson (1958) and Meier-Brook (1976a, b). Voucher specimens are deposited in collections of the Museo Nacional de Ciencias Naturales, Madrid, Spain. Bithynia tentaculata, Valvata piscinalis, Potamopyrgus jenkinsi, Lymnaea peregra, Pi- sidium amnicum and Corbicula fluminea were sampled from the Miño River (Ponteve- dra, Spain) in January 1992. Specimens of Unio sp. were captured at the Gasset reser- voir (Ciudad Real, Spain) in December 1991. Pseudamnicola cf. luisi and Ногайа sturmi were sampled at Pilar del Mono fountain (Granada, Spain) in February 1992. Speci- mens of Melanopsis sp. were collected at the Molí fountain (Alicante, Spain) in January 1992, and Ancylus fluviatilis from the Perales River (Madrid, Spain. November, 1991). Ani- mals were transported alive to the laboratory in plastic jars inside a portable refrigerator with ice, and artificial aeration was provided for five seconds every eight hours. At the lab- oratory they were kept in aquaria at 13°С with a 12-hour artificial day-night cycle and aeration until tests were carried out. The RELAXING TECHNIQUES FOR FRESHWATER MOLLUSCS 31 TABLE 1. Number of days each species was kept in aquaria REPEATED TEST TEST Melanopsis sp. 10 12-24 Valvata piscinalis 6 9 Bithynia tentaculata 27 31-39 Potamopyrgus jenkinsi 13 20 Pseudamnicola cf. luisi 16 24 Horatia sturmi 3 — Ancylus fluviatilis 2 3 Lymnaea peregra 6 10-18 Unio sp. 1 21 Pisidium amnicum 12 19-41 Corbicula fluminea 24 34 number of days each species was kept under this cycle varied, and 15 indicated in Table 1. Pretests indicated that drowning might be effective in some species. This was proved for specimens of Melanopsis sp, which were submerged in water in a covered jar avoiding air bubble formation. Eleven narcotics were tested (Table 2). Five of them, sodium pento- barbital, pentobarbital, MS 222, phenoxyeth- anol and urethane were used in different con- centrations (see Table 2). Doses of clove oil are explained in Table 2. Diethylether and chloroform were also tested with Melanopsis sp. because this species group proved to be very resistant to narcotization. Trials with menthol were made as in pretests, adding crystals to cover the water surface of the jar. No records were found in the literature indi- cating the use of valerian, lime tree, clove oil, diethyleter, chloroform, phenoxyethanol nor MS 222 in narcotization of freshwater mol- luscs. Tested products, and when necessary, commercial names and firms are: Pentobarbital and sodium pentobarbital (nembutal) are manufactured by Sigma Chemical Co. Ethyl M-Aminobenzoate is manufactured by Sandoz under the name MS 222. Phenoxyethanol is manufactured by Merck under the name Ethyleneglycolmonophe- nylether. Urethane is manufactured by Fluka Che- mie AG. Lime tree is Tilia sp. Valerian is Valeriana officinalis (Linnaeus, 1758). Menthol Cryst is manufactured by Merck. Clove tree oil is from Syzygium aromaticum (Linnaeus, 1758). Diethylether is manufactured by Riedel-de Haén. Chloroform is manufactured by Probus S.A. Water used for dilutions came from the same site as the species tested, except for Horatia sturmi and Pseudamnicola cf. luisi, where the water came from the Miño River and from dechlorinated tap water, respec- tively. Dilutions were prepared as soon as samples reached the laboratory and stored in a refrigerator at 10-13°C, being transferred to room temperature 12 hours before each ex- periment. The standard volume of dilutions was 8 ml, except for larger species, such as Unio sp. and Corbicula fluminea, where the volume was 150 ml and 37 ml, respectively. One specimen of Melanopsis sp. and Unio sp., two of Bithynia tentaculata, Horatia sturmi, Pisidium amnicum and Lymnaea pe- regra, and three ofthe remaining species were tested. Tests were made in covered glass jars with a total volume of 16 ml, except for Cor- bicula fluminea and Unio sp., in which jar vol- umes were of 82 and 240 ml respectively. In order to stardardize the experiment and to avoid maceration after full relaxation, we carried out experiments at 6-10°C in all cases except in Melanopsis sp., where, due to difficulties to relax it, tests were also done at 15-18°C. All tests were simultaneously done for each species. Once the complete set of tests was concluded with each spe- cies, only those drugs and concentrations yielding the best results were subsequently used to repeat the experiment. Finally, the most successful method was used in a third test to narcotize all the remaining specimens of each species sample. The experiments were regularly checked, as in pretests, in order to record any changes in the animals (with a stereomicroscope when necessary). After we checked for lack of response to mechanical stimulus, we fixed specimens in ethanol at —12°C (+2°С). Criteria for Evaluation The maximum extension (relaxation) achieved with each method was observed and quantified according to the following code (Tables 3-5): 4 = very good (animal fully extended and sometimes a little turgid or swollen), 3 = good (not fully extended and sometimes wrinkled), 2 = fair (only part ofthe foot visible outside the shell), and 1 = bad 32 ARAUJO, REMÓN, MORENO 8 RAMOS TABLE 2. Chemical products and concentrations as weight percentage’ Sodium Pento- Pento- Phenoxy- Lime Menthol Clove barbital barbital MS 222 ethanol Urethane tree Valerian cryst oil Diethylether Chloroform 0.400% 2.000% 0.20% 1.00% 20% 1.0% 1.0% (*) (&) (55) (5) 0.200% 1.000% 0.10% 0.50% 1.0% 15 drops 0.100% 0.500% 0.05% 0.5% 10 drops 0.050% 0.250% 5 drops 0.025% 0.125% "Phenoxyethanol is as volume percentage. (*): Enough to cover vial surface. (**): Except for Unio sp (60, 40 and 20 drops) and Corbicula fluminea (30, 20 and 10 drops). (+): Only tested with Melanopsis sp, see literature. (animal withdrawn inside the shell). These codes refer to all specimens tested, with the exceptions quoted in the results section. п the case of 2, 3 and 4 (fair, good and very good extension), the time consumed by each method (action time) was registered and quantified as follows: 4 < 24 hours, 3 = 24-48 hours, 2 = 48-72 hours and 1 > 72 hours. The response of specimens to the fixation after one minute (fixation |) and after 24 hours (fix- ation Il) was also codified in the case of 2, 3 and 4, as follows: 4 = very good (no retrac- tion, without modification), 3 = good (slight retraction), 2 = fair (large retraction) and 1 = bad (animal withdrawn). All data and inci- dences of the experiment were registered in specially designed forms. RESULTS Magnesium chloride, magnesium sulfate and chloral hydrate were not successful when tested on Pisidium amnicum, Corbicula fluminea, Unio sp., Bithynia tentaculata, Val- vata piscinalis and Potamopyrgus jenkinsi, and therefore were rejected for subsequent tests. Urethane (4%) was only successful for Unio sp. and was lethal for smaller species; in the remaining tests it was used at lower con- centrations. Freezing was also tried with some hydro- biid species. Frozen animals were fully ex- panded and did not retract when fixed in 70% ethanol, contrary to that observed by Runham et al. (1965) where calcium formalin was used. However, this technique was dis- carded since it resulted in serious damage to the skin of the animal and consequent loss of external morphological characters of taxo- nomic interest. From all the methods employed for killing specimens, sudden immersion of relaxed specimens in hot water (60°C) before fixa- tion, seemed most effective in avoiding ani- mal retraction. This weakens the columelar muscle of gastropods, enabling easy extrac- tion of the animal without breaking the shell. However, this method was not used in further tests as it is suspected of causing internal tissue disturbances. Submersion in liquid ni- trogen was also rejected since in species tested (Potamopyrgus jenkinsi, Unio sp., and Pisidium amnicum) it damaged the skin ofthe animal. For subsequent tests, cold ethanol was used in the fixation of specimens to avoid retraction caused by ethanol at room temperature. A complete reference to results can be found in Tables 3-5. Results on each species are discussed below for those cases when narcosis was good or very good, according to the codes specified in the previous section and in the Tables. For those species in which pretests were made, results are described following the results of the tests. Melanopsis sp. (Table 3) Specimens of Melanopsis sp. are difficult to relax. Results obtained under the same conditions as those of other species tested were very poor. Consequently, tests were re- peated at different temperatures trying such additional drugs as diethylether and chloro- form. The best results were obtained using 0.25% sodium pentobarbital and 0.05% pen- tobarbital, narcotization being good at 48 and 72 h, respectively. Sodium pentobarbital (0.25%), yielded a very good fixation | and a fair fixation |. When this test was repeated with 15 adult specimens and four juveniles in RELAXING TECHNIQUES FOR FRESHWATER MOLLUSCS 33 TABLE 3. Results with prosobranch gastropods Valvata Bithynia Potamopyrgus Pseudamnicola Melanopsis sp. piscinalis tentacula jenkinsi cf. luisi Horatia sturmi PENTOBARBITAL 0.400% 3.34% 2.1.22 3244 4.4.4.4 3.4.4.3 4.3.4.4 0.200% 2.3.4.1 4.1.4.4 3.2.4.3 4.4.4.4 4.4.4.4 4.3.4.4 0.100% 1.-.-.- 4.1.3.3: 4.1.4.8 4.3.4.4 4.3.4.4 4.3.4.4 0.050% 3.2.4.4 4.1.4.4. 23.344 4.3.4.4 1.-.-.- 4.3.4.4 0.025% 1.-.-.- 1.-.-.- 1.-.-.- 1.-.-.- 1.-.-.- 1.-.-.- SODIUM PENTOBARBITAL 2.000% 1.-.-.- 1.-.-.- 1.-.-.- 1.-.-.- 1.-.-.- 2.2.4.4 1.000% 2.3.4.4 1.-.-.- 1.-.-.- 1.-.-.- 1.-.-.- 2.2.4.4 0.500% 2.3.4.4 4.1.3.3 1.-.-.- 4.4.3.3 2.1.2.2 4.2.4.3 0.250% 3.3.4.4 4.1.4.4 4.1.4.3 4.3.4.4 4.2.4.4 4.2.4.4 0.125% 3.3.4.3 4.1.4.4 4.1.4.4 4.2.4.4 4.1.4.4 4.2.4.4 MS 222 0.20% 1.-.-.- 3.1.44. 2243 1.-.-.- 2.2.4.3 1.-.-.- 0.10% 1.-.-.- 1.-.-.- 4.1.3.2 1.-.-.- 2.1.3.3 3.2.4.3 0.05% 1.-.-.- 1.-.-.- Ale 1.-.-.- 29.3 2183.3 PHENOXYETHANOL 1.00% 1.-.-.- 1.-.-.- 1.-.-.- 1.-.-.- 2.2.3.3 23.8383 0.50% 1.-.-.- 1.-.-.- 1.-.-.- 1.-.-.- 32.838 33838 URETHANE 2.0% 1.-.-.- 1.-.-.- 4.1.4.4 3432 2.1.4.4 3.3483 1.0% 1.-.-.- 1.-.-.- 4.1.4.2 1.-.-.- 243.3 3.2.4.3 0.5% 1.-.-.- 1.-.-.- 1.-.-.- 1.-.-.- 1.-.-.- 1.-.-.- LIME TREE 1.0% 2.2.4.4 1.-.-.- 1.-.-.- 1.-.-.- 1.-.-.- 2.2.4.1 VALERIAN 1.0% 2.2.4.4 1.-.-.- 1.-.-.- 1.-.-.- 1.-.-.- 3.2.4.4 MENTHOL CRIST (5) 34.2.2 2.1227 02:24:83 1.-.-.- 3.4.4.3 4.3.4.4 CLOVE OIL 15 drops 1.-.-.- 2122. ® == 1.-.-.- 1.-.-.- 1.-.-.- 10 drops 1.-.-.- 1.-.-.- 1.-.-.- 1.-.-.- 1.-.-.- 1.-.-.- 5 drops 1.-.-.- 1.-.-.- 1.-.-.- 1.-.-.- 1.-.-.- -.-.- DIETHYLETHER (56) 4.3.2.1 CHLOROFORM (ae) 1.-.-.- DROWNING 1 1.-.-.- DROWNING 2 2188 Results are the best obtained including repetitions. When narcotization is bad (value 1) no results of time nor fixations | and | are registered. First column, under each species, shows narcotization values (4 = very good, 3 = good, 2 = fair and 1 = bad), second column shows action time (4 < 24 hours, 3 = 24-48 h, 2 = 48-72 h and 1 > 72h), third and fourth columns show fixation | and fixation Il respectively (code values as for narcotization). (*) Enough to cover vial surface. (**) Adding drop by drop. Drowning 1 results are with water from the same locality of the species. Drowning 2 results are with deionized water. a solution of 100 ml, good narcotization was obtained in 72 h as well as very good fixation | and fixation Il, thus equalling the results ob- tained with 0.05% pentobarbital. Good narcotization and fixation can be ob- tained in 48 h with 0.125% sodium pentobar- bital. Also good narcotization was with 0.4% pentobarbital in 48 h (poor fixation ll) and with menthol in 24 h (fair fixation Il). Results with chloroform were not satisfac- tory, but by adding diethylether drop by drop to the water, a proper narcotization can be obtained in 48 h. However, fixation was al- ways very unsatisfactory. 34 ARAUJO, REMON, MORENO & RAMOS TABLE 4. Results with pulmonate gastropods Ancylus Lymnaea fluviatilis peregra PENTOBARBITAL 0.400% 4.4.4.3 4.1.4.3 0.200% 4.4.4.4 4.1.4.3 0.100% 4.4.4.4 4.1.4.3 0.050% 3.4.4.2 4.1.4.3 0.025% 1.-.-.- 1.-.-.- SODIUM PENTOBARBITAL 2.000% 2.4.1.1 2.4.3.1 1.000% 4.4.4.4 2.4.4.1 0.500% 4.4.4.4 РЗ 0.250% 3.4.3.3 4.1.4.3 0.125% 4.4.4.3 4.1.4.3 М$ 222 0.20% 3.4.2.2 2.4.3.1 0.10% 3.4.3.3 1-.-.- 0.05% 1.-.-.- 1.-.-.- PHENOXYETHANOL 1.00% 2.4.3.1 1.-.-.- 0.50% 2.4.4.1 3.4.3.1 URETHANE 2.0% 4.4.4.4 3.1.4.4 1.0% 3.4.2.2 4.1.4.2 0.5% 3.4.2.2 1.-.-.- LIME TREE 1.0% 2.4.3.2 3.1.4.2 VALERIAN 1.0% 3.4.2.2 2.1.32 METHOL CRIST (0) 4.4.4.4 3.4.4.1 CLOVE OIL 15 drops 2.4.4.2 2.4.3.1 10 drops 3.4.4.2 3.4.3.1 5 drops 3.4.3.3 3.4.4.1 Conditions and codes as in Table 3. Valvata piscinalis (Table 3) Pentobarbital (0.05%) and 0.25% and 0.125% sodium pentobarbital at 96 h were successful. However, results of the same tests at 77 h were only fair. Pentobarbital (0.2%) worked initially very well, but in repe- tition it was successful in one out of three cases. Good results were also obtained with 0.5% sodium pentobarbital and 0.2% MS 222, although not all tested specimens were as well extended as with the previous meth- ods. Pentobarbital (0.1%) gave irregular re- sults among the specimens tested including the repetition; some specimens remained withdrawn while in others extension was very good. Time elapsed for narcotization in all these cases was between 77 h and 96 h. A third test carried out with 0.125% so- dium pentobarbital in 50 ml of solution with 48 specimens, gave very good results in 77 h (27 specimens fully extended, 13 uncom- pletly extended and 8 withdrawn). Pretests were made with specimens from the same locality but collected in August 1990. Menthol and drowning tests were made in 0.5 ml of water. Results with 0.05%, 0.1% and 0.2% MS 222 were fair in 49 h, except in the last case which was good in 25 п. Results with menthol at 12°C (25 h), 2% urethane (25 h) and drowning (72 h) were also fair, whereas tests with 1% and 2% sodium pentobarbital, 4% urethane, and menthol (24-28°C) were not satisfactory. Bithynia tentaculata (Table 3) The best results were obtained in 54 h with 0.125% sodium pentobarbital. In the second and third repetitions, time elapsed was over 100 h. The third repetition was carried out with 50 specimens in 50 ml of solution, but the result was not as good as in the first trials because the experiment could not be com- pleted. Fixation | and fixation ll were very good in all cases. At 72 h and 80 h, 0.1% pentobarbital and 0.25% sodium pentobarbital, respectively, yielded very good narcotization and fixation |, although animals withdrew slightly after 24 h (fixation Il). The same results were obtained in both repetitions but one specimen re- mained closed in 0.1% pentobarbital. Results were optimal at 80 h with 1% and 2% urethane, although fixation Il was not good with the former dilution. A subsequent repetition of tests did not yield good results. Very good narcotization and fair fixation Il was obtained in 78 h and in 100 h (in the repetition) with 0.1% MS 222. With 0.05% MS 222 (no repetition was made), narcotiza- tion was very good at 80 h, but fixations | and || were very poor. With 0.4% and 0.2% pentobarbital for 56 h results of narcotization were good for one specimen and unsatisfactory for the other, but no test was repeated. Likewise without repetition, 0.05% pentobarbital yielded good narcotization in only one specimen in 32 h. Fixations | and ll were very good. Pretests were also made with specimens from the same location but collected in June 1990. Results using 0.05% and 0.1% MS 222 were very unsatisfactory as in the case of 4% urethane. Good results were obtained using RELAXING TECHNIQUES FOR FRESHWATER MOLLUSCS 35 TABLE 5. Results with bivalves. Unio sp PENTOBARBITAL 0.400% 2.4.4.4 0.200% 3.4.4.3 0.100% 4.3.4.3 0.050% 4.3.4.3 0.025% 3.2.4.3 SODIUM PENTOBARBITAL 2.000% 2.4.4.4 1.000% 3.4.4.4 0.500% 2.4.4.4 0.250% 3.3.4.4 0.125% 2.3.4.4 М$ 222 0.20% 4.4.4.4 0.10% 4.4.4.3 0.05% 4.2.4.3 PHENOXYETHANOL 1.00% 1.-.-.- 0.50% 1.-.-.- URETHANE 2.0% 4.2.4.4 1.0% 4.2.4.4 0.5% 3.2.4.3 LIME TREE 1.0% 1.-.-.- VALERIAN 1.0% 1.-.-.- MENTHOL CRIST (6) 1.-.-.- CLOVE OIL 60 15 30 drops 3.4.4.4 40 10 20 drops 3.4.4.3 20 5 10 drops 3.4.4.4 Conditions and codes as in Table 3. 2% urethane. Menthol and drowning, tested in 4 ml of water, gave poor results. Potamopyrgus jenkinsi (Table 3) This is probably the easiest species to nar- cotize. Optimal results were obtained in ap- proximately 30 h using 0.4%, 0.2%, 0.1%, 0.05% pentobarbital and 0.25% sodium pen- tobarbital. After repetition of these tests some specimens remained withdrawn. So- dium pentobarbital (0.125%) gave very good narcotization in 72 h, with similar success in the repetition. A third test with 100 speci- mens using 50 ml of 0.1% pentobarbital yielded very good results in 29 h for all spec- imens. Pisidium Corbicula amnicum fluminea 2 Sal 31:33 2.1.4.2 4.1.4.2 3.1.4.3 4.1.4.2 3.1.4.3 4.1.4.3 3.1.4.3 1.-.-.- 1.-.-.- 1.-.-.- 3.2.4.4 1.-.-.- 4.3.4.2 3.3.4.3 3.1.4.3 4.2.4.3 2114.2 4.4.4.1 4.3.4.3 2.2.2.2 4.3.4.4 4.1.4.4 4.3.4.1 4.1.4.4 4.3.4.4 1.-.-.- 4.3.4.4 2:33 3.2.4.2 1.-.-.- 3.1.4.1 3.1.4.3 2.1.4.2 1.-.-.- 1.-.-.- 1.-.-.- 1.-.-.- 1.-.-.- 4.2.4.4 3.2.4.4 3.4.3.3 1.-.-.- 3.4.4.4 1.-.-.- 3.3.4.4 1.-.-.- 0.5% sodium pentobarbital yielded very good narcotization and good fixations | and Il after 24 h. Urethane (2%) worked well after 24 h, with good fixation | and fair fixation Il; a repetition gave only fair results after 48 hours. Pretests were made with specimens from the same locality captured in August 1990. Results were very poor using 0.5%, 0.1% and 0.2% MS 222, menthol (also at 25°C), 2% sodium pentobarbital, 4% urethane and drowning. Good and fair results of narcotiza- tion were obtained after 5 h with 1% sodium pentobarbital and 2% urethane respectively. Results of fixation in ethanol after submer- sion in liquid nitrogen for 10 sec were not satisfactory. 36 ARAUJO, REMON, MORENO & RAMOS Pseudamnicola cf. luisi (Table 3) Optimal results for narcotization and fixa- tions | and Il were obtained with 0.2% pen- tobarbital (21 h), 0.1% pentobarbital (30 h), 0.25% sodium pentobarbital (70 h) and 0.125% sodium pentobarbital (94 h). Similar results were obtained in repetitions of all these tests. A third test with 50 ml of 0.1% pentobarbital using dechlorinated tap water gave excellent results in 30 h for the remain- ing 88 specimens in the sample. Good narcotization and fixations | and Il were obtained after 24 h using menthol. Sim- ilar results were obtained in the repetition. Although no repetition was carried out, 0.4% pentobarbital after 22 h was a good narcotic, yielding very good fixation | and good fixation Il. Results of narcotization and fixation were good after 52 h with 0.5% phenoxyethanol. Horatia sturmi (Table 3) Several methods were very good with this species. After 28 h, results were very good for narcotization and fixations | and Il with 0.4%, 0.2%, 0.1% and 0.05% pentobarbital and menthol. Results were similar after 51 h with 0.125% sodium pentobarbital. They were generally good with 0.5% and 0.25% sodium pento- barbital, 0.1% MS 222 and 1% urethane also in 51 h, with 0.5% phenoxyethanol (although the specimens seemed to have the skin re- moved) and 2% urethane in 44 h, and in 67 h with valerian. Slight depigmentation was observed in black specimens after using menthol and urethane. Because no repetitions were car- ried out for this species, it is not possible to asses if such depigmentation was due to a direct effect of these products or to an ex- cessive exposure to them, resulting in a slight maceration. Good extension resulted for 54 specimens in the sample kept for 7-9 days inside the refrigerator in water from their original locality (33 fully extended, 10 good, 5 fair and 6 with- drawn). Ancylus fluviatilis (Table 4) Best results for narcotization and fixations | and Il were obtained with 0.1% pentobar- bital, 1% and 0.5% sodium pentobarbital, 2% urethane and menthol. The same results were obtained with 0.2% pentobarbital with- out repetition. Time of exposure was always between 4 h and 7 h. Pentobarbital (0.4% and 0.05%), sodium pentobarbital (0.25% and 0.125%), MS 222 (0.2% and 0.1%), urethane (1% and 0.5%), valerian (1%) and 10 and 5 drops of clove oil were good or very good narcotics with an action time between 4 h and 7 h, but fixation was not as good as with the former methods. No repetitions were made. Lymnaea peregra (Table 4) There were difficulties experienced with fixation Il. For example, after 5 h, 0.4% and 0.2% pentobarbital and 2% and 1% ure- thane gave very good narcosis, although specimens retracted with the fixative. How- ever, after 76-96 h, 0.4%, 0.2%, 0.1% and 0.05% pentobarbital and 0.125% sodium pentobarbital were very successful narcotics, with very good fixation | and good fixation Il, also in the repetitions. Sodium pentobarbital (0.25%) gave very good results at the same time, being only good in the second test. Two further tests were carried out using 0.1% pentobarbital and 0.125% sodium pentobar- bital, the results being exceptionally good at 93 h for the first dilution (30 specimens in 100 ml of water). In the second case, narcotiza- tion was good after 136 h. Narcotization with 1% urethane was very slow (162 h) though very good, but fixation Il was only fair. Urethane 2% also gave gener- ally good results in 94 h. Using 1% lime tree (100 h) and menthol (5 h) good results were obtained for the first tests (bad fixation II with menthol), but not for the repetition. First tests with 10 and 5 drops of clove oil and 0.5% phenoxyethanol were good for narcotization and fixation | in 5 h, but very poor for fixation Il. Repetitions of these tests yielded poor results in 72 h. Unio sp. (Table 5) Very good narcotization and fixations | and Il were obtained after 24 h using 0.2% MS 222, although after repeating the test, fixation Il was only fair. Very good results but with an action time of between 24 h and 80 h were also obtained with 0.05% and 0.1% pento- barbital, 0.1% and 0.05% MS 222 and 1% and 2% urethane. The same results were ob- tained in repetitions, except in the case of 0.1% pentobarbital in which narcotization RELAXING TECHNIQUES FOR FRESHWATER MOLLUSCS 37 was not as successful as in the initial test. Fixation | was very good in all cases, however the animal tended to close its valves 24 h later (except with 2% and 1% urethane in which fixation ll was very good). In 72 h the test with 0.5% urethane gave good relaxation and fixations | and II, but in repetition narcotization was fair and the ani- mal was withdrawn with valves open. Good results were obtained with 0.2% and 0.025% pentobarbital in 24 and 72 h, respec- tively, 1% and 0.25% sodium pentobarbital in 24 and 30 h, respectively, and in the three tests with clove oil (in 24 h). With clove oil, a slight disturbance of the epithelium was oc- casionally observed. Pretests were carried out with specimens from the Miño River collected in July 1990. Results were very good with 0.05%, 0.1% and 0.2% MS 222 after 143, 29 and 50 h, respectively. Only in the third case, valves remained open after immersion in cold etha- nol. Narcotization was also very good with 30 drops of clove oil and 4% and 2% urethane in about 6 h. Subsequently, specimens were submerged in liquid nitrogen for 10 sec, and only for one sec in the case of 2% urethane. Results of fixation were good in the first and third cases, and unsatisfactory in the second. In similar pretests with specimens from the Gasset reservoir collected in November 1989, animals were relaxed with valves open but with the foot withdrawn using 2% phe- noxyethanol (29 h) and 1% and 2% sodium pentobarbital, both for 6 h. In the last two cases, valves remained open after fixation, having previously been submerged in liquid nitrogen for 10 sec. Finally, two experiments were made with 150 ml of deionized water using 1% urethane at 20°С (and posterior fix- ation with ethanol) and at 5°C (fixation with cold ethanol). Results of narcotization and fixation were good for the former and fair for the second, both in 10 days. Pisidium amnicum (Table 5) Best results of narcotization and fixations | and ll were obtained with 0.5% and 1% phe- noxyethanol in 30 and 48 h, respectively. Repetition of both methods was very suc- cessful in 49 h. A third test carried out with 24 specimens in 50 ml of water with 0.5% phe- noxyethanol, yielded exceptionally good re- sults in 29 h. Narcotization was excellent with 0.5% sodium pentobarbital after 49 h, but in the repetition fixation ll was only fair. With 0.1% and 0.05% MS 222, narcotization was very good, time ranging between 48 and 56 h, while fixation ll was unsatisfactory in both tests. Initial results of narcotization and fixa- tions | and II with 0.1% MS 222 were very good in only one specimen in 30 h. Sodium pentobarbital (0.25%) gave good results in only one specimen, at 73 h, and also in the repetition. Sodium pentobarbital (1%), urethane (2%), 10 drops of clove oil and menthol yielded good or very good narcosis in 24-72 h, but in the repeated tests only one of the two spec- imens was successfully relaxed. Good re- sults were obtained for only one specimen using 15 drops of clove oil in 24 h. Fixation II with 2% urethane was only fair but very good with the other products and also in the rep- etitions. The only test made with 1% ure- thane gave good narcotization in 73 h, very good fixation | and a poor fixation Il. Using 0.05% pentobarbital, the minimum time neccessary to obtain good results of narcosis and fixation was 73 h. MS 222 (0.2%) produced very good narco- tization in 48 h in just one specimen, and good fixations | and Il. Results were very bad for repetitions. Results with 0.1% and 0.025% pentobar- bital in 96 h and with 5 drops of clove oil in 30 h, were satisfactory. Pretests were carried out with specimens from the same location collected in August 1990. Narcotization was excellent with 0.05%, 0.1% and 0.2% MS 222 in 5, 6 and 24 h, respectively, and with menthol (in 24 h in 4 ml of water). Fixation with ethanol was unsatisfactory in the first case and satisfac- tory in the others. Immersion during 10 sec in liquid nitrogen was not successful in the first case, and in the second, specimens re- mained open but the epithelium seemed to be broken. Sodium pentobarbital (1%) pro- duced good narcotization in 24 h. Fixation in ethanol was also good. Results were poor with 2% sodium pentobarbital and 2% and 4% urethane, as in the test with menthol at temperature between 24°С and 28°С. Corbicula fluminea (Table 5) The most satisfactory narcotic was 0.1% MS 222 in 130 and 100 h, fixations | and Il being excellent. The third test carried out with 49 specimens in 250 ml of solution pro- duced excellent results in 94 h. MS 222 (0.05%) produced a very good narcosis in 38 ARAUJO, REMÓN, MORENO 8 RAMOS 100 h but fixation Il was a little worse in the repetition. Pentobarbital (0.05%) in 100 h and 0.25% sodium pentobarbital in 71 h, pro- duced very good narcotization and fixation |, and good fixation Il. The same occurred in the repetition of both methods. For 24 h, 0.125% sodium pentobarbital was a very good narcotic, with very good fix- ation | but poor fixation Il. For 72 h, fixation Il was just fair. Menthol and 0.5% sodium pentobarbital were good narcotics after 48 h, being very good fixation | and good fixation Il. When in- creasing action time of narcotics to 72 h, fix- ation Il with menthol was very good, but re- sults with 0.5% sodium pentobarbital were unsatisfactory. Although repeated tests were not made, narcotization and fixation | were very good using 0.2% and 0.1% pentobarbital in 96 h, fixation II being only fair. Likewise without repetition, tests made with 0.4% pentobar- bital (in 96 h) and 1% urethane (in 168 h) demonstrated that both were good narcotics with good and very good fixation |, respec- tively, and good fixation Il. Using 2% and 0.5% urethane gave very unsatisfactory re- sults. Pretests were made with specimens from the same location collected in August 1990. The volume of dilution was between 5 and 7 ml. Narcotization and fixation were good us- ing 0.05%, 0.1% and 0.2% MS 222 and 2% urethane in 140 h. Fixation with cold ethanol was Satisfactory. One specimen, although well relaxed with 1% sodium pentobarbital in 3 h, closed its valves after fixation in ethanol. When menthol was used at a temperature between 24°C and 28°C during 50 h, the specimen was closed with the foot extended and remained so after fixation in ethanol. A similar test at 12°C was completely unsuc- cessful. DISCUSSION Among species tested in this study, pul- monate gastropods were the most suscepti- ble to narcotization, most of the methods tried giving good results; pentobarbital, so- dium pentobarbital, menthol and urethane were the most successful agents, MS 222 being good for Ancylus fluviatilis, phenoxy- ethanol and lime tree for Lymnaea peregra, and clove oil for both. The most universal agents for narcotizing freshwater proso- branchs were pentobarbital and sodium pen- tobarbital. The second most effective prod- uct was urethane (especially for Bithynia tentaculata, Potamopyrgus jenkinsi, and Ho- ratia sturmi), followed by menthol (results were fair with Valvata piscinalis and with Bithynia tentaculata and failed with Pota- mopyrgus jenkinsi) and MS 222 (good results with Bithynia tentaculata and Horatia sturmi). Phenoxyethanol was only good for Pseu- damnicola cf. luisi and Horatia sturmi. Re- garding bivalves, MS 222, urethane, pento- barbital and sodium pentobarbital were, in this order, the best products, with phenoxy- ethanol being the best method for Pisidium amnicum and clove oil for Unio sp. and Pi- sidium amnicum, although ineffective for Corbicula fluminea. This brief summary shows that there is a considerable variation between species in their susceptibility to narcotic agents, which is in agreement with Runham et al. (1965). This variation, which is also found between species belonging to close genera of the same group, may be observed in the degree of extension obtained and/or in the narcotic action time, and suggests that there are many factors involved in the response of freshwater molluscs to narcotization, among which can be: the physiological status of the animal (i.e. season of the year in which animal is captured, time living in aquaria), origin of the water used for dilutions, volume ratio of narcotic dilution versus animal, temperature and aging of narcotic dilutions. McCraw (1958) refers to the seasonal dif- ference of Lymnaea stagnalis in response to narcotic agents. We found some differences between pretests and tests in Bithynia ten- taculata with 0.1% and 0.05% MS 222, in Pisidium amnicum with 2% urethane, and in Potamopyrgus jenkinsi with 1% sodium pen- tobarbital. However, these differences do not seem to be relevant since changes registered were not always coincident and because re- sponses to the other methods tested were similar between two set of tests. This also suggests that origin of the water used for di- lutions is not likely to influence results, at least for species living in changing environ- ments. Differences between pretests and tests in Bithynia tentaculata may have been caused by any of these factors, although the most feasible hypothesis is that in pretests the animals died when exposed to the nar- cotics, because of their weakened condition after three months living in aquaria. RELAXING TECHNIQUES FOR FRESHWATER MOLLUSCS 39 Corbicula fluminea differed from other spe- cies in that results of all methods varied be- tween specimens collected in August (pre- tests) and in January (tests). In this species, results were contradictory as sometimes they improved (menthol) and sometimes they worsened (1% sodium pentobarbital and 2% urethane). Considering the adaptive potential of an invasive species such as C. fluminea, the water used for dilutions would not seem to be the cause of such differences. In this case, the observed narcotization results could be caused by physiological seasonal differences. Similar results were obtained in repetitions of the tests in most of the species, although in some cases poorer results were registered. This was the case of Pisidium amnicum, Po- tamopyrgus jenkinsi and Bithynia tentaculata. Even though contradictory results might be due to the time specimens lived in aquaria, to aging of the solutions or to both, the fact that changes were not observed in the remaining species studied would point to different physiological responses of the animals to the same narcotics. Several authors (van der Schalie, 1953; McCraw, 1958; Meier-Brook, 1976a) ob- served that time needed for relaxation was generally shorter for smaller animals. We found this trend for basommatophorans and prosobranchs tested, especially remarkable being the long time needed for the Bithynia and the Valvata species. Among the bivalves tested, no general rule could be observed re- garding size; for instance Unio sp. and Pisi- dium amnicum could be relaxed, with the same narcotic and concentration, in the same time. Giusti & Pezzoli (1980) recommend “drowning” for hydrobiids, although in our early experiments results were acceptable only with Lymnaea peregra and Melanopsis sp. Therefore, it does not seem to be useful, not only in terms of time, but also the diffi- culty in checking response to mechanical stimulus without disturbing the drowning process by introducing air into the jar. The most effective narcotics for freshwater molluscs were pentobarbital and sodium pentobarbital. Both relaxed all species tested in an excellent extended position, although the optimai concentration needed varied. The minimum effective concentration of pento- barbital was 0.025% for Unio sp. and Pisid- ¡um amnicum, although 0.05% was neces- sary for narcotize the rest of species. The time used by this drug for the last species is the same even if the concentration is higher. The independence between drug-concentra- tion and action time found in Pisidium amni- cum was also observed in pulmonates and many prosobranchs. For all species tested, a concentration between 0.1% and 0.2% pen- tobarbital is recommended, the results using the highest concentration (0.4%) being gen- erally worse. Overdosing with this drug seems unlikely (Meier-Brook, 1976a). Con- versely, Meier-Brook (1976a) points out the risk of overdosing using sodium pentobar- bital and recommends doses between 0.05% and 0.1%. We have obtained good results using higher concentrations, but be- yond the limit of 0.5%, results were irregular. Both pentobarbital and sodium pentobar- bital are expensive and classified as ‘‘con- trolled substances,” therefore not easily available, especially sodium pentobarbital. We agree with Meier-Brook (1976a), that pentobarbital is the most advisable relaxing agent for freshwater molluscs. With this product there is danger of deposits of white dust over specimens. However, this effect, due to low solubility of the product, disap- pears when specimens are cleaned or in the moment of fixation. According to Meier- Brook (1976a), pentobarbital has the advan- tage of raising the concentration gradually, avoiding shock produced in the animals by overdosing. The effectiveness of pentobarbital and the selected concentrations were further tested by us on samples of four other different spe- cies in the following aqueous solutions: 0.1% in 8 ml was employed to narcotize two spec- imens of Lymnaea truncatula (Muller, 1774) and 30 specimens of Neohoratia cf. corona- doi (Bourguignat, 1870), 0.2% in 10 ml was used for 36 specimens of Neohoratia schuelei Boeters, 1981, and 0.2% in 15 ml for five specimens of Physa acuta Draparnaud, 1805. With all the species the results ob- tained were good. Menthol has usually been employed as a narcotic for invertebrates in general and for molluscs in particular. It has been highly rec- ommended by Berry (1943) for Amnicolidae after trying ‘‘a dozen anesthetics.” From ex- perience of two of us (М. А. В. and D. M.) with species of different genera (Lymnaea, Physa, Ancylus, Ferrissia, Gyraulus, Planorbarius, Potamopyrgus, Pseudamnicola, Мегсипа, Horatia, Neohoratia, Belgrandia, Belgran- diella, and Theodoxus) successful results 40 ARAUJO, REMÓN, MORENO 8 RAMOS were generally obtained using menthol at room temperature, especially in winter. How- ever, results in the present experiments were not as good as expected, especially in the case of Potamopyrgus jenkinsi. This unpre- dictability has already been reported (van der Schalie, 1953; McCraw, 1958), although re- sults may be improved transferring animals to hot formalin (Van Eeden, 1958; Runham et al., 1965) or to hot water. Probably the tradi- tional use of this product is mainly due to the fact that К is available, cheap (also recycla- ble) and an easily handled product with ac- ceptable results over a wide range of species of the different groups of freshwater mol- luscs. Data in Michelson (1958) agree with our results regarding the swelling of parts of the anatomy of pulmonate snails with urethane, such swelling is easy to observe in such spe- cies as Ancylus fluviatilis and Unio sp. treated with this product. We have carried out no tests to check the further extension of narco- tized specimens in contact with distilled wa- ter as occurs in pulmonates. Problems with necrosis or autoamputation (Michelson 1958) have not been detected by us. 1% urethane seemed to be the minimum effective concen- tration of this product for use as a narcotic for prosobranchs and for Lymnaea peregra, agreeing with the observation of Runham et al. (1965) on L. stagnalis. This inexpensive and easily available product is harmful to hu- mans and must be handled with care. Lime tree and valerian, though easily ac- cessible, are not very successful narcotics. With the former, fair results may be obtained with Melanopsis sp., Horatia sturmi, and An- cylus fluviatilis, and good results with Lym- naea peregra. Valerian works fairly well with Melanopsis sp. and Lymnaea peregra and well with Horatia sturmi and Ancylus fluviati- lis. Both are not advisable for bivalves. Clove oil, is a cheap, non-toxic product that gives good results with pulmonates and bivalves (except for Corbicula fluminea). This product and phenoxyethanol (an inexpensive but toxic product) sometimes leave deposits over the specimens which are easily cleaned. MS 222 mixed with sodium pentobarbital was used by Joosee & Lever (1959) and Le- ver et al. (1964) to anaesthetize freshwater molluscs. Used for the first time as a narcotic by us, it gave excellent results in bivalves. However, as similar results can be obtained with other methods, it is not recommended because of its high price and special require- ments (it must be protected from light and stored in cold). Bad fixation obtained on Melanopsis sp. with diethylether after excellent relaxation suggest that it might be a good anaesthetic but not a narcotic. While looking for a universal method for narcotize freshwater molluscs, we designed the experiences here described to standard- ize as much as possible the procedures for narcotization. However, if only one species or a few of them are to be studied, or available time is restricted, it is possible to improve the results. Use of mixtures of some of the agents tested have been reported to yield good results sometimes (van der Schalie, 1953; McCraw, 1958; Runham et al., 1965). It is therefore advisable to carry out specific tri- als before starting long-term studies. If time is a problem, then it is also important to bear in mind that the process of relaxation can be considerably shortened by carrying out nar- cotization at room temperature. The risk in this case can be uncontrolled death or mac- eration of the animals, requiring a very close monitoring. While questions still remain, we hope we have established the choice of a useful nar- cotic method for each one of a wide range of species of freshwater molluscs, from a wide range of drug choices. ACKNOWLEDGEMENTS We are gratefully indebted to A. G.-Valde- casas and A. |. Camacho for their construc- tive criticism of the manuscript. B. Arano, B. Kelly and a anonymous reviewer improved the English. Thanks also to R. Sanchez and L. G. Alaejos for their help in the samples trips. This work received financtial support from the Project “Fauna Ibérica Il” (SEUI, DGICYT PB89 0081) LITERATURE CITED EMBERTON, K. C., 1989, Retraction/extension and measurement error in a land snail: effects on systematic characters. Malacologia, 31: 157- Ue BERRY, E. G., 1943, The Amnicolidae of Michigan: distribution, ecology and taxonomy. Miscella- neous Publications Museum of Zoology, Univer- sity of Michigan, 57: 68 pp., 9 pls. GIRDLESTONE, D., S. G. H. CRUICKSHANK & W. WINLOW, 1989, The actions of three volatile RELAXING TECHNIQUES FOR FRESHWATER MOLLUSCS 41 general anaesthetics on withdrawal responses on the pond snail Lymnaea stagnalis L. Compar- ative Biochemistry and Physiology (C-Compara- tive Pharmacology and Toxicology), 92: 39-44. GIUSTI, Е. 8 E. PEZZOLI, 1980, Gasteropodi, 2 (Gastropoda: Prosobranchia: Hydrobioidea, Pyrguloidea). In: Guide per il riconoscimento delle specie animali delle acque interne italiane. Consiglio Nazionale delle Ricerche AQ/l/47: 66 Pp. JOOSE, J. & J. LEVER, 1959, Techniques of nar- cotization and operation for experiments with Lymnaea stagnalis (Gastropoda, Pulmonata). Proceedings of the Academy of Sciences of Am- sterdam, 62: 145-149. LEVER, J., J. С. JAGER, 8 A. WESTERVELD, 1964, A new anaesthetization technique for fresh water snails, tested on Lymnaea stagnalis. Malacolo- gia, 1: 331-338. LINCOLN, В. J. & J. G. SHEALS, 1985, Invertebrate animals: collection & preservation. British Mu- seum (Natural History), London: 150 pp. MCCRAW, В. M., 1958, Relaxation of snails before fixation. Nature, 4608: 575. MEIER-BROOK, C., 1976a, An improved relaxing technique for mollusks using Pentobarbital. Ma- lacological Review, 9: 115-117. MEIER-BROOK, C., 1976b, The influence of varied relaxing and fixing conditions on anatomical Characters in a Planorbis species. Basteria, 40: 101-106. MICHELSON, E. H., 1958, A method for relaxation and immobilitation of pulmonate snails. Transac- tions of the American Microscopical Society, 77: 316-319. RUNHAM, М. W., К. ISARANKURA 4 В. J. SMITH, 1965, Methods for narcotizing and anaesthetiz- ing gastropods. Malacologia, 2: 231-238. SMALDON, С. & E. W. LEE, 1979, A synopsis of methods for the narcotisation of marine inverte- brates. Royal Scottish Museum (Edinburgh), In- formation Series, Natural History 6: 96 pp. VAN EEDEN, J. A., 1958, Two useful techniques in fresh water malacology. Proceedings of the Ma- lacological Society of London, 33: 64-66. VAN DER SCHALIE, H., 1953, Nembutal as a re- laxing agent for molluscs. American Midland Naturalist, 50: 511-512. Revised Ms. accepted 8 December 1993 MALACOLOGIA, 1995, 36(1-2): 43-66 LAND-SNAIL COMMUNITY MORPHOLOGIES OF THE HIGHEST-DIVERSITY SITES OF MADAGASCAR, NORTH AMERICA, AND NEW ZEALAND, WITH RECOMMENDED ALTERNATIVES TO HEIGHT-DIAMETER PLOTS Kenneth C. Emberton Department of Malacology, Academy of Natural Sciences, 1900 Benjamin Franklin Parkway, Philadelphia, Pennsylvania 19103-1195, U.S.A. ABSTRACT Basic data are presented for newly reported sites (= areas of 4 + 2 hectares) of highest known land-snail diversities for the tropics (Manombo, Madagascar [MDG]: 52 shelled species) and North America (Pine Mountain, Kentucky, U.S.A. [USA]: 42 shelled species) and are compared with Waipipi Reserve (= Jones Bush), New Zealand, the highest-diversity site known for the world (NZL: 56 shelled species, despite higher figures in the literature). MDG, with mostly new and endemic species in nine families vs. NZL’s five families, belies Solem’s (1984) opinion that tropical rainforests are not very diverse and adds great urgency to the need for collecting tropical land snails on the verge of extinction. Among the three sites, shell-size distributions differ conspicuously: minute species (diameter < 5 mm) are twice as dominant at NZL as at USA, with MDG intermediate; medium to large species (10-40 mm) are two to three times as prevalent at USA as at MDG, and are virtually absent at NZL; and only MDG has giant species (> 40 mm). Shell-shape distributions also differ markedly: USA and MDG are both Cainian bimodal, but with different secondary peaks at H/D 1.8 vs. 3.6, and NZL is strictly unimodal; flat-to-subglobose (H/D 0.4-0.8) is the most common shape at all three sites, but is twice as common at NZL as at MDG, with USA intermediate; only USA has very flat shells (H/D < 0.4), and only MDG has extremely tall shells (H/D > 3.2), whereas tall shells (H/D > 2.0) are entirely absent at NZL. Ecological and taxonomic differences among the three sites were used to construct simple models assuming pure natural selection and pure long-term phylogenetic constraints. Predictions of these models suggest that both natural selection and phylogenetic constraints are necessary to explain observed community morphologies, and also that addi- tional factors, including chance colonization history and short-term phylogenetic constraints on rapidly speciating clades, played important roles. Cainian height-diameter plots compound two mathematically independent variables—size (e.g. diameter) and shape (height/diameter)—that seem better treated separately. Height and diameter, however, miss much of the relevant variation in land-snail shells, which seem better defined by the coiling aperture's rates of expansion, downward translation, and cutward displacement (Raupian parameters modified for mathernatical independence); a simple method is presented for calculating these three vari- ables from five measurements taken from shell x-rays. Key words: tropical biodiversity, biogeography, Gastropoda, community morphology, natural selection, phylogenetic constraints. INTRODUCTION Land-snail communities occur nearly worldwide, with sympatric species diversities ranging from one (subantarctic islands) to a predicted 72 (Waipipi Reserve [= Jones Bush], Manakau Peninsula, North Island, New Zealand), and with more than 30 spe- cies believed to be extremely rare, especially in tropical rainforests, where ‘‘snails . . . gen- erally are neither diverse nor abundant’ (Solem, 1984). Recent collections in Mada- gascar (Emberton, unpublished), however, have brought to light a small patch of lowland rainforest (adjacent to the village of Ma- 43 nombo, south of Farafangana, Fianarantsoa Province) with at least 52 sympatric, native, shell-bearing species. In addition to this un- expectedly high diversity, species at Ma- nombo are strikingly variable in size, ranging in shell diameter from 1.2 mm to 70.4 mm. While collecting and sorting, the author was impressed by the difference in sheil-size dis- tribution of this site from others he had col- lected, most notably in eastern North Amer- ica and in New Zealand. Standard methods for comparing land-snail community mor- phologies (Cain, 1977, 1978a, 1978b, 1981a, 1981b; Cameron, 1988; Cameron & Cook, 1989; Heller, 1987) seemed inadequate for 44 EMBERTON testing this impression, so alternatives were used and investigated. The purposes of this paper are (1) to provide basic data on the Madagascar site and on the most diverse site known in eastern North America (Hubricht, unpublished), (2) to compare the shell-size and shell-shape distributions of these two sites and of New Zealand’s—and the world's—richest site (Solem, et al., 1981; Solem & Climo, 1985), (3) to evaluate the possible roles of natural selection and phylo- genetic constraints in producing differences among these sites, and (4) to show the need for and to recommend alternative methods for comparing land-snail community mor- phologies. MATERIALS AND METHODS Sites Compared Comparisons were made among three sites that had been collected with enough ex- perience, care, and intensity to assure that the entire shelled malacofauna—including the minute, usually under-represented spe- cies—was sampled close to its entirety. The Manombo, Madagascar, site had been sam- pled once for “macros” and once for “mi- cros.” For both collections, the villagers of Manombo were given instructions on meth- ods and search images for finding snails, and were offered good prices for all shells, with bonuses for live snails. The people of Ma- nombo proved to be ardent collectors, many of them individually besting the author. The “macro” collection was made 27 September 1990, 10:00 ат-12:00 noon, by an estimated 45 people, including the author's party, for a total of about 90 person-hours. The “micro” collection was made 16 September 1992 by 66 villagers collecting all day, for an esti- mated 450+ person-hours. The area covered by the collectors is un- known, but seems unlikely to have surpassed a few hectares. Sounding the horn after the “macro” collection brought everyone walk- ing to the off-road car within just a few min- utes. Microhabitats for the “micros” were densely distributed within the forest and re- quired long periods of collecting time, so the total area covered in a day—even by 66 peo- ple—would also probably not have been more than a few hectares. The total collecting area, judging from the fraction the author was able to see, was uniform lowland, hot-humid rainforest (Koechlin et al., 1974; Emberton, in press a) with some selective cutting but gen- erally intact, with flat terrain, many large smooth-barked trees, many lianas and epi- phytes, no outcrops or rocks, and broad- leafed litter of shallow-to-moderate depth. Sorting to highly conservative morphospe- cies by the author yielded a total diversity of 52 species (Emberton, 1994a, unpublished). This is surely an underestimate, because mi- croarboreal and subsoil habitats were un- dercollected and because both visits were during the dry season, but should give an adequate picture of shell-size and -shape distributions. This (52 species) seems to be the highest diversity report for a tropical land- snail collection site; the second highest is a rainforest station in New Caledonia, where repeated collections over the course of a year by expert collectors yielded 41 species (Tillier, 1989a). According to Leslie Hubricht (in litt.), who 15 the most experienced living collector of land snails of the eastern United States (Hubricht, 1985), that region's most diverse site is a small area (presumably less than four hect- ares) on Pine Mountain, Harlan County, Ken- tucky, which has yielded 44 land gastropods, of which 42 are shelled snails and 2 are slugs (Hubricht, unpublished). This diversity is un- surpassed by any known site in either the western United States (Barry Roth, pers. comm.), Mexico (Fred Thompson, pers. comm.), or Canada (Pilsbry, 1939-1945; Cameron, 1988), so it is the richest known in North America. (The previous North Ameri- can record was 41 gastropod species, taken from a larger area: Solem, 1984.) Hubricht visited the Pine Mountain site during the spring of at least four different years, each time collecting intensively with the express purpose of finding as many spe- cies as possible (Hubricht, pers. comm.). The author and John Petranka made an intensive, productive collection within 2 km of Hu- bricht’s site on 9 May 1982. The area has a dense, old-second-growth, mixed hardwood forest on a limestone base, with rich soil, a deep leaf-litter cover, and an uneven ground with many large rocks. New Zealand’s Waipipi Scenic Reserve (= Jones Bush, southwest of Auckland), a small (4.2 ha), remnant, partially degraded patch of temperate rainforest, was collected 3 Janu- ary 1977 by David Roscoe and Bruce Hazle- wood, and 10-14 and 17 February 1981 by Frank Climo, David Roscoe, and the late Alan LAND-SNAIL COMMUNITY MORPHOLOGIES 45 Solem (Solem et al., 1981). The long experi- ence and skill of these collectors, the thor- oughness of their methods (including litter sieving and flotation), and their primary ob- jective of getting all species, combined to as- sure an accurate representation of the true malacofauna. In total, the site yielded 56 na- tive shelled snails and at least one native slug (Solem et al., 1981: 462, appendix ЗА). This number indicates a lower diversity than cited elsewhere for Waipipi Reserve (Jones Bush): “about 72 native species is a probable real- Ку” (Solem et al., 1981: 453), “exceeds 70 species” (Solem, 1984: 12), “60 species [of native land snails and slugs] have been col- lected [in a 2 hectare patch]” (Solem & Climo, 1985: 1). Nevertheless, Waipipi Reserve re- mains the most diverse known site in the world, and its number of known species would probably be increased by additional collecting (Solem et al., 1981). The terrain, vegetation, and snail micro- habitats of Waipipi Reserve were described and partially illustrated by Solem et al., (1981). The present author collected with Climo and Roscoe in a similar patch of bush on the northwest South Island of New Zealand in June 1984. These patches lie within steep gulleys (hence their escape from wood-cutters, fires, and sheep-and cattle- grazers) and have a variety of trees forming a dense canopy, no outcrops or rocks, and a generally very deep and diverse leaf litter that includes curled fronds of palms and fern trees. Thus all three sites have been well sampled for all size categories of land snails by expe- rienced collectors. Exact sampling areas are unknown but all seem to be on the order of 4 + 2 hectares. Shell-Size Comparisons and Predictions For the Madagascan (MDG) and North American (USA) sites, an “average” shell of each species was measured for height and diameter using vernier calipers or an ocular micrometer on a dissecting microscope. No measurement data were readily available for the New Zealand site (NZL), but Solem & Climo (1985: table 2) had published shell-di- ameter distributions of 83 species of the Manakau Peninsula, given in 0.5 mm inter- vals. After subtracting the four introduced species, the diameter distribution of the re- maining 79 species was taken as indicative of that of NZL’s 56 species (= a subset of the 79). Shell diameter was used as an index of shell size. This index is advantageous for its simplicity, its ease of interpretation, and its ecological relevance in approximating the minimum-diameter opening through which a snail can carry its shell into shelter (see be- low). Because this index is so approximate, because shell size can vary tremendously within a single population of land snails (Goodfriend, 1986; Emberton, 1988a), and because NZL’s size distribution is repre- sented by an inflated number of species, shell size is treated in this paper in only very broad categories. Other indices that have been used for shell size—approximate vol- ume (Solem et al., 1981) and height-plus- diameter (Gould, 1984)—also have their shortcomings; it seems unlikely that using ei- ther of these alternative indices would signif- icantly change the results of this analysis. Shell-size histograms were based on the di- ameter intervals of 0.50-5.00 mm (minute), 5.01-10.00 mm (small), 10.01-20.00 mm (medium), 20.01-40.00 mm (large), and 40.01+ mm (giant). To simplify size distributions for modeling, the small and large size classes were de- leted. Thus predictions were made for three disjunct size classes: minute (< 5 mm), me- dium (10-20 mm), and giant (> 40 mm). Predictions based on pure natural selec- tion assumed that available size niches are filled with no constraints on selection other than those imposed by long-term climatic conditions. Each of the three sites was scored on a scale of one to three for its pos- session of physical niches (whether filled or not) of the three sizes minute, medium, and giant, and for its climatic aids toward filling those niches by long-term freedom from frost, drought, and severe storms. The impor- tance of each niche-filling aid to each size category of snails was then ranked from one to three. The resulting tables were then used to predict the representation of each snail size at each site. For example, to calculate the predicted micros at USA, the USA frost- free score was multiplied by the importance of frost-freedom to micros, and this product was added to the USA drought-free score times the importance of drought-freedom to micros, plus the USA storm-free score times the importance of storm-freedom to micros; the resulting sum was then multiplied by USA’s score for its number of physical niches 46 EMBERTON for micros. From the resulting table of predic- tions, a histogram of predicted shell-size dis- tributions was prepared for each site, scaled for direct comparison with the histogram of actual shell-size distributions. Predictions based on pure phylogenetic constraints assumed that (a) each family is no more size-constrained in these three sites than К is throughout its total world distribu- tion, and (b) within each site, species will vary randomly within their family’s world-wide size range, unrestricted by natural selection. Families (and groups of related families) rep- resenting 10% or more of the species at any of the three sites were arranged phylogenet- ically, following Nordsieck (1986; see Ember- ton & Tillier, 1994, regarding Tillier, 1989b). Each of these families (or groups of related families) was categorized by its total range of shell diameters (simplified as minute, me- dium, and giant), both within the three sites and worldwide. Worldwide family size .ges were determined using the collections of the Academy of Natural Sciences of Philadel- phia, guided by Zilch (1959-1960). For each site, the number of species falling into minute, medium, and giant size classes was counted for each of the dominant families (or groups of related families). This number was then divided equally among the size classes occupied by the family worldwide. The site’s predicted number of species in each size class was obtained by adding such results over all dominant families. A histogram of these predictions was prepared, scaled for direct comparisons with the site’s actual size distributions. Shell-Shape Comparisons and Predictions Shell height/diameter (H/D) was used as an index of shell shape (Cain, 1977). For the Madagascan (MDG) and North American (USA) sites, H/D was calculated for the ‘‘av- erage” shell of each species from height and diameter measurements described above. To represent the New Zealand site (NZL), H/D values for 77 native species of the Manakau Peninsula (including Waipipi Reserve = NZL) were taken from Solem & Climo (1985: fig. 6). Comparative histograms used ten intervals: H/D = 0.00-0.40 (very flat), 0.41-0.80 (flat to subglobose), 0.81-1.20 (globose), 1.21-1.60 (moderately elevated), 1.61-2.00 (elevated), 2.01-2.40 (moderately tall), 2.41-..0 (tall), 2.81-3.20 (very tall), 3.21-3.60 and 3.61-4.00 (both extremely tall). To simplify shell-shape distributions for modeling, four disjunct categories were used: very flat (H/D < 0.41), globose (H/D = 0.81-1.20), tall (H/D = 2.01-2.80), and ex- tremely tall (H/D > 3.21). Natural-selection predictions were based on the demonstrated tendencies for flat-shelled and tall-shelled species to forage on horizontal and vertical surfaces respectively, with globose-shelled species variable and more versatile (Cain & Cowie, 1978; Cameron, 1978, 1981; Cook & Jaffar, 1984; Heller, 1987); in addition, it was assumed that very flat shells are effective for escaping drought in narrow crevices. Each of the three sites was scored on a scale of one to three for its possession of physical niches (inclination angles of smooth surfaces; nar- row, unflooded crevices) for the four shape categories. The importance of each of three niche-filling aids (long-term freedom from frost, drought, and severe storms: see above) to each of these shell-shape categories was ranked from one to three. Predictions of the representation of each snail shape at each site followed the methods described above for shell size. Phylogenetic-constraints predictions for shell shape used the same methods de- scribed above for shell size, except that if any species in a dominant family at a site fell within one of the four shape categories, then all the site’s members of that family were т- cluded in the computations. Orthogonal Raupian Parameters Raup’s (1961, 1966) W, D, and T (= the coiling aperture’s rates of expansion, out- ward displacement, and downward transla- tion) are mathematically correlated when cal- culated from the periphery of the aperture (Emberton, 1986, 1994b). Several different modifications of Raupian methods, however, have made W, D, and T orthogonal by taking the geometric center of the aperture as the standard point of reference (Raup & Graus, 1972; Harasewych, 1982; Illert, 1983; Oka- moto, 1984, 1988). To demonstrate the ad- vantages of orthogonal W, D, and T over height and diameter, two tall shells and two flat shells of very different ontogenies but identical height/diameters were sketched in cross-section, based on recent experience with shell x-rays (Emberton, in press, in prep.), and an easy method was devised for calculating W, D, and T. The four shells were then plotted for comparisons in two different LAND-SNAIL COMMUNITY MORPHOLOGIES 47 morphospaces: Cainian (height vs. diameter) and Raupian (W vs. D vs. T). RESULTS Site Data Appendices 1-3 list the species of MDG, USA (with permission of L. Hubricht), and NZL in systematic order, with shell measure- ments given for MDG and USA. Systematics studies of MDG are in progress (Emberton, 1994a, unpublished), so the species, most of which are new, are simply numbered within tentative genera. All of Hubricht's collections from USA and the author's collections from near that site (author's station GS-119) are at the Field Museum of Natural History, Chi- cago. Shell-Size Comparisons Figure 1 compares the shell-size distribu- tions of USA, MDG, and NZL. Minute species (shell diameter 0-5 mm) are everywhere a major component, but contribute less than half as strongly in USA (40%) as in NZL (84%), with MDG intermediate (63%). Small species (5-10 тт) are roughly equivalent throughout: NZL 15%, MDG 13%, USA 22%. However, medium (10-20 mm) and large (20-40 mm) species are three times and two times as common in USA (24%, 14%) as in MDG (8%, 8%), and are virtually absent in NZL (0%, 1%). Among the three sites, only MDG has giant species (40-70 mm), which comprise 8% of its diversity. Selection-Based Size Predictions Table 1 shows the size-class model and its predictions assuming pure natural selection. Availability of minute physical niches (whether filled or not) was scored intermedi- ate for USA (deep, broad-leaf litter; crevices in rocks, logs, and rough-barked tree trunks) and MDG (shallow, broad-leaf litter; crevices in logs, palm-tree trunks, and numerous ep- iphytes and vines), but high for NZL (ex- tremely deep and complex litter; crevices in logs and palm-tree trunks). Physical niches for medium snails were scored intermediate at all sites, with the lack of rock shelters at MDG and NZL compensated for by fallen palm boles. Availability of giant physical niches seems to depend on ease of mobility among rare sheltering large logs (present at all sites), so was scored low for NZL (with dense, loose, rough-surfaced litter obstruct- ing movement), intermediate for USA (with a fairly smooth litter surface but many rough stones and rough-barked trees and logs), and high for MDG (smooth surfaces through- out, including most trees and logs). The climatic aids to filling these niches differs among sites. Freedom from frost is highest at tropical MDG, intermediate at maritime-temperate NZL, and lowest at continental-temperate USA. Freedom from drought is highest at NZL (nearly regular rain- fall augmented by temperate-coast fogs and streamside topographic shelter), intermedi- ate at MDG (short dry season under tropical sun), and lowest at USA (occasionally rain- less summers, intensified by rapid drainage through the limestone base). Freedom from the effects of major storms was scored high- est at NZL (in a well-sheltered gulley), inter- mediate at USA (exposed mountainside, thunderstorms and occasional blizzards and hailstorms), and lowest at MDG (exposed to fairly frequent cyclones and with yearly tor- rential rains). The importance of these climatic factors toward niche-filling ability by land snails de- pends on their size category. Frost is a greater obstacle to giants than to minutes (which can more easily escape into deep, narrow crevices and, because of their faster thaw times, can more easily evolve physio- logical adaptations to body freezing), with mediums intermediate. Drought, on the other hand, more rapidly and drastically affects minutes than giants (which with their lower surface-to-volume ratio can withstand desic- cation longer, and with their greater mobility can better seek saving shelter), with medi- ums intermediate. Severe storms not only knock snails from their physical niches, but also transport them, causing gene flow that can thwart natural selection adapting them to local niches. These effects are strongest on minute snails, weakest on giants, and inter- mediate on mediums. Table 1 summarizes these scores. Result- ing predictions of shell size classes at each site are also given in Table 1 (bottom). Phylogeny-Based Size Predictions Nine families (or groups of related families) contributed at least 10% of the species to at least one of the three sites; together they 48 EMBERTON 10 MDG 20 40 70 Shell Diameter (mm) FIG. 1. Shell-size distributions of the native land snail species in the most diverse known sites of North America (USA), Madagascar (MDG), and New Zealand (NZL). The vertical scale is the proportion of total species, rounded to 0.05. comprised 90%, 96%, and 84% of the total species diversities of MDG, NZL, and USA, respectively. Figure 2 lists these dominant families (or groups of related families) phylo- genetically, arranged top to bottom from most ancient (= plesiomorphic = “primitive”) to most recent (= apomorphic = “derived”. Figure 2 also arranges the three sites ac- cording to the overall phylogenetic age of their faunas. Of the three, MDG has the stron- gest representation of ancient taxa, including most of the prosobranchs (Cyclophoridae), all of the presumably more ancient families of the Achatinida (sensu Nordsieck, 1986), and most of the presumably more ancient fami- lies of the most recent Helicida. NZL 1$ next, with its complement of prosobranchs, its dominance by Achatinida, and its absence of Helicida; and USA follows, with all of the most recent Helicida (but also with the “‘prim- itive” Vertiginidae). As in other geographically isolated land-snail faunas (Cain, 1977, 1978a, 1980; Peake, 1978; Cameron & Cook, 1992), phylogenetic overlap among these LAND-SNAIL COMMUNITY MORPHOLOGIES 49 TABLE 1. A model to predict shell size distributions assuming pure natural selection Physical Niches Niche-Filling Aid Site Minute Medium Giant Frost-free Drought-free Storm-free USA 2 2 2 1 1 2 MDG 2 2 3 3 2 1 NZL 3 2 1 2 3 €) Niche-Filling Importance to Snails Aid Minutes Mediums Giants Frost-free 1 2 3 Drought-free 3 2 1 Storm-free 3 2 1 Predicted Snails Site Minutes Mediums Giants USA 20 16 12 MDG 24 24 36 NZL 60 32 12 USA = Pine Mountain, Kentucky, U.S.A.; MDG = Manombo, Fianarantsao Province, Madagascar; NZL = Waipipi Reserve (= Jones Bush), North Island, New Zealand. Physical-niche scores: 1 = rare, 2 = intermediate, 3 = common. Scores for historical presence of niche-filling aids: 1 = rare, 2 = intermediate, 3 = common. Scores for importance of niche-filling aids to size classes of snails: 1 = low, 2 = intermediate, 3 = high. See text for method of calculating predicted size-class representations. Shell size classes: minutes = 0-5 mm diameter, mediums = 10-20 mm, giants = 40-70 mm. Shell Size Shell Shape _ % of Species _ These World- These MDG NZL US Bites Wide Bites World-Wide _ Cyclophor-Liarei 29 4 0 min-med min-med glob-tall vflat-xtall Vertiginidae 0 o 14 min min tall glob-tall Subulinidae 12 0 0 min min-med tall-xtall tall-xtall+*+* | Btreptaxidae 10 o 0 min min-med tall-xtall vflat-xtall = Acavidae 10 0 0 gnt med-gnt glob vflat-tall = Punct-Charop-Disc 4 92 10 min-med min-med vflat-glob vflat-tall Helixarion-Eucon 25 0 5 min-med min-med glob globes a Zonitidae 0 0 29 min-med min-med* vflat-glob vflat-glob = Polygyridae _90 _ 0 26 med med* (flat) vflat-glob Total 90 96 84 FIG. 2. Evolutionary relationships, sizes, and shapes of the superfamilies dominating the most diverse land snail communities of Madagascar (MDG), New Zealand (NZL), and North America (USA). The evolutionary tree follows Nordsieck (1986). Abbreviated family names: Cyclophoridae and Liareidae; Punctidae, Charop- idae, and Discidae; Helixarionidae and Euconulidae. Shell sizes: min = minute = diameter 0-5 mm; med = medium = 10-20 тт; gnt = giant = 40-70 mm. Shell shapes: vflat = very flat = height/diameter (h/d) 0.0-0.4; (flat) = h/d 0.4-0.8; glob = globose = h/d 0.8-1.2; tall = h/d 2.0-2.8; xtall = extremely tall = h/d 3.2-4.0. * very rarely and only barely reaches giant size (Zonitidae: Aegopis, Poecilozonites; Polygyridae: Neohelix major). ** excluding the systematically problematic, very-flat-shelled Cupulella (?Subulinidae) and Roybellia (?Helixarionidae). three sites is minimal, with no species or gen- era and extremely few families in common (Appendices 1-3). In addition, Figure 2 gives the shell-size categories covered by these nine dominant families (or groups of related families), both within the three sites and worldwide. Size ranges are the same except in Subulinidae, 50 EMBERTON minute medium USA actual UBA predicted by natural selection USA predicted by phylogenetic constraints giant Shell Size FIG. 3. Shell-size distributions at Hubricht's (unpublished) site on Pine Mountain, Kentucky, U.S.A., as recorded and as predicted by models assuming pure natural selection (Table 1) and pure phylogenetic constraints (see text). Streptaxidae, and Acavidae sensu lato (Em- berton, 1990), which do not reach medium size at the three sites but do elsewhere. Numbers of species from the dominant families (and groups of related families) falling into minute, medium, and giant size catego- ries (from data in Appendices 1 and 2, and Solem & Climo, 1985: table 2) were: MDG 38, NZL (actually the Manukau Peninsula; see Methods) 64, and USA 20. Redistributing these species among size categories under the assumption of phylogenetic constraints free from natural selection yielded for MDG 17 minute, 19 medium, and 2 giant; for NZL 32 minute, 32 medium, and O giant; and for USA 10.5 minute, 9.5 medium, and 0 giant. Actual vs. Predicted Size Distributions Figures 3-5 compare distributions of minute, medium, and giant species as actu- ally sampled (top), as predicted from pure natural selection (middle), and as predicted from pure phylogenetic constraints (bottom). For USA (Fig. 3), both selection and phylog- eny were adequate predictors of minute and medium categories, but only phylogeny cor- rectly predicted the absence of giant species. For MDG (Fig. 4), selection predicted more giants than mediums, and phylogeny pre- dicted more mediums than giants, neither of them reflecting reality in themselves, but in combination predicting the natural equality between mediums and giants (Fig. 4: top). Neither predictor by itself or in combination, however, could account for the high natural representation of minute species. For NZL (Fig. 5), phylogeny, but not selec- tion, accurately predicted the absence of gi- ants; and selection, but not phylogeny, ac- curately predicted the predominance of minutes. Neither predictor, however, ac- counted for the absence of mediums. LAND-SNAIL COMMUNITY MORPHOLOGIES 51 MDG actual MDG predicted by natural selection MDG predicted by phylogenetic .4 constraints .3 .2 sol .0 minute medium giant Shell Size FIG. 4. Shell-size distributions at the site near Manombo, Fianarantsao Province, Madagascar, as recorded and as predicted by models assuming pure natural selection (Table 1) and pure phylogenetic constraints (see text). Shell-Shape Comparisons Figure 6 compares shell-shape distribu- tions among the three sites. USA and MDG are both Cainian bimodal with a primary peak at about H/D 0.6, but with the secondary peak at about H/D 1.8 for USA and about H/D 3.6 for MDG. NZL, in contrast, is strictly un- imodal. Flat-to-subglobose (H/D 0.4-0.8) 15 the most common shape at all three sites, but is 2.5 times as common at NZL (83%) as at MDG (33%), with USA intermediate (64%); only USA has very flat shells (H/D < 0.4), and only MDG has extremely tall shells (H/D > 3.2), whereas tall shells (H/D > 2.0) are en- tirely absent at NZL. Selection-Based Shape Predictions Table 2 shows the pure-selection model and its predicted shape-class distribution for each site. Very flat physical niches (whether filled or not) were scored common at USA 52 EMBERTON minute medium NZL actual NZL predicted by natural selecticn NZL predicted by phylogenetic constraints giant Shell Size FIG. 5. Shell-size distributions at Waipipi Reserve (= Jones Bush), North Island, New Zealand, as repre- sented by Manakau-Peninsula snails (Solem & Climo, 1985) and as predicted by models assuming pure natural selection (Table 1) and pure phylogenetic constraints (see text). (within rock crevices, under loose bark of logs, and under rocks) but rare at both MDG and NZL, where rocks are few, log bark is less detachable, and other narrow crevices (e.g. at the bases of palm-tree branches) are usually too wet to provide shelter. Niches for globose snails were scored intermediate at all sites, due to their varieties of inclinations of crawling surfaces: rocks, logs, and trees at USA; logs, trees, and epiphytes at MDG; and trees and deep, complex litter at NZL. Verti- cal foraging niches for both tall- and ex- LAND-SNAIL COMMUNITY MORPHOLOGIES 53 Shell Height/Diameter FIG. 6. Shell-shape distributions of the native land snail species in the most diverse known communities of North America (USA), Madagascar (MDG), and New Zealand (NZL). Dots signify non-zero proportions < 0':025. tremely tall-shelled snails were scored com- mon at MDG (many smooth-barked trees and big, broad, smooth leaves; smooth-surfaced litter allowing migration), rare at NZL (few smooth-barked trees; deep, uneven-sur- faced litter blocking migration), and interme- diate at USA (few smooth-barked trees, but smooth-surfaced litter). Long-term climatic aids to filling these niches were scored previously (Table 1). The importance of these climatic factors toward niche-filling ability by land snails varies ac- cording to shell shape. Although very flat- shelled snails can enter narrow, deep refuges to escape frost, drought, and major storms; tall- and extremely tall-shelled arboreal snails are openly exposed to (and are highly vulner- able to) these threats, with globose-shelled snails intermediate. Table 2 summarizes these scores and uses 54 EMBERTON TABLE 2. A model to predict shell shape distributions assuming pure natural selection VFlat Site USA 3 MDG 1 NZL 1 Physical Niches Globose Tall XTall 2 2 2 2 3 3 2 1 1 Importance to Snails Niche-Filling Aid Frost-free 1 2 3 3 Drought-free 1 2 3 3 Storm-free 1 2 3 3 Predicted Snails Site USA 12 16 24 24 MDG 6 24 54 54 NZL 8 24 24 24 Sites and scores are as in Table 1. See text for method of calculating predicted shape-class representations. Shell shape classes: vflat = very flat = height/diameter (h/d) 0.0-0.4, globose = h/d 0.8-1.2, tall = h/d 2.0-2.8, жа! = extremely tall = h/d 3.2-4.0. them to predict shell-shape distributions for the three sites. Phylogeny-Based Shape Predictions Figure 2 shows shell-shape ranges cov- ered by the nine dominant families (or groups of related families), both within the three sites and world-wide. Shape ranges are generally greater worldwide, with only subulinid, helix- arionid-euconulid, and zonitid worldwide ranges fully covered in the three sites. Redistributing all species of the nine dom- inant families (or groups of related families) among shape categories under the assump- tion of phylogenetic constraints free of natu- ral selection yielded for MDG 7.33 very flat, 20.33 globose, 10.33 tall, and 8.00 extremely tall; for NZL 16.75 very flat, 16.75 globose, 16.75 tall, and 0.75 extremely tall; and for USA 12.8 very flat, 17.8 globose, 4.3 tall, and 0.0 extremely tall. Actual vs. Predicted Shape Distributions For USA (Fig. 7), natural selection was a better predictor of the tall-shell category, but only phylogenetic constraints predicted the absence of extremely tall shells. Although both these factors predicted the presence of very flat and globose shells, neither predicted the natural predominance of very flat over globose shells. For MDG (Fig. 8), both factors successfully predicted the presence of tall, extremely tall, and globose shells, but phylogenetic con- straints was more accurate in predicting the natural ratio of globose to tall plus extremely tall. Furthermore, the natural-selection mod- el's success in these categories was entirely accidental, because all of the tall and very tall MDG species were not arboreals as pre- dicted, but ground dwellers (Hainesia sp. 1, “Subulina” sp. 2, “Edentulina” spp. 1, 2, Streptostele sp. 1), and all of MDG's known arboreals were not tall as predicted, but sub- globose to globose (Tropidophora spp. 1-3, Ampelita sp. 2, Helicophanta sp. 2, Kaliella sp. 1). Neither the natural-selection nor the phylogenetic-constraints model was able to predict the absence of very flat species from MDG. For NZL (Fig. 9), natural selection was bet- ter at predicting the dominance ofthe globose shells, and phylogenetic constraints was bet- ter at predicting the absence of extremely tall shells. Neither factor, however, succeeded in predicting the complete absence from NZL of both tall and very flat shells. Orthogonal W, D, and T vs. Height and Diameter Figure 10 compares two tall shells and two flat shells. The shells are hypothetical but bi- ologically plausible, with similarities to actual living species (Zilch, 1959-1960). The shells LAND-SNAIL COMMUNITY MORPHOLOGIES 55 . 10 USA actual .05 .00 .10 USA predicted by .05 natural selection 00 (es O ен о . 10 .05 . 00 уегу glo- flat bose USA predicted by phylogenetic constraints tall extremely tall Shell Shape FIG. 7. Shell-shape distributions at Hubricht’s (unpublished) site on Pine Mountain, Kentucky, U.S.A., as recorded and as predicted by models assuming pure natural selection (Table 2) and pure phylogenetic constraints (see text). Dots signify non-zero proportions < 0.0125. are shown in cross section as they would ap- pear in an x-ray (Hutchinson, 1989: fig. 3; Emberton, 1994a). The two standard mea- surements traditionally used in land-snail community morphology (Cain, 1977, ff.; this paper) are: height and diameter = maximum shell dimensions parallel to and perpendicu- lar to the shell’s central axis of rotation (Fig. 10b). Alternative measurements needed for cal- culating mathematically independent ver- sions of Raupian parameters (Raup, 1961, 1966) are also shown in Figure 10. Two ap- ertures w whorls apart are matched in area by circles with the same geometric centers (in Fig. 10 these circles are estimated by eye, but they could be generated more precisely by a computerized algorithm). Four measurements are then taken from the circles’ centers: r, and r, = radii of the smaller and larger circles (Fig. 10d), and d and t = distances between the circles’ centers perpendicular to and parallel to the shell’s axis of rotation (Figs. 10a, c). Orthogonal coiling parameters are calculated: W = (m/w) (rz° — r,°) D = d/w raw Table 3 provides all these measurements and calculated variables for Figure 10’s four shells. Figure 10e plots the four shells in Cainian two-dimensional morphospace Cain (1977, ff.). In this height-diameter plot, shells a and b occupy the same point in the upper (tall- shelled) region, and shells c and d occupy the same point in the lower (flat-shelled) region. Alternative size and shape plots (not figured) as used in this paper would show single val- ues for a and b at smaller and taller positions on the unidimensional scale line than the sin- gle values for c and d. Figure 10f plots the same four shells in the recommended alternative, Raupian three- dimensional morphospace (Raup, 1961, 1966). In this plot, each shell occupies a unique position, and the tall and flat shells b and c are closer to each other than either is to the other tall or flat shell a or d. DISCUSSION Tropical Land-Snail Diversity and Extinction The most important message of this paper is the urgent need to collect disappearing tropical land-snail faunas as quickly as pos- sible. Manombo Reserve, Madagascar (MDG), contains the tropics’ most diverse known site, which is close to and may even- tually surpass Waipipi Reserve, New Zealand (NZL), as the world’s most diverse known 56 EMBERTON -30 -25 .20 .15 .10 .05 .00 .15 -10 MDG actual MDG predicted by natural selection .05 . 00 .20 .15 -10 . 05 . 00 уегу glo- flat bose MDG predicted by phylogenetic constraints tall extremely tall Shell Shape FIG. 8. Shell-shape distributions at the site near Manombo, Fianarantsao Province, Madagascar, as re- corded and as predicted by models assuming pure natural selection (Table 2) and pure phylogenetic constraints (see text). Dots signify non-zero proportions < 0.0125. site. MDG, with mostly new and endemic species in nine—mostly phylogenetically an- cient—families vs. NZL's five families (Ap- pendices 1, 3), might even already rank higher than NZL in molluscan genetic diver- sity. These data, added to accumulating ev- idence of high land-snail diversities at rain- forest sites in Peru (R. Ramírez, pers. comm.), Costa Rica (С. Altaba, pers. comm.), mainland Africa (De Winter, 1992), and New Caledonia (Tillier, 1989a), refute Solem's (1984: 17) unsubstantiated claim that in trop- ical rain forests “snails . . . generally are nei- ther diverse nor abundant.’ The fact is that although tropical rain-forest snails generally are not abundant (thus re- quiring very intensive collecting), they can of- ten be extremely diverse. Vast regions of the tropics are uncollected or undercollected for land snails (Solem, 1984). Such high sympat- ric diversities, coupled with the patterns of tiny geographic ranges, high degrees of en- demism, and extreme ecological fragilities well known in land snails (Tillier & Clarke, 1983; Solem, 1984, 1990; Murray et al., 1988; Emberton, 1994a, in press a), means that no system of parks and reserves, no matter how extensive, can prevent major extinctions dur- ing the next few decades. Not only must a large fraction of tropical land-snail biodiver- sity lie outside of reserves, but reserve status is no guarantee of protection (Soulé, 1991). Manombo Reserve is a good case in point, with a village in its midst, and with defores- tation still actively taking place in May 1993 (personal observations). Besides the importance of collecting sheer numbers of disappearing tropical species, emphasis needs to be placed on oceanic is- lands as refugia for ancient clades that have already undergone some extinction on con- tinents. The two islands in this study, Mada- gascar and New Zealand, show this trend beautifully (Fig. 2). The additional tendency of LAND-SNAIL COMMUNITY MORPHOLOGIES 37 .10 NZL actual .05 .00 -10 .05 NZL predicted by natural selection ee... or a: ae .10 .05 . 00 уегу glo- flat bose NZL predicted by phylogenetic constraints tall extremely tall Shell Shape FIG. 9. Shell-shape distributions at Waipipi Reserve (= Jones Bush), North Island, New Zealand, as rep- resented by Manakau-Peninsula snails (Solem & Climo, 1985) and as predicted by models assuming pure natural selection (Table 2) and pure phylogenetic constraints (see text). Dots signify non-zero proportions < 0.0125. oceanic islands to have а phylogenetically depauperate fauna (Peake, 1978; Cameron 8 Cook, 1992) is obvious in New Zealand but only scarcely applies to Madagascar, the world's fourth largest island (Appendices 1, 3). Comparing Diversities For land-snail biodiversity, the lines are quite indistinct between alpha-, beta-, and gamma-diversities (Cameron, 1992; = sym- patric, mosaic, and allopatric diversities of Solem, 1984: 11). Certainly, to say that all the snails found within a four-hectare area of for- est are living in sympatry would be mislead- ing. Most Waipipi-Reserve species, for ex- ample, are well segregated by microhabitat (Solem & Climo, 1985); thus the 56-species figure for this site (Appendix 3) mixes alpha- and beta-diversity. There has yet to be a standardized sampling method that allows truly accurate biogeographic comparisons of land-snail diversities. Land snails have such high gamma-diversity (Solem, 1984), how- ever, that the most efficient tropical collect- ing requires frequent movement among sites, so that standardized, intensive quadrat sam- pling seems counterproductive. Land snails are so patchily distributed (beta-diversity) that the best way to collect a tropical site quantitatively is to spend set amounts of time (a) searching for micros in the right microhab- itats, (b) beating vegetation over inverted um- brellas for arboreal micros, (c) scanning the ground and digging out refuges for macros, and (d) scanning the trees for arboreal mac- ros. Others disagree, and get diverse collec- tions from ground-litter quadrats augmented by macro searches (R. Ramirez, pers. com- mun.) or from bagged litter samples (E. Naranjo Garcia, pers. comm.). Another problem in comparing land-snail diversities is that of differing species con- cepts, especially with regard to gamma- diversity. Land-snail species are notoriously variable in shell morphology (Goodfriend, 1986) and sometimes in genitalia as well (Tillier, 1989a). Polytypic species are com- mon, and their sometimes exuberant overde- scriptions by past taxonomists are only be- ginning to be cleared up using modern species concepts (e.g. Gould & Woodruff, 1986). The three sites compared in this pa- per, however, suffer from little if any such overdescription (Hubricht, 1985; Solem et al., 1981; Emberton, 1994a). Comparing and Predicting Community Morphologies Height-diameter plots could have been used to compare community morphologies of USA, MDG, and NZL, but such plots com- 58 EMBERTON 3 units e 8 *ab ht 4 *c,d о 0 4 8 dm 00 FIG. 10. Four hypothetical shells (a-d, shown as simplified x-rays) and their positions in Cainian (e) and Raupian (f) morphospaces. Cainian measurements (= parameters): dm = diameter, ht = height; Raupian measurements (taken from circles whose areas and centers coincide with apertural cross-sections): w = whorl count between circles, r, and r, = radii of circles, d and t = distances between circles’ centers perpendicular to and parallel to the axis of rotation; Raupian parameters: the coiling aperture's per-whorl rates of whorl expansion (W = [m/w] [r,2?—r,*)), outward displacement (D = d/w), and downward translation (T = Ум). Table 3 lists all measurements and variables for the four shells. pound two mathematically independent vari- ables—size (e.g. diameter) and shape (height/diameter)—that seem better treated separately. Thus important differences among the three sites were detected in both size (Fig. 1) and shape (Fig. 6). The sites also differ enormously in ecology (Tables 1, 2) and in phylogenetic composition (Fig. 2); predic- tions from simple models based on these dif- ferences (Figs. 3-5, 7-9) suggest that both natural selection and phylogenetic con- straints are necessary to explain observed distributions of both shell size and shape, but that even these two factors in combination are not always sufficient. Additional factors, therefore, must be involved. A complete set of factors controlling com- munity morphology should include proximate natural selection for both foraging-surface in- cline (Cain, 1977, ff.) and shelter site (Solem & Climo, 1985; this paper), climatic exclusion (Gould, 1970), long-term phylogenetic con- straints (this paper), chance colonization his- tory (Cameron, 1988; Cameron 8 Cook, 1992), speciation within clades with short- term phylogenetic constraints (Cameron 4 Cook, 1992), interspecific competition (Cain, 1977, ff.), direct environmental induction (Gould, 1970), and constructional соп- straints. Of these, climatic exclusion, chance colonization history, and short-term phyloge- netic constraints on rapidly speciating clades seem to have played important roles in form- ing the community-morphology differences LAND-SNAIL COMMUNITY MORPHOLOGIES 59 TABLE 3. Measurements and calculated variables for the four hypothetical shell x-rays shown in Fig. 10 a Cainian measurement: ht. 7.5 measurement: diam. 30 calculation: ht./diam. 2:5 Raupian measurement: w 2 measurement: г. 0.3 measurement: r, 1.0 measurement: d 0.5 measurement: t 4.2 calculation: W 1.4 calculation: D 0.3 calculation: T 21 among USA, МОС, and NZL. Climatic exclu- sion may explain better than available niche space (Table 1) the absence in eastern North America of giant snails, which seem to be restricted to the tropics (Zilch, 1959-1960) and which (family Camaenidae) occupied northern North America during its tropical phase in the Cretaceous and early Tertiary (Solem, 1978). Chance alone may have pre- vented the extremely tall-shelled clausiliids from colonizing eastern North America from ecologically similar western Europe, thus ex- plaining the absence of very tall shells at USA. Extensive radiations of New Zealand punctids and charopids and Madagascan cy- clophorids within phylogenetically con- strained genera may partially explain NZL's and MDG's dominance by minute, flat-to- subglobose shells. The absence from NZL of medium-sized snails may be due to chance colonization history, because the introduced, medium-sized Bradybaena similaris has found a niche there (Solem et al., 1981). Cli- matic exclusion, on the other hand, may have prevented the giant-shelled, New Zealand genus Paryphanta (Powell, 1979) from estab- lishing at NZL. It may also be climatic exclu- sion that explains the absence of very flat shells from both MDG and NZL, where all narrow crevices may be too wet to serve as niches for land snails. The absence of tall and very tall shells at NZL is probably due partly to climatic exclusion (since tall charopids are found elsewhere in New Zealand) and to chance colonization history (no subulinids, clausiliids, etc., invaded the islands). One can hope eventually to quantify the relative contributions of all these (and per- Shell b C d 7.5 2:3 2.3 3.0 7.0 7.0 2.5 0.3 0.3 vá 4 2 0.15 0.2 0.4 0.45 0.4 0.9 0.5 1.8 1.4 4.1 1.0 1.0 0.1 0.1 1.0 0.1 0.5 0.7 0.6 03 0.5 haps other) factors toward any given land- snail community morphology, in a way analagous to such efforts on individual mor- phology (Raup, 1972; Cheverud, 1982; Em- berton, in press b). In striving toward such a goal, it seems essential to alter Cain's (1977, ff.) protocol in two ways. First, individual, highly localized communities should be the standard units of comparison, rather than re- gional, whole-island, or continental faunas. There is as yet no standard definition of a land-snail community in terms of collecting area. А community could be defined (and measured) as the two- or three-dimensional area within which all land-snail individuals are capable of physical contact with all (or 90% or 80% of) other species during an average generation. Second, community-morphology compar- isons should include all land molluscs. The cited works have generally treated proso- branchs separately from pulmonates, on the argument that prosobranch radulae are fun- damentally so different that they must oc- cupy different feeding niches. There seems to be little hard evidence for this, and there 1$ considerable evidence that pulmonate radu- lae have adapted to a full range of niches, from the “prosobranch niche” (Cain, 1977) of scraping algal films from tree trunks (Achati- nella, Liguus) to snagging large living prey (Euglandina). To try to isolate components of the land-snail fauna based on feeding niche, therefore, seems unavoidably complex and artificial, especially since there are no data on most species of most faunas. Thus fixed size and shape differences between pulmonates and non-pumonates can be attributed to 60 EMBERTON long-term phylogenetic constraints but should not be segregated a priori. Gould (1970) established a valuable ap- proach toward determining the causes of dif- ferences in community morphology when he analysed a “naturally replicated experiment” in land-snail community structure that con- trolled for the factors of natural selection, phylogenetic constraints, colonization, and speciation. In that study of fossil land-snail assemblages of Bermuda, Gould (1970: fig. 10) discovered major effects from climatic exclusion and measurable effects from envi- ronmental induction. Despite such progress, the goal of “a sci- ence of form” (Gould, 1971; Raup, 1972) for land-snail community morphology still seems a long way off, because so little is known about any of the controlling factors. For ex- ample, although natural selection for taller shells to feed on more vertical surfaces seems reasonably well documented (Cain & Cowie, 1978; Cameron, 1978, 1981; Cook & Jaffar, 1984; Emberton, in press b), there is also a second adaptive shape for verti- cal-surface feeding: flat-shelled rock-cliff foragers that shelter in narrow rock crevices (Emberton, 1986, 1988b, 1991a; Heller, 1987). At MDG the tall-shelled snails are not vertical-surface feeders, but are restricted to the ground. Also at MDG, the arboreal spe- cies—like those of other rainforests in the Philippines and northern Australasia (Cain, 1978b)—are globose rather than high-spired. These few examples demonstrate how much remains to be discovered about natural se- lection on the functional morphology of shell shape. Evaluating phylogenetic constraints on land-snail shells requires robust phyloge- netic hypotheses, which unfortunately are al- most entirely lacking above the subfamilial level (Emberton et al., 1990; Emberton, 1991b; Bieler, 1993; Emberton & Tillier, 1994). Cain’s (1977) call for more ecological data on land-snail species remains strongly in ef- fect, especially for tropical faunas that are rapidly going extinct. Cameron & Cook (1989) set excellent standards for rapidly evaluating ecological differences within a community. Even when such studies are not possible (e.g. during fast-moving, labor-intensive tropical surveys), recording the positions and activity states of collected specimens would provide valuable new data. Much of the recent progress in land-snail community morphology has been due to the simplicity and elegance of height-diameter plots and the naturally bimodal patterns they detect (Cain, 1977, ff.). Height and diameter are correlated, however, and finer discrimi- nation among patterns can be made by treat- ing size and shape separately (Figs. 1, 6). A much higher refinement can be achieved at the cost of x-raying (Emberton, 1994b) and taking five measurements per shell (Fig. 10b- d). Mapping community morphologies by Raup’s W, D, and T (Fig. 10f) offers much more than simply adding a third dimension: these three variables, which can be further customized (Harasewych, 1982; Kohn & Riggs, 1975; Okomoto, 1988; Ackerly, 1989; Emberton & Chapman, unpublished), ele- gantly define much of the shell’s ontogeny (but see Gould, 1968; and Hutchinson, 1989, 1990), can yield reasonably precise calcula- tions of shell volume and surface area (Raup & Graus, 1972; in contrast to approximations by Solem & Climo, 1985), and are mathemat- ically orthogonal, so can be analyzed inde- pendently for controlling factors (Emberton, in press b, this paper) or can be used to de- fine a natural three-dimensional morpho- space with realized regions bounded by con- structional constraints, competitive exclusion, etc. (Raup, 1966). ACKNOWLEDGEMENTS | wish to thank the National Science Foun- dation (grant DEB-9201060) and the Ameri- can Philosophical Society (1990 travel grant) for supporting this work; Leslie Hubricht for allowing me to publish his Pine Mountain list; John Petranka for guidance and help at Pine Mountain, Kentucky; Owen Griffiths, Aldus Andriamamonjy, Ruffin Arijaona, and villagers of Manombo for their superb collecting; Pa- tricia Wright and Benjamin Andriamihaja of the Ranomafana National Park Project for in- valuable logistic support in Madagascar; Frank Climo, David Roscoe, and the late Alan Solem for arranging my field experience in New Zealand; and George Davis and two anonymous reviewers for helpful criticism of a previous draft. LITERATURE CITED ABBOTT, R. T. & K. J. BOSS, eds., 1989, A clas- sification of the living Mollusca. American Mala- cologists, Melbourne, Florida, 189 pp. LAND-SNAIL COMMUNITY MORPHOLOGIES 61 ACKERLY, S. C., 1989, Kinematics of accretionary shell growth, with examples from brachiopods and molluscs. Paleobiology, 15: 147-164. BIELER, R., 1993. Gastropod phylogeny and sys- tematics. Annual Review of Ecology and Sys- tematics, 23: 311-338. САМ, 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, (В) 277: 377- 428. CAIN, A. J., 1978a, The deployment of operculate land snails in relation to shape and size of shell. Malacologia, 17: 207-221. CAIN, A. J., 1978b, Variation in terrestrial gastro- pods in the Philippines in relation to shell shape and size. Journal of Conchology, 29: 239-245. CAIN, A. J., 1980, Whorl number, shape, and size in some pulmonate faunas. Journal of Conchol- ogy, 30: 209-221. САМ, A. J., 1981a, Possible ecological signifi- cance of variation in shape of Cerion shells with age. Journal of Conchology, 30: 305-315. САМ, А. J., 1981b, Variation in shell shape and size of helicid snails in relation to other pulmo- nates of the Palearctic region. Malacologia, 21: 149-176. CAIN, А. J. 8 В. Н. COWIE, 1978, Activity of dif- ferent species of land-snail on surfaces of differ- ent inclinations. Journal of Conchology, 29: 267- 212. CAMERON, R. A. D., 1978, Differences in the sites of activity of coexisting species of land mollusc. Journal of Conchology, 29: 273-278. CAMERON, R. A. D., 1981, Functional aspects of shell geometry in some British land snails. Bio- logical Journal of the Linnean Society, 16: 157- 167. CAMERON, R. A. D., 1988, Incomplete conver- gence of shell sizes and shapes in forest snail faunas from two continents: a relic of environ- mental history? Malacologia, 28: 147-157. CAMERON, R. A. D., 1992, Patterns of diversity in land snails: the effects of environmental history. Abstracts of the Eleventh International Malaco- logical Congress, Sienna: 169-170. CAMERON, R. A. D. & L. M. COOK, 1989, Shell size and shape in Madeiran land snails: do niches remain unfilled? Biological Journal of the Linnean Society, 36: 79-96. CAMERON, R. A. D. & L. M. COOK, 1992, The development of diversity in the land snail fauna of the Madeiran archipelago. Biological Journal of the Linnean Society, 46: 105-114. CHEVERUD, J., 1982, Phenotypic, genetic, and environmental morphological integration in the cranium. Evolution, 36: 499-516. COOK, L. M. 8 N. JAFFAR, 1984, Spire index and preferred surface orientation in some land snails. Biological Journal of the Linnean Society, 21: 307-313. DE WINTER, A. J., 1992, Gastropod diversity in a rain forest in Gabon, West Africa. Abstracts of the Eleventh International Malacological Con- gress, Sienna, 1992: 172-173. EMBERTON, К. C., 1986, The evolution of multiple sympatric homeomorphy among three genera of land snails. Ph.D. Thesis, University of Chicago. EMBERTON, К. C., 1988a, Shell variation in a pop- ulation of Polygyra septemvolva (Pulmonata: Po- lygyridae). Proceedings of the Academy of Nat- ural Sciences of Philadelphia, 140: 285-294. EMBERTON, К. C., 1988b, The genitalic, allozy- mic, and conchological evolution of the eastern North American Triodopsinae (Gastropoda: Pul- monata: Polygyridae). Malacologia, 28: 159-273. EMBERTON, K. C., 1990, Acavid land snails of Madagascar: subgeneric revision based on pub- lished data (Gastropoda: Pulmonata: Stylom- matophora). Proceedings of the Academy of Natural Sciences of Philadelphia, 142: 15-31. EMBERTON, К. C., 1991a, The genitalic, allozymic, and conchological evolution of the Mesodontini, trib. nov. (Gastropoda: Pulmonata: Polygyridae). Malacologia, 33: 71-178. EMBERTON, K. C., 1991b, Polygyrid relations: a phylogenetic analysis of 17 subfamilies of land snails (Mollusca: Gastropoda: Stylommato- phora). Zoological Journal of the Linnean Soci- ety, 103: 207-224. EMBERTON, K. C., 1994a, Thirty new species of Madagascan land snails. Proceedings of the Academy of Natural Sciences of Philadelphia, 145: 147-189. EMBERTON, K. C., 1994b, Polygyrid land-snail phylogeny: external sperm exchange, early North American biogeography, iterative shell ev- olution. Biological Journal of the Linnean Soci- ety, 52: 1-31. EMBERTON, K. C., in press a, On the endangered biodiversity of Madagascan land snails, in: А.С. VAN BRUGGEN & S. WELLS, eds., Molluscan Diver- sity and Conservation. EMBERTON, K. C., in press b, Partitioning a mor- phology among its controlling factors. Biological Journal of the Linnean Society. EMBERTON, K. C., G. S. KUNCIO, G. M. DAVIS, S. M. PHILLIPS, K. M. MONDEREWICZ & Y. H. GUO, 1990, Comparison of recent classifica- tions of stylommatophoran land-snail families, and evaluation of large-ribosomal-RNA se- quencing for their phylogenetics. Malacologia, 31: 341-366. EMBERTON, К. С. & $. TILLIER, 1994, Letter to the editor: clarification and evaluation of Tillier’s (1989) stylommatophoran monograph. Malaco- logia. GOODFRIEND, G. A., 1986, Variation in land-snail shell form and size and its causes: a review. Sys- tematic Zoology, 35: 204-223. GOULD, S. J., 1968, Ontogeny and the explanation of form: an allometric analysis. Journal of Pale- ontology, 42: 81-98. GOULD, S. J., 1970, Land snail communities and Pleistocene climates in Bermuda: a multivariate analysis of microgastropod diversity. Proceed- 62 EMBERTON ings of the North American Paleontological Con- vention, September 1969, Part E: 486-521. GOULD, $. J., 1971, D'Arcy Thompson and the science of form. New Literary History, 2: 229- 258. GOULD, $. J., 1984, Morphological channeling by structural constraint: convergence in styles of dwarfing and gigantism in Cerion, with a de- scription of two new fossil species and a report on the discovery of the largest Cerion. Paleobi- ology, 10: 172-194. GOULD, $. J. & D. $. WOODRUFF, 1986, Evolution and systematics of Cerion (Mollusca: Pulmo- nata) on New Providence Island: a radical revi- sion. Bulletin of the American Museum of Natural History, 182: 389-490. HARASEWYCH, M. G., 1982, Mathematical mod- eling of the shells of higher prosobranchs. Bul- letin of the American Malacological Union, 1981: 6-10. HELLER, J., 1987, Shell shape and land-snail hab- itat in a Mediterranean and desert fauna. Biolog- ical Journal of the Linnean Society, 31: 257-272. HUBRICHT, L., 1985, The distributions of the na- tive land mollusks of the eastern United States. Fieldiana, Zoology, (New Series) 24: 1-191. HUTCHINSON, J. М. C., 1989, Control of gastro- pod shell shape: the role of the preceding whorl. Journal of Theoretical Biology, 140: 431-444. HUTCHINSON, J. М. C., 1990, Control of gastro- pod shell form via apertural growth rates. Journal of Morphology, 206: 259-264. ILLERT, C., 1983, The mathematics of gnomonic seashells. Mathematical Biosciences, 63: 21-56. KOECHLIN, J., J.-L. GUILLAUMET, 8 P. MORAT, 1974, Flore et Vegetation de Madagascar. J. Cramer in der Gantner Verlag Kommanditgesell- schaft, FL-9490, Vaduz, Germany, 687 pp. KOHN, А. J. 8 A. С. RIGGS, 1975, Morphometry of the Conus shell. Systematic Zoology, 24: 346- 359. MURRAY, J., E. MURRAY, M. S. JOHNSON 4 B. CLARKE, 1988, The extinction of Partula on Moorea. Pacific Science, 42: 150-153. NORDSIECK, H., 1986, The system of the Stylom- matophora (Gastropoda), with special regard to the systematic position of the Clausiliidae, II. Im- portance of the shell and distribution. Archiv fúr Molluskenkunde, 117: 93-116. OKAMOTO, T., 1984, Theoretical morphology of Nipponites (a heteromorph ammonoid). Kaseki (Fossils), Paleontological Society of Japan, 36: 37-51 [In Japanese]. OKAMOTO, T., 1988, Analysis of heteromorph am- monoids by differential geometry. Paleontology, 31: 35-52. PEAKE, J., 1978, Distribution and ecology of the Stylommatophora. Pp. 429-526 in: V. FRETTER & J. PEAKE, eds., Pulmonates. Volume 2A. Systemat- ics, evolution and ecology. Academic Press, London. PILSBRY, H. A., 1939-1945, Land Mollusca of North America (north of Mexico). Academy of Natural Sciences, Philadelphia, Monographs, 3: 1-994, 1-1113. POWELL, A. W. B., 1979, New Zealand Mollusca. Marine, land and freshwater shells. Collins, Auckland, 500 pp; 82 pls. RAUP, D. M., 1961, The geometry of shell coiling in gastropods. Proceedings of the National Acad- emy of Sciences, U.S.A., 47: 602-609. RAUP, D. M., 1966, Geometric analysis of shell coiling: general problems. Journal of Paleontol- ogy, 40: 1178-1190. RAUP, D. M., 1972, Approaches to morphologic analysis. Pp. 28-44, in: Т. J. М. SCHOPF, ed., Mod- els т Paleobiology. Freeman, Cooper & Com- pany, San Francisco. ВАЧР, D. М. & В. В. GRAUS, 1972, General equa- tions for volume and surface area of а logarith- mical coiled shell. Mathematical Geology, 4: 307-316. RIEDEL, A., 1980, Genera Zonitidarum. W. Back- huys, Rotterdam, 197 pp. SOLEM, A., 1978, Cretaceous and early Tertiary camaenid land snails from western North Amer- ica (Mollusca: Pulmonata). Journal of Paleontol- ogy, 52: 581-589. SOLEM, A., 1984, A world model of land snail di- versity and abundance. Pp. 6-22, in: A. SOLEM 8 А. С. VAN BRUGGEN, eds., World-wide snails: bio- geographical studies on non-marine Mollusca. E. J. Brill/W. Backhuys, Leiden. SOLEM, A., 1990, How many Hawaiian land snail species are left?—and what can we do for them? Bishop Museum Occasional Papers, 30: 27-40. SOLEM, A. & Е. M. CLIMO, 1985, Structure and habitat correlations of sympatric New Zealand land snail species. Malacologia, 26: 1-30. SOLEM, A., F. M. CLIMO 4 D. J. ROSCOE, 1981, Sympatric species diversity of New Zealand land snails. New Zealand Journal of Zoology, 8: 453- 485. SOULÉ, M. E., 1991, Conservation: tactics for a constant crisis. Science, 253: 744-750. : TILLIER, S., 1989a, Patterns of diversity in New Caledonian charopid land snails. Abstracts of the Tenth International Malacological Congress, Túbingen: 248 pp. TILLIER, S., 1989b, Comparative morphology, phylogeny and classification of land snails and slugs (Gastropoda: Pulmonata: Stylommato- phora). Malacologia, 30: 1-303. TILLIER, S. 4 В. С. CLARKE, 1983, Lutte bi- ologique et destruction du patrimoine génétique: le cas des mollusques gastéropodes pulmonés dans les territoires francais du Pacifique. Géné- tique Sélection Evolution, 15: 559-566. ZILCH, A., 1959-1960, Gastropoda. Teil 2. Euthy- neura. Pp. 1-834 in: O. H. SCHINDEWOLF, ed., Handbuch der Palaeozoologie, Band 6. Ge- brúder Borntráger, Berlin. Revised Ms accepted 9 February 1994 LAND-SNAIL COMMUNITY MORPHOLOGIES APPENDIX 63 APPENDIX 1. Manombo Reserve, Madagascar: taxonomy and shell dimensions of native shelled land snails. Introduced Achatina is omitted. Higher classification follows Abbott 8 Boss (1989) for subclasses Proso- branchia and Gymnomorpha and Nordsieck (1986) for subclass Pulmonata: order Stylommatophora. Species Genus 4 Higher Classification Number Height Subclass PROSOBRANCHIA Order MESOGASTROPODA Superfamily CYCLOPHOROIDEA Cyclophoridae Boucardicus 1 9.1 2 5.3 3 2.9 4 312 5 3.8 6 2.8 7 3.0 8 2.5 9 2.3 10 3.5 1 1.6 12 1.9 13 Sl Cyathopoma 1 2.0 Hainesia 1 31.0 Superfamily LITTORINOIDEA Pomatiasidae Tropidophora 1 16.0 2 30.0 3 24.0 Superfamily RISSOIDEA Assimineidae Omphalotropis 1 3.9 2 5.6 Subclass PULMONATA: Order STYLOMMATOPHORA Superfamily BULIMINOIDEA Buliminidae (= Enidae) Rachis 1 11.4 Suborder SIGMURETHRA Infraorder ACHATINIDA Superfamily ACHATINOIDEA Subulinidae “Subulina” 1 3.5 2 ZU 3 4.9 4 2.9 5 5.1 6 13.6 Superfamily STREPTAXOIDEA Streptaxidae: Streptaxinae “Edentulina” 1 T2 2 5.2 3 4.8 Gulella 1 2.9 Streptaxidae: Enneinae Streptostele 1 25.2 Measurements (mm) Diam. 1.9 Ht./Diam. 52 1.94 2.08 DAT: 2.42 2.55 3.68 2:12 2.26 2.00 1.93 3.55 (continued) 64 APPENDIX 1. (Continued) Genus & Higher Classification Superfamily ACAVOIDEA Acavidae Ampelita Helicophanta Superfamily PUNCTOIDEA Charopidae Pilula Infraorder HELICIDA Superfamily HELIXARIONOIDEA Helixarionidae: Sesarinae Kaliella Helixarionidae: Microcystinae Microcystis Helixarionidae: Ariophantinae Kalidos Malagarion Helixarionidae: Macrochlamydinae Sitala EMBERTON Species Number ON — № — ND = © ND — № — Measurements (mm) Height 19.0 80.0 Diam. 46.0 65.0 Ht./Diam. 0.53 0.68 dant 0.69 APPENDIX 2. Pine Mountain, Kentucky, U.S.A.: taxonomy and shell dimensions of native shelled land snails. Two species of native philomycid slugs also occur: Philomycus venustus and Pallifera secreta. Classification as in Appendix 1, except that Riedel (1980) was followed for the Zonitidae. Polygyrid measurements were from Pine Mountain specimens, all others were from the literature for geo- graphically and ecologically relevant specimens. Genus 4 Higher Classification Subclass PROSOBRANCHIA Order ARCHAEOGASTROPODA Superfamily HELICINOIDEA Helicinidae Hendersonia Order MESOGASTROPODA Superfamily RISSOIDEA Hydrobiidae Pomatiopsis Subclass PULMONATA: Order ARCHAEOPULMONATA Suborder ELLOBIOIDEA Ellobiidae Carychium (E Subclass PULMONATA: Order STYLOMMATOPHORA Suborder ORTHURETHRA Superfamily COCHLICOPOIDEA Cochlicopidae Cionella Measurements (mm) Species occulta lapidaria clappi nannodes morseana Height 4.0 6.3 ate ate & © Thal Diam. 6.6 3.2 0.8 0.6 2.3 Ht./Diam. 0.61 197 2.38 2.33 3.09 LAND-SNAIL COMMUNITY MORPHOLOGIES 65 APPENDIX 2 (Continued) Measurements (mm) Genus & Higher Classification Species Height Diam. Ht./Diam. Superfamily PUPILLOIDEA Vertiginidae Vertigo gouldi 1.6 0 1.60 V. clappi 1.5 0.8 1.88 Columella simplex 2.2 1.4 1:57 Gastrocopta pentodon 1.8 141 1.64 G. contracta 2.5 1.4 1.79 G: corticaria 2:5 1.0 2.50 Suborder SIGMURETHRA Infraorder ACHATINIDA Superfamily RHYTIDOIDEA Haplotrematidae Haplotrema concavum 9.0 21.0 0.43 Superfamily PUNCTOIDEA Punctidae Punctum blandianum 0.7 1.2 0.58 Discidae Discus patulus 4.0 8.9 0.45 D. nigrimontanus 2.4 7.4 0.32 Anguispira mordax 6.0 18.0 0.33 Infraorder ELASMOGNATHA Superfamily SUCCINEOIDEA Succineidae Succinea ovalis 25:0 135 1.85 Infraorder HELICIDA Superfamily HELIXARIONOIDEA Euconulidae Euconulus fulvus 2.4 31 0.77 Guppya sterkii 0.8 12 0.67 Superfamily VITRINOIDEA Zonitidae: Gastrodontinae Gastrodonta interna 5.0 7.4 0.68 Ventridens collisella Re 87 0.83 Striatura meridionalis 1.0 17 0.59 Zonitidae: Zonitinae: Vitreini Paravitrea multidentata 2.2 4.0 0.55 P: subtilis 1.4 2.9 0.48 P. capsella 3.0 9:0 0.54 Zonitidae: Zonitinae: Zonitini Mesomphix inornatus 9.8 20.0 0.49 M. perlaevis 12.2 20.0 0.61 M. cupreus 14.4 28.4 0.51 Vitrinizonites latissimus en 1129 0.60 Glyphyalinia cumberlandiana 182 3.0 0.40 G. rimula 3.4 rer 0.44 Superfamily POLYGYROIDEA Polygyridae: Triodopsinae: Triodopsini Neohelix albolabris 2:5 33137. 0.68 Xolotrema denotata 8.8 16.3 0.54 Triodopsis tridentata 9.6 15.4 0.62 IE vulgata 8.3 14.3 0.58 Polygyridae: Polygyrinae Allogona profunda 15.0 28.6 0:52 Stenotrema edvardsi 516 7.8 0.72 5. stenotrema 7.1 9.9 0.72 Polygyridae: Polygyrinae: Mesodontini Patera appressa 8.7 16.4 0.53 Inflectarius inflectus 6.6 10.9 0.61 Mesodon zaletus 20.2 29.2 0.69 Appalachina sayana 19.0 24.2 0.78 66 EMBERTON APPENDIX 3. Waipipi Reserve (= Jones Bush), New Zealand: taxonomy of native shelled land snails. The slug Athoracophorus bitentaculatus is not included. Classification as in Appendix 1. Genus & Higher Classification Species Subclass PROSOBRANCHIA Order ARCHAEOGASTROPODA Superfamily HYDROCENOIDEA Hydrocenidae Georissa purchasi Order MESOGASTROPODA Superfamily CYCLOPHOROIDEA Liareidae Liarea hochsteteri Cytora cytora C. torquilla Subclass PULMONATA: Order STYLOMMATOPHORA Suborder ORTHURETHRA Superfamily ACHATINELLOIDEA Achatinellidae Lamellidea novoseelandica Suborder SIGMURETHRA Infraorder ACHATINIDA Superfamily RHYTIDOIDEA Rhytididae Delos coresia D. jeffreysiana Rhytida greenwoodi Superfamily PUNCTOIDEA Charopidae Caviella buccinella G: roseveari* Mocella eta M. aff. maculata “Charopa” pseudanguicula SC: chrysaugeia TE aff. pseudanguicula 1 ICH“ fuscosa CH pilsbryi Fectola mira iF unidentata Genus & Higher Classification Species |= infecta Huanodon pseudoleiodon H. hectori Allodiscus urquharti A. aff. granum Geminoropa cookiana Serpho kivi Flammulina perdita IF, chiron “‘Thalassohelix”’ ziczac Suteria ide Phenacohelix giveni 2 pilula P. n. sp. 1 Therasiella neozelandica ТЕ serrata TE aff. neozelandica Te celinda Punctidae “Laoma” mariae ler marina* le aff. marina 1 Sl leimonias “Phrixgnathus” erigone ED ariel APS elaiodes ee moellendorffi er conella ur n. sp. 59 NE poecilosticta “Paralaoma” п. sp. 38 re n. sp. 29 PDS lateumbilicata Bor n. sp. 1 Piz n. sp. 8 a ан. п. sp. 33 EIRE n. sp. 40* a serratocostata * collected by Roscoe and Hazlewood in 1977 but not included in Appendix 3A of Solem et al., (1981: 462). MALACOLOGIA, 1995, 36(1-2): 67-77 DISTRIBUTIONAL DIFFERENCES AMONG ACAVID LAND SNAILS AROUND ANTALAHA, MADAGASCAR: INFERRED CAUSES AND DANGERS OF EXTINCTION Kenneth C. Emberton Department of Malacology, Academy of Natural Sciences, 1900 Bejamin Franklin Parkway, Philadelphia, Pennsylvania 19103-1195, U.S.A. ABSTRACT Seven species of giant, acavid land snails occur in the region of Antalaha, Diego Suarez Province, northeastern Madagascar. Analysis of 4,446 specimens collected from 39 stations along three coast-to-inland transects in October 1990 revealed significant distributional pat- terns in five of the species. Clavator moreleti was more abundant at low elevations and away from the coast, and Ampelita xystera occurred at an intermediate distance from the coast. Both these species were virtually absent from the southern transect (Ankavia River valley), where A. julii was at its most abundant. Ampelita fulgurata was restricted to the inland middle transect (Ankavanana River valley), and A. soulaiana to intermediate-inland sites on the northern transect (Andempona River valley). In contrast, neither A. /атаге! nor Helicophanta amphibu- lima showed significant spatial heterogeneity, although the latter tended to be scarcer near the coast. Combining these data with older, published records supported the results and sug- gested that the distributional patterns can be explained by historical founding events (Clavator тогеей southward along the coast and Ampelita julii northward), by climatic exclusion and coastal land clearing (Ampelita xystera), and by local speciation events (А. soulaiana and pos- sibly A. fulgurata). The Antalaha area is undergoing rapid forest clearing, definitely endangering A. soulaiana and possibly endangering A. julii, A. fulgurata, and the genetically distinct local race of A. xystera. Key words: Gastropoda, Pulmonata, Stylommatophora, Acavoidea, rainforest biogeography, speciation, endangered species. INTRODUCTION This paper is the fourth in a long-term study on the phylogeny, morphological evolution, and biogeography of the acavoid land snails worldwide, beginning with Madagascan taxa because of that island's status as an environ- mental hotspot (Myers, 1988; Emberton, in review a). The first paper (Emberton, 1990) reviewed the acavoids as a monophyletic clade with a Gondwanan distribution and with an unusually great range of shell shapes— from globose to high-spired—wherever they occur, collated all taxonomic and distribu- tional data on Madagascan acavids, and performed a cladistic analysis of 21 species of Madagascan acavids based on five ana- tomical characters from the publications of Fischer-Piette and colleagues (see Emberton, 1990, for references). The second paper (Em- berton, in review b) presented a more rigorous cladistic analysis of 18 species of Madagas- can acavids in the genera Helicophanta, Am- pelita, and Clavator based on a new data set of 71 informative allozyme characters, and 67 predicted from preliminary dissections that acavid terminal genitalia will help resolve their phylogeny. The third paper (Emberton, in press) analyzed the estivation sites and ex- ternal body and shell morphologies of nine diverse acavid species in an evolutionary con- text. This paper analyzes the distributional patterns of the seven acavid species that oc- cur in the rainforest region of Antalaha, north- eastern Madagascar, and compares the re- sults with 20-year-old data from the same region (Fischer-Piette et al., 1973) to hypoth- esize causes for these distributions and to assess the danger of extinction for each spe- cies. MATERIALS AND METHODS Collecting stations are mapped in Figure 1. Transects were run along roads and paths that followed approximately the valleys of the Andempona, Ankavanana, and Ankavia riv- ers, in the vicinity of Antalaha, Diego Suarez Province, Madagascar. Collecting was dur- 68 EMBERTON Sambava FIG. 1. Map of stations. Adapted from Emberton (1994). ing the dry season to aid travel on the un- paved roads (Bradt, 1992). Transportation was by four-wheel-drive vehicle as far as possible (Fig. 1, solid lines), then on foot (dot- ted lines) for the two northern transects. Whenever a village was encountered, resi- dents were informed via translator of the au- thor's intent to purchase land snails collected from native forests, with bonuses for live specimens; a shell of the introduced and common Achatina was shown as the single kind not wanted. Purchases were made along the return trip, questioning the collec- tors for precise locality information (type of forest and its direction and distance from the village, plus date and time of collection). Ad- ditional collections were made away from vil- lages with the aid of two assistants and who- ever else could be recruited from the area. Land clearing, which eradicates native land ACAVID SNAILS OF ANTALAHA, MADAGASCAR 69 snails (Emberton, in prep.), was extensive along all three transects and limited the num- bers and distributions of collection stations; some villages made no collections because of local festivals or because forests were considered too inaccessible. The 39 stations listed in Appendix | pro- duced acavid snails. Station numbers (in the KCE series) provide access to the computer- cataloged vouchers at the Academy of Nat- ural Sciences of Philadelphia. Station coordi- nates and elevations were estimated from topographic maps. Collections were sorted to acavid species and counted for live and dead specimens. Stations were ranked for each of the three variables latitude, inland distance, and eleva- tion. Latitude distinguished among the three transects: (1) northern = Andempona River valley and environs, (2) middle = Ankavanana River valley plus coastal station 252, and (3) southern = Ankavia River valley and environs. Inland distance categorized minimum straight-line distance to the coast into (1) 0-2.9 km, (2) 3-11.9 km, and (3) 12+ km. El- evation was also assigned three ranks: (1) 0-19.9 m, (2) 20-49.9 m, and (3) 50-80 m. Species distributions were assessed by analyses of variance (ANOVAs) (Sokal 8 Rohlf, 1969). Station species counts (live plus dead) were transformed to proportions of to- tal station acavids. These proportions were used as dependent variables for (1) one-way ANOVAs with treatment = elevation, (2) two- way ANOVAs with treatments = latitude and inland distance, and (3) two-way ANOVAs with treatments = latitude and elevation. Three-way ANOVAs and two-way ANOVAs with treatments = inland distance and eleva- tion were not possible because of empty treatment cells. ANOVAs were by least- squares linear likelihood estimates, using SYSTAT multivariate general linear hypothe- sis programs (Wilkinson, 1988). Computer outputs of partitioned sums of squares were used to calculate the proportions of total species-distribution variance explained by treatments and their interaction, and unex- plained, using a hand calculator. ANOVA cell means for each species were computed us- ing SYSTAT (Wilkinson, 1988). Distribution maps for the seven acavid species were prepared by combining present data with localities listed in Fischer-Piette et al. (1973). Antalaha-region distributions were compared with total Madagascan distribu- tions as summarized by Emberton (1990). RESULTS Figure 2 illustrates (in a phylogenetic con- text) the seven acavid species that were found, and Table 1 lists the number of each species collected at each station. A total of 4,446 specimens was collected, with individ- ual station collections ranging from two to 2,297. Maximum acavid site-diversity was four species (stations 206, 245-249). (Non- acavid large land snails of the helicarionid genus Kalidos and the prosobranch genera Acroptychia and Tropidophora were also col- lected in abundance; several stations were also collected for small to minute snails [Em- berton, 1994].) Ranks of each station for latitude, inland distance, and elevation are given in Table 1. One-way ANOVAs for elevation showed sig- nificant treatment effects in only one species, Clavator moreleti, for which elevation ex- plained 31% of distributional variance (p < 0.001). Clavator moreleti was more abundant proportionally at lower elevations, with ANOVA cell means: Elev. Cell Mean Rank Count Proportion 1 10 0.36 2 21 0.01 3 8 0.08 Two-way ANOVAs with treatments = lati- tude and inland distance showed significant treatment effects for five of the seven acavid species (Table 2). Table 3 gives cell sizes and cell mean proportions for each species from these ANOVAs. Ampelita julii was signifi- cantly stratified by latitude (Table 2), repre- senting a fourth of all collections along the entire southern transect, but absent from all but some inland stations of the northern and middle transects, where it averaged less than a tenth of collections (Table 3). Ampelita julir s presumed sister species, А. soulaiana, was significantly (Table 2) restricted to the stretch between 3 and 11.9 km inland along the An- dempona River valley (= northern transect) (Table 3). Combining these two species as A. (Eurystyla) removed significant inland-dis- tance x latitude interaction effects but not significant latitudinal effects, which explained 26% of the distribution of this subgenus (Ta- ble 2). Ampelita xystera was significantly affected by inland distance, averaging a fourth or 70 EMBERTON FIG. 2. Camera lucida drawings of seven species of acavid land snails from the region of Antalaha, northeastern Madagascar (Fig. 1). All are to the same size scale; Cm is 70.0 mm in height. Al = Ampelita (Ampelita) lamarei (two shells showing variation, station 205, Academy of Natural Sciences of Philadelphia [ANSP] catalog number 391391), Aj = A. (Eurystyla) julii (206, ANSP 391399), As = A. (E.) soulaiana (248, ANSP 391411), Ах = A. (Xystera) xystera (206, ANSP 391429), Af = A. (X.) fulgurata (205, ANSP 391434), Cm = Clavator moreleti (206, ANSP 391452), Ha = Helicophanta amphibulima (204, ANSP 391482). The phy- logeny is based on allozymes for Ha, Cm, Al, Aj, and Ax (Emberton, in review b); and on shell similarity for As and Af (Emberton, 1990). more of collections 3-12 km from the coast, not statistically significant, despite the fact but virtually absent from both coastal and that no A. xystera were collected from the more inland stations; latitudinal effects were southern transect. The related A. fulgurata ACAVID SNAILS OF ANTALAHA, MADAGASCAR 71 TABLE 1. Numbers of acavid snails collected at 39 stations in the region of Antalaha, northeastern Madagascar Ampelita Clavator Helicophanta Sta lamarei julii soulaiana xystera fulgurata moreleti amphibulima # Lat In Ev L D L D L D L D L D L D L D Total 216 OO 0 290 оо 0 0 оо 0 0 13 0 3 45 250 1 1 3 0 оо оо 0 0 оо 0 11 26 0 22 59 244 1 2722 0 0 0 оо 0 0 оо 0 0 0 0 3 3 245 22а 0 8 0 оо 5 0 28 0 0 0 оо 2 43 248 i) 22 0 7 0 0 0 5 0 8 0 0 0 оо 5 25 249 112227, 22 0 4 0 0 0 6 0 9 0 0 0 0 1 72 92 246 W283 0 4 0 0 0 7 0 12 0 0 0 On 50 6 29 247 2 0 2 0 оо 0 0 оо 0 0 оо 0 2 223 1 3 2 0 00 оо 0 0 01 0 0 0 оо 2 21 225 as 22 0 4 0 1 “0 0 0 оо 0 0 0 0 0 5 226 ТЗ; 2 0 100 оо 0 0 оо 0 0 оо 0 10 229 Эх. 42 0 20 оо 0 0 оо 0 0 оо 4 6 232 ir 32 0 10 1.0 0 0 оо 0 0 0 0 8 10 233 № 23252 0 2 0 2 0 0 0 0.0 0 0 оо 25 29 234 19.783,22 Or 12-30 6 0 0 0 оо 0 0 0 1 97 116 235 Wes: 72 0 10 оо 0 0 оо 0 0 0.0 8 9 236 ln Se 2 0 7 0 2 0 0 0 070 0 0 0 0 1 10 237 Е 31.22 1 160 5 0 0 0 оо 0 0 0 0 43 65 238 11.43? 2 0 16 0 оо 0 0 оо 0 0 оо 1 17 239 13: $2 0 6 0 0 0 0 0 12 0 0 0 0 0 0 18 240 1 92 0 890 оо 0 0 оо 0 0 оо 41 130 241 IES IZ 0 41 0 оо 0 0 оо 0 0 оо 0 41 243 LS? 0 2-10 0 0 0 0 оо 0 0 02:0 0 2 218 SES 0 6 0 оо 0 0 оо 0 0 оо 1 и 224 1 23 3 0 8720 Ov 0 0 0 оо 0 0 0170 100 108 200 ZA 1. 0 0 0 оо 0 0 оо 0 % 2510 0 9 252 A: 2 00 o” 0 0 0 оо 0 0 оо 0 2 201 ZO del 0 оо оо 0 0 ¡NRO 0 1 1 0 0 3 206 EAN 0 224 0 202 0 0 42 614 0 0 97 1118 0 0 2297 207 IL EA 0 0 0 2 0 0 0 60 0 0 17 330 0 0 409 253 2 2 1 0 оо оо 0 0 46 0 0 40 230 0 109 204 2 3 3 6 60 0 0 0 0 0 оо 0 0 0 1 299 366 205 2 3 3 Зи 92 2153-10 0 0 оо 65 0 1 0 50 266 208 ЗАЗ 0 50 2770 0 0 оо 0 0 оо 0 7 210 | 0 lO оо 0 0 оо 0 0 оо 1 2 211 Зо 0 10 14 0 0 0 оо 0 0 о 0 21 36 214 3° 21 0 0 0 оо 0 0 оо 0 0 оо 8 8 215 3 2.2 0 2 0 4 0 0 0 оо 0 0 ONO 1 И 212 3 3 1 0 7 0 6 0 0 0 оо 0 0 оо 10 23 Total 12 669 2 300 0 23 42 79 0 65 173 1514 3 853 4446 Live/Dead 0.02 0.01 0.00 0.05 0.00 0.11 0.00 Stations are numbered as in Fig. 1 and are arranged geographically by latitude (Lat: 1, 2, and 3 = north, middle, and south transects), inland distance (Inl: 1, 2, and 3 = 0-2.9, 3-11.9, and 12+ km from coast), and elevation (Elv: 1, 2, and 3 = 0-19.9, 20-49.9, and 50—80 m). D = dead, L = live. was collected only at station 205 (middle transect, inland distance 12+ km), where it comprised 12% of the collection. Combining these two species as A. (Xystera) eliminated significant spatial heterogeneity (Table 2). Clavator moreleti was significantly affected by both latitude and inland distance as well as by their interaction. Although it was absent or virtually absent from the entire southern transect, from the far-inland middle transect, and from the mid- and far-inland northern transect, it averaged half or more of total col- lections from other regional divisions (Table 3). Thus С. moreleti т this study was re- stricted to and dominant at the mouth of the Andempona River valley (< 3 km from the 72 EMBERTON TABLE 2. Variances in species’ proportions explained by latitude, by distance from the coast, and by their interaction Latitude Inland Lat*Inl Unexpl. A. lamarei 0.01 0.09 0.04 0.85 А. juli 0355010240101 0.63 А. soulaiana 0.08 0108/0235 0.61 А. xystera 0.03 0.16* 0.07 0.74 A. fulgurata 04135 O0 2gE 0.48 C. moreleti 0232220550802 02 H. amphibulima 0.03 0.07 0.07 0.82 А. (Eurystyla) 0.26** 0.01 0.05 0.68 А. (Xystera) 0.05 0.13 0.06 0.76 * р < 0.05, ** р < 0.01, ** р < 0.001 (ANOVAs, df 2, 30 or [interactions] 4, 30). TABLE 3. Cell means from the ANOVAs of Ta- ble 2 Cell Sizes Inland 1 2 3 1 2 6 1174 Lat 2 2 4 2 3 1 4 1 Ampelita (A.) lamarei 1 0.32 0.28 0.49 Lat 2 0.50 0.02 0.27 3 0.71 0.20 0.30 А. (Eurystyla) julii 1 0.00 0.00 0.04 Lat 2 0.00 0.02 0.10 3 0.29 0.24 0.26 А. (E.) soulaiana 1 0.00 0.10 0.00 Lat 2 0.00 0.00 0.00 3 0.00 0.00 0.00 A. (Xystera) xystera 1 0.00 0.25 0.04 Lat 2 0.00 0.30 0.00 3 0.00 0.00 0.00 A. (X.) fulgurata 1 0.00 0.00 0.00 Lat 2 0.00 0.00 0.12 3 0.00 0.00 0.00 Clavator moreleti 1 0.46 0.00 0.00 Lat 2 0.50 0.66 0.00 3 0.00 0.00 0.00 Helicophanta amphibulima 1 0.22 0.37 0.43 Lat 2 0.00 0.00 0.50 3 0.00 0.56 0.44 coast) along the lower Ankavanana River val- ley (< 12 km from the coast). Neither Ampelita lamarei nor Helicophanta amphibulima showed significant effects of latitude or inland distance (Table 2). Ampelita lamarei was extremely widespread, with no tendencies toward geographical trends. He- licophanta amphibulima, on the other hand, was conspicuously (though not significantly in a statistical sense) more common at greater distances inland. This trend was most pronounced in the middle transect, where H. amphibulima was found only at the most in- land station; was distinct in the southern transect, where it did not appear in the most coastal station; and was subtle in the north- ern transect, where it graded from 22% to 37% to 43% of collections going inland. Two-way ANOVAs with treatments = in- land distance and elevation had the following cell counts: Inland Distance 1 2 3 1 2 1 2 Elevation 2 7 5 2 3 1 15 4 In these ANOVAs, no species showed signif- icant treatment effects from elevation or from inland x elevation interaction. Only two spe- cies showed significant effects from inland distance: Ampelita soulaiana (p < 0.01, 24% of variance explained) and Clavator moreleti (p < 0.05, 17% of variance explained). When A. soulaiana was combined with A. julii in A. (Eurystyla), no significant treatment effects remained. High but less than significant in- land-distance effects were evident for A. xys- tera (0.05 < p < 0.10, 16% of variance ex- plained). Figure 3 combines locality data from this study with those of Fischer-Piette et al. (1973) to produce distributional maps for the seven species. Adding the 20-year-old local- ities generally supports the results of this analysis. Thus Ampelita lamarei and Heli- cophanta amphibulima are both widespread and unlocalized in the region, although the former was found at more of the distant in- land sites. The northern coastal sites for H. amphibulima (Fig. 3) were on hills (Fischer- Piette et al., 1973; Table 1); no elevational data were given for this species’s two south- ern coastal sites. Clavator moreleti in the region of Antalaha is clearly a species of the coast and the lower river valleys (Fig. 3). Fischer-Piette et al.’s (1973) locality of ‘‘Antsiranamatso” is inter- ACAVID SNAILS OF ANTALAHA, MADAGASCAR 73 à Antalaha dAntongil \ Antalaha ‚ Baie Baie FIG. 3. Distribution maps based on this study and Fischer-Piette et al. lamarei, Aj = A. (Eurystyla) julii (dots), As = A. (E.) soulaiana (stars), Ax = A. (Xystera) xystera (1973). Al = Ampelita (Ampelita) (dots), Af = A. (Х.) fulgurata (stars), Ст = Clavator moreleti (open circle = dubious record), Ha = Helicophanta amphibu- lima. 74 EMBERTON preted here as spurious (Fig. 3, open cir- cle). Ampelita julis significant trend of occur- ring farther inland south of Antalaha (Table 3) is supported by the additional southwestern site of Antsambalahy, but is strongly negated by the extreme northwestern site of Amboa- hangibe (Fig. 3; Fischer-Piette et al., 1973). The seemingly related A. soulaiana is known from only three localites along a short stretch of the Andempona River valley, where it is closely bounded both upstream and down- stream by А. julii (Fig. 3). Ampelita xystera's inland distributional limit of approximately 12 km is well sup- ported by the seven additional localities pro- vided by Fischer-Piette et al. (1973) (Fig. 3). The coastal limit of 3 km (Table 3) is negated, however, by one of these localities: Antsera- nambidy (1 km south of Ampahana), a village within 0.2 km of the coast. Ampelita xystera's probable relative, A. fulgurata, is known from only five localities, four of which are mapped in Figure 3. The fifth locality, “Ambohitsitan- drona, 700 m” (Fischer-Piette, 1952), lies somewhere on the Masoala Peninsula, either south of Antalaha as mapped by Fischer- Piette (1952: fig. 1, #6), or on the peninsula's west coast (along the Baie d'Antongil, Fig. 3) near Mahalevona or near Ambanizana (Viette, 1991). There is a distinct geographical seg- regation between A. fulgurata and A. xystera: fulgurata occurs more inland and upland, but comes very close to xystera in the Anka- vanana River valley (Fig. 3). DISCUSSION The area around Antalaha is now Mada- gascar's most intensively surveyed region for acavid land snails. This study, by hiring native collectors as much as possible, includes the largest collections of acavids ever made in Madagascar. Acavoids are among the world’s largest, most ancient, relict, and K-selected non-orthurethran stylommato- phoran snails, and Madagascar contains the greatest surviving radiation of acavoids (Em- berton, 1990, in review a). Madagascar’s en- vironmental crisis (Myers, 1988; Green & Sussman, 1990) may put some of its acavids in danger of extinction, despite its system of Reserves and National Parks (Nicoll & Lan- grand, 1989). Helicophanta amphibulima seems safe from extinction at present. Although Ember- ton (1990) mapped its distribution along Madagascar’s entire western length (follow- ing Fischer-Piette, 1950), this species is also known in the east from Analamazoatra (= Perinet Reserve; Fischer-Piette & Garreau de Loubresse, 1965) and from the Antalaha area (Fischer-Piette et al., 1973; this study). Thus H. amphibulima is widespread and is pro- tected in several reserves. In addition, as dis- covered in this study, it has a broad ecolog- ical tolerance and a widespread local distribution. Clavator moreleti is probably endangered in the region of Antalaha, because of current deforestation of its coastal and river-mouth habitats. In the north, however, C. moreleti has been collected from high elevations that are under protection (Montagne d’Ambre, Mont Tsaratanana), as well as on Nosy Be, where it may be protected in Lokobe Reserve (Fischer-Piette & Salvat, 1963). The coastal distribution of this species in the Antalaha re- gion is enigmatic, and could be due to its colonization history: perhaps C. moreleti in- vaded the region relatively recently along the coast, so has not had enough time to pene- trate far inland. Ampelita (A.) lamarei is widely distributed in northern and eastern Madagascar (Fischer- Piette, 1952; Emberton, 1990), and its broad local distribution (this study) further protects it from extinction. Ampelita (Eurystyla) julii could be endan- gered. Its type—and sole western—locality, Ambanja (opposite Nosy Be; Fischer-Piette, 1952), is probably deforested by now, and none of its eastern localities (Fig. 3, plus Maroantsetra, at the head of Baie d’Antongil) falls under protection and will be deforested within a decade or two. It can be reasonably hoped, however, that A. julii occurs within Marojezy Reserve (northwest of Antalaha) or within what will hopefully become Masoala National Park (southwest of Antalaha). The significantly increasing abundance of A. ий toward the south in this study suggests that its range does extend more to the south, and that it has only relatively recently spread (and speciated) northward. Ampelita (E.) soulaiana is definitely endan- gered. This taxon apparently represents a re- cent speciation or subspeciation event within the range of A. ий, but no live-collected specimens are available to test this hypoth- esis. The extremely small range of A. soulai- ana (Fig. 3) will almost certainly be deforested in the near future. Ampelita (Xystera) xystera 1$ safe as a spe- ACAVID SNAILS OF ANTALAHA, MADAGASCAR 75 cies, as it has the widest known range of any Madagascan acavid (most of the northern half plus all the eastern rainforest). Its 12-km inland limit in the Antalaha region may mean its eventual local eradication; this limit may be due to climatic exclusion, “overcome” by speciation to form the inland A. (X.) fulgurata (see below). Ampelita xystera's current 3-km coastal limit in the Antalaha area suggests advancing eradication by coastal deforesta- tion. Antalaha A. xystera are well distin- guished genetically from other regions, and are more plesiomorphic phylogenetically (Emberton, in review b), so should be saved. Ampelita (X.) fulgurata is probably endan- gered. All of its four exactly known localities are unprotected and certain to be deforested within the next few years, despite its inland range (Fig. 3). The inland parapatry of this range relative to the related A. xystera sug- gests a parapatric speciation event as the genesis of A. fulgurata; other species with similar shells are widely separated geograph- ically (Fischer-Piette, 1952; Emberton, 1990). The range given for A. fulgurata in Emberton (1990) was too broadly interpreted. Ampelita fulgurata's fifth locality of “Ambohitsitan- drona” was interpreted by Fischer-Piette (1952) as an unprotected site south of Anta- laha, but there are two mountains so named on the western Masoala Peninsula, providing hope that A. fulgurata has a wide enough range to fall under the protection of a pro- posed Masoala National Park. The only local- ity where A. fulgurata has been collected in the past 20 years is station 205 (Fig. 1); this species has never been collected alive. In sum, the distributional patterns of Anta- laha-area acavid snails are significantly dif- ferent and can be explained by historical founding events (Clavator moreleti and Am- pelita juliñ, by climatic exclusion and coastal land clearing (Ampelita xystera), and by local speciation events (А. soulaiana and possibly A. fulgurata). This area is unprotected and 1$ undergoing rapid deforestation, definitely en- dangering A. soulaiana and possibly endan- gering A. julii, A. fulgurata, and the genetically distinct and plesiomorphic local race of A. xystera. ACKNOWLEDGEMENTS This work was supported by a 1990 travel grant from the American Philosophical Soci- ety and by National Science Foundation grant DEB-9201060. Dr. Patricia Wright, In- ternational Director of the Ranomafana National Park Project, helped arrange for col- lecting and export permits from Madagas- car's Department of Waters and Forests. Do- minike Ratova, Dominique Harison, Angus Schofield, and many Malagasy villagers as- sisted in the field; Aldus Andriamamonjy and his family provided lodging in Antananarivo and helped arrange transportation to Anta- laha; Mr. Tatienne helped find transportation in Antalaha; Lu Zhang helped sort collections to species; Elizabeth Perry researched sta- tion coordinates and elevations; and Andria Garback computer cataloged the collections. | also wish to thank the owner of the yellow four-wheel-drive in Antalaha for renting to me during the vanilla harvest and for his coura- geous driving on seemingly impassable roads. LITERATURE CITED BRADT, H., 1992, Guide to Madagascar, 3rd ed. Hunter Publishing, Edison, New Jersey, 230 pp. EMBERTON, K. C., 1990, Acavid land snails of Madagascar: subgeneric revision based on pub- lished data (Gastropoda: Pulmonata: Stylom- matophora). Proceedings of the Academy of Natural Sciences of Philadelphia, 142: 15-31. EMBERTON, K. C., 1994, Thirty new species of Madagascan land snails. Proceedings of the Academy of Natural Sciences of Philadelphia, 145: 147-189. EMBERTON, K. C., in press, Morphology and aes- tivation behaviour in some Madagascan acavid land snails. Biological Journal of the Linnean So- ciety. EMBERTON, K. C., in review a, On the endangered biodiversity of Madagascan land snails, in: A. C. VAN BRUGGAN & S. WELLS, eds., Molluscan diver- sity and conservation (submitted). EMBERTON, K. C., in review b, An allozyme-based phylogeny for 18 species of Madagascan acavid land snails. Veliger (submitted). EMBERTON, K. C., in prep., Effects of land clear- ing and introduced exotics on the land-snail fauna of northeastern Madagascar. FISCHER-PIETTE, E., 1950, Mollusques terrestres de Madagascar genre Helicophanta. Journal de Conchyliologie, Paris, 90: 32-106. FISCHER-PIETTE, E., 1952, Mollusques terrestres de Madagascar genre Ampelita. Journal de Con- chyliologie, Paris, 92: 1-59. FISCHER-PIETTE, E., М. CAUQUOIN & A. TES- TUD, 1973, Mollusques terrestres récoltés par M. Soula dans la région d'Antalaha (Madagas- car). Bulletin du Museum National d’Histoire Na- turelle, Zoologie, 94: 477-531. FISCHER-PIETTE, E. & М. GARREAU DE LOU- 76 EMBERTON BRESSE, 1965, Mollusques terrestres de Mada- gascar Famille Acavidae. Journal de Conchylio- logie, Paris, 104: 129-160. FISCHER-PIETTE, E. 8 F. SALVAT, 1963, Mol- lusques terrestres de Madagascar, genre Clava- tor. Journal de Conchyliologie, 103: 53-74. GREEN, С. М. & R. W. SUSSMAN, 1990, Defores- tation history of the eastern rain forests of Mad- agascar from satellite images. Science, 248: 212-215. MYERS, N., 1988, Threatened biotas: ““hotspots” in tropical forests. The Environmentalist, 8: 1-20. NICOLL, M. E. 4 O. LANGRAND, 1989, Madagas- car: revue de la conservation et des aires proté- gées. World Wide Fund For Nature, Gland, Swit- zerland, xvii + 374 pp. SOKAL, В. В. & Е. J. ROHLF, 1969, Biometry. W. H. Freeman, San Francisco, 776 pp. VIETTE, P., 1991, Chief field stations where insects were collected in Madagascar. Faune de Mada- gascar, Supplément 2: 1-88. WILKINSON, L., 1988, SYSTAT: the system for sta- tistics. SYSTAT, Inc., Evanston, Illinois, U.S.A. Revised Ms accepted 9 February 1994 APPENDIX | 200. 14.52.05S 50.14.40E, elev. approx. 10 m, approx. 1 km W of Andripika (approx. 6 km NW of Antalaha), remnant degraded for- est on partially cleared and planted hillside beside road and river, 13 Oct. 1990, 3.0 per- son-hours. 201. 14.51.40$ 50.13.40E, elev. approx. 10 m, approx. 2 km E of Valambana (approx. 7 km NW of Antalaha), talus and boulders beside road and river, 13 Oct. 1990, 1.5 per- son-hours. 204. 14.48.20S 49.59.10E, elev. approx. 50 m, northwest spur of Sarahandrano Peak, within bend of Ankavanana River, near An- dranofotsy (WNW of Antalaha), virgin forest on slopes and ridgetop, 14 Oct. 1990, >10 person-hours. 205. 14.50.00S 50.08.25E, elev. approx. 50 m, region of Antsahanoro (approx. 17 km WNW of Antalaha), native forest, 13-15 Oct. 1990, >20 person-hours. 206. 14.50.20S 50.12.45E, elev. approx. 10 т, vicinity of Malotrandrohely (approx. 17 km WNW of Antalaha), native forest, 13-15 Oct. 1990, >20 person-hours. 207. 14.50.50S 50.13.00E, elev. approx. 10 m, “Bimanary, 40 m W of Valambanina” (approx. 10 km NW of Antalaha), native for- est, 16 Oct. 1990, >10 person-hours. 208. 14.55.25S 50.16.25E, elev. approx. 50 m, region of Andrakarakani-ali, approx. 2 km SSW of Antalaha, 12-17 Oct. 1990, >20 person-hours. 210. 14.59.15S 50.12.50E, elev. approx. 10 m, region of Antserasera (approx. 12 km SW of Antalaha), 12-17 Oct. 1990, >10 per- son-hours. 211. 14.59.15S 50.12.50E, elev. approx. 10 m, region of Antserasera (approx. 12 km sw of Antalaha), 12-17 Oct. 1990, >10 per- son-hours. 212. 14.59.15S 50.12.00E, elev. approx. 10 m, region of Ambodimanga (approx. 15 km SW of Antalaha), 12-17 Oct. 1990, >10 person-hours. 214. 14.58.20S 50.13.25E, elev. approx. 10 m, “Mahatsara” (Mahasoa?), near Ant- serasera (approx. 10 km SW of Antalaha), 17 Oct. 1990, approx. 3 person-hours. 215. 14.56.25S 50.15.10E, elev. approx. 20 m, Andrakarakan’ 1, approx. 3 km NE of Ambohitsara (approx. 8 km SW of Antalaha), approx. 0.5 km south of road, patch of virgin forest on hillside and hilltop, 17 Oct. 1990, 4 person-hours. 216. 14.37.35$ 50.11.10E, elev. approx. 20 m, Ambohimanodina (hill), approx. 1.2 km S of Ambodipont-Sahana (approx. 35 km N of Antalaha), burn on edge of impacted for- est, 18 Oct. 1990, 0.5 person-hours. 218. 14.40.00S 50.03.50E, elev. approx. 80 m, approx. 2-3 km E of Andampibe (7 km S of Lanjarivo, and approx. 40 km NNW of Antalaha), virgin dry pandanus-palm forest on quartz sand, 18 Oct. 1990, approx. 8 per- son-hours. 223. 14.36.55S 50.03.25E, elev. approx. 30 т, vicinity of Mangatsahatsa (1.5 km $ of Lanjarivo, and approx. 40 km NNW of Anta- laha), native forest, 19 Oct. 1990, aprox. 6 person-hours. 224. 14.36.15S 50.02.50E, elev. approx. 50 m, 2 km NW of Mangatsahatsa (1.5 km S of Lanjarivo, and approx. 40 km NNW of Antalaha): native forest, 18 Oct. 1990, ap- prox. 6 person-hours. 225. 14.35.05S 50.03.20E (Lanjarivo), elev. approx. 35 m, approx. 4 km NW of Lan- jarivo (approx. 40 km NNW of Antalaha), na- tive forest, 18 Oct. 1990, approx. 9 person- hours. 226. 14.36.05S 50.03.20E (Lanjarivo), elev. approx. 35 m, approx. 2 km ENE of Lan- jarivo (approx. 40 km nnw of Antalaha), virgin forest, 19 Oct. 1990, approx. 4 person-hours. 229. 14.36.05S 50.03.20E (Lanjarivo), elev. approx. 35 m, approx. 2 km N of Lan- ACAVID SNAILS OF ANTALAHA, MADAGASCAR IH jarivo (approx. 40 km NNW of Antalaha), par- tially cleared forest, 19 Oct. 1990, approx 4 person-hours. 232. 14.36.05$ 50.03.20E (Lanjarivo), elev. approx. 35 m, approx. 3 km w of Lan- jarivo (approx. 40 km NNW of Antalaha), par- tially cleared forest in different place from 233, 19 Oct. 1990, approx. 2 person-hours. 233. 14.36.055 50.03.20E (Lanjarivo), elev. approx. 35 m, approx. 3 km W of Lan- jarivo (approx. 40 km NNW of Antalaha), par- tially cleared forest in different place from 232, 19 Oct. 1990, approx. 4 person-hours. 234. 14.36.05S 50.03.20E (Lanjarivo), elev. approx. 35 m, approx. 3 km w of Lana- jarivo (approx. 40 km nnw of Antalaha), virgin forest, 19 Oct. 1990, approx. 10 person- hours. 235. 14.36.055 50.03.20E (Lanjarivo), elev. approx. 35 m, approx. 3 km sw of Lan- jarivo (approx. 40 km nnw of Antalaha), virgin forest, 18 Oct. 1990, approx. 3 person-hours. 236. 14.36.05$ 50.03.20E (Lanjarivo), elev. approx. 35 m, approx. 2 km NW of Lan- jarivo (approx. 40 km nnw of Antalaha), virgin forest, 18 Oct. 1990, approx. 2 person-hours. 237. 14.36.05$ 50.03.20E (Lanjarivo), elev. approx. 35 m, region of Lanjarivo (ap- prox. 40 km NNW of Antalaha), forest, 18-19 Oct. 1990, >20 person-hours. 238. 14.36.30$ 50.04.40E (Ambodilalona), elev. approx. 30 m, between “Ankorakabe” and Ambodilalona (approx. 40 km NNW of Antalaha), along path: virgin forest, 19 Oct. 1990, approx. 5 person-hours. 239. 14.36.30$ 50.04.40E (Ambodilalona), elev. approx. 30 m, approx. 4 km N of Am- bodilalona (approx. 40 km nnw of Antalaha), virgin forest, 19 Oct. 1990, approx. 2 person- hours. 240. 14.36.30$ 50.04.40E (Ambodilalona), elev. approx. 30 m, region of Ambodilalona (approx. 40 km NNW of Antalaha), forest, 18-19 Oct. 1990, >20 person-hours. 241. 14.36.30S 50.04.40E (Ambodilalona), elev. approx. 30 m, approx. 2 km N of Am- bodilalona (approx. 40 km NNW of Antalaha), virgin forest, 19 Oct. 1990, approx. 4 person- hours. 243. 14.36.30S 50.04.40E (Ambodilalona), elev. approx. 30 m, approx. 2 km W of “Am- bosimila” (unmapped village approx. 1 km E of Ambodilalona, approx. 40 km nnw of Anta- laha), virgin forest, 19 Oct. 1990, approx. 1 person-hour. 244. 14.36.4085 50.07.15E (Ambinanifaho), elev. approx. 30 m, approx. 4 km SW of Am- binanifaho (approx. 36 km NNW of Antalaha), virgin forest, 18 Oct. 1990, approx. 2 person- hours. 245. 14.36.4058 50.07.15E (Ambinanifaho), elev. approx. 30 m, region of Ambinanifaho (approx. 36 km NNW of Antalaha), virgin for- est, 18 Oct. 1990, >20 person-hours. 246. 14.38.0058 50.07.25E, elev. approx. 70 m, approx. 3 km $ of Ambinanifaho (ap- prox. 36 km NNW of Antalaha), partially cleared forest, 19 Oct. 1990, approx. 4 per- son-hours. 247. 14.36.3085 50.09.00E, elev. approx. 50 m, Beramboa (hill), approx. 2 km E of Am- binanifaho (approx. 36 km nnw of Antalaha), partially cleared forest and burn, 19 Oct. 1990, approx. 9 person-hours. 248. 14.37.20$ 50.08.10E, elev. approx. 25 m, approx. 2 km E of Ambinanifaho (ap- prox. 36 km NNW of Antalaha), south side of road: lowland virgin forest, 19 Oct. 1990, ap- prox. 5 person-hours. 249. 14.37.20$ 50.08.10E, elev. approx. 25 m, 7 km w of Ambodipont-sahana (ap- prox. 35 km N of Antalaha), virgin forest, 19 Oct. 1990, approx. 4 person-hours. 250. 14.37.20$ 50.11.05E, elev. approx. 50 m, Ambohimanodina (hill), 1 km S of Am- bodipont-Sahana (approx. 35 km N of Anta- laha), virgin forest near coast, 19 Oct. 1990, approx. 12 person-hours. 252. 14.44.10$ 50.13.20E, elev. approx. 5 m, 1 КтЕ of Tampolo, near coast (approx. 20 km N of Antalaha), on trees near shore. 18 Oct. 1990, approx. 8 person-hours. 253. 14.51.25$ 50.13.00E, elev. approx. 10 m, 0.5 km W of Valambana (approx. 13.5 km NW of Antalaha), virgin forest, 18 Oct. 1990, approx. 6 person-hours. Cie MALACOLOGIA, 1995, 36(1-2): 79-89 EFFECT OF TEMPERATURE ON REPRODUCTION IN PLANORBARIUS CORNEUS (L.) AND PLANORBIS PLANORBIS (L.) THROUGHOUT THE LIFE SPAN Katherine Costil 8 Jacques Daguzan Laboratoire de Zoologie et Ecophysiologie (L. А. INRA), Université de Rennes I, Campus de Beaulieu, Av. du General Leclerc, 35042 Rennes Cedex, France ABSTRACT The reproduction of two planorbid species, Planorbarius corneus and Planorbis planorbis, was studied at 5, 10, 15, 20 and 25°С. All reproduction parameters were affected by temper- ature. Planorbis planorbis began to lay eggs from 10°C, whereas P. corneus reproduced from 15°С. Sexual maturity was earlier at higher temperatures. The snails spent most of their life span in reproduction (at least, 49% for P. corneus at 25°C, and 87% for P. planorbis at 15°C). Special attention was payed to abnormalities of eggs and egg capsules. Planorbarius corneus placed at 20°C and P. planorbis at 15°C produced the maximum number of descendants, 1,600 and 3,357 newly hatched snails per individual respectively. Key words: Planorbidae, Planorbarius corneus, Planorbis planorbis, reproduction, tempera- ture, ecophysiology. INTRODUCTION The reproduction of freshwater snails, and especially bilharziasis intermediate host snails, has been the subject of many papers (Cole, 1925; Chernin 8 Michelson, 1957; van der Schalie 8 Berry, 1973; Aboul-Ela 4 Bed- diny, 1980; Seuge & Bluzat, 1983; Vianey- Liaud, 1990). However, these studies were limited in time and, to our knowledge, little quantitative data about the descendants of pluriannual species are available. Planorbarius corneus and Planorbis plan- orbis are hermaphrodite freshwater snails. The complexity of their reproductive system is increased by internal fertilization, auto- and allosperm storage, egg capsule complexity, and autolysis of foreign sperm (Geraerts 4 Joosse, 1984). Eggs, which comprise a zy- gote surrounded by perivitelline fluid and membrane, are embedded in jelly and en- closed in a common egg capsule. Freshwater snails usually practise cross-fertilization (Dun- can, 1975; Vianey-Liaud, 1990). Planorbis planorbis cannot be considered as a self-fer- tile species, whereas isolated P. corneus pro- duces few capsules, eggs and egg cells (Cos- til, 1993). Little is known about the ecophysiology of reproduction in these planorbid species, and the aim of this study 1$ therefore to test the influence of temperature on: age of snails at onset of sexual maturity; reproduction-period duration; number of eggs per capsule; “abnormalities” of the capsules and the eggs; 79 variation of the numbers of capsules and eggs per individual; and fecundity (number of eggs per individual) and fertility (number of newly hatched snails per adult) throughout the life span. MATERIALS AND METHODS Snails were collected in spring 1987 from two ponds located near Rennes, Brittany, France. Brought to the laboratory, they laid egg capsules. Reproduction was studied on snails hatched in the laboratory. The newly hatched snails were grouped in equal sized sets of 17 individuals (P. corneus; ANOVA: significance level of 95%: N = 374, F = 1.84, p = 0.12) or 20 specimens (P. planorbis: N = 400, F = 0.01; p = 0.99) at five constant tem- peratures: 5, 10, 15, 20 and 25°C. In the case of P. corneus, 374 individuals were reared as five sets at 5°C and 10°C, and four sets at 15, 20 and 25°C. For P. planorbis, 400 individuals were divided into four sets at each tempera- ture. At every temperature and according to mortality, the groups were adjusted to con- stant density. Between sets at different tem- peratures, plastic partitions were placed in the aquaria so that each snail could be in the same volume of water. Up to the age of ten months, they had a volume of 45 ml then 150 ml of pond water per individual, and they were fed with fresh lettuce ad /ibitum in a 12/12 h light/dark photoperiod. Sexual maturity was considered acquired when the first egg mass was observed. Every 80 COSTIL 8 DAGUZAN week, the survivors were counted, the egg capsules collected, and eggs counted under a binocular microscope. Each experiment continued until the death of the last snail. Ab- normalities in the clutches were studied throughout the life of P. corneus reared at 20°C and 25°C, but only at fixed dates in the other cases. At each temperature, the following param- eters were computed: mean age (A,,) and, mean (D,,), minimum (D,,,,.) and maximum (D,,.,) shell diameters at onset of sexual maturity; reproduction period duration, T,., (in weeks); reproduction period duration in relation to max- imum life span, T,., (in %); mean numbers of egg capsules (N.) and eggs (Ме) per snail alive per two weeks; mean number of eggs per egg capsule (N.,.); fecundity of snails (Fec): cumulative number of eggs laid per snail throughout the life span; the fecundity per reproduction week is also calcu- lated (Fec/T,.,); proportion of capsules without eggs (in relation to N.) (in %); proportion of capsules containing one or more eggs without egg cells (in relation to N.) (C,,) (in %); proportion of eggs without egg cells (in relation to N.) (Ey) (in %); proportion of eggs including two or more egg cells (in relation to М.) (in %); and snail fertility (Fer): cumulative number of newly hatched snails produced per snail throughout the life span. Considering the survivorship (Costil, 1994) and the reproduction of the planorbids, Leslie matrices were constructed (Leslie, 1945). The descending vertical elements represented the successive age classes (from birth to maxi- mum four years). The probability of an indi- vidual surviving from age М to age М + 1 was stated on the diagonal, and mean individual fertility per year class on the top horizontal row. RESULTS Onset of Sexual Maturity Planorbarius corneus laid egg capsules at temperatures of 15, 20 and 25°С, whereas Р. planorbis reproduced from 10°С (Table 1). The higher the temperature, the faster the maturation was: 15th week at 25°С and 49th week at 15°С for P. corneus. In P. corneus, the minimum, maximum and mean diameters were similar at 20°С and 25°C. At 15°C these values were higher and, on average, snails began their reproduction when they reached 13 mm. For P. planorbis, the mean diameter ranged from 5.4 to 6.9 mm. Reproduction-Period Duration In the laboratory, egg laying occurs throughout the year though this is not true for field populations. For both species consid- ered together, the reproduction-period dura- tion was negatively correlated with tempera- ture (Kendall’s correlation: М = 7, t= —0.617, р = 0.05), and positively correlated with тах- imum longevity (т = 0.905, р = 0.04). At 15°C, P. corneus reproduced for 175 weeks, corre- sponding to 75.8% of its maximum longevity, whereas P. planorbis laid eggs for 125 weeks, corresponding to 87.4% of its life. Variation of the Number of Egg Capsules and Eggs Temperature strongly influenced the repro- duction of the two species, especially egg laying in P. corneus. At 20°С, a peak was observed from the 41st week to the 75th week, and the maximum production of eggs reached 8.5 capsules per snail per two weeks (Fig. 1). At 25°C, the snails produced a con- stant number of capsules (between 0.5 and 1/snail/2 weeks), and at 15°C, the maximum value was 3.2. At the latter temperature, an egg laying rhythm of 44 weeks was observed (periodogram method: Fourier's analysis, p < 0.05). Great variations of the number of eggs per capsule were noticed in P. corneus. At the two highest temperatures, the capsules had few eggs at the beginning of the repro- duction period. The numbers of eggs per mass then increased to 21.5 (20°С, 131st week) or 17.8 (25°C, 57th week). The maxi- mum number for snails reared at 15°C reached 15.4. There were significant Ken- dall's correlations between the number of laid capsules and the number of eggs per capsule at 20°C (positive correlation: t = 0.425, p < 0.001) and at 25°C (negative cor- relation: t = —0.304, p < 0.03). In P. planorbis, capsule production varied from one week to the next irrespective of temperature, with maximum values of 6.8 (10°C), 7.4 (25°C), 8.6 (15°C) and 9.7 (20°C) (Fig. 2). Nevertheless, rhythms of 56 weeks (10°C) and 58 weeks (15°C) were found by 81 REPRODUCTION IN PLANORBIDS 91 +06 8, + 98 61 + 68 0 8 + 7б ($ + %) eyes Bulyojey иеэи\| Le 97 9€ Oc 8? 0'S +98 1'9+6'8L Se+ 64! ДЕС А КАНО | ¿"EL 89r 60€ 98 LE 6c€ 9987 $985 0911 (A Ob Sé Le ol 50 85 OS! 897 LOL Zt GLO Gel Р 28 L'ZZ L'8r GE 58 ep! GLI ebb ve 19 Sc! gel GS 22 v'8 0'6 LZ O'eL ee ¿'S Эф 6d BL 6077S 90+8'9 60+69 80+65 cL+SOL 6 LE SL Sc SL 9095 9.08 991 9:01 993 SIQIOUEIA SIQLIOUEIT (цоцелер piepue]s + x) SNOUIOD зпиедиоив Ol +t6 LL сб Ly Le 87+ srl 67 +08 6'с1 Le $181 7Р9 60 go eck 18 S'69 891 €0c LEC LVL 941 gel 991 gL 66 80+1`01 РЕ 8 С | ZL 6r 903 951 aınsdes 66e led sBBa jo saquinu шпиихеи| S + (9/8N) ajnsdeo 66e Jod s6Ge jo saquinu ueayy Уээлм UONONPOAda Jad |leus sed pre; s66s jo лэашим (294) ueds 9}! ayy зпочбполц} ¡reus sad pre s6Ba jo лэашим 492M иоцопро.аэл Jad ¡reus sad pre] sainsdeo 66e jo saquuiny ueds aji ayi зпочбполцз ¡reus Jad pre, sainsdeo 66e jo лэашпм (syaem) (21) ‘AneBuo] шпцихеш 0} чоце|эл и! чоцелпр ромэа uoronpoidey ($\ээм) Анлэбио|! шпилхеи\ (syaem) (71) иоцелтр ромеа uoljonpoide# (ww) (*g) эзэшер unuIxen (ww) (a) лэзэщер шпииии S (Па) Ayınyew |епхе$ jo jesuo je лэзэшер ивэи| (syaam) (y) Ayunyew penxes jo jesuo ye эбе иеэи\ alnjeladwa | 9/mie18dula] о} иоцеел и! S/QUOUB/A зиоиеа pue зпешоэ snuequoue/g yO uolonpoidei ay] jo Seinj2e} шеи р JIGWL 82 COSTIL & DAGUZAN Nc Ne» 15°C Ne/c 10 9 8 7 6 " à Ñ \ Au ВИ xy, a 4 1 iM a 1 \/ Y ий "Val 3 | y ' > o — O 20 40 60 80 100 120 140 160 180 200 220 24 Age (weeks) Nc 20°C 0 2027407556077 804100» 71207 414072160 Age (weeks) Nc 25°C Ne/c S — NW Un Qs ~ 00 Age (weeks) FIG. 1. Reproduction of Planorbarius corneus in relation to temperature and age: variation in the number of egg capsules per individual per 2 weeks (Nc); variation in the number of eggs per capsule (Ne/c). periodogram analysis (p < 0.05). Between 15°С and 25°С, the number of eggs per сар- sule, low in the first weeks of the reproduc- tion, increased for 12 to 18 weeks before sta- bilizing. At the end of the reproductive period, the number of the eggs per capsule of P. planorbis reared at 10°С cannot be well es- tablished because of small number of laid capsules. The number of eggs per mass pro- duced by the snails at 15°C is positively cor- related with the number of laid masses (т = 0.348, p < 0.001). Nc Nc 10°C Ne/c 10 9 8 7 6 5 E 8 > 6 2 4 1 2 0 0 0 20 40 60 80 100 120 140 160 Age (weeks) Nc Nc 15°С Ne/c OA AS Ne/c 24 22 - 20 2 \ 18 d Zn Ales ANAT AE 16 6 / TAN YF 14 1 4 ' 5 1 1 12 4 / if 10 y 1 | Es 3 6 = 4 1 2 0 0 0 20 40 60 80 100 120 140 Age (weeks) 0 20 40 60 80 Age (weeks) Nc 25°C Ne/c © — DO BR LU A Jo 0 5 10 15 20 25 30 35 Age (weeks) FIG. 2. Reproduction of Planorbis planorbis in re- lation to temperature and age: variation in the num- ber of egg capsules per individual per 2 weeks (Nc); variation in the number of eggs per capsule (Ме/с). REPRODUCTION IN PLANORBIDS 83 0 20 40 60 30 100 120 140 160 Age (weeks) 0.20 40 60 80 100 120 140 160 Age (weeks) FIG. 3. Reproductive “abnormalities” of Planorbar- ¡us corneus reared at 20°C and 25°C: variation in the number of capsules including one or more eggs without egg cells (C,,); variation in the number of eggs without egg cells (E,,). “Abnormalities”” of Egg Capsules and Eggs Capsules were considered abnormal when they contained no egg, or when all or part of the eggs were without egg cells, ¡.e. zygotes. Moreover, the capsules with eggs containing more than one egg cell were also considered abnormal. Throughout the reproductive pe- riod of P. corneus, the numbers of capsules without eggs at 20°С and 25°С were respec- tively 125 (corresponding to 0.14%) and 17 (corresponding to 0.20%). The greatest per- centages of capsules containing eggs with- out egg cells occurred at the beginning of reproduction (Fig. 3). The mean percentages of eggs without egg cells at 25°С and 20°С were respectively 7.1% (standard deviation, $ = 7.5) and 5.5% ($ = 6.8). For P. corneus reared at 15°С (age: 49-81 weeks), this value was 3.9% (s = 3.3) (Fig. 4), and 21 eggs with more than one egg cell occurred as twin eggs (16), three egg cells (3), four egg cells (1) and nine egg cells (1). At 20°C, the total number of eggs with multiple egg cells was twice what it was at 25°С (Table 2). Twins were Cw (%) Ew (%) ] | Cw 1 404 |------- Ew [40 | | | 307 к 30 ] : = | 207 \ N p20 1 TT /- | / F 10 ] y AA NE 0 у—— A A (IRt[T[Tro>_nAA —— |) 45 20 3 60 65 =0 1 30 85 Age (weeks) FIG. 4. Reproductive ““abnormalities” of Planorbar- ¡us corneus reared at 15°C (from 49 to 81 weeks old): the number of capsules containing one or more eggs without egg cells (C,,); the number of eggs without egg cells (E,,). particularly numerous but rarely developed further. Whatever the species, no eclosion occurred in eggs containing more than two egg cells. Table 3 summarizes the results concerning P. planorbis. The percentage of capsules without eggs increased with time. Except at 10°C, the same phenomenon was observed for eggs without egg cells, which were espe- cially numerous at 25°C. Two capsules with- out proper eggs but containing 16 and 18 egg cells were laid by snails reared at 10°C. A significantly greater number of eggs with several egg cells was laid at 10°C compared with higher temperatures (H = 5.326, p = 0.15; Kruskall & Wallis test). Nevertheless, even at 10°C, these eggs were not often no- ticed. Dwarf or giant eggs were encountered particularly at 10°C. The number of egg cells per egg never exceeded four, and twin egg cells were the most numerous. Fecundity and Fertility Planorbarius corneus was most fecund at 20°C (Table 1). The 6,031 capsules laid by all the snails contained on average 14.8 eggs, whereas this number was 8.0 for clutches produced at 15°C. In P. planorbis, the high- est number of eggs per capsule was also ob- served at 20°C (18.9 + 6.1). The temperature of 15°C may have allowed the snails to lay the maximum capsule number due to a lengthening of the life span and reproduction period. Nevertheless, fecundity related to re- production week was maximum at 20°C (2.5 capsules and 46.8 eggs/individual/reproduc- tion week). The temperatures of 25°C and 10°C appeared to be especially unfavorable for the reproduction of P. planorbis. 84 COSTIL 8 DAGUZAN TABLE 2. “Reproductive abnormalities’ in Planorbarius corneus reared at 20°C and 25°C: number of egg cells per egg. 20°C Number of eggs containing such Number of egg a number of Percentage of eggs containing such a number 25°C Number of eggs containing such a number of Percentage of eggs containing such a number cells per egg egg cells of egg cells egg cells of egg cells 2 204 0.226 9 0.106 3 30 0.033 1 0.012 4 8 0.009 1 0.012 5 5 0.006 0 0 9 2 0.002 0 0 EMO Mal Sree 15 1 0.001 0 0 Total 253 0.280 1 0.130 The dynamics of pluriannual experimental populations was investigated using Leslie matrices (Table 4). At every temperature, the greatest reproductive effort was made by the penultimate age class. The total fertility of P. planorbis reared at 25°С was low (23 newly hatched snails per individual). The greatest fertility of P. corneus was at 20°С (1,600 т- dividuals/snail), whereas P. planorbis pro- duced the maximum number of newly hatched snails (3,357/adult) at 15°С. The eggs of P. corneus developed at 10°С, but this species did not reproduce at this tem- perature. DISCUSSION In contrast to P. corneus, P. planorbis de- veloped but did not lay eggs at 10°С. How- ever, for the embryonic development study, the eggs placed in incubation were laid at 20°С. It was possible that the temperature at which the eggs had been elaborated was im- portant for the future development. Eggs of P. planorbis laid at 10°C could develop if they were then incubated at this temperature. So, the hatching rate could be slightly lower than the rate of P. corneus (35%), as it was the case for the other temperatures. The number of newly hatched snails could be 404. The minimum threshold for the reproduc- tion of temperate freshwater snails appears to be between 7°C and 12°C (Vaughn, 1953; Boerger, 1975; Duncan, 1975; Eversole, 1978; Krkac, 1982), and 12°C was also stated for P. corneus by Precht (1936). Ac- cording to Joosse & Veld (1972), ovogenesis of Lymnaea stagnalis (L.) did not depend on temperature, and ovocytes of all stages were found in the ovotestis of infertile snails reared at 5°C or 8°C, although spermatogenesis stopped at these low temperatures. Eggs were produced by adults of Lymnaea obrussa Say at temperatures ranging from 10°C to about 26°C (Mattice, 1975). Krkac (1982) explained that egg laying of Physa acuta (Draparnaud) was stimulated by every temperature increase up to 30°C. At a tem- perature of 25°C, the maximum threshold for the reproduction of both studied species was not attained, and it could be slighly below 30°C. In the tropical species, Biomphalaria glabrata (Say), it reached 33°C (Vianey-Liaud, 1982). The temperature effect on animal repro- duction is both direct (existence of thresh- olds) and indirect (general development, growth). There is an acceleration of repro- ductive system maturation with temperature, a relationship not observed in some other freshwater pulmonates, including Bulinus truncatus (Audouin), Biomphalaria alexand- rina (Ehrenberg) and Helisoma duryi (Weth- erby) (El Eman & Madsen, 1982). Moreover, these snails laid their first egg masses when they attained a certain size but at very differ- ent ages (three weeks for the earlier Bulinus and five weeks for the earlier planorbids). In comparison with the two studied species, other species appear to reach sexual matu- rity earlier: P. acuta, five weeks (Perrin, 1986), Lymnaea truncatula (Múller), four weeks (Ho- dasi, 1976). Onset of sexual maturity de- pends on many factors. For example, in the laboratory, Lymnaea peregra (Müller) from exposed habitats initiated reproduction ear- lier and put more effort into it than snails from sheltered habitats (Calow, 1981). Basommatophoran snails often reproduce REPRODUCTION IN PLANORBIDS 85 TABLE 3. Reproductive “abnormalities’’ of Planorbis planorbis in relation to age and temperature. Proportion of capsules Age without eggs Temperature (weeks) (%) 29 0 31 0 33 0 35 0 37 0 51 0 10°C 59 0.9 77 3.4 105 8.9 113 17.2 129 19.1 141 20.7 153 30.8 31 0 49 0 57 0.5 73 0.3 85 0.6 93 157 15°С 117 122 123 125 127 52 131 33:3 133 DIES 135 33:3 21 0.2 35 0.4 51 4.6 20°C 53 2.9 59 8.5 61 11.9 63 alıka 65 8.8 15 0 25 1.5 25°C 27 0 29 0 31 0 33 0 Proportion of capsules including 1 or more eggs without egg cells Proportion of eggs including 2 or more egg Proportion of eggs without (%) egg cells (%) cells (%) 20:5 8.1 0.43 5.0 2.1 0.70 4.6 2.1 0 0 0 0.75 0 0 0.30 1.3 0.5 0.25 1.0 0.6 0 3.6 0.5 2.30 0 0 0.24 0 0 0.26 3.9 1.7 0 0 0 0 0 0 0 0.7 0.3 0 0.9 0.5 0 0:5 0.2 0.04 0.6 0.3 0 1.1 0.6 0.05 0.4 0.06 0.03 0.4 0.02 0.02 122 0.3 0.04 4.7 1.4 0 20.0 3.9 0 30.0 8.3 0 33.0 8.5 0 1.1 0.3 0.02 0.4 0.1 0.02 5.0 1:3 0 6.5 1.4 0 lat 2.9 0.04 8.6 3.0 0 19.3 5.6 0.11 25.8 9.4 0 1122 4.5 0 27; 128 0.10 39.4 21.4 0 90.6 652 0.60 95.7 93.0 0 100 100 0 before reaching half adult size, and growth then continues (Larambergue, 1939). Consid- ering the maximum size when the planorbids produce their first egg capsules, we con- clude that P. planorbis and P. corneus lay their first eggs at sizes of 7.1 mm and 12.5 mm respectively, and these sizes correspond to 0.5 or 0.4 times the observed maximum sizes. As for many freshwater pulmonates, plan- orbid reproduction in experimental condi- tions is continuous throughout the year. In P. planorbis and P. corneus, the mean durations of the reproduction period were 74.9% and 64.7% of the maximum longevity respec- tively, compared with 63.6% in H. duryi (Aboul-Ela & Beddiny, 1980). For P. planor- bis, the last capsule was laid shortly before death, whereas in P. corneus reared at 20°C and 25°C, reproduction stopped 40 weeks before. The latter snails might suffer from a gamete exhaustion and/or damage to the re- productive system. Although few malacologists have reported 86 COSTIL 8 DAGUZAN TABLE 4. Leslie matrices constructed for experimental populations of Planorbarius corneus and Planorbis planorbis reared at different temperatures. Fertility at the successive age classes (from birth up to 4 years) is on the horizontal axis; survivorship is on the diagonal. The numbers of the top vertical row represent the total fertility (including the new hatched snails produced at the end of the life, during the incomplete year). See “discussion” for the numbers in brackets stated for P. planorbis at 10°С. о SQ a o 15°С 636 0 0.71 0 128 0 0.06 0 (107) 0 0.63 0 1244 0 0.69 1760 0 Planorbarius corneus 20°С У © 2555 SQ № 10°С EN ь Planorbis planorbis 15°С © y 20°С So (Me) Sesa less come ooect on a abnormalities of eggs or egg masses (Bloch, 1938, for P. corneus and L. stagnalis; Bond- esen, 1950, for Ancylus fluviatilis (Múller); Bi- gus, 1981, for Physa acuta), it is important to take such abnormalities into account if we do not want to overestimate the snail fertility. In our study, most of the eggs without an egg cell were laid at the beginning of the repro- ductive period, so the start-up of reproduc- tion seemed to present some problems. However, the occurrence of capsules without eggs in P. corneus and P. planorbis, and of the eggs without egg cells in P. planorbis, increases with time. These eggs and cap- sules reflected a lack or a dysfunction in the formation of gametes and eggs. For individ- uals of P. planorbis of the same age, the re- production problems were more numerous with higher temperatures. Vianey-Liaud (1982) showed that at 33°С the reversible sterilization of B. glabrata was not due to a problem of reproductive system differentia- tion, but to a disruption of the system func- tioning. By contrast, Michelson (1961) showed that in B. glabrata low temperatures reduced fecundity without obviously damag- ing the reproductive system, whereas high temperatures reduced the female sexual or- 94 243 153 0 0 0 0 0 0 555 0.63 0 0 0 0.11 0 890 74 0 0 1600 0 0 0.27 0 22 0 150 0 0 (153) 0 (140) 0 0 0 (404) 0 0 0.21 0 1059 0 3357 0 2389 gans. The results reported here confirm the former rather than the latter author. In Ancy- lus fluviatilis, the percentage of abnormal capsules, eggs and egg cells varied from 10% to 24% according to year (Bondesen, 1950). Moreover, dwarf eggs, associated with the starting or the stopping of spawning, could be explained by a failure in the func- tioning of the albumen gland. In the case of the two eggs of P. planorbis that included 16 and 18 egg cells but no egg, the albumen gland may have ceased its production. The greatest number of egg cells per egg reached four for P. planorbis and 15 for P. corneus. This number was two for A. fluviatilis (Bond- esen, 1950), six for Physa fontinalis (L.) (per- centage of twins. 0.1-0.4%, De Witt, 1955), and 47 for P. acuta (Bigus, 1981). For the two species studied here, twins rarely hatched. However, six of the 15 embryos contained in the egg of P. corneus developed to early tro- chophore stage and moved energetically be- fore dying. The great majority of the capsules and eggs are normal. To what extent is the num- ber of eggs per capsule a species feature? Does this number vary according to parent state (age, size) and environmental factors? REPRODUCTION IN PLANORBIDS 87 The planorbid family is characterized by the production of a smaller number of egg cells per egg in comparison with lymnaeids or physids. In relation to European planorbids, P. corneus has capsules rich in eggs. For ex- ample, Bloch (1938) observed 136 eggs in a single capsule. Here, maximum values of 48 egg cells per capsule in P. corneus and 46 in P. planorbis were seen. Moreover, in P. cor- neus, the mean number of eggs observed per mass was lower than the numbers found by other authors: 18.2 (Cole, 1925); 27.8 (N = 63) and 33.3 (N = 39) (Oldham, 1930); 71 (N = 28) (Alyakrinskaya, 1981); 15.6 or 33.3 according to respectively the small and the large race (Precht, 1936). In Bithynia tentaculata (L.), geographic provenance is the principal source of capsule-richness variation (Vincent & Gaucher, 1983). The number of eggs per capsule depends also on food (Van der Steen, 1967), dissolved oxygen in water (Aly- akrinskaya, 1981), and crowding (Chernin & Michelson, 1957). According to Oldham (1930), there does not appear to be a relationship between egg laying dates of P. corneus and the number of eggs per capsule. Here, however, the num- ber of eggs per mass was generally low at the starting of spawning, and then increased. Af- ter this initial period, egg richness for P. planorbis was relatively constant, but was more variable for P. corneus. In the latter, a reduction in reproduction activity had a sim- ilar effect on the richness of capsules laid at 20°C, and an opposite result at 25°C. A negative correlation in Lymnaea catascopium (Say) (Pinel-Alloul & Magnin, 1979) and in L. stagnalis (Mooij-Vogelaar et al., 1970), or no correlation in В. glabrata (Vianey-Liaud, 1990) were found between the reproduc- tive intensity and richness of masses in eggs. Unlike Vianey-Liaud (1990) for B. glabrata, but as Precht (1936) for P. corneus and Mad- sen et al. (1983) for H. duryi, Boag 8 Pear- stone (1979) suggested that the biggest indi- viduals of L. stagnalis laid the richest capsules. In our study, we did not notice a greater number of eggs per capsule over the growing period. In L. fontinalis, capsule rich- ness was not proportional to snail size but depended on egg-laying date (Duncan, 1959). From field caged experiments, tempo- ral measurements and dissections of females of the ovoviviparus prosobranch Viviparus georgianus (Lea), Buckley (1986) concluded: “Spat size is positively correlated with female age irrespective of female size, though brood numbers increase with maternal size and growth rates.” A great variation of the capsule and egg production occurred for both species. Snails that produced a lot of capsules for two weeks should reduce their production after that, and we cannot envisage a lack of foreign sperm, because it could be stored in the sperma- theca. A period of intensive reproduction was only observed for individuals of P. corneus reared at 20°C during the first half of their life, when their physiological state allowed them to reproduce. Physa acuta at 20°C showed a similar reproductive pattern (Per- rin, 1986). Egg laying rhythms of 44 weeks (P. corneus at 15°C) and 56-58 weeks (P. plan- orbis at 10°C and 15°C) were observed using periodogram analysis. The rhythm of 44 weeks could be a multiple of a shorter rhythm (about 20 weeks according to Figure 1), and no obvious rhythm was found by the correl- ogram method (autocorrelation of time-se- ries). Further experiments should be per- formed to confirm this rhythm and also the rhythms concerning P. planorbis. Moreover, it would be interesting to know if such rhythms are endogenous or not. It is difficult to compare fecundities in dif- ferent species because the reproductive function is strongly affected by experimental conditions, which are variable depending on authors. Nevertheless, some results are available: B. glabrata: 9,000 eggs and 7,000 newly hatched per year (Vianey-Liaud, 1990); L. stagnalis: between 4,655 and 10,832 eggs depending on light conditions and during a reproduction period of 7-13 months (Seuge & Bluzat, 1983); H. duryi during its whole life: between 145 (60 snails/2 water liters; repro- duction period of 47 weeks) and 7,245 eggs (5 snails/2 |; reproduction period of 99 weeks) (Aboul-Ela & Beddiny, 1980). These three species are more fecund than the two stud- ied planorbids: P. corneus: 1,600 newly hatched (20°C; 141 weeks), P. planorbis: 3,357 newly hatched (15°C; 125 weeks). Moreover, in both planorbids, the newly hatched snail production, which is continu- ous, increases with time until the penultimate year of life. In iteroparous mollusc species, reproduction effort increases with successive breeding seasons (Browne & Russell-Hunter, 1978). In the field, the fecundity of Helisoma trivolvis (Say) has been estimated at 1,962 eggs per adult snail during the spring breed- ing period, and at 1,263 eggs per snail in au- tumn (Eversole, 1978). Considering the tem- 88 COSTIL 8 DAGUZAN perature for the minimum reproduction threshold (10°C for P. planorbis, 15°C for P. corneus) and the maximum numbers of eggs and newly hatched young produced, P. plan- orbis appears to require colder conditions than P. corneus. Such a result is also found for the optimum growth of these two species (Costil, 1994). From studies performed in seven species of aquatic snails, van der Schalie 8 Berry (1973) deduced that the lym- naeids reproduced and thrived best in cool (19°C to 22°C) conditions, whereas the plan- orbids required warmer water (22°C to 25°C), and the physids were highly tolerant, being able to maintain themselves in a much wider temperature range (12°C to 30°C). Our ex- perimental approach is completed by studies performed in field (ecology, life cycle) (Costil, 1993), and all this research should allow us to get in the future a thoroughly knowledge of the biology of these two species. ACKNOWLEDGMENTS We are grateful to Dr. G. Dussart (Univer- sity of Canterbury) for helpful comments on the present paper and for linguistic help. We also thank M. Foulon for technical assis- tance. LITERATURE CITED ABOUL-ELA, |. A. & E. А. М. BEDDINY, 1980, Fac- tors affecting the reproductive capacity and population dynamics of Helisoma duryi (Weth- erby). |: effect of temperature. Journal of the Egyptian Society of Parasitology, 10: 135-150. ALYAKRINSKAYA, |. O., 1981, Egg nutrient con- tent in gastropods. Akademiia Nauk SSSR, Dok- lady, Biological Sciences, 260: 472-474. BIGUS, L., 1981, Polyvitellinitát und Keimver- schmelzungen bei Physa асиа (Pulmonata, Basommatophora). Zoologische Jahrbücher Abteilung für Anatomie und Ontogenie der Tiere, 105: 526-550. BLOCH, S., 1938, Beitrag zur Kenntnis der Onto- genese von Süsswasserpulmonaten mit beson- derer Berücksichtigung der Mitteldarmdrüse. Revue Suisse de Zoologie, 45: 157. BOAG, D. A. & S. M. PEARSTONE, 1979, On the life cycle of Lymnaea stagnalis in southwestern Alberta. Canadian Journal of Zoology, 57: 353- 362. BOERGER, H., 1975, A comparison of the life-cy- cle, reproductive ecologies and size weight re- lationships of Helisoma anceps, H. campanula- tum and H. trivolvis (Gastropoda, Planorbidae). Canadian Journal of Zoology, 53: 1812-1824. BONDESEN, P., 1950, A comparative morpholog- ical-biological analysis of the egg-capsules of freshwater pulmonate gastropods, Hygrophila, Basommatophora, Pulmonata. Natura Jutland- andica, 3: 1-208. BROWNE, В. & W. RUSSELL-HUNTER, 1978, Re- productive effort in Molluscs. Oecología, 37: 23- 274 BUCKLEY, D., 1986, Bioenergetics of age-related versus size-related reproductive tactics in fe- male Viviparus georgianus. Biological Journal of the Linnean Society, 27: 293-309. CALOW, P., 1981, Adaptational aspects of growth and reproduction in Lymnaea peregra from ex- posed and sheltered aquatic habitats. Malacolo- gia, 21: 5-13. CHERNIN, E. & E. MICHELSON, 1957, Studies on the biological control of schistosome-bearing snails. Ш Further observations on the effects of crowding on growth and fecundity in Australor- bis glabratus. American Journal of Hygiene, 65: 71-80. COLE, W. H., 1925, Egglaying in two species of Planorbis. American Naturalist, 59: 284-286. COSTIL, K., 1993, Contribution à l'écologie des mollusques gasteropodes dulcicoles armoric- ains; recherches bio-écologiques et écophysi- ologiques sur deux especes de pulmonés Plan- orbidae: Planorbarius corneus (L.) et Planorbis planorbis (L.). Ph. D. Thesis, University of Rennes |, 421 pp. COSTIL, K., 1994, Effects of temperature on sur- vivorship and growth of two freshwater gastro- pods, Planorbarius corneus and Planorbis plan- orbis. Journal of Molluscan Studies, in press. DE WITT, W. F., 1955, The life-cycle and some other biological details of the freshwater snail Physa fontinalis (L.). Basteria, 19: 35-73. DUNCAN, С. J., 1959, The life cycle and ecology of the freshwater snail Physa fontinalis (L.). Journal of Animal Ecology, 28: 97-117. DUNCAN, C. J., 1975, Reproduction in pulmonates. рр. 309-365, in: V. FRETTER 4 J. PEAKE, eds., Pul- monates, Vol. 1, Academic Press, New York. EL EMAN, М. A. 8 H. MADSEN, 1982, The effect of temperature, darkness, starvation and various food types on growth, survival and reproduction of Helisoma duryi, Biomphalaria alexandrina and Bulinus truncatus (Gastropoda: Planorbidae). Hydrobiologia, 88: 265-275. EVERSOLE, A., 1978, Life-cycles, growth and pop- ulation bioenergetics in the snail Helisoma triv- olvis (Say). Journal of Molluscan Studies, 44: 209-222. GERAERTS, W. P. M. 8 J. JOOSSE, 1984, Fresh- water snails (Basommatophora) pp. 142-208, A. S. TOMPA, ed, in: The Mollusca, Vol. 7: Repro- duction, Academic Press, New-York. HODASI, J. К. M., 1976, The effects of low tem- perature on Lymnaea truncatula. Zeitschrift fúr Parasitenkunde, 48: 281-286. REPRODUCTION IN PLANORBIDS 89 JOOSSE, J. & С. J. VELD, 1972, Endocrinology of reproduction in the hermaphrodite gastropod Lymnaea stagnalis. General and Comparative Endocrinolology, 18: 599-600. KRKAC, N., 1982, Effect of sudden temperature change on behaviour, oviposition and mortality in Physa acuta Draparnaud (Gastropoda: Pulmo- nata). Malacologia, 22: 167-170. LARAMBERGUE, de M., 1939, Etude de l’autofe- condation chez les gastéropodes pulmonés; re- cherches sur l'aphallie et la fécondation chez Bulinus (Isodora) contortus (Michaud). Bulletin Biologique de la France et de la Belgique, 73: 19-231. LESLIE, P. H., 1945, On the use of matrices in population mathematics. Biometrika 33: 183- 212. MADSEN, H., W. Е. THIONGO & J. H. OUMA, 1983, Egg-laying and growth in Helisoma duryi (Wetherby) (Pulmonata: Planordidae): effect of population density and mode of fertilization. Hy- drobiologia, 106: 185-191. MATTICE, J., 1975, Effect of constant and varying temperature on egg production of Lymnaea obrussa Say (Mollusca: Gastropoda). Verhand- lungen Internationale Vereinigung für Theori- tische und Angewandte Limnologie, 19: 3174- 3178. MICHELSON, E. M., 1961, The effects of temper- ature on growth and reproduction of Australorbis glabratus in the laboratory. American Journal of Hygiene, 73: 66-74. MOOIJ-VOGELAAR, J. W., J. С. JAGER & W. J. VAN DER STEEN, 1970, Effects of density levels and changes in density levels on reproduction, feeding and growth in the pond snail Lymnaea stagnalis (L.). Proceedings of the Koninklijke Nederlandse Akademie van Wetenschappen, C, 76: 61-69. OLDHAM, C., 1930, Fecundity of Planorbis cor- neus. Naturalist, 801: 177-178. PERRIN, N., 1986, Les paramètres du cycle vital de Physa acuta (Gastropoda, Mollusca) en milieu expérimental. Revue Suisse de Zoologie 93: 725-736. PINEL-ALLOUL, В. & E. MAGNIN, 1979, Cycle de développement, croissance et fécondité de cinq populations de Lymnaea catascopium catascopium (Lymnaeidae) au lac Saint-Louis, Québec. Malacologia, 19: 87-101. PRECHT, H., 1936, Zur Kopulation und Eiablage einer Planorbiden. Zoologischer Anzeiger, 115: 80-89. SEUGE, J. 4 R. BLUZAT, 1983, Effets des condi- tions d'éclairement sur le potentiel reproducteur de Lymnaea stagnalis (gastéropode, pulmoné). Malacologia, 23: 321-331. VAN DER SCHALIE, H. 8 E. BERRY, 1973, The effects of temperature on growth and reproduc- tion of aquatic snails. Sterkania, 50: 1-92. VAN DER STEEN, W. J., 1967, The influence of environmental factors on the oviposition of Lym- naea stagnalis (L.) under laboratory conditions. Archives Neerlandaises de Zoologie, 17: 403- 468. VAUGHN, С. M., 1953, Effects of temperature оп hatching and growth of Lymnaea stagnalis apressa. American Midland Naturalist, 49: 214- 228: VIANEY-LIAUD, M., 1982, Effets des hautes tem- pératures sur la reproduction de Biomphalaria glabrata. Malacologia, 22: 159-165. VIANEY-LIAUD, M., 1990, Biologie de la reproduc- tion de Biomphalaria glabrata (Say, 1818) (Mol- lusque, Gasteropode, Planorbidae). These d'é- tat, University of Montpellier Il, 185 pp. VINCENT, B. & M. GAUCHER, 1983, Variations de la fecondite et de la structure des populations chez Bithynia tentaculata L. (Gastropoda: Proso- branchia). Canadian Journal of Zoology, 61: 2417-2423. Revised Ms. accepted 8 December 1993 MALACOLOGIA, 1995, 36(1-2): 91-95 COLOUR POLYMORPHISM IN THE MANGROVE SNAIL LITTORARIA INTERMEDIA IN SINAI L. М. Cook 4 J. Bridle School of Biological Sciences, University of Manchester, Manchester M13 9PL United Kingdom ABSTRACT Samples of Littoraria intermedia Philippi (Gastropoda: Littorinidae) have been examined from an isolated mangrove at the southern end of the Sinai Peninsula, Egypt. The snails are abun- dant оп the lower trunks and pneumatophores of Avicennia marina. Littoraria intermedia 1$ usually monomorphic in shell colour, but here it is polymorphic. There is an orange morph, similar to the orange morph of leaf-living Littoraria species, at a frequency of about 8%, and there may also be distinct morphs in the dark class. The trunk substratum also shows distinct patches of brown and black colour. A substantial amount of predation by crabs is inferred from the observed frequency of shell breakage. Polymorphism may therefore occur in species of Littoraria in populations subject to predation. It is not, as has been suggested, restricted to situations where predation may be assumed to be negligible. Key words: Littorinidae, predation, polymorphism, Littoraria. INTRODUCTION The genus Littoraria (Gastropoda: Littorin- idae) consists of some 30 species associated with mangrove trees. Some of these live on bark surfaces and are usually monomorphic for shell colour; others live on the leaves of the trees and commonly exhibit shell-colour polymorphism (Reid, 1986). This association is sufficiently general for it to be interesting to examine exceptional cases. Littoraria inter- media is a very widespread species, extend- ing from the east coast of Africa to the Pacific Ocean as far east as Hawaii, the Society Is- lands, and Samoa. Typically, it lives on roots and trunks of trees of the genera Rhizophora and Avicennia. Reid (1986) described the co- lour as “variable over the entire geographical range, but usually rather constant in each lo- cality. Ground colour usually grey, some- times pale brown, cream, whitish or rarely or- ange pink.” He also noted (1986) that orange-pink shells are to be found in samples from Arnhem-land, Northern Territory, Aus- tralia, and (personal communication) from the Red Sea and Aqaba. Observations on the north coast of Papua New Guinea, in Thai- land (Andaman Sea), and in Kenya indicated that the shells are uniform and very similar in colour in all these areas (Cook, 1986a, b; Cook 4 Garbett, 1992). On the other hand, Professor J. Heller (personal communication) suggested that the snails living on mangrove 91 trees at a site on the Sinai Peninsula were polymorphic, in the sense of having clearly discontinuous phenotypes. From their local- ity, these should be L. intermedia (Reid, 1986), and we therefore decided to examine this colony in more detail. THE MANGROVE SITE The mangrove consists of stands of Avi- cennia marina (Forskäl) on the coastal fan of the dry Wadi Kid. This is at the northern end of the Strait of Tiran, opposite the southern end of the Arabian side of the Gulf of Agaba and a few km north of the village of Nabk. The site may be approached by traveling north along the coast from the airport that serves Sharm el Sheikh and the Na'ama Bay resort. The site is described in detail by Por et al. (1977). It consists of a series of lagoons and strands landward of a shallow fossil coral shelf, which is only covered by a few cm of water. The trees are short and have thick trunks, frequently damaged, with red-brown bark and dark brown to black areas where the wood has been exposed. Groves of pneumatophores extend from the trees to seaward and into the lagoons. Samples of snails were collected from pneumatophores and from trunks just above the water level at three locations: to the north, at the centre, and to the south of the curved end of Wadi 92 COOK 8 BRIDLE TABLE 1. Frequencies (%) of different colours in samples of shells of Littoraria intermedia from Wadi Kid, Sinai Peninsula. The standard error is for arcsin transformed frequencies. Sample and Sample location size 1. North, trunk 232 2. North, trunk 297 3. North, pneumatophore 318 4. Central, pneumatophore 87 5. South, trunk Walz 6. South, trunk 83 7. South, pneumatophore 74 Standard error Kid. No snails were found on the leaves of the trees, and none of this species on rocks or stones. COLOUR VARIATION AND FREQUENCY Read (1986) records Littoraria intermedia from the Gulf of Aqaba as well as throughout the Red Sea. On the basis of shell shape and penis morphology, the mangrove littorinids collected at Wadi Kid belonged to this spe- cies. Three samples were taken from the north of the sequence, one from the centre, and three from the south. Three of these came from pneumatophores, and four from trunk surfaces (Table 1). This normally uni- form species showed a considerable range of shell colour. As in other members of the ge- nus, there appears to be a polymorphism for presence or absence of shell pigmentation and less clear-cut variation in the pattern of pigmentation, when present. In Table 1, four phenotypic categories are distinguished. The first column shows the frequency of shells with orange pigmentation. Most of these are bandless, but some have one or two grey- brown bands running along the whorls. The other three categories have the grey-brown pigmentation covering more or less the whole shell. These three groups vary in inten- sity of pigmentation from dark grey to a pale yellowish grey. Pigmentation is interrupted periodically as it is laid down during forma- tion of the shell, so as to produce transverse pale and dark striping. The overall colour has two components, comprising the pigmenta- tion and, where visible, the ground colour of the shell, which is yellowish. The colours of dry shells have been compared with stan- Grey/brown Orange dark mid light 1.5 30.1 57-3 5.0 7.8 24.5 61.6 6.1 6.4 16.8 74.2 РТ 7.9 7.0 58.4 25.8 UP 49.5 37.6 5:5 9.3 38.4 33.7 18.6 8.2 14.9 56.8 20.3 0.36 3.69 312 3.14 dards т a colour handbook (Когпегир & Wanscher, 1967). This has a range of hues modified by tone and intensity, cross-refer- enced to names given to them in common English usage. The modal value for the cat- egory called orange here is between greyish- orange and greyish-red. The banding pig- mentation is olive, and the yellow colour showing on banded shells is referred to as light yellow. These categories show different amounts of variation across the samples, the fre- quency of orange being almost invariant whereas the grey-brown types vary between samples. For the comparison of orange and non-orange between samples x° = 1.4 with 6 degrees of freedom (P > 0.9). For the other three categories, х? = 165.7 with 12 degrees of freedom, which is highly significant. An- other way of expressing this difference is to calculate the standard errors of the frequen- cies. When this was done using the arcsin transformed frequencies, the standard error for the grey-brown categories was found to be about ten times that of the orange cate- gory (Table 1). The implication is either that the environment affects the different morphs differently, or that it is not as easy to separate the grey-brown forms from each other as it is to distinguish them from orange. Variation in the grey-brown forms is not associated with position north to south or with bark or pneu- matophore substratum. In the species that live on foliage and are generally agreed to be polymorphic, a similar range of colour is found. There are yellow and orange forms, usually without bands but oc- casionally with narrow bands, and a category (dark), which is heavily banded. The orange forms in L. pallescens, L. filosa, L. philippi- MANGROVE SNAIL POLYMORPHISM 93 ana, L. luteola and L. albicans, illustrated by Reid (1986) in a colour plate, are the same colour as the orange class described here. The three categories are phenotypically dis- tinct, and there 1$ little difficulty in assigning shells to one or other of them, but variation in the banding patterns occurs. Thus, individu- als of L. pallescens from Papua New Guinea show little variation within each morph, whereas those from Thailand show more variation (illustrated by Cook & Garbett, 1989). There seems no doubt that genetically controlled morphs are involved that affect both ground colour of the shell and nature and extent of banding. For practical descrip- tive purposes, this allows us to distinguish three forms in L. pallescens: yellow, orange, and dark. However, there 1$ also variation in intensity of banding. Shells of the bark-living species L. intermedia and L. scabra are sometimes paler if the individuals are living on Avicennia than on the alternative preferred mangrove trees of the genus Rhizophora (Reid, 1986). Given this information from other species and the pattern of variation in frequency be- tween the samples of L. intermedia from Wadi Kid, we suggest that they are polymor- phic for at least two forms, an orange morph analogous to, and possibly homologous with, the orange form of other Littoraria species, and a “dark” category. This is different from orange, possibly genetically heterogeneous and probably also subject to environmentally induced variation in expression. EVIDENCE OF PREDATION Damage arising from crab predation is an important source of mortality in tropical lit- toral molluscs (Hamilton, 1976; Vermeij, 1978, 1982, 1992). Many species of crab at- tack the shells from the mouth, breaking out crescentic pieces of shell that are likely to reflect the shape of the chela of the attacker. If the prey is not killed, it is likely to carry an irregular scar on the shell that provides evi- dence of past attack. Reid (1992) carried out an extensive study of the effect of crab pre- dation on Littoraria species. Using exclusion arenas, he found direct evidence that the presence of predators is associated with the presence of scars on the shells of individuals that have survived attacks. High levels of damage occurred in L. intermedia and two other species living on bark but lower levels TABLE 2. Frequency of shells showing some breakage and repair in the samples, and mean shell height and shell strength estimated from sub-samples of 20 from each site. Strength 1$ estimated as log load (N) required to break the shell divided by log shell height. Damage Shell height Log load/ Sample (%) (mm) Log height 1 28.9 17.2 + 0.42 1.43 = 0.011 2 28.9 18.1 + 0.39 1.45 + 0.010 3 15.1 17.4 + 0.29 1.39 - 0.019 4 19.4 18.8 + 0.40 1.29 + 0.020 5 20.6 14.0 +0.43 1.38 + 0.018 6 12.3 15.7 +0.54 1.39 + 0.020 7 Tal 15.7 #0.35 1.28=0.014 in two leaf-living species. There was also a higher level on Rhizophora, where there were more crabs, than on Avicennia, where there were fewer. Damage to shells in mangrove snails is much more likely to be due to pre- dation than to accidental breakage (Vermeij, 1978), unlike intertidal species living in habi- tats where there is powerful wave action. Por et al. (1977) list 14 species of decapod Crus- tacea living in the area of the present study, some of which could be predators on the snails. Table 2 shows the frequencies of shells that show repaired cracks. There is variation between sites, but in some of the samples over a quarter of the shells are damaged. The heights of sub-samples of 20 shells from each site were measured to the nearest 0.1 mm with vernier callipers. The means and standard errors are given in Table 2. The load required just to break the shells was then es- tablished (measured in Newtons) using an In- stron 4301 table testing machine. The method is described in Cook & Kenyon (1993). Table 2 also shows the mean and standard error for log load divided by log height. The means vary between sites (F = 14.6, 4.1. 6 & 133, Р < 0.01), the strongest being in the two northerly sites (samples 1 4 2) and the weakest in site 7. Shell size is greater in the first four sites than in the last three. The samples from the first two sites have the highest incidence of predation, whereas sample 7 has the lowest. There may therefore be a correspondence between at- tack and robustness of shell, which varies between different parts of the Wadi Kid man- grove. 94 COOK & BRIDLE DISCUSSION Taken overall, there is a good association within the genus Littoraria of polymorphism with the foliage habitat and monomorphism with living on bark, on which the individuals are cryptic. There are two possible reasons for such an association. Either natural selec- tion, probably through the agency of apos- tatic predation, selects for polymorphism on the vivid and heterogeneous background of foliage, whereas predation favours crypsis on bark (the selective hypothesis); or living on leaves removes species from the attention of predators to such an extent that variant forms may accumulate through mutation (the neutral hypothesis). Rosewater (1970) ex- pressed the latter hypothesis as follows: “When snails leave the ground and ascend trees, they are immediately free of much of the danger from attacks by ground-living in- vertebrates and mammals which under ordi- nary conditions may select them for the fa- miliar subdued coloration usually evidenced by many exposed land, freshwater and ma- rine snails. lt may be theorized, therefore, that in L. scabra [in which he included the leaf-living species] colour variation is not un- der the control of selective forces usually ex- erted upon other species of Littorinidae and is, therefore, freely expressed in many of its populations.” It is not easy to decide he- tween these alternatives. There appears to be good evidence for predation by crabs (Reid, 1992), which is sometimes heavy but 1$ not necessarily selective. Reid (1987, 1992) has shown that apostatic selection, presum- ably through the agency of predation, may operate on leaf-living species. Some at- tempts to demonstrate selection have failed, however (Reid, 1987; Cook 8 Garbett, 1992), and a certain amount of selection does not in itself show that the polymorphism arises from selection. Provided the effective population is sufficiently large, a balance of mutation and accidental loss could be responsible (Cook, 1992). These species have a long- lived planktonic stage, so that large effective population size 1$ possible. One piece of evidence against the neutral hypothesis, although not a particularly strong one, is that different leaf-inhabiting species have similar morphs at similar frequencies. Darks are the most common (usually over 50%), yellows usually under 50% and orange morphs at 0-10% (Reid, 1986, 1987; Hughes & Jones, 1985; Cook, 1986a, 1992). Theoret- ically, this could come about if only a limited number of expressions of the genes con- cerned were possible and the morphs have the same relative mutation rates in the differ- ent species. However, the deterministic movement of gene frequency under mutation pressure is extremely slow. If we consider an allele at frequency q which mutates at rate u, and an alternative allele mutating at rate v, then the change in frequency per generation is Aq = v — q(u+v). Integration gives an ap- proximate expression for the number of gen- erations, n, required for a given change in frequency. This 15, п = In([q,(u+v) —v]/[q,(u+v) — v]}/(u+v) If we start with the allele at frequency v, then the number of generations to get to 63% of the equilibrium frequency v/(u+v) is 1/(u+v), and to get within 100f% takes —In(1—f)/(u+v) generations. Thus, the number of genera- tions required to get near to equilibrium is several times the reciprocal of the mutation rates, probably millions of generations. Dif- ferent species would be likely to exhibit dif- ferent frequencies by chance. The observations on the Sinai colonies constitute another, stronger, type of evi- dence. The argument suggests that bark-liv- ing species are usually monomorphic be- cause selective predation favours crypsis on a uniform grey-brown background. Some species are very difficult to see, for example L. strigata, which has a disruptive trans- versely striated pattern making it inconspic- uous on Avicennia trunks. Typically, L. inter- media is similar in colour and patterning to the background and does not stand out from it. No variant colours were seen, for example, among thousands of individuals examined on the north coast of Papua New Guinea. In the present instance, we have an isolated loca- tion in which the species is, exceptionally, polymorphic. The evidence indicates that it is also subject to predation and that robustness varies from sample to sample. The subjective impression is that the surface of the trees is broken into patches of bark and bare wood, which are much more distinct than usual and which have alternative grey and reddish co- lours like those of the morphs. It is possible to argue that, both here and in the case of foliage-living species, the heterogeneous background leads to selection for polymor- phism (Cook, 1986b). At present, this is speculation. What is demonstrated here, MANGROVE SNAIL POLYMORPHISM 95 however, is that polymorphism does not oc- cur only in the absence of predation, which is the neutralist conjecture. ACKNOWLEDGEMENTS We are grateful to Professor J. Heller for drawing our attention to this interesting col- ony, and to the Academic Study Group, Lon- don, for financial support. We thank Dr. D. G. Reid for comments on the typescript. LITERATURE CITED COOK, L. M., 1986a, Site selection in a polymor- phic mangrove snail. Biological Journal of the Linnean Society, 29: 101-113. COOK, (|. M., 1986b, Polymorphic snails on varied backgrounds. Biological Journal of the Linnean Society, 29: 89-99. COOK, L. M., 1992, The neutral assumption and maintenance of colour morph frequency in man- grove snails. Heredity, 69: 184-189. COOK, L. M. & S. D. GARBETT, 1989, Patterns of variation in mangrove littorinid molluscs on Phuket Island. Phuket Marine Research Bulletin, 53: 1-14. COOK, L. M. $. D. GARBETT, 1992, Selection in the polymorphic mangrove snail Littoraria pall- escens. Pp. 247-253, in J. GRAHAME, P. J. MILL & D. G. REID, eds., Proceedings of the 3rd International Symposium on Littorinid Biology. Malacological Society, London. COOK, L. М. & В. KENYON, 1993, Shell strength of colour morphs of the mangrove snail Littoraria pallescens. Journal of Molluscan Studies, 59: 29-34. HAMILTON, P. V., 1976, Predation on Littorina ir- rorata (Mollusca: Gastropoda) by Callinectes sapidus (Crustacea: Portunidae). Bulletin of Ma- rine Science, 26: 403-409. HUGHES, J. М. 8 М.Р. JONES, 1985, Shell colour polymorphism in a mangrove snail Littorina sp. (Prosobranchia: Littorinidae). Biological Journal of the Linnean Society, 25: 365-378. KORNERUP, A. 8 J. Н. WANSCHER, 1967, Meth- uen handbook of colour, 2nd ed. Methuen, Lon- don. РОВ, F. D., I. ООВ, & A. AMIR, 1977, The mangal of Sinai: limits of an ecosystem. Helgolánder Wissenschaftliche Meersuntersuchungen, 30: 295-314. REID, D. G., 1986, The littorinid molluscs of man- grove forests in the Indo-Pacific region. British Museum (Natural History), London. 288 pp. REID, D. G., 1987, Natural selection for apostasy and crypsis acting on the shell colour polymor- phism of a mangrove snail, Littoraria filosa (Sow- erby) (Gastropoda: Littorinidae). Biological Jour- nal of the Linnean Society, 30: 1-24. REID, D. G., 1992, Predation by crabs on Littoraria species (Littorinidae) in a Queensland mangrove forest. pp. 141-151, in J. GRAHAME, P. J. MILL 8 D. G. REID, eds., Proceedings of the 3rd Interna- tional Symposium on Littorinid Biology. Malaco- logical Society, London. ROSEWATER, J., 1970, The family Littorinidae in the Indo-Pacific. Part I. The subfamily Littorini- nae. Indo-Pacific Mollusca, 2: 417-506. VERMEIJ, С. J., 1978, Biogeography and adapta- tion: patterns of marine life. Harvard University Press, Cambridge, Massachusetts. xili+332 pp. VERMEIJ, С. J., 1982, Gastropod shell form, breakage and repair in relation to predation by the crab Calappa. Malacologia, 23: 1-12. VERMEIJ, С. J., 1992, Repaired breakage and shell thickness in gastropods of the genera Lit- torina and Nucella in the Aleutian Islands, Alaska. Pp. 135-139, in J. GRAHAME, P. J. MILL & D. G. REID, eds., Proceedings of the 3rd Interna- tional Symposium on Littorinid Biology. Malaco- logical Society, London. Revised Ms. accepted 1 March 1994 MALACOLOGIA, 1995, 36(1-2): 97-109 THE RELATIONSHIP BETWEEN SHELL-PATTERN FREQUENCY AND MICROHABITAT VARIATION IN THE INTERTIDAL PROSOBRANCH, CLITHON OUALANIENSIS (LESSON) Michael С. Gardner', Peter В. Mather”*, lan Williamson? & Jane М. Hughes? ABSTRACT A number of studies undertaken on the highly polymorphic intertidal mollusc Clithon ouala- niensis reported that colour-morph frequencies varied on a regional basis in the Indo-Pacific region (Gruneberg, 1976, 1978, 1979). Our study examined colour-morph variation on a local scale in the same species and demonstrated that a level of variation similar to the regional variation described by Gruneberg was present in Clithon populations collected from different microhabitats at a single locality in northeastern Queensland. An examination of genetic dif- ferentiation (using allozyme electrophoresis) of the same populations failed to identify an as- sociation between genotype and microhabitat and confirmed that Clithon populations at least on a local scale belong to a single gene pool. Factors that influence the distribution of morphs at particular sites are most likely to be such ecological factors as differential predation. The results of this study indicate that relationships between environmental variables on a regional scale and colour-morph frequencies in Clithon need to be reassessed and the extent of local variation studied intensively. INTRODUCTION Polymorphisms provide opportunities to investigate evolutionary events in natural populations (Gillespie & Tabashnik, 1990), and some understanding of selective pres- sures and the maintenance of variability is possible (Reimchen, 1979). Colour polymor- phisms have been investigated in many ani- mals, including insects (Brakefield, 1990), birds (Hughes, 1982), fish (Endler, 1988), and mammals (Kettlewell, 1973). In many mollusc species that show shell colour and pattern polymorphisms, a rela- tionship between shell colour/pattern and en- vironmental elements has frequently been suggested in the maintenance of the poly- morphisms (Etter, 1988, Nucella sp.; Green- wood, 1992, Cepaea sp.; Chang & Emlen, 1993, Cepaea sp.). Differential predation has been suggested as a factor maintaining poly- morphisms in many such species (Cook, 1983, Littorina sp.; review in Cook, 1986; Hughes & Mather, 1986, Littorina sp.; Reid, 1987, Littoraria sp.; review in Cook & Kenyon, 1991), although such other factors as tem- perature tolerance and area effects have also been reported (Jones et al., 1973). A particularly striking example of shell co- lour and pattern polymorphism in gastropods is provided by the intertidal prosobranch С/- thon oualaniensis (Gruneberg, 1976, 1978, 1979; Goodhart, 1987). The variation in C. oualaniensis is complex, with many shell co- lours and a large number of different banding patterns present (Gruneberg, 1976, 1978, 1979). Clithon are widely distributed in the Indo- Pacific region and are commonly found on muddy sand, stones or seagrass beds in the upper reaches of the tidal flats in sheltered localities often near mangroves, the inlets of lagoons or the mouths of rivulets (Gruneberg, 1976), where they feed non-selectively on deposits (Dye & Lasiak, 1987) and are sus- pected of burying in the substrate across the high tide. Sometimes populations can be very large, and hundreds of snails may be found per square metre of substratum. The sexes are separate, and it is unlikely that Cli- thon has a free-swimming pelagic larva (Gruneberg, 1976), although information on the presence or absence of a larval phase is sketchy despite breeding trials having been attempted (Gruneberg, 1976). In a number of studies, Gruneberg de- 'CSIRO Division of Fisheries, СРО Box 1538, Hobart, 7001 Australia “Centre for Biological Population Management, School of Life Science, Queensland University of Technology, Brisbane, 4001 Australia. “Faculty of Environmental Sciences, Griffith University, Nathan, 4111 Australia. “Person to whom correspondence should be addressed. 98 GARDNER ET AL. scribed shell colour and pattern polymor- phisms in C. oualaniensis over a large geo- graphic area of the Indo-Pacific (Gruneberg, 1976, 1978, 1979, 1982). Two regions were investigated: (a) populations from Ceylon and India (“western” Clithon); and (b) populations around Hong Kong and Malaysia (““eastern” Clithon). Gruneberg recognized snails from these two regions as being distinct “forms” of Clithon, found homogeneity with respect to the morph frequencies present within each of these regions, and suggested that this was equivalent to the “area effects” reported in the land snail Cepaea nemoralis by Cain & Currey (1963). However, Gruneberg generally collected only one sample from a single locality (beach), assuming it to be representative of the whole beach, and, although he recorded the substratum type, he did not consider the effects of within-locality habitat variability. Temporal variation in morph frequencies at the one site (other than a comparison be- tween juvenile and adult animals from the same collection) also was not investigated, although some studies of other polymorphic molluscs have shown morph frequency changes over time (Hughes & Jones, 1985, Littorina sp.; Hughes & Mather, 1986, Litto- rina sp.; Greenwood, 1992, Cepaea sp.). The morph categories that Gruneberg de- scribed may have led to his reporting relatively uniform morph frequencies over large dis- tances, because little attention was paid to the overall appearance of individuals and a great deal of attention was paid to relatively small differences in shell pattern that may have no selective advantage (assuming visual predation). In addition, Gruneberg’s approach to statistical validation of the data was un- usual because he compared frequencies of single morphs separately between regions, perhaps obscuring overall morph-frequency patterns between populations, instead of comparing frequencies of all morphs present in a population at the one time as is the more common approach (e.g. Reid, 1987). Studies of other polymorphic gastropods have found variation in morph frequencies at a much smaller scale than those described by Gruneberg for C. oualaniensis (Cepaea sp.; reviewed in Cook, 1986). Observations of populations collected from single localities (beaches) have suggested that there can be considerable within-locality morph variation present, and where these differences exist they may relate to microhabitat differences (Cook, 1986; Hughes & Mather, 1986; Reid, 1987). И similar associations exist for Clithon, then Gruneberg’s descriptions of between- region variation in morph frequencies need to be re-examined. This study aimed to reassess the signifi- cance of regional morph variation in Clithon oualaniensis described by Gruneberg by de- termining whether there are shell-morph fre- quency differences within a single locality. Data from allozyme electrophoresis was also used to determine if cryptic species were present—as has been found in the genus Lit- torina (Mastro et al., 1982; Ward, 1990)—that may confuse the comparison of shell-morph frequencies. If the differences in morph fre- quencies among samples were due merely to chance and limited movement between ar- eas, then significant differences in both shell- morph and allele frequencies at enzyme loci would be expected between sites. However, if the differences were due to differential re- moval of particular shell morphs by predators, then, as long as no linkage disequilibrium ex- ists between allozyme and shell-pattern loci, the differentiation in morph structure should not be reflected in the allozymes. MATERIALS AND METHODS A study area was selected at Dingo Beach and two adjacent bays (Nellie and Cham- pagne bays—approximately 20.8°S, 148.8°E) 40 km northeast of Proserpine, Queensland, Australia. The three localities showed similar microhabitat variation and had relatively high densities of Clithon that allowed for easy col- lection of large samples over short distances. Selection of Microhabitats Sampling was designed to assess the vari- ability present within a single locality and variability among localities on similar micro- habitat types. Three distinct microhabitat types were chosen that represented the major microhab- itats utilised by Clithon in the area: 1. seagrass—characterised by sand-mud substratum with a sparse covering of the seagrass Halophila ovalis. 2. rock/coral—fragments of dead coral, rocks and shells on a sand-mud sub- stratum. 3. shelly sand—sandy substratum con- taining fine shell fragments. SHELL PATTERN AND HABITAT ASSOCIATION IN CLITHON 99 TABLE 1. The location of populations, microhabitat types and sample sizes for all populations. Population Time of Microhabitat Sample size number collection Locality type (N) 1 Feb. 1992 Dingo Beach shelly sand 162 2 July 1992 Dingo Beach shelly sand 207 (От rep) 3 July 1992 Dingo Beach shelly sand 250 (15 m rep) 4 July 1992 Dingo Beach shelly sand 221 (30 m rep) 5 July 1992 Dingo Beach shelly sand 223 (45 m rep) 6 July 1992 Nellie Bay shelly sand 248 7 July 1992 Champagne shelly sand 287 Bay 8 Feb. 1992 Dingo Beach rock/coral 435 9 July 1992 Dingo Beach rock/coral 263 10 July 1992 Nellie Bay rock/coral 200 11 Feb. 1992 Dingo Beach seagrass 187 12 July 1992 Dingo Beach seagrass 244 The microhabitats consisted of a number of adjacent patches of approximately 100 to 150mY*, and beaches were a mosaic of differ- ent-sized patches of the various microhabi- tats. The seagrass microhabitat was available on Dingo Beach only. Rock/coral microhabi- tats were sampled on Dingo Beach and Nellie Bay, and shelly sand microhabitats were sampled on Dingo Beach, Nellie Bay and Champagne Bay. Detailed sampling of the shelly sand microhabitat was undertaken at Dingo Beach in February and July 1992 to assess spatial and temporal variation within individual microhabitats. Preliminary breeding studies on Clithon had shown direct genetic segregation of several characters (Gruneberg, 1976), although shell pattern 1$ presumably under polygenic control (Goodhart, 1987). Following conventions out- lined in Cain (1988) the term “form” will be used to describe different shell appearances (Gruneberg used ‘‘morph’’) because the spe- cific characters in the combinations used have not been tested for heritability. Population Sampling Snails were collected in February and July 1992 (Table 1), with populations collected at both times taken from the same position in the microhabitat. All snails within an area de- fined by four 1 m? quadrats dropped ran- domly in a microhabitat were collected. Care was taken to collect all animals found within the quadrats to avoid selection only of the most conspicuous forms. It was assumed that this collection would be representative of the population at a particular site. All snails from one population were collected during a single low tide. Individuals were placed in numbered cryogenic vials while still alive and stored in liquid nitrogen until they reached the laboratory, where they were transferred to a —80°С freezer and stored pending scor- ing of forms and electrophoretic analysis. Shell-Form Classifications The basic pattern of Clithon consists of fine transverse (“axial”) black lines on a coloured background (Gruneberg, 1976). This pattern is often complicated by the presence of tri- angles or “tongues.” The size of these tongues is variable, as is the distance be- tween the lines, although it is usually con- served within a single individual. A spiral pat- {ет (“ladders””) may be superimposed on the transverse lines exposing the background colour, and this region may contain tongues in “whorls” (Gruneberg, 1976). Usually there are three spirals, the thickness of which can vary, but some individuals may have only one thick central spiral. Variation in background colour ranges from white to green to deep orange. Occasionally, two background co- lours (in sharp zones like those of the spirals) may be present on a single individual. The pattern colour may be black or reddish pur- 100 GARDNER ET AL. ® % в a a 1 => > A 5 @ 4 0 e ® & 7 3 9 10 8 3 Pe, 11 12 13 10mm FIG. 1. Shell-form categories for Clithon oualaniensis. ple. In addition, a sharp, distinct character may be present, referred to as a “purple spi- ral” by Gruneberg (1976). This character was only found at a zone near the suture and at the opposite end to the suture. Individuals were scored for shell colour and pattern using a system modified from Gruneberg (1976). Juvenile shell patterns were scored using a stereo microscope, and when individuals were difficult to assign to specific forms they were excluded from the analyses. Gruneberg developed a system that was weighted heavily towards the type of patterning present on the shells, with a total of 17 main forms identified (Gruneberg, 1976). Our scoring method simplified the pat- terning emphasis by grouping similar catego- ries. The difference between fine transverse lines and coarse transverse lines (Gruneberg, 1976) was considered to be of little signifi- cance, and these two forms were subse- quently grouped. Similarly, ‘‘zebras’’ and “ti- gers’ have been pooled, and narrow “spirals” and “spiral tongues” have been in- cluded with “ladders.” Our approach in- volved a shift of focus from an emphasis on pattern type to a system combining pattern colour, shell colour and pattern type, which emphased the overall appearance of snails. Snails that had a similar overall appear- ance were grouped into forms, with 14 forms being recognised. The smaller number of forms recognised in the present study would, presumably, mask differences that would have been present in Gruneberg’s study, and if differences are found, they would be of a greater magnitude than those found by Gruneberg. The Clithon forms are shown in Figure 1 and described in Table 2. Electrophoretic Analysis Individuals were placed on ice immediately after removal from the freezer and were re- moved from their shells using forceps and a probe. Tissues were homogenised in 30 to 250 micro-litres of grinding buffer (Tris HCI pH 8 diluted 1:20 with distilled water plus 0.1% Triton X-100) (Richardson et al., 1986), de- pending on individual size. The homogenate was centrifuged at 6,000 rpm for 15 min at 4°C. The supernatant was removed and used immediately for electrophoresis on cellulose acetate plates (Titan Ш Zip Zone Cellulose Acetate Plates, Helena Laboratories, Texas USA) using a 75mM Tris citrate buffer pH 7.0 as an electrode buffer. The plates were then stained for the appropriate enzyme following SHELL PATTERN AND HABITAT ASSOCIATION IN СИТНОМ 101 TABLE 2. Description of Clithon oualaniensis forms at Dingo Beach. Morph No. Description form 1 characterised by closely spaced transverse lines with or without small tongues on a white or green background. form 2 has 3 spirals (the pattern colour of which is black) superimposed on the basic pattern with a green background and usually tongues in whorls. form 3 similar to Form 2 but with a red or purple pattern colour. form 4 relate to what Gruneberg (1976) called ‘tigers’ and ‘zebras’. They have widely spaced transverse lines on a green background with little or no tongues. form 5 may be any pattern on a green background but must have a purple spiral. form 6 similar to Form 4 but with large tongues covering the shell. form 7 a single large spiral is found in these snails and the background within this spiral is usually green with the rest of the background white. Tongues in whorls may be present. form 8 purple spiral оп an orange or yellow background. form 9 plain white or green snails with little or no pattern. form 10 plain yellow or orange snails with little or no patterning. form 11 yellow or orange snails with closely spaced transverse lines with or without small tongues. form 12 tiger or zebra with or without tongues on a yellow or orange background. form 13 spiralled orange snails usually found without tongues in whorls. form 14 jet black appearance. Actually this form type has very close, black transverse lines on a whitish green (never yellow or orange) background. procedures outlined in Richardson et al. (1986). Electrophoresis was performed at 4°C. From each sample, individuals were exam- ined for four polymorphic enzyme systems, representing six polymorphic gene loci: as- partate aminotransferase (E.C.2.6.1.1; Aat-1 and Aat-2 loci); esterase D (E.C.3.1.1.1; EstD- 1); isocitrate dehydrogenase (E.C.1.1.1.42; Idh-1 and Idh-2) and 6-phosphogluconate de- hydrogenase (E.C.1.1.1.44; 6pgd-1). Other polymorphic enzymes, including aconitase, adenylate kinase, and phosphoglucomutase, were not scored due to poor resolution, the enzymes denaturing after prolonged periods of freezing. For enzymes with more than a single locus, the most anodal locus was des- ignated as 1, whereas alleles at individual loci were designated with alphabetic characters from the anodal (a) to the most cathodal (2). Alleles d and e at the EstD-1 locus were grouped as the а allele as they had very similar mobilities, and they were difficult to resolve on some plates. Analysis of Genetic Data Except where indicated, the computer pro- gram BIOSYS-1 (Swofford 8 Selander, 1989) was used to analyse allozyme data. The goodness of fit of observed genotype fre- quencies was tested to those expected if the population was in Hardy-Weinberg equilib- rium using a y” test. Deviation of observed genotype frequencies from expected values was estimated by the F. statistic (Wright, 1965), in which F,, = 1 — (Н/2Мра,) where H is the observed frequency of heterozygotes in the sample and М 1$ the sample size. Shell Form and Allelic Frequency Comparisons (Using a y” Test for Independence) Frequencies of shell form and allelic varia- tion were compared between sites within one microhabitat, between like microhabitats from different localities, between different mi- crohabitats, and over time within one habitat. To ensure expected values were >5, it was necessary to group some forms in some anal- yses. All groupings were based on similarity of appearance. The same problem arose with electrophoretic data, and genotypes were pooled into two to three classes when three or more alleles were observed at a single locus. Allelic frequencies at the Aat-1 and Aat-2 loci could not to be compared due to the near fixation of alleles observed at these loci. RESULTS Within-Habitat Variation There was no significant variation in form frequencies between the four replicate sam- ples taken from different areas of the shelly 102 frequency (%) frequency (%) 40 w о № о — o 50 > о 10 GARDNER ET AL. Shelly sand replicates Dingo Beach От 15т п = 207 п = 250 45т п = 223 FIG. 2. Comparison of shell-form frequencies at four replicate sites in а shelly sand site. February VE LPS RE RAN A RE 53 mr Ta EAS form ‹ 4 seagrass u shelly М rock/coral > n= 187 п = 162 п = 435 FIG. 3. Shell-form frequencies in three microhabitats in February. SHELL PATTERN AND HABITAT ASSOCIATION IN СИТНОМ 103 sand microhabitat at Dingo Beach during July (Fig 2: X392 = 44.2, р > 0.05). The repli- cate samples from this analysis were thus considered as a single sample in further anal- yses. Variation at a Local Scale Analysis of the Dingo Beach data indicated significant differences in shell-form frequen- cies between micro habitats in both months (Fig 3: February хо? = 98.7, р < 0.01; Fig 4: July Xogz = 50.0, р < 0.01). (Due to low ex- pected values in some cells, forms 8 and 13, 10 and 11, and 5 and 6 were grouped for the Feb. analysis.) In the February collection, for example, the relative frequencies of forms 3, 11, 13 and 14 were lower, and forms 1 and 2 were a higher in the seagrass microhabitats than in other microhabitats. Forms 2 and 7 had higher relative frequencies in the sea- grass microhabitat in the July collection, and the relative frequencies of forms 4, 11 and 13 were lower in the seagrass locality for the same month. Temporal Variation in Shell-Form Frequency There was significant temporal variation in form frequencies in the shelly sand and seagrass microhabitats (shelly sand x%,>2 = 35.3, р < 0.01; seagrass 4,22 = 29.9, р < 0.01: forms 8 and 13 were grouped), but not in the rock/coral microhabitat (4,42 = 19.2, р > 0.05) (Figs. 3, 4). Temporal effects were most evident in the shelly sand microhabitat, where the relative frequencies of forms 1 and 4 had increased and the relative frequencies of forms 11 and 14 had declined between collection dates. Variation to a lesser extent was evident at the seagrass site, where a decrease in the rela- tive frequency of form 1 occurred over time, while the frequency of most other forms (ex- cept form 2) increased. The non-significant result for different collection times at the rock/coral locality suggested higher temporal stability at this site. Within Habitat Variation Among Localities There was no significant variation in form frequencies between like microhabitats at different localities (Fig 5: shelly sand X,. = 26.8, р > 0.05; Fig 6: coral/rock X,,2 = 12.5, p > 0.05: forms 8 and 13, 10 and 11 were grouped). Therefore, the relative frequencies of shell forms did not vary between the same microhabitats (shelly sand and rock/coral) at different localities in July 1992. Presence of Cryptic Species If cryptic species were present, deficien- cies of heterozygotes would be expected at sites where two or more species were repre- sented in significant proportions. Thus, at some sites heterozygote defficiencies would be expected across the range of loci, whereas at other sites, where a single spe- cies comprised most of the sample, no sig- nificant effect would be detected. In addition, at those sites where low heterozygosity oc- curred, linkage disequilibrium would be ex- pected. Allele frequencies for all loci investigated in each population are shown in Table 3. The F,, values for populations that deviated from Hardy-Weinberg equilibrium are shown in Ta- ble 4. Some significant values were ob- served, but these are unlikely to be due to the presence of cryptic species because there was no consistency within sites. At only one site was more than one locus out of Hardy- Weinberg equilibrium, and in this instance only two loci out of six possible loci showed this result. Polymorphic loci examined in this study were tested for linkage disequilibrium and no significant results were obtained. Genetic Differentiation Among Sampled Populations There was a significant difference in allelic frequency at the 6pgd-2 locus between pop- ulations on Dingo Beach collected in Febru- ary (6pgd-2:7,2 = 19.458, р < 0.01), but no significant difference was found т the July collections at this locus (6pgd-2:7,2 = 3.965, p > 0.05). No significant differences were found at any other loci in either the February or July collections (February: Idh-1:7,2 = 0.339; p > 0:05; Idh-2:7.2 = 0:55, p.> 0:05; EstD-1:7,2 = 3.331, р > 0.05) (July: Idh-1:722 = 1.226,9 > 0,05; 91-2 7-2 = 1-276. р > 905: EstD-1:xg2 = 7.247, р > 0.05). There were no significant differences in al- lelic frequencies for any locus (Idh-1:%32 = 3.601 p> 0:05. 91-27-26 р 0505; EstD-1:x = 6.953, р > 0.05; 6pdg-2:Xe2 = 3.57, р > 0.05), indicating that allelic frequen- cies are consistent within and between mi- crohabitats and that the Dingo Beach shelly 104 GARDNER ET AL. July 50 40 > = 30 ES о is = E Se (LE y 20 Él | 1048 0 form seagrass = shelly Be rock/coral п = 244 п = 250 п = 263 FIG. 4. Shell-form frequencies in three microhabitats in July. Shelly sand July 40 30 frequency (%) № — o В Dingo Beach Ss Nellie Bay Champagne Bay #| п = 901 п = 248 п = 287 FIG. 5. Relative shell-form frequencies in shelly sand microhabitats at three localities. SHELL PATTERN AND HABITAT ASSOCIATION IN СИТНОМ 105 Rock/coral July 40 frequency (%) N aw o oO — o п = 263 form 5 Dingo Beach Nellie Bay n = 200 FIG. 6. Relative shell-form frequencies in coral/rock microhabitats at two localities. sand populations (July 1992) can be consid- ered as a single population. Gene Flow Between Populations Table 5 shows the results of Fst estimates for each locus in the sampled populations. Values are very low, ranging from 0.011 to 0.018, with a mean of 0.016, indicating that gene flow is extensive between microhabitats at the same locality and between adjacent localities. This lack of genetic differentiation indicated that the significant differences in form frequencies described earlier are not the result of limited gene exchange between pop- ulations occupying different microhabitats but must be due to other factors. DISCUSSION No evidence for the presence of cryptic species was obtained in this study, but sig- nificant differences in shell-form frequencies related to microhabitat were evident in the intertidal snail Clithon oualaniensis within a single locality (beach). Shell-form frequen- cies were also found to change temporally in some microhabitats. Differences in form fre- quencies at a local scale have not been pre- viously reported for this species. There was, however, no evidence for a relationship be- tween genotype and microhabitat utilisation in the enzyme loci investigated in this study, and any small differences in allelic frequency among sampled populations possibly re- flects the lack of a larval dispersal phase (Gruneberg, 1976) or a sampling error due to the number of tests performed (Type | error). The lack of genetic differentiation among populations suggests a high level of gene flow. At any rate, the differences between sites in shell-form frequencies are much greater than any observed at allozyme loci. Evidence of microhabitat and temporal dif- ferences in colour-form frequencies at a sin- gle locality requires a reassessment of Gruneberg's results (Gruneberg, 1976, 1978, 1979). Gruneberg (1979) compared popula- tions from different geographic regions, in- cluding northeastern Queensland, Malaya- Singapore, Hong Kong, Bay of Bengal, and the Gulf of Mannar (Arabian Sea), and de- scribed significant differences in the frequen- cies of some forms between these areas. For 106 TABLE 3. Allelic frequencies for all populations. GARDNER ET AL. 10 Locus 1 2 3 4 5 Idh-1 (N) 22 47 45 48 48 A 0.00 0.01 0.00 0.01 0.01 B 0.86 0.85 0.87 0.82 0.71 С 0.00 0.01 0.00 0.00 0.00 DE 0.14 0.13 Ouls 0.17 0.27 E 0.00 0.00 0.00 0.00 0.01 Idh-2 (N) 22 45 45 48 48 A 0.91 0.91 0.89 0.96 0.96 B 0.09 0.09 0.11 0.04 0.02 С 0.00 0.00 0.00 0.00 0.02 D 0.00 0.00 0.00 0.00 0.00 Aat-1 (N) 32 48 48 48 48 A 0.00 0.00 0.00 0.00 0.00 B 0.00 0.02 0.00 0.00 0.02 C 0.98 0.98 0.99 1.00 0.98 D 0.00 0.00 0.00 0.00 0.00 E 0.02 0.00 0.01 0.00 0.00 Aat-2 (N) 32 48 48 48 48 A 0.00 0.01 0.00 0.00 0.00 B 0.00 0.00 0.03 0.00 0.02 С 0.00 0.01 0.00 0.00 0.01 D 0.98 0.95 0.96 1.00 0.97 E 0.00 0.03 0.00 0.00 0.00 F 0.02 0.00 0.01 0.00 0.00 6pgd-2 (N) 33 46 46 48 48 A 0.00 0.00 0.00 0.00 0.00 B 0.20 0.02 0.01 0.02 0.03 С 0.64 0.51 0.49 0.48 0.64 D 0.17 0.44 0.45 0.45 0.30 E 0.00 0.03 0.05 0.05 0.03 EstD-1 (N) 12 47 44 47 47 A 0.08 0.12 0.26 0.12 0.14 B 0.71 0.59 0.46 0.66 0.56 С 0.13 0.20 0.16 0.15 0.18 D 0.08 0.10 0.13 0.08 0.12 Population 6 7 8 9 11 12 48 47 72 59 48 58 45 0:03 0:00 001 0.01 0.00 0.01 0.01 0.79 0.77 0.85 080 0851 10:8280:88 0.00 0.00 0.00 0.00 0.02 0.0 0.00 0:17 0.23 0.14 020 Озоне 0.01 0.00 0.01 0.00 0.00 0.00 0.00 48 48 72 59 48 57 45 0.93 0.89 087 094 093 991 0.92 0:07 0.12 010 006 00/00 IE 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 003 000 0001100088000 48 48 72 60 48 60 46 0:00 000 0:00) 0100885001 0.00 0.00 0.00 0.00 0.00 0.00 0.01 0.00 0.00 1.00 1.00 1.00 1.00 0.98 1.00 0.98 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.02 48 48 72 60 48 60 46 0.02 0.00 0.00 0.00 0.01 0.00 0.01 0.01 0.00 0.01 0.02 002002010 0.00 0.00 0.00 0.00 0.01 0.00 0.01 0.95 1.00 0.97 0.97 0.96 0.96 0.97 0.00 0.00 0.00 0.0 0.00 0.00 0.00 0.02 0:00 002 0:02 бот 48 48 UE 58 46 59 45 0.01 0.00 0.00 0.00 0.01 0.00 0.00 0.01 0.00 0.01 0.01 0:07 000800) 0157250163062 0.50 0.46 0.62 0.66 0:37 (035 035 DAS 04 0.38 0.30 0.04 0.02 0.02 0.01 0.05 0.00 0.03 48 46 69 59 46 36 45 о ля 0.127 0.14 0131018086 0.58 0.55 0.49 0.49 0/69 MD IIS 0:62 0.097 023 0.23 023 OSO 05 eh? Si 0.17 014 70.0555 0 AOS *Represents a null allele at this locus example, the form “purple spiral” was re- ported to have an incidence of 15.5% in northeastern Queensland as compared to 0.015% in the Gulf of Mannar. With no de- tailed investigation of form-frequency differ- ences within local areas, however, Grune- berg could not know if similar differences were present on much smaller geographic scales. The present study has shown signif- icant differences between different micro- habitats on a local scale. Therefore, any in- ferences about frequency differences on larger scales without such information is open to question. Moreover, any correlation of form frequency with an environmental vari- able—e.g. surface salinity, as Gruneberg (1979) suggested—has to be viewed with caution unless the relationship can be shown to be related to local form-frequency differ- ences as well. Gruneberg's (1982) explanation for the variability in shell colour and pattern in C. oualaniensis as being ‘‘pseudo-polymor- phic” has also been criticised by some au- SHELL PATTERN AND HABITAT ASSOCIATION IN СИТНОМ 107 TABLE 4. Departure from Hardy-Weinberg as estimated by the fixation index. Observed Expected Fixation index Pop. number Locus heterozygotes heterozygotes (F) 1 Idh-1 2 5.182 0.614 2 Idh-2 4 7.289 0.451 3 Aat-2 2 3.865 0.482 Idh-2 4 8.889 0.550 4 EstD-1 20 24.670 0.189 5 Idh-2 2 3.875 0.484 12 Idh-1 9 11.967 0.248 thors (e.g., Goodhart, 1987). The delineation of “eastern” and “western” Clithon morph fre- quencies by Gruneberg (1976, 1979) was made on the basis that particular morphs oc- curred at significantly different frequencies between regions. For example, three differ- ent types of “ladders” were recognised by Gruneberg (1979): “spiral tongues” (tongues in whorls), “ladders proper” (spirals with no tongues), and “yellow spirals” (spiralled or- ange). He reported that the relative frequen- cies of each of these morph categories dif- fered considerably from one population to another (in “western Clithon’’). Gruneberg at- tributed this to the influence of environmental conditions that vary irregularly in time and space, and suggested that the “morphs” did not relate to separate genotypes. However, in Gruneberg’s “eastern Clithon,” ladders were not observed, although yellow spirals and spiralled tongues were present (Gruneberg, 1979). He attributed these differences to a founder event, with the variation in the phe- notypic expression of the ladder morphs in “western Clithon” absent in the eastern form. Thus, in one region all ladders were sup- posed to be the same genotype expressed differently due to environmental conditions, but in his eastern populations the different ladder morphs were supposed to be sepa- rate genotypes. All the morphs found in Gruneberg's eastern and western popula- tions were found in the present study, and there seems little basis for his separation of regional (eastern vs. western) classes of Cli- thon populations. Temporal variation in relative form frequen- cies found in the present study also suggests a further criticism of Gruneberg’s interpreta- tions. Temporal variation may be due to dif- ferences in mortality between forms in differ- ent microhabitats possibly caused by predation (Cook, 1986; Reid, 1987), variable temperature conditions (Cook, 1986; Green- TABLE 5. Fst values for all loci for February and July samples. LOCUS FEB Fst JULY Fst ldh-1 0.002 0.015 ldh-2 0.003 0.011 Aat-1 0.010 0.011 Aat-2 0.003 0.011 6pgd-2 0.034 0.018 EstD-1 0.026 0.017 Mean 0.021 0.016 wood, 1992) or variation in physiological stress (Etter, 1988) for example. The pres- ence of temporal changes in shell-form fre- quency add support to our suggestion of some external factor affecting form frequen- cies and not subpopulation structuring as would be the case if cryptic species were present. A number of alternative explanations have been proposed to explain shell-pattern poly- morphisms in gastropod species. The impor- tance of crypsis was stressed by Goodhart (1987), who considered that a generally cryp- tic appearance should be favoured by natural selection except in such special cases as warning colouration in many noxious crea- tures. Extremely polymorphic species (Such as Clithon) may contain different morphs that are at an equal selective advantage in differ- ent microhabitats, with a range of morphs able to exist in all situations where the spe- cies is found (Endler, 1988). However, some morphs may be more favoured in certain mi- crohabitats, which could lead to form-fre- quency differences as were found in this study. Thus, differences in form frequences on local and larger scales may be explained by a combination of frequency-dependent selection (Clark et al. , 1988) and selection for crypsis. Predator-mediated selection for crypsis is 108 GARDNER ET AL. thought to be the most plausible explanation for the relationship between shell-colour- form frequencies and habitat in Clithon. How- ever, detailed caging experiments, such as those of Hughes & Mather (1986), would be necessary before there is substantial evi- dence of selection for crypsis influencing the form frequencies of Clithon populations in different microhabitats. By eliminating all po- tential predators of this species and manip- ulating the form frequencies in caged and non-caged areas, the survival of various forms in different microhabitats could be as- sessed. Therefore, before any explanations that seek to use environmental gradients and variables to explain form-frequency differ- ences in shell-colour patterns in Clithon oualaniensis can be accepted, more detailed analyses of local variation in these characters needs to be undertaken. Our study would in- dicate that, at best, any positive correlations between environmental variables and colour- form frequencies on a regional scale may re- late to the fact that local populations were collected from different microhabitat types from localities within each region. So, signif- icant differences in form frequencies be- tween regions perhaps reflect differences in microhabitat availability between the regions, and where the same microhabitats are avail- able within an individual site, the same or re- lated significant form-frequency differences may also exist. ACKNOWLEDGMENTS This study formed part of an Honours de- gree (M. Gardner) and would not have been possible without a bursary provided by the Centre for Biological Population Manage- ment. Michael Gardner would like to thank his wife, Joyanne, for her inspiration, encour- agement and patience during the Honours year. Fellow honours students helped in var- ious ways, and the authors are also indebted to all those involved in the collection of snails. The authors would also like to thank the three anonymous reviewers of an earlier version of this paper for their helpful criticism and sug- gestions. LITERATURE CITED BRAKEFIELD, P. M., 1990, Genetic drift and pat- terns of diversity among colour-polymorphic populations of the homopteran, Philaenus spumarius in an island archipelago. Biological Journal of the Linnaen Society, 39: 219-37. CAIN, A. J., 1988, The scoring of polymorphic co- lour and pattern variation and its genetic base in molluscan shells. Malacologia, 28: 1-15. САМ, А. J. 4 J. D. CURREY, 1963, Area effects in Cepaea. Philosophical Transactions of the Royal Society of London, B, 246:1-81. CHANG, H. & J. M. EMLEN, 1993, Seasonal vari- ation of microhabitat distribution of the polymor- phic land snail Cepaea nemoralis. Oecologia, 93: 501-507. CLARKE, B. C., P. R. SHELTON & G. S. MANI, 1988. Frequency-dependent selection, metrical characters and molecular evolution. Philosophi- cal Transactions of the Royal Society of London, B, 319: 631-640. COOK, L. M., 1983, Polymorphism in a mangrove snail in Papua New Guinea. Biological Journal of the Linnean Society, 20: 167-173. COOK, L. M., 1986, Polymorphic snails on varied backgrounds. Biological Journal of the Linnean Society, 29: 89-99. COOK, L. М. & G. KENYON, 1991, Frequency-de- pendent selection with background heterogene- ity. Heredity, 66: 67-73. DYE, A. H. & T. A. LASIAK, 1987, Assimilation ef- ficiencies of fiddler crabs and deposit-feeding gastropods from tropical mangrove sediments. Comparative Biochemistry and Physiology, 87: 341-3. ENDLER, J. A., 1988, Frequency-dependent pre- dation, crypsis and aposematic coloration. Philosophical Transactions of the Royal Society of London, B, 319: 505-523. ETTER, R. J., 1988, Physiological stress and colour polymorphism in the intertidal snail Nucella lapil- lus. Evolution, 42: 660-680. GILLESPIE, R. G. 4 B. E. TABASHNIK, 1990, Main- taining a happy face: stable colour polymor- phism in the spider, Theiridion grallator (Araneae, Theridiidae). Heredity, 65: 67-74. GOODHART, C. B., 1987, Why are some snails visibly polymorphic, and others not? Biological Journal of the Linnean Society, 31: 35-58. GREENWOOD, J. J. D., 1992, In the pink (and yel- low). Nature, 357: 192-192. GRUNEBERG, H., 1976, Population studies on a polymorphic prosobranch snail (Clithon (Pic- toneritina) oualaniensis Lesson). Philosophical Transactions of the Royal Society of London, B, 275: 385-437. GRUNEBERG, H., 1978, Micro-evolution in a poly- morphic prosobranch snail (Clithon oualaniensis (Lesson)). Proceedings of the Royal Society of London, B, 200: 419-440. GRUNEBERG, H., 1979, A search for causes of polymorphism in Clithon oualaniensis (Lesson) (Gastropoda; Prosobranchia). Proceedings of the Royal Society of London, B, 203: 379-386. GRUNEBERG, H., 1982, Pseudo-polymorphism in Clithon oualaniensis. Proceedings of the Royal Society of London, B, 216: 147-57. SHELL PATTERN AND HABITAT ASSOCIATION IN CLITHON 109 HUGHES, J. M., 1982, An explanation for the asy- metrical hybrid zone between white-backed and black-backed magpies. Emu, 82: 50-3. HUGHES, J. М. 8 M. P. JONES, 1985, Shell colour polymorphism in a mangrove snail Littorina sp. (Prosobranchia: Littorinidae). Biological Journal of the Linnean Society, 25: 365-378. HUGHES, J. М. & P. В. MATHER, 1986, Evidence for predation as a factor in determining shell co- lour frequencies in a mangrove snail Littorina sp. (Prosobranchia: Littorinidae). Evolution, 40: 68- 77. KETTLEWELL, B., 1973, The evolution of melan- ism. Oxford University Press, London. MASTRO, E., V. CHOW & О. HEDGECOCK, 1982, Littorina scutulata and Littorina plena: sibling species status of two prosobranch gastropod species confirmed by electrophoresis. The Veliger, 24: 239-246. REIMCHEN, T. E., 1979, Substratum heterogene- ity, crypsis, and colour polymorphism in an in- tertidal snail (Littorina mariae). Canadian Journal of Zoology, 57: 1070-1085. REID, D. G., 1987, Natural selection for apostasy and crypsis acting on the shell colour polymor- phism of a mangrove snail, Littoraria filosa (Sow- erby) (Gastropoda: Littorinidae). Biological Jour- nal of the Linnean Society, 30: 1-24. RICHARDSON, В. J., Р. В. BAVERSTOCK & М. ADAMS, 1986, Allozyme electrophoresis. Аса- demic Press, Sydney. SWOFFORD, D. L. 8 В. В. SELANDER, 1989, Bio- sys-1; a computer program for the analysis of allelic variation in population genetics and bio- chemical systematics. Release 1.7 Illinois Natu- ral History Survey, Illinois. WARD, R. D., 1990, Biochemical genetic variation in the genus Littorina (Prosobranchia: Mollusca). Hyarobiologia, 193: 53-9. WRIGHT, S., 1965, The interpretation of population structure by F-statistics with special regard to systems of mating. Evolution, 19: 395-420. Revised Ms. accepted 2 March 1994 MALACOLOGIA, 1995, 36(1-2): 111-137 EL GÉNERO CANARIELLA HESSE, 1918, Y SU POSICIÓN EN LA FAMILIA HYGROMIIDAE (GASTROPODA, PULMONATA, HELICOIDEA)' Miguel Ibáñez, Elena Ponte-Lira & María В. Alonso Departamento de Biología Animal, Universidad de La Laguna, E-38206 La Laguna, Tenerife, España ABSTRACT Canariella is a poorly known genus of the Hygromiidae, endemic to the Canary Islands, with 18 nominal taxa of specific and subspecific rank. Until now, no information on the internal anatomy of its genital ducts was known, and the external morphology of the genital system, which lacks any trace of the dart-sac complex, was known for only five species. In another article (Groh et al., in press), four nominal taxa (= three species) of Canariella were described conchologically and anatomically. The present work treats the remaining known species. (We exclude Helix plutonia Lowe, 1861, which has been included in Canariella but really belongs in a new genus of the Hygromiidae.) Lectotypes are designated for the type species, Carocolla hispidula Lamarck, 1822; Helix bertheloti Férussac, 1835; H. everia Mabille, 1882; H. fortunata Shuttleworth, 1852; H. (Gonos- toma) hispidula subhispidula Mousson, 1872; and H. (Ciliella) lanosa Mousson, 1872. The holotypes of Helicodonta salteri Gude, 1911, and Helix (Gonostoma) beata Wollaston, 1878, are also studied. These eight taxa differ slightly from one another in shell morphology, but agree in the morphology of the genital system, and there is no geographical isolation among them (Fig. 39); therefore, they are considered to belong to a single species, and the last seven names are synonymized with Carocolla hispidula. However, six populations are conchologically distin- guishable, a sign of the beginning of radiation, and therefore we consider them with the rank of infrasubspecific varieties. Lectotypes of Helix (Gonostoma) gomerae Wollaston, 1878; Carocolla planaria Lamarck, 1822; Helix afficta Férussac, 1832; and H. eutropis Shuttleworth, 1860, are also designated. Applying the results of this study and the authors’ knowledge of further new species not yet described, the following new diagnosis of Canariella is proposed: Mantle collar with four small lobes (subpneumostomal, left dorsal, right dorsal and right lateral; as an exception, C. eutropis has also the left lateral lobe). Kidney sigmurethric, without secondary ureter. Central and first lateral radular teeth with small but evident ectocones. Right ommatophore retractor passing between penis and vagina. Genital system without the dart-sac complex and with several vaginal digitiform glands, each with an independent, very slender initial portion; they are crown-shaped when there are more than three. With a sheath surround- ing the distal male duct (between the atrium and the penis retractor muscle insertion). With internal differentiation penis-epiphallus (externally this differentiation can be undistinguishable). Penis retractor muscle with an epiphallar insertion. Penial nerve originating from the right cerebral ganglion (verified in the type and two additional species). Canariella is considered as “incertae sedis” within the Hygromiidae and is compared with several other hygromiid genera without the dart-sac complex. The most closely related genera are Montserratina, Ciliella (which is not present in the Canary Islands, in spite of published records), Schileykiella, Tyrrheniellina and Ciliellopsis. Other Hygromiidae species lack dart-sac complex, but they differ in the presence of vaginal appendages or in the morphology of the terminal parts of the male ducts. Key Words: Pulmonata, Hygromiidae, Canariella, systematics, Canary Islands. INTRODUCCIÓN peor conocidos en la actualidad, a pesar de que los primeros datos bibliográficos exis- El género Canariella Hesse, 1918, en- tentes sobre sus especies se remontan a démico del Archipiélago Canario, es uno de principios del siglo XIX. En efecto, Férussac los representantes de la familia Hygromiidae (1821), publicó los nombres de las dos pri- Notes on the Malacofauna of the Canary Islands, No. 28 (subvencionado con el proyecto 92/160 del Gobierno de Canarias). 111 Wes IBANEZ, PONTE-LIRA & ALONSO meras: Helix (Helicigona) afficta y H. (Helici- gona) lens, siendo ambos nombres “no dis- ponibles” (ICZN, Artículo 12). Al ano siguiente, Lamarck (1822) describid ambas especies y las denomino Carocolla planaria y Carocolla hispidula, respectivamente. Desde entonces, se han descrito (la mayoría sólo conquiológicamente) y asignado a este gé- nero 18 taxones nominales del nivel especie: 1. Carocolla hispidula Lamarck, 1822 2. Carocolla planaria Lamarck, 1822 3. Helix (Helicigona) afficta Férussac, 1832 4. Helix bertheloti Férussac, 1835 5. Helix leprosa Shuttleworth, 1852a 6. Helix fortunata Shuttleworth, 1852a 7. Helix discobolus Shuttleworth, 1852b 8. Helix eutropis Shuttleworth, in Pfeiffer, 1860 9. Helix (Macularia) plutonia Lowe, 1861 10. Helix (Ochthephyla) multigranosa Mousson, 1872 11. Helix (Ciliella) lanosa Mousson, 1872 12. Helix (Gonostoma) hispidula sub- hispidula Mousson, 1872 13. Helix (Gonostoma) gomerae Wollas- ton, 1878 14. Helix (Gonostoma) beata Wollaston, 1878 15. Helix everia Mabille, 1882 16. Helix pthonera Mabille, 1883 17. Helix (Gonostoma) рату! Ponsonby у Sykes, 1894 18. Helicodonta salteri Gude, 1911 Únicamente se ha descrito la anatomía ex- terna del aparato reproductor de cinco de ellos (números 1, 5, 6, 7 y 9), por Krause (1895), Hesse (1931) y Odhner (1931); y está en vías de publicación la redescripción de otros tres (números 10, 16 y 17, junto con una especie nueva para la ciencia: Groh et al., en prensa), incluyendo por primera vez datos de la anatomía interna de sus aparatos repro- ductores. Finalmente, está en vías de publi- cación la descripción de una nueva especie subfósil (Hutterer, en prensa) y la redescrip- ción del número 9, que en realidad pertenece a un nuevo género de Hygromiidae. El presente trabajo está dedicado a los restantes taxones nominales conocidos del género, entre los que se encuentra la especie tipo. En base a los resultados de este estudio y a los datos que poseemos de las otras es- pecies existentes del género (todavía no des- critas), en este artículo se corrige la diagnosis FIG. 1. Medidas tomadas en la concha (explica- ción, en el texto). de Canariella y se discute su posición entre los Hygromiidae. METODOLOGÍA Para hacerlas más objetivas, las descrip- ciones se han basado, en parte, en datos biométricos. Las medidas, realizadas con un calibrador digital electrónico conectado a una computadora, son las siguientes (Fig. 1): A: altura de la concha; B: diámetro de la con- cha; C: altura de la última vuelta de espira; D: altura del lado ventral de la concha; E: longi- tud de la abertura; F: anchura de la abertura; G: diámetro del ombligo (sin peristoma); H: altura del lado dorsal de la concha (= A-D). Con ellas se han calculado los valores má- ximo (M), mínimo (m) y medio (X), así como el coeficiente de variación (CV) de Pearson (ex- presado en %), indicándose en cada tabla el número de ejemplares medidos (n) del taxón correspondiente. Es necesario señalar que el diámetro del ombligo (G) es difícil de tomar en muchos casos, bien por sus pequeñas di- CANARIELLA Y SU POSICIÓN EN HYGROMIIDAE 113 mensiones, o porque el peristoma lo tapa parcial o incluso totalmente. Cualquier error por pequeño que sea o, simplemente, la exis- tencia de variabilidad en sus dimensiones, es más apreciable que en las otras medidas, que son de mayor magnitud. Esto queda claramente reflejado en los altos valores que presenta el coeficiente de variación de Pear- son en esta medida, por lo que ni ella ni el índice en el que participa son válidos para un estudio estadístico; sin embargo, sí lo son para dar información sobre su tamaño desde el punto de vista descriptivo. Con estas medidas se han calculado los parámetros e índices que se indican a con- tinuación. Los intervalos utilizados se han calculado teniendo en cuenta todas las es- pecies del género. — Tamaño de la concha (B). En función de su diámetro, la concha es: pequeña: < 11.50 mediana: 1150-1576 grande: 15.76 — 19.02 muy grande: > 19.02 —Indice de la forma de la concha (А/В). Re- laciona la altura de la concha con su diá- metro. Según su valor, la concha es: aplanado-lenticular: < 0.485 deprimida: 0.485 — 0.60 cónico-ovalada: > 0.60 — Indice de la forma dorsal (H/B). Relaciona la altura de la parte dorsal de la concha con su diámetro. Según su valor, la concha dor- salmente es: = 0225 > 0.225 aplanada: cónica: — Indice de la forma ventral (D/B). Relaciona la altura del lado ventral de la concha con su diámetro. Según su valor, la concha ventral- mente es: < 0.375 > 0.375 aplanada: ovalada: — ndice de la forma de la abertura (E/F). Re- laciona la longitud de la abertura con su an- chura. Según su valor, la abertura es: ovalado-deprimida: < 0.825 ovalada: 0.825 — 0.895 redondeada: > 0.895 — пасе del tamaño del ombligo (B/G). Re- laciona el diámetro de la concha con el del ombligo. Según su valor, el ombligo es: muy pequeño: > 29.095 pequeño: 20.57 — 29.095 mediano: 11.50 = 20:57 grande: <111,90 En cuanto al aparato reproductor, no se han tomado medidas de sus conductos al observar que existe variabilidad en sus di- mensiones dentro de una misma especie. Esta variabilidad probablemente se debe al diferente momento del desarrollo del repro- ductor con respecto a la fase de reproduc- ción. Además, en muchas especies el nú- mero de aparatos reproductores disponibles en estado adulto no era suficiente para que las medidas obtenidas fueran estadística- mente fiables. Con respecto a la terminología del aparato reproductor, se utiliza la palabra “proximal” para designar a la zona más cer- cana a la gonada, y “distal” para la más cer- cana al orificio genital. Abreviaturas: ANSP— Academy of Natural Sciences, Philadelphia; CGH— K. Groh pri- vate collection, Hackenheim; DMNH— Dela- ware Museum of Natural History, Greenville, FMNH— Field Museum of Natural History, Chicago, MHNG— Muséum d'Histoire Na- turelle, Geneve; MNHN— Muséum National d'Histoire Naturelle, Paris; NHM— Natural History Museum, London; NMB— Naturhis- torisches Museum, Bern; NMW— National Museum of Wales, Cardiff; TFMC— Museo de Ciencias Naturales de Tenerife, Santa Cruz de Tenerife; ZMZ— Zoologisches Mu- seum der Universitát, Zúrich. DESCRIPCIONES TAXONÓMICAS Familia Hygromiidae Tryon, 1866 Género Canariella Hesse, 1918 Especie tipo: Carocolla hispidula Lamarck, 1822. Designación: Hesse (1918: 106-107). Diagnosis Original “An dem stark geschwollenen Penis sitzt ein schmächtigerer, nach dem Vas deferens zu sich verjüngender Epiphallus und ein win- ziges Flagellum mit hakenfórmig umgebo- gener Spitze. Der Retractor ist an der Grenze von Penis und Epiphallus angeheftet; an der 114 IBÁÑEZ, PONTE-LIRA 8 ALONSO langen Vagina sitzen drei ziemlich lange, dünne, cylindrische Glandulae mucosae; Pfeilsack nicht vorhanden. Samenblase kugelig auf kräftigem Stiel. Uterushals etwa halb so lang wie die Vagina” (Hesse, 1918). [A continuaciön del pene, grueso, hay un del- gado epifalo que, despues del conducto de- ferente, se estrecha en un diminuto flagelo con la punta torcida a modo de gancho. El músculo retractor está sujeto en el límite en- tre el pene y el epifalo; en la vagina, larga, hay tres glándulas mucosas bastante largas, delgadas y cilíndricas; no existe saco del dardo. La bolsa copulatriz es esférica y está sobre un conducto grueso. El cuello del útero es casi la mitad de largo que la vagina.] Hesse (1918) se basó únicamente en la forma externa del aparato reproductor, por lo que se equivocó al considerar que el límite entre el pene y epifalo está a nivel de la in- serción del músculo retractor; en realidad, este músculo se inserta en el epifalo, hecho que fue reconocido posteriormente por este autor (Hesse, 1931) en Helix fortunata. Ade- más, hay otras características importantes que pueden ser incluídas en la diagnosis; por ello, a continuación se realiza una breve des- cripción del género, y de ella se extrae una nueva diagnosis. Descripción del Género Canariella La concha mide de 6 a 21 mm de diámetro y tiene de 4 a 6 3/4 vueltas de espira; nor- malmente es aplanada, angulada o aquillada y está cubierta por pelos periostracales menores de 1 mm de largo. El color más típico es marrón claro y la ornamentación de la te- loconcha está formada por costulaciones ra- diales en número y grosor variables, sobre las que se superponen crestas espirales muy fi- nas. El peristoma no está engrosado y forma un pequeño labio en las zonas columelar y basal; sus extremos normalmente se insertan alejados entre sí en la zona parietal y entre ellos, en los individuos más viejos, hay una zona provista de una tenue callosidad. El collar del manto (Fig. 2) tiene forma de “О? inclinada y el ángulo superior se encuen- tra cerca del pneumostoma, en cuyas proxi- midades se disponen cuatro lóbulos poco desarrollados (la nomenclatura está basada en Gittenberger & Winter, 1985). El lateral derecho es el más largo y grueso; es relati- vamente ancho en la zona de contacto con el ano y adelgaza progresivamente hasta la parte media del collar. El dorsal derecho está 2 FIGS. 2-3. Canariella hispidula var. hispidula (Tabaiba Alta, Tenerife). (2) Collar del manto. (3) Complejo paleal. Escala: 1 mm. situado encima del ano y es muy pequeño, casi inconspicuo. El dorsal izquierdo es más conspicuo que el derecho, terminando en un extremo prominente. El subpneumostomal tiene forma triangular y está apenas engro- sado. Salvo en C. eutropis, es destacable la ausencia del lóbulo lateral izquierdo, que está presente en otros Hygromiidae. CANARIELLA Y SU POSICIÓN EN HYGROMIIDAE 115 El complejo paleal (Fig. 3) ocupa la última vuelta de espira. El techo del pulmón pre- senta una serie de manchas oscuras, irregu- lares. El riñón es sigmurétrico, sin uréter se- cundario, la mitad de largo que el pulmón y el doble que el corazón. La mandíbula (Fig. 18), odontognata, está provista de un número variable de costillas, que pueden ser anchas o estrechas, salvo en los laterales, donde es casi lisa. La rádula (Figs. 19, 20) consta de 90 a 140 filas de dientes, sin que se aprecie delimitación clara entre los laterales y los marginales. El central tiene un mesocono de punta ligeramente aguda y dos ectoconos pequeños, pero níti- dos, en su base. Los primeros dientes la- terales son más grandes y robustos que el central y están provistos de un mesocono y de un pequeñco ectocono, ambos puntiagu- dos. Hacia los márgenes, la anchura del me- socono disminuye a la vez que aumenta la del ectocono, que puede tener su cúspide dividida en dentículos. Aparato reproductor (Figs. 28-38). El atrio es corto. El músculo retractor del pene se inserta en el epifalo. El pene es tubular y se diferencia del epifalo por su anatomía interna (externamente esta diferenciación puede no apreciarse); está envuelto por una vaina fina y traslúcida, que se suelda a él en su extremo distal (junto al atrio) y al epifalo a nivel de la inserción del músculo retractor. El conducto deferente desemboca lateroapicalmente en el epifalo, junto con el flagelo. El epifalo es tubular y alberga en su interior dos a cinco pliegues longitudinales, de los que general- mente se prolongan dos en la cavidad del pene, fusionándose parcialmente entre sí y formando una papila peneana acanalada. La papila en algunas especies ocupa poco es- pacio en la luz del pene (Figs. 28-350), o in- cluso puede faltar completamente, como en Canariella multigranosa (Mousson, 1872; Groh et al., en prensa), mientras que en otras llega a ser muy grande (Figs. 37-38c), sepa- rando la cavidad del epifalo de la del pene. Este posee además, en su interior, una serie de boceles longitudinales. La vagina es tubu- lar y también está provista internamente de boceles longitudinales. Carece de saco del dardo, sacos accesorios y apéndices vagina- les accesorios. En ella desembocan 1-8 glándulas vaginales digitiformes, normal- mente simples y dispuestas, cuando hay varias, formando una corona, cuyo diámetro se estrecha bruscamente en la base, antes de su unión con la vagina. El conducto de la bolsa copulatriz alberga gran número de pliegues irregulares. El músculo retractor del ommatóforo dere- cho pasa entre el pene y la vagina. El nervio peneano aparentemente se origina del gan- glio cerebroideo derecho (hasta ahora sólo ha sido confirmado en tres especies, entre ellas la especie tipo del género). Nueva Diagnosis del Género Canariella Collar del manto típicamente con cuatro lóbulos poco desarrollados (lateral derecho, dorsal derecho, dorsal izquierdo y subpneu- mostomal). Riñón sigmurétrico, sin uréter se- cundario. Dientes central y primeros laterales de la rádula provistos de ectoconos peque- ños, pero nítidos. Pene diferenciado del epi- falo por su anatomía interna y envuelto por una vaina penana. Músculo retractor inserto en el epifalo. Vagina tubular, sin trazas del aparato estimulador, con una o varias glán- dulas vaginales digitiformes, que tienen la base de menor diámetro que el resto y están dispuestas, cuando hay más de tres, for- mando una corona. El músculo retractor del ommatóforo derecho pasa entre el pene y la vagina. El nervio peneano se origina aparen- temente del ganglio cerebroideo derecho. Observaciones Los caracteres del collar del manto y del complejo paleal son similares en todas las especies (con la excepción de C. eutropis), omitiendose en ellas su descripción, así como la de otros caracteres comunes, para evitar repeticiones. Canariella hispidula (Lamarck, 1822) Helix (Helicigona) lens Férussac, 1821: 41 (Folio) o 37 (Quarto), no. 153 [nomen nu- dum; non Helix lens, — Deshayes, in Fé- russac 8 Deshayes, 1850 (= Lindol- miola lens,— Gittenberger & Groh, 1986)]; d’Orbigny, 1839: 66, lam. 2 figs. 7-9; Gray, 1854: 11; Chevallier, 1965: 489. Carocolla hispidula Lamarck, 1822: 99 [loc. typ.: Tenerife]; 1838: 148; Mermod, 1951: 713-715, fig. 68 [1: lectotipo; 2: paralectotipo]. Helix (Helicigona) barbata Ferussac, in Ferus- sac & Deshayes, 1832: lam. “66*,” fig. 4 [no. fig. 3; non H. barbata Férussac, 1821]. Helix hispidula,— Webb & Berthelot, 1833: 116 IBÁÑEZ, PONTE-LIRA 8 ALONSO FIGS. 4-11. Concha. (4-9) Canariella hispidula. (4) Lectotipo de Carocolla hispidula (MHNG, foto G. Dajoz; diámetro de la concha: 13 mm). (5) Holotipo de Helicodonta salteri (NHM). (6) Lectotipo de Helix bertheloti (MNHN). (7) Lectotipo de Helix everia (MNHN). (8) Lectotipo de Helix fortunata (ММВ). (9) Holotipo de Helix (Gonostoma) beata (NHM). (10-11) Canariella planaria. (10) Lectotipo de Helix (Helicigona) afficta (ММНМ). (11) Lectotipo de Carocolla planaria (MHNG, foto G. Dajoz). Escala: (5-11) 5 mm. CANARIELLA Y SU POSICION EN HYGROMIIDAE 117 TABLA 1. Datos biométricos (dimensiones en mm, e índices) de la concha en las diferentes variedades de Canariella hispidula. A: altura de la concha; B: diámetro de la concha; C: altura de la última vuelta; D: altura del lado ventral de la concha; E: longitud de la abertura; F: anchura de la abertura; G: diámetro del ombligo (sin peristoma); H: altura del lado dorsal de la concha (= A-D); п: número de ejemplares medidos. VALORES: M, valor máximo; m, valor mínimo; X, valor medio; CV: coeficiente de variación de Pearson (en %). A B С D Е Е G A/B H/B D/B E/F B/G n C. hispidula var. hispidula M 7.11 15.33, 9:06” 4.65 6.26 7.19 3.36 m Sa De er 4257 9 3:33 4.92 5.24 2.01 x 631 13.15 527.401 5.69 6.26 225 0.48 0.17 0.31 0.91 5,31 24 CV 5.00 4.49 3.78 6.63 4.19 5.84 10.61 5.96 15.94 5.63 2.85 9.23 var. bertheloti M #08 12.71 5.68: 74.53 5.85 6.47 2.06 m 5.45 9.41 4.52 3.08 4.34 4.50 0.96 X 6.15 10:95 5.03 3.73 4.99 5.58 1:33 0.56 0.22 0.34 0.90 8.38 28 CV 5.84 6.46 5.60 7.71 6.39 (Ge 31288 3:69 13:75 5.52 5.20 10:82 var. fortunata M 6.34 14.30 5.19 4.20 5.97 7:05 2295 m 5.40 12.07 4.57 3.42 4.93 5.74 1.83 x 5.94 13.21 4.90 3.92 5.47 6.49 2.44 0.45 0.15 0.30 0.84 5.48 15 CV 3.83 4.02 3.01 3.52 5.09 4.20 8.37 3.49 12.52 5:16: 3.55 7.92 var. beata M 6.06 12.82 4.86 3.60 5.49 5.1.0 2.47 m 4.71 10.07 3.92 2.87 4.28 4.73 1:33 X SAN eG: (42d = Sul 7 4.65 5:22 1.79 0.46 0.17 0.28 0.89 6.42 12 CV 4.96 5.49 4.80 5.53 5.32 4.92 15.46 5.927 13.01. 24.625.147 12.47 var. subhispidula M 7.07 11.99 562 4.18 5.35 6.43 2.19 m 5.41 9575 4.38. 3.21 4.44 4.87 ти x 6:32 11109. 5187 3.80 4.95 5:62 1.65 0.57 0.23 0.34 0.88 6.87 17 CV 5.48 3.98 4.56 4.89 4.41 5.47 11.81 3:10 8.62 3.75 3.35 10.29 var. lanosa M 5.40 8.98 4.44 3.44 4.18 4.66 0.39 m 4.61 7.56 3.70 2.74 3.24 3.81 0.05 X 4.96 8.22 4.04 3.09 8.78 4.30 0.21 0.61 0.23. 0.38 0.88. 53:73 13 CV 4.61 4.13 4.09 5.25 5.58 5.89 40.03 4.38 12.91 3.10 4.88 49.81 Conjunto de todas las variedades M fall 15.33. 5.68 4:65 6.26 755 3.36 т 4.61 7.56. 13.100! 274 3.24 3.81 0.05 x 5:98 1147 4:87 3.69 5.02 5.67 dora 0.52 0.20 0.33 0.89 12.26 109 CV 8.76 12.06 854 9.69 10.56 11.03 36.33 10.60 18.82 8.66 4.50 80.68 314; Pfeiffer, 1848: 209; 1868: 260; 1876: 294-295 [partim]; Deshayes, in Férussac 8 Deshayes, 1851: 372-373; Mabille, 1884: 69-70; Krause, 1895: 25, fig. 5; Kraepelin, 1895: 9 [partim (loc. = Güimar)]; Shuttleworth, 1975: lam. 2, fig. 6 Helix bertheloti Férussac, 1835: 90; d’Or- bigny, 1839: 65-66, lám. 2, figs. 4-6; Pfeiffer, 1876: 295; Mabille, 1884: 81. Helix fortunata Shuttleworth, 1852a: 141; 1975: lám. 2, fig. 4; Pfeiffer, 1853: 162; 1868: 260; 1876: 296; González Hidalgo, Helix (Gonostoma) hispidula,— Albers, 1860: 92; Mousson, 1872: 62-63 [partim]. Helix (Ciliella) lanosa Mousson, 1872: 61-62, lám. 3, figs. 34-36; Pfeiffer, 1870-76: 83, lám. 122, figs. 34-36; 1876: 273-274; Mabille, 1884: 67; Tryon, 1887: 223, lám. 53, figs. 30-32. Helix (Gonostoma) hispidula subhispidula Mousson, 1872: 63. Helix (Gonostoma) bertheloti,— Mousson, 1872: 63-64. Helix (Gonostoma) fortunata,— Mousson, 1872: 64 [partim]; Wollaston, 1878: 389- 1869: 37-38; Smith, 1884: 276; Mabille, 390 [partim]. 1884: 81-82 [partim]; Krause, 1895: 25. Helix (Hispidella) lanosa,— Wollaston, 1878: Helix berthelotii,— Gray, 1854: 11. 384-385. 118 IBÁÑEZ, PONTE-LIRA 8 ALONSO Helix (Gonostoma) beata Wollaston, 1878: 390-391; Mabille, 1884: 85. Helix everia Mabille, 1882: 147; 1884: 71, lám. 17, fig. 13. Helix (Anchistoma) hispidula subhispidula,— Tryon, 1887: 122. Helix (Anchistoma) everia,— Tryon, 1887: 123, lám. 38, figs. 4-6. Helix (Anchistoma) fortunata,— Tyron, 1887: 123, Im. 24, figs. 55-57. Helix (Caracollina) beata,— Tryon, 1887: 123. Hygromia (Ciliella) lanosa,— Pilsbry, 1895: 276; Gude, 1896: 18. Helicodonta (Caracollina) hispidula sub- hispidula,— Pilsbry, 1895: 288; Gude, 1896: 19. Helicodonta (Caracollina) everia,— Pilsbry, 1895: 289. Helicodonta (Caracollina) fortunata,— Pils- bry, 1895: 289; Gude, 1896: 19 [partim]. Helicodonta (Caracollina) beata,— Pilsbry, 1895: 289; Gude, 1896: 2: 19. Helicodonta (Caracollina) hispidula everia, — Gude, 1896: 19. Gonostoma hispidula, — 246-247. Gonostoma fortunata,— Boettger, 1908: 247. Helicodonta salteri Gude, 1911: 268. Canariella hispidula,— Hesse, 1918: 107; Odhner, 1931: 86, figs. 37C, 38C, 40A, 41A; Richardson, 1980: 422. Canariella fortunata,— Hesse, 1931: 54-55, lám. 8, fig. 67а-е; Odhner, 1931: 87, figs. 37C, 38D; Richardson, 1980: 422; Git- tenberger 4 Groh, 1986: 222-223. ?Canariella leprosa,— Hesse, 1931: 55, lam. 8, fig. 68a-b. Canariella everia,— Odhner, 1931: 14: 87. Canariella hispidula bertheloti,— Richardson, 1980: 422. Caracollina beata,— Richardson, 1980: 423. Caracollina everia,— Richardson, 1980: 424. Boettger, 1908: Es una especie polimorfa conquiológica- mente, con seis poblaciones diferenciadas, siendo esto un indicio de que se estan ini- ciando diversos procesos de especiación, aunque no hay aislamiento geográfico real entre ellas y el aparato reproductor es muy similar en todas (para su estudio, se han di- secado 64 ejemplares del total de las va- riedades: 13 de la forma típica; 17 de la var. bertheloti; 8 de la var. fortunata; 2 de la var. beata; 12 de la var. subhispidula; 12 de la var. lanosa). Por ello, consideramos que en la fase actual del proceso cada población no llega a alcanzar la categoría subespecífica y utilizamos sus nombres más antiguos para denominarlas, aunque dándoles el rango in- frasubespecífico de variedad. Como puede observarse en su mapa de distribución geográfica (Fig. 39), la var. ber- theloti (representada en el mapa con estre- llas) es la que ocupa una superficie mayor de la isla, en la vertiente Sur de las Cañadas del Teide y de la cordillera dorsal de La Espe- ranza, y también la que admite mayor varia- ción altitudinal (se encuentra entre 100 y 1,625 m de altitud), lo que implica que se encuentra en una gran variedad de biotopos, con vegetación fundamentalmente de piso basal en las zonas bajas y de pinar en las altas, habiéndose recolectado además en jarales y en los escasos enclaves de bosque de laurisilva de esta vertiente de la isla. Su área de distribución se solapa ligeramente con la de la forma típica (var. hispidula: cua- drados blancos, que alcanza altitudes mucho menores, entre 10 y 490 m, con vegetación de piso basal) y entra en contacto con otras dos: la var. subhispidula (círculos) que se situa en las zonas altas cercanas a la dorsal de La Esperanza (en la vertiente sur, entre 1,000 y 1,700 m de altitud y con vegetación de pinar y también de fayal-brezal) y la var. lanosa (cruces), en la vertiente norte, que al- canza hasta la zona norte del macizo mon- tañoso de Anaga (situado en el extremo Nor- deste de la isla), entre 450 y 1,500 m de altitud, con vegetación de laurisilva, fayal- brezal y pinar; en la actualidad, su área de distribución está interrumpida entre la dorsal de La Esperanza y Anaga, debido a la des- trucción por el hombre de su biotopo natural. Finalmente, al norte del área ocupada por la var. hispidula, en la vertiente sur del macizo de Anaga y con biotopos similares, se loca- lizan la var. fortunata (cuadrados negros, en- tre 50 y 550 m de altitud) y la var. beata (triángulos, entre 80 y 350 m de altitud). Material Examinado Material tipo (conchas vacías, de Tene- пе). —Lectotipo (MHNG 1092/28/2, col. La- marck, selec.: E. Ponte-Lira y M. Ibáñez) y un paralectotipo (MHNG 1092/28/1) de Ca- rocolla hispidula. Holotipo de Helicodonta salteri (NHM 1922.8.29.33). Lectotipo (selec.: K. Groh) y 2 paralectotipos de Helix bertheloti (MNHN, com. Webb, coll. Ferussac). Lec- totipo (selec.: K. Groh) y 5 paralectotipos de Helix everia (MNHN, 4, rec. Dr. Vernau y 1, rec. Bourgeau). Lectotipo (selec. E. Ponte- CANARIELLA Y SU POSICIÓN EN HYGROMIIDAE 119 Lira y M. Ibáñez) y 7 paralectotipos de Helix fortunata (NMB 307, Blauner, 1851), de Santa Cruz, y otros 2 (ZMZ 508672/partim, Blauner, 1852) de Santa Cruz. Holotipo de Helix (Gonostoma) beata (NHM 95.2.230, leg. Barón de Paiva), de “Betancuria, Fuerteven- tura” (localidad errónea). Lectotipo (selec.: E. Ponte-Lira у М. Ibáñez) y 9 paralectotipos de Helix (Gonostoma) hispidula subhispidula (ZMZ 508663, leg. Fritsch), de Paso Alto, y otros 6 (ZMZ 50866595, leg. Fritsch, 1863 y 508664/1). Lectotipo (selec.: E. Ponte-Lira y M. Ibáñez) y 1 paralectotipo de Helix (Ciliella) lanosa (ZMZ 506132, Tarnier, 1865). Otro material (todo, de Tenerife). —Var. hispidula.—5 conchas (ZMZ 508668/2, de Taganana y 508666/3, Fritsch, 1862); 1 (ANSP 248295/partim, P. Hesse, ex. Preston, 1912), de Santa Cruz; 15 (DMNH 15436/7, C. L. Richardson y 128654/8, R. Jackson), de Candelaria; 3 (MHNG 984/201); 2 (MNHN, Mauge); 5 (CGH) y 36 (TFMC), de Candelaria; 6 (TFMC, J. F. Guerra, 1953), de la calle En- rique Wolfson, Sta. Cruz. Var. bertheloti.—4 conchas (NHM 1854.9.28.39, Webb y Ber- thelot); 4 (ЕММН 94096, С. D. Nelson), de La Orotava; 2 (FMNH 37784, С. К. Gude, ex. С. S. Parry); 5 (FMNH 37783, С. К. Gude, ex. H. В. Preston); 1 (ZMZ 508669, Tarnier, 1864), de Güimar; 3 (ZMZ 508660, Wollaston, 1860), de La Orotava y Santa Cruz; 2 (ZMZ 508661, Fritsch, 1863); 4 (ZMZ 508658, Blauner, 1852); 1 (ANSP 248296, Hesse, ex. Preston), de СИтаг; 6 (ОММН 128692, В. Jackson), de СИтаг; 5 (ММВ, Blauner, 1851), de Gülmar; 5 (MNHN, Jousseaume, Letellier, 1949 y Bourgeau, 1856); 7 (TFMC, J. M. Fernández, 1965), de la Ladera de Güimar. Var. fortunata.—6 conchas (FMNH 158.206), de Candelaria; 6 (FMNH 158.213), de La Res- balada; 6 (NMW, Melvill-Tomlin, 4 de ellos, de Santa Cruz); 2 (ZMZ 508671, Wollaston, 1870), de Santa Cruz; 4 (ZMZ 508667, Fritsch, 1872); 1 (ANSP 97264, Wollaston); 4 (ANSP 248295/partim, Hesse, ex. Preston), de Santa Cruz; 1 (ANSP 1563, A. D. Brown); 1 (ANSP 1514, A. D. Brown), de “Gran Ca- naria”; 2 (DMNH 151668, R. Jackson); 2 (MHNG, Moricand); 2 (MHNG); 2 (MNHN, De- nis); 45 (TFMC, F. Guerra y J. M. Fernández), del Bco. Tahodio; 4 (TFMC), del Bco. de San- tos; 4 (TFMC), de Las Mesas; 11 (TFMC), de la Ladera de Pino de Oro. Var. beata.—5 con- chas (NHM 1854. 9.28.35, Webb y Berthelot), de Santa Cruz; 8 (MNHN/2, Locard, MNHN/2, М. Delaunay, 1882 y MNHN/4, Bourgeau, 1885); 6 (MNHN, Vernau, 1877- 78), del bosque de Las Mercedes; 1 (MNHN, Jousseaume); 1 (ANSP 5111, A. D. Brown). Var. subhispidula.—6 conchas (FMNH 158.211), de El Palmar; 1 (ZMZ 508670); 2 (ANSP 33232, Wollaston); 4 de “Helix berthe- lot” (МНМ 1854.9.28.34, d’Orbigny). Var. lanosa.—1 concha (TFMC, J. M. Fernández, 1955), del Monte Las Mercedes; 23 (TFMO), de Agua García, S. Diego y Bco. Carnicería. Además, 750 conchas y 243 ejemplares en alcohol, de las seis variedades, recolectados entre los días 5-05-1979 y 15-01-1994, en diversas localidades de la isla (Fig. 39). Forma típica: var. hispidula (Lamarck, 1822) Helicodonta salteri Gude, 1911 Descripciön (Tabla 1, Figs. 2-5, 28) El animal tiene el cuerpo de color gris con manchas mäs oscuras, alargadas, que se dis- ponen en filas longitudinales en el dorso de la cabeza. La concha es aplanado-lenticular, con 4 a 5/2 vueltas de espira, la última an- gulada. La sutura es nítida, el ombligo, grande y la abertura redondeada. El color es marrón claro, sin brillo. Las costulaciones ra- diales son muy suaves en la protoconcha y están algo más desarrolladas en las si- guientes vueltas de espira; en la última son irregulares, debido a que se deforman al rodear bases de pelos, que están presentes en gran número; por ello, en ocasiones se fusionan unas con otras o bien se inte- rrumpen; en el lado ventral son más suaves que en el dorsal. Superpuesta a esta orna- mentación hay otra, formada por crestas es- pirales muy finas y numerosas. La concha está densamente cubierta de pelos perios- tracales finos y largos, sobre y entre las cos- tulaciones, que se desprenden fácilmente junto con el periostraco, quedando en su lugar protuberancias pequeñas. Su longitud varía mucho en cada vuelta de espira. En el lado dorsal son pequeños, menores de 100 um, salvo en la zona de la sutura con la si- guiente o en la periferia de la última, donde son mucho mayores, alcanzando hasta 800 um. En el lado ventral son menores y su lon- gitud disminuye hacia el ombligo, en el que no sobrepasan los 160 um. Mandíbula con más de 15 costillas. La rá- dula tiene la siguiente fórmula: (С + 17—25L) x 95-115. Los dientes cercanos al borde ra- dular tienen el ectocono dividido en dos den- tículos. Aparato reproductor. La distancia del atrio 120 IBÁÑEZ, PONTE-LIRA 8 ALONSO a la inserción del músculo retractor del pene es ligeramente mayor que la del resto del epi- falo, que a su vez es más del doble de largo que el flagelo. Epifalo y flagelo son tubulares y adelgazan paulatinamente hasta el final del flagelo, que termina en forma de dedo de guante (no es puntiagudo). El pene tiene un pequeño estrangulamiento justo antes de desembocar en el atrio, y está ensanchado asimétricamente en su porción proximal, for- mando una pequeña protuberancia. El epi- falo alberga cuatro pliegues longitudinales, que están bien desarrollados entre el flagelo y la inserción del músculo retractor, mientras que hacia la zona distal uno de ellos se va atenuando y termina desapareciendo antes de la unión con el pene. De los tres que quedan, uno termina en el extremo proximal del pene y los otros dos se prolongan en su cavidad y se fusionan formando una papila acanalada, situada en el mismo lado que la inserción del músculo retractor y cuya lon- gitud equivale a casi la mitad del pene. A continuación de ella, el pene presenta 4-5 boceles cortos y gruesos, que lo recorren longitudinalmente hasta su inserción en el atrio. La vagina es tubular y en su interior posee 5 a 6 boceles longitudinales, que no conectan con los pliegues del conducto de la bolsa copulatriz. Un poco por debajo del orificio de comunicación con el oviducto, desembocan en ella un número variable de glándulas vaginales digitiformes (de 3 a 8), dispuestas formando un círculo a su alrede- dor. El nervio peneano se origina del ganglio cerebroideo derecho. var. bertheloti (Férussac, 1835) (Tabla 1, Figs. 6-7, 29). Helix everia Mabille, 1882 La concha es deprimida y algo más pe- queña, con la espira menos angulada y el ombligo más pequeño. Las costulaciones ra- diales de la teloconcha están más desarro- lladas y más próximas entre sí. El número de pelos periostracales es netamente inferior, variando su tamaño de 200 a 800 um de lon- gitud en el lado dorsal y no pasan de 110 um en el ombligo. Con 3 a 6 glándulas vaginales. var. fortunata (Shuttleworth, 1852) (Tabla 1, FIGs: 8, 25, 30): La concha es más aplanada, con la espira muy angulada y la abertura ovalada. Los pe- los periostracales están dispuestos desor- denadamente; los de la periferia alcanzan hasta 800 um de longitud, mientras que en el ombligo no sobrepasan los 100 um. Con ба 7 glándulas vaginales. var. beata (Wollaston, 1878) (Tabla 1, Figs. 9, Sil): La concha es mas aplanada y algo mas pequena, con la espira aquillada, la abertura ovalada y el ombligo más pequeño. Las cos- tulaciones radiales son muy numerosas en la teloconcha. En los ejemplares mejor conser- vados, la pilosidad es muy escasa y se loca- liza en la periferia de la concha, siendo los pelos muy cortos (no pasan de 350 um de longitud); los del ombligo son todavía más cortos, alcanzando hasta 100 um. Los dos reproductores examinados únicamente po- seen dos glándulas vaginales. Observaciones El holotipo (Fig. 9) de Helix beata (que es el único ejemplar sobre el que se basó Wollas- ton para describirla) tiene el lado dorsal de la concha muy aplanado, existiendo otros ejemplares con él más alto. Es probable que el error en el dato de la localidad típica (““Be- tancuria, Fuerteventura”) se deba a un cam- bio de etiqueta o bien de la caja en que el ejemplar estuviera almacenado, desde que fue recolectado por el Barón de Paiva hasta que fue estudiado por Wollaston. var. subhispidula (Mousson, 1872) (Tabla 1, Figs. 14, 32). La concha es deprimida y algo más pe- queña, con la espira ligeramente angulada, la abertura ovalada y el ombligo más pequeño. Las costulaciones radiales de la teloconcha están bien desarrolladas. Los ejemplares mejor conservados presentan una pilosidad muy escasa, situada en el lado dorsal de la última vuelta de espira y en la periferia de la concha, normalmente en los espacios inter- costales. Los más largos son los de la pe- riferia, que alcanzan hasta 700 um de longi- tud; los del ombligo, en cambio, son muchísimo más cortos, no pasando de 60 um. Con 4 a 6 glándulas vaginales. var. lanosa (Férussac, 1835) (Tabla 1, Figs. 15, 26, 33): La concha tiene forma cónico-ovalada y es bastante más pequeña, con la espira redon- deada. El ombligo es muy pequeño, estando en algunos ejemplares casi totalmente cu- CANARIELLA Y SU POSICIÓN EN HYGROMIIDAE 121 bierto por el peristoma, y la abertura es ova- lada. Toda la concha está cubierta de pelos periostracales finos y largos, que en el lado dorsal de cada vuelta son menores de 170 um, salvo en la zona de la sutura con la si- guiente o en la periferia de la última vuelta, en donde están los más grandes, que llegan a alcanzar cerca de 550 um de longitud. En el interior del ombligo, no sobrepasan los 100 um. Con 3 a 5 glándulas vaginales. Observaciones Es la población que más se diferencia con- quiológicamente de las otras, evidenciando que su proceso de especiación es el más avanzado. Canariella discobolus (Shuttleworth, 1852) Helix discobolus Shuttleworth, 1852b: 290 [loc. typ.: Gomera]; 1975: lam. 6, fig. 6; Pfeiffer, 1853: 643; 1868: 260; 1870-76: 84, lám. 123, figs. 1-2; 1876: 296. Helix afficta Ferussac,—d’Orbigny, 1839: 66, lám. 3, figs. 24-26; Gray, 1854: 11 [non H. afflicta Férussac, 1821]. Helix (Gonostoma) discobolus,—Albers, 1860: 92; Mousson, 1872: 66, lám. 4, figs. 1-2. Helix (Anchistoma) discobolus,—Tryon, 1887: 123, lám. 24, figs. 41-42. Helicodonta (Caracollina) discobolus, —Pils- bry, 1895: 289; Gude, 1896: 19. Canariella discobolus,—Hesse, 1931: 55-56, lám. 9, fig. 69a-b. Canariella discobola,— Richardson, 1980: 422. Material Examinado 1 concha (NHM 1854.9.28.46/partim, leg. Webb y Berthelot). Además, 64 conchas y 7 ejemplares en alcohol, recolectados en di- versas localidades, entre los días 29-07- 1982 y 30-01-1989. Hábitat y Distribución (Fig. 43) Endémica de La Gomera. Se distribuye en la zona sur de la isla, entre 175 y 800 m de altitud. Está ligada a zonas pedregosas con vegetación de piso basal. Descripción (Tabla 2, Figs. 12, 34) El animal fijado tiene el cuerpo de color blanquecino. La concha es aplanado-lenti- cular, con 6 a 6% vueltas de espira, las pri- meras anguladas y las últimas aquilladas. La sutura es nítida, el ombligo grande y la aber- tura ovalada, angulada en la unión de las zo- nas palatal y basal. El color es marrón claro, sin brillo. Las costulaciones radiales son muy suaves en la protoconcha (donde se inte- rrumpen dando la apariencia de una granu- lación) y en las primeras vueltas de espira, y están bien marcadas en las siguientes, siendo irregulares en la última. En el lado ventral son más regulares y más gruesas que en el dorsal, disminuyendo su grosor en las proximidades del ombligo y de la periferia. La pilosidad está restringida a la zona de la pe- riferia, donde los pelos periostracales alcan- zan hasta 500 um de longitud, y al ombligo, donde son mucho más numerosos y cortos, no pasando de 45 um. Mandíbula con 10 a 13 costillas. La rádula tine la siguiente fórmula: (С + 18-30L) x 98-130. Los dientes cercanos al márgen radular tienen el ectocono dividido de forma irregular. Aparato reproductor (se han disecado 3 ejemplares): la distancia del atrio a la inser- ción del músculo retractor del pene es igual que la del flagelo y menor que la del resto del epifalo. Las tres regiones son tubulares y van adelgazando paulatinamente desde el atrio hasta el final del flagelo. El epifalo alberga en su interior cuatro pliegues longitudinales, de los que dos terminan en su extremo distal y los otros dos se prolongan en la cavidad del pene; éstos, en dos de los reproductores examinados, engruesan ligeramente y termi- nan juntos, pero independientes entre sí, mientras que en el tercero se fusionan for- mando una pequeña papila acanalada, situa- da aproximadamente en el mismo lado que la inserción del músculo retractor. A conti- nuación de ella, el pene presenta cuatro bo- celes (dos más gruesos) que lo recorren lon- gitudinalmente hasta su inserción en el atrio. La vagina es gruesa y de tamaño similar al del pene; en ella desembocan dos glándulas vaginales digitiformes de tamaño variable y en su interior posee 6-7 boceles longitudina- les delgados. Su pared presenta un engro- samiento muy desarrollado con aspecto al- mohadillado, que ocupa casi toda su cavidad y está dividido en cuatro lóbulos por un surco longitudinal y otro transversal. Entre los dos lóbulos proximales, cerca de su extremo, se encuentra el orificio de comunicación con el oviducto. El inicio del conducto de la bolsa copulatriz presenta internamente gran nú- mero de boceles con bordes irregulares. 122 IBÁÑEZ, PONTE-LIRA & ALONSO FIGS. 12-20. Concha y SEM detalles. (12) Canariella discobolus (Barranco de la Rajita, La Gomera). (13) Canariella gomerae. Lectotipo de Helix (Gonostoma) gomerae (NHM; es un ejemplar pequeño dentro de la especie). (14-15) Canariella hispidula. (14) Lectotipo de Helix (Gonostoma) hispidula subhispidula (ZMZ). (15) Lectotipo de Helix (Ciliella) lanosa (ZMZ). (16) Canariella leprosa (El Draguillo, Tenerife). (17-18) Cana- riella eutropis. (17) Lectotipo de Helix eutropis (NMB). (18) Mandíbula de un ejemplar de Morro del Cava- dero, Fuerteventura). (19-20) Rádula de Canariella planaria (Benijo, Tenerife). (19) Diente central y primeros dientes laterales. (20) Dientes laterales próximos al margen radular. Escala: (12-17) 5 mm; (18) 200 um; (19-20) 20 um. CANARIELLA Y SU POSICIÓN EN HYGROMIIDAE 123 TABLA 2. Datos biométricos (dimensiones en mm, e índices) de la concha de las especies de Canariella. Símbolos, como en la Tabla 1. A B С D E F Canariella discobolus M Sill 22021025 Siew 5.56: MAN 978 m 6.62 16.08 5.14 4.21 5.67 3.56 x 7.16 18:14 5.61 5.03 6.41 7.93 CV 468 478 345 703 543 9.52 Canariella gomerae M 6.55 13.79 5.16 4.53 5.69 6.67 т 5.22 11.51 4.33 2.93 4.59 5.47 X 5.89 12.54 4.74 3.80 5.01 5.91 CV 4.07 4.29 5.28 7.30 5.83 4.58 Canariella planaria M 5.79 14.95 4.53 4.37 6.23 7.05 m 4.11 12.70 3.24 2.76 4.53 5.69 X 5.13 1396 4.03 3.44 5.48 6.31 CV 5.07 2.98 5.50 8.84 4.96 4.74 Canariella leprosa M 4.68 9.18 3.94 3.48 3.91 5.00 m 4.25 7.97 3.49 2.71 3.50 3.90 X 4.56 8.58 3.70 3.07 3.80 4.59 СУ 3.35 5.16 5.60 9.19 3.95 7.54 Canariella eutropis M 8.84 16.52 7.47 6.47 7.27 8.06 т 7.32 13.82 5.88 4.83 5.08 6.47 X 8.00 15.00 6.74 565 6.43 6.99 CV 4.33 2.81 4.36 5.70 3.95 4.06 Сапапе!а дотегае (Wollaston, 1878) Helix (Gonostoma) дотегае Wollaston, 1878: 392-393 [loc. typ.: Hermigua, La Go- mera]; Mabille, 1884: 84-85. Helicodonta (Caracollina) gomerae,—Pilsbry, 1895: 289; Gude, 1896: 19. Caracollina gomerae,—Richardson, 424. 1980: Material Examinado Material tipo (conchas vacío).—Lectotipo (selec.: E. Ponte-Lira y М. Ibanez) y 3 para- lectotipos de Helix (Gonostoma) gomerae (МНМ 95.2.216-19), de Hermigua (La Go- mera); y otro paralectotipo (FMNH 37781, colección de Gerard K. Gude, ex G. S. Parry; juvenil) de Hermigua. Otro material. —1 juve- nil (NMW; Melvill-Tomlin collection). Además, 84 conchas y 19 ejemplares en alcohol, re- colectados entre los días 3-01-1978 y 2-01- 1993, en diversas localidades de la isla. Hábitat y Distribución (Fig. 43) Endémica de La Gomera. Se distribuye por la zona norte de la isla, entre 100 y 1,200 m de altitud, en varios tipos de vegetación, desde tabaibales a fayal-brezal y laurisilva. G A/B H/B D/B E/F B/G n 4.94 2.95 3.95 0.40 0.12 0.28 0.85 4.65 17 9.95 5.65 26.42 6.36 14.52 7.88 2.73 2.07 2.36 0.47 0.17 0.30 0.85 5.85 11 9.36 2.24 16.21 7.96 2.30 8.93 3.29 РУ 2.78 0.37 0.12 0.25 0.87 5.07 32 8.29 4.52 18.17 8.31 4.78 8.94 0.30 0.20 0.25 0.53 0.17 0.36 0.83 35.88 4 20.00 2.49 15.89 4.05 4.09 21.81 3.00 1.62 2.22 0.53 0.16 0.38 0.92 6.89 29 12.03 4.26 16.96 4.68 4.80 11.29 Descripción (Tabla 2, Figs. 13, 21, 35) El animal fijado tiene el cuerpo de color blanquecino. La concha es aplanado-lenti- cular, con 5/2 a 6 vueltas de espira, las pri- meras anguladas y las últimas aquilladas. La sutura es nítida, el ombligo grande y la aber- tura ovalada, angulada en la unión de las zo- nas palatal y basal. El color es marrón claro, sin brillo. Las superficies dorsal y ventral están provistas de costulaciones radiales suaves y uniformes. Superpuesta a esta or- namentación hay otra, formada por crestas espirales muy finas y numerosas. Además, sobre las costulaciones y los espacios que hay entre ellas existen, en la última vuelta, pequeños gránulos redondeados, que son más patentes en la superficie ventral y posiblemente corrsponden a la base de los pelos periostracales. La pilosidad está res- tringida a la zona de la periferia, donde son finos y muy poco abundantes en los ejem- plares adultos, alcanzando hasta 400-450 um de longitud, y al ombligo, donde son mu- cho más numerosos y cortos (menores de 35 um). Mandíbula con más de 13 costillas. La rá- dula tiene la siguiente fórmula radular: (C + 25 — 31L) x 97 — 125. Los dientes cercanos 124 IBÁÑEZ, PONTE-LIRA & ALONSO al margen radular tienen el ectocono a veces dividido en dentículos. Aparato reproductor (se han disecado 3 ejemplares): la distancia del atrio a la inser- ción del músculo retractor del pene es ligera- mente inferior a la del resto del epifalo y mayor que el flagelo. Este, es largo y muy esbelto, con un diámetro ligeramente supe- rior al del conducto deferente. El pene es casi cilíndrico y está engrosado en su extremo distal, disminuyendo paulatinamente su diá- metro en sentido proximal. El epifalo alberga en su interior a cuatro pliegues longitudina- les, de los que dos terminan en el extremo proximal del pene y los otros dos se prolon- gan en su cavidad, fusionándose en una papila acanalada pequeña (menor que 1/3 de la longitud del pene), situada en el mismo lado que la inserción del músculo retractor. A continuación de ella, el pene presenta cuatro boceles alargados que lo recorren longitudi- nalmente hasta su inserción en el atrio. La vagina es gruesa y más corta que el pene y en su porción proximal desembocan, a la misma altura, tres glándulas vaginales digiti- formes; dos de ellas son de igual tamaño y la tercera es más corta. Internamente, la vagina posee 3-4 boceles longitudinales. En la parte proximal, justo debajo del orificio de comu- nicación con el oviducto, presenta un engro- samiento almohadillado de su pared, dividi- do en dos lóbulos por un surco longitudinal. El inicio del conducto de la bolsa copulatriz tiene en su interior gran número de boceles con bordes irregulares. Canariella planaria (Lamarck, 1822) Carocolla planaria Lamarck, 1822: 99 [loc. typ.: Tenerife, hic. restr.: Vertiente Norte de Tenerife, entre la Punta de Juan Blas y la Punta de Anaga]; 1838: 148; Mer- mod, 1951: 712-713, fig. 67 [paralec- totipo]. Helix afficta Férussac, 1821: 41 (Folio) o 37 (Quarto), n° 151 [nomen nudum; ICZN, Art. 12]; Férussac, in Férussac & Des- hayes, 1832 (Atlas), lám. 66*, fig. 5; Pfei- ffer, 1848: 211; 1853: 162; 1868: 260; Deshayes, in Férussac 4 Deshayes, 1851: 372; Tryon, 1887: 122, lám. 24, figs. 52-54 [partim]; Chevallier, 1965: 489. Helix afficta planaria,—Mousson, 1872: 65- 66; Pfeiffer, 1876: 296. Helix planaria,—Wollaston, 1878: 391-392; Mabille, 1884: 82-84; Tryon, 1887: 122, lám. 24 figs. 58-60 [partim]. Helicodonta planaria,—Gude, 1896: 19. Caracollina planaria,—Richardson, 1980: 425. Material Examinado Material tipo (conchas vacías, de Tene- rife).—Lectotipo (MHNG 1092/27/1, col. La- marck, selec.: E. PonteLira y M. Ibáñez) y un paralectotipo (MHNG 1092/27/2) de Ca- rocolla planaria. Lectotipo (selec.: K. Groh) y 3 paralectotipos de Helix (Helicigona) afficta Férussac, 1832 (MNHN; col. Férussac). Otro material.—3 conchas (MNHN); 4 (NMW); 4 (ZMZ 508677/2 y 508676/2) de Taganana; 1 (ZMZ 508678); 3 (ЕММН 37785) y otras 2 (FMNH 37782) de Almáciga; 2 (DMNH 78297) de Guayonga; 7 (MHNG); 1 (NHM 1854.9.28.46/partim); 1 (ANSP 397001), de Montaña Tafada; 2 (ANSP A17996 y 397002) de Benijo. Además, 283 conchas y 37 ejem- plares en alcohol, recolectados entre los días 3-03-1981 y 8-11-1993, en diversas loca- lidades. Hábitat y Distribución (Fig. 40) Endémica de Tenerife. Habita en la ver- tiente norte de la isla, desde Punta de Juan Blas a Punta de Anaga, entre 5 y 800 m de altitud. Vive en zonas con vegetación muy diversa, normalmente en tabaibales. Descripción (Tabla 2, Figs. 10, 11, 19, 20, 36) El cuerpo es de color gris claro con man- chas más oscuras, alargadas, dispuestas en filas longitudinales en el dorso de la cabeza. La concha es aplanado-lenticular, con 5 a 52 vueltas de espira, las primeras anguladas y las últimas aquilladas. La sutura es nítida, el ombligo grande y la abertura ovalada, angu- lada en la unión de las zonas palatal y basal. El color es marrón claro, con un ligero brillo. Las costulaciones radiales son muy suaves en las primeras vueltas de espira y ligera- mente más marcadas en las siguientes; en el lado ventral son más suaves que en el dorsal. Las crestas espirales en el lado ventral son finísimas y muy apretadas mientras que en el dorsal hay en su lugar algunas estrías espi- rales profundas, poco numerosas y espacia- das entre sí. La pilosidad está restringida a la zona del ombligo, donde los pelos son muy numerosos y cortos. Mandíbula con más de 20 costillas. La rá- _ — nn CANARIELLA Y SU POSICIÓN EN HYGROMIIDAE 125 FIGS. 21-27. SEM detalles de la concha. (21-23) Protoconcha y primeras vueltas de espira. (21) Canariella gomerae (Aguajilva, La Gomera). (22) Canariella eutropis (Morro del Cavadero, Fuerteventura). (23) Cana- riella leprosa (El Draguillo, Tenerife). (24) С. leprosa. Detalle de la última vuelta, lado dorsal. (25) C. hispidula var. fortunata (Cabezo de las Mesas, Tenerife). Detalle del ombligo. (26) Canariella hispidula var. lanosa (Agua García, Tenerife). Detalle de la penúltima y última vueltas. (27) C. eutropis (Morro del Cavadero). Quilla y ornamentación (última vuelta, lado dorsal). Escala: (21, 27) 500 um; (22, 23) 1 mm; (24, 25) 125 um; (26) 250 um. 126 dula tiene la siguiente fórmula: (С + 27 — 31L) x 100 — 130. Los dientes cercanos al márgen radular poseen el ectocono dividido general- mente en dos dentículos de diferente ta- maño. Aparato reproductor (se han disecado 10 ejemplares): la distancia del atrio a la inser- ción del músculo retractor del pene es 212 veces mayor que la del resto del epifalo. El flagelo es muy delgado y muy corto, casi ru- dimentario. El epifalo alberga en su interior cinco pliegues longitudinales (de los que dos son muy pequeños) que normalmente finali- zan en su extremo distal; en un ejemplar, es- tos pliegues conectan con los boceles del pene, aunque estrechándose en la zona de contacto. En el interior del pene hay 7-8 bo- celes longitudinales, algunos de ellos anas- tomosados. El que está en el mismo lado que la inserción del músculo retractor, tiene en su extremo proximal un engrosamiento papili- forme muy nítido (no existe la papila peneana típica de las otras especies), y el situado en el lado opuesto engruesa considerablemente en posición distal. La vagina alberga en su interior un número variable de boceles longi- tudinales, algunos anastomosados, que se prolongan sin interrupción (estrechándose) en el conducto de la bolsa copulatriz. En su zona media desemboca una glándula vaginal digitiforme pequeña (como excepción, en un ejemplar encontramos dos glándulas). Canariella leprosa (Shuttleworth, 1852) Helix leprosa Shuttleworth, 1852a: 142 [loc. typ.: Tenerife, hic restr.: zona norte de Anaga]; non Canariella leprosa,—Hesse, 1931: 55, lam. 8, fig. 68a-b]; 1975: pl. 3, fig. 10; Pfeiffer, 1853: 130; 1870-76: 82- 83, 14m: 122, figs. 31=33; 1876: 273; ? Mabille, 1884: 66-67. Helix (Ciliella) leprosa,—Mousson, 1872: 61, lám. 3, figs. 31-33; Tryon, 1887: 223, lám. 53, figs. 33-35. Hygromia (Ciliella) leprosa, —Pilsbry, 1895: 276; Gude, 1896: 18. Helix (Hispidella) leprosa,— Wollaston, 1878: 383-384. Material Examinado Una concha (ZMZ 506131; rec. Tarnier, 1865) de “Tenerife.” Además, 24 conchas y 17 ejemplares en alcohol, recolectados entre IBÁÑEZ, PONTE-LIRA 8 ALONSO los días 3-03-1981 y 15-01-1993, en diversas localidades. Hábitat y Distribución (Fig. 41) Endémica de Tenerife. Se distribuye por la zona norte de la isla, entre 400 y 800 т de altitud. Está ligada generalmente a bosques de laurisilva y fayal-brezal, habiéndose re- colectado también en piso basal. Descripción (Tabla 2, Figs. 16, 23, 24, 37) El animal fijado tiene el cuerpo de color blanquecino. La concha es deprimida, con 4 3/4 a 5 vueltas de espira. La sutura es nítida, la abertura ovalada y el ombligo muy pe- queño y casi completamente tapado por el peristoma. El color es marrón, sin brillo, pu- diendose desprender el periostraco parcial- mente, incluso en los animales vivos. Las costulaciones radiales son pequeñas, no equidistantes, muy suaves (prácticamente inexistentes) en la protoconcha y en las pri- meras vueltas de espira y bien marcadas en las siguientes. A partir de la tercera vuelta, tienen una serie de interrupciones, mucho más patentes en la última, donde son susti- tuidas por filas de gránulos alargados que dan a la concha un aspecto muy caracterís- tico, más acusado en el lado ventral. En las proximidades del ombligo, los gránulos están atenuados y desaparecen casi por completo. Superpuesta a esta ornamenta- ción hay otra, formada por crestas espirales muy finas y numerosas, que se han desgas- tado sobre los gránulos dorsales. La pi- losidad está restringida a la zona de la pe- riferia, donde son finos y escasos, alcanzando hasta 360 um de longitud. Aparato reproductor (se han disecado 5 ejemplares): El atrio es corto; la distancia desde él hasta la inserción del músculo re- tractor del pene es alrededor del triple que la del resto del epifalo. El flagelo es muy del- gado y muy corto, casi rudimentario. El epi- falo alberga en su interior tres pliegues lon- gitudinales, que en su extremo distal se fusionan en un grueso bocel del pene, que alcanza más de las 3/4 partes de la longitud del pene. La base de este bocel está unida a todo el perímetro interno del pene, realizán- dose la comunicación entre el epifalo y el pene a través de un pequeño conducto que - atraviesa la base del bocel y se abre inmedia- tamente después, en su zona proximal. El resto de la pared del pene está ornamentado CANARIELLA Y SU POSICIÓN EN HYGROMIIDAE 127 con otros seis boceles, mucho más delga- dos. La vagina presenta en su pared interna cuatro boceles longitudinales finos, que no conectan con los pliegues (más finos y nu- merosos) del conducto de la bolsa copulatriz. Una única glándula vaginal digitiforme, grande y relativamente gruesa, desemboca cerca del extremo distal de la vagina. Observaciones Conquiológicamente es similar a C. ptho- nera por el tamaño y la ornamentación de la concha, pero ambas se diferencian clara- mente por el ombligo y la forma del la aber- tura (C. pthonera tiene ombligo grande y abertura redondeada). Con respecto a la anatomía del aparato reproductor, se dife- rencia claramente de las demás especies del género por la sustitución de la papila del pene por un grueso bocel perforado en su base. Mousson (1872) creó el género Ciliella para agrupar, junto a Ciliella ciliata (Studer), dos especies de Canarias: Helix leprosa y Helix lanosa. Basándose probablemente en Mous- son, varios autores (entre ellos, Germain, 1930; Thiele, 1931; Zilch, 1960; y Schileyko, 1991) indican que Ciliella se encuentra en Eu- ropa y en Canarias. Pero Ciliella carece de glándulas vaginales, que están presentes en las dos especies de Canariella (H. leprosa y H. lanosa). Podemos afirmar, por tanto, que Ciliella no tiene representates en el Archi- piélago. Canariella eutropis (Shuttleworth, 1860) Helix eutropis Shuttleworth, in Pfeiffer, 1860: 237 [loc. typ.: montes de Jandía, Fuer- teventura (figura en la etiqueta del mate- rial tipo)]; Albers, 1860: 139; Pfeiffer, 1868: 371; 1870-76: 81, lám. 122, figs. 28-30; Mousson, 1872: 58-59, lám. 3, figs. 28-30; Wollaston, 1878: 380; Ma- bille, 1884: 89-90; Tryon, 1888: 36, lám. 7, figs. 16-18; Pilsbry, 1895: 289; Gude, 1896: 19. Material Examinado Material tipo (conchas vacías).—Lectotipo (selec.: E. Ponte-Lira у М. Ibáñez) y 1 para- lectotipo de Helix eutropis (NMB 139, “Dr. Bolle detexit’’), de los montes de Jandia. Otro material.—2 conchas (FMNH 37788); 3 (ANSP 397000/2 y A17995) del Morro del Ca- vadero (Fuerteventura). Además, 54 conchas y 114 ejemplares en alcohol, recolectados entre los días 18-08-1986 y 8-03-1990, en diversas localidades de los montes de Jan- día. Hábitat y Distribución (Fig. 44) Endémica de Fuerteventura. Habita en las zonas más húmedas de los montes de Jan- día, entre 250 y 800 m de altitud. Descripcion (Tabla 2, Figs. 17, 18, 22, 27, 38) El cuerpo es de color gris con manchas más oscuras, alargadas, que se disponen en filas longitudinales en el dorso de la cabeza. La concha carece de pilosidad; es deprimida, aplanada en el lado dorsal y ovalada en el ventral, con 42 a 514 vueltas de espira; a partir de la segunda, está provista de una quilla muy patente. La sutura es nítida, el om- bligo grande y la abertura redondeada, aun- que angulada en la unión de las zonas palatal y basal. El color es marrón amarillento, sin brillo, existiendo normalmente tres bandas espirales ligeramente más oscuras, dos dor- sales y una ventral. Las costulaciones radia- les son muy suaves en las primeras vueltas de espira, transformándose en costillas muy marcadas y distanciadas entre sí en las si- guientes. En el lado ventral son más numero- sas (alrededor de 45-50, frente a 35-37 del lado dorsal de la última vuelta). Algunas cos- tillas del lado dorsal y todas las del ventral se prolongan sobre la quilla, que además tam- bién tiene otras protuberancias propias. Las crestas espirales son muy finas y numerosas, estando desgastadas sobre la quilla. El collar del manto difiere de las demás especies descritas del género por la presen- cia del lóbulo lateral izquierdo, muy fino y poco conspicuo, que está situado en posi- ción opuesta al pneumostoma. Mandíbula con 3 a 9 costillas. La rádula tiene la siguiente fórmula: (С + 28-32L) x 130-140. Los dientes cercanos al borde ra- dular tienen el ectocono dividido en un nú- mero variable de dentículos, que le dan un aspecto aserrado. Aparato reproductor (se han disecado 4 ejemplares): la distancia del atrio a la inser- ción del músculo retractor del pene es similar a la del resto del epifalo y doble a triple que la del flagelo. El epifalo alberga en su interior cinco pliegues longitudinales delgados, lisos 128 IBÁÑEZ, PONTE-LIRA & ALONSO FIGS. 28-38. Aparato reproductor y detalles (escala: 1 mm); a: visión general; b: anatomía interna de la vagina y zona distal del conducto de la bolsa copulatriz; c: anatomía interna del pene y zona distal del epifalo; d: sección longitudinal del pene y transversales del pene y del epifalo (sin escala); e: disposición de las glándulas vaginales alrededor de la vagina; f: pene evaginado, mostrando la papila acanalada (*) (28) Canariella hispidula var. hispidula (Tabaiba Alta, Tenerife). (29) C. hispidula var. bertheloti (Arafo, Tenerife). (30) С. hispidula var. fortunata (Cabezo de las Mesas, Tenerife). (31) С. hispidula var. beata (Lomo Bermejo, Tenerife). (32) C. hispidula var. subhispidula (Montaña del Cascajo, Tenerife). (33) C. hispidula var. lanosa (Agua García, Tenerife). (34) Canariella discobolus (Barranco de La Rajita, La Gomera). (35) Canariella gomerae (Aguajilva, La Gomera). (36) Canariella planaria (Playa Fabián, Tenerife). (37) Canariella leprosa (Monte Tenejías, Tenerife). (38) Canariella eutropis (Morro del Cavadero, Fuerteventura). en su porción proximal y ondulados desde la zona de la inserción del músculo retractor hasta el pene. Todos ellos se fusionan entre sí en su extremo distal, prolongándose en una papila peneana que en su base está unida a todo el perímetro interno del pene, cerrando completamente la luz del conducto (salvo su propio conducto interno), y en su porción libre es acanalada. Esta papila ocupa aproximadamente la mitad de la longitud del CANARIELLA Y SU POSICIÓN EN HYGROMIIDAE 129 FIGS. 32 y 33. pene y de su base nace un grueso bocel lon- gitudinal, situado en el mismo lado que la inserción del músculo retractor. Este bocel llega casi hasta el extremo distal del pene y su superficie está plegada transversalmente. Al evaginarse el pene, el bocel forma una protuberancia longitudinal, que se dispone inmediatamente detrás de la papila y au- menta considerablemente el grosor del pene evaginado; probablemente su función es evi- tar que éste se separe accidentalmente del otro individuo durante la cópula. Finalmente, el resto de la pared del pene está ornamen- ES PET 7 et а р — к 19% / FIGS. 34 y 35. tado con pequeños pliegues irregulares, ge- neralmente transversales y a veces anasto- mosados entre sí. La vagina está recorrida en su interior por cuatro boceles longitudinales finos; cerca de su extremo proximal desem- bocan en ella 2-3 glándulas vaginales digiti- formes pequeñas y relativamente gruesas. El 130 IBÁÑEZ, PONTE-LIRA & ALONSO FIGS. 36 y 37. inicio del conducto de la bolsa copulatriz está provisto internamente de gran número de pliegues irregulares, anastomosados en- tre sí. DISCUSIÓN El género Canariella fue situado inicial- mente por Hesse (1918) en la subfamilia He- licodontinae, posición que ha sido mantenida por diversos autores hasta la actualidad (Hesse, 1931, 1934; Odhner 1931; Zilch, 1960; Richardson, 1980; Vaught, 1989). Por su parte, los helicondóntidos han sido con- siderados como una subfamilia de Helicidae (Hesse, 1918; Zilch, 1960; Gittenberger, 1968; Kerney et al., 1983) y también como una familia diferente (Helicodontidae), т- cluyédolos en la superfamilia Helicoidea (Damianov 8 Likharev, 1975; Schileyko, 1978) o en Helicondontoidea (Schileyko, 1979). Recientemente, Nordsieck (1987, 1993b) y Schileyko (1991) ubican a Canariella entre los higrómidos, que estaban considerados tradi- cionalmente como una subfamilia de Heli- cidae; pero, en base a una nueva interpreta- ción de sus caracteres anatómicos, fueron segregados como una familia diferente por Schileyko (1973), opinión compartida por di- versos autores (Nordsieck, 1987, 1988, 1993b; Giusti 8 Manganelli, 1987; Puente 4 Prieto, 1992). CANARIELLA Y SU POSICIÓN EN HYGROMIIDAE 131 Canariella hispidula Canariella planaria Canariella leprosa 40 41 FIGS. 39-44. Distribución geográfica. Mapas UTM con cuadrículas de 10 x 10 km, realizados según el procedimiento informático de La Roche 8 Barquín (en prensa). Los símbolos representan cuadrículas de 1 x 1 km; las letras BS, CS, CR y ES designan las cuadrículas de 100 x 100 km. (39-41) Isla de Tenerife. (39) Variedades de Canariella hispidula; cuadrado blanco: var. hispidula; estrella: var. bertheloti; cuadrado negro: var. fortunata; triángulo: var. beata; círculo: var. subhispidula; cruz: var. lanosa (la interrupción principal de su área de distribución se debe a la destrucción por el hombre del biotopo natural). (42) Mapa general del Archipiélago Canario (no está hecho a escala). (43) Isla de La Gomera. (44) Isla de Fuerteven- tura. 132 IBANEZ, PONTE-LIRA & ALONSO Canariella discobolus A Canariella gomerae 3 Canariella eutropis FIGS. 42-44. Los trabajos de Nordsieck y Schileyko han tenido la virtud de reactivar el interés de los malacologos en la sistematica de los Heli- coidea, pero estan basados en ideas perso- nales de sus autores, no confirmadas por técnicas objetivas, lo que ha conducido a re- sultados muy dispares. Nordsieck (1987) co- loca a Canariella en Ciliellinae (junto con los grupos de Oestophora Hesse, 1907, y de Trissexodon Pilsbry, 1895, más los géneros Caracollina Beck, 1837, y Ciliella Mousson, 1872), independiente de Helicodontinae, e incluye a ambas subfamilias en Hygromiidae y a ésta en la superfamilia Helicoidea. Por su parte, Schileyko (1991) incluye a Canariella en Ciliellinae (junto con Ciliella, Haplohelix Pilsbry, 1919, y Schileykiella Manganelli, Sparacio & Giusti, 1989), y a ésta en Ciliel- lidae, junto con Halolimnohelicinae y Vicarii- helicinae; y engloba a Ciliellidae en Hygromi- oidea, superfamilia no reconocida por Nordsieck (1987, 1993b) y tampoco por Giusti & Manganelli (1987, 1988, 1990) y por Manganelli et al. (1989), mientras que sí lo está por Prieto et al. (1993). Los caracteres de mayor relevancia utiliza- dos en ambas clasificaciones son los mis- mos (configuración y posición del aparato estimulador, posición de la bolsa copulatriz con respecto al ovoespermiducto y morfo- logía y posición de las glándulas vaginales), destacando los referentes al aparato estimu- CANARIELLA Y SU POSICIÓN EN HYGROMIIDAE 133 lador y a la topografía de la bolsa copulatriz. Pero difieren esencialmente por la concep- ción plesiomórfica que cada una de ellas atribuye a estos caracteres: e Bolsa copulatriz unida por una banda de tejido conjuntivo y muscular a la pared del pulmón [Schileyko] <> bolsa copula- triz situada junto al ovoespermiducto [Nordsieck]. e Con uno (Bradybaenidae, Xanthony- chidae) o cuatro sacos del dardo (= es- tilóforos) [Schileyko] <> saco del dardo sencillo [Nordsieck]. Ambas posturas están basadas en una ar- gumentación lógica. Por ejemplo, en relación con la posición de la bolsa copulatriz con respecto al ovoespermiducto, Schileyko (1991), basándose en Fraser (1946) consi- dera que en los pulmonados la bolsa copu- latriz se ha formado por la separación de una parte de la cavidad del manto, por lo que la forma inicial pre-helicoide se caracterizó por tenerla unida a la pared del pulmón. Por el contrario, Nordsieck (1987) afirma que la posición de la bolsa copulatriz junto al ovoespermiducto es plesiomorfa, porque se origina a través de su separación del ovi- ducto. Este autor indica que la mayoría de los Stylommatophora, particularmente los más primitivos, la tienen situada de esta forma, y considera que la posición apomorfa (libre) de la bolsa copulatriz podría tener la ventaja de que el contenido (productos de desecho de espermatóforos y esperma) puede ser reab- sorbido mejor. También manifiesta que esta posición tiene un origen múltiple evidente dentro de los Helicoidea, porque Brady- baenidae y Helicidae (que la tienen separada del ovoespermiducto) no forman ningún grupo monofilético. Finalmente, en relación con el aparato estimulador, Nordsieck (1985) justifica su opinión señalando que se origina del apéndice del pene de sus antepasados Orthurethra. Con respecto a la bolsa copulatriz, su ori- gen ovoespermiductal ha sido constatado por Visser (1973) en Gonaxis Taylor, 1877, y por Nel (1984) en Elisolimax Cockerell, 1893; además, Visser (1977), considera que la bolsa copulatriz de Basommatophora tiene origen diferente que en Stylommatophora; y Visser (1988) y Nordsieck (1993a) llegan a la conclusión de que ambos grupos se sepa- raron muy tempranamente en el proceso evolutivo de los pulmonados, por lo que la evolución de la bolsa copulatriz ha podido seguir caminos diferentes en ellos. Esta posibilidad está avalada, además, por el re- gistro fósil, ya que los Stylommatophora aparecieron en el Carbonífero, alrededor de 150 millones de años antes que los Basom- matophora, que lo hicieron entre el Jurásico posterior y el Cretácico (Solem, 1985). Los géneros aparentemente más próximos a Canariella son Montserratina Ortiz de Zárate López, 1946, Ciliella, Schileykiella, Tyrrheniellina Giusti 8 Manganelli, 1992 (sinonimia: Tyrrheniella Giusti & Manganelli, 1989) y Ciliellopsis Giusti 8 Manganelli, 1990. Estos seis géneros comparten las siguientes características: (А) Microescultura espiral de la concha for- mada por crestas muy finas y numero- sas. (В) Ausencia de estilóforos, apédices accesorios. (С) El nervio peneano aparentemente se оп- gina del ganglio cerebroideo derecho (no está constatado en Schileykiella). (D) El músculo retractor del ommatóforo derecho pasa entre el pene y la vagina. (E) Presencia de una vaina envolviendo al pene. sacos y En Montserratina la concha tiene una mi- croescultura similar a la de Canariella. Con respecto al aparato reproductor, el lugar de inserción del músculo retractor del pene es similar y ambos géneros comparten, por otro lado, el carácter plesiomórfico de presencia de glándulas vaginales digitiformes (de las que carecen los otros géneros), que tienen su porción inicial muy estrecha. La principal diferencia entre ellos consiste en la presencia en Montserratina de un pequeño músculo que conecta la vagina con el músculo co- lumelar; además, su papila peneana es per- forada típica, aunque el orificio está situado en posición subterminal, y posee una ca- vidad circular en su pared. Ciliella tiene la papila peneana muy pare- cida a la de algunas especies de Canariella, habiéndola descrito Manganelli et al. (1989) como una lengua arrugada que delimita a un surco espermático, cuya base abraza com- pletamente a la abertura del epifalo en el pene. Pero posee un lóbulo lateral izquierdo en el collar del manto, carece de glándulas vaginales digitiformes y el músculo retractor se inserta en el límite entre pene y epifalo; conquiológicamente se diferencia, además, 134 IBANEZ, PONTE-LIRA & ALONSO por la presencia de escamas “en forma de uña.” En Schileykiella y en Tyrrheniellina, como en Canariella, la concha tiene costulaciones radiales, además de crestas espirales y pi- losidad, y el músculo retractor del pene se inserta en el epifalo. Además, en el pene de Schileykiella hay una estructura parecida a la papila de Canariella planaria, descrita por Manganelli et al. (1989) como una protube- rancia maciza lateral que imita a una verda- dera papila peneana. Pero ambos géneros se diferencian claramente de Canariella por la ausencia de glándulas vaginales digitiformes. Ciliellopsis tiene también la microescultura de la concha similar a la de Canariella, pero se diferencia de ella por poseer un lóbulo la- teral izquierdo del collar del manto, por care- cer de glándulas vaginales digitiformes, por la posición del músculo retractor, que se in- serta en el límite entre el pene y el epifalo, y por la papila del pene, que es perforada tí- pica, aunque los pliegues del epifalo conti- nuan en su interior (Giusti & Manganelli, 1990: 273, fig. 36). Canariella también podría estar rela- cionada con los géneros Gasulliella Git- tenberger, 1980, Caseolus Lowe, 1852, y Haplohelix Pilsbry, 1919, que carecen com- pletamente de estructuras vaginales, según los dibujos de Gittenberger (1980: 207, fig. 5), Mandahl-Barth (1943: lám. 6 y lám. 7, fig. 2) y Verdcourt (1975: 936, figs. 1-6), respectiva- mente; y, aparentemente, el músculo retrac- tor del pene se inserta en el límite entre el pene y el epifalo, menos en Caseolus, en el que parece insertarse en el epifalo. Pero se desconoce en ellos la estructura interna del aparato reproductor, así como la posible presencia O ausencia de papila y de vaina peneana (en Gasulliella es muy probable que no exista la vaina peneana, ya que Gitten- berger, 1980, no menciona su presencia). Otros Hygromiidae carecen también de aparato estimulador y sus derivados, pero pertenecen a diversas subfamilias. Por ejem- plo, Ashfordia Taylor, 1917, y Szentgalia Pintér, 1977, están incluídos en Monachinae por la estructura del complejo peneano, sim- ilar a la del género Monacha Fitzinger, 1833, teniendo además el músculo retractor del ommatóforo derecho libre del pene y de la vagina. Metafruticicola lhering, 1892, Creti- gena Schileyko, 1972, y Caucasocressa Hesse, 1921, han sido incluídos por Nords- ieck (1987) en Hygromiinae por la constitu- ción de las vías finales masculinas (con una estructura muy compleja de la región peneana) y por la distribución, iniciándose además el conducto de la bolsa copulatriz casi en el atrio (la vagina es prácticamente inexistente). Otro género que también se po- dría comparar con Canariella es Cyrnotheba Germain, 1929, pero carece de vaina envolvi- endo al pene y tiene diferente anatomía de la papila peneana. Sin embargo, en una clasificación a nivel superior al del genéro es discutible la validez de los caracteres “A” a “E” anteriormente mencionados, ya que es posible que se hayan originado por convergencia, pues aparecen en otros géneros filogenéticamente distantes (Giusti, pers. com.): el carácter “А” se encuentra también, por ejemplo, en Xero- tricha apicina (Lamarck, 1822); el “B” tam- bién se presenta en Ashfordia y Metafrutici- cola; el “C” y el “D” son compartidos por muchos otros géneros de Hygromiidae y, fi- nalmente, sobre el “E” hay pocos datos, pero existen estructuras similares en géneros tan diferentes como Helicella Ferussac, 1821, y Helicodonta (Ferussac) Risso, 1826. Otros caracteres probablemente sean más útiles a la hora de buscar afinidades, como los referentes al collar del manto y al com- plejo paleal; pero tenemos pocos datos sobre ellos en otros géneros, por lo que de momento mantenemos a Canariella sin asig- nar a ninguna subfamilia dentro de Hygromi- idae. AGRADECIMIENTOS Deseamos expresar nuestro más sincero agradecimiento a Folco Giusti (Siena) por su detallada revisión crítica del trabajo y sus sugerencias sobre diversos aspectos del mismo; a Simon Tillier (MNHN), Yves Finnet (MHNG), Fred Naggs (NHM), Rúdiger Bieler (DMNH; actualmente en FMNH), Alan Solem (t) (FMNH), Trudi Meier (ZMZ), Alison Trew (NMW), Margret Gosteli (NMB) Klaus Groh (CGH) y Juan J. Bacallado (TFMC), por el envío del material de sus museos o de sus colecciones privadas; a G. Dajoz (MHNG), por sus fotografías del material tipo de Ca- rocolla hispidula y C. planaria, de la colección Lamarck; a F. La Roche (La Laguna) y R. Hut- terer (Bonn), por su ayuda en la traducción de textos del alemán; y a Fátima C. Henríquez y Manuel J. Valido, por su ayuda, entre otras cuestiones, en la recolección del material. CANARIELLA Y SU POSICIÓN EN HYGROMIIDAE 135 LITERATURA CITADA ALBERS, J. C., 1860, Die Heliceen nach natürlicher Verwandtschaft systematisch geordnet, 2nd ed. 139 pp., Leipzig. BOETTGER, O., 1908, Liste der Mollusken aus einem Sande im Barranco von Tegina auf Tene- rife (Canaren). Zeitschrift der Deutschen Geolo- gischen Gesellschaft, 60: 246-249. CHEVALLIER, H., 1965, Catalogue des collections du Museum correspondant a |’Histoire Naturelle des Mollusques de Ferussac. Bulletin du Mu- seum National d’Histoire Naturelle, (2)37: 476- 489. DAMIANOV, S. G. & I. M. LIKHAREV, 1975, Fauna Bulgarica, 4. Gastropoda terrestria, 526 pp., So- fia. DESHAYES, С. P., 1850. Ver FÉRUSSAC & DE- SHAYES, 1832. FERUSAG, J. В. DE, 1821, Tableaux systématiques des animaux mollusques, classés en familles na- turelles, dans lesquels on a établi la concordance de tous les systémes; suivis d'un prodrome gé- néral pour tous les mollusques terrestres ou flu- viatiles, vivants ou fossiles. Il. Tableau de la famille des limacons, 1-90 (Folio) o 1-95 (Quarto), Paris. FÉRUSSAC, J. B. DE, 1832, Ver FÉRUSSAC 8 DE- SHAYES, 1832. FÉRUSSAC, J. В. DE, 1835, Commentaires a le travail: synopsis Molluscarum terrestrium et flu- viatilium quas, in itineribus per insulas Canarias, observarunt P. B. Webb et S. Berthelot. Bulletin de Zoologie, Sec. 2(68): 90-91. FÉRUSSAC, J. В. DE & С. P. DESHAYES, 1830, 1851 [1819-1851], Histoire naturelle générale et particuliere des mollusques terrestres et fluvia- tiles tant des espèces que Гоп trouve aujourd'hui vivantes, que des depouilles fossiles de celles qui n'existent plus; classés d'après les ca- ractères essentiels que présentent ces animaux et leurs coquilles, vol. |: 402 pp. FÉRUSSAC, J. B. DE 8 G. P. DESHAYES, 1832, Histoire naturelle generale et particuliere des mol- lusques terrestres et fluviatiles, Atlas, 22-27: pl. 66* u. a. 4 Explications des planches des livrai- sons XXII, ХХШ, XXIV, XXV, XXVI et XXVII, Paris. FRASER, L. A., 1946, The embryology of the re- productive tract of Lymnaea stagnalis appressa Say. Transactions of the American Microscopical Society, 65: 279-298. GERMAIN, L., 1930, Mollusques terrestres et fluvi- atiles. Faune de France, 21: 1-477, 13 pls. GITTENBERGER, E., 1968, Zur Systematik der in die Gattung Trissexodon Pilsbry (Helicidae, He- licodontinae) gerechneten Arten. Zoologische Mededelingen, 43: 165-172. GITTENBERGER, E., 1980, Three notes on Iberian terrestrial gastropods. Zoologische Mededelin- gen, 55: 201-213. GITTENBERGER, E. & К. GROH, 1986, Zum Status der Férussac'schen Taxa Helix lens und Helix barbata (Pulmonata: Helicidae). Archiv fúr Mol- luskenkunde, 116:219-223. GITTENBERGER, E. & А. J. DE WINTER, 1985, A pyrenean Monacha species, Monacha (Meta- theba) atacis spec. nov. (Mollusca: Gastropoda: Pulmonata). Zoologische Mededelingen, 59: 197-207. GIUSTI, Е. 4 G. MANGANELLI, 1987, Notulae Ma- lacologicae, XXXVI. On some Hygromiidae (Gas- tropoda: Helicoidea) living in Sardinia and in Cor- sica (Studies on the Sardinian and Corsican Malacofauna, VI). Bollettino Malacologico, 23: 123-206. GIUSTI, F. & G. MANGANELLI, 1988, Notulae Ma- lacologicae XL. Montserratina Ortiz de Zárate López, 1946 (Pulmonata, Hygromiidae): a rede- scription. /berus, 8:89-100. GIUSTI, F. 4 G. MANGANELLI, 1989, Notulae Ma- lacologicae XLIV. A new Hygromiidae from the Tyrrhenian islands of Capraia and Sardinia with notes on the genera Xeromicra and Xerotricha (Pulmonata: Helicoidea) (Studies on the Sardin- ian and Corsican malacofauna, VIII). Bollettino Malacologico, 25:23-62. GIUSTI, F. & G. MANGANELLI, 1990, Ciliellopsis oglasae, a new hygromiid from Montecristo Is- land (Tuscan Archipelago, Italy) (Pulmonata: He- licoidea). Journal of Conchology, 33:269-277. GIUSTI, Е. 4 G. MANGANELLI, 1992, Tyrrheniellina nomen novum pro Tyrrheniella Giusti 8 Manga- nelli, 1989 (Pulmonata: Hygromiidae). Bollettino Malacologico, 28:231. GONZALEZ HIDALGO, J., 1869, Moluscos del viaje al Pacífico verificado de 1862 a 1865 por una comisión de naturalistas enviada por el Gobierno Español, Parte Primera. Univalvos terrestres.—I! + 152 pp., 8 pls, Madrid. GRAY, J. E., 1854, List of the shells of the Canaries in the collection of the British Museum, collected by MM. Webb and Berthelot. Described and fig- ured by Prof. Alcide d’Orbigny in the “Histoire naturelle des Illes Canaries.” 32 pp., London. GROH, K., C. E. PONTE-LIRA, M. R. ALONSO & M. IBANEZ (en prensa), Revision of the genus Ca- nariella P. Hesse 1918. |. Alvaradoa n. subgen., with description of a new species from El Hierro (Gastropoda: Pulmonata: Hygromiidae). Archiv für Molluskenkunde, 123. GUDE, G. K., 1896, Description of a new species of Vitrina and new forms of Helicidae, with a list of the helicoid shells hitherto found in the Canary Islands. Proceedings of the Malacological Soci- ety of London, 2: 15-22. GUDE, G. K., 1911, Description of a new species of Helicondonta from Tenerife. Proceedings of the Malacological Society of London, 9: 268. HESSE, P., 1918, Die Subfamilie Helicodontinae. Nachrichtsblatt der Deutschen Malakozoologi- schen Gesellschaft, 50: 99-110. HESSE, P., 1931, Zur Anatomie und Systematik palaearktischer Stylommatophoren. Zoologica, 31: 1-118, pls. 1-16. HESSE, P., 1934, Zur Anatomie und Systematik 136 IBÁÑEZ, PONTE-LIRA & ALONSO palaearktischer Stylommatophoren. Zoologica, 33: 1-59, pls. 1-9. HUTTERER, В. (en prensa), A new, possibly extinct species of Canariella from Tenerife, Canary Is- lands (Gastropoda: Hygromiidae). Vieraea. KERNEY, М. Р., В. А. О. CAMERON 4 J. H. JUNG- BLUTH, 1983, Die Landschnecken Nord-und Mitteleuropas. 384 pp., 24 pls., Hamburg. KRAEPELIN, K., 1895, Zoologische Ergebnisse einer Frühjahrs. Exkursion nach Madeira und den Kanarischen Inseln. Verhandlungen des Naturwissenschaftlichen Vereins in Hamburg, (3)2: 6-17. KRAUSE, A., 1895, Landschnecken von Tenerifa. Nachrichtsblatt der Deutschen Malakozoologi- schen Gesellschaft, 27: 20-29. LA-ROCHE, F. & J. BARQUIN (en prensa), A com- puting method for distribution mapping. Boletim do Museu Municipal do Funchal, Suppl. 2. LAMARCK, J. В. DE, 1822, Histoire naturelle des animaux sans vertebres, 6(2): 232 pp., Paris. LAMARCK, J. B. DE, 1838 (Rev: G. P. DESHAYES & H. MILNE EDWARDS), Histoire naturelle des ani- maux sans vertèbres, Deuxieme édition, 8: 660 pp., Paris. LOWE, R. T., 1861, Diagnoses of new Canarian land-Mollusca. Annals and Magazine of Natural History, (3)7: 104-112. MABILLE, J., 1882, Molluscornm [sic] novorum di- agnoses succintae. Bulletin de la Societe Philo- matique de Paris, (7)6: 132-147. MABILLE, J., 1883, Diagnoses testarum novarum. Bulletin de la Société Philomatique de Paris, (7)7: 115-132. MABILLE, J., 1884, Matériaux pour une faune ma- lacologique des Iles Canaries. |. Nouvelles Ar- chives du Museum d’Histoire Naturelle, (2)8: 17- 182. MANDAHL-BARTH, G., 1943, Systematische Un- tersuchungen über die Heliciden-Fauna von Ma- deira. Abhandlungen der Senckenbergischen Naturforschenden Gesellschaft, 469: 1-93. MANGANELLI, G., I. SPARACIO & F. GIUSTI, 1989, New data on the systematics of two sicil- ian land snails, Helix parlatoris Bivona, 1839 and Helix reinae L. Pfeiffer, 1856 and description of Schileykiella n. gen. (Pulmonata: Hygromiidae). Journal of Conchology, 33: 141-156. MERMOD, G., 1951, Les types de la collection La- marck au Museum de Geneve. Mollusques vi- vants. ll. Revue Suisse de Zoologie, 58: 693- 1,53: MOUSSON, A., 1872, Revision de la faune mala- cologique des Canaries. Neue Denkschriften der Allgemeinen Schweizerischen Gesellschaft für die Gesammten Naturwissenschaften, 25: 1- 176, 6 pl. NEL, N. E., 1984, The ontogeny of the genital sys- tem of Elisolimax flavescens (Keferstein) (Pulmo- nata, Urocyclidae). Annale Universiteit van Stel- lenbosch, 4: 1-130. NORDSIECK, H., 1985, The system of the Stylom- matophora (Gastropoda), with special regard to the systematic position of the Clausilidae, |. Im- portance of the excretory and genital systems. Archiv fúr Molluskenkunde, 116: 1-24. NORDSIECK, H., 1987, Revision des Systems der Helicoidea (Gastropoda: Stylommatophora). Ar- chiv fúr Molluskenkunde, 118: 9-50. NORDSIECK, H., 1988. Das System der tertiáren Helicoidea Mittelund Westeuropas (Gastropoda: Stylommatophora). Heldia, 1: 109-120, pls. 15- ur NORDSIECK, H., 1993a, Phylogeny and system of the Pulmonata (Gastropoda). Archiv fúr Mol- luskenkunde, 121: 31-52. NORDSIECK, H., 1993b, Das System der paläark- tischen Hygromiidae (Gastropoda: Stylommato- phora: Helicoidea). Archiv für Molluskenkunde, 122 (Adolf Zilch-Festschrift): 1-23. ODHNER, N. H., 1931, Beiträge zur Malakozoolo- gie der Kanarischen Inseln. Lamellibranchien, Cephalopoden, Gastropoden. Arkiv för Zoologi, 23A: 1-116. ORBIGNY, A. D’, 1839, Mollusques, echinoder- mes, foraminiféres et polypiers recueillis aux lles Canaries par MM. Webb et Berthelot, In: P. B. WEBB & $. BERTHELOT, Histoire naturelle des Iles Canaries, 2(2)(Zoologie): 152 pp., 14 pls., Paris. PFEIFFER, L., 1848-1876, Monographia heliceo- rum viventium sistens descriptiones systemati- cas et criticas omnium huius familiae generum et specierum hodie cognitarum, 1: 1-484, 1848; 3: 1-711, 1853; 5: 1-565, 1868; 7: 1-674, 1876; Lipsiae. PFEIFFER, L., 1860, Beschreibung neuer Heliceen. Malakozoologische Blätter, 7: 231-240. PFEIFFER, L., 1870-1876, Descriptions et figures de coquilles extramarines nouvelles ou peu con- nues. Novitates Conchologicae, (1)4: 1-171, pls. 109-137. PILSBRY, H. A., 1895, Helicidae, vol. 7. Manual of conchology, structural and systematic, (2)9(36): 161-366, pls. 41-71. PONSONBY, J. H. & Е. В. SYKES, 1894, Two new land-shells from Tenerife. Proceedings of the Malacological Society of London, 1: 55. PRIETO, С. E., А. I. PUENTE, К. ALTONAGA & В. J. GOMEZ, 1993, Genital morphology of Caracol- lina lenticula (Michaud, 1831), with a new pro- posal of classification of helicodontoid genera (Pulmonata: Hygromioidea). Malacologia, 35: 63-77. PUENTE, А. I. & С. E. PRIETO, 1992, Plentuisa vendía, a new genus and species from the Picos de Europa (North of the Iberian peninsula) (Gas- tropoda: Helicoidea: Hygromiidae). Journal of Conchology, 34: 159-168. RICHARDSON, L., 1980, Helicidae: catalog of spe- cies. Tryonia, 3: 1-697. SCHILEYKO, A. A., 1973, Comparative character- istics of the palaearctic families of molluscs from the superfamily Helicoidea. Zoologicheskii Zhur- nal, 54: 492-506. SCHILEYKO, A. A., 1978, Molluscs. Land molluscs CANARIELLA Y SU POSICIÓN EN HYGROMIIDAE 137 of the superfamily Helicoidea. Fauna SSSR (n.s.)17: 1-348. SCHILEYKO, A. A., 1979, The system of the order Geophila (=Helicida) (Gastropoda, Pulmonata). In: Morphology, Systematics and Phylogeny of Molluscs. Trudy Zoologicheskogo Instituta, Aka- demiia Nauk SSSR, 80: 44-69. SCHILEYKO, A. A., 1991, Taxonomic status, phy- logenetic relations and system of the Helicoidea sensu lato. Archiv für Molluskenkunde, 120: 187- 236. SHUTTLEWORTH, R. J., 1852a, Diagnosen einiger neuer Mollusken aus den Canarischen Inseln. Mittheilungen der Naturforschenden Gesell- schaft in Bern, 241/242: 137-146. SHUTTLEWORTH, R. J., 1852b, Diagnosen neuer Mollusken. Mittheilungen der Naturforschenden Gesellschaft in Bern, 260/261: 289-304. SHUTTLEWORTH, R. J., 1860. Ver PFEIFFER, 1860. SHUTTLEWORTH, R. J., 1975, Tabulae ineditae Molluscorum Insularum Canariensium, W. BACK- HUYS, edit. 43 pp., 8 pls.; Krefeld. SMITH, E. A., 1884, An account of the land and freshwater Mollusca collected during the voyage of the “Challenger” from December 1872 to May 1876. Proceedings of the Zoological Society of London, for 1884: 258-281, pls. 22, 23. SOLEM, A., 1985, Origin and diversification of pul- monate land snails. pp. 269-293, in: K. M. WILBUR, ed., The Mollusca, 10: Evolution Academic Press, London. THIELE, J., 1931, Handbuch der systematischen Weichtierkunde. Teil 2: Gastropoda: Opistho- branchia und Pulmonata: 377-788. Fischer, Jena. [reprint: A. Asher & Co., Amsterdam, 1963. Translated into English: Smithsonian Institution Libraries, Washington, 1992]. TRYON, G. W., 1887, Helicidae |. Manual of con- chology, structural and systematic, (2)3: 313 pp., 63 pls. TRYON, G. W., 1888, Helicidae Il. Manual of con- chology, structural and systematic, (2)4: 296 pp., 69 pls. VAUGHT, C., 1989, A classification of the living Mollusca. American Malacologists, Inc., R. T. AB- BOTT 4 K. J. BOSS, eds., 189 pp., Melbourne, Flor- ida. VERDCOURT, B., 1975, On the systematic posi- tion of some east African Helicidae—4. Revue de Zoologie Africaine, 89: 934-939, pl. X. VISSER, M. H. C., 1973, The ontogeny of the re- productive system of Gonaxis gwandaensis, with special reference to the phylogeny of the sper- matic conduits of the Pulmonata. Annale Univer- siteit van Stellenbosch, 48: 1-79. VISSER, М. H. С., 1977, The morphology and sig- nificance of the spermoviduct and prostate in the evolution of the reproductive system of the Pul- monata. Zoologica Scripta, 6: 43-54. VISSER, M. H. C., 1988, The significance of termi- nal duct structures and the role of neoteny in the evolution of the reproductive system of Pulmo- nata. Zoologica Scripta, 17: 239-252. WEBB, P. В. & Т. $. BERTHELOT, 1833, Synopsis molluscorum terrestrium et fluviatilium quas in itineribus per insulas Canarias, observarunt. An- nales des Sciences Naturelles, 28: 307-326. WOLLASTON, T. V., 1878, Testacea atlantica or the land and freshwater shells of the Azores, Ma- deiras, Salvages, Canaries, Cape Verdes and Saint Helena, 588 pp., London. ZILCH, A., 1960, Gastropoda, Euthyneura. In: w. WENZ, edit., Handbuch der Paláozoologie, 6: 601-835 Berlin. Revised Ms. accepted 21 March 1994 u ~ у ao 2 PCD A uw 7” dana CR A м re APO “YO q Bu RAR . 7 A o a, va Mi oe sn) u MALACOLOGIA, 1995, 36(1-2): 139-146 AGE-RELATED DIFFERENTIAL CATABOLISM IN THE SNAIL BITHYNIA GRAECA (WESTERLUND, 1879) AND ITS SIGNIFICANCE IN THE BIOENERGETICS OF SEXUAL DIMORPHISM N. Eleutheriadis & M. Lazaridou-Dimitriadou Section of Zoology, Department of Biology, University of Thessaloniki, 54006 Thessaloniki, Greece ABSTRACT Catabolic partitioning of carbon and nitrogen was investigated to clarify the sexual dimor- phism of bioenergetics in Bithynia graeca. Experiments involved post-breeding male and fe- male snails 1, 3, and 11 months old, grazing on Aufwuchs (epiphytic scum flora). Per-snail ingestion and partitioning rates were maximal for 11-month-old snails and declined with age in both sexes. Three-month-old males had lower weight-specific rates and efficiencies than fe- males. Eleven-month-old females had lower weight-specific rates and efficiencies than males. The youngest snails had higher rates in the various components of the energy budget than the older snails. Key words: Bioenergetics, Bithynia graeca, prosobranch snails. INTRODUCTION In actuarial (age- and sex-related) bioener- getics, it is generally accepted that the ana- bolic expenditures of females during the pro- duction of eggs or young will far outweigh any comparable effort involved in the pro- duction of male gametes. However, this is rarely proved, and particularly in the cases where the sexes do not differ markedly in adult biomass, it is often assumed that the production of eggs in females is compen- sated for by the greater kinetic expenditure of males (Aldridge et al., 1986). Direct investi- gation of catabolic partitioning of carbon and nitrogen in individual molluscs has comple- mented traditional assessment of changing C:N ratios of biomass in molluscan popula- tions (Russell-Hunter 4 Buckley, 1983) and thus provided a means for actuarial cross- checking or auditing of metabolic efficien- cies. Catabolic allocation can be studied by ex- periments involving nearly concurrent mea- surements of oxygen uptake and nitrogenous excretion (Aldridge, 1982; Russell-Hunter et al., 1983; Tashiro, 1982; Tashiro & Colman, 1982). Such investigation can reveal shifts from protein- to carbohydrate-based catab- olism. When discussing problems in terms of actuarial bioenergetics, quantification of both physiological rates and ecological efficiencies can be of value in a broadly adaptational ap- proach. 139 This paper deals with the catabolic parti- tioning of ingested nitrogen, and protein and nonprotein carbon, in the post-breeding sea- son of the semelparous freshwater proso- branch snail Bithynia graeca, in three age- classes—1 month, 3 months, and 11 months old—and in the last two age-classes for each sex. We have examined grazing-feeding in specimens of В. graeca, which 1$ an endemic species of Greek lakes; it has been examined for each sex and age class for a population in the artificial Lake Kerkini, Macedonia, Greece. The primary focus of this work was to assess ingestion rates and subsequent bioenergetic partitioning of ingested nitrogen and carbon. Information from feeding and as- similation studies was combined with respi- ratory and excretory measurements. In as- sessing differential catabolic allocation, techniques allow distinction between protein carbon and nonprotein carbon in partitioning and provide data that are of wider signifi- cance in discussing the complex differential bioenergetics of sexual dimorphism that can occur in this semelparous species. MATERIALS AND METHODS Specimens of the prosobranch mollusc Bithynia graeca were collected from the dam at the artificial Lake Kerkini in Macedonia, northern Greece. Animals in this population are semelparous and annual and have a 140 ELEUTHERIADIS 8 LAZARIDOU-DIMITRIADOU lifespan of approximately 12 months (there are no biennial populations in Kerkini). The sexes are separate, and male and female individuals occur in about the same numbers. Reproduction takes place in spring, and the majority of adults die after egg-laying. Growth rate is quick during spring, and life expectancy decreases with increasing age. Bithynia graeca is capable of true tissue de- growth with low mortality rate during the win- ter, when the snails can remain without food and out of the water for 6 to 7 months be- cause of the management of the lake level (Eleutheriadis 4 Lazaridou-Dimitriadou, sub- mitted). In 1992 from May to September, 1-month- old [shell height (H) = 1 mm + 0.25] (mean + standard deviation), 3-month-old (H = 1.40 mm + 0.75) and 11-month-old (Н = 4.2 mm + 0.80) snails from the same population and conditions were brought into the laboratory where they were used in separate experi- ments on catabolic partitioning, involving nearly concurrent measurements of oxygen uptake and nitrogenous excretion. Their shells were cleaned of Aufwuchs (epiphytic scum flora), and individuals were classified by age and sex. Age determination was made using shell height. Sex determination was based on the male’s reddish testis, which was noticeable as a dark area in the last whorls from the apex. The 11-month-old snails were collected in a post-breeding con- dition. After sorting, snails were placed in wa- ter from the Lake Kerkini. All water used in these studies was boiled and filtered prior to the beginning of the study. After allocation to their respective experi- mental group, snails were fed with dry Auf- wuchs consisting of particles smaller than 50 u and with a C:N (Carbon:Nitrogen) ratio 7.8: 12.8. Each group, consisting of 15, 10, and 5 individuals from 1-month-old, 3-month-old, and 11-month-old snails respectively, was placed in a culture bowl and fed with a known weight of food over a 48 h period. After 48 h the snails were removed and any uneaten food and faeces were collected from each culture bowl and separated using a filter of 50 u mesh, and then dried and reweighed. All weights were determined after drying to a constant weight at 60°С. Weights were re- corded with a Sauter AR 1014 microbalance (precision + 0.0001 9). For each experimental group, batches of food and faeces were analyzed for C and N to determine ingestion and egestion rates. Car- bon analyses were carried out using the “wet oxidation” method outlined by Russell- Hunter et al. (1968). Nitrogen analyses were done using the method of D'Elia et al. (1977). The principle of the latter is that nitrogenous compounds in the water are oxidized to ni- trate by heating with an alkaline persulphate solution under pressure. These batch analy- ses provided appropriate C and N values for food and faeces. The weights recorded for each experimental group could then be used to obtain individual rates of total C and N ingestion (Tl) and egestion (NA = not assim- ilated) and, hence total С and М assimilation rates (TA) for each snail (by subtracting the appropriate NA rate from TI) Oxygen uptake rates were determined us- ing a Digital Oxygen System Model 10 man- ufactured by Rank Brothers Ltd., Bottisham, Cambridge. The rate of oxygen uptake was calculated from the rate of depletion of oxy- gen from the water. In no case was this de- pletion allowed to exceed 40%. The de- crease in oxygen tension at the electrode was recorded on a Linseis flat-bed recorder series LS Model 0480L. Controls (in the ab- sence of snails) were run to determine the rate of depletion of oxygen due to the elec- trode and other possible factors. Measure- ments on blank chambers (without snails) were carried out before the start of the ex- periment and after every five determinations. Young, males and females were placed in groups of 25, 15, and 5 individuals from 1-month-old, 3-month-old and 11-month-old snails, respectively, in 3 ml filtered lake water. In the chamber, the magnetic stirring bar was located under the snails, which were sepa- rated from the stirrer by a nylon mesh glued to a plastic ring. Disturbance of the snails appeared to be minimal, as the snails were observed to crawl on the mesh. The mean values of control measurements were sub- tracted from the percent decrease of O, sat- uration obtained for measurements in the presence of snails. O, uptake rates were converted to values for standard tempera- ture, pressure and solubilities using the Stan- dard Methods (American Public Health Asso- ciation, 1976). Throughout the experiments the temperature of the inlet and outlet water of the waterjacket was maintained at 26°C. To assess N losses in catabolism, each snail was set up individually in a jar contain- ing 3 ml of water for 24 h at 26°C. After 24 h, the snail was removed and 0.5 ml urease, 15 U/ml (Cooper-Biomedical, Worthington) was AGE-RELATED DIFFERENTIAL CATABOLISM IN THE SNAIL 141 TABLE 1. Tissue dry weight, with per snail rates (micrograms per hour) of ingestion (Tl) and assimilation (TA) (mean + standard deviation) of carbon and nitrogen. Age Tissue dry Total C Total N Total C Total N (months) Sex n weight (mg) ingested ingested assimilated assimilated 1 — 51 0.34 + 0.09 1.74 + 0.47 0.18 + 0.07 0.80 + 0.37 0.10 + 0.05 3 fem. 33 0.86 + 0.12 2.34 + 0.86 0.25 + 0.08 0.88 + 0.25 0.12 + 0.02 3 mal. 39 0.87 + 0.21 2.00 + 0.84 0.21 + 0.08 0.70 + 0.33 0.09 + 0.02 11 fem. 26 2.08 + 0.73 6.36 + 4.33 0.64 + 0.38 2.65 + 1.86 0.28 + 0.09 11 mal. 23 1.66 + 0.42 5319 E65 0.60 + 0.15 2.80 + 1.28 0.32 + 0.05 added to hydrolyze any urea. The ammonia concentration of the water from each snail was then determined using Berthelot's colour reaction following the modification of Chaney 8 Marbach (1962). This reaction is based on the formation of the indole dye indophenol blue. Appropriate blanks and standards were used. After both oxygen uptake and nitrogen ex- cretion rates had been determined, snail shell dimensions were recorded. Snails were then decalcified т 8.5% НМО., and the perios- tracum was removed. The tissue was dried to constant weight at 65°С. This allowed the computation of all data in terms of both rates per individual snail (Table 1, Fig. 1) and as weight-specific values (Table 2). Comparisons were made between individ- ual rates of ingestion and assimilation (indi- vidual data were expressed per milligram of dry tissue) and were further computed so that they were all expressed in carbon terms (as units of protein carbon or nonprotein carbon). The conversion of various partitioning rates into a common carbon currency and the as- signment as protein carbon or nonprotein carbon (Russell-Hunter & Buckley, 1983; Russell-Hunter et al., 1983; Aldridge et al., 1985) were based on the following assump- tions: that all nitrogen excreted was derived from the breakdown of proteins, and that ox- ygen was consumed in proportion to the breakdown of organic carbon compounds, both protein and nonprotein. Weight-specific rates of ammonia excre- tion can be converted (multiplying by 0.827) to rates of nitrogen excretion. The conversion factor used to estimate the oxygen con- sumption for protein catabolism from this rate was 5.92 ul O,/ug N. Thus, a weight- specific rate of oxygen consumption for the protein fraction of catabolism could be esti- mated. Subtraction of this rate from the over- all weight-specific oxygen uptake rates gave the rate of oxygen consumption for the non- A PE #0 Oxygen uptake N exretion Age (months) FIG. 1 (A) Oxygen uptake (micrograms О2 per snail per hour) and (B) total nitrogen excretion (micro- grams М per snail per hour) for three ages (1, 3 and 11 months) of Bithynia graeca. №, females, U, males. Vertical bars represent standard errors. protein fraction. Equivalents for carbon mass consumed were derived from the appropriate amounts of CO, evolved (0.536 ug C con- sumed/ul CO,). The relation of CO, evolved to O, con- sumed differs for proteins and nonproteins. The CO, evolved from protein catabolism can be derived directly from the weight-spe- cific nitrogen excretion rate (4.75ul CO, evolved/ug N excreted). The CO, evolved from nonprotein catabolism can be esti- mated by multiplying the weight-specific ох- ygen consumption for the nonprotein fraction by an appropriate respiratory quotient. Food- component analysis for prosobranchs sug- gests an average of 10% fat and 90% car- bohydrate for the nonprotein fraction, giving a respiratory quotient of 0.95. Thus, separate estimates of catabolism (RA = respired as- similation) in carbon terms for protein and 142 ELEUTHERIADIS 8 LAZARIDOU-DIMITRIADOU TABLE 2. Mean weight specific partitioning rates (micrograms C per milligram per hour) of protein and nonprotein С sources into the various components of the energy budget. Non- Non- Non- Non- Age Protein С protein Protein С protein С Protein protein Protein С protein (months) Sex п ingested С ingested assimilated assimmilated Cin RA Cin RA in NRA Cin NRA 1 — 51 1.70 3.40 0.95 1.42 0.53 0.31 0.42 li 3 fem. 33 0.94 1.78 0.45 0.58 0.32 0.28 0.13 0.30 3 mal. 39 0.79 1.50 0.34 0.46 0.25 0.38 0.09 0.08 11 fem. 26 1.00 2.05 0.47 0.79 0.45 0.15 0.02 0.61 11 mar 2% ай 2:31 0.63 1.02 0.50 0.18 0.13 0.87 nonprotein could be computed. Rate values were also calculated for total ingestion (TI), egestion (NA) and assimilation (TA) deter- mined in both carbon and nitrogen terms, based on food and faeces, and from these were derived non-respired assimilation (NRA) values. Additionally, all rates (and relative ef- ficiencies) in terms of nonprotein carbon and protein carbon [carbon content being 3.25 times the nitrogen content of protein (Rus- sell-Hunter & Buckley, 1983)] could be com- puted. In order to find out if rate differences existed among or/and between experimental groups, analysis of variance (ANOVA) and Fisher LSD tests were executed, respectively (Daniel, 1991). Data was logarithmically transformed prior to analysis. RESULTS The basic data on catabolic allocation are shown as oxygen uptake and nitrogenous excretion rates in Figure 1. Values for oxygen consumption per individual (Fig. 1A) were statistically higher for 11-month-old snails compared to 3- and 1-month-old snails and for 3-month-old snails in relation to 1-month- old snails (P < 0.05). Values for nitrogenous excretion per individual were statistically higher in older than younger snails (P < 0.05). Further comparisons made below were weight-specific (i.e., the individual data were expressed per milligram of dry tissue) and were further computed so that data were ex- pressed in carbon terms (as units of protein carbon or nonprotein carbon). Mean tissue dry weight and individual rates of ingestion and assimilation in C and N terms, for each age and sex fed on Auf- wuchs, are shown in Table 1. The females of both ages had slightly higher grazing rates than males, but these were not statistically different. This occurred only for the smaller females for the assimilation rate but not for the older females. The older snails had higher ingestion and assimilation rates for C and N than the younger snails (P < 0.05). For each age and sex that fed on Auf- wuchs, mean weight-specific rates of protein and nonprotein carbon partitioning are shown in Table 2. Rates for males and fe- males were compared using ANOVA and Fisher LSD test (Table 3). Overall, young females had higher weight- specific rates of carbon (ingested and assim- ilated) than young males. In contrast, older males generally had higher weight-specific rates than females in the same age-class. The weight-specific carbon partitioning rates tended to decline with age from 1-month to 3-month-old snails only, and increased from 3-month-old to 11-month-old snails. The dif- ferences were statistically significant be- tween the 1-month-old snails and the other age classes for protein C ingested (P < 0.05) (Table 3), between 1-month-old and 3- month-old male snails for protein С assimi- lated (P < 0.05) (Table 3) and between 1-month-old and 3-month-old snails for non- protein C ingested (P < 0.05). There were no statistical differences for nonprotein C assim- ilated. The differences were significant be- tween 11-month-old male and 3-month-old male and between 1-month-old and 3- month-old snails for protein RA (P < 0.05) (Table 3). Categories of carbon partitioning rates per snail are shown in Figure 2. In Figure 2A in- gested and assimilated values are con- trasted, whereas Figure 2B presents the par- titioning between respired and nonrespired assimilation. Expressed in units per snail and time, all rates were highest in 11-month-old snails. Differences could be detected in both rates of ingestion and in the differential ca- tabolism of protein and nonprotein sub- strates. The protein carbon efficiencies in AGE-RELATED DIFFERENTIAL CATABOLISM IN THE SNAIL 143 TABLE 3. Comparisons between different age and sex groups (1-month-old, 3-months-old ог 11-month-old female (F) or male (М) snails) of Bithynia graeca concerning the protein С ingested, protein С assimilated and protein С in respired assimilation with Fisher LSD test (*:Р < 0.05). 1 ЗЕ 3M 11 F 11M Protein С ingested (РС) — Protein C — 1 assimilated (PCa) Protein in R.A. (PRA) — M РС! 0.174* 0.190 — M PCa 0.334* 0.366 — 3 M PRA 0.395* 0.257 — F РС! 0.174* — Е РСа 0.334* — Е РВА 0.390* — М РС! 0.163* 0.180 0.180* = M PCa 0.313 0.347 0.347 -— 11 М РВА 0.385 0.211 0.251" — Е РС! 0.163* 0.180 0.180 — 0.170 Е РСа 0.313 0.347 0.347 — 0.327 F PRA 0.386 0.243 0.249 — 0.235 Rate (micrograms С « snail"! h'!) (months) FIG. 2. Partitioning of nonprotein (O) and protein carbon (И) of Bithynia graeca. (A) Total ingestion (TI) and total assimilation (ТА). (В) Respired assim- ilation (RA) and nonrespired assimilation (NRA). The left pair of histogram bars are for females; the right pair, for males. non respired assimilation for 11-month-old snails are negative indicating that protein de- growth takes place in order to meet the needs of reproduction knowing that efficien- cies tend to decline with age. Females had higher ingestion rates for protein carbon than males in both age classes, but these were not statistically different. The values for protein C ingested and assimilated were derived from N ingested and assimilated, so the results of the statistical tests were the same. Statistical differences were detected in rates of inges- tion in differential catabolism of nonprotein substrates between 11-month-old snails and all the other age-classes (P < 0.05), and in rates of assimilation of nonprotein substrates between 11-month-old snails and the 3-month-old males and 1-month-old snails (P < 0.05). Protein RA were statistically dif- ferent between 11-month-old snails and all the other age-classes. Average assimilation efficiencies are shown in Table 4. Overall, young females had higher assimilation efficiencies than young males, while in contrast, older males had higher efficiencies than older females. The youngest snails had higher assimilation effi- ciencies than 3-month-old and 11-month-old female snails. The differences in processing rates with age and sex noted above could be expressed as percentage values of nonprotein assimila- tion over total ingestion (gross efficiencies) or over total assimilation (net efficiencies) (Rus- sell-Hunter & Buckley, 1983). For each age and sex, average gross and net efficiencies are presented in Table 5. Young females had higher gross and net growth efficiencies than 144 ELEUTHERIADIS 8 LAZARIDOU-DIMITRIADOU TABLE 4. Mean assimilation efficiencies (percent) for various budget components. Age (months) 1 Sex n 51 C assimilation efficiency 45 Protein C assimilation efficiency 57 Nonprotein C assimilation efficiency 42 3 3 11 11 females males females males 93 39 26 23 38 33 42 48 48 46 47 54 33 31 39 44 TABLE 5. Gross and net growth efficiency averages (percent). Gross growth efficiency Age (months) Sex n TotalC Protein С 1 — 51 30 25 3 fem 33 16 14 3 mal. 39 08 14 11 fem 26 21 02 11 mal 23 29 lil young males and the opposite occured in older snails. The youngest snails had higher gross and net efficiencies for total C than all the other age classes. DISCUSSION During a lifespan of 12-13 months, B. graeca reach up to 7 mm in length and are active from March to October. The active pe- riod consists of three parts: breeding (April- June), post-breeding (July-August) and prewinter period, lasting from September un- til the onset of diapause (Eleutheriadis & Laz- aridou-Dimitriadou, 1992). In this study, it was hoped that age- and sex-related differ- ences in catabolic allocation would be re- vealed. Bithynia graeca feeds both by radula grazing on detritus and Aufwuchs and by ctenidial filter-feeding on phytoplankton and other suspended organic material, as any other species of Bithyniidae. The frequency of filter-feeding may depend on the availabil- ity of other food sources and the nature of the suspended particles (Fretter 8 Graham, 1962). The prediction of sexual dimorphism in the metabolic processes of animals involves higher anabolic rates and efficiencies for fe- males and higher catabolic rates and efficien- cies for males. The relatively higher anabolic expenditures of females are invested in the production of eggs or young (Russell-Hunter 8 Buckley, 1983), and it is usually assumed Net growth efficiency Nonprotein С Total C Protein С Nonprotein С 32 64 04 78 17 42 20 52 05 23 30 17 30 50 30 77 38 60 44 85 that in males the greater kinetic energy ex- penditures are directed to extensive gene dispersion. К must be noted that the data for B. graeca were presented both as rates (individual and weight-specific) and as efficiencies (gross and net), and were derived from experiments limited to the grazing mode of feeding (ex- cluding the alternative filter-feeding mode) and to post-breeding snails (excluding the spring period of exponential growth in fe- males). The age- and sex-related differences in catabolic allocation revealed the theorized trade-offs against the known bioenergetic differences in rates and efficiencies between the sexes. Oxygen consumption and nitrog- enous excretion in older snails were statisti- cally higher than those in younger snails, and this also occured for Tl and TA in both carbon (protein and nonprotein) and nitrogenous terms. The youngest snails had higher rates in the various components of the energy budget than all the others. The reason for these dif- ferences could be the decrease in metabolic rhythm as the snails were growing up. This has also been observed in terrestrial gastro- pods (Jennings & Barkham, 1976; Lazaridou- Dimitriadou & Daguzan, 1978; Stern, 1968; Zeifert 8 Shutov, 1979; Charrier 8 Daguzan, 1980; Staikou 8 Lazaridou-Dimitriadou, 1989; Lazaridou-Dimitriadou 4 Kattoulas, 1991) and in freshwater gastropods (Aldridge et al., 1986). A A A AAA ew AGE-RELATED DIFFERENTIAL CATABOLISM IN THE SNAIL 145 Overall, young females had higher average assimilation efficiencies than young males because the anabolic expenditures for the production of eggs would probably far out- weigh any comparable effort involved in the production of male gametes. In contrast, 11- month-old males had higher assimilation ef- ficiencies than 11-month-old females, prob- ably because the males had greater kinetic expenditures than females. This species is semelparous and lives for 12 to 13 months, so the 11-month-old snails probably had lower energy expenditures in the production of gametes. For the same reasons, young fe- males had higher gross and net growth effi- ciencies than young males and the opposite in older snails. The ratio N-RA/TA in carbon terms (net growth efficiency) ranged from 23-64% for B. graeca. Hunter (1975) gave almost the same range (22-41%) for two populations of Lym- naea palustris, whereas McMahon (1975) re- ported 22-52% for four generations in three populations of Laevapex fuscus. In two pop- ulations of the European stream limpet An- cylus fluviatilis much lower values (11.2% and 11.8%) were found (Streit, 1975, 1976a, cited in Russell-Hunter 8 Buckley, 1983). Burky (1971) reported 19% for N-RA/TA in Ferrissia rivularis, and Aldridge (1982) re- ported 12.2%, 12.4% and 12.9% in carbon terms for three populations of Leptoxis cari- nata. Tashiro & Colman (1982) reported the highest values (more than 90%) for N-RA/TA in Bithynia tentaculata for filter-feeding. The effect of food quantity and quality must be the reason for these many-fold differences in net growth efficiencies as McMahon et al. (1974), McMahon (1975), and Aldridge et al. (1986) have also noted. LITERATURE CITED ALDRIDGE, D. W., 1982, Reproduction tactics in relation to life-cycle bioenergetics in three natu- ral populations of the freshwater snail, Leptoxis carinata. Ecology, 63: 196-208. ALDRIDGE, D. W., W. D. RUSSELL-HUNTER & D. E. BUCKLEY, 1986, Age-related differential ca- tabolism in the snail, Viviparus georgianus, and its significance in the bioenergetics of sexual di- morphism. Canadian Journal of Zoology, 64: 340-346. AMERICAN PUBLIC HEALTH ASSOCIATION, 1976, Standard methods for the examination of water and wastewater, 13th ed. American Public Association Inc., Washington, D.C. BURKY, A. J., 1971, Biomass turnover, respiration and interpopulation in the stream limpet, Ferris- sia rivularis (Say.). Ecological Monographs, 41: 235-251. CHANEY, A. & С. МАВВАСН, 1962, Clinical chem- istry, 8: 130. CHARRIER, M. & J. DAGUZAN, 1980, Consumma- tion alimentaire: production et bilan énergétique chez Helix aspersa (Müller) (Gasteropode pul- mone terrestre). Annalles de la Nutrition et de l’Alimentation, 34: 147-166. DANIEL, W. D., 1991, Biostatistics: a foundation for analysis in the health sciences. John Wiley & Sons, New York. D'ELIA, C. F., P. A. STEUDLER 4 N. CORWIN, 1977, Determination of total nitrogen in aqueous samples using persulphate digestion. Limnology and Oceanography, 22: 760-764. ELEUTHERIADIS, N. 8 M. LAZARIDOU-DIMITRIA- DOU, 1992, Contribution to the ecology of Bithy- nia graeca, Westerlund, 1879 Macedonia, Greece. Abstracts of the Eleventh International Malacological Congress, Siena, Italy, 1992, 175 pp. ELEUTHERIADIS, М. & М. LAZARIDOU-DIMITRIA- DOU, submitted, The life cycle, population dy- namics, growth and secondary production of the snail Bithynia graeca (Westerlund) (Gastropoda: Prosobranch) in Lake Kerkini (Serres, Mace- donia, Northern Greece). FRETTER, V. & A. GRAHAM, 1962, British proso- branch molluscs. Bartholomew Press, Dorking. HUNTER, R. D., 1975, Growth, fecundity, and bioenergetics in three populations of Lymnaea palustris in upstate New York. Ecology, 56: 50- 63. JENNINGS, T. J. 4 J. P. BARKHAM, 1976, Quan- titative study of feeding in woodland by the slug Arion ater. Oikos, 27: 168-173. LAZARIDOU-DIMITRIADOU, М. & J. DAGUZAN, 1978, Consummation alimentaire, production et bilan énergétique chez Euparypha pisana (Müller) (Gastéropode Pulmoné). Annalles de la Nutrition et de Г Alimentation, 32: 1317-1350. LAZARIDOU-DIMITRIADOU, M. & M. KATOULAS, 1991, Energy flux in a natural population of the land snail Eobania vermiculata (Múll.) (Gas- tropoda: Pulmonata: Stylommatophora) in Greece. Canadian Journal of Zoology, 69: 881- 891. McMAHON, В. F., 1975, Growth, reproduction, and bioenergetics variation in three natural pop- ulation of a freshwater limpet Laevapex fuscus (C. B. Adams). Proceedings of the Malacological Society of London, 41: 331-351. McMAHON, R. F., R. D. HUNTER & W. D. RUS- SELL-HUNTER, 1974, Variation in Aufwuchs at six freshwater habitats in terms of carbon bio- mass and of carbon: nitrogen ratio. Hydrobiolo- gia, 45: 391-404. RUSSELL-HUNTER, W. D., D. W. ALDRIDGE, J. S. TASHIRO & B. S. PAYNE, 1983, Oxygen uptake and nitrogenous excretion rates during overwin- 146 ELEUTHERIADIS 8 LAZARIDOU-DIMITRIADOU ter degrowth conditions in the pulmonate snail, Helisoma trivolvis. Comparative Biochemistry and Physiology, A, 74: 491-497. RUSSELL-HUNTER, W. D. & D. E. BUCKLEY, 1983, Actuarial bioenergetics of nonmarine mol- luscan productivity. Pp. 463-503, in W. D. RUS- SELL-HUNTER, ed., The Mollusca, Academic Press, Inc., N.Y. RUSSELL-HUNTER, W. D., R. T. MEADOWS, M. L. APLEY & А. J. BURKY, 1968, On the use of a “wet oxidation” method for estimates of total organic carbon in mollusc growth studies. Pro- ceedings of the Malacological Society of Lon- don, 38: 1-11. STAIKOU, A. & М. LAZARIDOU-DIMITRIADOU, 1989, Feeding experiments on and energy flux in a natural population of the edible snail Helix lu- corum L. (Gastropoda Pulmonata Stylommato- phora) in Greece. Malacologia, 31: 217-227. STERN, G., 1968, Recherches sur le bilan énergé- tique de la limace Arion rufus L. en période de croissance. Doctorat 3e cycle, Université de Paris IV, 75230 Paris. TASHIRO, J. S., 1982, Grazing in Bithynia tentac- ulata: age-specific bioenergetic patterns in re- productive partitioning of ingested carbon and nitrogen. American Midland Naturalist, 107: 133- 150. TASHIRO, J. 5. & S. D. COLMAN, 1982, Filter- feeding in the freshwater prosobranch snail Bithynia tentaculata. American Midland Natural- ist, 107: 114-132. ZEIFERT, D. V. & S. V. SHUTOV, 1979, Role of certain terrestrial mollusks in the transformation of leaf litter. Ecologyia (Sofia), 5: 58-61. Revised Ms. accepted 18 May 1994 MALACOLOGIA, 1995, 36(1-2): 147-153 DIETARY PREFERENCE OF THREE FRESHWATER GASTROPODS FOR EIGHT NATURAL FOODS OF DIFFERENT ENERGETIC CONTENT Heinz Brendelberger Zoologisches Institut der Universität zu Köln, Physiologische Ökologie, D-50923 Köln, Germany ABSTRACT The food preferences of three freshwater gastropods, Radix peregra (Pulmonata), Bithynia tentaculata and Bythinella dunkeri (Prosobranchia), have been examined in laboratory experi- ments. Eight natural foods, a green alga, a cyanobacterium, and leaves of sycamore maple, alder and oak with different conditionings, were tested in various combinations. The number of animals found on these foods during a three-hour period was counted. There were significant differences in the food choice of the three species: Bithynia tentaculata clearly preferred algae and cyanobacteria, Bythinella dunkeri selected leaves conditioned by microorganisms and discriminated against unconditioned leaves and algae, and Radix peregra showed intermediate food preferences for almost all foods. For Radix and Bithynia, the results of these experiments were correlated with the C:N ratio and the % nitrogen content of the food items: preference increased with decreasing C:N ratios and with increasing nitrogen content. There was no such correlation for Bythinella. The possible reasons for this are discussed. Key words: Optimal foraging, food choice, natural food, C:N-ratio, freshwater gastropods, Radix, Bithynia, Bythinella. INTRODUCTION One of the assumptions of optimal forag- ing theory is that “high fitness is achieved by a high net rate of energy intake” (Begon et al., 1990). If they are behaving as optimal foragers, grazing gastropods should therefore feed preferentially on foods of high energy content when presented with an array of potential food sources of equal quantity and accessa- bility. Discrimination by gastropods according to algal class (Calow, 1970; Imrie et al., 1990), food particle size (Levinton 4 DeWitt, 1989) and even taste (Daldorph & Thomas, 1988) has been described. In a related paper (Bren- delberger, 1992), | found that various com- ponents of natural food were contrastingly attractive to Radix peregra and Bythinella dunkeri. McMahon et al. (1974) was able to relate the uptake of periphyton and the rejec- tion of macrophyte tissue by gastropods to the carbon to nitrogen ratios of these two foods, low and high respectively. Despite the success of that study, a systematic investi- gation of a wide array of natural gastropod foods and their energy value has not yet been made. 147 Using a true food preference method (sensu Peterson & Renaud, 1989), | therefore tried to relate the food preferences of three freshwater gastropod species, Radix pe- regra, Bithynia tentaculata and Bythinella dunkeri, to the C:N ratios and the %-nitrogen content of the foods under consideration. MATERIAL AND METHODS The experiments were made with three freshwater gastropod species, Radix peregra (Múller) (Pulmonata, Lymnaeidae), Bithynia tentaculata (L.) (Prosobranchia, Bithyniidae), and Bythinella dunkeri (Frauenfeld) (Proso- branchia, Hydrobiidae). Radix and Bithynia were collected in shallow waters near the shore of the Rhine near Köln, whereas Bythinella was from a first order stream 30 km east of Köln, Germany. The animals were kept in the laboratory un- der constant conditions, a 12:12 h light-dark cycle at 18+2°С, for at least two months be- fore the experiments began. During this ac- climation period, they were fed with lettuce (Radix, Bithynia) or Fontinalis antipyretica (a moss) (Bythinella dunkeri) in non-limiting amounts. 148 BRENDELBERGER Food choice was tested in glass petri dishes, 22 cm diameter for Radix and Bithy- nia, 6 cm diameter for Bythinella, using ten animals per dish. Foods were always offered in pairs at opposite sides of the experimental dishes. The position of the animals in the dishes was recorded every ten minutes dur- ing the three-hour experimental period. The “preference” of a test food was calculated from the sum of all animals feeding on that food during the experimental period. The foods offered were a cyanobacterium (Synechococcus elongatus), a green alga (Chlamydomonas reinhardii) and leaves of three deciduous tree species with different conditioning. Leaf discs of sycamore maple (Acer pseudoplatanus), alder (Alnus gluti- поза) and oak (Quercus robur) were condi- tioned for 10 days. Additional food sources were watered alder leaves or ultrasonicated leaves (i.e. depleted of the microorganisms growing on the decaying leaf surface), or the microorganisms from the leaf surface. For details of the experimental set-up and preparation of the different foods, see Bren- delberger (1992). In order to obtain realistic results bearing as much similarity to field conditions as pos- sible, the following details were observed: — Some species of freshwater gastro- pods, such as Bithynia tentaculata, show a circannual rhythm of activity, even under constant laboratory condi- tions (Brendelberger & Júrgens, 1993). Therefore, all preference tests were performed in spring and summer when all species were active. — Food choices of freshwater gastropods are not constant, but may change over time or with the age of the animals (Skoog, 1978; Egonmwan, 1991). For the experiments, only juveniles of Radix peregra (6-8 mm shell length) and Bithynia tentaculata (5-6 mm shell length) were used. Juveniles of B. ten- taculata have been shown to have higher growth rates than adults (Tashi- ro, 1982). They should therefore search for food more constantly and intensely. The juvenile animals of Radix and Bithy- nia never showed any interactions that could interfere with the search for food. The maximum body length of Bythinella dunkeri is only 3 mm; therefore, adult animals were used for their food pref- erence tests. — Gastropods may feed preferentially on food items that they have been eating before (Bleakney, 1989; Imrie et al., 1990). Therefore, the maintenance food given during the acclimation period was excluded from the experiments. — Food preference has been found to change with hunger level (Calow, 1973). The effect of hunger level on food pref- erence was tested with Radix peregra. A moderate hunger level was created for all species in all experiments. — As gastropods react not only to the kind of food, but also to the quantity of food offered (Daldorph & Thomas, 1988; Madsen, 1992), all food items were given at similar, comparable con- centrations. Algae and cyanobacteria were fed at identical biovolumes of 1.2 x 10° fl per filter, and leaf discs had an area of 5 cm? (big petri dishes) or 1 cm? (small petri dishes). — Food was replaced by fresh food of the same kind and concentration every hour. This had two effects: food could not be depleted by the animals, i.e. it did not lose its attractivity, and the an- imals had to find and actively crawl on the food several times in order to pro- duce high preference values. — Animals found to be inactive, ¡.e. not having changed their position during the first ten minutes of an experiment, were replaced by other animals of the same species and hunger level. — All food combinations were tested with 3-5 replicates of ten animals each per species. In control experiments without food or with empty filters only, the homogeneity of the distribution of the animals was tested with chi-square statistics (p < 0.05). The experi- ments with detritus and algae were designed as true food preference experiments, sensu Peterson & Renaud (1989). As they recom- mended, when there is no depletion of food, t-tests for pairwise comparisons of the ex- perimental results were carried out. An elemental analysis was performed in or- der to characterize the nutritional value of each food: aliquots (+5 mg) of each food were combusted in a Carlo-Erba elemental analyser, and carbon and nitrogen content were recorded. The results were related to the food preference values found for the three gastropod species. FOOD PREFERENCE OF FRESHWATER GASTROPODS 149 preference > oO | foo a combinations FIG. 1. Pairwise results (mean + SD) of food preference experiments for three gastropod species and 8 food items; dotted (front): Radix peregra, dashed (medium): Bythinella dunkeri, dark grey (back): Bithynia ten- taculata; 1 = Synechococcus elongatus; 2 = Chlamydomonas reinhardii; 3 = Sycamore, conditioned; 4 = oak, conditioned; 5 = alder, conditioned; 6 = alder, ultrasonicated; 7 = alder, watered; 8 = microorganisms from alder; * : food choice of a species is significantly different for pairwise tested foods at р < 0.05; Vertical bars representing Standard deviations are scaled to half the SD values. RESULTS The results of food preference tests are in- fluenced by the hunger level of the animals; well-fed animals are more selective than hun- gry ones (Calow, 1973). Hungry snails show high activity searching for food and are thought to discriminate less between food items of different quality. The effect of hunger level has been tested with Radix peregra. Four sets of 3 x 10 animals each were starved for 1, 2, 4 or 7 days. After this time, the attractivity of Chlamydomonas cells at standard concentration was tested. The at- tractiveness increased from 11.5 + 5.8 (ani- mals found on this food per experimental pe- riod of 3 hours) after a one-day hunger period, to 22.3 + 10.4 (2 days), 40.8 + 8.7 (4 days), and 55.0 + 5.2 (7 days). After even longer periods, animals were found to behave very erratically: some showed higher activity, while others completely ceased moving around. The intermediate hunger level of 4 days was therefore chosen for the experi- ments. Without food or with only clear filters, Radix, Bithynia and Bythinella were homogeneously distributed in their petri dishes. But when nat- ural food was offered, the animals showed significant differences with respect to differ- ent foods, the three gastropod species be- having differently (Fig. 1). Radix peregra had preference values from 9.8 for watered alder leaves to 58.2 for Chlamydomonas cells. Me- dium attractivity was for conditioned leaves of alder, sycamore and oak, for Synechococcus and for microorganisms isolated from alder. Leaves deprived of their microorganisms, i.e. ultrasonicated leaves, were less attractive. Bythinella dunkeri, in contrast, was not at- tracted by algae and cyanobacteria: Chlamy- domonas and Synechococcus both gave preference values well below 10. Very high attractivity was shown to incubated leaves, with no statistical difference between alder, sycamore and oak. However, when watered leaves or ultrasonicated leaves were pre- sented, their attractivity was much lower than that of the conditioned leaves. Bithynia ten- taculata, the third species, preferred Chlamy- domonas and Synechococcus much more than (2 x greater) any kind of leaves. In general, Radix peregra fed on almost all foods at intermediate values, Bythinella dunkeri clearly preferred detritus and dis- criminated against algae and cyanobacteria, whereas Bithynia did just the opposite: it pre- ferred algae and cyanobacteria and discrim- inated against detritus. The results of the elemental analysis are presented in Table 1. The C:N ratio increased from about six for cyanobacteria and algae to 150 BRENDELBERGER TABLE 1. Carbon to nitrogen ratios and nitrogen content (percent) of eight food types (mean + SD; all determinations in triplicate): No Food type C:N %N 1 Synechococcus elongatus 5.4 + 0.1 8.0 + 0.9 2 Chlamydomonas reinhardii 6.0 + 0.9 7.3 = 3.9 3 Sycamore maple, incubated 18.8 + 5.0 252203 4 Oak, incubated 24.4 + 0.8 2.0 + 0.3 5 Alder, incubated 12.4+0.6 3.9+0.2 6 Alder, ultrasonicated 14.3+0.8 3.5 = 0.2 7 Alder, watered 15.4 + 0.3 32-5081 8 microorganisms from alder 22212 2.9+0.6 a maximum value of 24.4 for incubated oak leaves. Intermediate values were found for in- 30 = Radix | cubated alder and sycamore leaves, for wa- 2 Y APIO tered and ultrasonicated leaves of alder, and 20 À 6 for microorganisms from alder. As the carbon 4 content of these foods was generally fairly 10 constant (between 43% and 48% of dry E weight), changes in C:N ratios reflect changes CR ©. in nitrogen content. This is shown in the fourth < о column of Table 1. Absolute values were from = 30F a Bythinella |, = 2% to 8% nitrogen, with a minimum of 2% for 6 Y va © © dunkeri so oak and a maximum of 8% for cyanobacteria. 20+ . e168 As food with a low C:N-ratio and/or a high и a A nitrogen content is generally considered to 10: y № 7 = be better food (McMahon, 1974), | tried to .. sr = correlate these data with the results of the [ee |; food preference experiments. Figure 2 shows that there is a significant correlation between s0r „, Bithynia =: C:N ratio and food preference (solid line) for CAC BL tentaculata Radix peregra and for Bithynia tentaculata. 6 The corresponding regressions are: A Radix peregra: у = 21.421—0.248x; п = 14; r = -0.532; Bithynia tentaculata: y = 2 22.554—0.375х; п = 14; r= 0.722; (with y = О ep ee O 10 20 30 40 50 60 70 80 C:N-ratio and x = food preference value). The nitrogen content of these foods was also significantly correlated with food prefer- ence (Fig. 2, dashed line): Radix peregra: y = 1.190+0.096x; n = 14; г = 0.615; Bithynia tentaculata: у = 1.037—0.143x; п = 14; г = 0.876; (with y = %N-content and x = food preference value). The food preferences of Bythinella dunkeri, on the other hand, could not be related to either C:N-ratio or nitrogen content. There- fore, no lines have been drawn for this spe- cies in Figure 2. The possible reasons for this are discussed below. DISCUSSION Feeding is one of the main interactions be- tween an animal and its environment. If the quantitative and qualitative aspects of this preference FIG. 2. Relationships between food preference and C:N-ratio (dots, line) and between food preference and % nitrogen content (triangles, dashed) for Ra- dix peregra, Bythinella dunkeri and Bithynia tentac- ulata. Each symbol represents the mean of 3-5 replicates with ten animals each. For details of lin- ear regressions, see text. process are precisely known, the animal's position in the food web of its habitat can be evaluated. The first step in feeding 1$ the selection of suitable food items. For freshwater gastro- pods, this behaviour is governed by chemical and tactile stimuli (Masterson & Fried, 1992). Therefore, investigations of the attractive or FOOD PREFERENCE OF FRESHWATER GASTROPODS 151 deterrent properties of isolated substances, such as amino acids, sugars or phenolics (Norton et al., 1990), are helpful, but cannot be used directly to explain the performance of the animals in the field. | therefore concentrated upon different natural foods, detritus and its components. The results show that there are significant dif- ferences in the behaviour of the three snail species tested and also in their responses to different foods. Radix peregra showed intermediate prefer- ence values for many foods, with a slight preference for green algae. This is in accor- dance with previous studies, in which Radix peregra was shown to feed preferentially on green algae (Calow, 1973a,b; Knecht 8 Wal- ter, 1977; Lodge, 1986; Brendelberger, 1992), although in general, pulmonates are regarded as generalist herbivores (Madsen, 1992). Bythinella dunkeri, in contrast, selected strongly against algae, but preferred condi- tioned sycamore maple and alder leaves. This has also been observed in previous ex- periments (Brendelberger, 1992). A prefer- ence for conditioned rather than uncondi- tioned leaves has also been found for Gammarus pulex and Asellus aquaticus by Barlocher (1990). A possible explanation for this phenomenon 1$ the increased availability of amino acids produced by aquatic hypho- mycetes during conditioning. Bithynia tentaculata preferred green algae and cyanobacteria. This can be explained by Bithynia's tendency to feed on suspended food, mainly algae, whenever possible (Bren- delberger & Júrgens, 1993). Filtration of the green alga Chlorella has been shown to yield a higher net gain of carbon and nitrogen per respired cost than grazing (Tashiro 8 Col- man, 1982). Consequently, the uptake of fil- terable green algae should be favoured whenever possible. The various detrital com- ponents had almost no attractivity for Bithy- nia. Thus, it can be shown that these three spe- cies, even though two of them are from the same habitat, behave completely differently, and that they clearly discriminate between different food items. In the process of conditioning, bacteria and aquatic fungi degrade the organic parts of deciduous leaves. The maximum biomass of fungi on elm and oak leaves occurs in the second week (Findlay & Arsuffi, 1989). For the present study, leaves were conditioned for ten days. Aquatic hyphomycetes increase the protein and nitrogen content of leaves (Kaushik & Hynes, 1971), whereas Findlay & Arsuffi (1989) found that the biomass of con- ditioning microorganisms may equal 5.2% of the leaf biomass in terms of carbon. There- fore, carbon and nitrogen seem to be good indicators of the total energy content (C) and the share of easily available substances (N) (Aldridge, 1983), and they are suitable mea- sures to determine energy fluxes through in- dividuals or populations (Russell-Hunter & Buckley, 1983). Numerical values for these are given by Richardson (1990): he found a C:N-ratio of 19.4, 2.4% nitrogen in alder leaves. The alder leaves in this paper had C:N ratios of 18.8 and 2.5% nitrogen. Values for ash, which is less attractive to shredders (Richardson, 1990), are 35.6 (C:N) and 1.2% nitrogen. Oak leaves are known to degrade more slowly than sycamore leaves (Kaushik & Hynes, 1971); therefore, the microorganismal biomass on the leaves after ten days will probably be different, contributing to their contrasting attractivity. The green alga Chlorella, used success- fully in feeding experiments with Bithynia ten- taculata by Tashiro & Colman (1982), was found to have a low C:N-ratio of 12.0 and 4.6% nitrogen. These values are even sur- passed, in terms of food quality, by Chlamy- domonas reinhardii used in this study (C:N = 6.0; 7.3%N; cf. Table 1). The food preference of Radix peregra and Bithynia tentaculata could indeed be ex- plained by the C:N-ratios and nitrogen con- tent: food preference increased with increas- ing nitrogen content and with decreasing C:N-ratio. A preference for food rich in nitro- gen has also been found by Newman et al. (1992) for an amphipod, a caddisfly, and a physid snail. The food preference of Bythinella dunkeri, however, cannot be explained by carbon and nitrogen content. There are several possible explanations for this. In contrast to Radix and Bithynia, in which juvenile animals were tested, adult Bythinella had to be used be- cause of the species’ small size. But animals do not only select food that is of generally high quality, they also select food items to meet specific requirements. These require- ments may differ for adult Bythinella, which may be investing more energy in reproduc- tion, whereas juvenile Radix and Bithynia are investing in somatic growth. A switch in pref- erence from juvenile to adult snails, caused 152 BRENDELBERGER by changing specific requirements during an animal’s life history (Tashiro, 1992), is a pos- sible explanation. In previous experiments, Bythinella showed strong preference for diatoms (Bren- delberger, 1992). Diatoms (not tested here) are characterized by their frustules that con- sist mainly of silica. Their carbon content per unit weight is known to be lower than in green algae (Calow, 1970). Thus, Bythinella may not get the appropriate “cues” of good food when fed green algae and detritus only. Al- ternatively, tactile stimuli may be more im- portant for this animal, which occurs in fast- flowing, low-order streams. The importance of contact chemoreception has been shown for Ancylus fluviatilis (Calow, 1973), a snail occurring in the same kind of habitat as Bythinella. Radix and Bithynia, in contrast, may be guided predominantly by distant chemoreception. Assuming that good food is characterized by a low C:N-ratio and a high nitrogen con- tent per unit biomass, Radix peregra and Bithynia tentaculata were found to be optimal foragers. Bythinella dunkeri, in contrast, did not show optimal foraging based on these criteria. But these two variables, C:N-ratio and nitrogen content, are but two of many ways of characterizing natural foods. At cer- tain times, trace element content, vitamins, essential amino acids, and other factors may be more important for an animal than overall organic content alone. It remains to be tested whether the foods thus selected and eaten by the snails are also those that can be better assimilated. ACKNOWLEDGMENTS The patient help of R. Bieg and D. Lúdke, who traced numerous snail-runs, is greatfully acknowledged. W. Lampert made possible the elemental analyses. Many thanks go to E. J. Cox for correcting the English. The study was made possible by financial support from the Deutsche Forschungsgemeinschaft. LITERATURE CITED ALDRIDGE, D. W., 1983, Physiological ecology of freshwater prosobranchs. Pp. 329-358 in: K. M. WILBUR, ed., The Mollusca, Volume 6, Ecology. Academic Press, Orlando. BARLOCHER, F., 1990, Factors that delay coloni- zation of fresh alder leaves by aquatic hypho- mycetes. Archiv für Hydrobiologie, 119: 249- 255. ВЕСОМ, M., J. L. HARPER 4 С. В. TOWNSEND, 1990, Ecology, 2nd ed. Blackwell, Boston. BLEAKNEY, J. S., 1989, Morphological variation in the radula of Placida dentritica (Alder 8 Hancock, 1843) (Opisthobranchia: Ascoglossa/Sacoglos- sa) from Atlantic and Pacific populations. The Veliger, 32: 171-181. BRENDELBERGER, H., 1992, Food selection ex- periments with two freshwater gastropods. Ver- handlungen der Internationalen Vereinigung für Theoretische und Angewandte Limnologie, 25: 4260-4272. и BRENDELBERGER, H. & $. JURGENS, 1993, Sus- pension feeding in Bithynia tentaculata (Proso- branchia, Bithyniidae), as affected by body size, food and temperature. Oecologia, 94: 36-42. CALOW, P., 1970, Studies on the natural diet of Lymnaea pereger obtusa (Кобе!) and its possi- ble ecological implications. Proceedings of the Malacological Society of London, 39: 203-215. CALOW, P., 1973, The food of Ancylus fluviatilis (Müll.), a littoral stone-dwelling herbivore. Oeco- logia, 13: 113-133. CALOW, P., 1973b, Field observations and labo- ratory experiments on the general food require- ments of two species of freshwater snails, Plan- orbis contortus (Linn.) and Ancylus fluviatilis (Müll). Proceedings of the Malacological Society of London, 40: 483-489. DALDORPH, P. W. С. & J. D. THOMAS, 1988, The chemical ecology of some British UK freshwater gastropod molluscs; behavioral responses to short chain carboxylic acids and maltose. Fresh- water Biology, 19: 167-178. EGONMWAN, В. 1., 1992, Food selection in the land snail Limicolaria flammea Müller (Pulmo- nata, Achatinidae). Journal of Molluscan Studies, 58: 49-56. FINDLAY, 5. С. Е. С. & T. L. ARSUFFI, 1989, Mi- crobial growth and detritus transformations dur- ing decomposition of leaf litter in streams. Fresh- water Biology, 21: 261-270. IMRIE, О. W., С. В. MCCROHAN & $. J. HAWKINS, 1990, Feeding behaviour in Littorina littorea: a study of the effects of ingestive conditioning and previous dietary history on food preference and rates of consumption. Hydrobiologia, 193: 191- 198. KAUSHIK, М. К. 8 Н. В. М. HYNES, 1971, The fate of dead leaves that fall into streams. Archiv fúr Hydrobiologie, 68: 465-515. KNECHT, A. 8 J. E. WALTER, 1977, Vergleichende Untersuchungen der Diáten von Lymnaea auric- ularia und Lymnaea peregra (Gastropoda: Ba- sommatophora) im Zürichsee. Schweizerische Zeitschrift für Hydrologie, 39: 299-305. LEVINTON, J. S. 8 T. H. DEWITT, 1989, Relation of particle-size spectrum and food abundance to particle selectivity in the mud snail Hydrobia tot- FOOD PREFERENCE OF FRESHWATER GASTROPODS 153 teni (Prosobranchia: Hydrobiidae). Marine Biol- оду, 100: 449-454. LODGE, D. M., 1986, Selective grazing on periphy- ton: a determinant of freshwater gastropod mi- crodistribution. Freshwater Biology, 16: 831- 841. MADSEN, H., 1992, A comparative study on the food-locating ability of Helisoma duryi, Bi- omphalaria camerunensis and Bulinus truncatus (Pulmonata: Planorbidae). Journal of Applied Ecology, 29: 70-78. MASTERSON, С. & В. FRIED, 1992, Chemoattrac- tion and dietary preferences of Biomphalaria gla- brata (Gastropoda, Planorbidae) for leaf lettuce, Tetramin and egg yolk. Comparative Biochemis- try and Physiology, A, 103: 597-600. MCMAHON, В. F., В. D. HUNTER & W. D. RUS- SELL-HUNTER, 1974, Variation in Aufwuchs at six freshwater habitats in terms of carbon bio- mass and of carbon:nitrogen ratio. Hydrobiolo- gía, 45: 391-404. NEWMAN, В. M., Z. HANSCOM & W. С. KER- FOOT, 1992, The watercress glucosinolate-my- rosinase system: a feeding deterrent to caddis- flies, snails and amphipods. Oecologia, 92: 1-7. NORTON, T. A., S. J. HAWKINS, N. L. MANLY, G. A. WILLIAMS & О. С. WATSON, 1990, Scraping a living: a review of littorinid grazing. Hydrobio- logía, 193: 117-138. PETERSON, С. H. & Р.Е. RENAUD, 1989, Analysis of feeding preference experiments. Oecologia, 80: 82-86. RICHARDSON, W. B., 1990, A comparison of de- tritus processing between permanent and inter- mittent headwater streams. Journal of Freshwa- ter Ecology, 5: 341-358. RUSSELL-HUNTER, W. D. & D. E. BUCKLEY, 1983, Actuarial bioenergetics of nonmarine mol- luscan productivity. Pp. 463-503 in: К. М. WILBUR, ed., The Mollusca, Volume 6, Ecology. Academic Press, Orlando. SKOOG, G., 1978, Influence of natural food items on growth and egg production in brackish water populations of Lymnaea peregra and Theodoxus fluviatilis (Mollusca). Oikos, 31: 340-348. TASHIRO, J. S., 1982, Grazing in Bithynia tentacu- lata: age-specific bioenergetic patterns in repro- ductive partitioning of ingested carbon and ni- trogen. American Midland Naturalist, 107: 133- 150: TASHIRO, J. $. & $. D. COLMAN, 1982, Filter-feed- ing in the freshwater prosobranch snail Bithynia tentaculata: bioenergetic partitioning of ingested carbon and nitrogen. American Midland Natural- ist, 107: 114-132. Revised Ms. accepted 12 July 1994 2 | BR: : | (+ | to ie | = pt ay MALACOLOGIA, 1995, 36(1-2): 155-169 PATTERNS OF LAND SNAIL DISTRIBUTION IN A MONTANE HABITAT ON THE ISLAND OF HAWAII Robert H. Cowie?, Gordon M. Nishida?, Yves Basset? & Samuel М. Gon, ШЗ ABSTRACT A quantitative survey of a 35 km? area between 1,500 m and 2,100 m elevation on the island of Hawaii recorded at least 16 species of land snails. Fifteen of these are probably endemic to the island; one is indigenous but not endemic. Canonical correspondence analysis (CCA) of their local distributions in relation to substratum (i.e., lava) type, altitude, and a suite of vege- tation-related variables, explained 24% of the variance in distribution and abundance. The unexplained variance is probably related to a range of other abiotic, biotic, stochastic and scale factors. Of this 24% overall variance, 79% was explained by axes 1 and 2 of the CCA, which seemed related most strongly to lava type and altitude, respectively. The vegetation-related variables seemed relatively unimportant, although there was a hint that a number of species were negatively associated with the plant community characterized as “Dodonaea shrubland.” Military activities, the presence of introduced feral ungulates, and the increasing trend of invasion by non-native plants, all have the potential to damage this unique fauna. Key words: land snails, Hawaii, ecology, conservation, distribution patterns, canonical cor- respondence analysis. INTRODUCTION The native land snail fauna of the Hawaiian Islands, with 779 recognized species (Cowie et al., in press), is one of the most speciose in the world per unit area (cf. Solem, 1984). Sys- tematic monographs of most of the groups represented are available, but there remain great difficulties in dealing with the fauna be- cause many of these works adopted a now outdated, essentially conchological species concept, which led to considerable over- description of taxa. In addition, and despite this intensive study of some groups, a num- ber of groups remain much less well known, with large numbers of undescribed species (e.g., the Endodontidae: Solem, 1976, 1990). Apart from the notable studies by M. G. Had- field and his colleagues on the growth, de- mographics and population dynamics of cer- {ат species of Achatinellinae (Hadfield & Mountain, 1981; Hadfield, 1986; Hadfield 8 Miller, 1989; Hadfield et al., 1993), under- taken largely with a view to their conservation (see also Severns, 1981), virtually nothing is known of the ecology of the Hawaiian land snails, except what can be gleaned from scattered anecdotal comments in the taxo- nomic literature. The present study was initiated as a simple inventory survey (Cowie & Nishida, 1993), part of a wider environmental impact assess- ment that also included surveys of plants and vertebrates. It is the first study to survey the entire malacofauna of a particular area in the Hawaiian Islands, rather than to focus on par- ticular taxa. lt is also the first time distribu- tions of snails in the Hawaiian Islands have been assessed quantitatively in relation to habitat characteristics. Local patterning in the land snail distributions was analyzed by canonical correspondence analysis (Ter Braak, 1986) in relation to altitude, substra- tum (i.e., lava) type, and to a range of vege- tational variables derived from other parts of the wider environmental assessment. MATERIALS AND METHODS The Study Area The study took place on the island of Ha- waii, the largest and youngest island in the Hawaiian chain (Armstrong, 1983; Clague & Dalrymple, 1987). The study area (Fig. 1) is located in the saddle area between the two largest volcanoes, Mauna Loa and Mauna Contribution number 1994-006 to the Hawaii Biological Survey. “Bishop Museum, P.O. Box 19000-A, Honolulu, Hawaii 96817-0916, U.S.A. “The Nature Conservancy of Hawaii, 1116 Smith Street, Honolulu, Hawaii 96817, U.S.A. 156 COWIE ET AL. Island of Hawaii FIG. 1. Map of the study area showing the location of transects and sampling sites and the type of lava at each site that was surveyed for snails. Open circles — pahoehoe; closed circles — aa. LAND SNAIL DISTRIBUTION PATTERNS IN HAWAII! 157 Kea, just on the leeward (west) side of the saddle and on the slopes of Mauna Loa. It is considered in two parts: the main area and the approach road area, comprising a total of about 35 km”. It slopes gently from an alti- tude of approximately 1,500 m to 2,100 m, with a northwesterly aspect. Rainfall is low, around 500 mm/yr, and varies little over the study area, although declining somewhat with increasing altitude (Armstrong, 1983; Gi- ambelluca et al., 1986). Mean annual temper- ature on the island of Hawaii decreases with increasing altitude at a rate of approximately 5.5°C/1,000 m (Armstrong, 1983), giving a range from about 16°C to 13°C from the lower to the upper parts of the study area. The study area is criss-crossed by lava flows from Mauna Loa, some relatively recent and as yet essentially unvegetated, others older and with established vegetation. Most of the study area falls within a “kipuka,” the “Kip- uka Аза,’ which is an area of climax vege- tation that has not recently been covered with lava, although a number of relatively re- cent and virtually unvegetated lava flows do extend into the study area. Hawaiian lava is of two main types (Peterson & Tilling, 1980), known by their Hawaiian names, that con- ceivably offer different habitat characteristics for snails: smooth “pahoehoe,” derived from rapidly flowing lava, and jagged, broken “aa.” Both types are present in the study area (Fig. 1). The vegetation is montane and subalpine dry shrubland and forest (Gagné & Cuddihy, 1990). Four main vegetational com- munities were recognized (modified from Gagné & Cuddihy, 1990): (1) bare lava or very sparse pioneer vegetation, (2) Metrosideros forest (canopy dominated by Metrosideros polymorpha Gaud. [Myrtaceae]), (3) Sophora/ Myoporum forest (canopy codominated by Sophora chrysophylla (Salisb.) Seem. [Fa- baceae] and Myoporum sandwicense Gray [Муорогасеае)), and (4) Dodonaea shrubland (low canopy dominated by Dodonaea viscosa Jacq. [Sapindaceae)). The study area is part of a military training area. It has been modified locally by con- struction of military facilities, including roads, buildings, targets, and flattened areas sur- faced with crushed lava. Miltary activity in the area has probably also fostered the introduc- tion of non-native weeds, especially grasses (Gagné & Cuddihy, 1990). In addition, non- native feral ungulates have probably had a significant impact on the constitution of the plant communities. Nevertheless, the plant communities, particularly the trees and shrubs, remain essentially native. Sampling Sites The present study was part of a broader zoological and plant community assessment, for which a grid of transects 1.2 km apart was established. These transects were desig- nated M1-8 in the main area and A1-7 in the approach road area (Fig. 1). A total of 104 sites were located and flagged at 250 m in- tervals along these transects. All 21 sites in the access road area (sites 1-21 of the present study) and 37 of the 83 sites in the main area (sites 22-57, plus an additional site “x"—see below) were sampled for land snails (Fig. 1). Selection of sampling sites for snails in the main area was designed to give a relatively even rather than random cover- age of the area. None of the sampled sites was located in the immediate vicinity of an area highly modified by military construction. Sampling Protocol Field work took place between 8 March 1992 and 9 February 1993. Two samples, collected by searching through litter and soil, often in, around and under rocks (turning them over or removing them, especially in ar- eas of aa lava), were taken at each of the 58 sampling sites (except one site on transect M6—site ‘‘x’’—where only one sample was taken, and four sites—sites 15, 16, 31, 57— where no snails were found). Each sample was taken by one person working for 1 h, having identified appropriate habitat within 10 m of the flagged site. Although there may be unconscious differences in sampling strategy between collectors, sampling effort at every site was consistent because one sample was always taken by each of the first two authors. Most samples were taken from an area of about 0.5 m°, but if snails/shells were scarce the area covered in 1 h was larger. In addition, 15 min searches for tree snails, looking at leaves, trunks and under bark, and covering 10 trees or shrubs within 30 т of the flag, were conducted at every site sampled. Litter/soil samples from a 30 x 30 cm area were collected at five sites. Because sampling took place during day-time and most snail species are essentially nocturnal, snail distributions necessarily represent rest- ing sites rather than sites of activity. How- ever, given that most of the species are very 158 COWIE ET AL. small (2-10 mm) and probably do not travel far, the scale of the sampling protocol would be unlikely to allow resting and active sites to be distinguished. Sorting and Identification of Material All specimens collected during the field trips were sorted, counted, and identified as far as possible. Litter samples were sifted us- ing standard mesh screens of decreasing mesh size, followed by scanning of all sifted soil and litter. Identifications were made by reference to the extensive malacological col- lections of the Bishop Museum (Honolulu), and to appropriate literature. Specimens were recorded as “live,” “dead recent” (shell with at least half the periostracum still present, and the shell retaining its original color), and “dead old” (less than half the pe- riostracum still present, and/or shell opaque white). Only intact shells or fragments of shells containing the shell apex, and identi- fible non-apical fragments of species other- wise not represented in a particular sample, were counted (cf. Christensen & Kirch, 1986). This protocol removes the possibility of counting a single shell more than once, but does leave uncounted a number of readily identifiable specimens. Nevertheless, it is deemed a more rigorous approach, did not exclude a large amount of identifiable mate- rial, and therefore does not influence the overall conclusions of the study. Despite extensive previous systematic study of the Hawaiian land snail fauna, there remain many undescribed species. In partic- ular, the island of Hawaii is less well known malacologically than the other islands (Cowie, in press). In addition, the extent of intra-specific variation 1$ often unknown. Therefore, it is frequently difficult to identify material to a recognized species. This 1$ the case in the present study. However, most of the material collected could be assigned to distinct ““morphospecies,” even though a specific name could not be applied with cer- tainty. All material is deposited in the mala- cology collections of the Bishop Museum (TL-1994.050). Environmental Variables Six environmental variables were recorded at each site. Values of the vegetation-related variables were derived from data accumu- lated as part of the non-malacological as- pects of the overall environmental assess- ment. (1) Altitude: Taken from a 1:50,000 map of the Pohakuloa training area with contour in- tervals of 12.2 m (40 ft). (2) Canopy height: Scored in the field in 8 classes, from 1 for no canopy, to 8 for can- ору height greater than 10 т. (3) Canopy closure: Scored in the field in 12 classes, from 1 if the site was completely open, to 12 ifthe canopy was more than 50% closed. (4) Vegetational community: Four commu- nities were recognized, based on the domi- nant trees or shrubs (see introduction). They were coded as follows: 1—bare lava or very sparse pioneer vegetation, 2—Metrosideros forest, 3—Sophora/Myoporum forest, 4— Dodonaea shrubland. (5) Vegetational heterogeneity: A measure of combined canopy and understory hetero- geneity, reflecting the overall vegetational heterogeneity of the site, was obtained as fol- lows. For the canopy, a score of 1 was given when one canopy species was dominant and a score of two when two or more species were codominant. The understory was cate- gorized into bare substratum, native shrubs, native grasses, native ferns, native herbs, alien shrubs, alien grasses, alien herbs, alien vines. One or as many as four of these nine elements could be considered dominant or codominant at a particular site, giving a score of 1-4 for increasing understory heterogene- ity. By adding the canopy and understory scores, the combined heterogeneity score therefore ranged between 2 and 6. (6) Lava type: The substratum from which the actual samples were taken, ¡.e., pahoe- hoe (coded as 1) or aa (coded as 2) (Fig. 1). Statistical Analysis There were only minor differences in the presence/absence of particular species be- tween the two samples taken at each site. However, log-likelihood G statistic analysis (Rohlf & Sokal, 1969; Sokal & Rohlf, 1981) of the 46 sites at which there were sufficient numbers of specimens, indicated highly sig- nificant differences in relative abundances between the two samples (p < 0.001 in 30 cases, p < 0.01 in six cases, p < 0.05 in two cases), due perhaps both to different biases between the two people performing the sam- pling or to differences in microhabitat be- tween the two sample locations. Neverthe- LAND SNAIL DISTRIBUTION PATTERNS IN HAWAII 159 less, in order to obtain a more general picture of the fauna at each site and to relate the snail distributions to environmental variables at the meso-scale of the sites rather than the micro-scale of the individual samples, the data for the two samples at each site have been combined for the purpose of the follow- ing analysis. The program ADE 3.6 (Chessel 8 Dolédec, 1993) was used to carry out a canonical cor- respondence analysis (CCA; Ter Braak, 1986; Lebreton et al., 1988; Palmer, 1993) on the abundance (combined number of live and dead individuals) of each species at each site in relation to the six environmental variables indicated above. CCA 1$ particularly appro- priate when species show non-linear rela- tionships with environmental variables (Ter Braak, 1986), as is recognized may often be the case in studies of molluscs (Bishop, 1981). CCA is designed for gradient analy- ses, that is, analyses of species distributions along environmental gradients. Neither veg- etational community nor lava type is a gradi- ent variable. However, they have been incor- porated into the CCA as pseudo-gradient variables for exploratory purposes. The four sites at which no snails were found (sites 16, 17, 31, 57) were included in the CCA, but the single site (site “х’”) from which only one sample was available was excluded because snail abundance at that site would be under- estimated. Also, specimens referred to “Lep- tachatina sp.” were excluded as being uni- dentified specimens that probably belonged to the other Leptachatina spp. recorded. RESULTS Taxonomy At least 16 species are represented in the collections (Table 1). All but one of them are probably endemic to the Hawaiian Islands (Vitrina tenella is native but not endemic.) Their classification here follows Cowie et al. (in press). The taxonomy of many of the groups is uncertain; no previous collections have been made in the area of the study, and, with the often highly localized distributions of Hawaiian land snail species, it is probable that a number of the species found are un- described. Problematic taxa are now briefly discussed. Lamellidea sp.: Only three species of Lamel- lidea have been recorded from the island of Hawaii: L. gracilis (Pease, 1871), L. oblonga TABLE 1. Land snail taxa found during the survey. ACHATINELLIDAE Lamellidea sp. Tornatellides sp(p). AMASTRIDAE Leptachatina (L.) lepida Cooke, 1910 Leptachatina (Angulidens) anceyana Cooke, 1910 Leptachatina sp. A Leptachatina sp. B Leptachatina sp. C Leptachatina sp. PUPILLIDAE Nesopupa (Infranesopupa) subcentralis Cooke 8 Pilsbry, 1920 Pronesopupa sp. SUCCINEIDAE Succinea konaensis Sykes, 1897 HELICARIONIDAE Euconulus (Nesoconulus) sp. cf. gaetanoi (Pilsbry & Vanatta, 1908) Philonesia sp. ZONITIDAE Nesovitrea hawaiiensis (Ancey, 1904) Striatura (Pseudohyalina) sp. cf. meniscus (Ancey, 1904) ?Striatura Sp. Vitrina tenella Gould, 1846 (Pease, 1865) and L. peponum (Gould, 1847). Their shell morphology is similar, but material in the Bishop Museum shows a range of vari- ation both within and among individual lots, including type lots, with some overlap be- tween lots referred to different species. Both L. gracilis and L. oblonga have been consid- ered lowland species that probably do not reach the altitude of the study area (Cooke & Kondo, 1960), but it is not possible to identify the present material more precisely. Tornatellides sp(p).: The genus Tornatellides can be difficult to distinguish from other closely related genera, particularly Tornatel- laria. However, Tornatellides bears live young, whereas Tornatellaria lays eggs (Cooke & Kondo, 1960). The present material manifests some variation, especially in size, but is all tentatively referred to the genus Tor- natellides, because embryos were found in- side some individuals. The variation exhibited perhaps suggests more than one species, although this variation may yet be intra- specific, with only a single, rather variable species being represented. Referral to partic- ular species is not possible. 160 COWIE ET AL. Leptachatina sp. A: This species, although somewhat variable in size, appears to be dis- tinct. It is rather tall and narrow with a rela- tively large protoconch. It is somewhat simi- lar to L. imitatrix Sykes, 1900, but probably represents an undescribed species. Leptachatina sp. B: Specimens assigned to this species appear somewhat intermediate between L. lepida Cooke, 1910, and L. ko- naensis Sykes, 1900, being fatter than the former but thinner than the latter. They may belong to one of these species, or may rep- resent an undescribed species, but a firm de- cision depends on future revision of the group. Leptachatina sp. C: Specimens assigned to this species are similar to but appear distinct from L. lepida. They are large, with a rather straight, not convex, outline to the shell spire, and perhaps represent an undescribed spe- cies. Further taxonomic research will be re- quired to confirm their true status. Pronesopupa (Sericipupa) sp.: Referral of these specimens to subgenus appears fairly secure. However, they do not correspond precisely to any of the three species—P. ly- maniana Cooke 4 Pilsbry, 1920, P. orycta Cooke 4 Pilsbry, 1920, P. sericata Cooke and Pilsbry, 1920—described from the island of Hawaii and may represent an undescribed species. Euconulus (Nesoconulus) sp. cf. gaetanoi: The present material does not correspond precisely to anything in the Bishop Museum collections but is nevertheless very close to E. gaetanoi. It may or may not be a distinct species. Striatura (Pseudohyalina) sp. cf. meniscus: Type material of S. meniscus, held at the Bishop Museum, contains a range of mor- phological variation, especially in umbilicus width, and in fact seems to include two spe- cies. The holotype has a wide umbilicus, whereas the specimens from the present sur- vey correspond very closely to those para- types with the narrower umbilicus, which re- semble S. pugetensis Dall, 1895. Detailed taxonomic work, beyond the scope of this ecological study, is necessary to decide whether the present specimens are indeed S. meniscus or S. pugetensis, or whether they belong to a further, closely related, but pos- sibly undescribed species. Baker (1941) hinted at this confusion. ?Striatura sp.: Specimens from the present survey, distinct from the previous species, nevertheless appear closely related to it con- chologically and so are tentatively assigned to the genus Striatura. Material of this spe- cies in the Bishop Museum collections has been labeled S. meniscus, but incorrectly. The survey specimens do not correspond to anything in the type collections of Striatura in the Bishop Museum nor to the written treat- ment of Baker (1941). They cannot be iden- tified further and may belong to an unde- scribed species. Philonesia sp.: Although clearly belonging to the genus Philonesia, the present material, apparently of one species only, does not cor- respond closely with material of any of the Philonesia spp. from the island of Hawaii held in the Bishop Museum collections. It is pos- sibly an undescribed species. Overall Abundance A total of 12,273 specimens (252 live) was collected by hand searching in the field, with an additional 2,342 (563 live) being extracted from the soil/litter samples. No arboreal spe- cies were found, although ?Striatura sp. was found on tree trunks on one occasion at one site. The raw data are held by the first author and summarized in the appendices of Cowie 8 Nishida (1993). The vast majority of specimens collected were empty shells. Only 10 of the 16 taxa were recorded live, and then, except in the case of ?Striatura sp., which was the most common species found in the survey, only with very few live individuals. Species rich- ness at a particular site ranged from two to 12 species. In only one instance did the litter/soil sam- ples detect a taxon not represented in the regular field samples at a particular site (Tor- natellides sp. found in very low numbers at site 25). п all other cases, the regular sam- ples detected more species than the soil/lit- ter samples. Not unexpectedly, however, of the species recorded from the soil/litter sam- ples, relatively greater numbers of the smaller species were recorded in these samples compared to the regular samples. In one in- stance (site 45), the soil/litter sample de- tected a species alive (Pronesopupa sp. three live out of a total of 34) that was only represented by dead (dead recent) shells in the regular field samples at that site. LAND SNAIL DISTRIBUTION PATTERNS IN HAWAII 161 Eigenvalue FIG. 2. Partitioning of eigenvalues across the six axes extracted by the CCA. Lamellidea sp., Succinea konaensis and ?Striatura sp. were the most abundant (over 1,000 specimens of each), although in the case of Lamellidea sp. by no means ubiqui- tously distributed. The rarest species (less than 100 specimens each) were Leptachatina anceyana, Leptachatina sp. B, Leptachatina sp. C, Pronesopupa sp., Euconulus sp. cf. gaetanoi and Vitrina tenella. The remaining species were intermediate in abundance. Patterns of Distribution General Patterns Detected by the CCA: The total eigenvalue for the six axes extracted by the CCA is 0.474, partitioned according to Figure 2. Much of the variance (79%) in spe- cies abundance by site, as constrained by the environmental variables incorporated in the CCA, was explained by the first two axes. The ordination diagram (Fig. 3) describes the relations among species and sites, as related to the six environmental variables, on the first two axes of the CCA. In addition to the CCA, a correspondence analysis (CA) was performed on the abun- dance (combined number of live and dead individuals) of each species at each site, and the results compared to those of the CCA, as recommended by Ter Braak (1986). The spe- cies scores on axis 1 and axis 2 of the CA were not highly correlated with those of the CCA (г = —0.466, р = 0.069 and г = —0.039, р = 0.887, respectively, п = 16). The correla- tions for axes 3 and 4 were better (r = 0.488, р — 0.055 and fF =0:651; P= 0.006, respectively), but these axes only explained a small additional amount of the overall vari- ance (Fig. 2). This poor correlation for axes 1 and 2 weakens the robustness of the CCA, which must therefore be evaluated cau- tiously. Following Borcard et al. (1992), it is possible to estimate the proportion of the to- tal variance explained by the environmental variables by dividing the total eigenvalue of the CCA (0.474) by that of the CA (1.944). This indicates that 24% of the overall vari- ance in species abundances is explained by the current environmental variables and that 79% of this is explained by axes 1 and 2 of the CCA. Interpretation of the ordination diagrams generated by a CCA is clearly explained by Ter Braak (1986). The length and direction of an arrow representing an environmental vari- able indicates the importance of the variable in the formation of the axes: the smaller the angle between the arrow for a particular vari- able and an axis, the larger the contribution of the variable to that axis; the longer the arrow relative to other arrows, the greater the contribution. The score of a species or site on a particular environmental variable is deter- mined by dropping a perpendicular from the species or site point to the arrow (or to the imagined extension of the arrow) represent- ing that variable. A high score (positive or negative) on the arrow represents a strong association of the species or site with that variable. In the present case, axis 1 is closely related to lava type, and lava type is the most impor- tant variable among those included in the study (Fig. 3, Table 2). Both the canonical coefficients and intraset correlation coeffi- cients (Table 2) for lava type are distinctly greater than those for any other variable. In the ordination diagram (Fig. 3), all sites to the left of the altitude and canopy height arrows are on aa lava and all to the right are on pa- hoehoe. Axis 2 is more difficult to interpret but ap- pears most closely related to altitude (Table 2). Inspection of the ordination diagram (Fig. 3) reveals indeed that those sites falling in the upper part of the diagram are high altitude sites whereas those in the lower part are low altitude sites. The ordination diagrams for axes 3 and 4 are not presented, because these axes make only minor contributions to explaining the overall variance. They are difficult to interpret because neither is clearly related to just a single variable, although axis 3 may be a composite of the vegetational variables. Inter-specific Associations and Overall Strength of Environmental Associations: The ordination diagram (Fig. 3) indicates no clear clusters of species, suggesting that there are 162 COWIE ET AL. Altitude AXIS 2 Canopy closure AXIS 1 FIG. 3. Ordination diagram of site and species distributions on axes 1 and 2 of the CCA. Sites are indicated by the numbers as given in Figure 1. Snail species are represented as follows: A—Lamellidea sp., B—Tor- natellides sp(p)., C—Leptachatina lepida, D—Leptachatina anceyana, E—Leptachatina sp. A, F—Lepta- chatina sp. В, G—Leptachatina sp. С, H—Nesopupa subcentralis, l—Pronesopupa sp., J—Succinea ko- naensis, K—Euconulus sp. cf. gaetanoi, L—Philonesia sp., M—Nesovitrea hawaiiensis, N—Striatura sp. cf. meniscus, O—?Striatura sp., P—Vitrina tenella. The arrows representing the environmental variables are scaled up by a factor of 6.49 for clarity of presentation (see Ter Braak, 1986). no strong associations among snail species, that is, no clear sub-communities. The spe- cies all plot rather close to the center of the ordination diagram (Fig. 3), also indicating that none of them has a particularly strong association with any of the environmental variables incorporated in the analysis. Asso- ciations with the most significant variables are presented below. Local Rarity and Patchiness: Certain species were found only in very low numbers and/or at very few sites. This rarity may be only a local phenomenon and not reflect overall rar- ity on the island of Hawaii. Some apparent patterning in the distributions of these rare species may simply be due to sampling error. For instance, Pronesopupa sp. was recorded in very low numbers (< 10 at any one site) LAND SNAIL DISTRIBUTION PATTERNS IN HAWAII 163 TABLE 2. Canonical coefficients and intraset correlation coefficients (see Ter Braak, 1986) of the environ- mental variables with the first four axes of the CCA. Canonical coefficients Correlation coefficients Variable Axis 1 Axis 2 Axis3 Axis4 Axis 1 Axis 2 Axis 3 Axis 4 Altitude — 0.33 0.54 0.02 0.80 -0.22 0.82 0.09 0.56 Canopy height 0:07, 0:52 112 0.43 0.10 —0.44 0.63 0.01 Canopy closure — 0.08 —0.54 —0.16 —0.06 0.42 —0.20 0.33 —0.41 Vegetational community 0.38 0.15 1.33 —0.01 0.50 0.38 0.53 —0.34 Vegetational heterogeneity — 0.04 0.47 —0.43 —1.02 0.37 0.24 0.54 —0.43 Lava type —0.87 —0.09 0.35 —0.48 -092 —-0.12 —0.03 -0.20 from only four sites in the higher parts of the main study area (sites 41, 45, 52, 56) and from only three along the approach road (2, 9, 10). Of these seven sites, six were on aa lava. This apparently disjunct distribution, and the high score on the lava type arrow may therefore simply be sampling artifacts. The high scores of certain species on the altitude arrow are possibly reflections of their rarity. For instance, Leptachatina anceyana was only found, in low numbers, at five sites towards the lower part of the main study area (sites 24, 29, 30, 32, 37). Leptachatina sp. B was recorded from only six sites (15, 24, 43- 46), although four of these were on a single transect (M5) and might reflect real patchi- ness. Euconulus sp. cf. gaetanoi was re- corded at only two sites (27, 45) and the rel- atively high score on the altitude arrow probably reflects its greater abundance at the higher of these sites. Vitrina tenella was also collected at only two sites (44, 45). However, these were close together on a single transect (M5); both were on pahoehoe lava and in Sophora/Myoporum forest. If this spe- cies were widely but sparsely distributed over the whole study area, one would not have expected the two collection localities to be so close together, perhaps suggesting that this is indeed a very localized distribu- tion. The absence of rare species from certain vegetational communities (Table 3) may well also be a sampling artefact related not only to the overall rarity of the snail species but also to the different numbers of sites of each com- munity type that were sampled. Only Vitrina tenella is confined to a single vegetational community, but this may be an artefact of its extreme rarity in the study area. Absence of Leptachatina sp. C, Pronesopupa sp. and Euconulus sp. cf. gaetanoi from communities characterized as Dodonaea shrubland and as bare lava or very sparse pioneer vegetation TABLE 3. Presence (+) or absence (0) of land snail taxa with vegetational community, coded as fol- lows: 1—bare lava or sparse pioneer vegetation; 2—Metrosideros forest; 3—Sophora/Myoporum forest; 4—Dodonaea shrubland. Vegetational community Land snail taxa 1 2 3 4 Lamellidea sp. - + + + Tornatellides sp(p). + + = + Leptachatina lepida + + - - Leptachatina anceyana 0 + + + Leptachatina sp. A + + + 0 Leptachatina sp. B + + + 0 Leptachatina sp. С 0 + + 0 Leptachatina sp. + + + + Nesopupa subcentralis + + + + Pronesopupa sp. 0 + + 0 Succinea konaensis + + - + Euconulus sp. cf. gaetanoi 0 + 0 0 Philonesia sp. + = - - Nesovitrea hawaliensis - + - 0 Striatura sp. cf. meniscus + + + - ?Striatura sp. + + + + Vitrina tenella 0 0 + 0 may be a reflection of both the overall rarity of these species and the relatively few sites of these communities that were sampled. Ab- sence of Leptachatina sp. B from Dodonaea shrubland and Leptachatina anceyana from bare lava or very sparse pioneer vegetation may be explained in a similar way. However, the absence of the more common Leptacha- tina sp. A and Nesovitrea hawaliensis from Dodonaea shrubland may reflect a real phe- nomenon (see below). Lava Type: All but two species were found on both lava types. Euconulus sp. cf. gaet- апо! was only recorded from aa and Lepta- chatina anceyana only from pahoehoe. Both these species score highly on axis 1 of the CCA (Fig. 3), which appears closely related to 164 COWIE ET AL. lava type, but both are so rare (only recorded at two and five sites, respectively) that this apparent association may be due to chance and of no real biological meaning. Other spe- cies scoring highly on axis 1 are Leptachatina sp. A, Leptachatina sp. C, Pronesopupa sp., Philonesia sp., Nesovitrea hawaiiensis and ?Striatura sp. The distributions of Leptacha- tina sp. A, Philonesia sp. and Nesovitrea ha- waliensis are somewhat similar to each other and show similar associations with lava type, as follows. Sixteen of the 21 sites at which Leptachatina sp. A was found, 17 ofthe 23 at which Philonesia sp. was found, and 19 of the 26 at which N. hawaliensis was found, were on aa. All these associations are statis- tically significant (G tests; p < 0.005 in all cases). Leptachatina sp. C and Pronesopupa sp. were found at five sites (four of them aa) and seven sites (six of them aa), respectively, but this is too few for G statistic analysis. All these species that appear to show an asso- ciation with aa plot on the ordination diagram (Fig. 3) among the left hand cluster of sites, all of which are on aa. Two species, Lepta- chatina anceyana and ?Striatura sp., plot among the pahoehoe sites on the ordination diagram. Of these two species, L. anceyana occurred at very few sites (see above), but ?Striatura sp. was both widespread and abundant, and, although in terms of pres- ence/absence at aa or pahoehoe sites it showed no significant association, it oc- curred in significantly greater numbers on pa- hoehoe (G test: p < 0.001). Altitude: Taxa scoring highly on the altitude arrow (Fig. 3) are Lamellidea sp., Leptacha- tina anceyana, Leptachatina sp. B, Euconulus cp. cf, gaetanoi and Vitrina tenella. Of these, only Lamellidea sp. is at all common, being overall the second most abundant species recorded. It is completely restricted to the lower parts of the study area (all approach road transects and transects M1-3). The high scores of the other species are possibly re- flections of their rarity. Vegetation: The presence/absence of snail species, according to the vegetational com- munity at each site are presented in Table 3. The CCA indicated no strong influence of any of the vegetational variables on snail species distributions, although the ordination dia- gram (Fig. 3) clearly grouped the sites char- acterized as Dodonaea shrubland to the far right and those as bare lava or very sparse pioneer vegetation to the left, with the other two vegetation types scattered between them. None of the snail species appears par- ticularly associated with any vegetational community, nor with any other vegetational characteristic (but see below). Nevertheless, direct inspection of the data indicates that Leptachatina sp. A, Philonesia sp. and Nesovitrea hawaliensis have some- what similar distributions, being absent or nearly absent from the central and more northerly sites on transects M2-4. This gap in the distributions of these three species ap- pears roughly to correlate with the presence of Leptachatina anceyana (only recorded in the lower part of the main study area, at a total of five sites on transects M1-4) and with a concentration of Dodonaea shrubland along the more northern end of transect M3. (Casual observations suggested that this part of the study area also supported the highest concentration of feral sheep, which may have had an impact on both physical and chemical soil characteristics.) These patterns perhaps suggest some ecological interaction or differ- ential habitat preferences between the snail species. Leptachatina anceyana plots to the far right of the ordination diagram (Fig. 3), as do Dodonaea shrubland sites. Leptachatina sp. A, Philonesia sp. and Nesovitrea hawaliensis, as well as perhaps having a negative association with Dodonaea shrubland, are all positively associated with aa lava. In addition, all five sites at which Lep- tachatina anceyana was recorded were on pahoehoe. Because the CCA indicated a much greater influence of lava type, the rela- tionship with lava type may be more impor- tant than the apparent relationship with veg- etational community for these species. DISCUSSION Although the land snail fauna of the Hawai- ian islands is one of the most species-rich in the world for an area of comparable size (cf. Solem, 1984), the local species richness re- corded in the present study is not excep- tional. Only 16 species were found, with no more than 12 at any one collection site. While truly valid comparisons of species richness can only be made in terms of area (and sam- pling effort), this local species richness seems comparable to that found in similar sampling programs in many other parts of the world, that is, ranging up to about 30 species LAND SNAIL DISTRIBUTION PATTERNS IN HAWAII! 165 but usually fewer (e.g., Bishop, 1981; Solem, 1984; Cameron, 1992). Greater numbers might be found in particularly diverse areas, but the species might not be truly sympatric on a small scale. The most notable exception is the finding of over 70 species that were indeed suggested as being “microsympat- ric”” in small patches of forest in New Zealand (Solem et al., 1981; Solem, 1984; Solem & Climo, 1985); and a few other regions (in the Caribbean and Australia) are also known to support high levels of microsympatric snail species (Solem, 1984). The relatively low numbers in the present survey, combined with the extraordinary species richness in the Hawaiian islands as a whole (Cowie et al., in press), reflect the fact that most Hawaiian land snail species are highly localized either geographically (i.e., particular parts of an is- land, a valley, ridge, etc.) or ecologically (i.e., lowland, rainforest, etc.). Additionally there- fore, because collections have not previously been made in the area of the study, it is not surprising that much of the present material appears to represent undescribed species. The comparison of the total eigenvalues of the CA and the CCA indicates that the envi- ronmental variables incorporated in the study explained 24% of the overall variance in spe- cies abundance by site. The remaining vari- ance may be partly explained by other abiotic and biotic factors, as well as by stochastic variation, especially related to historical fac- tors (cf. Bishop, 1981). Such abiotic factors as pH, calcium availability and soil humidity, are known to influence snail distributions elsewhere, although their effects are not al- ways straightforward (Cameron, 1978; Peake, 1978; Bishop, 1981; Cain, 1983). Un- fortunately, it was not possible to obtain ap- propriate data to incorporate these variables into the present study. Such biotic factors as competition and predation, have only rarely been demonstrated as influencing the spatial distributions or abundances of land snails (Mordan, 1977; Peake, 1978; Cain, 1983; Cowie & Jones, 1987; Smallridge & Kirby, 1988). Historical factors have been shown to be important (Cameron 4 Dillon, 1984), but very few studies have addressed this ques- tion. It is not possible to speculate on the relative importance of these factors in rela- tion to the unexplained variance in the present study. It is not uncommon in studies of this kind for a relatively large part of the variance to remain unexplained; but the fac- tors that are found to have significant influ- ences may yet be important in structuring the community under study (Borcard et al., 1992). Nevertheless, it is also possible that the distributions of the species are heavily influenced by environmental variability on a much finer micro-scale than incorporated in the present analysis. This is suggested by the highly significant differences in relative abun- dances of species between the two samples taken at each site. Of the 24% of the overall variance at the meso-scale explained by the factors т- cluded in the study, 79% is explained by the first two axes of the CCA. These two axes appeared to be related most closely to lava type and altitude. A number of associations of certain spe- cies with lava type are clear. However, it is not clear exactly what 1$ the real variable, as- sociated with lava type, that 15 influencing these associations. The physical characteris- tics of the two types of lava are very different and may be important for the snails. The smooth surfaces of pahoehoe provide littie microhabitat for snails, and in areas of pahoe- hoe most snails/shells were found in places where there was shade, moisture and an ac- cumulation of litter and soil, such as in the cracks in the lava or at the bases of slabs of lava. The broken nature of aa lava provides much greater possibilities for shade, but of- ten any soil and litter was found only deep down after removing many chunks of lava. It could be argued that pahoehoe is more vari- able physically. Young pahoehoe is smooth, but as it ages it can break down into small rocks and boulders that perhaps offer habi- tats more akin to aa. This greater physical variability may be reflected in the wider spread of pahoehoe sites than of aa sites on axis 1 of the CCA (Fig. 3). There is no signif- icant difference in chemical composition be- tween the two lava types (Peterson 4 Tilling, 1980; Vitousek et al., 1992), so soil pH is not influenced directly by lava type. However, the vegetation supported on the two types of lava may differ (Karpa & Vitousek, 1994), as was suggested by the CCA. Different vege- tation might influence such factors as soil chemistry, and depth and decomposition rate of litter, but the vegetation-related variables incorporated in the study were not strongly related to the snail faunal composition. Vi- tousek et al. (1992), whose study area en- compassed that of the present study, found differences among sites in accumulation of carbon, nitrogen and phosphorus in soils, 166 COWIE ET AL. availability of soil nutrients, and in foliar nu- trients of Metrosideros polymorpha, the dom- inant tree of their study. Although some of these differences were related to altitude, lava type and flow age, there seemed to be no consistent pattern, and our understanding of variations in these factors on the dry, northwest slopes of Mauna Loa remains poor (Vitousek et al., 1992). Karpa 8 Vitousek (1994) hinted at other possible differences between the lava types (local flooding and susceptibility of the vegetation to fire), but it seems unjustified to speculate further on the importance of these rather poorly under- stood environmental variables in influencing the distributions of the snail species in the present study. Only one species, Lamellidea sp., is suffi- ciently abundant to suggest reliably that its recorded distribution relates to altitude. It is possible that the study area located the up- per altitudinal boundary of this species, since, although it could not be decisively identified, at least two of the three species of Lamellidea previously recorded from the is- land of Hawaii are lowland taxa (Cooke & Kondo, 1960). This clear relation of the dis- tribution of Lamellidea sp. to altitude and the more tentative overall association of the fau- nal composition with altitude may be related to such factors as temperature and rainfall. Cameron (1978) indicated decreasing land snail species richness at higher altitudes in his study area in England and implied a rela- tionship with local climate. Certainly climatic factors have frequently been considered of fundamental importance in determining snail distributions (e.g., Peake, 1978; Arad, 1990; Asami, 1993; Baur & Baur, 1993). There is a gradient, at least in temperature and perhaps in rainfall, related to altitude in the study site, although the range is small (Armstrong, 1983; Giambelluca et al., 1986), and variation on a microhabitat scale might be more important. But, in the absence of appropriate data on temperature tolerance, resistance to desic- cation, and other factors, of the snail species (cf. Baur & Baur, 1993), further speculation is not justified. The lack of a clear relationship with vege- tational community or any other vegetational variable, except for the extremely tentative associations (both positive and negative) of some species with Dodonaea shrubland, is a little surprising. This lack of relationship sug- gests that the significant differences between the two samples at each site might be related to environmental heterogeneity on a much smaller, micro-scale. The land snail fauna of the Hawaiian islands is recognized as being under serious threat of extinction, with many species already gone (Hadfield, 1986; Solem, 1990). The vast ma- jority of specimens collected were empty shells. The recording of empty shells as “dead old” and “dead recent” was done in an at- tempt to get some feel for the likely recent and perhaps continuing presence of species that were not recorded live (six out of 16). How- ever, it is not known how long it takes for shells to lose their periostracum and to turn white and opaque. It may well be a matter of years rather than weeks or months, and will probably differ among taxa and among local- ities according to such things as exposure to sunlight, rainfall and soil acidity. However, the fact that allspecies, even ifnot collected alive, were recorded as “dead recent” at at least one site suggests that all the species re- corded in the study area are probably still extant. Relative rarity in the study area may serve as an explanation of the absence of live individuals for Vitrina tenella (only 2 speci- mens found), Leptachatina sp. B (14 speci- mens), Euconulus sp. cf. gaetanoi (25), and perhaps for Leptachatina anceyana (46) and Leptachatina sp. C (69). Leptachatina sp. A was also not found alive but occurred in somewhat higher abundance (234 specimens in the field samples), although it could not be considered common. Species found in abun- dance but only, or almost only, as dead shells may be extinct or closer to extinction than rare species that were nevertheless found alive, or indeed than common species found in high numbers both dead and alive. For instance, ?Striatura sp. is the most abundant species in terms of both live snails and dead shells, but Lamellidea sp., the second most abundant in terms of dead shells, was among the rarer species in terms of numbers of live snails found (eighth out of the ten species collected alive). This high relative number of Lamellidea sp. shells might reflect a recent increase in the mortality of this species and the possibility that it may become locally extinct in the near future. But it might also reflect the possibility that Lamellidea sp. shells do not disintegrate as rapidly as those of ?Striatura sp. Such pos- sibilities can only be extremely speculative because nothing is known of such factors as relative differences in rates of shell weathering and breakdown among different species, dif- ferences in life histories and rates of mortality. LAND SNAIL DISTRIBUTION PATTERNS IN HAWAII 167 GENERAL CONCLUSIONS The land snail community of the study area is composed mostly of species endemic to the island of Hawaii. Significant relationships between their distribution within the study area and at least two environmental variables (lava type and altitude) have been demon- strated. Identification of survey material by reference to the Bishop Museum collections (there are no previous collections specifically from the survey area) and assessments of their previously known distributions suggest that the species recorded, while not unique to the study area, are representative of a fauna characteristic of the Kona side of the island of Hawaii and perhaps more specifi- cally to the Hualalai-Puuwaawaa area. How- ever, much of the material in the Museum was collected many years ago. Nothing is known of the current status of the species at those earlier collecting localities. With the in- creasing impacts of alien plants and animals introduced to the Hawaiian Islands (Cowie, 1992), it is quite possible that these snail spe- cies have declined or gone extinct in these localities. The present survey area, especially for those species recorded alive, is therefore an important part of their known distribution. It is the only area in the Hawaiian Islands for which such a detailed faunistic survey of land snails has been carried out. The finding of living Leptachatina lepida 1$ particularly noteworthy because the study area is the only known locality at which this species is known to be still extant. Leptacha- tina lepida is only one of as few as perhaps six or ten extant species of the Hawalian en- demic family Amastridae, which once num- bered over 400 species-group taxa (Cowie et al., 1994). Amastrids, and perhaps the genus Leptachatina in particular, seem highly sus- ceptible to habitat modification (Christensen & Kirch, 1986). The survey area 1$ therefore of particular significance in the preservation of what remains of this unique and once highly diverse family. The survey was conducted as part of an assessment of the potential environmental impact of military activities on the fauna. Mil- itary use of the study area has had and will continue to have direct impacts on the snails (e.g., explosions, construction work). Fur- thermore, and perhaps most significantly, it will probably result in habitat change that is likely to be highly detrimental to the fauna. In addition, other factors, such as grazing by introduced feral ungulates (Cameron, 1978), changes in the floral composition of the area due to the increasing numbers of introduced species (Karpa & Vitousek, 1994), and preda- tion by rodents (Hadfield et al., 1993), ants (Solem, 1976) and perhaps introduced galli- naceous birds (Buckle, 1989), are all serious threats to this unique community. ACKNOWLEDGMENTS We thank the Nature Conservancy of Ha- май (TNC), especially Theresa Cabrera, Bill Garnett, Luciana Honigman, Joan Yoshioka and Dan Zevin for facilitating the field work. Arthur Cain commented on the manuscript and Nancy Young assisted with the produc- tion of Figure 1. The U. S. Army, through TNC, funded the field work and the initial analysis and report writing as part of an as- sessment of the potential impact of military activities on a largely pristine Hawalian envi- ronment. LITERATURE CITED ARAD, Z., 1990, Resistance to desiccation and dis- tribution patterns in bush-dwelling snails. Jour- nal of Zoology, 221: 113-124. ARMSTRONG, В. W., ed., 1983, Atlas of Hawaii, 2nd edition. University of Hawaii Press, Hono- lulu. 238 pp. ASAMI, T., 1993, Interspecific differences in des- iccation tolerance of juvenile land snails. Func- tional Ecology, 7: 571-577. BAKER, Н. B., 1941, Zonitid snails from Pacific is- lands, parts 3 and 4. Bernice P. Bishop Museum Bulletin, 166: 205-370, pls. 43-65. BAUR, B. 8 A. BAUR, 1993, Climatic warming due to thermal radiation from an urban area as pos- sible cause for the local extinction of a land snail. Journal of Applied Ecology, 30: 333-340. BISHOP, M.J., 1981, Quantitative studies on some living British wetland mollusc faunas. Biological Journal of the Linnean Society, 15: 299-326. BORCARD, D., Р. LEGENDRE 8 P. DRAPEAU, 1992, Partialling out the spatial component of ecological variation. Ecology, 73: 1045-1055. BUCKLE, P., 1989, Monacha cantiana in the diet of the pheasant, Phasianus colchicus. The Con- chologists’ Newsletter, 111: 241-242. CAIN, A. J., 1983, Ecology and ecogenetics of ter- restrial molluscan populations. Pp. 597-647, In W. D. RUSSELL-HUNTER, ed., The Mollusca. Volume 6. Ecology. Academic Press, London. CAMERON, R. A. D., 1978. Terrestrial snail faunas of the Malham area. Field Studies, 4: 715-728. CAMERON, В. А. D., 1992, Land snail faunas of the Napier and Oscar ranges, Western Australia; di- 168 COWIE ET AL. versity, distribution and speciation. Biological Journal of the Linnean Society, 45: 271-286. CAMERON, В. А. D. & P. J. DILLON, 1984, Habitat stability, population histories and patterns of variation in Cepaea. Malacologia, 25: 271-290. CHESSEL, D. 4 S. DOLEDEC, 1993, ADE version 3.6: Hypercard Stacks and QuickBasic Microsoft Programme library for the analysis of environ- mental data. Users Manual. University of Lyon, Lyon. CHRISTENSEN, C. C. 8 P. V. KIRCH, 1986, Non- marine mollusks and ecological change at Bar- bers Point, Oahu, Hawaii. Bishop Museum Oc- casional Papers, 26: 52-80. CLAGUE, D. A. & ©. В. DALRYMPLE, 1987, The Hawaiian-Emperor volcanic chain. Part 1. Geo- logic evolution. United States Geological Survey Professional Papers, 1350(1): 5-54. COOKE, C. M., Jr. 8 Y. KONDO, 1960, Revision of Tornatellinidae and Achatinellidae (Gastropoda, Pulmonata), Bishop Museum Bulletin, 221: 1-303. COWIE, R. H., ed., 1992, The impact of alien spe- cies on island ecosystems: extended abstracts of a symposium, 30 May 1991, Honolulu, Hawaii, XVII Pacific Science Congress. Pacific Science, 46: 383-404. COWIE, В. Н., in press. Variation in species diver- sity and shell shape in Hawaiian land snails: in situ speciation and ecological relationships. Ev- olution. COWIE, R. H., N. L. EVENHUIS 8 C. C. CHRIS- TENSEN, in press, Catalog of the native land and freshwater molluscs of the Hawaiian Islands. Bishop Museum Bulletin. COWIE, R. H. 8 J. $. JONES, 1987, Ecological interactions between Cepaea nemoralis and Ce- paea hortensis: competition, invasion but no niche displacement. Functional Ecology, 1: 91- 97. COWIE, R. H. 8 G. M. NISHIDA, 1993, Malacolog- ical inventory survey in the multi-purpose range complex study area of the Pohakuloa training area, island of Hawaii. Bishop Museum Report to the Nature Conservancy of Hawaii. 56 pp. GAGNE, W. С. & L. W. CUDDIHY, 1990, Vegeta- tion. Pp. 45-114 In W. L. WAGNER, D. R. HERBST 8 S. H. SOHMER, eds., Manual of the flowering plants of Hawaii, Vol. 1. University of Hawaii Press & Bishop Museum Press, Honolulu. GIAMBELLUCA, T. W., M. A. NULLETT 4 T. A. SCHROEDER, 1986, Rainfall atlas of Hawaii. Re- port R76, Department of Land and Natural Re- sources, Honolulu, vi + 267 pp. HADFIELD, M. G., 1986, Extinction in Hawaiian achatinelline snails. Malacologia, 27: 67-81. HADFIELD, M. G. & S. E. MILLER, 1989, Demo- graphic studies on Hawaii's endangered tree snails: Partulina proxima. Pacific Science, 43: 1-16. HADFIELD, М. G., $. E. MILLER & А. H. CARWILE, 1993, The decimation of endemic Hawaiian tree snails by alien predators. American Zoologist, 33: 610-622. HADFIELD, М. С. & В. $. MOUNTAIN, 1981, A field study of a vanishing species, Achatinella mustel- ina (Gastropoda, Pulmonata), in the Waianae mountains of Oahu. Pacific Science, 34: 345- 358. KARPA, D. M. 8 P. M. VITOUSEK, 1994, Succes- sional development of a Hawaiian montane grassland. Biotropica, 26: 2-11. LEBRETON, J. D., D. CHESSEL, В. PRODON & М. YOCCOZ, 1988, L'analyse des relations espe- ces-milieu par l'analyse canonique des corres- pondences. Acta Oecologica/Oecologia Gener- alis, 9: 53-67. MORDAN, P. B., 1977, Factors affecting the distri- bution and abundance of Aegopinella and Nesovitrea (Pulmonata: Zonitidae) at Monks Wood National Nature Reserve, Huntingdon- shire. Biological Journal of the Linnean Society, 9: 59-72. PALMER, M. W., 1993, Putting things in even bet- ter order: the advantages of canonical corre- spondence analysis. Ecology, 74: 2215-2230. РЕАКЕ, J., 1978, Distribution and ecology of the Stylommatophora. Pp. 429-526, In V. FRETTER & J. PEAKE, eds., Pulmonates. Volume 2A. Systemat- ics, evolution and ecology. Academic Press, London. PETERSON, D. W. & В. I. TILLING, 1980, Transition of basaltic lava from pahoehoe to aa, Kilauea Volcano, Hawaii: field observations and key fac- tors. Journal of Volcanology and Geothermal Re- search, 7: 271-293. ВОНЕЕ, Е. J. 4 В. В. SOKAL, 1969, Statistical ta- bles. Freeman, San Francisco, xi + 253 pp. SEVERNS, R. M., 1981, Growth rate determination of Achatinella lila, a Hawaiian tree snail. Nautilus, 95: 140-144. SMALLRIDGE, М. A. 8 G. С. KIRBY, 1988, Com- petitive interactions between the land snails Theba pisana (Müller) and Cernuella virgata (Da Costa) from South Australia. Journal of Mollus- can Studies, 54: 251-258. SOKAL, В. В. 8 Е. J. ROHLF, 1981, Biometry, 2nd Edition. Freeman, San Francisco, xviii + 859 pp. SOLEM, [G.] A., 1976, Endodontoid land snails from Pacific Islands (Mollusca: Pulmonata: Sig- murethra). Part 1. Family Endodontidae. Field Museum of Natural History, Chicago, xii + 508 pp. SOLEM, [G,] A., 1984. A world model of land snail diversity and abundance. Pp. 6-22, In A. SOLEM & А. С. VAN BRUGGEN, eds., World-wide snails. Е. J. Brill/Dr. W. Backhuys, Leiden. SOLEM, [G.] A., 1990, How many Hawaiian land snail species are left? And what we can do for them. Bishop Museum Occasional Papers, 30: 27-40. SOLEM, [G.] A. & Е. М. CLIMO, 1985, Structure and habitat correlations of sympatric New Zealand land snail species. Malacologia, 26: 1-30. SOLEM, [G.] A., Е. М. CLIMO & D. J. ROSCOE, 1981, Sympatric species diversity of New Zea- LAND SNAIL DISTRIBUTION PATTERNS IN HAWAII 169 land land snails. New Zealand Journal of Zool- VITOUSEK, P. M., G. APLET, О. TURNER & J. J. ogy, 8: 453-485. LOCKWOOD, 1992, The Mauna Loa environ- TER ВВААК, С. J. F., 1986. Canonical correspon- mental matrix: foliar and soil nutrients. Oecolo- dence analysis: a new eigenvector technique for gía, 89: 372-382. multivariate direct gradient analysis. Ecology, 67: 1167-1179. Revised Ms. accepted 5 August 1994 € MALACOLOGIA, 1995, 36(1-2): 171-184 GENETIC HETEROZYGOSITY AND GROWTH RATE IN THE SOUTHERN APPALACHIAN LAND SNAIL MESODON NORMALIS (PILSBRY 1900): THE EFFECTS OF LABORATORY STRESS Alan E. Stiven Department of Biology, University of North Carolina, Chapel Hill, North Carolina 27599-3280, U.S.A. ABSTRACT Mesodon normalis hatchlings (totaling 569) exposed to a stress involving excess moisture, mucus, and feces exhibited a reduced mean growth rate, an increased mortality rate, and a significant positive association between juvenile growth and genetic heterozygosity. No signif- icant genotype-dependent growth and mortality occurred in an unstressed control cohort of 458 offspring. The significant r? value 0.053 in the stressed cohort is consistent with values from comparable studies in marine bivalves. Two of the five enzyme loci (ALA and РС!) contributed significantly to the association in the stressed cohort, but neither heterozygous deficiencies nor level of heterozygosity were associated with the growth-heterozygosity correlation. When the growth-heterozygosity association was examined in each of the eight broods comprising the stressed cohort, only one clutch from one parent showed a significant growth-heterozygosity association. In six of the remaining seven broods in the stressed cohort and in five of the eight broods in the unstressed cohort, the trend was in the direction of enhanced growth of het- erozygotes over homozygotes or failure of the homozygote class to survive. By comparing the genotypic structure of the parent to its offspring, it was determined that selfing did not occur and that multiple paternity was common during reproduction in this sample of the monoecious Mesodon normalis. These findings have significance for previous work on the population biology of this species in that genotype-dependent growth and survivorship appear to influence more the timing of adulthood and reproduction rather the usual body size-dependent reproductive output re- ported for many European helicids. Key words: growth rate, survival, genetic heterozygosity, stress. INTRODUCTION One consistent finding from research on the regulation of helicid land snail abundance is an inverse association between population density and such fitness components as ju- venile growth rate and survivorship, and adult body size (e.g. Dan 8 Bailey, 1982; Baur, 1988; Perry 4 Arthur, 1991). However, the re- lationship of reproductive output to growth, body size and adult age 1$ less well estab- lished (e.g. Carter 8 Ashdown, 1984; Baur, 1988; Baur & Baur, 1990; Foster 8 Stiven, in press, in review). Considerable research ex- ists on genetic variation in land gastropod populations (e.g. the European Cepaea spp.) (Cain, 1983), and thus any investigation of mechanisms controlling the growth, body size, and age connections cannot ignore the possible importance of a genotype and growth rate association. This study focuses on the importance of laboratory induced 171 stress in the southern Appalachian land snail Mesodon normalis on this association. Positive associations between shell size or growth rate and genetic heterozygosity are common among mollusks (Zouros & Foltz, 1987). This association often appears when loci show heterozygote deficiencies, when the sample comes from a natural rather than a laboratory population (Zouros, 1987), when the sample consists of a random population sample ratherthan offspring of a single mating (Mallet et al., 1986), when the sample consists of a young rather than an older cohort (Diehl 8 Koehn, 1985), or when the number of loci examined is large (Koehn et al., 1988). Expla- nations range from overdominance (Zouros et al., 1980, 1983), to the presence of null alleles (Foltz, 1986), and to different responses of genotypes to environmental stress (Koehn 8 Shumway, 1982; Hawkins et al. 1986; Holley & Foltz, 1987). Attempts have also been made to assess the contribution of individual loci to 172 STIVEN the growth rate-heterozygosity relationship by comparing fitness components between homozygotes and heterozygotes for individ- ual loci, with glycolytic and protein catabolism enzyme loci emerging as being most signifi- cant (Koehn et al., 1988; Gentili 4 Beaumont, 1988; Borsa et al., 1992). Even though stress was predicted to en- hance the heterozygous advantage (Koehn 8 Shumway, 1982; Mitton 8 Grant, 1984), allo- zyme heterozygosity and fitness have been inadequately studied under stressful and op- timum conditions (Hoffmann 8 Parsons, 1991). Most evidence supporting the impor- tance of stress has come largely from bivalves (e.g. Diehl 8 Koehn, 1985; Gentili 8 Beaumont, 1988 [but see Gaffney's 1990 reanalysis], Koehn & Bayne, 1989; Scott 8 Koehn, 1990; Borsa et al., 1992). In gastropods, evidence for the significance of stress in enhancing the relationship between genetic heterozygosity and fitness is scarce (Komai 4 Emura, 1955; Booth et al., 1990). There are a few studies of aquatic and marine snails in which no asso- ciation was found in either the stress or con- trol treatment (Fevolden & Garner, 1987; Foltz et al., 1993), or the association persisted over a range of environments in which some could be labeled stressful (e.g. low to high salinity levels; Garton, 1984). Heritability of shell size in land snails can range from 50% to 70% (Goodfriend, 1986). For growth rate, a major determinant of adult size in land snails, heritability can range from 40% to 60% in Cepaea nemoralis (Ooster- hoff, 1977). Emberton (in press) also reported a heritability component of 30% and an en- vironmental component of 50% for labora- tory growth rate in the southern Appalachian land snail Mesodon normalis. Mesodon normalis (Pilsbry, 1900), which was formerly considered a subspecies of Mesodon andrewsae (see Pilsbry, 1940) (Sty- lommatophora: Polygryidae), is one of the larger endemic land snails of the deciduous forests of the southern Appalachian Moun- tains of the U.S.A. (Hubricht, 1985). Adult population densities are generally low (< 2/10 m?) (Foster & Stiven, in review), especially when compared to the European helicid Ce- paea nemoralis, for which densities range from 5-40/10 m? (Perry & Arthur, 1991). Me- sodon normalis is active on the forest floor, predominately at dawn and dusk (Asami, 1993) from about late April to mid-October. Monoecious adults mate during the spring and early summer, then lay several clutches of up to 110 eggs in the leaf litter. Young emerge from June through August. This spe- cies is a determinant grower with a recurved shell lip forming as shell growth ceases and adulthood occurs. Juvenile snails require ap- proximately two years to reach an adult size (27 to 36 mm diameter), and they often con- tinue to live three or more years as adults (Foster & Stiven, in review). Prior work on Mesodon normalis has shown that juveniles reared at lower densities grew faster and became larger adults than did snails reared at higher densities. Survivorship was density-dependent, and the slower grow- ing juveniles had a higher probability of dying younger than their faster growing counter- parts (Foster & Stiven, in review). This study examines the association between juvenile growth rate and genetic heterozygosity in clutches of stressed (a period during juvenile growth in the laboratory of increased mois- ture, feces, and mucus) and unstressed co- horts of Mesodon normalis. In particular, this investigation tests the hypothesis that stress (defined as an “environmental change that causes some response by the population” (Underwood, 1989)) causes a genotype- dependent response(s) in an environmentally stressed population. Parent and offspring genetic data also per- mitted an assessment of mating patterns (self or cross fertilization, single or multiple pater- nity) (Anderson & McCracken, 1986; Gaffney & McGee, 1992). For example, Foltz et al. (1984) predicted that successful gastropod colonizers of North American by European species would be self-fertilizing, and that most endemic North American forms found in relatively undisturbed habitats would be out- crossing. In addition, data on shell and tissue color patterns in the offspring are presented as possible markers in future population re- search. METHODS Collections and Laboratory Techniques To obtain egg clutches for the “stressed” and “unstressed” cohorts, adult Mesodon normalis were collected from sites near High- lands, North Carolina, and the U.S. Forest Service Coweeta Hydrologic Laboratory and Experimental Forest (about 25 km east of Highlands) in late May of 1988 and 1989. MESODON GROWTH AND HETEROZYGOSITY 173 Specifically, Highlands snails came from a site (Chin Site) along Highway 106 about 6 km southwest of Highlands (1,170 m eleva- tion), and Coweeta snails came from Water- shed 28 (1,100 m elevation) (Stiven 1989; Foster 8 Stiven, in review, respectively, have more detailed descriptions of these sites). Adult snails were placed individually in par- tially covered plastic cylindrical containers (15 cm, diameter, 6 cm deep) which con- tained a 1 cm layer of soil covered by about 2 cm of leaf litter from the respective sites. Snails were fed daily rations of lettuce dusted with calcium carbonate, and the soil and leaf litter were changed weekly (Cowie 8 Cain, 1983). Egg clutches were removed immedi- ately after laying and placed into plastic 9 cm petri dishes lined with moist paper toweling. Only first clutches were used in the experi- ment. After hatching, each of the experimen- tal clutches was placed into 15 cm diameter partially covered plastic chambers lined with moist paper toweling and covered with a 0.5 cm layer of soil and a thin layer of fragmented leaf litter. Snails were fed as above. Cham- bers were cleaned and soil and leaf litter re- newed weekly, except in the stress treatment described below. After one month of snail growth, the chamber was replaced by one of 20.3 cm diameter, and after two months by one of 25.3 cm diameter. All snails that died were removed. The experiment was termi- nated after 100-115 days of growth. Labora- tory temperature ranged from 21°С to 24°С, and the light-dark regime approximated nat- ural conditions. Experimental Design and Response Variables The “stressed” and “unstressed” juvenile cohorts were established as follows. Of the 24 adults collected in 1988, eight produced first clutches, six from Coweeta adults and two from Highlands adults. These clutches were the “stressed” cohort and were ex- posed to a 2-week “stress” period in mid- July, in which chambers were not cleaned and soil and leaf litter were not replaced. Feeding was continued during this period and excess feces, mucus, and moisture was apparent. Individual growth rates were slowed and significant mortality occurred in all chambers of the “stressed” treatment. From the 15 adults collected in 1989, seven first clutches came from Coweeta and one from Highlands snails. These clutches were not stressed during the juvenile period and were the control or ““unstressed” cohort. The remainder of the adults failed to produce clutches. The shell diameters of newly hatched snails of each clutch were measured with a calibrated reticle on a dissecting microscope, and a mean diameter for each clutch was calculated. At the end of the experiment, the shells of surviving individuals were measured with dial calipers to the nearest 0.1 mm, their soft tissue processed for electrophoresis, and shell fragments saved. Growth was ex- ponential during the first 100 days of growth, and the instantaneous growth rate ([In final size — In initial size]/days of growth) was cal- culated for each surviving individual utilizing the mean shell size for each clutch at hatch- ing and its final size and standardizing the growth rate to 100 days of growth. In the stressed cohort, tissue color was noted at time of sacrifice. For the unstressed survi- vors, tissue color, tissue mottling and shell color were recorded. Tissue colors were light tan, dark brown, and intermediate. Shell col- ors were either brown or grayish brown, and head tissue mottling or spotting was either present or absent. Genetic Analysis Genetic data for the experimental young and adults came from starch gel electro- phoresis, as described in Emberton (1988) and Stiven (1989). The entire tissue of young was processed, but the genotype of the adults was assessed by cutting off a small piece of the foot muscle and washing in dis- tilled water before processing. Five polymor- phic loci (PGM-2, PGI, MPI, ALA-2, and ALG) were used for the 1988 surviving experimen- tal young. An additional PGM locus (PGM-1) was resolved in the 1989 young. There were two visible loci for the peptidase-leucyl ala- nine gels but only one (ALA-2) was clear enough to use. All loci except MPI were run on а LiOH buffer. МР! was run on a TEB9/8 buffer. The genotypic and allelic data for the adult group and for each experimental cohort were summarized using Swofford 8 Seland- er's (1989) BIOSYS-I program. Measures of genetic heterozygosity included the number of heterozygote loci per individual, mean het- erozygosity as direct count and Nei’s (1978) unbiased estimate. Departures of genotypic frequencies from Hardy-Weinberg expecta- tions were tested by chi-square with one de- 174 STIVEN TABLE 1. Comparisons of clutch size, hatching success, survival, and instantaneous growth rate over the first 100 days of growth for first clutches produced by adult snails under the “unstressed” and “stressed” treatments. Values are means + SE per clutch (data for eight clutches/treatment). Unstressed Stressed t P Clutch Size (no.) 62.88 + 4.13 DERE 7.60 1.83 0.09 Hatching Success 0.91 + 0.02 0.91 + 0.02 0.15 0.88 Survival 0.81 + 0.02 0.34 + 0.05 8.79 <0.0001 Growth Rate (100 days) 0.68 + 0.03 0.41 + 0.03 5.74 0.0001 gree of freedom using a pooling procedure RESULTS giving three genotype classes, homozygotes for the most common allele, heterozygotes for the most common allele and one of the other alleles, and all other genotypes. Het- erozygote deficiencies were expressed as DRE tle)/ his, where fic andes were observed and Hardy-Weinberg expected heterozygosities respectively. To examine genetic differentiation of the adults between the two sites, F-statistics (Nei, 1977) were also computed for the adult genetic data and significance of Fay tested by chi-square. Statistical Procedures The association between multilocus ge- netic heterozygosity and growth rate was as- sessed by regressing the instantaneous growth rate (for the 100 days) with the num- ber of heterozygous loci for an individual in each of the “stressed” and “unstressed” co- horts. The association was then examined within individual clutches (family effects) and between the two sites of origin of parents. The relative contribution and significance of each locus to the fitness-heterozygosity as- sociation was determined by a multiple linear regression model of growth rate as a function of heterozygosity of each locus as an inde- pendent effect (scoring of the genetic state of each locus for each animal as homozygous or heterozygous) following Koehn et al. (1988). Significance of these effects for each locus was assessed by F-tests. The effect of site of origin of the parent, parent-offspring genetic similarity, and tissue or shell color- specific differences utilized either the chi- square test (or G-test), analysis of variance (ANOVA), or analysis of covariance (AN- COVA) when appropriate. Bartlett’s test was used to assess homogeneity of variances. All statistical analyses were done using SYSTAT (Wilkerson, 1990). The significance level was 201053 Treatment Conditions Young hatched after 13-20 days, and eggs of any one clutch hatched within 24 hours of each other. For the 1988 snails, the first clutches used in the experiment appeared between June 14 and June 17, and hatching occurred between June 27 and July 5. For 1989 snails, the comparable dates were May 24 through May 29 and June 9 through June 15. Second and third clutches were pro- duced by some snails into early August. The conditions associated with the treatments (stressed and unstressed) produced the ex- pected differences in levels of survival and growth rate (Table 1). Of the 569 initial num- ber of hatchlings in the eight clutches in the 1988 “stressed cohort,” only 192 or 34% survived to the end of the experiment. The comparable figures for the 1989 “unstressed cohort” were 458 hatching from eight clutches, with 373 or 81% surviving. In addi- tion, the growth rate of the survivors in the unstressed treatment was about 65% greater than in the stressed treatment. Clutch size (number per clutch) and hatching success did not differ between the treatments. Adults from the Highland’s site did have a signifi- cantly greater mean clutch size than those from the Coweeta site (95.0 vs. 65.2, ty 14 = 3.15, P = 0.007), but hatching success, growth rate, and survival did not differ be- tween sites. Genetic Structure of Adults The 1988-89 field collection of adults yielded 26 from Coweeta and 13 from High- lands. The level of heterozygosity (direct count) in Coweeta snails was over twice that of Highlands snails based upon five loci (0.292 and 0.129 respectively). All loci, with one exception, had genotypic frequencies MESODON GROWTH AND HETEROZYGOSITY 0.6 STRESSED u © > a o 0:5 = e | o - 2 + | © B _] 0.4 ; « 5 = + = О 0.3 + Regression Е = 10.62 *** 4 r2 = 0.053 *** n= 192 IO a a ee ee О 1 2 3 4 Number Heterozygous Loci 103 0.8 UNSTRESSED 0:7. | $ 4 0.6 - ] 0.5 + Regression F = 0.17 ns - r2 = 0.0004 ns n = 373 0.4 | — =} | __ je l O 1 2 3 4 Number Heterozygous Loci FIG. 1. Correlation and regression of individual juvenile growth rate and level of heterozygosity for the stressed and unstressed experimental cohorts. Data are depicted as means (+ 1 SE) to illustrate the patterns, but the analysis is on individuals. corresponding to Hardy-Weinberg expecta- tions. The exception, ALA in Coweeta snails, exhibited a deficiency in heterozygotes (D = — 0.363; Х? = 6.03, Р = 0.014). Two loci, МР! and ALA, were also monomorphic in High- lands but not Coweeta snails, contributing to the lower P-value in the Highlands’ popula- tion. The overall Fa, (Nei 1977) for all loci (0.121) was significant (XP y; 43 = 35.4, P < 0.001). Mean F,, and Е- were positive and high (0.101 and 0.210 respectively). Association Between Heterozygosity and Growth Rate Effect of Stress at Level of Population (Co- hort): The association between an individu- al's growth rate and its number of heterozy- gous loci was highly significant for the stressed cohort but not for the unstressed cohort (Fig. 1). Because of the large number of points in each treatment level, only the growth rate means (+ 1 SE) for each het- erozygote frequency value are shown in Fig- ure 1. However, the values of г? and F are _ from the analysis of individuals, not means. Adding the data of the 6th locus (PGM-1) did not change the nonsignificant growth rate- heterozygosity association in the unstressed cohort. In addition, the growth rate of het- erozygous individuals was 17% higher than fully homozygous individuals (F, ¿99 = 6.622, P = 0.011) in the stressed cohort. In the un- stressed cohort, growth rates did not differ between the two genetic groups (homozy- gous and heterozygous rates were 0.666 + 0.015 and 0.667 + 0.011 respectively, P = 0.94). Effect of Stress at Brood Level (Clutch-Sibs): Of the eight clutches in the stressed treat- ment, only one from a Highlands parent (H3) exhibited a significant association between individual growth rate and number of het- erozygous loci (Table 2). Three broods in the stressed treatment, however, had no surviv- ing homozygotes. The trend was towards greater growth rates of heterozygotes over homozygotes. In the unstressed cohort, only one brood lacked surviving homozygotes. One brood had a borderline significantly faster growth rate-heterozygosity association (Table 2), and again, there was a trend to- wards enhanced growth of heterozygotes over homozygotes in three of the remaining broods. ANCOVA of individual growth rate among clutches (heterozygosity differences among clutches as covariate) indicated significant variation for both stressed and unstressed broods (F7 4383 = 26.62, P = < 0.0001; F7 364 = 14.61, P < 0.0001 respectively). This sug- gests a strong parent or genotype effect on growth rate when differences in brood het- erozygosity are controlled. Parent effects also significantly influenced the level of sur- 176 STIVEN TABLE 2. Results of analysis of variance of growth rate by number of heterozygous loci for each Mesodon normalis brood in stressed and unstressed cohorts. The Differ. column represents the percentage difference in growth rate of heterozygous over fully homozygous individuals (+ value). Stressed Differ Brood (%) F Р ie C1 no surviving homozygotes C2 +8.3 0.04 0.85 0.001 C3 no surviving homozygotes C4 +20.6 0.36 0.57 0.056 C6 — 1.0 2.84 0.10 0.098 C7 no surviving homozygotes H3 +20.0 5.93 0.02* 0.165 H4 +12.0 0.587 .451 0.022 vival among broods in the stressed treat- ments (X° 47 = 73.3, Р < 0.0001) but not in the unstressed broods (P = 0.61). The Influence of Site of Origin Because genetic heterozygosity of adults differs between the two sites of origin, and level of survival in the stressed treatment was influenced by the variable parent, it is of in- terest to know if growth rate and heterozy- gosity ofthe young were also effected by site of origin. ANOVA confirms that both variables in both treatments are significantly higher in young from Coweeta parents than from High- lands parents (Fig. 2) (Growth: F, 499 = 96.96. P=<0:0001 57-827, R=0:004; Hetloci: F; 499 = 66.11, Р < 0.0001; F, 37, = 25.34, P < 0.0001). The Contribution of Individual Loci The relative contribution of each locus to the multiple-locus positive association of growth rate and heterozygosity in the stressed treatment was assessed by multiple regression (Table 3). The Type Ш sum of squares for a given locus is a measure of the association of heterozygosity with growth rate at that locus. The rankings of loci by their SS indicates the importance of two loci, ALA and PGI, with both being significant (Table 3). The highest ranking loci are not necessarily those with the highest level of heterozygosity. An analysis of differences in mean growth rates between homozygotes and heterozy- gotes at each locus (Fig. 3) confirmed the significant contribution of ALA and PGI to the positive fitness-heterozygosity association. A comparable multiple regression analysis Unstressed Differ. Brood (%) F Р r C1 — 2.6 0.43 0.52 0.009 C2 +27.5 3.78 0.06 0.078 C3 — 4.3 1.01 0.32 0.032 C4 +11.8 1.36 0.25 0.027 C5 no surviving homozygotes C6 +39.9 4.07 0.05 0.085 C7 +12.8 4.42 0.04* 0.121 H1 9 0.48 0.49 0.00 was also performed on the unstressed data set. Two loci, ALG and ALA, showed signifi- cant F-values. Further analysis indicated that in the ALA locus homozygotes exhibited a higher growth rate than heterozygotes, but for the ALG locus the converse was true. Heterozygote Deficiencies and Offspring-Parent Genetic Relationships Parents and their offspring from the same site displayed comparable levels of heterozy- gosity (Fig. 4). In the stressed cohort, only PGI exhibited a significant deficiency in het- erozygotes (Table 4). However, in the un- stressed cohort there were two loci (PGI and ALG) that had deficiencies, both in Coweeta snails and one (PGI) in Highlands snails. On pooling over sites, PGM-2 and ALA also be- came deficient. In adults from both sites, only one of the five loci showed a heterozygote deficiency (D-value for ALA was —0.417**, by chi-square) (Table 4). It appears, there- fore, that in Mesodon normalis the finding of a significant association between growth rate and genetic heterozygosity is not associated with increased levels of heterozygote defi- ciency among loci. However, one of the two loci that showed a significant contribution to the correlation was also deficient in heterozy- gotes. When the genetic structure of surviving off- spring and their corresponding monoecious parent was compared, there was no evi- dence of selfing in any of the 16 adults; that is, non-parent alleles were present in the off- spring of at least one locus in every adult. Evidence of multiple paternity was found in eight of the 16 adults by comparing the locus-specific genotype of the parent with all MESODON GROWTH AND HETEROZYGOSITY 177 Stressed 2.0 O о 1.5 — Y) 3 = 1.0 > N O = 2 0.5 ® E O > 0.0 Pr = o Unstressed A 1.4 © : — 1.2 Yy HETL 12 = O GRO Ф ed © 1.0 cc = > 0.8 0.8 O = (5 0.6 0.4 0.4 FIG. 2. Mean growth rates (left X-axis) and number of heterozygous loci (right X-axis) for Coweeta and Highlands clutches in the stressed and unstressed treatments. Error bars are 1 SE. 178 STIVEN genotypes of all corresponding offspring. Multiple paternity may actually be higher be- cause the genotypes of young that died dur- ing the experiment are not known. Variation in Tissue and Shell Colors The frequencies of the three tissue color classes did not differ between the Coweeta and Highlands young in either the stressed or unstressed cohort (Table 5). However, in the stressed cohort higher growth rates and higher levels of genetic heterozygosity were associated with the light tan morph, and sig- nificantly lower values with the intermedi- ate and dark morphs. If these tissue color morphs have a simple genetic basis (Table 5), then the color morph frequencies conform to Hardy-Weinberg expectations in the stressed cohort but not in the unstressed cohort. Shell color and tissue mottling were re- corded only for individuals in the unstressed cohort. Brown shell morphs had their highest frequency (73%) in snails derived from Coweeta, and grayish brown morphs were the exclusive shell color in the Highlands snails. Genetic heterozygosity (number of heterozygous loci per individual) was signifi- cantly higher in brown shelled and mottled tissue morphs (F, 37, = 43.57, Р < 0.0001; Fıazrı = 5.69, Р = 0.018, respectively). Growth rate did not differ between shell color morphs, but was higher in mottled than in non mottled morphs (F, 3,ı = 46.60, Р < 0.0001). DISCUSSION The Growth Rate-Heterozygosity Association The finding of a positive association be- tween growth rate and allozyme heterozy- gosity in M. normalis at the population level parallels that of many similar studies (Allen- dorf 4 Leary, 1986; Zouros & Foltz, 1987). However, the positive association in M. nor- malis was found only in the stressed cohort, in which higher mortality and reduced mean growth rate occurred. In the control popula- tion in which survivorship and growth rate were significantly higher, growth rate was in- dependent of genetic heterozygosity. In many of the studies depicting a positive as- sociation, information on possible stress conditions is often not available. As in many TABLE 3. Results of multiple regression analysis on growth rate-heterozygosity association for each locus utilizing all individuals in the stressed cohort. H is the mean direct-count heterozygosity value for all individuals for each. Loci are ranked in their importance by value of the Type Ш sum of squares (SS). В? = 0.103 and overall regression Fy 186 = 4.28, P = 0.001. Locus H SS F Р АРА 0.54 0.148 10.54 О РС! 0.34 0.068 4.86 0.029* РСМ-2 0.06 0.010 0.71 0.400 МР! 0.02 0.002 0.14 0.710 АРС 0.55 0.001 0.07 0.793 `Р<10:05, EP 01001 of these studies, the positive association in Mesodon occurred during the vulnerable early life stage of the cohort, a time when most energy 15 being allocated to somatic growth in mollusks (Zouros et al., 1980). However, for М. normalis, it is not known if the association eventually disappears as the snail cohort ages, as has been shown in some marine bivalves (Diehl & Koehn 1985). In Mesodon, the amount of variance in growth rate explained by variation in het- erozygosity 1$ small (5.3%), but corresponds to г? values from similar studies, even those in which the number of loci sampled was three times that of this study (Koehn et al., 1988). The Mesodon results are also consistent with those of the few studies in which envi- ronmental stress was described or was part of a planned experimental treatment. In some studies, increased cohort mortality was as- sociated with stress (Samollow 8 Soule, 1983). Such was the case in this study. Also individuals that have a greater probability of dying younger are those that are slower growers and hence smaller (Foster & Stiven, in review), thus leaving a greater proportion of larger, more heterozygous and faster growing individuals. These results for M. nor- malis appear to be consistent with the over- dominance model, where multiple locus het- erozygotes exhibit superior fitness to their associated homozygotes (Zouros & Foltz, 1987; Zouros et al., 1988). For this laboratory scenario to be relevant to natural Mesodon populations, a group of hatchlings or juveniles would have to be ex- posed to a period(s) of environmental stress or crowding that causes mortality (e.g. a cold, wet period with reduced dispersal ac- MESODON GROWTH AND HETEROZYGOSITY 179 0.5 []Нот [/] Het OD > 0.45 N O O = ns NE ns a | à i poc = N б 0.35 / = ТИ И PGM-2 y u ALA-2 U FIG. 3. Mean growth rates for homozygous and heterozygous individuals for each locus in the stressed cohort. Error bars are 1 SE, ns means not significant at = 0.05, *** means significant at P < 0.001 (t-tests). 0.5 04 stressed 0.3 0.2 Heterozygosity (H) COW HIGH COW HIGH FIG. 4. Mean heterozygosity (direct count) for Coweeta and Highlands adults (AD) and corresponding offspring (YN) for the stressed and unstressed treatments. Error bars are 1 SE. 180 STIVEN TABLE 4. Heterozygote deficiencies (—0) for loci in the stressed and unstressed cohorts of surviving young. Significance by chi-square. COW and HIG are the Coweeta and Highlands sites. Cohort PGM-2 PGI MPI ALA ALG Stressed: COW 0.028 0 AUS 0.260 0.169 = 0.10 HIG 0.017 —0.740** 0.037 —0.098* 0.293 POOLED 0.022 = 0872353 0.193 0.033 = 05105 Unstressed: COW 0.009 = 0.835255 — 0.022 — 0.074 = 0.203 HIG 0.153 =] {OOO 0 0 0 POOLED = 0.3935 —0.417^** —0.037 —0.098* — 0.3987 < 0:05, OO

189 = 921, PE 0006 1E2 2=8 Heterozygosity among colors ns F> 189 = 4.21, P = 0.016 ES. 2 =8 Hardy-Weinberg conformity** X? = 67.5, P < 0.001 ns Shell Color* Between sites X? = 57.62, P < 0.0001 not recorded Growth between colors ns not recorded Heterozygosity between colors Tissue Mottling* Between sites Growth between classes Heterozygosity between classes *Tissue color: 1 = light tan, 2 = intermediate, 3 = dark *Shell color: 1 = brown, 2 = greyish brown *Tissue mottling: 1 = none, 2 = mottled “Assuming A,A, = color 1, A,A> = color 2, АА» = color 3 tivity, mucus accumulation and high litter moisture, or a late frost). Unfortunately, little is known of mortality and its causes in Mes- odon normalis in the field. Mortality in the lab- oratory is density dependent, and in the field mean adult size is larger in low density envi- ronments (Foster & Stiven, in review), as is the case in Capaea nemoralis (Oosterhoff, 1977). When the growth rate-heterozygosity as- sociation was examined separately in each of the eight clutches in the stressed treatment, only one was positive. However, in three of the remaining clutches, only heterozygotes were found at the end of the experiment, suggesting that the homozygotes died, al- though the initial frequencies of genotypes could not be assessed. The non-significant Fr 371 = 42.61, Р < 0.0001 | = 72 X? = 8.69, Р = 0.003 F; 371 = 46.6, Р < 0.0001 21 F; 371 = 6.69, Р = 0.018 1 72 not recorded not recorded not recorded not recorded trend in the remaining clutches was for higher growth rates in heterozygotes. A smaller sample size (brood vs. cohort) may also be partly responsible for the lack of significance (Beaumont et al., 1983; Zouros & Foltz, 1987), or there may be a true absence of the association at the brood level. In a number of other studies of sibling cohorts in marine bi- valves (Beaumont et al., 1983; Gaffney & Scott, 1984; Beaumont et al., 1985; Mallet et al., 1986), positive relationships were also absent. Gaffney & Scott (1984) point out that many of the positive associations between growth and heterozygosity in marine bivalves come from large populations, from which in- dividuals were sampled at random, and that individuals coming from a single mating may not show the positive association. The prob- MESODON GROWTH AND HETEROZYGOSITY 181 lem of detecting the positive relationship at the brood level might also be affected by the number of different matings of the parent with different adults, as well as differences in survivorship among clutches (Zouros 4 Foltz, 1987). In Mesodon, the removal of possible effects of differential clutch survivals levels by ANCOVA did not abrogate the positive as- sociation at the population level in the stressed cohort, nor did it change the out- come when the level of analysis was by brood rather than across individuals. In the stressed Mesodon cohort, two of the five loci (ALA, PGI) were significant contribu- tors, with heterozygotes having faster growth rates than homozygotes for both loci. PGI functions in the glycolytic pathway, and ALA is involved in protein catabolism. In the Koehn et al. (1988) study of the coot clam, Mulinia loci, ALA was significant but PGI was not; a total of eight out of 15 loci had signif- icant effects, including PGM, MPI and three nonspecific AP loci. Gentili & Beaumont (1988) reported significant contributions of only two out of eight loci in a high density treatment cohort of Mytilus edulis, and Borsa et al. (1992) found that only one locus (PGM) out of seven had a higher heterozygosity in the survivors of a marine bivalve exposed to anoxic stress compared to the control. In the unstressed Mesodon cohort, two loci had contrasting relationships; growth rate was higher in heterozygotes in one locus but lower in heterozygotes in the other. Most studies of the assessment of the comparative contribution of loci have utilized up to five or six loci. Koehn et al. (1988) warned that this number may be inadequate. They argued that a large enough sample of diverse poly- morphic genes should be assayed to encom- pass various metabolic roles, that the linkage relationship among loci be known, and that the correlation between a fitness parameter and heterozygosity be established. Whereas the last assumption is met in the Mesodon study, and the polymorphic loci used cover a range of functions, the number (5) is small, and linkages among the loci are not known. Thus, it is premature to draw definitive con- _ clusions about which loci are more significant contributors to the heterozygosity-growth rate relationship in the Mesodon system. The significant Fs, value of 0.121 for the adult sample (those producing clutches plus those that did not) from Coweeta and High- lands sites suggests sizeable differentiation (Wright, 1978; Hartl & Clark, 1989). The high positive values of F,, and F,- also reflect the high levels of homozygosity in the popula- tions, expected in a species with limited dis- persal and probably inbreeding. These Fay values, although derived from relatively low sample numbers are similar to those reported for two other land snails, Mesomphix an- drewsae and Mesomphix subplanus, from separate watershed in the Coweeta forest (Stiven, 1989). Zouros & Foltz 1987 suggest that the pres- ence of many loci with heterozygous defi- ciencies is associated or even enhances the chance of finding a positive correlation of growth rate and heterozygosity. In the stressed Mesodon cohort, which exhibited the positive growth rate-heterozygosity asso- ciation, only one locus (PGI) showed geno- typic frequencies that did not conform to Hardy-Weinberg expectations, and this locus was a significant contributor to the associa- tion (Table 3). In contrast, two loci showed heterozygote deficiencies in the unstressed control treatment. Therefore, in this study the presence of loci that are deficient in het- erozygotes was not a necessary condition for a significant fitness-heterozygosity associa- tion. How balanced are the stressed and un- stressed experimental treatments from the two different years with regard to site of ori- gin and genetics of contributing parents? The ratio of Coweeta to Highlands parents was very similar, 8:1 and 8:2 for the stressed and unstressed treatments, respectively. As noted earlier, the heterozygosity level in Coweeta adults was over twice that in High- lands adults. However, mean heterozygosi- ties for the stressed and control treatments were similar, 0.233 and 0.250. It appears, therefore, that initial parental genetic diver- sity and site contribution are essentially equivalent in the two treatments. However, more population genetic and ecological work should focus on understanding the causes of the different levels of heterozygosities as well as the different mean clutch sizes found be- tween Coweeta and Highlands. Significance to Population Processes in Mesodon normalis Part of the population regulation process in the terrestrial gastropod M. normalis comes from the effects of population density on ju- venile growth and their subsequent rate of 182 STIVEN development to adults. In the field, adult den- sities and adult sizes are negatively corre- lated, with Coweeta sites having higher den- sities and smaller adult sizes than Highlands sites (Foster & Stiven, in review). As in many invertebrates, growth rates are quite variable among juveniles, and this variability has sig- nificance for both final size and time and age at maturity. From our laboratory experiments on the effects of density and food (Foster & Stiven, in review), a small fraction (5.1%, or 11 animals) ceased growth and showed adult characteristics after one year of growth. These were the faster growers. Also those growing longest tended to be those that were smallest at the beginning of the experiment. In the field, juveniles that were larger at the start of the second winter also became the larger adult the following summer, the usual time of first breeding (Foster & Stiven, in review), and these would be the faster growing juveniles. They may produce their first offspring that summer, and may live up to three years more as adults (Foster & Stiven, in review). The smaller and slower growing juveniles are also more prone to die younger under increased environmental stress (i.e. density; Foster & Stiven, in review) and are also the more ho- mozygous individuals (this study). “Older” adults, regardless of size, also produce twice as many clutches but fewer eggs per clutch as do the “younger” adults, but the total number of eggs and hatching success do not differ with age (or adult body size) (Foster & Stiven, in press). Adult age cannot be precisely de- termined, but the older adults have exten- sively eroded periostraca, whereas younger adults have intact periostraca. If, under stress conditions, the faster growing more heterozy- gous juveniles have better survivorship, ma- ture and reproduce sooner (possibly even at the end of the second summer, but at least early the next summer), their lifetime fitness would obviously be greater, and they may be- come larger adults. It is not known if the smaller adults derived from the slower grow- ing, more homozygous juveniles would also be more prone to higher mortality after reach- ing adulthood. In the European helicid snail Cepaea nemoralis, the larger faster growing juveniles tend to reach sexual maturity earlier and to be the larger adults (Oosterhoff, 1977). In addition, parental body size, egg size and fecundity, and juvenile growth rate are posi- tively correlated (Oosterhoff, 1977; Carter & Ashdown, 1984; Baur, 1988). In contrast, in M. normalis, juvenile growth rate and repro- ductive output are independent of the size of the parent, even though clutch number and clutch size are related to adult age (Foster € Stiven, in press). Therefore, the significance of the variable and genotype-dependent growth and mortal- ity rates in M. normalis may lie not so much with differential fecundity, but with a coupling of increased survival and earlier reproduction for the more heterozygous faster growers, especially when adverse environmental con- ditions occur during the period of juvenile growth. Breeding System No evidence of selfing was apparent from a comparison of parent and offspring geno- types in M. normalis from either site, confirm- ing the speculation of Foltz et al. (1984) that native gastropods in relatively undisturbed environments would be outcrossers, with self-fertilization the more likely mode of the colonizer (European). There is also strong evidence (62.5% of the parents) for multiple paternity in M. normalis. While perhaps widespread in pulmonate mol- lusks, most reports, such as this, come from studies with another focus (Murray, 1964). Tissue and Shell Color Morphs Whereas the frequencies of the three tis- sue color morphs did not differ between the stressed and unstressed treatments, higher growth and heterozygosity levels were char- acteristic of the light tan morph, with signifi- cantly lower values for the dark and interme- diate forms. The significance of this is not known, but the morph frequencies in both Coweeta and Highlands offspring from the stressed cohort conform to Hardy-Weinberg expectations, suggesting a genetic mecha- nism. Tissue color, tissue mottling, and shell color variation in these populations require further work, especially as markers in popu- lation work. ACKNOWLEDGMENTS Suppont for this study was provided by fac- ulty research grants from the University of North Carolina. Special thanks for various forms of help are also due the Highlands Biological Station's Director, Dr. Richard Bruce, personnel at the Coweeta Hydrologic MESODON GROWTH AND HETEROZYGOSITY 183 Laboratory, and my former graduate student, Bradley Foster. Comments on the manu- script by Kenneth Emberton and one anony- mous reviewer are gratefully appreciated. LITERATURE CITED ALLENDORF, F. W. & В. F. LEARY, 1986, Het- erozygosity and fitness in natural populations of animals. Pp. 57-76, In: M E. SOULE, ed. Conser- vation biology. Sinauer Associates, Stampford, Connecticut. ASAMI, T., 1993, Divergence of activity patterns in coexisting species of land snails. Malacologia, 35: 399-406. ANDERSON, J. В. 8 G. Е. MCCRACKEN, 1986, Breeding system and population genetic struc- ture in philomycid slugs (Mollusca: Pulmonata). Biological Journal of the Linnean Society, 29: 317-329. BAUR, B., 1988, Population regulation in the land snail Arianta arbustorum: density effects on adult size, clutch size and incidence of egg cannibal- ism. Oecologica, 77: 390-394. BAUR, В. & А. BAUR, 1990, Experimental evidence for intra- and interspecific competition in two species of rock-dwelling lands snails. Journal of Animal Ecology, 59: 301-315. BEAUMONT, А. R., С. М. BEVERIDGE & М. D. RUDD, 1983, Selection and heterozygosity within single families of the mussel Mytilus edulis (L.). Marine Biology Letters, 4: 151-161. BEAUMONT, A. R., E. M. GOSLING, C. M. BEV- ERIDGE, М. О. RUDD & С. М. BURNELL, 1985, Studies оп heterozygosity and size in the scal- lop, Pecten maximus (L.). Pp. 443-455, In: P. E. GIBBS, ed., Proceedings 19th European Marine Biology Symposium. Cambridge University Press, England. BOOTH, С. L., D. $. WOODRUFF & $. J. GOULD, 1990, Lack of significant associations between allozyme heterozygosity and phenotypic traits in the land snail Cerion. Evolution, 44: 210-213. BORSA, P., Y. JOUSSELIN & В. DELAY, 1992, Re- lationship between allozyme heterozygosity, body size, and survival to natural anoxic stress in the palourde Ruditapes decussatus L. (Bivalvia: Veneridae). Journal Experimental Marine Biology and Ecology, 155: 169-181. САМ, A. J., 1983, Ecology and ecogenetics of ter- restrial molluscan populations. Pp. 597-647, In: W.P. RUSSELL-HUNTER, ed., The Mollusca, Vol. 6, Ecology. Academic Press, NY. CARTER, М. A. & М. ASHDOWN, 1984, Experi- mental studies on the effects of density, size, and shell color and banding phenotypes on the fecundity of Cepaea nemoralis. Malacologia, 25: 291-302. COWIE, В. H. & А. J. САМ, 1983, Laboratory main- tenance and breeding of land snails, with an ex- ample from Helix aspersa. Journal of Molluscan Studies, 49:176-177. DAN, М. & 5. E. В. BAILEY, 1982, Growth, mortal- ity, and feeding rates of the snail Helix aspersa at different population densities in the laboratory, and the depression of activity of helicid snails by other individuals, or their mucus. Journal of Mol- luscan Studies, 48: 257-265. DIEHL, W. & В. К. KOEHN, 1985, Multiple-locus heterozygosity, mortality, and growth in a cohort of Mytilus edulis. Marine Biology, 88: 256-271. EMBERTON, К. C., 1988, The genitalic, allozymic, and conchological evolution of the tribe Mesod- ontini (Gastropoda: Pulmonata: Polygyridae). Malacologia, 33: 71-178. EMBERTON, K.C., in press, Partitioning a mor- phology among its controlling factors. Biological Journal of the Linnean Society. FEVOLDEN, S. E. & $. P. GARNER, 1987, Environ- mental stress and allozyme variation in Littorina littorea (Prosobranchia). Marine Ecology Progress Series, 39: 129-136. FOLTZ, D. M., 1986, Ми! alleles as possible causes of heterozygote deficiencies in the oyster Cras- sostrea virginica and other bivalves. Evolution, 40: 869-870. FOLTZ, D. M., H. OCHMAN 8 В. К. SELANDER, 1984, Genetic diversity and breeding systems in terrestrial slugs of the families Limacidae and Arionidae. Malacologia, 25: 593-605. FOLTZ, О. M., $. Е. SHUMWAY & D. CRISP, 1993, Genetic structure and heterozygosity-related fit- ness effects in the marine snail Littorina littorina. American Malacological Bulletin, 10: 55-60. FOSTER, B. A. & A. E. STIVEN, in press. The ef- fects of age, body size, and site on reproduction in the southern Appalachian land snail Mesodon normalis (Hubricht, 1985). American Midland Naturalist. FOSTER, B. A. & A. E. STIVEN, in review. Regula- tion processes in populations of the land snail Mesodon normalis (Hubricht 1985): testing the interrelationships of site, body size, density, food, growth and survival. GAFFNEY, P. M., 1990, Enzyme heterozygosity, growth rate. and viability in Mytilus edulis: an- other look. Evolution, 44: 204-210. GAFFNEY, P. М. & В. MCGEE, 1992, Multiple pa- ternity in Crepidula fornicata (Linaeus). The Veliger, 35: 12-15. GAFFNEY, Р. М. & T. SCOTT, 1984, Genetic het- erozygosity and production traits in natural and hatchery populations in bivalves. Aquaculture, 42: 289-302. САВТОМ, D. W., 1984, Relationship between mul- tiple locus heterozygosity and physiological en- ergetics of growth in the estuarine gastropod Thais haemastoma. Physiological Zoology, 57: 530-543. GENTILI, M. R. 4 R. A. BEAUMONT, 1988, Envi- ronmental stress, heterozygosity, and growth rate in Mytilus edulis L. Journal of Experimental Marine Biology and Ecology, 120: 145-153. 184 STIVEN GOODFRIEND, G. A., 1986, Variation in land-snail shell form and size and its causes: a review. Sys- tematic Zoology, 35: 204-223. HARTL, D. L. 8 A. G. CLARK, 1989, Principles of population genetics, 2nd ed. Sinauer Associates, Sunderland, Massachusetts. 682 pp. HAWKINGS, A. J., В. L. BAYNE & A. J. DAY, 1986, Protein turnover, physiological energetics and heterozygosity in the blue mussel, Mytilus edulis: The basis of variable age-specific growth. Pro- ceedings of the Royal Society of London, (B) 229: 161-176. HOFFMAN, А. A. & P. A. PARSONS, 1991, Evolu- tionary genetics and environmental stress. Ox- ford University Press. Oxford, England. 284 pp. HOLLEY, М. E. & D. W. FOLTZ, 1987, Effects of multiple-locus heterozygosity and salinity on clearance rate in a backish-water clam, Rangia cuneata (Sowerby). Journal Experimental Marine Biology and Ecology, 111: 121-131. HUBRICHT, L., 1985, The distribution of the native land mollusks of the eastern United States. Fiel- diana, (n.s.) 24: 1-191. KOEHN, R. K. 8 B. L. BAYNE, 1989, Towards a physiological and genetical understanding of the energetics of the stress response. Biological Journal of the Linnean Society, 37: 157-171. KOEHN, R. K., W. J. DIEL 8 T. M. SCOTT, 1988, The differential contribution of individual en- zymes of glycolysis and protenin catabolism to the relationship between heterozygosity and growth rate in the coot clam, Mulinia lateralis. Genetics, 18: 121-130. КОЕНМ, В. К. & $. E. SHUMWAY, 1982, A genetic physiologic explanation for differential growth rate among individuals of the oyster Crassostrea virginica (Gmelin). Marine Biology Letters, 3: 35- 42. KOMAI, T. & S. EMURA, 1955, A study of popula- tion genetics of the polymorphic land snail Bradybaena similaris. Evolution, 9: 400-418. MALLET, A. L., E. ZOUROS, K. E. GARDNER-KEP- KAY 8 K. R. FREEMAN, 1986, Genetics of growth in the blue mussel: family and enzyme- heterozygosity effects. Marine Biology, 92: 475- 482. MITTON, J. В. & М. С. GRANT, 1984, Associations among protein heterozygosity, growth rate, and developmental homeostasis. Annual Reviews of Ecology and Systematics, 15: 479-499. MURRAY, J., 1964, Multiple mating and effective population size т Cepaea nemoralis. Evolution, 18: 283-291. NEI, M., 1977, F-statistics and analysis of gene diversity in subdivided populations. Annuals of Human Genetics, 41: 225-233. NEI, M., 1978, Estimation of average heterozygos- ity and genetic distance from a small number of individuals. Genetics, 89: 583-590. OOSTERHOFF, L. M., 1977, Variation in growth rate as an ecological factor in the land snail Ce- paea nemoralis (L.). Netherlands Journal of Zo- ology, 27: 1-132. PERRY, R. & W. ARTHUR, 1991, Shell size and population density in large helicid land snails. Journal of Animal Ecology, 60: 409-421. PILSBRY, H. A., 1900, Mollusca of the Great Smoky Mountains. Proceedings of the Academy of National Sciences, Philadelphia, 52: 110- 150. PILSBRY, H.A., 1940, Land Mollusca of North America. Academy of Natural Sciences of Phila- delphia, Monographs, 1: 994 pp. SCOTT, Т. М. & В. К. KOEHN, 1990, The effect of environmental stress on the relationship of het- erozygosity to growth rate in the coot clam Mulina lateralis (Say). Journal Experimental Ma- rine Biology and Ecology, 135: 109-116. SAMALLOW, P. B. 8 M. E. SOULE, 1983, A case of stress related heterozygosity in nature. Evolu- tion, 37: 646-649. STIVEN, A. E., 1989, Population biology of two southern Appalachian land snails (Mesomophix spp): variation among six sites with differing dis- turbance histories. Oecologica, 79: 372-382. SWOFFORD, D. L. & В. В. SELANDER, 1989, Bio- sys-1: a computer program for the analysis of allelic variation in genetics, release 1.7. User's manual. University of Illinois, Urbana. UNDERWOOD, A. J., 1989, The analysis of stress in natural populations. Biological Journal of the Linnean Society, 37: 51-78. WILKERSON, L., 1990, SYSTAT: the system for statistics. SYSTAT Inc. Evanston, Illinois. WRIGHT, S., 1978, Evolution and the genetics of populations, Vol. М. University Chicago Press, Chicago, Illinois, 580 pp. ZOUROS, E., 1987, On the relationship between heterozygosity and heterosos: An evaluation of the evidence from marine mollusks. /sozymes: current topics in biology and medical research, 15: 255-270. Liss, New York. ZOUROS, E. 8 D. W. FOLTZ, 1987, The use of allelic isozyme variation for the study of hetero- sis. /sozymes: current topics in biology and med- ical research, 13: 1-59. Liss, New York. ZOUROS, E., М. ROMERO-DOREY & А. L. MAL- LET, 1988, Heterozygosity and growth in marine bivalves: further data and possible explanations. Evolution, 42: 1332-1341. ZOUROS, E., S. M. SINGH, D. W. FOLTZ 4 A. L. MALLET, 1983, Post-settlement viability in the American oyster (Crassostrea virginica): an over- dominant phenotype. Genetical Research, 41: 259-270. ZOUROS, E., S. M. SINGH 4 H. E. MILES, 1980, Growth rate in oysters: an overdominant pheno- type and its possible explanations. Evolution, 31: 856-867. Revised Ms. accepted 29 August 1994 MALACOLOGIA, 1995, 36(1-2): 185-202 DIAGNOSIS OF THE GENUS CRASSOSTREA (BIVALVIA, OSTREIDAE) David R. Lawrence Department of Geological Sciences, Marine Science Program, and Belle W. Baruch Institute for Marine Biology and Coastal Research, University of South Carolina, Columbia, South Carolina 29208, U.S.A. ABSTRACT The oyster genus Crassostrea is the only valid genus in the subfamily Crassostreinae. Char- acters or tendencies considered diagnostic of other crassostreine genera are either environ- mentally controlled or can be found in the type species, С. virginica, or in its direct forerunner, С. gigantissima. Late larval forms of this genus all possess distinctive convexities and ligament placement; adults have a right side promyal passage and a nonorbicular adductor muscle scar. A large size and elongate outline, a cupped left valve, an umbonal cavity, posterior and/or ventral displacement of the muscle scar, large void chambers, and significant nonvesicular chalky deposits are skeletal characters that may be present in some populations or species of this extremely variable taxon. Crassostrea as rediagnosed has few living species; the taxon 1$ evolutionarily conservative. Applying this conservatism to the fossil record, recognizing that chomata are a part of the history of the genus, and realizing that similar evolutionary changes have not been synchronous throughout the geographic range of the genus, are all essential to deciphering the geologic history and evolution of Crassostrea. Ongoing and future biological work should contribute significantly to the understanding of this history. Key words: Bivalvia, Ostreidae, Crassostreinae, Crassostrea, taxonomy, classification. “The first prerequisite in oyster classification is availability of ample material.” (Stenzel, 1971: №1094) INTRODUCTION Stenzel (1971) suggested a diphyletic origin for the oysters and recognized two families, the Gryphaeidae Vyalov, 1936, and the Os- treidae Wilkes, 1810, within the superfamily Ostreoidea Wilkes, 1810 (Table 1A). Polyphyly might be inferred from the subsequent cre- ation of some additional ostreoidean families [e.g. for Crassostrea and its allies by Scarlato 8 Starobogatov (1979a, 1979b) and for the pycnodonteine oysters by Torigoe (1981)]. However, these and other proposals have not been accompanied by clear indications of the evolutionary significance of the taxa involved. By contrast, Malchus (1990), in a work em- phasizing Cretaceous oysters, erected a third ostreoidean family, Palaeolophidae (Table 1B), and used interpretations of shell struc- tures to suggest the historical significance of his taxon. Although arguments for monophyly have been made (Nicol, 1984), and the most recent synopsis of living oysters (Harry, 1985) maintains a two-fold division of the group, Malchus's still more recent work proffers tri- phyletic origins for the oysters. Harry (1985) recognized three subfamilies within the Ostreidae: Lophinae Vyalov, 1936; 185 Ostreinae Wilkes, 1810; Crassostreinae Scarlato & Starobogatov, 1979 (Table 1A). Harry further pointed out that the separation of Crassostrea and its associates at the sub- family level had been presaged by the ar- rangement of taxa in Stenzel (1971). Malchus (1990) erected a fourth subfamily for Juras- sic-Cretaceous ostreid oysters: subfamily Li- ostreinae Malchus, 1990 (Table 1B). The present study questions the groupings of genera within this latter subfamily, and also queries Malchus's placement of genera within the Crassostreinae. One traditional evolutionary view has the crassostreine oysters derived from the os- treine members of the family during the Cre- taceous Period as oysters supposedly moved into more inshore and coastal set- tings (Yonge, 1960: 97). Yet the Crassostrei- nae possess characters that many workers deem to be primitive in nature (discussion in Stenzel, 1971: N1958), suggesting that the geologic history of these two subfamilies may be the reverse of that proposed by Yonge. Malchus (1990) has begun the appli- cation of primitive-versus-derived character recognition (Henning, 1966; Ax, 1987; Funk & Brooks, 1990) to the oysters. A clear under- 186 LAWRENCE TABLE 1. Families and subfamilies within the superfamily Ostreoidea Wilkes, 1810, with their general geologic ranges. Comparison of pre-1990 classification (A) with that proposed by Malchus in 1990 (B). Subfamilies with living representatives marked by an asterisk (*). A. Pre-1990 oyster classification. After Stenzel (1971), Torigoe (1981), Freneix (1982), and Harry (1985). Family Gryphaeidae Vyalov, 1936; Triassic-Neogene Subfamily Gryphaeinae Vyalov, 1936; Triassic-Jurassic Subfamily Exogyrinae Vyalov, 1936; Jurassic-Cretaceous Subfamily Gryphaeostreinae Freneix, 1982; Cretaceous-Neogene “Subfamily Pycnodonteinae Stenzel, 1959; Cretaceous-Neogene Family Ostreidae Wilkes, 1810; Triassic-Neogene “Subfamily Lophinae Vyalov, 1936; Triassic-Neogene “Subfamily Crassostreinae Scarlato & Starobogatov, 1979; Cretaceous-Neogene “Subfamily Ostreinae Wilkes, 1810; Cretaceous-Neogene B. Classification of Malchus (1990: p. 196, adapted from table 17). Family Palaeolophidae Malchus, 1990; Triassic-Cretaceous Subfamily Palaeolophinae Malchus, 1990; Triassic-Cretaceous Family Gryphaeidae Vyalov, 1936; Triassic-Neogene Subfamily Gryphaeinae Vyalov, 1936; Triassic-Jurassic Subfamily Exogyrinae Vyalov, 1936; Jurassic-Cretaceous Subfamily Gryphaeostreinae Freneix, 1982; Cretaceous-Neogene “Subfamily Pycnodonteinae Stenzel, 1959; Cretaceous-Neogene Family Ostreidae Wilkes, 1810; Triassic-Neogene Subfamily Liostreinae Malchus, 1990; Triassic-Neogene “Subfamily Lophinae Vyalov, 1936; Paleogene-Neogene *Subfamily Crassostreinae Scarlato & Starobogatov, 1979; Paleogene-Neogene *Subfamily Ostreinae Wilkes, 1810; (?)Paleogene-Neogene standing of the taxa involved, and most es- pecially of within-taxon variability, is critical to this technique. Hopefully, this paper will encourage the use of phylogenetic view- points through an analysis and diagnosis of the genus Crassostrea. BACKGROUND: CHARACTERS OF THE CRASSOSTREINAE Stenzel (1971), in the Treatise on Inverte- brate Paleontology, provided an historical summary of many of the most important works in ostreid taxonomy. Stenzel, with a background of work including studies by Or- ton (1928), Nelson (1938), Ranson (1943, 1948), Stenzel (1947), Gunter (1950), Thom- son (1954), and Sohl & Kauffman (1964), con- tended that both the anatomy of living oysters and the shell characters of living and fossil oysters must be considered in the develop- ment of any meaningful classification. When dealing with exoskeletons alone, Stenzel ar- gued for primary reliance upon shell microar- chitecture and for the relative taxonomic im- portance of features on the surface of the shells’ internal cavity, because changes in the latter characters may often reflect alterations in the position of soft tissues. By the time Stenzel wrote the Treatise vol- ume, workers had recognized numerous os- treid characters with real or potential value in unraveling the systematic and evolutionary relationships of the group. These features in- cluded larval development and form; adult shell outlines and relative sizes; valve geom- etries; the presence or absence of a promyal passage on the right side of the soft tissues; the geometry and placement of the adductor muscle; and the degree of development of left valve umbonal cavities and also of valve chambers, both void and containing chalky deposits. Among these characters, nonincubatory larval development; a late larval or prodisso- conch II shell with a distinctly more convex left valve and a ligament developed far anterior to any tooth precursors; the presence of a right side promyal passage; and a nonorbicular ad- ductor muscle scar are presently recognized as invariant attributes of crassostreine oysters (Stenzel, 1971; Torigoe, 1981; Егепех, 1982; Harry, 1985; see later section for discussion of larval form; contrast Malchus, 1990: 82, table 9, for muscle scar). Other soft-tissue and uni- fying characters of the Crassostreinae include thickened and food-storing mantle lobes and an accessory heart that “does not receive GENUS CRASSOSTREA 187 TABLE 2. Nominal genera within subfamily Crassostreinae Scarlato 8 Starobogatov, 1979, and characters that have been suggested as distinguishing the other genera from Crassostrea Sacco, 1897. Discussion of these character differentiations in text. Data from Stenzel (1971), Chiplonkar & Badve (1979), Torigoe (1981), Harry (1985), Chinzei (1986), and Moore (1987). CRASSOSTREINE GENERA 1—Crassostrea Sacco, 1897; Cretaceous-Holocene 2—Pseudoperna Logan, 1899; Cretaceous 3—Acutostrea Vyalov, 1936; Cretaceous-Eocene 4—Gyrostrea Mirkamalov, 1963; Cretaceous 5—Indostrea Chiplonkar & Badve, 1976; Cretaceous 6—Bosostrea Chiplonkar & Badve, 1978; Cretaceous 7—Soleniscostrea Chiplonkar 8 Badve, 1979; Cretaceous 8—Cussetostrea Chiplonkar & Badve, 1979; Cretaceous 9—Konbostrea Chinzei, 1986; Cretaceous 10—Striostrea Vyalov, 1936; Eocene, Holocene 11—Saccostrea Dollfus & Dautzenberg, 1920; Miocene-Holocene COMPARISON OF OTHER GENERA WITH CRASSOSTREA Character in Comparison Genus, by numbers above with Crassostrea 2 Denticles or chomata present Adult shell size smaller/larger Greater valve massiveness Umbonal/ligamental areas different x Conical shell form present External ornamentation different Muscle scar placement different Muscle scar geometry different хх IONMOOD> adjacent neobranch units as tributaries” (Harry, 1985: 149). The Crassostreinae may also display a rel- atively large size and generally elongate out- line among oysters, a cupped left valve and flatter right valve, an umbonal cavity under the ligamental area of the left valve, and a pos- terior and/or ventral displacement of the ad- ductor muscle and its scar; they have the abil- ity to produce internal and large valve chambers, which may be filled with nonve- sicular chalky deposits (Gunter, 1950; Tori- goe, 1981; Harry, 1985; Malchus, 1990). [Chalky deposits of crassostreine oysters are nonvesicular at any magnification of light mi- croscopy, but apparently display “‘microves- icles”” under the electron microscope; Car- riker et al., 1980; Harry & Dockery, 1983).] Yet - none of these latter characters is constant in expression within the taxa of this subfamily. Extreme within-taxon variability of crassos- treine oysters had been recognized for many years prior to the appearance of the Treatise volume (for example, Korringa, 1952; Gunter, 1954), and parts of this variability were ac- knowledged by Malchus (1990) in the most recent summary classification of the oysters. жж | C x 4 5 6 it 8 9 10 11 X X X X X X X X X X X X X X X X X X X X X X X X Unfortunately, Malchus (1990: 99, 101) did not formally define his concept of the Cras- sostreinae, but various text discussions and tables are present (esp. pp. 68-98, 196) from which his notions can be extracted. Malchus used shell microstructure as his primary tax- obasis. Within this framework, the Crassos- treinae were characterized by having primarily simply foliated layers in the inner ostracum (non-prismatic layer) of their shells; these lay- ers are arranged in an extremely lenticular fashion and form half or less of a typical valve cross-section. The remaining spaces are chambers, typically large, which may be void or be filled with chalky deposits. The Cras- sostreinae share these basic characters with the Ostreinae, even though minor structural differences within and between the subfam- ilies are present (Malchus, 1990: 69, 87). This microstructural basis caused Malchus to reassign genera within the Ostreidae. Prior to 1990, at least eleven genera had been pro- posed that could be assigned to the Cras- sostreinae (Table 2, top, 1-11). The majority of these taxa had been erected for Creta- ceous representatives. Malchus (1990: 196; Table 1B) apparently removed all of these 188 LAWRENCE Cretaceous genera from the crassostreinae and referred many of them to his new sub- family Liostreinae; this latter subfamily he characterized as having no or few, mostly small chambers that may be void or chalk- filled. Malchus (1990: 201) did not treat the Cretaceous Crassostrea-like genus Konbos- trea Chinzei, 1986, or the similar-aged genera Soleniscostrea Chiplonkar 8 Badve, 1979, and Cussetostrea Chiplonkar & Badve, 1979. Konbostrea is characterized by pre-eminent, large, lenticular areas filled with altered chalky deposits and cannot be assigned to Malchus's Liostreinae. The Crassostrea-like genera that Malchus removed to the Liostreinae [Pseudoperna Lo- gan, 1899; Acutostrea Vyalov, 1936; Gyros- trea Mirkamalov, 1963; Indostrea Chiplonkar 8 Badve, 1976; Bosostrea Chiplonkar & Badve, 1978] all include taxa with relatively small adult stages. If lineages of oysters are marked by increases in adult sizes (see fol- lowing section) then Malchus's (1990: 75-76) own arguments, involving efficiency of cham- ber use in the building of larger or more elon- gate oyster exoskeletons, can be used to suggest that chambering (chalk-filled or void) should also increase in prominence and size through time. At least in the Crassostrea-like oysters, to arbitrarily subdivide them, using the degree and size of chambering as one primary taxobasis, is to create a “horizontal” classification (Newell, 1965) that does not aid the reconstruction of phylogenetic histories. Concomitantly, accepting the placement of Turkostrea Vyalov, 1936, and its allies, in the Crassostreinae (Malchus, 1990: 196) must await more formal reanalyses of these Cen- zoic and largely Eastern Hemisphere taxa. These latter taxa have form characters differ- ent from those in previously recognized members of the subfamily Crassostreinae. Thus, the rest of this paper's discussion does focus upon the 11 genera of Table 2. CRASSOSTREINAE AND CRASSOSTREA Introduction What distinguishes Crassostrea from the other ten genera on Table 2? Or stated con- versely, what criteria or characters have been proposed to separate these other genera from Crassostrea? Distinguishing traits, as suggested by original definers or subsequent and major revisers, are outlined in Table 2 (bottom, A-H). These latter characters form the basis for comparisons with Crassostrea in the following portions of this section. The supposedly distinctive features of the other ten genera include characters men- tioned previously as variable within the sub- family, and involve shell size, massiveness, outline and form; external sculpture of both valves; the presence of denticles or chomata; and such features of the internal shell cavity as muscle placement and muscle scar out- line. Are these valid distinguishing charac- ters? To answer this question, the type spe- cies of Crassostrea must be examined, for “the type species must be elucidated fully first, else the genus [will] remain obscure” (Stenzel, 1971: N1095). Crassostrea virginica and its History Gunter (1950) and Stenzel (1947, 1971) have summarized the nomenclatural history of the species now known as Crassostrea vir- ginica (Gmelin, 1791). By original definition, this extant species is the type species of Crassostrea Sacco, 1897. Crassostrea virgin- ica is rather widespread in the western Atlan- tic Ocean, occurring from the coasts of the Maritime Provinces of Canada south through the Gulf of Mexico and Caribbean Sea to the coast of Brazil (Harry, 1985). Because of its economic importance, this species has been intensively and extensively studied for well over 100 years, with one summary of much of this work on the American (or Atlantic) oyster provided by Galtsoff (1964). Whenever prac- ticable in the following discussions, terms and examples or illustrations are drawn from Galtsoff (1964) and Stenzel (1971) for living oysters, and from Stenzel (1971) and Mal- chus (1990) for fossil representatives; hope- fully, the arguments may then be followed with the minimum of outside sources. In examining the proposed distinguishing characters of crassostreine genera, the phy- logeny of the type species is important, for living taxa are indeed the products of history. Only two proposals for the ancestry of Cras- sostrea virginica have been proffered. Sohl & Kauffman (1964) argued that the American oyster is the extant member of a lineage— their C. soleniscus (Meek, 1871) lineage— that began during the Cretaceous Period, and that the large and massive Tertiary oyster C. gigantissima (Finch, 1824) was the direct precursor of the extant C. virginica. Hopkins (1978), in a published abstract, suggested GENUS CRASSOSTREA 189 FIGS. 1-2. Chomata on right valves of Crassostrea gigantissima (Finch) from (1) the late Eocene of Burke County, Georgia, and (2) the late Oligocene- early Miocene of Onslow County, North Carolina. Localities in text. Views of dorsoposterior right valve margins; valve interior up and dorsal axis to right. 1. Specimen N1-301, showing relict chomata on margin of valve adjacent to, but outside of, plane of commissure, bar = 2.0 mm. 2. Specimen LBC-28, view onto commissural shelf, showing raised rims and slit-like appearance of chomata, bar = 2.0 mm. that the ancestry of С. virginica included the Late Cretaceous species С. glabra (Meek & Hayden, 1857). Among other reasons, Hop- kins chose C. glabra because of its wide- spread occurrence in settings interpreted as brackish water, and he suggested that this ancestry may explain the extreme tolerance of C. virginica, among oysters, to waters of lowered salinities. Unfortunately, there is no written record of Hopkins' views of the immediate precursor of C. virginica. But, to assume static environmental tolerances - through time is tenuous (in an evolutionary sense) and 15 counter to well-documented cases of Cenozoic oysters assigned to Cras- sostrea that lived in rather normal marine settings (e.g. Jimenez et al., 1991). These arguments aside, Crassostrea gigan- tissima is the only crassostreine oyster avail- able to serve as a direct forerunner for the American oyster. This giant, fossil oyster 1$ widespread in the Western Hemisphere and its synonymy when completed (Lawrence, in preparation) should approach at length that of its related Eurasian species С. gryphoides (Schlotheim, 1813; see references in Stenzel, 1971: N1082). That transatlantic migrations provided the first stocks of С. virginica is highly unlikely for two reasons. First, the dis- tances involved are too great for direct mi- gration. The maximum cited larval dispersal distance in oysters is 1,300 km (Stenzel, 1971: N1035); the nearest coastal regions of the Eastern Hemisphere in Europe or Africa to the critical middle-Atlantic coast of North Amer- ica during early Miocene times (see following section on chomata) are farther distant ac- cording to recent Atlantic Ocean basin recon- structions (Allmon, 1990: 111). Secondly, it is clearly nonsimplistic to call upon some spe- cial case of dispersal to happen at an exact and prescribed time in the past, especially when there is no evidence to support these migrations. Endemic Western Hemisphere crassostreine oysters are much more likely predecessors of C. virginica. Thus, the sug- gested differences between and among cras- sostreine oyster genera are examined below using both С. virginica and С. gigantissima. Because of their previous emphasis in the definitions of genera, denticles or chomata (Table 2, bottom, A) are addressed first. Purported Diagnostic Characters of Other Crassostreine Genera Chomata: Chomata are ridgelets or tuber- cules (right valves; Figs. 1, 2) and pits (left valves) that occur on or near the borders of inner valve surfaces (Stenzel, 1971: N1029, figs. J7, J30, J31, J84, J113, J127, J128, J129). Their origin and significance are still not understood. Within the Ostreidae, Mal- chus (1990: 87) recognized thin and un- branched steg-chomata and the reduced pustular chomata, with these features ap- pearing on or near the free growing valve margin or lateral to the ligamental area (in the latter case, as relict chomata). Malchus's useful phrase “relict chomata' may be for- mally redefined to include chomata that do not appear on the latest formed lamellar lay- ers, occupy very marginal positions, and are typically (but not invariably) lateral to the lig- amental area. Stenzel denied the presence of chomata in Crassostrea; this diagnosis has been ac- 190 LAWRENCE FIG. 3. Saccostrea sp., showing smaller, lid-like right valve and elongate left valve with prominent umbonal cavity and numerous void chambers in- ternal to the ligamental area. Adapted from part of Chinzei (1982: fig. 15), bar = 5.0 cm. cepted by Torigoe (1981), Harry (1985), and Malchus (1990) and has been used to distin- guish Crassostrea from numerous other cras- sostreine genera (Table 2, bottom, A). But not all workers have shared this perspective upon the genus Crassostrea. Among neon- tologists, Thomson (1954: especially pp. 162-163) included chomata-bearing taxa in Crassostrea. Stenzel (1971: N1094) dis- missed this view by calling the work of Thomson a “lumping” classification. More pointedly, other paleontologists with access to large collections of Western Hemi- sphere Cretaceous and Cenozoic oysters have not considered the absence of chomata to be a distinguishing character of Crassos- trea (e.g. Sohl & Kauffman, 1964; Woodring, 1982). Sohl & Kauffman’s view of the С. so- leniscus lineage has the taxa increasing in size and massiveness through mid-Cenozoic times, with the more recent C. virginica being smaller and more variable, and a decrease or gradual loss of chomata in this lineage through time. Stenzel (1971: N1028, N1193) was most certainly aware of Sohl 8 Kauff- man's work but remained curiously silent on this differing concept of the genus Crassos- trea, maintaining his position that “the under- lying idea that chomata are a feature impor- tant to classification is sound” (Stenzel, 1971: N1088). More recent work has supported this con- trary view of the genus: “It would be unreal- istic to suppose that the chomata-bearing valves represent a different species, much less a different genus. Wherever they were found, they are associated with valves that lack chomata. Three chomata-bearing valves are attached to the exterior of a right valve that lacks chomata. . . . Aside from the chomata, the two sets of valves are indistin- guishable” (Woodring, 1982: 611-612; dis- cussion of the Panamic and Cenozoic spe- cies Ostrea cahobasensis Pilsbry & Brown 1917, which Woodring assigned to Crassos- trea). Against this backdrop, evidence from С. gigantissima can be examined at that spe- cies” type locality and elsewhere. By original definition (Finch, 1824; Howe, 1937), the type locality of Crassostrea gigan- tissima is Shell Bluff along the Savannah River in Burke County, Georgia (Shell Bluff Landing Ga.-S.C. 1:24000 Quadrangle, 1980 edition, NE 1/4 of NW 1/4). The oysters there occur in late Eocene strata, and Veatch 8 Stephenson (1911: 245) have provided a general description of the stratigraphy in the upper part of the exposures at Shell Bluff. Crassostrea gigantissima is most prominent in a sequence of three superposed oyster- bearing beds, in a partially lithified sand ma- trix. The oyster remains in each bed “coarsen upward,” and the top of each bed is charac- terized by an intact framework of entire valves and/or articulated shells. Because of its prominence as a ledge-former, the basal bed in this sequence has been the focus of collecting over the years. The present owners of Shell Bluff have not allowed collecting at the site for a number of years. However, the same sequence of strata can be recognized 6.7 km SSW of the Shell Bluff exposures, as channeled and eroded remnants exposed in roadcuts on the NW side of Ben Hatcher Road (Shell Bluff Land- GENUS CRASSOSTREA 191 ing Ga.-S.C. 1:24000 Quadrangle, 1980 edi- tion, NW 1/4 of SW 1/4), between Fairfield Church and the Newberry Creek bridge. Be- cause of induration, systematic sampling of these exposures has been impossible. But collections do include oysters with chomata on both right (Fig. 1) and left valves; these features are confined to juvenile stages of growth; of fourteen chomata-bearing right valves presently in hand, ten display relict chomata. Thus, C. gigantissima in its type beds displays chomata. Chomata are more prominent in older members of the Crassostrea soleniscus lin- eage. The Cretaceous species Crassostrea cusseta Sohl & Kauffman, 1964, is another member of this lineage. Specimens of this latter taxon also display both left and right valve chomata; these features were present in a majority (85%) of the valves available to Sohl & Kauffman (1964: H10) for their original description. Chomata in adult C. cusseta do not occur merely as relict chomata; they were produced during more extensive periods of ontogeny and, on the margins of internal cav- ities range ventrally to mediolateral positions (Sohl & Kauffman, 1964: H10). The youngest known Atlantic Coastal Plain occurrences of Crassostrea gigantissima are in latest Oligocene-earliest Miocene strata of east-central North Carolina (Ward et al., 1978). In systematically collected oysters (Lawrence Belgrade Collection or LBC) from Belgrade, Onslow County, North Carolina (Maysville, North Carolina 1:62500 Quadran- gle, 1948 edition, SW 1/4 of NW 1/4; locality described by Lawrence, 1968, 1975), chomata occur infrequently (19 of 141 spec- imens), like the Eocene occurrences are re- stricted to juvenile life stages, with one ex- ception (Fig. 2) do occur solely as relict chomata on adult individuals, and appear as slits with raised rims on right valves only (Fig. 2): In Atlantic Coastal Plain strata, C. virginica first appears in units of early Miocene age from New Jersey (Whitfield, 1894; Richards & Harbison, 1942) and the Delmarva Peninsula. Recent construction along U.S. Highway 13, -8.7 km south of Smyrna, Delaware (Dover, Delaware 1:24000 Quadrangle, 1982 edition, NE 1/4 of NW 1/4), uncovered extensive оуз- ter-bearing beds of late early Miocene age (L. W. Ward, personal communication, 1992). Ongoing examination by the writer of a large collection (> 300 individuals of both right and left valves) of these C. virginica has not yet recorded the presence of chomata. Thus, Malchus's (1990) notion of the ontogenetic loss of chomata is expressed phylogeneti- cally within this one lineage of crassostreine oysters, suggesting that the presence or ab- sence of these features can be a poor hall- mark for the definition of genera within the subfamily Crassostreinae. Hence, along the Atlantic coast of North America, chomata had disappeared in Cras- sostrea by about 18 million years ago (age from L. W. Ward, personal communication, 1992). There are, however, no reasons to as- sume that this event was synchronous throughout the entire geographic range of the genus, or even throughout the range of the lineage which includes the type species. Crassostrea, with its near-cosmopolitan range, has persisted in a wide variety of en- vironments that could have influenced rates of evolutionary change. In a phylogenetic and temporal sense, the lack of chomata 15 not a distinguishing character of the genus Cras- sostrea. Shell Size and Massiveness: The maximum size of shells has been cited as a diagnostic feature of the genera Pseudoperna, Acu- tostrea, Indostrea, and Konbostrea (Stenzel, 1971; Chiplonkar & Badve, 1976; Chinzei, 1985; Table 2, bottom, B). This maximum size is largely a function of growth rates and life span or longevity in individual crassos- treine oysters, and both of these attributes are dependent upon many extrinsic, environ- mental factors (Stenzel, 1971: N1027). With- out analyzing living oysters of known and dif- fering ages, or without detailed examination and interpretation of periodic growth fabrics (e.g. those of ligamental areas; Stenzel, 1971: N1014-N1016) growth rates and life spans cannot be traced in space and/or time. Such data for crassostreine oysters are meagre at present (Stenzel, 1971: N1014-N1016). Even if patterns of life span are found, they in turn must be interpreted in an acceptable fashion. For example, a preliminary and unpublished analysis by the writer suggests that the early Miocene transition from С. gigantissima to С. virginica along the Atlantic Coast of North America did involve decreases in maximum life span and a resulting overall smaller size for adult American oysters, but the reason or reasons for these changes remain obscure. Stenzel (1971: N1027) suggested that fossil oysters tend to be larger than still-living ones primarily because of oyster fishing pressures 192 LAWRENCE by humans, but certainly more than a lack of human intervention is responsible for the larger sizes of many fossil crassostreine oys- ters. Among shell sizes the extreme elongation of Konbostrea warrants special note, be- cause valve heights of over one meter have been cited for this taxon (Chinzei, 1986). But Crassostrea gigantissima was itself appropri- ately named. Heights of over 50 cm for C. gigantissima were measured by the writer in the outcrops at Belgrade, North Carolina; valve heights over 66 cm (26 in) have been recorded for other North Carolina occur- rences of С. gigantissima (Loughlin et al., 1921: 126); and indistinct molds in late Eocene blocky, calcareous clays at Griffins Landing, Burke County, Georgia (Girard, Ga.- S.C. 1:24000 Quadrangle, 1964 edition, NE 1/4 of NW 1/4) suggest even greater heights for this precursor of the American oyster. In summary, maximum adult size is a poor ge- neric designator among crassostreine oys- ters. The use of maximum size may be useful for species differentiation within a genus, so long as these size differences can be related to reasonable interpretations of life histories and the influence of external, environmental controls. Significant left valve thickening, largely through chamber formation, has been cited as a common characteristic in taxa of the ge- nus Striostrea (Harry, 1985: 150; Table 2, bot- tom, C). But Crassostrea gigantissima dis- plays this same trait. In the LBC collection of C. gigantissima, left valve thicknesses range to over 7 cm and estimated valve height to thickness ratios are 3 and lower (compare Harry, 1985: 149-150). Void chambers are prominent in cut sections of thick C. gigan- tissima valves. In some of the LBC individu- als, valve thickening occurred without appar- ent and significant increases in valve height, and this same situation was described by Harry (op. cit.) for Striostrea. This attribute, extreme valve thickening, is by no means confined to members of Striostrea because it occurs in the lineage including the type spe- cies of Crassostrea. Geometries of Umbonal and Ligamental Ar- eas: Freneix (1972: 98-99; 1982) pointed out the crassostreine characters of Gyrostrea Mirkamalov, 1963, and removed that genus from the Exogyrinae, where it had been placed by Stenzel (1971: N1125). Spirally coiled growth during postlarval and immature stages, reflected in ligamental and umbonal area outlines, is one important distinction of Gyrostrea (Stenzel, op. cit; Table 2, bottom, D). However, Crassostrea gigantissima in the LBC materials includes individuals with sim- ilar coiling characters, with spiraling ranging to over three-quarters of a volution. Illustra- tions of Gyrostrea (Stenzel, 1971: fig. J99) do not all display preserved coiling, and these figures exhibit only part of the variability that may be found in the systematically collected suite of C. gigantissima from Belgrade, North Carolina. Other differences in the outlines of the lig- amental or cardinal area (Table 2, bottom, D) have been cited as diagnostic for numerous crassostreine genera by Chiplonkar & Badve (1979: 445). But these outlines are quite vari- able in Crassostrea virginica (Galtsoff, 1964: figs. 18, 19, 21, 22, 34, 407429540222: 385) and can be influenced by age class and a variety of extrinsic and environmental pa- rameters (Galtsoff, 1964: 16). The LBC spec- imens of C. gigantissima display ostreoid, gy- rostreoid, and turkostreoid ligamental area outlines (Siewert, 1972; Malchus, 1990: 77) and also exhibit the variability in overall shell form (plate, triangular, and near-sickle shapes; Malchus, 1990: 89-91) that accom- panies this spectrum of dorsal region fea- tures. Valve profiles (convex versus concave) strongly depend upon substratum geometry, crowding, and other extrinsic factors and, in the LBC materials, several left valves of C. gigantissima display concave profiles; the differences in profiles among genera outlined by Malchus (1990: 95) thus become moot points. In summary, valve form in dorsal and other regions shows considerable within- taxon variability in crassostreine oysters, and these differences cannot be used to separate other proposed genera of the subfamily from Crassostrea. Conical Shell Form: Conical or cup coral-like shell form has been stated to be distinctive of three crassostreine genera—in the external form of older individuals of Striostrea and some ecomorphs of Saccostrea, and in the internal cavity of gerontic adults of Konbos- trea (Stenzel, 1971; Harry, 1985; Chinzei, 1986; Table 2, bottom, E). Evolutionary con- vergence toward a cone-shaped form has been well documented in a number of bi- valved organisms, including both molluscs (Yonge, 1962; Perkins, 1969) and brachio- pods (Rudwick, 1961; Williams & Rowell, GENUS CRASSOSTREA 193 1965). Crassostreine shell characters in- volved in this conical form include the devel- opment of left valve cupping and umbonal cavities, and the production of prominent chambers, both void and with chalky depos- its. As explanations, these conical shell forms have been related to substrata and shell crowding (Stenzel, 1971: N1135; Chinzei, 1986), and the suggestion that adult oysters may continue to accrete their exoskeletons without changing the volume of either their soft tissues or their shell's internal cavity (Stenzel, 1971: N1014). Conical external form in Striostrea and Saccostrea is caused by the production of rather prominent left valve umbonal cavities during the development of elongate ligamen- tal areas on that valve. Numerous chambers lie beneath the ligamental area; in dorsoven- tral sections of left valves, these chambers have a strongly convex/concave outline (Fig. 3; Chinzei, 1982). But these same growth patterns, including the development of strik- ing umbonal cavities, occur in populations of Crassostrea (Stenzel, 1971: fig. J101, 1b, 2a); only the degree of development of these fea- tures separates Crassostrea from Striostrea and Saccostrea. That such differences help to confer separate generic status is doubtful. Conical internal cavity form was achieved in a very different fashion in the elongate Konbostrea. During elongation, growth of these oysters in dorsal interior regions of the shells was strongly directed toward the op- posing valve. This growth pattern involved the production of chalky deposits, strikingly reducing the internal cavity volume in dorsal shell regions; soft tissue connections with the ligament were maintained through a small conical opening; in gerontic forms of Kon- bostrea, the ligament was most certainly dys- functional (Chinzei, 1986). These latter growth patterns occur in more elongate specimens of Crassostrea gigantis- sima from the Belgrade collection. In some C. gigantissima, ventral displacement of the in- ternal cavity was accomplished by means of the formation of void chambers, but, with their less extreme shell heights, the distinctive ‚ conical and dorsal ends of the internal cavity did not develop. Two individuals (LBC-118, 128) apparently maintained a functional liga- ment by producing a discontinuous, saltated ligamental area during ontogeny. Aspects of growth related to environmentally controlled and strong elongation are thus similar in Crassostrea and Konbostrea, given the gen- eral plasticity of oysters (Gunter, 1954). Only growth details, and the degree of their ex- pression, separate the two genera. External Ornamentation: The absence of coarse radial ornamentation (‘‘ribs’’) on left valves was cited by Chiplonkar & Badve (1976, 1979) as diagnostic for their genera Indostrea, Bosostrea, and Cussetostrea (Ta- ble 2, bottom, F). However, left valve ribs may or may not be present in Crassostrea virgin- ica, and the development of these valve fea- tures in the American oyster is strongly influ- enced by local environmental factors (Galtsoff, 1964: 18, figs. 4, 15, 21). In the Chesapeake Bay area of North America, for example, commercial oyster fishers have recognized that ribbing is most common in “sand oysters” from intertidal or high sub- tidal firm bottoms, and in “reef oysters” from intertidal clusters (Kent, 1988). The presence or absence of left valve ribbing is by no means a character worthy of use in distin- guishing other crassostreine oysters from Crassostrea. The occurrence of right valve riblets in the prismatic layer of that valve was noted as a marker for Striostrea by Stenzel (1971: N1136; Table 2, bottom, F). Although pris- matic shell layers may not be commonly pre- served in fossil crassostreine oysters, these very riblets appear in prismatic layers on dor- sal regions of right valves in Crassostrea gi- gantissima from the LBC materials. In my opinion, one junior synonym of C. gigantis- sima is the taxon Ostrea alabamiensis Lea, 1833, described from the Eocene of its namesake state. Right valve radial riblets from juvenile stages have been cited as di- agnostic for this taxon (Dall, 1898: 679). Per- haps the Alabama occurrences of this oyster led Stenzel (1971: N1136) to extend the range of Striostrea back into the Eocene, be- cause earlier life stages of this latter genus have the typical ostreiform (and not conical) habit. Otherwise Stenzel recognized Strios- trea aS a present-day genus. Regardless of this interpretation, the presence of right valve riblets cannot be used to separate other crassostreine genera from Crassostrea. Muscle Scar Position: Placement of the muscle scar can be influenced by the pres- ence of the promyal passage in crassostreine oysters. This passage is essentially an exten- sion of the epibranchial chamber lying be- tween the mouth and the adductor muscle; the development of the passage may involve 194 LAWRENCE a posterior displacement of the muscle (Nel- son, 1938; Gunter, 1954). The passage may be further accommodated by one or more of: increased (relative) shell height, increased left valve cupping, increased development of the left valve umbonal cavity, and migration of adductor muscle attachment to a more ven- tral position (Gunter, 1950; Sohl & Kauffman, 1964). Thus, it is possible to correlate anat- omy with shell features of the Crassostreinae (Sohl 4 Kauffman, 1964; Stenzel, 1971). The position of the muscle scar along both dorsoventral and anteroposterior axes has been cited as diagnostic for some genera of crassostreine oysters (Stenzel, 1971; Chip- lonkar & Badve, 1979; Table 2, bottom, G). Chiplonkar 8 Badve (1979) erected the genus Cussetostrea (type species Crassostrea cus- seta Sohl & Kauffman, 1964), and cited a dor- soposterior muscle insertion as one charac- teristic of their taxon. This diagnosis is based upon a misinterpretation of Sohl & Kauff- man’s discussion of Crassostrea cusseta. |n- deed dorsoposterior muscle scar vestiges were cited by Sohl & Kauffman (1964: H10). However, these are interior or within-valve remnants of the oblique track of successive muscle insertion areas (the hypostracum) produced by the oyster during its ontogeny. No discernible muscle scars appear on the internal cavity surfaces of large Crassostrea cusseta specimens, in ventroposterior or other positions, and Sohl & Kauffman inter- preted this condition to reflect “atrophy of the adductor muscle and subsequent cover- ing of the last formed scar by additional lay- ers of calcite during late maturity and old age” (Sohl & Kauffman, 1964: H11). When formed, the muscle scars of Crassostrea cus- seta were most likely posterior and medial or ventral in position. A medial position of the adductor along the dorsoventral axis has been cited for some forms of Striostrea (Stenzel, 1971: N1136). But measures of oyster shell form, measures that relate muscle position to that of the soft tissue mass, have not been made (compare Galtsoff, 1964: fig. 42). Numerous Crassos- trea virginica valves and shells occur along the South Carolina coast (in both present-day and archaeological contexts), the muscle scar placement of which may be qualitatively described as posterior and dorsoventrally medial. Biologically meaningful valve mea- sures (Such as a three-dimensional shell mid- line; Sohl & Kauffman, 1964: H4-H5) might be developed and used to quantify muscle scar placement along both dorsoventral and anteroposterior body axes. Muscle Scar Geometry: A kidney-shaped muscle scar 1$ cited as diagnostic for the ex- tant genera Striostrea and Saccostrea by Chiplonkar & Badve (1979: 445; Table 2, bot- tom, H). However, “reniform” is but one qualitative descriptor for the gibbous, “соп- cave,” or nonorbicular muscle scar outlines known in the crassostreine oysters (Stenzel, 1971: N963). These outlines are, in turn, strongly dependent upon overall shell form. In Crassostrea, the “typical” scar outlines with abrupt dorsal ends are most obvious in the relatively common and elongate forms. Yet both Galtsoff (1964: fig. 22) and Stenzel (1971: fig. J8) figured subovate/trigonal valves of С. virginica, the scars of which have rounded dorsal ends and a distinctly kidney- shaped outline. Galtsoff (1964: fig. 50) illus- trated the extreme variability of form within C. virginica muscle scars, and Harry (1985: 154) interpreted the muscle scars of all Crassos- treinae as reniform in outline. Furthermore, in the fossil record, recognition of scar outline details may be hampered by exfoliation of the surrounding lamellae of the internal shell cav- ity. Muscle scar outline has no inviolable role in the differentiation of crassostreine genera. Summary: The fossil genera Pseudoperna, Acutostrea, Indostrea, Bosostrea, Solenis- costrea, Cussetostrea, and Kombostrea can- not be separated from Crassostrea, because all of the tendencies or diagnostic characters proposed by original definers or subsequent and major revisers of these taxa are either environmentally controlled or can be found in Crassostrea virginica, the type species of Crassostrea, or its immediate ancestor Cras- sostrea gigantissima. Very likely, the reported fossil occurrence of Striostrea includes spec- imens referable to Crassostrea gigantissima. Other Aspects of Living Crassostreine Oysters Introduction: The genera of living crassos- treine oysters (Crassostrea, Striostrea, and Saccostrea), merit additional comments based upon developmental, genetic, and biogeographic-evolutionary viewpoints. A complete presentation of crassostreine life, biogeography, and evolutionary history is be- yond the scope of the present work; the fol- lowing sections are intended to show some of the problems and prospects in developing GENUS CRASSOSTREA 195 these histories from a taxonomic point of view. Larval Shells: Larval shells have been ac- cepted as useful in oyster classification since the studies of Ranson (1939, 1943, 1948, 1967) on living species. Ranson recognized only one genus (Crassostrea) in the presently acknowledged members of the Crassostrei- nae, but Dinamani (1976) used studies suc- ceeding those of Ranson (Pascual 1971, 1972; Dinamani, 1973, 1976) to differentiate larvae of Saccostrea from those of Crassos- trea within the subfamily. Critical to this anal- ysis were larval shell form elements in the prodissoconch || (late larval) stage. Dinamani (1976) noted that larval Saccos- trea has a hinge margin that remains unmod- ified throughout larval development and that includes two equal groups of tooth pre- cursors (two to a group) and orthogyrate um- bones. Conversely, Crassostrea prodisso- conch Il larvae display inequilateral growth in the hinge margin, with posterior tooth precur- sors decreasing in prominence and umbones tending toward the opisthogyrate condition. Dinamani (1976: 99) further pointed out that “early larval stages of the Crassostrea type have an additional pair of teeth in the left valve.” By contrast Waller (1981: 47-48) chose to accent general similarities in hinge margins among all the Ostreidae during early prodis- soconch | stages, and he pointed out that Crassostrea starts the prodissoconch | phase with a relatively small size. In compar- ison to Ostrea, prodissoconch ll growth in Crassostrea contributes more significantly to the overall shape of mature larvae. Because left valve convexity differences appear to be- gin with prodissoconch Il growth, mature Crassostrea larvae have left valves that more greatly exceed the right valves in convexity, and left valve umbones that extend farther over the hinge margin, in comparison with their ostreine relatives (Pascual, 1971, 1972; Carriker & Palmer, 1979; Waller, 1981). These same and distinctive geometries also apply to species assigned to Saccostrea (Dina- «mani, 1973, 1976). Antero-posterior differ- ences in growth among the Crassostreinae only develop subsequent to the onset of these other and shared traits. Are such late larval differences significant enough to help confer separate generic status upon Saccos- trea? The answer to this question need not be affirmative. Perspectives of Genetics: Studies of genet- ics provide additional insights into the taxon- omy of crassostreine oysters. Hybridization tests have been widely applied to these oys- ters, with seminal studies by Galtsoff & Smith (1932) and Imai & Sakai (1961). Indeed Sten- zel (1971: N1135) argued for separating Sac- costrea from Crassostrea because ‘‘species of these two genera cannot be made to crossfertilize each other.” More recently Chanley 8 Dinamani (1980) have questioned the gametic incompatibility of taxa assigned to Crassostrea and Saccostrea and have noted (Chanley 8 Dinamani, 1980: 120) that “it is known that sometimes one population of Crassostrea will hybridise with other spe- cies while another population of the same species will not.” These writers, however, did not present evidence to support that last statement. Even if this claim 1$ true, and the results of single hybridization experiments must be evaluated with caution, the failure to achieve partial or complete success in re- peated hybridization studies cannot be used to support the recognition or creation of sep- arate crassostreine genera. For reproductive isolation, however achieved, has been one hallmark of biologists’ species concept for many decades (Mayr, 1988). The presence of this isolation does not demand the creation of higher, supraspecific taxa; rather, genetic cohesiveness may be viewed as one logical consequence of speciation within genera. Using crossed immuno-electrophoresis techniques for determining genetic dis- tances, Brock (1990) claimed to substantiate not only the separation of Ostrea from the crassostreine oysters, but also the separa- tion of Saccostrea from Crassostrea. These analyses used pooled tissue homogenates from about 50 individuals from each of six species, two species from each of the three genera. Genetic distances determined were 0.25-0.29 between Ostrea and Crassostrea, 0.32-0.33 between Ostrea and Saccostrea, and 0.22-0.26 between Crassostrea and Saccostrea (Brock, 1990: 61). These data can be interpreted in a variety of ways. First of all, arbitrary values of indices of dissimilarity (or similarity) cannot be used to differentiate taxa at the various hierarchical levels; nor can dis- similarities deemed appropriate (for whatever reason) for use in one taxon necessarily be transferred to another. With the same general range in genetic distances, it can be argued that Brock's data fail to support the separa- tion of these three genera into more than one 196 LAWRENCE family. Also, because of accepted subfamilia! biological differences between Ostrea and Crassostrea, it can be argued conversely, us- ing these same data, that all three genera belong in separate families. Brock (1990: 62) pointed out some of these problems in inter- pretation. Furthermore, the use of relayed and introduced oysters (Japanese oysters from Oregon, USA) without further explana- tion does decrease the credibility of Brock's data set. Endemic, natural, and “wild” pop- ulations of oysters, collected far from re- search stations that have a history of han- dling introduced species, should be actively sought in such studies of genetics. Difficul- ties in data gathering and interpretation in studies of oyster genetics abound. Data (Buroker et al., 1979a, b; number of individuals analyzed unpublished) gathered using gel electrophoresis techniques indicate that average heterozygosities for three pre- sumed species assigned to Saccostrea (17- 19%) fall within the range of six presumed species assigned to Crassostrea (6-22%), and that the taxa assigned to Saccostrea have greater “between-species” genetic similarities than do taxa assigned to Crassos- trea. Buroker et al. (1979b: 179) used this ob- servation to suggest, following the fossil record, that “the Saccostrea genus 1$ the more recently evolved oyster lineage of the two,” but they did not address arguments for or against the separation of Saccostrea from Crassostrea (see next section of this study). In attempts to determine genetic relatedness, both nuclear and cytoplasmic genetic struc- tures should be considered, and caution should be used in “inferring population ge- netic structure and gene flow from any single class of genetic markers” (Karl 8 Avise, 1992: 102). In the future, expanded taxonomic and geographic studies of crassostreine mito- chondrial DNA (Reeb 8 Avise, 1990; Avise, 1992) should provide meaningful data for the recognition of vicariants and the analysis of paleobiogeographic events. Used in concert with other studies of oyster genetics, such data should help to quell past unnecessary and unfounded speculation about oyster migrations (Stenzel, 1971: N1027; Durve, 1986). Spatio-Temporal Viewpoints: The genus Crassostrea, as recognized by Stenzel (1971), Torigoe (1981), and Harry (1985), has few liv- ing species or superspecies (for superspecies concept, see Buroker et al., 1979b). In his synopsis, Harry (1985: 153, 156), assigned only four species to Crassostrea: the Atlantic or American oyster, C. virginica; the Portu- guese oyster, C. angulata (Lamarck, 1819) from the eastern North Atlantic; the Japanese oyster, C. gigas (Thunberg, 1793), from west- ern Pacific and Indian Oceans; and C. colum- biensis (Hanley, 1846) from the eastern Pa- cific. But work predating Harry’s has indicated that the Portuguese and Japanese oysters are the same species, based upon likenesses in both adult and larval shells, easily achieved hybridization, and normal meiosis and mitosis in the hybrids (Imai & Sakai, 1961; Menzel, 1974). This finding was foreshadowed by the work of Rutsch (1955), who combined many fossil Cenozoic Crassostrea of the Eurasian Tethyan Seaway into the single fossil taxon C. gryphoides (Stenzel, 1971: N1081-N1082). The geological record does not demand that the Portuguese and Japanese oysters are separate species, and the fossil record does not promote the notion that the Japanese oys- ter may have been imported into the eastern Atlantic by humans (compare discussion in Buroker et al., 1979a). That long-separated and present-day populations of the Portu- guese and Japanese oysters can hybridize is testament to the conservative nature of the genus Crassostrea. Phylogenetically, the first appearances of still-living Crassostrea species are marked by decreases in maximum adult size, and the reasons for these changes are still not clear. The timing of these appearances, and hence rates of evolutionary change, have not been the same throughout the geographic range of Crassostrea; reasons for these diachronous changes also need to be explored. By the end of the early Miocene, Crassostrea in the northwestern Atlantic Ocean had the essen- tial appearance of C. virginica, while mem- bers of the lineage to which C. virginica be- longs, in Panamic regions, still bore chomata on both left and right valves (Woodring, 1982: 612, pl. 94). Crassostrea of similar age from North African and/or European Tethyan realms still included extremely elongate and massive-shelled populations, with some indi- viduals bearing chomata (Hoernes & Reuss, 1870; Newton & Smith, 1912). The originally designated type species of the chomata- bearing genus Saccostrea, S. saccellus (Du- jardin, 1835), = S. cuccullata (Born, 1778), appeared in Europe during the Miocene (Doll- fus & Dautzenberg, 1920), where and when GENUS CRASSOSTREA 197 undoubted Crassostrea species Боге chomata. These same types of Crassostrea may have been present then in other parts of the world, and this taxon in the Neogene (Mi- ocene-Holocene) fossil record needs to be closely re-examined in Sub-Saharan African, northern and southern South American, In- dian Oceanic, Asian, and Pacific Oceanic ar- eas. Interestingly, on mangrove forest or rocky coastlines, living crassostreine oysters of many parts of the world are regularly as- signed to Saccostrea and Striostrea (Stenzel, 1971: N1135; Harry, 1985: 150), whereas similar oysters of the Caribbean and West Indies regions are referred to Crassostrea [C. rhizophorae (Guilding, 1828), = С. virginica; Newball 8 Carriker (1983), Littlewood & Donovan (1988)]. This difference in assign- ment, and the lack of living taxa assigned to Saccostrea and Striostrea in the western North Atlantic, likely reflect the geographi- cally varying rates of evolution within the ge- nus Crassostrea. Regardless of details, Crassostrea is the only crassostreine oyster available to yield the taxa presently known by Striostrea and Saccostrea. The pivotal question becomes: are differences between and among the spe- cies significant enough to warrant placement in separate genera? One response is obvi- ous: because the lineage of the type species of Crassostrea is so varied in space and time, should anything less be expected of the ge- nus with regard to both soft tissues and ex- oskeletons? Recognition of but one crassos- treine genus would still include, very likely, less than ten living species within Crassos- trea (Harry, 1985: 149-156). This latter rec- ognition, without subgenera, will help to pro- vide focus upon a number of critical aspects of this taxon. Paleontologists have continued to define numerous typological species of crassos- treine oysters using inadequate samples from the fossil record. If the “expanded” ge- nus has fewer than ten living species, nothing different should be expected at a given “time plane” of the Cenozoic fossil record of this ‘conservative and variable taxon. Our geolog- ical knowledge of the crassostreine oysters 1$ limited by fragmentary preservation of strata representing shallow marine and near-shore life environments, but enough productive lo- calities exist (Lawrence, 1968; Laurain, 1980; Moore, 1987; Jimenez et al., 1991) to provide the materials for thorough analysis of vari- ability in form, thus leading to revised synon- ymies and geographic ranges for fossil spe- cies of the genus. Recognition that chomata are a part of the history of Crassostrea, and that evolutionary changes in the genus have been diachronous over the face of the earth, are keys to deciphering the geologic history of these oysters. Because chomata are so varied in their occurrence and expression, very large suites of specimens, from con- trolled intervals of time, are necessary for this geologic work. Conclusions In sum, from both paleontologic and neon- tologic points of view, there are no compel- ling reasons to recognize more than one genus within the crassostreine oysters. Dis- tinguishing one conservative, plastic taxon may be the only way to focus attention upon the evolutionary history of these oysters. SYSTEMATICS OSTREIDAE WILKES, 1810 CRASSOSTREINAE SCARLATO 8 STAROBOGATOV, 1979 CRASSOSTREA SACCO, 1897 Synonymy Taxa newly added to the synonymy of Stenzel (1971: N1128) are marked by an as- terisk (*). Crassostrea Sacco, 1897: 15 Gryphaea Fischer, 1886: 927 [non Lamarck, 1801: 398] *Pseudoperna Logan, 1899: 95 Crassostrea (Euostrea) Jaworski, 1913: 192 *Saccostrea Dollfus 8 Dautzenberg, 1920: 471 Dioeciostrea Orton, 1928: 320 Crasostrea Koch, 1929: 6 (nom. null.), fide Stenzel, 1971: N1128 Dioeciostraea Thiele, 1934: 814 (nom. null.) *Saxostrea lredale, 1936: 269 *Striostrea Vyalov, 1936: 17 Angustostrea Vyalov, 1936: 18 *Acutostrea Vyalov, 1936: 18 Grassostrea Vyalov, 1948: 23 (nom. null.) Somalidacna Azzaroli, 1958: 115 Crassotrea Miyake & Моча, 1962: 599 (пот. null.) 198 LAWRENCE *Sanostrea Miyake & Noda, 1962: 599 (nom. null.) *Gyrostrea Mirkamalov, 1963: 152 *Indostrea Chiplonkar 4 Badve, 1976: 245 *Bosostrea Chiplonkar 4 Badve, 1978: 106 *Cussetostrea Chiplonkar 4 Badve, 1979: 443 *Soleniscostrea Chiplonkar & Badve, 1979: 444 *Striostrea (Parastriostrea) Harry, 1985: 151 *Konbostrea Chinzei, 1986: 140 [non Seilacher, 1984: 217 (nom. nud.)] Diagnosis Nonincubatory larvae. Late larval shells with left valve greatly exceeding right valve in convexity; left valve umbone significantly overhanging hinge line; ligament developed far anterior to all tooth precursors. Adults with right side promyal passage, which may be accommodated by one or more of: a left valve umbonal cavity, left valve cupping, dorsoventral elongation, and poste- rior and/or ventral displacement of the ad- ductor muscle; none of the latter characters constant in its expression within the genus. Adductor muscle scar nonorbicular. Valve chambers ranging to very prominent. Chalky deposits, when present, nonvesicular under light microscopy. Remarks This diagnosis must not be interpreted as merely a “lumping” one, nor will it necessar- ily lead to the reinstatement of only three genera of living oysters (Ranson, 1943, 1948). Recognition of but one crassostreine genus, with very likely fewer than ten living species, does not require the consideration of separate subgenera, and should help to provide new focus upon the total natural his- tory of this taxon. This diagnosis also does not ignore or dis- miss the many fine studies on oyster biology of the past century. Rather, it restates what oyster biologists have emphasized for many years, the extreme variability of the shells of oysters in the genus Crassostrea (Galtsoff, 1964: 27), and extends this concept to soft tissues as well. In understanding phyloge- nies, nothing has been gained by using the abundance of anatomical data to erect new genera and subgenera of living crassostreine oysters. This revised taxonomy should serve to highlight some of the past and ongoing biological works that are vital to our under- standing of the history of these oysters. One guidepost for future studies, in both present-day and ancient settings, must be a continuing realization that the ultimate and unshakable classification of oysters cannot be constructed by a single investigator (Vy- alov, 1948; Stenzel, 1971: N1093). Workers should also remember that “oysters are among the most plastic organisms known” (Gunter, 1954: 134). ACKNOWLEDGMENTS My knowledge of crassostreine oysters has been aided by the ability to study living, archaeological, and fossil representatives, in both field and laboratory/museum contexts, in North America, Europe, and Africa. Since 1963 at various times, this work has been financially supported by my parents; H. H. Hess and the (then) Department of Geology, and the Boyd Fund, at Princeton University; the U. S. National Science Foundation through its Graduate Fellowships Program, Science Faculty Fellowships Program, and International Program; the Society of the Sigma Xi; the Belle W. Baruch Institute for Marine Biology and Coastal Research, Ma- rine Science Program, Department of Geo- logical Sciences, and Research and Produc- tive Scholarship Committee at the University of South Carolina; the DuPont Company through funds provided by the U. $. Depart- ment of Energy; Viro Group, ETE Division; and lastly numerous governmental and pri- vate organizations dealing with the conserva- tion and preservation of the cultural history of southeastern United States coastal regions. The unreferenced collections of fossil cras- sostreine oysters cited in this paper are pres- ently under the writer's control; they will be permanently curated at The Charleston Mu- seum, Charleston, South Carolina. | Thank E. V. Coan, B. C. Coull, G. M. Davis, A. C. Lawrence, and three anonymous reviewers for thoughtful comments on this work. This 1$ Contribution 1020 of the Belle W. Baruch In- stitute for Marine Biology and Coastal Re- search, University of South Carolina. LITERATURE CITED ALLMON, W. D., 1990, Review of the Bullia group (Gastropoda: Nassariidae) with comments on its evolution, biogeography, and phylogeny. Bulle- tins of American Paleontology, 99(335): 1-179. GENUS CRASSOSTREA 199 AVISE, J. C., 1992, Molecular population structure and the biogeographic history of a regional fauna: a case history with lessons for conserva- tion biology. Oikos, 63: 62-76. AX, P., 1987, The phylogenetic system. New York, Wiley, 698 pp. AZZAROLI, A., 1958, L'Oligocene e il Miocene della Somalia, stratigrafia, tettonica, paleontolo- gia (microforaminiferi, coralli, molluschi). Paleon- tographica ltalica, new series, 22: 1-143, 36 pls. BORN, |., VON, 1778, Index rerum naturalium musei Саезаге! Vindobonensis, Pars l.ma, Testacea. Verzeichniss der natúrlichen Seltenheiten des K. К. Naturalien Kabinets zu Wien, Erster Theil, Schalthiere. Vienna, Officina Krausiana, 458 pp. BROCK, V., 1990, Intergeneric distances between Ostrea, Crassostrea, and Saccostrea, studied by means of crossed immuno-electrophoresis. Ma- rine Ecology Progress Series, 68: 59-63. BUROKER, N. E., W. К. HERSHBERGER & К. К. CHEW, 1979a, Population genetics of the family Ostreidae. |. Intraspecific studies of Crassostrea gigas and Saccostrea commercialis. Marine Bi- ology, 54: 157-169. BUROKER, М. E., W. К. HERSHBERGER & К. К. CHEW, 19796, Population genetics of the family Ostreidae. Il. Interspecific studies of the genera Crassostrea and Saccostrea. Marine Biology, 54: 171-184. CARRIKER, M. R. & R. E. PALMER, 1979, Ultra- structural morphogenesis of prodissoconch and early dissoconch valves of the oyster Crassos- trea virginica. National Shellfisheries Association, Proceedings, 69: 103-128. CARRIKER, M. R., R. E. PALMER & R. S. PREZANT, 1980, Functional ultramorphology of the dissoconch valves of the oyster Crassostrea virginica. National Shellfisheries Association, Proceedings, 70: 139-183. CHANLEY, P. & P. DINAMANI, 1980, Comparative descriptions of some oyster larvae from New Zealand and Chile, and a description of a new genus of oyster, Tiostrea. New Zealand Journal of Marine & Freshwater Research, 14: 103-120. CHINZEI, K., 1982, Morphological and structural adaptations to soft substrates in the Early Juras- sic monomyarians Lithiotis and Cochlearites. Lethaia, 15: 179-197. CHINZEI, K., 1986, Shell structure, growth, and functional morphology of an elongate Creta- ceous oyster. Palaeontology, 29: 139-154. CHIPLONKAR, G. W. & R. M. BADVE, 1976, Palae- ontology of the Bagh Beds, Part V, Ostreina (i). Indian Academy of Sciences, Proceedings, (B) 83: 244-256. CHIPLONKAR, G. W. &R. M. BADVE, 1978, Palae- ontology of the Bagh Beds, Part V, Ostreina (ii). Indian Academy of Sciences, Proceedings, (B) 87: 105-117. CHIPLONKAR, G. W. & R. M. BADVE, 1979, Tax- onomic comments on subfamily Ostreinae Rafinesque. Indian Academy of Sciences, Pro- ceedings, (B) 88: 441-447. DALL, W. H., 1898, Contributions to the Tertiary fauna of Florida, with especial reference to the Silex-beds of Tampa, and the Pliocene beds of the Caloosahatchie River, including in many cases a complete generic revision of the generic groups treated of and their American Tertiary species, Part IV: 1, Prionodesmacea: Nucula to Julia. 2, Teleodesmacea: Teredo to Ervilia. Wag- ner Free Institute of Philadelphia, Transactions, 3(4): 571-947, pls. 23-35. DINAMANI, P., 1973, Embryonic and larval devel- opment in the New Zealand rock oyster, Cras- sostrea glomerata (Gould). The Veliger, 15: 295- 299: DINAMANI, P., 1976, The morphology of the larval shell of Saccostrea glomerata (Gould, 1850) and a comparative study of the larval shell in the Ge- nus Crassostrea Sacco, 1897 (Ostreidae). Jour- nal of Molluscan Studies, 42: 95-107. DOLLFUS, G.-F. 4 P. DAUTZENBERG, 1920, Con- chyliologie du Miocene moyen du Bassin de la Loire, pt. 1: Pelecypodes (Suite et fin). Société Géologique de France, Mémoires, 27, Paleontol- ogie, 22(2-4): 379-500, pls. 34-51. DUJARDIN, F., 1835, Memoire sur les couches du sol en Touraine et description des coquilles de la craie et des faluns. Société Géologique de France, Mémoires, 2(2): 211-311, pls. 15-20. DURVE, V. S., 1986, On the ancestry and distribu- tion pathways of three species of Indian oysters. Indian Journal of Marine Sciences, 15: 56-58. FINCH, J., 1824, Geological essay on the Tertiary formations in America. American Journal of Sci- ence, 7: 31-43. FISCHER, P., 1886, Manuel de conchyliologie et de paléontologie conchyliologique ou histoire na- turelle des mollusques vivants et fossiles suive d'un appendice sur les brachiopodes par D. P. Oehlert, Fascicule 10. Paris, F. Savy. Pp. 897- 1008. FRENEIX, S., 1972, Les Mollusques bivalves cré- tacés du Bassin cótier de Tarfaya (Maroc mérid- ional). Notes et Memoires du Service Ge- ologiques (Maroc), 228: 49-255. FRENEIX, S., 1982, Disparité microstructurale du test entre Gryphaeostrea et Gyrostrea (Huitres du Crétacé-Miocene). Nouvelle classification des genres. Annales de Paléontologie, 68: 223- 240. FUNK, V. A. 8 D. В. BROOKS, 1990, Phylogenetic systematics as the basis of comparative biology. Smithsonian Contributions to Botany, 73: 1-45. GALTSOFF, P. S., 1964, The American oyster Crassostrea virginica Gmelin. U. S. Fish and Wildlife Service, Fishery Bulletin, 64: 1-480. GALTSOFF, Р. $. & В. O. SMITH, 1932, Stimulation of spawning and crossfertilization between American and Japanese oysters. Science, 76: 371-372. GMELIN, J. F., 1791, Systema naturae per regna tria naturae. Editio decimo tertia, aucta, refor- mata, 1(6): 3021-3909. Leipzig, G. E. Beer. GUILDING, L., 1828, Observations on the zoology 200 LAWRENCE of the Caribbean Islands. Zoological Journal (London), 3: 403-408, 527-544. GUNTER, G., 1950, The generic status of living oysters and the scientific name of the common American species. American Midland Naturalist, 43: 438-449. GUNTER, G., 1954, The problem in oyster taxon- omy. Systematic Zoology, 3: 134-137. HANLEY, 5. С. T., 1846, A description of new spe- cies of Ostreae, in the collection of H. Cuming, Esq. Zoological Society of London, Proceedings, 13: 105-107. HARRY, H. W., 1985, Synopsis of the supraspe- cific classification of living oysters (Bivalvia: Gry- phaeidae and Ostreidae). The Veliger, 28: 121- 158. HARRY, Н. W. 8 D. T. DOCKERY, Ш, 1983, Notes on the lower Oligocene oysters of Mississippi. Mississippi Geology, 4: 7-14. HENNIG, W., 1966, Phylogenetic systematics. Ur- bana, University of Illinois Press, 263 pp. HOERNES, M. & A. E. REUSS, 1870, Die Fossilen Mollusken Tertiaer-Beckens von Wein. Abhand- lungen der Kaiserlich-Koniglichen Geologischen Reichsanstalt, 4, pls. 68-82. HOPKINS, S. H., 1978, An oyster family tree: an- cestry of Crassostrea virginica. National Shell- fisheries Association, Proceedings, 68: 81-82. HOWE, H. V., 1937, Large oysters from the Gulf Coast Tertiary. Journal of Paleontology, 11: 355- 366. IMAI, T. & S. SAKAI, 1961, Study of breeding of Japanese oysters. Tohoku Journal of Agricultural Research, 12: 125-171. IREDALE, T., 1936, Australian molluscan notes, no. 2. Records of the Australian Museum, 19: 267- 340, pls. 20-24. JAWORSKI, E., 1913, Beiträge zur Stammesge- schichte der Austern. Zeitschrift fur Induktive Abstammungs-und Vererbunglehre, 9: 192-215. JIMENEZ, A. P., J. C. BRAGA & J. M. MARTIN, 1991, Oyster distribution in the Upper Tortonian of the Almanzora Corridor (Almeria, SE Spain). Geobios, 24: 725-734. KARL, S. A. & J. C. AVISE, 1992, Balancing selec- tion at allozyme loci in oysters: implications from nuclear RFLPs. Science, 256: 100-102. KENT, B. W., 1988, Making dead oysters talk. Ar- nold, Maryland Historical Trust, 107 pp. KORRINGA, P., 1952, Recent advances in oyster biology. Quarterly Review of Biology, 27: 266- 308, 339-365. LAMARCK, J. В.Р. DE M., 1801, Systeme des ап- imaux sans vertébres, ou table général des classes, des ordres et des genres de ces ani- maux. Paris, Deterville, 432 pp. LAMARCK, J. B. P. DE M., 1819, Histoire naturelle des animaux sans vertebres, présentant les char- acteres généraux et particuliers des ces ani- maux, leur distribution, leurs classes, leurs familles, leurs genres, et la citation des princi- pales езресе$ qui s'y rapportent, 6(1): 195-220, Paris. LAURAIN, M., 1980, Crassostrea gryphoides et C. gingensis (Schlotheim, 1813). Deux expressions morphologiques d'une méme espece (Miocene, Bivalvia). Geobios, 13: 21-43. LAWRENCE, D. R., 1968, Taphonomy and infor- mation losses in fossil communities. Geological Society of America, Bulletin, 79: 1315-1330. LAWRENCE, О. R., 1975, Paleoenvironmental set- ting of Crassostrea gigantissima (Finch) commu- nities, Coastal Plain of North Carolina. South- eastern Geology, 17: 55-66. LAWRENCE, D. R., in preparation, The oyster Crassostrea gigantissima (Finch) in the Cenozoic of the Western Hemisphere. LEA, |., 1833, Contributions to geology (Tertiary formation of Alabama; new tertiary fossil shells from Maryland and New Jersey; new genus of fossil shell from New Jersey; tufaceous lacustrine formation of Syracuse, Onondaga Co., N.Y.). Philadelphia, 227 pp., 6 pls. LITTLEWOOD, D. T. J. & S. K. DONOVAN, 1988, Variation of Recent and fossil Crassostrea in Ja- maica. Paleontology, 31: 1013-1028. LOGAN, W. М., 1899, Some additions to the Cre- taceous invertebrates of Kansas. Kansas Univer- sity Quarterly, (A) 8(2): 87-98, pls. 20-23. LOUGHLIN, G. F., E. W. BERRY & J. A. CUSH- MAN, 1921, Limestones and marls of North Carolina. North Carolina Geological and Eco- nomic Survey, Bulletin, 28: 1-211. MALCHUS, N., 1990, Revision der Kreide-Austern (Bivalvia: Pteriomorphia) Agyptens (Biostratigra- phie, Systematik). Berliner Geowissenschaftliche Abhandlungen, (A) 125: 231 pp., Berlin. MAYR, E., 1988, Toward a new philosophy of bi- ology: observations of an evolutionist. Cam- bridge, Massachusetts, Harvard University Press, 564 pp. MEEK, F. B., 1871, A preliminary list of fossils col- lected by Dr. Hayden in Colorado, New Mexico, and California, with brief descriptions of a few of the new species. American Philosophical Soci- ety, Proceedings, 11: 425-431. MEEK, Е. В. & F. V. HAYDEN, 1857, Descriptions of new species and genera of fossils, collected by Dr. F. V. Hayden in Nebraska Territory, under the direction of Lieut. G. K. Warren, U. S. Topo- graphical Engineer; with some remarks on the Tertiary and Cretaceous formations of the North- west, and the parallelism of the latter with those of other portions of the United States and Terri- tories. Academy of Natural Sciences of Philadel- phia, Proceedings, 9: 117-148. MENZEL, R. W., 1974, Portuguese and Japanese oysters are the same species. Journal of the Fisheries Research Board of Canada, 31: 453- 456. MIRKAMALOV, Kh. Kh., 1963, Classification of the exogyras. Moskovskoye Obshchestvo Ispytate- ley Prirody, Byulleten, Otdel Geologicheskiy, 38: 152-153. [in Russian] MIYAKE, M. & H. NODA, 1962, Vitamin B group in the extracts of Mollusca. II. On vitamin B6, inosi- GENUS CRASSOSTREA 201 tol, pantothenic acid, biotin, and niacin. Bulletin of the Japanese Society of Scientific Fisheries, 28: 597-601. MOORE, E. J., 1987, Tertiary marine pelecypods of California and Baja California: Plicatulidae to Os- treidae. United States Geological Survey, Pro- fessional Papers, 1228-C: 1-53, 34 pls. NELSON, T. C., 1938, The feeding mechanism of the oyster. |. On the pallium and the branchial chambers of Ostrea virginica, O. edulis and O. angulata, with comparisons with other species of the genus. Journal of Morphology, 63: 1-61. NEWBALL, S. & М. В. CARRIKER, 1983, System- atic relationship of the oysters Crassostrea rhizo- phorae and C. virginica: a comparative ultra- structural study of the valves. American Malacological Bulletin, 1: 35-42. NEWELL, N. D., 1965, Classification of the Bivalvia. American Museum Novitates, 2206: 1-25. NEWTON, R. B. 8 E. A. SMITH, 1912, On the sur- vival of a Miocene oyster in recent seas. Records of the Geological Society of India, 42: 1-15, 8 pls. NICOL, D., 1984, Critique on Stenzel's book on the Ostracea. The Nautilus, 98: 123-126. ORTON, J. H., 1928, The dominant species of Os- trea. Nature, 121: 320-321. PASCUAL, E., 1971, Morphología de la charnela larvaria de Crassostrea angulata (Lmk.) en difer- entes fases de su desarrollo. Investigación Pesqueras, 35: 549-563. PASCUAL, E., 1972, Estudio de las conchas lar- varias de Ostrea stentina Payr. y Ostrea edulis L. Investigación Pesqueras, 36: 297-310. PERKINS, B. F., 1969, Rudist morphology. In: R. C. MOORE, ed., Treatise on Invertebrate Paleontol- ogy, Part N(2), Mollusca 6 Bivalvia, Lawrence, Kansas, Geological Society of America and Uni- versity of Kansas Press, pp. N751-N764. PILSBRY, H. A. 8 A. P. BROWN, 1917, Oligocene fossils from the neighborhood of Cartagena, Co- lombia, with notes on some Haitian species. Academy of Natural Sciences of Philadelphia, Proceedings, 69: 32-41. RANSON, G., 1939, Le provinculum de la prodis- soconque de quelque ostréidés. Bulletin du Mu- séum National d'histoire Naturelle, Paris, (2)11: 318-332. RANSON, G., 1943, Note sur la classification des ostréidés. Société Géologique de France, Bulle- tin, (5)12: 161-164. RANSON, G., 1948, Prodissoconques et classifi- cation des ostréidés vivants. Bulletin du Musée Royal d'Histoire Naturelle de Belgique, 24(42): 1-12. RANSON, G., 1967, Les especes d’huitres vivants actuellement dans le monde, défines par leurs coquilles larvaires ou prodissoconques. Etude des collections de quelques-uns des grands mu- sées d'histoire naturelle. Revue des Travaux de l’Institut des Pêches Maritimes, 31: 127-199. ВЕЕВ, С. A. & J. С. AVISE, 1990, A genetic dis- continuity in a continuously distributed species: mitochondrial DNA in the American oyster, Cras- sostrea virginica. Genetics, 124: 397-406. RICHARDS, Н. С. 8 A. HARBISON, 1942, Miocene invertebrate fauna of New Jersey. Academy of Natural Sciences of Philadelphia, Proceedings, 94: 167-250, pls. 7-22. RUDWICK, М. J. S., 1961, The feeding mechanism of the Permian brachiopod Prorichthofenia. Palaeontology, 3: 450-471. RUTSCH, R. F., 1955, Die fazielle Bedeutung der Crassostreen (Ostreidae, Mollusca) im Helvétien der Umgebung von Bern. Eclogae Geologicae Helvetiae, 48: 453-464. SACCO, F., 1897, | molluschi dei terreni terziarii del Piemonte e della Liguria. Bollettino dei Musei di Zoologia e di Anatomia Comparata della R. Uni- versita, Torino, 12(298): 99-102. SCARLATO, O. A. 8 Ya. |. STAROBOGATOV, 1979a, The system of the suborder Mytileina (Bi- valvia). Pp. 22-25, in |. M. LIKHAREV, ed., Sixth Meeting on the Investigation of Molluscs. Mol- luscs. Main Results of their Study. Abstracts of Communications, Nauka, Leningrad. [in Rus- sian]. SCARLATO, O. А. & Ya. |. STAROBOGATOV, 1979b, General evolutionary patterns and the system of the Class Bivalvia. Zoological Institute of Leningrad, Trudy, 80: 5-38. [in Russian]. SCHLOTHEIM, E. F., VON, 1813, Beiträge zur Naturgeschichte der Versteinerungen in geog- nostischer Hinsicht. Taschenbuch Gesammte Mineralogie, C. C. LEONHARD, ed., 7: 3-134. SEILACHER, A., 1984, Constructional morphology of bivalves: evolutionary pathways in primary versus secondary soft-bottom dwellers. Palae- ontology, 27: 207-237. SIEWERT, W., 1972, Schalenbau und Stammes- geschichte von Austern. Stuttgarter Beitrage zur Naturkunde (B) 1: 1-57. SOHL, N. F. 8 E. G. KAUFFMAN, 1964, Giant Up- per Cretaceous oysters from the Gulf Coast and Caribbean. United States Geological Survey, Professional Papers, 483-H: 1-22, 5 pls. STENZEL, H. B., 1947, Nomenclatural synopsis of supraspecific groups of the family Ostreidae (Pelecypoda, Mollusca). Journal of Paleontology, 21: 165-185. STENZEL, H. B., 1959, Cretaceous oysters of southwestern North America. International Geo- logical Congress, 20th, Mexico, 1956, [Traba- jos], El Sistema Cretacico, 1: 15-37. STENZEL, H. B., 1971, Oysters. In: R. C. MOORE, ed., Treatise on Invertebrate Paleontology, Part N(3), Mollusca 6 Bivalvia, Lawrence, Kansas, Geological Society of America and University of Kansas Press, pp. N953-N1224. THIELE, J., 1934, Handbuch der Systematischen Weichtierkunde, Pt. 3: 779-1022 Jena, Gustav Fischer. THOMSON, J. M., 1954, The genera of oysters and the Australian species. Australian Journal of Ma- rine and Freshwater Research, 5: 132-168, pls. 1-11. 202 LAWRENCE THUNBERG, C. P., 1793, Tekning och beskrifning pá еп stor Ostronsort ifrán Japan. Kungliga Svenska Vetenskapsakademiens Handlingar, 14: 140-142. TORIGOE, K., 1981, Oysters in Japan. Journal of Science of Hiroshima University, (B) [1-Zoology] 29: 291-419. VEATCH, J. O. & L. W. STEPHENSON, 1911, Pre- liminary report on the geology of the Coastal Plain of Georgia. Georgia Geological Survey Bul- letin, 26: 1-466. VYALOV, O. S., 1936, Sur la classification des huîtres. Comptes Rendus (Doklady) de ГАсад- emie des Sciences de l'URSS, Nouvelle Serie, 4(13): 17-20. VYALOV, O. S., 1948, Printsiply klassifikatsii sem- eyistva Ostreidae. Lvovskoe Geologicheskoe Obshchestvo, Trudy, Paleontologicheskaya Seriya, 1: 30-40. WALLER, T. R., 1981, Functional morphology and development of veliger larvae of the European oyster, Ostrea edulis Linne. Smithsonian Contri- butions to Zoology, 328: 1-70. WARD, L. W., D. R. LAWRENCE & B. W. BLACK- WELDER, 1978, Stratigraphic revision of the middle Eocene, Oligocene, and lower Miocene- Atlantic Coastal Plain of North Carolina. United States Geological Survey, Bulletins, 1457-F: 1-23. WHITFIELD, R. P., 1894, Mollusca and Crustacea of the Miocene formations of New Jersey. United States Geological Survey, Monographs, 24: 1-195, 24 pls. WILKES, J., 1810, Encyclopaedia Londinensis; or, Universal dictionary of arts, sciences, and litera- ture. Vol. 5. London, J. Adlard. WILLIAMS, A. & D. J. ROWELL, 1965, Morphology. In: R. C. MOORE, ed., Treatise on Invertebrate Pa- leontology, Part H, Brachiopoda 1, Lawrence, Kansas, Geological Society of America and Uni- versity of Kansas Press, pp. H57-H138. WOODRING, W. P., 1982, Geology and paleontol- ogy of Canal Zone and adjoining parts of Pan- ama. Description of Tertiary mollusks (Pelecy- pods: Propeamussiidae to Cuspidariidae; Additions to families covered in P 306-E; Addi- tions to gastropods; Cephalopods). United States Geological Survey, Professional Papers, 306-F: 541-759, pls. 83-124. YONGE, C. M., 1960, Oysters. London, Collins, 209 pp. YONGE, C. M., 1962, On Etheria elliptica Lam. and the course of evolution, including assumption of monomyarianism, in the Family Etheriidae (Bi- valvia, Unionacea). Philosophical Transactions of the Royal Society of London, (B) 244: 423-458. Revised Ms. accepted 21 July 1994 MALACOLOGIA, 1995, 36(1-2): 203-208 LETTER TO THE EDITOR CLARIFICATION AND EVALUATION OF TILLIER’S (1989) STYLOMMATOPHORAN MONOGRAPH Kenneth С. Emberton' & Simon Tillier? When KE (Emberton, 1991) performed an independent phylogenetic analysis of 17 of the 111 subfamilies of stylommatophoran land snails previously analyzed by ST (Tillier, 1989), his results differed so greatly that he wrote a detailed critique of ST’s monograph (Emberton, unpublished), initiating an ex- change between KE and ST that accumu- lated to about 100 manuscript pages, which we condense into this letter. We are very grateful to Rúdiger Bieler for his useful com- ments on an earlier draft. Tillierrs (1989) phylogenetic characters, character states, and transformation series (= “morphoclines”) were not clearly defined, making his work irreproducable. Here we remedy the problem by defining and illustrat- ing his transformation series (Figs. 1, 2). 1. BM = buccal mass: 1 = spheroidal to ovoidal tending toward cylindrical; 2 = clearly cylindrical (Fig. 1: BM). 2. OC = esophageal crop: 1 = absent; 2 = separated from gastric crop by a distinct por- tion of the esophagus; 3 = separated from gastric crop by a simple constriction; 4 = as in 3 but extending forward to the nerve ring (Fig. We OG). 3. SC = gastric crop: 1 = cylindrical; 2 = median portion inflated; 3 = anterior region inflated; 2’ = funnelform, widening from esophagus to stomach; 0 = unscorable (e.g. semislugs), so eliminated (Fig. 1: SC). 4. PS = gastric pouch: 1 = joining the gas- tric crop without any constriction, distinctly wider than the gastric crop; 2 = joining the gastric crop without any constriction, slightly wider or no wider than the gastric crop; 2’ = separated from gastric crop by a constric- tion, distinctly wider than the gastric crop (Fig: 1 PSL 5. IL = intestine length (relative to the com- bined lengths of the gastric crop and stom- ach): 1 = intestinal loops reaching a level be- tween the distal limit of gastric pouch and the middle of gastric crop; 2 = intestine shorter, but intestinal loops distinct; 3 = intestinal loops reduced to an almost flat sigmoid; 2” = intestinal loops long, reaching proximally at least the level of the distal limit of the gastric pouch (Fig. 1: IL). 6. LR = ratio of kidney length to lung length: 1 = 0.45-0.7; 2’ = 0.7-1.0; 2 = 0.36- 0.45; 3 = 0.25-0.36; 4 = 0.0-0.25; with semi- slugs and slugs not scored at all (Fig. 1: LR). 7. UR = degree of closure of the ureter: 1 = no closed retrograde ureter; 2 = closed ureter reaching at most lung top; 3 = ureteric tube reaching a point between lung top and pneumostome; 4 = ureteric tube reach- ing the pneumostome (full sigmurethry) (Fig. 1: UR). 8. RR = kidney internal morphology: 1 = either two distinct regions (the distal one usu- ally lacking lamellae) or three distinct regions (the median one either lacking lamellae or with lamellae different in appearance from those in the proximal region); 2 = kidney ho- mogeneous in internal morphology, with lamellae reaching the distal region and the level of the kidney pore (Fig. 1: RR). 9. CC = length of cerebral commissure: 1 = greater than 1.1 x right cerebral ganglion width; 2 = between 1.1 and 0.9 x right cere- bral ganglion width; 3 = less than 0.9 х right cerebral ganglion width (Fig. 2: CC). 10. CPD = length of the right cerebro-pedal "Department of Malacology, Academy of Natural Sciences, 1900 Benjamin Franklin Parkway, Philadephia, Pennsylvania 19103-1195, U.S.A ?Laboratoire de Biologie des Invertébrés Marins et de Malacologie, Museum National d'Histoire Naturelle, 55 Rue Buffon, 75005 Paris, FRANCE 204 EMBERTON & TILLIER BM OC SC FIG. 1. Diagrammatic illustrations of Tillier’s (1989: appendix E) cladistic characters and their character- state transformations. See text for names and definitions. STYLOMMATOPHORAN PHYLOGENETICS 205 СС СРО CPR Er iv A = wy Paci 3h FIG. 2. Diagrammatic illustrations (cont.) of Tillier’s (1989: appendix E) cladistic characters and their char- acter-state transformations. See text for names and definitions. connective: 1 = longer than twice the width of 3 = shorter than right cerebral ganglion width the right cerebral ganglion; 2 = between one (Fig. 2: CPD). and two times right cerebral ganglion width; 11. CPR = ratio between the lengths of the 206 EMBERTON & TILLIER cerebro-pedal connectives (left/right): 1 = less than 0.9; 2 = from 0.9 to 1.1; 3 = from 1.1 to 1.5: 4 = Ноа sito 25 (56-2: CPR). 12. PLD = position of the right pleural gan- glion: 1 = closer to the pedal ganglion than to the cerebral ganglion (hypoathroid); 2 = closer to the cerebral ganglion than to the pedal ganglion (epiathroid); O = unscorable, so eliminated (Fig. 2: PLD). 13. PLG = position of the right pleural gan- glion: 1 = closer to the pedal ganglion than to the cerebral ganglion (hypoathroid); 2 = closer to the cerebral ganglion than to the pedal ganglion (epiathroid); O = unscorable, so eliminated (Fig. 2). 14. VG = position of the center of mass of the visceral ganglion relative to the median plane of the pedal ganglia: 1 = on the right side; 2 = in the middle; 3 = on the left side (Fig. 2: VG). 15. PAD = right parietal and pleural gan- glia: 1 = separate; 2 = in contact or fused (Fig. 2: PAD). 16. PAG = position of the left parietal gan- glion: 1 = in contact with left pleural, or closer to the left pleural than to the visceral, and separated from both by a distinct connective; 2 = closer to the visceral than to the left pleu- ral, and separated from both by a distinct connective; 3 = in contact with the visceral ganglion alone, and separated from the left pleural by a distinct connective; 4 = in con- tact or fused with both left pleural and vis- ceral ganglia (Fig. 2: PAG). 17. FG = fusion of the visceral ganglion: 1 = none; 2 = with the right parietal ganglion; 3 = with both parietal ganglia; 2’ = with the left parietal ganglion (Fig. 2: FG). ST stresses that these character states were not always the same as those used in his “factor analyses” (Tillier, 1989: text-figs. 5-7, 10-18). Character states were first ten- tatively defined for correspondence analysis (see below), evaluated, and then re-defined and re-scored for phylogenetics. For exam- ple, when a character state such as gastric crop shape (SC) in semislugs could not be placed within a transition series, it was rede- fined as state O and eliminated from phylo- genetic analysis. ST's “factor analysis” 15 not the statistical method known to most American workers as factor analysis, but the method of Benzécri (1973: “analyse factorielle des correspon- dances””) that is better translated as “corre- spondence analysis.” Correspondence anal- ysis, unlike factor analysis, requires no mul- tivariate-normal assumption (Fénelon, 1981; Jambu 4 Lebeaux, 1979). KE disagrees with most of ST’s character- state choices because they (a) oversimplify a complex character into a single measure- ment, ratio, or quality (BM, LR, RR, CC, CPD, СРВ, PLD, PLG); (b) apply arbitrary cutpoints to continuous variation (BM, IL, LR, UR, CC, CPD, CPR, PLD, PLG); and/or (c) include possible artifacts of fixation, preservation, and dissection (food bolus in OC, SC, PS; stretching in CC, CPD, CPR, VG [Emberton, 1989: fig. 4)). KE and ST agree that ganglionic fusion (FG) is an important but difficult-to-score charac- ter. ST rechecked his dissections of An- guispira, Sagda and Thysanophora and found that he had scored them incorrectly for gan- glionic fusion: correct scorings are as in Em- berton (1991). KE also scored Acavus, Brady- baena, and Polygyrella differently from ST, who did not recheck these genera. KE dis- agrees with ST’s opinion that ganglionic con- tact (short of fusion) is a reliable character. KE advocates use of structurally complex characters divisible into discrete, qualitative states. ST rebuts that this is impractical for soft-part molluscan anatomy, and that all the transformations KE (Emberton, 1991) used, with the possible exceptions of his charac- ters 4 and 5, would prove to be continuous if more taxa were included. KE doubts that his characters 2-5 and 8-18 will prove continu- ous, but agrees that 1, 6, and 7 may; KE counters that ST made many of his (ST’s) characters artificially continuous by reducing them to measurements and ratios. Users of ST’s anatomical figures are cau- tioned that sinistral species (with all organs right-to-left reversed) are not indicated as such in the captions (e.g. Tillier, 1989: figs. 111, 484, 512). As discussed more recently by Tillier 8 Ponder (1992), the Otinidae were used by ST as the stylommatophoran sister group be- cause they share with the Stylommatophora (a) monotremy, (b) kidney not surrounded by lung, and (c) five ganglia in the ventral chain. This position is not accepted by Nordsieck (1992), who proposed the Ellobiidae as the sister-group of the Stylommatophora. ST’s (Tillier, 1989) “phylogenetic analysis” was performed in 1984 (Tillier, 1985), using the then unpublished algorithm of Delattre (1988), which does not permit reversals, STYLOMMATOPHORAN PHYLOGENETICS 207 which may not find the most parsimonious tree(s), and whicn KE believes 15 a phenetic rather than cladistic method. Obviously, the more recent availability of more efficient cla- distic algorithms has made this рай of ST’s work obsolete; reanalysis using Hennig86 (Farris, 1988) yields a very different topology (Richard Lamb, personal communication). KE contends that reanalysing ST’s tabulated data is unproductive because of numerous defects in the conception and scoring of characters. The Orthurethra/non-Orthurethra split pro- posed by Pilsbry (1900) remains the only as- pect of stylommatophoran phylogeny sup- ported by ST’s morphological data. The division of the order Stylommatophora pro- posed Бу ST (Tillier, 1989) into two subor- ders, Brachynephra and Dolichonephra, was submerged by Nordsieck (1992) correctly in KE’s opinion, although ST believes that ad- ditional data may support these taxa. The multiplication of phylogenetic hypoth- eses in the past 15 years, as summarized by Bieler (1993), shows in our opinion that: (a) monophyly remains undemonstrated for most families and suprafamilial taxa; (b) there are high levels of homoplasy in all known an- atomical characters; (с) any worthwhile fur- ther morphological study should include nu- merous taxa and characters, should include careful character analyses, and should clearly define and illustrate all character states and suggested transformations; (а) histological sections will be required to re- solve some potentially important characters, such as fusion among ventral-chain ganglia, kidney internal morphology, and the fertiliza- tion pouch-seminal receptacle complex. Other anatomical characters worth compar- ing may be the ureteric interramus and nearby structures, the position of the proxi- mal hermaphroditic duct, the fusion of the free retractor muscles, and genital accessory organs. Molecular characters may prove use- ful, as shown by Emberton et al. (1990) and Tillier et al. (1992, and in press). Thus despite over a century of work, only _ two suprafamilial clades of the Stylommato- phora, Orthurethra and Sigmurethra, may be resolved, although recent molecular studies by ST cause him to question even the mono- phyly of the Sigmurethra. Resolution of sty- lommatophoran higher phylogeny remains a tremendous but hopefully not impossible challenge. LITERATURE CITED BENZECRI, J. P., 1973, L'analyse des données. Dunod, Paris: Volume 1, La taxinomie: 612 pp.; Volume 2, L'analyse des correspondances: 619 pp. BIELER, R., 1993, Gastropod phylogeny and sys- tematics. Annual Review of Ecology and Sys- tematics, 23: 311-338. DELATTRE, P., 1988, Sur la recherche des filia- tions en phylogénese. Revue Internationale de Systémique, 2: 479-504. EMBERTON, K. C., 1989, Retraction/extension and measurement error in a land snail: effects on systematic characters. Malacologia, 31: 157-173. EMBERTON, K. C., 1991, Polygyrid relations: a phylogenetic analysis of 17 subfamilies of land snails (Mollusca: Gastropoda: Stylommato- phora). Zoological Journal of the Linnean Soci- ety, 103: 207-224. EMBERTON, K. C., G. S. KUNCIO, G. M. DAVIS, S. М. PHILLIPS, К. М. MONDEREWICZ & Y. H. GUO, 1990, Comparison of recent classifica- tions of stylommatophoran land-snail families, and evaluation of large-ribosomal-RNA se- quencing for their phylogenetics. Malacologia, 31: 327-352. FARRIS, J. S., 1988, Hennig86, Version 1.5. James S. Farris, Port Jefferson Station, New York, NY 11776. FENELON, J. P., 1981, Qu'est-ce que l’analyse des donnees? Lefonen, Paris: 311 pp. JAMBU, М. 8 М. O. LEBEAUX, 1979, Classification automatique pour l'analyse des données. II. Logiciels. Dunod, Paris: 400 pp. NORDSIECK, H., 1992, Phylogeny and system of the Pulmonata. Archiv für Molluskenkunde, “1990”, 121: 31-52. PILSBRY, H. A., 1900, On the zoological position of Partula and Achatinella. Proceedings of the Academy of Natural Sciences of Philadelphia, 3: 561-567. TILLIER, S., 1985, Morphologie comparée, phylog- énie et classification des gastéropodes pul- monés stylommatophores (Mollusca). These de Doctorat d’Etat, Museum National d’Histoire Na- turelle and Université Paris 6: 236 pp. TILLIER, S., 1989, Comparative morphology, phy- logeny and classification of land snails and slugs (Gastropoda: Pulmonata: Stylommatophora). Malacologia, 30: 1-303. TILLIER, S., М. MASSELOT, H. PHILIPPE 4 A. TILLIER, 1992, Phylogenie moleculaire des Gas- tropoda (Mollusca) fondée sur le sequencage partiel de ГААМ ribosomique 28$. Comptes- Rendus de l’Académie des Sciences, Paris, série Ш, 314: 79-85. TILLIER, S., MASSELOT, M., GUERDOUX, J. & A. TILLIER, in press, Monophyly of major gastro- pod taxa tested from partial 28$ rRNA se- 208 EMBERTON & TILLIER quences, with emphasis on Euthyneura and hot- vent limpets Peltospiroidea. The Nautilus. ; The editor-in-chief of Malacologia welcomes let- TILLIER, $. 8 W.F. PONDER, 1992, New species of ters that comment on vital issues of general im- Smeagol from Australia and New Zealand, with a portance to the field of Malacology, or that com- discussion of the affinities of the genus (Gas- ment on the content of the journal. Publication is tropoda: Pulmonata). Journal of Molluscan Stud- dependent on discretion, space available and, in jes, 58: 135-155. some cases, review. Address letters to: Letter to the Editor, Malacologia, care of the Department of Malacology, Academy of Natural Sciences, Revised Ms. accepted 1 January 1994 19th and the Parkway, Philadelphia, PA 19103. MALACOLOGIA, 1995, 36(1-2): 209-215 INDEX Page numbers in /ta/ics indicate figures of taxa. No new taxa appear in this number of Malacologia. Acavidae 50, 64 Acavoidea 64 Acavus 206 Achatina 68 Achatinella 59 Achatinellidae 66, 159 Achatinellinae 155 Achatinelloidea 66 Achatinida 48, 63, 65 Achatinoidea 63 Acroptychia 69 acuta, Physa 39, 84, 86, 87 Acutostrea 187, 188, 191, 194, 197 Adula (Botula) 1 falcata 5 Aegopis 49 Aequipecten 15 afficta, Helix 121,124 afficta, Helix (Helicigona) alabamiensis, Ostrea 193 albicans, Littoraria 93 albolabris, Neohelix 65 alexandrina, Biomphalaria 84 Allodiscus granum 66 urquharti 66 Allogona profunda 65 Amastridae 159 Amepelita xystera 69 amnicum, Pisidium 30-32, 35, 37-39 Ampelita 54, 64, 67 Ampelita (Ampelita) lamarei 70, 72, 73, 74 Ampelita (Eurystyla) 72 juli 70, 72-74 soulaiana 70, 72-74 Ampelita (Xystera) 71, 72 fulgurata 70, 72, 73, 75 xystera 70, 70, 72-74 Ampelita fulgurata 71, 72, 74, 75 joli 69,71, 72, 74; 75 lamarei 71, 72 soulaiana 69, 71, 72, 75 xystera» 71, 72.14, 75 amphibulima, Helicophanta 70, 71-74 Amyqdalum 10 `’апсеуапа, Leptachatina 161-164 anceyana, Leptachatina (Angulidens) Ancylus 39 Ancylus fluviatilis 30, 31, 34, 36, 3-40, 86, 145, 152 andrewsae, Mesodon 172 andrewsae, Mesomphix 181 Anguispira 206 Anguispira mordax 65 112, 776, 124 159 angulata, Crassostrea 196 Angustostrea 197 apicina, Xerotricha 134 Appalachina sayana 65 appressa, Patera 65 Archaeogastropoda 64, 66 Archaeopulmonata 64 Arcuatula 5, 12 Argopecten irradians 15, 23, 24 ariel, Phrixgnathus 66 Ariophantinae 64 Ashfordia 134 Assimineidae 63 Athoracophorus bitentaculatus 66 barbata, Helix 115 barbata, Helix (Helicigona) 115 Basommatophora 133 beata, Canariella hispidula var. 119, 120, 129 beata, Caracollina 118 beata, Helicodonta (Caracollina) 118 beata, Helix (Caracollina) 118 beata, Helix (Gonostoma) 112, 116, 118, 119 Belgrandia 39 Belgrandiella 39 bertheloti, Canariella hispidula 118 bertheloti, Canariella hispidula var. 120,728 bertheloti, Helix 112, 116, 117-119 bertheloti, Helix (Gonostoma) 117 berthelotii, Helix 117 Biomphalaria alexandrina 84 Biomphalaria glabrata 84, 86, 87 bitentaculatus, Athoracophorus 66 Bithinella 149 Bithynia 149 graeca 139-146 tentaculata 29-34, 38, 39, 87, 145, 147, 148, 150-152 blandianum, Punctum 65 Bosostrea 187, 188, 191, 193, 194, 198 Botula 1,12 Botula cinnamomea 5,8 Boucardicus 63 Brachidontes 1 darwinianus 2,5,10 solisianus 2,5,10 Brachynephra 207 Bradybaena 206 similaris 59 Bradybaenidae 133 buccinella, Caviella 66 Buliminidae 63 Buliminoidea 63 Bulinus truncatus 84 Bythinella dunkeri 147-152 cahobasensis, Ostrea 190 118 209 210 INDEX Camaenidae 59 Canariella 111-137 discobola 121 discobolus 121, 122, 123, 129, 131 eutropis 122, 123, 125, 127-131, 130 everia 118 fortunata 118 gomerae 122, 123-124, 125, 129, ven hispidula 115-122, 122, 131 hispidula bertheloti 118 var. beata 119, 120, 129 var. bertheloti 119, 120, 128 var. fortunata 119, 120, 125, 129 var. hispidula 114, 119-120, 128 var. lanosa 119.120; 1252729 var. subhispidula 119, 120, 129 leprosa 118; 123; 122, 123, 125, 126-127, 730, 131 multigranosa 115 р/апапа 116, 122, 123, 124, 130, 131, 134 pthonera 127 capsella, Рагауйгеа 65 Caracollina beata 118 everia 118 gomerae 123 planaria 124 Cardium 23 Carocolla hispidula 112, 113, 115, 776, 118 planaria 112, 116, 124 Carychium clappi 64 nannodes 64 Caseolus 134 catascopium, Lymnaea 87 Caucasocressa 134 Caviella buccinella 66 roseveari 66 celinda, Therasiella 66 Сераеа 97, 98 nemoralis 98, 180, 182 Cerastoderma edule 23 Charopa chrysaugeia 66 fuscosa 66 pilsbryi 66 pseudanguicula 66 Charopidae 49, 64, 66 charruana, Mytella 1,4,5, 10-12 chiron, Flammulina 66 Choromytilus 4 chrysaugeia, Charopa 66 ciliata, Ciliella 127 Ciliella 132,133 ciliata 127 Ciliellidae 132 Ciliellinae 132 Ciliellopsis 133, 134 cinnamomea, Botula 5, 8 Cionella morseana 64 clappi, Carychium 64 clappi, Vertigo 65 Clavator 67 moreleti 69, 70, 71, 73-75 Clithon oualaniensis 97-109, 100 Cochlicopoidea 64 collisella, Ventridens 65 columbiensis, Crassostrea 196 Columella simplex 65 concavum, Haplotrema 65 conella, Phrixgnathus 66 contracta, Gastrocopta 65 cookiana, Geminopora 66 Corbicula fluminea 30-32, 35, 37-40 coresia, Delos 66 corneus, Planorbarius 79-89 coronadoi, Neohoratia 39 corticaria, Gastrocopta 65 Crasostrea 197 Crassostrea 185-202 Crassostrea (Euostrea) 197 Crassostrea angulata 196 columbiensis 196 gigantissima 188, 189, 190-194 gigas 196 gryphoides 189, 196 rhizophorae 197 soleniscus 188 virginica 24, 188-189, 196 Crassostreinae 185-202 Crassotrea 197 Cretigena 134 cuccullata, Saccostrea 196 cumberlandiana, Glyphyalinia 65 cupreus, Mesomphix 65 Cupulella 49 Cussetostrea 187, 188, 193, 194, 198 Cyathopoma 63 Cyclophoridae 48, 49, 63 Cyclophoroidea 63, 66 Cyrnotheba 134 cytora, Cytora 66 Cytora cytora 66 Cytora torquilla 66 darwinianus, Brachidontes 2,5, 10 Delos coresia 66 jeffreysiana 66 demissus, Modiolus 10 denotata, Xolotrema 65 Dioeciostraea 197 Dioeciostrea 197 Discidae 49, 65 discobola, Canariella 121 discobolus, Canariella 121, 122, 123, 1297 131 discobolus, Helicodonta (Caracollina) 121 discobolus, Helix 112,121 discobolus, Helix (Anchistoma) 121 discobolus, Helix (Gonostoma) 121 Discus nigrimontanus 65 patulus 65 Dolichonephra 207 Dreissena polymorpha 15-27 dunkeri, Bythinella 147-152 INDEX duryi, Helisoma 84-86 Edentulina 54, 63 edule, Cerastoderma 23 edulis, Mytilus 2,4,5,7-11, 22-24, 181 edvardsi, Stenotrema 65 elaioides, Phrixgnathus 66 Elasmognatha 65 Elisolimax 133 Ellobiidae 64, 206 Ellobicidea 64 Endodontidae Enidae 63 Enneinae 63 erigone, Phrixgnathus 66 eta, Mocella 66 Euconulidae 49, 65 Euconulus fulvus 65 gaetanoi 161-163 Euconulus (Nesoconulus) gaetanoi 159, 160 Euglandina 59 Eurystala 69 eutropis, Canariella 122, 123, 125, 127- 131,730 eutropis, Helis 112 eutropis, Helix 122,127 everia, Canariella 118 everia, Caracollina 118 everia, Helicodonta (Caracollina) 118 everia, Helicodonta (Caracollina) hispidula 118 everia, Helix 112, 116, 118, 120 everia, Helix (Anchistoma) 118 Exogyrinae 186 falcata, Adula (Botula) 5 Fectola infecta 66 mira 66 unidentata 66 Ferrissia 39 rivularis 145 filosa, Littoraria 92 Flammulina chiron 66 perdita 66 fluminea, Corbicula 30-32, 35, 37-40 fluviatilis, Ancylus 30, 31, 34, 36, 3-40, 86, 145, 152 fontinalis, Physa 86 fortunata, Canariella 118 155 fortunata, Canariella hispidula var. 119, 120: 725; 129 fortunata, Gonostoma 118 fortunata, Helicodonta (Caracollina) 118 fortunata, Helix 112, 114, 116, 117, 119 fortunata, Helix (Anchistoma) 118 fortunei, Limnoperna 5 fulgurata, Ampelita 71, 72, 74, 75 fulgurata, Ampelita (Xystera) 70, 72, 73, 75 fulvus, Euconulus 65 fuscosa, Charopa 66 fuscus, Laevapex 145 gaetanoi, Euconulus 161-163 211 gaetanoi, Euconulus (Nesoconulus) 159, 160 Gastrocopta corticaria 65 pentodon 65 Gastrodonta interna 65 Gastrodontinae 65 Gasulliella 134 Geminopora cookiana 66 georgianus, Viviparus 87 Georissa purchasi 66 gigantissima, Crassostrea 188, 189, 190- 194 gigas, Crassostrea 196 giveni, Phenacohelix 66 glabrata, Biomphalaria 84, 86, 87 Glyphyalinia cumberlandiana 65 rimula 65 gomerae, Canariella 122, 123-124, 125, 129, 131 gomerae, Caracollina 123 gomerae, Helicodonta (Caracollina) 123 gomerae, Helix (Gonostoma) 112, 122, 123 Gonaxis 133 Gonostoma fortunata 118 hispidula 118 gouldi, Vertigo 65 gracilis, Lamellidea 159 gracilis, Lithophaga 5, 8 graeca, Bithynia 139-146 granosissimus, Modiolus demissus 9 granum, Allodiscus 66 Grassostrea 197 greenwoodi, Rhytida 66 Gryphaea 197 Gryphaeidae 185, 186 Gryphaeinae 186 Gryphaeostreinae 186 gryphoides, Crassostrea 189, 196 Gulella 63 Guppya sterkii 65 guyanensis, Mytella 1,2 Gyraulus 39 Gyrostrea 187, 188, 192, 198 Hainesia 54, 63 Halolimnohelcinae 132 Haplohelix 132, 134 Haplotrema concavum 65 Haplotrematidae 65 hawaliensis, Nesovitrea 159, 162-164 hectori, Huanodon 66 Helicarionidae 159 Helicida 48, 64, 65 Helicinoidea 64 Helicodonta 134 planaria 124 salteri 112, 116, 118, 119 Helicodonta (Caracollina) beata 118 discobolus 121 everia 118 fortunata 118 contracta 65 212 gomerae 123 Helicodontidae 130, 132 Helicodontoidea 130 Helicoidea 130, 132 Helicophanta 54, 64, 67 amphibulima 70, 71-74 Helis eutropis 112 Helisoma duryi 84-86 trivolvis 87 Helix afficta 121,124 afficta planaria 124 barbata 115 bertheloti 112, 116, 117-119 berthelotii 117 discobolus 112,121 eutropis 122,127 everia 112, 116, 118, 120 Tortunatar 12: 114, Пл» та hispidula 115 lanosa 127 lens 115 leprosa 112,126, 127 planaria 124 pthonera 112 Helix (Anchistoma) discobolus 121 everia 118 fortunata 118 Helix (Caracollina) beata 118 Helix (Ciliella) lanosa 112, 117, 119, 122 leprosa 126 Helix (Gonostoma) beata 112, 116, 118, 119 bertheloti 117 discobolus 121 дотегае 112, 122, 123 hispidula 117 рагу! 112 Helix (Helicigona) а ста 112, 116, 124 barbata 115 lens 115 Helix (Hispidella) lanosa 117 leprosa 126 Helix (Macularia) plutonia 112 Helix (Ochthephyla) multigranosa 112 Helixarionidae 49, 64 Helixarionoidea 64, 65 Hendersonia occulta 64 hispidula, Canariella 115-122, 122, 131 hispidula, Canariella hispidula var. 114, 119-120, 728 hispidula, Carocolla 112, 113, 115, 116, 118 hispidula, Gonostoma 118 hispidula, Helix 115 hispidula, Helix (Gonostoma) hochsteteri, Liarea 66 Horatia 39 Horatia sturmi 30, 31, 33, 36, 38-40 Huanodon hectori 66 pseudoleiodon 66 Hydrocenidae 66 Hydrocenoidea 66 in INDEX Hydromiidae 134 Hygromia (Ciliella) lanosa 118 leprosa 126 Hygromiidae 111-137 Hygromiinae 134 Hygromioidea 132 ide, Suteria 66 imitatrix, Leptachatina 160 Indostrea 187, 188, 191, 193, 194, 198 infecta, Fectola 66 Inflectarius inflectus 65 inflectus, Inflectarius 65 inomatus, Mesomphix 65 intermedia, Littoraria 91-95 interna, Gastrodonta 65 irradians, Argopecten 15, 23, 24 Ischadium recurvum 9 jeffreysiana, Delos 66 jenkinsi, Potamopyrgus 30-35, 38-40 Juli, Ampelita 69, 71, 72, 74, 75 julii, Ampelita (Eurystala) 70, 72-74 Kalidos 64, 69 Kaliella 54, 64 kivi, Serpho 66 konaensis, Leptachatina 160 konanensis, Succinea 159, 162, 163 Konbostrea 187, 188, 192-194, 198 Laevapex fuscus 145 lamarei, Ampelita 71, 72 lamarei, Ampelita (Ampelita) 74 Lamellidea 159, 161-163, 166 gracilis 159 novoseelandica 66 oblonga 159 peponum 159 lanosa, Canariella hispidula var. 12051255129 lanosa, Helix 127 lanosa, Helix (Ciliella) 122 lanosa, Helix (Hispidella) lanosa, Hygromia (Ciliella) Laoma leimonias 66 тапае 66 marina 66 lapidaria, Pomatiopsis 64 lateumbilicata, Paralaoma 66 latissimus, Vitrizonites 65 leimonias, Laoma 66 lens, Helix 115 lens, Helix (Helicigona) lens, Lindholmiola 115 lepida, Leptachatina 160, 162, 163 lepida, Leptachatina (Leptachatina) 159 leprosa, Canariella 118, 123, 122, 123, 125120127 13013 leprosa, Helix 112,126, 127 leprosa, Helix (Ciliella) 126 leprosa, Helix (Hispidella) 126 leprosa, Hygromia (Ciliella) 126 70,22% 13, 119 112, 11223 112 118 MS INDEX Leptachatina 159, 160, 164 anceyana 161-164 imitatrix 160 konaensis 160 lepida 160, 162, 163 Leptachatina (Angulidens) anceyana 159 Leptachatina (Leptachatina) lepida 159 Liarea hochsteteri 66 Liareidae 49, 66 Liguus 59 Limnoperna 1,11 fortunei 5 Lindholmiola lens 115 Liostreinae 185, 188 Lithophaga 12 gracilis 5,8 nasuta 5 Littoraria 97 albicans 93 filosa 92 intermedia 91-95 luteola 93 pallescens 92 philippiana 92 scabra 93 strigata 94 Littorina 97,98 Littorinoidea 63 Lophinae 185, 186 luisi, Pseudoamnicola 30, 31, 33, 36, 38, 39 /uteola, Littoraria 93 lymaniana, Pronesopupa 160 Lymnaea catascopium 87 obrussa 84 palustris 145 peregra 29-31, 34, 36, 38-40, 84 stagnalis 23, 29, 40, 84, 87 truncatula 39, 84 Macrochlamydinae 64 maculata, Mocella 66 major, Neohelix 49 Malagarion 64 mariae, Laoma 66 marina, Laoma 66 Melanopsis 30-33, 39, 40 meniscus, Striatura 160, 162, 163 meniscus, Striatura (Pseudohyalina) 160 Mercuria 39 meridonalis, Striatura 65 Mesodon andrewsae 172 normalis 171-184 zaletus 65 Mesodontini 65 Mesogastropoda 63, 64, 66 Mesomphix andrewsae 181 cupreus 65 inomatus 65 perlaevis 65 subplanus 181 Metafruticicola 134 159, 213 metcalfei, Modiolus 4, 5 Microcystinae 64 Microcystis 64 mira, Fectola 66 Mocella eta 66 maculata 66 modiolus, Modiolus 5, 7 Modiolus 1,12 demissus 10 demissus granosissimus 9 metcalfei 4, 5 modiolus 5, 7 squamosus 9 striatulus 5 undulatus 5,8 moellendorffi, Phrixgnathus 66 Monacha 134 Montserratina 133 mordax, Anguispira 65 moreleti, Clavator 69, 70, 71, 73-75 morseana, Cionella 64 Mulinia 181 multidentata, Paravitrea 65 multigranosa, Canariella 115 multigranosa, Helix (Ochthephyla) Musculista 1, 10 senhausia 5, 11 Musculus 1,10 Mytella charruana 1, 4, 5, 10-12 guyanensis 1,2 speciosa 1 strigata 1-14; 3, 6, 7 tumbezensis 1 Mytilus 1,11,12 edulis 2,4,5, 7-11, 22-24, 181 nannodes, Carychium 64 nasuta, Lithophaga 5 nemoralis, Cepaea 98, 180, 182 Neohelix albolabris 65 major 49 Neohoratia 39 coronadoi 39 schuelei 39 neozelandica, Therasiella 66 Nesopupa subcentralis 162, 163 Nesopupa (Infranesopupa) subcentralis 159 Nesovitrea hawaliensis 159, 162-164 nigrimontanus, Discus 65 normalis, Mesodon 171-184 novoseelandica, Lamellidea 66 Nucella 97 oblonga, Lamellidea 159 obrussa, Lymnaea 84 occulta, Hendersonia 64 Oestophora 132 Omphalotropis 63 Ortheurethra 64, 66, 206 orycta, Pronesopupa 160 Ostrea 196 Ostrea alabamiensis 193 cahobasensis 190 112 214 Ostreidae 185, 195 Ostreinae 185, 186 Otinidae 206 oualaniensis, Clithon 97-109, 100 ovalis, Succinea 65 Palaeolophidae 185, 186 Palaeolophinae 186 pallescens, Littoraria 92 palustris, Lymnaea 145 Paralaoma lateumbilicata 66 serratocostata 66 Paravitrea capsella 65 multidentata 65 subtilis 65 parryi, Helix (Gonostoma) Paryphanta 59 Patera appressa 65 patulus, Discus 65 pentodon, Gastrocopta 65 peponum, Lamellidea 159 perdita, Flammulina 66 peregra, Lymnaea 29-31, 34, 36, 38-40, 84 12 peregra, Radix 147-152 perlaevis, Mesomphix 65 perna, Perna 11 Perna 12 perna 11 viridis 5,8, 11 Phenacohelix giveni 66 piula 66 philippiana, Littoraria 92 Philonesia 159, 160, 162, 163 Phrixgnathus ariel 66 conella 66 elaioides 66 erigone 66 moellendorffi 66 poecilosticta 66 Physa 39 acuta 39, 84, 86, 87 fontinalis 86 pilsbryi, Charopa 66 Pilula 64 piscinalis, Valvata 30-34, 38, 39 Pisidium amnicum 30-32, 35, 37-39 piula, Phenacohelix 66 planaria, Canariella 116, 122, 123, 124, 130: 131, 134. planaria, Caracollina 124 planaria, Carocolla 112, 116, 124 planaria, Helicodonta 124 planaria, Helix 124 planaria, Helix afficta 124 Planorbarius 39 corneus 79-89 planorbis, Planorbis 79-89 Planorbis 29 planorbis 79-89 plutonia, Helix (Macularia) 112 poecilosticta, Phrixgnathus 66 Poecilozonites 49 INDEX Polygyrella 206 Polygyridae 49, 65 Polygyrinae 65 Polygyroidea 65 polymorpha, Dreissena 15-27 Pomatiasidae 63 Pomatiopsis lapidaria 64 Potamopyrgus 39 jenkinsi 30-35, 38-40 profunda, Allogona 65 Pronesopupa 159-163 lymaniana 160 orycta 160 sericata 160 Pronesopupa (Sericipupa) Prosobranchia 63, 64, 66 pseudanguicula, Charopa 66 Pseudoamnicola 39 juisi 30731 33, 36, 38:39 pseudoleiodon, Huanodon 66 Pseudoperna 187,188, 191, 194, 197 pthonera, Canariella 127 pthonera, Helix 112 pugetensis, Striatura 160 Pulmonata 63, 64, 66 Punctidae 49, 65, 66 Punctoidea 64-66 Punctum blandianum 65 Pupillidae 159 Pupilloidea 65 purchasi, Georissa 66 Pycnodonteinae 186 Rachis 63 Radix peregra 147-152 recurvum, Ischadium 9 rhizophorae, Crassostrea 197 Rhytida greenwoodi 66 Rhytididae 66 Rhytidoidea 65, 66 rimula, Glyphyalinia 65 Rissoidea 63, 64 rivularis, Ferrissia 145 roseveari, Caviella 66 Roybellia 49 saccellus, Saccostrea 196 Saccostrea 187, 190, 194-197 cuccullata 196 saccellus 196 Sagda 206 salteri, Helicodonta 112, 116, 118, 119 Sanostrea 198 Saxostrea 197 sayana, Appalachina 65 scabra, Littoraria 93 Schileykiella 132-134 schuelei, Neohoratia 39 senhausia, Musculista 5, 11 Septifer 11 sericata, Pronesopupa 160 Serpho kivi 66 serrata, Therasiella 66 Sesarinae 64 160 INDEX Sigmurethra 63, 65 similaris, Bradybaena 59 simplex, Columella 65 Sitala 64 Soleniscostrea 187,188, 194, 198 soleniscus, Crassostrea 188 solisianus, Brachidontes 2,5, 10 Somalidacna 197 soulaiana, Ampelita 69, 71, 72, 75 soulaiana, Ampelita (Eurystala) 70, 72-74 speciosa, Mytella 1 squamosus, Modiolus 9 stagnalis, [утпаеа 23, 29, 40, 84, 87 stenotrema, Stenotrema 65 Stenotrema edvardsı 65 stenotrema 65 sterkii, Guppya 65 Streptaxidae 50, 63 Streptaxinae 63 Streptostele 63 striatulus, Modiolus 5 Striatura 159, 160 meniscus 160, 162, 163 meridonalis 65 pugetensis 160 Striatura (Pseudohyalina) meniscus 159, 160 strigata, Littoraria 94 strigata, Mytella 1-14; 3, 6, 7 Striostrea 187, 192-194, 197 Striostrea (Parastriostrea) 198 sturmi, Horatia 30, 31, 33, 36, 38-40 Stylommatophora 63, 64, 66, 133, 206, 207 subcentralis, Nesopupa 162, 163 subcentralis, Nesopupa (Infranesopupa) 159 subhispidula, Canariella hispidula var. 119, 120, 129 subhispidula, Helicodonta (Caracollina) hispidula 118 subhispidula, Helix (Anchistoma) hispidula 118 subhispidula, Helix (Gonostoma hispidula 119 subhispidula, Helix (Gonostoma) hispidula 1122722 subplanus, Mesomphix 181 subtilis, Paravitrea 65 Subulina 54, 63 Subulinidae 49, 63, 65 Succinea konanensis 159, 162, 163 ovalis 65 Succineidae 159 Succineoidea 65 Suteria ide 66 Szentgalia 134 tenella, Vitrina 159, 161-163, 166 tentaculata, Bithynia 29-34, 38, 39, 87, 145, 147, 148, 150-152 Thalassohelix ziczac 66 Theodoxus 39 215 Therasiella celinda 66 neozelandica 66 serrata 66 Thysanophora 206 Tomatellaria 159 Tornatellides 159, 160, 162, 163 tridentata, Triodopsis 65 Triodopsinae 65 Triodopsini 65 Triodopsis tridentata 65 vulgata 65 Trissexodon 132 trivolvis, Helisoma 87 Tropidophora 54, 63, 69 truncatula, Lymnaea 39, 84 truncatus, Bulinus 84 tumbezensis, Mytella 1 Turkostrea 188 Tyrrheniella 133 Tyrrheniellina 133, 134 undulatus, Modiolus 5,8 unidentata, Fectola 66 Unio 30-32, 35-40 urquharti, Allodiscus 66 Valvata piscinalis 30-34, 38, 39 Ventridens collisella 65 Vertiginidae 48, 65 Vertigo clappi 65 gouldi 65 Vicariihelicinae 132 virginica, Crassostrea 24, 188-189, 196 viridis, Perna 5,8, 11 Vitreini 65 Vitrina tenella 159, 161-163, 166 Vitrinoidea 65 Vitrizonites latissimus 65 Viviparus georgianus 87 vulgata, Triodopsis 65 Xanthonychidae 133 Xenostrobus 1,5 Xerotricha apicina 134 Xolotrema denotata 65 xystera, Amepelita 69 xystera, Ampelita 71, 72, 74, 75 xystera, Ampelita (Xystera) 70, 70, 72- 74 zaletus, Mesodon 65 ziczac, Thalassohelix 66 Zonitidae 49, 65, 159 Zonitinae 65 Zonitini 65 VOL. 36, NO. 1-2 1995 MALACOLOGIA International Journal of Malacolog y Revista Internacional de Malacologia Journal International de Malacologie Международный Журнал Малакологии Internationale Malakologische Zeitschrift Vol. Vol. Vol. Vol. Vol. Vol. Vol. Vol. Vol. Vol. Vol. Publication dates 28, No 29, No. 29, No 30, No 31, No 31, No 32, No 33, No 34, No 35, No 35, No 2 1 2 2 => в a2 He 12 SEZ Ba 2 19 January 1988 28 June 1988 16 Dec. 1988 1 Aug. 1989 29 Dec. 1989 28 May 1990 7 June 1991 6 Sep. 1991 9 Sep. 1992 14 July 1993 2 Dec. 1993 VOL. 36, NO. 1-2 MALACOLOGIA CONTENTS В. ARAUJO, J. М. REMÓN, D. MORENO & М. A. RAMOS Relaxing Techniques for Freshwater Molluscs: Trials for Evaluation of Different ео als ados JOST BORCHERDING Laboratory Experiments on the Influence of Food Availability, Temperature and Photoperiod on Gonad Development in the Freshwater Mussel Dreissena ОПОРА Seren бы они ею see enone! ehe el ot een dope er ea ель ame chan HEINZ BRENDELBERGER Dietary Preference of Three Freshwater Gastropods for Eight Natural Foods of Different Energetie СООО ae LM. COOK & J. BRIDLE Colour Polymorphism in the Mangrove Snail Littoraria intermedia in Sinai ..... KATHERINE COSTIL & JACQUES DAGUZAN Effect of Temperature on Reproduction in Planorbarius corneus (L.) and Plan- erbis.planorBis WE) Throughout: the: Life Sparte, sa as ROBERT H. COWIE, GORDON М. NISHIDA, YVES BASSET & SAMUEL М. GON, Ш Patterns of Land Snail Distribution in a Montane Habitat on the Island of nihil ancora mon Samii neato и ее N. ELEUTHERIADIS & M. LAZARIDOU-DIMITRIADOU Age-Related Differential Catabolism in the Snail Bithynia graeca (Westerlund, 1879) and its Significance in the Bioenergetics of Sexual Dimorphism ........ KENNETH C. EMBERTON Distributional Differences Among Acavid Land Snails Around Antalaha, Mada- gascar: Inferred Causes and Dangers of Extinction .......................... KENNETH C. EMBERTON Land-Snail Community Morphologies of the Highest-Diversity Sites of Mada- gascar, North America, and New Zealand, with Recommended Alternatives to Reight-Diameler PIOUS as ee. aa re ee ea daras KENNETH C. EMBERTON & SIMON TILLIER Clarification and Evaluation of ТШегз (1989) Stylommatophoran Mono- 6 LÉ 6 ee cree ea ee a ee eae getel MICHAEL G. GARDNER, PETER B. MATHER, IAN WILLIAMSON & JANE M. HUGHES The Relationship Between Shell-Pattern Frequency and Microhabitat Variation in the Intertidal Prosobranch, Clithon oualaniensis (Lesson) .................. MIGUEL IBÁÑEZ, ELENA PONTE-LIRA 8 MARIA R. ALONSO El Género Canariella Hesse, 1918, y su Posición en la Familia Hygromiidae (Gastropoda, Pulmonata; :HelicGidGa) ока ксеро amid sees te an DAVID R. LAWRENCE Diagnosis of the Genus Crassostrea (Bivalvia, Ostreidae) .................... - ALAN E. STIVEN Genetic Heterozygosity and Growth Rate in the Southern Appalachian Land Snail Mesodon normalis (Pilsbry 1900): The Effects of Laboratory Stress ..... MARÍA VILLARROEL Y JOSÉ STUARDO Morfología del Estomago y Partes Blandas en Mytella strigata (Hanley, 1843) NENE UN e A EE A emma 1995 29 15 147 e 79 155 139 67 43 203 97 111 185 171 AWARDS FOR STUDY AT The Academy of Natural Sciences of Philadelphia The Academy of Natural Sciences of Philadelphia, through its Jessup and McHenry funds, makes available each year a limited number of awards to support students pursuing natural history studies at the Academy. These awards are pri- marily intended to assist predoctoral and immediate postdoctoral students. Awards usually include a stipend to help defray living expenses, and support for travel to and from the Academy. Application deadlines are 1 March and 1 October each year. Further information may be obtained by writing to: Chairman, Jessup- McHenry Award Committee, Academy of Natural Sciences of Philadelphia, 1900 Benjamin Franklin Parkway, Philadelphia, Pennsylvania 19103-1195, U.S.A. WHY NOT SUBSCRIBE TO MALACOLOGIA? ORDER FORM Your name and address Send U.S. $26.00 for a personal subscription (one volume) or U.S. $45.00 for an institutional subscription. Make checks payable to “MALACOLOGIA.” Address: Malacologia Department of Malacology Academy of Natural Sciences 1900 Benjamin Franklin Parkway Philadelphia, PA 19103-1195, U.S.A. MALACOLOGIA, 1995, 36(1-2) 36(1-2) INSTRUCTIONS FOR AUTHORS 1. MALACOLOGIA publishes original re- search on the Mollusca that is of high quality and of broad international interest. Papers combining synthesis with innovation are par- ticularly desired. While publishing symposia from time to time, MALACOLOGIA encour- ages submission of single manuscripts on diverse topics. Papers of local geographical or systematic interest should be submitted elsewhere, as should papers whose primary thrust is physiology or biochemistry. Nearly all branches of malacology are represented on the pages of MALACOLOGIA. 2. Manuscripts submitted for publication are received with the tacit understanding that they have not been submitted or published elsewhere in whole or in part. 3. Manuscripts may be in English, French, German or Spanish. Papers in lan- guages other than English must include a translation of the Abstract in English. Authors desiring to have their abstracts published in other languages must provide the transla- tions (complete with main titles). Include all foreign accents. Both American and British spellings are allowed. 4. Unless indicated otherwise below, contributors should follow the recommenda- tions in the Council of Biology Editors (CBE) Style Manual (ed. 5, 1983) available for U.S. $24.00 from CBE, 9650 Rockville Pike, Bethesda, MD 20814, U.S.A. 5. Be brief. 6. Manuscripts must be typed on one side of good quality white paper, double- spaced throughout (including the references, tables and figure captions), and with ample margins. Tables and figure captions should be typed on separate pages and put at the end of the manuscript. Make the hierarchy of headings within the text simple and consis- tent. Avoid internal page references (which have to be added in page proof). 7. Choose a running title (a shortened version of the main title) of fewer than 50 letters and spaces. MALACOLOGIA 1995 8. Provide a concise and informative Ab- stract summarizing not only contents but re- sults. A separate summary generally is super- fluous. 9. Supply between five and eight key (topic) words to go at the end of the Abstract. 10. Use the metric system throughout. Mi- cron should be abbreviated um. 11. Illustrations are printed either in one column or the full width of a page of the journal, so plan accordingly. The maximum size of a printed figure is 13.5 x 20.0 cm (preferably not as tall as this so that the cap- tion does not have to be on the opposite page). 12. Drawings and lettering must be dark black on white, blue tracing, or blue-lined paper. Lines, stippling, letters and numbers should be thick enough to allow reduction by 2 or 3. Letters and numbers should be at least 3mm high after reduction. Several drawings or photographs may be grouped together to fit a page. Photographs are to be high contrast. High contrast is especially im- portant for histological photographs. 13. All illustrations are to be numbered sequentially as figures (not grouped as plates or as lettered subseries), and are to be ar- ranged as closely as possible to the order in which they are first cited in the text. Each figure must be cited in the text. 14. Scale lines are required for all nondi- agrammatic figures, and should be conve- nient lengths (e.g., “200 um,” not “163 um”). Magnifications in captions are not accept- able. 15. All illustrations should be mounted, numbered, labeled or lettered, i.e. ready for the printer. 16. A caption should summarize what is shown in an illustration, and should not du- plicate information given in the text. Each let- tered abbreviation labeling an individual fea- ture in a figure must either be explained in each caption (listed alphabetically), or be grouped in one alphabetic sequence after the Methods section. Use the latter method if many abbreviations are repeated on different figures. 17. Tables are to ben ised sparingly, ‘and vertical lines not at all. 18. References cited in the ее must be in the Literature Cited section and vice versa. Follow a recent issue of MALACOLOGIA for bibliographic style, noting especially that se- rials are cited unabbreviated. Supply pagina- tion for books. Supply information on plates, etc., only if they are not included in the pagination. 19. In systematic papers, synonymies should not give complete citations but should relate by author, date and page to the Literature Cited section. - 20. For systematic papers, all new I specimens must be deposited in museums - where they may be studied by other scien- _tists. Likewise MALACOLOGIA requires that voucher specimens upon which a paper is based be deposited i in a museum where they may eventually be reidentified. “ 21. Submit each manuscript in triplicate. ductions. | ty Г 22. Authors who want illustrations re- turned should request this at the time of or- dering reprints. Otherwise, illustrations will be maintained for six months only sta т pon: АЙ NR у | REPRINTS AND PAGE COSTS AUN 23. When 100 or more reprints are or- dered, an author receives 25 additional cop- | ‘ies free. Reprints must be ordered at the time proof is returned. to the Editorial Office. Later - orders cannot be considered. For each au- thors’ change i in page proof, the cost is U.S. $3.00 or more. Pop 124. When ап. article is 10 or more printed ‚ pages long, MALACOLOGIA requests that an ~ author pay part of the publication costs if, grant or! institutional ре available. и Е N ! \ 3 ñ 4 Ai т DOES y О \ SUBSCRIPTION PURCHASE 25, Effective ian 1995, Personal ар tions are U.S. $36. 00. Institutional, agency or © dealership ‘subscriptions are U.S. $55.00. - + Back issues and single volumes: $36.00 for — j individual purchaser; $55.00 for institutional ( Pe or agency purchasers. There is a one dollar. - handling charge per volume for all purchases _ of single volumes. Address inquiries, to the | Subscription Office. Dr Te eet) DANS Ne AR RTL SoM Yo RAR 2 us. Kur y Wie a A Rr Е 14 x Pr er ws] Franks PA un EN ue E en 5 RATS 2 Pare A RN N AA dr, О CN PCT u ARE; у “à An a Но \ ire AN 4 : | The second and third copies can be repro- | | a) VOL. 36, NO. 1-2 _ MALACOLOGIA CONTENTS MARÍA VILLARROEL Y JOSÉ STUARDO BA, Morfologia del Estomago y Partes Blandas en Mytella stigata (Hanley, 1843) (Bivalviaz Myllidachs 2 о la slo PÈRES BR JOST BORCHERDING | en Laboratory Experiments on the ee of Food Availability, Temperature and Photoperiod on Gonad Development in the aa” Mussel Dreissena Polymorpha 22 я а Nr eee AE bobos. R. ARAUJO, J. М. REMÓN, О. MORENO & М. A. RAMOS a A Relaxing Techniques for Freshwater Molluscs: Trials for Ellen of Different » A A ES A A N A Ge CLO A KENNETH C. EMBERTON Land-Snail Community Morphologies of the Highest-Diversity Sites of Made | gascar, North America, and New Zealand, with Recommended Alternatives to o Height-Diameter.Rlöts UT con ee Re La ali à CLS STORE Ss jae KENNETH C. EMBERTON ; | Rss | a Distributional Differences Among Acavid Land Snails And Antalaha, Mada- | gascar: Inferred Causes and Dangers of Extinetion LD. ee А wey E COSTIL & JACQUES DAGUZAN | SS oe - Effect of Temperature оп Reproduction in Planorbañus corneus (L.) and Plan- ( orbis planorbis (L.) Throughout the Life Span .......:.:......... FRE een Ya’ L M. COOK & J. BRIDLE Bas a eh а Colour Polymorphism in the Mes Snail Littoraria intermedia in п Sinai. at Va 5 MICHAEL E: GARDNER, PETER В. MATHER, IAN WILLIAMSON & JANE М. HUGHES The Relationship Between Shell-Pattern Frequency and Microhabitat Variation - in the Intertidal Prosobranch, Clithon oualaniensis (Lesson) .......... Кенни С. (8 MIGUEL IBÁÑEZ, ELENA PONTE-LIRA 8 MARÍA В. ALONSO | } LT El Género Canariella Hesse, 1918, y su Posición: en (a Familia Hyaromidae Me. | (Gastropoda, Pulmonata, Helicoidea):..:....... nie wa aes + oe ER SPA AÑ "N. ELEUTHERIADIS & М. LAZARIDOU-DIMITRIADOU q Age-Related Differential Catabolism in the Snail Bithynia graeca (Westertund, ^ ey 1879) and its Significance in the Bioenergetics of Sexual Dimorphism зай м _ HEINZ BRENDELBERGER | G à it Dietary Preference of Three Freshwater lad sb for Eight Natural Foods of ‘Different Energetic Gontent .......................,.......... и en ROBERT H. COWIE, GORDON M. NISHIDA, YVES BASSET & SAMUEL M. GON, mo \ Patterns of Land Snail Distribution in, a MRS Habitat on the ‘Island of > Hawaii .......... ни nets a ato о 15 ALAN E. STIVEN cc apts ) Ane TE ÓN \ Genetic HET bons and Growth Rate in Le Southern Appalachian Land a J Snail Mesodon normalis (Pilsbry 1900): The Effects of Laboratory Stress mess DAVID В. LAWRENCE | fh i EN. Diagnosis of the Genus Crassostrea (Bivalvia, Ostreidae) . ки eo O KENNETH С. EMBERTON & SIMON TILLIER | Ji pd, A Clarification and Evaluation of Tillier's. (1989) Shin tenté ‘ee 10) LUE de Mme en O EOS RS oak EC RS : YAA . MU 3 2044 072 160 567 Рут +. , oe # LU , : . + ь . pated ; rs Dean e г : ed 4 : + , , rd 4 , . Fer rue? nern art , EL AR > г ; : р Е : pS 17 р Lise À F . : we й к # À : ; ? Foy SE pe ++ r , , мы 4 diet gid : p , p 4 + { I gn pr JE re р ми , res Е , , ; я ‘ ; + ri р ` e | ; ar Len CAE ts f р a ; er? à LB ; . APCE 0 1 ; sur , Âge db ro) й ‚ y f Es A , tier : у x £ La > of ‘ , . > A OX, é F , nents i ; О ‘ L ’ ot ov ‘ , > m e? , E , ‘ , e 2 r à 4 : : , ‘ > : Hp peta , ore ant er e ’ e ' y р ' toa J ra , м 2 + О A e 4 0 4 ‘ 4 , , . , . ы + ы 4 2 , * , “ À pi , i О + f , 4 ’ , A ' # ‘ , , > oF , Le , к О ” e ., О 4 ‘ , of , , ' re A eee " , я , , wre ’ ’ Ben р rar , р vo , ' fy tet ; ime se ou ge pl HAUT: , pes ? ’ : à "ue Ку у , a care ARAS Vert RAC d } e : / N . y Pa | f RTE A A pére, 3 . В ’ 1 + ” , vr иг“ La nt, [2 > $ , nl и у № р > +. м ye ; ’ 1e 5 + > Г ‘ p ’ РР r , г п, 1 " ; ele r р ' pe . 2 gag у 14 o. pet PCR AS , ñ Pro р има и: gn tee A LE pes 4 te В ‘ , 2 ’ Kr E ыы 1 ern s 4 y bebo Lee la Ú , mara get 3 st , я , 4 a: svat р ne r y 4 f Ri я » Y + , в 049 р re $ ; 5 + Fern vo IR 4 #2 : Я y р , 0 oF , nos , e, , ‘ , ’ # , er ae ‘ e , Le AA ” ’ и * ee Г Й d y y y... , y + as + , р 0 je ve р , , у и, ' ' ” ' DE PEN af! y A у 4 .... O A “be? oe , ’ nf ’ pas ee 1? J и w н.в ’ 0 , # , * BT ‘ , ум 4 nee y Рем 4 ALES 0 y > . pra a >. ¿nn ere, 1 0 Ap Aw А ne ’ res er. Ce LAVE ; as ’ 4 ..os y A + TON vn Vs Nr 14 ; r u 7 - РО o q. 48 y. Pz Pes # ’ nr ww 4 7 } , / t pl re , р Arf ns? , УЖО PE A 1! En sd oi у te vies es dés ff A ” , 1 Ps sh D й , RTE nm. sa one , ' ‘ Se he O ESO a? er + A 4 : ’ ‘ a eee? ADULTE pure a rf ine eile ree AeA SY ' ; Ha: rn 1.4 di dde eus FE и pe . : A Я гении A re re и ; d / Им НОАК, J ur Plan у ¢ tens ea +e - obs ers Far ae Prop: REN 4 vr , 1 mo mos + nl , 114 roo ¿Le : fr 2 .. в -ч - ver ’ ии or Saw ‘ aay " 4 e у er en CR ‘ 2 ren Fat a A 1 ’ , 1 inate à . ‚ #1 pet + ‘ , 5 - / rt ve. € 11. are 208 PA, pha, ape р г у у ; , , see edhe pnt x af fahr , й у AAA A oad ET de pete ПСТ ds dut НР meee pi ke e. ¡AA we PA ARE FAR, 5 ’ ae р we , rasp rn ey , AA FR) | ef N ” Er pe + Er on. e , / "му AR ие , HA AUTRES SALE ct ER tt PT LT FAHREN р о р nee AAA de , ee va , Hin ‘ , ñ , 3 id , + oe rd HAMAS у t von men heel tates kr te ; 7 ‘ ee A LUN ne oy 6d ’ РИ su + ur ine AT , tee ae р A Me 4 | ‘ +s РС ANT et f , PRA A 1 epee 1 CC pecs anche TN CUT wf , 5 ; A ' NIV AN y nn feat ; , ” й RAP OR EL KL ER ’ MES 4 entr ty rato” TIE , i pee ги ‘ here , ИИ нь Pb О * oa [271 1 , eure ni ’ é ovr ype ninas , phot taped . 11 И rad ; : FA Mage 2 ore , ‘ we a penises a М en re 5, , , у О nn er an prt ; ИР ¢ / e PR. eher Ago 114 u... Le ‘ mM. oF 2 no ide + AU A en и fi LA t vers и РИ Ae " oe х ' vor . ; ; DARSTELLT vor р ый ; . ОИ APA TN ad-e4 ev y Ben A Fi; nally pple DAA ts wer SA > M ENTES SE PAT ’ f tty . y A a pane Я ARAN AAA TORIO PP TRE LI DIRA AU o” Ри ter 14 y e. В MU nt а ern Y A ИН A Line ne п ar ’ oni À +, ELA sm ty! oo ” Ferien “i m ’ . " НИЯ И Pp. ‘ A A pr Neue ‘1 La ee oer AA? 4... £ ИИА hue tor AN ge ate TEL PONT With fy ete ee ne ß ren tA АКА ner A ARDEN AL A , r nu. f , e DR) MTL Hy? 7 SATA г , er A IS e AAA at PTE vat rn + 2 ns oe bape O ПТИ RO RA ee y NE PTE aOR EU TR A ok MR PA est TN ET ERA AU tu ARTE mn TA их ; sper? WT a в raed a , ; IA ent „ и AA DIR ВН $ nase DEIA we DB , , , e у у ’ net, PES, НЯ АНЯ У р 1 ; run Ly bee u Y 1 Y ии “we id DIR EEE a F1 ve A TT AU ARR ES SR PA e. ws oye ee уе partes poll A PP EU e A 14 AA SEA meat EA IA ME Е rauhen erh A nro nr ” , fs LH: rue? ‘ + UT ANT ARS + АЯ РИ , A ARTIE Ve и AR PATENT ey wr YUL OS ent ‘1 = y PAT DIR, RARA RN TE ON ITA pr eg A mus pert 11 dy QU SA A e. ‘9 afr be ANE AR и Ir; , nage ye ( ' ‘ Г te Pe ae ЧИ 11 РИА Я en PT #1 DIO ont Г иг, LAA И a TAS N e , д o AAA TR AAA AAA y Y má RER ; CARE MODE ET PAST AAA IU LA PAROI ALIAS QU as , МИ a hotel pad adit eet CU RE Ed hid NR ERST HH IEEE A AAA A AMA NAAA DS Panne e, Ly ts И CA Г НИИ Г. И У vor PAPA TIN BEHRINGER, este , Fast AH : veer 14 nes RAI AE Maa! DREH BEI LUE her ee pie ae 11 ARA ns LA DA ET ARE 54 hoes Me ra en et Tht teat, Fat etal rent whew BT LEA 2 à ot pre e ВА ИИ ЯНА Де setae PAPAS PILE DAA un ИИ RN gt A ПИЯ tgs wt ARO AL A el ta D 8 AP 1 7 И tthe gas pee ИТ три PAT 2 PM ENT WO ee TA ah ead un ae RTL ALT rt PRIE EL beet edie AA AGS AE TRE NL ALL LIRE DR AGS EAT ser ре И ИТ ‘ 1 PCE ES UE A VAN A rt 44) AA AS ИИ LINEAS MIS TADA pie? DE 2 AAC ut pp ara eed ABRA EN AAA ARTE NT AMIS Kinn oD Zu Uns PAT ИИ H 57 ; я АЛЯ И ны RER, etd / И, fa inne in ee Ee tbe AR sat ARA ITALIA OA pes ди И tree WAH PIR Ra ji Hat t/t ee Met pra AAA nee RE АИ DARIA AA НАЧ НЯ AA ит 4 РАТИ pr ИИ ИИ ИЕ НВА TE ры ИИ CY Anat ene! AA Lf FTE aap as Y re ae 14 TA A Рут Я Ир Wien bey ИЯ AAA ASS Pts A Ne rara emir, lt AL TRES A nga ИД , AR чи ВИ и ни TIA ene EAN TS И ИЯ В OA M М RER ЖАН PA РАО ERA CTL ha AAA WAY TELE AT at EE EMA. RE NA A A A irs TN] И ние TON И ‚ ys Ares WR CPP НЯ ATARI IAS ART ee AGAR LIS AS ARI IA BEE ME RT Pr TR И; A AULA AA A US ИИ AO AT 1 PA САНА ИИД AAA m nto” ATTE HA LTDA AAA ws РТИ РАЙ RR RUHR, РИ TEL ER AE Lk ПИТ А eed АЗИИ, ne AA ИЕН Я я, vr г, /A АМИ PAT PILE LEUR ARA FA POP APA IT het Я И VA AIT PAT AL erage te «¿Alea A IA Td ив ни, rte RIED: ней, PRE M A TE DEP INIA ЧАИ АЯ ALT И НЯ RARE EE TER 1 и A AT ИНО RAT MAT LA PR APT ИИ, И ИЯ НЕ RT, ia РИН 6e koe os 517 SAA UM AIT И PAA TON Pha bar O Ll И НИИ В ot gh Le RER PER AENA AN EL Pr PT ALI ИНН АИ, PAR E A RA ИИ О АН НО ИРИ TURE rR A ey , пен, pm À 4 A $ Pt ie nS Е PELE h TEILEN, ‘ р АДА и EU FE tat 4 i er ? (LPRA ре" РИА er, АЯ ВНИИ АНА НА АА Bes PTT APM LT ИЕ ИНАЯ ИЕ, ile at у АИ ИЯ т Ни ИИ ap AU AAA TA PHP Uno is EI eee en TH RR TRE ; y RA TY AA AA / AA Ир ИИ 0 т OR I LA NT TOOL NEIL AA A ия, A PATENTS ATOS PATATE ПАТТИ И ИИ ODA EME à MARRAINE RR ИИ И ИИ a ghana НИИ у 22 TU AA En ra ЧАН TEAL AAA YA Y ad | } f PER Ue И ИИ, uf! oe ve ei eee eget ‚ DENE PAT , UT CARA AA AS р И MATIN ИЯ nde NE AU Par RAU RA + wea ET RE TANT, И 4 As ade # an НИ НЯ в se Wht rt И, 142 y dl A SE fs AA aah iter AAA LOL EE nr Beuel rt A sai { Lo yee p IRA OA A WER nl Annan ИН, НИ ПАЯ ВИН AA PTE ИИ ИНН НИ PAM PILE RAA AO TA PAT PE ENTE RE LL AAT EERE SNS lagi RR PER OAI LEE ART AIT e ISI TOA Ad TEER TP ANR AT A et MALUS AA PAT EE A AA A A ST M MP PA И HA HAT НИ wees / И ИНАЯ AAA AER IIA Е A TNA A AA ИИ А НИ Ta PL AS MAT RL ER AAA A A RER TEE A I OA AM) ри PCI DIU gr HT Er ААА Ч A LAS SAA DA A Ch AIT. RAR ВОРОН НЯ ИЯ AAA ПИ ти Vocus ALICATES AMA AAA ИИ an MAD ОРВИ RE i IAEA DAA AVERIA ARAS Dr th RR TIAS PAPE LIT NAN OUR Passe MA A gin TRA BsAs 4 4 EAE RE р. рн AAN AAA # В РИ TE RTE A A AA AA НАЧАЛ as PPT BR AA AAA TT НИССА (EPL LL AA TOTAL PA RAIN MEER ИИ IE AH A MELUN MATUTE PA РНИИ PEL ALLO pEet O A Ad PME PAIN i na tet LEI (4 et р Я, Hart; и ete CAS 4 LE Er TAR TATE a м АТИ PA 21 EL ARA rs ALS АЛИЕВА йе PER TM Pe AE PT de AS HA TE IIA TARA ds. / PAT RA AA AE AT EE A IE a AA y PARIO ART TL A EA PANDA PP AI PT SG AA A GE APT ; PA ARAS ATA ei aay ИЯ ИИ, RAA TT AP ET ee Apot BAAR OO Mee NREL ey TO AAA AT PA laye AA PP APT EE ЗА rn ATAR TEA A A НОТ ИИ СОНАТА PAIE PPT р И OMAR AI ПРИ ARA ARE, EN IRALA CALLA RS TN AAA AU НИ A rey ates tle COIS A DR a A) Wine НЯ НА НИ О A A APT DA ANR ини AA Ton AA A II TU A ИНН АИ т WII a a И ре EA ales ДИ ИНН Л АСИ у A ре ARAN PIAA AE LEGA ELLE APE AT sari Ebest te ENA AAA TARE TR EY nt ruht i D? HAS ne PAT SER AAA AAA tarts tgs AT EEE Kerr wen UA PE ine PR, Ed RA APE AA iV pr AB AAA ИТ D ART AOS ñ det PAID AAN RAIN CIA A PARA DIOR TA a aL PTS IA de AA A нате ИН ИН ЧИ те НН 4 BER nz у. ; ARIAS PRA EII RI MENT ААА, al bite’ a ИИ yd A AAA Y AOS ye у y A AE AIR - ИАС vr. ту ЧАИ И НИНЕ DR RE ER PAT Gal RA DIA Ye ий HAT DATA RATS НИ те pa nd ОТ: ИО ИИ) Иер’, th AU ANA RR ARA RIA RE IATA AR AAA AAA BRETT AA AOS TA Petes , ИИ А ah HE ARI ERT LY A Я RUHR ИИ A e A A AO LU VE THREE AA ER DAA LA ER, й УИ A НБ 4 и HAL OUEN AE LILIA РИ DT AL ACTA PARDON TA LA Пе ИМО ТАКОЙ ААА Deh hee EET Ed Y 440 PE LORIE APA MIEL eA VAT UTIL AA A ATI AAA AMP LARA IRE lA PATAS A ges TARA A LE RAS AO AY ah MALA AA va PAC AR IE TEE RAA MALI AA AAA AI Sy DATA РИА НАСА PPT EC UE ART PTE ATLAS USA APS MTL RTE 0 NAAA AAA TALARA A ITA ИНАЯ art er te PUTA MR ГИДА AAA GE ти, MAC D M НА NPA AÑ Пенни ний, ИАТА IA TI WI ANA NAT TAN AO PAU CAEN et RIRS y PNA CAT LL TOR RIA RL EL A D AT M Le MATIN A CPAM PAT UAH TA ИО ИНАЯ НЙ ИЯ / OLLA Ph A IA OEA AIDA AT AA AN) PTT EE aT TILA ПОРКА ИЯ IAE NRA РИ II COTE AD AIRE AT НИХ НИИ ИАА И a ALLA A spilt A AR PAR RAE AA eV, MATE TR A TA MONS PARU LE TT AA TALA A LAR ИАА INTI AECL, AA IRA A ARR ERA TE PA ITA mut ; AAA AIDA HA ee } ИАН HALL CAE (RDA PATATE MEL OS ANA AT ‘ TIA О A AA ty (41 HOT rear Le Tite, о ИИА и у e ALT OP AA PEL OUR ANT D PA OA PET PA PET ete RTL PNA] AU WERDEN ATARI ARS АЯ ARI URN ИО À 4 Y DAA AN PAR ET RL A PA AT AT НИИ И 7 TER FREE, ри КАТЯ AR RDA ee НЯНИ UAE ATT PUR PA те OR ARA A Я VRR TH HR АНЯ НИИ vum ЧАИ Е а RAT НИЙ A DATA A UL A AIR IL VA ENT rt AAA PA CARE A RI A PA TT TLL A i PR TT EN AAA O RAMA AAA wt ATTIRENT ПЛОВА a N a PN ee A; AA I ragen ЧАЯ и; a I TA) AE NE TE A Al AA PA TE here AA ARAL AME ROTA a НИИ АЛИЯ CHARTE PAIN AL A CARAS TIA AS IRA EA A RANA PARLAIT НИЯ nn: И 7 bet VI MAE PA ALP HAN AAA wur, nt API AAA AA ив we AAA a OA VET TETAS ALIAS ПВА ATP A DT IA PAR AA RCN en (HA PEGA RADA ИЯ AA ADA IU ENT AA ИИ as DI A ACID M Oe EH ME ИН АИ ТАН CARTE AA de ARR ET M ANA AP AE ARTE E NT И AN ET D Lite AO dois ' = WATT TL III AAA D A AA UA НИ ИН AA AA FAIL AAA AP TAL IN LARA AT AAA AAA AAA РИАННА IIA AAA ET DURE Я уни" HO , nen REDET EDER N ви ИЯ A AR A A RATE AA MER LE TC AP НЯ ИИ в MOI Я $ AAA TT TO ИЯ НИИ, ACTUA Ad M PA И AAA LEE APN A A TIA AY, Hf AAA A AAN AE: me AAA AAA САННИ И er hay ИЯ И, ИНН, ИИ НИ ИИ Be ИЯ RAT 7 у , mw... aft 1 IU + RL LA EUR er vera, ur. or ПУСТОГО UTA TIT ое Е, Ah 140 # 404 LATA TAN ORAR AA GUS AE ИИ AA A TRA RA DIA RATA AA AA Y ОИ f ИИ, HE hehe PAL AA AL PPE AAA PELL EE HORE ALTER ALES PLAT LAA MA OU PAU TARA ARTO 112 ДИТ WALT ALY НАНТ A PAR NO bg AA НИ И ANNE И TCL a Ин CANO ULLAL a ИН Y ЛИР TS ver 14 Vib chy OL UTP PAPA CAO PAIE PAP NT ET tt + AAA 4 WATT Ad A APE Y RL ON ARAS ИЯ pen RRM O NE SA TDMA IS DA A WA PA ys RAT OMA AAA A ALAA # ELE MLL Я ! FELLATE AMIE И АТИ HA И НИЙ À 4 ute Bi th ait Vp tt ER АЯ И ee RATA PR И И НИИ 7 ИАТА Whee ett PATEL в AL 4 i, iy alae bite dU le i Wht VY RATA AS YI ET A NEN. (YE IIED A AAA ПД AAA P Ir gar fe dee porno тв IT MITA AR ARIAS eut PAUL LOT AU A LU UE Pere bene y ‘ ANA „ Mi ИНН ее ТА Jets NY VIA ADA ” IRA AA ‘ Ave VIEIRA PRETO ВИ LU ИВО UN 4 aa e , ПИТАНИИ КРИ (444 AU MALE AIT AIT) НИНА Vy 2 AAA ия, НИЯ ИИА АЛИ Us a E ALLA OY VETO) ASA ITA PILA AAN EEE IE RAA EMO, KARRIERE II НИИ ARAN OO A PP AR НА ARO OL AAT AMD PTET Aad aa ПИЯ ME AE a VA IIA DARA PR AA НИТ IE PRE ИН TP AAA A OL UT PE PEN PU DA! À UE A TE PAS AH (ATP) И IIA A PR TA Wr 1 TH . А a COS CAIRO ARE TE Cay spied es Woe A AAA) ИИ И nm я и e IEEE RIN Gite ee A 4: CAS re Ye LAIA SARA PITT TT War нот ИИА AAA ИИ, ‚444% he NIET IT LL